European Coal Geology and Technology
Geological Society Special Publications Series Editor A. J. FLEET
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 125
European Coal Geology and Technology
EDITED BY
R. GAYER Department of Earth Sciences, University of Wales, Cardiff AND
J. PESEK Charles University, Prague, Czech Republic
1997 Published by The Geological Society London
THE GEOLOGICAL SOCIETY The Society was founded in 1807 as The Geological Society of London and is the oldest geological society in the world. It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society is Britain's national society for geology with a membership of around 8000. It has countrywide coverage and approximately 1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including professional matters. The Society has its own publishing house, which produces the Society's international journals, books and maps, and which acts as the European distributor for publications of the American Association of Petroleum Geologists, SEPM and the Geological Society of America. Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least two years' relevant postgraduate experience, or who have not less than six years' relevant experience in geology or a cognate subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may apply for validation and, subject to approval, may be able to use the designatory letters C. Geol. (Chartered Geologist). Further information about the Society is available from the Membership Manager, The Geological Society, Burlington House, Piccadilly, London W1V 0JU, UK. The Society is a Registered Charity, No. 210161. Published by The Geological Society from: The Geological Society Publishing House Unit 7, Brassmill Enterprise Centre Brassmill Lane Bath BA1 3JN UK (Orders: Tel. 01225 445046 Fax 01225 442836)
Distributors
USA AAPG Bookstore PO Box 979 Tulsa OK 74101-0979 USA (Orders: Tel. (918) 584-2555 Fax (918) 560-2652)
First published 1997
Australia The publishers make no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility for any errors or omissions that may be made. 9 The Geological Society 1997. All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with the provisions of the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 9HE. Users registered with the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970, USA: the item-fee code for this publication is 0305-8719/97/$10.00.
Australian Mineral Foundation 63 Conyngham Street Glenside South Australia 5065 Australia (Orders: Tel. (08) 379-0444 Fax (08) 379-4634)
India Affiliated East-West Press PVT Ltd G- 1/ 16 Ansari Road New Delhi 110 002 India (Orders: Tel. (11) 327-9113 Fax (11) 326-0538)
Japan British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 1-897799-86-1 ISSN 0305-8719 Typeset by Aarontype Ltd, Unit 47, Easton Business Centre, Felix Road, Easton, Bristol, UK Printed by The Alden Press, Osney Mead, Oxford, UK
Kanda Book Trading Co. Tanikawa Building 3-2 Kanda Surugadai Chiyoda-Ku Tokyo 101 Japan (Orders: Tel. (03) 3255-3497 Fax (03) 3255-3495)
Contents
Preface
Regional coal reserves, coal basin tectonics and stratigraphy DORUSKA, J. The Czech Republic Energy Policy: conception and implementation in a market economy PEgEK, J. & DOPITA, M. Coal production and usage in the Czech Republic KUMPERA, O. Controls on the evolution of the Namurian paralic basin, Bohemian Massif, Czech Republic KRS, M., PEgEK, J., PRUNER, P., SKO~EK, V. & SLZPI~KOVA, J. The origin of magnetic remanence components of Westphalian C to Stephanian C sediments, West Bohemia: a record of waning Variscan tectonism DREESEN, R., BOSSIROY, D., SWENNEN, R., THOREZ, J., FADDA, A. OTTELLI, L. & KEPPENS, E. A depositional and diagenetic model for the Eocene Sulcis coal basin of SW Sardinia INANER, H. & NAKOMAN, E. Turkish lignite deposits KARAYIGIT, A. I. & WHATELEY, M. K. G The origin and properties of a coal seam associated with continental thin micritic limestones, Selimoglu-Divrigi, Turkey KARAYIGIT, m. I. & WHATELEY, M. K. G. Chemical characteristics, mineralogical composition and rank of high sulphur coking coals of Middle Miocene age in the G6kler coal field, Gediz, Turkey TICLEANU, N. & DIACONITA, D. The main coal facies and lithotypes of the Pliocene coal basin, Oltenia, Romania SI~KOV, G. D. Bulgarian low rank coals: geology and petrology STUKELOVA,I. E. Coal petrology and facies associations of the South Yakutian Coal Basin, Siberia
Coal petrology and palaeontology GAYER, R. A., FOWLER, R. & DAVIES, G. Coal rank variations with depth related to major thrust detachments in the South Wales coalfield: implications for fluid flow and mineralization DvoIL~K, J., HON~K, J., PE~EK, J. & VALTEROVA, P. Deep borehole evidence for a southward extension of the Early Namurian deposits near N6m~i6ky, S Moravia Czech Republic: implication for rapid coalification KOSTOVA, I., MARKOVA, K. & KUNTSCHEV, K. M6ssbauer spectroscopic investigation of low rank coal lithotypes PREMOVIC, P. I., NIKOLIC, N. D. & PREMOVIC, M. P. Comparison of solid state ~3C NMR of algal coals/anthracite and charcoal-like fusinites: further evidence for graphitic domains SYKOROVA, I., (~ERN~', J., PAVLIKOVA, H. & WEISHAURTOV~,, Z. Composition and properties of North Bohemian coals STEFANOVA, M. & MAGNIER, C. Aliphatic biological markers in Miocene Maritza-Iztok lignite, Bulgaria SYBRYAJ,S. Floristic characters of the upper coal-bearing formation in the Transcarpathians
vii
1 3 13 29
49
77 101 115
131 141 149
161
179
195 201 207 219 229
vi
CONTENTS
Mineral matter in coal and the environment
BAQR1, S. R. H. The distribution of sulphur in the Palaeocene coals of the Sindh province of Pakistan CAVENDER,P. F. & SPEARS,D. A. Sulphur distribution in a multi-bed seam BOUSKA, V., PESEK, J. & ZAK, K. Values of ~34S in iron disulphides of the North Bohemian lignite basin, Czech Republic JANKES, G., CVETKOVIC, O. & GLUMICIC, T. Determination of different forms of sulphur in Yugoslav soft brown coals PREMOVI(~, P. I., NIKOLIC, N. D., PAVLOVIC, M. S., JOVANOVIC, LJ. S. & PREMOVIC, M.P. Origin of vanadium in coals: parts of the western Kentucky (USA) No. 9 coal rich in vanadium SPEARS, D. A. Environmental impact of minerals in UK coals
237 245 261 269 273 287
Mining geophysics
GREGOR, V. & TI~2KY, A. A well logging method for the determination of the sulphur contents in coal seams by means of deep gammaspectrometry MACH, K. A logging correlation scheme for the main coal seam of the North Bohemian brown coal basin, and the implications for the palaeogeographical development of the basin HOLU~, K. Seismic monitoring for rock burst prevention in the Ostrava-Karvinfi coalfield, Czech Republic KALA~, Z. An analysis of mining induced seismicity and its relationship to fault zones OPLU~TIL, S., PE~EK, J. & SKOPEC, J. Comparison of structures derived from mine workings and those interpreted in seismic profiles: an example from the Ka~ice deposit, Kladno Mine, Bohemia
297 309 321 329 337
Coal technology and coalbed methane
BARRAZA, J., CLOKE, M. & BELGHAZI, A. Improvements in direct coal liquefaction using beneficiated coal fractions ALEKSI(~, B. R., ERCEGOVAC, M. D., CVETKOVlC, O. G., MARKOVlC, B. Z., GLUMI(~IC, T. L., ALEKSIC, B. D. & VITOROVIC, D. K. Conversion of low rank coal into liquid fuels by direct hydrogenation ASMATULU, R., ACARKAN, N., ONAL, G. & CELIK, M. S. Desulphurization of low-rank coals by low-temperature carbonization WHATELEY, M. K. G., GENCER, Z. & TUNCALI, E. Amelioration of high organic sulphur coal for combustion in domestic stoves STANOJEVI(~,P., JANKES,G., KUBROVIC,M., STANOJEVI(~,M. & BLAGOJEVI(~,P. The use of pulverized lignite/natural gas mixed fuels in the high-temperature process of a cement rotary kiln DOUCHANOV, D. & MINKOVA, V. The possibility of underground gasification of Bulgarian Dobrudja's coal BOARDMAN, E. L. & RIPPON, J. H. Coalbed methane migration in and around fault zones HOLUB, V., ELIAg, M., HRAZD[RA, P. & FRANCU, J. Geological research into gas sorbed in the coal seams of the Carboniferous in the Mgeno-Roudnice basin, Czech Republic GRZYBEK, I., GAWLIK, L., SUWALA, W. & KUZAK, R. Estimation method for methane emission from Polish coal mining TAKLA, G. & VAVRUS~.K, Z. Methane emissions and its utilization from Ostrava-Karvinfi collieries in the Upper Silesian coal basin, Czech Republic
349
Index
441
357
365 371 379
385 391 409 425 435
Preface Despite the major reduction in the coal mining industry that has taken place in Europe over the last decade, most European countries remain strongly dependent on utilizing coal for both power production and in the steel industry. There is an increasing tendency to import cheaper coal from sources outside Europe and this trend is likely to continue and even expand. However, the need to use indigenous coal is essential and by improving knowledge of coal geology and technology, more efficient and competitive use of existing proven and indicated reserves will be possible. This volume contains some 40 papers describing new research into coal geology and coal technology. These have been grouped into five sections dealing with separate aspects of the subject, so that related papers are placed together in the volume. However, some important coal basins have been researched by several different techniques, and papers on these topics have been included in the appropriate different sections. For example, the Upper Silesian basin, one of the most important Upper Palaeozoic coal basins in Europe, is covered by six papers in four of the sections of the volume. Similarly, the North Bohemian lignite basin is described in four papers placed in four different sections. Coal deposits from twelve countries are covered in the volume, with the majority of papers (34) covering deposits in Central and Eastern Europe. Nevertheless, the geology and technology described, despite having a geographical bias, is of general applicability. The deposits together with the associated concepts and methods may not be well known in the west so that the papers and included references should provide an invaluable data source. Thus the volume can be seen as a companion volume to European Coal Geology (Whateley & Spears 1995) which concentrated on coal deposits in western Europe. The present volume also describes new and important research in western Europe, updating the coal geology provided in the earlier volume. Section One includes 11 papers describing regional coal reserves, coal basin tectonics and stratigraphy. The regions covered include Bulgaria, the Czech Republic, Romania, Sardinia, Siberia, and Turkey. Amongst these interesting accounts are a paper by the late Professor Otto Kumpera, which relates the coal accumulation in the Upper Silesian basin to processes related to foreland basin tectonics, and a paper by Krs et ai. documents the waning effects of the Variscan orogeny in the Bohemian Massif by a detailed study of palaeomagnetism. Dreesen et al. describe an unusual coal basin in Sardinia in which coal forming environments are closely associated with carbonates and evaporites. The section also contains an important paper by Pesek & Dopita discussing the present and future energy requirements and associated environmental issues of the Czech republic, as an example of one of the developing eastern European countries. Section Two covers various aspects of coal petrology and palaeontology in seven papers. These include papers describing unusual variations of coal rank with depth in Moravia (Dvorak et al.) where coals remain at relatively low rank despite being buried beneath the Carpathian thrust sheets, and in South Wales (Gayer et al.), where high levels of heat flow and reversals in rank increase with depth are attributed to fluid flow within the basin. Other authors describe the results of various analytical approaches to the study of coal petrology, including solid state 13C NMR studies of fusinites (Premovic et al.), M6ssbauer spectroscopy of low rank coal lithotypes (Kostova et al.), and biochemical analysis of lignite (Stefanova & Magnier). Section Three deals with mineral matter in coal and the environment. The six papers include the sulphur contents of Pakistan coals (Baqri), of Yugoslavian lignites
viii
PREFACE
(Jankes et ai.) and of a multi bed coal in the UK (Cavender & Spears). Bouska et aL discuss the sulphur isotopic composition of North Bohemian lignites and Premovic et aL present the results of vanadium analysis in Kentucky coals. Section Four contains five papers concerned with mining geophysics. These include well logging techniques applied to the North Bohemian lignite basin (Mach) and the use of a deep gamma spectrometer (Gregor & Tezky). Seismic monitoring for rock bursts (Holub) and mining induced seismicity (Kalab) are two aspects of seismic investigation covered in the section. The final Section Five includes papers describing coal technology and coalbed methane. Liquefaction is discussed in two papers; one by Aleksic et aL using direct hydrogenation of low rank coals and the other describing experiments on beneficiated coal fractions (Barraza et aL). Desulfurization is also covered in two papers; one by Asmatulu et al. and the other by Whateley et aL, both dealing with unusual techniques to treat high sulphur Turkish coals. Gassification and coalbed methane generation from mines is covered by Douchanov & Minkova, Gryzbek et aL and Holub et aL, whilst Boardman & Rippon present an analysis of the influence of faults in coalbed methane production. The editors would like to thank all the authors for submitting the papers which represent a selection of those originally presented at the Second European Coal Conference in 1995 in Prague. We would also like to thank the many geologists who reviewed the papers: Mesdames & Messieurs Austin, Bouska, Brabham, Bright, Bryant, Cloke, Cole, Cornford, Davidson, Davies, Dopita, Drozdzewski, Ellison, Frodsham, Gayer, Gillespie, Glover, Goulty, Guion, Harris, Hathaway, Hemsley, Holub, Honek, Jelinek, Jones, Juch, Karayigit, Konecny, Kostova, Kropacek, Kumpera, McLean, Malan, Martinec, Miliorizos, Moore, Oplustil, Patrick, Pesek, Premovic, Querol, Rhodes, Rippon, Rosa, Simunek, Skocek, Spears, Spiker, Thomas, Turner, Wagner, Wakefield, Whateley. Many of the papers were written by authors whose first language is not English and this represented a problem not only for the authors but also for the reviewers. Both worked very hard to produce the present results. We have been continually amazed at the language skills of European geologists and hope that any slight errors remaining in the texts do not detract from the value of the volume. Sadly, one of the authors, Professor Kumpera, died before completing the final version of his major work on the geology of the Upper Silesian basin. Although his widow, Anna Kumperova, continued with the drafting of the diagrams, the conclusions have been added by the editors who accept responsibility for any errors inadvertently produced. We would also like to thank David Ogden, the staff editor at the Geological Society Publishing House for his continuing support and editing of this volume. Dr Rod Gayer, Cardiff Professor Jiri Pesek, Prague
Reference WHATELEY, M. K. G. & SPEARS, D. A. (eds) 1995. European Coal Geology. Geological Society, London, Special Publication, 82.
Preface Despite the major reduction in the coal mining industry that has taken place in Europe over the last decade, most European countries remain strongly dependent on utilizing coal for both power production and in the steel industry. There is an increasing tendency to import cheaper coal from sources outside Europe and this trend is likely to continue and even expand. However, the need to use indigenous coal is essential and by improving knowledge of coal geology and technology, more efficient and competitive use of existing proven and indicated reserves will be possible. This volume contains some 40 papers describing new research into coal geology and coal technology. These have been grouped into five sections dealing with separate aspects of the subject, so that related papers are placed together in the volume. However, some important coal basins have been researched by several different techniques, and papers on these topics have been included in the appropriate different sections. For example, the Upper Silesian basin, one of the most important Upper Palaeozoic coal basins in Europe, is covered by six papers in four of the sections of the volume. Similarly, the North Bohemian lignite basin is described in four papers placed in four different sections. Coal deposits from twelve countries are covered in the volume, with the majority of papers (34) covering deposits in Central and Eastern Europe. Nevertheless, the geology and technology described, despite having a geographical bias, is of general applicability. The deposits together with the associated concepts and methods may not be well known in the west so that the papers and included references should provide an invaluable data source. Thus the volume can be seen as a companion volume to European Coal Geology (Whateley & Spears 1995) which concentrated on coal deposits in western Europe. The present volume also describes new and important research in western Europe, updating the coal geology provided in the earlier volume. Section One includes 11 papers describing regional coal reserves, coal basin tectonics and stratigraphy. The regions covered include Bulgaria, the Czech Republic, Romania, Sardinia, Siberia, and Turkey. Amongst these interesting accounts are a paper by the late Professor Otto Kumpera, which relates the coal accumulation in the Upper Silesian basin to processes related to foreland basin tectonics, and a paper by Krs et ai. documents the waning effects of the Variscan orogeny in the Bohemian Massif by a detailed study of palaeomagnetism. Dreesen et al. describe an unusual coal basin in Sardinia in which coal forming environments are closely associated with carbonates and evaporites. The section also contains an important paper by Pesek & Dopita discussing the present and future energy requirements and associated environmental issues of the Czech republic, as an example of one of the developing eastern European countries. Section Two covers various aspects of coal petrology and palaeontology in seven papers. These include papers describing unusual variations of coal rank with depth in Moravia (Dvorak et al.) where coals remain at relatively low rank despite being buried beneath the Carpathian thrust sheets, and in South Wales (Gayer et al.), where high levels of heat flow and reversals in rank increase with depth are attributed to fluid flow within the basin. Other authors describe the results of various analytical approaches to the study of coal petrology, including solid state 13C NMR studies of fusinites (Premovic et al.), M6ssbauer spectroscopy of low rank coal lithotypes (Kostova et al.), and biochemical analysis of lignite (Stefanova & Magnier). Section Three deals with mineral matter in coal and the environment. The six papers include the sulphur contents of Pakistan coals (Baqri), of Yugoslavian lignites
viii
PREFACE
(Jankes et ai.) and of a multi bed coal in the UK (Cavender & Spears). Bouska et aL discuss the sulphur isotopic composition of North Bohemian lignites and Premovic et aL present the results of vanadium analysis in Kentucky coals. Section Four contains five papers concerned with mining geophysics. These include well logging techniques applied to the North Bohemian lignite basin (Mach) and the use of a deep gamma spectrometer (Gregor & Tezky). Seismic monitoring for rock bursts (Holub) and mining induced seismicity (Kalab) are two aspects of seismic investigation covered in the section. The final Section Five includes papers describing coal technology and coalbed methane. Liquefaction is discussed in two papers; one by Aleksic et aL using direct hydrogenation of low rank coals and the other describing experiments on beneficiated coal fractions (Barraza et aL). Desulfurization is also covered in two papers; one by Asmatulu et al. and the other by Whateley et aL, both dealing with unusual techniques to treat high sulphur Turkish coals. Gassification and coalbed methane generation from mines is covered by Douchanov & Minkova, Gryzbek et aL and Holub et aL, whilst Boardman & Rippon present an analysis of the influence of faults in coalbed methane production. The editors would like to thank all the authors for submitting the papers which represent a selection of those originally presented at the Second European Coal Conference in 1995 in Prague. We would also like to thank the many geologists who reviewed the papers: Mesdames & Messieurs Austin, Bouska, Brabham, Bright, Bryant, Cloke, Cole, Cornford, Davidson, Davies, Dopita, Drozdzewski, Ellison, Frodsham, Gayer, Gillespie, Glover, Goulty, Guion, Harris, Hathaway, Hemsley, Holub, Honek, Jelinek, Jones, Juch, Karayigit, Konecny, Kostova, Kropacek, Kumpera, McLean, Malan, Martinec, Miliorizos, Moore, Oplustil, Patrick, Pesek, Premovic, Querol, Rhodes, Rippon, Rosa, Simunek, Skocek, Spears, Spiker, Thomas, Turner, Wagner, Wakefield, Whateley. Many of the papers were written by authors whose first language is not English and this represented a problem not only for the authors but also for the reviewers. Both worked very hard to produce the present results. We have been continually amazed at the language skills of European geologists and hope that any slight errors remaining in the texts do not detract from the value of the volume. Sadly, one of the authors, Professor Kumpera, died before completing the final version of his major work on the geology of the Upper Silesian basin. Although his widow, Anna Kumperova, continued with the drafting of the diagrams, the conclusions have been added by the editors who accept responsibility for any errors inadvertently produced. We would also like to thank David Ogden, the staff editor at the Geological Society Publishing House for his continuing support and editing of this volume. Dr Rod Gayer, Cardiff Professor Jiri Pesek, Prague
Reference WHATELEY, M. K. G. & SPEARS, D. A. (eds) 1995. European Coal Geology. Geological Society, London, Special Publication, 82.
The Czech Republic energy policy: conception and implementation in a market economy JOSEF DORUSKA
Ministry of Industry and Trade of the Czech Republic, 11015 Prague 1, Czech Republic Abstract: In 1992 the government of the Czech Republic approved the 'Energy Policy of the
Czech Republic'. It was directed to the legislative and ecological respects which are compatible with European Union countries.
In the period since February 1992, when the energy policy was approved by the Government of the Czech Republic there have been many changes. For example, in July 1992 the Government approved its programme; the former Czech and Slovak Federal Republic was divided into the Czech and Slovak Republics; a major part of the energy companies was privatized; price adjustment of a considerable part of the energy commodities was abrogated and the Ingoldstadt oil pipeline construction was started. The Government has considered many other aspects which have a substantial influence on the energy sector including documents on: Governmental policy concerning the environment of the Czech Republic, of principles of the governmental mineral policy; European agreement on incorporation of the Czech Republic into the European Union; European Energy Charter; results of the Uruguay round of GATT; Convention on climatic change; and others. Because of the changes and new agreements there is a need to update the energy policy of the Czech Republic. The updated energy policy that is being elaborated by the Ministry of Industry and Trade, is created in such a way that the transition of the power industry would lead t o - i n technical, legislative and ecological respects - a compatibility with the power industries of the advanced countries of the European Union. The basic long-term objectives of the updated energy policy are: 9 to ensure sufficient energy supplies for the economy at acceptable prices; 9 to minimize negative impacts of energy production, distribution and consumption on the environment in order to reach a common level in the countries of the European Union; 9 to prepare the power economy of the Czech Republic for entry into the European Union in legislative and technical respects.
With respect to the above-mentioned principles, a programme of desulphurization of power plants has been accepted. By 1998 the following power plants will be equipped with machinery for flue gas desulphurization:
Tugimice II Prun6~ov Po6erady Tisovfi Chvaletice M~lnik II M61nik III D6tmarovice
800 MW 1490 MW 1000 MW 110 MW 600 MW 220 MW 500 MW 800 MW
The realization of this programme will significantly contribute to improvement of the environment. The government of the Czech Republic realizes the importance of: 9 ensuring the energy for the national economy 9 sustaining ecological limits resulting from the impacts on the environment 9 ensuring the permanently sustainable development of the national economy 9 fulfilling obligations of the Czech Republic resulting from the Energy Charter. The government is ready to react to changing conditions in energy supplies. The government realizes that with respect to: (1) the level of national reserves of fossil energy sources, and (2) the negative impacts of utilization of fossil energy sources on the environment; it is necessary to develop ways that will respect both the Energy Charter and conditions of permanently sustainable development of the national economy. The Czech Republic is one of a group of countries of Central and Eastern Europe where the transition process is taking place. In historical times the territory of the Czech Republic was part of the Roman Empire, whereas in the Middle Ages borders were
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 1-2.
output output output output output output output output
2
J. DORUSKA
difficult to define between dozens of kingdoms and principalities. At the beginning of the 21st century we are on the threshold of transnational integration of the energy sector. The basis of the economic prosperity of the European Union countries was in steel and coal, but the development of human knowledge has extended this to include information and communication areas. In these areas the Czech Republic is ready to play its full role.
References Resolution of the Government of the Czech Republic on the Power Policy of the CR, Prague, February 1992. Updated Energy Policy of the CR, Ministry of Industry and Trade of the Czech Republic, Prague, June 1994. The Basic Principles of the Governmental Mineral Policy of the CR (work version), Ministry of Economy of the Czech Republic, Prague, 1994.
Coal production and usage in the Czech Republic J. P E S E K 1 & M . D O P I T A ~
1Faculty of Science, Charles University, 128 43 Praha 2, Albertov 6, Czech Republic 2 Vysok{l s~kola bgt~skgt, 708 O00strava-Poruba, ti.17, listopadu, Czech Republic Abstract: Coal mining in the industrialized countries of Europe including the Czech Republic is witnessing a prolonged recession in the production of bituminous coal and lignite in particular due to reduced demands. Recession in numerous fields of industry has resulted apparently in lower production of metallurgical coke as well as in power generation. In addition, some countries have substituted the burning of solid fossil fuels with petroleum or natural gas or with other energy sources such as nuclear power, hydroenergy or geothermal energy. The Czech Republic is facing similar problems.
Judging from various scenarios presented by different institutions, it is becoming apparent that the production of bituminous coal in the Czech Republic even after the year 2000 will not drop dramatically below the level of production during 1993 and 1994. It is expected that about 14-16 x 106 metric tons of bituminous coal will be extracted in the year 2000 while the production in 1994 was about 17 x 106metric tons. Limited coal reserves in the workable levels of coal mines which will still be in operation in 2000 and whose prospects are promising, will require the development of new mine levels (e.g. Darkov and CSM mines in the Upper Silesian basin) and, around the year 2010, even the sinking of some new shafts in the Beskydy piedmont part of the Upper Silesian basin. Financial and time demands will play an important role when establishing such a scenario. The cost of developing a new mine level, taking into account the extent of the mining space required, the depth and the mining method, may be from 1.5 to 3 x 109 K6 at the current prices. Costs in developing a new mine can be as much as 20 x 109 K~ (see the Frengtfit mine) but the anticipated output from such a mine can only be achieved 10 to 15 years after commencing its construction. Some extra time is also required for conceptual issues, designing and for negotiations with legal entities operating in the region. These delaying factors argue for an intensification of studies to produce a long-term plan for solid and other fuel consumption in the Czech Republic in order to provide alternatives in the time span of 25 to 30 years. We consider the role of government to be paramount in this issue which is supported by numerous documents from industrialized European countries as well as from the USA. Our view, similar to that of Formfinek (1994) is that the role and importance of coal in the structure of primary energy sources has been underestimated. It is to be
noted that the Czech Republic always has been and will continue be more dependent on the output of coal in power generation in the year 2000 than any neighbouring country.
Electric power generation vs coal production and coal reserves The whole spectrum of problems related to coal mining can be divided basically in two groups. The first involves issues related to mining and necessary protection of coal reserves, whereas the second group involves issues related to improvement of the environment badly affected by mining operations. The policy to develop heavy industry following the communist coup d'gtat in February 1948, resulted in considerable increase in pig iron, steel and other energy demanding products. This resulted in a rather high consumption of electricity by the former Czechoslovakia (Fig. 1). As more than 90% of Czechoslovakia electricity was generated from lignite in the early 1960s, lignite production in the years 1946 through 1984 increased from 19.5 x 106metric tonnes to more than 101 x l06., and generation of electricity increased from 5.6 x 109 kWh in 1946 to 89 x 109kWh in 1989. Nuclear power stations at Jaslovsk6 Bohunice and Dukovany came into operation in the 1970s. The gradual introduction of nuclear power supplied only a part of the increase in energy demands of the former Czechoslovakia. However, in the 1980s, the * Similar considerable increase in production of lignite after World War II was recorded in former Yugoslavia, and in production of bituminous coal in former USSR, Poland and Australia. By contrast, a completely opposite trend in mining for coal in the same period of time was recorded for instance in the USA, Great Britain, France and Spain.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 3-12.
4
J. PESEK & M. DOPITA 1001
2 ......... 8Q-
..~ 60QO0e
t,..O 40-
20-
O. 1935
4'0 4's
~'0
5'~ 6~
6'~ 7~
7~
~0
~s
~0 ~9'95
Fig. 1. Electrical power generation in the period 1937-1994 (1) in Czechoslovakia; (2) in the Czech Republic. Note: Czech Republic constituted about two thirds of the territory of former Czechoslovakia.
patterns of production of electricity and coal (particularly lignite) began to differ considerably (see Figs 1-3). Whereas production of lignite reached its peak in 1985, the power generation culminated later in 1989. The relatively great difference between the decreasing coal production but continuing increase in power generation up to 1989 can be attributed to the electricity supply from nuclear power stations. Apart from the major decrease in demand for solid fuels in the former Czechoslovakia which is also evident in the Czech Republic, the coal mining industry remains the largest and most
important sector of the mineral raw materials mining industry. Coal has a prominent position among fossil fuels because natural hydrocarbon resources in the Czech Republic are negligible. Extraction of coal and its utilization has had a continuingly harmful influence on the environment, not only in coal mining districts (Ostrava region, Kru~n6 hory piedmont basin) but also in areas where its consumption has been concentrated such as around large coal-burning power stations and in large cities (e.g. North Bohemian basin, the M6lnik region, Prague agglomeration and other large cities).
120 2 .........
100
80 "~ ~176 ".~
~- 6o-
20
0
4935 4'0 4~,
5'0
5'5
r~
6's
7b
75
e:o
~
9'0 19'95
Fig. 2. Production of lignite (in metric tonnes) in the period 1937-1994 (1) in Czechoslovakia; (2) in the Czech Republic.
COAL PRODUCTION IN THE CZECH REPUBLIC
5
25
~Z Is
10-
0
-----r
193s 4'0 45
s'0
&
60
&
7'0
7's
8b
8's
ag'gs
Fig. 3. Production of bituminous coal both in former Czechoslovakia and in the Czech Republic (in metric tonnes) in the period 1937-1994.
It is obvious that the high coal consumption in former Czechoslovakia has resulted in a significant depletion of coal reserves with associated adverse impacts. Economic reform introduced after 1989 accelerated further reduction in demand for bituminous coal and lignite (Figs 2 & 3). This trend led to a decrease in coal production to economically and technically feasible levels and also to an abandonment of selected (and by that time) inefficient mines. The intentions of the Czech Government and pressure by local authorities are oriented towards a reduction of the negative impacts of both coal mining and its combustion on the environment (cf. Formfinek 1994, Reichmann 1994, Spousta 1995). These factors are reflected in the present position of the coal mining industry and in its search for the most suitable methods for a rapid inprovement of its situation. The solution lies not only in economical, technical and technological parameters, but also in averting potentially grave social and thus political consequences. The decrease in production of fossil fuels in the Czech Republic has resulted not only from the above factors but also from the restructuring of the economy which has led to a gradual reduction of metallurgical production and a concentration on the manufacturing of energetically less demanding products. However, it should be noted that the decline in industrial production in the Czech Republic and in the former Czechoslovakia in the years 1990-1993 has not been matched as yet by a similar decrease in the power demands. To manufacture products worth 1 USD, Austria and France needed in 1990, 0.18 and 0.2 kg of oil equivalent
respectively, whereas in the former Czechoslovakia the equivalent requirement was 1.6kg! These figures perhaps do not need any comment.
Durability of workable coal reserves in mined levels of the Ostrava-Karvinfi coal district and of other bituminous coal basins The Czech coal mining industry has reacted to reduced demand for bituminous coal and lignite in the same way as any other country's mining industry. Several underground mines have already been abandoned with the result that in the Rosice-Oslavany, Plzefi and Intra Sudetic basins, mining activities have ceased completely. Lignite production in the single underground mine in the Sokolov region as well as extraction of bituminous coal in the Ostrava part of the Upper Silesian basin has also terminated. Reduction of coal mining in the Peffvald part of the Upper Silesian basin is scheduled to commence during 1995. Various volumes of workable reserves were left behind in all the above-mentioned mines (Table 1). If these volumes are included in an overall diminution of workable coal reserves in the Czech Republic and when taking into consideration minimum profit achieved by the coal mining companies, it is necessary to discuss again the fate of workable reserves which our country has currently at its disposal. Ostravsk6 doly a.s., our largest bituminous coal mining company which together with the Dill CSM mine produce more than 90% of the annual output in the Czech Republic showed a profit of only
J. PESEK & M. DOPITA Table 1. Workable reserves in 10 6 t o n n e s left in abandoned bituminous coal mines due to closure programmes in the years 1990 through 1995 Sverma mine (OKR) Hehnanice mine Ostrava mine Odra mine J. Fu6ik mine Krimich mine Dobr6 ~t6sti mine Jind~ich mine Z. Nejedl2~ mine J. Sverma mine (VUD) Kate~ina mine
32.002 26.863 30.496 25.284 13.727 2.5 0.5 7.2 121.14 8.12 46.5
by January by January by January by January an estimate an estimate by January by January by January by January by January
147 x 106K6 (approximately 5.5 x 106USD) in 1992, and 31.4 x 106K6 (approx. 1.2 x 106USD) in 1993. The company applied selective but not always well-advised mining measures to improve mining methods. Nevertheless, it is necessary to separate strictly workable reserves which occur in the operating levels of existing mines from those whose development and extraction would require huge investment, particularly in the Czech part of the Upper Silesian basin (Table 2). However, individual mines of the Ostrava-Karvinfi coal district (OKR) have not at present sufficient financial resources for the required investment. It should be noted that one more underground bituminous coal mine in the Kladno region is still in operation. Its coal reserves, however, constitute less than 10% of workable reserves occurring in mined and developed levels of the Ostrava-
l, 1994 l, 1994 1, 1994 1, 1994 by the date of expected shut-down in 95 by the date of expected shut-down in 95 1, 1991 1, 1994 1, 1994 1, 1992 1, 1994
Karvinfi coal district. Table 2 shows that the coal reserves in currently operating mines will last without any large investment on average until 2010 or 2016, depending on the percentage lost during recovery, unless some unexpected event in the mining industry occurs in the Ostrava region e.g. abandonment of more mines, isolation of currently workable reserves due to regional or ecological limitations (see intensions of local authorities to outline a safety pillar under the city of Karvinfi and/or under the spa of Darkov). Introduction of these or other measures would reduce the lifespan of workable coal reserves in the Czech Republic (Pe~ek & Pe~kovfi 1993; Pe~ek et al. 1993). It is to be hoped that these alarming figures should provoke the relevant authorities into appropriate action. Almost 130 x 106 metric tonnes of workable reserves have been left in
Table 2. Workable reserves of bituminous coal in I06 tonnes in operating mines in the Czech Republic registered by December 31, 1994, and their duration in operating and developing levels Upper Silesian basin
total workable reserves of which confined to: operating levels developing levels designed levels total
586.6 149 164 164 454
Kladno basin
workable reserves in the Kladno mine Total workable reserves confined to operating and developing levels Anticipated output in the Czech Republic in 1995 (sine 1992)
19 332 16.4-16.7 i.e. average about 16.5 anticipated recovery factor 60% 100%
Decline in reserves of operating mines in 1995 Anticipated production in CR in 2000 (sine 1992) Decline in reserves of operating mines in 1996 through 2000 Anticipated volume of reserves by January 1, 2001 Life of mineable reserves in operating mines in operating and developing levels at the yearly anticipated output of 14.5 x 10 6 i.e. till the year
-23.1 14.5 -101.5 201.8
-16.5 14.5 -76.5 239.0
9.9 years 2010
16.4 years 2016
COAL PRODUCTION IN THE CZECH REPUBLIC prematurely closed mines of the O K R which represents 6 to 8 years of coal production in the Czech Republic. Further reserves appear to be irretrievably lost in the Rosice-Oslavany basin (RUD mines) and particularly in mines of the Intra Sudettic basin (VUD).
Issues related to profitability of underground coal mining with particular reference to bituminous coal Termination of bituminous coal production in the Rosice-Oslavany basin, then in NE Bohemia, and abandonment of coal mines in the Ostrava region and in western Bohemia (see above) was motivated either by reduced consumption or by unprofitability of underground mining. To date it has been questionable whether discussion of profitability (i.e. not subsidized) of coal production is meaningful under regulated (until recently) prices of coal and the still regulated prices of energy. This is the major issue from which the majority of partial problems are derived. According to an EC commission report, the average expenditure related to extraction of 1 metric ton of bituminous coal in EC countries in 1990 was equal to 200DM (3460K6, i.e. 102ECU). Expenditure in Germany was 260DM (132ECU). In 1994, their figures were 289 DM (147 ECU) in the Ruhr basin and 265DM (135ECU) in the Saar basin but the price of 1 metric tonne of an equivalent of bituminous coal imported into Germany was 70 DM (36 ECU) in 1994. Expenditure in Great Britain as only 150DM (76ECU) in 1994. Consequently, prices in Germany were subsidized by 54.5 ECU per metric tonne, i.e. about 1853 K6 (1 E C U = 3 4 K 6 ) . In Spain the equivalent subsidy was 26.4 ECU per tonne. However, in 1993 the subsidy in Germany increased to 69.5 ECU per tonne, and in 1994 to 215 DM (109ECU) per tonne in the Ruhr basin and 210DM (107ECU) in the Saar basin which is K6 3720 and 3633 respectively. In contrast, expenditure related to the extraction of 1 tonne of bituminous coal in Spain were reduced to ECU 19.8 per tonne in 1993, when numerous unprofitable mines were shut down. Comparing geological and mining conditions, the Ruhr basin appears to be unambiguously very similar to the OKR. Thus the intention to make underground mining bituminous coal in the Czech Republic profitable, seems to be highly problematic. It may be too late to reverse, but closing so-called non-profitable mines remains
7
highly questionable from the viewpoint of a long-term mineral policy. It may have been more realistic to compare the costs involved in closing selected mines with those related to subsiding operating mines. There can be no doubt that if underground mining for bituminous coal had been subsidized at a comparable rate to that applied in countries of the European Economic Community (see above) the so-called unprofitable Czech mines could have continued to operate. It is obvious that outlay of capital related to the development of new levels and/or a new mine would be higher than any mining company could afford. The question is whether it is more rewarding to extract the easily accessible coal reserves in mines scheduled to be shut down, but in which some investment has already been made and operation costs incurred, in order to develop some part of the mining space. The relatively small volume of workable reserves of bituminous coal should provoke the country's planner to reconsider the future of coal mining and to assess if, within the next ten to fifteen years, there will be any coal left to be extracted. There is no doubt that in the long term some revitalization of coal demand will occur. Coal should be considered not only as a traditional fuel for energy generation and production of coke but also and in particular as an irreplaceable raw material for the chemical industry. There is a requirement for the thoughtful manipulation of the coal reserves because present mining methods do not allow the remaining coal to be extracted from prematurely abandoned mines. Consequently, coal reserves of abandoned basins (coal districts) are lost forever including elimination of mining skills in the region. Reference to abandonment of numerous mines in industrialized western countries appears to be irrelevant when considering our specific situation. The fundamental difference between the Czech Republic and for instance Great Britain, Germany and/or other countries is in the fact that some of these countries including USA and Canada have considerably larger yet untouched coal reserves which can be exploited in the event of revitalized demands for bituminous coal.
Where to obtain energy after exhaustion of coal reserves? A considerable reduction in coal mining (e.g. France) or even its complete liquidation (e.g. Belgium) has taken place in several west European countries. The generation of electricity from classical sources is either substantially
8
J. PESEK & M. DOPITA
(France) or partially (.numerous west European countries) replaced by nuclear energy, indigenous or imported noble fuels or the partial substitution of indigenous coal by imported cheaper coal. The absence of large deposits of crude oil and gas in the Czech Republic, the slow and expensive construction of the Temelin nuclear power plant which, together with the high level waste repository, is opposed by both the Czech and foreign public, suggest that the Czech Republic could in the future be dependent on importing a large volume of coal for the generation of electricity, once the domestic fossil fuels have been exhausted. Consequently, prior to the complete exhaustion of the Czech coal reserves, the republic should either plan the construction of further nuclear power stations or make advance arrangements with potential coal exporting countries (such as Poland) for the supply of the necessary volume of coal. These negotiations should involve not only contracts containing long-term financial guarantees but also specifications and data on basic technological parameters of the imported coal including limits on harmful substances, etc. Published data suggest that coal from the Polish part of the Upper Silesian basin has for instance a higher content of sulphur. It is also necessary to determine the volume of coal that can be imported via the international transport network, particularly through the present railroad system, and also to consider boat transport, etc. However, if the coal is imported from other than neigbouring countries, then its import will be limited by the transport capacity of the transit countries and would also incur transit charges which show an increasing trend. Another issue involves payments for imported electricity or fuels. The administration would need to consider the financial sources required to pay for them. Such considerations are not premature for longterm planning. The future price of imported electricity or fuels should also be considered. The republic's present trade balance is static and the import of large volume of coal could seriously destabilize the situation. Despite the short-term fluctuations of prices we consider that the risk of a significant rise in price of coal and/or other fossil fuels should not be underestimated. This reasoning is especially pertinent in the event of further reductions or a complete shut down of coal mining in central and western Europe, bearing in mind that the present low wage manpower in South African Republic, Ukraine and other countries will not last much longer. In the case of Ukraine the present problems in coal mining may lead to a reduction of exported coal, particularly if Ukraine meets
its obligations and shuts down the Chernobyl nuclear power station by the year 2000. It is anticipated that this nuclear capacity will be replaced by the construction of new coal burning power stations.
Is coal a strategic raw material for the Czech Republic? We believe that this question deserves an unambiguous positive answer. Provided that more than 30-35% electricity requirement is generated by combusting mostly lignite from opencast mines (in the year 2000 still about 48%), then there can be no other answer. We recall the economic break-down resulting from the sudden extreme drop in temperature which occurred between December 31, 1978 and January 1, 1979. If we look at coal from another angle, the Czech administration should at least create the conditions and apply appropriate measures to secure enough coal reserves for operating power and heating plants at large agglomerations. This is required to prevent a reduction in power and heat generation leading to a complete breakdown of the whole economy because of, for instance, extreme climatic changes or other reasons such as long-lasting strikes of miners or railroad workers.
Regional and enviromental limitations stimulated by the Czech administration The Government of the Czech Republic has since 1991 passed several decrees constraining the limits of mining within the current coal mining areas particularly in the Kru~n~ hory piedmont coal basins (see e.g. S~korova et al. this volume, Bou~ka et al. this volume). The mining areas are delimited, according to law No. 44/1988 Coll. of the Czech National Council and in the wording of decree No. 172/92 from 16 March 1992 of the Czech Board of Mines, by the district Boards of Mines. The new mining law under preparation will attempt to reflect this situation. In our view, the situation is paradoxical, as the Government of the Czech Republic in order to reduce the negative impacts of coal mining on the population and environment, has developed its own decrees on the regional and ecological limits before making laws. There is no doubt that the destruction of tens of communities after 1948, including the ancient town of Most in northern Bohemia, resulted from originally useful objectives, i.e. making coal
COAL PRODUCTION IN THE CZECH REPUBLIC reserves available for mining (as we now know, wanton coal mining), but it also essentially affected the destiny of thousands of families that had to abandon their homes and native land. The current protection of other communities against liquidation unfortunately leads to other paradoxes. On the one hand, everyone, particularly the citizens in the North Bohemian basin, are right in calling for improvement of the environment in the basin area, but on the other hand, mining for coal with the lowest sulphur content in the North Bohemian basin at Chaba~ovice has been rather prematurely reduced. The same aspect applies to the activity of the 'rescuers' of the Libkovice village, considering that it is the underground mines in the North Bohemian basin which usually extract essentially better-quality coal than do the open cast mines (the coals from the former show usually lower sulphur and ash contents). The most controversial decision in this respect is the decree No. 441 from 1991 of the Government of the Czech Republic which constrains the development of the CSA open cast mine in the North Bohemian basin. This decree confines the extraction in this mine to the limits of the so-called first phase of its development. This reduces its reserves to such an extent that extraction will come to an end in 2007. The original mining scheme suggested that it would operate until about 2050, which was projected for the second phase of its development. The scheme would, nevertheless, entail the destruction of the villages of Ji~etin and t~ernice. The problem of a shorter or longer life for this mine, however, requires that its solution cannot be postponed until the first years of the next century, when a qualified decision could be taken based on actual needs. Extraction technology (turning of a face) requires the problem to be solved before the end of 1996. If the Government of the Czech Republic changes its earlier decision, not to allow the second phase, after this date the second phase would only be possible (if at all?) with huge financial losses and with considerable losses in coal recovery. The fact that the sterilization of these reserves will not only reduce the life of the mine by more than 40 years, but will also markedly influence that of the whole Basin should also be taken into consideration. We are of the opinion that the problems associated with destroying the communities were unnecessarily politicized. None of the large-scale open cast mining in densely populated Europe could have taken place without destroying those communities that were in the path of the mine developments. For instance, the Lower Rhine
9
basin in Germany, where more than ten communities had to give way to coal mining is not different from the CSA mine. Unlike the common practice introduced in former Czechoslovakia after 1948, the stoped out workings are immediately restored after abandonment, completely new villages are built and the inhabitants of the abandoned communities are offered adequate housing.
The problem of improvement of the environment Emissions of sulphur and nitrogen oxides that often exceed the limits from time to time give rise to air conditions approaching smog. This is particularly frequent in many of the densely populated towns where it has caused a reduction in life expectancy and calls for a radical solution that has been the subject of a number of crucial government decrees. It is a very serious problem with political overtones. The measures aimed at mitigating the negative influences on the environment should take place at two parallel levels. Whereas some solutions can be put into practice almost 'from day to day', the second group of problems can be solved only within a longer time interval and at considerably higher costs.
The medium- to long-term solutions Desulphurization of thermal power plants. According to the agreement on the atmosphere, all power plants in the Czech Republic will have to comply with emission limits that correspond to European standards by 1998. Until then, the Czech Energy Company must shut down in the thermal power plants obsolete units with a capacity totalling 2280 MW. Until now, 11 units with an output of 1225MW have been shut down. In five smaller power plants efficient fluid bed boilers will be installed, whereas the remaining 31, with a total output of 5730 MW, will be desulphurized by 1998 (Otava 1994). Installations of desulphurization systems in Czech thermal power stations will require extraction, preparation and transportation of a relatively large quantity of limestone (wet limestone washing), for the treatment. This will have a harmful influence on both the environment elsewhere, and also the limestone reserves. It will also be necessary to establish a market for the gypsum bi-product generated by the treatment, whose annual production will be 5 x 10 6 metric tonnes (M. Ku~vart pets. comm.), which is around five times the country's current consumption of gypsum.
10
J. PESEK & M. DOPITA
Fluid combustion of medium- to high-ash coals or high-sulphur coals. Fluid combustion (particularly under circulation or under pressure) would result in higher efficiency, and the use of reserves of coal with a heating value equal to or greater then 6 MJ kg -1 . These coals have not been considered for suitable mining due to their low quality. This could markedly extend the length of life not only of some of the mines, but of whole districts (basins) where these reserves are currently not extracted. There would also be an associated negative impact on the environment elsewhere resulting from the extraction and transportation of limestone which needs to be mixed with the high-sulphur coal prior to burning. Use of natural gas in major towns and power stations. A substantial or complete substitution of coal combustion by gas in the thermal power plants, heating plants and in the domestic heating in major towns in the Czech Republic provides a real possibility of reducing harmful emission in the most exposed regions. This concept has been both approved and initiated within an environmental program supported by 6.1 • 109K~ of government finance. It is, however, necessary to bear in mind the long-term dependence on natural gas from Russia with 37% of the world gas reserves. The construction of a new gas pipeline through Byelorus and Poland to western Europe, planned to transport 2-3 • 109m 3 gas to Frankfurt an der Oder by 1996, confirms a necessary diversification of gas sources as a safeguard against possible changing attitudes in the transit states. It will be necessary to explore the possibility of linking into this gas pipeline, whose capacity, however, is not planned to increase until long after the year 2000. We are not the only ones who take the view that, in the longer term, the prices of natural gas on the world markets will increase. If gas subsidies in the Czech Republic are completely removed by 1998, it may put a considerable financial strain on consumers. Should the prices of gas be higher for smallscale consumers than for large-scale consumers, as is suggested in some reports, we would consider this approach to be unjustfied.
Relatively short-term solutions Obligatory preferential supplies to large-scale consumers in major towns and in the North Bohemian basin with low-sulphur fuel. In the immediate future, we will only be able to monitor but not solve the major problem of S- and
N-oxide pollution that causes smog. Nevertheless, the problem could be solved by suitable legislative measures. The creation of suitable reserves of low-sulphur coal at the thermal power stations and urban heating plants would preclude unforeseen shutdowns of these facilities (see above). Alternatively, it might be possible to substitute high-sulphur lignite by the lowersulphur bituminous coal. This would, however, be possible only after the costly upgrading of furnaces in the heating plants and power stations. This would, nevertheless, bring some benefits. Owing to the higher heating value and lower ash content of the bituminous coal, the volume of the combusted bituminous coal would be lower as would be the volume of ash and/or clinker. Neither would it be necessary to carry out a costly desulphurization of the emissions from medium-size consumers (50 to 300 MW), since the sulphur contents in the coals used in power plants usually does not exceed 0.6%
Conclusions Public opinion in the Czech Republic, as in most developed countries is more and more concerned to see a rapid improvement in the environment. This should provide a driving force for change. It is clear that many of the current ecological problems are linked with coal mining and the burning of solid fuel. Therefore many countries are interested in developing new technologies for coal use covered under the heading of clean coal technology. Its principal goal is to make the use of coal more environmentally friendly and at the same time to increase the efficiency of its combustion. Several technologies have been industrially established, others are being tested. In the USA, which has the largest viable bituminous coal reserves in the world, a great deal of attention is paid to the problems of bituminous coal liquefaction and to the production of coal bed methane. It has been estimated that about 20 to 25% of coal reserves in the USA are high sulphur coals (1.5% and more). In our opinion, coal liquefaction is not an option in the Czech Republic in the foreseeable future. This paper therefore, has concentrated on several solutions which should be adopted in a systematic way. Some of these would lead to an improvement in the environment and to an increase of reserves of a material that might already be exhausted by 2020. As reserves dwindle coal will no longer be the principal source of environmental pollution, but will remain an important raw material for the chemical industry.
COAL P R O D U C T I O N IN THE CZECH REPUBLIC Everyone of us should bear in m i n d that this will be the time of our children's generation. We do not want them to deplore the imprudence of their parents! But we must not be unrealistic in thinking that whatever our e c o n o m y will need, can be easily bought or imported. This raw material (as well as others) will be hard to pay for and it will be necessary to create (and finance) the conditions enabling to realize this idea.
References FORMANEK,V. 1994. The need for well-advised energy policy in the Czech Republic. (In Czech). UhliRudy-Geologickf~ prdzkum, 1, 405-408. OTAVA, B. 1994. Clear clouds of Po6erady made by CEZ (Czech Energy Company). (In Czech). Koruna, LN III, 3 (Dec. 8, 1994).
11
PEgEK,J. & PESKOV~,,J. 1993. Prospects of mining and decline in coal reserves of the Czech Republic. (In Czech). Uhli-Rudy, 41, 136-138. & 1995. Coal production and coal reserves of the Czech Republic and former Czechoslovakia. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds) European Coal Geology. Geological Society, London, Special Publication, 82, 189-194. - - , DOPITA, M., OPLUSTIL,S., PEgKOV.A,,J., KOULA,J. & ZELENKA,O. 1993. Real volume of coal reserves to be extracted in operating mines of the Czech Republic. Part I. Bituminous coal. (In Czech). Uhli-Rudy, 41, 307-316. REICHMANN,F. 1994. Ecology and mining for mineral raw materials. (In Czech). Uhli-Rudy-Geologick) prdzkum, 1, 414-416. SPOUSTA, Z. 1995. A commentary to the document: "The role of coal industry in energy policy and the role of state in mining". (In Czech). UhliRudy-Geologick~ pr~zkum, 2, 20-21. -
-
Controls on the evolution of the Namurian paralic basin, Bohemian Massif, Czech Republic O. K U M P E R A
VSB-Technical University of Ostrava, Institute of Geological Engineering, Ostrava-Poruba, tK17.listopadu, 708 33, Czech Republic Abstract. The Namurian A paralic molasse deposits of the Upper Silesian Coal Basin form erosion remnants of an extensive foreland basin located in the eastern part of the Bohemian Massif. This basin represents the latest stage of development of the Moravian-Silesian Paleozoic Basin (Devonian-Westphalian). The paralic molasse stage of the foreland basin evolved from foreland basins with flysch and with marine molasse. The deposition of the thick paralic molasse (Ostrava Formation) started in the Namurian A. In comparison with other coal-bearing foreland basins situated along the Variscan margin in Europe, this is characterized not only by earlier deposition, but also by a different tectonic setting. It is located in the Moravian-Silesian branch of the Variscan orocline striking NNE-SSW, i.e. perpendicularly to the strikes of more western European foreland basins. In the Vis6an and Namurian, the foreland basin developed rapidly under the influence of the western thrustfold belt in the collision zone. The deposition was influenced by contrasting subsidence activities of the youngest and most external trough - Variscan foredeep - and the platform. The Upper Silesian Basin shows therefore a distinct W-E lithological and structural polarity and zonation.
The Upper Silesian Coal Basin (Fig. 1) represents one of the most important European paralic and limnic hard coal basins. The boundaries of the basin are not completely known as its coal-bearing sediments are mostly covered with younger sediments and can be seen only in small outcrops. The rocks are mainly known either from deep exploration and/or structural boreholes or from mines. The total known area of the basin is approximately 6500km 2, of which more than two thirds are situated in Poland. The Czech part of the basin, the Ostrava-Karvinfi coalfield (Dopita & Kumpera 1993a), is located in the southern parts of the basin with a known area of around 2000km 2. However, the actual extent of the Czech part of the basin is far greater as shown by prognostic studies (Zeman 1977) and paleogeographic analyses (Turnau 1962, 1970; Dopita & Kumpera 1993b). Coal-beating sediments in the south of the district were found at mineable depths in deep boreholes only in the area shown in Fig. 4. According to geophysical data, further south they plunge steeply south under the nappes of the Outer Carpathians in the zone of the E - W striking Sulov faults, where they have been found in the Jablfinka 1 borehole at a depth of 2985-3870m under the nappes northwest of Vsetin (Polick~, & Hon6k 1984). The extent of the Upper Carboniferous coal-bearing sediments beneath the Carpathians nappes must be even greater as proved by deep boreholes in the
surroundings of N~m6i6ky in Southern Moravia (Purkyfiovfi 1978). In these boreholes, the coalbearing Carboniferous formations were shown to be below mineable depths with the top of the formations being at 2711 m and the base greater than 4787m. These data suggest that before denudation, the Czech part of the Upper Silesian Basin covered a great area in the south and southeast, which was not limited by the Czech boundary, and that the erosional remnants of the coal-bearing Carboniferous formations are preserved in half-grabens south of the zone of the Sulov faults, buried beneath the nappes of the Outer Carpathians.
The paralic molasse (Ostrava Formation Namurian A) in the framework of the Moravian-Silesian Paleozoie Basin The molasse formations of the Upper Silesian Paleozoic Basin form part of the thick DevonianCarboniferous accretionary wedge, which is preserved as erosional remnants of a large basin at the eastern margin of the Bohemian Massif (Fig. 1). The basin formed as a result of a continental plate collision at the eastern border of the Bohemian Massif from Devonian through Westphalian times (Kumpera & Foldyna 1992). In the collision zone several units of the Czech Massif meet. These units are defined in the
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 13-27.
14
O. KUMPERA
Fig. 1. Schematic geological map of the Moravian-Silesian Paleozoic Basin (compiled after J. Dvo~fik, A. Kotas, M. Dopita, O. Kumpera, J. Foldyna) 1, Devonian; 2, Permian; 3, plutonic complex of Brunovistulian basement; 4, crystalline complex of Brunovistulian basement; 5, Namurian A-predominantly coal-bearing paralic molasse; 6, Namurian B-Westphalian-predominantely coal-bearing continental molasse; 7, Lower Carboniferous; 8, remnant basin relic; 9, borehole Krfisnfi; 10, locality of Fig. 3.
Report of the Working Group for Regional Geological Classification of the Bohemian Massif at the former Czechoslovak Stratigraphic Commission (Chulpa6 & Vrfina 1994). In the west, internal orogenic zones of the Bohemian Massif are interpreted as the hangingwall to this collision zone (Fritz et al. 1993). The eastern Cadomian block of the Brunovistulian basement formed the footwall which gradually disintegrated and subsided (Kumpera 1988) during an oblique collision (Grygar 1992). The deeply eroded roots of the collision suture are located in the Silesian and Lugian units in the north and in the Moravian and Moldanubian units in the south. The collision of the two plates of a contrasting crustal character resulted in a rapid uplift in the central parts of the Bohemian Massif and the formation and evolution of subsiding and migrating foreland basins, the final basin being located on Brunovistulian basement (Fig. 2). This consequently led to the development of a thick sedimentary accretionary
wedge with a complex composition and structure. The preserved filling of the basin represents a rather small erosional remnant of a far larger basinal structure. The Moravian-Silesian Paleozoic Basin underwent a complicated evolution in the course of the collision covering several types of basins (Klein 1987). These are described by Kumpera (1983), Hladil (1988), Dopita & Kumpera (1993a), and Kumpera & Martinec (1995). Carboniferous foreland basins with molasse represent development during the latest stage of oblique collision.
Partial troughs within the Moravian-Silesian Paleozoic Basin The basin depocentre migrated from the collision zone towards the foreland (Kumpera 1971; Dvo~fik 1973). The sedimentary wedge mainly consists of siliciclastic flysch and molasse sediments (a smaller part belongs to platform
NAMURIAN PARALIC MOLASSE IN BOHEMIAN MASSIF
15
Fig. 2. Schematic conception of the partial flysch troughs and molasse fordeep and of the surrounding source areas in the Moravian-Silesian Paleozoic Basin during the Late Vis6an and Namurian A stages of development. Two of the troughs and depressions are in Fig. 3. carbonates and rift d e p o s i t s - K u m p e r a & Martinec 1995). The total thickness of the Carboniferous sedimentary wedge (after compaction) is more than 12km, although, due to the prograding wedge, this is never developed in one location. The average rate of sedimentation is therefore about 270 m Ma -1 . The accretionary wedge has its greatest thicknesses in the west, in the vicinity of a broad collision zone, and above all in the northwest. The preserved maximum thickness is about 6 km, whereas the total thickness decreases gradually to 200 m in the extreme eastern parts of the basin. The decrease is not continuous, but the thickness distribution varies within narrow partial troughs and forebulges (Figs 2 & 3) which developed during the Carboniferous (Kumpera 1983). Some of these troughs and elevations have been proved conelusively by isopach studies. Two partial basins
and forebulges of different ages are well documented in the eastern part of the Paleozoic basin (Fig. 3). In the western trough (eastern flysch basin in Fig. 3), the thickness of Upper Vis6an flysch is reduced from 2500m in the western part to 100m, or less, in the eastern part. In the eastern elevation, stratigraphic gaps at different Vis6an levels have been even observed. The eastern trough is filled with thick uppermost Vis6an and Upper Carboniferous molasse deposits.
The Variscan foredeep and platform in the development of the Upper Silesian Coal Basin The initial ideas of uniform geotectonic development of the Upper Silesian Basin have been
w
E
FORELAND BASIN EASTERNFLYSCH BASIN Budi~ov
PLATFORM
1
0
I
....
.'.
.....
FOREDEEP Odry-Hranice ; ......
,
._.;
Vala~sk~ Mezi~I~f
..
FOREBULGE I " Orlov~ s t r u c t u r e Fren~t~t p. Radh.
;
" i - . 9~ 9. . . . . . , . 9 ,..~r
am " ur
i
,
]
A ~-.~-~
~
I
2
3
4
5
Fig. 3. Palinspastic cross section showing the thickness of stratigraphic units at the transition from the foreland basin with flysch to the foreland basin with molasse in the Moravian-Silesian Paleozoic Basin during Late Visban (Goa-~) and Namurian A. l, carbonates; 2, predominantly shaly deposits; 3, predominantly graywackes; 4, conglomerates; 5, coal-bearing paralic molasse.
16
O. KUMPERA
modified in terms of a polytype basin (Havlena 1982) as a result of deep exploration boreholes. The first boreholes sunk in the southern and southeastern parts of the basin within the coalbearing Carboniferous showed considerable differences between the area of active mining in the north and areas under exploration in the south. With continual exploration, it has been determined that thicknesses of parts of the Carboniferous sequence diminish towards the east and southeast. In addition, their coal capacity and the thickness and number of seams also decrease. Extensive borehole exploration throughout the area of the basin together with new data from the deeper levels of the mines has proved that in the course of the Carboniferous development, the coal basin was divided into two parts with distinct developments and structures (Fig. 3): 9 The younger Variscan foredeep, represented by a narrow mobile zone along the western margin of the basin. 9 The Upper Silesian stable block, an extensive platform in the eastern part of the basin. This division is only apparent in the Upper Vis~an and Upper Carboniferous levels of the basin. During the Devonian and the earliest Carboniferous, the whole preserved part of the basin was a platform. Thus, the Devonian and Lower Carboniferous carbonates have a similar thickness throughout the basin (up to 700 m in the south and up to 1100m in the northern parts of the basin in Poland). By Early Vis~an time, Variscan deformation had reached the eastern boundary of the basin. It is only the Upper Vis6an formations that are largely of a clastic character. Their thickness reaches 1000-1500 m in the foredeep but eastwards, towards the platform, decreases to 100m (in the Krfisn~, 1 borehole- Roth 1979). Correlated stratigraphical units of the molasse vary considerably in thickness across the basin. The foredeep is filled with marine and paralic molasse sediments, whose compacted thickness is up to 4500m (before compaction, the thickness may have reached more than 6000m) in the depocentre, whereas it decreases to 200m or less over the easternmost platform forebulge. In addition to variations in thickness, the foredeep and platform differ markedly, especially in the development of the paralic molasse (Ostrava Formation - lower Namurian - Ez zone). The main feature of the coal-bearing paralic molasse deposition was that subsidence was compensated by clastic supply. Nevertheless, the sedimentation was influenced by the
contrasting subsidence rates of the foredeep and the platform (Fig. 4). Thus the foredeep is characterized by a full subsidence compensation and the platform by a retarded subsidence.
The main lateral changes in the iithological development of the Namurian paralic molasse (Ostrava Formation) See Table 1 of Dvo~ik et al. this volume, for the stratigraphic classification of the mollasse-filled foreland basin.
Thickness of the Ostrava Formation First of all, the foredeep and the platform differ in the thickness of the paralic molasse. In the foredeep, the thickness of the Ostrava Formation reaches up to 3200 m and decreases to about 100m or less in the platform forebulge (in the vicinity of the Kop~ivnice-T~inec anticlinorium). These changes are well illustrated on isopach maps of the total thicknesses of individual lithostratigraphic units of the Ostrava Formation, namely the Pet~kovice and Jaklovec Members (Figs 5 & 6). All members show maximum thicknesses in the foredeep (especially in the north), but thicknesses decrease to the east, particularly the southeast, in the area of the forebulge. This contrast in thickness gradually diminishes in successively younger stratigraphic units, from the oldest (Peffkovice Member) to the youngest fully preserved unit (Jaklovec Member). The youngest unit of the Ostrava F o r m a t i o n Poruba Member - has not been studied because the upper part of the member is nowhere preserved. Estimates of sediment accumulation rates across the whole basin vary from 250350 m Ma -1 in the Late Vis6an flysch depression, through a surprisingly high 9 0 0 m M a -1 in the foredeep during Namurian A, to zero in the platform forebulge.
Lateral changes in coal accumulation The foredeep and the platform differ not only in the thickness of the paralic molasse, but also in the number and thickness of coal seams developed. Coal accumulation decreases markedly towards the platform (Dopita & Kumpera 1993b). The Ostrava Formation contains more than 170 coal seams with an average thickness of 73 cm in the foredeep, whereas the number of coal seams in the same stratigraphic interval
N A M U R I A N PARALIC MOLASSE IN BOHEMIAN MASSIF
17
OL4N b " " -lZ N
/ ,~,j-,-,.,--~.,~
|
~J
I, - - . . .
.9
\ ," ..."~
ZN "-. .....
1
o-'*
7n j"
FB
~
a b c
"- .'7'.
o,,
:'"..
L. "1."
~
5 .....
!---
~
.No.
".
:.'|
'.. /
i,!
-
7
/
-
i C~F'~,-r
9
,.
,
.L!.."| "
-
-@ \~,--~
~,
>"
>
>
10 11
|
12
13 I~
os-
0 L
~
50km
14
Fig. 4. Outline of the early Namurian palaeogeography in the Czech part of the Upper Silesian Coal Basin. 1, paralic molasse in known regions (with compensated subsidence); 2, postulated pre-erosional extent of paralic molasse; 3, paralic molasse in regions with retarded subsidence; 4, source areas (a-lowlands, b-hills, c-mountains), ZN-prograding thrust zones; 5, present erosional limits of the known basinal regions; 6, postulated original limits of the basin; 7, depocentre axis; 8, postulated forebulges; 9, directions of marine trangressions; 10, directions of clastic transport; 11, rivers and deltas; 12, humid climate; 13, postulated volcanic centres; 14, state boundaries, FB-postulated forebulge. decreases to 40, or even 20, in the platform, where some parts of the Ostrava Formation are even non-productive. This is illustrated in the maps of total coal accumulation of the Pettkovice Member (Fig. 7) and the Jaklovec Member (Fig. 8) with maximum values in the foredeep and minimum values in the vicinity of the platform forebulge. These maps show the division of the basin into sub-areas, where subsidence was more or less compensated by sedimentation. Spatial variations in coal accumulation in the sub-areas are thought to have been controlled by the interplay between tectonics and clastic supply. The isopach maps suggest a slight shift of the maximum coal accumulation eastwards with the younger units, probably due to an eastward migration of the depocentre. As with total sediment thickness, the contrast between the foredeep and the
platform diminishes towards the younger stratigraphic units, particularly in the southern areas, where an extensive platform forebulge was gradually uplifted in the vicinity of the present Koptivnice-Ttinec anticlinorium.
Lateral changes in the distribution of other lithotypes Simultaneous changes in the number, thickness and character of various correlatable stratigraphic units can be observed between the foredeep and platform. Up to 80 marine and brackish bands, representing marine transgressions, have been recorded altogether in the northern part of the foredeep (Reho~ & l~eho~ov~ 1972) These show progressively less
18
O. K U M P E R A
2F,
0I,
!
1(2
20 ,,I
km
Fig. 5. Isopach map of the Pet~kovice Member (lower part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin). 1, present limits of the basin; 2, limit of the Frengt~tt relic of the Karvinfi Formation; 3, isopach in metres. marine influence or disappear towards the platform so that only the four most important, thick bands are present in the eastern part of the platform. The thickness of those bands with the most stable areal distribution is up to 180 m in the foredeep but is markedly reduced towards the platform. The faunal content changes significantly also from north to south in the area of the foredeep. Towards the south, elements of a brackish fauna, or even a freshwater fauna, occur more often at the expense of elements of a marine fauna in a considerable number of marine bands (l~eho~ & l~eho~ovfi 1972). These changes are even more marked in a W - E direction. This indicates that marine
conditions transgressed from north to south and west to east through the foredeep and, from time to time, even reached the area of the Upper Silesian platform block. The foredeep, in which each marine transgression resulted in at least 22m of accumulated clastic sediments, represents an area regularly flooded by the sea, whereas the platform was an area only sometimes flooded (Havlena 1982). In addition, volcanoclastic rocks present in the f o r e d e e p altered tuffites in terrigenous siliclastic sediments, kaolinite tonsteins in coal seams (up to 16 beds in the f o r e d e e p - Dopita & Krfilik 1977), and kaolinized tuffites redeposited as 'whetstone' rocks (30 beds in the western part)
NAMURIAN PARALIC MOLASSE IN BOHEMIAN MASSIF
Q
10 -,
i
19
20 km
Fig. 6. Isopach map of the Jaklovec Member (Upper part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin). For 1-3, see Fig. 5.
disappear towards the platform. This can be seen in the isopach map (Fig. 9) of the s.c. Main Ostrava Whetstone, which forms an important marker bed in the lower part of the Ostrava Formation over large areas of the basin. Figure 9 shows a decrease in its thickness from 12m in the west to 0 m in the east. It also shows marked
local variations in the thickness of the Main Whetstone that probably represent a combination of subsidence and fluvial control. Volcanic material was largely redeposited into the foredeep from the platform by complex sedimentary processes. A map of composite thickness of all the volcanogenic beds in the Ostrava Formation
20
O. KUMPERA
?'g'l
-/ /
0
z9
Fig. 7. Total coal accumulation map of the Pet~kovice Member (lower part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin). For 1-3 see Fig. 5. (not shown), shows a similar pattern. It must be stressed that some of these beds have an areal extent ranging from 102-103 km 2. The thickness of individual cyclothems, the grain size and the number of sandstone and conglomerate layers increase towards the southeastern part of the platform indicating a source area in the vicinity of the Kop~ivnice-T~inec anticlinorium - Fig. 10 (Jansa 1967).
A similar pattern of the main trends in the development of the basin is given by the distribution of coalification intensity (Adamusovfi et al. 1992). The more modern methods of coal quality determination have not been sufficiently and equally applied to the whole area of the Czech part of the Upper Silesian Coal Basin so that the degree of coalification has been characterized by volatile matter V aaf. The map
N A M U R I A N PARALIC MOLASSE IN BOHEMIAN MASSIF
21
2#";71 F"-t //
/ ,o
I
i
I
,o 1
0
10
20kin
Fig. 8. Total coal accumulation map of Jaklovec Member (Upper part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin). For 1-3 see Fig. 5.
of coalification of the upper surface of the Ostrava Formation again illustrates the basic difference between the area of the Variscan foredeep and the Upper Silesian platform. In the area of the Variscan foredeep on its west margin, the V daf values fall to less than 10~ whereas the
lowest V daf values in the area of the Upper Silesian platform are only 20% and the maximum is more than 35% V daf in the youngest part of the paralic molasse sequence. Generally, the degree of coalification is controlled by stratigraphy; the coalification is lower
22
O. KUMPERA
,t _l
/
0
10
.I__
2,0kin
i.,
Fig. 9. Isopach map of the Main Ostrava whetstone horizon (lower part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin),For 1-3 see Fig. 5.
in the younger members of the formation. However, at the same time, the intensity of coalification depends on the thickness of the basin fill in any region. At comparable stratigraphic levels, the degree of coalification in the Upper Silesian platform is lower than that in the Variscan foredeep. The lowermost degree
of coalification occurs in the Namurian A in South Moravia (up to 40.9% Vaaf). Doubtless, this is connected with a generally low subsidence rate in the southern areas of the MoravianSilesian Paleozoic Basin. By contrast, the highest degree of coalification in the western part of the foredeep is probably
NAMURIAN PARALIC MOLASSE IN BOHEMIAN MASSIF
23
9
4r,-
I--3 1 2 P _z, l
-,
L..._./
0 I
10 I
20 km ,.J
Fig. 10. Isolines of sandstone index in the Jaklovec Member (Upper part of the Ostrava Formation, Czech part of the Upper Silesian Coal Basin).l, limits of the basin; 2, limits of the Fren~tfit relic of the Karvinfi Formation; 3, sandstone index (%).
partly a result of the deepest burial of coalbearing strata during the coalification processes, and partly due to the greatest heat flow at the Variscan front in the west of the basin. This is in good agreement with the results from the whole Moravian-Silesian Paleozoic Basin (Sko6ek 1976).
Structural zonation The Upper Silesian Basin is stratigraphically and paleogeographically (e.g. Havlena 1982) as well as geotectonically and structurally (Kumpera 1971, 1980; Kotas 1985) asymmetric. The main manifestation of tectonic asymmetry in a
24
O. KUMPERA
W - E direction is the greater intensity of folding and the development of more complex structures in the western part of the basin. In the eastern part of the basin within the area of the Upper Silesian platform, the prevailing structures are taphrogenic. Structural asymmetry parallel to the Variscan orogenic trend is expressed in somewhat more complicated tectonic styles in northern areas. Thus, the regional tectonic scheme, like the other features described above, prove a generally greater mobility of the basin in the west and in the north. A detailed analysis of basinal tectonic structures indicates the presence of structural zonality within the foredeep area. Here, parallel to the axis of the foredeeep, narrow zones trending in a NNE-SSW direction and characterized by unusual fold-fault structures are present. Thus, it is possible to delineate a zone containing a holomorphic style of folding along the western margins of the basin. To the east, this zone is fringed by the zone of western brachystructures. After that, the area of idiomorphic (ejective) folds follows, within which the parallel narrow zones of the Mich~ilkovice-Rybnik and OrlovfiBohugovice anticlinal fold-fault structures are developed. Between these two anticline-fault structures, the zone of eastern brachystructures is located. The tectonic zones are characterized by specific fold-fault structures (Foldyna & Kumpera 1991).
The main vertical changes in the development of the paralic molasse (Ostrava Formation) Vertical changes in some lithological parameters point to important processes in the development of the basin as well as to the changing basinal regime.
The abundance of marine facies Marine conditions greatly influenced the sedimentation of the lower part of the Ostrava Formation but became less significant upwards. This can be demonstrated by the number of marine units in successive members of the Ostrava Formation. In the Pet~kovice Member, up to 32 bands with marine or brackish faunas are known, in the Hrugov Member 26 and in the Jaklovec Member, marine or Lingula-bearing bands are represented only by the Susta marine band and the Barbora group of 4 bands. Marine influence increased again in the Poruba Member, where up to 20 bands (Reho~ & l~eho~ov~ 1972) have been found. By comparing the maximum
thickness of a member with the maximum number of marine occurrences documented by bands in that member with a marine or brackish fauna, the frequency of marine influence can be quantified. The greatest effects occur in the Peffkovice Member (1 transgression per 22m thickness of sediments). Lesser effects are recorded in the Hru~ov and Poruba Members (1 transgression per 94 m and 49 m of sediments, respectively). These data indicate that although considerable variations in subsidence rate occurred (as shown by the thick nonproductive sequences with a prevailing marine influence upon the megacycle boundaries), the rate of subsidence gradually decreased.
Periodical variation between marine molasse and paralic molasse Thick and laterally extensive marine units occur, widely separated by 300 to 500 m thick sequences of paralic molasse containing great coal accumulations. The most important are the marine bands associated with the seams Naneta, Franti~ka, Enna, Barbora, Roemer and Gaebler. These marine bands have a rather constant faunal content over the whole Czech part of the basin and have a considerable thickness up to 180 m. Periods of extensive marine flooding are associated with considerable falls in coal accumulation or even in the development of barren measures (Dopita & Havlena 1980). Closely associated with the most stable faunal bands in the Ostrava Formation are nonproductive or only weakly coal-bearing sequences, whose thickness varies between 100-240m. They divide the richly coal-bearing sequences of the Ostrava Formation. The lithological nature and mainly a marine origin of nonproductive parts of the sequence indicate periods, when subsidence remained uncompensated over almost the whole of the Czech part of the basin. Thus the Ostrava Formation can be classified as a polyfacial sequence, in which the prevailing paralic molasse was several times interrupted by the development of marine molasse. The marine molasse was formed, in contrast to the paralic molasse, during periods of uncompensated subsidence. This suggest changes in subsidence rates and tectonic activity both within the basin and in the source area.
Changes in volcanic activity The layers of volcanogenic sediments reflect intensive volcanic activity in volcanic centres, which have yet to be identified in detail, although
NAMURIAN PARALIC MOLASSE IN BOHEMIAN MASSIF they are probably in the western source area. The most important is the Ostrava Whetstone, whose thickness is up to 12 m, and which is developed over almost the whole area of the Czech part of the basin with the exception of the easternmost areas. Another important unit is the stratigraphically lower whetstone of the Leonard seam. The early Namurian A was the most volcanically active period in the near source area, producing frequent and thick layers of whetstone in the paralic basin. Later, the volcanic activity waned and mostly only produced layers of tonstein in the swamps and peat moors which were quiet sedimentary environments protected from resedimentation. Thin pyroclastic layers falling into a high energy environment would have been kaolinized and dispersed amongst clastic materials. Nevertheless, in the lower Namurian A sequence up to 46 units of volcanogenic sediments have been preserved at various stratigraphic levels. Their maximum total thickness is 16 m indicating that the Namurian A represents a period of strong and frequent volcanic activity in the Variscides of Central Europe. The diminution in the total thickness of volcanogenic sediments to the east suggests their redeposition from the platform to the foredeep.
Changes in the petrographic and geochemical composition Changes can also be observed in the petrographic composition of psammites through the stratigraphic sequence. Among them, graywacke sandstones prevail. In the upper part of the paralic molasse, the number of graywacke layers, arkosic sandstones and arkoses increase (Fialovfi et al. 1978). This relates to important changes in the source area. Whereas in the lower part of the Ostrava Formation, the major source lay in the western orogenic area together with resedimentation of older Carboniferous clastics; in the upper part, the influence of the eastern source area of the stable Upper Silesian block (e.g. from the area of the forebulge) gradually manifested itself. This is also indicated by the geochemistry of some claystones suggesting an increasing supply from morphologically flatter source areas which were exposed to long-term chemical weathering.
The influence of eustatic movements upon sedimentation Some major changes in the lithology of the Namurian paralic sequence can be related to
25
glacio-eustatic changes of the sea level as well as to climatic oscillations during the Late Carboniferous. They can be correlated with the mesothems described by Ramsbottom (1979) from northwestern Europe (Sko6ek 1991). A shortage of radiometric age data makes it difficult to interpret the influence of climatic changes on sedimentation in the Carboniferous of the Upper Silesian Basin. However, it would expected that these changes could contribute to lithological changes in cyclothems, or they could be reflected in marine transgressions which might result in a complex interplay with tectonics.
Some paleogeographic features of the basin As with the earlier stages of the development of the Moravian-Silesian Basin, the Namurian A paralic molasse basin developed on continental crust in the foreland under a compressional regime. Therefore, the basin was filled by erosion of rocks of the overthrust lithospheric plate both in the inner parts of the Bohemian Massif and in the foreland thrust zone. This latter consisted of the waning collision zone and also the western areas of the Paleozoic rocks in the Moravian-Silesian Basin, which were, progressively, included into the thrust-fold zone. Also sedimentation in the paralic molasse took place partly at the expense of synsedimentary uplift within the platform foreland. Active tectonic development both in the source area and in the area of the basin itself, resulted in a significant resedimentation of elastics connected with the processes of basinal cannibalism (Kumpera & Martinec 1994). The Czech part of the Upper Silesian Basin that is preserved today, represents an erosional remnant of the former basinal structure that was originally substantially larger than today. The western part of the basin-fill had been already eroded by the end of the development of the Variscan fold-thrust structure and the uplifts of the Rheno-Hercynian and Sub-Variscan zones. The axis of the maximum compensated subsidence is thought to be situated to the west of the existing erosional western boundary of the basin. This axis plunged to the north. Towards the western margin of the basin, all indications of mobility become more marked: thicknesses of stratigraphic units (Figs 5 & 6), coal accumulation (Figs 6 & 7), degree of coalification, thickness and number of faunal bands, etc. Along its western boundary, the basin is amputated by tectonics and erosion. It can be assumed that within the early Namurian, the basin extended
26
O. KUMPERA
along the whole eastern margin of the Bohemian Massif (Fig. 4). The original southern margin of the basin is not known, but on the evidence from deep boreholes located at Nrm~i~ky (Dvo~fik et al. 1997) as well as an analysis of the tectonic position, it is probable that the basin extended originally as far as the Austrian border, where its remnants are still preserved. Likewise, it is possible that the coal-bearing deposits covered a considerable area to the southeast and east beneath rocks of the Styrian (Carpathian) nappes (Dopita & Kumpera 1993b). As described above various source areas can be postulated for the basin fill. The western source area comprised both the areas formed by the crystalline complex and cover sediments. The geological composition of the denuded surface changed in the course of sedimentation during the Namurian A. A great amount of arkosic sandstones and arkoses in psammites in the upper part of the Ostrava Formation and a considerable content of K feldspars indicate that the level of erosion had cut down into the larger granitoid bodies. During the later Namurian A eastern source areas joined those in the west. In contrast to the orogenic western source, these less extensive source areas were intraplatform elevations in the foreland. Their intermittent influence was first seen in the sedimentation of the Hru~ov Member by the Kopfivnice-Tfinec uplift. In the later Namurian A low elevations within the platform areas sourced short streams that fed the more eastern part of the basin. The relief of source areas became gradually less and less flat and produced more chemically mature solid products of weathering (Kumpera & Martinec 1993).
onto the forebulge; (b) the development of marine bands and the degree of marine influence diminish, indicating marine transgression towards the south and east; (c) the thickness of volcanogenic coal tonsteins and whetstonesboth show a decrease in thickness towards the forebulge; (d) the development of individual cyclothems, showing a general increase in both grainsize and number of sandstone and conglomerate beds towards the forebulge indicating a source area in the southeast; (e) degree of coalification, which decreases at any one stratigraphic level from west to east and southeast, attributable to a combination of shallower burial and lower heatflow in the stable area of the forebulge; and (f) intensity of fold/fault structure with a generally asymmetric pattern developed, verging to the east. 4. Vertical changes in the sediment fill of the foreland basin reflect a diminishing tectonic source from the Variscan hinterland in the west. These are documented as: (a) diminution in the number of marine occurrences upwards, indicating a decreasing subsidence rate with time; (b) common interruptions of marine conditions by coal-rich parallic molasse suggesting intermittently changing subsidence rates and tectonic activity in the source areas; (c) decrease in total thickness of volcanogenic ash, suggesting a waning volcanic activity in the mountain belt from Namurian A times; and (d) changing geochemistry of claystones, indicating a gradual lowering of relief in the source area.
Conclusions
References
1. The Visran Upper Carboniferous coalbearing molasse of the Upper Silesian Coal Basin represents the latest and most eastward development of a Variscan foreland basin, formed as a result of oblique continental collision in the west. 2. The molasse-filled foreland basin is divided into a western foredeep with up to 4.5km of sediment and an eastern platform (forebulge) with less than 200m of sediment. Decreasing thickness from foredeep to forebulge is recorded for each of the stratigraphic units but younger units show less marked thickness changes. 3. Contrasts between foredeep and forebulge are documented in: (a) coal accumulationboth the number of coal seams and the total coal thickness diminish towards the east and south
ADAMUSOVA, M., DOPITA, M., FOLDYNA, J., KALENDOVA, J., KUMPERA, O. (~ STRAKOS, Z. 1992. The isopachous and coalification maps of coal-bearing molasses in the Czechoslovak part of the Upper Silesian black coal basin. Sbor. Vdd. Praci Vys. Sk. Bdfi, Ostrava, 28, 1~. HG. (1), 27-38. CrtLUPA~, I. & VRANA, S. (eds) 1994. Regional Geological Subdivision of the Bohemian Massif on the Territory of the Czech Republic. Journal of the Czech Geological Society, 39, 127-144. DOPITA, M. & HAVLENA, V. 1980. Geology and mining in the Ostrava-Karvin6 Coalfield. OKD, Ostrava. -& KRALiK, J. 1977. Coal tonsteins in OstravaKarvin6 Coal Basin. OKD, Ostrava, 1-213. - - & KUMPERA,O. 1993a. Geology of the OstravaKarvinfi coalfield, Upper Silesian Basin, Czech Republic, and its influence on mining. International Journal of Coal Geology, 23, 291-321.
Professor Kumpera died before corrections to the manuscript had been completed. Final corrections were made by the editors
N A M U R I A N P A R A L I C MOLASSE IN B O H E M I A N MASSIF & - - 1 9 9 3 b . Contribution to the Paleogeography of Namurian A in the Bohemian Massif (Czech). Sbor. v6d. Praci Vys. Sk. bfin, Ostrava, 39, R. HG, 1104, 41-51. Dvo0akK, J. 1973. Synsedimentary tectonics of the Palaeozoic of the Drahany Upland (Sudeticum, Moravia, Czechoslovakia). Tectonophysics, 17, 359-391. --, HONEK, J., PESEK, J. & VALTEROVA, P. 1997. Deep borehole evidence for a southward extension of the Early Namurian deposits near Nemcicky, S. Moravia, Czech Republic: implication for rapid coalifaction. This volume. FIALOVA, V., POLICK~(, J. & HONI~K, J. 1978. Petrography and lithology of the Jaklovec Member (in Czech.). Ostrava, Sbor. GPO, 16, 5-38. FOLDYNA, J. & KUMPERA, O. 1991. Tectonic zones and areas in Subvariscan Zone of the Bohemian Massif (Upper Silesian Basin). Acta Univers. Carol. Kettner, Praha, 3-4, 165-281. FRITZ, H., DALLMEYER, R. D., NEUBAUER, F. & URBAN, M. 1993. Thick-skinned versus thinskinned thrusting: Mechanism for the formation of inverted metamorphic section in the SE Bohemian Massif. Journal of the Czech Geological Society 38, 33-34. GRYGAR, R. 1992. Kinematics of Lugosilesian orocline accretion wedge in relation to the Brunovistulian foreland. Sbor. Vdd. Praci Vys. Sk. bdd, Ostrava, I~. HG, 38, 1, 49-72. HAVLENA, V. 1982. The Namurian deposits of the Upper Silesian Coal Basin. Rozpr. Cs. Akad. V(d., R. mat. p(ir. vdd, 92, 7, 1-79. HLADIL, J. 1988. Zonality in the Devonian carbonate sediments in Moravia (CSFR). Proc. 1st Int. Conf. Bohemian Massif, Praha, 121-126. JANSA, L. F. 1967. Sedimentological evolution of Carboniferous strata in southern part of Upper Silesian Coal Basin. PhD Thesis, Charles University, Praha, MS, (in Czech). KLEIN, G. DE V. 1987. Current aspects of basin analysis. Sedimentary Geology, 50, 95-118. KOTAS, A. 1985. Structural Evolution of the Upper Silesian Coal Basin (Poland). C. R. 10. Congr. Int. Strat. Geol. Carb., Madrid, 3, 459-469. KUMPERA, O. 1983. Lower Carboniferous geology of Jesenlky Block (in Czech). Knih Ust[. Ust. geol. 59, Praha.
--,
27
1988. Brunovistulicum in Variscan development (in Czech). Acta Univ. Carolinae, Geol. Praha, 401-410. - - & FOLDYNA,J. 1992. Development of MoravianSilesian Paleozoic Basin. Sbor. Vdd. pracl, Vys. Sk. b6~. Ostrava, I~. HG, 38. -& MARTINEC, P. 1993. V~voj sedimentfi karbonsk6ho akre6niho klinu moravskoslezsk6 pfinve. Sbor. 1. desko-polskO konf. o sedimentol. karbonu, UG Ak. Vfid CR, Ostrava, 125-166. & 1994. The development of the Carboniferous accretionary wedge in the Moravian-Silesian Paleozoic Basin. Journal of the Czech Geological Society, 39, 1, 63-64. POLICK~/, J. & HONI~K,P. 1984. Produktivni karbon ve vrtu Jablfinka 1. Gas. Mineral. Geol., Praha, 29, 4, 445. PURKYlqOVA, M. 1978. Fl6ra svrchniho karbonu (namuru A) v paleozoiku JV svahfi Cesk6ho masivu u N6m6i6ek na ji~ni Morav~. Gas. Slez. Muzea, Opava, A 27. RAMSHOTTOM, W. H. C. 1979. Rates of transgression and regression in the Carboniferous of NW Europe. Journal of the Geological Society, London, 136, 147-153. ROTH, Z. 1979. The Krfisn~i 1 borehole in the central part of the Moravskoslezske6 Beskydy Mountains. Vdst. (Jstf . (lst. geol. Praha, 55, 2, 75-83. ]~EHOI~, F. & I~EHOI~VA, M. 1972. Makrofauna uhlonosn~ho karbonu deskoslovenskd (6sti hornoslezsk~ phnve, Ostrava, Profil. SKO~EK, V. 1975. Regional and geological interpretation of organic matter coalification in the late Palaeozoic sediments of the Bohemian Massif. Vdst. (Jst(. Ust. geol. Praha, 51, l, 13-25. 1991. Indications of the Late Carboniferous eustatic and climatic oscillations in the Upper Silesian Basin. Vdst. (lst(. tQst. geol. Praha, 66, 2, 85-96. TURNAU, E. 1962. The Age of Coal Fragments from the Cretaceous Deposits in the Outer Carpathians, Determined on Microspores. Bull. Acad. Polonaise Sci. Krak6w, gOol.-g~ogr., 10, 2, 85-89. 1970. Mikroflora i paleogeografia karbonu produktywnego v polskiej czesci Karpat. Biul. Ins. geol., Warszawa, 235, 13, 163-229. ZEMAN, J. 1977. Progn6za roz~i~eni uhlonosnOho karbonu pod vn6jgim flygem Z/tpadnich Karpat. Geol. Prdzk., Praha, 19, 12, 353-357.
The origin of magnetic remanence components of Westphalian C to Stephanian C sediments, West Bohemia: a record of waning Variscan tectonism M I R O S L A V K R S ~, JIt~I P E S E K 2, P E T R P R U N E R ~, V L A D I M [ R
SKO(~EK 2
& JANA SLEPICKOVA 2
1 Geological Institute, Academy of Sciences of the Czech Republic, Rozvojov6 135, 165 O0 Prague 6, Czech Republic 2 Faculty o f Sciences, Charles University, Albertov 6, 128 43 Prague 2, Czech Republic Abstract. Petromagnetic and magnetomineralogical investigations of Westphalian C to
Stephanian C rocks from the Central and Western Bohemian Late Palaeozoic Basins were undertaken to determine the origin of magnetic remanence components and of ferrimagnetic minerals - carriers of respective remanence components in a variety of rock types including tufts, tuffites, siltstones, sandstones, grey and red claystones. Multi-component analysis allowed the separation of remanence components and a selection of petromagnetic and/or magnetomineralogical methods were used to determine the ferrimagnetic minerals. Magnetite was found to be the principal carrier of Variscan remanence components in most non-red-coloured rocks and haematite in red claystones originated during diagenesis in the Carboniferous period. In several samples, haematite, goethite and other Fe-oxides were found to result from recent weathering. In some samples, the Variscan remanence components were separated at relatively low temperatures, from about 150 ~ onwards. Variscan virtual pole positions have been derived from relatively small sets of samples. Nevertheless, they show that the rocks of the Westphalian C and D ages were more intensively deformed than those of the Stephanian, agreeing with an overall decrease in intensity of deformation in the final stages of the Variscan orogeny. This paper aims to establish a geologicalhistorical succession for the generation of the remanence components in Carboniferous rocks of the Late Palaeozoic Basins in Central and Western Bohemia and to determine the minerals - carriers of their respective remanence
components. Distinguishing epigenetic remanence components from those of a syngenetic origin on the basis of a multi-component analysis is possible only where these components mutually differ in direction. A Variscan overprint took place in the Bohemian Massif in the Latest
Table 1. Units of Central and Western Bohemian Late Paleozoic Basins Age
Formation C
Member
Lin6 Otruby Slan~,
B
Stephanian
Malesice Jelenice
Carboniferous
,
A?
T~nec
Cantabrian o
9
e
e
.
~
9
o
t
9
9
.
e
n
9
o
e
9
9
,,
D
.
9
9
9
o
N2;'~any
:
Kladno
Westphalian C (1) Double line indicates gap in sedimentation
From Gayer, R. & Pegek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 29-47.
Radnice
,
30
M. K R S E T AL.
k~
0
<
<
-H
-H
-H
k~
0or
-H-H c--I
rm
oo
c~
-H~
-H
-H
O ~
~'h r
r r
r
Cq
-H-H C',l c'-I
r
0
I-~
oo o
o~
ko
o~
o~
o~
r162
,.el . o. o
-8
.,..~ o
0
O0
0
0
0
MAGNETIC
REMANENCE
COMPONENTS
r~
op
~.~
o
o
~
~.~
.<
b E
< 0p.
~
t"-I
0
-H t~
-H
m
-H
-H
m
~
t~
t~
C'-I ~r
t"r
O0 n~ 0
<<
<<
~
O 0
n9 9
~ ,
.o
E
.o
~
co~
~-~
~,.tn
o~
,~
~ o~
o
o~
o~
o
oo
o~
31
32
M. KRS E T AL.
~"'~
' , , f ~ . ~,,, __
CZECH REPUBLC I /
L. \
J ~. ,.,.?'
RB~
~v
/r TLUSTICE
.
~HoR M R IOAO~
LeR
MR i
I
N
~777<-qB~
~Skm
Fig. 1. A sketch map of the Western Bohemian Carboniferous and selected Carboniferous Relics in the broader region of the Western Bohemian and Central Bohemian Basins, incl. localities of collected oriented samples. 1, Lin6 Formation; 2, Slan~, Fm; 3, T~,nec Fm; 4, TS,nec and Kladno Fms (undifferentiated); 5, Kladno Fm (N~any and Radnice Members); 6, Late Proterozoic and Early Palaeozoic rocks (crystalline rocks and sediments); 7, fructure line; 8, uncertain boundary between lithostratigraphic units. Basins: MB, Man6tin; ZB, Zihle; PB, Plzefi; RB, Radnice. Carboniferous Relics: LeR, Letkov; MiR, Mirogov; HoR, Ho~ovice; ZeR, Zebr~k.
Carboniferous to the Earliest Permian, which caused directions of the overprint components to be close or even identical with the palaeomagnetic directions of the analyzed Carboniferous rocks. Study of the origin of the remanence components therefore required the use of other methods. A combination ofpalaeomagnetic, petromagnetic and magnetomineralogical methods appeared to be the most suitable one. The different values of unblocking temperatures and the varying spectra of micro-coercive forces reflect the variable mineral composition and grain size of the ferrimagnetics. Thermoremanent magnetization in volcanic rocks and the detrital magnetization carried by finely disseminated magnetite grains in sandstones, claystones and siltstones are mostly syngenetic. Under the term of syngenesis we include magnetization related to cool-
ing or deposition of the rocks; all other components generated later (chemoremanent, viscous magnetization) are epigenetic. The processes of syngenesis and epigenesis in redbeds is not precisely dated, since these rocks, besides the detrital (syngenetic) magnetization, show a larger proportion of the chemoremanent magnetization which most probably is due to a gradual dehydration of unstable Fe-oxides into the final stable haematite phase (Krs & Pruner 1995). Even haematite is frequently remagnetized, most probably because of the circulation of water with dissociated oxygen, or it may be generated by an intensive weathering of magnetite or of some Fe-sulphides (Krs 1967). Biostratigraphically well dated rocks of Westphalian C D and Stephanian A B and C ages were chosen for our study. The previously derived and
MAGNETIC REMANENCE COMPONENTS Mt/Mn -3 ~---~-Jn = 112 x l 0 Aim
No. 7199A3
~.+.+~_~,,
1]
~
"t
0.16 x 103AIm
[
,
e...XY
,
,
,
, ,~-
200 N....~400 500~ f _ ~ , , , , , "-4~t
O...XZ I -2-
33
o NSI 3 0 0 . / 100 C'----TR-
-
~,60
.oo oom , /
._:
Iooooo 9 }o~ -,,~ ///-
w4,,,,,,-/_+--,,,, \ ~oo~-:
,~o~
~oo~
~oX
/
/12o
:
W 9 ' -4 1unit=79.Tx10 Aim
Up
Down
210 ~
No.7199A1
"/
~Mts/~t 200
400
~'t/~n
No.7201A2
\ 500~ ~t
~t ~n--729x165ts,l I
200
I
400
I
150 S
0,1, 200
21
"-t" t
~
/
I
500~
400
600~ t
|
200
400
I
500~ ----t
Fig. 2. Bil~ Hora near Plzefi. Westphalian C, tuffite. The upper half of the figure: results of thermal demagnetization of the natural sample. The uppermost plot shows normalized values of Mt/Mn in relation to temperature t~ Mt is the remanent magnetic moment of the sample demagnetized at temperature t~ Mn is the remanent magnetic moment of the sample in natural state. Beneath the plot of Mt/M, in relation to t~ Zijderveld's diagram and stereographic projection of remanent magnetization vectors are given for the sample in natural state (NS) and after progressive thermal demagnetization at temperature t~ The lower half of the figure presents results of thermal demagnetization of the samples subjected to saturation magnetization (prior to thermal treatment). The plots show normalized values of Mt,s/Ms and ~t/Jt~ in relation to temperature t~ Mt,s is the remanent magnetic moment of the sample in saturated state and thermally demagnetized at temperature t~ Ms is the remanent magnetic mament with saturated magnetization Js under the room temperature; ~ t is the volume magnetic susceptibility of the sample demagnetized at temperature t~ ~n is the volume magnetic susceptibility of the sample at room temperature.
statistically averaged palaeomagnetic directions within the Westphalian-Stephanian range (Krs 1968) were thus verified by refined techniques. Another objective was to verify and to date more accurately the genesis of the palaeotectonic rotations that had already been identified in the Bohemian Massif early in the palaeomagnetic research of the region (Birkenmajer, Krs & Nairn 1968).
Outline of geology of the areas investigated The Late Palaeozoic Basins of the Central and Western Bohemia consist of terrigenous ('limnic') deposits. These form an almost uninterrupted area of about 3500km 2. The Central and Western Bohemian Basins are parts of the Bohemian Massif megahorst. Their origin was connected with tectonic
34
M. KRS E T AL. Mt I'Mnj~Jn =253 x 1(] 6 Aim 0 5-]
~-h~
"[
o...XZ
11.9 x I0-6A Im
~"+-+'+~'; '
o;oo .h ~
9~
-
I ~ ), ) I ~'o---540oC Down "~,.,...500 oc
t/"
--q-,
_
oo./ l
:-
[ I
\
-
200 oc 0J--~-24 \ 150~ 250~
MtslMs
No. 7 1 9 0 A 2
0.4
\
I
,ooo
1unit=31.Sx1()6A/m
--,,-)
50
120
2 1 0 ~ S
150
No. 7 1 9 6 A 3
Mr,siMs
'~
200 400 600~ Mt/Mn "-'~t 2" I / ~ n = 12 x10-6 [SI] I
I
200
I
!
400
I
I
600 ~ "--"t
200 400 600~ ~ft/;Fn "-" t 24 ~fn=7.5x 1C~6[ SI ] I
I
200
1
I
400
l
I
600 ~ --" t
Fig. 3, Tlustice near Zebrfik. Westphalian C, tuff, tuffite. See caption to Fig. 2.
activity during the Variscan orogeny, reflected in a progressive but uneven subsidence of the Variscan intramontane area in which most of the limnic basins of Western and Central Europe were formed. Initial sedimentation in the Central and Western Bohemian Basins began in the Westphalian C and with some breaks in deposition, probably reflecting the Variscan orogeny, along with several diastems, sediment accumulation continued at least up to the Stephanian C or, more likely, Autunian time. However, Autunian clastics have nowhere been found preserved in the Central and Western Bohemian Basins. The stratigraphic sequence is divided into four formations, some of which are subdivided into members and horizons (Table 1). The first two, Kladno and T~,nec Formations, consist
dominantly of fluviatile deposits, while the others are chiefly lacustrine. Mudstones and claystones of the Kladno and Slan~r Formations are mostly grey, with a few significant coal seams, whereas those of the two remaining formations are mostly red. The Carboniferous deposits in general dip primarily at only a few degrees. However, the dip may have been steepened by uneven subsidenced blocks. The strata of some basins show definite, although general, dips. Large outcrops in the Plzefi Basin consist of deposits dipping mainly to the E to SE, whereas in the southern part of the Central Bohemian Basin the deposits dip generally to the N. Reconstruction of the faults and fault systems in the Central and Western Bohemian Basins poses serious problems. Here, with the exception of the northern parts of the Central
MAGNETIC REMANENCE COMPONENTS
35
1.43 1.29. Mt/Mn/1.52 1.31 5.81]
N,/ s -{ jn=498•
45ooq
/
400% ~ NS--~)
e...XY
.~.~.M~'
n .
k-*
1o3 x 16 Aim ----'
|
Up
\ 6
~ ...... 2 0, 0....j . _,~, J ~ 0 ,0
," 600~ ,
3
; o
400oC ",>Q~--._.100o C~r " 1 unit=239x1(55Alm S
No. 7 2 5 0 A 3
200
400
500oc
..... 400
150
No. 7 2 5 6 A 4
--~t
200
210 ~ S
200
,
,~n.70•
5OO~
200
~t
400
~ o ~
600o(2 -',-t
5 s r] 400
600% --"- t
Fig. 4. Bilfi Hora near Plzefi. Westphalian C, laminated sandstone, reddish claystone. See caption to Fig. 2.
Bohemian Basins, subsidence faults predominate and generally strike N W - S E to N N E SSW and E-W. In the Ohe River district the Carboniferous is cut by NE to SW trending faults parallel to the Litom6f'ice Deep Fault. Furrow-like depressions owe their origin to recurrent movements that took place along N N E - S S W trending basement faults. All basins exhibit, besides disjunctive tectonic structures, conjuctive phenomena. A close relationship exists between these two types of tectonic factors. Some faults were produced by displacement along one plane, while others form zones up to several hundred metres wide. The faults trend in a curved pattern; the throw and dip vary. The fault throw ranges from several cm to several hundreds of metres. Most frequently movements along the faults are of
the order of l0 -~ to 102m. Fault planes are estimated to dip on average 65 ~ to 70 ~, but very low angle faults dipping about 10~ to 25 ~ are also present together with steep to vertical faults. Faults often virgate and vicariate, the isolated fractures becoming Y-shaped. A number of faults were active in both synsedimentary and postsedimentary conditions, while others formed after the termination of deposition.
Laboratory procedures Laboratory procedures were combined in a way that enabled the derivation of both the palaeomagnetic directions and the determination of the m i n e r a l s - carriers of the respective
36
M. KRS E T AL.
Mtl~..~J n=7719x 165Aim ' 1-"~H"N--~| 4, , ~ 0.5t 56.5x 166Aim
No.7244A2
o...xz
N I
.T /
'260'
e...XY 200oc
~NS
3oooc..~c~ 3o~
60o 8oooc "-"" t
/
_
:
/
/
7oooc
//o~
~
~
"k.~o
L~s-52ooc/
w-I'""
' " : + ! ' " " " ~E -I /
U/5zooc E \ W___~F_! ; : I I I '. : \ -4 / Up ~ B50~ Down 2 4 0 ~ 4 ~ 68/0~ ">-~\ ',LZ...-'C l~ j680~ S lunit=Tg6xl()6AIm 210 ~ " 150 No. 7240 A1 M t . l ~ _ j s = 1091x,(]4Aim
i
200
1
i
1
i
I
I
I
I
No. 7244A1 Nt'ls]~-~ " Js=17432x ,(]4Alrn
--I-
400 600~ 200 400 600~ ~r _ ~m.m~o --4- t --4. t ~'t I~,n 2 t ~ 2 gtn=8X1()6 [ S,] I (~6[SI l ltt~m-11"~4~~ 200 400 600 oC 200 400 600 ~ I
1
---'-- t
i
i
1
I
I
1
--"
t
Fig. 5. Mirogov, 'Lomy na JanovE. Westphalian D, grey siltstone. See caption to Fig. 2.
remanence components. This method provides data to be used for the reconstruction of the history of the origin of the magnetization components. The following approach was adopted: Hand samples were collected in the field from the localities mentioned in Table 2 and at locations shown in Fig. 1. Laboratory specimens in the form of small cubes were prepared from the hand samples to be measured on spinner magnetometers JR-4 and JR-5 (Jelinek 1966). The hand samples are designated by both numbers and by the letter A, e.g. 7190A. Laboratory specimens are designated by indexes 1, 2, 3, etc., for example 7190A1, 7190A2, 7190A3, etc. Laboratory rock specimens in their natural state were subjected to progressive thermal demagnetization by using the MAVACS equip-
ment (P[ihoda et al. 1989) securing generation of a high magnetic vacuum in a medium of thermally demagnetized specimens. The remanent magnetization of specimens in their natural state is identified by the symbol Jn, the corresponding remanent magnetic moment by the symbol Mn. The remanent magnetic moment of the rock specimen demagnetized at temperature t~ is denoted by Mt. Graphs of normalized values of M t / M n = ~ ( t ) were constructed for each analyzed specimen and they provided primary information on the unblocking temperatures of the minerals - carriers of remanent magnetization. The directions of Jn and those of the remanent magnetization of the thermally demagnetized specimens in the course of a progressive thermal demagnetization are shown in stereographic projection. The orthogonal projection of the
MAGNETIC REMANENCE COMPONENTS
100[ % ]
~
9ao ~
Roudn6 near Plze6
37
o,-' * ' ~ * ~
0/o"
271100[pT1
No. 7217A3
l
l
1
2
I
' [
| I
-'
I
4 6 810
"nil
20
tIu|'l
4060
'
I
II
Ill'l
n
100 200 4(33 1000[rnT] 80 600 800
fi___.
lOO-%] 90" 80I 70"
Pit between Ledce end Zitov
~'13413[nT]
/
60~" SO"
No.7335A5 ~IRM/t~H
-~ 4o30" 20" 10I
I
1
2
, , ~
'
,,
4 6 810
'
I
20
''I
'I'I
I
'
I
'I
I
'l'l'
40 60 100 200 4.00 lO00[mT] 80 600 800
Fig. 6. Isothermal remanent magnetization (IRM) in dependence on direct magnetic field (H). No. 7217A3: a sample of grey claystone from the Roudn/t locality near Plzefi; No. 7335A5: a sample of red claystone from a pit between Ledce and Zilov.
remanent magnetization vectors is shown by the Zijderveld's diagram, where a full circle indicates projection onto a horizontal plane (XY) and a blank circle indicates projection onto a north-south vertical plane (XZ). The natural state of the specimens is designated by NS. Phase or mineralogical changes of magnetically active (mostly ferrimagnetic) minerals frequently occur during the laboratory thermal tests. These changes can be clearly derived from the graphs of the normalized values of ~,/~t =f(t), where Jg, designates the volume magnetic susceptibility of specimens in the natural state and ~ t the susceptibility of samples demagnetized at temperature t~ The 9 and ~ , values were measured on a kappabridge KLY-2 (Jelinek 1973). Because of an excessive number of data, these graphs are not shown in the present paper, only typical examples of pilot specimens are presented. In order to determine the unblocking temperatures of minerals in low magnetic rocks (with a low content of ferrimagnetic minerals)
with appropriate accuracy, pilot samples were selected for the respective localities. They were subjected to isothermal progressive magnetization by a direct magnetic field up to the saturation state with the use of a direct field with a maximum intensity of 1000mT (10 000 Oe). Dependence of the isothermal remanent magnetization (IRM) on the direct magnetic field (H) was tested for 28 specimens. Again, owing to the large amount of data these graphs are not shown here. However, this information will be used for data interpretation, see below. The specimens with the saturated remanent magnetization Js and with the corresponding remanent saturation moment Ms, were subjected to a progressive thermal demagnetization by using the MAVACS equipment. The plots of the normalized values Mr,s~ Ms in relation to the temperature t~ of the demagnetization for the pilot specimens are shown in Figs 2 to 5 and in Figs 7 to 12. The symbol Mt,s indicates the moment of a specimen that had a remanent magnetic saturation
M. KRS ET AL.
38
H
~
Mr/ ?n+~---~k J n :555 x 10 5A/m
,\
No. 7222 A3 E
N
W
I
I
I
/
~5.5x10
Aim
I
'Oo n
540oC
L
,
,
200
;'--s
N
400
600~
-----300oC
V
S
Q,..XY 210 ~
Mt,s/MsNO. 7217A3_L
150
S
1unit = 79.6 xlOSA/m--
~,,t,s/MsNO.7219 A1-4
/--
/ \
o
2 0 0 ~ 400
t
~t/~n 1
500~
--~'t
x16 s [,Sll i
200
i
400
200
~t/~n
400
600%
----~ t
1 :.~ i
600~ -'~t
i
i
200
i
i
400
i
i
600~ --~t
Fig. 7. Roudnfi near Plzefi. Westphalian D, grey claystone. See caption to Fig. 2.
moment of Ms in the initial state and was subsequently demagnetized at temperature t~ Separation of the remanent magnetization components was carried out by using the multi-component analysis of Kirschvink (1980). The statistics of Fisher (1953) were used for both the derivation of mean palaeomagnetic directions from the data of progressive thermal demagnetization for selected sets of rocks with suitable physical properties and for calculation of mean directions of the pertinent remanence components derived by the multi-component analysis.
Results of laboratory measurements The heterogeneous petrographic rock types selected for laboratory treatment represent a natural material with a wide spectrum of
magnetic properties, with varied geological history, origin of the remanence components, and with variable magnetically active minerals. This paper describes typical examples of the measurement results.
Bil5 Hora near Plzefi, 7198A-7201A, tuffite, Westphalian C Various Fe-oxides with markedly unstable properties were identified. Specimen No. 7199A3 shows a single-component remanence with haematite generated by recent weathering as its carrier. This specimen with recently generated haematite shows a narrow spectrum of microcoercive forces, the saturation state was reached in a high field of 900 mT in intensity. Another investigated specimen No. 7201A2 with various Fe-oxides shows a wide spectrum of micro-
MAGNETIC REMANENCE COMPONENTS Mt/Mn
.+"+~.
I \ Jn=84 * I(]4Alm~.
NO. 7 2 0 6 A1
0.54
I
12"8x1()4A/m~--~ I
o...xz
N
I
200
!
!
I
|
N 400
600~
e...XY 200~ 300~ \,.,
--
660~
.
~'+-+._
1-.~-+/
t
39
o
".1~ 450~ S ,, 585oC ~ -~-,,.-, -o'--\ 100~
/ft /
: 71
"~-%oooc , ooc
: : :_: : : : : Ej
Up lunit=Zg6xl(~4Ai m
Down
MLslM s No. 7 2 0 2 A 2
I
I
2 ~/~n 200/
J
I
! _ 200
400
210 " ' - - ' r " ' S
500~ "-~'t
150
MI,slMs N o . 7 2 0 6 A 2
I C 215-~ ~
400 "-~'t 600~
1
J
~
it/~r
I
-
200 400.,, 600~ /L~" --~'t
/ 200
400
600~ "-~'t
Fig. 8. Rad6ice near Plzefi. The Westphalian D/Earliest Stephanian, grey claystone. See caption to Fig. 2.
coercive forces, the saturation state was reached at 200mT. Samples of this group did not preserve their Variscan magnetic directions owing to intensive weathering. Some samples also revealed a conspicuous instability during their thermal treatment (Fig. 2).
Tlustice near Zebrdk, 7190A-7197A, tuff- tuffite, Westphalian C Two-component magnetization predominates. Goethite is the carrier of the remanent magnetization with the direction of the recent geomagnetic field (influence of weathering). The carrier of the characteristic magnetization is a minor magnetite proportion. The Variscan remanence component is well separable at an interval of 250~176 In four samples the increase in
IRM on a magnetizing field H was investigated and in three samples a saturation state was reached in a direct magnetic field of 700mT, which is characteristic of magnetically hard samples. Unblocking temperatures for goethite _<100~ and for magnetite below 580~ can be derived from the graphs of Mt, s/Ms plotted against temperature (Fig. 3).
Bild Hora near Plze~, 7247A-7259A, laminated sandstone, reddish claystone, Westphalian C Samples show physically unstable properties. The carriers of the magnetization are the Fe-oxides with unblocking temperatures mostly below 500~ generated by weathering. Only two specimens (No. 7250A1- laminated sandstone and
M. KRS ET AL.
40
MilMn
No. 7215AI o...xz
~..
i...XY
/
I l~/"\+'-,~_.
T *-^o^
9
+
~4u u
E
up
" ] J.=684x16 A/m.w.+..~ / \ ,J, / . . . . . 200 N 400 600~ ~..~.-J--.~-.30
Oo n 3 0 0 /
-/--/
/
"~'
/
:
~
24o1\
{ - ,oooo , '
I'!__3oo~
/~2o
200oC S
No.7211A1
No.7215A3
11,~
Js=2651xl0-4AIm
|
i
,
200
!
!
|---
SO0~
400
"gt I~n --D t 2 t 0tn=lggxlOB[ Sll I t -i:
' 't E
-
1 unit =7g.6x lOBA/m Mt,slMs
.60
/-WT.'__~_'_'.]~I,~';"''''
f%.
,
-'~t
-
--
-
-
-~. ---'~,i
l
200
~ ' 4
0
-'i
,
B00~
"-~t
200
~ftlafn
400
B00~ ~t
1 |
|
200
i
i
400
i
i
B00~ ---"t
Fig. 9. Rad6ice near Plzefi. Stephanian A, green siltstone. See caption to Fig. 2.
No. 7256A2 - red claystone) yielded palaeomagnetic directions. Phase changes, derivable from the graphs of Mt, s/Ms plotted against temperature, point to a low reliability of derivation of the palaeomagnetic directions (Fig. 4).
MiroEov, 'Lomy na Janovd' quarries, 7236A-7246A, grey siltstone, Westphalian D The samples were mostly unsuitable for palaeomagnetic analysis, only in three of them were Variscan palaeomagnetic directions within temperature interval of 100~176 (300~ obtained. A low proportion of magnetite has been proved, e.g. in specimen No. 7240A1. The intensively weathered samples showed haematite, with the direction of remanent magnetization close to the present-day geomagnetic field.
In specimen No. 7244A2, a Variscan direction of remanence at high demagnetization temperatures of 650~ and 680~ is visible (Fig. 5).
Roudn6 near Plzefi, 7217A-7223A, grey claystone, Westphalian D Minerals with a wide spectrum of unblocking temperatures mostly below 400~ and of microcoercive forces are the carriers of the magnetization. Figure 6 examplifies development of isothermal remanent magnetization in dependence on direct magnetic field for samples from the Roudnfi locality near Plzefi (No. 7217A3) and from a red claystone pit between Ledce and Zilov (No. 7335A5). A broad spectrum of micro-coercive forces for the sample from Roudnfi is evident. Haematite has not been
MAGNETIC REMANENCE COMPONENTS
41
M t l M n / J n = 8 6 0 x l(~6AIm
No. 72 24 A3 o...XZ Q XY "'" w
,:
up
1]+'--+'+\4..~ 0"5t /
N t
t :[.45ooc
E
,,:%~,,
t:: : : / ~ 3oo~ 0ow. s-
\ 33.4x 1(~6A/m ~4---:-lJ , , ,,, ' ' 200 '~ 400 ^~' _ 600~ ~ ~ ~ . 3 u ~ /'I "% " 300/ \6o
/
:
W
illlil
50oC
IV
I_a..---~176
No.7227A2
1 ] ~ J s :4171x164Alrn
I
200
i
/
400
I
E
~,~>%._Z_.--~,=~ =,u / ~S 200~
Mt.s/M s
200 400 600~ ~'t/~n ---"t 2t ~'n=151x166 Is1] I
i
S
S 150oC 1unit =7g.6 x 106Aim
Mr,sIMs
\
i
I
No.7228A2 Js :4768x1154A/rn
200 400 600~ ~t/a'n -'-'- t 2t ~n=18xld6[sl] i
600~ " "
I~
,-,~
i
'1
200
t
I
400
i
!
600~ "-'-t
Fig. 10. Dolni Vtk~. Stephanian B, siltstone. See caption to Fig. 2.
proved, the proportion of magnetite is low. Ferrimagnetic minerals are represented by Fe-oxides, by goethite and probably by one of the "r-Fe203 or rl-Fe203. The Variscan remanence component is separable at a relatively low temperature, within the interval of (100) 150~176 (Fig. 7).
Raddice near PlzetL 7202A-7207A, grey claystone, Westphalian D/the Earliest Stephanian Multi-component remanence is caused by syngenetic magnetite and by secondary Feoxides, ranging from goethite to haematite, formed by recent weathering (specimen No 7206A1). Magnetite is the carrier of the
Variscan remanence component, being separable within a temperature range 250~176 Samples containing exclusively haematite, generated by recent weathering, show a narrower spectrum of microcoercive forces and they reach the saturation state in high intensity magnetic fields (Fig. 8).
Rad~ice near Plzeti, 7208A-7216A, green siltstone, Stephanian A No haematite was detected in this sample group. Otherwise, magnetic properties similar to samples from the previous locality have been proved. Magnetite is, again, the carrier of the Variscan remanence component, which is separable in the temperature interval 250~ (Fig. 9).
42
M. K R S E T AL.
Dolni Vlkf~Y, 7224A-7231A, siltstone, Stephanian B Mostly minerals with a low unblocking temperature below 300~ are the remanence carriers. The saturation state was reached at a field intensity of 350 mT; these are therefore mediummagnetic-hardness minerals. Graphs of Mt,s/Ms plotted against temperature show a higher proportion of minerals with a lower unblocking temperature, but even a lower content of magnetite has been proved with unblocking temperature below 580~ The Variscan remanence component is already separable at relatively low temperatures, within an interval of 150~ to 300~ (Fig. 10).
Zihle, a red claystone pit, 7260A-7272A, red claystone, Stephanian C The remanence consists of two components, a smaller proportion represented by viscous magnetization. The carrier of the principal remanence component is an extraordinarily stable
haematite with unblocking temperature below 680~ Three sites were sampled at this locality. The group of samples 7263A-7272A shows palaeomagnetic directions that correspond well to those from the clay pit between the villages of Ledce and Zilov (samples No. 7232A-7235A), see below. The high stability of palaeomagnetic remanence is also well documented by the results of progressive thermal demagnetization of the specimens (Table 3). The structure of Zijderveld's diagrams, graphs of Mt/Mn and Mt,s/Ms plotted against temperature, and the extraordinary stability of magnetic susceptibility vs. the thermal fields from the Zihle locality (Fig. 11) entirely correspond to the values established for the clay pit locality between the villages of Ledce and Zilov (cf. Fig. 12).
A red claystone pit between the villages of Ledce and Zilov, 7232A-7235A, red claystone, Stephanian C The two-component remanence is due to a small proportion of viscous (recent) component
Table 3. Results of progressive thermal demagnetization using the MAVACS apparatus. A red claystone pit, locality Zihle. Stephanian C, red claystone, Nos of samples 7263A-7267A Temperature (~
20 100 150 200 250 300 350 400 450 500 540 580 620 650 680 700
Mean direction of remanent magnetization D(~
I(~
215.5 206.5 206.4 206.5 206.1 206.2 205.6 205.5 205.3 204.5 204.7 204.6 203.4 208.6 176.9 .
19.9 11.1 9.8 9.0 9.2 8.7 8.6 8.2 8.9 8.7 8.4 8.2 8.3 8.0 - 1.2 .
.
.
a9s (~
k
11.8 10.6 10.4 10.6 10.7 10.6 9.8 9.9 9.6 10.1 10.9 10.7 10.2 12.9 12.5
43.2 52.8 54.6 52.9 51.7 53.5 61.7 60.6 64.2 58.8 50.5 52.1 57.7 35.9 38.2
5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
D, I, declination, inclination of the remanent magnetization after dip correction; a9s, semi-vertical angle of the cone of confidence calculated according to Fisher (1953) at the 950 probability level; k, precision parameter; n, number of the samples analyzed.
MAGNETIC REMANENCE COMPONENTS
Mt/~In
W
I
:
I
650~
0.5x 164Aim
1'
:
Up 700oC~ 680~ /
43
'o'
~
/
58~c4 500~
\
NS ~,, c -~-~_ 400~ [-
300~
b'~200oC 100~ 1 unit =15.9xl(~4A/m /M NO.7263A2
!
|
200
!
i
!
!
400
S
MtdM s NO. 7 2 7 0 A 2
200 400 600~ n =221x10-6[SI] --"t
aft/~'n
21U~_----q'50
'
600~ --"t
200 400 600% ~ft~.n.~_~n=249x166 [sI] --"- t !
!
200
i
i
400
i
i
600~ "----t
Fig. 11. Zihle, a red claystone pit. Stephanian C, red claystone. See caption to Fig. 2.
and to a prevailing Variscan component brought about by syngenetic haematite. The haematite is characterized by a narrow spectrum of micro-coercive forces and by high magnetic hardness (the lower part of Fig. 6). The stability of palaeomagnetic directions is extraordinarily high. Table 4 shows the mean remanence directions using Fisher's (1953) statistics of four samples submitted to progressive thermal demagnetization. The Variscan remanence directions are derivable within an interval of 150~176 and optimum cleaning was achieved at a temperature of 620~ (Fig. 12).
Palaeomagnetic directions and virtual pole positions The objective of our work was not merely to increase the data base of palaeomagnetic
directions, but to determine on several representative rock samples, the genetic history of the remanence components or the overprint components. This approach has, nevertheless, partially verified the results of previous palaeomagnetic measurements (Krs 1968) by using a refined methodology, particularly by introducing the multi-component analysis of remanence and by using the thermal demagnetizaion in a high magnetic vacuum. Table 5 summarizes the derived palaeomagnetic directions and the corresponding virtual pole positions. Higher values of palaeomagnetic declination have been proved at several localities (Rad6ice near Plzefi- grey claystone, Bil~ Hora near Plzefi and also at the red claystone localities between the villages of Ledce and Zilov, Zihle). Similar higher values of palaeomagnetic declination in western Bohemia were reported earlier (Birkenmajer et al. 1968). The data presented here indicate that the higher values
44
M. KRS E T AL. MtlMn
11~"+'~, -4 I\ 'N-,,.~.~.0.8x10 Aim
NO. 7232AI
w
u#
: " =~f' "~0o~o/ 7 ~ 6sooc
" / . . . .
200
o...xz
f
o...XY
/
~
/ ~ , T 585o C
\
150~C S 1unit = 23.9 x 1()4A/rn
~
.
,/'g~
600
800~
:
uu/
hoooc,l
.,
-
/
"k,,~u
:
210 ~
/
150 S
No.7235A5 Mr,s/Ms 1 1 ~ ~ 1 ~ 4 0 2 3 x 1(}4A/m
200
~tlMn
!
i
200
400
!
i
400
600~ ~ t
I
!
600~ --~.f
Fig. 12. Between villages Ledce and 7,ilov, a red claystone pit. Stephanian C, red claystone. See caption to Fig. 2.
for palaeomagnetic declination in the Carboniferous are not constant across the studied area, but differ from one locality to another. On the other hand, the palaeomagnetic directions for Early Permian rocks are homogeneous in the territory of the Bohemian Massif, indicating an extraordinary tectonic stability (consolidation of blocks of the newly forming supercontinent Pangea, Krs & Pruner 1995). Figure 13 shows the virtual pole positions for the studied Westphalian and Stephanian rocks. Although the pole positions have been calculated from a small number of samples, the Westphalian rocks exhibit a higher scatter of palaeomagnetic data than do those of the Stephanian. For rocks of the Westphalian the ~,p = 36.25 ~N; Ap : 164.87 ~ E; Q{95: 15.4~ k=36.4; N = 4 ; for rocks of the Stephanian
qap = 37.46 ~N; Ap--163.76 ~ E; 0 { 9 5 = 6.7~ k = 132.7; N - 5 . The difference in magnitude of the precision parameter k is significant. The mean pole positions calculated in this paper correspond well to those derived on a larger data set for the whole Bohemian Massif, on rocks dated biostratigraphically (Krs & Pruner 1995).
Conclusions Palaeomagnetic, petromagnetic, magnetomineralogic analyses and the multi-component analyses of remanence of rocks of Westphalian C to Stephanian C ages in the Central and Western Bohemian Late Palaeozoic Basins have provided the following information:
MAGNETIC
REMANENCE
COMPONENTS
45
Table 4. Results of progressive thermal demagnetization using the MAVACS apparatus. A red claystone pit, between villages Ledce and Zilov. Stephanian C, red claystone, Nos of samples 7232A-7235A Temperature (~
20 100 150 200 250 300 350 400 450 500 540 580 620 650 680 700
Mean direction of remanent magnetization D(~
I(~
208.8 206.8 206.3 206.1 205.7 205.8 205.5 205.4 204.6 204.9 204.9 205.0 204.3 204.9 177.4 38.2
-9.9 -12.9 - 14.4 - 14.4 - 13.9 - 14.1 -12.5 - 11.8 - 10.0 -10.1 -8.3 -7.1 -5.2 - 5.3 -2.9 -35.5
o~95 (~
k
14.0 14.4 14.5 14.3 14.2 14.3 14.4 14.4 14.0 13.6 13.9 13.3 12.8 13.2 5.2 49.4
44.3 41.6 40.9 42.1 42.7 42.4 41.7 41.9 44.2 46.4 44.8 48.8 52.3 49.2 307.8 4.4
4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
D, I, declination, inclination of the remanent magnetization after dip correction; a95, semi-vertical angle of the cone of confidence calculated according to Fisher (1953) at the 95% probability level; k, precision parameter; n, number of the samples analyzed. Table 5. Virtual pole positions, mean palaeomagnetic directions. Plzet[ Basin, Westphalian C to Stephanian C,
cf Table 2 Site
Locality
Mean palaeomagrnetic directions
Dp (o)
a95 (~
k
n
Virtual pole positions
/p (~
1
Bilfi Hora near Plzefi
unstable sample
2
Tlustice near Zebrfik
201.4
18.6
Ovals of confidence
~p(~
Ap(~
dm (~
dp (~
27.69
169.89
10.93
5.68
3 10.5
34.2
7
3
Bil~i Hora near Plzefi
218.5
-12.8
19.7
162.3
2
36.04
143.52
20.07
10.23
4
Miro~ov 'Lomy na Janov6'
200.8
2.8
10.5
131.8
3
36.27
167.51
10.50
5.25
5
Roudnfi near Plzefi
190.7
-5.7
6.5
87.1
7
42.22
178.90
6.52
3.27
6
Rad6ice near Plzefi
212.2
-10.8
3.6
653.8
4
38.07
150.98
3.65
1.85
7
Rad6ice near Plzefi
196.5
3.0
14.4
15.7
8
36.81
172.56
14.41
7.21
8
Dolni V l k ~
197.4
-0.6
6.8
79.1
7
38.32
170.88
6.80
3.40
9
Zihle, a red claystone pit in operation
205.0
2.3
7.2
45.6
10
34.40
162.57
7.20
3.60
10
A red claystone pit, between Ledce and Zilov
204.3
-5.2
12.8
52.3
4
38.47
161.66
12.84
6.44
Dp, Ip, mean palaeomagnetic declination, inclination; a95, semi-vertical angle of the cone of confidence calculated at the 95% probability level; k, precision parameter; n, number of the samples analyzed; ~p, Ap, palaeolatitude, palaeolongitude of the virtual pole position; dm, dp, ovals of confidence calculated at the 95% probability level.
M. KRS E T AL.
46
Westphalian
Step hanian
Fig. 13. Stereographic projection of virtual pole positions. Virtual pole positions are denoted by small full circles. The mean pole position calculated from virtual pole positions is denoted by a crossed small full circle, it is circumscribed by a circle of confidence calculated according to Fisher (1953) at the 95% probability level. Westphalian: 2, Tlustice near Zebr~k, Westph. C; 3, Bil~ Hora near Plzefi, Westph. C; 4, Miro~ov, 'Lomy na Janov6', Westph. D; 5, Roudnfi near Plzefi, Westph. D. Stephanian: 6, Rad6ice near Plzefi, the Westph. D/Earl. Steph.; 7, Rad6ice near Plzefi, Steph. A; 8, Dolni Vlk~,~, Steph. B; 9, Zihle, a red claystone pit, Steph. C; 10, Between villages Ledce and Zilov, a red claystone pit, Steph. C.
(1) Magnetite is the principal carrier of the Variscan remanence component in most nonred-coloured rocks. In tufts and tuffites the magnetization is of thermoremanent and detrital origin, and it is also of detrital origin in grey, green claystones and siltstones. In these cases, magnetite is syngenetic with the rock, i.e. the direction of its Variscan palaeomagnetic remanence component corresponds to the time of deposition and compaction of the rock. (2) In the red-coloured rocks, haematite is the carrier of palaeomagnetization which clearly originated during the diagenesis of the rock (red claystone). Haematite of this type is physically stable; it has been proved to contain a slightly viscous component and its spectrum of unblocking temperatures is wide. (3) In some rocks other than red, e.g. in the grey siltstone from the Miro~ov locality, in the 'Lomy na Janov6' quarries and in the grey claystone from the Rad6ice near Plzefi locality, haematite has been found to be a product of weathering. Haematite of this type shows a narrow spectrum of unblocking temperatures;
its direction of remanent magnetization corresponds to that of the field of a recent (theoretical, co-axial, geocentric) magnetic dipole. (4) Except for the Bilfi Hora locality near Plzefi (highly weathered tuffite), all rocks of the localities studied (Tables 2 and 5) show a multicomponent remanence and some of them also a Variscan remanence component. (5) In some rocks, the Variscan remanence component was already separable at low temperatures, e.g. in the grey claystone from Roudnfi near Plzefi from (100) 150~ upwards, and in a claystone from Dolni V l k ~ from 150~ In red claystones from the localities of 2;ihle and from the clay pit between the villages of Ledce and Zilov, the Variscan component was separable from (100) 150~ upwards. These data indicate that these localities were not significantly reworked chemically, thermally or by other processes in postVariscan times. It was the recent rock weathering that caused either a partial or a complete obliteration of the Variscan remanence component in some of the rocks.
MAGNETIC REMANENCE COMPONENTS (6) Virtual pole positions have been derived from relatively small sets of data. Nevertheless, the Westphalian rocks seem to have undergone palaeotectonic deformation more intensively than those of the Stephanian. This is in accordance with an overall decrease in palaeotectonic deformation in the final phase of the Variscan orogeny, which finally terminated by Early Permian times (Krs & Pruner 1995). The authors wish to thank Professor R. Gayer and Dr. V. Kropfirek for reviewing the paper and suggestions. They are also grateful to Professor R. Gayer for improvement of the English. The authors would like to thank Mrs. Marta Krsov/t and RNDr. Daniela Venhodovfi for their help in laboratory works. They also acknowledge the financial support through the Grant No. 113/94 of the Charles University in Prague.
References BIRKENMAJER,K., KRS, M. & NAIRN,A. E. M. 1968. A palaeomagnetic study of Upper Carboniferous rocks from the Inner Sudetic Basin and the Bohemian Massif. Bulletin of the American Geological Society, 79,589-608.
47
FISHER, R. 1953. Dispersion on a sphere. Proceedings of the Royal Society, A217, 295-305. JELiNEK, V. 1966. A high sensitivity spinner magnetometer. Studia geophysica et geodaetica, Praha, 10, 58-78. - - 1 9 7 3 . Precision A.C. bridge set for measuring magnetic susceptibility and its anisotropy. Studia geophysica et geodaetica, Praha, 17, 36-48, KIRSCHVINK, J. L. 1980. The least-squares line and plane and the analysis of palaeomagnetic data. Geophysical Journal of the Royal Astronomical Society, 62, 69%718. KRS, M. 1967. Research Note: On the palaeomagnetic stability of Red sediments. Geophysical Journal of the Royal Astronomical Society, 12, 313-317. 1968. Rheological aspects of palaeomagnetism? International Geological Congress Prague, XXXIII Session, 19-28 August 1968, Proceedings, Section 5, 87-96. & PRUNER, P. 1995. Palaeomagnetism and palaeogeography of the Variscan formations of the Bohemian Massif, comparison with other European regions. Journal of the Czech Geological Society, Praha, 40/1-2, 3-46. PI~iHODA, K., KRS, M., PESINA,B. & BLAHA,J. 1989. MAVACS- a new system creating a non-magnetic environment for palaeomagnetic studies. Special Issue Cuadernos de Geologia Ibdrica, Madrid, 12, 223-250.
A depositional and diagenetic model for the Eocene Sulcis coal basin of SW Sardinia ROLAND
DREESEN'*,
DOMINIQUE
B O S S I R O Y ~, R U D Y S W E N N E N ~,
JACQUES
T H O R E Z 3, A U R E L I O
F A D D A 4,
LUCIANO
OTTELLI 5 & EDDY
KEPPENS 5
l Institut Scientifique de Service Public, 200 Rue du Chdra, B-4000 Li@e, Belgium 2 Katholieke Universiteit Leuven, Afdeling Fysico-chemische Geologie, Celestijnenlaan 200 C, B-3001 Leuven, Belgium 3 Universit~ de Lidge, Service de G~ologie GOnOrale, GOologie des Argiles et Sddimentologie des Silicoclastiques, AllOe du 6 Aofft, B18, B-4000 Sart Tilman par Lidge 1 Belgium 4 Carbosulcis, Miniera Monte Sinni, 1-09010 Cortoghiana (Ca), Sardinia, Italy 5 Vrije Universiteit Brussel, Eenheid Geochronologie, Laboratorium voor Stabiele Isotopen Geochemie, Pleinlaan 2, B-1050 Brussels, Belgium * Current address." VITO, Boeretang 200, 2400 Mol, Belgium
Abstract: Detailed sedimentological, mineralogical and petrographical analysis of closely
spaced cored boreholes has enabled the development of a revised depositional model for the early Eocene coal-bearing Produttivo Formation of the Sulcis Basin. The deposition of autochthonous-hypautochthonous palustrine-lacustrine coals and associated carbonates was interrupted episodically by sedimentation of allochthonous lithocalcirudites and lithocalcarenites. The latter clastics display characteristic upward shoaling tidal flat sequences related to marine incursions. This interpretation is in contrast to the previously accepted fluvial origin of the detrital episodes. The coarse basal transgressive lag deposits consist of various carbonate intraclasts, dolosparite, grains consisting of calcite cement and euhedral-subhedral evaporite-bearing quartz grains. Combined stable isotope and cathodoluminiscence analysis has revealed a complex diagenetic history for the clastic deposits. The potential extrabasinal or intrabasinal provenance of the clasts, in particular the origin of the evaporite relicts, is discussed. The subtropical-tropical coastal marshes of the Florida Everglades (USA) are proposed as a possible modern analogue for the subbituminous Sulcis coals.
The Sulcis coal basin in southwest Sardinia (Fig. 1) represents the only subbituminous coal deposit in Italy. Between 1979 and 1992 over two hundred exploration boreholes were drilled from the surface and from underground galleries by Carbosulcis S.p.A. Recently, a number of these cored drillholes have been reinvestigated as part of an international research project funded by the European Commission (ECSC). The main objective of this work is to enhance the resolution or reliability of intrabasinal lithostratigraphical correlations of coal seams through an integrated sedimentological and sequence stratigraphical approach. Previous sedimentological analyses include scientific case studies conducted under the supervision of Italian universities (Siena, Cagliari) and consultancy studies completed by national or international experts. The former dealt with micropaleontological, palaeoecological or microfacies characteristics of the marine and lacustrine
carbonates, which are closely associated with the coals. The latter concentrated on the search for local lithological and palaeontological marker beds. Most of these data are unpublished and were made available to us by courtesy of the Board of Directors of Carbosulcis. The coal-bearing sequence (the so-called 'Produttivo') displays complex interfingering of marine-influenced and non-marine sediments, including coals, carbonaceous mudstones, marls, brackish to fresh-water limestones, and carbonate-rich detrital rocks. The latter 'hybrid' clastic rock types (quartz-rich lithocalcirudites, lithocalcarenites) represent minor or major clastic intercalations, which can be used for subdividing the coal-bearing sequence into successive lithological intervals. This paper will focus on the origin and the role of the latter 'detrital' intervals in the context of a revised depositional model for the coal-bearing deposits of the Sulcis Basin.
From Gayer, R. & Pegek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 49-75.
50
R. DREESEN E T AL.
Fig. 1. Location map of Sulcis mining project area (dark shaded area) in SW Sardinia.
Regional geological setting, palaeogeography and stratigraphy The coal-bearing Sulcis Basin formed during early Tertiary times as a result of extensional tectonic events which affected the south-eastern edge of the Iberian Plate (Assorgia et al. 1992). After an initial sea level rise at the base of the Palaeogene, a sea level fall was induced in the early Eocene by the Pyrenean orogeny in the western Mediterranean area. Subsequently, the synsedimentary graben-like Sulcis Basin was formed and reached equilibrium between subsidence and infill (Fadda et al. 1994). The Sulcis Basin was infilled by various sedimentary and volcano-sedimentary deposits: Palaeogene marine, brackish and continental (coal-bearing) deposits, Oligo-Miocene calc-alkaline volcanics (ignimbrites); and Neogene fluviatile and fluviolacustrine deposits. The aeral extent of the Sulcis Basin is estimated at about 200 km 2. It is bordered to the east by Palaeozoic basement outcrop and to the west by the sea (Fig. 2). The Eocene coalbearing formation ('Formazione produttiva a lignite' or 'Produttivo') has a known subsurface areal extent of more than 100 km 2. It dips to the SSW with an average dip of 8-10 ~. The thickness of the Produttivo and the number and thickness of the coal seams gradually increase to the SSW. The Produttivo Formation reaches a maximium thickness of 70 m. In the mining project area of Monte Sinni (Fig. 1) its
average thickness is 40-50 m at depths between 200 and 400m. The estimated coal reserves exceed some 250 million tons. The Sulcis coal can be classified as a low rank non-caking coal with very high volatile matter, low reflectance and calorific values (sub-bituminous A-B coal; Glanzbraunkohle). The heterolithic coal-bearing Caenozoic sediments of the Sulcis Basin have been affected by E-W and N N W - S S E oriented block faults (Fig. 2), which can be related to the anticlockwise rotation of the Corso-Sardinian microplate (Orsini et al. 1980) during successive tectonic phases of the Alpine orogeny. Eocene normal growth (listric) faults created halfgraben structures and were responsible for thickness variations of the coeval sedimentary deposits (Assorgia et al. 1992). The coal seams and the associated 'barren rocks' (mostly limestones and marls) are often affected by smallscale folds. The latter occur in narrow belts and they originated as a result of differential lateral movements (gravitational sliding or slumping; Bandelow & Gangel 1993) related to rollover phenomena (Cocozza et al. 1989; Fadda et al. 1994). According to Plaziat's (1981) palinspastic palaeogeographic reconstruction of the periPyrenean region, the Sulcis area represented a coastal embayment during early Eocene times. This area was bordered by a shallow sea in the SE and was intermittently affected by terrigenous influxes derived from a continental source, supposedly located in the NW. The latter
EOCENE SULCIS COAL BASIN
51
Fig. 2. Simplified geological map of the Sulcis Basin, SW Sardinia (modified after Fadda et al. 1994). Heavy lines correspond to major faults.
Fig. 3. Palaeogeographical map of the peri-Pyrenean region during Ilerdian (Ypresian; early Eocene) times (modified after Plaziat 1981). source area would correspond to the PyreneoProvengal mountain chain (Fig. 3). Tambareau et al. (1989) have stressed the analogies between the continental microfaunas and palynofloras of the Ilerdian (Ypresian) deposits of Sardinia
and the Pyrenean foreland (Languedoc area) which corroborate the continental continuity between the Pyrenean area and Sardinia. The Palaeogene deposits of the Sulcis Basin reach a maximum thickness of 140m and
52
R. DREESEN E T A L . DEPOSITIONAL
STAGES
FORMATIONS
MAJOR
LITHOLOGIES
COCOZZA et al. 1989
CONSULTANCY STUDIES
ASSORGIA et al. 1991 FADDA et al. 1994 tu u
......
O z
Volcano sedimentary
~ ......
pyroclastlcs
volcanics interbedded within continental terdgenous successiona
'
.............
rhyolitic Ignlmbrttas
.............
andesitic basaits
complex 0
I
~_) . . . ~ .
0
polygenic cong,o.~
Cixerri Fm.
alluvial fans
claystones
~ t o ~ o ~ .,
..........
Fm.
'1r 7~ rr. ~ Miliolitic
-
,-
30- 40 m B~S~L CONGLO.U.., --
PALAEOZOIC~ BASEMENT
~ custdne
freshwater Iimest. marls claystones
parallc lagoonal
~
~-=-
PERMO- " ~ J f / J \ TRIAS.~C ~ v j / / /
.
.
.
.
.
.
.
littoral
~
hypersaline & mesohallne lagoons
marls
-
.
.
.
.
w
......
lacustrine palustrine episodically interrupted by channellized tidal clastic deposits reworking pseudomorphosed
fluvial channels supralittoral paludal lagoonal
fluvial channels
b i o - calcarenites
~
Limestone Fm. P-,-~_':-CF.~
cgl.
"-" siltat. ~ r . ~ L
Produttivo
i ~
not studied
braided plains
alltatones
i
not studied
alluvial fans
sandstones
LU
0
THIS PAPER
450 m ,~, b . . , h . . ,, ,,., ,~
0 ~
Z
ENVIRONMENT
lag
folded metasedl.,mentary
evaporites
sublittoral
\ \ '
notstudied
restricted marine with fluctuating salinities to
ah\...allowmarine ~
\
Fig. 4. Stratigraphic scheme, major lithologies and depositional settings of the Palaeogene deposits in the Sulcis area. display a characteristic transgressive-regressive megasequence with marine carbonates above a polymict conglomerate at the base; paraliccontinental heterolithic sediments in the middle; and continental facies at the top of the sequence (Fig. 4). A marked reduction in diversity of the microfaunas and an inferred salinity anomaly about 35m above the base of the marine carbonates marks the onset of a regression, which reached its acme with the deposition of continental coal-bearing sediments. The resulting heterolithic paralic formation, the so-called 'Produttivo' is sandwiched between shallow marine limestones ('Calcare a Miliolidi') at the base and coarse detrital fluvial deposits at the top ('Cixerri Formation') (Fig. 4). Most of the boreholes drilled by Carbosulcis reached the top of the Miliolitic Limestones Formation only. Good biostratigraphic markers are restricted to the marine carbonates. The basal part of the Miliolitic Limestones contains large foraminifera (Alveolinids and Orbitolitids) which suggest an Ilerdian age (early Ypresian, Early Eocene; Cherchi 1983). The age of the coal-bearing formation is more difficult to determine due to obvious palaeoecological constraints. However, the lowest coal-bearing strata yield charophytes
and palynomorphs suggesting a Cuisian (late Ypresian) age (Pittau 1977; Salvadori 1979). The Produttivo Formation is unconformably overlain by the Cixerri Formation, the basal part of which has been assigned to the earliest Lutetian (Middle Eocene) on the basis of palynomorph and charophyte content (Pittau Demelia 1979).
Previous work The first detailed sedimentological-palaeoecological work on the coal-bearing Tertiary Sulcis Basin was an unpublished study conducted by the University of Siena (Cocozza et al. 1989). In this study, four major lithostratigraphical units were recognized within the 'Eocene series'. The basal unit corresponds to the shallowingupward coastal-lagoonal carbonates of the Miliolitic Limestone Formation (Fig. 4). The second and third units roughly correspond to the coal-bearing ('Produttivo') sequence. The second unit was interpreted as a gradual transition from coastal-lagoonal carbonates through brackish-lagoonal and heterolithic coal-bearing lacustrine-palustrine facies. The latter unit is interrupted by an important fining-upward clastic deposit, which has been interpreted as
EOCENE SULCIS COAL BASIN fluvial in origin. The third unit is lithologically similar to the second, but the coals tend to be thicker. Its upper limit is marked by a second clastic episode. The fourth and final unit corresponds to the Cixerri Formation (partita) and is characterized by the lack of coal and by the dominance of coarse clastics. The latter have been interpreted as braided fluvial channel deposits. The carbonates of the basal unit comprise a variety of microfacies types, indicating lowenergy, shallow lagoonal environments of a tropical to subtropical arid coastal zone (Cocozza et al. 1989). The benthic foraminiferal faunas indicate dominant hypersaline lagoonal conditions, with intermittent freshwater influxes (temporary hyposaline conditions; Cocozza etal. 1989). Foraminiferal wacke/packstones with ostracodes, molluscs, green algae and echinoderms are the dominant microfacies type. Crossbedded grainstones with peloids, low-energy fibro-radiated ooids, intraclasts and cyanobacterial oncoids are less frequent. The latter grainstones enclose corroded larger foraminifera with open-marine affinities (Alveolina, Orbitolites), the presence of which can been related to storm-induced transport. All microfacies types show more or less important bioturbation. Near-emergence is indicated by plant rootlets, incipient pedogenesis, and mottling whereas micro-karst phenomena suggest temporary subaerial exposure. The carbonates in the second and third unit comprise freshwater-influenced restricted-marine ('paralic') and lacustrine limestones (Cocozza et al. 1989). The paralic mud/ wacke/packstones contain abundant smoothshelled ostracodes, brackish foraminifera (e.g. Ammonia), freshwater molluscs, charophytes and dwarf miliolinids. Rare gastropod-bivalve coquinas (Cyrena, Potamides) occur in their basal part. The lacustrine limestones consist of mudstones, wackestones and (pseudo) packstones with variable amounts of skeletal grains such as those of freshwater gastropods, freshwater bivalves and charophytes. Plant root pedoturbation and other pedogenetic features are common. The clastic or detrital facies consist of finingupward sequences of rudites, arenites and siltstones. These strata have been related to braided fiver deposits in a distal alluvial plain setting (Cocozza et al. 1989). Other unpublished reports (e.g. by RIMIN 1990-1991) contain less sedimentological information. These studies focussed on the stratigraphical framework of the Produttivo. The clastic episodes have been again interpreted as channelized fluvial deposits. A major result of these
53
studies has been the subdivision of the coalbearing sequence into informal lithostratigraphic units, and the regrouping of the numerous tiny coal layers into 10-12 multi-seam coal horizons. This lithostratigraphic scheme was based on lithological predominance criteria and on the presence of local marker beds. However, bed-bybed correlation is not generally possible, except in the case of closely spaced boreholes. Even then, correlation of individual coal seams is problematic because of marked seam irregularities. The wedge-shaped sandstones which erode underlying coals, were interpreted as fluvial channels. The coarser clastics of the overlying Cixerri Formation were attributed to alluvial fan deposits. A stratigraphical study recently conducted by Montan Consulting GmbH (Bandelow & Gangel 1993) lead to similar conclusions. Although potential marker beds were identified (including questionable bentonite layers) and the correlation of coal seams over larger distances between closely spaced boreholes was possible, a bedby-bed correlation of the coal seams remained uncertain, due to the combined effect of synsedimentary tectonics (difference in thickness of coeval strata) and sedimentological events (e.g. wash-out phenomena).
Facies spectrum and distribution The 40-70 m thick heterolithic sequence of the Produttivo Formation is subdivided into four informal lithostratigraphic units, based on a detailed study of 50 closely spaced cored boreholes in the area of Monte Sinni (Fig. 1). Each unit contains one or more parasequences, bounded by unconformities which correspond to the erosional bases of the 'detrital episodes' (Fig. 5). The boundary between the Miliolid Limestones and the Produttivo is not clear-cut, but consists of a gradual transition from restricted marine to lacustrine-palustrine limestone facies. This boundary represents not an unconformity but rather a palaeoecological/palaeoenvironmental change within the coastal lagoon setting. The sudden occurrence of 'paralic' microfaunas (mixohaline or hyposaline conditions) and the first occurrence of (autochthonous) coal seams, can be used as criteria for defining the base of the Produttivo Formation. This event is characterized by the mass occurrence of (pyritized) smooth-shelled ostracodes and by an important decrease or even disappearance of miliolinids (dwarf forms). The Cixerri Formation which unconformably overlies the Produttivo consists of conglomerates, coarse sandstones and var-
54
R. DREESEN E T A L .
Fig. 5. Ideal parasequence within the Produttivo Formation, with evolution of relative water depth and relative sea level (MFS: marine flooding surface; TST: transgressive system tract; HST: highstand system tract; LST: lowstand system tract; EV: evaporite; PAL/LAC: palustrine / lacustrine; TF: tidal flats. Not to scale.
iegated mudstones. The lack of coals and limestones is used here as a criterion to distinguish it from the Produttivo, although a gradual transition cannot be excluded. The ideal parasequence or depositional unit within the Produttivo Formation is an upwardshoaling unit, several metres to about 10 metres thick. It is bounded by marine flooding surfaces, coinciding with an erosional unconformity (Fig. 5) and marking the base of a 'detrital' episode. This depositional unit corresponds to a parasequence, in the sense of Van Wagoner et al. (1988, 1990), and the Produttivo Formation is composed of a parasequence-set formed by
the stacking of analogous parasequences. In contrast with the unpublished reports referred to earlier, we consider the detrital episodes not as fluvial deposits but rather as marine sediments. They are interpreted as shallow marine tidal deposits, displaying characteristic subtidalintertidal-supratidal upward-shoaling sequences (Fig. 6). The 'detrital' sediments represent the lowermost part of each parasequence. The upper part consists of 'continental' supratidal deposits, which grade vertically (and laterally) into lacustrine and palustrine coal-bearing facies. The tidal character of the clastics is indicated by a vertical suite of characteristic sedimentary structures, whereas tidal subenvironments may be distinguished on the basis of lithologies and associations of structures (Terwindt 1988). Marine fossils are apparently lacking in the studied shallow marine tidal deposits, despite indirect evidence for in-situ organic activity such as intense bioturbation and the presence of escape structures. Reworked cyanobacterial mats or oncoids (stromatolites) and fragmented thick-shelled (mixohaline?) molluscs occur in the basal part of each parasequence. We suggest that this apparent lack of in situ organisms must be related to the extreme harsh environmental conditions (abnormal salinities) during deposition of the subtidal and intertidal sediments. Analogous observations have been made in tidal sequences from the Late Devonian Psammites du Condroz Group of the Ardennes in Belgium (Thorez et al. 1988). The subtidal facies is represented by relatively thin (5-20cm) coarse-grained lag deposits ('microconglomerates') overlaying erosional unconformities. The lags are composed of grey quartz-rich (litho-)calci-dolorudites and (litho-) calci-doloarenites displaying some grading and oblique or cross stratification. The subangular, strongly packed pebbles or granules include various intraclasts such as grains consisting of calcite and dolomite cement, lacustrine mud/ wackestones, pedogenic carbonates, black pebbles, oncoids, stromatolitic crusts, chert, coal and mollusc fragments. Euhedral quartz grains are common and locally abundant. This subtidal facies is interpreted as a thin transgressive unit. It coincides with a transgressive lag, which resulted from the reworking of underlying or lateral deposits, i.e. the winnowing of finegrained sediments and the accumulation of coarse-grained sediments on the ravinement surface, during shoreface erosion (Swift 1975). The ravinement surface corresponds here to the flooding surface (MFS; Fig. 5). Because of the reduced sediment influx (as the shelf area expands the volume of sediment being supplied
EOCENE SULCIS COAL BASIN
55
Fig. 6. Borehole 67-91: Lithologies, sedimentary structures and inferred tidal subenvironments within clastic deposits of the Produttivo Formation. Logs a and b refer to basal parts of lithostratigraphic intervals D and C respectively (see Fig. 7). per unit area decreases) one of the principal sources of coarse material for transgressive deposition is cannibalization of previously deposited sediments (Arnott 1995). The intertidal facies consist of a sequence, a few to several metres in thickness, of grey
'sandstones', 'siltstones' and 'mudstones', which correspond to intensly bioturbated, alternating calcarenites and carbonaceous calcilutites. Individual beds are between ten and several tens of cm in thickness, not exceeding 1 m. The latter calcarenites and calcilutites vertically grade into
56
R. DREESEN E T AL.
dolarenites and dololutites: stained acetate peels (Friedman 1959; Katz & Friedman 1965; Dickson 1966) and X-ray diffraction indicate nonferroan and ferroan dolomite (the latter as a cement) increasing from bottom to top (Fig. 7). Kaolinite is omnipresent. Lenticular, flaser and ripple bedding are most common, whereas pervasive pedoturbation (due to plant roots?) frequently induces destratification. Lower and upper intertidal subenvironments can be distinguished on the basis of sedimentary structures (Fig. 6). Vertically recurrent microconglomeratic or litharenitic levels suggest stacking of the tidal sand bodies. The supratidal facies consist of a sequence, several meters in thickness, of grey dololutites displaying a rather massive aspect due to intense bio- and pedoturbation. Palaeosols are present as proved by the occurrence of mottling and illuviation-oxidation phenomena along rootlets and by the presence of caliche nodules. The change from marine to continental supratidal conditions is marked by a change in colour (grey versus tan/beige) and by pedogenetic features: incipient calcretes, rootlets or seatearths related to overlying coals. The overlying palustrine-lacustrine facies are characterized by irregularly alternating coals, carbonaceous mudstones, freshwater limestones and marls. The beige to characteristically hazelnut-brown coloured ('nocciola') limestones have been deposited in shallow, nearshore lacustrine environments (littoral carbonates). They consist of bioclastic and phytoclastic material: plant remains and skeletal debris of fresh-water molluscs (including limnic gastropods e.g. Planorbis, Paludina, Melanopsis; Cocozza et al. 1989), smooth-shelled ostracodes and charophyte gyrogonites. Lignite laminae are common and all intermediate lithologies exist between relatively pure, non-carbonaceous limestones and impure coals. Staining shows that the limestones consist exclusively of non-ferroan calcite. Strongly packed monospecific mollusc shell debris may account for the characteristic pseudopackstone texture. Although deep water facies have not been recorded, some limestones are poor in skeletal debris and display a varvoid texture. Dissolution residues reveal only a subordinate amount or even a lack of siliciclastic material (Fig. 7). The limestones contain numerous roots and there is evidence for subaerial exposure, such as the presence of rhizoliths, Microcodium, desiccation cracks, micro-karst phenomena and nodular fragmentation or 'micro-brecciation'. Marls are interbedded with the limestones. They are generally dark-coloured, carbonaceous, rooted and contain occasional shells (mollusc,
ostracodes) as well as pyritized charophyte gyrogonites. Individual coal seams never exceed lm in thickness, but coal-limestone associations build up sequences several metres thick. The coals are composed of leaf cuticles, spores, pollen and sub-ordinate algae. Petrographic analysis of several hundreds of samples gives an average of 73.3% vitrinite, 11% liptinite, and 5% inertinite (Fadda et al. 1994). The liptinites frequently contain alginite, whereas fusinite and semifusinite are common in the inertinites of the lower coal seams. Other physico-chemical parameters include: volatile matter content of 42%, vitrinite reflectance values of 0.45 to 0.5%, fixed carbon of 48 to 52%, average ash content of 10% and average S content of 6% (Fadda et al. 1994). The ash analysis shows a low silica ratio with high iron (24%) and sulphate (17%) content. The general lack of well-developed seatearths or rooted horizons and stump or stem horizons (Fadda et al. 1994) suggests that some of the Produttivo coals are allochthonous: they result from the reworking of plant remains from swampy-marshy zones into subaqueous, nearshore lacustrine areas. Moreover, the presence of alginite in most of the coal seams would indicate a partial algal origin. The palynomorph content of the coal-bearing strata is rather poor and badly preserved. The associations are dominated by herbaceous plant pollen typical of both warm palustrine-savanna and moderate steppe-prairie type environments (Cocozza et al. 1989). The good preservation of fresh-water mollusc shells and charophyte gyrogonites in the coals and the close association with carbonates suggest that water acidity was very low. Decimetric coquinas are sometimes interbedded with the coal-limestone sequence in the lowermost part of the Produttivo. These coarse mollusc packstones locally display erosional bases and contain coal clasts. They are almost exclusively composed of gastropod (Potamides) and bivalve (Cyrena) packstones, suggesting mixohaline environments (Cocozza et al. 1989). The scoured bases, orientation of the shells and indistinct grading of the coquinas point to a possible storminduced origin. For four selected lithostratigraphical intervals (A through D), cumulative thicknesses of lithology classes have been processed into isopach maps, using simple kriging contour routines. The contour maps (Fig. 8) cover an area of about 16 km 2, which approximately corresponds to the mining project area (Fig. 1). The following lithologies have been selected and grouped for processing: coals (including all dirty
Fig. 7. Borehole 67-91: Lithological-sedimentological log showing mineralogy, clay minerals (right) and interpreted depositional environment (left). A-D: lithostratigraphical intervals; S.M.: shallow-marine, R.M.: restricted marine; SB: subtidal; SB/IT: subtidal/intertidal, IT: intertidal, SP (m): (marine) supratidal, PL/LC: palustrine/lacustrine, PS: palaeosol; Ca: Calcite; Do: Dolomite; Qtz: Quartz; Cl: Clays; KF: K-Feldspar; K: Kaolinite; I: Illite; C: Chlorite; Sm: Smectite; (10-14C) and (10-14Sm): irregular mixed layers. Shaded lines represent allochtonous coals. Arrows correspond to upward shoaling trends. Left-hand side of log corresponds to carbonates, right-hand side of log to clastics and coals.
58
R. D R E E S E N E T AL.
o
,,.._.,,
o
8 o
e.-.
ff ..=
o
8
8~
EOCENE SULCIS COAL BASIN
59
60
R. DREESEN E T AL.
coals and carbonaceous mudstones), lacustrine limestones (including all marly limestones), and the clastics (including 'siltstones', 'sandstones' and 'conglomerates'). The regressive trend of the Produttivo is evidenced by a progradation, in time, of coal from the SW to the NE. Furthermore, a comparison of the coal and the limestone isopach maps reveals a good correlation between both lithologies, except for the earliest interval A. This suggests that the limestones are genetically related to the coals (corroborating the inferred lacustrine-palustrine environment). However, the limestones of interval A are not true lacustrine but rather paralic in origin: i.e. they were deposited under hyposaline, restricted marine conditions indicated by their microfacies and microfaunal content. Their maximum thickness shows a N N W - S S E to N N E - S S W orientation, which parallels or corresponds to the former coastline. The coals and the clastics are mutually exclusive, as shown by the location and orientation of their zones of maximum thickness. The area of maximum development of the clastics roughly corresponds to that of the paralic belt, which would corroborate their relation with the marine environment. The stacking of zones of maximum thickness of clastic rocks in the NW corner of the studied area (compare the successive siltstone/sandstone/conglomerate isopach maps in Fig. 8) might indicate the existence of a narrow depression or a preferential pathway, such as a channel or a 'slough', for the clastics in this area. The clastic bodies have a variable
thickness and lateral extent (Fig. 9): the small and thin lenticular sand bodies are interpreted as shallow marine channel or gully fills whereas the larger sand bodies correspond to stacked sand sheets. The latter have a lateral extent up to 1 km and a maximum thickness of 5 m.
Petrography and geochemistry Methods More than 100 thin-sections have been studied with conventional and cathodoluminescence (C.L.) petrography. C.L. petrography was carried out with Technosyn Cold Cathodo Luminescence Model 8200 Mark II. Operating conditions were 16-20 kV gun potential, 420 #A beam current, 0.05 Torr vacuum and 5 m m beam width. After careful petrographic characterization of individual authigenic minerals, a microscope mounted micro-drill assembly (with a drill-bit of 0 . 5 - 1 m m ) was used to obtain carbonate powders of 1-10mg for isotopic analysis. However, complete separation of individual phases was not always possible. Isotopic analysis of carbon and oxygen was performed on a Finnigan Mat delta E stable isotope ratio mass spectrometer. Carbonate powders were dissolved in >100% orthophosphoric acid at 25~ All data have been corrected following procedures modified from Craig (1957). The isotopic compositions are expressed as 0 values in per mil (%0) difference from the PDB
Fig. 10. (1) Photomicrograph of non-stained calci-dolomite. Dolomite and monocrystalline quartz rock. Locally a chert particle (ch) as well as some spherical chalcedony bioclasts (s) occur. Scale = 160 #m; PPL; (borehole 4/B 76~tr-47.50 m). (2) Photomicrograph of an euhedral quartz particle with carbonate (c) and lath shaped anhydrite (A) inclusions, surrounded mainly by monocrystalline quartz and algal micrites (M) particles. Notice that the latter are severely affected by compaction. Scale = 80 #m; PPL; (borehole 4/B 76~tr-38.30 m). (3) Photomicrograph of quartz particles consisting of different phases which locally possess a euhedral outline. A lath shaped internal arrangement is accentuated by the presence of elongated calcite inclusions. Locally some minute lath shaped anhydrite inclusions (A) occur. This particle is surrounded by micritic algal clasts. Scale = 80 #m; PPL; (borehole 43/90-433.25 m). (4) Photomicrograph of quartz particles with ghosts of pseudomorphosed lath shaped crystals displaying a felted texture. This particle is dominantly surrounded by dolomite as well as quartz particles. Notice that some of the latter possess a euhedral to subhedral outline. Furthermore some micrite algal particles (M) are present. Scale = 80 #m; NPL; (borehole 43/90-447.70 m). (5) Photomicrograph of different types of particles, namely a length slow chalcedony particle (CH) with dolomite inclusions (D), a laminated micritic algal particle (M) and a micrite particle with relict sponge spines (S). Scale = 80#m; PPL; (borehole 43/90-433.25 m). (6) Photomicrograph of well-rounded polycrystalline quartz particle surrounded by dolomite. Typical is the undulose extinction as well as the trails of minute inclusions within the quartz phases. In the lower left corner part of a monocrystalline quartz grain with sometimes lath shaped calcite inclusions occurs. Scale = 20 #m; PPL; (borehole 47/90-477.60 m). (7) Photomicrograph of intensely compacted micritic algal clasts next to dolomite and quartz particles. Notice the microsparitic nature (MS) of the algal clast on the left side of the picture. Scale = 85 #m NPL; (borehole 67/91-377.30 m). (8) Cathodoluminescence photomicrograph of Fig. 10.7. The fine tubular texture within the algal clast as well as recrystallisation textures become more apparent. Notice the presence of a zoned calcite particle with truncated edges between the algal clasts. Furthermore dull red dolomite particles are easily distinguishable from the non- to darkbrown luminescing quartz particles. Around some of the dolomite particles a dull luminescing cement (D1) is present. It precedes a deep red phase (D2). Scale = 85#m; CL; (borehole 67/91-377.30 m).
EOCENE SULCIS COAL BASIN
61
62
R. DREESEN ET AL.
international standard. Reproducibility, determined by replicate analysis of samples NBS 19 and NBS 20, is better than 0.1%o for oxygen and 0.05%o for carbon. No correction for dolomite or siderite dissolution by phosphoric acid has been applied.
Petrography of sedimentary particles A detailed study of the (litho-)calci-dolorudites and (litho-)calci-doloarenites of the Produttivo Formation revealed that the detrital grains mainly consist of quartz and carbonate (dolomite and calcite) particles (Fig. 10.1). Well rounded chert as well as feldspar particles (K-feldspar and subordinate plagioclase) are present in low concentrations (<5%). Cements are in general scarce and dominantly consist of dolomite, ferroan dolomite and locally some calcite. Authigenic kaolinite and pyrite content locally can grade up to 5%. Intracrystalline porosity between dolomite phases as well as secondary porosity locally can be in the order of 15%. Monocrystalline quartz particles exceed 95% of the total quartz population. Two dominant types can be differentiated. Firstly, rounded to subrounded quartz grains with minute (< 1 #m) inclusions which often are arranged along intraparticle planes. Most of these grains possess a uniform extinction and their origin is not determinable. Secondly, euhedral to subhedral quartz phases which are characterized by lathshaped inclusions (Fig. 10.2-3). Few rounded examples also exist. The inclusions, which can be 30 by 5 #m in size, either are aligned parallel to each other, or are scattered within their host. A zonal arrangement is less common. Petrographic characteristics as well as SEM-EDX data, indicate that these lath-shaped inclusions consist of anhydrite or calcite. Other inclusions are framboidal pyrite and rhomb-shaped dolomite. In some cases, lath-shaped inclusions have been completely replaced by mega-quartz, their outline is often still visible (Fig. 10.4). In many cases a typical felted texture is present. Important here is the fact that in none of the cases has a dust rim been identified in the inclusion-bearing part of the quartz grains. Thus the euhedral to subhedral outline does not relate to post-depositional authigenic quartz overgrowth. Furthermore, some grains clearly show transportation features. Broken edges crosscut both quartz and the lath-shaped inclusions. These curious quartz phases correspond to transported early diagenetic replaced evaporites. These textures are similar to the megaquartz phases described by Folk & Pittman (1971), Milliken (1979), Arbey
(1980) and Swennen & Viaene (1986). According to these authors such evaporite replacement silica diagenesis preferentially develops within schizohaline settings. A similar origin can also be proposed for some of the microflamboyant, concentric and spherulitic chalcedony particles (Fig. 10.5) containing anhydrite inclusions or their pseudomorphosed outlines. However, not all chalcedony particles can be shown to have an evaporite related origin. The overall proportion of 'chert' particles is less than 3%. Within some polycrystalline quartz grains anhydrite-related relicts have also been observed. However, most of the polycrystalline quartz particles consist of a mosaic of intricate or well defined quartz crystals. Sometimes, these crystals display a parallel alignment and an undulose or gradual phase extinction (Fig. 10.6). The grains are often well rounded and are most likely of metamorphic origin. Under cathodoluminescence the majority of these quartz particles possess a dull to dull brown luminescence (Fig. 10.8; Fig. 11.5-6). Carbonate grains, which locally make up to 80% of the detrital content (Fig. 10.1) dominantly consist of dolosparite. Dolomicrite, micrite and sparite grains are also present. Most dolosparite particles have a grain size which is similar to that of neighbouring quartz grains. In transmitted light a cloudy core can often be differentiated from a transparent rim. They are both in optical continuity. Under C.L. this corresponds to a reddish brown to brown central part and a dull brown rim (Fig. 11.6). The centre is often broken in different pieces and rehealed by the dull brown or deep reddish dolomite phase. This, together with the fact that the rim is missing where adjacent grains touch each other, points towards a post-compactional origin for the rim. Most of the larger carbonate grains, which often are severely affected by compaction, consist of micrite or microsparite (Fig. 10.7). Within most of these pyrite- and organic matterenriched micrite grains faint radial and concentric microtextures are recognizable. Under C.L. these textures become more distinct (Fig. 10.8). Bright yellow tubes within a dull brown matrix testify that these micrite particles correspond to disrupted and transported porostromate algae. Within these clasts lozengeshaped calcite pseudomorphs after gypsum are locally present (Fig. 11.1). C.L. characteristics of the microsparite particles clearly indicate that these correspond to cemented and/or partly recrystallized porostromate algae (Fig. 10.8). Some of the cements within these porostromate algae reflect a complex cementation history. In
EOCENE SULCIS COAL BASIN addition to these algal fragments micrite clasts with sponge spicules have also been identified (Fig. 10.5). Within lithostratigraphic unit D some micrite clasts with pedogenetic textures, such as circumgranular cracks and glaebules are present (Fig. 11.2). The sparite grains are most easily recognizable after staining (Fig. 11.3) or under C.L. In the latter case they display a number of bright luminescing textures typical for cements such as different types of zonation patterns, sector zonation, variation in cement-type succession (e.g. acicular followed by blocky, etc.) (Fig. 11.5-6). The truncated particle terminations indicate that cementation did not occurred in situ but that these are sparite clasts which have been eroded and transported. Partly broken and compacted cement particles (Fig. 11.6) also support this interpretation. Many of the intensively zoned cement grains reflect a broad spectrum of variations in cementation conditions, most of which are compatible with shallow meteoric diagenesis in their source area. Also textures which may correspond to marine diagenesis (acicular non-luminescent crystal crusts), as well as speleothem-like textures have been identified. Large blocky uniformly dull or bright yellow particles may represent reworked shallow burial cements formed either in reducing or suboxic conditions or burial cements. A broad spectrum of cemented lithologies was evidently being eroded during Eocene times and their excellent preservation state suggest a nearby source area. Other unusual particles observed include: mono-axial chalcedony sponge spicules, up to 20#m in diameter (Fig. 11.3) which are also incorporated in larger chert particles; oolites, which commonly enclose a quartz nucleus; in some cases lath shaped anhydrite crystals or their calcitized pseudomorphs occur within the quartz grain; bioclasts such as gastropods, ostracodes, charophytes, bivalves, which occur within discrete layers. Some of these particles, such as poly- and monocrystalline quartz, are extrabasinal in origin and most likely are derived from the Palaeozoic sedimentary and intrusive igneous hinterland. However, for the dolomite particles, the calcite cement grains ('sparites') and the evaporite-bearing quartz grains, an extrabasinal origin is less obvious. An extrabasinal origin can be argued since the immediate (actual) surroundings of the Eocene Sulcis basin consist mainly of Cambro-Silurian formations including Cambrian carbonates. The dolomite particles could be derived from the Lower Cambrian Dolomia Rigata or the overlying Grey Dolomite Formation and Black Limestones Formation or
63
the nearly time equivalent Planu Sartu Member (Bechst~idt & Boni 1989). The complex zonation pattern recorded in the calcite cement particles would then reflect the complex diagenetic history recorded within these source area rocks. Evaporite moulds and some gypsum platelets have been described from the 'Dolomia Rigata' (Schledding 1985). However, their occurrences are minor and they are not silicified. Whether quartz pseudomorphs after evaporites occur in the nearby hinterland is not clear. In this scenario one would expect also to find many more Cambrian allochems (e.g. oolites, intraclasts) and reworked Cambrian fossil debris (e.g. echinoids, Archaeocyathids, trilobites) which occur in time-equivalent strata, which is not the case. Euhedral quartzes could also have been derived from nearby Triassic outcrops in SW Sardinia (but no longer exposed because of extensive erosion) or from the Triassic in southern France (given the palaeogeographic position of Sardinia, close to the Pyrenees; Fig. 3). Quartz bipyramids with anhydrite inclusions are quite common in the Keuper of the Corbi6res area (Debelmas 1974; Jalfrezo 1977). These Triassic evaporite-bearing sediments were unroofed by the uppermost Cretaceous, and certainly by the Palaeocene, and were thus providing clasts. Another scenario would be that these particles are quasi synsedimentary in origin. This would imply that during Eocene times (time equivalent with the deposition of Miliolitic Limestone Formation or Produttivo Formation) a carbonate-evaporite succession was deposited which underwent complex diagenesis. This diagenesis would be characterised by differentiated calcite cementation, dolomitisation and selective silicification of evaporites. This scenario is supported by the existence of algal clasts displaying a complex calcite cement zonation pattern and the presence of gypsum pseudomorphs (Fig. 11.1). Development of such a hypersaline setting with evaporite ponds would fit into the general palaeo-environmental context, but whether sufficient time would be available for such diagenetic overprinting can be questioned. The alkaline depositional conditions, however, are favourable for silicification (Arbey 1980). Within such a setting dolomitization could also be one of the dominant diagenetic processes: in fact concomittant dolomitization and silicification of evaporites has often been reported in literature (e.g. Swett 1965; Tucker 1976 and others). In order to generate clasts, these strata would subsequently have to be eroded, which could be explained by block-faulting which has been reported during Eocene time in the Sulcis
64
R. D R E E S E N E T AL.
EOCENE SULCIS COAL BASIN basin. This mechanism may thus have given rise to a nearby emergent source area for the considered particle types. This block-faulted barrier could perhaps also explain the development of a lacustrine-palustrine basin. Within this scenario the major difference in shape between the euhedral to subhedral evaporitebearing quartz grains and the well to subrounded mono- and polycrystalline quartz particles could be easily explained. However, in the extra-basinal scenario this difference might simply relate to the degree of reworking of quartz particles from older sandstones.
Diagenesis Unravelling the diagenetic history of the 'Produttivo' strata is complicated not only by the existence of diagenetic particles such as dolomite and 'sparite' grains (Fig. 11.5-6) but also by the severe compaction of most lithologies (Fig. 10.7; Fig. 11.2). Furthermore, cements are not very common. In general a distinction between very localised early diagenetic products and postcompactional phases can be made. In Fig. 12 a generalized paragenetic succession is given.
Fig. 11.
65
Early diagenetic products In the lower part of the Produttivo Formation but much more common in the Miliolitic Formation, biomold development followed by circumgranular intraparticle calcite cementation is common in the bioclastic horizons. Also recrystallization phenomena can be observed and most likely relate to the stabilization of aragonite and high Mg-calcite components and/or neomorphism of fine crystalline shell textures. Within these layers as well as higher in units B and C (Fig. 9), framboidal pyrite may locally exceed 5%. Its origin is here tentatively related to the activity of sulphate reducing bacteria which implies a regular supply and infiltration of marine water to provide the sulphate ions. Higher upwards in the sequence (unit D) pyrite is less common. Here, siderite cement and siderite nodules are more common and occur close to the development of pedogenetic textures (glaebules, circumgranular cracks, desiccation features, etc.). The fact that siderite instead of pyrite preferentially formed suggest a shift towards methanogenic related processes. Around some
(1) Photomicrograph of a micritic algal clast with calcite pseudomorphs after lozenge-shaped gypsum crystals. Scale = 80 #m; NPL; (borehole 54/90-332.65 m). (2) Photomicrograph of glaebule particle with circumgranular cracks cemented by several generations of calcite. Small disortic textures (Do) are locally discernible. This particle is surrounded by quartz and micritic algal particles. Scale = 80 #m; NPL; (borehole 58/90-503.85 m). (3) Photomicrograph of calci-dolorudite where the four most common particles are visible namely: a) quartz (Qz), b) dolomite (D: brownish hue, not stained), c) calcite (C: red stained), d) micritic algal particles (M) which became very dark due to staining. One circular section of a chalcedony bioclast (S; sponge spicule) is also discernible. Notice the presence of a blue ferroan dolomite phase which is affecting especially the dolomite particles. This ferroan dolomite phase also occurs in compactional cracks. Scale--85 #m; NPL; (borehole 67/91-377.60 m). (4) Cathodoluminescence photomicrograph with non- to darkbrown luminescing quartz (Qz) and dull red dolomite particles (D). In the central part some dolomite rhombs (DR) occur. The bright yellow (B) and orange yellow (O) particles corresponds to calcite particles. The following cements are present: a) dull luminescing dolomite (D 1) which only develops around some dolomite particles, b) red luminescing dolomite cement (D2), c) brown yellow to yellow calcite cement (C) characterized by sector zonation, d) purple luminescing kaolinite (Ka). Notice that between D2 and the calcite a corrosive contact exists (see arrows). Scale = 85 #m CL; (borehole 67/91-377.30 m). (5) Cathodoluminescence photomicrograph of calci-dolorudite with in the central part nicely zoned calcite cement particle. Here the truncated texture of the cement is clearly visible. Apart from non- to darkbrown luminescing quartz (Qz) and dull red dolomite (D) particles a bright blue luminescing K feldspar (Ka) is present as well as many micritic algal clasts (M) and other bright luminescing sparite (B) particles. Kaolinite (Ka) is the only easily recognizable cement. Scale = 85 #m; CL; (borehole 67/91377.30 m). (6) Cathodoluminescence photomicrograph of calci-dolorudite with similar composition as plate 3.5. Here the central calcite cement particle displays compactional cracks which are cemented by brown yellow calcite (C). Post-compactional dull luminescing dolomite (D 1) and red luminescing dolomite (D2) cement is also present. Between these dolomite cements and the calcite cement a corrosive contact exists (see arrows). Scale = 85 #m; CL; (borehole 67/91-377.30m). (7) Cathodoluminescence photomicrograph of contact between a quartz (Qz) calcidolorudite in the upper part consisting of dolomite (D) particle and a lower part composed of rhombic dolomite crystals which most likely developed within a sheltered pore. The position of the bioclast is now taken over by the bright yellow luminescent calcite (C). Between the dolomite rhombs as well as within the calcite cement some purple luminescent kaolinite (Ka) occurs. Notice that the kaolinite booklets in the calcite clearly are floating and are corroded (see arrows). Scale = 85 #m; CL; (borehole 67/91-377.60 m). (8) Cathodoluminescence photomicrograph with the classical particles of a calci-dolorudite (see plate 1.5 for abbreviations). In the central part a calcite cement particle (CC) is present. Scale = 85 #m; CL; (borehole 67/91-363.16 m).
R. DREESEN ET AL.
66
DIAGENESIS EARLY
.-
BURIAL
SEDIMENTATION PEDOGENESlS (L) SIDERITISATION (L) FRAMBOIDAL PYRITISATION (L) CALCITE I PRECIPITATION (L) BIOMOLD DEVELOPMENT & RECRYSTALISATION (L) AUTHIGENIC QUARTZ DEVELOPMENT COMPACTION 1 STYLOIJTISATION & COALIFICATION NON-FERROAN DOLOMITE (NFDI) FRACTURING EPISODE I FERROAN DOLOMITE (FDI) NON-FERROAN DOLOMITE (NFDII) SECONDARY POROSITY DEVELOPMENT KAOLINITE FORMATION FRACTURING EPISODE II BLOCKY BRIGHT YELLOW TO BROWN CALCITE II L ) = only locally
developed
Fig. 12. Generalized paragenetic succession of diagenetic events in the Produttivo Formation.
of the monocrystalline quartz particles a thin authigenic rim is present, however it is not determinable whether it formed before or after compaction.
Compaction Interpenetrating to sutural grain to grain contacts in addition to the existence of sutured and non-sutured seam solutions testify to an important episode of compaction. Low amplitude stylolites are present locally. The micrite/microsparite particles are affected, while dolomite and 'sparite' particles show similar textures to the quartz grains (Fig. 10.8). The period during which burial gave rise to the compaction features most likely corresponds with the major period of coalification.
Late diagenetic products Syntaxial non-ferroan dolomite (NFDI) overgrowths affecting only dolomite particles is often present. This nearly inclusion free transparent
phase possesses a dull lunminescence (Fig. 11.6). It is only present within pores which survived compaction and is absent at grain to grain contacts. This cement is often covered by ferroan dolomite which luminesces deep red (Fig. 11.6; FDI). The irregular to wispy contact between both cement phases as well as the fact that ferroan dolomite spots also occur within the dolomite grains and along cleavage planes (Fig. 11.5) indicate that most of the dolomite phases were partly to entirely recrystallized by ferroan dolomitizing solutions. The ferroan dolomite also occurs as rehealing phases in the compactional cracks of broken dolomite grains. The many thin fractures which are cemented by ferroan dolomite testify to an important phase of fracturing preceeding ingress of these solutions. In the central part of some of these fractures a non-ferroan blocky dolomite cement (NFDII) occurs which has not been recognized outside the fractures. Whether this was linked to a fracturating episode is not obvious. A relatively important late diagenetic episode was the development of secondary porosity which is most likely related to feldspar dissolution. This is supported by the fact that the
EOCENE SULCIS COAL BASIN feldspar content in the siderite nodules is an order of magnitude higher than in the surrounding lithologies. Early diagenetic siderite cementation and nodule formation are believed to have sealed these parts of the rock so that they were unaffected by late diagenetic processes. A relationship exists between lithologies with high secondary porosity and abundant kaolinite matrix, as recorded by Giles & Marshall (1986). Kaolinite booklets occur within primary intercrystalline pores as well as in secondary pores (Fig. 11.2-3). A genetic link between feldspar dissolution and kaolinite precipitation seems likely, especially within a setting where coalification occurred. According to Giles & Marshall (1986) organic acids liberated during the maturation of organic matter, especially coal, are efficient agents of feldspar dissolution. In the case of a closed to semi-closed diagenetic system ions, liberated during dissolution can give rise to the precipitation of kaolinite and other authigenic minerals such as illite and authigenic quartz. The latter, however, have not been recognised in the studied strata. Finally, after a renewed fracturing episode, fluid channeled along these fractures caused calcite fracture fill and limited cementation. A corrosive contact between this bright yellow to brown luminescent calcite and other carbonate phases, especially the ferroan dolomite cement (FDI; Fig. 11.4), testifies to the aggressive nature of these fluids or to the circulation of undersaturated fluids preceeding cementation. Within some larger cavities corroded kaolinite booklets appear to float in this late diagenetic calcite (Fig. 11.7), indicating that this calcite cementation postdates kaolinite precipitation. A remarkable feature of this calcite cement is that it only occurs in fractures or large (>300 #m) pores.
Stable isotopes Table 1 is split in two parts, and gives the 013C and 0180 of the analysed samples. Part A groups all the samples which consisted of rather pure sedimentary or diagenetic phases, or where based on microscopic examination, the relative proportions of different contributing phases can be estimated. It was not possible to sample calcite I and the non-ferroan dolomite I seperately. In part B samples are grouped where such an estimation could not be carried out due to the presence of too many different components. An essential first step in diagenetic studies is the estimation of the original isotopic composition of marine water. This starting composition
67
can serve as a standard value against which the diagenetic products can be evaluated. Reported estimates for Lower Eocene marine carbonates vary around +0.5 + 5%o Ox80 and +2.3 • 0130 (bulk sediment data reported by Shackleton 1986). The coquina debris and its marine limestone matrix (Miliolitic Formation) which possess values of 0i80 of -5.0 + 0.3%0 and 013C of -1.30+0.05%o are clearly depleted with respect to Lower Eocene marine values. This also accounts for stable oxygen and carbon data from the Miliolitic Limestone of the Sulcis Basin reported by Perna et al. (1994) ( 0 1 8 0 : - 5 . 7 to -9%o and 013C: -0.6 to -3.3%o). Recrystallization by meteoric and/or warm fluids with involvement of depleted CO2, most likely derived either from soil-gas CO2 or from decarboxylation reactions within the coal layers of the Produttivo Formation are possible explanations. This seems also to account for most isotopic signatures of the calcarenites and other detrital calcite/dolomite dominated lithologies sampled in this study. These strata cluster within an area defined by 013C of -4.5 • and 0180 of -7.2 +0.8%o. Sample 400.00m however is more depleted in 0180 (-9.82%0). This may relate to the presence of minute calcite II veinlets and possibly to recrystallization due to interaction with calcite II bearing solutions. One of the larger calcite II veins analysed was collected from this sample. 013C and 0180 values of the glaebules respectively plot around -9.2%o and -5.4%o (Table 1 and Fig. 13). The influence of soil-gas CO2 in the soil formation is clearly reflected in the depleted 013C of these pedogenic carbonates (Salomons et al. 1978; Cerling 1991). The oxygen isotope signature is in agreement with a meteoric water dominated system. The few sampled siderites are characterized by a 013C varying between -0.5 to +2.8%0 and 0180 varying between -4.9 to -8.0%0. The carbon signature could be interpreted to reflect siderite formation in equilibrium with atmospheric CO2. However, based on the bacterial micro-textures observed under high magnification as well as on literature data (e.g. Curtis et al. 1986; Moore et al. 1992), this signature most likely reflects a mixture of different CO2 sources of which atmospheric CO2 could be one source. A depleted sulphate reduction CO2 type is less likely to be involved since these siderites occur in unit D where marine incursions less frequently occurred. This is also deduced from the virtual absence of framboidal pyrite. CO2 derived from bacterial fermentation could be involved, however this is not the only CO2 source since the carbon isotopic composition is less enriched in comparison to the + 15%o which
R. D R E E S E N E T AL.
68
Table 1. Oxygen and carbon isotope data of diagenetic and sedimentary components of the Eocene Sulcis Basin (FD: ferroan dolomite; NFD: non-ferroan dolomite)
013C
PART A
0180
Glaebules in claymatrix 374.00 Glaebules (pure)
-9.25 -9.29
-5.57 -5.35
Siderite (nodule centre) 359.9 Siderite (nodule edge) 358.90 Siderite nodule
+2.78 +2.38 -0.47
-7.98 -6.38 -4.97
400.00 Calcite vein 403.35 Calcite vein (with pyrite)
-4.51 -4.05
-10.46 - 11.16
403.35 400.00 394.80 394.80 390.40 383.20 377.60 363.38 355.70 355.40
-6.05 -2.30 -1.35 -1.25 -3.88 -2.67 -5.89 -4.45 -4.43 -3.59
-7.78 -9.82 -4.68 -5.29 -8.06 -7.13 -6.42 -7.10 -6.55 -6.42
+0.11 -8.34 +0.68 -8.82 -1.58 -1.81 -0.87 +0.11 -0.86 -0.70 -9.04 -10.96
-4.14 -7.95 -7.42 -7.62 -6.97 -7.44 -6.04 -5.62 -7.55 -8.56 -8.13 -8.33
-3.63 -4.16 -6.11 -4.40 -2.34 -3.67 -4.54 -3.44 -5.43 -5.78 6.69 -7.54 -3.05 -4.60
-6.13 -5.33 -6.44 -6.40 -4.21 -5.26 -7.62 -8.04 -7.01 -7.07 -8.34 -7.29 -7.57 -6.50
-3.30 -0.66 -1.87 -2.07
-7.01 -8.97 -5.10 -6.88
Ferroan limestone matrix Limestone matrix Bivalve shells + limestone (70% + 30%) Limestone around bioclasts Limestone matrix Limestone matrix Limestone components Limestone matrix Limestone matrix + < F D Limestone matrix, slightly ferroan
Ferroan dolomite matrix 360.30 Ferroan dolomite matrix 359.89 Ferroan dolomite matrix 357.70 Ferroan dolomite matrix (>80% FD) 355.40 Ferroan matrix between FD vein 355.00 Ferroan dolomite matrix (>90% FD) 355.00 Ferroan dolomite vein + 20% matrix 355.00 Ferroan dolomite vein (pure) 355.00 Ferroan dolomite vein (pure) 354.00 (Ferroan) dolomite vein 362.70 Non ferroan + ferroan dolomite vein 359.89 Non ferroan § ferroan dolomite vein PART B Limestone/dolomite matrix + FD 377.30 Dominantly FD+limestone/dolomite matrix 377.30 Limestone/dolomite + FD + NFD? 377.30 Dominantly FD + limestone/dolomite 377.10 Dominantly FD + matrix 377.10 FD + limestone § NFD? 375.29 Limestone matrix with FD vein 365.90 Dominantly limestone/dolomite matrix + FD 363.16 Dominantly limestone/dolomite matrix 362.70 Dominantly FD 361.75 Dominantly FD + dolomite matrix 358.90"Dominantly FD + limestone/dolomite matrix 358.25 Dominantly FD in organic rich mud 354.00 Dominantly FD + limestone/dolomite matrix ML-1 ML-2 ML-3 ML-4
Miliolitic Miliolitic Miliolitic Miliolitic
limestone limestone limestone limestone
(Perna (Perna (Perna (Perna
et et et et
al. al. al. al.
1994) 1994) 1994) 1994)
EOCENE SULCIS COAL BASIN
69 a13C
/
FERROAN DOLOMITE (GROUP 1)
SIDERITE T+2
_[+,
a180
-11
-10
-4
MILIOLITIC LIMESTONE
%
11"
.
-3
-2
-1
11=
CALCARENITE (CALCITE& DOLOMITEGRAINS)
A, 4, 9
CALCITEVEINS
FERROAN+ NON FERROAN DOLOMITE (GROUP2) ~
.~...- GI.AEBULES 9-10 9 ~= 9 A ~k 9 X 1~ 9
GLAEBULES SIDERITES CALCITEVEINS CALCARENITE FERROANDOLOMITE NON-FERROANDOLOMITE MILIOLITICLIMESTONE MILIOLITICLIMESTONESULClS BORENOLES MIXEDSAMPLES
Fig. 13. Plot of carbon and oxygen stable isotope data from the Produttivo Formation. is typical for the anaerobic (bacterial) carbonate reduction and fermentation processes (Irwin et al. 1977; Irwin 1980 and others). Ferroan dolomite veins as well as intensively ferroan dolomitized strata have been sampled at different stratigraphic levels within the Produttivo Formation. Their 013C-0180 plot into two distinct areas. Group I clusters around a 013C of - 1 +2%o and 9180 o f - 6 4 - 2 % 0 . The second group is characterized by depleted 013C values varying around -8.5 4- 0.5%o and 9180 values of -7.8 + 0.3%o. In fact this group clusters close to the values of the sampled non-ferroan dolomites II with even more depleted 013C values down to -10.96%o and with rather similar olSO-values. Whether both dolomite types are genetically
linked is not yet clear. It is therefore proposed that at least two post-compactional dolomitization stages should be differentiated. Both could be related to circulation of meteoric water, however their ferroan post-compactional and dolomitizing nature point to circulation of evolved fluids. According to their 013C signature depleted CO2 was not or only slightly involved in the group I ferroan dolomites, whilst in the group II ferroan + non ferroan dolomites, CO2 derived from decarboxylation reactions should be taken into consideration. The calcite II cements and veins are characterized by moderately depleted 013C values (-4.254-0.30%o) and highly depleted 9180 values (-10.80 + 0.40%o). Such depleted oxygen
70
R. DREESEN E T AL.
values are characteristic of high temperature fluids. However, fluid inclusion data are needed to correctly evaluate the significance of their isotopic signature.
Depositonal setting The Sulcis coals most probably originated as a response to rising sea level. They represent in fact the upper end member of a highstand system tract (Fig. 5). The coarse lag at the base of each clastic episode corresponds to a transgressive system tract followed by the intertidal-supratidal lower member of the highstand system tract. The lag is composed of reworked material, including various limestone intraclasts, silicified evaporites and dolomites. The latter possibly represent the only relicts of a lowstand system tract, which has been completely eroded and winnowed by hurricanes (?) before the next marine flooding event. A working model illustrating the successive stages in the development of the ideal parasequence of the Produttivo Formation is depicted in Fig. 14. It is interesting to note that the successive flooding events ('detrital episodes') within the Produttivo Formation apparently coincide with the main transgressive cycles (3th order cycles 2.3 to 2.6) of Haq et al. (1988) for the Lower Eocene (Ilerdian/Cuisian). Alternatively, the silicified evaporites have been eroded and reworked from emerged Palaezoic rocks (Cambro-Silurian basement) or Triassic rocks in the near hinterland and episodically swept into the fresh-water marsh. Within the Produttivo Formation two lithology types can be differentiated. Within the first type the allochthonous calci/doloarenites dominate. Here, particles clearly have been transported, however, most of them only over limited distances. The second autochthonous
lithology-type is dominated by the pure palustrine/lacustrine carbonates. Most of the coal layers do not occur in situ but limited transport is indicated (hypautochthonous coals). The absence of in situ evaporites, the limited development of microkarst and desiccation cracks and the fact that the successions are capped with coal horizons point towards sub-humid climatic conditions (Platt & Wright 1992). Root development within the autochthonous carbonates is extensive. The limited pedogenetic features, such as glaebules and 'micro-nodular' structures indicate occasional emergence but they become more frequent in the upper part of the Produttivo Formation. Well developed calcretes, terrestrial gastropods and desiccation breccias which reflect prolonged exposure are absent. All these features point towards a complex environment of marginal marine, brackish and fresh-water settings. The Florida Everglades provides a potentially useful modern analogue. This area is characterized by intensely vegetated fresh-water marshland, swamp and fresh-water to brackish lagoons. The coastal mangrove region may provide an analogue for the Sulcis coal swamp. The Everglades has a very low topographic gradient and minimal relief (<5m) where large areas show dense fresh-water wetland vegetation (marsh, prairie and swamp forest) (Olmsted & Loope 1984). Areally the saw grass marsh is the most important environment and also the most important peat-forming environment in southern Florida. The filtering effect of this vegetation possibly explains the pure nature of the autochthonous carbonates. Lakes, forming areas of algal carbonate sedimentation also occur, however, they only make up a small proportion of the total area. Wide marshy zones occur in the lower lying areas forming ponds and broad, low-gradient vegetated or open channels termed 'sloughs'. Water
Fig. 14. Idealized block diagram (not to scale) and successive stages in the development of an ideal parasequence in the Produttivo Formation (Sulcis Basin, Early Tertiary). 1: LST (lowstand system tract): during a fall of the relative sea level, several small areas became emerged. Marginal marine facies (with oncoids and stromatolite mats) as well as evaporitic ponds developed during periods of extreme desiccation (dolomite, anhydrite, gypsum), whereas the evaporite sediments were affected by early diagenetic silicification. The latter ponds might have been fed by intermittent water supply (either ephemeral streams or marine incursions). Palustrine facies extended offshore producing an intermittent marshy vegetation (rooting). 2: TST (transgressive system tract): a rise of the sea level (vertical arrow) induced a reworking of all the inshore sediments, and triggered their transport (large horizontal arrow) and accumulation in a more landward position. The reworked sediments cover (displaying a fining upward trend) consists of residual dolomite (partial dissolution), calcite and limestone clasts, silicified evaporites, oncoids, fragmented stromatolitic crusts and peat clasts. 3: HST-TF (highstand system tract - tidal flats): subsequent tidal conditions resulted in the temporary development of subtidal, intertidal and supratidal subenvironments. The (channelized) tidal sands prograded somewhat farther seaward (large horizontal arrow) thanks to the available accomodation space. 4: HST (highstand system tract): extensive supratidal conditions, a relatively high water table in the area and a lack of detrital input lead to the development of extensive lacustrinepalustrine facies with hypautochtonous peat formation and fresh water carbonates.
EOCENE SULCIS COAL BASIN
71
72
R. DREESEN E T A L .
drainage occurs mainly along these sloughs, but water flow is sluggish. A similar type of'sloughs' may have existed in Eocene time in the Sulcis area. These depressions then could have been filled by allochthonous sediments. Consequently these sandbodies may not correspond to channel fills of braided rivers as previously described (Cocozza et al. 1989; Assorgia et al. 1992; Fadda et al. 1994). Thin (millimetre thick) evaporite layers actively precipitate within the surficial desiccated algal crusts in the central areas of the mud banks in Florida Bay. Precipitation of thin evaporites and peat formation thus occur simultaneously in the Florida setting today, but they form in different sedimentary environments. However, the evaporites are not silicified and they would most likely be destroyed if they were transported. If the evaporites were silicified it is possible that they could be ripped up by tropical storms and hurricanes and transported and deposited along the coastal levee in the coastal swamp and in the sloughs. This may explain the irregular distribution of the sandbodies shown in Fig. 9. The marl layers or lenses found below the saw grass peat in the open marshes of the Florida Everglades may represent a modern analogue for the fresh water carbonates of the Eocene Sulcis coals. The calcium contained in the surficial waters was precipitated, either by lime secreting algae (algal mats) or by the combined effect of all herbaceous plants present on the carbon dioxide content of the flowing surface water (Spackman et al. 1964). The precipitation of marl as opposed to peat reflects a rather delicate balance between the rate of subsidence (or differential peat compaction or ground water table rise) and the rate of peat accumulation. If the rate of subsidence (or water table rise) is too fast, the saw grass peat surface will soon be covered by an algal mat producing a layer of marl. A few centimetres of difference in water depth is all that appears to separate the two environments but the size of the areas involved is often to be measured in terms of square kilometres (Spackman et al. 1964). This development will continue until the area has been built up to a level that would permit repopulation by saw grasses. Although the hypothesis of reworked Cambrian allochems from the hinterland is not entirely dismissed, and euhedral quartzes might well have been derived from unroofed silicified Triassic evaporites, we still favour the idea of coeval evaporitic lacustrine and palustrine carbonate facies in the Eocene Sulcis Basin. The reworked calcite allochems are probably not
Cambrian in origin as the Cambrian limestones were generally too strongly recrystallized and homogenized (showing twinned and bent crystals; Maria Boni, personal communication). The hypothesis of silicified evaporites reworked from remote Triassic outcrops is not plausible because of the freshness and the size of the quartz bipyramids, excluding transport over larger distances. A nearby Triassic source area is also unlikely: if quartz bipyramids were supplied as clasts to the Sulcis basin, why are they totally lacking in the Miliolitic Limestone Formation? The idea of a synsedimentary origin for the allochems is supported by the presence of gypsum pseudomorphs within the porostromate oncoids. Also the complex cement characteristics within these algal mats could, after reworking, generate the 'sparite' particles. Finally, the nearly non-abraded aspect of the euhedral to subhedral quartz grains replacing evaporites also supports a very limited transport. Periods of extreme desiccation, leading to the establishment of hypersaline conditions within the lacustrine ponds (evaporitic carbonate flats) may have led to evaporite precipitation, associated with diagenetic silicification and dolomitization. Subsequently, bypramidal evaporitebearing quartz and dolomite grains could have easily been reworked by episodic high-energy events and redistributed as clasts. Diagenetic silicification of coeval evaporitic lacustrine carbonate facies might have taken place by leaching of biogenic siliceous mineral matter from the peats, a phenomenon already reported from the Okefenokee Swamp and the Everglades (Cohen et al. 1987; Andrejko et al. 1983; Cohen 1995). The biogenic silica commonly consists of sponge spicules and plant phytoliths. Siliceous sponge spicules have frequently been observed in the calci-dolorudites and -arenites of the Produttivo Formation. Silicophytoliths (opaline silica) are commonly produced by sedges and grasses (Andrejko et al. 1983; Cohen 1995): the latter represented a dominant plant community in the Eocene Sulcis peats. Biological degradation processes play a significant role in silica dissolution and mobility in a peat-forming environment. The apparent lack of observed spicules and the occurrence of chert lenses in Tertiary lignites deposited under similar conditions to the Ekofenokee peats might be a function of such bioerosional activities (Andrejko et al. 1983). The high SIO2/A1203 (%) ratios (average 1.95) of the Sulcis coal ashes, suggest an excess in silica possibly related to authigenic biogenic silica (the dominant clay mineral being kaolinite with a Si/A1 ratio of about 1; Szymanski et al. 1990).
EOCENE SULCIS COAL BASIN Facies assemblages characteristic of fluctuating, evaporitic and non-evaporitic lacustrine episodes have been recorded in the Tertiary Narbonne Basin from the northern Pyrenean foreland in southeastern France (Szulc et al. 1991). Silicified evaporite horizons have been recorded in the basal Cretaceous palustrine deposits of the Rupelo Formation in North Spain (Platt 1989). Minor lenticular gypsum and nodular anhydrite have been described from the carbonate dominated coal-bearing lacustrine facies of the Oligocene Calaf and Mequinenza sequences in the Ebro Basin, NE Spain (Cabrera & Saez 1987). Here, paludal carbonate marshes or swamps and evaporitic carbonate flats developed in the outer lacustrine fringes at times of lowered water level. Similarly, evaporites have been recorded from shallow carbonate lacustrine facies in the Permian of the Aragon-Bearn Basin (Pyrenees) (Garc6s & Aguilar 1992): pseudomorphs after gypsum and evaporite replacement textures formed in restricted or littoral palustrine areas during low stand lake levels.
Conclusion Detailed sedimentological-petrographical analysis of cores from autochthonous and allochthonous lithology types of the Produttivo Formation, results in a revised depositional model for the Eocene coal-bearing strata of the Sulcis Basin in Southwest Sardinia. The Florida Everglades are considered as a modern analogue of the Produttivo Formation: a tropicalsubtropical coastal marsh, adjacent to restricted lagoons with low tidal fluctuations. Mangroves lining the muddy coastlines allowed grass marshes to expand behind them and formed an efficient trap for sediments carried by the flood tide. After initially prevailing restricted marine to brackish water conditions (topmost Miliolitic Limestone, basalmost Produttivo Fro), freshwater marshes gradually developed landward of the tidal influence, giving rise to paludal and lacustrine conditions. The marine influence left its mark in the relatively high pyrite content of the coals. Episodically these marshes became invaded by the sea, possibly as a result of regional subsidence related to extensional tectonics. The latter marine transgressions in turn (re)initiated peat formation by inland migration of the marsh belt, in front of the invading sea. This is reflected by the upward shoaling marine clastics (tidal sand bodies) and by the palustrine conditions returning at the end of each clastic sequence. The abundance of evaporite-bearing euhedral to subhedral quartz
73
grains and that of dolomite cement clasts in the latter deposits suggests reworking of evaporitic facies. Petrographic evidence favours the idea of a synsedimentary origin for these particles, although reworking of silicified evaporite-bearing Cambrian or Triassic rocks from the nearby hinterland cannot be totally rejected. This working model implies periods of extreme desiccation during the Eocene, leading to the establishment of evaporitic conditions within the lacustrine ponds of the Sulcis Basin marshes. Palustrine carbonate marshes and evaporite carbonate flats have been described from elsewhere in the stratigraphic record, and simultaneous evaporite precipitation, fresh water carbonate deposition and peat formation can actually be observed in the Florida Everglades, which is believed to be the most appropriate modern analogue. This study has been conducted with the financial support of the European Commission (Convention E.C.S.C. n~ 7220-AF/214). We thank It. Marco Slavik and Ir. Gianfranco Manconi (Carbosulcis S.p.A., Cortoghiana) for access to data. We are most indebted to MM. Giorgio Sardu and Giancarlo Contini (Carbosulcis S.p.A.) for help with sampling and description of the cores, and to MM. Marc Bauwens, William Vaesen and Pierre Blavier (I.S.Se.P., Li6ge) for assistance with thin section preparation, figures and photographs. Dr. Romeo Flores (USGS, Denver, Colorado) is thanked for the stimulating discussions on the origin of the clastic sand bodies. A special note of thanks goes to Paul Guion (Oxford), Michael Dusar (Brussels) and Stanislav Oplustil (Prague) for their careful reading, constructive criticism and helpful suggestions, improving the content of the original manuscript. Finally, we thank Maria Boni (Naples) for the information on the Cambrian carbonates of SW-Sardinia.
References ANDREJKO, M. J., RAYMOND, R. & COHEN, A. D. 1983. Biogenic silica in peats: possible sources for chertification in lignites. In." RAYMOND,R. Jr. & ANDREJKO,M. J. (eds). Proceedings of the Workshop on Mineral Matter in Peat." its Occurrence, Form and Distribution, September 26-30, 1983,
Los Alamos National Laboratory, Los Alamos, New Mexico, 25-37. ARBEY,F. 1980. Les formes de la silice et l'identification des bvaporites dans les formations silicifi~es (Silica forms and evaporite identification in cherts). Bulletin des Centres de Recherches ExplorationProduction Elf-Aquitaine, 41,309-365.
ARNOTT, R. W. C. 1995. The parasequence definition-are transgressive deposits inadequately addressed? Journal of Sedimentary Research, 65, 1-6.
74
R. D R E E S E N E T AL.
ASSORGIA, A., BARCA, S., COCOZZA, T., DECANDIA, F. A., FADDA, A., GANDIN, A. & OTELLI, t . 1992. Characters of the Caenozoic sedimentary and volcanic succession of Western Sulcis (SW Sardinia). In: CARMIGNANI, L. & SASSI, F. P. (eds). Contributions to the Geology of Italy with special regard to the Palaeozoic Basements. IGCP N ~ 276, Newsletter, 5, 17-20. BANDELOW, F.-K. & GANGEL, L. 1993. Die terti~ire Braunkohlenlagerst/itte yon Monte Sinni im Sulcisbecken (Sardinien). Zeitschritt fur Angewandte Geologie, 39, 2, 90-95. BECHSTADT, T. & BONI, M. 1989. Tectonic control on the formation of a carbonate platform: The Cambrian of Southwestern Sardinia. In. Controls on Carbonate Platform and Basin Development. Special Publication of the Society of Economic Paleontologists and Mineralogists, 44, 107-122. CABRERA, L. & SAEZ, A. 1987. Coal deposition in carbonate-rich shallow lacustrine systems: the Calaf and Mequinenza sequences (Oligocene, eastern Ebro Basin, NE Spain). Journal of the Geological Society, London, 144, 451-461. CHERCHI, A. 1983. Presenza di Ilerdiano a Alveolinid a e e Orbitolitidae nel bacino paleogenico del Sulcis (Sardegna SW). Bollettino della Societgl Sarda Scienze Naturali, 22, 107-119. CERLING, T. E. 1991. Carbon dioxide in the atmosphere. Evidence from Cenozoic and Mesozoic paleosols. American Journal of Science, 291, 377-400. COCOZZA, T., DECANDIA, F. A. & GANDIN, A. 1989. Studio geologico, stratigrafico e paleogeografico del bacino carbonifero del Sulcis, nel programma di ricerche minerarie di base. Convenzione Societa Carbosulcis e Universita di Siena. Rapporto Interno Carbosulcis. COHEN, A. D. 1995. Peat deposits of the Ekofenokee Swamp: their significance in interpreting the characteristics of coal seams. A Field Guidebook for Amoco Production Company's Coal ResourceRock Field Seminar, May 1995, 128-133. - - , SPACKMAN,W. & RAYMOND, R. 1987. Interpreting the characteristics of coal seams from chemical, physical and petrographic studies of peat deposits. In: SCOTT, A. C. (ed.) Coal and Coalbearing Strata: Recent Advances. Geological Society, London, Special Publication, 32, 107-126. CRAIG, H. 1957. Isotopic standards for carbon and oxygen correction factors for mass spectrometric analysis of carbon dioxide. Geochimica et Cosmochimica Acta, 12, 133-149. CURTIS, C. D., COLEMAN, M. L. & LOVE, L. G. 1986. Pore water evolution during sediment burial from isotopic and mineral chemistry of calcite, dolomite and siderite concretions. Geochimica et Cosmochimica Acta, 50, 2331-2334. DEBELMAS, J. 1974. G~ologie de France, 2, 307. DICKSON, J. A. D. 1966. Carbonate identification and genesis as revealed by staining. Journal of Sedimentary Petrology, 36, 491-505. FADDA, A., OTTELI, L. & PERNA, G. 1994. I1 Bacino Carbonifero del Sulcis. Geologia, Idrogeologia, Miniere. Carbosulcis S.p.A., Cagliari.
FOLK, R. L. & PITTMAN, J. S. 1971. Length-slow chalcedony: a new testament for vanished evaporites. Journal of Sedimentary Petrology, 41, 1045-1058. FRIEDMAN, G. M. 1959. Identification of carbonate minerals by staining methods. Journal of Sedimentary Petrology, 29, 87-97. GARCES, B. L. & AGUILAR, J. G. 1992. Shallow carbonate lacustrine facies models in the Permian of the Aragon-Bearn Basin (Western spanishfrench Pyrenees). Carbonates and Evaporites, 7, 94-107. GILES, M. R. & MARSHALL,J. D. 1986. Constraints on the development of secondary porosity in the subsurface: re-evaluation of processes. Marine and Petroleum Geology, 3, 243-255. HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level changes. Society of Economic Paleontologists and Sedimentologists, Special Publication, 42, 71-108. IRWIN,H. 1980. Earlydiageneticcarbonateprecipitation and pore fluid migration in the Kimmeridge Clay of Dorset, England. Sedimentology, 27, 577-591. --, CURTIS, C. & COLEMAN, M. 1977. Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, 269, 209-213. JALFREZO, M. 1977. Pyr6n6es-Orientales, Corbi&es. Guides G~ologiques R~gionaux, Masson, 42. KATZ, A. & FRIEDMAN,G. M. 1965. The preparation of stained acetate peels for the study of carbonate rocks. Journal of Sedimentary Petrology, 35, 248-249. MILLIKEN, K. L. 1979. The silicified evaporite syndrome - two aspects of silicification history of former evaporite nodules from Southern Kentucky and Northern Tennessee. Journal of Sedimentary Petrology, 49, 245-256. MOORE, S. E., FERRELL, R. E. & AHARON, P. 1992. Diagenetic siderite and other ferroan carbonates in a modern subsiding marsh sequence. Journal of Sedimentary Geology, 62, 357-366. OLMSTED, I. & LOOPE, L. L. 1984. Plant communities in Everglades National Park. In" GLEASON, P. J. (ed.) Environments of South Florida Present and Past II: Miami Geological Society, Coral Gables, FL, 167-184. ORSINI, J. l . , COULON, C. • COCOZZA, T. 1980. D6rive c6nozoique de la Corse et de la Sardaigne et ses marqueurs g6ologiques. Geologie en M(jnbouw, 59, 385-396. PERNA, G., TURI, B. & VESICA,P. 1994. Le calcite delle cavitfi carsiche del Calcare Miliolitico. In: FADDA, A., OTTELLI, L. & PERNA, G. Il Bacino Carbonifero del Sulcis. Geologia, Idrogeologia, Miniere. Carbosulcis, Cagliari. PITTAU, P. 1977. Palynological investigation of the Lower Tertiary Sardinia coal layers. Bollettino della Societd Geologica Italiana, 93, 937-943. PITTAU DEMELIA, P. 1979. Palinologia e datazione della sezione di Tanca Aru nelle Valle del Cixerri (Sardegna sudoccidentale). Bollettino della Societh Paleontologica Italiana, 18, 303-314.
E O C E N E SULCIS COAL BASIN PLATT, N. H. 1989. Lacustrine carbonates and pedogenesis: sedimentology and origin of palustrine deposits from the Early Cretaceous Rupelo Formation, W. Cameros Basin, N Spain. Sedimentology, 36, 665-684. - & WRIGHT, V. P. 1992. Palustrine carbonates and the Florida Everglades: towards an exposure index for the fresh-water environment. Journal of Sedimentary Petrology, 62, 6 1058-1071. PLAZIAT, J. C. 1981. Late Cretaceous to Late Eocene palaeogeographic evolution of Southwest Europe. Palaeogeography, Palaeoclimatology, Palaeoecology, 36, 263-320. RIMIN, 1990--1991. Descrizione, interpretazione e correlazione relative al 'Produttivo' dei sondaggi a carotaggio continuo della miniera di M.Sini. Relazione conclusiva sui lavori svolti. Internal report Carbosulcis S.p.A. SALOMONS, W~ GOUDIE, A. & MOOK, W. G. 1978. Isotopic composition of calcrete deposits from Europe, Africa and India. Earth Surficial Process, 3, 43-57. SALVADORI, A. 1979. II Bacino Paleogenico del Sulcis. Ricerca di correlazioni biostratigrafiche. PhD thesis, Dipartimento Scienza della Terra, Univ. Cagliari, Sardinia. SCHLEDDING, T. 1985. Fazies, Geochemie und Palgiogeographie der unter- bis mittel-kambrischen Gonnesa Formation und der basalen Cabitza Formation des Sulcis (SW-Sardinien, Italien): Der Zerfall einer Karbonatplatform. PhD thesis, University of Freiburg (Germany). SHACKLETON, N. J. 1986. Paleogene stable isotope events. Palaeogeography, Palaeoclimatology, Palaeoecology, 57, 91-102. SPACKMAN, W., SCHOLL, D. W. & TAFT, W. H. 1964. Field Guidebook to the Environments of Coal in Southern Florida. Geological Society of America. Pre-Convention Fieldtrip, November 16-18, 1964. SWENNEN, R. & VIAENE, W. 1986. Occurrence of pseudomorphosed anhydrite nodules in the Lower Visean (lower Moliniacian) of the Verviers Synclinorium, E-Belgium. Bulletin de la Socidtd belge de Gdologie, 95, 89-99. SWETT, K. 1965. Dolomitization, silicification and calcitization patterns in Cambro-Ordovician oolites from Northwest Scotland. Journal of Sedimentary Petrology, 35, 928-938. SWIFT, D. J. P. 1975. Barrier-island genesis: evidence from the Central Atlantic shelf, eastern USA Sedimentary Geology, 14, 1-43.
75
SZULC, J., ROGER, Ph., MOULINE, M. P. & LENGUIN, M. 1991. Evolution of lacustrine systems in the Tertiary Narbonne Basin, northern Pyrenean foreland, southeast France. Special Publication of the International Association of Sedimentologists, 13, 279-290. SZYMANSKI, R., LOEBER, L. & DURAND, D. 1990. Local composition of reference clays as studied statistically by analytical scanning transmission electron microscopy. In: FARMER, V. C. & TARDY, V. (eds) Proceedings, 9th International Clay Conference, Strasbourg 1989. Sciences Gbologiques M~moire, 89, 149-158. TAMBAREAU, Y., FEIST, M., GRUAS-CAVAGNETTO, C. & MURRU, M. 1989. Caract&isation de l'Ilerdien continental dans le domaine ouest-mrditrrranren. Comptes Rendus de l'Aeaddmie des Sciences, Paris, 308, Srrie II, 689-695. TERWINDT, J. H. J. 1988. Palaeo-tidal reconstructions of inshore tidal depositional environments. In: DE BOER, P. L., VAN GELDER, A. & NIO, S. D. (eds) Tide-influenced Sedimentary Environments and Facies. Sedimentary and Petroleum Geology, Dordrecht, Reidel, 233-264. TUCKER, M. E. 1976. Replaced evaporites from the Late Precambrium of Finmark, Arctic Norway. Sedimentary Geology, 16, 193-204. THOREZ, J. GOEMAERE, E. & DREESEN, R. 1988. Tideand wave-influenced depositional environments in the Psammites du Condroz (Upper Famennian) in Belgium. In: DE BOER, P. L., VAN GELDER, A. & NIo, S. D. (eds) Tide-Influenced Sedimentary Environments and Facies. Sedimentary and Petroleum Geology, Reidel, Dordrecht, 389-415. VAN WAGONER, J. C . , POSAMENT1ER, I-I. W., MITCHUM, R. M., VAIL, P., SARG, J. F., LOUTIT, T. S. & HARDENBOL,J. 1988. An overview of the fundamentals of sequence stratigraphy and key definitions. In: W1LGUS, C. K., HASTINGS, B. S., KENDALL, C. G. St. C., POSAMENTIER, H. W., Ross, C. A. & VAN WAGONER, J. C. (eds) SeaLevel Changes: An Integrated Approach. Special Publication of the Society of Economic Paleontologists and Mineralogists, 42, 39-45. , CAMPION, K. M. & RAHMANIAN, V. D. '1990. Siliciclastic sequence stratigraphy in well logs, cores and outcrops. American Association of Petroleum Geologists, Methods in Exploration, 7.
Turkish lignite deposits H. I N A N E R
& E. N A K O M A N
Faculty of Engoleering, Department of Geology, Bornova, Izmir, Turkey Abstract: The distribution of lignite deposits in Turkey is such that in general the Eocene lignites are in northern Turkey, Oligocene lignites are in northwestern Turkey, Miocene lignites are in western Turkey, and Pliocene-Pleistocene lignites are in eastern Turkey. Only the Oligocene lignites are paralic deposits, the rest being formed in a limnic environment. Turkey has about 8.4 Gt of lignite reserves of which 3.9 Gt are the exploitable reserves. Most of the known lignite deposits in Turkey are of low calorific value and have high contents of ash, moisture and total sulphur. Almost 80% of the total reserves have calorific values below 2500 kcalkg-1. The lignites having low calorific values are generally consumed in power plants. The lignites having relatively high calorific values are exploited for domestic and industrial use in the country. The majority of Turkish lignite deposits are worked in open-pit mines, but there are also some underground operations. In this article general information will be given about the stratigraphies, reserves, qualities and mining methods for the major lignites deposits of Turkey. These deposits have been explored and evaluated by the General Directorate of Mineral Research and Exploration (MTA), and exploited using various mining methods by the Turkish Coal Enterprises (TK|). Coal exploration is ongoing with geological mapping and drillings in the several lignite fields. Generally, large deposits are mined by T K i and low reserve (small) deposits are mined by private enterprises. According to recent studies (MTA 1993), there are 181 well explored lignite deposits and 98 lignite deposits which require further investigation. There is very little information about many of the privately small lignite deposits. Up to now, 8.4Gt of lignite reserves have been
_._-, k /~Demirhanh ~)~,6...*..v)'e
>
.
/
M
a
e L A c K
_
_
BURSA 9
9 De l~lrpllar
i
l
a
13uk
m
~
General information about lignite deposits of Turkey The distribution of lignite deposits in Turkey is such that generally the Eocene lignites are in northern Turkey, Oligocene lignites are in northwestern Turkey, Miocene lignites are in western Turkey, and Pliocene-Pleistocene lignites are in eastern Turkey (Fig. 1). Only the Oligocene lignites were formed in a paralic environment. The rest of the Turkish lignite deposits were deposited in a limnic environment.
,,
~
~
determined in the whole of Turkey. Some 115 000 km 2 have been explored, 40 000 km 2 have been investigated in detail, and more than 1.1 Mm have been drilled during evaluation.
~_ A
~
V
A
~ir
AI
ikorK~l 9K 9
) ) /'~
s E A
9
Refahffe o
-
"~"a
"Ollu
9 o
._./.~.
~.
,_,
Homs~an ERZURUM
~.~ j
....
t~"
YOZG,AT5o(" url
~l~lll
~2
_V~AN~oVAN
. . . . . . ;..; % <~ ~ S,~keeAYOiN L~. . . . ~ V ~ oKONYA I v~-~'I v ~ \ ki in ~ahinali i ~ B~ROUR ]/JBeysehir i Hale . . .~. .' ~ 9 Vataw ~ HARAMAN
Tufonbeyfi ADIyAMAN KAHRAMANMARA~ ~ O o C~l~si
~-.-v"
o ITERRANE
Fig. 1. Distribution of lignite deposits in Turkey.
From Gayer, R. & Pe~ek, J. (eds), 1997, European CoalGeologyand Technology, Geological Society Special Publication No. 125, pp. 77-99.
/'-.. r
/.~-'~..-J
/I l'~"
~
Er--~ek
I
II E o c e n e L i g n i t e s 9 Oligocene Lignites 'P
Pliocene
Lignites
"
) (-~
9"'-'1 t . ./" "v"
-'~ .~
v,
b
0
o
~o5~
~ ~
~
OZ
Z~
o +
e~
"c
N=
C,'~ ~ . ~
C',.1
~
r
.
~tNI
~
"~" .,.,~
,,.a
m
0a
aa
,o '-'
~"~i
C"
~ z ~o5~
~
Z
Z
Z
Z
Z
0
0
Z
Z
e~
~
~
o
t~
w167167167167
,"~,"~,"~
C'4C'4C",1
t'~
I ~-~
~2
N ~N
~
~
~
t~ -
~
D
80
H. iNANER & E. NAKOMAN
Eocene lignites
the locations suitable for coal deposition. These are the deposits of: Mengen, Merke~ler, G6yniik, Sorgun, (Teltek and Tosya (Fig. 1). Eocene lignites were deposited in closed basins with a basement generally made up of the Palaeozoic or Mesozoic rocks. Coal seams are developed usually between marls. These coals are laterally restricted and usually not thick. There are one or two coal seams 0.90-6.00m thick. Lignites are bright, brittle and have good
Eocene lignites are located in very limited areas in Turkey, because marine conditions prevailed until the end of the Oligocene over much of the country. Coal formation occurred in only small islands in the Mesogene Sea. Eocene lignites occur in northern Turkey parallel to the boundaries facing Anotalia within the Pontide system which was unaffected by tectonic movements, at
j . / Oemirhanh
"~" 2
,,.....,~'__ O. _ OKIRKLAREli \ 9 ",~ EDIRNE V~e k
\
~
f "J J
9
\j"
f"
)
BLACK
0 Saray
~
SEA ZON GULOAK
Edir k~ ~ $ a f ~ , i a r ~ " Hask~y TEKIRDAr ~
9 ,,AlOMalkara~
Ahmctpasa
J
5TANBUL
0 Mengen
~"~
~
,T
n _
_
OMerke~ler
_
MARMARA SEA
|
BO. A
E)
0 ~ lrpl l ar
'~
,,,
~
AQ 0 Beypazam
Sa~,rlar _murmu Devecikona#l ~ - i v i l i
Or
B9 9Altda rnar
Kele$
BA LIKESiR Hamzaok 9 Harmanalan O "~_ 9 Davutlar 9Oomaniq Cak.r~oursunbey O TUncbilek . TLi'kpiyal, Tav.~nh 0 Seyit6mer 9 ~Kozlui~ren Cinge t'-~- neni~ I | ""t~ ~-.--Beni~. 11 KUTAHYA ua'ar - - e _ ~ e . ~ Evci,er =rhala I.~ktar Soma Ey~ez Merkez 0
O B a h crek ~y
Gediz
d 0 MANiSA
0 Hararnikby
:::C:;bIZMiR
9KurucjOI I/gin
z
,O'keO
AYDIN O
DENiZLi
O~ahinali
~
Q
BcyTechir G. ~'~ Avdanclk
J
~
KONYA E)
9
Beysehir ~
~Ka'-
Eskihi asa~_ p'~Jrcjut
l,~ M
itas,__._.-'~_, o vataOa,,
.. v, _. v ~ S y . ~~ Karacannsar.
Bayw
Eki zkOy
~" ~
~ " L ~~/~---'
....
. ', I~1 STUDY ~:1 AREA
kq alar
....
~ M ' '~=' -
~ ~4ralan~~"~
~__-~/~ ...........
Karadik'n 9
M
E 0
f~ I
~ T
Fig. 2. Lignite deposits and sectors of Western Turkey.
E
R
R
A
,)
N
'-1 0 D e p o s i t !e S e c t o r E A
N
I
I
o
Zsk
"
TURKISH LIGNITE DEPOSITS quality. Some of deposits have a high total sulphur content (between 1.25% and 9.60%). Reserves are small (Table 1).
Oligocene lignites The Oligocene paralic lignites are generally found in northwest Turkey. Among these deposits are: Saray, Malkara, Demirhanh, Harmanh, ile, and Vize. There are some other small lignite deposits at Kale in southwest Turkey and Sereflikoqhisar central Turkey which probably formed on islands (Fig. 1). Eustatic sea level changes took place during the Oligocene, resulting in the formation of these paralic lignite deposits. Coal seams are numerous but thin in these deposits. Reserves are higher than the Eocene lignites, but the quality is poorer (Table 2).
Miocene lignites The Miocene lignites are generally located in fault bounded basins in western Turkey (Fig. 1).
Extensional tectonics started in the Middle Miocene and still continues in this area. The age of lignites is overwhelmingly the Middle Miocene (Akgtin & Akyol 1992). Intensive flora cover which grew in the tropical-subtropical climatic conditions resulted in the accumulation of thick peat deposits in structually active limnic basins. The flora cover became poorer with the cooler climatic conditions from the Late Miocene up to the end of Pliocene (Akgfin & Akyol 1992). Because of this, there are very few Late Miocene lignites in western Turkey. Some of the most important Miocene lignites are: Muf~la region (Yata~an-Milas) deposits, Soma, Beypazan, Seyit6mer, Tunqbilek, Orhanell, Keles, Devecikona~l, ~an, ~lrpdar, Sahinali, S6ke, Gediz, Dursunbey, Alpagut, Oltu, A~kale, Bahqek6y, Kemaliye, and Ermenek (Table 3). Miocene lignites have very high reserve capacities. These lignite-bearing formations lie unconformably on basement rocks of the Palaeozoic and Mesozoic. Lignite seams are underlain by a fining-upward sequence of conglomerates, sandstones, and siltstones, and are overlain by marls, limestones, tuffites and also
~..~_....~._j~s,,oP KASI"AMONU
~ ~
~
~
o
~
B&lkaya
"-
,
0
0
0
Ayvak~r |
|
ANKARA |
5orgun | O ~ZGAT
Baheeh~y o
SiVAS Q
Kangal
Kemaliye
| KAu
,
~...,..~..~,\
O
i ~.~'
Erci~
Karhova
o
| BiNGOL
o
L
AN
|
enek
/ ADANA
.-'"
Tp t
.~ !
,.. .~. "~ 9 /'~" .v-.~..,.j ~ . / r \
p ADIYAMAN
KAHRA'NANM ARA,~ KARAMAN | ' "
t.- 3
~-.. -!
Afsin- Elbistan 0 0 Tufanbeyll Go'lba~
KONYA O
"\.,,
",~
0
et eflikochi~r
~.~
~r .~
ERZURUM O ERZ~NCAN
Hamal ~ Etyemez Kalbur'~aym
~
o
o HOrasan
0
Refahiye O
CORUM
K.-. ~
9 5utkans ..
Askale
Q
CANKIRI
otP~ e
/sp,r O
Tosya Alpagut Cettek
Orta
81
o Province o Deposit 9 Sector
!
0
Z I ~
Fig, 3. Lignite deposits and sectors of Eastern Turkey.
25 krn,
STUDY AREA
,,..a
o= , ~ o
b
~
..~Z
~05~
0
~0~
~0
D~O
~ 0 ~
0
~ 0 0
~Z
0
o
goo
~o o
~ g o
o g o o o
o
o g o o
go
o
~ o ~ 0
o o o o 0 0 0 ~
o O 0
.0
+
~ ~+ ~,~ ooo
o
0 0 0 0
ooo O 0
ooo
o
0 0 0 0 0 0
o ooo 0 0 0 0 0
~C
o~
o~
o~g~
0"~ ~
r~
~o
~ o ~
~
~
~
~
~
~ ~ - ~
r/J
u
8~
~_.~
~.~ ,
,.~
>
:~ ,.~ ,.~
c~ c . ~
Z
,-1
.=o
..
s
ZZ
0
~
~gg
googoo
0
0
0
o~
oo~176go~
~ZZZ
9
0
~Z
o~
o~
o~
g I
~
m~
+
~
O~
Io~
O~
o~ ~-~
0'.~ .~
<,~
~3
.~ o
.
m
~
.
~
.
mo
o o o
. o o ~
~
0
~
~a 0
O~
0
0
I~ .
~.~0~
O~
~
9
9169
+
<,._.
,x
8~
~
gggg
gggogoo
~ O~
.
.
.
~lr~
o
.
D
o~176
.
0
',~
,
TURKISH LIGNITE DEPOSITS recent alluvium. Although there is a great variation in the lignitic quality of the region, (between 1335 kcal kg- 1 and 4500 kcal kg- 1) the Miocene aged lignites are generally good in quality. Lignite seams are few, usually one or two and rarely three, and their mineable thickness is 1.00-25.00metres. The majority of these deposits are worked as open pit mines, although there are also underground mines in the region. The lignite reserves of western Turkey.make up 33% of the total Turkish reserves (Inaner & Nakoman 1993).
Pliocene lignites The Pliocene lignites are generally located in eastern Turkey (Fig. 1), where subtropical microclimatics enabled some peat to accumulate in limnic basins. The Pliocene lignite deposits are: Af~in-Elbistan, Bey~ehir, Ilgan, Kangal, Orta, G61ba~l, Karhova, Tufanbeyli, Erci~, Ispir, Horasan, and Refahiye. These are limnic deposits. The basement is generally Palaeozic or Mesozoic limestones, ophiolotic rocks, recrystallized limestones, metamorphic schists and serpentinites. Coal seams are very few, usually one or two. Mineable seams are thick (between 1.00 m and 39.58 m) but there are sterile partings within the seams. Most of these lignites have high moisture and ash contents (25.31-53.42% and 11.38-36.37%), and low calorific values (1083-2239 kcal kg-1). The average calorific values are about 1000kcalkg -1. However, these lignites have usually large reserves (Table 4). These lignites are generally consumed in power plants. Deposits such as Ispir, Erci~ and Refahiye which have low reserves are used for regional domestic heating.
Descriptions of selected Turkish lignite deposits The following descriptions include examples with higher reserves from the Eocene, Oligocene, Miocene, and Pliocene.
Mengen deposit This Eocene deposit is situated in the northeast of Bolu Province in northwest Turkey. The basement is made up of Palaeozoic metamorphic and igneous rocks, and Mesozoic sandstones and
85
recrystallized limestones. The Paleocene is represented by marls and limestones. The Eocene is divided into two, with the ipresian made up of sandy and marine limestones, and sandstones and the Lutecian, comprising seven units (Kaya & Dizer 1984). The Lutecian starts with fossiliferous limestones at the base, followed by Marls and limestones. Stratigraphically the lignite bearing bituminous shale layers are found near the base. The upper levels are made up of smelly limestones, highly fossiliferous marl-limestones and platy marls, whereas the Upper Eocene is represented by agglomerates, conglomerates and sandstones. These rocks have a Quaternary cover of slope debris (Fig. 4). There are two coal seams, an upper and lower, with an average thickness of 5.00 m. The percentage of average moisture, ash, and total sulphur in the original coal are 9.70% 20.98% and 9.25% respectively and a calorific value of 4755 kcalkg -1. The proven, probable, possible, and total reserves are; 23539760; 39736697; 14690000; and 78016350 tonnes respectively (Table 1). It is worked by longwall mining.
Malkara deposit The Oligocene Malkara coal deposits are located in the northwest of Tekirda~ Province in European Turkey (Fig. 1). The sectors are Ahmetpa~a and Hask6y (Fig. 2). Thirteen coal seams are developed in a sandstone-marl series, but only three seams are mineable (Fig. 5). The coal has an average quality (Table 2), with contents of moisture, ash, and total sulphur in the original coal of 25.30% 33.29% and 1.29% respectively, and calorific value varying between 2014 to 2317kcalkg -~. The total reserves are 11.595357 • 10 6 tonnes. The coals are worked by both open-pit and underground mining. There is currently no production in the area.
Mu~la region (Yata~an-Milas) deposits Lignite-bearing Neogene sediments cover large areas in the districts of Yata~an and Milas situated in the province of Mu~la in southwest Turkey (Fig. 1). These deposits are Turgut, Eskihisar, Bayar, Ba~yaka, Tlnaz, Karacahisar, Sekk6y, Ekizk6y, (~aklralan, Htisamlar and Alatepe (Fig. 2). Stratigraphically these 11 deposits are very similar and can be summarized together: The Kerme Formation lies unconformably on a basement of Palaeozoic gneisses and schists,
o
~'~ ~ ,, ~o
~
N
o
~z
~o5~
0
Z
OZ
9
0
Z
Z
ZZZ
Z~
Z
0 0 0
0 0 0
O 0 O 0 O 0
~ 0 ~ 0 ~ 0
0 0 0
0 0 0
0 0 0
0 0 0 0 0 0 0 0 0
O 0 ~ 0 ~ 0
0 0 0
O 0 O 0 O 0
0
0
0 0 0
~ 0
0
0
0
0 0 o 0 ~ 0
0 0
O 0 O 0
~
~
9
,.o
~ ~.+ ~
o~ ~-~
i
~9 ~~
~_,~ I
~
O
o
l
l
l
l
l
~
l
~'~
o
o~
~
~o
8~
~
. ~
=
=
~
~
=
o
~
.
=
~
Inl
TURKISH LIGNITE DEPOSITS
AGE
LI THOLOGY
87
EXPLANATIONS
QUATERNARY
Slope
debris
. . . . . . . . , .
UPPER
.
.
.
.
.
.
.
.
,
i . . . . . . .... . . . . . . . .
EOCENE.
o
ooooo
o ~,
_Tl.lT
--
T - i- -
-iT)T1 I
9 I I I I
Conglomerate
ooo o
o o o
oc
I
I
I
J I I
I
I I I I I. I. I, I
. . . . . .
PALAEOZ01C
section
limestone
Smelly limestone Marl
~
Coat s e a m ( t o p )
~
Coal seam(
bottom
Bituminous
mar[
~
)
Marine limestone
~
Recrystollized
CRETACEOUS
stratJgraphic
Fossiliferous ~ -
SQodstone I I I I-~ I l l l J I 1 I Limestone
PALAEOCENE
Schematic
Agglomerate
I-L-i T2 t --~
T IT
EOCENE
F i g . 4.
Sandstone
of
limestone
S a n d s t o n e
~--~=~-~
Metomor~ic and
"--'1
Igneous
rocks
the Mengen Deposit.
and Mesozoic marbles. This formation contains occasionally 1.50-2.00 m thick lignite seams, the majority of which are not economically exploitable, and are not developed across the whole region. The overlying Turgut Formation is made up of conglomerates, claystones, siltstones, sandstones and gravelstones. A 20m thick lignite seam occurs between the Turgut Formation and the overlying Sekk6y Formation. This in turn is succeeded by the Yataf~an Formation consisting of gravelstones, claystones, sandstones, tufts, marls and limestones, and by the Miler limestones which are only locally developed. Alluvium overlies the Neogene sediments (Fig. 6). Only one seam of 3.30-17.50m mineable thickness. This seam has the following average quality: moisture 19.38-39.19%; ash 18.5450.14%; and total sulphur 2.35-5.00% all in the original coal, and a calorific value varying from 1434-2671kcalkg -1. There are 706.159 x 10 6 tonnes total proven reserves and 60000 X 10 6 tonnes probable reserves. Total reserves are 766.159x 10 6 tonnes (Table 3). There are open-pit mining operations in some of these sectors, using either dragline, or excavator and truck methods.
Soma deposit This Miocene deposit is in the district of Soma which is in the north of Manisa Province in western Turkey (Fig. 1). The sectors of this deposit are; Eynez, Merkez, I~klar, Tarhala, ~inge, Dualar, Evciler, Deni~ I Deni~ II, Kozlu6ren and Tfirkpiyale (Fig. 2). Stratigraphically these deposits in the Soma lignite region show similar characteristics (Fig. 7). The basement is made up of Palaeozoic metamorphic schists and greywackes, and Mesozoic crystalline limestones. The lignite-bearing Miocene sediments, consisting of units of gravelstones-sandstonesclay, marl and limestones, lie unconformably on the basement. There are two coal units; the main seam at the base of the marl and the middle seam in the middle-top parts of the limestone. The Pliocene sediments, lying unconformably on the Miocene sediments, are made up of coloured clayey sandstone, tuff-marl-agglomerate, gravelstone varved clay and silicified limestone-tuffite (Nakoman 1971). The upper coal seam unit is present in the coloured clay-sandstone units. Unconformably overlying the Pliocene sediments are Holocene units which are made up of terrace gravels, alluvium and slope debris.
88
H. INANER & E. NAKOMAN
AGE
LITHOLOGY
EXPLANATIONS
QUATERNARY
Alluvium
T
T
T T T]" Sandstone, ctaystone, marl, tuff
PLIOCENE A A A A AA A A A
A A
A
A
A
A
m
A
Basalt
A
Coal seam( top )
Sandstone, marl
OLIGOCENE
Coo[ s e a m ( m i d d l e ) Mctrt Coal seam ( bottom ) Sandstone Conglomerate
PALAEOZOIC
Gneiss
Fig. 5. Schematic stratigraphic section of the Malkara Deposit.
Andesites and basalts cover very large areas in the south and north of the Soma Region. The three coal seams in the Soma region, described above, vary in mineable thickness from deposit to deposit. The bottom, main seam, makes up the major part of mineable deposits in the region, except in the Deni~ I and Deni~ II sectors, where only the upper seam is mined. The coal properties are given below for the Soma coal region. Average coal thickness is 2.00-25.00m. On an original coal basis the average moisture is 12.24-30.28%, the ash 30.00-46.54%, the total sulphur 0.93-3.52% and the calorific value 1486-3428 kcal kg -1. The total reserves for these sectors are 601.255725 x 106 tonnes proven, 54.571977 x 106 tonnes probable and 22.764500 x 106 tonnes possible reserves, giving a total reserve of 678.592202 x 106 tonnes. Both open-pit (truck and shovel) and underground mining operations (longwall mining with sublevel caving) are employed in the area.
Beypazarl deposit Beypazari is 100 km northwest of Ankara and contains Miocene lignite deposits (Fig.l). The sectors of this deposit are; A B, and Altdamar (Fig. 2). Pre-Neogene rocks in the region are represented by Palaeozoic metamorphic schists and Palaeocene-Eocene Klzdbaylr Formation. The coal-bearing t~oraklar Formation is made up of cyclic sequences of sandstones, agglomerates and tuffites. Four formations containing volcano-sedimentary rock units above the t~oraklar Formation are the Hlrka Formation, with bituminous shales and natural soda, the Karadoruk Formation of solely limestones, the Pliocene Softa-1 Formation of sandstones and claystones, and the Softa-2 Formation of chalks and clayey limestones. Pleistocene unconsolidated gravels and Holocene terrace gravels, slope debris and alluvium overlie the Neogene volcano-sedimentary rocks (MTA 1993).
TURKISH LIGNITE DEPOSITS
LITHOLOGY
AGE
89
EXPLANATIONS i,
Alluvium ( 20 m.)
QUATERNARY
Limestone ( 50 m.) ~ " _ ~ " 7 --"TL";'.o--. . . . .
r T T, .'7.2-............ . ~ o o T ;-'." 7 . -
Grovelstone, claystone, sandstone, conglomerate
~
marl (150m.)
MIOCENE Coal seam
................... .L'.'--" : - ' . " ' . ' - -
"':-
:.-7._.-7 ....~_7_" -7-.~: .................. .......... . ......... o~,ooo=,,- 0 %,'o-eo o-o-o-,, * o--- 9 '{ i "ii i i ! )':'.:.: :':']ii:. ................... .............
claystone ( 200 m.) Sandst0ne, marl ( 200 m.) Coal seams Marine limestone
i .............
MESOZOIC
Conglomerat e, sandstone,
ILi[~
Mot ble
--r
PALAEOZOIC Gneisses
Fig. 6. Schematic stratigraphic section of the Mu~la Region (Yata~an-Milas) Deposit.
Neogene volcanics erupted from the begining of the Neogene intercalate with all formations (Fig. 8). Two coal seams are present, varying in thickness between 2.64 and 6.00 metres. The percentages of average moisture, a s h , and total sulphur in the original coal are 14.83-26.44%, 25.36-48.70% 2.79-4.04% respectively, and the calorific value is 1989-2839 kcalkg -1. The total proven and workable reserves are 390.3175 • 106 tonnes and 236 • 106 tonnes respectively. The coal is worked by underground mining, using a fully-mechanized longwall method.
Tunfbilek deposit The Miocene Tungbilek coal region covers large areas in the west and the north of the Tav~anll district belonging to the Kiitahya Province
(Fig. 1). The Domani~ region, which also has coal, is to the north of this region. The oldest rocks in the Tun~bilek area, which form the basement are Palaeozoic metamorphic schists and crystallized limestones, and Cretaceous ultrabasic rocks. The basement is unconformably overlain by the Miocene and Pliocene units. The Miocene units start with compact clastic formations and continue with a clay-marl sequence. The Tungbilek coal seam occurs in the lower parts of clay-marl sequence. Fresh water limestone and silicic limestone layers are present at higher levels (Nakoman 1988). The Pliocene is 300m thick and starts with clastic sediments which are transitional upwards to agglomerates and tuftites (Fig. 9). The fresh water limestones divide this series into two parts; andesitic volcanic rocks overlying tufts, tuffites and agglomerates which in turn are overlain by basaltic lavas. The Quaternary is made up of slope debris. There is a single Miocene coal seam,
90
H. INANER & E. NAKOMAN
AGE
L[ THOLOGY
EXPLANATIONS Alluvium, slope debris
QUATERNARY
VV V VvVvVvVvV V PLEISTOCENE T T T T ,'~": 2 '~-:" : : .~':'- ' "-~":'~ Q o o o o o T T T T TT TTT T T
PLIOCENE
Volcanic r o c k Tuff, tuffite, l i m e s t o n e Claystone, s a n d s t o n e A g g l o m e r a t e , tuff, tuffi te Marl Coal s e a m ( t o p ) Sandstone, c laystone Coal s e a m ( m i d d l e )
I
1 I
I
I
Limestone
1 I
I
1
MIOCENE
Mar[
Coal seam ( bottom )
-o
-
o
Claystone, sandstone
_o_o_o
MESOZOIC
Limestone
PALAEOZOIC
Metamorphic
schists
Fig. 7. Schematic stratigraphic section ofthe Soma Deposit.
which is 0.90-14.75m thickness and contains several laterally impersistent layers and lenses. The percentages of water, ash and total sulphur in the original coal are 10.65-14.35%, 38.0848.85% and 1.46-2.25% respectively and the calorific value is 2021-2657kcalkg -1. The proved and probable reserves are 270.850 x 106 tonnes and 46.882x 106 tonnes respectively making the total reserves of 317.732x 106 tonnes for the region. It is worked both by open-pit and underground mining operations. The opencast mining uses dragline, and excavator and truck systems. Underground mining is mainly by longwall with sublevel caving.
Seyit6mer deposit This area of Miocene coal is near Seyit6mer city in the west of Turkey (Fig. 1). The basement of the Seyit6mer region is generally made up of serpentinized ultrabasic rocks (gabbro, amphibolite, diorite, etc.), radiolarites and crystalline limestones (Fig.10). The Miocene which disconcordantly overlies the basement, starts with basal
conglomerates and sandstones followed by blue green coloured basal clays and by the main seam (bottom seam). The main seam is covered by clays with diatoms and marls with bituminous marls. The top seam overlies these claystone and marls. The Pliocene concordantly overlies the top seam seam, and starts with marl, tuffite and limestone (Nakoman 1988). The youngest unit is alluvium in the region. The top seam is too thin to be mineable. The bottom seam averages 16.00m mineable thickness. The percentages of water, ash and total sulphur in the original coal are 32.98%, 31.18% and 1.21% respectively, and the calorific value is 1900kcalkg -1. The proved reserves are 198.666 x 106 tonnes in the Seyit6mer region. The seam is worked by openpit mining methods with a dragline, excavator and truck system.
Can deposit These Miocene coals are found in and around (~an within the county of (~anakkale in northwest Turkey (Fig. 1). The basement is made up of
TURKISH LIGNITE DEPOSITS
LITHOLOGY
AGE
91
EXPLANATIONS ,,
HOLOCENE
Terrace gravel 0
0
0
O
PLEISTOCENE
o
gravel
o
C h a l k , c l a y e y limestone I - l(Softa-2 Forrootion) ._L,&__.:=,=.. S a n d s t o n e ~ c l a y s t o n e Sof t a - 1 F o r m a t i o n )
I 9 .
T
T
T I
Unconsotidated o
0
T
_
o o
0
T
0
b
o
=I-
0
o
o
"'~
PLIOCENE
0
0
o
T l
i
L",'
T
Ctaystone,
marl,
tuff
T l
Limestone
Is
MIOCENE
Ist I 9 I s l 1"~1)
( Karadoruk
Formation) ,,,
Bituminous shale, natural soda --
( Hirka Formation
)
Coat seam
..a_s.~..:_L.. r~" -,
. T .
.T_. S a n d s t o n e , a g g l o m e r a t e , t u f f i t e
o o o o o ~ o ~0 _%
( (~ o r a k l ~ " F o r m o t i o n ) C l a y s t o n e , s a n d s t o n e , gravelstor~
PALAEOCENE-EOCENE . . . . . =
=
o
PALAEOZOIC
o
( Kizilbaytr =
Formation)
,,
Metamorphic schists ......
Fig. 8. Schematic stratigraphic section of the Beypazan Deposit.
Palaeozoic phyllites, Mesozoic arkoses, limestones, spilite and diapsidic diabases (MTA 1993). The Miocene is represented by lignitic clays and tuffite layers reaching up to 400 m in thickness above the basement. The Pliocene consists of agglomerates about 300 m in thickness. The Plio-Pleistocene comprises a young andesitic volcanic suite (Fig. 11). The Quaternary takes the form of gravel terraces and alluvium. There is only one seam with an average mineable thickness of 16 m. The percentages of water, ash and total sulphur in the original coal are 18.21%, 27.90% and 4.20% respectively with a calorific value of 2994 kcal kg- 1. The proven reserves which can be exploited by both open-pit and underground mines, are 85.387 x 106 tonnes and 1.5 x 106 tonnes making a total of 86.887 x 106 tonnes, of which only 73.7 x 106 tonnes are workable (Table 3). The deposit has been worked by both open-pit and underground
mining methods and is currently mined by truck and shovel methods in an open pit mine at Can.
Orhaneli deposit The sectors of Burmu, ~ivili and Sa~lrlar show similar stratigraphic sequences at the northwest of Orhaneli in the Province of Bursa (Fig. 12). The pre-Neogene rocks are made up of schist, marbles and ophiolites and the Neogene formations of detrital rocks with conglomerates, coaly marl and tuffites at the base, and volcanic tufts and lava flows at the top. The post-Neogene sediments are Pleistocene gravels and valley fill alluvium (Fig. 12). The only coal seam is Miocene in age and hard and differs in quality from sector to sector.
92
H. INANER & E. NAKOMAN
AGE
LITHOLOGY
i
EXPLANATIONS
Slope debris
QUATERNARY i
A AA AA A A A
Basaltic [avos
T T T T T T T
Tuff
V
V
V
V
V
V
V
Andesite
PLIOCENE Fresh water limestone T T T T T T T ~ Tuffite, agglomerate 9
,
oo
.
o o o o o .. s l s
1
sl-
'_J.' Silicic timestone
[
I ,
J,
Claystone, marl
MIOCENE
Coal seam ( 2-14.75m, )
s
5
s
;
CRETACEOUS
UItrabasic rock
sSs 55 S's; s PALAEOZOIC
~ ' ~'-"-"--x.. ~ ,
Metamorphic schist Crystallized hmestone
Fig. 9. Schematic stratigraphic section of the Tun~bilek Deposit.
A C~
LITHOL O G Y
QUATERNARY
Alluvium
l PLIOCENE
EXPLANATION S
I
l
T
1
1
T
T
T
L imestone Tuffite Marl Cacti seam ( top ) Diotomic claystone Bituminous mar[
MIOCENE
Coat s e a m ( b o t t o m ) Blue green coloured, basal claystone Basal conglomerate
PALAEOZOI C
Crystallized limestone R adiolari t e. serpantinize( uitmbasic rock
Fig. 10. Schematic stratigraphic section of the Seyit6mer Deposit. This deposit has been worked by both open-pit and underground mining methods. In the sectors which are mined by open-pit, the percentages of water and ash in the original coal are 21.2924.45% and 22.47-42.96% respectively, and the
calorific value is 2134-2850 kcal kg -1 . The proven and workable reserves are 32.340 • 106 tonnes and 29.100 • 106 tonnes respectively. In the sectors which have underground mining, the percentages of water and ash in the original
TURKISH LIGNITE DEPOSITS
AGE QUATERNARY
EXPLANATIONS
LITHOLOGY
' o"9 . .o, . -,0 "O. O*"'0' " .0 V
V
V
-,'o'
V
V V
Alluvium
" 0 '*.
V
V
PLIO-PLEISTOCENE
V V
V
PLIOCENE --T
--
T
T
--
yu
T
-_
T
T
--
--T--T--T
--
Tuff, Lignltic CIoystone
C(aystone, tuff Cool seam ( bottom )
T r
T
T
T
T
T T T T T T T T T T
T
T
Andesitic Tuff
ConglomeraCe
Arkose, spil/t e ond diopsitic di abase Phyl lit e
M ESOZOIC PALAEOZOIC
Fig. 11.
Andesite
Agglomerate
-T
UPPER MIOCENE
93
Schematic stratigraphic section of the (~an Deposit.
coal vary are 21.37-25.29% and 14.4637.88% respectively and the calorific value is 2294-3412kcal kg -1. The proven and workable reserves are 13.712 x 10 6 tonnes and 9.6 x 10 6 tonnes respectively. The total proven and workable reserves are 46.052 x 10 6 tons and 38.70 x 10 6 tonnes respectively (Table 3). The mining is by dragline.
Af~in-Elbistan deposit The Pliocene Af~in-Elbistan lignite deposit is located in southeastern Turkey (Fig. 1). This deposit is the biggest in Turkey and covers 120 km 2. The Pre-Neogene basement consists of Permo-Carboniferous limestones, Eocene limestones and ophiolites (MTA 1993). Neogene sediments are limnic in character and the sediments below the coal are made up of claystone, marls, and gravelstone. Freshwater limestones and Post-Neogene formations, about 80m thick, consist of gravels and sandstones overlying the coal bearing formations (Fig. 13). The coal has a minumum thickness of 4.00 m, and a maximum of 58.00 m, averaging 39.58 m. The coal lies between 10.00 and 150m beneath the surface, with an average depth of 50 m. The
percentages of water, ash and total sulphur in the original coal are 50.00%, 20.00% and 1.46%, respectively, and the calorific value averages 1050 kcal kg -1. There are 3 357.340 x 10 6 tonnes of proved and 2.115 x l 0 6 tonnes of workable reserves (Table 4). It is being worked by open pit mining. The Af~in-Elbistan Lignite Establishment was formed to feed the largest coal power station with a capacity of 20 mtpy in Turkey. As the largest open cast mining project in the country, six bucket wheel excavators, five spreaders, five reclaimers and a belt conveyor system of approximately 55 km in a total length are utilised.
Bey~ehir deposit This Pliocene lignite deposit is in the west of Konya Province in southcentral Turkey (Fig. 1). The sectors of this deposit are: Karadiken, AvdancN, Ak~alar (Fig. 2). Stratigraphically these sectors show smiliar characteristics. The basement is made up of Palaezoic metamorphic schists and Mesozoic crystalline limestones. Pliocene sediments lie unconformably on the basement. These are made up of the generally loose and occasionally uncemented gravels,
94
H. INANER & E. NAKOMAN
AGE
9
PLEISTOCENE
EXPLANATION
LITHOLOGY
~
" ' '
,
.o
.
.
.
.
: ai" o..'o.
o'
"
"
.
,. o .' :o
o"
'
. . .
AUuvium, gravel
NEOGENE
9 - o
POST
,
9
Andesite ./
I .. \.
T
/\/
T
T
T T
PLIOCENE
T T
T
T
T
T
T
T T
T T
T T
T
T
T T
T
T
g
T
Tuffite
T
~ ~ ~ ~-
T
MIOCENE
u~ z 0
T
T
T
~ u ~
T
T
T
Tuff .u_
T T
T
UPPER
T
T
T
T
T
T T
T
T
T T
T T
T
T T
Basalt
T
T
T T
i
w (..9 o
m
._ ~l-Iol-l-Iol-Il--I--I--l-I Ii1--1~1
UPPER
A
CRETACEOUS
~
PALAEOZOIC
~
A A
Bosal congtomerate
A A
A_~ ~ m
Vg~ ~
/
Marl
I~
I--l-l'l I-I-
A A
Coo[ seom
~
Ophio[ite
Marble
~(y:)n rr Metamorphi c schists
Fig. 12. Schematic stratigraphic section of the Orhaneli Deposit.
sands and clays (MTA 1993). Lignitic beds, limestones and marls lie concordantly over these units. The top levels of the Pliocene are made up of limestones. The youngest units are alluviums in the region (Fig. 14). The seam thickness varies between 3.20 to 8.14m. The average moisture, ash, and total sulphur contents in the original coal are: 45.3553.42%, 16.55-27.00%, and 1.09-1.10%, respectively and the calorific value is 10831430kcalkg -1. There are 10.276 • 106 tonnes of proven reserves which can be worked by underground mining method in Ak~alar sector. The total proved and workable reserves which could be worked by open-pit mining, are 218.590 • 106 tonnes and 160.326x 106 tonnes respectively (Table 4). There is no production in the area.
Kangal deposit
This Pliocene deposit which has three sectors, namely Kalburqaym, Etyemez, and Hamal, is situated 25km south of Kangal city in east central Turkey (Figs 1 & 3). Stratigraphically these sector are very similar. The basement is made up of the Mesozoic low grade metamorphic limestones and ophiolites. Neogene sediments lie unconformably on the basement. These are divided into two formations: The Kalburqayln and Bicir Formations. The Kalburqaym Formation begins with gravels at the base, and is overlain by coaly units. The coal and coaly clays contain gastropod fossils. The Bicir Formation lies conformably on the Kalburqaym Formation. The Bicir Formation can be easily
TURKISH
LIGNITE
DEPOSITS
LITHOLOGY
AGE
-o
POSTNEOGENE
m
9
.
~,..
EXPLANATIONS
o
o
95
.
Gravel, s a n d s t o n e ( 8 0 m thick)
o
o
o
9
o
1 I ~ ~ I I ] l I I 1 [ 1 I ] I I I [ I i
Fresh w a t e r limestones
NEOGENE Coot seom ( 4 - 5 8 m. thick) Cloystone, marl, grovelstone oo_Oo
I
EOCENE
go_
o.go_o
I
I
I
I
Limestone, o p h i o l i t e
l
Limestone
P ERMOCARBONIFEROUS
Fig. 13. Schematic stratigraphic section of the Af~in-Elbistan
AGE
Deposit.
LIT HOLOGY
EXPLANATIONS
OUATERNARY
Alluvium
Mort I
I
1 1
1
PLIOCENE
I
1 I
I
i
l
I I
Limestone
I
Cool
seom
--
o ~ o . ~ o - o . . o - o - . o .o 9 e. o - . o - -
o
-
-
- -
.
o - - - - 0 - o
9 o.
o .--.o - ~ . o . - ; o o_.o o.~.--=.__o - -
Oo ~ 1 7~6o
- -
----.o'.
o
Uncemented grovels, sonds and days
.
MESOZOIC
Crystallized limestone
PALAEOZOI C
Metamorphic
schists
Fig. 14. Schematic stratigraphic section of the Bey~ehir Deposit.
distinguishable by its yellowish white coloured limestones, clayey limestones and marls. Post Neogene rocks are volcanics and the Quaternary is made up of terrace gravels and slope debris (Fig. 15). There are two coal seams, the bottom and upper. The average thickness of the bottom seam is 8.00 to 15.00 m and the average thickness of
upper seam is 1.90 to 12.00m. The percentages of average moisture, ash, and total sulphur in the original coal are 49.83-52.09%, 19.0421.00%, and 2.02-3.57% respectively, and the calorific value is 1207-1494kcal kg -1. The total proven reserves are 202.607 x 106 tonnes. It is being worked by open-pit (truck and shovel) and underground methods (Table 4).
96
H. iNANER & E. NAKOMAN
AGE
LITHOLOGY
EXPLANATIONS
QUATERNARY
Terrace
A A
A
gravel
A
A
A
Basalt
A
POSTNEOGENE T
T
T T
T
T
T T
T
Tuff
Yellowish white coloured timestone, c l a y e y limestone.
,'T~ ITli!TL,
marls
-lJ - I - l--l-l-l-l1 I I I I I I i i I I I I I I I I I I
( Bicir F o r m a t i o n )
I J I I I I NEOGENE Clayey mar[ Coal seam ( top ) Marl Coal seam ( b o t t o m )
o
.
o D
o
L'.~
9
~ ~ o
a
9
o
Gravelstone
( Kalbur~ay=r= Formation )
o
Semi- metamorphic
MESOZOIC
limestone
Ophiolite
Fig. 15. Schematic stratigraphic section of the Kangal Deposit.
Orta deposit This Pliocene deposit is near (~ankm in north central Turkey (Fig. 1). The basement consists of Lower Creteceous crystalline limestone. Volcanism, which continued from the Eocene to the end of the Miocene, produced a volcanic cover over large areas of the region and form the base of coal beds. The coal bearing formation lies disconcordantly over the basement volcanics and is assumed to be Pliocene in age (MTA 1993). It is made up of sandy claystones and conglomerates, containing thick but low calorific value lignites. The youngest sediments are modern alluvial deposits (Fig. 16). Two coal seams are developed with an average thickness of 20 m. The percentages of moisture and ash are; 48.47%, and 28.59%, respectively, and the calorific value is 868 kcal kg -1 . The total proven reserves are 123.165 • 106 tonnes, which
could be worked only by open-pit mining which is not yet operational (Table 4).
Karhova deposit This Pliocene deposit is located in the northeast of Bing61 Province (Fig. 1). The basement is made up of the andesites and the basalts which are Upper Miocene in age. Above these andesites and basalts, cyclic sequences, consisting of 10-15m of Pliocene tufts and tuffites and 20-25m of siltstone, sandstone, gravelstone, conglomerates and clays, were deposited. The boundary between this series and the coaly series is made up a band of 2.5 m thick agglomeratestufts. The coaly series follows concordantly above basement rocks. The thickness of this series which is made up of clays, clayey tuffites,
T U R K I S H LIGNITE DEPOSITS
EXPLANATIONS
LITHOLOGY
AGE
97
Alluvial deposits
QUATERNARY ~.-
.
.
O
.
O
.
O
.
o
o
o
Coal seam ( t o p )
PLIOCENE
Sandy claystone Coal seam ( bottom ) ," ~
EOCENE
~
0
~
O
O
O
V V V V V V V V V V V V V V V V V V V V V V V V V V V V V v V
Conglomerot e
Volcanic rocks
Cryst at[ized |imestone
LOWER CRETACEOUS
-1Fig. 16. Schematic stratigraphic section of the Orta Deposit.
[
AGE
EXPLANATIONS
LITHOLOGY
Alluvium, grovel, travertines
QUATERNARY v V V V v V V V V v V T TTT T TT T o
"
o
" "o','
ii " o.:
Andesite, basalt Tuff, agglomerate Coat seam ( top )
,o..
PLIOCENE
Sandstone, gravelstone Coal seam(bottom) Tuff. tuffite V V V
UPPER MIOCENE
V
V V V V V V V V V V V V V V V V V V V V V V V V V V V
CRETACEOUS
Andesite, basalt
Basement rock
Fig. 17. Schematic stratigraphic section of the Karhova Deposit.
sands and gravels is about 350-400m and contains two coal seams. A younger volcanic unit of agglomerates, tufts, andesites and basalts overlies them. The quaternary is represented by gravels with travertines and alluvium (Fig. 17).
The average thickness of the bottom seam is 8.50m. The percentages of moisture ash and total sulphur in the original coal in the underground mining area, are; 43.00%, 24.63% and 0.57% respectively, with a calorific value
98
H. INANER & E. NAKOMAN
of 1663kcalkg -1. The proven reserves are 53.884 x 10 6 tonnes. The percentages of moisture ash and total sulphur in the original coal of the open-pit mining area, are; 46.56%, 24.08%, and 0.47%, respectively, with a calorific value of 1458 kcal kg -1. The proven reserves are 30 x 10 6 tonnes. The coal can be worked by both open-pit and underground mining methods. There is no current production (Table 4).
General economic evaluation and results Estimates of the total lignite reserves in Turkey from studies to date are about 8.4Gt, out of which 3.9 Gt are the exploitable reserves. Turkish lignite reserves represent around 2% of the world total (K6ktiirk 1994). TKI owns 6 Gt of these lignite reserves which amount to 72% of the total reserves of the country. 3.2 Gt of the TKi reserves are in the Pliocene Af~in-Elbistan coal basin. There have been no serious investigations during the past decade into the development of the lignite reserves in Turkey. Additional exploration must be carried out to find new deposits. A range of information is required before the coal reserves could be utilized in the power plants. Most of the known lignite deposits in Turkey are of low calorific value and have high contents of ash, moisture and total sulphur. 6 Gt of lignite reserves (72%) are concentrated within 70 important deposits. The lignite reserves, both worked and unworked, owned by TKI, are classified according to the calorific values as shown in Table 5. Almost 80% of the total reserves have calorific values below 2500kcalkg -1, 13% are
in the range 2500-3000 kcal k g -1, while only 7% are over 3000 kcal kg -1. The calorific value for industrial and household use must be greater than 3000 kcal kg -1 (Table 5). Thus 93% of total reserves are unsuitable for industrial and household purposes. These low calorific value lignites can only be used in for power plants. Electrical energy in Turkey is produced in both thermal and hydro power plants, with a total installed capacity of 20125 MW. Thermal power plants represent about 52% (10443 MW) of the total installed capacity, the rest being hydro-electrical plants. The installed capacity of lignite fired power plants is 5450MW. These are Af~inElbistan, (~aylrhan, Kangal, Kemerk6y, Orhaneli, Seyit6mer, Soma, Tunqbilek, Yata~an and Yenik6y. The total saleable lignite production of Turkey is 45.4Mt, 38.7Mt of which are produced by TKi and the rest by private enterprise (K6kti.irk 1994). Almost 90% of Turkish lignite deposits are worked as open-pit mines, the remainder being underground mines. In terms of the technology employed in mining lignites, TKI is well equipped for open-pit mining compared with underground mining. Underground mining is being carried out either by fully mechanized or semi mechanized systems depending on the formation and coal seam conditions. Coal mining in the Beypazan underground mine is performed by a fully mechanized system and similar systems are envisaged for other underground operations. Depending on the geological and mining conditions different mining systems such as dragline-excavator-truck, excavatortruck or bucket wheel excavator-belt conveyor systems are used in the opencast mines for coal production and overburden removel. The
Table 5. According to the lower calorific values, the proved and total lignite reserves in Turkey
Lower calorific value (kcalkg-1)
Proved reserves (tonnes • 10 6)
(%)
Total reserves (tonnes • 10 6)
(%)
>1000 1001-1500 1501-2000 2001-2500 2501-3000 3001-3500 3501-4000 4001-4500 >4501 Total
123.2 4367.1 870.3 674.0 927.0 324.4 4.3 11.8 36.9 7339.0
1.6 59.5 11.9 9.2 12.6 4.4 0.1 0.2 0.5 100.0
265.7 4519.1 972.2 932.2 1105.3 401.0 28.5 31.8 118.6 8374.4
3.2 54.0 11.6 11.1 13.2 4.8 0.3 0.4 1.4 100.0
TURKISH LIGNITE DEPOSITS
99
Table 6. Areas of utilization for the Turkish/ignites
References
Consumption
%
Thermal power stations Domestic heating Industrial factories Internal consumption
67.4 16.5 15.7 0.4
AKGUN,F. & AKYOL,E. 1992. Palynostratigraphy of the coal-bearing Neogene deposits in Bfiyfik Menderes Graben, Western Anatolia. 1st. International symposium on Eastern Mediterranean Geology, Proceedings and Abstracts, Adana, Turkey. ALTAr, M., (~ELEBI, E. & FIKRET, H. 1994. Development of energy sector of Turkey and projections of supply and demand (1970-2010), 6th National Energy Congress. izmir, Turkey (in Turkish). INANER, H. 8r NAKOMAN,E. 1993. Lignite deposits of the western T~irkiye, Bulletin of the Geological Society of Greece, 28]2, 493-505. KAYA, O. & DiZER, A. 1984. Stratigraphy of Mengen Coal Basin, MTA Publication, No. 97/98 Ankara, Turkey (in Turkish). KOKTORK, A. 1994. Lignite Sources and Utilization in Turkey, Turkish Energy Day, 6th National Energy Congress, Izmir, Turkey (unpublished, in Turkish). MTA, 1993. Turkish Coal Inventory, Ankara (in Turkish). NAKOMAN E. 1971. Coal, MTA Educational Series No. 8, Ankara (in Turkish). - - 1 9 8 8 . Coal Deposits of Turkey, Postgraduate Lecture Notes, Izmir (unpublished, in Turkish).
machinery and equipment used in open cast operations mostly reflects the latest technologies. Draglines of various capacities are utilized in major open cast operations. Lignite is used mainly to generate electricity in thermal power station (67.4%), for domestic heating (16.5%), industrial factories (15.7%), and internal consumption (0.4%) (Alta~ et al, 1994). The lignites with low calorific values are generally consumed in power plants under state ownership. Private lignite companies usually supply lignite for domestic and industrial uses (Table 6).
The origin and properties of a coal seam associated with continental thin micritic limestones, Selimoglu-Divrigi, Turkey A. I. K A R A Y I G I T 1 & M. K. G. W H A T E L E Y 2
1Department of Geological Engineering, Hacettepe University, Beytepe-Ankara, Turkey 2Department of Geology, University of Leicester, University Road, Leicester LE1 7RH, UK Abstract: The Selimoglu coalfield is situated at the southeastern part of Divrigi, which is geographically distinct from the major coalfields of Turkey. A number of thin and lenticular coal seams and sedimentary rocks occur in the Selimoglu unit of the Ogulbey Formation of Upper Miocene age. Only one seam, which is produced by an underground method, is associated with continental thin micritic limestones that have a total thickness of about 150 cm. It has a macroscopically bright appearance and a working thickness of 50-90 cm (70 cm average). A total of 32 channel samples were collected from the mined seam for proximate, mineralogical and petrographic analyses. The results of proximate analyses on an air-dried basis show that the coal is characterized by low moisture content (1.68% average), high ash yield (24.04% average), high total sulphur content (up to 8.92%) and high calorific value (5606kcalkg-1 average). The coals average 65.5% vitrinite, 4.5% liptinite, 2.5% inertinite and 27.4% mineral matter. Desmocollinite is the dominant maceral in the vitrinite group and calcites in the whole-coal minerals. Carbonate minerals with irregular shapes are in general early-diagenetic, and occur in desmocollinites in all the samples. The micritic texture of the limestone shows both diagenetic and authigenic origin in subaquatic conditions with high pH, but organic debris is allochthonous in origin. The reflectance values of telocollinite (0.77% Rr) show a high volatile bituminous coal rank. The random reflectance of telocollinite and spectral maxima (625-661 nm) of sporinites, and proximate analyses reveal that the thermal history may have been affected by volcanic activity that occurred in the coal field.
Most Miocene coals in Turkey are of lignite or subbituminous coal rank and are generally associated with claystone, marl and rarely sandstone. However, there is a lack of information about this type of formation in Turkey. The coal seam, which is produced in an underground mine, forms a useful example to assess the formation of coal associated with continental micritic limestones in the Selimoglu coal field. This paper summarizes the geological setting and stratigraphy of the coal field, and presents proximate analyses, the mineralogicpetrographic composition, spectral properties of some liptinite macerals and the rank of the coal seam. The Selimoglu coal field is geographically remote from the major coalfields of Turkey, and is situated 20 km southeast from Divrigi (Fig. 1). About 30-50 t/day of coal is produced in Coal Mine II (for location see Fig. lb and lc). The region containing the coal field was first investigated by Wedding (1965), who studied the stratigraphy of the region. This was followed with studies by Keskin et al. (1984), who revised the geological map and determined a similar stratigraphy for the coal field as that of Wedding
(1965). The basic geological characteristics of the Divrigi region, including the northern small part of the study area have been investigated by Tunc et al. (1991). This present paper represents an extension of a preliminary study by Karayigit (1993).
Methods of study Representative rock, coal and coaly bituminous shale samples were collected; from which thin sections of rock samples were prepared to determine the petrographic composition. X-ray powder diffraction (XRD) analyses were performed to determine the mineralogical composition of limestones. In addition, for age determination, limestones, marls and claystones were sieved for ostracoda, and palynological investigations were made on coals and coaly bituminous shale. A total of 32 (31 samples from Coal Mine II and 1 sample from Coal Mine I for locations see Fig. lc) fresh, channel coal samples that represent the full thickness, including dirt bands ( < l c m thick) within the seam, were
From Gayer, R. & Pe~ek, J. (eds), 1997, EuropeanCoalGeologyand Technology, Geological Society Special Publication No. 125, pp. 101-114.
102
A. I. K A R A Y I G I T
& M. K. G. W H A T E L E Y
Fig. 1. (a) The stratigraphical sequence of the Selimoglu coal field, (b) simplified geological map around the coal mines, (e) some macroscopical seam sections, lateral extend of the mining seam and the underground map on a more detailed geological map (modified after Karayigit 1993).
CONTINENTAL MICRITIC LIMESTONES, TURKEY collected from the coal seam for proximate, mineralogical and petrographic analyses. Proximate analyses (moisture, ash, volatile matter) as well as total sulphur analyses and calorific values of all samples were performed and the reported results expressed as weight percentages, except calorific values, and made in accordance with the ASTM (1991) procedure. The whole-coal minerals of all coal samples were identified by X-ray powder diffraction (XRD). After identifying all peaks on the every X R D diagram, the net area under each diagnostic peak was determined and converted to a percentage for each mineral. In order to determine the chemical composition of carbonates and some silicate minerals, two coal briquettes were selected and examined on a JEOL 8250 electron microprobe. Maceral analyses were determined using a reflected light microscope (Leitz MPV II) with a 32x objective and oil immersion (noil :1.518) on polished briquettes. Ordinary white light from a tungsten lamp and blue light, K510 barrier filter for determination of liptinite macerals were used for illumination. The analyses are based on counting 500 points on each sample and the reported results are expressed as volume percentages of the various macerals and minerals. Petrographic constituents of the coals were determined using the information given by ICCP (1963; 1971) and Stach et al. (1982). Random reflectances of vitrinite (telocollinite and desmocollinite) were measured with a minimum of 50 points on every briquette using the same microscope with a 50• oil immersion objective, sapphire (0.551%R) and glass (1.23% R) standards for calibration. During spectral fuorescence emission measurements of liptinite mecerals, a Leitz MPVSP microscope fitted with a high-pressure 100WHg light source, a BG38 and a BG1 filter, and a K460 barrier filter was used. Spectral intensities in the range of 460-700 nm were measured and corrected spectral intensities were automatically produced using a connected computer. Some numerical parameters, such as relative intensities, wavelength of maximum intensity ()~max) and the logarithmic ratio (Q) of the relative intensity of red (650nm) and green (500nm), were then generated from them. The principles, basic calibration techniques and some applications of measuring fluorescence in geological samples are documented in the work of Jacob (1964; 1973), Pflug (1966), Ottenjann et al. (1975), Teichmtiller & Ottenjann (1977), Teichmiiller & Wolf (1977), Robert (1981), Crelling (1983), and Teerman et al. (1987).
103
Geological setting and stratigraphy The location map of the study area, the stratigraphical sequence of the coal field and simplified geological map around the coal mines are shown in Fig. 1. The Giines ophiolite forms the basement in the coal field (Fig. la). It is of Upper Cretaceous age and contains generally serpentinized rocks (Tunc et al. 1991). The Ogulbey Formation rests unconformably on the basement. It has an extensive areal distribution and is subdivided into three informal units; from base upward, Hantepe unit, coal-bearing Selimoglu unit and Hanioglu unit (Fig. 1a). The Hantepe unit, which has an average thickness of 90m, contains mainly thin bedded lacustrine micritic limestones (Fig. la) that are composed of calcites. The Selimoglu unit only hosts coal seams throughout the coal field and the thickness can reach up to a maximum of 295 m. In the lower part of the unit, reddish mudstones and minor sandy limestones and thin sandstones (<20cm thick) are more common. In the middle and upper part, the unit consists of laterally discontinuous rocks ranging in grain size from conglomeratic sandstone to claystone, and also coal seams and limestones. At the top, a laterally extensive coaly bituminous shale (about 100cm thick), which can be used as a marker in the coal field, is present. The conglomeratic sandstone occurs as a cross-stratified channel fill sequence. Most of the thick-bedded sandstones contain pebbly channel lags. Both the conglomeratic and thick-bedded sandstones display erosional bases, and they are interpreted as fluvial deposits. Thinner sandstones, with large amounts of calcitic cement, show no sedimentary structures. Only one coal seam, which is exposed around Selimoglu village (Fig. 1b and c), is produced by underground mining. This seam is associated with continental thin micritic limestone lenses and bands (Fig. lc), that have a total thickness of about 150 cm. The seam has a macroscopically bright appearance and a working thickness of 50-90 cm (70 cm average). Its thickness within the limestones decreases from Coal Mine III to Coal Mine I (Fig. lc), which presumably relates to subsidence, carbonate precipitation and organic matter supply in the peat environment. The limestones are well cemented and composed of calcite. The micritic texture of the limestones shows both diagenetic and authigenic origin (Gierlowski-Kordesh et al. 1991), but organic debris is probably allochthonous in origin. Up to 7 other different coal seams (<50 cm thick) in the unit have no economic significance.
104
A. I. KARAYIGIT & M. K. G. WHATELEY
The lower part of the Selimoglu unit was interpreted to have been accumulated in a lacustrine environment, but its middle and upper parts are in general thought to represent an alluvial-fluvial environment in a prograding lacustrine deltaic system. The mined seam was probably formed in a small, shallow lake or pond on an alluvial plain, whereas the other seams occurred in small swamps between fluvial channels. The Hanioglu unit, which has an average thickness of 500m, contains alternations of limestone-marl, and claystone-limestone-marl of lacustrine origin. The limestones have a similar petrographic composition to those of the Hantepe unit. An Upper Miocene to Pliocene (?) age, based on ostracod studies, was given to the Ogulbey Formation. In addition, charas and gastropods were also found in some samples. All the fossils appear to be of non-marine origin. The mined seam in the Selimoglu unit does not contain sufficient diagnostic palynomorphs to determine its age. However, in some samples collected from the other coal seams and the coaly bituminous shale, some spores and pollens were determined as Miocene in age. It was concluded that the Ogulbey Formation was of Upper Miocene age. The Dejdekar volcanics of Upper Miocene (or Lower Pliocene?) age are usually massive with rarely columnar jointing, and were intruded into the Ogulbey Formation. These rocks are petrographically identified as andesite, trachyandesite and dacite. The Yamadag lava flows are of Lower Pliocene age (Tunc et al. 1991) and they were petrographically determined as olivine-basalt. Over these units, Quaternary deposits contain alluvium, landslides and rockfalls (Fig. la and b). During field studies, it was determined that the nearly horizontalyl bedded Ogulbey Formation in the area close to the Dejdekar volcanic intrusion was tilted up to 70 ~. Even though the rocks of the Ogulbey Formation above the intrusion have been mainly eroded, the formation has a circular outcrop (representing a dome fold) produced by the Dejdekar volcanic intrusion. A number of normal faults in the Ogulbey Formation were formed by the volcanic intrusion (Fig. lb).
Results and discussion P r o x i m a t e analysis
Table 1 summarizes the results of the proximate and X R D analyses of the coal samples on an air-
dried basis. The moisture contents are low and average 1.68% (Table 1). In addition, it was determined that the moisture contents of the coals on an as-received basis are less than 5% according to unpublished reports of the coal mining company. This means that the coals are characterized by low moisture contents. The coals are also characterized by high values of ash yields, volatile matter contents and calorific values (Table 1). The average value of the volatile matter contents on a dry, ash-free basis are calculated as 52.82% from the results on an air-dried basis. This value shows a subbituminous coal rank (Unsworth et al. 1991). Whereas the coals, during the volatile matter determination, produced an agglomerate button showing swelling. This character, as known, suggests at least a bituminous coal rank in the ASTM (1991) classification. It appears that the volatile matter contents was increased from the micritic limestones that formed in the coals. The coals are characterized by high values of total sulphur contents (Table 1). High sulphur contents in coals are commonly explained by the proximity of the original peat to marine waters during deposition, as the sulphate ions in seawater provide an abundant source of sulphur (Casagrande et al. 1977; Cohen et al. 1984; Given & Miller 1985). Whereas, as mentioned earlier, there are no marine overburden deposits
Table 1. Range and average values of the proximate analyses and minerals identified by X-ray diffraction studies of all the coal samples on an air-dried basis Analyses Proximate analysis Moisture (%) Ash (%) Volatile matter (%) Total sulphur (%) Calorific value (kcal kg-1) Calorific value (MJ kg-1 )
Range
Average
1.20-4.00 1.68 17.86-32.46 24.04 36,08-47.10 39.12 4.14-8.92 5.64 4284-6312 5606 17.94-26.43 23.47
XRD analysis of the whole coal (%) Clay minerals (30)* 0-8 Gypsum (8) 0-8 Quartz (32) 1-16 Feldspar (27) 0-28 Calcite (32) 33-96 Dolomite (27) 0-7 Pyrite (32) 1-19 Marcasite (3) 0-2
2 1 9 7 70 2 10 0
* Figures in brackets are the number of samples determined on X-ray powder diffraction (XRD) diagram.
CONTINENTAL MICRITIC LIMESTONES, TURKEY in the coal field. The high sulphur contents of the coals can be related to possibly high water table, high pH and low Eh in the calcium-rich peat-forming environment (Stach et al. 1982; Roberts 1988). These conditions will be discussed below in the light of the results of the other analyses.
The dominant mineral in the coals is calcite (Table 1 and Fig. 2). Pyrite is by far the dominant sulphide present in the samples; marcasite is extremely rare. Quartz, feldspar including plagioclase (Fig. 3a) and K-feldspar, clay minerals, dolomite and gypsum make up the other constituents (Table 1). Plagioclase and most of the quartz (Fig. 3c) may be of detritial origin, eroded and transported from the granitic and dacitic basement rocks that are exposed in the Divrigi region. Gypsum was probably of secondary origin; the circulation of more recent oxidizing meteoric waters may have caused the oxidation of the pyrites and the resulting sulphate is precipitated as gypsum as secondary minerals in the coal seams.
Mineralogy The major mineral phases found in the coals on an air-dried basis were determined by XRD analysis. The minerals, their range and average values are given in Table 1. Two examples of XRD traces are presented in Fig. 2.
Cm:
Clay m i n e r a l s Feldspar Quartz Calcite Dolomite Pyrite
F: Q: C: D: Py:
A; C
Q Q C C ~ i ; ~C :
I
105
PY .Co.
9- Z L , . L ~ . . ~'
Q
c
~
Cm
Q ~ . CC~ c
~ qC
~
Q ][ ' II 5
10
. . . .
I . . . .
15
I .... 1 .... 20
25
I .... 30
I .... 35
I .... 40
[ .... 45
o%
I .... I .... I .... 50
55
60
I
20
65
(CuKot)
Fig. 2. Two examples of X-ray powder diffraction (XRD) traces from the mining seam in the Selimoglucoal field.
Fig. 3. (a) Plagioclases (P) showing some alterations with white zonal structures (A) determined by a scanning electron microscope image, (b) early-diagenetic corbanate minerals (C), pyrite (Py) within desmocollinite (Dc), (e) early-diagenetic corbanate minerals (C), quartz (Q) and inertinite (I) within desmocollinite (Dc).
CONTINENTAL MICRITIC LIMESTONES, TURKEY Table 2. The detection limits of electron microprobe for carbonate analysis, range and average values of the chemical compositions of early diagenetic carbonate minerals analysed on two coal briquettes selected from the coal samples Carbonate analysis
Detection limits
Range
Average
(%) (n ----43)
CaO 0.02 MgO 0.02 FeO 0.05 MnO 0.05 SrO 0.08 BaO 0.15 CO2 (by difference)
22.46-49.21 0.27-4.61 0.01-0.36 0-0.08 0.02-0.68 0.01-0.22 49.00-76.44
39.17 1.43 0.08 0.03 0.13 0.11 59.04
n: The number of measurements
The detection limits of the electron microprobe, and the range and average values of the elements in the early diagenetic carbonate minerals in the coals, are presented in Table 2. The dominant constituent is CaO which has a range of 22.46-49.21% (39.17% average). It is thought that the lower C a O % values are from organic materials and/or clay minerals mixed in the calcium carbonate. MgO is less than 5% in the calcium carbonates. FeO, MnO, SrO and BaO are mainly below detection limits. Minor amounts of dolomite were identified by X R D analyses in the coal samples (Table 1). It is possible that Mg required for dolomite and found in some calcium carbonates may have
107
been derived from the serpentinized Gfines ophiolite. The chemical composition of plagioclase, K-feldspar and quartz and the relevant detection limits on the electron microprobe are shown in Table 3. The analysis results of the plagioclase indicate an andesine-labradorite composition. In addition, some alterations showing white zonal structures were determined within plagioclases in a scanning electron microscope image (Fig. 3a). The alteration zone within the large plagioclase grain shown in Fig. 3a was analysed at four points and the total oxide composition was found to be less than 61.87%, in which 43.39-51.49 SIO2% 0.01-0.05 TiO2, 4.26-5.36 A1203%, 0.01-0.04 CrO3%, 0.04-0.08 FeO%, 0.02-0.03 M n O % 0-0.01 M g O % , 1.58-2.89 C a O % , 2.16-2.39 N a 2 0 % , 0.08-0.19 K 2 0 % and 0.01-0.03 N i O % were determined. It is possible that the alteration zone was formed by chemical weathering of plagioclase in the peat environment, and was later infilled with mainly quartz and small amounts of clay minerals that occurred in peat environment. The chemical composition of the quartz minerals indicates that some quartz minerals are almost pure SiO2 (Table 3) and it is possible that these quartz minerals may be of detrital origin.
Petrographic composition and depositional environment The range and average values of the results obtained from the petrographic analyses are
Table 3. The detection limits of electron microprobe for silicate analysis, range and average values of the chemical compositions of plagioclase, K-feldspar and quartz on two coal briquettes elected from the coal samples Silicate analysis (%) SiO2 TiO2 A1203 Cr203 FeO MnO MgO CaO Na20
K20 NiO Total
Detection limits 0.02 0.04 0.02 0.04 0.05 0.05 0.02 0.02 0.02 0.02 0.04
n: number of measurements
Plagioclase (n = 16)
K-Feldspar (n = 1)
Range
Average
53.54-57.91 0-0.04 25.52-29.10 0-0.04 0.22-0.47 0.01-0.03 0-0.05 8.47-12.51 4.67-6.54 0.17-0.37 0-0.04 97.67-100.52
56.23 0.02 26.70 0.03 0.29 0.03 0.02 9.59 5.87 0.28 0.02 99.05
63.99 0.00 17.72 0.04 0.10 0.04 0.00 0.01 0.82 15.63 0.02 98.37
Quartz (n = 7) Range
Average
81.66-101.01 0-0.62 0-0.72 0.01-0.04 0.02-1.61 0.01-0.03 0-0.22 0.01-0.32 0-0.56 0-0.09 0-0.04 82.68-101.11
89.51 0.11 0.13 0.02 0.30 0.02 0.05 0.17 0.15 0.03 0.02 90.51
108
A. I. KARAYIGIT & M. K. G. WHATELEY Vitrinite
"7 *i "7 ~-/
\* \*
o
\*
. . . . . . . . .
o 4 Liptinite
8
12
16
20
z4
28
'~ 32
3~ 40 Inertinite
Fig. 4. Locations o f petrographic constituents in the form o f a ternary diagram on a mineral-matter free basis.
given in Table 4, and the constituents are also presented in the form of a triangular diagram on a mineral-matter free basis (Fig. 4). The average maceral content of all the samples on a mineralmatter free basis is 90.5% vitrinite, 6.0% liptinite, 3.5% inertinite (Table 4). As shown by Fig. 4 the petrographic compositions of all but one sample (which was from Coal Mine I, for location see Fig. lc) are similar to each other. The dominant vitrinite group maceral is desmocollinite (Fig. 3b and c; Table 4). Desmocollinite is the dominant vitrinite type of reed peats, which accumulated in reed swamps with grasses, sedges and ferns and in general require a higher water table than do forest swamps (Stach et al. 1982). The vitrinite group macerals derive from organic precursors formed at relatively low redox potentials (Eh). This suggests that the coals studied were developed in a low Eh. Total liptinite varies from 1.7 to 18.8 % on a mineral matter-free basis with the most important maceral of this group being liptodetrinite (Table 4). In addition to the liptinite macerals presented in Table 4 a minor amount of alginite (or telalginite) was found. Sporinites have generally weak yellowish orange and brown fluorescence. Cutinites are very thin in the samples and show very weak orange brownreddish brown fluorescence. Resinites have yellow-orange brown fluorescence and in general occur together with corpocollinite and fluorinite surrounded by cutinites. Fluorinite, which probably forms from plant oils (Teichmfiller 1974), shows strong greenish yellow-yellow fluores-
cence colour. Exsudatinite is generally observed as yellow, yellowish orange, orange and brown. Alginite shows very strong green fluorescence colour. Bituminite shows yellow-yellowish orange, and occurs as a groundmass to alginite and the other macerals, and is mainly found in a finely dispersed form mixed with mineral matter. Liptodetrinite is composed of small masses of liptinite group macerals which are too small to identify as a particular maceral; in many cases in the coals such masses may be small decomposition products of alginite. The presence of alginite in coal has an important paleoenvironment significance in reconstructing the ancient coal swamps. In general, alginite and its decomposition products indicate wet conditions in intermontane and platform basins rather than in foredeep basins, and alginite-bearing coals can be interpreted as subaquatic (Hagemann & Wolf 1989). More aerobic conditions and oxidative biodegradation produce maceral precursors of the inertinite group from mainly macerals in the vitrinite group (Stach et al. 1982). However, the amount of inertinite macerals in the coals is negligible, and inertodetrinite is slightly more abundant than the other inertinite macerals (Table 4). This result indicates that aeorobic microorganisms and fires were not significant factors in the peat environment. In the coals, some minerals show similar amounts by both XRD analyses and coal microscopy (Table 1 and 4). During petrographic analyses, carbonate minerals, which are mainly calcite and minor amounts of dolomite determined by XRD analyses (Table 1), were evaluated in two groups, as early diagenetic (syngenetic) and late diagenetic (epigenetic). The early diagenetic carbonate minerals are generally brownish-gray in reflected light and have weak yellowish fluorescence colour and irregular shape in a clearly microlayered desmocollinite in all the samples (Fig. 3b and c). The weak fluorescence colour probably indicates organic materials within them. Whereas, the late diagenetic carbonate minerals were formed in coal cleats, and they are black-gray with common polysynthetic twinning in reflected light and no fluorescence colour. They were not included in the petrographic analysis. Thus, the carbonate minerals given in Table 4 are only early diagenetic. Clay minerals and quartz were black in reflected light. Some quartz grains are rounded and subrounded, have a 20-40#m diameter and are of detrital origin. Framboidal pyrite mostly form circular shaped framboids between 10-20 #m in diameter. Crystalline pyrite consists
CONTINENTAL MICRITIC LIMESTONES, TURKEY
109
Table 4. The petrographic constituents of the coals, range and average values of their volume percentages, and also random reflectances of telocollinite and desmocollinite
Macceral analysis and %Rr measurement
Range
Average
Telocollinite Desmocollinite Corpocollnite Vitrodetrinite
1.8 (2.4)*-14.1 (15.6) 29.6 (52.9)-64.6 (81.6) 3.6 (5.9)-15.2 (18.8) 0.7(0.8)-5.6 (9.2)
5.7 (7.9) 49.1 (67.8) 8.0 (10.9) 2.7 (4.0)
Vitrinite
43.6 (74.5)-85.0 (95.3)
65.5 (90.5)
Sporinite Cutinite Resinite Fluorininte Exsudatinite Bituminite Liptodetrinite
0.0-1.0 (1.4) 0.0-2.9 (3.4) 0.0-1.5 (1.9) 0.0-1.7 (2.0) 0.0-1.8 (2.3) 0.0-1.0 (1.1) 0.9 (1.3)-9.8 (11.2) 0.9 (1.7)-16.4 (18.8)
Liptinite
Fusinite Semifusinite Micrinite Macrinite Sclerotinite Inertodetrinite Inertinite
0.3 (0.4) 0.6 (0.7) 0.2 (0.2) 0.1 (0.1) 0.3 (0.4) 0.1 (0.1) 3.1 (4.1) 4.5 (6.0)
0.0-2.2 (2.7) 0.0-0.4 (0.6) 0.0-1.5 (1.8) 0.0-1.4 (2.4) 0.0-0.5 (0.7) 0.0-2.5 (2.8)
0.6 (0.8) 0.0 0.5 (0.7) 0.4 (0.6) 0.1 (0.1) 0.8 (1.2)
0.9 (1.1)-5.9 (6.7)
2.5 (3.5)
Carbonate mineralst Clay minerals + quartz Framboidal pyrite Crystal pyrite Massive pyrite
2.4-44.8 1.3-16.9 0.7-7.8 0.5-3.8 0.0-4.1
17.2 5.1 3.2 1.6 0.4
Mineral matter
7.9-53.7
27.4
Random reflectance measurement ( % Rr )
%Rr in telocollinite %Rr in desmocollinite
0.68-0.79 0.50~.65
0.77 0.61
* Figures in brackets are the values in a mineral matter-free basis. t Carbonate minerals include mainly calcite and minor amounts of dolomite. %Rr: Mean random reflectance of vitrinite for every coal sample
of individual euhedral shaped pyrite 1-5 #m in diameter. Massive pyrite was observed mainly in fracture surfaces of the macerals. Pyrite is mainly found as framboidal and crystal pyrite with relatively lower amounts of massive pyrite (Table 4). The formation of the massive pyrite and late diagenetic carbonate minerals can be related to the late diagenetic cleat mineralization, which was probably formed by deep burial during the deposition of the Hanioglu unit and later modified by volcanic activity in the coal field. Finally, the abundant desmocollinites in the vitrinite group, minor amounts of the inertinite group macerals and the presence of alginite (telalginite) and bituminite (or lamalginite) in the coals indicate that the conditions remained
subaqueous throughout the coal and carbonate sedimentation in the peat swamp. It is possible that the Upper Carboniferous-Lower Cretaceous(?) Munzur limestones, and limestone-marl alternations of the Kozluca Formation of Ypresian-Lutetian age (Tunc et al. 1991), which are located around the study area, may have played an important role as source rocks in the formation of the limestone-bearing sequence in the coal field. The influx of calcium-rich waters into the swamp reduced the acidity of the peat to a much greater degree than would have sea water. Bacterial activity is accelerated, resulting in increased degradation of plant remains. Most calcium-rich coals almost always show the characteristics of subaquatic genesis and are remarkably high in
110
A. I. KARAYIGIT & M. K. G. WHATELEY
organic sulphur and syngenetic pyrite, probably due to severe bacterial activity and abundant supply of protein-rich substances (Stach et al. 1982). In the light of this, the Selimoglu swamp environment may have had high pH and low Eh conditions, and an accelerated bacterial activity resulting in HzS reacting with Fe z+ and organic peat components to form pyrite and organic sulphur compounds. In addition, it is also possible that the HzS formation could have been accelerated by ground water carrying sulphate derived from the thick gypsum-bearing rocks around the Divrigi region. An additional study on sulphur isotopic ratios for evaporites and pyrite in the coals would be needed to solve the problem in detail.
maxima (/~max) values of sporinites are less than 580nm in lignites and subbituminous coals, whereas for bituminous coals this value is higher (Robert 1981). In this study, the spectral maxima of 17 different sporinites are between 625-661 nm, their fluorescence intensities at spectral maxima are medium or weak (0.029-0.248 when calibrated to unity), and their Q (red/green ratio) values show a relatively broad range (1.35-2.69) (Table 5). Based on the coal classification given by Robert (1981), the range values of the "~max of sporinites indicate a high volatile bituminous coal stage. This rank, as explained below, is similar to the coal stage determined by %Rr of telocollinite.
Fluorescence spectroscopy
Reflectance measurement and rank
The fluorescence spectra of sporinite, cutinite, resinite, fluorinite, exsudatinite and alginite were measured, and Table 5 summarizes the range of values of /~max and red/green ratio (Q). The lowest values of the '~max and red/green ratios were measured in alginite (Table 4), because of its very strong green fluorescence. Typical examples of fluorescence spectral curves of some liptinite macerals are presented in Fig. 5. Similar curve trends are seen only between cutinite and resinite; the other curves are generally different for each liptinite maceral (Fig. 5). It appears that it is possible to identify some liptinite macerals in Turkish coals using these given typical spectral curves. The fluorescence of sporinite varies steadily with rank from green to yellow, then to brown red; the fluorescence measurement constitutes a coal rank parameter, for example, the spectral
The random reflectance of telocollinite and desmocollinite, which are submacerals of the vitrinite group, were measured mainly at a minimum 50 points on each coal sample and the mean values for every coal sample were calculated (Table 4). The range and average values of the random reflectance of desmocollinite are lower than for telocollinite. The relatively weak reflectance of desmocollinite is probably due to decomposition products of cellulose, and cellulose-rich huminites/vitrinites showing particularly weak reflectances (Stach et al. 1982). The average value of all the measurements for desmocollinite and telocollinite is 0.61%Rr and 0.77%Rr, respectively (Table 4). The 0.77% Rr value of telocollinite shows a high volatile bituminous coal rank in the ASTM classification given by Stach et al. (1982).
Table 5. The range values of the wavelength of the maximum intensity in nm (Amax)and red/green ratio (Q) of some liptinite macerals Maceral Sporinite Cutinite Resinite Fluorinite Exsudatinite Alginite
n
17 4 8 14 12 2
Range
Red/Green Ratio, Q Range
625-661 645-676 592-657 593-632 598-651 475-553
1.35 (3.00)-2.69 (16.00) 1.54 (4.00)-1.96 (8.50) 1.31 (2.58)-2.20(10.20) 1.09 (1.43)-1.42 (3.79) 1.25 (2.23)-2.31 (13.50 0.81 (0.41)-0.89 (0.60)
/~max
n: number of measurements. Amax:The wavelength of the maximum intensity in nm. Figures in brackets are the arithmetic values.
CONTINENTAL MICRITIC LIMESTONES, TURKEY 100
111
1~ / Cutinite
Spodnite
8O .~_~70
~00
_=
~50 .~ 40 -~aO or 20 10t0 ,
i
0 450
500
550 600 Wavelength (nm!
650
,
450
700
500
9
550
~
6O0
650
7O0
Wavelength (nrn) 100
100
90-
Resinite
~80
o~80
70
70-
~00-
~50
~>=4o-~30n~ 2 0 -
rr 20
100
0 450
5O0
550
6O0
650
7OO
500
450
Wavelength (nm)
550 600 Wavelength (nm)
650
7OO
100
100
90-
90 8O
70
70 -
~00-
_~50 >=4o
_50-
~r 3 0 of 20
N30.
_,
2010-
0 450
I 500
P P 550 600 Wavelength (nm)
,, ' ~" , 650
0 700
I
450
500
i
;
550 600 Wavelength(nm)
;
650
7~
Fig. 5. Typical examples of fluorescence spectral curves of some liptinite macerals from the mining seam in the Selimoglu coal field.
Most Miocene coals in Turkey are of lignite or subbituminous coal rank, whereas the investigated coals have a high volatile bituminous rank. The results obtained from %Rr of telocollinite and spectral maxima (625-661 nm) of sporinites, the low moisture content (1.68% average), high calorific values (5606kcalkg -1 average) and agglomerating characters of the coals suggest that the thermal history may
have been affected by the volcanic activity that occurred in the coal field.
Conclusions In the Selimoglu coal field, only one coal seam, which is associated with thin continental micritic limestones, is exploited by underground mining.
112
A. I. KARAYIGIT & M. K. G. WHATELEY
It has a macroscopically bright appearance and a working thickness of 50-90cm (70 cm average), and it was probably accumulated in a small, shallow lake or pond on an alluvial plain. The coal seam is characterized by low moisture content (1.68% average), high ash yield (24.04% average), high total sulphur content (up to 8.92%) and high calorific value (5606kcal kg -1 average). The maceral and mineral matter contents on a mineral-matter free basis average 90.5% vitrinite, 6.0% liptinite and 3.5% inertinite. It is thought that peat formation may have been developed in subaquatic conditions with high pH and low Eh and an accelerated bacterial activity. Most Miocene coals in Turkey are of lignite or subbituminous coal rank, whereas the investigated coals are of a high volatile bituminous rank. It is possible that the thermal history may have been affected by volcanic activity that occurred in the coal field. We acknowledge the Turkish Scientific and Research Council (TUBITAK) for supported first writer's research project (TBAG/YBAG-948), the British Council of Turkey who supported Karayigit's expenses in UK and for the help given by S. Toprak with the microscope for spectral analyses and C. Tunoglu with osctracod studies. Our thanks also to R. Wilson for the help with the electron microprobe studies.
References ASTM 1991. Annual Book of A S T M Standards, Gaseous Fuels; Coal and Coke. 1916 Race Street, Philadelphia, PA 19103, 05.05. CASAGRANDE, D., SIEFERT, L., BERSCHINSKI, C. & SUTTON, N. 1977. Sulfur in peat forming systems of Okefenokee Swamp and Florida Everglades: Origins of sulfur in coals. Geochimica et Cosmochimica Acta, 41, 161-167. COHEN, A. O., SPACKMAN, W. & DOLSEN, P. 1984. Occurrence and distribution of sulfur in peatforming environments of southern Florida. International Journal of Coal Geology, 4, 73-96. CRELLING, J. C. 1983. Current uses of fluorescence microscopy in coal petrology. Journal of Microscopy, 132, 132-147. GIERLOWSKI-KORDESH, E., GOMEZ FERNANDEZ,J. C. & MELI~NDEZ, N. 1991. Carbonate and coal deposition in an alluvial-lacustrine setting: Lower Cretaceous (Weald) in the Iberian Range (east-central Spain). International Association of Sedimentology, Special Publication, 13, 109-125. GIVEN, P. H. & MILLER, R. N. 1985. Distribution of forms of sulfur in peat from saline environments in the Florida Everglades. International Journal of Coal Geology, 5, 397-409. HAGEMANN, H. W. & WOLF, M. 1989. Paleoenvironments of lacustrine coals- the occurrence of algae in humic coals. In: LYONS, P. C. & ALPERN, B.
(eds) Peat and CoaL" Origin, Facies, and Depositional Models. International Journal of Coal Geology, 12, 511-522. ICCP, 1963; 1971. Internationales Lexikon Ffir Kohlenpetrologie. Centre National de la Recherche Scientifique 15, Quai-Anatole-France, Paris. JACOB H. 1964. Neue Erkenntnisse auf dem Gebiet der Lumineszenmikroskopie fossiler Brennstoffe.Fortschr. Geol. Rheinld. u. Westf. 12, 569-588, Krefeld. 1973. Kombination yon Fluoreszenz-und Reflexions-Mikroskopphotometrie der organischen Stoffe yon Sedimenten und Boden.-Leitz-Mitt. Wiss. u. Techn. VI, 1 21-27, Frankfurt. KARAYIGIT, A. I. 1993. Geological and sedimentological investigation of the Selimoglu (Divrigi-Sivas) basin, and chemical-petrographic properties of the coals. Project no: TBAG 948/YBAG 15, TUBITAK, Earth Sciences Research Grant Committee [in Turkish]. KESKIN, E., GI3RSOY, N. & GI2RSOY, B. 1984. Geology of the Sivas-Divrigi (Selimoglu-Mursal) area. MTA Report No: 7616. [In Turkish] OTTENJANN, K., TEICHMOLLER,M. & WOLF, M. 1975. Spectral fluorescence measurements of sporinites in reflected light and their applicability for coalification studies. In: ALPERN, B. (ed.) P~trographie de la MatiOre Organique des Sediments, Relation avec la Paleotemperature et le Potentiel Petrolier, Paris, 67-95. PFLUG, H. D. 1966. Fluoreszenzmessungen an Gesteinen und Fossilien.-Leitz-Mitt. Wiss. u. Techn. II, 6 Frankfurt, 183-188. ROBERT, P. 1981. Classification of organic matter by means of fluorescence; application to hydrocarbon source rocks. #International Journal of Coal Geology, l, 101-137. ROBERTS, D. L. 1988. The relationship between macerals and sulphur content of some South African Permian coals. International Journal of Coal Geology, 10, 399-410. STACH, E., MACKOWSKY,M.-TH., TEICHMULLER,M., TAYLOR, G. H., CHANDRA,D. & TEICHMf3LLER,R. 1982. Stach's Textbook of Coal Petrology. Gebruder Borntraeger, Berlin. TEERMAN, S. C., CRELLING,J. C. & GLASS, G. B. 1987. Fluorescence spectral analysis of resinite macerals from coals of the Hanna Formation. Wyoming, USA. International Journal of Coal Geology, 7, 315-334. TEICHM(]LLER, M. 1974. Uber neue Macerale der Liptinit-Gruppe und die Entstehung des Micrites.Fortschr. Geol. Rheinld. u. Westf 24, 37-64. OTTENJANN, K. 1977. Liptinite und lipoide Stoffe in einem ErdO'lmuttergestein.-ErdG1 u. Kohle, 30, 387-398. -~ WOLF, M. 1977. Application of fluorescence microscopy in coal petrology and oil exploration. Journal of Microscopy, 109, 49-73, London. TUNG, M., OZCELIK, O., TUTKUN, Z. & GOKCE, A. 1991. Basic geological characteristics of the Divrigi-Yakuplu-Ilic-Hamo (Sivas) area. Doga Turkish Journal of Engineering and Environmental Sciences, 15:2, 225-245. [In Turkish]
C O N T I N E N T A L MICRITIC LIMESTONES, T U R K E Y UNSWORTH, J. F., BARRATT, D. J. & ROBERTS, P. T. 1991. Coal quality and combustion performance. An International Perspective. Coal Science and Technology, 19, Elsevier, Amsterdam.
113
WEDDING, E., 1965. A report on the Divrigi (Sivas) lignite basin. Directorate of Mineral Research and Exploration, Report No: 3774, [in Turkish translation].
Chemical characteristics, mineralogical composition and rank of high sulphur coking coals of Middle Miocene age in the G6kler coal field, Gediz, Turkey A. I. K A R A Y I G I T l & M. K. G. W H A T E L E Y 2
1Department of Geological Engineering, Hacettepe University, Beytepe-Ankara, Turkey 2 Department of Geology, University of Leicester, University Road, Leicester LE1 7RH, UK Abstract: Most Miocene coals in Turkey are of subbituminous to lignite rank. The G6kler
coal field in the western part of Turkey contains mainly high sulphur coals of bituminous rank. The chemical characteristics, mineralogical composition and rank of high sulphur coking coals of Middle Miocene age in the coal field are investigated for the first time. A total of 46 channel and core samples were collected from underground mine workings and from boreholes drilled in the coal field. The results of the proximate analyses as well as total sulphur analyses and calorific values on an air-dried basis show on average 1.2% moisture, 22.9% ash, 34.6% volatile matter, 6.9% total sulphur contents and 5850kcalkg -1 calorific value. X-ray powder diffraction studies of the coal samples on an air-dried basis show quartz, pyrite and calcite to be the dominant minerals; kaolinite, hydromuscovite, dolomite, gypsum, iron sulphate hydrate and rarely illite/smectite and feldspar constitute the remainder. Secondary calcite in random fractures surfaces of the coals is especially abundant in samples obtained from an area adjacent to the fault zones. The mean random reflectance values (%Rr) of telocollinite vary between 0.50 and 0.95%. These values show that the rank can be determined as a high volatile bituminous stage. In addition, these coals can form isotropic coke. Fluorescence intensities of sporinite are weak to very weak. The mean random vitrinite reflectance values within the coal field generally increase towards southern parts of the coal field. It is thought that this increase can be related to the recent hydrothermal antimony mineralization in the southeastern parts of the coal field.
The coal field is located east of Gediz in the northwestern part of the Muratdagi region (Fig. 1). In Turkey, the coal and coal-bearing strata have been studied extensively by the General Directorate of Mineral Resource and Exploration (MTA) and Turkish Coal Enterprise. Most Miocene coals of Turkey, for example Krtahya-Seyit6mer and Tuncbilek, Canakkale-Can, Manisa-Soma, Mugla-Yatagan and Ankara-Beypazari are of subbituminous to lignite coal rank depending on their chemical properties with calorific values which range from 1900 to 3500kcalkg -1. The investigated G6kler coals have a calorific value greater than 5200kcalkg-aand sulphur contents of more than 5% according to unpublished reports prepared by MTA. The Gediz region contains important mineral deposits like borax and antimony, as well as coal deposits. Some studies on the regional geology, coal geology and antimony mineralization have been made by Atabek (1939), Kalafatcioglu (1961), Akkus (1962), Lebktichner (I965), G6kmen (1970), Grin (1975), Bing61 (1977), K6ksoy & Ileri (1977), G6kce (1987), K6ksoy et al. (1987) and Aral (1989). The stratigraphy, petrological properties and geochronology of
rocks of the Muratdagi region were investigated by Bing61 (1977) in detail, and also, for the first time, Bing61 determined the age of the G6kler coals as Middle Miocene age using palynological studies. G6kce (1987) investigated the geology of the antimony mineralization found in the Muratdagi region. He proposed that the antimony mineralization formed from the hydrothermal solutions which are still precipitating antimony at the present time. Aral (1989) indicated that all host rocks were first strongly silicified and open spaces were lined with crystalline quartz prior to mineralization. Mineralization is of two types: antimonite with pyrite and marcasite as in the G6yntik mine, and antimonite with no other sulphides as in the Derek@ mine including the Sakarciburnu and Karciolukpinari mineralizations (Fig. 1). Mercury minerals are absent. Trace amounts of arsenopyrite and sphalerite are present at the G6ynfik mine. Gangue minerals observed in both mineralization stages are quartz, chalcedony, opal, calcite, clay minerals, and fuchsite. Major secondary antimony minerals are antimony oxides such as valentinite, cervantinite, kermesite, and metaantimonite (Aral 1989). The detailed geology and economic potential of the
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 115-130.
116
A. I. K A R A Y I G I T
& M. K. G. W H A T E L E Y
o . w
9~.,
~
co
~o
{~
o
Eo~
.~=
~
~
121
v
~
.
|
._
~.~,
~o ~ o~ ~- i ~o ~.
E
0
".~_
Jaddt~)
(~.ssl~JnF)
~ E
0
l--
oo 0 o
s
N
N~
0
F~ "6
oo I
N_e
:0
0
.=.
o
~
m
m
m
m
~
m
mmmmmm immm
m m ~ m m
E o o o o,-~l 0.0 : o
. ,...~ ,,~
~N
%o
.~ g
HIGH SULPHUR COKING COALS, TURKEY northeastern part of the coal field was studied by K6ksoy et al. (1987), and they proposed the generalized stratigraphical sequence of the coal field (Fig. 2). In the coal field, total coal production is about 1 Mt per year, which is sold mainly to cement
z 0
03 u) uJ
o LL
"1I---
LU
; o .o,ooY
o
0
"
~
.
I~ ~ - . ' - ' . -
factories. The overall objective of this study was to determine the chemical properties, mineralogical composition and rank of the high sulphur G6kler coals, and explain why this coal, which has high calorific values and swelling indices between 5 and 8.5, is used for industrial purposes.
EXPLANATION Alluvium Gray-white sandstone with cross-bedding and sericites UNCONFORMITY
7i+ r ~
Gray-white conglomerate interbedded with sandstone, claystone and limestone
." -,'~
UNCONFORMITY
Fossiliferous, yellow-white, lacustrine limestone ,
!
Gray sandstone-claystone with gypsum (Gyp)
..
iii
z
;--I~
Z l,i,l
0
L;~+...~-., ~-_-_,+..*+..-_
0
~;
ILl J 0 E3
" n; ,,,
_
.
E',
I--
12
Light brown, thin bedding sandstone
...................
: ~, it)
Greenish claystone
--
'~,
~
.
.
.
Claystone-marl with common seficites
.
B(~yOk seam Claystone Clayey dolomitic limestone with ostracode KBcLik seam Gray-light brown sandstone-claystone with sericites Reddish poligenic conglomerate UNCONFORMITY Kmk-1 :Mafic and ultramafic rocks Kmk-2: Metedetrital rocks Kmk-3: Limestone and marble o
t.<s,
s
117
i
Fig. 2. The generalized stratigraphical sequence of the G6kler coal field (modified after K6ksoy et al. 1987).
118
A. I. KARAYIGIT & M. K. G. WHATELEY
Geological setting Rocks which range in age from Jurassic to Quaternary crop out in the Muratdagi region. In this region (Fig. 1), the Asagibelova Formation of Jurassic age, Muratdagi Melange of Cretaceous age, Paleocene Baklan Granite, Miocene Karacahisar volcanics (mainly rhyolite) and coal-bearing G6kler Formation, Pliocene Karsakatepe deposits and Quaternary G6cfiktepe sediments and alluvium were identified by G6kce (1987). The Karacahisar volcanics were also considered as Middle Miocene age and were deposited contemporaneously with the coalbearing sediments in the Muratdagi region (G6kce 1987). However, our field studies have failed to locate the volcanics in the coal field. In the basement immediately below the coal field, serpentinized mafic and ultramafic rocks, metadetrital rocks, that consist mainly of quartz and muscovite, and different types of limestone
and marble of the Muratdagi Melange crop out. The bedding dips of the G6kler Formation are 30-50 ~ at the boundary of basement rocks, and 10-15 ~ below the Karsakatepe deposits. All the basement rocks and the coal-bearing G6kler Formation were cut by a number of normal faults which strike NW-SE, and by less frequent strike-slip faults cutting the normal faults (Fig. 1) (G6kce 1987). In the northeastern part of the coal field a small syncline and anticline were determined from underground maps and shown on iso-reflectance map (see Fig. 8). Tectonically the area is still active as evidenced by the occurrence of strong earthquakes in the region; the last one occurred on March 28, 1970. The coal seams are located at the base of the G6kler Formation (Fig. 2), and examples of their macroscopic seam sections are shown in Fig. 3. There are two coal seams, the upper Bfiyfik and the lower Kticfik seams. They are separated by about 3 m of black claystone and
(roof)
[~
Coal Brecciated coal Clayey dolomitic limestone Claystone Claystone with coal Borehole missing
5O
0
1
(floor)
BOyi3k seam
Ki.icCik seam
Unnamed seam
Fig. 3. Examples of macroscopical seam sections of the Biiyfik, Kiiciak and unnamed seams.
HIGH SULPHUR COKING COALS, TURKEY brown to dark brown clayey dolomitic limestone with ostracode that is laterally extensive and 50-100 cm thick (Fig. 2). The presence of clayey dolomitic limestone as an intra-seam unit is some indication of limited acid conditions during peat accumulation. Towards the coal field margins this parting interval gradually decreases and the coal seams amalgamate into a single coal seam. In this study, this single seam, coal seams in some boreholes and some undefined coal seams in the underground mines which cannot clearly be differentiated are referred to as the unnamed seam (Fig. 3). The B~iyiik seam appears to be dull because of black claystone partings in the seam (Fig. 3). The Kticfik seam is bright and contains less mineral partings than the Btiytik seam (Fig. 3). The Biiytik and Kfictik seams average 170 cm and 120 cm thick, respectively. The floor rock of the Ktictik seam is green claystone that includes rare sericites. The roof rock of the Bfiyiik seam is claystone-marl with common sericite.
Methods A total of 46 channel and core coal samples were collected from the Grkler coal field (for sample locations, see Fig. 8). Thirty seven samples were collected from the underground mine workings and 9 from core samples from the 6 boreholes. All samples were mechanically crushed to 1 mm size and split into two subsamples (ASTM 1991). One sample was used for petrographic briquette preparation and embedded in cold setting epofix resin, and then polished. The other was further reduced to 60 mesh (250#m) for proximate analysis (moisture, ash and volatile matter) as well as total sulphur analysis and calorific value of all samples on an air-dried basis. In addition, for some evaluations, volatile matter and calorific values on a dry, ash-free basis were calculated from the results on an airdried basis. Sulphur forms and Free Swelling Index (FSI) tests were determined on only six samples selected from the samples. The analyses were made according to the ASTM (1991) procedure and presented as weight percentages, except calorific values. Random reflectances of the maceral vitrinite group telocollinite were measured at a minimum of 50 points on every coal polished briquette (ICCP 1963; 1971; Stach et al. 1982). A Leitz MPV II microscope with a 50x oil immersion (n:1.518) objective, sapphire (0.551%R) and glass (1.23%R) standards for reflectance calibration were used. During spectral fluorescence measurements of some liptinite macerals, a Leitz
119
MPV-SP microscope equipped with a highpressure 100WHg light source, a BG38 and a UG 1 filter, and a K460 barrier filter was used. This microscope system was computerised so that spectral intensities in the range of 460-700nm could be measured and corrected spectral intensities automatically produced. Some numerical parameters, such as relative intensities, wavelength of maximum intensity (Amax) and the logarithmic ratio (Q) of the relative intensity of red (650nm) and green (500 nm), were then generated from them. The principles, basic calibration techniques and some applications of measuring fluorescence in geological samples are given by Jacob (1964; 1973), Ottenjann et al. (1975), Teichmtiller & Ottenjann (1977), Teichmtiller & Wolf (1977), Crelling (1983) and Teerman et al. (1987). All coal samples on an air-dried basis, except two samples with insufficient coal, were analysed by X-ray powder diffraction (XRD) methods. After identification of the whole-coal minerals on each XRD diagram, the nett area under each diagnostic peak for every mineral was determined and then the nett area value was converted to a percentage. In order to determine the chemical composition of some silicate minerals, two coal polished briquettes were selected, one from the Btiyfik (sample no: GM-10) and one from the Kticfik (sample no: C-94) seams, which were examined on JEOL8250 electron microprobe. The mineralogic composition (by weight percentage) determined on the electron microprobe was grouped according to our interpretation of the minerals. The information given by Newman & Brown (1987) was used for interpretation of the clay minerals by their chemical composition.
Results and discussion Chemical character&tics
Table 1 summarizes the results of the chemical characteristics and XRD analyses on an airdried basis and random reflectance measurements of all coal samples. In addition, the histograms of some analytical results are shown in Fig. 4. The moisture contents of the coals on an airdried basis are low and average 1.2% (Table 1). In addition, it was determined that the moisture content of the coals on an as-received basis is less than 5% according to unpublished reports of the coal mining companies. The ash yields show a broad variation ranging from 10.951.2% but they are generally less than 40%. The
120
A. I. KARAYIGIT & M. K. G. WHATELEY
Table 1. The chemical characteristics, mineralogical composition and vitrinite reflectance measurements of the Bfiyfik, Kficfik and unnamed seams in the G6kler coalfield ii
BiJy~Jkseam Analyses
KOcOkseam
19 sample Range
m
16 sam lep~__~ Range
L
~
~sh %
~
10.9-35.3 ~
1 2 . 7 ~
tolatile matter %
~. 23,1-41,2 (33.7)
33.6-41,7~
28.1-43.8 (32.7)
34.6
0.7-8.7 ~
6.1-10.1 (8.0)
6,9
~
% ~'
%
Calorific value, kcal k -1
~
2.2-3.5
1.14.1
2.1-2.9
2.7
0.5-0,8
0.2-1.1
0.3-0.7
0.6
2.2-2.5
3.8-4.0
3,1-3.7
3.2
2977-7169 (5392)
585O
tI
3591-6887 5668 ~
15.03-28.83 (23.73) 22.80-30.40 (26.7._~ 12.46-30.02(22.58)
Calorific value, MJ index*
+
7.0-7.5
1.0-5.0
~
Ilite/smectite
90-..__3.3 (0.~_)
~"
~dromuscovite
90-..._.._~7 (4)
E
~yn~um
)
~uartz
9
12-86(59)
)oiomite e._.yrite
~
~ d r a t e
+
n
~
%Rr
~
8.0-8.5 0-9 2) 0
~
0-7 ( 2 )
a_2_
~
0-6
20-75 ~ ~
1
3-84 ( 5 0 ~ _ 0
0
0-82 (10)
~
0-94 (16)
15
0-20 (2)
~
0-31 ~6)
3
~
2-~..~~
1_L.6
0-17 (5)
5
0-16 (4)
0-13 ~
~ ~
50-60 0.50-~
o~ ;~ Standard deviation I 0.02-0.05 (0.03) 0.02-0,05(0.04) . 0.03-0.05,(0.04) *: Analysis of only six samples; Hydromuscovite here describes illite, muscovite and illite Figures in brackets are the average values n: Number of the random reflectance measurements measured on every coal sample %Rr: Mean random reflectance of telocoUinite for every coal sample
r
24.49
0
3___.~.~____~ 0-
0
Calcite
t
0-6 (3
o-6~
Felds ar
=~ ~
1.2
0 . 5 - ~
Or anic sul hut* %
;
~ for~e , coal field
0.7-2.6~
~ h u r *
~
11 sample Range
~
~ u r *
~
Average
Jloisture %
Total sulphur %
~,
iiiii
?
Unnamedseam
5a 0.7__~2 0.04
9
Kticiik seam has a relatively lower average ash yield than the others (Table 1 and Fig. 4a). The volatile matter contents of the coal samples, like their ash yields, show a broad range of 23.1-43.8% (average 34.6%), but most of them are 27.5-42.5% (Fig. 4c). The calcu-
lated volatile matter contents on a dry, ash-free basis for the Biiytik, Ktictik and unnamed seams range 37.6-62.9% (average 46.4%), 40.8-57.1% (average 45.6%) and 36.8-71.1% (average 46.2%), respectively. These values in general vary between 35-55% (Fig. 4d), and the range of
HIGH SULPHUR COKING COALS, TURKEY 50
121
50-
45
Ash %, (adb)
Total sulphur %,
IB BQy(Jk s e a m
45 .
(~Ib)
I-I KQcLik s e a m 9 Unn, a m e d s e a m
35 ~ 30 20
1
35. 30. ~
20'
15
15 -
10
10 9
,
o
IH
, 15
20
25
30
(a)
35 Class
',
40
45
il 50
o. 55
1
2
3
4
5
(b)
6 Class
7
8
9
10
11
60Volatile matter %, (adb)
45 4O
Volatile matter %, (daf)
50-
35
~
4 0 -
25 u_ 20
~20.
15 10
10-
5 22.5
25
27.5
(c)
30
32.5 35 Class
37.5
40
42.5
35
45
40
45
(d)
50
55 Class
60
65
70
75
6045
Calorific v a l u e
55-
Calorific value
40
kcal kg ' , (adb)
50-
kcal kg-1, (dad)
45-
35
40-
~'35.
~L. -
~25
25" ~20. 15105O-
"20 15 10 5 0
(e)
cl=,,~
(f)
c=~
'adb): A i r - d r i e d basis; (daf): Dry, a s h - f r e e basis
Fig. 4. Histograms of (a) ash yields, (b) total sulphur contents, (c) and (d) volatile matter contents, (e) and (f) calorific values.
values suggests a subbituminous to bituminous coal rank according to the classification given by Unsworth et al. (1991). On the other hand during the volatile matter determination, the coals produced an agglomerate button showing swelling. The total sulphur contents reach up to 10.1% (Table 1), and most of them are 5-10%
(Fig. 4b). Relatively higher-total sulphur contents were found in the samples collected from the area around the syncline. This indicates that the conditions for formation of sulphur in the swamp environment probably changed laterally. The sulphur forms show that the sulphate sulphur content is only 9% (average 0.6%) of the total sulphur content while the pyritic
122
A. I. KARAYIGIT & M. K. G. WHATELEY
sulphur (average 2.7%) and organic sulphur (average 3.2%) make up respectively 42% and 49% (Table 1). There are many possible origins of sulphur in the precursor peat, such as marine roof rocks (Horne et al. 1978), marine influences during deposition (Casagrande et al. 1977), accelerated microbial degradation as the result of increased swamp water (pH > 6) (Renton & Bird 1991). The coal field is situated in an intermontane basin and there are no marine rocks in both the coal-bearing G6kler Formation and the Karsakatepe deposits. It is possible that the origin of the original sulphate of the G6kler swamp environment related to the volcanic rocks that formed contemporaneously in the Muratdagi region and/or accelerated microbial degradation in increased swamp water (>pH 6). These conditions would also be ideal for the precipitation of the iron disulphide minerals (Renton & Bird 1991). This problem and lateral suphur variation in the coal field are the subject of on-going investigations. The calorific values, like ash yield and volatile matter content, show a broad range of 29777260kcalkg -1 (average 5850kcalkg-1), and most calorific values are 5000-7500kcalkg -1 (Fig. 4e). The Kficfik seam, because of relatively lower ash yields, has a higher average calorific value than the other seams (Table 1). Most of the calorific values on a dry, ash-free basis are 7000-8500kcalkg -1 (Fig. 4f) and the values suggest a bituminous coal rank. FSI or crucible swelling number is a smallscale test for obtaining information about the free-swelling properties of a coal; the results may be used as an indication of the caking characteristics of the coal when burned as a fuel (ASTM 1991). The method is one of the most commonly used tests in the coking industry to determine whether a coal will coke. Coals are generally considered to have coking properties if their FSI is over four, whereas a FSI of seven or more indicates a high quality coking coal. The test is significantly affected by factors such as particle size distribution, heating rate, oxidation, weathering, petrographic composition, and mineral (or ash) content (Carpenter 1988). All the investigated coals, except one low FSI value that was found in a sample with a high ash yield (51.2%) on an air-dried basis, display FSI values between 5 and 8.5 (Table 1). The FSI results indicate that the investigated coals have both caking and quality coke-forming properties, which is unusual in Turkish coals of this age. Further evidences of coking properties of the G6kler coals are seen where spontaneous combustion has resulted in coke formation locally in the underground mine in the north-
eastern part of the coal field. This coke is very strong with good pore structure. A polished briquette of this coke was prepared for petrographic investigation. It was observed that this coke shows isotropic coke structure. Its proximate analysis on an air-dried basis is: Moisture % Ash %: Volatile matter %: Fixed carbon %: Total sulphur %: Combustible sulphur %: Sulpur in ash %: Calorific value (kcal kg-1)
4.4 21.1 12.1 62.4 7.6 5.8 1.8 5541
This coke, naturally, has a lower volatile matter than the coals on an air-dried basis, but total sulphur and moisture contents are slightly higher. The higher moisture content is presumably from pore waters. The combustible sulphur content is still high. These results indicate that these coals are suitable for industrial purposes and not for coking coal in iron and steel manufacture, because the total sulphur content of a metallurgical coke is usually preferred to be less than 1%. In the literature, some similar high volatile, high-fluidity Indian caking coals of Permian and Tertiary age have been referred to as 'abnormal coal' by Chandra et al. (1984). Some characteristics of these abnormal coals (Chaudhuri & Ghose 1990) are compared with the G6kler coals (Table 2). The moisture and volatile matter contents of the Indian and G6kler coals are quite similar. Although the G6kler coals have much higher ash yields than the Indian coals, they show higher FSI values.
Mineral
matter
X R D studies of the whole-coal samples show quartz, pyrite, and calcite in some samples to be the dominant minerals; kaolinite, hydromuscovite, dolomite, gypsum, iron sulphate hydrate, and rarely illite/smectite and feldspar constitute the remainder (Table 1). The chemical composition of silica, kaolinite and hydromuscovite and the relevant detection limits on the electron microprobe are shown in Table 3. The hydromuscovites interpreted from the microprobe results include illite, muscovite or sericite. The chemical compositions of the individual silica particles vary across a broad range (Table 3), but some have a very high SiO2 content.
HIGH SULPHUR COKING COALS, TURKEY In this study it is thought that the particles that have a high SiO2 content are quartz grains in detrital origin. They are believed to have been transported from the metadetrital basement rocks that have the very high quartz and muscovite contents. The remaining silica particles with low SiO2 contents are intimately intergrown Table
123
with small amounts of clay minerals which contain a maximum of 7.49% A1203 and 5.19% FeO content and minor amounts of CaO, Na20 and K20. They may have been derived from an alteration of detrital feldspar and/or muscovite, which are common in basement clastic rocks, and/or plant-derived silica in the mire environ-
2. Comparison of some characteristics of the abnormal Indian coals with the G6kler coals
Moisture
Ash
Voiatile
Free
%
%
Matter %
Swelling
,, (adb)
(dmmf)
Index
(adb)
%Rr
G6kler coals (Middle Miocene) B(lyUk seam
3.4
51.2
44.1
1.0
0.67
BOyfJk seam
1.7
38.8
46.3
5.0
0.74
1.1
14.7
41.3
7.0
0.74
K0c0k seam ,|
9
Ktlctlk s e a m
2.6
15.2
40.5
7.5
0.83
unnamed seam
1.1
30.1
46.2
8.0
0.70
unnamed seam
1.6
28.0
36.0
8.5
0.95
2.0
3.1
43.9
4.0
0.59
1.3
23.0
37.6
5.5
0.86
1.5
16.8
37.5
4.5
0.82
,.
,,
.,
,
,i
Indian coals B
Tikak coal-Tertiary ,.
Bhatdih coal-Permian Ranipur coal-Permian
..
|111
i
i
(adb): Air-dried basis; (dmmf): Dry, mineral matter-free basis %Rr: Mean random reflectance of telocollinite for every coal sample Table 3. The detection limits of electron microprobe for silicate analysis and results of the chemical analyses of silica, kaolinite and hydromuscovite ==
..
Silicate
Detection
Silica (n:22)
Analysis, %
Limit
Range
1
2
3
1
2
3
4
6
SiO2
0.02
45.76-99.67 (81.15)
43.25
46.24
46.46
38.56
42.45
45.43
52.69
54.48
TiO2
0.04
0.00-0.04 (0.02)
0.54
0.31
0.01
0.13
0.20
0.17
0.36
0.17 26.03
Hydromuscovite (n:5)
Kaolinite (n:3)
AI203
0.02
0.00-7.49 (1.36)
30.38
33.05
36.76
27.91
25.57
31.74
24.78
Cr20 ~
0.04
0.00-0.05 (0.02)
0.03
0.02
0.01
0.04
0.02
0.02
0.03
0.04
FeO
0.05
0.03-5.19 (0.43)
1.86
0.39
0.18
1.62
2.06
2.09
1.46
0.89
MnO
0.05
0.00-0.04 (0.02)
0.04
0.03
0.02
0.02
0.04
0.00
0.01
0.03
MgO
0.02
0.00-0.21 (0.03)
0.48
0.04
0.08
0.83
1.31
1.06
0.80
1.06
CaO
0.02
0.00-0.40 (0.09)
0.52
0.23
0.04
0.41
0.51
0.29
0.19
0.52
Na20
0.02
0.00-0.32 (0.10)
0.06
0.21
0.04
0.04
0.09
0.08
0.08
0.11
K20
0.02
0.00-0.43 (0.07)
0.62
0.01
0.56
2.70
4.32
3.00
2.36
2.53
NiO
0.04
0.00-0.07 (0.02)
0.05
0.02
0.00
0.01
0.01
0.07
0.06
0.02
54.13-99.84 (83.31)
77.83
80.55
84.16
72.27
76.58
83.95
82.82
85.88
Total
,.,
,.
....
n: Number of measurements; Fiqures in brackets are the average values
124
A. I. KARAYIGIT & M. K. G. WHATELEY
ment. However, very few feldspars have still been identified in the coal samples (Table 1). The calcite contents determined in 25 samples range from 2-94% and the higher values (>58%) were only obtained from seven samples (C-56, 57, 106, 108; 86-1, 86-3/2, 86-10, for sample location see Fig. 8) located in an area adjacent to the fault zones. Petrographically, these calcites are epigenetic forming in random fracture surfaces of the coals, and so they are termed breccioid coals (Fig. 3). Dolomite was determined in only 11 samples ranging from 2 to 31%, and only 4 samples contain both more calcite and low dolomite, whereas, 7 samples include only dolomite. Petrographically, these dolomites and/ or calcites (>24%) occur within coal macerals and associated with clay minerals. It appears that these are early diagenetic (syngenetic) minerals. The coals contain large amounts of pyrite (Table 1). In addition, gypsum and iron sulphate hydrate have also been determined in a few samples. The sulphate minerals are not normally found in fresh coals, but are commonly seen on mine faces that have been exposed for some time (Ward 1978). This means that the circulation of more recent oxidizing meteoric waters may have caused the oxidation of the pyrite and the resulting sulphate is precipitated as iron sulphate hydrate as secondary minerals in the coal seams. In addition, sulphuric acid produced during the process may react with any carbonate minerals present to form gysum. Kaolinite is the most abundant of the clay minerals determined in the coal samples (Table 1). Kaolinite minerals interpreted from the results of the electron microprobe analysis have slightly lower A1203 contents than ideal kaolinites. The kaolinite is thought to have formed by weathering of feldspars in the acid waters of the coal swamp. The chemical composition of the hydromuscovites (Table 3) contains lower K20 contents than ideal mica or illite. It is thought that these hydromuscovites are detritial in origin and they were transported from the metadetrital rocks. Sericites are also common in the overburden deposits of the coal seams (Fig. 2).
Reflectance measurements, classification and compar&on with some Turk&h coals Random reflectance of telocollinite was measured at a minimum 50 points on each coal sample and the mean value and its standard deviation were calculated (Table 1). The mean random reflectance (%Rr) of telocollinite for all coal samples ranges from 0.50 to 0.95 % (0.72%
average) and the coals can be classified as 'High Volatile Bituminous C-A' and the average value as 'High Volatile Bituminous B' according to the ASTM classification presented by Stach et al. (1982). The average value (0.72%Rr) is compared with the some other coals in Turkey studied by Dogru (1978), Karayigit (1983), Yagmurlu & Karayigit (1984), Demirel (1989), Karayigit & Cicioglu (1994), Karayigit & Eris (1994), Whateley & Tuncali (1995). The distribution and average random reflectance values of huminite/ vitrinite of the compared coals in Turkey are presented in Fig. 5a and b respectively. The G6kler coals have a higher average reflectance value than some very important Turkish lignites, such as Seyit6mer, Tuncbilek and Beypazari. Only the Askale coals of Miocene age have a similar %Rr to the G6kler coals. The Askale coal field is in a very complex tectonic system and located near the Northern Anatolian Fault Zone, which is a major, plate bounding strikeslip fault in Turkey. Dogru (1978) indicated that the Askale coals have been affected by this fault, producing high reflectance values. The high %Rr values of the G6kler coals are discussed below.
Fluorescence spectroscopy Within this study, the fluorescence spectra of sporinites were measured in order to obtain some additional information on the rank of the G6kler coals. During petrographic examination with blue light and an orange barrier filter, it was found that sporinites fluoresce a weak yellowish orange and reddish brown. In addition, some samples having unusually high reflectance values contain exsudatinite maceral showing orange, reddish orange and reddish brown fluorescence. In the samples, exsudatinite is a secondary maceral and it is difficult to identify by reflected light with an oil immersion lens because of its dark colour. Exsudatinites appear to fill voids and cracks in vitrinite. Occasionally, they fill edges of pyrite, especially massive pyrite. The spectral maxima of 17 different sporinites taken from six samples with reflectance values of 0.50-0.76%Rr are between 605-670nm and their fluorescence intensities at spectral maxima are weak to very weak (0.01-0.07 when calibrated to unity), and their Q ratios show a broad range (1.12-4.28). With increasing rank (>0.76% Rr) the sporinites in the samples do not fluoresce. On some samples, some sporinites have variable fluorescence, for example, the
HIGH SULPHUR COKING COALS, TURKEY
BLACK SEA
(!1)
N
MEDITERRANEAN SEA
RANK DIN
125
%Rr ASTM 1.0 0.9-
v,. -
(Glanz-)
Dull Soft
- ~ Sub~ ~; Bit. ~
Lignite
0.11 -
o.7
/
B C
-~
0.6 ~ 0.5
I
'
U
0.4 ""
0.3
0.2
Peat
(To~
Peat I I
r
I
I l
r
I
I
I
I
I
I
I
f
I
I
|
,!| (b)
Fig. 5. (a) The distribution of the compared coals in Turkey, (b) locations of the average %Rr values of huminite/vitrinite for every coal fields in the DIN and ASTM classifications.
spectral maxima and Q values of five sporinites on a single sample (sample no: 86-6) show a range of values between 641-670nm and 1.12-2.64, respectively. This fresh coal sample was taken from the core of the unnamed seam and has the lowest vitrinite reflectance value (0.50% Rr) which has not been influenced by weathering. Weathering of sporinites has been shown to produce different fluorescence values. It is believed that the high fluorescence spectral values may be related to a relatively irregular increase in rank of the sporinite macerals in the same sample at the begining of the bituminous coal rank, which was described as the first coalification jump of liptinites (Stach et al. 1982). It may also be due to different botanical origins or syn-or post-depositional conditions.
Examples of the selected spectral curves of sporinites and exsudatinites are shown in Fig. 6. Sporinite shows many sharp peaks (Fig. 6a). The coal sample with lowest vitrinite reflectance value (0.50% Rr) has a higher wavelength of maximum intensity (666nm) than the sample (621 nm) with a higher vitrinite reflectance value (0.76% Rr). This means that there is not a good relationship between the wavelengths of maximum intensity of sporinites and %Rr, which are used as coal rank parameters. This may be related to the hydrothermal alteration of the coal samples under post-depositional conditions in the coal field. The spectral maxima of exsudatinite measured at 7 points on four different coal samples, which have 0.56-0.84%Rr, are between 602 and
126
A. I. KARAYIGIT & M. K. G. WHATELEY
100 90+ 80 + 70 §
spomize ......
60 +
~i
un (0.76%Rr) un (0.50%Rr)
.',~ 1l
' '=
i
, '~i!
50 t40 § 30 t20 +
"r
10+ I
0 , 500
450
(a)
550 600 Wavelength (nm)
650
700
100 Exsudatinite
,; ',
~, ,'
";
80 70
......
un (0.84%Rr)
;II'\
j' 'V
60 50 40 30 20
,
0 ,
(b)
450
500
550 600 Wavelength (nm)
650
700
Fig. 6. Examples of two spectral curves of sporinite (a) and exsudatinite (b).
654nm. Their fluorescence intensities at the spectral maxima are higher (0.112-0.343) than sporinites. Compared to the sporinites, exsudatinites form fewer sharp peaks (Fig. 6a and b). The dashed line spectrum in Fig. 6b shows a typical spectral curve for exsudatinite. It appears that it is possible to identify exsudatinite in Turkish coals using this typical spectral curve.
Relations among coal rank parameters and coal rank variation in the coalfield Graphical comparison of some chemical and X R D analyses, and % R r results are given in Fig. 7. There is no clear relationship between volatile matter and calorific value on a dry, ashfree basis (Fig. 7a). A plot of calcite and dolomite versus volatile matter content on a dry, ash-free basis (Fig. 7b) shows two different groups. The first group includes lower calcite and dolomite contents, whereas the second group has higher. The early diagenetic carbonate minerals are situated in the first group, but there
is no clear relationship between calcite and dolomite, and volatile matter contents because of analytical errors and X R D analyses. On the other hand, the second group consists of higher volatile matter contents because of epigenetic calcites. Figure 7c shows a weak relationship between volatile matter content on an a dry, ashfree basis and % R r of telocollinite. Finally, volatile matter content and calorific value, which are used as coal rank parameters, in some samples are mainly affected by the high ash yield and carbonate minerals in the coals. The )k values and intensities of the spectral maxima in sporinites are high, and weak-very weak, respectively and their Q ratios show a broad range. With increasing coal rank the sporinites in the samples do not fluoresce. The % R r values in this field are more useful indicators of lateral variations in rank than the fluorescence measurements of sporinites and the chemical analyses of the coal samples. The Bfiyfik and Kficfik seams obtained from the same borehole have very similar reflectance values (for example 0.76-0.77%Rr at 86-2;
....
H I G H SULPHUR COKING COALS, T U R K E Y 9000
n
i
9
I
"I
8000
127
- 7000 9
8 6000
L
i
>-
+ I
O 5O0O
4000
i
|I
55
60
~ 30
35
40
45
(a)
50
65
70
75
80
90
100
Volatile matter %, (daf) 80 75
7o " " 65
t IoB0y0ksea 1
Group IIo
/ 9 Kuc0k seam / ~a Unnamed seam~
~ 6o N 55
~
Group I
~ so N 45 "~ >
40 35
I 40
30
0
10
20
30
(b)
I 50
I 60
I 70
80
Calcite + Dolomite % 1.00
o
0.90
n
3
! o Buyuk seam l 9Kucuk seam I n Unnamed seam
a ~ ,_ rr
J
Co
0.80
~
0.70
o 9
o
o
9
# ~
0.60
9
O
o
O
n
[]
0.50 0.40 0,30
(c)
30
I
I
I
I
35
40
45
50
I
I
I
55 60 65 Volatile matter %, (daf)
:
:
70
75
BO
(adb):Air-dried basis; (daf): Dry, ash-free basis; CV: Calorific value %Rr: Mean random reflectance of telocollinite for every coal sample
Fig. 7. Graphical comparisons of the chemical analyses, X-ray powder diffraction (XRD) analyses, and %Rr measurements of telocollinite. (a) relationship between volatile matter contents versus calorific values, (b) calcite + dolomite contents determined by XRD versus volatile matter contents, (c) volatile matter contents versus %Rr.
0.76-0.74% Rr at 86-3; for location see Fig. 8). This shows that the parting sediments (3 m thick) between the two coal seams do not clearly influence the reflectance values. Figure 8 shows the iso-reflectance map of the coal field. Sample locations, reflectance values, faults and folds,
which are determined from a number of underground mine maps, and antimony mineralization in the Deliktas and Karacatepe areas also illustrated on the map. The iso-reflectance values for the seams over the whole coal field in general increase toward southern parts of the
128
A. I. K A R A Y I G I T & M. K. G. W H A T E L E Y
0
/'
~,
+*
!
9
8
C:'--
+
/ ""
c-
/ ./..
~
........
~-
0
:;
~.
~
~..,o.. ~
o
.
9
,!-:
:
-//
--~
..~
=
Z:..v:
-~
:=
:~
'-
3:',
et"
m ,0
o
:;
\
\ \
~,.
~
~
~
.-
~i7 _
,',A
:
.v
0.
~
,
,, \
,,,
}\
~
\
~
\
~
o
-~
~
~
--~
~o
"
N
|
="I
~, .;
._~ 1
~
~I
~
~
o- ~ t
~ ~ o ~
_
:1~.+
~
~
~
ul
~
.
-
~,
.,. " "~ ,- .~ ~
,-
~--
,.~ + \
_ ~
.~
/.i + ~
I
o,il._o.
~,
~.
,~,
i
~ " . ~,<~ . 0
-<;
t 6
~
..;'.. "'~Z~
"~
,=,
~
"..". \ ~\ +
.
.~,~..i~~o
~-i~ ..~-
Q-
~"
~
II
-~ _~''. ~,.'~L=,_ ~/ :. ~ _ ~ ~" ~ "
=
~
i,-
o~..-
3
9 [] <
._=~
\
.
~
~ \
,,
9 9" ..- . ~~ ~ .-
~
" d~_
~
~ ~..,
~. ....
/I T
.;l
-
\
\
:j--'~
9
..-,.
~+
i o'/. .
.
6
0
_..
,
". ...... ~ . . ~ . .
.~..
/"..
,o o . . . /
,,
.,~-0~:
.."/
-/ 6
HIGH SULPHUR COKING COALS, TURKEY coal field (Fig. 8). In addition, the reflectance values seem to increase toward the Karakaya fault zone (Fig. 8), and also in the same fault zone the values increase from 0.74-0.76% Rr (borehole 86-3) to 0.93% Rr (GM-11) near the silicifed zone of antimony mineralization. This indicates that this fault zone probably was affected by the antimony mineralization or high heat flow. On the other hand, the brecciated coals including higher calcite and dolomite contents have relatively low %Rr values (for example: C-106/0.58; C-108/0.57; C-56/0.65; C-57/0.65 and 86-10/0.59; Fig. 8). These results imply that the normal faults and the brecciated coals were probably formed prior to antimony mineralization. During field studies antimony mineralization and associated silicification were seen along the strike-slip fault zone, which is identified by G6kce (1987) and around Karacatepe antimony mineralization in the limestone-marble of the Muratdagi Melange (Fig. 8). In addition, a number of calcareous cones were formed probably by hot-springs in the Muratdagi region. K6ksoy & Ileri (1977) determined that some cones of recent hot-springs in the Ilica, about 40 km west of study area, contain high antimony values. G6kce (1987) proposed that the mineralization in the Muratdagi region was formed by the precipitation of antimony from hydrothermal solutions which vented along the Pre-Pliocene fault zones, especially where the calcareous rocks provide the optimum physicochemical environment. The increase of the vitrinite reflectance values, high calorific values and decreasing fluorescence properties of sporinites in the coal field can be related to the high heat flow resulting from recent hydrothermal mineralization. Thus the degree of coal rank and its lateral variation in the G6kler coal field is very advanced due to the high heat flow adjacent to the boundary faults in which antimony mineralization occurred. The temperature can be estimated from the reflectance values. The maximum value of the mean random reflectance of vitrinite of all the coal samples is measured as a 0.95% Rr. As we can not consider the effective heating time, this 0.95% Rr shows that the coals may have been theoretically affected by maximum temperatures of as high as 200~ according to Bostick et al. (1979).
Conclusion Mean random vitrinite reflectance values (%Rr) of between 0.50% and 0.95% were recorded in
129
the coal field. The %Rr values within the coal field generally increase towards the southern parts of the coal field. It is thought that this increase can be related to the recent hydrothermal antimony mineralization. The associated high heat flow has also resulted in a weak devolatilization of the coals (Fig. 8c), and an increase in the wavelengths of the spectral maxima of sporinites, to an optimum degree producing the coking properties in these coals. We acknowledge the British Council of Turkey who supported Karayigit's expenses in UK and the Research Foundation of Hacettepe University, Ankara, Turkey for supporting the reseach project, the help given by S. Toprak for spectral analyses of liptinite maceral groups and R. Wilson for evaluation of the microprobe results. We also thank O. Mal~n and M. Miliorizos who critically reviewed the manuscript.
References AKKUS, F. M. 1962. Geology of the area between Kiitahya and Gediz. MTA Bulletin, 58, 21-30 [in Turkish with English abstract]. ARAL, H. 1989. Antimony mineralization in the Northern Muratdagi (western Turkey). Economic Geology, 84, 780-787. ASTM, 1991. Annual Book of ASTM Standards, gaseous fuels; coal and coke. 1916 Race Street, Philadelphia, PA 19103, 05.05. ATABEK, S. 1939. A report on iron and antimony deposits of Oysu (Gediz). MTA Report E-990 [in Turkish]. BINGOL,E. 1977. Geology of Muratdagi and petrology of the main rock units. Bulletin of Geological Society of Turkey, 20/2, 13-67 [in Turkish with English abstract]. BOSTICK, N. H., CASHMAN,S. M., MCCULLOH, T. H. & WADDEL, C. T. 1979. Gradients of vitrinite reflectance and present temperature in the Los Angeles and Ventura Basins, California. In: OLTZ, D. F. (ed.) Low Temperature Metamorphism of Kerogen and Clay minerals, Los Angeles, 65-96. CARPENTER, A. M. 1988. Coal classification. IEA Coal Research, London. CASAGRANDE, D., SIEFERT, L., BERSCHINSKI, C. & SUTTON, N. 1977. Sulfur in peat forming systems of Okefenokee Swamp and Florida Everglades: Origins of sulfur in coals. Geochimica et Cosmochimica Acta, 41, 161-167. CHANDRA, D., GHOSE, S. & CHAUDHURI,S. G. 1984. Abnormalities in the chemical properties of Tertiary coals of upper Assam and Arunachal Pradesh. Fuel, 63, 1318-1323. CHAUDHURI, S. G. & GHOSE, S. 1990. Fluorescence technique-its application for better prediction of properties for certain Indian coals. International Journal of Coal Geology, 14, 237-253. CRELLING, J. C. 1983. Current uses of fluorescence microscopy in coal petrology. Journal of Microscopy, 132, 132-147.
130
A. I. K A R A Y I G I T & M. K. G. W H A T E L E Y
DEMIREL, I. H. 1989. Geology and sedimentology of the Tertiary aged sequences at the Ermenek (Konya) region and detailed examination of the coal seams. PhD thesis, Hacettepe University, Ankara [in Turkish]. DOGRU, R. 1978. Physical and chemical properties of some Turkish lignites. PhD thesis, Hacettepe University, Ankara [in Turkish]. GOKCE, A. 1987. Geology of the antimony mineralization in the Muratdagi (Gediz-Kiitahya) region. Bulletin of the Faculty of Engineering, Cumhuriyet University, Serie A-Earthsciences, 4, 65-85 [in Turkish]. GOKMEN, V. 1970. Report on the Neogene deposits around Gumelekoy (Gediz-Kfitahya). MTA Report 6183 [in Turkish]. GI)N, H. 1975. Geology of the Gediz (Kfitahya) Neogene basin and its southern part. MTA Report 6276 [in Turkish]. HORNE, J. C., FERME, J. C., CARUCCIO, F. T. & BAGANZ, B. P. 1978. Depositional models in coal exploration and mine planning in Appalachian Region. AAPG Bulletin, 62, 2379-2411. ICCP, 1963;1971. Internationales Lexikon Far Kohlenpetrologie. Centre National de la Recherche Scientifique 15, Quai-Anatole-France, Paris. JACOB, H. 1964. Neue Erkenntnisse auf dem Gebiet der Lumineszenmikroskopie fossiler Brennstoffe.Fortschr. Geol. Rheinld. u. Westf. 12, 569-588, Krefeld. - - 1 9 7 3 . Kombination yon Fluoreszenz-und Reflexions-Mikroskopphotometrie der organischen Stoffe yon Sedimenten und Boden.-Leitz-Mitt. Wiss. u. Techn. VI, 1 21-27, Frankfurt. KALAFATCIOGLU, A. 1961. Geological Report of the Area Between Gediz-Usak. MTA Report 2818 [in Turkish]. KARAYIGIT,A. I. 1983. Geology of Bahcekoy coal basin ( GOlbasi-Ankara ) and petrographic characteristics of coals. MSc thesis, Hacettepe University, Ankara [in Turkish]. & CICIOGLU, E. 1994. Geology, depositional environment and coal petrology of Sorgun Eocene coals, Yozgat-Turkey. International Conference, & Short Course on Coalbed Methane and Coal Geology, University of Wales Cardiff, UK, 12th-16th September 1994. - & ERIS, E. 1994. Geological setting and petrology of Celtek Eocene coals (Amasya), & the influence of the North Anatolian Fault and volcanism on the coal rank. International Conference, & Short Course on Coalbed Methane and Coal Geology, University of Wales Cardiff, UK, 12th-16th September 1994. KOKSOY, M. & ILERI, S. 1977. Genetic relationships between Gediz-Simav-llica hot-springs and antimony deposits formed in this area. TUBITAK, Turkey, Project No: TBAG-199 [in Turkish].
--,
GORMI3S, S., SAHBAZ, A., KARAYIGIT, A. I. & CERIT, O. 1987. Geological investigations of G6"kler (Gediz-Kfitahya) coalfield and determination of economical potential. Report of Earth Sciences Application and Research Centre of Hacettepe University, Project code: Yuvam/86-11 [in Turkish]. LEBKI3CHNER,R. F. 1965. A report on the coal geological investigations in the G6"kler coal basin, GedizKfitahya. Directorate of Mineral Research and Exploration, Report 3601 [in Turkish translation]. NEWMAN, A. C. D. & BROWN, G. 1987. The chemical constitution of clays. In: NEWMAN, A. C. D. (Editor): Chemistry of Clays and Clay Minerals. Longman Scientific & Technical, Mineralogic Society Monograph No. 6, 1-129, England. OTTENJANN, K., TEICHMISLLER,M. • WOLF, M. 1975. Spectral fluorescence measurements of sporinites in reflected light and their applicability for coalification studies. In: ALPERN, B. (ed.) Pdtrographie de la Mati&e Organique des Sediments, Relation avecla Paleotemperature et le Potentiel Petrolier, pp. 67-95, Paris. RENTON, J. J. & BIRD, D. S. 1991. Association of coal macerals, sulfur, sulfur species and the iron disulphide minerals in three columns of the Pittsburgh coal. International Journal of Coal Geology, 17, 21-50. STACH, E., MACKOWSKY,M.-TH., TEICHM~LLER, M., TAYLOR,G. H., CHANDRA,D. & TEICHM(3LLER,R. 1982. Stach's textbook of coal petrology. Gebriider Borntraeger, Berlin. TEERMAN, S. C., CRELLING,J. C. ~; GLASS, G. B. 1987. Fluorescence spectral analysis of resinite macerals from coals of the Hanna Formation. Wyoming, USA International Journal of Coal Geology, 7, 315-334. TE1CHMI~ILLER, M. & OTTENJANN, K. 1977. Liptinite und lipoide Stoffe in einem ErdO'lmuttergestein.Erd61 u. Kohle, 30, 387-398. -& WOLF, M. 1977. Application of fluorescence microscopy in coal petrology and oil exploration. Journal of Microscopy, 109, 49-73, London. UNSWORTH, J. F., BARRATT, D. J. & ROBERTS, P. T. 1991. Coal quality and combustion performance. An International Perspective. Coal Science and Technology 19, Elsevier, Amsterdam. WARD, C. R. 1978. Mineral matter in Australian bituminous coals. Proc. Australas. Inst. Min. Metall. No. 267, 7-25. WHATELEY, M. K. G. & TUNCALI, E. 1995. Origin and distribution of sulphur in the Neogene Beypazari lignite basin, Central Anatolia, Turkey. In: WHATELEY, M. K. G & SPEARS, D. A. (eds), European Coal Geology. Geological Society, London, Special Publication, 82, 307-323. YAGMURLU, F. & KARAYIGIT,A. I. 1984. Petrographic characteristics of Citak (Akhisar) coals. Turkish 4th Coal Congress, Publication of Chamber of Mining Engineering, 111-122 [in Turkish].
The main coal facies and lithotypes of the Pliocene coal basin, Oltenia, Romania NICOLAE
TICLEANU
& DORINA
DIACONITA
Geological Institute o f R o m a n i a , 1 Caransebe,~ Street, 78344 Bucure~ti - 32, R o m a n i a
Abstract: This paper presents the results of palaeobotanical studies correlated with coal petrography research carried out by the authors in the western part of the Dacic Basin (Oltenia) coal deposits. Of the seven coal facies that have been distinguished so far in the lignites of Oltenia, the present paper deals with the five main facies generated by: swamp deciduous forest, Carex ssp. grassy marsh, swamp with Glyptostrobus, reed swamp and aquatic macrophyte prairie. The palaeoecological and palaeophytocoenoical studies have led to the reconstruction of the factors that characterize the major coal facies. General geological setting
Previous research
The Pliocene coal basin of Oltenia contains the most important coal deposits in Romania and it is situated south of the South Carpathians, between the Danube and the Olt Valley. The coal precursor, peat accumulated in the western area of the Dacic Basin. Structurally, the coal basin overlies the Carpathians Foredeep and the Moesian Platform (Fig. 1). According to Ticleanu et al. (1988) and Ticleanu & Andreescu (1988) the tectonicstructural units have played a very important role in the distribution of the coal deposits. Thus the greatest thickness of most of the 22 coal layers is situated on the internal side of the fore deep. On the external side of the foredeep both the number and the thickness of the coal layers decrease so that in the Neogene cover on the Platform there are only 1-2 beds with thickness exceeding 2 m, and they occur as local pockets. According to Andreescu et al. (1985) the sequence of the Pliocene coal deposits consists of three lithostratigraphic units (Fig. 2):
According to Teichmfiller (in Stach et al. 1982) 'the term "coal facies" refers to the primary genetic types of coal, which are dependent on the milieux under which the peat originates.' At the same time, the coal facies also includes the petrographical composition and the physicalchemical properties of coal. Using these factors, Ticleanu & Bi~oianu (1989) have described seven facies in the Pliocene coal of Oltenia. The petrographical coal structure (Bi[oianu & Ilie 1967, Ilie & Bi!~oianu 1967), the connections between the petrography and the physical-chemical characteristics of coal) Ticleanu et al. 1989, 1992) and the identification of coal-generating vegetation and environmental conditions (Ticleanu 1986, 1992a, b) have also been studied.
9 the Berbe~ti Formation (Uppermost PontianLower Dacian, according to Andreescu (oral communication) is mainly psammitic and in the upper half contains clayey intercalation and six coal layers (A, B and I-IV) that constitute the Valea Vi~enilor coal complex; 9 the Jiu-Motru Formation (Upper DacianMiddle Romanian) is mainly pelitic-psammitic with eight coal layers (V-XIII) that constitute the Motru coal complex, the most important coal complex in the basin; 9 the Cfinde~ti Formation (Middle RomanianPleistocene) is mostly psammitic-psephitic that includes clayey intercalation and five coal layers (XIV-XVIII). These constitute the B~lce~ti coal complex, developed only in the central area of the basin.
Methodology This study is designed to improve the knowledge of the characteristic flora of the main coal facies by reconstructing the environments in which the parental material accumulated in the Pliocene. For a better understanding of plant communities we have identified the flora using the methodology and the terminology presented by Ticleanu (1992a, b, 1995a, b), and we have combined this with the results of taphonomical, palaeophytocoenotical, palaeocarpological and cuticulary analyses carried out over the last few years. The fossil plant material that we studied for the palaeoecological and palaeophytocoenotical interpretations, came from over 12 500 m of cores from 85 boreholes and from the working faces of 22 coal open-pit mines. The results of this research identified over 800 accumulations of fossil vegetable remains (AFVR), of which 585 are autochthonous, 153 are hypautochthonous, 28 are allochthonous and 24 are of mixed origins.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 131-139.
132
N. TICLEANU & D. DIACONITA
9
,
,'
'
9 'i
'
9
"
, ' . k
"
. .IINT.FRN, AL}.'~ "-~,' ....
"
'
"
"
' '.'."
,(EXTER'NAL)'
CRAIovA\ MOES[AN
PLAZrFORM
t
/
t
l
LEGEND ...............
1 - The Oltenia
.......
2 - Pericarpathian
. . . . L_J_ ,
coal basin
boundary
fault
3 - Flexure
Fig. 1. Sketch map of the delimitation of the Oltenia coal basin and its structural framework.
Besides these, 68 AFVRs of fossil fruits and fructifications and over 560 AFVRs of fossil wood accumulations (xylite), were identified. This wide range of AFVRs indicates the autochthony of the coals in Oltenia and enables the reconstruction of the plant communities to the level of palaeophytocoenoses. The knowledge about coal-generating plant communities and environment has been extended on the basis of correlation of taphonomical investigations, ecological demands and coenotical characteristics of modern plant equivalents of fossil plants and fossil plants associations. The use of the principle of uniformitarianism has been facilitated by the relatively young age of coal deposits (2.5-4.5 Ma) and by the fact that most of the fossil plants (41) and palaeophytocoenoses (18) of the Pliocene swamps in Oltenia are also to be found today, or still have direct modern equivalents in today's swamps in Romania.
The authors have carried out many petrographical analyses with a view to establishing connections between coal facies and coal lithotypes generated by these.
Brief palaeogeographical, palaeoclimatical and palaeophitocoenotical description Palaeogeographically speaking, the coal deposits in Oltenia accumulated on the western side of the Dacic Basin (Fig. 1), which, most of the time, in the Pliocene was a large alluvial-lacustrine plain. Because of vertical tectonic movements, the morphological aspect of this accumulation plain changed cyclically (Ticleanu 1992a), each cycle having four stages: fluviatile, fluvio-lacustrine, telmatical and lacustrine. The cyclical variations of the subsidence rates have determined the cyclical distribution of stages, each of them being characterized by a
PLIOCENE COAL FACIES AND LITHOTYPES IN ROMANIA
I
Bk, LCE.STI COALY COMPLEX
HOTRU COALY COMPLEX
VALEA VISENILOR COALY COMPLEX Coals
f----'~cioy~ Sillic r162
~
R$~mml~'Q P~.ephile
Fig. 2. Synthetic lithostratigraphic column of the Pliocene coal deposits from Oltenia. prevailing lithofacies: mainly sandy in the fluviatile stages, sandy-silty-clayey in the fluviolacustrine stages, with peats in the telmatical stages and mainly clayey in the lacustrine ones. During the telmatical stages the basin subsidence was low and equalled plant accumulation rates, allowing most of the accumulation plane to be covered by eutrophical low moors, with thicknesses ranging between 0 and few metres. Climate type that allows large swamp generation is characterized by a positive hydrological regime, where precipitation exceeds evapotranspiration. According to Ticleanu (1995c), during the Pliocene peatmoor generation the mean annual temperature fluctuated between 14 and 15.5~ and the precipitation quantity exceeded 1200 mm per year, almost uniformly distributed with only a slight summer peak. In these conditions, the temperature of the peat was probably 16-22~ with a maximum in Upper Dacian. The presence inside the coal layers of some centimetre-scale argillaceous intercalations with some fossil vegetable remains (FVR) concentrations, some with much pollen and others with many fossil leaves, shows the existence of spring and autumn floods.
133
The above climatic premises have encouraged the development of a hygro-hydrophyte vegetation relatively poor in species but very rich in specimens. Ticleanu (1992a) has identified more than 70 species. The most frequent species are
Byttneriphyllum tiliaefolium, Phragmites oeningensis, Glyptostrobus europaeus, Salix ssp., Stratiotes dacicus, Carex ssp. Typha latissima, Trapa ssp., Acer tricuspidatum, Nelumbo protospeciosa and Osmunda regalis, the other species being found in less than 5 AFVR each. The first six taxa constituted plant communities (palaeophytocoenoses) alone but they also generated associations with other taxons. The most frequent palaeophytocoenoses were identified in autochthonous AFVRs. These are (ordered by their frequency): Byttneriphyllum
tiliaefolium (122), Glyptostrobus europaeus (108), Phragmites oeningensis (84), Salix ssp. (43), Byttneriphyllurn tiliaefolium & Glyptostrobus europaeus (19), Phragmites oeningensis & Typha latissima (19), Stratiotes dacicus (18), Trapa ssp. (18), Byttneriophyllum tiliaefolium & Salix ssp. (12) and Carex ssp. (12). Using modern swamps in Romania as analogues, the Pliocene swamp plant communities are believed to be distributed first of all by their demands for water. From this point of view, using a similar method to Teichmtiller's (1958) several areas (palaeobiotope) can be distinguished (Fig. 3): marginal; seasonally fooded; almost permanently flooded; permanently covered by water, between 0 and 2 metres deep, permanently covered by water, between 0 (2) and 3 metres deep, depending on the stage of vegetation's evolution and open water area. Each of the above areas (palaeobiotope) includes one or more vegetal associations (phytocoenoses). These associations constitute ecological series from the centre to the margins of the swamp. In this way, in the modern swamps in the Danube Delta, Rudescu et al. (1965) shows the presence of one ecological series that contains the following vegetal associations (from the open water to the marginal area): Characetum,
Nupharetum, Phragmitetum, Typhaetum, Scirpetum, Carecetum, Juncetum, Salicetum. Except the Juncetum association, not yet found, all the others have corresponding palaeophytocoenoses in the Pliocene swamps. Modern analogues suggest that the flooded areas of vegetation in the Pliocene are dominated by grass associations, especially Carex (prevailing herbaceous stage). Later these grass dominated associations were replaced by palustral forest associations (prevailing forest stage). Unlike the modern swamps, where the flooded
134
N. TICLEANU & D. DIACONITA
Zts
~ -
0~. ~q r..)
, II
II
II
II
II
II
II
II
t
r i
i
,,
g Z;>-
>-
1 ,,1
~
~.
M
~.~~e
-~1
1-,
~..~1
-~--'--
~
~-i
> ~- ~
m
9
~ "~
m
II
m
tl
II
II
II
II
t~. =
II
".~
e
t 5
~
o
o ~
-~, . -
~
~
,
~ ~d ~
~"
'~
PLIOCENE COAL FACIES AND LITHOTYPES IN ROMANIA
135
forest areas are dominated by Salix, Populus, Alnus and Betula, the Pliocene swamps forest
Table 1. The main physical and chemical characteristics of the lithotypes of the Olternia.
palaeophytocoenoses
Lithotype
were
dominated
by
Byttneriophyllum tiliaefolium, Salix ssp., Glyptostrobus europaeus, as well as other tree species. Because rainfall distribution controls various factors (such as swamp bottom morphology, annual rainfall quantity, evapotranspiration etc.), the areas occur as more or less parallel bands. The general aspect of the vegetation was that of a huge mosaic such as that presented by McCaffrey & Hamilton (in Cohen et al. 1984), for the Okefenokee swamp map of the southeast United States.
The main coal facies and the lithotypes According to Ticleanu & Biloianu (1989) the Pliocene coals in Oltenia contain the following facies: Sequoia abietina forest, swamp deciduous forest, Carex spp. grassy marsh, swamps with Glyptostrobus, reed swamps, floating vegetal formation and aquatic vegetation. Considering the fact that the first facies played an insignificant role in the coal-generating phytomass constitution and the sixth one led only to coaly clays, in this study we shall refer only to the remaining facies which constitute those of real significance, together with an addititional facies - aquatic macrophyte prarie. To describe the primary characteristics of coal layers in each coal facies we have used Teichmfiller's method (in Stach et al. 1982).
The coal faciesmgrassy marsh Carex ssp. Plant communities: palaeophytocoenoses with Carex ssp. (Carex flagellata, Carex cf nigra etc.) Associated dements: Scirpus rnaeotica, Cladium mariscus, Cladium palaeomariscus, Sparganium neglectum, Najas pliocenica, Butomus umbelatus, Oenanthe aquatica, Trichosanthes fragilis, Pedicularis sp., Lythraceae ( ?Lythrum salicaria), etc. Most of these taxa are still living today and examples are found in actual Carex ssp. swamps in Romania (Pop 1960). In the palustrian vegetation in the Danube Delta, Popescu et al. (1981) have identified the association with
Aanh (%)
Vd (%)
Xylite 1.26-3.35 47.7-68.4 Xylitic coal 8.58-17.82 39.0-52.2 Weak xylitic 16.79-28.0 36.2-46.9 coal Detrital coal 22.11-48.45 27.8-38.8
Qd (kcal/kg) 5042-6034 4340-5378 3874-4871 2503-4012
Aanh, Anhydrous ash; V~ Volatile matter (air-dried basis); Qa Upper calorific value. In the first stages of the evoluion of the Carex swamp's vegetal cover, there were only small clumps of bushes of Salix cinerea and Salix pliocenica distant from the open water but they developed into bigger and bigger clusters, as they grew nearer to the central open water area of the swamp where they occurred beside Byttneriphyllum tiliaefolium, Nyssa, Glyptostrobus, etc. Palaeobiotope: floodable areas, but only in the first stages of vegetation evolution. Type of deposition: mainly autochthonous, partially hypautochthonous by vegetable remains brought by floods and less allochthonous (pollen, leaves fragments and fruits brought by wind). Depositional milieux: telmatic. Environmental characteristics: the peat is generated especially by roots and less by aerial remains of plants, deposited under aerobicanaerobic conditions with a moderate to weakly acid pH(5.0-6.8), rarely from weak acid to neutral (6.0-7.2), and extremely rarely in an alkaline environment. Lithotype: detrital coal, i.e. a groundmass with a fine detritic texture, more or less layered, more than 50mm thick, frequently with more than 30% ash content (Table 1). According to Pop (1960) the actual Carex peat contains more than 10% ash, and where flooded, this percentage increases because of the mineral material suspended in water. We consider that transformation of peat into lignite can lead to a relative enrichment of the ash content up to 30%. The petrographical composition of the detrital coal lithotype shows its high humodetrinite content (Table 2).
Claudietum mirisci. The presence of taxa that generate actual phytocoenoses ( Carecetum, Scirpetum and Claudictum) and the relatively high density of their seeds in the clays and the clayey silts that accompany the coal layers prove the development on large areas of coal facies of swamps with sedge (Carex ssp.).
The coal facies---deciduous forest Plant communities: Palaeophytocoenoses with Byttneriophyllum tiliaefolium, Byttneriophyllium tiliaefolium-Glyptostrobus europaeus, Salix ssp.Glyptostrobus europaeus, Salix ssp. (Salix abla, Salix fragilis, Salix grandifolia, etc.).
136
N. TICLEANU & D. DIACONITA,
Table 2. Petrographical composition of the Pliocene coals between the Danube and the Amaradia Valley after Bitoianu (in Ticleanu et al. 1989, 1992).. Group of macerals Ubgroup of erals
Huminite % Humotelinite
Humodetrinite Humocollinite
68.2-87.0 49.0-49.6 22.0-27.8 8.0-18.5
8.0-12.7 25.8-32 51.8-54 59.0-61.7
Liptinite (%)
Inertnite (%)
0.15-0.20 1.30-3.0 1.80-3.5 1.10-1.5
0.31-1.2 2.40-3.2 2.70-4.4 2.07-4.2
Lithotype xylite xylitic coal weak xylitical coal detrital coal
Associated elements: Acer tricuspidatum, Populus populina, Nyssa disseminata, Liquidambar europaeum and Carya cf. aquatica. Palaeophytocoenoses with Byttneriophyllum tiliaefolium, Glyptostrobus europaeus and Nyssa were located in the central areas of the swamp, and Salix ssp. in the marginal ones. Palaeobiotope: seasonally flooded areas. Type of deposition: frequently autochthonous, rarely hypautochthonous, and extremely rarely allochthonous (pollen, flying fruits). Depositional milieux: telmatic. Environmental characteristics: acid pH (3.5-5), deposited in aerobic-anaerobic conditions, to explain the rapid loss of most of the cellulose. Lithotypes: detrital coal, weak xylitical coal and rarely xylite coal. The weak xylitic coal was probably generated in the internal part of the seasonally flooded area, to the limit of the almost permanently flooded area, from palaeophytocoenoses with Byttneriophyllum tiliaefolium-Glyptostrobus europaeus and Salix ssp.-Glyptostrobus europaeus. In the same areas a very small part of the lithotype xylite also accumulated, included in detrital coal, resembling rare xylite lenses and bands with more than 50 mm thick. Generally, the ash content is relatively low, because of the protection granted by the surrounding palaeophytocoenoses against water with suspended clay content. The coal facies--forest swamp with Glyptostrobus Plant communities: palaeophytocoenoses with Glyptostrobus europaeus. Associated elements: very rare Taxodium dubium and Nyssa disseminata representing Miocene relicts. In open horizontal structured palaeophytocoenoses, trees were scarce, and between them were areas covered by water, even when the water level was at its lowest. In these prairie
0.1-0.13 0.6-0.7 0.3-1.3 0.2-3.7
areas occasional Stratiotes dacicus were found. This explains the relatively high frequency of seeds of this species in the xylitic coal. Surrounding Glyptostrobus europaeus trunks, and between these trunks and pneumatophores fern bushes (Osmunda regalis) were growing similar to Osmunda lignitum (Petrescu & Givulescu 1986), in Chattian swamps in the Petro~ani Basin. The role of Glyptostrobus europaeus species in the genesis of coals, has been treated in many papers. The most recent is the paper by Boulter et al. (1993). Palaeobiotope: almost permanently flooded areas, where water withdraws for only one or two months in a year. Type of deposition: mainly autochthonous. Depositional milieux: telmatic to subaquatic (limnic). Environmental characteristics: in closed horizontally structured palaeophytocoenoses (pure Glyptostrobus forest) the pH was probably low (3.5-5) and in open structured ones the pH was higher but still weakly acidic (5.6--6.6) and less probably neutral (6.8-7.2), as in modern swamps with Stratiotes aloides, the modern equivalent of Stratiotes dacicus species. Considering that the palaeophytocoenoses with Glyptostrobus europaeus were covered by water between ten and eleven months in a year, the plant material accumulated under mostly anaerobic conditions. Lithotypes: xylitic coal, xylite and weak xylitical coal. The xylitic coal lithotype is composed of alternating bands of fine coal mass, weakly banded, and bands and lenses of fossil wood (xylite) both less than 50 mm thick. The xylitic coal is second in importance (30% of the entire reserves) after detrital coal in the soft brown coals in Oltenia. Many coal layers consist only of this lithotype. Pieces of charcoal (inertinite) with a variety of different sizes occur relatively frequently, in the detrital of xylitic coal. In our opinion, these
PLIOCENE COAL FACIES AND LITHOTYPES IN ROMANIA show the presence of frequent natural fires in the forest swamp. These fires have played the same role in the Pliocene swamps as they do in modern swamps for the vegetation in Okefenokee, as shown by Izlar (in Cohen et al. 1984). The ash content of the xylitic coal lithotype is relatively low, because the clayey material is kept out by the surrounding palaeophytocoenoses. The ash content of the xylitic coal lithotype is also proportional to the amount of xylite. The ash content xylite is very low (Aanh: 1.26-3.35%) and is of primary origin (Table 1). Another important lithotype generated in the same facies is xylite that represents between 5 and 20% of entire coal volume, and is represented by lenses and bands that can reach up to a few meters long and 50-350 mm thick, sometimes even more. More than 80% of the xylite represents branches, trunks and roots of Glyptostroboxylon tenerum, identified by Petrescu (oral communication). The xylite lithotype is characterized by the highest humotelinite content (Table 2).
The coal facies--reed swamp Plant communities: palaeophytocoenoses with Phragmites oeningensis and with Typha latissima. Associated elements: Sparganium noduliferum, Stratiotes dacicus, Butomus umbelatus, Najas lanceolata, Equisetum sp., Carex ssp. Palaeobiotope: permanently covered by water with a depth of less than 2 m, at maximum. In deeper water the plants are drowned. Type of deposition: autochthonous, partially hypautochthonous up to the limit with the aquatic macrophyte prairie, because of the action of streams and storm waves. Depositional milieux: subaquatic (limnic). Environmental characteristics: similar to modern swamps with Phragmites australis, where the pH is mainly neutral to alkaline. The plant material accumulates in anaerobic conditions. Modern peats with Phragmites contain from 15 to 20% ash content. By residual enrichment, the resulting coals could reach between 30-50% ash content. Ph. australis, is growing today in the Danube Delta on a surface of more than 20000 ha, in which a surface of more than 100000ha is covered by 'plaur', a floating vegetal mass, mainly generated by the rhyzome of this species. Lithotype: detrital coal. Distinguished from the detrital coal from swamps with sedge (Carex ssp.) and from hygrophite deciduous forest,
137
because this facies contains frequent seeds of
Stratiotes dacicus. The quantity of this coal shows the importance of palaeophytocoenoses with Phragmites in coal generating phytomass constitution.
Aquatic macrophyte prairie Plant communities: palaeophytocoenoses dominated by one of the main taxa: Stratiotes dacicus, Trapa urceolata, Trapa expectata, Myriophyllum nagavicum, Hydrocharis morsus-renae, Potamogeton corniculatus, Ceratophyllum demersum, Ceratophyllum submersum, Nymphaea alba, Nuphar pliocenicum and Nelumbo protospeciosa. The relative high number (68) of AFVR of aquatic plants, and their palaeocarpological content has helped us to identify 15 palaeophytocoenoses in the Pliocene swamps. From the 27 aquatic plant associations found by Popescu et al. (1981) in the Danube Delta, at least eight also existed as such in the Pliocene
(Hydrocharitetum morsus-ranae, Stratiotetum, Cerato_phylletum demersi, Myriophyllo-Potametum, Najadetum, Myriophyllo-Nupharetum, Trapo-Nymphoidetum and Trapetum). Associated elements: Salvinia sp., Spirematospermum wetzleri, Brasenia tanaitica, Myriophyllum spinosum, Trapa givulescui, Trapa victoriae, Trapa horrida, which could also generate monocoenoses. The aquatic macrophyte prairies are the first in an ecological succession (Fig. 3). Vertically, in the genetical series, the aquatic macrophyte prairies are replaced by swamps with Phragmites, and these by forest swamps. Such successions can be foune in many of the coal open-pit mines in Oltenia. Palaeobiotope: permanently flooded areas, with a depth of more than 2 m (in the initial stage of the vegetation evolution the depth can be between 0 and 3 m). Type of deposition: autochthonous, but in many ways hypautochthonous because of water currents and storm waves than move the plant material from low deep swamps. Depositional milieux: subaquatic (limnic). Environmental characteristics: Similar to swamp lakes in the Danube Delta and the Romanian Plain where the water is introduced into the lake via rivers, but almost double the rainfall and with a pH from neutral to alkaline. The plant remains accumulated in an anaerobic environment. Lithotypes: frequently, in this environment coaly clays and clayey coals were generated. Nevertheless, the frequency of Nelumbo leaves, strictly
138
N. T I C L E A N U & D. DIACONIT-~
autochthonous on the X layer level in Pinoasa and Plo~tina open-pit mines, lead us to the conclusion that this species has played an important role in the phytomass constitution. In the same way that in the Okefenokee swamp (Cohen et al. 1984) Nymphaea peat together with Taxodium peat constitutes more than 80% of this peat. Palaeophytocoenoses with Stratiotes and with Trapa could generate such peats and also detrital coals.
Conclusions The Pliocene coals in Oltenia were generated in five distinct coal facies (grassy marsh Carex ssp., swamp deciduous forest, swamp with Glyptostrobus, reed swamps and aquatic macrophyte prairie) with a time and spatial distribution controlled, first of all, by the hydrological level, which has generated palaeobiotopes with different environmental factors (pH, Eh, etc.). Each facies has a characteristic plant community, composed of one or more palaeophytocoenoses generated by plants belonging to major taxonomic groups (angiosperm, mono and dicotyledon and gymnosperm) distinguished by the content in the main coal-generating substance (cellulose, pentosane, lignin, etc.). In the primary peat swamps different environmental conditions from one palaeobiotope to another and important quantities of vegetal phytomass with different chemical composition accumulated. This determined the development of distinct lithotypes, characterised by physicalchemical properties (Table 1) and certain petrographical characteristics (Table 2), that enabled differentiation of the main lithotypes detrital coal, xylite and xylitic coal. Phytogeographical considerations indicate that the coal-generating flora are distinguished by a strong Pliocene characteristic feature determined by: the fact that corresponding modern plant assemblages are prevalent in the eutrophical low moors in Romania, the important role of species Glyptostrobus europaeus and Byttneriophyllum tiliaefolium; the sporadic presence (relict) of the species:
Taxodium dubium, Nyssa disseminata, Acer tricuspidatum, Carya cf aquatica and Liquidambar europaeum, all of them characteristic of the Miocene flora in Europe, and with modern correspondents in the south-east of North America.
References ANDREESCU, I., TICLEANU, N., PANA, I., PAULIUC, S., PELIN, M. & BARUS, T. 1985. Stratigraphie des d6p6ts pliocenes a charbons. Zone est d'Oltenie (Secteur Olt-Jiu). Analele Universitdlii Bucure~ti, Geologie, 34, 87-96. BITOIANU, C. & ILIE, S. 1967. Contribu~ii la studiul petrografic al c~trbunilor de la Valea Motrului (Oltenia). Studii tehnice #i economice, 7, A, 165-173. BOULTER, M. C., HUBBARD, N. L. B. R. & KVACEK,Z. 1993. A comparison of intuitive and objective interpretations of Miocene plant assemblages from north Bohemia. Palaeogeography, Palaeoclimatology, Palaeoecology, 101, 81-96. COHEN, A. D., CASAGRANDE,D. J., ANREJKO, M. J. & BEST, G. R. 1984. The Okefenokee Swamp: its natural history, geology and geochemistry. Wetland Surveys, 709. ILIE, S. & BITOIANU, C. 1967. Studiul petrografic al c~rbunilor de la Rovinari. Studii tehnice #i economice, 7, A, 177-185. PETRESCU, I. tYr GIVULESCU, R. 1986. Flore et vegetation de la Valea du Jiu (Basin Petro~ani). Revue de Palaeobiologie continentale XIV, 2, 385-395. PoP, E. 1960. Mla~tinile de turb~ din R. P. Romgm~t. Editura Academiei, 511. POPESCU, A., SANDA, V. & NEDELCU, G. A. 1981. Allgemeine Ubersicht fiber die Vegetation des Donaudeltas. Horti Bucurestiensis, Acta Botanica, 175-191. RUDESCU, L., NICULESCU, C. & CHIVU, I. P. 1965. Monorafia stufului din Delta Dun~rii. Editura Academiei Romdne Bucure~ti. STACH, E., MACKOWSKY, M. Th., TAYLOR, G. H., CHANDRA, D., TEICHMULLER, M. t~r TEICHMULLER, M. & TEICHMI~ILLER,R. 1975. Coal Petrology, Berlin. TEICHMIJLLER, M. 1958. Rekonstruktionen verschidener Moor typen Hauptfl6zes der Niederrheinischen Braunkohlen. Fortschritte in der Geologie yon Rheinland und Westfalen, 2, 599-612. TICLEANU, N. 1986. Date preliminare privind studiul palaeobotanic al unor foraje de referin~t, 70-71/ 3, Palaeontologie, 235-248. 1992a. Studiul genetic al principalelor zficfiminte de cfirbuni neogeni din Romfinia pe baza palaeofitocenozelor caracteristice, privire speciafft la Oltenia. Tezd de doctorat, 339. Universitatea Bucureqti, Romfinia. - - 1 9 9 2 b . Main coal-generating palaeophytocoenoses in the Pliocene of Oltenia. Romanian Journal of Palaeontology, 75, 75-80. 1995a. An Attempt to Reconstitute the Evolution of the Mean Annual Temperature in the Neogene of Romania. Romanian Journal of Palaeontology, 76/3, 137-144. - - - 1 9 9 5 b . Taphonomic Researches on the Fossil Plants from the Pliocene Coal Deposits in Oltenia. Romanian Journal of Palaeontology, 75/3, 153-160. - - 1 9 9 5 c . Utilisation of the Palaeobotanical Data in the Study of the Coal Deposits. Romanian Journal of Palaeontology, 76/3, 145-152.
PLIOCENE COAL FACIES A N D LITHOTYPES IN R O M A N I A & ANDREESCU, [. 1988. Considerations on the development of Pliocene coaly complexes in the Jiu-Motru Sector (Oltenia). Ddri de searnd, Institutul de Geologie l'i Geofizicd, 72-73/2. Zdcdminte, 227-244. & BITOIANU,C. 1989. Coal Facies, Characteristic Palaeophytocoenoses and Lithotypes of Pliocene from Oltenia. Studia Universitatis Babeq - Bolyai, 34, 2, 89-93. - - , ANDREESCU, l., BITOIANU, C., PAULIUC, S., NICOLAE, Gh., NI[COLAE,V., POPESCU,A., BARUS, T., PASLARU,T., GRIGORESCU, Gh. & TICLEANU, M. 1988. Remarks on the relationship between the spatial distribution of the coal complexes in -
-
-
-
139
the Olt-Jiu sector and the structural-genetic factors. Ddri de seamd, Institutul de Geologie ~i Geofizicd, 72-73/2. Z&'dminte, 215-226. - - - , BITOIANU,C., MUNTIU, O. & NAGAT, F1. 1989. Palaeofitocenozele carbogeneratoare, petrografia ~i chimismul litotipilor din c~rbunii plioceni dintre Valea Jiulu ~i Valea Amaradiei. Ddri de seamd. Institutul de Geologie ~i Geofizicd, 74/2, Zdcdrninte-Geochimie, 115-129. TICLEANU, N., BITOANU,C., NICOLAE, Gh., POPESCU, A., TICLEANU,M., MUNTIU, O. & PROD,~,NESCU,I. 1992. P~trographie et propri6t6s physico-chimiques des charbons pliocenes du secteur Jiu-Danube. Romanian Journal Mineral Deposits, 75, 107-115.
Bulgarian low rank coals: geology and petrology GEORGE D. SISKOV Sofia University 'St. Kliment Ohridski', 15 Tsar Osvoboditel Blvd, 1504 Sofia, Bulgaria
Abstract: The largest coal-forming maximum in Bulgaria took place during the Neogene. Fifteen coal deposits are located in four coal-bearing provinces. The coal deposits south of the Balkan Mountains were formed in small grabens and depressions filled with molasse. Only the coals in Northern Bulgaria were formed in a small palaeodelta. The coal measures are of varying thickness and contain a few coal seams, with compact to complex structure and a range of thickness. Three groups are defined on the basis of maceral composition, allowing a reconstruction of the coal-forming ecosystems and the genesis of the genotypes during biochemical coalification. According to Alpern's classification the coals have middle to high ash content (ashy to coaly facies). They are of huminite type with low liptinite and inertinite content, and of low rank - lignite and mat brown coals.
Bulgaria contains more than 50 coal deposits but most of them are of no industrial value due to the complex conditions affecting their exploitation, the nonprospective character of the resource and the low quality of the coals (high ash and sulphur content). Their formation coincides with the world coal-forming maxima during Carboniferous, Early Jurassic, Late Cretaceous, Paleogene and Neogene. The geological potential of Bulgarian coal resources is about 8 x 109 tonnes, of which 85 % are low rank, 15 % middle rank, and < 1% high rank coals. According to the Bulgarian Standard they are divided into four g r o u p s lignites, brown coals (mat and bright), hard coals, and anthracites (Si~kov & Valceva 1983). Lignites are concentrated in the Neogene sediments and are one of the main energy resources in Bulgaria.
the beginning of the Late Oligocene (Si~kov et al. 1986). In that period typical marine sedimentation was gradually replaced by limnic sedimentation caused by progressive regression (Panov 1982). The coal formation occurred in highly peneplaned coastal areas covered with eutrophic swamps.
24~
!
26~
2
.
zt
__U
) V 4 ",-s
-
~
.,*'*
\
l
Geology of the Neogene coal deposits On the basis of topographical, morphotectonic, lithologic and genetic characteristics the Neogene coal deposits are located in four coalbearing provinces, three of them being south of the Balkan Mountains (Fig. 1, Table 1). The coal formation process started in Middle Miocene and ended in Pliocene. It took place within the period of the Pyrenian and Styrian phases of the Late Alpine orogeny and was concentrated in local depressions and grabens filled with molasse (Ivanov 1983). Features representing continuous coal formation during the Paleogene maximum have been established in the region of the Thracian Valley. Here, after the Savic phase, a post-tectonic depression was formed (Ivanov 1983) in which there were favourable conditions for coal formation from
mmm
mm mmmmmm
mmmm
5OO
Om1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Fig. 1. Location of the coal-bearing provinces, geolgical age and structure of the coal-bearing strata of the Neogene coal deposits in Bulgaria. A, Dacian coal, bearing province: (1) Lom; (2) Kozloduj; B, Thracian coal-bearing province: (3) Elhovo; (4) Mariza East; (5) Mariza West; C, Sofia coal-bearing province: (6) Sofia; (7) Beli Brjag; (8) Aldomirovzy; (9) Stanjanzy; (10) Kovachevzy; (11) Karlovo; (12) Chukurovo; D, Strimon-Mesta coal-bearing province: (13) Kjustendil; (14) Oranovo; (15) Razlog; (16) Goze Delchev.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 141-148.
142
G . D . SISKOV Table 1. Neogene coal deposits in Bulgaria Coal-forming maximum
Coal-bearing provinces
Neogene
Dacian
Coal deposits
Lom Kozloduj Thracian Elhovo Mariza East Mariza West Sofia Sofia Beli Brjag Aldomirovzy Stanjanzy Kovacevzy Karlovo Chukurovo Strimon-Mesta Kjustendil Oranovo Razlog Goze Delchev
Thracian coal-bearing province This province is situated in the area of the Thracian depression and is filled with sediments of the lower (Paleogene) and the upper coalbearing molasse (Neogene). The coal formation started in the Late Oligocene and continued through the Miocene and into the Pliocene. Gradual younging of the coal-bearing sediments occurred from west (Mariza West) to east (Elhovo). According to Kojumdgieva (1983) this is due to the gradual progression of the regression in this direction as well as to the formation of the Black Sea. Under these circumstances the plain had been raised relatively following the marine regression. The period of relatively compensated sedimentation led to the formation of thick peat deposits. The Second coal seam of the Mariza East deposit (maximum thickness of 25 m) is a good illustration of the long duration of the peat formation process. The coal-bearing strata consist of terrigenous sediments varying in thickness from 45 m (minimum at Mariza East) to 390m (maximum at Elhovo) where 3-4 coal seams form coal-bearing measures of varying thickness (up to a maximum of 25 m). Paleogeographical reconstruction shows that the water table in the eutrophic swamps was generally high, falling very low only during seasonal desiccations. An interesting phenomenon is the small coal deposit in the area of the Gulf of Sozopol, where a thin coal layer with preserved root systems
Number of coal seams
Geological age
3 2 3 3 4 3 1-5 1-3 1 1 3 16 1 4 8 14
Pliocene Pliocene Pliocene Early-Middle Miocene Early-Middle Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene Middle Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene Pliocene-Late Miocene
under the silicified 'Stone Forest' has been found. Geological and petrological studies have shown that the coal formation occurred in a swamp situated in the crater of an extinct volcano which was destroyed by a storm during which marine water influx produced a rapid change of geochemical conditions. The coal layer is covered with pyritized wood tissue with upright silicified Taxodium stumps mostly covered by recent sands (Si~kov et al. 1988).
Sofia coal-bearing province This province covers the area of Western Bulgaria between the Balkan and the Sredna Gora Mountains. The coal-bearing sediments fill subequatorial grabens and small depressions in which coal formation started in the Middle Miocene (Chukurovo) and lasted till the beginning of the Pliocene (Stanjanzy). In this highly active tectonic zone continental lakes were formed, later to be rapidly filled with terrigenous sediments. Gradually these lakes were transformed into eutrophic swamps with a high degree of mineralization and highly dynamic ground waters. The coal-bearing strata consist of clastic sediments - conglomerates, sands, clays. Their thickness varies from 25 to 100 m while in the outlying parts of the deposit it reaches 900 m (Sofia). The number of coal seams varies from 1 to 5, except at Chukurovo where 16 coal seams are present in the coal-bearing sequence. The thickness of the coal seams varies up to a maximum of 15-20 m.
BULGARIAN LOW RANK COALS
Strimon-Mesta coal-bearing province This province is located in the southwestern part of Bulgaria. Several small coal deposits are situated along the valleys of the Strimon and Mesta rivers in almost meridional orientation. The coal deposits were formed in restricted and small grabens filled with coarse clastic terrigenous sediments. The thickness of the coalbearing strata varies reaching a maximum of 800m (Goze Delchev). In the highly active tectonic zone favourable conditions for the formation of mezotrophic to oligotrophic swamps occurred resulting in a variable number of coal s e a m s - 1 (Kjustendil) to 14 (Goze Delchev). Laterally their thickness changes very rapidly. The coal seams very often wedge out and split into benches and lenses of coaly clay and clay.
Dacian coal-bearing province This province is located mostly in the northern areas of the Moesian Platform including the Lom and the Kozloduj deposits, the latter representing the southern fragments of the Oltenia basin in Romania. The coal formation is connected with the desiccation of the Pontian Basin during a prolonged regression in the Pliocene (Kojumdgieva 1983). The coal sediments are of Pliocene age - Dacian-Romanian, with the exception of
143
some small coal deposits which are older (Kojumdgieva & Popov 1988). In various parts of the Lom deposit the deltaic sedimentary complex differs in structure and thickness as illustrated by the irregular alternation of the lithological bodies, represented by coarser to finer clastic sediments. These sediments contain unevenly developed benches and lenses which vary in number; a typical feature of a subaeral delta. The finer pelitic sediments of the Lom area are typical of the lower-upper delta plain (Si~kov & Angelov 1984). In the Lom depression extensive but uneven peat accumulation occurred associated with a vast lower-upper delta plain in which lateral facies migration was developed by a northward shift of the deltaic front and contained fluvial channels. The coal formation in the zone of the Kozloduj deposit developed in a coastal plain environment. The swamp was large. In contrast to the Lom deposit the coals were found in interdeltaic environments with a high water table and associated with fine-grained clastic sediments. The coal seams are up to 4 m thick and cover an area of about 1 km 2 (Si~kov & Angelov, unpublished data).
Petrology of the Neogene coals Over 6200 coal samples were collected from opencast and underground mines, exploratory boreholes and outcrops.
Table 2. Maceral composition of the Bulgarian low rank coals No.
Coal deposits
Ash Maceral composition (%) A d (%) Total H
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
Lom Kozloduj Elhovo Mariza East (briqueting) Mariza East (energetic) Mariza West Sofia Beli Brjag Aldomirovzy Stanjanzy Karlovo Kovachevzy Chukurovo Kjustendil Oranovo Razlog Goze Delchev
42.6 20.1 38.3 20.2 39.7 23.8 16.6 15.0 25.4 21.9 17.7 33.2 22.3 23.1 20.1 20.8 17.6
53 71 60 74 61 63 77 66 64 74 70 68 75 36 63 75 80
L
I
3 7 2
1 1 2
1
2
2 8 7 7 8 8 8 4 14 3 7 6 11
1 4 2 8 6 5 5 6 1 15 3 1 3
Org. matter
M
H
43 21 36 23 36 25 14 19 22 13 17 22 10 46 27 18 6
92 90 96 96 95 82 89 87 88 85 83 89 84 67 88 92 85
L 5 8 2
I 3 2 2
1
3
3 12 9 9 11 8 10 5 15 5 10 7 11
2 6 2 4 1 7 7 6 1 28 2 1 4
144
G . D . SISKOV
Petrological data show that the homogeneous genotypes gelide, peptide, liptide and fiside form the coal matter in various proportions, though gelide and peptide predominate. The heterogeneous genotypes gelofuside, fusoliptide, etc. are also represent. The wide genotype differentiation is caused by: (1) the different chemical, anatomical, species and plant ecosystem compositions of the coal-forming plants and their behaviour during peat formation; (2) the physico-chemical parameters (Eh, pH) of the ground water environment; (3) the type and rate of the chemical reactions and petrological processes depending on the oxygen supply; (4) the activation energy involved in the phytogenic matter transformation during microbial metabolic processes (Si~kov 1988). In the lithotype balance humoclarite is prevalent (up to 85%). A higher quantity of xylain and liptain(the latter is a specific lithotype formed by impregnated coniferous fragments with resins - Si~kov 1976) fragments of various size are found mainly in the coal deposits from the Thracian and the Sofia coal-bearing provinces. They are distributed chaotically. Semifusain and fusain are found in small amounts, and occasionally mark bedding in the coal seams of some deposits from the Thracian (Elhovo, Mariza East) and Strimon-Mesta provinces (Kjustendil, Goze Delchev). All mono-, bi- and trimacerites are represented in the microlithotype composition. Humoclarite prevails and, along with carbargilite represents partially to completely disintegrated plants, consisting of variously fragmented atrinite, and more rarely, of densinite. The average compositions of the maceral analyses are given in Table 2. The macerals of the three groups are in different proportions, with the huminite macerals predominating. They form the groundmass in which other fragmentary macerals of the huminite, liptinite and inertinite groups are chaotically dispersed. Atrinite and densinite are present in differing amounts. Liptinite macerals are represented mainly as resinite and more rarely as cutinite, sporinite and suberinite. The mineral components are represented by widely varying amounts of clays and pyrite. The clay minerals consist of kaolinite, illite and montmorilonite. The maceral amount recalculated per organic matter shows that the Neogene coals are relatively homogeneous in composition containing more than 80% of huminite, with the exception of the Kjustendil coals (Table 2, Fig. 2). The average petrological data show
H
RANK
80%
80%
A~-/. SHALES 75 SHALEY
W m o ,< u.
50 35
~A~
,~,1I 4
ASHY
1(
,I
1
I,'i5i 'r IIII
PURE Rr %
BRIGHT 16 17
15
MAT Z 0.3 < n"
LIGNITE
0.2
~Io I
10,S 862 -Tg 4 1~3[ 11 |
fILIal1
100% H TYPE
80% L
Fig. 2. Triangular diagram and petrological position by type, facies and rank of the Neogene coals in Bulgaria according to Alpern's classification: 1, Lom; 2, Kozloduj; 3, Elhovo; 4, 5, Mariza East; 6, Mariza West; 7, Sofia; 8, Beli Brjag; 9, Aldomirovzy; 10, Stanjanzy; 11, Karlovo; 12, Kovachevzy; 13, Chukurovo; 15, Oranovo; 16, Razlog; 17, Goze Delchev. that the Neogene coals in Bulgaria are of huminite type with a minimum content of liptinite and inertinite. The principal components are determined on the basis of a correlation matrix using the algorithm reported by Wehlstedt & Davis (1968). The Q-dendrograph was plotted by the method of McCammon&Wenninnger (1970) (Si~kov & Andreev 1987). The cluster analyses indicate that the Neogene coals are very distinctly differentiated into three groups (Fig. 3, Table 3).
BULGARIAN LOW RANK COALS
I 0.06
I B
I
[
J
0.04
I I
] A2
i
I B2
B~
0.02
2 7 16 8 17 15 9 40
13
6
11 10
12
50
1 4 5 3
60
1PC = 0.74 H - 0.08 L - 0.66 I (79%) 2PC = - 0.34 H - 0.81L - 0.48 I (21%)
HII
1Q ~3 \
/ ./
14
A
\
-20
- 30
Fig. 3. Q-dendrograph and grouping of the Neogene coal deposits by their maceral composition: 1, Lore; 2, Kozloduj; 3, Elhovo; 4, 5, Mariza East; 6, Mariza West; 7, Sofia; 8, Bell Brjag; 9, Aldomirovzy; 10, Stanjanzy; 11, Karlovo; 12, Kovachevzy; 13, Chukurovo; 14, Kjustendil; 15, Oranovo; 16, Razlog; 17, Goze Delchev.
Groups A and B are each subdivided into two subgroups. Coals, having huminite 82-92%, are referred to Group A. Coals, containing huminite above 92%, are assigned to Group B. A significant difference is observed in the Kjustendil deposit due to the higher content of inertinite (Group C). The grouping of the coals in accordance with the ratios (H/I) and (L/I, H) is illustrated in Fig. 3, where the H/I ratio has an average value of 79%. The statistical analysis of the maceral content provides a means of reconstructing the peatforming ecosystems in combination with the
tectonic position of the coal deposits, the physico-chemical environment in the peat bogs and the hydrodynamics of the groundwater table. Coals in which the coal-forming communities are equally divided between forest and herbaceous populations are referred to Group A. The petrological investigations show that the forest vegetation consists of angiosperm and coniferous species. The coniferous species in combination with herbaceous vegetation form Subgroup A~ where the amount of densinite groundmass considerably predominates. The amount of well preserved wood tissue producing textinite and textoulminite filled with resinite also increases. Coals from three deposits form Subgroup A2 (Table 3) in which herbaceous vegetation predominates while the number of angiosperm and coniferous species is equivalent. The amount of densinite is prevalent and the content of euulminite and gelinite increases. The coals from the Chukurovo deposit are distinct. They contain mainly species, Taxodiaceae being particularly typical. Due to the prevalence of coniferous vegetation the amount of resinite also increases markedly. Coals predominantly formed by herbaceous vegetation in peat-producing ecosystems with a lower participation of forest species are assigned to Group B. In this case huminite macerals are highly gelified and the groundmass consists of atrinite. Their division into two subgroups is based on the amount of inertinite. Another important petrological factor, related to the genesis of coal macerals is the nature of the peat swamps and their hydrodynamics. Where major tectonic activity and enhanced hydrodynamics and aeration of ground water occurs, possible seasonal desiccation would result in oxidation leading to a deviation from the general maceral balance. This is the reason
Table 3. Petrological groups of the Neogene coal deposits in Bulgaria Group Subgroup Maceral composition (%) Coal deposits H A
B
C
L
I
A1
85-92
7-11
1-4
A2
82-85
8-12
6-7
B1
> 92
B2
89 67
1-5 5 15
2-3 6 28
145
Kozloduj (2), Sofia (7), Beli Brjag (8), Aldomirovzy (9), Oranovo (15), Razlog (16), Goze Delchev (17) Mariza West (6), Stanjanzy (10), Karlovo (11), Chukurovo (13) Lom (1), Elhovo (3), Mariza East (4, 5) Kovachevzy (12) Kjustendil (14)
Table 4. Rank parameters of the Neogene coals in Bulgaria Coal deposits
Relectance Rr (%)
Bed moisture W r (%)
Carbon Volatile matter C daf (%) VM daf (%)
Calorific value Osdaf (kJ/kg)
Lom Kozloduj Elhovo Mariza East Mariza West Sofia Beli Brjag Aldomirovzy Stanjanzy Kovachevzy Karlovo Chukurovo Kjustendil Oranovo Razlog Goze Delchev
0.13 0.22 0.18 0.20 0.21 0.22 0.22 0.21 0.21 0.22 0.19 0.20 0.33 0.34 0.39 0.41
50.0 51.8 63.6 64.4 43.1 50.0 47.5 51.6 52.5 48.6 44.0 33.0 26.6 30.4 32.5 43.5
64.7 65.1 63.8 65.0 62.3 64.6 63.8 64.1 61.6 63.9 64.8 64.3 66.9 66.0 64.8 67.9
22.52 22.68 22.76 22.31 22.87 23.51 23.27 22.92 19.42 22.78 23.77 23.38 23.98 23.85 22.92 23.02
LOW RANK
67.7 64.1 53.0 55.8 58.8 52.0 53.4 52.6 61.0 67.3 58.8 57.0 51.0 48.3 55.4 49.1
I TRANSITIONzoNE ]
MEDIUMRANK .x
I -24 MJ/Ro
.--
Xx
,
[ ~
.
~
~
~ -
21.
Q = 21.569+ 6.316Rr - 3.183Rr2 r = 0.653, n = 27
19
.
.
0.4
0.6
l~ls~3~tl
L
0.8 ,
,,
1~0 ,
%RF
l,,
Stand. LIGNITE
ASH
HIGHVOL BITUMINOUS
SUB. ~ B I T ' ~ - ~
r xx xx x ~ ~
0.2 o~
6al x
x
x
~
0.4
~ ,,~mll-"~ 0.2
0.8
1.0
VM = 69.187
r = 0.836, n-26~7"15Rr + 26"885Rr2
x
x x
_x
0.2 o2
x
0.6
~ x
75 70
W = 76.602 - 159.664Rr + 101.913Rr 2 r = 0.862, n=27
x
0.4
2 )(~'
0.6
x
x x
x
x 0.4
0.8
1.0
xx r = 0.924, n=27
0.6
O.8
1.0
%Rr
Fig. 4. Rank classification of the Cenozoic coals in Bulgaria and the position of the Neogene coals according to Alpern's classification, Bulgarian Standard and ASTM as well as relationship of reflectance (Rr, %), bed moisture (W r, %), volatile matter (VM daf, %), carbon content (C daf, %) and calorific value (Qdaf, MJ/kg).
BULGARIAN LOW RANK COALS
The author would like to thank Dr A. Andreev (Geological Institute of the Bulgarian Academy of Sciences) for computation of the maceral data. The author is also indebted to Dr I. Todorov (Bulgarian Research & Services Group Ltd) for the preparation of the figures.
for the difference in the coals from the Kjustendil and Kovachevzy deposits. The results of the maceral analysis of our study appear to be in agreement with the published indices (Diessel 1986; Calder et al. 1991). The coal facies determined on the basis of the ash content and maceral composition are illustrated in Fig. 2, where the position of macerals is projected on the abscissa. The content of ash is between 17.6 wt % (Goze Delchev) and 42.6 wt % (Lore). Most of the coals belong to the ashy humic facies. Only coals from the Lom and Elhovo deposits as well as a part of the Mariza East deposit are coally humic facies. The average classification parameters - huminite (gelinite) reflectance (Rr, %), calorific value (Qsdaf, MJ/kg), bed moisture (W r, %), yield of volatile matter ( V M ~ %) and carbon content (C dav, %), are given in Table 4. The position of the Neogene coals is plotted in Fig. 4. There are good correlations between the huminite reflectance and the other classification parameters.
Conclusion The Neogene coals in Bulgaria belong to the low rank coals of Alpern's classification. On the ASTM scheme they are determined as lignites. According to the Bulgarian Standard the coals are divided into lignites (Class O1) and mat brown coals (Class O2), with the exception of the coals from the Goze Delchev deposit which are referred to the bright brown coals (Class 03). The low coalification of the coals from the Thracian, Sofia and Dacian coal-bearing provinces is a consequence of the relatively small thickness of the overburden sediments and the short geological time for the development of the coalification process. The coalification degree of the coals from the Strimon-Mesta province is higher and depends on: (1) the thickness of the overburden sediments (to 500m) which has caused the reorientation of the coal genotypes (lithotypes, microlithotypes, macerals) and the appearance of macro- and microbedding; and (2) an anomalous geothermal gradient. Velinov&Bojadgieva (1981) have stated that the temperature measured in the boreholes at 300 m depth is 37-38~ and at 500m depth, 66~ (3) the tectonic mobility - rapid tectonic movements occurring along faults, typical of the zone of the Kraishte lineament.
147
References ALPERN, B. 1981. Pour une classification syntetique universelle des combustibles. In: La geologie des charbons, des schistes bitumineux et des kerogenes, 271-290. CALDER, J. H., GIBBING,M. R. • MUKHOPADHYAY,P. K. 1991. Peat formation in a Westphalian B piendment setting, Cumberland Basin, Nova Scotia: implications for the maceral-based interpreparation of rheotrophic and raised paleomires. Bulletin de la Societe geologique de France, 162, 2, 283-298. DIESSEL, C. F. K. 1986. On the correlation between coal facies and depositional environments. In. Advances in the Study of the Sydney Basin. Proceedings of 20th Newcastle Symposium~ 19-22. |VANOV, Z. 1983. Apercu general sur l'evolution geologique et structurale des Balkanides. In: Guide de l'exeursion. University Press of Sofia, 3-26. KOJUMDGIEVA,E. 1983. Paleogeographic environment during the desiccation of the Black Sea. Palaeogeography, Palaeoclimatology, Palaeoecology, 43, 195-204. & PoPov, N. 1988 . Lithostratigraphy of the Neogene sediments in Bulgaria. Palaeontology, Stratigraphy, Lithology, 25, 3-26. MCCAMMON, R. I. t~ WENNINNGER, G. 1970. The dendrograph. Kansas Geological Survey Computer Control, 48. PANOV, G. 1982. Tertiary coal sedimentation in the Upperthracian depression. PhD thesis. SlgI~OV, G. 1976. Liptain- properties and genesis. Annuaire de l'Universite de Sofia, Faculte de Geoloque et Geographie, 67, 1, 151-169. 1988. Theoretical fundamentals of biochemical eoalification. Kliment Ohridski University Press, Sofia. - - & ANDREEV,A. 1987. A way to reconstruct coalforming peleoplant communities based on the micropetrographic composition of Bulgarian Neogene coals. Comptes rendus de l'Academie bulgar des sciences, 40, 4, 77-80. & ANGELOV, A. 1984. Delta-plain model of sedimentation of the Lom lignite basin. Comptes rendus de l'Academie bulgar des sciences, 37, 11, 1531-1533. - - & VALCEVA,S. 1983. Petrological nomenclature of lignites and brown coals. Comptes rendus de l'Academie bulgar des sciences, 36, 6, 799-801. SISKOV, G., STEFANOVA, U. t~ ZLATEV, I. 1986. Petrological characteristics of the coals from the Brod member in the West Mariza basin. Annuaire de l'Universite de Sofia, Faculte de Geologie et Geographie, 76, l, 40-53.
-
-
148
G.D.
SISKOV, G., VALCEVA, S. & PIMPIREV, H. 1988. Preconditions for the formation of the 'Stone forest' and coal deposits in the Gulf of Sozopol. Annuaire de l'Universite de Sofia, Faculte de Geologie et geographie, 77, 1, 190--200.
SISKOV VELINOV, T. & BOJADGIEVA, K. 1981. Geothermal investigations in Bulgaria. Technika, Sofia. WEHLSTEDT, W. C. & DAVIS, J. C. 1968. Fortran IV program for computation and display of principal components. Kansas Geological Survey Computer Control, 21.
Coal petrology and facies associations of the South Yakutian Coal Basin, Siberia I. E. S T U K A L O V A
Geological Institute of Russian Academy of Sciences, Pyzhevsky per., 7, Moscow, 109017, Russia Abstract: The South Yakutian Coal Basin consists of isolated depressions in the south of the Yakutian region of Eastern Siberia in Russia, which are filled with Jurassic and Cretaceous coal-bearing sediments. The basin contains considerable resources of highquality bituminous coals, with a caking index (y) of 6-21 mm. The coal-bearing formation contains coal seams of various thickness, from 0.5-2.5m up to 15.0m and above. Some of them are near the surface and can be mined by opencut methods. Genetic types and facies of the Mesozoic coal-bearing strata have been recognised as proluvial, alluvial, delta, lacustrine-swamp and peat bogs sediments. The petrographic composition of the sandstones identified four terrigenous-mineral associations, which are represented by true arkoses, greywacke arkoses, feldspar-quartz and quartz-feldspar greywackes. Humic coals consists of vitrinite (70-90%), inertinite (10-20%) and liptinite (0-10%) maceral groups. In the vitrinite maceral group there is a high percentage of telinite macerals, up to 60-65%. The basin is characterized by a high alteration of Mesozoic sediments and organic matter. Coals are of middle and high rank (0.65-2.15% R0). There are high concentrations of bitumen in chloroform extracts of organic matter from the coals, from 0.0938% to 3.6466%. The mineral matter of terrigenous rocks is altered to the catagenetic stage and the metagenetic stage. The South Yakutian coals are of high quality because of their rank and composition.
The South Yakutian Coal Basin is located in the south of the Yakutian region of Eastern Siberia in Russia, between 56~ ~ North and 120~ ~ East (Fig. 1). The basin extends from the Oliokma river in the west to the Uchur fiver in the east and covers an area of 25 100 square km. The towns of Aldan, Nerjungri and Chulman
-
126~
120 0
Q i
132.* E
75 150Km *
.
;OON
I
Fig. 1. Sketch map of the South Yakutian Coal Basin (revised after Bredihin 1973). 1, Mesozoic coal-bearing deposits; 2, Coal-bearing regions: (1) Usmun, (2) Aldano-Chulman, (3) Tokin.
are situated in the region. It is a very important coal basin in Russia, and contains considerable resources of high-quality bituminous coals. The coal-bearing formation contains coal seams varying in thickness from 0.5-2.5m to greater than 15.0 m. Some of the seams are near the surface and can be mined by opencut methods. The South Yakutian Coal Basin includes three main coal-bearing regions, namely: Usmun, A l d a n o - C h u l m a n and Tokin regions (Fig. 1). The Nerjungri coalfield in the A l d a n o - C h u l m a n region contains caking coals which are now opencut mined, producing about 9 Mt per year. The Nerjungri coalfield is situated near the rail station at Chulman town and the coals are exported to the Far East region of Russia and to Japan. The Elga coalfield in the Tokin region also contains caking coals, but the coalfield is situated far from the rail system and it will be opencut mined in the future. The stratigraphy and coal-bearing potential of the sediments in the South Yakutian Coal Basin have been investigated by many authors (Prosviryakova 1961; Mokrinsky 1961; Waltz 1961; Fatkulin et al. 1970; Bredihin 1973; Prilutsky 1979; Zhelinsky 1980; Markovich 1981; Vlasov 1981; Nazarov & Stukalova 1991) and are now well known; but many problems have yet to be solved. At present there is no information about reasons for the high degree of metamorphism of the South Yakutian coals.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 149-160.
150
I. E. STUKALOVA
There is no concensus about the tectonic position of the South Yakutian Coal Basin. Some investigators consider that the basin formed in intermontain depressions (Bredihin 1973; Terentyev 1979; Vlasov 1981); others that the depressions are paleorifts (Nazarov & Stukalova 1991).
Geological setting and stratigraphy The South Yakutian Coal Basin consists of isolated depressions in the south of the Siberian platform, which are filled with Jurassic and Cretaceous coal-bearing sediments (Fig. 2). The Mesozoic aulacogens are related to major region faults. The Mesozoic sediments are deposited on Archean-Proterozoic igneous and high-grade metamorphic basement. Precambrian granites and gneisses are widespread in the region and metamorphosed. Lower Cambrian carbonate
rocks also occur. Mesozoic rocks are more than 3500m thick and are represented by various members of terrigenous strata, ranging from Early Jurassic to Early Cretaceous in age. Lower Cretaceous terrigenous sediments without coals of the Nagornaya Member are developed in the south of the region as are Cretaceous volcanic sediments of the Karaulov Member. Jurassic-Cretaceous alkaline intrusive rocks also occur (Fig. 2). Coal-bearing strata are developed in the Usmun and Aldano-Chulman regions, where Early Jurassic deposits of the Juhta Member consist of conglomerates and coarse and medium sandstones of proluvial origin (Fig. 3). Middle Jurassic sediments of the Duraji Member are represented by coarse, medium and fine ground alluvial sandstones with many coal seams between 0.5-2.0m thickness. Upper Jurassic sediments comprise three Members: Kabakta, Berkakit and Nerjungra. The Kabakta
Fig. 2. Geological map of the South Yakutian Coal Basin, Siberia (revised after Zhelinsky, 1980). 1-4, Mesozoic coal-bearing deposits: (1) Lower Jurassic, Juhta Member (J1); (2) Middle Jurassic, Duraji Member (J2); (3) Upper Jurassic, Kabakta, Berkakit and Nerjungra Members 03); (4) Lower Cretaceous, Holodnican Member (Crl); (5) Lower Cretaceous terrigenous sediments of Nagornaya Member; (6) Cretaceous volcanic sediments of Karaulov Member; (7) Jurassic-Cretaceous alkaline intrusive rocks; (8) Lower Cambrian carbonate metamorphic rocks; (9) Precambrian igneous and metamorphic rocks; (10) faults; (11) Coal-bearing regions: 1, Usmun, 2, Aldano-Chulman, 3, Tokin; (12) profiles with boreholes in regions: (AB) Usmun region, (CD) Aldano-Chulman region, (EF) Tokin region.
SOUTH YAKUTIAN COAL BASIN
/./sare/t re eian
,4,~#erzoClqu[azo'/t /'e.Cte/t
151
@~i~ r e j i ~#
.
~9":'. .."~
~K. , , ~~:% ~ m ~
~.
~9 .
.... '7
o~
- ~..
~
.~--~
Fig. 3. Stratigraphy, lithology and coal rank in coal-bearing strata in three regions of the South Yakutian Coal Basin.
and Berkakit Members consist of nearshore medium and fine grained sandstones and coarse and fine grained siltstones and mudstones with paralic coal seams 1.0-2.5 m in thickness. The Nerjungra Member is formed of alluvial and deltaic coarse, medium and fine-grained sandstones and coarse and fine-grained siltstones with coal seams of great thickness up to 25.030.0m in the Aldano-Chulman region. The high-quality caking coals of the Nerjungry coalfield in the Aldano-Chulman region are mined by opencut methods. In the Usmun region the Nerjungry Member is missing. In the Tokin region the Nerjungra Member is represented by proluvial and alluvial coarse and medium grained sandstones with thin coal seams. Early Cretaceous sediments of the
Holodnican Member contain, in the Tokin region, alluvial conglomerates, coarse, medium and fine-grained santstones and coarse and finegrained siltstones with thick coal seams (5.0-10.0 m). In the Aldano-Chulman region the Early Cretaceous sediments consist of alluvial conglomerates and sandstones without coals, and in the Usmun region Early Cretaceous sediments are absent (Fig. 3). The strata are cyclic and contain a complex of floral fragments and palinology (Mokrinsky 1961; Bredihin 1973). The age of coal-bearing strata has been determined by the flora and other fossils. The Jurassic sequence is characterized by the flora Annulariopsis microphylla Vassil., Neocalamites sp., Phlebopteris cf. polypodiodes Brougn., Czekanowskia Setacea Heer,
152
I. E. S T U K A L O V A
SOUTH YAKUTIAN COAL BASIN Raphaelia diamensis Sew, Cladophlebis serrulata Sam and others (Prosviryakova 1961; Bredihin 1973; Markovich 1981). The cretaceous rocks contain the flora Equisetites asiaticus Pryn., Ctenis yokoyamai Kr. et Pryn., cf. burejensis Pryn., Coniopteris nymphrum Heer, Czekanowskia rigida Heer, Pityophyllum nordenskioldii (Heer) Nath., and others (Prosviryakova 1961; Markovich 1981). The palaeoenvironment and facies of the Mesozoic coal-bearing strata have been determined as proluvial, alluvial, delta, lacustrineswamp and peat bogs (Nazarov & Stukalova 1991).
153
too"/.
s
too%
F
1s
Petrographic composition of coals One of the aims of the study of the South Yakutian coals was to explain the high caking characteristics of the coals. We investigated coals in the Usmun, A l d a n o - C h u l m a n and Tokin regions of the South Yakutian Coal Basin. The method was to define the maceral composition, and the technological and chemical properties of the very thick coals, such as the Upper Jurassic coal seams in the A l d a n o Chulman region and the Lower Cretaceous coal seams in the Tokin region. The coals were analyzed microscopically, and by chemical and organic geochemical methods. We investigated the maceral composition of the coals using transmitted light with thin sections and incident light with polished sections with magnifications of 20-600x. Humic coals consist of vitrinite (70-90%), inertinite (10-20%) and liptinite (0-10%) maceral groups. The vitrinite maceral group contains a high percentage of telinite macerals, up to 60-65% (Fig. 4). For example, bituminous coals with high volatile matter content in the Elga coalfield of the Tokin region, using an oil immersion objective, was shown to contain: inertinite components (I) representing transformed wood fragments (sample 375.0/128 and sample 410.0/128, Fig. 4, photo 1,2); a high percentage, up to 70-80%, of vitrinite components (Vt) represented by telinite (Vtl) and collinite (Vt2) (sample 505.0/128 and sample 320.0/131, Fig. 4, photo 3,4); and liptinite
Fig. 5. Terrigenous-mineral associations of sandstones in the coal-bearing strata of the South Yakutian Coal Basin (revised after Nazarov & Stukalova, 1991). Diagram by Shutov 1972. Explanation: Q, quartz; F, feldspar; R, rock fragments; (1) true arkoses and feldspar~luartz sandstones with a low content of quartzite, granite and gneiss fragments; typical of the Lower Jurassic association in the Aldano~hulman and the Usmun regions; (2) feldspar~luartz greywackes and greywacke arkoses with a high content of quartzite, granite and gneiss fragments; typical of the Middle and Upper Jurassic association in the three regions; (3)'greywacke arkoses, feldspar-quartz greywackes and quartz-feldspar greywackes with high content of quartz porphyry, felsitic porphyry, andesite, granite and trachyte fragments; typical of the Upper Jurassic and Lower Cretaceous association in the Aldano-Chulman region; (4) greywacke arkoses and quartz-feldspar greywackes with a high content of rhyolite, felsitic porphyry, trachyte and their tuff fragments; typical of the Lower Cretaceous association in the Tokin region. components (L) represented by cutinite and resinite (sample 225.0/131 and sample 325.0/ 131, Fig. 4, photo 5,6).
Proximate analyses and vitrinite reflectance values Chemical analyses of the South Yakutian coals gave moisture contents (W daf) from 0.20% to
Fig. 4. Petrographic composition of bituninous coals with high volatile matter in the Tokin region (Elga coalfield) of the South Yakutian Coal Basin, Lower Cretaceous, Holodnikan Member, incident light, oil immersion, magnification x300. Indices: Vtl, telinite; Vt2, collinite; I, inertinite; L, liptinite. Photo 1. Sample 375.0/128. Inertinite components, borehole 128, depth 375.0 m. Photo 2. Sample 410.0/128. Inertinite components, borehole 128, depth 410.0 m. Photo 3. Sample 505.0/128. Vitrinite components (collinite), borehole 128, depth 505.0 m. Photo 4. Sample 320.0/131. Vitrinite components (telinite), borehole 131, depth 320.0m. Photo 5. Sample 225.10/131. Liptinite components (cutinite and resinite), borehole 131, depth 225.0 m. Photo 6. Sample 325.0/131. Liptinite components (cutinite) and vitrinite components (telinite), borehole 131, depth 325.0 m.
154
I. E. STUKALOVA
J-W ,
A/-E
A
B 2z
i
177
50
51]
53
::: x ,~.~,
~'~
. ~
9~
21
~ [ ~ [7~]~. 7 0,5~ o
[r/ s
Fzz:I,
IX.xls
-B]e
Fig. 6. Lithology and terrigenous-mineral associations of the coal-bearing strata in the Aldano Chulman region of the South Yakutian Coal Basin, Siberia. Expiation for figures 6-8: 1, conglomerates; 2, sandstones; 3, siltstones and mudstones; 4, coal seams; 5, true arkoses and feldspar-quartz sandstones with a low content of quartzite, granite and gneiss fragments; typical of the Lower Jurassic association in the Aldano Chulman and the Usmun regions; 6, feldspar-quartz greywackes and greywacke arkoses with a high content of quartzite, granite and gneiss fragments; typical of the Middle and Upper Jurassic association in the three regions; 7, greywacke arkoses, feldspar-quartz greywackes and quartz-feldspar greywackes with a high content of quartz porphyry, felsitic porphyry, andesite, granite and trachyte fragments; typical of the Upper Jurassic and Lower Cretaceous association in the Aldano-Chulman region; 8, greywacke arkoses and quartz-feldspar greywackes with a high content of rhyolite, felsitic porphyry, trachyte and their tuff fragments; typical of the Lower Cretaceous association in the Tokin region; 9, isoreflectance lines; 10, boreholes. 2.80%, volatile matter contents (V daf) ranging from 17.0% to 40.0%, ash content (A daf) varying from 1.57% to 25.92%, caking index (y) of 6-21 ram. Vitrinite reflectance was measured according to ICCP standards (Stach 1982) on polished sections of the coals using a microscope-photometer 'MRE-Leitz', with a magnification of 600x. Jurassic and Cretaceous coals in the South Yakutian Coal Basin have middle and high rank, with vitrinite reflectance values (R0)
of 0.65-2.15%. In all the boreholes investigated (more than 50 boreholes with more than 500 samples) the vitrinite reflectance values increase with depth. In the Usmun region the vitrinite reflectance (Ro) of organic matter varies from 0.55% to 1.00%. On the profile A - B of the Usmun region (Fig. 6) the isoreflectance lines of 0.55% R0 and 0.85% R0 are indicated. These are high volatile bituminous coals (Fig. 3). In the A l d a n o - C h u l m a n region the vitrinite reflectance varies from 1.15% to 2.15%. The profile C - D of
SOUTH YAKUTIAN COAL BASIN
I
G 2531
1868
' 253~ %,~
155
_
\
LL2811m
44
O
2,
'
4 X~ I
2785
'
--
.
.
.
.
.
Z 7 8 4 277S
-~i
-~~i ~2777 ,.'.N ~
,,-%
4)
2. ,
..-~, " :-~X' ~"
A ~
'x'~r~
,0
11
',-2. "-I,55R ~
~
'-2" 9
ii:l~
o,
t-I
-Ro
Fig. 7. Lithology and terrigenous-mineral associations of the coal-bearing strata in the Aldano-Chulman region of the South Yakutian Coal Basin, Siberia. the Aldano-Chulman region (Fig. 7) shows isoreflectance lines of 1.55% R0 and 2.00% Ro. These are low volatile bituminous coals (Fig. 3). In the Tokin region of the South Yakutian Coal Basin vitrinite reflectance ranges from 0.65% to 1.15%. On the profile E-F of the Tokin region (Fig. 8) isoreflectance lines of 0.75%Ro, 0.85% R0 and 1.00% R0 are shown. These are high volatile bituminous coals (Fig. 3).
Bitumen
analyses
In order to better understand the effects of metamorphism on the technological and chemical properties of the coals (Teichmuller 1974,
1990; Puttmann et al. 1985) bitumen analyses and column chromatography were carried out. Samples from different coalfields in the South Yakutian Basin were investigated. In the Usmun region we investigated high volatile coals from three boreholes, No 50, 165 and 203 in the Syllach coalfield. The vitrinite reflectance (R0) is 0.75%, the volatile matter is about 33-40% and the caking index is 6-21 ram. In the Tokin region high volatile coals were investigated with a vitrinite reflectance of about 1.0%, volatile matter of about 27-40% and a caking index is 6-21 mm. The coals are from boreholes No 13, 108, 110, 146, 158 and 160 in the Elga coalfield. In the Aldano-Chulman region low volatile coals were investigated with 1.5%R0 and
156
I. E. STUKALOVA
N-s F
E 128 l,~176
".[~ ~
0
,
77
.', 9
p
9
.=. ,~!v ~ / /
'
131 .~-
ltIO)~
0 85 "
!::I
1 I
2Kin I
12
15
,'-;
~
""
~
8
~176
I~1
z_z]
0 I
-" g,o '
9
,~/
X *~,o-
I
i
:.:~< .r
~
1B ~
9
i
/
//~
~;
.,'
-'-' 9
~
. "'" ~1
.
1,0 Ro
Fig. 8. Lithology and terrigenous-mineral associations of the coal-bearing strata in the Tokin region of the South Yakutian Coal Basin, Siberia. 17-22% volatile matter. The caking index is up to 6-21 mm. The samples were from borehole No 3455 in Nerungri coalfield. According bitumen analyses, the fifteen investigated samples contain water (H20) from 0.08% to 2.71%. The contents of ash varies from 9.20% to 40.75% and CO2 from 0.80% to 10.00%. Insoluble organic matter varies between 77.88%-95.05%. No humic acids were detected due to the high alteration of sediments. Extracts of organic matter in chloroform from the coals
in the basin yielded high levels of bitumen. Concentrations of chloroformic bitumen vary from 0.0938% to 3.6466%, of alcohol-benzolic bitumen (A) from 0.0345% to 3.7266% and of alcohol-benzolic bitumen (C) from 0.0322% to 1.886%. Concentrations of total bitumen range from 0.1635% to 7.9838%. Chloroform extracts were separated into different fractions by column chromatography. The concentrations of methano-naphthene oils are from 6.25% to 26.21%, and of polyaromatic
Fig. 9. Petrographic composition of sandstones in coal-bearing strata in the Aldano-Chulman region of the South Yakutian Coal Basin, transmitted light, crossed polars, magnification • Indices: Q, quartz; P1, plagioclases; F, feldspars; R, rock fragments. Photo 1. Sample 27.5/1868. Lower Cretaceous, Holodnican Member, borehole 1868, depth 27.5 m. Greywacke sandstone with high content of quartz and felsitic porphyry, quartzites and granites, sericitization of feldspar. Photo 2. Sample 74.5/1868. Lower Cretaceous, Holodnican Member, borehole 1868, depth 74.5 m. Greywacke sandstone with high content of quartz and felsitic porphyry, quartzites and granites, chloritic and laumontite cement, sericitization of feldspar. Photo 3. Sample 560.1/2777. Middle Jurassic, Duraji Member, borehole 2777, depth 560.1 m. Quartz-feldspar greywacke with high content of felsitic and quartz porphyry, granites, gneisses and rare trachytes, hydromica and laumontite cement. Photo 4. Sample 590.4/2777. Middle Jurassic, Duraji Member, borehole 2777, depth 590.4 m. Feldspar-quartz greywacke with high content of granites and gneisses, hydromica and quartz cement. Photo 5. Sample 837.9/2777. Lower Jurassic, Juhta Member, borehole 2777, depth 837.9 m. Feldspar-quartz sandstone with few quartzite, granite and rarely dolomite fragments, hydromica and laumontite cement. Photo 6. Sample 847.9/2777. Lower Jurassic, Juhta Member, borehole 2777, depth 847.9 m. Feldspar-quartz sandstone with few quartzite and granite fragments.
SOUTH YAKUTIAN COAL BASIN
157
158
I. E. STUKALOVA
oils from 4.92% to 15.95%. The total content of methano-naphthene, aromatic and polyaromatic oils is from 12.50% to 34.13%, with a maximum of 43.51%, representing a high concentration of hydrocarbons. The extracts contain 7.38-21.90% of benzolic resins 0.41% to 1.88% of alcohol resins and 0.49-1.15% of alcohol-benzolic resins present in three samples. The concentration of high molecular weight resins and combinations are from 4.92% to 21.39%. The contents of all resins varies from 13.30% to 49.90%. The extracts also contain asphaltenes ranging from 20.34% to 69.95%. The high concentration of bitumen in the chloroform extracts of organic matter of the South Yakutian coals, is probably responsible for the high caking properties of the coals.
Petrographic composition of sandstones The petrographic composition of the sandstones in coal-bearing strata of the South Yakutian Coal Basin identified four terrigenous-mineral associations, according to the scheme proposed by Shutov (1972). The first is represented by true arkoses and feldspar-quartz sandstones with a low content of quartzite, granite and gneiss fragments (Fig. 5). It is a typical Lower Jurassic association in the Usmun and AldanoChulman regions (Figs 6 & 7). For example, sample 837.9/2777 from a depth of 837.9m in borehole 2777 in the Aldano-Chulman region, is a feldspar-quartz coarse sandstone containing 60% of quartz (Q), 20% of feldspar (F) and 20% of rock fragments (R), represented by quartzite, granite and dolomite fragments. These sandstones contain a hydromica and laumontite pore cement (Fig. 9, photo 5). Another example, sample 847.9/2777 from a depth of 847.9m in the same borehole 2777, is a feldspar-quartz medium sandstone containing 55% of quartz (Q), 20% of feldspar (F) and 25% of rock fragments (R), represented by quartzite and granite fragments. Feldspar is represented by plagioclases (P1). These sandstones contain a hydromica pore cement (Fig. 9, photo 6). The second association features greywacke arkoses, feldspar-quartz and quartz-feldspar greywackes with a high content of felsitic and quartz porphyries, granites, gneisses and rare trachytes (Fig. 5). It is characteristic of the greater part of the Middle and Upper Jurassic coal-bearing strata in three regions (Figs 6, 7, 8). For example, the Duraji Member in the Aldano-Chulman region is represented by feldspar-quartz greywackes (Fig. 9, photo 3,4).
Quartz-feldspar greywacke of sample 560.1/ 2777 from 560.1m in borehole 2777 contains 20% of quartz (Q), 40% of feldspar (F) and 40% of rock fragments (R), represented by felsitic and quartz porphyries, granites, gneisses and trachytes. These sandstones contain a hydromica and laumontite pore cement (Fig. 9, photo 3). Another example, sample 590.4/2777, is represented by feldspar-quartz greywacke from 590.4m in borehole 2777. It contains 30% of quartz (Q), 35% of feldspar (F) and 35% of rock fragments (R), represented by granites and gneisses. These sandstones contain a hydromica and quartz pore cement (Fig. 9, photo 4). The third association of true arkoses, greywackes with high content of quartz and felsitic porphyries, quartzites and granites (Fig. 4), is typical of the Upper Jurassic and Lower Cretaceous deposits in the Aldano-Chulman region (Fig. 7). For example, the greywacke medium sandstone of sample 27.5/1868 from 27.5m in borehole 1868 contains 20% of quartz (Q), 30% of feldspar (F) and 50% of rock fragments (R), represented by quartz and felsitic porphyries, quartzites and granites. The sericitization of feldspar is widespread (Fig. 9, photo 1). Another example, sample 74.5/1868 is represented by greywacke medium sandstone from 74.5m in borehole 1868 in the Aldano-Chulman region from the Lower Cretaceous, Holodnican Member. It contains 25% of quartz (Q), 25% of feldspar (F) and 50% of rock fragments (R), represented with high content of quartz and felsitic porhpyries, quartzites and granites. These sandstones contain a chloritic and laumontite pore cement. The sericitization of feldspar is widespread (Fig. 9, photo 2). The fourth association of feldpathic greywackes and greywacke with a high content of rhyolites, felsitic porphyries, trachites and their tufts (Fig. 4), is mostly encountered in the Lower Cretaceous deposits of the Tokin region (Fig. 8).
Mineral alteration Mineral alterations were investigated both by light microscopy and by X-ray diffraction analyses of clay minerals. The mineral matter within the terrigenous rocks of the South Yakutian Coal Basin has been altered within the catagenetic and metagenetic stages. The terminology of the stages and periods of lithogenesis are those proposed by Vassoevich (1962). The metagenetic stage is seen in highly altered Lower Jurassic deposits in the basin (Figs 5 & 7). In this stage of lithogenesis the blastic structures
SOUTH YAKUTIAN COAL BASIN and welded joints are widely spaced. X-ray diffraction analyses of clay minerals show that kaolinite, dickite, and mica of the 1 Md polytype are present. The catagenetic stage affects Middle and Upper Jurassic and Lower Cretaceous deposits in the basin. This stage of lithogenesis is characterized by alteration of allogenic minerals and formation of new structures. The catagenetic stage includes two subzones: the smectite-mica subzone and the laumontite subzone. X-ray diffraction analyses of clay minerals show that the rocks from the smectite-mica subzone contain up to 5-15% mixed-layer phase smectite-mica and chlorite. X-ray diffraction analyses of the clay nminerals show that in the laumontite subzone laumontite, chlorite and mixed-layer phase smectite-mica with packages up to 10-15% are present. Laumontite forms mainly in the central parts of pores or substitutes for other minerals such as plagioclase, hornblende, biotite and pyroxene (Fig. 9). Other authigenic minerals typical of the laumontite subzone are epidote, sphene and quartz. The characteristics of the three regions of the South Yakutian Coal Basin were compared. Different parts of the basin demonstrate different alteration of organic and mineral matter. Lower Jurassic deposits in the basin are in the metagenetic stage. In the Usmun region high volatile bituminous coals occur and in the Aldano-Chulman region there are low volatile bituminous coals. The middle-Upper Jurassic and Lower Cretaceous deposits of the basin are in the catagenetic stage. The coals in the A l d a n o - C h u l m a n region are low volatile bituminous rank and in the Tokin region they are high volatile bituminous coals.
Conclusions Investigation of the coals in the Usmun, Aldano-Chulman and Tokin regions of the South Yakutian Coal Basin demonstrate that humic coals consist of vitrinite (70-90%), inertinite (10-20%) and liptinite (0-10%) maceral groups. There is a high percentage of telinite macerals, up to 60-65%, in the vitrinite maceral group. According to chemical analyses the coals contain moisture (W daf) from 0.20% to 2.80%, volatile matter (V daf) ranging from 17.0% to 40.0%, ash content (A daf) between 1.57% to 25.92%, caking index (y) is 6-21 ram. Coals are of middle to high rank, with vitrinite reflactance values (R0) of 0.65-2.15%.
159
The South Yakutian coals are of high quality because of their rank and composition. The secondary bitumen macerals are probably responsible for the high caking index of the coals and their coking properties, as compared with coals of the same rank from others coalfields and basins. Sandstones in the coal-bearing formation are represented by true arkoses, greywacke arkoses, feldspar-quartz and quartz-feldspar greywackes. The mineral matter of terrigenous rocks are altered to the catagenetic stage and the metagenetic stage. Vitrinite reflectance and mineralogical parameters were used to evaluate the stages of alteration of sediments.
References BREDIHIN, I. S. 1973. South Yakutian (Aldan) Coal Basin. Geology of coalfields and shales of the USSR., Vol. 9, Nedra Publishing, Moscow, 5-117 (in Russian). FATKULIN, I. Ya., GEBLER, I. I. & RESHETKO,A. N. 1970. Statistic correlation between vitribite reflectance and quality of the coals in the AldanoChulman region. Chemistry of fuels, Vol. 3, 141-143 (in Russian). MARKOVICH, E. M. 1981. Palaeobotanic considerations of stratigraphy and correlation. South Yakutian coal-bearing formation, Leningrad, Nauka Publishing, 33-43 (in Russian). MOKR1NSKY,V. V. 1961. Metamorphism of coals in the South Yakutia. South Yakutian Coal Basin, Leningrad, Publishing of Academy of Sciences of the USSR, 382-420 (in Russian). NAZAROV,V. I. & STUKALOVA,I. E. 1991. Catagenetic alterations of the Jurassic and Lower Cretaceous deposits in South Yakutia. Geology of the coalfields, Ekaterinburg, 100-112 (in Russian). PRILUTSKY, A. M. 1979. Petrographic composition and quality of coals in the South Yakutian Coal Basin. Stratigraphy, paleoenvironment and lithology of the South Yakutian Coal Basin. Transections, VSEGEI, Vol. 306, Leningrad, Nauka Publishing, 78 82 (in Russian). PROSVIRYAKOVA,Z. P. 1961. Palaeobotanic characteristics of the coal deposits in the South Yakutia. South Yakutian coal-bearing formation, Publishing of USSR Academy of Sciences, Vol. XI, pp. 122 175 (in Russian). PUTTMANN, W., WOLF, M. & WOLFF-FISCHER, E. 1985. Chemical characteristics of liptinite macerals in humic and sapropelic coals'. Advances in Organic Geochemistry, Vol. 10, 625-632. STACH, E., MACKOWSKY,M. Th., TEICHMULLER,M., TAYLOR, G. H., CHANDRA, D. & TEICHMULLER, R. 1982. Stach's Textbook of Coal Petrology'. 3rd edn. Gebruder Borntraeger, Berlin. SHUTOV, V. D. 1972. Classification of the terrigenous rocks and greywackes. Greywackes, Nauka Publishing, Moscow, 9-29 (in Russian).
160
I. E. S T U K A L O V A
TEICHMULLER, M. T. 1974. Generation of petroleum like substances in coal seams as seen under the microscope. In: TISSOT, B. & BIENNER, F. (eds) Advances in Organic Geochemistry, Paris, 321-348. - - 1 9 9 0 . The genesis of coal from the viewpoint of coal geology. International Journal of Coal Geology, 16, 121-124. TERENTYEV, E. V. 1978. Tectonics of the coalfields of the USSR. Geology of coalfields and shales of the USSR, 12, Nedra Publishing, Moscow, 94-162 (in Russian). VASSEOVICH, N. B. 1962. About terminology for the stages and periods of lithogenesis. Transactions, VNIGRI, 190, Leningrad, 220-230 (in Russian).
VLASSOV, V. M. 1981. Usmun, Tokin and Gonam reg(ons in the South Yakutian Coal Basin. South Yakutian Coal Formation, Leningrad, Nedra Publishing, 24-32 (in Russian). WALTZ, I. I. 1961. Petrographic composition and structure of coal seams in South Yakutia. South Yakutian Coal Basin, Leningrad, Publishing of Academy of Sciences of the USSR, 176-277 (in Russian). ZHELINSKY, V. M. 1980. Mesozoic coal-bearing formation of South Yakutia. Novosibirsk, Nauka Publishing (in Russian).
Coal rank variations with depth related to major thrust detachments in the South Wales coalfield: implications for fluid flow and mineralization ROD
GAYER,
RICHARD
FOWLER
& GARETH
DAVIES
Department of Earth Sciences, University of Wales Cardiff, PO Box 914, Cardiff, CF1 3YE, UK Abstract: Coal maturity data in the form of volatile matter (daf and dmmf) and random vitrinite reflectance have been analysed for the South Wales coalfield. They show that in general coals increase in rank with depth, obeying Hilt's law, and increase in rank laterally from high volatile bituminous coals in the south and east of the coalfield to anthracite in the northwest of the coalfield. The lateral increase in rank does not coincide with the basin depocentre which was located to the southwest of the coal basin during Westphalian times. The rank pattern with depth in the Westphalian A-Lower Westphalian C Coal Measures of the eastern half of the coalfield suggests a palaeogeothermal gradient of approx. 310~ -1, equivalent to a basal heatflow of 295mWm -2. Investigation of vitrinite reflectance in a coal sequence repeated by intense Variscan thrusting indicates that coal rank was acquired both pre- and syn-thrusting. Detailed analysis of the volatile matter data reveals the presence of excursions from Hilt's law present in one or more coal seams close to the boundary between Westphalian A & B. Of the 154 data sets analysed from the coalfield, 94 (61%) show one or more excursion. It is shown that the excursions correlate with thrust detachments within the coal seams, and it is argued that the excursions represent an increase in maturity temperature caused by fluids carrying heat into the coal seam along the seismically active thrusts. The fluids may also have been responsible for the carbonate, oxide and sulphide mineralization of the coalfield. Preliminary comparisons with the Ruhr coal basin in Germany suggest that future studies involving computer generated thermal models are required to understand the thermal evolution of both basins. The South Wales coalfield represents a major Late Carboniferous coal basin developed on the Variscan foreland. Mining in the coalfield has long recognised the presence of coals ranging in rank from high volatile bituminous coals in the south and east of the coalfield, through intermediate ranks, into anthracite in the northwest of the coalfield. Attempts to explain this rapid rank variation, laterally over c. 50km, have ranged from those invoking differences in the original coal-forming plant communities or depositional environments (Strahan & Pollard 1915; MacKenzie-Taylor 1926; Fuchs 1946), through those resulting from differential burial (Jones 1949; Wellman 1950; White 1991) to those associated with differing heat flows brought about by magmatic heating (Firth 1971), frictional heating along thrusts (Trotter 1948, 1950, 1954), differing basement regimes (Gill et al. 1979), and effects of hot fluids (Davies & Bloxam 1974; Gayer et al. 1991; Austin & Burnett 1994). Most of these have been discussed in detail by White (1991) and by Austin & Burnett (1994), but no convincing proof of the process producing the rank variation has been forthcoming. This paper presents some additional data on the age of rank development relative to Variscan thrusting and the presence of variations in
maturity with depth in the coalfield. By converting maturity indices to temperature, using the formula devised by Barker & Goldstein (1990), values of palaeo-geothermal gradients both for specific stratigraphic intervals within the Coal Measures succession and for different localities within the coalfield are derived. It is argued that the magnitudes and variations of these geothermal gradients are difficult to reconcile with a burial model alone and that the presence of excursions in maturity values coincident with thrust detachments in the coals suggests a causal link, possibly associated with the flow of hot mineralizing fluids guided by thrusts.
Regional geology of the South Wales coalfield Stratigraphy The sediments of the South Wales coal basin are preserved in a structurally complex E - W trending Variscan synclinorium extending from SW Dyfed to the western flank of the Usk antiform (Fig. 1). The coal basin overlies a southward thickening (0-1 km) Lower Carboniferous platform carbonate sequence (Wilson et al. 1987)
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 161-178.
R. GAYER ET AL.
162
+
+
+ lI
./71 !
+
/ /
§
0
km
+\
,\*
10
//
/
+
+
S
,.;'-CCA v + \+
+
+
Upper Pennant Measures
I i q
1 LowerPennant Measures
~
Faults
Lower & Middle Coal Measures
" ~...
Antitorm
Pre- and Post-Coal Measures
"-4-
Synform
Fig. 1. Map of South Wales coalfield, showing major structures and generalized stratigraphy. BTS, Betws-Tonyrefail Synform; CCA, Cardiff-Cowbridge Antiform; CCD, Careg-Cennen Disturbance; GS, Gelligaer Synform; LCS, Llantwit-Caerphilly Synform; LLD, Llanon Disturbance; MA, Maesteg Antiform; MGF, Moel Gilau Fault; ND, Neath Disturbance; PA, Pontypridd Antiform; SVD, Swansea Valley Disturbance; TD, Trimsaran Disturbance; UA, Usk Antiform; 1, Ffos Las OCCS; 2, Treforgan No. 2 borehole; 3, Treforgan No. 3 borehole; 4, Park Slip OCCS; 5, Ffaldau Colliery; 6, Park Colliery; 7, Ffyndaff OCCS; 8, Maerdy Colliery; 9, Llanharan Colliery; 10, Coedely Colliery; 11, Lewis Merthyr Colliery; 12, Cwm No. 4 shaft; 13, Lady Windsor Colliery; 14, Merthyr Vale Colliery; 15, Windsor Colliery; 16, Nantgarw Colliery; 17, Penalta Colliery, 18, Brittania Colliery; 19, Bedwas Colliery; 20, Oakdale Colliery; 21, Nine Mile Point Colliery; 22, Celynen North Colliery; 23, Celynen South Colliery; 24, Blaenserchan No. 2 shaft and Underground Borehole. that passes conformably downwards into a thick (3 km) Old Red Sandstone unit and a shallow marine Lower Palaeozoic succession. Crystalline basement underlies the coal basin at depths ranging from 3.5km in the northwest of the main coalfield to over 6 km in the east (Hillier 1989). The Coal Measures sequence is up to 3.5 km thick in the centre of the basin and ranges in age from basal Namurian to early Stephanian, covering a time span of some 21 Ma (according the timescale of Lippolt et al. (1984)). The basin was initiated during the early Namurian, following a regional compressional event that resulted in the breakup of the Dinantian carbonate platform, relocation of the basin depocentre and the influx of clastic detritus (Hartley & Warr 1990). The basin has been interpreted as a Late Carboniferous foreland basin at the northern margin of the Variscan orogenic belt (Kelling 1988; Gayer & Jones 1989) thought to have been formed by lithospheric downflexure as a
response to a Variscan tectonic load to the south in SW England. The basin was filled by sediment derived both from the north but mainly from the erosion of the tectonic load to the south (Kelling 1988, Jones 1989a, Hartley & Warr 1990). Throughout Silesian sedimentation the basin depocentre was oriented approximately E-W (varying from NE-SW to NW-SE) and centred on the Swansea-Gower area (Fig. 2). Stratigraphical thicknesses decrease markedly away from the depocentre, particularly to the east but more gradually to the north and west (Hartley 1993). The basin-fill sequence coarsens and shallows upwards from marine mudstones and sandstones (Namurian A-lower Westphalian A), through coastal plain coalbearing mudstones and sandstones (upper Westphalian A-lower Westphalian C) to coarse grained sandstones and conglomerates deposited in an alluvial braidplain (upper Westphalian CStephanian) (Fig. 3, Jones 1989b, Hartley 1993).
COALIFICATION EXCURSIONS AND T H R U S T - G U I D E D FLUIDS
2,5
a. Shale Gp., Namurian R2-G1
I
75
b. Upper Westphalian A
~
-- _
_
~
..~.~
~
~
10
dlf 120 11
c. Lower Westphalian B
~ ~ _
~
:30
~,,,f~-
=1oo
d. Brithdir Beds, Upper Westphalian C Fig. 2. Isopach maps for four time slices to show
depocentrc in south of the coalfield.
Contours in metres.
163
164
E T AL.
R. GAYER
NN
< Z
< -r
Mynyddis lyn
114
~.-_-.-_-.-.-.-,
Swansea
U.I
Three F e e t
m
Hughes
m
if?
----.-~ ~
-.-.-.-.-.,
Z
Brithdir
<
98
Captain
i
._1 < -r a.
N
No2
Upper
1.1.1
.... .....
Z
Cefn
< Z
~-~
<
<
Z <
-r ~_
Z
~o ua
m
<
o
42 Stephanian 2.5
westphalian D
Five
Feet
Garw
I0
13
1.5
wes tphalian
......~
(3
~
2
M.B.
Yard Seven F e e t 2 6
i
~"
(km)
3
Feet Nine Four F e e t 4 0 S i x F e e t 37 C a e r a u 36 N i n e F e e t 31 Bute 30 Amman Rider
i
SCALE
60
Two
n
M.B.
65
Coed M.B.
Pentre
< n,' m
91
Cwmgorse
Abergorky
i
Z
Rhondda
"
C
1
Gastrioceras Westphalian i s u b c r e n a t u m M.B. A & B i
sandstone
0.5 [ NaRurian
............
marine
band
dominantly i
mudstone
]
coal
Fig. 3. Generalized stratigraphical column (not to scale) of the South Wales Coalfield, showing the main sandstone units in the lower part of the Lower Coal Measures and in the Upper (Pennant) Coal measures, and the principal coal seams with their seam numbers, as used in the text. The seam numbers follow an unpublished report by Mr Robin Thewlis, formerly of British Coal Opencast. The inset (lower right) shows the stratigraphical column to scale. The majority of the 125 coal seams present in the coalfield occur in the productive Coal Measures of late Westphalian A-Westphalian B age. Although no younger solid formations overlie the Upper Carboniferous Coal Measures, to the south of the coal basin, in the Vale of Glamorgan, a relatively thin (<300 m) Mesozoic sequence (Upper Triassic-Lower Jurassic) rests unconformably on Variscan deformed Dinantian-Namurian rocks. Vitrinite reflectance values of 0.51%Rm in the Lower Jurassic rocks (Cornford 1986) suggest a possible younger Mesozoic cover subsequently eroded.
Structure The basin-fill has been affected by Variscan deformation, the principal elements of which are: (i) approx. E - W trending basin-scale northverging folds; (ii) strike-parallel thrusts (and associated lag faults), with southward directed transport along the southern margin of the coalfield, but northwards in the centre and north of the coalfield; (iii) E N E - W S W trending zones of fold and thrust disturbance, thought to represent reactivated basement Caledonoid structures as Variscan thrust ramps (Jones
COALIFICATION EXCURSIONS AND THRUST-GUIDED FLUIDS 1991; Brooks et al. 1994); and (iv) NNW-SSE striking cross faults, commonly showing evidence of early strike-slip movement and later normal dip-slip movement (Fig. 1, Cole et al. 1991). All of these structures have been interpreted as the effects of a northwards propagation of Variscan fold and thrust deformation into the coal-bearing foreland basin (Gayer et al. 1991). Jones (1989a, b) showed that sedimentation throughout the Westphalian was affected by incipient development of the main E - W folds in the coalfield, demonstrating the close timing between basin subsidence and compressive deformation. Of particular relevance to this paper are the thrusts, which have been analysed in some detail within the working opencast coal mines (e.g. Gayer et al. 1995). This analysis has shown that thrusts occur either as major detachments
165
along coal seams with imbricate thrusts branching upwards into the hangingwall sequence, or as isolated thrust ramps which in some cases are linked downwards to the thrust detachments in coal seams (Jones 1991; Hathaway & Gayer 1994). Thrust detachments commonly occur within the coals of the lower part of the productive Coal Measures, so that at Park Slip opencast coal site (OCCS) (Fig. 1), on the southern margin of the coalfield and at Ffyndaff OCCS (Fig. 1), on the northern margin, detachments occur in the Nine Feet group (seams 30-34) and Six Feet group (seams 36-39) of seams (Cole et al. 1991; Jones 1991 and Fig. 4). At Ffos Las OCCS (Fig. 1) detachments occur in the Big (Four Feet seam 40) Kings (Two Feet Nine seam 42) and Green (Upper Two Feet Nine seam 43) (Frodsham et al. 1993). Where thrusting has been intense, e.g. at Ffos Las OCCS
(a)
S
N Thrust 1 ----.-"--
Driver Seam Upper
6' Seam
f
"
Bute Seam \ Thrust 2
~
#
Red S e a m
_
Thrust 3
~
"
Red S e a m
- - i ~
J
h
1
i
~ Thrust 4
,
~
__~~---___~"
~
,~,,v//____~,.
Antithetic 0
100m
(b)
S
N
I
t
/
/
/.
0
m
100
i
Fig. 4. Cross sections through the productive Coal Measures to show the style of thrust structure in the South Wales coalfield. (a) Section drawn parallel to the thrust transport direction through Ffyndaff OCCS, based on coal extraction sections. Note the thrust detachments through the Nine feet seam (seam 31) and Red seam (seam 36). Modified from Gayer et al. (1994). (b) Section drawn from British Coal borehole data through Park Slip West OCCS, showing thrust detachments at the level of the Six Feet seam (seam 37). See Fig. 1 for site locations.
166
R. GAYER E T AL. during thrusting, as the coal seams matured by compaction and dewatering (Gayer et al. 1991; Gayer 1993).
in the northwest of the coalfield, with thrust shortening estimated at 70% (Cole et al. 1991), it has been demonstrated that detachments within several coal seams have moved simultaneously producing a style of deformation that appears to be unique to coal-bearing sequences. This has been termed Progressive Easy Slip Thrusting (PEST) by Frodsham et al. (1993), and explained in terms of fluid overpressuring
~
~
.
,
~
~
a. No.2 RhonddaSeam
Metamorph&m
The sediments of the coalfield have been affected by very low grade metamorphism within the
~
23
~
'
~
7
27
~
C__~
1
1
31
b. Fou i ~ . , ~
~
19 ,,~ 23 27
1
11
15
31
Fig. 5. Maps of coal rank (%Vm) for the South Wales coalfield, after White (1991). (a) Five feet seam (Lower Coal Measures); (b) Four Feet seam (Middle Coal Measures); (c) Rhondda No 2 seam (Upper Coal Measures). Note the location of the anthracite area (less than 9% Vm) in the northwest of the coalfield and compare with the Coal Measure isopach maps (Fig. 2).
COALIFICATION EXCURSIONS AND THRUST-GUIDED FLUIDS
coalfield structure; in particular she showed that the coal isovols were folded by the coalfield synform and also by the Pontypridd antiform. The isovols do not however parallel stratigraphical boundaries precisely, cutting gently down stratigraphy to the southeast. This may reflect the thinner stratigraphy in this direction causing a particular horizon to be less deeply buried and therefore of lower rank in the thinner succession in the southeast. It might also reflect a syndeformational age for rank development with the rocks on the southern flank of the coalfield being uptilted before the imposition of rank. To test the possibility of syn-deformational rank development and specifically to determine the timing of rank formation relative to thrusting, a study of the vitrinite reflectance of coals repeated by several thrusts in the Ffos Las OCCS has been undertaken. The structural geology in the Ffos Las OCCS has recently been described by Frodsham et al. (1993). The site lies across the Llanon and Trimsaron disturbances which are represented by four major north directed thrusts each with displacements greater than 100m, and many meso- and minor-scale thrusts (Fig. 6). The combined throw of the thrust-related structure is 1175 m and thrust shortening has been estimated as 70%. Coal samples were collected from the Big (seam 40), Kings (seam 42), Green (seam 43) and Graigog (seam 44) coal seams in the different thrust sheets exposed by the opencast workings at the time of the study. The results of the vitrinite reflectance analyses are shown in Fig. 7 where Rm has been plotted against the
diagenetic grades of metamorphism, based on illite crystallinity characteristics (White 1991). The metamorphism ranges into the lower anchizone facies in the extreme northwest of the coalfield, based on the presence of pyrophyllite (Bevins et al. 1996). Coal rank, based on volatile matter contents (%Vm dmmf) and vitrinite reflectance (%Rm) , increases from high volatile bituminous coal in the south and east of the coalfield to anthracite in the northwest of the coalfield (Fig. 5, White 1991), coinciding with the illite crystallinity metamorphic pattern. White (1991) demonstrated that the coal isovols (lines of equal volatile matter) are parallel to stratigraphical boundaries around major fold structures in the coalfield, suggesting a preVariscan folding origin for the coal rank development. The rapid lateral increase in coal rank has been discussed in detail by White (1991) and by Austin & Burnett (1994), who suggested that burial beneath a now eroded sedimentary load (White 1992) or inflow of hot fluids either from the Variscan mountain belt to the south (Gayer et al. 1991) or along deeply penetrating faults in the underlying basement (Austin & Burnett 1994) are the mechanisms in most agreement with the observations.
Local coal rank development and its relationship to thrusting White (1991), working on a coalfield scale, showed that the coal rank in the coalfield was developed before the formation of the major
N
167
Llannon Disturbance
Trimsaran Disurbance
S
.... 7- _ _ : : " ) .
"~-~
~
~..-~
~ .
Key:
"--.. - - . .
-~
"" ~ Graigog Rider, Graigog, Green and Kings seams
100 metre
Big Seam .........
Other clastic rocks Superficial deposits Thrusts
Fig. 6. Cross-section of structure at Ffos Las OCCS based on borehole data (after Frodsham et al. 1993).
168
R.
E T AL.
GAYER
2 9, ~ .w.,+ ,,.-,
"
I
++
'
+I
F
I
i
-+| ~++
I
+ '
I ,
++_ . _ ~
~~
~N,
....-- ..---""--+ ' ' -
~_-~
rj I-, ..--.. 0 IW
t-+
k+
m..
j
o.,
,,r:::::J
,
ol
~,..:u
I(
#....~----
+ +~ooo++++o+o++++ ++++- +++++++---. " "+++++++ " ++++++++++
++++++++
+.m
++-"++
2'= o
o>
+::I
q::I
..,_+
=
l ,+...
o
J
.~ u,,.B
+~L.+
Ak
\
uO
i
r
~
._...._ ~
,_..-~ " - ' ~
f '.'-
0
J
'" +~ mm ' " <
m+
:~ '''
'''
'''
'''
'''
-+u(..+u++++++++ . . . . . .
m 0
"m--; <,.,
I
J '"
b
r',
c~
_.+_+++
o
0
(O
+++
(.b (.b (.b 0 0
b
C)
0
~ + + 9
~
C)
(..+
s~ sz 9+ ~
c.)
COALIFICATION EXCURSIONS AND THRUST-GUIDED FLUIDS coal seams arranged (a) stratigraphically and (b) tectonically, in ascending thrust sheets. There is no clear decrease in Rm with increased stratigraphical height, suggesting that the rank was not developed pre-thrusting, but there is no clear decrease in R m with higher tectonic position suggesting that the rank was not developed post-thrusting (Fig. 7). There is a general decline in Rm upwards with stratigraphy in each thrust sheet but with anomalous values in the vicinity of thrusts. The R m pattern at Ffos Las OCCS is therefore complex and can only be explained if the vitrinite reflectance was developed both before and during thrusting.
Variations in coal rank with depth
Establishment of the local palaeogeothermal gradient White (1991) demonstrated that the volatile matter (Vm dmmf) decreases constantly with depth across the coalfield, and therefore suggested that a constant geothermal gradient operated during coalification, without specifying a value for this gradient. Barker & Goldstein (1990) have developed an empirical equation relating R m to temperature: T(~
= [(In Rm) + 1.26]/0.0081.
This relationship can be used to reconstruct the local palaeogeothermal gradient in the South Wales coalfield. Values of vitrinite reflectance (Rm) from 12 collieries in the east of the coalfield (Fig. 1) have been plotted against the local depth below seam 91, as recorded in the Btaenserchan borehole (Fig. 8a). Figure 8b shows Vm (dmmf) from the same collieries and the same seams plotted against local depths beneath seam 91. The two plots show very similar, but reversed linear trends which suggests a linear increase in rank with depth. Using Barker & Goldstein's (1990) relationship, the variation in temperature with depth below seam 91 is plotted in Fig. 8c which gives a reconstructed palaeogeothermal gradient for the eastern part of the coalfield of approx. 310~ -1. However, these plots are based on the assumption that seam 91, close to the top of the Coal Measures, was isothermal, (i.e. reached the same maximum temperature, and thus the same values of R m and Vm) across the relevant area of the coalfield. In fact Vm for seam 91 varies from 28% to 33%, an estimated temperature difference of 18~ which is a likely explanation for the scatter of data points in the plots.
169
The palaeogeotherm of 310 ~ km-1 is approximately four times that of 5075~ -l determined by Alderton & Bevins (1996) from fluid inclusions within quartz crystals grown in an ironstone nodule within the Middle Coal Measures in the central eastern part of the coalfield. Alderton & Bevins (1996) assumed a constant geothermal gradient between the sample position and the palaeosurface. However, our plots of temperature against depth suggest that the geothermal gradient was much higher within the mud and coal dominated succession of the Middle Coal Measures and became appreciably less within the sand dominated Upper Coal Measures. Although there are insufficient vitrinite reflectance data from the northwest of the coalfield to produce similar plots against depth, the higher temperatures required for the generation of anthracite with Rm values of up to 4% (327~ within an only marginally thicker Middle Coal Measures sequence suggest even higher geothermal gradients may have operated during the rank development in this part of the coalfield.
Excursions from Hilt's law in the South Wales Coal Measures The detailed variation of rank (and therefore temperature) with depth can be studied using the coal rank determinations from collieries, shafts and boreholes throughout the coalfield. The vast majority of recorded coal maturity data for the coalfield is in the form of volatile matter (dmmf or daf). This was the standard proximate analysis carried out by the former British Coal Corporation and, their predecessors, the National Coal Board. Within the last ten years vitrinite reflectance measurements were also occasionally recorded, but these represent only a small fraction of the total data set. Since almost all the mines are now closed and relatively few coal samples are available, it is not possible to carry out new vitrinite reflectance analyses. Thus volatile matter content rather than vitrinite reflectance has been used for the analyses. The plots of Fig. 8a & b suggest a close inverse relationship between Vm (dmmf) and Rm, which has been extensively analysed by McCartney & Teichmialler (1972), Bartenstein & Teichmiiller (1974), and Teichmfiller & Teichmfiller (1982). Using the Barker & Goldstein (1990) equation, it has been possible to relate both to temperature (Fig. 8c). The data used were collated from British Coal archives (White 1992) and from the Coal Survey
> 9r
///r
9 9 /
/
/
/
/-" /-
/
/
J
e~ >
/
o
./
./
g
~
g (w) '16 m ~ ~o|oqqldoG
04,
>
@
9 0
~m
(tu) 'I 6 ure~ ~OlOq 41d~G
4
~o
4
(m) 'I6 u*~s ,~opq 41d~(l
COALIFICATION EXCURSIONS AND THRUST-GUIDED FLUIDS Seam Records of the National Coal Board. Data sets were collected from 154 collieries, shafts and boreholes in which Vm values from five or more seams were available. For these localities Vm values were plotted against depth. In general a regular decrease in Vm occurs with increasing depth, indicating an increase in rank with depth and that Hilt's law is obeyed. However, 94 data sets show an excursion from Hilt's law. Figure 9 shows 12 plots of variations of Vm (dmmf) with depth (after White 1992), representing the range of data available from locations across the coalfield. Three main types of pattern can be seen: (i) Vm decreases regularly with depth throughout, with no excursion (e.g. Treforgan No. ! borehole & Lady Windsor colliery); (ii) a single excursion occurs to lower Vm content, shown by one or more seams (e.g. Cwm colliery); and (iii) two or more excursions to lower Vm content occur (e.g. Coedely colliery). The excursion in Vm is mirrored by similar excursions in the volatile elements Sulphur and Phosphorous (Fig. 10). Where an excursion occurs the regular pattern of decreasing Vm with depth is perturbed, with an increase in Vm occurring with depth above the excursion (e.g. Nantgarw colliery). The geothermal gradient, calculated by converting %Vm to %Rm, using the graphs in Stach et al. (1982), and thence to temperature, using the Barker & Goldstein (1990) equation, is also highly variable with values ranging from 8-15~ -~ at Nantgarw colliery to 133-193~ k m -1 at nearby Cwm colliery (Fig. 9). The excursions from Hilt's law most commonly occur centred on seam 37 (48% of excursions) but ranges from seam 80 (< 1%) to seam 13 (2%). The excursion is thus not seam specific and hence is unlikely to be related to the original coal composition. There does, however, appear to be a strong correlation between the stratigraphical level of the excursion and the presence of thrusting within the seam. Although it is difficult to obtain information on the presence of in-seam thrusting from the abandoned mine records, they can be directly observed in the working opencast coal mines. Here thrust detachments in coal seams commonly produce a pervasive, oblique, sigmoidally shaped fabric that has been described as a cleavage duplex (Frodsham et al. 1993; Gayer !993). The detachments also develop true duplexes with roof thrusts immediately above the seam roof and floor thrusts in the seat earth below the seam. The duplexes interleave seat earth and carbonaceous roof rock into the seam and can be recognised in boreholes and shafts as 'rashings', a miners' term for this type of
171
structure (Woodland & Evans 1964). Thrust detachments commonly occur in seams 30-43, with the greatest incidence in the Nine Feet group (seams 30-34) and the Six Feet group (seams 36-39) in the east of the coalfield and in seam 40 in the west of the coalfield. At Llanharan colliery, in the hangingwall of the major Llanharan thrust (Woodland & Evans 1964), two excursions from Hilt's law are recorded in the Vm (daf) data at the level of seams 37 and 27. These are both associated with in-seam thrust detachments revealed by rashings, which also occur within or adjacent to coals 40, 36, 32, 30 and 13 (Fig. lla). In the New Shaft at Park colliery, rashings occur associated with coals between seams 31 and 36 with a Vm (dmmf) excursion present at seam 36. The Cockshot Rock, a persistent fluvial sandstone, lies 45m beneath the excursion at seam 36 (Fig. 11b) In this latter case, the absence of Vm values for the seams between 31 & 36, in which the rashings are present and immediately beneath which the permeable Cockshot rock occurs, means that the precise location and extent of the excursion cannot be determined precisely.
Discussion
Palaeogeothermal gradients and heat flow in the South Wales coalfield One of the outstanding problems of the South Wales coal basin is the explanation for the major lateral rank variation, with seams at the base of the Coal Measures varying from a Rm of 1.0% in the southeast of the coalfield to 4.0% in the northwest of the coalfield (White 1991). This represents a variation of maximum maturity temperature from approx. 150~ in the southeast to 325~ in the northwest. It is difficult to explain this variation by burial depths alone, since the depocentre for the preserved Coal Measures succession lies to the south of the area of highest rank (compare Figs 2 & 5). It would require a considerable thickness of younger sediments deposited before the onset of Variscan deformation to develop the required burial temperatures (e.g. White 1992). Any such sediments have since been completely eroded. Calculation of the required burial depth to produce the maximum maturity temperatures in the northwest of the coalfield is not simple as it depends on a number of undefined variables that include the magnitude of the palaeo-heat flow during the Late Carboniferous, and the thermal conductivity of the now-eroded sedimentary
(a) -250
Manly 4237~40
Nhte MilePeat
-150 ,,~ -200 ~ 91
36 ~ 3 1
~-~ -300
~ "250 t 4 ~ _ - 3 0 0 -350 1-
-351)
f
9
10
-400
I
11 12 Volatile% dmmf
Lady WhMsor
-450
~
I
-5oo[
40
~
13.1/~/ 26 --
-650
I
28
28.5
13
14 15 Volatile% dmmf
I
I
I
29 29.5 30 Volatile% dmmf
-150
30.5
37
-200 -250
26~,,,,,/30 , 17 19 21 Volatile% dmmf
13~
16
Merthyr Vale
23
Lewis Merthyr
-400 -420t 37~
-300 ~ 3 0
-460 -480
I
Ffaidaa
-300 12
"-e-A"2
--13#r26
!
-100
~
-550 -600
31
40----"-'-'-~"~31
-350 t 306 ~
26
13
-500
-450
I
12
12.5 13 Volatile% dmmf
13.5
I
16
I
16.5
I
t
I
17 17.5 18 Volatile% dmmf
19
18.5
(b) Treforl~m2
~
"~-25o
40
3 / _ _ ~ 6
~-300
-310 -330
-350 9 4
t
t
4.5
t
6.5
7
~
-600"550 I -650 -700 26
C-
t
26
26.5
]"
31 i
t
i
27 27.5 28 Volatile% dmmf
D
-800
28
29
..-..----"~31 W" t
30 31 Volatile% dmmf
t
I
6.2
-700
37 t
23
36 t
25
P
27 Volatile% dmmf
29
Coedely
-300i -
30
-650 -250
65
37 ,
29
I
5.9 6 6.1 Volatile% dmmf
Cwm 4
-850
28.5
Naatgarw
-700
~
2 37
-750 13
~
5.8
-550 91~4
t
32
42
37
~ 31
-350 5.7
-
t
5 5.5 6 Volatile% dmmf ][~dwm
400
Treforf,~ 1
-200
-250 -230 t 3 6 ~ "~ -270
-511t)
-650 20
3 3 7 ~ t
31 I
25 30 Volatile% dmmf
6.3
COALIFICATION EXCURSIONS AND THRUST-GUIDED FLUIDS cover. Whatever values might finally be reached for these variables and the thickness of the cover, it would be necessary for the palaeogeothermal gradient imposed by the model to match the palaeogeothermal gradients measured in the various preserved stratigraphical successions. The geothermal gradient indicated in this study of 310~ -1 for the preserved Lower and Middle Coal Measures in the eastern half of the South Wales coalfield, and the assumed even higher gradient in the anthracite zone to the northwest imply high levels of heat flow, as indicated by the following calculation: heat flow (m Wm -2) = geothermal gradient (~ km-1) x thermal conductivity (Wm -1 ~ -l) Thus, with an average thermal conductivity of Coal Measures of 0.9Wm -l ~ -l, a palaeogeothermal gradient of 310~ km -a implies that the South Wales Late Carboniferous heat flow was 295 m Wm -2. This level of heat flow is comparable with values associated with oceanic spreading ridges (Sclater et al. 1980) and is completely unrealistic for continental foreland basins. Preliminary attempts to model the situation in the South Wales coalfield using the commercial basin maturity computer software package BasinMod (Platte Rivers Associates 1996) suggest that such gradients are unlikely to have been generated by burial alone, and it seems likely that the normal continental foreland basal heatflow was enhanced by heat carried into the basin by transient hydrothermal fluid flow.
Role of fluids in the South Wales coalfield The recognition of excursions from Hilt's law in the Lower and Middle Coal Measures of the South Wales coalfield provides strong evidence for localised increase in temperature at specific stratigraphical levels. This is precisely the form of thermal depth profile to be expected where heat has been transferred laterally into the basin along a permeable conduit (Duddy et al. 1994). The temperature would have been raised to
173
produce higher than normal values immediately along and above the conduit, and higher than normal thermal gradients in the overlying sequence. Below the conduit temperatures would have reverted to values associated with the normal heat flow in a normal thermal gradient. Conodont CAI values for the Dinantian limestones beneath the foreland basin show broadly equivalent thermal conditions as for the overlying Coal Measures, but with one major exception. This is in the southwest of the basin in the Swansea and Gower Peninsular area, where very high values of CAI have been recorded (Austin & Burnett 1994) in an area where the overlying Coal Measures show their lowest values of maturity. It is also broadly the site of the foreland basin depocentre (cf. Fig. 2). The origin of this relationship is unclear, but it seems unlikely to have developed from normal burial maturation and may imply the passage of fluids through the Dinantian limestones. The correlation of excursions with in-seam thrust detachments suggests that the thrusts were fluid pathways. Evidence from the Caribbean accretionary prism indicates that fluids are expelled along the basal thrust detachment of the accretionary wedge (Bangs et al. 1990). Similarly, thrust detachments associated with the destructive continental margin west of Vancouver Island channel fluids and produce abnormal heat flows (Westbrook et al. 1993). In recent years there has been a growing recognition that thrusts are able to guide fluids into sedimentary basins (Lawrence & Cornford 1995), and indeed the seismic action of the thrusts may well have increased the rate of fluid flow by seismic pumping (Sibson 1994). In the case of the thrusts within the South Wales coalfield it is clear that the strains associated with thrust deformation has produced extensional and contractional fracture systems that have allowed the volume surrounding the thrust to become permeable and thus to allow fluid ingress (Hathaway & Gayer 1996).
Mineralization in the South Wales coalfield The fracture systems in the coals of the South Wales coalfield are extensively mineralized and more than 50 different mineral species have so far
Fig. 9. Plots ofVm% (dmmf) against depth for 12 sites in the South Wales coalfield, located in Fig. 1 to show the three main types of variation. See text for explanation. Calculated palaeogeothermal gradients between specific seams are shown for seven data sets as follows: Coedely, 117-140~ -l (seams 91-13); Cwm 4, 133193~ km -l (seams 42-13); Nantgarw, 008-016~ km -1 (seams 91-31); Bedwas, 035-053~ km -1 (seams 91-13); Mardy, 056-079~ km -1 (seams 42-13); Lady Windsor, 056~ km-1 (seams 40-13); Faldau, t33-164~ -1 (seams 42-13).
174
R. GAYER E T A L . a)Volatile M a t t e r
-200 -250
91
-300 8O
-350
-400
65
4
6O
-450 -500 .
36
_
-550 -600 -650 20
22
26
24
28
I
t
30
32
34
Volatile % d m m f
b)Phosphorous and Sulphur
-200
-250
91
___...--4
-300
-350
-400 J
-450
-500
42
4o
37
-550
9
-600
26
9
!
-650 0
0.5
1
1.5
2
2.5
--
,,, I 3
3.5
4
% Element
Fig. 10. Vm, sulphur and phosphorous variations with depth at Coedely colliery (located in Fig. 1). (a) Vm% (dmmf), lines represent +1 standard deviation from the mean, and show two excursions from Hilt's Law. (b) Variations in S (squares) and P (circles) showing similar excursions to those in (a). been identified (Gayer & Rickard 1994). These include an early carbonate and oxide phase consisting of Ca, Mg, Fe and Mn carbonates, Ba and Ca sulphates, clays (kaolinite, various illites and mixed layer clays), and quartz, followed by a base metal sulphide and selenide phase consisting of Fe, Co, Ni, Cu, Zn, Mo, Cd and lead sulphides and clausthalite. Gold has
also been discovered in the coal cleat system in some coals, associated with the late stage of the earlier mineral paragenesis (Gayer & Rickard 1994). It has been argued by Gayer et al. 1991 that this mineralization was a result of fluid movements along thrusts carrying exotic ions into the basin and leaching elements from the compacting sediments.
(a)
Lianharan Colliery
-50 -100 Oe
-150
repeat
37
Llanharan Thrust
,.-.,
g
-200
-250 42 -300
37
-
4~_0
it
A36 3 1 G ~
-350
O/~
Red
30
A Amman
~
Yard.
26
-400
13
-450 32
=
~.
!
t
,
I
i
:
33
34
35
36
37
38
Volatile % daf
(b)
Park
42 -300 4O 37
-400
26
-50C 12
t 13
t 14
I 15
16
Volatile % dmmf Fig. 11. Plot of Vm against depth to show relationship between excursions from Hilt's law and in-seam thrust detachments revealed by rashings bands, and the C o c k s h o t Rock sandstone; (a) L l a n h a r a n colliery, lines represent +1 standard deviation from the mean; (b) Park colliery (located in Fig. 1).
176
R. GAYER ET AL.
Comparison with the Ruhr coal basin It is interesting to compare the situation in the South Wales coal basin with that in the Ruhr coal basin, another coal-bearing foreland basin along the northern Variscan margin (Gayer et al. 1993). Computer generated thermal models for 11 localities in the Ruhr coal basin of Germany have suggested that the Ruhr Coal Measures were buried beneath an additional 2.2-3.5 km of younger Carboniferous sediments that were completely eroded before the deposition of the Mesozoic cover (Littke et al. 1994). In order to achieve a match between calculated and observed geothermal gradients the Late Carboniferous heat flow was calculated to have been between 64 and 83 m Wm -2, very high values for downflexed continental crust in an orogenic foreland which in modern situations have low heat flow values with average geothermal gradients of 22~ km -1 to 24~ -l (Allen & Allen 1990). In the Ruhr coal basin the modelled average palaeogeothermal gradient was between 36-47~ -~ but the observed gradient in the preserved Coal Measures succession is 63-65~ km -1, reflecting the lower thermal conductivity of the Coal Measures (and the assumed higher thermal conductivity of the now eroded cover). The above suggests that the Ruhr basin experienced high Carboniferous values of heat flow and a thick Late Carboniferous cover, subsequently eroded. This is in contrast to the South Wales basin where locally high heat flow values appear to be related to fluid inflow along thrusts. In the Ruhr basin there is little evidence for in-seam thrust detachments; the thrusts are commonly ramps and are intimately associated with the folds (Kunz & Wrede 1985). However, as in South Wales, the coal rank was developed at the time of deformation. Excursions from Hilt's law have also been observed in the volatile matter data, although it is unclear how these have been interpreted Ouch 1991). It seems at least possible that fluid inflow has had some role in the development of the Ruhr basin. Computer generated thermal modelling involving possible transport of heat into the basins by fluids is required to understand the thermal evolution of both basins.
Conclusions 1. Coals in the South Wales coal basin show an increase in rank not only with stratigraphical depth, obeying Hilt's law, but also laterally towards the northwest of the coalfield. 2. Vitrinite reflectance (Rm) studies in a thrust repeated succession indicate that rank was developed both before and during thrusting.
3. The geothermal gradient within the Lower and Middle Coal Measures of the eastern part of the coalfield is 310~ -1 and is presumed to be higher in the northwest of the coalfield. Preliminary thermal modelling of the basin suggests that burial alone cannot be responsible for this gradient which would require Late Carboniferous basal heat flow values of 295 m Wm -2. 4. Excursions from Hilt's law occur in locally developed zones associated with one or more coal seams, and most commonly with seam 37 near the base of the Middle Coal Measures. These excursions are observed in plots of volatile matter variations with depth and are interpreted as localised zones of higher temperature. Thermal gradients associated with the excursions vary from 8~ -1 to 193~ -I 5. The excursions are correlated with in-seam thrust detachments, seen as cleavage duplexes and thrust duplexes in working opencast coal mines and as rashings in the colliery records. 6. It is argued that fluids, associated with seismic activity along the thrusts, have carried heat into the coal seams, causing a local increase in the heat flow and a resultant perturbation of the thermal gradient. The fluids have also introduced minerals into the coals. 7. Comparisons with the Ruhr coal basin imply possible similarities as well as differences between the two basins. It is suggested that thermal modelling of both basins may provide the solution to an understanding of the thermal histories of the Variscan foreland basins. The maturity data collated for this study were made available by the former British Coal Corporation. We are extremely grateful to British Coal Opencast and to Celtic Energy for allowing access to opencast coal mines in South Wales and for providing plans and sections from which the thrust structure within the South Wales coalfield has been deduced. The final version of the manuscript has been greatly improved by suggestions made by Ron Austin and Chris Cornford.
References ALDERTON, D. H. M. & BEV1NS, R. E. 1996. P-T conditions during formation of quartz in the South Wales coalfield: evidence from coexisting hydrocarbon and aqueous fluid inclusions. Journal of the Geological Society, London, 153, 265~75. ALLEN, P. A. & ALLEN, J. R. 1990. Basin Analysis." Principles and Applications. Blackwell, Oxford. AUSTIN, R. L. & BURNETT, R. D. 1994. Preliminary Carboniferous conodont CAI data South Wales, the Mendips and adjacent areas, United Kingdom. M~moires Institut Gdologique de l'Universit~ Catholique de Louvain, 35, 137-153.
COALIFICATION EXCURSIONS AND THRUST-GUIDED BANGS, N. B., WESTBROOK, G. K., LADD, J. W. & BUHL, P. 1990. Seismic velocities from the Barbados Ridge Complex: indicators of high pore fluid pressures in an accretionary Complex. Journal of Geophysical Research, 95, 8767 8782. BARKER, C. E. & GOLDSTEIN, R. H. 1990. Fluid inclusion technique for determining maximum temperature in calcite and its comparison to the vitrinite reflectance geothermometer. Geology, 18, 1003-1006. BARTENSTEIN, H. & TEICHMULLER, R. 1974. Inkohlungsuntersuchungen, ein Schlfissel zur Prospektierung yon Palfiozoischen KohlenwasserstoffLagertsfitten? Fortschritte in der Geologie von Rheinland und Westfalen, 24, 129-160. BEVINS, U. E., WHITE, S. C. & ROBINSON, D. 1996. The South Wales Coalfield: low grade metamorphism in a foreland basin setting? Geological Magazine, 133, ?39-749. BROOKS, M., MILIORIZOS, M. & H1ELIER, B. V. 1994. Deep structure of the Vale of Glamorgan, South Wales, UK. Journal of the Geological Society, London, 151,909 917. COLE, J. E., MILIORIZOS, M., FRODSHAM, K., GAYER, R. A., GILLESPIE,P. A., HARTLEY,A. J. & WHITE, S. C. 1991. Variscan structures in the opencast coal sites of the South Wales Coalfield. Proceedings of the Ussher Society, 7, 375-379. DAVIES, M. M. & BLOXAM, T. W. 1974. The geochemistry of some South Wales coals. In: OWEN, T. R. (ed.) The Upper Palaeozoic and PostPalaeozoic Rocks of Wales. University of Wales Press, Cardiff, 225-261. DUDDY, I. R., GREEN, P. F., BRAY, R. J. & HEGARTY, K.A. 1994. Recognition of the thermal effects of fluid flow in sedimentary basins. In: PARNELL, J. Geofluids: Origin, Migration and Evolution q[ Fluids in Sedimentary Basins. Geological Society, London, Special Publication, 78, 325-335. FIRTH, J. N. M. 1971. The Mineralogy of the South Wales Coalfield. PhD thesis, University of Bristol. FRODSHAM, K., GAYER, R. A., JAMES, J. E. PRYCE, R. 1993. Variscan thrust deformation in the South Wales Coalfield- a case study from Ffos-Las Opencast Coal Site. In: GAYER, R. A., GREILING, R. O. & VOGEL, A. (eds) The Rhenohercynian and Subvariscan Fold Belts. Earth Evolution Science Series, Vieweg, Braunschweig., 316-348. FUCHS, W. 1946. Origin of coal and change in rank in coalfields. Fuel in Science and Practice, 25, 132. GAYER, R. A. 1993. The effect of fluid over-pressuring on deformation, mineralisation and gas migration in coal-bearing strata. In: PARNELL, J., RUFFELL, A. H. & MOLES, N. R. (eds) Contributions to an International Conference on fluid evolution, migration and interaction in rocks. Geofluids '93 Extended Abstracts, Torquay, 186-189. --, COLE, J., FRODSHAM, K., HARTLEY, A. J., HILLIER, B. MILIORIZOS, M. & WHITE, S. 1991. The role of fluids in the evolution of the South Wales Coalfield foreland basin. Proceedings of the Ussher Society, 7, 380-384. - - , GREILING, R. O., HECHT, C. & JONES, J. A. 1993. Comparative evolution of coal bearing
FLUIDS
177
foreland basins along the Variscan northern margin in Europe. In: GAYER, R. A., GREILING, R. O. & VOGEL, A. (eds) The Rhenohercynian and Subvariscan Fold Belts. Earth Evolution Science Series, Vieweg, Braunschweig, 47-82. - - , HATHAWAY,T. M. & DAVIS, J. 1995. Structural geological factors in open pit coal mine design, with special reference to thrusting: case study from the Ffyndaff sites in the South Wales Coalfield. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds) European Coal Geology. Geological Society, London, Special Publication, 82, 233-249. & JONES, J. 1989. The Variscan foreland in South Wales. Proceedings of the Ussher Society, 7, 177 179. -& RICKARD, D. 1994. Colloform gold in coals from southern Wales. Geology, 22, 35-38. GILL, W. D., KHALAE, F. I. & MASSOUD, M. S. 1979. Organic matter as indicator of the degree of metamorphism of the Carboniferous rocks on the South Wales Coalfields. Journal of Petroleum Geology, 1, 39-62. HATHAWAY, T. M. & GAYER, R. A. 1994. Variations in the style of thrust faulting in the South Wales Coalfield and mechanisms of thrust development. Proceedings of the Ussher Society, 8, 279 284. & ----1996. Thrust-related permeability in the South Wales coalfield. In. GAYER, R. A. & HARRIS, I. H. (eds) Coalbed Methane and Coal Geology. Geological Society, London, Special Publication. 109, 121-132. HARTLEY, A. J. 1993. A depositional model for the Mid-Westphalian A to late Westphatian B Coal Measures of South Wales. Journal of the Geological Society, London, 150, 1121-1136. & WARR, L. M. 1990. Upper Carboniferous basin evolution in SW Britain. Proceedings o[" the Ussher Society, 7, 21-216. JONES, J. A. 1989a. The influence of contemporaneous tectonic activity on Westphalian sedimentation in the South Wales coalfield. In: ARTHURTON, R. S., GUTTERIDGE, P. & NOLAN, S. C. (eds) The Role of Tectonics in Devonian and Carbonferous Sedimentation in the British Isles. Special Publication of the Yorkshire Geological Society, Wigley, 243-253. 1989b. Sedimentation and Tectonics in the Eastern Part of the South Wales Coalfield. PhD thesis, University of Wales, Cardiff. 1991. A mountain front model for the Variscan deformation of the South Wales coalfield. Journal of the Geological Society, London, 148, 881-891. JONES, O. T. 1949. Hilt's Law and the volatile content of coal seams. Geological Magazine, 86, 303-364. JUCH, D. 1991. Das Inkohlungsbild des RuhrkarbonsErgebnisse einer Ubersichtsauswertung. Glrickauf Forschungshefte, 52, 37 47. KEELING, G. 1988. Silesian sedimentation and tectonics in the South Wales basin: a brief review. In: BESLY, B. & KEELING, G. (eds) Sedimentation in a Synorogenic Basin Complex, the Upper Carboniferous of Northwest Europe. Blackie, London, 38 42. -
-
-
-
-
-
178
R. G A Y E R E T AL.
KUNZ, E. & WREDE, V. 1985. Exploration und Aufschluss des Nordfeldes der Zeche Haus Aden aus geologischer Sicht. Fortschritte in der Geologie yon Rheinland und Westfalen, 33, 11-32. LAWRENCE, S. R. 8r CORNFORD, C. 1995. Basin geofluids. Basin Research, 7, 1-7. LITTKE, R., BUKER, C., LI3CKGE, A., SACHSENHOFER, R. F. & WELTE, D. H. 1994. A new evaluation of palaeo-heat flows and eroded thicknesses for the Carboniferous Ruhr basin, western Germany. International Journal of Coal Geology, 26, 155-183. LIPPOLT, H. J., HESS, J. C. & BURGER, K. 1984. Isotopische Alter yon pyroklastischen Sanidinen aus Kaolin-Kohlentonsteinen als Korrelationsmarken fiir dan mitteleuropfiische Oberkarbon. Fortschritte in der Geologie yon Rheinland und Westfalen, 32, 119-150. MACKENZIE-TAYLOR, E. 1926. Base exchange and its bearing on the origin of coal. Fuel in Science and Practice, 5, 195. MCCARTNEY, J. T. & TEICHMULLER, M. 1972. Classification of coals according to degree of coalification by reflectance of the vitrinite component. Fuel, 51, 64-68. SCLATER, J. G., JAUPART, C. ~; GALSON, D. 1980. The heat flow through oceanic and continental crust and the heat loss of the Earth. Reviews in Geophysics and Space Physics, 18, 269-311. SmSON, R. H. 1994. Crustal stress, faulting and fluid flow. In" PARNELL, J. (ed.) Geofluids: Origin, Migration and Evolution of Fluids in Sedimentary Basins. Geological Society, London, Special Publication, 78, 69-84. STACH, E., MACKOWSKY, M-Th., TEICHMI]LLER, M., TAYLOR, G. H., CHANDRA,D. & TEICHMI3LLER,R. 1982. Stach's Textbook of Coal Petrology. Gebruder Borntraeger, Berlin.
STRAHAN, A. & POLLARD,W. 1915. The coals of South Wales, with special reference to the origin of anthracite. Memoir of the Geological Survey of Great Britain. TEICHMULLER, M. • TEICHMULLER, R. 1982. The geological basis for coal formation. In: STACH, E., MACKOWSKY, M. Th., TEICHMOLLER, M., TAYLOR, G. H., CHANDRA, D. d~ TEICHMULLER, R. (eds) Stach's Textbook of Coal Geology (3rd edition). Gebriider Borntraeger, Berlin, 5-86. TROTTER, F. M. 1948. The devolatilization of coal seams in South Wales. Quarterly Journal of the Geological Society, London, 104, 387-437. 1950. The devolatilization equation for South Wales coals. Geological Magazine, 87, 196-208. --1954. The genesis of the high rank coals. Proceedings of the Yorkshire Geological Society, 29, 267-303. WELLMAN, H. W. 1950. Depth of burial of South Wales coals. Geological Magazine, 87, 305-323. WESTBROOK, G., CARSON, R., MUSGRAVE, R. & Shipboard Scientific Party. 1993. Fluid flow within a convergent continental m a r g i n - results from ODP Leg 146, Cascadia margin. Geofluids '93 Extended Abstracts, Torquay, 178-180. WHITE, S. 1991. Palaeo-geothermal profiling across the South Wales Coalfield. Proceedings of the Ussher Society, 7, 368-374. WHITE, S. C. 1992. The Tectono-Thermal Evolution of the South Wales Coalfield. PhD Thesis, University of Wales Cardiff. WOODLAND, A. W. & EVANS,W. B. 1964. The Geology of the South Wales Coalfield (Part IV). The Country around Pontypridd and Maesteg (3rd edition). Memoir of the Geological Survey of Great Britain.
Deep borehole evidence for a southward extension of the Early Namurian deposits near N6m~i~ky, S. Moravia, Czech Republic: implication for rapid coalification J. D V O I ~ A K l, J. H O N I ~ K 2, J. P E S E K 3 & P. V A L T E R O V A 4
1 Czech Geological Survey, Leitnerova 22, 658 69 Brno, Czech Republic 2 Hongk Co. Ltd, Opavsk[t 4150/9, 70800 Ostrava 4-Pustkovec, Czech Republic 3 Faculty of Science, Charles University, Albertov 6, 12843 Praha 2, Czech Republic 4 Geofond, Kostelni 26, 170 O0 Praha 7, Czech Republic Abstract: Unexpected Early Namurian (Namurian A) sediments were identified in several
boreholes in the vicinity of N6m6i~ky (the N6m6i~ky basin), south Moravia. Coal fragments were recovered from the boreholes N6m 1, 2, 5 and 6. These fragments come partly from in situ coal seams, partly from eroded coal seams and partly from coalified logs. Although these fragments were recovered from depths of 2690.9 m (N~m 5) to 4803 m (N~m 1), their mean reflectance (R0) is 0.57% up to 0.9% which corresponds to subbituminous to high volatile bituminous coal. The very low rank of the coal at these depths argues for very fast coalification of the coal fragments most likely during the Carboniferous. The rank of the coal is believed not to have been affected by later burial beneath Jurassic sediments or by tectonic burial under Carpathian nappes. The presence of Early Namurian (Namurian A) sediments has been proved below the Carpathian Flysch nappes at relatively great depths in several boreholes drilled by Moravsk6 naftov~ doly Co. during exploration for oil and
natural gas in the vicinity of N6m~i6ky, SE of Brno (Fig. 1). In addition to rocks, some of which strongly resemble sediments of the Czech part of the Upper Silesian coal basin, fragments of isochronous coal were also identified. The
v POLAND
..s
Fig. 1. Schematic geological map of Moravia-Silesian Devonian and Carboniferous. 1, unfolded rocks on the platform; 2, volcano-sedimentary formations on the surface; 3, basinal formations below the flysch; 4, carbonate formations on the surface or below the flysch and molasse; 5, Upper Silesian basin; 6, borehole Jablfinka 1.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 179-193.
180
J. DVOI~.AK E T A L . A
Age
Formation
Westphalian A
Member Doubrava Such~i
Hubert fresh water group of bands
Karvinfi Saddle
Namurian
Poruba
Gaebler group of marine bands
Jaklovec
Barbora group of marine bands
Ostrava
Enna group of marine bands Hrugov Frantigka group of marine bands Pet~kovice
Main whetstone horizon
1)hiatus Fig. 2. Lithostratigraphical scheme of the Upper Silesian basin (A) and proved age of sediment filling of the N6m6i6ky basin (B).
first record of fragments of Carboniferous coal found in drilling mud was that of the N6m 1 borehole (Reh/tk 1975). However, no systematic coring of the boreholes was undertaken because the programme was aimed at testing the oil and gas potential in the area. The discovery of coal bearing Carboniferous strata was completely unexpected at such depths. The samples obtained from the drilling mud were subjected to technological tests and petrological investigation despite some uncertainty about the true number of coal seams penetrated during the drilling. It is also possible that some fragments have come from eroded and redeposited coal seams or from
coalified logs. The total length of the analyzed borehole section was from 4250 m up to 4535 m. Palynological studies suggest a Carboniferous age of the examined coal (Knobl-Jachowicz in I~eh~k et al. 1973). Another Carboniferous coal was found in the N6m 2 borehole. A bituminous coal seam, 1700mm thick, was penetrated by the borehole at a depth of 3374.30 to 3376m. Numerous coal fragments were also found in drilling mud in addition to the above mentioned coal seam. The great thickness of the coal seam was the main reason for detailed investigation of both the coal and adjacent sediments which were found
EARLY NAMURIAN COAL-BEARING DEPOSITS, CZECH REPUBLIC in the N~m 1 and 2 boreholes (Hon~k et al. 1978, Hon~k-Vrbovfi 1980, Hon6k et al. 1980, Polick)-Fialovfi 1980) and in other boreholes (see below). Phytopalaeontological studies proved the Late Carboniferous age of these sediments (Early Namurian) which appear to be the same age as the Ostrava Formation in the Upper Silesian basin. Specifically they appear isochronous with the upper part of the Ostrava Formation, commencing with the Enna marine horizon and including the Jaklovec and Poruba members (Purkyfiovfi 1978a, b). Palynological studies by Valterovfi (1978, 1982) also proved the Carboniferous age of the unit (Fig. 2). Later finds of coal fragments from the N6m 5 and 6 boreholes were studied to a lesser extent. No gamma-gamma logging was undertaken in the boreholes for technical reasons. Consequently, the true number of coal seams penetrated during the drilling remains unclear. The source of some coal fragments is also questionable. Despite this uncertainty, the results indicate the extension of the Upper Silesian coal basin into the area beneath the Carpathian Flysch nappes and have important implications for the timing and process of coalification.
Geology of the N~m~i~ky area The region under consideration belongs to the M~nin block (cf. Dvof'fik 1993) which is the southernmost part of the Paleozoic Drahanskfi Vrchovina Plateau, located south of the city of Brno. The basement beneath its sediments consists of the Precambrian Brno-granitoid massif. Two sub-blocks (Fig. 3) were distinguished in the area: the western sub-block consists of the Basal Clastic Formation of Old Red facies resting on the granodiorites. This formation is more than 1400m thick and likely Early Devonian in age. It is overlain by relatively thin (400m) reef limestones (Middle Devonian and Frasnian). The eastern sub-block has suffered greater subsidence. The sedimentation also starts with terrestrial red-purple arkoses of the Old Red facies, with unknown thickness. These were deposited on weathered granitoids of the Precambrian Brno massif. The boundary between the sub-blocks is formed by a N-S trending fault which was penetrated by the N~m 5 borehole (see Figs 3 & 4). The marine transgression recorded in the eastern sub-block reached this area at approximately the boundary between the Eifelian and Givetian. The transgression was followed by
181
deposition of dark, partly dolomitic, limestones of the La~finky Limestones which grade into light grey very pure Vil6movice Limestones. Both types of limestones belong to the Macocha Formation which is of Givetian-lowermost Fammenian age (Fig. 5). This formation is characterized by reef-building coral- and stromatoporoid faunas. The thickness of this formation is about 800 m in the west, gradually thinning to 480 m in the east. A major regression occurred at about the Middle Fammenian, following which the Vil6movice Limestones were karstified. Dark biodetrital Hfidy-l~i~ka Limestones, locally with corals and brachiopods (Gigantoproductus) were deposited after a second marine trans-gression which occurred in the Late Vis6an. These limestones are about 120 to 130 m thick in the west (N~m 2 and 5 boreholes) but thin out to the east. The Hfidy-l~i6ka Limestones are transitional toward the top into dark, locally calcareous silty shales, which represent here the Myslejovice Formation. The shales which were penetrated by N6m 5, N6m 2 and N6m 1 boreholes were 23, 47 and 38 m thick respectively. The onset of coarse-grained sedimentation which outpaced the basin subsidence occurred at the boundary between the Early and Late Carboniferous. A large body of coarse-grained petromict (polymict) conglomerates about 500 m thick was deposited along the western margin of the rapidly subsiding eastern sub-block which is only 150 thick in the N6m 2 borehole. These conglomerates are completely missing in the N6m 1 borehole. The Early Namurian is represented by a cyclic series in which grey to black-grey sandstones, locally with numerous fragments of fossil flora, are the dominant sediments. The N~m 1 borehole penetrated the following rock sequence from the bottom to the top: greywacke sandstone about 80m thick; pink arkoses about 160m thick, and finally feldspathic sandstones more than 250 m thick which represent the last member of the whole sequence terminating the sedimentation. Intercalations of black-grey siltstones and shales and also coal seams, confined to the upper part of the sequence, are much less abundant. Intercalations consisting of tuffites, often mixed in sandstones, represent a typical constituent of this sequence. The N~m 5 borehole revealed also a layer of coarse-grained conglomerates about 200m thick which must thin out towards the N6m 1 borehole. The preserved thickness of the whole sequence reaches 1100 m thick in the west. It is reduced to 600 m toward the N~m 1 borehole some 5 km
182
J. DVOI~.,~K E T AL.
§
,I§
|174
.
|
§
',1-
"t9I"
4.1 '4" t -j §
,k 9 91"
.4-
e+ ,,,t.
"t"
'tl.
"#"
.f. 4"
"f"
§
"f"
+"
.d':'l
+
',*, +.
"F. -i,,. +,. .41-
§
"#" "l-
|
|
4"41- 4-4.
+
§
§
9-, i ,
"1" 4. "t" ~ -S" -I,4"*-
~
-I! I~,J
"1,.
i I ,
+
o%'~""
++ O . o 0 . /
o
~ F,..T,~
1
-Ik 4-
~+~+~ + : / .
~IHHO+-,-
II
.t-
+
+
~+++
I
'1"
-+-
§
't'
'4"
4,. "0-
All.
-t-
"41-
-+.
.
§
§
4. "l',
+ § +
.I.
§ 't1"
§ ,~
(B~'+"
I .t91"
§
.0-| |
-~
; II
"I:'Z-Z
l I l,l~ I ~ ,,
il
I
11 II
I
tll I I11 Ii I 1I I
,
I
I
i ~ i i I I I I
~
!
I I 1
I I I
I
I
I
I i
1.1
I
I I
I I
,11 Ill I kill I
Ii
~!11 , ~ .
~.
9
9"'
9, " 9 9 9 9
9 9
"*.
:" O'
0
o o
o 0 0
O
o/ 2
o
0
0 :
l)
0
00
-
o.': .- ~
.
. , 9 .
Oj
'~
~ - l .
941" §
=E
~SL~1 ~.~ ~~
'4"
"4-
~.
Q §
"0" +
.
4" +.41.
,1.
.9. ~
4-
.O-
4.
..o= ~+" @ r
,.C3
§
O 9 ,-~ §
4.
.a-
~.9 .~
,~-
.
9 '.,
§
+ ~-~
L
. . . .
+
~
9
9
o
0
§ §
o o .-~
r:
§
4"
.1"
"~
"
~
IiIII
!
ill*~ I I IH. I
~
EARLY
NAMURIAN
COAL-BEARING
DEPOSITS,
CZECH
REPUBLIC
183
~9 9 ,
IIi
![ .... ... :. . . . . /-../ "'--.: .
9
9
.
I
I..IJ
f
,
9
~ ~
E.,
:
"
~II . . . .:.
-
"
Z
9
.
.....
.
-
9'
:
"
,
:
:
"
9
"
"
"
-
f
.
.
.
:
..
..." 9
,
:'-:-:
.. .
.
. .
." .-..
C"M
......
.
s..-:
9
.
~
" "
"-
.
."-_
.
.
.
9
.
~.~ ~.~ o
.
.
~ , ~
oo=~o
.= ~ .,~
9
..' -
o
!:il i:
. .
.
.i!
- . , '.
i9 ~~ i :~. q i
. " . . "
o
~.o~
~-~
9
O
....
: ""
~~ ' ~~=
.
:---; !..:
~
.
j
~
.'..:'":"
i'-
!...!
--.:
9 -
,."
.~
9
"
; -7~ !- i. ! . . . . :
,
o
E~ ~
:_.
-
9
9 --.."
-
-
9 ,
.
,.
--"':'";/ ::i: . /.
..._"--"
.
9 ,I
~'~
"
/..::
E
*0.,1 Z
o
::../...../ /.../:." ....
i
,
.
9
.
- . . . . ". . . . . J
. .'"-:"" :'...: "":
~
-~
: .... . .'
:"-.
- . :
-
.
:...
F--:
r:
~
-
/-.-...-
:.:...-/
~
Io
:
:
"
~-~
"
:
:-,._,
:--.q
F,."
/-.. .:~
-
::." ' ". ' . ". . ' :'- " :---:
"
.
~
-~ . . . .
:
~
"
...
,~
,
.....
.:" ..
~....:
~...~
,0,,I
j
9
;
-
1
9
""
9-.
I
:..
.
o
~ .H "H
o
:~~
"
.
,
~
/
.~0
9
,~
.
"
.:..../; !.-i i...i::. 9
,~
I
"
Z
"
".
~;:
i,/
"":: ...... l, ;".._.: '--j :i:. : -"-. : ,:..'
."
."-..:...:...:..-'~ "
9 '
9 ."
:....
i, .
,
""
""
""
"'.''.'"
,..~
":~
;
-
t
~
o
4-
o
11.
+
i" 9
~
..',.
Z
'~
~
+
§
+
#-~
~
~
.
. 9
§
4-
~
~
.2 ~ . ~ _
.
.
9
I 9
..~
~
~
<~.:~
.
.
%
'
*
+
9
,
9
~
*
4"
.
.
9
,
4"
9
.
.~
§
o
, ~ o ~ ~ J +'~ ~
*
.... . . . .
~:a~
G~,
"'i
-..~ '" i . j " "
"i
"
~
I
,':;..:
~
*, 9
~
~
~-~
~
~
~-
~
o .r ~
,.~, ~
~ ~. ~~
"~
~ r *..
9
,
.
9
~
.
.
+
Z
"
o~
0
0
o'~-~
9 ~
J. DVO15,AK ET AL.
184
SE
NW "
[
I
tower .
IGo 7 IGo B Go ,~. [Go
lm% II.ower
Mystejovice Formation
I
t
:~Jupper upper 1
Formation ............
_.........-"'5'~oJ --''~'' 7 , o
f"~//
I I
,
Macocha Formation Fig. 5. Lithostratigraphical scheme of the Devonian and Early Carboniferous of the Drahany Upland. O, Ostrov Shales; B, Brezima Formation; K, Krtiny Limestones; D, Drazovice Formation.
ESE in which the thickness of Carboniferous sediments is believed to be doubled due to an overthrust fault. The Ostrava Formation in the vicinity of the N6m~i6ky boreholes forms the southernmost and the youngest known alluvial fan belonging to the Moravian Paleozoic (Dvo~fik 1995).
Comments on tectonics The M~nin block has a tectonic boundary with the Nesva~ilka block in the north. This fault trending NW-SE was active during the Devonian and Carboniferous sedimentation. The N-S fault within the M6nin block was active during the deposition of the Ostrava Formation (Fig. 5). Many other normal faults and thrust faults can also be deduced on the basis of changes in mean reflectivity values (R0) which were measured on samples from the N6m 1, 2 and 6 boreholes. Westerly verging thrust faults are thought to be Neogene in age, accompanying the emplacement of the Carpathian nappes. Their existence has been proved palaeontologically suggesting that 100-200m of the sequence has been repeated. The eastern part of the Jurassic autochthonous sediments, together with the underlying Paleozoic, were also affected by thrust faults.
The age of coal bearing sediments and coal fragments Palynological and coal petrological investigations of coal fragments from the N6m 1 and 2 boreholes show clearly that these fragments are of Carboniferous age (cf. Knobl-Jachowicz in I~,ehfik et al. 1973; l~ehfik 1975; Valterovfi 1978). However, Elifig (1974) and Strakog-l~ehfik (1975) considered these fragments to have been redeposited into Jurassic clastic sediments. In contrast, Rehfik (1975) interpreted the coal fragments in N~m 1 borehole as evidence of the occurrence of seven coal seams whose properties, particularly their degree of coalification, correspond to the uppermost part of the Ostrava Formation (Early Namurian), specifically to the Poruba Member of the Upper Silesian basin, l~ehfik's opinion was later supported by macrophytopaleontological (Purkyfiovfi 1978a, b) and microphytopaleontological studies (Valterovfi 1978, 1982).
Macrophytopaleontological investigations Purkyfiovfi (1978a, b) found in the N6m 1 and 2 boreholes twenty plant species some of which are of Late Vis6an to Early Namurian age and others are of Namurian and Westphalian age. The assemblage indicates that the sediments
EARLY NAMURIAN COAL-BEARING DEPOSITS, CZECH REPUBLIC belong to the uppermost Early Namurian which lithologically correspond to those occurring in the upper part of the Ostrava Formation, i.e. to the Jaklovec and/or Poruba members of Upper Silesian basin. These units are characterized by the occurrence of the following species: Lyginopteris larischii, L. bartonecii, L. cf. stangeri, Sphenopteris adiantoides, Rhodeopteris stachei, Pecopteris aspera, Neuropteris cf. bohdanowiczii, Sphenophyllum tenerrimum, Mesocalamites roemeri, M. costiiformis, Lepidodendron cf. veltheimii, L. cf. obovatum and L. cf. aculeatum.
185
sediments found in several boreholes near N6m6i6ky. The occurrence of Early Namurian clastics in the N~m-1 borehole, at depths of 4253-4704m, is supported by finds of the following species of miospores: Bellispores (Artuz) Sullivan, Ahrensisporites Potonie et Kremp, Savitrisporites Bhardwaj and Tripartites (Schemel) Jachowicz. A similar assemblage of miospores was found at depths of 3274.954402.0m in the N6m 2 borehole. Among important miospores, the following species were found: Rotaspora Schemel, Schulzospora Kosanke, Verrucosisporites (Ibrahim) Smith and Butterworth.
Palynological studies Valterov~ (1978, 1982) following earlier investigations by Knobl-Jachowicz (in l~eh~tk et al. 1973), provided palynological evidence for the Early Namurian age of coal fragments and
Coal petrology The coal fragments from the N~m 1, 2, 5 and 6 boreholes are of various provenances:
Table 1. Relative proportions of macerals and microfithotypes in % in samples of a coal seam penetrated by the Ngm 2 borehole Maceral groups Sample T366 1"367 T368 T783 T369
(%)
Depth (m) 3374,30-3375,00 3375,00-3375,50 3375,50-3376.00 3374,30-3376,00 3375,60
Macerals and minerals
Telinite Collinite
T366 T367 T368 T783 T369 Sample
1,8 87,3 1,2 75,2 1,6 72,2 1,0 77,3 0:6 79,3 Micdnite SemiMacrinite fusinite 0,9 1,0 4.6 3,1 3,9 2,8 4,3 2,5 2,6 1.4
Microlithotypes Sample
T366 T367 T368 "1"783 T369
L 5,3 12,4 9,4 10,7 10,1
I
2,8 10,0 12,0 9,3 8,9
Note upper part of seam middle part of seam lower part of seam average sample lump sample
(%)
Sample
T366 T367 T368 "1"783 T369
V 91,9 77,6 78,6 80,0 81,0
L Macro- Micro- Resinite Cutinite Alginite V Total sporinite spodnite Total 0,1 0,2 5,1 89,1 0,3 4,5 0,1 0,3 12.2 76,4 0,7 11,1 -, 9,2 76,8 0,8 8.5 0,3 10,5 78,3 0,8 9,4 0,1 t0,0 80,1 0,7 9.3! M Fusinite SderoI Clay Sulphid~ Carbo- Other nates miner. Total Unite Total minerals 0,2 3,1 0,5 0,3 2,7 2,9 1,6 -. 0,1 1,7 0,5 9.9 1,5 0,2 0,1 2,2 4,3 0,8 11,7 1,9 0,1 2,0 9,21 2,0 1,8 0,5 0,1 1,1 8,81 1,0 4,4 0,4
(%)
Vitdte 67,0 43,6 45,1 45.1 45,9
Liptite
Inertite Clarite
1,3 2,1 0,5 2,7
17,5 27,0 16,0 21,4 29,7
Dudte
0,8 2,9 1,0 0,5 -I
V'ddTdCartx)- Shale nertite macerite minerite 1,1 7,3 6,3 2,O 21,4 1,81 3,2 29,4 3,2 2,3 27,3 2,9 0,8 1,0 19,8
(a)
Fig. 6. (a)-(e) Relative proportions of macerals and microlithotypes in samples from a coal seam penetrated by the N6m 2 borehole.
(c)
Fig. 6. (continued)
188
J. DVOI~AK ET AL.
Fig. 6. (continued) (i)
(ii)
(iii)
Coal fragments in drill cuttings were brought to the surface in the drilling mud. The coal particles were separated by washing the drill cuttings in a heavy liquid or by hand picking. These samples provided information on the average grade of the coal coming from a certain depth interval. Thin bands or small coal seams or seam with a minimum thickness of 170 cm was identified in the N6m 2 borehole and layers of coal claystones to coal siltstones were recovered from drill cores. Tiny fragments of organic substance (organoclasts) in clastic sediments of core runs. These samples served almost exclusively for the determination of their rank.
Coal petrology was carried out on samples from two boreholes. Drill cuttings samples from the N~m 1 borehole were prepared by maceration. The maceral composition corresponds to an average petrographic composition of the coal substance from a certain depth interval. By contrast, the samples from the N~m 2 borehole were from core and represent a coal seam. These samples were examined by maceration combined with microlithotype analysis. The results thus
represent the true coal petrographic composition from a specified coal seam. Samples from the N6m 1 borehole show that macerals of the vitrinite group vary between 78 and 84% those of the liptinite group between 10 and 16%; and the inertinite macerals from 5 to 13%. Detailed analysis of coal samples from the N6m 2 borehole show vitrinite group macerals ranging between 77.6 and 91.9%. Some samples also show a slightly increased content of liptinite group macerals (5.3-12.4%) but mostly a lower content of inertinite group macerals (2.8-12%) (Table 1 and Figs 6a-e). In contrast, a few isolated fragments from the drill cuttings of the N~m 2 borehole, examined as bulk samples, showed higher contents of inertinite. One sample contained more than 50% inertinite. This composition resembles coal of the Karvinfi Formation of the Upper Silesian basin. Unfortunately, the depth from which these particular fragments come from is unknown.
Chemical-technological analyses The samples were subjected to basic technological and chemical analyses. Samples showing ash
EARLY NAMURIAN COAL-BEARING DEPOSITS, CZECH REPUBLIC contents exceeding 10% A d were washed in a mixture of trichlorethylene and bromoform having a specifc gravity 1500kgm -3 in order to reduce the ash content. Only samples from the N6m 6 borehole were not washed because of their small volume and low coal content, i.e. analysed without reducing the ash content. This affected the results of the chemical-technological analyses particularly as far as the content of volatile matter, coking and other properties are concerned. However, the A s values of samples from the Nfim6i6ky area should be considered approximate only because only a part of the drill core and coal fragments were sampled, the rest being left for further investigations. The ash content of samples from the drill cuttings (the N6m 1 and 6 boreholes) can be uses only for the recalculation of analyses. The chemical-technological analyses show that the coal fragments were, in general, only slightly coalified. This is supported by the contents of volatile matter (V daf) which in samples from the N6m 1, 2 and 5 boreholes vary between 37.1 and 39.9%. Coal samples from the N6m 6 borehole, in fact, showed a greater range from 37.3 to 41.5% but these samples were not washed to reduce the content of ash (see above). The maximum coalification, occurred in the coal fragments derived from the greatest depth in samples from the N6m 1 borehole. The minimum coalification similarly occurred in samples observed from the N6m 5 borehole. Combustion heat values (Qdaf) of samples from the boreholes N6m 1 and 2 vary between 32.69 and 34.81MJkg -~. The lowest value Q ~af s was established in a sample from the N6m 5 borehole giving 31.14 MJ kg -1 (Table 2). A very low rank of coal results in a worsening of coking properties of the coal. Values of the swelling index vary between 0.5 to 2.0. Only the two deepest samples from the N6m 1 borehole showed negative expansion during the dilatation test. The other samples showed only some contraction. Total sulphur contents Std in coal of all the samples are rather low, ranging between 0.35 and 0.96%. The sulphur content is high only exceptionally. Samples from 2692.3m in the N~m 5 borehole, and from 3060-3080m in the N~m 6 borehole showed 3.5 and 5.35% Std respectively. These extreme values are believed to have been caused by the possible occurrence of nodules of iron disulfides (Table 2). The results of elemental analysis support the character of coalification which follows from the contents of volatile matter: the maximum content of carbon in volatile matter (C aaf) was
189
found in the N~m 1 borehole (82.84-84.32%), lower contents were established in samples from the N6m 2 borehole (80.08-82.14%), whereas the lowest concentrations were found in samples of the N6m 6 borehole (78.89-83.01%) (Table 2). The content of hydrogen in the volatile matter (H ~af) varies between 5.34 and 6.38%. The lowest content of 4.81% was found in a sample from the N6m 5 borehole (Table 2). The results of the technological tests indicate that the Carboniferous coal from the vicinity of N6m6i6ky ranges from subbituminous to high volatile bituminous coal. This is in contrast with the relatively great depth of about 3000 to 4000m from which the coal samples were recovered. Only finds from the N6m 6 borehole are from depths less than 3000 m.
Vitrinite reflectance The low degree of coalification of the organic matter established on the basis of technological tests (V daf, cdaf) is supported also by the results of vitrinite reflectance (R0) measurements. The minimum value of R0 is 0.57% (the N6m 5 borehole, depth 2690.9m) whereas the maximum value is 0.9% (the N6m 1 borehole, depth 4803.5m). This sample together with a coal fragment from the depth of 4801.5m comes from the Early Carboniferous basement of the coal bearing unit. The majority of values of Ro vary between 0.6 and 0.85% (Tables 2 and 3). According to Hilt's law, the rank of coal increases with increasing stratigraphic depth. Graphic expression of this law is a general plot of coalification (Patteisky-Teichmfiller 1960). The plot of coalification for the Czech part of the Upper Silesian basin and its use for interpretation of faults in boreholes was established by Weiss (1976). A palaeogeothermal gradient of between 70~ to 90~ per km has been calculated, using Buntebarth's method, in the Jablflnka 1 borehole, situated on the axis of a Variscan foredeep (see Fig. 1). The borehole penetrated a coalbearing formation of the Early Namurian and also rocks of the Early Carboniferous and Devonian. Boreholes situated to the west show steeper gradients of Rmaxvalues (Dvo~fik 1989).
Discussion
Timing of coalification The coal in the N6m6i6ky boreholes was coalified prior to burial beneath the Carpathian nappes because the nappe cover thickens towards the
190
J. D V O R A K E T AL.
9
I'--
~ ~ ' - ' - ' ' - ~
~
I"",,,l''~l~"'l"'~i
~
o"
0
o
o
z ~
.
.
O N , O O ~
~
,
o
o
0
0
~
,? ~ ~
0
~
~
~ ~
~ ~ N
o" o ' I o ' o " o "
o
o
o
o
o
~0
~ 0 0 0
o
~
0 0 0 0
o
~e
N N I o o
O h O '
c~ ~
~
N
'
~
~
~
l
~
i
~
a
~
~ h N ~ N
h ~ o . ,...~
N
N
h
N
N',-"
~
~
~
~- ~-
N
04 ~1 ' , - 04
r
I1~ ~
0 4
9
"~
r
I-IE
E
Eo~o
~oE
:u
:_u
E
~o~-~ ,.~
0
9 ~
Z~x~
m
EARLY NAMURIAN COAL-BEARING DEPOSITS, CZECH REPUBLIC
191
Table 3. Mean relectance (Ro) measured on coal fragments from the NYm 1, 2, 5 and 6 boreholes Borehole
I
Depth (m) N~rnEi~,ky 1 4248-4250 4310-4325 14495 4524 4530-4535 4204.50 4206.20 4254.20 4316.90 4317.00 4319.50 4382.60 4434.80 .... 4801.50 4803.50 N~m6iEky 2 3204.50 13271.50 13370.10 [3373.50 [3375.50 13478.30 [3481.00 13624.50 13625.80 [3740.20 1385g.50 13860.50
R (%) 0.75 0.80 0.77 0.76 0.79 0.70 0.71 0.74 0.78 0.81 0.79 0.83 0.79 0.90 0,90 0.60 0.60 0.68 0.63 &73 0.74 0.70 0.70 0.74 0.84 0.83 0.85
V'~ iI Borehole (~~176i1 37.2 ,NlmEiEky2 36.0~ 36.7 ,, 37.9 36.3 i,Nlm(EiEky 5 38.4 38.2 ~' 37.4 I I 36.5 F'N~mEi6J~ 35.8 36.3
I Depth I (m) I 3862,30 13863.10 14401.50 3480.00 2690.90 2692.20 2692.20 2692.30 3349.50 I 3350.80 I 3351.10 1"3843.00 13060- 3080 13095 13105- 3110 13145 13160 3185 13190-3250 13250 - 3300 t3300- 3315 i3315 13315-3325 13330 - 3345 13350-3365 J3510- 3525 13545-3550 . 13760
n
35.3]: 36.3 ,, 33.6 ,, 33.6 ,, 40.9 ,, 40.9 ,, 39.2 ,, 40.2 ,, 37.7 ,, 37.5 ,, 38.5 38.5 "'j 37.5:1 35.1 1 35.3 :i 34.9 ,,
R (%) 0.83 0.80 0.74 0.75 0.58 0.57 0.57 0.60 0.69 0.68 0.72 0.76 0.58 0.60 0.61 0.63 0.71 0.74 0.77 0.76 0.78 0.79 0.78 0.77 0.77 0.79 0.81
V n=f (%) 35.3 36.0 37:5 37.2 9 40 9 40 > 40 9 40 38.3 38.5 37.3 36.8 ~, 40 9 40 9 40 9 40 37.9 37.2 36.5 36.7 36.3 36.1 36.3 36.5 36.5] 36.1 35.7 i
Values V daf are derived from relationship R0 - Vaar, established by Weiss (1976) for the Czech part of the Upper Silesian basin. ESE (Fig. 4) and this has no effect on the coal rank in the underlying Carboniferous (c.f Fig. 4 and Table 3). By analogy with the coalification process in the Upper Silesian basin it is likely that the rank was developed during the late Carboniferous.
Conclusions
The discovery of previously unknown Early Namurian (Namurian A) sediments has been proved palaeontologically in the N~m6i6ky 1, 2, 5, 6 boreholes, drilled in the search for oil and natural gas in SE Moravia (Purkyfiovfi 1978a, b; Valterovfi 1978, 1982). However, neither systematic coring, because the program was aimed at testing the oil and gas potential of the area, nor logging of the boreholes for technical reasons was undertaken. (1)
The partially cored N6m 2 borehole penetrated part of a coal seam at a depth of 3374.3-3376.0m. This borehole and some
(2)
(3)
other boreholes (e.g. N6m 1, 5 and 6) provided the majority of coal fragments found in drilling mud. Some of them likely represent redeposited coalified logs and coal clasts eroded from coal seams. With the exception of the coal seam identified in the drill core from a specific depth in the N~m 2 borehole the remaining coal fragments brought up in drilling mud may not have originated from the indicated depths. Lithological similarities between sediments from the boreholes and those occurring in the Upper Silesian basin indicate that Early Namurian occurrences in the N~m6icky basin were linked with the Upper Silesian foredeep. Purkyfiovfi (1978a, b) considered that these sediments are isochronous with the Jaklovec and/or Poruba Members of the Ostrava Formation of the Upper Silesian basin, suggesting an Early Namurian (Namurian A) age. Palynological studies of coal and sediments also support a similar age (Valterovfi 1978, 1982).
192 (4)
(5)
(6) (7)
J. DVOI~,h,K E T AL. Coal samples from the N6m 2 borehole show the following composition (data in parenthese show the composition of coal fragments from the N6m 1 borehole): vitrinite 77.6-91.9% (78-84%), liptinite 5.3-12.4% (10-16%), inertinite 2.8-12% (5-13 %). A few samples of drill cuttings analysed as bulk samples from the N6m 2 borehole (unknown depth) show an increased content of inertinite, which in one fragment exceeds 50%. Their composition corresponds to that of coal coming from the overlying Karvin~ Formation of the Upper Silesian basin. Basic technological parameters of the coal fragments from the N6m 1, 2, 5 and 6 boreholes are as follows: A a 5.8-37.9%, V daf 37.1-41.5% Qoaf S 31.1434.81MJkg -1, S d 0.37-5.35, SI 0.5-2.0. The mean reflectance (R0) measured on Namurian coal fragments is 0.57-0.81%. Mean values of Rmax obtained from all layers of the Late Carboniferous penetrated by the N6m 1 borehole (total 11 samples) are equal to 0.7%. The calculated gradient of all boreholes is 0.03% Rmax.
Coal fragments identified in the drilling mud from these boreholes and samples of a coal seam penetrated by the N~m 2 borehole at a depth of 3374.3-3376.0 m show that their reflectance (R0) corresponds to that of high volatile bituminous, and occasionally to subbituminous coal. The very low rank of coal found in these boreholes argues for very fast coalification of peat which is in agreement with coal clasts found both in the Ostrava Formation of the Upper Silesian basin which come from eroded Early Namurian coal seams and the Westphalian C and D units of the South Wales basin (Gayer-Pe~ek 1992, Gayer et al. 1996). The rank of coal from N~m6i~ky was not affected by the deposition of overlying Jurassic sediments and a flysch cover nor by its burial under Carpathian flysch nappes. Very fast coalification provides evidence that the temperature gradient in the Variscan foredeep varied between 70 and 90~ -1 (Dvo~'~tk 1990).
References BUNTEBARTH, G., KOPPEL, J. & TEICHMOLLER, M. 1982. Palaeogeothermic in the Ruhr Basin. In: (~ERM~.K, V. & HAENEL, R. (eds) Geothermics and Geothermal Energy, 45-55. DrolL&K, J. 1980. Geotectonic condition of the forming and the extinction of the reef complex, notably in the Devonian of Moravia. Vdstnik Usoredniho fistavu geologick~ho, 55, 203-208.
1989. Anchimetamorf6za ve varisk6m tektog6nu st~edni Evropy -jeji vztah k tektogenezi. Vdstnik (/st(edniho ftstavu geologick~ho, 64, 17-20. 1990. Geology of Palaeozoic sediments of the deep borehole Jablfinka 1 (Beskydy Mts, NE Moravia)- comparison with the deep borehole Mfinsterland- 1. Sbornlk geologick~eh vdd, 45, 65-90. - - 1 9 9 3 . Moravsk~ paleozoikum. Geologie Moravy a Slezska. Sbornik pfisp6vkfi k 90. v~,ro~i narozeni prof. dr. K. Zapletala, 41-58. 1994. Varisk~ flydovf~ v~voj v NizkOm Jesenlku. Czech Geological Survey, Special Papers 3. 1995. Moravo-Silesian Zone. Stratigraphy. In: DALLAMEYER, R. D., FRANKE, W. & WEBER, K. (eds) Pre-Permian Geology of Central and Eastern Europe. Springer, Berlin, 447-489. ELIAg, M. 1974. Mikrofaci~lni v~,zkum karbonfit6 naftonad6jn~,ch oblasti na p[iklad6 autochtonni jury jihov~,chodnich svahfi Cesk6ho masivu. Zemni plyn a nafta, 19, 359-374. GAYER, R. & PEgEK, J. 1992. Cannibalisation of Coal Measures in the South Wales Coalfield- significance for foreland basin evolution. Proceedings of the Ussher Society, 7, 380-384. - - , S~'KOROV,k,I. & VALTEROVA,P. 1996. Coal clasts in the Upper Westphalian sequence of the South Wales coal basin: implications for the timing of maturation and fracture permeability. In: GAYER, R. & HARRIS, I. (eds) Coalbed Methane and Coal Geology. Geological Society, London, Special Publication, 109, 103-120. HON~K, J. & VRBOVA, V. 1980. Chemicko-technologick6 a uheln~-petrografick6 vyhodnoceni karbonsk6ho uhli z vrtfi N6m~i6ky 1 a N~m6i6ky 2. Sbornlk GeologickOho prdzkumu Ostrava, 21, 51-77. et al. 1978. Zhodnoceni vrtnf~ch jader paleozoic~ch hornin z vrtu Ndmdidky 2. MS Geofond, Praha. --, POLICK?, J. & WEISS, G. 1980. Diageneze karbonsk~ch hornin z vrtfi N6m6i6ky 1 a N6m8i6ky 2. Sbornik Geologickkho pr~tzkumu Ostrava, 21, 77-79. PATTEISKY, K. & TEICHMOLLER,M. 1960. InkohlungsVerlauf, Inkohlung-Masstabe und klassifikation der Kohlen auf Grund von Vitrit-Analysen. Brennstoff-Chemie, 41, 79-84, 97-104, 133-137. POLICKY, J. & FIALOVA, V. 1980. Petrografick~, a litologick~ charakter karbonu ve vrtech N6m6i6ky 1 a N6m6i6ky 2. Sbornik Geologickgho pr~zkumu Ostrava, 21, 49-51. PURKY]qOVA, E. 1978a. F16ra svrchniho karbonu (namuru A) v paleozoiku jv. svahfi (~esk6ho masivu u N6m6i6ky na ji~ni Morav6. Casopis SlezskOho Muzea Opava, A27, 77-86. 1978b. Makrofloristick~i korelace sedimentfi karbonu ve vrtech Zaro~ice - 1, Uhfice - 1 a 2 a N~m6i6ky 1 a 2. Zemnf~ plyn a nafta, 23, 555-566. I~EHAK, J. 1975. Carboniferous coal from M~m6i6ky 1 deep borehole near Hodonin in southern Moravia. Vdstnik (/st(edniho ~stavu geologick~ho. 50, 179-182.
E A R L Y N A M U R I A N C O A L - B E A R I N G DEPOSITS, C Z E C H R E P U B L I C et al. 1973. Zhodnocenl uhli z hlubokf:ch vrt~ v okoli Velk~ch Pavlovic. MS Geofond, Praha. STgAKO~, Z. & REHAK, J. 1975. Diskuse k vfskytu uhli karbonsk~ho st6?i na ji~ni Moravd. Sbornik II. uheln6 geologick+ konference Pfirodov~deck+ fakulty UK, 137-141. VALTEROVA, P. 1978. Palynologick~ v2~zkum ve vrtu N6m6i6ky 2. Zemnf: plyn a nafta, 23, 597-618.
--
193
1982. Zji~tdni karbonsk~ch miospor v hlubok~ch vrtech jv. svah~ Cesk~ho maslvu na jiYni Moravd. Sbornik IV. uheln6 geologick6 konference P~irodov6deck~ fakulty UK, 151-154. WEiss, G. 1976. K prfib6hu zm6n stupn6 prouheln~ni s hloubkou v 6s. 6~isti hornoslezsk6 p~inve. Sbornik Geologick~ho pr~zkumu Ostrava, 11, 9-34.
M6ssbauer spectroscopic investigation of low rank coal lithotypes IRENA
K O S T O V A l, K A L I N K A
MARKOVA 2 & KRASIMIR
KUNTCHEV
1
l Institute of Applied Mineralogy', Bulgarian Academy of Sciences, 92, Rakovska Str., 1000, Sofia, Bulgaria 2 St Kliment Ohridski University of Sofia, Tzar Osvoboditel Blvd 15, Sofia, 1000, Bulgaria Abstract: Low rank coal lithotypes- xylain, humovitrain, semifusain, fusain and liptain sampled from the Maritsa Iztok coal basin (Bulgaria) have been examined by M6ssbauer spectroscopy with no pre-concentration procedures. The results are used to identify three iron species in coal lithotypes and show that covalent iron (Fe n) related to pyrite, is the main iron species in xylain, while in humovitrain ferric iron is dominant. The total quantity of iron species in semifusain, fusain and liptain is about the same, but their distribution is different. Ferric iron dominates in all the three lithotypes. Ferrous iron, although present in smaller quantities, has a higher content in fusain than in semifusain. Our results illustrate the type of oxidation processes which formed the coal lithotypes. A transformation of Fez+ to Fe 3+ has occurred out as a result of differing oxidation processes. The intensity of that transformation increases during the destructive microbial oxidation and decreases during thermal oxidation and direct oxidation processes. The opposite transformation of ferric to ferrous iron has been achieved during both thermaloxidation and direct oxidation processes.
Detailed study of the mineral matter in coal is very important for the preservation of the environment since mineral matter may cause air, water and soil pollution as a result of the combustion of coal. There are various techniques which have been used for the analysis of mineral matter in coal. X-ray diffraction, infrared spectroscopy, thermal and microscopic analysis are the most commonly used techniques, but in some cases, due to their low sensitivity, pre-concentration procedures are required. The high sensitivity and noninterference characteristics of the M6ssbauer effect allows it to be used for the determination of several iron species without any pre-concentration procedure. Bituminous and subbituminous coal and their lithotypes have been examined in detail using M6ssbauer spectroscopy (Smith et al. 1978; Melchior et al. 1982; Martinez-Alonso et al. 1987). We have used this method for the investigation of iron species in low rank coal lithotypes which had not previously been attempted. The main iron-bearing minerals found in the Maritsa Iztok coal basin are illite, pyrite, siderite and dolomite. Some pyrite in coal is very unstable and is converted into iron sulphates with different number of hydrous water molecules.
Experimental method
Coal samples The subject of the present study is coal lithotypes of low rank: xylain, humovitrain,
semifusain, fusain and liptain sampled from the Troianovo mine No. 1 in the Bulgarian Maritsa Iztok coal basin (Fig. 1) Their characteristics are presented in Table 1. The investigated lithotypes belong to three genetic series. The first genetic series is primary plant m a t t e r x y l a i n - humovitrain; the second genetic series is primary plant m a t t e r - s e m i f u s a i n - fusain; and the third genetic series is primary plant matter - liptain. The coal lithotypes have been sampled because of their more homogeneous nature than the trivial coal molecular structure. This is the reason these petrographic ingredients have been selected to develop a model in our study.
M6~sbauerspectroscopy M6ssbauer spectra of the samples were obtained using a purpose-made spectrometer with constant acceleration and a resolution of about 0.1 mm/s per channel. The radiation source used was 57Co in a palladium matrix, while the isomer shifts were measured with respect to the center of the spectrum of a reference sample of c~-Fe. Owing to the low content of iron in the samples the spectra were accumulated to reach a signal of 10 6 counts per channel. To determine the line parameters the spectra were processed by the computer program 'M6sspec' for iterative approximation of the experimental points through a sum of Lorentzian profiles using the least-square method.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 195-199.
196
I. KOSTOVA E T AL.
> k~
BULGARIA
(
- - -
GREECE ~'~
Varna et'/"J" ~
.
.
.
.
..J
.
45km
MaritsaIztokCoal Basin
Fig. 1. Location of Maritsa Iztok coal basin in Bulgaria.
Table 1. Characteristics of lithotypes Lithotypes
Xylain Humovitrain Semifusain Fusain Liptain
Proximate analysis (wt%)
Elemental analysis (%)
Moisture
Volatile matter V daf
Carbon
Hydrogen Nitrogen
Sulphur
Oxygen
Wa
Ash content Ad
cdaf
H daf
N daf
sdaf
odaf
7.2 10.0 9.4 8.8 4.0
2.6 5.9 9.2 4.4 2.4
61.1 56.4 35.1 17.0 70.9
68.7 66.9 71.1 87.2 70.8
6.8 5.2 4.3 3.1 7.0
0.7 1.2 0.8 0.8 0.8
4.7 3.2 3.3 1.9 3.7
19.1 23.5 20.5 7.0 17.7
a, analytical; d, for dry basis; daf, for combustible basis.
Results The results obtained from the M6ssbauer study of low rank coal lithotypes are presented in Fig. 2 and Table 2. Ferric (Fe 3+) and ferrous (Fe z+) iron, including covalent iron (FeII) were established in different quantities in the five examined lithotypes - xylain, humovitrain, semifusain, fusain and liptain. With Fe u we identify the covalent iron, connected with pyrite in contrast to the other ferrous iron, which may be related to other mineral phases such as siderite and dolomite, which have been detected with X R D analysis, or organic compounds. The ferric iron is connected mainly with the clay mineral illite. The observed iron species can be subdivided in
two groups (a) and (b) with similar M6ssbauer parameters. These groups have centre shifts (6 mm/s) and quadrupole splittings (A mm/s) as follows: group (a) 0.32-0.42 and 0.62-0.85 group (b) 1.05-1.33 and 2.11-2.99 Ferric or ferrous covalent iron, or a combination of both, belong to group (a). It is not possible to determine separately ferric and covalent iron due to a covalent deposition of iron in pyrite on the one hand and the close M6ssbauer parameters (centre shifts of 0.32-0.35 and quadrupole splitting of 0.60-0.65mm/s) of both iron species on the other hand. An additional real
LOW RANK COAL LITHOTYPES
197
(1)
(2)
(3) ..,~_ (-'1
.> .1-,, m
(4)
fl) n"
(5)
9
-a
t
I
-2
1
I
I ,,,
1
2
o
1
r
~
[
I
(;
J
I
+3
l
I
I0
Velocity (mm/s)
Fig. 2. M6ssbauer spectra of (l) xylain; (2) humovitrain; (3) semifusain; (4) fusain and (5) liptain. Velocity axis is with respect to ~Fe. difficulty is the insignificant content of iron in these samples. Ferrous iron belongs to group (b). It may be related to the carbonate minerals, siderite and dolomite, in coal. Iron from group (a) has been established in xylain and humovitrain in the first genetic series (Table 2). With the help of other parallel studies (XRD, SEM and TEM), and according to unpublished data it can be demonstrated that in xylain iron is connected mostly with pyrite, i.e. covalent iron (FeII) dominates. The expected presence of Fe 3+ is in a subordinate quantity, about 8 to 10%. With the second representative of the genetic series, humovitrain, because of the higher value of quadrupole splitting, the iron present is Fe 3+ (Fig. 2).
In the members of the second genetic series, semifusain-fusain, iron of both group (a) and group (b) has been established. Fe 3+ and Fe 2+ can be clearly distinguished in semifusain (Fig. 2). A very small quantity of covalent iron can be masked by Fe 3+. Ferrous iron is present in considerably smaller amounts (11.4%), and ferric iron predominates. It is very likely that Fe 2+ is connected with carbonates, siderite and dolomite, which have been established in this type of coal under X-ray diffraction and SEM analysis. With the second representative of the genetic series, fusain, the iron forms observed coincide with those established in the semifusain, but their quantities are different (Table 2). The amount of Fe z+ is greater (group b), and the
198
I. KOSTOVA E T AL. Table 2. M6ssbauer parameters a for lithoO'pes low rank
Sample
Group b
Centre shift c
Quadrupole splitting
Line widthd
Xylain
a
0.32 4- 0.05
0.62 4- 0.05
0.56 4- 0.08
100+8
Humovitrain
a
0.35 4- 0.015
0.66 4- 0.1015
0.41 4- 0.025
100 4-4
Semifusain
a b
0.334-0.15 1.124-0.03
0.694-0.015 2.11 4-0.03
0.564-0.03 0.26+0.07
88.64-2 11.44-4
Fusain
a b
0.424-0.01 1.334-0.01
0.854-0.01 2.994-0.01
0.554-0.02 0.424-0.02
67.6 4- 3 32.4 4- 3
Liptain
a b
0.36 4- 0.06 1.054-0.10
0.65 4- 0.06 2.73+0.1
0.30 4- 0.08 0.384-0.06
80.04- 10 20.0 4- 10
a All parameters are in mm/s. b a is assigned as pyrite (FeII) or Fe 3* or combination of both. b is Fe 2+. c The centre shifts are reported relative to c~Fe. d Width at half maximum of the peak.
amount of the iron of group (a) less (Fe 3+ or/ and Fe n) in comparison with semifusain, as can be demonstrated by the high quadrupole splitting of the first peak (Fig. 2). The basic iron present (67%), is Fe 3+. If there is any admixture of covalent iron, its amount will a be minimal, (about 5 to 7%). One third of the iron established in fusain is bivalent. In semifusain, and especially in fusain, an increased content of macropores is observed (Markova et al., 1992) with 98.4% in semifusain and 99% in fusain. A large proportion of the clay minerals and pyrite is found in these pores. The porous structure of these lithotypes is related to the high ash content (Table 1). It is likely that a great part of the Fe 3+ and Fe z+ is connected with mineral matter. In liptain, from the third genetic series, iron of both (a) and (b) groups has also been established, Fe 3+ (80%) being predominant (Table 2).
Discussion
Ferrous iron was found to be present in the low rank lithotypes of the first genetic series: x y l a i n - humovitrain formed as a result of gelefication under microbial oxidation destruction conditions (Sigkov 1988). However, in xylain covalent iron (Fe n) is dominant indicating pyrite. Pyritic iron in humovitrain is probably present at rather low concentration, in the range 10-15 wt%. Salts of metals with variable valency have been identified in peat bogs (Garrels & Maskenty 1974). It is suggested that these salts have acted catalitically during the
decomposition of the peroxides and hydroperoxides to free radicals: R O O H + Fe 2+ ~ RO" + Fe 3+ + OH"
(1)
These peroxides and hydroperoxides have been produced by the oxidation of organic matter. The resultant free radicals RO" and OH" are very active and are the reason for the polimerization process (Kucher et al. 1980). The formation of lithotype maceral of the first genetic series represents a continuation of the destruction by microbial oxidation (Si~kov 1988). Therefore, the probable reason for the high Fe 3§ content in the final product of this genetic series, humovitrain, is the continuous oxidation process represented by mechanism (1). According to Sigkov (1977) this lithotype appears to be a huminic polymer. The lithotypes of the second genetic series: semifusain - fusain contain both Fe 3§ and Fe 2§ These lithotypes are a product of fusanization which has taken place as a result of thermal processes in a strongly acidic medium with high oxygen fugacity (Sigkov 1988). It can be assumed that due to this intensive oxidation process the peroxide and hydroperoxide groups have disintegrated under the action of the salts of the transition metals by a combination of reaction (1) and mechanism (2) (Ivanov 1970): R O O H + Fe 3+ ---+ROO" + Fe 2+ + H +
(2)
However, our results indicate that the processes of formation of ferrous iron {mechanism (1)} are dominant not only xylain and humovitrain but also in semifusain and fusain. With the
LOW RANK COAL LITHOTYPES progressive increase in the fusanization process, from semifusain to fusain, the intensity of reaction (1) decreases, while the intensity of a mechanism (2) increases (i.e. the ferric iron content increases). The lithotype of the third genetic s e r i e s liptain, which was formed under the direct action of oxygen, contains both iron species - Fe 3+ and Fe 2+ but the ferrous iron predominates. Consequently, it can be concluded that the peroxides and hydroperoxides produced by oxidation have been decomposed by the action of metals with variable valency according to the two mechanisms as discussed above. The results illustrate the processes of oxidation associated with the formation of the coal lithotypes.
Concusions Low rank coal lithotypes contain three iron species which can be identified using M6ssbauer spectroscopy without pre-concentration procedures. Covalent iron (Fe n ) related to pyrite is the main iron species in xylain. However, in humovitrain ferric ison is dominant. Almost 100% ferric and ferrous iron in different quantities has been determined in the three lithotypes - semifusain, fusain and liptain. Ferric iron dominates in all three. Ferrous iron is present in smaller quantity but it increases in fusain. Our study demonstrates that a transformation of Fe 2+ to Fe 3+ has been carried out as a result of different oxidation processes. The intensity of that transformation increases during microbial oxidation destruction and decreases during
199
thermal oxidation and direct oxidation processes. The opposite transformation of ferric to ferrous iron occurs during both thermal oxidation and also direct oxidation processes.
References GARRELS, P. & MACKENZY, F. 1974. Evolusia osadachnix porod. Ser. Earth Sciences, Vol. 58, Mir, Moskwa. IVANOV, S. 1970. Verishni radicalovi reactsii. Nauka I izkustvo, Sofia. KUCHER, R.V., KOMPANETS,V. A. & BUTUZOVA,L. F. 1980. Structura iskopaemix uglei i ix osobenost k okisleniu. Naukova dumka, Kiev. MARKOVA, K., RADEV, G. & KOSTOVA, N. 1992. Razpredelenie por v ugolnix litotipax niskogo ranga. Ximia tv topliva, 3, 20-22. MARTINEZ-ALONSO, A., GRACIA, M., GANCEDO, R., GONZALEZ-FLIPE, A. R. & TASCON, J. M. D. 1987. The roles of organic and mineral matter in aerial oxidation of brown coal. In: MOULUN,J. A. et al. (eds) International Conference on Coal Science 1987.
MELCHIOR, D. C., WILDEMAN, T. R. • WILLIAMSON, D. L., 1982. Mrssbauer investigation of the transformations of the iron minerals in oil shale during retorting. Fuel, 61, 516-522. SIgKOV, G. D. 1988. Teoretichni osnovi na biohimichnata vaglefikatsiya. Univ. Izd. St. Kliment Ohridski. SMITH, G. V., Liu, J. H. & SAPOROSCHENKO,M. 1978. Mrssbauer spectroscopic investigation of iron species in coal. Fuel, 57, 41-45. VOITKEVICH, G. V., KIZILTSTEIN,L. I. & HALODKOV, J. I. 1983. Rol organicheskogo vechtestva v konsentrasii metalov v zemnoi kore. Nedra, Moskva.
Comparison of solid state 13C NMR of algal coals/anthracite and charcoal-like fusinites: further evidence for graphitic domains P. I. P R E M O V I ( ~ , R. S. N I K O L I ( ~ & M. P. P R E M O V I ( ~
Laboratory for Geochemistry and Cosmochemistry, Department of Chemistry, University of NiY, P.O. Box 91, 18000 NiY, FR Yugoslavia 13C NMR crosspolarization (CP)/magic angle spinning (MAS) spectrometry. We examined two charcoallike fusinites from Serbia: the Jerma (Jerma mine) and Miro~ (mine 'Aliksar') seams. This examination revealed that atomic H/C ratios calculated (on the basis of the CP/MAS parameters) for fusinites studied are higher by 68% (Jerma) and by 64% (Miro6) than the H/C values which are determined by elemental analysis. Calculated H/C values infer that either more carbon or less hydrogen is required for the fusinite structures than is contained in the samples. We conclude that the differences in the estimation of H/C for bituminous charcoal-like fusinites between solid state 13C NMR and elemental analysis can be explained by graphitic domains within the maceral 'invisible' in the CP/MAS experiment. Abstract: Carbon distribution in coals and coal macerals was studied by
Coals have been subjected to many magnetic resonance studies, and many parameters have been measured to obtain information about coal molecular structure. During the last decade researchers have focused their attention on solid state 13C N M R spectroscopy with CP/ MAS because in principle this technique provides a non-destructive way to measure the aromatic/alkenic carbon fraction of coals fa (the ratio of aromatic/alkenic carbon Car to total carbon C), one of the key parameters which characterize the coal structure (Wilson & Vasallo 1985). Usually in 13C N M R spectra of most coals two broad lines can be distinguished belonging to the aromatic/alkenic and aliphatic carbon atoms respectively: fa of coal is defined as the ratio of the integrated line intensity for aromatic/alkenic carbon atoms to the total integrated line intensity (Speight 1994). In addition to aromaticity fa, dipolar dephasing (DD) experiments provide estimates of other structural parameters of coals including the aliphatic (s) and the aromatic/alkenic (p) fraction which are protonated (Wilson & Vasallo 1985). Yet major problems exist concerning the use of solid state 13C N M R spectrometry in coal research. One such problem is that the 13C N M R experiment gives inadequate quantitative estimation of carbon distribution in coal which strongly contradicts other geochemical data (Premovi6 et al. 1992). This contributes to the difficulty in the unequivocal interpretation of the N M R data. The purpose of this report is to show that the estimation of H/C for fusinite (maceral.of the inertinite group) with solid state 13C N M R is not in agreement with elemental data. For the sake of clarity, we will consider only two fusinite materials from two Serbian
seams: Jerma (the Jerma mine) and Miro6 (the Aliksar mine) with high maceral purity (>90%). For comparison, two fresh-water algal coals (torbanites) (Scotland and S. Africa) and, marine algal coal tasmanite (Australia) and the Vrgka Cuka anthracite (Serbia) were also examined. The earliest spectroscopic work of which we are aware which discusses chemical structure of algal coal-torbanites is that of Millais & Murchison (1969). These authors investigated five torbanite samples from: S. Africa, France and Scotland. Their petrographic examination indicates that these freshwater coals contain alginite (maceral of the exinite group) in excess of 90% by volume. Cane & Albion (1971) have proposed that alginite is an oxidative polymer of straight-chain alkadiens of molecular formula: CH2 = CH(CH2),,CH = CH(CH2)4CH3 (n = 15, 17 and 19). Allan et al. (1979) analyzed three torbanites (S. Africa, Australia and Scotland) using various geochemical and optical techniques. They concluded that the torbanites are composed of polymeric materials which contain relatively high proportion of aliphatic structures. According to Allan et al. (1979) the evidence for aromaticity is conflicting but the total Car is suggested to be small on the basis of the elemental analysis and infrared (IR) spectra. For a number of years, this laboratory has been engaged in the structural elucidation of coals and kerogens. Premovi6 et al. (1987) studied two torbanites from Scotland and S. Africa by both 13C N M R CP/MAS technique combined with DD experiments and 1H N M R MAS technique. This examination has shown that these coals have predominantly both aliphatic carbon and protons ( > 9 5 % of total
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 201-205.
202
P. I. PREMOVI(~ E T AL.
organic carbon and hydrogen) incorporated into polymethylene (-CH2-) skeleton structures.
Experimental procedure The isolation procedure was similar to that used by Premovi6 (1984) and Premovi6 et al. (1986). Powdered rock (50g) was extracted with benzene methanol (3:lv/v) for 96h in a Soxlet apparatus. The residue remaining in the Soxlet thimble was treated with boiling hydrochloric acid (HC1, 4 M) to remove most of the carbonates. Carbonate removal was checked by IR analysis. The insoluble residue was further demineralized by repeated treatment with boiling hydrofluoric/hydrochloric acids (HF/HC1, 22 M and 0.25 M, respectively). This acid mixture removes silicates and the removal was checked by IR analysis. The final residue is the coal sample. It contained only small traces of inorganic minerals, including pyrite, as confirmed by the electron microprobe analysis. Algal coals (Table 1) are of Permian age (about 250 Ma). The Vrgka (~uka antracite and the maceral concentrates from Serbian seams (Table lb) are of Jurassic age (about 200 Ma). All 13C N M R spectra of the coal samples were recorded at 25.15 MHz on a Bruker CXP-100 as previously described (Premovi6 et al. 1986). The
1 K FIDS, acquired with a 3 ms contact time, 0.35 second recycle time and a rotor frequency of c. 4 kHz, were zero filled to 8 K before Fourier transformation. The pulse sequence employed for obtaining the dipolar dephasing (DD) spectra is described elsewhere (Premovi6 et al. 1987). Proton N M R spectrum of tasmanite was taken at 270MHz with MAS and BR-24 at room temperature on an N M R pulse spectrometer constructed in the laboratories of the Friedrich Schiller University, Jena (Germany) (Premovi6 et al. 1987). For F T I R analysis, the sample was mixed with anhydrous potassium bromide and pressed into the disc (2.5 mg/150 mg KBr) with a load of 200 MPa. The spectra were recorded at room temperature on a Bruker ISF l13V F T I R spectrometer.
Results and discussions In addition to torbanites, we have studied by 13C CP MAS marine algal coal: tasmanite containing more than 90% by volume sporinite (maceral of the exinite group). Figure 1 shows typical 13C and 1H N M R spectrum of tasmanite which indicate a presence of a strong aliphatic carbon (Fig. l a) and proton (Fig. l b) bands
Table 1. Geochemical data on the coals." (a) algal coals," (b) bituminous coal macerals (Serbian seams) b (a) Location
Maceral
C
H
(O,N) a
)ca
s
H C
H C
0.09 0.12 0.10
0.89 0.92 0.90
1.66* 1.57" 1.53"
1.61t 1.62t 1.62t
fa
S
p
H C
H C
1.00 0.90 0.80
0.00 0.80 0.80
0.25 0.55 0.65
0.25* 0.40* 0.50*
0.25~ 0.66~ 0.85~
(mol/kg) South Africa Scotland Tasmanite
alginite alginite sporinite
67 69 60
111 108 92
5 4 1
(b) Location
Maceralc
C
H
(O,N) a
(mol/kg) Vr~ka6uka Jerma Miro~
vitrinite inertinite inertinite
76 78 76
19 32 38
0.7 2 3
* Experimental. t Calculated using expression 2. Calculated using expression 1. a Dry, ash-free corrected maceral data. b Separated by sink-and-float procedures by heavy liquids starting with hand-picked lithotypes that were rich in the desired maceral. Predominant maceral (>95%) component.
SOLID STATE 13C NMR
203
--CH 2--
a
~i = I
.
9
l
160
,
I
140
~
l
120
i
I
,
1
~
!
1
,
100 80 60 40 CHEMICAL SHIFT (Plan)
1
20
9
1
0
~
1
-20
Aliphatlc
b
A
1
10
~
1
5
0
CHEMICAL SHIFT
L
(ppcn)
Fig. l. 13C (a) and 1H NMR (b) spectra of the powdered sample of the Vrgka Cuka anthracite. inferring that more than 95% of both organic carbons and protons are aliphatic. Thus these results suggest that the tasmanite is an aliphatic material which also contains a relatively high proportion of polymethylene chains and rather low amount of aromatic/alkenic groups in the structures. It is likely that the best estimated fa value for these materials is close to 0.10 (Premovid et al. 1987). If total coal carbon is apportioned to both aliphatic (Ca0 and aromatic/alkenic carbons C~r, then C = C a l + C a r assuming an overall H/C
value of: 2 for aliphatic portion and 1 for aromatic/alkenic part, we may write H = 2SCal +pCar were H is the total hydrogen of coal. Combining these two equations we obtain H ~ - = 2s(1 - f ~ ) +Pfa-
(1)
As N M R study indicates that fa values for the algal coals in question is small (c. 0.1) and that
204
P. I. PREMOVI(~ E T AL.
most of aromatic/alkenic carbons are nonprotonated (Premovi6 et al. 1987) then the product Pfa is small and can be neglected. In this case, formula (1) is simplified into the form: H
~ - = 2s(1 -fa).
in elemental data tend to imply) there is a good correspondence between experimental (obtained by elemental analysis) and calculated (through NMR data) H/C ratios for algal coals considered here. We have also studied the Vrgka Cuka anthracite using both the 13C N M R CP/MAS (Fig. 2a), 1H N M R MAS (Fig. 2b). The results show that this coal has predominantly polyaromatic structures withfa = 1.0 and consists chiefly of the vitrinite maceral (>90%). It has been suggested that all X3C atoms in these structures
(2)
The H/C values (calculated using the expression 2) of algal coals studied are listed in Table la. Unless the calculations are more seriously in error (than the stated uncertainly
Aromatic
a
I
160
,
I
140
,
I
1
120
100
,
1
80
,
l
,
60
I
,
40
1
20
CHEMICAL SHIFT (ppm)
Aromatic
b
-*:
1
I
lo o CHEMICAL SHIFT (OOm)
,
Fig. 2. 13C (a) and ]H NMR (b) spectra of the powdered sample of tasmanite.
~
1
0
,
i
- 20
SOLID STATE 13C NMR are not equally cross polarized with IH nuclear spins. Most researchers of the subject now agree, however, that the number of ~3C atoms in these coal structures that are not observed by 13C N M R is small (Speight 1994). Since the anthracite is wholly polyaromatic, the total carbon can be expressed as C = C a r . If this is correct then the total hydrogen is given by H = pCar = pC. Hence p = H/C. The calculated value of H/C(= p) shown in Table 1b is in excellent agreement with the experiment. Table 1b lists atomic H/C ratios of Serbian (Jerma and Miro~) fusinite samples computed using expression (1). An examination of this table reveals that calculated H/C values for the fusinite samples are higher by 68% (Jerma) and by 64% (Miro6) than those experimentally determined values. Thus, the H/C values calculated on the basis of the CP/MAS parameters suggest that either more carbon or less hydrogen is required for the fusinite structures than it is contained in the sample. If this notion is valid then there are only two reasonable explanations for the contradiction (between experimental and calculated H/C value given in Table lb): (1) the fusinite carbons are extensively substituted e.g. by O or N for which there is, however, no persuasive geochemical evidence; and, (2) in the fusinite structures there are carbon atoms which do not show their resonancies in the 13C N M R CP/MAS spectrum. In general, the CP/MAS experiment relies on the presence of organic structures abundant in protons in order to observe the 13C N M R resonancies. Consequently, the 13C N M R spectra do not show signals from carbon atoms in structural domains within coal lacking protons, such as graphite. On the other hand, physical, chemical and other studies indicate that the coal fusinites are similar to natural charcoals which is consistent with the view that these macerals had been exposed to elevated temperatures and charred before incorporation in the sediment (Panti6 and Nikoli6 1973). If this concept is true then fusinites as natural charcoal materials would undoubtedly contain a high amount of graphitic components which are inactive for the CP/MAS approach. Thus, we suggest that the differences in the estimation of H/C for charcoal-like fusinites (Serbian seams) between solid state 13C N M R
205
and elemental analysis can be explained by graphitic domains* within these macerals invisible in the CP/MAS N M R experiment. This work is supported by a grant to PIP from Ministry of Science (Serbia), Project 0206. Special thanks to: the late D. Urogevi6 who supplied the fusinite samples (Serbia), the Vrgka Cuka mine (Serbia) for providing the anthracite sample, and Buerau of Mineral Resources, Geology & Geophysics (Australia) for supplying the tasmanite sample.
References ALLAN, J., BJOROY, M. & DOUGLAS, A. G. 1979. A
geochemical study of the exinite group maceral alginite, selected from three Permo-Carboniferous torbanites. In: DOUGLAS, A. G. & MAXWELL, J. R. (eds) Advances in Organic Geochemistry 1979. Technip, Paris, 599-618. CANE, R. F. & ALBION,P. R. 1971. The phytochemical history of torbanites. Journal of the Proceedings of the Royal Society New South Wales, 104, 31-37. MILLAIS, R. & MURCHISON,D. G. 1969. Properties of coal macerals: infrared spectra of alginites. Fuel, 48, 247-258. PANTIE, N. & NIKOLIC,P. 1973. Ugalj. Nau6na knjiga, Belgrade. PREMOVlC, P. I. 1984. Vanadyl ions in ancient marine carbonaceous sediments. Geochimica et Cosmochimica Acta, 43, 873-877. --, PAVLOVIC, M. S. & PAVLOVIC, N. Z. 1986. Vanadium in ancient sedimentary rocks of marine origin. Geochimica et Cosmochimica Acta, 50, 1923-1931. , STOJKOVIC, S. R., PUGMIRE, R. J., WOOLFENDEN, W. R., ROSENBEREG, H. & SCHELER, G.
1987. Spectroscopic evidence for the chemical structure of algal kerogens. In: RODRIGUEZCLEMENTE, R. R. & TARDY, Y. (eds) Proceedings of the International Meeting 'Geochemistry of the Earth Surface and Process of Mineral Formation '. C.S.I.C., Madrid, 421-430. , JOVANOVIC, Lj. S., & MICHEL, D. 1992. SolidState 13C and 1H NMR in kerogen research: Uncertainty of aromacity estimation. Applied Spectroscopy, 46, 16-18. SPEIGHT, J. R. 1994. Application of spectroscopic techniques to the structural analysis of coal. Applied Spectroscopy Review, 29(2), 117-169. WILSON, M. A. & VASALLO, A. M. 1985. Developments in high resolution solid state 13C NMR spectroscopy of coals. Organic Geochemistry, 8, 299-312. * Consists chiefly of amorphous charcoal.
Composition and properties of North Bohemian coals IVANA HELENA
SYKOROV,~
PAVLiKOV,~
l, J A R O S L A V
3 & ZUZANA
( ~ E R N ' ~ 2,
WEISHAUPTOV,~
1
l Institute of Rock Structure and Mechanics, Czech Academy of Sciences, V HoleYovidk(tch 41, 182 09 Prague, Czech Republic 2 Department of Petroleum Technology and Petrochemistry, Institute of Chemical Technology, Technickdt 5 166 28 Prague, Czech Republic 3 N M R Laboratory, Institute of Macromolecular Chemistry, Czech Academy of Sciences, Heyrovskdho ndtm. 2, 162 06 Prague, Czech Republic Abstract: This work presents the mean chemical, micropetrographic, surface and other characteristics of coal seams from western, central, and eastern parts of the North Bohemian brown coal basin. Attention was especially paid to the elemental composition, ash content, content and forms of sulphur, occurrence of syngenetic and epigenetic sulphides, maceral composition, and degree of gelification and decomposition of components in the huminite maceral group. Some other coal characteristics were also assessed, such as pore texture, extractability and solvent swelling of the coals. The coals examined were huminitic with a variable xylite and detrite content. Huminite reflectance varied between 0.33 and 0.39%. Substantial differences in pore texture of the coals were found in the range of meso- and macropores. These differences largely affected the extractability of the North Bohemian coals. The coals also exhibited extremely high swelling ratios in basic solvents, such as pyridine.
The N o r t h B o h e m i a n basin is the most important b r o w n coal basin in the Czech Republic. It is situated south of the Kru~n6 H o r y m o u n t a i n s and has an area of approximately 1400 k m 2.
The main coal seam belongs to the M i o c e n e Most F o r m a t i o n . The only mineable seam in the N o r t h Bohemia coal basin is the m a i n coal seam with an average thickness of 30 m a n d m a x i m u m
i
,_.r-'/% L
S' '73.75"bs7
@~
.....
.+f'
~"'"'"<,.<:.,. ,,
Fig. 1. Map of North Bohemian coal basin.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 207-217.
~,~
208
I. SS(KOROV~, E T AL.
thickness of 60 m. It consists mainly of huminites. Liptobiolites and liptodetrites occur rather rarely. Very stable interlayers divide the seam into a few, mostly three, beds of different quality. The seam is approximately subhorizontal except where it was deformed by postsedimentary tectonism or where it is draped over crystalline or volcanic basement. The seam is best developed at Most and in the western part of the basin near Chomutov (See Fig. 1). Coal with a higher ash content occurs in the main seam in the vicinity of Chomutov. The lower part of the seam has a higher content of pyrite and so tends towards a pyritic coal clay. In the P6tipsy-~Zatec part of the North Bohemian coal basin (Zatec delta), the main seam commonly splits into a few almost non-mineable leaves. In the Bilina delta, the main seam is divided into more irregular beds. Sudden splitting, pebbles in coal and an increase in mineral matter content occur near the Kru~n6 Hory fault, (Elznic 1963). The seam is only rarely fully developed. Some parts of the seam, i.e. the highest part in the Most area and the basal part in the Teplice area, are uneconomic due to a high mineral content. Generally, the main seam has a higher mineral content in the south than in the north. The main seam, with a lower thickness is also known in some isolated deltaic lobes in the Doupovsk6 Hory and (~esk6 Sffedohofi mountains (Bou~ka et al. 1995). Xylodetritic and semidetritic coals prevail in the North Bohemian coal basin. Less frequent are xylitic and detritic coals. The economic coals are massive, glossy and semiglossy soft coals, and partly oxyhumolites. Para- and orthotypes,
and in the deepest areas (Hrdlovka, Osek, Dfinov) even metatypes, of brown coal were formed in relation to their depth of burial. Coal quality varies across the basin. In open pit mining, the whole seam is occasionally worked. In some open pit mines, remnant pillars from deep mines are exploited. The average thickness of the mined seam varies between 12 and 26 m Q~ between 11 and 20MJ/kg, ash content between 7.4 and 39.8 wt% dry and total sulphur content between 0.4 and 2.9wt% dry (Pe~ek 1993). Coal composition, huminite properties and degree of gelification of North Bohemian brown coals were extensively studied in the 1960s and 1970s in relation to the coal processing and briquetting (Kurz 1981; Malfin 1962). Higher liptinite content in coal increases the hydrogen, carbon and volatile matter contents, heat of combustion and tar yield (Svoboda 1953; Zelenka 1974) as well as the coal reactivity during gasification, pyrolysis, combustion, and liquefaction (Hrn6i~ & B~rta 1982; Dehmer 1989; Furimsky et al. 1990; Martinez-Tarazona et al. 1994). Coal mined in the North Bohemian basin is used for the power industry. Previously, it was used for low temperature pyrolysis to produce synthetic liquid fuels and gas. The aim of our work was to evaluate the composition and structure of North Bohemian coals. In addition, evaluations such as coal extractability, solvent swelling, pore texture, and spectroscopic analysis were carried out. Most of the coal samples were from the main seam of the North Bohemian basin.
Table 1. Overview of samples and their description Coal
Coal description
Western part
Libou~ Nfistup
clay-like coal with strips of xylite, clearly visible sulphate forms xylitic-detritic coal, brown-black colour, low occurrence of finely dispersed sulphide, partly weathered, whitish sulphate forms
Central part
Vr~any Sverma (~SA 92 CSA 93 Le~fiky Bilina Centrum Kohinoor
xylitic-detritic coal without apparent sulphides, isolated strips of clay mineralization xylitic-detritic coal without massive and visually observable mineralization of sulphides and clay minerals xylitic-detritic coal, partly clay-like, isolated tiny flat sulphide concretions xylitic-detritic coal, brown-black colour, without visually observable sulphides xylitic-detritic coal, brown-black colour, without visually observable sulphides xylitic-detritic coal with isolated fusite strips, sulphides in the form of tiny flat concretions and thin veins xylitic-detritic coal without apparent mineralization xylitic-detritic coal with sparse occurrence of sulphides in the form of flat concretions and tiny massive fillings of cracks
Eastern part
Chaba~ovice
xylitic-detritic coal without apparent mineralization
NORTH
BOHEMIAN
COALS
209
Table 2. Technical analysis of North Bohemh~n coals Coal
Ash
Sd
Sdo4
Sd
Sdo
Volatile matter (wt%, d a f )
32.9 8.6 15.7 15.7 21.8 4.7 19.0 3.4 4.1 8.6 3.7
0.9 1.8 2.5 0.9 2.4 0.7 1.7 1.4 0.5 1.5 0.3
0.2 0.5 0.1 <0.1 0.4 <0.1 <0.1 <0.1 <0.1 <0.1 -
0.4 0.2 1.2 <0.1 0.5 <0.1 0.8 0.7 <0.1 0.5 -
0.3 1.1 1.2 0.8 0.6 0.7 0.8 0.6 0.4 0.9 -
51.0 50.5 49.8 46.8 48.8 55.9 49.1 47.9 50.3 48.2 48.3
wt%, dry
Liboug Nfistup Vrgany Sverma CSA 92 t~SA 93 Le2fiky Bilina Centrum Kohinoor Chaba~ovice
Calorific value (MJ/kg) 27.5 28.0 28.1 30.0 29.4 32.1 29.0 29.8 30.1 29.9 28.8
Std, ado4 , S d and Sao denote the total, sulphate, pyritic, and organic sulphur, respectively.
Table 3. Elemental analysis of North Bohem&n coals (wt%, daf) Coal
C
H
N
So
O (diff)
Liboug Nfistup Vrgany Sverma (~SA 92 (~SA 93 Le~fiky Bilina Centrum Kohinoor Chaba~ovice
68.0 70.2 67.9 71.5 71.1 74.6 69.9 72.1 72.5 71.7 70.7
5.4 4.9 5.4 5.9 5.3 6.2 5.5 5.4 5.6 5.6 5.3
1.4 1.8 1.3 1.0 0.9 1.0 1.3 1.0 1.0 1.2 1.1
1.6 1.2 1.4 0.9 2.0 0.9 1.1 0.7 0.5 1.0 0.3
23.6 21.9 24.1 20.7 20.8 17.3 22.2 20.9 20.4 20.6 22.6
Table 4. Petrographic analysis of North Bohemian coals and characteristics of iron sulphides (%) Coal
Libou~ Nfistup Vrgany Sverma (~SA 92 (~SA 93 Le~fiky Bilina Centrum Kohinoor Chaba~ovice
R0
0.33 0.35 0.33 0.36 0.38 0.37 0.35 0.36 0.39 0.39 0.37
H
64 87 78 78 71 84 74 90 86 83 88
L
5 4 6 7 6 13 5 6 10 7 6
I
1 1 1 3 1 1 5 1 1 1 1
MM
30 8 15 12 22 2 4 3 3 9 5
Microscopic forms of FeS2 FeS2
dispersed
clusters
inclusions massive
3.1 2.2 3.6 0.6 0.4 0.7 0.6 2.0 1.3 2.2 0.3
1.8 0.8 1.4 0.2 0.4 0.2 0.2 0.0 0.3 1.1 0.3
1.3 1.2 1.9 0.2 0.0 0.3 0.2 0.0 0.7 0.5 0.0
0.0 0.2 0.3 0.2 0.0 0.2 0.2 1.0 0.3 0.6 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0
R0, light relectance of huminite; H, huminite content; L, liptinite content; I, inertinite content; MM, mineral matter content.
210
I. SYKOROVA ET AL.
Experimental procedure Samples Samples represented the typical development of the main seam in the North Bohemian coal basin. The western part of the seam was represented by the coal samples from the N~stup and Libou~ mines, central part of the seam by coal samples from the Vr~any, Sverma, CSA, LO.~iky, Bilina, Centrum, and Kohinoor
mines, and eastern part by the coal sample from the Chaba~ovice mine. An overview of the samples is given in Table 1.
Methods of examination of the brown coal samples The ash content (Ad), volatile matter (vdaf), total sulphur (STY), individual sulphur forms
Fig. 2. Types of macerals in North Bohemian coals: a, ulminite; b, densinite; c, textinite; d, liptodetrinite; e, inertinite; f, syngenetic pyrite.
N O R T H B O H E M I A N COALS 1.
Sdso4' Soa), heat of combustion, and biogenous elements (C, H, N, S, O) in the organic matter were determined according to national standards. Petrographic evaluation of the coal matter samples consisted of the determination of reflectance of gelified huminite macerals (R0), determination of maceral group contents of huminite, liptinite, and inertinite, and determination of the mineral content with emphasis on the various forms of iron sulphides. U n d e r the same conditions as the maceral analysis, the syngenetic and epigenetic types of sulphide mineralization were determined based on the following four FeS2 classes (Grady 1977; Frankie & H o w e r 1987; R e n t o n and Bird 1991; H o n 6 k 1992; S~korovfi & Vodi6kovfi 1992): ( sd'
2. 3. 4.
211
crystals and framboids dispersed t h r o u g h the coal matter crystals and framboids forming clusters in the coal matter microscopic fillings of cellular walls, compartments and microcracks macroscopic massive grains.
Microscopic and m i c r o p h o t o m e t r i c measurem e n t of the polished surface of the coal grains was performed on an U M S P 30 Petro microscope-microphotometer (Zeiss-Opton) in oil immersion ( n = 1.518) and in reflected light at a wavelength of 546 nm. Total magnification was 450 • Fluorescence analysis was carried out with the same instrumentation using a halogen discharge lamp and an F109 filter set.
Table 5. Maceral composition of huminite (%) and factors of gelification and tissue preservation in North Bohemian coals Coal
Attrinite Densinite Textinite Ulminite Gelinite Corpo- Huminite Gelification Tissue preservation huminite index index
Liboug 6 Nfistup 6 Vrgany 4 Sverma 19 CSA 92 9 0SA 93 15 Le~fiky 9 Bilina 6 Centrum 3 Kohinoor 3 Chaba~ovice 6
16 36 28 8 15 15 11 24 18 21 19
0 2 4 2 1 traces 2 2 1 0 3
38 38 37 47 46 48 48 50 54 50 52
2 4 4 1 1 4 3 6 7 8 7
1 1 1 1 traces 2 1 2 3 1 1
64 87 78 78 72 84 74 90 86 83 88
2.4 4.4 2.7 1.1 2.0 2.0 2.0 3.7 4.8 3.9 2.6
1.8 0.9 1.2 1.8 1.9 1.5 2.1 1.5 2.1 1.6 1.8
Table 6. Pore texture characteristics of the North Bohemian coals (daf bas&)
Coal Libou~ Nhstup Vr~any Sverma ~;SA 92 C;SA 93 Le~/tky Bilina Centrum Kohinoor Chaba~ovice
Vmm (cm3 "g-l)
Vmicro (cm3 "g-l)
Smm (m 2 "g-l)
Smicro (m 2 "g-l)
Prmm (%)
Prmicro (%)
0.166 0.092 0.118 0.122 0.170 0.222 0.136 0.094 0.160 0.105 0.083
0.072 0.076 0.073 0.063 0.065 0.051 0.079 0.074 0.072 0.067 0.076
9.6 3.6 8.3 10.8 10.7 21.6 9.5 7.8 13.3 11.2 6.3
260 277 263 224 230 180 276 259 249 238 267
20.7 11.5 14.1 14.4 18.7 25.3 15.9 10.7 17.0 11.9 10.2
9.1 9.5 8.8 7.5 7.1 5.9 9.3 8.5 7.6 7.6 9.3
Vmm , m e s o - and macropore volume; Vmicro,micropore volume; Smm , m e s o - and macropore surface; Smi.... micropore surface; Prmm, porosity based on meso- and macropores; Prmicro, porosity based on micropores.
212
I. SS(KOROVA ET AL.
The apparent density (da), volume (Vmm),and surface area (Smm) of mesopores and macropores were determined by high-pressure mercury porosimetry on a Carlo Erba Porosimeter 2000. The volume of micropores (Vmicro) was determined by the Dubinin-Raduskiewic method, which is based on a CO2 isotherm measured at 25~ (Carlo Erba Sorptomatic 1800). The surface area of micropores (Smicro) was calculated by the Medek method. The true density (dr) was calculated from the apparent density by subtracting the pore volume from the total volume. Coal extractability was examined by using Soxhlet extraction with chloroform, tetrahydrofuran and pyridine at the boiling point of the respective solvent. Soxhlet extraction with methanol was applied after the coal had been extracted with pyridine to remove pyridine from coal pores. Volumetric coal swelling in toluene, tetrahydrofuran and/or pyridine was determined after an equilibration time of seven days. The height of dry and/or swollen coal in the tube was measured after centrifugation at 3000 rpm for 3 min. Swelling index, i.e. the ratio of swollen to dry coal volume, was calculated on a daf basis using densities which resulted from porosimetric measurements. Average density of minerals was considered as 2.3 g/cm 3. FTIR spectra of HC1/HF demineralized coals in a form of KBr pellets were measured on a Bruker IFS 88 spectrometer. One hundred scans were acquired at 2 cm -1 resolution to obtain an average spectrum. Spectral intensity was adjusted to a sample concentration of 1 mg/cm 2. Quantitative evaluation of FTIR spectra was done by an assessment of the 3000-2750cm -~ and 900700cm -a spectral regions using Solomon's
Fig. 3. Meso- and macropore distribution in the Nfistup coal.
absorptivities (Solomon & Carangelo 1988). 13C CP MAS N M R spectra were measured on a Bruker MSL 200 spectrometer at a frequency of 50.32MHz and a contact time of 1 ms. The SP MAS spectra were obtained with a pulse delay of 100 s. About 3000 scans were acquired to obtain an average spectrum.
Results and discussion
Chemico-technological character&tics o f the samples The results of the chemico-technological analyses are consistent with the data of North Bohemian coals obtained by Hub~i6ek (1964) and Malkovsk~ et al. (1985). The composition and properties of the coals strongly reflect their mineralization, degree of coalification, and petrographic composition. In this work, mineralization of the samples was examined by measuring their ash content, total sulphur, individual sulphur forms, and by microscopic evaluation of the pyritic-marcasitic mineralization. Generally, mineralization in the North Bohemian coal basin increases from north to south (Malkovsk2~ et al. 1985; Havlena 1964). Data in Table 2 show that coals with a high ash content, between 15.7 and 32.9wt% occurs in the western (Liboug) and central (Vr~any, (~SA, Le2~iky) part of the basin. Most abundant are clay minerals, quartz, and pelosiderites; carbonates are less frequent (Bou~ka 1981). Mineral matter is finely dispersed through the detrite, fills the cellular tissue compartments,
Fig. 4. Meso- and macropore distribution in the Centrum coal.
NORTH BOHEMIAN COALS
213
or forms distinct bodies. Iron sulphides are stable mineralization components in the basin. Increased total sulphur content up to 2.5 wt% (Table 2) was found in samples from the western and central parts of the basin. The sulphur content was mostly formed by organic (up to 1.1 wt%, dry) and pyritic (up to 1.2wt%, dry) sulphur. The sulphate content was low (below 0.5 wt%, dry) and it was presumably associated with the epigenetic oxidation of iron sulphides. With respect to the ash content, the more coalified coals are in the central part of the basin, where the coals exhibited rather a high
calorific value (between 29.0 and 32.1 MJ/kg, Table 2) and somewhat higher carbon content (between 70.7 and 74.6wt%, Table 3). Brown coal metatypes with higher calorific value and carbon content frequently occur in that region, particularly at Osek (Zelenka 1973). The petrographic composition, particularly a high liptinite content, is a further parameter affecting the chemical composition of coal (Wolfrum & Wawrzinek 1981). This is well documented by the CSA 93 coal, for which the highest carbon and hydrogen content, calorific value and volatile matter content were found.
Fig. 5. Extractability of North Bohemian coals in chloroform.
Fig. 7. Extractability of North Bohemian coals in pyridine.
Fig. 6. Extractability of North Bohemian coals in tetrahydrofuran.
Fig. 8. Relationship between tetrahydrofuran extractability and surface of meso- and macropores for North Bohemian coals.
I. SYKOROV,~ ET AL.
214
Petrographic characteristics of the samples The samples examined were huminitic coals with variable xylite, detrite, and mineral contents. Huminite reflectance varies from 0.33 to 0.39% corresponding to the brown coal orthotype. Huminite concentration varies from 64% to 90% and is the most abundant maceral group (Table 4). The macerals, ulminite (Fig. 2a) and densinite (Fig. 2b) contributes to the high huminite content (Tables 4 and 5), a characteristic of orthotypic brown coals. The concentration ofattrinite and gelinite is substantially lower, i.e. 3-19% and 1-8% respectively. Textinite (Fig. 2c) and corpohuminite contents does not exceed 4%. Liptinite is formed by waxy and resinous substances. The maceral composition of liptinite was determined by a fluorescence measurement. Sporinite and liptodetrinite (Fig. 2d) predominate particularly in the CSA 93 and Centrum samples (Table 4). The remaining liptinite macerals, i.e. cutinite, resinite, exsudatinite, bituminite, suberinite, and fluorinite, are accessory.Inertinite (Fig. 2e) is more abundant in only the Sverma and Le~fiky coals (Table 4), where inertodetrinite and macrinite are found in addition to sclerotinite and fusinite. The results presented in Table 4 indicate that the syngenetic pyrite type (Fig. 2f) with typical fine grained forms - euhedral crystals and framb o i d s - predominate in the coals. The crystals and framboids can occur dispersed through the organic as well as inorganic coal matter. More frequently, however, they form clusters which were rather extensive and compact in the coal samples from Liboug and Vrgany. The fillings of
25 "o
~ -~
Pyridine
Porometric analysis of brown coals The results of porometric analysis of the coals (Table 6) indicate that the values characterizing the microporous texture, i.e. volume (Vmicro) and surface area (Smicro), oscillated within a very narrow range. It confirms that the microporous texture is a characteristic of natural coals, in which the microporous phase forms the basis of the coal matrix. Appreciable differences in the porous texture were found for meso- and macropores. Their maximum volume (Vmm) differs from the minimum value by a factor of three, their maximum surface area (Smm) is up to sevenfold with respect to the minimum value. This indicates that different total porosity values, which are 19-31% of the coal volume (Table 6), primarily mirror differences in the meso- and macropore distribution. Except the N/tstup coal, the pore distribution is almost identical for all coals with
O
20-
Table 7.
0
15
0
0
~-
0
0
0
5-
0 0
I 5
CoalswellingindexesJbr North Bohemian coals
Coal
oC~ 6 ~
~., 9 10
uJ
microcracks and microscopic grains by massive FeS2 were also classified as syngenetic iron sulphides. The macroscopic grains by massive FeS2 (Bilina) were classified as the epigenetic type. Diessel (1986) has introduced two petrographic indices, i.e. the gelification index (GI) and tissue preservation index (TPI). These indices well characterize a coal depositional environment. In our study, slightly modified GI and TPI (Kalkreuth et al, 1991; Whateley & Tuncali 1995) were assessed for North Bohemian coals, and they are presented in Table 5. Values of GI and TPI indicate that North Bohemian coals had been developed in a limno-telmatic and/or telmatic environment of deposition.
J 10
t 15
t 20
25
S u r f a c e of m e s o - a n d m a c r o p o r e s (m2/g daf)
Fig. 9. Relationship between pyridine extractability and surface of meso- and macropores for North Bohemian coals.
Liboug Nfistup yrgany Sverma I~SA 93 Le2fiky Bilina Centrum Kohinoor Chaba~ovice
Swelling index (daf) in solvent Toluene
Tetrahydrofuran
Pyridine
1.05 1.07 1.04 1.14 1.19 1.10 1.07 1.10 1.04 1.15
1.6 1.5 1.8 1.8 2.2 1.7 1.9 1.8 1.9 2.0
2.8 2.3 2.5 2.8 3.0 2.6 2.6 2.7 2.6 3.1
NORTH BOHEMIAN COALS a maximum in the region of mesopores. The Nfistup coal has an almost constant relationship between pore radius and pore volume. Comparison of a typical pore distribution, represented by the Centrum coal, with the pore distribution of the Nfistup coal is apparent from Figs 3 & 4.
Coal extractability As expected, the extractability of coals increased from chloroform to pyridine. The chloroform extract accounted for 1.2 to 8.6 wt% daf of the coal, for tetrahydrofuran and pyridine the extract yields were found in the range 4.0 to 13.1wt% daf and 8.0 to 25.0wt% daf, respectively. As evidenced from the broad yield ranges, large differences in extractability have been found between individual coals. Extraction yields are graphically presented in Figs 5 to 7 for the respective solvents. The highest differences in extractability between coals were found for chloroform. An average extractability was found between 2 and 3wt% daf with an exception of the Sverma, (~SA and Centrum coals. Chloroform mostly extracts highly aliphatic waxes and resins, whose composition is quite different from that of the bulk of coal. Stronger solvents, such as tetrahydrofuran and pyridine, are able to extract compounds which are a part of the coal macromolecular network or can be considered as clusters from which the coal network is formed. Consequently, differences in pyridine extractability seem to be less than for chloroform, however, the (~SA 93 coal still exhibited exceptionally high pyridine extractability. Another coal with an exceptional
215
extractability is the Nfistup coal. This coal in all cases gave the smallest extraction yield. When compared with porometric characteristics, the extractability was in a close relationship with a surface of meso- and macropores. The relationship is presented in Figs 8 & 9 for tetrahydrofuran and pyridine extractions. As expected, a relatively close relationship was also found between liptinite content and extractability of coal in either solvent, because liptinite is known to consist of easily extractable material.
Coal swelling Coal swelling is an important investigative tool into the thermodynamic interactions and physicochemical structure of coal. When simplified, swelling is a measure of the extent of some of the intermolecular interactions, which account for crosslinking the coal macromolecular network (Quinga & Larsen 1988). For the North Bohemian coals, extremely high values of coal swelling were found, especially in pyridine. Coal swelling indexes for the coals are listed in Table 7. It can be supposed that coal swelling has some relation to coal extractability. For instance, pyridine is known to break hydrogen bonding, which contributes to coal network crosslinking. Simultaneously, it can liberate coal clusters bonded to the coal network only through hydrogen bonding and extract them from the coal. It can be seen from Figs 10 & 11 that there is a general tendency for increasing extractability with coal swelling. However, the individual points are highly spread in the plots and the relationships are not good.
15 t~ -o
Tetrahydrofuran
"0
~
6
C
0
o
o
Pyridine
0
20-
o o
v
"~,
25-
0
12
3
v
"o
0
15-
Q~
0
C
o
o
10-
0
o
5-
X
w
0 1.0
I
I
1.5
2.0
Swelling index Fig. 10. Relationship between swelling index
and tetrahydrofuran extractability for North Bohemian coals.
LU 2.5
0 2.0
I
I
2.5
3.0
Swelling index Fig. 11. Relationship between swelling index and tetrahydrofuran extractability for North Bohemian coals.
3.5
216
I. SYKOROVA E T AL. Table 8. Structural parameters of North Bohemian coals derived from FTIR and solid state 13C NMR
Coal
fa,cv
Liboug Nfistup Vrgany Sverma (2SA 93 Le~iky Bilina Cen trum Kohinoor Chaba~ovice
0.53 0.57 0.52 0.50 0.51 0.57 0.55 0.50 0.54 0.58
fa,SP 0.72 0.68 0.61 0.65 0.70 0.68 0.69
HaLIR
LH,IR
0.16 0.17 0.14 0.14 0.14 0.17 0.17 0.14 0.14 0.16
0.15 0.14 0.13 0.14 0.14 0.16 0.15 0.13 0.13 0.14
faaH
(H/C)al
0.19 0.19
2.15 2.22
0.23 0.25 0.21 0.19
2.02 1.95 2.18 2.24
0.20
2.13
fd, carbon aromaticity; Har, hydrogen aromaticity; fa,H, ratio of hydrogen bearing
aromatic carbon to total carbon; faaH, ratio of hydrogen bearing aromatic carbon to aromatic carbon; (H/C)ab atomic H/C ratio in aliphatic structures. Spectroscopic characteristics
Quantitative spectroscopic characterization of coal is rather a difficult task. It is a consequence of a high heterogeneity of coal and its multicomponent nature. An attempt was made to assess some basic structural parameters of North Bohemian coals by FTIR and solid state 13C NMR spectroscopy. The latter was performed by using the cross-polarization (CP) as well as single pulse (SP) techniques. It should be noted that the 13C CP MAS NMR discriminates against aromatic carbon in coals and the SP technique is often recommended instead. However, the SP technique is time-consuming and acquisition of one coal spectrum can take over 24 hours. Quantitative FTIR spectroscopy deals with problems of average absorptivities of CH bonds, especially aliphatic (Cern~ 1995). So, aromatic hydrogen concentration was only evaluated from FTIR spectra. Some structural parameters derived from FTIR and solid state NMR spectra of North Bohemian coals are presented in Table 8. The values of fa and Hat parameters were directly obtained from the NMR and FTIR spectra, respectively. Other structural parameters were derived by using simple equations without any assumption dealing with coal structure (Cern~ & Pavlikovfi 1995).
occur in the western and central parts of the basin. The eastern part of the basin is characterized by low sulphur content. Main organic components are formed by macerals of huminite and liptinite groups, of which ulminite, densinite and liptodetrinite occur most frequently. Pore texture analysis of the coals showed very little differences in the range of micropores. However, substantial differences were found in the range of meso- and macropores. These differences and/or content of liptinite largely affect the extractability of the North Bohemian coals. The coals also exhibit extremely high swelling ratios in basic solvents, such as pyridine. Authors gratefully acknowledge the financial support of the Grant Agency of Czech Republic (Grant No. 104/94/1791) and the Grant Agency of Czech Academy of Sciences (Grant No. 246101 and 3046607).
References
BOUgKA, V. 1981. Geochemistry of Coal, Elsevier, Amsterdam. , PEgEK, J. et al. 1995. Mineralizace uheln~,ch sloji. Report for Ministry of Environment, Prague. (~ERNV, J. 1995. Structural dependence of CH bond absorptivities in FTIR spectra of fossil fuels. Abstracts of the lOth Spectroscopic Conference, Lanskroun, Czech Republic, p. M1-P-8. - -
Conclusion
The overall composition and properties of North Bohemian coals depend slightly on their geological position in the basin. The analyzed coals were orthotype brown coals. Higher ranked coals were found in the central part of the basin. Some coals with a higher ash content
- -
& PAVLiKOVA, H. 1994. Structural Analysis of Low-Rank-Coal Extracts and Their Relation to Parent Coals. Energy & Fuels, 8, 375-379. & 1995. Quantitative solid state NMR and FTIR spectroscopy of low rank coals and reliability of structural parameters. In: PAJARES, J.A. & TASCON, J. M. D. (eds) Coal Science, Vol. 1, Proceedings of the 8th International Conference on Coal Science, Oviedo, Spain. Coal Science and Technology, 24, 111-114.
N O R T H B O H E M I A N COALS DEHMER, J. 1989. Petrographical and organic geochemical investigation of the Oberpfalz brown coal deposit, West Germany. International Journal of Coal Geology, 11, 273-290. DIESSEL, C. F. K. t986. The correlation between coal facies and depositional environments. In: Advances in the Study of the Sydney Basin, Proceedings of 20th Symposium, University of Newcastle, 19-22. ELZN1C, A. 1963. Severozfipadni omezeni chomutovsko-mostecko-teplick6 pfinve. Reports of the Central Geological Institute, 38, 245-251. FRANKIE, K. A. & HOWER, J. C. 1987. Variation in pyrite size, form and microlithotype association in the Springfield (No. 9) and Herrin (No. 11) coals. Western Kentucky. International Journal of Coal Geology, 7, 349-364. FURIMSKY, E., PALMER, A. D. et al. 1990. Prediction of coal reactivity during combustion and gasification by using petrographic data. Fuel Processing Technology, 25, 135-151. GRADY, W. C. 1977. Microscopic varieties of pyrite in West Virginia coals. Society of Mining Engineers, AIME, 262, 268-274. HAVLENA, V. 1964. Geologie uhelnf:ch lo3isek. 2. CSAV, Praha, 293-353. HON~K J. 1992. Fe-disulphides in the brown coal of the North Bohemian brown coal basin. Aeta Montana, Praha, 86, 45-49. HRN~il~, J. & BARTA, V. 1982. Macerfily skupiny exinitu (liptinity)- vlastnosti, fluorescen6ni mikroskopie, chemickfi struktura a vztah ke zkapalfiovfini. Hnddd uhli, 4, 31-38. HUaA~EK, J. 1964. Pasportizace a klasifikace hn6d~ch uhli (~SSR a jejich popelfi. Pr6ce (?stavu pro vfzkum a vyu3itl paliv, 9, 3-138. KALKREUTH, W., KOTIS, T. et al. 1991. The geology and coal petrology of a Miocene lignite profile at Meliadi Mine, Katerini, Greece. International Journal of Coal Geology, 17, 51-67. KURZ, R. 1981. Eigenschaften der rheinischen Braunkohle und ihre Beurteilung als Roh- und Brennstoff. Fortschritte in der Geologic yon Rheinland und Westfalen. Krefeld, 29, 381-425. MAL~,N, O. 1962. Die Entwicklung der optischen Methoden zur Beurteilung yon Braunkohle. Freiberger Forschungshefte, A253, 5-20.
217
MALKOVSK~-, M. 1985. Geologie severo(esk~ hnddouhelnO pdmve a jeff okoli. Ust~edni Ust~av Geologick~, Praha. MARTiNEZ-TARAZONA,M. R., MARTiNEZ-ALONSO,A. & TASCON, J. M. D. 1994. Characterization of common lignite, xylitic lignite and pyropissite varieties of low-rank coals. Fuel, 73, 11, 1723-1728. PEgEK, J., KOULA, J. et al. 1993. Refiln6 vyt~!teln6 z~isoby uhli v otev~en2~ch dolech v Cesk+ Republice.(~ist II. Hn~d~ uhli. Uhll, Rudy, 2, 337-345. QUINGA, E. M. Y. & LARSEN, J. W. 1988. Solvent Swelling of Coals. In: Y~3RI3M, Y. (ed.) New Trends in Coal Science. Kluwer, Dordrecht, 85-116. RENTON, J. J. & BraD, D. S. 1991. Association of coal macerals, sulphur, sulphur species and the iron disulphide minerals in the three columns of the Pittsburgh coal. International Journal of Coal Geology, 17, 21-50. SOLOMON, P. R. & CARAYGELO, R. M. 1988. FTIR analysis of coal. 2. Aliphatic and Aromatic Hydrogen Concentration. Fuel, 67, 949-959. SVOBODA, J. V. 1953. Geologicko-petrografickdt studie v hnddouheln~ oblasti komo(anskd. CSAV, Geotechnica, Praha, 17, 1-63. SYKOROVA, I. & VODICKOV.h,,A. 1992. Microscopical forms of Fe disulphides. Acta Montana, Praha, 85, 93-104. WHATELEY, M. K. G. & TUNCALI, E. 1995. Origin and distribution of sulphur in the Neogene Beypazari Lignite Basin, Central Anatolia, Turkey. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds), European Coal Geology. Geological Society, London, Special Publications, 82, 307-320. WOLFRUM, E. & WAWRZINEK, J. 1981. Beziehungen zwischen petrographischen und chemischen Eigen-schaften rheinischen Braunkohle. Braunkohle-Hefte, 11, 381-386. ZELENKA, O. 1973. Chemicko-technologick6 ukazatele stupn6 prouheln~ni severo6esk6 hn~douheln6 pfinve. Uhli, 22, 59-63. 1974. Vliv voskfi a pryskyfic na chemickotechnologick~ vlastnosti uhli v severo~esk~ hn6douheln6 pfinvi. Uhli, 23, 8 321-325.
Aliphatic biological markers in Miocene Maritza-Iztok lignite, Bulgaria MAYA
STEFANOVA 1 & CAROLINE
MAGNIER 2
l Inst. Org. Chem. Bulg. Acad. Sci. Acad. G. Bonchev str. bl. 9 Sofia 1113, Bulgaria 2 Inst. Francais du P~trole, 1 & 4, Av. de Bois-Pr~au, 92506 Rueil-Malmaison, France Abstract: Chromatographic separation and mass spectral studies of bitumens extracted from
lithotypes (humovitrain, xylain, liptain and humoclarain) from the Miocene Maritza-Iztok lignite reveal the following: (1) A preponderance of the c~-phyllocladane skeleton over the pimaranes/abietanes. The variety of tricyclic diterpanes confirms that the generation biota included gymnosperms; (2) The presence of dicotyledonous angiospermaes in the generation biota is indicated by the occurrence of widely distributed des-A-lupane structure; (3) The
dominance of terpanes over steranes confirms the prevailing contribution of terrigenous input; (4) The preponderance of C27f~hopane and the presence of/3/3 hopanes indicates a low degree of thermal maturation.
The early to mid-Miocene 'Maritza-Iztok' lignite is located in southeastern Bulgaria (see Fig. 1 in Si~kov, this volume for location). It is described as having the following maceral composition: 92% huminite, 6% liptinite and 2% inertinite (5]i~kov 1988). The basin is included in the 'A sub-group' of the petrological classification of Neogene deposits (Si~kov & Andreev 1987) characterized by an almost equal contribution to coal formation by gymnosperm and angiosperm plants with a slight dominance of angiosperms. The vitrinite reflectance (R0) of the lignite is 0.18 +0.02% (Si~kov et al. 1986). This study is of the humovitrain (I), xylain (II), liptain (III) and humoclarain (IV) lithotypes and involves the acquisition of information concerning the molecular composition of the aliphatic portions of the extracted bitumen. This allows the possible plant precursors to the lignite organic matter to be determined.
Experimental methods Ultimate and proximate analyses of the samples under study are summarized in Table 1. The maceral composition of the lithotypes has been described in our previous study of the products after Nail treatment of the lithotypes (Velinova et al. 1993; Stefanova et al. 1996). Bitumens were exhaustively extracted by CHC13 (1:10) and separated into aliphatic and aromatic portions by column chromatography using an analytical glass column (100 • 5 mm) dry packed with silica gel (Davison, 100-200 mesh) and previously washed with methylene chloride (CH2C12). The adsorbent was activated for 16 hours at 180~ Bitumen was dissolved in a minimum of CH2C12 and placed at the top of the column. The sample was eluted with CH2C12 (fraction 1), and MeOH/CHCI3/H20 (25:60:4 v/v) (fraction 2). The first fraction was subjected to
Table 1. Proximate and ultimate analysis of lithotypes
Characteristics
I
II
III
IV
Proximate analysis (wt%) Moisture Ash (db) VM (daf)
16.0 9.2 52.8
9.2 3.6 58.5
6.9 3.0 68.1
16.9 22.6 60.5
Ultimate analysis (wt% daf) C H N S Odi~
65.8 5.5 1.0 4.0 23.7
67.3 6.1 0.9 3.3 22.4
72.1 7.2 0.8 3.4 16.5
65.1 6.5 1.6 3.8 23.0
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 219-228.
220
M. STEFANOVA & C. MAGNIER
thin layer chromatography (TLC) separation for isolation of the paraffinic/naphthenic compounds. TLC plates were developed with CH2C12. Berberine sulphate (Fluka) in methanol was applied as a spraying reagent to make the bands visible in day light. The hydrocarbon fraction was separated to branched/cyclic and normal alkane sub-fractions by molecular sieve adduction (5A). Total hydrocarbon fractions and branched/cyclic alkane sub-fractions were characterized by gas chromatography (GC) and by gas chromatography-mass spectrometry (GC-MS). GC analyses were conducted using a Varian 3600 instrument and a DB1 capillary column (30 m, 0 0.32 mm, 0.1 #m film thickness). The program consisted of 50~176 at 10~ 110~176 at 3~ rain, and isothermal hold for 20min. Samples were introduced with an oncolumn injector heated from 50-300 ~ at 20~ The detector was operated at 300~ and the data were acquired by a Hewlett Packard 1000 integrator. G C - M S analyses were conducted using a Hewlett-Packard 5890 gas chromatograph coupled to a VG70-250SE mass spectrometer. Scans were acquired over the mass range of 60-550da. Mass spectra were corrected for background by subtraction. A DB-5 (30m, 0.25mm, film thickness 0.5#m) fused silica capillary column was used. GC conditions were: column program 50~176 at 3~ flow rate 1 ml/min, splitless for first 80 s.
lithotypes. The highest bitumen content is in liptain (39wt%) with 2.34wt% saturates composed of branched/cyclic alkanes. As the content of linear alkanes is low the total extract from the liptain sample was not adducted. Figure 1 shows the GC separations of extracted bitumens. The high peak of ct-phyllocladane eluted by n-C20 alkane is recognized in all samples. Distribution curves for the homogeneous lithotypes, humovitrain (I) and xylain (II) are smooth and calculated carbon preference index (CPI) values are close to unity. A similar pattern of alkane distribution was determined by Markova et al. (1993) for lithotypes from the same deposit, n-alkane distributions without a
a)
28
26
3O 26 22 20
,,[1
R e s u l t s and discussion
-r-
Table 2 summarizes the contents and compositions of the extracted bitumens from different
Table 2. Characteristics of the bitumens from lignite lithotypes Content (wt%)
I
Yield of bitumen Saturates Branched/cyclic alkanes * ** PhyUocladane * **
c~
27
Lithotype II
III
1.1 0.06
10.3 0.29
39.0 2.34
8.9 0.38
50.3 0.03
54.9 0.16
98.0 2.29
53.8 0.20
1.10 1 5 . 0 37.10 0.0003 0.024 0.847
IV
2.76 0.006
* wt% from the sample subjected to separation; **wt% from initial lignite lithotype; Lithotype I, humovitrain; II, xylain; III, liptain; IV, humoclarain.
2325 ]2i31
TIME
Fig. 1. Gas chromatograms of the total bitumen fractions of lithotypes (a) humovitrain; (b) xylain; (e) humoclarain; (numbers over peaks refer to carbon numbers of n-alkanes).
BIOLOGICAL M A R K E R S IN M I O C E N E LIGNITE well m a r k e d odd/even p r e d o m i n a n c e have been attributed to bacterial lipids (Chaffee et al. 1986), while Disner & H a r o u n a (1994) assumed that such patterns represent a signature of the contribution of non-flowering plants to coal formation. In the heterogeneous lithotype, humoclarain, a preponderance of the o d d homologues is
221
registered, CPI = 1.6 (Table 3). The abundance of t/-C25, n-C27, and rt-C26 is high and the m a x i m u m is at n-C2s homologue, and is typical for a contribution of epicuticular waxes from higher plants (Chaffee et al. 1986; W a n g & Simoneit 1990). G C separations of the branched/cyclic alkanes are d o m i n a t e d by o~-phyllocladane (Fig. 2, peak 3,
Table 3. Maturity parameters Parameter
Lithotype
CPI* Pr/Ph 'bio'/N 'geo' hopanes ** C279/, Tm/Ts C29~o~/o~/~
C3~c~/3[S/(S + R)]
I
II
III
IV
0.99 0.33 1.13 6.43 2.86 1.41 0.17
0.97 1.12 0.44 3.02 1.71 0.66 0.53
0.50 --
1.65 0.51 1.05 6.60 3.67 1.20 0.20
* CPI = E % odd alkanes/E % even alkanes; ** E C29-C31/~/~/~ C29-C310!/~.
See Table 2 for Lithotypes. -
'
a]
j
I~,3
~,!
c)
i
!fl
3
I
I
ILl t~ Z
C~
~
0 O-
f
800
E~s~b9 i000
1200
i
t~ l.tJ er"
~O0
600
900
zOO:)
Z600
20O0 2200
Z~O0 2s~o 3OCO -~ZCO_.JOmO ~
600
iO0
zOO0 1.200 Z4O0 ~6 0 ~ aO
2000 ~ 22C0 ~ O0 2600 2SOO .~000 3~00
lad
bl
d]
l--
n~
6111 oo
s
Boo 1000 !200 1400 160~i 1800 2000 2200 2r
2~E.O2~3~ ~2CC .
---
:
o
560
~00 . ~1000 Z200 ZeO0 2~00 ~00~."~:o" .~_~ _ L ~14CO _ ~ _ -ZSO0 - ~ ' . .ZOO0 - ~ - - k2ZOO ~ , 2400 Z600 _
SCAN
Fig. 2. Total ion currance (TIC) of the branched/cyclic alkanes (a) humovitrain; (b) xylain; (e) liptain; (d) humoclarain; (numbers over peaks refer to Table 4). Pr, pristane; Ph, phytane.
Table
4.
Compounds' identified by mass spectrometry in hydrocarbon fractions
Peak
Composition
Compound designation
M+
Base peak
Structure*
1 2 3 4 5 6 7 8 9 l0 11 12 13 14 15
C19H34 C2oH32 C2oH34 C24H42 C27H46 C20H34 C15H26 C]5H26 C19H28 C2oH32 C19H34 C2oH36 C2oH34 C2oH32 C2oH34
norpimarane isophyllocladene c~-phyllocladane Des-A-lupane 17fl(H)-22,29,30-trisnorhopaane dihydrorimuene methyl, perhydro-phenanthrene cedrane 18-norabiatane-18,11,13-triene AS'l~ fichtelite sandracopimarane #-phyllocladane sandaracopimaradiene
262 272 274 330 370 274 206 206 256 272 262 276 274 272 274
233 120 123 123 149 259 191 82 241 257 109 247 123 257 123
I II III IV V VI VII VIII IX X XI XII
ent-beyerane
XIII XIV
* structures in Fig. 5.
1 12
Ph
! '56o
I
i
-
5gr
660
6~o
760
7So
s6o
830
960
9g0 logo loso
111o?"11'50
1~oo
o)
g_ 21
t
22
ILl
o
23
i,
Z
19
o
N
('
I:l..
22 20
I/1 ILl
,
o _B26 23 :
27
";C19
I / I101/+ 1.
I :
I
ll: b0 ' " ~o'ob ' iz'o6
i4o6
" " i6'oo
" t,'0o
2o'o5'
~2'ob
' ,,
~ o
q
o22
iJ_l >
21
b)
123
I,..,J
26
i,a_l re"
~_
22!
19
25 ~
I
, //
[L
....
5~0
' '
looo
15oc
--
"
2~,'o~
25co
"
2 ./'\
12 ~.
6 ./\
11//'~,\ .,,.....
13 .
3cac
SCAN
Fig. 3. Mass fragmentograms m/z 183 of branched/ cyclic alkanes (a) xylain; (b) humoclarain; 9 isoalkane; 0 , anteiso-alkane; Sq, squalane.
SCAN
Fig. 4. TIC of tri/tetracyclic diterpanes (a) humovitrain; (b) xylain; (e) liptain (numbers over peaks refer to Table 4).
BIOLOGICAL M A R K E R S IN M I O C E N E LIGNITE
I. n o r p i m a r a n e
II.imophyllocladene
III.o-phyllocladane
223
IV. D e m - A - l u p a n e
Me
V. 1 7 F ~ H ) - 2 2 , 29, 3 0 - VI. D i h y d r o r i m u e n e Tri s n o r h o p a n e
VII. M e t h y l , P e r h y d r o phenanthr ene
IX. 1 8 - N o r a b i a t a n e -
XI. F i c h t e l i t e
X. A~'IO-rimuene
VIII. C e d r a n e
XII. Eandara-
8, 1 1 , 1 3 - t r i e n e
coplm~rane
XIII. S a n d a r a c o p i mar a d i e n e
X I V. e n t - B e y e r a n e
XV. C - T r i c y c l i c 23triterpane
--
XVI.C24 17, 2 1 - s e c o hopane
EI X. 1 7 o (
H), 21 ~
H) -
Fx)mohopane. Csl
Fig. 5. Chemical structures cited.
X V I I 91 7 ~ ( H ) - 2 2 , ~ g , 3 0 trisnorhopane, C27
XX. 17F)( H ) , 21 ~ ( H ) -
homo~or et ane, C31 '~'-'
X V I I I . 18~x(H) -P2p 2 9 , 3 0 t r isnor neohopane, C27
X X I . 1 7 t ? ( H ) p 21 ~ ( H ) -
h o . . , ~ h o p a n e , C31 t?~:
224
M. STEFANOVA & C. MAGNIER
o) E27
tlJ I.,"I Z o
uq u/ 4
I,i >
C2~
12
C29
,~i
_.J ILl
660
860
1000
1200
~,~, L%~, ~~
~
1400
311~
1'60'0''i'80,0 2000
2200
2400
2600
28~00 3000
t
32100
b)
ILl Z o
a. u~
1,1.,I > l,,,-
._J r~
[27
I 10'00
12'00
1400
1600
1800
S C A N
2000
2200
-----
2400
2600
2800
BIOLOGICAL MARKERS IN MIOCENE LIGNITE
225
c)
C2 7
Vl Z 0 IX.
n-
c29~k L~
>
._J
SCAN Fig. 6. Mass fragmentograms m/z 191, key ion for tricyclic terpanes, 17,21-secohopanes, pentacyclic triterpanes (numbers over peaks refer to Table 4). O, tricyclic terpanes; O, 17,21-secohopanes; - - , 'geo' hopanes; 'bio' hopanes; (a) humovitrain; (b) xylain; (c) humoclarain.
Table 4, Str.III), already mentioned for the separations of the total extracts. ~-Phyllocladane is less common in humovitrain than in liptain in which it reaches 0.85 wt% of lithotype (Table 2). Traces from n-C16 to n-Cl8 alkanes occur in the sample of non-adducted liptain (Fig. 2c). Homologues are accompanied by the isoprenoids pristane and phytane. These also occur in xylain and humoclarain samples where iso- and anteiso-alkanes are indicated (Fig.3). Figure 4 illustrates the regions of tri/tetracyclic diterpanes. The mass spectral identifications are based on published spectra (Philp, 1985). In Table 4 some of the proposed structures, illustrated in Fig. 5, are gathered. The pimarane/abietane structures are tentatively assigned with a prevalence of pimarane compounds: norpimarane (Fig. 2, and Fig. 4, peak 1, Str.I); sandaracopimarane (Fig. 2c, d, Fig. 4a, c, peak 12, Str.XII); and unsaturated homologues, i.e. sandaracopimaradiene (Fig. 4b, peak 14, Str.XIII), dihydro-rimuene (Fig. 2b,
Fig. 4a, c, peak 6, Str.VI) and rimuene (Fig. 2c, Fig. 4b, peak 10, Str.X). Abietane structures are represented by fichtelite (Fig. 2c, d and Fig. 4a, b, c, peak 11, Str.XI) and 18-norabiatane-18,11,13-triene (Fig. 2c, peak 9, Str.IX). The preponderance of phyllocladane skeleton is obvious in all samples. Noble et al. (1986), in their study of tetracyclic diterpanes with entbeyerane, phyllocladane and e n t - k a u r a n e structures, noted that the ratios of C-16 epimers were dependent on maturity. Phyllocladane is a common constituent in low rank coals and is thought to be derived from phyllocladene (Str.II) which is a component in a number of conifers (ten Haven et al. 1992). Phyllocladane is well studied and correlated with maturity by Simoneit (1977), Alexander et al. (1987), Hazai et al. (1988), Hazai & Alexander (1991), Ten Haven et al. (1992), Disnar & Harouva (1994). The relative quantity of the thermodynamically preferred /3-phyllocladane increases with thermal maturity. It is registered in negligible
226
M. STEFANOVA & C. MAGNIER
Calculated values for Pr/Ph, C27~/c~and C29flc~/c~/3 in the case of humovitrain and humoclarain are similar but those for xylain differ noticeably, moreover the content of hopanes in xylain is negligible compared with to the other two lithotypes. In humovitrain, xylain and humoclarain a significant quantity of Des-A-lupane (peak 4, Fig. 2 and Fig. 6, Str.IV) is determined. This sesterterpenoid is considered to be a photochemical or photomimetric degradation product of certain 3-oxygenated triterpenoid precursors from higher plants (Wang & Simoneit 1990; Stout 1992). Des-A-lupane accompanied by an unknown series of C24H40 species, was recorded by the above cited authors in extracts from Tertiary coal. Its occurrence indicates input of material from higher (particularly dicotyledonous) plants. Coals with high proportions of terrestrial 9 tricyclic terpane (Str.XV), C22-C26 , with a input contain low quantities of steranes (Aref'ev maximum content of C23 homologue; 9 tetracyclic terpanes (17,21-secohopanes) et al. 1992). The m/z217 fragmentograms produced in this study do not indicate the (Str.XVI), C22-C24; and occurrence of steranes. 9 hopanes (Str.XVII-XXI), C27-C 31; The comparative study of the aliphatic comTriterpanes were not registered in the liptain pounds in the Maritza-Iztok lignite lithotypes bitumen sample, which is a peculiarity of the reveals some peculiarities of their biomarker lithotype. assemblage. A considerable difference was regisPentacyclic triterpanes are indicated in the tered in the amount of extractable bitumens, the m/z 191 fragmentograms (Fig. 6). Calculated highest content was determined for liptain, where maturity parameters are shown in Table 3. All it reached 40% and the lowest- for humodistributions are dominated by C27/3 hopane, vitrain, as low as 1.1%. The values determined trisnorhopane (Fig. 2, peak 5, Str.V). The ratios for xylain and humoclarain are comparable. A C27/3/a are >>1, an indication that C27 hopa- similar relationship was recognized for the noids are represented predominantly by the contents of saturates. The GC separations of all branch/cyclic biologically synthesized constituent. C29/3/3 hopanoid is the most abundant species, accom- alkane portions were dominated by the highly panied by the converted C29/3a isomer. Calcu- expressed contents of a-phyllocladane and lated moretane/hopane ratios (C29/3a/a/3) are amounted 0.85 wt% in the case of liptain. A prehigher in the humovitrain and humoclarain ponderance of the a-phyllocladane skeleton over samples and half as great in the xylain sample. pimaranes/and abietanes was registered for all Hopanes are divided into 'bio'- and 'geo'- samples. The variety of tricyclic diterpanes hopanes. Biohopanes have /3/3 stereochemistry. indicated that the generation biota included Increased maturity leads to the transformation gymnosperms. The presence of cedrane in liptain of these to the more thermodynamically stable bitumen was an additional proof for the family a/3 configuration of geohopanes (Tissot & Cupressaceae (order Coniferales). Welte, 1984, Ourisson et al. 1987, Aref'ev et Our results have shown that n-alkanes disal. 1992, Peters & Moldowan 1993) allowing the tribution depended on the type of lithotype. In stereochemistry of hopanes to serve as an the case of the homogeneous lithotypes, humoindicator of degree of maturation. The values vitrain and xylain, the CPI was close to 1.0 and higher than unity calculated for the ratios could be explained by bacterial lipids or nonE 'bio'/E 'geo' hopanes in the case of humovi- flowering plants contributions. For the heterotrain and humoclarain, supported by the high geneous lithotype, humoclarain, a prepondervalues for Tm/Tsand C31a/3[S/(S + R)] ratios are ance of odd homologues was registered. The taken as an unequivocal indication of a low other maturity parameters were closer for degree of geochemical transformation. Certain humovitrain and humoclarain, while according differences among the hopane distribution are to them xylain could be described as being at a reflected in the maturity parameters (Table 3). higher level of geochemical transformation.
quantities compared to the a-partner (Fig. 2c, Fig. 4a, c, peak 13). Ent-beyerane (Fig. 4c, peak 15, Str.XIV) coelutes with pimarane and contributes to the creation of the asymmetric peak. Its retention time is a bit less than the value for pimarane. The overlapping of the two compounds is obvious in Fig. 4c. The liptain sample contains a relatively large amount of cedrane (Fig. 2c, peak 8, Str.VIII). This sesquiterpenoid was found in the fossil resin retinellite, and in a montan wax by Grantham & Douglas (1980). The occurrence of cedrane suggests that the liptain was originally derived from plants of the family Cupressaceae (order Coniferale). Mass fragrnentograms m/z 191, key ion for tricyclic diterpanes and triterpanes reveal the following distributions of series (Fig. 6):
BIOLOGICAL MARKERS IN MIOCENE LIGNITE A lack of hopanes was registered for the liptain bitumen. In the other lithotypes the hopane pattern of distribution is similar, determined by the dominance of C27~ hopane and relatively high contents of/3/3 a n d / 3 a isomers. The prevalence of hopanes with <30 carbon atoms suggests C30-diploterol-like structures as a potential precursor of hopanes. Des-A-lupane was detected in all lithotypes, and may represent equivocal evidence for the input of angiospermous organic matter in lignite formation. The presence of 17,21secohopanes is supplementary proof for a terrestrial contribution. The presence of tricyclic triterpanes, C22-C25, is an indication for microbial activity. Tricyclics appear to be more stable thermally than pentacyclic triterpanes. Thus with increasing maturity the ratio of tri- to pentacyclic can be expected to rise (Waples & Machihara 1991). The lithotypes studied are at a low degree of maturation consistent with the negligible content of tricyclics compared with pentacyclic triterpanes.
Conclusions The aliphatic biomarker compositions of homogeneous and heterogeneous lithotypes of the 'Maritza-Iztock' lignite were studied by chromatographic and spectral methods. Some of the identified compounds confirm the previously described peculiarities of this deposit (Si~kov 1988). Resinites derived from conifers are frequently easily recognized in lignite by microscopic techniques. Our results confirm that forest and sedge-grass populations took part in the formation of the lignite and indicate aspects of the chemical composition of the plant debris. Aliphatic compounds, with precursors gymnosperms and angiosperms were detected in different proportions depending on the lithotype.
References ALEXANDER, G., HAZAI, I., GRIMALT, J. O. & ALBAIGES, J. 1987. Occurrence and transformation of phyllocladane in brown coal from Nograd Basin. Hungary. Geochimica et Cosmochimica Acta, 51, 2065-2073. AREF'EV, O. A., ZABRODINA,M. N., GULJAEVA,N. D. & PETROV, A1.A. 1992. Policiklicheskie biomarkeri tverdich kaustobiolitov. Khimia tverdogo topliva, Bdd 1, 12-35. CHAFFEE, A. L., HOOVER, D. S., JOHNS, R. B. & SCHWEIGHARDT, F. K. 1986. Biological markers extractable from coal. In: JOHNS, R. B. (ed.) Biological Markers in the Sedimentary Record. Elsevier, Amsterdam, 311-347.
227
DISNAR, J. R. & HAROUNA,M. 1994. Biological origin of tetracyclic diterpanes, n-alkanes and other biomarkers found in Lower Carboniferous Gondwana coals (Niger). Organic Geochemistry, 21, 143-152. GRANTHAN, P. J. & DOUGLAS,A. G. 1980. The nature and origin of sesquiterpenoids in some Tertiary fossil resins. Geochimica et Cosmochimica Acta, 44, 1801-1810. HAZAI, I. • ALEXANDER, G. 1991. Occurrence and transformation of tricyclic aromatic hydrocarbons in low rank coals. Fuel, 70, 971-978. , ESSIGER, B. & SZEKELY, f. 1988. dentification of aliphatic biological markers in brown coals. Fuel, 67, 973-982. MARKOVA, K., STOYANOVA, G. & PEEVA, N. 1993. Study of the effect of autoxidation processes on alkanes from low rank coal. Oxidation Communication, 16, 289 298. NOBLE, R. A., ALEXANDER,R., KAGI, R. I. & KNOX,J. 1986. Identification of some diterpenoid hydrocarbons in petroleum. Organic Geochemistry, 10, 825-829. OURISSON,G., ALBRECHT,P. & ROHMER,M. 1987. The hopanoids. Pure & Applied Chemistry, 51,709-729. PETERS, K. E. & MOLDOWAN,J. M. 1993. Terpanes. In: The Biomarker guide to interpreting molecular .fossils in petroleum and ancient sediments. Prentice-Hall, Inc. Englewood Cliffs, New Jersey, 142-264. PHILP, R. P. 1985. Diterpenoid hydrocarbons. In: Fossil fuel biomarkers. Applications and spectra. Elsevier, Amsterdam, 133-153. SIMONErT, B. R. T. 1977. Diterpenoid compounds and other lipids in deep-sea sediments and their geochemical significance. Geochimica et Cosmochimica Acta, 41,463-476. SIgKOV, G. 1988. Taxonomic composition of coalforming plants. In: Theoretical fundaments of biochemical coalification. University Press, Sofia, 29-35 (in Bulgarian). -& ANDREEV, A. P. 1987. A way to reconstruct coal-forming paleoplant communities based on the micropetrographic composition of Bulgarian neogene coals. Compt. rend Acad. Bulg. Sci., 40, 77-80. --, KEHAJOVA, M. & STAJKOVA, S. 1986. Brief geological description of coal deposits in Bulgaria. In: Coals and coal deposits in Bulgaria. A bibliography, University Press, Sofia, 19-32 (in Bulgarian). STEFANOVA, M., MAGNIER, C. & VELINOVA,D. 1995. Biomarker assemblage of some Miocene-aged Bulgarian lignite lithotypes. Organic Geochemistry, 23, 1067-1084. STOUT, S. 1992. AIiphatic and aromatic triterpenoid hydrocarbons in a Tertiary angiospermous lignite. Organic Geochemistry, 18, 51-66. TEN HAVEN, H. L., LITTKE, R. & RULLKOTTER, J. 1992. Hydrocarbon biological markers in Carboniferous coals of different maturities. In: MOLDOWAN, J. M., ALBRECHT,P. & PnluP, R. P. (eds) Biological Markers in Sediments and Petroleum. Prentice-Hall, Englewood Cliffs, New Jersey, 142-155. i
228
M. STEFANOVA & C. MAGNIER
TtssoT, B. P. & WELTE, D. 1984. Petroleum Formation and Occurrence. 2nd edn., Springer, Berlin. VELINOVA, D., LAZAROV,L. • ANGELOVA,G. 1993. N.M.R. investigation of low-rank coal lithotypes pretreated with Nail. Fuel, 72, 991-996. WANG, T-G. & SIMONEIT, B. R. T. 1990. Organic geochemistry and coal petrology of Tertiary
brown coal in the Zhoujing mine Baise Basin, South China 2. Biomarker assemblage and significance. Fuel, 69, 12-20. WAPLES, D. W. & MACHIHARA,T. 1991. Biomarkers as organic-facies indicators. In: Biomarkers for Geologists, American Association of Petroleum Geologists, Tusla, Oklahoma, 41-63.
Floristic characters of the upper coal-bearing formation in the Transcarpathians SVETILANA
SYABRYAJ
Institute of Geological Sciences NASU, Chkalova srt. 55b, 252054 Kiev, Ukraine Abstract: The results of palynological studies of samples from outcrops, boreholes and
mines in the Transcarpathians are presented. Megafloral remains from the Ilnitsa suite show the age of the upper coal-bearing formation to be Romanian. The palynological analysis allows the reconstruction of changes in the plant cover during the Romanian, and the determination of the climatic and palaeogeographic conditions. The Upper coal-bearing formation of the Transcarpathians belongs to the Ilnitsa Suite of Late Pliocene age and consists of alternating aleurolites, coaly clays, tuffites and five coal-beds. The clays and tuffites are characterized by ostracodes, some molluscs, leaf remains, wood, seeds and fruits. The megaflora remains were studied by
Iljinskaja (1968). Palynological studies on some sections in the Neogene deposits of the Transcarpathians have been published previously (Chmarsky 1954; Shchekina 1960; Rybakova 1964, 1968, 1975). The present author studied 13 sections from Ilnitsa, Rokosovo, Gorbki, Berezinca Veliky Racovets, Uzhgorod (Syabryaj
POLAND
SLOVAKIA
%
9Uzhgorod 5
Irshava ~
1
._______J
HUNGARY
Rakhov ~
ROMANIA
Fig. 1. Location of boreholes and outcrops in the Transcarpathians; 1, Ilnitsa; 2, Veliky Rakovets; 3, Rokosovo; 4, Berezinka; 5, Vuzhgorod; 6, Gorbki.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 229-236.
230
S. SYABRYAJ
1969, 1975, 1986) (Fig. 1). Sediment samples were taken from numerous boreholes, outcrops and mines. The Ilnitsa section is the most complete. It contains five coal seams with thicknesses as follows: 0.11-5.0m; 0.1-2.1m; 0.25-1.4m; 0.41.5m; 0.25-1.3m. Other localities contain one or two of the coal beds seen at Ilnitsa. The Ilnitsa section was used as a standard.
Palynomorph records The Upper coal formation contains two coal horizons. The lower one comprises the fourth and fifth layers and the other three beds are in the upper horizon. Two different palynomorph assemblages were obtained. Both show a high proportion of angiosperm pollen but there is also an important component of coniferous pollen, mainly belonging to the Taxodiaceae. The first assemblage is from the fourth and fifth seams. The pollen diagram (Fig. 2) presents a picture of forest vegetation with a high proportion of deciduous angiosperm trees: Carya, Juglans, Pterocarya, Engelhardtia, Ulmus, Celtis, Zelkova, Quercus, Fagus, Castanea, Liquidambar, Alnus, Parrotia, Carpinus, Nyssa, Magnolia. There is a lower proportion of shrubs which are represented mainly by Myrica, Moraceae, Rhus, Corylus, Cornaceae, Caprifoliaceae and Rosaceae. Coniferous tree pollen is also abundant. These are identified as Pinus (predominantly P.diploxylon), Picea, Abies, Cedrus, Keteleeria (a single find) and Tsuga. The Taxodiaceae are represented by Taxodium, Glyptostrobus and Sequoia (the latter beeng rare). Single finds of Ginkgo, Sciadopitys and Podocarpus pollen are reported as well. As for the herbaceous plants, the representatives of hydro- and hygrophytic group are the most abundant. They belong to the Cyperaceae, Hydroharitaceae, Sparganiaceae, Nymphaeaceae, Typhaceae, Potamogetonaceae, and Gramineae. In the coal bands of the fourth seam Rynchospora (Cyperaceae), was found to be abundant. This genus is characteristic of mesotrophic peat-bogs. Just in these spectra, Sphagnum is most abundant. The pollen of other herbaceous plants (Plantaginaceae, Polygonaceae, Compositae, Chenopodiaceae, Leguminosae, Ranunculaceae) are relatively rare. Spore finds are attributed to the ferns Osmunda, Polypodium, Dryopteris, Onoclea. There are also many fungi and diatoms. Most of these algae are inhabitants of freshwater basins (Vodop'jan 1979).
The second assemblage corresponds to the first, second and third coal seams. The pollen spectrum of these deposits is rather uniform, and the pollen diagram (Fig. 2) presents a picture of forest vegetation. Deciduous tree pollen is dominant, but their proportion has diminished. They are represented by Carya (the most abundant), Juglans, Pterocarya, Quercus, Fagus, Castanea, the latter being single finds in only a few samples. Quercus pollen is more abundant and more varied in this assemblage than in the lower one. Three species were identified: Q. cf. pubescens Will.; Q. cf. petrea Liebl.; Q. cf. robur L. Ulmus and Zelkova were discovered in the same quantity as in the first assemblage, whilst Celtis decreased. No significant change was noted in the composition of the Betulaceae. Liquidambar incieased proportionaly and became more diverse, as also happened with Tilia, which is represented by three species: T. cf. tomentosa Moench., T. cf. cordata Mill., and T. platyphylla Scop. The pollen of thermophile plants are rare. Single finds of Platycarya, Engelhardtia, Magnoliaceae, Aralia, Ilex, Nyssa are recorded. Coniferous tree pollen is more abundant than in the first assemblage: Pinus, Picea, Tsuga (T. cf. canadensis (L.) Carr., T. cf. diversifolia (Maxim.) Mast.), Abies (A. cf alba Mill.). Keteleeria is absent and Cedrus, Podocarpus, and Ginkgo are represented with single finds. The Taxodiaceae are represented by Taxodium, Glyptostrobus, and Sequoia. The first is the most abundant and the last is a single find. The proportion of sporomorphs of shrubs is small and they are chiefly represented by Rhus, Cornaceae, Celastraceae, Caprifoliacae, Elaeagnaceae, Rosaceae, Myrica. Herbaceous forms are represented by numerous hydroand hygrophile taxa belonging to the Gramineae, Labiatae, Caryophyllaceae, Plantaginaceae, Ranunculaceae, Gentianaceae, Umbelliferae etc. Abundant and diverse polypodiacean spores are found; their number increases in the upper part of the formation. The spores of Osmunda, Equisetum, Salvinia, Lycopodium, Sphagnals, Bryales are also recorded. Lusatisporis punctatus Krutzsch and L. perinatus Krutzsch were found only in bands of the second coal seam. They appear to characterize this coal seam.
Discussion Analysing the composition of the two palynocomplexes we note that the first assemblage is the more varied one. The second complex
PALYNOLOGICAL STUDIES OF THE TRANSCARPATHIANS
231
~J
,s II.J.d
~
I
.
.
.
T
.
~
.,
m.
~
.
m ~
ira..,.,
d-
.
.n
~ .
Z . ,,..~
B
~,..d=.-"
.,
'..
o
E
~ i
II.,.,,!
r
I
'
I
, h
.--I
~
8 8
8
[
,
,x, , - - i
I
II
L,
--=
o
i
.ll,--.d
n,,
d,n
~
N
~
q ~a_ilIM
9
I.
I=
.d
I
I
'
=~
=eL.
9
I
'
..
dll
i
'
=! ~}
,.,ILl.
.
,,,
z
I,.
I!i 9
8
~
8
o ,,.'-r
,,,-.,
~io
.. . . . , . .
,
.-.
~
I Ir~l,;:,i!i _
=--
..
>
_
=_-_.
~
== >_.
>
~=
,-4 E
232
S. SYABRYAJ
Table 1. Comparison of the composition of mega- and microfloral remains from the Upper Pliocene of the
Transcarpathians Pollen and spores
Megaflora
Taxa
Number of palynomorphs
Taxa
Number of megafloral remains
1
2
3
4
Ilnitsa, mine N 1 Polypodiaceae Osmunda Pinus Pinus cf. sylvestris P. tertiaria P. cf. strobus P. cf. Baileyana Picea Picea media
2 1 49 1 1 1 5 6 1
Taxodiaceae Glyptostrobus
4
Sciadopitys Poaceae
1 2
Liliaceae Juglandaceae Juglans J. cf. cinerea Carya C. cf. aquatica Alnus Carpinuas Zeilkova Ulmus Celtis Liquidambar Tilia cordata
Cornaceae Rosaceae Oleaceae (Fraxinus) Ericaceae Gentianaceae Veliky Rakovaets Pinus Picea Abies Tsuga
2 2 3 1 3 2 21 2
Equisetum parlatorii Dryopteris denticulata D. linneaneiformis Osmunda heeri Pinus (wood)
Cupressinoxylon Taxodium dubium Glyptostrobus europaeus Sequoia langsdorfii Phragmites oeningensis Cyperacites zollikoferi
1 7 1 5
4 29 12 4
1
Betula Quercus pontica miocenica
10
Acer Cercediphyllum crenatum Alangium tiliaefolium
21 2 16
Fraxinus paviifolia
18
1
1 6 2 1 2
1 3 1 1 2
Loranthus
1
Taxodium dubium
1
Arundo anomale Phragmites oeningensis
1 6
Cyperus Phyllites cf. Myrica
1 1
35 10 6 2
Sciadopitys
2
Poaceae (cf. Phragmites) Potamogeton Sparganium
2 1 3
P A L Y N O L O G I C A L STUDIES OF T H E T R A N S C A R P A T H I A N S Table 1.
233
(continued)
Pollen and spores Taxa
Number of palynomorphs 2
1
Veliky Rakovats
Megaflora Taxa 3
Number of megafloral remains 4
(continued) Salix rozmarinifolius S. varians
Juglans cf. regia Juglans Carya C. cf. elegans Pterocarya Betula Alnus Corylus Fagus Quercus cf. Castanea Celtis Ulmus Moraceae Liquidambar Elaeagnus Tilia cf. cordata Tilia
1
5 15 3 5 3 22 2 17 3 4 9 9 2 3 2 2
Carpolites cf. Carya Pterocarya paradisiaca
Aesculus hippocastanea
34
1
Acer integerrimum A. subcampestre A. trilobatum A. campestre Cedrella sarmatica Alangium tiliaefolium Fraxinus paviifolia
4 27 17 3 2 234 53
Rosaceae Nymphaeaceae Labiatae Ranunculaceae Urticaceae Lythraceae Franceniaceae Primulaceae Rubiaceae Solanaceae Berezinka Polypodiaceae Polypodium Cyatheaceae Pinus Picea Larix Tsuga Taxodiaceae Taxodium Poaceae
Salix
4 1
2 44 3 1
2 4 2 1
Glyptostrobus europaeus Phragmites oeningensis Typha latissima Cyperus reticulata Arundo goepperti Salix varians S. subaurita Salix Populus rhamnifolia
5 1
2 4 25 6 5 2
234 Table 1.
S. SYABRYAJ
(continued)
Pollen and spores Taxa
Megaflora
1
Number of palynomorphs 2
Berezinka (continued) Juglans Carya Alnus Corylus Betula Quereus Fagus Castanea Ulmus Moraceae Nymphaeaceae Liquidambar
1 2 4 2 4 2 13 3 5 3 1 1
Nyssa Rhus
Oleaceae Fraxinus cf. ornus Rosaceae Legominiosae Ranunculaceae Valerianaceae Thymeliaceae
Taxa 3
Number of megafloral remains 4
Quercus pontica-miocenica
6
Castanea atavia Ulmus longifolia
3 3
Nelumbo protospeciosa Liquidambar europea Platanus aceroides
8 3 16
Acer aegopodifolium A. hungaricum A. sancta-crucis A. trilobatum Acer cf. sinense cf. A. pseudoplatanus Trapa transcarpatica Alangium tiliaefolium
2 2 6 10 2 8 13 6
1 1
2 1 2 1 2 1 1
contains a lesser quantity of thermophile plants and more abundant hydrophile angiosperms, as well as dark coniferous elements and more numerous herbaceous taxa. It is necessary to concentrate on the third coal seam, which is poorer than the lower and upper ones. It does not contain thermophile elements, including light-dependent ones like Celtis and it shows fewer sporomorphs of angiosperm pollen whilst Taxodium and Osmunda are abundant (Fig. 2). The third seam may be regarded as the boundary between the lower and upper horizons, but its palynological composition is more akin to that of the upper horizon. The spectra of the first, second and third coal seams are to be considered togather. The spectrum from the coal seam of Gorbki, the upper seam of Uzhgorod, and several small seams from Rocosovo are identical to the second palynocomplex. The composition of taxa and their numerical content is an argument in favour
of correlating these coal seams with the upper horizon. The presence of Luzatisporis punctatus and L.perinatus in the Gorbki coal bed allows its correlation with the second seam of Ilnitsa. The palynological contents of the Ilnitsa Suit can be compared with megafloral remains from the Berezinca, Ilnitsa, and Veliky Racovets localities. This comparison between the palynological and megafloral composition highlights similarities on the generic level. However, the palynological contents are more diverse (Table 1). The general picture of vegetation throughout the coal bearing formation shows generally a rather monotonous warm-temperate deciduous forest cover with some subtropical elements. However, coniferous forests existed in the Trancarpathians as well as in marshy conditions as is shown by the presence of Taxodium, Nyssa, Alnus and Osmunda which were found to be widespread in some of the wetter areas. Sphagnales were important in the swamps, especially during the
PALYNOLOGICAL STUDIES OF THE TRANSCARPATHIANS accumulation of the fourth coal seam. Some species of pine also inhabited swampy grounds, as can be seen in the contemporary Mshana swamp in the Transcarpathians. Brush swamps with Myrica and Salix invaded the outer parts of marshes. Wet forests with Carya, Liquidambar, Pterocarya, Alnus, Acer, and Ulmus occured in the vicinity. Mixed and deciduous forests with Pinus, Abies, Cedrus, Tsuga, Betula, Fagus, Quercus and Castanea developed in the drier habitats on higher areas. The climate was warm-temperate, rather humid. Only when the third coal seam accumulated did the broad-leaved forests become more impoverished. The third coal seam overlies a thick volcanic ash band and the coal contains much tuffaceous material. Volcanic activity taking place immediately before the third coal bed accumulated, evidently influenced the floral composition. Some thermophile and lightdependent elements (such as Celtis) vanished from the plant cover. However the volcanicity also had a positive influence. The broad-leaved forests on the Vigorlat-Gutin mountain ridge probably expanded due to the volcanic products accumulating at the foot of the volcanoes (Syabryaj 1991). Such processes can be observed today in areas of modern volcanic activity, e.g. Kamchatka and Indonesia. The enrichment of plant communities is observed during the period of formation of the second and first coal seams, when volcanism diminished. Taxa such as Castanea, Platycarya, Engelhardtia, Magnoliaceae and the lightdependent Celtis reappeared in the deciduous forests, and the area of Quercus communities increased. Modern beech forests are of importance in separate regions of Carpathians and on the Vigorlat-Gutin mountain ridge, where the rich original beech forests are preserved today as forest reservations. Only the marsh forests remained unchanged during the period of Upper coal-bearing formation. At the end of the Pliocene the flora became poorer, but there were no fundamental changes in plant cover and its structure. The Late Pliocene flora of the Transcarpathians was warm-temperate, essentialy wet, with predominantly thermophile deciduous forest elements, producing swamp forests. Coniferous and mixed forests occurred in the mountain areas.
Conclusions With regard to the floral composition, the entire period of coal formation of the lower and upper
235
horizons shows marked similarities. The entire interval is referred to the Romanian. The climate was warm-temperate, rather humid with a tendency to become cooler prior to the Gunz glaciation. The flora from the coaly sediments of the Ilnitsa suite formed different communities similar to the ones of the same age in neighbouring areas. Thanks are due to Dr. I. Iljinskaja for kindly providing the samples of the Ilnitsa Suite with megafloral remains from the Berezinka, Ilnitsa and Veliky
Rakovets localities. I am grateful to the anonymous reviewers for suggesting several corrections.
References CHMARSKY, N. Z. 1954. Some data about fossil flora of the Tertiary of the Transcarpathians. In: Scientific notes of Dnepropetrovsk Institute. Collected works of geol.-geogr, faculty and Institute of Geol., 39, 127-132 (in Russian). ILJINSKAJA, I. A. 1968. The Neogene flora of the Transcarpathian region of the UkrSSR. Publ. house 'Nauka', Leningrad (in Russian). RYBAKOVA, N. O. 1964. The new data about the Upper Neogene flora of the Transcarpathian region of UkrSSR (by palynological study). Bull. Moscow Society of test of nature, Geol. Series, 64, 241-245 (in Russian). - - 1 9 7 5 . Palynological description of the Miocene and Pliocene deposits in the Transcarpathians (UkrSSR). Paleontol. collect., 1-2, N 12, 142-147 (in Russian). SCHEKINA, N. O. 1960. History of the Neogene flora of the Ukrainian Carpathians and Precarpathians. Flora and fauna of the Carpathians. Publ. house AS USSR, 58-74 (in Russian). SYABRYAJ,V. T., LEVITSKY,B. P., SYABRYAJ,S. V. & EMETS, T. P. 1969. Substantial and palynological composition of coal-bearing formation of the geosinclinal part of UkrSSR. Publ.house 'Naukova dumka', Kiev (in Russian). SYABRYAJ,S. V. 1975. Description of the Levantinian
flora and vegetation of the Transcarpathians. In: Flora. Taxonomy and Phylogeny of plants, 279-288 (in Russian). - - 1 9 8 6 . Evolution of the Neogenian flora and vegetation of the Carpathians. Doct. hab.thesis (in Russian). - - 1 9 9 1 . Vegetation and volcanism in the Neogene of the Transcarpathians. Palaeovegetational development of Europe and region relevant to its palaeofloristic evolution. Proc. PEPC. Vienna, 231-234. VODOP'JAN,N. S. 1979. The diatoms in the Pliocenian deposits of the Transcarpathians. Ukr. Botan.journ. 35, 141-146 (in Ukrainian).
The distribution of sulphur in the Palaeocene coals of the Sindh Province of Pakistan S. R. H. B A Q R I
Pakistan M u s e u m o f Natural History, Garden Avenue, Shakarpyran, lslamabad, Pakistan
Abstract: 46 samples of coals were collected, representing eastern and western coalfields of the Sindh Province of Pakistan to investigate any systematic quantitative changes in total sulphur. The coal is Palaeocene in age and the number and thickness of coal seams decreases from east to west. The sulphur content was determined as total sulphur on an elemental analyser Carlo Erba model EA 1108 on a dry basis. The sulphur content displays systematic variations and increases gradually from east to west. It is about 1.44% in the eastern coalfields at Tharparkar, 2.55% in the central coalfields at Badin and 4.95% in the western coalfields at Lakhra. The Palaeocene coalfields of the Sindh Province were deposited in shallow continental lagoonal areas towards Tharparkar in the east and comparatively down dip deeper areas with brackish waters (deltas, estuaries) towards the Lakhra in the west. The distribution of sulphur displays a good theoretical example in the change of depositional environment from fresh water to brackish waters environments. It is concluded that the coalfields with low sulphur occur in the east as compared to the coalfields with high sulphur in the west.
The Palaeocene coals of the Sindh Province of Pakistan occur in the Bara Formation and have been discovered from east to west at Tharparkar, Badin and Lakhra (Fig. 1). Stratigraphically, coal bearing Palaeocene sediments in the Sindh area have been studied by several authors (Shah 1977). Frederiksen et al. (198990: in Shah et al. 1992) named the Palaeocene rocks, from base to top the Khadro Formation, the Bara Formation, the Lakhra Formation and the Sonahri Formation (Fig. 2). The coal bearing Bara Formation consists of shales, sandstones, marls and coal bends in the western side at Lakhra, while it consists of clays and sandstones of probably fresh water origin in the east at Tharparkar. The presence of marls and lenticular argillaceous limestone bands at Lakhra in the west (Fig. 2) reflect comparatively more saline and deeper water conditions as compared to the east at Tharparkar. There are about 11 coal seams in the Tharparker coalfield that range in thickness from 1.42 to 27 m. In addition, there are numerous thin coal partings intercalated with the shales. There are three workable coal seams in the Badin area and the thickness varies from 0.55 to 3.1m (Khan et al. 1992). Three coal seams have been reported from the Lakhra Coalfield which range in thickness from 0.3 to 3.00m. The number and total thickness of coal seams generally decrease from east to west, indicating
a source of the coal vegetation towards east where the Precambrian granitic basement rocks are also exposed at Nagarparkar. The Bara Formation is unconformably overlain by Pleistocene conglomerates and sandstones of probable fresh water origin in the eastern coalfields (Tharparkar and Badin coalfields) and conformably overlain by Eocene marls and clays of probable bracklish water to marine origin in the western coalfields, located in the Lakhra area. Figure 2 gives a generalized section of the Palaeocene rocks in the Lakhra Coalfield of the Sindh Province. The depths of the respective samples are provided in Table 1 and the details about the locations drilling cores lithology etc are provided by Shah et al. (1992). 46 coal samples were collected to investigate the total distribution of sulphur and to understand the palaeo-environments of the depositional basin. The sulphur content was determined as total sulphur in an elemental analyser carlo Erba model EA 1108 on dry basis.
Distribution of sulphur Figures 3 and 4 and Table 1 give the distribution of total sulphur in two boreholes (TP-3, TP-4) in the Tharparkar area and in some coals from the Badin and the Lakhra coalfields. The sulphur in most of the samples is less than 1%
From Gayer, R. & Pegek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 237-243.
238
S. R. H. BAQRI
' 72"
I
6,~"
L
U.S.S. R
/
AFGHANISTAN
NW R#, - ~ ::)NTIE~ ~ ~'\'~:~3) ~,--~
~32"
JAMMU a KASHMIR
/ ,~'.~s ,6 ,7 ( PAN JAB
~1 ~9 ,7
/ , ~
BALUCHISTAN
/
~ --.'~-.
//
Lf.,,.,..
(SINDH. /
IRAN
INDIA
24"...-
4 /~ 4 ~ ' / 4 /r ,-e~-.4
0 } 0
200 krn I.,
TP4
~
J~
1
e.~. 200mi I
Fig. 1. The location map of the coalfields of Pakistan. The circles display the locations of the samples collected for the present work: 1, Indus East Badin; 4, Lakhra; 26, Tharparkar.
with the exception of eight samples and indicates the lowest concentration of sulphur in Pakistani coals. The average concentration of sulphur in 40 samples of the two boreholes in the Tharparkar coalfield is about 1.44% and is the lowest of all coalfields of Pakistan (Baqri 1993). Figure 4 and Table 1 show that the sulphur in most of the samples (with the exception of one sample) from borehole TP-3 remains even less than 1% and
indicates a good prospect for an area of low sulphur coal. The distribution of sulphur with depth in bore hole TP-3 does not show any systematic variation and probably reflects the uniformity of the processes of deposition or diagenesis. The distribution of sulphur in bore hole TP-4 is interesting and is not uniform. It does not show any systematic variation with depth but appears as a random distribution and
S U L P H U R IN P A L A E O C E N E COALS, P A K I S T A N
MEMBER
>-
AND
U
o I.d
_ ~.m
o :z:
COAL ZONE
EXPLANATION
I,=-
OUJ
..=Z
AI luvi urn
9
'LI 9 i
:----;. . . . .
i
LIJ r,.)
Shale and ctay stone ;:.~:: Sandy shale .. ; ... _.:.-, Siltstonr and sift Shaly sandstont Sandstone sctr~d~ Coat Undorctay
100
Meting Shale
__ %--
Meting Limr
TM'sohnaricoat zone
i
0 W EL
ZOO
;~'i'~-L':~
300
t
Coat zone W Z Ld
Upper Dhaduri Ohaduri Upper stra y~ z o Inyatabad ~ i Upper Sonda I Lower Sonda Upper Wassi ~ Wassi ~[
n,.
Ru bbty time stone. Even be dded limestone Calcareous shale Calcar~ou s siltstone Calcareous sandstone
Laki Limestone
i~ :z:
239
:-..'-'-- ..... ~
,
C?'."" ~'-" . . . . . . . 4100
.,,~r..4~.~ ..... ,
":
9 .,,
Lower s t r a y .--
..
50O
7-::. . . . .
~-_
9 ." . . ' . . . - . ".',-...
SCALE IN NETERS
Fig. 2. The stratigraphic section of the Palaeocene/Eocene rocks from the Lakhra Coalfield in Sindh Province
240
S. R. H. BAQRI
Table 1. The weight percentage of sulphur (wt%) in Palaeocene coal samples with respective borehole numbers and depths in metres. Tharparkar (TP4, TP3) Badin (BN1) and Lakhra (LS4, SOs)
Sample
Sample No and depth in metres
Sulphur (wt%)
1. 2. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
LS4, 156.97-157.64 LS4, 190.98-191.28 BN1, 0.36m Bara Fm. BN1, 101.7 BN1, 122.85-123.31 TP4, 180.91-181.25 TP4, 181.25-181.6 TP4, 181.6-181.93 TP4, 192.05-193.02 TP4, 193.2-193.58 TP4, 193.58-194.31 TP4, 194.51-195.16 TP4, 200.68-201.12 TP4, 203.45-203.75 TP4, 205.73-206.08 TP4, 223.78-224.66 TP4, 229.01-229.71 TP4, 229.71-230.18 TP4, 230.18-230.73 TP3, 148.6 TP3, 153.92-154.22 TP3, 154.22 TP3, 155.7-156.72 TP3, 156.72-156.97 TP3, 157.64-158.24 TP3, 158.34-158.94 TP3, 158.52-160.02 TP3, 158.94-160.02 TP3, 161.52-162.42 TP3, 162.42-163.07 TP3, 163.07-162.42 TP3, 164.30-164.75 TP3, 164.75-165.52 TP3, 165.52-166.11 TP3, 166.11-167.66 TP3, 167.66-168.21 TP3, 168.21-169.21 TP3, 169.21-169.90 TP3, 169.90-170.72 TP3, 170.73-171.83 TP3, 171.38-172.26 TP3, 172.26-172.83 TP3, 172.93-173.83 TP3, 173.83-175.23 TP3, 197.25-198.35
0.92 6.27 1.65 5.29 0.73 3.53 7.81 16.94 0.43 0.43 0.37 0.42 2.97 2.93 0.56 0.3 0.44 1.36 2.88 0.82 0.71 1.36 0.85 0.71 0.84 0.69 0.72 0.57 0.62 0.69 0.49 0.66 0.59 0.54 0.1 0.51 0.59 0.45 0.2 0.69 0.71 0.38 0.58 0.7 0.76
therefore reflects probably variations during the depositional and diagenetic history. Seven samples contain less than 1% sulphur, five samples display sulphur between 1-4% and only two samples display more than 7% sulphur (Table 1).
Table 2 provides the average distribution of sulphur in the Badin and Lakhra coalfields of the Sindh Province. The sulphur in the Badin coalfield is lower (2.55%) than the Lakhra (4.95%). The low sulphur in the Badin coalfield is due to leaching of the sulphate ions through meteoric waters, poor in sulphate ions, most likely fresh/brackish waters as the Bara Formation with intercalating coal seams is overlain by conglomerates and sandstones of Pleistocene age.
Discussion
Querol et al. (1991) studied the total sulphur in coals of Teruel District, Spain and concluded that the distribution of sulphur in the coals is influenced by the depositional and diagenetic environments. Querol et al. (1989) carried out detailed investigations on the iron sulphide precipitation sequence in Albian coals from the Maestrazgo basin, southeastern Iberian Range, northeastern Spain. They proposed five stages of iron sulphide precipitation which were controlled by the coalification evolution. The lowest concentration of sulphur in the coalfields is generally due to either their deposition in the proximal part of the depositional basin (fresh water ponds, Fig. 5) or due to the leaching of the sulphate ions during early or late diagenetic stage through the waters poor in sulphate ions. It is most likely that the eastern coalfields of the Sindh Province experienced both processes. The eastern coalfields were likely deposited in the proximal part (fresh water, Fig. 4) of the Palaeocene depositional basin, shallower in the east and comparatively deeper towards the west (representing brackish to delta front environments). The basement rocks are still exposed further east at Nagarparkar in Tharparkar district. The second cause for the low sulphur concentration in the eastern coalfields (Tharparkar area and Badin area) is due to the leaching of the sulphate ions by meteoric waters from the overlying Pleistocene conglomerates and sandstones of probably fresh/brackish water origin as compared to the western coalfields at Lakhra where the Palaeocene rocks are conformably overlain by Eocene marls, clays and limestones of marine origin. Figures 3 and 4 display the increasing distribution of sulphur from east to west and indicate a basin with its proximal end towards the east (Tharparkar) and the distal end towards the west where more sulphate rich waters acted during the deposition of the Palaeocene coal field at Lakhra.
S U L P H U R IN P A L A E O C E N E COALS, P A K I S T A N
Distribution
241
of SuLphur
5
;"4
~3 m
2
0 WEST
LAKHRA
BADIN
T H A R P A R K A R EAST
Fig. 3. The average distribution of sulphur in the Sindh coals from east to west (Table 2).
Distribution of sulphur in Sindh coots I sulphur% ) 17 16-
1513 12 I1 tO 8 8
7 G 5 /,.
3 2 1 0
12 3 4 5 6 7 8 9 101112131/,1516171M92021ZZ?..~/,2r WEST
3~f33435363"/N 4243/,4/,S&6 EAS~I
Seriol no of somples(Tobte | ) . Fig. 4. The distribution of sulphur in the individual samples of coals from the coalfield of the Sindh Province.
242
S. R. H. BAQRI
Table 2. The distribution of sulphur in the Palaeocene coals of the Sindh Province Name of coalfield
Average sulphur (wt%)
SD
SE
No of samples
Lakhra Badin Tharparkar
4.95 2.55 1.44
3.55 2.41 2.85
2.05 1.39 0.45
3 3 40
3. The sulphur in the eastern coalfields or the Sindh Privince is low and therefore the eastern coalfields are the g o o d prospects for low sulphur coals and m a y be exploited for industrial purposes. I am grateful to the EEC for providing a financial grant as a post-doctorate Marie Curie Senior Research Fellowship for the carrying out of these studies. I wish to thank Professor Dr N. Hamilton, University of Southampton, for allowing me to conduct these studies in the Department of Geology, I am also grateful to Dr Bashir Ahmad Sheikh, Chairman of the Pakistan Science Foundation, Islamabad, for his encouragement and granting permission for me to carry out these studies in the UK. I am thankful to Dr Shahzad A. Mufti, Director General, Pakistan Museum of Natural History, who encouraged the publication of this work. I am grateful to Mr. Sher Akbari for his help in the laboratory. Finally, I pay my regards to Dr John Marshall for his keen interest, supervision and encouragement of these studies, and to Mr Abbas Ali Shah, Director, Geological Survey of Pakistan for providing the borehole samples for these studies.
SD, Standard deviation; SE, Standard Error.
Summaryand conclusions 1. The sulphur content in Palaeocene coals of the Sindh Province increases from east to west. It is 1 . 4 4 w t % at T h a r p a r a k a r , 2 . 5 5 w t % at Badin and 4.95 w t % at L a k h r a . 2. The Palaeocene coals were deposited in fresh water p o n d s towards the east and in comparatively m o r e saline brackish water conditions towards the west.
17 3
-~ wFre.ph water amp
Tidal flat= klangrovr swarr~Fre~ , and . _..and water Delta Front_ _ cnannets lid61- channel swam p
.f
/
~
:r
,Fresh water awamp ~" ~'
~ .'-
-
/
ki~,. . 11 -
H i g h w"a........... ter ........ :"
~ """'-'-"'-'~-'" ,
1-2 3 /. 5 6 West
=-
. . . . . .
7~
...
"
" ~.~ ~.'-_"~-'~_~.~. ~ .
:"
"~'~:
9
Z:~.:~'r..:~,= " - ' i - : - ".~.~'~." C LAY
"~B A S E M E N T .
. . . .
j
R 0
C K S
~
!1 l '
m-I i--am mm I Immmimmii=mmmm_-m._lm=mil ' mIO1112131/~lS16t711H920212223Zt, Z5262"/2~Z93031 32.~3635.,t6373839/.O41~&3~St,6 East
Serial no of sarnples
( T a b l e 1)
Fig. 5. The palaeoenvironments of depositional basin and the distribution of sulphur in the Palaeocene coals of the Sindh Province of Pakistan.
S U L P H U R IN P A L A E O C E N E COALS, P A K I S T A N
References BAQRI, S. R. H. 1993. Research progress report for Marie Curie Bursary, No. B/Cl1"-923195, University of Southampton, UK. KHAN, S. A., KHAN, I. A., ABBAS, S. G. & KHAN, A. L. 1992. Coal Resources Potential of Pakistan. Information Release 533, Geological Survey of Pakistan. QUEROL, X., CHIUCHONS,S. • LOPEZSOLER,A. 1989. Iron Sulphide precipitation sequence in Albian coals from the Maestrazgo basin, Southeastern Iberian Range, northeastern Spain. International Journal of Coal Geology, 11, 171-189.
243
, FERUANDEZ-TURIEL J. L., LOPEZ-SOLER, A., HAGEMANN, H. W., DEHMER, J., JUAN, R. & Ruiz, C. 1991. Distribution of Sulphur in coals of the Termel mining district, Spain. International Journal of Coal Geology, 18, 327-346. SHAH, A. A., KHAN, S. A., TAGER,M. A., CHANDIO, A. H. & LASHARI, G. S. 1992. Drilling and Coal Resources Assessment in Southern Sindh, Pakistan. Information Release 537, Geological Survey of Pakistan. SHAH, S. M. I. (ed.) 1977. Stratigraphy of Pakistan, Memoir Geology Survey of Pakistan, 12, 138.
Sulphur distribution in a multi-bed seam PAUL
F. C A V E N D E R
& D. A L A N
SPEARS
Department of Earth Sciences, University of Sheffield, Sheffield $3 7HF, UK Abstract: The Parkgate Seam (Langsettian) is an important seam for the coal industry in the UK. It extends over a considerable part of the East Pennine Coalfield and a large database of information on the seam has been complied by (the former) British Coal for the Nottinghamshire area from many sampling locations. The seam has been subdivided into a series of mappable units termed plies. This paper demonstrates how sulphur concentrations may vary considerably on a vertical scale and that spatial distributions within individual plies may be markedly different to other plies within the seam. Often seam sulphur maps for industrial use are produced for a composite of the whole thickness of the seam, typically as the whole seam-less-dirt. The neglect of the vertical variation in seam sulphur content may lead to the production of maps which are of limited industrial use. Variations in sulphur distributions also has implications for the origin of the sulphur in the coal. Distribution maps have allowed the controls on sulphur in the Parkgate Seam to be investigated, and these are dominated by complex depositional processes.
Mapping of geochemical variables (sulphur and chlorine contents), ash content, seam splitting and, perhaps most importantly, seam thickness has been carried out throughout the history of the coal mining industry. This has been achieved by interpretation of data collected over the period of exploration, development and mining of seams, resulting in an in depth knowledge of the seam both vertically and spatially. Data are usually accumulated by seam sampling from borehole cores or pillar sections; see Ward (1984) for discussion of sampling techniques. Since economically viable seams are commonly in excess of 1 metre in thickness, several samples are usually collected and analysed. These analyses have allowed the accurate subdivision of seams, which are usually formed from a number of distinctive, laterally persistent units, termed plies. Historically data were manipulated by hand and profiles and isolines were then plotted by eye, rather than by a sophisticated technique. However, this work is now done as routine by computer, creating the opportunity of accessing and manipulating an extensive database in a way not previously feasible. Access to the British Coal database (British Coal 1992a) has allowed much investigation of the coal. The database includes extensive information (on 3569 samples) resulting from the exploitation, observation and analysis of seams for many locations. The Parkgate Seam, from the Langsettian (Westphalian A) of the Nottinghamshire Coalfield, UK, has been studied in detail and forms the basis for the work presented here. The seam is also the subject of other work on the prediction of sulphur in coal (Cavender & Spears 1995a) and an alternative analytical technique for the determination of forms of sulphur and
trace elements in coal, based on the use of ICPAES (Cavender & Spears 1995b). The present paper describes and discusses the mapping of sulphur in coal and investigates the use of ply-by-ply mapping in contrast to the mapping of the whole seam, taking into account also the vertical distribution of the sulphur content. It then goes on to discuss the controls on sulphur in the Parkgate Seam.
The Parkgate Seam The Parkgate Seam has been worked in the East Pennine Coalfield, UK, and substantial reserves remain at the present day. The seam was regarded by (the former) British Coal as one with an important future well into the next century. A considerable amount of data are available on the characteristics of the seam within the seam database (British Coal 1992a). Figure 1 shows the stratigraphical position of the seam within the Westphalian, together with the subdivision of the seam as used in this work. Seam plies are referred to by ply codes in the text of this work and the reader is referred to Fig. 1 for their intra-seam positions. The seam lies in the upper part of the Langsettian and is a considerable stratigraphical interval from any marine sediments in the sequence, a common prerequisite for high sulphur coals (Williams & Keith 1963). Several workers have discussed the depositional environments of the coal-bearing deposits of the East Pennine Coalfield, the most notable recent publications being by Guion & Fielding (1988), Guion et al. (1995) and Flint et al. (1995).
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 245-260.
246
P. F. CAVENDER & D. A. SPEARS
Westphatian
Parkgate Seam
L a n g s e t t i a n ( W e s t p h a t i a n A) m- m-~
Clay Cross Marine Band
i (Vanderbeckei)
t s
]right coal
t
C
t
Bolsovian
,'
,
I DEEP SOFT
,
f PARKGATE
ledium sulphur
t
/ i
t
/
)cc. dirty bright
I /
B
ledium ash with Ih ash coal, Dase i, occ. high
TUPTON / COCKLESHELL THREEQUARTERS
t
Duckrnantian t t
! SILKSTONE (YARD / BLACKSHALE)
i i z
I
/
'
~ KILBURN
bright and dull
~hur 1st / 2nd
]inly mudstone 3al, often dirty
A
UPPER BAND 9m. m-~ Norton Marine Band - ' - ' - - 1 NORTON FORTY YARDS . m- m-] Alton Marine Band ALTON ( G a s t r i o c e r a s listeri) i BELPER LAWN - Pot Clay Marine Band
Langsettian
...............
/
ash content very high sulphur )yrite common
m . m -~ ( G a s t r i o c e r a s s u b c r e n a t u m )
Fig. 1. Stratigraphical position of the Parkgate Seam within the Westphalian of Nottinghamshire and the
simplified subdivisions of the Parkgate Seam (redrawn from Cavender & Spears 1995a).
Correlation of the Parkgate Seam Before considering the sulphur distribution in the seam, the correlation of the seam is outlined, both from a historical perspective and for the purpose of this work. The Parkgate Seam extends over a considerable area, and is present throughout the East Pennine Coalfield. Figure 2 shows the extent of the seam and also the areas in which the intraseam correlation has been considered. It also shows the main area of the study of this work, which lies in the northern part of the Nottinghamshire Coalfield. Many workers have suggested subdivisions for the Parkgate Seam over the whole East Pennine Coalfield, and a summary of these subdivisions is given in Table 1 together with references. It should be noted that the most important and easily identifiable ply of the seam is the l st/2nd Piper split (X2P), this subdivides the seam into the 1st Piper (1P) and the 2nd Piper (2P). Originally the 'Pipers' of Derbyshire were not considered to be the same seam as the Parkgate of Yorkshire. This was noted by Edwards (1951), who introduced the term 'Dukeries' coal which represented a composite of the Deep Hard and 1st Piper Seams. This term was subsequently dropped by Edwards
(1967) when the Parkgate Seam (of Yorkshire) and the 1st and 2nd Piper Seams (of Nottinghamshire and Derbyshire) where proved to be continuous. The seam is termed the Parkgate Seam in this paper. The most up-to-date and relevant correlations of the Parkgate Seam in Nottinghamshire have been made by British Coal (1992a, b). These correlations show differing terminologies with the British Coal (1992a) division being used in this work. This subdivision of the Parkgate Seam has been made using a number of variables. Historically, a number of samples have been collected by British Coal at each sampling location. The boundaries between the samples are selected by either changes in the character of the coal, the presence of partings, or by a maximum sample thickness (usually around 200 mm). This means that several samples may be collected from a single ply of the seam. The following information is recorded for each sample collected: the thickness; the coal lithology and characteristics; the ash content and the total sulphur content. These variables, together with the presence of seam splits are used to define the individual plies of the seam and all of these data have been included in the seam database by British Coal (1992a). The seam subdivision is shown on Fig. 1, together
S U L P H U R D I S T R I B U T I O N IN A M U L T I - B E D SEAM
1320
[340.^.
,= . . . . . .
o~sea_m......
T ~-
247
, ,, --.r-_ Be~;ercotes Colliery i
,."':'"-':
~elbeck Colliery ; ~100m
,, i
2
/'/ /
Q_O erton Co ery
Ba~nsley
O' ', "" ; Thoresby Colliery ;, ~'-'
9
i '<'
/
t
'-)~
s
Ollerton Colliery 3's |
," Ollerton~Collierylg'sA&B
i~,'t, ," ,,,,,
400 % ' ,t',
seam present at depth
I/
'Olierton-Col ery 18's
I "!,"Bilsthorpe
\, 4
o~,O~
"
Z
S a m p l i n g p~176
: ",
She,, ,d
,/
!5
",i ~
380
............. ~ - - ~
/
i ",
Lr4 v~ j"
,..
6
i
erfie d
i
",
/
......lil;iiiii!iiiiiiiii, iiiiiiiiii!iiiiiii [ I !ii!
study !i::iiiiiiii!!i:i!i:i:il i :::.iarea :h:i::ili::iiii::!ii::il
I|ii:iii
ii:.i:.ii
Main
~!
,
seam present at depth
i Mansfield II
360
/ 380
9
i
} ~ ~
o) Pillar section Borehole section .~utt'ne ~ w~ /
1
i i : i!iii::ili :::iiii!i!!|
360
9~
i
)
seam present at depth
Nottingham
II
34O
\
\
. . . . . . . . . . . . . . . . . . . . . .
~ .
\ N
0 kilometres I
1440
10 t
/
~ |
/ \
,
1460
--... _
~ "- I"
/ 32_0
I
Fig. 2. The extent of the Parkgate Seam in the East Pennine Coalfield showing the main area of study, areas where correlations have been made (see Table 1) and sampling locations within the main area of study.
248
P. F. CAVENDER & D. A. SPEARS i i
:'
i
.~
o~
i
-~
~
~,
E
o
~
~
~
~
_o
~
~
I
~,. :
_
i
E
e",
(J
.~
:
~ o
I
:: _o
leOO Jaddn
-~ ....
(:':
,
',
........
.01
I
9
t-
~
O
I'~.~
e-.
~.
h--
E
8
c-.
.~_ 9
~iiill~
"~
Z
=~ --
4.%
_
sdoi Jaded ;s L
::
,
_-_ !!!~
~
I
9-
Jedld
I
Puz
::
o
~o
~
~E
_~~
~"
~
9~ -
1~-I2
UJ
:
l i:
:
:,
m ~ r- ~
~
~
0
o.
~
ll
,
"0
C
eO~
'
~
~
_
e'" ~
o
~.
"-
:
~?.
UaA!6 UO!S!A!pqns pel!e;ep
~'go
~
ou
e,-
,
i
9
,_. ....
~-
,..,
c-.
o
~,!
h" ~
iE m
~
0..
,,
~ ~
:
a. -o
03 Tr
~ ~
I
r
..~-
' ~ CO '-'~r
.
= o = | .~ = ~
='
.~
E ="
=,~1 9-
"
~
1
is
I
~
> , ~ - o ~ = _.e
~,-
~
~
. - ~
,,=
S U L P H U R D I S T R I B U T I O N IN A M U L T I - B E D SEAM with a brief description of the characteristics of each ply. The 1PUB and 1PM plies are difficult to distinguish where seam splitting is absent and therefore are considered together in this work. The 2nd Piper is c o m m o n l y split, but the splits tend to be laterally impersistent and therefore the ply has not been subdivided. In addition to the plies m e n t i o n e d above a r o o f and a floor coal are sometimes present (Table 1), separated from the m a i n b o d y of the seam by splits.
Generation of seam sulphur maps M a p s showing the distribution of seam sulphur content have been p r e p a r e d using i n f o r m a t i o n from the British Coal seam database (British Coal 1992a). Table 2 shows an example of the seam data for one sampling location within the database. F r o m the database, for a ply or g r o u p
249
of plies (specified by the ply code, Table 2), data were extracted for a given variable, in this case the sulphur content. These r a n d o m l y spaced data were then gridded using the kriging technique (see Swan & Sandilands 1995) and then s m o o t h e d sulphur isolines were constructed. All m a t h e m a t i c a l procedures a n d plotting were carried out using a personal computer. If, as in m a n y cases, m o r e t h a n one sample had been analysed for an individual ply, then a m e a n value for this ply was calculated using a weighing related to the thickness of the individual samples.
Sulphur distribution in the Parkgate Seam The only sulphur data available in the seam database are for total sulphur contents and therefore distribution maps are for total sulphur
Table 2. Format of British Coal seam database
1
2
3
FLASH LANE
468 644
364 663
693.48 693.49 693.59 693.78 693.79 693.80 693.94 694.14 694.25 694.26 694.41 694.44 694.65 694.68 694.76 694.81 694.83 695.04 695.14 695.22 695.34 695.40 695.53 695.57 695.70
10 1 10 19 1 1 14 20 11 1 15 3 20 3 8 5 2 21 10 8 12 6 13 4 13
BRIT DULL BRIT BRIT DIRT DBRT BRIT BRIT BRIT DBRT BRIT DULL BRIT DULL BRIT DULL DBRT DIRT BRIT DIRT DBRT DBRT BRIT DIRT BRIT
1PUT 1PUT 1PUT 1PUT 1PUXB 1PUB 1PUB 1PUB 1PM 1PM 1PM 1PH 1PH 1PH 1PH IPH 1PH X2P 2PU 2PXL 2PL 2PL 2PL XFC2P FC2P
3.4 3.4 3.4 3.2 55.6 20.8 3.3 4.1 7.1 28.4 3.7 8.1 3.3 3.7 4.9 10.1 15.6 79.5 14.0 84.0 16.1 25.0 10.9 53.0 14.2
1.20 1.20 1.20 1.69 0.80 1.52 1.84 1.79 2.07 1.00 2.01 0.91 1.04 0.87 1.00 1.04 1.30 0.62 3.39 0.66 8.70 16.40 3.02 1.09 2.53
4
5
6
7
8
9
Table indicates data for one location. 1 - name of location, 2 - grid reference (eastings), 3 - grid reference (northings), 4 - depth to the base of unit from borehole origin, 5 - thickness of unit (cm), 6 - coal lithology code (BRIT Bright Coal, DULL - Dull Coal, DBRT - Dirty Bright Coal, DIRT - Dirt Band), 7 - ply-code (see Fig. 1), 8 - ash content (%), 9 - total sulphur content (%).
250
P. F. CAVENDER & D. A. SPEARS
distribution only. However, a number of samples have been collected for this study and 80 samples analysed, from all plies of the seam, in order to determine the forms of sulphur in the coal. Samples from seven complete seam sections were collected from the northern part of the Nottinghamshire Coalfield (see Fig. 2). Analysis of the samples for forms of sulphur was made using an alternative technique, based on the British Standard method, but using ICP-AES to determine sulphur in the digestion solutions prepared (see Cavender & Spears 1995b for details). Total sulphur was also analysed, using the British Standard high-temperature method (British Standards Institution 1977). The analyses have allowed the vertical distribution of the forms of sulphur within the seam to be studied. In addition the vertical distribution of total sulphur for a wider area has been studied using information in the seam database.
Forms of sulphur in the Parkgate Seam From the analyses carried out for this work, it has been demonstrated that the variation in total sulphur in the Parkgate Seam is almost entirely due to the differing concentration of pyrite (Cavender & Spears 1995a). This relationship can be seen clearly when total and pyritic sulphur are plotted against each other (Fig. 3a), and also in individual seam profiles (Fig. 4). Table 3 shows the general distribution of the forms of sulphur in the seam. Pyrite shows a considerable variation (1.17• in the seam with very high percentages in some samples being due to the presence of pyrite nodules. Organic sulphur is relatively constant with only a small variation
(a)
(1.09 + 0.36%). It has been suggested that organic sulphur increases with pyritic sulphur but to a lesser degree (Wandless 1959). This relationship is shown in Fig. 3b, but in this case the relationship between the forms of sulphur is unclear.
Vertical variation of total sulphur and forms of sulphur." analysed seam sections Figure 4 shows the vertical distribution of the forms of sulphur in the seam at the selected sampling locations (see Fig. 2). In all cases the 2P has the maximum total sulphur concentration, although in the Ollerton 18's (Fig. 4b), Ollerton 19's B and Thoresby sections, the total sulphur is not as high as in the other sections. Both the Welbeck (Fig. 4c) and Ollerton 3's (Fig. 4d) sections show very high concentrations of total sulphur in the 2P, with a maximum at the base of this ply (the base of the seam). In addition to the 2P, the 1PUB/1PM ply also has high total sulphur but this is much less prominent than in the 2P. The 1PH ply shows a low in the total sulphur profile. This ply is mainly composed of banded bright and dull coal, the latter often being consistent with a low sulphur content (Wandless 1959).
Vertical variation of total sulphur." seam database The vertical variation of total sulphur in the seam is summarized in Figure 5. The results have been calculated as the mean of all the mean total sulphur contents at each location for individual plies. The results are similar to those obtained
(b)
10
10 r2=0.09
9
~
8 7 ~
6
it~
0 ,r>,
-5
5
.tO
4 3
9 ~o
9
9" ~ " i
tr22 = 0.9596 i
i
[
]
i
i
I
I
I
2 1 0
~
,~"" - ' . .9. t," "~ 9 9
L'~.,_
1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 total sulphur % organic sulphur % Fig. 3. (a) Pyritic sulphur plotted against total sulphur for coal samples analysed (redrawn from Cavender & Spears 1995a). (b) Pyritic sulphur plotted against organic sulphur for coal samples analysed.
SULPHUR
DISTRIBUTION
IN
A
MULTI-BED
SEAM
251
8 "5
5,
r~
o
"S rO
E
o
I
3o
E
b "-6 r,,,,,)
rl
t., 4
8
iii! !!!! !!!i !!!!!!!!!!!!!!!!-iiiii~i-_iiiii]i!i Z!]iiiiii~
-~~ -
/
r-O Q-~.
+
"6 rO
'
o
~"- .......................................................................... i-i}iii......i--ii--ii i ----ii ............ --ii
~. ..........................._._...................... __ jf~!!~!!~ i~ ~5 I.-
~
en
.,~
"r
"r
I.--
r
9
~,~,~,~~{~~./~
~
X
X
"6 o
u
~ ~
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......................................................................
.~
2o . . . . . . . . . . . . . .
~o o
r21
,......... P - -
.... t
.........................................................................................................................................
o
'S
~~'~] iiii ......................................... ............. iiiiiiiiiiiiiiiiiiiiiiiii-i
0
~6 ~
.
.
.
.
n
~
E o
~,,=,
252
P. F. CAVENDER & D. A. SPEARS
T a b l e 3. Forms of sulphur in the Parkgate Seam (data from Cavender & Spears 1995a)
Mean S.D. Min. Max.
total
pyritic
sulphate
organic
2.34 1.83 0.84 11.05
1.! 7 1.79 0.00 9.61
0.08 0.11 0.00 0.58
1.09 0.36 0.12 1.71
greater number of samples ( n = 192 except; 1PM, n = 179; 2P, n = 1 5 9 ) in the database. The maximum sulphur content is at the base of the seam (2P). Many ply descriptions from British Coal records indicated that pyrite nodules are commonly present in this ply, emphasising the importance of pyritic sulphur in this seam. The lowest concentration of sulphur is seen the 1PH ply, with the 1PUT ply also low in sulphur, and the 1PUT/1PM ply with a mean of just over 3%. From the analyses done for this work and also the information in the seam database it is clear that the vertical variation in the total sulphur is considerable.
n=75 from the seven analysed seam sections (above), although results from the latter are in all cases lower than the corresponding result from the seam database. It is likely that this is due to the
80
mean: 1.84 s.d: 0.55
60
1PUT
40 20 0
SOT
~ 1mean:3.15
40
.:
1PUB
.
20 10 0
~
50
mean: 3.25 s.d: 1.50
40
1PM
30 20
on IIINnn o .
10 0
i
90 ] ~ - ~ m e a n :
1.51
1PH
25
mean: 4.19
20
2P
15 10 5 0
,
,
0
,
,
,
1
IIl!lOOnnnOn
n 2
3
4
5
6
7
8
9
10
total sulphur %
Fig. 5. Vertical distribution of total sulphur in the Parkgate Seam (data from British Coal 1992a).
SULPHUR DISTRIBUTION IN A MULTI-BED SEAM
Ash-sulphur relationships in the Parkgate Seam Although the data on the forms of sulphur within the Parkgate Seam are restricted to analyses done in this work, a vast number of total sulphur analyses exist for the samples in the seam database. These samples also have been analysed for ash content. This ash content, determined by the high temperature ignition of coal, is a reflection of the mineral matter present in the coal in situ. The ash content is therefore controlled by various mineral fractions. By X-ray diffraction analysis, the main components of the mineral matter in the Parkgate Seam have been identified as quartz, illite, kaolinite and pyrite. Pyrite is transformed to iron oxide during the ashing process, and the reaction may be described as follows: 4FeS2 + 1102 - - + 2Fe203 + 8SO2
(1)
This reaction means that there is a minimum possible amount of iron oxide within the resulting ash for a given amount of pyrite within a coal sample. This amount can be calculated and a theoretical relationship between ash and pyritic sulphur determined. This relationship is: 0.8032 x ash
Spy =
(2)
253
where Spy is the percentage of pyritic sulphur in the coal sample and 'ash' is the percentage of ash in the coal sample. Since only total sulphur data are available in the seam database the theoretical relationship between total sulphur and ash is: St =
(0.8032 x ash) + So + Ss
(3)
where S t is the percentage of total sulphur in the coal sample, So is the percentage of organic sulphur in the coal sample and Ss is the percentage of sulphate sulphur in the coal sample. Since the values of organic and sulphate sulphur have been determined and they are both relatively constant in the Parkgate Seam (Table 2) they can be included in the theoretical equation: St = (0.8032 • ash)+ 1.18.
(4)
Figure 6 includes this line (4) on a graph of ash plotted against total sulphur, for all samples present within the seam database. Of note are the very small number of samples which lie outside this theoretical line (that is with high sulphur contents and low ash contents), indicating that equation (4) applies to the samples in the seam database. The outlying samples may be due to abnormally high percentages of organic sulphur or analytical errors. The plot also indicates another trend in the data. The minimum amount of sulphur present
35 t h e o r e t i c a l line of m i n i m u m ash a g a i n s t sulphur:
30
.
y = O . 8 0 3 2 x + 1.09
25
o~ z,..
= 20 (.-
C~
O)
O 9
9
10 9
5
L
]
,
9
/ .d~JSg~'~
0 /
9
9 o
, 9 ."
8
*O* ,4
~
o 9
..
**
.
.e.
,
. 9
" .........
. . . . 9
9
, .........
9
9149
' " ;;~" . . . . .
,
"-" " ""
/
0 /20 p r o g r e s s i v e 'dilution' of s u l p h u r by detrital fraction of coal ash
40
60 ash %
Fig. 6. Total sulphur plotted against ash for data within the seam database.
80
100
254
P. F. CAVENDER & D. A. SPEARS
decreases with increasing ash content, in contrast to the relationship already noted. In this case it is clear that the increase in ash content is not due to pyrite, although pyrite may still be present in small amounts in these samples, but silicate minerals such as detrital clays and quartz. The decrease in minimum sulphur content is caused by the relative decrease in the amount of carbonaceous material in the samples which are either dirty coals or from seam splits. Some samples on Fig. 6 show very high percentages of total sulphur. These samples are likely to include pyrite nodules and therefore abnormally high values of are obtained (>25%).
Spatial variation of total sulphur Sulphur distribution maps have been produced for the main plies of the seam. A composite map of the whole seam-less-dirt sulphur (Fig. 7) is shown in addition to the total sulphur in the 1PUT (Fig. 8), 1PUB/1PM (Fig. 9), 1PH (Fig. 10) and 2P (Fig. 11) plies. These figures show data other than sulphur distributions and are referred to in the discussion of controls on sulphur in the seam. The Parkgate Seam composite map (Fig. 7) shows a variation of sulphur contents from < 1 to >3%, with most of the seam being between
Fig. 7. The distribution of total sulphur in the Parkgate Seam (whole seam-less-dirt), showing fault trends.
Fig. 8. The distribution of total sulphur in the IPUT ply of the Parkgate Seam and the sandstone units above the seam.
Fig. 9. The distribution of total sulphur in the 1PUB/1PM plies of the Parkgate Seam.
Fig. 10. The distribution of total sulphur in the 1PH ply of the Parkgate Seam and the relationship to seam splitting (redrawn from Cavender & Spears i995a)
Fig. 11. The distribution of total sulphur in the 2P ply of the Parkgate Seam and the relationship to seam splitting (redrawn from Cavender & Spears 1995a).
SULPHUR DISTRIBUTION IN A MULTI-BED SEAM 2 and 3%. The spatial variation is considerable with, for example, Ollerton Colliery showing a generally lower sulphur content than Bevercotes Colliery. However, this map does not accurately represent the whole-seam sulphur as it would be mined since, it excludes the seam splits (dirt bands). Maps of the main plies of the seam show that the location of the area of maximum concentration of sulphur for each ply does not correspond. This is important since most maps used in the coal industry are usually composites, indicating whole seam-less-dirt (e.g. Fig. 7). Therefore these maps do not show the considerable vertical variation which may exist in seams and hence, are of limited industrial use in multibed seams of this type. Successful ply-by-ply mapping has been achieved in most of the area studied due to the high density of data points, however in some places (e.g. the easternmost extent of the 2P ply) data density is low and the distribution is interpolated over several kilometres.
Controls
o n s u l p h u r c o n t e n t d i s t r i b u t i o n in
the Parkgate
Seam
The fundamental controls on sulphur in coal have been outlined by a number of authors (Altshuler et al. 1983; Berner 1984; Casagrande, 1987; Chou 1990; Calkins 1994). These are essentially the availability of sulphate, iron and organic matter. Sulphate is found naturally in sea water, and to a much lesser degree in freshwater, therefore water salinity is an important factor affecting the sulphur content in peat and subsequently coal. Iron is required for the fixation of sulphur in the form of pyrite, which as already indicated is an important fraction of the total sulphur in the Parkgate Seam. The presence of organic matter, especially in a reactive form, is necessary to control the bacterial reduction of sulphate, which converts the sulphate to a form in which it may be fixed in the sediment. Both depositional and post-depositional controis may effect the sulphur content within a coal seam. The presence of pyrite nodules, which are of depositional origin, and pyrite in coal cleat, of clear post-depositional origin demonstrates this. As indicated above little relationship is seen between the sulphur contents of adjacent plies in the Parkgate Seam (Fig. 12). This may be used to demonstrate the structural control on sulphur in the seam is probably negligible. The two plies plotted are adjacent to each other in the seam succession but are commonly divided by the X2P split. The lack of any linear trend in these data indicates that the factors controlling the
257
12 10 i
~ 8 ,~ 6 -~ -~ 4 2
9
9 9
S~" oo 0
* 9
-
:.. ~ ~': :+ u
4 _"~.~ 1'.
s.-. 9 .,,
+7,.. 9 ,..'.--
0 1 2 3 total sulphur (1PH) % Fig. 12. Total sulphur in the 2P ply plotted against total sulphur in the 1PH ply of the Parkgate Seam (from Cavender & Spears 1995a).
sulphur distribution in these plies are operating differently, and therefore cannot relate to a geological structure which is geographically specific (Cavender & Spears 1995a). Plots of other adjacent plies in the seam also show a similar relationship. Lack of control by faulting may be shown by comparing the sulphur distribution with the actual pattern of faulting. Figure 7 shows the faulting and the whole seam-less-dirt sulphur content. Two sets of faults are observed, trending WNW-ESE and NE-SW with varying throws of up to 60 m. Although fault planes may provide pathways for pore waters within sediments (Goodarzi et al. 1993), even the area with the most concentrated faulting shows no relationship to the pattern of sulphur distribution. The seam is not a horizontal unit varying from about 350 to 850m below O.D. over the area studied. Prominent within the area is the crest of the Eakring Anticline (shown in Fig. 11) which trends from the central southern part of the area to the NNW. If percolation of groundwater by structural processes occurs, it will encounter either the roof or the floor of the seam first, and most likely the floor due to dewatering of sediments lower in the succession. Figure 11 shows a sulphur high in the area of Bilsthorpe Colliery in the 2P ply of the seam. This is near to the crest of the Eakring Anticline indicating a possible control. However other plies of the seam do not show a corresponding sulphur high in this area indicating that this control is unlikely. Movement of porewater in sediments after burial may be a potential source of sulphate to a
258
P. F. CAVENDER & D. A. SPEARS
coal seam (Spears 1991). A comparison of the presence of permeable beds, such as sandstones, near or adjacent to the base of the seam and the sulphur distribution in the floor of the seam may give some indication whether this process has been in operation. Although not discussed in detail here a study of sub-seam sandstone units indicates that there is no apparent relationship between their occurrence and the sulphur content in the 2P ply. In most documented cases, depositional controls are the most important on the sulphur distribution in the resulting seam. Although the depositional environment of the Parkgate Seam succession has not been discussed here, it has already been noted that previous publications indicate that there is apparently little marine influence. The major seams in the Langsettian/ Duckmantian succession of the coalfield have mean total sulphur contents of between 1.3 and 3.0% (Allen 1995). The identification of depositional controls is complicated by the multi-ply division of the seam, and especially by the sulphur variations from ply-to-ply. The base of the seam (2P) may be expected to be influenced by the mire environment and geochemistry during peat accumulation, flooding events during and after the cessation of peat deposition, and various later diagenetic processes. The sulphur content in the 2P shows a great variation in sulphur content from <3% to >9%. The presence of pyritic nodules, as noted earlier, would tend to suggest that pyrite is of a depositional stage origin, and is therefore related to processes operating in the mire. The occurrence of high sulphur contents at the base of seams, as shown for the 2P ply has been noted by several other workers. Wandless (1959) indicates that sulphur is often high at the roof and floor of a seam, as well as adjacent to dirt bands or seam splitting. This may be due to the less than optimum geochemical conditions for peat accumulation and organic preservation in these circumstances (Renton & Bird 1991). The flood events which caused seam splitting may have been possible mechanisms for the incorporation of sulphate into the mire during deposition, which may subsequently be fixed as sulphur in the resulting seam. The presence of seam splitting may therefore be a variable which can be used to demonstrate the occurrence of high sulphur coal. A broad relationship is seen at the southern part of Fig. 11, with a greater sulphur concentration where seam splitting (X2P) has developed. However in the northern part of the map where the split is in the form of a 'channel' feature, no such relationship is
seen. The 1PH (Fig. 10) ply also shows isolated sulphur highs, where seam splitting occurs below it. These examples do not show a clear positive relationship between sulphur content and seam splitting, although some indication of the role of flooding events in controlling sulphur contents is observed. Sandstone units above the seam may indicate the presence of post-depositional channels which developed after the cessation of peat accumulation. These channels may have been a source of sulphate allowing sulphur incorporation into the resulting seam. Figure 8 shows the sulphur concentration in the uppermost ply of the seam (1PUT), together with the sandstone channels and units above the seam (as mapped by British Coal 1992b). However little relationship between the sulphur content and the channels is seen. This is probably due to the fresh water nature of these channels which were thus an unlikely source of sulphate. Figure 8 also shows the trend of a swilley in the 1st Piper which also appears to have no influence on the sulphur characteristics of the seam, this swilley is likely to represent the course of a palaeochannel which was later abandoned with the re-establishment of the mire (Elliot 1965). The above discussion has excluded all postdepositional and a number of depositional controls as likely influences on the sulphur distribution in the Parkgate Seam. However although mappable depositional controls have not been identified complex controls which operated in the mire are likely to have caused the sulphur variations observed in the seam. At the beginning of this discussion, sulphate availability was emphasised as a fundamental prerequisite for sulphur incorporation into a seam. Flooding of the coal-forming mire with sulphate bearing waters therefore appears to be a likely control. Evidence for flood events is seen by the presence of seam splits within the seam and Fig. 1 and Table 1 show that a number of these exist within the Parkgate Seam. However the flood events which cause seam splitting may extend beyond the mappable boundary of the split and therefore their extent is difficult to map. These flooding events may also be the source of Fe 2+ in detrital form, another important input for the retention of sulphur in coal as pyrite. This does not necessarily mean that coal with a high pyrite content will be high in ash. The Fe 2+ is likely to be sourced from other external sources such as porewaters, providing the right geochemical conditions persist (Spears 1991). The swamp waters during this stratigraphical interval have commonly been suggested as being of fresh-water fluvial type, with deposition in an
SULPHUR DISTRIBUTION IN A MULTI-BED SEAM upper delta plain environment (Guion & Fielding 1988; Guion et al. 1995). However due to the lowlying nature of this type of environment and nearness to base level there is a possibility of a minor brackish influence, by mixing of saline and fresh water inland from the base level itself. This may be investigated further by using elements other than sulphur which are indicators of palaeosalinity. One such element is boron. In the Parkgate Seam it has been demonstrated that boron and sulphur show a high positive correlation (Cavender & Spears 1995b). In addition boron tends to be higher at the base of the seam and sometimes the roof. Although boron is commonly in concentrations of < 5 0 p p m in the seam, representing fresh water influenced environments (Banerjee & Goodarzi 1990) some samples have boron contents >100ppm and rarely >150ppm, indicating a brackish influenced environment. Therefore a rise in base level may have caused the increased availability of sulphate to the mire, from more brackish conditions, and resulted in plies with high sulphur contents within the seam. The spatial variation within individual plies is more complex, since for example although the 2nd Piper ply is generally high in sulphur content, there is a two to three-fold variation in the area studied. Intra-ply variations are likely to be controlled by more subtle controls which are unlikely to be preserved in the geological record as mappable features, such as palaeotopography and localized variations in iron availability and swamp salinity/sulphate availability.
Conclusions The initial part of this work demonstrates that seams may be subdivided using different variables into a number of plies and also that the distribution of sulphur in a seam plies may be mapped given sufficient data points. This ply-byply mapping is of value since this study indicates there is a considerable vertical and spatial variation of total and forms of sulphur in the Parkgate Seam. Therefore mapping on a ply-byply basis will provide more in-depth information than a composite map of the whole seam, which is commonly used in the coal industry. It has been demonstrated that the sulphur distribution in the Parkgate Seam is controlled mainly by depositional processes, although some late stage incorporation of sulphur does occur. These processes are highly complex and no specific mappable controls have been identified. However these depositional processes are likely to be dominated by flooding events.
259
The authors wish to thank a number of people who have aided this research, especially John Rippon. Also acknowledged are Mike Cooke, Allan Goode and John Raine. Acknowledgement is made to the British Coal Utilization Research Association Ltd. and the UK Department of Trade and Industry for a grant in aid of this research, but the views expressed are those of the authors and not necessarily those of BCURA of the DTI.
References ALLEN, M. J. 1995. Exploration and exploitation of the East Pennine Coalfield. In: WHATELEY,M. K. G. • SPEARS, D. A. (eds) European Coal Geology, Geological Society, London, Special Publication, 82, 207-214. ALTSHCULER,Z. S., SCHNEPEE,M. M., SILBER,C. C. & SIMON, F. O. 1983. Sulphur diagenesis in Everglades peat and origin of pyrite in coal. Science, 221, 221-227. BANERJEE, I. & GOODARZI, F. 1990. Palaeoenvironment and sulfur-boron contents of Mannville (Lower Cretaceous) coals of southern Alberta, Canada. Sedimentary Geology, 97, 297-310. BERNER, R. A. 1984. Sedimentary pyrite formation: an update. Geochimica et Cosmochimica dcta, 48, 605-615. BRITISHCOAL, 1992a. Parkgate Seam database. British Coal internal document. - - 1 9 9 2 b . Piper/Parkgate Seam, summary of geology. British Coal internal document (map). BRITISH STANDARDS INSTITUTION, 1977. Methods for Analysis and Testing of Coal and Coke, part 6. Total sulphur in coal. BS1016: Part 6: 1977. CALKINS, W. H. 1994. The chemical forms of sulphur in coal: a review. Fuel, 73, 475-484. CASAGRANDE,D. J. 1987. Sulphur in peat and coal. In: SCOTT, A. C. (ed.) Coal and Coal-bearing Strata." Recent Advances. Geological Society, London, Special Publication, 32, 87-105. CAVENDER,P. F. & SPEARS,D. A. 1995a. Assessing the geological controls on sulphur in coal seams: progress towards predictive mapping. In: Desulphurisation, IChemE Symposium Series No. 138, 197-205. -& 1995b. Analysis of forms of sulphur in coal, and minor and trace element associations with pyrite by ICP analysis of extraction solutions. In: PAJARES,J. A. & TASCON,J. M. D. (eds) Coal Science, 1653-1656. CHOU, C. L., 1990. Geochemistry of sulphur in coal. In: ORR, W. L. & WHITE, C. M. (eds) Geochemistry of Sulphur in Fossil Fuels. American Chemical Society Symposium Series No. 429, 30-52. EDEN,R. A., STEVENSON, I. P. & EDWARDS, W. N. 1957. Geology of the country around Sheffield. Memoirs of the Geological Survey (England and Wales). 3rd Edition. EDWARDS, W. N. 1951. The Concealed Coal)qeM of Yorkshire and Nottinghamshire. Memoirs of the Geological Survey (England and Wales). 3rd Edition.
260
P. F. C A V E N D E R & D. A. SPEARS
- - 1 9 6 7 . Geology of the Country' around Ollerton. Memoirs of the Geological Survey (England and Wales). 2nd Edition. ELLIOT, R. E. 1965. Swilleys in the Coal Measures of Nottinghamshire interpreted as palaeo-river courses. Mercian Geologist, 1, 133-142. FLINT, S., AITKEN, J. & HAMPSON, G. 1995. Application of sequence stratigraphy to coal-bearing coastal plain successions: implications for UK Coal Measures. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds) European Coal Geology. Geological Society, London, Special Publication, 82, 1-16. FROST, D. V. & SMART, J. G. O. 1979. Geology of the Country' around Derby. Memoirs of the Geological Survey (England and Wales). GOODARZI, F., VAN DER FLIER-KELLER, E., BEATON, A. P. & CALDER, J. 1993. Influence of groundwater on the geochemistry of Canadian coals. In: MICHAELIAN K. H. (ed.) Proceedings of the 7th International Conference on Coal Science, Alberta, Canada, 1, 156-159. GUION, P. D. & FIELDING, C. R. 1988. Westphalian A and B sedimentation in the Pennine Basin, UK. In: BESLY, B. M. & KEELING, G. (eds) Sedimentation in a Synorogenic Basin Complex: the Upper Carboniferous of NW Europe. 153-177. - - , FULTON,I. M. & JONES, N. S. 1995. Sedimentary facies of the coal-bearing Westphalian A and B strata north of the Wales-Brabant High. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds)
European Coal Geology, Geological Society, London, Special Publication, 82, 79-97. MITCHELL,G. H., STEPHENS,J. V., BLOMEHEAD,C. E. N. WRAY, D. A. 1947. Geology of the Country around Barnsley. Memoirs of the Geological Survey (England and Wales). NATIONAL COAL BOARD, 1957. Yorkshire Coalfield Seam Maps, Parkgate Seam Folio. Coal Survey Laboratory. SMITH, E. G., RHYS, G. H. & EDEN, R. A. 1967. Geology of the Country around Chesterfield, Matlock and Mansfield. Memoirs of the Geological Survey (England and Wales). & GOSSENS, R. F. 1967. Geology of the Country around East Retford, Worksop and Gainsborough. Memoirs of the Geological Survey (England and Wales). SWAN, A. R. H. & SANDILANDS,M. 1995. Introduction to Geological Data Analysis. Blackwell, Oxford. SPEARS, D. A. 1991. Pyrite in some U.K. coals. In: DUGAN, P. R., QUIGLEY, D. R. & ATRIA, Y. A. (eds), Processing and Utilisation of High Sulphur Coals IV 85-93. WANDLESS, A. M. 1959. The occurrence of sulphur in British Coals. Journal of the Institute of Fuel, 32, 258-266. WARD, C. R. (ed.) 1984. Coal Geology and Coal Technology. Blackwell, Oxford. WILLIAMS, E. G. & KEITH, M. L. 1963. Relationship between sulfur in coal and the occurrence of marine roof beds. Economic Geology, 58, 720-729.
262
V. BOUSKA E T AL.
Table 1. Content of sulphur in the lignites of the North NO. 4460
}
I
=2.006 =1.608
=1.578 ~O =1,927
min =0.030 __ normal. max -19.930 - - Iognorm.
Bohemian Basin
Sample
4014 I 3568 [ 3122 I
Ad
Sd
(%)
(wt%) (wt%) (wt%) (wt%)
Sdo,
Sd
Sd
Western part
2676 I 2230 I
7s4 I 1338 I 892 I 446 I
Merkur 3-333/94 1-332/94 232/94 16-328/94 65/211
17.59 5.99 8.18 1.68 8.60 1.81 80.49 0.58 27.80 2.99
2.11 0.09 0.49 0.01 1.33
2.26 0.09 0.21 0.03 0.98
1.62 1.50 1.11 0.54 0.68
Liboug 2-342/94 66/211 67/211 233/93
13.30 25.50 34.09 32.93
2.87 3.64 2.93 0.92
1.22 1.38 1.52 0.24
0.32 0.81 1.25 0.35
1.33 1.45 0.16 0.33
CSA 46/211 47/211 48/211 234/63
10.93 20.85 29.52 4.69
5.19 0.82 8.67 0.73
1.51 0.32 2.35 0.01
3.13 0.16 4.45 0.05
0.55 0.34 1.87 0.67
Bilina 10-338/94 12/340-94 8-339/94 51/211 53/94 28-343/94 52/211
16.61 8.33 4.94 3.89 3.37 14.56 40.94
0.52 1.26 0.71 1.24 1.35 3.60 0.88
0.01 0.13 0.02 0.37 0.02 0.69 0.24
0.02 0.50 0.07 0.62 0.69 1.84 0.43
0.49 0.63 0.62 0.25 0.64 1.07 0.21
Hlubina (1.M6j mine) 53/211 5.07 1.12 54/211 4.94 1.24 55/211 10.84 0.76 57/94 4.13 1.45 59/211 4.63 0.92 60/211 3.71 1.01 61/211 27.29 0.87 58/94 8.59 1.45 56/211 14.99 0.51 57/211 7.50 1.23 58/211 2.83 1.39
0.27 0.33 0.23 0.04 0.24 0.27 0.25 0.04 0.21 0.39 0.40
0.52 0.68 0.17 0.52 0.35 0.54 0.50 0.52 0.19 0.72 0.69
0.33 0.26 0.36 0.89 0.33 0.20 0.12 0.89 0.11 0.12 0.30
Eastern part 43-363/94 44-364/94 62/211 54/94 45-365/94
0 0 0.21 0 0
0 0 0.11 0 0
0 0 0.04 0 0
^
A NO.
I
~
-1.728
s -0.934 GD =1.774
2090 t 18.2 CM "1"4"85
1672 [
11' /11
,,631 1254
I
I II
N =11477 rain =0.O30 - - normal. max -19.930 - - Iognorm.
\
Central part
1045 [ 836 I 627 I
418 I
9
B Fig. 1. Histograms of sulphur constructed for total number of samples n = 14993 (A) and for the most frequently occurring values (B). Sulphur values correspond to wt%. to occur within the whole basin but the lowest contents are confined to the eastern and southeastern margin of the NBB (i.e. contents not exceeding 1 wt% S in coal). Contents between 1 and 1.5 wt% S in coal are most frequent in the southern segment of the central part of the NBB. Lower contents were also identified in the area of the Zatec delta and its northern extension. In contrast, high contents of sulphur (exceeding 1.5 wt%) occur along the northern margin of the basin at the foot of the Kru~nhory Mts and in the westernmost part of the NBB. The content of the organic sulphur is rather low (0.65-1.0 wt%; Hokr 1975) but an unusually high content was reported from the Hrabfik mine (2.95wt%, Bou~ka 1981). Concentrations of sulphate sulphur range mostly between hundredths to tenths of a per cent, with a maximum of 0.2wt%. Higher contents of sulphate sulphur indicate local oxidation processes have occurred (Hubfi6ek 1964; Bou~ka
4.67 14.60 5.66 3.65 45.78
0 0.24 0.36 0.30 0.38
1981). Elemental sulphur was found only in burnt-out parts of the coal seam. Its origin is derived from reduction of iron disulphides at high temperatures at the centre of the fire. The largest component of sulphur in coal of the NBB is pyrite (aggregates of microscopic dimensions) with a significant proportion of marcasite (Zelenka et al. 1970). The content of sulphide
~345 OF IRON DISULPHIDES sulphur varies considerably. Some local enrichment with disulphides of iron has been observed particularly in the lower part of the lower bench of the coal seam. Another local enrichment occurs in the upper bench in the central part of the basin, along its northern margin. Higher values of St~ were reported from the western part of the basin whereas in the eastern part, in the vicinity of Chabafovice, the lowest concentrations of St~ = 0.2-0.8 wt% occur (Zelenka 1993). Our data support this pattern (see Table 1).
Geological setting
numerous volcanic bodies and pyroclasticsclose to the basin (Doupovsk~ hory Mts, Cesk6 stfedohofi Mts). Volcanic activity preceded and also followed the formation of lignite.
Iron disulphides confined to the coal seam of the NBB Iron disulphides, the most abundant of which are pyrite and marcasite, occur in the coal seam and are represented by three major genetic types (Bougka 1981, Dubansk~ & Gottstein 1990): (i)
The North Bohemian Basin (NBB) represents the most important lignite basin in the Czech Republic (Fig. 2). It is filled with Eocene to Miocene, exclusively continental, mostly fluvial and lacustrine sediments which were deposited in a subsiding rift zone on the Variscan consolidated basement. A large peat bog developed in the North Bohemian Basin in the Early Miocene. A high to low volatile lignite seam, 30 to 40, locally even 60 to 70 m thick, was formed. Its formation was accompanied by extensive mostly basic volcanism which gave rise to
263
Synsedimentary type. The origin of disulphides of this type is isochronous with the origin of coal. They originated during the biochemical and partly also geochemical stage of coalification (Bou~ka 1981, Dopita et al. 1985). These disulphides are therefore of syngenetic or early diagenetic origin. Amorphous monosulphides were precipitated during this stage in the form of FeS.H20 or melnikovite, greigite, mackinawite or even pyrite and marcasite. It is known that the bacterium Desulfovibrio desulphuricans preferentially reduces
.r
9
f c ~ . , . t .''~
_
\
~
,,
,, ~\\\\ , ~ \ ~
~.~
"
..~,~
'~'~,-~,>~J-"
"kl\
'-'"~
0
50km
Fig. 2. Geological sketch map of the North Bohemian Basin and its vicinity: 1-4 Tertiary formations: 1, sediments of the Sokolov and Cheb Basins; 2, sediments of the North Bohemian basin; 3, effusive rocks and pyroclastics of the Doupovsk6 hory Mts; 4, effusive rocks and pyroclastics of the (~esk6 stfedohofi Mts.
264
V. BOUSKA ET AL.
the lighter ion 32802- than the heavier ion 34802-. Consequently, the hydrogen monosulphide which forms at the beginning of coalification is richer in the lighter isotope of sulphur whereas the heavier isotope concentrates in the residual sulphate solution. Pyrite, in particular, represents the characteristic synsedimentary variety of disulphide. The mineral was identified by X-ray diffraction. It forms finely dispersed microscopic framboids and tiny veinlets arranged mostly parallel with the bedding of coal and also layers of pyrite-bearing sandstone which occur at the base of the coal seam in the western part of the NBB. (ii) Diagenetic type. This disulphide formed together with the coal seam, during its burial. Tiny veinlets of iron disulphides usually filled desiccation cleats. (iii) Epigenetic type. This type is developed in already formed coal seams in which the iron disulphides were concentrated along faults in form of marcasite twins and clusters of crystals or as crystalline aggregates which filled cleats and fractures. Dubansk2~ & Gottstein (1990) anticipated the existence of iron disulphides of hydrothermal origin whose source was in the basement of the coal seam. However, this type of iron disulphides has not yet been found in the coal seam of the NBB even though low-temperature solutions could have occurred in the coal seam where some exothermal reactions might have taken place during the process of coalification (Bougka 1981). Volcanic activity could have also played some role in accumulation of this type of disulphide (Dubansk~, 1984).
Analytical methods Separated samples of iron disulphides were oxidized under vacuum using CuO at a temperature of 790~ The sulphur isotopic composition of the generated SO2 gas was measured using a Finnigan MAT251mass spectrometer. Overall analytical error is +0.15%.
Values of 6345 in lignite of the North Bohemian Basin Hokr et al. (1972) provided the first data on isotopic composition of sulphur in coal of the NBB. They gave 634S values from - 1 to 0%0 with
accuracy -4-1.0% for total sulphur in lignite. Values of 634S of-0.6%0 correspond to pyrite and 634S of +3.2%o correspond to sulphate sulphur. The amount of organic sulphur (0.852.9wt%) in lignite of the NBB is almost negligible, and therefore the isotope 6348 value equal to +1.2%o is of little importance for the general assessment of isotopic composition of local sulphur. Dubansk~, & Gottstein (1990) indicated that 634S values of synsedimentary and diagenetic types of iron disulphides in the NBB are similar to each other. These authors believe that the majority of disulphides in the NBB could have originated through thiobacterial reduction of sulphates. As the parent solution consisted of sulphates, then the disulphides showing negative 634S values are considered to be relatively younger than those exhibiting positive 634S, thus higher values. However, these authors do not provide an unambiguous explanation for markedly negative 634S values found in epigenetic disulphides among which marcasite is most abundant. They assume that a sublimation process could have been involved. Rainswell (1982) considered that pyrite from Jet Rock, which showed values -40%0 up to -43%0, formed through a reduction of sulphates in an open marine environment with an unlimited supply of sulphate ions. During the kinetic bacterial fractionation of sulphur isotopes the residual sulphates became isotopically heavier (the 634Svalues of the residual sulphate and later sulphides increase). The kinetic fractionation in a relatively fast low-temperature process of bacterial reduction of sulphates causes pyrite originated through such a process to be depleted in the heavier isotope of sulphur. The shift in isotopic composition of sulphur between the source sulphate and the final sulphides which originated through bacterial reduction may vary considerably depending on specific conditions; if the system for sulphate is closed or open, if there is enough nutrition for bacteria, if the originating H2S is immediately fixed in sulphides or remains free, etc. In general, the major shifts in isotopic composition of sulphur between sulphate and sulphide are characteristic of an open environment for sulphate or where there is unlimited supply of sulphate ions (seas) which the bacteria drain to form sulphides. In contrast to that, a minimum shift, sometimes almost non-existant, can be observed in the environment where the supply of sulphate is limited, i.e. when all incoming sulphate is continually reduced. There is, of course, a wide spectrum of various environments between these two extremes.
~34S OF IRON DISULPHIDES
265
Table 2. Variation in the sulphur isotopic composition o f the Krugnk hory Mts sulphides and sulphates and disulphides of the North Bohemian Basin (NBB)
Samples
(~348 range
References
of values Sulphides of the crystalline rocks of the Krugn6 hory Mts, Smr6iny and Slavkovsk~, les Mrs
-1 to +5%0
Smejkal et al. (1978)
Sulphates of the Miocene lake, thenardite (NazSO4) and gypsum as well as the sulfates of the Western Bohemian Spa mineral waters
+2 to +6%0
Smejkal et al. (1978)
Estimated value for the sulphate supplying the Basin
+ 5%o
Smejkal et al. (1978)
Volcanic sulphur from Cesk~ sffedohofi and Doupovsk6 hory Mts (estimate)
,-~0%o
Smejkal et al. (1978)
St for the coal of the NBB Pyrite from the coal of the NBB Sulfate sulphur from the NBB Organic sulphur in the coal of the NBB
-1 to 0%o -0.6%o +3.2%o +1.2%o
Hokr Hokr Hokr Hokr
Synsedimentary (early diagenetic) pyrite (sample No. 23) in claystone from the upper part of the coal seam, Doly Bilina open-cast mine Synsedimentary (early diagenetic) pyrite (sample No. 34) in claystone from the upper part of the coal seam, Doly Bilina open-cast mine Framboids of synsedimentary pyrite (sample No. 65/291) in coal, lower seam, Merkur open-cast mine Very fine radially divergent aggregates of early diagenetic pyrite (sample No. 37b) on the base of a big sandy lens, Doly Bilina open-cast mine Small crystals of synsedimentary pyrite (sample No. 37a) on the base of a big sandy lens, Doly Bilina open-cast mine
-0.8%o
this work
+0.8%o
this work
-0.5%o
this work
+0.5~
this work
0.0%o
et et et et
al. al. al. al.
(1972) (1972) (1972) (1972)
this work
Synsedimentary pyrite, Velkolom (2SA open-cast mine Synsedimentary marcasites, Velkolom CSA open-cast mine Diagenetic marcasites, Velkolom CSA open-cast mine Diagenetic marcasites, M. Gorkij I, Brafiany, today Doly Bilina open-cast mine Diagenetic marcasite, 1. M~ij mine, H~ije at Duchov, NBB
-4.2%o +3.5%0 +2.8%o 3.1%o -3.3%o -3.3& -4.8%o -3.5%0
Dubansk~ & Gottstein (1990) Dubansk~ & Gottstein (1990) Dubansk~ & Gottstein (1990) Dubansk~ & Gottstein (1990) Dubansk~ & Gottstein (1990) Dubansk9 & Gottstein (1990) Dubansk~,& Gottstein (1990) & Gottstein (1990) DubanskSI
Epigenetic layer of marcasite from the base of the sandy clay layer (sample No. 39), Doly Bilina open-cast mine Epigenetic pyrite concretion, the base of the 3 seam (sample No. 42), Doly Nfistup Tu~imice, Libou~ open-cast mine, NBB Crystals of marcasite (size up to 6 cm), Bilina tectonic fault (sample No. 46), Doly Bilina open-cast mine Radially divergent aggregate of epigenetic pyrite (cubes of 2-3 mm, sample No. 33), Bilina tectonic fault, Doly Bilina open-cast mine Crystals of marcasite, near the Bilina tectonic fault (sample No. 35a), Doly Bilina open-cast mine Crystals of marcasite on the fine grained aggregate of pyrite (sample No. 36), base of the coal seam, Doly Bilina open-cast mine
+4.2%o
this work
+5.8%o
this work
+4.9~
this work
+11.5%o
this work
+12.0%o
this work
+ 12.6%0
this work
Sulphate aerosols + flying ash from the power plant Chvaletice Sulphate aerosols + flying ash from the power plant Chvaletice
+0.1%o -0.8%0 -0.7%o
~ern~, (1982) Cern~, (1982) Buzek & Sr~imek (1985)
-
266
V. BOUSKA E T AL.
The isotopic composition of sulphur was studied on disulphides from the NBB, particularly on pyrite and marcasite representing various genetic types which were examined by petrographic and geochemical methods prior to isotopic studies (Bou~ka & Pegek 1995). The results obtained, together with other data are summarized in Table 2 and Fig. 3. The dissolved sulphate which supplied the NBB during the Miocene is believed to have been derived from weathered sulphides of the Krugn6 hory crystalline complex, Smr6iny unit and Slavkovsk~, les region (Smejkal et al. 1974). Consequently, based upon the knowledge of isotopic composition of sulphides of these units (-1 up to +5%o) and the fact that during syngeneous oxidation of sulphides no significant shift in isotopic composition occurs, the sulphate supplying the NBB should exhibit similar slightly positive 634S values. Volcanogenic sulphur is considered to be less important. Its isotopic composition should be close to 0%o. The TeplfiBarrandian region, located to the south of the NBB is characterized by accumulations of sulphides with low ~534Svalues averaging below 0% (Smejkal et al. 1974). Obtained sulphur isotopic compositions of sulphides from the NBB indicate that this unit was not the dominant supply of sulphur into the basin. Shifts in the isotopic composition of sulphur in the studied samples compared with anticipated values in source sulphates are rather small. In general, the sulphides showing lower values (around 0%0) were formed in larger reservoir of sulphates, and are believed to be of earlier origin and synsedimentary. Sulphides exhibiting higher values were formed during later stages of bacterial reduction when the isotopic composition of sulphates had already shifted towards higher ~348 values due to the previous drain of sulphur isotopes into earlier sulphides. Consequently, the sulphides with high ~348 values are
thought to have been formed during later stages of the whole process, i.e. during the diagenesis or as epigenetic sulphides. Smejkal (1978) estimated the ~534S values in sulphates of a Miocene lake at the onset of sedimentation of the Cypris Formation in the Cheb Basin to be +2 up to +5%o. Rather positive ~534S values in biogenetic pyrite of the Cheb Basin, in comparison with those of the original sulphates, indicate that pyrite was formed during an advanced stage of sulphate reduction in a closed basin, most likely in an undrained lake. Pyrites of the Cypris Formation showed (see Smejkal et al. 1974) 634S values varying between +7 and +45%0, the mean value being +16%o. In contrast to that, thenardite efflorescence and gypsum showed values exhibiting a conspicuous peak between +5 and +6%0. These values are close to those of sulphates which occur in the majority of mineral waters of western Bohemia which argues for their similar origin or source. Similar values were obtained in our samples no. 39, 42 and 46 (see Table 2). Some epigenetic sulphide types show higher ~534S values in the range from +4.2 to +12.6%. The formation of these sulphides is probably related to bacterial (or organic-matter related) reduction of the sulphate content of basinal porewaters. These pore-waters represent a residual sulphate reservoir already shifted to higher ~34S values by previous stages of sulphate reduction. Synsedimentary disulphides, mostly pyrites, which form microframboids or veinlets parallel with bedding, show values close to 0%o (our samples no. 23, 34, 37a, 37b and 65/291- see Table 2). These mostly control the overall isotope composition of sulphur in common lignite samples. The synsedimentary disulphides, showing various relative proportions, are distributed through the whole section of the coal seam, except in the eastern part of the NBB where they are less abundant. Sulfide.s of the crys'(alline rocks (Kru~,n~ h.ory M~s] Volcanic sulfur {Cesk~ st~edohoH, Doupovsk~ hory M'(s)
I
Organic sulfur Sulfate sulfur Synsedimentary ( 'early diagene'(ic) disulfides Epigene'dc disulfides Aerosols * flying ash, ChvaLe~ice
I
I
J
I
t
910~
Fig. 3. Diagram of plotted data from Table 2.
~" S
~34S OF IRON DISULPHIDES The mean 6348 values established in the coal of the NBB when compared with those of atmospheric SO2 and sulphate ions in water, suggest that some fractionation and a certain shift towards heavier values took place in the atmosphere and water environments (Cern~, 1982). Nevertheless, the combustion products (ash and sulphate aerosols) of the Chvaletice power plant which burns lignite of the NBB showed /534S values equal to +0.1 and -0.8 (Cern~ 1982). Several measurements at the same power plant (Buzek & Srfimek 1985) showed a mean value of ~534S corresponding to -0.7%o which is close to the mean value of coal and pyrite from the NBB. A certain but not too considerable shift may be caused by organic sulphur in the coal of the NBB. However, the concentrations are small and its isotope composition is known from only one measurement (Hokr et al. 1972).
Conclusions Among the genetically different types (synsedimentary, diagenetic and epigenetic) of iron disulphides (pyrite, marcasite) which occur in lignite of the North Bohemian Basin, the synsedimentary disulphides appear to be most abundant. They form fine dispersed microscopic framboids or veinlets mostly parallel with bedding. These disulphides show c534S values fluctuating around 0%0. These values when compared with those of sulphur dioxide in gases of the Chvaletice power plant, which burns lignite of the North Bohemian Basin, indicate that they correspond to mean values of synsedimentary disulphides. This also
267
suggests that no considerable shifts in isotopic composition of sulphur take place during thermal oxidation process in power plants.
References BOUgKA,V. 1981. Geochemistry of Coal. Elsevier, New York. - - & PEgEK,J. 1995. Mineralizace uhelnf~eh slojL MS Faculty of Science, Charles University, Prague. BUZEK, F & SRAMEK,J. 1985. Sulphur isotopes in the studies of stone monument conservation. Studies in Conservations, London, 171-176. (~ERN'I', J. 1982. Sledovdni pomdru stabilnlch izotopft slry ve srd~kdch v Praze. MS Faculty of Science, Charles University, Prague. DOPITA, M., HAVLENAV. & PE~EK, J. 1985. Lo~iska fosilnich paliv. SNTL, Prague. DUBANSKY, A. 1984. Sulfidick~ mineralizace v uhli severo6esk+ hn6douheln~ p~nve. Uhli-Rudy, 32, 223-231. -& GOTTSTEIN, O. 1990. Izotopick6 slo2eni siry Fe-sulfidh z SHR. Uhli, 38, 301-305. GOTrSTE1N, O. 1985. Vliv antropogenni (innosti na geochemii izotop~ siry. MS Institute of Applied Geology, Czechoslovac Academy of Sciences. Report No II-6-2/03.102. HOKR, Z., KOHOUT, J., HLADiKOVA,J. 1972. ~34S v hnJd~ch uhllch severo(esk~ho baz~nu. MS Geofond, Prague. RAINSWELL,R. 1982. Pyrite texture, isotopic composition and the availability of iron. American Journal of Science, Washington, 282, 1244-1263. SME~KAL,V. 1978. Isotopic geochemistry of the Cypris Formation in the Cheb basin. VJstnik Ust(ednlho flstavu geologick~ho, Praha, 53, 3-18. , HAUR, A., HLADiKOVA,J. & VAVIS.IN,I. 1974. ISOtOpic composition of sulphur of some sedimentary and endogenous sulphides in the Bohemian Massif. Casopis pro Mineralogii a Geologii, Praha, 19, 225-237.
Determination of different forms of sulphur in Yugoslav soft brown coals G. J A N K E S 1, O. C V E T K O V I ( ~ 2 & T. G L U M I ( ~ I C 2
1Faculty of Mechanical Engineering, University of Belgrade, 27.marta 80, 11000 Belgrade, FR Yugloslavia 2 IChTM, Center of Chemistry, Njegogeva 12, 11000 Belgrade, FR Yugoslavia Abstract: This paper presents the results of determination of different forms of sulphur in
two Yugoslav soft brown coals, Kolubara (KOL) and Kostolac (KOST). The forms of sulphur were determined as sulphate, monosulphide, pyrite and organic sulphur containing compounds by step-wise oxidation with perchloric acid. The tests were carried out in an inert atmosphere in Bethge's apparatus. KOL and Kost were quite different in sulphur content, KOL containing 0.71wt% of sulphur and KOST 3.11wt%, as well as in the distribution of each type of sulphur compound. Pyrite was dominant in KOST (1.89 wt%) and organic sulphur in KOL (0.47wt%). The analytical method applied was simple and good reproductivity of results was shown. It was important that all types of sulphur were determined in a single and small sample, using the same oxidizing agent. The Kolubara (KOL) and Kostolac (KOST) deposits of soft brown coal are of utmost importance for electricity production in Yugoslavia. They are located in northern and northeastern Serbia respectively (Fig. 1), over an area of several hundred square kilometres. Geological research has shown that the carbonaceous layer in Kolubara was formed sometime between the upper and lower Pontian, on the edges of the Panonian sea, created from various types of marshland vegetation. The carbonaceous rocks of Kostolac consist of two geological units, one of them dating from the boundary between the Panonian and Pontian, while the other formed during the Pontian and early Pliocene. According to their chemical, mineralogical and petrographic features, both are typical soft brown coals with a low degree of carbonation and a variable petrographic composition (Petkovi6 & Novkovi6 1975). There are six power plants located close to the Kolubara and Kostolac open cast coal mines near Belgrade, in the middle of an agricultural and densely populated area. The total output of these plants is 4860 MW and average emission of SO2 is 308 000 tonnes/year (Antic et al. 1992). This means that the use of coal in electricity production in the future will require limiting SO2 emissions. There are no emmision standards in Yugoslavia. At present emission standards for SO2, NO~, particulates, CO etc. are at a proposed stage, but their adoption has been postponed due to the existing economic situation. Currently, average SO2 emissions from Kolubara and Kostolac coals exeeds the limits determined by EEC environmental regulations (Federal Hydrometeorological Institute 1994).
A number of techniques are available for SO2 control: (a) coal cleaning, (b) gasification (or pyrolysis) with combustion and gas cleaning prior to combustion, (c) flue gas desulphurization, (d) combination of these techniques. Numerous commercial processes exist for each of these tehniques. The price of electricity under the emission limit in force determines the possibility of applying these processes. Proper choice depends on sound knowledge of the quality and forms of sulphur in coal and possible ways of their transformation in different prosess conditions. There was a serious lack of information about this matter for domestic coals in Yugoslavia. Sulphur compounds in coals are inorganic compounds, mainly sulphates, monosulphides and pyrite, and organic compounds such as thiophenes, aryl-, cyclic-, aliphatic-sulphides, as well as disulphides, which are minor components (Atar 1978). Because of the diversity of sulphur forms, estimation of various types of sulphur compounds in coal has always been an interesting, important and difficult analytical problem. Several analytical methods for the determination of inorganic forms of sulphur have been proposed (ASTM 1982; Tuttle et al. 1986). According to analytical procedures most often used, the monosulphide-type of sulphur was not determined (its content was usually small). On the other hand, organic sulphur was always determined indirectly, as the difference between total sulphur and the sum of inorganic forms of sulphur. In the determination of sulphur forms for small samples of KOL and KOST coals step-wise
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 269-272.
270
G. JANKES E T AL.
i
i
L.~ ' ~
0 i
50km
'~ 9 % 'I. ".,,. .....
i
r.-.i
gOST
~'~
KOL
~,,r
"~" ~,,
(
'
-,%.
\
5 .7
"~"% % 9
p-
Cv i 9 9 ".-
,r,
9s " . . . .
k g ) ? i9 .~./
Fig. 1. Map of Serbia showing the location of Kolubara and Kostolac coal fields. oxidation with perchloric acid was used (McGovan & Markuzevski 1988) 9 The results are presented in this paper.
Experimental procedure The samples used in the experiment were obtained from Kolubara and Kostolac coal mines and from coal seams which have been for the past few years in exploitation. The samples were obtained by a standard sampling method (ISO 5069-1), from the power plant conveyer belt. By means of a standard method of sample preparation (ISO 5069-2), 200 kg of coal yielded 0.5 kg of sample each, with grain
size - 125 + 90 m. The characteristics of these samples are given in Table 1 and the major constituents of the ash are presented in Table 2 (standard procedure according to the ISO 1171 was used for the ash analysis). The same samples were prepared for wire-mesh devolatilization experiments where only 0.01 g of coal was used for each test. Therefore, it was important to determine the proportion of various forms of sulphur (sulphate, monosulphide, pyrite and organic sulphur-containing compounds) in a very small sample 9This was done by successive treatment of the sample (300-500rag) with perchloric acid solutions of different concentrations, i.e. of different boiling points, as selective oxidizing agents 9The tests were carried out in an inert atmosphere (N2) in a somewhat modified
FORMS OF SULPHUR IN YUGOSLAV BROWN COALS Table 1. Characteristics of KOL and KOST coal samples
Parameters
Samples KOL
KOST
Ash (wt%, dry basis) Sulphur, total (wt% dry basis)
35.12 0.72
37.12 3.06
Ultimate analysis (wt% dry ash free) Carbon Hydrogen Nitrogen Sulphur + oxygen (by differences)
54.81 5.47 0.80 38.92
52.70 4.82 0.79 41.69
Heating value (kJ/kg-1) HHV LHV
15880 14585 15030 13625
Bethge's apparatus (McGovan & Markuzevski 1988). To determine the monosulphide and sulphate forms of sulphur the samples were treated with 40% HC104 at 115~ for 40min. The hydrogen sulphide originating from monosulphides, was introduced into 30% H202, where it oxidized and was determined as sulphate. Sulphate sulphur was determined from a filtrate obtained by filtering and rinsing the sample at the end of the reaction. Pyritic sulphur was determined in the sample residue. The oxidation was carried out by 55% HC104 at 145~ for 90min. The sulphur-containing gases were introduced into 30% H20 2. After the termination of the reaction the sample was filtered, rinsed and dried. The filtrate was added to the H2Oz-solution and the sulphate content was determined in the combined liquid mixture. Finally, the organic sulphur was determined by oxidation of the residual sample with a mixture of 75% HCIO4 and concentrated H3PO4 (9: 1) at 200~ for
271
60 min. At the end of the reaction the content of the flask was filtered and the filtrate was added to the 30% solution of hydrogen peroxide. In all cases the concentration of sulphate was determined turbidimetrically and calculated as the content of sulphur in the dry sample. Using this method, all the types of sulphur compounds present were determined in a single sample with the same oxidizing agent.
Results and conclusions
The results of direct determination of sulphur compounds in coal samples KOL and KOST are presented in Table 3. They were compared with the results of ultimate analyses of the same samples presented in Table 1. The sum of all sulphur forms determined by step-wise oxidation (Table 3) is assumed to be the total sulphur. This value for total sulphur ranged within the limits of microanalitical experimental errors. Moreover, the results were found to be reproducible (McGovan & Markuzevski 1988; Cvetkovi6 et al. 1995). It is shown that KOL and KOST are quite different in sulphur content, as well as in distribution of each type of sulphur compound. For example, KOL contains 0.71 wt% of sulphur while KOST has 3.11 wt% (Table 1). The samples do not show significant differences in ash content, but they are quite different in their content of pyrite, and organic sulphur (Table 3). The difference in total sulphur, as well as pyrite sulphur content (KOL 25%, KOST 61% of total sulphur), in the samples analyzed may point to difference in precursory matter and to different conditions in the depositional environment. A considerably higher content of pyrite sulphur (Table 3) and approximately identical
Table 2. Major constituents of ash (wt% ) Sample
KOL KOST
Constituents SiO2
A1203 Fe203 CaO
MgO
K20
Na20
SO3
Total
55.8 45.9
22.4 20.9
2.0 2.8
0.4 0.4
0.5 0.3
2.9 4.1
100.0 100.0
11.3 16.4
4.7 9.2
Table 3. Sulphur forms in Yugoslav soft brown coals (wt%, d.b.)
Sample
Monosulphide
Sulphate
Pyritic
Organic
Total
KOL KOST
0.01 0.06
0.06 0.76
0.18 1.89
0.47 0.40
0.71 3.11
272
G. J A N K E S E T AL.
amounts of organic matter in both samples (Table 1) indicate that KOST coal was formed in a more pronounced reducing environment than KOL coal. Regardless of the amount of Fe in the samples under consideration, which differs significantly (Table 2), in the KOL sample most of the sulphur (66% of total suphur) is bound to organic matter, because the reducing environment was not as pronounced. The results of organo-geochemical analysis of aliphatic hydrocarbons isolated from the soluble part of the organic matter of the above coals have also shown up difference in the character of the depositional environment. Reducing conditions in the depositional environment are charactered by a predominance of phytane as opposed to pristane, i.e. (Pr/Ph < 1 (Didyk et al. 1978)). The ratio of Pr/Ph<_ 1 in the KOL sample, and Pr/Ph<< 1 in the KOST sample (Table 3) points to a more pronounced reducing condition in the depositional environment of organic matter at KOST (unpublished results). Pyrite and organic sulphur transform to SO2 during combustion and contribute to air pollution. The share of these two forms of sulphur in total sulphur is 91% for KOL and 74% for KOST. The difference in suphur forms indicates different behaviour during combustion. It also points, however, to different possible desulphurization strategies. Pyrite sulphur, which predominates in KOST, indicates the possibility of applying coal-cleaning techniques. KOL, where organic sulphur is dominant, requires more detailed research into ways of transforming the sulphur compounds. This should provide necessary data for the selection of the gas desulphurization technique which could be applied to achieve the required reduction of SO2 emission.
References ANTIC et al. 1992. Study: Analyses of possible application of flue gas cleaning processes - FGD and NOx control (in Serbian). Mechanical Faculty, University of Belgrade, Mining Institute, Belgrade,
Energoproject, Belgrade AYrAR, A. 1978. Chemistry, thermodynamics and
kinetics of reactions of sulphur in coal-gas reactions: A Review. Fuel, 57, 4, 201-212. BERNER, A. R. 1984. Sedimentary pyrite formation: An update. Geochimica et Cosmochimica Acta, 48, 4, 605-615. CVETKOVI(~, O., GLUMI(~It~, T., DRAGUTINOVIC, V. & VITOROWC, D. 1995. Determination of different forms of sulphur in Aleksinac oil shale and evaluation of their pollution potential. I Regional
Symposium:
Chemistry
and
Environment,
Vrnja6ka Banja, 107-110. DIDYK, B. M., SIMONEIT, B. R. T., BRASSELL,S. C. EGLINTON, G. 1978. Organic geochemical indica-
tors of paleoenvironmental conditions of sedimentation. Nature, 272, 216-222. FEDERAL HYDROMETEOROG1CAL INSTITUTE 1994. National review on strategy and policy of air polution abatment in Federal Republic of Yugoslavia, Belgrade. JANKES, G., CVETKOVIC, O., GLUMI(~I(~, T., MILOVANOVIC, N. 1995. Rapid devolatilization of lignite Kolubara (in Serbian), Chemical industry, Belgrade, 49, 7-8, 317-321. McGOWAN, W. C. & MARKUZEWSKI,R. 1988. Direct
determination of sulphide, pyritic and organic sulphur in a single sample of coal by selective, step-wise oxidation with perchloric acid. Fuel, 67, 8, 1091-1095. PETKOVIC, K. & NOVKOVlC, M. 1975. Geology of
Serbia. VII Kaustobiolites (in Serbian), Faculty of Mining and Geology, University of Belgrade, Belgrade, 144-186. TUTTLE, M. L. et al. 1986. An analytical scheme for determining forms of sulphur in oil shales and associated rocks. Talanta, 33, 953-961.
Origin of vanadium in coals: parts of the Western Kentucky (USA) No. 9 coal rich in vanadium P. I. P R E M O V I C 1, N. D. N I K O L I C l, M. S. P A V L O V I C 2, LJ. S. J O V A N O V I C 1 & M . P. P R E M O V I C 1
1 Laboratory for Geochemistry and Cosmochemistry, Department o f Chemistry, University o f NiY, P.O. Box 91, 18000 Nig, FR Yugoslavia 2 Institute o f Nuclear Sciences Vin(a, 11000 Beograd, FR Yugoslavia Abstract: The existence of vanadyl (VO2+)-non-porphyrins (P) in a thin band (enriched with
vanadium) of the Western (W.) Kentucky (KY) No. 9 coal seam was shown by electron spin resonance (ESR). The ESR analysis indicates that VO2+-non-P are associated with the coal organic insoluble fraction. ESR parameters show that VO2+ ion is in an environment with approximately axial symmetry and chelated possibly by carboxylic/phenolic oxygen ligand donor atoms. These parameters are compared with those of VO2+-fulvic acid complexes and the model complexes with salicylic/phthalic acids reported by others. It is concluded that the vanadylation of W. KY No. 9 thin coal band occurred during its diagenetic (peat-forming) stage in the Pennsylvanian swamp. The extraordinary V enrichment of the top 15 cm, relative to the lower parts of the W. KY No. 9 coalbed is interpreted by a high V concentration of the past swamp water attained through a sudden and exceptional external supply. The predominant source of the metal was probably volcanic ash on the land that was weathered/leached of its vanadium. The association of V/Cr (together with their enrichment in the top of the seam) and Ni implies that the volcanic ash was derived from basalts. Abundant organic (humic) materials (with the high V enrichment factor) and low rate of deposition were the primary factors responsible for the high vanadium content of the coal. From the chemistry of VO2+, FeS2 and CrOH 2+ it is deduced that the oxidation potential Eh and pH of the ancient peat interstitial water were approximately -0.2 to -0.3 V and 5-6, respectively.
The presence of vanadium in US coals is well documented. Much analytical data (Zubovi6 et al. 1961; Zubovi6 1966; Cahill et al. 1976; Swaine 1976, 1977; Valkovi~, 1983) have been compiled because the metal has a significant effect on coal conversion processes (catalyst poisoning) (references in Maylotte et al. 1981). In addition, vanadium contributes to harmful physiological effects (e.g. lung disease) arising from the industrial combustion of coal and the resultant ejections of vanadium derivatives into the atmosphere (Rehder 1991). The discovery of the ESR signals of voZ+-P in the petroleums and asphaltenes by O'Reilly (1958) prompted researchers to apply the technique for the detection and approximate quantification of low concentrations of VO2+-P in various carbonaceous geological materials without recourse to extraction (Premovi6, 1978). Hocking & Premovi6 (1978) were the first to use this technique to study voZ+-P in the coal/coallike inclusions of the Athabasca tar sand. Premovi6 (1984) and Premovi6 et al. (1986) applied ESR to estimate the distribution of VO2+-P in the bitumen and kerogen fractions of ancient shaly-type sediments: the La Luna Mara
(LLM) limestone and Serpiano (Se) marl. Finally, Nissenbaum et al. (1980) detected VO2+-P in the DS asphalt float (Israel) using ESR. The average V content of US coals is 20 ppm (Valkovi6 1983). However, the concentration in certain parts of some seams exceeds 2000 ppm (Zubovi6 1966), e.g. the Western K Y No. 9 seam. Recently, Maylotte et al. (1981) applied the V-XAFS to probe the chemical and structural environment of V in the parts of this coalfield (Providence mine, Union County, Fig. 2) enriched with V. According to these authors the metal exists as V 3+ and VO 2+ in both of which it is bound to oxygens. Maylotte et al. (1981) also found that the the predominant form of V in the so-called heavy fraction of the K Y coal (specific gravity>_1.4gcm -3) is very similar to roscoelite of V203 with V 3+ with octahedral oxygen coordination. On the other hand, V in the light fraction (_
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 273-286.
P. I. PREMOVIC ET AL.
274
parts (rich in V) of the W. KY No. 9 seam. It was expected that these studies might yield additional information of interest as a contribution to the general understanding of the origins of the coals enriched with V.
distilled water to pH 7. After drying, the insoluble organic concentrate was again exhaustively extracted with benzene/methanol azeotrope until the solvent siphoning to the flask was clear. Material was dried at 80~ and stored in a desiccator.
Experimental procedure Elemental analysis
Sample preparation and pyrolysis The coal samples were ground to a fine powder (200-400 mesh) with a ball mill and Soxlet extracted exhaustively with benzene/methanol azeotrope to remove soluble organic material (bitumen). The extracted rock was treated with 20% hydrochloric acid (HC1) to remove carbonates. After filtration and washing, the remaining minerals were acid leached by digestion for 72 hours at room temperature using 1:1 by volume mixture of concentrated hydrofluoric acid (HF): 48% and HCI: 20%. The mixture was filtered and the residue washed successively with boiling
Elemental analysis of the coal organic insoluble fraction (Table 1) for C, H and N was done on a LECO Model 600 C H N Determinator while S was determined on the L E C O Model SC 32 Sulfur Analyzer. O was determined by difference.
Reflectance measurement The coal sample was mixed with an epoxy binder in a plastic mould and cured overnight. The grain mount was ground and polished. Maximum reflectance measurement (%R0, max) was
Table 1. Geochemical analyses of KY 9 (a) Chemical analysis Fraction
Cold HCI*
Boiling HCI
HF*
Organic
:kl% Organic fraction +1%
16.5 soluble 6
3 insoluble 94
3
77.5
* The HCI/HF fraction (b) Distribution of V[-4-1Oppm], voe+[zklOOppm], Cr[-4-1Oppm] and Ni[+lOppm] V
VO2+
Cr
Ni
1000 1800 800
500 n.d.* 650
70 135 50
40 130 15
Whole coal fraction HC1/HF Organic insoluble * n.d. not detected
(c) Elemental analysis (+0.5%, moisturefree): organic insolublefraction* C
H
H/C
N
S
Mineral matter
O (diff.)
67.5
4.5
0.8
1.0
7.0
5.0
13.0
* Total acidity 1 • 0.4 g eq kg -~ (d) M6ssbauer analysis Form of Fe
FeS2
Jarosite
+0.5%
96.0
4.0%
Sample number: 15
ORIGIN OF VANADIUM IN COALS carried out by means of a Zeiss MPM II microscope, fitted with white halogen/UV HBO light sources and using an Epiplan (Neofluor) oil immersion objective (noil = 1.518 at 546nm). (%R0, max) was recorded on vitrinite.
275
nitrosodisulfonate (Fremy's salt) for which g=2.0055+0.0001 and the nitrogen hyperfine splitting, aN---1.3094-0.001mT (Faber & Fraenkel 1967). A quartz sample tube (approximately 2mm o.d., 0.8mm i.d.) was used for Fremy's solution (c. 10-3 M) that was taped on the exterior of the tube.
Acidity determination Total acidity of the coal organic insoluble fraction was determined by exchange with barium hydroxide measuring the amount of barium uptake (Schafer 1970).
Acid leaching 1 g of the coal organic insoluble fraction was refluxed for 30 days with 100 ml of 6 M HCI. The leached residue was removed by centrifugation, carefully washed with distilled water (again by centrifugation) until free of CI-, and then thoroughly dried over P205 in a vacuum desiccator at room temperature. The dried residue was then analyzed for VO 2+ by ESR.
Scanning electron microscopy ( S E M ) and electron microprobe analysis The coal sample was examined by SEM and energy-dispersive X-ray (EDX) spectroscopy with a JEOL JSM 5300 electron microscope equipped with a Link System QX 20003 EDXspectrometer. Operating conditions for EDX analysis were 30 keV accelerating voltage, 0.1 #A beam current and a beam spot diameter of approximately 3 #m.
X-ray absorption fine structure (XAFS) spectroscopy
Atomic absorption spectrometry (AAS)
The S XAFS experiments were conducted at beam-line X-19A at the National Synchrotron Light Source (NSLS) at Brookhaven National Laboratory (New York, USA). Electron energies were 2.53GeV and beam currents were typically 90-200mA. A silicon (111) doublecrystal monochromator was used to vary the X-ray energy from approximately 50 eV below to 300 eV above the S K-shell absorption edge (2472eV). The experiments were done in the fuorescent mode, using a fluorescent ionization detector described elsewhere (Huffman et al. 1991). Further details concerning XAFS are found in Huffman et al. (1991).
A Perkin-Elmer model 4000 atomic absorption spectrometer was used with a Perkin-Elmer platinum hollow-cathode lamp and a nitrous oxide/acetylene burner head.
M6ssbauer spectroscopy
Emission spectrometry A PGS-2 plane grating spectrograph (Carl Zeiss, Jena) was used with an attachment for photoelectric detection, an arc plasma excitation source, and a Bausch and Lomb diffraction grating as the monochromator (Marinkovi6 & Vickers 1971).
Electron spin resonance (ESR) ESR measurements were performed on finely ground powders of (unheated and heated) geological samples which were transferred to an ESR quartz tube (4 mm o.d., 3 mm i.d.). Spectra were recorded on a Bruker ER 200D ESR spectrometer employing 100kHz modulation and a nominal frequency of 9.5GHz. The g-values and hyperfine coupling constants were determined relative to a solution of potassium
The M6ssbauer absorption spectrum was obtained using a constant-acceleration M6ssbauer spectrometer of standard design at US Steel Corporation, Research Laboratory, Monroeville (USA). The multichannel analyzer featured a dual-input module enabling simultaneous accumulation of a sample and calibration spectrum of an Fe foil; isomer shifts were measured with respect to metallic Fe at room temperature. Source consisted of c. 30 to 80 mCi of 57Co in a Pd matrix. Further details concerning M6ssbauer analysis are found in Huggins & Huffman (1979).
276
P. I. PREMOVIC ET AL.
Results and discussion system
Depositional environments, and the coal samples' thermal history The Springfield (W. KY) No. 9 coal of the Carbondale Formation (Middle Pennsylvanian, Fig. 1) in the W. KY coalfield (Fig. 2) of the Illinois Basin (correlative with the Illinois No. 5 and Indiana No. V coals) is the most abundant coal in the W. KY coalfield. The W. KY No. 9 coal was deposited when coastal/deltaic swamp environment (peat-forming system) covered large areas of the W. KY Basin (Rice et al. 1979). The climate at the time of the peat/coal deposition was tropical to subtropical (Hower & Wild 1982). Coal rank varies from high volatile A bituminous (hvAb) to hvCb bituminous. V (and Cr) is enriched in the ash at the top of the coal in the western portion of the field. Ni, Zn, Cu, and Co can also be concentrated in the top benches, but the trend is not as consistent as the V/Cr enrichment (Hower et al. 1990b). Hower et al. (1983, 1990a, b) considered that hydrothermal metamorphism generated some of hvAb coals in the Union coalbed (near the W. KY Fluorspar District, Fig. 2) against the background of hvCb rank. In addition, these authors suggested that the hydrothermal fluids deposited relatively high concentrations of metals such as Ba and Zn in this coalfield. The coal samples used in this study came from the top 15 cm of the W. KY No. 9 seam at Providence mine containing 1000ppm V (Table 1). Parts of the seam are exceptionally high in V up to 1800ppm (Maylotte, pers.
formation
z D
3[ tYuJ
o.
Z
:~
m
z <
>
.u
>,.
-----'Carbondale'
-KY9
Q,,, ..I
z < n ft. i if)
Fig. 1. KY 9 within the Pennsylvanian system. comm.). The elemental analysis of the coal is shown in Table 1. The coal is rich in vitrinite (around 80%). Vitrinite maximum reflectance measurement (%Ro, max=0.60) indicates that the coal under study belongs to hvCb rank.
Fig. 2. The W. KY and Illinois (Saline and Gallatin counties) coalfields (approximate (shaded) outline of the lateral distribution of coal samples rich in V). KY: U(nion), W(ebster), H(opkins), D(avies), He(nderson), Mc(lean), M(uhlenberg) and O(hio) counties; Illinois: G(allatin) and S(aline) counties.
ORIGIN OF V A N A D I UM IN COALS Throughout this paper the top 15cm part of the W. KY No. 9 coal (enriched with both V and VO 2+) will be referred to as K Y 9 unless otherwise specified.
E S R of VO 2+ and V-XAFS study Figure 3 shows the ESR spectrum of VO 2+ incorporated into the KY 9 matrix. One sees five of the weak parallel components of the spectrum at the extremites, two at low field and three at high field. The remaining ones are masked by the much stronger perpendicular components in the centre of the spectrum. There are two obvious points. First, all of the coal VO 2+non-P sites have the same magnetic parameters since there is only one set of lines in the spectrum. Second, the absence of any small splittings of the perpendicular components of
277
the spectrum implies that the VO2+-non-P sites have axial symmetry, aside from any possible rhombic distortions much smaller than the linewidth of 1 mT. In this case all the spinHamiltonian parameters can be derived from this ESR spectrum using the axially symmetric spin-Hamiltonian
=/3o[gflHzSz + g•
+ HySy)]
+ Syly)
+ AII(S.Iz) + A•
where gll, g• AIjand A l are the parallel (z) and perpendicular (x,y) components of g and 51V hyperfine coupling tensors, respectively. Hi, Si and L represent the vector components of the magnetic field, electron spin, and 51V nuclear spin along the i(= x, y, z) axes. Experimental ESR spin-Hamiltonian parameters for the VO 2+ compounds in KY 9 (given in Fig. 3), however, differed significantly from
]S,v
~20 mT
I
..I
,
51~11
1
Fig. 3. First derivative, room temperature, X-band spectrum of VO 2§ within the insoluble organic fraction of KY 9. ESR parameters: All = 17.6 4- 0.2 mT, A• = 5.7 • 0.4 mT; gLL= 1.951 + 0.003, and g• = 1.985 :t=0.010 (for VO2+-P, Premovi6 1984); All --- 19.2 + 0.3 mT and A• = 6.9 + 0.5 mT; gll --- 1.937 • 0.005, and .
.
~111_-11~931m~0a~d (Af~176
-
.
;
~
.
~
~
.
~93ur~Tc Aa~d in7 t5hmT(f~
-
"
Te~
.
,
"
.
McBa~eel9~8~b);
AII = 19.9roT and A• = 7.5mT (for the VO2+-fulvic acid in the deep peat, Abdul-Halim el al. 1981); and All -- 19.2 mT, A• = 6.8* mT (for the VO2+-phthalate/salicylate mixture, Templeton & Chasteen 1980). * Calculated as mean value of hyperfine couplings (A:,x and Ayy) derived from a non-axial spin-Hamiltonian.
278
P. I. PREMOVIC ET AL.
those of VO2+-P (Fig. 3). It has been shown that these parameters are particularly sensitive to direct ligand substitution in VO 2+ complexes (Holyk 1979). Thus, the differences in ESR parameters, especially All (which is the most sensitive parameter to the bonding) can represent the differences in the bonding ligands around VO 2+ in a VO 2+ complex. Model compound studies have shown this to be valid (Holyk 1979). For this reason, a comparison of the spin-Hamiltonian parameters for VO 2+ in KY9 and voZ+-P of various bituminous sedimentary rocks (Fig. 3) implies that VO 2+ compounds in the coal are of non-P type. Our ESR signal intensity indicates that the concentration of VO2+-non-P incorporated into KY9 is around 600ppm of VO 2+, that most of the metal (60%) resides in an organic-insoluble phase and that 80% of this V is in the VO 2+ form (Table 1). The high value for All (19.2mT, Fig. 3) indicates that the VO 2+ ion incorporated into KY9 coal is probably complexed with oxygenated functional groups such as carboxylic/ phenolic. The All and A l of voZ+-non-P are very similar to those reported VO 2+ ions incorporated into the structure of soil humic acid (McBride 1978) and fulvic acid isolated from either a podzol soil (Templeton & Chasteen 1980) or an organic-rich deep peat (90% organic matter, Abdul-Halim et al. 1981) (Fig. 3). It is suggested by these authors that these ions are bound to (carboxylic/phenolic) oxygen ligand donor atoms in the humic/fulvic acid structures. It is interesting to note that most of V is concentrated in the fulvic fraction of modern peat (Cheshire et al. 1977). In an effort to model the binding environment, ESR spectra of many fulvic acid solutions containing a variety of ligand mixtures were studied by Templeton & Chasteen (1980). Particular emphasis was placed on the salicylate/phthalate mixture because, it is generally thought that carboxylic/phenolic structures are the probable functional groups present in fulvic acid. In addition, as these authors pointed out the fulvic acid used in their investigation are characterized by a preponderance of such groups. For this reason the ESR (All and A• hyperfine parameters for this model salicylate/ phthalate complex are given in Fig. 3. The similarity in the (AlL and A• values (Fig. 3) suggests that the ligand fields about VO 2+ are comparable for VO2+-non-P and for the fulvic acid complexes on the model (salicylic/phthalate) compound. The ESR data, of course, alone do not constitute proof that carboxylic/phenolic groups make up the first coordination sphere of
VO 2+ in VO2+-non-P. However, they are certainly consistent with this interpretation. Further support for this notion comes from XAFS investigation of V in KY 9 by Maylotte et al. (1981). This study shows that there is no evidence of V in the N environment (such as voZ+-P). The limit of detection using the V XAFS method was about 50 ppm for VO 2+. Therefore, one may safely conclude that VO 2+non-P are located within the KY9 organic structure and that they are coordinated with the oxygen ligand donor atoms. These atoms are arranged in a nearly octahedral system with a strong tetragonal compression along the V-O bond of VO 2*. It is probable that coordination is primarily by carboxylate/phenolate groups with their four oxygen ligand donor atoms in the equatorial plane of VO2+-non-P, which concurs with the known chelating functional groups of unoxidized/oxidized coals. In the free axial position, perhaps, there is one water molecule/ hydroxyl ion (OH-).
Pyrite ( F e S e ) and other S compounds
EDX analysis shows that the KY9 sample contains relatively high Fe (_>2% of total sample weight); SEM and chemical analysis indicates that most of this Fe is in an unoxidized form. M6ssbauer spectroscopy reveals that 96% of total Fe present in KY 9 is pyritic Fe and only 4% appears as jarosite (the iron sulphate mineral). This mineral is usually present in weathered coals; presumably weathering product of FeS2 (Huggins & Huffman 1979). According to Smith & Batts (1974) the Fe sulphates (as weathering products of FeS2) in the coals are only of significance in relationship to very recent secondary processes. Very recently, we have initiated in situ XAFS measurements of KY 9 part. Analysis of the data are still under way and only a few preliminary results will be discussed here. The S K-edge XAFS spectrum of KY 9 (Fig. 4) can be resolved in terms of two major general form components, unoxidized and oxidized. One of the unoxidized forms is the inorganic sulphide derived principally from FeS2 (associated with the organic insoluble part). The SEM and EDX analyses indicate that all macerals of KY9 contain substantial FeS2, especially inertinite in which FeS2 is the dominant S form. The other unoxidized forms are aliphatic sulphides and aromatic thiophenes. Oxidized forms include sulphate which can conceivably be derived from both inorganic and organic S compounds. The fact that sulphate is present in KY9 (Fig. 4)
ORIGIN OF VANADIUM IN COALS
279
Origin o f VO2+-non-P
t-
Py T
si4
L ul
O
,Q ,<
Zo -8
,~"
, i
-4
0
i
'
I 4
'
I 8
11
115
Energy, eV (Elemental sulfur)
Fig. 4. XAFS spectrum of S in KY 9: Py (FeS2); Su (sulphide); T (thiophene); and, 804 (sulphates).
strongly suggests that this part has been altered by natural weathering or induced oxidation. Experimental evidence suggests that primary sedimentary FeS2 forms in anoxic depositional environments provided that organic matter, dissolved sulphate, S-reducing bacteria and a source Fe co-occur in sufficient quantities (Berner 1970). Many authors have noted that peats accumulating in brackish to marine environments tend to be enriched with S content, while fresh-water, peat-forming systems tend to produce coal with a lower S content. Studies of modern peat-forming environments, however, show a substantial increase in the FeS2 content of peats forming in marine-influenced environments (such as ancient W. KY swamp) (Casagrande et al. 1977; Altschuler et al. 1983). A 150 to 600 mm thick marine bituminous shale overlies the W. KY No. 9 coal over much of its extent. Since the marine sediment lies directly above the roof of W. KY No. 9 coalbed, it is probable that encroachment by the sea occurred very soon after, or even terminated, the final peat-forming stage. In this event, the organic material could be expected to be extremely reactive and a very rapid and complete reduction of sulphate to sulphide (HzS/FeS2) by organic/biological reactions could occur. It is, also, probable that some sulphate necessary for bacterial production of HzS/FeS2 in the W. KY No. 9 ancient swamp had arisen as a result of the downward diffusion of the dissolved sulphate in the overlying seawater. High contents of FeS2, VO 2+ and polyaromatic paramagnetic structures (PPS): c. 45 x 1019 spins g-a in the coal, as determined by ESR, is consistent with this depositional model (Premovi6 1992; Premovi6 et. al. 1993). Thus it seems reasonable to conclude that FeS2 is formed during early diagenesis of KY 9, especially that incorporated into the maceral matrixes.
Coals have two major stages of formation: (a) a diagenetic or peat-forming stage that is controlled by biological activities; (b) coalification stage in which temperature, time and pressure are important. In the coalification sequence: peat ~ lignite ~ subbituminous coal ---, bituminous (hvCb ---,hvBb ~ hvAb) coal the content of oxygenated functional groups dramatically decreases, reaching its minimum with bituminous material. Hence, the first three members of this sequence have a much higher capacity of complexing VO 2+ ions from aqueous solution than the fourth one (Szalay & Szilagyi 1967). Thus, VO 2+ ions could be incorporated into KY 9 in any of these sequential phases. The hvb coals subjected to air oxidation at moderate temperatures (< 150~ are characterized with good cation-exchange properties. This is attributed to acidic (carboxylate/phenolate) groups (the cation-exchange sites) formed during the oxidation process (Chandra 1982). In fact, these groups are excellent coordinating sites in the coal structure which would be rapidly filled by VO 2+ ions (and other cations) through uptake from aqueous solution under suitable physicochemical conditions. Preliminary measurements indicate that total (carboxylic/phenolic) acidity of the K Y 9 organic insoluble fraction is 1-t-0.4geqkg -1 (Table 1) which may (theoretically speaking) bond up to 33 500ppm of VO 2+ ions from the aqueous solution. Hower & Davies (1981) estimated that the W. KY Pennsylvanian coals attained maximum burial (2-3 km) by the end of the Permian and were uplifted to near the present surface by the middle of the Cretaceous. It is clear that prolonged weathering of K Y 9 had to be initiated/advanced during this near-surface stage of its burial. In aerated natural (subsurficial/ meteoric) waters (such as those which have been in contact with KY 9 since its uplifting by the middle of the Cretaceous), V is predicted to occur in the +5 oxidation state as the vanadate H,VO~ -3 ion (Wanty & Goldhaber 1992). As a consequence, the V species involved in the adsorption process in an oxic milieu appear to be anionic, resulting in a relatively low affinity for the cation-exchange (carboxylate/phenolate) sites in the coal (Van der Sloot 1976). At this stage it is more likely that the vanadylation of KY 9 occurred during its peatforming stage. It is unlikely that a process would be introduced via subsurface water during the lignite/subbituminous phases of the coal-forming process. The subsurface water contains no appreciable amount of V, regardless whether
280
P. I. PREMOVIC ET AL.
the source of water is meteoric (including vadose), connate or juvenile (White 1965; Overton 1973). Casagrande & Erchull (1977) in their study of metals (including V) in the subtropical Okefenokee (comprised of a wooded swamp environment) peat-forming system (Georgia, USA) pointed out that the peat-forming environment is of major importance for the distribution of metals ultimately found in coal. It is well-established that the Okefenokee represents a suitable model system that approximates ancient peat-forming systems that have ultimately given rise to coals. In fact, the tropical/subtropical low-land paralic swamps were at their maximum development in the Pennsylvanian (Johnson 1980). Under typical swamp/peat physicochemical conditions, the most stable form of V in aqueous solution is generally vanadate ions (HnVO~]-3) but the peat humic/fulvic components can reduce HnVO~]-3 to VO 2+ in an aqueous phase (Wilson & Weber 1979). VO 2+ (and other V 4+) ions can then form stable complexes with many biogeochemical compounds (e.g. humic/fulvic acids, lignins, porphyrins etc.) through chelation, metal exchange reactions and redox reactions. The fact that >70% of VO 2+ ions are incorporated into the organic insoluble fraction of KY 9 infers two important points: (a) the swamp/peat milieu was highly reducing (see below) and with a high reducing/complexing capacity through a mediation of its humic substances (i. e. highly enriched with humics); and, (b) VO 2+ ions must be bonded to geochemically stable oxygenated groups of the coal in order to survive both the coal-forming process (about 200 Ma) and prolonged weathering (100Ma). The lack of change of All of voZ+(-non-P) upon prolonged (six months) heating at 100~ (and accompanied dehydratation) of KY 9, is evidence that VO 2+ is strongly bound to the coal insoluble organic structure. The fact that extensive laboratory leaching with 6M HC1 had no effect on the ESR signals attributed to voZ+-non-P incorporated into insoluble organic part of KY9 lends further support to this interpretation.
Source o f V, volcanism and the origin of fusinites Zubovi6 (1966) found that V was enriched in the thin top block (generally _<15 cm) of six sites of the Springfield (W. KY No. 9, Illinois No. 5) coal in a relatively large area (>50000km 2) of the southern Illinois Basin (KY counties: Union, Webster, Hopkins, Davies, Henderson, McLean, Muhlenberg and Ohio; Illinois counties: Saline, Gallatin), (Fig. 2). Zubovi6 (1966) pointed out
that there is no indication that the bituminous shale which overlies W. KY No. 9 coal is a likely source of the coal V enrichment. He argued that this bituminous rock 'of comparable thickness is present throughout Illinois, but there is no comparable V enrichment of the underlying coals'. Published analyses by Hower et al. (1990b) show the V enrichment in the top benches relative to lower benches at 42 of the 44 sites of W. KY No. 9 coal where the bench samples were collected. It is quite unlikely that a source of V in the W. KY No. 9 coal was ordinary shallow (<100m) swamp waters (the from within theory). The levels of metals in the modern swamp (stagnant) waters are indeed very low (< 1 ppm) (Casagrande & Erchull 1976, 1977). The principal arguments against the hydrothermal origin of V are: (a) hydrothermal fluids contain a rather low concentrations of V similar to seawater (Wedepohl 1971; Jeandel et al. 1987); (b) numerous intensive hydrothermal sources are necessary to supply the metal for about 104 years (a time interval sufficient for the deposition of the _<15cm of the coal for an assumed coal deposition rate of 1-2cmky -1) over a large area (in the range of 5 • 10 4 105 km 2) which is quite unlikely; Fe oxides that precipitates from the hydrothermal fluids are an excellent sink for V (Dymond & Roth 1988; Trefry & Metz 1989) and (c) it is difficult to believe that a thin coal band, extending over ten thousands square kilometres would be enriched by (circulating) hydrothermal fluids while the underlying (and much thicker) parts of the same coal remain unaffected. Local hydrothermal/ surface-water activities, however, might have caused some epigenetic enrichments. The coalfields in Union and Webster counties are adjacent to the W. KY Fluorspar District which is the known area of extensive hydrothermal activities which occurred possibly as late as Upper Cretaceous (Brecke 1962). We must, therefore, postulate an (extraordinary) external supply of V of the past swamp/peat water of the W. KY No. 9 coal. We propose two views on the V source problem which we term from below and from above. According to the first concept, V was derived from either volcanic water or volcanic ash. Extensive and intense volcanism is known to have occurred in the Middle Carboniferous of North America. The other concept, perhaps more actualistic, presumes that the surface (water) processes were adequate to extract, concentrate and transfer V through weathering/ leaching volcanoclastic materials from adjacent land areas. This view was formulated by Premovi6 et al. (1986, 1993) to explain the
ORIGIN OF VANADIUM IN COALS abnormal V enrichment of some ancient sedimentary rocks of marine origin. The appeal to volcanism as a source of V arises from the presumed inadequacies of ordinary processes to supply and transport V in sufficient quantities. The evidence for a volcanic source is the presumed close association in time and space of the V enrichment in a particular sediment and volcanism (Premovi6 et al. 1993). As far as we are aware, there is no direct mineralogical (or other) evidence in favor of volcanism. However, the fact that this coalbed enriched with V cover an area of at least >50000km 2 (Fig. 2) may suggest a relatively intensive source. In this case, the main source of V would be the ash containing the ejected materials which were carried downwind from the distant eruptive centre. It is, then, the exception rather than the rule to find these fine volcanic materials in the conditions in which they were when freshly deposited. Their porosity and the (physical/ chemical) instability of their constituents make them prone to alteration, especially in (chemically speaking) acidic environment such as the swamp/peat-forming milieu (see below). Hower & Wild (1982) observed (macroscopically) the increase in fusinite in the top benches relative to the middle/bottom benches. Many authors consider that fusinites achieved their high carbon contents before deposition and most probably by exposure to elevated temperatures. A commonly held view of the origin of some fusinites is that natural pyrolysis occurred as a result of forest fires, ignited by lightning or meteorites. If this concept is adequate for the W. KY No. 9 coal fusinites then it is difficult to escape the conclusion that forest fires swept through the ancient swamp of the W. KY Basin in the later stages (Des Moinesian) of the peat accumulation. It is clear, then, that these fires could be readily triggered by volcanic eruptive materials. In addition, some of the erratic lateral trends observed for V (and Cr) distribution of the W. KY No. 9 coal could be due to the V concentrating effects of the forest fires through the formation of the charcoal/ash materials. The W. KY No. 9 coalbed at the Providence mine is about 210cm thick and has a marine black shale roof immediately on top of the coal (Maylotte, pers. comm.). According to this author, only coal samples from the top 15 cm have high V content and even then not all samples from that level showed it. For this reason, we analyzed a coal sample of the top 10cm part from another (near-by) site. The absolute value of the V (100 ppm) concentration (and, of course, VO 2+ content: < 10 ppm) in this sample is substantially less then at the first
281
location (Table 1). This erratic horizontal distribution of V in the top 15cm bench samples, which were accumulated in the same freshwater basin though separated only by less than 10km (Hower et al. 1990b), suggests that perhaps the reducing/complexing capacity of humics in the particular locations of past swamp/peat forming basin may have played an important role in concentrating V. This, on the other hand, may reflect the local divergence of vegetational cover within this area of ancient forest swamp (Casagrande & Erchull 1977). In addition, a sample of bituminous roof shale covering K Y 9 contains <100ppm of V, supporting Zubovid's suggestion (Zubovid 1966) that V in this and other V enriched coal(s) is not derived from this sedimentary cover.
P r i m a r y and secondary V As noted above, V occurs in high concentrations at the top (<_150 mm) of the coalbed throughout much of the western half of the coalfield. A typical example is the V enrichment (1240 ppm) in the 126 mm top bench at a site in the Hopkins County, as opposed to the rest of six (lower) benches (total thickness c. 154 cm) in which the V content (Fig. 5) varies between 9 to 39 ppm (24 ppm used in discussion) (Hower et al. 1990b) and it is similar to the level of V (20ppm) reported by Valkovi6 (1983) for the average US coals. Thus, V (1240 ppm) in the top bench rises over the V level (24 ppm) in lower benches by a factor of >50. The most spectacular example of the V enrichment (Zubovi6 1966) was in the top 150mm section of the near-by Hopkins site, where it reached 2080ppm and dropped to < 9 p p m below the top. Such enormous V enrichments in the top portions of the W. KY No. 9 coal could only be explained by an abrupt and high influx of V into the W. KY Pennsylvanian swamp/peat-forming system and its subsequent localized chemical accumulation into the peat humics. This proposition, on the other hand, can only support the eruptive volcanism view. If this concept is correct then the volcanic event must have taken place during the Middle Pennsylvanian (Des Moinesian) about 300Ma ago and therefore preceded the deposition of the uppermost portion of the W. KY No. 9 coal. The volcanic ejecta cannot, however, have stayed aloft for more than 6 months, and by Stokes's Law should have settled through <100m of the Pennsylvanian swamp water in less than one week. These time intervals are much shorter than those assumed (c. 104 years) for the sedimentation of the top 150 mm section
282
P. I. PREMOVIC E T A L .
1400 I 1200 I 1000 i 8O0
12C
I
c3 40 )
01 0
~ . . . . . 5. . . . . . . . . . . 50
100
150
Thickness (crn)
Fig. 5. The V/Cr distribution for benches at site 19, Hopkins County (Hower et al. 1990b): , V; .....
, Cr
of the W. KY No. 9 coal. When volcanic ash fell weathering/leaching of the metal should have occurred in 104-106 years (Zielinski 1979). That is, removal of V must have taken place during, geologically speaking, in a short time interval, relatively soon (104-106 years) after the Des Moinesian volcanism that produced the ash. Consequently, the high V concentration in this 150ram top section apparently represents an admixture of primary V (derived directly from the primary ash fallout) and secondary V (which could be derived from other primary fallout deposits and eroded from elevated land sites near the W. KY Pennsylvanian forest swamp). It seems reasonable to suppose: (a) both kinds of V came from a single volcanic eruption and (b) the V contribution of the secondary ash fallout to the total coal V is much larger than that of the primary material. Thus, we suggest that the main mass of V in the top benches of W. KY No. 9 coal appears to have been added to the Pennsylvanian swamp/ peat basin as a product of volcanic activity mainly through geochemical alteration of volcanic ash. The absence of readily recognizable volcanic material (such as ash) explains, in a sense, the existence of these V-rich upper
benches themselves: explosive volcanic activity was at a minimum during most of the KY 9 time and geochemical volcanic activity at a maximum. Had the activities been reversed the uppermost portions of the W. KY No. 9 coal would be associated with a relatively large mass of volcanic ash. In brief, volcanic ash and V-rich solution were the products of volcanic activity during the KY 9 time; ash was product of brief period (<1 year) of explosive activity; V-rich solution was the product of long sustained period (104-106 years) of geochemical activity. The ultimate cause of volcanic activity of a type resulting in large-scale discharge of ash rich in V remains a volcanological problem. The highest V contents (1000-2000ppm) in the top benches of the W. KY No. 9 coal (Zubovi6 1966; Hower et al. 1990b; this work) are much higher than those of average volcanic ash-fall (200ppm, Leventhal et al. 1983). This suggests that V in the Des Moinesian volcanic ejecta was greater than normal and/or that the humics of Pennsylvanian W. KY swamp/peatforming basin had a great reducing/complexing capacity for the corresponding V ions from the swamp aqueous solution. Relative to ordinary US coal (Valkovi6 1983), the top _<150mm portions of the W. KY No. 9 coal are enriched by a factor >50 for V (Zubovic 1966; Hower et al. 1990b; this work). If the 'normal' concentration of V in the Pennsylvanian swamp water was enhanced by a factor 100, most of the extraordinary V concentrations of these sections would be easily explained by involvement of three factors: (a) high V geochemical enrichment factor (>50000: 1) of peat (Szalay & Szilagyi 1967); (b) relatively low sedimentation rate; (c) relatively high input of organic matter and V. Consequently, it is reasonable to assume that the V concentrations in this ancient swamp had not exceeded 0.2 ppm. In general, volcanic ashes have a chemical composition similar to that of the igneous rocks of the same family. Vanadium in these rocks is present predominantly as V(III) which ionic species are, however, relatively immobile. On the other hand, H,VO] -3 are readily soluble and can migrate far in the ash weathering/leaching solution and over a wide range of pH. Thus we suggest that V reached reducing bottom of ancient W. KY swamp/peat as HnVO] -3 where it was reduced to VO 2+ by humic/ fulvic components of the peat (see above). The fact that the HC1/HF soluble fraction of K Y 9 contains exceptionally high V content (1800ppm, Table 1) i.e. that 40% of total V in the coal resides in this fraction (Table 1) indicates that substantial amount of V(III)
ORIGIN OF VANADIUM IN COALS (initially located in ash-fall) was released and altered into hydrated oxides and/or vanadates which may be adsorbed on the clay particles or precipitated in CaCO3 (Evans 1978). These species are relatively labile and, as such, soluble in cold HC1. The release and associated alteration had to occur during the weathering/leaching process of ash in the O2 saturated aqueous solution.
283
~-O. O ul
Cr, V, Ni and basaltic volcanic ash Apart from V, W KY No. 9 coal is also enriched in other metals, notably Cr. 60% of total Cr in K Y 9 resides in the organic insoluble fraction (Table 1). Cr follows the V enrichment pattern in the top benches of the 42 sites of W. KY No. 9 coalfield. For instance, there is a 15-fold increase of the Cr content in the top section (120 ppm) of the Hopkins site over the mean value (8 ppm) for six benches below (total thickness 154 cm, Fig. 5) (Hower et al. 1990b). In fact, V/Cr/Ni are expected to be concentrated from basaltic volcanic ashes (Wedepohl 1971). Thus, we tentatively suggest that the V/Cr/Ni enrichment of the HC1/HF soluble fraction of K Y 9 (Table 1) is associated with basaltic volcanic activitity which characterizes the Middle Carboniferous volcanism of (Eastern) USA.
Physicochemical conditions of deposition of Pennsylvanian W. K Y peat-forming system Numerous Eh-pH diagrams for V are present in the literature. A critical review of thermodynamic data for aqueous V species has been presented by Wanty & Goldhaber (1992). They focused their attention on the results of experimental studies of V chemistry, especially on those for which experimental physicochemical conditions are similar to naturally occurring conditions. According to Baas-Becking et al. (1960), who made a thorough study of Eh and pH in many natural aqueous environments, the acidity of the swamp/peat waters and the interstitial waters in the peat is unlikely to vary outside the range pH 4-8. Moreover, they found that the Eh values of most peat-forming systems do not exceed +0.6V. We have constructed the Eh-pH diagram for total V concentrations of 0.2 ppm (as previously discussed) in an aqueous solution (Nikoli6 1993). This diagram is essentially identical to that of Wanty & Goldhaber (1992) but, for the sake of simplicity, we present only a part of the diagram (Fig. 6) for which
I I
"0'54
1,
6
8
pH Fig. 6. Eh-pH diagram for V O 2+ and its associated humic/fulvic acid complexes in the presence of molecular (FeS2) and ionic (CrOH 2+) species in the Pennsylvanian W. KY swamp. Hatched area represents approximate stability field for interstitial water within the swamp/peat-forming system.
physicochemical conditions attributed to a peatforming system are close to natural conditions i.e. those defined by Baas-Becking et al. (1960). We may further restrict the Eh values by the existence of FeS2 in natural marine (aqueous) environments which should be in the region +0.1 V to -0.2 V for the pH range given above. Thermodynamic data used for the FeS2 field are those reported by Wagman et al. (1982). The total dissolved element concentrations in the construction of the FeS2 stability field are: 280 ppm of Fe (an arithmetic mean value for the swamp waters of the Okefenokee system); and 90 ppm of inorganic S (found in the average seawater) (Goldberg 1961). Although this diagram has been prepared for arbitrarily selected values for V, Fe and S, the critical boundary between the VO 2+ and FeS2 fields is not significantly affected by modifying these values ten-fold in either direction. Thus this diagram may be used to estimate possible physicochemical conditions of the swamp/peat-forming system at the time when the Pennsylvanian W. KY peat is formed. It is apparent from Fig. 6 that VO 2+ species are stable thermodynamically only at high Eh conditions (>0V). Therefore, natural solutions with Eh lower than 0 V would not be expected to contain VO 2+ or its complexes thereof. Nevertheless, such solutions may contain VO 2+ (down to - 0 . 2 V ) because of the formation of stable
284
P. I. PREMOVIC E T AL.
complexes between VO 2+ and organic acid ligands (such as salycylate) (Breit & Wanty 1991). It is quite certain that we may expect similar enlargment of the VO 2+ stability field for voZ+-humic/fulvic acid complexes in the W. KY Pennsylvanian swamp/peat water (Fig. 6). Wilson & Weber (1979) observed that the VO 2+ concentration in the reaction solution containing a ten-fold molar excess of (soil) fulvic acid (under strictly anaerobic conditions) is rather low due to the formation of minor (diamagnetic) (VOOH) and major solid VO(OH)2 above c. pH4.5. In fact, their data fitted exactly the Francavilla & Chasteen (1975) experimental curve obtained for VO 2+ in aqueous solution in the absence of air O2. Although there are some differences in the chemical behaviour/structure between soil fulvic acid and the corresponding peat component, these results clearly indicate that the peat is probably the most effective for the VO 2+ complexation at pHs < 6. Bacterial HzS production is pH dependent and one of the main sulphate reducing bacteria is Desulfovibrio desulfuricans. These organisms are predominantly active at pH 6.5-8.0 (Alexander 1967), though there are a few cases where they grow at pH 5.5. The marine peat-forming milieu has a pH near neutrality (7), while the freshwater swamp/peat is at pH4. Thus higher pH(5-6) may be the cause of high H2S production in the ancient W. KY swamp/peat basin and of the consequent higher FeS2 production. Under the deduced physicochemical conditions of deposition of W. KY No. 9 coal (Fig. 6) the bulk of Cr present in the Pennsylvanian swamp water should be present as CrOH 2+ ions. The Okefenokee swamp waters contain <60 ppb Cr (Casagrande & Erchull 1977). If the Pennsylvanian swamp water of W. KY was enriched with Cr by a mediation of the Des Moinesian volcanic ash (likewise V) than it is quite reasonable to assume that Cr of this water exceeded >60ppb. On the other hand, much of the stability field of Cr is occupied by insoluble Cr203. This species dissolves to form CrOH 2+ below pH 5 for the Cr concentrations >1 ppb (Brookins 1988). Hence, physicochemical conditions of the W. KY Pennsylvanian swamp interstitial water within the upper part of the Des Moinesian peat are probably best represented by a shaded area (Fig. 6). It is worth emphasizing that on Earth there are at present no such enormous forest swamps like those during Carboniferous time (subsequently leading to the formation of the largest coalfields); whose geomorphological and climatic conditions seem to be rather incompatible with these at present. It is
therefore difficult to reconstruct the physicochemical conditions of sedimentation in the Pennsylvanian swamp of the W. KY Basin when the geomorphology and climate were quite different.
Conclusions 1. VO 2+-non-P in a thin coal band (rich in V) of the Western KY No. 9 bed are detected by ESR. 2. The ESR spectral parameters of VO 2+non-P indicate that this ion is bound to oxygen ligand donor atoms, possibly carboxylate/ phenolate groups. 3. The vanadylation occurred during the peat-forming stage of coal formation in the Middle Pennsylvanian swamp of W. KY. 4. The V enrichment in the top benches relative to lower benches of the W. KY No. 9 coalbed is caused by abrupt and high influx of V into the Pennsylvanian swamp of W. KY. It is suggested that the main source of the metal was from volcanic ash on the land by weathering/ leaching which remove V from the ash to the ancient swamp. 5. From the chemistry of VO 2+, FeS2 and CrOH 2+, it is deduced that the oxidation potential Eh and pH of the ancient peat interstitial water was approximately -0.2 to -0.3 V and 5 to 6, respectively during the W. KY No. 9 coal formation. This research was supported by grant number 0206 from the Ministry of Science (Serbia). We thank D. H. Maylotte (KY) for supplying geological samples. We are grateful to G. F. Huffman and F. E. Huggins for the XAFS and M6ssbauer analyses. We also thank M. Miljkovid for SEM and EDX examinations, J. C. Hower for providing us with reprints/copies of his publications and Ana de Pablo for reviewing the manuscript with skill and patience.
References ABDUL-HALIM,A. L., EVANS,J. C., ROWLANDS,C. C. THOMAS, J. H. 1981. An EPR spectroscopic examination of heavy metals in humic and fulvic acid soil fractions. Geochimica et Cosmochimica Acta, 45, 481-487. ALEXANDER,M. 1967. Soil Microbiology. Wiley, New York. ALTSCHULER,Z. S., SCHNEPFE,M. M., SILBER,C. C. & SIMOr~, F. O. 1983. Sulfur diagenesis in Everglades peat and origin of pyrite in coal. Science, 221,221 226.
O R I G I N OF V A N A D I U M IN COALS BAAS-BECKING,L. G. M., KAPLAN,I. R., & MOORE, D. 1960. Limits of the natural environment in terms of pH and oxidation-reduction potential. Journal of Geology, 68, 243-284. BERNER, R. A. 1970. Sedimentary pyrite formation. American Journal of Science, 268, 1-23. BRECKE, E. A. 1962. Ore genesis of the Cave-in-Rock fluorspar district, Hardin County, Illinois. Economic Geology, 57, 449-535. BREIT, G. N. & WANTY, R. B. 1991. Vanadium accumulation in petroleum and carbonaceous rocks: A review of geochemical controls during deposition and diagenesis. Chemical Geology, 91, 83-97. BROOKINS, D. G. 1988. Eh-pH Diagrams for Geochemistry. Springer, New York, 48-49, 68-72, 102-103. CAHILL, R. A., KUHN, J. K., DREHER, G. B., RUCH, R. R., GLUSKOTER,J. H. & MILLER, W. G. 1976. Occurrence and distribution of trace elements in coals. Division of Fuel Chemistry, American Chemical Society, 21(7), 90-93. CASAGRANDE, O. J. & ERCHULL, L. D. 1976. Metals in Okefenokee peat-forming environments: relation to constituents found in coal. Geochimica et Cosmochimiea Acta, 40, 387-393. & 1977. Metals in plants and waters in the Okefenokee swamp and their relationship to constituents found in coal. Geochimica et Cosmochimica Acta, 41, 1391-1394. --, SIEFERT, K., BERSCHINSKI, C. & SUTTON, N. 1977. Sulfur in peat-forming systems of the Okefenokee Swamp and Florida Everglades: origin of sulfur in coal. Geochimica et Cosmochimica Acta, 41, 161-167. CHANDRA, D. 1982. Oxidized coals. In: STACH, E., MACKOWSKY, M.-Th., TEICHMIJLLER, M., TAYLOR, G. H., CHANDRA, D. & TEICHMULLER, R. (eds) Stach's Textbook of Coal Petrology, Gebriider B6rntrager, Berlin 198-205. CHESHIRE, M. V., BERROW,M. L., GOODNAN, B. A. & MUNDIE, C. M. 1977. Metal distribution and nature of some Cu, Mn and V complexes in humic acid: An electron paramagnetic resonance study. Geochimica et Cosmochimica Acta, 39, 1711-1713. DYMOND, J. & ROTH, S. 1988. Plume dispersed hydrothermal particles. A time-series record of settling flux from the Endeavor Ridge using moored sensors. Geochimica et Cosmochimica Acta, 52, 2525-2536. EVANS, H. T. Jr. 1978. Vanadium. In: WEDEPOHL, K. H. (ed.) Handbook of Geochemistry, Vol. II/2, Sect. 23-A, Springer, New York. FABER, R. J. & FRAENKEL, G. K. 1967. Dynamic frequency shifts and the effects of molecular motions in the electron spin resonance spectra of dinitrobenzene anion radicals. Journal of Chemical Physics, 47, 2462-2476. FRANCAVILLA,J. & CHASTEEN,N. D. 1975. Hydroxide effects on the electron paramagnetic resonance spectrum of aqueous vanadyl (IV) ion. Inorganic Chemistry 14, 2860-2862. GOLDBERG, E. D. 1961. Marine geochemistry. Annual Review of Physical Chemistry, 12, 29-48.
-
-
285
HOCKING, M. B. & PREMOVIC, P. I. 1978. Coal inclusions of the Athabasca tar sands: Characterization and direct determination of vanadyl porphyrin content by electron spin resonance. Geochimica et Cosmochimica Acta, 42, 359-365. HOLYK, N. M. 1979. An electron paramagnetic resonance stud)' of model oxovanadium (IV) complexes in aqueous solution: Correlation of magnetic properties with ligand type and metal chelate structures. MSc Thesis, University of N. Hampshire. HOWER, J. C. & DAVIS, A. 1981. Application of vitrinite reflectance anisotropy in the evaluation of coal metamorphism. Geological Society of America Bulletin, 59, 980 996. & WILD, G. D. 1982. Petrographic variation in the Springfield (No. 9) coal in Western Kentucky. International Journal of Coal Geology, 2, 17-30. --, FIENE, F. L., WILD, G. D. & HELFRICH, C. T. 1983. Coal metamorphism in the upper portion of the Pennsylvanian Sturgis Formation in western Kentucky. Geological Society of America, Bulletin, 94, 1475-1481. , RIMMER, S. M., WILLIAMS, D. A. & BEARD, J. G. 1990a. Coal rank trends in the Western Kentucky coalfield and relationship to hydrocarbon occurrence. In: NuccIo, V. F. & BARKER, C. E. (eds)Applications of thermal maturity studies to energy exploration, Soc. Econ. Miner. Paleont, 127-138. --, WILD, G. D., POLLOCK, J. D., TRINKLE, E. J., BLAND, A. E. & FIENE, F. L. 1990b. Petrography, geochemistry, and mineralogy of the Springfield (Western Kentucky No. 9) coal bed. Journal of Coal Quality, 9(3), 90-100. HUFFMAN, G. P., MITRA, S., HUGGINS, F. E., SHAH, N., VALIDYA, S. & Lu FULONG. 1991. Quantitative analysis of all major forms of sulfur in coal by X-ray absorption fine structure spectroscopy. Energy & Fuels, 5, 574-581. HUGGINS, F. E. & HUFFMAN, G. P. 1979. M6ssbauer analysis of iron-bearing phases in coal, coke, and ash. In: KARR, C. Jr. (ed.) Analytical Methods for Coal and Coal Products, Vol. 3. New York, Academic, 371-423. JEANDEL, C., CASSIO, M. & MINSTER, J. F. 1987. Vanadium behaviour in the global ocean and in the Mediteranian Sea. Marine Chemistry, 21, 51 74. JOHNSON, G. A. L. 1980. Carboniferous geography and terrestrial migration routes. In: PANCHEN, A. I. (ed.) The Terrestrial Environment and the Origin of Land Vertebrates. Academic, London, 39-54. LEVENTHAL,J. S., BRIGGS, P. H. & BAKER, J. W. 1983. Geochemistry of the Chattanooga shale, Dekalb County, Central Tennessee. Southeastern Geology, 24, 101-116. MARINKOVI(~, M. & VICKERS, T. 1971. Free atom production in a stabilized dc arc device for atomic absorption spectrometry. Applied Spectroscopy, 25, 319-324.
286
P. I. P R E M O V I C E T AL.
MAYLOTTE, D. H., WONG, J., PETERS, R. L. St., LYTLE, F. W. & GREEGOR, R. B. 1981. X-ray absorption spectroscopic investigation of trace vanadium sites in coal. Science, 214, 554-556. MCBRIDE, M. B. 1978. Transition metal bonding in humic acid: An EPR study. Soil Science, 126, 200-208. NIKOLIC, N. D. 1993. Vanadium and physicochemical conditions of sedimentation. PhD thesis, University of Nig. NISSENBAUM, A., AIZENSHTAT, Z. & GOLDBERG, M. 1980. The floating asphalt blocks of the Dead Sea Basin. In: DOUGLAS, A. G. & MAXWELL, J. R. (eds) Advances in Organic Geochemistry 1979, Pergamon, Oxford, 157-161. O'REILLY, D. E. 1958. Paramagnetic resonance of vanadyl etioporphyrin. Journal of Chemical Physics, 29, 1188-1189. OVERTON, H. L. 1973. Water chemistry analysis in sedimentary basin. 14th Annual Logging Symposium Society Professional Well Logging Analyst Association, Paper L. PREMOVIC, P. I. 1978. ESR methods for trace metals in fossil fuels. In: PREMOVIC, P. I. (ed.) Proceedings of the Seventh Yugoslav Conference on General and Applied Spectroscopy, Serbian Chemical Society, 11-16. - - 1 9 8 4 . Vanadyl ion in ancient marine carbonaceous sediments. Geochimica et Cosmochimica Acta, 48, 873-877. - - 1 9 9 2 . Organic free radicals in Precambrian and Palezoic rocks: Origin and significance. In: SCHIDLOWSKI, M. (ed.) Early Organic Evolution: Implications for Mineral and Energy Resources, Springer, New York, 241-256. --, PAVLOVIC, M. S. & PAVLOVId, N. Z. 1986. Vanadium in ancient sedimentary rocks of marine origin. Geochimica et Cosmochimica Acta, 50, 1923-1931. - - , PAVLOVIt~,N. Z., PAVLOVIC, M. S. & NIKOLI(~, N. D. 1993. Physicochemical conditions of sedimentation of the Fish Clay from Stevns Klint, Denmark, and its detrital nature: Vanadium and other supportive evidence. Geochimica et Cosmochimica Acta, 57, 1733-1746. REHDER, D. 1991. The bioinorganic chemistry of vanadium. Angewandte Chemie International Edition 30, 148-167. RICE, C. L., SABLE,E. G., DEVER,G. R. & KEHN, T. M. 1979. The Missisipian and Pennsylvanian (Carboniferous) Systems in the United States-Kentucky: US Geological. Professional Papers, 1110-F. SCHAFER, H. N. S. 1970. Carboxyl groups and ion exchange in low-rank coals. Fuel, 58, 197-213. SMITI4, J. W. & BATTS, B. D. 1974. The distribution and isotopic composition of sulfur in coal. Geochimica et Cosmochimica Acta, 38, 121-133. SWAINE, D. J. 1976. Trace elements in coal. In: TUGARINOV, A. E. (ed.) Recent Contributions to Geochemistry and Analytical Chemistry, Wiley, New York, 539-550.
1977. Trace elements in coal. Trace Subst. Environ. Health, 11, 107-116. SZALAY, A. & SZILAGYI, M. 1967. The association of vanadium with humic acids. Geochimica et Cosmochimica Acta, 31, 1-6. TEMPLETON, III G. D. & CHASTEEN, N. D. 1980. Vanadium-fulvic acid chemistry: conformational and binding studies by electron spin probe techniques. Geochimica et Cosmochimica Acta, 44, 741-752. TREFRY, J. H. & METZ, S. 1989. Distribution and chemistry of suspended particles from an active hydrothermal vent site on the Mid-Atlantic Ridge at 26~ Earth Planet. Sci. Lett., 88, 1-15. VALKOVIC, V. 1983. Trace Elements in Coal, CRC Press, 64-77. VAN DER SLOOT, H. A. 1976. Neutron activation analysis of trace elements in water samples after preconcentration on active carbon. PhD thesis, Free University of Amsterdam. WAGMAN, D. D., EVANS, W. H., PARKER, V. B., SCHUMM, R. H., HALLOW, I., BAILEY, S. M., CHURNEY, K. I. & BUTALL, R. L. 1982. The NBS tables of chemical thermodynamic properties. Selected values for inorganic and C1 and C2 organic substances in SI units. Journal of Physical Chemistry, Reference data 11 (supplement 2), 392. WANTY, R. B. & GOLDHABER, M. B. 1992. Thermodynamics and kinetics of reactions involving vanadium in natural systems: Accumulation of vanadium in sedimentary rocks. Geochimical et Cosmochimica Acta, 56, 1471-1483. WEDEPOHL, K. H. 1971. Environmental influences on the chemical composition of shales and clays. In: AHRENS, L. H., PRESS, F., RUNCORN, S. K. & UREY, H. C. (eds) Physics and Chemistry of the Earth, Pergamon, 307-333. WHITE, D. E. 1965. Saline waters in sedimentary rocks. In: Fluids in Subsurface Environments. American Association Petroleum Geologists, Memoirs No. 4, 346-366. WILSON, S. A. & WEBER, J. H. 1979. An EPR study of the reduction of vanadium (V) to vanadium (IV) by fulvic acid. Chemical Geology, 26, 345-354. ZIELINISKI, R. A. 1979. Uranium mobility during interaction of rhyolite obsidian, perlite and felsite with alkaline carbonate solution. Chemical Geology, 27, 47-63. ZUBOVIC, P. 1966. Physico-chemical properties of certain minor elements as controlling factors in their distribution in coal. In: GIVEN, P. H. (ed.) Coal Science, American Chemical Society Advances Chemistry Ser. 55, Washington D.C., 211 231. --, STADNICHENKO, T. • SHEEPEY, N. B. 1961. Geochemistry of minor elements in coal of the Northern Great Plains Coal Province. United States Geological Survey Bulletin, ll17-A, 57.
Environmental impact of minerals in UK coals D. A. S P E A R S
Department of Earth Sciences, University of Sheffield, Brookhill, Sheffield $3 7HF, UK Abstract: Both detrital and diagenetic minerals are quantitatively important in UK coals. The detrital minerals include quartz, and the clay minerals illite, mixed-layer illite-smectite, kaolinite and chlorite. The diagenetic minerals are dominated by pyrite/marcasite, carbonates, mainly ankerite and calcite, and kaolinite. Pyrite makes an important contribution to the S contents of the coals and is largely responsible for the variation in total S. The S content of coals delivered to power stations could be reduced by selective extraction during mining and improved physical coal-cleaning. Pyrite is a major location for trace elements of environmental concern and the elimination of pyrite would reduce both SO2 and trace element emissions. Pyrite is also important in the weathering environment due to its instability. Its breakdown in coal stocks and colliery discard heaps may lead to spontaneous combustion. In discard heaps, pyrite breakdown is also responsible for acid porewaters and toxic elements in solution which inhibit colonization by vegetation and thus hinder reclamation of derelict land. If carbonates are sufficiently abundant, acid conditions may be eliminated. Acid porewaters are one facet of the general problem of acid mine drainage, which has a major impact on water quality, particularly in streams and rivers. In the UK acid mine drainage is a problem of growing concern due to reduction of deep-mine pumping following closure of collieries. Although pyrite breakdown in colliery spoil has an important influence on porewater compositions, it does not greatly influence the engineering stability because the major component, the mudrocks, contain very little pyrite. Upon exposure, the mudrocks break down relatively rapidly due to sedimentary structures to form an aggregate. Loss of interparticle bonding is a much slower process but there are a few mudrocks associated with low rank bituminous coals with both a high mixed-layer clay content and porosity which disintegrate rapidly during cycles of wetting and drying. There are also regional variations in the relative proportions of the clay minerals which influence the engineering properties. Kaolinite is dominant in the northern coalfields. This is also detected in the composition of fly ash produced by coal-fired power stations. Fly ash is enriched in those trace elements which have an environmentalimpact. The volatile elements, many of which were sulphide-associated in the coal, are enriched on the surface of the ash particles where they are accessible to leaching in the weathering environment, resulting in the contamination of natural waters.
In the British Coal Measures the bituminous coals vary widely in ash content through the seam. Minimum ash contents, measured at 850~ (HTA), are usually around 2%. However, major seams are usually complex, containing gradational carbargillites and inter-seam mudrocks with high ash contents. Hence the coal delivered to U K power generators averages approximately 16% ash. Based on the chemistry of the ash and quantitative mineralogy of lowtemperature plasma ashes (LTA) it is possible to demonstrate that the coal ash is essentially derived from the minerals in the coal (Pike et al. 1989; Ward 1989). The connate fluids make a minor, but nonetheless significant, contribution towards Na and C1, although it should be noted that C1 is volatilized in the ashing procedure. The ash content of these coals is therefore a direct reflection of the mineral content. This paper will briefly describe the main minerals present in U K coals and their origins before dealing in greater detail with their environmental impact. The composition of minerals is an important factor on the environmental
impact and, as composition is closely related to the origin and mode of occurrence, these aspects are summarized in this paper. Not considered here are the impacts that minerals have on utilization processes such as their influence on combustion properties.
Minerals in UK coals Sulphides, carbonates and silicates are the common minerals in the East Pennine Coalfields but oxides and phosphates also occur (Spears 1987). Minerals may be dispersed or concentrated and the grain size ranges from clay to sand-sized, but with some minerals aggregating to form megascopic concretions. The composition and mode of occurrence of the minerals are a function of the origin.
Detrital minerals The composition of the detrital sediment incorporated into the peat in the swamp environment
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 287-295.
288
D. A. SPEARS
is comparable to that found in the mudrocks, siltstones and sandstones in the sequence as a whole. However, the coal precursor swamp was generally a low-energy depositional environment and consequently the detrital sediment was finegrained consisting essentially of clay minerals with minor quartz present as silt-sized grains. The composition of the clay fraction is a function of the climate and the bedrock composition in the source area from which the sediment was derived. In the East Pennine Coalfields, illite and kaolinite are the major detrital clay minerals with minor amounts of a chlorite mineral. The illite is associated with mixed-layer illite-smectite. Discrete smectite has not been identified. Rootlets in seatearths demonstrate that soil-forming conditions also existed in the basin of deposition and in a few cases extreme alteration has produced a kaolinite rich (fireclay) bed. On the other hand, chlorite, which is susceptible to weathering, occurs in many seatearths (Rippon and Spears 1989) and also in coals, thus demonstrating that clay transformations at the time of deposition and during early diagenesis
were not of major importance. Regional differences in the clay mineralogy have been described, as, for example, in the work of Taylor & Spears (1970), which showed kaolinite was dominant in the coalfields of Scotland and northern England. Figure 1 is based on that X-ray diffraction study of tailings samples from coal preparation plants in all areas of the coalfields. These preparation plants usually dealt with several coal seams from more than one colliery, and the tailings samples are therefore representative and thus ideal for a study of regional variations.
Precipitate (diagenetic) minerals In low ash coal samples (<5% ash) kaolinite is usually the only clay mineral detected. Based on textural evidence seen under the microscope this is thought to be diagenetic. The X-ray diffraction traces show the structure is well ordered, unlike the fine-grained detrital kaolinite. In high-ash samples the diagenetic kaolinite may be obscured by the detrital kaolinite. Pyrite and its dimorph marcasite are the most common sulphide minerals. Pyrite is present as a
7A Kaolinite k
9 Scotland, Northumberland & Durham l e Yorkshire & Lancashire v East & West Midlands
i
10,/k Mica
ML Illite-Smectite
Fig. 1. To show the clay mineral proportions in representative clay samples (tailings samples) from all coalfields in the UK. Also shown are three mudrocks known to have engineering instability. Chlorite is a minor component and is not illustrated, but its abundance is proportional to that of illite.
ENVIRONMENTAL IMPACT OF MINERALS IN COALS number of textural types which formed over a long period of geological time (Love et al. 1983; Kneller & Maxwell 1985; Frankie & Hower 1987). The formation of FeS2 is controlled by the availability of iron and reduced sulphur species. The latter is linked to the concentration of sulphate in the water in the depositional basin, not only for the early diagenetic pyrite but also for the late diagenetic pyrite although the links are less clear. A number of elements occur in the pyrite in both solid solutions and as discrete sulphide phases. A simple relationship of composition with time of formation was not observed in the work of White et al. (1989). Sulphides other than iron do occur, and may even be recognized in hand specimens, as well as under the microscope. However their volumes are small compared with the iron sulphides. The cleat carbonate consists of ankerite and calcite. Textural relationships in thin section demonstrate a later stage of formation for the calcite compared with the ankerite (Spears & Caswell 1986). Based on the work of Fellows (1979) and Caswell (1983) most of the carbonate would appear to be in the cleat. Siderite occurs spasmodically in minor amounts in the coal but not in the cleat.
Weathering of minerals associated with coals Physical weathering: influence on engineering stability Mudrocks are intimately associated with coal seams. The mudrocks extracted with the coal are disposed of underground and on the surface in discard heaps. The engineering stability of the discard heaps is very much dependent on the behaviour of mudrocks during weathering. The composition, classification and weathering behaviour of Coal Measures mudrocks was comprehensively reviewed by Taylor (1988). Physical breakdown of many mudrocks takes place rapidly and is largely controlled by sedimentary structures. Mudrocks were observed to break down rapidly to conglomerate and sand-sized aggregates during emplacement on the discard heap. Once the material was buried, further physical breakdown appeared minor. Sedimentary textures are also important for some mudrocks, particularly with respect to air-breakage and slaking. If fragments of mudrocks are allowed to desiccate, air is drawn into voids in response to negative suction pressures. On subsequent saturation, the water drawn into the mudrock by capillarity creates pressures which
289
are capable of shattering the mudrock. Slaking tests conducted in vacuo restricted, but did not eliminate the breakdown indicating the presence of another controlling factor. This was identified as the intraparticle expansion of smectite. Although discrete smectite does not occur, smectite is present in mixed-layer clay minerals. The presence of an effective porosity and a significant smectite component within the mixed-layer clay mineral are both favoured by a high clay content. During deep-burial diagenesis porosity is reduced and mixed-layer illite-smectites are progressively illitized, therefore the stability of mudrocks is increased. It is for this reason that in the UK the unstable mudrocks which created problems, particularly in the coal preparation plans, are associated with the lower rank bituminous coals. Three unstable mudrocks are shown plotted on Fig. 1. For all three mudrocks not only is the relative proportion of mixed-layer clay high, but as these are clay-rich mudrocks the total clay content is also high. The shear strength of mudrocks is also a function of rank. Taylor (1988, table 5) demonstrated that the shear strength of British colliery discards fell into three groups, which were also rank groupings. Exceptions to this grouping, as for example with the Scottish samples (Fig. 1), are thought to be related to the much higher kaolinite content in the more northern coalfields and the more 'geotechnically' inert nature of kaolinite. The relative clay mineral proportions influence the strength parameters, but an even more important variable is the total clay content, particularly as an increase in the clay content means there is a decrease in the quartz content.
Chemical weathering. influence on engineering stability The detrital minerals were subjected to one or more weathering regimes before they were deposited and have therefore achieved some measure of chemical stability in the weathering environment. The minor diagenetic changes undergone by the detrital minerals do not fundamentally change the stability. The detrital minerals will therefore react slowly, if at all, during exposure in discard heaps. This does not, however, apply to the diagenetic minerals that formed in restricted chemical environments. Sulphides and carbonates are all potentially reactive when exposed to surface waters. Pyrite oxidation takes place rapidly when exposed to weathering at the surface. There are
290
D. A. SPEARS
a number of steps in the oxidation of pyrite, some of which are catalysed by the Thiobacillus bacteria. The reactions and their influence of the engineering stability of mudrocks, including the effect of the acidity generated on carbonates and clay minerals, are reviewed by Taylor (1988) and Pye & Miller (1990). The latter authors noted that chemical alteration of a Carboniferous mudrock was a rapid process and that significant post-emplacement alteration had taken place in an embankment. This was mainly attributable to the oxidation of pyrite and the knock-on reactions involving carbonates and clay minerals. The sulphuric acid produced in the breakdown of pyrite is consumed principally by the dissolution of carbonates and to a lesser extent by the clay minerals. There is loss of interlayer K and octahedral Mg, Fe and A1 from the illite, leading towards the formation of kaolinite. Gypsum and more complex sulphates such as jarosite are precipitated. The engineering properties are changed in the mineral transformations and textural changes. The pyritic Carboniferous mudrocks considered by Pye & Miller (1990) are marine whereas the Carboniferous mudrocks dealt with by Taylor (1988) are non-marine and thus essentially non-pyritic. The latter are from the Westphalian Coal Measures whereas the pyritic mudrocks are of Namurian age. This explains why in UK colliery discard heaps chemical weathering has been found to be relatively minor and not therefore a major factor in the long-term stability of the colliery spoil. Pyrite is not restricted to marine shales and the colliery discard will often include nodular and other concentrated forms of pyrite such as pyritized coal. If this material is isolated by inert mudrocks, and particularly if movement of oxygenated surface-derived porewaters is restricted by compaction at the time of emplacement, there should be minimum impact on the long-term stability of the engineering structure. Pyrite weathering at Yorkshire Main discard heap, cited by Taylor (1988), was found to be restricted even though tipping from an aerial ropeway meant that the optimum compaction had not been achieved. With a more open structure there is a risk of spontaneous combustion triggered by the reactive and exothermic pyrite oxidation. Although burnt shales is more stable, combustion of discard heaps is highly undesirable.
Chemical weathering: acid mine drainage Acid mine drainage, that is pollution by acidic and/or ferruginous waters, is a long-standing
problem in the older parts of the coalfields. In the UK there is considerable concern that the widespread programme of colliery closures and the associated cessation of pumping could widen the impact of acid mine drainage. The problem arises from the underground oxidation of pyrite by oxygenated waters originating from the surface. The pyrite reactions are the same as those occurring in colliery discard and described by Taylor (1988) and Pye & Miller (1990). Furthermore, the associated reactions involving carbonates and clay minerals are also important as they influence the pH and the composition of the ochre precipitate. The pyrite is present in the roof and floor measures and also in the residual coal left as roof supports. It has been noted (Morrison et al. 1990; Younger 1994) that problems of acid mine drainage are mainly associated with coals deposited in marine or brackish-water strata rather than with coals deposited in terrestrial fluvial environments. The association of pyrite with marine conditions, and particularly the control of SO4 availability has been noted earlier in this paper. Another aspect of the pyrite weathering and the resultant acid mine drainage is the concentration of ions in solution. These may indicate the involvement of other minerals in the reactions, for example Ca and Mg from carbonates and A1 from clay minerals, but on the other hand pyrite is an important host for a number of trace elements, which will also be released into solution under acid conditions. Coprecipitation of toxic metals with iron oxyhydroxides could create problems, for example in the functioning of treatment facilities (Murdock et al. 1994).
Restoration of derelict land Deep mining in the UK has created derelict land. Although such activity accounts for less than 20% of the total area of derelict land in the UK, it is associated with regions of high population density and therefore there is an awareness of the environmental impact. Opencast mining in the UK has not led to the same problems because it is a more recent development with, therefore, greater concern for the environment. Colliery discard heaps are a major contributor to derelict land, a significant fraction of which have been reclaimed by colonization with vegetation, often after regrading. Most of the problems associated with establishing a flourishing vegetation relate directly to the nature of the colliery spoil and the constituent minerals. The engineering behaviour of the material is clearly important for slope
ENVIRONMENTAL IMPACT OF MINERALS IN COALS stability. However the relative chemical stability of the clay minerals, a factor noted in the previous section, together with the inertness of the more carbonaceous material produces very little in the way of nutrients for plant growth (Bradshaw & Chadwick 1980). According to these authors, colliery spoil may become colonized with vegetation naturally, but there is not a classical ecological succession with more species appearing as a function of age. The number of species present would appear to be more a function of pH, irrespective of age. As in acid mine drainage, the pH of the porewaters is related to the breakdown of pyrite. Bradshaw & Chadwick (1980) note that not only is it the amount of pyrite which is important but also the proportion of the more rapidly reacting fine-grained pyrite. Ambient temperatures are also important as would be predicted and observed oxidation rates in summer months are 5-10 times higher than those recorded during the winter months (Backes et al. 1993). The discard is dark coloured, and therefore surface temperatures in direct sunlight will be higher, creating a more inhospitable environment for the establishment of the vegetation. The pH of the porewaters also depends on whether or not carbonate is available to neutralize the acid solution from the pyrite breakdown. If there is an excess of reactive pyrite, porewaters within the colliery discard will remain acid, releasing A1 into solution from the alteration of the clay minerals. Even at very low concentrations A1 in solution is extremely toxic to plant growth. The Mn released from the carbonates is also toxic in low concentrations. Bradshaw & Chadwick (1980) also note that some freshly deposited colliery discard shows salinity problems due to water soluble salts originating from groundwater. The resultant problems for plant growth are relatively short lived in a humid climate as the salts are rapidly leached. This problem is probably associated with the connate water and restricted to deep mine, high C1 coals. There is a contrast: chemical weathering is important in creating acid conditions within porewaters which adversely effect the vegetation whereas chemical weathering apparently has little influence on the stability of the discard heaps. However, the former is restricted to the zone of root activity, essentially the surface of the discard heap, whereas the stability relates to the behaviour of the whole. Also, chemical weathering of a minor component, and pyrite breakdown in particular, can dominate the chemistry of the porewaters and have a major impact on the vegetation, whereas the bulk engineering properties are little changed.
291
S02 emissions and the role of pyrite In 1988 agreement was reached in Europe on the Large Combustion Plant Directive (8/609/EEC) which required a 60% reduction in SO2 emissions from existing plant in the U K by 2003. The planned flue gas desulphurization (FGD) capacity in the UK has been reduced to 6000 MW (Cooper & Kyle 1995) in response to changes in the balance of fuels used, notably a decrease in coal and increase in gas. In the European Union as a whole, the 40% reduction in SO2 emissions achieved in 1990 compared with the level in 1980 is seen as a direct consequence of the sharp increase in F G D capacity (Bolt 1995). According to Harrison (1995) F G D is currently the only practicable solution for existing coal-fired power stations with respect to SO2 emissions. This author also notes that, in the UK, although conventional coal-cleaning does significantly reduce the total sulphur content of the coal, in principle only half of the sulphur is amenable to such treatment, as the other half occurs as organic sulphur which can only be removed by more expensive chemical means. These organic S-pyritic S proportions are in agreement with analyses obtained for the Parkgate Coal in the East Pennine Coalfield (Cavender & Spears 1995), where the average organic S was 1.09+0.36% and the pyritic S was 1.17 • 1.79 (n = 75). Sulphate S is a minor component, 0.08 +0.11% and the total S was determined as 2.344-1.83%. Although the average organic S and pyritic S values are comparable, the standard deviations are significantly different. The organic S varies over a relatively small range compared with the pyritic S and the latter is responsible for most of the variation in the total S content. The samples analysed had a minimum total S value of 0.89% and a maximum value of 11.05%, which corresponds to the minimum and maximum pyrite S values. It therefore follows that improved physical coal-cleaning would reduce the S content of many of the coal samples by rather more than 50% suggested by Harrison (1995). However, to separate all the pyrite would require fine-grinding, as much of the pyrite is micron-sized and intimately associated with coal macerals (Kneller & Maxwell 1985). A representative collection of 22 U K coals (Burchill & Way 1993) has an average organic S content of 0.82 4- 0.23% (broadly comparable with the Parkgate Coal) and an average pyritic S content of 0.35 4- 0.31% - significantly lower than for the Parkgate Coal. However, the coal samples are not entirely representative because mineral contents are untypically low. This is
292
D. A. SPEARS
demonstrated by the low ash contents which average 5 . 3 6 i 4 . 1 0 % and 4.65+2.4% if one sample with a 20.4% ash content is excluded. The latter is a commercial coal grade more typical of that delivered to the power generators. Nevertheless the selected seam samples do demonstrate that the distribution of mineral matter through seam profiles is less uniform than organic S, and low pyrite S coals are present in most seams. In some of the coals analysed the organic S is also low, which is a feature of specific coalfields. In the work on the Parkgate Coal (Cavender & Spears 1995), the S distribution has been mapped on a plie by plie basis through the seam and over a wide geographical area. This has been done, not only to establish the geological controls on the S distribution and thus improve prediction by an improved understanding of processes, but also to identify areas of high S coal which might be rejected in a mining strategy designed to reduce the average S content of the mined coals. The importance of pyrite to the total S variation also highlights the importance of improved physical coal-cleaning to separate pyrite and reduce the S content of the coal and consequent SO2 emissions.
The role of minerals on element emissions from coal-fired power stations Finkelman (1982) listed the following elements with a potential sulphide association: Sb, As, Bi, Cd, Co, Cu, Ga, Ge, In, Pb, Hg, Mo, Re, Se, Te, Sn and Zn. A number of these elements are of environmental concern, particularly linked to atmospheric emissions. It has been estimated on a global scale that the power producers are the largest source of Hg (38%), Ni (52%) and V (74%) and an important source of Sb (21%), Cd (9%) and Se (23%) (Clarke & Sloss 1992). Coal is a major source of As, Cd, Hg, Sb and Se with Ni, Sn and V mainly derived from oil combustion. Elements of greatest environmental concern linked to coal utilization are As, B, Cd, Hg, Mo, Pb and Se, and of moderate concern Cr, Co, Ni, V and Zn. It should be noted that a significant number of these elements of concern are thought to have a sulphide association and therefore if globally the coal-derived input into the atmosphere is significant, then pyrite is an important contributor. In the work of White et al. (1989), synchrotron radiation X-ray fluorescence (SXRF) was used for the direct determination of trace elements in pyrite (n = 206) and marcasite (n = 25). The predominance of pyrite analyses reflects its greater abundance. Although detection limits are several
orders of magnitude better than can be obtained with an electron microprobe, the relatively large beam diameter utilized in the SXRF at that time (~ 20 #m) meant that not all textural types of pyrite could be analysed. In particular it was not possible to analyse the micron sized framboidal pyrite which is a major textural type. Nevertheless the SXRF analyses of the sulphides demonstrated significant concentrations of many of the elements listed by Finkelman (1982). No systematic differences were noted in trace element concentrations between early and late diagenetic sulphides. This also suggests that the composition of the small framboids is comparable with other textural types. Arsenic, Se, Pb, Cu and Ni were detected in most sample, T1 in about half and Zn and Mo in about a quarter. The analyses show log-normal distributions with mean values influenced by extreme values, probably due to sub-micron sized inclusions of other sulphides such as galena, chalcopyrite and blende. White et al. (1989) noted a highly significant correlation between As and Se, but not between other elements. This lack of correlation can be attributed to the inclusion of subsurface inclusions of other sulphides in the analyses and also the inferred variability of the solutions and the precipitation processes. In follow-up work by Spears & Martinez-Tarazona (1993) pyrite was concentrated using density fractions and analyses comparable to those of White et al. (1989) were obtained. The analyses also demonstrated that not only does pyrite contain elements recognized as of environmental concern, but that it is the main location for these elements in the coal. The importance of pyrite as a major host for trace elements of concern in the environment is an additional factor which needs to be taken into account in a cost/benefit analysis for the reduction of SO2 emissions. Although improved physical coal-cleaning will in theory eliminate the pyritic S, there remains the organic S and hence the requirement for FGD. While there could be less emphasis on increasing the efficiency of pyrite removal prior to combustion if FGD is to be fitted, the benefit is not only of additional S removal prior to combustion but also in the reduction of trace element concentrations in combustion products and emissions.
The influence of coal minerals on combustion residues The ash produced during combustion is essentially derived from the minerals present in the original coal. It therefore follows that the ash
ENVIRONMENTAL IMPACT OF MINERALS IN COALS composition reflects the relative proportions and compositions of the minerals. In a conventional coal-fired power station two forms of ash are obtained, the furnace bottom ash (~20%) is a high density aggregate whereas the fly ash (~80%) mainly consists of fine-grained, spherical particles. Both forms of ash are in demand in the construction industry but there is an excess production of fly ash, which is disposed of in lagoons, landfill sites and mounds. In the UK the excess, non-marketed, fly ash production was approximately 6.5 • 106 tonnes in 1990. The composition of fly ash is a function of furnace efficiency and the ash composition of the coal, which is directly controlled by the mineralogy. The level of unburnt carbon should be very low, ideally zero, otherwise the fuel loss is economically important. In the work of Hubbard et al. (1985) sampling of fly ash at 26 UK stations showed that while the ash composition was relatively constant at one location over a period of days, there were important regional differences. As noted earlier kaolinite is more important in northern coalfields in the UK (Fig. 1). This means that in terms of chemical composition A1203 is more important and K20 less so. These chemical differences are inherited by the fly ash and thus the composition of the new minerals and the glass is influenced. A high kaolinite percentage in the original coal is equated with a higher mullite content in the ash, and the generation of cenospheres (most of the amorphous silicates) is largely the result of partial fusion of illite (Hubbard et al. 1985). According to these authors it is the amorphous silicate component which constitutes a potential pozzolana (i.e. not having direct cementitious properties but which will nevertheless react with Ca 2+ released during cement hydration to form low solubility compounds of cementitious character). The coal mineralogy therefore has a major influence on the use of fly ash in cement and the elimination of a potentially waste byproduct. There is an important element fractionation between furnace bottom ash and fly ash (Clarke & Sloss 1992). The latter is in contact with the gas stream as temperatures fall. The volatile elements depleted in the furnace bottom ash are incorporated into the glass phase and precipitated on the surface of fly ash particles. Because of the surface association element concentrations are highest for the smallest fly ash particles, which are also the most difficult particulate emissions to eliminate. Volatile elements, which are only partially retained by the particulates, also contribute to atmospheric emissions. The proportion of these present in
293
the gaseous emissions is a function of volatility, thus the sulphide associated elements figure prominently. The on-land disposal of fly ash has environmental impacts (reviewed by Carlson & Adriano 1993). Elements associated with the surfaces of fly ash particles are liable to be leached in the weathering environment and released into solution in concentrations which are a function of the availability and aqueous chemistry. Most studies have been conducted on fly ash lagoons because of the immediate impact on the environment. Those elements incorporated into the glass phase may also be leached but at much slow rates than the sublimates. Nevertheless, significant amounts could enter porewaters over a period of time and these too could contribute an anthropogenic input into surface waters and groundwaters. The contamination of surface waters may be less critical because of dilution and dispersion. However, porewaters originating from fly ash disposal sites could have a significant impact on groundwater quality. In the case of such pointsource contamination there is less opportunity for dilution and groundwater remediation schemes are expensive. A study of a longestablished ash mound (Lee & Spears 1995) provided information on the longer-term weathering of PFA and potential input into aquifers over a period of 17 years. Equilibrium with respect to gypsum was noted in the deepest and oldest porewaters, but the concentration of other elements in the porewaters increased with depth indicating continued reaction of the fly ash and the non-attainment of equilibrium. Trace elements which were found to increase with depth were, with maximum concentrations, As (--~50#g/1), Se (~40#g/l), Pb (0.6mg/1), Mo (3mg/1), Ni (0.15mg/1), Li (Zing/l), B (20mg/1) and Cr (0.2 rag/l). These elements are all thought to be pyrite-associated in the coal with the exception of Li, B and Cr. The porewater in contact with the fly ash in the cited study was slightly alkaline with a pH of 7.96+0.38 (Lee & Spears 1995). This alkaline pH is attributable to hydrolysis of the surface Ca, Mg, Na and K (Talbot et al. 1978). Prediction of the leachate pH has been based on the (CaO + MgO) to (SO3 + 0.04 A1203) ratio in the bulk ash (Van der Sloot et al. 1985). With the exception of Na derived from connate fluids in the coal and part of the S originating from organic S, all the other elements are primarily associated with the coal minerals; Ca and Mg originate from carbonates, K, A1 and some of the Na comes from the clay minerals and the non-organic S from pyrite. The pH is very important in controlling the solubility of
294
D. A. SPEARS
many of the trace elements in leachates derived from fly ashes. Under alkaline conditions a number of elements occur as soluble anionic forms including arsenate, selenate, molybate, borate, chromate and vanadate whereas the more metallic cations tend to precipitate as pH increases (Hjelmar 1990). The porewater concentrations observed by Lee & Spears (1995) conform with predicted solubilities.
Conclusions 1. Minerals are quantitatively important in UK coals and are responsible for the ash, which for power station coals averages approximately 16%. 2. Detrital minerals present include quartz, illite, mixed-layer illite-smectite, kaolinite and chlorite and the non-detrital, diagenetic minerals include pyrite, kaolinite, and the carbonates ankerite and calcite. 3. The detrital minerals are relatively stable during chemical weathering with chlorite the least stable, but the smectite component in the mixedlayer clay is capable of expansion. This effect is potentially greatest in clay-rich mudrocks associated with low rank coals, which also have higher porosities and greater propensity for air-breakage, due to capillary forces. 4. There are regional variations in the composition of the clay mineral fraction associated with UK coals. This is detected in the geotechnical behaviour and also in the composition, and hence properties, of fly ash produced by power stations. 5. Pyrite makes a significant contribution to the total S of the coal and is largely responsible for the variation in the total S. 6. Pyrite is also an important location in the coal for environmentally sensitive trace elements. Improved physical coal-cleaning would not only reduce the S content of the coal but would reduce emission and disposal problems arising from trace elements. 7. Trace elements, many originating from pyrite in the coal, show significant concentrations in fly ash. The volatile elements not lost in the gaseous emissions, are enriched on the surfaces of the fly ash particles. Maximum concentrations are therefore recorded for the finest size ash particles, which are the most difficult to prevent escaping into the atmosphere from power stations. 8. The excess, non-marketed, fly ash is disposed of in lagoons, landfill sites and mounds. The surface associated trace elements may be readily leached into natural waters. Trace elements incorporated into the glass may be released
into porewaters over a longer time period. Fly ash is therefore a potential contaminant of surface and groundwaters. 9. Pyrite is unstable in the weathering environment. The rapid reaction may trigger spontaneous combustion in coal stocks and discard heaps. In general, the oxidation of pyrite in UK colliery discard does not create problems of engineering instability, mainly because pyritic mudrocks are uncommon in the Coal Measures. 10. The near-surface breakdown of pyrite in discard heaps creates acid conditions in porewaters. Carbonates and to a lesser extent clay minerals will react and toxic elements may be released into solution in addition to those originating from the pyrite. 11. If the acid porewaters escape from the discard heaps, contamination of surface waters results. This is one manifestation of the problem of acid mine drainage. More important are the acid discharges from abandoned underground workings, particularly in coalfields where pumping has ceased.
References BACKES, C. A., PULFORD, I. O. & DUNCAN, H. J. 1993. Seasonal variation of pyrite oxidation rates in colliery spoil. Soil Use and Management, 9, 30-34. BOLT, N. 1995. The overseas perspective on desulphurisation. In: Desulphurisation No. 4, SheffieM 1995. ICHEME Symposium Series No. 138, 15-20. BRADSHAW, A. D. 8s CHADWICK,M. J. 1980. Restoration of Land. University of California Press,
Berkeley, Los Angeles. BtJRCmLL, P. & WAY, D. S. 1993. The CRE Coal Sample Bank: A Users Handbook. British Coal Corporation, Coal Research Establishment, Cheltenham. CARLSON, C. L. & ADRIANO,D. C. 1993. Environmental impacts of coal combustion residues. Journal of Environmental Quality, 22, 227-247. CASWELL, S. A. 1983. Geochemistry and Mineralogy of Coal and Coal-Bearing Strata from the Cannock Coalfield, with Special Reference to Chlorine. PhD
Thesis, University of Sheffield. CAVENDER, P. F. 8s SPEARS, D. A. 1995. Assessing the
geological controls on sulphur in coal seams: Progress towards predictive mapping. In: Desulphurisation No. 4, Sheffield 1995. ICHEME Symposium Series No. 138, 197-204. CLARKE, L. B. & SLOSS, L. L. 1992. lEA Coal Research. Report IEACR/49, London. COOPER, J. R. & KYLE, W. S. 1995. FGD in the UK the historical perspective. In: Desulphurisation No. 4, Sheffield, 1995. ICHEME Symposium Series No. 138, 59-73. FELLOWS, P. M. 1979. An Investigation of the Water Soluble Elements in Coal and Coal Bearing Strata.
PhD Thesis, University of Sheffield.
E N V I R O N M E N T A L I M P A C T OF M I N E R A L S IN COALS FINKELMAN, R. B. 1982. Modes of occurrence of trace elements and minerals in coal: an analytical approach. In: FILBY, R. H., CARPENTER, B. S. & RAGAINI, R. C. (eds) Atomic and Nuclear Methods in Fossil Energy Research, Plenum, New York, 141-149. FRANKIE, K. A. & HOWER, J. C. 1987. Variation in pyrite size, form and microlithotype association in the Springfield (No. 9) and Herrin (No. 11) coals, western Kentucky. International Journal of Coal Geology, 7, 349-364. HARRISON, J. S. 1995. Reduction of sulphur emissions- the economic equation. In: Desulphurisation No. 4, SheffieM 1995, ICHEME Symposium Series No. 138, 5-13. HJELMAR, O. 1990. Leachate from land disposal of coal fly ash. Water Management and Research, 8, 429-444. HUBBARD, F. H., DHIR, R. K. & ELLIS, M. S. 1985. Pulverized fuel ash for concrete: Compositional characterisation of United Kingdom PFA. Cement and Concrete Research, 4, 185-198. KNEELER, W. A. & MAXWELL,G. P. 1985. Size, shape and distribution of microscopic pyrite in Ohio coals. In: ATTIA, Y. A. (ed.) Processing and Utilization of High Sulfur Coals. Elsevier, Amsterdam, 141-165. LEE, S. & SPEARS, D. A. 1995. The long-term weathering of PFA and implications for groundwater pollution. Quarterly Journal of Engineering Geology, 28, S 1-S 15. LOVE, L. J., COLEMAN, M. L. & CURTIS, C. D. 1983. Diagenetic pyrite formation and sulfur isotope fractionation associated with a Westphalian marine incursion, northern England. Transactions of the Royal Society of Edinburgh: Earth Sciences, 74, 165-182. MORRISON, J. L., SCHEETZ, B. E. STRICKLER, D. W., WILLIAMS, E. G., ROSE, A. W., DAVIS, A. & PARIZEK, R. R. 1990. Predicting the occurrence of acid mine drainage in the Allegheman coal-bearing strata of western Pennsylvania; an assessment by simulated weathering (leaching) experiments and overburden characterization. In: CnvI, L. L. & CHOU, C. L. (eds) Recent Advances in Coal Geochemistry. Geological Society of America, Special Paper, 248, 87-89. MURDOCK, D. J., Fox, J. R. W. & BENSLEY,J. G. 1994. Treatment of acid mine drainage by high density sludge process. Proceedings of the International Land Reclamation and Mine Drainage Conference and the Third International Conference on the Abatement of Acid Mine Drainage, Vol. l: Mine
295
Drainage. Pittsburgh, P.A., April 24-29, 1994. U.S. of Mines Special Publication SP 06A-94, 241-249. PIKE, S., DEWISON, M. G. & Spears, D. A. 1989. Sources of error in low temperature plasma ashing procedures for quantitative mineral analysis of coal ash. Fuel, 68, 664-668. PYE, K. & MILLER, J. A. 1990. Chemical and biochemical weathering of pyritic mudrocks in a shale embankment. Quarterly Journal of Engineering Geology, 23, 365-381. RIPPON, J. H. & SPEARS, D. A. 1989. The sedimentology and geochemistry of the sub-Clowne cycle (Westphalian B) of north-east Derbyshire, U.K. Proceedings of the Yorkshire Geological Society, 47, 181-198. SPEARS, D. A. 1987. Mineral matter in coals with special reference to the Pennine Coalfields. In: SCOTT, A. C. (ed.) Coal and Coal-Bearing Strata: Recent Advances, Geological Society, London, Special Publication, 32, 171-185. & CASWELL,S. A. 1986. Mineral matter in coals: Cleat minerals and their origin in some coals from the English Midlands. International Journal of Coal Geology, 6, 107-125. -& MARTINEZ-TARAZONA,M. R. 1993. Geochemical and mineralogical characteristics of a power station feed coal, Eggborough, England. International Journal of Coal Geology, 22, 1-20. TALBOT, R. W., ANDERSON, M. A. & ANDREN, A. W. 1978. Qualitative model of heterogeneous equilibria in fly ash pond. Environmental Science and Technology, 12, 1056-1061. TAYLOR, R. K. 1988. Coal Measures mudrocks: composition, classification and weathering processes. Quarterly Journal of Engineering Geology, 21, 85-99. & SPEARS, D. A. 1970. The breakdown of British Coal Measures rocks. International Journal of Rock Mechanics and Mining Science, 7, 481-501. WARD, C. R. 1989. Minerals in bituminous coals of the Sydney basin (Australia) and the Illinois basin (USA). International Journal of Coal Geology, 13, 455-479. WHITE, R. N., SMITH, J. V., SPEARS, D. A., RIVERS, M. L. & SUTTON, S. R. 1989. Analysis of iron sulphides from UK coal by synchrotron radiation X-ray fluorescence. Fuel, 68, 2480-1486. VAN DER SLOOTT, H. A., WIJKSTRA,J., VAN STIGT, C. A. & HOECKE, D. 1985. Leaching of trace elements from coal ash and coal-ash products. Wastes in the Ocean, 4, 467-498. YOUNGER, P. L. 1994. Minewater pollution: The revenge of Old King Coal. Geoscientist, 4(5), 6-8. -
-
-
-
A well logging method for the determination of the sulphur content of coal seams by means of deep gammaspectrometry ViT GREGOR ~ & ANTONiN
TI~2KY 2
1Podpgrova 5, Brno-Medl6nky, 62100, Czech Republic 2 Lacinova 2, Brno-l~edkovice, 62100, Czech Republic Abstract: This contribution describes new spectral logging equipment developed at Geofyzika Brno which is designed to remotely determine the sulphur content in brown coal. The method is based on analyses of the prompt gamma ray spectrum induced in coal seams by a neutron source. Theoretical studies have demonstrated the possibility of using the 5.42MeV peak of gamma radiation from sulphur. The measured spectrum is analysed in three independent parallel ways. The analyser is calibrated in the range from 0.6% to 10% of sulphur by weight. The lower detection limit is believed to be 0.3-0.5%, and the measurement sensitivity 0.1%. The undesirable consequences of the presence of sulphur in the raw materials of energy production are well known and therefore it is not necessary to explain the significance of such a problem at a time when the ecological aspects of the industry are being monitored very closely. At the beginning of the chain of energy production from brown coal there is the emission of sulphur dioxide. Therefore the task arose how to determine the amount of sulphur in specific parts of brown coal beds. Only on the basis of such information is it possible to decide whether particular parts of coal seams are suitable for mining and which method of desulphurization to use. The most convenient means of obtaining this information is from geological boreholes. Such direct measurement of sulphur has only become feasible with the advance of the theoretical and technical tools of nuclear physics during the last twenty years. Nargolwalla et al. (1977) stimulated interest in the possibility of such measuring equipment, but in spite of verification of the principles of his method in the laboratory, a successful field logging application has not yet been realized. In 1992 a special gammaspectrometer for detection of the prompt gamma-ray spectrum generated by the capture of thermal neutrons was developed as the result of an initiative by VGP (Borehole and Geological Prospection) Osek and Geofyzika Brno. This spectrometer probe was certified with very good results on artificially prepared laboratory samples. Research project No. 105/93/1272, supported by the Grant Agency of the Czech Republic continued in 1993 with the work mentioned above. The object of this project was to develop a practical well-logging measurement system for the determination of the sulphur content of coal beds. The project considers both the field applications of the spectrometer and the
development of new mathematical methods of interpretation of the spectra, with the purpose of improving considerably the selectivity and the accuracy of the measurements. In view of the limited extent of this paper we do not publish all the data that form the basis of our conclusions. This was done in the final publication of the project.
Detection of sulphur concentration in coal seams: accuracy and representation The basic consideration in the research of each new measuring method must start from a known and/or hypothetical distribution of the material to be measured in the volume under study, with a set target in terms of the quality of results of the measurements to be achieved (accuracy, reproducibility, sensitivity). If the interpreted results of the measurements have to represent real properties of the volume under study with sufficient accuracy the two above-mentioned conditions must be mutually dependent. Information about the distribution of sulphur in each ten-centimetre long segment of the coal bed in both the horizontal and vertical directions can be obtained from the following data, which illustrate a one-metre long interval of the randomly chosen coal seam. The concentrations of sulphur in the horizontal direction are (in %): 1.9; 2.13; 2.28; 1.99; 1.27; 1.16; 0.83; 0.88; 1.05; 1.12 (the average concentration in the whole one metre long interval is 1.5%-I-0.6%). The concentrations of sulphur in vertical direction are (in %): 2.96; 5.26; 4.49; 5.94; 7.12; 2.65; 4.15; 4.48; 3.40; 2.21 (the average concentration of sulphur in the whole one metre interval is 4% • From this data it can be seen that the individual values fluctuate considerably and the probability
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 297-307.
298
V. G R E G O R & A. TI~2K~:
(a)
..-.-.. S A M P L E S ...... P R O B E
SULPHUR CONTENT
(%)
12-
I0'
I l
!
i
o
j.
l
J
i
l
I
I
,
2
!
J
j
!
!
8
!
J
!
J
j 5
4
i
a 0
DEPTH. (m)
(b)
SAMPLES PROBE
SULPHUR CONTENT
(~)~
4-
I \ 2.
" ~
"~."~
~_._~
0 0
HORIZONTAL
Fig. 1. Sulphur distribution.
DISTANCE
(m)
~Z~
7
SULPHUR CONTENT BY GAMMASPECTROMETRY that any of them would represent the average value of the sulphur content in the whole one-metre segment is very low. In the case of five-centimetre segments, the representation of individual values is much more unfavourable. For laboratory analyses on core samples it is typical to measure the investigated volume in only the first cubic decimetre. Because of the lateral and vertical variations in the sulphur content of coal seams it is necessary to take much bigger sections of the coal seam for sulphur analyses. One result of the study is that through increasing the number of the analysed volumes a much better representation of the results is obtained, because the average value of the sulphur concentration in such big segments is more constant. This can be reached in practice only by means of well-logging. Furthermore, it is unreasonable to consider as valid for seam quality evaluation those values in lower orders than units of wt% of sulphur. This is because of the real distribution of sulphur in coal seams. V. Prokop (1994, pers. comm.) carried out a statistical evaluation of the mutual dependence of sulphur content data that were obtained from laboratory analyses of cores and from welllogging of the same boreholes. Figure la shows the sulphur distribution in coal matter when sampling at equidistant depth steps was used. There is an example of a function in the same diagram which shows how well-logging measurement with its better radius of investigation reflects the same coal bed. Figure l b shows the situation in horizontal direction of a coal seam in a similar way (the samples were taken from a surface of the bed). Our own observations confirm published information that coal seams are formed by layers in a predominantly horizontal attitude. Such a geometrical relationship is another important reason for using a well-logging method which has a much larger radius of investigation. The obtained evidence indicates the following design requirements. The tool used in welllogging for sulphur concentration measurements must provide results from approximately onemetre long segments of the coal seam. The concentrations of sulphur in coal beds of SHR (North Bohemia Coal Basin) vary within a range from 0.1% to about 20%. The calibration range of the tool must cover this range. Also, the tool must be capable of achieving reproducibility to within one tenth of one per cent of sulphur content, which is one numeric order lower than is necessary for interpretation purposes. It is assumed that the measurements will take place both in dry and in water-filled, cored or uncored
299
boreholes. The ability to use logging technology in uncored boreholes will lead to a considerable reduction in costs and to an increase of the speed of prospecting operations. In consequence it will be possible to carry out prospecting operations using a denser net of boreholes and by those means to obtain results that are more representative of the actual sulphur concentration in the coal seams under study. To conclude this chapter it is necessary to emphasize the fact that any logging equipment for the direct detection of sulphur content cannot be calibrated from the results of laboratory analyses of samples from boreholes drilled through a coal seam. From the point of view of accuracy it is also impossible to compare the results of these two methods of sulphur content measurements properly. These two methods have very different radii of investigation and operate on different parts of the coal stratum. This fact, together with the above-mentioned variability of sulphur concentration in the seam results in a very low degree of correlation of the results of these two methods. Carrying out the laboratory analyses is convenient in such a case, when we must know exactly the concentration of sulphur in a given sample of coal. Well-logging brings a result representing an average value of sulphur concentration in the vicinity of the borehole.
The physical basis of the method The first experiments to determine coal seam sulphur concentration carried out by means of classic well-logging methods were unsuccessful. At that time the best results were obtained by magnetic methods, thanks to the connection between sulphur content and the presence of iron. Only some of the nuclear methods among which are included theoretical possibilities of sulphur detection remained in the complex of logging methods. X-ray fluorescence analysis seems to be of limited use due to the very low radius of investigation and its very high dependence on a clean and mechanically sound borehole surface. The activation analysis has its spectrum in a higher level but its count rates are very low. The application of spectral analyses of prompt gamma rays generated by the capture of thermal neutrons proved to be the most convenient method. In this method characteristic gamma rays of high energy, up to 10MeV, and of significant count rates are emitted from the chemical elements under study. As a consequence of high-energy gamma radiation only a little loss of energy
300
V. GREGOR & A. Tt~ZKY
takes place and so an adequately large radius of investigation is achieved in this case. The atomic nucleus is excited by the absorption of a thermal neutron and it returns to the stable state in nearly 10-14s with the simultaneous emission of gamma quanta. It is necessary to note here that gamma quanta in a similar range of energies can also arise from the inelastic scatter of neutrons on their path from the neutron source. Some publications (for example Serra 1984) say that in the case of inelastic scatter there are influences especially from such elements as C, O, Si, Ca and possibly Fe and S, which could complicate the analysis of the resulting spectrum in the general case. Using a 252Cf neutron source, the excited neutrons have average energies about 2.3MeV, which is probabilistically too low to cause inelastic scatter. We have never observed peaks corresponding to the energies of inelastic scatter, even in case of very high carbon concentrations. The situation in the spectra corresponding to the mechanism of thermal neutron capture, which we are detecting in coal beds is not quite simple. The following particular influences are included here. (1) The peaks corresponding to separate energies are not straight lines. The reason for this lies in the resolution power of the detector used (in the case of detector BGO (Bi4Ge3012) 2"• 2" the resolution power for 137Cs is approximately 11%). (2) Each characteristic energy of the gamma rays emitted by the elements present is manifested in the spectrum by three peaks: a peak of
full absorption, and two escape peaks. The distance between the full peak and the two escape peaks is one and two times the annihilation energy respectively (so-called single and double escape). (3) Besides the peaks of sulphur, which are our main interest, peaks from other elements, which are present in significant quantities in coal (for example Fe, Si, A1, Mn, Ti, H, Zn, K, Ca), are included in the spectrum. These peaks have a different position and energy heights when compared with the spectra of sulphur. As the consequence of above-mentioned influences there are significant interferences generated in the spectrum. These interferences cannot be resolved without using complicated mathematical procedures. These procedures will be described later. A particularly significant interference is that between the full peak of sulphur, corresponding to an emission energy of 5.42 MeV, and the escape peaks resulting from the presence of Fe. In particular we would mention the single escape of Fe, whose full energies are 6.018 and 5.920MeV. Such a situation is shown in Fig. 2, which is a representation of a spectrum which was measured from a coal seam with a high concentration of sulphur. Sometimes it is not possible to ignore the influence of the contribution of Ti and Mn (There are no visible peaks of Ti or Mn in Fig. 2.). The selection of detectors closely determines the physical basis of the measuring method. The range of detectors available for selection is very limited by the working conditions of the logging
PULSES
i =
**
~
,o
',136
'
',~d
' ' 'l~O
' ' 't~
CHANNEL
Fig. 2. Spectrum from a coal seam with high concentrations of sulphur.
' ~"tl~
' ' '2b," ' ' ' z ~ d
:"
'.'~a'
'
SULPHUR CONTENT BY GAMMASPECTROMETRY probe. It is not possible to use a detector of unsuitable proportions, nor a detector requiring controlled operating conditions in order for it to function. In practice there are two types of spectrometric detectors, either a semiconductor or a scintilator. In spite of some advantages of semiconductor detectors (amongst the advantages is a high-resolution power and a good thermal stability of the spectrum); their use is less convenient especially for the following reasons: all semiconductor detectors in current production need cooling in cryostats; the small volume and density of the detector means that its effectiveness is small particularly in the range of energies that need to be detected. In the case of scintillation detectors, crystals of either NaI(T1) or BGO types may be considered. Crystals of NaI(T1) have much worse spectrometric properties in the prompt gamma rays energy range of interest (approximately 3-10MeV) and much lower detection efficiency in comparison with crystals of BGO. For these reasons and also for the reason of the limited diameter of the logging probe, the 2 " x 2" crystals of BGO have been chosen for our working detector.
The measuring equipment The equipment used in the research of a welllogging method for sulphur detection was constructed at Geofyzika Brno continuing the tradition of the manufacture of nuclear technology in this company. A special spectrometric logging probe, described in greater detail by Gregor & Kagparec (1993, pers. comm.) became the basis of the measuring equipment. The equipment includes a surface module with power supply and an interface for serial communication with the probe, it also includes a computer for checking the measurements and for spectra recording and processing. On the basis of many considerations, calculations and practical experimental attempts we chose the following elements for our work. For the neutron s o u r c e - 252Cf and the scintiblock with a 2 " x 2" BGO crystal (produced by Bicron) as a detector. This crystal has a several times greater volume than the crystals used earlier at FJFI VUT (Faculty of Nuclear Sciences and Physical Engineering, Technical University of Prague) and at GIP (Geoindustria Prague) - it is nearly 10 to 15 times bigger. This fact gives another advantage: lower heights of escape peaks in comparison with the full peaks. This results in a decrease in mutual interference between the peaks of the elements present in the
301
coal matter. Another advantage is in the increase in detection efficiency of the system. There is a further important difference between the probe made by Geofyzika Brno and earlier probes: all the resulting spectrum is measured directly in the borehole now. The spectrum is sent in digital form by means of a serial channel directly to the computer after its detection. A 256 channel analyser controlled by a microcomputer is used for distinguishing the energies of the impulses coming from the detector in the logging tool. The analyser is equipped with a progressive correction for the dead time of the converter and with a time meter for limiting the time of the spectrum measurement. The analyser is controlled by the computer via the serial channel. Such a system has two main advantages; the transmission of the analogue signal from the logging tool to the surface is eliminated reducing potential electronic interference with the signal, and it is not necessary for the probe to be linked to the surface with a coaxial cable. In the case of transmitting the analogue impulses from the borehole to the surface the use of coaxial cable is a necessity, mainly because of the large bandwidth of signal carried. Using the customary four-wire, steelplated logging cables for the operation with our spectrometric probe is sufficient. An important precondition for the successful processing of the measured spectra by means of mathematical methods is perfect spectra stabilisation. The equipment used carries out this stabilization by reference to the position of the very expressive full absorption peak of hydrogen at the energy of 2.223 MeV. The measurement process for the interpretation of sulphur content and for the data manipulation are controlled by an IBM PC computer (a notebook model). The software runs under the MSDOS operating system which is a further advantage of this equipment.
The methods of processing the spectra All the information pertaining to the radiation that originated from the elements present in a coal seam is contained in the spectrum which must be correctly measured and placed on the energy axis. Figure 2 shows an example of a spectrum obtained from a coal seam with a high concentration of sulphur. It can be seen that there is a full peak from sulphur over channel No. 170, and it can further be seen that there is a last significant peak from iron over channel No. 240 and over channel No. 70 there is a peak from hydrogen. If we use only the count rates
302
V. GREGOR & A. TI~ZKY
from selected energy windows for the quantitative interpretation of the spectra we receive a very limited quality of the results - in the same manner as Nargolwalla et al. (1977) and later Drahofiovsk~ et al. (1986). In the latter work there is an assumption that the limit of detection of sulphur is about 2% (if the content of Fe is no more than 1%) in the case of using BGO detectors and where the sulphur concentration is around 3% the relative accuracy of interpretation is expected to be in the range 20-30%. Nargolwalla et al. (1977) came to a similar conclusion. In accordance with this conclusion it is possible to achieve the accuracy around 8% with his equipment but only for sulphur contents greater than 1%. The goal is therefore to find a method for the interpretation of the spectra which is capable of processing a more significant part of the information present in the spectra. A method involving additional mathematical processing of the measured data will primarily make a possibility of the elimination of interferences between different elements and will make possible the interpretation of such quantitative changes in the spectrum that cannot be detected by the usual process of evaluation. Our effort is directed towards obtaining and comparing the results from the following three methods of mathematically processing the spectrum. (a) The method of empirical correction of the standardized areas of the peaks. (b) The multicomponent statistical analysis. (c) The method of deconvolution of the spectra. In detail they are as follows: (a) Minima values of the spectra of prompt gamma rays induced in a coal seam can be expressed by a polynomial function. The constants of such a function vary from one spectrum to another, especially the constant A0 which expresses the shift in the y-axis. The above-mentioned function describes a basic line which is fixed on characteristic minima of the real spectrum. The area of the analysed peak of sulphur is then standardized by means of the function value of the basic line in the x-coordinate under the maximum of the peak. The value obtained is corrected further according to the height of the nearest peaks in the spectrum. In this way it is possible to remove interference and to find the corrected value of the peak corresponding to sulphur. In practice the influence of Fe (5.920, 6.018MeV) and in some cases also that of Ti and Mn is eliminated, which is of great advantage. (b) The multicomponent statistical analysis is based upon the unit spectra of all those components under investigation (one component is sulphur, the other components are elements
present in the ash) which are detected in a coal of known composition. Such a condition is fulfilled by means of a set of standards with one component of the spectrum predominant in each of the standards. Then every new spectrum analysed is fitted to these unit spectra to provide a minimalization of the deviations of derived linear combinations of unit spectra from the spectrum being analysed. By such a process we can derive coefficients for each component that are proportional to the concentrations of the elements present in the coal. The main advantage of this process is the opportunity it provides for removing a significant amount of the mutual interferences created by the various elements present. As the method makes it possible to take broader ranges of spectra for calculations, and the quality of the fit increases as a consequence of improved statistics, there is a significant increase the uniqueness of the solution. In such a manner we can lower the detection threshold to concentrations of sulphur in coal of under 1%. This method was successfully used by Geofyzika Brno in solving the problem of determination of U, Th and K, and it is still used in their field gamma spectrometer model GS 256. (c) The method of spectra deconvolution is based on theoretical calculations which are based on the unique physical/mathematical parameters of the detection system (present in the probe). A matrix of the responses to all the spectral energies belonging to the isotropic field of the emitter is calculated. By means of the deconvolution we try to create undistorted primary spectra from the measured spectra which have been distorted by measuring tool influences. In an ideal case it means that the spectrum detected by means of the BGO detector will be changed by means of deconvolution into a line spectrum in which the height of the lines is directly proportional to the concentration of any given element. In the real situation the escape peaks and the interferences connected with them are removed from the spectrum and a considerable increase in the ability of the whole measuring system to differentiate individual peaks is evident. Common to all these methods is the problem of standardization in respect of the intensity of the average flow of thermal neutrons which produce the gamma radiation in the rock matter under study. This average flow depends on the activity of the neutron source used, on the density and moderating characteristics of the strata in the vicinity of the probe, and on the borehole geometry in the vicinity of the probe. The average flow of thermal neutrons cannot be measured directly with sufficient accuracy in the field. It is considered that the only possibility of solving the
SULPHUR CONTENT BY GAMMASPECTROMETRY problem is by deducing this quantity directly from appropriately chosen intervals of the spectrum obtained. We have successfully verified this in principle whilst obtaining the first measurements from the models during the testing the logging probe (Gregor & Kagparec 1993). In the case of two last methods of interpretation it will be possible to add to the data concerning the sulphur concentration some information about the concentration of other elements in the ash, such as for example Fe, Si, and possibly Ti.
Discussion of results achieved on artificially prepared models and in the field The logging probe developed by Geofyzika Brno was calibrated according to the sulphur content on six artificially prepared models of coal material. Each of these models had a nominal and uniform concentration of sulphur throughout the volume. The values of the concentrations in the artificial models were between 0.33% and 10% sulphur by weight. The spectra of prompt
STANDARDIZED
303
gamma radiation detected from the separate models were used in the construction of calibration curves for the three interpreting methods described above. (a) The calibration curve for the method of empirical corrections is shown in Fig. 3. There the concentration of S (in %) is plotted on the x-axis, and the standardized response read from spectra on the y-axis. Also shown are the graphs which represent an approximate line of the calibration curves for the cases of logging in dry and in water-filled boreholes. The points for drawing these curves are based on average values of the standardized responses read at the depths of samples for which the results of laboratory analyses on sulphur lay in a narrow range close to some chosen values of sulphur concentration. Experimental measurements took place at the experimental base at Most (the coal mine CSA). It will be necessary to continue with these experimental measurements to obtain better accuracy of the calibration curves. We anticipate reaching an accuracy of interpretation of around 0.3-0.5% in the range from 1% to 15% by weight of sulphur content using the analogue
RESPONSE
1.4
1,3 models
1,2
1 dry
I~
J
1.0 0
i
I
i
I
i
I
i
I
I
i
i
2
3
4
5
6
7
8
9
I0
SULPHUR
CONCENTRATION
Fig. 3. Calibration curves - empirical corrections method.
(%)
well
304
V. GREGOR & A. TI~ZK'Y
method of evaluation with the calibration plot (Fig. 3). We expect much higher accuracies in the case of computer interpretation. (b) The calibration graph obtained by means of the multicomponent statistical analysis is shown in Fig. 5. The interpretation of the obtained spectra was carried out by I. Ka~parec (Geofyzika Brno). The concentration of sulphur in the artificial samples is plotted on the x-axis, similarly to the above-mentioned example. The sulphur concentration evaluated by means of the given method is plotted on the y-axis. We evaluated four different sets of measurements of the spectra from the individual models. The course of the calibration curve itself and the relatively small deviation of the data from the line prove this analysis to be correct. We assume the lower limit of resolution is around 0.3% of sulphur concentration and the sensitivity of measurement is round 0.1%. This method also enables the determination of the content of some other elements present in the ash Fe, Si, Ti - because the concentrations of these elements are the remaining components needed for calculations of the sulphur content. (c) The calibration graph obtained by means of the method based on deconvolution is shown in Fig.v4. Calculations were made by J.Klusofi (FJFI CVUT). To achieve higher accuracy in the
results it is necessary to consider increasing the quality of the primary spectrum (in terms of its energetic stability), achieving a more precise analytical expression of the energetic dependence of the resolution power of the detection system, a more detailed study of the speed of convergence, and the removal of the influence of irregular jumps at the ends of spectra. Figure 6 shows an indisputable advantage of this method. The primary spectrum obtained from the model with a 10% concentration of Ti is shown in Fig. 6a and the same spectrum after the deconvolution is shown in Fig. 6b. As we can see from these graphs, the superposition of the escape peak has disappeared from the spectrum and an increase in resolution power has occurred. In consequence the status of the interval of the Ti peaks (6.76MeV and 6.42MeV) has closely approached the form of a line spectrum. The spectrum after deconvolution therefore raises the possibility of distinguishing and quantitative evaluating of more chemical elements than the primary spectrum. The reproducibility of measurements is an important criterion for the use of the equipment in the field. A basic set of such attempts was made at the coal mine (~SA. It was found that an acceptable reproducibility could be achieved if the time of measurement of one spectrum was
RESPONSE
30-
20-
-,1..
10-
0
1
2
3
SULPHUR
4
5
CONCENTRATION
Fig. 4. Calibration graph - deconvolution method.
6 (%)
7
8
9
10
SULPHUR CONTENT BY GAMMASPECTROMETRY 400 s. The stage of reproducibility will depend on the method of processing the spectra. We found that the reproducibility did not noticeable decrease even if the time taken for the measurement of one spectrum was shortening by a half, to 200 s. The effect of statistical fluctuations on the spectrum becomes noticeable when the measurement time is reduced below 200 s. We consider the fact that results from viewpoint of sulphur content determination obtained from both dry and water-filled boreholes are apparently equivalent to be a relatively surprising one. Certain particular properties of the spectra (total number of impulses, the distance of a chosen peak of sulphur from the base line) are different but the interpretation results are practically the same in both cases. This fact, on its own, demonstrates that the probe receives data from a relatively large radius around the detector. A water filled borehole is, for neutrons, an environment with high moderation and absorption ability, but in spite of that there is no noticeable loss of important information from the vicinity of the borehole. Neither was there a decrease in the measure of correlation between the measurements made in the same borehole with and without water.
305
We have not been able to determine exactly an accurate value of the radius of investigation under field conditions. Besides the technical parameters of the probe, an accurate value of the radius of investigation depends to a substantial degree on the physical properties of the surrounding matter (density, chemical composition of the matrix, porosity, saturation). Nargolwalla et al. (1977) considers the radius of investigation to be in the range of 40-50 cm for his probe, Serra (1984) considers the radius to be in the range of 25-37cm for his probe. We registered radiation from a distance of about 50 cm during our experimental measurements. The sensitivity of measurements to changes in the diameter of the borehole is also connected to the radius of investigation of the probe. Changes in the borehole diameter as it passes through coal seams are usually significant. Our conclusions can be demonstrated on measurements taken in a water-filled borehole in the area of Strup6ice. Spectral measurements were taken from between 32.0m and 41.8m depth in the borehole. The diameter of the borehole varies between 0.13m and 0.26m, the greatest diameters of between 0.21 m-0.26 m being recorded in the pelitic parts of the profile. Two minimum
RESPONSE
10 _
8765 4
3 ................................................................................................................................................................................
1 i
0
1
l
2
3
4
5
SULPHUR CONCENTRATION
Fig. 5. Calibration graph - multicomponent statistical analysis.
!
i
6
7
(%)
8
9
10
306
(a)
V. GREGOR & A. TI~ZK'Y PULSES
Ti
....
....
6, 7 6 M e V 6, 42. H e Y
,
2~0
24 0
CHANNEL
(b)
Spectrum
after
deconvolution
CHANNEL RESPONSE 1OOOOO
fl
!
9
i!
"
4OOOO
20000
......
0 1
2
3
4 ENERGY
5
(HeV)
Fig. 6. (a) Primary spectrum and (b) the same spectrum after deconvolution.
values in the range 0.13m-0.14m are in one case in coal of good quality (at c. 36m depth) and in the second case in coal with high ash content (at c. 39.5m depth). It can be seen that the differences between cavitation in good quality coal seam and that in overlying and underlying pelitic rocks are significant and that it therefore helps us to express our opinion on the influence of borehole diameter on the
recorded spectra. The shapes of spectra are considered to be practically independent of changes in borehole diameter. Changes in the shifts of spectra along the y-axis may be compensated by means of the standardization of spectra during its further mathematical processing. Other changes of a similar type resulting from an unequal flow of thermal neutrons may be similarly compensated. Nargolwalla et al. (1977)
SULPHUR CONTENT BY GAMMASPECTROMETRY also indicated only a rather small influence exerted by the borehole diameter on the results of measurements.
Conclusions The results obtained during the solving of problem of sulphur content determination by means of well-logging measurements based on analyses of prompt gamma radiation generated by the capture of thermal neutrons may be summarized in following manner. 1. A deep gammaspectrometer working with detector BGO and with neutron source 252Cf was developed and constructed. Working geometrical and mechanical parameters of the probe were optimized during the laboratory and field experiments. 2. The physical basis of the problem of welllogging sulphur content determination by means of prompt gamma radiation was studied. The peak of sulphur at 5.42 MeV energy was chosen for evaluation. 3. An optimal method of measuring and processing of spectra was designed. (a) The quality of results (sulphur content determination) seems to be equivalent both in dry and water-filled boreholes. (b) The spectra do not depend substantially on the borehole diameters under common field conditions (diameters in the range from 0.1 to 0.25 m). (c) With respect to considered technological conditions the measurements should be done stationary for the time from 200 to 400 s for each measurement. 4. Several ways of interpreting spectra were studied. Three methods were worked out to the
307
stage of calibration curves. (a) The method of empirical corrections of the standardized areas of the peaks based on the elimination of influence of escape peaks of Fe and Ti from the full absorption peak of sulphur. (b) The multicomponent statistical analyses based on fitting of measured spectra from unit spectra of analysed components. (c) The method of deconvolution of the spectra which solves the return to primary, undistorted by measuring instrument, spectra by means of mathematical processing. 5. Achieved results confirm the possibility of sulphur content determination from about 0.3 to 0.5wt% of sulphur with the sensitivity of 0.1% and the variation coefficient of 0.2% provided of measurement in coal seams of a good quality (low content of ash material). 6. The presence of significant peaks of Fe, Si, Ti in measured spectra shows the possibility of detemination of these elements concentrations in ash material in future.
References DRAHOlqOVSKY,R. et al. 1986. Kvantitativni stanoveni siry v hn~d6m uhli SHR primo z karot~2niho m6?eni. Geoindustria Tuchlovice. GREGOR, V. & KASPAREC,I. 1993. V~voj karot~2niho analyz~toru obsahu siry v uhli. MS-Geofyzika, a.s. Brno. NARGOLWALLA, S. S. et al. 1977. Nuclear Melalog Grade Logging in Mineral Deposits. International Symposium of Nuclear Technique in Exploration. Vienna, Austria, March. SERRA, 0. 1984. Fundamentals of Well-logging Interpretation.
The Acquisition
Elsevier, New York.
of Logging Data.
A logging correlation scheme for the Main coal seam of the North Bohemian brown coal basin, and the implications for the palaeogeographical development of the basin KAREL
MACH
North-Bohemian Mines JSC, OHMG of Bilina Mines, 418 29 Bilina, Czech Republic Abstract: A method for correlating the Main coal seam in the North Bohemian brown coal
basin by interpreting borehole logs is described. The most important well log was the gamma-gamma density log, but in some boreholes it was also necessary to use the natural radioactivity and electric resistance logs or technological sampling of the core. All these methods, at various levels of sensitivity, reflect changes in the amount of organic matter in the borehole profile. Vertical changes of coal matter content characterize variations of sedimentation conditions. In close boreholes gamma-gamma density logs are easily correlatable so that changes of sedimentation can be correlated over great distances. The method has been used to correlate the main seam along the axis of the basin. The main result of this correlation is that coal-bearing sedimentation occurred uninterruptedly along the main axis of the basin, with the development of swamp and lake conditions. During the swamp conditions peat accumulated over the majority of the present basin area, whilst during the shallow lake conditions the region of clastic sedimentation was controlled by the distance from the inflowing sediment source. The coal-bearing sedimentation cycle started by a levelling of the palaeorelief by infilling the depressions, and was followed an extension of the sedimentation area. Compaction of the previously deposited layers affected the ensuing strata. No significant variations in stratal thickness were found along the length of the main borehole line nor in any of the auxiliary lines whole, indicating that tectonic controls in basin subsidence were unimportant. Thus changes of sedimentation conditions dominantly of climatic conditions caused water levels in the basin area to fluctuate. For many reasons it is concluded that the correlatable boundaries are isochronal. The correlation scheme represents the first really objective correlation of the main coal seam over the area of the whole North Bohemian basin.
Variations in the physical properties of coal seams offer a broad range of possibilities for correlating an individual coal seam across a coal basin. Amongst the many measurable properties, are coal micropetrography, analyses of ash content, combustible matter, water, sulphur, arsenic and many other trace elements, micropalaeontological analyses, isotopic analyses - all of which result from analyses of samples taken either from the drill core or from outcrop. The main problem hindering successful utilization of this data for the purposes of a detailed correlation of seam strata is the degree of subjectivity involved in sample taking. The seam is usually sampled according to its macroscopically distinct properties, and this process will often unify microscopically distinct rocks into greater intervals, and thereby result in a loss of important distinguishing detail. This practice is encouraged both by the instantaneous requirements of production practice and by the pressure to save costs on the geological survey. Some of the sampling methods used do attempt to remove the subjectivity by systematic sampling at set intervals, for instance by chip samples. Such point sampling methods will work only with
some rock types, and for that reason it may not be possible to extend the analysis results far from the sample point. A further problem is that of core recovery and core quality, in that the core presented to the geologist often bears little relationship to the strata in the ground. The small number of boreholes from which samples have been taken for analysis also makes correlation of the seam by its properties difficult. M a n y of the above-mentioned deficiencies may be overcome by logging methods involving nondestructive 'sampling' of borehole walls. The subjectivity factors diminish here and the network of logging bores within the framework of the North Bohemian brown coal basin (NBB) is very dense. Geophysical logging methods, in contrast to sampling for analysis, are dependent on measurement of such physical properties of rocks as electrical resistance, magnetic susceptibility, natural radioactivity, thermal conductivity, reaction to irradiation by various types of incident radiation etc. The basis of the interpretation of the geophysical logging records of holes from which no core was taken has to be the geophysical logs of those boreholes from which
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 309-320.
310
K. MACH
core was taken for analysis. Only where such correlation is possible throughout an area can geophysical logging methods be used with confidence as a method to replace coring. Because of the utilization of coal from the NBB as a major source of energy, there exists a dense network of boreholes with analysis data relating to the energy generating properties of the coal, e.g. ash content, water and calorific value. From the usability standpoint for correlation, the most suitable analyses result from the detailed sampling for ash content (calorific value and water content are to a certain measure functions of the ash content). Many other physical properties that are measurable by geophysical logging methods are also a function of the components of the ash material in coal. These are above all the density and natural radioactivity and to great measure the electrical resistance. Other geophysical logging methods are complementary and their use is restricted to strata differentiation outside the coal seam. Observation of seam strata in any open pit indicates that attempts to achieve a detailed subdivision of individual coal seams will be best aided by those geophysical logging methods whose results are proportional to the ash contents. The changes of sedimentation conditions in the NBB were recorded, above all, on the basis of the changes in ash content of the
coal seam. Because of the problems of subjectivity in core sampling it is considered that the determination of the variation in ash content through the seam section is more effectively achieved by the method of density gammagamma logging. The network of density gamma-gamma logged boreholes is comparable with the network of cored boreholes and the interval of logging measurement in the seam is of the order of centimetres (compared to sampling intervals in the cored seam sections of the order of decimetres or metres). Fluctuations in seam ash content are therefore mapped with maximum detail both in plan and vertically. For this reason the method of comparison of curves of the density gamma-gamma logs is used as the principal method of correlation of coal and noncoal strata in the coal seam in the NBB. The potential of the method is illustrated by reference to several boreholes in the vicinity of Bilina mine (Fig. 1). The profiles of the curves of adjoining boreholes are very similar so that the correlation of single coal and non-coal strata is possible despite fluctuations in thicknesses of individual strata. The factor that limits the distance over which reliable correlation might be is the high variability of the thickness of the succession in which the coal occurs. In the NBB the thickness varies from several tens to several thousand metres. The potential of the method
B
B PA179
E,AE5
JU535
JU$3?
JU$t5
JU~li3
JU/e~
E3"T75
SS'l"/;
I
LP237
Fig. 1. Opencast Bilina part (B-B ~) of logging correlation scheme (curves of density GAMMA-GAMMA logging from several bores with marked corellatable boundaries): 1, various correlatable boundaries; 2, gamma-gammadensity log (density growths to the left).
NORTH BOHEMIAN BROWN COAL BASIN LOGGING CORRELATION for correlation is limited by underground mining of the seam, tectonics affecting the seam, and mining of the seam by opencast methods before geophysical logging was introduced. In such areas it is necessary to use whatever information is available from any of the other above-mentioned methods (ash contents, electroresistance logging).
D a t a used
The course of the line of boreholes used for construction of the logging scheme section is the result of their selection from a great amount of available data according to several conditions. The adjacent profiles of boreholes had to be distinctly correlatable, the line should cross the major seam structures and at the same time, at least by its branches, run through the mining areas of operating giant opencast mines. The seam sections in areas of underground workings were, as far as it was possible, determined from boreholes that intersected pillars. Areas of total extraction of the seam, if they did not provide at least the results of detailed sampling of drill core, were left out. For tactical reasons the areas of so-called abnormal seam development at the Most-Bilina-Duchcov margin of the basin were not included on the main section line. Abnormally great thicknesses of the seam in these areas often created coal bonanzas that attracted the miners during the early stages of the coal mining industry in the NBB. Unfortunately, this early extraction of the thicker coal has lead to great problems during the recent collection of survey data. In spite of this shortage of data comment is made in the conclusion. An attempt has been made to include on the section the line of all the borehole profiles that are considered to cover the coal-bearing strata down to the apparently pre-Miocene underlying strata. Most boreholes in the NBB do not, however, fulfil this condition so that it was necessary to use also many incompletely drilled boreholes. The final course of the section line after the fulfilment of all the conditions could not be a straight line and it approaches the axis line of the basin only in a zigzag way (Fig. 2).
Result of correlation - isochrons
The result of this study is the first objectively constructed correlation scheme of the main coal seam across the entire NBB. The full section, because of its size, is not capable of publication
311
and for that reason the paper will summarize the consequences of the scheme for our knowledge of the formation of coal-bearing strata in the NBB. The scheme produced is confined to the period of origin of the coal seam and for that reason it does not represent later geological events (besides the diagenetically conditioned compaction of the sediments). The main finding is that it shows a significant palaeogeographical basin infilling. The presence of similar logged sequences of strata in adjacent boreholes and the ability to correlate these sequences over distances comparable with size of the whole sedimentary basin show that the vertical variation in ash contents in the seam is not random from borehole to borehole but systematic across the NBB. Above all it is evident that the changing sequence of sedimentary conditions that is correlatable across such great distances have isochronal character. In other words while the occurrence of one clay or coal strata in the boreholes in the area does not need to be interpreted as the consequence of one palaeogeographical event since it can be accounted for by a gradual facies change, the repeated occurrence of such distinctly comparable strata as that indicated by the scheme, forces such a conclusion. This conclusion is intended to resolve a discussion that has for many years involved geologists engaged in determining the geology of the NBB. The principle has for some time been acknowledged in the Chomutov area of the NBB where the coal seam is divided by clay interlayers (Zima 1986; Hor6i~ka 1988; Ov6arov 1988). However, it is not as easily seen in the area of 'uniform' development of the seam to the northeast of a line from Most to Albrechtice. The detailed development of the coal seam is concealed here by the diminishingly small thickness of the clay interlayers visible in the Chomutov area, and the geological context is in places complicated by the above-mentioned 'abnormal' seam development, both of which have resulted in the formulation of incorrect hypotheses. Correlation schemes in this area, for those reasons, mostly consider the seam to be split into three leaves according to the technology of 'mining' methods (Vficl 1989; Brus et al. 1987; Hokr in Malkovsk) 1985). This is not the first attempt to apply the above-mentioned principle in this area of the NBB, but the work of specialist geophysicist has been overlooked by the volume of work of geologists (D/tfia in V~icl 1989). Information from the thousands of logging curves was consequently not properly utilized in this area. The present correlation scheme was only applied to this area in 1993, and it is considered that the submitted scheme
312
K. MACH
--q'-,g"
,
.,
~:~sw,Tz.-..,._-..,*usTl'A4 .u,~A~v/ ITALq ~.~ -" J
9
I
~ ~ --~ 9
m,~
North-Bohemian
I
'
~"~, i ~"~ .,
t,J
brown-coal basin
~
o,
/1'
- ~-"-ZY-"~
( r~
o
~o =,,,,
Fig. 2. Orientation map of North-Bohemian brown-coal basin map with marking of bore lines, used in submitted correlation scheme: 1, territories of cities; 2, borehole used in scheme; 3, borehole profiles presented in this paper; 4, borehole profiles used in shorted scheme (Fig. 3). I-IX, other mine localities I, Bilina Mine; II, Nfistup mines; III, Libou~; VI, Merkur; V, Sverma; VI, VrMny; VII, VCSA; VIII, Kohinoor; IX, Chabafovice. X-XIV, villages X, Albrechtice; XI, Hole~ice; XII, Nov+ Sedlo; XIII, Kundratice; XIV, Vysok~t Pec; XV, Kfem~; XVI, VIII, Marifinsk6 Rad6ice).
redresses previous imbalances in the consideration of the available data and resolves the longterm dispute in a satisfactory way. The conclusion that the whole sequence represents a single depositional event presupposes that single strata that are part of the given sequence and can be correlated for great distances could be considered as isochronous horizons. The profiles of the logging curves enable single strata to be traced from borehole to borehole even where the thickness changes. At the same time the gradual changes in the petrographical composition of single strata do not prevent their being traced for considerable distances. The ability to trace
given strata ends when their thickness reduces to zero or their petrographical composition closely approaches that of the overlying or underlying strata. From the correlation scheme it is evident that the thickness of each coal stratum is very consistent over great areas while the thickness of non-coal strata can quickly fluctuate from centimetres to several tens of metres. It is also the case that the petrographical composition of coal strata is more consistent. The coal strata consist of various types of coal, clay coal and coal claystones. On the other hand, the non-coal strata cover the broad spectrum of clastic rocks
N O R T H BOHEMIAN BROWN COAL BASIN LOGGING CORRELATION AHIOO
KB107
. I,.ERKUR"
,, LIBOL~"
•
2
B
DO 31B7
PE.30
OROUZtqOVICEPEer'ICE
V1~194 ,,, VR.~NY"
DNZOO .. ~.S~"
MR56
,1O425
, KOHINCX]R" ,, B ~ . I I ~ { ~ J ) "
~ 26 TEPLICE
313 20~
., C H J ~ V I C E "
E
~ 7
~o
72
111111115 NITt]I 773,
7--110
Fig. 3. Shorted logging correlation scheme of main coal seam of North-Bohemian brown coal basin: 1-9 typical correlatable parts of coal seam (this dividing of coal seam is used by miners in central part of basin; 1, upper bench; 2, upper part of middle bench; 3, 4, 5 - 1., 2. and 3. subbenches of lower part of middle bench; 6, upper part of lower bench; 7, lower part of lower bench; 8, upper part of 'lower seam'; 9, lower part of 'lower seam'; 10, thick stratas of lake phase sediments; 11, underground mining; 12, number of borehole; 13, name of locality or 'name' of mine. Position of boreholes see Fig. 2. from claystones to coarse sands including their mineralized f o r m s - sandstones and pelocarbon a t e s - and include types with varying content of material of plant origin. The growth of thickness of non-coal strata is directly proportional to the increase in coarse clastic material in the sediment. From a palaeogeographical view the scheme illustrates growth and development of the coal-forming swamp that existed in the NBB area during the Miocene Period. During the whole period of the formation of the seam
strata in the coal-forming swamp, there was a regular alternation between the more or less undisturbed accumulation of material of plant origin and whole-area flooding of the area by sediment bearing streams. During both the peatbog phases and the very shallow lake phases, the basin area was supplied by streams whose catchment areas are relatively well known from the association of heavy minerals in the clastic sediments (Cadek et al. 1985). The dominant southerly source of clastic material
314
K. MACH
into the basin did not change during the life of the coal-forming swamp in significant detail. What did change, however, were the positions of entry of the stream mouths into the swamp and lake.
Progress of lake phases The positions of channel inflows into the swamp during lake stages is shown by the maximum thicknesses of clay interlayers connected by sand accumulation. At each clay interlayer there can theoretically be found at least one such channel. The documentation from the mining locality of Libou~ and the results of drilling operations in the Sverma a r e a - w e s t (Ov6arov 1988) show that the inflows change their path in the same way as a meandering river. The maximum thickness of an interlayer is present in the channel so
RC135
RC132
KBI08
the axis of sediment contribution is revealed by compression of the former over-sedimented strata sequences. The most typical example of such filling of a channel can be seen in the Libou~ part of the scheme section (Fig. 4). The relationship between the sand-clay and the siltclay sediments in one such stratal thickening is shown in Fig. 5. Figure 5 shows that the attitude of the sand bedding inside the interlayer is unconformable against the clay sediments. The stratification and composition of the sand sediments correspond to the filling of a channel cut below the level of the water into which the stream is flowing (Reineck & Singh 1973). So in the section we can see the filling of the channel that brought into the basin the previously deposited clay and silt material. The sand bed shows a distinct tendency towards an increase in the proportion of coarse clastic material in the earliest filling
KBI0?
NS117
NSl14
X
A
t0lOre Fig. 4. Opencast Libou~ part (A- A') of logging correlation scheme (curves of density GAMMA~3AMMA logging from several bores with marked well correlatable boundaries). Position of this part see Fig.2. *l, various correlatable boundaries; 2, gamma-gamma-density log (density growths to the left).
NORTH BOHEMIAN BROWN COAL BASIN LOGGING CORRELATION
0
20
~0
I
I
I
60 m
I
~1
V-I
315
m 5 Czl6 U-d7
Fig. 5. Documentation of lake phase from the walls of coal faces of Liboug mine: l, sands; 2, sandy clays; 3, claystones; 4, coal claystones; 5, coal; 6, coalyfied stumps; 7, diagenetical faults.
that decreases downstream, with the gradual filling of the channel as stream speed decreases. The following depositional sequence can be deduced: (1)
(2)
(3)
(4)
Flooding; Sedimentation of mostly clay material with the occasional return of peatbog conditions where the water level decreases; Just below the sand bed itself; Subaqueous erosion of already deposited material; formation of bed profile; Further advance of channel face; Sedimentation of channel filling; gradual decrease of rate of bed cut-through; Channel choking up; Decreasing of inflow; start of peatbog conditions.
In a given area these processes repeat one after another, with short periods of flooding followed by the deposition of an interlayer and sand tongue. The significance of this flooding was not great, however, but it documents, together with the coal claystones, stump horizons, and horizons of pelocarbonate concretions, the fact that against the background of a major flood phase there were many small fluctuations in the hydrological regime of the area. The grain, composition and granularity of clay sediment, away from the channel on the bottom of the greater part of the lake are indirectly proportional to the distance from the tongues of the psammites, which is the consequence of a loss of flow speed across the broader lake area. Apart from the differentiation of clastic material according to granularity, the mineralogial differentiation of the clay minerals found in the sediments has been affected by the reaction between lake water and inflow water, as has also the differential laying down of some chemical elements carried in solution. This phenomenon significantly complicates the geochemical correlation of
strata and over areas comparable in size with the NBB basin area it becomes a practical impossibility. The length of tongues of psammites is apparently directly proportional to the length of time for which the lake phase lasted and indirectly proportional to the degree of compaction of the previously deposited strata. The largest most complex tongue formation in the Holegice- Nov~ S e d l o - Kundratice line extended to Kundratice and Vysokfi Pec, even to the margin of the basin. The development of this inflow is substantially more complicated and less well documented than the inflow described from the Liboug locality. The greater complexity is caused above all by the longer period of existence of this inflow. The lake stages were here suppressed many times by abundant vegetation and the psammite beds themselves as a result of this changed position and channel dimensions several times. Although the single beds here are often of greater dimensions and their filling more sandy, their construction is very similar to others elsewhere and they show all the features described from the Liboug locality. Within the sand strata filling the beds it is common to find cross stratification, but in contrast to the Libou~ locality the sand sediments also occur outside the channel beds themselves in the form of concordantly laid strata, passing laterally into silty claystones. The Vrgany sand-clay accumulations may then be considered as the development of a few psammite tongues and associated beds in the framework of one great lake phase. Generally, the described phenomena represent a very specific case of a river delta. The specificity lies in the extreme shallowness of the basin reservoir, in the flow leads, and in the extremely high compaction of the beds filling the base of the reservoir. While the classic delta of the Mississippi type (Reineck & Singh 1973) builds into a practically unlimited sedimentation area in
316
K. MACH
the form of a sea, the delta of a river leading into a coal-forming swamp in its lake phase had at its disposal only a several decimetres deep reservoir and any further sedimentation area it formed itself by compaction of the previously deposited sediments, primarily the peats and clays. The very low water level in the flooded swamp is indicated by the low height of the preserved stump of trees that were covered by the lake sediment in the form of very broad extensions of the intraformation breccias (Mach 1993a) inside the single interlayers. The depth of the lake ranged from decimetres to a maximum of 2 metres. The occurrence of intraformational breccias and coal strata at the same time attest to occasional fluctuations of the lake level leading to the drying-up of extensive areas of the lake bottom. The dynamics of the water flow resulted in clastic material not being deposited in some parts of the lake, so that in those areas the individual coal strata are not divided by a non-coal interlayer. In these cases the inundation event can sometimes be proved by coal micropetrography (Malfin in Vficl 1989; Malfin in Zima 1986), which has revealed the occurrence of thin strata of 'allochthonous coal' at this horizon. Apart from this evidence it is only the weathering of outcrops of the coal seam in opencast mines that demonstrates any stratification in the seemingly monotonous coal. On detailed inspection there is usually nothing on the strata boundary-line that would indicate its presence, although a very thin stratum of more clayey coal or a thin horizon of 'breaking' fusinite is occasionally found. The existence of a clay interlayer over a great area or the continuing of this interlayer as a 'non-visible boundaryline' between the coal strata demonstrates the exceptionally flat surface of the swamp before the flooding. The occurrence of allochthonous coal implies penecontemporaneous erosion of minor elevations in the peatbog by wave activity in the encroaching lake. It is not easy to understand what caused the flooding of such an extensive area of peatbog. And in order to do so it is necessary to return to the peatbog (swamp) phase of the development of the coalforming swamp.
The progress of peatbog phases The extensive development of most coal seams over great areas comparable with the area of the entire coal-forming swamp demonstrates the practically continuous extent of peatbog across the area of swamp. This requires a levelling activity of the preceding lake phase. The water
flow into the basin must have been reduced so that the flora managed relatively quickly to colonise the quite flat drying floor of the swamp. The effect of water flows leading into the peatbog was changed. As a result of the rising mass of peat and plant cover only relatively pure water penetrated into the interior of the peat swamp while the clastic material was laid down soon after inflow into the peatbog. In these marginal areas, in the direction of the water inflow a ring ledge of finely mixed sediments formed at the tongue, indicating regular changes in sedimentation type. The lateral transition of these sediments towards the basin into coal strata of equivalent age is gradational and relatively rapid so that the borehole logs commonly note the sudden change. The characteristic signs of sediments of this continuous inflow are, besides the rhythmicity, the alternation between clastic sediments and sediments of mixed material. The latter are composed of plant fragments and clastics, with the presence of stump horizons, prints of water plants including root parts, root soils, and the occasional tests of fresh-water bivalves together with abundant lenticular sideritic sections with thickness up to 50 cm. The stump horizons are accompanied by sub-horizontal to obliquely laid stem parts of trees, bearing the evidence of a long-term of decomposition in air (the charred stems are hollow and lack the rind). From a palaeobotanical standpoint it is the growth which periodically occurs on this tongue that stretches into bog and is the same as the growth on the adjacent peatbog. The occurrence of these sediments laterally splitting the coal strata, was found in the correlation scheme at two places. However, only one can be demonstrated in outcrop at an opencast mine. At the present time in the detailed borehole survey there is another feature of this type in the foreland of Vr~any mine. It is considered to be of the same character by analogy with the Bilina mine, where the phenomenon is observable over a length of several hundred metres. The character of the sediments is very similar; this inflow existed for a period several times longer and persisted over many lake phases of swamp development. By interpretation of the changes on the geophysical logging curves it is possible to detect the beginning of this type of sedimentation. The reliable identification of the termination of this phase, however, is not possible without the detailed description of drill core, because the accumulation of sediment from permanent flows in the peatbog phase can pass without noticeable change into the accumulation of sand-clay material in the lake phase (and vice versa) and
NORTH BOHEMIAN BROWN COAL BASIN LOGGING CORRELATION this transition is not manifested in any way on logging curves. The behaviour of sediment in the vicinity of such features during compaction under the load of overlaying strata causes the development of obliquely bowed strata, diagenetical polished surfaces and other effects that might otherwise be explained by erroneous hypotheses.
Causes of regime changes There still remains the question of why the quiet peatbog regime is so often disturbed by the lake regime. By analogy with some recent local peat swamps (Titov 1952) it follows that the development of a peatbog of a size corresponding to the NBB area must be interrupted by occasional flooding for a very simple reason. The marginal parts of a peatbog are so efficient a preventing the inflow of mineral nutrients to its interior that during the amassing of peat, nutrients become exhausted and increments of plant mass are near to zero. The degradation of plant cover leads to changes in the rate of evaporation from peatbog and to terrain differentiation, caused by small fluctuations in mineral contents in the substrate. The peatbog stagnates and on its surface lakes form that become connected into great water areas acting by surge erosion on the formerly relatively dry places. The increase of the area of clear water surface further lowers the evaporation rates from the area and the lake gradually extends to cover the whole area of the peatbog. At this stage the inflow will still be divided from the lake by the barrier formed by that area of the swamp still provided with nutrients by the inflow, but the breaking of this barrier could cause an increase of water level across the whole swamp area. The dimensions of the basin are such that an insignificant growth of water inflow could also result in a substantial increase in the water level of the swamp. The long-term flooding of the swamp floor covering the stump parts of trees and scrub leads to their dying and to the quick decomposition of those parts of the plants that are above the water level. The final destruction of the peatbog could be achieved by surge erosion by the lake. The thin unconnected horizons of fnsinitic coal, often occuring especially in profiles of seams without interlayers are considered to be the result of fires, and fire could be added to the list of factors that might lead to the destruction of the swamp vegetation mentioned above. If it was the case that the basin was not isolated and that an outflow existed somewhere, discharging the water from the swamp, then it could be that the movements of water level in the swamp were caused by
317
changes in the erosion base of this outflow. No such outflow is either visible in the present area of the basin, or has been reliably proved. With regard to the effect of tectonic phenomena on the course of coal-bearing sedimentation, much has been written, but there is little evidence of their ability to effect the relatively quick changes of sedimentation conditions within the framework of the whole basin. Above all the fluctuations in thicknesses of both coal and non-coal strata are mostly explicable in other ways than syngenetic tectonism in the underlaying strata, and it is clear that sudden regionally limited fluctuations of stratal thicknesses would have to accompany these movements. The correlation scheme, however, dooes not give a cover of the whole basin so that it cannot be completely excluded that there exist places where, during the coal-bearing sedimentation, such movements were manifested.
Climatic changes Briefly summarized, there is much evidence that changes in the hydrological regime of the swamp, or in the watershed of the inflows into the swamp, were the originators of changes in sedimentation conditions in the NBB. As a main cause of such hydrological changes, climatic changes must be considered the most probable. Periodical and relatively radical changes of climate have been proved in the Quaternary by many methods (Cilek 1993) and it is unreasonable to believe that such changes did not occur in the Miocene. Some hydrological factors such as for example water evaporation could be in many cases influenced by internal processes, caused by the stagnation of successive increments of plant mass, or from the insufficiency of mineral nutrients, or from fires of the peat swamp vegetation.
Some applications The alternations of peat swamp conditions with very shallow lake conditions is accompanied by the existence of coal-forming swamp from the initial formation of the NBB until the definitive end. Some sections of the lines of the correlation scheme objectively document the fact that the cyclical exchanging of both regimes in the coalforming swamp is accompanied by a gradual extension of the whole basin area. From logging profiles it can be seen that the lowermost strata are limited in the direction of increasing palaeorelief. This phenomenon is the natural consequence of the gradual infilling of primary
318
K. MACH
depressions in the palaeorelief. It is assumed that the total thickness of peat mass and included strata of clay materials in the area of uniform development of the seam in the Most area immediately before the end of the coal-forming swamp conditions reached 250 metres or seven times the present seam thickness (Hurnik 1972). It is no wonder that the total area of the NBB was at that time much greater than today and that the seam found in erosional relicts of this basin contains only its topmost elements. These relicts occur in the Pohradice and K~em)2 area, in areas of volcanic elevations in the Most and Bilina area, but also in other areas. At Bilina mine this phenomenon has been documentated directly in the coal face (Mach 1994). From the correlation scheme presented follow the answers to many questions concerning the course of coal-forming sedimentation. The following represent some of the more important points: where and when the coal-forming sedimentation began and when, where and how it finished (Fig. 6). It is clear that the first coalforming swamps appeared in the areas of Mines Nfistup Tugimice and in the Most area. The cessation of coal-forming sedimentation in the whole basin happened in three stages. The first stage ended the existence of the coal-forming swamp in the Chomutov area, the second in the majority of the rest of the basin and the third destroyed the peatbog in the area of MostRad6ice. Another problem resolved by the scheme is the problem of areas of so-called abnormal development of the seam which are a special case of psammitic tongues, arising in the lake phases of the life of the coal-forming swamp. They owe their development to: (1) (2) (3)
(4)
their position on a divided basin margin; their position in the bottom part of the seam; that besides the period of tongue development of further lake phase they did not bring the greater sedimentation of clastics ('uniform development of seam'); the concentration in these areas of the mouths of inflows carrying the clastic material that ended the coal forming conditions.
The differential compression of the underlaying strata of the coal seam, the close proximity of compacting sand-clay accumulations forming the strata overlying the coal seam, and the compression effect of the psammitic tongues of these inflows led to horizontal movements of the partially consolidated peat mass in the early stages of the end of coal-bearing sedimentation. Although it would be possible to outline a
detailed solution here, it is considered to be beyond the scope of this paper. Finally, with reference to the 'geochemical isochron' proposed by Elznic et al. (1986), its position only locally follows some of the boundary-lines on the proposed scheme section, but over greater areas of the NBB the isochron and the boundary lines intersect. It is considered that the given boundary line, marking the change in contents of some chemical elements and at the same time a change in composition of clay minerals, characterizes not a change of source area but rather a change of position of the inflow source of clastic material relative to a given place. In consequence practically the whole area to the west of a line from Most to Albrechtice is above this isochron and many areas of uniform development of the seam are below the given boundary line. In transitional areas this geochemical boundary line lies within the seam strata and often it cannot be unequivocally placed. In this regard it is relatively well documentated that many chemical elements are either bound to the mineral matter (the ash) or organically to the coal mass (Bougka et al. 1972) and differentation of the clastic material in the process of sedimentation of clay minerals in the basin can be considered as proved (Sloupskfi 1985; Divokfi 1987; Ruck~, et al. 1990). It is considered very probable that the differentiation of many chemical elements occurred at the same time as the sedimentary differentiation of various minerals during the lake stages of the life cycle of the coal-forming swamp . Similar differentiation could also have occurred with the development of some authigenic minerals (carbonates etc.). It is to be hoped that the proposed scheme, intentionally presented without a new scheme of stage names, will be accepted as having proved the technical aspects and will, with further improvement, form the basis of any new division of the NBB. It is hoped that the final products of this scheme will be a series of new palaeogeographical maps documenting the development of the NBB and explaining much that is currently disputed. In perspective a further understanding of the processes leading to the development of coal seams can be seen to be closely connected with the geological processes in this scheme. In this respect, a reinterpretation of existing studies, carried out in accordance with the proposed scheme can bring not only the verification of the scheme, but also increased precision and enrichment of the submitted theory. There is no doubt that a mechanistic use of the scheme method would reveal some mistakes in the detail of the correlation as result of insufficient data.
R
6
Fig. 6. Main stages of rise of coal-bearing sediments in North-Bohemian brown-coal basin: 1, crystalline rocks; 2, Cretaceous sediments; 3, volcano-clastics; 4, volcanites; 5, underlying clastic sediments; 6, peat to coal; 7, overlying and coal seam clastic sediments; 8, direction of water stream; 9, water level; 10, peat forming swamp. (a) Levelling of depression of palaeorelief; rise of first swamps. (b) The regime of coal-forming swamp; gradual increasing of its area. (c) Supppression of coal-forming sedimentation by start of lake conditions.
320
K. MACH
Conclusions Under the conditions of the NBB the interpretation of g a m m a - g a m m a - d e n s i t y logs has proved to be the most useful method of stratigraphic correlation within the main coal seam. The qualitative character of logs gives the possibility of correlating logs measured under different conditions (various method of boring etc.) and correlation of technologically unequal strata. This method eliminates most of the potential subjective aspects, which arise from sampling core. The method with the best results reflects variations of ash contents in vertical profiles of the coal seam. This shows variations of sedimentation conditions with time at the point of boring. Using this method in the NBB it is possible to determine isochronous horizons within the main coal seam stratigraphic interval and thus derive more precise knowledge about the internal structure of the coal deposit. My thanks for help in acquiring logging documentation of boreholes and for accompanying me on visits to NBB mining localities goes to colleagues working in various mining and NBB survey organizations; above all to: I. Strbfifi, P. Sulcek, J. RehoL O. Jane6ek, F. Folt)n, P. Coufal and J. Prochfizka. Special thanks are due to J. Zima and L. Hor~i6ka for enabling me to have access to the adjusted sections of partial correlation logging schemes of some parts of the Chomutov NBB area, and assisting me in their elaboration.
References BOUSKA, V., JAKES, P., PACES, T., POKORN~"J. et al. 1980. Geochemie. Academia Praha, Praha. BRUS, Z., ELZNIC, A., HURN~K, S. & ZELENKA, O. 1987. Geologie oblasti. - XXVI. Celostfitni konference CSMG,VUHU Most, Most.
CADEK, J., DUSEK, P. & ELZNIC, A. 1986. Nov6 poznatky o geologick6m @voji komplexu mioc6nnich sedimentfi severo6esk6 pfinve. - Sbor. V. uhel. konf. pfirodov, fak. (Praha), 21-25, Praha. CILEK, V. 1993. V~,sledky ledov~ho vrtu SUMMIT v Gr6nsku. Vesmir, 72/11,624-627, Praha. DIVOKA, H. 1987. Zhodnoceni nerost6ho slo~eni sedimentfi DJS z hlediska dobyvatelnosti. VI)HU Most, Most. ELZNIC, A. 1986. Problematika fizemniho 6len6ni severo6esk6 hnadouheln6 pfinve. Sbor. VII. uhel. konf. pfirodov, fak. (Praha), 71-74, Praha. HOR(~ICKA, L. 1988. Dil6i zfiv6re6nfi zprfiva fikolu Prun6fov - Chomutov. Geoindustria, n.p. Praha, Praha. HURNiK, S. 1972. Koeficient sednuti n6kter~ch sedimentfi v Severoeeskb hnadouhelnb pfinvi. Casopis pro mineralogii a geologii, 17/4, 365-372, Praha. MACH, K. 1993a. Intraformaeni brekcie na VMG v Bilinfi. Zpravodaj SHD, 3/1993, 11-16, VUHU Most, Most. 1993b. Korelace vrstev hlavni uheln6 sloje mezi lomy Bilina, Kopisty a V(~SA. Zpravodaj SHD, 4/1993, 31-40, VUHU Most, Most. - - 1 9 9 4 . Dokumentace rozgi~ov~.niuhlotvorn+ ba~iny na VMG. Zpravodaj SHD, 3/1994, 24-30, VUHU Most, Most. MALKOVSKY, M. et al. 1985. Geologie severoeesk6 hn6douheln6 pfinve a jejiho okoli. UUG, Academia, Praha. REINECK, H. E. & SINGH, I. B. 1973. Depositional Sedimentary Environments. Springer, Berlin. OV(~AROV, K. et al. 1988. Zfiv6re6nfi zprfiva fikolu Sverma - zfipad. Geoindustria, st.p. Praha, Praha. RUCKY, P., SLOUPSK.~,M. & THIELEV. 1990. Zhodnoceni dob~vac!ch podminek na lomu Vrgany do roku 1990. VUHU Most, Most. SLOUPSKA, M. 1985. Nerostn6 slo~.eni terciernich sedimentfi SHR. VI)HU Most, Most. TtTOV, I. A. 1952. Vzaimod6jstvie rastit61nych soobg6estv i uslovij sredy. Sovjetskaja Nauka, Moskva. VACL, J. et al. 1989. Zfiv6re6n/t zpr~tva fikolu velkolom Maxim Gorkij. Geoindustria, st.p. Praha, Praha. ZiMA, J. et al. 1986. Zfiv~re6nfizprfiva fikolu LibouL Geoindustria, n.p. Praha, Praha.
Seismic monitoring for rockburst prevention in the Ostrava-Karvinfi Coalfield, Czech Republic KAREL
HOLUB
Institute of Geonics, Academy of Sciences of the Czech Republic, Studentskd 1768, 70800 Ostrava-Poruba, Czech Republic Abstract: Seismic activity in the Ostrava-Karvinfi Coalfield is continuously monitored by networks of seismographic stations located both underground and on the surface. Software has been implemented for statistical analysis and various types of data display to aid interpretation. Practical experience obtained in long-term seismic monitoring has proved that it is an indispensable part of the rockburst prevention scheme applied in underground mines in this region. Continuous seismic monitoring enables us to delimit seismically active areas, observe the trends of seismic activity in time and space so that we can make more accurate estimates of local rockburst hazard, and check the effectiveness of the rockburst prevention methods used.
The Ostrava-Karvinfi Coalfield is the southwestern part of the Upper Silesian Coal Basin which extends from southern Poland into the Czech Republic (Fig. 1). This mining district is situated in a geologically complicated region around the contact between two large tectonic units, the older Bohemian Massif and the younger Carpathians Mountains. Seismic activity induced by underground mining in the Ostrava-Karvinfi Coalfield has been observed since 1912, and is superimposed on a background of weak natural seismic activity in the area. The foci of these weak earthquakes are mostly concentrated in the Jeseniky Mountains and in the neighbourhood of Opava and 0esk~ T~in (K/trnik et al. 1954; Prochfizkovfi 1994), as shown in Fig. 1. Long-term seismological observations using a single recording station started in the OstravaKarvinfi Coalfield in 1977, and over the next two decades 42 stations equipped with digital instrumentation were installed. Since seismological monitoring started in this region, the association between mining-induced seismic events and rockbursts causing great damage to mine workings has been proved. Interpretation of observations has made it possible to develop new procedures for objective assessments of rockburst hazard, and consequently to apply rockburst prevention measures in underground coal mines. This paper reports on the procedures currently used as part of the rockburst prevention scheme applied by individual mines located in the eastern part of the Ostrava-Karvinfi Coalfield.
Local and regional seismographic networks Original plans for the distribution of seismographic stations were based on the expected lifetimes of mines in the regions prone to rockbursts. Since 1983, when the strongest recorded rockburst occurred in the (~SA Mine (E-- 10 l~ J) the basic network has been gradually expanded with new types of digital seismic systems. The automation of procedures for detection, recording and processing of seismic events provided by these systems has made it possible to determine the locations of hypocentres and to quantify the seismic energy released by individual events within a broad-band energy scale.
Local seismographic network At present, this network of vertical seismometers represents a widespread system which operates in all mines except for the Franti~e and CSM Mines (see Fig. 2). The seismometers are mostly installed underground, although there are a few on the surface. All stations are equipped with digital instrumentation for application automatic procedures. These include data recording at individual mines, and also determination of arrival times and amplitudes of P and S-waves. After this pre-processing, the data are transmitted by modem connection to the operational centre at the (~SA Mine. Further processing is done there to obtain the hypocentral coordinates, estimated energy and origin time of seismic events, which are then stored in a
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology', Geological Society Special Publication No. 125, pp. 321-328.
322
K. HOLUB %
|
%
~,
%% "%
--NMO VOS ES BASn
Fig. 1. Schematic structural map of the Bohemian Massif (after Dopita & Kumpera, 1993) also showing location of the Ostrava-Karvimi Coalfield.
database. The data resulting from the final interpretation are daily transmitted back to individual geophysical and geomechanical laboratories at the respective mines. In parallel with the geophysical database, a technological database stores data describing the daily positions of the coal faces and the blasting operations performed (location and explosive charge sizes). A more detailed description concerning the monitoring and analysis of induced seismicity observations is given by Holub et al. (1995).
Regional network The regional network consisting of ten threecomponent stations equipped with Lennartz Electronic instrumentation covers the whole mining district. Three stations are situated underground and seven are located at the surface (Fig. 2). The OKC seismic station, which doubles as one of the ten monitoring stations of this network, is located about 20 km west of the centre of the mining area. The instrumentation at each station provides for automatic data recording, pre-processing and radio transmission
to a central laboratory. The results of continuous observations of a regional character are used to create a database of current information, constraining arrival times of P and S-waves, hypocentral coordinates, seismic energy and/or magnitude estimate of the recorded seismic events. This database represents an essential source of information for the geomechanical service of the Mine Survey and Safety institute, based at Paskov, near Ostrava. First arrival times of P-waves recorded by the regional network are also included in the database for the local seismographic network at the (~SA Mine, and are used for hypocentral locations by this monitoring system.
Long-term observations of induced seismicity Spatial distribution of hypocentres Many major faults with different throws and dips, define a complicated block-structure of the Ostrava-Karvin/t Coalfield. The blocks delineated by these major tectonic faults were used as elementary units in the planning of the layout
SEISMIC MONITORING FOR ROCKBURST PREVENTION
!
/
~'~
~ ""-
9 9 .~.
underground
,,
/ DOUBRAVA".] /
; . . . . . i" .-J . 01,~ SA ~.j":
"
9 !
9
'k
i ,
~ 9
9
e-
/
I
i
"
j
9
........ j ,
9
"'"(
..
\..
t
I
I
9 ''-
9.
"x 9
"-/" 7 " N.
,.
\.
0.
I
---...
i P
i
(~SM [3
\.
\
~.KV~TENI
.~km ""' 9 ,
9 '""
i
L....DUKLA . F R A N T I ~ E K
L
F
]
/
0
9
[]
,
9 ~_ . . . . . . . . "_...i" / 9 LAZY .Zl . .~. . . . . . l,---, 9 !' " 9 9 D A R K O V I I . . . . . . . . .
./
[30KC
o
/........
9
F'"-"
?
9
surface
323
9
I
i
.....
,,2"
Fig. 2. Distribution of seismographic stations in the Ostrava-Karvin~ Coalfield: 1, local seismographic network; 2, regional seismic network.
of mines. Each mine take consists of several tectonic blocks, in which individual coal seams are gradually being mined by longwall methods 9 In the mines of interest here, only longwall retreat faces are used, either with caving and/or the low-pressure stowing method 9In contrast to the distribution of earthquake hypocentres along tectonic faults, analysis of the hypocentres of induced seismic events has proved that the great majority of them were concentrated in areas of current mining activity. Only rarely were hypocentres located in the vicinity of major faults, and in those cases could the events be attributed to tectonic movements induced by mining 9 According to the location plots from longterm observations over the whole region, areas with higher concentrations of hypocentres exist. For the correct interpretation of a particular location plot, comparison with contemporary mining activities in the area investigated is necessary. As a result, it has been possible to identify isolated areas of seismic activity where the hypocentres are clustered around the active mine workings within a single tectonic block, e.g. blocks 5 and 7 in the CSA Mine, blocks 2, 3 and 4 in the Dukla Mine, block 7 in the Lazy Mine and block 6 in the Darkov Mine (Fig. 3). By contrast, there were other highly active areas which overlapped several tectonic blocks, e.g. in the vicinity of block 3 in the CSA Mine (Fig. 3). Under the latter circumstances, it was practically
impossible to infer the causes of individual seismic events. This may imply that some tectonically distinct blocks are combined in units of greater size in which mining-induced stresses interact, while other blocks demarcate self-contained areas within which seismic events are induced only by local mining activity.
Temporal changes in the distribution of hypocentres and release of se&mic energy In addition to the spatial distribution of hypocentres, the time-dependent changes in distribution provide very important complementary information concerning the seismic regime. Benioff graphs are usually used to show how seismic energy release varies with time. In these graphs, the so-called strain release is calculated as the square root of the amount of seismic energy released, and is assumed to be proportional to the amount of elastic deformation released from the rocks in the vicinity of each hypocentre. The overall trend is represented by plotting the cumulative amount of strain release, whilst the details are more clearly seen by plotting the daily amount of strain release 9 Distinctively different patterns of seismic energy release occur as successive longwall panels are worked 9 A particularly obvious example is given in Fig. 4, during mining of
324
K. HOLUB
Fig. 3. Location plot of seismic event (E > 102 J) epicentres located during 1992. Mine takes and tectonic blocks are demarcated by thick and thin lines, respectively. In the upper right corner the energy scale is given. adjacent longwall panels in the 7th tectonic block at the CSA Mine. Mining panel no. 17331 induced a very low level of seismic activity and was completed without any occurrence of a large seismic event. In contrast, extraction of the adjacent panel, no. 17332, induced a high level of seismic activity, manifested by a steep slope on the graph of cumulative strain release, exceeding the value of S l i m = 15 J1/2/day in the graph of diurnal increment (Fig. 4) whose definition will be given below. During coal extraction by longwall face no. 17332, only a single strong rockburst ( E = 107j) occurred, in January 1992. Almost a month later, this longwall face was stopped and the release of seismic energy rapidly decreased. This example is in good agreement with general geomechanical considerations, for the operation of the first longwall in the appropriate tectonic block usually proceeds without any problems. A higher degree of seismic hazard may be expected subsequently when mining neighbourhood longwall panels, especially the third or fourth panels in an area when a substantial part of the coal seam in the investigated tectonic block has been extracted and the area affected has been enlarged.
Another example of the process of seismic energy release during longwall operation comes from face no. 13933 at the CSA Mine (Fig. 5). Without respect to the fact that this longwall panel was the first one, mined in the seam no.39 in the 3rd tectonic block, the situation from the viewpoint of geomechanics was here very hazardeous, all preventative measures are intensified. It is worthwhile to mention that this area was affected by the strongest rockburst (E=101~ which occurred in the whole Ostrava-Karvinfi Coalfield in 1983. Further details concerning this longwall panel operation and results of seismological observations have been given by Kalenda et al. (1992).
Frequency-energy distribution This distribution can be described by a formula given by Gutenberg and Richter (1954), which for our purposes can be written as log N - - a b l o s E where N is the cumulative number of events with energy within the prescribed energetic window, E is the amount of seismic energy released in a single event, and a and b are numerical constants. This formula
SEISMIC MONITORING FOR ROCKBURST PREVENTION
325
Fig. 4. (a) Benioff graph and (b) its diurnal increment during mining operations for longwall faces nos. 17331 and 17332 in the 7th tectonic block at the CSA Mine. The period of working for each panel is shown in (a) by lines parallel to the time axis. represents a straight line, whose slope b characterizes the number of weak seismic events relative to the number of strong ones. For fitting real data sets, two different approaches to b-value determination were applied. The first method is based on the calculation of slope b of the straight line using least-squares regression, while the other known as the maximum likelihood method, was suggested by Aki (1965) and Utsu (1965) for analysing earthquakes. The advantage of the latter method is that single strong seismic events (induced events and/or earthquakes) have less influence on the estimated slope b of the regression straight line. The software in use at the operational centre enables the b-values for different regions (the whole coal mine district, mine, or tectonic block) to be calculated. In our computations, all data sets are strictly limited to events having a minimum energy value of 102j. Due to different approaches in the b-value calculation, only slight differences in the resulting values appear to have occurred (Slavlk et al. 1992; Holub 1996).
A typical example of the time-space variation of b-values was the effect of mining operations on seismic activity when the single longwall panel no. 13933 was mined in the 3rd tectonic block at the (~SA Mine in 1990-1991 (Fig. 6). An examination of the b-value variations has revealed that the higher b-values were found in 1989 before starting the mining operations ( b l - 0 . 7 4 - 0 . 8 0 and b 2 = 0 . 6 7 - 0 . 8 0 ) . After coal winning started during 1990-1991, lower values were established (b~--0.38- 0.55 and b 2 = 0 . 2 9 - 0.41). The b~-values were computed by using least squares regression, and the b2-values by means of the maximum likelihood method.
Utilization of long-term seismological observations in geomechanical practice In the Ostrava-Karvinfi coal mines, all seams and mine workings are classified in one of three categories of the rockburst hazard according to the prevailing geological (e.g. type of
326
K. HOLUB
Fig. 5. (a) Benioff graph and (b) its diurnal increment for the 3rd tectonic block at the (~SA Mine before and during mining of longwall panel no. 13933. The period of working is shown in (a) by a line parallel to the time axis. sedimentary rocks, tectonics, depth of the coal seam) and geomechanical (e.g. mechanicalphysical properties of sediments, mined out area and edges of unmined seams in the roof, occurrence of the rockbursts in the past) conditions: 9 (1) Rockburst occurrence is not expected during mining operations even if no preventative measures are applied 9 (2) Rockburst occurrence cannot be excluded during mining operations unless preventative measures are applied 9 (3) Rockburst occurrence is to be expected unless preventative measures are applied. This category includes mine workings where rockbursts have already occurred For mine workings in the third category, extensive active preventative measures are obligatory, e.g. destressing blasting in the seam, drilling tests, camouflet shotfiring of a large amount of explosives in the roof above a longwall face, water infusion, and others. In contrast to active preventative measures, the passive ones, are aimed at specific activities (e.g. mine design, mine field development,
choice of mining equipment and others) which could prevent rockbursts occurrence and/or mitigate their consequences. The criteria for assigning areas to third category are not always in agreement with the real rockburst hazard, resulting in an overestimate of the number of regions in this category. Therefore it was recommended that mine workings should be classified for rockburst hazard using the results of long-term seismological observations. At present, applications of the seismic observations are directed towards the determination of seismically active regions, monitoring the development of seismic activity to assess rockbursts hazard, and checking the efficiency of the measures applied to prevent rockbursts. The objectives of these efforts are: 9 to determine regions where active and passive preventative measures would have to be applied 9 to check the effectiveness of the active preventative measures applied 9 to reclassify mine workings according to the degree of rockburst hazard
SEISMIC M O N I T O R I N G FOR ROCKBURST PREVENTION
[bl = 0,74_+ 0.07 I
1189 * 6189
1bz=0.67 +-0.101 x
x
x
3~
x x x x
bl I
2
3 4 log E ( J )
5
I
I
bz
i
1
6
+ 0.03 "~176176176 I b~bl = 0.44 0.37 _+0.04II
'
2
x
x
'
'
'
'
'
6
[ b l = 0.38 +_ 0.02
I b2 = 0.29 + 0.03
Z X X
'
5
7/9o. 12/s0
x x
'
3 4 log E ( J )
=
x
0.80+0.19[
,b:
}
2
"~
327
X x
2,
0
0
1
2
3 4 IogE(J)
5
6
q
x
x
t
I * 1
I
I
( I ! I 3 4 IogE(J)
:
:
',
!
',
I
5
!
~
1
!
:
:
2
',
t
t
I
3 4 IogE(J)
v9t. 12/91
J
i
I
5
I
l
[
I
=
6
[ bl = 0.55 __.0.03] i
I b~ = 0.41 + 0,031
x
~ I 2
:
1
9 9
I
:
b~ = 0.48 + 0.03[ bz = 0.35 - 0.021
x
:
!
1191 6/91
X
,
0
x
x
x x x
x
~
:
6
0
1
2
3 4 log E ( J )
5
6
7
Fig. 6. Frequency-energy distributions of seismic events located in the 3rd tectonic block at the CSA Mine before mining (January-December 1989) and during operation of longwall face no. 13933 (January 1990-December 1991). bl and b2 were calculated using least-squares regression and the maximum likelihood method, respectively.
On the negative side, all efforts aimed at predicting the times of rockbursts with greater precision are very questionable, in common with the general experience for earthquake prediction. The analysis of long-term observations has shown that for any region of the coalfield a value Sli m c a n be defined. This value represents the maximum diurnal increment in strain release (in units of J1/2/day) for which the probability of rockburst occurrence is negligible. It can generally be stated from our investigations of the Ostrava-Karvin~i coal mines, that the value of Sli m is 15 j1/2/day, i.e. this is considered to be a safe value for continued mining provided that it has not been exceeded during the two preceding months. Once the value S=15j1/2/day is exceeded, the occurrence of a rockburst cannot
be excluded and all active preventative measures must be taken. Values of diurnal increment greater than 15j1/2/day were found in all areas where a rockburst had previously occurred (Holub et al. 1991). As an example of this approach, the reclassification of individual mine workings into lower categories of rockburst hazard has been proposed and in several cases this has been done. However, the diurnal increments in strain release are not the only parameter to be considered. When making such a decision, one take into account many factors, such as the level of the seismic activity, location plots, geological and tectonical situation, and the effects of old workings in adjacent areas. Destressing blasting and camouflet shotfiring of a large amount of explosives are the principal
328
K. HOLUB
preventative measures employed to reduce rockburst hazard. Their effectiveness can be checked by using seismic methods. Whereas we have primarily investigated the effectiveness of the latter method, the investigations in Polish mines are aimed at checking the effectiveness of destressing blasting with charges of up to 300 kg, as reported by Filipek et al. (1992) and Dubinski & Syrek (1994). For our work with camouflet shotfiring in the Ostrava-Karvinfi coal mines, an empirical formula is used for estimating the efficiency of the large explosions (up to 3000kg): ~7 = E s / 2 . 6 Q
where Es is the seismic energy in joules determined by using interpreted data from seismograms and Q is the amount of explosives in kg. It is considered that if V > 3, then the blasting has been effective in releasing accumulated strain energy. The application of b-value estimates to rockburst prediction in the Ostrava-Karvinfi coal mines is still under evaluation at the present time. Preliminary results have confirmed the general validity of the b-value criterion, which hypothesizes lower values in regions of higher rockburst hazard. At this stage of investigation, only qualitative changes in time dependent b-values have been established, as reported, for example, by Gibowicz (1979) and Holub (1996).
Concluding remarks The advantages of long-term continuous seismological observations for assessing rockburst hazard in the Ostrava-Karvinfi coal mines are as follows. Routine application of rockburst prevention measures is based on reliable predictions of hazardeous conditions within the rock mass for those mines where the rockburst hazard exists. All possible measures are taken to increase the safety levels in mine workings where rockbursts represent a serious threat. Efficient application of preventative measures at appropriate times minimizes stoppages in coal production, and represents substantial reductions in costs which have been confirmed by several individual mines. This work was carried out as part of research project No.105/93/2409, which was financially supported by the Grant Agency of the Czech Republic. This manuscript benefited from critical and thoughtful review by N. R. Goulty. The technical assistance of J. Rugajovfi in the preparation of the manuscript is also appreciated.
References AKI, K. 1965. Maximum likelihood estimate of b in formula log N = a bM and its confidence limits. Bulletin of the Earthquake Research Institute, Tokyo University, 43, 237-239.
DOPITA, M. & KUMPERA, O. 1993. Geology of the Ostrava-Karvinfi Coalfield, Upper Silesian Basin, Czech Republic, and its influence on mining. International Journal of Coal Geology, 23, 291-321. DUBINSKI, J. & SYREK, B. 1994. The effectiveness of destressing blasts performed in the Wujek Coal Mine. In: RAKOWSKk Z. (ed.) Geomechanics 93. Balkema, Rotterdam, 59-62. FILIPEK, M., MITREGA, P. & SYREK, B. 1992. An attempt to assess the efficiency of destressing blasting performed in longwalls with caving under high rockburst hazard conditions. Publications of the Institute of Geophysics of the Polish Academy of Sciences, M-16 (245), 319-332 (in Polish). GmowIcz, S. J. 1979. Space and time variations of the frequency magnitude relation for mining tremors in the Szombierki Coal Mine in the Upper Silesia, Poland. Acta Geophysica Polonica, XXVII, No. 1 39-49. GUTENBERG, B. & RICHTER, C. F. 1954. Seismicity of the Earth and Associated Phenomena. 2nd edn, Princeton University Press. HOLUB, K. 1996. Space-time variations of the frequency-energy relation for mining-induced seismicity in the Ostrava-Karvinfi Mining District. Pure and Applied Geophysics, 146, 265-280. - - , VAJTER, Z., KNOTEK, S. t~ TRAVNICEK,L. 1991. Application of results of seismologic monitoring during the operation of mine workings in the Ostrava-Karvinfi Coal Basin. Publications of the Institute of Geophysics of the Polish Academy of Sciences, M-15 (235), 219-228. - - , SLAViK,J. & KALENDA,P. 1995. Monitoring and analysis of seismicity in the Ostrava-Karvin~t Coal Mine District. Acta Geophysica Polonica, XLIII, No. l, 11-31. KALENDA, P., SLAVIK, J., HOLUB, K. ~r VAJTER, Z. 1992. Statistical analysis of induced seismicity parameters in the Ostrava-Karvinfi Coal Basin with regard to the 3rd tectonic block of the CSA colliery. Acta Montana, 84, 85-96. KARNiK, V., MICHAL, E. & MOLNAR, A. 1958. Erdbebenkatalog der Tschechoslowakei bis zum Jahre 1956. Travaux Gdophysiques, 69, N(~SAV, Praha, 411-598. PROCHAZKOVA, D. 1994. Earthquakes in the Jeseniky Mts. in 1986. Travaux Gdophysiques, XXXVI (1988-1992), Geophys. Inst. of CAS, Praha, 28-38. SLAViK, J., KALENDA,P. & HOLUB, K. 1992. Statistical analysis of seismic events induced by the underground mining. Acta Montana, Series A, No. 2 (88), 133-144. UTSU, T. 1965. A method for determining the value of b in formula log N = a - bM showing the magnitude-frequency relation for earthquakes. Geophysical Bulletin, Hokkaido University, 13, 99-103.
An analysis of mining induced seismicity and its relationship to fault zones ZDENl~K
KAL,~B
Institute of Geonics, Academy of Sciences of the Czech Republic, Studentskd 1768, Ostrava-Poruba, 70800, Czech Republic Intense induced seismicity has resulted from long-standing mining activity in the Karvin/t part of the Ostrava-Karvin~ Coal District, Upper Silesian coal basin. Interpretation of mining induced seismic events in combination with other knowledge (geological, tectonic, geomechanical, technological) aids the understanding of failure processes in the rock mass. Seismological observations over a three-year period were analysed. Four sets of mining-induced seismic events have been tested to evaluate the seismicity of important fault zones. It follows from the analysis that the seismic activity on important fault zones occurs only as a consequence of mining activities. Accumulations of mining induced seismic events occur on stress concentrators, which may be geological and/ or anthropogenic structures. Abstract:
The geological and tectonic structure of the rock mass of a deposit is an important factor that influences the origin of seismic events induced by mining activities (henceforth referred to as seismic events). This influence is present in the Czech part of the Upper Silesian basin, where the Ostrava-Karvinfi Coal District (OKR) is located (Fig. 1). The underground mining of black coal has taken place for more than 100 years and during this period, a complicated network of worked-out and caved mine spaces has been created. As a result, a complex induced stress field, variable with time, has been formed.
480000
470000
In places, where the strength of the rock mass is exceeded, failure occurs and energy is emitted in the form of seismic events. Information gained from seismic interpretation (e.g. time of seismic event, origin, location of foci, energy, dimensions of the focal area, mechanism of the focus) and relation to the geological and tectonic structure of the area, the geomechanical situation and existing mining activities produce a complex pattern of mininginduced seismic activity in the area. These data can be used to provide information on the importance of individual structures, the actual
460000
450000
REPUBLIC
I POLAND
/"s" ""'~. ~ ......... .)
,I
9
Is
~
.11ooooo
,"
'~..... _
9
,-..,;
:~ .... .
r-,-'"T,,.;
z~ .....
""]"'-.
9. . . .
~,.,.,;"'_,,.,,-,7~. ;-1-h ..... i
~
i" 6 ~ 7 I 8 ...... ~' ], ' . . . . . iI_ . . . .
"~
,i/?
"~,
ji
.r . . . . . . . .
~.-.~
t
I'
(
i
. . . . .
9
11 oooo
i/
r"
120000
L ..... .'J
plan showing the position of collieries in Ostrava-Karvinfi Coal District in the K[ovfik co-ordinates km) (Collieries: 1, Doubrava; 2, Lazy; 3, Dukla; 4, CSA; 5, Darkov; 6, Franti~ek; 7-9, kv6ten; 8, CSM).
F i g . 1. A
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 329-335.
(in
330
Z. KALAB
mechanism of the failure process and, significantly, the relation between seismic events and current geological and mining conditions. Subsequently, this leads to the possibility of specifying the probability of rockburst occurrences. The aim of this paper is to observe the connection of seismic activity with important fault zones in the area most affected by induced seismicity, i.e. the Karvinfi part of the OKR.
Tectonic and geomechanical situation The Czech part of the Upper Silesian basin is formed by Carboniferous sediments (a detailed description of the geological structure is given in Dopita & Kumpera 1993). The Karvinfi part of the OKR has the character of a basin with principal fault systems trending both in N-S and E-W directions. It is possible to model the region as a partial block structure consisting of beds of medium-thick psephito-psammitic layers that have a typically subhorizontal orientation. The thickness of individual beds ranges from metres up to 10-20m. Laterally, the dimensions of partial geological blocks vary from hundreds to several thousands of metres. The boundaries of the blocks are usually formed by fracture dislocations, largely filled with plastified materials. Intrablock tectonics is not very marked, but it exists and can divide the layers within the block into smaller partial blocks. Areas of varying intensity of tectonic dislocation differ in mininginduced seismic activity (Rakowski 1989). A geological analysis of important tectonic structures in the Karvin/l part of the OKR has been made. From a geomechanical point of view, the analysis has given the following results: 9 the thickness of faults (fault zones) ranges from several metres to some tens of metres 9 throws vary from metres to several tens of metres 9 the majority of faults (especially those with greater thickness), contain a filling consisting of plastified cataclastic materials 9 faults are commonly, although not always, wet The detailed evaluation of both tectonic structure and palaeostress conditions has proved the existence of dominant structuro-dynamic conditions favouring rockbursts. Critical sites were identified in the massif (e.g. the most important fault zones, hanging corner structures, sections with a small frequency of faults and others) where rockbursts could occur given the existence of disadvantageous, especially mechanical conditions (Kumpera et al. 1991). From a
geomechanical point of view, an analysis of the intensity of tectonic dislocations has resulted in the following relationships (Rakowski 1989): 9 rockburst areas are situated usually in places little affected by intrablock fault tectonics 9 in areas with sudden changes in the intensity of tectonic dislocations, a greater intensity of rockburst events can be expected 9 in areas with a smaller intensity of intrablock tectonic dislocation, the existence of tectonic discontinuous concentrators of stress and residual tension cannot be excluded
Seismicity and mining situation An analysis of mining induced seismic events, in relation to mining activities, has proved the existence of two important groups. The first is closely connected in space and time with the advance of mine openings, whereas, in the second group a more or less random relationship exists between seismic events and mining activities. These seismic events are induced at greater distances and are likely to be the effects of several workings. Moreover, shifts in time between mining activities and the initiation of events can also occur (Gibowicz & Kijko 1994; Rudajev 1989). An example where the development of seismic activity in a given area is influenced by the driving of mine workings is presented by Kone6n~, (1994). An empirical relationship between the process of massif failure in a limited area and the probability of the origin of anomalous rockbursts can be established. This is based on the relation between the number of mining-induced seismic events, the amounts of emitted seismic energy and the intensity of coal fracturing. The relationship is not simple because, in addition to the effects of mining operations, it is necessary to take into account 'natural' factors. The behaviour of all the factors given below must be studied to understand the development of seismic activity (Kone~n2~ 1989): 9 the primary stress field and geological structure (factors unaffected by human activity) 9 the geomechanical structure of the massif (given by natural factors, but partly influenced by active interference) 9 the secondary stress field and changes in it (impacts of mining activity) From the above, it follows that for mining purposes, it is desirable to study the development of induced seismicity under the geological
AN ANALYSIS OF SEISMICITY and geomechanical circumstances of the primary stress field by monitoring the stress changes that occur as a result of mining activities.
Seismological monitoring Induced seismicity results from anthropogenic activities causing changes in the rock massif that lead to its failure. Failure is accompanied by the generation of seismic waves that can be monitored by seismic stations. Events due to blasting operations have a special signifcance. If the intensity of an event corresponds to the size of the charge, the seismicity simply recording the blasting operation. However, if an event with an
331
energy greater than that corresponding to the charge is recorded, the blasting operation must have initiated failure of the massif in an area where stress conditions were close to critical. To monitor the seismic activity in the OKR, three levels of seismic networks are used. Local seismic networks of individual collieries represent a basic level of monitoring. The second level is ensured by a regional network named Seismic Polygon of the Ostrava-Karvinfi Collieries. The station Ostrava-Kr~isn6 Pole, which is a part of both the state and world seismological networks, is also situated in the Ostrava region. The development of seismological monitoring in the O K R as well as its utilization in the fight against rockbursts is described by Kalfib et al. (1994).
4.0 41 4.2 4.,344 45 46 47 48 49 50 51 52 53 54. 55 56 57 58 59 60 61 62 63 64. 65 66 67 68 69 70 71 72 7,3 74 75 E F" G H I d K L M N 0 P Q
Nzmi l, ml llmmmm an lan lml immBnui s liOi
R S T U V W X Y Z M BB CC DD lEE FF C-~ HH II JJ KK LL MM NN
E F G H I J K L M N 0 P Q R S T U V W X Y Z ,AA BB CC DD EE FF GG HH II JJ KK It MM NN
4.0 4.1 4.2 43 4.4 4.5 4.6 4.7 4.8 49 50 51 52 53 54, 55 56 57 58 59 60 61 62 63 64, 65 66 67 68 69 70 71 72 73 74. 75
Fig. 2. Intensity of exploitation Ik in the central part of the Ostrava-Karvin~t Coal District in 1990 (in square 250 • 250 m network; lines in network are the most important fault zones).
332
Z. KALAB
Modern methods of interpreting digital seismic data also test for tectonic effects. Seismic events recorded only in the focal area are represented by data from the experimental seismic network of a modular system in the Lazy Colliery with the possibility of connecting with up to eight surface or underground stations. The network monitors a seismically active part of the mine, covering an area of about 3 km 2 (Knejzlik et al. 1992). The layout of the stations of the Seismic Polygon of the OKD corresponds with its position as a regional network. Seven surface stations surround the Ostrava-Karvinfi District. The remaining three stations are located in underground workings. The network covers an area of 200km 2, using Lennartz Electronic GmbH equipment. The Fren~tfit Seismic Polygon is an autonomous unit situated about 30 km southwest of the Karvinfi part in an area of the Fren~t~it colliery under projection. It contains five surface stations that are telemetrically connected with the central recording station (Knejzlik & Zamazal 1992).
Data sets To evaluate the seismicity of significant tectonic zones in the Karvinfi area, four sets of mining induced seismic events have been identified: 9 9 9 9
all recorded mining induced seismic events intense rockbursts mining shocks intense seismic events induced by blasting operations
The sets cover seismic events that occurred from 1989 to 1991. Altogether, the localizations of about 14000 mining induced seismic events with an energy E > 100J (according to the energy scale in the OKD) were used in the first set, but due to limited data, the focal depths were not defined. The second set contained rockbursts with an energy E > 105 J. There were 94 events within the three year period. The set of mining shocks (weak surface shocks) comprises 183 records. This set is spatially inhomogeneous because it is based upon weak seismic events that can be observed on the surface and reported by the public to the processing centre. The last set consists of seismic events induced by blasting operations when 'shooting a charge'. The energy of these seismic events was higher than that stated because of the correlation between the charge weight and the energy of the induced
event. The correlation ratio was determined on the basis of a 100kg charge for all blasting operations. The bulletins 'Seismologickfi aktivita OKR' (Seismological Activity in the OKR) processed at the seismological centre in the CSA Colliery served as the basic information. A tectonic schematic map of the Karvinfi part of the OKR was used to create a visual analysis of the seismicity associated with tectonic zones. The map was processed in the DPB Paskov company (Mine Exploration and Safety) by plotting the distribution of tectonic elements at a depth of 500m. This depth corresponds roughly to the supposed z-coordinates of the foci of seismic events. With regard to the fact that the tectonic structure of the OKR is very complex, only the most important fault zones were taken into account. Nevertheless, it is not possible to determine unambiguously, whether a given seismic event originated or did not originate on an existing tectonic plane. To obtain unambiguous results, it would be necessary to establish data sets, in which the focal depth is also known to test whether the dislocation plane contains the focus. This will soon be possible using data recorded by seismic networks which provide three-component digital records of the wave patterns of seismic events. For the purpose of assessing the extent of mine activities, synthetic maps of mining intensity for the observed and previous periods were produced. The intensity of exploitation was evaluated by means of the Ik index (Kone6n~, 1989), which represents the thickness of the seam that could be extracted provided that exploitation is realized across the whole area under evaluation (in accordance with the mine network, a square 250 • 250 m network is used, see Fig. 2). Worked-out areas are determined from the mine maps within individual time intervals.
Discussion of results and conclusions As a consequence of potential inaccuracies, (e.g. in the localization of faults, the determination of their dip and the focal depth of the seismic event) a non-quantitative method was used to assess the seismic activity of significant fault zones in the Karvinfi part of the OKR. A visual comparison of the horizontal position of the focus with the fault at a depth of 500 m was used as a basis. In contrast to research in areas without mining-induced seismicity (Prochfizkovfi 1985) in this area, it is necessary to consider induced stresses that are variable in time. These arise as a result of mining activities.
AN ANALYSIS OF SEISMICITY Data from individual sets were, with half-year intervals, compared with the tectonic sketch plan and the intensity of mining in the area under study (e.g. see Figs 3 & 4). The data can be interpreted as follows: 9 No seismic event of natural origin (e.g. tectonic earthquake) has been identified in the data under analysis. No mining-induced seismic event has been recorded which was located within an area of a significant fault zone, where no mining activity had occurred. 9 The induction of seismic events is wholly dependent upon the space-time distribution of mine workings, especially of active faces.
333
This is valid also for the set of intense rockbursts that are not connected directly with mining activities. If an important fault zone occurs in the vicinity of the face, seismic events do not originate preferentially on it or in its surroundings. This result contradicts the results of Spi~fik & Zimovfi (1988) which suggested an increased seismicity along some fault zones. However, the authors warned that actual mining activities and the distribution of worked-out spaces were not considered. Significant concentrations of seismic events are observed in areas, where mining operations took place under extremely complicated
40 41 42 43 44 45 46 47 48 4.9 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74- 75
E F G H I J K L M N 0 P Q
P O R S T U V W X Y Z AA BB CC DD EE FF C,(; HH II Jd KK LL Mkl
R S T U V W X Y Z M BB CC DD EE FF GG HH II JJ KK LL MM NN
NN 4.0 41 42 4.3 44. 4.5 4.6 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
Fig. 3. Number of mining-induced seismic events (data set No. 1) in central part of the Ostrava-Karvinfi Coal District in 1990 (in square 250 • 250 m network).
334
Z. KALAB 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
"
/I
oF'E:
-~/\~.
H
/
d
~ - , . . -.
\ ~--'"i ' , -"\.
'I
\ . ~ - - < W
~ J
:
~ /
R s
/
/
\ ~
v
/
\
/
/
/ ...
3 6
/
Xts
/11
DD
\
EE
2
N
/
i/
oo
,
:'-~
~
u.-~
,-...__
/ /
I
I
311 "
R ~s
"_'>Lq- '
/ u \v
k r ~.q:
:/7~-.:~._
:~:
i w ~
z BB
3/
r I :i :\ :
~ .
.
.
.
~
. "
(~
1
I..~i..,.:
.
.,:-"'-,
:'/ .
\
"
~
,
/
DD
I
EE
Ii
'
'
~ i'~
"
~ ..
:
NN
t
/:\~\
I I ii i// .
"------f
P o
i~
Y,,
\
l/
\
I\M i
k\
ff ' I~,
.
~J(
/
//
KK
)!
i,:. ) /
"7/... ~_....--"
ee
H
L3:-:.;L ) ./~Zc:!/ ?i~'Y"~7"--% " ].'I
~:....
I
\~
/i
I
/
z-
r t;F
,
::,\
~
-
"X
\
.
"k
) '1 7 2 2 1 k Ii
i\ $/
/
u
" " ~ v ]
//
--,),--~,II
/
"'\
.
//
/
p Q
I"-
,
"-.-~
"
o
. '-'/Yk~
- - ":%:~
\_~.
M
w
-"~'-,.[]/i
-"\
..
(---'
oo li
KK
I J
u. ;
[ [ ] " NN 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 -'-" --Z.~ ..._
Fig. 4. Number of rockbursts (data set No. 2) in central part of the Ostrava-Karvinfi Coal District in 1990 (in square 250 x 250 m network).
geological, geomechanical and technological conditions. These concentrations are connected with the existence of important stress concentrators, on which seismic events are induced. 9 The connection between mining shocks and tectonics has been studied previously (Mtiller 1989; Veseki 1993). The results are not unambiguous, there is no demonstrable significance in the concentration of foci of these shock. It is possible to produce similar results as for previous studies. 9 Seismic events induced by blasting operations in workings are usually localized in the rocks overlying the working, in which the
blasting operation was undertaken (release of stress in the roof). In only a very small number of seismic events has a shift of foci towards fault zones been demonstrated. It is often not possible to distinguish a seismic manifestation of the blasting operation from an induced seismic event. Analysis of the data recorded within the threeyear period from 1989 to 1991 by the stations of mine network showed: 9 No incidence of the generation of seismic events on important fault zones or in their close vicinities has been proved in areas where no mining operations have occurred.
AN ANALYSIS OF SEISMICITY
9 Significant sources of induced seismicity are generated in zones of stress concentration where the spatial distributions of both geological and anthropogenic structures is critical. 9 The influence of the characteristics of the massif on the generation of mining-induced seismic events, can be determined by interpreting the digitally recorded data, with particular emphasis on assessing the depth of the focus and parameters of the plane of dislocation, or other physical parameters of the focus (see e.g. Swanson 1992; Tepper et al. 1992). The author acknowledges financial support from the Grant Agency of Czech Republic (reg. No. 105/93/ 2904 and 105/95/0474) and from the Czech-American Scientific and Technical Program (reg. No. 930 65).
References DOPITA, M. & KUMPERA, O. 1993. Geology of the Ostrava-Karvin~ coalfield, Upper Silesian Basin, Czech Republic, and its influence on mining. International Journal of Coal Geology, 23, 291-321. GIBOWlCZ, S. J. & KIJKO, A. 1994. An Introduction to Mining Seismology. Academic, San Diego. KAL,~B, Z., KNEJZL]K, J. & M~LLER, K. 1994. Seismological monitoring in Ostrava area. Exploration Geophysics, Remote Sensing and Environment, 1, 26-33. KNEJZLiK, J., GRUNTORAD,B. & ZAMAZAL,R. 1992. Experimental local seismic network in the A. Z~tpotock~, Mine of the Ostrava-Karvin~ Coal Field. Acta Montana, 84, 97-104. -& ZAMAZAL, R. 1992. Local seismic network in southern part of the Ostrava-Karvin~ Coalfield. Acta Montana, 88, 211-220. KONE~N~, P. 1989. Mining-induced seismicity (rock bursts) in the Ostrava-Karvin6 Coal Basin, Czechoslovakia. Gerlands Beitr. Geophysik, Leipzig, 986, 525-547.
335
1994. Mining induced seismicity in the Czech part of Upper Silesian Coal Basin depending on mining conditions. In: RAKOWS~I, Z. (ed.) Geomechanics 93 Proceedings. Balkema, Rotterdam, 63-68. KUMPERA, O., GRYGAR, R., KALENDOV,~,J., ADAMUSOV~,,M. & VONDRAKOVA,J. 1991. The evaluation method of structure and tectonic setting and palaeostress conditions in relation to rockbursts prognosis. MS Report, Technical University, Ostrava (in Czech). MI2LLER, K. 1989. Location of mining shock in Karvinh part of OKB. In: Seismology in engineering and mining practice. Proceedings, Technical University, Ostrava, 33-36 (in Czech). PROCHAZKOVA, D. 1985. Space-and-time pattern of seismicity. Proceedings of symposium, Geophysical Institute of CAS, Prague, 46-53. RAKOWSKI, Z. 1989. The conception of a physical model of rockburst prone areas in OstravaKarvin~ Coal Basin. Proceedings of symposium, ECE of the United Nations, Ostrava, Czechoslovakia, A21. RUDAJEV, V. 1989. Major causes of rockbursts and the role of seismology in their research. Proceedings of symposium, ECE of the United Nations, Ostrava, Czechoslovakia, A23. SWANSON, P. L. 1992. Mining-induced seismicity in faulted geologic structures: An analysis of seismicity-induced slip potential. PAGEOPH, 139, No. 3/4, 657-676. SPIC,~K, A. & ZIMOV,/k, R. 1988. Seismic activity in Karvin6 part of OKB and its reasons. MS Report, Geophysical Institute of CAS, Prague (in Czech). TEPER, L., IDZIAK,A., SAGAN,G. & ZUBEREK,W. M. 1992. New approach to the studies of the relations between tectonics and mining tremors occurrence on example of the Upper Silesian Coal Basin (Poland). Acta Montana, Ser. A, 88, 161-178. VESEL,~, V. 1993. An elementary analysis of mining shock. In: KALAB, Z. (ed.) Seismology and the Environment. Proceedings, Institute of Geonics, CAS, Ostrava, 146-152 (in Czech).
Comparison of structures derived from mine workings and those interpreted in seismic profiles: an example from the Ka~ice deposit, Kladno Mine, Bohemia STANISLAV OPLUSTIL, JIl~I PESEK & JIl~i SKOPF,C Faculty of Science, Charles University, Albertov 6, 128 43 Prague 2, Czech Republic
Abstract: Five seismic profiles across the Ka6ice coal deposit were reinterpreted and compared with observations in mine galleries. The comparison shows that approximately 80% of normal faults with displacement exceeding 5 m detected on the seismic profiles really exist. In contrast, in only two cases have mine workings shown faults (vertical displacement 10-15m) that have not been identified by seismic measurements. Discrepancies may be mostly explained by: (i) misinterpretation of the fault with slope of the presedimentary palaeorelief accompanied by differential compaction; (ii) virgation of faults and misinterpretation of a fault zone composed of several small faults individually below the detection limit but whose aggregate displacement is detected, giving the appearance of a single fault; (iii) faults which die away toward the overburden indicating synsedimentary movements in the deposit. Reflection seismic has become a common and useful method of exploration of coal-bearing deposits in the central Bohemia, especially in the coalfields of the Kladno Basin (Kadle6ik et al. 1979, 1985, 1986). However, almost none of these coalfields has been mined until now. Exploitation of the Ka~ice deposit is the only exception which enables comparison between seismically derived tectonic interpretation and reality.
Moreover, at the end of 1970s and the begining of 1980s, reflection seismics was carried out in the central and northern parts of the deposit. Advanced exploitation allows us to compare observations in the galleries with the results of seismic interpretation.
Stratigraphy of the Kladno Basin and Kadice deposit
History of the deposit For more than 150 years, thick coal seams of the Radnice Member have been exploited in the Kladno coalfield located along the southern margin of the Kladno Basin. Later, in the mid1950s the Ka6ice deposit was discovered NW of Kladno coalfield, beyond an area of postsedimentary erosion of the coal seams (Fig. 1). Exploratory drilling of the deposit during the 1960s (Salava 1960; Richter 1964, 1966, 1969) allowed its opening in 1969 through a gallery from the Kladno Mine. Consequently, coal exploitation followed from 1975. The annual coal production from the deposit has varied between 400 and 450 x 103 tons during 1980s, however, in 1994 it was only 311 • 103 tons. From 1986 to 1992, refractory claystone was mined, but its exploitation was abandoned due to economic reasons. During the last 20 years of exploration and exploitation of the Ka6ice deposit a large amount of new data has been collected, the concentration of which is greater than for any other part of the Kladno Coalfield (results until 1980 are summarized in Spudil et al. 1980).
The Kladno Basin with the Kladno Coalfield including the Ka6ice deposit in central Bohemia is only a small part of a WSW-ESE elongated complex of Upper Carboniferous continental and partly coal-bearing sediments, which extends from western through central to eastern Bohemia with a length exceeding 250kin (Fig. 1). In western and central Bohemia, the Carboniferous sediments are divided into four lithostratigraphic formations. From the bottom these are: Kladno, T~,nec, Slan~, and Lin6 Fins based on alternation of red and grey sediments. Deposition began in mid-Westphalian and, including several hiatuses, lasted at least to the end Carboniferous. The coal reserves of the Kladno Basin are concentrated mainly within the Radnice Member at the base of the Carboniferous infill. It contains up to five mineable coal seams grouped into the Radnice (with Lower and Upper Radnice seams) and younger Lubnfi (with Lower, Middle and Upper Lubnfi seams) group of seams. Most of them have been mined in the study area. There are no other workable seams in overlying units. In the area of the Ka~ice deposit only the Kladno and T~nec Fins are fully present, whilst
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 337-347.
338
S. OPLUSTIL E T AL.
~>i'ONIcE
DEPRESSION
150km
A VE LVARY RIDGE
BILICHOV RIDGE
RIDGE
N
(LADNO
Y ~2
Y ls
,o
,,Skm
B
Fig. 1. Generalized map the Upper Carboniferous deposits in the Czech Republic (A) and location of the study area within the southern part of the Kladno Basin (B) in Central Bohemia (after Klener 1982, modified). 1, Kladno Basin; 2, other Carboniferous deposits in the Czech Republic; 3, coalfields; 4, outcrop of the Proterozoic basement; 5, present distribution of the Radnice Member. Numbers represent the names of different coalfields and deposits (1, Rynholec coalfield; 2, Ka6ice deposit; 3, Brand~sek d.; 4, T~ebusice d.; 5, Jan d.; 6, Minice d.; 7, Kokovice d.)
only the basal part of the Slan2~ Formation has been recognized. The preserved thickness of the Upper Carboniferous deposits varies from 450 m to 650 m depending on palaeotopography, erosion and tectonics. These deposits are underlain by Upper Proterozoic basement composed mainly of a monotonous complex of folded shales and uncommon volcanic rocks and cherts. Deep erosion of the basement created a significant palaeotopography with differences in elevation between the paleohighs and paleovalleys of up to 150 m in the study area and its close vicinity (Oplu~til-Vizdal 1995). The axes of the main elevations and depressions are in a good agreement with structural elements (foliation, fold axes, cherts and volcanic belts) of the basement and both follow a predominantly WSW-ENE direction which is the trend of the Kladno Coalfield itself. Workable coal seams are developed in palaeovalleys. The Ka6ice depression is the northern protrusion of larger Kladno depression. It is surrounded by two significant palaeohighs from which protrude minor ridges
with variable directions into the Ka6ice Deposit itself. The slope of the elevations commonly reaches 10-20 ~ locally even more. In the Ka6ice Depression the coal-bearing Radnice Member is dominated by siltstones and mudstones with four mineable coal seams (Fig, 2). Near the base, mudstones commonly interfinger with breccia derived from weathered basement surrounding the depression. In the upper part of the Radnice Member sandstone bodies also occur, the number and thickness of these increasing significantly to the south. The thickness of coals decreases upward; while the Upper Radnice and Lower Lubn~i coals commonly exceeds 3m (max. 7m) the remaining coals (Middle and Upper Lubnfi) rarely reach 2m. The thickness of the Radnice Member is only erosional and varies greatly from 0 to 185 m (average thickness around 90 m) within the study area and its close proximity. The overlying unit, the N~,~any Member, was deposited after a hiatus and reaches an average thickness of 350m with only slight variations.
STRUCTURES IN MINEWORKINGS AND SEISMIC PROFILES
Lrl
(...) ._I
L.,.. .
c
->,
O
.
Fli:r uloy M,
.
. .
. .
. .
. .
. .
. .
.
.
. . . . . . . . . . . . . . . . . . . . . . . .
. . .
. .
.
. .
. .
. .
- . . . . .
. .
. .
.
.
.
.
.
;.--3::...::.?-:: :.72: ~ j c-:-~-1---<-3"-T:--'T- 7 7::, -. -:-- :5:-~-,-,-
Moteslce M L,._ JetenicaM,
g~
k.
c
...
,
. . . . _.-.~.- _. -._.-._ .- _.-._.- _ - _ - _ - _ - .
--------'--UPPER LUBNA CDAL
-_-_-_-.-_-_-_-_-_ _ - ~CI
c-
q_.
E~
--
i,
-~
i
/
0..,
/
1.-.
s
U
,-
L
+,
".>.,
..Q
I.--
.t.a
(_.)
.
.
2-'-'2:-'_. . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
"-----"--. T" - - . - " ,
~
~
_
.
.
.
.
.
.
.
.
.
.
.
. 7 .-
.-
T
-
-
.~ . ~ . -
- -. - 7 . 7 - -. - . - 7 . , ."-~ ~ 2 ~ o ~ ; o ; ~
.
-2-_--_--_--7.-_-----_--_-_--_--Z~-1-..... ~
/ /' /
.I
i
(D (~
339
/
.
co
N~'any M.
--
c
~
e-
I
LOVER LUBN~ COAL
' / UPPER RADNICE COAL
'
o
n
+
CL
m
g
t"
i> o tn
(3
LOVEP I~ADNICE COAL
Rodnlce M.
v
,
,
,
.
......
N N
,
II
'
8
{t'
x
>'',
Fig. 2. Stratigraphy of the Kladno Basin and lithology of the Radnice M e m b e r in the study area. 1, mudstones; 2, (arkosic) sandstone; 3, (arkosic) conglomerate; 4, breccia; 5, refractory claystone; 6, tuff, tuffite; 7, coal; 8, basement.
The fluviatile sediments of the N)~any Member are arranged into cycles, the thickness of which varies between 1 and 10m. The lower parts of the cycles dominate with medium-to-coarsegrained sandstones occasionally with conglomerates. These cycles are grouped into six 40 to 60m thick mesocycles which exhibit finingupward trends with a thin coal seam in the highest cycle of each mesocycle (Spudil 1982). The following unit, the T2)nec Formation, shows a very similar lithology and sedimentary architecture to the previous unit, making it rather difficult to distinguish. It is composed of three mesocycles with predominantly red coloured sediments without coal seams. The T)nec Formation passes gradually into the Slan) Formation. The Slan~, Formation is preserved only in the northern part of the Ka6ice Deposit with an erosional thickness of 30-40 m. Apart from the
predominantly grey coloured mudstones it is very similar to the underlying T~Tnec Fro. and N~7~any Member. Within the study area, the Carboniferous sediments are covered by Upper Cretaceous deposits, the thickness of which varies in relationship to the pre-Cretaceous relief and intensity of post-Cretaceous denudation. Usually, the Cretaceous deposits are less than a few tens of meters thick. They are composed of continental fluviolacustrine sediments (conglomerates, sandstones and mudstones, occasionally with a thin seam of dirty coal) at the base grading upward into marine sandstones and siltstone. They are overlain by marl.
Coalfield The concentration of data in the area of the Ka~ice deposit is greater than that of the other
340
S. OPLUSTIL E T AL.
coalfields in the Kladno Basin. Since its discovery in the 1950s there have been more than 50 deep boreholes drilled from the surface to the basement and over 500 mining boreholes, penetrating usually only a part of coal-bearing succession of the Radnice Member. In addition, there are 47 kilometres of galleries, which have been the most useful for the construction of the tectonic map of the deposit. An independently created tectonic map has been derived from seismics. All the galleries and mining boreholes are located in the approximatly 100 m thick basal coal-bearing complex. The remaining 350-500m of overburden is known only from deep boreholes.
Reflection seismics The commonly used borehole spacing is insufficient to determine the fault tectonics of the Kladno basin. The CMP (common-midpoint) method of reflection seismics can be used to aid the fault analysis. The important advantage of the CMP method is its ability to detect separate horizonts even at a depth of several kilometres. The results of the CMP method are commonly presented either as time or depth sections, where continuous reflecting horizons are often clearly visible. The irregularities that exist in the course of these horizons can be interpreted as evidence of faults. The rocks filling a continental sedimentary basin may be developed as a cyclic sedimentatary sequence consisting of many sandy and clay layers. The lithological interfaces between neighbouring sedimentary layers create reflecting boundaries for seismic waves generated at the Earths surface. The parameter describing the reflectivity of a medium is called the reflection coefficient, which can be defined as the amplitude ratio of the incident and reflected waves. In practice the reflection coefficient is expressed as a simple function of the densities and longitudinal wave velocities in the overlying and underlying media. In the case of a cyclic sequence, the time differences between separate reflections are so small that many of these reflections arrive within the time interval corresponding to the wavelet of an individual reflection. Under such conditions, instead of separated true reflections, rather random interference patterns of numerous reflections coming from the individual boundaries may be expected. The amplitudes of these reflections are usually small, as are those from the summary reflections. The total amplitude of the reflections depends not only on the reflection coefficient, but also on the number of reflections
affecting the signal, on the thickness and lithological stability of separate layers, etc. Since these factors are more or less variable, the extent of such reflections is usually limited and the whole wave field can be characterized as rather irregular. The amplitude of a reflected wave depends on the difference in physical properties on both sides of a given interface. In coal-bearing sedimentary basins, coal seams appear as layers of an anomalous physical behaviour because of their low values of density and seismic velocity, which affects the reflection coefficient positively. Therefore, a group of beds containing coal seams is usually characterized by strong reflections that make it possible to determine fault positions where reflection horizons have been displaced. This shift can vary from several hundred seconds to a few milliseconds and may not be observed at all if the throw is too small. Seismic measurements in the Kladno Basin were carried out in 1979 and 1983 by Geofyzika Brno. The Vibroseis measurements with 12-fold coverage in the neighbourhood of Ka~ice village were processed by using standard procedures including wave migration. Strong reflections in the coal-bearing basal Radnice Member have allowed the detection of faults which can be compared with the results of later geological mapping. Within the overlying sediments, the fault structures affect the wave field less evidently, causing local and disconnected shifts of rather weak reflections. As an example of traceable faults causing time shifts in the coal-bearing Radnice Member, the part of seismic line 69/83 is presented in Fig. 3a. A number of normal faults with vertical throws of several tens of metres is obvious in the middle part of the seismic depth section. In Fig. 3b, a part of seismic line 70/83 is shown, where the sedimentary beds are only a little faulted by few minor normal faults. The time shifts are negligible, but correspond to faults located in the coal mine adits. On both figures, the different wave field can be observed for overlying Carboniferous sediments and underlying shales of Proterozoic age, while within the Carboniferous beds the part containing the coal seams is quite different from the overlying cyclic sediments without coal layers.
Structure derived from mine The deposit is affected by post-sedimentary normal faults with a general N W - S E strike. Faults with other strike directions are rare.
Depth (m)
69/83
leg
ee@
1oo
2ee
3ee
4ee
.~88
Depth (m)
70183
lee
eee
lee
2oe
3ee
4ee
500
Fig. 3. Part of the depth seismic profiles with interpreted faults. 3A, depth seismic profile 69/83; 3B, depth seismic profile 70/83. For location see Fig. 6.
S. OPLUSTIL ET AL.
342
~D
0 0 ~.~
\ \ "x
'\ ,,..I"
,,,.,
/
/
~.~ ~m
.~
k
9.. f.
e~
o 7
~
~"
E~-
r-
\
I"~ 84
~.1
~N ~
.~ .~.
~ ~ .~.
o~
e6P!~l Aoqo!lJe
\
.~ ~ z ~ \
\~
\\
/i--.-
~.~ ,~ "~ ~
STRUCTURES IN MINEWORKINGS AND SEISMIC PROFILES Moreover, they have only small vertical displacement with a maximum of a few metres. The majority of these faults dip SW. Reverse faults with throws exceeding several tens of centimetres have not been observed in the galleries. In the early 1960s, palaeo-ridges with a NW-SE direction were misinterpreted in seismic profiles as normal faults, because until that time only those ridges running NE-SW had been known. The throw of the normal faults varies greatly from several centimentres to over a hundred metres. Approximately 80% of all documented faults have throws less than 2m and only 10% exceed 5 m; thus they are around or above the detection limit of reflection seismics. Larger faults with more than 40m downthrow occur at a spacing of about 1.5km in the Kladno Coalfield (Fig. 4). They dip at approximately 75 ~ (Spudil 1982) and extend to the Proterozoic basement but are truncated by the pre-Cretaceous surface. Smaller faults (10-20m) are traceable for a distance of several hundred metres up to one kilometre and their dip is usually 55-75 ~. They also continue into basement but it is uncertain whether they reach the top of Carboniferous sequence. They are typically sinuous. The faults often virgate. They commonly occur as systems of antithetic faults, creating grabens several tens up to one hundred metres wide (near the base of the Carboniferous) running across the coalfield (Fig. 4). Small faults (throw less than 10 m) usually do not disturb the whole section of the Radnice Member and diminish as they approach basement. The most distinct reflection corresponds with the boundary between subhorizontally layered undeformed Upper Carboniferous sediments of various lithologies and the folded lithologically monotonous complex composed mainly of Proterozoic shales. The general dip of the basement surface is 5-8 ~ to NNE. Therefore the shallowest occurrence of this boundary is in SW part of the deposit (borehole K6 11, -38.8m); whilst the average altitude of the surface reaches 430 m. To the NE it falls to a depth below -280 m (borehole So 8, -287.7m surface 400m). In detail (Fig. 5), however, the general dip is superimposed on the slopes angles of the pre-Carboniferous ridges. Therefore the final angle may exceed 15~. The study area is affected by three significant normal fault systems. The largest one runs along the eastern margin of the deposit where it continues from the Kladno Coalfield to the south. The vertical displacement reaches 100 m in the SE corner of the Ka6ice Deposit. Further north the fault gradually diminishes and virgates
343
into several smaller faults, the downthrow of which varies between 10 and 20m. They are traceable over a distance of 1 to 2 km in seismic profiles and exceptionally in galleries. Another significant fault zone composed of two major antithetic faults limits the western part of the Ka~ice Deposit. In the SW it creates a 70 m wide graben at the level of the Upper Radnice Seam (about 50 m above the Carboniferous basement). It has a throw of 6-10m. Northward its vertical displacement increases up to 80 m (eastern fault, cross cut 243) in NW edge of the deposit. A vertical displacement of about 60 m on the western fault is estimated from the discrepancy in basement altitude between two boreholes located in the upthrown and downthrown blocks. A similar graben is known from the southwestern margin of the deposit, where it continues from the Kladno Coalfield. In the Ka6ice Deposit, however, only its eastern normal fault was proved in the gallery.
Comparison of structures observed in the mine and those interpreted in seismic profiles Five reinterpreted seismic profiles are compared with the structural map derived from mine. These profiles run mainly from WSW to ENE being more or less perpendicular to prevailing direction of the faults. The location of all seismic fault indications in Fig. 6 if not stated, corresponds with the base of the Carboniferous. Seismic profile 1C/78 runs from SW to NE through the study area (Fig. 6). Two normal faults have been inferred in the area of mining activity (vertical movement 20 and 5 m) and, just behind the eastern end of the galleries, a 200 m wide tectonic zone has been located. Within this zone two principal normal faults have been recognized. A larger one (80 m) on its western margin and a smaller one in its eastern margin. Both dip to the SW. These indications are in good agreement both in size and dip with a significant normal fault proved in a cross-cut at the SE margin of the deposit as a non-branched fault. Seismic measurement supports the idea of their virgation and gradual diminution to the NNW. The remaining two indications correspond only partly to observations in the mine. While the smaller one (5m) has been known from several galleries north of the profile, the larger one (c. 20m) dipping NE probably has no equivalent in the mine. However, the distance between the detection site and the closest gallery
344
S. O P L U S T I L E T AL.
-12 80'
!
-60
/
\
\ \
Le3\
N
\
\\\\\
\
\
I/I/
~oX~ LADN0 O
\
~" #~l~ JILt el~~~~~~ N
o
~oo
~u3~ '
"7
Fig. 5. Tectonic sketch of the study area and contours of the surface of the Late Proterozoic basement; 1, contours of the surface of the Proterozoic; 2, significant normal faults observed in mine; 3, borehole.
S T R U C T U R E S IN M I N E W O R K I N G S
A N D SEISMIC PROFILES
345
Le| 6 t I t t I
I
|
\
t,
,
z~
z~
/8:
\
Le 3
,
70183
\ \
83
\ \ \ \
1183
\
\\~0
'
t
\
|
\
,
N
%3
[]
\
:/78
\ \ \
/
2,2
L/;g V~ n 1 8 3
9 ~
soz t |
L
i
0
l
\
~,5
~m
a
,.J
>.3
\
32 \
\
K~12
\ \
\ \ \ \
\ .
.
.
.
Le 3 4
,
l
|
Fig. 6. Comparison of structures derived from mine with those interpreted in seismic profiles. 1, normal faults observed in gallery and its expected continuation; 2, indication of normal faults in depth seismic profiles the existence of which have been proved (simple frame) and which have not been proved (double frame) in mine; 3, depth seismic profiles, thick part corresponds with location of profiles depicted in Fig. 3; 4, borehole.
346
S. OPLUSTIL ET AL.
in the direction of the fault is 500 m and it has been observed that a 5m fault may terminate over a distance of 100m. Nevertheless, larger faults are usually more persistent. Detection of the fault in seismic profile through the whole Carboniferous sequence excludes misinterpretation due to the vicinity of a pre-sedimentary ridge or sudden facies changes within the Radnice Member. Seismic profile 69/83 crosses the middle part of the deposit in a WSW-ENE direction. The two westernmost indications lie outside the area proved by the galleries. They are interpreted as antithetic normal faults with vertical displacement of about 20-30 m. They match well with the hypothetical continuation of the graben proved in the mine l k m to SE. Its eastern limit, which is well known from several galleries, has been detected 240 m further east. 250 m further east on the profile a large fault (up to 100 m at the base of the Radnice Member, c. 30-50 m about 90 m higher) dipping eastward has been detected. This normal fault has no equivalent proved in the surrounding galleries (distance 100m). It is believed, that this discrepancy is induced by close proximity of a pre-sedimentary ridge protruding into the deposit from the west. The inclination of its slope probably exceeds 20 ~ (locally 30 ~ or more) as it results from palaeo-relief reconstruction. Moreover, misinterpretation could be affected also by rapid facies and thickness changes within the Radnice Member near the ridge and also by compaction. It is supported by a decreasing value of the throw to the top of the Radnice Member. There are seven other indications with sizes varying from 5 to about 15(20)m. Only two of them (throw 5-10m) remain unproven in the mine. They probably correspond to a narrow fault zone composed of small normal faults under the detection limit, the aggregate throw of which could affect the seismic reflections. Seismic profile 70A/83 runs SW-NE c. 300500m north of profile 69/83. There are seven indications of normal faults with vertical displacements between 5 and 15 m. Only two of them probably do not correspond to the fault system proved in the mine. The westernmost indication dips westward with an estimated throw of about 10(15)m. Its parameters are comparable with the normal faults observed in the mine, but, c. 80-100 m eastward at the level of the Lower Lubnfi Coal, i.e. 50m above the basement. It is believed that this discrepancy could be due to different stratigraphic levels. Otherwise the faults proved in the mine lack any indication on the seismic profile.
The second indication without a proved equivalent is an east-dipping normal fault with an estimated displacement of c. 5-10m (Fig. 6). It is situated in a zone with an increased number of small normal faults whose aggregate displacement may resemble a single fault in seismic profile. Seismic profile 70/83 crosses the northern part of the deposit from WSW to ENE. Five indications of normal faults have been recognized within the seismic profile. Their throw varies from 5 to 25 m. Four of them are situated in the mining area and have been observed in galleries. The fifth indication is located behind the eastern margin of the galleries in the proximity of borehole Le 3. It is interpreted as an east-dipping normal fault with estimated throw of 25m. Probably on the same fault, gallery 1008 terminated at the NE margin of the deposit. The second indication from the east has been detected only at the base of the Carboniferous (throw c. 10 m); it seems to be absent higher but is indicated again around the boundary of the Kladno and the T~nec Fms approximately 400 m higher. Seismic profile 71/83 runs NNW-SSE along the eastern margin of the deposit, just behind the eastern end of the galleries. It is nearly parallel with the main faults of the study area. Therefore all of the indications belong to the main fault (fault zone) of the Kladno Mine. They dip either to the W or to the E due to undulations of a fault with a low angle of dip, 30-40 ~. Their throw decreases gradually northward from 20 to 10m.
Conclusion Comparison of structures interpreted in seismic profiles with those observed in mines show good agreement. Approximatly 75% of seismic indications correspond with the observations in mine galleries in both dip and throw. The seismic data have indicated nearly all of the observed normal faults above the detection limit, which in the central and western Bohemian Carboniferous is between 5 and 10m. The number of normal fault indications in seismic l:rofiles slightly exceeds the number of observed faults. Most of them are around the detection limit. However, in one case a large fault with a throw of several tens of metres has been interpreted with no equivalent in the mine. The possible explanations for most of the discrepancies between seismic data and observations in mine are as follows:
STRUCTURES IN MINEWORKINGS AND SEISMIC PROFILES 9 misinterpretation of the slope of a presedimentary ridge, which could exceed 150m elevation accompanied by sudden facies change. The influence of presedimentary palaeotopography and different compaction could persist up to the level of the T~nec Fm. (Spudil et aL 1980) 9 several smaller faults under the detection limit whose aggregate throw is interpreted as a single fault. The apparent discontinuous character of some faults in seismic sections can be induced by the coincidence of the throw of the normal faults with the thickness of the cycles. The resulting reflections appear to be uninterrupted. It could be a common phenomenon in the N ~ a n y Member and the T2?nec Formation where the average cycle thickness varies between 7 and 10m. In one case a normal fault dying out towards overlying beds indicates an occurrence of synsedimentary movements in the deposit. Despite the above mentioned discrepancies, reflection seismics has been proved to be a useful method for exploration of new coalfields in the Upper Carboniferous coal basins of western and central Bohemia. It indicates that a proper density of seismic profiles allows the construction of a reliable structural plans where even faults with small vertical displacements can be indicated by the geophysical method.
347
References KADLE~iK, J., SKAROV,~, M. & JIHLAVEC, F. 1979. Geofyzik6lnY - geologick6 zhodnoceni reflexnYseismickfwh pracl S R B technologii VIBROSEIS na ~kolu Peruc-Slapanice. Written final report. --
Archives Geofond, Brno. et aL 1985. Seismickf: prdzkum na lo~isku Slanj~ v r. 1983-1985. Written final report. Archives Geofond, Brno. et al. 1986. Z6vJre(nd zprdva o reflexnJseismick6m prdzkumu SRB v oblasti PerucKokovice v r. 1983. Written final report. Archives
Geofond, Brno. OPLUSTIL, S. & VIZDAL, P. 1995. Pre-sedimentary palaeo-relief and compaction: controls on peat deposition and clastic sedimentation in the Radnice Member, Kladno Basin, Bohemia. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds), European Coal Geology. Geological Society, London, Special Publication, 82, 267-283. RICHTER, V. 1964. Kadice. Zdvdrednd zprdva pJredb~n6 etapy prdzkumu. Written final report. Archives Geofond, Praha. - - 1 9 6 6 . Dopln~k zdv#rednd zprdvy Kadice. Written final report. Archives Geofond, Praha. 1969. Dopln~k zdvgredn~ zprdvy Ka~ice sever. Written final report. Archives Geofond, Praha. SALAVA,J. 1960. Z6vYre?n6 zprdva Ka?ice. Vyhleddvaci etapa. Written final report. Archives Geofond, Praha. SPUDIL, J. 1982. Strukturn6 geologickfi charakteristika lo2iska Ka6ice. In: HAVLENA,,V.,~ PESEK,J. (eds) Sbornik IV, uhelng geologickd konference pFidovYdeck~ fakulty, Praha, 133-142. et al. 1980. Z6v~re?n6 zpr6va ~kolu Ka{ice.
Written final report. Archives Geofond, Praha.
Improvements in direct coal liquefaction using beneficiated coal fractions J. B A R R A Z A ,
M. C L O K E
& A. B E L G H A Z I
Coal Technology Research Group, Chemical Engineering Department, University o f Nottingham, Nottingham NG7 2RD, UK.
Abstract: Beneficiated coal fractions from Point of Ayr coal (North Wales) were liquefied in order to determine their effect on conversion, product and metal distribution in coal extract solutions. The coal fractions were obtained in a dense medium cyclone separation unit, using aqueous solutions of Ca(NO3)2, as medium, of relative density 1.26. The original coal and the coal fractions were liquefied in an autoclave with hydrogenated anthracene oil (HAO) as solvent. Liquefaction results show an improvement in conversion for the overflow fractions over the feed coal, together with a shift in the net product distribution toward higher oils content and lower asphaltenes and preasphaltenes material in the liquid products. A marked decrease in the proportion of A1, Mg, Mn and Si was found in the extracts using overflow coal. However, Ti and Ca, which are deactivating elements of the hydrocracking catalyst used to upgrade the coal liquids, increased their proportion.
During the last decade, work has been carried out in order to separate and concentrate macerals for use in liquefaction processes (Dyrack & Horwitz 1982; Cronauer & Swanson 1991). Investigators seem to agree that the coal characteristics, particularly petrographic and mineral compositions are important parameters in the coal liquefaction process, affecting overall conversions, product and metal distribution in the coal liquids. It has been established (King et al. 1984; Steller 1987) under a wide range of liquefaction conditions that the more reactive maceral is liptinite followed, in decreasing order, by vitrinite and inertinite. Also, some studies (Keogh & Poe 1987; Oner et al. 1994) have shown that the presence of mineral matter in coal has a catalytic effect towards oils production in liquefaction carried out in the presence of hydrogen and at conditions of high severity. One of the liquefaction techniques which could process coals of a wide range of maceral and mineral compositions is the British Coal two-stage coal liquefaction. In the first stage, the coal is digested in a hydrogen-donating solvent in the absence of hydrogen, at a low pressure (20-30 barg), and the resulting mixture is filtered to produce a low-ash extract solution. In the second stage the extract solution is catalytically hydrocracked in the presence of hydrogen to upgrade the products. Catalytic deactivation, during the hydrocracking stage was found (Robatt & Finseth 1984; Kovach et al. 1978) due to the deposition of metals such as titanium, calcium and magnesium. Because of this, attempts have been made to reduce these undesired elements in the extract
before the hydrocracking process. Cloke (1986), reported that an increase in the digestion pressure produced a significant reduction in the extract ash levels. Also, the addition of toluene before the filtration process, precipitates heavy organics producing extracts of low concentration in mineral matter (Cloke et al. 1993). However, very little work has been reported in order to attempt to decrease the metal content in coal extracts using beneficiated coal fractions, which are simply clean coal fractions with high concentration of organic matter and low concentration of mineral matter. The purpose of this study was to ascertain the effect of liquefying beneficiated coal fractions, obtained by a dense medium cyclone unit, on conversion, product distribution and element content in the extract solutions.
Experimental Materials A bituminous coal from Point of Ayr (North Wales) was used in the study. Coal samples, runof-mine and uncrushed, were supplied by British Coal. The dense medium used for the separation was an aqueous solution of Calcium Nitrate Tetrahydrate, obtained from Berk Ltd, UK. Advantages of this medium include that: it can produce solutions of a wide range of densities; it has low toxicity; and it is moderately inexpensive (Rhodes et al. 1993). The solvent used for the liquefaction was a process-derived Hydrogenated Anthracene Oil (HAO), supplied by the
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 349-356.
350
J. BARRAZA ET AL. overflow and the under flow fractions. The product of the digestion was filtered to produce a filter cake and a coal extract solution. Details of the liquefaction procedure has been previously reported in the work of Cloke et al. (1987).
British Coal Liquefaction Project. Standards and controls used to determine the metal element concentration were made from B.D.H. 'Spectrosol' solutions.
Procedures
The fresh coal was crushed, wet screened, filtered and air-dried at room temperature to obtain coal samples of particle size -250 + 63 #m. Coal samples were processed in a dense medium cyclone unit, which is a closed circuit sumppump-cyclone. A diagram of the process is shown in Fig. 1. Water and Calcium Nitrate (powder) were added to the feed sump and the density determined and adjusted. A relative density of 1.26 was used in the experiment, which was obtained by increasing the medium concentration or diluting it with tap water. The coal was added to the media, producing a slurry, which was agitated with a stirrer driven by air. The slurry was pumped (2 barg) to the cyclone and samples of overflow and underflow were collected. The recovered samples were filtered and air-dried in the laboratory at room temperature ready to liquefy. Liquefaction runs were made in a 2 litre autoclave with HAO as solvent in a ratio HAO/Coal: 2/1, w/w. Approximately 350g of coal (as received) was fed to the autoclave. The liquefaction was carried out for 1 hour and a temperature of 420~ Digestion pressures of 10barg and 40barg were used for the original coal, while a pressure of 40 barg was used for the
Analysis
Analysis for moisture and ash were carried out using the Standard BS1016 methods, while maceral analysis was performed by the procedure described elsewhere (Cloke et al. 1994). The coal extract solution was analyzed for hexane, toluene and tetrahydrofuran (THF) insoluble, in order to determine the product distribution. In the present study, the product distribution is defined as: oils, (100-hexane insolubles); asphaltenes, (hexane insolubles- toluene insolubles); preasphaltenes, (toluene insolubles - THF insolubles) and heavy organics, (THF insolubles). The coal conversion (on a dry mineral matter free, dmmf, basis) is defined in terms of coal ash, filter cake ash and quinoline insoluble. The extract solution and filter cake were analyzed for ash content at a temperature of 815~ in platinum crucibles. Nine elements (A1, Ca, Fe, K Mg, Mn, Na, Si and Ti) concentrations were determined in the original coal, overflow, underflow, coal extract liquid and HAO, using a Perkin-Elmer model 2380 Atomic Absorption Spectrophotometer. The metal concentrations were determined on the basis of the work of Hamilton (1986).
Overflow
Air Coal+ medium
IJ
T---~ (~ d---~~ ] - - - Cycloneunit
Underflow
I
i~
/ Cleaj,
! Stirrer
'\,,, \,
SumpTank J
Gaugepressure
//
ReceptorMedium Tank
/'
~/// Feedp u ~
|
Recirculation pump
9
Fig. 1. Diagram of the dense medium cyclone process.
DIRECT COAL LIQUEFACTION
Results and discussion Dense medium cyclone separation Mass yields, ash, macerals and element analysis of the original coal and coal fractions obtained in the separation are shown in Table 1. Results show that the overflow fraction gives a lower vitrinite concentration than the original coal, however, it has the lowest ash content and the highest concentration of liptinite. By contrast, the underflow fraction shows the lowest vitrinite and the highest inertinite content. Clearly, the findings indicate that the cyclone separation had
Ash, maceral and major element analysis of coal samples
Table 1.
Coal samples as received Original
Overflow Underflow
Relative Density Yield mass (% w/w) Ash (%, db) 14.5
1.26 43.0 1.2
1.26 57.0 24.0
Maceral analysis, mmf (% v/v) Vitrinite Liptinite Inertinite
77.0 15.7 7.3
73.6 7.7 18.7
7.85 8.40 9.00 1.22 1.65 0.02 0.85 18.74 1.35
9.07 3.10 9.10 2.61 1.89 0.16 0.23 29.38 0.67
80.9 10.1 9.0
Element analysis in ash (% w/w) A1 9.31 Ca 2.94 Fe 8.70 Mg 2.46 Mn 1.69 K 0.13 Na 0.34 Si 30.16 Ti 0.69
351
a positive effect to concentrate the organic matter and to decrease the mineral matter in the overflow fraction. The higher concentration of liptinite in the overflow fraction would be of benefit to the liquefaction process, while the higher concentration of non-reactive macerals in the underflow would be detrimental. Results of concentration of the elements in ashes of coal samples show that in general, the elements with the highest concentration are Si, Fe, A1, Ca and K. In order to analyze changes in the concentration of the elements in the overflow fractions relative to the original coal fed to the cyclone, the concentrations are expressed in terms of proportions. Figure 2 shows the proportion values. In the overflow, it was found that Ca has the highest increase followed by Na and Ti, while Fe and Mg show the same proportions, and the rest of the elements show a decrease. The above results suggest that Ca, Na and Ti have a tendency to be associated with the organic matter, and A1, Si, K and Mn with the mineral matter. In agreement with our results, a previous study (Barraza et al. 1994) has shown that Ca shows organic affinity using float-sink separations. The above changes in the concentration proportion suggest that some elements have been removed more than others. In order to examine this trend, the results are evaluated in terms of the masses of the elements present in the original coal, overflow and underflow. Results of the elemental masses obtained in Table 2 show that the biggest decrease of all the elements occurs in the overflow fractions, however some elements show greater reductions than others. In order to compare the removal of the elements, the masses were transformed into proportions, relative to the original coal feed. Figure 3 gives the mass proportion of elements for the overflow coal fraction. Ca, Na and Ti have the highest proportion, which in terms of
3.00 1 2.50 -' 2.00 1.50 1.00 0.50
-[ 1-
0.00
AI
Ca
Fe
K
r , Mg
Mn
Na
Si
Ti
Element
Fig. 2. Concentration proportion of elements is ashes of overflow relative to original coal.
J. BARRAZA E T AL.
352
Table 2. Global and major element mass in original feed
and the coal fractions. The overflow fraction produced the highest conversion, while the underflow fraction gave the lowest conversion. The highest conversion from the overflow fraction would be associated with its high concentration of reactive macerals (liptinite+ vitrinite) as well as to its low mineral matter content, and the lowest conversion in underflow may be due to the high inertinite level. These results are consistent with those reported for Parkash et al. 1985; Joseph et al. 1991. Performing the digestion at higher pressure appears to produce a slight increase in conversion in the original coal. Higher content of light compounds, which were not released from the autoclave, may explain this increase in conversion. Despite the high conversion values obtained from the overflow fraction, it gave lower oils content and higher heavy organics material in the coal extract liquid compared to the original coal and the underflow fraction. The digestion pressure appears to affect the oils level in the original coal. Oils concentrations are higher at 10barg than 40barg, however the latter pressure, appears to reduce the concentration of asphaltenes and preasphaltenes in the product from the overflow fraction. Low heavy organics content in the underflow compared with the overflow fraction also was achieved. The product distribution obtained above includes the oils from the HAO and the oils produced from coal. HAO represents a large amount of the oils in the final product and it would have affected the distribution obtained. Therefore, an analysis is carried out on the basis of products formed from the coal alone. In order to evaluate the net product distribution, a mass balance is performed taking as basis 100kg of coal, which was separated into an overflow
cyclone, overflow and underflow coal fractions Coal samples Original Mass (g) Ash (g)
Overflow
Underflow
Global mass (g) 100.00 132.56 1.20 31.81
232.56 33.72
Element mass (g) Element AI Ca Fe K Mg Mn Na Si Ti
3.14 0.99 2.93 0.83 0.57 0.04 0.11 10.17 0.23
0.09 0.10 0.11 0.01 0.02 0.00 0.01 0.22 0.02
2.89 0.96 2.39 0.73 0.55 0.04 0.09 9.47 0.20
Basis: 100.00g of overflow as received. removal means that they were partially removed, while A1, K Mn and Si show a significant degree of removal. Again, the notable proportion of Ca and Ti in the overflow is observed. These variations in the mass of the elements suggest that they may show different behaviour during liquefaction.
Liquefaction o f original coal and coal fractions Results of conversion and product distribution in the coal extract liquids are shown in Fig. 4. Note that the oils figure includes the original HAO solvent. Differences were observed in liquefaction conversion from the original coal
0.12 - 0.10 ~-
0.08 0.06 0.04 0.02
0.00
I A1
I Ca
i [ Fe
] K
t
,
Mg
r--1
Mn
1[
I
Na
Element
Fig. 3. Mass proportion of elements of overflow fraction relative to original coal.
Si
Ti
DIRECT COAL LIQUEFACTION
353
100 90 u
80
m !
70
60 50
ji
!
40 30
~'
20 10
i
N
0o
Feed lO
0
Feed-dO
OF-I 26 40
DciOn.ois ...
I
nll Asphaltens
L
mHeavy Organics
UF-I26-40
Feed to autoclave
Fig. 4. Conversion and product distribution.
fraction and then liquefied in the autoclave. For this mass balance, it was assumed that the amount of gas produced is 2% of the coal converted and all the HAO during the liquefaction finish as oils. Results of the net product distribution are presented in Fig. 5. These findings show that, in general, the overflow fraction produced the highest net percentage of oils, with small reductions in the asphaltenes and preasphalteues. However, the heavy organics content is the highest. The underflow produced
the lowest oils percentage value. These results show that the overflow fraction has a beneficial effect towards production of oils. With regard to the element distribution during the digestion process, results of the concentration of the elements and ash content of the coal extracts are shown in Table 3. The concentration of the element and ash of the HAO are also reported in the same table. These findings show that for digestion at higher pressure, a reduction in the ash content for the
50.0 45.0 r'loils mAsphaltenes mPreasphaltenes
40.0
mHeavy Organ es
35.0 ,5
30.0
g~ 25.0 ~'
2
1. Original coal, 10 bar 2. Original coal, 40 bar 3. Overflow, 40 bar 4. Underflow, 40 bar
20.0 15.0 10.0 5.0 0.0 !
2
3 Feed to auloclave
Fig. 5. Net products distribution oi] a basis of 100kg as received.
354
J. BARRAZA E T AL. Table 3. Major element analysis in ashes of coal extract liquids and HAO Feed to autoclave
Original
Digestion Pressure, bar Extract ash value, % w/w
10 0.045
Concentration of element in ash (% wt) A1 Ca Fe K Mg Mn Na Si Ti
2.96 17.26 5.9 0.04 4.6 0.52 0.42 1.21 1.37
Overflow
Underflow HAO
40 0.023
40 0.048
40 0.041
1.76 17.53 9.9 0.05 0.62 1.9 0.35 2.92 1.09
0.37 22.24 9.4 0.03 0.22 0.18 0.19 1.21 1.37
0.55 20.83 10.5 0.03 0.65 1.17 0.22 0.59 0.68
0.003
8.11 1.65 10.9 1.07 0.67 0.16 1.85 15.2 N/D
N/D: Not detected.
extract solution from the original coal is achieved. These findings are in agreement with the results obtained by Cloke (1986). Thus, a digestion pressure of 40 barg was used with the overflow coal fractions, since it was expected that this would give a low ash content in the coal extracts. However, the results of ash content in the extracts from the coal fractions show that the beneficiated overflow did not produced a coal extract with ash content lower than the coal extract from the original coal. This may be due to the differences in heavy organics material content in the coal fraction as is shown in Fig. 5. Differences in the proportion of elements found in the extracts, defined as the concentration of the element in the ash of the extract divided by the concentration of the element in the ash of the feed to the autoclave, are shown in Fig. 6. Ca and Mn show the highest increase in all the coal extract liquids, however, Ca gives the lowest proportion in the extract solution from the overflow fraction, while Mn has the lowest
~9
t~
s
proportion in extract solution from the original coal digested at 10 barg. With regard to Na and Mg, both show a high proportion in the extract liquid from the original coal and a low proportion in the extract from the overflow. Ti and Fe do not show variations, while A1, K and Si give a great reduction in the majority of extract solutions. Kovach et al. (1978), have shown that the alkali metals, Ca and Na, and the acidic metals, Ti and Si, are greater deactivators of the hydrocracking catalysts used in a two-stage liquefaction process. In the present study, a reduction in the concentration of these elements using overflow fractions has been achieved, which would be beneficial in prolonging the hydrocracking catalyst life. The above results show that there are changes in the major element proportions between the extracts produced using different types of feed to the autoclave. In order to examine this trend the results are recalculated in terms of the masses of the elements in each extract and coal samples fed
16.00 14.00
ii
12.oo 10.00
~
o
s.oo
6.00 4.00 2.00 0.00
[] Original10 I~1Original40 IlllOF 1.2640 [] UF 1.26 40
_ _ _
AI
Ca
Fe
K
Mg
Mn
Na
Si
Element
Fig. 6. Concentration proportions of elements in coal extracts relative to feed autoclave.
Ti
355
DIRECT COAL LIQUEFACTION Table 4. Mass of major element in coal extract solutions
Feed to autoclave
Overflow
Original coal
Digestion Pressure, bar
10
40
Mass of coal converted, dmmf (g) 65.90 Mass of extract (g) 260.58
Underflow 40
40
71.60 266.16
88.63 282.86
62.98 257.72
Mass of element in extracts (mg) Element A1 Ca Fe K Mg Mn Na Si Ti
3.47 20.24 6.92 0.05 5.39 0.61 0.49 1.42 1.61
to the autoclave. For this, a basis of 100g of coal fed to the autoclave was used and the material vented was estimated 2% of the coal converted. The results for each type of coal fed to the autoclave are shown in Table 4. Differences were observed in both the amount of coal converted and the mass of the elements in the extracts. The highest mass of coal converted was obtained with the overflow fraction, which is due to the high conversion achieved. Also, it was found that the majority of masses of the elements show a decrease in the extracts from the original coal as the pressure is increased. However at 40 barg, the masses of some elements in the extract from the overflow were not reduced compared to the masses of the same elements in the extract from the original coal. Therefore, in order to analyse which elements give a reduction in the extracts from overflow relative to the original coal, the masses of the elements are transformed into the proportions shown in Fig. 7. Results show a decrease in the proportion for A1, Mg, Mn and
1.08 10.73 6.06 0.03 0.38 1.16 0.21 1.79 0.67
0.50 30.20 12.76 0.04 0.30 0.24 0.26 1.64 1.86
0.58 22.01 11.09 0.03 0.69 1.24 0.23 0.62 0.72
Si. However, Ti, Ca and Fe have the highest increase and K and Na show approximately the same proportions. It indicates that the elements Ca, Ti and Na, which are considered to be the deactivating elements of the hydrocracking catalyst, were not reduced in the extracts using the overflow fraction compared to the original coal. These findings again may suggest the association of Ca and Ti with the organic matter, such as has been shown in other studies (Robatt et al. 1984; Cloke 1986)
Conclusions
1. The dense medium cyclone unit produced an overflow coal fraction of high concentration in organic matter and low concentration in mineral matter. 2. In general, liquefaction results show an improvement in conversion for the overflow
J
2.5
1.5 ~-
1
0.5 +--F , [---] ~
0 AI
Ca
Fe
K
Mg
Mn
Na
Si
Ti
Element
Fig. 7. Mass proportion of element in extract from overflow relative to extract from original coal.
356
J. BARRAZA E T AL.
fractions over the original coal. By contrast, the underflow fraction gave the lowest conversion. Given the high mineral matter content to remove in the filtration process, the underflow fraction is not a material desired for liquefaction purposes. 3. A shift in the net product distribution towards higher oils and lower asphaltenes and preasphaltenes were obtained in the liquid product using the overflow fraction. However, it gave the highest heavy organics content, which would be detrimental for the hydrocracking stage. 4. The beneficiation did not reduce the ash content in the filtered coal extract solution compared to the ash content in the extracts from the original coal. However, the quantity of mineral matter to be removed at the filtration stage was reduced. 5. Reduction in the proportion of masses of A1, Mg, Mn and Si, in the extracts from the overflow relative to the original coal were obtained. However, some of the strongest deactivating elements such as Ti and Ca show a large increase in proportion. The authors gratefully acknowledge the award of a grant in aid of research from the European Coal and Steel Community (ECSC), the Colombian Institute of Science and Technology (COLCIENCIAS), the British Coal Utilization Research Association and the United Kingdom Department of Trade and Industry. The assistance of British Coal Liquefaction for provision of samples is acknowledged. The views expressed are those of the authors and not necessarily those of the funding bodies.
References BARRAZA, J., GILFILLAN,A., CLOKE, M. & CLIFT, D. 1994 International Coal Conference on Coal Bed Methane, Cardiff, Wales, September. CLOKE, M. 1986. Fuel, 65, 417. - - 1987. PhD. Thesis, University of Nottingham - - , BELGHAZI,A., MARTIN, S., KELLY, B., SNAPE, C. E., MCQUEEN, P. & STEEDMAN, W. 1993. International Conference on Coal Science, Banff, Canada, - - , CLIFT, D., GILFILLAN,A., MILES, N. & RHODES, D. 1994. Fuel Processing Technology, 38, 153. CRONAUER, D. & SWANSON, A. 1991. 201 ACS National Meeting, Atlanta, Georgia, 14. DYRKACK, G. R. & HORWlTZ, E. P. 1982. Fuel, 61, 3. HAMILTON, S. 1986. Thesis M. Phil, University of Nottingham. JOSEPH, J. T., FISHER, R. B., MASIN, C. A., DYRKACZ, G. R. & BLOOMQUIST, C. A. 1991. Energy and Fuels, 5, 724. KEOGH, R. A. & POE, S. H. 1987. International Conference on Coal Science, The Netherlands, 289-294. KING, H. H., DYRKACKZ, G. R. & WlNANAS, R. E. 1984. Fuel, 63, 341. KOVACH, S. M., CASTLE, L. J. & BENNETT,J. V. 1978. Industrial and Engineering Chemistry; Production, Research and Development, 17, 1 62-67. ONER, M., ONER, G., BOLAT, E., YATIN, G., KAVLAK, C. & DINCER, S. 1994. Fuel, 73, 10. PARKASH, S., LALI, K., HOLUSZKO,M. & DU PLESSIS, M. P. 1985. Liquid Fuel Technology, 3, 3. ROBBATT,A., Jr., FINSETH, D. H. & LETT, R. G. 1984. Fuel, 63, 1710-1714. RHODES, D., HALL, S. T. & MILES, N. J. 1993. XVIII International Mineral Processing Congress, 23-28. STEELER, M. 1987. International Conference on Coal Science, The Netherlands, 115-118.
Conversion of low rank coal into liquid fuels by direct hydrogenation B. R. A L E K S I ( ~ ~, M. D. E R C E G O V A C ,
O. G. C V E T K O V I ( ~ 3,
B. Z. M A R K O V I ( ~ ~, Y. L. G L U M I ( ~ I ( ~ 3, B. D. A L E K S I ( ~ ~ & D. K. V I T O R O V I ( ~ 3
'IChTM, Center of Catalysis and Chemical Engineering, NjegoYeva 12, 11000 Belgrade, FR Yugoslavia 2 Faculty of Mines and Geology, University of Belgrade, DjuYina 7, 11000 Belgrade, FR Yugoslavia 3 IChTM, Center of Chemistry, Njegogeva 12, 11000 Belgrade, FR Yugoslavia Abstract: A study of low-rank coal conversion into liquid products by direct catalytic hydrogenation was undertaken. A soft brown coal from the 'Tamnava' field of the Kolubara mines characterized by a huminite reflectance of 0.27+0.03%RR, ash content of 10.4wt%(db), carbon content of 64.0wt% (daf), and volatiles ca. 50wt%(db), was submitted to liquefaction in a batch reactor. The effect of reaction parameters on both the yield and nature of liquefaction products was studied for temperatures ranging from 365 to 440~ and pressure from 13.5 to 16.5 MPa, with process duration from 1 to 8 hours. The total coal conversion was high at all applied reaction conditions (84-93%, daf coal basis), pointing to a high reactivity of this coal. The yield of particular liquid products varied markedly depending on temperature and residence time. The yield of light-oil (n-heptane soluble products) increased and that of asphaltenes decreased by increasing the temperature and prolonging the residence time. Changes in petrographic composition of the coal were examined by microscopic analysis. At more severe reaction conditions the content of semicoke and coke increased. With the increase of temperature at mild conditions, the proportions of both the reacted coal and granular residue increased, while the cenospheres and mineral matter decreased.The nature of the changes observed in the organic and mineral components of the coal grains was used to correlate the degree of coal conversion with the experimental conditions.
Contrary to the previous conception that only bituminous and subbituminous coals might be used for liquefaction, much research effort has recently been directed toward determining the susceptibility of low-rank coals to liquefaction. It has been reported (Derbyshire & Stansberry 1987; Mondragon et al. 1988) that certain soft brown coals liquefy more readily than high-rank coals. There is a number of research papers dealing with the correlation between liquefaction behaviour and various properties of coals, especially coal petrography (Artemova et al. 1989; Given et al. 1980; Hower et al. 1991; Parkash et al. 1984), but there is a lack of generally valuable correlations. One of the reasons is the chemical and petrographic heterogeneity of coals within the same rank. The other reason might be the complex interaction of coal mineral components with the organic part of coal during liquefaction. Therefore, data obtained under certain experimental conditions of a particular coal liquefaction cannot be considered reliable for other types of coals or even for the same coal treated by a different liquefaction procedure (Tischer & Utz 1983).
The work presented here was aimed at obtaining information on the behaviour of soft brown coal from the 'Tamnava' field of the Kolubara mines (Serbia) during liquefaction by direct catalytic hydrogenation. The effect of the reaction parameters on the changes in the coal petrographic composition was examined and correlated with the changes observed in the degree of coal conversion and in the liquefaction product yields.
Experimental method Liquefaction of pulverized coal (< 160 #m) slurry in tetralin in the presence of a granulated cobaltmolybdenum hydrodesulphurization catalyst was performed by a stream of hydrogen in a batch reactor. The reaction parameters were varied in the following ranges: temperature, from 365 to 440~ pressure, from 13.5 to 16.5 MPa; residence time in the reaction conditions, 1 to 8 hours. The choice of particular parameters was based on preliminary experi-
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 357-363.
358
B. R. ALEKSIC E T A L .
mental work. After cooling, the liquefaction products were separated according to a previously established procedure (Vitorovi6 et al. 1991). After filtration the solid residue and the catalyst were rinsed several times with the filtrate itself, then the catalyst was separated by sieving and dried to constant weight at 300~ The liquid liquefaction products were separated from tetralin by distillation. The light oil was dissolved in n-heptane, the undissolved part consisting of asphaltenes. The liquid products were characterized by ultimate analysis and gas chromatography (30m Supelco capillary SPB-1 column). The solid residues were characterized by ultimate and micropetrographic analyses. The methods have been described in more detail elsewhere (Vitorovi6 et al. 1991, 1994). The degree of coal conversion (X) was calculated on the basis of the dry, ash free solid residue (R) according to: X = lO0(mo-R)/mc, where mc denotes the initial mass of the coal (dry, ash free). Microscopic examination of the solid residues and the initial coal sample was carried out on two polished blocks prepared from each sample, involving 1000 measurements according to ICCP standards (1963, 1971).
Results and discussion The characteristics of the initial coal sample are given in Table 1. The coal substance of the 'Tamnava' coal sample is in a relatively low stage of humification and gelification (0.73). The proportion of the non-gelified macerals is 42.0vo1%, and that of gelified (homogenous) macerals 31.0vo1%. According to huminite
reflectance, 0.27 + 0.03% RR, and volatiles, 49.5 (wt%, dry basis), the average coal sample belongs to soft brown coals, designated as M2 coals (Ercegovac 1986) with a xylite content of 42.0 wt%. The liquefaction yields representative of different reaction conditions (temperature, T; pressure, p; residence time, rr) are shown in
Table 1. Characteristics of Tamnava coal sample Equilibrium moisture (wt%) 12.3 Ash (wt%, dry basis) 10.4 Sulphur, total (wt%, dry basis) 1.0 Fixed carbon (wt%, dry basis) 26.1 Volatiles (wt%, dry basis) 49.5 Ultimate analysis (wt%, dry basis, ash free) Carbon 64.0 Hydrogen 5.7 Nitrogen 1.3 Oxygen (by difference) 29.0 Heating value (kJ kg-1) HHV 19 327.3 LHV 18 425.5 Macerals and minerals (vol%) Huminite 67.0 Textinite 25.0 Ulminite 21.0 Atrinite 11.0 Densinite 4.0 Gelinite 6.0 Liptinite 3.5 Inertinite 6.0 Minerals 23.5 Clay 20.0 Pyrite 2.5 Carbonates 1.0 Gelification index 0.73 Huminite reflectance, RR (%) 0.27 Xylite (wt%) 42.0
Table 2. Liquefaction yields (wt% daf coal) and coal conversion (%) Test No 1 2 3 4 5 6 7 8 9 10 11 12
Conversion
Reaction conditions
Products
T (~
p (MPa) rr (h)
Oil
Asphaltenes
Solid residue
365 365 365 365 365 400 400 420 440 440 440 440
13.5 13.5 13.5 13.5 16.5 13.5 16.5 15.0 13.5 13.5 15.0 15.0
22.6 21.8 16.8 16.5 22.2 26.6 25.5 52.5 68.2 74.4 65.4 73.0
25.5 26.3 17.0 15.7 27.0 27.8 23.5 7.2 7.8 5.1 6.2 3.3
15.7 11.5 14.5 6.7 10.3 13.8 15.3 9.2 11.5 11.3 10.8 13.6
1 4 6 8 4 4 4 4 4 8 4 8
84.3 88.5 85.5 93.3 89.7 86.2 84.7 90.8 88.5 88.9 89.2 86.4
C O N V E R S I O N OF LOW R A N K COAL INTO L I Q U I D FUELS
359
Table 3. Petrographic composition and optical characteristics of liquefaction residues of Kolubara coal ("Tamnava')
Categories of grains (vol%) No. 1 2 3 4 5 6 7 8 9 10 11 12
Unreacted and partly reacted coal Reacted coal Isotropic humoplasts Asphaltenes (pitch-like material) Cenospheres (Iso.) Semi-coke (Iso.) Coke (Aniso.) Isotropic grains (porous, A-type) Homogenous isotropic grains (high % RR) Granular residue (partly porous structure) Fragments (<0.010ram) Mineral Clay Matter Pyrite Carbonates
Test 1 2 3 4 5 6 7 8 9 10 11 12 No T (~ 365 365 365 365 365 400 400 420 440 440 440 440 p(MPa) 13.5 13.5 13.5 13.5 16.5 13.5 16.5 15.0 13.5 13.5 15.0 15.0 7-r (h) 1 4 6 8 4 4 4 4 4 8 4 8 -
-
0.5
-
0.5
0.5
2.0
-
1.0
0.5
1.0
1.0
12.5 13.0 40.0 12.0 ll.0 23.0 27.0 ll.5 28.0 24.5 18.0 10.5 4.0 6.5 tr. 0.5 6 . 0 1.0tr. tr. tr. 0.5 tr. 0.5 0.5 1.5 1.0 2.0 10.0 1.0 -
5.5 -
2.5
1.0 1.5
-
-
-
tr.
-
2.0
0.5
1.0
2.0
1.0
0.5
3.5
0.5 -
3.5 tr. 1.5 tr. 2.0 3.0 1.0
1.5 5.0 0.5 1.5
2.5 tr. 3.5 0.5 2.5 2.5
0.5 tr.
0.5
2.0 3.0 4.5 2.0 2.0 ll.0 4.0 1.0 4.0 0.5 tr. 1.0 0.5
0.5
0.5
42.0 45.0 35.5 47.0 61.0 54.0 44.0 69.0 52.0 46.0 61.0 59.0 3.0
2.0
3.5
1.0
2.0
4.0
4.5
21.0 24.0 12.0 31.5 3.0 3.0 1.0 3.0 . . . . .
9.0 3.0
9.5 1.5 2.0
8.0 4.5 2.5 4.0 2.0 -
Table 2. High coal conversion was observed at all investigated conditions but the relative yields of liquefaction products varied. By prolonging the residence time at lower temperature and pressure (test 4 vs. test 1) the a m o u n t of solid residue decreased, the conversion was higher, but the relative yields of liquid products decreased probably due to formation of gases. At m o r e severe reaction conditions (tests 9-10 and 11-12), the yield of n-heptane soluble liquid product (oil) increased as a result of a higher degree of h y d r o g e n a t i o n of coal particles dissolving at the beginning of the process. The effect o f temperature m a y be seen if tests 2, 6 and 9 are compared. The overall coal conversion and the solid residue did not change markedly, but the yield of oil increased, at the expense of asphaltenes whose yield decreased. Results of tests 2 and 5, or 9 and 11, respectively, showed that the effect of pressure on the liquefaction yields was less pronounced. The categories of grains presented in Table 3 were identified and estimated by micropetrographic analysis of the solid residues. The residue grains were placed in different categories (Guyot & Diessel 1981; Ercegovac 1986; Vitorovi6 et al.
1.0
3.0
7.5
0.5
1.5
8.0 9.0 9.5 3.0 1.5 3.5 3.5 4.0 1.5 -
1994) as follows: unreacted and partly reacted coal (cat. 1; up to 2.0 vol%), reacted coal (cat. 2; 10.5-40.0vo1%), isotropic humoplasts (cat. 3; traces to 6.5 vol%), asphaltenes (pitch-like material; cat. 4; traces to 2.0 vol%), cenospheres (iso., cat. 5; traces to 10.0 vol%), semi-coke (iso., cat 6; 1.0-11.0 vol%), coke (aniso., cat. 7; traces or 0.5-4.0vo1%), isotropic grains (A type; cat. 8; traces to 3.0vo1%), h o m o g e n e o u s isotropic grains (cat. 9; traces to 3.5vo1%), granular residue (partly porous structure; cat. 10, 35.5-69.0vo1%), fragments ( < 0 . 0 1 0 m m , cat. 11; 0.5-7.5vo1%) and mineral matter (cat. 12; clay, 3.0-31.5vo1%; pyrite, 1.0-4.0vo1%; and carbonates, max. 2.0vo1%). The microscopic appearance of the solid liquefaction residues is shown in Figs 1 & 2. The effects of residence time and temperature on the proportions of the particular grain categories are shown in Tables 4 & 5 respectively. The extended time o f liquefaction (from 1 to 8 hours) at mild conditions (365~ and 13.5 MPa) (Table 4 tests 1, 2 and 4) affected the coal petrographic composition as follows: the proportions of the reacted coal and semi-coke plus coke did not change greatly; the a m o u n t of
360
B. R. ALEKSIC E T AL.
Fig. 1. Microscopic view of solid liquefaction residues of Kolubara coal, field Tamnava: (a) partly reacted coal, the beginning of decomposition of huminite (test 6); (b) reacted coal with pitch-like material (? asphaltenes) and small inclusion of inertinite (test 9); (c) huminite plasticity and the beginning of formation of humoplasts after batch hydrogenation (test 1); (d) individual and coalesced humoplasts (solvent affected material) and granular residue with fine grained mineral matter (heat affected material - test 2); (e) isotropic humoplasts with higher reflectance (test 12); (f) fused and partly carbonized material with degasification pores (test 7); (g) granular residue (down) and reticulated isotropic cenosphere with thick wall (test l); (h) mixed grains: reacted coal (upper), granular residue (left) and weak - anisotropic semi-coke (test 12); (i) highly reflecting isotropic homogenous grain (left), porous weak - anisotropic semi-coke and highly reacted coal (gray) (test 7). Reflected light, oil, • 360.
C O N V E R S I O N OF LOW R A N K COAL INTO L I Q U I D FUELS
361
Fig. 2. Microscopic view of solid liquefaction residues of Kolubara coal, field Tamnava: (a) characteristic view of specific category of oxidized grains with cracks (isotropic grains - type A; test 7); (b) same category of grains with a carbonized outer wall (aniso.; test 7); (c) porous isotropic grain showing development of vacuoles (black holes) - 'transition stage' to semi-coke formation ? (test 6); (d) "transition stage' to semi-coke, formation between the granular residues (test 12); (e) porous isotropic semi-coke (test 12); (f) highly anisotropic grain of coke with high degree of reflectance (test 12); (g) specific art of agglomeration of reacted coal grains (test 11); (h) mixed grains: simple cenosphere and pyrrhotite between the granular residues (test 1); (i) mixed grains: slightly altered huminite particle displaying indications of partial softening, single thin walled cenospheres and pyrite-pyrrhotite agglomerate between the granular residues (test 1). Reflected light, oil, x360.
362
B. R. ALEKSIC E T AL.
Table 4. Change in petrographic" composition of the solid residues with the residence time at different reaction conditions Categories of grains (vol%)
A. T= 365~ p-- 13.5 MPa
Tr=
No. 2 5 6+ 7 10 12
Reacted coal Cenospheres (Iso.) Semi-coke & coke Granular residue Mineral matter
1 12.5 10.0 1.0 42.0 24.0
1h
B. T= 440~ p = 13.5 MPa
C. T= 440~ p = 15.0 MPa
rr=4h
rr=8h
rr=4h Tr=8h Test No.
Tr=4h
Tr=8h
2 13.0 5.5 45.0 27.0
4 12.0 1.0 1.5 47.0 34.5
9 28.0 tr. 52.0 11.0
11 18.0 3.0 3.0 61.0 13.0
12 10.5 4.5 15.0 59.0 7.0
10 24.5 2.0 6.0 46.0 12.5
Table 5. Change in petrographic composition of the solid residues with temperature Categories of grains (vol%)
No. 2 Reacted coal 5 Cenospheres (Iso.) 6 + 7 Semi-coke & coke 10 Granular residue 12 Mineral matter
A. Tr=4h; p = 13.5MPa
B. r r = 4 h ; p = 15.0 MPa
C. % = 8 h ; p = 13.5 MPa
T= 365~ T=400~
T= 440~
T=420~ T=440~ Test No.
T= 365~
T=440~
2 13.0 5.5 45.0 27.0
9 28.0 tr.
8 11.5 2.5 4.0 69.0 8.5
4 12.0 1.0
10 24.5 2.0 6.0 46.0 12.5
6 23.0 tr. tr. 54.0 13.0
isotropic cenospheres decreased considerably (from 10.0 to 1.0 vol%); both the granular residue and the amount of mineral matter increased, in accordance with the observed higher conversion of the organic part of coal at longer residence time (Table 2). The effect of the residence time at the most severe conditions (440~ and 15.0 MPa) was demonstrated by an increased formation of semi-coke and coke (Table 4, tests 11 and 12). The observations confirmed the postulated mechanism of coal liquefaction (Shinn 1984). The observed decrease of the mineral matter content was due to the increaase of the amount of coke and semicoke in the liquefaction residue. In contrast to experiments at mild conditions, the prolonged residence time at higher temperature and pressure caused a slight increase in the cenosphere content. The mechanism of particular maceral changes during the coal liquefaction is poorly understood at present. Therefore, it is difficult to draw any definite conclusion on the experimental conditions promoting formation or decay of cenospheres.
-
52.0 11.0
11 18.0 3.0 3.0 61.0 13.0
1.5
47.0 34.5
By increasing the temperature at mild conditions, the proportions of both the reacted coal and the granular residue increased, while those of cenospheres and mineral matter decreased (Table 5, tests 2, 6 and 9). Semi-coke and coke were not observed at these conditions. These results were in accordance with the increase of oil yield with temperature (Table 2, tests 2, 6 and 9). At a higher pressure, semi-coke and coke appeared at higher temperatures (Table 5, tests 8 and 11). By increasing the pressure from 13.5 to 16.5 MPa an effect was observed at mild conditions (365~ 4 h) in tests 2 and 5 (Table 3). The maceral composition of the two residues was similar except for the higher proportion of granular residue and the lower proportion of mineral matter in the sample obtained at higher pressure. It is interesting to note that an increase in the proportions of mezophase products, especially the semi-coke and coke, was not observed by increasing the pressure at mild conditions. However, at more severe reaction conditions the proportion of semi-coke and coke
CONVERSION OF LOW RANK COAL INTO LIQUID FUELS increased probably as a result of simultaneous effects of the temperature, pressure and residence time (Table 3 tests 10 and 12). The high yield of oil and the low yield of asphaltenes, observed at 440~ and 13.5 or 15.0MPa, indicated that polymerization of dissolved and partly hydrogenated coal fractions did not occur even during a prolonged reaction time. Hence, the coking was a solid particles process, involving the remaining carbonaceous part of the coal. The results obtained indicate that at 440~ the reactions in coal yielding liquid and gaseous products (dissolution, volatilization, decomposition of the coal substance) end in a period of time shorter than four hours, under the applied experimental conditions. During prolonged retention time, hydrogenation of the liquid phase and thermal processes in the remaining solid phase continue with no signs of any significant interaction. At lower temperatures and pressures the conversion of coal did not seem to have been terminated during the applied residence times. Further investigations are necessary for a more reliable correlation of petrographic compositional changes of the 'Tamnava' coal during the liquefaction process with the liquefaction yields.
Conclusions
A high degree of conversion (>84%) of the soft brown coal 'Tamnava' was observed during liquefaction by direct catalytic hydrogenation. Varying the reaction conditions (the temperature, pressure and residence time), high yields of liquid products were obtained. The high reactivity of the coal was confirmed by petrographic analysis which showed that there was no unreacted coal in the solid residues with most of the liquefaction runs. The petrographic composition of the residues depended on the reaction conditions, but a more reliable correlation requires additional investigations. Nevertheless, petrographic analyses produced valuable data concerning the changes in the 'Tamnava' soft brown coal during liquefaction. This work was supported in part by the Research Fund of Serbia (Project No 0816). The authors are grateful to the Kolubara Mine for providing the coal 'Tamnava' samples.
363
References
ARTEMOVA, N. J. KASATOCHKINA, L. J. CHIZHEVSKAYA, V. R. & SHULAKOVSKAYA,L. V. 1989. The effect of the petrographic composition of coals on their hydrogenation. Khim. Tverd. Topl. (Chem. Solid Fuels), 4, 75-79 (in Russian). DERBYSHIRE, F. & STANSBERRY,P. 1987. Comments on the reactivity of low-rank coals in liquefaction. Fuel, 66, 1741-1742. ERCEGOVAC, M. 1986. Brown and black coal hydrogenation in comparative studies of their petrographic composition and solid residue. An. Gdol. de la P~nins. Balkanique, 50, 419-441 (in Serbian). GIVEN, P. H., SPACKMAN,W., DAVIS, A. & JENKINS, R. G. 1980. Some proved and unproved effects of coal geochemistry on liquefaction behaviour with emphasis on U.S. coals. In: WHITEHURST, D. D. (ed.) Coal Liquefaction Fundamentals. American Chemical Society Symposium Series, 139, 3-34. GUYOT, R. E. & DIESSEL,C. F. K. 1981. Petrographic studies on insoluble residues of hydrogenated coals. International Journal of Coal Geology, 1, 197-207. HOWER, J. C., KEOGH, R. A. & TAULBEE,D. N. 1991. Petrology of liquefaction residues: maceral concentrates from a Pond Creek durain, eastern Kentucky. Organic Geochemistry, 17, 431-438. ICCP - Internationales Lexikon ffir Kohlenpetrologie 2. Ausgabe, Paris, Centre National du Recherche Scientifique, 1963. und Erg~inzungen Band zur 2. Ausgabe, 1971, Paris. MONDRAGON, F., QUINTERO, G., ACOSTA, R. & JARAMILLO, A. 1988. Liquefaction characteristics of some Columbian coals: 1. Reactivity in catalytic hydrogenation. Fuel, 67, 1709-1711. PARKASH, S., DU PLESSIS, M. P., CAMERON,A. R. & KALKREUTH, W. O. 1984. Petrography of low rank coals with reference to liquefaction potential. International Journal of Coal Geology, 4, 209-234. SHINN, J. H. 1984. From coal to single-stage and twostage products: a reactive model of coal structure. Fuel, 63, 1187-1196. TISCHER, R. E. & UTZ, B. R. 1983. Comparison of normal and rapid heat-up modes in a batch screening test for coal liquefaction catalysts. Industrial and Engineering Chemistry, Product Research and Development, 22, 229-233. VITOROVI(~, D., ALEKSIC, B. R., KONTOROVIC, S. I., ALEKSIC, B. D., ERCEGOVAC, M., MARKOV1C, B. Z., BOGDANOV,S. S. & CVETKOVlC,O. G. 1991. Liquefaction of brown coal prepared by grinding under different conditions. Fuel, 70, 849-855. VITOROVIt~, D. K., ALEKSIt~, B. R., ERCEGOVAC, M. D., ALEKSI{~, B. D., KONTOROVIC, S. I., MARKOVlC, B. Z., CVETKOVI~, O. G. & MITROVSKI, S. M. 1994. Liquefaction behaviour of Kolubara soft brown coal. Fuel, 73, 1757-1765.
Desulphurization of low-rank coals by low-temperature carbonization R. A S M A T U L U , N. A C A R K A N ,
G. O N A L & M. S. C E L I K
Istanbul Technical University, Mining Engineering Department, Coal and Minerals Processing Section, Ayazaga, 80626 Istanbul, Turkey Abstract: A lignite sample from the Istanbul region with 14.18% inherent moisture, 11.01%
ash, 1.86% total sulphur, 47.24% volatile matter, 41.75% fixed carbon and 5590kcal kg-1 calorific value as dried basis has been subjected to a set of systematic low-temperature carbonization tests. The tests have been carried out as a function of particle size, temperature and heating time. A semicoke product containing 15.22% volatile matter, 16.67% ash, 68.04% fixed carbon contents with 63% desulphurization on the basis of total sulphur has been obtained under the optimum conditions of 650~ temperature and 50 minutes of heating time. The product has a calorific value of 6403 kcal kg-1 but does not have enough strength for use as a fuel and thus needs to be improved by briquetting.
Istanbul, one of the major populated cities in the world, is currently encountering a severe air pollution problem in the winter season, partly caused by the burning of low-rank coals. The better quality coals extracted from coal mines in the vicinity of Istanbul are generally not upgraded by coal preparation processes. These coals are produced about 200-400mm in size with 30-40% moisture, 35-40% volatile matter, 1-3% total sulphur contents and calorific values of 2400-3800 kcal kg -1 as received basis. Coal used for household heating makes up about 60% of the total heating requirement in Istanbul. Lignites produced from the Istanbul region constitute about 80% of this total consumption. The high moisture, volatile matter and sulphur contents are the major causes of coal-based air pollution in Istanbul. High moisture not only leads to low combustion efficiencies but also the discharge of unburned fines into the atmosphere. In addition, since domestic boilers and stoves are not specifically designed for burning low-rank coals, the volatile matter is emitted into the air before it is fully burned. This problem results in hydrocarbon and particulate matter emission. All these problems mentioned above led to the development of new technologies for the minimization of air pollution originating from coal. Since sulphur is a major pollutant endangering the human life, various desulphurization techniques, i.e. desulphurization prior to combustion, desulphurization during combustion and, postcombustion desulphurization processes need to be systematicaly tested (Celik & Somasundaran 1994). One of these processes involving low temperature carbonization at temperatures in the range of 400-700~ aims to reduce the sulphur content of coal (Gazanfer 1983; Lowry 1945; Von 1982; Onal et al. 1995; Sciazko et al.
1993). Low-temperature drying processes have been employed since 1920s and the first patents by Fleissner appeared a few years later (Fleissner 1927 and 1928). In this process hot air or gas is sent to accomplish drying of either surface or inherent moisture of coal. Since the inherent moisture in low-rank coals is distributed in the form of very fine capillaries, high temperatures are necessary. This leads to fragmentation and consequently production of fines. Also, the dried coal is sensitive to spontaneous combustion. Work done in this area in the last two decades (Koppelman 1977; Cole & Ness 1977; Verschuur et al. 1976; Wash 1977; Evans & Sieman 1979) have focused on steam or hot water drying. The use of hot water or steam drying has advantages of (i) production of low moisture coal, (ii) reduction in the tendency of coal to reabsorb moisture, and most importantly, (iii) reduction in fragmentation and spontaneous ignition. A pilot plant erected in Wyoming in 1992 and Ceska Palivova commercial plant built now in the Czech Republic working under 80 atmosphere pressure and 400~ represents the most up to date technology in this area (Gentile 1995). It is the objective of this study to conduct lowtemperature carbonization tests on a typical Istanbul region lignite with the aim of reducing the sulfur levels in the coal. Table 1.
Analysis of the lignite sample on dry basis
% Moisture % Ash % Total sulphur % Combustible sulphur % Volatile matter % Fixed carbon Upper calorific value, Kcal kg-1
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 365-369.
32.50 11.01 1.86 1.05 47.24 41.75 5590
366
R. ASMATULU E T AL.
Experimental procedure Materials
Low-temperature carbonization tests were carfled out on representative samples taken from Istanbul-Yenikoy region lignites. The complete analysis of the lignite sample on dry basis is presented in Table 1. The original sample of minus 100mm in size was crushed below 50ram and then classified into three size fractions. The analyses of these size fractons are shown in Table 2. Methods
The carbonization tests were performed in a 10 cm diameter cylindrical retort with an interior
volume of 1500 cm 3. The discharged gases were transferred to a cooling system via a screwed gate. The interior volume of the retort limited the upper size of the coal sample to 50 mm. The tests were conducted with 500 g samples brought to an inherent moisture of 14%. The furnace was preheated to desired temperature and the sample placed for a certain period. After the sample was taken out and cooled, the weight loss, volatile matter and total sulphur contents, size analysis, and drum strength measurements were performed (Table 2). The experiments were performed as a function of particle size ( 1 9 x 5 0 , 10x 19, and 1 x 10mm), temperature (400-700~ and heating period (20-120min). The physical and chemical properties of the semicoke product were also determined.
Table 2. The analysis of coal sample on dry basis as a function of particle size Size fraction (mm)
% by weight
% volatile % fixed matter carbon
% total sulphur
% ash content
calorific value (kcal/kg)
19 x 50 10 x 19 1 x 10
69.8 16.2 14.0
47.82 46.47 45.60
1.80 1.84 2.10
10.30 11.68 12.96
5598 5470 5270
41.88 41.85 41.44
Table 3. Total sulphur contents in semicoke products under different conditions Temperature ( ~
Heating time (min)
Size (mm) 19x50
10x 19
1 x 10
400
20 40 60 80 100 120
1.84 1.76 1.61 1.57 1.51 1.44
1.91 1.83 1.76 1.64 1.51 1.42
2.29 2.22 2.10 2.03 1.98 1.85
500
20 40 60 80 100 120
1.77 1.56 1.42 1.36 1.31 1.25
1.83 1.67 1.53 1.38 1.31 1.28
2.17 2.07 1.92 1.89 1.80 1.68
600
20 40 60 80 100 120
1.70 1.61 1.39 1.32 1.27 1.2
1.74 1.59 1.45 1.39 1.28 1.21
2.11 1.98 1.85 1.65 1.52 1.46
700
20 40 60 80 100 120
1.44 1.31 1.28 1.21 1.14 1.06
1.60 1.50 1.42 1.31 1.18 1.10
2.06 1.94 1.80 1.58 1.41 1.35
DESULPHURIZATION OF LOW-RANK COALS Results and discussion The total sulphur contents obtained upon low temperature carbonization of Yenikoy lignite at three different size fractions (19 x 50, 10 x 19, and 1 x 10mm) are presented in Table 3. The percent sulphur removal for each size fraction is respectively illustrated in Figs 1 through 3 as a function of heating time. The most important factor in the carbonization and also in the desulphurization process appears to be the level of temperature. For instance, while a 60% desulphurization is achieved on the semicoke product at 700~ after 40 rain and this accounts for about 90% of the total desulphurization, at 100
.... -50+19
9O
> 0
15 9 V.M.
mm
80
o 400 ~ o 5 0 0 *C zx 6 0 0 *C
l J ]
1 O0
120
70
i,i r~ 60 rY
A
u_ _J
5O
~o pZ
40
0 nL~ IX.
3O
2O I0
20
40
60
CARBONIZATION
80 TIME,
rain.
Fig. 1. The percent sulphur removal of coal as a function of heating time for 19 • 50 mm size fraction. 100 rnm
1oo F
15 ~ V . l d . /
90 iI .< > 0 "~ i,i
400~ even after 120 min only about 50% of the desulphurization is attained. The slope of the curves in Figs 1-3 also shows this trend, i.e. the slope increases with increasing the temperature. The results reveal that at low temperatures and initial heating periods only the removal of moisture is achieved. Only above 400~ does the volatile matter and sulphur begin to separate from the solid. The sulphur removal at 700~ and 120 min of heating attains its peak value of about 69.4%. Carbonization of coal is generally found to exhibit a parallel behaviour to that of desulphurization. For instance, the fixed carbon level of 41% has been raised to 66-70% with calorific values in the range of 63006500 kcal/kg upon carbonization. At this level of carbonization about 70% of sulphur removal is achieved. All the three size fractions exhibit a similar trend in that at 400~ and in the beginning of heating the sulphur removal remains at low levels and increases with increasing temperature. The lower sulphur removal levels from the finer particle sizes can stem from two contributing effects. First, as the coal bed used is fixed, the diffusion of heat through the particle takes some time and in such cases, coarser particles are more advantageous. Second, the systematic increase in the fixed carbon content as a function of size and temperature indicates that with increasing the temperature coal becomes more amenable to sudden shocks which in turn lead to higher carbonization levels. This difference is minimized as the temperature decreases (Asmatulu et al. 1995). The net affect appears to govern the results presented in Figs 1-3.
/
-19+10
o D "0
80
mm o a ~, O
I
70 _
,~ >
80
o
7o
400 ~ 500 ~ 600 ~c 700 ~
r 6O
50
2
5o
40
(/3
40
~
30
._J
~: 3o
U ~
. . . . 15 % V.M. -i0+1
I 90
400 500 600 700
60 u_ _J
367
0
C)
u
20
20 n
10
10
0-
0 0
20
40
60
80
IO0
CARBONIZATION TIME, rn[n.
Fig. 2. The percent sulphur removal of coal as a function of heating time for -19 + 10mm size
120
0
20
40
60
80
1O0
CARBONIZATION TIME, rain.
Fig. 3. The percent sulphur removal of coal as a function of heating time for -19 + 10ram size
120
368
R. ASMATULU E T AL.
Table 4. Physical and chemical properties of the semicoke product obtained at 650~ Size fraction (mm)
Moisture Ash CombustibleTotal Volatile Fixed Calorific Drum test Oversize Ignition (%) content sulphur sulphur matter carbon v a l u e oversize (%) temp. (%) (%) (%) (%) (%) (kcal/kg) (%) (~
-50 + 19 3.87 -19 + 10 3.94 -10 + 1 4.02
19.95 17.20 19.59
0.37 0.34 0.62
1.39 1.41 1.79
Coal undergoes fragmentation upon carbonization due to the removal of moisture and volatile matter from coal. This is strongly dependent upon the particle size and temperature. For example, the 19 • 50 mm size fraction treated at 400~ for 20min resulted in fragmentation producing only 13% below 19 mm, whereas at 700~ 45.6% of the coal passed below 19mm after 20 min heating. The fragmentation of coal also reduces the strength of coal. Low temperature coking on non-coking Mckinley and Crown lignites has been tested at 600-700~ and found to increase the fixed carbon content from 45 to 66%. These tests were performed in the size range of -75 + 20 mm. 20% of the coal was found to pass below 20 mm. Similar results have been also obtained on German and Turkish lignites (Lowry 1945; Von 1982; Onal et al. 1995; Asmatulu et al. 1995). Low temperature coking of Polish bituminous coals is now exploited on commercial scale (Sciazko et al. 1993). Low-temperature carbonization increases the calorific value of coal from 5200 kcal kg-I to 6000-6600kcalkg -1, depending upon the temperature and residence time of the process. The results reveal that 650~ and 50 rain are respectively the optimum conditions for Istanbul Region lignites. The physical and chemical properties of the semicoke product obtained under optimum conditions and for different particle sizes are given in Table 4. Coals with high moisture and volatile matter contents such as Istanbul Region coals usually cause air pollution when burned in simple combustion systems. Combustion of such coals must be done in specially designed systems in order to minimize air pollution or else lignite should be converted to a semicoke product by low-temperature coking. The optimum conditions established in this study is only valid for fixed bed coking furnaces. The mode of coking and the dimensions of the furnace certainly affect the time and temperature of the carbonization process. An important consideration from the viewpoint of air pollution is the removal of tar in
15.01 15.99 15.37
69.04 6459 66.33 6324 65.04 6221
11.28 40.91 90.55
52.41 65.41 90.63
320 300 290
lignite. The experiments showed that if semicoking is continued until 15% volatile matter remains, the tar in Istanbul Region lignites is fully removed. Another important advantage of carbonization is that in addition to upgrading of coal in terms of moisture and volatile matter, a significant portion of the combustible sulphur is also removed. Under optimum conditions 63% of desulphurization is achieved. The released gases can be purified for further use as a utility gas or portion of it can be recycled as an energy source for the process itself. A disadvantage of the low-temperature carbonization process is the fragile nature of the semicoke product. It is possible to use the plus 10 mm fraction for household heating while the finer fractions can be utilized in cement and ceramic industries. However, if feasible, the fine product can be briquetted both to improve its strength as well as its utilization. Conclusions Low-temperature carbonization studies conducted on the desulphurization of the Istanbul region lignite sample can be summarized as follows. 1. The Istanbul region lignite is amenable to low-temperature carbonization tests. Tests conducted as a function of temperature and heating time revealed the optimum conditions to be 650~ and 50 min of heating time. 2. A semicoke product containing 15.22% volatile matter with 6403kcalkg -1 has been obtained. 3. The sulphur removal increases with increasing the temperature and reaches a value of 63% at 650~ 4. Coal undergoes fragmentation upon carbonization process due to the removal of moisture and volatile matter from coal. This is strongly dependent on the particle size. As the particle size increases the tendency for particles to fragment increases and this in turn reduces the strength of coal. The strength can be improved by briquetting of coal.
D E S U L P H U R I Z A T I O N OF L O W - R A N K COALS
References ASMATULU, R., ACARKAN, N., ONAL, G. & CELIK, M. S. 1995. Upgrading of low-rank coals by lowtemperature carbonization. Proceedings of Technologies for Mineral Processing, Baia Mare, Romania, 29-35. CELIK, M. S. • SOMASUNDARAN,P. 1994. Desulfurization of coal. In Kural, O. ed., Coal: Resources, Properties, Utilization and Pollution, Kurtis Press, Istanbul, 253-269. COLE, E. L. & NESS, H. V. 1977. Treatment of Solid Fuels. U.S. Patent 4,018,571, April 19, 1977 and U.S. Patent 4,052,169, October 4, 1977 (Texaco). EKINCI, E. 1982. Production Methods of Metallurgical Coke and Smokeless Fuel and its Application to Turkish Coals (In Turkish). Proceedings of International Coal Technology Seminar, 127-137. EVANS, D. G. & SIEMAN, S. R. 1979. Separation of Water from Solid Organic Materials. U.S. Patent 3,552,031, January 5, 1979. FLEISSNER, H. 1927-1928. Drying of Coal. U.S. Patent 1,632,829,1927 and U.S. Patent 1,679,078, July 31, 1928. GAZANEER, S. 1983. Smokeless Fuel Experience of Seyitomer Lignites. International Coal Utilization Conference, Sept. 6-10, Istanbul.
369
GENTILE, R. H. 1995. Clean Fuel Technology: The Contribution of K-Fuel. Proceedings of 3rd Coal Technology and Utilization Seminar, Cayirhan, Turkey. KOPPELMAN, E. 1977. Process for Upgrading Lignitetype Coal as a Fuel. U.S. Patent 4.052,168, October 4, 1977. LOWRY, H. H. 1945. Low-Temperature Carbonization Chemistry of Coal Utilization. Wiley, New York. ONAL, G., MUSTAFAEV,I., ASMATULU,R., YILDIRIM,I., ACARKAN,N. 8s CELIK, M. S. 1995. Desulphurization of Turkish Lignites by Low temperature Coking, ECOS'95, July 11-14, Istanbul, 640-645. SCIAZKO, M., KUBICA, C. & RZEPA, S. 1993. The Smokeless Fuel-Properties and Testing Methodology. Fuel Processing Technology, 36, 123-128. VERSCHUUR, E. et al. 1976. Thermal Dewatering of Brown Coal, U.S. Patent 3,992,784, November 23, 1976 (Shell). VON, H. H. 1982. Thermich Veredelung Der Kohel mit Ausnahme von Kokereien Technische Mitteilunge 75. Jahrgang, Heft 2/3, p. 117-128. WASH, E. J. 1977. Upgrading Subbituminous Western Coal. Canadian Patent 1,020,477, November 8, 1977. ZIELINSKI, H., KACZMARZYK, G., SCIAZKO, M. & SECULA, M. 1992.2nd Int. Cokemaking Congress, London, 28-30 Sept. 1992, Inst. Mater., London, 551-554.
Amelioration of high organic sulphur coal for combustion in domestic stoves M I C H A E L K. G. W H A T E L E Y 1, Z A F E R G E N C E R 2 & E R T E M T U N C A L I 2
1Geology Department, University of Leicester, Leicester LE1 7RH, UK 2 Directorate of Mineral Research and Exploration (MTA), Ankara, Turkey Abstract" Despite the proximity to Ankara, raw coal mined in the Beypazari basin cannot be used as a domestic fuel in Ankara because the high combustible sulphur content would add to the already severe pollution problems in that city. This study investigated a method of reducing the SO2 emissions by adding lime (CaO) to the coal prior to combustion. Sorbent, as lime, was added to the combustion chamber (a domestic stove), by pretreating the coal by mixing crushed coal with lime and molasses and turning the coal into briquettes. Lime was added to lump coal for comparative purposes. The ratio of Cafree : S was varied for a series of combustion tests, and the heating efficiency and the amount of sulphur fixed in the ash were determined. The results of the tests show that optimum sulphur retention and heating efficiency were obtained when the Cafree :S molar ratio was between 0.95 and 1.15 for the briquettes and between 1.00 and 1.25 for the lump coal. For lump coal, 50% of the total sulphur could be fixed in the ash and at least 75% of the heating efficiency retained. During combustion of the briquettes, at least 57.5% of the sulphur could be fixed in the ash and at least 85% of the heating efficiency could be achieved. This suggests that the addition of lime to briquettes may be a feasible way of reducing the SO2 emissions for domestic stoves.
Ankara, the capital of Turkey, has a population of around 3.25 • l06 people, most of whom rely upon coal as a domestic fuel supply for cooking and heating. The climate is such that in winter temperature inversion on the high level plateau (elevation 1500m) often traps the sulphurous
and nitrous emissions from these stoves creating an e n v i r o n m e n t a l hazard ( D u r m a z et al. 1993). As m u c h as 0.8 M t of coal is b r o u g h t into A n k a r a from s u r r o u n d i n g areas a n d from outside T u r k e y to supply the domestic market. As far as is possible, low sulphur coals are used, but
{ ~
,t:t Jit:t:t:t t~~:t:[:t:l%, ::
:
v
v
-
Neogene
v v
v
v
Palaeocene{ ~
I
~
,
i :
9
V
9
V
Late Jurassic F ret : Taceous :TTIc
l ~ "~ "~ ~ " ~ ' ' "" "" "~ \ "~)liKg~ unfa~'l~l-] % "" ,~, "~
0
(~aylrhan lignite field// +
~ + \
PalaeozoicI ~
I++
Miocene sedimentaryunits Teke volcanics Kizil~zayGroup
Ophiolites Nardin Formation (flysch) So'uk~ am
limestone Granite
[ [-~'-] Mh;at;:~ics
lOkm
Fig. 1. Location of the Beypazari basin showing the position of the (~ayirhan lignite field and the main rock units in the region (from Whateley & Tuncali 1995b, modified after Yagmurlu et al. 1988). From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 371-377.
372
M. K. G. WHATELEY E T A L . desulphurization (FGD) plant, which effectively removes 99% of the SO2 from the flue-gas. FGD is economic at this scale, but would probably be inefficient and not cost effective on domestic stoves. This study was designed to determine whether the Beypazari lignite could be beneficiated with the addition of lime to the coal in order to reduce the SO2 emissions from domestic stoves to acceptable levels.
most of Turkey's cheaply mined lignites have a high sulphur content. To alleviate this problem, low-sulphur coal is imported into Turkey from USA, South Africa, Italy, Colombia and Australia, some of which is transported to Ankara. Importation and transportation costs result in raised prices of domestic fuel. One potential source of lignite for the Ankara market lies in the Beypazari basin (Fig. 1), only 100 km NW of Ankara. Coal, of Miocene age, is known in two areas of the basin, namely the ~ayirhan and the Koyunagili lignite fields (Fig. 1). Coal from the (~ayirhan field only was used in this study. The (~ayirhan field contains some 400 Mt of coal, which is currently mined to supply a 300 MW thermal power station (TPS). Unfortunately, this lignite is characterized by high sulphur content (maximum 8.2% on an air dried basis), up to 75% of which has been shown to be combustible sulphur (Whateley & Tuncali 1995a). The TPS is fitted with a flue-gas
Lignite characteristicas There are two separate lignite seams in the (~ayirhan lignite field (Fig. 1). The lower lignite seam was deposited in the lower part of the ~oraklar Formation (Fig. 2). It has variable thickness, is laterally impersistent and is of very poor quality. It was not used in this study. The thicker, economically important,
LITHOLOGY
FORMATIONS
AGE J
__
Z
__
~
- -
- -
claystone,
9
Kirmk-rgormaation~
mudstone & gypsum limestone conglomerate, sandstone
claystone, mudstone, fine-grained sandstone
-
Upper Miocene .......
=i i --
_
HirkaFormation
'
~
....
.... _.,--/
Upper ligniteseam . . . . (~oraklar Formation
~'~
--b
~
~ Vv "E
~(~
9
0
o Lowerlignites e a m ~ "~
PreNeogene f ~ ~
"~
+
+
K-~-7~/
~
+ ~
> v
+\ ,
%
%
~
'-,..,,
"'-' ".-, ",-, ,-,,_,
Basementrock )
~
silicified claystone & limestone, chert shale, bituminous shale, trona & tuff cross-bedded conglomerate, sandstone & mudstone Metamorphics, ophiolites, granites, limestones & clastic sediments
Fig. 2. Schematic stratigraphic section of the ~ayirhan basin (from Whateley & Tuncali 1995b, modified after Yagmurlu et al. 1988 and Inci 1991).
AMELIORATION OF HIGH ORGANIC SULPHUR COAL upper lignite seam was deposited at the top of the formation. It is laterally persistent and has a reasonably uniform thickness of about 3.0 m. A one metre thick, tuffaceous, siltstone parting, with cherty nodules, splits the upper seam into two lignite beds, referred to as the first (Tv) and second (Tb) seams (Whateley & Tuncali 1995a) Detailed descriptions of these two seams are given by Whateley & Tuncali (1995a), who examined proximate and ultimate analyses, calorific value, combustible sulphur and pyritic sulphur contents, palynological descriptions, petrographic analysis, reflectance values and ash oxide analyses derived from coal samples collected from boreholes and underground mine faces. Reflectance measurements (Rmax [%]) varied from 0.34 to 0.38, putting the coal into the lignite rank category. The first and second seams are mined and mixed together for combustion in the TPS and Table 1 gives the average quality of both seams.
373
Mineral matter chacteristics The inorganic fraction of the raw coal, the mineral matter, was examined using XRD and SEM EDX analytical techniques. The major minerals present are the zeolites, analcime and clinoptilolite, and pyrite, with variable amounts of minor minerals such as gypsum, albite/ anorthite, marcasite, quartz, illite, dolomite and apatite. Vertical variation between the top and bottom seams is recognized because of the almost exclusive presence of clinoptilolite in the first seam, with minor amounts of analcime and almost exclusively analcime in the second seam, with minor amounts of clinoptilolite. Most Turkish lignites contain quartz, kaolinite, illite, smectite and pyrite as the main minerals with varying minor amounts minerals which include calcite, dolomite, hydromuscovite, feldspar, barite, celestite and marcasite. No other Turkish lignite is known to contain the abundance of zeolites found in this coal.
Table 1. Average values of the proximate, ultimate, sulphur, calorific value and ash oxide analyses from the upper seam in the Beypazari basin, Turkey
As received Air dried results results Proximate analyses Moisture content (%) Ash content (%) Volatile matter (%) Fixed carbon (%)
23.38 23.27 28.74 24.61
9.00 35.88 29.12 26.00
2.52 0.84 3.38
3.81 0.77 4.58
Ultimate analyses Carbon (%) Hydrogen (%) Nitrogen (%) Oxygen (%) Total sulphur (%)
31.12 2.39 7.32 55.73 3.44
40.37 2.98 9.12 43.26 4.27
Calorific value Calorific value (kcal kg- 1)
2318
3391
Grindability index
64.21
Ash melting point (~
1250
Ash analyses SiO2 (%) A1203 + TiO2 (%) Fe203 (%) CaO (%) MgO (%) Na20 (%) K20 (%) SO3 (%)
41.99 16.32 11.96 8.71 4.19 4.95 1.17 9.30
Sulphur analyses Combustible sulphur (%) Sulphur in ash (%) Total sulphur (%)
Ash characteristics When coal is burnt the inorganic residue is referred to as the ash and is measured as a percentage of the original mass of the raw coal, either on an as received or air dried basis. The as received ash content changes vertically both in and between the first (Tv) and second (Tb) seams at Beypazari (Whateley & Tuncali 1995b). The average ash content of the second seam is 27.4%, compared to the first seam which averages only 19.2%. The ashes were analysed by ICP-AES after sample digestion. The oxide analyses were tabulated and correlation coefficients presented by Whateley & Tuncali (1995a, Table 5). They found that there was a strong positive correlations between CaO and SO3, of 0.84 and 0.87 in the first and second seams respectively. This was due to the anhydrite in the ash which developed during the combustion of the coal. The pyritic and organic sulphur forms were oxidized and the the SO2 combined with Ca found mainly in the clinoptilolite, but also in the organic matter, to form the anhydrite. Whateley & Tuncali (1995a) showed that an average of 28% of the total sulphur in the raw coal reported in the ash in the first seam and 25% of the sulphur reported in the ash of the second seam. The remainder of the sulphur was reported as combustible sulphur, and was lost to the atmosphere as SO2. Recent work has shown that the Beypazari coal consists of the following forms of sulphur
374
M. K. G. WHATELEY E T AL.
(British Standard 1977), namely sulphate sulphur, 0.23%, pyritic sulphur, 2.34%, and organic sulphur 2.03%. The organic sulphur is contributing some 45% towards the total sulphur. As more than 75% of the sulphur is lost as combustible sulphur, this suggests that the pyritic sulphur is contributing to the combustible sulphur as well. The ASTM and British Standard tests for measuring the ash content expect the coal to be ashed at between 750 and 850~ At these temperatures pyritic sulphur as well as the organic sulphur will oxidize, resulting in the release of SO2. It is apparent that the high sulphur coals in Turkey that are used as domestic fuels, contribute to the SO2 emissions that are a major factor in urban air pollution (Durmaz et al. 1993). This project intended to establish whether a practical solution could be found whereby the addition of Ca to the coal in forms other than that found in zeolites and organic matter would lead to a lowering of the SO2 emission from the Beypazari coals.
Results of sorbent addition
and molasses. The Cafree:S molar ratio was varied for a series of combustion tests (Table 2). The free Ca contained in the coal was not included in these ratios. A fixed mass of coal (7kg) was burnt in each experiment and the central grate and the flue gas temperatures were monitored continually (Fig. 3) and graphs of these variations with combustion time were drawn (Figs 5 and 6). For clarity only the raw coal (L1 and B1) and the penultimate experiment (L5 and B5) are shown. The raw coal, both as lump coal and as briquettes, burnt fiercely and within 25 minutes produced central grate temperatures as high as 1300~ The lump coal produced the highest central grate temperatures, but the stove lost heat rapidly. The briquettes retained their heat longer (Fig. 4). In general there was not a great difference in the combustion time between the lump coal and the briquettes in each experiment. As the Cafree :S molar ratio increased, so the time taken for both the lump coal and briquettes to reach their maximum temperature increased, e.g. 90 and 70minutes respectively in experiment 5 (Fig. 4). The combustion time increased in the stove as the Carree:S ratios increased,
One of the most freely available forms of Ca is in lime (CaO). Experimentation has shown (Gencer 1988) that in a domestic stove the coal burns at an average temperature of about 780~ although the maximum may reach as high as 1300~ At these temperatures, sulphur is released from oxidized pyrite and organic material. The CaO reacts with the pyritic S during combustion in the following way, 2FeS2 + 4CaO + 702 ~ 4CASO4 + 2FeO resulting in capture of some of the sulphur as anhydrite in the ash. Sorbent, as lime, was added to the combustion chamber (a domestic stove), firstly in briquettes and secondly with the lump coal. The briquettes were produced by mixing crushed coal with lime
Table 2. Cafree'S molar ratio used in the combustion tests when lime was added to lump coal and briquettes of Beypazari coal, Turkey
Briquettes
Lump coal
Cafree:S molar ratio
B1 B2 B3 B4 B5 B6
L1 L2 L3 L4 L5 L6
0 (no lime added) 0.50 0.75 1.00 1.50 2.00
Fig. 3. Photograph showing the domestic stove used in the experiment.
AMELIORATION OF HIGH ORGANIC SULPHUR COAL
i
1400
350
1200
300
1000
o~
250
800
~
200
600
~ E
150
375
"
'
Ls B5
E
L5
400t/t
~ 100
200
50
0
, 0
25
',"
,
, L1,
,
0
, 0
50 75 100 125 150 175 Time (minutes)
,
25
,
,
,
50 75 100 125 150 175 Time (minutes)
Fig. 4. Variations in the central grate temperature with combustion time in a domestic stove burning coal from the Beypazari basin, Turkey. The graph shows raw coal (L1 and BI) and coal mixed at a Cafree: S molar ratio of 1.5 : 1.
Fig. 5. Variations in the flue gas temperature with combustion time in a domestic stove burning coal from the Beypazari basin, Turkey. The graph shows raw coal (L1 and B 1) and coal mixed at a Cafree: S molar ratio of 1.5:1.
e.g. up to 175 minutes in experiment 6 (Table 3). The average central grate temperatures ranged from 924 to 506~ for the lump coal and between 856 and 702~ for the briquettes. There is no clear pattern to be seen in the grate temperatures to compare the different fuel types, except that in general, the average central grate temperatures decreased as the Cafree:S ratios increased (Table 3). The flue gas temperatures increased rapidly to maxima of around 310~ when raw coal was burnt both as lump coal and as briquettes (Fig. 5).
The flue gas temperatures decreased rapidly past the maxima. As the Cafree:S molar ratio increased, so the time taken for the lump coal to reach its maximum temperature increased, e.g. to 125 minutes in experiment 5 (Fig. 5). The flue gas temperatures were raised rapidly when the briquettes were burnt in experiment 5 (Cafree :S ratio of 1.5:1), but the temperature dropped more slowly past the maximum. In general, there was very little change in the average flue gas temperatures as the Cafree:S ratio increased (Table 3). The average flue gas temperatures
Table 3. Results of combustion tests in a domestic stove for a series of Beypazari lignite (L) and briquette (B) samples burnt with various amounts of lime (see Table 2) Average temps Heat losses (%) Sample Combustion time Stove No. (mins) grate
Flue gases
Heat loss Heat loss Heat loss Thermal Sulphur by flue due to in grate efficienty fixation gas CO in (%) (%) flue gas
L1 L2 L3 L4 L5 L6
105 120 130 150 170 175
790 894 798 924 793 506
200 215 215 210 215 214
21.76 27.00 30.29 30.30 35.25 38.00
7.69 10.25 4.27 7.12 3.42 11.00
4.94 4.87 6.94 7.82 21.00 12.00
65.6 57.9 58.5 54.7 40.3 39.0
20.7 36.5 40.8 50.3 59.1 60.6
B1 B2 B3 B4 B5 B6
145 135 140 150 160 175
856 827 833 724 736 702
205 217 224 220 225 227
25.50 26.90 29.80 30.50 33.60 36.50
5.78 7.35 5.30 2.60 6.00 11.30
0.63 2.90 4.20 6.60 3.20 5.60
69.1 62.8 60.7 60.3 57.2 46.6
20.9 43.7 52.5 53.6 63.7 69.3
376
M. K. G. WHATELEY E T AL. 70
70
./,,~~,,~
y=22.699x + 25.078
60 r = 0 . 9 4 7 8 ~ _ _
-
50-
~ 55" 50- y=-14.442x + 66.507 " ~ . I--'c:45" ~ r-'-0"9843 ~ ~ =
40
~ '
r=0.9629
_
~ 30" -3 f~ 20-
[] Lump coal samples 9 Briquette samples
03
10
35 3O
y=20.44 + 2s.078 40-
0 0.25 015 0.:75 i
1.25 1~5 1.75 2
Cafree :total S molar ratio
o o.;,s ols o. ,s
i
1.i,s lls 1.?s 2
Cafree:total S molar ratio
Fig. 6. Variations in the thermal efficiency of a domestic stove burning coal from the Beypazari basin, Turkey, with respect to the Carree : S molar ratios of the various feeds.
Fig. 7. Variations in the sulphur retention in the ash plotted as a percentage of the total sulphur content of the raw coal from the Beypazari basin, Turkey, with respect to the Carree : S molar ratios of the various feeds.
measured when the briquettes were burnt were, on average, some 5 to 10~ higher than those measured when lump coal was burnt. By measuring the temperatures in the grate and flue gases it was possible to calculate the heating efficiency of the central stove during each experiment (Table 3 and Fig. 6). With the addition of lime to the stove fuel, the heating efficiency dropped markedly at the low Cafree : S molar ratio of 0.5:1 (Fig. 6). The lump coal dropped from 66% efficiency with raw coal to 58% and briquettes dropped from 69% to 63% efficiency. Thereafter the rate of change slowed between ratios 0.5 : 1 and 1:1. At these ratios, the variation in efficiency was between 58 and 55% for lump coal and between 63 and 60% for the briquettes. As more lime was added with the lump coal (1.5:1 ratio) the heating efficiency dropped rapidly to 40%. A similar effect was seen with the briquettes, although the drop in heating efficiency was less noticeable at the 1.5:1 ratio (57%). The final heating efficiency recorded was 47%. The ash was analysed for total sulphur content and this was reported as a percentage of the total sulphur in the raw coal (Table 3 and Fig. 7). This figure reflects the amount of sulphur retained in the ash. Graphs of sulphur retention plotted against Carr~ :S molar ratios (Fig. 7) show a rapid increase in the amount of sulphur retained in the ash immediately lime was
added both to the lump coal (from 21% to 37%) and briquette fuel (from 21% to 44%). In both fuel types the rate of sulphur retention in the ash slows beyond the 1.5:1 Carrel: S ratio. In all cases the amount of sulphur retained in the ash was greater when the lime was mixed with the coal as briquettes.
Discussionof results The time taken to reach the maximum temperature in the grate and in the flue gases increased as the Carree :S molar ratio increased (Figs 4 & 5). The maximum temperature reached is lower as the ratio increases. The high sulphur content of the coal meant that lime had to be added in large proportions. This resulted in reduced operating temperatures (Fig. 4) and a reduced heating efficiency (Fig. 6). There is an immediate loss of heating efficiency as soon as lime is added to the coal, but between the 0.5:1 and 1 : 1 ratio of Cafree : S there is no significant change in loss of heating efficiency. This suggests that there is no additional benefit to be gained by increasing the ratio of Carrie : S to more than 1:1 as far as heating efficiency is concerned. Sulphur retention increased from a base level of 21% when no lime was added to over 60% when the Carrer : S molar ratio was 2:1 (Fig. 7). The retention of sulphur in the ash when raw
AMELIORATION OF HIGH ORGANIC SULPHUR COAL coal is burnt and no lime is added is a function of the presence of the clinoptilolite/heulandite zeolites. The effect of the addition of lime was an immediate increase in the amount of sulphur retained in the ash. The rate of retention remained high between a 1: 1 and 1.5 : 1 ratio, thereafter the retention rate slowed. This suggests that there is no particular gain in sulphur retention beyond the 1.5 : 1 ratio of Carree : S. From Figs 6 and 7 it can be concluded that the optimum sulphur retention and heating efficiency were obtained when the Cafree: S molar ratio was between 1.00 and 1.25 for the lump coal and between 0.95 and 1.15 for the briquettes. At these ratios for lump coal, 50% of the total sulphur could be retained in the ash and at least 75% of the heating efficiency achieved. During combustion of the briquettes, at least 57.5% of the sulphur could be retained in the ash and at least 85% of the heating efficiency could be achieved. The added advantage of the reduced operating temperatures of the stove once lime was added with the coal is that the anhydrite in the ash would be more likely to be retained in the ash and not broken down. Chinch6n et al. (1991) have shown that anhydrite is stable up to temperatures of 1060• 10~ beyond which anhydrite decomposes into CaO and SO2. This study showed that significant reduction of the SO2 emissions could be obtained with the addition of lime to the raw coal which, during combustion, would convert some of the SO2 derived from oxidation of pyritic and organic sulphur in the coal into anhydrite in the ash. At the optimum Cafree : S molar ratios there was no significant reduction of the heating efficiency of the coal. The briquettes proved to be slightly more thermally efficient and retained a greater proportion of the sulphur in the ash than the lump coal and lime mixture. This suggests that by pretreating the Beypazari coal by crushing and mixing the coal with lime and molasses to form briquettes for use as a domestic fuel, the SO2 emissions can be reduced without significantly reducing the heating efficiency of the fuel.
377
The authors wish to acknowledge their colleagues and the Director at MTA, Ankara and their colleagues in the Geology Department, University of Leicester who provided technical and scientific support during the project. They would also like to thank the Director of TKI, Ankara who gave permission for access to the coal site and the management and staff at the (~ayirhan mine for their assistance during the collection of the samples. Considerable help was given by Professor A. Spears and Dr X. Querol, who read and commented on earlier versions of this manuscript. Their help is very much appreciated.
References BRITISH STANDARD1997. Analysis and testing of coal and coke, Part I1, Forms of sulphur in coal, BS1016, part 11. CHINCHON, J. S., QUEROL, X., FERNANDEZ-TURIEL. J. L. & LOPEZ-SOLER, A. 1991. Environmental impact of mineral transformations undergone during coal combustion. Environmental Geology and Water Science, 18, 11-15. DURMAZ, A., DOGU, G., ERCAN, Y. & SIVRIOGLU,M. 1993. Investigation of the Causes of Air Pollution in Ankara and Measures for its Reduction. NATO Science for Stability Program. GENCER, Z. 1988. An Investigation of Methods for Fixation with Lime of Sulfur Dioxide Formed by the Combustion of Ankara-Beypazari Lignites. MSc thesis, Gazi University. INCI, U. 1991. Miocene alluvial fan-alkaline playa lignite-trona bearing deposits from an inverted basin in Anatolia: sedimentology and tectonic controls on deposition. Sedimentary Geology, 71, 73-97. WHATELEY, M. K. G. & TUNCALI, E. 1995a. The origin and distribution of sulphur in the Neogene Beypazari lignite basin, Central Anatolia, Turkey. In: WHATELEY,M. K. G. & SPEARS, A. (eds) European Coal Geology. Geological Society, London, Special Publication, 82, 307-323. & - - 1 9 9 5 b . Quality variations in the highsulphur lignite of the Neogene Beypazari Basin, Central Anatolia, Turkey. International Journal of Coal Geology, 27, 131-151. YAGMURLU, F., HELVACI, C. & INCI, U. 1988. Depositional setting and geometric structure of the Beypazari lignite deposits, Central Anatolia. International Journal of Coal Geology, 10, 337-360.
The use of pulverized lignite/natural gas mixed fuels in the high-temperature process of a cement rotary kiln M. bIANOJEVIC', (5. JANKES', M. K U B U R O V I C ' , M . S T A N O J E V I C 2 & P. B L A G O J E V I ( ~ 2
1Faculty of Mechanical Engineering, University o f Belgrade, 27 marta 80, 11000 Belgrade, Yugoslavia 2 Beodin Cement Factory. Beodin, Yugoslavia Abstract: This paper presents the results of industrial tests of low grade lignite combustion in the 500t/day wet-process rotary cement kiln. For the use of pulverized lignite, the following process characteristics are determined: combustion process parameters, flue gas properties, and the influence of coal ash properties on clinker quality. Industrial tests have shown that it is possible to substitute 50-80% of the natural gas with pulverized low-grade lignite, while the output of the kiln, specific heat consumption, and the quality of cement clinker remain unchanged.
The cement industry faced an increase in the prices of liquid and gaseous fuels used for clinker production after the energy crisis of 1973. This was the reason for the increasing introduction of coal as a fuel, which is the dominant fuel in the West European cement industry today. Anthracites and other highquality coals are commonly in use. The introduction of precalcination in new cement production technologies has enabled the use of low-grade fuels, such as soft brown coal. Soft brown coals (lignites) are the main energy resource in Yugoslavia, and it was of great importance to determine the conditions for their use in cement rotary kilns. The investigation focused on determining all the necessary process parameters and technical limitations for the use of 'Kolubara' lignite. This paper presents the experimental results of mixed lignite/natural gas fuel combustion for cement production, in a wet-process rotary kiln (in the 'Beo6in' cement factory).
Process and experimental details
The coal usually used in the cement industry should comply with the following requirements (Duda 1976): LHV: rain. 21 (MJ/kg) Ash content: 12-15 (wt%) Volatile matter: 18-22 (wt%) Moisture content: up to 12 (wt%) as delivered. The high moisture content of Yugoslav lignites (up to 50%) requires two-step drying. The first
step is performed in drying plants (near the open coal fields) and the second in the factory's drying-grinding plant. The coal ash remains as part of the clinker material, thereby reducing the amount of raw material. The chemical composition and ash content of coal determine the necessary corrections in the raw material used for the required cement clinker quality to be achieved.
Table 1. Production process parameters for cement
rotary kiln in Beodin cementfactory Nominal kiln output Range of output Specific kiln output Kiln dimensions: length (L) shell diameter length-to-diameter ratio (L/D) Inner volume Inclination Speed of the kiln Internal heat exchangers (chain curtain): length surface Clinker grate cooler (Follax) Fuel Specific heat consumption Secondary air temperature Flue gas temperature Outlet clinker temperature Flue gas dust content (% of raw material consumption)
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 379-383.
t/day t/day kg/m 3h
500 450-520 15-17.2
m m
135 3.6-4.0 35
-
m3 % rpm m m2 m
1260 4 max. 1.2
20 1400 2.6 x 14.9 natural gas/ pulverized coal kJ/kg el. 6000-6700 ~ 550-600 ~ 150 ~ 50 % 1.5
380
M. STANOJEVI(~
Experiments were performed in the wetprocess rotary kiln in the 'Beo6in' cement factory. Technical data and process parameters of this plant are shown in Table 1. The scheme of the clinker production plant used for industrial tests is shown in Fig. 1. The plant consists of the following parts: wet-process rotary kiln with combined fuel burner (for pulverized coal and natural gas), a clinker grate cooler, and a tube ball mill with a direct coal-drying system. The plant was originally designed for use with high quality coal. The material balance scheme for this kiln is shown in Fig. 2. The mathematical model of the
material balance includes the following process parameters: consumption of natural gas, pulverized lignite, and raw material slurry; natural gas composition; flue gas composition; raw material, dust, and cement clinker composition; original and pulverized coal composition; dust concentration in flue gas; flow of hot air for the tube ball mill; primary air flow; moisture flow separated from coal. Four industrial tests were carried out. The aim was to determine the influence of the coal/ natural gas ratio on the clinker production process. The quality and grain size of 'Kolubara' dried lignite were different in each test. The
? 6
,-~
po21
to2ii~:o21k.~_
~01
l~
Fig. 1. The scheme of clinker production plant and the locations of measuring points. 1, rotary kiln; 2, grate cooler; 3, flue gas chamber; 4, flue gas fan; 5, chimney; 6, burner; 7, chain curtain
VVA'IERVAI=,~ 8EPARA'TEI:> I=ROMr I~.
DU6T IN FLLE~
A
MILL
l
~
ROTARY KILN
I r'~
CLINKER COOLER
~-
AIR Fig. 2. The material balance scheme of the wet process rotary kiln.
r
r
-
-
J
X=-
P U L V E R I Z E D L I G N I T E IN C E M E N T R O T A R Y K I L N m e a s u r e m e n t of characteristic parameters in each test was carried out in steady-state condition, which lasted a m i n i m u m of 6 to 8 hours. Seeing that the material remains in the kiln for a b o u t 2.4 hours the m e a s u r e m e n t periods were long e n o u g h to provide reliable data on the kiln operation.
Results The quality of the coal used in the experiment is presented in Table 2. F o r each test, the table gives data a b o u t the proximate and ultimate analyses of coal as it enters the mill and of pulverized coal as it enters the b u r n e r of the kiln.
The table shows that the moisture content of the original coal varied from 28.5 to 30.5 w t % while L H V varied between 13.7 and 15.7 MJ/kg. The ash c o n t e n t in tests a b, a n d c where coal with particle size of - 15 + 0 m m a n d - 15 + 5 m m was used, turned out to be 9.8 to 13.8wt%; in test d, using coal with a particle size o f - 5 + 0 m m the figure was 1 8 . 4 w t % . The moisture content o f the pulverized coal in tests a b, and c r a n g e d from 12.1 to 1 2 . 4 w t % . In test d the o u t p u t of the coal mill was reduced due to the use of lower-quality coal, which resulted in a moisture content of 9.8 w t % . The L H V of the pulverized coal was r o u g h l y the same in the samples p r o d u c e d by tests a and d i.e. 18.2MJ/ kg, for test b it was 19.3 MJ/kg, and for test c it had the highest value - 19.85 MJ/kg.
Table 2. Coal properties Industrial test in roatry kiln: Coal size distribution (tube ball mill inlet) Carbon Hydrogen Oxygen Nitrogen S. comb.
C H O N S
Moisture Ash S. total S. in ash S. comb. Coke C~ Volatile Combustible Low heat value Carbon Hydrogen Oxygen Nitrogen S. comb Moisture Ash S. total Coke Cfax Volatile Combustible Low heat value
C H O N S
a - 15 + 5 mm
b -15+0mm
c -15+5mm
d -5+0mm
wt% wt% wt% wt% wt%
Ultimate analysis of coal: 38.61 40.14 3.18 3.31 13.91 14.51 0.54 0.35 0.31 0.36
41.92 3.55 15.47 0.54 0.22
36.84 2.92 11.83 0.44 0.32
wt% wt% wt% wt% wt% wt% wt% wt% wt% kJ/kg
Proximate analysis of coal: 30.45 29.54 13.00 11.79 0.90 0.89 0.59 0.54 0.31 0.35 38.09 38.16 25.10 26.37 31.46 32.30 56.56 58.66 14 342 14954
28.50 9.80 0.73 0.51 0.22 37.01 27.20 34.49 61.70 15 717
29.27 18.38 0.96 0.64 0.32 44.14 25.77 26.59 52.36 13 698
wt% wt% wt% wt% wt% wt% wt% wt% wt% wt% wt% wt% kJ/kg
Ultimate analysis of pulverized coal: 48.64 50.10 51.41 4.01 4.13 4.35 17.52 18.11 18.97 0.68 0.44 0.67 0.39 0.43 0.27 Proximate analysis 12.39 16.37 1.13 47.98 31.62 39.63 71.24 18 171
381
of pulverized coal: 12.07 12.31 14.72 12.02 1.11 0.89 47.62 45.39 32.90 33.37 40.31 42.30 73.21 75.67 19284 19845
47.10 3.73 15.12 0.56 0.41 9.59 23.49 1.23 56.42 32.94 33.99 66.93 18209
382
M. S T A N O J E V I C
Table 3. Rotary kiln process parameters Industrial test in rotary kiln: Coal size distribution (tuble ball mill inlet)
a -15+5mm
b -15+0mm
c -15+5mm
d -5+0mm
Pulverized coal Low heat value
kJ/kg
18 171
19 284
19845
18209
Natural gas Low heat value (at ~
kJ/m 3
Raw material slurry Moisture content Specific consumption Consumption
% kg/kg cl. t/h
43.32 2.7850 58.82
44.62 2.8546 58.63
44.70 2.8678 57.36
44.00 2.8234 60.00
Dry raw material Ignition loss Specific consumption
% kg/kg cl.
34.01 1.5785
34.11 1.5809
34.32 1.5859
34.12 1.5811
Clinker Production
t/h
21.12
20.54
20.00
21.25
Fuel consumption Natural gas
35466
ma/h m 3/kg cl. kg/h kg/kg cl.
1741 0.0824 3390 0.1605
Specific energy consumption In natural gas In pulverized lignite Total
kg/kg cl. kg/kg cl. kg/kg cl.
2922 3004 5926
Natural gas/pulverized lignite ratio Natural gas Pulverized lignite
% %
49 51
Pulverized lignite
1323 0.0644 3918 0.1907
714 0.0357 4768 0.2384
1396 0.0657 4245 0.1998
2 284 3 678 5962
1266 4731 5997
2330 3637 5967
38 62
21 79
39 61
Table 4. Cement clinker properties Industrial test in rotary kiln: Natural gas/pulverized lignite ratio
referent 100/0
a 49/51
b 38/62
C
21/79
d 39/61
Mineralogical composition: 3CaO.SiO2 2CaO.SiO2 3CaO.A1203 4CaO.A12OyFezO3
wt% wt% wt% wt%
66.66 15.26 3.76 11.02
51.23 32.76 2.93 9.05
66.11 15.90 5.99 9.50
50.35 29.88 5.18 11.29
67.28 16.82 3.50 10.98
Oxides content: ignition loss SiO2 A1203 Fe203 CaO MgO free CaO
wt% wt% wt% wt% wt% wt% wt%
0.6 20.77 5.67 3.45 66.36 2.13 0.54
0.4 21.05 5.67 3.45 65.80 2.13 0.84
0.52 21.10 5.93 3.45 65.66 2.13 0.69
0.33 21.15 5.80 3.45 65.65 2.13 1.08
0.34 21.25 5.80 3.45 66.22 2.13 0.72
Characteristic modules: hydraulic module (HM) aluminate module (AM) silicate module (SM)
-
2.22 1.64 2.28
2.18 1.64 2.30
2.15 1.72 2.25
2.16 1.68 2.28
2.17 1.68 2.29
Bulk density
kg/m 3 1428
1383
1398
1388
1502
PULVERIZED LIGNITE IN CEMENT ROTARY KILN Rotary kiln process parameters determined in industrial tests are given in Table 3. It is shown that the energy from coal in total energy consumption for all industrial tests was 50 to 80%. The factors limiting the substitution of gas by coal were: a decrease in temperature in the kiln sintering zone, and the output of the coal mill. In order for the required clinker quality to be maintained, it is necessary to have a temperature of 1550~ to 1650~ When only natural gas was used, the mean temperature was around 1600~ dropping to 1540 to 1590~ in tests with coal, i.e. the maximum temperature drop was 60~ In tests a b, and d the degree of substitution of natural gas by pulverized coal was determined by this maximum drop in temperature. In test c the quality of pulverized coal was such that the degree of substitution reached (79%) was not accompanied by a drop in temperature below the above limits. However, a higher degree of substitution was impossible to attain due to limitations in the mill's output. Specific energy consumption and the rotary kiln output shown in Table 3 were roughly the same in all tests. They were similar to mean values attained when only natural gas is used. Table 4 shows clinker quality (as defined by mineralogical composition, bulk density, free CaO, characteristic modules) for all coal tests and the test which used only natural gas. All parameters indicate that clinker quality in coal test remained unchanged compared to the natural-gas test.
Conclusion The industrial tests of drying and grinding of lignite were performed in the factory's existing tube ball mill, and the pulverized lignite, together with natural gas, was used in the 500 t/day wet-process rotary kiln.
383
During the experiment the natural gas substitution ratio was 50-79% of total energy without affecting clinker production. The use of pulverized lignite did not affect specific energy consumption in the clinker production process. This could lead to the conclusion that the main process features, such combustion quality, heat transfer, and kiln output remained within the limits which did not influence the overall kiln production process. Higher substitution of natural gas with pulverized lignite was not possible during the industrial experiments. The main reason was the insufficient output of the tube mill and the pneumatic transport system which had been constructed for the use of high-quality coal. The results of the described experiment have shown that high-quality coal is not the only solid fuel utilisable in clinker production, and that there is a future for carefully pretreated lowgrade lignite as the main fuel for the Yugoslav cement industry. The same is possible for other coals of similar quality.
References DUDA, W. 1976. Cement Data Book. Bauverlag GmbH, Wiesbaden and Berlin, 279-283. PERKOVlC, B., STANOJEVIC, M., DOKI~, S. 1994. Industrial Tests of Substitutions of Anthracites with Dry Pulverized Lignite 'Kolubara' in Cement Factory Beo(in, Final Report. Mining Institute, Beograd, Faculty of Mechanical Engineering, Beograd. STANOJEVI(~, M. & KARAN, M. 1994. The Use of Yugoslav Solid Fuels (Lignites) in Rotary Kilns in Cement Industry. International Conference 'Energy for industry '94', Beograd, Proceedings, 218-223. , PETROV,A. & KUBUROVId:,M. 1993. Influence of lignite 'Kolubara' properties on the production of pulverized lignite. Termotehnika, Beograd, XIX, 1-2, 55-64. VULETII~,g. 8r STANOJEVIC,M. 1987. Possibilities and conditions for the use of lignite 'Kolubara'. Mining Journal, Beograd, 1, 64-67.
The possibility of underground gasification of Bulgarian Dobrudja's coal DOUCHKO
DOUCHANOV
& VENECIA
MINKOVA
Institute of Organic Chemistry, Bulgarian Academy of Sciences, Acad.G.Bonchev str., bl.9, 1113 Sofia, Bulgaria Abstract: The gasification of coal within underground coal seams and using the combustible
gas as a fuel is an idea that has attracted scientists for many years. Some success in gasifying thick coal seams near the surface has been demonstrated in recent experiments in the USA. Tests on underground coal gasification in Belgium and France have been carried out supported by the Energy Commission of the European Communities since 1978. The latest tests at Thullin, Belgium carried out by a joint Belgium and German team are thought to be promising. The depth of the seams has been selected as representative of Southern European and some Mid-European coals as an essential first attempt before moving to a further stage, at around 900-1000 metres depth. Work on this programme is at an early stage and its progress will be watched with much interest. For Bulgaria it is of vital importance to develop underground coal gasification on Dobrudja's coal (1.5 109 t ) - as an important energetic strategy and the possibility of environmental utilization of this resource.
The first tests on underground coal gasification (UCG) were performed in 1912. They were followed by a number of tests in the former USSR and France. However, these attempts on U C G were discontinued due to the low calorific value of the gas produced and the difficult control on the combustion processes and the chemical reactions in the underground gas generator. After the experiment carried out in Gorlovka (1932), the former USSR re-commenced the U C G tests. Following the completion of the test period, at the end of the 1950s, a series of U C G stations was built in: Tula (130kin south of Moscow), Yushno-Abinsk (Kuznets mine basin, Siberia), Shatsky (80 km southeast of Tula) and Angren (120kin southeast of Tashkent). These stations were designed to use lignite coals. The U C G installations at Lisichansk (Donets basin) and Kamenskaya (130 km from Rostov) operated in bituminous and anthracite coals. The gas station at Tula operated in lignite coals in horizontal seams between 0.3 and 5 m thick, lying at a depth of 50m. In 1958 a total of 400 x 106m 3 of low calorific gas (750-850 kcal/m 3) was produced. Its composition is as follows: CO, 5.5%; H2, 13.5%; CO2, 17%; CH4, 1.6%; CnHm, 0.2%; O2, 0.5%; HzS , 1.0%; N2, 60.7%. The seams developed in Yushno-Abinsc were 8-22 m thick at a depth of 250 m and with a dip of about 45-70 ~ (Antonova et al. 1967). In this installation a total of 100 x 106 m3/year gas at a calorific value of 1000 kcal/m 3 was produced and it is known that about 700 engineers and 3000 technicians were involved in its production in 1957. The gas produced is characterized by the
following composition: CO, 13.4%; He, 13%; CO2, 11.8%; CH4, 3.6%; CnHm, 0.1%; 02, 0.2%; H2S, 0.01%; N2, 57.9%. The U C G station at Shatsky, built in 1959 (lignite seams of 2.9 m thickness, lying at a depth of 40m), has produced 200 x 106 m 3 gas/year at a calorific value of 800 kcal/m 3. Initially this gas was used mainly for local plants after which it was employed to drive two 1 2 M W turboalternators in an electricity power station. The station at Angren is situated on lignite coal seams that are 20m thick, at a depth of l 1 0 - 1 5 0 m and dipping at 30 ~ The area of the coal deposit is 150 km 2 and the coal ash is 11%. The station has produced 600-800 x 106 m 3 gas at a calorific value of 700-800 kcal/m 3. The gas, whose composition was: CO, 5.6%; H2, 15.2%; C02, 19%; CH4, 2.5%; CnHm, 0.2%; O2, 0.5%; H2S, 0.4%; N2, 56.1.% (Lavrov et al. 1971) was used mainly for electricity generation in a 200 M W power-steam station. The station at Kamensky(Rostov) has operated in coal seams dipping at an angle of 40-50 ~ with a gas production rate of 2 x 106 m3/day at a calorific value 900kcal/m 3 and that at Lisichansk has operated in coal seams which are 0.5-1 m thick, with a dip of 30-40 ~ and a gas production rate of 100-120 x 106m3/year at a calorific value of 850 kcal/m 3 (Skafa 1960). The stations at Yushno-Abinsk and Angren were the only ones remaining in operation after 1980. Parallel to the experiments on U C G in the former USSR in 1945-1965, tests in USA, England, Belgium, Poland, Czechoslovakia (Prasek & Koranda 1989), in the Tatabanya
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geologyand Technology, Geological Society Special Publication No. 125, pp. 385-390.
386
D. DOUCHANOV & V. MINKOVA
mine in Hungary during 1979-1980 (Szechy et al. 1988), Italy, Japan and Morocco (Djerada, 1947-1950) were carried out. The tests performed in England (Newman Spinney 1949) and in USA-Gorgas, Alabama (1946-1958) gave the most promising results. In USA the development of underground coal gasification is considered as a prospective trend. Taking into account the fact that the coal reserves of USA amount to 87% of the mineral resources, the eventual realization of UCG could increase almost threefold the exploitation of the coal reserves. The oil crisis in the 1970s has initiated again experiments on UCG in USA. The Energy Center of USA undertook gasification experiments in Hanna, Wyoming to test the possibility of underground gasification of 10m thick subbituminous coal seams, lying at a depth of 120 m. The coal permeability in the seams was enhanced by hydrofracturing. In 1973 a total of 0.24 x 10 6 m 3 gas/day was produced using air as a gasification agent. A team at the Energy Center-Laramie has carried out UCG tests in Hanna for 55 days. The gas composition in vol% is as follows: H2, 17.3; CO, 14.7; C02, 12.4; CH4, 3.3; C2H4, 0.6; 02, 0.5; HzS, 0.1; N2, 51.0; At, 0.6. The Lawrence Livermore Laboratory at Wyoming has performed a number of tests on UCG in seams that are 15 m thick, at a depth of 150-900 m by applying a steam-oxygen mixture feeding. The gas production rate was 5 x 104 m3/day at a calorific value of 2350 kcal/m 3. The Morgantown Energy Technology Center (METC) has carried out underground gasification on 2m thick bituminous coals, at a d e p t of 275 m, in West Virginia (Strickland 1977). The gas production rate was 0.1 x 10 6 m 3 at a calorific value of 1100 kcal/m 3. It should be noted that this experiment has a dual importance. On the one hand, the reserves of bituminous coals amount to 817 gigatonnes and on the other, about 30% of these reserves are found at a great depth. In addition the majority of these reserves are highballast and their exploitation by deep mining would result in serious environmental problems (Clean Act, USA, 1970). These experiments on U C G in the United States should be supplemented by three additional tests. Firstly, he results obtained by the Los Alamos Scientific Laboratory (LASL) in the Fruitland mines (New Mexico) on the separate runs of the two processes accompanying UCG, i.e. pyrolysis and semicoke gasification in different areas of the mine, aimed at reaching an optimal control on the processes. Secondly, the UCG performed by the Basic Resources Inc. in Fairfield (Texas), following a Soviet pattern, in order to build two 20MW
steam-powered stations, and thirdly the UCG tests carried out in the Reno Junction, Wyoming by the Atlantic Richfield Company (ARCO).
The present UCG situation in Europe In Northern Belgium, between the Campine Basin and the border with Holland the hypothetical coal reserves amount to 15-20 gigatonnes which is equal to the total amount of coals produced in Belgium for the last 150 years. The German geologists consider that coal reserves in the Ruhr lying at a depth of 1200 m and those at a depth up to 5000m are 10 and 1000 gigatonnes, respectively. Holland possesses 1500 gigatonnes coal resources, with the majority of them lying at a depth of 1500-3000 m. On the basis of preliminary studies of the National Mine Institute (NMI)-Belgium and those of Professor Wenzel from AIX University in Germany a joint project between Belgium and Germany was agreed in 1986. In Thullin (Belgium) UCG tests were carried out at depths of 860 m. Air was used as a gasification agent in the underground gas generator and a low calorific generator gas was produced. The utilization of oxygen-steam mixture at a pressure of 10-20 MPa yielded a product with mean C H 4 content (32.6% without N2). This content is two times higher than that obtained in the pilot installation 'Ruhr-100' of the Lurgi Company, Germany. Thus after CO removal the calorific value of the gas has attained a value similar to that of natural gas. Figure 1 shows the principal scheme of operation of this installation. The air supplied by compressor 2 enters the basic seam layer through the pressure hole 3; the low calorific gas produced by UCG, is released through the outlet hole 4. The latter is cooled to protect it from excessive increases in temperature and a water steam is produced simultaneously as a result of the heat of the removed gas. The low calorific gas reaches the surface at high pressure with a temperature of about 250-300~ after which it is washed in scrubber 5 before its inlet into the combustion chamber of the steam generator 6. The gasification of the base coal layer yields a high calorific gas (grison) which is contained in the upper layers. In the programme described above the following partners participate: from BelgiumThe Geological Department at Brussels, the Department of Chemical Engineering at the University of Liege, the Mining Department at Mons Polytechnic, Distrigaz, Brussels; from
REVIEW OF HISTORY OF UNDERGROUND COAL GASIFICATION
Ai~
E
x h~
387
~,~////////////////" '/.--6 5; :-_ -
-
E O
,--
.z
"///t~//////-///z///~'~ O ,r-
i
g:
Fig. 1. Scheme of UCG carried out in Thullen, Belgium: 1, motor; 2, compressor; 3, injection hole; 4, gas removal hole; 5, scrubber for low calorific gas purification; 6, steam boiler with combustion chamber; 7, alternator; 8, turbine; 9, heat-exchanger; 10, condensation vessel; 11, pump; 12, grison outlet; 13, scrubber for high calorific gas purification: 14, coal seams. Germany - Rheinisch-Westfalische Technische Hochschule (AIX), Saarberg-Interplan (Saarbrucken), Bergbau-Forschung GmbH (EssenKray), Messerschmitt (Munchen). England and the former USSR were among the first to show great interest in development of UCG. Thus after a 30 years interruption the National Coal Board undertook in Newman Spinney (1949) the production of gas by UCG for electricity generation in a 100-200MW steam-power station. In France, a research group for underground coal gasification, i.e. Groupe d'Etude de la Gaz~ification Souterraine (GEGS) including Charbonnage de France, Gas de France and Institut Fran~ais du P~trole was founded in 1976. Its goal was the development of coal reserves amounting to 2 x 109 tonnes at a depth of more than 800m. In the 1980s about 22 million dollars were ensured for the realization of this project. For the first stage which was carried out in Bruay-en-Artois (Nord-Pas-deCalais) about 4.3 million dollars were spent (Ferreti 1982). The first stage of the test began from a derelict mine tunnel at a depth of 1000 m and with two boreholes (injection and gas
Fig. 2. Scheme of UCG carried out in Pas-de-Calais, France: 1, existing mine gallery at a depth of 1000m; 2, injection hole; 3, internal hole diameter (60 ram); 4, external hole diameter (110ram); 5, closing valve; 6, gas removal hole; 7, coal seam; 8, metal grid; 9, intrusion of water under 800 bars pressure. removal) at a depth of 170m, separated by 65m drilled in a 1.2m thick coal seam. The experiment proceeded in the following stages: first, water injection (hydrofracturing) at a pressure of 100-300 bars, followed by air feeding for expansion of the hole, in the third stage electrocombustion is carried out and in the last stage oxygen is supplied to produce a substitute of the natural gas (SNG) (Fig. 2). The very first objective of the Belgium and French projects was to produce a low calorific gas for electricity generation. Further it was aimed to produce a gas at a calorific value of 2600kcal/m 3 which after concentration could substitute for the imported natural gas (SNG) with a calorific value of 9500 kcal/m 3. In Bulgaria, since the discovery of the Dobrudja's coal basin, intensive geological investigations on the origins and petrographic composition of the coal have been carried out. Detailed study of the coalification of this coal has revealed the presence of a thick series of bituminous coals. Ultimate and proximate analyses of the coals show that according to the degree of coalification they are all volatile coals (V daf =35-40%; C d a f = 80-83% and swelling i n d e x - 1 ) . On average coals are characterized by a low sulphur content-0.6-1.5% and a low content of mineral matter-Ad = 6-12%. Individual seam samples have a higher content of mineral matter-A d = 13-31%, sulphur = 0.8-4.2% and volatile matter yield V daf = 3 7 - 4 5 % . The increased content of volatiles (up to 45%) and
388
D. DOUCHANOV & V. MINKOVA
2.1
l
~-a b " ~ d '
e
r
r
3
I Gas
Fig. 4. Countercurrent combustion in UCG: 1, injection hole; 2, gas removal hole; 3, compressor; 4, coal seam; 5, combustion flow direction.
Fig. 3. Electrolinking method: 1, electrodes; 2, injection and removal holes; 3, insulation: ab,de-highly conducting channel sections, c-unheated coal seam region. of cdaf-76-80% shows that some of the samples can be referred to high volatile bituminous coals (Minkova et al. 1983). In the 1970s it was considered that these volatile coals could be used in a mixture (up to 55%) for coking in the Kremikovtsi Metallurgical Plant (Trayanov, 1979). However, it turned out that the coal deposits are situated at a depth of 1500 m which requires large investments for mining. The gasification of coal seams at shallow and intermediate depths can be undertaken by directional drilling whereby the interaction between neighbouring holes (for injection and gas removal) is a result of the natural coal permeability. Often the latter is enhanced by: water feeding (hydrofracturing); air injection under pressure (electrolinking) (Fig. 3); blasting (Klimentov, 1964), or by using derelict mine tunnels from which holes of up to 200m are drilled, etc. The most complicated problems in this respect are: the realization of the linking along the coal seam whereby the countercurrent combustion ensuring sufficient gas flow between the holes is most often employed for its stabilization (Fig. 4); the pack compression of the rocks disturbing hermeticity facilitates the water penetration from the water-carrier layers; the control of the gasification front flow from the injection to the gas removal hole, etc. Some
environmental problems such as the air pollution caused by various chemical derivatives evolved during U C G should be also considered. In Table 1 are shown data from the studies on the pollution during U C G testing in Hoe Creek, Wyoming, USA (Mead et al. 1977). During U C G at greater depths, as will be required for the Bulgarian coal, similar pollution has not been registered: the clays, for example, take part in the neutralization of H2S and NH3. It should be noted that the coal seams in Europe are characterised by great depths and small thickness. Under these conditions the successful performance of U C G requires that a number of technical difficulties are overcome such as: the effective linking in the coal seam; the impeded control of the gasification front; and the control of the multistage gasification process as a whole. The reaction agents and the basic products used in UCG, as in all gasification processes of coal, besides coal are: oxygen, water, hydrogen, carbon oxides, methane, hydrogen sulphide, etc. (Douchanov & Angelova 1982). The composition of the gas produced is influenced by catalytic reactions which occur on the surface of the coal matrix and the inorganic salts which are found in abundance around the underground gas generator. The coal characteristics, the geometry of the coal seams (depth, dip, etc.), the amount, type and the quality of the gasification agent, its pressure and temperature, the geometry of the holes are some of the basic
Table 1. Study of the pollution by UCG during the Hoe Greek test Permissible concentration limit (mg/1)
Pollution Pregasification Inside burn zone species value (mg/l) Concentration increase (mg/1)
Outside burn zone Concentration (mg/1)
increase
Phenols CNNH~Pb 2+ SO~-
500 300 70 0.04 1000
5 • 105 0.001 3 K 10 4 0.20 100• 0.5 40 K 0.05 5K 250
0.001 0.01 0.5 0.001 200
0.1 0.4 20 0.001 2000
100K 40K 40x 10•
REVIEW OF HISTORY OF UNDERGROUND COAL GASIFICATION
389
to a 'new edition' of the interest towards this huge resource. This project includes the following participants: the Mine Geological Technological Institute in Madrid, the ENDESA and OCICARBON Companies, the English Nuclear Energy Board and the Belgium Council for UCG research. The main area of coal mined in Bulgaria belongs to the lignites from the M a r i t s a - East field. They are the main feed stock of the thermal power plants. It is envisaged that during the next 5-10 year the Elhovo lignite field (500 x 106 tonnes) will be set into operation (Douchanov & Angelova 1982). The characteristics of the coal from the aforementioned deposits are as follows: Marits-East (W, 5060%; A d, 30-60%; S d, 4.0-5.0%; C daf, 64%; H daf, 6.8%; Combustion heat of working fuel, 6.6 MJ/Kg) and Elhovo (W, 55-55%; A d, 30-37%; S d, 7.0-8.0%; C dar, 62%; Combustion heat of working fuel -6.3 MJ/Kg). During recent years, stand-tests were performed for coal gasification in a fluidized bed, at pressures of 1-2 MPa, both with steam-air and steam-oxygen mixes (Lazarov 1986). Besides the advantages, these experiments have certain shortcoming. For example, during gasification of high-ash coal, agglomeration of the particles and disturbance of the normal fluidized bed in the gas generator might occur. Such phenomena have been observed with steam-air gasification
parameters that determine the effectiveness of the UCG process. All these complex interactions indicate the necessity for various models and mathematical modelling (Gunn & Whitman 1976) to optimise conditions prior the realization of the underground experiments. The financial aspects of the problem are also of essential importance. The cost of a hole in the 1980s was two million French francs. The greater distance between the two holes (injection and gas removal) is decisive for the optimal amount of the burned coals utilized by them. This distance, imposed by the experiment, has a basic meaning for evaluating the economical efficiency of the process. French scientists (Pottier & Chaumet 1978) have derived the so called 'integral coefficient t(L)' (Fig. 5) which reveals the relationship between the following parameters: Q, E, L, S. As seen from Fig. 5, these are: the amount of the coal utilised by a pair of holes (injection and gas removal) Q = 6000 tons, coal seam thickness E of 2 m burned coal seam area S = 2 5 0 0 m 2 and the distance between the pair of holes L = 60 m the integral coefficient t(L), is 0.7. The fact that in 1991 (Furfari 1992) the Energy Council of the European Communities decided to invest 18 million ECU in a large-scale Spanish-English-Belgium test on UCG in the Alcorisa region of Spain with a depth of the holes of 600 and 900 m, points unambiguously
"d':O,~ ~-0,6 7"=0J3 "U=I,0
l
90
1/111I ~ t(L)
5
#' Ill / I n~3
1,0 OH 0,6 04
I
9 15 21
,'
7 9 11 13 15 f
"
,', '.
I
~
f
I
|
S ( 1000 m 7)
f
i
.-1
27 10~ 5 E t~E 3
E 2
O,[x1OOOf/ pair h o l e s )
Fig. 5. Curves for calculation of the integral coefficient t(L): Q, quantity of the coal matrix utilized by a pair holes (injection and gas removal); E, thickness of the coal seam; L, distance between the holes.
390
D. DOUCHANOV & V. MINKOVA
of the Elhovo coal in a quasi boiling layer, under pressure, at gasification temperatures of 920930~ (Lazarov & Douchanov 1988). Due to the lower reactionability, compared to the MaritsaEast coal the gasification of the Elhovo coal does not run at a satisfactory speed when the temperature is below 950~ To clarify the possibility of overcoming the particle agglomeration during gasification of Elhovo coal, the following studies were performed: (i) on the cause of particle agglomeration (Douchanov et al. 1993) and (ii) on the impact of some catalysts. The results from study (i) on the Elhovo coal during gasification in the fluidized bed with a steam-air blow, under pressure, confirm the necessity of running the process at a lower temperature and at conditions avoiding local overheating, i.e. at intensive quasi-boiling layer. The effect of some catalysts on increasing the speed of gasification of the Elhovo coal was tested at temperatures below 900~ The results obtained, indicated that the carbonates of the alkaline metals are active catalysts during gasification of the Elhovo coal at 750-800~ (Angelova & Douchanov 1983; Douchanov & Lutskanov, 1996). Since proposals for mining the Dobrudja's coal basin have been abandoned, underground coal gasification remains an important strategy for utilizing this resource. During the past decade and particularly in the last few years a revitalization in the interest in U C G has occurred both in the United States and in other countries. Considering the large reserves of the Dobrudja's coal basin amounting to about 1.5 x 109t and the insufficient resources of petrol and natural gas in Bulgaria, we conclude that ways should be investigated including cooperation with other countries with experience in this field, to carry out research, mathematical modelling and pilot experiments on U C G in the Dobrudja's basin.
DOUCHANOV, D. & ANGELOVA,G. 1982. Issledvane vazmozchnostite za intenzifikatsiya na gazifikatsiyata na vuglischtata, Izvestiya na BAN, 15, 393-399. DOUCHANOV,D., LUTSKANOV,L., MARINOV,S. P., MINKOVA, V. & YOSSIEOVA,M. 1997. Catalytic effect of ZnC12 in the pyrolysis of lignites, Fuel, in press. DOUCHANOV, D., MINKOVA, V., MARTINEZ-ALONSO, A., PALAClOS, J. M. & TASCON, J. 1993. Low temperature ashing of Bulgarian lignites, Erdol und Kohle, 12, 461-467. GUNN, R. & WHITMAN, D. 1976. Packed-bed models for the in situ gasifier, Laramie Energy Research Center, Reprint No LERC/RI-762. FERRETI, M. 1982. La valorisation du charbon, Paris, 219-237. FURFARI, S. 1992. Gasification and 1GCC within the European Communites, Erdol und Kohle, 45, 292-293. KLIMENTOV, P. P. 1964. Gidravlicheskii razriv dlya podzemnoi gazifikatsii zalezchei uglei, Izvestiya
visshikh uchebnikh zavedenii-geologicheskaya razvedka, Moskva, 1097-105. LAVROV, N. B., KULAKOVA,M. A., KAZACHKOVA, S. T., ZHIRENYI,A. E., ANTONOVA,R. I. & VOLK, A. F. 1971. O podzemnoi gazifikatsii angrenskogo burougolnogo mestorozchdeniya, Khirniya tverdogo topliva, Moskva, 1, 73-76. LAZAROV, Y. 1986. Habilitazionen trud, MINPROEKT, Sofia. LAZAROV, L., DOUCHANOV, D., MARINOV, S. P. & STEFANOVA, M. 1988. Agglomeration of Elhovo's coal in the pressurised fluidized bed gasifier, Symposium 'Physico-technical problems of energetics', Moscow, Reprint 12. MEAD, S., CAMPBELL,J. H. & NTEPHENS,D. R. 1977. Environmental tests in Hoe Greek UCG. Proc. of 3rd Annual Underground Coal Conversion Symposium, Lawrence Livermore Laboratory, Reprint 770652, 475-489. MINKOVA, V., ANGELOVA, G., GORANOVA, M. & RAZVIGOROVA, M. 1983. Varhu khimicheskiya sastav na vaglischtata na Dobrudjanskiya basein, Khimiya i industriya, Sofia, 5, 207-210. POTTIER, M. & CHAUMET, P. 1976. Gaz6ification souterraine profonde du charbon; problemes et perspectives, L'Industrie du P~trole, 498, Septembre, 53-57. PRASEK, K. & KORANDA, J. 1989. Stav vyzkumu pdzemniho zplynovani uhli, PLYN, Praga, 69/5, 141-145. References SKAFA, P. V. 1960. Podzemnaya gazifikatsiya, Gosgortehizdat, Moskva, 110-115. ANGELOVA,G., DOUCHANOV,D. & RAZVIGOROVA,M. & LAZAROV, L. 1983. The role of catalysts in STRCLAND,L. 1977. In situ gasification of West Virginia coal by 'long wall generator', Proc. of intensifying the process of gasification of lignite to 3-rd Annual Underground Coal Conversion Symproduce synthesis gas, Seminar on Chemical from posium, Lawrence Livermore Laboratory, Reprint Synthesis Gas, Economic Commission for Europe, Geneva, Chem/Sem. 12, R. 17. No 770652, 81-85. ANTONOVA, R. I., GARKUSHA, I. S., GERSHEIV1CH, SZECHY, G., KISS, J. & AZENBEGI, J. 1988. UnderE. G., KREININ, E. V., LAVROV,N. V., SEMENKO, ground Gasification of Coal, MagyarKemikusok Lapja, Budapest, 8, 289-295. D. K. & FEDOROV,N. A. et al. 1967. Issledovanie TRAYANOV B. 1979. Kam Voprosa za Termichnata nekotorikh zakonomernostei protsessa podzemPodgotovka na Vuglischtata za Koksuvane, PhD noi gazifikatsii uglei. Khimiya tverdogo topliva, thesis, Sofia. Moskva, 1, 86-90.
Coalbed methane migration in and around fault zones E. L. B O A R D M A N
& J. H. R I P P O N
International M&ing Consultants Limited, PO Box 18, Common Road, Huthwaite, Sutton-in-Ashfield, Nottinghamshire, NG17 2NS, UK Abstract: One of the characteristics of all operating coalbed methane fields is the considerable variation in producibility success within these fields and even between adjacent wells, suggesting that very site-specific controls are operating. The most likely control in many coalfields is geological structure, particularly faults, which can divide the ground into fluid migration pathways and zones with bypassed, retained gas. Apart from the faulted zone itself adjacent ground will have been subject to dilational or contractional strain, and the strain profiles on either side of the fault will have their own individual permeability characteristics which may be further modified by subsequent burial history. Although any one well will be site-specific, this introductory paper seeks to describe the general ways in which modern understanding of faults and their associated strained ground can contribute to better well spacing and detailed siting and therefore a greater proportion of successful completions.
Methane from coal seams (coalbed methane) is potentially an important source of natural gas worldwide given the large volumes of gas contained in the coals and associated rocks in the world's coal basins. To date successful economic production of coalbed methane from virgin seams is limited to a few areas in the USA. More recently attention has turned to other coalfields, particularly in the United Kingdom, eastern and western continental Europe, China, South Africa, France, Spain and Australia. A common feature of coalbed methane production has been the co-existence of both low and high production wells in the same field and often in close proximity to one another. Current activity in the USA is addressing the problems of poorly producing wells, in particular focusing attention on why certain wells are not producing at their perceived potential even though many are located near to high productivity wells. For example, a well in the Cedar Grove Field of the Black Warrior Basin, which has been in production since March 1990 and currently averages 100 000 cuft/day (CFD) has offset wells in the same area producing between 200000 and 225 000 CFD (Kuuskraa et al. 1994). Surveys in the wells have provided evidence that some of the perforated horizons have taken little or no sand during hydraulic fracture treatment and some perforations are blocked by coal fines. The extent to which these factors are controlled by variation in the coal and strata properties or whether these are purely mechanical/treatment problems is not known. However, in other areas, increased gas production is attributed to the proximity of wells to tectonic structures. Additional fracturing and enhanced permeability have been attributed to the proximity of wells to the Crystal Creek
anticline in the Grand Valley Field in the central part of the Piceance Basin, Colorado (Stevens 1993). The detailed geology and hydrology of this reservoir are still not well understood and it is clear that a better understanding of cleat geometry and the structural controls on natural fracturing will help to improve characterization of the coals in terms of gas producibility. Coal seams constitute unconventional natural gas reservoirs. They not only store gas but they are also a source of gas. They are a more complex form of reservoir than a 'conventional' natural gas reservoir both in terms of their geology and in the mechanisms involved in gas production. Gas producibility from coal seams is generally controlled by the interplay between the following factors: coal distribution, rank, gas content, permeability, hydrogeology, depositional and structural setting. Local variations in the geology and reservoir characteristics must be expected to influence the feasibility of producing methane from them. One of the key local conditions will be permeability. It is aspects of local variation that are addressed by this paper, in particular the way in which faults and their associated strain zones may impose variations in the permeabilities and connectivities of the coals and their surrounding rocks. The paper is based largely on mining experience in the United Kingdom Carboniferous coalfields but is considered to be generally relevant. No single paper can adequately address all productivity settings with respect to faults. This contribution is seen as a general introduction, and discusses general concepts that can then be considered for site-specific use. These concepts are based on the most up-to-date geological models wherever appropriate. While compatible
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 391-408.
392
E. L. BOARDMAN & J. H. RIPPON
TECHNICAL CONTROLS
INHERENT GEOLOGY
* l
1
INVESTIGATION AND INTERPRETATION
PRODUCTION WELL PATTERN
INDIVIDUAL WELL COMPLETION DESIGNS
discussed in this paper
*
Fig. 1. Controls on coalbed methane producibility. Apart from 'external' controls, eg financial regime and marketing, a CBM prospecrs success will depend on the geology, and on the techniques involved in its interpretation and engineering. Geological interpretation is tabulated separately to emphasise the importance of detailed understanding where the geology itself is very variable. with these, it is the further intention of the paper to minimize descriptions and terminologies that are too detailed for an introductory paper.
Geological controls on coalbed methane producibility The ultimate producibility of any individual well will reflect both the site's geology and the design and technology of the investigative and extractive processes (Fig. 1). Individual well design is beyond the scope of this paper. Many geological controls may be considered significant for the ultimate producibility and these are reviewed in the literature. A resum6, with particular reference to two North American Cretaceous formations is
provided by Kaiser et al. (1994). A British contribution is provided by Baily et al. (1995). Figure 2 lists these controls in terms of geological history, from original depositional setting, through the varied phases of burial history, to the present stress setting. In any individual prospect, producibility will frequently reflect several of these controls and their interactions. The depositional setting will determine the thickness, lateral extents, degrees of connectivity, and vertical frequencies of the coals themselves, and also important inter-coal sediments particularly sandstones, which can form associated conventional hydrocarbon reservoirs. Also, ultimate coal strength and fracture permeability may be influenced by the nature and setting of the
I GEOLOGICAL CONTROLS I
I
I I
associated conventional reservoirs
I
net coal thicknees and types
I
max. burial depth
1
uplift and unconformities
I
folding
H'ST~ I igneous effects
I
'i i
fluid migrations
I
joint systems
~
I
stress ratios and vectors
L
I i
interaction with inherited features
coat rank factors
permeability factors
II.
.
]=,,
retained gas content factors
* discussed in this paper
Fig. 2. Geological controls. Many geological factors interact to affect producibility. The present paper only discusses fault systems in any detail.
COALBED METHANE AND FAULT ZONES
393
Time
c o n t e m p o r a r y within depositional cycle
contemporary w i t h overall coal measures deposition
SEDIMENTARY PROCESSES
I faultlets
1
I compaction faults
I rotational slips
_
IGNEOUS PROCESSES
I
I ? delta lobe features
I (e.g)
sill emplacement faults
" growth
Millimet r es
v
I
post-depositional
I
I gas phase faults
TECTONIC PROCESSES
I synsedimentary faulting
I
1
1
post-depositional faulting
faults "
Increase in m a x i m u m possible displacements
Kilometres "11,=._
(sedimentary and tectonic processes)
Post-depositional tectonic faulting is by far the m o s t c o m m o n t y p e , and is the m o s t important economically
Fig. 3. Types of faulting in British Carboniferous Coal Measures. A large generic range is found, and each type will have geometries and linkage characteristics which reflect the very different modes of origin. precursor vegetation itself. Most identified geological controls relate to the long burial history that characterizes many hard coals, especially Carboniferous coals. As Fig. 2 illustrates, several of these factors interact to give the coal rank, and also the basic permeability of the deposit. The present stress setting provides a final control, with the principal horizontal stress vector in particular influencing the directional permeability of the inherited structure. Because of the many interacting geological processes, it is very unlikely that all relevant factors presented by the geology of a particular prospect will be sufficiently known and understood for complete analysis. Figure 1 therefore includes both inherent geology and its investigation and interpretation, as separate controls; in other words, the suitability of investigative techniques, and the applicability of the interpretation, will be critical in maximizing producibility. Of all the geological controls, the most site-specific, and currently the least understood in terms of coalbed methane producibility is considered to be faults and their associated structures (Figs 3 & 4).
Fluid migration through unfaulted ground Before considering the effects of fluid migrations in faulted ground, it is suitable to refer briefly to
the broad aspects of migration through unfaulted ground, specifically coal-bearing sequences. These are laid down in many depositional settings and characteristically comprise very varied rock types, usually reflecting river system migrations. The proportions and geometries of these rock types are also very variable. In most hard coal deposits, the sandstones can be significantly permeable via their well developed joint systems, with the claystones acting as aquicludes. (However, in some high rank coalfields, claystones may themselves be upranked sufficiently to allow passage of groundwaters at depth). Coal can form another permeable rock type, for example, in the near-surface where destressed, and in the tensile strain zones caused by mineworkings; in general, in situ coals in the UK can only rarely be considered permeable at any practical scale, particularly for water, but also for methane. The sandstones can form reservoirs for conventional hydrocarbons, and significant water inflows to mine workings usually involve sandstones as either source or migration pathway. As the sandstones are also the most intricate in geometry, fluid migration pathways cannot successfully be modelled without good borehole data control. A key factor in fluid migration through coal is the jointing; this includes 'cleat' and 'slip' in the British coalfields. Pervasive cleating usually lies roughly normal to the coal bedding, with fracture densities
394
E. L. BOARDMAN & J. H. RIPPON
|FAOL;"SYSTEMS I ..,
j ,,
I
normal
[ L ~
MODE
I
reverse
1
thrust
l
sWike slip
I
"blind"
I 1
intersects
free
surface
I 'DAMA!E ZONE L
gouge rocks sealing
/ non-sealing
fracture
linkages
I LINKAGES
I
I
I
number relay volumetric subsidiary o zones variations faults joint trends fault populations densities orientation (fault densityl rnineralisation
Fig. 4. Fault system attributes. Those factors which describe the compartmentalization of the ground by faulting and which affect the general permeability. See text for details.
sometimes several tens/m; cleat densities vary according to structural setting, and also with coal lithotype and rank. The high rank coalfields of southern England and South Wales have greater cleat densities than those to the north. In general terms, the cleat orientations tend to reflect Variscan compressional trends, with deviations from the regional on the approach to individual faults. The degree of mineralization reflects the geological history of each coalfield, and in the U K can vary from the well mineralized cleats (carbonates and sulphides) of the eastern Pennine Basin to the poorly mineralized coal of South Wales. Many detailed cleat studies of British coals were undertaken in the 1950s and 1960s as part of studies on the introduction of plough coal cutters; these commonly recorded all cleats in terms of density, orientation, and relationship to coal lithotypes for a variety of seams and geological settings. However, there has been little published on cleat in British coals, and for an overall geological appreciation of cleating, and jointing in general, reference may be made to the wider geological literature, e.g. Rawnsley et al. (1992). Further comment on cleat is provided later, with respect to Longannet mine in Scotland. Regarding slip, this is typically a fracture system lying at around 45 ~ to the coal bedding and characteristic of the southern British coalfields of Kent and South Wales. These fractures are usually unmineralized, and form conjugate sets with fracture strike tending to parallel the local 'cross faults' - i.e. those of overall normal fault mode which intersect the Variscan compressional trends of these coalfields at a high angle; however, variations from this parallelism are known. Although often polished, there is only rarely minor displacement of the seams at
these fractures, the density of which is typically much less than cleating, perhaps 1/m being characteristic. Similar slip has been recorded locally in other British coalfields, mainly adjacent to large faults. Fluid migrations through coal-bearing sequences will also depend upon bedding dip, hydraulic gradients, and any mining extractions that have modified the sequence through voids and associated strains.
Faults and fault zones There is a very extensive geological literature on faults and fault zones, both general and also highly technical. The intention of the present paper is to discuss those aspects which relate to fluid migration and/or retention, specifically coalbed methane. Figure 4 tabulates the faultrelated factors. Faults may originate in various ways, some relating to the depositional processes of the host formation but the majority being tectonic and post-depositional. These are invariably categorized as normal, reverse/thrust or strike-slip depending on the dominant slip direction with respect to the fault plane. However, this is very much a simplification, especially when discussing zones of faults. For example a strike-slip fault zone may include many apparently normal and reverse faults. Individual faults may well show both dip-slip and strike-slip characteristics giving oblique slip with extension or contraction. Normal, reverse and strike slip faults and their variants grow in response to changes in the magnitude and ratios of the principal stresses. These may themselves alter throughout the geological history of a fault, from it initiation to its final extent, producing
COALBED METHANE AND FAULT ZONES variations in fault plane dip and curvature. Events in subsequent geological history may reactivate a fault, quite possibly under a different stress regime. The fracture or fracture zone itself will reflect these formative and subsequent stresses together with the physical characteristics of the host rocks and any lubricating clays and fluids. A wide 'damage zone' may result, with a hierarchy of interlinked faults with rotated or crushed pieces of host rock; alternatively a single neat fracture with minimal damage zone width may result. In particular contexts, there may be a relationship between fault throw and the width of the damage zone (e.g. Robertson 1983; Knott 1994). In this paper the term 'damage zone' refers to the immediate faulted volume across which there is measureable displacement and which includes the gouge and the normal drag. For a simple fault, the variation in throw (in normal faults, throw is the vertical component of the displacement) is known to be systematic, with the greatest throw - ideally corresponding to the point of initiation - located centrally on a fault plane that has essentially ellipsoidal limits when viewed in strike projection (e.g. Rippon 1985; Barnett et al. 1987; Walsh & Watterson 1990). Figure 5 illustrates the general principle, showing contours of displacement, the fault limit (tip line) being zero displacement. Following from Fig. 5 it should be noted that a fault cannot be viewed as a fracture independent of the adjacent ground. The variation in throw across the fault necessitates a related strain in the rock volume; in the simplest case of an isolated normal fault, the upper hangingwall and lower footwall of an ideal normal fault will be dilated above regional stratigraphic thickness with a corresponding compression in the lower hangingwall and upper footwall. Such strain zones will approximate to ellipsoidal volumes, and for large faults (hundreds of metres) may show very significant volumetric differences between adjacent hangingwalls and footwalls. Within these strains, the rock fabric will also vary and the most obvious variable is likely to be joint densities. Again, using the idealized simple fault (Fig. 5) the upper hangingwall and lower footwall may be expected to have much better developed joints; this may or may not translate into enhanced fracture permeability depending on geological history.
Fault modification of fluid migration From the above introductory account of faults and their associated damage and strain zones,
395
the following features are seen to affect the permeability of the fractured and strained ground.
Fault plane and damage zone Apart from the simple juxtaposing of permeable and impermeable formations, structural traps in conventional hydrocarbon reservoirs frequently depend on the sealing characteristics of a fault damage zone; oil and gas may be retained by a clay-rock dominated 'gouge', or may have migrated by leakage through 'windows' provided by, e.g. the fault juxtaposing sandstones, with minimal clay-rocks in the gouge (see e.g. Lindsay et al. 1993). Sandstone-to-sandstone situations may, however, still have reduced communication across the fault because of grain size reduction/recrystallization. Such considerations apply to water migrations as well as conventional hydrocarbons. The sealing characteristics will, however, be very site-specific and will depend on the following main factors: the lithologies actually present within the faulted sequence; the number of faulting events and their style; the diagenesis of the faulted rocks; and the displacement relative to the thickness of the bed in question. Potential migration along the fault zone itself will depend on similar considerations. However, there will be differences in the special case presented by coalbed methane, where the reservoir rock is coal, the permeability of which may be very low by comparison with sandstones, and in which producibility commonly requires direct stimulation. In this case, the fault plane and damage zone will aid methane depletion over geological time if the actual fracture pattern extends laterally into the coal seams for a significant distance, i.e. tens of metres.
The adjacent strained volume The general nature of the strained volume (Fig. 5) was discussed above, with the potential for extra joint permeability identified for the upper hangingwalls and lower footwalls of idealised normal faults. Although many examples of these strained volumes are known, actual field observations of the necessary rock fabric changes are rare. At Longannet mine in Scotland, a number of large (tens to hundreds of metres throw) faults are intersected by the mineworkings, and proved/ imaged by boreholes and high-resolution reflection seismic. Integration of all these data has allowed the mapping of the practical limits of the
396
E. L. BOARDMAN & J. H. RIPPON
fault-adjacent strained volume, at mined horizons. Overall, this is expressed as prominent dilation/contraction of the hangingwall and footwall as appropriate, together with some extra minor faulting, mainly in the footwalls; gradient change at the mined horizon is the most
T
"--" "'I-- ......
9-----" i I 1 I I
I"
I I
1"
I ~. I~
~
tool
t t
v /
/
'
/
/
;
"
A
l I I
9"9 I I
I',
1 I
B
Ptan of fault intersecting successive horizons, A,B,C
c
ti
~
. : ~ c , ement (fault
B
obvious visible effect. Figure 6 illustrates the extents of the mappable strains adjacent to one of these, the Abbey Craig East Fault, an essentially isolated structure which is well-proved by mineworkings and by explorations. Changes in the coal, in terms of strength and joint orientation/
I""
- i
/
\
\\ x /~
I
line"
" " .~.
~
\
.\ .L
f ~'--"~ -h-
..~ _
\
X
"
---I--"
'\
'
/./
//
/
n---/:'~g7
/hangi , _ w . , , , - ,oo,-
/
_
f
/
wa~l
/
J
B
/
1
/
/
//~ //// / / ,
Contours of displacement
Cross-section
Projection of fault displacement data on to a ~ : r t i c a l p l a n e , with horizons A , B , C p l o t t e d for the footwall intercept.
of fault and
adjacent strained ground.
\\
./ "~._ ._~.__~ /
Normal ( tensional
'\, )'\, \}~
\ Fig. 5. Idealized simple normal fault. The fault's mapped intersections with three horizons, A B, C are shown in plan view; as projected onto a vertical plane and contoured for displacement; and as a cross section which illustrates the strain in the adjacent ground necessitated by the logically varying displacement. (Displacement = throw/cosine hade).
COALBED M E T H A N E A N D FAULT ZONES
397
o
,i,-
:l
|
1,1.1
.,
7" ~<~
} , /
/
/
,,~/' /, /
//
/
/
,~,.'~ ,
/
,'. ~,
/
9 /t
._ m
/"
.@ |
o,1
.r 0
',-
~
&
o~
.=
]
,- _~.=_
o
o
(.~ ~ 9 Q" " "
"~ t" N
S
0
o~
~Z
oi~~~
'~-
.,~.: 9 ~
g
~-~
~,!~! ~
~
"~.~'n
:...~
"!~.~ ~._~,,-.....~
w o ~Z
m
t "?
~,= ~.~ ~ ~o~
0
...o ~
~
~D
~ ' ~ o
398
E. L. BOARDMAN & J. H. RIPPON e
/ -__.__._ 9
/"
2'
,<.t
9
./
-.__L ~_. 9
T180m
;
0-n,' . "
9
.J
/
-e-.-e-.
/
o
'~.
.,,.~
.....-:.-.o0,,,
~30.i \" ~\
.-._. ~
~ '" "
\',..
\\:-.,
,.
.i m
"'
....
e
all d a t a at Upper Hirst horizon
I
j~-~
-__
% \
.~,,<,.,,~
o
_~,~.
. ,
9
...v / 2 / 9.'I /
Borehole 1 km
I
~ . . -
..6v.
"~.
o
CRAIG
EA.S,~
~-'~" ~ ' ~ . ~ ' ~ .
~.~.~- ~'.-:-.-~.-~-LL~.-9- ~4,~OLv
_.f.-..~"
"~..~
\
"---~--
..._
"'~"~
~.
9
' \i =~-___ ~
"
.
. . ~ - - ~ . ~ ~ ,, ~. ~ ~ ,,-'*
"]
cleat
;
/"
-
"~
~ .---.-~.. ~o-~--~ ~,. ">"~ ~
Main c l e a t a d j a c e n t t o f a u l t s ( n o t all f a u l t s are s h o w n )
x /
....,,.. ~o~:-.,..
~Lo~.._ c.~~.~. ~/
....___.
"\
.._z_.'~..~\.
------
Main m i n e r a l i s e d
",,,
'~
.~
~
o
~"
", \ .~.',<,
...~.\
....
a.
"~,
""-T _._
/
/ ,.,<-~ /-/-
~.~ ,7~
~--.c,,'~_,.. "-<-->,.
--" ,
.y
o
~,,>,..~/..----:_:~ / "---- "~ / - ' - 7
?~f
/
/ /
I~ . - ~ ' -~ ~ .,<.,~.__-.---
/
\
\
ALLo~'.~...~................~ ~"~" ~ "~
1,. "
..
b. ---)4- P r i n c i p a l h o r i z o n t a l
...~ ~ F~u~'E-"
|
\ ..~.."Je~ . -'-- "~.
"~,~
.~ . ~ - ~ . e
-*~-
">,~.
',~,
Im
o
stress away from faults I
Borehole 1 km
1
~ " "~"
,
COALBED METHANE AND FAULT ZONES frequency, are difficult to assess visually apart from very local swings in the main cleat; however, there are signifcant character changes noted from in-seam seismic transmission and reflection surveys, and reduced integrity of surface seismic reflectors is also prominent across these zones. The example in Figure 6 is drawn at a coal horizon, but it should be noted that the dilations/ contractions adjacent to these larger faults may actually relate more to the prominent sandstones and claystones that characterise the Longannet sequence, as, in bulk terms, these present the greater opportunity for modifications. In most British coalfields, an increase in coal cleat density is sometimes recorded in the immediate approach to a significant fault. This is not always the case at Longannet. The regional cleat trend there (Fig. 7) subparallels the main fault trend, and is prominently mineralised with carbonates, the secondary cleat generally being subordinate in density and less mineralized. The situation at Longannet is locally complicated by marked reduction in cleat density in areas subject to mild thermal metamorphism, and by intense fracturing adjacent to minor intrusions; however, the general picture shows only local increases in cleat density on the approach to faults. Much more noticeable is that the main cleat orientation can swing towards both smaller and larger faults, sometimes lying normal to the fault within a few metres of it; this recalls the noding of faultadjacent joints reported by Rawnsley et al. (1992). The Longannet descriptions are included here only to illustrate the potential geological relationships; there is no specific coalbed methane content. Given this background, it is assumed that for coalbed methane, fracture permeability will be enhanced in the dilated volumes in the coals, as well as the sandstones, but that the emphasized fracture systems in the coals will be more subtle. This probably represents the best conditions for producibility, with good retention and a latent fracture system that can be stimulated to encourage methane release. Those zones adjacent to the fault which have suffered compressional strain should, by contrast, have a much less pervasive fracture system. Field measurements of the bulk strength of in-place coals are notoriously difficult to standardise, mainly because of fracture systems; however, in such compressed settings, overall
399
stronger and 'tighter' coals are to be expected, probably with very good seismic propagation and reflection characteristics. Unless modified by later geological history, coals in these zones may be expected to have significant retained methane, but also poor producibility because of reduced fracture permeability. The fracture aperture effects of these strains are currently unknown, and will be difficult to detect directly in mined exposures, where disturbance is inevitable. In general it must be assumed that apertures will be greater where the fracture lies parallel to the local principal stress; open jointing will be prone to mineralizations through later geological history.
Fault linkages and populations The degree of connectivity of a fault fracture system will influence the potential for any fluid migration. Fault linkages may be considered in the following categories: (i)
(ii)
kinetically-related linkages, in which all the fractures result from a common geological deformation phase; other linkages, in which faults from various geological deformation phases interact and intersect to give a more complex situation.
As with the fault-adjacent strained volumes described above, there will be a fine balance between migration loss through fractures, a n d ultimate producibility through stimulation of latent fractures. In these circumstances, it is proposed that particularly beneficial coalbed methane producibility may be found in those zones where faults overlap. Fault overlaps (Peacock & Sanderson 1994; Childs et al. 1995) occur where the overall deformation associated with an individual fault - including its adjacent strained volume - interacts with that of another. Again, this particular form of 'linkage' may be kinematically consistent, or not, with differing detailed geologies as a result. Within these overlaps, ground strain may be considerable, with changes in horizon dip, and fabric change including extra faulting and jointing. Such zones may therefore provide the required permeability for producibility, without the overall fracture system depleting the methane over geological
Fig. 7. The Abbey Craig East Fault: coal cleat and present stress. This structure lies in the Scottish Midland Valley, some 40 km northwest of Edinburgh: it is essentially an isolated fault allowing easy study of its displacement and strain patterns. Fault displacement in metres. (a) The regional mineralised cleat shows general parallelism with the gross fault trend except immediately adjacent to faults. See text for discussion. (b) The present principal horizontal stress trend. See text for discussion; variations close to larger faults are known and on-going work is seeking to define these.
400
E. L. BOARDMAN & J. H. RIPPON
time because the fracture pattern is not fully linked. Figure 10 illustrates an overlap/relay zone for a strike slip fault system. The concept of fault populations, well known in coal mining for many decades, has recently been developed for the assessment of fracture patterns in conventional hydrocarbon reservoirs. A fault population is a system of faults of all sizes in which there is some numerical relationship between the size categories (e.g. Walsh & Watterson 1992). For tectonic faults, power law size distributions are sometimes applicable. However, in the authors' experience there are significant gaps in fault size distributions in many British coalfields and power law size distributions may not be universally applicable. Ideally a fault population should be based on one geological deformation phase, but at least some published sets appear to include faults of several phases. The potential for describing conventional hydrocarbon prospectivity using fault population studies- in which fracture numbers below seismic resolution are assessed from those revealed by seismic- is very dependent on choice of the most suitable mathematical model, and is beyond the scope of this paper. For coalbed methane, it is considered that fault population studies will rarely have applicability, as they are essentially regionally-based whereas coalbed methane producibility requires much more site precision. Those aspects of fault population studies that may have coalbed methane application are as follows: (i)
(ii)
fault damage zone: specific fault hierarchies and linkages making up the broken ground itself; preferred faulted horizons: faults initiated preferentially at a particular depth, and propagating largely within a particular depth interval. For example, Rippon & Raine (1986) implied that fault populations lay at differing preferred horizons (Westphalian A/Westphalian B) at two adjacent collieries in Derbyshire, England. It is thought likely that such horizon preferences are common in many coalfields, with each 'cell' having its own fracture characteristics (orientations, size relationships, preferred nucleation depths) on a scale of 25 km2-100 km 2.
From this listing of fault-related features, it will be seen that fluid migration and retention characteristics are likely to vary significantly along, and adjacent to faults. However, any idealized pattern may well be modified by subsequent geological history.
Fault zone burial history In UK coalfields, the main deformation and faulting phase is generally considered to be later Carboniferous/early Permian, with some later reactivations. In certain coalfields, the later movements have both complicated the existing faults and initiated new faults. By contrast in other coalfields, existing faults have been reused, with renewed but essentially uncomplicated growth. These differences must partly reflect the orientation of the existing faults to the new stress pattern. In addition to reactivation, fault zones will be modified with respect to their permeability attributes by burial/uplift history; and by regional fluid migrations and mineralizations resulting from igneous replacements, or driven by tectonic uplift (e.g. Daniels et al. 1990) or from other geochemical redistribution mechanisms. It should be expected that such fluid migrations in the geological past will themselves have taken advantage of any structurally controlled permeability patterns presented by a fault zone, to the extent that a key fracture system, e.g. the main coal cleat, may be largely sealed by mineralization effected by these fluids. The burial history of a fault zone should also include consideration of the recent/continuing present stress field. It is known from British coalfields that stress magnitudes and ratios can vary over an area as small as one mine lease, and also vary through lithology and stratigraphy. For example, at Asfordby mine southeast of Nottingham (Whitworth et al. 1994), the variations indicated some response to ground heterogeneities produced at least partly by inherited, or 'fossil' fault structures. Current work by the authors is investigating the response of the present principal horizontal stress to the fossil strain zones adjacent to the Abbey Craig East Fault at Longannet mine (Fig. 6). As described previously the strain zones have been investigated and mapped for the mined horizon, and the volumetric differences between the lower footwall and hangingwall strains assessed using their fabric differences, which are thought to be largely joint orientations and frequencies. Ongoing mapping aims to define any refractions of the stress vectors, both fossil and present, within the strain zones. The modifications produced by 'fossil' heterogeneities on the present stress field are not currently well understood, and again are likely to be highly site-specific. However, the orientation of the principal horizontal stress with respect to that of the inherited fracture pattern will have considerable influence on well design and producibility for coalbed methane.
COALBED METHANE AND FAULT ZONES
Models for coalbed methane retention and migration One of the principles behind the present paper is that coalbed methane producibility is very sitespecific and will be strongly affected by fault zones. However, it is appropriate to the understanding of individual prospects and well sites that the general principles be understood. This can be attempted by combining the following: 9 idealization of normal, thrust, and strike slip fault geometries; 9 assessing the physical changes imposed on the adjacent rock ('fabric changes') with special reference to fracture permeability; 9 assessing any burial history particulars; 9 assessing the interaction between the 'fossil' fault structures and the present stress fields, drawing upon mining experience.
401
progressively greater subsidence towards the fault plane, directly influencing net coal thickness on a very local scale. It should be noted that in some circumstances, the greater coal thickness might alternatively form in the footwall. For example, the hangingwall may be subsiding too fast for peat generation or retention, with the footwall offering a better water-table balance within an overall basin subsidence setting. The fault plane in Fig. 8.4 is shown as concave to the hangingwall, or listric, only to illustrate another possible shape for normal faults. The listric shape is not essential for the 'growth' aspectactive during deposition - illustrated here. Also, the increasing inter-horizon thicknesses towards the fault should be recognised as potentially including both depositional increases and dilational effects from subsequent fault growth.
Thrust faults and zones Normal faults Figure 8 illustrates some of the interactive factors which will influence ultimate coalbed methane producibility. Given the very varied geology of faults and fault zones already described, only a few interactive factors can be dealt with in such simplified drawings, which are chosen more to illustrate general principles than provide immediately usable templates. Figure 8.1 illustrates an idealized normal fault. The stylized bedding is shown at a wider spacing across the faulted interval merely for clarity. In this simple case, the fault is seen as essentially unmodified by later geological events; it did not intersect the original free surface, and it is not exposed by the present erosion level. The idealized distribution of permeability characteristics, with respect to coalbed methane migration, retention and producibility are described above. In Fig. 8.2 the upper part of the idealized fault has been eroded, with likely coalbed methane depletion in the upper hangingwall. It is thought that widespread unconformities result in wholesale degassing over significant vertical ranges (Creedy 1988). However, subsequent reburial by later basin development may allow secondary gas generation (Fig. 8.3). The potential for retention of this methane will depend on many burial history aspects, including depth ranges, and the nature of the unconformity and overlying formations. Figure 8.4 shows a fault that was growing during deposition of the coal-bearing sequence. Here, coals are thickening into the fault on the hangingwall side, where vegetation growth and preservation were able to keep pace with the
Reverse f a u l t s - taken here to have fault plane dips characteristically greater than 45 ~ to bedd i n g - are considered by the authors to be relatively rare except in strike slip fault zones (see e.g. Fig. 10). They only rarely form in isolation, and very rarely form in regionally integrated systems; some will be reactivations of originally normal faults. These are, therefore, not considered in any detail in this paper. Thrust faults (Fig. 9) in the southern British coalfields, with Variscan deformation, typically have low fault plane dips, <_5~ commonly to 30 ~ with steeper dips where the propagation of the fracture is through stronger rocks (ramps). In strongly folded coalfields such as the Ruhr and the Appalachians, thrust fault plane dips may be much more pronounced. A detailed consideration of thrust-related permeability is provided by Hathaway & Gayer (1996). A simple thrust is commonly associated with folding in the immediate footwall and hangingwall (Fig. 9.1). Higher permeability is to be expected in the hangingwall from well-developed cleating/jointing and maybe thrust-parallel shears. Locally in the South Wales coalfield, the footwall is 'tight', presumably reflecting downwards variation in displacement; such settings may have characterised some of the gas outburst situations in the western anthracite field. The detailed distribution of very low permeability anthracite in relation to thrusts or gas outbursts is, however, not known and it would be inappropriate to draw any definitive conclusion. The great majority of thrust faults occur in linked systems as in Fig. 9.2. These may be idealized in various ways, partly depending
402
E. L. BOARDMAN & J. H. RIPPON
1. Simple normal fault, "blind", minimal linkages to other faults, little degassing
(| \J'
I
(~) volume increase; increased permeability (~ volume decrease; decreased permeability
| .i- /
/
/
/,/I
2. Erosion : essentially degassed, especially in upper hangingwaU (| ? residual gas retention in lower footwall ((~))
3. Early degassing by uplift / unconformity; secondary gas generation from subsequent reburial, with potential retention in lower footwall
4. "Growth" fault w i t h thicker coals along fault hangingwall, with
probably enhanced permeability (| Probably low-permeability footwall with generally thinner coals ((~))
Fig. 8. Normal fault examples. Many faults will undergo modifications over geological time and their related permeability aspects will also be altered. Only some examples are illustrated. See text for further comment. on the closeness to the contemporary free surface and on the lithologies involved. No work is known on the relative prospectivity of differing thrust systems. Because of their generally low fault plane dips, thrust fault propagations are frequently modified by lithological contrasts. Figure 9.3 shows a thrust ramping up
when intersecting a sandstone. Here it is suggested that the hangingwall coal, lying between the fault plane and a constraining strong sandstone, will be highly sheared, giving the possibility of coal-derived gas accumulating in the sandstone as a conventional gas reservoir, assuming a suitable clay-rock cap.
COALBED METHANE AND FAULT ZONES
403
1. Simple thrust displacing coal seam (~) high permeability in hangingwall, potential degassing (~ reduced permeability in footwall with potential for (gas) outbursts
|
(based on mining experience in South Wales coalfield)
2. Thrust systems
imbricate fan; ? potential for degassing
thrust duplex; ? potential for gas retention
3, Lithological control on thrust affected ground impermeable cap
~..,.,,--0
(~) likely to be highly sheared; ? gas accumulation in conventional reservoir
Fig. 9. Thrust fault examples. The permeability characteristics of the ground adjacent to thrust fault systems are particularly poorly understood, reflecting conventional hydrocarbon interests in generally extensional regimes. However, many permeability patterns may be envisaged. See text for further comment, and Hathaway & Gayer (1996) for detail.
Strike-slip faults and zones Strike-slip fault systems present very varied patterns, both in map and section view. The simplest form, idealized as a vertical or nearvertical fault plane, with very subordinate
vertical displacement, is rarely seen in British coalfields, although well-documented examples are known to the authors. More frequently, identifiable strike-slip zones include a significant dip slip displacement and fault plane dips at up to 45~ with such attributes, their strike-slip
404
E. L. BOARDMAN & J. H. RIPPON
'~~
~
~ 9~ ~ 1 7 6
~o "~
Maps of strike slip faults
~o ~~176 ~
1. Early stage fault zone with minimal linkage but high strain / permeability between individual faults, potential high gas retention
9. . .
2. Highly linked fully formed
9
fault zone with potentially extensive degassing
3. S c h e m a t i c strike slip s y s t e m ( m o d i f i e d f r o m W o o d c o c k and Fischer, 1986)
.
.
.
.
- -
~
9
~
9
"
releasing offset 9~ (relay zone)
", ~ " ~"
'k ~s'=lIt
"ly
releasing bend
restraining bend
....
!
/ /
9
.
.
r i
Complex gas retention potential, probably maximissd in releasing offset (high strain / permeability, with minimum linked system fractures)
Fig. 10. Strike-slip fault examples. A very wide range of fault styles will represent strike slip faults and zones, with correspondingly very complex strains and permeability patterns. See text for comment.
character may be difficult to discern. In these cases, individual component faults will appear to be normal, or reverse (Fig. 10) with the overall zone showing prominent lateral movement. These differences may at least partly reflect erosion levels, as the cross section of Fig. 10 implies. British coalfield fault mapping tends to record only the dip slip throw component, any strike-dip displacement being difficult to document because of generally low bedding dips. Because of the complexity of many strike slip zones and their considerable variation (see e.g. Naylor et al. 1986) it is difficult to generalize in any useful way about their attributes with
respect to fluid migrations, and particularly coalbed methane. Hence in Fig. 10, the diagrammatic map which is based on Woodcock & Fischer (1986) shows a range of fault settings. It was suggested above that the fault overlap zone illustrated by this may offer a greater producibility, combining strained ground with no through-going fracture. Depending on the relative depth of the fracture pattern with respect to the coal-bearing sequence, it is further suggested that the zones adjacent to the 'root' faults (see Fig. 10 cross sections) may also be prospective. Strained volumes adjacent to strike slip faults will be more difficult to identify unless the
COALBED M E T H A N E AND FAULT ZONES
405
Fig. 11. Detailed interpretation of FMS. Detailed interpretations of image data, such as this example of FMS, are routinely carried out to resolve seam boundaries, coal cleat, natural and induced fractures and sedimentary features. For this coal seam, one individual cleat among many is shown bold, in the form of a 'sine' curve on the 'unrolled' image of the borehole wall.
406
E. L. BOARDMAN & J. H. RIPPON
bedding is steeply dipping, because of the absence or randomness of observable dip slip. Such a fault will be mistaken for a normal fault of small throw, whereas its strike slip component may be hundreds of metres, with compatible large strains.
Investigative and analytical techniques Mining areas, sometimes with laterally extensive workings, often at many levels, provide valuable data on the subsurface geology. The East Pennine coalfield of the U K for example extends over an area of some 11000 km 2, the productive Carboniferous coal measures attain a maximum thickness of about 1500m and contain up to 30mineable seams in parts of the coalfield. Coalfields provide unique data sets on the intensity, style and extent of faulting, and detailed analysis is essential to an understanding of the controls on the subsurface flow of fluids. UK mining data sets result from a legal obligation on mining surveyors to record faults encountered underground, or in opencast mines. For most 20th Century mineworkings, these are recorded on the detailed 1:2500 scale, with smaller-scale plans for mine planning purposes derived from these. It is thus possible to investigate fault patterns to a very intricate degree. At this scale of investigation, fault patterns and populations may be seen to vary significantly between mined horizons (Rippon & Raine 1986) and the need for specialist geological interpretation is emphasized.
Seismic surveys Surface seismic reflection surveying has been used as an exploration tool in the British coal industry for more than 20 years with very significant refinements in both acquisition and interpretation in recent years. Today the surface seismic technique is used for delineating the positions of sedimentary disturbances and fault throws that are little greater than seam thickness in the best data areas. Faulting of less than 5 m in throw has been imaged by modern 2D surface seismic lines and faults of around 2 m throw by recent 3D surveys. In virgin areas the surface seismic, often integrated with borehole geophysical techniques, is used to reduce risk for immediate and long-term mine planning and provides an additional large source of detailed structural coalfield data.
Borehole surveys Downhole geophysical wireline logging in the UK coalfields was initiated in the 1950s when several holes were logged using Schlumberger resistivity, natural gamma and temperature tools. It was not, however, until the 1960s with the development of slim-hole tools by British Plaster Boards (BPB) Instruments Ltd that geophysical logging of boreholes became a routine feature of coal exploration. The early logging package comprised natural gamma, density, caliper and sometimes a neutron log and was primarily utilized to interpret seam thickness and relative quality (principally ash content). A rapid expansion in coal exploration in the early 1970s and the introduction of surface reflection seismic into the industry to evaluate geological structure was linked with the development of dipmeter and sonic logs and the running of seismic reference surveys to allow calibration of the sonic logs for synthetic seismogram generation. The processed dipmeters allowed detailed interpretation of structural information from a borehole, which coupled with the surface seismic data led to the accurate mapping of faults with throws above 10-20m. The problem of the detection of minor structures by downhole logging was, however, not resolved until the later introduction of the Schlumberger Formation Micro Scanner (FMS) in 1986. This was developed from the standard Schlumberger High Resolution Dipmeter (SHDT) by adding a large number of small microresistivity buttons to the SHDT pads. The data from these buttons are now captured digitally which allows interactive processing and interpretation to be carried out. The detection of fractures and faulting from the interpretation of FMS data sets has proved to be very successful (Onions & Whitworth 1995). It is only in certain lithologies such as weak seatearths where detection has proved difficult. Several types of natural fractures have been observed on the FMS. These include: (i)
(ii)
Coal cleat: orthogonal joint system specific to the coal seam consisting of closely spaced subvertical fractures together with any mineralization (Fig. 11). Fracturing and faulting: faulting identified when either a change in lithology occurs at the identified feature or there is clear evidence of displacement of laminae.
Data sets already exist from boreholes which penetrate faulted ground. It should be possible to analyse fracture intensity within faulted volumes and compare the intensities both
COALBED METHANE AND FAULT ZONES within and outwith the faulted zone. In addition, in-situ permeability data from suitably located holes in similar situations could be acquired to provide additional data from subsequently drilled holes. Downhole borehole testing equipment is now available which can not only carry out in situ permeability testing but can also be used for in situ stress testing. Permeability tests are carried out by isolating part of the formation in the borehole by means of packers. The natural flow and pressures can then be measured, or injection of fluids can be undertaken to measure permeability. Using higher pressures the formation interval can be hydrofractured and measurements of in situ stress obtained. The orientation of induced fracturing and hence the orientation of the principal horizontal stress direction can be subsequently derived from FMS images run over the fractured part of the borehole wall.
Discussion The foregoing introductory review has emphasised the site-specific nature of many coalbed methane well settings. A general interpretation scheme is therefore inappropriate, but the following principles result from the overall intricacies and interactions of various geological factors. (i)
It is likely that, in faulted ground in particular, generalizations of practical value will be inappropriate over areas greater than 5 km 2. (ii) All aspects of coal geology (Fig. 2) should be considered for overall prospect appraisal, with their relative importances refined for individual wells, or groups of wells. Given the cost of data acquisition, it is important to honour all available information. Extensive practical experience in coal geology is required to achieve this. (iii) The integration of all data resources, from historical mine plans through to new seismic reflection data, is also essential, particularly where fault plane dips are low. The construction of cross sections normal to the fault trends will often be necessary to achieve this. These cross sections (1:5000 is suggested as suitable for such data integrations) constrain the interpretations and allow accurate assessment of throw variations and the extent of strained ground. (iv) Specifically regarding faulting, interpretation should seek to combine all available
(v)
407
data (on faults, joints, dips; field-mapped faults and seismic interpretations) with the most appropriate theoretical understanding. Neither of these elements on their own will give adequate understanding of any particular prospect. It is unlikely that consistent success will be achieved, particularly in Carboniferous prospects, without refined exploration techniques, tailored to site conditions.
Conclusions The cost-sensitivities of many coalbed prospects, and the intricacies of their geology, require detailed appraisal for optimized production. This is particularly the case in faulted areas, where juxtaposed sequences may have very different structural and permeability characteristics. It is hoped that the present introductory paper will serve to stimulate interest in the structural contribution to well siting. Modern coal exploration techniques have been developed to an extent where it is now possible to resolve structural and permeability data to a high degree of precision. Many existing data sets have already been acquired in areas affected by faulting and it is considered that a number of these could be subjected to detailed interpretation to provide information on the fabric changes associated with different fault settings. The authors acknowledge the approval of International Mining Consultants Ltd for this paper, and for the company's support in its production. The Scottish Coal Company Ltd, and Rock Mechanics Technology Ltd are thanked for approval to use material related to Longannet Mine. R. A. Gayer, P. Gillespie and T. Hathaway have suggested valuable modifications to an earlier draft. Views expressed in this paper are those of the authors and not necessarily those of any of the above-mentioned companies.
References BAILEY, H. E., GLOVER, B. W., HOLLOWAY, S. YOUNG, S. R. 1995. In: WHATELEY, M. K. G. & SPEARS, D. A. (eds) European Coal Geology. Geological Society, London, Special Publication 82, 251-265. BARNETT, J. A. M., MORTIMER, J., RIPPON, J. H., WALSH, J. J. & WATTERSON, J. 1987. Displacement geometry in the volume containing a single normal fault. AAPG Bulletin, 71, 925-937.
408
E. L. B O A R D M A N & J. H. R I P P O N
CHILDS, C., WATTERSON,J. & WALSH, J. J. 1995. Fault overlap zones within developing normal fault system. Journal of the Geological Society, London, 152, 535-549. CREEDY, O. P. 1988. Geological controls on the formation and distribution of gas in British Coal Measures strata. International Journal of Coal Geology, 10, 1-31. DANIELS, E. J., ALTANER, S. P., MARSHAK, S. & EGGLESTON, J. R. 1990. Hydrothermal alternation in anthracite from eastern Pennsylvania: implications for mechanisms of anthracite formation. Geology, 18, 247-250. GIBSON, J. R., WALSH, J. J. & WATTERSON, J. 1989. Modelling of bed contours and cross sections adjacent to planar normal faults. Journal of Structural Geology, 11, 317-328. HATHAWAY, T. M. & GAYER, R. A. 1996. Thrustrelated permeability in the South Wales Coalfield, UK. In: GAYER, R. & HARRIS, I. (eds) Coalbed Methane and Coal Geology. Geological Society, London, Special Publication, 109, 121-132. KAISER, W. R., HAMILTON, D. S., SCOTT, A. R., TYLER, R. & FINLEY, R. J. 1994. Geological and hydrological controls on the producibility of coalbed methane. Journal of the Geological Society, London, 151, 417-420. KNOTT, S. D. 1994. Fault zone thickness versus displacement in the Permo-Triassic sandstones of NW England. Journal of the Geological Society, London, 151, 17-25. KUUSKRAA,V. A., LAMBERT,S. W. & SCHRAUFNAGEL, R. A. 1994. Black Warrior coalbed methane productivity improvement project. Quarterly
Review of Methane from Coal Seams Technology, 11, Nos 3, 4, 50-55. LINDSAY, N. G., MURPHY, F. C., WALSH, J. J. & WATTERSON, J. 1993. Outcrop studies of shale smears on fault surfaces. Special Publication of the International Association of Sedimentologists, 15, 113-123. NAYLOR, M. A., MANDL, G. & SIJPERSTEIJN,C. H. K. 1986. Fault geometries in basement-induced
wrench faulting under different initial stress states.
Journal of Structural Geology, 8, 737-752.
-
ONIONS, K. R. & WHITWORTH, K. R. 1995. Applications of electrical borehole imaging to mining design. Scientific Drilling, 5, 69-75. PEACOCK, D. C. P. & SANDERSON, D. J. 1994. Geometry and development of relay ramps in normal fault systems. AAPG Bulletin, 78, 147-165 RAWNSLEY, K. D., RIVES, T., PETIT, J. P., HENCHER, S. R. & LUMSDEN,A. C. 1992. Joint development in perturbed stress fields near faults. Journal of Structural Geology, 14, 935-951. RIPPON, J. H. 1985. Contoured patterns of the throw and bade of normal faults in the Coal Measures (Westphalian) of northeast Derbyshire. Proceedings of the Yorkshire Geological Society, 45, 147-161. & RAINE, J. D. 1986. Some methods of assessing fault pattern variations in the East Midlands Coalfield. The Mining Engineer, April 1986. ROBERTSON, E. C. 1983. Relationship of fault displacement to gouge and breccia thickness. Mining Engineer, 35, 1426-1432. STEVENS, S. H. 1993. Coalbed methane - state of the industry, Piceance Basin, Colorado. Quarterly -
Review of Methane from Coal Seams Technology, ll, 23-27. WALSH, J. J. • WATTERSON,J. 1990. New methods of fault projection for coalmine planning. Proceedings of the Yorkshire Geological Society, 48, 209-219. & 1992. Populations of faults and fault displacements and their effects on estimates of fault-related regional extension. Journal of Structural Geology, 14, 701-712. WHITWORTH, K.R 1994 (ed.) Investigation into the
Effects of Lithology on the Magnitude and Ratio of in-situ Stress in Coal Measures. Commission of the European Communities, Directorate General for Energy, ECSC Agreement No. 7220-AF/845. WOODCOCK, N. & FISCHER, M. 1986. Strike slip duplexes. Journal of Structural Geology, 8, 725-736.
Geological research into gas sorbed in the coal seams of the Carboniferous in the M~eno-Roudnice Basin, Czech Republic V. HOLUB ~, M. ELI.AS l, P. H R A Z D I R A ~ & J. F R A N C U 2 1 Czech Geological Survey, Kldrov 3, Prague 1, 118 21, Czech Republic 2 Czech Geological Survey, Leitnerova 22, Brno, 602 00, Czech Republic
Abstract: In 1994 research in the Czech Republic began into coalbed methane (CBM). The target of this research is the coal seams of the M~eno-Roudnice Basin, in the region between M~lnik and Benfitky nad Jizerou. This region was chosen as a model area for evaluating the CBM potential in the Permo-Carboniferous continental basins in the Bohemian Massif. The first stage of the work was directed towards obtaining data for basin analysis, the completion of lithological, stratigraphic and structural surveys particularly concentrated on fault deformation of the basin filling. A general study was conducted into the problem of the sorbed gases with respect to the occurrence of CBM. The latest interpretation of laboratory analysis of the M61nik Main coal seam shows a methane content of 4-10m3/t in the coal at a depth of 250 to 800m.
The Czech Geological Survey in collaboration with Energie Kladno a.s., the Faculty of Natural Science of Charles University, Prague and the Institute of Geonics of the Academy of Sciences of the Czech Republic, Ostrava, in 1994 began research into coalbed methane (CBM) potential. The target of this research is the coal seams of the M~eno-Roudnice Basin, in the region between M61nik and Benfitky nad Jizerou (Fig. 1). This region was chosen as a model area for evaluating the CBM potential in the Permo-Carboniferous continental basins of the Bohemian Massif. This paper presents the results of the initial stage of research.
Geological setting The M~eno-Roundnice Basin was chosen as a model area because geophysically and through drilling it is currently the most extensively investigated continental Permo-Carboniferous basin in the Bohemian Massif, with total proved geological coal reserves of 1 286922kt (Bosfik & Zbfinek 1992). The results of research into this basin to date have been summarized by Holub & Tfisler (in Malkovsk~, et al. 1974), Holub & Tfisler (1981) and Bosfik & Zbfinek (1992). In accordance with Holub & Pegek (1992) we consider the M~eno-Roudnice Basin as one of the basins of the Central Bohemian region of the continental Late Palaeozoic basins of the Bohemian Massif, lying to the east of the K l a d n o Rakovnik Basin and to the west of the Mnichovo Hradi~t6 Basin. It is situated NNE
of Prague. The M~eno-Roudnice Basin is shown in the map of the continental type of the PermoCarbinerous basins of the Bohemian Massif (Fig. 1). The crystalline basement of the MgenoRoudnice Basin is predominantly composed of epi- to mesozonal metamorphosed Proterozoic, which in the eastern and south-eastern parts of the basin is partly overlain with slightly metamorphosed Ordovician rocks belonging to the Caradocian. These units form a part of the Bohemicum in the sense used by Mahel' et al. (1984). A strip of volcanic rock of granodiorite to diorite type extends below the southern wing of the Carboniferous basin, classified by Mahel' et al. (1984) as Moldanubicum. The sedimentation of the Late Palaeozoic in the major section of the M~eno-Roudnice Basin began with the N ~ a n y Member (Westphalian D) of the Kladno Formation. The strata sequence continues through to the Lin~ Formation (Stephanian C to lower A u t u n i a n ) - as shown by the profile of the Zd6tin borehole Zd-1 (Table 1). The Radnice Member was identified in only a small part of the basin, near its south-eastern border in the vicinity of M61nik. In most of the southern part of the basin, the N~[any Member forms the basin fill but it wedges out to the north and north-east. In this area, sediments of the T~nec Formation are the basal lithostratigraphic unit of the Carboniferous basin fill. The N~,[any Member is the oldest coal-bearing unit in most of the basin. The major development of the Slan~ Formation distinguishes the M~eno-Roudnice Basin from the other regional-geological units
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 409-423.
410
V.
t ./'..:...:
9. . . . : .
f , .~.~
.~,
~
, r -;.
~
--.,~s -"~,~,~, " .
j
~.~
- -...-r
,~..a,.~" . . - . ""/ -". ' ". .'+.'. ~ "~
~~.- -o
r
I?:".".-'/:-".":
.3
-'. " " ".'--."" "- ~,
9
E T AL.
.,~'7 ~ - : . : : S ' ; --~
i.~-" . . .-,._ <.....'.~
HOLUB
"~
.If!i-.'. -. :-.- .r
.~
L . / : : :.'.'." : : " - : "?."~
,%_. :-.':-.: . : . :-.-.:-,:;
9
F'"::
~".'~-
".'::'!~,
['::': : : , ' ~
! t ,'~ ~
+,.
~:I9
~-
9..0
9. .::..-..'~,,,d . . --t )/, 9
." 0 " "
11t~--'+
IIIII U I b
:';, A , . ~ ~llli."
-.Z'::.'.:~
J'
~-
p
a
. =
i
o
-,
z
2'
-.,
N H 0
\
~2
i \
%
~.
.
x-
-." /
E
.X
O.
(D. O. (D, ,I-
o
_-
/
.
_~-}"
GAS SORBED IN CZECH
REPUBLIC
COAL SEAMS
411
Table 1. Stratigraphy and coal-bearing sequences in the Zddtin (Zd-1) borehole (Mgeno-Roudnice Basin)
(Holub 1994) s~ra%
i g r a p h y [
general stratLgraph, scale I l o c a l
I .... s%ra%igr, scale
seams (in metros)
coal
[
Qua~er.! C R E T A C E O U S
I
!
iIT I u p p e r
I l
u
P
Ir[ m i d d l e lo i r~! 1 . . . .
-"
lol
iu"
r
icier
,
[
Teplice Jizcra
-
B i lakn d Horn
depth
(fros -
to)l
!:!-*'* I ~t ~resh-
ai nl et uerrocpael lai tt ieos n s with coal
Peruc
subst.admixture
thickness (in metres)
I I o.o-i.3
I I
1,3
I 1~
1
4,5
i-! I 5,8-239,4 ti I! !
233,6
1239.4-281,4
Xot'Tc~ny
ne
! I
1
42,0
! I l 1281.4-289"1 I
7~
l [
! !
I
'
lower iPermian
u
!i
lower
I'. I
C
e r b o n
__
] 589,&-596,0 i
barren e q u i v a l e n t os ~he Kounov C o a l s
i
I
u
~
-
I
Htedle
-
1639,2-761', 8 I
i ! I I
I
Jelentce
! !
I
I.
!
Slant
.
.
m
-
I
,
I I l I
I
-
Ledce
s
4
294,45
t i 1 f
I
14.45 i m
I
I I I
s
! f
1598"0-613"4 i ] 1 . - )613,4-639,2 ]
'!i
I
I
! !
l
i
i Kounov
eI
I I I I I lal inl
126901-583,6
I
Kamenn~ Mos~
t.sl r
-
Lln~
761,'8-770,9
u p p e r sea 770.90-771.78 (0,88) 1 c o o t s / gun punch c o r e :768.30-796,25 (0.94) l n t e r b ~ l c o a l seam 778.27-778,34 (0,07) coring gun punch c o r e :77R,20-778060 (0,40)
15,35
I
25,80 122.60
9,15
I
70,1
770,9-841,0
! (
main seam
7 8 4 , 2 7 - 7 8 7 , 6 2 (3,35) c o r i n g gun punch c o r e : 7 8 1 , 5 0 - 7 8 5 , 2 0 (3,70)
1
I
I 841,0-910.1
T~nec
69,1
1
I t
,Iel
'
Iii
IPi
,.! IPROTEROZOIC
I ~
i
i t!
'[
*-~
Nttaay
1
with barren e(Dlivalents os t h e Nt~any C o a l s carbonised stems
(968,fl5-969,15)
I
I
79,3
910,1-989,4
I
,
i i
! I
I
989,4-1042,2
52.85
Fig. 1. Permo-Carboniferous basins in the Bohemian Massif (of continental character with the area studied) (Holub 1994). 1. Permo-Carboniferous basins of the Sudetic region: a, Cesk~ Kamenice Basin; b, Mnichovo Hradi~t6 Basin; c, Krokono~e-Piedmont Basin; d, Intra-Sudetic Basin; e, occurrences in the Orlick6 hory Mrs; f, Orlice Basin. 2. Permo-Carboniferous basins of the Central Bohemian region: a, Plzefi Basin; b, Man6tin Basin; c, Radnice Basin; d, Zihle Basin; e, K l a d n o - R a k o v n i k Basin; f, M~eno-Roudnice Basin; g, occurrences in the neighbourhood of Krava~e. 3. Late Palaeozoic in the Krugn6 hory Mts: a, occurrence at Brandov; b, occurrences at Teplice and Moldava. 4. Permo-Carboniferous sediments of the Furrows: a, Blanice Furrow; b, Boskovice Furrow; c, Jihlava Furrow. 5. Researched area.
412
V. HOLUB E T AL.
of the Permo-Carboniferous of the central and western Bohemian region. Its oldest unit, the Jelenice Member, is most fully developed in the Mgeno-Roudnice Basin of all the central Bohemian basins, both with respect to maximum thickness and relative coal-bearing richness. The most significant correlation horizon of the entire Permo-Carboniferous filling is the Mgec Member which forms the roof of the M~lnik Coals. In the eastern section of the basin, the thickness of the upper part of the Slan~ Formation decreases markedly; considering the nature of sedimentation, it is clear that it is caused by the primary reduction of the later units of the Slan~ Formation strata. The Ledce and Kounov members which locally are only occasionally coal-bearing, are reduced to only a few metres of coal-rich layers and the Kamenn2) Most Member is notably reduced or, in some places, totally absent. The Lin~ Formation is the thickest lithostratigraphic unit. It forms the transition between the characteristic features of the Central Bohemian and Sudetic Permo-Carboniferous regions: in the eastern section rock types which are also characteristic of the Krkonoge Piedmont Basin can be observed. In the lower section of the Lin~ Formation, the correlative Zd~tin Horizon occurs. It is overlain by conglomerates containing pebbles of Silurian and Devonian limestones from the Barrandian area. In all formations of the basin the products of explosive volcanic activity are present. These are found most frequently in the Lin6 and Slan~, formations, where they are predominantly of rhyolitic composition, and rarely of intermediate to basic character. The Mgeno-Roudnice Basin is structurally asymmetric with a basinal axis trending WNW-ESE. The cross section through the basin is shown in Fig. 2. The Permo-Carboniferous basin-fill is tectonically disturbed by frequent faults, predominantly radial in character. Their vertical throw ranges from tens to hundreds of metres. The Permo-Carboniferous filling of the basin is overlain by an Upper Cretaceous cover which overlaps it with a slight discordant unconformity. The Cretaceous sequence is stratigraphically limited at its base by freshwater Cenomanian, and at its roof by deposits belonging to the lower Coniacian. Their lithology is very varied depending on the fact that the Cretaceous filling of the Mgeno-Roudnice Basin is located on the borders of four facies regions- Prague (Vltava-Berounka), Lu~ice, Jizera and Labe. The predominant section of the Cretaceous basin-fill belongs to the sediments of the middle Turonian; the upper Turonian and Coniacian deposits occur only as denudation relics.
Tectonic Pattern An outline of the tectonic characteristics of the basin has been given in the previous section. Opinions concerning the fault structure of the Mgeno-Roudnice Basin depend on the level of knowledge and the divergence in conceptual approach of individual authors. The basic characteristics have been presented by Holub & Tfisler (1981). They distinguished two basic fault systems: one striking NE-SW (transversal to the structural lines and in places, especially to the east of the studied area, twisting towards the Rhine direction) and the other generally younger than the Sudetic system (NW-SE, twisting in places towards WNW-ESE) striking parallel to or obliquely with the direction of the axis of the basin. Many faults are shown in the tectonic map by E. Stanik (in Zbanek et al. 1991). In contrast with the above outline there is a conspicuously large number of faults trending W N W - E S E in the southern section of the basin. A preliminary suggestion of the possibility of block dissection of the Mgeno-Roudnice Basin into basement blocks was put forward by St~p/mek (in Holub et al. 1994). Three basic zoned blocks are characterised by various levels of tectonic fractures, while one of the distinguishing phenomena is also the varied homogenity of the Carboniferous basement. The eastern and western blocks are notable for the presence of narrow trough structures restricted by steeply dipping faults in the Rhine direction. The central block is relatively less disturbed with fewer subsiding dislocations of both main systems. The most eastern block is characterised by greater morphological dissection of the basement with some crystalline uplifts with no Carboniferous cover. Partial depressions continue to the eastern Mnichovo Hradigt6 Basin and, from a regional facies point of view, belong to the Sudetic region. The essential characteristics of the tectonic structure of the MgenoRoudnice Basin based on a synthesis of knowledge to date, are shown in Fig. 3.
Coal deposits Bituminous coal seams occur in the basin at four stratigraphic levels. The most significant is the M61nik group of seams located in the Jelenice Member of the Slan~, Formation (Stephanian B), followed in importance by the N~,~any group of seams in the N2)~any Member (Westphalian D) of the Kladno Formation; the Kounov group of seams (Stephanian B - upper section) was noted
t~
I
l~l Z '0 0 H
E
~§ +'4---
E E
=~
i
E 0
Z
0 H
m
~ ~~
o
0 t~ I
o
I ' ~
~"-
o~
~D
I,~;-
~,~
0
~
~
0
~..,~
~ =
~'~
~-~
"~
~ ~
~
0
=.~
'~
~
~
~
~
~
-4"
~
~;-~
~'~
f~
D ~ ~ ~ co
~
~
0
~
V. H O L U B E T AL.
414
/ I , t .i'
o
I O s /
z~
, ,,,,~ O t'-
~.Q,J Z , , ~ . ~ . . , m
.a
GAS SORBED IN CZECH REPUBLIC COAL SEAMS in a small area. In the lower section of the Lin~ Formation, thin coal seams occur in the Zd6tin Horizon (see Table 1). It is possible to divide the Mgeno-Roudnice Basin into the three coal deposit areas: the Mfilnik area, the central part of the basin and the Benfitky nad Jizerou area. In the N)~any Member (Kladno Formation) it is possible, according to Bosfik & Zbfinek (1992), to recognize six groups of coal seams which they name N R 0 to N R 5 . Mostly these contain 1-2 thin coal seams, and in places only coal representatives or coal seam equivalents. Coal formation in the N ~ a n y Member gradually extended from the SW section of the M61nik district in a NE direction, so that the seam group N R 4 has already expanded to cover almost the entire area. Seams N R 0, N R 2, N R 3, N R 4 are identified as having a thickness greater than 0.4 m. Altogether, they are however without any great significance. In the Slan~ Formation there are two coalbearing horizons: the M61nik and Kounov groups of coal seams. Bosfik & Zbfinek (1992) zoned five groups of coal seams within the M61nik Coals in the Jelenice Member. The base and lower M~lnik seam had a greater thickness in the eastern part of the Benfitky nad Jizerou area. The most important is the M61nik Main seam extending regularly almost throughout the entire area examined at a thickness of more than 1 m. From the point of view of coal production as well as of the occurrence of CBM, this seam is of the greatest significance. It is lacking in only a few places of the middle part in the deposit area where it either was originally absent or has been since eroded. Additionally, it was not found in several boreholes at Kropfi6ova and M~lnickfi Vrutice, where it is most likely missing because of faulting. Apart from this, the character of the M~lnik seam in the basin is regular. From its southern border, the basin slopes to the north at the southern border, it occurs at a depth of between 160-200 m; in the north, it has a depth of 900-1000m. At the southern border, it is limited through erosion or tectonic faulting. In the W and N directions the seam slowly wedges out or diminishes. The Jizera fault system has probably contributed to its being restricted in the east part of the basin. So far it has not been possible to determine its continuation towards the NW (towards the Lib~chov borehole), and further in the east from the borehole at Brodce in the vicinity of Ben~itky nad Jizerou. A connection of the M~lnlk Main seam from the M~lnlk area to the Benfitky nad Jizerou area was found in the central coal deposit-area along the
415
N and S borders of the basin (Bosfik & Zbfinek 1992). The M61nik Interjacent coal seam was found in its usual thickness only in the SE section of the Benfitky nad Jizerou area, the M~lnik Upper seam in the NW and NE sections of the M61nik area, and in maximum thickness in the southern section of the Benfitky nad Jizerou area. In the Kounov Member it is possible to delineate the Kounov Coals with five seams. The Kounov seam was located at its maximum thickness only in the NW section of the M61nik region, and mainly after that in the centre of the Central and Benfitky nad Jizerou areas. Scant coal formation was noted in the Zd6tin Horizon of the Lin~ Formation in the lower section of the strata unit. The Zd6tin Coals are formed in five seams; however, only on the N W border of the M61nik area and in the Roudnice section of the basin was the Zd6tin coal seam found at maximum thickness. From an economic point of view it is, however, insignificant. The remaining stratigraphic units of the Permo-Carboniferous of the M~eno-Roudnice Basin are barren with respect to coal deposits. Rare thin seams occur only in the T~,nec Formation and in the H~edle Member of the Slan~, Formation. They do not, however, have significance as resource deposits.
Technological properties of coal Earlier technological exploration up to the present time was directed towards the characterization of the coal from the viewpoint of classic mining and utilization. Technological characteri s t i c s - ash content, calorific value, specific sulphur content with average values, review of the thickness of individual seams, as well as area and tonnage of recorded reserves of bituminous coal are shown in Table 2. It should be noted that perhaps 40% of the coal reserves are located in the protected water sources area. From the point of view of a total technological evaluation, the main resource is the M61nik Main seam, which is both the best in quality and the most significant in quantity. Its quality is mostly very good and raised sulphur content occurs only locally. However, some harmful elements are present in significantly high proportions, e.g. arsenic (average content 154 g/t), zinc (average 307 g/t), mercury (average 0.25 g/t) and fluorine (average 0.018%). Technologically the coal is evaluated as energetic (useful for power plants), combustible,
416
V. HOLUB E T AL.
Table 2. Average technological parameters and coal geological reserves of individual seams in the MYeno-Roudnice
Basin (after Bos6k-Zbdmek et al. 1992)
Coal seams
Zd6tin Kounov M61nik Upper M~lnik Intrabed M61nik Main M61nik Lower M61nik Basal N~,~any 4 N~any 3 N2~any 2 N~any 0
Average technological parameters
(Ad)
Thickness (m)
Ash (%)
0.42 0.68 0.58 0.48 2.22 0.49 0.44 0.58 0.61 0.51 0.58
51.35 53.20 44.26 42.48 31.98 32.31 17.97 34.81 34.69 38.47 43.27
Calorific value (Q ~MJ/kg)
Specificsulphur (g St/M J)
13.13 11.81 15.26 10.13 19.26 19.16 21.21 17.82 17.99 15.89 14.82
1.30 2.07 0.68 0.53 0.95 1.30 0.37 3.44 3.44 3.84 4.27
poorly sintering, commercial group V I A with average numeric code 622, easily adjustable, with good floatability, but without independent coking capacity. It is predominantly susceptible to self-ignition and its dust is evaluated as explosive. The ash of the M~lnik Main coal seam is of medium fusibility, with an average value of around 1350 ~. Special exploration for natural gas or CBM was not undertaken earlier in the M~enoRoudnice Basin, as previous exploration was directed towards evaluation of the coal deposit and its possible suitability for coal production. The occurrence of methane in coal seams was observed only superficially and non-systematically. The method used for setting technological and geochemical parameters did not enable the measurement of the sorption properties of the coal seams, their porosity and other properties essential to the evaluation of possible CBM reserves. It was thus necessary to undertake special research to measure the sorption properties of the coal seams, their porosity, permeability and other properties essential for the estimation of CBM reserves. The main target of the research was the M~lnik Main seam. Samples of coal were taken from 23 exploration boreholes throughout the area representing the variable quality of coal from the M61nik Main seam over the entire studied area. Concurrently, samples of coal from the Kounov and N~,[any seams were also taken from several of these boreholes. These coal seams are found in only some sections of the basin and represent only completion of the coal reserves, with coal which has different properties from that of the M61nik Main seam. Coal from the Kounov seams is of lower rank
Area (103 m 2)
Tonnage (kt)
2 544 27 862 51 020 13 064 308 618 12 851 1 918 71 249 81 327 17866 6 008
1 818 32 588 47911 9 958 1 010 044 9 308 1 140 62116 74428 14071 5 540
than coal from the M61nik Main seam from the same locality. In contrast, coal from the N ~ a n y seams is of higher rank than coal from the M61nik Main seam in the same locality. Analyses of coal seams and associated rocks of the M61nik Coals from boreholes in the Benfitky nad Jizerou area are presented in Table 3. The characterization of the coal properties from samples taken all over the M~eno-Roudnice Basin is shown in the Table 4. The following properties were determined for each sample: the thickness of the M61nik Main seam; depth in the borehole; ash content; vitrinite reflectivity; maceral characteristic; volume of volatile matter; total volume and surface of pores; porosity; potential gas content with respect to methane; pore size and permeability. To enable mutual correlation, all tests were carried out on identical samples. Contour maps were constructed from the data obtained: (a) (b) (c) (d)
(e)
(f)
depth of deposition and the thickness of the M61nik Main seam; ash content in dry matter and the thickness of the M61nik Main seam; porosity, potential gas bearing content and the thickness of the M~lnik Main seam; vitrinite reflectivity R0, content of volatile matter in combustible coal material and the thickness of the M61nik Main seam; total pore volume, potential gas bearing content and the thickness of the M~lnik Main seam; total pore surface, potential gas bearing content and the thickness of the M61nik Main seam.
GAS SORBED IN CZECH REPUBLIC COAL SEAMS
417
Table 3. Analyses of coal seams and associated rocks of the MYlnlk Coals from boreholes in the vicinity of Benftky nad Jizerou (Klibfni, NYmec et al. 1994) No. of sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Collection Seam label depth 686.5 687.5 688.2 688.6 689 691.7 692.2 692.3 692.5 692.7 693 693.5 693.7 693.9
overlying sandstone
Reflectance R0
M61nikUpper 0.59 coal seam 0.59 sandstone M~lnik Main 0.56 coal seam 0.55 0.55 0.57 0.58 0.58 0.54
Mineral part 85 80 82 25 30 79 35 38 32 26 28 33 74 23
From the contour map of vitrinite reflectivity R0 and volatile matter in combustible coal material it can be seen that with an increase in depth of the seam deposit there is a rise in vitrinite reflectivity and a decrease in the content of volatile matter in the combustible coal material and vice versa. Similarly it follows from the contour maps of total porosity, total pore volume and surface area in relation to potential gas content of the M61nik Main coal seam, that increase in depth leads to a significant reduction in the porosity of the coal material which is also associated with a lessening of the total pore volume and surface area and simultanoeusly also of the potential gas-bearing property of the coal. These changes in reflectivity, volume of volatile matter and porosity of the coal material indicate changes in coalification which increases with depth of the M~lnik Main seam. Where the M~lnik Main seam can be found at shallow depth, which is in the south of the M~eno-Roudnice Basin, values were obtained that showed greater porosity, lesser reflectivity and a raised content of volatile matter and simultaneously a higher gas content. The character of the coal material in the M61nik Main seam, together with the values for reflectivity, total porosity and potential gas content show that this seam has the necessary qualities required for the sorption of CBM produced during the coalification process. Whether that quality of the coal material in itself is adequate for the seam to hold a significant volume of producable methane (desorbable), depends on factors which determined the presence and
Maceral analyse Vitrinite (%)
Inertinite (%) Liptinite (%)
70 65 70 67 65 69 70 68 63
30 35 25 26 31 28 24 28 28
10 5 7 4 3 6 4 9
geological development of the basin. These are primarily the origin and development of postCarboniferous formations including heat, tectonic and hydrogeological effects. Concurrently with the geochemical evaluation of the coal samples from the M~eno-Roudnice Basin, a complex evaluation of knowledge obtained from an exploration borehole drilled at Zd6tin was also undertaken (borehole Zd-2). The results showed that the Upper and Main M61nik seams at a depth of 690m represent a similar coal type and are differentiated from each other by ash content, which is in agreement with the results of borehole log measurement. The coal cores obtained were tested not only for coal quality, but also examined in detail for the evaluation of the quantity and quality of desorbable gas ('container test'), from which was ascertained a gas-bearing value of 10.8m 3 t -1. The composition of the desorbed gas was consistent with the composition of natural gas. The results of this desorption test in relation to the seam depth and other properties of the coal and in correlation with results obtained from analysis of the coal from the M61nik Main seam from further boreholes, suggests the existence of prognostic reserves of methane held in the M~lnik Main seam. Confirmation of this prediction is the subject of further research. The results of the analysis of the coal seams of the M61nik Coals and associated rocks from the newly-drilled borehole at Zd6tin, and samples taken from earlier boreholes drilled N of Benfitky nad Jizerou, are presented in Tables 3 and 4.
418
V. HOLUB E T AL. Table 4. Basic characterization of the coal properties from samples of the entire MgenoRoudnice Basin (Klib6ni & N~mec 1994) Sample
Depth (m)
Coal
w a ( % ) Aa(%) vdaf(%) R0(%)
JB-1 Jabkenice
820 823 515 550 770 460 1080 250 907 927 724 745 749 907 288 315 490 683 825 327 500 693 284 793 948 748 660 844 843 1025 1035 944 155 250 234 241 356
M61nik M~lnik M61nik M61nik N~any Kounov N2?~any M~lnik M61nik M~lnik M~lnik M~lnik M~lnik NS,~any M~lnik M61nik N2}~any M61nik N~,~any M~lnik M~lnik NS,~any M~lnik M~lnik N~,~any M61nik M~lnik N~:~any Kounov M61nik M61nik M61nik M61nik M61nik M61nik M61nik M61nik
2.01 2.11 2.01 3 3.13 1.33 1.58 3.08 3.12 0.75 1.07 2.33 1.83 1.71 3.25 3.9 2.31 2.29 3.43 3.96 3.06 1.7 2.73 1.87 1.99 2.76 2.07 2.74 1.17 2.23 0.65 1.17 4.46 4.71 4.58 3.68 3.97
BC-1 Brodce Hu-1 Hrugov CHT-1 Chot~tov KBL-1 Kbel SVK- 1 Sovinky SS-1 Su~no
DS-1 Dolni Silvno KV-1 Kropfi6ova Vrutice MV-2 M61nik6 Vtelno SZ-1 Stfi~ovice HS-1 Hostin KRP-1 Krpy CHO-1 Choru~ice VUJ-1 VelkS,Ujezd RD-1 Radoufi SI~-I St~emy VS-1 Vysokfi MUJ-1 Mal~: Ujezd LBL-1 Liblice LBL-2 Libllice BS-3 By,ice BS-4 By,ice
Thermal history of the M~eno-Roudnice Basin Prognosis of CBM and its recognition for possible production requires detailed information concerning the properties of the seams and of the surrounding rocks. A m o n g the required characteristics are particularly the physical qualities of the environment (especially those which affect the migration of fluid), and knowledge of the conditions affecting the balance between the solid stage (the coal material of the seam) and the liquid and gas phases (methane, carbon dioxide, carbon monoxide, and nitrogen etc.). Conditions for the state of balance changed during geological development with changes of pressure and tempera-
9.8 18.77 22.36 12.6 20.71 39.01 28.54 18.8 16.03 18.57 53.15 26.94 13.47 18.04 35.09 24.22 22.77 18.16 20 7.15 11.41 22.43 11.14 21.77 21.12 36.84 30.77 12.29 38.81 27.16 34.28 19.85 12.52 24.36 22.59 21 31.07
41.43 39.93 34.5 38.18 45.33 65.12 45.39 43.66 48.46 67.66 46.94 41.92 43.21 38.28 43.07 43.17 44.92 41.96 42.77 41.61 45.52 42.44 44.24 49.69 45.59 45.36 50.3 38.01 35.04 47.03 47.58 42.18 43.56 53.33 44.09 44.33 41.83
0.62 0.69 0.7 0.69 0.73 0.54 0.65 0.57 0.66 0.62 0.71 0.7 0.68 0.75 0.53 0.56 0.57 0.59 0.62 0.58 0.58 0.6 0.53 0.6 0.68 0.57 0.59 0.67 0.73 0.72 0.7 0.72 0.48 0.52 0.53 0.55 0.55
ture. If that led to a disturbance of balance, it could have caused desorption of gas from the coal and its migration from the coal seams, or conversely it could have produced suitable conditions for its sorption into the coal material through migration. Although quantification of these data is difficult, a rough model is proposed to outline the trends of heat development, the maturing of organic material, the formation of gas and the progress of subsidence and erosion. The model of of burial and thermal history of the M~eno-Roudnice Basin is based on: (a)
analysis of the thermal history of the M~eno-Roudnice Basin (Francfi et al. 1994, Kone~n~ et al. 1994),
GAS SORBED IN CZECH REPUBLIC COAL SEAMS (b)
analysis of compaction of sediments on the basis of changes in the volume density and porosity of rocks in the borehole profiles (Kone6n~ et al. 1994), analysis of bio-markers (Francfl et al. 1994).
(c)
According to (~ermfik (in Ibrmajer, Suk et al. 1989), the M~eno-Roudnice Basin has, from the point of view of the regional character of the heat flow of the Bohemian Massif, an increased fluid temperature which is typical for the region of the Lu~ice-Labe line. This can be seen from depth-temperature data obtained during studies of the boreholes St~emy MB-7, Brodce BC-1, Kropfi6ova Vrutice KV-1, M61nick6 Vtelno MV1 and Sedlec MB-21. This anomalous of the heat flow follows a NW-SE (Sudetic) direction. The thermal gradient of the borehole MB-7 is 31.5~ km -l, of the borehole MB-21 31.8~ -1 and of the boreholes BC-1, KV-1 and MV-1 33~ -l. These gradients are used as a starting point for the model of organic material and for orientational assessment of the sorption process in the methane-coal system.
Well: HPivno- 1 B u r i a l history (thickness
Geothermal data are used from the borehole Sedlec MB-21, where the relationship between temperature (t) and depth (d) is expressed: t = 0.0318d+ 9.5(~ and where the heat flow q --79.6mWm -2. For construction of the model, the evolution of the basin was divided into 'events', which have either the character of sedimentation, erosion or without sedimentation. Each event is characterised by the thickness of the deposits, time interval, lithology of the rock, hydro-chemical qualities of the water and palaeo-climatic data at the interface between the sediment and the water. The heat flow for the given section and its change with time is based on the geotectonic position of the basin. To calibrate a model of the thermal history of the studied area of the M~eno-Roudnice Basin, the thermal maturity of the organic matter was measured. According to these data, Francfi et al. (1994) proposed a model of burial, erosion and geothermal history of the Late Palaeozoic and Cretaceous sediments for the studied section of the basin (Fig. 4).
Temperature
decompacted)
Jurassic
419
,,
us
(C)
Plle0ge
I,~
W~lt seoUon t 1"~ K~T~Ih
~K~c -3 rC2"sln 2-
Cgtno C:-nm
E
.10
g r
|
Q
~]
IZ3 [~ I]ITII! [~
3OOO
noo
BELOW
30"C
30- 60 6 0 - B0 90-12o
1 ~ 0 - 160
2~o
1~o
Time
lko
(mabp)
ob
o
IE$
Fig. 4. Burial and thermal history of the Hfivno-1 borehole: Sedimentation in the Late Palaeozoic was followed by erosion during the Permo-Triassic to the Late Cretaceous. Additional post-Turonian sedimentation is assumed followed by erosion of these missing units (Francfl 1994). PDIT M software has been used.
420
V. HOLUB E T AL,
The basis of this model was analysis of organic material from rock and coal. From these samples, mineral and total organic carbon was analysed and Rock-Eval pyrolysis was performed. Further analysis of bituminous extracts with the aid of gas chromatography and mass spectrometry was carried out on selected samples. This analysis was directed towards identification of 'bio-markers', i.e. in the group of organic materials, particularly the polycondensed structure of sterane and triterpane type. The main goal of this analysis was to identify the trend of maturation of organic materials so that it would be possible to test a model of the development of the studied section of the Mgeno-Roudnice Basin from the point of view of the possible presence of CBM. Examples are given in Fig. 5 which shows the maturity of organic material in relation to depth, shown as the index of isomerisation of biomarkers. The following conclusions can be derived. 1. At the boundary between the Cretaceous and Permo-Carboniferous there is an obvious partial jump in maturity. 2. The increase of maturity with depth in the Hfivno Hi-1 and Skuhrov Sh-1 boreholes shows two parallel continuing trends in coalification, while the profile of the second of the examined boreholes appears to be more uplifted and more heavily eroded.
~ 200
3. Thermal maturation of organic matter connected with the formation of hydrocarbons is also evident in the distribution of extracted saturated hydrocarbons. With increasing depth there is a proportional increase in the content of lighter homologues over heavier (C17-C21)/ (C27-C31). It is therefore possible to deduce that in the Carboniferous sequences, which are today at a depth below 300 m the formation of liquid hydrocarbons occurred. Conditions for their preservation were and are better than in the shallower sediments of mainly Upper Cretaceous age which are less influenced by compaction. Results from analysis of organic matter and from the model of the geothermal history of the studied section of the Mgeno-Roudnice Basin can be summarized as follows. 1. According to the vitrinite reflectance and the index of isomerisation of bio-markers, the level of maturity (coalification) increases with depth in the borehole. At the boundary between the Cretaceous and Permo-Carboniferous there is a jump in maturity from which it may be deduced that after deposition of the PermoCarboniferous and before Cretaceous sedimentation there was deep burial, and subsequent erosion of a part of the section. 2. Erosion in the SW part of the basin (in the vicinity of the Skuhrov Sh-1 borehole) appears to be greater than in the central eastern region (Hfivno Hi-1 borehole).
x~~
Skuhrov-1
--
_____X
Hrivno- 1 E
v "(: 9 <..., EL (D
i 400
'
a
<~~ Cretaceous [] Carboniferous
i. 800
i 0.00
I 0.20
'
I 0.40
'
I 0.60
' 0.80
22S/(22S+22R) ~13C31Homohopanes Fig. 5. Biomarker maturity increase with depth in two boreholes: Hfivno-1 (in the E) and Skuhrov-1 (in the W) (Franc6 1994).
GAS SORBED IN CZECH REPUBLIC COAL SEAMS 3. Maturity of organic matter in the Cretaceous sediments reveals that these were also buried to a greater depth than at present. 4. The model of burial and thermal history of the basin based on the above suggests the following events: 9 sedimentation of the N ~ a n y Member, T~nec and Slan~ formations (Westphalian D Stephanian B) 9 sedimentation of a sequence perhaps 1400 m thick (Stephanian C to Autunian) and its erosion in the Upper Permian and Triassic and to a restricted degree in the Jurassic and Cretaceous 9 sedimentation of the Peruc and Korycany members and of the Bilfi Hora Formation (Cenomanian - Lower Turonian), 9 hypothetical sedimentation of a sequence perhaps 1100m thick in the uppermost part of Cretaceous, and its subsequent erosion. 5. The formation of hydrocarbons in the deeper Carboniferous sequences took place mainly at the close of the Palaeozoic and also, however, in part at the end of the Upper Cretaceous. Analysis of the compaction of sediments of the Permo-Carboniferous (Konern~, et al. 1994) led to similar results.
Hydrogeology Knowledge of the hydrogeological relationships of the M~eno-Roudnice Basin makes it possible to predict or to reject places as having a likely concentration of CBM. The PermoCarboniferous sedimentary filling of the basin can be hydrogeologically characterised as a collection of irregularly alternating rock strata of varying levels of permeability (from high rock permeability, e.g. conglomerate and sandstone layers, to impermeable rocks: claystone and siltstone). According to the hydrogeological properties of the rock there are restricted hydrogeological aquitards and aquicludes in the profile of the basin (see Table 5) Aquicludes prevent the communication of groundwater and CBM between aquitards. The comunication however permits tectonic disturbance and changeability of facies of the rock complexes. In the hydrogeological aquitards it is possible to predict the accumulation of CBM above all in aquitards within formations with M~lnik Coals. The enrichment of the Permo-Carboniferous aquitards occurs at the aquitards outcrops outside of the basin. The
421
Table 5. Geological and hydrogeological definition of the MYeno-Roudnice Basin (elaborated by Hrazdira 1994, after to the data by Boshk-Zbdmek et al. 1992)
Formation
Member Coals
Kt Kj Kb Kpk
Teplice Jizera Bilfi Hora Korycany Peruc
PC1 PC s
PC t PC k
Hydrogeological unit aquitard K1 aquiclude K1-K2 aquitard K2 aquiclude K2-C1
Kamenn~ Most Kounov Kounov Ledce H~edle M~ec Jelenice M~lnik N~any Radnice
N~any
aquitard C1 aquiclude C1-C2 aquitard C2 aquiclude C2-C3 aquitard C3
PA o PT Legend: K t, Teplice Formation (Cretaceous); K j, Jizera Formation (Cretaceous); K b, Bilfi hora Formation (Cretaceous); K pk, Peruc-Korycany Formation (Cretaceous); PC1, Lin6 Formation (Permian and Carboniferous); PC s, Slan~, Formation (Carboniferous); PC t, T2~nec Formation (Carboniferous); PC k, Kladno Formation (Carboniferous); PAo, Ordovician; PT, Proterozoic.
descending flow of groundwater is made more difficult by vertical and horizontal facies changes, and by the decrease in permeability of aquitards in relation to the depth of deposition. Some hydrochemical properties of groundwater in the Permo-Carboniferous aquitards (e.g. low values of T.D.S. in comparison with groundwater in Cretaceous aquitards) or higher values for rock filtration parameters in aquitards indicating communication of groundwater and CBM between hydrogeological complexes. The migration of both groundwater and CBM via these communicational paths can be assumed.
Conclusions After evaluation of the available geological data and determining more precisely the lithostratigraphic and tectonic structure with a view to the
422
V. HOLUB E T AL.
possible presence of CBM, it has been shown to be most important to concentrate attention on: 9 the conditions for the formation and development of CBM in coal seams dependent on the geological conditions and the history of the Mgeno-Roudnice Basin, particularly between M61nik and Ben/ttky nad Jizerou area (lithology, tectonics, basin analysis) 9 geochemical and mechanical properties of the M~lnik Main seam as a major source and reservoir of CBM and on the characteristics of the surrounding strata as a sealing horizon 9 analysis of the hydrogeological relationships within the area with respect to the presence of CBM A lithostratigraphic model of the filling of the basin was produced and modifications were made to the current tectonic model (Fig. 3). Stress fields in the basin were orientationally evaluated. A basic contribution of this stage is the further evaluation of post-sedimentational history of the M~eno-Roudnice Basin using analysis: 9 of compaction of sediments (evaluation of changes in the volume, density and porosity of rock at depth), 9 thermal maturation of organic matter (biomarkers). Both methods show that on top of the PermoCarboniferous, a sequence of strata roughly 1400m thick was deposited, and eroded before sedimentation of the Upper Cretaceous sequence. From the research carried out it was further shown that the Cretaceous sequence reached a total thickness of perhaps l l00m. Hydrocarbons were generated in the more deeply deposited strata of the Carboniferous mainly at the end of the Palaeozoic and also at the end of the Late Cretaceous. Primary attention was given to the depositional and geochemical evaluation of the studied region of the Mgeno-Roudnice Basin, from the viewpoint of the presence of CBM. 39 coal samples were taken from 23 earlier-drilled boreholes, and these were fully studied geochemically. Areal evaluation of these results is shown in the Table 4. Analysis of the condition for sorption and desorption of methane in coal is difficult because of insufficient available data. Despite this, taking into consideration current work of similar character, we can state that the effect of reduction of pressure, e.g. as a result of erosion of the cover, is a reduction of the original volume by only 20%, while a reduction in temperature from 44~ to 20~ could by contrast increase the sorption of methane by
18%. With respect to sorption and desorption changes, the volume of methane bonded to the coal basically should not change. With a sufficient quantity of produced gas, the sorption capacity, depending on the type of coal and coalification (coal rank), is roughly 9 m 3 t -1. Starting from this point of evaluation, the difference between current gas volume and sorbed gas volume may be considered to be losses caused by diffusion and transportation via porous environments. It is not possible to determine the dynamics of diffusion without a more detailed study of the transporting properties of the rock. Calculations already made, however, show that uplift of the massif has occurred, and that therefore its permeability may be also locally increased. During the preparation of the final version of this work we used data supplied by the organizations which cooperated on the project of the Czech Geological Survey, Prague entitled 'Geological Research into Gas Sorbed in the Coal Seams of the Carboniferous in the Mgeno-Roudnice Basin'. The following co-workers of the following organizations took part in the project: Energie Kladno, a.s. (Klibfini, L., N6mec, J., including Medek, Holub/t~, Dopita, Kozfik, Kraus and H)ka), Faculty of Natural Science of the Charles University Prague (Hrfich, S. including Blecha, V., Kn~z, J., Kobr, M. and Skopec, J.), Institute of Geonics of Academy of Science of the Czech Republic, Ostrava (Kone6n~,, Petr, Martinec, P., Holub, K. and Kone6n~, Pavel).
References BOS,~K, P. & ZB,~,NEK, J. 1992. Geologiek~ pom~ry lo(iska (erndho uhli M~lnik - Bendtky nad Jizerou.
Geologick2) Prfizkum. 1992, 11, 325-330. Praha. FRANCU, J. et al. 1994. Anal/~za biomarker~ a model geotermickd historie sediment~t m~ensko-roudniekd pfnve. MS Archiv (;esk6ho geologick6ho fistavu.
Praha. HOLtJB, V. et al. 1994. Geologiek~ v~zkum sorbovanf~eh plynd v uhelnf~ch sloj[ch karbonu mYensko-roudniekd pdnve. MS Archiv Cesk+ho geologick6ho fistavu.
--
Praha. & PEgEK, J. 1992. H. Svrehni karbon a perm. In: CHLUPA(~, J. & STORCH,P. (eds) RegiondJnd geologick~ dgleni Cesk~ho masivu na fizemi Cesk~ republiky (Regional geological division of the Bohe-
--
mian Massif on the territory of the Czech Republic). 12as. Mineral. Geol., 37, 4, 263-267. Praha. & TASLER, R. 1974. Mladgipaleozoikum a spodni trias v podloi Cesk~ kfidov~ p6nve (Late Palaeozoic and lower Triassic in the basement of the Bohemian Cretaceous Basin). In: MALKOVSK'I', et al. Geologie CeskO k(idovO p6nve a jejiho podloYi
(Geology of the Bohemian Cretaceous Basin and its basement). V~d. Ust~edni flstav geologiek~ v Academii, nakl. CSAV Praha.
GAS SORBED IN CZECH REPUBLIC COAL SEAMS --
& 1981. Geologie mgenskO p6nve a ?ernouheln~ch lofisek mezi M(lnikem a Ben6tkami nad Jizerou. Sbor. geol. V6d, 1o~. Geol., Mineral., 22, 7-78. Praha. HRACH, S. et al. 1994. Mgensko-roudnick6 p~mev. Hodnoceni st6vajlcich geofyzik6lnich pracL MS Pfirodov~deck6 fakulty University Karlovy. Praha., Archiv Cesk6ho geologick6ho fistavu. Praha. IBRMAJER, J., SUK, M. et al. 1989. Geofyzikflni obraz CSSR. Ust~edni lJstav Geologick~. Praha. KLmANI, L., N~MEC, J. et al. 1994. V~skyt hoFlavOho zemniho plynu v uhelnf:ch slojlch karbonu Mgensko-roudnickd p6nve. MS Energie Kladno a.s., Archiv (~esk6ho geologick6ho fistavu. Praha.
423
KONE(~NY,P. et al. 1994. Studium mo~nosti identifikace stressovdho pole MYensko-roudnick~ pdmve. MS of the Institute of Geonics of Academy of Sciences of the Czech Republic, Ostrava, Archiv Cesk6ho geologick6ho flstavu. Praha. MAHEL, M., KODYM, O. & MALKOVSK'?, M. 1984. Tektonick6 mapa CSSR. Geologick~, I0stav Dion~,za St6ra. Bratislava. MALKOVSKY, M. et al. 1974. Geologie (eskd k(idovO p6nve a jejiho podlo~i. 0st~edni 0stav Geologick~. Praha. ZB,~NEK, J. et al. 1991. Zdtv(re(n6 zprfva vyhled6vaciho prdzkumu M ( l n l k - Benftky nad Jizerou. MS Geofond. Praha.
Method for estimating methane emissions from Polish coal mining IRENEUSZ WOJCIECH
G R Z Y B E K 1, L I D I A G A W L I K 2, SUWALA 2 & RYSZARD
KUZAK 3
1Polish State Mining Authority, ul. Poniatowskiego 31, 40-956 Katowice, Poland 2 Mineral and Energy Economy Research Centre, ul. Wybickiego 7 31-261 Krak6w, Poland 3 The Silesian University, Institute o f Earth Sciences, ul, B~dzihska 60, 41-200 Sosnowiec, Poland Abstract: To improve the accuracy of estimations of methane emissions from coal mining, a study of the Polish mine-specific method of methane measurement has been undertaken. To carry out the study, the following assumptions have been made: (1) Methane emissions are proportional to the gas content of coals; (2) The single mine average gas content is a function of both exploitation depth and gas distribution; (3) The total volume of methane released during mining is in proportion to the total gas content, while the volume emitted from post-mining processes and from waste rock storage to the residual gas content. The study has shown that: (1) There are four sources of methane emission from coal mining: mining and post-mining processes, degassing systems and barren rocks storage; (2) Emission from degassing systems should be measured directly, while the emissions from the other sources could be estimated and characterized by release factors; (3) The release factors for mining processes are described by parabolic equations, which have been defined by comparing total measured emissions from venting and degasing systems with in-situ methane content. For the remaining emission sources the factors are equal either to the average residual gas content or to the total gas content, depending on the proportion of residual to the total gas content; (4) Release factors and emissions, calculated for Polish coal mining were found to be four times less than had been previously suggested.
Methane is one of the most aggressive greenhouse gases and a major source of its emission is coal mining (Pyka 1993). Many attempts have been undertaken to evaluate the amount of methane emitted resulting from coal exploitation. The assumption and results of those works are described in detail in Smith & Sloss (1992) and in the report of U.S. Environmental Protection Agency (EPA 1994). One of the best known is the evaluation by Boyer II (Boyer et al. 1990), the results of which are the basis of the methodology of greenhouse gases emission evaluation approved by the Organization for Economic Cooperation and Development (OECD), during the Intergovernmental Panel on Climate Change (IPCC) in Paris (OECD 1991). This methodology in its basic form assumes a linear relation between in situ methane content (G) and the so called 'release factor' (W), which is a volume of methane emitted by a mass unit of coal exploited. This dependence is described by a regression equation in the form: W--- 2.04G + 8.16.
(1)
Where local data on the methane content are not available, the OECD/IPCC methodology suggests application of the global average release
factor (Win), which, multiplied by the amount of coal produced (Q), enables evaluation of the global methane emission (E):
E = WmQ.
(2)
At first, the global average release factor for underground coal mines, evaluated on the basis of American data, was assumed to be 27.1 m3/Mg (Boyer et al. 1990; OECD 1991). Then, after more data were analysed (CIAB 1992), the suggested release factor for high emission (for mining and post-mining processes together) was 29m3/Mg and for low emission 10.9m3/Mg (OECD 1991; EPA 1994). In all of the above mentioned cases, when the basic form of OECD/IPCC methodology (termed 'global-average method') was applied to Polish coal mines (Boyer et al. 1990; CIAB 1992; Pilcher et al. 1991; Radwafiski et al. 1991), the methane emission was overestimated. This was, inter alia, the reason for undertaking the development of the methodology of methane emission estimation on the basis of specific data for individual Polish coal mines. As a result the so-called 'mine-specific method' has been developed, and its theory and implications are presented in this paper.
From Gayer, R. & Pe~ek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 425-434.
426
I. GRZYBEK E T A L .
Basic assumptions and the data used in the study
(1)
In coal-bearing formations, methane is contained both in coal seams and in surrounding barren rocks. When coal is mined most of methane is released, but some is retained in the coal. The methane retained is termed residual gas content. In coal, methane is mainly sorbed; in barren rocks it exists as a free gas. In both cases the reservoir pressure is one of the principal agents in controlling gas content. The reservoir pressure can be assumed to be in continuity between the coal and the surrounding barren rocks (Tarnowski & Struzik 1978), so the amount of free gas in rocks is proportional to the methane content in coal. Some examples show that in Polish conditions the quotient of the methane volume contained in barren rocks (Vp) to the total volume of methane contained in the coal formation (V) is similar to the quotient of the residual methane content (Gr) to the total methane content (G) of coal: Vp V
Gr -
G
(3)
Coal exploitation causes a decline in reservoir pressure in the area influenced by underground workings. This results in methane emission from rocks and coal. The methane contained in rocks is released dynamically, with a rate that is proportional to the gradient of pressure between the underground opening and the unaffected rock mass. Methane release from coal is somewhat different. At an early dynamic stage, its release is the same as from barren rocks (Grzybek 1993a). When the pressure gradient approaches close to zero, the velocity of methane release depends on rate of diffusion through coal (Mazzsi 1992). Usually, dynamic release is very quick, while the diffusion is very slow. It was estimated that the complete diffusion of methane from in situ coal may take many years (Smith & Sloss 1992). In contrast, diffusion from crushed coal may only take tens of hours to tens of days (Gawraczyflski & Borowski 1986; Seidle & Arri 1990). The amount of methane that is released from a mass unit of coal in the process of diffusion is usually comparable with the residual methane content (e.g. Grzybek et al. 1994). Therefore, the amount of coal released in the dynamic process (Gd) corresponds with the difference between total and residual methane content: Gd = G - Gr.
(4)
Thus, in general, the following can be assumed.
(2)
(3)
(4)
(5)
Amount of methane emitted by a mass unit of coal produced, in all processes of coal mining, is proportional to the average methane content of exploited coal seams. Average methane content of exploited coal is a function of the depth-dependent distribution of methane content (Kotas 1994; Nie6 1993; Smith & Sloss 1992) and the depth of exploitation. The difference in the rates of methane release implies that during coal exploitation methane is released dynamically, whereas diffusion takes place after the mining process, from coal in post-mining processes and also from dispersed coal material in waste rock, mined together with coal. Emission of methane from mining processes per mass unit of coal produced is, therefore, proportional to the difference of the average total methane content and the average residual methane content. Emission of methane from post-mining activities and from waste rock heaps is proportional to the residual methane content.
In the light of the above assumptions, to evaluate the emission from Polish coal mining the following data were used: methane contents and residual methane contents of coal in Polish coal deposits, depths of coal exploitation in coal mines, amounts of coal produced, amounts of deposited waste rocks and the percentage of coal material in the waste rock heaps. The evaluation was compared with data on emission, capture and use of methane in coal mines. Data on the methane content and the rate of desorption - used to evaluate the residual methane content (see below) were gathered from 25% of total available boreholes spaced regularly in Polish coal basins (Fig. 1). Data were collected only for those boreholes where the laboratory tests on gas content were made by the same laboratory and where the methods employed were: two-phase vacuum degassing method (Kobiela et al. 1992) for methane content tests and rate of desorption measured in manometric desorbometer type DMC-2 (Grzybek 1993b). Data on coal production, the amount of methane emitted by ventilation shafts ('ventilation emission'), the amount of methane captured in degassing systems and the amount of methane used, and also the maximum and minimum depth of exploitation in each particular coal mine were collected for the years 1990-1992 directly from coal mines. Special attention was paid to the reliability of data concerning the
ESTIMATION METHOD FOR METHANE EMISSIONS
,/
f........
]
I
I
I - > ~ J "" -
~
~i"
(
""~.l.
~
I
P. Jl
\
I
'
/ -
.
/
f' I
/
"t
\
.P-~-- ~
i ~
427
,.., ~ t"
I
/I 0
*
i.
9
I0
'
1-/r
~
. "'~llloswu~c~M
,---.. "-"--<---__7
/
/I
Y \~-~-\
2-.,J
IIl
3-"
x_
~,j~.
"'-..
4-~/ "
5 - " ~ 6 " - / " ~i.
",, '
"~ 7-~"
Fig. 1. Location of boreholes chosen for analysis, location of coal mines and division of the Upper Silesian Coal Basin into regions: 1, main faults; 2, main overthrusts; 3, boreholes; 4-6, borders of: the basin (4), coal mines (5) and identified regions (6); 7, region number. ventilation emission. The data on amount of gas captured and used are fully reliable since each degassing system is attached to high quality measuring instruments; but the amount of methane released by ventilation shafts is usually calculated on the basis of measurements of air velocity in ventilation shafts and concentration of methane in the air, which are made only once a month. The accuracy of those calculations depends, primarily, on the accuracy of methane concentration measurements, which is 0.01%. Taking into account typical volumes of air that usually go through ventilation shafts of a single mine (on average about 16 x 109m 3 per year) the expected, average calculation error is about 0.8 x 106 m 3 of methane per year and mine. So, for the purpose of this study, data on ventilation emission for 1990-1992 were taken into account only from mines where the calculated emission was a minimum of two times the expected calculation error (i.e. no less than 1.6 x 106 m 3 per year). Other data necessary to develop the minespecific method of methane emission estimation were found and gathered from unpublished reports of research institutes and the Civil Service.
Programme of research and the results Bearing in mind the above mentioned assumptions, the following programme of research was followed: (1) (2)
(3) (4) (5)
(6)
(7) (8)
Identification of sources of methane emission in coal mining. Analysis of the natural conditions under which methane occurs and appropriate classification of parts of Polish coal basins into regions of uniform conditions. Depth standardization of the methane content distribution in each of the regions. Evaluation of the average methane content and the average residual methane content. Calculation of the specific emissions of those coal mines, in which ventilation emission had been assumed to be reliable. Correlation and regression analysis of the specific emissions and the average methane contents. Evaluation of the release factors from each source of emission. Calculation of the total emission from the coal mining system for the 1992.
428
I. GRZYBEK E T A L .
Sources of emission in coal mining The system of hard coal mining includes the following technological processes during which emission of methane may occur: (a) (b) (c) (d) (e)
The opening up of the coal deposit and associated development. Exploitation of coal, its haulage and lifting up to the surface. Preparation, storage, transport and crushing of coal prior to its final use. Degassing of the deposit before, during and after the exploitation. Storage of wastes (waste rock and preparation refuse).
The mining processes mentioned in points (a) and (b) are carried out underground. Methane released in the course of those processes and not captured by degassing systems is transferred to the atmosphere through mine ventilation systems. These ventilation emissions are thus the first source of methane emission. The remainder of the methane is contained in the extracted coal and extracted waste rock and is released in the course of post-mining processes mentioned in point (c) and at spoil heaps (point (e)). These are the second and the third sources of methane emissions respectively. The fourth source of methane emission is the degassing systems of coal mines. They use only part of the gas captured; the rest is released to the atmosphere. Emission of methane other than from the sources of hard coal production system mentioned above is likely to be very small.
Classification of coal basins into regions and standardization of the existing distribution of methane content On the basis of the geological data from Polish coal basins and available published material, a number of regions were distinguished, for which different gas conditions were expected. The regions in the Upper Silesian Coal Basin are shown in Fig. 1. In each region gas data were standardized to 100 metre depth intervals, defined in relation to the roof of coal-bearing Carboniferous formation, as well as (separately) to the depth at which methane content for the first time exceeds 4.5 m3/Mg of dry and ash free coal (excluding of values of methane content within zones of gas traps below the Carboniferous overburden; see: Kotas 1994). The choice of the methane content value can be justified by observed change in gas conditions for deposits
of methane content higher than 4.5m3/Mg (Nied 1993). In 100metre intervals identified in this way, the average and maximum methane contents were calculated for each region distinguished and standardized distributions of average and maximum methane contents were estimated (Fig. 2). The distributions for each region were compared with each other graphically (Fig. 3) and using statistical methods (Kolmogorov-Smirnov test, two-sample test, Spearman range correlation). As a result, on the basis of similarity, some regions could be merged. Further research was then carried out for each region.
Evaluation of the total and residual methane contents Evaluation of the average methane content (GK) was made for each coal mine, by comparing the standardized mines' exploitation depth interval with the standardized distribution of average methane content in the region where the coal mine is located:
~ikl
Gini
(5)
where: Gi is the average methane content of the ith 100-metre interval; ni is the number of measurements of methane content in the ith interval of the distribution and k is the number of intervals that lay in between minimum and maximum depth of exploitation in any given coal mine. The average residual methane contents were established for each region. The analytical method, described by Kandora & Grzybek (1992) was applied. The method is based on observed correlation between rate of desorption (AP2) and the total methane content (Fig. 4), where the coefficient a0 is equal to the residual methane content in the regression equation: G = a0 + al AP2.
(6)
The results of residual gas content calculations have been partially published (Grzybek et al. 1994).
Calculation of the specific emissions and release factors from each emission source Specific emissions (Wek) were calculated by dividing the sum of ventilation emission (Ew)
ESTIMATION METHOD FOR METHANE EMISSIONS
429
0
.
.
.II
i
.
.
4'4
" ..........
0
tF,,n
0
,,,.
,
,
,
'~. ":. '~.
[ ~ 'uoq.no jo ,loo~o~ ,.uo~!
o
=
9
=1
'
"~"
:l Ill ,,,il -t ~ nlumnow~nm|nnn! 9-
,
, ,.
,~,~.
!lmunnnn~lq||~,|l|~
9
. . . . .
:t .~lumnw-,~l~lmltq, n.,!!
0
P
"
8
m|muumn~qlU||t~n .,..
,
.
,
":. "~.":.
~~ ~
430
I. GRZYBEK E T A L .
b)
o) 0
4
O
12
18
0
G max
4 J
I
700'
*700'
300.
*300
-100,
-I00
-500
-SO0
-900
"900
z
Z
c)
d)
J
0
4
t
i
8
,
12
I
I
.10o
-200 -500
~2
!
|
18 G I ' 1 1 0 1 [ &
"~''"~v.
.,..~:. .......
lS Gmox
~ I00,
~
8
0
/,
8
J
l
I
12 ,
i
16 Gmox !
-200
~,...:::.:......
i
-50(>
.__(,,,.:..':~'~._ .,,:~..-~ 9 .,
-800
-800
-1100
/
"'--
-1100"
,. ......
III IV
.--- ~ . .....
V u
Fig. 3. Comparison of standarized distributions of maximum methane content (Gmax,m3/Mg), defined in relation to the roof of the Carboniferous (a & b), as well as to the depth at which methane content for the first time exceeds 4.5 m3/Mg (e & d); Z (m) is the relative depth, which is equal to zero for the roof of the Carboniferous (a & h) or for the depth where Gmax ~ 4.5 m3/Mg for the first time (e & d); numbers below the figure explain particular regions of the Upper Silesian Coal Basin.
plus the amount of methane captured by degassing systems (E0) of each coal mine (only reliable measurements) by the coal mine output (Q): Ew+E0 Wek - - - -
Q
(7)
According to the assumption (3) of the work, the sum given in the numerator of equation (7) is the amount of methane that is released from coal
and surrounding rocks dynamically and so, taking into account the equations (3) and (4), it is proportional to G. On this basis, many variants of regression between the average methane content (GK) and the specific emissions (Wek) were analysed, and finally the best regression equations were chosen (Fig. 5) - those characterized by the least error of ventilation emission estimation (5-15%). Estimated values of the specific emission were
ESTIMATION METHOD FOR METHANE EMISSIONS
431
2e
G
10
0"
0
20
40
60
80
100
120
140
160
180
200
Ap, Fig. 4. Distribution of rate of desorption (AP2, 1 • 10-3 m of H20) versus methane content (O, m3/Mg) in one of the Upper Silesia coal mines.
assumed to be the release factors from underground (mining) processes (We). The resulting regression equation has the form: We = aa Gg~ + a2G~
(8)
where: 9 for coal mines 10m3/Mg: al = 9 for coal mines 10m3/Mg: al =
of specific emission 14%k< 3.776 and a2 = -0.605; of specific emission Wek _> 21.452 and a2 = -3.346.
In further work the established equations were extrapolated to all coal mines, with an additional assumption that for coal mines of average methane content lower than the residual methane content of the region, the final release factor is twice as low as the one calculated by equation (8). Release factors from post-mining processes and from waste heaps were not estimated. It is
assumed that all residual methane is released from coal before its final use, so the release factors from post-mining activities are equal to: 9 residual mines in is higher 9 methane mines.
methane c o n t e n t - in those coal which the average methane content than the residual methane content; c o n t e n t - in the remaining coal
However the above assumption is not accurate, since there are no data to identity the volume of methane burned during the final use of the coal. The release factors for waste heaps were established taking additionally into account, that the calculated average content of dispersed coal material in waste rock is equal to 15% (comp. Bolewski & Gruszczyk 1989; PIG 1988). Release factors from degassing systems were also not estimated, as the amounts released to the atmosphere can be precisely measured in coal mines.
432
I. GRZYBEK E T A L .
a) t Wek
[m3/Mgl 5O !
9
4 0 84 Wt"
21.542 G - 3.346 G :
30
~ _ 1,-----,"__'A" _-----"--2"-7-:-'<-" :" ,--o
1
2
3
4
b) Wek
/
[rn3/Mgl
C,,,I,,,3/Mgl
/
/
8 7
Wj"
.
3.7"/6
G - 0.60S G 3
#
6
.
:
.'
"9
.
5 4-1
2 1 0 o
1
2
~'
---4
Fig. 5. Comparison of the linear equation by Boyer et aL (1990) and Polish parabolic equations ilustrating regression of specific emission (Wek, m 3/Mg) in relation to average methane content (GK, m 3/Mg) for Polish coal mines of high (a) and low (b) emissions; dots sign data from particular coal mines for 1990 -1992, except those where Wek < 1.6 x 106 m3/Mg.
Calculated release factors (in m3/Mg of coal produced) for each emission source are:
Evaluation of total methane emission from
9 for mining processes: 0,000-31.108 (average 6.005); 9 for post-mining processes: 0,000-1.907 (average 1.481); 9 for deposition of waste rock: 0,032-0.212 (average 0,065).
The estimated release factors for each source of methane emission were applied to equation (2) in place of the global average release factor, and the emission from Polish hard coal system was calculated. The amount of methane captured by degassing systems was subtracted from the
h a r d c o a l m i n i n g in P o l a n d
ESTIMATION M E T H O D FOR M E T H A N E EMISSIONS
emission calculated for mining processes. Total methane emission in 1992 was evaluated to be 935.2 x 106m 3, out of which 697.65 x 106m 3 was emitted from mining processes.
Conclusions The result of the presented research is the development of a method of methane emission evaluation. The method is adjusted to the specific conditions that exist in Polish coal mines. It is different from the global average method, suggested by OECD/IPCC, mainly due to the application of a parabolic equation that describes release factors for mining processes. The form of the equation is different for coal mines of high and low specific emission and proves that improvements introduced by OECD/IPCC in their methodology are in the right direction. Nevertheless, the modified OECD methodology suggests linear regression equations, which under Polish conditions would lead to improper evaluation of methane emission. In the light of this new method of Polish evaluation, the OECD methodology gives overestimated results. The methane emission for Polish hard coal mining systems, estimated by the method presented here is four times lower than the emission calculated on the basis of linear equations. Although the parabolic equation gives the best approximation of methane emissions from the whole Polish coal mining operation, it does not explain the emissions from each coal mine. It shows that emissions of methane are control led by unidentified factors, which are mine and, probably, coal basin specific. Therefore, the results of any global estimation of methane emissions need to be interpreted with caution.
References BOLEWSKI,A. & GRUSZCZYK,H. 1989. Geologia Gospodarcza. Wydawnictwa Geologiczne. Warszawa. BORER II C. M., KELEFANT,J. R., KUUSKRAA,V. A., MANGER, K. C. & KRUGER, D. 1990. Methane Emissions from Coal Mining." Issues and Opportunities for Reduction, U.S. Environmental Protection Agency Report, EPA[40019-90[O08. CIAB (Coal Industry Advisory Board) 1992. Global Methane Emissions from the Coal Industry. Coal Industry Advisory Board, Global Climate Committee/International Energy Agency Draft Report. EPA (Environmental Protection Agency) 1994. International Anthropogenic Methane Emissions. Estimates for 1990. U.S. Environmental Protection Agency Report, EPA 230-R-93-010.
433
GAWRACZYIqSKI, Z. & BOROWSKI, J. 1986. Zmiany zawartogci metanu w czasie w pr6bkach pobranych z urobionego w~gla. In: Metody rozpoznawania zagrokenia metanowego w kopalniach wcgla kamiennego. Katowice, czerwiec 1986. Katowickie Gwarectwo W~glowe- Zarz~d Oddziatu SITG w Katowicach, Conference Proceedings, 135-150. GRZYBEK, I. 1993a. Introduction to stimulation of coalbeds. Coalbed Methane Newsletter, 4-6, 10-14. 1993b. The Polish methods of the coalbed methane content testing and its reserves estimating. In: Proceedings of the 1993 International Coalbed Metane Symposium, Birmingham, Alabama, USA, May 17-21 1993, I, The University of Alabama, Tuscaloosa, 61-68. , GAWLIK, L., SUWALA,W. & KUZAK, R. 1994. Wst~pne wyniki implementacji analitycznej metody okreglania metanonognogci resztkowej w GZW. In: Wcgiel kamienny- wtasnoJci, akumulacja, uwalnianie i pozyskiwania gazdw kopalnianych, Krak6w, 11-12 paddiernik 1994. Akademia G6rniczo-Hutnicza w Krakowie, Workshop Proceedings, 7-9. KANDORA, P. & GRZYBEK, I. 1992. On the Criteria of the Possibility of Balancing and Exploitation of Coalbed Methane. United Nations Economic Comission for Europe Workshop on the Recovery and End-Use of Coal-Bed Methane, Katowice, Poland. Gt6wny Instytut G6rnictwa, Katowice. KOBIELA, Z., MODRZEJEWSKI,Z., SIMKA,A., SOBALA, E. • WRONA, B. 1992. The Methods of Sampling, Laboratory Methods of Determining the Gas Content in Coal Seams and Methods of Predicting Methane Emission in the Mine Workings Used in Poland. United Nations Economic Comission for Europe Workshop on the Recovery and End-Use of Coal-Bed Methane, Katowice, Poland. Gt6wny Instytut G6rnictwa, Katowice. KOTAS, A. (ed.) 1994. Coal-Bed Methane Potential of the Upper Silesian Coal Basin, Poland. Prace Paflstwowego Instytutu Geologicznego, CXLII, Warszawa. MAZZSI, D. 1992. Cavity Stress Relief Method for Recovering Methane from Coal Seam. United Nations Economic Comission for Europe Workshop on the Recovery and End-Use of Coal-Bed Methane, Katowice, Poland. Gt6wny Instytut G6rnictwa, Katowice. NIEC, M. 1993. Z[oza metanu w formacjach w~glonognych. In: Szkota Eksploatacji Podziemnej '93, Ustrofi 1-5marca 1993, Materiaty, 2. Centrum Podstawowych Problem6w Gospodarki Surowcami Mineralnymi i Energi~, Polska Akademia N a u k - Akademia G6rniczo-Hutnicza, Krak6w, 281-301. OECD (Organization for Economic Cooperation and Development) 1991. Estimation of Greenhouse Gas Emissions and Sinks. OECD Experts' Meeting, Paris, France, 18-21 February 1991 Final Report for Intergovernmental Panel on Climate Change.
434
I. GRZYBEK E T A L .
PIG (Pafistwowy Instytut Geologiczny) 1988. Zestawienie opracowanych technologii i metod wykorzystania odpad6w. Polish Geological Institute Report of CPBP 04.10.04 Subprogramme, Warszawa. PILCHER, R. C., BIBLER, C. J., GLICKERT, R., MACHESKY, L. & WILLIAMS, J. M. 1991. Assessment of the Potential for Economic Development and Utilization of Coalbed Methane in Poland. U.S. Environmental Protection Agency Report,
EPA[40011-911032. PYKA, M. 1993. About the energy of methane. Silesian Coalbed Methane Newsletter, 4-6, 18-26.
RADWAIqSKI, E., SKOWRO/QSKI,P. & TWAROWSKI,A. 1991. Uwarunkowania inwentaryzacji emisji i wychwytu gaz6w cieplarnianych w Polsce w 1988 roku. Polish Foundation for Energy Efficiency Report. SEIDLE, J. P. & ARRI, L. E. 1990. Use of Conventional Reservoir Models for Coalbed Methane Simulation. International Meeting of the Petroleum Society of CIM and SPE, Calgary, CIM/SPE 90. SMITH, I. M. & SLOSS, L. L. 1992. Methane Emissions from Coal. Perspectives, IEA Coal Research. TARNOWSKI,J. & STRUZIK,A. 1978. Opracowanie pola gazono~no~ci projektowanego obszaru g6rniczego. Przeglqd G6rniczy, 3, 95-105.
Methane emission and its utilization from Ostrava-Karvinfi Collieries in the Upper Silesian coal basin, Czech Republic G. T A K L A
& Z. V A V R U S , ~ K
Ddlni prdzkum a bezpednost Paskov, a.s. (Underground Exploration and Mine Safety, Inc.) 73921 Paskov, The Czech Republic Abstract: The Ostrava-Karvinfi part (OKR) of the Upper Silesian coal basin is the most
important hard coal basin in the Czech Republic. This basin extends over an area of 1600km2. The Carboniferous basin fill contains 255 seams with a net coal thickness of 150m. The methane content in the coal is estimated between 4.4 to 20m3/t. In the OKR, about 120million cubic meters per year of methane are produced by gas drainage plants from a number of collieries. The methane is used in local industries. Several measures have been taken to increase the methane extraction in active mines and in areas of closed collieries. Additionally, pilot projects for coalbed methane production have been started in virgin coal bearing areas with no previous coal production. The authors' company is progressing with such a pilot project, with the first coalbed methane wells completed recently. Results to date are encouraging. All these activities are designed to increase the safety in mines, to reduce methane emission to the atmosphere, and to develop new sources of energy. The hard coal reserves in the Czech Republic are found in one major and three minor coal mining districts areas: Ostrava-Karvinfi, Northeast Bohemia, Kladno and in West Bohemia. The Ostrava-KarvinS, Coalfield (OKR) contains the most significant of the coal deposits. It covers an area of about 1600 km 2, within which about 335 km 2 belong to the mining claims of individual collieries. Hard coal production in the O K R represents more than 95% of the total hard coal production in the Czech Republic. More than 99.8% of methane emission from mines in the Czech Republic is related to mining operations in the OKR. Coal mining in the O K R started about 200 years ago. The earliest record of the quantity of methane drained by mine ventilation relates to the year 1910 when the coal production increased to 7.67 million metric tons per year, accompanied
by a release of methane averaging 589,000 cubic metres per day. Historical data on coal production and total methane released in the O K R are summarized in Table 1. In the course of the advance of mine mechanization and increase of coal mining intensity, it became necessary to introduce a system of methane drainage to isolate part of the methane liberated during coal mining, for safety reasons. In some longwalls, applying this method, it is possible to reduce the amount of methane released into the atmosphere by up to 50% or more. During the 35 years of operating mine methane drainage, an efficiency of about 30% of all the methane released has been reached. This means that about 30% of the total amount of methane released from mining operations in the whole mining district is collected and drawn into the methane drainage plants.
Table 1~
Coalproduction in OKR and methane exhalation
Year
Coal production (106 t/year)
Methane content in outlet shafts (thousands m3/day)
Methane quantity drained (thousands m3/day)
Total methane recovery (thousands m3/day)
1910 1920 1930 1940 1950 1960 1970 1980 1990 1994
7.67 7.59 10.67 16.25 13.72 20.87 23.86 24.69 20.06 15.80
589 856 873 869 910 957 1526.4 1381.4 999.6 743.32
-
589 856 863 869 910 989.9 2080.1 1949.7 1379.1 1022.6
32.9 553.7 568.3 379.7 279.4
From Gayer, R. & Pegek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 435-440.
436
G. TAKLA & Z. VAVRUS/~K
However, due to leakages in rock strata, imperfection in the conductor casing strings of boreholes etc., some dilution of the mine gas occurs. Because the whole system works below atmospheric pressure, some air may be sucked in, so that the final methane concentration becomes reduced to about 50-55%. Within a short period of time, after further development of the mine gas drainage network, ways of using this source of energy were investigated. In most mines heating plants were modified in such a way to enable them to use mine gas for heating and for the production of hot water. For the remaining amount of mine gas, which reached up to 200 million cubic metres per year, gas pipelines were laid, in collaboration with regional gas utilities. Also, the use of methane in local industry was established, mainly in metallurgical plants and power stations.
Overview of the present state At present, the mine gas drainage system in the O K R consists of 21 individual mine gas drainage plants with a total of 115 vacuum pumps installed. An extensive underground pipeline network connects about 5000 mine gas sources,
mainly mine boreholes to this system. In addition to these boreholes, abandoned mine and 'oldman' workings are connected to the system as well, separated from active mine workings by seals. From individual mine gas drainage plants, pipelines feed into the main pipeline system. The gas mixture collected is then Oelivered to consumers in local industry, mainly heating plants and steelworks. The gas delivered consists of a mixture of methane with air, with a methane content of about 50-55%. The equipment used to burn this gas has to be designed to account of this methane/air mixture. The modification means that the equipment cannot be used for a different type of gas. The typical gas mixture consists of 50-55% of methane, 2-4% of carbon dioxide, 1-3% of oxygen, and 38-47% of nitrogen. The heating value of such gas is about 19-20 Megajoule per cubic metre. The present gas pipeline network forms a system of about 92km total length. Figure 1 shows a schematic drawing of the gas pipeline system and mine gas drainage plants. Table 2 gives an overview on the methane production from mine gas drainage plants and on the consumption of gas within the mines themselves or in the regional industry. Due to the recent reduction in the intensity of coal mining and improvements in the efficiency
mNOV~- HU~" ..~ KARVIN,A
9
OOUBRAVA CS~A
9
/ -I -
... .....
....
I .......... ~ I ~ " ........ FUC~Ky " " " ' OOR*'vRo~./'"" .......... /G~.O.OUKLA ......
,.o0" '~
'~
"
/
/
\
M.
X,~.~....~csH
"~
e ......... SVOBOOA
,v,
ov
@OSTRAVA --OS TRAVA 9 ~]1
' "EXPLANATION
GA$ORAINAGEPLANT N. METALLURGICAL PLANT ~
@ ........
.............
COALBEOt,,IETHANEPIPELINE
"
/ J
/
mm
PASKO~
I"
I
T ~
. ,.
STARi'C
III.
PASKOV
P I P E L I H E FOR COKE OVEN GAS
Fig. 1. Schematic drawing of pipeline system and mine gas drainage plants in OKR, Czech Republic.
METHANE EMISSION FROM OKR MINES
Table 2. Methane production from mine gas drainage and its utilization (in thousands m3/year)
Year Total Own Deliveries to Vented production utilization industry 1958 1960 1965 1970 1975 1980 1985 1990 1994
4300 14300 76233 239713 235700 225028 207459 150395 110324
542 41 617 61 158 69843 64230 65023 52441
2100 12100 61 966 127 907 168 970 136533 120261 69085 44977
2200 2200 13 725 70 189 5571 18 652 22969 16287 12 905
of mine ventilation systems, the importance of mine gas drainage for the safety of mine operations is gradually decreasing. In view of this development our company DPB, Inc. has established new methods for collecting methane from active mines. Based on a detailed study of geological data, new locations and rock sequences were explored, where methane could have accumulated in greater quantities. At such locations specially designed and completed mine boreholes are drilled, leading to increases in total coalbed methane production. Locations favourable for methane production could be found in the upper part of mine claims, in weathered Carboniferous strata, in faulted intervals, and also in areas with unconsolidated rocks covering coal mine areas. According to results up to date, a development of such opportunities could considerably increase mine gas production from gas drainage plants in active mines. In the mine CSM in Stonava for example, four mine boreholes of up to 150 m depth were drilled into beds close to the top of the Carboniferous. After a few months of operation, production of methane from these holes rose to up to 12 000 cubic metres per day. This new method is termed 'Supplementary Mine Gag Drainage', because, unlike traditional mine gas drainage, it is not concerned mainly with mine safety. Within less than two years of introducing this method the production from 16 boreholes with a total length of 2563 m has reached 7.6 • 106 m 3 of methane. In addition to conventional gas drainage in coal mines, DPB has also been involved in two fundamental development programmes for exploration, production and utilization of coalbed methane from 9 abandoned coal mines 9 virgin coal seams
437
These programmes are based on a study initiated by the US Environmental Protection Agency (EPA), Global Change Division. This study recommended an integrated approach to coalbed methane recovery in the OKR.
Gas recovery from abandoned coal mines Due to the political and social changes since 1989/1990 in the Czech Republic, coal production has decreased as the coal industry was restructured and unprofitable coal mines were closed down or are being closed. In the Ostrava part of the mining district of the OKR, four collieries have been closed. Mine claims over these collieries cover an area of more than 100 square kilometres extending partly over very densely populated parts of the city of Ostrava. After abandonment operations, the hoisting and outlet shafts are being filled in. The ventilation systems, that diluted and vented the mine gases, were shut down after more than 200 years of mining operations. Water levels in the mines would rise as predicted, gradually flooding the mine workings to the expected level. During this stage some mine gas may leak, out of control, to the surface via old mine shafts and other potential leaks in their vicinity, and also via natural migration channels, such as fissures and tectonic faults. This migration is also influenced by rapid drops in atmospheric pressure which lead to a pressure differential between the gas in the mine workings and the atmosphere above. A risk of explosion or fire would occur, with such gas possibly finding its way into basements of buildings or industrial plants in the densely populated city of Ostrava. At three former mine sites, mine gas drainage systems and plants had been in operation at the surface before the decision to close the mines was taken. The authors' company, DPB, Inc., prepared a study, showing the necessity to maintain mine gas drainage from shut-down mines even after the termination of mining in the OKR. For this reason the company bought mine gas drainage plants from two mines before their closure. DPB is now converting these plants for the permanent extraction of mine gas from closed mine workings after the termination of coal mining in the area. Before the mine shafts were filled in, boreholes were drilled at mine locations noted for their maximum presence of methane. These boreholes were connected to an existing pipeline. Precautions had to be taken to avoid damage of the mine gas pipeline while the mine shafts were filled in. Some of the abandoned shafts were selected to adapt them as gas
438
G. TAKLA & Z. VAVRUSAK
production shafts, by filling in their upper section only and leaving the bottom parts open and connected to former mine working levels, so that those could act as a reservoir for mine gas. In the upper filled part of such shafts, pipes had to be installed so that the mine gas collected could be pumped into the mine gas drainage plant. DPB believes that mine gas of a quality sufficient for industrial use can be pumped out from closed mines of the Ostrava mining district. The methane content of the gas mixture produced varies from 40 to 80% but an increase in methane content can be expected once the filling of the abandoned shafts is completed. A similar situation to the case described above caused DPB in 1992/93 to put its first surface gob wells in operation in an area of abandoned coal mines, for the purpose of verifying: 9 the possibility of recovering gas from abandoned coal mines, 9 the technical and technological requirements for drilling, completing and testing such boreholes, and 9 the possibility of commercial use of this gas. Gob wells in our project are wells extracting methane released from strata surrounding coalbeds being mined and captured in the rock massive in micropore-macropore structures and fractures. The gob wells were located in abandoned mine workings of Vrbice and Rychvald in the Ostrava part of the OKR, where active mining was already discontinued in the 1950s. These wells were drilled into the mined multiple coal seams of the Ostrava Formation. The design and technical performance of gob wells fulfilled the following requirements: 9 Low drilling and completion costs in comparison to conventional gas wells, to ensure profitability despite relatively low production rates. 9 Prevention of contamination of gob zones by drilling fluids. 9 Maintenance of safety standards in drilling and production in accordance to regulation issued by the Czech Bureau of Mines. The gob wells were tested to yield fundamental information on production pressure, volume of gas production and methane concentration, to allow estimates of profitability and to provide a basis for planning to tie in the wells into the production system of gas drainage plants already existing.
The wellhead pressure of our gob wells varied between 101.21-101.62kPa, depending on barometric pressure. The technique of well testing was therefore modified in such a way to take this peculiarity into account. A mobile vacuum pump was constructed by modifying a conventional vacuum pump so as to create the underpressure required to stimulate the flow of gas from the gob zones.The vacuum unit consisted of a vacuum pump, a measuring line allowing gas sampling, and an electric motor. The results of a short-term production test at the V-3 gob well are given in Figure 2. After the evaluation results of the short-term production tests, a flow line was laid connecting the boreholes with the existing gas drainage plants. Figure 3 displays the characteristics of the gas production from the V-3 well during its first year of operation. The first pilot gob wells of DPB demonstrated the feasibility of gob gas extraction from abandoned coal mines.
Coalbed methane development in virgin coal seams Coalbed methane (CBM) development in virgin coal seams in the Czech Republic dates back to early 1992. This development was encouraged for several reasons: 9 The occurrence of methane in active mines of the OKR. 9 The high density of previous exploration throughout the OKR region, including exploration wells, evaluation and assessment of coal reserves, studies of coal deposits, etc. 9 The government policy for the diversification of gas resources. 9 The interest of the Czech Government and of private companies regarding the exploration and production of coalbed methane. According to Czech Mining Law, the area of virgin coal seams in OKR was divided among four CBM concessionaires 9 D~lni prfizkum a bezpe6nost Paskov, a.s. (DPB, Inc.) 9 Energie Kladno, a.s. 9 Geologick~ prflzkum Ostrava, a.s. (GPO, Inc.) 9 Unigeo Ostrava, a.s. A permit to explore for CBM was granted to all four companies. According to Czech Mining Law, the companies may expect to obtain a production concession in the case of successful results.
M E T H A N E EMISSION F R O M O K R MINES
I
+oo
~-3 ~oa WELL
~+.o.,,,,. ,,,oouc,.o..,,.
439
II
+o_
.
150 - - - ; - - - / / i
.
.
.
IS"
.
1
2
.
.
m~--~-~.o__.__...~ , P-
. . .
.
3
.
.
4
WELLHEAD PRESSURE
IkP-I
"EJ9 UNDERPRESSURE
~p.]
.
5
.
+, ~ _ = . . < d',--W_~l .....
.
6
7
8
10
9
:
CH4 C O N C E N T R A T I O N
I""I -~-" ~ P R O D U C T I O N PERIOD Imd'n~leel
- " ~ - GOB GAS PRODUCTION
l"~'ml
_
Fig. 2. Short-term production test of the V-3 gob well.
IGV-3 GOB WELL AS P R O D U C T I O N
Iio0. sc,,,It,J 100000
X
IIECALCULATEO TO 100% CH4
10000 -
X
if(
, . , . ~ ~
C:l----lil
--~" - " ' - +K-+-----I~----..#+_.+--~.--..~,~ .,.
1000 "%'~,:
.....
X:
100 jl.
10
.
. . . .
DAYS
_
.
.
. . . . . ~ . ~/.:.4,+.+ " ~-~-wj - w j ~ - - - ' - m -. - -. m. . . . . ..
0 -
.
I. . . .
,oo....._.+//_. 50
.
i"
.jk...~
'
' i
9
__"
%"
.I.
"
.I.
.t..
I
....%'-
!
11=
.
.Jl..
.
.
.
"+
.
,lb.
.
7/93 0/93 9/93 10;931 lj93121931194 2/94 3/94 4/94 5/94 6/94' 71~14 MONTHS '
- - , k - AVE.CI-14 C O N C E N T R .
I~1 -I'~- TOTAL CH4 PRODUCT.
I ]
'
'
- - I - - AVE.UNDERPRESSURE
I'+e'l
- ~ . ' - A V E . C l I 4 PRODUCTION
Fig. 3. Gas production of the V-3 gob well during first year of production.
I,,,,~/,+od
440
G. TAKLA & Z. VAVRUSAK
According to the energy resource policy of the Czech Government, CBM exploration is financially supported by the Ministry of Economy. By the end of 1994, nine CBM wells were drilled. Two of the wells were fractured, with promising results for future production. In April/May 1995, more four wells were fractured and long-term pumping tests were started. Four to six new wells are expected to be drilled by the end of 1995. DPB owns a concesion for CBM exploration in ten prospection areas, covering a total of 240km 2 of virgin coal fields that contain 4.1 • 10 9 tonnes of hard coal reserves. These areas are characterized by a high density of exploration wells, drilled for coal, gas and water prospection on a spacing of some 1.5 km. Several geological evaluations and studies were carried out to assess the importance of the coal reserves and to build a database. Accordingly, the coal deposits of these areas are well documented, with very good geological information about each of the prospection areas. However, the occurrence of methane in virgin coal seams is not well documented so far. For most of the prospection areas, data on gas content of the coal are either not available at all, or very incomplete at best. For this reason, a CBM exploration campaign with a chance of success would first have to identify the most favourable areas and coal seams by drilling pilot wells. The CBM pilot project proposed by DPB, comprises six pilot test wells in the most promising exploration areas, which are named as follows: 9 9 9 9
D6tmarovice Petrovice - Vficlavovice Pf'ibor - zfipad T r o j a n o v i c e - Fren~t~it.
Each of the wells is continuously cored in the Carboniferous interval to obtain primarily information on the gas content in the rock sequence. Core samples of the coal are subjected to standard canister desorption tests (according to the Standard of the US Bureau of Mines). Simultaneously, coals and rocks surrounding the coal sampling points are tested to obtain information required for the design and evaluation of hydraulic fracturing of the coal. Drilling, coring and completion are assigned to local contractors, as is hydraulic fracturing of the coal. The first pilot well was completed in the Karvinfi Formation's coal seams in the area of Ditmarovice at the begining of 1995. Successful hydrofracturing opened two coal seams (4.3m and 1.4 m thick). A beam pump was installed to test water and gas production. The second pilot well in the Vficlavovice area (Ostrava Formation's coal seams) did not realise the expected coal gas content and gas in place. The third pilot well of DPB' s project is now being drilled in the area of Fren~tfit (Karvinfi Formation). An additional one or two pilot wells are planned for spudding in 1995.
Summary The paper presents information on some methods of coal gas processing and utilization in the largest hard coal basin of the Czech Republic., i.e. the Ostrava-Karvinfi Mining District. Improvements in the methods of mine gas drainage, increased recovery of methane from closed mines, and new coalbed methane production from coal fields of the region could contribute to the worldwide aim at improving the use of clean energy sources and reducing the emission of methane into the atmosphere.
Methane emission and its utilization from Ostrava-Karvinfi Collieries in the Upper Silesian coal basin, Czech Republic G. T A K L A
& Z. V A V R U S , ~ K
Ddlni prdzkum a bezpednost Paskov, a.s. (Underground Exploration and Mine Safety, Inc.) 73921 Paskov, The Czech Republic Abstract: The Ostrava-Karvinfi part (OKR) of the Upper Silesian coal basin is the most
important hard coal basin in the Czech Republic. This basin extends over an area of 1600km2. The Carboniferous basin fill contains 255 seams with a net coal thickness of 150m. The methane content in the coal is estimated between 4.4 to 20m3/t. In the OKR, about 120million cubic meters per year of methane are produced by gas drainage plants from a number of collieries. The methane is used in local industries. Several measures have been taken to increase the methane extraction in active mines and in areas of closed collieries. Additionally, pilot projects for coalbed methane production have been started in virgin coal bearing areas with no previous coal production. The authors' company is progressing with such a pilot project, with the first coalbed methane wells completed recently. Results to date are encouraging. All these activities are designed to increase the safety in mines, to reduce methane emission to the atmosphere, and to develop new sources of energy. The hard coal reserves in the Czech Republic are found in one major and three minor coal mining districts areas: Ostrava-Karvinfi, Northeast Bohemia, Kladno and in West Bohemia. The Ostrava-KarvinS, Coalfield (OKR) contains the most significant of the coal deposits. It covers an area of about 1600 km 2, within which about 335 km 2 belong to the mining claims of individual collieries. Hard coal production in the O K R represents more than 95% of the total hard coal production in the Czech Republic. More than 99.8% of methane emission from mines in the Czech Republic is related to mining operations in the OKR. Coal mining in the O K R started about 200 years ago. The earliest record of the quantity of methane drained by mine ventilation relates to the year 1910 when the coal production increased to 7.67 million metric tons per year, accompanied
by a release of methane averaging 589,000 cubic metres per day. Historical data on coal production and total methane released in the O K R are summarized in Table 1. In the course of the advance of mine mechanization and increase of coal mining intensity, it became necessary to introduce a system of methane drainage to isolate part of the methane liberated during coal mining, for safety reasons. In some longwalls, applying this method, it is possible to reduce the amount of methane released into the atmosphere by up to 50% or more. During the 35 years of operating mine methane drainage, an efficiency of about 30% of all the methane released has been reached. This means that about 30% of the total amount of methane released from mining operations in the whole mining district is collected and drawn into the methane drainage plants.
Table 1~
Coalproduction in OKR and methane exhalation
Year
Coal production (106 t/year)
Methane content in outlet shafts (thousands m3/day)
Methane quantity drained (thousands m3/day)
Total methane recovery (thousands m3/day)
1910 1920 1930 1940 1950 1960 1970 1980 1990 1994
7.67 7.59 10.67 16.25 13.72 20.87 23.86 24.69 20.06 15.80
589 856 873 869 910 957 1526.4 1381.4 999.6 743.32
-
589 856 863 869 910 989.9 2080.1 1949.7 1379.1 1022.6
32.9 553.7 568.3 379.7 279.4
From Gayer, R. & Pegek, J. (eds), 1997, European Coal Geology and Technology, Geological Society Special Publication No. 125, pp. 435-440.
436
G. TAKLA & Z. VAVRUS/~K
However, due to leakages in rock strata, imperfection in the conductor casing strings of boreholes etc., some dilution of the mine gas occurs. Because the whole system works below atmospheric pressure, some air may be sucked in, so that the final methane concentration becomes reduced to about 50-55%. Within a short period of time, after further development of the mine gas drainage network, ways of using this source of energy were investigated. In most mines heating plants were modified in such a way to enable them to use mine gas for heating and for the production of hot water. For the remaining amount of mine gas, which reached up to 200 million cubic metres per year, gas pipelines were laid, in collaboration with regional gas utilities. Also, the use of methane in local industry was established, mainly in metallurgical plants and power stations.
Overview of the present state At present, the mine gas drainage system in the O K R consists of 21 individual mine gas drainage plants with a total of 115 vacuum pumps installed. An extensive underground pipeline network connects about 5000 mine gas sources,
mainly mine boreholes to this system. In addition to these boreholes, abandoned mine and 'oldman' workings are connected to the system as well, separated from active mine workings by seals. From individual mine gas drainage plants, pipelines feed into the main pipeline system. The gas mixture collected is then Oelivered to consumers in local industry, mainly heating plants and steelworks. The gas delivered consists of a mixture of methane with air, with a methane content of about 50-55%. The equipment used to burn this gas has to be designed to account of this methane/air mixture. The modification means that the equipment cannot be used for a different type of gas. The typical gas mixture consists of 50-55% of methane, 2-4% of carbon dioxide, 1-3% of oxygen, and 38-47% of nitrogen. The heating value of such gas is about 19-20 Megajoule per cubic metre. The present gas pipeline network forms a system of about 92km total length. Figure 1 shows a schematic drawing of the gas pipeline system and mine gas drainage plants. Table 2 gives an overview on the methane production from mine gas drainage plants and on the consumption of gas within the mines themselves or in the regional industry. Due to the recent reduction in the intensity of coal mining and improvements in the efficiency
mNOV~- HU~" ..~ KARVIN,A
9
OOUBRAVA CS~A
9
/ -I -
... .....
....
I .......... ~ I ~ " ........ FUC~Ky " " " ' OOR*'vRo~./'"" .......... /G~.O.OUKLA ......
,.o0" '~
'~
"
/
/
\
M.
X,~.~....~csH
"~
e ......... SVOBOOA
,v,
ov
@OSTRAVA --OS TRAVA 9 ~]1
' "EXPLANATION
GA$ORAINAGEPLANT N. METALLURGICAL PLANT ~
@ ........
.............
COALBEOt,,IETHANEPIPELINE
"
/ J
/
mm
PASKO~
I"
I
T ~
. ,.
STARi'C
III.
PASKOV
P I P E L I H E FOR COKE OVEN GAS
Fig. 1. Schematic drawing of pipeline system and mine gas drainage plants in OKR, Czech Republic.
METHANE EMISSION FROM OKR MINES
Table 2. Methane production from mine gas drainage and its utilization (in thousands m3/year)
Year Total Own Deliveries to Vented production utilization industry 1958 1960 1965 1970 1975 1980 1985 1990 1994
4300 14300 76233 239713 235700 225028 207459 150395 110324
542 41 617 61 158 69843 64230 65023 52441
2100 12100 61 966 127 907 168 970 136533 120261 69085 44977
2200 2200 13 725 70 189 5571 18 652 22969 16287 12 905
of mine ventilation systems, the importance of mine gas drainage for the safety of mine operations is gradually decreasing. In view of this development our company DPB, Inc. has established new methods for collecting methane from active mines. Based on a detailed study of geological data, new locations and rock sequences were explored, where methane could have accumulated in greater quantities. At such locations specially designed and completed mine boreholes are drilled, leading to increases in total coalbed methane production. Locations favourable for methane production could be found in the upper part of mine claims, in weathered Carboniferous strata, in faulted intervals, and also in areas with unconsolidated rocks covering coal mine areas. According to results up to date, a development of such opportunities could considerably increase mine gas production from gas drainage plants in active mines. In the mine CSM in Stonava for example, four mine boreholes of up to 150 m depth were drilled into beds close to the top of the Carboniferous. After a few months of operation, production of methane from these holes rose to up to 12 000 cubic metres per day. This new method is termed 'Supplementary Mine Gag Drainage', because, unlike traditional mine gas drainage, it is not concerned mainly with mine safety. Within less than two years of introducing this method the production from 16 boreholes with a total length of 2563 m has reached 7.6 • 106 m 3 of methane. In addition to conventional gas drainage in coal mines, DPB has also been involved in two fundamental development programmes for exploration, production and utilization of coalbed methane from 9 abandoned coal mines 9 virgin coal seams
437
These programmes are based on a study initiated by the US Environmental Protection Agency (EPA), Global Change Division. This study recommended an integrated approach to coalbed methane recovery in the OKR.
Gas recovery from abandoned coal mines Due to the political and social changes since 1989/1990 in the Czech Republic, coal production has decreased as the coal industry was restructured and unprofitable coal mines were closed down or are being closed. In the Ostrava part of the mining district of the OKR, four collieries have been closed. Mine claims over these collieries cover an area of more than 100 square kilometres extending partly over very densely populated parts of the city of Ostrava. After abandonment operations, the hoisting and outlet shafts are being filled in. The ventilation systems, that diluted and vented the mine gases, were shut down after more than 200 years of mining operations. Water levels in the mines would rise as predicted, gradually flooding the mine workings to the expected level. During this stage some mine gas may leak, out of control, to the surface via old mine shafts and other potential leaks in their vicinity, and also via natural migration channels, such as fissures and tectonic faults. This migration is also influenced by rapid drops in atmospheric pressure which lead to a pressure differential between the gas in the mine workings and the atmosphere above. A risk of explosion or fire would occur, with such gas possibly finding its way into basements of buildings or industrial plants in the densely populated city of Ostrava. At three former mine sites, mine gas drainage systems and plants had been in operation at the surface before the decision to close the mines was taken. The authors' company, DPB, Inc., prepared a study, showing the necessity to maintain mine gas drainage from shut-down mines even after the termination of mining in the OKR. For this reason the company bought mine gas drainage plants from two mines before their closure. DPB is now converting these plants for the permanent extraction of mine gas from closed mine workings after the termination of coal mining in the area. Before the mine shafts were filled in, boreholes were drilled at mine locations noted for their maximum presence of methane. These boreholes were connected to an existing pipeline. Precautions had to be taken to avoid damage of the mine gas pipeline while the mine shafts were filled in. Some of the abandoned shafts were selected to adapt them as gas
438
G. TAKLA & Z. VAVRUSAK
production shafts, by filling in their upper section only and leaving the bottom parts open and connected to former mine working levels, so that those could act as a reservoir for mine gas. In the upper filled part of such shafts, pipes had to be installed so that the mine gas collected could be pumped into the mine gas drainage plant. DPB believes that mine gas of a quality sufficient for industrial use can be pumped out from closed mines of the Ostrava mining district. The methane content of the gas mixture produced varies from 40 to 80% but an increase in methane content can be expected once the filling of the abandoned shafts is completed. A similar situation to the case described above caused DPB in 1992/93 to put its first surface gob wells in operation in an area of abandoned coal mines, for the purpose of verifying: 9 the possibility of recovering gas from abandoned coal mines, 9 the technical and technological requirements for drilling, completing and testing such boreholes, and 9 the possibility of commercial use of this gas. Gob wells in our project are wells extracting methane released from strata surrounding coalbeds being mined and captured in the rock massive in micropore-macropore structures and fractures. The gob wells were located in abandoned mine workings of Vrbice and Rychvald in the Ostrava part of the OKR, where active mining was already discontinued in the 1950s. These wells were drilled into the mined multiple coal seams of the Ostrava Formation. The design and technical performance of gob wells fulfilled the following requirements: 9 Low drilling and completion costs in comparison to conventional gas wells, to ensure profitability despite relatively low production rates. 9 Prevention of contamination of gob zones by drilling fluids. 9 Maintenance of safety standards in drilling and production in accordance to regulation issued by the Czech Bureau of Mines. The gob wells were tested to yield fundamental information on production pressure, volume of gas production and methane concentration, to allow estimates of profitability and to provide a basis for planning to tie in the wells into the production system of gas drainage plants already existing.
The wellhead pressure of our gob wells varied between 101.21-101.62kPa, depending on barometric pressure. The technique of well testing was therefore modified in such a way to take this peculiarity into account. A mobile vacuum pump was constructed by modifying a conventional vacuum pump so as to create the underpressure required to stimulate the flow of gas from the gob zones.The vacuum unit consisted of a vacuum pump, a measuring line allowing gas sampling, and an electric motor. The results of a short-term production test at the V-3 gob well are given in Figure 2. After the evaluation results of the short-term production tests, a flow line was laid connecting the boreholes with the existing gas drainage plants. Figure 3 displays the characteristics of the gas production from the V-3 well during its first year of operation. The first pilot gob wells of DPB demonstrated the feasibility of gob gas extraction from abandoned coal mines.
Coalbed methane development in virgin coal seams Coalbed methane (CBM) development in virgin coal seams in the Czech Republic dates back to early 1992. This development was encouraged for several reasons: 9 The occurrence of methane in active mines of the OKR. 9 The high density of previous exploration throughout the OKR region, including exploration wells, evaluation and assessment of coal reserves, studies of coal deposits, etc. 9 The government policy for the diversification of gas resources. 9 The interest of the Czech Government and of private companies regarding the exploration and production of coalbed methane. According to Czech Mining Law, the area of virgin coal seams in OKR was divided among four CBM concessionaires 9 D~lni prfizkum a bezpe6nost Paskov, a.s. (DPB, Inc.) 9 Energie Kladno, a.s. 9 Geologick~ prflzkum Ostrava, a.s. (GPO, Inc.) 9 Unigeo Ostrava, a.s. A permit to explore for CBM was granted to all four companies. According to Czech Mining Law, the companies may expect to obtain a production concession in the case of successful results.
M E T H A N E EMISSION F R O M O K R MINES
I
+oo
~-3 ~oa WELL
~+.o.,,,,. ,,,oouc,.o..,,.
439
II
+o_
.
150 - - - ; - - - / / i
.
.
.
IS"
.
1
2
.
.
m~--~-~.o__.__...~ , P-
. . .
.
3
.
.
4
WELLHEAD PRESSURE
IkP-I
"EJ9 UNDERPRESSURE
~p.]
.
5
.
+, ~ _ = . . < d',--W_~l .....
.
6
7
8
10
9
:
CH4 C O N C E N T R A T I O N
I""I -~-" ~ P R O D U C T I O N PERIOD Imd'n~leel
- " ~ - GOB GAS PRODUCTION
l"~'ml
_
Fig. 2. Short-term production test of the V-3 gob well.
IGV-3 GOB WELL AS P R O D U C T I O N
Iio0. sc,,,It,J 100000
X
IIECALCULATEO TO 100% CH4
10000 -
X
if(
, . , . ~ ~
C:l----lil
--~" - " ' - +K-+-----I~----..#+_.+--~.--..~,~ .,.
1000 "%'~,:
.....
X:
100 jl.
10
.
. . . .
DAYS
_
.
.
. . . . . ~ . ~/.:.4,+.+ " ~-~-wj - w j ~ - - - ' - m -. - -. m. . . . . ..
0 -
.
I. . . .
,oo....._.+//_. 50
.
i"
.jk...~
'
' i
9
__"
%"
.I.
"
.I.
.t..
I
....%'-
!
11=
.
.Jl..
.
.
.
"+
.
,lb.
.
7/93 0/93 9/93 10;931 lj93121931194 2/94 3/94 4/94 5/94 6/94' 71~14 MONTHS '
- - , k - AVE.CI-14 C O N C E N T R .
I~1 -I'~- TOTAL CH4 PRODUCT.
I ]
'
'
- - I - - AVE.UNDERPRESSURE
I'+e'l
- ~ . ' - A V E . C l I 4 PRODUCTION
Fig. 3. Gas production of the V-3 gob well during first year of production.
I,,,,~/,+od
440
G. TAKLA & Z. VAVRUSAK
According to the energy resource policy of the Czech Government, CBM exploration is financially supported by the Ministry of Economy. By the end of 1994, nine CBM wells were drilled. Two of the wells were fractured, with promising results for future production. In April/May 1995, more four wells were fractured and long-term pumping tests were started. Four to six new wells are expected to be drilled by the end of 1995. DPB owns a concesion for CBM exploration in ten prospection areas, covering a total of 240km 2 of virgin coal fields that contain 4.1 • 10 9 tonnes of hard coal reserves. These areas are characterized by a high density of exploration wells, drilled for coal, gas and water prospection on a spacing of some 1.5 km. Several geological evaluations and studies were carried out to assess the importance of the coal reserves and to build a database. Accordingly, the coal deposits of these areas are well documented, with very good geological information about each of the prospection areas. However, the occurrence of methane in virgin coal seams is not well documented so far. For most of the prospection areas, data on gas content of the coal are either not available at all, or very incomplete at best. For this reason, a CBM exploration campaign with a chance of success would first have to identify the most favourable areas and coal seams by drilling pilot wells. The CBM pilot project proposed by DPB, comprises six pilot test wells in the most promising exploration areas, which are named as follows: 9 9 9 9
D6tmarovice Petrovice - Vficlavovice Pf'ibor - zfipad T r o j a n o v i c e - Fren~t~it.
Each of the wells is continuously cored in the Carboniferous interval to obtain primarily information on the gas content in the rock sequence. Core samples of the coal are subjected to standard canister desorption tests (according to the Standard of the US Bureau of Mines). Simultaneously, coals and rocks surrounding the coal sampling points are tested to obtain information required for the design and evaluation of hydraulic fracturing of the coal. Drilling, coring and completion are assigned to local contractors, as is hydraulic fracturing of the coal. The first pilot well was completed in the Karvinfi Formation's coal seams in the area of Ditmarovice at the begining of 1995. Successful hydrofracturing opened two coal seams (4.3m and 1.4 m thick). A beam pump was installed to test water and gas production. The second pilot well in the Vficlavovice area (Ostrava Formation's coal seams) did not realise the expected coal gas content and gas in place. The third pilot well of DPB' s project is now being drilled in the area of Fren~tfit (Karvinfi Formation). An additional one or two pilot wells are planned for spudding in 1995.
Summary The paper presents information on some methods of coal gas processing and utilization in the largest hard coal basin of the Czech Republic., i.e. the Ostrava-Karvinfi Mining District. Improvements in the methods of mine gas drainage, increased recovery of methane from closed mines, and new coalbed methane production from coal fields of the region could contribute to the worldwide aim at improving the use of clean energy sources and reducing the emission of methane into the atmosphere.
Index
References in italics are to Figures or Tables Abbey Craig East Fault 396, 397, 398-9, 399, 400 acid mine drainage 290, 294 adjacent strained volume 395-9 Af~in-Elbistan deposit 93, 95 air pollution 261,365, 368, 371,374 A1, in solution, toxicity of 291 algal coals 201-6 algal mats 72 alginite 56, 108, 110, 201 Alpine Orogeny 50 Andano-Chulman coal-bearing region 149, 150, 151, 151, 153, 154, 155-6, 158 vitrinite reflectance 154-5 ankerite 289 aquitards and aquicludes, M~eno-Roudnice basin 421 Asagibelova Formation 118 Askale coalfield 124 asphaltenes 358, 363 atrinite 145 Badin coalfield 237, 240, 241, 242 Baklan Granite 118 B~lce~ti coal complex 131 Bara Formation 237 Barbora units 24 basin-fill M~eno-Roudnice basin 412 S Wales coalfield 162, 164-5 Belgium, underground coal gasification 386-7 Ben~itky nad Jizerou area 415 Berbe~ti Formation 131 Berkakit Member 150, 151 Bevercotes Colliery 257 Beypazari deposit 88-9, 91, 98, 124, 371, 372 Bey~ehir deposit 93-4, 95 Bicir Formation 94-5 Bilfi Hora Formation 421 Bilfi Hora (nr Plzeh) sandstone/claystone, Fe-oxides 35, 39-40 tuffite, single-component remanence 33, 38-9 Bilina delta 208 Bilina mine 310, 310, 316, 318 Bilsthorpe Colliery, sulphur high 257 biomarkers aliphatic, Maritza-Iztok lignite 219-28 M~eno-Roudnice basin 420, 422 bitumen 226 aliphatic and aromatic portions 219-20 analyses, South Yukutian coal basin 155-8 bituminite, Selimoglu coal 110 bituminous coal 36 former Czechoslovakia/Czech Republic 5, 5-7 M~eno-Roudnice basin 412-17, 418 Point of Ayr 349-56 Selimoglu unit 107 UK, ash content reflecting mineral content 287, 292-4 Bohemian Massif 29, 32, 33, 321,322 Namurian paralic basin 13-27
borehole diameter, changes in 305-6 borehole surveys 406-7 boron, positive correlation with sulphur 259 British Coal database 245, 249 Brno granitoid massif 181 brown coal liquefaction by hydrogenation 357-63 North Bohemian coal basin 208, 210-16 Yugoslav, different forms of sulphur in 269-72 see also lignite BS 1016 350 Bulgaria, low rank coals 141-8 genesis of coal macerals 145, 147 geology of neogene deposits 141-3 Dacian coal-bearing province 143, 147 Sofia coal-bearing province 142, 147 Strimon-Mesta coal-bearing province 143, 147 Thracian coal-bearing province 142, 147 petrology of neogene coals 143-7 Biiyiik seam 118, 119, 120, 123, 126-7 caking and coking properties G6kler coals 122 South Yakutian coals 153, 158, 159 calcite 69, 107, 124, 289 camouflet shotfiring 326, 327-8 Can deposit 90-1, 93 Turkey 90-1, 93 Cfinde~ti Formation 131 carbargilite 144 13C NMR spectroscopy (CP/MAS technique) 201-5 carbonates 53, 110, 287, 289 carbonization tests, Istanbul-Yenikov region coals 366-8 carbons aromatic/alkenic 201 and protons, aliphatic 201,201-2, 202-4 Carpathian Mountains 321 Carpathian Nappes 184 catalytic deactivation 349, 355 ~ayirhan lignite field 371,372, 372-4 cedrane 226 cement clinker production plant 380, 380, 382, 383 cenospheres 293, 362 Central and West Bohemian Basins 29 outline geology 33-5 C~sk6 Stredohofi Mountains 208, 263 chalcedony 62 channel inflows, lake stages, N Bohemian Basin 314 charcoal see inertinite chlorite 288 Chukurov deposit 142, 145 Chvaletice power station 267 Cixerri Formation 52, 53, 53-4 clay 137 clay minerals 144, 288, 294 clean coal technology 10 cleat density, near to faults 399 cleat mineralization, late diagenetic 110 cleating 393-4
442
INDEX
coal breccioid 124, 129 clayey 137 detrital (detric) 135, 137, 208 effects of low-temperature drying 365 forms of sulphur in 261-3 fragmentation upon carbonization 368 high organic sulphur, amelioration for burning in domestic stoves 371-7 humic (huminitic) 153, 214 hypautochthonous 70 low rank Bulgaria 141-8 conversion to liquid fuels by direct hydrogenation 357-63 desulphurization by low-temperature carbonization 365-9 low rank lithotypes 195-9 origin of vanadium in 273-86 as raw material for chemical industry 7 stages of formation 279 sulphur compounds in 269 xylite/xylitic 136, 137, 208 xylodetric and semidetric 208 see also bituminous coal; brown coal; lignite coal cleaning 269, 291 coal extractability 212, 213, 215 coal facies, term 131 coal gasification 269, 385-90 coal liquefaction, improvements in 349-56 dense medium cyclone separation 350, 351-2 liquefaction, original coal/coal fractions 352-5 coal minerals and element emissions from power stations 292 influence on combustion residues 292-4 coal mines, abandoned, gas recovery from 437-8 coal mining, methane emissions from, Poland 425-34 coal rank 393 S Wales coalfield 166, 176 relationship to thrusting 161, 167-9 variations with depth 161, 169-71 coal strata, split by sediments 311,314, 315, 316 coal swelling 215 coalbed methane development in virgin coal seams 391,438-40 geological controls on producibility 391,392, 392-3, 399, 401,404 migration in and around fault zones 391-408 models for retention and migration 401-6 research in the M~eno-Roudnice Basin 409-23 secondary gas generation 401,402 coalification 279 Bulgarian low rank coals 147 increasing with depth, M~eno-Roudnice basin 420 intensity, Upper Silesian Coal Basin 20-3, 26 N~m~i6ky area 189, 190, 191 Cockshot Rock 171,175 coking coal, high sulphur, G6kler coalfield 115-30 colliery spoil, UK 290-1 condont CAI values, S Wales 173 coquinas 56 ~oraklar Formation 88, 372-3 corpohuminite 214
cutinite 110 Cypris Formation, pyrites 266 Czech Republic coal production and usage 3-12 constraints on limits of mining 8-9 energy policy 1-2 workable reserves, bituminous coal 6, 6-7 deformation 16, 47, 165 degassing 401,402 degassing systems, capture of methane 426, 430 Dejdekar volcanics 104 deltas, in coal-forming swamps 315-16 dense medium cyclone separation 350, 351-2 conversion and product distribution 352, 353 element distribution during digestion 353-4, 354 overflow fraction 351, 351,352, 353, 355-6 underflow fraction 351,352, 356 densinite 145, 214, 216 density gamma-gamma logging 310-11 depositional environments and coalbed methane producibility 392-3 different, Kolubara and Kostolac coals 271,272 KY 9 coal seam 276, 279-80 derelict land, restoration of, UK 290-1 Des-A-lupane 226, 227 desiccation Eocene 72, 73 Pontian Basin 143 desmocollinite 103, 107-8, 110, 112 destressing blasting 326, 327-8 desulphurization 9, 272, 297 by low-temperature carbonization 365-9 detrital minerals, UK coals 287-8, 289, 294 detrite 214 diagenesis deep-burial 289 early, formation of FeS2, KY 9 seam 279 marine 63 replacement of evaporites 62 Sulcis coal basin 65-7 and sulphur distribution in Sindh coals 240 diagenetic minerals, UK coals 288-9 dipolar phasing (DD) experiments 201 Dobruja coal, possible underground gasification 387-8 Dolni Vlk~, siltstone 41, 42 dolomite 63, 66, 69, 124 dolomitization 72 post-compactional 69 Doupovske Hory Mountains 208, 263 Drahanskfi Vrchovina Plateau 181 Dukeries coal 246 Duraji Member 150, 158 Eakring Anticline 257 East Pennine Coalfield 406 EEC, Large Combustion Plant Directive 291 electrolinking 388 Elga coalfield 149, 152-3, 153, 155 Elhovo deposits/lignite 142, 147, 389 gasification in a fluidised bed 390 engineering stability, and weathering 289-90 Enna marine horizon 181 epigenesis 32
INDEX Europe, present underground gasification position 386-90 evaporites, Sulcis coal basin 63, 72, 73 exsudatinite Grkler coals 124, 125-6, 126 Selimoglu coal 110 fault damage zones 395, 400 fault linkages 399-400 fault populations 400 fault systems 34-5, 343, 394, 415 Mgeno-Roudnice Basin 412, 414 strike-slip 403-6 fault zones 129, 394-5 burial history 400 and induced seismicity 329-35 faults/faulting 257, 394-5 antithetic 343 detected on FMS data sets 406 Karice deposit 340, 342, 343, 347 normal 50, 396, 401,402 reverse 343, 401,404 strike-slip 404 virgate 343 Fe-oxides 41 FeS2, formation of 279, 289 see also iron disulphides; pyrite Ffos Las OCCS 165-6, 167, 167-9 Ffyndaff OCCS 165 flood events, and seam splitting 258 flue gas desulphurization 1269, 291,372 fluid combustion 10 fluid migration fault modification of 395-400 through unfaulted ground 393-4 fluorinite 110 fly ash 293, 294 foreland basins 14, 26 fracture systems, and permeability 399 fractures, detected on FMS data sets 406 fragrnentograms 222, 225, 226 France, underground coal gasification 387 Free Swelling Index (FSI) 119, 122 Fren~tfit Seismic Polygon 332 fulvic acid 278 furnace bottom ash 293 fusain 144, 195, 196, 197, 198, 199 fusanization 198, 199 fusinite 201,205, 281,317 gelification, huminite 211 gelification index (GI) 214 geophysical logging 406 Germany, underground coal gasification 386 gob wells 438, 439 Grc~iktepe sediments and alluvium 118 goethite 39, 41 Grkler coalfield 115-30 caking/coking properties of coals 122 geological setting 118-19 iso-reflectance map/values 127-9 mineral matter of coals 122-4 G6kler Formation 118 seams in 118, 118-19
443
Goze Delchev deposit 143, 147 graphitic domains 205 gravitational sliding/slumping 50 greenhouse gas emissions evaluation 425, 433 grison 386 G(ines ophiolite 103 gypsum 107 1H NMR MAS technique 201,204 Hfidy-Rirka Limestones 181 haematite 32, 38, 41, 43, 46 Hanioglu unit 103, 104 Hantepe unit 103 heat flow, S Wales coalfield 173 heavy industry, Czechoslovakia 3-5 Hilt's Law 189 excursions from, S Wales 169, 171,174, 175, 176 Hirka Formation 88 Holodnican Member 151, 158 hopanes 226, 227 Hrabfik mine 262 H~edle Member 415 Hru~ov Member 24, 26 humic acid 278 huminite 144, 145, 219 N Bohemian coals 208, 211,216 huminite reflectance 147, 214, 358 humoclarain 219, 221,226 humoclarite 144 humotelinite 137 humovitrain 195, 196, 197, 198, 199, 219, 220, 226 hydrofracturing 387, 388, 407, 440 hydrothermal fluid flow 173, 176 hydrothermal fluids 276 Iberian Plate 50 illite crystallinity 167 Ilnitsa Suite 229 inelastic scatter 300 inertinite 110, 113, 136-7, 144, 188, 219, 278, 349, 351 N Bohemian coals 211,214 South Yukutian coal basin 153, 159 see also fusinite inertodendrite 110 iron 257 ferric, ferrous, covalent 196-9 iron disulphide/sulphide 211,213, 214 6345 values in 261-7 diagenetic 264, 266, 292 epigenetic 264, 266 synsedimentary 263-4, 266, 267 Jaklovec Member 16, 17, 21, 24, 181, 191 jarosite 278 Jelenice Member 412, 415 Jerma seam, fusinite 201,205 Jeseniky Mountains 321 Jiu-Motru Formation 131 Jizera fault system 415 jointing, and fluid migration through coal 393-4 Juhta Member 150
444
INDEX
Kabakta Member 150, 151 Ka~ice deposit coalfield data 339-40 history of 337-9 post-sedimentary faulting 340, 342, 343 Ka6ice depression 338 Kalbur~;ayiri Formation 94 Kammenn~, Most Member 412 Kangal deposit 94-5, 96 kaolinite 124, 293 kaolinite precipitation 67 Karacahisar volcanics 118 Karadoruk Formation 88 Karakaya fault zone 129 Karliova deposit 96-8 Karsakatepe deposits 118 Karvin/t Formation 188 Kerme Formation 85, 87 Khadro Formation 237 Kjustendil coals 143, 144 Kladno Basin 6 reflection seismics 340, 341 stratigraphy of 337-9 Kladno coalfield 337, 342, 343 Kladno Formation 34, 337-8, 409, 412 Kladno mine 337-47 Kolubara coal/lignite 269, 270, 271, 271,272, 379 Kolubara mines, 'Tamnava' field, brown coals, liquefaction by catalytic hydrogenation 357-63 Kop~ivnice-T~inec uplift 26 Korycany Member 421 Kostolac coal 269, 270, 271, 271-2 Kounov coals/seams 412, 415, 416 Kounov Member 412, 415 Kozloduj coal deposits 143 Kozluca Formation 110-11 Krugn6 Hory Fault 208 Krugn~hory Mountains 262 supplying sulphate to N Bohemian Basin 266 Kru~n6hory piedmont coal basins 8 KiJc[ik seam 118, 119, 120, 122, 123, 126-7 lake phases, peat-swamps 313, 314-16 Lakhra coalfield 237, 239, 240, 241, 242 Lakhra Formation 237 La2finky Limestones 181 Ledce Member 412 Ledce-Z;ilov, red claystone pit, two-component magnetization 42-3, 44 Leonard seam, lower whetstone 25 Libou~ mine 314, 314 lignite Bulgaria 147 Czechoslovakia/Czech Republic 4, 4, 5 Istanbul-Yenikov region, low-temperature carbonization tests 365-9 Kolubara, use in cement rotary kilns 379-83 Maritza-Iztok coal basin 219-28 Turkey 77-99 lime, as sorbent for amelioration of high sulphur coal 374-6, 377
limestone dolomitic 119 micritic 101, 103 reef 181 Lin6 Formation 337, 409, 412, 415 lipids, bacterial 221 liptain 144, 195, 196, 199, 225 bitumen content 220, 220 liptinite 119, 144, 153, 159, 188, 219, 349, 351 N Bohemian coals 211,214, 216 Selimoglu coal 103, 107, 108, 111, 113 liptobiolite 208 liptodetrinite 110, 208, 214, 216 Litom6fice Deep Fault 35 Llanharan Colliery, excursions from Hilt's law 171, 175 Llannon Disturbance 167, 167 logging correlation scheme, N Bohemian Basin 309-20 Lom coal deposits 143, 147 Lom depression 143 Longannet mine, Scotland 395-9 Lubnfi seams 337, 338 Lu2ice-Labe line 419 Macocha Formation 181 magnetic remanence isothermal (IRM) 37, 40-1 multi-component 39, 41, 46 single-component 33, 38-9 magnetic remanence components, W Bohemia 29-47 magnetite 39, 41, 46 magnetization chemoremanent/thermoremanent 32 two-component 39 viscous 42 Main coal seam, Most Formation 207-8 logging correlation scheme for 309-20 Main Ostrava Wetstone 19, 22, 25 Malkara deposit 85, 88 marcasite 262, 263, 266, 288 marine facies, Ostrava Formation 24 Maritza East deposit 142, 147, 389 Maritza-Iztok coal basin 195-9 lignite, aliphatic biological markers in 219-28 maximum maturity temperatures, S Wales coalfield 171, 173 megasequences, transgressive-regressive 51-2 M61nik Coals 412, 415 M61nik Interjacent coal seam 415 M~lnik Main Seam 415, 415-16, 422 estimation of CBM reserves 416-17 Mengin deposit 85, 87 M6nin block 181, 184 metamorphism 166-7, 276 methane emission Polish coal mining 425-344 from mining processes 426, 428 from post-mining activities 426, 431 utilization, Ostrava-Karvinfi coalfield 435-40 methane release, from coal 426, 430 micro-karst 53, 56 Mililolitic Limestone Formation 52, 67 Mililolitic Limestones/Produttivo Formation boundary 53
INDEX mine gas drainage from abandoned mines 437-8 Ostrava-Karvin~t coalfield 435-7 mineral alteration, South Yakutian coal basin 158-9 mineralization antimony, G6kler coalfield 115, 116, 117, 129 hydrothermal 129 S Wales coalfield 173:-4 sulphide, syngentic and epigenetic 211 mining longwall, seismic energy release during 323-4, 325, 326 opencast/opencut/open pit 998, 149, 151,208, 290 Miro6 seam, fusinite 201,205 Miro~ov, 'Lomy na Janovr' quarries, siltstone 36, 40, 46 Mn, toxicity of 291 Moesian Platform 143 monosulphides 269, 271 Moravian-Silesian Paleozoic Basin paleogeographic features 25-6 paralic morasse 13-14, 16-24 partial troughs within 14-15 Mrssbauer spectroscopy 195-9, 278 Most Formation 207-8 Motru coal complex 131 M~ec Member 412 M~eno-Roudnice Basin coal deposits 412-15 geochemical evaluation of samples 417, 418 technological properties of coals 415-17 geological setting 409-12 hydrogeology 421 research into coalbed methane 409-23 thermal history 418-21 mudrocks behaviour during weathering 289 chemical alteration a rapid process 290 Mu~la region (Yata~an-Milas) deposits 85, 87, 89 multicomponent statistical analysis 302, 304, 305 Munzur limestones 110-11 Muratdagi Melange 118, 129 Muratdagi region 122, 129 Myslejovice Formation 181 Namurian paralic molasse, Bohemian Massif 13-27 natural gas, use of 10 with pulverized lignite in cement rotary kiln 379-83 Nrmri~ky area 13 age of coal bearing rocks/coal fragments 184-5 coal found in deep boreholes 179-81 coal petrology 185-9 geology of 181-4 S extension of Early Namurian deposits 179-93 Nerjungra Member 150, 151 Nerjungry coalfield 149, 156 Nesvarilka Block 184 Netherlands, underground coal gasification 386 North Bohemian Coal Basin 207-17 analyses, elemental, petrographic and technical 209, 212-16 constraints on mining 8-9 634S values in iron disulphides 261-7 geological setting 263 logging correlation, Main coal seam 309-20
445
Northern Anatolian Fault Zone 124 N ~ a n y Member 338-9, 347, 409, 412, 421 coal seam groups 415, 416 Ogulbey Formation 103, 104 Ollerton Colliery 257 Oltenia coal basin 131-9 main characteristics of coal facies 134 main coal facies and lithotypes aquatic macrophyte prairie 137-8 deciduous forest 135-6 forest swamp 136-7 grassy marsh Carex spp. 135 reed swamp 137 Orhaneli deposit 91-3, 94 orogenic trend, Variscan 24 Orta deposit 96, 97 Ostrava Formation 13-14, 16-24 changes in petrographic/geochemical composition 25, 26 Nrmrirky boreholes 181, 184, 191 vertical changes in development of 24-5 Ostrava-Karvinfi coalfield 13 induced seismicity related to fault zones 329-35 seismicity and mining situation 330-1 seismological monitoring 331-2 tectonic and geomechanical situation 330 Karvinfi area evaluating seismicity of tectonic zones 332 fault systems in 330 methane emission and its utilization 435-40 coalbed methane development, virgin coal seams 438-40 gas recovery from abandoned mines 437-8 seismic monitoring for rockburst prevention 321-8 induced seismicity, long-term observations 322-5 local seismographic network 321-2 regional seismographic network 322 use of observations in geomechanical practice 325-8 Seismic Polygon 331-2 workable coal reserves 5-7 workable reserves 5-7 oxidation 198, 199, 289-90 oxides 287 oxyhumolites 208 Pakistan, distribution of sulphur in Sindh coals 237-43 palaeobiotopes 133, 134 palaeoclimate Oltenia coal basin 133 Transcarpathians 234-5 palaeogeography development of N Bohemian Coal Basin 311-20 of Oltenia coal basin 132-3 palaeogeothermal gradients, S Wales coalfield 161, 169, 170, 171-3, 176 palaeomagnetic directions 43-4 palaeophyocoenoses, Oltenia coal basin 133-5 palaeosalinity 259 palaeosols 56 palaeostress, Ostrava-Karvin~ basin 330 palaeovalleys, containing coal seams 338
446
INDEX
palynomorph assemblages, Upper coal formation 230-6 parasequences, Produttivo Formation 54, 70-1 Park Colliery, rashings in New Shaft 171,175 Park Slip OCCS 165 Parkgate Coal, S organic and pyritic 291,292 Parkgate seam 247, 248 ash-sulphur relationships 253-4 subdivisions, E Pennine Coalfield 248 sulphur distribution 249-57 controls on 257-9 partial block structure, Ostrava-Karvinfi basin 330 peat accumulation, Dacic Basin 131 peat mire, complex controls in causing sulphur variations 258 peatbog phases, peat swamps 313, 316-17 peatbogs final destruction of 317 metal salts in 198 permeability, and coalbed methane production 391, 393 Peruc Member 421 Pet~kovice Member 16, 17, 20, 24 phosphates 287 phyllocladane 225 c~-phyllocladane 221,225, 226 /3-phyllocladane, and thermal maturity 225-6 Plzefi Basin 30-1, 34 pollution 4 by underground coal gasification 388, 388 see also air pollution polyaromatic structures 204-5 porewaters 291 from fly ash sites, contamination source 293, 294 potential source of sulphate to coal seams 257-8 a residual sulphate reservoir 266 porosity, secondary, and feldspar dissolution 66-7 Poruba beds/Member 24, 181, 184, 191 power industry, coal from N Bohemian coal basin 208 Produttivo Formation 50, 52, 67 diagenetic products 65-7 intertidal facies 55-6 overlying palustrine-lacustrine facies 56 parasequences in 54, 70-1 regressive trend 60 subtidal facies 54-5 supratidal facies 56 Progressive Easy Slip Thrusting (PEST) 166 psammite tongues, N Bohemian Basin 315, 318 Pyrenean orogeny 50 pyrite 107, 144, 197, 198, 199, 250, 257, 271,272, 288 containing elements of environmental concern 292, 294 framboidal 67, 110 G6kler coals 124 in KY 9 seam coals 278-9 massive 110 N Bohemian Basin 262, 263, 264, 266 oxidation of 289-90 and S02 emissions 291-2 syngenetic 214 underground oxidation of 290 see also iron disulphides
quartz mineral 122-4 Radrice (nr Plzefi) claystone, multi-component remanence 39, 41 siltstone 40, 41 Radnice Member 337, 338, 409 fault detection in 340, 341 'rashings' 171, 175 reflection seismics 406 common-midpoint method, Kladno Basin 340, 341 reservoir pressure, controlling gas content 426 resinites 110, 144, 227 rockbursts 330, 334 seismic monitoring for prevention of 321-8 Rosice-Oslavany vasin 57 rotary kiln, wet process 380, 380, 382, 383 Roudnfi (nr Plzefi), claystone, isothermal remanent magnetization 37, 40-1 sealing, by fault damage zones 395 seam plies Parkgate seam 245, 258 ply-by-ply mapping 257, 259 seam splitting, and sulphur content 258 seam sulphur maps, generation of 249 sedimentary particles, petrography of, Sulcis Basin 62-5 sedimentation 310 coal-forming 311-20 cyclic Early Namurian 181, 184 N~,~'any Member and T~nec Formation 339 Oltenia coal basin 132-3 influence of eustatic movements on 25 S Wales coal basin 162, 164 seismic activity, natural and imposed 321 seismic monitoring for rockburst prevention 321-8 seismic reflection surveying 406 seismicity, induced, Ostrava-Karvin/t coalfield 322-5 relationship to fault zones 329-35 seismological monitoring 331-2 sources generated in stress concentration zones 335 spatial distribution of hypocentres 322-4 tectonic and geomechanical situation 330 Sekk6y Formation 87 Selimoglu coal field 101-14 geological setting and stratigraphy 102, 103-5 mineralogy 107 petrographic composition/depositional environments of coals 107-11 Selimoglu unit 103-4 semifusain 144, 195, 196, 197, 198, 199 Seyit6mer deposit 90, 92, 124 siderite 65, 67 silicates 287 silicification 129 diagenetic 72 Slan2~ Formation 34, 337, 339, 409, 412, 421 coal seam groups 415 slip, in UK coalfields 393, 394 smectite, inteparticle expansion of 289
INDEX smog, Czech Republic 9-10 Smr6iny unit sulphides 266 SO2 emissions 269, 374, 377 from pyrite and organic sulphur 272 and Large Combustion Plant Directive 291 and role of pyrite 291-2 soft coal 208 Softa-1 and Softa-2 formations 88 Soma deposit 87-8, 90 Sonahri Formation 237 sorbent addition, amelioration of high sulphur coal 374-6, 377 South Wales coalfield 161-78, 401 coal rank development related to thrusting 161, 167-9 variation with depth 161, 169-71 comparison with Ruhr coal basin 176 metamorphism 166-7 role of fluids 173 stratigraphy 161-4 structure 164-6 South Yakutian coal basin, Siberia 149-60 bitumen analyses 155-8 Cretaceous flora 153 geological setting and stratigraphy 150-3 Jurassic flora 151, 153 mineral alteration 158-9 proximate analyses and vitrinite reflectance values 153-5 Sozopol, Gulf of, coal deposit 142 sparites 63 spectra, processing methods 301-3 spectra deconvolution 302, 304, 304, 306 spectrometric detectors 301 spectrometric logging probe 301 radius of investigation of 305 sphagnales 234-5 sporinite 110, 111,202, 214 G6kler coals 124-5, 126, 129 stable isotopes, Sulcis coal basin 67-70 steranes 226 strain release, Ostrava-Karvinfi coal mines 327 stress fields 400, 422 primary and secondary 330 subbituminous coal, Italy 49 sublimation 264 subsidence basinal 25-6 N6m6i6ky area 181 Variscan intermontane area 34 Sulcis coal basin, SW Sardinia 49-75 depositional setting 70-3 diagenesis 65-7 facies spectrum and distribution 53-60 geological setting, palaeogeography and stratigraphy 50-2 petrography and geochemistry 60-5 revised depositional model 73 stable isotopes 67-70 sulphate 257, 269, 271,278-9 reduction to sulphide 279 sulphate ions, leaching by meteoric waters 240 634S values, in iron disulphides 261-7 sulphides 278, 287
447
sulphur 264, 266 Beypazari coal 373-4 distribution multi-bed seam (Parkgate seam) 245-60 in Sindh coals 237-43 elemental 261,262 forms of 261-3, 269-72 organic 261,269, 271,272, 291,374 pyritic 269, 271,291,374 sulphate 261,271,291,374 sulphide 261,262-3 see also iron disulphide; pyrite volcangenic 266 sulphur content affected by depositional and post-depositional controls 257-9 determined through gammaspectroscopy 297-307 high Gfkler coking coal 115-30 Selimoglu coal 107, 113 Istanbul-Yenikov region lignite 366, 366, 367 Main coal seam (Most Formation) 213 N~m6i~ky area coals 189 total, G6kler coals 121-2 /534S values, in lignite of N Bohemian Basin 264-7 Susta marine unit 24 swamp environment Dacic Basin 131-8 G6kler coal 122 N Bohemian brown coal 313-18 Selimoglu coal 107, 110, 111 Syllach coalfield, bitumen anayses 155 syngenesis 32 tasmanite (algal coal) 201,202-4 tectonic dislocations, Ostrava-Karvinfi basin 330, 332 tectonics extensional 50 tectonic zones, Bulgaria 142, 143, 147 telinite 159 telocollinite 103, 112, 119, 124 textinite 214 Tharparker coalfield 237, 238, 240, 241, 242 thermal demagnetization 36, 42, 42, 43, 45 thermal neutron capture, interferences generated in the spectrum 300 Thracian valley, coal formation in post-tectonic depression 141 thrust detachments, Variscan 165-6, 171,173, 175, 176 thrust faults 184, 401-2, 403 thrust ramps 164, 165, 402, 403 thrusting, S Wales coalfield relationship to coal rank development 167-9 in seam 171, 173, 176 tissue preservation index (TPI) 214 Tlustice (nr Zebr~ik), tuff-tuffite, two-component magnetization 34, 39 Tokin coal-bearing region 149, 151, 151, 152-3, 153, 155, 158 toluene, use of in coal liquefaction 349 torbanites (algal coal) 201-2 trace elements, in fly ash 293-4, 294 transgressive cycles, Lower Eocene 70 trimacerites 144
448
INDEX
Trimsaron Disturbance 167, 167 triterpanes 226, 227 Tunqbilek deposit 89-90, 92, 124 Turkey 89-90, 92, 124 Turgut Formation 87 Turkey, lignite deposits 77-99 Eocene lignites 80-1 Oligocene lignites 81 Miocene lignites 81, 82-4, 85 Pliocene lignites 85, 86 named deposits described 85-98 T~,nec (Tinec) Formation 34, 337, 347, 409, 415, 421 cyclic sedimentation in 339 UK mining data sets 406 underground coal gasification 387 UK coals, environmental impact of minerals in 287-95 ulminite 214, 216 Upper coal formation, Ilnitsa Suite, floristic characters of 229-6 Upper Silesian Coal Basin 6, 179, 191 Beskydy piedmont area 3-5 lithostratigraphy 180 Polish 8 methane content and rate of desorption 426 regional division according to methane content 428, 429, 430 see also Ostrava Formation; Ostrava-Karvinfi coalfield USA coalbed methane production 391 origin of vanadium, KY 9 seam 273-86 underground coal gasification 386, 388, 388 Usmun coal-bearing region 149, 150, 151, 154, 155, 158 USSR (former), underground coal gasification stations 385 Valea Vi~enilor coal complex 131 vanadium, origin of in US coals (KY 9 seam) 273-86 depositional environments, and thermal history 276-8, 279
KY peat-forming system, physicochemical conditions of deposition 283-4 origin of VO2+-non-P 279-80 origins of V from within and from above theories 280-1 primary and secondary V 281-3 pyrite (FeS2) and other S compounds 278-9 vanadylation, of KY 9 seam 279-80, 284 ventilation emissions of methane 426, 428, 435 Vil~movice Limestones 181 virtual pole positions, Westphalian and Stephanian rocks 43-4, 45, 46, 47 vitrinite 153, 159, 188, 349, 351 Selimoglu coal 103, 107, 108, 113 vitrinite reflectance 129, 154, 167, 176, 276, 417, 420 Ffos Las OCCS 167-9 N6m6i6ky area coals 189, 191 volatile matter contents, G6kler coal 120-1 volcanic ash 235 volcaniclastic rocks, in Variscan foredeep 18-19 volcanism Des Moinesian, and V in KY 9 seam 280-1,282 Neogene, Turkey 88-9 Vrgany mine 316 Vrganyvsand-clay accumulations 315 Vrgka Cuka anthracite 201,204-5 weathering 46 minerals associated with coal 289-90 well logging, and use of gammaspectroscopy 297-307 wireline logging 406 xylain 144, 195 Maritza-Iztok coal basin 195, 196, 197, 198, 199, 219, 220, 226 xylite 214 Yamadag lava flows 105 Yata~an Formation 87 Zatec delta 208, 262 Zd6tin coals 415 Zd6tin Horizon 412, 415 Zihle (red claystone pit), claystone, viscous magnetization 42, 43
European Coal Geology and Technology edited by R. A. Gayer (University of Wales, Cardiff, U.K) and J. Pesek (Charles University, Prague, Czech Republic) Despite the decline in the coal mining industry across Europe during the last decade, coal continues to supply a major part of the growing global energy requirements, particularly in the developing countries. During this period there has been a shift in coal extraction techniques from deep underground mines to open pit mines, reflecting cheaper production costs in a market economy. It seems probable that the European coal industry is now entering a period of stability. However, the change to opencast mining has highlighted environmental issues and the need to solve these problems. This volume presents 39 papers written by coal scientists in the forefront of European coal research. The papers cover a wide spectrum of coal geology and technology, with sections on regional coal reserves, coal basin tectonics and stratigraphy, coal petrology and palaeontology, mineral matter in coal and the environment, mining geophysics, coal technology and coalbed methane. Many of the studies describe coal deposits from Central and Eastern Europe, some of which are not well known in the west, so that the papers and included references will provide an invaluable data source. This book will be of value to all coal scientists in both the extraction and energy industries and to academics. It will also be a useful reference for students at both undergraduate and postgraduate levels.
• 448 pages • 229 illustrations • 39 papers @ index ISBN
Cover illustration: Rotating bucket excavator (TC2-K800) of the North Bohemian Basin cutting lignite of Main Seam in Vrsany Open Pit near Most, Czech Republic.
!
i.-;97799
86
1
