Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe (Geological Society Special Publication No. 313)

Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe

The Geological Society of London Books Editorial Committee Chief Editor

BOB PANKHURST (UK) Society Books Editors

JOHN GREGORY (UK) JIM GRIFFITHS (UK) JOHN HOWE (UK) PHIL LEAT (UK) NICK ROBINS (UK) JONATHAN TURNER (UK) Society Books Advisors

MIKE BROWN (USA) ERIC BUFFETAUT (FRANCE ) JONATHAN CRAIG (ITALY ) RETO GIERE´ (GERMANY ) TOM MC CANN (GERMANY ) DOUG STEAD (CANADA ) RANDELL STEPHENSON (UK)

Geological Society books refereeing procedures The Society makes every effort to ensure that the scientific and production quality of its books matches that of its journals. Since 1997, all book proposals have been refereed by specialist reviewers as well as by the Society’s Books Editorial Committee. If the referees identify weaknesses in the proposal, these must be addressed before the proposal is accepted. Once the book is accepted, the Society Book Editors ensure that the volume editors follow strict guidelines on refereeing and quality control. We insist that individual papers can only be accepted after satisfactory review by two independent referees. The questions on the review forms are similar to those for Journal of the Geological Society. The referees’ forms and comments must be available to the Society’s Book Editors on request. Although many of the books result from meetings, the editors are expected to commission papers that were not presented at the meeting to ensure that the book provides a balanced coverage of the subject. Being accepted for presentation at the meeting does not guarantee inclusion in the book. More information about submitting a proposal and producing a book for the Society can be found on its web site: www.geolsoc.org.uk.

It is recommended that reference to all or part of this book should be made in one of the following ways: EVANS , D. J. & CHADWICK , R. A. (eds) 2009. Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313. STONE , H. B. J., VELDHUIS , I. & RICHARDSON , R. N. 2009. Underground hydrogen storage in the UK. In: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 215 –224.

GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 313

Underground Gas Storage Worldwide Experiences and Future Development in the UK and Europe

EDITED BY

D. J. EVANS

AND

R. A. CHADWICK

British Geological Survey, Keyworth, UK

2009 Published by The Geological Society London

THE GEOLOGICAL SOCIETY The Geological Society of London (GSL) was founded in 1807. It is the oldest national geological society in the world and the largest in Europe. It was incorporated under Royal Charter in 1825 and is Registered Charity 210161. The Society is the UK national learned and professional society for geology with a worldwide Fellowship (FGS) of over 9000. The Society has the power to confer Chartered status on suitably qualified Fellows, and about 2000 of the Fellowship carry the title (CGeol). Chartered Geologists may also obtain the equivalent European title, European Geologist (EurGeol). One fifth of the Society’s fellowship resides outside the UK. To find out more about the Society, log on to www.geolsoc.org.uk. The Geological Society Publishing House (Bath, UK) produces the Society’s international journals and books, and acts as European distributor for selected publications of the American Association of Petroleum Geologists (AAPG), the Indonesian Petroleum Association (IPA), the Geological Society of America (GSA), the Society for Sedimentary Geology (SEPM) and the Geologists’ Association (GA). Joint marketing agreements ensure that GSL Fellows may purchase these societies’ publications at a discount. The Society’s online bookshop (accessible from www.geolsoc.org.uk) offers secure book purchasing with your credit or debit card. To find out about joining the Society and benefiting from substantial discounts on publications of GSL and other societies worldwide, consult www.geolsoc.org.uk, or contact the Fellowship Department at: The Geological Society, Burlington House, Piccadilly, London W1J 0BG: Tel. þ44 (0)20 7434 9944; Fax þ44 (0)20 7439 8975; E-mail: [email protected]. For information about the Society’s meetings, consult Events on www.geolsoc.org.uk. To find out more about the Society’s Corporate Affiliates Scheme, write to [email protected] Published by The Geological Society from: The Geological Society Publishing House, Unit 7, Brassmill Enterprise Centre, Brassmill Lane, Bath BA1 3JN, UK (Orders: Tel. þ44 (0)1225 445046, Fax þ44 (0)1225 442836) Online bookshop: www.geolsoc.org.uk/bookshop 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. # The Geological Society of London 2009. 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 0305-8719/09/$15.00. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-86239-272-4 Typeset by Techset Composition Ltd., UK Printed by MPG Books Ltd, Bodmin, UK Distributors North America For trade and institutional orders: The Geological Society, c/o AIDC, 82 Winter Sport Lane, Williston, VT 05495, USA Orders: Tel. þ1 800-972-9892 Fax þ1 802-864-7626 E-mail: [email protected] For individual and corporate orders: AAPG Bookstore, PO Box 979, Tulsa, OK 74101-0979, USA Orders: Tel. þ1 918-584-2555 Fax þ1 918-560-2652 E-mail: [email protected] Website: http://bookstore.aapg.org India Affiliated East-West Press Private Ltd, Marketing Division, G-1/16 Ansari Road, Darya Ganj, New Delhi 110 002, India Orders: Tel. þ91 11 2327-9113/2326-4180 Fax þ91 11 2326-0538 E-mail: [email protected]

Acknowledgements The authors wish to thank the following who willingly offered their time to provide reviews of the manuscripts and without whose support this volume would not have been possible: Mark Abbott, Pierre Be´rest, Benoit Brouard, Dennis Coleman, Geoff Dutton, Gerard Durup, Eckard Faber, John Gale, Irina Gaus, Joachim Gottsmann,

Dave Highley, MBE, Laurie McLenahan-Heitter, Richard Metcalfe, John Milsom, Vernon Milton, Brent Miyasaki, Nikos Papinakolau, Mike Piggin, Professor I.R. Price, Joe Ratigan, John Voigt, Tom Welch, Curtis Williams, Glyn Williams Jones, Malcolm Wilson, Brian Withington and Don Woods.

Contents Acknowledgements

vi

EVANS , D. J. & CHADWICK , R. A. Underground gas storage: An introduction and UK perspective

1

HAVARD , J. & FRENCH , R. The importance of gas storage to the UK: The DECC perspective

13

FERNANDO , A. & RAMAN , A. Gas storage: An onshore operator’s perspective

17

PLAAT , H. Underground gas storage: Why and how

25

EVANS , D. J. & HOLLOWAY , S. A review of onshore UK salt deposits and their potential for underground gas storage

39

LAIER , T. & ØBRO , H. Environmental and safety monitoring of the natural gas underground storage at Stenlille, Denmark

81

LUX , K.-H. Design of salt caverns for the storage of natural gas, crude oil and compressed air: Geomechanical aspects of construction, operation and abandonment

93

VON

TRYLLER , H., REITZE , A. & CROTOGINO , F. New procedure for tightness tests (MIT) of salt cavern storage wells: Continuous high accuracy determination of relevant parameters, without the need to use radioactive tools

129

HIETTER , L. M. Environmental issues in permitting gas storage: The Wild Goose case history

139

DAVIDSON , M. Underground gas storage project at Welton oilfield, Lincolnshire: Local perspectives and responses to planning, environmental and community safety issues

149

MIYAZAKI , B. Well integrity: An overlooked source of risk and liability for underground natural gas storage. Lessons learned from incidents in the USA

163

EVANS , D. J. A review of underground fuel storage events and putting risk into perspective with other areas of the energy supply chain

173

STONE , H. B. J., VELDHUIS , I. & RICHARDSON , R. N. Underground hydrogen storage in the UK

217

RIDING , J. B. & ROCHELLE , C. A. Subsurface characterization and geological monitoring of the CO2 injection operation at Weyburn, Saskatchewan, Canada

227

CHADWICK , R. A., ARTS , R., BENTHAM , M., EIKEN , O., HOLLOWAY , S., KIRBY , G. A., PEARCE , J. M., WILLIAMSON , J. P. & ZWEIGEL , P. Review of monitoring issues and technologies associated with the long-term underground storage of carbon dioxide

257

Index

277

Underground gas storage: An introduction and UK perspective D. J. EVANS* & R. A. CHADWICK British Geological Survey, Keyworth, Nottingham, NG12 5GG *Corresponding author (e-mail: [email protected]) Abstract: Rising demand and the depletion of its offshore reserves has resulted in the UK becoming a net importer of natural gas. An increased reliance on imports and limited current storage availability mean that the UK faces increasing energy bills and risk of disruption to supply. Because of this the UK government has set about ensuring security of energy supply. Steps taken include the construction of major new pipelines from Norway and Holland and improvements to interconnectors in the southern North Sea. The Government also recognizes that improvements to the gas supply infrastructure are required, including the need for significant increases in gas storage capacity; best met by the construction of underground storage facilities. Focus on energy security has also raised the likelihood of a new generation of coal-fired powerstations. For such a step to be environmentally viable, clean-coal technologies with near-zero greenhouse gas emissions will be required. Underground CO2 storage will be a key element of this strategy. This volume reviews the technologies and issues involved in the underground storage of natural gas and CO2, by means of case-studies and examples from the UK and also from overseas. The potential for underground storage of other gases such as hydrogen, or compressed air linked to renewable sources is also reviewed.

In October 2004, the Geological Society convened a two-day conference at the Aberdeen Conference Centre: ‘The Future Development and Requirements for Underground Gas Storage in the UK and Europe’. The conference was held at a time when attention was turning to the imminent import dependency facing the UK and was attended by representatives from not only industry and academia, but also local government departments who were dealing with applications to develop underground gas storage (UGS) facilities onshore UK. The conference took place shortly after the Moss Bluff incident in Texas (July 2004), which, in the light of the infamous Hutchinson incident in Kansas, highlighted one of the main concerns of residents in the areas of proposed storage facilities — safety (see Miyazaki; Davidson; Evans). This volume arose from the 2004 Aberdeen meeting. In the present gas and energy supply climate, this review and appraisal of the technologies of underground gas storage and future UK requirements is particularly timely.

The UK requirements for UGS and Government views on ‘need’ In 2004, Government predictions were that the UK, despite the historical riches of the North Sea, would become a net importer of gas sometime during 2006 (DTI 2003). In fact, the situation was worse than predicted, with the UK becoming a net importer of

gas during 2004. Current predictions are that the UK will import over 80% of its gas by 2020, which brings with it an increased reliance upon foreign supplies and the possibility of rising gas and energy costs and disruption to supply (DTI 2006a, b, 2007a). The conference was opened by John Havard, from the former Department of Trade and Industry (DTI), now the Department of Energy and Climate Change (DECC) and previously Business Enterprise and Regulatory Reform (BERR), who outlined Government’s view on UK gas supply and the need for increased storage infrastructure. The opening paper (Havard & French) summarizes Government’s views on the need for increased gas supply infrastructure, and a regulatory environment to allow such infrastructure to be delivered to the market in a timely fashion. UK requirements and storage operations as seen from the perspective of a UGS operator (Star Energy) are also presented (Fernando & Raman). Star Energy operates one underground gas storage facility, having converted the depleting Humbly Grove Oilfield to storage in February 2004, and is currently evaluating several other locations in the south of England and in the East Midlands (Fernando & Raman; Evans & Holloway). UK annual gas consumption is around 103 billion m3 (bcm), but current storage availability of approximately 4 bcm is only about 4% of annual consumption; equivalent to approximately 14 days supply (Table 1). This is a much smaller

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 1– 11. DOI: 10.1144/SP313.1 0305-8719/09/$15.00 # The Geological Society of London 2009.

2

D. J. EVANS & R. A. CHADWICK

Table 1. Comparison of annual gas consumption, gas storage volumes and approximate days storage in the UK, other European countries and the USA (based upon Fernando 2005; IGU 2006) Country

Annual consumption (bcm)

Storage capacity (bcm)

Storage capacity relative annual consumption (%)

Days storage (approx.)

UK Germany Italy France USA

103 101 81 46 631

4 19 13 11 114

4 19 16 24 18

14 69 59 87 66

proportion than many European countries and the USA. Forecasts are for rising demand for gas in all sectors, perhaps reaching 135 –140 bcm by 2020 (DTI 2006a, b). If ageing nuclear and coalfired power stations are not replaced by similar new plant, then they are likely to be predominantly replaced by gas-fired plants, further increasing demand. In 1990, the domestic sector accounted for 50% of UK gas usage, with gas-fired generation virtually non-existent (National Grid 2007). Low gas prices and market liberalization in the 1990s led to the ‘dash-for-gas’ as gas-fired power generation increased dramatically. Over the next fifteen years it produced a split between the market sectors, such that by 2004, power generation accounted for 33% of UK gas consumption, with domestic demand making up 36%, and the industrial/manufacturing sector 13% (National Grid 2007). Estimates suggest that the minimum share of gas in electricity generation will rise to 46% by 2012 with some analysts suggesting that this figure could exceed 60% (POST 2004). Storage of gas both onshore and offshore is only possible in suitable geological structures or formations, which are present in a limited number of locations. Potential reservoir rocks have to be of a certain minimum depth and a proven trapping configuration, including a caprock must be present. Similarly, halite beds suitable for developing storage caverns have to be greater than a certain thickness and depth. Presently, however, UK UGS applications are subject to numerous and lengthy planning consent processes; both local planning controls, currently overseen by the Department for Communities and Local Government (CLG), and specialist development consent regimes currently administered by the DECC. In addition, and as noted by Government (DTI 2006b, 2007a), local communities close to proposed facilities strenuously oppose UGS development. Opposition is based mainly on the fear of a repetition of rare but major incidents seen at UFS facilities, most notably in the USA, where fatalities have occurred (Evans). As a consequence, almost every UGS application

appears destined to undergo lengthy delays and a Public Inquiry. The situation is highlighted by the proposals for cavern storage facilities at Byley (Cheshire) and Preesall (Lancashire) and for conversion of the depleting Welton Oilfield (Lincolnshire). Byley was eventually approved by the Secretary of State in May 2004, two years after the initial application. At Preesall, the original application was submitted in November 2003 and, on appeal, went to a Public Inquiry (October 2005–May 2006). The Secretary of State for DCLG finally refused planning permission and hazardous substance consent on 16 October 2007, almost 18 months after the close of the Inquiry. Refusal was based partly upon safety aspects, including risk of gas migration and explosion. Similarly, the application submitted by Star Energy in November 2003 to convert and develop the depleting Welton Oilfield in Lincolnshire as a UGS facility has faced a difficult passage. Despite the widespread opposition to the plan from local residents and parish councils, Lincolnshire Council officials and planners had recommended in favour of the scheme, pointing out that it was in accord with national energy policy and was supported by both the energy regulator and the DTI. No major concerns were expressed over the project by the Environment Agency, English Nature or the Health and Safety Executive. However, at a public meeting in early 2006, councillors went against their officials’ recommendation and refused planning permission, citing local fears over health and safety as a main reason for refusal. The arguments and final decision were influenced by perceived risk being based upon analogy with incidents at two American gas storage facilities (LCC 2006). These incidents took place at American salt cavern UGS facilities, circumstances entirely different to that proposed at Welton, a producing oilfield with proven trap and retention capabilities. Although originally scheduled to proceed to Public Inquiry, the operators of the field may attempt to proceed under the existing 1965 Gas Act legislation (Star Energy 2006).

INTRODUCTION AND UK PERSPECTIVE

However, the progress of two UGS proposals is noteworthy. An application for a salt cavern storage facility at Stublach (Cheshire) was submitted in December 2005 during the Preesall Public Inquiry. It appeared to signal a change in local government views and had achieved full planning permission and hazardous substance consent by June 2006: both being granted on the basis of national need (Ineos 2006). Similarly, an application to convert the depleted Caythorpe gas field submitted in December 2005, was rejected by the East Riding of Yorkshire Council (ERYC). It went to a Public Inquiry in April–May 2007 and in February 2008 was approved by the ministers of state for CLG and at that time BERR, based in part, on national need (CLG 2008). Even prior to the Canatxx decision, the UK Government had become concerned by the fact that supply infrastructure developers were being faced with increasing risk and delay to proposals and no guarantee that a project would proceed. Government observed ‘It is all too easy to suggest that the need can be met in some other way, or that the project could be located elsewhere. All localities have a part to play in national energy policy. Just as some locations are more suitable for wind farms due to factors such as wind speed, so other localities will be more suitable for gas storage’ (DTI 2006b). Of interest in the Caythorpe inquiry is that case law (the Newport case) was held to show that public perception of fear (risk) is capable of being a material consideration in determining planning applications (core documents cited in Newman 2007). A number of factors were seen to demonstrate that local people’s fears for their safety were neither baseless nor unfounded. First, the site would not be COMAH regulated and secondly, ERYC submitted a record of incidents associated with underground gas storage (core documents cited in Newman 2007). However, Ministers of State (CLG/BERR), in line with the Inspector, reached a different conclusion to the Preesall decision, stating ‘that risk cannot be entirely eliminated, but that is true of any form of onshore storage, and there were not such risks to human health and safety as to warrant rejecting this particular site’ (DCLG 2008). In the event of supply disruption, the lack of storage and increased consumption would leave the UK in a position of great vulnerability. The potential for such a situation was highlighted by difficulties experienced during the winter of 2005/2006, which resulted from a combination of events (Lowery 2006). First, in January 2006, a gas dispute between Russia and the Ukraine led to difficulties in supply to a number of European countries and caused volatility in prices. Within hours, Austria, France, Germany, Hungary, Italy, Poland and Slovakia had all reported pressure

3

drops in pipelines of 30– 40%. Secondly, immediately following a fire and explosion at the offshore Rough storage field (the UKs largest gas storage facility) in February 2006, wholesale prices rose by 40%, but dropped back when the extent of the problem on the platform was clarified. The need for additional gas storage was highlighted following the decision of National Grid to call a Gas Balancing Alert on Monday 13 March 2006 (Lowery 2006). The situation resulted from uncertainty over imports from Norway due to unplanned maintenance on platforms and pipelines and other incidents on the Norwegian Continental Shelf, as well as the prospect of supply disruptions in France due to industrial action. Rising demand and an increased dependence on gas imports when coupled with limited gas storage volumes pose an additional problem to the UK. The lack of gas storage volume means that storage facilities normally filled during off peak summer months (when gas prices are lower) fail to provide an adequate buffer of stored gas. The UK will, therefore, have to buy gas during the high demand winter months, or on the short-term market, when prices are higher. The general public is already suffering from this shortfall in storage volume in rising energy bills and more people could be forced into what the Government call fuel poverty: when households spend more than 10% of their income on fuel to heat their homes adequately (DTI 2003; DEFRA 2004). If the UK encounters harsh winters then the problem could be made much worse. There are also important safety issues. A shortage of gas, if not managed properly, could give rise to a gas supply emergency. For safety reasons a minimum pressure must be maintained within the National Transmission System (NTS), which requires a balance between gas supply and demand. Put simply, gas taken from the network by consumers has to be replaced by gas flowing into the network from producers, gas processing facilities, storage facilities, interconnector pipelines and LNG import facilities (HSE 2008). It is dangerous if the pressure in the network is too low because appliances may not burn the gas properly, resulting in incomplete combustion and noxious fumes. Alternatively, if the appliance does not have an effective protective device, flames may go out only for the gas to be re-ignited when pressure is restored resulting in a fire or explosion (HSE 2008). In addition, air may enter the system. In the event of supply disruption and a drop in pressure in the NTS, supplies to businesses will be interrupted before domestic customers. In a worst case scenario, whole towns and villages could be progressively disconnected and this could lead to further problems as air may enter the system to mix with the gas resulting in a potentially explosive mix and fatal explosions (House of Lords

4

D. J. EVANS & R. A. CHADWICK

2004; HSE 2008). This would result in manpowerintensive and time-consuming procedures, as every supply point must be shut off at the meter. On reinstating the system when sufficient gas is again available, great care would have to be taken with each sub-network, each main and each meter. Furthermore, when a property is reconnected, the fitness of the appliances would also have to be checked (House of Lords 2004). This would result in a slow reconnection of gas supplies, as all premises would require inspection prior to the restoration of supplies. Government, aware of potential supply problems and that businesses and homes in the UK require a reliable supply of energy that also maintains safety of the operating system, is concerned about delays to delivering gas supply infrastructure. It has already taken steps to increase import capacity and secure supply. This includes improvements to compressors on the existing Zeebrugge interconnector, the construction of the major new Langeled South pipeline from Norway (Ormen Lange field) and, in the southern North Sea, an interconnector pipeline from Bacton to Balgzand (Groningen, Holland). However, Government recognizes that UK applications to develop import and gas storage facilities in suitable areas face major delays in the delivery of important gas supply infrastructure (DTI 2003, 2006b, c, 2007b). Currently these include both local planning controls overseen by the CLG and specialist development consent regimes currently administered by DECC. The need for increased gas supply infrastructure and a regulatory environment to allow such infrastructure to be delivered to the market in a timely fashion was set out by the Government in the Energy White Paper of February 2003: ‘Our Energy Future — creating a low carbon economy’ (DTI 2003). This was re-iterated in two further Government papers: ‘The Energy Challenge: Energy Review Report’ of July 2006 (DTI 2006b) and ‘Meeting The Energy Challenge: A White Paper On Energy’ of May 2007 (DTI 2007a). In 2006, the Secretary of State announced plans to review the current regulatory framework in the UK for gas supply infrastructure both onshore and offshore (DTI 2006b, c). In November 2006, the Government also issued a consultation paper ‘Offshore natural gas storage and liquefied natural gas import facilities: Improving the regulatory framework for offshore natural gas storage and offshore LNG unloading’ (DTI 2006c), aimed at finding solutions to potential regulatory and licensing problems for the construction of offshore import and storage facilities. Following consideration of responses to this consultation and given the number of parties interested in offshore developments and the UKs need for this new infrastructure, DECC intends that draft legislation be prepared and considered by Parliament as soon as Parliamentary time allows (DTI 2007b).

Onshore, Government recognizes the importance of local democracy in the decision-making process and that stakeholders’ views must be taken into account. However, it also states that a rigorous planning system must enable decisions to be taken in a reasonable time, with a balance being struck between the concerns of local authorities and those that they represent and the national need for infrastructure that will provide us all with secure energy supplies. Indeed the Office of the Deputy Prime Minister (ODPM and now the CLG) Planning Policy Statement 1 recommends that planning authorities should ‘recognise the wider sub regional, regional or national benefits of economic development and consider these alongside any adverse local impacts’ (ODPM 2005). Government further states ‘New energy infrastructure projects may not always appear to convey any particular local benefit, but they provide crucial national benefits, which all localities share. In particular, projects add to the reliability of national energy supply, from which every user of the system benefits’ (DTI 2006d). However, any changes to the system designed to speed up the process must ensure that the safety of operations is addressed at all times. These operations are covered by existing HSE and COMAH regulations (HSE 2006a, b). The Government is, therefore, consulting on proposals to address the need for a simplification of the onshore gas planning regime as part of the 2007 White Paper, Planning for a Sustainable Future (DCLG 2007). This sets out proposals for the new planning system and for rationalizing the regime for nationally significant gas supply infrastructure projects in England. Government took a significant step in this process on 25 June 2008 when Members of Parliament voted to speed up the planning process by creating a new infrastructure planning commission (IPC). The commission will have key powers on big projects such as nuclear plants or airport extensions (Webster & Elliot 2008). National policy statements will set out the country’s requirements for key infrastructure projects and will be the subject of a ‘national debate’, with public and parliamentary involvement, which would then provide a framework for the commission to make decisions about individual projects. The commission could, therefore, also be required to rule on proposed UGS facilities.

Geological storage options There are a number of ways in which hydrocarbons (including natural gas) and other energy carriers such as hydrogen or compressed air can be stored underground. The main forms relevant to natural gas storage in the UK are pore storage in reservoir rocks (depleted oil/gas fields or aquifers) and bulk

INTRODUCTION AND UK PERSPECTIVE

storage in man-made caverns, primarily in salt, although storage has been successfully undertaken in other rocks. Plaat describes and reviews the nature of storage offered by the various forms of underground geological repository. In addition to those outlined by Plaat, other less frequently used options are available including abandoned mine storage. Although hydrocarbons have been stored in the latter, for example in America (abandoned coal mine) and France (abandoned iron ore mine), most facilities have closed due to leakage of the stored product through the caprock (e.g. Raven Ridge Resources 1998; Piessens & Dusar 2003). A little known storage facility was constructed about 180 m below ground in the Chalk at Killingholme in North Lincolnshire. Opened in 1985, liquefied petroleum gas (LPG) is stored and retained in man-made caverns by hydrostatic pressure (Trotter et al. 1985; Geological Society 1985). For a number of reasons, halite (rock salt) represents a unique host material for the development of large man-made caverns at depths of 300– 2000 m. These caverns offer important storage space for materials that do not themselves dissolve salt. Worldwide, salt mines and many thousands of caverns are used for the storage of a variety of products. The British and European standard for gas storage BS EN1918-3:1998 (BS 1998b) recognizes the efficacy of salt storage, due to its impermeable and visco-plastic properties. Caverns specifically engineered and constructed in halite offer important storage volumes that may be used for the storage of liquid (oil, natural gas liquids (NGLs and liquefied petroleum gas (LPG)) or gaseous hydrocarbons, hydrogen, compressed air (e.g. Crotogino et al. 2001; Leith 2001; Cheung et al. 2003), paper and magnetic records. Thousands of caverns are being used to store hydrocarbons worldwide with around 100 in France alone (Be´rest & Brouard 2003). A total of 66 facilities with around 396 caverns are used for storing natural gas (IGU 2006). They provide an important component of a storage portfolio, offering short to medium term, high deliverability options within and complementing the longer-term seasonal storage offered by pore storage facilities (Plaat). Salt caverns may also be used for the disposal of (generally solid) waste materials and radioactive waste (e.g. Veil et al. 1998). In the United States and Russia caverns have been used for the underground testing of munitions and nuclear weapons (Thoms & Gehle 2000; Leith 2001). Salt caverns may also have a use in the storage of carbon dioxide (Dusseault et al. 2001, 2002; Shi & Durucan 2005), although this does not represent a particularly sustainable use of the halite storage resource. The UK has important oilfields and salt deposits both onshore and offshore. Onshore, two operational

5

gas storage facilities are developed in the depleted Hatfield Moors gas field, converted to a gas storage facility during 2000 (Ward et al. 2003) and the Humbly Grove oilfield, which commenced operation in November 2005. A variety of hydrocarbons are also stored in operational salt cavern facilities in the Cheshire Basin and in NE England. UK geology would thus permit a significant volume of natural gas to be stored underground in a variety of subsurface facilities, providing a blend of longerand shorter-term storage to meet the differing supply demands. The development and distribution of halite beds onshore UK and the locations of currently operational and planned gas storage cavern facilities are outlined by Evans & Holloway.

Environmental issues Alongside safety issues, initial environmental review is key to gaining public acceptance of new underground gas storage facilities. Experience of UGS in the UK is limited and there are few published examples relating to environmental issues in other UGS developments. The Wild Goose Gas Storage Field in California illustrates the benefits of carrying out a thorough environmental review. In this case, permits were approved without opposition (McClenahan-Hietter). Key factors for success are described and the approach to environmental review can be a model for the process of granting permits to other fields, serving to reduce schedules, costs and risks when considering new storage fields. A case study is presented for the environmental and safety monitoring of a natural gas underground storage facility at Stenlille, Denmark (Laier & Øbro). For safety reasons and to protect the environment it is necessary to monitor the storage operation carefully. Occasional higher concentrations of dissolved methane have been encountered in shallow observation wells. However, stable isotope analyses and radiocarbon dating show that the gas does not originate from the underground gas storage operations but is instead the result of local microbial activity. It shows the benefits of soil gas monitoring above a UGS site, with implications for safety (see below).

Gas tightness, safety and monitoring of underground storage operations As stated previously, many local residents close to proposed underground gas storage facilities raise questions over the safety of UGS operations. As part of this opposition, one or two isolated incidents that have led to a number of fatalities are commonly cited. Most infamously at Hutchinson (Kansas), gas escaped from a salt cavern via a damaged well and

6

D. J. EVANS & R. A. CHADWICK

migrated some 14 km to the town of Hutchinson, where it emerged at the surface via old disused brine wells, causing gas geysers around the town. One at a caravan park ignited, killing two people (Allison 2001a, b; Watney et al. 2003; Nissen et al. 2004). Hutchinson and other incidents at salt cavern storage facilities, generally in the United States, are cited by most action groups opposed to the development of storage facilities (including pore-space ones) as reason enough to refuse permission on safety grounds. However, the two storage options represent very different storage environments with operational differences and should not be compared in such a way. Gas tightness, safety and the monitoring of underground storage infrastructure, particularly wells, are all of great importance. If most of the rock formations around the wellbore are impermeable, the situation is favourable. Soft-impermeable formations can have a very beneficial effect in that they creep naturally and tend to tighten around the well, improving the bond between the cement and the casing. For example, the salt layers in which the Tersanne natural gas facility is developed in France are overlain by 600 m of predominantly clayey strata. Geophysical logs have revealed a significant improvement in well conditions over time, which is attributed to clay creep (Be´rest et al. 2001). In depleted oil or gas fields, exploration and development wells form the principal breaches of the original caprock, and provide the main potential pathways for stored gas to migrate back to surface. This has proved to be a problem in California where many wells were drilled during the early years of exploration at the end of the nineteenth and beginning of the twentieth centuries. Their locations are often not well documented and many were not properly completed, even to past regulatory standards. Their condition may well be poor and permit the leakage of gas. The Mont Belvieu incident in 1980 was related to a well that dated from 1958 and illustrates a lesson that concerns the lifespan of wells. The well performed satisfactorily for 22 years with a leak occurring subsequently (Be´rest et al. 2001). Miyazaki reviews well problems and explains that even modern well completions suffer deterioration perhaps more quickly than might be expected. In the UK, oil and gas exploration onshore does not have the history of intense activity seen in parts of the United States. The systematic search for oil in Britain commenced at Hardstoft (Derbyshire) in October 1918. This was the first of a number of exploration wells drilled by D’Arcy (the forerunner of BP) in the period 1918–1920. Prior to this, drilling had taken place for water and mineral (coal, iron ore etc.) exploration purposes. A second phase of onshore drilling took place in the

mid –late 1930s when many of the oilfields in the East Midlands were discovered. Since then, exploration has continued sporadically and the locations of exploration and production wells are generally well known. Many recent onshore hydrocarbon fields do not have the density of exploration wells as is often found in earlier phases of exploration. Modern drilling techniques have allowed multiple wells to be drilled directionally from just one or two platforms and from which the fields are produced. Provided locations are known, remediation of old wells can be undertaken to improve their integrity, prior to storage operations. Salt caverns are cavities that are connected to the surface through a cased and cemented well. There may be one or more casing strings set in the well to allow injection or withdrawal of fluids into or from the cavern. The well represents the main problem in the escape of hydrocarbons both when injecting into and producing from the cavity (Be´rest et al. 2001). Gas tightness is clearly a fundamental prerequisite for these cavities and storage wells are generally completed to a higher standard than is typical in ordinary oil-industry operations. This is achieved through carefully designed drilling programs and brining operations, and good grouting and fixing of the well casing and cement, particularly in the vulnerable salt roof above the cavern. Many old brine wells and caverns created during brine extraction for the chemical industry were not designed or constructed with gas storage in mind, and were not subject to the design criteria of modern day gas storage caverns. The latter can be constructed within very exact tolerances, with sizes and shapes monitored accurately by sonar techniques. Lux reviews the current developments in salt cavity design, with sections covering the geomechanical characteristics of storage cavities, principal safety demands for their design and recent design concepts. Also covered are geotechnical methods for assessing cavern safety including criteria for determining operational pressure ranges and safety margin and abandonment procedures. Halite beds or salt domes present a corrosive environment to both well casings and cements and may lead to problems during drilling or completion of a gas tight well. This is particularly so if water is present. Shallow halite beds may be dissolved by circulating groundwater resulting in a zone of wet rock head with collapse breccias. Salt domes in the United States are often overlain by a very permeable zone (caprock) of solution breccia, through which brine circulates and that may lead to a number of problems with any wellbores. Problems may occur due to ground instability and there may be issues with achieving good cementing and completion of wells. Important bedded halite deposits are present onshore in the UK but there are no

INTRODUCTION AND UK PERSPECTIVE

known salt domes. These are restricted to the offshore in, for example, the southern North Sea, and any potential problems might need to be evaluated if such deposits are to be exploited. The depth and development of wet rock head in onshore salt deposits is reviewed in Evans & Holloway, providing an indication of locations where halite beds are probably too shallow to be developed for gas storage purposes. For salt cavern storage, therefore, an important pre-storage stage is the mechanical integrity test (MIT). During this stage of development, the gas tightness of the newly created cavern (in effect a large pressure vessel) is tested by raising the pressure in the cavern and shutting-in the well to observe any variations in pressure over a period of time. Pressure cycling is also required to subject the cavern to storage conditions and test for damage to the cavern walls. The selection of maximum pressure is based upon rock mechanical tests of the halite, the depth of the cavern (thickness of overburden) and temperature. The depressurization rate during withdrawal must also be carefully calculated in order to avoid inducing large tensile stresses that can be damaging to the rock formation or cemented wells. A variety of integrity tests are available with the most widely used method worldwide being the In Situ Balance method (ISB). However, this can be associated with errors. Tryller, Reitze & Crotogino present a method (SoMIT) based on ultrasonic techniques in which the interface depth, the temperature and the differential pressure at the interface depth can be measured continuously during the tightness test, thereby achieving much greater levels of accuracy to verify gas tightness than was previously the case. Leakage of stored gas cannot always be ruled out and operators accept there are likely to be losses from storage, with operations and budgets calculated accordingly. From the economic viewpoint, viability of storage depends fundamentally on the speed of the stock rotation and the nature of the products stored. When storing compressed air to absorb daily excess electric power, for example, a loss of 1% per day is considered reasonable. But when storing oil for strategic reasons, a loss of 1% per year would be regarded as a maximum acceptable value (Be´rest et al. 2001). From the safety viewpoint, it must be shown that any hydrocarbon product migrating from storage does not accumulate to dangerous levels in other strata, or leak to surface where it may build up in buildings. It is, therefore, important to have detection equipment and procedures in place for dealing with any leakage of stored product. As described above, monitoring of the natural gas aquifer storage facility at Stenlille, Denmark not only illustrates the importance of monitoring with respect to safety, but also to

7

environmental issues (Laier & Øbro). The facility, operational since 1989, provides an important example of the role of baseline studies and ongoing monitoring for gases following the commencement of storage operations. The soil gas monitoring programme at the Weyburn Oilfield (Riding & Rochelle) where CO2 is injected both for storage and for enhanced oil recovery, has obtained baseline and post-injection soil gas survey data. These indicate a shallow biological origin for the measured CO2 in soil gases and no evidence for leakage of the injected CO2 to ground level. Furthermore, the long-term safety and performance of CO2 storage was assessed by the construction of a ‘features, events and processes’ (FEP) risk database that provides a comprehensive knowledge base for the geological storage of CO2. A number of incidents at underground fuel storage (UFS) facilities have led to fatalities, casualties and damage to property, either connected to the storage facility or to surrounding developments (Evans). The incidents have been highlighted by objectors to the development of UGS facilities and are used to question the safety of virtually all proposed UGS facilities onshore UK. This volume carries a paper by a Local District Councillor (Davidson) who has been involved in the process of assessing a proposal for UGS at the depleting Welton Oilfield. Davidson is able to convey first hand experience of not just the current planning and application process, but the feelings and fears expressed by local residents opposed to such developments in close proximity to their communities. Regarding the relative safety of UFS and UGS, Evans provides an extensive literature and websourced review of documented incidents or problems that have been encountered at UFS facilities. The numbers of incidents, fatalities and casualties reported at UFS facilities are compared with those sustained in other areas of the energy supply chain, particularly in connection with above ground infrastructure. The casualty figures for UGS are orders of magnitude smaller than those found in other areas of the energy supply chain. This perhaps lends support to claims that salt caverns provide the safest form of storage for large volumes of hydrocarbons (Be´rest et al. 2001; Be´rest & Brouard 2003), and that underground gas storage as well as oil and gas production can be conducted safely if proper procedures are followed (Chilingar & Endres 2005). Far graver consequences and much higher death rates are associated with incidents at, for example, above ground fuel storage tanks (Persson & Lo¨nnermark 2004) or elsewhere in the energy supply chain (Evans). These figures may help to allay some of the fears expressed by local residents opposed to UGS facilities.

8

D. J. EVANS & R. A. CHADWICK

Other gas storage scenarios Pore storage and salt cavern facilities offer potential for the storage of a number of other gases, including compressed air energy storage (CAES), possibly linked to renewable sources such as wind or water. The technological concept of CAES is more than 30 years old (Glendenning 1981), with the first CAES facility commissioned in Germany in 1978, using caverns created in the Huntorf salt dome near Hamburg for storage (Glendenning 1981; Thoms & Gehle 2000; Crotogino et al. 2001; Cheung et al. 2003). Hydroelectric power plants have, for many years, been used to store excess off-peak (night-time and weekends) power and provide increased peak time output. CAES facilities likewise provide the potential to store energy and could be used alongside, for example, wind turbines. Though instances of this technology are not numerous, it is likely that CAES will assume a greater importance as energy markets evolve. If widespread renewable energy is to become reality, then the utility industry might have to consider more options for energy storage including compressed air (Schaber et al. 2004). Distributed generation and microgrids in which small CAES plants play an important role might facilitate such a system. Although a highly mobile molecule, hydrogen may also be stored underground, with former brine caverns already in use on Teesside. Stone et al. examine the potential for large-scale underground hydrogen storage in halite (rock salt) deposits onshore in the UK. They consider the technical, geological and physical issues of storage, the locations of salt deposits and both legal and economic aspects. Given the greater flexibility in terms of injection and withdrawal rates, most potential for compressed air and hydrogen storage is probably provided by salt caverns. However, storage in highly porous and permeable rocks is possible and is being considered in the UK. The UK Government has recently announced targets for stringent reductions in the UKs greenhouse gas emissions in the coming decades. In addition to this, the recent renewed focus on the security of energy supply has raised the likelihood that a new generation of coal-fired power stations will be built. For such a step to be environmentally viable, clean-coal technologies with near-zero greenhouse gas emissions will be required. A key component of this strategy is the large-scale deployment of underground CO2 storage. A number of current research and demonstration projects are investigating the injection and long-term underground storage of CO2 worldwide. Most current research and operational programmes are focused on pore-space storage, including depleted oil fields

(where CO2 might provide a period of enhanced oil recovery), gas fields and saline aquifers. Aspects of two industrial-scale CO2 injection projects currently in operation and using reservoir pore space are described, with an emphasis on how such storage sites can be monitored to ensure storage security and safety. Both studies illustrate the feasibility of underground storage and how injected volumes can be monitored and verified with the application of sophisticated geophysical and geochemical techniques. Anthropogenic CO2 is being injected as part of a commercial enhanced oil recovery (EOR) operation into a carbonate reservoir at Weyburn in Canada. Riding & Rochelle describe various aspects of the storage operation related to containment integrity: geological characterization, long-term geochemical performance of the caprock and results from a suite of monitoring surveys. A diverse portfolio of potential monitoring tools is available for monitoring CO2 storage sites, some tried and tested in the oil industry, others as yet unproven. Chadwick et al. describe the type of monitoring programme that may be required by future regulatory regimes to prove efficacy in emissions reduction and to ensure site safety and integrity. Specific reference is made to monitoring at the ongoing CO2 injection and storage operations at the Sleipner field in the North Sea. Here time-lapse 3D seismic and time-lapse gravimetry are proving successful in imaging and characterizing the CO2 plume in the storage reservoir. It is concluded that the technical nature and duration of storage site monitoring programmes are likely to be highly site specific, but they are essential to provide an acceptable basis for site closure.

Summary In the current energy climate, with rising concerns about security of supply, environmental degradation and rising costs, this publication is most timely. The development of UGS, underground CO2 storage and the storage of other gases including compressed air and halogen, will play a key role as energy technology evolves in the coming decades. Many of the issues that arise during the planning and construction phases of UGS sites are covered, including the important and related aspects of safety and public confidence. The lessons learned from ongoing research into CO2 storage may provide valuable input into the planning, development and ultimately the abandonment procedures of UGS facilities that are only just being developed in the UK. This publication should, therefore, prove of interest to developers, planners and local communities as we encounter new issues in the energy landscape of the twenty-first century.

INTRODUCTION AND UK PERSPECTIVE

We are grateful to M. Stephenson and R. Evans for their helpful and constructive comments and reviews of the manuscript. This paper is published with the permission of the Executive Director, British Geological Survey, Natural Environment Research Council.

References A LLISON , M. L. 2001a. Hutchinson, Kansas: a geologic detective story. Geotimes, American Geological Institute, October, 2001. A LLISON , M. L. 2001b. The Hutchinson Gas Explosions: Unraveling a Geologic Mystery. Kansas Bar Association, 26th Annual KBA/KIOGA Oil and Gas Law Conference, 1, p3-1 to 3-29. B E´ REST , P. & B ROUARD , B. 2003. Safety of salt caverns used for underground gas storage. Oil & Gas Science and Technology, 58, 361– 384. B E´ REST , P., B ROUARD , B. & D URUP , J. G. 2001. Tightness tests in salt cavern wells. Oil & Gas Science and Technology, 56, 451 –469. BS. 1998. BS EN 1918-3: Gas Supply Systems — Underground Gas Storage — Part 3: Functional Recommendations for Storage in Solution-Mined Caverns. British Standards Institution, London. C HEUNG , K. Y. C., C HEUNG , S. T. H., N AVIN DE S ILVA , R. G., J UVONEN , M. P. T., S INGH , R. & W OO , J. J. 2003. Large-scale Energy Storage Systems. Imperial College Research Report, ISE2 2002/2003. World Wide Web Address: http://www.doc.ic.ac.uk/~mpj01/ ise2grp/energystorage_report/storage.html. C HILINGAR , G. V. & E NDRES , B. 2005. Environmental hazards posed by the Los Angeles Basin urban oilfields: an historical perspective of lessons learned. Environmental Geology, 47, 302–317. C ROTOGINO , F., M OHMEYER , K.-U. & S CHARF , R. 2001. Huntorf CAES: More than 20 years of successful operation. Paper presented at Spring 2001 Meeting, Orlando, Florida, USA, 15–18 April 2001. World Wide Web Address: AKE2003H03c_Crotogino_ea_ HuntorfCAES_CompressedAirEnergyStorage.pdf. DCLG. 2007. Planning for a Sustainable Future. White Paper Presented to Parliament by The Secretary of State for Communities and Local Government (DCLG), The Secretary of State for Environment, Food and Rural Affairs, The Secretary of State for Trade and Industry, The Secretary of State for Transport, 21 May 2007. World Wide Web Address: http://www.communities.gov.uk/documents/ planningandbuilding/pdf/planningsustainablefuture. DCLG. 2008. Letter dated 6 February 2008 to Kim Eccles from departments of Communities and Local Government (DCLG) and Business and Regulatory Reform (DBERR) regarding Town and Country Planning Act 1990, Sections 78 & 266 Appeals by Caythorpe Gas Storage Ltd. Caythorpe Well Site, Rudston, YO25 4JD. Appeal Refs: APP/E2001/A/06/2024097 & APP/E2001/A/07/2037845. World Wide Web Address: http://www.communities.gov.uk/ documents/planningcallins/pdf/686469.pdf DEFRA. 2004. Fuel Poverty in England: The Government’s Plan for Action. Department for Environment,

9

Food and Rural Affairs (DEFRA). World Wide Web Address: http://www.defra.gov.uk/environment/ climatechange / uk/ household/fuelpoverty/pdf/ fuelpov_actionplan.pdf. DTI. 2003. Our Energy Future — Creating a Low Carbon Economy. Department of Trade and Industry (DTI), February 2003. World Wide Web Address: http:// www.berr.gov.uk/files/file10719.pdf. DTI. 2006a. Understanding Gas Supply this Winter: A Pack to Help Improve Understanding of the Key Issues and Where to Find Out More. Department of Trade and Industry (DTI), January 2006. World Wide Web Address: http://www.berr.gov.uk/files/ file27821.pdf. DTI. 2006b. The Energy Challenge: Energy Review Report. Department of Trade and Industry (DTI), July 2006. World Wide Web Address: http://www. berr.gov.uk/files/file31890.pdf. DTI. 2006c. Offshore natural gas storage and liquefied natural gas import facilities: Improving the regulatory framework for offshore natural gas storage and offshore LNG unloading — a consultation. Department of Trade and Industry (DTI), November 2006. World Wide Web Address: http://www.berr.gov.uk/files/ file35073.pdf. DTI. 2006d. Energy statement of need for additional gas supply infrastructure: Ministerial Written Statement, 16 May 2006. Secretary of State for Trade and Industry (DTI). World Wide Web Address: http://www.berr. gov.uk/files/file28954.pdf. DTI. 2007a. Meeting the Energy Challenge: A White Paper on Energy. Department of Trade and Industry (DTI), May 2007. World Wide Web Address: http:// www.berr.gov.uk/files/file39387.pdf. DTI. 2007b. Government response to public consultation: Offshore natural gas storage and liquefied natural gas import facilities. Department of Trade and Industry (DTI), May 2007. World Wide Web Address: http:// www.berr.gov.uk/files/file38982.pdf. D USSEAULT , M. B., B ACHU , S. & D AVIDSON , B. C. 2001. Carbon Dioxide Sequestration Potential in Salt Solution Caverns in Alberta, Canada. Solution Mining Research Institute, Fall 2001 Technical Meeting Albuquerque, New Mexico, USA, October 7– 10. D USSEAULT , M. B., R OTHENBURG , L. & B ACHU , S. 2002. Sequestration of CO2 in Salt Caverns. Petroleum Society, Canadian Institute of Mining, Metallurgy & Petroleum. Canadian International Petroleum Conference 2002, Calgary, Alberta, Canada, 11–13 June 2002. Paper 2002–237. F ERNANDO , A. 2005. Gas storage: perspectives on investment in the new paradigm. Presentation to IEA seminar, June 2005. World Wide Web Address: http:// www.iea.org/textbase/work/2005/Gashubs/Hub_4.2. pdf. G EOLOGICAL S OCIETY 1985. Photographic feature: Liquified petroleum gas caverns at South Killingholme. Quarterly Journal of Engineering Geology & Hydrogeology, 18, ii –iv. G LENDENNING , I. 1981. Compressed air storage. Physics in Technology, 12, 103– 110. H OUSE OF L ORDS . 2004. Gas: Liberalised Markets and Security of Supply. House of Lords, European Union

10

D. J. EVANS & R. A. CHADWICK

Committee 17th Report of Session 2003– 2004. World Wide Web Address: http://www.publications.parliament.uk/pa/ld200304/ldselect/ldeucom/105/105.pdf. HSE. 2006a. Salt cavity and natural gas storage – consent and operational issues. Health and Safety Executive Gas and Pipelines Unit Paper, February 2006. World Wide Web Address: http://www.hse.gov.uk/supply/ saltcavity.htm HSE. 2006b. The health and safety risks and regulatory strategy related to energy developments: An expert report by the Health and Safety Executive contributing to the Government’s Energy Review, 2006. Health and Safety Executive (HSE) report, 28 June 2006. World Wide Web Address: http://www.hse.gov.uk/ consult/condocs/energyreviews/energyreport.pdf HSE. 2008. Gas Supply Emergencies: The Health & Safety Executive’s (HSE) Role in Gas Supply Emergencies. Wide Web Address: http://www.hse.gov.uk/gas/ supply/emergencies.htm. IGU. 2006. International Gas Union (IGU) Website on Underground Gas Storage Facilities. World Wide Web Address: http://www.igu.org/html/wgc2006/ WOC2database/Excel/Report_Tab_Summary_UGS_ Key_Data_2006_in_operation_english.xls. I NEOS . 2006. I NEOS Enterprises secures planning permission for new strategic UK gas storage facility in Cheshire. I NEOS press release, 29 June 2006. World Wide Web Address: http://www.ineos.com/new_ item.php?id_press=132. LCC. 2006. Decision on Welton Gas Storage Application by LCC’s Regulatory Committee on 22 February 2006. Lincolnshire County Council (LCC). World Wide Web Address: http://www.planningportal.gov. uk/england/professionals/en/1115313894518.html. L EITH , W. 2001. Geologic and Engineering Constraints on the Feasibility of Clandestine Nuclear Testing by Decoupling in Large Underground Caverns. U.S. Geological Survey Open File Report 01-28 (this report can be downloaded from the USGS Eastern Region Publications World Wide Web Address: http://geology.usgs.gov/open-file/2001.html or the USGS Eastern Region Earth Surface Processes Publications World Wide Web Address: http://geology.er. usgs.gov/eespteam/EESPT_PUB.html. L OWERY , C. 2006. The Government’s Energy Review — Increasing dependence on gas imports. EIC Letter to the Energy Review Team, Department of Trade and Industry, 31 March 2006. Energy Information Centre Ltd. World Wide Web Address: http://www.berr. gov.uk/files/file30786.pdf. N ATIONAL G RID , 2007. Gas Transportation Ten Year Statement 2007. World Wide Web Address: http:// www.nationalgrid.com/NR/rdonlyres/F085FC32-8C534999-AF88-80388A29AE2C/22103/TYS2007.pdf. N EWMAN , A. S. 2007. Town and Country Planning Act 1990, Planning (Hazardous Substances) Act 1990, Acquisition of Land Act 1981, Gas Acts 1965 and 1986, East Riding of Yorkshire Council, Appeals and Applications by Caythorpe Gas Storage Ltd. Report to the Secretary of State for Communities and Local Government and the Secretary of State for Trade and Industry, 6 July 2007. World Wide Web Address: http://www.communities.gov.uk/documents/ planningandbuilding/pdf/685826.pdf.

N ISSEN , S. E., W ATNEY , W. L., B HATTACHARYA , S., B YRNES , A. P. & Y OUNG , D. 2004. Geologic Factors Controlling Natural Gas distribution related to the January 2001 gas explosions in Hutchinson, Kansas. Poster display, American Association Petroleum Geologists 2004 — 3 posters. World Wide Web Address: http://www.kgs.ku.edu/ PRS/publication/2004/AAPG/NG_Migration/index. html. ODPM, 2005. Planning Policy Statement 1: Delivering Sustainable Development. Office of the Deputy Prime Minister, HMSO. World Wide Web Address: http:// www.communities.gov.uk/documents/planningandbu ilding/pdf/planningpolicystatement1. P ERSSON , H. & L O¨ NNERMARK , A. 2004. Tank fires: Review of incidents 1951– 2003. SP Fire Technology, B RANDFORSK Project 513-021 SP Report 2004:14, SP Swedish National Testing and Research Institute, Bora˚s, Sweden. P IESSENS , K. & D USAR , M. 2003. CO2 sequestration in abandoned coalmines. Proceedings of the International Coal Bed Methane Symposium, May 5 –9, Tuscaloosa, Alabama. Paper No. 346. POST. 2004. The future of UK gas supplies. The Parliamentary Office of Science and Technology (POST) Postnote, October 2004, Number 30. World Wide Web Address: http://www.parliament.uk/documents/ upload/POSTpn230.pdf. R AVEN R IDGE R ESOURCES , 1998. Gas storage at the abandoned Leyden Coalmine near Denver, Colorado. Report prepared under work assignment 3-1 of the US Environmental Protection Act Contract 68-W50018 by Raven Ridge Resources, Incorporated and Penn, Stuart and Eskridge. World Wide Web Address: http://www.cmpdi.co.in/cbm/CBM%20 papers/Mine%20Studies/stu001.pdf. S CHABER , C., M AZZA , P. & H AMMERSCHLAG , R. 2004. Utility-scale storage of renewable energy. Electricity Journal, 17, 21–29. S HI , J. Q. & D URUCAN , S. 2005. CO2 storage in caverns and mines. Oil & Gas Science and Technology, 60, 569–571. S TAR E NERGY . 2006. Gas Storage Update. Star Energy plc Press Release, 1118 October 2006. World Wide Web Address: http://www.starenergy.co.uk/fileadmin/ pdfs/Gas_Storage_Press_Release_18_10_06.pdf. T HOMS , R. L. & G EHLE , R. M. 2000. A brief history of salt cavern use (keynote paper). In: G EERTMAN , R. M. (ed.) Proceedings of 8th World Salt Symposium, part 1, Elsevier B.V. 207– 214. T ROTTER , J. T., T HOMPSON , D. M. T. & P ATERSON , T. J. M. 1985. First Mined Underground Storage in Great Britain. Tunnelling 85. Institution of Mining and Metallurgy, London. V EIL , J. A., S MITH , K. P., T OMASKO , D., E LCOCK , D., B LUNT , D. L. & W ILLIAMS , G. P. 1998. Disposal of NORM – contaminated oilfield wastes in salt caverns. Report prepared for the US Department of Energy, Office of Fossil Energy, contract W-31-109Eng-38. Argonne National Laboratory. W ARD , J., C HAN , A. & R AMSAY , B. 2003. The Hatfield Moors and Hatfield West Gas (Storage) Fields, South Yorkshire. In: G LUYAS , J. G. & H ITCHENS , H. M. (eds) United Kingdom Oil and Gas Fields,

INTRODUCTION AND UK PERSPECTIVE Commemorative Millennium Volume. Geological Society, London, Memoir, 20, 905–910. W ATNEY , W. L., N ISSEN , S. E., B HATTACHARYA , S. & Y OUNG , D. 2003. Evaluation of the role of evaporite karst in the Hutchinson, Kansas Gas Explosions, January 17 and 18, 2001. In: J OHNSON , K. S. & N EAL , J. T. (eds) Evaporite Karst and Engineering/

11

Environmental Problems in the United States. Oklahoma Geological Survey Circular, 109, 119–147. W EBSTER , P. & E LLIOT , F. 2008. Planning rebels pacified but there’s trouble ahead over road duty. F r o m T h e T i m e s, June 26, 2008. World Wide Web Address: http://www.timesonline.co.uk/tol/news/politics/ article 4215035.ece.

The importance of gas storage to the UK: The DECC perspective J. HAVARD* & R. FRENCH Energy Markets Unit, Department of Energy and Climate Change (DECC), formerly Business, Enterprise and Regulatory Reform (BERR), London *Corresponding author (e-mail: [email protected])

A changing gas supply picture The UK economy faces a major challenge; our indigenous gas supplies are in decline and we are moving towards increasing import dependence on gas. By the end of the decade, the UK will have an import dependency of around 30%, and by 2020 it could rise to some 80%. To manage this change, new gas supply infrastructure is needed to increase our capacity to import, store and transport gas efficiently. A regulatory environment that enables the development of timely and appropriately sited infrastructure projects is vital. The need for increased gas supply infrastructure, and a regulatory environment to allow such infrastructure to be delivered to the market in a timely fashion, was set out by the Government in the Energy White Paper of 2003 (DTI 2005a): Our Energy Future — creating a low carbon economy. It identified four challenges, one of which was securing the reliability of energy supplies. This remains integral to an energy policy that meets the needs and expectations of all energy consumers. It was considered as part of the DTI’s Energy Review (DTI 2006a) and will be addressed again in the forthcoming Energy White Paper. It is clear that any weakness in infrastructure could push up gas prices, or result in interruptions to supply, with harmful consequences for both UK industry and UK consumers.

The importance of gas storage A central feature in creating a reliable gas market is balancing physical supply and demand. There are a number of reasons for this: (1) there is a very strong seasonal element in gas demand (largely arising from household demand for central heating purposes in winter) — the ‘swing factor’, i.e. the ratio of winter peak daily gas demand to annual average daily demand, is over 30%; (2) there is also a strong within-day element in gas demand: gas demand is not ‘flat’ over each period of 24 hours, but, especially in winter,

tends to peak in the evening and to fall in the very early morning; (3) gas ‘fails to danger’ (i.e. in the event of a supply failure there is a risk of air entering the gas supply pipes thereby creating a highly explosive and dangerous situation); because of this it is highly important to maintain continuity of supply to (especially) the household market; (4) and yet gas (unlike electricity) travels along the supply system to final consumers relatively slowly; this increases the difficulty of balancing gas supply and demand at the point of consumption, and therefore adds a geographical element to the balancing problem. We have three kinds of tool for providing the flexibility to balance the physical gas market:

On the supply side † †

by varying (increasing or decreasing) supplies of gas into the market; by drawing on close-to-market gas storage facilities;

On the demand side †

through demand management; in practice, by relying on the commercial arrangements between gas market players and gas consumers to reduce gas demand at times of tight supply.

There is some demand-side flexibility, but the much higher winter demand compared to summer demand means there is inevitably an important role for the supply side in balancing the market in the winter, and thereby preventing a supply failure that could result in interruptions to users or even air entering the gas supply system, creating a potentially dangerous situation. Therefore close-to-market storage infrastructure is substantially more important to us now than it was when we had significant indigenous supplies. Our gas production from the North Sea previously reduced our need for gas storage, as compared to some of our European neighbours who lack indigenous supplies. Now, to replace the capability of the southern North Sea gas fields and the Morecambe

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 13–15. DOI: 10.1144/SP313.2 0305-8719/09/$15.00 # The Geological Society of London 2009.

14

J. HAVARD & R. FRENCH

field in Morecambe Bay, which have traditionally provided increased gas supply to meet seasonal peaks in demand (e.g. in winter), we need more storage than we have required in the past. However, the storage of gas onshore and offshore is only possible in certain geological structures, which are present in a limited number of locations in Great Britain: the main such structures are salt formations and partially depleted oil and gas fields. (In continental Europe aquifers are also used for underground gas storage purposes.) The kinds of gas storage project that these structures support are significantly different. Salt caverns typically hold a much smaller volume of gas than partially depleted petroleum fields, but they have an advantage in their response time — the gas can be withdrawn very quickly. On the other hand, partially depleted petroleum fields typically have a very large storage space; and (for example) the important ‘Rough’ storage field, off the Yorkshire coast, is an important seasonal storage facility, which fills in summer and is used to provide ‘endurance capability’ in winter. Due to geological limitations, applications from developers to construct such facilities may be more common in some parts of Great Britain than others. Although such facilities may not always appear to convey a local benefit, they do provide crucial national benefits, in which all localities share. In particular, they add to the reliability of national energy supply, from which every user of the system benefits.

The development of the UK gas storage market The market is responding actively to the challenge of increased import dependency, with actual and planned investment in gas import infrastructure, storage and related transportation of some £10 billion over the period 2005–2010. The projects have the potential to make a real difference to our gas supply infrastructure; by 2010, our storage capacity could more than double, and our import infrastructure is planned to more than triple. However, all this depends on a regulatory environment that enables the development of timely and appropriately sited new gas supply infrastructure projects. The current consents regime is only now starting to deal with this new tranche of gas supply infrastructure projects; and it is important to consider how avoidable delays may be prevented, both now and in the future, to ensure that these projects, and those that follow them, can commission on time if approved. This means a planning consent regime that offers more clarity for developers about processes and timescales, thereby contributing to a lower overall level of risk for developers.

DTI’s focus is therefore on reducing the regulatory barriers to maximizing gas supply, a view shared by the Trade and Industry Committee (DTI 2005b), and in line with this Government’s focus on better regulation. Measures to help streamline and simplify onshore gas infrastructure consents regimes are being considered and will be consulted on in due course. Measures to improve the offshore gas infrastructure consents regime are currently under consultation (DTI 2007). In addition, we have identified and are trying to meet a need amongst local decision makers for improved information about the changing gas supply situation so that they may make decisions on applications.

Benefits to end-consumers Ultimately an insufficiency of gas storage facilities could lead to a physical gas supply shortage. This could have serious consequences for endconsumers. Shortages have always been a possibility, particularly on the coldest winter days. Fortunately so far there has never been a serious overall gas supply shortage. But it would be wrong to be complacent. A shortage could result from such factors as disruption elsewhere in the physical gas supply chain, or severe winter weather. It is significant that 37% of electricity is currently generated from gas (DTI 2006b), and that, at times of shortage of gas, power stations may not continue to operate because they are such large consumers. Consequently, a gas supply failure could also have a serious impact on electricity customers. In the medium term, our dependency on gas is unlikely to diminish significantly, even though use of renewables will probably increase. Around 90% of the UK’s energy needs are currently met by fossil fuels, and they will continue to be the dominant source of energy for decades to come. Importing gas from further afield through lengthened supply chains brings with it risk for the UK. Gas storage and related infrastructure will help us meet this challenge in the decades ahead. In summary, we need timely and appropriately sited gas supply infrastructure, including gas storage to be delivered by the market, because: †

† †

Great Britain is becoming increasingly dependent on gas imports, and requires new gas supply infrastructure to help ensure security of supply; new projects enable extra supply and storage options in a timely fashion if they proceed without avoidable delays; there are limited locations currently suitable for much needed gas storage projects, and they support different kinds of storage facility;

UK GOVERNMENT PERSPECTIVE

†

†

gas storage is needed to enable effective balancing of the gas market in the winter, ensuring that slow-moving gas is available close to market when consumers require it; and new energy infrastructure projects provide national benefits, shared by all localities.

References DTI. 2005a. Energy White Paper, February 2005, Cm. 5761, Section 6.51. The Stationary Office.

15

DTI. 2005b. Trade and Industry Committee First Report on Security of Supply, 13 December 2005. DTI. 2006a. Energy Review report, ‘The Energy Challenge’. DTI, July 2006. http://www.dti.gov.uk/files/ file31890.pdf. DTI. 2006b. Energy Review: A Report. DTI, July 2006, p. 25. DTI. 2007. Government response to public consultations: offshore natural gas storage and liquified natural gas import facilities. Department of Trade and Industry (DTI), May 2007. World Wide Web Address: http:// www.berr.gov.uk/files/file38982.pdf

Gas storage: An onshore operator’s perspective A. FERNANDO1,2* & A. RAMAN1 1

Star Energy Group plc, Ground Floor, Burdett House, 15/16 Buckingham Street, London WC2N 6DU

2

Present address: Ecosecurities, 1st Floor, Park Central, 40/41 Park End Street, Oxford OX1 1JD *Corresponding author (e-mail: [email protected]) Abstract: Operating 28 onshore oil and gas fields, Star Energy Group plc is the leading onshore operator in terms of fields. The company believes that with the decline in flexible gas production facilities close to market in the North Sea and Irish Sea, Great Britain (GB) will need a significant expansion in gas storage capacity. To meet these demands, Star Energy aims to construct underground gas storage (UGS) facilities by converting depleted onshore hydrocarbon reservoirs both on- and offshore GB. This paper provides an overview of the UK position with respect to UGS, Star Energy’s strategy and their planned developments.

Star Energy Group plc operates 28 UK onshore oil and gas fields, making it the leading onshore operator in terms of fields. The company’s primary focus is to establish a significant multi-site gas storage business. This gas storage business will be developed by converting suitable, partially depleted, onshore hydrocarbon reservoirs into gas storage facilities. In addition, Star Energy is looking at acquiring partially depleted reservoirs offshore that will further augment its vision for a multi-site gas storage business. Star Energy believes that with the decline in flexible gas production facilities close to market in the North Sea and Irish Sea, Great Britain (GB) will need a significant expansion in gas storage capacity. The company considers that an optimum mix of offshore and onshore gas storage in partially depleted hydrocarbon reservoirs is the most economic and efficient method of expanding gas storage capacity in GB. The gas storage facilities currently in use in the UK have been formed by the conversion of depleted hydrocarbon reservoirs, by the leaching out of salt caverns and by storing liquefied natural gas (‘LNG’) in peak shaving facilities at the extremities of the gas network.

Storage duration, short-term and seasonal storage The duration of a storage facility is defined as the amount of storage space divided by the capacity to deliver gas. In general, LNG peak shaving facilities have storage cycles (durations) of just a few days, salt cavity storage facilities have cycles of between

5 and 30 days, and storage cycles within depleted field storage facilities are up to about 100 days (Plaat 2009). Each type of facility has its own load curve and is able to meet differing demands (Fig. 1). Storing gas means that it is possible to average out the seasonal variations in gas demand by injecting gas into storage during the summer period, when gas demand is lowest, and storing it for withdrawal during the winter period, when gas demand is highest. Figure 1 is derived from National Grid data and illustrates Long Range (i.e. seasonal) Storage (LRS), which comprises a pattern of injection over the summer and withdrawal over the winter, with variations to take into account demand/price levels. Many storage facilities are also used to smooth out short-term variations in the supply and demand for gas. These variations can either be within day, day-to-day or week to week etc. Some facilities inject or withdraw gas at very short notice to balance their customers’ actual gas supply and demand positions and to help maintain the integrity of the gas network. The medium range storage (MRS) and short range storage (SRS) lines — Figure 1 illustrates the pattern of injection and withdrawal for medium (MRS) and short-range storage (SRS) respectively. The various types of storage facility provide flexibility, with the difference between the medium and long-range storage particularly apparent with each type providing a balanced portfolio of storage volume able to respond to the various short and medium–long term seasonal demands of a system (Fig. 1): the medium range storage permits greater injectivity and withdrawal. Such facilities provide

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 17–24. DOI: 10.1144/SP313.3 0305-8719/09/$15.00 # The Geological Society of London 2009.

18

A. FERNANDO & A. RAMAN

Fig. 1. Patterns of injection and general storage flows for long range (depleted fields), medium range (salt cavern) and short-range (salt cavern and LNG) storage types (based upon National Grid data). The graph illustrates the flexibility provided by the various storage types.

greater flexibility, and a high degree of variation, and are often operated in a different manner to the long-range storage facilities.

Overview and recent developments in the GB gas market Gas is a crucial part of GBs energy mix, which excluding transportation, accounts for over half of primary energy demand (National Grid 2007a). Although primary energy demand has been growing relatively slowly over the past 25 years, the share of gas has grown rapidly due to the expansion of gas-fired power generation (largely at the expense of coal). In 2006, 31% of GBs electricity was generated from gas (BERR 2007a table 5.4). This illustrates that security of gas supply extends beyond the narrow gas sector and into electricity. The use of gas in power generation may allow for this sector of the market to provide an alternative to gas storage. However, this is more likely to be used for peak storage rather than seasonal storage as developers are unlikely to use gas-fired plants to run predominantly in summer. Security of gas supply is important not just because of its role in electricity supply. Unfortunately when gas supply fails, it fails to ‘unsafe’ and requires considerable manpower and time to re-establish supply (House of Lords 2004, p.34);

‘Once gas supplies to consumers are disrupted, every supply point must be shut off at the meter. When sufficient gas is again available, the system must be commissioned with great care — each sub-network; each main; each meter. Necessarily, this is manpower-intensive and time-consuming; and to add to this, UK Safety Regulations require that when a property is reconnected, the fitness of the appliances is also checked.’ Traditionally the GB market has been supplied by highly flexible gas and oilfields located in the North Sea and Irish Sea. However, these fields are now in a period of rapid decline and the GB market is heading for significant dependence on imported gas. National Grid (the operator of the gas National Transmission System) presented a supply demand view (National Grid 2007b figure 4.2B). The graph (Fig. 2) illustrates the rapid decline in indigenous production from the UK continental shelf (UKCS), which along with slowly growing demand, results in a very rapid growth in the gas import requirement, reaching almost 50% by 2011/12, and 80% by the end of National Grid’s ten year planning horizon (National Grid 2007b). However, Figure 2 does not tell the whole story. Although indigenous production has been declining in aggregate terms, it appears that the more flexible sources of production are declining even more rapidly. This is illustrated by published

ONSHORE OPERATOR PERSPECTIVE

19

Fig. 2. UK supply-demand curves and the growing import requirement (National Grid 2007a).

gas production data (BERR 2007b), split between dry gas production (generally representing gas that comes from flexible gas fields) and associated gas production (which in general comes from oil fields, or gas condensate fields where gas is produced in association with oil or condensate and hence tends to be less flexible from the buyers point of view). Figure 3 (BERR 2007a; table F.2) illustrates how the share of (generally more flexible) dry gas production in total gas production has been declining (BERR 2007b). Furthermore, the ‘swing’ from dry gas production has also been declining. According to these data, in the late 1990s winter dry gas production made up almost two thirds of total gas production. By 2005/6 (the latest year for which data are available) it was at 50%. Both of these trends illustrate a declining level of flexibility to meet the seasonal swing between the summer and winter demand. Historical quarterly gas demand, illustrates that despite the expansion of gas fired power generation, demand still has a marked seasonal pattern (Fig. 4). The coldest winter quarter exhibits a demand of approximately twice that of the warmest summer quarter. Data published by National Grid (2007b)

show that in the period since gas market liberalization the winters have tended to be warmer than average, with consequently lower winter demands (Fig. 5). Despite the series of warm winters there has been a significant ‘tightness’ in the winter gas market. The temperatures during the so called ‘cold snap’ at the end of February/early March 2006 were in fact close to (or even lower than) seasonal normal, but the gas price rose extremely rapidly and briefly topped £1/therm due to the outage at the Rough gas storage facility. That said, the winters of 2006–2007 and 2007–2008 have shown actual demand exceeding seasonal normal demand, but prices have been relatively benign due to the introduction of new supply sources from Norway and LNG imports. However, the diversity of supply resources has also led to volatile price swings in the prompt market. This trend raises an issue: are supply constraints the new market drivers as compared to the traditional demand led model? A considerable amount of new importation infrastructure has been planned to meet gas import needs, some of which is under construction. However, it is unclear how much ‘swing’ these new import projects will provide. Unlike dedicated fields that are

Fig. 3. UK gas production illustrating the declining share of dry gas production (based upon BERR 2007).

20

A. FERNANDO & A. RAMAN

Fig. 4. UK quarterly gas consumption figures (based upon BERR 2007a table 4.1).

geographically close to the market, these import sources make available capacity rather than physical gas flows. The physical gas flows may well depend on the price in the GB market relative to alternative markets. One indication of an appropriate level of gas storage is to look at the experience of other large gas markets, particularly those that have a high level of import dependence. Table 1 summarizes data published by the IEA (IEA 2007) illustrating levels of gas consumption, net imports and gas storage capacity for a number of major gas markets. Germany, Italy and France all show a high level of import dependence at 80% or over. They also have storage capacity of between 16% and 24% of their annual gas consumption. The

USA has a far lower level of import dependence, but has gas storage capacity equivalent to 18% of gas consumption, largely because of the size of the country and the distance between the producing and consuming areas. The Netherlands is still a very significant gas exporter, but despite this has a greater proportion of storage capacity than GB. GB is shown to have a very low level of storage capacity in relation to gas consumption, and as noted above is heading towards high levels of import dependence over the decade from 2010 onwards. This would suggest either that all the other major gas markets have far too much gas storage, or that GB needs to build very significant volumes of gas storage as it moves from selfsufficiency to import dependence.

Fig. 5. The seasonal normal and actual demand curves (based upon National Grid 2007b).

ONSHORE OPERATOR PERSPECTIVE

21

Table 1. Levels of gas consumption, net imports and gas storage capacity for a number of major gas markets (based upon IEA 2007) A B C ¼ B/A D E ¼ D/A Annual consumption Net imports/ Net M Storage Storage cap/Annual (bcm) (exports) (bcm) consumption (%) capacity (bcm) consumption (%) United Kingdom USA Germany Italy Netherlands France

102 634 102 81 51 46

2 97 81 68 235 44

The main types underground gas storage There are three main types of underground gas storage: depleted hydrocarbon reservoirs, aquifers and salt caverns. The first two involve the storage of gas in natural porous strata (i.e. within pore spaces in rock). These storage facilities tend to have a relatively large amount of storage capacity but relatively low rates of gas injection and withdrawal. As a result, these facilities are well suited to seasonal gas storage. Salt caverns involve the leaching of salt deposits underground to create a cavern. These caverns tend to have relatively small amounts of storage capacity but with high rates of injection or withdrawal. Salt caverns are, therefore, well suited for peak storage or as trading tools for short-term arbitrages due to demand changes (for example weekend demand versus weekday demand) or weather changes. Not all regions have access to all types of underground gas storage. However, where suitable hydrocarbon reservoirs exist they tend to form the bulk of gas storage capacity. This is because they have a number of inherent advantages. First, ‘they are known to be capable of storing natural gas or oil for geological timescales — in many cases millions of years’ (BGS 2004). They also often require less ‘cushion gas’ than other types of gas storage. Sometimes the storage of natural gas can result in

2 15 79 84 269 96

4 115 19 13 3 11

4 18 19 16 6 24

incremental oil production by re-pressuring the reservoir. Finally, they are commonly close to existing infrastructure and constitute ‘brownfield’ development given the existence of hydrocarbon extraction and processing facilities. Table 2 shows that for many developed gas markets depleted hydrocarbon reservoirs provide the bulk of gas storage capacity (IEA 2007). They are also the most prevalent type of gas storage facility in countries with appropriate geology (Table 3).

Current and future GB gas storage capacity, and obstacles to development Britain has a very low level of gas storage capacity compared to other import dependent gas markets. Like other major gas markets most of Britain’s gas storage capacity is in depleted hydrocarbon reservoirs. However, Britain is different to the other major gas markets analysed here in that not only has it very few gas storage facilities, but it is also highly dependent on a single facility (Rough gas storage facility, see Table 4). Excluding LNG peak shaving facilities in all cases, Britain currently has five operational gas storage facilities, compared to 43 in Germany whose gas market is almost the same size (Table 1) and 15 and 10 in France and Italy

Table 2. Gas storage capacity by storage type and working capacity (based upon IEA 2007)

UK USA France Italy Germany Netherlands

Depleted reservoirs (%)

Salt cavern (%)

Aquifer (%)

Total (%)

89 87 0 100 58 100

11 4 8 0 33 0

0 9 92 0 9 0

100 100 100 100 100 100

Note: excludes LNG peak shaving facilities

22

A. FERNANDO & A. RAMAN

Table 3. Underground gas storage facilities by type of facility (based upon IEA 2007) Depleted reservoir Salt cavern Aquifer Total UK USA France Italy Germany Netherlands

2 318 0 10 14 4

2 30 3 20 0

43 12 8

4 391 15 10 42 4

Note: Excludes LNG peak shaving facilities

onshore underground gas storage facilities. In general, gas storage provides national infrastructure that is important in terms of national security of supply. The local benefits of gas storage are perceived to be limited. The situation is summarized in a memorandum by the British Geological Survey to House of Lords Select Committee on European Union (BGS 2004) ‘There is no obvious upside to living above, or near, an underground gas storage facility, but neither is there any significant downside, as the surface facilities are very unobtrusive, quiet and easily hidden by careful landscaping and/or tree planting.’

Financing respectively and where gas markets are somewhat smaller than Britain (Table 1). The other striking fact is that Britain is highly dependent on a single facility, the Rough gas storage facility, an offshore depleted gas field that accounts for 81% of current storage capacity. In contrast, the largest facilities in Germany, Italy and France account for 22%, 23% and 32% of capacity respectively. Storage is a long lead-time activity due to the time taken for the technical work, followed by the planning process and the detailed engineering and construction processes. Salt caverns require leaching that further add to timescales. In the case of major salt projects (for example the proposed Preesall storage facility) the leaching process may take a decade to complete for all storage caverns (Canatxx 2003, p. 5). Developing gas storage facilities presents a number of differing challenges.

Technical Not all hydrocarbon reservoirs, or salt deposits, are suitable for gas storage. A substantial amount of technical work and investment is required in order to determine the technical viability of potential projects.

Planning Planning has been and continues to present major problems and delays in the development of

The development of a gas storage facility is a capital-intensive exercise. With a liberalized gas market and mandatory third party access (TPA), it may be necessary to seek exemption from TPA to contract the facility to a third party in order to provide a contracted revenue stream against which finance could be raised.

Access to monopoly networks Gas storage projects require access to the gas grid, and may also require reinforcement of the electricity infrastructure. The process for gaining access to the gas grid is both time consuming and subject to some regulatory uncertainty (Ofgem 2005; Star Energy 2005). If reinforcement is required the grid operators are allowed a considerable amount of time to make this capacity available, which may further delay projects.

Gas storage outlook There are a number of gas storage facilities being developed or planned in Britain (Tables 4–6, based upon BERR 2007a). The projects that have received planning permission will potentially add capacity equivalent to about 30% of current capacity. However, given that the gas market is expected to continue to grow, the result is that in 2014, Britain will have approximately the same

Table 4. Existing underground gas storage facilities in Britain Project National Grid LNG Hatfield Moor Holehouse Farm Hornsea Humbly Grove Rough Total

Type

Capacity GWh

Status

LNG4 Depleted field salt cavity Salt cavity Depleted field Depleted field

2806 1260 600 3495 3146 36 117 47 424

In operation In operation In operation In operation In operation In operation

ONSHORE OPERATOR PERSPECTIVE

23

Table 5. Underground gas storage projects with planning permission Project Aldbrough Aldbrough Expansion Byley Stublach Total

Type

Capacity GWh

Status

Salt cavity Salt cavity Salt cavity Salt cavity

4400 4400 1860 5000 15 660

Under construction — Phase 1 onstream Q4’08 Planning permission received May 2007 Under construction — Phase 1 onstream Q4’08 Planning permission received 2006

Table 6. Proposed underground gas storage projects at various stages in the planning application process Project

Type

Capacity GWh

Albury Phase 1 Albury Phase 2 Bletchingly Caythorpe

Depleted field Depleted field Depleted field Depleted field

1751 7825 9850 2600

Gainsborough Portland Preesall Saltfleetby Welton White Hill Total

Depleted field Salt cavity Salt cavity Depleted field Depleted field Salt cavity

1550 10 944 12 396 6566 4761 4596 62 839

amount of gas storage, as a proportion of gas demand, as it has now. Figure 6 (based upon National Grid 2007b) illustrates the evolution of gas storage capacity relative to demand based upon all the projects in Tables 4–6 being granted planning permission and developed on the timescales outlined. For salt cavern projects assumptions (based upon information in the public domain) have been used to model the leaching process (which determines the speed with which capacity is developed). Figure 6 reveals that even if all the projects currently

Status Pre-planning Pre-planning Pre-planning Planning permission rejected in 2006, Public inquiry completed 2007, permission granted February 2008 Pre-planning Planning permission granted May 2008 Planning permission refused October 2007 Planning application submitted Planning application rejected Planning application submitted

identified are developed in a timely fashion, Britain will still have gas storage capacity of less than 9% by 2014; at a time when gas imports are expected to make up around 80% of demand (DTI 2006, 2007).

Conclusions The British gas market is currently undergoing a fundamental shift from self-sufficiency to import dependence. The move from highly flexible local production to import dependence is likely to require a significant expansion of gas storage capacity. Although some projects have come through the planning process, it remains a substantial obstacle. Onshore gas storage in depleted hydrocarbon reservoirs provides the bulk of gas storage facilities by number and capacity where geology allows. This type of gas storage has several advantages and can make a significant contribution to British security of supply.

References Fig. 6. Gas storage relative to total annual demand to 2015 estimated by National Grid (2007b). Note: Excludes LNG peak shaving facilities.

BERR. 2007a. Digest of UK Energy Statistics (DUKES) 2007. Department of Business and Regulatory Reform (BERR). World Wide Web Address: http:// stats.berr.gov.uk/energystats/dukes07.pdf.

24

A. FERNANDO & A. RAMAN

BERR. 2007b. Department of Business, Environment and Regulatory Reform (BERR) website for production statistics. World Wide Web Addresses: http://www. og.berr.gov.uk/information/statistics.htm and http:// www.og.berr.gov.uk/pprs/full_production.htm. BGS. 2004. Memorandum by British Geological Survey to the House of Lords Select Committee on European Union. Gas: Liberalised Markets and Security of Supply. House of Lords, European Union Committee 17th Report of Session 2003–2004, 87–90. World Wide Web Address: http://www.publications.parliament.uk/ pa/ld200304/ldselect/ldeucom/105/105we04.htm. CANATXX . 2003. Proposed Natural Gas Storage Facility, Preesall Saltfield. 2003. Canatxx Gas Storage Limited, Non-Technical Summary of the Environmental Statement. World Wide Web Address: http://www.lancashire. gov.uk/environment/documents/DevC/Extra/02.04. 1415nts.pdf. DTI. 2006. The Energy Challenge: Energy Review Report. Department of Trade and Industry, July 2006. World Wide Web Address: http://www.berr.gov.uk/files/ file31890.pdf. DTI. 2007. Meeting the Energy Challenge: A White Paper on Energy. Department of Trade and Industry, May 2007. World Wide Web Address: http://www.berr. gov.uk/files/file39387.pdf. H OUSE OF L ORDS . 2004. Gas: Liberalised Markets and Security of Supply. House of Lords, European Union Committee 17th Report of Session 2003–2004. World Wide Web Address: http://www.publications. parliament.uk/pa/ld200304/ldselect/ldeucom/105/ 105.pdf.

IEA. 2007. Natural Gas Information: 2007 edition. International Energy Agency (IEA). World Wide Web Address: http://www.iea.org/Textbase/publications/ free_new_Desc.asp?PUBS_ID=1080. N ATIONAL G RID . 2007a. Gas Transportation ten year statement 2007. National Grid Transco. World Wide Web Address: http://www.nationalgrid.com/NR/rdonlyres/ F085FC32-8C53-4999-AF88-80388A29AE2C/22103/ TYS2007.pdf. N ATIONAL G RID . 2007b. Gas Transportation ten year statement 2007. National Grid Transco. World Wide Web Address: http://www.nationalgrid.com/NR/rdonlyres/ ABA258D7-17D2-4357-BCF7-C9C492201806/22104/ TYS_2007Charts.xls. O FGEM . 2005. Gas transmission: new NTS entry points, reserve prices in auctions and unit cost allowances (UCAs). Office of Gas and Electricity Markets (Ofgem) consultation document, May 2005. World Wide Web Address: http://www.ofgem.gov.uk/Networks/Trans/ PriceControls/Transco/Documents1/10847-13905. pdf . P LAAT , H. 2009. Underground gas storage — why and how. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 25– 37. S TAR E NERGY . 2005. Comments on the Ofgem consultation document, May 2005. Star Energy Group plc. June 2005. World Wide Web Address: http://www.ofgem.gov.uk/ Networks/Trans/PriceControls/Transco/ Documents1/ 11214-13905_StarEnergy.pdf.

Underground gas storage: Why and how HANS PLAAT HP Petroleum Engineering Services, Buitenes 51, 9407 CS Assen, The Netherlands Corresponding author (e-mail: [email protected]) Abstract: Although it appears surprising that gas is put back into the ground after expending so much time, effort and money on extracting it in the first place, underground gas storage (UGS) plays an important role in the management of the gas supply chain. UGS has been used effectively for nearly a century to balance the mismatch in gas supply and demand. Its use continues to grow and with the advent of gas market liberalization, additional uses of UGS have been introduced. In several countries some 20–30% of the annual gas consumption is supplied through the use of UGS. This paper provides an overview of the most common use of UGS, the current status of UGS in the world and the main characteristics of the various types of facility: such as gas fields, aquifers and salt caverns. Aspects related to the planning and performance of gas storage facilities are also discussed.

The purpose of gas storage The gas supply chain is characterized by large hourly, daily and annual imbalances between supply and demand. Although on the one hand, producers and transporters prefer to deliver the gas at the same constant rate at all times, users only need gas at certain times, e.g. during cooking, in winter for heating, harvest season for sugar factories etc. Gas storage plays an important role in bridging this gap between supply and demand. A typical daily demand profile for a year is shown in Figure 1. In this example the maximum daily demand is about twice the annual average demand. Without balancing measures, producers and transporters would have to double the production and transport capacity to satisfy demand. This could be extremely expensive if the gas needs to be transported over large distances. Underground gas storage (UGS) allows the full demand to be met, in a case where the supply transport (pipeline) capacity is only slightly larger than the annual average demand (Fig. 2). When demand is larger than the supply capacity, gas is withdrawn from the UGS facility (‘send-out’) and in times of low demand, the facility is refilled by injecting gas back into storage. This example is a typical case of ‘seasonal storage’, whereby the storage is filled during summer months and emptied during winter months. The same principles as described above for seasonal fluctuations during the year are also applicable for hourly fluctuations during a day. In some domestic markets, these fluctuations can even be larger than daily fluctuations over a year. This calls for ‘daily storage’, whereby the gas is injected during

the night and produced mainly during the morning and evening hours. Since daily and hourly gas demand generally has a strong correlation with the ambient temperature, and given the irregular weather pattern in most countries, there will be occasional high peaks, where the demand is even higher then during a normal peak. This calls for the use of so-called ‘peakshaver’ facilities. A peakshaver delivers gas at relatively high rates for a short period of time, usually a few days. These are generally used only at peak periods during winter and need not be refilled immediately. Refilling will then take place during the summer period and can take six months or longer. Storage is also used as a safeguard against supply interruptions. These can be of a technical nature, such as failures of the production facilities or in the pipeline system. In case of cross-border gas transport, political events in the producing or transit countries might result in temporary reduction or total loss of gas supply; an example being the dispute between the Ukraine and Gazprom in the winter of 2008/09. To safeguard against such events, use is made of ‘strategic storage’. The gas in strategic storage may not be used for many years. The design criterion for these facilities is usually based on an interruption of two to six weeks duration. Re-injection at high rate is generally not required. In addition to storage requirements for the market, storage is also sometimes found near or within the gas producing areas. This so-called ‘production storage’ is used by producers to provide a stable gas supply to the pipeline system. In these cases the gas production itself is usually not very constant, such as gas production that is associated

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 25–37. DOI: 10.1144/SP313.4 0305-8719/09/$15.00 # The Geological Society of London 2009.

26

H. PLAAT

Fig. 1. Typical gas demand profile.

with oil production and/or prone to frequent supply interruptions; for example as may be caused by hurricanes in the Gulf of Mexico. In most markets there is a need for all or most of these types of storages and generally one finds that a particular UGS facility combines more than one of the above-mentioned purposes. The purposes described above are the so-called traditional storages uses. With the introduction of market liberalization and the associated ‘unbundling’ of the various components that make up the supply chain. Market liberalization and ‘unbundling’ have occurred at varying times around the

Fig. 2. Illustration of the use of underground gas storage.

world: mid-nineties in the USA, and end nineties in the UK, this century in continental Western Europe and is only just starting in Eastern Europe. With these changes come additional needs for gas storage. In the past, the whole gas supply chain was, for a large part, provided by integrated companies, which were able through total system optimization, to provide efficient balancing of the mismatch in demand and supply. Following unbundling, the gas chain has divided into a large number of independent parts, e.g. producers, shippers, traders, pipeline operators, storages operators and distribution companies. Each of these is responsible for

ROLE OF UGS

its own part of the chain and, in particular for the balancing of its own demand and supply situation. Although the overall storage requirement of the total gas market is still the same as in a nonliberalized market, the fact that individual players now need to take care of their own balancing has led to an increased demand for specialized storage services, notably those of a short-term nature (Plaat 2007). The most recent developments, notably in the USA, are plans to use UGS as an alternative to tank storage at liquefied natural gas (LNG) receiving terminals (McCall 2004). Plans exist to construct several new LNG receiving terminals providing the capability for regasification and immediate storage in salt caverns over the next few years. The use of caverns in such a manner will be completely different to the traditional use of gas caverns for peak shaving, where it is usually sufficient for the working volume to be injected over several weeks or months. At LNG terminals the caverns will be required to receive the total working volume in less than 5 days and permit the production of the full working volume again over a period of 10 to 30 days, in order that they are ready for the arrival of the next LNG carrier. Operating a storage cavern in this manner, especially the rapid pressure changes resulting from the very high injection rates, adds significant complexity to the design of such a facility (Crotogino et al. 2006).

UGS terminology A UGS facility is characterized by the working volume, cushion gas, deliverability, injectability and duration. Throughout the industry these terms are not always used consistently, which can lead to confusion. The working volume, sometimes called the working gas, is the maximum volume of gas available for withdrawal, when the facility has been completely filled. The working gas volume may be cycled more than once a year. Following liberalization, the working gas is usually owned by the clients of the storage facility. The cushion gas or base gas, is gas that stays in the UGS; it is not available for withdrawal and is required to ensure adequate minimum pressure to provide the required deliverability, even when all the working gas has been withdrawn. In aquifer and cavern storages all the cushion gas needs to be injected initially. In caverns, the cushion gas is also required to sustain mechanical stability. In oil and gas fields the cushion gas volume may consist of recoverable and non-recoverable in-situ gas volumes and injected gas volumes. The cushion

27

gas is usually owned or leased by the operator of the UGS facility. The sum of cushion and working gas is often called the inventory. The deliverability is the amount of gas that can be delivered (withdrawn) from a storage facility per unit of time. It is usually constant for a large part of the working volume, but may decline when a large part of the working volume has been produced as pressures in the reservoir or cavern decrease. The relationship between deliverability and the amount of gas in storage is described by the withdrawal profile. The injectability is the complement of the deliverability; it is the amount of gas that can be injected into a storage facility per unit of time. Similarly, an injection profile is used to describe the relationship with the volume of gas in storage. Consequently, the injectability generally declines as the stored amount of gas increases towards the maximum. The duration is the relationship between the working volume and either the deliverability or injectability; it corresponds to the time required to produce the working volume at a rate equal to the deliverability (¼ working volume/deliverability). It is similarly so for the injectability. The duration is an effective parameter to characterize the purpose and/or properties of a facility. The duration of seasonal storages is generally quite long (60 to 100 days), in peakshavers it is relatively short (1 to 20 days) and it is only a few hours in daily storages. There is no commonly accepted definition of the capacity of a UGS facility. Most frequent use of the term is in the context of working volume; sometimes it is used for the maximum inventory, i.e. the sum of working and cushion gas. In the Netherlands it commonly means deliverability. Recently, Guidelines for Good TPA Practice for Gas Storage System Operators (GGPSSO) were published by the European Energy Regulators Group for Electricity and Gas (ERGEG 2005), in which capacity was defined as follows: ‘storage capacity is space (expressed in normal cubic metres or energy), injectability and deliverability (expressed in normal cubic metres or energy per time unit).’ During the 2006 World Gas Conference, the International Gas Union published a glossary of relevant UGS terminology (IGU 2006), which defines capacity as ‘total ability of a storage facility to provide working gas volume, withdrawal rate and injection rate’.

Types of UGS Storage in gas fields Most UGS facilities are constructed in depleted gas fields. This is because their ability to contain the gas

28

H. PLAAT

over a prolonged period of time has been proven and, due to the usual presence of a large number of wells and a long production history, the reservoir and its behaviour is well understood. However, this does not necessarily mean that any gas or oil field is suitable for UGS. Major considerations with regard to the suitability of a gas field for conversion to UGS are: Containment. Although gas reservoirs have proven ability to contain gas on a geological timescale, this does not necessarily mean that when converted to UGS no gas losses will occur. There are several possible reasons for the possible loss of gas: † overfilling and spilling into a nearby geological structure; † poor cement bonds or casing-related failures in existing wells; † reactivation of faults; and † exceeding the capillary threshold entry pressure of the caprock, or even fracturing, when the reservoir pressure is increased above the original reservoir pressure.

being directly proportional to the amount of gas in the reservoir. However, in other reservoirs, the water that is situated adjacent to, and sometimes beneath the gas-bearing part of the reservoir (the aquifer) will flow into the gas-bearing area. In some cases, this flow can be so strong that water will bypass the cushion gas and towards the end of the production season will reach the wells. Gas wells are able to handle limited amounts of water, but too large amounts will ‘kill’ them. In several instances it has been observed that when aquifer water reaches the UGS wells, deliverability declines due to the deposition of scale within the wellbores (Dereniewski 2006). On the other hand, since the aquifer flow in these cases contributes to maintaining the pressure, less cushion gas could be required for this purpose. There is a precarious balance between the amount of cushion gas that is required to prevent too much water reaching the wells and the amount that is sufficient to maintain reservoir pressure at the desired levels.

Well productivity and injectivity. These are strongly dependent on reservoir properties, reservoir depth and tubing size. Although from a purely technical point of view low well productivity does not preclude a reservoir for UGS purposes, the high number of wells required in such a case may make a project uneconomical. For these reasons storage wells tend also to have larger diameter production/injection tubing: 75800 and even 95800 are now commonly used in the more highly productive reservoirs.

Condensate content of the original gas. As dry pipeline gas is injected, and comes in contact with the water present in the reservoir, some water evaporates and humidifies the stored gas. Consequently, when the gas is next withdrawn, it needs to be treated to remove the water before it is returned to the transport pipeline system. If the original gas in the reservoir was rich in condensate, then with a lowering of the pressure during depletion, a significant part will condense out of the gas phase and be left behind in liquid form in the reservoir. As with water, some of this condensate will re-evaporate and will end up in the produced gas stream. Removal of both water and condensate requires a different type of gas treatment plant than that for the removal of water only. For water removal, glycol absorption systems are commonly used and are relatively cheap, with a pressure loss of only a few bars over the treatment facility. Such systems, however, are not suitable for gases containing condensate. In such cases, plants based on the Joule-Thompson effect, so called low temperature separation (LTS) plants, are often used. The disadvantage of this type of plant is a pressure drop of 20 –40 bars, which results in lower productivity per well. Therefore, modern UGS facilities have started to use silica gel adsorption units, which have a similar pressure drop to glycol units, but require higher capital investments than both the glycol and LTS units. The choice is thus whether to use LTS with a large pressure drop or the more expensive silica gel but with a lower pressure drop.

Aquifer strength. Many UGS reservoirs don’t experience significant water flow and behave more or less like a tank, with the reservoir pressure

Distance to existing infrastructure and market. Since large diameter, high pressure pipelines with sufficient capacity are required to transport the gas to and

Size. Obviously the reservoir is limited geologically, and should be large enough to accommodate the working volume and the necessary cushion gas. But it should not be too large; otherwise the amount of required cushion gas becomes excessive and may render the project uneconomical. Reservoir properties. Porosity, permeability and reservoir thickness not only influence the well performance, but are also important for the movement of the gas towards and away from the wells deeper into the reservoir. Gas velocities in a UGS storage reservoir are many times higher than during the original gas production period. Typically, a similar volume of gas to that produced over perhaps ten to twenty years during field depletion, may be produced and injected again in periods of just three to six months during UGS.

ROLE OF UGS

from the UGS facility, proximity to a main pipeline system can be a major economic consideration.

Storage in oil fields When gas is stored in oil reservoirs, use is generally made of a gas cap that is already present above the oil in the reservoir; this can be the primary gas cap or a secondary one formed during oil production. The behaviour of such a UGS facility is very similar to that of a UGS facility in a gas reservoir. An additional benefit that may accrue using the gas cap of an oil reservoir as the storage area, is that a period of higher oil recovery may sometimes be achieved compared to when no gas storage was undertaken. In the storage case, the average pressure in the reservoir can be maintained, while wells producing with a high gas–oil ratio (GOR) do not need to be closed in; instead the excess gas can be re-injected straight away or these high GOR wells are produced only when the UGS is in send-out mode. Oil reservoirs without an initial gas cap have also been used for gas storage. Development of full working volume and cushion gas takes many years, since the gas needs to displace the oil, which as a consequence needs to be produced. In these reservoirs the containment of gas requires special attention. Although such a reservoir is able to contain an oil accumulation, this does not necessarily mean it can also contain gas, even at the same pressure. In some cases the absence of an initial gas cap may have been caused by a caprock that is permeable to gas but not to oil. Any free gas in the reservoir will already have escaped from the structure. A telltale sign of such a situation is the presence of a gas plume in the shallower, overlying formations. Furthermore, although old wells in gas fields were designed to provide gas isolation between the various reservoirs, this is not always the case in oil fields. Hence in oil fields the mechanical condition of old wells could pose a larger risk to gas containment than might be the case in gas fields.

Storage in aquifers An aquifer is a porous reservoir filled with water (generally saline at depth). By injecting gas into this type of reservoir, it can be converted to a gas storage reservoir. Once in operation the behaviour and the operation of the facility is similar to storage in gas reservoirs. However, when selecting an aquifer for UGS, there are additional points that require attention. There is more uncertainty regarding the ability of the reservoir to contain the gas. Although for a gas field this capability has already been proven over millions of years, for an aquifer,

29

containment needs to be newly established. Amongst other things, this requires special core analysis of the caprock and gas injection field tests. The maximum pressure in many gas reservoirs is often limited to the original reservoir pressure; in an aquifer the pressure needs to be higher than the original pressure in order to be able to inject any gas. Furthermore, knowledge of the reservoir characteristics and an adequate model is often lacking, due to the lack of existing wells and a production history. This all means that the lead time to develop UGS in an aquifer is longer than that for depleted gas fields, and that also the upfront costs tend to be higher.

Storage in salt caverns Salt caverns are widely used for high deliverability/ relatively low volume UGS. A salt cavern or cavity is created by dissolving (‘leaching’) the salt, which is present in the subsurface in rock form (halite or rocksalt) using solution mining techniques. A well is drilled into a suitable salt formation and, after running and cementing the last casing, two concentric tubulars ‘leaching strings’ are run in the hole for the purposes of leaching the cavern (Fig. 3). The annulus between the outer leaching string and the casing is filled with either a diesel or nitrogen blanket to prevent leaching of salt around and above the bottom of the casing (casing shoe). To dissolve a cavern in salt, fresh water is circulated into the well through one of the leaching strings and higher density brine produced (withdrawn) through the other leaching string. The shape of the cavern is controlled by changing the depth of these two strings, the depth of the blanket, and the rate and direction of circulation. The shape of the cavern is monitored by periodic running of a sonar survey in the cavern (e.g. von Tryller et al. 2009). Leaching of a cavern takes many months to several years. There are many considerations that need to be fulfilled to make a gas cavern suitable for gas storage. These include: † sufficient size/volume; † short and long term structural stability; † limited volume decreases (wall convergence due to salt creep); and † safe containment of the stored material (no gas losses). These requirements mean that not all salt deposits are suitable for gas storage purposes. The halite deposit should be large enough to accommodate a salt cavern. The preferred location for cavern UGS is in salt domes, such as found in NW Europe and around the Gulf of Mexico. These

30

H. PLAAT

Fig. 3. Salt cavern leaching process and system components.

salt domes are generally more than 1000 m high and several kilometres in diameter. Many salt storage caverns have been created in bedded halite deposits, for instance in France, Canada and the northern and eastern USA, and are under development in China. Bedded halite consists of thinner (100 –300 m) halite beds commonly with interbeds of insoluble material, such as mudstone, sandstone, limestone, dolomite and anhydrite. Due to the thinner salt bed geometries, the volume of these caverns will be smaller than the volumes of caverns in salt domes, but can be maximized by using different shapes with a larger width/height ratio. In salt domes the preferred cavern shape is that of a vertical cylinder several hundred metres high, with a diameter of some 50 –80 m. Such caverns typically have a volume of 300 000 to 700 000 m3, although smaller and larger sizes occur. Caverns in bedded salt usually have volumes in the range 100 000 to 300 000 m3. The stability requirement of the gas caverns puts limitations on the shape, the height and the maximum and minimum operating pressures. Maximum pressure must be below lithostatic pressure, and below the pressure at which the salt would start fracturing. The minimum pressure is governed by the need to keep cavern wall convergence (salt creep) to acceptable levels (see below). The minimum allowable pressure lies usually around 20–35% of the maximum pressure. This

means the cushion gas will be around a quarter to a third of total gas stored in a full cavern, with the working gas comprising around two thirds. The shape of caverns created in halite, can be well controlled during the solution mining process. However, if there are layers of magnesium and/ or potassium salts, these layers are leached preferentially, potentially leading to irregular and unsuitable cavern shapes. The construction of salt caverns for UGS requires adequate geological site characterization to define impurities, anomalous zones, and geological structure. For normal salt mining (extraction) operations, the cavern shape and height are not important and are not strictly controlled. Therefore, caverns created for the sole purpose of salt mining only are not necessarily suitable for gas storage. Salt is a plastic material that will move slowly under large pressure differences (salt creep). This plasticity increases at higher temperatures. Since the pressure in a gas storage cavern is lower than the pressure exerted on the salt by the overlying rocks, the salt tends to squeeze towards the cavern centre, gradually reducing the cavern volume (cavern convergence). For caverns at a depth of 1000–1400 m, the convergence rate is usually below 1% of cavern volume per year. At greater depths both the temperature and the overburden pressure increase, so that the convergence rate increases rapidly. For example, near Harlingen, in

ROLE OF UGS

the Netherlands, in a solution mining operation at depths up to 3000 m, convergence in the order of 70% of the cavern volume per year has been observed. (Breunese et al. 2003). This cavern closes nearly as fast as the salt is mined. To keep cavern convergence within acceptable limits, shallow depths are preferred. On the other hand, since the pressure at which the gas can be stored increases with depth, it is desirable to create a cavern as deep as possible to maximize both the amount of gas that can be stored and the deliverability per cavern. This need to balance different storage factors means that salt caverns for gas storage are most commonly developed at depths less than 1700 m depth, with the optimal range usually between 1000 –1500 m. Recently a plan was announced to create a salt cavern for gas storage in southern England at a depth of around 2400 m (Egdon Resources Plc 2006). In order to control cavern convergence and to minimize the volume of cushion gas, it is planned to maintain the pressure in the cavern at a high level by injecting saturated brine when gas is being produced and withdraw the brine when gas is injected. The brine will then be temporarily stored in a nearby underground sandstone reservoir. This project received planning permission in May 2008 (see Evans & Holloway 2009). For stability reasons there needs to be a sufficiently large distance between storage caverns, generally in the order of 300–400 m. Since even a large cavern, such as those in salt domes, typically has a working volume of only 40 –70  106 m3, several caverns will be required for a sizeable project. The salt deposit should therefore extend over at least several square kilometres to be able to accommodate such a project. The creation of salt caverns will result in a large volume of brine (salt saturated water), which has to be disposed of. The preferred way is to use the brine in salt factories for salt production. Other ways include disposal in surface waters, such as rivers or the sea, although this is often not permitted, or the re-injection into an underground porous layer. Brine disposal limitations have in some cases stopped otherwise economic storage projects. An overview of recent developments in salt cavern storages has been presented by Gilhaus (2007).

Other types of underground gas storage facilities Some abandoned room and pillar mines have been converted for use as gas storage facilities. The first one was a limestone mine near Lawrenceburg, Indiana, in 1952. In 1970, a salt mine near Burggraf–Bensdorf in former Eastern Germany

31

was converted to gas storage service. Gas storage in coal mines has also been tried in the Leyden mine, near Denver, Colorado (1959– 2003) and in two mines in the Belgian Ardennes between 1975 and 2000. Maintaining gas tightness has been a major problem in the coal mines and all three have now been abandoned following gas leakage to surface. Two gas storage facilities have been created in man-made hard rock caverns. Near Ha´je in the Czech Republic, a complex of tunnels was excavated in granite at a depth of around 1000 m and with a total length of 45 km, using the shaft of a nearby abandoned uranium mine for access. Thereafter this tunnel complex was sealed and since 1998, it has been used to store some 54  106 m3 of working gas. In Sweden, gas has been stored in a lined rock cavern (LRC) since 2004. Excavated in granite at a depth of 100 to 200 m, the cavern has a volume of 40 000 m3 and is lined with steel plating to ensure gas tightness. Its working volume is 8.5  106 m3, with a deliverability of 960  103 m3/day (Mansson et al. 2006). There has been a recent proposal to create caverns for gas storage in limestone formations through dissolution by acid, using solution mining techniques in a similar way to which salt caverns are created (Castle et al. 2004). This process is still in the research phase and there are many factors that require further investigation, including gas tightness of the caverns, cavern stability and disposal of the large amounts of CO2 (with attendant greenhouse gas effects) that will be generated by the dissolution of the limestone.

History and statistics Underground gas storage commenced in 1915 with the convertion of a depleted gas field in Welland county, Ontario, Canada, followed the year after by the Zoar field in western New York state, USA. This field is still in use as a UGS facility after 90 years of service. Gradually more and more gas fields in the USA were converted to UGS. In 1942 the first oilfield, Playa del Rey, California, was converted to storage and in 1946 in Kentucky the first storage in an aquifer was created. In Europe gas storage started in Germany in 1953 with the commissioning of the Engelborstel aquifer storage. Gas storage in salt caverns is a more recent development. The first UGS salt cavern was commissioned in 1961 in Michigan (USA), followed by the first facility in Canada in 1964, and in Europe in 1970 (Tersanne, France). By early 2007 there were 628 UGS facilities operational worldwide, with a total

32

H. PLAAT

Table 1. Overview of existing UGS facilities in the world, January 2007 Area

Number of UGS facilities Gas & oil fields Aquifers Salt caverns Other Total

Europe Former Soviet Union USA Canada South America Asia Australia Total

68 36 318 45 2 5 4 478

23 13 45

29 1 26 9

3

81

65

4

1

123 50 390 54 2 5 4 628

Working volume Deliverability (109 m3) (106 m3/d) 79 109 108 20 0.2 1.6 1.0 318

1560 980 2560 315 2 11 10 5440

Planning of storage facilities

Fig. 4. Example of gas demand/temperature relationship.

working volume of 318  109 m3 and a deliverability of 5.4  109 m3/day. Most of these (390) are located in the USA. Table 1 gives an overview of UGS by area.

The functionality that a gas storage facility needs to provide is dependent on its intended use. Usually the storage needs to bridge the gap between a limited flexibility on the supply side, which is usually well known since it is stipulated in the gas supply contracts, and larger fluctuations on the demand side, which are market dependent. The first step in estimating storage requirements is an analysis of the market characteristics. In many markets the gas demand is strongly related to the ambient temperature and commonly shows a strong correlation with it (Fig. 4). Using such correlations and incorporating other non-temperature effects, including weekday/ weekend differences and special demand patterns of large customers, forecasts of annual gas demands can be generated. A useful tool in analysing these

Fig. 5. Load duration curves. (a) and (b) represent examples of different curves.

ROLE OF UGS

Fig. 6. The use of a load duration curve to determine UGS functionality.

gas forecasts is the ‘load duration curve’ (LDC), which represents a graph of the hourly demand for a year sorted in decreasing order of gas demand (Fig. 5). The load factor is defined as the annual average demand divided by the maximum hourly demand. Figure 5a illustrates the effect of different load factors. Even if the load factor is equal, the shape of the LDC can vary, resulting in significantly different storage requirements (Fig. 5b). Figure 6 illustrates how an LDC is used in combination with the flexibility offered by the supply contracts to determine the required UGS functionality. The areas (shaded in the figure) between the LDC and the lines representing the limits of the supply flexibility correspond to the working volume. The maximum difference between the LDC and these lines determines the required deliverability and injectability respectively. Underground gas storage is not the only flexibility measure that can be used. Other possibilities include interruptible sales contracts, linepack, pipestorage, LNG and increase in supply flexibility (production swing). Each of these have a different duration, involving a different working volume/ deliverability relationship, while different types of

Fig. 7. Position of UGS in the load duration curve.

33

UGS facilities also provide different duration and each of these have their own optimal place within the load duration curve. The LDC is usually met in the higher end with the shortest duration facilities and from the lower end with those having the longest durations. Figure 7 illustrates an ‘ideal’ mix of facilities, based on their duration. The above description has been simplified to illustrate the basic principles. In practice there are more factors influencing the desired UGS functionality, notably with the advent of liberalization, which has increased the need for short duration measures that can be cycled several times per year (Plaat 2007). Another characteristic of UGS facilities that needs to be taken into account is the decrease in deliverability when more and more working volume is extracted, i.e. as the storage empties. Similarly, injectability decreases when the total gas in the storage gets close to its maximum (the reasons for this are explained below). Such behaviour is tolerated for seasonal storage facilities, since the gas demand tends to reduce towards the end of the winter, and injection requirement reduces towards the end of the summer (Fig. 8). However, for peakshaving purposes, such operational behaviour might not be desirable and needs to be designed out.

UGS performance The performance characteristics of a UGS facility are determined by the components that make up the facility. A schematic picture of a UGS facility is shown in Figure 9. Starting at the pipeline that takes the gas from the main transport system to the UGS, after metering, the gas reaches one or more compressors to increase the pressure of the gas to a level sufficient for injection. From the compressors

34

H. PLAAT

Fig. 8. Typical UGS requirement: seasonal storage.

the gas flows through pipelines to the wells and down into either a salt cavern or a storage reservoir. In pore storage facilities (gas fields, aquifers) the gas then travels away from the wells, further into the reservoir via the interconnected pore spaces in the rock. During this process the pressure in the cavern/reservoir increases proportionally with the total amount of gas injected. During the withdrawal cycle the gas moves in the opposite direction first through the reservoir, then back through the wells and pipelines to a treatment facility for the removal of water and, when present, condensate. Sometimes compression is also used (usually during the end of the withdrawal season) to increase pressure to the

Fig. 9. Components of a UGS facility.

required pressure of the pipeline system. During this entire flow path the gas is subject to friction, which causes pressure losses. The main sources of pressure loss are: (1) The reservoir: a. friction loss is dependent on permeability and thickness of the reservoir; b. this is not applicable for salt caverns, which is the reason that caverns have higher deliverability than pore-space storages and with their limited volume are better suited for short duration peakshavers than reservoirs.

ROLE OF UGS

35

Fig. 10. UGS performance (a) determination of achievable gas rate; (b) effect of decreasing reservoir pressure.

(2) The wells: a. static loss, which is caused by the weight of the gas column and is thus mainly dependant on the depth. During production this is a pressure loss, during injection this results in a pressure gain; b. dynamic loss, which depends on the depth and diameter of the well. (3) Surface pipelines: a. friction loss. (4) Treatment facilities a. friction loss; b. plants based on the LTS process have an additional pressure drop to cool the gas to dewpoint temperature. Friction is thus the main source of pressure loss in a UGS facility and is approximately proportional to the square of the gas flow rate. Although at low flow rates friction loss can be quite small, it

Fig. 11. UGS withdrawal profile.

becomes more important as gas rates increase. Figure 10 illustrates the principles behind determining the deliverability of a UGS facility. The vertical axis in the graph represents the reservoir pressure and the horizontal axis the gas flow rate. The two dashed horizontal lines in Figure 10a represent the boundary conditions: the upper line corresponds to the pressure in the reservoir/cavern, and the lower line to the pressure at which the gas needs to be delivered to the pipeline system. Starting from the reservoir pressures to calculate the pressure at the wellhead, there will be a small static pressure drop due to the weight of the gas column in the well (independent of the gas flow rate). In addition there are the frictional losses (which are dependent on the flow rate), resulting in the downward curving line showing the dependency of the wellhead pressure on the gas rate. Starting at the other end, the pipeline pressure, the gas has to overcome an increasing pressure drop over the treatment facility with increasing rates as shown by the upward curving

36

H. PLAAT

Fig. 12. Examples of withdrawal and injection profiles for Third Party Access.

line in the figure (plant inlet pressure). The intersection of these two lines yields the achievable gas rate for a particular reservoir/cavern pressure. As gas is withdrawn from a UGS facility, the pressure in the reservoir/cavern decreases. In most cases this decrease is more or less linear with the gas remaining in storage. This causes a gradual drop in the wellhead pressure (Fig. 10b). The achievable gas rate at the intersection with the facilities curve will gradually reduce. In addition, other design factors often put a limit on the maximum achievable gas rate, including limitations due to erosional velocities, prevention of hydrates, and maintenance of dewpoint specifications. For these and other reasons, gas treatment facilities have a maximum allowable gas rate that is usually substantially less than the theoretical achievable rate when the storage is full. This limitation is represented by the vertical line in Figure 10b. The reservoir pressure at which the intersection of the wellhead pressure line with the facilities line coincides with this maximum is called the critical reservoir pressure. At higher reservoir pressures the deliverability of the UGS facility is determined by the maximum allowable throughput of the facilities; at lower pressures, the deliverability decreases and is mainly governed by the well performance. In such circumstances compression can be used to increase deliverability. Using compression is effectively the same as lowering the outlet pressure of the plant. But the same principles for determining deliverability apply. The critical reservoir pressure will be lower and there might be different maximum and minimum flow limitations, which are governed by the compressor design. The use of compression during the latter part of the production season is often undertaken as an efficient and relatively cheap way to increase the working volume and to reduce the cushion gas requirements. Since the reservoir pressure is directly related to the amount of working volume extracted, this

behaviour can be translated into a withdrawal profile (Fig. 11). As long as the total volume of gas withdrawn is less than critical working volume, the deliverability of the UGS is at its maximum value, which corresponds to the facilities constraint. When more gas is extracted, the deliverability reduces in line with the achievable well potentials. This characteristic allows the operator to increase the total working volume of the UGS, but at reduced deliverability. During injection a similar approach can be translated into an injection profile. Whether such behaviour is acceptable during the operation of a UGS facility depends on its purpose. In many cases this characteristic can be accommodated. For example demands on a seasonal storage during March and April are lower than those in January and February. Similarly injection requirements in September are less than those in the summer. Many Third Party Access (TPA) storage contracts define facility specific withdrawal and injection profiles, usually using a step function to simplify accounting (Fig. 12).

Conclusions For nearly a century, underground gas storages have proven their value in balancing the fluctuations in gas demand with the limitations of the production and transport sectors. UGS has become indispensable in most current gas markets. With the advent of gas market liberalization the need for and different types of UGS use has only increased. The design of a UGS facility is very much a compromise between what geology and technology allow, how this is reflected in the costs and what functionality and flexibility is required by the market. Although UGS is a flexible tool, once built, there are still significant constraints in the way these can be used, notably when market conditions are changing.

ROLE OF UGS

References B REUNESE , J. N., VAN E IJS , R. M. H. E., DE M EER , S. & K ROON , I. C. 2003. Observation and prediction of the relation between salt creep and land subsidence in solution mining. The Barradeel Case. Solution Mining Research Institute Fall 2003 Conference, 5–8 October, Chester, United Kingdom. C ASTLE , J. W., B RUCE , D. A., B RAME , S. E., B ROOKS , D. A., F ALTA , R. W. & M URDOCH , C. 2004. Design and Feasibility of Creating Gas-Storage Caverns by Using Acid to Dissolve Carbonate Rock Formations. Society of Petroleum Engineers (SPE) Paper 91436, SPE Eastern Regional Meeting, 15–17 September 2004, Charleston WV, USA. C ROTOGINO , F., K O¨ CKRITZ , V. & R EINHOLD , S. 2006. Conceptual Design of Storage Caverns for an LNG Receiving Terminal in Europe. Solution Mining Research Institute Spring 2006 Conference, 30 April– 3 May, Brussels, Belgium. D ERENIEWSKI , E. 2006. Horizontal Well Performance in the Michigan Six Lakes Storage Field. 2006. Paper No. 2.4EF.04, World Gas Conference, 5– 9 June 2006, Amsterdam, Netherlands. EGDON RESOURCES PLC . 2006. Portland Gas Storage Project Update. Press release 28 June 2006. ERGEG. 2005. Guidelines for Good TPA Practice for Gas Storage System Operators (GGPSSO) 23 March 2005. (European Regulators Group for Electricity and Gas, World Wide Web Address: www.ergeg.org). E VANS , D. J. & H OLLOWAY , S. 2009. A review of onshore UK salt deposits and their potential for underground gas storage. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and

37

Europe. Geological Society, London, Special Publications, 313, 39– 80. G ILHAUS , A. 2007. Natural Gas Storage in Salt Caverns – Present Status, Developments and Future Trends in Europe. Solution Mining Research Institute Spring 2007 Conference, 29 April– 2 May 2007, Basel, Switzerland. IGU (I NTERNATIONAL G AS U NION ) 2006. Glossary of relevant technical Underground Gas Storage Terminology. Report of Working Committee 2, Underground Storage of Gas — Study Group 2.1 — Base UGS Activities, Attachment 4, presented during the World Gas Conference, 5 –9 June, Amsterdam, Netherlands. M C C ALL , M. M. 2004. Critical components of salt cavern-based liquefied natural gas receiving terminal undergo field tests. Gas Technology Institute. Gas Tips, Summer 2004, 25– 28. M ANSSON , L., M ARION , P. & J OHANSSON , J. 2006. Demonstration of the LRC Gas Storage Concept in Sweden. 2006. Paper No. 2.2CS.03, World Gas Conference, 5– 9 June, Amsterdam, Netherlands. P LAAT , H. 2007. The Effects of Liberalization on Gas Storage Facilities. The Global Gas Village — UGS workshop, 24–25 April 2007. World Wide Web Address: http://www.energywise.nl. Past Workshops. VON T RYLLER , H., R EITZE , A. & C ROTIGINO , F. 2009. New procedure for tightness tests (MIT) of salt cavern storage wells: continuous high accuracy determination of relevant parameters, without the need to use radioactive tools. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 129–137.

A review of onshore UK salt deposits and their potential for underground gas storage D. J. EVANS* & S. HOLLOWAY British Geological Survey, Keyworth, Nottingham NG12 5GG *Corresponding author (e-mail: [email protected]) Abstract: The UK faces a major change in the nature of its gas supply as North Sea production declines and the country becomes increasingly reliant upon gas imports. As a result the UK Government recognizes that significant investment in gas supply infrastructure is required to maintain security of supply and manage the gas market. Part of that infrastructure will be additional underground gas storage capacity in specially designed and engineered salt caverns. This paper summarizes the distribution and nature of halite (rock salt) deposits in England and Northern Ireland, and reviews the details of existing and planned storage sites in salt caverns. There is considerable potential for further salt cavern development. However, not all of the UK salt fields are suitable, with the halite beds being too shallow, thin or impure.

The UK became a net importer of natural gas during 2004 (DTI 2006a) and will become increasingly reliant upon imports. UK Government predictions are that, by 2020, over 80% of the gas consumed in the UK will be imported (DTI 2006b, c). This could threaten UK market deliverability, price stability and security of supply. Gas imports via new pipelines and liquefied natural gas (LNG) terminals will partially fill the shortfall but there will be a requirement for seasonal and strategic gas storage capacity. Government recognizes the need for greater storage capability in the UK (DTI 2006c, d, 2007), which can be provided by underground gas storage (UGS) facilities. These have and will be developed both in depleted oil and gas fields and solution-mined salt caverns, both onand offshore UK. Renewable energy sources may also be developed in association with compressed air storage in salt caverns, where off peak energy is used to ‘charge’ the caverns; the compressed air being released through turbines during periods of high demand. Furthermore, if in the future the UK moves towards a hydrogenbased energy economy, salt caverns could be used as storage sites for hydrogen, which is already stored in salt caverns on Teesside (BGS 2006; Evans 2008). Salt (sodium chloride, NaCl) is found in nature as solid rock salt (halite) or in solution as brine. Halite is present in massive bedded deposits of both onshore and offshore UK, and large halokinetic structures are present offshore in the southern North Sea (e.g. Cameron et al. 1992). The physical properties of halite make it a very good storage medium with solution-mined caverns constructed

in salt being regarded as prime underground gas storage facilities (BS 1998). Halite is gas tight due to high entry pressures and low permeabilities. The distance between salt (NaCl) lattice units is 2.8  10210 m, whereas the smallest molecular diameter of a methane molecule is 3.8  10210 m (Warren 2006). Halite entry pressures are typically greater than 3000 psi, and impure evaporite beds have entry pressures greater than 1000 psi. These contrast with most water-bearing shales which typically have entry pressures of 900–1000 psi (Warren 2006). Permeabilities of halite typically lie in the range 1026 –1029 millidarcies (mD), with anhydrite c. 1025 mD. These compare to shale permeabilities that are 1021 –1025 and rarely as low as 1028 mD (Beauheim & Roberts 2002; Warren 2006). By comparison, the permeability of well-formed cement in boreholes is of the order of 1025 mD (Nelson 1990; Celia et al. 2006). Halite also deforms viscoplastically, i.e. under normal burial temperatures and (lithostatic) pressures it undergoes deformation by ductile, crystal plastic flow mechanisms. As a consequence any faults, fractures or damaged areas that may have developed in the body of the halite will anneal (‘heal’). These factors coupled with halite deposits generally having large lateral extents, mean that massive halite deposits represent excellent rocks in which to create carefully designed and operated caverns for storage of gas and other hydrocarbons. UGS facilities will need to be located close to the market or users, and therefore will require developments onshore. The UK has a number of salt basins both onshore and offshore. Onshore, however, prospective areas and sites are constrained firstly by the presence of salt beds and thereafter by the

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 39–80. DOI: 10.1144/SP313.5 0305-8719/09/$15.00 # The Geological Society of London 2009.

40

D. J. EVANS & S. HOLLOWAY

minimum depth and salt thickness required for the safe design, construction and commercial operation of caverns and, importantly, their proximity to urban developments. This paper provides an overview of the location and geology of salt deposits onshore in the UK to enable an assessment of their suitability and potential for UGS development. The existing salt cavern gas storage facilities and those projects under construction or in the planning stage are then reviewed, providing background to the development of salt cavern storage facilities in the UK to date.

UK salt beds, natural brines and wet rock head The UK possesses important bedded halite deposits which are commonly associated with mudstone, anhydrite and/or other evaporitic minerals. The most extensive deposits, underlying the NE of England, are of Permian age (Figs 1 & 2). Thick halite beds of Triassic age occur in NW England and the Cheshire, Stafford, Worcester, Somerset and Wessex Basins, with thin halite beds in NE England (Figs 1, 3 & 4). Further salt deposits of Permian and,

Fig. 1. Distribution of the main halite-bearing basins in Britain and the location of operational and proposed UK underground gas storage sites, including depleting oil and gas fields and mined Chalk facilities.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

41

Fig. 2. General location map showing Permian halite-bearing basins and borehole geophysical logs through representative Permian sections. KB, Kelly Bushing level; GR, gamma ray; BHCS, borehole compensated sonic log; SONL, sonic log.

42

D. J. EVANS & S. HOLLOWAY

Fig. 3. The Triassic halite-bearing basins of England, with representative borehole geophysical logs illustrating the presence of non-halite interbeds (mudstone, anhydrite and other evaporitic units). Abbreviations as for Figure 2. GL, ground level; OD, Ordnance Datum; RT, rotary table; SON, sonic log.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

43

Fig. 4. Stratigraphic nomenclature of the Mercia Mudstone Group and Triassic halite beds (following Howard 2006 pers. comm.).

most importantly, Triassic age, are found in Northern Ireland (Fig. 5), as proved in the Larne boreholes (Manning & Wilson 1975; Penn 1981). There are no known salt deposits in Scotland or Wales. Individual salt beds vary from a few centimetres to tens of metres in thickness and salt-dominated sequences may be hundreds of metres thick. Onshore there is little or no evidence for the salt beds having undergone halokinetic movement; however, offshore in the southern North Sea,

pillows, domes and walls have formed from thick sequences of late Permian (Zechstein) salts (e.g. Cameron et al. 1992). Onshore, the halite beds occur only in the subsurface, defining a region of ‘dry rock head’. They do not crop out at or near the surface due to dissolution by circulating groundwaters. The region where the salt would have cropped out had it not been dissolved and is defined by a zone of collapse breccias of mudstones that originally overlaid, or were interbedded with

44

D. J. EVANS & S. HOLLOWAY

Fig. 5. Map showing the Larne salt field in Northern Ireland, and geophysical logs through the Triassic halites proved in the Larne Nos. 1 and 2 boreholes.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

45

Fig. 6. Schematic of wet rock head development (based upon Earp & Taylor 1986; Wilson & Evans 1990).

the halite (Fig. 6) and is known as ‘wet rock head’. Wet rock head conditions are recorded in the Cheshire, Worcester, Needwood, North and South Stafford basins and in the Preesall and Walney Island salt fields in NW England. In late glacial times, when sea level was unusually low, the solution rate in UK salt fields was probably accelerated temporarily as hydraulic gradients were increased and surface streams and most likely the brine seepages were rejuvenated. During glacial periods, penetration of fresh water may be hundreds of metres deeper (e.g. McMillan et al. 2000; Heathcote & Michie 2004). Such fresh water invasion would have potentially significant consequences on the development of wet rock head and has been discussed in relation to the halite beds in the Cheshire Basin (e.g. Howell & Jenkins 1976, 1980). Conversely, during times of high sea level the system would be sluggish, with movement virtually ceasing over wide areas of rock head. The modern-day UK conditions, where there has been no pumping, appear to be closer to the latter condition than the former (e.g. Earp & Taylor 1986). The base of the collapse breccia is usually marked by the presence of natural brines and the effects of wet rock head may lead to problems of metal corrosion, collapse of boreholes and importantly in the context of this paper, leakage at gas storage sites. Problems related to salt dissolution (wet rock head) have been encountered at the Conway Underground East Storage facility in Kansas, USA (Ratigan et al. 2002). Leaks at the Conway facility had been known since the late 1950s, but the presence of propane and hydrocarbons in local groundwater and in a well drilled in 2000 required detailed investigations. These indicated that loss of circulation at the top of the Permian Hutchinson Salt Member was due to wet rock head, which was subsequently found to be

present across large areas of the storage facility site. Up to 10 m of the upper salt bed is now missing, which has caused the development of a collapse breccia from the overlying Upper Wellington Shale. As a result, fissures and voids have formed, which have permitted further entry of water and dissolution of the halite beds and into which stored hydrocarbon products have migrated. In areas of the UK the brines at the top of salt beds have historically been exploited by pumping to the surface (see e.g. Earp & Taylor 1986; Wilson & Evans 1990). However, this process led to major subsidence problems, particularly in the Cheshire Basin (e.g. Taylor et al. 1963; Evans et al. 1968; Earp & Taylor 1986) and brine pumping is no longer permitted on any significant scale. Subsidence features often include linear depressions, which were known as ‘brine-runs’ by the early brine miners (Howell 1984; Cooper 2002). These are channels cut in the salt surface by fresh water that replaced the extracted saturated brine. The channels, which generally form branching networks, widened and deepened away from the pumping stations where the fresher water first came into contact with the salt and where dissolution was most rapid and extensive. The unsupported mudstones overlying the channels fissured, which allowed more water to pass down and so intensified the cutting of channels that ultimately led to collapse and subsidence of the overlying beds. Subsidence was not always close to the points of abstraction, there being a tendency for the most distant parts of the induced brine streams to exhibit the most active subsidence, often many kilometers away (Earp & Taylor 1986). Amongst the deepest recorded onshore occurrences of wet rock head are those in the Stockport-Knutsford district, where the salt subcrops against an undulating surface marking the base of the collapse breccia (not base drift) that varies

46

D. J. EVANS & S. HOLLOWAY

between about 61 m and 122 –152 m below present ground surface (Taylor et al. 1963). In the ChesterWinsford area, wet rock head generally extends down to between 50 and 60 m. However, some brine channels adjacent to larger faults may be between 70 and 80 m below sea level, perhaps 100 m below ground level (Earl & Taylor 1986; Fig. 21). Collapse breccias overlying the top of the salt in the Winsford No. 1 Borehole indicate that wet rock head lies at a depth of 162.5 m (119.2 m below OD) with up to 12 m of salt possibly having been dissolved (Earp & Taylor 1986, p. 51). There are also reports that wet rock head may have been identified down to a depth of 180 m in parts of Cheshire (Howell 1984; Cooper 2002). In the Worcester Basin, a zone of collapse over the wet rock head extends to a depth of around 90 m (Cooper 2002). In NW England, wet rock head is present in both the Preesall and Walney Island salt fields (Wilson & Evans 1990; Rose & Dunham 1977). At Preesall an area of wet rock head was mapped extending up to 0.75 km to the west of the Preesall Fault Zone, and to a depth of 50 –70 m below the base of the drift (Wilson & Evans 1990). Dry rock head is recorded in a number of old wells at 116–119 m below ground level including the access shafts to the Preesall mine, which operated dry until forced to close due to flooding caused by nearby brine wells (Thompson 1908, 1927 (unpublished report); Landless 1979). To the north, the Walney Channel could be the site of a former brine run and associated solution subsidence feature (Jackson et al. 1995). Offshore, beneath the East Irish Sea, wet rock head may develop down to depths exceeding 220 m, e.g. in well 110/9-1 and in the Calder and Morecambe fields (e.g. well 110/ 2a-8; Jackson et al. 1987, 1995).

Triassic halite-bearing strata Most of the halite deposits in the UK onshore area occur within the Triassic Mercia Mudstone Group (MMG) now referred to as the Sidmouth Mudstone Formation (Howard, pers. comm. 2006). The MMG comprises a succession mainly consisting of interbedded red-brown siltstone and mudstone, with gypsum or anhydrite and, in places, halite beds. Strata of the MMG extend eastwards from Northern Ireland, across the Irish Sea into England, continuing eastwards beneath the North Sea, the Netherlands, central Germany and Poland with different lithostratigraphical nomenclature. The sequences represent ancient desert sediments deposited in a semi-arid environment consisting mainly of flat, low-lying plains, which were frequently flooded by seawater. Halite was deposited as a result of the evaporation of shallow brine pools that developed, with salt crystals likely forming both at the

water surface and on the bed of the pools (Arthurton 1973). Salt polygons in the Northwich Halite Member are indicative of emergent conditions between frequent incursions of seawater across the plains in which the brine pools were situated. Mudstone beds are generally representative of aeolian deposits (Rees & Wilson 1998). At the time of deposition there was significant differential subsidence across this vast area and several individual subsiding basins can be identified, with thicker and cleaner halite deposits being found in the basin centres. Well logs through the major halite-bearing formations in the MMG are shown in Figures 3, 7, 8, 9, 12, 15, 16 & 17. The lithostratigraphy of the MMG has been revised recently (Howard 2006 pers. comm.) and is summarized in Figure 4.

Cheshire Basin The MMG of the Cheshire Basin (Figs 3, 4 & 8) comprises predominantly reddish-brown mudstones and siltstones with halite beds, the latter being largely restricted to two sequences: a lower Northwich Halite Member and an upper Wilkesley Halite Member (Rees & Wilson 1998; Howard 2006, pers. comm.). Beds of halite vary from nearly pure salt to a mudstone rock with varying proportions of halite crystals known as ‘haselgebirge’ facies. The Cheshire Basin provides over 85% of the total UK salt output (BGS 2006). The MMG in the Cheshire Basin is almost entirely concealed beneath thick glacial deposits. Consequently, knowledge of the stratigraphy of the group (Figs 4 & 8) and its salt deposits is based mainly upon mining, borehole and seismic reflection data, and the distribution of salt solution subsidence hollows. The ‘outcrop’ of the halite is associated with linear subsidence features caused by natural dissolution. The halite beds can be traced for a distance of about 59 km in a north – south direction and they extend roughly 24 km east –west. The full halite sequence has been penetrated in the IGS Wilkesley borehole where it was first conclusively proved that the ‘Keuper’ saliferous beds were in two distinct formations (Evans 1970; Wilson 1993). Well logs through the saliferous members are shown in Figure 3. The Northwich Halite Member (NHM), formerly the Northwich Halite Formation and before that the Lower Keuper Saliferous Beds (Pugh 1960), occurs between the Bollin Mudstone Member and the Byley Mudstone Member. Mudstones form up to 25% of the NHM (Earp & Taylor 1986), the base of which is placed at the base of the lowest halite bed greater than 2 m thick (Wilson 1993). The member attains a maximum known thickness of 283 m in the Byley Borehole, 15 km north of Crewe, but thins rapidly eastwards such that near

POTENTIAL OF ONSHORE UK SALT DEPOSITS Fig. 7. Location map and borehole geophysical logs through representative boreholes proving the approximate onshore limits of the Wessex Triassic salt basin. Abbreviations as in Figure 2. 47

48

D. J. EVANS & S. HOLLOWAY

Fig. 8. Map of the Cheshire Basin, showing the distribution of the main Triassic halite beds, main locations and boreholes referred to in the text.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

Alsager, it has reduced to 110 m (Evans et al. 1968). The halite-bearing strata occur beneath an area of more than 30 km2. Present-day rock salt production is almost entirely from the Bottom Bed of the NHM at a depth of about 140 m in the Winsford (Meadowbank) Mine at Winsford, near Northwich where it is around 30 m thick. The mine is dry and stable (no surface subsidence results from the operations) and has been the main source of rock salt in the UK since 1928. The rock salt is extracted by room and pillar mining, galleries being 8 m high and 20 m wide, separated by pillars 20 m by 20 m. Serious flooding of the mine occurred during mining operations in 1968, when a borehole was accidentally intersected. As a result, protection barriers of 75 m are now left around boreholes (BGS 2006). The Wilkesley Halite Member (WHM), formerly the Wilkesley Halite Formation and before that the Upper Keuper Saliferous Beds (Pugh 1960), is in places both purer and thicker than the NHM. Its full thickness has been proved only in the Wilkesley Borehole, where it is 404.5 m thick, comprising halites up to 25 m thick. Interspersed are numerous partings of mudstone up to 12 m thick that contain crystals and veins of halite (Wilson 1993). In contrast to the NHM, the WHM contains sandstones that may be more than a metre thick. Halite in the Cheshire Basin is affected by groundwater and ‘wild’ or natural brine has long been pumped commercially. However, ‘brine-runs’ in the top of the salt develop, often very quickly (Earp & Taylor 1986), and unpredictable subsidence can occur, commonly many kilometres from the point of extraction. Natural brine pumping operations are now severely restricted, with only minor quantities produced at Wincham, near Northwich (BGS 2006). However, controlled brine pumping is undertaken in the Northwich Halite Member at the Holford and Warmingham brinefields in Cheshire (Fig. 8). These carefully controlled and monitored operations produce large caverns within the salt, which have been used for storage of hydrocarbons (BGS 2006).

Stafford and Needwood Basins Salt springs and extraction sites have been referred to in parts of Staffordshire since 1686 (Sherlock 1921). Two halite-bearing basins occur in Staffordshire. A north –south outcrop of the Sherwood Sandstone Group immediately east of the Hopton Fault separates the MMG of the Stafford Basin to the west from that of the Needwood Basin to the east (Fig. 9). The halite-bearing strata are referred to as the Stafford Halite Member, which is the equivalent

49

of the Wilkesley Halite Member of the Cheshire Basin (Fig. 4; Warrington et al. 1980). In the Needwood Basin, salt springs and inland salt marshes are indicative of salt-prone sequences coming near to crop at Tixall and Burton on Trent (Stevenson & Mitchell 1955). Salt beds have been proved in the Stafford Corporation, Chartley, Bagot’s Park and Hanbury 1 boreholes (Fig. 9). The Chartley Borehole proved saliferous mudstones at depths between 95 m and 162.4 m (Sherlock, 1921; Stevenson & Mitchell, 1955). Within this succession, two beds of halite 3.2 m and 5.8 m thick were encountered. The Bagot’s Park borehole proved sporadic rock salt in saliferous mudstone between 160.7 m and 227.7 m, with a prominent dolomitic mudstone horizon at 168.5 m (Stevenson & Mitchell 1955). The thickest salt bed is around 2 m between c. 173 m and 175 m depth. The Hanbury Borehole proved the Stafford Halite Member between 127 m and 158 m with four prominent non-saliferous horizons (Figs 3 & 9). The Stafford Halite Member is relatively shallow and between the Stafford Corporation, Chartley and Bagot’s Park boreholes, the halite beds show a general easterly degradation in quality (Fig. 9). In the eastern part of the Stafford Basin, beneath Stafford, several halite beds are present at depths greater than 60 m in a sequence 50 –65 m thick. They are preserved in a synclinal structure, bounded on its eastern side by the Hopton Fault and have been proved in various water borings made by the Stafford Corporation and a series of brine wells completed in the 1880s. Records from the corporation boreholes (Sherlock 1921; Whitehead et al. 1927) show two beds of halite: an upper bed 12.8 m thick and a lower bed 2 m thick, separated by 8.5 m of saliferous mudstones with beds up to 15 cm thick. Elsewhere, halite beds forming the Stafford Halite Member have an aggregate thickness of 21 m over a 46 m interval and probably represent marginal basin facies (Stevenson & Mitchell 1955; Notholt & Highley 1973). A single, thin bed of halite was reported at a depth of 63.4 m in the Hurst Farm Borehole, indicating that the southern limit of the main Stafford Halite Member lies to the south of this (Bridge & Hough 2002). Highly saliferous water flow was also encountered in the Ashflats borehole. The Stafford Halite Member is thus relatively shallow, contains few thick beds of salt and shows rapid lateral variations in the quantity and quality of the salt.

Worcester Basin Springs, boreholes and workings at Droitwich, and Stoke Prior (Fig. 10) prove that halite beds, though not recognized at outcrop, are present within the MMG of the Worcester Basin (Sherlock 1921;

50

D. J. EVANS & S. HOLLOWAY

Fig. 9. Geological map of the Stafford and Needwood Basins, showing main locations and boreholes referred to in the text.

Wills 1970, 1976; Mitchell et al. 1961; Poole & Williams 1980; Old et al. 1991). Indeed, salt has been extracted from brine in Droitwich from as early as 816 AD, with salt pits containing salt springs operating in the seventeenth century (Sherlock 1921). In 1828, salt was discovered and mined at Stoke Prior, though the extraction method later changed to brine pumping. Brine continued to be extracted until the 1970s (Notholt & Highley 1973).

The halite-bearing beds are up to 90 m thick and contain around 40% siltstone and mudstone units (Cooper 2002). They form the Droitwich Halite Member, thought to be equivalent to the Wilkesley Halite of the Cheshire Basin (Fig. 4; Barclay et al. 1997). As in the Stafford Basin to the north, the halite deposits probably represent a marginal basin facies, being contaminated with mudstone (Notholt & Highley 1973).

POTENTIAL OF ONSHORE UK SALT DEPOSITS

51

Fig. 10. Geological map of the Worcester Basin, showing main locations and boreholes referred to in the text.

The exact limits and extent of the Droitwich Halite Member are uncertain. Boreholes in the Droitwich area prove an aggregate thickness of 21 m of halite in six main beds at depths between 91 m and 125.5 m. More could be present at greater depths. Several salt beds at Stoke Prior, varying in thickness from 2 m to 12 m and totalling 26 m or more, occur at depths greater than 90 m. The halite may extend eastwards as far as the Lickey End and Stoke Pound faults (Old et al. 1991). However, at Redditch (11.25 km to the east) and at Stratford-on-Avon (32 km to the ESE), no halite has been found in the MMG (Notholt & Highley 1973). In 1945 an ICI borehole at Saleway, c. 5 km to the SE of Droitwich, proved an aggregate halite thickness exceeding 30 m (Wills 1976) and though halite remains unproven to the south, it is thought likely that it extends southwards at depth into the Worcester Basin (Barclay et al. 1997). To the east of the Smite-Pirton-Tewkesbury

fault system, the MMG is up to 1200 m thick and subsidence features are found in the north of the district. Seismic reflection data also reveal strong reflectors that may be halite beds. The area of halite deposition was probably controlled by the Smite and Inkberrow faults, which may be borne out by the absence of salt in the Kempsey borehole to the west of the Smite-Pirton-Tewkesbury fault system (Barclay et al. 1997). In the Worcester Basin, wet rock head conditions exist in a NE-trending belt 1–2 km wide and 12 km long, parallel to the ‘outcrop’ through the Droitwich to Stoke Prior area (Cooper 2002).

Somerset Saliferous deposits known as the Somerset Halite Member occur within the small east–west trending Central Somerset Basin (Fig. 11). The basin contains a thick sequence of Lower Jurassic rocks

52

D. J. EVANS & S. HOLLOWAY

and Mercia Mudstone Group overlying thin Sherwood Sandstone Group, which are flanked by Palaeozoic rocks to the north and south. To the west the basin continues offshore beneath the Bristol Channel. A coal exploration borehole drilled at Puriton in 1910 proved some 427 m of MMG and encountered halite beds between 183 m and 219 m (McMurtrie 1911; Usher 1911; Whittaker 1971). Following tests, brine was extracted for a period of 11 years

until the works closed down in 1922, eight years after being taken over by the Salt Union Company. It is thought that Puriton lies near the southern edge of the Somerset salt field and subsequent deep boreholes have provided more detail on the occurrence and distribution of halite across the basin. The IGS Burton Row borehole, drilled in 1971 on Brent Knoll, about 11 km north of Puriton, proved 484 m of MMG, with the main halite beds occurring between 693.8 m and 742.4 m (Fig. 3; Whittaker

Fig. 11. Geological map of the Somerset Basin, showing main locations and boreholes referred to in the text.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

1972, 1973; Whittaker & Green 1983). Halite veins and stringers were also encountered between 643 m and 797 m (Notholt & Highley 1973). The halite occurred at a similar stratigraphical level to that in the Puriton borehole.

Wessex Basin (Dorset) The MMG occurs in the Wessex-Channel Basin in southern Britain, much of it being concealed beneath a cover of Jurassic and younger rocks. Prior to oil exploration in the 1970s, halite was not known in this area. However, since then, a series of boreholes have proved the existence of a main saliferous unit underlying large areas of Dorset and the area immediately offshore. This was commonly referred to as the Dorset Halite Formation (e.g. Stewart et al. 1996; Harvey & Stewart 1998), and is now known as the Dorset Halite Member. The approximate limit of the onshore salt basin is shown in Figure 7. The absence of halite in the Norton, Wilmingham and Arreton 2 wells on the Isle of Wight indicates that the eastern limit lies offshore in Bournemouth Bay. The southern and southwestern limits are poorly constrained, but lie offshore in the Channel Basin, perhaps defined by a line of graben affecting Jurassic and Cretaceous rocks in an ENE line from south of Swanage to south of Lyme Bay (Harvey & Stewart 1998). The Dorset Halite Member in general thickens southwards towards the Weymouth Anticline. Until recently, the thickest drilled succession was c. 350 m, proved by the Chickerell borehole (Fig. 12). A younger halite 20–65 m thick occurs, separated from the main halite by 60– 80 m of the Branscombe Mudstone Formation. However, some of the thickening here could be related to salt movement in the core of the anticline with the halite beds acting as a de´collement for faulting in the overlying succession (Fig. 13). Depths to the uppermost halite bed generally increase to the south, from c. 422 m and 1036 m in the vicinity of Marshwood and Ryme Intrinseca respectively, to 2141 m at Southard Quarry (Fig. 7) and continue into the offshore area. In between there are a number of sub-basins in which the top of the halite is proved at depths of around 1699 m (Spetisbury), 1936 m (Winterborne Kingston) and 1428 m (Martinstown). The Portland 1 borehole on the Isle of Portland was drilled in 2006 to the south of the axis of the Weymouth Anticline to confirm the presence and thickness of the halite sequences for possible gas storage purposes (see below). It proved a saliferous sequence 470 m thick (c. 41 m thicker than forecast), with the main halite bed encountered at a depth of 2365 m, being 135 m thick and approximately 43 m thinner than forecast (Egdon 2006a, b). Generally, the thickness of individual units of relatively pure halite range from 20–41 m, but can

53

exceed 60 m in thickness, as proved in the Martinstown, Chickerell and Southard Quarry boreholes (Figs 7 & 12).

Preesall (Lancashire) and Walney Island (Cumbria) Halite occurs in the MMG in several locations around the margins of the East Irish Sea Basin and Solway Basin (Fig. 14). The Preesall salt field near Fleetwood, Lancashire, was proved during drilling for iron ore (haematite) in 1872 (Thompson 1908; Landless 1979; Wilson & Evans 1990). A shaft was subsequently sunk near one of these boreholes in 1885, and in 1889 the Preesall Salt Company commenced pumping brine from the shaft. In 1890, the company became part of the United Alkali Company Ltd who sank two further shafts and mined rock salt. In 1926, the company became one of the founding component companies of Imperial Chemical Industries Limited (ICI). The mine worked salt from two levels, though the area of mining was not large (c. 350  500 m; Wilson & Evans 1990) and, as a result of flooding, was eventually abandoned in the early 1930s (Landless 1979). However, brine extraction for use as chemical feedstock in the production of chlorine and caustic soda at ICI’s Hillhouse plant continued at the Preesall salt field until 1992, when changing conditions in the electricity market and other factors meant that the Hill House plant was closed down and the brinefield mothballed. The field, though small, has played a significant role in the development of the technique of controlled brine pumping in the UK (Wilson & Evans 1990). The dominantly mudstone MMG succession in the Fleetwood area attains a thickness of over 800 m and contains a number of halite horizons (Fig. 3). However, the distribution of halite beds is not well constrained everywhere, due to a thick drift cover across the area, and is known mainly from borehole provings (Wilson & Evans 1990; Evans et al. 2005). The lowermost mudstones (Hambleton Mudstones) where proved in boreholes, contain no halite beds. The overlying Singleton Mudstones comprise dominantly reddish-brown mudstones with impersistent beds of halite, mainly to the west of Preesall in the central and western Fylde area. Two halites have been proved in the ICI B8 borehole to the west of the Wyre Estuary; the Rossall Halite near the base of the formation and the Mythop Halite near the top (Wilson & Evans 1990). The Rossall Halite is about 11.5 m thick and occurs 20 m above the base of the Singleton Mudstones, whereas the Mythop Halite occurs within a 59 m section of mudstones with halite veins up to 10 cm thick. Haslegebirge facies are found within this sequence and beds of pure halite up to

54 D. J. EVANS & S. HOLLOWAY

Fig. 12. North–south borehole geophysical log correlation across the western areas of the Wessex Basin, showing the development of Triassic halite beds. Abbreviations as in Figure 2.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

55

Fig. 13. Seismic reflection line illustrating the thickening of the (Dorset) halite of the Wessex Basin in the Weymouth Anticline. Also shown, sketch of the level of proposed gas storage caverns within the halite beds (after Egdon 2006a).

56

D. J. EVANS & S. HOLLOWAY

Fig. 14. Mercia Mudstone and halite occurrences around the East Irish Sea Basin and main locations and structures referred to in the text.

2 m thick are encountered at the top of the succession. Breccias above and below the Mythop Halite are the likely correlatives of thicker salt beds present offshore in the East Irish Sea (Wilson & Evans 1990). The succeeding Kirkham Mudstones Formation comprises three members: a lower (Thornton) and an upper (Coat Walls) mudstone member with a

thick middle member known as the Preesall Halite Member (Figs 3, 4 & 15). Onshore, a few thin halite beds are encountered towards the top of the Thornton Mudstones. The Preesall Halite is the equivalent of the Northwich Halite in the Cheshire Basin (Wilson 1990, 1993) and extends north to Walney Island (see below) and offshore into the East Irish Sea (Jackson et al. 1987). The Preesall

POTENTIAL OF ONSHORE UK SALT DEPOSITS 57

Fig. 15. General north–south borehole geophysical log (gamma ray) correlation of the Preesall Halite along the western boundary of the existing Preesall brinefield. Top Preesall Halite in brinewells (BW) taken from ICI borehole log descriptions.

58

D. J. EVANS & S. HOLLOWAY

Halite is preserved in a broad synformal structure now thought to represent a down-faulted graben (Evans et al. 2005). This is bounded to the east by the down-west Preesall Fault Zone and to the west by the NE-trending, down-east Burn Naze Fault (Fig. 15), lying beneath the Wyre Estuary and antithetic to the Preesall Fault. The Preesall Halite is not now thought to reach subcrop beneath the Wyre Estuary on the western limb of the Preesall Syncline as depicted by Wilson & Evans (1990). Nearly 100 boreholes have penetrated and been used to extract brine from the Preesall Halite, which over the worked brinefield is generally between 100 and 130 m thick. Thickening of the salt beds eastwards into the Preesall Fault indicates syndepositional movements. To the east, the top of the Preesall Halite progressively rises and once within 50–75 m of the base of the drift, is subject to dissolution. In these areas it is overlain by collapse breccias of Coat Walls Mudstone, defining an area of wet rock head. Where the drift is thicker, the wet rock head may reach depths of up to 125 m (Wilson & Evans 1990). Along the western edge of the worked brinefield, the top of the Preesall Halite is at depths between 225 m and 325 m, and varies in thickness between c. 140 m in BW-123 and an unbotttomed 289 m in BW-130 (Evans et al. 2005). Thus the base of the halite may reach depths of greater than 500 m (Wilson & Evans 1990; Evans et al. 2005). In 2004, Canatxx Gas Storage Company Limited drilled two geophysically logged salt exploration wells at The Heads in the south of the unworked salt field and Arm Hill in the north. The latter was fully cored. The boreholes proved the top of the Preesall Halite to lie at 226 m and 366 m below ground level respectively (Fig. 15). The thickness of the halite beds was 210 m in the vicinity of the Heads and c. 240 m at Arm Hill. In the Arm Hill borehole the Preesall Halite consists almost entirely of rock salt with generally only thin mudstone and anhydrite interbeds present. Three more prominent zones of salt and non-salt interbeds 5 m, 6.19 m and 6.54 m thick were also encountered (Fig. 15), with the maximum individual non-halite bed thicknesses in these zones between 1.04 m and 1.76 m (Ratigan 2005). Gamma ray logs from the older ICI brine wells and the Canatxx wells reveal a consistent halite stratigraphy that can be traced along the western boundary of the brinefield (Fig. 15). These indicate that the Preesall Halite contains fewer thick mudstone interbeds than the other Triassic salt fields in the UK (Fig. 3). Westwards of the Canatxx wells, approaching the Burn Naze Fault, the Preesall Halite rises to shallower levels but is not thought to crop out beneath the estuary (Evans et al. 2005). North of the Fleetwood-Preesall area, across Morecambe Bay, the MMG is present on, and restricted to, an area on Walney Island and a

narrow coastal strip near Barrow-in-Furness (Figs 14 & 16). This is covered entirely by thick drift, however, saliferous sequences have been proved by borings at a number of levels within the MMG succession (Sherlock 1921; Rose & Dunham 1977; Johnson et al. 2001). The MMG is preserved within a narrow NNW-trending syncline in the hangingwall of the Haverigg Fault (Fig. 16; Rose & Dunham 1977). This fault partly replaces the eastern limb of the syncline and probably represents the southward continuation of the Lake District Boundary Fault (Jackson & Mulholland 1993). It has been correlated with the Preesall Fault (Sherlock 1921). However, the Gleaston or Yarlside faults are the more likely correlatives, with the faults probably representing a set of en echelon structures within the Lake District Boundary Fault Zone. The syncline, the axis of which underlies the Walney Channel, shows similarities to the Preesall Syncline to the south. It preserves a 550–600 m thick sequence of MMG, comprising a basal 15 – 20 m of grey and sandy (Hambleton) mudstones passing successively upwards into a 300–400 m thick red (Singleton) and red/green (Kirkham) mudstones sequence (Rose & Dunham 1977). Thin salt beds and ‘rotten marls’ are present within the Singleton Mudstones, proved in the Walney Island No. 2, British Gypsum No. 1 and Biggar No. 1 boreholes, the latter drilled by Ultramar Exploration Ltd in 1990, in the south central part of the island (Fig. 16). They are thought to equate with the thin Rossall and Mythop salts in the Fleetwood area to the south of Morecambe Bay (Rose & Dunham 1977; Jackson et al. 1995). The halites are readily identified on the geophysical logs of the Biggar No. 1 borehole, which show the uppermost (Mythop) halite unit between c. 205 m and 238 m and the lower (Rossall Halite) between 368 m and 415 m (Figs 3 & 16). The geophysical log character indicates a number of non-halite interbeds within each halite, presumed to be predominantly mudstones. The lower half of the intervening mudstone is variably saliferous. Beneath the lower (Rossall) halite, the borehole proved a further 97 m of mudstones, before entering, at around 512 m, and terminating (at 645 m) within sandy sequences of the Ormskirk Formation (SSG). The highest strata present on the island consist of a 100 m thick saliferous sequence, proved near Biggar and which is equated with the Preesall Halite to the south of Morecambe Bay (Rose & Dunham 1977). The uppermost halite beds are proved in the Walney Island Nos. 2, 3, 4 and 5 wells, drilled in the late 1880s, and which are overlain by around 50 m of red mudstones representing, in part, collapse breccias of the higher Kirkham (Coat Walls?) mudstones (Rose & Dunham 1977). Indeed, the Walney Channel could be the site of a

POTENTIAL OF ONSHORE UK SALT DEPOSITS

Fig. 16. Geological map of the Walney Island area, showing the boreholes and locations referred to in the text.

59

60

D. J. EVANS & S. HOLLOWAY

former brine run and associated solution subsidence feature (Jackson et al. 1995). The Walney Island No. 5 borehole terminated in a 10 m thick rock salt and mudstone sequence, indicating further halite beds could be present at depth (Sherlock 1921). Three main halite beds with an aggregate thickness of just over 72 m were recognized, the thickest of which is 46 m (Sherlock 1921). The borehole succession, reclassified by Rose & Dunham (1977), reveals the top of the Preesall Halite at 110.64 m below ground (surface being þ10 m OD) and the base at c. 227 m. The halite beds contain a sequence of mudstones and mixed (haselgebirge) facies around 24 m thick between 164.13 m and 187.77 m depth. The British Gypsum No. 1 well (Fig. 16), drilled in 1971, proved a thick mudstone breccia, which was interpreted by Rose & Dunham (1977) as the solution breccia of a major halite formation indicating the former presence of the Preesall Halite in that area. The area of the salt field is only around 5 km2 onshore, and it is overlain by wet rock head throughout (Rose & Dunham 1977). The main halite beds supported a small and long since defunct brining industry near Biggar. Brine was pumped from the wet rock head and piped to a salt plant on the south of the Island (Rose & Dunham 1977). The wet rock head conditions, shallow depths and thin salt beds make controlled solution-mining impracticable, with a risk of surface subsidence. The same conditions, plus limited volumes of halite, also mean that conventional mining of the lower halite beds is impossible (Rose & Dunham 1977).

Southport (Lancashire) At least two boreholes to the NE of Southport in Lancashire have proved halite beds beneath driftcovered areas (Fig. 14). The composite log and sample descriptions from the Banks 1 exploration well, drilled by Clyde Expro plc in 1992, describe the top of halite at about 86 m, with the base of the halite placed at 209.7 m (Fig. 17). The overlying strata are described as the Coat Walls Mudstone Member overlain by 16.8 m of drift. The halite unit is 93.6 m thick and assigned as the Preesall Halite. Casing extends to around 358 m BOD and geophysical logs over the halite interval may Fig. 17. Geophysical logs of the Banks 1 borehole NE of Southport (Lancashire), illustrating the company composite interpretation of the Preesall Halite (KB, 10.67 m above OD). The blocky nature of the sonic log may indicate that some halite-dominated beds extend to at, or near, the base of the superficial deposits. Older halites equivalent to the Mythop and/or Rossall halites may also be present at depths between 300 m and 370 m.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

reflect poor borehole wall conditions. However, the gamma ray log and blocky nature of the sonic log would appear to indicate a thick halite interval with some interbedded mudstone horizons. Halite beds may in fact extend to at or near subcrop at the base of the superficial deposits. In addition, similar blocky sonic log character indicates that older halites may be present at depths between 264 and 334 m BOD: the likely equivalents to the Mythop and/or Rossall halites (Fig. 17). The second proving of halite is the Southport Deep Shaft borehole, drilled in 1991 (Fig. 14). This borehole proved the top of a halite unit at 62.2 m BOD and reached TD within the same halite at 96.1 m BOD. The borehole proved greater than 33.9 m of what is likely to be the Preesall Halite. The well logs describing the top of the halite unit between 62 m and 75 m, make no mention of wet rock head conditions. The extent of the halite beds hereabouts are not known in detail but there are early records of ‘brine stronger than seawater’ from an area NE of Southport (Sherlock 1921). Dickinson (1882) postulated that the subsidence of the seashore at Southport might be related to the solution of rock salt thereabouts, a theory thought unlikely by Sherlock (1921).

Carlisle Basin The MMG crops out in the Carlisle Basin in NW Cumbria, and a bed of halite has been proved within it in the Silloth 1A borehole (Figs 3 & 14). Onshore the MMG is represented by a succession of red-brown shales (the Stanwix Shales), which attain a thickness of approximately 500 m. The 6 m thick halite in the Silloth 1A borehole lies approximately 200 m above the base of the Stanwix Shales. It has been correlated with a thin halite within the gypsumrich mudstones in the Kelsick Moss borehole and it is thought likely to be the lateral equivalent of the Preesall Halite (Jackson et al. 1995). The Carlisle Basin extends offshore into the Solway Basin, the SW margin of which crops out on the northern end of the Isle of Man (Fig. 14). Within the Solway Basin the MMG is over 800 m thick and contains significant halite beds, often with mudstone intercalations (Jackson et al. 1987, 1995; Jackson & Mulholland 1993; Jackson & Johnson 1996; Chadwick et al. 2001).

Isle of Man A number of boreholes drilled between 1891 and 1906 in the exploration for coal at the Point of Ayre, Isle of Man (Fig. 14), revealed the presence of Permian and Triassic strata concealed beneath a thick drift cover and lying below sea level (Lamplugh in Sherlock 1921; Notholt & Highley

61

1973). These strata were deposited near the SW margin of the Solway –Carlisle Basin (Jackson & Mulholland 1993; Jackson et al. 1995). Triassic strata proved included MMG up to 246 m thick, at depths between 183 and 274 m (Lamplugh in Sherlock 1921; Notholt & Highley 1973). The MMG contains up to 21 recognized halite beds totalling 23 m in thickness, the five thickest of which add up to 17 m. In addition to the main halite beds, a further 39 m of strata comprise intercalations of salt and mudstone in varying amounts. The halite beds have a restricted occurrence in the north of the island, covering an area of about 7.75 to 10.4 km2. They are known to extend offshore into the Solway Basin and are thought to be the equivalent of the Rossall Halite of the Walney Island and Preesall areas, onshore NW England (Jackson et al. 1995; Chadwick et al. 2001). Following the discovery of rock salt, the Manx Salt and Alkali Company Ltd began producing salt on a commercial scale in 1903 and continued to do so until the late 1950s (Notholt & Highley 1973).

NE England (Cleveland Basin) Mercia Mudstone Group strata crop out in a belt stretching across eastern and NE England between Leicester and Middlesbrough. Boreholes in the Whitby area including Lockton East 1 (Fig. 3) have proved a 30 –40 m thick halite member (the Eskdale Evaporite Member) lying near the base of the MMG (Notholt & Highley 1973). This is the lateral equivalent of the Ro´´t Halite Member of the southern North Sea (Warrington et al. 1980). The onshore halite beds represent the marginal equivalents of much thicker halite beds deposited offshore in the Southern North Sea Basin (see Cameron et al. 1992).

Northern Ireland Halite forms beds of variable thickness within the MMG between Carrickfergus and Larne in the SE corner of County Antrim, to the NE of Belfast (Fig. 5). Their presence underlying the area has been known since the mid nineteenth century and they have been worked for over 100 years (Griffith & Wilson 1982). Halite was first encountered in a borehole at Glynn in 1839 during the search for coal (e.g. Doyle 1853; Notholt & Highley 1973; Griffith & Wilson 1982) and working of salt took place mainly in three areas: Woodburn or Duncrue, Eden and Red Hall, although other shafts or mines are recorded (e.g. French Park Mine, Burleigh Hill Mine and Maidenmount). Production of salt was variable and that from Red Hall only ever by solution mining from boreholes. Most of the mines had ceased operations by the end of the 1920s, although

62

D. J. EVANS & S. HOLLOWAY

some continued production into the 1950s. The only remaining operational mine is at Kilroot, 4 km ENE along the coast from Carrickfergus. The mine was opened by the Irish Salt Mining and Exploration Company Limited in 1967, and is accessed by an inclined adit. It produces rock salt by the roomand-pillar method. The MMG and associated halite beds occur on the southeastern margin of a Mesozoic basin centred on the Glenarm– Larne area. The Carrickfergus area lay closer to its margin, resulting in first thinner and then absent salt beds to the SE. Halite beds occur at three levels (Fig. 5), forming the Larne Halite, Carnduff Halite and Ballyboley Halite members (Manning & Wilson 1975; Penn 1981). Total thicknesses and lateral extent of the halite beds are not known with certainty. The only two boreholes to penetrate the entire MMG succession in the Larne Basin are at Larne and although only 750 m apart, they proved markedly differing sequences (Manning & Wilson 1975; Penn 1981; Griffith & Wilson 1982). Larne No. 1 proved 967 m of MMG that includes c. 538 m of bedded halite in (Manning & Wilson 1975) the Larne Halite Member (481 m thick, from 363 m), the Carnduff Halite Member (31 m thick, from 908 m) and the Ballyboley Halite Member (26.34 m thick, from 1001 m). Larne 2 (Fig. 5) ESE of Larne 1, proved similar halite beds, but of strikingly different thicknesses (Penn 1981): the Larne Halite Member being 178 m thick, the Carnduff Halite 180 m thick and the Ballyboley Halite 41 m thick. The Larne Halite contains three major halite beds up to 30 m thick with thinner (1.5–4.6 m) interbeds of halite and mudstone/siltstone (Penn 1981). The Carnduff Halite Member comprises five halite beds of more than 4.6 m thickness, the greatest of which is 70 m thick. The Ballyboley Halite Member contains four principal halite beds of less than 7.6 m thickness, with the lowermost bed being a mixed halite/mudstone basal part passing down into anhydrite/mudstone at the base of the member (Penn 1981). A c. 10 m thick dolerite sill occurs at the top of the Larne Halite in the Larne 2 borehole. The intrusive rock gives rise to a geophysical log response similar to that of the halite, hence the top of the halite is not taken at the obvious log break (Fig. 5). The thick-bedded halites of the Larne area diminish southwards into the worked halite fields, where they vary in thickness from 9 m to 27 m (Mitchell 2004). Lateral correlation is at present impossible due to the lithological changes occurring towards the basin margin (Griffith & Wilson 1982). At Carrickfergus, bedded salts are only 40 m thick, whilst to the west of Duncrue the halite beds die out and they are absent in the Belfast area.

Permian halite-bearing strata Halite-bearing strata of Permian age lie concealed at depth beneath much of eastern England, from Teesside southwards through Yorkshire into northern Lincolnshire (Figs 1 & 18). The halite beds form part of the Zechstein Group, which was deposited in a vast basin that extended eastwards from the United Kingdom, across the southern North Sea, into Germany and Poland. NE England lay not far from the western edge of this basin. Within the Zechstein Group, halite formations are interbedded with thick dolomite, mudstone and anhydrite formations in five cycles (Z1 –Z5). Each cycle represents a flooding of the southern Permian Basin from the north, followed by evaporation and drying out of the basin (Cameron et al. 1992). The halite beds thin rapidly westwards, but thicken and deepen to the east and south, east of the eastern limit of the predominantly carbonate sequence that lies along a roughly north –south line through central Yorkshire.

NE England These halite deposits have been exploited in two areas, referred to here as the Yorkshire and Teesside provinces. In the Yorkshire province, the Z2 Fordon Evaporites generally contain a thick halite succession (see well logs on Fig. 2) and at appropriate depths, provide suitable conditions for constructing caverns for gas storage. Cavities have been created in the Z2 salts at Atwick, near Hornsea in east Yorkshire at between 1710 and 1840 m depth and have been utilised for natural gas storage since the 1970s (Dean 1978, 1985). Around Hornsea and Atwick, the Fordon Evaporites are about 280 m thick, with the top at around 1660 m below ground level (bgl) and the base at c. 1940 m. South of Hornsea, the Great Hatfield No. 1 exploration well proved that the top of the Fordon Evaporites lie at depths of 1626 m west of the proposed Aldbrough storage site. To the north, around Robin Hood’s Bay, the Z2 salt is about 1220 m deep, deepening to about 1450 m at Scarborough. The thickness varies from about 75 to 90 m. At Fordon, the Z2 salt is deeper (1850 m) but also thicker (120 m). Further north, the search for water near Middlesbrough in 1859, led to the discovery, beneath about 220 m of Sherwood Sandstone, of the Teesside province of the Permian salt field. The Teesside salt field is formed by the (Middle or Main) Boulby Halite Formation (Z3) overlying the Billingham Main Anhydrite. Brine was produced for both salt and chemical processes around Greatham, NE of Billingham. Small caverns were developed at Saltholme north of the river Tees and Wilton to the south (Cooper 2002). Hereabouts, the halite varies

POTENTIAL OF ONSHORE UK SALT DEPOSITS

63

Fig. 18. Details of the Permian stratigraphy in NE England (based upon Gaunt et al. 1992; Berridge & Pattison 1994; Smith 1994; Frost 1998).

in thickness between 30 and 45 m, and lies at depths between 274 m and 366 m. At Saltholme the halite is approximately 340 m below the surface and varies in thickness between 27 m and 40 m. Over a hundred solution-mined caverns have been formed, varying in final leached volumes from 10 000 m3 to 100 000 m3 capacity. At Wilton, solution-mined caverns were developed in salt beneath a cover of MMG and SSG strata approximately 650 m below the surface in salt of approximately the same thickness as at Saltholme. Extraction of brine for use as chemical feedstock continued until at least 2002. Existing cavities were used solely for storage purposes, storing ethylene, butane and nitrogen. ICI also converted smaller caverns in the Saltholme and Wilton brine fields into storage sites for light hydrocarbons, propane, propylene, crude oil, gas oil, naphtha, ethylene, nitrogen and hydrogen. Permian saliferous sequences do not come to crop in NE England because in places their facies is restricted to deeper parts of the depositional basin (Smith 1989), or elsewhere due to a dissolution front and wet rock head conditions (Cooper 2002). The dissolution front extends much deeper (down

to 300–400 m) than in the Cheshire Basin, Devensian glaciation having caused elevated groundwater head conditions that flushed out the saline waters (Howell & Jenkins 1976; Cooper 2002).

Northern Ireland The Larne No. 2 Borehole proved at least 1264 m of Permian rocks in the Larne Basin (Penn 1981; Mitchell 2004; Evans et al. 2006). These strata include a 113 m thick bed of halite within the White Brae Mudstone Formation (‘Permian Upper Marls’) towards the top of the Permian sequence and lying directly above the Magnesian Limestone Formation (Fig. 19). The top of the halite at 1688 m is marked by a bed of anhydrite approximately 19 m thick and is overlain directly by c. 53 m of reddish brown mudstones and thin siltstones that are typical of the White Brae Mudstone Formation. The areal extent and compositional variation of the Permian halite is poorly constrained, with no other boreholes having proved the halite. Halite has only been proved in the Larne Basin in the area of south and east Antrim that lies south of

64

D. J. EVANS & S. HOLLOWAY

Fig. 19. Geophysical logs of the Permian halite in Larne No. 2, illustrating the presence of a 113 m thick halite bed towards the top of the Permian immediately overlying the Magnesian Limestone Formation and overlain by a 19 m thick anhydrite and siltstone division. The uniform sonic log response indicates a very pure halite bed with no apparent non salt (anhydrite or mudstone) interbeds.

the Sixmilewater Fault in its hanging wall block (Fig. 5). Gravity data indicate that the Permian salt is likely to be present in the deepest part of the basin, marked by the lowest gravity anomaly values, which is centred over Larne and extends c. 2.5 km to the south and west of the town and under the north of Larne Lough (Carruthers et al. 1999; Evans et al. 2006). To the north of the Sixmilewater Fault, the Ballytober Borehole proved over 400 m of Permian strata with 100 m of the White Brae Mudstone Formation representing the top of the Permian sequence. The sequence at Ballytober is similar to the succession in the Larne No. 2 Borehole, except for the absence of the bed of halite at the base of the White Brae Mudstone Formation. Towards the southern edge of the Larne Basin in the Newmill Borehole, an attenuated Permian succession the top of which is represented by 22 m of anhydrite and shale, is presumed to represent the Magnesian Limestone. The ‘Permian Upper Marls’ are present elsewhere in the Newmill area and the Permo-Triassic units, particularly the halite beds,

show thinning in this direction away from the Larne area. (Evans et al. 2006). Where proved, the geophysical log response of the Permian halite is very uniform (Fig. 19) and with the mean %NaCl calculated as 91.7%, is similar to that of pure NaCl. This indicates that any impurities are dispersed throughout the halite rather than as mudstone or siltstone interbeds which may be important for gas storage (Evans et al. 2006).

Fractures and infilling halite associated with non-halite interbeds Halite-bearing formations in the UK contain non-halite interbeds infilled vein systems representing old fractures which are almost exclusively filled with evaporite minerals (e.g. Wilson & Evans 1990; Smith 1996). Such infilled fractures also occur in the enclosing mudstone formations. The presence of fractures is an important issue in the assessment of gas tightness of the enclosing

POTENTIAL OF ONSHORE UK SALT DEPOSITS

and caprock sequences, and thus the safety of an underground storage facility. Halite/non-halite layer interfaces are the weakest points in the halite deposits and fractures may develop and propagate from them, generally parallel to the interface (KDHE 1997). It is pertinent, therefore, to assess briefly the possible presence and origin of any fractures and their infill material in evaporite sequences and the likelihood that any fractures will remain open. Cracks and fractures in non-halite interbeds and mudstone caprock may develop in a number of ways. They may be sedimentary and early burial features such as desiccation cracks in silty mudstone horizons, most likely representing material introduced to the salt basin during floods and which subsequently dried out. Fractures may also develop post burial as a result of diagenesis or faulting. Of significance are the effects of late mesodiagenetic and telodiagenetic processes, often associated with uplift of basins, whereby circulating groundwaters may lead to the rehydration of anhydrite cements and their replacement by gypsum in evaporite sequences (e.g. Strong et al. 1994; Milodowski et al. 1999). The hydration process leads to a major volume increase (over 40%; Cooper 1988; Warren 2006) that leads to a progressive crack-seal mechanism of hydration with overpressuring and hydrofracturing. It causes expansive disruption of the rocks that is associated with the development of cross-fibre or ‘satin spar’ gypsum veining, which is known in the Cheshire Basin (Milodowski et al. 1999). This forceful expansion can create extreme pressures and has been responsible for rock explosions and local uplifts in America (Brune 1968; Cooper 1988). Anhydrite cementation is widespread and perhaps even characteristic of Permo-Triassic successions in the Cheshire Basin, west Cumbria, Yorkshire– Lincolnshire, the Preston area and Wessex Basin (Milodowski et al. 1986, 1987, 1999; Strong & Milodowski 1987; Strong 1993). In the Ripon area, circulating groundwaters control the rehydration of anhydrite to gypsum, a transition that takes place at about 100 m depth (Cooper 1986, 1988). Continued invasion and circulation of meteoric groundwaters may then dissolve carbonate cements and gypsum vein/fracture infills, leading to the formation of voids and even collapse of strata. In addition, the fractures may be related to the plastic deformation of salt and brittle deformation of more competent non-halite interbeds (thin claystone/mudstone/evaporite rhythms), which show joints and fractures and pull-apart boudinage structures (i.e. potential voids are created). Continued ‘flow’ of the enclosing halite deposits leads to the non-halite beds eventually forming layers of breccias with rotated blocks (e.g. Smith 1996; Warren

65

2006). The halite infilling the fractured non-halite interbeds could have a number of origins.

Infill of fractures by precipitation from brines In addition to fluids expelled from siliciclastic sequences in sedimentary basins during burial, fluids exist in evaporite sequences and these may become mobilized during burial. Fibrous infilling material is suggestive of precipitation from saturated brines that remain within the halite deposits during burial (e.g. Borchet & Muir 1964; Sonnenfield 1984; Warren 2006). The fracture infills may have originated during early burial, with autobrecciated sequences and dessication cracks infilled with minerals such as halite and anhydrite precipitated out from the saltwater body or substrate. Hypersaline brines that do not dissolve the original precipitate move through the precipitate along intercalations, encountering cracks or voids in the non-halite beds in which halite and/or anhydrite may be precipitated. Hypersaline brine movement may occur as a result of gravitational instability allowing the brine to penetrate down, up or laterally into the precipitate (halite) body during its burial (e.g. Borchet & Muir 1964; Sonnenfield 1984), or during burial the brine is expelled due to increasing pressure. Halite deposits naturally also contain small volumes (typically 0.05 wt%) of original saturated brine, contained as either inclusions within the salt crystal lattices or along crystal/grain boundaries as films or in minor voids (refer Carter & Hanson 1983; Urai et al. 1986; Spiers et al. 1986, 1988, 1989, 1990; van Keken et al. 1993). These minor amounts of fluid migrate gradually, under overburden or tectonic pressures, through the salt crystals and salt body. Studies on salt in the USA have estimated brine inflow rates of 0.5 to 3.0 ml/ day at monitored boreholes, which amounted to between 2 and 10 litres per hole after about 20–30 years, approaching zero with time (Bradshaw & McClain 1971; Carter & Hanson 1983). Migration paths of fluid inclusion ‘bubbles’ within rock salt specimens subject to temperature gradients in laboratory tests are described (Carter & Hanson 1983). Small volumes of liquid may, therefore, have moved through and along grain boundaries, to potentially crystallise in cracks and voids in non-halite beds and seal fractures. The dehydration of any interbedded gypsum to anhydrite above 42 8C or at depths of around 1000 m (dependent upon the geothermal gradient), may also release water into a salt deposit that may lead to some dissolution of the halite deposit and the formation of a brine solution from which halite might be precipitated elsewhere. Equally, as described above, the rehydration of anhydrite to

66

D. J. EVANS & S. HOLLOWAY

gypsum by invasion of (meteoric) water causes fracturing and contemporaneous infill of the fractures by gypsum in an ongoing crack –seal process.

Non-brine infill origins (and fractures)

been dissolved, or intersected by human activity in the form of boreholes or mining operations (Warren 2006). The Boulby Potash Mine in NE England has suffered blowouts during mining operations due to the presence of high-pressure gas pockets in shaly parts of the potash (Corbett 1996). The presence of such gas in the Permian deposits of NE England may be of significance when considering the construction and operation of gas storage caverns. There is also mention of gas in the well logs for one brine well in the Preesall Halite of the Preesall salt field, Lancashire. The occurrence was likened to similar gas provings in the Cheshire Basin, which were analysed as mainly nitrogen.

Some fracture infills may have arisen from processes other than precipitation from saturated brine(s). As halite deforms by crystal plastic deformation mechanisms, more competent non-halite interbeds fracture and potential voids are created. As the fractures in the non-halite beds appear and propagate, the halite ‘flows’ (in a process akin to extrusion) around separated or boudinaged sections of competent materials, infilling the fractures. The resulting blocks of competent non-halite beds may be forced upwards or downwards into the enclosing halite deposits and may be rotated through any angle. It is also conceivable that under burial and increasing pressure, or due to some tectonic event causing an increase in pressure, some pressure solution of the salt may take place, similar to the process of fluid assisted grain boundary diffusional creep. If this occurs, small amounts of liquid may move along grain boundaries, migrate to and crystallize in, cracks and voids in non-halite beds, potentially sealing fractures.

The UK salt industry has a long history, with salt being derived from natural brine since at least 816 AD . In 1670 rock salt was discovered at depth in the Cheshire Basin, which ultimately led to commercial extraction by mining, solution mining and drilling to the wet rock head in order to pump natural or ‘wild’ brine (BGS 2006). Large-scale salt mining and industrial solution mining commenced in the nineteenth Century and continues in the Cheshire Basin and areas of NE England today.

Gases in halite deposits

Rock salt

Although rock salt is effectively gas tight, gas is known to occur in halite beds and in massive halokinetic salt structures. Most domal salts in Louisiana are classified ‘gassy’ (Hinkebein et al. 1995) and miners have long known about the dangers of gas within salt formations: several fatal accidents in conventional salt mines have been caused by outbursts of methane gas and associated saltfalls (Golden 1981; Molinda 1988; Hinkebein et al. 1995). The migration of methane gas over long distances through major salt deposits, facilitated by fractures has also been postulated in Portugal, Yemen and NE Brazil (Terrinha et al. 1994; Davison et al. 1996). The source of the gas is generally attributed to the fact that many evaporite sequences contain organic rich layers, thought to have been produced during algal blooms (Warren 2006). On burial, gases are produced from these layers by bacterial sulphate reduction at low temperatures and nonbiogenically by thermochemical sulphate reduction at higher temperatures. Thin impurity-rich salt beds interlayered with beds having some porosity and permeability (so-called ‘carrier beds’) leak small amounts of volatiles. Movement out of the salt, however, is unlikely until a pathway is available, which is generally as a result of the salt having

In 2005, salt production in the UK was around 6 million tonnes, over 95% of which was produced in England, with the remainder mined in Northern Ireland (BGS 2006). Two companies presently produce rock salt in England. Salt Union Ltd operates the Winsford (Meadowbank) Mine at Winsford, near Northwich in Cheshire, which commenced extraction in 1844, and Cleveland Potash Ltd, produces rock salt as a by-product of potash mining at the Boulby Mine in the North York Moors National Park. ICI developed the Boulby reserves in NE England during the 1960s as fertiliser demand grew and the UK was reliant on imported potash. Construction started in 1969 and the first product was produced in 1973 from depths between 1200 and 1500 m. In Northern Ireland, the Kilroot Mine near Carrickfergus is operated by the Irish Salt Mining and Exploration Company Limited.

Current extraction of salt in the UK

Brine Three companies, all based in the Cheshire Basin area, produce brine: the privately owned (and largest) group INEOSChlor Ltd, British Salt (a subsidiary of US Salt Holdings) and the smaller New Cheshire Saltworks Ltd (BGS 2006). The producing

POTENTIAL OF ONSHORE UK SALT DEPOSITS

brinefields are at Holford (Lostock Gralam), Warmingham and Wincham, near Northwich (Figs 3 & 8). Brine is extracted from the Northwich Halite Member, with at Holford, up to 50 cavities currently used for brine extraction (BGS 2006). Salt caverns produced by controlled brine extraction (for chemical feedstock and salt manufacturing) have been used in the UK for storage purposes since the late 1950s and early 1960s, when the storage of manufactured gas at Teesside was reported (Northolt & Highley 1973; Beutel & Black 2005). Caverns are numerous and are currently used to store natural gas, hydrogen, various liquid hydrocarbons, other fluids, solids and waste products. To date, salt caverns in the UK have not been used to develop compressed air energy storage facilities, although they are in use and under development elsewhere for this purpose at, for example, Huntorf in Germany and McIntosh, Alabama, USA; these facilities having been constructed in 1978 and 1991 respectively (Bary et al. 2002).

Natural gas storage potential in the UK salt fields Although there are no set minimum depths or thicknesses of halite beds that can be used for gas storage, economic limits will ultimately dictate where gas storage cavern facilities can be developed. The depth of the halite limits not only the height and size of the cavity, but the pressure at which the gas can be stored: the shallower the cavity the smaller it can be and the lower the confining pressures then the lower the storage pressures. Also the shallower the salt, the more likely it is that wet rock head conditions could impact on the site. Clearly, certain UK salt fields do not provide adequately thick or deep enough beds of halite in which to develop natural gas storage caverns, e.g. those in the Carlisle Basin and Staffordshire. However, the Triassic Northwich Halite in the Cheshire Basin, the stratigraphically equivalent Preesall Halite in west Lancashire, the Dorset Halite in Dorset and the Permian Z2 salt in NE England have potential because they are deep enough or thick enough. These areas either contain existing natural gas storage caverns or have current planning applications for storage cavities. Of the other fields, the Larne area in particular, has good prospects that have yet to be fully appraised. There may also be potential in the Worcester Basin but this is speculative at present. The Somerset salt field as proven to date is unlikely to be prospective because the salt is mainly in haselgebirge facies rather than thick beds of pure salt. However, a gas storage cavern was leached out in haselgebirge facies salts of Permian age at depths between 1305 m and

67

1400 m at Kiel in Germany (Coates et al. 1981; Be´rest & Brouard 2003). Due, however, to the high insolubles content of the halite the effective cavern volume was reduced to less than 60% of the initial 68 000 m3 cavern volume. Prospects on Walney Island in NW England are probably unrealistic given the likely limited extent of the proven halite beds, their shallow depth, wet rock head conditions and the fact that the area is poorly understood. The thickest salt bed is around 46 m, but its top is only a maximum of 226 m below ground level. This is shallower than the caverns in the Northwich Halite in Cheshire and the unmined Preesall Halite being considered for storage caverns in the area of the Wyre Estuary. Similarly, the salt beds proved to the NE of Southport are poorly known and are likely to be too shallow for underground gas storage purposes.

Operational, developing and planned salt cavern gas storage sites onshore UK There are a number of planned or currently operational gas storage facilities onshore in the UK (Table 1). These are reviewed briefly below. Government is also keen that salt deposits present in the East Irish Sea and southern North Sea and suitable for gas storage development are developed to complement the requirement for gas infrastructure (Smith et al. 2005; DTI 2006 h). The first project is now being planned in thick Preesall Halite deposits of East Irish Sea: the ‘Gateway Storage Project’, around 30 km offshore to the SW of Barrow-in-Furness (Stag Energy 2006, 2007).

Holford (H165), Cheshire (Triassic halite) A former brine production cavern (H165) in the Northwich Halite Member in the Holford Brinefield, Cheshire (Fig. 8), was converted by ICI in 1984 into a gas storage cavern. It was originally leased to Transco, providing diurnal storage, and operation has since been transferred to INEOSChlor; it is used by National Grid for daily balancing in the local distribution zone. Following a 10-year inspection, INEOS recommenced operation in conjunction with EDF Trading in late January 2007 (INEOS 2007). EDF Trading owns the nearby Hole House gas storage facility and has gained significant experience in developing, operating and trading fast cycle salt cavity storage.

Hole House, Cheshire (Triassic halite) The Hole House facility, west of the village of Warmingham near Crewe, Cheshire is a gas storage facility developed in the Triassic Northwich

68

Table 1. Summary of operational and planned underground gas storage facilities on and offshore the UK, including depleted oil/gasfield, salt caverns and the chalk LPG storage facilities

D. J. EVANS & S. HOLLOWAY

69

(Continued)

POTENTIAL OF ONSHORE UK SALT DEPOSITS

Table 1. Continued

70

D. J. EVANS & S. HOLLOWAY

POTENTIAL OF ONSHORE UK SALT DEPOSITS

Halite Member (Figs 8 & 20) of the Warmingham Brinefield, which is owned and operated by British Salt. Permissions and consents were originally granted to Aquila Energy Limited in 1995. Commercial operations began in February 2001. The facility was acquired by EDF Trading Limited in October 2002, since when it has been operated by EDFs subsidiary company Energy Merchant Gas Storage (UK) Limited (Beutel & Black 2005). It is linked to the Transco gas transmission system. Phase I of the project saw the construction of two cavities, each of approximately 30 mcm (million

71

cubic metres, mcm; c. 150 GWh) and a gas processing plant, which became operational in March 2003. The caverns have been designed to provide a high flexibility facility, capable of supplying ‘peak gas’. British Salt uses the brine produced during the cavern washing process. The salt is up to 230 m thick and the tops of the caverns are at about 300– 400 m depth, slightly shallower than at the Holford and Byley sites (Beutel & Black 2005). Gas can be delivered at a rate of 2.8 mcm/ day and injected at 5.6 mcm/day (UK Gas Report 2005). Phase II saw two additional salt cavities

Fig. 20. Schematic sections of the cavern development in the Northwich Halite for the Byley and Hole House areas in the Cheshire Basin (based upon Evans et al. 1968; Earp & Taylor 1986; Beutel & Black 2005).

72

D. J. EVANS & S. HOLLOWAY

constructed, providing a further 30 mcm (150 GWh) storage, and an upgrade of the gas processing plant with gas again delivered at 2.8 mcm/day and injected at 5.6 mcm/day. EDF expects the first of these two cavities to be commissioned by the end of 2006, and the second by the end of 2008 (UK Gas Report 2005).

Byley/Holford Cheshire (Triassic halite) The Byley gas storage scheme is being developed in the Triassic Northwich Halite Member (Figs 8 & 20), around the Drakelow Lane area, south of the Holford Brinefield, owned by INEOSChlor. It is also referred to as the Holford storage scheme. Planning consent was not granted when originally applied for by Scottish Power in 2002. Scottish Power appealed the decision, which led to a Public Inquiry in late 2002. Following the Inspector’s decision and after an intervention by the then Deputy Prime Minister John Prescott, consent was granted in May 2004 in the national interest. It was confirmed after a legal challenge against the intervention (filed in June 2004) failed in December 2004, and work commenced clearing and preparing the site in March 2005. In July 2005 Scottish Power sold its rights to Eon UK. In August 2005 work on the infrastructure commenced and Eon had planned to start the brining process for the caverns in the summer of 2006. The facility will be connected to the National Transmission System (NTS) by a 4 km long pipeline. INEOSChlor owns the brinefield/salt and will undertake the solution mining, with Eon leasing the caverns and owning/operating the infrastructure. The plans are for eight cavities providing around 170 mcm storage capacity, with a deliverability of 16 mcm per day emptying in around 10 days, and injectability of 8 mcm/day. It was estimated that Phase 1 cavern washing would commence in summer 2006, providing about half the space (4 caverns) and would be completed by 2008. The second phase would be completed (and full capacity achieved), by 2010 (UK Gas Report 2005). Cavern tops will be between 630 m and 730 m below ground (Beutel & Black 2005) and the latest designs indicate the base of the caverns are likely to be at the level of the ‘Thirty Foot Marl’, which will form the cavern ‘sump’. Earlier designs had indicated that the ‘Thirty Foot Marl’ would lie at a level between half to two thirds of the way up the caverns (e.g. Beutel 2002).

Stublach, Cheshire (Triassic halite) The Stublach gas storage facility is located south of the Holford Brinefield between Drakelow Lane and Lach Dennis and immediately north of the Byley

facility (Fig. 8). The proposal to develop the Stublach facility, comprising 28 caverns in the Northwich Halite, will provide around 540 mcm capacity. The cavities will be bell-shaped, approximately 100 metres in height, with their tops at around 550 metres depth. INEOS Enterprises Limited submitted the planning application to Cheshire County Council in December 2005. Following a council meeting, planning permission was granted in June 2006, subject to the Government not calling it in (Cheshire Council 2006; INEOS 2006). In July 2006 the DTI ‘confirmed that whilst recognising that each case must be decided on its own merits, the Energy Markets Unit of the DTI believes that new gas storage projects would be invaluable from an energy policy perspective’ and in December 2006 the Government confirmed that it would not be calling for an Inquiry and gave the go ahead for development. Hazardous substance consent was also granted. The project is thus fully consented and development can commence. The statements by council members following the meeting said that ‘As far as [the council] are concerned this application is a very difficult thing to refuse on local grounds . . . Byley was a long time ago and things are now vastly different . . . Members felt that the need for gas storage had been more clearly identified than when considering the Byley application . . . And today felt that that national need was more important than all other planning considerations.’ However, council Members also resolved that should planning permission be granted, a legal agreement be entered into with the applicant and that over 80 stringent planning conditions would be required. These views and statements are very different to the opposition and attitudes widely displayed towards earlier proposed schemes. It may illustrate a change in the perception of underground gas storage by local government, perhaps linked to the publicity surrounding issues such as security of supply and volatile/higher gas prices for consumers. In late August 2007, Gaz de France announced that an agreement had been signed with INEOS Enterprises for the commercial development of the proposed salt cavern storage facility (GDF 2007). INEOS will continue to be involved in the development of the facility, which will involve the construction of up to 28 caverns by solution mining. The brine will be used for industrial purposes. The first phase of the Stublach gas storage facility remains on track to commence cavity development in 2009, with commissioning of the first caverns anticipated in 2013 and the remaining caverns developed through to 2018. The Group will operate the infrastructures under a 30 year lease agreement running until 2037.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

King Street, Cheshire (Triassic halite) NPL Estates, through its wholly-owned subsidiary King Street Energy Ltd, is proposing to develop a salt cavern gas storage facility near Rudheath in Cheshire (NPL 2007). The site will be to the north of other sites being developed or proposed at Byley and Stublach (Fig. 1). The proposed site, referred to as the King Street development, was formerly part of the operational Holford brinefield and has an existing planning consent for brining and underground waste disposal. NPL proposes leaching nine caverns, each with a volume of 400 000 m3, in the Northwich Halite some 400 m below the surface. It is reported that up to 216 mcm of gas will be stored in total, of which up to 126 mcm will be working gas during normal operations. The site will require further detailed geological investigation involving drilling and other exploration activities. The supporting gas processing facility will be located on the former Octel Site on the northern edge of the Holford brinefield near Lostock Gralam. The site extends to about 16 acres and is remote from the local community. Once completed the wellheads will be secured in small compounds and fully screened from the surrounding area. NPL proposes construction of a twin pipeline system between the Mersey Estuary and the King Street site to supply leaching water and to discharge the weak brine. Other gas storage projects in the district take water from the local rivers and pass the brine to process users. However, the rivers have little remaining abstraction capacity and there is no scope for local companies to process more brine for some time to come. The pipeline system will include pumping stations at both ends of the line and one set at approximately the half-way point. These facilities will be largely underground. There will be a need for an intermediate storage tank system at the King Street end to provide a buffer between the brining and pipeline operations. (Note added in proof: Planning permission was refused in December 2008.)

Preesall/Wyre, Lancashire (Triassic halite) Canatxx Gas Storage Ltd planned to develop a salt cavern storage facility in the Preesall Halite of the Lancashire salt field (Figs 1 & 14), which was worked by both mining and solution mining until the final brine extraction operations closed in 1993 (Thompson 1908; Landless 1979; Wilson & Evans 1990; BGS 2006). In 2003, Canatxx submitted a planning application to develop up to 24 caverns, providing storage space for between 1200 and 1700 mcm of gas, in the unworked Preesall Halite to the west of the existing brinefield, beneath areas

73

of the River Wyre Estuary. Pre-existing salt cavities arising from the brine extraction process have already been used by ICI for the storage of hazardous materials. Limitations exist on the area of development due to the salt coming to crop in the east and south, former brineworks and by large areas of new housing in areas of unworked halite in the north of the saltfield. The unworked halite is between 140 m and 240 m thick and the top of the caverns will be at depths between 350 and 400 m. Caverns, the heights of which will depend upon local geological conditions and where in the salt field they are sited, will be accessed via deviated wells from clusters of wellheads. The facility has an anticipated operational life of 25 years and will be connected to the National Transmission System by a pipeline, with deliverability estimated at as much as 114 mcm/day. The planning application received strong opposition, both from the local residents and the planning authority and in December 2004, Lancashire County Council voted to oppose the scheme. The application went to a Public Inquiry that ran from October 2005 to May 2006. On 17 October 2007, Hazel Blears, Secretary of State for Communities and Local Government (CLG), announced that the Government was dismissing the Canatxx appeal and refusing Planning Permission and Hazardous Substances Consent for development of a natural gas storage facility. The plans were rejected mainly on the grounds of the impact on the local environment and on safety issues (DCLG 2007). Canatxx has indicated it will not appeal the decision but may resubmit revised plans.

Isle of Portland, Dorset (Triassic halite) In April 2005, Egdon Resources announced plans to develop a high deliverability salt cavern storage facility in Triassic salts beneath the Isle of Portland in Dorset (Figs 7, 12 & 13). In February 2005 Portland Gas Limited was established as a wholly owned subsidiary of Egdon Resources. Portland Gas signed an agreement with Portland Port Limited in April 2005 to lease a 5 hectare ‘brownfield site’ at the former naval base HMS Osprey for a period of up to 90 years. On 16 January 2008, Portland Gas plc demerged from Egdon Resources Plc. If the project gains approval, Portland Gas Storage Limited, forming a wholly owned subsidiary of Portland Gas plc, would own and operate the Portland facility. Work began with German cavern design experts DEEP on a feasibility study of the potential capacity and operating parameters for a gas storage facility in the Triassic salts of the Weymouth and Portland

74

D. J. EVANS & S. HOLLOWAY

area. A seismic line was acquired in May 2005 and an exploration well (Portland 1) was drilled and completed in June 2006. This proved the presence and thickness of the Triassic saliferous beds (Egdon 2006a). Initial estimates were that the storage facility would have potential to provide up to 10% of the UK gas demand on a typical winter day and provide a storage facility for 1% of the UK annual consumption. The storage facility was planned for development in three phases of six caverns. Each phase would bring a working storage volume of 330 mcm. The project was designed with gas export capabilities to the national gas grid increasing from 18 to 54 mcm/day through the three phases. In September 2006 a technical feasibility study confirmed the potential for cavern storage over an approximate 20 km2 area. Plans were revised with the proposal to construct 14 caverns storing up to 1000 mcm of natural gas. Injection and withdrawal rates are planned to be around 20 mcm/day, which would permit the filling and emptying of the entire storage volume in 50 days (Egdon 2006b). The caverns will be at depths greater than 2000 m and up to 100 m high (Fig. 13) and operate in brine compensated mode. As a result of the feasibility study, the Environmental Statement, planning and pipeline construction authorization applications were submitted in March 2007. Planning permission was granted by Dorset County Council on 16 May 2008. It is anticipated that initial storage capacity will be available during the winter of 2011, with full capacity being available by the winter of 2015 (Portland Gas 2008).

Hornsea/Atwick (Permian halite) The Hornsea gas storage facility in East Yorkshire (Figs 1 & 2) was built originally by British Gas Corporation and became operational in 1979, providing storage and peak shaving supply to the National Transmission System (NTS). The facility was bought by US energy company Dynergy in 2001 and sold the following year to the current owners and operators SSE Hornsea Ltd, part of Scottish & Southern Energy plc (UK Gas Report 2005). The facility comprises a central processing area and nine salt cavities leached into the main salt of the Fordon Evaporites (Z2) at depths of between 1730 km and 1800 m below the surface (Beutel & Black 2005). The size and volume of the caverns varies due to the variations in thickness of the salt, with the facility providing a total of about 325 mcm of gas storage space. Gas can be injected at about 2 mcm/day and withdrawn at up to 18.5 mcm/day (UK Energy Report 2005).

Aldbrough South (Permian halite) Plans for an underground gas storage facility to the south of SSE’s Hornsea (Atwick) gas storage site, at Aldbrough in East Yorkshire (Figs 1 & 2) have been ongoing since 1997. In 1997, two separate planning applications were submitted, one by British Gas (BG) at Aldbrough ‘North’ (six caverns) and a second by Intergen at Aldbrough ‘South’ (three caverns). Both applications were rejected, due to strong local objections and a further application was also refused precipitating, in 1999, the first Public Inquiry to be held into the planning and siting of such facilities (Beutel & Black 2005). The Inquiry resulted in the Government granting permission to both BG and Intergen to proceed with plans to develop the two facilities. In 2001, ownership of Aldbrough ‘North’ passed to Dynergy with its acquisition of BG assets, which they then sold to Scottish & Southern Energy (SSE) in 2002. In 2003 Intergen sold Aldbrough ‘South’ to Statoil (UK) Ltd. The new owners combined the two projects in late 2003 (referred to here as Aldbrough South, Phase I; Table 1). The joint venture, estimated to cost £225 million, will generate a total storage capacity of around 420 mcm, with SSE owning 280 mcm storage space and Statoil 140 mcm. Injectability is thought likely to be around 20 mcmd and deliverability c. 40 mcmd (UK Energy Report 2005). Site work commenced in March 2004 and leaching of the first of the nine planned caverns began in March 2005. Cavern tops will be between 1800 and 1900 m below ground, with the first five caverns expected to be ready for commercial use around October 2007. The remaining four are likely to be completed by 2009 (UK Energy Report 2005). The plant will be operated remotely, the intention being to use above ground facilities already on site. Aldbrough South Phase II will develop nine further caverns on land adjacent to Phase I operations and will double the amount of gas to be stored (Rigzone 2006). Permission was granted in May 2007.

Aldbrough North: Whitehill (Permian halite) During 2006, Eon UK carried out geological investigations into the suitability of the Permian salt for underground gas storage to the north of Aldbrough in East Yorkshire. In mid January 2007, a planning application to build the Whitehill facility comprising 10 caverns, was submitted to East Riding of Yorkshire Council. The facility will provide an anticipated total working gas capacity of 420 mcm of gas, with a deliverability of 40 mcm/day.

POTENTIAL OF ONSHORE UK SALT DEPOSITS

Approval was granted in late 2008 (EYDG 2005; Wingas 2008). The first phase could be operational in 2010, with completion of work planned by 2013.

Teesside: Billingham (Saltholme) and Wilton (Permian halite) The Teesside salt field in south Durham is formed by the (Middle or Main) Boulby Halite Formation (Z3) which overlies the Billingham Main Anhydrite (Figs 1 & 18). Salt was extracted at Greatham by controlled brine pumping from around 1822 until at least 1969. As early as 1959, the Northern Gas Board used a solution-mined cavity to store town gas (Notholt & Highley 1973). Small ICI brine production caverns to the north of the River Tees at Saltholme and south of the river at Wilton, have been converted and used for storing light hydrocarbons and the various fluids and gases associated with oil refining since 1960. In the Teesside area, the Boulby Halite lies at depths of between 274 m and 366 m, deepening eastwards offshore and is up to 45 m thick (Notholt & Highley 1973). At least four caverns at Saltholme owned by INEOSChlor, with volumes of 10 000 m3 –30 000 m3 (providing a total net cavern volume of 0.08 mcm), were converted to gas storage by Huntsman (Beutel & Black 2005). Injection and withdrawal rates are not available. Additionally, caverns in the Teesside (Billingham) area have, for many years, been used to store other liquids and gases such as nitrogen and hydrogen.

Northern Ireland (Triassic and Permian halite) Interest is being shown in the potential for developing salt cavern storage facilities in halite beds of both Triassic and Permian age in the Larne Basin, Northern Ireland. As described above, the main halites are of Triassic age and have been proved in the Larne 1 and 2 boreholes (Figs 5 & 19). On 24 July 2007, Egdon Resources Plc announced that a wholly-owned subsidiary, Portland Gas NI Limited was being granted an exploration licence from The Crown Estate to evaluate the suitability of the Permian halite sequence to create gas storage caverns below Larne Lough, County Antrim (Egdon 2007). The halite sequence was proved by the Larne No. 2 borehole, which was drilled in 1981 close to the docks in Larne, encountering a 113 m thick sequence of Permian halite at a depth of 1688 m (Fig. 19). Portland Gas carried out a seismic survey during October 2007, to establish the extent of the halite proved in the borehole.

75

Conclusions The clear requirement for increased natural gas storage capacity in the UK is reflected in the plans for the development of several (underground) salt cavern storage facilities. The best prospects for the development of salt storage caverns onshore are the Triassic Northwich Halite in the Cheshire Basin, the Preesall Halite in west Lancashire and the salt of the Permian Fordon Evaporites (Z2) in NE England. Of the other halite-bearing strata, the Dorset salt field in particular, and the Permian and Triassic halite beds of the Larne area, offer good prospects. However, the Wessex Basin halite onshore is of limited extent and the useful area of halite may already be bound up in the licence areas held by Portland Gas. The Triassic and Permian halites of the Larne area have yet to be fully appraised. Potential may also exist in the deeper parts of the Worcester Basin, but extensive halite is unproven at present. The Somerset and Staffordshire salt fields and the Ro´´t Halite Member in eastern England appear to offer little prospectivity. The halite formations on Walney Island have yet to be fully appraised, but are likely to be either too shallow, too thin, of limited extent and affected by wet rock head conditions as are the poorly known halites in the Southport area of Lancashire. We are grateful to D. Highley, MBE for his support of this work and helpful discussions. A. S. Howard, G. A. Kirby and N. J. Riley, MBE are thanked for their helpful and constructive comments and reviews of the manuscript. Thanks are also due to E. Hough, A. Cooper and R. G. Crofts for helpful discussions and to K. Sanderson (SABIC UK Petrochemicals) and R. Stevenson (INEOS) who provided information relating to salt caverns in the Teesside and Cheshire Basin areas, respectively. The paper was improved by the careful reviews of the two referees D. Highley, MBE and M. A. W. Abbott to whom we extend our thanks. L. Noakes and R. Demaine are thanked for diagram production. This paper is published with the permission of the Executive Director, British Geological Survey, Natural Environment Research Council.

References A RTHURTON , R. S. 1973. Experimentally produced halite compared with Triassic layered halite-rock from Cheshire, England. Sedimentology, 20, 145–160. B ARCLAY , W. J., A MBROSE , K., C HADWICK , R. A. & P HARAOH , T. C. 1997. Geology of the country around Worcester. Memoir British Geological Survey (England and Wales), sheet 199, HMSO, London. B ARY , A., C ROTOGINO , F., P REVEDEL , B., ET AL . 2002. Storing natural gas underground. Oilfield Review, Summer 2002, 2 –17. B EAUHEIM , R. L. & R OBERTS , R. M. 2002. Hydrology and hydraulic properties of a bedded evaporate formation. Journal of Hydrology, 259, 66–88.

76

D. J. EVANS & S. HOLLOWAY

B E´ REST , P. & B ROUARD , B. 2003. Safety of salt caverns used for underground gas storage. Oil & Gas Science and Technology, 58, 361–384. B ERRIDGE , N. G. & P ATTISON , J. 1994. Geology of the country around Grimsby and Patrington. Memoir British Geological Survey, England & Wales, Sheets 81, 82, 90, 91. HMSO, London, for the British Geological Survey. B EUTEL , T. 2002. Assessment of Geological and Geotechnical Site Suitability. Proof of Evidence presented to the Byley Gas Storage Public Inquiry, APP/P4/1. B EUTEL , T. & B LACK , S. 2005. Salt deposits and gas cavern storage in the UK with a case study of salt exploration in Cheshire. Oil Gas European Magazine, 1/2005, 31–35. BGS, 2006. Mineral Planning Factsheet — Salt. British Geological Survey, for the Office of the Deputy Prime Minister. http://www.bgs.ac.uk/mineralsuk/ downloads/mpfsalt.pdf B ORCHERT , H. & M UIR , R. O. 1964. Salt Deposits: the origin, metamorphism and deformation of evaporates. University Series in Geology, D. Van Nostrand Company, Ltd., London. B RADSHAW , R. L. & M C C LAIN , W. C. 1971. Project salt vault: A demonstration of disposal of high activity solidified wastes in underground salt mines. Oak Ridge National Laboratory, Tennessee, ORNL-4555. B RIDGE , D. MC C. & H OUGH , E. 2002. Geology of the Wolverhampton and Telford district. Sheet description of the British Geological Survey 1:50 000 Series Sheet 153 (England & Wales). Keyworth, Nottingham. B RUNE , G. 1968. Anhydrite and gypsum problems in engineering geology. Engineering Geology (Sacremento), 2, 26–33. BS. 1998. BS EN 1918-3: gas supply systems — underground gas storage — Part 3: Functional recommendations for storage in salt cavities. British Standards Institution (BSI), London. C AMERON , T. D. J., C ROSBY , A., ET AL . 1992. United Kingdom Offshore Regional Report: The Geology of the Southern North Sea. London, HMSO for the British Geological Survey. C ARRUTHERS , R. M., B EAMISH , D., ET AL . 1999. Regional interpretation of gravity and aeromagnetic data from Northern Ireland. Geological Survey of Northern Ireland Technical Report 99/1. C ARTER , N. L. & H ANSEN , F. D. 1983. Creep of rock salt. Tectonophysics, 92, 275– 333. C ELIA , M. A., B ACHU , S., N ORDBOTTEN , J. M., K AVETSKI , D. & G ASDA , S. 2006. A risk assessment tool to quantify CO2 leakage potential through wells in mature sedimentary basins. In: R ØKKE , N. A., B OLLAND , O., O’B RIEN , D. ET AL . (eds) The 8th International Conference on Greenhouse Gas Control Technologies, 19–22 June 2006, Trondheim, Norway, Elsevier, Abstracts volume. C HADWICK , R. A., J ACKSON , D. I., B ARNES , R. P., ET AL . 2001. Geology of the Isle of Man and its offshore area. Keyworth, Nottingham. British Geological Survey and Treasury Isle of Man. CHESHIRE COUNCIL . 2006. Major Scheme to Store Gas in Salt Cavities Given the Go Ahead. Cheshire County Council News Release, 20 June 2006, Ref 212/06. World Wide Web Address: http://www.cheshire.gov.

uk/NR/exeres/2CE02F3E-F2D4-420D-818C-B6944 F16173B,frameless.htm? C OATES , G. K., L EE , C. A., M C C LAIN , W. C. & S ENSENY , P. E. 1981. Failure of man-made cavities in salt and surface subsidence due to sulfur mining. Topical report RSI-0131, Submitted to Strategic Petroleum Reserve, Geotechnical Division, Sandia National Laboratories, Albuquerque, NM 87185. C OOPER , A. H. 1986. Subsidence and foundering of strata caused by the dissolution of Permian gypsum in the Ripon and Bedale areas, North Yorkshire. In: H ARWOOD , G. M. & S MITH , D. B. (eds) The English Zechstein and Related Topics. Geological Society, London, Special Publications, 22, 127–139. C OOPER , A. H. 1988. Subsidence resulting from the dissolution of Permian gypsum in the Ripon area: its relevance to mining and water abstraction. In: B ELL , F. G., C ULSHAW , M. G., C RIPPS , J. C. & L OVELL , M. A. (eds) Engineering Geology of Underground Movements. Geological Society, London, Engineering Geology Special Publications, 5, 387– 390. C OOPER , A. H. 2002. Halite karst geohazards (natural and man-made) in the United Kingdom. Environmental Geology, 42, 505 –512. C ORBETT , G. R. 1996. The Development of a Coal Mine Portable Microseismic Monitoring System for the Study of Rock Gas Outbursts in the Sydney Coal Field, Nova Scotia. MSc Thesis submitted to Department of Mining and Metallurgical Engineering, McGill University, Montreal, March 1996. World Wide Web Address: http://www.mining-technology. com/projects/boulby/index.html D AVISON , I., A LSOP , G. I. & B LUNDELL , D. J. 1996. Salt tectonics: some aspects of deformation mechanisms. In: A LSOP , G. I., B LUNDELL , D. J. & D AVISON , I. (eds) Salt Tectonics. Geological Society, London, Special Publications, 100, 1 –10. DCLG. 2007. Town and Country Planning Act 1990 (section 78), Planning (Hazardous Substances) Act 1990, Appeals by Canatxx Gas Storage Limited application refs: 02/04/1415 and hsc/05/01 land at Preesall salt field, Stalmine, Wyre Estuary, Lancashire. Department of Communities and Local Government (DCLG) letter to Mr R. Glover, 16 October 2007, Refs APP/Q2371/A/05/1183799 & APP/ HSC/05/07. World Wide Web Address: http:// www.communities.gov.uk/documents/planning-callins/ pdf/decision-appealcanatxx D EAN , F. 1978. Salt cavity storage. Gas Engineering and Management, Sept 1978. 291– 304. D EAN , F. 1985. The Hornsea salt cavity storage installation. Gas Engineering and Management, January 1985, 14, 20–26. D ICKINSON , J. 1882. ‘Reports of Inspectors of Mines’ for 1881, p. 53. D OYLE , J. B. 1853. Notes on the salt mine of Duncrue. Transactions of the Geological Society of Dublin, 5, 232–236. DTI. 2006a. Digest of United Kingdom Energy Statistics (DUKES) 2006. Department of Trade and Industry, URN No. 06/87. DTI. 2006b. Our Energy Challenge: Securing clean, affordable energy for the long term. Department of Trade and Industry, Consultation Document, January 2006.

POTENTIAL OF ONSHORE UK SALT DEPOSITS DTI. 2006c. The Energy Challenge: Energy Review Report 2006. Department of Trade and Industry, July 2006. DTI. 2006d. Understanding gas supply this winter. Department of Trade and Industry, January 2006. DTI. 2006e. Energy statement of need for additional gas supply infrastructure. Ministerial written statement, Secretary of State for Trade and Industry, 16 May 2006. DTI. 2006f. Joint Energy Security of Supply Working Group (JESS): Sixth Report. Department of Trade and Industry, April 2006. DTI. 2006g. Secretary of State’s second report to Parliament on security of gas and electricity supply in Great Britain. July, 2006, The Stationary Office. DTI. 2006h. Undersea salt caverns to boost UK gas supply. Department of Trade and Industry press release 12th January 2006. DTI. 2007. Meeting the Energy Challenge: A White Paper on Energy. Department of Trade and Industry, May 2007. World Wide Web Address: http://www.berr. gov.uk/files/file39387.pdf E ARP , J. R. & T AYLOR , B. J. 1986. The Geology of the Country around Chester and Winsford. Memoir of the British Geological Survey, Sheet 109. E GDON , 2006a. Portland Gas Storage Project Update. Egdon Resources Plc Press Release 28 June 2006. Wide Web Address: http://www.portland-gas.com. E GDON , 2006b. Portland Gas Storage Project Update. Egdon Resources Plc Press Release 4 September 2006. Wide Web Address: http://www.portland-gas. com. E GDON , 2007. Investigation of a site for a potential new gas storage facility in Northern Ireland. Egdon Resources Plc Press Release 24 July 2007. World Wide Web Address: http://www.portland-gas.com/ index2.php?option=com_content&do_pdf=1&id=65 E VANS , D. J. 2008. An appraisal of Underground Gas Storage technologies and incidents, for the development of risk assessment methodology. Report prepared by the British Geological Survey for the Health & Safety Executive 2008. Health & Safety Executive (HSE) Research Report RR605. World Wide Web Address: http://www.hse.gov.uk/research/rrpdf/rr605.pdf E VANS , D. J. 2009. A review of underground fuel storage problems and putting risk into perspective with other areas of the energy supply chain. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 173– 216. E VANS , D. J., H OUGH , E., C ROFTS , R. G., T ERRINGTON , R. & W ILIAMS , G. 2005. The geology of the Preesall salt field area. British Geological Survey, Internal Report, Number CR/05/183N. E VANS , D. J., R EAY , D. M., M ITCHELL , W. I. & B USBY , J. 2006. Appraisal of underground energy storage potential in Northern Ireland. British Geological Survey, Internal Research Report IR/06/095, prepared for the Department of Enterprise, Trade and Investment (DETI), Northern Ireland. E VANS , W. B. 1970. Triassic salt deposits of north-western England. Quarterly Journal of the Geological Society of London, 126, 103–123.

77

E VANS , W. B., W ILSON , A. A., T AYLOR , B. J. & P RICE , D. 1968. The Geology of the Country around Macclesfield, Congleton, Crewe and Middlewich. Memoir of the Geological Survey of Great Britain, Sheet 110. EYDC. 2008. East Yorkshire District Council Website. World Wide Web Address: http://www.eastriding.gov. uk/PublicAccess731c/tdc/dcapplication/applicationdetailview.aspx? case no=JCD5IMBJFA000 F ROST , D. V. 1998. Geology of the Country around Northallerton. Memoir British Geological Survey, England and Wales, Sheet 42, HMSO, London. G AUNT , C. D., F LETCHER , T. P. & W OOD , C. J. 1992. Geology of the Country around Kingston upon Hull and Brigg. British Geological Survey Memoir, Sheets 80 and 89 (England and Wales), HMSO, London. GDF. 2007. Gaz de France to Develop Large Gas Storage Facility in England. Gaz de France Press Release, Tuesday 28 August 2007. World Wide Web Address: http://www.oilvoice.com/Gaz_de_France_ to_Develop_Large_Gas_Storage_Facility_in_Engla/ 10561.htm G OLDEN , F. 1981. Hideaways for Nuclear Waste. Time, Monday March 16. World Wide Web Address: http://www.time.com/time/magazine/article/0,9171, 922475-1,00.html G RIFFITH , A. E. & W ILSON , H. E. 1982. Geology of the Country around Carrickfergus and Bangor. Memoir of the Geological Survey of Northern Ireland, Sheet 29, HMSO, Belfast. H ARVEY , M. J. & S TEWART , S. A. 1998. Influence of salt on the structural evolution of the Channel Basin. In: U NDERHILL , J. R. (ed.) Development, Evolution and Petroleum Geology of the Wessex Basin. Geological Society, London, Special Publication, 133, 241– 266. H EATHCOTE , J. A. & M ICHIE , U. MC L. 2004. Estimating hydrogeological conditions over the last 120 ka: an example from the Sellafield area, UK. Journal of the Geological Society, London, 161, 995–1008. H INKEBEIN , T. E., B AUER , S. J., E HGARTNER , B. L. ET AL . 1995. Gas intrusion into SPR caverns. Albuquerque NM: Sandia Laboratories. Report SAND94-0023. World Wide Web Address: http:// www.osti.gov/energycitations/servlets/purl/2064923f3rvB/webviewable/206492.pdf H OWELL , F. T. 1984. Salt karst of the Cheshire Basin, England. In: C ASTANY , G., G ROBA , E. & R OMIJN , E. (eds) Hydrogeology of karstic terranes. International Association Hydrogeologists, International Contributions to Hydrogeology, 1, 252– 254. H OWELL , F. T. & J ENKINS , P. L. 1976. Some aspects of the subsidences in the rock salt districts of Cheshire, England. In: Proceedings Anaheim Symposium, December 1976. International Association of Hydrological Science Publications, 121, 507–520. H OWELL , F. T. & J ENKINS , P. L. 1980. Interaction between saline interfaces and rock salt beds, Cheshire, England. In: 3rd International Symposium on Water Rock-Interfaces, Edmonton, July 1980. Alberta Research Council and International Association of Geochemistry 93–94. INEOS . 2006. INEOS Enterprises secures planning permission for new strategic UK gas storage facility

78

D. J. EVANS & S. HOLLOWAY

in Cheshire. INEOS press release, 29 June 2006. World Wide Web Address: http://www.ineos.com/ new_item.php?id_press=132 INEOS . 2007. INEOS Enterprises and EDF Trading start operation of new gas storage facility. Press Release 29 January 2007. World Wide Web Address: http:// www.ineos.com/new_item.php?id_press=149 J ACKSON , D. I. & M ULHOLLAND , P. 1993. Tectonic and stratigraphic aspects of the East Irish Sea Basin and adjacent areas: contrasts in their postCarboniferous structural styles. In: P ARKER , J. R. (ed.) Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. Geological Society, London, 791– 808. J ACKSON , D. I. & J OHNSON , H. 1996. Lithostratigraphic nomenclature of the Triassic, Permian and Carboniferous of the UK offshore East Irish Sea Basin. British Geological Survey, Keyworth, Nottingham. J ACKSON , D. I., M ULHOLLAND , P., J ONES , S. M. & W ARRINGTON , G. 1987. The geological framework of the East Irish Sea Basin. In: B ROOKS , J. & G LENNIE , K. W. (eds) Petroleum Geology of North West Europe. Graham & Trotman, London, 191–203. J ACKSON , D. I., J ACKSON , A. A., W INGFIELD , R. T. R., B ARNES , R. P. & A RTHUR , M. J. 1995. The Geology of the Irish Sea. British Geological Survey, United Kingdom Offshore Regional Report, HMSO, London. J OHNSON , E. W., S OPER , N. J. & B URGESS , I. C. 2001. Geology of the Country around Ulveston. British Geological Survey, Memoir, Sheet 48 (England and Wales), HMSO, London. KDHE. 1997. Letter to Dr L. C. Gerhard, Director Kansas Geological Survey, from J. J. O’Connell, Secretary Kansas Department Health and Environment (KDHE) January 22 1997. World Wide Web Address: http://www.kgs.ku.edu/Hydro/Hutch/GasStorage/jan_22_97.pdf L ANDLESS , V. 1979. Preesall salt mines. Memoirs of the Northern Mine Research Society, British Mining, 11, 38–43. M C M ILLAN , A. A., H EATHCOTE , J. A., K LINCK , B. A., S HEPLEY , M. G., J ACKSON , C. P. & D EGNAN , P. J. 2000. Hydrogeological characterisation of the onshore Quaternary sediments at Sellafield using the concept of domains. Quarterly Journal of Engineering Geology and Hydrogeology, 33, 301–323. M C M URTRIE , J. 1911. On a boring at Puriton, near Bridgwater, in search of Coal south of the Mendips. Proceedings of the Somerset Archaeology and Natural History, 57, 25–53. M ANNING , P. I. & W ILSON , H. E. 1975. The stratigraphy of the Larne Borehole, County Antrim. Bulletin of the Geological Survey of Great Britain, 50, 1 –50. M ILODOWSKI , A. E., S TRONG , G. E., W ILSON , K. S., A LLEN , D. J., H OLLOWAY , S. & B ATH , A. H. 1986. Diagenetic influences on the aquifer properties of the Sherwood Sandstone in the Wessex Basin: investigationof the geothermal potential of the UK. British Geological Survey, Keyworth, Nottingham. M ILODOWSKI , A. E., S TRONG , G. E., W ILSON , K. S., A LLEN , D. J., H OLLOWAY , S. & B ATH , A. H. 1987. Diagenetic influences on the aquifer properties of the Sherwood Sandstone in the East

Yorkshire-Lincolnshire Basin: investigation of the geothermal potential of the UK. British Geological Survey, Keyworth, Nottingham. M ILODOWSKI , A. E., S TRONG , G. E., S HEPHERD , T. J., ET AL . 1999. Diagenesis of the Permo-Triassic rocks. In: P LANT , J. A., J ONES , D. G. & H ASLAM , H. W. (eds) The Cheshire Basin: Basin evolution, Fluid Movement and Mineral Resources in a Permo-Triassic Rift Setting. Nuffield Press (Abingdon) for British Geological Survey, Nottingham. M ITCHELL , G. H., P OCOCK , R. W. & T AYLOR , J. H. 1961. Geology of the Country around Droitwich, Abberley and Kidderminster. Memoir British Geological Survey, England and Wales, Sheet 182, HMSO, London. M ITCHELL , W. I. 2004. The Geology of Northern Ireland — Our natural foundation. Geological Survey of Northern Ireland, Belfast. M OLINDA , G. M. 1988. Investigations of methane occurrence and outbursts in the Cote Blanche domal salt mine, Louisiana. US Bureau of Mines, Report of Investigations, RI-9186, Washington, DC. N ELSON , E. B. (ed.) 1990. Well Cementing. Developments in Petroleum Science, 28, Schlumberger Educational Services, Elsevier, Amsterdam, Netherlands. N OTHOLT , A. J. G. & H IGHLEY , D. E. 1973. Mineral Dosier No. 7, Salt. Mineral Resources Consultative Committee, HMSO, London. NPL. 2007. Proposed gas storage, King Street, Cheshire. NPL Estates press release 2007. World Wide Web Address: http://www.nplestates.com/NPLGroup/ KingStreetEnergy/tabid/213/Default.aspx O LD , R. A., H AMBLIN , R. J. O., A MBROSE , K. & W ARRINGTON , G. 1991. Geology of the Country around Redditch. Memoir British Geological Survey, Sheet 183, HMSO, London. P ENN , I. E. 1981. Larne No. 2 Geological well completion report. Report of the Institute of Geological Sciences, 81/6. Keyworth. P OOLE , E. G. & W ILLIAMS , B. J. 1980. The Keuper saliferous beds of the Droitwich area. Institute of Geological Sciences, Report No. 80/2, HMSO, London. PORTLAND GAS . 2008. Portland Project Dorset: project description. World Wide Web Address: http://www. portland-gas.com/index.0php?option=com_content &task=view&id=22&Itemid=101 P UGH , W. J. 1960. Triassic Salt: discoveries in the Cheshire-Shropshire Basin. Nature, 187, 278– 9. R ATIGAN , J. L. 2005. Core logging, well logging, well testing, and laboratory core testing of Canatxx exploratory wells at Fleetwood, United Kingdom. PB Energy Storage Services Inc., Topical Report PB-0104, April 2005. R ATIGAN , J. L., M C C LAIN , A., N ICHOLS , J. & H OFFINE , S. 2002. Fugitive NGL at the Williams Midstream Natural Gas Liquids Conway Underground Storage Facility — Geologic and Hydrogeological Studies, Cavern Well Testing and NGL Recovery. Technical Paper, Solution Mining Research Institute Spring 2002 Meeting, 28 April — 1 May, Banff, Alberta, Canada. R EES , J. & W ILSON , A. A. 1998. The Geology of the Country around Stoke-on-Trent. British Geological

POTENTIAL OF ONSHORE UK SALT DEPOSITS Survey Memoir, England and Wales, Sheet 66, HMSO, London. RIGZONE . 2006. Statoil seeking to extend UK gas storage facility. Press report, Monday September 25, 2006. World Wide Web Address: http://www.rigzone. com/news/article.asp?a_id=36489 R OSE , W. C. C. & D UNHAM , K. C. 1977. Geology and Hematite Deposits of South Cumbria. British Geological Survey England and Wales, Economic Memoir for 1:50 000 geological sheet 58, and northern part of sheet 48, Sheet 66, HMSO, London. S HERLOCK , R. L. 1921. Volume XVII Rock salt and brine. Special Report on the Mineral Resources of Great Britain. Memoirs of the Geological Survey, HMSO, London. S MITH , D. B. 1989. The late Permian palaeogeography of north-east England. Proceedings of the Yorkshire Geological Society, 47, 285– 312. S MITH , D. B. 1994. Geology of the Country around Sunderland. Memoir British Geological Survey, England and Wales, Sheet 21, HMSO, London. S MITH , D. B. 1996. Deformation in late Permian Boulby Halite (EZ3Na) in Teesside, NE England. In: A LSOP , G. I., B LUNDELL , D. J. & D AVISON , I. (eds) Salt Tectonics. Geological Society, London, Special Publication, 100, 77–87. S MITH , N. J. P., E VANS , D. J. & A NDREWS , I. J. 2005. The Geology of Gas Storage in Offshore Salt Caverns. BGS Confidential Report to the Department of Trade and Industry, No. GC05/11. S ONNENFIELD , P. 1984. Brines and Evaporites. Academic Press, Orlando, California. S PIERS , C. J., U RAI , J. L., L ISTER , G. S., B OLAND , J. N. & Z WART , H. J. 1986. The influence and fluid-rock interaction on the rheology of salt rock. Nuclear Science and Technology, EUR 10399 EN. Office for Official Publications of the European Communities, Luxembourg. S PIERS , C. J., U RAI , J. L. & L ISTER , G. S. 1988. The effect of brine (inherent or added) on rheology and deformation mechanisms in salt rock. In: H ARDY , H. R. & L ANGER , M. (eds) The Mechanical Behaviour of Salt. Proceedings of the Second Conference. Trans Tech Publications, Claustahal-Zellerfield, 89– 102. S PIERS , C. J., P EACH , C. J., B RZESOWSKY , R. H., S CHUTJENS , P. M. T. M., L IEZENBERG , J. L. & Z WART , H. J. 1989. Long-term rheological and transport properties of dry and wet salt rocks. Nuclear Science and Technology, EUR 11848 EN. Office for Official Publications of the European Communities, Luxembourg. S PIERS , C. J., S CHUTJENS , P. M. T. M., B RZESOWSKY , R. H., P EACH , C. J., L IEZENBERG , J. L. & Z WART , H. J. 1990. Experimental determination of constitutive parameters governing creep of rock salt by pressure solution. In: K NIPE , R. J. & R UTTER , E. H. (eds) Deformation Mechanisms, Rheology and Tectonics. Geological Society, London, Special Publications, 45, 215–227. STAG ENERGY . 2006. Irish Sea offshore LNG import terminal and gas storage project will improve security of gas supply for the UK and Ireland. Press statement: 22 February 2006.

79

STAG ENERGY . 2007. Gateway Gas Storage Project Environmental Impact Consultation Document. A report prepared for Gateway Storage Co. Ltd. by Rudall Blanchard Associates Ltd. S TEVENSON , I. P. & M ITCHELL , G. H. 1955. Geology of the Country between Burton upon Trent, Rugeley and Uttoxeter. Memoirs of the Geological Survey of England and Wales, Sheet 140, HMSO, London. S TEWART , S. A., H ARVEY , M. J., O TTO , S. C. & W ESTON , P. J. 1996. Influence of salt on fault geometry: examples from the UK salt basins. In: A LSOP , G. I., B LUNDELL , D. J. & D AVISON , I. (eds) Salt Tectonics. Geological Society, London, Special Publication, 100, 175– 202. S TRONG , G. E. 1993. Aspects of the petrology and diagenesis of Sherwood Sandstone Group rocks, Preston, Lancashire, UK: a possible evaporitic cement precursor to secondary porosity? In: N ORTH , C. P. & P ROSSER , D. J. (eds) Characterisation of Fluvial and Aeolian Reservoirs. Geological Society, London, Special Publication, 73, 279– 289. S TRONG , G. E. & M ILODOWSKI , A. E. 1987. Aspects of the diagenesis of the Sherwood Sandstones of the Wessex Basin and their influence on reservoir characteristics. In: M ARSHALL , J. D. (ed.) Diagenesis of Sedimentary Sequences. Geological Society, London, Special Publications, 36, 325– 337. S TRONG , G. E., M ILODOWSKI , A. E., P EARCE , J. M., K EMP , S. J., P RIOR , S. V. & M ORTON , A. C. 1994. The petrology and diagenesis of Permo-Triassic rocks of the Sellafield area, Cumbria. Proceedings Yorkshire Geological Society, 50, 77– 89. T AYLOR , B. J., P RICE , R. H. & T ROTTER , F. M. 1963. Geology of the country around Stockport and Knutsford. Geological Survey of Great Britain Memoir, England and Wales, Sheet 98. T ERRINHA , P., R IBEIRO , A. & C OWARD , M. P. 1994. Mesoscopic structures in a salt wall. The Loule salt wall diapir, Algarve Basin, south Portugal. Salt Tectonics meeting, Geological Society, London, 14–15 September 1994/abstract. T HOMPSON , F. J. 1908. The rock-salt deposits at Preesall, Fleetwood, and the mining operations therein. Transactions of the Manchester Geological and Mining Society, 30, 105–116. T HOMPSON , F. J. 1927. An account of the salt deposits at Preesall — Fleetwood and their development. Unpublished ICI Report. UK GAS REPORT . 2005. UK gas storage and LNG capacity set to soar. UK Gas Report, Issue 290, 8 August 2005, 18– 24. World Wide Web Address: http://www.platts.com/content/Natural%20Gas/ highlights/2006/UKGasTracker.pdf U RAI , J. L., S PIERS , C. J., Z WART , H. J. & L ISTER , G. S. 1986. Weakening of rock salt by water during longterm creep. Nature, 324, 554 –557. U SSHER , W. A. E. 1911. Excursion to Dunball, Burlescombe, Ilminster, Chard, Ham Hill and Bradford Abbas. Proceedings of the Geologists’ Association, 22, 246 –263. VAN K EKEN , P. E., S PIERS , C. J., VAN DEN B ERG , A. P. & M UYZERT , E. J. 1993. The effective viscosity of rock salt: implementation of steady-state creep laws in

80

D. J. EVANS & S. HOLLOWAY

numerical models of salt diapirism. Tectonophysics, 225, 457– 476. W ARREN , J. K. 2006. Evaporites: Sediments, Resources and Hydrocarbons. Springer-Verlag, Berlin, 1035 pp. W ARRINGTON , G., A UDLEY -C HARLES , M. G., E LLIOTT , R. E., ET AL . 1980. A correlation of Triassic rocks in the British Isles. Special Report of the Geological Society, London, 13. W HITEHEAD , T. H., D IXON , E. E. L., P OCOCK , R. W., R OBERTSON , T. & C ANTRILL , T. C. 1927. The Country between Stafford and Market Drayton. Memoirs of the Geological Survey of England and Wales, Sheet 139, HMSO, London. W HITTAKER , A. 1971. The salt industry at Puriton, Somerset. Proceedings of the Somerset Archaeology and Natural History Society, 114, 96–99. W HITTAKER , A. 1972. The Somerset Salt Field. Nature 238, 265– 266. W HITTAKER , A. 1973. The Central Somerset Basin. Proceedings of the Ussher Society, 2, 585 –592. W HITTAKER , A. & G REEN , G. W. 1983. Geology of the Country around Weston-super-Mare. Memoir

Geological Survey Great Britain, England and Wales, Sheet 279 with parts of sheets 263 and 295, HMSO, London. W ILLS , L. J. 1970. The Triassic succession in the Central Midlands in its regional setting. Quarterly Journal of the Geological Society, London, 126, 225–283. W ILLS , L. J. 1976. The Trias of Worcestershire and Warwickshire. Institute of Geological Sciences, Report No. 76/2, HMSO, London. W ILSON , A. A. 1993. The Mercia Mudstone Group (Trias) of the Cheshire Basin. Proceedings of the Yorkshire Geological Society, 49, 171– 188. W ILSON , A. A. & E VANS , W. B. 1990. Geology of the country around Blackpool. British Geological Survey Memoir, England and Wales, Sheet 66, HMSO, London. WINGAS 2008. Saltfleet by natural gas storage facility, Lincolnshire: Planning support statement. Wingas storage UK Ltd report, October 2008. World Wide Web Address: http://www.e-lindsey.gov.uk/e-docs/ 155754_2.pdf

Environmental and safety monitoring of the natural gas underground storage at Stenlille, Denmark T. LAIER1* & H. ØBRO2 1

Geological Survey of Denmark and Greenland (GEUS), Øster Voldgade 10, DK-1350 Copenhagen, Denmark

2

Dansk Olie & Naturgas A/S, Agern Alle´ 24– 26, DK-2970 Hørsholm, Denmark *Corresponding author (e-mail: [email protected]) Abstract: The Stenlille natural gas underground storage is located 70 km SE of Copenhagen and has been in operation since 1989. For safety reasons and to protect the environment it is necessary to monitor the storage carefully. Natural gas is being stored in an anticlinal structure with an expected gas storage capacity of about 3 billion Nm3 (volume under ‘normal’ conditions) in the upper Triassic Gassum Sandstone Formation 1500–1600 m below the surface; it replaces saline formation water. So far, nineteen deep wells have been drilled on and around the structure. The 300 m thick clay sequence of the Lower Jurassic Fjerritslev Formation above the gas storage reservoir has acted as an efficient seal, since no sign of gas leakage has been observed in the monitoring well located in a sand stringer 15 m above the gas reservoir. Other monitoring wells have been located in order to check for possible lateral escape of natural gas. A baseline study on naturally occurring hydrocarbons performed before the natural gas storage came into operation indicated the presence of only trace amounts hydrocarbon gases in the subsurface of the Stenlille area. Results of analysis by the headspace and sorbed gas methods on drill-cuttings suggest that low-temperature thermal generation of hydrocarbon gases (d13C1: 247 to 242‰; d13C2: 234 to 230‰) has taken place in organic-rich marine shale below 1300 m. Low concentrations of dissolved methane (,0.5 mg/l) of bacterial origin (d13C1: 290 to 262‰) were found in shallow groundwater that is used for water supply in the Stenlille area. After the start of injection of natural gas in 1989 (C1:C2:C3 ¼ 91:6:2; d13C1: 247‰), no increase in methane concentration and no higher hydrocarbon gases were observed during the regular analysis of groundwater from 10 shallow wells located above the underground natural gas storage. However, a sudden increase in dissolved methane concentration from 0.02 to 27 mg/l was measured in a 130 m deep observation well after a minor gas leakage had been detected at a new deep drilling into the natural gas storage in 1995. Nonetheless, no increase in methane was observed in shallow groundwater at the same locality. Occasional higher concentrations of dissolved methane (up to 15 mg/l) were encountered in shallow observation wells in low permeability layers. Stable isotope analyses (d13C1: 269 to 252‰) and radiocarbon dating show that the gas does not originate from the underground gas storage because the methane was less than 300 years old, but it may have formed due to local microbial activity.

Following the decision in 1979 by the Danish parliament to use natural gas from the North Sea, a transmission network was established in 1981–1984. The primary network, including transmission plant and gas storage facilities, is operated by the state-owned Danish Oil and Gas company that is also in charge of buying natural gas from the producers. Transmission of natural gas began in 1984 and the throughput of gas has increased steadily since then, and is now c. 7.5 billion Nm3 (volume under ‘normal’ conditions) per year. In order to buffer the supply of gas to consumers, two natural gas underground storage facilities were established; one in salt caverns at Ll. Torup and the other in a deep aquifer near the town of Stenlille (Fig. 1). Storage of natural gas at the two

sites started in 1987 and 1989 respectively. Possible impacts on the environment as a result of the construction and operation of the natural underground storage were considered carefully, particularly for the populated Stenlille area. The local communities have been kept informed about the plans and operations of the natural gas underground storage since the start of investigations. To be able to detect any trace of natural gas that may have leaked from the underground storage, it is imperative to know the characteristics of the hydrocarbon gases that may have been present in that environment prior to the injection of natural gas into the underground storage. A baseline study on naturally occurring hydrocarbon gases in the subsurface of the Stenlille area was therefore performed. The study comprised

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 81–92. DOI: 10.1144/SP313.6 0305-8719/09/$15.00 # The Geological Society of London 2009.

82

T. LAIER & H. ØBRO

Fig. 1. The Danish natural gas transmission network.

analysis of hydrocarbon gases, including stableisotopes, in cuttings from a number of deep wells and analysis of dissolved gas in shallow groundwater. Stable isotope ratios of different constituents, e.g. methane and ethane, may give useful information about the origin of the hydrocarbon gas. Since isotope ratios change insignificantly during migration, they are potentially useful for detecting leakage of gas from an underground gas storage (Coleman 1987; Coleman et al. 1977). Unforeseen loss of gas from the underground storage is of economic as well as safety and environmental concern. However, for the local community the latter two aspects are more important. Results of the various monitoring activities are therefore reported regularly to the local authorities as well as the ministries of energy and environment. The actions taken for safety and environmental protection include pressure monitoring in specifically designed wells, which will give immediate information on major losses of gas. Minor leaks, which could develop over time, may escape detection by pressure monitoring and would require analysis of subsurface fluids. So as an extra precaution, shallow groundwater from a number of water wells and observation wells was analysed regularly during the whole period of operation of the gas underground storage near Stenlille. Stable isotope

analysis and radiocarbon dating was performed in the event of higher than normal concentrations of dissolved methane in order to deduce the origin of the gas. This paper presents an overview of the development of the Stenlille natural gas underground storage and the various activities carried out to detect any unforeseen loss of gas to the environment. The results of the baseline study on naturally occurring hydrocarbon gases are presented and the chemical and isotopic composition of the gases is discussed to determine their origin. A summary of the pressure monitoring results since the start of gas storage operation is given and the results of regular analyses of dissolved gas in groundwater is presented. The chemical and isotopic composition of natural gas injected through time is also presented and the possibility of detecting minor leakage of this gas over time is discussed using experience gathered from the regular analysis of shallow groundwater.

Stenlille natural gas underground storage: overview No commercial accumulations of oil and gas have so far been found in onshore Denmark, so the common practice of using depleted oil and gas fields for

SAFETY MONITORING AT STENLILLE, DENMARK

Fig. 2. Sketch of the natural gas underground storage at Stenlille. A Zechstein salt pillow is located c. 2800 m below the surface. Sand, Late Triassic Gassum Formation; claystone, Early Jurassic Fjerritslev Formation.

storage of natural gas was never an option. Furthermore, no salt domes that may be used for storage of gas in salt caverns exist in the subsurface of the eastern part of the country. Storage of natural gas

83

in an aquifer therefore appeared to be the most suitable option for the densely populated Copenhagen area. A potential storage site near the town of Stenlille, c. 70 km SW of Copenhagen, was identified using older seismic mapping (Fig. 1). Investigations began in 1979 with the drilling of the first deep well into the Stenlille structure. Coring and test-pumping of fluids from the Gassum Sandstone Formation confirmed its potential for natural gas storage. Six additional wells were drilled for further development of the underground storage that permitted more tests of the future reservoir sandstone as well the caprock above. The storage at Stenlille finally came into operation in 1989 with the large-scale injection of natural gas into the sandstone reservoir, 1500 m below surface (Fig. 2). The fairly long period of development was due to the other underground storage facility in salt caverns near Ll. Torup (Fig. 1) being developed and completed at the same time. Nineteen wells have been completed to date; 12 wells for injection and withdrawal of gas and seven observation wells for monitoring pressure variation in the aquifer around the gas reservoir and in the caprock above (Fig. 3). The Gassum Sandstone Formation is c. 140 m thick, but only the upper 40 m is used for storage. The upper 40 m of the sandstone formation is divided into 5 gas storage zones separated by thin shale beds (Fig. 4). It is estimated that a total of 3 billion Nm3

Fig. 3. Well locations above the Stenlille underground gas storage and the extent of the different gas zones. Gas zone 5 is served by gas wells ST-2, -8, -14, -16 and -18; the remainder of the gas wells serve gas zones 1 –3. Contour lines indicate depth in metres below sea level to the top Gassum Sandstone Formation.

84

T. LAIER & H. ØBRO

the natural gas in the Stenlille structure. A number of interbedded thin sand layers exist over the gas reservoir allowing for the efficiency of the caprock to be checked against vertical leakage of gas. The Gassum Sandstone Formation at Stenlille forms an anticlinal structure with a vertical closure of c. 35 m covering an area of 14 km2 (Fig. 3). The structure formed as a result of movements of Zechstein salt forming a pillow approximately 2800 m below surface (Fig. 2).

Monitoring methods

Fig. 4. Schematic cross section (SW–NE) of the Stenlille natural gas underground storage including numbering of the different gas zones. Well locations are indicated at the top of the figure.

of natural gas can be stored in the Gassum Sandstone Formation.

Geological setting The good reservoir quality of the Gassum Sandstone Formation was known from a number of onshore wells in Denmark. Mapping the extent of this sandstone sequence had been carried out prior to the Stenlille evaluations as part of a geothermal energy feasibility study (Bertelsen 1978). The large amount of geological information obtained from drilling and coring the Stenlille structure allowed for a detailed description of the depositional environment responsible for the formation of the Gassum Sandstone Formation (Hamberg & Nielsen 2000). The Gassum Sandstone Formation was formed in Rhaetian times when the Danish Basin was narrow and semi-enclosed. The Danish Basin itself is an intracratonic feature located in the eastern part of the North Sea rift system. It formed as a result of Late Palaeozoic rifting followed by Mesozoic thermal subsidence. The Gassum Formation consists of cyclically interbedded sharp-based shoreface sandstones and offshore marine mudstones interrupted locally by fluvio-estuarine and lagoonal deposits (Hamberg & Nielsen 2000). The Gassum Formation represents part of the general long-term second-order transgression of the Danish Basin, starting from continental to shallow marine deposits of the underlying Upper Triassic sediments, and ending in the overlying fully-marine claystones of the Lower Jurassic Fjerritslev Formation (Bertelsen 1978). The Fjerritslev Formation, 300 m thick, forms the caprock of the sandstone reservoir that holds

For the baseline study of naturally-occurring hydrocarbon gases in the subsurface, samples from the drilling of three observation wells, ST-4 to ST-6 (Fig. 3) were analysed (Laier 1989b). Analysis of shallow groundwater was also performed, as part of the baseline study and for monitoring after injection of natural gas had begun. Finally, the chemical and isotopic composition of the natural gas was analysed every second year.

Drill cuttings Unwashed cuttings were collected during drilling every 10 m and sealed in half litre cans after bactericide had been added. Headspace analysis with respect to hydrocarbon composition was performed by gas chromatography. Analysis of sorbed gases was also performed by treating cuttings with acid in an evacuated system (Faber & Stahl 1983). Cuttings were washed and iron filings, derived from the drilling operation, were carefully removed using a magnet before adding the acid (Laier 1989b) in order to prevent light hydrocarbons from being generated as a result of iron–acid interaction (Jeffrey & Kaplan 1988).

Groundwater A number of the wells used for dissolved gas analysis are private water wells, so considerable care was taken to ensure reproducibility with respect to pumping and sampling of the wells. Information concerning any changes that may affect the results of the dissolved gas analysis was also obtained from the landowner. Samples for analysis of dissolved hydrocarbon gases were collected in either 12 ml evacuated test tubes (Venojectw) or in 15 ml serum bottles sealed with Teflon-coated crimp caps. Test tubes were half-filled with water using a 10 ml syringe to transfer water from the water well, whereas serum bottles were completely filled with water. Before analysis, half the volume of water in the serum bottle was replaced with helium and the water and gas phases were allowed to equilibrate. Hydrocarbon analysis of the gas phase was performed

SAFETY MONITORING AT STENLILLE, DENMARK

by gas chromatography and the amount of dissolved gas was calculated knowing the partitioning of the gas component between the water phase and gas phase. Isotopic analysis required larger quantities of gas, so water for this purpose was sampled in double-valve cylinders with volumes of 100– 300 ml. Water was then transferred to an evacuated bottle allowing the major part of dissolved gas to move into the gas phase for further processing before isotopic analysis.

85

Stable isotopic and radiocarbon analysis Gas components were separated using a gas chromatograph with thermal conductivity detector (TCD) and oxidized over CuO at 900 8C. CO2 resulting from combustion was then isolated in glass ampoules that were sent to the University of Copenhagen for stable isotopic analysis. A few ampoules were sent to the AMS laboratory at the University of Aarhus for radiocarbon analysis (Laier et al. 1996).

Natural gas North Sea natural gas for chemical and isotopic analyses was tapped from a valve at the site of dewpoint measurements. Samples of gas were collected in 300 ml double-valve pressure bottles and stored at 5 bars. The composition of the hydrocarbon gas was determined annually or biannually by gas chromatography. A few samples of gas withdrawn from the underground storage were also analysed.

Hydrocarbon gases in the subsurface Headspace gas analyses were performed on 174 samples of cuttings collected from the drilling of 3 observation wells, ST-4 to ST-6 (Fig. 3). Hydrocarbon concentration and composition versus depth were fairly similar for all three wells (Laier 1989b), therefore only the results from the ST-5 well are shown (Fig. 5). Control analysis of sealed samples of drilling

Fig. 5. Hydrocarbon and isotope analyses of drill cuttings from the ST-5 monitoring well.

86

T. LAIER & H. ØBRO

fluid indicated that only a minor fraction of the hydrocarbon gas was being re-circulated with the fluid. An almost hundred-fold increase in methane concentration was observed when comparing samples from Cretaceous carbonates into the more organic-rich Jurassic rocks c. 1200 m below the surface. A marked increase in the iso- to normal butane ratio from ,1 to .2 was also observed across this transition (Laier 1989b), whereas the gas wetness, (C2 –C4) * 100/(C1 –C4)  20%, did not show a significant change (Fig. 5). Although the hydrocarbon gas content present in the rocks below 1200 m does not appear to be insignificant, the amount of dissolved hydrocarbon gas measured in the formation water recovered from pumping tests (1370 m and 1510 m below the surface) was very low, 30 mg/l or less (Laier 1989a). Dissolved gas consisted mostly of nitrogen and methane with only traces of higher hydrocarbon gases, wetness 1–4%. Stable isotope analysis of dissolved methane (d13C1: 265) from the Gassum Formation at another location (Laier 1989a), but approximately from the same depth, indicated that bacterial methane may be generated at these depths. The Gassum Formation temperature at Stenlille, c. 54 8C, does not preclude bacterial activity. Thus, it is possible that the dissolved hydrocarbon gas consisted of a mixture of bacterial and thermogenic gas.

Sorbed gases The sorbed gas content was analysed in 40 cutting samples from ST-5 and 12 samples from ST-4 and ST-6. The results showed a moderate sorbed methane content, 500– 800 ppb, in most rocks penetrated (Fig. 5). No particular trend in methane content with depth was found, but a marked increase in gas wetness from 25 to almost 50% was observed at approximately 1200 m depth (Fig. 5). Stable isotope analyses of sorbed methane and ethane (d13C1: 247 to 242‰; d13C2: 234 to 230‰) suggested that the hydrocarbon gases are thermogenic in origin. Low amounts of thermogenic gases may have been generated in the more organic-rich part of the Upper Triassic Gassum Formation and the Lower Jurassic Fjerritslev Formation. Even though vitrinite reflectance measurements, Ro ¼ 0.43–0.49% (Table 1), found organic matter in both formations to be thermally immature, thermogenic gases could still have formed (Rowe & Muehlenbachs 1999). The isotope ratios of methane and ethane changed systematically towards more negative values by 4 –5 parts per thousand though the upper part of the Fjerritslev Formation (1400–1300 m) (Fig. 5). The same shift in carbon stable isotopic ratio was observed in organic matter in this interval (Buchardt 1988, pers. comm.). For a given

Table 1. TOC and vitrinite reflectance measured on well cuttings from ST-5 well Depth (m b KB*) 1419.5 1552.5 1559.3

TOC (wt%)

Ro (%)

1.37 2.64 3.40

0.43 0.46 0.45

*m below Kelly Bushing TOC, total organic carbon Ro, vitrinite reflectance

thermal maturity, thermogenic gases reflect the isotope ratio of the organic matter from which they have been generated. Therefore, co-variation in isotope ratios of organic matter and thermogenic gases support the assumption that the gases were formed locally. The sorbed gas observed in the organically lean Cretaceous rocks are more likely to be due to migration of hydrocarbons from below, indicated by their insignificant variation in chemical and isotopic composition (Fig. 5).

Pre-storage methane in shallow groundwater Normally, analysis for methane in groundwater is only undertaken if methane is suspected to occur. This was not the case for the Stenlille area, so only four wells had been sampled for methane and only one of them showed the presence of methane. Consequently, in order to obtain a better overview of the methane content in groundwater, water samples were collected for methane analysis from all existing water wells in the area (Fig. 6). Twenty-one wells were sampled and methane in low concentrations (0.01–0.49 mg/l) was found in all of them except two (Table 2). No higher hydrocarbon gases were detected in any of the samples. Stable isotope analyses, performed on nine of the samples (d13C1: 290 to 262‰) (Table 2) showed that the methane was bacterial in origin. Bacterial methane derives most likely from the peat layers known to exist in the area. Most water wells in the area pump water from glacial melt-water sands underlying glacial till, 20– 30 m below the surface (Fig. 7). Before injection of natural gas began, observation well K1 was established adjacent to the first gas injection well (Fig. 6). K1 allowed water samples to be taken from two levels, at 36 m and 98 m in melt-water sand and Paleocene calcareous sand respectively (Fig. 7). Samples from both levels were collected and analysed (Table 2) before natural gas injection began. A second observation well, K2, was established in

SAFETY MONITORING AT STENLILLE, DENMARK

87

Fig. 6. Location map of the Stenlille area including water wells and monitoring wells for the shallow aquifers.

1993, when new gas injection wells were being drilled (Fig. 6).

Storage of natural gas Storage of natural gas, coming mainly from the Tyra gas field (Fig. 1) began in July 1989. The gas consisted mainly of methane (91%) with a little ethane (5.5%) and propane (2.0%). The isotope ratio of methane was 246.6 per mil. The composition of the natural gas being stored in the Stenlille structure was thus very different from that of the gas dissolved in shallow groundwater (Fig. 8). This makes it fairly easy to discriminate between the two types of gas and to identify even minor leaks from the deep gas storage reservoir at the near surface. The composition of the natural gas from the North Sea has changed a little over time; the relative methane content decreasing by only a few percent (Table 3). The change may have been due to the contributions of gas from additional platforms at a later stage (Fig. 1). Gas withdrawn from the gas storage reservoir had identical composition, within sampling and analytical error (Table 3), to the gas

injected. This is to be expected since no gas existed in the reservoir prior to injection. Leak detection from the Stenlille gas storage is thus much simpler than the situation where a depleted oil/gas reservoir is being used for gas storage (Buzek 1992). The amount of gas injected into and withdrawn from the storage is shown in Figure 9. The volume of gas stored currently is 1.1 billion Nm3, approximately one third of the maximum storage capacity at Stenlille. At the start, gas was injected only into the three upper gas zones (Fig. 4). However, after gas was detected in observation well ST-4 in 1994, the deeper gas zone 5 was also used for storage (Fig. 4), since then no gas has been detected in ST-4.

Methane in groundwater 1989– 2004 Regular analyses for dissolved hydrocarbon gases in shallow groundwater have been performed since 1990, monthly during the first year and quarterly thereafter. Only methane was found in all of the shallow groundwater samples taken and no sign of

88

T. LAIER & H. ØBRO

Table 2. Methane in groundwater prior to gas injection Well No. 46 257 265 292 298 323 389 399 400 407 428 478 482 494 497 502 507 518 520 529 558 K1 K1 K2

depth (m bs)

CH4 (mg/l)

14 36 31 49 50 29 14 32 25 32 26 52 28 36 32 19 27 33 18 31 21 36 98 26

0.010 0.035 0.020 0.030 0.045 ,0.005 ,0.005 0.090 0.095 0.045 0.005 0.260 0.490 0.095 0.470 0.030 0.190 0.180 0.030 0.230 0.075 0.400 0.020 0.045

d13CCH4 (‰)

268.3

282.5 285.7 283.0 290.0 261.8 272.7 277.7 285.6 264.5

most wells. Larger variations were seen in well no. 520 (Fig. 6), owned by Stenlille water works. Good sampling for methane analysis requires that the sampled water is pumped continuously during sampling. This was not always possible for the fully automated water works at Stenlille. When the water storage tank was nearly full, the pumps stopped and this situation occurred occasionally, particularly for well no. 520 which was the last well to be sampled. A fairly large variation in dissolved methane concentration was observed in the upper levels of the K1 observation well (0.4 to 3.7 mg/l) in the early 1990s, but not in the deeper levels, which showed very low values (0.02 mg/l) (Fig. 12). If the variation in methane concentration was due to a leakage around the nearby gas injection well (Fig. 6), the opposite might be expected, an increase in methane at the deeper level first. In order to deduce the origin of the gas, isotope analysis was performed regularly during the first year of sampling. The isotope ratio of the methane (d13C1: 262 to 252‰) together with the failure to detect higher hydrocarbon gas indicates that the gas was bacterial in origin (Fig. 8), although the source remains unknown.

Gas leakage, September 1995 leakage of gas from the underground storage was observed. The minor variations in dissolved methane concentration (Figs 10 & 11) were within generally accepted sampling and analytical error for

Further development of the natural gas underground storage required drilling of new injection/ withdrawal wells into the potential gas storage zones in the Stenlille structure. In August 1995, a minor gas leakage occurred in a new well, ST-14,

Fig. 7. Geological cross-section of the upper layers of the Stenlille natural gas underground storage. The locations of observation wells K1 and K2 are shown on Figure 6.

SAFETY MONITORING AT STENLILLE, DENMARK

Fig. 8. Classification diagram indicating the chemical and isotopic composition of natural gas from the Danish North Sea (open circle); dissolved gas in shallow aquifers (solid circle) and dissolved gas in glacial till (square). Triangles represent dissolved gas in the upper aquifer of the K1 observation well. The C1/C2 þ C3 ratio for the dissolved gases represents a minimum value as both ethane and propane were below detection limit for all dissolved gases. Solid circle in thermogenic box represents dissolved gas in lower aquifer of K1 after the gas leakage incident in 1995.

during gas injection. Later investigation revealed that the leak was due to gas seeping from the tubing into the water filled annulus between tubing and casing. Due to unfortunate circumstances, water in the

89

annulus was displaced by gas which then escaped through a small leak in the casing out into the cement and confining rocks, c. 780 m below surface. Having escaped, gas then migrated towards the surface where it was recorded on 13 September as tiny bubbles in the pit around the wellhead of the new well, ST-14 (Fig. 6). Chemical and isotopic composition of gas collected at the surface (Table 4) was similar to that of natural gas (Table 3), so there was little doubt that it had leaked from the well. The injection tubing was then plugged and filled with water, after which gas bubbling ceased within two weeks. The total volume of gas lost was estimated to be c. 5000 Nm3. The casing was repaired and the tubing replaced by a new one. The well was put back in service and no leakage of gas has been observed since. Leakage of natural gas to the environment might have been avoided if pressure monitoring of annulus had been carried out in ST-14 as is normally done for wells at Stenlille. However, gas injection had begun before all of the monitoring equipment had been installed into the new ST-14 well. Safety procedures concerning new wells in addition to the existing wells were improved after the gas leakage incident in order to prevent future leakage. Pressure testing of casing pipes and tubing had been carried out according to international (API) standards prior to installation in the ST-14 well.

Reservoir gas observed at the deep level in observation well K1 A week after the leak had been observed at the surface, a significant increase in dissolved gas concentration was measured at the deep level (98 m) in the K1 well, located 250 m from the ST-14 well. This gas had almost the same composition as the reservoir gas (Table 5, Fig. 8). No free gas was observed while taking a sample of water from this

Table 3. Natural gas composition Date 30-10-1989 11-05-1993 21-06-1993* 21-10-1994 29-09-1995 13-08-1997 15-07-1999 04-01-2001 16-12-2003* 02-09-2004

CH4 vol%

C2H6 vol%

C3H8 vol%

d13CCH4‰

91.4 91.5 91.8 91.5 91.5 91.5 88.7 89.3 88.8 89.5

5.48 5.08 5.32 5.40 5.77 5.40 6.93 6.96 6.92 6.90

1.96 1.69 1.80 1.82 1.60 1.80 3.08 2.68 3.02 2.67

246.6 247.5 247.5 247.3§ 247.0 247.2 246.5 246.6 246.5 246.5

*gas withdrawal § 13 d CC2H6 ¼ 231.9‰; d13CC3H8 ¼ 228.4‰

90

T. LAIER & H. ØBRO

Fig. 12. Methane concentration in groundwater from the K1 observation well. The sudden increase in dissolved methane was noted one week after the gas leakage at ST-14. Fig. 9. Gas in-place in the Stenlille natural gas underground storage since its start of operation in July 1989.

level, so it was concluded that all of the gas existed as dissolved gas at this time. However, for the gas to migrate from ST-14 to K1 within a week, a free gas phase must have existed at some point, although free gas was never observed in the samples from any of the wells. The concentration of dissolved gas has decreased since the gas leakage occurred and is presently below 2 mg/l, though its composition has changed little.

Young methane in low-permeable layers at K2

Fig. 10. Methane concentration in groundwater from Stenlille water wells.

After the gas leak had occurred, additional weekly sampling and analysis was performed at the K2 observation well site. Samples from the three shallow wells completed in the low permeability layer of glacial till (Fig. 7) were also analysed for dissolved hydrocarbon gases. This had not been done before the leakage incident, since the shallow wells were established mainly to monitor a chloride plume following a minor spill of salt. Due to slow recovery of the water-table, the samples from the three shallow wells represent water present in the wellbores at the time of sampling. It was therefore surprising to see that water from these wells actually contained methane (Fig. 13), up to 13 mg/l in the 14.5 m deep well (Table 5). No higher hydrocarbon gases were measured and stable carbon isotope analyses indicated that the gas was bacterial in origin (Fig. 8). In order to be certain that the gas was not reservoir gas that had been modified by extreme fractionation, radiocarbon analysis was also performed on dissolved methane from the three shallow wells. These analyses documented clearly that the gas did not come from the reservoir, since two of the samples contained post-nuclear bomb

Table 4. Seep gas at ST-14 well head

Fig. 11. Methane concentration in groundwater from private water wells.

Date

CH4 vol%

C2H6 vol%

C3H8 vol%

d13CCH4‰

13-09-1995

76.8

4.68

1.55

247.3

Sample contained c. 16% air

SAFETY MONITORING AT STENLILLE, DENMARK

91

Table 5. Dissolved gas in groundwater three weeks after gas leakage Well K1 K1 K2 K2 K2 K2

Depth (m bs) 36 98 4.5 9.0 14.5 26

CH4 (mg/l)

C2H6 (mg/l)

C3H8 (mg/l)

0.05 14.7 2.9 0.58 13.3 0.04

– 1.93 – – – –

– 0.87 – – – –

d13CCH4 (‰) n.a. 242.2 268.8 260.3 252.2 n.a.

D14C (PMC) n.a. n.a. 95 101.9 104.9 n.a.

Age (a)

280 ,50 ,50

“ 2 ”, below detection limit 0.001 mg/l; n.a., not analysed; PMC, percent modern carbon

Fig. 13. Methane concentration in groundwater from the K2 observation well. Note the different scales for methane in the sandy aquifer (25– 39 m) and the three shallow wells in the glacial till.

test material (Table 5). The source of the methane is still unknown, but may be due to bio-corrosion of iron filings (Daniels et al. 1987) produced during drilling of the wells. The fairly high concentration of dissolved methane may be due to the stagnant conditions in the glacial till, which may have enabled the accumulation of locally-generated, bacterial gas. The reason for the gradual decrease during the intense sampling period 1996–1998 may then be explained by replacement of water due to frequent pumping.

Pressure monitoring around the gas reservoir To ensure that any vertical gas migration from the storage reservoir will be detected as quickly as possible, gas saturation measurements across the caprock, via pulse neutron logging, is performed annually in selected wells. Pressure is also monitored continuously in the ST-3 caprock observation well (Fig. 3) in two sand stringers 15 m and 50 m above the storage reservoir. Migration of even small volumes of gas into these sand stringers will cause an increase in pressure that would be detected immediately. Finally, open-hole logging, including pressure profile measurements across the caprock, is performed when new wells are drilled. The logs

are designed to reveal the presence of gas in the caprock. Any pressure increase above initial pressure indicates gas migration. So far, none of these activities has revealed any vertical gas migration since the start of gas injection in 1989. The lateral extent of the stored gas is being monitored by observation wells placed in the aquifer surrounding the gas reservoir. The wells are placed within structural closure, so that gas approaching the edge of the structure will be detected before any gas is spilled from the structure. The observation wells have been filled with fresh water to allow pressure monitoring at surface, the pressure being a few bars at the wellhead. Observation wells have been placed north, south, NE and SW of the stored gas (Fig. 3). If gas reaches an observation well, water will be replaced by gas and the surface pressure will increase from a few bars to 130–140 bars. It is therefore easy to detect gas in the peripheral observation wells. Monitoring the surface pressure in the peripheral observation wells has not revealed any gas loss from the structure since the start of gas injection in 1989. However, in 1994 gas appeared in the observation well ST-4 towards the SW earlier than expected, so one extra storage zone (zone 5) was brought into service. Gas has so far not reached any other of the peripheral observation wells.

Conclusions Pressure monitoring around the natural gas underground storage reservoir indicates that no leakage of gas has occurred through the natural (geological) barriers. Regular analysis for dissolved hydrocarbons in shallow groundwater showed only the presence of bacterial methane that was also present prior to injection of natural gas. A minor leakage of gas occurred while injecting gas via a new well into the Stenlille structure. Only a deeper aquifer, not used for potable-water supply, was affected by this incident. Regular analysis of this aquifer showed that it is possible to detect leakage gas with great sensitivity, by combining hydrocarbon and stable isotope analyses.

92

T. LAIER & H. ØBRO

The composition of the natural gas injected for storage has changed very little over time. Gas withdrawn from the gas reservoir also has the same composition as gas being injected. Rarely, higher concentrations of presumably bacterial methane have been found in upper layers. Radiocarbon analyses documented that this methane was not reservoir gas that experienced extreme fractionation. The source of the methane is unknown, but may be related to drilling.

References B ERTELSEN , F. 1978. The Upper Triassic–Lower Jurassic Vinding and Gassum Formations of the Norweigian-Danish Basin. Danmarks Geologiske Undersøgelse Series B, 3. B UZEK , F. 1992. Carbon isotope study of gas migration in underground gas storage reservoirs, Czechoslovakia. Applied Geochemistry, 7, 471 –480. C OLEMAN , D. D. 1987. Gas identification by geochemical fingerprinting. Operating Section Proceedings, American Gas Association, 568–573. C OLEMAN , D. D., M ENTS , W. F., L IU , C. L. & K EOGH , R. A. 1977. Isotopic identification of leakage gas from underground-storage reservoirs — progress report. Illinois State Geological Survey, Illinois Petroleum 111. D ANIELS , L., B ELAY , N., R AJAGOPAL , B. S. & W EIMER , P. J. 1987. Bacterial methanogenesis and growth from

CO2 with elemental iron as the sole source of electrons. Science, 237, 509– 511. F ABER , E. & S TAHL , W. J. 1983. Analytical procedure and results of an isotope geochemical surface survey in an area of the British North Sea. In: B ROOKS , J. (ed.) Petroleum Geochemistry and Exploration of Europe. Blackwell Scientific Publications, 51–63. H AMBERG , L. & N IELSEN , L. H. 2000. Shingled, sharpbased shoreface sandstones: depositional response to stepwise forced regression in a shallow basin, Upper Triassic Gassum Formation, Denmark. In: H UNT , D. & G AWTHORPE , R. L. (eds) Sedimentary Responses to Forced Regressions. Geological Society, London, Special Publications, 172, 69–89. J EFFREY , A. W. A. & K APLAN , I. R. 1988. Hydrocarbons and inorganic gases in the Gravberg-1 well, Siljan Ring, Sweden. Chemical Geology, 71, 237 –255. L AIER , T. 1989a. Mapping of low enthalpy brines in Denmark for geothermal exploitation. In: M ILES , D. L. (ed.) Water– Rock Interaction WRI-6. Balkema, Rotterdam, 409–412. L AIER , T. 1989b. Stenlille Gas Storage — Study of naturally occurring hydrocarbon gases before injection. Geological Survey of Denmark. DGU report No. 15. L AIER , T., K UIJPERS , A., D ENNEGA˚ RD , B. & H EIER -N IELSEN , S. 1996. Origin of shallow gas in Skagerrak and Kattegat — Evidence from stable isotopic analyses and radiocarbon dating. Norges Geologiske Undersøgning Bulletin, 430, 119– 125. R OWE , D. & M UEHLENBACHS , A. 1999. Lowtemperature thermal generation of hydrocarbon gases in shallow shales. Nature, 398, 61– 63.

Design of salt caverns for the storage of natural gas, crude oil and compressed air: Geomechanical aspects of construction, operation and abandonment K.-H. LUX Clausthal University of Technology, Erzstraße 20, 38678 Clausthal-Zellerfeld, Germany Corresponding author (e-mail: [email protected]) Abstract: Salt cavities for the storage of natural gas in bedded or domal salt structures are an important element of current and future energy supply management. In Germany, the mechanical design of salt cavities has a history of more than 35 years. This paper gives a personal view of current salt cavity design. It discusses the geomechanical characteristics of storage cavities and principle safety demands for their design as well as recent design concepts and methods for providing geotechnical proof of safety with specialized criteria, limit values and safety margins. In view of the uncertainties inherent in the design of geotechnical mechanical structures, monitoring of excavation and operation are essential parts of underground geotechnical constructions. A new monitoring software code is presented that will help both to document that past and to plan future cavern operation. Cavern abandonment is an object of current research, especially the basic understanding of mechanisms acting or becoming active at elevated fluid pressures (gas or brine) at the level of primary (lithostatic) rock mass pressures. This paper presents some basic knowledge and a computer code for analysing the long-term behaviour of sealed liquid-filled salt cavities with simulation of pressure build-up, infiltration and following seepage flow.

Gas storage in salt cavities is a tried and tested technology in Germany with a history of more than 35 years. The first cavity used at the end of the 1960s, had a geometric volume of about 35 000 m3, more recently cavities with a geometric storage volume of more than 106 m3 have been designed and constructed. Figure 1 illustrates the various locations of storage cavity facilities and the number of cavities in Germany (Sedlacek 2005). From a geological point of view the storage cavern sites include domal salt structures as well as bedded salt structures. In total at 19 different sites gas storage facilities are currently operational within about 169 cavities, the largest facility having about 53 storage cavities. Individual sites may have several different geological structures, and cover a wide range of depth; from more compact cavities in bedded salt at a shallow depth range of 500–650 m, and going down to relatively deep cavities in the range 1400– 1850 m with very high and slim cavities in domal salt structures. Figure 2 illustrates cavern geometries and cavern locations within the respective site-specific rock mass (Lux 1984). The progressive improvement in cavern design is based on a continuously developing understanding of the mechanical behaviour of saliferous rocks as well as technical constructions in salt rock mass, secondly in improved lab testing facilities, modelling and numerical simulation techniques, and thirdly in excellent practical experiences over the years with

this type of geotechnical structure. The latter cover geotechnical aspects such as structural stability, tightness and surface protection as well as economic aspects. The operation pattern of gas storage cavities is changing from seasonal storage (one storage cycle per year) to a frequent turnover and high deliverability (several storage cycles per year) to satisfy the need for improved economic benefits and future gas markets. The demand for sufficient load-bearing capacity of the rock mass at the same level of geotechnical safety remains. Therefore the requirement for reliable cavern design as well as for current monitoring of rock mass stress during cavern operation is also increasing, with the complexity of cavern operation modes. Finally, the operational lives of the cavities will end one day. Therefore consideration needs to be given to long term surface and environmental protection. This aim can only be met with a good understanding of abandoned and sealed salt cavities either filled with solids or liquids. Despite the technical feasibility of refilling the cavities with appropriate (inorganic) solid waste material prior to abandonment, the preferred method for cavern abandonment in Germany today is to fill the cavity with water and create a permanent borehole sealing. The long-term behaviour of sealed liquid-filled cavities is an important area of geotechnical safety analysis and safety assessment.

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 93–128. DOI: 10.1144/SP313.7 0305-8719/09/$15.00 # The Geological Society of London 2009.

94

K.-H. LUX

Fig. 1. Gas and oil storage facilities in Germany (from Sedlacek 2007).

This paper gives a comprehensive overview of salt cavern design with special consideration of natural gas storage. It summarizes the author’s opinion and experience of the subject gained over the last 25– 30 years.

Geomechanical characteristics of salt cavities In comparison to other geotechnical underground constructions, cavities in salt rock mass show

GEOMECHANICS OF SALT CAVERN DESIGN

95

Fig. 2. Variety of salt cavities used for storage of natural gas and crude oil (after Lux 1984).

some specific characteristics with influence on cavern design and safety assessment. Figure 3 illustrates the differing geotechnical characteristics of a bedded salt structure including some cavities (Fig. 3a) and a representative cavern field within a domal salt structure (Fig. 3b). The following geotechnical characteristics are clear: (1) The solution mining method is used to excavate the cavities in the salt rock mass which are not directly accessible to humans;

(2) The only direct access to site-specific rock mass is via cavern boreholes. Thus rock mass structure in general and particularly next to individual boreholes can only be explored indirectly using geophysical methods. This leaves some uncertainties. From these more general characteristics the following geomechanical conclusions can be drawn: (1) The predetermined geometrical dimensions of storage caverns (shape, diameter, shape and

96

K.-H. LUX

Fig. 3. Sketch 3D cut away views of salt cavern storage sites (a) example of a bedded salt structure including gas storage cavities (b) example of a domal salt structure including gas storage cavities (after KBB 2005).

GEOMECHANICS OF SALT CAVERN DESIGN

roof configuration) must be reliably achieved during solution mining (monitored via sonar survey); (2) Load-bearing elements are the rock salt mass as well as non-halite rock salt formations (especially in bedded salt) and the cavern fluid via its pressure; additional artificial support of the rock mass is not feasible (mechanical safety without artificial support); (3) Sufficient and reliable tightness of a cavity must be realized and maintained during cavern operation, by the salt rock mass itself perhaps together with appropriate non-halite rock formations; artificial sealing of the cavern contour is not possible; (4) Reliable monitoring of the load-bearing behaviour of storage cavities is essential, but possible only via indirect measurements of contour failure and contour (rock mass) deformation (sonar or laser survey), apart from deformation measurements at the surface. The caverns are excavated by salt solution mining processes (Fig. 4a), which may permit the initial operation of gas storage. The rather unfavourable boundary conditions with respect to design and monitoring, specific to salt cavern construction and operation mentioned above, show that the mechanical design of cavities must have met relatively high standards prior to excavation and operation to avoid later disadvantages in safety or economics (Fig. 4b).

General demands for cavity design Following the basic rules of German mining authorities, cavities for the storage of crude oil and natural gas have to meet some general demands, summarized in general as geotechnical safety: These demands are: sufficient static stability (on a local and global scale); reliable tightness (rock mass as well as drillhole); acceptable surface subsidence; and environmentally safe abandonment. Furthermore, from the operator’s point of view, the cavities should have a maximal gas storage capacity (minimum cushion gas, maximum working gas), high deliverability and low convergence i.e. effective and long-term usability. These general demands have to be fulfilled during the life-time of any cavity i.e. construction, operation and post-operation phases. In principle, the fulfilment of these demands must be documented during the design and licence process, in advance of construction. Geomechanical models and numerical simulations used to predict the rock mass behaviour (taking into account all the loads that are to be expected during construction and operation as well as after abandonment) are the instruments required for the documentation of geotechnical safety and

97

economic usability. For safety reasons, appropriate design concepts have to be developed that take into account the site-specific rock mass structure and rock mass properties as well as the structural behaviour of the load-bearing elements, regarding possible failure mechanisms and including in-situ experiences. These design concepts differ in their detailed criteria, limit values and safety margins depending on the experience of the designer. The general demands and the type of issues that have to be addressed by geomechanical modelling and numerical simulations as part of the design process are summarized in Figure 5.

Design concepts and geotechnical proof of safety including safety criteria Although dealing with the same geotechnical subject, both general and more detailed procedures for cavity design and proof of safety vary throughout the world. In the past decades several different design concepts have been developed and practised (Lux 1984). Furthermore, a new design concept has been developed integrating current scientific knowledge and practical experiences. This design, DC-Lux-CM, is based on the framework of continuum mechanics. At the end of the 1990s a more sophisticated concept (DC-Lux-KM/CDM) was developed, taking into account several aspects of continuum damage mechanics, especially the damage and healing behaviour of rock salt (Lux et al. 1999a, b, 2002a, b). The procedure and the main elements of a generalized design concept are illustrated in Figure 6. In the case of bedded salt layers with fault zones in non-halite strata perhaps located next to cavities, the mechanical and hydraulic impacts of these fault zones have to be integrated into this scheme. From this point of view, fault zones will be handled like an individual rock mass layer with specific material properties and parameters. The load-bearing behaviour of this additional geotectonic element has to be taken into account both with respect to its influence on the cavern behaviour as well as with the mechanical reaction of this geotectonic element on cavern operation. Therefore, additional assumptions and criteria are necessary to describe the mechanical behaviour of fault zones on the one hand and to evaluate the mechanical behaviour of fault zones in their reaction on the cavern operation on the other. Figure 7 illustrates a cavern field with a cavity next to a fault zone in a bedded salt structure. An adequate geomechanical model should be 3-dimensional, and in certain cases can be idealized as a rotationsymmetric configuration, whereas the fault zone

98

K.-H. LUX

Fig. 4. Salt storage cavern construction and cavern characteristics. (a) solution mining techniques (after KBB 2005); (b) characteristics of solution mined salt cavities with respect to geotechnical design.

GEOMECHANICS OF SALT CAVERN DESIGN

99

Fig. 5. General demands for geotechnical safety and resultant derived design related questions (Lux 1984).

Fig. 6. Significant elements in the design of salt cavities (in green: elements with special consideration in this paper).

100

K.-H. LUX

Fig. 7. Cross-section illustrating salt cavity locations in the region of a fault zone (above) and the geomechanical model developed (below) based upon Lux et al. (2004).

will have a planar geometry with an estimated thickness. Considering the excavation and operation phases of storage cavities, the proof of geotechnical safety for these constructions must have two components: proof of static stability and proof of tightness. The proof of static stability has two (or three) main

aspects, especially for rocks with significant creep behaviour: (1) limitation of stress intensity in dependence of the foreseen minimum cavern pressure and operation time to exclude macroscopic fractures and spalling at the cavern contour;

GEOMECHANICS OF SALT CAVERN DESIGN

(2) limitation of creep strain to exclude creep rupture; and perhaps (3) limitation of the annual mean cavern convergence. Furthermore, due to cyclic loading over several decades, a progressive micro- to macro-fracturing of the rock mass in the immediate vicinity of the cavern has to be excluded or limited to an acceptable degree. Therefore, the proof of static stability for the load-bearing elements is as follows in general terms: (a) Rock formations with significant creep behaviour no tensile fracture cal sz  0 MPa ¼ bz no shear fracture—rock matrix

(1)

cal h (t, x, min s)  adm h(min s, t) ¼ cal s v(min s)=b(min s, t) (2) no creep rupture cal D1vv (min s, x, Dt)  adm D1vv (min s, Dt) ¼ 1 f (min s)=j1 (3) (b) Rock formations without creep behaviour no tensile fracture cal sz  0 MPa ¼ bz no shear fracture—rock matrix cal s1 (min s)  adm s1 (min s) ¼ b1 (min s)=js no shear fracture—rock joints cal t  adm t ¼ tf =jt:

(4)

(5) (6)

Because geologically impermeable rock mass formations are needed to create the necessary gas tightness of the cavities, proof of tightness has to show that this primary impermeability is retained even under the impacts of cavern operation with the related secondary stress and strain fields. Therefore, the proof of tightness is transferred into a proof of guaranteeing the adherence of certain stress and perhaps strain conditions in the rock mass next to the cavity during the different operation phases. Thus the proof of tightness is as follows in general terms: cal sz  0 MPa (no tensile fractures) as well as

(7)

cal s2,3 . max pi þ Ds (no fluid infiltration) (8a) in a rock mass area Masz with pre-given necessary extension around the cavity (8b) Furthermore, depending on the site-specific situation, induced surface subsidence due to cavern

101

operation or even cavern creation may be subject to acceptability limits. In this case the expected surface subsidence has to be quantified, by calculation or direct field measurement, for the period of operation or in total. From this aspect it follows: max cal u_ s  adm u_ in mm=a max cal us  adm us in m

(9) (10)

These criteria are neither complete nor valid in any arbitrary underground situation. It is necessary to modify or amend them with regards to the specific geotechnical characteristics of the site under consideration and to determine the safety margins that are implemented in each of these criteria.

Laboratory tests for determination of material properties and material parameters There is a close interdependence between material properties, constitutive laws and material parameters and consequently between these items and the planning of a laboratory test programme. Taking into consideration that the general material behaviour of salt rocks is well-known, the following procedure seems to be appropriate: (1) Assumption that the relevant rock facies will follow the general experience, i.e. † rocks such as rock salt and carnallite or sylvinite will show elastic-plastic/ viscous (ductile) behaviour; † rocks such as anhydrite, chalk or sandstone show elastic-pseudoplastic (brittle) behaviour; † rocks such as claystone, shale or marlstone (argillaceous rocks) will show a behaviour between these two modes, depending on the degree of compaction during diagenesis. (2) Selection and determination of appropriate constitutive laws for the rock mass formations relevant in the specific case from a rock mechanical viewpoint. (3) Design of a lab test program with respect to the relevant rock formations and the expected material behaviour, the available core sample material, the available test equipment and the rock parameters necessary in connection with proposed material laws. (4) Development of lab tests, evaluation of lab test data and determination of material parameters. (5) Transferring lab test parameters onto the rock mass by taking into account rock-type related scale effects, e.g. joint systems and their mechanical or hydraulical impacts as well as bond properties.

102

K.-H. LUX

According to results, additional tests may have to be undertaken to clarify or reduce existing uncertainties. Typical results of short-term and long-term (creep) tests for rock types relevant to salt rock mass and neighbouring rock formations have been published (e.g. Lux & Rokahr 1980, 1982; Lux & Heusermann 1983; Lux 1984). During the last ten years, elements of continuum damage mechanics (CDM) have been introduced in salt mechanics which were previously based on continuum mechanics theory (CM), especially: the onset of damage and dilatancy strength (limit), damage propagation as well as dilatation and healing of damage. The basic assumption within the framework of CM was that rock salt behaved in a ductile manner without micro-fissuring and related volume changes during visco–plastic flow (creep) deformation and failure were based on macroscopic creep rupture (tensile fracs or shear fracs depending on stress conditions). However, views on the mechanical behaviour of rock salt have changed. Upon reaching a certain level of stress intensity and loading rate, mainly intercrystalline micro-fractures open; with increasing stress or deformation the micro-fractures grow and connect as well. From a macroscopic view the material shows volume increase, a process known as dilation. The stressrelated onset of dilation is called the dilation limit or dilation strength. As a consequence of the microfracture development several material properties of rock salt change. Figure 8 illustrates the change in physical parameters and their qualitative characteristic as stress increases. The development of fractures in the walls of underground cavities e.g. drifts or workings (with negligible confining stress) and solution mined cavities (with substantial confining stress) in a rock salt mass is illustrated in Figure 9. At first, isolated micro-fractures develop, then more and more connected meso-fractures and finally macrofractures and joints that can lead to spalling and rock fall. This rupture process will not only lead to a reduction of mechanical properties, but also an increase in hydraulic properties, i.e. an increase in porosity and permeability. Therefore, advanced constitutive models include models for dilation limit and damage induced deformations and also a poroperm-model, which transduces the induced porosity in so-called secondary permeability (Hou 2002; Lux et al. 2003). Furthermore, boundary conditions and the kinetics of healing processes for recreation of the initial (undamaged) rock mass quality are analysed in lab tests very intensively (Du¨sterloh & Lux 2003). The following section presents a short overview on laboratory testing and characteristic material properties.

Fig. 8. Short-term behaviour of salt rock and the change of physical properties related to increasing stress (after Schulze 2001).

The rock mechanics laboratory and test equipment at the Clausthal University of Technology has been designed and constructed by the laboratory staff (Fig. 10). Both standard tests and highly specialized tests with modification of existing test devices can be performed (see: www.ifa.tu-clausthal. de/deponie/). Typically, during a laboratory test program, material behaviour is first tested under short-term loading conditions. Numerous lab tests (e.g. Fig. 11a) show rock salt to have a markedly plastic– ductile material behaviour when compared to other rocks like anhydrite or marlstone, which in contrast show a more or less linear elastic – brittle behaviour. While the elastic material parameters Young’s modulus E and Poisson’s ratio v have to be determined from loading –unloading cycles, the failure strength b can be derived from the initial loading curve corresponding to the maximum stress at failure. In the general case of plastic– ductile rock salt, no mechanically important joint systems exist in the rock mass. Consequently, the lab test results must not be modified to take into account their negative influence on strength and deformability as well as permeability. In contrast, for elastic –brittle rock types this effect has to be investigated very carefully if there are any joints or regular joint systems initially existing in the rock mass. If this is the case, the rock strength derived from rock samples would have to be

GEOMECHANICS OF SALT CAVERN DESIGN

103

Fig. 9. Development of micro- and macrofractures in rock salt (micro- and macroscopic scale).

reduced significantly to obtain the pertinent rock mass strength. Samples of rock salt tested in triaxial tests (Fig. 11b) demonstrate the impressive deformation that can be sustained until failure, both in triaxial compression tests (TC) and triaxial extension tests (TE). Typical failure limits and their bandwidths for rock salt are shown in a pffiffiffi bD =(I1 = 3) diagram (Fig. 11b). Thereafter, based on the data for its short-term material properties, the long-term behaviour of the

material under investigation has to be analysed. It is well known that rock salt shows pronounced timedependent stress-deformation behaviour. Figure 12 shows representative results of creep tests with rock salt. In detail, a creep curve with transient, stationary and accelerated creep strains until creep failure and the related creep strain rate curve are presented. The loading stress with sv ¼ 25 MPa (min s ¼ 1 MPa) seems to be relatively high, compared to the related short-term strength.

Fig. 10. Rock mechanics laboratory at the Clausthal University of Technology.

104

K.-H. LUX

Fig. 11. Compressive and extensional testing and behaviour of rock salt and other rock samples. (a) Stress-strain behaviour of different rock types; (b) rock salt samples tested under triaxial compressive and extension stress conditions (left) and failure limits as well as dilation limit of rock salt (right).

Therefore, the time until creep failure occurs, of about two days is relatively short. Nevertheless, the wellknown three distinct creep phases are clearly visible, especially with a view at the creep strain rate curve. This curve also shows the onset of the creep-rupture phase, recognizable by acceleration of the creep rate. In this case, the creep-rupture strain amounts to about 1f  216% (compressive strain). Constitutive laws in the framework of continuum mechanics are based on elastic deformation as well as transient and stationary creep deformation with the assumption of constant volume of rock (no damage, no dilatancy). Creep rupture is treated by

the use of a macroscopic-based creep rupture criterion, related to creep rupture strain depending on rock salt facies, minimal stress and temperature. In this context, Figure 12 (right) shows a diagram with stationary creep rates of different rock salt facies versus effective stress sv . The stationary creep rate is dependent on rock salt facies as well as a highly non-linear dependency on effective stress sv . Although not discussed here, the material behaviour of rock salt is also influenced by the rock mass temperature (Lux 1984). In contrast to the classic constitutive laws, advanced constitutive laws try to model the time

GEOMECHANICS OF SALT CAVERN DESIGN

105

Fig. 12. Results of creep tests (left) and related stationary creep strains versus effective stress (right).

and/or strain dependent weakening process of the material from the onset of damage up to macroscopic failure on a microscopic scale i.e. introduction of a so-called damage parameter D, which is a basic element of continuum damage mechanics. In addition to the well-known material parameters related to continuum mechanics this parameter describes the development and the actual intensity of micro-fissures in the rock. Representative results of triaxial short-term testing of volume and ultrasonic wave velocity are illustrated in Figure 13a. The change from compressive to extensional volume change as well as the change from increasing to decreasing ultrasonic wave velocity shows the onset of weakening, i.e. micro-fissuring of the polycrystal fabric. The creation of the fissures and pore spaces can be seen in Figure 13b. The fissured areas are related to regions with relatively high stress compared to failure strength, which is dependent on the actual minimal main stress and the respective short-term strength. Figure 14 illustrates a triaxial creep test with transient, stationary and tertiary (accelerated) creep strains. The onset of tertiary creep is commonly identified with the onset of creep failure, based upon increasing damage. Laboratory tests show that the time from the onset of tertiary creep due to microscopic fissuring until final macroscopic creep failure may be weeks or even months, depending on the stress state. Figure 14 also shows some details of creep test result preparation for determination of material parameters. The first step in this procedure consists of quantitatively identifying

the different portions on the creep deformation i.e. transient (¼ decreasing), stationary (¼ constant) and tertiary (¼ accelerated) creep rates. Experience indicates that a clear identification of different creep rate modes (portions) is only possible in a multi-step creep test with stepwise increased deviatoric stress. In this case the evaluation of the first two creep phases at load levels with stresses below dilatancy limit gives the parameter for transient and stationary creep of the sample, without any damage induced creep rate fractions. Starting from this basis and calculating the undamaged creep deformation for the stress conditions of third load steps with deviatoric stress above dilatancy limit will permit characterization and quantification of the difference between calculated and measured creep deformation in this third load step as tertiary creep deformation. The increase of stationary, damageless creep deformation at the end of the load step shortly before creep rupture is a consequence of weakening of rock salt, i.e. increasing deviatoric stress in the increasingly damaged sample at constant stress calculated for the undamaged sample size, which finally gives the load intensity. Together with the online recorded data for volume and ultrasonic wave velocity change, the necessary data for the constitutive models are available. The latest results of laboratory investigations on the behaviour of rock salt subjected to conditions that will lead to healing effects in the damaged crystalline fabric, i.e. recreation of rock salt mechanical quality following damage, are illustrated in Figure 15. The diagram shows the mechanical

106

K.-H. LUX

Fig. 13. Observation of weakening of a rock salt sample via volume change measurement and ultrasonic wave velocity measurement (a) and weakened salt rock fabric (b).

Fig. 14. Triaxial creep test up to creep rupture with identification of damage induced creep strain (third load step of a multi-stage creep test) after Du¨sterloh & Lux (2003).

GEOMECHANICS OF SALT CAVERN DESIGN

107

Fig. 15. Damage and healing of rock salt (Du¨sterloh & Lux 2003).

behaviour of a rock salt sample during deviatoric loading (¼ damage) and unloading (healing) phase. The confining stress amounts to s23 ¼ 30 MPa every time, whereas the axial stress starts at s1 ¼ 29.5 MPa, decreases down to s1 ¼ 1.0 MPa for a short time, and then increases up to s1 ¼ 25 MPa for some weeks. The mechanical condition of the rock salt sample has been observed via ultrasonic wave velocity measurements. The diagram shows the time-dependent development of ultrasonic wave velocity vP , relating to the initial value vP0 determined before testing. The ultrasonic wave velocity is a measure of the intensity of damage within the rock salt sample. Decreasing values of vP =vP0  1 mean increasing damage and vice versa. Figure 15 shows that during the loading phase, with increasing deviatoric stress, vP =vP0 is decreasing and damage is increasing, as expected. At constant maximal deviatoric stress, damage is also increasing, but at slower rate. The following reloading phase is characterized by two different parts, first an increase of wave velocity ratio at relatively high rate along with deviatoric stress decrease and second a further increase of wave velocity ratio towards vP =vP0 ¼ 1:0, but with definitely lower rate than before. Transforming these macroscopic observations to micro-mechanical mechanisms, it is supposed, that at first, micro-fracture closure takes place and secondly a kind of fracture healing may be occurring. An alternative interpretation is that this second part is related to further, but now time-dependent micro-fracture closure and that

there may be a third part sometime later with pressure influenced geochemical healing of micro-fissures. In this context the first two mechanisms may cause reduction of permeability (no or only few interconnected micro-fissures) and some upgrade of mechanical quality, whereas the third mechanism will lead to initial mechanical quality of rock salt. These phenomena will have to be investigated in the future. Thus, it is possible with today’s equipment to identify elastic, plastic, viscous and clastic material behaviour as well as corresponding material properties. Furthermore, damage-induced porosity and corresponding permeability can be identified. Finally it is possible to observe the process of healing of damage and to prepare the necessary data for modelling using laboratory tests.

Rock mass model, constitutive laws and rock mass parameters Rock mass model The rock mass structure explored, described and documented by geologists has to be transformed to a more or less simplified, but adequately idealized rock mass model from a geotechnical safety point of view (Fig. 7). The necessary simplifications therefore should be conservative with respect to the different criteria for proof of safety, but not too conservative with respect to economics. Also the idealization process should not eliminate

108

K.-H. LUX

mechanically important effects. Besides the idealization of the rock mass structure itself (by identifying and taking into account rock layers that are different in their properties from a mechanical point of view), the rock mass model contains approaches for the primary stress state (isotropic, anisotropic) and for the rock mass temperature. Finally the material behaviour of the different rock layers is modelled (as described below) via constitutive models. An example of the idealization process is shown in Figure 7. In this case with special respect to the fault zone, whose mechanical properties are more or less unknown, a 3D-rock mass model was chosen. This model gives an adequate approximation for modelling the load-bearing behaviour of the cavity next to the fault zone and the fault zone itself.

Constitutive models (material models) In the case of bedded salt structures there are layers both with and without creep behaviour existing in the relevant rock mass area. The following models can be used for modelling the mechanical behaviour: Rock salt layer: constitutive model (Hou/Lux). Lab test data form the basis for the development of advanced constitutive models, which are able to simulate numerically the time dependent weakening process. On the basis of these tests it is also possible to differentiate between undamaged and damaged rock mass areas, to estimate the creep rupture time in relationship to the existing stress state and to determine the intensity of damage with the help of

the damage parameter D. This damage parameter D is defined as follows:

D¼

Ad A  A0 ¼ A A

(11)

where Ad , actually damaged part of cross-sectional area in m2; A0 , actually non-damaged part of cross-sectional area in m2; A, actual cross-sectional area in m2. One import task is the classification of certain degrees of damage of the rock mass to calculated values of the damage parameter D, which amounts to 0  D  1. Recommendations have been made in the literature e.g. Lux et al. (1998) and Xie (2002). Figure 16 provides a schematic overview on the essential submodels of the Hou/Lux constitutive law. This has been developed on the basis of the constitutive law Lubby2, and now includes the above-mentioned elements of CDM, together with a new model for simulation of damage reduction, i.e. fracture closure and fracture healing. Four main equations describing the material behaviour of rock salt are related to creep strain rate, dilation strength, time dependent strength and damage evolution. These are as follows: _ eij þ 1_ vp _ dij þ 1_ hij 1_ ij ¼ 1_ eij þ 1_ ie ij þ 1 ij ¼ 1

bDil ¼ hD (s3 )  b(s3 , u)

Fig. 16. Overview of the constitutive Hou/Lux law with sub-models and their combination.

(12) (13)

GEOMECHANICS OF SALT CAVERN DESIGN

b(s3 ,u) ¼ bTC (s3 )  kb (s3 , u) hD (s3 ) ¼ 1  a4  exp (a5  s3 )

(13a) (13b)

bTC (s3 ) ¼ a6  a7  exp (a8  s3 ) (13c)  exp (a10s3 ) 1 kb (s3 , u) ¼ cos (u þ p6 ) þ a9  sin (u þ p6 ) (13d) b(t) ¼ b(D(t))  ds   dz a16 F F þ  F F D_ ¼ a15  (1  D)a17

b(s3, u) bTC(s3) kb(s3, u)

s3 sv u a0, a4  a17 D_ F ds, F dz F*

b(t)

states (Lux et al. 1998). The damage process itself is modelled with help of a kinetic equation for the evolution of the damage rate dependent on the numerical value of special yield functions F (following the theory of plasticity and describing the onset of dilatancy), accumulated damage D and several material parameters, equation (15). (For further information see Hou & Lux 2000).

(14)

(15a)

with 1_ ij bD(s3, u) hD(s3)

109

strain rate (1/d); damage boundary (MPa); ratio between the fracture strength and the dilatancy strength, with dependence on minimum stress; effective stress in state of fracture (MPa); ln-corrected effective stress in state of fracture (MPa); ratio between the compression and the extension fracture strength; minimum stress (MPa); effective stress (MPa); Lode’s angle (8); material parameters (with various dimensions); damage rate (1/d); yield functions with respect to dilatancy limit (MPa); unit factor (¼ 1 (MPa) for creation of dimensionless fraction); time-dependent failure strength (MPa).

In the framework of the classical design method (Lux-CM) the damage-induced effects on the mechanical behaviour of the rocks are not treated explicitly (ideal viscoplastic material). This means neither a kinetic function for the damage evolution nor a damage or dilatancy strength have to be defined and quantified. Therefore, equation (12) reduces to the first three components, representing elastic strain as well as transient and stationary creep strain. The onset of dilation is related to the stress state and follows equation (13). If this stress intensity level is reached or exceeded, microfractures start to develop and therewith damage in the rock fabric. Equation (14) describes the failure strength dependent on loading time for arbitrary stress states in a formulation similar to that of Mohr based on triaxial compression and extension short-term tests, but enlarged to arbitrary stress

Hard rock layers and fault zone: constitutive model (Hooke/ Mohr-Coulomb). This combined model is used where not only the salt rock mass, but also non-halite rock mass layers have to be taken into account and additional constitutive laws are needed. For rock types with more elastic –brittle material behaviour a linear-elastic constitutive model is taken, including the material parameters deformation modulus Ev and Poisson-ratio v. The material strength is described with the well known Mohr-Coulomb failure criterion, either in the dependence of shear stress equation (16a) or main normal stress equation (16b). In either case, it has to be decided whether the rock mass layer under consideration is fissured/jointed in its natural state and, if so, what influence the fissures will have on the mechanical behaviour (i.e. degree of stiffness as well as strength reduction). The intensity of the geogenic joint system has to be interpreted into the numerical values of the shear parameters (internal friction and cohesion). The constitutive laws are as follows: 1_ ¼ (E)1  s_

tf ¼ s  tan f þ c or s1f ¼

½2  c  cos f þ min s  (1 þ sin f) (1  sin f)

(15b)

(16a) (16b)

with: 1_ s_ (E)1 tf s

s1f f c

elastic strain rate vector; stress rate vector; inverse elasticity matrix; shear failure stress (MPa); normal stress perpendicular to failure plane (MPa); main normal failure stress (MPa); angle of internal friction (8); cohesion (MPa).

Preceding material parameters have to be determined or estimated depending on rock type.

110

K.-H. LUX

Fig. 17. Rock mass models and 3D cavern field modelling. (a) discretized 3D-rock mass model including different rock formations and a tectonic fault zone next to a storage cavity; (b) example for 3D cavern field modelling.

GEOMECHANICS OF SALT CAVERN DESIGN

Fig. 18. Idealized cavern operation process/cavern inside pressure versus time.

Numerical simulation of geomechanical cavern behaviour Simulation model and discretization Rock mass structure, cavern geometry, material behaviour and cavern operation patterns require

111

complex physical models to describe the loadbearing behaviour of the geomechanical structure. The determination of stress, strain and deformation in the rock mass due to cavern construction and operation is only possible with help of numerical tools such as the Finite Element Method (FEM) or Finite Difference Method (FDM). In this case, the computer Codes MISES3 (TDV 2002) and FLAC3D (Itasca 2005) are used for numerical simulations. However, these commercial software codes have been improved with our own subroutines to consider the specific items of salt rock mechanics and storage caverns in salt rock mass described in the previous chapters. These are important to provide sufficiently realistic yet idealized and simplified, geomechanical modelling. Figure 17a shows an example of a rock mass model discretization (FDM, 96192 zones and 105394 grid-points), in this case a spatial 3D-rock mass area including a bedded salt layer with a fault zone and a storage cavity next to this fault zone (see Fig. 7). A particular aim of this numerical investigation is to analyse the interference between cavern operation and fault zone reaction. Furthermore, Figure 17b presents another example of

Fig. 19. Cavern model and effective (von Mises) stress distribution in a horizontal cross-section.

112

K.-H. LUX

3D-modelling and discretization (FDM, 67458 zones and 74158 grid points). In this case, a 3D section of the rock mass surrounding a small cavern field is shown. The cavern field consists of four cavities at different distances and with different configurations. In this example of relatively closely implanted cavities, the main consideration is the mechanical interaction of the cavities as well as the pillar stability. Geomechanical data are not determined values but show some scatter or can only be estimated within a reasonable bandwidth. Therefore, during the real design procedure, several calculations with a parameter variation are necessary to provide an overall impression of the mechanical behaviour of the load-bearing system and its sensitivity to input-parameter uncertainty.

Idealized load cases Storage cavities or more precisely the rock mass surrounding the individual excavations is exposed in principle to the following loadings: † primary rock mass stresses (dependent on the regional geological stress field, rock mass structure and depth); † internal cavern pressure (dependent on excavation process, operation pattern and kind of abandonment); † temperature changes (dependent on brine temperature during excavation as well as gas injection and withdrawal rates and related thermodynamic processes during operation). At the time of the cavern designing process the exact loading conditions, especially for cavern operation phase, are not known, because they are related to gas demand and gas supply in their time and pressure related development. Therefore idealized load-cases have to be determined and analysed with a view to their geotechnical admissibility. Figure 18 shows an example of such idealized load cases that are derived for the seasonal storage operation mode limited to one annual storage cycle. The basic philosophy behind this idealization is that natural gas storage caverns should be operated for some decades. In principle, no degradation of the mechanical or hydraulic quality of the loadbearing elements is allowed with operation time and damage of salt rock mass during the annual storage cycle must be excluded. If there is any damage to rock salt, for example at cavern contour zone during operation at minimum inside-cavern pressure, then within the same storage cycle this damage must be eliminated via healing processes, i.e. a decrease of rock mass quality during operation must be eliminated and the original rock mass quality must be recreated.

At the beginning of a new annual storage cycle the original mechanical rock mass quality is available, so only a reference operational year has to be simulated and analysed numerically. Bearing this precondition in mind, in connection with Figure 18, the following load cases have been investigated: †

† †

Ordinary load cases (1, 2) with representative operation cycles with extreme internal pressure difference between min pi and max pi ! admissible min pi and corresponding waiting time at min pi; admissible max pi; admissible depressurization rate. Additionally, in load case (2) the necessary maximal pressure is specified and the related time for the required healing process to take place. And extraordinary load case (disaster case) (3) with internal pressure drop down to atmospheric pressure level (blow out) ! rock mass damage intensity versus time and allowable time for restabilizing the mechanical structure bearing in mind the precondition for only limited extension of rock mass fractures and spalling (no harmful impacts on neighbouring cavities and surface).

In load case (3) temperature changes (rock mass temperature decrease) have to be taken into account, whereas in load cases (1) and (2) the effect of thermo-mechanical stresses maybe negligible, especially in situations of relatively deep cavities and low depressurization rates. Furthermore, in practice, the reinjection of natural gas takes place at a certain pressure change rate, whereas theoretically an instantaneous pressure increase from minimum to maximum pressure level is modelled. This assumption is based on the fact that unloading stress paths will lead to reduced or even diminishing creep rates and thus the pressure change can be modelled with the linear-elastic constitutive model, independently of the real pressure increase rate. Switching over to other operational patterns, with several storage cycles per year, more or less independent of the seasonally-oriented winter – summer gas demand, it has to be decided, how many cycles can be carried out without damaging the rock mass too much and what is required to ensure restoration of rock mass quality. In general, it is permissible to allow limited weakening of salt rock mass during operation up to macro-fractures and contour spalling, assuming that all the other geotechnical safety criteria besides local stability, i.e. global static stability and tightness as well as third party protection, are fulfilled. In the case of tolerated rock damage up to rock failure a healing phase may not need to be foreseen. Otherwise, the design procedure would need to

GEOMECHANICS OF SALT CAVERN DESIGN 113

Fig. 20. Numerical simulation model and zones with extreme effective stress in the rock mass next to the cavity related to different constitutive laws (without and with damage) as well as damage intensity distribution at cavern contour.

114

K.-H. LUX

Fig. 21. Monitoring of cavern construction process via sonar survey measurement during and after excavation (left) and spatial view (right) on final cavern configuration (after Lux 1984).

include damage and fracture evolution to prepare a proper quantitative prognosis of rock mass behaviour and of failure process. Furthermore, the cavern behaviour during operation must be monitored very carefully. For cavities at shallow depths this relaxation of design philosophy may not be recommended from safety reasons.

Safety analysis and cavern design concepts Around the world within recent decades, design concepts have been developed based on different basic assumptions, in particular, concerning the physical models and the definition of safety. Following the very early approach of Dreyer (1974) the author suggested an initial design concept based on continuum mechanics in the early 1980s and a more sophisticated second design concept based on continuum mechanics but including elements of continuum damage mechanics (Lux 1984; Lux et al. 1999a, b, 2002a, b). The modification of the design concepts with time reflects progress in scientific and technical development. The main steps in this progressive development began with model mechanics, switched to analytical models (CM) based on laboratory investigations (strength and creep); they were followed by numerical

simulations (CM), improved by laboratory investigations (damage) and finally reached the present models with geometric and physical nonlinear (3D) numerical models including damage and healing (CM/CDM) as well as temperatureinduced stress and changes of material properties. Furthermore, it is also now possible to calculate damage-induced dilatency and its influence on the permeability of the salt rock mass (Lux et al. 2000). Bearing in mind all these modelling possibilities it is necessary to decide which design concept should be chosen and which safety definition the cavity design should be based on. Related questions may be the general tolerance of damage in the rock mass and if so, its limitation, or the exclusion of progressive damage up to macroscopic failure and, if not, the extent of allowed rock fall with time.

Fig. 22. Different cavern operation patterns.

GEOMECHANICS OF SALT CAVERN DESIGN

115

Fig. 23. Documentation of measured wellhead pressures versus time (have to be incremented on a daily scale as input data for COSP).

McClain & Fossum (1981) gave a very basic definition of geomechanical safety, ‘In spite of the fact that everybody understands what the word means, defining it turned out to be more than a straightforward task and we finally produced two definitions. From a rock mechanics or analysis point of view, stability may be defined as the intrinsic capability of a rock structure to maintain adequate structural integrity when disturbed from a condition of equilibrium. . .’ Following this definition for geotechnical safety of underground mechanical structures, there is a need for a more precise definition of static stability with a site-specific clarification by the responsible rock mechanics experts and the responsible mining authority. Yet, this is only valid considering

the cavern contour and the related local static stability. Additional safety demands for global static stability, tightness, third party protection and environmental protection give the overall framework and do not show a comparable variety of design freedom. Figures 19 and 20 show the mechanical effect on the load-bearing behaviour when taking into account the damage of rock salt in the design concept. Considering the stress distribution at the cavern wall, Figure 19 shows that in the case of a constitutive model with no damage element the extreme stress values are valid at the cavern contour, whereas in the case a constitutive model with consideration of damage and damaged-induced deformation the extreme stress values at the cavern

Fig. 24. Transformation of wellhead pressure into underground cavern pressures and comparison with admissible minimal and maximal pressures.

116

K.-H. LUX

Fig. 25. Determination of gas withdrawal and injection rates and comparison with admissible limit values.

contours are reduced due to weakening of rock salt and the stress redistribution caused by creep and damage processes. The extreme values now exist at some distance to the cavern contour in the rock mass where the minimal stress component is increased. Figure 20 shows this stress redistribution in terms of effective stress in an areal plot for the different constitution models. The distribution of

maximal stress from contour into the rock mass can be clearly recognized. Figure 20 also shows the development and distribution of the damage intensity (D) around the cavity with its maximum value at the cavern contour in the lower part of the cavity. This numerically determined damage intensity must be assessed according to related failure modes (Lux et al. 1998).

Fig. 26. Determination of integrated pressure-related operation times and comparison with pressure-dependent admissible operation times (linear interpolation).

GEOMECHANICS OF SALT CAVERN DESIGN

117

Fig. 27. Development of effective creep strains related to time-dependent cavern operation and comparison with admissible effective creep strains.

Thus in safety analysis and cavern design, advanced design concepts demand for appropriate design criteria. In the case of a design concept based on CM/CDM theory, modified or additional criteria are necessary, compared to the criteria mentioned above. These improved criteria must take into account the allowable intensity of damage and the healing of damaged rock mass areas or their gradual failure with continued cyclic cavern operation.

Monitoring of cavern construction and operation Since salt cavities are not directly accessible, neither their construction nor their behaviour during operation can be observed directly. To control the construction process, the excavation of a salt cavity by solution mining is usually continuously monitored via mass balance calculations and

Fig. 28. Development of effective creep strains related to pressure-dependent cavern operation and comparison with pressure-dependent admissible effective creep strains.

118

K.-H. LUX

Fig. 29. Development of operated gas volumes (integrated withdrawal and injection volumes) versus operation time.

step-wise monitored via sonar measurement. The main objective of this monitoring is to guarantee the adherence of the required dimensions of the cavity, i.e. cavern diameter and cavern roof configuration. The result of such a monitoring is shown in Figure 21. Assuming that the pre-defined configuration has been met during construction (proved via sonar survey measurements), further monitoring is necessary during cavern operation. Considering natural gas storage cavities and bearing in mind that the cavern design is just a prognosis, the mining authority will prescribe a certain time period after which the actual cavern state (contour behaviour, convergence) has to be demonstrated. These time periods may vary between five and ten years. Modern sonar surveys operate not only in brine, but also in natural gas; however other measuring systems using laser modules may also be used. The main point of these in-situ measurements is to demonstrate a stable cavity with respect to static stability as well as a reliable design. Otherwise additional design work is necessary, the operational conditions have to be modified and monitoring has to be intensified. Further hints on a reliable design can be derived from the results of the surface subsidence measurements, which are a consequence of the convergence of the salt rock mass and this convergence is closely dependent on the creep characteristics of the site-specific rock salt and the operation pattern. These more traditional instruments for monitoring cavern behaviour, i.e. documentation of well head pressure versus time, annual or less frequent

surface subsidence measurements, and underground laser measurements every 5 to 10 years, may be insufficient in some cases to demonstrate local and global static stability as well as tightness and long-term surface protection. Therefore, a special monitoring strategy has been developed, based on the real cavern operation and taking special account of rock mechanics. There are three main reasons for the development of a computer-aided programme for documenting and monitoring cavern operation. First, cavern operation patterns will become more and more complex in the future, changing from seasonal storage on a normal basis to a more flexible storage strategy with significantly shorter cycle times, characterized by terms such as high deliverability and frequent turnover. The coherence with rock mechanical criteria as well as the determination of used rock mass load-bearing capacity (convergence, damage) related to the cavern operation is no longer possible in this basic manner. Furthermore, in some cases the rock mass deformation is limited in order to prevent undetectable progressive weakening and thus a certain operational phase (withdrawal, injection) has to be followed by a so-called recovery phase over a predetermined period of time at a predetermined internal cavern pressure range. Therefore actual data of cavern status seem to be necessary at any time of operation. Second, complex cavern operation patterns will demand careful planning of the existing possibilities within the predetermined limits to ensure most effective economic cavern operation, for example the combining of existing working gas

GEOMECHANICS OF SALT CAVERN DESIGN

volumes and tolerable withdrawal rates with operation times for specific internal pressures or necessary gas injection rates to refill the cavity in time. Third, in some cases the mining authority demands a yearly documentation of past cavern operation including adhering to rock mechanics criteria. This documentation will prove both cavern operation within the approved limits as well as geotechnical safety of the underground storage facility. Figure 22 shows internal cavern pressures versus time on a more or less two-yearly basis for seasonal storage and frequent turnover storage. In the latter case the demands of rock mass load-bearing capacity is much more stringent and therefore adequate instruments for documentation (past as well as recent operation) and planning (future operation) are necessary. To support cavern operation with respect to documentation and planning, software called Cavern Operation Survey and Planning (COSP) has been developed in recent years. This software is based on daily representative cavern pressures (well head or underground) derived from real cavern operation. These operated or planned cavern pressures/operation times are transformed into cavern-specific geomechanically-related data such as pressure change rates, transient and stationary creep strain-increments as well as cavern convergence increments. For this transformation, the results of the FEM- or FDM-calculations developed for cavity design are used. The pressures themselves and their assigned waiting times as well as the derived data with relevance to rock mechanics are compared with site-specific criteria/admissible limit values. This comparison will show admissible and inadmissible operation states with respect to geomechanical criteria. In total, the following criteria will be checked: minimal and maximal admissible pressures; maximal withdrawal and injection rates; operation time(s) at minimum pressure(s); necessary time as well as pressure range for recreation of rock mass quality; admissible local deformation (creep strain) with respect to internal cavern pressure. Furthermore, withdrawal and injected gas volumes are calculated on a daily basis and integrated gas volumes are determined. Changing cavern storage volume due to rock mass convergence can be taken into account. It is also possible to enlarge the EDV-program to give a prognosis on future surface subsidence or to take into account thermodynamic processes with respect to more precise gas mass balances. Figures 23 to 28 give an overview on the input/ output data of the COSP software. In detail these figures show the results of a retrospective analysis

119

of cavern operation for one operation year (artificial operation patterns with three equal operation cycles): † documentation of measured wellhead pressures versus time (incremented on a daily scale as input data for COSP; Fig. 23); † transformation of wellhead pressure into underground cavern pressures and comparison with admissible minimal and maximal pressures (Fig. 24); † derivation of withdrawal and injection rates and comparison with admissible rates (Fig. 25); † determination of integrated pressure-related operation times and comparison with pressuredependent admissible operation times (linear interpolation; Fig. 26); † development of effective creep strains related to time-dependent cavern operation and comparison with admissible creep strains (Fig. 27); † development of effective creep strains related to pressure-dependent cavern operation and comparison with pressure-dependent admissible strains (Fig. 28); and † development of operated gas volumes (integrated withdrawal and injection volumes) versus operation time (Fig. 29). For further information on the COSP software, its theoretical background and handling, as well as its application features see Du¨sterloh & Lux (2005). A demonstration CD is available via the author’s e-mail address.

Cavern abandonment: hydromechanical behaviour of a sealed liquid-filled cavity The question of how to abandon salt cavities with respect to long-term surface and environmental protection has been discussed for more than twenty years. The main aspiration is a safe, maintenancefree and economical solution. Several possibilities have been considered: (1) air-filled cavities under atmospheric pressure; (2) refilled cavities with water or brine, unsealed; (3) refilled cavities with sand or gravel; (4) refilled cavities with an organic non-liquid more or less toxic wastes, for example from mining or non-mining industries as well as from incineration plants (municipal wastes); and (5) refilled cavities with water or brine with a permanently sealed borehole. Whereas with options (1) and (2) neither the longterm static stability nor surface protection can be guaranteed, it seems that case (3) is technically feasible but the industry considers that it is not economically viable. In principle, case (4) is also technically feasible, even as an underground repository for the

120

K.-H. LUX

final disposal of chemo-toxic waste regulated by law. In reality, at least in Germany, no such waste disposal facility has yet been created despite existing legal regulations. However, it is questionable whether this method should be the standard one for salt cavern abandonment, even without any other proven alternative. Currently in the industry, option (5) seems to be the accepted basic solution, for both economic reasons and public acceptance. Many scientific investigations with regard to technical feasibility have been carried out over the past twenty years or so. The main question still relates to the hydromechanical behaviour of a sealed liquid-filled salt cavity in the long term, this involves the identification and characterization of the basic hydraulic and mechanical processes as well as existing possibilities for physical modelling and numerical simulation of the coupled processes associated with this question. Assuming that long-term tight borehole sealing exists, the rock mass deformation due to creep properties of rock salt together with different liquid and rock salt densities will in principle cause a certain fluid-overpressure at the cavern roof area compared to the lithostatic rock mass pressure. Figure 30 shows the calculated increase of the fluid pressure with time with the assumption of an impermeable rock mass (Lux 2005). Lab tests and field experience show, that assuming these near lithostatic pressures, rock salt is no longer impermeable, but will become permeable to

a limited degree. The basic mechanism of this process is called infiltration and works as follows: †

† †

the pressurized fluid at first creates microfractures in the polycrystalline rock salt matrix along the grain boundaries against the resistance of the acting rock mass stresses; the fluid infiltrates into these newly created fractures; and the fluid pressure is built up again due to cavern convergence and the micro-fractures propagate into the rock mass and may also be connected.

The infiltration process characterized in this way is the basis for a new physical model that has been developed in recent years (Lux 2005, 2006). The main elements of this infiltration model are: onset of infiltration; direction of infiltration; infiltration rate; infiltration progress during time increment; intensity of infiltration and propagation of infiltration. To determine the physical context and to determine the necessary material parameters, test equipment has been built and used to model and analyse the infiltration process under laboratory conditions. Figure 31 gives an example of the infiltration of a fluid into a rock salt sample, starting at an artificial damage in the salt fabric. In this case, the pressurized fluid is a special tracer fluid that can be visualized via black light. The sample is loaded in axial as well as lateral directions. The acting fluid pressure is several bars larger

Fig. 30. Pressure build-up in a liquid-filled and sealed salt cavity (after Lux 2005).

GEOMECHANICS OF SALT CAVERN DESIGN

121

Fig. 31. Infiltration test: salt rock sample with infiltration of liquid.

than the lateral mechanical stress. The infiltration process can be indirectly monitored by the additional fluid volume necessary to keep the fluid pressure constant during the test time. After finishing the test at a certain time, dismantling of the

test sample and its slicing in the axial direction, the infiltrated rock salt area is clearly visible. After a certain time, the infiltration front will reach the opposite end of the sample. At this point, the infiltration process will change into a seepage process

122

K.-H. LUX

Fig. 32. Infiltration test with test conditions and main results: infiltration phase followed by seepage phase.

Fig. 33. Measured mean infiltration rates versus effective hydraulic pressure.

GEOMECHANICS OF SALT CAVERN DESIGN

123

Fig. 34. Measured mean specific infiltration volume versus effective hydraulic pressure.

following the pathways created before. Figure 32 shows the result of a typical infiltration test. Measured parameters are the mechanical pressures and the fluid pressure together with the fluid volume necessary to maintain the fluid pressure at a constant level during the infiltration and seepage flow stages. During the infiltration stage this fluid volume is equal to the total infiltration volume. The infiltrated rock volume can be estimated from the examination of the salt cores. Therefore, a mean infiltration rate and a specific infiltration volume can be derived from the test data. Figures 33 and 34 show the mean infiltration rate versus effective hydraulic pressure and the mean specific infiltration volume versus effective hydraulic pressure. Based on these lab test data it is now possible to determine the necessary material parameters. To simulate the infiltration process with regard to sealed liquid-filled salt cavities a special computer code called INFIL has been developed. Main elements of this computer code are the infiltration model introduced above; a subcode for the calculation of geomechanical state variables in the rock mass (pressure build-up phase and infiltration phase); a subcode for calculation of hydraulic processes in the infiltrated rock mass area (seepage flow phase), subroutines for coupling these different processes and a poroperm model that offers the transformation of secondary porosity due to infiltration into secondary permeability. The application of the computer code INFIL to analyse the long-term load-bearing behaviour of sealed liquid-filled salt cavities leads to some surprising new results. Figure 35 shows a representative salt cavity for storage of natural gas with a configuration typical for storage cavities in domal salt structures. Some basic parameters are

given. More detailed geometric and mechanical parameters are not of special importance in the following context. Within the salt rock mass there is an isotropic primary stress field existing on the level of the depth-dependent lithostatic (overburden) pressure. The numerical simulation process starts with the brine-filled cavity at hydrostatic pressure and the related stationary stress-strain field. The cavity (more precisely, the borehole) is then closed and, generally, due to the load difference between rock mass pressure and brine pressure the rock mass creeps inwards. The integral creep deformation leads to a so-called volume convergence and this convergence caused by brine compressibility

Fig. 35. Representative salt cavity model with discretization and some basic data.

124

K.-H. LUX

creates an increase in brine pressure. This brine pressure increase induces stress redistributions in the rock mass with the consequence of reduced creep rates. This hydraulic-mechanical coupled process is continued and numerically simulated on a daily incremental basis. Eventually, a situation is reached with an increased hydraulic pressure in the cavern roof area that is slightly larger than the associated tangential rock mass stresses. At this moment according to equation (10.1) the infiltration process starts. Propagation rates and specific infiltration intensity follows the developed physical models, whereas the direction of infiltration is related to the local stress field

Fig. 36. Propagation of infiltration front at different times.

and follows the basic assumptions of hydrofrac theory. The infiltration of brine into the rock mass causes the internal cavern pressure to decrease and it takes time to increase the brine pressure by cavern convergence again at the necessary level for further infiltration. This infiltration process numerically also runs on a daily incremental basis. The redistribution of the stress field according to changing brine pressure is also taken into account, resulting in a coupled simulation of infiltration process and mechanical process. This coupled process continues after some time with accelerated rates due to decreasing brine pressure according to the decreasing rock

GEOMECHANICS OF SALT CAVERN DESIGN

mass pressure, related to the progressively shallower depth of the infiltration front. Finally, the infiltration front reaches a porous water or brine-saturated rock formation, for example above the salt roof. After breakthrough of the infiltration front, the brine pressure decreases to the formation pressure of the porous formation, for example hydrostatic pressure. At this moment, the infiltration process will change into a seepage flow process, physically modelled with Darcy’s flow law. The seepage flow is restricted to the infiltrated rock mass area with its pathways, quantified as secondary porosity. The related permeability is calculated with the help of a poroperm model. The seepage flow process is also simulated on a daily incremental basis. Simplifying assumptions are saturated rock mass as well as laminar one-phase flow. Furthermore, stationary conditions are assumed for each time increment. Due to brine outflow into the porous rock formation the brine pressure in the cavity decreases, the convergence rate increases, the stress field changes and secondary permeability may change due to its stress dependency. Furthermore, the hydraulic pore pressure in the infiltration zone will change according to the changing hydraulic potential field. Finally, a stationary seepage flow state will be reached with equilibrium between cavern volume decrease according to creep convergence and outflow of brine, determined by pore pressure distribution and transmissivity of the created infiltration zone in the salt rock mass. To summarize, the complex hydromechanicalcoupled processes starting at cavern sealing and ending somewhere in the far future can be differentiated into three main phases: a pressure-built up phase in the closed cavity; an infiltration phase with creation of secondary pathways in the primary impermeable rock mass creating then a so-called infiltration zone; and a seepage flow phase with brine outflow on the microfractureinduced pathways within and only within the infiltration zone. Characteristic results of the numerical simulation are illustrated. Figure 36 shows the propagation of the infiltration zone in the salt rock mass. The infiltration zone is restricted to the cavern roof area and generally will develop in a limited area along the cavern axis towards the top of the salt structure, following the decreasing lithostatic rock mass pressure. The development of the infiltration front versus time is therefore shown in Figure 37. The propagation of the infiltration front occurs initially relatively slowly, due to the high internal cavern pressure according to the cavern induced stress field and low convergence rate, respectively. Yet after some time the propagation rate increases significantly and continuously. This change in the propagation rate takes place at the

125

Fig. 37. Development of infiltration front versus time.

moment when the infiltration front reaches the rock mass area which is no longer influenced by the secondary stress field of the cavity, i.e. the primary rock mass stress field above the cavity roof. Figure 38a & b shows the development of the brine pressure at cavern roof versus time. This process covers the initial pressure build-up phase (1), infiltration phase (2) as well as seepage flow phase (3). Together with Figure 37, which gives, with time, the position of the infiltration front, the corresponding pore pressures in phase (2) and (3) in the infiltrated rock mass can be calculated at each point in time and for every rock mass

Fig. 38. Development of brine pressure at cavern roof. (a) development of brine pressure at cavern roof versus time, initial pressure build-up phase (1) of closed liquid filled cavity; (b) development of brine pressure at cavern roof versus time with infiltration phase (2) and seepage flow phase (3).

126

K.-H. LUX

element. Figure 38a shows that the brine pressure increases at first in the pressure build-up phase more or less up to the level of lithostatic pressure in the cavern roof. Neglecting the mechanical impacts of the thermal expansion of brine (precondition: thermal equilibrium between rock mass and brine before cavern sealing) the process of pressure build-up in the cavity is finally dependent on cavern convergence and brine compressibility. In this example it takes about 100 years for the internal pressure to reach the level of primary rock mass pressure at the roof area of the cavity. Figure 38a (logarithmic scale) shows that the pressure buildup rate decreases with time, a consequence of the diminishing effective rock mass pressure due to increasing cavern inside pressure. The pressure rate at the final time period of phase (1) amounts to 1.76 1024 bar/day (to an order of magnitude). In this circumstance no hydrofracturing with macrofractures will occur, following current laboratory

test and field test experience. But an infiltration of brine into the polycrystalline salt rock fabric can not be prevented in this situation with slowly increasing brine pressure. The onset on infiltration phase (2) and its propagation can be seen from Figures 37 and 38b, showing both development of infiltration front into the rock mass and related pressure evolution versus time. According to Figure 38b the infiltration front reaches the top of the salt (salt mirror) about 530 years after onset of infiltration and the fluid breaks through into a porous water-saturated rock mass formation above the salt. At this time point seepage flow starts along the pathways just created by infiltration process. The transmissivity of the infiltration zone, hydraulic heads of brine in the cavity and water in the watersaturated porous rock mass formation as well as cavern convergence due to rock mass creep finally determine the stationary field of hydraulic potential and the level of brine pressure during phase (3), in

Fig. 39. The long-term behaviour of a liquid-filled closed salt cavity: fundamental aspects.

GEOMECHANICS OF SALT CAVERN DESIGN

this case about pi ¼ 22.1 MPa at the cavern roof. For further information see Lux (2006). Finally the results of numerical simulations give an excellent basis for drawing fundamental conclusions with respect to cavern abandonment. Figure 39 shows the principal long-term behaviour of a sealed, liquid-filled salt cavity. Significant aspects include the following questions: † Does an infiltration zone develop with progressive enlargement to permeable rock mass formations? † How much time does the infiltration process take to build up to hydraulic breakthrough? † Do the mechanical properties of salt rock mass in the infiltration zone change? † What kind of hydraulic situation would be expected after hydraulic breakthrough in the long term (pore pressures, internal cavern inside pressure)? † Are there any further hydraulically induced microfissures after breakthrough? † Does the infiltrated salt rock mass above cavity roof maintain its mechanical quality? † Is it possible to drain the released brine into the deep groundwater ecologically and safely? † Are there any consequences to the construction of the well bore seal? † How much time does it take to reach the final convergence of the cavity? † Are the subsidence rate and the maximum subsidence acceptable at the surface? † Is there a risk of developing sinkholes in the long term?

127

properties, characterized by elastic, plastic, viscous and clastic stress-strain relationships (constitutive laws). Furthermore, to analyse failure mechanisms with respect to loss of tightness of salt rock mass, numerical modelling of geomechanical and geohydraulic processes, and their coupling, is necessary. This article presents some of the important, recent developments in rock salt mechanics and storage cavern design. The selection of items and their specification is subjective and reflects the author’s views and experience. Further specification is available using the cited literature. Taking into account basic knowledge and some experiences dealt with in this paper the span of life of a salt cavity can be subdivided into the following major phases: excavation phase (solution mining); operation phase; abandonment phase; and maintenance-free long-term phase. There is currently a great effort to extend the natural gas supply system in Britain. Part of this extension is the construction of underground storage cavities in salt deposits. This paper may lead to a better understanding of these geotechnical constructions which cannot be observed or even inspected by the human eye, but must be excavated, operated and finally closed, after several decades with sufficient safety for surface (third party) protection as well as economics. Therefore appropriate and reliable methods for the design and monitoring of these underground constructions are needed. The selected aspects of the current situation presented here may give an overview on today’s knowledge and necessary future development to improve understanding and economics as well.

Concluding remarks This paper presents a personal view of the principles of rock mechanics behind creating, operating and abandoning salt cavities, especially gas storage cavities, in a safe and economical manner. The past thirty years have led to a much deeper insight into salt cavern behaviour in the field as well as in theory — based on increased practical experience and scientific knowledge worldwide. The main topics of the last ten years are the introduction of elements of damage mechanics in salt mechanics and the improvement in laboratory testing methods, both of them aiding a deeper understanding of mechanical failure and healing processes and leading to better physical modelling. On the other hand advanced computer techniques have enabled the numerical simulation of complex mechanical load-bearing structures, for example multilayer continua, irregular geometric configurations of relevant underground excavations, multicycle load cases, geometric nonlinear behaviour (large deformation) and physical nonlinear material

The author would like to express his thanks to R. Wolters and O. Czaikowski for their kind help in the preparation of this paper, both of them being members of staff of the Professorship for Waste Disposal and Geomechanics at Clausthal University of Technology.

References D REYER , W. 1974. Gebirgsmechanik im Salz. F. Enke Verlag, Stuttgart. D U¨ STERLOH , U. & L UX , K.-H. 2003. Geologische und geotechnische Barrieren - Gedanken zur Nachweisfu¨hrung. Clausthaler Kolloquium zur Endlagerung 2003, 8– 9. Mai 2003, Clausthal-Zellerfeld. Schriftenreihe Professur fu¨r Deponietechnik und Geomechanik TU Clausthal. H14, S.89–120. H OU , Z. 2002. Geomechanische Planungskonzepte fu¨r unterta¨gige Tragwerke mit besonderer Beru¨cksichtigung von Gefu¨gescha¨digung, Verheilung und hydromechanischer Kopplung. Schriftenreihe Professur fu¨r Deponietechnik und Geomechanik, TU Clausthal, H13. H OU , Z. & L UX , K.-H. 2000. Ein Scha¨digungsmodell mit Kriechbruchkriterium fu¨r duktile Salzgesteine bei

128

K.-H. LUX

langzeitiger Beanspruchung auf der Grundlage der Continuum-Damage-Mechanik. Bauingenieur, 75, (13). ITASCA . 2005. Flac 3D. Itasca Consulting Group, Inc., Minnesota, 2005. KBB. 2005. KBB Underground Technologies. World Wide Web Address http://www.kbbnet.de. L UX , K. H. 1984. Gebirgsmechanischer Entwurf und Felderfahrungen im Salzkavernenbau. F. Enke Verlag, Stuttgart, 1984, ISBN 3-432-94171-4, Habilitationsschrift. L UX , K.-H. 2005. Zum langfristigen Tragverhalten von verschlossenen solegefu¨llten Salzkavernen – ein neuer Ansatz zu physikalischer Modellierung und numerischer Simulation. Theoretische und laborative Grundlagen. Erdo¨l Erdgas Kohle, 11, 414–422, Urban Verlag, Hamburg. L UX , K.-H. 2006. Zum langfristigen Tragverhalten von verschlossenen solegefu¨llten Salzkavernen – ein neuer Ansatz zu physikalischer Modellierung und numerischer Simulation. Rechnerische Analysen und grundlegende Erkenntnisse. Erdo¨l Erdgas Kohle, 2. L UX , K. H. & R OKAHR , R. B. 1980. Dimensionierungsgrundlagen im Salzkavernenbau. Taschenbuch fu¨r den Tunnelbau, S. 240–312, Verlag-Glu¨ckauf, Essen. L UX , K. H. & R OKAHR , R. B. 1982. Laboratory investigations and theoretical statements as basis for the design of caverns in rock salt formations. Proceedings of the First Conference on the Mechanical Behaviour of Salt, Penn State University, Trans Tech Publications. L UX , K.-H. & H EUSERMANN , S. 1983. Creep tests on rock salt with changing load as a basis for the verification of theoretical material laws. Proceedings 6th Symposium on Salt, Toronto. L UX , K.-H., H OU , Z. & D U¨ STERLOH , U. 1998. Some new aspects for modelling of cavern behavior and safety analysis. Proceedings of SMRI Fall Meeting 1998, Rome, Italy. L UX , K.-H., H OU , Z. & D U¨ STERLOH , U. 1999a. Neue Aspekte zum Tragverhalten von Salzkavernen und zu ihrem geotechnischen Sicherheitsnachweis, Teil 1: Theoretische Ansa¨tze. Erdo¨l Erdgas Kohle, 3. L UX , K.-H., H OU , Z. & D U¨ STERLOH , U. 1999b. Neue Aspekte zum Tragverhalten von Salzkavernen und

zu ihrem geotechnischen Sicherheitsnachweis, Teil 2: Beispielrechnungen mit dem neuen Stoffmodell Hou/Lux. Erdo¨l Erdgas Kohle, 4. L UX , K.-H., H OU , Z., D U¨ STERLOH , U. & X IE , Z. 2000. Approaches for validation and application of a new material model for rock salt including structural damages. Proceedings of 8th World Salt Symposium, May 2000, The Hague. L UX , K.-H., D U¨ STERLOH , U. & H OU , Z. 2002a. Erho¨hung der Wirtschaftlichkeit von Speicherkavernen durch Anwendung eines neuen Entwurfs- und Nachweiskonzeptes (Teil I). Erdo¨l-Erdgas- Kohle, 6. L UX , K.-H., D U¨ STERLOH , U. & H OU , Z. 2002b. Erho¨hung der Wirtschaftlichkeit von Speicherkavernen durch Anwendung eines neuen Entwurfs- und Nachweiskonzeptes (Teil II). Erdo¨l-Erdgas-Kohle, 7. L UX , K.-H., H OU , Z. & X IE , Z. 2003. Ein Kopplungskonzept zur Beschreibung der hydromechanischen Wechselwirkungen in den aufgelockerten Konturzonen um ein Abdichtungsbauwerk im Salzgebirge. Bauingenieur, Band 78(11). L UX , K.-H., W ERMELING , J. & B ANNACH , A. 2004. Determination of allowable operating pressures for a gas storage cavern located close to a tectonic fault. Technical Conference Paper, SMRI – Fall 2004 Conference, Berlin. M C C LAIN , W. C. & F OSSUM , A. F. 1981. The evaluation of room stability. Proceedings of the 1st Conference on the Mechanical Behaviour of Salt. Trans Tech Publications, Clausthal. S CHULZE , O., P OPP , T. & K ERN , H. 2001. Development of damage and permeability in deforming rock salt. Engineering Geology, 61. S EDLACEK , R. 2007. Untertagespeicher 2007. Niedersa¨chsisches Landesamt fu¨r Bodenforschung (LBEG) Hannover. Erdo¨l Erdgas Kohle 124. Jg. 2008, Heft II. TDV. 2002. Mises Benutzerhandbuch TDV Ges.m.b.H., Rev. 12.01, January 2002. X IE , Z. 2002. Rechnerische Untersuchungen zum mechanischen und hydraulischen Verhalten von Abdichtungsbauwerken in Untertagedeponien fu¨r den Fall eines Lo¨sungszutritts. Schriftenreihe Professur fu¨r Deponietechnik und Geomechanik, TU Clausthal, H 12.

New procedure for tightness tests (MIT) of salt cavern storage wells: Continuous high accuracy determination of relevant parameters, without the need to use radioactive tools HARTMUT VON TRYLLER1, ANDREAS REITZE1* & FRITZ CROTOGINO2 1

SOCON Sonar Control Kavernenvermessung GmbH, Giesen, Germany 2

KBB Underground Technologies GmbH, Hannover, Germany *Corresponding author (e-mail: [email protected])

Abstract: In performing the mechanical integrity test (MIT) on salt cavern storage wells the most used method worldwide is the In-Situ Balance method (ISB). The principal sources of errors in the execution and evaluation of the test are the depth change of the gas/liquid interface and the surface area of the interface. In the past, the interface depth has been predominantly determined using radioactive methods, i.e. gamma-gamma, neutron-gamma and neutron-neutron tools. The disadvantages of these methods are the low measurement accuracy and the need to perform several tool runs during the test period, which introduces an additional source of error, because cost factors normally prevent continuous measurements. A new method (SoMIT) is presented based on ultrasonic techniques in which the interface depth, the temperature and the differential pressure at the interface depth can be measured continuously during the tightness test while achieving much greater levels of accuracy than was previously the case. In the SoMIT method a tool is fixed in place throughout the entire test period such that the problems associated with several tool runs are also avoided. The advantages for users are to be found in the greater accuracy available to verify the tightness of the well and also in the reduced test period.

An important prerequisite for assuring the secure and reliable operation of salt storage caverns is the verification of mechanical integrity of the cased and cemented borehole, particularly in the area of the final casing seat, against the host rock, including in water-bearing formations. This provides the proof that leakage will not result in: † the uncontrolled escape of flammable storage products to the surface or the contamination of usable drinking water sources, e.g. aquifers; or † unacceptable and costly loss of stock. Interface tests have developed worldwide into the standard method of verification; this involves injecting a limited volume of a test gas (which is lighter than the liquid) into the cavern. By measuring and monitoring various parameters such as head pressure and interface depth over a period, it is possible to evaluate and test the gas tightness of the well. In practice, three test methods have established themselves of which, worldwide, the most commonly used is the In-Situ Balance method (ISB). However, the levels of accuracy of this method may be compromised by the ability to determine the interface depth accurately. This paper presents

an improved ISB method involving the use of a sonar tool capable of achieving much greater accuracy and also of providing quasi-continuous interface depth measurements.

In-Situ Balance method and comparison with alternatives After pressurizing the cavern with brine or liquid product, and in order to set an interface in the open borehole, gas is injected into the annulus of the well to a level below the casing shoe of the final cemented casing. In the case of caverns completed with a packered production well string, the test is concentrated on the final cemented casing seat area (Fig. 1). Otherwise, the cemented casing to surface is included in the test. Prior to the commencement of the test phase, time is allowed for short-term pressure and temperature fluctuations in the system to approach equilibrium. To achieve a quantitative determination of the leakage rate, then determination of the well geometry (cross-sectional area versus depth) down to the planned interface depth is required. Once

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 129–137. DOI: 10.1144/SP313.8 0305-8719/09/$15.00 # The Geological Society of London 2009.

130

H. VON TRYLLER ET AL.

Fig. 1. Basic concept of a gas interface test (using nitrogen, N2) with and without brine withdrawal string (ISB method).

obtained, the following measurements are necessary at the start and at the end of a test interval: † wellhead pressure of test gas; † interface depth; and † temperature in well (optional to increase accuracy). To date, the disadvantage of the In-Situ Balance (ISB) method has been the lack of accuracy obtained in measurement of the well geometry and, in particular, the varying depth of the gas/liquid interface. This means that, particularly in the case of large borehole diameters, long and thus costly test periods are required. In addition, running the various necessary well logs introduces an additional source of error. The advantages are simple installation, execution, and evaluation. The test can be carried out without any major conversion during first fill or even during breaks in operations on liquid filled caverns. The accuracy of the ISB method is dominated by the cross-sectional area at interface depth and the error in determining the interface depth using gamma logs (i.e. approx. +0.1 m or 1/3 ft). An alternative method used predominantly in Germany is the In-Situ Compensation (ISC) method (Edler et al. 2003), which successfully avoids the errors in determining the interface depth by installing a special test string in which the interface depth is defined using a peephole. The higher level of accuracy achievable with this method is at the cost of installing special test equipment. In France, the Above Ground Balance (AGB) method is favoured in which a defined volume of a liquid test medium (gas oil) is injected into the test annulus to below the last cemented casing and is subsequently recovered. By comparing the volume recovered with that injected the tightness of the

well can be evaluated. For a detailed comparison of all three methods see also Crotogino (1995).

Wireline-based method to determine interface depth in MITs Radioactive methods In the past, the use of wireline-based methods to determine the interface level during MITs has focused mainly on radioactive methods. In recent years, the classic gamma-gamma tool has been increasingly replaced by ING tools (impulse neutron-gamma). In gamma-gamma probes a permanently radioactive substance (e.g. Cs137) is used and high levels of care are required in handling radioactive materials with corresponding risks. These are subject to strict authority regulations. In the case of ING probes the immediate risk for the workforce from radiation is avoided, none the less these probes require special permits and are subject to corresponding regulations and stipulations. However, both of these methods are subject to the same basic sources of error in their mode of employment in MITs. The interface depth is derived from geophysical logs (Fig. 2), which are recorded during the tool run and where the depth of the gas– liquid interface is derived from the change in intensity of the recorded level of gamma radiation. The gamma ray log in Figure 2 shows the gamma intensity suddenly increases at the brine–nitrogen interface due to the lower density of the nitrogen compared to the brine. A further source of error is in determining the reference depth. In cases where a log uses the surface as reference depth the interface depth is

TIGHTNESS TESTS FOR SALT CAVERN STORAGE

131

Fig. 2. Gamma- and CCL-log (abbreviations: cpm, counts per minute; CCL, casing collar locator tool).

automatically subject to all errors relating to the length measurement of the cable and hence is no longer suitable for precise depth calculations. A gamma tool suspended from the last cemented casing shoe or a collar near the interface depth offers some advantages in comparison. The cable length error is, for example, much smaller in this case. Having said this, the use of a casing shoe or collar (determined using a CCL, casing collar locator, or a multiple casing collar locator; MCCL) is subject to a further source of error, since this reference depth itself has also been derived from a log (Fig. 3). As a consequence of these sources of error as described previously, the interface depth when using radioactive-based probes and those with reference depths determined with CCL or MCCL have a margin of error of approximately þ/210 cm. This value is only achieved in cases where the logs are digitally recorded and evaluated. Where graphic methods are used the error margin is greater by a factor of two.

Fibre optic technology The distributed fibre optic temperature-sensing technique makes it possible to measure temperatures

over extended time periods and lengths of up to several kilometres while allowing an exact allocation of temperature data. A spatial resolution of 0.25 m and a temperature resolution of 0.05 K are possible. The fibre optic temperature-sensing system operates without any electronic circuits along the fibre. It is thus totally insensitive to electromagnetic fields and can be used even in hazardous areas subject to stringent explosion protection requirements. In terms of tightness tests, the fibre optic temperature sensing techniques offers primarily the following advantages: † † †

† †

simple installation of sensing cable in well; cable remains in well throughout entire test period; temperature field is not influenced by repeated logging runs such that measurements can take place continuously throughout the selected time window; simultaneous measurement of temperature along the entire length of the cable; and no electronic/electric components or modules are located along the length of the temperature sensing cable.

The evaluation of the fibre optic logs achieved in a cavern shows that it is possible in principle to

132

H. VON TRYLLER ET AL.

Fig. 3. Evaluation of logs.

determine the depth of the gas/liquid interface using fibre optic temperature logs through the brine string (Grosswig et al. 2003). Figure 4 shows an example of a recorded temperature profile with the interface at the intersection point of the two compensating lines. The measurements within the brine withdrawal string represent the worst case for the fibre optic sensing principle. More favourable measurement

conditions would apply if measurements were performed in the casing without the brine string, since this would allow the sensing cable to be in direct contact with the gas/liquid interface. The fibre optic temperature sensing technique allows the temperature –depth distribution to be determined continuously throughout the tightness test period across the entire length of the well with a high level of local resolution, generating data to

Fig. 4. Deriving interface depth from a temperature profile (fibre optic technology).

TIGHTNESS TESTS FOR SALT CAVERN STORAGE

allow temperature-dependent changes in the test gas volume to be recorded and corrected if required.

Concept for implementation of sonar tool during MITs Basic aspects of the SoMIT method The new SoMIT method concept is a further option for determining the interface depth (Fig. 5). This

Fig. 5. The SoMIT concept.

133

method is particularly well suited for gas caverns in cases where the test is to be performed prior to installing final completion and before the brine withdrawal string has been installed. In the SoMIT method the interface depth is no longer derived from log data but is determined directly from a position just above the interface using ultrasonics. The distance between a vertically arranged ultrasonic sensor, part of the SoMIT tool and the interface is determined by using the interface as a reflector. This distance can be evaluated

134

H. VON TRYLLER ET AL.

extremely precisely and has a resolution of between 1 and 2 mm. To avoid possible sources of error as are present in traditional wireline-based methods, for example as a result of cable lengthening, the SoMIT tool is fixed in place above the interface in the inner casing string during the entire duration of the MIT. This avoids problems associated with several tool runs. Since in this method the interface is observed continuously, any changes in interface depth are due entirely to ultrasonic measurements and have an accuracy of 5 mm, i.e. have a higher order of precision than the methods used previously (see also above). Concurrent to the continuous monitoring of the interface depth using ultrasonics, the temperature is also recorded continuously as is the pressure in the area of the interface. Figure 6 shows for a partition of a SoMIT test the continuously recorded interface depth as well as the temperature versus test time. An additional channel also allows a continuous high resolution pressure measurement to be undertaken at the cavern head. The recording of the pressure development using special pressure sensors allows a resolution of 1 mbar (c. 0.015 psi).

The evaluation of the SoMIT test uses the tried-and-tested algorithms known from the ISB method. However, the SoMIT method generates greater volumes of more precise data than are gathered from the use of the previously available technologies. For example, a temperature log is recorded during the run in and pull out of the SoMIT tool, providing documentation of the initial and the final situation. During the test the SoMIT software is capable of providing ongoing information of the current interface depth at any time, as well as temperature and differential pressure. This information allows trends to be detected early on.

SoMIT tool technology In order to implement the SoMIT concept, a tool was built which can be left in place in the cavern throughout the execution of the MIT and provides quasi-continuous monitoring of the interface (Fig. 7). For this purpose the tool has an ultrasonic transducer at its lower end capable of recording the position of the interface with very high accuracy, i.e. a resolution of approximately 5 mm (c. 1/5 inch).

Fig. 6. Interface depth and temperature logs obtained during the SoMIT test.

TIGHTNESS TESTS FOR SALT CAVERN STORAGE

135

The tool is also fitted with a multiple casing collar locator (MCCL), a temperature sensor and a high resolution pressure sensor such that during the evaluation of the tightness test the technicians have access to continuous recordings of other crucial parameters recorded near the interface depth. The SoMIT tool is fixed in place using a pressing device such that the tool can be clamped into place in a position on the inner casing just above the interface from which it can monitor the interface (Fig. 8).

Minimum detectable leak rate General In the ISB method, the quantitative verification of the tightness of the well or of any possible leak rate is derived from a determination of the mass (and not the volume) of the test gas enclosed in the well at various points in time: the mass difference divided by the time period gives the theoretical leak rate. The mass of gas enclosed can be calculated based on the standard gas equation: pV ¼ mZRT

(1)

m ¼ pV=ZRT

(1a)

or

where

m, gas mass p, gas pressure R, gas constant T, gas temperature V, geometrical volume Z, compressibility factor The difference in the enclosed gas mass Dm used to determine the theoretical leak rate over the measurement interval Dt is gained by rearranging the equation and assuming a number of simplifications (Z-factor and gas temperature are constant) giving: Dm ¼ ( pDV þ VDp)=(ZRT)

(2)

Since the pressure, the temperature and the geometry vary with depth, the well needs to be divided into vertical segments to which P constant values are assigned. The total gas mass m is then derived by summation of the individual masses mi.

Estimate of errors

Fig. 7. The SoMIT tool (diameter: 50 mm; length: 2.3 m without cable head and sinker bars).

The basic procedure when determining the errors of measurement or the theoretical accuracy when determining the various gas masses is independent of the tool chosen to determine the interface depth

136

H. VON TRYLLER ET AL.

Fig. 8. Arrangement of the pressing device for the SoMIT tool and the separation of sinker bars.

or its theoretically possible precision. The principal parameters when determining the errors are as follows: A dl p Dp dp dDp

cross sectional area of gas –brine interface accuracy of interface depth detection test pressure pressure change during test period error of pressure measurement error during determination of gas pressure change Dp R gas constant T gas temperature dT error of temperature determination V geometrical test gas volume DV gas volume change at interface dV error of volume determination dDV error during determination of gas volume change DV Z gas compressibility factor dZ error of compressibility factor determination. The maximum error (minimum detectable leak (MDL) in terms of gas mass difference) resulting from the individual variables in the overall result is calculated as the sum of the absolute values of the partial derivatives of the determinant expression in equation (2); for details see Crotogino (1995). Some of these variables are functions of other measured values, e.g. volumes V and DV are dependent upon measured lengths and diameters. Ultimately however, the accuracy of the overall result depends both on the accuracy of the instruments as

well as on the well geometry and the test parameters such as the maximum measurement error; the accuracy cannot be determined generally but has to be calculated on a case-by-case basis. In the following, the influence of the SoMIT tool on the theoretical accuracy of the ISB method (the minimum detectable leak rate, MDLR) is explained using a representative example. The values calculated for any individual case do, of course, depend on specific circumstances. Figure 9 presents the influence of the achievable accuracy during interface measurement on the overall minimum detectable leak rate (MDLR); the parameter in this case is the actual surface area of the gas–brine interface in the cavern neck. As predicted: † the MDLR increases with increasing errors in determining the interface depth; in the example shown the area of the gas–brine interface is

Fig. 9. Minimum detectable leak rate (MDLR) versus accuracy of interface detection.

TIGHTNESS TESTS FOR SALT CAVERN STORAGE

137

The advantage (in this case a factor of four) grows with increasing area of well cross-section. In practice this means the SoMIT method becomes preferable as the well configuration becomes less favourable, i.e. as often found in cases with older wellbores.

Summary and outlook Fig. 10. Minimum detectable leak rate (MDLR) versus test duration.

A ¼ 1 m2, producing, for example, a value of 3.8 kg/day when using the SoMIT probe (accuracy of interface depth detection is 0.01 m) and 21 kg/day when using standard gamma-gamma probes (accuracy of interface detection is 0.1 m), i.e. a factor of approximately five; † the ratio changes in favour of the SoMIT probe, the larger the area of the interface: when the interface has an area of 3 m2 the factor by which the accuracy of the MIT method is improved by using the SoMIT tool rises to eight. Figure 10 illustrates the influence of the test period on the measurement accuracy when determining the theoretical leak rate. The leak rate is calculated by dividing the minimum detectable leak (difference of enclosed gas mass at two points in time) by the time interval Dt. The error in recording the time is found to be negligible. The graph of the accuracy in determining the leak rate for an example well and completion design shows the plot of a measurement using the SoMIT tool, with a second plot using a gamma-gamma probe. Under the same accuracy demands, the SoMIT tool achieves the required results within a much shorter measurement period with corresponding reductions in costs for equipment and personnel.

The SoMIT method generates considerably greater levels of accuracy when performing mechanical integrity tests using the frequently employed In-Situ Balance method, without at the same time raising the benchmark for test equipment or costs. At the same time the SoMIT method, unlike the spot calculations of the gamma-gamma measurement, allows quasi-continuous recording of the relevant parameters interface depth, temperature as well as differential pressure and hence continuous calculation of the theoretical leak rate. This information flow allows the early detection of trends, e.g. in falling leakage rates, which can be extrapolated to predict leakage rates and hence allow drastic reductions in test duration.

References E DLER , D., S CHELER , R. & W IESNER , F. 2003. Real time interpretation of tests investigating the tightness of the cementation of final cemented casings in gas storage cavern wells. SMRI Spring 2003 Meeting, 27–30 April 2003, Houston, TX, USA. C ROTOGINO , F. 1995. SMRI Reference for External Well Mechanical Integrity Testing/Performance, Data Evaluation and Assessment. SMRI R&D Project 95-0001-S 1995, Solution Mining Research Institute (SMRI). G ROSSWIG , S. & H URTIG , E., ET AL . 2003. Mechanical integrity testing using the fibre optic temperature sensing technique. SMRI Fall (5– 8th October) 2003 Meeting, Chester, United Kingdom.

Environmental issues in permitting gas storage: The Wild Goose case history LAURIE MC CLENAHAN HIETTER1,2 1

Previous address: MHA Environmental Consulting, Inc. (MHA)

2

RMT, Inc., 4 West Fourth Avenue, Suite 303, San Mateo, California USA 94402 Corresponding author (e-mail: [email protected]) Abstract: With recent fires and explosions at Moss Bluff, Texas (2004) and Hutchinson, Kansas (2001), initial environmental review is becoming more critical for public acceptance of new underground gas storage facilities; the environmental review is especially important for gas storage fields that use depleted oil and gas fields. Components of successful environmental review and permitting include an experienced and capable project manager and technical team, sufficient and thorough data for use in understanding affected resources, monitoring data for analysing well conditions and a holistic approach to public relations. The Wild Goose Gas Storage Field, located in Butte County, California, was developed originally in 1999, using a depleted gas field. Its initial storage capacity was 396.5 million cubic metres (Mcm, or 14 billion cubic feet, bcf). A subsequent expansion to 821.3 Mcm (29 bcf ) and construction of a 41.8 km long gas pipeline were subject to a thorough environmental review and permits were approved without opposition. An interdisciplinary team experienced in environmental review of oil and gas fields successfully completed a comprehensive environmental review of the Wild Goose Gas Storage Field expansion. Keys to successful environmental review and permitting for the Wild Goose expansion project included assembling a qualified and independent technical team, identifying potential local emission sources, understanding well integrity, defining vertical and lateral containment of the gas field and developing a thorough understanding of affected parties’ concerns and the affected local environment. The thorough approach to environmental review undertaken in the Wild Goose Gas Storage Field project can be used as a model for the permitting process applied to other gas storage fields, serving to shorten schedules, reduce project costs and lessen risk when considering new underground storage facilities.

The Wild Goose Gas Storage Field, owned by EnCana Corporation, was initially developed in 1999. An expansion was proposed and this paper describes the approach to environmental review and the key aspects of the environmental and permitting programme that can lead successfully to obtaining permits for underground gas storage projects. The lessons learned during the successful and unopposed permitting of the Wild Goose expansion project can be applied to other underground gas storage projects throughout the world. The issues addressed include:

(1) description of the project; (2) overview of environmental impact assessment and review process; (3) environmental issues; (4) mitigation and monitoring; (5) public participation; and (6) keys to success in permitting.

Wild Goose Gas Storage Field project description Initial project The California Public Utilities Commission (CPUC) initially granted Wild Goose Gas Storage, Inc. (Wild Goose) a permit on 25 June 1997 to develop, construct and operate an underground natural gas storage facility in Butte County, California (Fig. 1) and to provide firm and interruptible storage services at market-based rates. Firm service describes natural gas service that is continuously available and that will not be disrupted in times of heightened demand. Interruptible service describes natural gas service that may not be continuously available but that may be disrupted in times of heightened demand. The approval entitled Wild Goose to: (1) construct and operate (including the injection and withdrawal of natural gas) a new well

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 139–148. DOI: 10.1144/SP313.9 0305-8719/09/$15.00 # The Geological Society of London 2009.

140

L. M. HIETTER

Fig. 1. Wild Goose Gas Storage project location.

pad above the depleted Wild Goose Gas Field, as shown by the project storage area and wells in Figure 2; (2) construct a bi-directional pipeline (loop pipeline) from the well pad to a new remote operating facility; and (3) construct a remote operating facility (remote facility site) from where all storage field operations could be managed and monitored. Initial development and construction of the Wild Goose project was completed in April 1999. The Commission’s initial approval of the project authorized the use of one of the twelve gas storage zones (L4) in the field (Fig. 3) and the construction of a 41.8 km long gas pipeline. Zone L4 is authorized for a maximum storage of 396.5 Mcm of natural gas. The initial project approval also limited the daily injection and withdrawal of gas into and from the field to 2.27 million cubic metres per day (Mcmd) and 5.66 Mcmd, respectively.

Expansion project The proposed project was designed to expand the permitted storage capacity at Wild Goose Gas

Storage Field from 396.4 to 821.3 Mcm (14–29 bcf), and increase daily injection and withdrawal rates to 12.7 Mcmd and 19.8 Mcmd (450 Mmcfd and 700 Mmcfd), respectively (Table 1). Four project components were proposed to expand storage capacity and increase injection and withdrawal rates: (1) expansion of the existing well pad site; (2) construction of a second storage loop pipeline; (3) expansion of the remote operating facility site; and (4) construction of a pipeline to connect the field to one of Pacific Gas and Electric (PG&E) Company’s main transmission pipelines (Line 400/401) and an interconnect facility at Delevan on the west side of the upper Sacramento Valley. Well pad site. The existing well pad at the Wild Goose Gas Field was expanded to provide for the added storage and injection/withdrawal capacity. Expansion of the well pad was designed to accommodate drilling of up to 16 new wells. The new wells would be used for injection/withdrawal and observation and would be drilled into the reservoir

WILD GOOSE CASE HISTORY

141

Fig. 2. Wild Goose Gas Storage area and wells.

zones L1, U1 and U2 (Figs 2, 3). The well pad expansion would displace approximately 0.57 hectares (c. 1.4 acres) of wetland and would require up to 19 877 cubic metres (26 000 cubic yards) of

Fig. 3. Gas storage reservoir cross section.

structural fill material and 765 cubic metres (1000 cubic yards of soil for elevation of the well pad site and construction of a perimeter berm that would protect the pad from winter flooding.

142

L. M. HIETTER

Table 1. Wild Goose maximum storage, injection and withdrawal limits

Storage Injection Withdrawal

Existing

Proposed

396.5 Mcm (14 bcf) 2.27 Mcmd (80 Mmcfd) 5.66 Mcmd (200 Mmcfd)

821.3 Mcm (29 bcf) 12.7 Mcmd (450 Mmcfd) 19.8 Mcmd (700 Mmcfd)

Storage loop pipeline. The addition of a second 18-inch (0.46-metre) diameter bi-directional loop pipeline to transport the additional gas volumes between the reservoir and the remote facility site was proposed. A fibre optic cable would be installed with the pipeline. Both the pipeline and cable would be installed under the same right-of-way as the existing loop pipeline. Remote facility site. The added capacity of the Wild Goose reservoir would require expansion of the remote facility site. The site serves as the operational base for the Wild Goose facilities and houses the equipment required to receive gas from the local utility (PG&E) pipeline system, to inject and withdraw gas from the reservoir and to prepare it for reintroduction into the PG&E system. The project would add three additional natural gas-fuelled engines with three additional compressors. Expansion of the remote facility site would include: (1) an increase of the land use area by 2.35 hectares to a total of 4.81 hectares (5.8–11.9 acres); (2) three additional natural gas-fuelled engines and compressors producing a total of up to 10.7 mW (mW, or 10.7 horsepower); (3) up to six additional produced-water storage tanks with a total capacity of 757.1 cubic metres (200 000 gallons); (4) dehydration units and reboilers; (5) natural gas coolers; (6) a relief vent for pressure relief from the compressor station piping; (7) a new 3.78 cubic metres (1000 gallons) glycol supply/drain tank; and (8) a standby generator. Line 400/401 connection pipeline and Delevan interconnect facility. Gas would be conveyed to and from the Wild Goose facilities from PG&E’s Line 400/401 gas transmission pipeline, which runs in a north–south direction along the west side of the upper Sacramento Valley. The proposed pipeline, which would be up to 36 inches (0.91 metres) in diameter, would connect the remote facility site to the PG&E’s Line 400/401 pipeline at the Delevan Compressor Station. Two fibre optic communication cables, one primary and one backup, would be installed in the pipeline trench to allow remote

operation of valves and data acquisition (such as pressures and flows to monitor field operation) by the project applicant. A new interconnect facility with valves, metering and pressure monitoring equipment would be constructed adjacent to PG&E’s Delevan Compressor Station. The Delevan interconnect facility would consist of a level area covered with gravel and with a small pre-engineered metal building to house instrumentation electronics and monitoring equipment.

California Environmental Quality Act review Environmental laws The California Environmental Quality Act (CEQA) is a state law requiring state or local agencies to consider the environmental effects of their actions, such as implementing a project or issuing a permit to a project developer and to disclose the nature of projects and the environmental effects of the projects. CEQA was modelled after the federal National Environmental Policy Act (NEPA), which has similar requirements for federal agencies. NEPA applies when a project proposed by a private developer is located on federal land, or if a project proposed by a developer or another government agency receives federal funding. A federal project on federal land located in California requires CEQA review if the project requires an air quality, water or waste discharge permit from a state or local agency. If a federal project is planned on nonfederal lands, CEQA review may be required. CEQA applies to projects proposed by agencies or applicants if the project is not located on federal land and requires state or local agency permit approval. The Wild Goose project required a permit from a state agency, the California Public Utilities Commission. CEQA takes NEPA requirements several steps further in environmental protection and disclosure. The key goal of both CEQA and NEPA is to inform decision-makers and the public about: (1) the proposed project or action; (2) the setting where the proposed project would be located; (3) the physical environmental effects of the proposed project;

WILD GOOSE CASE HISTORY

(4) mitigation measures that would reduce or avoid significant environmental effects; and (5) alternatives to the project that would reduce or avoid significant environmental effects.

Environmental documents CEQA requires the preparation of an environmental impact assessment document, known as an Environmental Impact Report (EIR), which identifies the environmental effects of a proposed project. An EIR was required for the Wild Goose Gas Storage expansion project. The primary components of the EIR required under CEQA art points 1–5 of the previous section which are described in more detail below. Introduction. The introduction to the EIR provides a summary of the project, the environmental review process, the results of public and government agency consultations and the list of required permits. Project description. The project description defines all project elements. Topics addressed include construction methods and equipment, construction workforce, schedule, operational procedures (water requirements, air emissions, waste discharge, etc.), maintenance and decommissioning. Also included are all components to be built as a result of a project such as roads, electrical transmission lines or pipelines.

143

If effects exceeding significance thresholds could occur, then mitigation must be defined, if possible, to reduce or avoid such occurrences. Mitigation measures must be enforceable and verifiable to ensure that significant impacts are avoided. Well-written mitigation measures clearly identify the action to be taken, when and by whom. Alternatives to the proposed project. Alternative projects or alternative elements of a project, such as design or operation, should be defined if significant environmental effects could result from implementation of the proposed project. Alternative projects should be designed to reduce or avoid such effects and may include designation of alternative locations, reduction in project size, or variation in construction and/or operational procedures.

Public notification and disclosure A primary aspect of CEQA is providing public notification to government agencies and members of the public who may be affected by a proposed project or who may have regulatory authority over a proposed project. CEQA mandates public notice that an agency is initiating preparation of an EIR and a public review period of environmental documents for 30 to 45 days.

Wild Goose environmental analysis

Existing environment. The existing environment section describes the existing physical and human environment that may be affected by project construction and operation. The section describes geology and soils (including geological hazards), hydrology and water quality, vegetation, wildlife, air quality, prehistorical and historical cultural resources, hazards and hazardous materials, noise, visual resources and aesthetics, land use (including agricultural), recreation, demographics, housing and traffic.

Approach to environmental review

Environmental impacts and mitigation measures. The environmental impacts and mitigation measures sections include a description of the environmental effects that will occur if the project is implemented. The review focuses on the significant environmental effects of the proposed project (i.e. effects that exceed established significance thresholds, such as state or federal ambient air quality or water quality protections standards). The environmental documents identify and focus on the significant environmental effects, unavoidable significant environmental effects, significant irreversible environmental changes and growthinducing impacts, such as increased urbanization, that may result from a project.

Interdisciplinary team. EIRs address many environmental disciplines and social issues. CEQA practitioners include planners and scientists. The project manager for preparation of an environmental document could be either a planner or scientist, depending on the type of project to be addressed. Experienced project managers usually have a working knowledge of all of the disciplines addressed in the EIR and are supported by scientists, planners and technical writers with varying levels of experience of the topic or section of the EIR that they author. The California Public Utilities Commission was the lead agency to conduct environmental review under CEQA. The Commission selected MHA

The key elements in the approach to environmental review of the Wild Goose project focused on using a very experienced interdisciplinary team to prepare the environmental analysis, lead public and agency consultation, prepare a thorough definition of the project and existing environments (baseline conditions) and define impacts and mitigation strategies related to the issues of migrating gases, well integrity and monitoring.

144

L. M. HIETTER

Environmental Consulting, Inc. (part of RMT, Inc. since 2007) to prepare an EIR addressing the environmental effects of construction and operation of the Wild Goose expansion project. MHA assembled an interdisciplinary team of experts to analyse the effects of the project and prepare the environmental document. To address geological issues and potential hazards, the team included several geologists with considerable experience in both oil and gas exploration and gas storage projects. The team also included scientists and planners able to address issues relating to areas of biology, hydrology, air quality, agriculture, cultural resources, land use and public relations. Scoping and consultation. One of the key goals of CEQA is to notify and inform the public of pending discretionary decisions to be made by a public agency. CEQA requires that a Notice of Preparation be published to notify the public and involved agencies that the lead agency intends to prepare an EIR. The Notice of Preparation describes the project, the lead agency, the permit applications that will be considered, the types of impacts that may occur and the date and time of the scoping meeting. The scoping meeting is a public meeting that is held to introduce the proposed project to the local community and agencies and to solicit public comment on the scope of the environmental analysis. A scoping meeting for the Wild Goose project was held early in the environmental review process to describe the project to the public and hear their concerns related to the project. The issues raised during scoping were then used to refine the scope of the environmental analysis. Several governmental agencies had regulatory permit approval authority over aspects of the Wild Goose expansion project. Permits were required by agencies such as the California Department of Fish and Game, the local air quality management district and the Regional Water Quality Control Board. Applicable regulatory agencies were contacted to solicit their concerns early in the process to avoid surprises in mitigation measures and other requirements. Thorough project definition. Defining the project is an essential element of a thorough and accurate environmental review of any project, and especially an underground gas storage field. In many cases, projects are not thoroughly defined, and can lead to uncertainty and misunderstanding by permitting agencies and the public. An incomplete project description may result in impacts being identified late in the process, which may cause schedule delays and increased costs. A gas storage project may include well construction and operation, storage field operation, gas

processing equipment, new pipelines and decommissioning. A thorough project description must have a complete definition of: (1) surface disturbance (total amount of area disturbed); (2) access (routes of access for construction and operation); (3) construction (construction equipment, manpower, schedule of operations, equipment storage and staging areas); (4) design (equipment layout, aesthetic presentation); (5) operational procedures (schedule of operation, equipment noise, operation, air and waste emissions, water use); (6) labour (size of workforce, source of workforce (local or imported); (7) decommissioning (methods for closing the field). Benefits to the project sponsor, regulatory agencies and the public from a thorough project description include: early identification of potential risks and environmental impacts; reduced permitting time; the making of better decisions by well-informed agency personnel and the public; reduced risk of project elements that were not expected by the agencies or the public, new concerns by agencies and the public and schedule delays that could be caused by analysis of unknown project elements or project redesign; and reduced liability. The Wild Goose project is located in an area of largely agricultural land. Construction of project facilities required removal of crops for pipeline construction. Important to landowners was an understanding of the pipeline routes, access road locations and facility locations and their ability to understand the effects the project would have on their crop management. It was necessary to include a description of the history of the field development, production and injection zones, well history and abandoned wells in order to assess the potential for gas to migrate to the surface and thereby to pose a risk of explosion or health effects. Baseline data collection. The purpose of a baseline data collection program is to ensure a thorough understanding of resources that might be affected in the proposed project area. Conducting environmental reviews under CEQA often includes site-specific baseline data collection for biological resources, hydrology and water quality, geology and soils, hazards, noise, air quality, public services and utilities. It is especially important to understand the baseline geological environment for underground gas storage fields in order to identify possible gas migration pathways and thus avoid potential safety hazards.

WILD GOOSE CASE HISTORY

The baseline data collection plan for the Wild Goose expansion project was established after conducting initial scoping and consultations and following the completion of a thorough project description identifying all components of the project. Baseline data was collected for biology, cultural resources, land use, geology and hydrology. Information regarding location of underground utilities was also collected to identify potential effects of construction and to identify potential horizontal gas migration pathways.

Environmental and safety issues of gas storage projects The environmental effects associated with drilling and producing oil and gas are well documented at oil and gas fields around the world. However, environmental effects associated with the longterm operation of underground gas storage fields are not as well studied and sometimes unique. Some of the critical issues, such as geological structures and migration pathways, may not be considered in environmental reviews of oil and gas drilling projects. The explosions and fires at other gas storage fields (e.g. Hutchinson, Kansas and Moss Bluff, Texas; see Evans 2009; Miyazaki 2009) prompt concern for anticipating potential future land uses and possible effects on health and safety in the area. Potential situations that may occur at either depleted oil or gas storage facilities include: (1) gas leaking along man-made pathways such as wells (abandoned or active wells) due to damaged casing, or leaking seals or utility pipelines; (2) migrating gases that travel along natural pathways such as faults and fractures; (3) emissions of hydrogen sulphide, methane, carbon dioxide and other thermogenic gases; and (4) both chronic and acute health effects associated with exposure to emissions of gases such as hydrogen sulphide, methane, benzene and toluene. Table 2 identifies some of the gases of concern that may be present in underground gas storage fields and the potential effects the emission of these gases have on human health. Some of the key issues relating to the safety of underground gas storage fields that must be assessed include: integrity of the storage field caprock; integrity and safety of wells in the region; potential pathways for migrating gases; storage zones and confining layers; shallow gas zones or aquifers; potential for fires and explosions; identifying the locations and integrity of old and abandoned wells.

145

Table 2. Underground gas storage gases and health effects Potential migrating gas Hydrogen sulphide Methane Benzene Toluene

Health effects Neurological damage Asphyxiation Cancer Birth defects

Defining baseline conditions is important when assessing potential impacts. Both the operator and regulatory agencies need to determine the potential environmental effects of a gas storage project in order to avoid costly future retrofits or accidents. For economic and safety reasons, geological structures and potential gas migration pathways must be delineated and analysed. Understanding the local and regional geology and ground conditions is vital to identifying potential gas migration pathways through conduits such as faults and fractures, active and abandoned wells and pipelines. A thorough review of possible project impacts allows for early project modifications or mitigation measures to reduce the risk of potentially significant impacts on the environment and the public. The Wild Goose project proponent’s well construction and completion data were reviewed to confirm conclusions regarding storage and safety. The geological review for the Wild Goose expansion project included the identification and analysis of faults in underlying formations, but also incorporated: regional and local stratigraphy, natural gas reservoirs and production (storage gas, biogenic and thermogenic gas and active and abandoned wells), mineral resources and extraction (nearby sand and gravel mining), soil characteristics (stability, contamination, potential for liquefaction, expansion and compaction) and geological hazards (such as faulting and seismicity including maximum credible earthquakes, subsidence, groundshaking, liquefaction, lateral spreading and shallow groundwater). To conduct an analysis of potential hazards, it was necessary to understand the existing conditions. Thus the key areas addressed in the Wild Goose expansion project EIR were: the hazardous materials stored and used at the existing underground gas storage facility, the hazardous wastes in the area, the natural gas storage field and well conditions, the integrity of abandoned wells and dry holes, the natural gas pipeline hazards due to local and regional faulting and project safety and inspection programmes. Construction impacts and mitigation measures. The effects of construction are commonly addressed

146

L. M. HIETTER

during the preparation of an EIR. The degree of surface disturbance related to project construction is often a key element of concern. Other construction concerns can include the effects of air emissions from drilling and traffic and the effects of construction traffic (noise, disruption etc.) on the community. The effects of construction associated with the Wild Goose expansion project were considered in the environmental review. Some of the issues of concern from the construction stage of the Wild Goose project included the effects of storage facility and pipeline construction on agricultural operations, biological resources, air quality and the Sacramento River. The proposed project is located in an agricultural area with rice fields, row crops (e.g. carrots, corn and cotton) and cattle grazing. Building a new well pad, remote facility and pipeline involved construction activities within a specified footprint on agricultural land. Impacts of concern included constructing temporary dykes in rice fields, restoring appropriate grades in the fields following pipeline construction, temporary loss of crops and potential interference with harvesting. The project was designed to minimize the effects on agricultural operations. The environmental impact assessment resulted in defining mitigation measures that included notification of farmers and ranchers, temporary fencing during construction and compensation for arable land removed from production by construction or operation. The project area also includes a wildlife refuge and construction noise can affect birds during the breeding season. A mitigation measure required that construction activities avoid the nesting season. Other sensitive biological resources in the area included wildlife species with special federal or California state regulatory status. Mitigation measures were specified to delineate and avoid sensitive habitats and to replace affected habitat. Biological monitors were used during construction to ensure that there were no adverse effects to sensitive species. The pipeline route crossed the Sacramento River, and was to be achieved by the drilling of directional boreholes, taking the pipeline beneath the river. However, boring has the potential to increase turbidity in the river, which could affect sensitive fish species if there is an unexpected situation. Such an event might be a ‘frac-out’, which can arise when the bedrock surrounding the borehole is fractured, allowing drilling mud to migrate into the stream water. A mitigation measure specified that the boring should take place outside the fish-spawning season in order to minimize any potentially adverse effects on the fish population.

Construction activities have the potential to contribute significant amounts of particulate matter (dust) to the atmosphere, which may contravene state and federal ambient air quality standards. As mitigation, dust control measures and a 24 km/ hour speed limit on unpaved roads were specified to reduce dust emissions. Operation impacts and mitigation measures. Various operation impacts must be considered in analysing the effects of a project. An issue of concern for gas storage fields is increased pressure in gas storage zones, which occurs when depleted oil and gas fields are used for gas storage. Gas pressure is often reduced as gas and oil are produced from reservoirs, and is increased when the field is used for storage. Higher gas storage field pressure may increase the potential for shallow (thermogenic) or deeper storage gases to be emitted through a number of different pathways, such as natural fractures or faults, other geological structures including formation boundaries and landslides, old or recent wells with compromised annular seals or casing and pipelines. Monitoring pressures and injection and withdrawal rates can help to identify any issues related to leaking gases. Avoiding leakage or migration of gases and delineation of any potential gas migration pathways requires evaluation of the integrity of all old wells in the gas storage area. If necessary, and to avoid unintended gas emissions, poor well completions must be remediated. The effectiveness and gas tightness of the caprock is also important to maintain the safety and integrity of the gas storage field. The interdisciplinary team assembled for the Wild Goose expansion project documented and evaluated baseline data identifying and considering potential operational impacts, including impacts to nearby residents. The Wild Goose project is located in a sparsely populated rural, agricultural environment. In certain areas, however, the land use could change with time to support a higher population density. Two central issues analysed that are unique to underground gas storage fields included the effects of increased pressure in gas storage zones (beyond those in existence during oil or gas discovery and during the initial development and production of the field) and the potential for vertical and lateral gas migration caused by increased pressure in the formation during gas injection. Other important issues also assessed included surface subsidence or uplift and deformation and the potential to affect the precisely levelled agricultural fields used for growing rice. Mitigation measures were designed to address potential impacts from ground surface movements that could affect the rice fields.

WILD GOOSE CASE HISTORY

Core tests were recommended for the Wild Goose field to evaluate the competency of the caprock. Soil gas monitoring around abandoned wells, well testing and monitoring were also recommended to detect possible gas leaking from wells. Potential health effects from the emissions of compressors and other project facilities were also addressed. Long-term monitoring and mitigation. When the effects of projects are uncertain, monitoring may be prescribed with mitigation defined if certain outcomes are identified. Mitigation measures that include monitoring must also include data evaluation and reporting. This ensures that mitigation is implemented if significant effects are identified. In addition to preconstruction mitigation measures for the Wild Goose project that specified caprock and well integrity testing and evaluation, the environmental analysis resulted in defining several long-term monitoring and mitigation requirements for geology, hazards and wetlands. Mitigation monitoring and reporting measures were developed during environmental review after defining and analysing existing geological, soil and hazard conditions. A mitigation measure specified that well inspections, testing and leak surveys be conducted at the end of each gas injection cycle to evaluate the integrity of abandoned wells. The mitigation measures specify that if gas is detected, samples are to be collected to determine the origin of the gas (biogenic, thermogenic, or storage gas). If the gas proves to be injected gas, then implementation of remedial actions is required to address any leakages. A documented southward-dipping normal fault is present along the southern edge of the gas storage field and could potentially act as a conduit for the escape of storage gas. A requirement is that soil gas probes are to be installed and monitored in the areas overlying this fault in order to detect as early as possible any escape of gas along the fault or through well bores. Any gas found in the probes should be analysed for source (biogenic, thermogenic, or storage gas). If storage gas is detected in these monitoring probes, operation at the field would be evaluated to redefine storage boundaries and possibly reduce allowable storage volumes to prevent storage gas from reaching the fault zone. The project was located in environmentally sensitive areas for various species of plants and wildlife, wetlands and agricultural resources. The project required the creation of wetlands to replace those lost during construction and monitoring to document the effect construction of facilities had on wetlands. A monitoring plan was designed to ensure the successful establishment of the new wetland vegetation.

147

Public and agency consultation Early consultation with the public and regulatory agency staff is critical to successful and timely permitting. Key steps in the public and regulatory affairs programme were: (1) identifying all affected parties, both government agencies and the public; (2) consulting with the parties early and often; (3) identifying and resolving issues early; (4) conducting information workshops; and (5) collaborating with parties to reduce or avoid environmental effects. In the Wild Goose expansion project, implementation of these steps by the applicant’s representatives and the environmental review team led to a successful public relations effort. Results of the consultation programme and thorough environmental impact assessment were that the schedule for permit approval was met, limited comments on the EIR were filed with the CPUC, permitting costs were reduced, permits were issued with no unexpected conditions and no appeals or litigation arose.

Conclusions The Wild Goose project is an example of successful environmental review and permitting of a natural gas storage field expansion. Thorough environmental impact assessment and review is a benefit to the project applicant because it can expedite the permit approval process and assist the applicant in gaining public consent. There were four key elements to success. Qualified project team. The interdisciplinary project team for the Wild Goose expansion project was successful in conducting a thorough environmental review under the defined schedule because of depth of experience of the team members in this type of environmental review. The project director, project manager and team geologists all had previous experience in working on gas storage fields. This resulted in the identification of data needs early in the process, which allowed for a timely environmental review. Potential hazards and risks were identified and addressed, which reduced the risk and potential liability to the project owner. Thorough project description and comprehensive data collection. Enables permitting agencies and the public to understand a project fully, hopefully guarding against doubts about the project. An incomplete project description may result in late identification of project impacts and may lead to delays in the schedule and increased costs.

148

L. M. HIETTER

The environmental review should begin with a comprehensive baseline data collection and analysis programme, designed to assess biological resources, geological conditions, well integrity, potential local gas migration pathways and identification of local emission sources. Once the project was defined and baseline conditions established, the Wild Goose environmental review process was completed very quickly, given the complexity of the project and the permitting process. Focusing initial efforts on defining the project allowed rapid preparation of the draft environmental document and completion of the permit approval process. The thorough project description and complete baseline data expedited the environmental analysis process and reduced the number of comments received on the draft environmental impact assessment document. Fewer comments meant that the process could be completed faster and with less expense. As a result, the environmental document was completed within three months from the time when the application was deemed complete. Comprehensive data analysis and proper monitoring. A complete analysis of the geological structures, gas field conditions and well integrity is an important component of a thorough environmental impact assessment of an underground gas storage facility. These elements allow for cost-effective field management by identifying issues of concern early in the project design phase. Resources are conserved and risks are avoided when potential hazards are identified and addressed early in the project. Delineating the baseline conditions ensures that future investigations can more easily identify whether issues are related to the project or to natural conditions. Thorough understanding of the field conditions can also shorten permitting and construction schedules and reduce construction costs by avoiding unanticipated problems during construction and operation. A fully developed understanding of potential risks and hazards allows better field management and helps avoid accidents that might

lead to gas leakage. These elements all serve to reduce financial risk and liability. Ongoing monitoring and data analysis ensures continued safe field operation. If proper monitoring is not undertaken, the result might be leakage of storage gas, odours, risk of explosions, negative health effects and financial loss. Coordinated approach to public relations. Lack of information, either about the project or environmental effects, is often a primary cause of delays in project permitting. A complete environmental analysis provides the decision makers and the public with the necessary information to make informed judgments. Agency staff and members of the public are usually more comfortable with projects when they believe that all environmental effects have been analysed thoroughly. The coordinated approach to public relations resulted in a successful permitting process at Wild Goose. Regular communication with landowners and agency personnel served to remove opposition and alleviate concerns about the project. The approach described here can be successfully applied to permitting underground gas storage fields throughout the world.

References E VANS , D. J. 2009. A review of underground fuel storage problems and putting risk into perspective with other areas of the energy supply chain. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 173 –216. M IYAZAKI , M. 2009. Well integrity: an overlooked source of risk and liability for underground natural gas storage: lessons learned from incidents in the United States. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 163 –172.

Underground gas storage project at Welton oilfield, Lincolnshire: Local perspectives and responses to planning, environmental and community safety issues MEG DAVIDSON 23 Beech Avenue, Nettleham, Lincoln LN2 2PP Corresponding author (e-mail: [email protected]) Abstract: This paper describes the research undertaken by two local councillors to assess the proposed underground gas storage (UGS) scheme at Welton oilfield, Lincolnshire and the implications for the local community. Compared to the US, where over 300 operational gas storage facilities in onshore depleted fields exist, this form of gas storage is in its infancy in the UK. This has proved to be a challenge for the various agencies involved in the planning process. The councillors have identified a number of planning and safety issues which merit further consideration, including the applicability of the control of major accident hazards (COMAH) regulations, well integrity and the effects of re-pressurization of the oilfield. The proposed UGS facility is in close proximity to residential areas and concern amongst local residents is considerable. The councillors believe that their research has shown that insufficient attention has so far been given to community safety issues. A full and thorough assessment of all potential risks associated with this particular type of UGS needs to be undertaken and appropriate safeguards put in place to protect the local community.

This paper presents an account of the proposed underground gas storage (UGS) project at the Welton oilfield in Lincolnshire, as seen through the eyes of two local councillors. They are members of West Lindsey District Council (WLDC) and represent wards in close proximity to the Welton oilfield. For the purposes of this paper they are referred to as the ‘ward councillors’. As a result of growing concerns from local residents about the proposed UGS scheme, and conflicting accounts of what this would entail, the ward councillors decided to research the subject of UGS and try to establish the truth about the safety implications for the local community. Their research has highlighted a number of important planning and safety issues which warrant further examination and which may be of relevance to the consideration of future UGS developments in the UK.

Description of the proposed UGS scheme History of welton oilfield The Welton oilfield is situated five miles north of Lincoln and, geologically, lies within the East Midlands Basin. There are currently six producing fields in the immediate area: Welton, Scampton South, Scampton North, Stainton, Nettleham and Cold Hanworth. There is a substantial gathering and processing facility adjacent to Welton oilfield.

The oil is transported from there by train to third party storage facilities at Immingham (Fig. 1). Oil was discovered in 1981. After further exploratory work, BP, the original owners, submitted a planning application to Lincolnshire County Council (LCC) in 1983. Planning permission was granted and production at the Welton oilfield started in November 1984. The current owners of the Welton oilfield are Star Energy, a UK based company established in 1999. Star Energy is the UKs second largest onshore oil producer, behind BP, with onshore UK producing assets in the Weald Basin of southern England and in the East Midlands.

Planning applications Star Energy submitted three planning applications in connection with the proposed UGS facility at the Welton oilfield in November 2003, as follows: † 14 wind turbines at various sites on the six oilfields; † construction of an 11 km gas pipeline from the Transco grid to the gathering centre adjacent to Welton oilfield; † gas storage for 56.6 million cubic metres (Mcm; 2 billion cubic feet [bcf]) gas and processing facilities, to be built on a 5.72 hectare site adjacent to the gathering centre. In the 12 months prior to submitting the planning applications for the Welton project, Star Energy

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 149–161. DOI: 10.1144/SP313.10 0305-8719/09/$15.00 # The Geological Society of London 2009.

150

M. DAVIDSON

Fig. 1. Map showing Welton oilfield, Lincolnshire and surrounding area.

WELTON OILFIELD, UK LOCAL ISSUES

devoted a considerable amount of money and effort to publicizing their wind turbine proposals. This attracted a lot of attention and much media interest. The first mention of the proposed pipeline and gas storage facility did not come until November 2003 when Star Energy submitted its planning applications to LCC. At the same time, the company sent a newsletter to 15 000 households in the area surrounding the Welton oilfield, informing residents for the first time about the gas storage project. It was placed firmly in the context of the much publicized wind turbine proposals, ‘the gas storage project is being planned alongside previously announced proposals for a cluster of wind turbines on or close to Star’s network of small oil production sites in the area’ (Star Energy 2003). During 2004, Star Energy first reduced the number of wind turbines and later withdrew this application altogether. The reason was never made clear. Local residents speculated that the Ministry of Defence had raised objections. There are several RAF airbases in the area, and wind turbines can cause problems with radar systems (Adam 2004).

Local community reaction By the end of 2003, public awareness of the wind turbine proposals was high. A few people knew about the application to build the gas pipeline, but hardly anyone was aware of the application relating to gas storage. Local parish councils had received details of the application for the wind turbines and the pipeline, but only Sudbrooke and Reepham Parish Councils received details of the gas storage application. The proposed extension to the gathering centre for gas processing facilities lies within the boundaries of the two parishes: theoretically other parishes in the vicinity were not affected so they were not notified of this particular application. Sudbrooke Parish Council believed that the gas storage application was of concern to a much wider area and in January 2004 called a meeting for members of neighbouring parish councils and local district and county councillors. At this meeting a local action group, formed the previous year to oppose the wind turbines, gave a presentation focusing on the gas storage proposals. A member of the group, an engineer with 30 years experience in the oil and gas industry, informed the meeting about the UGS proposals. He claimed that using a depleted onshore oilfield for UGS was an untried and untested technique in the UK and outlined his serious concerns about the proposed scheme. Reference was made to a catastrophic blow-out of ‘sour’ gas that had occurred in a gas field in China a few weeks earlier (see below). The speaker expressed fears that a major disaster could occur at the Welton oilfield should the proposed scheme go ahead.

151

This had a profound impression on everyone present. To that point, hardly anyone had been aware of the extent of the oilfields, or that they were ‘sour’ oilfields. No one had encountered the concept of onshore UGS before. It was decided that the local county councillor together with district colleagues should speak with officers at WLDC, LCC and the local MP to brief them about these proposals and the concerns that had been expressed. It was agreed that LCC planning officers should be urged to obtain external independent advice, especially on geological issues. As the use of a depleted onshore oilfield for UGS had at that time never been attempted before in the UK, there was some doubt as to whether agencies such as the Health and Safety Executive (HSE) and the Environment Agency (EA) would have the necessary knowledge and expertise. Other residents, not involved in the action group, but with experience in the oil and gas industry, were by this time also expressing alarm at the prospect of a sour oilfield being used for UGS. Many people including planners, council officers and officials from various government agencies had not heard of the proposals and when informed about them, saw no significance in them. Others had some awareness of the plans to store gas in the Welton oilfield but assumed it did not represent a change in its existing functions and operations. Frequent comments included, ‘there’s gas there anyway’ and ‘they’ll only be storing gas where gas has been for millions of years’. Others, including some local councillors and business representatives, acknowledged that the idea of using an oilfield to store up to 424.7 Mcm (15 bcf) of gas did, on the face of it, sound risky. However, they believed that the proposed scheme must be safe, otherwise it would not have been put forward. ‘Surely they would not allow it if it weren’t safe’ was a commonly expressed view. Residents living in the area around the oilfields mostly reacted with horror at the thought of vast quantities of gas being stored under their homes. ‘How will they know exactly where the gas will go?’ and ‘Isn’t there a risk of leaks and explosions?’ summed up many concerns. These were not unreasonable questions. Official agencies can be rather dismissive of public concerns. The prevailing attitude to local residents and the action group was to assume that they were ‘neurotic NIMBYs’ (not in my backyard) whose main concern was the value of their houses. In fact, the ward councillors found that a genuine fear of danger was uppermost in local residents’ minds. As one resident of Sudbrooke, the village closest to the oilfield said, ‘I’m not worried about the price of my house: I’m worried that I might not have a house left standing’.

152

M. DAVIDSON

Ward councillors’ research The ward councillors were present at the meeting organized by Sudbrooke Parish Council in January 2004 and they immediately realized that the proposed UGS scheme was potentially of huge significance for residents in the wards they represented. Two very different views on the safety of the scheme had been presented. On the one hand, Star Energy had stated in their community newsletter that storing gas in this way did not pose a danger. In contrast, a man with 30 years professional experience in the oil and gas industry, who appeared to know what he was talking about, was warning that the scheme could be a ‘disaster waiting to happen’. Neither of the ward councillors had previous knowledge of UGS and had limited knowledge of the existing oilfield operations. One of them had a background in business and finance and the other in biomedical sciences research. They decided to draw on their skills and experience to investigate the subject of UGS and to try to establish the facts on the potential risks of a UGS facility in a depleted oilfield. Having been presented with such widely divergent views, they worked on the basis of not accepting any opinion unless it could be corroborated. The first claim the ward councillors quickly verified was the incident at the Chinese gas field that had resulted in an escape of sour gas. There had been a gas well blow-out in Gao Quiao, Chongquing, China on 23 December 2003, when 243 people died, 9000 were injured and 64 000 were evacuated from their homes (APELL 2004). This incident occurred during gas exploration: this was not a UGS facility. The accident was due to negligence and failure to observe safety procedures (People’s Daily 2004). However, the information proved two important points: first, the action group’s account of this incident had been correct and second, it demonstrated the horrific effects of sour gas. The ward councillors felt that their initial research and discussions had shown that, to use a legal analogy, ‘there was a case to answer’. The next steps were to find out more about UGS in general and to look at the actual planning applications and the 75 page dossier, ‘The Case Against Gas Storage and Gas to Electricity Generation in the Welton Oilfield, Lincolnshire’ which had been produced by the action group and submitted to LCC as a formal objection to the planning applications. The ward councillors learned that the US has over 300 UGS facilities in onshore depleted fields (EIA 2006), by far the highest number in the world. UGS is far less prevalent in Europe and very much in its infancy in the UK. US facilities are classified as salt cavern, aquifer or depleted

field (EIA 2006). Depleted gas fields are by far the most common type of storage: 75% of all gas storage facilities worldwide are in depleted hydrocarbon fields (Plaat 2004). There are no readily available figures for the number of UGS facilities in depleted oilfields: they are grouped with depleted gas fields under the heading ‘depleted field’. In the UK there are currently two onshore UGS sites operational in depleted fields, with a further eight planned, including the proposal to convert the depleting Welton Oilfield in Lincolnshire (refer Evans & Holloway 2009). The Welton UGS facility would have a capacity of about 425 Mcm and would make a significant contribution to the overall gas storage capacity in the UK. The ward councillors’ research led them to the website of OFGEM (Office of Gas and Electricity Markets; OFGEM 2005), which carries a section on gas storage. The Department for Business, Enterprise & Regulatory Reform (previously Department of Trade and Industry or DTI) website (DTI 2005) carries JESS (Joint Energy Security of Supply Working Group) reports. The Welton oilfield is listed on both websites as a proposed UGS site, with its storage capacity integral to the calculations for future UK gas supplies. As far as local councillors and residents were concerned, the project was just at the drawing board stage, whereas in fact it seemed to be a fait accompli long before the planning applications had been submitted. The ward councillors felt that a map showing the extent and location of the oilfields in relation to what is on the ground was an essential starting point for all concerned. However, no such map was available. The planning application included an Ordnance Survey (OS) map showing the proposed 11 km pipeline route, but it did not show the wider area around the processing centre or the extent of Welton oilfield. There was a sketch map giving an indication of the proposed gas storage area in the southern end of Welton oilfield. The application was for storage of 56.6 Mcm (2 bcf) of gas, but larger quantities of 283 Mcm (10 bcf) and 425 Mcm had been mentioned in other documents. It was not clear what effect the larger volumes of gas would have on the storage area. It was suggested that maps showing the extent of the oilfields would be confidential. This turned out to be incorrect. The DTI website (a particularly valuable source of information), carried a map showing all of the onshore oil and gas fields in the UK. It also had data on all the UK oil and gas fields and all oil wells, both used and disused. The DTI maps of onshore fields did not show their relationship to what is at ground level. After studying these maps, the ward councillors concluded that it should be possible to superimpose the oil and gas field data upon an OS map. This proved beyond the

WELTON OILFIELD, UK LOCAL ISSUES

capabilities of WLDCs mapping software and for a time it looked as if it was not going to be possible to produce the required map. Unexpectedly, a map using this exact technique was published in the Lincolnshire Echo, a local newspaper. The map, showing the approximate aereal extent of the six producing oilfields on an OS basemap, came from Star Energy along with a letter from their Chief Executive (Wessel 2005). The company had asked for the map to be printed to refute what they claimed was an inaccurate depiction of the oilfields published the previous week alongside a leading article on the proposed UGS project. That map showed a large area encompassing most of the villages north of Lincoln and resembling a giant lake. Star Energy’s letter stated, ‘The article of 13 June included a map purporting to illustrate ‘where the oilfield lies’. My map of proven oil-bearing strata by comparison can be verified against the DTI’s official website. The stored gas will lie within a small portion of the Welton oil-bearing strata away from residential centres’. The only drawback to the map was that the oilfields were depicted as solid shapes that obscured the underlying OS map. Fortunately, it was a simple process to depict the oilfields in outline instead. At last, there was a map showing with reasonable accuracy the location of the six oilfields and the two proposed UGS sites in the Welton and Scampton North oilfields. This confirmed that the Welton oilfield lies directly beneath the villages of Sudbrooke and Scothern and part of Scampton North oilfield lies underneath RAF Scampton, a large military airbase (Fig. 1). The saga of the map confirmed the ward councillor’s view that an accurate map showing the position of the oilfields in relation to the local area was important for everyone’s understanding of the proposals. It would have been better for such a map to have been included in the planning applications rather than issued in response to hostile comments in the media.

Possible safety issues There were a number of safety issues arising from the planning applications and the environmental statement, which the ward councillors felt needed further examination. These included: † The claim that there would not be any ‘adverse geological impacts resulting from the injection, storage and extraction of gas since the pressure will not exceed pressures historically exerted in the reservoir’. This is a very crucial assumption, but no supporting evidence was given. † The presence of H2S in the oil and gas was alluded to very briefly, but no mention was

153

made of the levels of H2S, which might be present or of its potentially lethal effects. A lay person (which in this context would include planning officers as well as members of LCCs Planning Committee) might be unaware of its significance. † The concept of cold venting in the event of an emergency shutdown was worrying, especially in view of the presence of H2S. Although such a scenario was envisaged as an ‘infrequent occurrence’, this was quantified as once in 10 years. This compares to flood risk where more than once in 100 years frequency is considered unacceptably high in planning contexts. The safety debate took a dramatic turn in May 2004 when Star Energy floated on the London Stock Exchange. The ward councillors were alerted to the company’s Admission Document (Star Energy 2004). It turned out be very informative, giving many details that were not present in the planning applications. It revealed that Star Energy was intending, eventually, to store up to 425 Mcm of gas (compared to the 56.6 Mcm mentioned in the planning application) and to develop a second UGS site in the Scampton North oilfield, with a pipeline linking it to the Welton oilfield. This came as a surprise, but it was the section on risk factors that attracted most attention. Under the heading ‘Exploration, production, gas storage and general operational risks’, it was stated: † ‘The business of exploration for production of oil and gas, and the conversion of existing oil and gas fields to gas storage involves a high degree of risk.’ † ‘Old well bores could leak and/or result in explosion when under the levels of pressure demanded of a gas storage facility.’ † ‘The use of depleted reservoirs to store gas also requires the processing of the gas withdrawn from the reservoir. Failure of the processing equipment could expose the Group to substantial liquidated damages.’ † ‘. . . there may be circumstances where Star Energy’s insurance will not cover or be adequate to cover the consequences of the events described above.’ This was in striking contrast to Star Energy’s assertions in its community newsletters published in November 2003 and July 2004 that ‘gas storage is safe’. The ward councillors, as did many others, felt that the information on risks outlined in the Admission Document confirmed fears that UGS in a depleted oilfield could potentially pose significant risks. It demonstrated why more work was needed to ensure that all potential risks were fully identified and assessed before the UGS scheme at Welton proceeded.

154

M. DAVIDSON

It later transpired that neither LCC planning officers nor HSE officials were aware of the contents of the Admission Document. However, even when it was drawn to their attention, neither felt that this information was relevant to the planning process or HSEs own risk assessment. Star Energy subsequently stated, ‘as is standard practice, we point out all of the conceivable risks, however remote, of the oil and gas production business and the gas storage business’. This seemed to imply that the risks outlined were not that serious and played down the importance of the Admission Document. In fact, statements made in an Admission Document to the London Stock Exchange carry considerable weight. Company directors are under a legal obligation to disclose all relevant facts. If a risk is mentioned in such a document, it has significance. It would appear that the planning system does not require the level of candour demanded by the London Stock Exchange. This illustrates a weakness in the system in that no single document contains all the available information making it difficult to obtain the complete picture.

Proximity of proposed UGS facility to populated areas Discussion with different agencies revealed that many people were unaware that the Welton oilfield is situated in a populated area. A cursory glance at a map might give an impression that the area north of Lincoln is all open countryside. In fact, there are approximately 2400 people living within a one-mile radius of the proposed gas processing centre and 15 000 people living within a 4.8 km (3 mile) radius (Fig. 1). The villages on the northern fringe of Lincoln have expanded considerably over the past 20 years, with some having doubled in size since the oilfields started production in the 1980s. Amongst the many documents relating to UGS found by the ward councillors was a 2004 House of Lords Select Committee report on UGS (House of Lords 2004). In written evidence to the Committee, the British Geological Survey (BGS) described the potential for using depleted oil and gas fields for UGS in the UK (BGS 2004). The BGS submission touched briefly on local community concerns. Under the heading ‘Barriers to deploying UGS in the UK’, BGS states, ‘Public perception is an issue in UGS . . . Opposition is likely to be at the local level, given that the main perceived risks are local (e.g. risk of fire or explosion from a leaking facility . . .)’ and, ‘There is no obvious upside to living above, or near a UGS facility, but neither is there any significant downside’.

This reflects two widespread assumptions: first that UGS in a depleted field is straightforward and poses no added risks or complications to those pertaining to oil or gas field operations. Secondly, provided the geology is suitable, everything else will be okay. Local residents’ concerns are referred to as ‘perceived risks’, but no evidence is presented to explain why the possibility of leaks and explosions is a perceived rather than an actual risk. The ward councillors believe that the close proximity of the proposed UGS facility to residential areas adds another dimension to the assessment of risk. Not only is there a need to quantify the risks, but account must also be taken of the consequences of a major accident. The risk might be very low, but if the consequences were catastrophic, then even a low level of risk becomes unacceptable. Evans (2009) provides an analysis of previous problems encountered at underground fuel storage sites, including gas storage to assist in risk assessment. In the US, the country with by far the most experience of UGS facilities in depleted hydrocarbon fields, UGS near residential areas is not recommended, although it is not prohibited. There are areas in Los Angeles where new housing has been built above depleted oilfields used for UGS. The Playa Vista development, located over the Playa del Rey Oilfield (California) provoked fierce debate, with a number of agencies and experts advising strongly against the building of the development. A US oil and gas environmental expert advised categorically against locating UGS facilities near an urban setting or building over existing gas storage sites (Endres 2000): ‘Experience has shown that underground gas storage facilities can create a serious risk of explosion or fire and should not be placed under urban settings’. He also advises that, ‘no structures should be built over abandoned gas or oil wells’. On the other hand, another US expert, G. V. Chilingar, in an addendum to a paper entitled ‘Gas migration from oil and gas fields and associated hazards’, whilst acknowledging the safety problems posed by gas and oil operations in urban areas, expressed his personal belief that ‘oil and gas production in urban areas can be conducted safely if proper procedures are followed’ and that ‘after recognition of the existing problem proper safety operating procedures can be easily developed’ (Gurevich et al. 1993). However, even though Chilingar reiterated this addendum in a further paper published in 2005, the paper itself demonstrated that 12 years on, many of the problems in the Los Angeles Basin urban oilfields had persisted and much work was still needed to implement proper safety operating procedures (Chilingar & Endres 2005). Even if one takes the more optimistic view from the US, establishing the safety and minimizing the

WELTON OILFIELD, UK LOCAL ISSUES

risks of a UGS facility near a populated area requires considerably more evaluation, expertise and emergency planning than has been available so far for the Welton project. The whole issue of proximity to hazardous sites is, at the time of writing, undergoing a review in the UK as a result of the massive explosion at the Buncefield oil depot at Hemel Hempstead on 11 December 2005 (Powell 2006a, b; Health and Safety Executive 2006). The explosion occurred at 6 am on a Sunday morning and resulted in 43 people being injured. Sixty major commercial buildings on the nearby industrial estate were destroyed. Had the blast occurred during working hours, there would have been 30 000 people within the blast zone (BBC News 2005). The Health and Safety Executive (HSE) acknowledged that, ‘had the buildings surrounding the Buncefield site been occupied, significant numbers of people could have been killed or injured by the explosions.’ Also, ‘previous assumptions about people’s ability to respond to a similar incident to Buncefield, and their safety at given distances from it, may now be open to challenge’ (Health and Safety Executive 2007). HSE has also stated that the implications of the Buncefield explosion need to be considered for other types of major hazard sites as well.

Leaking wells At the international conference on UGS held at Aberdeen in 2004, presentations on US experiences alerted the ward councillors to the issue of leaking wells. This confirmed the risk information Star Energy had given in its Admission Document to the Stock Exchange in 2004. Well bores represent ‘the primary potential conduit for vertical gas migration’ from UGS facilities in depleted fields (Evans 2004). Not only producing wells and injection wells need to be assessed and monitored: disused wells and boreholes are also potential pathways for gas to migrate to the surface (Miyazaki 2004, 2009; Chilingar & Endres 2005). It is not just old wells that have leaked at US UGS facilities; there have been leaks from more recently constructed wells which have been plugged in accordance with current US regulations. A number of surveys have found that 10% of wells plugged and abandoned in the early 1990s developed leaks within 10 years (Miyazaki 2004, 2009). Miyazaki also asserts that most abandoned oil and gas wells develop leaks over time and that well integrity problems are amplified with UGS because of the higher pressures involved. Other US experts concur: a report to the California Public Utilities Commission (CPUC) concerning the Playa del Rey UGS facility stated that, ‘structural integrity of well components is not permanent. Over extended

155

periods of time, they eventually deteriorate’ (CPUC 2004). Another expert even went so far as to say that even if a UGS facility does not leak initially, it will with time: ‘The important question is not ‘if’ the storage facility will leak, but rather ‘when’ (Endres 2000). Using the GIS data on the BGS website, the ward councillors found that there are numerous boreholes throughout the area of the Welton oilfield, some in residential streets. In addition there are about 50 disused wells, mostly in clusters near producing wells. It is not clear whether detailed analysis of all these wells and boreholes will be undertaken or which agency would have responsibility for ensuring that this was done.

Presence of hydrogen sulphide The presence of hydrogen sulphide (H2S) in the oilfield is a major concern to local residents. Few people were aware before the UGS proposals were put forward that the six oilfields in the area are sour. The original planning applications made just two references to the presence of H2S. They did not give any figures, merely referring to the gas currently extracted with oil as being ‘rich in H2S’. Supplementary information submitted to LCC by Star Energy gave a figure of 3079 ppm for H2S in the Welton oilfield. This was described as ‘low by industry standards’ and requiring ‘only minimal processing’. In the context of community safety, 3000 ppm is not low, bearing in mind that 300 ppm is classified as ‘immediately dangerous to life and health’. Research conducted at the University of Southern California Medical Facility has established nervous system damage can occur even at concentration in air as low as 1 ppm (Chilingar & Endres 2005). Trials of a compact catalytic convertor conducted at the Welton oilfield in 1987 recorded much higher H2S levels: typically between 5000 ppm and 10 000 ppm, with peaks as high as 15 000 ppm (Eddington & Carnell 1991). The gas imported into the oilfield would be ‘sweet’, having been previously processed for the National Grid, and therefore would have an H2S level no higher than 3 ppm. Star Energy stated in the supplementary planning information and in their newsletter published in July 2004 that injecting sweet gas into the field would dilute the H2S to harmless levels. The estimated level would be no higher than 31 ppm. The ward councillors learned that predicting precise H2S levels is not straightforward. Although, theoretically, importing a large quantity of sweet gas could have a diluting effect, the gas may be stored for 6 months or more, which raises the

156

M. DAVIDSON

question of whether H2S could build up during that time. The ward councillors also learned that H2S is constantly produced and operators fight an ongoing battle trying to keep H2S levels down. It has been stated explicitly that the injection process would affect the whole field and enhance oil recovery. Presumably the naturally occurring sour oil or gas could be displaced in the course of re-pressurization of the oilfield. As well as the harmful and potentially fatal effects on humans, H2S is highly corrosive. US experts who have carried out extensive research on the problems encountered with the oilfields in the Los Angeles area have concluded that H2S, ‘will eventually destroy the integrity of both the steel and cement relied upon to prevent gas migration, including abandonments performed to the current standards of the DOGGR [Division of Oil, Gas and Geothermal Resources]. The corrosive conditions of hydrogen sulphide are well known, and have defied engineering solutions’ (Chilingar & Endres 2005). Star Energy pointed out in their supplementary planning submissions that H2S is a recognized problem, with established procedures to mitigate it. Nevertheless, this is an added risk factor for use of a sour depleted field for UGS, requiring constant vigilance, not only in the short term but for decades to come. Unfortunately, maintenance schedules can be vulnerable to cutbacks when operators face financial pressures, sometimes with disastrous results. An example of this is the BP Texas oil refinery disaster in 2005 in which 15 people died. An important contributory factor was BPs ‘cost-cutting and failure to invest’ which left the refinery ‘vulnerable to disaster’. BP targeted budget cuts of 25% in 1999 and another 25% in 2005; even though much of the refinery’s infrastructure and process equipment was in disrepair. Also, operator training and staffing had been downsized (US Chemical Safety Board 2007).

Use of depleted oilfields for gas storage From the outset, members of the local action group and others with experience in the oil and gas industry expressed concern at the proposal to use an oilfield for gas storage. They claimed that there were significantly more risks associated with this type of gas storage compared to the use of depleted gas fields or salt caverns. In contrast, official agencies in the UK seem to work on the assumption that there is little difference between gas fields and oilfields for gas storage. Hatfield Moors, a UGS facility in a depleted gas field in Yorkshire, which has operated successfully since 2000 (Ward et al. 2003), is commonly cited as proof that UGS is safe.

Logically, one might expect gas storage in a depleted gas field to be more straightforward than use of a depleted oilfield, especially where oil extraction is still taking place. From 1915, when the first UGS facility was established in a depleted gas field in Canada, until the 1950s, nearly all gas storage was in depleted gas fields (Energy Information Administration 1995). In this context, use of depleted oilfields is comparatively recent and represents a small percentage of UGS facilities worldwide. Seventy five percent of existing UGS facilities are in depleted gas fields (Plaat 2004); depleted gas and oilfields combined account for 81.6% (Energy Information Administration 2006). There is evidence to suggest that there is a higher risk of gas migration from oilfields compared to gas fields. Gas held in solution in oil reservoirs is not free to migrate, whereas free gas is capable of migrating upwards to the caprock. Storage gas represents a large volume of free gas and according to a group of US experts, ‘the caprocks of most oilfield reservoirs do not have sufficient sealing capacity to stop the migration of free gas’ (Gurevich et al. 1993). Gas storage in depleted oilfields is considered particularly vulnerable to explosion since gas leaks up the outside walls of all oil wells (Tek 2001). Also, although an oilfield may be proven to trap and hold fluids, it might not safely hold ‘pressurized’ gas (Endres 2000).

Incidents Welton oilfield Welton oilfield does not have an incident-free operational history. The most serious incident occurred in 1994 when there was an oil blowout from a well at the village of Scothern. A fine mist of oil was sprayed over a swathe of the village, including children in a school playground situated 200 m from the oil well. A vivid eye-witness account of the incident was published in the Lincolnshire Echo in a letter written in response to a letter from Star Energy’s Chief Executive which claimed, ‘our exemplary 20-year production record at Welton demonstrates that we have been more than capable of mitigating all of the risks associated with our business’ (Wessel 2005). The eye-witness had been a pupil who had been sprayed with oil in the playground: ‘I remember playing on the school field at lunchtime and looking up to see a large fountain of oil and a big black cloud of spray rushing towards us . . . All of my friends and I went as fast as we could towards the school building but were sprayed with oil before we got there.’ She concluded, ‘I dread to think what the situation would have been like if there had been any possibility, no matter how remote, that it could have been a gas

WELTON OILFIELD, UK LOCAL ISSUES

cloud, or worse still, that the cloud could have been ignited’ (Bates 2005). The 1994 incident did not result in any injuries, but clothes, cars and buildings were damaged. Candecca, the operators of the oilfield at the time paid out compensation. More recently, in 2003 and 2004, there have been two significant leaks of oily water from a pipeline at Langworth, a village on the edge of the oilfield.

157

risks of daily life that everyone faces. Also, we have a degree of choice about the hazards of daily life. We can, for example, choose whether to travel by plane and we can exercise a degree of control to manage day-to-day risks: we can cross the road at a pedestrian crossing, and we can make sure our gas appliances are checked and serviced. For those living in the vicinity of a UGS facility, there is neither choice nor the means to control or mitigate that risk.

UGS facilities Most of the catastrophic events at UGS facilities have occurred in US salt cavern storage sites (Hopper 2004). The Playa del Rey oilfield situated under Los Angeles provides the closest similarity to Welton. This is an example of gas storage in depleted oilfields situated under a populated area. The whole area has been dogged by problems caused by gas migration and subsidence (Gurevich et al. 1993; Chilingar & Endres 2005). There have been serious accidents such as the gas explosion and fire in the Fairfax area of Los Angeles in 1985, which demolished the Ross Department Store and injured over 23 people. There was a ‘near miss’ involving the Belmont school construction in downtown Los Angeles. The project had to be abruptly halted when gas seepage was detected in the main electrical vault of the project, just before the power was to be turned on (Chilingar & Endres 2005). A review of the records of the Southern California Gas Company, the owners of the Playa del Rey oilfields and gas storage facilities, indicated leaks and surface seepage documented in ten wells located in the Del Rey Hills area (CPUC 2004). It is worrying to note that following repairs in the ten identified wells, four experienced recurrences or new leaks.

Assessment of risk The Star Energy planning applications contained claims that the safety systems in place ‘will ensure that people are at greater risk from everyday life than the operation of the proposed gas processing facilities’. This seems to refer only to the above ground processing activities. There is no mention of the actual UGS process itself, namely the injection of gas at high pressure and re-pressurization of the oilfield, and no assessment of risk for these activities. The argument that ‘crossing the road/using a gas cooker/travelling in a plane are more hazardous than a UGS facility’ has been used by a number of people. Whatever the degree of risk of living on top of a UGS facility, and that remains debatable, the fact is that this is a risk in addition to all the

Regulatory and planning framework There are strict rules concerning planning applications which district and county councillors are required to observe. A councillor needs to choose between expressing an opinion and/or speaking on behalf of residents and taking part in the decisionmaking process on a planning application. Star Energy’s applications for the UGS scheme at Welton initially involved both WLDC (the planning authority for the wind turbines application) and LCC for the applications for the pipeline and gas storage. The ward councillors were not members of WLDCs Planning Committee so they were in a position to speak publicly on the proposals and represent the concerns of residents in their wards. The Welton planning applications filled three large volumes. On looking through them, the ward councillors were struck by the dearth of information on safety compared to the amount of information on environmental issues: the local great crested newts received more attention than the humans. The applications complied with the requirements of LCCs Minerals Application policy, which in turn adhered to the Mineral Planning Guidance Note (MPG1) laid down by government. The planning process takes into account environmental issues such as the welfare and protection of protected animal and plant species. Human safety is deemed to be the province of HSE and EA. The rationale is to avoid duplication by different agencies. The drawback is that this can mean that safety issues are not examined fully until the planning application has been approved and construction work is underway. In practice, the process of approvals and licensing by HSE/EA is less open to public scrutiny and consultation than planning applications submitted to local councils. HSE/EA do not necessarily deal with community safety issues and may be required to deal only with on-site safety plans. In early 2004, the planning applications went out for consultation and were presented to WLDCs Planning Committee for comment as a statutory consultee to LCC. The committee decided that the issue was of such magnitude and concern that WLDCs response should be decided by a meeting of the full council. The council meeting was held

158

M. DAVIDSON

in May 2004 at the Lincolnshire Showground and was immediately preceded by a public meeting attended by nearly 500 people. The outcome was that WLDC formally objected to the proposals, citing the overwhelming concerns about community safety, and also requested the Office of the Deputy Prime Minister (ODPM) to hold a public inquiry. This was subsequently refused. From the outset, the ward councillors did not share the confidence of those who asserted ‘they would not allow it if it were dangerous’. It is difficult to establish who, in this case, ‘they’ are. Which agency is the guardian of the safety of the local community in this instance? As far as the planning process is concerned, there is some limited scope for considering issues of community safety. Both public safety and public fear can be material considerations in determining a planning application. Public fear, even if shown to be unjustified, may continue to be a material consideration, but members of the Planning Committee then have to determine what weight to accord to any material considerations (Lincolnshire County Council 2006). For the most part, local planning authorities are expected to rely on advice on safety issues provided by HSE/EA. If a proposed development does not come under the Control of Major Accident Hazards (COMAH) Regulations 1999, HSE advice may be very limited and deal only with on-site safety as opposed to wider community safety issues. The ward councillors sought information and advice on local emergency planning from LCCs Civil Protection Unit (CPU). At that stage, CPU had not been called on to give advice on the actual UGS proposals and did not have detailed knowledge of them. Their remit covers pipelines, above ground gas installations and top tier COMAH sites, for which they draw up specific emergency plans. They will not have a specific remit for UGS in an onshore oilfield unless it is designated as a top tier COMAH site. At the time of writing, HSE has ruled that the Borehole Regulations will apply, not COMAH (see later discussion). This means that the CPU will not draw up a specific emergency plan for the Welton UGS facility. In the event of a major incident, they implement their Disaster Response Plan, drawn up in conjunction with the emergency services and local councils. In the words of one local resident, ‘They have a plan for picking up the bodies, but they are not allowed to take measures to prevent an accident in the first place’. At the outset of their research, the ward councillors expected that government agencies would draw up specific safety regulations to cover onshore UGS in the UK. They were surprised to find that this is not the case. Specific regulations tend to come into existence following a major disaster. For instance,

the regulations for offshore oil and gas operations were introduced in 1992 following the Piper Alpha disaster in 1988 (United Kingdom Offshore Operators Association 2006). The EU Seveso Directive is named after the location of a disastrous explosion at a chemical factory in Italy in 1976 (De Marchi et al. 1996). The approach being used for onshore UGS in the UK seems to be to try and fit it into existing legislative frameworks that were designed with other processes in mind. This is most clearly illustrated by the Boreholes Regulations versus COMAH debate that has taken place, mostly behind the scenes, at HSE. For the purposes of both the planning application and the boreholes/COMAH question, the Welton UGS proposals have been regarded as a minerals exploration/extraction process. In reality, this does not meet the particular demands and potential risks of UGS in a depleted oilfield. Gas storage does not involve exploration, nor does it involve mineral extraction in the sense of taking a mineral out of its original environment. It is a completely different use for an oilfield to inject, store and subsequently re-process gas, which was originally extracted elsewhere. The oilfield strata are a temporary storage location. This was not envisaged either by the Mineral Planning regulations or the Borehole Regulations. At one stage HSE was indicating that the COMAH regulations would apply to the Welton UGS facility. With at least 56.6 Mcm/40 000 tonnes of gas stored, this would have made it a top tier COMAH site. Subsequently HSE stated that it would not be classified as a COMAH site because the gas would be contained in natural strata and the site contains boreholes. HSE quoted the COMAH Regulations 3.(3)c, ‘These Regulations shall not apply to the activities of the extractive industries concerned with exploration for, and the exploitation of minerals in mines and quarries or by means of boreholes.’ In other words COMAH sites and borehole sites are mutually exclusive. The Seveso II (Amendment) Directive came into force on 1 July 2005 in the UK. This extended the scope of COMAH and the mutual exclusivity between borehole and COMAH regulations no longer exists. Despite this, HSE is still designating Welton as a borehole site and has said that COMAH will not apply, citing the same clause of the COMAH regulations as before. However, for salt cavity gas storage sites, HSE has applied both COMAH and the Borehole Regulations. HSE has acknowledged that this could appear to be ‘anomalous’, especially as other EU countries interpret the Seveso Directives as applying to strata storage (HSE 2006; p. 34 –35). Until July 2005, HSE was also advising both WLDC and LCC that the Planning (COMAH)

WELTON OILFIELD, UK LOCAL ISSUES

Regulations did not apply either, for the same reason, namely because Welton was a borehole site and the above ground inventory did not exceed 50 tonnes. The wording of the Planning (COMAH) Regulations puzzled ward councillors who thought it clear that the gas stored below ground should count as part of the hazardous substance inventory. It also proved very difficult to establish whether Hazardous Substances (HS) licensing was the responsibility of the District or County Council. Eventually they found out that the Planning (COMAH) Regulations are separate from the main COMAH regulations. The Planning Authority is responsible for the former and HSE for the latter, each being the other’s statutory consultee. HS licensing is dealt with by either the District or County Council; it is the responsibility of whichever council is the Planning Authority in that particular case. In the case of Welton, LCC, as the Minerals Planning Authority, was dealing with the planning applications and thus was also responsible for the HS licensing. Not until July 2005 was it confirmed that the Welton project would require Hazardous Substance Consent (HS Consent). There was a great deal of confusion leading up to this point and it is not clear whether the change in HSEs guidance was due to an actual change arising from the COMAH (Amendment) Regulations coming into force or because of a misunderstanding of the Planning (COMAH) Regulations. It seemed that planning officers at both WLDC and LCC were having to operate in ‘uncharted waters’. LCCs Mineral Local Plan dates from 1991 and UGS does not feature at all. LCC is now required to draw up Local Development Plans, a section of which will be devoted to an updated Minerals Development Plan. It has been a concern that LCC has not yet formulated a land use policy for UGS. Dealing with applications for UGS on a piecemeal basis will not be satisfactory. The updating of the Mineral Local Plan needs to include a debate on appropriate locations for UGS in Lincolnshire. HS Consent brings some scope for consideration of community safety and for public consultation that is not available with planning consent. A site designated as a hazardous installation will have an effect on future development and this will have implications for the District Council, which handles most of the planning applications. HSE is required to define ‘consultation distances’ or zones around a hazardous installation (this includes pipelines) and will need to advise accordingly on applications for future developments. Logically one could use the principles of the consultation distances in reverse in considering whether it is appropriate to site a hazardous installation near existing housing. The search for reassurances on the safety of UGS took one of the ward councillors and WLDCs Head

159

of Planning to the International Conference on UGS held at Aberdeen in October 2004. Here they learned much about UGS, but heard little about any proposed safety regulations for the UK. Three US speakers covered a range of safety issues and problems that have arisen in US facilities, particularly where there has been no specific, consistent legislative framework. The general response from the delegates was, ‘That’s America – we have not had those problems in Europe’. Seveso II seemed to be regarded as providing the necessary regulatory framework for safe operation of UGS. In the UK, Seveso II has been enacted through the COMAH (Amendment) Regulations. However, because HSE has ruled that they will not apply to UGS in natural strata (Health and Safety Executive 2006; p. 34–35), the safeguards that Seveso II would provide by designating the Welton scheme as a top tier COMAH site are not going to be deployed.

Discussion and conclusions The ward councillors’ research has identified a range of planning and safety issues which need further attention including well integrity, gas migration, land use planning, implications of HS Consent and the scope of the COMAH (Amendment) Regulations. The sense of urgency to increase UK gas storage capacity being imposed by the highest levels of the UK government means that there is pressure to put new processes into operation before all these planning and safety issues have been fully assessed and resolved. HSEs decision not to apply the COMAH Regulations to UGS in onshore depleted fields whilst applying it to UGS in salt cavities seems to be illogical. It has caused considerable confusion and delay in the planning process for the Welton oilfield proposals. The ward councillors believe that gas storage facilities at Welton and other depleted onshore fields should be designated as top tier COMAH sites. This would provide for both on-site and off-site safety assessments and offer considerably more protection and reassurance to the local community. The Welton oilfield was identified as a future gas storage site and its storage capacity added to the inventory of UK gas supplies at a very early stage, long before the planning applications were submitted. This has created an assumption that all the safety issues have been dealt with, which is not the case. There has so far been little or no assessment of the implications for community safety and there is uncertainty as to what extent this will be done. The Welton oilfield sits directly beneath a populated area, but no one in authority has even

160

M. DAVIDSON

posed the question as to whether a UGS facility should be sited near a built-up area. To date, the various agencies have given the impression of being in collective denial that there could be any additional risks posed by the use of a depleted oilfield for UGS close to a populated area. This is a source of great frustration to the local community. There is clear evidence that there are risks associated with UGS in onshore depleted fields. The risks detailed in Star Energy’s Admission Document are not just hypothetical: there are well documented incidents in the US, particularly arising from leaking wells and gas migration (Gurevich et al. 1993; Evans 2004, 2009; Chilingar & Endres 2005). Between 1990 and 1993, seven UGS facilities in the US had to be abandoned because of problems with leaks or substantial migration losses and a further three for unspecified safety reasons. This would suggest that problems with leaks or substantial migration losses can be hard to overcome (Energy Information Administration 1995). What seems to have been lacking after the Welton oilfield was identified in principle as a potential UGS site was a process of ‘due diligence’ to confirm whether this particular site was in fact suitable as a UGS facility. A thorough characterization and evaluation of a proposed site is essential (Cappel & Miyazaki 2004; Miyazaki 2004, 2009). A detailed assessment of the geological issues was carried out during the course of the planning process and various safety and operational issues were due to be dealt with by HSE once planning permission had been obtained. This still leaves some crucial issues that have not been fully evaluated, including establishing the extent of the oilfield and the proposed gas storage area, identifying the location and condition of all wells and boreholes and addressing other well integrity issues. Had this ‘due diligence’ taken place before the planning applications were submitted, it would have been possible to assess the community safety implications in a clear, transparent way in the course of the planning process. The UK has limited experience of onshore UGS and should take heed of the problems experienced in the US, the country with by far the most experience of UGS. The evidence would suggest that gas leaks from old wells are by no means uncommon. The presence of H2S in the Welton oilfield makes the task of maintaining the integrity of the wells and other infrastructure significantly more challenging. The key question is: bearing in mind that this facility would sit beneath a populated area, is it possible to provide sufficient mitigation of the risk should there be a gas leak from the field at some point? The ward councillors do not believe that there has been sufficient recognition of the difficulties of maintaining well integrity in UGS facilities so far.

There is no sign yet of a sufficiently robust and rigorous system being put in place in the UK to monitor, detect and deal with leaking wells and gas migration. No matter how urgent the need for additional gas storage capacity in the UK, this does not absolve the relevant authorities from addressing these vital issues before pressing ahead with new UGS schemes.

References A DAM , D. 2004. Why do turbines confuse military radar? The Guardian, 4 March 2004. APELL. 2004. Gas well blowout in Gao Qiao, Chongqing, China. World Wide Web Address: http://www.uneptie. org/pc/apell/disasters/china_well/china.htm. B ATES , C. 2005. Oil well blowout is still fresh in my memory (letter). Lincolnshire Echo, 21 June 2005. BBC NEWS . 2005. Q & A: Planning and hazardous sites. Could the planning laws change as a result of the Buncefield disaster? 14 December 2005. World Wide Web Address: http://www.news.bbc.co.uk/1/ hi/uk4529218.stm. BGS. 2004. Memorandum by British Geological Survey to the House of Lords Select Committee on European Union. Gas: Liberalised Markets and Security of Supply. House of Lords, European Union Committee 17th Report of Session 2003– 2004. Submission by the British Geological Survey (BGS) to the House of Lords Select Committee, 87–90. World Wide Web Address: http://www.publications.parliament.uk/pa/ ld200304/ ldselect/ ldeucom/105/105we04.htm. C ALIFORNIA P UBLIC U TILITIES C OMMISSION (CPUC) 2004. Sale of surplus SCG property at Playa del Rey and Marina del Rey, Initial Study: Appendix F Hazards and hazardous materials background information. World Wide Web Address: http://www.cpuc. ca.gov/Environment/info/esa/playa/pdf/apndx-f C APPEL , A. & M IYAZAKI , B. 2004. Flaws in regulatory over-sight and review of underground natural gas storage: financial liability and lessons learned from recent US incidents and disasters. In: E VANS , D. J., C HADWICK , R. A. & R OWLEY , W. J. (Convenors) ‘The Future Development and Requirements for Underground Gas Storage in the UK and Europe’. Geological Society, London, Conference, Aberdeen, 19–20 October 2004, Abstract. C HILINGAR , G. V. & E NDRES , B. 2005. Environmental hazards posed by the Los Angeles Basin urban oilfields. Environmental Geology, 47, 302–317. D E M ARCHI , B., F UNTOWICZ , S. & R AVETZ , J. 1996. Seveso: a paradoxical classical disaster. In: M ITCHELL , J. K. (ed.) The Long Road to Recovery. United Nations University Press, 1996, 86– 120. DTI. 2005. Department of Trade and Industry (DTI). World Wide Web Address: http://www.dti.gov.uk. E DDINGTON , K. & C ARNELL , P. 1991. Compact catalytic controls vent odours from oil field. Oil and Gas Journal, 89, 69–70. EIA. 1995. The value of underground storage in today’s natural gas industry. Energy Information Administration (EIA), Office of Oil and Gas, DOE/ EIA-0591(95).

WELTON OILFIELD, UK LOCAL ISSUES EIA. 2006. U.S. underground natural gas storage developments 1998– 2005. Energy Information Administration (EIA), Office of Oil and Gas, October 2006. World Wide Web Address: http://www.eia.doe.gov/pub/ oil_gas/natural_ gas/feature_articles/2006/ngstorage/ ngstorage.pdf. E NDRES , B. 2000. Hazards of gas storage fields. In: K HILYUK , L. P., C HILINGAR , G. V., R OBERTSON , J. O. & E NDRES , B. (eds) Gas Migration. Gulf, Houston, Texas, 301– 308. E VANS , D. J. 2004. Appraisal of an underground gas storage proposal at the Welton oilfield, Lincolnshire. British Geological Survey Internal Report CR/ 04/122. E VANS , D. J. 2009. A review of underground fuel storage problems and putting risk into perspective with other areas of the energy supply chain. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publication, 313, 173–216. E VANS , D. J. & H OLLOWAY , S. 2009. A review of onshore UK salt deposits and their potential for underground gas storage. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publication, 313, 39– 80. G UREVICH , A. E., E NDRES , B., R OBERTSON , J. O. & C HILINGAR , G. V. 1993. Gas migration for oil and gas fields and associated hazards. Journal of Petroleum Science and Engineering, 9, 223– 238. H OPPER , J. M. 2004. Gas storage and single-point failure. Natural Gas, Hart Energy Publishing, LP, 4545 Post Oak Place, Ste. 210, Houston. World Wide Web Address: www.falcongasstorage.com/pdf/article.singlepointfailure.pdf. H OUSE OF L ORDS . 2004. Gas: Liberalised Markets and Security of Supply. House of Lords, European Union Committee 17th Report of Session 2003–2004, 125 pp. World Wide Web Address: http://www.publications.parliament.uk/pa/ld200304/ldselect/ldeucom/105/105.pdf. H EALTH AND S AFETY E XECUTIVE . 2006. The health and safety risks and regulatory strategy related to energy developments: an expert report by the Health and Safety Executive (HSE) contributing to the Government’s Energy Review. World Wide Web Address: http://www.hse.gov.uk/consult/condocs/ energyreview/energy report.pdf. H EALTH AND S AFETY E XECUTIVE . 2007. Proposals for revised policies for HSE advice on development control around large-scale petrol storage sites. Health and Safety Executive (HSE) Consultative Document CD211 7– 10. L INCOLNSHIRE C OUNTY C OUNCIL . 2006. Report to Planning and Regulation Committee 22 February 2006. Development of a natural gas storage and refining facility, Welton oilfield. Lincolnshire County Council (LCC). M IYAZAKI , B. 2004. Well integrity: an overlooked source of risk and liability for underground natural gas storage. Lessons learned from incidents in the United States. In: E VANS , D. J., C HADWICK , R. A. & R OWLEY , W. J. (Convenors) ‘The Future Development

161

and Requirements for Underground Gas Storage in the UK and Europe’. Geological Society, London, Conference, Aberdeen, 19– 20 October 2004, Abstracts. M IYAZAKI , B. 2009. Well integrity: an overlooked source of risk and liability for underground natural gas storage. Lessons learned from incidents in the USA. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 163–172. OFGEM. 2005. Office of Gas and Electricity Markets (OFGEM) website. World Wide Web Address: http://www.ofgem.gov.uk. P EOPLE ’ S D AILY . 2004. Investigators identify causes of gas blowout disaster in Chongqing, 3 January 2004. World Wide Web Address: http://english.people. com.cn/200401/03eng20040103_131748.shtml. P LAAT , H. 2004. What is the best functionality for my UGS? In: E VANS , D. J., C HADWICK , R. A. & R OWLEY , W. J. (Convenors) ‘The Future Development and Requirements for Underground Gas Storage in the UK and Europe’. Geological Society, London, Conference, Aberdeen, 19– 20 October 2004, Abstracts. P OWELL , T. 2006a. The Buncefield Investigation: progress report. Report prepared for the Buncefield Major Incident Investigation by the Health and Safety Executive (HSE) and the Environment Agency (EA), 2 February, 2006. World Wide Web Address: http://www.buncefieldinvestigation.gov.uk/ report.pdf. P OWELL , T. 2006b. The Buncefield Investigation: third progress report. Report prepared for the Buncefield Major Incident Investigation by the Health and Safety Executive (HSE) and the Environment Agency (EA), 10 May, 2006. S TAR E NERGY . 2003. Gas storage plan to boost oilfield. Energy Matters, Star Energy (East Midlands) Ltd, Lincoln, Newsletter 21, November 2003. S TAR E NERGY . 2004. Admission Document to AIM, a market of the London Stock Exchange plc. Star Energy Group Plc. T EK , M. R. 2001. Professional gas migration experts sound the alarm. World Wide Web Address: http://www. saveballona.org/expert.html. U NITED K INGDOM O FFSHORE O PERATORS A SSOCIATION (UKOOA) 2006. Health and safety in the UK oil and gas industry. United Kingdom Offshore Operators Association (UKOOA). World Wide Web Address: http://www.oilandgas.org.uk/issues/ health/faq.htm. US C HEMICAL S AFETY B OARD . 2007. Final investigation report. Refinery explosion and fire BP Texas City, Texas, March 23 2005. US Chemical Safety and Hazard Investigation Board Report No. 2005-04-1-TX March 2007. W ARD , J., C HAN , A. & R AMSAY , B. 2003. The Hatfield Moors and Hatfield West Gas (Storage) Fields, South Yorkshire. In: G LUYAS , J. G. & H ITCHENS , H. M. (eds), United Kingdom Oil and Gasfields, Commemorative Millennium Volume. Geological Society, London, Memoirs, 20, 905– 910. W ESSEL , R. 2005. Oil storage in a big rock sponge (letter). Lincolnshire Echo, 22 June 2005.

Well integrity: An overlooked source of risk and liability for underground natural gas storage. Lessons learned from incidents in the USA BRENT MIYAZAKI Innovateur International, Inc., Post Office Box 377, Pasadena, California 91102-0377, USA Corresponding author (e-mail: [email protected] or [email protected]) Abstract: Safety is a primary concern at underground gas storage (UGS) sites. Thorough evaluation of all potential migration pathways is critical to ensure UGS containment and public safety. Substantial risk is directly associated with inaccurate technical evaluations, that may result in subsequent gas migration from a UGS facility. Existing wellbores, including abandoned oil and gas wells, old dry exploration wells and water wells represent primary potential vertical gas migration conduits, which are not always thoroughly analysed during UGS site evaluation studies. Most abandoned oil and gas wells develop leaks over time, even when plugged in accordance with current (US) government regulations. Leaking wells in urban areas represent significant health and safety hazards. Many cities including Los Angeles permit construction of new homes directly over abandoned wells, even though state agencies recommend against this practice, thus placing residents at risk. Explosions and fires, along with possible exposure to substances such as benzene and toluene, are possible when gas reaches the surface through leaking wells and accumulates inside building voids. These potential problems are amplified with UGS fields, where operating pressures, already raised above the declining pressures of the field, fluctuate when alternating gas injection and extraction induce cyclic stress on wellbores and caprock sequences.

Safety is a primary concern at underground gas storage (UGS) sites. Across the United States, hundreds of underground storage facilities hold natural gas. In 2004, US Department of Energy (DOE) data indicated 415 UGS sites distributed across the country in 30 different states (Fig. 1). Approximately 83.9% of these UGS facilities were in depleted oil and gas fields (Tobin & Thompson 2001). The remaining 16.1% were divided between mined caverns (6.5%) and aquifers (9.6%). Various technical and geological factors determine UGS suitability and potential safety hazards. A thorough evaluation of geological and related technical considerations is critical to ensure UGS containment and public safety. Substantial risk is directly associated with inaccurate technical evaluations that result in subsequent gas migration from a UGS facility. Leaking wells can be an overlooked source of significant risk and liability. Recent incidents in the US (Moss Bluff, Yaggy) illustrate spectacular releases from UGS facilities (see below, also Evans 2009). Volatile organic compounds (VOCs) create additional potential health and safety risks. Leaking storage gas may carry VOCs, thereby transferring VOCs to the surface or shallow groundwater. At the surface, these VOCs represent a potential exposure pathway. Possible VOC migration and

potential exposure issues could also apply to carbon dioxide (CO2) sequestration in depleted oil and gas fields. From a geological perspective, there are two types of physical storage: porosity and man-made caverns. Potential risks for these two types of UGS are dramatically different, and importantly, examples from one type are not representative of possible risks associated with the other. Porosity storage includes UGS primarily in depleted oil and gas fields, as well as some aquifers, where matrix porosity is the storage reservoir. In contrast, caverns are open storage spaces created by various mining operations. These man-made storage areas include solution caverns in thick-bedded halite beds or salt structures, as well as abandoned underground mine networks.

Typical UGS field operations Typical field operations at UGS facilities in the United States involve injecting natural gas when supplies are abundant (lower costs), and extracting stored gas when supplemental supplies are needed (higher gas prices). UGS operations balance supply and demand, addressing seasonal demand fluctuations, transmission pipeline capacity

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 163–172. DOI: 10.1144/SP313.11 0305-8719/09/$15.00 # The Geological Society of London 2009.

164

B. MIYAZAKI

Fig. 1. Map showing distribution of UGS sites across the United States (modified from Tobin & Thompson 2001).

restrictions, and surplus capacity associated with liquefied natural gas (LNG) deliveries. When an LNG tanker arrives at port and offloads its cargo, gas supply temporarily exceeds local and regional demand or transfer capacity. UGS facilities moderate temporary excess supply associated with LNG deliveries. Thus, UGS facilities near LNG ports play an important role in stabilizing natural gas prices. UGS facilities may address issues caused when peak demands exceed local supplies or deliverability limitations. High natural gas peak demand occurs both seasonally and daily. Commonly, seasonal peak demand occurs during winter months, and early mornings represent daily peak demand. UGS operations in southern California present some unique differences in peak demand periods compared to facilities in other parts of the United States. In southern California, there are two distinct seasonal peaks: winter for space heating and hot water supply, as well as summer for electrical generation (gas turbines) associated with space cooling (air conditioning). In addition to seasonal peaks, daily peak demands are also common with natural gas usage rising each morning, generally the result of greater hot water use (natural gas water heaters) as people prepare for work, school and begin their daily routines. During winter months, space heating also increases each morning, amplifying these daily peaks.

Daily gas injection and extraction cycles at some southern California UGS facilities induce stress on the UGS system. As a result, wells (both active and abandoned) and reservoir containment features (including caprock) are subjected to multiple pressure fluctuations (injection and extraction cycles) and show a higher probability of developing leaks or other types of containment failures. Porosity storage and mined caverns exhibit substantial dissimilarities in physical storage conditions. As a result, hazards associated with these UGS facilities differ dramatically from each other. Therefore, potential risks and hazards associated with porosity storage and mined caverns are not directly comparable to each other, and past releases at UGS caverns are not representative of hazards associated with porosity storage facilities. Internal gas flow in the porosity storage zone is restricted by physical reservoir characteristics, such as permeability. As a result, leaks or releases are typically small volumes at low rates. But a minor release may represent substantial financial consequences, such as at Leyden in Colorado. In contrast to porosity storage, UGS in caverns represents higher risks and greater potential consequences. Caverns and ‘abandoned’ mines are large open containers without internal physical flow restrictions. Uncontrolled releases from UGS facilities in caverns are restricted by well capacity flow rates. The uncontrolled release at Moss Bluff

WELL INTEGRITY

Texas, in 2004, represents the spectacular accidents portrayed by the public media.

Selected American UGS release examples Examples presented below provide background information on several releases from UGS facilities associated with well releases (see also Evans 2009). In some instances, examples illustrate past regulatory conflicts or failures.

Magnolia facility, near Grand Bayou, Louisiana, 2003 (Associated press 2004) In late December 2003, leaks were detected from a UGS well at the Magnolia storage facility near Napoleonville, Louisiana. This necessitated evacuating about 30 residents from 20 homes in the Grand Bayou community approximately 3 km (2 miles) from the site. Residents required temporary housing for several weeks while repairs were completed, and the UGS field operator had to pay for all food and lodging. Fortunately, no fires or explosions resulted. This facility had received federal government approval to begin operations in November 2003. Its history was: † UGS deep underground in natural caverns within salt domes; † November 2003, federal certification to operate issued; † December 2003, gas escaped from ruptured well casing required evacuation; † cracked well casing at 442 m (1450 feet), possible operator error; † January 2004, residents allowed to return home; † the incident caused direct and indirect financial losses.

Yaggy UGS facility, Kansas, 2001 (Allison 2001, 2004; Nissen et al. 2004; KGS 2006) The January 2001 explosions and fires in Hutchinson, Kansas provide the most graphic examples of hazards and liability associated with gas leaks at a UGS site. Natural gas escaped from the storage cavern and migrated laterally from the Yaggy UGS facility, located in a rural area of Kansas. An estimated 143 million cubic feet (mmcf) or 4.05 million cubic metres of gas escaped and travelled underground over 11 km (7 miles) through geological units. The gas reached Hutchinson and erupted in several areas. Explosions and fires in two sections of the city destroyed several businesses and homes, and resulted in two deaths. Many residents were evacuated for approximately 2.5 months while investigations were conducted. Government

165

agencies assessed the UGS operator fines, and lawsuits were filed. Case details are: † developed in early 1980s for propane storage; † vertical containment (overlying caprock) evaluated; † caverns constructed at shallow depth in salt bed (wells 200–277 m deep); † closed in late 1980s and wells partially plugged with concrete; † converted to natural gas storage in early 1990s; † acceptable confining pressure increased by 17% in 1997; † pressure in Pod Number 1 (storage cavern number 1) above maximum allowable operating pressure on 14 January 2001; † January 2001, over 4.05 cubic metres (143 mmcf ) of gas escapes over several days; † release through damaged casing and annular well seal (probably damaged in 1992); † gas migrates through aquifers over 11 km (7 miles) to City of Hutchinson; † storage gas reaches surface through old abandoned brine wells in many areas; † 17 January: explosions and fires in two areas destroy buildings, causing two fatalities; † many residents evacuated for 2.5 months while investigations conducted; † Kansas Gas Service settles with City of Hutchinson for US$372 731; † Kansas Department of Health Services fines Kansas Gas Service US$180 000; † class-action law suit filed (pending) seeking US$800 million in damages for lost business, declining property values, vacant rental properties; † consequences: two fatalities, fines and lawsuits.

Leyden UGS Facility (about 22 km NW of Denver), Colorado, 1998 (COGCC 2004; Raven Ridge Resources 1998) Near the rural community of Leyden, Colorado, a UGS facility operated successfully for almost 40 years. In 1998, a property owner adjacent to the UGS filed a lawsuit claiming that the facility contaminated a primary local groundwater aquifer. A small amount of storage gas was found in the aquifer, but was confined to the UGS property. Even though storage gas had not migrated laterally away from the facility, the jury decided in favour of the plaintiff and the operator abandoned this facility. Decommissioning activities began in 2000, and the mine was converted to an underground water storage facility beginning operations in 2005 (Evans 2009). A brief history of Leyden UGS is summarized below: † coal mine abandoned in 1950, acquired in 1957 by Public Service of Colorado; † detailed geological evaluation and testing conducted in 1958;

166

† † † † † † † † † † †

B. MIYAZAKI

beginning in 1960, UGS operated for almost 40 years in former coal mine; storage capacity c. 2.5 bcf or 708 cubic metres at 1.72 MPa (250 psig); vertical gas migration, probably through injection well annular seal; 1998: groundwater contaminated by leaking well, storage gas identified in aquifer; no state agency had direct jurisdiction over the problem identified; lawsuit filed by adjacent land owner in 1998; jury decision in favour of landowner, awards 1.8 million dollars (USD) in damages; award upheld on appeal in January 2000; Colorado Supreme Court upholds decision in June 2000; storage activities cease, field decommissioning begins in 2000; consequences: financial damages awarded, UGS facility closure.

Moss Bluff, Texas, 2004 (Bardwell & Horswell 2004; Clark 2004; Horswell 2004a, b) On 19 August 2004, an explosion and fire damaged surface piping at the Moss Bluff UGS facility about 65 km (40 miles) NE of Houston. Flames from the initial explosion reached about 60 m (200 feet) high. Heat from this fire damaged the wellhead, causing wellhead failure and a second explosion. Estimated value of lost gas inventory was between 30– 36 million dollars (USD). The history of the storage site incident being: † 19 August: well casing separation in cavern allowed storage gas to enter above-ground brine piping; † fire and explosion at surface piping, flames 45–61 m (150 –200 feet) high; † about 30 families in 1.6 km (one-mile) radius evacuated; † 20 August: heat caused wellhead failure, second explosion, flames 305 m (1000 feet) high; † uncontrolled release with no relief wells and no down-hole shut-off valve; † about 100 people evacuated from 4.8 km (threemile) radius; † 24 August: flames decreased to 91 m (300 feet), some residents allowed to return home; † flames burned until 25 August when cavern was empty; † consequences: direct and indirect financial losses.

Montebello Field (Los Angeles area), California, 1970 to current (MHA 2001) During the 1970s, soil-gas was detected in a residential area on Michael Collins Circle, a small

cul-del-sac along the east side of Montebello Boulevard and north of Avenida de la Merced. Two abandoned wells are present on one plot along the south side of the cul-de-sac over the NE part of the Montebello Gas Storage Field. Analytical test results identified helium, indicating the presence of imported, processed storage gas. Several homes were purchased and demolished. Seven extraction wells were installed to a depth of 12.3 m (c. 40 feet) in the early 1980s, along with a collection system and blower to extract soil-gas. This system successfully reduced soil gas concentrations to undetectable levels and maintained these levels for 20 years. Currently, cushion gas is being extracted from Montebello Gas Storage Field in preparation for subsequent decommissioning. Although the migration pathway from storage zone (approximate depth 5000–7500 feet or 1530–2290 m) to surface was not confirmed, nearby leaking abandoned wells are suspected. In addition to storage gas reaching the surface, storage gas was also found migrating to ‘shallow’ oil-producing zones. Water from ‘shallow’ oilproducing zones was also ‘flooding’ storage reservoir. It is suspected that poor annular seals allowed downward water migration as pressures in the storage zone were reduced. Background details to the storage site are: † oil field, produced from multiple zones in Pico and Puente Formations W Pico production: zones 1–4 from about 915–1615 m deep W Puente Formation: zones 5– 9 from 1615 m to over 2290 m deep; † gas storage in depleted zones 8-1, 8-2 and 8-3, depth from about 1530–2280 m; † limited technical oversight by multiple state regulatory agencies; † hundreds of wells and abandoned wells in field and UGS; † several leaking oil wells, abandoned wells in UGS zones; † consequences: direct and indirect financial losses, decommissioning initiated.

UGS well releases and financial impacts Releases from wells at UGS facilities create direct and indirect adverse financial impacts for operators. Lost gas inventory and operational closures represent direct impacts. In contrast, indirect financial liability results from evacuations and associated compensation, fines, lawsuits, or when structures are damaged by a gas explosion and fire. Substantial value may be associated with lost gas inventory resulting from well leaks. For the 2004 Moss Bluff release, the lost gas inventory amounted

WELL INTEGRITY

to approximately US$30–36 million. For comparison purposes, using 2004 natural gas prices of US$5.50 per BTU, inventory lost at Yaggy was valued at approximately US$0.75 million. Since it is difficult to quantify lost inventory resulting from long duration minor well leaks, the accurate data needed to estimate values is typically not readily available. Permanent or temporary UGS facility closures may result from releases. Dependent upon duration, lost operating income during temporary facility closures, caused by well releases, ranges from minor to substantial. Temporary closures at both Magnolia and Moss Bluff lasted about two weeks, whereas at other facilities, closures lasted for several months (e.g. Hiltfold; Evans 2009). Loss of future income caused by permanent facility closure is substantial as illustrated by the closing of the Leyden and decommissioning of the Montebello UGS facilities. Evacuations during the past decades range from a few weeks at Magnolia in 2003 and Moss Bluff in 2004, to several months at Yaggy in 2001 and Mount Belvieu in 1980. Cost for food and temporary lodging may range from thousands to several million dollars. These costs are directly dependent upon displacement duration and total families evacuated. Permanent relocation costs may occur if homes are demolished, as was required at Montebello. In addition to operational losses, UGS well releases may result in fines, penalties and legal settlements. The release from Yaggy, with subsequent explosions and fires in Hutchinson, Kansas, generated over US$0.5 million in fines and penalties. A law suit seeking several hundred million dollars (USD) in damages is also pending. Even minor releases may create significant financial liability. At Leyden, Colorado, a jury awarded US$1.8 million to the adjacent property owner. These payments do not include associated legal fees, which may be considerable.

Wells: vertical migration conduits In addition to open vertical fractures and other permeable geological structures, wellbores represent primary potential vertical gas migration conduits. Even when plugged in accordance with current government regulations, most abandoned oil and gas wells eventually develop leaks. Due to poor construction practices and deterioration over time, old wells are especially prone to leak development. Abandoned oil and gas wells, old dry holes, undocumented abandoned wells and dry holes, water wells and brine wells, newly constructed UGS wells with defective annular seals or casing all contribute to gas migration in the US. In California, prior to constructing new homes or buildings near or over abandoned wells, the Division of Oil, Gas and Geothermal Resources (DOGGR) requires new leak testing and inspection.

167

This process is set out in the Construction-Site Review Program (Gamache & Frost 2003) implemented during the mid – late 1980s. Approximately 10% of these well inspections detected leaks (see summary below). DOGGR (Santiago pers. comm. 2004) suspects that most of these leaks resulted from inferior material used by contractors during well abandonment. Relevant information on leak frequency from abandoned wells in the Los Angeles Basin is summarized below: (1) In the early 1990s, a major oil company plugged and abandoned many wells in depleted oil fields across the basin. A review of these abandoned wells prior to new urban development of the area found about 10% of local wells developed leaks within approximately 10 years (Blalock pers. comm. 2004). The company initiated remedial actions to re-abandon these wells, using additional materials not required by existing regulations. (2) City of Long Beach: of about 10 wells abandoned in the 1990s, six were found to be leaking in 2000 (Clarke pers. comm. 2004). (3) California DOGGR data indicates approximately 10% of abandoned wells develop leaks within one year (Pandolfi pers. comm. 2004). (4) City of Los Angeles (1999)‘. . . Non-leaking wells may suddenly begin to leak, as we have experienced in several locations in the City, particularly after earthquakes, periods of rising groundwater, and even hot spells.’ (5) DOGGR (Evans et al. 2003) ‘Many wells that were abandoned to the standards at the time are found to be leaking gas upon testing.’ New wells are not immune to leaks and releases. Construction problems or difficulties could damage casing. Defective materials may also contribute to future leaks. The casing failure and resulting gas release at the Magnolia UGS facility occurred just one month after receipt of certification to operate. Geological factors, including shallow and intermediate depth high pressure gas or water-bearing zones, may adversely affect annular seals and contribute to leaks. In addition, environmental factors, such as the presence of hydrogen sulphide (H2S) and carbon dioxide (CO2) may increase deterioration of annular seals and steel casing. Old wells in depleted oil or gas fields converted to natural gas storage facilities present potential leakage risks. This is especially notable in the United States, where many of the wells were originally drilled in the early to mid 1900s, with some wells as early as the late 1800s. Well integrity typically deteriorates over time, especially for old wells completed prior to modern design and construction practices. The potential for leaking old wells is

168

B. MIYAZAKI

amplified when abandoned oil or gas fields are converted to UGS facilities, where operating pressures are substantially higher than residual pressures in depleted oil or gas fields (see discussion above for Montebello Field). Repeated gas injection and extraction subject old well casings to alternating pressure cycles, thus increasing the potential for creating an environment that compromises casing and annular seals. As a result, old wells at UGS facilities converted from abandoned oil and gas fields experience more leaks and larger releases than comparable wells in depleted oil and gas fields without UGS facilities. These issues may also apply to CO2 sequestration in depleted oil and gas fields where old abandoned wells are present and injection pressures routinely fluctuate. Potential vertical migration pathways have not always been thoroughly analysed during UGS site evaluation studies. When leaking wells intersect permeable geological units, such as aquifers or fractured zones, complex potential migration pathways may exist. Numerous leaking wells, in some cases representing very high potential risk, are present at some American UGS sites. Examples provided above illustrate the level of liability associated with selected historical releases. As illustrated by recent release incidents, geological and other technical factors may directly influence risk level associated with UGS locations. In turn, these factors represent varying degrees of financial liability corresponding to site-specific factors and overlying land use. Several primary factors contributing to leaks in operating or abandoned wells are listed below: Common causes of casing leaks: † holes in casing (corrosion); † casing damage (Yaggy & Magnolia); † rusted steel casing (affecting annular seals); † casing shoe leaks; † channelling in annular seal; † poor cement bonding with pipe or borehole; and † deteriorated annular seals. Geological factors affecting annular seals and confinement: † ‘shallow’ and intermediate depth gas-bearing zones, especially when overpressured; † aquifers, especially artesian aquifers; and † permeable fractured zones encountered in wellbores. Wellbore factors affecting annular seals and confinement: † ‘mud cake’ on borehole wall; † poor quality cement (common in old wells); † scale on outside of casing; and † ‘sidetracking’ or other drilling problems.

As noted above, many factors contribute to well leak potential. In the Los Angeles Basin area, DOGGR indicates that many wells abandoned to the current standards, upon testing are found to be leaking gas (Evans et al. 2003). This gas is emanating from sands above the upper hydrocarbon zone, and in some cases from very shallow depths less than about 310 m (1000 feet). There is growing evidence that these shallow sands were charged with gas migrating up the annulus from deeper oil producing zones, at pressures as high as the original pressure of the oil reservoir below (Evans et al. 2003).

Hazards associated with leaking wells Leaking wells in urban areas represent significant health and safety hazards. Fires and explosions are possible when gas reaches the surface through leaking wells and accumulates inside structures. Exposure to possible toxic substances, such as benzene and toluene, is also a relevant concern. During the 1980s, three dramatic incidents associated with leaking wells related to producing oil and gas fields occurred. These incidents are not representative of most well leaks, nor were they associated with UGS facilities, but they illustrate a rare and extreme level of potential risk and liability that leaking wells at gas storage facilities may create. Near LaSalle, Colorado, during the mid-1980s, natural gas from a leaking oil and gas well reached an aquifer. Gas migrated laterally through the aquifer to a nearby abandoned water well before rising to the surface and accumulating inside a commercial building above this water well. An explosion and fire destroyed this structure. In March 1985, an explosion and fire destroyed the Ross Dress for Less Department Store, located in the Los Angeles Fairfax District. Natural gas migrated vertically through abandoned wells and along fracture systems beneath the area. Gas accumulated within the store, reaching explosive limits. In February 1989, gas above explosive levels was detected in buildings on the north side of Third Street, across from the department store explosion site, prompting new evacuations. Fortunately, no explosions or fires occurred in 1989 (Hamilton & Meehan 1992). In 2000, eight landowners in Weld County, Colorado filed complaints with the Colorado Oil and Gas Conservation Commission (COGCC), alleging gas in their domestic water wells was caused by oil and gas production activities in the area. COGCC initiated sample collection from 14 oil and gas wells, along with 15 domestic water wells. Laboratory analytical results indicated thermogenic gas present in six water wells. The source of gas in

WELL INTEGRITY

four water wells was traced to three nearby oil and gas wells. These oil and gas wells were repaired and additional mitigation measures implemented. Investigation to identify potential sources of thermogenic gas present in the other two wells continued (COGCC 2002). Potential risks associated with leaking wells at UGS facilities are dependent, in part, on the type of storage: porosity versus caverns. Since caverns have ‘infinite’ permeability, the maximum release rate is dependent upon physical well constraints. At Moss Bluff, Texas, flames from escaping gas reached over 300 m high. This incident illustrates an uncontrolled release through a well that was not equipped with a down-hole shut-off valve. In contrast to caverns, porosity storage facilities exhibit limited permeability dependent upon host reservoir rock and associated engineering properties. As such, release rates are limited by physical rock characteristics, such as near-well effective porosity and permeability. Although sufficient gas could slowly accumulate inside structures over time and create explosion hazards, dramatic releases similar to those from caverns are not generally recorded at porosity UGS facilities. The selected examples of leaking wells at several UGS facilities, presented above, with their resulting consequences, illustrate migration conduits through both wells and natural geological pathways. In addition to gas migration, possible toxic substances associated with leaking gas, and resulting human exposure to these substances, also warrant concern for assessing potential future liability.

Regulatory oversight and technical review Evaluating original well completions, abandonment procedures and well casing/cementing integrity are critical to minimize the potential for UGS migration losses through wells and to maintain safety at UGS sites. Environmental issues associated with leaking wellbores represent possible financial liability for UGS operators if not fully defined and mitigated. Generally, American regulatory agencies have not implemented adequate criteria for comprehensive well integrity analysis to ensure safety at UGS facilities. Thus, it becomes the responsibility of prudent UGS operators to conduct thorough evaluations to minimize potential financial liability and losses. An environmental impact assessment report for the Wild Goose storage facility expansion is an example of the environmental review and permitting process in California where the operator conducted technical studies beyond those required to meet minimum regulatory requirements (MHA 2002; Hietter 2009).

169

Regulatory agencies Many different federal, state and local regulatory agencies may hold jurisdiction over various aspects of UGS and releases from leaking wells. Typically, agency roles and responsibilities are narrowly defined. As a result, some issues may not fall clearly under one particular jurisdiction. This was the situation with a 1998 UGS release near Leyden, Colorado. In 2000, the Colorado Attorney General wrote a letter pertaining to the Leyden UGS facility stating, ‘No agency of the State of Colorado appears to have explicit statutory authority to assure a permanent, safe and environmentally proper closure of this operation’ (Colorado Department of Law 2000). Action by the Colorado legislature was required to designate a state oversight agency. In California, dependent upon compounds discharged, multiple local and state agencies may hold jurisdiction over evaluating UGS releases. State agencies may include: California Public Utilities Commission (CPUC), Division of Oil, Gas and Geothermal Resources (DOGGR), Department of Toxic Substances Control (DTSC), Regional Water Quality Control Board (RWQCB), and Air Quality Management District (AQMD). Since the primary responsibility of DTSC, RWQCB and AQMD is focused on other environmental issues (soil, water and air, respectively), they do not maintain technical staff with specific expertise to evaluate well leaks and associated hazards. The CPUC regulates UGS facilities in California. In addition to state and federal regulatory agencies, various city or county agencies may also hold jurisdiction over specific permit requirements associated with construction near abandoned wells. In Los Angeles and many other California cities, local regulations allow the construction of new homes and other structures directly over abandoned wells. This commonly occurs even though the DOGGR recommends against this practice. In the City of Signal Hill, a Starbucks outlet and a Food 4 Less supermarket were constructed over recently abandoned wells (Daniels 2004). By allowing home and commercial construction near and directly over abandoned wells, cities may inadvertently place residents and the public at risk, and create potential future liability for operators. In the early 1970s, concerns arose regarding building construction over or in near proximity to oil and gas wells (Evans et al. 2003). For example, during the 1970s, several houses in Montebello, California were evacuated, purchased and demolished when storage gas was identified migrating beneath these homes from nearby old abandoned wells. In response to various problems and rapid building encroachment into historic oil production

170

B. MIYAZAKI

areas, the California Laws for Conservation of Petroleum & Gas was enacted in 1986, establishing the Construction-Site Plan Review Program. Prior to implementation of the Construction-Site Plan Review Program established by the 1986 legislation, DOGGR did not review development plans. For urban development completed prior to this period, plugged and abandoned oil wells were not evaluated. Overlying some oil fields, such as Los Angeles City and Salt Lake, many structures may have been constructed directly over or very near oil wells that were not abandoned to current standards (Gamache & Frost 2003). Since DOGGR does not regulate construction over or near abandoned wells, they are limited to making recommendations. Generally, depending on site and building layout, between approximately 1.5–3 m clearance adjacent to structures is recommended. These minimum distances allow equipment access to abandoned wells that may require remedial actions if found leaking. Recommended DOGGR building clearance surrounding abandoned wells is discussed by Evans et al. (2003). Following the 2001 release from the Yaggy UGS facility and subsequent explosions in Hutchinson, Kansas, the Kansas Geological Survey (KGS) provided technical assistance to the Kansas Department of Health and Environment (KDHE), Kansas Gas Service and City of Hutchinson during an extensive investigation. Based on findings from the KGS studies, regulations governing UGS were revised to include additional technical review and evaluation criteria. Since August 2003, Kansas has required a comprehensive evaluation of potential vertical and horizontal migration pathways, including all abandoned wells (oil and gas, brine wells, water wells, etc.) in the area around proposed UGS facilities. In Kansas, two different state agencies regulate UGS facilities. KDHE provides oversight for storage in salt solution caverns, whereas the Kansas Corporation Commission (KCC) regulates underground porosity storage in depleted oil and gas reservoirs. Copies of regulations are available at the respective agency websites for salt solution caverns (KDHE 2006) and porosity storage (KCC 2006). The KDHE (2006) UGS regulations replace regulations adopted in 1981. Since injection wells for CO2 sequestration may develop leaks and contaminate groundwater aquifers, the federal government is proposing regulations for a new class of well. In July 2008, the US Environmental Protection Agency (EPA) issued its first proposed regulations for CO2 injection wells (Cappiello 2008). A final rule is expected in late 2010 or 2011. Copies of fact sheets and the proposed rule are available at the EPA website (EPA 2008).

Environmental impact assessment Potential health and safety hazards caused by leaking wells are typically discounted during environmental impact assessments. As discussed above, abandoned oil and gas wells, including old dry holes, are possible vertical migration conduits and may represent high risk. In most states, old wells are not thoroughly analysed by regulators during environmental impact analysis and evaluation for new surface development projects (both commercial and residential). Inadequate environmental impact assessment of potential leaking wells at UGS facilities is also common. Environmental impact assessment reports commonly indicate that since wells were abandoned in accordance with existing state regulations, no hazards exist. In addition, environmental impact assessment reports typically state that local methane codes protect homes and structures from methane seepage and intrusion contrary to physical evidence. As illustrated above, about 10% of abandoned wells in southern California develop leaks within a relatively short time frame (about 10 years or less). Although municipal building codes typically reduce the potential for VOC intrusion from abandoned wells into occupied buildings, these codes do not prevent gas migration in all cases. Some measures are not feasible, especially where shallow groundwater is present, resulting in design flaws that limit protection. Thus, conclusions indicating no hazards as stated in these environmental reports are based on flawed logic, and therefore, incorrectly discount potential hazard levels. Evans et al. (2003) describe California well abandonment requirements, along with historic problems and issues recognized.

Technical review Situations discussed above illustrate the importance of using a project team, with appropriate experience and expertise, to conduct a comprehensive review and evaluation of abandoned wells and other factors affecting gas containment in order to minimize risk and potential future financial liability. If abandoned oil and gas wells, along with old dry holes, are not thoroughly analysed by regulators during UGS site evaluation studies, these vertical migration conduits may represent potential highrisk gas migration conduits. Without adequate planning and data analyses by a qualified team, the technical review process is incomplete, and reports may incorrectly understate potential hazards or impacts. Commonly, the technical expertise of staff at regulatory agencies providing UGS oversight, as well as many of their environmental consultants, is inadequate to fully delineate potential hazards

WELL INTEGRITY

associated with UGS. In California, the CPUC conducts regulatory reviews and issues permits for UGS without maintaining staff with the appropriate technical backgrounds and expertise to evaluate hazards and impacts fully. Thus, UGS operators may face unidentified potential risk and financial liability. A widespread incorrect assumption is that old oil and gas wells do not leak when plugged and abandoned in accordance with current regulatory requirements. As discussed above, minor leaks from abandoned wells are fairly common.

Conclusions If thorough technical evaluations are conducted, underground natural gas storage is feasible under various geological conditions. In addition to boreholes, geological and reservoir conditions must be well defined to ensure both lateral and vertical containment of hazardous and explosive gases. Evaluating original well completions, abandonment procedures and well integrity are critical to prevent UGS migration losses and maintain safety. UGS site selection should also consider economic and environmental factors. Environmental issues associated with leaking wellbores represent possible financial liability if not fully defined. Remedial actions should be implemented, as necessary, to address potential risks delineated during analysis. Commonly, environmental consultants lack the necessary experience and expertise to evaluate hazards and risks at UGS facilities adequately. As a result, some issues are ignored or never evaluated, leaving the potential for future incidents unchecked. A multi-disciplinary technical team, comprising geologists, engineers, attorneys and regulatory experts is necessary for thorough data analysis and project evaluation. The project team must have appropriate experience and expertise. Current regulatory oversight in many parts of the USA does not protect the public or UGS facility operators from financial liability associated with UGS releases. Although regulations assist in controlling potential hazards and accidents, they do not address critical technical issues identified during recent studies. Multiple agencies may hold jurisdiction over different aspects of a UGS facility, and conflicts may arise over which agency is responsible for specific hazards and risks. In addition, regulatory agencies do not usually maintain technical staff with the expertise required to evaluate potential hazards associated with UGS. Possible economic and environmental liabilities are associated with UGS sites in or near urban areas. As such, adequate technical oversight and review is critical to maintaining public safety in these situations. Both existing and projected future land use

171

should be considered when evaluating UGS sites. During site evaluation, the potential financial liability associated with technical factors not addressed by existing regulations must be considered. Each UGS site exhibits unique characteristics. Therefore, incidents or release at one facility may not represent possible conditions or risks at other facilities. Dramatic differences between porosity storage and caverns preclude direct comparison of risks between these two vastly different types of UGS.

References A LLISON , M. L. 2001. Hutchinson, Kansas: A geologic detective story. Geotimes, October 2001, 46, 14– 18. A LLISON , M. L. 2004. Hutchinson gas explosions — unraveling a geologic mystery. Kansas Bar Association, 26th Annual KBA/KIOGA Oil and Gas Law Conference, 1, pp. 3-1 to 3-29. A SSOCIATED P RESS . 2004. Gas leak forces 30 to evacuate, underground storage well has a crack. The Sun Herald on-line, posted 2 January 2004. World Wide Web Address: http://www.sunherald.com/ mld/sunherald/news/state/7615998.htm B ARDWELL , S. K. & H ORSWELL , C. 2004. Valve failure sends flames into sky at moss bluff storage facility. Houston Chronicle, 20 August, 2004. B LALOCK , B. 2004. Chevron-Texaco. Personal communication on 6 February 2004. C APPIELLO , D. 2008. EPA unveils first rules on carbon dioxide storage. Associated Press, 15 July 2008. World Wide Web Address: http://www.usatoday.com/ news/washington/2008-07-15-2408097237x.htm C ITY OF L OS A NGELES . 1999. Inter-Departmental Correspondence. C LARK , J. 2004. Second moss bluff explosion accesses 6 bcf of gas, feeds larger fire. Oil and Gas Journal Online, 20 August 2004. C LARKE , D. 2004. City of Long Beach, Department of Oil Properties (retired). Personal communication on 28 April 2004. COGCC. 2002. Monthly Staff Report, December 2, 2002 (Report of the Colorado Oil and Gas Conservation Commission). COGCC. 2004. Monthly Staff Reports, Feb 10 2004 and May 24 2004 (Report by the Colorado Oil and Gas Conservation Commission). World Wide Web Address: http://cogcc.state.co.us C OLORADO D EPARTMENT OF L AW . 2000. Letter to Colorado Senator Powers and Speaker George from Ken Salazar, Colorado Attorney General, July 31, 2000. D ANIELS , C. 2004. Closure of oil wells raises issue of safety. Los Angeles Times, 25 April 2004. DOE (US Department of Energy) website 2004. World Wide Web Address: http://www.fossil.energy.gov/ programs/oilgas/storage/ D IETER , J. A. & P URSELL , D. A. 2000. Underground Natural Gas Storage, Gas Storage Supply and Demand Relief Valve. Simons & Company International, June 28, 2000.

172

B. MIYAZAKI

EPA (US Environmental Protection Agency) website 2008. World Wide Web Address: http://www.epa. gov/ogwdw/uic/wells_sequestration.html E VANS , B. L., S AILOR , R. P. & S ANTIAGO , E. 2003. Well Abandonment in the Los Angeles Basin: A Primer. SPE 83443. Society of Petroleum Engineers, Long Beach. E VANS , D. J. 2009. A review of underground fuel storage problems and putting risk into perspective with other areas of the energy supply chain. In: E VANS , D. J. & C HADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. Geological Society, London, Special Publications, 313, 173–216. G AMACHE , M. T. & F ROST , P. L. 2003. Urban Development of Oil Fields in the LA Basin Area, 1983 to 2001. SPE paper 83482. Society of Petroleum Engineers, Long Beach. H AMILTON , D. H. & M EEHAN , R. L. 1992. Cause of the 1985 Ross Store explosion and other gas ventings, Fairfax District, Los Angeles. In: P IPKIN , B. W. & P ROCTOR , R. J. (eds) Engineering Geology Practice in Southern California. Association of Engineering Geologists, Special Publications, 4, 143–157. HIETTER , L. M. 2009. Environmental issues in permitting gas storage: The Wild Goose case history. In: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Developments in the UK and Europe. Geological Society, London, Special Publications, 313, 139–148. H ORSWELL , C. 2004a. Gas from underground cavern explodes in Liberty County. Houston Chronicle, 19 August 2004. H ORSWELL , C. 2004b. Preparations to cap natural gas well blaze continue. Houston Chronicle, 20 August 2004. KCC. 2006. Kansas Corporation Commission (KCC) website 2006. World Wide Web Address: http:// www.kcc.state.ks.us/conservation/

KDHE. 2006. Kansas Department of Health and Environment (KDHE) website 2006. World Wide Web Address: http://www.kdhe.state.ks.us/geo/ KGS. 2006. Kansas Geological Survey (KGS) website 2006. World Wide Web Address: http://www.kgs. ku.edu/Hydro/Hutch/Background/ MHA. 2001. Mitigated Negative Declaration, Initial Study, and Mitigation Program for Decommissioning and Sale of Southern California Gas Company’s Montebello Gas Storage Facility, Montebello California. Report by MHA Environmental Consulting, Inc. (MHA) for California Public Utilities Commission, 11 May 2001. MHA. 2002. Draft Environmental Impact Report for the Wild Goose Storage, Inc. Expansion Project. Report by MHA Environmental Consulting, Inc. (MHA) for California Public Utilities Commission, March 2002. N ISSEN , S. E., W ATNEY , W. L., B HATTACHARYA , S., B YRNES , A. P. & Y OUNG , D. 2004. Geologic Factors Controlling Natural Gas Migration from the Yaggy Underground Gas Storage Site. AAPG Annual Meeting, 18–21 April 2004, Dallas, Texas. P ANDOLFI , L. 2004. Geoscience Analytical, personal communication August 2004. R AVEN R IDGE R ESOURCES , I NC ., 1998. Gas Storage at the Abandoned Leyden Coal Mine near Denver, Colorado. U.S. Environmental Protection Agency, contract 68-W5-0018. S ANTIAGO , E. 2004. California Division of Oil, Gas and Geothermal Resources, personal communication 29 April 2004. T OBIN , J. & T HOMPSON , J. 2001. Natural Gas Storage in the United States in 2001: A Current Assessment and Near-Term Outlook. Energy Information Administration, 9 pp. World Wide Web Address: http://tonto.eia.doe.gov/ftproot/features/storage.pdf

A review of underground fuel storage events and putting risk into perspective with other areas of the energy supply chain D. J. EVANS British Geological Survey, Keyworth, Nottingham NG12 5GG, UK Corresponding author (e-mail: [email protected]) Abstract: The UK became a net importer of gas during 2004 and faces an increasing dependency on imports, yet has very little gas storage capacity. The UKs capacity to import, transport and store gas and liquid natural gas (LNG) has to be improved, requiring greater investment in new gas supply infrastructure. Construction of appropriately sited onshore underground gas storage (UGS) facilities is needed. However, local groups oppose most proposed UGS sites on the grounds of safety, citing the dangers of gas migration and rare fatal events, mostly in America. This paper summarizes 228 reported events of widely varying cause, nature and severity at underground fuel storage (UFS) facilities; the majority at USA SPR facilities. Since UGS was first undertaken in 1915, reports of 13 fatalities, around 72 injured and the evacuation of at least 6700 people are found at UFS sites. Some communities have experienced multiple evacuations. In the context of the danger posed to the general public, three of those killed were staff at two UFS facilities. UGS (including LPG) has led to 10 civilian deaths, 25 injured and c. 1250 evacuated. In other areas of the energy supply chain, casualties are orders of magnitude greater, with at least 1525 dead, 6826 injured and the evacuation of over 7000 at incidents involving above ground fuel storage tanks since 1951. When considering UK UGS applications, the risk of UGS and wider UFS experiences should be put into context. Worldwide, over 90 years experience in UGS now exists, with around 630 facilities of different types currently operational. Technologies used are often those of, or derived from, a well-regulated oil and gas exploration industry. In contrast to public perception, industry and academia recognize that UGS has an excellent health, safety and environmental record. Although it should not be claimed that gas will never be found outside the intended well or storage facility area, UFS casualty figures appear to corroborate claims by supporters of the technologies that salt caverns provide one of the safest answers to the problem of storing large amounts of hydrocarbons and that even in urban areas underground gas storage, oil and gas production can be conducted safely if proper procedures are followed. If gas is found outside the intended system, then after recognition of the problem, mitigation and safe operating procedures can and have been developed.

Large gas reserves on the UK continental shelf (UKCS) meant that for many years the UK was a net exporter of gas. During that time it had also shown an increasing reliance on gas as an energy source in all aspects of industrial and domestic life. This was highlighted during the mid-1990s by the switch from coal to gas in electricity production (the so-called ‘dash-for-gas’; National Grid 2005). However, from a position in 2001 where the UK was the world’s fourth largest gas producing country, the UKCS reserves and production are now showing rapid decline such that during 2004 the UK became a net importer of gas (DTI 2006a, b, 2007a). In the future we will become increasingly dependent upon gas imports to meet demands and it is predicted that by 2020, as much as 90% will be imported (Fig. 1; DTI 2006a, b). The Government recognizes that the UK economy and gas users face major challenges in the face of continued growth in demand. Any weakness in infrastructure could result in higher gas

prices, or interruptions to supply with damaging consequences for both UK markets and consumers (DTI 2005, 2006a, b, 2007a). To meet these challenges, manage the changes and lessen impacts on UK users, the Government believes that there will be a need to increase substantially Great Britain’s capacity to import, transport and, most importantly, store gas and LNG efficiently (DTI 2006a–c, 2007a, b). The situation has been addressed partially with the construction of new pipelines (e.g. Langeled) and improved interconnectors such as Zeebrugge and Balgzand into Bacton having been commissioned (DTI 2006b, c, 2007a, b). However, the UK will require greater investment in new, timely and appropriately sited gas storage infrastructure. This may be best met through the development of onshore underground storage in salt caverns or pore storage facilities (depleted oil/gas field reservoirs or aquifers). UK onshore geology would permit a significant volume of natural gas to be stored underground in

From: EVANS , D. J. & CHADWICK , R. A. (eds) Underground Gas Storage: Worldwide Experiences and Future Development in the UK and Europe. The Geological Society, London, Special Publications, 313, 173–216. DOI: 10.1144/SP313.12 0305-8719/09/$15.00 # The Geological Society of London 2009.

174

D. J. EVANS

Fig. 1. Anticipated UK gas demand and declining production rates (after DTI 2006a).

a variety of subsurface facilities, providing a blend of longer and shorter-term storage to meet the differing supply demands. This would include storage in what the British Standards (functional recommendations) BS EN 1918 documents (BS 1998a, b) view as two of the more desirable scenarios: depleted (or depleting) oil and gas fields (having a proven ability to trap and retain hydrocarbons) and salt caverns (halite is impermeable and viscoplastic, such that faults and fractures anneal under normal burial temperatures and pressures). However, the potential areas for the development of underground gas storage (UGS) onshore in Great Britain are limited to where suitable geological strata or trapping structures are present. In these areas a number of projects are currently under development and construction, or are planned (Fig. 2; Evans & Holloway 2009). Presently, UK UGS applications face uncertain progress, being subject to numerous and lengthy planning consent processes; both local planning controls, currently overseen by the Department for Communities and Local Government (CLG), and specialist development consent regimes currently administered by the Department for Business, Enterprise and Regulatory Reform (BERR, formerly the Department for Trade & Industry or, DTI). As noted by Government, local communities close to proposed facilities also strenuously oppose UGS applications (DTI 2006b, 2007a). Opposition is based mainly upon the fear of a repetition of rare but serious incidents seen at UFS facilities, most notably in the USA, where fatalities have occurred (Fig. 3). The Preesall cavern storage and depleting Welton Oilfield proposals illustrate the degree of opposition to UGS in the UK (Evans & Chadwick 2009). After a lengthy Public Inquiry, planning permission was refused at Preesall based partly on perceived risk. Of interest was the outcome of the Inquiry into developing the depleted Caythorpe gas field for

UGS. The Inspector concluded that case law (the Newport case) shows public perception of fear (risk) is capable of being a material consideration in determining planning applications (core documents cited in Newman 2007). A number of factors were seen to demonstrate that local people’s fears for their safety were neither baseless nor unfounded. These included the submission by East Ridings of Yorkshire Council of a record of Incidents associated with underground gas storage (core documents cited in Newman 2007). However, ministers of state (CLG/BERR), following the Inspector, reached a very different ruling to the Preesall decision (DCLG 2007), stating ‘that risk cannot be entirely eliminated, but that is true of any form of onshore storage, and there were not such risks to human health and safety as to warrant rejecting this particular site’ (DCLG 2008). Planning permissions were granted. Confusion thus exists regarding safety and the perceived levels of risk at UFS/UGS facilities. Regulations and procedures must ensure that the safety of operations is addressed at all times and stages. This paper reviews and outlines problems encountered at UFS facilities alongside accident and casualty rates seen elsewhere in the energy supply chain and the petrochemicals industry. These data allow an objective assessment of the relative safety and risks of UFS/UGS and thereby help to place underground storage technologies in context. This may assist in improving public awareness of the safety record of, and issues surrounding, the development of UFS and UGS facilities.

The types and numbers of UFS facilities UFS provides reliable high volume storage and for detailed descriptions of the type and function of UGS facilities, refer to Plaat (2009). In 1996/1997 there were 580 underground storage sites worldwide, almost triple the number in 1970 and providing 262.4 bcm (billion cubic metres) volume of working gas capacity (Chabrelie et al. 2003). This was equivalent to 11.7% of the world’s consumption (2250 bcm; BP 2006). By 2005/2006 the number of storage facilities had increased to 627, providing working gas capacity of 319.3 bcm (Fig. 4), yet with worldwide consumption at 2750 bcm, the percentage of storage to consumption remained almost unchanged at 11.6% (BP 2006). Currently, around 630 UGS facilities are operational worldwide (Fig. 4). The USA operates the greatest number of UGS facilities, with 394, although 37 were classified as marginal at the end of 2005 (i.e. no injections or withdrawals, or withdrawals only were made; EIA 2006). Europe has around 120 facilities

PROBLEMS AND RISKS OF UFS

175

Fig. 2. GB and Ireland operational and proposed underground gas storage sites, with town gas storage sites investigated during the late 1950s and early 1960s and the Killingholme LPG chalk cavern storage site.

currently operational, mostly located in France and Germany. Generally, storage is in salt caverns, aquifers or depleted oil and gas fields: depleted fields provide the highest volumes and represent the most cost effective process. Large underground salt caverns provide secure environments for the containment of materials that do not cause dissolution of salt. They may be used to store a range of materials including liquid (oil, liquefied petroleum gas (LPG) and natural gas liquids (NGLs)), gaseous hydrocarbons and hydrogen. They are also used

for compressed air storage (Crotogino et al. 2001; Leith 2001; Cheung et al. 2003), or the disposal of solid and oilfield waste and radioactive waste materials (Veil et al. 1998; Thoms & Gehle 2000). Some caverns in the USA and Russia have been used for the underground testing of munitions and nuclear weapons (Thoms & Gehle 2000; Leith 2001). Abandoned mines (coal, ore and salt), unlined and lined rock cavities (LRC) also provide potential for the development of natural gas storage and compressed air storage facilities. Examples of gas

176

D. J. EVANS

Fig. 3. Major salt basins of the North America, including the main salt cavern storage facilities referred to in the text and images of the major UFS release incidents at Hutchinson (downtown fire, courtesy Chief Forbes, Hutchinson Fire Brigade; gas geyser, CUDD Drilling/Shannon Pope of RPC Inc.), Moss Bluff (after Dave Ryan, Beaumont Enterprises http://www.thefortressweb.com), Brenham (courtesy of NTSB 1993), Fort Saskatchewan (after Liz Nayowski, Fort Saskatchewan Record: http://www.chem.queensu.ca/chembook/articles/ethane_fire_in_fort_saskatchewan.htm). Also shown, night time image of the Carlsbad gas pipeline incident that caused a crater 15.5 m by 34.5 m. Bridge towers (estimated 10 m high) arrowed for scale (courtesy of NTSB 2003).

PROBLEMS AND RISKS OF UFS

177

Fig. 4. Number of underground natural gas storage facilities, working volumes and deliverability, total gas consumption both worldwide and in the USA (based upon Chabrelie et al. 2003; Favret 2003; IGU 2003; Plaat 2004, 2009; BP 2006; EIA 2006; WOC 2006).

storage in abandoned mines are rare (none are currently operational) but have included the Leyden coalmines (Jefferson County, Denver, Colorado) and coalmines in the Anderlues and Pe´ronnes mines in the Hainaut coalfield of southern Belgium (Piessens & Dusar 2003). An old iron mine in May-sur-Orne, northern France, with a capacity of 5 mcm was also used to store hydrocarbons but was decommissioned several years ago (P. Be´rest, 2008 pers. comm.). Leakage from such facilities is reported and has caused their closure in the past (see below). Unlined rock cavities may provide modest storage capacities in some countries with several dozen operational in Northern Europe, including five in France (P. Be´rest, 2008 pers. comm.). The Killingholme LPG storage facility commissioned in 1985 comprises purpose-mined storage caverns 180–190 m underground in the chalk of North Lincolnshire. It was the first and is thought to remain the only mined void

LPG storage facility outside of salt beds in the UK. Product containment is reliant upon hydrostatic pressure (Geological Society 1985; Trotter et al. 1985).

Events at UFS facilities and leakages of stored product The following sections outline events experienced at UFS facilities around the world (Figs 5–10), classified by storage type (salt cavern, reservoir etc). An earlier stage of the work reported 65 identified cases of failure or problems encountered at UFS facilities (Evans 2007). Subsequent work has found an additional 163 events involving various storage types. Some have reported casualties and most are from American SPR oil storage facilities. Figure 5 summarizes the type of stored product, country or American state in which the incident or

Fig. 5. Summary of main processes leading to events and processes at underground fuel storage facilities.

Fig. 6. Summary of documented events reported at underground hydrocarbon storage facilities developed in depleting oil/gas fields, some of which have led to leakage and/or failure. For Figures 6 –10, the numbers in brackets in column 6 relate to the numbers in column 2 of Figure 5.

Fig. 7. Summary of documented events reported at underground hydrocarbon storage facilities developed in aquifers, some of which have led to leakage and/or failure.

Fig. 8. Summary of documented events reported at underground hydrocarbon storage facilities developed in salt caverns that have led to leakage and/or failure resulting in deaths/injuries or evacuees.

problem occurred, the main contributory factors that led to the event and any reported casualty figures (and/or those evacuated), arising.

Events at depleted (or depleting) reservoir storage facilities The first storage operations were undertaken at an operating gas field in Welland County, Ontario (Canada) in 1915 (EIA 1995; WGC 2006). The first gas storage facility in a depleted reservoir was built in 1916, using a gas field in Zoar near Buffalo, New York (USA), and remains the oldest operational facility. Gas storage in depleted oil/ gas fields represents around 76% of the total number of UGS facilities (Fig. 4). Twenty-seven documented events at depleted oil and gas fields used for UGS purposes were found (Figs 5 & 6 and references therein). Between 1990 and 1993, seven UGS facilities in the USA had to be abandoned because of problems with leaks or substantial migration losses and a further three for unspecified safety reasons (EIA 1995). However, no further details could be ascertained to verify if these are included in Figure 6 and they are not included in figures here. Five incidents have involved injuries (five) and/or evacuees (c. 110; Figs 5 & 6). Of the incidents, 23 have occurred in North America (including Canada), of these, 11 were in California. The main cause of events relates to well or casing problems, including blowouts (60). This figure reflects perhaps the most dramatic event at the Leidy Field (Pennsylvania) in 1969 (A. Theodos, 2008 pers. comm.). Gas initially migrated from storage via five storage wells at 850 m3 (30 000 ft3) per day and later from 13 others. This led to a high pressure build up of gas in sandstones at shallower levels that caused fracturing of the rock over several square miles and led to the severing of casing in 30 other wells. Extensive surface blowouts in gas and water wells occurred, with gas continuing to escape for up to six weeks. Wells vented up to 56 634 m3 (2 mcf ) per day and an estimated 113.3 mcm (4 bcf) of gas was ultimately lost until pressures were reduced and down-hole plugs were set in wells. This appears to have been the first of a number of leaks at the Leidy Field: the most recent having occurred in September 2008. Elsewhere in America, a completion failed and storage operations were suspended until repairs were completed at Fort Morgan. In California, six incidents were related to problems with wells. In one case, wells were improperly plugged and had to be re-completed (Montebello; Gurevich et al. 1993; Khilyuk et al. 2000; Chilingar & Endres 2005). Three others were related to repairs of wells (two)

Fig. 9. Summary of documented events reported at underground hydrocarbon storage facilities developed in salt caverns that have not led to deaths/injuries or evacuees but have in some cases resulted in significant financial losses or closure. (a) events from1990s to 2000s; (b) events from 1950s to 1980s.

Fig. 9. (Continued)

184 D. J. EVANS Fig. 10. Summary of documented incidents or problems reported at underground hydrocarbon storage facilities developed in abandoned mines or unlined rock caverns constructed for underground gas storage. Also included are three leakage events at unconfirmed underground storage facility types in Canada and the USA.

PROBLEMS AND RISKS OF UFS

and damage during an earthquake (one). Directional, sidetracked wells were drilled to isolate and plug the three damaged well sections. In Europe, an incident in 2003 resulted from a problem with the casing of a well at the Breitbrunn/Eggstatt gas field (Bavaria, Germany). Repairs were quickly and successfully undertaken. Problems with above ground infrastructure were found reported in only three incidents: McDonald Island and Playa del Rey (PDR) California and the offshore Rough storage facility, UK Southern North Sea. Apart from well-related escape, migration of gas from the injection footprint was reported at 11 facilities (Figs 5 & 6), due to either overfilling beyond the spill point ((East) Whittier, Epps and Castaic and Honor Rancho), or leakage through the caprock (Castaic Hills and Honor Rancho, PDR, Montebello, Huntsman and El Segundo). At El Segundo, leakage was due to the caprock not being gas tight and probably faulted, and led to the facility being shut-in prior to 1993 (Gurevich et al. 1993; Khilyuk et al. 2000). At PDR, gas leaking from the reservoir into the adjoining Venice Beach accumulation was known from the earliest days of operation, with up to 25% of gas injected in 1953 having migrated (see below; Reigle 1953; Gurevich et al. 1993; Chilingar & Endres 2005). The SW