Nitrification in Saline Industrial Wastewater

NITRIFICATION IN SALINE INDUSTRIAL WASTEWATER

Nitrification in Saline Industrial Wastewater DISSERTATION Submitted in fulfilment of the requirements of the Board for Doctorates of Delft University of Technology and of the Academic Board of the UNESCO-IHE Institute for Water Education for the Degree of DOCTOR to be defended in public on Monday, 29 March 2004 at 10:30 hours in Delft, The Netherlands by

MOUSTAFA SAMIR MOUSSA born in Cairo, Egypt Master of Science, UNESCO-IHE

This dissertation has been approved by the promotor Prof.dr. H.J.Gijzen Prof.dr.ir. M.C.M.van Loosdrecht Members of the Awarding Committee: Chairman

Rector Magnificus Delft University of Technology

Co-chairman

Director UNESCO-IHE, Delft

Prof.dr. H.J.Gijzen

UNESCO-IHE, Delft, promotor

Prof.dr.ir. M.C.M.van Loosdrecht

Delft University of Technology, promotor

Prof.dr. J.G.Kuenen

Delft University of Technology

Prof.dr. P.Wilderer

Technical University München, Germany

Dr.ir. A.Klapwijk

Wageningen University

Dr. H.J.Lubberding

UNESCO-IHE, Delft

This research was sponsored by BTS Senter (BTS99130), Shell Global Solutions International, The Hague, Heiploeg Shrimp Processing, Zoutkamp and Ecco Tannery, Dongen. The project was carried out at the departments of Environmental Resources, (UNESCO-IHE, Delft) and of Biotechnology (Delft University of Technology). Copyright © 2004 Taylor & Francis Group plc, London, UK All rights reserved No part of this publication or the information contained herein may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording or otherwise, without written prior permission from the publisher. Although all care is taken to ensure the integrity and quality of this publication and the informationherein, no responsibility is assumed by the publishers nor the authors for any damage to property or persons as a result of operation or use of this publication and/or the information contained herein. Published by A.A.Balkema Publishers, a member of Taylor & Francis Group plc. http://www.balkema.nl/ and http://www.tandf.co.uk/ This edition published in the Taylor & Francis e-Library, 2006. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.” ISBN 0-203-02454-0 Master e-book ISBN

ISBN 90 5809 671 8 (Print Edition) (A.A.Balkema Publishers)

Contents Symboles

vii

Summary

x

Chapter1 Introduction Chapter2 Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures Chapter3 Short term effects of various salts on ammonia and nitrite oxidisers in enriched bacterial cultures Chapter4 Long Term Effects of Salt on Activity, Population Structure and Floc Characteristics in Enriched Bacterial Cultures of Nitrifiers Chapter5 Modelling Nitrification, Heterotrophic growth and Predation in Activated Sludge Chapter6 Nitrification activities in full-scale treatment plants with varying salt loads Chapter7 Model-based evaluation of the upgrading of a full-scale industrial wastewater treatment plant Chapter8 Evaluation and Outlook

4 32 48 69 95 124 140 156

Samenvatting (Summary in Dutch)

160

Acknowledgements

162

Curriculum

164

Symbols ASM

Activated Sludge Model

BOM

Biological Oxygen Monitor

BR

Batch Reactor

bNH4

Aerobic decay rate of ammonia oxidisers (day−1)

bNO2

Aerobic decay rate of nitrite oxidisers (day−1)

bH

Aerobic decay rate of heterotrophic biomass (day−1)

bpredators

Aerobic decay rate of predators (day−1)

Ci

Concentration of ionic species, i (mole)

Ci−

Concentration of anion species, i− (mole)

Ci+

Concentration of cation species, i+ (mole)

COD

Chemical Oxygen Demand

DO

Dissolved Oxygen

Fxi

Fraction of inert COD generated in biomass lysis

FISH

Fluorescent In Situ Hybridisation

HRT

Hydraulic Retention Time (day)

iN,XI

Nitrogen Content of XI (g N/g COD)

iN,BM

Nitrogen Content of biomass (g N/g COD) Affinity constant for ammonia of ammonia oxidisers (mg N/L) Affinity constant for oxygen of ammonia oxidisers (mg O2/L) Affinity constant for nitrite of nitrite oxidisers (mg N/L) Affinity constant for oxygen of nitrite oxidisers (mg O2/L) Affinity constant for nitrate of heterotrophic biomass (mg N/L) Affinity constant for organic carbon of heterotrophic biomass (mg COD/L) Affinity constant for oxygen of heterotrophic biomass (mg O2/L) Affinity constant for oxygen of predators (mg O2/L)

MLSS

Mixed Liquor Suspended Solids (mg/L)

MLVSS

Mixed Liquor Volatile Suspended Solids (mg/L)

mNH4

maintenance coefficient of ammonia oxidisers (mg NH4–

N/gXN– COD.day) mNO2

maintenance coefficient of nitrite oxidisers (mg NO2– N/gXN– COD.day)

mH

maintenance coefficient of heterotrophic biomass (mg COD/gXH– COD.day)

MNH4

Monod term for ammonia in bacterial growth

OUR

Oxygen Uptake Rate (mg O2/L.h)

pKa

The negative logarithm of stoichiometric dissocial constant

RNH4

volumetric uptake rate of ammonia (mg NH4–N/L.h)

RNO2

volumetric uptake rate of nitrite (mg NO2–N/L.h)

Rmax,NH4

maximum volumetric uptake rate of ammonia (mg NH4– N/L.h)

Rmax,NO2

maximum volumetric uptake rate of nitrite (mg NO2– N/L.h)

SBR(s)

Sequencing Batch Reactor (s)

SRT

Sludge Retention Time

SO2

Concentration of Oxygen (mg O2/L)

SNH4

Concentration of ammonia (mg NH4–N/L)

SNO3

Concentration of nitrate (mg NO3–N/L)

SNO2

Concentration of nitrite (mg NO2–N/L)

SN2

Concentration of nitrogen (mg N/L)

SS

Concentration of organic substrate (mg COD/L)

T

The temperature in °K

VFA

Volatile Fatty Acids

VSS

Volatile Suspended Solids

WWTP(s)

Wastewater Treatment Plant (s)

XNH4

Concentration of ammonia oxidisers (mg-VSS/L)

XH

Concentration of heterotrophic biomass (mg-VSS/L)

Xpredators

Concentration of predators (mg-VSS/L)

XI

Concentration of particulate inert (mg-VSS/L)

YNH4

Yield coefficient of ammonia oxidisers per NO2–N (g COD/g NH4–N)

YNO2

Yield coefficient of nitrite oxidisers per NO3−–N (g COD/g NO2–N)

YH

Yield coefficient of heterotrophic biomass on SS (g COD/g COD)

Ypred

Yield coefficient of predators on bacteria (g COD/g COD)

Zi

Charge of species, i

Zi−

Charge of anion species, i−

Zi+

Charge of cation species, i+

ηNH4

Anoxic reduction factor for ammonia oxidisers decay

ηNO2

Anoxic reduction factor for nitrite oxidisers decay

ηH

Anoxic reduction factor for heterotrophic growth Maximum growth rate of ammonia oxidisers (day−1) Maximum growth rate of nitrite oxidisers (day−1) Maximum growth rate of heterotrophic biomass (day−1)

µpredators

Growth of predators, presented in the model as predation rate (day−1)

Summary Biological nitrification-denitrification is one of the most common processes for nitrogen removal from wastewater. However, nitrification, the rate limiting step in biological nitrogen removal, proved to be one of the most difficult processes to design and control in wastewater treatment plants, because nitrifying bacteria are slow-growing and very sensitive to environmental factors (temperature, pH, dissolved oxygen concentration, toxic and inhibitory compounds). Researchers so far have concentrated mainly on nitrification in domestic wastewater treatment and achieved broad knowledge and practical experience about the process. The result is that biological nitrogen removal is widely and successfully applied for municipal wastewater. However, these experiences are not directly applicable to industrial wastewater due to its specific composition. Several industries are dealing with a high salt concentration in their wastewater. Also the policy of more economic use of water and water reuse will result in an increase of salt content of the ultimately produced wastewater. High salt levels may negatively affect nitrification, demonstrating the need for improved understanding of the precise effects of salt on nitrification. The response of the nitrification process to saline conditions and the adaptation mechanisms of nitrifying bacteria towards these conditions are still unknown. The available studies on the effect of salt on nitrification show a decline in activity for ammonia and nitrite oxidisers. However, it does not give clear answers on: what are the main inhibiting factors causing the effects, do all salts have similar effects, what is the maximum acceptable salt level, are ammonia oxidisers or nitrite oxidisers most sensitive to salt stress, can nitrifiers adapt to long term salt stress and are some specific nitrifiers more resistant to salt stress than others? The main focus of this dissertation is the understanding of the effects of salinity on nitrification considering all these questions. The research was carried out in two phases. In the first phase, laboratory scale activities were conducted to obtain fundamental data to determine the relationship between salinity and nitrification. In the second phase the results collected from the laboratory experiments were compared and validated with the results collected from full-scale treatment plants. Modelling was employed in both phases to provide a mathematical description for salt inhibition on nitrification and to facilitate the comparison. First phase: A method to measure the activity of ammonia and nitrite oxidisers in mixed bacterial cultures was developed and applied in the research as standard method to determine the inhibition effects of salt on ammonia and nitrite oxidisers. The short-term effects of various types of salt on the activity of ammonia and nitrite oxidisers were studied. Different types of salts appeared to have different inhibition effects on the ammonia and nitrite oxidisers. Non-adapted and adapted (to 10 g NaCl–Cl/L for one year) enriched cultures of nitrifiers were used to investigate the long-term effect of salt (gradually increased with 5 g Cl−/L up to 40 g Cl−/L). No difference in activity was observed between the adapted and non-adapted sludge. At 40 g Cl−/L inhibition reached

95% of salt free activity for ammonia and nitrite oxidisers in both adapted and nonadapted reactors. Nitrosomonas europaea and Nitrobacter sp were the only nitrifiers present at high salt levels. Increased salt concentrations resulted in better settling characteristics of the nitrifying sludge. At the same time the protozoan and metazoan predators in the laboratory scale experiments were found to be affected by salt. This effect was used to develop a mathematical model to describe the interaction between nitrifiers, heterotrophs and predators. Second phase: Nitrifier activities and population structure in full-scale domestic and industrial wastewater treatment plants (WWTPs) operated under various salt levels were investigated and compared with results obtained from laboratory scale activities. Finally, the activated sludge model No. 1 (ASM 1) was modified and applied to simulate COD and nitrogen removal in a full-scale industrial WWTP operated under salt stress. The research has lead to an improved understanding of the effect of salinity on nitrification. The results obtained within the course of this research can be used to improve the sustainability of the existing WWTPs operated under salt stress. The findings also form a guideline for more economical and sustainable design and start up of WWTPs dealing with salt in future.

Chapter 1 Introduction

Abstract The global situation for nitrogen is getting out of hand. There is a serious imbalance between the influx and efflux of N in the biosphere. The direct cause is the rapidly increasing production of chemical fertilisers. The annual production of fertiliser N has increased 9 fold over the past 40 years and amounts currently to some 37% of the world-wide biological Nfixation. Such a massive introduction of reactive forms of nitrogen into the environment over a relatively short period of time has numerous deleterious consequences, causing environmental and public health problems, both locally and at a global scale. The response to increasing pollution problems necessitated the promulgation of effluent standards for nutrients. In this framework environmental legislation in most countries includes stringent limitations for nitrogen to be discharged. However, the implementation of effluent standards at a global scale is limited due to the phenomenal costs of the high-rate wastewater treatment technology. It remains a challenge to come up with nitrogen pollution control strategies, which are effective and low cost. Other sources of nitrogen pollution than domestic should be considered. Industrial wastewater not only represents twice the volume of domestic wastewater, but also is usually more concentrated. Having a cost-effective N-removal technology in industry is still a target and needs more attention. Biological nitrificationdenitrification is the most common processes for nitrogen removal from wastewater; nitrification is the rate-limiting step in biological nitrogen removal. Nitrification in industrial wastewater presents a number of difficulties, including a wide range of different and varying temperatures, pH, presence of toxic compounds and salinity. Studies on the effect of salt on nitrification show a decline in activity for ammonia and nitrite oxidisers. However, no information is available on the maximum acceptable salt level and which nitrifying group is most sensitive to salt stress. The need for understanding the precise effects of salt on nitrification was addressed, as the main aim of this study.

1 Introduction 1.1 The nitrogen cycle Nitrogen is an essential component to all living organisms, as it is an important atom of DNA, RNA, proteins and other key organic molecules. In general, living organisms contain between 10–15% of their biomass as nitrogen. Although N represents only a minor constituent of living matter, it has been and continues to be the main limiting factor for biomass production on a global scale. Also in agricultural production, it appears that the other two limiting nutrients, potassium and phosphorus, are less frequently the prime limiting factor (Smil, 1997). Nitrogen is present on earth in many forms and huge amounts are stored in sediment and rock deposits and in the atmosphere. Nitrogen is present in a variety of compounds with different oxidation states. The movement and transformation of these nitrogen compounds through the biosphere is characterised by the nitrogen cycle (Figure 1.1). The atmosphere serves as a reservoir of nitrogen in the form of nitrogen gas, which makes up about 78% of the atmosphere, but nitrogen in this form is too inert to play a direct role in ecosystems. Plants and animals cannot use nitrogen gas directly from the air as they do with carbon dioxide and oxygen. It is only accessible to N2—fixing bacteria. The nitrogen must be available in a reactive form with hydrogen or oxygen before it can be assimilated by plants or used by other organisms. The plants, in turn, can be consumed by animals for the generation of animal protein.

Figure 1.1 The nitrogen cycle Transformation of these nitrogen compounds can occur through several mechanisms. Those of importance include N-fixation, ammonification, synthesis, nitrification, and denitrification. Each can be carried out by particular microorganisms.

Introduction

5

Nitrogen fixation Fixation of nitrogen (physical, chemical or biological) means the incorporation of inert, gaseous nitrogen into chemical compounds that eventually can be used by living organisms. Biological fixation of N2 is prominently accomplished by specialized microorganisms: cyanobacteria, symbiotic and free-living bacteria. Lightning also indirectly transforms atmosphere nitrogen into nitrate, which rains onto soil. Finally, N2 can be fixed industrially by the Haber-Bosch process, invented in 1913. At present the industrial fixation of nitrogen into ammonia plays a significant role, because it is responsible for 30% of the total nitrogen influx into the biosphere (Gijzen and Mulder 2001). Ammonification In most ecosystems nitrogen is preliminary stored in living and dead organic mater. Ammonification is the process responsible for the change of organic nitrogen compounds into the ammonia form. In general, ammonification occurs during decomposition of animal and plant tissue and animal faecal matter by bacteria; after hydrolysis of the proteins, the amino acids are either reused or the amino groups are converted into ammonia. Also the nitrogen present in urine is—via urea—converted into ammonia. Nitrification Nitrification is the biological oxidation of ammonium. This is done in two steps, first to the nitrite form, then to the nitrate form. Both steps can be carried out by different genera, both using CO2 as their source of cellular carbon. These transformation reactions are generally coupled and proceed rapidly to the nitrate form; under normal conditions nitrite levels are usually very low. The produced nitrate is used either by plants in the assimilation process or reduced by denitrification to N2. Denitrification Denitrification is the biological reduction of nitrate to nitrogen gas. It can proceed through several steps in the biochemical pathway, with the ultimate production of nitrogen gas. A fairly broad range of heterotrophic bacteria is involved in the process, requiring an organic carbon source for energy (Kuenen and Robertson 1994; Schmidt et al 2003). Nitrate reduction to ammonia In contrast to denitrification, the process of dissimilatory nitrate reduction to ammonia (DNRA) does not have N2 but NH4+ as final product. Apart from a nitrate reductase, a nitrite reductase, which reduces nitrite to ammonia, is involved in this process. Denitrification and DNRA can occur simultaneously and DNRA can be of quantitative importance in environments with high carbon/nitrate ratio or high sulphide concentration (Brunet and Garcia-Gil 1996; Cole 1996; Simon 2002).

Nitrification in saline industrial wastewater

6

ANAMMOX The denitrifying bacteria (as described above) are not the only bacteria producing nitrogen gas. Ammonia can be oxidized under anaerobic conditions also leading to N2 and it became clear that slow growing autotrophic bacteria belonging to the order of the Planctomycetales are carrying out this process. This process, in which both ammonia and nitrite are converted to N2, is called ANAMMOX, an acronym for ANaerobic AMMonia OXidation (Mulder et al 1995; Schmidt et al 2003). Assimilation Assimilation is the process in autotrophic organisms in which nitrogen compounds (NH4+, NO3−) are incorporated into cell material for growth, a biochemical mechanism that uses ammonia or nitrate. Animals and other heterotrophic organisms require protein from plants and other animals as their nitrogen source. They are not capable of transforming inorganic nitrogen into an organic nitrogen form. 1.2 The nitrogen cycle out of balance The influx and efflux of N in the biosphere has been kept in balance by nature. Several decades ago this balanced situation started to undergo a radical change mainly due to binding of atmospheric nitrogen gas for the manufacturing fertilisers (the invention of ammonia synthesis by Fritz Haber). The first commercial ammonia factory started its operations in 1913 in Germany, but production levels at a global scale remained low until the process became more energy efficient due to technological innovations in the 1960s. Since then the production of industrial nitrogen fertiliser via the so-called Haber-Bosch process showed a sharp increase. This process has removed the fundamental restriction on food production and therefore on population growth. Indeed, the doubling of the world population over the last 40 years would not at all have been possible without the intensive agriculture and animal production systems which primarily depend on nitrogen fertiliser. The increase in production of nitrogen fertiliser has been much faster than population increase. While population doubled between 1960 and 2000, the annual production of fertiliser nitrogen increased nine-fold from 1×1010 to 9×1010 kg. Current production is equivalent to about 37% of the total amount of nitrogen input achieved via terrestrial and marine biological N2 fixation (about 24×1010 kg per year). There is probably no other elemental cycle where the human impact has been so dramatic as the case for nitrogen (Gijzen and Mulder 2001). The massive introduction of reactive forms of nitrogen into the environment over a relatively short period of time has numerous deleterious consequences, causing environmental and public health problems, both locally and at a global scale (Scheible and Heidman, 1994; Vitousek et al 1997; Wiesmann 1994): • The formation of blooms of toxic cyanobacteria in fresh waters is of considerable concern with respect to human and animal health (e.g. potable water supply, fish production). Eventually the produced cyanobacteria, algal and plant biomass will die

Introduction

7

and become subject to biodegradation, causing substantial oxygen depletion and biodiversity loss in water bodies. • The oxidation of ammonia released into the environment, either directly (eutrophication) or from biomass degradation, will result in low oxygen levels of affected water bodies (theoretical consumption 4.57 g O2/g−N). The released toxins from cyanobacteria and the lower levels of dissolved oxygen will obviously affect many species of aquatic life. • The lower oxygen levels in water bodies may also result in incomplete nitrificationdenitrification and therefore stimulate the formation of NO and N2O gasses. Together with increased methane production from the decomposition of plant biomass in the sediments, this could contribute significantly to the global greenhouse effect. • The effect of high levels of nitrite and nitrate in drinking water may cause the so-called ‘blue baby’ disease (methemoglobinemia) in infants. The relation between nitrates and some forms of cancer has also been reported. Despite these negative impact of the fertiliser use there is, at present, no other substitute available and therefore mankind will develop an increasing dependency on the HaberBosch synthesis while population grows over the coming decades. There is some hope that the nitrogen fixing capacities of Rhizobium could be incorporated directly into plant species via genetic engineering, but it seems realistic to assume that this ‘solution’ is still several decades away from becoming reality. Therefore, reliance on chemical fertiliser must further increase to provide sufficient protein for the growth of the additional 2 to 3 billion people that will be born during the next 50 years (Gijzen and Mulder 2001). At the same time however, finding a sollution for the removal of reactive forms of nitrogen from the envrionment poses another urgent challenge. 1.3 Actions for rebalancing the nitrogen cycle When analysing the nitrogen pollution, care must be taken that the source of pollution is identified, in order to take measures to solve the problem. Different sources can contribute to the nitrogen problem, differing from location to location. The challenge is to find out the most effective approach to tackle the most urgent and most important one. Moreover, the approach for the same problem in one specific location may not be applicable to another. 1.3.1 Sources of nitrogen pollution The sources of the nitrogenous compounds in water can be of human, industrial, as well as agricultural origin. Natural sources can be atmospheric precipitation, dust, non-urban and non-agricultural leachates and biological fixation. Nitrogenous compounds of human origin can be for example treated and non-treated sewage, agricultural leachates (for instance from excess addition of fertilisers), some industrial wastewaters and surface runoff. Sewage always contains nitrogenous compounds, typically in concentrations between 25–50 mg N/L. The nitrogen consists of approximately 60% ammonia nitrogen, 40% organically bound nitrogen, and a small amount of nitrate (Scheible and Heidman, 1994). When sludge from sewage treatment plants is digested, a rich nitrogenous flow is

Nitrification in saline industrial wastewater

8

produced, with ammonia concentrations of 700 to 1000 mg-N/L. Agricultural leaching can result in large amounts (5–25 kg per hectare) of nitrogenous compounds released per year (Scheible and Heidman, 1994). In industrial effluents ammonia concentrations are often much higher (Wiesmann, 1994); the ammonia and nitrate levels in some industrial wastewaters is given in Table 1.1.

Table 1.1 Ammonia and nitrate concentrations in industrial wastewaters (Wiesmann, 1994). Industry/Products sludge digestion

Ammonia concentration range (g N/L) 1

tannery

0.35

cokery

0.45–4.1

oil refinery

0.02–0.9

coal gasification

1–2.5

fertiliser

0.2–1

synthetic fibre

0.8

slaughterhouse

0.15

livestock: swine

2.3

livestock: cattle

0.5–2.3

rendering plant

0.8

dairy

0.6

distillery

1.5

cellulose and paper

0.25

Pharmaceuticals

1–6

0.1–0.4

explosives

glass

Nitrate concentration range (g N/L)

2–12.5

0.3–0.65 0.48

electronics

0.5–2

pectin

1–2.7

uranium processing

4–11.3

Not only the nitrogen concentration, but also the amount of water polluted with nitrogen compounds is playing a role. The quantities of the three main sources of pollution (agriculture, industry and municipalities) are presented in Table 1.2. It illustrates that 70% of the total water use is for agriculture, 20% for industry and only 10% for domestic purposes. Industries generate more wastewater with higher quantities of nitrogen than municipalities.

Introduction

9

Table 1.2 Global water use in the 20th century (Cosgrove and Rijsberman, 2000). Use (Cubic kilometres) Agriculture

Industry

Municipalities

1900

1950

1995

Withdrawal

500

1,100

2,500

Consumption

300

700

1,750

Withdrawal

40

200

750

Consumption

5

20

80

Withdrawal

20

90

350

Consumption

5

115

50

0

10

200

Withdrawal

600

1,400

3,800

Consumption

300

750

2,100

Reservoir (evaporation) Total

1.3.2 Establishing nitrogen limits for emission The best way to prevent the release of nitrogenous compounds into surface waters is avoiding production of these substances. Integral process improvements are necessary to decrease the amounts of nitrogenous compounds produced. Furthermore, if possible the nitrogenous compounds should be recycled and not broken down into the constituting elements. However, end-of-pipe techniques (like wastewater treatment facilities) remain necessary to get rid of the nitrogenous compounds. Treatment objectives and priorities in industrialised countries have been gradually tightened over the past decades. This resulted in the so-called first, second and third generation of treatment plants (Table 1.3).

Table 1.3 The phased expansion and upgrading of wastewater treatment plants in industrialised countries to meet ever stricter effluent standards (WHO/UNEP, 1997). Decade Treatment objective

Treatment Operations included

1950– 60

Suspended/coarse solids removal

Primary

Screening, removal of grit, sedimentation

1970

Organic matter degradation

Secondary

Biological oxidation of organic matter

1980

Nutrient reduction (eutrophication)

Tertiary

Reduction of total N and total P

1990

Micro-pollutant removal

Advanced

Physicochemical removal of micropollutants

Nitrification in saline industrial wastewater

10

As a consequence, a number of treatment technologies, unit operations and processes have been developed to achieve the required treatment level (Table 1.4).

Table 1.4. Classification of common wastewater treatment processes according to their level of advancement (WHO/UNEP, 1997). Primary

Secondary

Tertiary

Advanced

Bar or bow screen

Activated sludge

Nitrification

Chemical treatment

Grit removal

Extended aeration

Denitrification

Reverse osmosis

Primary sedimentation

Aerated lagoon

Chemical precipitation

Electrodialysis

Comminution

Trickling filter

Disinfection

Carbon adsorption

Oil/fat removal

Rotating bio-discs

(Direct) filtration

Selective ion exchange

Flow equalisation

Anaerobic treatment/UASB

Chemical oxidation

Hyperfiltration

pH neutralisation

Anaerobic filter

Biological P removal

Oxidation

Imhoff tank

Stabilisation ponds

Constructed wetlands

Detoxification

Constructed wetlands

Aquaculture

Aquaculture

The fast population growth, urbanisation and industrialisation, all of which impose high demands on local water resource quality and quantity, while simultaneously generating pollution, which affects the very same water resource. The response to increasing pollution problems in receiving waters, and the growing concern about water quality protection, necessitated the promulgation of effluent standards for nutrients, especially for sensitive areas. In this framework environmental legislation in most countries includes stringent limitations for nitrogen to be discharged. National technology-based standards were established, moving all wastewater treatment facilities to secondary level at minimum. EU Policy on nutrients emissions Pollution and degradation of Europe’s waters as well as an increasing awareness by citizens and policy makers led to increased efforts to address water pollution. This resulted in a “second wave” of EU water legislation. Its first results were, in 1991, the adoption of • The Urban Wastewater Directive, addressing water pollution from all settlements except the small villages, as well as a range of industries with biodegradable wastewater; and, • The Nitrates Directive, addressing water pollution by nitrates from agriculture.

Introduction

11

Urban Wastewater Directive The urban wastewater directive has set ambitious objectives; • Wastewater collection and treatment for all settlements above 2000 population equivalents (p.e.) • Biological (secondary) treatment as a general rule, plus nutrients removal where the affected receiving waters show an elevated level of nitrates and/or eutrophication. The deadlines for achieving these objectives are 1998, 2000 and 2005, respectively (depending on the size of the discharge and the character of the receiving waters). Member states may choose between nutrient control by effluent concentration or by removal efficiency (Table 1.5). One or both nutrient parameters may apply depending on the local and regional situation.

Table 1.5 Discharge limits for treatment plants in EU. Total phosphorus

Effluent concentration

Minimum removal efficiency

2 mg/L for plants≤100,000 p.e.

80%

1 mg/L for plants>100,000 p.e. Total nitrogen

15 mg/L for plants≤100,000 p.e.

70–80%

10 mg/L for plants>100,000 p.e.

The Nitrate Directive The main objective of the nitrate directive is to reduce water pollution caused or induced by nitrates from agriculture and to prevent further nitrogen pollution. To ensure this objective, EU members have to identify waters (surface waters and ground waters) affected by nitrate pollution, and waters, which could be affected by nitrate pollution. The EU member states have the choice either to designate individual vulnerable zones in accordance with these criteria, or to apply the more stringent provisions of the Directive over all their territory (this option has been taken up by Denmark, Germany, Luxembourg, the Netherlands and Austria). Outside those affected areas (vulnerable zones) they have to promote codes of good agricultural practice on a voluntary basis (Blöch 2001). Implementation of EU policy on nutrient emissions A good practice of the EU policy implementation has been noticed in many member states. For instance, in the Netherlands, the water boards started in 1999 to take initiatives to benchmark their treatment performances. This shows the performance of the sewage treatment plants as a whole, with regard to the removal of oxygen-binding substances, phosphate and nitrogen. A treatment performance of 100 % means: 75% phosphate removal, 75% nitrogen removal and 90% removal of oxygen-binding substances. The treatment performance of the Dutch water boards has improved from 86% in 1996 to more than 91% in 2002. These water boards have to achieve 75% phosphate-removal and 75% nitrogen-removal by 2006. Phosphate removal of 79% was achieved in 2002.

Nitrification in saline industrial wastewater

12

Considering N-removal, in 2002 the removal efficiency averaged 69.4%. It is not certain, whether the 75% target for nitrogen removal will be achievable by 2006, taking the limited planned investments into consideration. In 2002, 40% of all sewage treatment plants failed at some time to meet the individual discharge requirements. Rapid improvement of the nitrogen removal in the next few years (before 2006) is a point of attention (Postma et al. 2003). Due to the successful implementation of the Directive in many EU states European waters have started to change. Reports by the European Environment Agency (1998, 1999a) clearly show improvements. The number of heavily polluted rivers has declined significantly, in particular as the pressure from organic matter and phosphates in urban wastewater has decreased. While progress has been made in many areas, others are still in a deplorable situation. After 25 years of European water legislation, Europe’s waters are in need of more protection, in need of increased efforts to get them clean or to keep them clean. This is a demand not only from the scientific community and other experts, but also to an ever-increasing extent from citizens and environmental organisations. 1.3.3 New Technologies National water quality goals have influenced the development of advanced treatment technologies especially in the area of nutrient control. With respect to nitrogen control, four recent developments have to be mentioned: The SHARON process SHARON (Hellinga et al 1998; van Loosdrecht and Jetten, 1998) is an acronym for Single reactor High activity Ammonia Removal Over Nitrite. In this process, a completely mixed reactor is operated at short residence time (1–1.5 days) and high temperature (30–40°C) leading to the selective wash out of nitrite oxidisers. This results in only partial oxidation of ammonia to nitrite and subsequently reduction of the latter to nitrogen gas in the denitrification process. This route is more favourable than the conventional route due to savings of 25% in oxygen supply and 40% reduction in COD demand. The ANAMMOX process Another recently discovered process allows nitrite reduction with ammonium as electron donor to nitrogen gas. This anaerobic ammonium oxidation (ANAMMOX, Mulder et al 1995; Schmidt et al 2003) process can be combined with partial nitrification (SHARON) leading to a direct net conversion of ammonium to N2 gas. Which makes complete autotrophic ammonia removal possible as a sustainable pathway of nitrogen removal from wastewater. The CANON process CANON is an acronym for Completely Autotrophic Nitrogen removal Over Nitrite (Strous et al 1997). This concept is the combination of partial nitrification and

Introduction

13

ANAMMOX in a single, aerated reactor. This process has been tested extensively on laboratory scale (Slikers et al 1998, 2003). Although ANAMMOX requires strict anoxic conditions, nitrifiers and ANAMMOX organisms are able to coexist under oxygenlimited conditions. Therefore, CANON would need process control to prevent nitrite build-up by oxygen excess under ammonia limitation (fluctuation of ammonia load). The OLAND process The OLAND process (oxygen-limited nitrification and denitrification) is described as a new process for one-step ammonium removal without addition of COD (Kuai and Verstraete 1998). Recently, it was confirmed that OLAND is based on the CANON concept. (Pynaert et al 2004) The formation of thick biomfilm could create a favourable condition for nitrifiers and ANAMMOX organisms to coexist even under normal oxygen conditions. A good overview of recent nitrogen removal technologies can be found in Schmidt et al (2003). 1.3.4 Future activities The implementation of effluent standards is so far limited to developed countries. In developing countries effluents remain largely untreated due to the phenomenal costs of sewerage systems and of high-rate wastewater treatment technology (Gijzen 2003). The challenge for these countries will be to come up with sewage management strategies, which are effective and low cost. The urgency of reconsidering the current practises (conventional technologies of domestic wastewater treatment) in the light of sustainability becomes evident. Implementation of cleaner production approaches is promising to achieve sustainable urban water and nutrient management. This approach incorporates pollution prevention or minimisation, treatment for reuse and stimulation of natural self-purification capacity of the receiving environment. The implementation of this approach is challenging (Gijzen 2003; Gijzen and Bijlsma 2000), but governments should put more efforts to bring it into practice. Nutrients from other sources than domestic are still a major challenge in nitrogen pollution control. Industrial wastewater is not only the double amount of domestic wastewater, but also in most cases it is much more concentrated (Table 1.1, 1.2). Concurrently, industrial activity is growing very fast to meet the human demands. Coming up with a cost-effective N-removal technology in industry is still a challenge and needs more attention. The main difficulties of industrial wastewaters are temperature, pH, presence of toxic compounds, salinity and fluctuations in flow and composition. This study focuses on nitrogen removal in industrial wastewaters emphasising on nitrification under salt stress, which is a common destabilising factor in industrial wastewaters. Full understanding of nitrification in saline wastewater will lead to a costeffective technology for industries reducing the pollution problems in the receiving waters.

Nitrification in saline industrial wastewater

14

1.4 Nitrification in saline industrial wastewater 1.4.1 Background Industry is an essential engine of economic growth worldwide and requires adequate resources of good quality water as a key raw material. Global annual water use by industry is expected to rise from an estimated 725 km3 in 1995 to about 1,170 km3 by 2025, by which time industrial water usage will represent 24% of all water abstractions. Industrial use of water increases with country income, ranging from 10% for low-and middle-income countries to 59% for high-income countries (World Bank, 2001). A number of industrial categories (petroleum refining, coke processing, dairy, chemical production, tannery, fish processing) contain significant amounts of nitrogen in their wastewater (Wiesmann 1994). Before discharge to the water body (according to the stricter regulations), almost full nitrogen removal is necessary. This has led to increasing activities in the field of development and optimisation of biological nitrogen removal. Biological nitrogen removal is conventionally achieved by making use of processes of the natural nitrogen cycle, namely through nitrification in an aerobic environment followed by denitrification in an anoxic environment. Biological nitrification-denitrification is the most common processes for nitrogen removal from wastewater; nitrification is the ratelimiting step in biological nitrogen removal, because nitrifying bacteria are growing slowly and are very sensitive to environmental factors (e.g. temperature, pH, dissolved oxygen concentration, toxic and inhibitory compounds) (Antonious et al 1990; Wagner et al 1996; Wagner and Loy 2002). Studies so far have concentrated mainly on domestic wastewater treatment and the results obtained may not be directly applicable to industrial wastewater due to their specific composition (high temperatures, sub-optimal pH values, presence of toxic compounds or high salinity). Thus, special attention for and concern with the design and operation of nitrogen removal for industrial wastewater treatment systems is necessary. The effect of salts on nitrogen removal is a major concern, especially in industrial wastewater treatment. Industries such as pickling, cheese manufacturing, seafood processing, tanning, productions of chemicals and Pharmaceuticals, oil and gas recovery, produce high inorganic salt concentrations in their wastewater. Other sources of saline wastewater include infiltration of subsurface water in the coastal areas into the sewer system, landfill leachates and contaminated ground water and ballast water for marine vessels or offshore installations. In future, waste minimisation practices are expected to generate brines via effective water reuse and recycling schemes. Also the use of saline water for toilet flushing due to the scarcity of fresh water will increase the wastewater salinity that reaches the treatment plants (Campos et al 2002; Dahl et al 1997; Woolard and Irvine 1995; Yu et al 2002). Nitrification is the bottleneck of the nitrogen removal process under salt stress, while denitrification has proved to be more stable under salt stress (Vredenbregt et al 1997; Dahl et al 1997). Studies on the effect of salt on nitrification are contradictory and difficult to be interpreted.

Introduction

15

1.4.2 Nitrification The nitrification process—an important process in the nitrogen cycle in nature—is defined as the biological transformation of reduced forms of nitrogen into nitrite and subsequently to nitrate. Generally, two absolutely different types of nitrification must be distinguished (Schmidt et al 2003): • Lithotrophic nitrification, in which the oxidation of inorganic, reduced nitrogen compounds serves as energy source for growth. Lithotrophic nitrification is carried out by two groups of bacteria, the ammonium-oxidizers and nitrite-oxidizers • Heterotrophic nitrification, in which nitrification is a co-oxidation and does not serve as an energy source. It is carried out by diverse groups of microorganisms (bacteria, fungi, and algae). In natural environments, the chemolithotrophic nitrifiers are the only group of microorganisms producing considerably high amounts of nitrite and nitrate from ammonia. The heterotrophic nitrifiers’ specific activity is estimated to be around 103– 104 times lower than that of lithotrophic nitrifiers and therefore heterotrophic nitrification is of minor ecological significance (Kuenen and Robertson 1994; Richardson et al 1998). Originally, the lithoautotrophic nitrifying bacteria altogether were grouped within one family, named Nitrobacteraceae and composed of two physiologically distinct groups of bacteria that are not phylogenetically related (ammonia-oxidizing bacteria and nitriteoxidizing bacteria). However, phylogenetic investigations made evident that a lot of distinct groups of organisms exist, which are not closely related to each other (Koops and Röser 2001). Cells of both groups are able to aggregate in clusters (flocs), which is common in wastewater treatment plants (Stalely et al 1989). 1.4.2.1 Ammonia oxidisers After the first reports on successful isolation of chemolithoautotrophic ammoniaoxidizing bacteria at the end of the 19th century, researches have continued to investigate their diversity in natural and engineered systems by applying enrichment and isolation techniques. These efforts resulted in the description of numerous species of ammonia oxidisers, now with the modern molecular biological techniques more species have been discovered. Chemolithoaotutrophic ammonia-oxidizing bacteria comprise two monophyletic lineages within the class Proteobacteria (Table 1.6). One group is located within the γ subclass, which contains only the Nitrosococcus oceanus and Nitrosococcus halophilus. The second group belongs to the β subclass, which contains two clusters, the Nitrosospira cluster and the Nitrosomonas cluster (Koops and Röser 2001; Purkhold et al 2000).

Nitrification in saline industrial wastewater

16

Table 1.6 The cultured ammonia oxidising bacteria and information on ecophysiological parameters and preferred habitats (Koops and Röser 2001). Species

Ecophysiological parameters

Preferred habitat

Salt requirements

Substrate (NH3) affinity

Nitrosomonas europea Nitrosomonas eutropha Nitrosomonas halophila Nitrosomonas mobilis

halotolerant or moderately halophilic

30–61 µM

sewage treatment plants, eutrophic freshwater and brackish water

Nitrosomonas communis Nitrosomonas sp.I Nitrosomonas sp.II

no salt requirement

14–43 µM

soils (not acid)

Nitrosomonas nitrosa

no salt requirement

19–46 µM

eutrophic freshwater

no salt requirement

1.9–4.2 µM

oligotrophic freshwater natural solis

Nitrosomonas marina Nitrosomonas sp.III Nitrosomonas aestuarii

obligatory halophilic

50–52 µM

marine environment

Nitrosomonas cryotolerans

obligatory halophilic

42–59 µM

βProteobacteria Nitrosomonas ureae Nitrosomonas oligotropha

Nitrosolobus multiformis Nitrosovibrio tenuis Nitrosospira sp.I Nitrosococcus γoceani Proteobacteria Nitrosococcus halophilus

no salt requirement

soils (not acid) Soils, rocks and freshwater

obligatory halophilic

marine environment

Introduction

17

Although the basic metabolism is more or less uniform within the physiologically defined groups of lithoautotrophic ammonia oxidizing bacteria, ecophysiological differences exist between the distinct representatives. Different members of these genera have been found to dominate different wastewater treatment plants or natural ecosystems, but general relationships between the ecological niche and evolutionary position are often still obscure (Schmidt et al 2003). Salt requirement is an ecophysiologically relevant discrimination factor. All isolates of the two species of Nitrosococcus (γ subclass of the Proteobacteria), Nitrosococcus oceani and Nitrosococcus halophilus, are obligately halophilic. The group located in the β subclass of the Proteobacteria, comprises obligately halophilic species and moderately halophilic or halotolerant species, respectively, together with species missing salt requirement or being salt sensitive. Within the genus Nitrosomonas, these differences are well reflected by the pronounced formation of phylogenetic lineages (Koops and Röser 2001). Physiology The physiology of conventional, ‘aerobic’ ammonia oxidizers is not completely understood. Only recently, it was discovered that these organisms also have an anaerobic metabolism. The proteobacterial ammonia oxidizers can obtain their energy for growth from both aerobic or anaerobic ammonia oxidation. Most likely ammonia (NH3) and not ammonium (NH4+) is the substrate for the oxidation process (Suzuki et al 1974; Bock et al 1991). The main products are nitrite under oxic conditions (DO> 0.8 mg O2/L), while under anoxic conditions (DO<0.8 mg O2/L) nitrogen gas, nitrite and nitric oxide are the main products (Schmidt and Bock 1997). 1.4.2.2 Nitrite oxidisers The second step of nitrification, the oxidation of nitrite to nitrate, is performed by nitrite oxidizing bacteria. Four phylogenetically distinct groups of nitrite-oxidizing bacteria have been described (Table 1.7). The major group, which belongs to the α subclass of the Proteobacteria, is represented by a single genus, Nitrobacter. Four species, Nitrobacter winogradskyi, Nitrobacter hamburgensis, Nitrobacter vulgaris and Nitrobacter alkalicus, have been described. Two marine species, Nitrococcus mobilis and Nitrospina gracilis, were assigned to the γ and the δ subclass of the Proteobacteria, respectively. The two species of the genus Nitrospira, Nitrospira marina and Nitrospira moscoviensis, are members of a distinct phylum close to the δ subclass of the Proteobacteria (Koops and Röser 2001). Physiology For nitrite oxidising bacteria the oxidation of nitrite to nitrate is the energy generation process. There is some evidence that Nitrospira is the more specialized nitrite oxidizer. The other genera are more versatile, are all able to use organic energy sources beside the major source nitrite, being facultative autotrophs and anaerobes, able to grow on heterotrophic substrates such as pyruvate and also capable of the first step of denitrification (the reduction of nitrate to nitrite) (Schmidt et al 2003; Koops and Röser 2001).

Nitrification in saline industrial wastewater

18

Table 1.7 The cultured nitrite oxidising bacteria and information on ecophysiological parameters and preferred habitats (Koops and Röser 2001). Species

Ecophysiological parameters Preferred habitat Salt requirements Nitrobacter alkalicus

α

Supclass of Proteobacteria

Nitrobacter winogrodskyi

alkali-and halotolerant

soda lakes

no salt requirement fresh water soils and rocks

Nitrobacter vulgaris Nitrobacter hamburgensis

γ Nitrococcus mobilis

obligatory halophilic

Nitrospina gracilis

obligatory halophilic

δ Nitrospira marina Nitrospira moscoviesis

obligatory halophilic no salt requirement

marine environment freshwater

Nitrococcus mobilis and Nitrospina gracilis are both obligatory halophilic, all isolates originate from marine habitats. The genus Nitrospira comprises obligatory halophilic species (N. marina) together with nonhalophilic species (N. moscoviensis). With Nitrobacter-isolates, obligate salt requirement has not been observed, although some strains were isolated from marine environments or from soda lakes (Koops and Röser 2001). The application of molecular methods revealed that yet uncultured Nitrospira-like microorganisms and not Nitrobacter spp., are the dominating nitrite oxidisers in most WWTPs. Nitrospira-like nitrite oxidisers are also of major importance in other ecosystems like drinking water distribution systems or soil (Wagner and Loy 2002). A recent study on the ecophysiology of the uncultured Nitrospira-like nitrite oxidisers in activated sludge has shown that these bacteria are able to fix bicarbonate and to simultaneously take up pyruvate (Daims et al 2001). Nitrospira-like nitrite oxidisers are probably K-strategists (with high substrate affinities and low maximum activity or growth rate) for oxygen and nitrite and thus outcompete Nitrobacter under substrate-limiting conditions in WWTPs. This hypothesis would also explain why Nitrobacter and Nitrospira co-exist in reactors with temporarily higher nitrite concentrations (Wagner and Loy, 2002).

Introduction

19

1.4.3 Nitrification under salt stress 1.4.3.1 Influence of salinity on physical-chemical processes The presence of inorganic salts in solution produces what is called non-ideal behaviour of ions and molecules in solution. It was recognized that the activity coefficient for ions in an electrolyte was related to the concentration of the charged particles in the solution. The activity of a cation is set equal to its concentration multiplied by an activity coefficient (γ+) and the activity of an anion is written as the product of its concentration and an activity coefficient (γ−). Experimentally, only the product (γ+γ−) could be determined since one cannot isolate individual ions, but must always deal with electrically neutral solutions. Therefore, the activity coefficients of salt solutions are given as the geometric mean (γ±). This means activity coefficient (γ±=(γ+γ−)1/2) is found to depend on the total ionic strength of the solution (Butler 1964). Ionic strength is an empirical measure of the interactions among all the ions, which causes deviation from ideal behaviour (Snoeyink et al 1980). Ionic strength is a general property of the solution and not a property of any particular ion in the solution. It was described by Lewis and Randall (1921) by: (1.1) Ci Concentration of ionic species, i (mole) Zi Charge of species, i The summation extend over the ions in the solution An example is that ionic strength affects the ratio, at equilibrium, of the molar concentrations of carbonate to bicarbonate. In a solution with an ionic strength of 10–3, the ratio (CO3−2/HCO3−)=0.58, while if we had not taken the effect of ionic strength into account, the ratio (CO3−2/HCO3−)=0.501. Another example is the effect of decreasing the solubility of molecular species, such as dissolved oxygen, by increasing the salt concentration known as the “salting out effect”. This phenomenon could be illustrated from the oxygen solubility for distilled and ocean water. The dissolved oxygen concentration for distilled water, and ocean water with ionic strength 0.7 was found to be 8.4 and 6.75 mg/L respectively (Snoeyink and Jenkins 1980). Effect of ionic strength on the NH3/NH4+ equilibrium The actual form of substrate used for ammonia oxidation by ammonia oxidisers is its undissociated form NH3, rather than ammonium ion NH4+ (Suzuki et al 1974; Bock et al 1991). Ammonia dissolved in water exists as an equilibrium of un-ionised ammonia (NH3) and ionised ammonia (NH4+) which is represented by the following equilibrium reaction: NH4++OH−↔NH3+H2O

Nitrification in saline industrial wastewater

20

The term total ammonia refers to the sum of NH3 and NH4+. The concentration of NH3 is dependent on a number of factors in addition to total ammonia concentration. Most important among these are pH and temperature; the concentration of ammonia increases with increasing pH and with increasing temperature (Emerson et al 1975). The stoichiometric dissociation constant (Ka) is defined as: Ka=(NH3)(H+)/(NH4+) Where the brackets represent molar concentrations and the following equation was used to calculate Ka at all temperatures: pKa=0.09018+2729.92/T Where pKa

the negative log of stoichiometric dissociation constant and

T

the temperature in °k

The percentage of unionised ammonia (NH3%) can be calculated from the solution pH and pKa by the following equations: NH3%=100(1+10(pKa−pH))−1 Ionic strength of a solution has a noticeable effect on the percent of unionised ammonia. The fraction of ammonia in the unionised form decreases with increasing ionic strength in hard water and saline water. In most natural fresh water systems the reduction of unionised ammonia attributable to dissolved solids is negligible. In saline or very hard waters there will be a small but noticeable reduction in unionised ammonia fraction (Emerson et al 1975; Johansson and Wedborg 1980; Clegg and Whitfield 1995). Effect of ionic strength on the oxygen solubility The oxygen solubility in fermentation media can deviate from that in pure water due to salts, substrates, and other solutes. Quicker et al (1981) have published a number of experimental values as well as an empirical correlation. They show that the solubility decreases with solute concentration. Ionic solutes have an influence that is different from nonionics. An impression of the order of magnitude of these effects: A 25% decrease in solubility is reached at about 200 g/L glucose or 50 g/L CaCl2 or 30 g/L NaHPO4 (van ’t Riet and Tramper 1995). 1.4.3.2 Effect of salinity on nitrifying organisms Eventually, the effect of salt on physical-chemical processes will consequently lead to impacts on microorganisms in a saline environment. As explained above, salt has a direct impact on the availability of many compounds. These compounds might be main substrates of several microorganisms. Any reduction in the availability of these compounds will negatively affect their biological activities. Besides the physicalchemical effect of salt, salt itself may have a direct impact on the physiology of living organisms.

Introduction

21

The microorganisms under saline conditions require to adjust and adapt physiologically and structurally (osmoadaptation) to defend their cytoplasm against external osmotic pressure, which decreases the free water activity inside the cytoplasm leading to subsequent cell dehydration and cessation of growth. Generally, the amount of bound water is independent of the osmolarity (0.4 mL/g dry matter), approximately 20% of the total water in fully hydrated cells. Therefore, the immediate effect of high osmolarity is a direct reduction in the volume of free water inside the cells rather than affecting the bound water (Galinski 1995). Hence, it can affect the biochemical reactions within the cytoplasm, which are mainly dependent on the free water present, but not on the three dimensional structure of the functional proteins. Adaptation mechanisms Mechanisms for survival and growth in high-osmolarity conditions must aim at maintaining osmotic equilibrium across the membrane (physiological means). Adaptation to a saline cytoplasm usually covers only a relatively narrow range of physiological reactions, whereas compatible solute producers have developed a more flexible type of adaptation. It is also apparent that seemingly salt-sensitive organisms may display this phenotypic response only because of the lack of suitable solute. Hence the composition of the growth medium has a great influence on the individual salt tolerance of bacteria. This is especially pronounced with media containing complex components that may serve as a source for compatible solutes. It should be noted that osmoadaptation at higher levels is determined not only by the composition of the cytoplasm. It also requires major changes in the composition of the membrane structure and arrangement (structural adaptation), and although the external phase of the cytoplasmic membrane as well as the periplasmic space are always in contact with a saline environment it still requires the necessary adjustments. Most of the mechanisms known for salinity adaptation are by physiological means (Galinski and Truper 1994). Nitrifying bacteria under salt stress The response of the nitrification process to saline conditions and the adaptation mechanisms of nitrifying bacteria towards these conditions are still unknown, autotrophic oxidation of NH4+ to NO2− does not seem to occur above 150 g salt/L, and the limit for the oxidation of NO2− to NO3− may even be lower (Oren 1999; Sorokin et al 2001),. Till so far, no nitrifying bacteria isolate seem to grow above 10% salt. A halophilic isolate named Nitrosococcus halophilus has its growth optimum at 4%NaCl and grow up to 9.4% NaCl (Oren 1999). Life at high salt concentrations is costly from a bioenergetic point of view. Since nitrifying bacteria gain very small amounts of energy from their dissimilatory metabolism, surviving in saline conditions would be difficult for nitrifiers which would further tax their already constrained energy conservation mechanism. It was estimated that Nitrosomonas has to oxidise 30g of ammonia for the production of 1g cell material Thus, it seems that adaptation by known physiological mechanisms is not feasible from a

Nitrification in saline industrial wastewater

22

bioenergetic point of view for chemoautotrophic nitrifying bacteria living at high salinity (Oren 1999). Studies on the effect of salt on nitrification show a decline in activity for ammonia and nitrite oxidisers. However, it does not give a clear answer on the maximum acceptable salt level and which nitrifying group is most sensitive to salt stress: ammonia oxidisers (Campos et al 2002; Hunik et al 1992, 1993) or nitrite oxidisers (Catalan-Sakairi et al 1996; Dincer and Kargi 1999; Furumai et al 1988; Oren 1999; Vredenbregt et al 1997). Moreover, results have a wide range, are difficult to compare and even show contradictory effects (Table 1.8). Reasons for these contradictions might be: • the system configuration and instability in the experimental conditions with respect to temperature, pH, presence of inhibitory compounds or factors; • the way of salt introduction to the system, as a pulse or by gradual increase; • the species involved, use of pure or mixed cultures and of adapted or non-adapted bacteria.

Table 1.8 Reported results on the impact of salt on the nitrification activity and settling characteristics in various systems and under different environmental conditions. Impact of salt on Activity of Nitrifiers

Salt

Environmental conditions

Nitrifiers

System

Reported observation

References

Temp.°C Medium*1 Seed*2 Adapted used*3

Inhibition (%)

Type

[gCl–/L]

pH

65–70

seawater

3.5–6.5

–

20–30

DW

DA

no

LA

70% Yu et al inhibition of (2002) µmax nitrifiers (0.25 day−1)

100

NaCl+ NH4Cl

18

7.8

20

SW

EC

no

LA

SVI not Campos et al affected due (2002) to initial high biomass (20gVSS/L) NO2 accum ulation due to DO limitation, not to salinity

5–60

NaCl

6,18, 30,36

8

25

SW

EN

no

LA

>18 gCl−/L Dincer and SRTmin is 25 Kargi (1999) days, at 0gCl−/L 12 days

Introduction

10–20

NaCl

31–55

NaCl

20–43

SW

EN

no

LA

3,6,12,18 nm*5

27–33

SW

DA

no

LA

NaCl

3,3,12,18

nm

27–33

SW

SA

to 5 g Cl−/L

LA

<5

NaCl

3

nm

20–22

SW

DA

no

LS

Intrasungkha et al (1999)

83

NaCl

18–20

8

28–30

SW

MS

to seawater

AQ

NO2− accum Catalanulation due Sakairi et al to limitation (1996, 1997) of trace elements and CO2

<5

NaCl

1–4

nm

nm

DW

DA

no

BA

Andreadakis et al (1997)

NaCl

10

7–8

30

IW

SA

to 10g Cl−/L

PF

Vredenbregt et al (1997)

0%(compared to 10Cl)

NaCl

20

7–8

30

IW

SA

to 10g CL−/L

PF

NO2− was the only product>20 gCl−/L

57%(compared to 20gCl)

NaCl

34

7–8

30

IW

SA

to 10g Cl−/L

PF

below 20 gCl−/L good fluidizable particles are formed

NaCl

10

7– 8.3

25–30

IW/SW

DA

no

PA

Shock load caused major inhibition

NaCl

20

7– 8.3

25–30

IW/SW

DA

no

PA

*1

8

NO2− accum Dincer and ulation Kargi (2002) above 12 gCl−/L

25

33%(compared to 10gCl)

18,30

23

Panswad and Anan (1999 a, b)

Dahl et al (1997)

DW=Domestic Wastewater; SW=Synthetic Wastewater; IW=Industrial Wastewater from a coal-fired power plant *2 DA=Domestic Activated sludge performing nitrification; EC=Enriched Culture of nitrifying bacteria; EN=Nitrosomonas and Nitrobacter in mixed culture; SA=Salt Adapted activated sludge performing nitrification; MS=Marine Sediment *3 LA=Lab-scale Activated sludge unit; LS=Lab-scale Sequencing batch reactor; AQ=Nitrifiers immobilised in macro-porous cellulose carrier; PF=pilot-scale Fluid-bed; PA=Pilot-scale Activated sludge unit

Nitrification in saline industrial wastewater

24

1.5 Scope and outline of the thesis The increasing pollution problems in receiving waters and the growing interest in water quality protection necessitate more stringent effluent criteria. Especially with respect to the discharge of nutrients like nitrogen effluent criteria are becoming more stringent. Biological nitrification-denitrification is one of the most promising methods to remove nitrogen from wastewater. However, nitrification, the rate limiting step in biological nitrogen removal, proved to be one of the most difficult processes to design and control in wastewater treatment plants, because nitrifying bacteria are slow-growing and very sensitive to environmental factors (temperature, pH, dissolved oxygen concentration, toxic and inhibitory compounds). Researchers so far have concentrated mainly on nitrification in domestic wastewater treatment and achieved broad knowledge and practical experience about the process. The result is that biological nitrogen removal is widely and successfully applied for municipal wastewater. However, these experiences are not directly applicable to industrial wastewater because of the specific composition of the wastewater. Several industries are dealing with a high salt concentration in their wastewater. Also the policy of more economic use of water and water reuse will result in an increase of salt content of the ultimately produced wastewater. High salt levels may negatively affect nitrification, demonstrating the need for improved understanding of the precise effects of salt on nitrification. The main focus of this dissertation is on the effect of salinity on nitrification. The objective of the research was to generate an understanding, based on laboratory scale experiments, modelling and full-scale investigation, of the sensitivity of the nitrification process to sub-optimal salt concentrations in combination with other sub-optimal environmental conditions expected to occur at full scale treatment plants. The dissertation is organized in 8 chapters. Chapter 1 introduces the subject and presents a general literature review on the global nitrogen cycle and nitrogen control technologies with special emphasis on the nitrification process. The contradictory results available on the impact of salt on nitrification are discussed. Chapter 2 presents a new method to measure the activity of ammonia and nitrite oxidisers in mixed bacterial cultures. The developed method allows measuring of the short-term effect of an inhibitor on both the ammonia and nitrite oxidisers in one test without the use of additional metabolic inhibitors. This method was applied in the research as a standard method to determine the inhibition effects of salt on ammonia and nitrite oxidisers. Chapter 3 presents the effect of various types of salt on the activity of ammonia and nitrite oxidisers using enriched cultures of nitrifiers obtained from two lab-scale sequencing batch reactors (SBRs) operated at different sludge ages. Basic equations for the impact of salts on the activities of ammonia and nitrite oxidisers were proposed. The long-term effects of salt on the activity and composition of nitrifiers are presented in Chapter 4. The long-term effects of salt on the flock characteristics of nitrifying sludge using salt-adapted (one year) and non-adapted enriched cultures of nitrifiers are also presented.

Introduction

25

A mathematical model describing the interaction between nitrifiers, heterotrophs and predators in laboratory-scale SBRs was developed and is presented in Chapter 5. The developed model considered multi-substrate consumption and multi-species growth, maintenance and decay in a culture where nitrifiers, heterotrophs and grazers (protozoa and metazoa) are coexisting. Furthermore the model was used to quantify the active biomass fraction of ammonia oxidisers and nitrite oxidisers which were used to calculate the actual specific activities of ammonia and nitrite oxidisers activity the in the lab-scale SBR reactors. In Chapter 6 nitrifier activities and population structure in full-scale domestic and industrial wastewater treatment plants (WWTPs) operated under various salt levels are investigated and compared with results obtained from laboratory-scale reactors. A modified version of the activated sludge model No. 1 (ASM 1) applied to model COD and nitrogen removal in a full- scale industrial WWTP operated under salt stress is presented in Chapter 7. The final chapter, Chapter 8, comprises the overall conclusions of the thesis and the recommendations for further research concerning nitrification. References Andreadakis AD, Kalergis CM, Kartsonas N, Anagostopoulos D (1997) Determination of the impact of toxic inflows on the performance of activated sludge by wastewater charachterization. Wat. Sci. Technol. 37:135–142. Antoniou P, Hamilton J, Koopman B, Jain R, Holloway B, Lyberatos G, Svoronos SA (1990) Effect of temperature and pH on the effective maximum specific growth rate of nitrifying bacteria. Water Res. 24:97–101. Blöch H (2001) EU ploicy on nutrients emission: legislation and implementation Water Sci. Technol. 44:1–6. Bock E, Koops H-P, Harms H, Ahlers B (1991) The biochemistry of nitrifying organisms. In: Variations of Autotrophic Life (Shively, J.M., Ed.): 171–200. Academic Press, London. Brunet RC, García-Gil LJ (1996) Sulfide-induced dissimilatory nitrate to ammonia in anaerobic freshwater sediments. FEMS Microbiol. Ecol. 21:131–138. Butler JN (1964) Ionic Equilibrium-A mathematical Approach. Addision-Wesly publishing company, INC. 48–439. Campos JL, Mosquera-Corral A, Sánchez M, Méndez R, Lema JM (2002) Nitrification in saline wastewater with high ammonia concentration in an activated sludge unit. Water Res. 36:2555– 2560. Catalan-Sakairi MAB, Wang PC, Matsumura M (1997) Nitrification performance of marine nitrifiers immobilized in polyester and macro-porous cellulose carriers. Fermentation and Bioeng. 84:563–571. Catalan-Sakairi MAB, Yasuda K, Matsumura M (1996) Nitrogen removal in seawater using nitrifying and denitrifying bacteria immobilized in porous cellulose carrier. Water Sci. Technol. 34:267–274. Clegg SL, Whitfield M (1995) A chemical model of seawater including dissolved ammonia and the stoichiometric dissociation constant of ammonia in estuarine water and seawater from −2 to 40°C. Geochimica et Cosmochimica Acta 59:2403–2421. Cole JA (1996) Nitrate reduction to ammonia by enteric bacteria: redundancy, or a strategy for survival during oxygen starvation. FEMS Microbiol. Letters 136:1–11.

Nitrification in saline industrial wastewater

26

Cosgrove WJ, Rijsberman FR (2000) World Water Vision, Making Water Everybody’s Business. The World Water Council, Earthscan Publ. Ltd., UK. Dahl C, Sund C, Kristensen GH, Vredenbregt L (1997) Combined biological nitrification and denitrification of high-salinity wastewater. Water Sci. Technol. 36:345–52. Daims H, Nielsen JL, Nielsen PH, Schleifer KH, Wagner M (2001a) In situ characterization of Nitrospira-like nitrite-oxidizing bacteria active in wastewater treatment plants. Appl. Environ. Microbiol. 67:5273–5284. Dincer AR, Kargi F (1999) Salt inhibition of nitrification and denitrification in saline wastewater. Environ. Technol. 29:1147–1153. Dincer AR, Kargi F (2001) Salt inhibition kinetics in nitrification of synthetic saline wastewater. Enzyme and Microbial Technology 28:661–665. Emerson K, Russo RC, Lund RE, Thurston RV (1975) Aqueous Ammonia Equilibrium Calculation: Effect of pH and Temperature. J. Fish. Res. Board Can. 32:2379–2383. Furumai H, Kawasaki T, Futawatari, T, Kusuda T (1988) Effects of salinity on nitrification in a tidal river. Water Sci. Technol. 20:165–174. Galinski EA, Truper BJ (1994) Microbial behaviour in salt stressed ecosystems. FEMS Microbiol. Rev. 15:95–108. Galinski EA (1995) Osmoadaptation in bacteria. Adv. Microb. Physiol. 37:273–327. Gijzen HJ, Bijlsma M (2000) Strategy options for sewage management to protect the marine environment—Technical. Measures. Chapter 3, report UNEP/GPA, The Hague, 2000. Gijzen HJ (2001) Aerobes, anaerobes and phototrophs: a winning team for wastewater manangement. Wat. Sci. Technol. 44:123–132. Gijzen HJ, Mulder A (2001) The global nitrogen cycle out of balance. Water 21, August 2001:38– 40. Gijzen HJ (2003) A 3-step strategic approach to sewage management for sustainable water resources protection. Wat. Sci. Technol. (in press). Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water Sci. Technol. 37:135–142. Hunik JH, Meijer HJG, Tramper J (1992) Kinetics of Nitrosomonas europaea at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 37:802–807. Hunik JH, Meijer HJG, Tramper J (1993) Kinetics of Nitrobacter agilis at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 40:442–448. Johansson O, Wedborg M (1980) The Ammonia-Ammonium equilibrium in sea water at temperatures between 5–25 °C. J. Solution Chemistry 9:37–44. Juretschko S, Timmermann G, Schmid M, Schleifer K-H, Pommerening-Roser A, Koops H-P, Wagner M (1998) Combined molecular and concentional analyses of nitrifying bacteriumin activated sludge: Nitrosococcus mobilis and Nitrospira-like bacteria as dominant populations. Appl. Environ. Microbiol. 64:3042–51. Koops H-P, Pommerening-Röser A (2001) Distribution and ecophysiology of the nitrifying bacteria emphasizing cultured species. FEMS Microbiol. Ecology. 37:1–9. Kuai L, Verstraete W (1998) Ammonium removal by the oxygen-limited autotrophic nitrificationdenitrification system. Appl. Environ. Microbiol. 64:4500–4506. Kuenen JG, Robertson LA (1994) Combined nitrification-denitrification processes. FEMS Microbiol. Rev. 15:109–117. Lewis GN, Randall M (1921) J. Am. Chem. Soc. 43:1111. Mulder A, van der Graaf AA, Robertson LA, uenen JG (1995) Anaerobic ammonia oxidation discovered in a denitrifying fluidised bed reactor. FEMS. Microbiol. Ecol. 16:177–183. Oren (1999) Bioenergetics aspects of halophilism. Microbiol. Molecul. Biol. Rev. 63:334–348. Postma GC, van Dijk JW, Linker PJ, Jagt BM, Geerts HM, Sietsma T (2003) Bedrijfsvergelijking zuiveringsbeheer 2002 (Comparison of the waterboards in the Netherlands 2002). www.vertis.nl/nederlands/bedrijjfsvergelijking2002htm.

Introduction

27

Purkhold U, Pommerening-Röser A, Juretschko S, Schmid MC, Koops H-P, Wagner M (2000) Phylogeny of all recognized species of ammonia-oxidizers based on comparative 16S rRNA and amoA sequence analysis: Implications for molecular diversity survey. Appl. Environ. Microbiol. 66:5368–5382. Pynaert K, Windey K, smets B, de Bo I, Verstraete W, Wyffels S, Boeckx P, van Cleemput (2004) Lab-scale oxygen=limited autotrophic nitrification-denitrification (OLAND): from concept to niche applications EU 5th Framework IcoN symposium Anammox: new sustainable N-removal from wastewater. 21–23 January 2004, Ghent, Belgium Quicker GA, Schumpe A, Konig B, Deckwer WD (1981) Comparison of measured and calculated oxygen solubilities in fermentation media. J.Biotechnol. And Bioeng. 23:635–650. Richardson DJ, Wehrfritz JM, Keech A, Crossman LC, Roldan MD, Sears HJ, Butler CS, Reilly A, Moir JWB, Berks BC, Ferguson SJ, Thomson AJ, Spiro S (1998) The diversity of redox proteins involved in bacterial heterotrophic nitrification and aerobic denitrification. Biochem. Soc. Trans 26:401–408. Scheible OK, Heidman J (1994) Nitrogen Control. U.S. Environmental Protection Agency (EPA) Washington D.C., EPA/625/R-93/010. Schmidt I, Bock E (1997) Anaerobic ammonia oxidation with nitrogen dioxide by Nitrosomonas eutropha. Arch. Microbiol. 167:106–111. Schmidt I, Sliekers O, Schmid M, Bock E, Fuerst J, Kuenen JG, Jetten MSM, Strous M (2003) New concepts of microbial treatment processes for the nitrogen removal in wastewater. FEMS Microbiol. Rev. 27:481–492. Simon J (2002) Enzymology and bioenergetics of respiratory nitrite ammonification. FEMS Microbiol. Rev. 26:285–309. Sliekers A, Derwort, Kuenen JG, Strous M, Jetten MSM (1998) Completely autotrophic ammonia removal over nitrite in one single reactor. Water Res. 14:23–45. Sliekers A, Third K, Abma W, Kuenen JG, Jetten MSM (2003) CANON and Anammox in a gaslift reactor. FEMS. Lett. 218:339–344. Smil V (1997) Global population and the nitrogen cycle. Scientific American 277:58–63. Snoeyink VL, Jenkins D (1980) Water chemistry. John Wiley & Sons, New York:74–82. Sorokin D, Tourova T, Markus C, Wagner M, Koops H-P, Kuenen JG, Jetten MSM (2001) Isolation and properties of obligate chemlithoautotrophic and extremely alkali-tolerant ammonia-oxidising bacteria from Mongolian soda lakes. Arch. Microbiol. 176:170–177. Stalely JT, Bryent MP, Pfennig N, Holt JG (1989). Aerobic chemolithotrophic bacteria and associatd organisms. Bergey’s Manual of Systematic Bacteriology (vol.3), Williams and Wilkins, Boltimore, USA: 1807–1834. Strous M, van Gerven E, Ping Z, Kuenen JG, Jetten MSM (1997) Ammonium removal from concentrated waste streams with the Anaerobic Ammonium Oxidation (Anammox) process in different reactor configurations. Water Res. 31:1955–1962. Suzuki I, Dular U, Kwork SC (1974) Ammonia or ammonium ion as substrate for oxidation by Nitrosomonas europea cells and extracts. J. Bacteriol. 120:556–558. van Loosdrecht MCM, Jetten MSM (1998) Microbiological conversions in nitrogen removal. Water Sci. Technol. 38:1–7. van ‘t Riet K, Tramper J (1991) Basic Bioreactor Design. Marcel Dekker, inc. New York. Vitousek PM, Aber J, Howarth RW, Likens GE, Matson PA, Schindler DW, Schlesinger WH, Tilman GD (1997) Human alteration of the global nitrogen cycle: Causes and consequences. Issues in Ecology No1, Ecological Society of America, Washongton DC. pp.17. Vredenbregt LHJ, Nielsen K, Potma AA, Kristensen GH, Sund C (1997) Fluid bed biological nitrification and denitrification in high salinity wastewater. Water Sci. Technol. 36:93–100. Wagner, M, Rath G, Koops H-P, Flood J, Amann R (1996) In situ analysis of nitrifying bacteria in sewage treatment plants. Water Sci. Technol. 34:237–44. Wagner M, Loy A (2002) Bacterial community composition and function in sewage treatment system. Environ. Biotechnol. 13:218–227.

Nitrification in saline industrial wastewater

28

WHO/UNEP (1997) Water Pollution Control—A Guide to the Use of Water Quality Management Principles. Wiesmann U (1994) Biological nitrogen removal from wastewater. In: Fiechter A (ed.) Advances in biochemical engineering biotechnology 51:113–154. Woolard CR, Irvine RL (1995) Treatment of hypersaline wastewater in the sequencing batch reactor. Water Res. 29:1159–1168. World Bank. (2001) World Development Indicators. Yu SM, Leung WY, Ho KM, Greenfield PF, Eckenfelder WW (2002) The impact of sea water flushing on biological nitrification-denitrification activated sludge sewage treatment process. Water Sci. Technol. 46:209–216.

Chapter 2 Improved Method for Determination of Ammonia and Nitrite Oxidation Activities in Mixed Bacterial Cultures Previously published as: Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003) Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures. Appl. Microbiol. Biotechnol. 63:217–221.

Abstract A simple and reliable method to measure the activity of ammonia and nitrite oxidisers in mixed bacterial cultures was developed. This method considers nitrification as a two-step process and can also be applied to measure the effects of specific inhibitors on the activity of nitrifiers. It allows measuring of the short-term effect of an inhibitor on both the ammonia and nitrite oxidisers in one test under controlled environmental conditions (pH, temperature). The developed method differentiates between the ammonia and nitrite oxidisers by consecutive injection of NO2− and NH4+. The main advantage of this method is avoiding the use of metabolic inhibitors for ammonia or nitrite oxidisers, as used by other methods. The method was applied in two different procedures, both using an enriched culture of nitrifiers. In the first procedure a small reactor of 10 mL in which the oxygen consumption rate (OUR) was used to determine the ammonia and nitrite oxidisers activities. This procedure only takes a few minutes per sample and therefore is suitable for screening of a large number of inhibitors in a short time period. In the second procedure a reactor of 500 ml was used in which the ammonia and nitrite consumption rate was determined. This procedure takes several hours. The advantage compared to the other procedure is, however, that the obtained substrate consumption rate can be used to determine the kinetic parameters of the ammonia and nitrite oxidisers. Both procedures were used to determine the inhibitory effects of salt (NaCl up to 15 g Cl−/L) on an enriched culture of nitrifying bacteria at lab-scale. The results obtained with the small reactor of 10 mL and the 500 mL reactor in the case of NaCl are very similar and in agreement with the results obtained for pure cultures of ammonia and nitrite oxidisers. The results of the method demonstrate its potential to accurately determine the individual activities of nitrite and ammonia oxidisers.

2 Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures 2.1 Introduction The nitrification process is sensitive to environmental factors such as temperature, dissolved oxygen concentration, pH, available substrate and product inhibition and inhibitory compounds (Antoniou et al 1990; Hellinga et al 1998; Sharma and Ahlert 1977). Inhibitory conditions can occur in many industrial wastewater treatment systems, which might lead to adverse effects on one or both steps of the nitrification process. The nitrification activity of a treatment plant, which receives wastewater with potentially inhibitory compounds, can be determined with help of a laboratory experiments, such as a respirometric test. Respirometric techniques, measuring the oxygen uptake rate (OUR) as function of substrate consumption, have been used to obtain kinetic parameters, wastewater characteristics and as a toxicity detection tool (Spanjers et al 1998). Quantifying the inhibitory effect of toxicants on nitrification in activated sludge systems was early investigated using respirometry, considering nitrification as one-step process. This assumption simplifies nitrification, the ammonia oxidation to nitrate via nitrite to one step only, namely direct oxidation of ammonia to nitrate (Nowak and Svardal 1993; Nowak et al 1995; Kong et al 1996; Spanjers et al 1998). The main disadvantage of this approach is neglecting the production of nitrite when different response of ammonia and nitrite oxidisers to the toxicant occurs. Nitrite accumulation itself is presumed to result in a toxic effect linked to pH on the biomass and especially on ammonia oxidisers (Anthonisen et al 1976; Antoniou et al 1990). However, partial oxidation of ammonia to nitrite and subsequent reduction of the latter to nitrogen gas in the denitrification process was seen as a favourable short cut, 25% saving in oxygen and 40% reduction in chemical oxygen demand (COD) requirements (Hellinga et al 1998; van Loosdrecht and Jetten 1998). Therefore, measuring the kinetic parameters of the nitrification under the prevailing conditions as two-step process is an essential approach for optimal design of the nitrogen removal in wastewater treatment plant. Selective metabolic inhibitors for ammonia and nitrite oxidisers have been used to allow separation of the different activities. Chlorate has been used to stop nitrite oxidation in soil, sediments, and activated sludge systems (Belser and Mays 1980; Hynes and Knowles 1983; Sumacz-Gorska et al 1995, 1996). However, doubts concerning the slow and non-specific action of chlorate limit its usefulness in discriminatory respiratory assays with mixed cultures. Ginestet et al (1998) developed a protocol allowing differentiation between bacterial activities in a mixed culture containing nitrifiers using respirometry. In this method concentrations of 86 mM allylthiourea and 24 mM azide

Improved method for determination

33

were used to completely and instantaneously inhibit ammonia and nitrite oxidisers, respectively, without affecting other activities of for instance, endogenous respiration, ammonia oxidisers, and heterotrophic bacteria. However, potential interaction of a selective metabolic inhibitor for ammonia or nitrite with the other toxic compounds to be tested makes these methods unreliable. Such interaction might lead to unrealistic effects of the toxicant on the nitrification activity. Therefore the aim of this research was to develop a reliable, simple, robust and quick procedure to measure shock-load effects of specific inhibitors on the activity of ammonia and nitrite oxidisers separately for a mixed culture. 2.2 Material and Methods 2.2.1 Cultivation of nitrifying bacteria Nitrifying bacteria used in this study were cultivated in sequential batch reactor (SBR) systems (2.5 L) with automated operation, control (pH: 7.5, T: 30°C) and monitoring. Nitrifying activated sludge obtained from a domestic wastewater treatment plant was used to inoculate the SBR systems. The SBR systems were operated for four years in cycles of 6 hours including a 10 minutes fill period, 4 hours reaction period, 80 minutes for settling, and 30 minutes for effluent discharge. A synthetic medium containing mainly ammonia and nutrients to enhance the microbial growth was used as SBR feeding. 1.5 L of medium was fed at the filling period and the effluent was removed at the end of the settling period. 2.2.2 Media Synthetic medium prepared with demineralized water had the following composition: (NH4)2CO3 857.95 mg/L as ammonia source, NaH2PO4.H2O 167.5 mg/L, MgSO4.7H2O 90 mg/L, CaCl2.H2O 14 mg/L, KCl 36 mg/L, yeast extract 1 mg/L, nutrient solution 0.3 mL/L. The nutrient solution was added to the medium in order to enhance the microbial growth and was prepared with the following chemicals mixed in one liter of demineralized water: 1.5g of FeCl3.6H2O, 0.15g of H3BO3, 0.03g of CuSO4.5H2O, 0.18g of KI, 0.12g of MnCl2.4H2O, 0.06g of Na2MoO4.2H2O, 0.12 g of ZnSO4.7H2O, 0.15 g of CoCl.6H2O, and 10g of EDTA. The very low COD in the influent medium (10 mg/L mainly due to the result of the yeast extract) was to enhance the growth of nitrifiers over heterotrophs. 2.2.3 Procedures for the assessment of nitrification activity The principle of the developed method is to determine the activity of ammonia and nitrite oxidisers separately by consecutive injection of NaNO2 (only nitrite oxidisers activity) and NH4Cl (both ammonia and nitrite oxidisers activities). This enables calculation of the nitrite oxidiser activity from the results of the first injection and the ammonia and nitrite oxidisers activity from the results of the second injection. By substracting the activity of the nitrite oxidisers from the activity obtained for the nitrite and ammonia oxidisers it is

Nitrification in saline industrial wastewater

34

possible to obtain the separate activities. The method was implemented and tested in two different procedures. In the first one, oxygen uptake rate (OUR) measurements were used, in which it takes a few minutes to quantify the nitrification activity under normal or inhibition conditions in a small sample volume (10-mL). In the second one, ammonia and nitrite removal rate measurements were used to assess the nitrification activity in a 500mL sample volume. The ammonia and nitrite measurements will take several hours in the bigger reactor but it allows the quantification of the kinetic parameters of nitrification under normal or inhibition conditions (short-term effects). 2.2.4 Assessment of nitrification activity using biological oxygen monitor (BOM) A biological oxygen monitor (BOM) is a batch type of respirometer for oxygen uptake rate (OUR) measurements with the possibility to inject the required substrate or the inhibitor directly into the reaction chamber (10ml, see Figure 2.1).

Figure 2.1 Biological Oxygen Monitor (BOM) for determination of oxygen uptake rates. Fresh biomass samples were withdrawn directly from the SBR (at the end of the reaction period) shortly before testing. The samples were washed and re-suspended to remove any remaining traces of substrate and to have a sufficient buffering capacity. Washing and resuspending of nitrifiers is an essential step in this procedure, where external pH-control is not possible due to the small sample volume. Therefore, several media (Table 2.1) were tested to select a medium with sufficient buffering capacity and without any effects on the activity of ammonia and nitrite oxidisers. Washing was conducted within a 50-mL analytical syringe. The suspension was allowed to stand until the biomass was settled. Hereafter the supernatant was gently removed. The washing procedure was repeated for 3–4 times. When re-suspending the bacteria in the same washing medium, a 5–10 times dilution was usually required in order to avoid having an activity faster than the response time of the oxygen electrode. 10 mL of the washed cells suspended in medium were transferred to the stirred BOM vessel and aerated for 10 min. by means of an air diffuser. CO2 gas was introduced with the aeration when the pH went beyond 7.5 (aeration might lead to stripping of CO2 and

Improved method for determination

35

consequently cause an increase in pH). The dissolved oxygen concentration was kept above 4.0 mg/L. Temperature was controlled at 30°C by means of a water bath. When the testing condition was stable (pH 7.5; DO above 4 mg/L; T 30 °C), the oxygen probe was sealed in the BOM vessel in such a way that no air bubbles remained in the liquid. The decrease in oxygen concentration was monitored and recorded by a computer. The endogenous respiration rate of the nitrifiers was determined first. Hereafter, NaNO2 was injected as substrate for the nitrite oxidisers and after 3 minutes NH4Cl was injected as substrate for the ammonia oxidisers. The recorded OUR in these three phases, no substrate, after nitrite injection and after ammonia injection are used to calculate the nitrite and ammonia oxidisers activity (Figure 2.2). The biomass content of the tested samples was determined as dry weight (g VSS/L) and the activity was expressed as specific oxygen consumption rate (mg O2. (g VSS)−1.h−1) The optimal initial concentration of NaNO2 to be added was determined. The concentration should not be below the substrate affinity constant of nitrite (KNO2) and not as high as to inhibit ammonia oxidisers (Anthonisen et al 1976; Antoniou et al 1990; Sharma and Ahlert 1977). Ammonia oxidising activities were measured with different NO2− concentrations (up to 50 mg-N/L) to determine the highest NO2− concentration without inhibition. Moreover, the developed procedure was compared with another method, in which 24 of µM azide were used to completely and instantaneously inhibit nitrite oxidisers (Ginestet et al 1998).

Figure 2.2 Nitrite and ammonia oxidising activities based on oxygen uptake profile by enriched nitrifiers (30°C, pH 7.5). The profiles were recorded using a BOM (10-mL). NO2− (10mg-N/L), NH4+ (100 mg-N/L) and azide (24 µM) were added as indicated in the figure. The nitrite and ammonia oxidation rates calculated with azide method were 10–15 % lower than with the developed method.

Nitrification in saline industrial wastewater

36

2.2.5 Assessment of nitrification activity using batch reactor (BR) A double-jacketed batch reactor (BR) with a maximum operating volume of 0.5L (enough for interval sampling) was used. Ammonia and nitrite removal was measured over several hours, allowing quantification of the kinetic parameters of nitrification. The batch experiments were performed under similar conditions as in the BOM (pH 7.5±0.05 and T 30±1°C). A 50–100 mL of biomass was manually transferred from the SBR to the BR and diluted with the SBR effluent at the beginning of each experiment. However, no washing and re-suspending steps were needed. The pH was maintained at 7.50±0.05 automatically by dosing of 0.1N HCl or 0.1N NaOH. At the beginning of the experiments the biomass was aerated for about 30 minutes to be sure that all substrates were consumed (endogenous respiration). NaNO2 was added to reach an initial concentration of 10–15 mg NO2−–N/L. Samples were withdrawn from the BR every 10 minutes and filtered for measuring the NO2− content to determine the nitrite uptake rate. After the consumption of NO2− was completed, NH4Cl was added to reach an initial concentration of 15–25 mg NH4+–N/L and the sampling procedure was repeated to determine the ammonia and nitrite uptake rate (Figure 2.3). To estimate the ammonia and nitrite uptake rate a simple double Monod mathematical model to describe the two-step nitrification was used. Maximum growth rate, yield coefficient and biomass concentrations were lumped into overall parameters, which represent the volumetric oxidation rate of ammonia and nitrite as follows: (2.1) (2.2) where: SNH4, SNO2

ammonia and nitrite concentration (mg-N/L)

XNH4, XNO2

ammonia and nitrite oxidisers (mg-COD/L) maximum specific growth rates for ammonia and nitrite oxidisers (L/h) affinity constant of ammonia and nitrite (mg-N/L)

YNH4, YNO2

yield coefficient for ammonia and nitrite oxidisers

RNH4, RNO2

volumetric uptake rate of ammonia and nitrite (mg-N/L.h)

Rmax,NH4, Rmax,NO2

maximum volumetric uptake rate of ammonia and nitrite (mg-N/L.h)

Improved method for determination

37

Figure 2.3 Typical profiles for nitrite and ammonia uptake rate by an enriched culture of nitrifiers (200 mg VSS/L) in a batch reactor (500-mL). NO2− (10mg-N/L) was added first followed by NH4+ (15 mg-N/L) after complete consumption of NO2− Estimation of affinity constants and uptake rate of ammonia and nitrite oxidisers is done automatically using a software program (Aquasim, Reichert et al 1994). The activity was expressed as specific nitrogen consumption rate (mg N (g VSS)−1 h−1) or expressed as a percentage of the activity obtained under reference conditions (pH 7.5, T 30°C and no inhibitor). 2.2.6 Procedure application The developed procedures were applied to determine the inhibition effect of salt (NaCl) on nitrifiers. In case of the biological oxygen monitor (BOM), 10 mL of biomass suspension (washed and re-suspended in SBR influent+Na2CO3 as described in Table 1) was added to the BOM reaction vessel together with a concentrated NaCl solution to reach the desired inhibitor concentration (5, 10, 15 g Cl−/L). Similarly, the test was conducted using the BR procedure performed under similar conditions as in the BOM (pH 7.5±0.05 and T 30±1°C). A certain amount of biomass was manually transferred from the SBR to the BR, diluted with the SBR effluent and NaCl was added (5, 10, 15 g Cl−/L) at the beginning of each experiment. The activity was expressed as the specific oxygen consumption and specific nitrogen consumption rate or expressed as a percentage of the activity obtained under reference conditions (pH 7.5, T 30°C and no inhibitor).

Nitrification in saline industrial wastewater

38

2.2.7 Analytical procedures Ammonia and nitrite were measured spectrophotrometrically in accordance with Standard Methods (APHA 1995). Nitrate-nitrogen was determined using Dionex 4500i series and Shimadzu C_R5A ion-chromatograph. The mixed liquor volatised suspended solid (VSS) was used to measure the biomass content in the tested samples. The VSS determination was performed after filtration of a 10ml sample of mixed liquor on a Whatman glass micro fibre filter (GC/F) filter. Dry weight was determined after the filter was dried for 24 h at 105°C and weighted on a microbalance. The ash content was calculated after incinerating the dried filter in an oven for 1 h at 550°C. 2.3 Results 2.3.1 Procedures for the assessment of nitrification activity The washing and re-suspending medium to be chosen should have a good buffering capacity, because nitrification leads to H+ production. Shift in pH during the reaction might affect ammonia and nitrite oxidation rate. Of the four media tested, the SBR influent medium in which ammonium carbonate was replaced by sodium carbonate to maintain the same ionic strength, gave the best result (see Table 2.1). This medium was therefore used for all further tests.

Table 2.1 Composition of the washing and resuspension media and its effect on the ammonia and nitrite oxidisers. Washing media

Composition

Ionic(1) pH change(2) Inhibition effect of media strength Before After

NH4 NO2 oxidisers oxidisers

Tris phosphate buffer

50mMTris/HCl+50 mM phosphate buffer (7)

0.08

7.5

7.5

>50%

Nil

SBR effluent

As SBR influent medium(3)

0.035

7.5

7.0

<10%

Nil

SBR influent medium SBR influent−(NH4)2CO3 without amm ammonium carbonate.

0.003

7.5

7.0

<20%

Nil

SBR influent medium with sodium carbonate used instead of ammonium carbonate to maintain the same

0.03

7.5

7.5

Nil

Nil

SBR influent+Na2CO3

Improved method for determination

39

ionic strength. (1)

ionic strength=0.5∑ (Ci Zi2), where Ci is concentration of ionic species (M) and Zi is the charge of species (2) pH change refers to the change in the pH within the test time (5 minutes) due to the H+ production (3) Mind that all NH4+ is converted into NO3−, the ionic strength increased due to addition of NaOH for pH control.

The oxygen uptake measurements using the BOM resulted in profiles shown in Figure 2.2. The initial slope is used for determination of the endogenous respiration. From the next part of the OUR profile the nitrite and ammonia activity can be calculated. A concentration of 10 mg-N/L was chosen since this had no effect on the ammonia oxidisers and is higher than the KNO2 (0.5−1 mg-N/L). The ammonia concentration used was similar to the concentration in the SBR cycle (100 mg-N/L) and high enough to avoid any ammonia limitation. The coefficient of variation of the measured activity is between 10–20%, which illustrates the reliability of the developed procedure. The results were compared with an alternative method, using azide as metabolic inhibitor for nitrite oxidisers. Lower nitrite oxidation rates were observed. Therefore, the inhibitory effect of azide on the nitrite oxidiser was tested. NaNO2 was used as a substrate for the nitrite oxidisers. The endogenous OUR was compared with the OUR after injecting azide in the presence of NaNO2 (Figure 2.4). The instantaneous inhibition of 24 µM azide to the nitrite oxidisers was about 80–90% (8 samples). The test was repeated (6 samples) with a higher azide concentration of 48 µM, but full inhibition of the nitrite oxidisers could not be reached.

Figure 2.4 Effect of azide on the oxygen uptake rate of the nitrite oxidisers using enriched nitrifying biomass (30°C, pH 7.5). The profile

Nitrification in saline industrial wastewater

40

was recorded using a BOM (10-mL). After assessing the endogenous respiration NO2− (10 mg-N/L) and subsequently azide (24 µM) was added. 2.3.2 Procedure application The optimised procedure was applied to investigate the effect of salt (NaCl) on the nitrification process. A biological oxygen monitor (BOM) with a volume of 10mL and a batch reactor (BR) with a volume of 500mL were used. The results are shown in Figure 2.5. A reduction of ammonia oxidiser activities as the result of a shock-load of NaCl was observed (Figure 2.5a). The increase in the inhibition of ammonia oxidiser activity with the increase of NaCl concentrations was noticed in both BOM and BR reactors. Moreover, an agreement between the reduction in activity measured within a few minutes (BOM) and this was measured within a couple of hours (BR) was found. A similar inhibition pattern for the nitrite oxidiser activities with the increase of NaCl concentrations was noticed and an agreement between BOM and BR results was demonstrated (Figure 2.5b).

Figure 2.5 Effect of NaCl on the activity of ammonia and nitrite oxidisers from an enriched culture (30°C, pH 7.5). Activities are expressed as specific oxygen uptake rates in the biological oxygen monitor (BOM, 10-mL) and as specific nitrite/ammonia uptake rate in batch reactor (BR, 500-mL). Error bars indicate standard error, only applicable for the BOM (n=4)

Improved method for determination

41

2.4 Discussion 2.4.1 Procedures for the assessment of nitrification activity Estimating the nitrification activity using the BOM has the advantage that a large number of samples can be analysed within a short time period. However, the pH control remains a crucial factor to ensure the test stability. Washing and re-suspending of bacterial cells in a medium, which has sufficient buffering capacity and no adverse affect on the bacterial activity, appeared to give good stability. Tris/phosphate (50 mM) was used by a number of researchers as a medium for washing and re-suspending bacterial cells (Hunik 1993; Leenen et al 1997). Their results did not show the reduction in activity of ammonia oxidisers as observed in our results and by van Ginkel et al (1983). This reduction in activity could be explained by the difference in ionic strength between the growth medium used in bacterial cultivation and the tris/phosphate buffer (Table 1). The sudden change of ionic strength of the re-suspension medium might disturb the osmotic pressure resulting in activity reduction. The growth medium with the replacement of ammonium carbonate by sodium carbonate to maintain the same ionic strength as a washing/resuspending medium has a high buffering capacity. Therefore it prevents the sudden decrease in ammonia oxidisers activity. In case of Hunik (1993) and Leenen et al (1997), the osmotic pressure of their Tris/phosphate buffer was more or less the same as the osmotic pressure of the growth medium they used, therefore a sudden decrease in activity probably did not occur. The effect of washing/re-suspending medium clearly shows the different behaviour of ammonia and nitrite oxidisers. An effect was observed only for ammonia oxidisers and not for the nitrite oxidisers. Differentiation between ammonia and nitrite oxidiser activity using azide did not show a clear difference between both activities. An instantaneous inhibition of nitrite oxidisers was observed as a result of injecting 24 µM azide. Full inhibition, however, could not be accomplished. The observed inhibition (80–90%) of azide to nitrite oxidisers was lower than the reported value (98±5%) by Ginestet et al (1998). One of the factors causing the deviation seems the difference in temperature used in this study (30°C) and the one used in their test (20°C). Moreover, a difference in the nitrifying community could also attribute to that deviation. Non-complete inhibition might lead to misleading results when investigating the separate activities of nitrite and ammonia oxidisers. If it were assumed that this inhibitory value was equivalent to 100% inhibition, there would be an over-estimation of the ammonia oxidiser activity and an under-estimation of the nitrite oxidisers. Thus, the use of azide shows a limitation of its application to fully suppress the nitrite oxidisers within this range of temperature. Moreover, potential interaction of azide with the tested toxicant cannot be excluded. The results of the developed procedure (consecutive injection of NaNO2 and NH4Cl) demonstrate its potential to accurately determine the individual activities of nitrite and ammonia oxidisers. However, the NO2− and NH4+concentrations have to be cautiously set to allow maximum activity without any toxicity phenomena at high substrate concentrations. A concentration of 10 mg-N/L of NO2− and 100 mg-N/L of NH4+ was found to satisfy these conditions. The observed nitrite concentration (10 mg-N/L) was in

Nitrification in saline industrial wastewater

42

agreement with the value reported by Ginestet et al (1998). The ammonia concentration (100 mg-N/L) used was higher than the ones reported by Ginestet et al (1998), Kong et al (1996) and Nowak and Svardal (1993) which were 50, 2.5 and 20 mg-N/1, respectively. The ammonia concentration used in this study (100 mg-N/l of NH4+) appeared to have no toxic effect on both ammonia and nitrite oxidisers. The enriched bacterial culture was adapted to this concentration because it was the same as used for cultivation in the pulse fed SBR systems, which were running for more than two years at the start of the experiments. 2.4.2 Procedure application The results obtained by our method during the investigation of the shock loads effects of salt (NaCl) on the nitrification process are comparable with the results reported by Hunik (1993). He used the same temperature, pH and range of salt level, but obtained his results for a pure culture (Figure 2.6). The agreement between the results confirms the reliability of the developed test to distinguish between the ammonia and nitrite oxidisers in a mixed culture in the presence of an inhibitor. The NaCl inhibition test showed a higher sensitivity of the ammonia oxidisers over the nitrite oxidisers (see Figures 6a and 6b). A similar difference in sensitivity was observed when testing the washing and re-suspending procedure. The osmotic shock had a higher effect on the ammonia oxidisers than on the nitrite oxidisers. The results obtained with the BOM and BR in the case of NaCl are very similar. This suggests that both methods are sufficiently accurate. The BOM procedure, however, should be preferred in case of a large concentration range of inhibitors or when different inhibitors need to be investigated. An advantage of the BR procedure is that the obtained substrate consumption rate can be used to determine the kinetic parameters of the ammonia and nitrite oxidisers. Moreover, the developed method avoids the use of any metabolic inhibitor for ammonia or nitrite oxidisers for activity differentiation.

Figure 2.6 Effect of NaCl on the activity of ammonia and nitrite oxidisers from an enriched culture (30°C, pH 7.5) using the developed method and the results reported from a

Improved method for determination

43

pure culture at similar conditions (Hunik 1993). Activities are expressed as specific OUR in BOM (10-mL) and as specific nitrite/ammonia uptake rate in BR (500-mL). Error bars indicate standard error, only applicable for the BOM (n=4). 2.5 Conclusions A simple and reliable method to measure the activity of ammonia and nitrite oxidisers in mixed bacterial cultures was developed. The main advantage of this method is avoiding the use of metabolic inhibitors for ammonia or nitrite oxidisers, as used by other methods. The method was successfully applied to determine the inhibitory effects of salt (NaCl up to 15 g Cl−/L) on nitrifying bacteria at lab-scale in two different procedures: a small reactor of 10 mL in which the oxygen consumption rate (OUR) was determined and a reactor of 500 mL in which the ammonia and nitrite consumption rate was determined. The results obtained from both procedures are very similar and in agreement with the results obtained for pure cultures of ammonia and nitrite oxidisers. The developed method could be extended to determine the kinetics of other environmental samples containing nitrifiers, such as samples from activated sludge systems and biofilms provided the following precautions are taken: • Select a proper washing/re-suspending medium (in case of BOM), which has the same ionic strength as the growth medium and a high buffering capacity to prevent pH shift; • Avoid having an activity faster than the response time of the oxygen electrode (in case of BOM) and take into account the required corrections for the calibration of the of oxygen electrode (pressure, temperature, salt or any other interference); • Test the suitable NO2− and NH4+ concentration to be used (maximal activity without toxicity); • Inject consecutively NO2− and NH4+ to fully aerated samples and to control the environmental conditions (pH, temperature) during the test time period.

References Anthonisen AC, Loher RC, Prakasam TBS, Srinath EG (1976) Inhibition of nitrification by ammonia and nitrous acid. J. Water Pollut. Control Fed. 48:835–852. Antoniou P, Hamilton J, Koopman B, Jain R, Holloway B, Lyberatos G, Svoronos SA (1990) Effect of temperature and pH on the effective maximum specific growth rate of nitrifying bacteria. Water Res. 24:97–101. APHA (1998) Standard methods for the examination of water and wastewater, 20th edn. American Public Health Association/American Water Works Association/Water Environment Federation, Washington D.C.

Nitrification in saline industrial wastewater

44

Belser LW, Mays EL (1980) Specific inhibition of nitrite oxidation by chlorate and its use in assessing nitrification in soil and sediments. Appl. Environ. Microbiol. 39:505–510. Ginestet P, Audic JM, Urbain V, Block JC (1998) Estimation of nitrifying bacterial activities by measuring oxygen uptake rate in the presence of the metabolic inhibitor allylthiouria and azide. Appl. Environ. Microbiol. 64:2266–2268. Ginkel CG van, Tramper J, Luyben KChAM, Klapwijk A (1983) Characterisation of Nitrosomonas europaea immobilised in calcium alginate. Enzyme Microb. Technol. 5:297–303. Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water Sci. Technol. 37:135–142 Hunik JH (1993) Engineering aspects of nitrification with immobilised cells. PhD Thesis. Wageningen Agricultural University, The Netherlands. Hynes RK, Knowles R (1983) Inhibition of chemoautotrophic nitrification by sodium chlorate and sodium chlorite: a re-examination. Appl. Environ. Microbiol. 45:1178–1182. Kong Z, Vanrolleghem P, Willems P, Verstraete W (1996) Simultaneous determination of inhibition kinetics of carbon oxidation and nitrification with a respirometer. Water Sci. Technol. 30:825–836. Leenen EJTM, Boogert AA, van Lammeren AAM, Tramper J, Wijffels RH (1997) Dynamic of artificially immobilised Nitrosomonas europaea: effect of biomass decay. Biotechnol. Bioeng. 55:630–641. Nowak O, Svardal K (1993) Observations on the kinetics of nitrification under inhibiting conditions caused by industrial wastewater compounds. Water Sci. Technol. 28:115–123. Nowak O, Svardal K, Schweighofer P (1995) The dynamic behaviour of nitrifying activated sludge systems influenced by inhibiting wastewater compounds. Water Sci. Technol. 31:115–124. Reichert P, Ruchti J, Simon W (1994) Aquasim 2.0. Swiss Federal Institute For Environmental Science and Technology (EAWAG), Dübendorf, Switzerland. Smolders GJF, van Loosdrecht MCM, Heijnen JJ (1994) Stoichiometric model of the aerobic metabolism of the biological phosphorus removal process. Biotechnol. Bioeng. 44:837–848. Spanjers H, Vanrolleghem P, Olsson G, Dold PL (1998) Respirometry in control of the activated sludge process: principles. Int. Water Assoc. Q, Scientific and Technical Report 7. Sumacz-Gorska J, Gernaey K, Demuynck C, Vanrolleghem P, Verstraete W (1995) Nitrification process control in activated sludge using oxygen uptake measurements. Environ. Technol. 16:569–577. Sumacz-Gorska J, Gernaey K, Demuynck C, Vanrolleghem P, Verstraete W (1996) Nitrification monitoring in activated sludge by oxygen uptake (OUR) measurements. Water Res. 30:1228– 1236. van Loosdrecht MCM, Jetten MSM (1998) Microbiological conversions in nitrogen removal. Water Sci. Technol. 38:1–7.

Chapter 3 Short Term Effects of Various Salts on Ammonia and Nitrite Oxidisers in Enriched Bacterial Cultures Submitted as: Moussa MS, Song Y, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003) Short-term effects of various salts on ammonia and nitrite oxidisers in enriched bacterial culture Appl. Microbiol. Biotechnol.

Abstract The effect of various types of salt on the activity of ammonia and nitrite oxidisers was investigated. Enriched cultures of nitrifiers obtained from two lab-scale sequencing batch reactors operated at sludge age (SRT) of 30 and 100 days were used in this study. The fluorescent in situ hybridisation technique was used to identify the presence of ammonia and nitrite oxidisers in both reactors. Respiration activity tests were used to determine the shock load effects of salt on ammonia and nitrite oxidisers under controlled conditions (pH 7.5, T: 30°C). At the same molar concentration the divalent cations (CaCl2, MgCl2) have a stronger inhibitory effect than monovalent cations both on ammonia and nitrite oxidisers. The effect of different salt ions was evaluated and quantified. Based on this a basic equation for the impact of salts on nitrification processes was proposed. SRT has no effect on the tolerance of ammonia oxidisers for shock loads of salt, nor on the type of ammonia oxidisers present. Moreover, the different ammonia oxidising species seem to have a similar response to salt stress at similar environmental conditions. In contrast, SRT has a significant impact on salt tolerance of nitrite oxidisers: the longer the sludge age the stronger the inhibition. The results demonstrate that Nitrobacter agilis is more resistant to shock loads of salt than Nitrospira under the same environmental conditions.

3 Short term effects of various salts on ammonia and nitrite oxidisers in enriched bacterial cultures 3.1 Introduction Nitrifying bacteria and the process of nitrification are sensitive to environmental factors such as temperature, dissolved oxygen concentration, pH, available substrate, product inhibition and inhibitory compounds (Antoniou et al 1990; Hellinga et al 1998; Sharma and Ahlert 1977). Understanding the effect of these factors will lead to better design and operation of wastewater treatment plants (WWTP). Keller et al (2002) have summarised the existing literature on nitrification and stressed the use of new technologies and the use of modelling. Additionally, a better understanding of the ecology and physiology of the prevailing organisms in nitrifying reactors may also help to improve the treatment efficiency (Dabert et al 2002; Wagner et al 2002). The effect of salts on nitrification is a major concern nowadays, especially in industrial wastewater treatment. Industries such as pickling, cheese manufacturing, seafood processing, tanning, chemicals and pharmaceuticals productions, oil and gas recovery produce high inorganic salt concentrations in their wastewater. Other sources of saline wastewater include infiltration of subsurface water in the coastal areas, landfill leachates and contaminated ground water. Waste minimisation practices are expected to generate brines in future via effective water reuse and recycling schemes. Also the use of saline water for flushing due to the scarcity of fresh water will increase the wastewater salinity that reaches treatment plants (Campos et al 2002; Dahl et al 1997; Woolard and Irvine 1995; Yu et al 2002). It is not clear what the maximum acceptable salt level is and which nitrifying group is most sensitive to salt stress: ammonia oxidisers (Campos et al 2002; Hunik et al 1992, 1993) or nitrite oxidisers (Catalan-Sakairi et al 1996; Dincer and Kargi 1999; Furumai et al 1988; Vredenbregt et al 1997). Contradictions from literature reports on this issue can be explained by the involvement of different microbial species, the use of either pure culture or mixed culture, differences in the environmental conditions, type of salt and the way of salt application. This study aims at qualifying and quantifying the short-term effect of different types of salt on ammonia and nitrite oxidisers. Short-term effects represent the sudden increase of salt concentration for non-salt adapted species, which is similar to salt fluctuation occurring in practice. In addition, the influence of sludge age on the salt inhibiting effect will be assessed. Long sludge age is commonly applied in industrial WWTPs as safety approach to avoid the washout of nitrifiers and to reduce the sludge production. For this

Short term effects of various salts

49

purpose stable enrichment cultures of nitrifying bacteria were subjected to various salts and salt concentrations. The respiration activities were monitored to assess salt effects. 3.2 Materials and methods 3.2.1 Operation of the Sequenced Batch Reactors The study was carried out in two laboratory scale reactors with a working volume of 2.5L at 30°C. The two reactors were operated automatically as Sequenced Batch Reactors (SBRs). The pH was maintained at pH 7.5±0.05 using 0.25 M NaOH and 0.25 M HCl (BIO controller ADI 1030 coupled with BioXpert 1.1x data acquisition and control program; Applikon b.v.Schiedam, The Netherlands). Nitrifying activated sludge obtained from a domestic wastewater treatment plant was used to inoculate the SBR systems. The SBR systems were operated for 4 years in cycles of 6 h including a 10 min fill period, 4 h reaction period, 80 min for settling, and 30 min for effluent discharge. The characteristics of the operating conditions are summarised in Table 3.1. A synthetic medium containing mainly ammonia and nutrients to enhance the microbial growth was used as SBR feeding. 1.5L of medium was fed during the filling period and the effluent was removed at the end of the settling period. The Sludge Retention Time (SRT), which was desired, was set by the amount of wasted sludge, which was removed from the mixed reactor during each cycle and the biomass in the effluent. Aeration was provided during the reaction period with airflow of 120 L/h. The two reactors were continuously monitored (on-line measuring of DO, pH, addition of NaOH) and sampled (NH4+, NO2−, NO3−) during several cycles. On-line cyclic measurements of DO, the amount of base solution consumed and constant biomass concentration (MLVSS) in the reactors confirmed a steady state condition. The sludge age in each reactor was initially set at 100 days. When steady state was reached, the second reactor was switched to operate at a lower SRT (30 days).

Table 3.1 Operational conditions of the sequenced batch reactors (SBR). N-Load

1200 mg-N/L.day

PH

7.5

COD Load

60 mgCOD/L.day

Temperature

30 °C

HRT

10h

Stirrer speed

650 rpm

(SBR30days)

30 days

Aeration

120 L/h

(SBR100days)

100 days

SRT

3.2.2 Media Synthetic medium prepared with de-mineralised water had the following composition: (NH4)2CO3 857.95 mg/L as ammonia source, NaH2PO4.H2O 167.5 mg/L, MgSO4.7H2O 90 mg/L, CaCl2.H2O 14 mg/L, KCl 36 mg/L, yeast extract 1 mg/L, nutrient solution 0.3 mL/L. The nutrient solution was added to the medium in order to enhance the microbial

Nitrification in saline industrial wastewater

50

growth and contained 1.5g of FeCl3.6H2O, 0.15g of H3BO3, 0.03g of CuSO4.5H2O, 0.18g of KI, 0.12g of MnCl2.4H2O, 0.06g of Na2MoO4.2H2O, 0.12 g of ZnSO4.7H2O, 0.15 g of CoCl.6H2O, and 10g of EDTA in one litre of de-mineralised water. 3.2.3 Respiration activity assay The biomass activity as well as viability was estimated by measuring the oxygen uptake rate (OUR) in a biological oxygen monitor (BOM). This is a batch type of respirometer with the possibility to inject the required substrate directly into the reaction chamber of 10mL. Fresh biomass samples were withdrawn directly from the SBR (at the end of the reaction period), washed and re-suspended in medium before testing. Washing and resuspending of bacterial cells in a medium, which has sufficient buffering capacity and no adverse effect on the bacterial activity was required to remove any remaining substrate and to stabilise the pH during the test. When re-suspending the bacteria in the same medium, a 5–10 times dilution was usually required in order to avoid having an activity faster than the response time of the oxygen electrode. 10 mL of the washed cells suspended in medium were transferred to the stirred BOM reaction vessel together with the concentrated salt solution to reach the desired inhibitory concentration. Two groups of salts were investigated: one with the same anion but with different cations (NaCl, KCl, CaCl2 and MgCl2); the other group of salts with other anions (KI, NaF, Na2SO4, K2SO4). The tested sample was aerated for 10 min; the dissolved oxygen concentration was kept above 4.0 mg/L and the pH at 7.5. Temperature was controlled at 30°C by means of a water bath. The oxygen probe was sealed in the BOM vessel in such a way that no air bubbles remained in the liquid. The decrease in oxygen concentration was monitored and recorded via a data acquisition system. In order to differentiate between the activity of the different biomass fractions (nitrite oxidisers, ammonia oxidisers), OUR was measured in the presence of relevant substrates. Different substrates were injected in the reaction chamber through a seal in the oxygen probe using an analytical syringe (Moussa et al 2003). 3.2.4 Oligonucleotide probes and fluorescent in situ hybridisation (FISH) To identify the population of nitrifying bacteria in both SBR reactors a set of rRNA targeted oligonucleotide probes for Fluorescence In Situ Hybridisation (FISH) was used. Samples were taken from the reactors at steady state conditions and immediately fixed with paraformaldehyde. In situ characterization of microbial populations followed a top to bottom approach. First the samples were hybridised with a probe set (EUB338, EUB338-II, EUB338-III) designed to target almost all bacteria (Daims et al 1999). Then the following group of specific probes was used: ALF968 and BET42a for the alpha and beta subclasses of Proteobacteria, respectively (Manz et al 1992 1996). The ammoniaoxidising and nitrite-oxidising bacteria were identified using previously published probes as described by Nogueira et al (2002). Oligonucleotide probes were purchased as derivatives labelled with the fluorescent dyes Cy3, Cy5, and 5(6)-carboxyfluorescein-Nhydroxysuccinimide-ester (FLUOS), respectively (Interactiva, Ulm, Germany). FISH was performed using the hybridisation and washing buffers as described by Manz et al (1992). The hybridised samples were analysed with a Zeiss Axioplan2 Imaging microscope.

Short term effects of various salts

51

3.2.5 Analytical procedures Ammonia and nitrite were measured spectrophotrometrically in accordance with Standard Methods (APHA, 1998). Nitrate-N was determined using Dionex 4500i series and Shimadzu C_R5A ion-chromatograph. The mixed liqueur volatile suspended solids (MLVSS) determination was performed after filtration of a 10mL sample of mixed liquor on a Whatman glass micro fibre filter (GC/F) filter. Dry weight was determined after the filter was dried for 24 h at 105°C and weighted on a microbalance. The ash content was calculated after incinerating the dried filter in an oven for 1 h at 550°C. The sludge retention time (SRT) of the reactor was calculated from the biomass concentration (MLVSS) in the reactor and the biomass concentration in the effluent. The floc size in the reactors was followed using image analysis. 3.3 Results 3.3.1 SBR performance +

−

The change in NH4 , NO2 , NO3− concentrations for the 2 reactors at steady state are shown in Figure 3.1. Ammonia was consumed within 95 and 115 minutes at SRTs of 30 and 100 days, respectively. Accumulation up to 40 mg/L of NO2–N was observed in the reactor operated at 30 days, while only 4 mg/L NO2–N was detected at an SRT of 100 days. Full oxidation of ammonia and nitrite occurred within 2 h, the rest of the cycle was a starvation period for the ammonia and nitrite oxidisers. Biomass concentrations of 1140 mg MLVSS/L and 3268 mg MLVSS/L were measured in the reactor operated at SRT of 30 days and 100 days, respectively.

Figure 3.1 Concentrations of NH4, NO2 and NO3 during a representative cycle in two nitrifying SBR reactors at steady state conditions (30°C, pH 7.5).

Nitrification in saline industrial wastewater

52

The two reactors were operated at different SRTs (30 and 100 days). 3.3.2 Diversity of nitrifying bacteria in the reactors With respect to ammonia oxidisers 4 species could be detected in both reactors (Table 3.2): Nitrosomonas.oligotropha, N. europaea, Nitrosococcus mobilis and Nitrosospira sp. Both Nitrosomonas species were dominant. Nitrosococcus mobilis at 100 days SRT might be below the detection limit. Concerning nitrite oxidisers, Nitrospira spp. was the dominant species in both reactors. Nitrobacter sp. could only be detected in the 30 days SRT reactor.

Table 3.2 Population structure of nitrifying bacteria dominant at the two SBR reactors operated at SRTs of 30 and 100 days. Reactor Ammonia-oxidizersa Nitrite-oxidizersb (sludge Nitrosomonas Nitrosomonas Nitrisococcus Nitrosospira Nitrobacter Nitrospirsa age) europaea oligotropha mobilis spp. spp. spp. (30 days)

++

+++

+

+

+

++

(100 days)

+++

++

−

+

−

++

a

compared with the total ammonia-oxidizers (Nso190 probe); compared to the total Eubacteria (EUB338 probe) − not detected + present ++ present in relatively high number +++ abundant

b

3.3.3 Effect of different types of salt on ammonia oxidisers (at 30 days SRT) The effect of different types of salt on ammonia oxidisers in an enriched culture of nitrifiers cultivated at 30 days SRT is shown in Figure 3.2. For all salts the activity is reduced with increasing salt concentrations, but there are significant differences between the different types of salt. Figure 3.2a shows the effect of four types of salt with the same anion (Cl−), but with different cations (Na+, K+, Ca++ and Mg++). The chloride salts with divalent cations have a stronger effect (Ca++, Mg++: 50% inhibition at 88 mM) than with monovalent cations (Na+, K+: 50% inhibition at 336 mM) on ammonia oxidisers. In Figure 3.2b the effect of other anions (I−, F−, SO4−−) is shown. Sodium fluoride and potassium iodide were more inhibitory (NaF, KI: 50% inhibition at 137 mM) than sodium and potassium sulphate (Na2SO4, K2SO4:50% inhibition at 440 mM).

Short term effects of various salts

53

Figure 3.2 Effect of different types of salt on the activity of ammonia oxidisers from an enriched culture operated at 30 days SRT, 30°C and pH of 7.5. The 100% is 380±22 mg O2/h.gVSS 3.3.4 Effect of different types of salt on nitrite oxidisers (at 30 days SRT) A similar reduction in the activity of nitrite oxidisers was observed as for ammonia oxidisers (Figure 3.3). The chloride salts with divalent cations have a stronger effect on nitrite oxidisers (Ca++, Mg++: 50% inhibition at 183 mM) than with monovalent cations (Na+, K+: 50% inhibition at 350 mM) (Figure 3a). The effect of other anions (I−, F−, SO4−−) is shown in Figure 3b. Sodium sulphate, potassium sulphate and potassium iodide were more inhibitory (Na2SO4, K2SO4, KI: 50% inhibition at 230 mM) than sodium fluoride (NaF: 50% inhibition at 470 mM).

Nitrification in saline industrial wastewater

54

Figure 3.3 Effect of different types of salt on the activity of nitrite oxidisers from an enriched culture operated at 30 days SRT, 30°C and pH of 7.5. The 100% is 50±4 mg O2/h.gVSS 3.3.5 Influence of sludge age on the salt tolerance of ammonia and nitrite oxidisers With respect to ammonia oxidisers no significant difference in salt tolerance was observed between the two reactors operated at 30 and 100 days SRT (Figure 3.4). However, the nitrite oxidisers in the 100 days reactor were significantly more inhibited than the ones in the 30 days reactor (Figure 3.5).

Short term effects of various salts

55

Figure 3.4 Effect of sludge age on the salt tolerance of ammonia oxidisers from enriched cultures operated at SRTs of 30 days and 100 days under similar environmental conditions (30°C, pH of 7.5).

Nitrification in saline industrial wastewater

56

Figure 3.5 Effect of sludge age on the salt tolerance of nitrite oxidisers from enriched cultures operated at SRTs of 30 days and 100 days under similar environmental conditions (30°C, pH of 7.5). 3.4 Disscusion In this study the impact of different types of salt on activities of ammonia and nitrite oxidisers was explored in two nitrifying SBR reactors differing in sludge age. The composition of the nitrifying community was monitored using FISH aiming to correlate this composition with the observed activities. 3.4.1 Composition of nitrifying population in the reactors With respect to the population of ammonia oxidisers, Nitrosomonas oligotropha and Nitrosomonas europaea were the dominant species in both reactors. This agrees with the results of Purkhold et al (2000), who found nitrosomonads (mainly N.oligotropha cluster and N. europaea/Nitrosococcus mobilis cluster) in 11 nitrifying wastewater treatment plants. Our findings are also consistent with quantitative FISH analysis of the composition of the ammonia oxidiser communities in other lab-scale systems (Daims et

Short term effects of various salts

57

al 2001b; Gieseke et al 2001; Juretschko et al 1998; Liebig et al 2001; Nogueira et al 2002). The fact that Nitrospira sp. was the dominant nitrite oxidiser in both reactors is in agreement with many former observations (Dams et al 2001a; Dionisi et al 2002; Gieseke et al 2001; Juretschko et al 1998; Schramm et al 1999). The authors reported that not Nitrobacter sp., but Nitrospira sp forms the dominant nitrite oxidiser in most wastewater treatment plants. In the present work Nitrobacter sp. was detected in small numbers only in the 30 days SRT reactor, which is characterised by high nitrite accumulation (40 mgNO2−N/L). This result is consistent with the recently published hypothesis that Nitrospira species are K-strategists (and thus thrive at low nitrite concentrations), while Nitrobacter is considered as r-strategist and can compete successfully only in environments with relatively high nitrite concentrations (Schramm et al 1999, 2000). This hypothesis can also explain the co-existence of Nitrobacter and Nitrospira in reactors with temporarily higher nitrite concentrations (Daims et al 2001; Nogueira et al 2002). 3.4.2 Impact of salt on the activity of ammonia and nitrite oxidisers Ammonia oxidisers. Hunik et al (1992) investigated the inhibitory effect of substrate (NH4Cl), different salts (KCl, NaCl) and the end product on a pure culture of Nitrosomonas europaea (ATCC 19718) at similar environmental conditions (30°C, pH 7.5). They observed no significant distinction between the different salts, substrates or end products and concluded that osmotic pressure, due to the high salt concentration, is the explanation for the reduction in activity. They proposed a formula to describe the inhibition of salt on the activity of ammonia oxidisers as a function of salt concentration (equation 3.1). (3.1) The inhibition effects of NaCl, KCl, Na2SO4 and K2SO4 found in this study are in agreement with the formula reported by Hunik et al (1992). However, the proposed formula failed to describe the inhibition effect of CaCl2, MgCl2, NaF and KI, salts not included in their study. Our results showed that at the same molarity, the divalent cations have a higher inhibitory effect than the monovalent cations. Thus the inhibitory effect of salts on ammonia oxidisers was likely due to the ionic strength rather than to molar concentration. The inhibition of salt on ammonia oxidisers cultivated at different sludge ages (30 and 100 days) is in agreement with the inhibition in pure culture (Hunik et al 1992) as shown in Figure 3.6. This suggests that Nitrosomonas (dominant in our systems and in most wastewater treatment plants) responds in a similar way to salt stress at similar environmental conditions. This hypothesis could also explain the results of CatalanSakairi et al (1997); Dahl et al (1997); Vredenbregt et al (1997), who worked in different systems, but found similar inhibition effects. Other researchers claim different responses of ammonia oxidisers to salt stress (Campos et al 2002; Dincer and Kargi 2001; Panswad and Anan 1999 a, b). This disagreement is probably caused by differences in the environmental conditions, the way of introducing salt to the system, and the presence of destabilising factors. The sensitivity

Nitrification in saline industrial wastewater

58

of ammonia oxidisers as well as the presence of destabilising factors during the tests (uncontrolled pH, Temperature) might lead to unrealistic inhibition behaviour (Moussa et al 2003).

Figure 3.6 Comparison of the effect of NaCl on ammonia and nitrite oxidisers from enriched cultures of nitrifiers (this study) with results from a pure culture of Nitrosomonas europaea and Nitrobacter agilis (Hunik et al. 1992, 1993) under similar environmental conditions (30°C, pH of 7.5). Error bars indicate standard error (n=4). Nitrite oxidisers. The shock load effects of KCl, Na-acetate, NaCl, and Na2SO4 on a pure culture of Nitrobacter agilis (ATCC 14123) at similar environmental conditions

Short term effects of various salts

59

(30°C, pH 7.5) was investigated by Hunik et al (1993). Similar to ammonia oxidisers they concluded that osmotic pressure could explain the inhibition in activity. They also proposed a formula that describes the inhibition of salt on nitrite oxidisers as a function of salt concentration (equation 3.2). (3.2) The inhibition effects of the NaCl, KCl and NaF on nitrite oxidisers in the 30 days SRT reactor agrees with this formula. However, again the proposed formula failed to describe the inhibition effect of CaCl2, MgCl2, Na2SO4, and KI. The results of these salts show that at the same molarity, the divalent cations have a higher inhibitory effect than the monovalent cations. So also for nitrite oxidisers the inhibitory effect of salt is likely due to the ionic strength rather than to the osmotic pressure. A significant difference was observed between sludge cultivated at 100 days on the one hand and sludge cultivated at 30 days and pure culture of Nitrobacter agilis used by Hunik et al (1993) on the other hand. Higher sludge age led to more inhibition (Figure 3.6). These results suggest that an increase of the sludge age resulted in reduction of Nitrobacter sp. and consequently in a reduction of resistance of nitrite oxidisers to shock loads of salt. So at similar environmental conditions Nitrobacter sp. might be more resistant to shock load effects of salt than Nitrospira sp. It remains unclear, whether the uncultured Nitrospira sp. has also less resistance to salt stress than Nitrobacter sp. under prolonged salt stress. More research is needed to correlate the microbial population with the physiology of nitrite oxidisers. 3.4.3 Unifying the effect of different types of salt (regression analysis) The difference in the inhibition effects of divalent and monovalent cations on ammonia and nitrite oxidisers at the same molarity has been the subject of speculation. This difference is either due to a unified general salt-related parameter or due to the type of salt. In other words each type of salt or group of salts has its own inhibition pattern. Nine parameters were used to explore the possibility of unifying the inhibition effect of salts. These parameters are: (1) ionic strength (2) concentration in g/L; (3) concentration in mM; (4) ionic strength of cation; (5) concentration of cation in g/L; (6) concentration of cation in mM; (7) ionic strength of anion; (8) concentration of anion in g/L; and (9) concentration of anion in mM. The obtained inhibition results were plotted as a function of these different parameters. Linear regression analysis was conducted to find the parameter with highest correlation coefficient. Ammonia oxidisers: The cation ionic strength was the best parameter to describe the shock load effects of different salts on ammonia oxidisers. Representing the ammonia oxidation activity as a function of the cation ionic strength of the salt instead of the salt concentration (mM) eliminated the deviation between the effect of divalent and monovalent cations (Figure 3.7). A linear regression analysis (r2= 0.88) of the effect of NaCl, KCl, CaCl2, MgCl2, Na2SO4 and K2SO4 on ammonia oxidisers activities operated at sludge ages of 30 and 100 days resulted in equation (3.3). (3.3)

Nitrification in saline industrial wastewater

60

Ci+ Concentration of cation species, i+ (mole) Zi+ Charge of cation species, i+ The constants in equation (3.3) are estimated to be 94 (95% reliability interval of 88 to 101) and −107 (95% reliability interval of −122 to −92). The fluoride and iodide salts have higher inhibition effects compared with the other salts. Thus a linear regression analysis (r2=0.84) of the effect of NaF and KI on ammonia oxidiser activities operated at sludge ages of 30 and 100 days was done. It resulted in equation (3.4). (3.4) The constants in equation (3.4) are estimated to be 92 and (95% reliability interval of 71 to 112) and −227 (95% reliability interval of −454 to −139).

Figure 3.7 Fitted values of the activity of ammonia oxidisers (line) with 95% confidence interval (grey area) as a function of the ionic strength of salt cations. Measured values are presented by dots. Nitrite oxidisers: The ionic strength of anions was found to best describe the shock load effects of different salts on nitrite oxidisers. Representing the nitrite oxidiser activity as a function of anion ionic strength instead of the salt concentration (mM) eliminated the deviation between the effects of the divalent and monovalent cations (Figure 3.8). A linear regression analysis (r2=0.96) of the effect of all tested salts (NaCl, KCl, CaCl2,

Short term effects of various salts

61

MgCl2, Na2SO4, K2SO4, NaF, and KI on nitrite oxidiser activities operated at 30 days SRT resulted in equation (3.5). (3.5) Ci− Concentration of anion species, i− (mole) Zi− Charge of anion species, i− The constants in equation (3.5) are estimated to be 101 (95% reliability interval of 96 to 105) and −142 (95% reliability interval of −159 to −125). A linear regression analysis (r2= 0.93) of data at 100 days SRT was conducted and resulted in equation (3.6). (3.6) The first constant is estimated to be 91.6 and with 95% reliability interval of 82 to 100.8 and the second one is estimated to be −218.9 and with 95% reliability interval of (−254 to −183.74).

Figure 3.8 Fitted values of the activity of nitrite oxidisers (line) with 95% confidence interval (grey area) as a function of ionic strength of salt anions under different sludge ages (SRTs of 30 and 100 days). Measured values are presented by dots. The approach to come to an overall description for the salt inhibition on both ammonia oxidisers and nitrite oxidisers does not seem to be successful. So, we propose separate explanations for the salt effects on the two groups of nitrifying organisms.

Nitrification in saline industrial wastewater

62

Ammonia oxidisers: All the inhibitory effects of various salts on ammonia oxidisers were explained as a function of ionic strength of salt cations, which leads to the following two hypotheses for salt stress: 1) Interaction of the positive salt ions with the enzymes catalysing the ammonia oxidation reaction (ammonia-mono-oxygenase and/or hydroxylamine oxidoreductase). 2) The positive salt ions may affect the dissociation constant of ammonia (Clegg and Whitfield 1995) resulting in a reduction of the unionised form of ammonia (the main substrate of ammonia oxidisers). Nitrite oxidisers: In contrast with the ammonia oxidisers the inhibitory effects of various salts on nitrite oxidisers were explained as a function of ionic strength of salt anions, which leads to the following two hypotheses for salt stress: 1) Interaction of the negative salt ions with the enzyme catalysing the nitrite oxidation reaction (nitrite oxidoreductase). 2) Competition of the negative salt ions with NO2−, the main substrate for nitrite oxidisers. Besides, a general effect of cations and anions on the membrane processes of ammonia and nitrite oxidisers, leading to an impaired protonmotive force, can also not be excluded. 3.5 Conclusions 1- The shock load effects of monvovalent and divalent salt ions on ammonia and nitrite oxidiser activities can be described as a function of ionic strength of salt cations for ammonia oxidisers and of ionic strength of salt anions for nitrite oxidisers; formulas describing this inhibition are proposed; 2- Fluoride and iodide salts have a stronger inhibitory effect (two times higher) on ammonia oxidisers than other salts tested in this study. Inhibition of fluoride and iodide on nitrite oxidisers was in the same range as the other salts; 3- SRT has no effect either on the tolerance of ammonia oxidisers for shock loads of salt or on the type of ammonia oxidisers present. In contrast, it has an effect on the type of nitrite oxidisers present and consequently on the tolerance of nitrite oxidisers for shock loads of salt. 4- Nitrobacter is detected in the 30 days SRT reactor is correlated with high nitrite levels (40 mg NO2–N/L). This result confirms the hypothesis that Nitrobacter can compete successfully only in environments with relatively high nitrite concentrations. 5- Nitrobacter is more resistant to the shock loads of salt than Nitrospira.

References Antoniou P, Hamilton J, Koopman B, Jain R, Holloway B, Lyberatos G, Svoronos SA (1990) Effect of temperature and pH on the effective maximum specific growth rate of nitrifying bacteria. Water Res. 24:97–101. APHA (1998) Standard methods for the examination of water and wastewater, 20th edn. American Public Health Association/American Water Works Association/Water Environment Federation, Washington D.C.

Short term effects of various salts

63

Campos JL, Mosquera-Corral A, Sánchez M, Méndez R, Lema JM (2002) Nitrification in saline wastewater with high ammonia concentration in an activated sludge unit. Water Res. 36:2555– 2560. Catalan-Sakairi MAB, Wang PC, Matsumura M (1997) Nitrification performance of marine nitrifiers immobilized in polyester and macro-porous cellulose carriers. Fermentation and Bioeng. 84:563–571. Clegg SL, Whitfield M (1995) A chemical model of seawater including dissolved ammonia and the stoichiometric dissociation constant of ammonia in estuarine water and seawater from −2 to 40°C. Geochimica et Cosmochimica Acta. 59:2403–2421. Dabert P, Delgenes J-P, Moletta R, Godon J-J (2002) Contribution of molecular microbiology to the study in water pollution removal of microbial community dynamic. Re/View in Environmental Science and bio/Technology 1:39–49. Dahl C, Sund C, Kristensen GH, Vredenbregt L (1997) Combined biological nitrification and denitrification of high-salinity wastewater. Water Sci. Technol. 36:345–52. Daims H, Nielsen JL, Nielsen PH, Schleifer KH, Wagner M (2001a) In situ characterization of Nitrospira-like nitrite-oxidizing bacteria active in wastewater treatment plants. Appl. Environ. Microbiol. 67:5273–5284. Daims H, Purkhold U, Bjerrum L, Arnold E, Wilderer PA, Wagner M (2001b) Nitrification in sequencing biofilm batch reactors: lessons from molecular approaches. Water Sci. Technol. 43:9–18. Daims H, Nielsen P, Nielsen JL, Juretschko S, Wagner M (2000) Novel Nitrospira-like bacteria as dominant nitrite-oxidizers in bio.lms from wastewater treatment plants: diversity and in situ physiology. Water Sci. Technol. 41:85–90. Dincer AR, Kargi F (1999) Salt inhibition of nitrification and denitrification in saline wastewater. Environ. Technol. 29:1147–1153. Dincer AR, Kargi F (2001) Salt inhibition kinetics in nitrification of synthetic saline wastewater. Enz. and Micro. Technol. 28:661–665. Dionisi HM, Layton AC, Harms G, Gregory IR, Robinson KG, Sayler GS (2002) Quantification of Nitrosomonas oligotropha-like ammonia-oxidizing bacteria and Nitrospira spp. from full-scale wastewater treatment plants by competitive PCR. Appl. Environ. Microbiol. 68:245–253. Furumai, H, Kawasaki T, Futawatari, T, Kusuda T (1988) Effects of salinity on nitrification in a tidal river. Water Sci. Technol. 20:165–174. Gieseke A, Purkhold U, Wagner M, Amann R, Schramm A (2001) Community structure and activity dynamics of nitrifying bacteria in a phosphate-removing biofilm. Appl. Environ. Microbiol. 67:1351–1362. Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water Sci. Technol. 37:135–142. Hunik JH, Meijer HJG, Tramper J (1992) Kinetics of Nitrosomonas europaea at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 37:802–807. Hunik JH, Meijer HJG, Tramper J (1993) Kinetics of Nitrobacter agilis at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 40:442–448. Juretschko S, Timmermann G, Schmid M, Schleifer KH, Pommerening-Roser A, Koops HP, Wagner M (1998) Combined molecular and conventional analyses of nitrifying bacterium diversity in activated sludge: Nitrosococcus mobilis and Nitrospira-like bacteria as dominant populations. Appl. Environ. Microbiol. 64:3042–3051. Keller J, Yuan Z, Blackall LL (2002) Integrating process engineering and microbiology tools to advance activated sludge wastewater treatment research and development. Re/View in Environmental Science and bio/Technology 1:83–97. Liebig T, Wagner M, Bjerrum L, Denecke M (2001) Nitrification performance and nitrifier community composition of a chemostat and a membrane-assisted bioreactor for the nitrification of sludge reject waters. Bioprocess Biosyst. Eng. 24:203–210.

Nitrification in saline industrial wastewater

64

Manz W, Amann R, Ludwig W, Wagner M, Schleifer KH (1992) Phylogenetic oligonucleotide probes for the major subclasses of Proteobacteria: problems and solutions. Syst. Appl. Microbiol. 15:593–600. Manz W, Amann R, Ludwig W, Vancanneyt M, Schleifer KH (1996) Application of a suite of 16S rRNA specific oligonucleotide probes designed to investigate bacteria of the phylum Cytophaga-Flavobacter-Bacteroides in the natural environment. Microbiol. 142:1097–1106. Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003) Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures. Appl. Microbiol. Biotechnol. 63:217–221. Nogueira R, Melo LF, Purkhold U, Wuertz S, Wagner M (2002) Nitrifying and heterotrophic population dynamics in biofilm reactors: effects of hydraulic retention time and the presence of organic carbon. Water Res. 36:469–481. Panswad T, Anan C (1999a) Impact of high chloride wastewater on an anaerobic/anoxic/aerobic process with and without inoculation of chloride acclimated seeds. Water Res. 33:1165–1172. Panswad T, Anan C (1999b) Specific oxygen, ammonia and nitrate uptake rates of a biological nutrient removal process treating elevated salinity wastewater. Bioresource Technol. 70:237– 243. Sharma B, Ahlert RC (1977) Nitrification and nitrogen removal. Water Res. 11:897–925. Schramm A, De Beer D, Gieseke A, Amann R (2000) Microenvironments and distribution of nitrifying bacteria in a membrane-bound biofilm. Environ. Microbiol. 2:680–686. Schramm A, de Beer D, van den Heuvel JC, Ottengraf S, Amann R (1999) Microscale distribution of populations and activities of Nitrosospira and Nitrospira spp. along a macroscale gradient in a nitrifying bioreactor: quantification by in situ hybridization and the use of microsensors. Appl. Environ. Microbiol. 65:3690–3696. Vredenbregt LHJ, Nielsen K, Potma AA, Kristensen GH, Sund C (1997) Fluid bed biological nitrification and denitrification in high salinity wastewater. Water Sci. Technol. 36:93–100. Wagner M, Loy A (2002) Bacterial community composition and function in sewage treatment system. Environ. Biotechnol. 13:218–227. Woolard CR, Irvine RL (1995) Treatment of hypersaline wastewater in the sequencing batch reactor. Water Res. 29:1159–1168. Yu SM, Leung WY, Ho KM, Greenfield PF, Eckenfelder WW (2002) The impact of sea water flushing on biological nitrification-denitrification activated sludge sewage treatment process. Water Sci. Technol. 46:209–216.

Chapter 4 Long Term Effects of Salt on Activity, Population Structure and Floc Characteristics in Enriched Bacterial Cultures of Nitrifiers Submitted as: Moussa MS, Sumanasekera DU, Ibrahim SH, Lubberding HJ, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM (2003) Long Term Effects of Salt on Activity, Population Structure and Floc Characteristics in Enriched Bacterial Cultures of Nitrifiers.

Abstract The effect of salinity on the activity, the composition of nitrifiers and floc characteristics of nitrifying sludge was studied. Non-adapted and adapted (to 10 NaCl– Cl−/L for one year) enriched cultures of nitrifiers were tested in three sequencing batch reactors. Salt was increased gradually with 5 g Cl−/L up to 40 g Cl−/L. No difference in activity was observed between the adapted and nonadapted sludge. The activities of ammonia and nitrite oxidisers dropped 36% and 11%, respectively, at salt concentrations of 10 g Cl−/L. At 40 g Cl−/L inhibition reached 95% of salt free activity for ammonia and nitrite oxidisers in both adapted and non-adapted reactors. Nitrosomonas europaea and Nitrobacter sp. were the only nitrifiers present at high salt levels (using Fluorescent In Situ Hybridisation). Increased salt concentrations resulted in better settling characteristics of the nitrifying sludge. After 118 days the sludge was brought back to the initial conditions (0 g Cl−/L for non-adapted and 10 g Cl−/L for adpted). Despite the change in population composition similar kinetics as before the salt stress were observed.

4 Long Term Effects of Salt on Activity, Population Structure and Floc Characteristics in Enriched Bacterial Cultures of Nitrifiers 4.1 Introduction Nitrification is the biological oxidation of ammonia to nitrate via nitrite by two groups of chemolithotrophic bacteria, ammonia oxidisers and nitrite oxidisers; both groups have low specific growth rates (Bock et al 1991; Prosser 1989). Nitrifying bacteria and the process of nitrification are sensitive to environmental factors such as temperature, dissolved oxygen concentration, pH, available substrate, product inhibition and inhibitory compounds (Antoniou et al 1990; Hellinga et al 1998; Sharma and Ahlert 1977). There is considerable interest in understanding the ecology of nitrifying bacteria, because nitrification is the bottleneck for biological nitrogen removal in many wastewater treatment plants (WWTPs) and the causes for a sub-optimal performance or even absence of the nitrification process are not always clear. Once nitrifiers have been washed out of a WWTP, recovery of the nitrification process can take long time due to the slow growth rates of the nitrifiers. There is an urgent need for interdisciplinary research at the interface between molecular microbial ecology and process engineering to understand the links between microbial diversity, process efficiency and process stability (Dabert et al 2002; Nogueira et al 2002; Wagner et al 2002). Nowadays salt is considered as a common stress factor in WWTPs, especially in the industrial sector. Industries such as pickling, cheese manufacturing, seafood processing, tanning, the production of chemicals and pharmaceuticals, oil and gas recovery produce high inorganic salt concentrations in their wastewater. Other sources of saline wastewater include infiltration of subsurface water in the coastal areas, landfill leachates and contaminated ground water. Waste minimisation practices are expected to generate brines in future via effective water reuse and recycling schemes. Also the use of saline water for flushing due to the scarcity of fresh water will increase the wastewater salinity that reaches the treatment plant (Campos et al 2002; Dahl et al 1997; Woolard and Irvine 1995; Yu et al 2002). Studies on the effect of salt on nitrification are difficult to compare and show contradictory results (Campos et al 2002; Catalan-Sakairi et al 1996; Dahl et al 1997; Dincer and Kargi 1999; Hunik et al 1992, 1993; Intrasungkha et al 1999; Panswad and Anan 1999 a, b; Vredenbregt et al 1997; Yu et al 2002). Reasons for these contradictions might be: (1) the system configuration and instability in the experimental conditions with respect to temperature, pH, presence of inhibitory compounds or factors; (2) the way of salt introduction to the system, as a pulse or by gradual increase; (3) the species involved, use of pure or mixed cultures and of adapted or non-adapted bacteria.

Nitrification in saline industrial wastewater

70

The work presented here describes the effect of NaCl on the activity, population structure and floc-forming characteristic of nitrifying sludge, adapted or not adapted to salt. A systematic approach was followed in which pH and DO were kept constant and nitrifying sludge was exposed to a gradual increase of NaCl (when full inhibition was reached, the conditions were brought back to the initial stage). 4.2 Materials and methods 4.2.1 Experimental set-up The study was carried out in three laboratory scale reactors with a working volume of 2.5 L each. The three reactors were operated automatically as Sequenced Batch Reactors (SBRs) at 30°C (Figure 4.1). The pH was maintained at 7.5±0.05 using 0.25 M NaOH and 0.25 M HCl (BIO controller ADI 1030 coupled with BioXpert 1.1x data acquisition and control program; Applikon b.v.Schiedam, The Netherlands). The SBR systems were operated identically in cycles of 6 hours including 10 minutes fill period, 4 hours reaction period, 80 minutes for settling, and 30 minutes for effluent discharge. The SBR was fed with a synthetic medium containing mainly ammonia and nutrients to enhance the microbial growth. 1.5 L of medium was fed during the filling period. A similar volume was pumped out of the reactor at the end of the settling period, resulting in a volumetric exchanging ratio of 0.6 and hydraulic residence time of 10 h. The Sludge Retention Time (SRT) in the three reactors was adjusted at 30 days by the amount of wasted sludge removed from the mixed reactor during each cycle and the biomass in the effluent. Aeration was provided during the reaction period with airflow of 120 L/h. The three reactors were continuously monitored (on-line measuring of DO, pH, addition of NaOH) and sampled (MLSS, MLVSS, NH4+, NO2−, NO3−) during several cycles. The main characteristics of the three reactors were as follows: SBR reactor one (R1) was operated for four years as a reference reactor (no salt added) and as a donor of nitrifiers during the whole research period. Nitrifying activated sludge was brought from a domestic wastewater treatment plant and was used as an inoculum to seed the reactor. SBR reactor two (R2) was inoculated from R1 and operated similar to R1. After reaching steady state conditions, salt was increased from 0 to 10 g NaCl–Cl−/L in one step. The reactor was continuously operated at a salt concentration of 10 g–Cl−/L for a period of one year before starting the experiments. The nitrifiers in R2 are representing well-adapted nitrifiers to 10 gCl−/L. SBR reactor three (R3) was seeded with sludge from R1 and operated at zero salt level for a period of four months before starting the experiments to establish steady state conditions. The nitrifiers in R3 are representing non-salt adapted nitrifiers. The three reactors were operated during the experimental period in three phases (Figure 4.1). The main activities performed in each phase are described as follows: Phase I: The activity of nitrifiers in all reactors was determined as the basis level at the beginning of the experiments. Once the base measurements were completed, NaCl was added to R3 to increase the salinity level from 0 to 10 g Cl−/L in two steps; each step with an increment of 5 g Cl−/L. An adaptation period of 2 weeks was allowed before

Long term effects of salt on activity

71

increasing the salt to the next level. The biomass activity was measured immediately after increasing the salt level (shock load effect) and again after 2 weeks of adaptation. The activities of nitrifiers in both reactors R2 (one year at 10 g Cl−/L) and R3 were monitored and compared four weeks after increasing the salt level to 10 g Cl− /L in R3. Phase II: The salt concentration was further increased from 10 g Cl−/L NaCl (5 g − Cl /L per step) up to almost full inhibition level in R2 and R3. An adaptation period of two weeks was allowed at each step before increasing the salt to the next level. The biomass activity was measured immediately after increasing the salt level and again after two weeks of adaptation. Whenever NO2− concentrations during the cycle became limiting for optimal activity of nitrite oxidisers, NaNO2 was added manually up to the concentration needed (25–30 g N–NO2−/L).

Figure 4.1 Schematic representation of the experimental set-up of the three sequencing batch reactors and the experimental phases. (1) Fermentor 2.5L; (2) Biocontroller; (3) P.C. with Bioxpert; (4) Connecting cables; (5) Digital (on/off) outputs; (6) pH and O2 electrodes; (7) Stirrer, engine; (8) Aeration; (9) Influent; (10) Base pump; (11) Acid pump; (12) Effluent; (13) Excess sludge; (14) Water bath

Nitrification in saline industrial wastewater

72

Phase III: When almost full inhibition level (95%) was reached, the salt level was adjusted to its original level (R2:10 g Cl−/L and R3:0 g Cl−/L) by feeding them with their original media. The activities of nitrifiers were monitored during the recovery period. 4.2.2 Medium Synthetic medium was prepared with de-mineralised water had the following composition: (NH4)2CO3 857.95 mg/L as ammonia source, NaH2PO4.H2O 167.5 mg/L, MgSO4.7H2O 90 mg/L, CaCl2.H2O 14 mg/L, KCl 36 mg/L, yeast extract 1 mg/L, nutrient solution 0.3 mL/L. The nutrient solution was added to the medium in order to enhance the microbial growth and contained 1.5g of FeCl3.6H2O, 0.15g of H3BO3, 0.03g of CuSO4.5H2O, 0.18g of KI, 0.12g of MnCl2.4H2O, 0.06g of Na2MoO4.2H2O, 0.12 g of ZnSO4.7H2O, 0.15 g of CoCl.6H2O, and 10g of EDTA in one litre of de-mineralised water. 4.2.3 Analysis Ammonia and nitrite were measured spectrophotrometrically in accordance with Standard Methods (APHA, 1998). Nitrate-N was determined using Dionex 4500i series and Shimadzu C_R5A ion-chromatograph. The mixed liqueur volatile suspended solids (MLVSS) determination was performed after filtration of a 10 mL sample of mixed liquor on a Whatman glass micro fibre filter (GC/F) filter. Dry weight was determined after the filter was dried for 24 h at 105°C and weighted on a microbalance. The ash content was calculated after incinerating the dried filter in an oven for 1 h at 550°C. The sludge retention time (SRT) was calculated from the biomass concentration (MLVSS) in reactors and biomass concentration in the effluent. The floc size in the reactors was followed using image analysis. The average floc diameter was measured using a representative sample, in which at least 500 particles were analysed (Tijhuis et al 1994). 4.2.4 Microscopic analysis The impact of salt on the presence of protozoa, rotifers and nematodes was followed with a light microscope (Olympus) according to the Eikelboom (2000) manual. Fluorescent in situ hybridisation (FISH) To identify the structure of population of the nitrifying bacteria in both SBR reactors a set of rRNA targeted oligonucleotide probes for Fluorescence in situ Hybridization (FISH) was used. Samples were taken from the reactors at steady state conditions and immediately fixed with paraformaldehyde. In situ characterization of microbial populations follows a top to bottom approach. First the samples were hybridized with a probe set (EUB338, EUB338-II, EUB338-III) designed to target almost all bacteria (Daims et al 1999). Then the following group of specific probes was used: ALF968 and BET42a for the alpha and beta subclasses of Proteobacteria respectively (Manz et al 1992, 1996). The ammonia-oxidising and nitrite-oxidising bacteria were identified using previously published probes (Table 4.1). Oligonucleotide probes were purchased as derivatives labeled with the fluorescent dyes Cy3, Cy5, and 5(6)- carboxyfluorescein-N-

Long term effects of salt on activity

73

hydroxysuccinimide-ester (FLUOS), respectively (Interactiva, Ulm, Germany). FISH was performed using the hybridization and washing buffers as described by Manz et al (1992). The hybridised samples were analysed with a Zeiss Axioplan2 Imaging microscope.

Table 4.1 Fluorescently labelled rRNA targeted probes used in this study probe

Target site on rRNA

Target organism

sequence

Reference

Ammonium-oxidizers NEU

653–670

Nm75

CCCCTCTGCTGCACTCTA

Halophilic and Wagner et al halotolerant ammonia (1995) oxidizers belonging to β-proteobacteria of Nitrosomonas spp.

CGGCAGCGGGGGCTTCGGCC Nitrosomonas genus

Hiorns et al (1995)

CGATCCCCTGCTTTTCTCC

Ammoma-oxidizers β-proteobacteria

Mobarry et al (1996)

Nso190

190–208

NmV

174–191 TCCTCAGAGACTACGCGG

Nitrosococcus mobilis

PommereningRöser et al (1996)

Nse1472

1472– 1489

ACCCCAGTCATGACCCCC

Nitrosomonas europaea

Juretschko et al (1998)

NOLI191 191–208

CGATCCCCCACTTTCCTC

Various members of Nitrosomonas oligotropha lineage

Gieseke (2001)

ALLSPIR 443–462

CCGTGACCGTTTCGTTCCG

Nitrosolobus Mobarry et al multiformis, (1996) Nitrosospira briensis, and Nitrovibrio tenuis

Nitrite-oxidizers NIT3

1035– 1048

CCTGTGCTCCATGCTCCG

Nitrobacter spp.

Wagner et al (1996)

CNIT3

1035– 1048

CCTGTGCTCCAGGCTCCG

Competitor for NIT3

Wagner et al (1996)

Ntspa 662

662–679 GGAATTCCGCGCTCCTCT

Genus Nitrospira

Daims et al (2000)

Nitrification in saline industrial wastewater

74

4.3 Results 4.3.1 Performance of the reactors Phase I: Under salt free conditions (R1 and R3) ammonia was consumed during the reaction period (4 hours) within 100 min. Immediately after increasing the salt level in R3 up to 10 g Cl−/L ammonia depletion took 150 min; this depletion rate did not change during the next 4 weeks. Nitrite accumulated up to 40 mg/L NO2–N under salt free conditions (R1 and R3) and to 25 mg/L NO2–N immediately after increasing the salt in R3 up to 10 g Cl−/L; also these levels stayed constant during the next 4 weeks (Figure 4.2). When ammonia and nitrite were depleted (after 150 min) the rest of the cycle was a starvation period. The average biomass concentrations in the reactors (1150 mg MLVSS/L) were independent of salt concentrations. The SRT also remained constant indicating that the net growth yield was not affected at 10 g Cl−/L.

Figure 4.2 Concentration of NH4+ and NO2− during cyclic measurements in a nitrifying SBR reactor (R3) under different conditions: 0 g Cl−/L, 10g Cl−/L (shock effect) and 10 g Cl−/L (after 4 weeks). The reactor was operated at pH 7.5, 30°C and 30 days sludge age. A comparison between the specific activity of ammonia and nitrite oxidisers for R2 (adapted to 10 g Cl−/L for a year) and R3 (adapted to 10 g Cl−/L for one month) is presented in Table 4.2. The shock load effects of 10 g Cl−/L on both ammonia and nitrite oxidisers were more pronounced in R2 (salt was added in one step) than in R3 (salt was

Long term effects of salt on activity

75

added in two steps). Ammonia oxidisers were more sensitive to short and long term stress of 10 gCl−/L (36–39% drop in activity) than nitrite oxidisers (5–13% drop in activity). The effect of 10g Cl−/L on both ammonia and nitrite oxidisers did not show significant differences between the one-year adapted (R2) and the one-month adapted (R3) sludge.

Table 4.2 Effect of NaCl (10g Cl−/L) on the activity of ammonia and nitrite oxidisers after shock load (during the 1st cycle) and in steady state. Both reactors were operated at 10 gCl−/L and were in steady state, R2 during one year, R3 one month. The 10 g of salt/L was added to R2 in one step, to R3 as in two steps of 5 g Cl−/L salt each. Standard deviations are indicated between brackets. Ammonia oxidisers

Reactor

Fresh water Steady state

Nitrite oxidisers

−

10 gCl /L Shock load

10 gCl−/L

Fresh water

Steady state

Steady state Shock load

Steady state

Spe Acti Spec Acti Spe Acti Specific Activ Spec Acti Spe Acti cific vity ific vity cific vity Rate ity % ific vity cific vity Rate % Rate % Rate % mg-N/g Rate % Rate % mgmgmgVSS.h mgmgN/g N/g N/gV N/g N/g VSS.h VSS.h SS.h VSS.h VSS.h R2

44 96±10

R3

45

59±6

61 59±6

100 62

64

62±1

64

34

58

56±6

95

49

84

51±5

87

100

Phase II: The effects of the stepwise salt increase in R2 and R3 (5 g Cl−/L) on ammonia and nitrite oxidisers are shown in Figure 4.3. Increase of salt led to lower activities. Up to 10 g Cl−/L the strong shock load effect was relieved after 2 weeks of adaptation. At 15 g Cl−/L the inhibition effect took some time to reach its full effect; the long-term effect was more severe than the short-term. There was no difference in behaviour between the longterm (one year) and short-term (one month) acclimatised nitrifying sludge. Up to 15 g Cl−/L the nitrite oxidisers were less affected by the salt, while at higher salt levels the inhibition was comparable for both ammonia and nitrite oxidisers. More than 95% reduction in activity was reached for both ammonia and nitrite oxidisers at a salt level of 40 gCl−/L in both reactors (R2 and R3).

Nitrification in saline industrial wastewater

76

Figure 4.3 Effect of salt on the activity of ammonia oxidisers and nitrite oxidisers in two nitrifying SBR reactors operated at pH 7.5, 30°C and 30 days sludge age. R2 was adapted to 10 gCl−/L for a year and R3 was not adapted. Single data points refer to activity in the 1st cycle after changing the salt level. Data points on the line refer to the activity 2–4 weeks after changing the salt level. On-line cyclic measurements of dissolved oxygen (DO), the amount of base solution consumed and the stable biomass concentration (MLVSS) in the reactors indicate steady

Long term effects of salt on activity

77

state conditions under the new salt levels. Within a week under the same salt stress, nitrifiers started to adapt to the new saline environment and stabilised their activities under the new conditions, but never reached the initial activities. Phase III: Two stages of recovery were observed (Figure 4.4): 1) Throughout the elimination of salt during the first 1–2 days (9 cycles) the activity for both types of nitrifiers increased 4 times. 2) During the following period at the initial salt levels (0 in R3, 10 in R2) the activity increased but slowly. After 2 weeks the specific activities in R2 were 23 mg-N/gVSS.h for both ammonia and nitrite oxidisers, which is 40% of the initial activity at 10gCl−/L. The specific activity for both ammonia and nitrite oxidisers in R3 reached 32 mg-N/gVSS.h after 2 weeks, which is 33% (ammonia oxidisers) and 53% (nitrite oxidisers) of the initial activities at 0 g NaCl–Cl−/L.

Figure 4.4 Recovery of nitrifying bacteria after gradual decrease of the salt concentrations from 40 g Cl−/L to 10 gCl−/L (R2) or 0 g Cl−/L (R3). Reactor conditions are similar to figure 2.

Nitrification in saline industrial wastewater

78

4.3.2 Diversity of nitrifying bacteria The composition of nitrifiers during the whole experimental period was stable in R1 and R3, as long it was operated at 0 g NaCl−Cl−/L (Table 3). Phase I: Four species of ammonia oxidisers were observed in all 3 reactors up to 10 g Cl−/L: Nitrosomonas oligotropha, Nitrosomonas europaea, Nitrosococcus mobilis and Nitrosospira sp.. Nitrosomonas oligotropha was the dominant ammonia-oxidiser at 0 g NaCl–Cl−/L, while N. europaea was the dominant at 10 g Cl−/L salt. In all reactors up to 10 gCl−/L Nitrospira sp. was the dominant nitrite oxidiser and Nitrobacter sp. could only be detected in small numbers. No difference in the composition of ammonia and nitrite oxidisers could be observed between R2 (adapted to 10 g Cl−/L for 1 year) and R3 (adapted to 10 g Cl−/L for only 1 month). Phases II&III: Above 10 gCl−/L and up to 30 gCl−/L only N. europaea and N. mobilis were found in both reactors (R2 and R3), while no nitrite oxidisers could be detected. At 40 g Cl−/L only N. europaea was observed. After 2 months of continuous operation at the initial salt level only N. europaea and Nitrobacter sp. were detected. This suggests that other organisms were not able to survive under these elevated salt levels. The quick reappearance of Nitrobacter sp. after recovery suggests that it was still present even at 40 gCl−/L, but it was below detectable levels (Table 4.3).

Table 4.3 Composition of the nitrifiers during the different phases of reactor operation under varying salt concentrations (0–40 g Cl−/L). In phase I R2 and R3 were adapted to 10 g Cl− /L for one year and one month respectively. Phases

Phase I

Phase II

Phase III Back to

−

Salt conc. (gCl L) Reactor

0

10

20

30

40

R1/R3 R2 R3 R2 R3 R2 R3 R2 R3

10

0

R2

R3

Ammonia oxidisers: Nitrosomonas europaea

+

+++ +++ +++ +++ +++ +++ +++ +++ +++ +++

Nitrosomonas oligotropha

+++

++

++

−

−

−

−

−

−

−

−

Nitrosococcus mobilis

++

+

+

+++

++

+++

+

−

−

−

−

Nitrosospira sp

+

++

++

−

−

−

−

−

−

−

−

++

++

++

−

−

−

−

−

−

−

−

Nitrite oxidisers: Nitrospira sp

Long term effects of salt on activity

Nitrobacter sp

+

+

+

−

−

−

79

−

−

−

+

+

− not detected + present ++ present in relatively high number +++ abundant

4.3.3 Higher organisms At 0 g NaCl/L protozoa, rotifers, nematodes and water mites were observed in both reactors (R1, R3). When the salt level was increased to 5 g Cl−/L (R3), these higher organisms stopped moving and started swelling and bursting. Probably, the increase of osmotic pressure caused this effect. A complete absence of all these organisms was observed when operating the reactor continuously under salt stress. However, nematodes were present in R2 (adapted to 10 gCl−/L for 1 year), but they immediately disappeared after further increase of the salt level. After the recovery period nematodes were observed again in R2, while both rotifers and nematodes were observed again in R3. However, protozoa didn’t reappear neither in R2, nor in R3. 4.3.4 Settling characteristics Increasing salt concentrations resulted in the formation of larger flocs and in improved settling characteristics of the nitrifying sludge. Average floc diameters increased from 100 µm at fresh water conditions to 200 µm at increased salt levels. Settling velocity also increased, while SVI reduced from 80 mL/g at 0 g NaCl/L to 40 mL/g for salt treated sludge. A slight increase in effluent suspended solids was observed only in the first cycle after changing the salt level. 4.4 Discussion The results demonstrated the complexity of the effect of elevated salt levels on the nitrification process. Nitrification activity, population structures and settling characteristics are interrelated and are all directly affected by salt. Investigating the impact of salt on one of these aspects and not considering the others might lead to wrong conclusions and interpretations. Most of the literature reports emphasise only on the effect of salinity on nitrifying activity resulting in contradictory interpretation of results. 4.4.1 Effect on activity The decline in activity for ammonia and nitrite oxidisers found in this study are in line with literature (Table 4). The wide range in results illustrated in the table could be ascribed to difference in systems and environmental conditions, the way of salt introduction to the system or to the presence of other destabilising factors. Some of these results are system dependent. Also some results are based on removal efficiencies, not on specific activities, which make it difficult to compare. Nevertheless, Table 4.4 gives an overview of the survival range of nitrifiers under salt stress. Nitrite oxidisers were less affected by salt stress if compared with ammonia oxidisers at salt levels below 10 gCl−/L in both the long-term and the short-term acclimated system.

Nitrification in saline industrial wastewater

80

Similar results were found in a previous enriched culture study by Moussa et al (2003 a, b) and in a pure culture study by Hunik et al (1993). In contrast with these results, Dincer and Kargi (1999) and Vredenbregt et al (1997) concluded from the accumulation of nitrite that nitrite oxidisers were more affected. Probably, nitrite accumulation in their experiments is not caused by salt stress, but due to oxygen limitation, phosphorous limitation, and/or the presence of toxicants. Also Campos et al (2002) explained the accumulation of nitrite in their system under salt stress to oxygen limitation. It is known from nitrification studies under non-salt stress, that oxygen limitation is crucial and that it leads to an incomplete nitrification process and to accumulation of nitrite (Garrido et al 1997; Pollice et al 2002; Picioreanu et al 1997). Salt affects directly the maximum solubility of oxygen and the oxygen transfer rate (van ’t Riet and Tramper 1991), which can lead to limited oxygen availability. The inhibition of salt on long-term and short-term adapted nitrifying sludge— measured as NH4+consumption, NO2− and NO3− formation (this study)—was similar to the inhibition—assessed as oxygen uptake rate—of non-adapted nitrifying sludge (Moussa et al 2003 a, b) and of pure cultures of nitrifiers (Hunik et al 1992, 1993) (Figure 4.5). Therefore, short-term activity measurements can be used as a quick tool to assess the inhibition pattern of salt on nitrifiers. The constant SRT and biomass levels during the experiments suggest that net yield remains unchanged at different salt concentrations, which was also found by Dincer and Kargi (2002) and Yu et al (2002). Elevated salt levels are leading to reduction in specific activity, but not to changes in biomass content, which suggests that salt causes inactivation of nitrifiers and/or an increased decay rate. This increased decay rate could be attributed only to salt, since there are no grazers present under these conditions.

Table 4.4 Reported results on the impact of salt on the nitrification activity and settling characteristics in various systems and under different environmental conditions Impact of salt on Activity of Nit rifiers

Salt

Type

Environmental conditions

Nitrifiers

System used*3

Settling character ristics

Repo rted observ ation

Refs*4

[gCl−/L]

Inhi bition (%)

pH

Temp .°C

Med ium*1

Seed*2 Ada pted

SVI

Effl uent SS

65–70

seawater

3.5–6.5

–

20–30

DW

DA

no

LA

>125

+

100

NaCl+ NH4Cl

18

7.8

20

SW

EC

no

LA

nm

nm

70% inhibition of µmax nitrifiers (0.25 day−1)

11

SVI not affected

2

Long term effects of salt on activity

81

due to initial high biomass (20g VSS/L) NO2 accumu lation due to DO limitation, not to salinity 5–60

NaCl

6,18,30,36

8

25

SW

EN

no

LA

nm

nm

>18 gCl−/L 5 SRTmin is 25 days, at 0gCl−/L 12 days

10–20

NaCl

18,30

8

25

SW

EN

no

LA

nm

nm

NO2− accum ulation above 12 gCl−/L

6

31–55

NaCl

3,6,12,18

SW

DA

no

LA

na*5

+

MLSS decreased with increased NaCl

8,9

20–43

NaCl

3,6,12,18

nm

27–33

SW

SA

to 5g Cl−/L

LA

na

+

<5

NaCl

3

nm

20–22

SW

DA

no

LS

nm

nm

83

<5

NaCl

18–20

nm*5 27–33

8

28–30

SW

MS

to sea water

AQ

nm

nm

7 −

NO2 accumu lation due to limitation of trace elements and CO2

3

NaCl

1–4

nm

nm

DW

DA

no

BA

nm

nm

1

NaCl

10

7–8

30

IW

SA

to 10g Cl−/L

PF

nm

nm

10

0% (com pared to 10gCl)

NaCl

20

7–8

30

IW

SA

to 10g Cl−/L

PF

nm

nm

NO2− was the only product>20 gCl−/L

57% (compared to 20gCl)

NaCl

34

7–8

30

IW

SA

to 10g Cl−/L

PF

nm

nm

below 20 gCl−/L good fluidizable particles

Nitrification in saline industrial wastewater

82

are formed

33% (compared to 10gCl)

*1

NaCl

10

7– 8.3

25–30 IW/SW

DA

no

PA

10

+

Shock load caused major inhibition

NaCl

20

7– 8.3

25–30 IW/SW

DA

no

PA

10

+

Good sludge stability due to gypsum precip itation

4

DW=Domestic Wastewater; SW=Synthetic Wastewater; IW=Industrial Wastewater from a coal-fired power plant DA=Domestic Activated sludge performing nitrification; EC=Enriched Culture of nitrifying bacteria; EN=Nitrosomonas and Nitrobacter in mixed culture SA=Salt Adapted activated sludge performing nitrification; MS=Marine Sediment *3 LA=Lab-scale Activated sludge unit; LS=Lab-scale Sequencing batch reactor; AQ=Nitrifiers immobilised in macro-porous cellulose carrier; PF=pilot-scale Fluid-bed; PA=Pilot-scale Activated sludge unit; BA=Batch Reactor *4 1=Andreadakis et al 1997; 2=Campos et al 2002; 3=Catalan-Sakairi et al 1996, 1997; 4=Dahl et al 1997; 5=Dincer and Kargi 1999; 6=Dincer and Kargi 2002; 7=Intrasungkha et al 1999; 8, 9=Panswad and Anan 1999 a, b; 10=Vredenbregt et al 1997; 11=Yu et al 2002 *5 nm=not measured; na=not affected *2

Long term effects of salt on activity

83

Figure 4.5 Comparison of the NaCl effect on ammonia and nitrite oxidisers of nitrifying enriched cultures. The adapted and non-adapted results from this study are based on nitrogen depletion rates. The reported results use the oxygen uptake measurements either in non-adapted enriched cultures of nitrifiers (Moussa et al 2003) or in pure cultures of Nitrosomonas

Nitrification in saline industrial wastewater

84

europaea and Nitrobacter agilis (Hunik et al 1992, 1993). Error bars indicate standard error (n=4). All reactor conditions were similar to Figure 2. The relatively rapid increase in activity as soon as the salt concentrations were restored to the initial levels could be related to reactivation. The gradual increase afterwards (from day 2 until day 60, when the original activity was totally restored-data not shown) can be attributed to growth of the surviving organisms of the salt stress. A similar recovery after NaCl stress was described by Panswad and Anan (1999 a, b). 4.4.2 Microbial populations The dominance of Nitrosomonas oligotropha and Nitrosomonas europaea at 0 g NaCl−Cl−/L and up to 10 g Cl−/L is consistent with results reported for ammonia oxidiser communities in WWTP and lab-scale systems (Daims et al 2001b; Gieseke et al 2001; Juretschko et al 1998; Liebig et al 2001; Nogueira et al 2002; Purkhold et al 2000). With respect to nitrite oxidisers, Nitrospira sp. was dominant at 0 g NaCL–C−/L and up to 10 g Cl−/L as well, which is in agreement with many former observations (Daims et al 2001a; Dionisi et al 2002; Gieseke et al 2001; Juretschko et al 1998; Schramm et al 1999). Also the presence of Nitrobacter sp. in small numbers is in agreement with the recently published hypothesis that Nitrobacter is considered as r-strategist and can compete successfully only in environments with relatively high nitrite concentrations as occurred in our systems (Daims et al 2001b; Nogueira et al 2002; Schramm et al 1999, 2000). Salt inhibition seems not to be dependent on adaptation time. Not only long-term and short-term adapted nitrifiers (this study), but also pure cultures from marine sediments (Catalan-Sakairi et al 1996, 1997) and nitrifiers in activated sludge (Dahl et al 1997) give a similar inhibition pattern. A selection of N. europaea and Nitrobacter sp. as a result of gradual increase of NaCl above 10 g Cl−/L was observed in both long-term (one-year) and short-term (one-month) adapted sludge. This selection was underlined by the re-growth of only these two species after the recovery period. Under elevated salt levels (up to 30 g Cl−/L) only N. europaea and Nitrosococcus mobilis were present and above 30 g Cl−/L only N. europaea survived. N. europaea remained the only detectable ammonia oxidiser species after the recovery period. These results show that N.europaea is more resistant to gradual salt increase than N.mobilis, although both species are known to be halotolerant or moderately halophilic ammonia oxidisers (Koops et al 2001; Wagner et al 1995). Hovanec et al (1996) also found that N.europaea was present, and presumably the active ammonia oxidiser, in all nitrifying seawater aquarium biofilters, which is consistent with the dominance of N.europaea in nitrifying biofilms seeded with marine sediments (Catalan-Sakairi et al 1996, 1997). The quick re-growth of Nitrobacter sp. during the recovery period suggests that it is more resistant to salt stress than Nitrospira sp., which is in agreement with previous studies (Hunik et al 1992; Moussa et al 2003b). However, both genera Nitrospira and

Long term effects of salt on activity

85

Nitrobacter comprise obligatory halophilic species together with non-halophilic species (Koops et al 2001). Moreover, these results demonstrate that both systems (short-term and long-term adapted nitrifying sludge) were able to fully recover their activity with less diversity in the population of nitrifiers. Sludge of a conventional domestic nitrifying WWTP is the best source for nitrifiers to become adapted to salt stress via gradual adaptation. This way of seeding gives high potential population diversity in contrast with salt adapted sludge with less diversity of nitrifiers. Increase of the diversity of the crucial functional groups of bacteria (e.g. nitrifiers) within WWTP is recommended to render the microbial community more resistant against perturbations (Nogueira et al 2002; Wagner et al 2002). The absence of higher organisms (protozoa, rotifers, nematodes and water mites) above 5 g Cl−/L indicates that increase of osmotic pressure due to salt has a severe impact. The re-growth of rotifers and nematodes after the recovery period to 0 g NaCl− Cl−/L could be explained by survival of salt tolerant stages in their life cycle. Only nematodes were observed under 10 g Cl−/L, both under long-term adapted reactors and after recovery. Nematodes are known to tolerate large variations in salt levels (APHA 1998). Despite the fact that protozoa are bioindicators for good settling and for good effluent quality with low suspended solids (Al-Shahwani and Horan, 1991; Martin-Cereceda, et al 1996; Madoni, et al 1993; Salvadó, et al 1995), the results of this study demonstrate that good settling and good effluent quality was also possible without protozoa in the presence of salt. The use of salt to remove protozoa, rotifers, nematodes and water mites selectively could be used to investigate the role of these organisms in the activated sludge system. More research is needed on the interaction between bacteria and other organisms in the activated sludge community. 4.4.3 Settling characteristics Elevated salt levels are leading to an increase in water density, which could have a negative impact on the settling characteristics. Besides, dynamic salt levels have been reported to result in unstable sludge blankets in the settler and reduced the separation efficiency (Ekama et al 1997). Panswad and Anan (1999 a, b) concluded that turbid effluent at elevated salt levels was an indication for deterioration of settling properties. However, this study showed improved settling at higher salt levels, which was also shown by Dahl et al (1997) and Yu et al (2002). Also in our experiments effluent turbidity was high, but only immediately after increase of the salt level. The experimental results can be explained by the following two mechanisms: (1) Due to increased salt levels and consequently an increased water density, lighter flocs will be washed out (turbid effluent), while the dense flocs remain, which leads to selection of bigger flocs in the reactor. This phenomenon was successfully applied to develop aerobic granular sludge in sequencing batch airlift reactors (Beun et al 2002). In such a system settling time was chosen to allow only granular sludge to retain, while density plays a similar role in systems under salt stress. (2) The increase of floc size is caused by a combination of electrostatic and hydrophobic interactions with the floc. Christopher et al (1998) observed the same change in floc characteristics due to increased wastewater salinity. In their study the floc achieved new characteristics in about 15 minutes after

Nitrification in saline industrial wastewater

86

increasing salinity. The increase in salt concentration reduced the electric double layers, thereby reducing the overall repulsive force between particles. The microbial aggregates then approached close enough so that increased hydrophobic interactions resulted in increased aggregation and the formation of larger flocs. These two mechanisms can be utilised to achieve good floc characteristics. Gradual increase in salinity will stimulate the selection of dense flocs with minimum washout. On the other hand, sudden increases in salinity increases water density, causing excessive wash out of biomass. Our attempt to directly increase salinity from 0 to 20 g Cl−/L resulted in a severe reduction of biomass due to biomass wash out and biomass flotation. This observation agrees with the dramatic decrease in biomass content as a result of the direct increase in salinity from 0 to 18 g Cl−/L as reported by Panswad and Anan (1999). 4.5 Conclusions Effect of salt on activity • At 10 gCl−/L: Ammonia oxidisers were more sensitive to short and long-term stress (36–39% drop in activity) than nitrite oxidisers (5–13% drop in activity). • At 40 g Cl−/L: inhibition reached 95% of salt free activity for ammonia and nitrite oxidisers in both adapted and non-adapted reactors. • No significant differences in activities were observed between the one-year adapted and the one-month adapted sludge. Effect of salt on microbial population • Nitrosomonas oligotropha was the dominant ammonia-oxidiser at 0 g NaCl– Cl−/L, while Nitrosomonas europaea was the dominant at 10 g Cl−/L salt. Nitrospira sp. was dominant nitrite oxidiser up to 10 g Cl−/L and Nitrobacter sp. could only be detected in small numbers. • Only N. europaea and Nitrobacter sp. were able to survive under high salt levels (40 g Cl−/L salt) • Even with a clear shift in population both systems (one-year adapted and the one-month adapted sludge) were able to fully recover their activity with less diversity in the population of nitrifiers. • Sludge of a conventional domestic nitrifying WWTP is a good source for nitrifiers to become adapted to salt stress via gradual adaptation. • Salt stress above 5 g Cl−/L caused a severe impact on higher organisms (protozoa, rotifers, nematodes and water mites). Starting from this salt level these higher organisms stopped moving and started swelling and bursting. Effect of salt on settling characteristics • Gradual increases in salinity will stimulate the selection of dense flocs and improve settling characteristics.

Long term effects of salt on activity

87

References Al-Shahwani SM, Horan NJ (1991) The use of protozoa to indicate changes in the performance of activated-sludge plants. Water Res. 25:633–8. Andreadakis AD, Kalergis CM, Kartsonas N, Anagostopoulos D (1997) Determination of the impact of toxic inflows on the performance of activated sludge by wastewater charachterization. Wat Sci. Technol. 37:135–142. Antoniou P, Hamilton J, Koopman B, Jain R, Holloway B, Lyberatos G, Svoronos SA (1990) Effect of temperature and pH on the effective maximum specific growth rate of nitrifying bacteria. Water Res. 24:97–101. APHA (1998) Standard methods for the examination of water and wastewater, 20 th edn. American Public Health Association/American Water Works Association/Water Environment Federation, Washington D.C. Belser LW, Mays EL (1980) Specific inhibition of nitrite oxidation by chlorate and its use in assessing nitrification in soil and sediments. Appl. Environ. Microbiol. 39:505–510. Beun JJ, van Loosdrecht MCM, Heijnen JJ (2002) Aerobic granulation in a sequencing batch airlift reactor. Water Res. 36:702–712. Bock E, Koops HP, Harms H, Ahlers B (1991) The biochemistry of nitryfying organisms. In: Shively JM, Barton LL Variation in autotrophic life. Academic press limited, London. Campos JL, Mosquera-Corral A, Sánchez M, Méndez R, Lema JM (2002) Nitrification in saline wastewater with high ammonia concentration in an activated sludge unit. Water Res. 36:2555– 2560. Catalan-Sakairi MAB, Wang PC, Matsumura M. (1997) Nitrification performance of marine nitrifiers immobilized in polyester and macro-porous cellulose carriers. Fermentation and Bioeng. 84:563–571. Catalan-Sakairi MAB, Yasuda K, Matsumura M. (1996) Nitrogen removal in seawater using nitrifying and denitrifying bacteria immobilized in porous cellulose carrier. Water Sci. Technol. 34:267–274. Cousin CP, Ganczarczyk JJ (1998) Effects of Salinity on Physical Characteristics of Activated Sludge Flocs. Water Quality Research Journal of Canada 33:565–587. Dabert P, Delgenes J-P, Moletta R, Godon J-J (2002) Contribution of molecular microbiology to the study in water pollution removal of microbial community dynamic. Re/View in Environmental Science and bio/Technology 1:39–49. Dahl C, Sund C, Kristensen GH, Vredenbregt L (1997) Combined biological nitrification and denitrification of high-salinity wastewater. Water Sci. Technol. 36:345–52. Daims H, Brühl A, Amann R, Schleifer K-H, Wagner M (1999) The domain-specific probe EUB338 is insufficient for the detection of all bacteria: development and evaluation of a more comprehensive probe set. Syst. Appl. Microbiol. 22:434–44. Daims H, Nielsen P, Nielsen JL, Juretschko S, Wagner M (2000) Novel Nitrospira-like bacteria as dominant nitrite-oxidizers in biofilms from wastewater treatment plants: diversity and in situ physiology. Water Sci. Technol. 41:85–90. Daims H, Nielsen JL, Nielsen PH, Schleifer KH, Wagner M (2001a) In situ characterization of Nitrospira-like nitrite-oxidizing bacteria active in wastewater treatment plants. Appl. Environ. Microbiol. 67:5273–5284. Daims H, Purkhold U, Bjerrum L, Arnold E, Wilderer PA, Wagner M (2001b) Nitrification in sequencing biofilm batch reactors: lessons from molecular approaches. Water Sci. Technol. 43:9–18. Dincer AR, Kargi F (1999) Salt inhibition of nitrification and denitrification in saline wastewater. Environ. Technol. 29:1147–1153. Dincer AR, Kargi F (2001) Salt inhibition kinetics in nitrification of synthetic saline wastewater. Enz and Micro Technol. 28:661–665.

Nitrification in saline industrial wastewater

88

Dionisi HM, Layton AC, Harms G, Gregory IR, Robinson KG, Sayler GS (2002) Quantification of Nitrosomonas oligotropha-like ammonia-oxidizing bacteria and Nitrospira spp. from full-scale wastewater treatment plants by competitive PCR. Appl. Environ. Microbiol. 68:245–253. Eikelboom DH (2000) Process control of activated sludge plants by microscopic investigation. IWA publishing Alliance House, London, UK. Ekama GA, Barnard JL, Günthert F.W, Krebs P, McCorquodale JA, Parker DS, Wahlberg EJ (1997) Secondary Settling Tanks: Theory, modelling, design and operation. IAWQ Scientific and Technical Reports, IAWQ London, U.K. Furumai, H, Kawasaki T, Futawatari, T, Kusuda T (1988) Effects on salinity on nitrification in a tidal river. Water Sci. Technol. 20:165–174. Garrido J M, van Benthum WAJ, van Loosdrecht MCM, Heijnen JJ (1997) Influence of dissolved oxygen concentration on nitrite accumulation in a biofilm airlift suspension reactor. Biotechnol. Bioeng. 53:168–178. Gieseke A, Purkhold U, Wagner M, Amann R, Schramm A (2001) Community structure and activity dynamics of nitrifying bacteria in a phosphate-removing biofilm. Appl. Environ. Microbiol. 67:1351–1362. Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water. Sci. Technol. 37:135–142. Hiorns WD, Hastings RC, Head I M, McCarthy A J, Saunders JR, Pickup RW, Hall GH (1995) Amplification of 16S ribosomal RNA genes of autotrophic ammonia-oxidizing bacteria demonstrates the ubiquity of nitrosospiras in the environment. Microbio. 141:2793–2800. Hockenbury MR, Jamro ES (1977) Total Dissolved Solids Effects on Microbiological Treatment. Proceedings of the Industrial Waste Conference:56–62. Hovanec TA, De Longe E (1996) Comparative analysis of aitrifying Bacteria associated with freshwater and marine aquaria. Appl. Environ. Microbiol. 62:2888–2896. Hunik JH, Meijer HJG, Tramper J (1992) Kinetics of Nitrosomonas europaea at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 37:802–807. Hunik JH, Meijer HJG, Tramper J (1993) Kinetics of Nitrobacter agilis at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 40:442–448. Intrasungkha N, Keller J, Blackall LL (1999) Biological nutrient removal efficiency in treatment of saline wastewater. Wat. Sci. Technol. 39:183–190. Juretschko S, Timmermann G, Schmid M, Schleifer K-H, Pommerening-Roser A, Koops H-P, Wagner M (1998) Combined molecular and concentional analyses of nitrifying bacteriumin activated sludge: Nitrosococcus mobilis and Nitrospira-like bacteria as dominant populations. Appl. Environ. Microbiol. 64:3042–51. Kargi F, Dincer AR (1996) Enhancement of biological treatment performance of saline wastewater by halophilic bacteria. Bioprocess Eng. 15:51–58. Koops H-P, Pommerening-Röser A (2001) Distribution and ecophysiology of the nitrifying bacteria emphasizing cultured species. FEMS Microbiol. Ecology 37:1–9. Liebig T, Wagner M, Bjerrum L, Denecke M (2001) Nitrification performance and nitrifier community composition of a chemostat and a membrane-assisted bioreactor for the nitrification of sludge reject waters. Bioprocess Biosyst. Eng. 24:203–210. Manz W, Amann R, Ludwig W, Wagner M, Schleifer KH (1992) Phylogenetic oligonucleotide probes for the major subclasses of Proteobacteria: problems and solutions. Syst. Appl. Microbiol. 15:593–600. Manz W, Amann R, Ludwig W, Vancanneyt M, Schleifer KH (1996) Application of a suite of 16S rRNA specific oligonucleotide probes designed to investigate bacteria of the phylum Cytophaga-Flavobacter-Bacteroides in the natural environment. Microbiol. 142:1097–1106. Madoni P, Davoli D, Chierici E (1993) Comparative-analysis of the activated-sludge micro fauna in several sewage-treatment works. Water Res. 27:1485–91.

Long term effects of salt on activity

89

Martin-Cereceda M, Serrano S, Guinea A (1996) A comparative study of ciliated protozoa communities in activated-sludge plants. FEMS Microbiol Ecol. 21:267–76. Mobarry BK, Wagner M, Urbain V, Rittmann BE, Stahl DA (1996) Phylogenetic probes for analysing abundance and spatial organization of nitrifying bacteria. Appl. Environ. Microbiol. 62:2156–2162. Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003a) Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures. Appl. Microbiol. Biotechnol. 63:217–221. Moussa MS, Lubberding HJ, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM (2003b) Short term effects of various salts on ammonia and nitrite oxidisers in Mixed Bacterial Cultures. Appl. Microbiol. Biotechnol. (submitted). Nogueira R, Melo LF, Purkhold U, Wuertz S, Wagner M (2002) Nitrifying and heterotrophic population dynamics in biofilm reactors: effects of hydraulic retention time and the presence of organic carbon. Water Res. 36:469–481. Panswad T, Anan C (1999a) Impact of high chloride wastewater on an anaerobic/anoxic/aerobic process with and without inoculation of chloride acclimated seeds. Water Res. 33:1165–1172. Panswad T, Anan C (1999b) Specific oxygen, ammonia and nitrate uptake rates of a biological nutrient removal process treating elevated salinity wastewater. Bioresource Technol. 70:237– 243. Picioreanu C, van Loosdrecht MCM, Heijnen JJ (1997) Modelling the effect of oxygen concentration on nitrite accumulation in biofilm airlift suspension reactor. Water Sci. Technol. 36:147–156. Pollice A, Tandoi V, Lestingi C (2002) Influence of aeration and sludge retention time on ammonia oxidation to nitrite and nitrate. Water Res. 36:2541–2546. Pommerening-Röser A, Rath G, Koops H-P (1996) Phylogenetic diversity within the genus Nitrosomonas. Syst. Appl. Microbiol. 19:344–351. Prosser JI (1986) Nitrification, special publication of the society for general microbiology, Oxford IRL Press, Volume 20. Purkhold U, Pommerening-Röser A, Juretschko S, Schmid MC, Koops H, Wagner M (2000) Phylogeny of all recognized species of ammonia-oxidizers based on comparative 16S rRNA and amoA sequence analysis: Implications for molecular diversity survey. Appl. Environ. Microbiol. 66:5368–5382. Salvadó H, Gracia M P, Amigó JM (1995) Capability of ciliated protozoa as indicators of effluent quality in activated-sludge plants. Water Res. 29:1041–50. Sharma B, Ahlert RC (1977) Nitrification and nitrogen removal. Water Res. 11:897–925. Schramm A, de Beer D, van den Heuvel JC, Ottengraf S, Amann R (1999) Microscale distribution of populations and activities of Nitrosospira and Nitrospira spp. along a macroscale gradient in a nitrifying bioreactor: quantification by in situ hybridization and the use of microsensors. Appl. Environ. Microbiol. 65:3690–3696. Schramm A, de Beer D, Gieseke A, Amann R (2000) Microenvironments and distribution of nitrifying bacteria in a membrane-bound biofilm. Environ. Microbiol. 2:680–686. Tijhuis L, van Loosderecht MCM, Heijnen JJ (1994) Solid retention time in spherical biofilms in a biofilm airlift suspended reactor. Biotechnol. Bioeng. 44:867–879. van ‘t Riet K, Tramper J (1991) Basic Bioreactor Design. Marcel Dekker, inc. New York. Vredenbregt LHJ, Nielsen K, Potma AA, Kristensen GH, Sund C (1997) Fluid bed biological nitrification and denitrification in high salinity wastewater. Water Sci. Technol. 36:93–100. Wagner M, Rath G, Amann R, Koops H-P, Schleifer KH (1995) In situ identification of ammoniaoxidizing bacteria. Syst. Appl. Microbiol. 18:251–64. Wagner, M, Rath G, Koops H-P, Flood J, Amann R (1996) In situ analysis of nitrifying bacteria in sewage treatment plants. Water Sci. Technol. 34:231–44. Wagner M, Loy A (2002) Bacterial community composition and function in sewage treatment system. Environ. Biotechnol. 13:218–227.

Nitrification in saline industrial wastewater

90

Woolard CR, Irvine RL (1995) Treatment of hypersaline wastewater in the sequencing batch reactor. Water Res. 29:1159–1168. Yu SM, Leung WY, Ho KM, Greenfield PF, Eckenfelder WW (2002) The impact of sea water flushing on biological nitrification-denitrification activated sludge sewage treatment process. Water Sci. Technol. 46:209–216.

Chapter 5 Modelling Nitrification, Heterotrophic growth and Predation in Activated Sludge Submitted as: Moussa MS, Lubberding HJ, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM Modelling Nitrification, Heterotrophic growth and Predation in Activated Sludge

Abstract A mathematical model describing the interaction between nitrifiers, heterotrophs and predators in wastewater treatment has been developed. The inclusion of a predation mechanism is a new addition to the existing activated sludge models. The developed model considered multi-substrate consumption and multi-species growth, maintenance and decay in a culture where nitrifiers, heterotrophs and predators (protozoa and metazoa) are coexisting. Two laboratory-scale Sequenced Batch Reactors (SBRs) operated at different sludge ages of 30 and 100 days for a period of four years were used to calibrate and validate the model. Moreover, to assess the predators activity a simple procedure was developed, based on measuring the respiration rate with and without the presence of the predators. The model successfully described the performance of two SBRs systems. The fraction of active biomass (ammonia oxidisers, nitrite oxidisers and heterotrophs) predicted by the proposed model was only 33% and 14% at SRT of 30 and 100 days, respectively. The high fraction of inert biomass predicted by the model was in accordance with the microscopic investigations of biomass viability in both reactors. The presented model was used to investigate the effect of increasing sludge age and the role of predators on the biomass composition of the tested SBR system.

5 Modelling Nitrification, Heterotrophic growth and Predation in Activated Sludge 5.1 Introduction Activated sludge systems form complex microbial ecosystems, comprising of bacteria, bacteriophages, protozoa and metazoa. Bacteria clearly play a vital role in the conversion of the wide diversity of organic compounds present and in the removal of nitrogen and phosphorus. This fact attracted researchers to acquire knowledge on the bacterial conversion processes in wastewater treatment over the last decade, which resulted for example in the successful design of nutrient removal plants and in the activated sludge models (ASM models, Henze et al 2000). The current model and design concepts consider bacteria as the sole active biomass. The activities of all other microbial community members (protozoa, metazoa, phages etc.) are hidden in a simple decay process responsible for the reduction of active biomass. This decay process is the sum of several independent processes like maintenance, lysis due to phage infection and predation (van Loosdrecht et al 1999). The successful use of present day activated sludge models does not show a need for including predation. However, the clear presence of predators in wastewater treatment plants can not be neglected. The role of the predators might influence the performance of the treatment plant and also lead to a change of the kinetic parameters of the micro-organisms from one plant to another. The role of protozoa in activated sludge systems was first investigated in detail by Curds and coworkers (1970; 1971 a, b; 1973). They showed that the role of protozoa is crucial in obtaining a good effluent quality with low suspended solids. Recent work has pointed to the use of protozoa as bio-indicators linked to process performance and effluent quality (Martin-Cereceda et al 1996; Salvadó et al 1995; Madoni et al 1993; Al-Shahwani and Horan 1991). Despite the fact that a lot of work has been conducted in both mathematical modelling and in the study of the microbial ecology of activated sludge systems, little work has been reported on the interaction between bacteria and other micro-organisms in the activated sludge microbial community. Especially the role of protozoa is still largely under-exposed in the academic research (van Loosdrecht et al 1999). The work presented here was carried out to quantify the interaction between nitrifiers, heterotrophs and predators by developing a model for multi-substrate consumption and multi-species growth, maintenance and decay in which nitrifiers, heterotrophs and predators are coexisting. The model was applied to: 1- Describe the performance of an Sequenced Batch Reactor (SBR) with an enriched nitrifying culture at different sludge retention times (SRTs);

Nitrification in saline industrial wastewater

96

2- Investigate the role of sludge age on the different fractions of biomass (nitrifiers, heterotrophs, predators) and on the accumulation of inert biomass in the SBR system; 3- Determine the impact of predators on the presence and activity of the nitrifiers and heterotrophs.

5.2 Theory 5.2.1 Interactions between bacteria and predators A nitrifying laboratory-scale SBR system was operated for over four years. A constant presence of different types of higher organisms (protozoa and metazoa) (Figure 5.1) was observed microscopically, indicating the stability of such an ecosystem. The relative abundance of the protozoa and metazoa was microscopically determined to be stable over the full experimental period. A schematic diagram of a food web describing the relation between the different species is given in Figure 5.2. The influent ammonia (SNH4) and organic carbon (SS) are oxidised to nitrate (SNO3) and CO2 by nitrifying and heterotrophic bacteria, respectively. These bacteria form the prey for predators (protozoa, metazoa), which excrete faecal pellets that contain inert COD (Schlimme et al 1997). Besides growth, bacteria and higher organisms are subjected to decay by viral attacks and other factors, which have in this paper been neglected. The decay products of nitrifiers, heterotrophs and predators consist of degradable organic carbon (available substrate for heterotrophs only) and inert organic carbon.

Figure 5.1 Examples of protozoa and metazoa forming a stable population the nitrifying the SBR sludge. A: Sessile ciliates B: Rotifers C: Nematodes D: Mite

Modelling nitrification, heterotrophic growth

97

Figure 5.2 Schematic diagram of the proposed food web illustrating the flow of the external and the internal substrate in addition to the interaction between the involved bacteria and predators. 5.2.2 Description of the Model The dynamic model is based on the schematic diagram of the food web, presented in the form of a matrix similar to the ASM models (Henze et al 2000). This matrix includes the process kinetics and stoichiometry and can be found in Appendix 5.1. The model includes six soluble compounds (dissolved oxygen, nitrogen gas, ammonia, nitrite, nitrate and COD) and five types of biomass (ammonia oxidisers, nitrite oxidisers, heterotrophs, predators and inert biomass) as particulate compounds. Stoichiometric and kinetic parameter values are given in Appendix 5.1. Key features and assumptions of the model are summarised below.

Nitrification in saline industrial wastewater

98

1. Nitrification is considered as a two-step process, carried out by two types of nitrifying biomass: ammonia and nitrite oxidisers. Since the SBR was operated at a temperature above 20°C this two-step approach is necessary in order to predict possible nitrite accumulation in such system (Hellinga et al 1998, 1999; Brouwer et al 1998; Nowak et al 1995). 2. Heterotrophic organisms are responsible for COD utilisation under both aerobic and anoxic conditions (denitrification). These two processes are similar, except that under anoxic conditions nitrate is used as electron acceptor instead of oxygen (Henze et al 2000). Nitrite was in this paper not considered in the denitrification process. Since nitrite was only present for a short period under aerobic conditions and since the anoxic period was not significant in the studied SBR system (only occurred shortly at the end of settling period where no nitrite is present). 3. Growth, maintenance and decay processes are described as suggested by Beeftink et al (1990) and applied by De Gooijer et al (1991); Hunik et al (1994); Leenen et al (1997). This means that the substrate (SNH4, SNO2, SS) is utilised for growth and maintenance of the ammonia oxidisers, nitrite oxidisers and heterotrophs. 4. The kinetic expressions in the model are based on switching functions (Monod equations) for all soluble compounds consumed (Henze et al 2000). 5. Two decay rates are considered in the model: aerobic decay occurs when the bacteria are starving in the presence of oxygen, while anoxic decay occurs when the bacteria are starving in the absence of oxygen and presence of NO3− (Leenen, et al 1997; Siegrist et al 1999). 6. A predation mechanism is considered in the model by introducing predators as active biomass. Predators grow aerobically on the degradable (1−Fxi) fraction of the three types of bacteria available. The predation rate is a function of the bacterial concentration and equal preference for predation of different bacterial types was assumed. Due to lack of data on the behaviour of predators under different environmental conditions, the reduction of the active predators is lumped together into one decay process. This process represents the sum of all decay and loss processes of the predators like lysis due to phage infection, predation by metazoa or destruction due to mechanical mixing. 7. The decay and predation processes of the active biomass result in the generation of inert biomass that is not further metabolised (Fxi). 5.3 Materials and methods 5.3.1 Continuous operation of the Sequenced Batch Reactors The study was carried out in two laboratory scale reactors with a working volume of 2.5L at 30°C. The two reactors were operated automatically as Sequenced Batch Reactors (SBRs). The pH was maintained at 7.5+0.05 using 0.25 M NaOH and 0.25 M HCl (BIO controller ADI 1030 coupled with BioXpert 1.1x data acquisition and control program; Applikon b.v.Schiedam, The Netherlands). Nitrifying activated sludge obtained from a domestic wastewater treatment plant was used to inoculate the SBR systems. The SBR systems were operated for four years in cycles of 6 hours including a 10 minutes fill

Modelling nitrification, heterotrophic growth

99

period, 4 hours reaction period, 80 minutes for settling, and 30 minutes for effluent discharge. The characteristics of the operating conditions are summarised in Table 5.1. A synthetic medium containing mainly ammonia and nutrients to enhance the microbial growth was used as SBR feeding. 1.5L of medium was fed during the filling period and the effluent was removed at the end of the settling period. The Sludge Retention Time (SRT) which was desired, was set by the amount of wasted sludge, which was removed from the mixed reactor during each cycle and the biomass in the effluent (Figure 5.3). Aeration was provided during the reaction period with airflow of 120 L/h. The two reactors were continuously monitored (on-line measuring of DO, pH, addition of NaOH) and sampled (MLSS, MLVSS, NH4+, NO2−, NO3−) during several cycles. On-line cyclic measurements of DO and the addition of the amount of base solution consumed and constant biomass concentration in the reactors confirmed a steady state condition. The sludge age in each reactor was initially set at 100 days. When steady state was reached, the second reactor was switched to operate at a lower SRT (30 days).

Table 5.1 Operational conditions of the sequenced batch rectors (SBR) N-Load

1200mg-N/L.day

pH

7.5

COD Load

60 mg-COD/L.day

Temperature

30 °C

HRT

10 h

Stirrer speed

650 rpm

(SBR30days)

30 days

Aeration

120 L/h

(SBR100days)

100 days

SRT

Nitrification in saline industrial wastewater

100

Figure 5.3 Schematic representation of the experimental set-up of the two sequencing batch reactors (SBRs). (1) Fermentor 2.5L; (2) Biocontroller; (3) P.C. with Bioxpert; (4) Connecting cables; (5) Digital (on/off) outputs; (6) pH and O2 electrodes; (7) Stirrer, engine; (8) Aeration; (9) Influent; (10) Base pump; (11) Acid pump; (12) Effluent; (13) Excess sludge; (14) Water bath. 5.3.2 Media Synthetic medium prepared with de-mineralised water had the following composition: (NH4)2CO3 857.95 mg/L as ammonia source, NaH2PO4.H2O 167.5 mg/L, MgSO4.7H2O 90 mg/L, CaCl2.H2O 14 mg/L, KCl 36 mg/L, yeast extract 1 mg/L, nutrient solution 0.3 mL/L. The nutrient solution was added to the medium in order to enhance the microbial

Modelling nitrification, heterotrophic growth

101

growth and was prepared with the following chemicals mixed in one litre of demineralised water: 1.5g of FeCl3.6H2O, 0.15g of H3BO3, 0.03g of CuSO4.5H2O, 0.18g of KI, 0.12g of MnCl2.4H2O, 0.06 g of Na2MoO4.2H2O, 0.12 g of ZnSO4.7H2O, 0.15 g of CoCl.6H2O, and 10 g of EDTA. The very low COD in the influent medium (10 mg/L) was to enhance the growth of nitrifiers over heterotrophs. 5.3.3 Respiration activity assay The biomass activity as well as viability was estimated by measuring the oxygen uptake rate (OUR) in a biological oxygen monitor (BOM). This is a batch type of respirometer with the possibility to inject the required substrate directly into the reaction chamber of 10 mL. Fresh biomass samples were withdrawn directly from the SBR (at the end of the reaction period), washed and re-suspended in medium before testing. Washing and resuspending of bacterial cells in a medium, which has sufficient buffering capacity and no adverse effect on the bacterial activity, was required to remove any remaining substrate and to stabilise the pH during the test. When re-suspending the bacteria in the same medium, a 5–10 times dilution was usually required in order to avoid having an activity faster than the response time of the oxygen electrode. 10 mL of the washed cells suspended in medium were transferred to the stirred BOM vessel and aerated for 10 min. The dissolved oxygen concentration was kept above 4.0 mg/L and the pH at 7.5. Temperature was controlled at 30°C by means of a water bath. The oxygen probe was sealed in the BOM vessel in such a way that no air bubbles remained in the liquid. The decrease in oxygen concentration was monitored and recorded via a data acquisition system. In order to differentiate between the activity of the different biomass fractions (nitrite oxidisers, ammonia oxidisers and heterotrophs), respiration was measured in the presence of relevant substrates. Different substrates were injected in the reaction chamber through a seal in the oxygen probe using an analytical syringe (Henze et al 2000; Cronje et al 2001; Ziglio et al 2002). 5.3.4 Determination of the predators activity The respiration under starvation condition with and without the presence of the predators was measured to determine the activity of predators. A shock load of NaCl was used to eliminate the predators. The impact of salt on the presence of the predators (protozoa, rotifers and nematodes) was followed in phase contrast mode using a light microscope (Olympus). The minimum dose of NaCl required for full elimination of the predators was 5 g NaCl–Cl−/L. The respiration activity of the samples treated with salt was measured using the BOM and expressed as percentage of the non-treated sample. 5.3.5 Analytical procedures The data acquisition BioXpert 1.1x (Applikon b.v.Schiedam, The Netherlands) was used to continuously store the monitored information (DO, pH) from the SBR system. Ammonia and nitrite were measured spectrophotrometrically in accordance with Standard Methods (APHA 1998). Nitrate-nitrogen was determined using Dionex 4500i series and Shimadzu C_R5A ion-chromatograph.

Nitrification in saline industrial wastewater

102

The MLSS and MLVSS determination was performed after filtration of a 10 mL sample of mixed liquor on a Whatman glass micro fibre filter (GC/F) filter. Dry weight was determined after the filter was dried for 24 h at 105°C and weighted on a microbalance. The ash content was calculated after incinerating the dried filter in an oven for 1 h at 550°C. The sludge retention time (SRT) of the reactor was calculated from the biomass concentration (MLSS/MLVSS) in the reactor and the biomass concentration in the effluent. The respiration measurements were performed in an on-line respirometer (Smolders et al 1994). The respirometer is a stirred, non-aerated, thermostated, 25-mL vessel with a DO electrode, connected to the SBR. Sludge was pumped directly from the SBR through the respirometer. The pump was automatically switched on and off within a cycle time of 5 min. When the pump was switched on (1 min), the SBR content was pumped (circulated) through the respirometer. When the pump was switched off (4 min), the decrease in oxygen concentration in the respirometer due to respiration was measured. Via this way the respiration rate could be calculated (using linear regression) every 5 minutes during one SBR cycle. The floc size in the reactors was followed using image analysis. The average floc diameter was measured using a representative sample, in which at least 500 particles were analysed (Tijhuis et al 1994). 5.3.6 Staining techniques (live-dead stain) Molecular Probes’ LIVE/DEAD® BacLight™ Bacterial Viability Kits (Moleculer Probes, chemical no. L-7007 and L-7012) were used to discriminate between viable cells and dead cells (Ziglio et al 2002; Barbesti et al 2000; Weinbauer et al 1998). Staining was carried out by adding 3µL of the stock stain solution (prepared by adding 10 µL of SYTO® 9 green-fluorescent nucleic acid stain; 10µL of propidium iodide and 180 µLof Mill Q) to 100 µL of the microbial sample suspension, after washing the sample three times with DDW. The mixture was then incubated at room temperature in the dark for 15 minutes after well mixing. After the incubation, the stained microbial suspension was examined with a fluorescence microscope. 5.3.7 Simulation model The developed model was applied to analyse the performance of the nitrifying SBR systems considered in this study. Aquasim (Reichert et al 1994) was used as modelling tool. The model was used to simulate the conversion processes that occur in the laboratory SBR, to get insight in the biomass composition in order to understand the role of predators and sludge age on the population structure and function. 5.3.8 Reactor description in the model The SBR in the model was described with a variable volume (1.0 L min- 2.5 L max) mixed reactor compartment (SBR compartment). In order to simulate periodically influent addition, effluent removal and aeration a modulo function was used. Within each cycle this function increases linearly from 0 to the total cycle time and jumps back to 0 at

Modelling nitrification, heterotrophic growth

103

the beginning of the next cycle. For the simulation of the solids separation (settling) another completely mixed compartment (constant volume) was introduced (effluent compartment) and connected to the main one with an advective link with a purification recycle to return only the solid content in the outflow back to the SBR compartment. An additional process for solids removal from the SBR compartment as function of the required SRT was used to simulate the biomass wasting from the system. Aeration was simulated by introducing a gas compartment (completely mixed compartment) connected with the SBR compartment using a diffusive link. A schematic drawing of the SBR configuration in Aquasim is given in Figure 5.4.

Figure 5.4 Schematic drawing of the SBR as implemented in the Aquasim simulation model. 5.3.9 Model calibration and validation A step-wise approach for model calibration was applied as proposed by Meijer et al (2001). Firstly, the solid content of the reactor was fitted under steady state conditions (simulation time is larger than 6 times the SRT). Three parameters, predation rate, predators decay rate and fraction of inert COD generated in biomass lysis (bpred, bG and Fxi), were used to fit the solid content of the reactor, the respiration rate at the end of the cycle when there is no external substrate present any more and the predators respiration. Other stoichiometric and kinetic parameters needed for the model are given in Appendix 5.2, and were mainly derived from literature. Secondly, the forms of nitrogen and respiration rate measurements within the SBR cycle were fitted by calibrating the affinity constant for oxygen, nitrite and ammonia of the ammonia and nitrite oxidisers

Nitrification in saline industrial wastewater

104

All calibration steps were performed on the SBR operating at 30 days SRT. The calibrated model was validated by checking its capacity to predict the measured data from a SBR operating at SRT of 100 days. 5.4 Results 5.4.1 Steady state operation During the first week of operation, the initial concentration of inoculation sludge (3349 mg MLVSS/L) dropped to 2000 mg MLVSS/L due to removal of unsettled sludge via the effluent. Nitrification activity was observed from the first cycle. Gradually the effluent became clear and the amount of well settling biomass (with a brownish colour) increased in the reactors. After 170 days, constant and similar volumetric ammonia and nitrite removal rates for the two reactors at 100 days SRT were observed. After 200 days, the SRT of one of the reactors was switched to a lower SRT (30 days). Constant on-line measurement results of DO, a constant base addition and a constant VSS concentration in the reactors were used as an indication of a steady state condition. The steady state cyclic profiles of NH4+, NO2−, NO3− and respiration rate in the two reactors are shown in Figure 5.5. Ammonia was consumed within 95 and 115 minutes at SRT of 30 days and 100 days, respectively. Accumulation up to 40 mg/L NO2–N was observed in the reactor operated at lower SRT (SBR30 days), while only 4 mg/L NO2–N was detected at SRT of 100 days. Full oxidation of ammonia and nitrite occurred within 2 hours, the rest of the cycle was a starvation period for the ammonia and nitrite oxidisers. The measured respiration rate in this period represented the heterotrophic growth as function of the decayed material and predation of the predators. A higher biomass content and bigger floc diameter were observed in the reactor operated at SRT of 100 days. The operational parameters describing the reactor performance are represented in Table 5.2. Regular microscopic investigation of the two reactors clearly showed a stable presence of protozoa and metazoa (as shown in Figure 5.1) over four years of continuous operation.

Table 5.2 Average measurements over the operation period of 4 years for the two reactors operated at 30 and 100 days SRT. Parameters

unit

R30days

R100days

MLSS

mg/l

1220

±30

3268

±57

MLVSS

mg/l

1140

±29

2995

±84

Floc diameater

µm

70

±4

90

±5

Modelling nitrification, heterotrophic growth

105

Figure 5.5 Concentration profiles of NH4, NO2, NO3 and OUR during a representative cycle in the two reactors SBR30 days and SBR100 days at steady state conditions. Model calibrated at steady state for SRT=30 days. Calibrated model used to simulate steady state for SRT =100 days (validation). 5.4.2 Determination of the Predators Activity Different types of inhibitors for predators (Metronidazole, cycloheximide and nystain, etc.) were tested to eliminate the predators (data not shown). These inhibitors have been reported and applied as specific agents acting on eucaryotic cells only (Novitsky and Dalhousie 1990; Lee and Welander 1994). In general it took a long time (couples of hours) before any effects were observed. Moreover, only some types of predators were affected. A chemical compound that affects all higher organisms present could not be found in these tests. Meanwhile, the nitrifying biomass cultivated in the SBR system was used to investigate the effect of shock loads of salt on the nitrification. A serious impact of salt on the presence of the predators was observed during these tests. It was interestingly enough to draw the attention to investigate the use of salt as an inhibitor for predators. The addition of 5g NaCl–Cl−/L had an immediate effect on the protozoa and metazoa in samples from both reactors. Microscopic inspection made clear that they stopped moving and started swelling and bursting (Figure 5.6), probably due to an increased osmotic pressure. A reduction of about 12–15% in respiration activity under starvation condition was observed after exposure to salt. The activity of ammonia and nitrite

Nitrification in saline industrial wastewater

106

oxidisers and of heterotrophs was not affected (reversible) in a respirometric activity test when the salt was washed away (data not shown).

Figure 5.6 Effects of salt (5 g NaCl– Cl−/L) on ciliates and rotifers. 5.4.3 Simulation model results Model calibration and validation The model was calibrated for the SBR reactor operated at 30 days SRT. In the first step of the calibration procedure, the values of 0.01 day−1 for the predation rate (bpred), 0.15 day−1 for the predators decay rate (bG) and 0.15 for the fraction of inert COD generated in biomass lysis (Fxi) were estimated. With these values the model successfully predicted the VSS, respiration at the end of the cycle and predators respiration as well. Consequently, the affinity constants for oxygen, nitrite and ammonia of ammonia and were estimated (Appendix 5.2). The model nitrite oxidisers under steady state conditions correctly described the NH4, NO2, NO3 and respiration rate profiles. The calibrated model was validated by using it to describe the second reactor operated at 100 days SRT. The validation steps were performed with the same input where a value of 3 mg O2/L model parameters as used for 30 days SRT except for was used instead of 1 mg O2/L to fully describe the nitrification and oxygen uptake (Figure 5.5). This difference in the apparent oxygen affinity constant resulted from a stronger diffusion gradient. Floc size measurements showed a larger floc size (90 µm) at SRT 100 days compared to the floc size at SRT 30 days (70 µm). The increased floc size

Modelling nitrification, heterotrophic growth

107

might result in an increased diffusion gradient reflected by a higher apparent affinity (Beccari et al 1992). constant The fraction of the active bacteria (ammonia oxidisers, nitrite oxidisers and heterotrophs) predicted by the proposed model was 33% and 14% at SRT 30 and 100 days, respectively. The remaining fraction (67% and 86%) represents the dead biomass (particulate inert fraction) and predators. These results were in accordance with the livedead stain tests in which the microscopic investigation of the biomass viability in both reactors confirms the high fraction of inert biomass predicted by the model (see figure 5.7). However, quantitative differentiation between the two reactors using live-dead stain tests alone was not possible, as the difference is not significant enough to be microscopically quantified.

Figure 5.7 Bacterial viability in the SBR 30 days, where LIVE/DEAD BacLight Bacterial Viability Kits was applied. The viable cells are green and the dead cells red. 5.5 Discussion 5.5.1 Determination of the predators activity The present study successfully applied a simple procedure in which the activity of the predators could be measured in suspended mixed cultures. Salt shock was found to be a lethal agent acting only on the protozoa and metazoa and not affecting the bacterial activities. A 5 g NaCl–Cl−/L salt solution was used to eliminate the protozoa and metazoa

Nitrification in saline industrial wastewater

108

immediately. The results of using other types of inhibitors (Metronidazole, cycloheximide and nystain, etc.) besides salt were in agreement with the findings of Lee and Welander (1994). They reported that the addition of cycloheximide and nystatin to a biofilm reactor caused a dramatic decrease in the amount of rotifers and nematodes but that the number of attached ciliates and flagellates remained unchanged. Also attempts to introduce alternating anoxic/ anaerobic periods or lowering pH to a value below 6 were not successful to eliminate all protozoa and metazoa. This is in accordance with the findings of van Dongen et al (2001). The quick and direct effect of salt on both protozoa and metazoa makes salt very efficient to use to determine the predators activity. Measuring the contribution of the predators to the respiration under starvation condition gave sufficient information to assess the contribution of the predators in a mixed suspended culture. 5.5.2 Modelling of the predators activity The developed model simplifies the complex reality of the predation-prey relationships: 1. All the different types of predators (protozoa and metazoa) are described as one type (predators, XG). This overcomes the lack of information on the kinetic and stoichiometric parameters for the different types of predators and simplifies the complexity of the interaction between the bacteria and predators and between predators itself (Ratsak et al 1996). This simplification means that the kinetic coefficients will have to be calibrated for each different system. 2. The predation process does not discriminate between the different types of bacteria present but is a function of the bacterial concentration. This assumption is in agreement with Griffiths (1989), who did not find any protozoa that would graze specifically on nitrifiers. Moreover, the reduction of the active predators is simplified in one process (decay of the predators). No other choice could be made since the interactions between the predators are rather complex and adequate information to explicitly describe it is still lacking. 5.5.3 Simulation of the effect of SRT on the active biomass fraction The validated model was used to simulate the effect of SRT on the active biomass fraction. The simulations were conducted up to SRT of 100 days. The model results were compared at the steady state of each simulated SRT. For the case of 30 days SRT, the active biomass fraction reaches its steady state within one SRT where the inert biomass takes at least 4 times the SRT to reach a steady state. Consequently, volumetric oxidation rates expressed per volume are reaching a constant value faster than the biomass specific oxidation rates (Figure 5.8). The impact of increasing the sludge age on the nitrifying SBR system is illustrated in Figure 5.9. It is apparent that the biomass content increases with increasing SRT. However, the increase in ammonia/nitrite oxidisers, heterotrophs and predators content follows a saturation curve and reaches its maximum value at 40 days SRT. The inert biomass concentration increases linearly as long as the SRT increases. Consequently, the volumetric ammonia and nitrite oxidation rate follow a saturation curve as well while the specific oxidation rate decreases as a result of the accumulation of inert biomass. These

Modelling nitrification, heterotrophic growth

109

model results are in agreement with experimental results reported by Pollice et al (2002). They mentioned that the specific ammonia oxidation rate dropped to 14% of its initial value as the consequence of increasing the SRT from 3 days to 24 days in a nitrifying SBR system operated at 32°C and pH above 7.2. These results make the role of SRT in nitrifying SBR systems clear. Increasing the SRT increases the active biomass fraction and the oxidation rate up to a maximum level (40 days SRT). Any further increase will not lead to any volumetric improvement. When increasing the SRT till 40 days (the maximum volumetric oxidation rate), the active biomass concentration will increase in the SBR, resulting in a faster volumetric substrate utilisation rate and a longer starvation period per cycle. The active biomass reaches a saturation level and only the inert biomass increases with increasing SRT. In these model simulations and experimental tests the cycle length was kept constant. Operating the SBR system at a shorter cycle length will lead to an increased volumetric oxidation rate. Therefor, other systems operated at a long SRT (such as membrane bioreactors) need to be optimised carefully, to avoid accumulation of high amount of inert biomass and high operational cost without gaining any volumetric improvement.

Figure 5.8 Change in variables during simulations of SBR at SRT=30 days towards steady state.

Nitrification in saline industrial wastewater

110

The proposed model predicted the presence of heterotrophic bacteria in the system as results of the influent COD (10 mg/L) and generated COD in the decay mechanism. The contribution of the influent COD and the SRT on the heterotrophic concentration is illustrated in Figure 5.10. The influent COD is responsible for about 40% of the total formed heterotrophic biomass. About 60% resulted from the decay mechanism. This can be calculated by taking the heterotrophic biomass concentration at 10 mg/L and subtract the biomass concentration at 0 mg/L. Furthermore, the heterotrophic biomass increased by 11% as a consequence of increasing the SRT from 30 to 100 days. These results show the significance of a low input of COD on the formed heterotrophic biomass in an autotrophic system. Such a low value of COD (5–10 mg/L) could be indirectly introduced to a similar system via aeration (contaminated air with traces of COD) and via media preparations (water and nutrient compounds can contain traces of organic COD). The presence of heterotrophs were also detected in non-sterilised chemostat systems used to cultivate ammonia and nitrite oxidisers, despite the fact that original cultures were free of heterotrophs and sterilised media were used (Rittman et al 1994). They claim that ammonia and nitrite oxidisers produce soluble microbial products (SMP) that can support the detected heterotrophic population. In order to describe this, high yield values for ammonia and nitrite oxidisers (0.44 and 0.12 mg CODcell mg−1, respectively) were used to predict proper biomass composition of autotrophs and heterotrophs.

Figure 5.9 Effect of sludge age on the active fraction of the biomass at steady

Modelling nitrification, heterotrophic growth

111

state in SBR reactors as predicted by the developed model.

Figure 5.10 Effect of influent COD concentration on the content of the heterotrophic biomass fraction in the nitrifying SBR system fed with ammonia as sole energy source and nutrient to enhance the microbial growth (the source of low input COD, 10 mg/L to the system) The hypothesis of the soluble microbial production by nitrifiers can not be excluded. However, the use of the death-regeneration concept in addition to a small input COD in the influent quantified the presence of the heterotrophs proper in our model. This simplified the parameter estimation and prevented the use of unrealistic high yield parameters for nitrifiers as in case where using the SMP concept in modelling such systems. 5.5.4 Simulation of the role of predators on the active biomass fraction The role of predators on a nitrifying SBR system was investigated using the developed model. The simulations were conducted under similar operational conditions for the two reactors but with switching off all the predators’ processes. The model results of the biomass fractions at steady state in both cases (with and without predators) are illustrated in table 5.3. No significant change in the total biomass content (MLVSS) was observed in both reactors. This could be explained by the fact that the main fraction of the total

Nitrification in saline industrial wastewater

112

biomass is particulate inert (60 and 80%) and reduction in the amount of predators was compensated by an increased active biomass fraction. Meanwhile, a dramatic increase in the active biomass fraction was observed in both reactors when there were no predators present. An almost double increase of the active biomass fraction was calculated for the reactor of 100 days SRT, gaining significant higher oxidation rate. In ASM 1 model (Henze et al 2000) such an effect would have been corrected for by increasing the lysis rate.

Table 5.3 Model simulation results on the effect of predators on the active fraction of biomass in the two reactors operated at 30 and 100 days SRT. Parameters

R30days unit

MLVSS

mg/L

XA

mgCOD/L

R100days

No predators With predators No predators With predators 1186

1111

3171

3011

395

265

442

290

23

17

18

7

166

127

190

153

10

8

8

4

168

129

199

142

10

8

8

3

mgCOD/L

−

97

121

%

−

6

3

955

960

1684

3570

%

57

61

67

83

%

43

33

33

14

% Dead fraction (Inert+predators)

57

67

67

86

% XN

mgCOD/L %

XH

mgCOD/L %

Xpredators XI Active bacteria

mgCOD/L

A number of researchers already studied the influence of predators on bacterial activity in activated sludge systems. Curds (1971) presented a simple mathematical model of the activated sludge system. It considers the fate of a single substrate compound, two types of bacteria (flocculent and non-flocculent), and two forms of ciliated protozoa. The application of the model to study the population dynamics of these organisms in completely mixed reactor experiments could predict the correlation between the presence of ciliated protozoa and effluent quality. However, this attempt was mainly emphasising the correlation between ciliated protozoa and effluent turbidity. Lee and Welander (1994) investigated the role of predators on the nitrification in lab-scale aerobic biofilm reactors. They concluded that the addition of nystain and cycloheximide to the test reactor lead to a rapid decrease in the quantity of biofilm-consuming predators, most of them rotifers

Modelling nitrification, heterotrophic growth

113

and nematodes, and a simultaneous increase in nitrification which finally stabilised at a level twice as high as in the reference reactor. Bouchez et al (2000) showed that predation by protozoa was the main reason for the failure of bioaugmentation in a nitrifying SBR system. The disappearance from the reactor of added bacteria coincided with the overgrowth of protozoa. The results of this study present the possibility for increasing the nitrification activity by suppressing the growth of predators in a nitrifying system or other systems in which slow-growing bacteria play an important role. However, the use of salt as a selective inhibitor for predators under practical conditions needs to be further investigated. Suitable dosage strategy and the effect of the loss of predators on the whole system performance should be carefully considered. 5.6 Conclusions A rather simple procedure for the determination of the predators activity in suspended mixed culture has been developed. Salt shock was used as a selective agent acting only on the protozoa and metazoa and not affecting bacterial activities. The procedure was successfully tested, verified and applied on a lab-scale nitrifying SBR system. A model was developed to describe a mixed culture in which nitrifiers, heterotrophs and predators (protozoa and metazoa) are coexisting. The developed model proved to be capable of describing the interaction between nitrifiers, heterotrophs and predators in two laboratory-scale nitrifying (SBRs) systems operated at different sludge ages. The model was a helpful tool to get insight in the system and to investigate the impact of increasing sludge age and the role of predators on the performance of the nitrifying SBR system. The model results showed the need for careful optimisation of systems operated at long SRT (such as membrane bioreactors), to avoid accumulation of high amounts of inert biomass and to avoid high operational costs without gaining any volumetric improvement. The model showed its capacity to elucidate the biological processes in activated sludge systems by including the effect of the predators. The practical application of the developed model and assessments of predators activity needs to be verified under full scale activated sludge plant operation. References Al-Shahwani SM, Horan NJ. (1991) The use of protozoa to indicate changes in the performance of activated-sludge plants. Water Res.25:633–8. APHA (1998) Standard methods for the examination of water and wastewater, 20th edn. American Public Health Association/American Water Works Association/Water Environment Federation, Washington D.C. Barbesti S, Citterio S, Labra M, Baroni MD, Neri MG, Sgorbati S (2000). Two and three-color fluorescence flow cytometric analysis of immunoidentied viable bacteria. Cytometry. 10:214–8. Beccari M, DI Pinto AC, Ramadori R, Tomei MC (1992) Effect of dissolved oxygen and diffusion resistances on nitrification kinetics. Water Res. 26:1099–1104.

Nitrification in saline industrial wastewater

114

Beeftink HH, van der Reijden RTJM, Heijnen JJ (1990) Maintenance requirements: Energy supply from simultaneous endogenous respiration and substrate consumption. FEMS Microbiol. Ecol. 73:203–210. Bouchez T, Patureau D, Dabert P, Juretschko S, Dore J, Delgenes P, Moletta R, Wagner M (2000) Ecological study of a bio-augmentation failure. Environmental Microbiology 2:179–190. Brouwer H, Bloemen M, Klapwijk B, Spanjers H (1998) Feed-forward control of nitrification by manipulating the aerobic volume in activated sludge plants. Wat. Sci. Tech. 38:245–254. Cronje CL, Beeharry AO, Wentzel MC, Ekama GA (2002) Active biomass in activated sludge mixed liquor. Water Res. 36:439–444. Curds CR, Cockburn (1970) Protozoa in biological sewage-treatment process-I. A survay of the protzoan fauna of British percolating filters and activated sludge plants. Water Res. 4:225–236. Curds CR (1971a) A computer-simulation study of predator-prey relationships in a single-stage continuous-culture system. Water Res. 5:793–812. Curds CR (1971b) Computer simulation study of microbial population dynamics in the activatedsludge process. Water Res. 5:1049–1066. Curds CR. (1973) A theoretical study of factors influencing the microbial population dynamics of the activated sludge process-I. Water Res. 5:1269–1284. De Gooijer CD, Wijffels RH, Tramper J (1991) Growth and substrate consumption of Nitrobacter agilis cells immobilized in carrageenan: Part 1. Dynamic modelling. Biotechnol. Bioeng. 38:224–231. Griffiths BS (1989) The effect of protozoan on nitrification-implications from application of organic wastes applied to soils. In: Nitrogen in organic wastes applied to soil. Eds. J.AA.Hansen, K Henriksen. Academic Press, London:37–46. Gujer W, Larsen TA (1995) The implementation of biokinetics and conservation principles in ASIM. Wat. Sci. Technol. 31:257–266. Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Wat. Sci. Tech. 37:135–142. Hellinga C, van Loosdrecht MCM, Heijnen JJ (1999) Model based designed of a novel process for ammonia removal from concentrated flow. Mathematical and Computer Modelling of Dynamic Systems 5:351–371. Henze M, Gujer W, Mino M, van Loosdrecht MCM (2000) Activated Sludge Models ASM1, ASM2, ASM2d and ASM3. IWA Publishing, London. Hunik JH, Bos CG, van den Hoogen MP, De Gooijer CD, Tramper J (1994) Co-immobilized Nitrosomonas europea and Nitrobacter agilis cells: Validation of dynamic Model for simultaneous substrate conversion and growth in K-carrageenan gel beads. Biotechnol. Bioeng. 43:1153–116. Lee NM, Welander T (1994) Influence of predators on nitrification in aerobic biofilm processes. Wat. Sci. Tech. 29:355–363. Leenen EJTM, Boogert AA, van Lammeren AAM, Tramper J, Wijffels RH (1997) Dynamics of artificially immobilized Nitrosomonas europaea: Effect of biomass decay. Biotechnol. Bioeng. 55:630–641. Lesouef A, Payraudeau M, Rogalla F, Kleiber B (1992) Optimizing nitrogen removal reactor configurations by on-site calibration of the IAWPRC activated sludge model. Wat. Sci. Tech. 25:105–123. Martin-Cereceda M, Serrano S, Guinea A (1996) A comparative study of ciliated protozoa communities in activated-sludge plants. FEMS Microbiol. Ecol. 21:267–76. Meijer SCF, van Loosdrecht MCM, Heijnen JJ (2001) Metabolic modelling of full-scale biological nitrogen and phosphorus removing WWTPs. Water Res. 35:2711–2723. Madoni P, Davoli D, Chierici E (1993) Comparative-analysis of the activated-sludge microfauna in several sewage-treatment works. Water Res. 27:1485–91.

Modelling nitrification, heterotrophic growth

115

Novitsky JA, Dalhousie U (1990) Protozoa Abundance, Growth, and Bacteriovory in the Water Column, on Sedimenting Particles, and in the Sediment of Halifax Harbor. Geographical Rev. 36:859–863. Nowak O, Svardal K, Schweighofer P (1995) The dynamic behaviour of nitrifying activated sludge systems influenced by inhibiting wastewater compounds. Wat. Sci. Tech. 31:115–124. Pollice A, Tandoi V, Lestingi C (2002) Influence of aeration and sludge retention time on ammonia oxidation to nitrite and nitrate. Water Res. 36:2541–2546. Ratsak CH, Maarsen KA, Kooijman SALM (1996) Effects of protozoa on carbon mineralization in activated sludge. Water Res. 30:1–12. Reichert P, Ruchti J, Simon W (1994) Aquasim 2.0. Swiss Federal institute For Environmental Science and Technology (EAWAG), Duebendorf, Switzerland. Rittman BE, Regan JM, Stahl DA (1994) Nitrification as source of soluble organic substrate in biological treatment. Wat. Sci. Tech. 30:1–8. Salvadó H, Gracia MP, Amigó JM (1995) Capability of ciliated protozoa as indicators of effluent quality in activated-sludge plants. Water Res. 29:1041–50. Schlimme G, Marchiani M, Hanselmann K, Jenni B (1997) Gen transfer between bacteria within digestive vacuoles of protozoa. FEMS Microbiol. Ecol. 23:239–247. Siegrist H, Brunner I, Koch G, Leinh Con Phan, Van Chieu Le (1999) Reduction of biomass decay rate under anoxic and anaerobic conditions. Wat. Sci. Tech. 39:129–137. Smolders GJF, van loosdrecht MCM, Heijnen JJ (1994) Stoichiometric model of the aerobic metabolism of the biological phosphorus removal process. Biotechnol. Bioeng. 44:837–848. Tappe W, Laverman A, Bohland M, Braster M, Ritteshaus S, Groeneweg J, van Verseveld HW (1999) Maintenance Energy Demand and Starvation Recovery Dynamics of Nitrosomonas europaea and Nitrobacter winogradskyi Cultivated in a Retentostat with Complete Biomass Retention. Appl. Environ. Microbiol. 65:2471–2477. Tijhuis L., van Loosdrecht MCM, Heijnen, JJ (1994) Solid retention time in spherical biofilms in a biofilm airlift suspended reactor. Biotechnol. Bioeng. 44:867–879. van Dongen LGJM, Jetten MSM, van Loosdrecht MCM (2001) The Combined Sharon/Anammox Process, A sustainable method for N-removal from sludge water. IWA Publishing, London: 29– 32. van Loosdrecht MCM, Henze M (1999) Maintenance, endogenous respiration, lysis, decay and predation. Wat. Sci. Tech. 39:107–117. Weinbauer MG, Beckmann C, Höfle MG (1998) Utility of green fluorescent nucleic acid dyes and aluminum oxide membrane filters for rapid epifuorescence enumeration of soil and sediment bacteria. Appl. Environ. Microbiol. 64:5000–3. Ziglio G, Andreottola G, Barbesti S, Boschetti G, Bruni L, Foladoria P, Villa R (2002) Assessment of activated sludge viability with flow cytometry. Water Res. 36:460–468.

Nitrification in saline industrial wastewater

116

APPENDIX 5.1 Process Kinetics and stoichiometry for enriched culture of nitrifiers in SBR system

Modelling nitrification, heterotrophic growth

117

Appendix 5.2 Stoichiometric and kinetic parameters at 30°C used to simulate the nitrifying SBR sysstem. Symbol

Expression

Unit

Definition

Reference

SO2

−

mg O2/L

Concentration of O2

SN2

−

mg N/L

Concentration of N2

SNH4

–

mg N/L

Concentration of NH4

SNO2

–

mg N/L

Concentration of NO2

SNO3

–

mg N/L

Concentration of NO3

SS

–

mg COD/L

Concentration of organic substrate

X

–

mg COD/L

Concentration of ammonia oxidisers

XNO2

–

mg COD/L

Concentration of nitrite oxidisers

XH

–

mg COD/L

Concentration of heterotrophic biomass

Xpredators

–

mg COD/L

Concentration of predators

XI

–

mg COD/L

Concentration of participate inerts

YNH4

0.18

g COD/gNO2–N

Yield of ammonia oxidisers Hunik et al (1994) per NO2−_N

YNO2

0.06

g COD/g NO3–N Yield of nitrite oxidisers per NO3−_N

Hunik et al (1994)

YH

0.63

g COD/g COD

Yield of hetrotrophic biomass on SS

Henze et al (2000) ASM2d

Ypred

0.5

g COD/g COD

Yield of predators on bacteria

Curds (1971, 1973)

Fxi

0.15

g COD/g COD

Fraction of inert COD generated in biomass lysis

This study

iN,XI

0.02

g N/g COD

nitrogen content of XI

Henze et al (2000) ASM2d

iN,BM

0.07

g N/g COD

nitrogen content of biomass Henze et al (2000) ASM2d

1.4

day−1

maximum growth rate of ammonia oxidisers

NH4

Hunik et al (1994)

Nitrification in saline industrial wastewater

118

1&3

mg O2/L

affinity constant for oxygen This study (for SRT of ammonia oxidisers 30 &100 days)

5

mg NH4−N/L

affinity constant for ammonia of ammonia oxidisers

This study

mNH4

0.35

mg NH4−N/g XA–COD.day

maintenance coefficient of ammonia oxidisers

Tappe et al (1999)

bNH4

0.3

day−1

aerobic decay rate of ammonia oxidisers

Measured according to Lesouef et al (1992)

ηNH4

0.5

–

anoxic reduction factor for ammonia oxidisers decay

Siegrist et al (1999)

0.9

day−1

maximum growth rate of nitrite oxidisers

(Hunik et al (1994)

1

mg O2/L

affinity constant for oxygen This study of nitrite oxidisers

2

mg NO2−N/L

affinity constant for nitrite of nitrite oxidisers

This study

mNO2

1.15

mg NO2−N/g XN–COD.day

maintenance coefficient of nitrite oxidisers

Tappe et al (1999)

bNO2

0.2

day−1

aerobic decay rate of nitrite Measured according oxidisers to Lesouef et al (1992)

ηNO2

0.5

–

anoxic reduction factor for nitrite oxidisers decay

Siegrist et al (1999)

12

day−1

maximum growth rate of heterotrophic biomass

Henze et al (2000) ASM2d

0.2

mg O2/L

affinity constant for oxygen Henze et al (2000) of heterotrophic biomass ASM3

2

mg COD/L

affinity constant for organic Henze et al (2000) carbon of heterotrophic ASM3 biomass

0.5

mg NO3−N/L

affinity constant for NO3 of Henze et al (2000) heterotrophic biomass ASM3

mH

0.12

mg COD/g XH– COD.day

maintenance coefficient of heterotrophic biomass

Meijer et al (2001)

ηH

0.8

–

anoxic reduction factor of heterotrophic growth

Henze et al (2000) ASM2d

bH

0.8

day−1

aerobic decay rate of heterotrophic biomass

Henze et al (2000) ASM2d

µpredators

0.2

day−1

predation rate

This study

Modelling nitrification, heterotrophic growth

119

0.2

mg O2/L

affinity constant for oxygen This study of predators

bpredators

0.15

day−1

decay rate of predators

This study

MNH4

SNH4/(0.01+SNH4) –

Monod term for ammonia in bacterial growth

Henze et al (2000) ASM3

Chapter 6 Nitrification activities in full-scale treatment plants with varying salt loads. Submitted as: Moussa MS, Garcia Fuentes O, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ. Nitrification activities in full-scale treatment plants with varying salt loads.

Abstract The effect of salt on the nitrification activity in full-scale wastewater treatment plants (WWTP) was investigated. Not only the activity of ammonia and nitrite oxidisers was measured, but also the nitrifying population was assessed (by Fluorescent In Situ Hybridisation)—in fullscale domestic and industrial WWTPs, operated under various salt levels. The results demonstrate a decline in the activity of ammonia and nitrite oxidisers with an increase in salt content: the domestic WWTP with the lowest salt level (0.13 gCl−/L) had the highest specific activity of ammonia and nitrite oxidisers (4.3 and 2.4 mg-N/gVSS.h, respectively), while the lowest specific activities of ammonia and nitrite oxidisers (1.1 and 0.5 mg-N/gVSS.h) were measured in WWTP with highest NaCl concentration (16 gCl−/L). However, comparing the nitrification activity of different types of sludge developed under different operational conditions with the reported values was not directly possible. The activated sludge model (ASM) was used to translate the routine operational data into parameters to enable the calculation of the actual fraction of nitrifiers and consequently the actual specific activity of ammonia and nitrite oxidisers. Expressing the activity of ammonia oxidisers in terms of actual specific activity makes the results from pure cultures, enriched cultures, pilot scale and full scale WWTPs comparable. Moreover, this will confirm the behaviour of nitrifiers under salt stress and validate the results obtained from pure and enriched cultures to be extrapolated to full scale.

6 Nitrification activities in full-scale treatment plants with varying salt loads. 6.1 Introduction Ammonia is the predominant nitrogen compound in wastewater and is removed in wastewater treatment plants (WWTPs) by conversion to gaseous nitrogen via nitrification and denitrification. With the dramatic increase in nitrogenous wastes due to the expansion of animal husbandry, food processing, nitrogen-producing industries, and other human activities, the handling of nitrogenous wastes has become a critical factor in environmental management. The removal of nitrogen from wastewater is of extreme environmental importance, because the release of untreated wastewater can result in devastating eutrophication of the environment. Nitrification, the first step of nitrogen removal, is the biological oxidation of ammonia to nitrate via nitrite by two groups of chemolithotrophic bacteria, ammonia oxidisers and nitrite oxidisers; both groups have low specific growth rates (Bock et al 1991; Prosser 1986). Nitrifying bacteria are crucial in the microbial communities of nitrifying wastewater treatment systems. Once nitrifiers have been washed out of a WWTP, recovery of the nitrification process can take a long time. This is not only due to their slow growth rates but also to their sensitivity to environmental factors such as temperature, dissolved oxygen concentration, pH, available substrate, product inhibition and inhibitory compounds (Antoniou et al 1990; Hellinga et al 1998; Sharma and Ahlert 1977). Achieving a stable, reliable and cost-effective nitrification process within a WWTP is of major importance (Kowalchuk and Stephen 2001; Rittmann et al 1999; Wagner et al 1998; Wagner and Loy 2002). Therefore, the nitrification process has been the main focus of most of the microbiological studies carried out so far on nitrogen removal (Dabert et al 2002). However, most of these studies are concentrated on domestic wastewater treatment and reports of nitrification from industrial full-scale WWTP are very scarce (Wiesman 1994). Extreme conditions— e.g. extreme pH conditions, presence of toxicant compounds, salinity—typically prevailing in industrial WWTPs were not given much attention. Although the understanding of the microbiology in pure and enriched cultures of nitrifiers and in activated sludge systems for treating domestic wastewater has increased significantly, the results obtained are not directly applicable to industrial wastewater. There has been little cross-linking between the findings from laboratory-and pilot-scale studies on the microbiology and biochemistry of nitrifiers and their presence in full scale WWTPs, especially within the industrial sector. This paper describes the activities and population structure of ammonia and nitrite oxidisers in four full-scale WWTPs (1 domestic and 3 industrial) operated under different salt levels. To our own measurements, combined with the collected routine operational data, we have added literature results to demonstrate a relationship between the salt level

Nitrification activities in full-scale treatment

125

and the nitrification activity, not only in full scale WWTPs, but also in pure and enriched cultures. 6.2 Materials and methods The aim of this research was to measure the nitrification activity using activated sludge from domestic and industrial wastewater treatment plants operated under various salt levels. The study covers one domestic and three industrial wastewater treatment plants, all are in the Netherlands. Hoek van Holland is the domestic WWTP (freshwater-no salt), while Heiploeg (shrimp processing), Ecco (tannery) and Seafarm (production and processing of sea-fish) represent industries with saline wastewater. Within a period of 3 months, each plant was visited at least two times for measurements and sludge characteristics. Measurements included: reactor volume, wastewater flow rate, mixed liquor suspended solids (MLSS), mixed liquor volatile suspended solids (MLVSS), influent and effluent COD, TKN, pH, temperature and salt concentration. Sludge characteristics include the assessment of the activity of ammonia and nitrite oxidisers and the population structure. 6.2.1 Activity of ammonia and nitrite oxidisers Sludge samples (1 L) were directly transferred to a double-jacketed batch reactor (BR) in the UNESCO-IHE laboratory (Delft, The Nehterlands) and were kept aerated over the night to be sure that all ammonia and nitrite were consumed. The BR had a maximum operating volume of 1 L (enough for interval sampling). Ammonia and nitrite removal were measured over several hours, allowing quantification of the kinetic parameters of nitrification. The batch experiments were performed under standardised conditions (pH 7.5±0.05; T 20±1°C). The pH was maintained at 7.5±0.05 automatically by dosing of 0.1 N HCl or 0.1 N NaOH. NaNO2 and NH4Cl were consecutively injected to estimate the nitrite and ammonia uptake rate. For preconditioning of the sludge similar injection steps were performed ahead of the real measurements. A simple double Monod mathematical model was used to describe the two-step nitrification process. Maximum growth rate, yield coefficients and biomass concentrations were lumped into overall parameters, which represent the volumetric oxidation rate of ammonia and nitrite as described by Moussa et al (2003). The activity was expressed as specific nitrogen consumption rate [mg N (g VSS)−1 h−1] or expressed as a percentage of the activity obtained under reference conditions (pH 7.5, T 20°C). 6.2.2 Oligonucleotide probes and fluorescent in situ hybridisation (FISH) To identify the population of nitrifying bacteria in the investigated WWTP a set of rRNA targeted oligonucleotide probes for Fluorescence In Situ Hybridisation (FISH) was used. Samples were taken from each plant and immediately fixed with paraformaldehyde. In situ characterization of microbial populations followed a top to bottom approach. First the samples were hybridised with a probe set (EUB338, EUB338-II, EUB338-III) designed to target almost all bacteria (Daims et al 1999). Then tsets of specific probes

Nitrification in saline industrial wastewater

126

were used: (ALF968 and BET42a) were used for the alpha and beta subclasses of Proteobacteria, respectively (Manz et al 1992 1996). The ammonia-oxidising and nitriteoxidising bacteria were identified using previously published probes as described by Nogueira et al (2002). Oligonucleotide probes were purchased as derivatives labelled with the fluorescent dyes Cy3, Cy5, and 5(6)-carboxyfluorescein-N-hydroxysuccinimideester (FLUOS), respectively (Interactiva, Ulm, Germany). FISH was performed using the hybridisation and washing buffers as described by Manz et al (1992). The hybridised samples were analysed with a Zeiss Axioplan2 Imaging microscope. 6.2.3 Analysis Ammonia, nitrite, total kjeldahl nitrogen (TKN), chemical oxygen demand COD, Chloride (Cl−), sulphate (SO4−2), pH and conductivity were measured in accordance with Standard Methods (APHA, 1998). Nitrate-N was determined using Dionex 4500i series and Shimadzu C_R5A ion-chromatograph. The MLSS and MLVSS determination was performed after filtration of a 10 mL sample of mixed liquor on a Whatman glass micro fibre filter (GC/F) filter. Dry weight was determined after the filter was dried for 24 h at 105°C and weighted with a microbalance. The ash content was calculated after incinerating the dried filter in an oven for 1 h at 550°C. 6.3 Results The different treatment plants represent a wide range of flow (100–21000 m3/day), influent COD (300–2175 mg-COD/L), influent TKN (20–557 mg-N/L) and salt level (0.13–16 gCl−/L) (Table 6.1). All treatment plants perform nitrification as they have high ammonia removal efficiency (>95%). However, a significant difference among treatment plants in activity of ammonia and nitrite oxidisers was observed.

Table 6.1 Operating data and measured influent and effluent parameters of the 4 full-scale WWTP. Plant Name Parameter

Unit

Type

System Volume Flow Temperature

Hoek van Holland

Heiploeg

Ecco

Seafarm

Domestic

fish processing Industry

tannery Industry

marine aquaculture Industry

AS

SBR

AS

BF

3

(m )

5700

2800

8000

3

21000

440

843

(°C)

14

14–18

22

(m day−1)

100

Nitrification activities in full-scale treatment

127

COD

(mg/L)

420

2050

2175

301.0

Influent TKN

(mg/L)

40

557

519

20.0

7

7.6

7.6

7.3

pH COD

(mg/L)

40

74

117

Effluent TKN

(mg/L)

1.8

12

24

7

7.5

7.6

7.3

pH Reactor

MLSS

(mg/L)

4445

4700

8300

9960

MLVSS

(mg/L)

3200

3390

5870

3540

(ms/cm)

1

10

24

44

(mg/L)

130

3000

8000

16000

(mg/L)

47

68

4000

Conductivity Cl

− 2−

SO4

6.3.1 Activity of ammonia and nitrite oxidisers The results demonstrate a decline in the activity of ammonia and nitrite oxidisers with an increase in salt content (Figure 6.1). The domestic WWTP with the lowest salt level (fresh water, 130 mgCl−/L) had the highest specific activities of ammonia and nitrite oxidisers (4.3 and 2.4 mg-N/gVSS.h, respectively). The lowest specific activities of ammonia and nitrite oxidisers (1.1 and 0.5 mg-N/gVSS.h) were measured at the highest NaCl concentration in the sludge from Seafarm. Under all experimental conditions the specific activity of ammonia oxidisers was always two times higher than that of the nitrite oxidisers.

Figure 6.1 Activity of ammonia and nitrite oxidisers measured at different salt concentrations for sludge collected from 4 full-scale WWTP. The

Nitrification in saline industrial wastewater

128

activities were measured under standard conditions: pH 7.5 and T 20°C. 6.3.2 Population diversity of ammonia and nitrite oxidisers Four species of ammonia oxidisers (Nitrosomonas oligotropha, Nitrosomonas europaea, Nitrosococcus mobilis and Nitrosospira sp.) were detected within the collected samples of the 3 industial WWTPs (Table 6.2). No nitrite oxidisers were detected with the two available oligonucleotide probes (NIT3 and Ntspa662).

Table 6.2 Population structure of nitrifying bacteria in industrial full scale WWTPs operated at different salt levels. WWTP

Heiploeg

Ecco

Seafarm

Salt conc. (gCl /L)

3

8

16

Conductivity (ms/cm)

10

24

44

+++

−

+++

Nitrosomonas oligotropha

−

+++

+

Nitrosococcus mobilis

+

−

−

Nitrosospira sp.

+

+

−

Nitrospira sp.

−

−

−

Nitrobacter sp.

−

−

−

−

Ammonia oxidisers: Nitrosomonas europaea

Nitrite oxidisers:

6.4 Discussion 6.4.1 Activity of ammonia and nitrite oxidisers The results showed the adverse effect of salt on the nitrification process in full scale WWTPs. However, the results demonstrated the complexity of comparing nitrification activity of different types of sludge developed under different operational conditions. The operational conditions of industrial WWTPs vary widely and reports on the nitrification performance of their wastwater are very scarce (Wiesmann 1994). Moreover, only few studies distinguish between ammonia and nitrite oxidisers (Hunik et al 1992, 1993; Moussa et al 2003 a, b, c). The decrease in specific activity of ammonia oxidisers with increase of the salt content is in line with the reported results (Dahl et al 1997;

Nitrification activities in full-scale treatment

129

Vredenbregt et al 1997; Panswad and Anan 1999 a, b). However, there is much deviation under domestic wastewater conditions (Pollice et al 2002, Salem et al 2003) (Figure.2).

Figure 6.2 Comparison of specific ammonia oxidisers activities (mgN/gVSS.h) measured in the 4 full scale WWTP (taken from figure 1) with the results reported in literature. The results were all calculated to temperature 20 °C, with the formula R20=RT. exp(0.094(T−20)). The specific activity of pure or enriched cultures of nitrifiers under salt-free conditions is 10–15 higher as compared to full scale WWTPs (Table 6.3). This could be explained by the difference in active fraction of ammonia and nitrite oxidisers present in the investigated sludge. The sludge is comprised of inert particulates, biomass of heterotrophs (produced through COD conversion) and nitrifiers (produced through ammonia oxidation), grazers (protozoa and metazoa) and inert particulates produced through the decay of heterotrophs, nitrifiers and grazers. The ratio of each fraction to the total biomass is a function of the operating conditions, such as, solid retention time (SRT), hydraulic retention time (HRT), COD removal and TKN removal.

Nitrification in saline industrial wastewater

130

Table 6.3 Reported values of ammonia (AOB) and nitrite (NOB) oxidisers activities of different type of biomass developed in different system and different operational conditions (Temperature, salt level, COD/N ratio, SRT). Specific rate AOB

System*1 Biomass*2 CODin NH4in SRT Reference

Temp.

Salt

°C

gCl−/L

NOB

mg mg NH4– NO2– N/gvss.h N/gvss.h

mg/L

50

days

3.63

22

0.1 FA

DA

22.9

32

0.1 PA

DA

10

32

0.1

5

4.8

32

0.1

14

3.2

32

0.1

24

4.76

28

0.1 LA

3.29

28

2.48

8 Salem et al (2003) 3 Pollice et al (2002)

500

25

10

5

500

25

10

28

10

500

25

10 Panswad and Anan (1999)

2.43

28

20

500

25

10

2.14

28

30

500

25

10

3.5

30

20

SA

42.8

200

2.35

30

20

SA

395.9

35

3

25

10 PF

840

120

2

30

20

840

120

220

20

660

FB

DA

280

mgN/L

PCA

Vredenbregt et al (1997) Dahl et al (1997) Copp and Murphy (1995)

20

PCN

160

20

MC

444

30

LC

PCA

0.3 Hunik et al (1992)

30

LC

PCN

0.3 Hunik et al

1412

Nitrification activities in full-scale treatment

131

(1993) 96

58.5

30

0.1 LSBR

EC

10

250

30 Moussa et al (2003c)

58.5

55.7

30

10

EC

10

250

30

17.7

19.6

30

20

EC

10

250

30

14.7

12.2

30

30

EC

10

250

30

3

1.9

30

40

EC

10

250

30

*1

FA=Full-scale Activated sludge WWTP; PA=Pilot-scale Activated sludge unit; LA=Lab-scale Activated sludge unit; PF=pilot-scale Fluid-bed; FB=lab-scale Fed-Batch mode; LC= sterile Lab-scale Chemostat; LSBR=Lab-scale Sequencing batch reactor; PA=Pilot-scale Activated sludge unit *2 DA=Domestic Activated sludge performing nitrification; SA=Salt Adapted activated sludge performing nitrification; PCA=Pure Culture of Ammonia oxidisers Nitrosomona europea; PCN =Pure Culture of Nitrite oxidisers Nitrobacter agilis; MS=Marine Sediment; MC=Nitrosomonas and Nitrobacter in Mixed Culture; EC=Enriched Culture of nitrifying bacteria

The activated sludge model (ASM1) was used to translate the routine operating data into parameters to enable the calculation of the fraction of nitrifiers and consequently to recalculate the actual specific activity of ammonia and nitrite oxidisers (equation 6.1, 6.1). (6.1) (6.2) The ASM1 model was applied to recalculate the actual activity of ammonia oxidisers both in this study and in other reported studies. Ammonia oxidisers: The assessment of the fraction of ammonia oxidisers demonstrates the influence of the above-mentioned operating parameters on the portion of the nitrifiers to the total biomass. This is in agreement with the results reported by Rittmann et al (1999) and Wiesmann (1994). Expressing the activity of ammonia oxidisers in terms of specific activity eliminates the deviation between the specific activity from different systems and makes the results from pure cultures, enriched cultures, pilot scale and full scale WWTPs comparable (Figure 6.3). Moreover, the results of activities of pure and enriched cultures can now be extrapolated to full scale WWTPs. The activity seems to be dependent on the salt concentration, irrespective of the dominant species of ammonium oxidisers, which was also demonstrated by Moussa et al (2003c).

Nitrification in saline industrial wastewater

132

Figure 6.3 Specific activity of ammonia oxidisers based on active fraction of the sludge (mg-N/gVSS ammonia oxidisers.h) as a function of salt concentrations. The results of this study are compared with reported data, all recalculated for 20 °C. Nitrite oxidisers: Similar recalculation of the activity of nitrite oxidisers in terms of specific activity agrees with results from pure cultures (Copp and Murphy 1995; Hunik et al 1993) and enriched cultures (Moussa et al 2003c). 6.4.2 Population of nitrifiers Ammonia oxidisers: Both Nitrosomonas oligotropha and Nitrosomonas europaea are normally present together under low salt levels, not only in full scale WWTPs (Daims et al 2001b; Gieseke et al 2001; Juretschko et al 1998; Liebig et al 2001; Purkhold et al 2000), but also in laboratory-scale systems (Nogueira et al 2002; Moussa et al 2003c). At elevated salt levels in the industrial WWTPs only one of the Nitrosomonas species, N. europaea, became dominant, at 3 and 16 gCl−/L (Heiploeg and Seafarm). This is in line with Hovanec and De Longe (1996) and Catalan-Sakairi et al (1996, 1997). Also in enriched cultures with several ammonia oxidising species present at low salt concentrations, N. europaea became dominant when salt was increased above 10 gCl−/L (Moussa et al 2003c). Surprisingly, N. oligotropha was dominant at 8 gCl−/L (Ecco), which could be attributed to the high sulphate concentration (4g SO4−2/L) in this WWTP. It is not surprising that N.europaea becomes dominant under elevated salt levels, because it is halotolerant or moderately halophilic (Koops et al 2001; Wagner et al 1995).

Nitrification activities in full-scale treatment

133

Nitrite oxidisers: Despite the fact that nitrite oxidisers were active (oxidation of nitrite to nitrate), neither of the two probes (NIT3 and Ntspa662) were able to detect Nitrobacter sp. or Nitrospira sp. The reason could be that the fraction of these nitrite oxidisers was below the detection limit, since these WWTPs are operated at high sludge age and/or high organic load. Alternatively, nitrite oxidisers might be present, but are not detectable with the two available probes. A number of studies have demonstrated a relationship between the presence of distinct nitrifiers and specific environmental conditions (Kowalchuk and Stephen 2001; McCagig et al 1999; Rittmann et al 1999; Stephen et al 1999). However, a relation between salt and population of nitrifiers cannot be drawn, since within the limited number of WWTPs with elevated salt levels, two different ammonia-oxidising species became dominant. Nevertheless, the use of the model to quantify the specific activity and to correlate this to the presence of specific populations within different systems was promising and could be applied in future to confirm this relation. 6.5 Conclusions • The results showed the adverse effect of salt on nitrification in full scale WWTPs. The domestic WWTP with the lowest salt level (0.13 g Cl−/L) had the highest specific activity of ammonia and nitrite oxidisers (4.3 and 2.4 mg-N/gVSS.h, respectively). The lowest specific activities of ammonia and nitrite oxidisers (1.1 and 0.5 mg-N/g VSS.h) were measured in the WWTP with the highest NaCl concentration (16 g Cl−/L). • It is rather complex to compare between the specific nitrification activity of different types of sludges developed under different operational conditions. This is due to the variation in active fraction of ammonia and nitrite oxidisers present in the investigated types of sludge. • The use of the activated sludge model to calculate the active fraction of nitrifiers and consequently recalculate the actual specific activity of ammonia and nitrite oxidisers makes the results from pure cultures, enriched cultures, pilot scale and full scale WWTPs comparable. • Model application to quantify the actual specific activity validates the previous results of salt on nitrification obtained at laboratory-scale. The ammonia and nitrite oxidisers activity seems to be dependent on the salt concentration, irrespective of the dominant species of ammonium oxidisers. • Both Nitrosomonas oligotropha and Nitrosomonas europaea are normally present under low salt levels, not only in full scale WWTPs but also in laboratory-scale systems. At elevated salt levels only Nitrosomonas europaea became dominant. • A clear identification of the nitrite oxidisers present in the investigated WWTPs was not possible. The reason could be that either the fraction of these nitrite oxidisers was below the detection limit or they were present, but are not detectable with the two available probes.

Nitrification in saline industrial wastewater

134

References Antoniou P, Hamilton J, Koopman B, Jain R, Holloway B, Lyberatos G, Svoronos SA (1990) Effect of temperature and pH on the effective maximum specific growth rate of nitrifying bacteria. Water Res. 24:97–101. APHA (1998) Standard methods for the examination of water and wastewater, 20th ed. American Public Health Association/American Water Works Association/Water Environment Federation, Washington D.C. Bock E, Koops HP, Harms H, Ahlers B (1991) The biochemistry of nitryfying organisms. In: Shively JM, Barton LL Variation in autotrophic life. Academic Press, London. Catalan-Sakairi MAB, Wang PC, Matsumura M. (1997) Nitrification performance of marine nitrifiers immobilized in polyester and macro-porous cellulose carriers. Fermentation and Bioeng. 84:563–571. Catalan-Sakairi MAB, Yasuda K, Matsumura M. (1996) Nitrogen removal in seawater using nitrifying and denitrifying bacteria immobilized in porous cellulose carrier. Water Sci. Technol. 34:267–274. Copp JB, Murphy KL (1995) Estimation of the active nitrifying biomass in activated sludge. Water Res. 29:1855–1862. Dabert P, Delgenes J-P, Moletta R, Godon J-J (2002) Contribution of molecular microbiology to the study in water pollution removal of microbial community dynamic. Re/View in Environmental Science and bio/Technology 1:39–49. Dahl C, Sund C, Kristensen GH, Vredenbregt L (1997) Combined biological nitrification and denitrification of high-salinity wastewater. Water Sci. Technol. 36:345–52. Daims H, Brühl A, Amann R, Schleifer K-H, Wagner M (1999) The domain-specific probe EUB338 is insufficient for the detection of all bacteria: development and evaluation of a more comprehensive probe set. Syst. Appl. Microbiol. 22:434–44. Daims H, Purkhold U, Bjerrum L, Arnold E, Wilderer PA, Wagner M (2001b) Nitrification in sequencing biofilm batch reactors: lessons from molecular approaches. Water Sci. Technol. 43:9–18. Gieseke A, Purkhold U, Wagner M, Amann R, Schramm A (2001) Community structure and activity dynamics of nitrifying bacteria in a phosphate-removing biofilm. Appl. Environ. Microbiol. 67:1351–1362. Hellinga C, Schellen AAJC, Mulder JW, van Loosdrecht MCM, Heijnen JJ (1998) The SHARON process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water. Sci. Technol. 37:135–142. Hovanec TA, De Longe E (1996) Comparative analysis of Nitrifying Bacteria associated with freshwater and marine aquaria. Appl. Environ. Microbiol. 62:2888–2896. Hunik JH, Meijer HJG, Tramper J (1992) Kinetics of Nitrosomonas europaea at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 37:802–807. Hunik JH, Meijer HJG, Tramper J (1993) Kinetics of Nitrobacter agilis at extreme substrate, product and salt concentrations. Appl. Microbiol. Biotechnol. 40:442–8. Juretschko S, Timmermann G, Schmid M, Schleifer K-H, Pommerening-Röser A, Koops H-P, Wagner M (1998) Combined molecular and concentional analyses of nitrifying bacteriumin activated sludge: Nitrosococcus mobilis and Nitrospira-like bacteria as dominant populations. Appl. Environ. Microbiol. 64:3042–51. Koops H-P, Pommerening-Röser A (2001) Distribution and ecophysiology of the nitrifying bacteria emphasizing cultured species. FEMS Microbiol. Ecology 37:1–9. Kowalchuk GA, Stephen JR (2001) Ammonia-oxidising bacteria: a model for molecular microbial ecology. Annu. Rev. Microbiol. 55:485–529.

Nitrification activities in full-scale treatment

135

Liebig T, Wagner M, Bjerrum L, Denecke M (2001) Nitrification performance and nitrifier community composition of a chemostat and a membrane-assisted bioreactor for the nitrification of sludge reject waters. Bioprocess Biosyst. Eng. 24:203–210. Manz W, Amann R, Ludwig W, Wagner M, Schleifer KH (1992) Phylogenetic oligonucleotide probes for the major subclasses of Proteobacteria: problems and solutions. Syst. Appl. Microbiol. 15:593–600. Manz W, Amann R, Ludwig W, Vancanneyt M, Schleifer KH (1996) Application of a suite of 16S rRNA specific oligonucleotide probes designed to investigate bacteria of the phylum Cytophaga-Flavobacter-Bacteroides in the natural environment. Microbiol. 142:1097–1106. McCaig AE, Phillips CJ, Stephen JR, Kowalchuk GA, Harvey M (1999) Nitrogen cycling and community structure of β-subgroup ammonia oxidising bacteria within polluted, marine fishfarm sediments. Appl. Environ. Microbiol. 65:213–20. Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003a) Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures. Appl. Microbiol. Biotechnol. 63:217–221. Moussa MS, Lubberding HJ, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM (2003b) Short term effects of various salts on ammonia and nitrite oxidisers in Mixed Bacterial Cultures. Appl. Microbiol. Biotechnol. (submitted). Moussa MS, Lubberding HJ, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM (2003c) Long Term Effects of Salt on Activity, Population Structure and Floc Characteristics in Mixed Bacterial Cultures of nitrifiers. Appl. Environ. Microbiol. (submitted). Nogueira R, Melo LF, Purkhold U, Wuertz S, Wagner M (2002) Nitrifying and heterotrophic population dynamics in biofilm reactors: effects of hydraulic retention time and the presence of organic carbon. Water Res. 36:469–481. Panswad T, Anan C (1999a) Impact of high chloride wastewater on an anaerobic/anoxic/aerobic process with and without inoculation of chloride acclimated seeds. Water Res. 33:1165–1172. Panswad T, Anan C (1999b) Specific oxygen, ammonia and nitrate uptake rates of a biological nutrient removal process treating elevated salinity wastewater. Bioresource Technol. 70:237– 243. Pollice A, Tandoi V, Lestingi C (2002) Influence of aeration and sludge retention time on ammonia oxidation to nitrite and nitrate. Water Res. 36:2541–2546. Prosser JI (1986) Nitrification, special publication of the society for general microbiology, Oxford IRL Press, Volume 20. Purkhold U, Pommerening-Röser A, Juretschko S, Schmid MC, Koops H, Wagner M (2000) Phylogeny of all recognized species of ammonia-oxidizers based on comparative 16S rRNA and amoA sequence analysis: Implications for molecular diversity survey. Appl. Environ. Microbiol. 66:5368–5382. Rittmann BE, Laspidou C.S, Flax J, Stahl, DA, Urbain V, Harduin, H, van der Waarde JJ, Geurkink B, Henssen MJC, Brouwer H, Klapwijk A, Wetterauw (1999) Molecular and modeling analyses of the structure and function of nitrifying activated sludge. Water Sci. Technol. 39:51–59. Salem S, Berends DHJG, van der Roest HF, van der Kuijl RJ, van Loosdrecht MCM (2003) Fullscale application of the BABE process. Water Sci. Technol. (in press). Sharma B, Ahlert RC (1977) Nitrification and nitrogen removal. Water Res. 11:897–925. Stephen JR, Chang Y-J, Macnaughton SJ, Kowalchuk GA, Leung KT (1999) Effect of toxic metals in indigenous soil β-subgroup proteobacterium ammonia oxidizer community structure and protection against toxicity by inoculated metal-resistant bacteria. Appl. Environ. Microbiol. 65:65–101. Vredenbregt LHJ, Nielsen K, Potma AA, Kristensen GH, Sund C (1997) Fluid bed biological nitrification and denitrification in high salinity wastewater. Water Sci. Technol. 36:93–100. Wagner M, Rath G, Amann R, Koops H-P, Schleifer KH (1995) In situ identification of ammoniaoxidizing bacteria. Syst. Appl. Microbiol. 18:251–64.

Nitrification in saline industrial wastewater

136

Wagner M, Nogueira DR, Juretschko S, Rath G, Koops H-P, Schleifer KH (1998) Combining fluorescent in situ hybridisation (FISH) with cultivation and mathematical modelling to study population structure and function of ammonia oxidising bacteria in activated sludge. Water Sci. Technol. 31:441–449. Wagner M, Loy A (2002) Bacterial community composition and function in sewage treatment system. Environ. Biotechnol. 13:218–227. Wiesmann U (1994) Biological nitrogen removal from wastewater. Advances in biochemical Engineering/biotechnology 51:113–154.

Chapter 7 Model-based evaluation of the upgrading of a full-scale industrial wastewater treatment plant Previously published as: Moussa MS, Rojas AR, Hooijmans CM, Gijzen HJ, van Loosdrecht MCM (2004) Model-based evaluation of nitrogen removal in a tannery wastewater treatment plant. Accepted for the IWA Conference on wastewater treatment for nutrient removal and reuse. Bangkok, Thailand (January 26–29, 2004).

Abstract Computer modelling has been used in the last 15 years as a powerful tool for understanding the behaviour of activated sludge wastewater treatment systems. However, computer models are mainly applied for domestic wastewater treatment plants (WWTP). Application of these types of models to industrial wastewater treatment plants requires a different model structure and an accurate estimation of the kinetics and stoichiometry of the model parameters, which may be different from the ones used for domestic wastewater. Most of these parameters are strongly dependent on the wastewater composition. In this study a modified version of the activated sludge model No. 1 (ASM 1) was used to describe a tannery WWTP. Several biological tests and complementary physicalchemical analyses were performed to characterise the wastewater and sludge composition in the context of activated sludge modelling. The proposed model was calibrated under steady-state conditions and validated under dynamic flow conditions. The model was successfully used to obtain insight in the existing plant performance, possible extension and options for process optimisation. The model illustrated the potential capacity of the plant to achieve full denitrification and to handle a higher hydraulic load. Moreover, the use of a mathematical model as an effective tool in decision-making was demonstrated.

7 Model-based evaluation of the upgrading of a full-scale industrial wastewater treatment plant 7.1 Introduction The activated sludge system is currently the most widely used biological wastewater treatment process, treating both domestic and industrial wastewater. The process requires a high degree of operational control and management. In order to obtain maximum removal efficiency from the activated sludge plant, the operator must have a full understanding of this complex process. Much research has been done over the last 15 years to understand the behaviour of activated sludge systems using computer modelling. A common language for all modellers in this field regarding concepts and nomenclature is provided by the ASM models developed by the IWA task group (Henze et al 2000). The ASM models have proved to be a useful tool for the dynamic simulation of activated sludge systems treating domestic wastewater. However, application of these models to industrial wastewater treatment plants remains limited. To apply these models to industrial wastewater treatment plants it could be necessary to extend the ASM models by additional kinetic reactions (Nowak et al 1995). The use of models especially in the field of industrial wastewater will support plant operators. The physical, chemical and biological properties of industrial wastewater and their variation in flow and composition make the operation more complicated. Moreover, the model could be used as a quantitative way to predict the effect of different production scenarios on their wastewater treatment plant (van Zuylen 1993), supporting the operator in decision making. The tannery industry is one of the industries generating high amounts of polluted water while the industry has a low profit margin. Therefore, purification of the generated wastewater has a high impact on the overall production costs. The work presented here emphasises on modelling of a tannery activated sludge wastewater treatment plant and its practical application. The main objectives of this study are: • To modify activated sludge model ASM1 to satisfactorily describe the COD and N removal in the tannery wastewater treatment plant; • To evaluate the plant performance using the modified model; • To investigate the required modifications for the plant optimisation and future extensions on plant capacity.

Model-based evaluation of the upgrading

141

7.2 Materials and methods 7.2.1 Plant and process description The study covers the wastewater treatment plant of Ecco Tannery Holland B.V., a tannery factory located at Dongen, The Netherlands. The WWTP is in operation since 1987, treating wastewater generated from different steps of the tannery plant and designed for COD and N removal the general lay-out of the WWTP plant is presented in figure 7.1. The configuration of the plant consists of primary and secondary wastewater treatment and sludge treatment. The primary treatment treats the segregated stream containing Crtotal, which is removed by chemical precipitation. The supernatant liquid is pumped to the next step, where it is mixed with the rest of the generated wastewater in a covered equalisation tank (1750 m3). The equalisation tank buffers the dynamic flow generated during the week (5 working days/week) and provides the plant with a minimum flow during the weekend. In the equalisation tank a dose of Fe(OH)3 (iron sludge from a drinking water treatment plant) is mixed with wastewater to remove the S2− compounds. The top gas layer of the equalisation tank is pumped off, washed and used for aeration in the second stage. Eight primary clarifiers (8×50 m3) are used for particulate solid separation at the end of the primary stage.

Figure 7.1 Plant lay-out of the wastewater treatment plant Ecco Tannery B.V. The secondary treatment

Nitrification in saline industrial wastewater

142

is a conventional plug flow activated sludge system with predenitrification The secondary treatment consists of a conventional plug flow activated sludge system, which is the main focus of this study. The system involves a plug flow reactor of a total volume of 8000 m3. The first part of the reactor (1000 m3) is non-aerated to ensure denitrification while the rest of the reactor is aerated. An internal recycle flow (Qint) supplies the denitrification zone with nitrate. To avoid phosphorus limitation for microbial growth, a flow of 25L/day (Qin3) of H3PO4 of 75% concentration is dosed in the aerated zone. The outlet of the reactor is connected to a secondary settler of 800 m3, where the settled sludge (Qreturn) is pumped to the denitrification zone and the excess sludge (Qex) is pumped to the equalisation tank of the primary treatment. Finally, the treated effluent is pumped via a 37 km force main to the water authority gravity line, which conveys the wastewater to the domestic WWTP Rilland Bath. The sludge treatment deals with primary and chromium sludge. These two different types of sludge are produced in the plant. The primary sludge, which is collected from the primary settlers, is conditioned in a buffering tank (200 m3), de-watered by a filter press, resulting in a sludge cake having a dry content of 30–35%. This sludge cake and the sludge generated from the chromium removal (chromium sludge) are transported by trucks for final disposal. The rejected water resulting from the primary sludge treatment is pumped to the bioreactor (Qin3). The reactors volumes, the hydraulic and operational collected data are summarised in Table 7.1.

Table 7.1 Operational flow data and reactor volumes of WWTP Ecco Tannery B.V. The values printed italic are obtained from mass balances and are used in the model. Flow

Average flow rates m3/d

Reactor

Volume m3

Influent, Qin1

710

Equalisation tank

1750

Influent, Qin2

125 (133)

Primary settler (8×50)

400

Influent, Qin3

0.025

Unaerated zone

1000

Effluent, Qeff

755

Aerated zone

7000

Return Sludge, Return

1920 (650)

Total reactor volume

8000

Internal recycle, Qint

5760

Secondary settler

800

Excess sludge, Qex

80 (90)

Sludge buffering tank

200

7.2.2 Measurements The staff of WWTP Ecco Tannery Holland B.V. provided the routinely collected operational data of the bioreactor and its performance over the year 2000. A detailed sampling and experimental program was conducted in October and November 2000. The pseudo steady state measurements of the WWTP were performed during two sampling

Model-based evaluation of the upgrading

143

runs (11 October and 24 November 2000). During each run samples were collected from the influent, return sludge and effluent. The samples were analysed for T (temperature), pH, Alkalinity, DO (dissolved oxygen), CODtot (total COD), CODS (COD of the microfiltrated fraction, 0.45mm pore diameter), NH4–N, NO3–N, TKN (Total Kjeldal Nitrogen), Ptot (total phosphorus), VSS (Volatile Suspended Solids) and TSS (Total Suspended Solids). In addition, six different sampling points over the length of the bioreactor were defined and sampled during the second run (24 November 2000). These six sampling points were used to describe the hydraulic regime and biological conversion as function of the reactor length. The average values of the routinely collected data for the year 2000 and the average measurements of the sampling program are presented in Table 7.2. Several biological batch tests were performed at the UNESCO-IHE laboratory to determine the influent and sludge characteristics (Ekama et al 1986; Orhon et al 1999a, b). The tests were performed at controlled temperature 20°C, pH of 7.5±0.05 and under aerobic and anoxic conditions. Nitrification batch tests were performed under aerobic conditions, in which NaNO2 and NH4 were consequently injected (Moussa et al 2003a). This test allows measuring the kinetic parameters of nitrite and ammonia oxidisers and was used for model calibration.

Table 7.2 Measured influent and effluent and the influent composition required for the model of WWTP Ecco Tannery B.V. The yearly average values are printed in Italic, these values were used to calculate steady-state influent composition. Measurements

Model Influent Composition

Value

Value

Description Influent Effluent

units

Description

Symbol Dynamic Steadystate

units

Total COD, CODTotal

2525 (2920)

167 (190)

gCOD/m3

Soluble compounds

Soluble COD, CODS

1785

143

gCOD/m3

Dissolved oxygen

SO2

0.3

0.3

gCOD/m3

Total N–Kj

488 (515)

7 (11)

gN/m3

Readily biodegradable COD

SS

840

972

gCOD/m3

Soluble N– Kj

454

6

gN/m3

Soluble inert COD

SI

177

205

gCOD/m3

Ammonium, NH4+

438

10

gN/m3

Ammonium

SNH4

438

448

gN/m3

Nitrite, NO2−

0

0 (0.1)

gN/m3

Nitrite

SNO2

0

0

gN/m3

Nitrate NO3−

0

99 (50)

gN/m3

Nitrate

SNO3

0

0

gN/m3

Nitrification in saline industrial wastewater

Total phosphorus, PTotal

0.4

3.3

gP/m3

Ortho phosphate, PO42−

0.1

2.8

gP/m3

Total Chromium, Cr3+

0.23

0.2 (0.1)

gCr/m3

Calcium, Ca2+

200

320 (400)

gCa/m3

Particulate compounds

Chloride, Cl−

8066

7960 (7380)

gCL/m3

Sulfate, SO42−

3700

3900 (3100)

Total Suspended Solids, TSS

1593

Volatile Suspended Solids, VSS Alkalinity Temperature pH Dissolved Oxygen, DO

Alkalinity

144

SAlk

50

50

mole HCO3/l

Inert particulate COD

XI

454

525

gCOD/m3

gSO4/m3

Slowly biodegradable COD

XS

1060

1226

gCOD/m3

500

g/m3

Heterotrophics

XH

0

0

gCOD/m3

255

127

g/m3

Ammonia oxidisers

XNH4

0

0

gCOD/m3

2030

570

gCaCO3/m3 Nitrite oxidisers

XNO2

0

0

gCOD/m3

22

22

°C

9 (9.4)

7.5 (7.1)

0.3

67

gO2/m3

7.2.3 Process model (selection and adjustment) The simulations of the secondary treatment of Ecco Tannery Holland B.V. WWTP were performed with AQUASIM® (Reichert 1998), a computer software package used for simulation. The process was modelled according to the flow scheme in Figure 1, with the hydraulic and operational parameters as presented in Table 1. The plug flow condition of the bioreactor was modelled as six completely mixed stirred compartments in series with an internal recycle flow. In the model the secondary clarifier separates solids and water ideally. The amount of suspended solids discharged in the effluent was considered in the sludge age (SRT) calculation. Oxygen concentrations in each compartment were controlled with a PI-controller in accordance to the measured values. A modified version of the activated sludge model No. 1 (ASM 1) proposed by the IWA task group (Henze et al 2000) was used to calculate the biological conversions in

Model-based evaluation of the upgrading

145

each compartment (Appendix 7.1 and 7.2). The main modification incorporated in ASM1 to simulate the WWTP was that the nitrification was considered as a two-step process (Nowak et al 1995). The reason for this is that ammonia oxidisers grow faster than nitrite oxidisers at temperatures above 15–20 °C (Hellinga et al 1999). Moreover, inhibiting compounds present in industrial wastewater might lead to an adverse effect on one or both steps of the nitrification process. Thus describing the nitrification in two steps enables the identification of any inhibition and the detection of partial nitrification (NO2 accumulation). 7.2.4 Influent measurement and characterisation For the use of the model to simulate the WWTP a detailed wastewater characterisation to determine the model components is required. Laboratory tests involving biodegradation were conducted to determine the influent characteristics. The total influent COD can be described as: CODtotal=SS+SI+XS+XI The readily biodegradable COD (SS) was determined by two different approaches using an aerobic and anoxic batch test as described by Ekama et al (1986). The soluble inert COD (SI) was determined according to the approach suggested by Orhon et al (1999a, b). The Influent XI fraction (XI/CODtotal) ratio was estimated as a result of model calibration fitting the solid COD balance as proposed by Meijer et al (2001) and consequently the rest will represent the XS fraction. Average influent measurements and the calculated model influent compositions are presented in Table 7.2. 7.3 Balancing operational data and measurements 7.3.1 Estimation of sludge age, Q recycling and Qin2 For a reliable simulation study the sludge age (SRT) should be known within 95% accuracy (Brdjanovic et al 2000; Meijer et al 2001). Therefore a check on the SRT (or sludge production) is strongly recommended. For the evaluation of sludge production the overall phosphorus balance was used as proposed by Nowak et al (1999). Three balances were formulated (see Figure 7.2): the overall P balance (equation 1), the overall flow balance (equation 2) and the P balance over the settler (equation 3). Qin1Pin1+Qin2Pin2+Qin3Pin3=QeffPeff+QexPex (1) Qin1+Qin2+Qin3=Qeff+Qex (2) (Qin1+Qin2+Qin3)Pr=QeffPeff+(Qreturn+Qex)Pex (3)

Nitrification in saline industrial wastewater

146

Figure 7.2 Phosphorous and overflow balance over the bioreactor and settler. The mass balances are presented in equation (1)–(3). Concentration and flow are used from table 1, 2 and 3. indicate the sampling points. The yearly average measurements of Qin1, Qin3 and Qef in addition to the two runs average measurements of Pin1, Pin2, Pin3 (chemical P addition) and Pex were used to evaluate the Qin2, Qreturn and Qex and consequently the SRT. The calculated Qin2, Qex and SRT (70 days) were in agreement with the recorded value of the plant. However, the balanced Qreturn (650 m3/d) was found to be inconsistent with the reported value (1920 m3/d). The reported value is clearly inconsistent with the measured sludge concentrations in the reactor and return sludge, indicating that the reported value is wrong. Therefore we used the balanced value in simulating the treatment plant. 7.4 Model calibration and simulation After the determination of the main operational parameters and the influent characterisation, the model of the WWTP was calibrated. A step-wise approach was applied as proposed by Meijer et al (2001). First the solids were fitted (P, COD and TKN) on the basis of yearly average measurements. Next the nitrification and denitrification were calibrated on the basis of yearly average measurements. 7.4.1 Calibration of the solids The solids balance is a non-conserved balance. An incorrectly assumed COD load or sludge production will generally be compensated by the simulated oxygen consumption of the process. Because the SRT is fixed according to the PTOT balance, the sludge-COD concentration in the process is mainly determined by the influent XI/CODtotal ratio (fXIin). Inert COD (XI) accumulates in the process. Increasing fXIin therefore leads to increasing the COD in the process and vice versa. By adjusting the influent ratio fXIin to 0.18 the model described the measured MLVSS in the reactor (using a conversion factor of 1.4

Model-based evaluation of the upgrading

147

gCOD/gMLVSS). When fXSin is used to fit the COD balance, all model uncertainties related to the production of XI and the influent characterisation of XS and XI are lumped in the influent fXIin fractionation. As a result of adjusting the fXIin fraction, the model predicted accurately the soluble COD in the effluent and total P and TKN in the bioreactor. 7.4.2 Calibrating nitrification and denitrification Nitrification and denitrification were calibrated on the basis of yearly average measurements. Adjusting the oxygen half saturation coefficient of the ammonia oxidisers and heterotrophic biomass was used to fit the nitrification and denitrification in the bioreactor according to Meijer et al (2001). To simulate the measured effluent ammonium and nitrate concentration, a value of 1 mgO2/l and 0.75O2/l and respectively. The calibrated values of and are linked was used for to oxygen diffusion limitation within the sludge floc and oxygen concentration gradients in the tanks caused by non-ideal mixing, processes that are not accounted for in the and are expected to be simulations. Because of differences in mixing intensities, different for each compartment but the same values were used in all compartments for simplicity. Since these values are most sensitive in the anoxic compartment this will not affect the outcome of the simulations significantly. The nitrification batch tests were used in this study as a last step in the calibration procedure. Ammonia and nitrite oxidising activities in the batch test were not predicted very well by the model; observed ammonia conversion and nitrite conversion rates were approximately 20% lower than predicted. Due to the fact that the WWTP is under-loaded, these differences could not be observed when simulating the full-scale yearly average values. There are three possible reasons for this lower experimental nitrification conversion rates in the batch tests than predicted by the model: either the amount of nitrifiers predicted by the model in the plant is too high or the growth rate in the model is too high or both. This means that re-calibration was required. Since the decay rate is the most uncertain parameter we choose to calibrate on this coefficient. Increasing the decay rates by 30 % (from 0.15 and 0.10 day−1 to 0.20 and 0.13 day−1 for the ammonia and nitrite oxidisers respectively) resulted in a good fit of the predicted data to the measurements. When this newly calibrated model was used for the full-scale simulation of the treatment plant, the nitrogen content in the effluent did not change. In a separate test similar results were obtained when the growth rate was reduced because of the high correlation of the parameters decay rate and growth rate. 7.5 Model validation Model validation was performed via validating the capacity to predict the measured concentrations of NH4+, NO2− and NO3− along the bioreactor length using the dynamic influent data. The recorded average daily influent flow during the period 13–26 November 2000 (Figure 7.3) was used as input flow for the model. The simulated value of NH4+, NO2− and NO3− of day 24 November (the 12th day of the 14 days dynamic simulation period) were compared with the values obtained from sampling over the

Nitrification in saline industrial wastewater

148

length of the bioreactor conducted on the same day. The model provides a rather accurate prediction of the NH4 and NO3 along the bioreactor (Figure 7.4).

Figure 7.3 Measured influent flow (Qin1) during 13–26 November 2000, on day 11/24/2000 samples were taken and used for model validation (see Figure. 7.4)

Figure 7.4 Measurement results in terms of nitrogen over the length of the bioreactor of WWTP Ecco (markers) on 11/24/2000. The curves are calculated by the calibrated model.

Model-based evaluation of the upgrading

149

7.6 Model application This study was conducted in the first year of changing the ownership of the company. The new owner performed various viability studies on increasing the hide production and application of different tannery process and operational strategies. The main question, which arose, was whether the existing WWTP could cope with these future requirements. Based on the plant performance and effluent quality, an additional volume requirement (25% of the bioreactor volume) was proposed by the operator. The old equalisation tank was selected to be the additional volume, and to be operated as an anoxic reactor. This upgrading aimed to increase the plant capacity and to achieve full denitrification. Evaluating the upgrading concept for the WWTP was the main focus of our modelling study. The validation step demonstrated the capability of the proposed model to describe the WWTP correctly under both steady state and dynamic conditions. Hereafter, the model was used as a tool to obtain insight in the existing plant performance, possible extension and ways of process optimisation. 7.6.1 Evaluation of the existing plant performance and possible extension The model visualised the existing conditions and the potential capacity of each process involved in the purification system. Nitrification as the most sensitive process was accomplished within 60% of the total bioreactor volume under both steady state and dynamic conditions (Figure 4). Moreover, denitrification seemed incomplete (only 80%). The same effluent quality was also predicted even up to a load increase by a factor two. This illustrates that the WWTP Ecco has been under-loaded and that the bioreactor volume is not the limiting factor if the load is expanded. It is possible that the aeration capacity and the sludge treatment units will be limiting factors in the treatment process in possible expansion. Despite the high concentration of NO3 in the effluent (50 mg-N/l) the present set-up of the plant has a high denitrification potential. Within the existing process configurations better effluent quality with less operational costs could be achieved via process optimisations. 7.6.2 Process optimisations One of the major achievements of plant optimisation is enhancing the denitrification activity to achieve full denitrification in the system. The model was used to quantify the main limiting factors, which hinder the denitrification. Anoxic zone volume, internal recirculation flow and the availability of easily degradable COD were investigated as limiting factors. Different modifications were simulated under steady-state conditions to study their effect on the realisation of full denitrification. COD limitation seemed to be the most crucial factor for the denitrification limitation. Different modifications were simulated under steady-state conditions to study their effect on the realisation of full denitrification. Introducing the reject water (Qin2) into the denitrification zone in addition to increasing the denitrification zone (up to 4000 m3) and increasing the internal recycle flow (up to 10000 m3/h) will directly lead to full denitrification. However, increasing the internal recycling flow should be carefully performed to avoid a highly aerated recycling flow to the denitrification zone. Therefore, oxygen should be monitored. Improving the

Nitrification in saline industrial wastewater

150

mixing condition in the denitrification zone was recommended to improve the denitrification capacity of the system. This results in a reduction in oxygen consumption, reduction in additional alkalinity to be provided and a reduction in the effluent pollutant (NO3) concentration. This would therefore, result in a reduction in the over-all operational costs. 7.7 Discussion The influent characterisation results of WWTP wastewater do agree with the reported values of the tannery wastewater (Orhon et al 1999a). The modified ASM1 model proposed in this paper for COD and N removal described the performance of the WWTP An accurate and correct well with the adjustment of only two parameters description of the system configurations, balancing the operational data with the measurements to accurately calculate the SRT and the use of the stepwise calibration proposed by Meijer et al (2001) simplified the complexity of the model calibration. The use of batch tests for model re-calibration was useful because the plant was under loaded and therefore the effluent concentrations profiles were not sensitive for the rate parameters. Model validation under dynamic conditions is of great significance in industrial WWTP because the plant is usually working under highly dynamic conditions in comparison with domestic WWTP. The observed reduction in ammonia and nitrite oxidising activity is attributed to the presence of salt (7.5 g Cl−/L). This decline in the activity for ammonia and nitrite oxidisers is in agreement with earlier results of salt inhibition effects on nitrifying sludge (Moussa et al 2003b). The use of a higher decay rate of nitrifiers to mathematically describe the salt impact simplifies the model description of salt inhibition (only one parameter to calibrate). Other parameters for calibration could have been chosen but the decay rate was considered as the most uncertain coefficient in the model. The potential use of the model was illustrated in the evaluation of the upgrading of the WWTP. During this evaluation the bioreactor capacity and process configuration were investigated. The model provided a better and quantitative understanding of the plant operation and treatment process. The investigation of several modifications to reach further plant optimisation was very time-effective and cheap (Salem et al 2002). Because it gives quantifiable results, modelling supported the decisions to be taken with respect to the plant extension. It was originally proposed to increase the plant volume (25% with the use of old equalisation tank) to cope with anticipated load expansion. The simulations clearly indicated that with the present system and future increase in load good effluent quality can be reached even combined with an increased denitrification. An additional application of the model is the usage by the plant manger or the plant staff. These people use the model in a different way, namely, as a tool to quantitatively predict the effect of certain decisions on the treatment process (Salem et al 2002). This application of the model is very useful in case of industrial WWTP, like the tannery studied here. The cost of treating the wastewater generated during the production process has a crucial impact on the overall production costs. This also increases the awareness of the impact of each pollutant term in each part of the process and stimulates the practice of waste minimisation to have an environmentally friendly production.

Model-based evaluation of the upgrading

151

7.8 Conclusions Activated sludge models commonly applied for domestic wastewater treatment plant could also be used for industrial WWTPs if the following steps are carefully considered: 1. An accurate description of the system configurations; 2. Balancing the operational data with the measurements to accurately calculate the main important reactor input parameters (flow rates and SRT); 3. Selection of model process and components which are significant and dynamic in this system configuration; 4. Complementary analyses to assess the wastewater and the sludge characterisation; 5. Stepwise calibration under steady-sate conditions and finally model validation under dynamic conditions. The modified ASM1 model proposed in this paper for COD and N removal proved to be able to describe the performance of Ecco Tannery Holland B.V. WWTP wastewater treatment plant. The model was successfully used to evaluate and optimise the plant performance. In addition it was demonstrated that the model could be used by the plant manager to support his decisions quantitatively resulting in saving time and money. References Brdjanovic D, van Loosdrecht MCM, Versteeg P, Hooijmans CM, Alaerts GJ, Heijnen JJ (2000) Modelling COD, N and P removal in a full-scale WWTP Haarlem Waarderpolder. Water Res. 34:846–858. Ekama GA, Dold PL, Marais GvR (1986) Procedures for determining influent COD fractions and the maximum specific growth rate of heterotrophs in Activated sludge systems. Wat. Sci. Tech. 18:91–114. Hellinga C, van Loosdrecht MCM , Heijnen JJ (1999) Model based designed of a novel process for ammonia removal from concentrated flow. Mathematical and Computer Modelling of Dynamic Systems 5:351–371. Henze M, Gujer W, Mino T, van Loosdrecht MCM (2000). Activated sludge models ASM1, ASM2 and ASM3. Scientific and Technical Report, IWA Publishing, London. Meijer SCF, van Loosdrecht MCM, Heijnen JJ (2001) Metabolic modelling of full-scale biological nitrogen and phosphorus removing WWTPs. Water Res. 35:2711–2723. Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003a) Improved method for determination of ammonia and nitrite oxidation activities in mixed bacterial cultures. Appl. Microbiol. Biotechnol. 63:217–221. Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003b). Short term effects of various salts on ammonia and nitrite oxidisers in Mixed Bacterial Cultures. Appl. Microbiol. Biotechnol. (Submitted). Nowak O, Svardal K , Schweighofer P (1995) The dynamic behaviour of nitrifying activated sludge systems influenced by inhibiting wastewater compounds. Wat. Sci. Tech. 31:115–124. Nowak O, Franz A, Svardal K, Muller V, Kuhn V (1999) Parameter estimation for activated sludge models with the help of mass balances. Water Sci. Tech. 39:113–120. Orhon D, Ates E, Ubay Cokgor E (1999a) Modelling of activated sludge for tannery wastewaters. Water Environment Research, 71:50–63.

Nitrification in saline industrial wastewater

152

Orhon D, Ubay Cokgor E, Sozen S (1999b). Experimental basis for the hydrolysis of slowly biodegradable substrate in different wastewaters. Wat. Sci. Tech. 39:87–95. Reichert P (1998) AQUASIM 2.0 Computer Program for the Identification and Simulations of Aquatic Systems. Swiss Federal institute For Environmental Science and Technology (EWAG), Dübendorf, Switzerland. Salem S, Berends D, van Loosdrecht MCM, Heijnen JJ (2002) Model-based evaluation of a new upgrading concept for N-removal. Wat. Sci. Tech. 45:169–176. van Zuylen HJ (1993) From scientific computation to decision support. Knowledge-based system 6:3–10.

Appendix to Chapter 7

APPENDIX 7.1 Stoichiometry matrix of the activated sludge model (modified ASM1) applied for the industrial WWTP Ecco Tannery B.V.

APPENDIX 7.2 Process kinetics of the activated sludge model (modified ASM1) applied for the industrial WWTP ECCO Tannery B.V.

Chapter 8 Evaluation and Outlook

8 Evaluation and Outlook 8.1 Introduction The aim of this research was to achieve a better understanding of nitrification under saline conditions. The research was carried out in two phases. In the first phase, laboratory scale activities were conducted to obtain fundamental data to determine the relationship between salinity and nitrification. In the second phase the results collected from the laboratory experiments were compared and validated with the results collected from full-scale treatment plants. Modelling was employed in both phases to provide a mathematical description for salt inhibition on nitrification and to facilitate the comparison. The research has lead to an improved understanding of the effect of salinity on nitrification, while subjects for further research were also identified. The research findings and challenges are described in more detail below. 8.2 Detection of nitrification inhibition In industrial activated sludge wastewater treatment there is always the risk of inhibitory compounds in the influent due to spilling or other incidents in the industry. An effective control in order to maximise the use of the plant volume is needed. Therefore an early detection of inhibitory compounds is essential. The nitrification process is one of the most sensitive processes within modern WWTPs and any disturbance might lead to washout of nitrifiers. Moreover, it requires a long time period before these microorganisms are fully re-established in the plant. Thus, a high priority is always given to the nitrification parameters when control strategies are applied to the activated sludge system designed for COD and N removal. One of the achievements of this study was the development of a simple and reliable method to measure the activity of ammonia and nitrite oxidisers separately in mixed bacterial cultures. The main advantage of this method is the possibility to differentiate between the activities of ammonia and nitrite oxidisers without the use of metabolic inhibitors. The method can be applied to measure the effects of specific inhibitors on the activity of both groups of nitrifiers. It can, therefore, also be used as a detection method for early diagnosis of nitrification problems. Incorporating the method in the existing protocol of the automatic respirometer/on-line is promising and could be a helpful tool to optimise nitrogen removal processes.

Evaluation and outlook

157

8.3 Nitrification under salt stress Initially, short-term effects of several salts on nitrification were studied. Different types of salts appeared to have different inhibition effects on the ammonia and nitrite oxidisers. Therefore, formulas were developed to unify these differences in inhibition effects on both ammonia and nitrite oxidisers. These formulas can be used in the design, operation and control of the WWTPs that operate under salt stress. These findings facilitate the mathematical description of salt inhibition on nitrifiers from a process-engineering point of view, although the real inhibition mechanisms are not yet clearly understood. Investigating the actual inhibition mechanisms in depth (cell membrane, energy requirement, enzyme inhibition) might not be required for optimal process design. Nevertheless, it could be an interesting subject for further research, since it might lead to a more accurate description of inhibition. Acclimatisation is a term, which is commonly used to describe the recovery of bacterial cultures when exposed to a certain inhibitor over extended periods of time. Acclimatised cultures are recommended for seeding a system that will be operated under similar conditions to shorten the start-up period. Salt acclimatisation was investigated within the course of this research. The effect of salinity on the activity, the composition of nitrifying populations and floc characteristics was observed. The main finding was that acclimatised and non-acclimatised nitrifying sludge were behaving similarly with respect to activity and population selection in response to different salt concentrations. This means in practice that sludge from a conventional domestic nitrifying WWTP can be used as a source for nitrifiers to become adapted to salt stress via gradual adaptation. This avoids the necessity of seeding a system that is expected to operate under salt stress with salt acclimatised sludge. The monitoring of the activity and population composition of a full-scale reactor during the start-up period, after seeding according the proposed approach could be an interesting subject for further research. The selection of Nitrosomonas europaea and Nitrobacter sp. as a result of gradual increase in salinity was shown by the re-growth of only these two species after lowering the salt concentration to zero values. This result shows the high resistance of these species to salt stress. Interestingly, the systems were able to fully recover their nitrification activity with reduced diversity in the population of nitrifiers. These findings support the hypothesis that the conversion rate depends on environmental conditions and not on the type of species. In other words, different types of nitrifiers behave kinetically similar under similar environmental conditions. This hypothesis remains to be verified and could be an interesting topic for further research. 8.4 Role of predators in nitrifying activated sludge systems The current design concepts of activated sludge systems consider bacteria as the sole active biomass. The activity of all other microbial community members (protozoa, metazoa, bacteriophages etc.) is hidden in a simple decay process responsible for the reduction of active biomass. This decay process is the sum of several independent

Nitrification in saline industrial wastewater

158

processes like maintenance, lysis due to bacteriophage infection and predation. Commonly, protozoa in activated sludge systems were used as bio-indicators linked to process performance and effluent quality and crucial in obtaining a good effluent quality. In this research the presence of protozoa and metazoa in the SBRs (Sequencing Batch Reactors) was monitored in order to provide information to understand their role in these systems. A mathematical model describing the interaction between nitrifiers, heterotrophs and predators has been developed. The model successfully describes the performance of SBRs and predicts the fraction of active biomass (ammonia oxidisers, nitrite oxidisers and heterotrophs). The developed model is a first step in generating a better understanding of the role of predators in activated sludge systems and it opens the black box in which all predator processes are put under the decay process. Further investigations are needed to verify the laboratory scale findings under full-scale activated sludge conditions. Moreover, further development of microbiological measurements able to quantify the active biomass for each individual group of organisms is highly needed. These developments can help in bridging the gap between the microbiologist, biochemists, engineers and process modellers. 8.5 Application of activated sludge model in industrial WWTPs Activated sludge models have been successfully applied in the last two decades for domestic wastewater treatment plants. However, their application to industrial WWTPs is still limited for several reasons. The main reason is the complexity of industrial WWTPs in terms of dynamic flow, waste composition, etc. Another important reason relates to the lack of information of model parameters under extreme conditions prevailing in industries (pH, temperature, toxicant presence). Besides, large variations exist between similar industries, which make the treatment process specific for each individual plant. All these factors have limited the applicability of the existing activated sludge model and the transfer of existing experiences between industries. Therefore, the use of large reactor volumes, excess aeration end extra chemical dosages is commonly practised in industries to ensure compliance with effluent quality requirements. Mathematical modelling has shown to increase the confidence of operators in WWTP design and thereby limits the use of an unnecessary high safety factor for new or upgraded treatment plants. This will directly be translated into lower treatment costs. In this study a modified version of the activated sludge model No. 1 (ASM 1) was applied under static and dynamic conditions. Firstly, the steady state model was used to calculate the actual specific activity of ammonia and nitrite oxidisers of different WWTPs investigated in the course of this research. The routine operating data were used to assess the fraction of nitrifiers in the sludge. The results from different types of nitrifying populations (pure cultures, enriched cultures) and different scale reactors (lab scale, pilot scale, full scale WWTP) were compared. The steady state model confirmed the behaviour of nitrifiers under salt stress and validated the results obtained from laboratory scale (this study) to be interpreted on full scale. The use of the model to quantify the specific activity and to correlate this to the presence of specific populations within different systems was promising and could be applied in future research. Secondly, the dynamic

Evaluation and outlook

159

model was used to describe a full-scale tannery WWTP. The model was used to obtain insight in the plant performance, possible extension and options of upgrading. The model illustrated the potential of the plant to have a better effluent quality and to handle a higher hydraulic load with simple modifications. The model was used as an effective tool to visualise and quantify the plant conditions. The results demonstrate the application of the model as a decision support tool especially in cases when industrial treatment plants are exposed to substantial fluctuations in production. Based on the experience gained in the course of this research it is concluded that more efforts should be made to motivate designers and operators of industrial WWTPs to become familiar with activated sludge models. A good approach would be to develop a standard influent characterisation procedure for industrial wastewater similar to the one successfully developed and applied for domestic wastewater. Moreover, the model parameters need to be extended to satisfy the operational conditions in industries (e.g. high temperature and long SRT). Disturbance factors such as pH, salinity, etc could be incorporated in the model. All this in addition to producing simple and reliable detection sensors could promote the application of activated sludge modelling in the industrial field. In conclusion, the research has lead to an improved understanding of the effect of salinity on nitrification. The results obtained within the course of this research can be used to improve the sustainability of the existing WWTPs operated under salt stress. The findings also form a guideline for more economical and sustainable design and start up of WWTPs dealing with salt in future.

Samenvatting Biologische verwijdering van stikstof via nitrificatie en denitrificatie wordt zeer algemeen toegepast in de afvalwaterzuivering. Nitrificatie, de snelheidsbepalende stap in het hele proces, blijkt echter moeilijk te sturen, omdat nitrificerende bacteriën enerzijds traag groeien en anderzijds erg gevoelig zijn voor allerlei omgevingsfactoren (temperatuur, pH, zuurstofconcentratie, remmende stoffen). Tot nu toe heeft wetenschappelijk onderzoek zich voornamelijk gericht op nitrificatie in huishoudelijk afvalwater en er is op dit gebied dan ook veel kennis en practische ervaring opgedaan met als resultaat de algemene toepassing van biologische stikstofverwijdering uit huishoudelijk afvalwater. Helaas zijn deze positieve ervaringen niet direct toepasbaar op industrieel afvalwater vanwege de specifieke samenstelling ervan. Veel industrieën hebben ook nog te maken met hoge zoutconcentraties in hun afvalwater. Het zoutgehalte van industrieel afvalwater zal in de toekomst verder toenemen, omdat de politiek een efficiënter gebruik van water eist. Hogere zoutconcentraties zullen een negatieve invloed hebben op de nitrificatie. Een beter begrip van de effecten van zout(en) op de nitrificatie is dan ook hard nodig. Weliswaar is bekend dat de activiteit van zowel ammonium- als nitrietoxideerders afneemt bij hogere zoutconcentraties, maar veel details zijn nog niet bekend: Geven alle zouten dezelfde remming, wat is de maximaal getolereerde zoutconcentratie, zijn ammonium- of nitietoxideerders het meest gevoelig, is adaptatie aan zout mogelijk, zijn sommige nitrificeerders beter bestand tegen zoutbelasting dan andere? Het begrijpen van de effecten van zout op de nitrificatie staat centraal in dit proefschrift, waarbij de bovengenoemde vragen de leidraad vormen voor het onderzoek, dat in twee fasen is uitgevoerd. De proeven op laboratoriumschaal (eerste fase) leverden fundamentele gegevens over de invloed van zout op de nitrificatie; in de tweede fase werden de resultaten van de laboratoriumproeven vergeleken met en gevalideerd met proeven in industriële afvalwaterzuiveringen op praktijkschaal. Computer modellen zijn gebruikt in beide fasen, zowel voor een goede mathematische beschrijving van de relatie tussen zout en nitrificatie als om de onderlinge vergelijking te vergemakkelijken. In de eerste fase is er een methode ontwikkeld om simultaan de activiteit van ammonium- en nitrietoxideerders in mengpopulaties van bacteriën te meten; deze methode is gedurende het hele onderzoek toegepast om de remmende effecten van zout op zowel ammonium- als nitrietoxideerders vast te stellen. Ammonium- en nitrietoxideerders reageerden verschillend op de aangeboden zouten. Vervolgens werden de lange-termijn effecten van aanpassing aan 10 g NaCl per liter bestudeerd, niet alleen op de activiteit van de nitrificeerders, maar ook op de samenstelling van nitrificerende bacteriepopulaties en op de eigenschappen van het bacterieslib. Er werd geen verschil gevonden tussen de wel en niet aan 10 g NaCl geadapteerde nitrificeerders; in beide gevallen werd bij 40 g NaCl een remming van 95% gevonden. De enig overgebleven

Samenvatting

161

nitrificeerders bij hoge zoutconcentraties waren Nitrosomonas europaea en Nitrobacter sp. Hogere zoutconcentraties hadden een betere bezinking van het bacterieslib tot gevolg. Al bij lage zoutconcentraties werden de aanwezige bacterie-etende protozoën en metazoën gedood. Op basis van dit effect is een mathematisch model opgesteld om de interacties tussen nitrificeerders, heterotrofe bacteriën en bacterievore organismen aanschouwelijk te maken. Tijdens fase 2 werd duidelijk dat het effect van zout op de activiteit van nitrificeerders in huishoudelijk en industrieel afvalwaterzuiveringen niet verschilde van de effecten op laboratoriumschaal. Het bestaande actief-slib model (ASM1) is aangepast om CZV en stikstofverwijdering in industriële afvalwaterzuiveringen onder zoutbelasting te simuleren. Het onderzoek heeft geleid tot een beter begrip van het effect van zout op nitrificatie. De verkregen resultaten kunnen worden gebruikt om de duurzaamheid van de huidige afvalwaterzuiveringen met verhoogd zoutgehalte te verbeteren.

Acknowledgments Thanks to Allah the exalted, the most merciful, for giving me the strength and persistence to keep going with this research even during the most difficult moments. May Allah accept this work and count it as a good deed. My deep thanks to my country Egypt where I grew up and had the first lessons and experiences in my life, I hope I can pay it back some day. I would like to express my thanks to my promoters: Prof. Huub Gijzen who stimulated and supported the formulation of my PhD joint project. Huub, you had an important role in guiding and challenging not only during the research but also after the completion; Prof. Mark van Loosdrecht for his ideas in the earlier stage of the research, valuable feedback, enthusiasm and friendship. Mark, you were always inspiring, educating and available when I needed you. I am also grateful to my supervisors: Tineke Hooijmans who supported me especially during the difficult initial period of this study and for her efforts in establishing the project; Henk Lubberding who guided me during the research, inspired my microbiological experience and spent lots of time helping in finalizing this work. I would like to acknowledge the MSc students, Samir Ibrahim, Deepthi Sumanasekera, Alejandro Rivera Rojas, Orleans Garcia, Aboubakar Gomina, Jochem Smit, Yan Song, Said Rehan, Akram Botorous and Hala Elsadig who were involved in this study for their valuable contribution. I would like to express my appreciations for the endless help and support of the laboratory staff at UNESCO-IHE: Fred Kruis, Frank Wiegman, Kees Bik, Peter Heerings,. Special thanks to the staff of Kluyver Laboratory of TU Delft: Sjaak Lispet, Stef van Hateren and Udo van Dongen for their continuous help during this study. Great thanks to my colleagues at IHE Saleh and Saber for their continuous encouragement and inspiring discussions on our related research topics. Special thanks to my friend M.Fiala for all what he did and what he is still doing for me. These acknowledgments would not be complete without expressing my gratefulness to GOSD for expanding my background as a structural engineer with environmental experience and also my appreciation to UNESCO-IHE for providing the best working environment. I remain very grateful and gratified to my family, especially my parents, my wife, my brothers and sister and the two young researchers Adham and Yusuf for their support, patience, understanding and prayers throughout the period of this work.

Curriculum Vitae Moustafa Samir Moussa was born in Cairo, Egypt on August 11th, 1965. He graduated in 1987 as a Civil Engineer from the faculty of Engineering, University of Ain Shams, Cairo. He was awarded his BSc with general grade “very good” and “distinction” for his awarding project. After finishing his military service in 1989 he started his professional career in GOSD, General Organization for Sanitary and Drainage, greater Cairo. Here he was responsible for the sanitary, hydraulic and structural design of Shoubra El Khimma wastewater treatment project. In addition he was appointed as instructor for different training courses in wastewater collection and treatment for the new work orientation training programs. In October 1995, he studied at the International Institute for Infrastructural, Hydraulic and Environmental Engineering (IHE) in Delft (now called UNESCO-IHE Institute for Water Education). In September 1996 he obtained a post-graduate diploma in Sanitary Engineering and was awarded a scholarship from Shell to continue his MSc research. In 1997 he obtained his MSc degree and continued his research through an additional fund from Shell. This research formed the starting point of his PhD research, a joint project of UNESCO-IHE, TU Delft, Shell Global Solutions B.V., Ecco Tannery Holland B.V., Heiploeg B.V. and BTS Senter (an agency of the Dutch Ministry of Economic Affairs). Since Januari 2004 the author works as a researcher/lecturer at UNESCO-IHE.