(Moustafa Samir Moussa) Nitrification in Saline in PDF

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 Mnchen, 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 Routledges 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
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 (day1)
bNO2
Aerobic decay rate of nitrite oxidisers (day1)
bH
Aerobic decay rate of heterotrophic biomass (day1)
bpredators
Aerobic decay rate of predators (day1)
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 NH4N/L.h)
RNO2
volumetric uptake rate of nitrite (mg NO2N/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 NH4N/L)
SNO3
Concentration of nitrate (mg NO3N/L)
SNO2
Concentration of nitrite (mg NO2N/L)
SN2
Concentration of nitrogen (mg N/L)
SS
Concentration of organic substrate (mg COD/L)
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 NO2N (g

COD/g NH4N)
YNO2
Yield coefficient of nitrite oxidisers per NO3N (g COD/g

NO2N)
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+
NH4
Anoxic reduction factor for ammonia oxidisers decay
NO2
Anoxic reduction factor for nitrite oxidisers decay
Anoxic reduction factor for heterotrophic growth

Maximum growth rate of ammonia oxidisers (day1)
Maximum growth rate of nitrite oxidisers (day1)
Maximum growth rate of heterotrophic biomass (day1)
predators
Growth of predators, presented in the model as predation

rate (day1)
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 NaClCl/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 1015% 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 N2fixing 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
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 isvia ureaconverted 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
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 11010 to 91010 kg. Current production is equivalent to about
37% of the total amount of nitrogen input achieved via terrestrial and marine biological
N2 fixation (about 241010 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
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/gN). 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
2550 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
produced, with ammonia concentrations of 700 to 1000 mg-N/L. Agricultural leaching

can result in large amounts (525 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
0.454.1
oil refinery
0.020.9
coal gasification
12.5
fertiliser
0.21
synthetic fibre
0.8
slaughterhouse
0.15
livestock: swine
2.3
livestock: cattle
0.52.3
rendering plant
0.8
dairy
0.6
distillery
1.5
cellulose and paper
0.25
Pharmaceuticals
16
0.10.4
explosives
glass
Nitrate concentration range (g

N/L)
212.5
0.30.65
0.48
electronics
0.52
pectin
12.7
uranium processing
411.3
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
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
20
80
Withdrawal
20
90
350
Consumption
115
50
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
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 Europes 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 plants100,000 p.e.
80%
1 mg/L for plants>100,000 p.e.

Total nitrogen
15 mg/L for plants100,000 p.e.
7080%
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 (Blch 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.
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, Europes 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 (11.5 days) and high
temperature (3040C) 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.
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 processan important process in the nitrogen cycle in natureis
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
Rser 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
subclass, which contains only the Nitrosococcus oceanus and Nitrosococcus halophilus.
The second group belongs to the subclass, which contains two clusters, the Nitrosospira
cluster and the Nitrosomonas cluster (Koops and Rser 2001; Purkhold et al 2000).
16
Table 1.6 The cultured ammonia oxidising bacteria

and information on ecophysiological parameters
and preferred habitats (Koops and Rser 2001).
Species
Ecophysiological parameters
Preferred habitat
Salt requirements
Substrate
(NH3)
affinity
Nitrosomonas
europea
Nitrosomonas
eutropha
Nitrosomonas
halophila
Nitrosomonas
mobilis
halotolerant or
moderately
halophilic
3061 M
sewage treatment plants,

eutrophic freshwater and
brackish water
Nitrosomonas
communis
Nitrosomonas
sp.I
Nitrosomonas
sp.II
no salt requirement
1443 M
soils (not acid)
Nitrosomonas
nitrosa
no salt requirement
1946 M
eutrophic freshwater
no salt requirement
1.94.2 M
oligotrophic freshwater
natural solis
Nitrosomonas
marina
Nitrosomonas
sp.III
Nitrosomonas
aestuarii
obligatory
halophilic
5052 M
marine environment
Nitrosomonas
cryotolerans
obligatory
halophilic
4259 M
Proteobacteria Nitrosomonas
ureae
Nitrosomonas
oligotropha
Nitrosolobus
multiformis
Nitrosovibrio
tenuis
Nitrosospira sp.I
Nitrosococcus
oceani
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 ( subclass of the Proteobacteria), Nitrosococcus oceani
and Nitrosococcus halophilus, are obligately halophilic. The group located in the
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 Rser
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 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 and the 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 subclass of the Proteobacteria (Koops and
Rser 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 Rser
2001).
18
Table 1.7 The cultured nitrite oxidising bacteria and

information on ecophysiological parameters and
preferred habitats (Koops and Rser 2001).
Species
Ecophysiological parameters Preferred habitat

Salt requirements
Nitrobacter
alkalicus
Supclass of
Proteobacteria
Nitrobacter
winogrodskyi
alkali-and halotolerant
soda lakes
no salt requirement
fresh water soils
and rocks
Nitrobacter
vulgaris
Nitrobacter
hamburgensis
Nitrococcus
mobilis
obligatory halophilic
Nitrospina gracilis
Nitrospira marina
Nitrospira
moscoviesis
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 Rser
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 (+) and the activity of an anion is written as the product of its concentration
and an activity coefficient ().
Experimentally, only the product (+) 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 (). This means
activity coefficient (=(+)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 103,
the ratio (CO32/HCO3)=0.58, while if we had not taken the effect of ionic strength into
account, the ratio (CO32/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++OHNH3+H2O
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
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(pKapH))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
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
6570
seawater
3.56.5
2030
DW
DA
no
LA
70%
Yu et al
inhibition of (2002)
max
nitrifiers
(0.25 day1)
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
560
NaCl
6,18,
30,36
25
SW
EN
no
LA
>18 gCl/L Dincer and

SRTmin is 25 Kargi (1999)
days, at
0gCl/L 12
days
Introduction
1020
NaCl
3155
NaCl
2043
SW
EN
no
LA
3,6,12,18 nm*5
2733
SW
DA
no
LA
NaCl
3,3,12,18
nm
2733
SW
SA
to 5 g
Cl/L
LA
<5
NaCl
nm
2022
SW
DA
no
LS
Intrasungkha
et al (1999)
83
NaCl
1820
2830
SW
MS
to
seawater
AQ
NO2 accum Catalanulation due Sakairi et al

to limitation (1996, 1997)
of trace
elements
and CO2
<5
NaCl
14
nm
nm
DW
DA
no
BA
Andreadakis
et al (1997)
NaCl
10
78
30
IW
SA
to 10g
Cl/L
PF
Vredenbregt
et al (1997)
0%(compared
to 10Cl)
NaCl
20
78
30
IW
SA
to 10g
CL/L
PF
NO2 was
the only
product>20
gCl/L
57%(compared
to 20gCl)
NaCl
34
78
30
IW
SA
to 10g
Cl/L
PF
below 20
gCl/L good
fluidizable
particles are
formed
NaCl
10
7
8.3
2530
IW/SW
DA
no
PA
Shock load
caused
major
inhibition
NaCl
20
7
8.3
2530
IW/SW
DA
no
PA
*1
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
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. 37:135142.
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. 24:97101.
Blch H (2001) EU ploicy on nutrients emission: legislation and implementation Water Sci.
Technol. 44:16.
Bock E, Koops H-P, Harms H, Ahlers B (1991) The biochemistry of nitrifying organisms. In:
Variations of Autotrophic Life (Shively, J.M., Ed.): 171200. Academic Press, London.
Brunet RC, Garca-Gil LJ (1996) Sulfide-induced dissimilatory nitrate to ammonia in anaerobic
freshwater sediments. FEMS Microbiol. Ecol. 21:131138.
Butler JN (1964) Ionic Equilibrium-A mathematical Approach. Addision-Wesly publishing
company, INC. 48439.
Campos JL, Mosquera-Corral A, Snchez M, Mndez 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. 84:563571.
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.
34:267274.
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
40C. Geochimica et Cosmochimica Acta 59:24032421.
Cole JA (1996) Nitrate reduction to ammonia by enteric bacteria: redundancy, or a strategy for
survival during oxygen starvation. FEMS Microbiol. Letters 136:111.
26
Cosgrove WJ, Rijsberman FR (2000) World Water Vision, Making Water Everybodys 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. 36:34552.
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. 67:52735284.
Dincer AR, Kargi F (1999) Salt inhibition of nitrification and denitrification in saline wastewater.
Environ. Technol. 29:11471153.
Dincer AR, Kargi F (2001) Salt inhibition kinetics in nitrification of synthetic saline wastewater.
Enzyme and Microbial Technology 28:661665.
Emerson K, Russo RC, Lund RE, Thurston RV (1975) Aqueous Ammonia Equilibrium
Calculation: Effect of pH and Temperature. J. Fish. Res. Board Can. 32:23792383.
Furumai H, Kawasaki T, Futawatari, T, Kusuda T (1988) Effects of salinity on nitrification in a
tidal river. Water Sci. Technol. 20:165174.
Galinski EA, Truper BJ (1994) Microbial behaviour in salt stressed ecosystems. FEMS Microbiol.
Rev. 15:95108.
Galinski EA (1995) Osmoadaptation in bacteria. Adv. Microb. Physiol. 37:273327.
Gijzen HJ, Bijlsma M (2000) Strategy options for sewage management to protect the marine
environmentTechnical. 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. 44:123132.
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. 37:135142.
Hunik JH, Meijer HJG, Tramper J (1992) Kinetics of Nitrosomonas europaea at extreme substrate,
product and salt concentrations. Appl. Microbiol. Biotechnol. 37:802807.
Hunik JH, Meijer HJG, Tramper J (1993) Kinetics of Nitrobacter agilis at extreme substrate,
Johansson O, Wedborg M (1980) The Ammonia-Ammonium equilibrium in sea water at
temperatures between 525 C. J. Solution Chemistry 9:3744.
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. 64:304251.
Koops H-P, Pommerening-Rser A (2001) Distribution and ecophysiology of the nitrifying bacteria
emphasizing cultured species. FEMS Microbiol. Ecology. 37:19.
Kuai L, Verstraete W (1998) Ammonium removal by the oxygen-limited autotrophic nitrificationdenitrification system. Appl. Environ. Microbiol. 64:45004506.
Kuenen JG, Robertson LA (1994) Combined nitrification-denitrification processes. FEMS
Microbiol. Rev. 15:109117.
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. 16:177183.
Oren (1999) Bioenergetics aspects of halophilism. Microbiol. Molecul. Biol. Rev. 63:334348.
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-Rser 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.
66:53685382.
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. 2123 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. 23:635650.
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 26:401408.
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. 167:106111.
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
Simon J (2002) Enzymology and bioenergetics of respiratory nitrite ammonification. FEMS
Sliekers A, Derwort, Kuenen JG, Strous M, Jetten MSM (1998) Completely autotrophic ammonia
removal over nitrite in one single reactor. Water Res. 14:2345.
Sliekers A, Third K, Abma W, Kuenen JG, Jetten MSM (2003) CANON and Anammox in a gaslift
reactor. FEMS. Lett. 218:339344.
Smil V (1997) Global population and the nitrogen cycle. Scientific American 277:5863.
Snoeyink VL, Jenkins D (1980) Water chemistry. John Wiley & Sons, New York:7482.
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. 176:170177.
Stalely JT, Bryent MP, Pfennig N, Holt JG (1989). Aerobic chemolithotrophic bacteria and
associatd organisms. Bergeys Manual of Systematic Bacteriology (vol.3), Williams and
Wilkins, Boltimore, USA: 18071834.
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. 31:19551962.
Suzuki I, Dular U, Kwork SC (1974) Ammonia or ammonium ion as substrate for oxidation by
Nitrosomonas europea cells and extracts. J. Bacteriol. 120:556558.
van Loosdrecht MCM, Jetten MSM (1998) Microbiological conversions in nitrogen removal.
Water Sci. Technol. 38:17.
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. 36:93100.
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. 34:23744.
Wagner M, Loy A (2002) Bacterial community composition and function in sewage treatment
system. Environ. Biotechnol. 13:218227.
28
WHO/UNEP (1997) Water Pollution ControlA 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 51:113154.
Woolard CR, Irvine RL (1995) Treatment of hypersaline wastewater in the sequencing batch
reactor. Water Res. 29:11591168.
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.
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. 63:217221.
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: 30C) 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
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
34 times. When re-suspending the bacteria in the same washing medium, a 510 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
35
consequently cause an increase in pH). The dissolved oxygen concentration was kept
above 4.0 mg/L. Temperature was controlled at 30C 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.h1)
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 M 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
(30C, pH 7.5). The profiles were
recorded using a BOM (10-mL). NO2
(10mg-N/L), NH4+ (100 mg-N/L) and
azide (24 M) were added as indicated
in the figure. The nitrite and ammonia
oxidation rates calculated with azide
method were 1015 % lower than with
the developed method.
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 7.50.05
and T 301C). A 50100 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 7.500.05
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 1015 mg NO2N/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 1525 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)
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 h1) or expressed as a
percentage of the activity obtained under reference conditions (pH 7.5, T 30C 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 7.50.05 and T 301C). 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 30C and no inhibitor).
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 105C and weighted on a microbalance. The ash content was
calculated after incinerating the dried filter in an oven for 1 h at 550C.
2.3 Results
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
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 (0.51 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 1020%, 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 M azide to the nitrite oxidisers was about 8090% (8 samples). The test was
repeated (6 samples) with a higher azide concentration of 48 M, 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 (30C, pH 7.5). The profile
40
was recorded using a BOM (10-mL).

After assessing the endogenous
respiration NO2 (10 mg-N/L) and
subsequently azide (24 M) was
added.
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
(30C, 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)
41
2.4 Discussion
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 M azide. Full inhibition, however, could not be
accomplished. The observed inhibition (8090%) of azide to nitrite oxidisers was lower
than the reported value (985%) by Ginestet et al (1998). One of the factors causing the
deviation seems the difference in temperature used in this study (30C) and the one used
in their test (20C). 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
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.
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
(30C, pH 7.5) using the developed
method and the results reported from a
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. 48:835852.
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.
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. 39:505510.
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.
Ginkel CG van, Tramper J, Luyben KChAM, Klapwijk A (1983) Characterisation of Nitrosomonas
europaea immobilised in calcium alginate. Enzyme Microb. Technol. 5:297303.
Technol. 37:135142
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. 45:11781182.
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.
30:825836.
Leenen EJTM, Boogert AA, van Lammeren AAM, Tramper J, Wijffels RH (1997) Dynamic of
artificially immobilised Nitrosomonas europaea: effect of biomass decay. Biotechnol. Bioeng.
55:630641.
Nowak O, Svardal K (1993) Observations on the kinetics of nitrification under inhibiting
conditions caused by industrial wastewater compounds. Water Sci. Technol. 28:115123.
Nowak O, Svardal K, Schweighofer P (1995) The dynamic behaviour of nitrifying activated sludge
systems influenced by inhibiting wastewater compounds. Water Sci. Technol. 31:115124.
Reichert P, Ruchti J, Simon W (1994) Aquasim 2.0. Swiss Federal Institute For Environmental
Science and Technology (EAWAG), Dbendorf, Switzerland.
Smolders GJF, van Loosdrecht MCM, Heijnen JJ (1994) Stoichiometric model of the aerobic
metabolism of the biological phosphorus removal process. Biotechnol. Bioeng. 44:837848.
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.
16:569577.
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.
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: 30C).
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 30C. The two reactors were operated automatically as Sequenced Batch Reactors
(SBRs). The pH was maintained at pH 7.50.05 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:
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 510 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 30C 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.
51

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 105C and weighted on a microbalance. The ash content was
calculated after incinerating the dried filter in an oven for 1 h at 550C. 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 NO2N was observed in the
reactor operated at 30 days, while only 4 mg/L NO2N 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 (30C, pH 7.5).
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)
+++
++
++
compared with the total ammonia-oxidizers (Nso190 probe);

compared to the total Eubacteria (EUB338 probe)
not detected + present ++ present in relatively high number +++ abundant
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).
53
Figure 3.2 Effect of different types of

salt on the activity of ammonia
operated at 30 days SRT, 30C and pH
of 7.5. The 100% is 38022 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).
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, 30C and pH of 7.5. The
100% is 504 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).
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
(30C, pH of 7.5).
56

salt tolerance of nitrite oxidisers from
enriched cultures operated at SRTs of
30 days and 100 days under similar
environmental conditions (30C, 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
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
mgNO2N/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 (30C, 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
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 (30C, 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
59
(30C, 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)
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,
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.
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 NO2N/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
Washington D.C.
63

2560.
Catalan-Sakairi MAB, Wang PC, Matsumura M (1997) Nitrification performance of marine
Bioeng. 84:563571.
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
40C. Geochimica et Cosmochimica Acta. 59:24032421.
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 1:3949.
Microbiol. 67:52735284.
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.
43:918.
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. 41:8590.
Enz. and Micro. Technol. 28:661665.
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. 68:245253.
Furumai, H, Kawasaki T, Futawatari, T, Kusuda T (1988) Effects of salinity on nitrification in a
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. 67:13511362.
Technol. 37:135142.
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. 64:30423051.
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
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. 24:203210.
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. 15:593600.
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. 142:10971106.
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. 63:217221.
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. 36:469481.
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. 33:11651172.
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. 11:897925.
Schramm A, De Beer D, Gieseke A, Amann R (2000) Microenvironments and distribution of
nitrifying bacteria in a membrane-bound biofilm. Environ. Microbiol. 2:680686.
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. 65:36903696.
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.
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).
4.2.1 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 30C (Figure 4.1). The pH was maintained at 7.50.05 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 NaClCl/L in one
step. The reactor was continuously operated at a salt concentration of 10 gCl/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 (2530 g NNO2/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
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
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 105C and weighted on a microbalance. The ash
content was calculated after incinerating the dried filter in an oven for 1 h at 550C. 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-
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
653670
Nm75
CCCCTCTGCTGCACTCTA
Halophilic and
Wagner et al
halotolerant ammonia (1995)
oxidizers belonging to
-proteobacteria of
Nitrosomonas spp.
CGGCAGCGGGGGCTTCGGCC Nitrosomonas genus
Hiorns et al
(1995)
CGATCCCCTGCTTTTCTCC
Ammoma-oxidizers
-proteobacteria
Mobarry et al
(1996)
Nso190
190208
NmV
174191 TCCTCAGAGACTACGCGG
Nitrosococcus
mobilis
PommereningRser et al
(1996)
Nse1472
1472
1489
ACCCCAGTCATGACCCCC
Nitrosomonas
europaea
Juretschko et al
(1998)
NOLI191 191208
CGATCCCCCACTTTCCTC
Various members of
Nitrosomonas
oligotropha lineage
Gieseke (2001)
ALLSPIR 443462
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
662679 GGAATTCCGCGCTCCTCT
Genus Nitrospira
Daims et al
(2000)
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 NO2N under salt free
conditions (R1 and R3) and to 25 mg/L NO2N 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, 30C 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
75
added in two steps). Ammonia oxidisers were more sensitive to short and long term stress
of 10 gCl/L (3639% drop in activity) than nitrite oxidisers (513% 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
9610
R3
45
596
61
596
100
62
64
621
64
34
58
566
95
49
84
515
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).
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, 30C 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 24 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
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 12 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 NaClCl/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.
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 NaClCl/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
NaClCl/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 (040 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
10
20
30
40
R1/R3 R2 R3 R2 R3 R2 R3 R2 R3
10
R2
R3
Ammonia oxidisers:
Nitrosomonas europaea
+++ +++ +++ +++ +++ +++ +++ +++ +++ +++
Nitrosomonas oligotropha
+++
++
++
Nitrosococcus mobilis
++
+++
++
+++
Nitrosospira sp
++
++
++
++
++
Nitrite oxidisers:
Nitrospira sp
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 didnt 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 m at fresh water conditions to 200 m 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.
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 inhibitionassessed as oxygen uptake rateof 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
6570
seawater
3.56.5
2030
DW
DA
no
LA
>125
100
NaCl+
NH4Cl
18
7.8
20
SW
EC
no
LA
nm
nm
70%
inhibition
of max
nitrifiers
(0.25
day1)
11
SVI not
affected
81
due to
initial high
biomass
(20g
VSS/L)
NO2
accumu
lation due
to DO
limitation,
not to
salinity
560
NaCl
6,18,30,36
25
SW
EN
no
LA
nm
nm
>18 gCl/L 5
SRTmin is
25 days, at
0gCl/L 12
days
1020
NaCl
18,30
25
SW
EN
no
LA
nm
nm
NO2
accum
ulation
above 12
gCl/L
3155
NaCl
3,6,12,18
SW
DA
no
LA
na*5
MLSS
decreased
with
increased
NaCl
8,9
2043
NaCl
3,6,12,18
nm
2733
SW
SA
to 5g
Cl/L
LA
na
<5
NaCl
nm
2022
SW
DA
no
LS
nm
nm
83
<5
NaCl
1820
nm*5 2733
2830
SW
MS
to
sea
water
AQ
nm
nm
NO2
accumu
lation due
to
limitation
of trace
elements
and CO2
NaCl
14
nm
nm
DW
DA
no
BA
nm
nm
NaCl
10
78
30
IW
SA
to
10g
Cl/L
PF
nm
nm
10
0% (com
pared to
10gCl)
NaCl
20
78
30
IW
SA
to
10g
Cl/L
PF
nm
nm
NO2 was
the only
product>20
gCl/L
57%
(compared
to 20gCl)
NaCl
34
78
30
IW
SA
to
10g
Cl/L
PF
nm
nm
below 20
gCl/L
good
fluidizable
particles
82
are formed
33%
(compared
to 10gCl)
*1
NaCl
10
7
8.3
2530 IW/SW
DA
no
PA
10
Shock
load
caused
major
inhibition
NaCl
20
7
8.3
2530 IW/SW
DA
no
PA
10
Good
sludge
stability
due to
gypsum
precip
itation
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
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
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
NaClCl/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 NaCLC/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
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; Salvad, 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
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
(3639% drop in activity) than nitrite oxidisers (513% 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.
87
References
Al-Shahwani SM, Horan NJ (1991) The use of protozoa to indicate changes in the performance of
activated-sludge plants. Water Res. 25:6338.
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. 37:135142.
APHA (1998) Standard methods for the examination of water and wastewater, 20 th edn. American
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. 39:505510.
Beun JJ, van Loosdrecht MCM, Heijnen JJ (2002) Aerobic granulation in a sequencing batch airlift
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.
2560.
Catalan-Sakairi MAB, Wang PC, Matsumura M. (1997) Nitrification performance of marine
Bioeng. 84:563571.
Catalan-Sakairi MAB, Yasuda K, Matsumura M. (1996) Nitrogen removal in seawater using
34:267274.
Cousin CP, Ganczarczyk JJ (1998) Effects of Salinity on Physical Characteristics of Activated
Sludge Flocs. Water Quality Research Journal of Canada 33:565587.
Daims H, Brhl 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. 22:43444.
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. 41:8590.
Microbiol. 67:52735284.
43:918.
Enz and Micro Technol. 28:661665.
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. 68:245253.
Eikelboom DH (2000) Process control of activated sludge plants by microscopic investigation.
IWA publishing Alliance House, London, UK.
Ekama GA, Barnard JL, Gnthert 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
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. 53:168178.
Microbiol. 67:13511362.
process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water. Sci.
Technol. 37:135142.
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. 141:27932800.
Hockenbury MR, Jamro ES (1977) Total Dissolved Solids Effects on Microbiological Treatment.
Proceedings of the Industrial Waste Conference:5662.
Hovanec TA, De Longe E (1996) Comparative analysis of aitrifying Bacteria associated with
freshwater and marine aquaria. Appl. Environ. Microbiol. 62:28882896.
Intrasungkha N, Keller J, Blackall LL (1999) Biological nutrient removal efficiency in treatment of
saline wastewater. Wat. Sci. Technol. 39:183190.
Juretschko S, Timmermann G, Schmid M, Schleifer K-H, Pommerening-Roser A, Koops H-P,
Kargi F, Dincer AR (1996) Enhancement of biological treatment performance of saline wastewater
by halophilic bacteria. Bioprocess Eng. 15:5158.
emphasizing cultured species. FEMS Microbiol. Ecology 37:19.
Madoni P, Davoli D, Chierici E (1993) Comparative-analysis of the activated-sludge micro fauna
in several sewage-treatment works. Water Res. 27:148591.
89
Martin-Cereceda M, Serrano S, Guinea A (1996) A comparative study of ciliated protozoa

communities in activated-sludge plants. FEMS Microbiol Ecol. 21:26776.
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.
62:21562162.
Moussa MS, Lubberding HJ, Hooijmans CM, van Loosdrecht MCM, Gijzen HJ (2003a) Improved
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).
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.
36:147156.
Pollice A, Tandoi V, Lestingi C (2002) Influence of aeration and sludge retention time on ammonia
oxidation to nitrite and nitrate. Water Res. 36:25412546.
Pommerening-Rser A, Rath G, Koops H-P (1996) Phylogenetic diversity within the genus
Nitrosomonas. Syst. Appl. Microbiol. 19:344351.
Prosser JI (1986) Nitrification, special publication of the society for general microbiology, Oxford
IRL Press, Volume 20.
Purkhold U, Pommerening-Rser A, Juretschko S, Schmid MC, Koops H, Wagner M (2000)
66:53685382.
Salvad H, Gracia M P, Amig JM (1995) Capability of ciliated protozoa as indicators of effluent
quality in activated-sludge plants. Water Res. 29:104150.
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. 65:36903696.
Schramm A, de Beer D, Gieseke A, Amann R (2000) Microenvironments and distribution of
nitrifying bacteria in a membrane-bound biofilm. Environ. Microbiol. 2:680686.
Tijhuis L, van Loosderecht MCM, Heijnen JJ (1994) Solid retention time in spherical biofilms in a
biofilm airlift suspended reactor. Biotechnol. Bioeng. 44:867879.
van t Riet K, Tramper J (1991) Basic Bioreactor Design. Marcel Dekker, inc. New York.
Wagner M, Rath G, Amann R, Koops H-P, Schleifer KH (1995) In situ identification of ammoniaoxidizing bacteria. Syst. Appl. Microbiol. 18:25164.
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. 34:23144.
90
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; Salvad 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);
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.
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 20C 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 (1Fxi) 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).
5.3.1 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 30C. 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
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
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:
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 510 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 30C 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 NaClCl/L. The respiration activity of the samples treated with salt was measured
using the BOM and expressed as percentage of the non-treated sample.
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.
Standard Methods (APHA 1998). Nitrate-nitrogen was determined using Dionex 4500i
series and Shimadzu C_R5A ion-chromatograph.
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 105C and weighted on a
microbalance. The ash content was calculated after incinerating the dried filter in an oven
for 1 h at 550C. 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 3L of the stock stain solution (prepared by adding 10 L of
SYTO 9 green-fluorescent nucleic acid stain; 10L of propidium iodide and 180 Lof
Mill Q) to 100 L 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
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
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 NO2N was observed in
the reactor operated at lower SRT (SBR30 days), while only 4 mg/L NO2N 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
70
90
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 NaClCl/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 1215% in respiration activity under starvation
condition was observed after exposure to salt. The activity of ammonia and nitrite
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 day1 for the predation rate (bpred),
0.15 day1 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 m) at
SRT 100 days compared to the floc size at SRT 30 days (70 m). The increased floc size
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 NaClCl/L salt solution was used to eliminate the protozoa and metazoa
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
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 32C 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.
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 (510 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 mg1, respectively) were used to
predict proper biomass composition of autotrophs and heterotrophs.

active fraction of the biomass at steady
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
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
166
127
190
153
10
168
129
199
142
10
mgCOD/L
97
121
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
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.25:6338.
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. 10:2148.
Beccari M, DI Pinto AC, Ramadori R, Tomei MC (1992) Effect of dissolved oxygen and diffusion
resistances on nitrification kinetics. Water Res. 26:10991104.
114
Beeftink HH, van der Reijden RTJM, Heijnen JJ (1990) Maintenance requirements: Energy supply
from simultaneous endogenous respiration and substrate consumption. FEMS Microbiol. Ecol.
73:203210.
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 2:179190.
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. 38:245254.
Cronje CL, Beeharry AO, Wentzel MC, Ekama GA (2002) Active biomass in activated sludge
mixed liquor. Water Res. 36:439444.
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. 4:225236.
Curds CR (1971a) A computer-simulation study of predator-prey relationships in a single-stage
continuous-culture system. Water Res. 5:793812.
Curds CR (1971b) Computer simulation study of microbial population dynamics in the activatedsludge process. Water Res. 5:10491066.
Curds CR. (1973) A theoretical study of factors influencing the microbial population dynamics of
the activated sludge process-I. Water Res. 5:12691284.
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.
38:224231.
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:3746.
Gujer W, Larsen TA (1995) The implementation of biokinetics and conservation principles in
ASIM. Wat. Sci. Technol. 31:257266.
process: an innovative method for nitrogen removal from ammonia-rich wastewater. Wat. Sci.
Tech. 37:135142.
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 5:351371.
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.
43:1153116.
Lee NM, Welander T (1994) Influence of predators on nitrification in aerobic biofilm processes.
Wat. Sci. Tech. 29:355363.
Leenen EJTM, Boogert AA, van Lammeren AAM, Tramper J, Wijffels RH (1997) Dynamics of
artificially immobilized Nitrosomonas europaea: Effect of biomass decay. Biotechnol. Bioeng.
55:630641.
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.
25:105123.
Martin-Cereceda M, Serrano S, Guinea A (1996) A comparative study of ciliated protozoa
communities in activated-sludge plants. FEMS Microbiol. Ecol. 21:26776.
Meijer SCF, van Loosdrecht MCM, Heijnen JJ (2001) Metabolic modelling of full-scale biological
nitrogen and phosphorus removing WWTPs. Water Res. 35:27112723.
Madoni P, Davoli D, Chierici E (1993) Comparative-analysis of the activated-sludge microfauna in
several sewage-treatment works. Water Res. 27:148591.
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.
36:859863.
Nowak O, Svardal K, Schweighofer P (1995) The dynamic behaviour of nitrifying activated sludge
systems influenced by inhibiting wastewater compounds. Wat. Sci. Tech. 31:115124.
Ratsak CH, Maarsen KA, Kooijman SALM (1996) Effects of protozoa on carbon mineralization in
activated sludge. Water Res. 30:112.
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. 30:18.
Salvad H, Gracia MP, Amig JM (1995) Capability of ciliated protozoa as indicators of effluent
quality in activated-sludge plants. Water Res. 29:104150.
Schlimme G, Marchiani M, Hanselmann K, Jenni B (1997) Gen transfer between bacteria within
digestive vacuoles of protozoa. FEMS Microbiol. Ecol. 23:239247.
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. 39:129137.
Smolders GJF, van loosdrecht MCM, Heijnen JJ (1994) Stoichiometric model of the aerobic
metabolism of the biological phosphorus removal process. Biotechnol. Bioeng. 44:837848.
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. 65:24712477.
Tijhuis L., van Loosdrecht MCM, Heijnen, JJ (1994) Solid retention time in spherical biofilms in a
biofilm airlift suspended reactor. Biotechnol. Bioeng. 44:867879.
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. 39:107117.
Weinbauer MG, Beckmann C, Hfle 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. 64:50003.
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. 36:460468.
116
APPENDIX 5.1
Process Kinetics and stoichiometry for enriched
culture of nitrifiers in SBR system
117
Appendix 5.2
Stoichiometric and kinetic parameters at 30C
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
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/gNO2N
Yield of ammonia oxidisers Hunik et al (1994)

per NO2_N
YNO2
0.06
g COD/g NO3N 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
day1
maximum growth rate of

ammonia oxidisers
NH4
Hunik et al (1994)
118
1&3
mg O2/L
affinity constant for oxygen This study (for SRT

of ammonia oxidisers
30 &100 days)
mg NH4N/L
affinity constant for

ammonia of ammonia
oxidisers
This study
mNH4
0.35
mg NH4N/g
XACOD.day
maintenance coefficient of
ammonia oxidisers
Tappe et al (1999)
bNH4
0.3
day1
aerobic decay rate of

ammonia oxidisers
Measured according
to Lesouef et al
(1992)
NH4
0.5
anoxic reduction factor for

ammonia oxidisers decay
Siegrist et al (1999)
0.9
day1

nitrite oxidisers
(Hunik et al (1994)
mg O2/L
affinity constant for oxygen This study

of nitrite oxidisers
mg NO2N/L
affinity constant for nitrite

of nitrite oxidisers
This study
mNO2
1.15
mg NO2N/g
XNCOD.day
nitrite oxidisers
Tappe et al (1999)
bNO2
0.2
day1
aerobic decay rate of nitrite Measured according

oxidisers
to Lesouef et al
(1992)
NO2
0.5
anoxic reduction factor for

nitrite oxidisers decay
Siegrist et al (1999)
12
day1

Henze et al (2000)
ASM2d
0.2
mg O2/L
affinity constant for oxygen Henze et al (2000)

of heterotrophic biomass
ASM3
mg COD/L
affinity constant for organic Henze et al (2000)

carbon of heterotrophic
ASM3
biomass
0.5
mg NO3N/L
affinity constant for NO3 of Henze et al (2000)

ASM3
mH
0.12
mg COD/g XH
COD.day
Meijer et al (2001)
0.8
anoxic reduction factor of

heterotrophic growth
Henze et al (2000)
ASM2d
bH
0.8
day1
aerobic decay rate of

Henze et al (2000)
ASM2d
predators
0.2
day1
predation rate
This study
119
0.2
mg O2/L
affinity constant for oxygen This study

of predators
bpredators
0.15
day1
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
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, salinitytypically
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
125
and the nitrification activity, not only in full scale WWTPs, but also in pure and enriched
cultures.
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
7.50.05; T 201C). The pH was maintained at 7.50.05 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 h1] or expressed as a percentage of the activity obtained under
reference conditions (pH 7.5, T 20C).
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
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 (SO42), pH and conductivity were measured in accordance with
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
105C and weighted with a microbalance. The ash content was calculated after
incinerating the dried filter in an oven for 1 h at 550C.
6.3 Results
The different treatment plants represent a wide range of flow (10021000 m3/day),
influent COD (3002175 mg-COD/L), influent TKN (20557 mg-N/L) and salt level
(0.1316 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
(m )
5700
2800
8000
21000
440
843
(C)
14
1418
22
(m
day1)
100
127
COD
(mg/L)
420
2050
2175
301.0
Influent TKN
(mg/L)
40
557
519
20.0
7.6
7.6
7.3
pH
COD
(mg/L)
40
74
117
Effluent TKN
(mg/L)
1.8
12
24
7.5
7.6
7.3
pH
Reactor
MLSS
(mg/L)
4445
4700
8300
9960
MLVSS
(mg/L)
3200
3390
5870
3540
(ms/cm)
10
24
44
(mg/L)
130
3000
8000
16000
(mg/L)
47
68
4000
Conductivity
Cl
SO4

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
128
activities were measured under

standard conditions: pH 7.5 and T
20C.
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)
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
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;
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(T20)).
The specific activity of pure or enriched cultures of nitrifiers under salt-free conditions is
1015 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.
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
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
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
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
25
10 PF
840
120
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
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
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).
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 SO42/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).
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.
134
References
APHA (1998) Standard methods for the examination of water and wastewater, 20th ed. American
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
Bioeng. 84:563571.
Catalan-Sakairi MAB, Yasuda K, Matsumura M. (1996) Nitrogen removal in seawater using
34:267274.
Copp JB, Murphy KL (1995) Estimation of the active nitrifying biomass in activated sludge. Water
Res. 29:18551862.
Daims H, Brhl 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. 22:43444.
43:918.
Microbiol. 67:13511362.
process: an innovative method for nitrogen removal from ammonia-rich wastewater. Water. Sci.
Technol. 37:135142.
Hovanec TA, De Longe E (1996) Comparative analysis of Nitrifying Bacteria associated with
freshwater and marine aquaria. Appl. Environ. Microbiol. 62:28882896.
Juretschko S, Timmermann G, Schmid M, Schleifer K-H, Pommerening-Rser A, Koops H-P,
emphasizing cultured species. FEMS Microbiol. Ecology 37:19.
Kowalchuk GA, Stephen JR (2001) Ammonia-oxidising bacteria: a model for molecular microbial
ecology. Annu. Rev. Microbiol. 55:485529.
135

McCaig AE, Phillips CJ, Stephen JR, Kowalchuk GA, Harvey M (1999) Nitrogen cycling and
community structure of -subgroup ammonia oxidising bacteria within polluted, marine
fishfarm sediments. Appl. Environ. Microbiol. 65:21320.
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).
243.
Prosser JI (1986) Nitrification, special publication of the society for general microbiology, Oxford
IRL Press, Volume 20.
Purkhold U, Pommerening-Rser A, Juretschko S, Schmid MC, Koops H, Wagner M (2000)
66:53685382.
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. 39:5159.
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).
Stephen JR, Chang Y-J, Macnaughton SJ, Kowalchuk GA, Leung KT (1999) Effect of toxic metals
in indigenous soil -subgroup proteobacterium ammonia oxidizer community structure and
protection against toxicity by inoculated metal-resistant bacteria. Appl. Environ. Microbiol.
65:65101.
Wagner M, Rath G, Amann R, Koops H-P, Schleifer KH (1995) In situ identification of ammoniaoxidizing bacteria. Syst. Appl. Microbiol. 18:25164.
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. 31:441449.
Wiesmann U (1994) Biological nitrogen removal from wastewater. Advances in biochemical
Engineering/biotechnology 51:113154.
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 2629, 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

7.2.1 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 (850 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
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 3035%. 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 (850)
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
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), NH4N, NO3N, 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 20C, pH of 7.50.05 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 NKj
488
(515)
7 (11)
gN/m3
Readily
biodegradable
COD
SS
840
972
gCOD/m3
Soluble N
Kj
454
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 (0.1)
gN/m3
Nitrite
SNO2
gN/m3
Nitrate NO3
99 (50)
gN/m3
Nitrate
SNO3
gN/m3
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
gCOD/m3
255
127
g/m3
Ammonia
oxidisers
XNH4
gCOD/m3
2030
570
gCaCO3/m3 Nitrite
oxidisers
XNO2
gCOD/m3
22
22
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
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 1520 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)
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
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 day1 to 0.20 and 0.13 day1 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 1326
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
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 1326 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.
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
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.
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.
34:846858.
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.
18:91114.
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 5:351371.
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. 35:27112723.
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. 31:115124.
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. 39:113120.
Orhon D, Ates E, Ubay Cokgor E (1999a) Modelling of activated sludge for tannery wastewaters.
Water Environment Research, 71:5063.
152
Orhon D, Ubay Cokgor E, Sozen S (1999b). Experimental basis for the hydrolysis of slowly
biodegradable substrate in different wastewaters. Wat. Sci. Tech. 39:8795.
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),
Dbendorf, 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. 45:169176.
van Zuylen HJ (1993) From scientific computation to decision support. Knowledge-based system
6:310.
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
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 bacterin 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 industrien
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
efficinter 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 industrile 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 bacterin 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 protozon en
metazon gedood. Op basis van dit effect is een mathematisch model opgesteld om de
interacties tussen nitrificeerders, heterotrofe bacterin 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 industrile 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.

(Moustafa Samir Moussa) Nitrification in Saline in PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Moustafa Samir Moussa) Nitrification in Saline in PDF

Uploaded by

Copyright:

Available Formats

NITRIFICATION IN SALINE

Nitrification in Saline Industrial

MOUSTAFA SAMIR MOUSSA

This dissertation has been approved by the promotor

Director UNESCO-IHE, Delft

UNESCO-IHE, Delft, promotor

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

Delft University of Technology, promotor

Delft University of Technology

Technical University Mnchen, Germany

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

Samenvatting (Summary in Dutch)

Activated Sludge Model

Biological Oxygen Monitor

Aerobic decay rate of ammonia oxidisers (day1)

Aerobic decay rate of nitrite oxidisers (day1)

Aerobic decay rate of heterotrophic biomass (day1)

Aerobic decay rate of predators (day1)

Concentration of ionic species, i (mole)

Concentration of anion species, i (mole)

Concentration of cation species, i+ (mole)

Chemical Oxygen Demand

Fraction of inert COD generated in biomass lysis

Fluorescent In Situ Hybridisation

Hydraulic Retention Time (day)

Nitrogen Content of XI (g N/g COD)

Nitrogen Content of biomass (g N/g COD)

Mixed Liquor Suspended Solids (mg/L)

Mixed Liquor Volatile Suspended Solids (mg/L)

maintenance coefficient of ammonia oxidisers (mg NH4

maintenance coefficient of nitrite oxidisers (mg NO2

maintenance coefficient of heterotrophic biomass (mg

Monod term for ammonia in bacterial growth

Oxygen Uptake Rate (mg O2/L.h)

The negative logarithm of stoichiometric dissocial constant

volumetric uptake rate of ammonia (mg NH4N/L.h)

volumetric uptake rate of nitrite (mg NO2N/L.h)

maximum volumetric uptake rate of ammonia (mg NH4

maximum volumetric uptake rate of nitrite (mg NO2

Sequencing Batch Reactor (s)

Sludge Retention Time

Concentration of Oxygen (mg O2/L)

Concentration of ammonia (mg NH4N/L)

Concentration of nitrate (mg NO3N/L)

Concentration of nitrite (mg NO2N/L)

Concentration of nitrogen (mg N/L)

Concentration of organic substrate (mg COD/L)

Volatile Fatty Acids

Volatile Suspended Solids

Wastewater Treatment Plant (s)

Concentration of ammonia oxidisers (mg-VSS/L)

Concentration of heterotrophic biomass (mg-VSS/L)

Concentration of predators (mg-VSS/L)

Concentration of particulate inert (mg-VSS/L)

Yield coefficient of ammonia oxidisers per NO2N (g

Yield coefficient of nitrite oxidisers per NO3N (g COD/g

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

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

Charge of anion species, i

Charge of cation species, i+

Anoxic reduction factor for ammonia oxidisers decay

Anoxic reduction factor for nitrite oxidisers decay

Anoxic reduction factor for heterotrophic growth

Temp.C Medium1 Seed2 Adapted used*3