Oct 19, 2014
THE REMOVAL OF NITROGEN COMPOUNDS FROM WASTEWATER
This Page Intentionally Left Blank
Studies in Environmental Science 54
THE REMOVAL OF NITROGEN COMPOUNDS FROM WASTEWATER
by
B. Halling-Sarensen
and
S.E. Jargensen
DFH, lnstitut A Environmental Chemistry Section Universitetsparken 2,2100 Copenhagen 0, Denmark
ELSEVIER Amsterdam - London - N e w York - Tokyo 1993
ELSEVIER SCIENCE PUBLISHERS B.V Molenwerf 1 P.O. Box211,IOOOAE Amsterdam,The Netherlands
ISBN: 0-444-891 52-8
0 1993 Elsevier Science Publishers B.V. All rights reserved.
No part ofthis publication may be reproduced, stored in a retrieval system ortransmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, withoutthe prior written permission of the publisher, Elsevier Science Publishers B.V., Copyright & Permissions Department, P.O. Box 521,1000 AM Amsterdam,The Netherlands.
Special regulationsfor readers in the USA -This publication has been registered with the Copyright Clearance Center Inc. (CCC), Salem, Massachusetts. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA. All other copyright questions, includ- ing photocopying outside of the USA, should be referred to the publisher.
No responsibility isassumed bythe publisher for any injuryand/ordamageto personsor propertyas a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, in- structions or ideascontained in the material herein.
This book is printed on acid-free paper.
Printed in The Netherlands
Studies in Environmental Science
Other volumes in this series
1 2
3
4 5 6
7
8 9
10 11 12 13
14
15 16 17
18
19
20 21 22 23
24 25 26
27 28 29
30 31 32 33
Atmospheric Pollution 1978 edited by M.M. Benarie Air Pollution Reference Measurement Methods and Systems edited by T. Schneider, H.W. de Koning and L.J. Brasser Biogeochemical Cycling of Mineral-Forming Elements edited by P.A. Trudinger and D.J. Swaine Potential Industrial Carcinogens and Mutagens by L. Fish bein Industrial Waste Management by S.E. J~rgensen Trade and Environment: ATheoretical Enquiry by H. Siebert, J. Eichberger, R. Gronych and R. Pethig Field Worker Exposure during Pesticide Application edited by W.F. Tordoir and E.A.H. van Heemstra-Lequin Atmospheric Pollution 1980 edited by M.M. Benarie Energetics and Technology of Biological Elimination of Wastes edited by G. Milazzo Bioengineering, Thermal Physiology and Comfort edited by K. Cena and J.A. Clark Atmospheric Chemistry. Fundamental Aspects by E. MBszaros Water Supply and Health edited by H. van Lelyveld and B.C.J. Zoeteman Man under Vibration. Suffering and Protection edited by G. Bianchi, K.V. Frolov and A. Oledzki Principles of Environmental Science and Technology by S.E. J~rgensen and I. Johnsen Disposal of Radioactive Wastes by Z. Dlou h y Mankind and Energy edited by A. Blanc-Lapierre Quality of Groundwater edited by W. van Duijvenbooden, P. Glasbergen and H. van Lelyveld Education and Safe Handling in Pesticide Application edited by E.A.H. van Heemstra-Lequin and W.F. Tordoir Physicochemical Methodsfor Water and Wastewater Treatment edited by L. Pawlowski Atmospheric Pollution 1982 edited by M.M. Benarie Air Pollution by Nitrogen Oxides edited by T. Schneider and L. Grant Environmental Radioanalysis by H.A. Das, A. Faanhof and H.A. van der Sloot Chemistry for Protection of the Environment edited by L. Pawlowski, A.J. Verdier and W.J. Lacy Determination and Assessment of Pesticide Exposure edited by M. Siewierski The Biosphere: Problems and Solutions edited by T.N. Veziroglu Chemical Events in the Atmosphere and their Impact on the Environment edited by G.B. Marini-Bettolo Fluoride Research 1985 edited by H. Tsunoda and Ming-Ho Yu Algal Biofouling edited by L.V. Evans and K.D. Hoagland Chemistryfor Protection of the Environment 1985 edited by L. Pawlowski, G. Alaerts and W.J. Lacy Acidification and its Policy Implications edited by T. Schneider Teratogens: Chemicals which Cause Birth Defects edited by V. Kolb Meyers Pesticide Chemistry by G. Matolcsy, M. Nddasy and Y. Andriska Principles of Environmental Science and Technology (second revised edition) by S.E. Jprrgensen and I. Johnsen
34
35
36
37 38 39
40 41
42 43
44 45 46 47
48
49 50 51
52
53
Chemistry for Protection of the Environment 1987 edited by L. Pawlowski, E. Mentasti, W.J. Lacyand C. Sarzanini Atmospheric Ozone Research and its Policy Implications edited by T. Schneider, S.D. Lee, G.J.R. Wolters and L.D. Grant Valuation Methods and Policy Making in Environmental Economics edited by H. Folmer and E. van lerland Asbestos in Natural Environment by H. Schreier How to Conquer Air Pollution. A Japanese Experience edited by H. Nishimura Aquatic Bioenvironmental Studies: The Hanford Experience, 1944-1984 by C.D. Becker Radon in the Environment by M. Wilkening Evaluation of Environmental Data for Regulatory and Impact Assessment by S. Ramamoorthy and E. Baddaloo Environmental Biotechnology edited by A. Blazej and V. Privarova Applied Isotope Hydrogeology by F.J. Pearson Jr., W. Balderer, H.H. Loosli, B.E. Lehmann, A. Matter, Ti. Peters, H. Schmassmann and A. Gautschi Highway Pollution edited by R.S. Hamilton and R.M. Harrison Freight Transport andthe Environment edited by M. Kroon, R. Smit and J. van Ham Acidification Research in The Netherlands edited by G.J. Heij and T. Schneider Handbook of Radioactive Contamination and Decontamination by J. Severa and J. Bar Waste Materials in Construction edited by J.J.J.M. Goumans, H.A. van der Sloot and Th.G. Aalbers Statistical Methods in Water Resources by D.R. Helsel and R.M. Hirsch Acidification Research: Evaluation and Policy Applications edited by T. Schneider Biotechniques for Air Pollution Abatement and Odour Control Policies edited by A.J. Dragt and J. van Ham Environmental Science Theory. Concepts and Methods in a One-World, Problem-Oriented Paradigm by W.T. de Groot Chemistry and Biology of Water, Air and Soil. Environmental Aspects edited by J. Tolgyessy
There is an increased interest for nitrogen removal in waste water
treatment all over the world. We have therefore found it useful to give an overview
of the wide spectrum of nitrogen removal processes available today. Part A gives a very brief overview of nitrogen pollution sources, the global
nitrogen cycle and the treatment methods. Part B presents details of all biological methods for nitrogen removal, while Part C treats the physico-chemical nitrogen removal methods. Design examples related to Parts 6 and C are given in
appendices.
The volume is not a textbook written for engineers, but is rather written for a
wide spectrum of environmentalists who would like to have an overview of the
available methods from a biological and chemical point of view. Design equations
are given in the text, but more emphasis has been laid on the profound
understanding of the biological and chemical processes and the basic factors that
influence these processes. Parameters and regression equations for a quantitative
description of these factors and their influence on the key processes are presented
in several tables. This feature makes the volume very useful as a handbook on
nitrogen removal processes.
The authors, Copenhagen, 24 June 1993
vi i
This Page Intentionally Left Blank
TABLE OF CONTENTS
PART A NITROGEN IN THE ENVIRONMENT
CHAPTER 1 NITROGEN COMPOUNDS AS POLLUTANTS S. E. Jarrgensen and B. Halling-Sarrensen
1.1 The Role of Nitrogen in the Environment 1.2 The Nitrogen Cycle 1.3 Sources of Nitrogen Pollution 1.4 The Effect of Nitrogen Discharge 1.5 Treatment Processes for the Removal of Nitrogen 1.6 The Major Processes in the Removal of Nitrogen 1.7 Summary 1.8 Some Useful Definitions
3
3 6 9 12 21 22 25 39
PART B BIOLOGICAL UNIT PROCESSES FOR THE REMOVAL OF NITROGEN 41
CHAPTER 2 BIOLOGICAL NITRIFICATION AND DENlTRlFlCATlON 43 B. Halling-Sarrensen
2.1 Introduction 43 2.2 Classification of Unit Processes in Nitrification
and Denitrification 44 2.3 Terminology Used in Waste Water Treatment 46 2.4 Comparison of the Biofilm and Activated Sludge Unit
Processes 48 2.5 Comparison of the Nitrification Rate for Different Unit
Processes 50 2.6 Conclusions 53
CHAPTER 3 PROCESS CHEMISTRY AND BIOCHEMISTRY OF NITRIFICATION B. Halling-S~rensen
3.1 Introduction 3.2 Nitrification 3.3 The Biochemical Pathway in the Nitrification Process 3.4 The Energy and Synthesis Relationship 3.5 Kinetics of the Nitrification Process 3.6 The Kinetic Expressions for the Nitrification Process 3.7 Relationship Between Growth Rate and Oxidation Rate 3.8 The Influence of Temperature on the Nitrification Rate
55
55 55 56 58 61 61 66 71
ix
3.9 The Influence of Dissolved Oxygen on the Nitrification Rate 3.10 The Influence of pH on the Nitrification Rate 3.1 1 A Kinetic Expression Combining Several Limiting Factors of the
3.12 Bacterial Population Dynamics Applied in the Nitrification
3.13 Effects of Inhibitors on Nitrification
84 87
Nitrification Process 93
Process 95 102
CHAPTER 4 PROCESS CHEMISTRY AND BIOCHEMISTRY OF DENlTRlFlCATlON B. Halling-Smensen
119
4.1 Introduction 119 4.2 Types of Bacteria Accomplishing Denitrification 120 4.3 Biochemical Pathways 122 4.4 Energy and Synthesis Relationship 124 4.5 Alternative Electron Donors and the C/N Relationship 128 4.6 Kinetic Expression for the Denitrification Process 137 4.7 Relationship Between Growth and Removal Rate 138 4.8 Kinetic Constants in the Denitrification Process 138 4.9 The Influence of Oxygen on the Denitrification Rate 139 4.10 The Influence of Temperature on the Denitrification Rate 139 4.1 1 The Influence of Carbon Concentration on the Denitrification Rate 142 4.12 The Influence of pH on the Denitrification Rate 145 4.13 Combined Kinetic Expression for the Denitrification Process 147 4.1 4 Bacterial Population Dynamics for the Denitrification Bacteria 149 4.15 Influence of Toxic Substances on the Denitrification Process 150 4.16 Conclusion 151
CHAPTER 5 ATTACHED GROWTH REACTORS 153 B. Halling-Smensen
5.1 Introduction 153 5.2 The Biofilm 154 5.3 The Development of a Bacterial Biofilm 155 5.4 Modelling the Transport and Reactions within a Biofilm 159 5.5 A Mass-balance Equation for a Biofilm Plant 165 5.6 The Nitrifying Trickling Filters (NTF) 170
5.6.1 The Performance of Trickling Filters 171 5.6.2 Equations for Modelling the Nitrifying Trickling
Filter (NTF) 177 5.6.3 The Application of the Trickling Filter 182 5.6.4 Recent Developments in the Technology of the Nitrifying
Trickling Filters (NTF) 186 5.6.5 Nitrogen Loading Capacity and Removal Efficiency of the
Different NTF-Applications 186 192 5.6.6 Advantages and Disadvantage of the NTF
X
5.7 Rotating Biological Contactors (RBC) 193 5.7.1 The Performance of the RBC 193
5.7.3 The Application of the RBC 206 5.7.4 Recent Development in the RBC Technology 207 5.7.5 Nitrogen Loading Capacity and Removal Efficiency of the
RBC Process 208 5.7.6 Advantages and Disadvantages of the Nitrifying RBC 213
5.8 Submerged Filters 215 5.8.1 Case Study; Simultaneous Nitrification and Denitrification
(SND) as Tertiary Treatment Step, using a Submerged Biofilter of Clinoptilolite 216
5.7.2 Equations for Modelling the RBC Reactor 198
CHAPTER 6 SUSPENDED-CULTURE REACTORS 235 B. Halling-Serrensen
6.1 Activated Sludge Unit Processes 235 6.2 Process Design 237 6.3 Activated-sludge Process Configurations 245 6.4 The Kinetics of the Activated Sludge Process 252 6.5 Modification of Activated Sludge Plants for the Biological
Removal of Nitrogen 252 6.6 Modelling the Activated Sludge Process 256 6.7 Advantages and Disadvantages of the Separate and Combined
Activated Sludge Processes 256
PART C
S. E. Jmgensen PHYSICO-CHEMICAL UNIT PROCESSES FOR THE REMOVAL OF NITROGEN
CHAPTER 7 AIR STRIPPING S. E. Jorgensen
7.1 Physico-chemical Principles of Air Stripping 7.2 Process Variables 7.3 Gas Transfer 7.4 Design of Stripping Tower 7.5 Practical Experience 7.6 Application of Stripping
261
261 265 270 275 288 292
xi
CHAPTER 8 BREAKPOINT-CHLORINATION 295 S. E. J~rgensen
8.1 Principles of Breakpoint-Chlorination 295 8.2 Process Variables 299 8.3 Design of Breakpoint-Chlorination Units 300 8.4 Application of Breakpoint-Chlorination for the Removal of Nitrogen 301
CHAPTER 9 ION-EXCHANGE 305 S. E. J~rgensen
9.1 Principles of Ion Exchange 305 9.2 Process Variables 31 3 9.3 The Sequential and Continuous Ion Exchange Operation 320
334 9.4 Application of Ion Exchange
CHAPTER 10 MEMBRANE PROCESSES S. E. J~rgensen
10.1 Principles of Membrane Processes 10.2 Process Variables 10.3 Design of the Reverse Osmosis Unit 10.4 The Reverse Osmosis System 10.5 Application of Reverse Osmosis and Ultrafiltration
337
337 34 1 348 350 352
CHAPTER 11 PRECIPITATION 355 S. E. Jgrgensen
1 1.1 Principles of Precipitation 355 1 1.2 Process Variables 372 1 1.3 Design of Plants for Precipitation of Nitrogen Compounds 378 11.4 Application of Nitrogen Removal by Precipitation 389
APPENDICES of PART: B Determination of Kinetic Coefficients for Activated Sludge Processes 393
The Stripping Column 40 1 The Ion Exchange Column 403 The Reverse Osmosis Unit 407 The Sedimentation Tank 409
C Design Examples for: 399
References 41 3
index 439
xii
Part A
NlTROG EN IN THE ENVIRONMENT
Nitrogen in the Environment The Nitrogen Cycle Sources of Nitrogen Effects of Nitrogen Processes for Nitrogen Removal
This Page Intentionally Left Blank
1. NITROGEN COMPOUNDS AS POLLUTANTS
1.1. The Role of Nitrogen in The Environment
Nitrogen compounds are becoming increasingly important in waste water
management, because of the many effects that nitrogenous material can have on
the environment. Nitrogen, in its various forms can deplete oxygen due to
nitrification, fertilize aquatic plant growth, exhibit toxicity toward aquatic life, affect chlorine disinfection efficiency and present a public health hazard. These effects
will be reviewed further in Section 1.4.
This volume is about the nitrogen removal processes applied in
environmental technology. A detailed description of all processes, biological as
well as physical-chemical, will be presented, but obviously the selection of
environmental technology must be based not only upon what the technology can offer, but also upon which problems we need to solve. Before we can make the final selection of the proper technology, we need to answer a chain of questions:
1. What are the concentration and form(s) of the pollutants in the
ecosystem in focus? 2. Are these concentrations and forms changed over time? If “yes” due to
what processes? Can these processes be controlled? 3. What are the sources of the pollutants? Which sources are point sources
and which sources are non-point or diffuse sources? 4. What is the geographical distribution of the problem(s)? Are the problems local, regional or global? The answers to this question require in most cases that a local, regional or global cycle of the pollutant is set up.
Figure 1.1. gives a flow chart of a procedure showing how to get from
emission of mass and energy to a solution of the related environmental problems.
Emission is translated into imission and concentration. The effect and impact of a concentration of a compound or energy is found by considering all the chemical, physical and biological processes that take place in the ecosystem. This step will often require the application of ecological models as a management tool, as the
3
processes are interactive and an overview is difficult to obtain without a
synthesizing instrument as modelling; see J~rrgensen (1 988). This evaluation leads
us hopefully to an acceptable ecological solution by the use of ecological
engineering or environmental technology. The former attacks the problem in the ecosystem, which is often needed, when the cause of the problem is coming from
non-point sources; the latter attempts to reduce or dilute the emission at the point
source. This volume focuses on the environmental technological solutions to
nitrogen removal, although the application of biological removal processes in
nature, i.e., in the ecosystems, also will be touched upon, as these processes are in
principle the same processes, such as the biological processes characterizing the
biological nitrogen removal processes in environmental technology.
The three steps in the procedure presented in Fig. 1.1 concerned with
ecosystems are the most complex ones. They require a very comprehensive
ecological and environmental scientific knowledge, which often must be syn-
thesized in a model in order to give applicable answers to the crucial
environmental problems. It would require a second volume of this book to discuss
in detail the nitrogen in the environment, but a brief discussion of these problems
seems necessary to present the frames of the environmental technology available
for nitrogen removal - or to touch upon the problems behind the possible solutions
presented in parts B (removal of nitrogen by biological methods) and C (removal of
nitrogen by physical-chemical methods) of the volume. Section A is a necessary
part of this volume to avoid the separation of “the problem” and “the solution,”
which should always be avoided in environmental management. An integration of
the two sides of environmental issues should always be attempted, and it is
facilitated in the case of nitrogen pollutions by the fact, that many of the biological,
chemical and physical processes used for nitrogen removal in environmental technology are, as mentioned above, the same as the processes that take place in
nature. The following sections of this introductory chapter will consequently focus
on : - the global nitrogen cycles, to understand the global effects of our nitrogen
emissions,
- the regional and local nitrogen cycles and budgets, - the sources of nitrogen emission and their relative quantitative importance - the typical effects of elevated nitrogen concentrations in aquatic ecosys-
4
t
I
I . C
3
Further trans- 0 formation by processes
:: 'a
c
a E
b
Solution by use of eco logial engineering
Solution by use of environmental technoloqy
Figure 1.1. The flow chart illustrates a procedure which can be used to get from emission to the solution of the problem.
5
tems and in portable water. This includes a brief presentation of the toxicity of
nitrogen compounds.
1.2. The Nitrogen Cycle
Figure 1.2 illustrates the global nitrogen cycle. The amount of nitrogen in the
various pools and the transfer flows are mainly based upon the figures from Bolin
and Cook (1983). As seen from Fig. 1.2 many of the numbers are indicated as
ranges due to uncertainty in the estimation. The figures have steadily been
adjusted due to new measurements and new knowledge gained during the last two
decades. Further changes of our knowledge about the nitrogen cycle can be
expected in the coming years.
The cycle is not in balance due to human activity. The production of nitrogen fertilizer the conversion of gaseous nitrogen as dinitrogen (N2) into ammonia and
nitrate, which are deposited in the lithosphere. The major part is washed out to the
hydrosphere, where no major global change of the concentration of nitrogen is
observed, but where the nitrogen together with phosphorus may cause
eutrophication on a local or regional scale; see also Section 1.4.
Another unbalance in the nitrogen cycle is the transfer of nitrogen by
combustion from the lithosphere to the atmosphere. These fluxes would influence
the atmospheric concentrations of inorganic nitrogen radically, if the inorganic nitrogen was not deposited in the hydrosphere and lithosphere. It implies that also
from this source there is a net, diffuse input of nitrogen to the lithosphere and
hydrosphere, contributing to the eutrophication of aquatic ecosystems on a local or
regional scale.
Figure 1.3 shows another important nitrogen cycle, namely in soil and
ground water. All nitrogen compounds and in particular nitrate show an increasing concentration in the root-zone, due to the increasing nitrogen input to the
lithosphere from dry and wet deposition and from industrial fixation; see also Fig.
1.2. A part of this nitrate will leach to the groundwater and an elevated nitrate
concentration in this potable water source is observed as a result of the above
mentioned unbalance in the nitrogen cycle. As the time needed for the nitrate to
reach the ground water table is several decades, the final results of the increasing
nitrogen pollution during the last 25 years have not yet been reflected in the nitrate
6
concentration of ground water. Elevated nitrate concentrations are already a threat to the ground water quality in many industrialized countries due to the increased
ATMOSPHERE
I- -
I
To sediment 40 PEDOSPHERE
brust + sediment 1.9*181
Figure 1.2. The global nitrogen cycle is shown. Values in compartments are in
Pg N, while the fluxes are in Tg N I yr.
nitrogen consumption and pollution over hte last 30-40 years, but we can expect
that the problem will increase very rapidly in the coming years due to the above
mentioned time lag.
7
Figure 1.4 gives the nitrogen cycle in many aquatic ecosystems. The
increasing inflow of nitrate and ammonium to the aquatic ecosystems due to the in-
Evaporation
A
Dentrification
Leaching to the ground water
Figure 1.3. The nitrogen cycle in soil and ground water. Note that the processes
causing the global unbalance are included in the figure: the industrial production of
fertilizers and the wet and dry deposition. These two processes are causing the
nitrate pollution of the ground water sources.
creased production of fertilizers and the increase in nitrogenous emission from human activities in general to the atmosphere and further on to the lithosphere and
atmosphere, accelerate directly the growth of plants. This process, named
eutrophication, may cause several other problems as will be touched upon in the
next sections, dealing with the sources and effects of nitrogen pollution.
a
DENIT 4
17 PHOTO NFIX
Inflow
Outflow
SN
Figure 1.4. The nitrogen cycle in an
uptake of inorganic nitrogen by algae aquatic ecosystem. The processes are: 1)
(and plants), 2) photosynthesis, 3) nitrogen fixation, 4) grazing, 6) predation 5) and 7) loss of undigested matter 8) mortality, 9)
mineralization, 10) and 11) settling, 12) excretion of ammonia, 13) release of
ammonium from sediment, 14) nitrification, 15) 16) and 18) input / output, and 17)
denitrification.
1.3. Sources of Nitrogen Pollution
The abatement of nitrogen pollution must be based upon a knowledge of the
quantities of nitrogen from the various sources. Mass balances must be set up for
ecosystems and for entire regions. Table 1.1 gives an example. The estimated
nitrogen loadings for the San Francisco Bay Basin (from “California “, 1974) are
shown. The. mass balance shows clearly that major problems are rooted in the discharge of waste water and nitrogen from dairies and feedlots. The abatement
should therefore concentrate primarily on these sources of nitrogen emissions.
Table 1.2 gives the nitrogen balance for Denmark, which is a country
characterized by intensive agriculture and many food processing industries. The
balance shows that a comprehensive reduction of the nitrogen emission requires
that the non-point sources are included in the abatement scheme. It is, however, in
most cases more difficult to find good solutions to the reduction of nitrogen emission from non-point sources than from point sources. Some of the
9
ecotechnological methods briefly touched upon in part B may be applied effectively
to reduce these sources, but a wide spectrum of methods, which include planning
and legislation, is needed in practical environmental management to achieve an
acceptable result.
It would be going too far in this context to present these methods in detail in
this context, but it should be mentioned that the following components have been
included in the management of nitrogen pollution in Denmark:
1. Legislation concerning manure application schemes and storage capacity
for surplus manure.
2. Use of wetlands close to sensible aquatic ecosystems.
3. Legislation on the needs for green fields during the winter season.
4. Encouragement of limited use of fertilizers. Were this not be successful,
the imposition of a tax on the use of fertilizers will be considered.
5. Financial support for erection of biogas-plants for treatment of manure
and other animal wastes.
Table 1.1.
Estimated nitrogen loadings for the San Francisco Bay Basin*) Nitrogen source t per year Percent of total
Municipal waste water
Industrial waste water
Dry deposition
Wet deposition
Urban run-off
Non-urban run-off Nitrogen from irrigated
agricultural land
Nitrogen from dairies
and feedlots
26 000
16 000
590
390
1 400
1 900
900
6 000
49
30
1.1
0.8
2.7
3.6
1.7
11.1
Total 53 000 100
*) Source “California”, 1974
10
The methods for nitrogen removal presented in Parts B and C are aimed at the reduction of nitrogen in the effluents from industries and municipalities, i.e., the
point sources. The few ecotechnological methods, mentioned in Part B can be
used for both point sources and non-point sources. These methods are defined as the application of ecologially sound engineering methods for restoration of
ecosystems. They are included in this volume, as they are based on the same principles as the other methods, presented in Part 8. All the methods presented
otherwise can be considered as environmental technological methods.
Table 1.2.
Source Nitrogen loadings for Denmark *)
1000 t N / yr.
Municipal Waste water 24.1
Villages, summer houses
without sewage systems 2.9
Industries 5.3
Run-off 0.8
Ferti I izers 115-130
Animal waste, manure 45- 1 09 Dry and wet .deposition 12-30
Biological nitrogen fixation 10-28
Removed by harvest 115-130
Denitrification 25-43
Run-off, agriculture (difference) 42-1 24
Total loadings of nitrogen 75-1 57
*) Sources: SJVF (1 985) and Miljmtyrelsen (1 984).
11
1.4. The Effect of Nitrogen Discharge
The effects of nitrogen discharge will be mentioned briefly in this section to
be able to relate the methods of nitrogen removal with the expected effects of their
application.
The four major effects are:
1. Fertilization (eutrophication) of aquatic ecosystems
2. Oxygen depletion in aquatic ecosystems.
3. Toxicity to aquatic life.
4. Contamination of ground water by nitrate and its effect on the public
health.
The word eutrophic generally means "nutrient rich." Naumann introduced
in 1919 the concepts of oligotrophy and eutrophy. He distinguished between
oligotrophic lakes containing little planktonic algae and eutrophic lakes containing
much phytoplankton.
The eutrophication of lakes in Europe and North America has grown rapidly during the last few decades due to the increased urbanization and the increased
discharge of nutrients per capita.
The production of fertilizers has grown exponentially in this century as
demonstrated in Fig. 1.5, and the concentration of nutrients in many lakes reflects
the same exponential growth, (AmbUhl, 1969).
The word eutrophication is used increasingly in the sense of the artificial
addition of nutrients, mainly nitrogen and phosphorus, to water. Eutrophication is
generally considered to be undesirable, although it is not always so. The green color of eutrophic lakes makes swimming and boating less safe
due to increased turbidity. Furthermore, from an aesthetic point of view the
chlorophyll concentration should not exceed 100 mg m-3. However, the most
critical effect from an ecological viewpoint is the reduced oxygen content of the
hypolimnion, caused by the decomposition of dead algae. Eutrophic lakes might
show high oxygen concentrations at the surface during the summer, but low
oxygen concentrations in the hypolimnion, which may cause fishkill.
On the other hand an increased nutrient concentration may be profitable for
shallow ponds used for commercial fishing, as the algae directly or indirectly form
food for the fish population.
12
1900 1950 1980 Year
Figure 1 .C.The production of fertilizers (t yr-l), as demonstrated for N and P,O,,
has grown exponentially (the y-axis is logarithmic).
About 16-20 elements are necessary for the growth of freshwater plants, as
shown in Table 1.3, where the relative quantities of essential elements in plant
tissue are shown.
The present concern about eutrophication relates to the rapidly increasing
amounts of phosphorus and nitrogen, which are normally present at relatively low concentrations. Of these two elements phosphorus is often considered the major
cause of eutrophication, as it was formerly the growth-limiting factor for algae in the
majority of lakes but, as demonstrated in Fig. 1.5, its usage has greatly increased during the last decades. Nitrogen is a limiting factor in a number of East African
lakes as a result of the nitrogen depletion of soils by intensive erosion in the past.
Nitrogen may, however, become limiting to growth in lakes and in coastal zones as a result of the tremendous increase in the phosphorus concentration caused by discharge of waste water, which contains relatively more phosphorus than nitrogen. While algae use 4-10 times more nitrogen than phosphorus, waste water
generally contains only 3 times as much nitrogen as phosphorus.
13
Table 1.3.
Average fresh-water plant composition on wet basis
Element Plant content (percentage)
Oxygen 80.5
Hydrogen 9.7
Carbon 6.5
Si I icon 1.3
Nitrogen 0.7
Calcium 0.4
Potassium 0.3
Phosphorus 0.08
Magnesium 0.07
Sulfur 0.06
Chlorine 0.06
Sodium 0.04
Iron 0.02
Boron 0.001
Manganese 0.0007
Zinc 0.0003
Copper 0.0001
Molybdenum 0.00005
Cobalt 0.000002
Nitrogen accumulates in lakes to a lesser extent than phosphorus and a
considerable amount of nitrogen is lost by denitrification (nitrate to gaseous N2).
The growth of phytoplankton is the key process in eutrophication and it is
therefore of great importance to understand the interacting processes regulating its
growth.
Primary production has been measured in great detail in many large lakes.
This process represents the synthesis of organic matter, and can be summarized as follows:
Light + 6C02 + 6H20 = CSH,206 + 602 (1.1)
14
This equation is necessarily a simplification of the complex metabolic
pathway of photosynthesis, which is dependent on sunlight, temperature and the
concentration of nutrients. The composition of phytoplankton is not constant (note
that Table 1.5 only gives an average concentration), but reflects to a certain extent
the chemical composition of the water. If, for example, the phosphorus
concentration is high, the phytoplankton will take up relatively more phosphorus - termed the luxury uptake.
The sequence of events leading to eutrophication often occurs as follows.
Oligotrophic waters often have a N:P ratio of more than or equal to 10, which
means that phosphorus is less abundant relative to the needs of phytoplankton
than nitrogen. If sewage is discharged into the lake the ratio will decrease since,
the N:P ratio for municipal waste water is about 3:l. Consequently, nitrogen will be
less abundant than phosphorus relative to the needs of phytoplankton. Municipal
waste water contains typically 30 mg 1'' N and 10 mg I-' P. In this situation,
however, the best remedy for the excessive algal growth is not necessarily to
remove nitrogen from the sewage, because the mass balance might show that
nitrogen-fixing algae would produce an uncontrollable input of nitrogen into the
lake.
It is necessary to set up a mass balance for the nutrients. This will often
reveal that the input of nitrogen from nitrogen-fixing blue green algae, dry and wet deposition and tributaries is already contributing too much to the mass balance for
any effect to be produced by nitrogen removal from the sewage. On the other hand
the mass balance may reveal that most of the phosphorus input (often more than
95%) comes from the sewage, and so demonstrates that it is better management to
remove phosphorus from the sewage rather than nitrogen. It is, therefore not a
matter of which nutrient is limiting, but which nutrient can most easily be made to
limit the algal growth.
These considerations have implied that the eutrophication process can be
controlled by a reduction in the nutrient budget. For this purpose a number of
eutrophication models have been developed, which take a number of processes
1978), into account. For details, see Jsrgensen (1976), Jsrgensen et al.,
Jsrgensen et al., (1 986) and Jorgensen (1 988).
Generally however, it is possible to conclude that reduction of the
15
eutrophication in aquatic ecosystems requires a solution which is tailored to the
particular case. Some will require reduction in the phosphorus inputs, some in the
nitrogen inputs and some will require reductions in inputs of both nutrients.
Nitrogen reductions seem to be most important for the eutrophication control in
lakes and marine environment during the summer time, while spring run-off often
transfers large amounts of nitrogen to the aquatic environment, making it difficult to control nitrogen as the limiting factor.
Maintenance of a high oxygen concentration in aquatic ecosystems is
crucial for survival of the higher life forms in aquatic ecosystems. At least 5 mg / I is
needed for many fish species. At 20-21 OC this corresponds to 519 = 56%
saturation. The oxygen concentration is influenced by several factors, of which the
most important are the decomposition of organic matter, and the nitrification of
ammonia (ammonium) according to the following process:
Ammonia is formed by decomposition of organic matter. Proteins and other
nitrogenous organic matter are decomposed to simpler organic molecules such as
amino acids, which again are decomposed to ammonia. Urea and uric acid, the
waste products from animals, are also broken down to ammonia. Nitrifying
microorganisms can use ammonia as an energy source, as the oxidation of ammonia is an energy-producing process. This decomposition chain is illustrated in Fig. 1.6, where it can be seen that the free energy (chemical energy) is
decreased throughout the chain.
The nitrification process can be described by the following first order kinetic
expression:
dN = - K n * t dt
(1.3)
16
or
where
Nt = concentration of ammonium at time = t
NO = concentration of ammonium at time = 0
Kn = rate constant, nitrification
L Proteins w I Amino acids I> 1
I I Urea,uric acid y
1 1 44 Ammonia Ammonia
1 Nitrite - e Nitrate
Figure 1.6. Decomposition chain: from protein to nitrate.
Nt and No may here be expressed by the oxygen consumption
corresponding to the ammonium concentration. Values for K, and No are given for
some characteristic cases in Table 1.4. Kn is dependent on the temperature as
illustrated in the following expression:
17
Kn at T = ( Knat 2OoC) ' K,(T-20) (1.6)
where T = the temperature (OC), KT = a constant in the interval 1.06-1.08.
Table 1. 4.
Characteristic values, Kn, and No (20 OC)
Kn (1 / 24h) No
Municipal waste water 0.15-0.25 80-130 Mechanical-treated muni-
cipal waste water 0.10-0.25 70-1 20
Biological-treated muni-
cipal waste water
Potable water
River water
0.05-0.20 60- 1 20
0.05 0- 1
0.05-0.10 0-2
The relation between ammonium concentration and oxygen consumption
according to (1.2) may be calculated as (2 * 32)/14 = 4.6 mg 0, per mg NH,+ - N,
but due to bacterial assimilation of ammonia this ratio is reduced to 4.3 mg 0, per
mg NH,+- N in practice.
It is easy to see from the values of ammonium nitrogen or total nitrogen in municipal waste water that the oxygen consumption for nitrification is significant. If
a total nitrogen concentration of 28 mg N / I is presumed, the oxygen consumption
for nitrification becomes 128 mg / I , which may be compared with the BOD5 of
municipal waste water on about 200 - 250 mg / I . The growth of nitrifying
microorganisms is, however, relatively slow, which implies that the nitrification is
not completed in aquatic ecosystems with short retention times. Ecological models
(Jprrgensen, 1988 and Jprrgensen and Johnsen, 1989) can be used to characterize the role of the oxygen depletion caused by nitrification and therefore the
18
consequences for the aquatic life of nitrifying ammonium in waste water before
discharge. The conclusion will, however, generally be that nitrification of municipal
waste water is required for all discharge into inland water ecosystems. Many industrialized countries have therefore introduced an effluent standard for ammonium and organic nitrogen concentrations.
While nutrients are necessary for plant growth, they may produce a
deterioration in life conditions for other forms of life. Ammonia is extremely toxic to fish, while ammonium, the ionized form is harmless. As the relation
between ammonium and ammonia is dependent on pH: (see also Section 7.1)
NH4+ = NH3 + H+ (1.7)
where pK = -log K and K = equilibrium constant for process (1.7).
The pH value as well as the total concentration of ammonium and ammonia
is thus important. This is demonstrated in Table 1.5. This implies that the situation is very critical in many hypereutrophic lakes during the summer, when photosynthesis
is most pronounced, as the pH increases when the acidic component CO, is
removed or reduced by this process. The annual variations of pH in a
hypereutrophic lake are shown in Fig. 1.7. pK is about 9.24 - 9.30 in distilled water
at 18 - 25OC, but increases with increasing salinity. It implies that the
concentrations shown in Table 1.5 are higher in sea water.
It is a clear conclusion from these considerations that ammonium discharge
into aquatic ecosystems, in particular inland waters, is not desirable and that municipal waste water therefore must be nitrified before discharge.
The pubic health hazard is associated with nitrate in groundwater, which occur
due to leaching of ritrate; see Fig. 1.3. Nitrate in drinking water is associated with
methemoglobinemia, which affects infants less than three months, because of their
lack of an enzyme capable of oxidizing nitrite.
19
Table 1.5
Concentrations of ammonium nitrogen (ammonium + ammonia), in mg per.1, which contains an unionized ammonia concentration of 0.025
mg NH3 / I at various pH and temperatures
O C' pH =7.0 pH =7.5 pH =8.0 pH =8.5 pH =9.0 pH =9.5
5 19.6 6.3 2 0.65 0.22 0.088
10 12.4 4.3 1.37 0.45 0.16 0.068
15 9.4 5.9 0.93 0.31 0.12 0.054
20 6.3 2 0.65 0.22 0.088 0.045
25 4.4 1.43 0.47 0.17 0.069 0.039
30 3.1 1 0.33 0.12 0.056 0.035
9.5
0.5
7 .5 ' 4 7 10 1 2 4
Month
Figure 1.7. The seasonal variation in pH in a hypereutrophic lake (Lake Glumsa, Denmark).
20
When water with a high concentration of nitrate is used in preparing infant
formulas, nitrate is reduced to nitrite in the stomach after ingestion. The nitrites
react with hemoglobin in the blood to form methemoglobin, which is incapable of carrying oxygen in comparison to hemoglobin. The result is suffocation
accompanied by bluish tinge to the skin, which explains the use of the term “blue
babies” in conjunction with methemoglobinemia. From 1945-1 975 about 2000 cases of methemoglobinemia were reported in
the US. and Europe with a mortality rate of 7-8%. Because of the difficulties in
diagnosing the disease and because no reporting is required, the actual incidence
may be many times higher (Kaufman, 1974).
WHO and most countries have set up standards for nitrate in drinking water.
Typical standards are: US. 10 mg nitrate- N I I and in most European countries 30 - 100 mg nitrate / I.
1.5 Treatment Processes for the Removal of Nitrogen In the past several years a number of processes have been developed with the
specific purpose of transforming nitrogen compounds for removing nitrogen from waste
water.
The processes considered in this book are presented as follows; in part 6, the
biological removal methods, nitrification and denitrification; and in part C, the physical
and chemical methods, Stripping, Break-point Chlorination, Ion Exchange, Membrane
Processes and Precipitation.
In determining which method is most suitable for a particular application, the
following aspects must be considered:
1) Form and concentration of the nitrogen compounds in the process influent.
2) The required effluent quality of the waste water.
3) Other treatment processes to be applied for the removal of other compounds.
4) The construction and running costs for the process.
5) The reliability of the process.
6) The flexibility of the process.
As an short introduction to parts B and C of this text book, the follwhg section will, present a brief description of the various nitrogen removal processes des
cribed. The process characteristics, compound selectivity, and normal range of
efficiency are presented.
21
1.6 The Major Processes in the Removal of Nitrogen The major processes considered in the removal of nitrogen in this text on for
are: Biological nitrification and denitrification (Part B) and Stripping, Break-Point
Chlorination, Ion Exchange, Membrane Processes and Precipitation (Part C). These
processes are technically and economically the most suitable at the present time.
Biological nitrification and denitrification The principal effect of the nitrification process is to transform ammonia-nitrogen
into nitrate by the use of nitrifying bacteria under aerobic conditions. Denitrification
converts nitrate to nitrogen gas by use of denitrifying bacteria, under anoxic conditions.
The efficiency of the nitrification process depends on the extent to which organic
nitrogen is transformed into ammonia-nitrogen. Chapters 3 and 4 present, in detail, the
different factors governing the nitrification and denitrification processes. Nitrification can
be carried out in conjunction with secondary treatment (combined oxidation of organic
material and nitrification) or as a tertiary treatment (seperate stage nitrification ) see
Chapters 5 and 6. In both cases, either attached-growth reactors or suspended-growth
processes can be used. Denitrification can also be carried out in either attached or
suspended growth reactors. For the denitrification process to be carried out, a carbon
source and an anoxic environment are required. Chapter 5 explains the biofilm theory
used in the attached-growth technology and shows the application of some of the most
frequently used attached-growth processes: trickling filter, rotating biological contactor
and submerged filters. The application of the submerged filter is mainly described as
a case study on the use of clinoptilolite as a submerged biobed, for the simultaneous
nitrification and denitrification processes. Chapter 6 shows the practical use of the
activated sludge process.
The overall removal efficiency in a nitrification and denitrification plant ranges
from 70 to 95 per cent for tertiary treatment, and down to 10-20 per cent for secondary
treatment. The costs of attached-growth biological removal plants are moderate
compared with activated sludge plants.
22
Air stripping The stripping process (Chapter 7) is used to remove volatile gases such as
hydrogen sulfide, hydrogen cyanide and ammonia by blowing air through the waste
water. The process is, therefore, to be considered as a transfer of a compound from
a liquid phase to a gas phase. The basic principle of this process of nitrogen removal
is illustrated in Figure 7.1.
The rate at which ammonia can be removed by air stripping is highly
dependent on pH, because the exchange between the two forms, ammonium which
is the ion form, and ammonia, which is a highly water soluble gas, is an acid base
reaction. High efficiency in ammonia removal requires adjustment of the pH to about
11 .O prior to the stripping process.
The principal problems associated with ammonia stripping are its inefficiency
in cold weather, required shut down during freezing conditions, and the formation of
calcium carbonate in the air stripping tower.
The best practical results are achieved by the use of countercurrent packed
towers (0degaard 1988). As the amount of air needed is roughly independent of the
ammonia concentration, the cost per kg of ammonia removed is much lower at high
ammonia concentrations. Stripping is, therefore, most attractive for industrial waste
water with a high concentration of ammonium.
Break-Point Chlorination Breakpoint chlorination is accomplished by addition of chlorine to the waste
stream in an amount sufficient to oxidize ammonia-nitrogen into nitrogen gas (see
Chapter 8). In practice, approximately 9-10 mg/l of chlorine is required for every 1 mg/l
of ammonia-nitrogen. In addition, the acidity produced by the process (equation 8.2)
must be neutralized. The chemicals add greatly to the total dissolved solids and result
in substantial operating expenses.
1) By using sufficient chlorine it is possible to obtain effluents reduced in ammonia
concentration to near zero.
2) The low spatial requirement makes it particularly suitable for certain applications,
including addition to an existing facility, where nitrogen removal is required, but space
is limited. Nitrite and nitrate are not removed by this method, which is a major
disadvantage.
The method has, however, two advantages:
23
Ion Exchange Ion exchange is a process in which ions on the surface of a solid are
exchanged for ions of a similar charge in a solution with which the solid is in contact
(Chapter 9). When all the exchange sites have been replaced, the resin must be
regenerated.
Both natural solids, such as the natural clay mineral clinoptilolite, and synthetic
ion exchange, can be used in the removal of ammonium ions.
pH control is crucial in the ion exchange process, as the form of the ion
exchanger is dependent on the pH, see equations (9.1) to (9.3), unless the ion
exchanger is a strong acid or base, and also because the form of the ions to be taken
up is dependent on pH. The optimum ammonium exchange by clinoptilolite occurs
within an influent pH range of 4 to 8. If the pH drops below this range, hydrogen ions
begin to compete with ammonium for the available ion exchange capacity. As the pH
increases above 8, a shift in the ammonia-ammonium equilibrium toward ammonia
begins. Consequently, any operation outside the pH range 4 to 8 results in a decrease
in the exchange capacity. Neither ammonia, nitrate or nitrite or organic nitrogen can
be bound to clinoptilolite.
Ion Exchange is very effective in removing ammonium from waste water, but
is, however, not a very attractive treatment method for removal of high ammoinum
concentrations. This is because the regeneration becomes more frequent. The
operational costs, therefore, become high due to the elution frequency. Using
clinoptilolite clay as matrix in a submerged bio-bed as presented in Chapter 5,
diminishes this problem because the micro-organisms (nitrifying bacteria) regenerate
the ion exchanger. A combination of ion-exchange and nitrification seems, therefore,
to be attractive, as presented in Chapter 5.
Membrane Processes Membrane separation, electrodialysis, reverse osmosis, ultrafiltration and other
such processes play an increasingly important role in the treatment of waste water
(Chapter 10).
A membrane is defined as a phase that can act as a barrier between other
phases. It can be a solid, a solvent-swollen gel, or even a liquid.
Osmosis is defined as the spontaneous transport of a solvent from a diluted
solution to a concentrated solution across a semi-permeable membrane. At a certain
24
pressure, the so-called osmotic pressure, equilibrium is reached. The osmotic pressure
varies with the concentration and the temperature, and depends on the properties of
the solution.
Nitrogen compounds treated in such systems are mainly in the form of
ammonium or nitrate. Electrodialysis can be expected to remove about 40 per cent of
these forms; in reverse osmosis, 80 per cent.
Today the application of membrane techniques is still limited, but waste water
engineers and scientists in the field of membrane processes expect a rapid growth in
the use of these technologies in the very near future.
Precipitation Precipitation, in a strictly chemical sense, is the transition of a substance from
the dissolved state to the non-dissolved state by the addition of other reagents that
lead to the formation of precipitates.
Most nitrogen compounds are, unfortunately, readily dissolved in water, which
implies that precipitation cannot be used as an easy solution to the problem of nitrogen
removal. Nitrogen removal by the use of precipitation may, however, be carried out by
the two processes shown as equations (1 1.33) and (1 1.34) in Chapter 1 1. The nitrogen
needs to be in form of either proteins or ammonium.
The application of precipitation in the removal of nitrogen requires a three-step
plant. Addition of chemicals is the first step. The second step is flocculation and as the
third step follows some sort of separation process to separate the suspended matter
(precipitate) from the clear water phase.
1.7 Summary Table 1.6 summarizes the effect, advantages and disadvantages of the various
processes presented in this volume for the removal of nitrogen from waste water. The
effect that each process has on each of the three major forms of nitrogen, organic
nitrogen, ammonium and nitrate are shown. Average removal percentages which can
be expected from the different processes are also indicated.
Table 1.7 shows an estimation of costs for the different processes compared
with the efficiency. The processes are divided into three categories; expensive,
moderate and low cost processes, because it is difficult to estimate exact figures. Also,
the efficiency is divided into low, medium and high content of nitrogen in the influent
.25
waste water.
The overall removal in a particular treatment plant will depend on the types of
unit processes employed and their relation to each other.
In general, the reliability of a given treatment process is higher for the physico-
chemical treatment processes than for the biological treatment processes.
On the other hand, costs are generally higher for the physico-chemical
methods than for the biological methods. It is, therefore, important to find a balance
between costs and reliability for the process, used for each of the types of waste
water. This can only be found by conducting pilot-plant studies of the specific waste
water before deciding which application is to be used.
26
Table 1.6 The effect, advantages and disadvantages of various treatment processes on nitrogen compounds
Effect on nitrogen form Treatment process
Organic N NHflH4+ NO;
Removal Process Process Reference of total Advantages Disadvantages Chapter nitrogen entering process %
Biological treatment Processes
Attached-growth processes
- Nitrification (separate stage) Limited effect +NO,' No effect
Nitrification (combined oxidation and nitrdication) Limited effect +NO,- No effect
70-90
5-20
Good protection against most toxicants. Stable operation. Stability not linked to secon- dary clarifier as organisms are attached to media.
Combined treat- ment of carbon and ammonia in a single stage not linked to a secondary clari- fier as biomass attached to media
Greater number of 5 unit processes required than for combined carbon oxidation and nitri- fication.
No protection against 5 toxicants. Only mode- rate stability of ope- ration. Cold weather operation impracticable in most cares.
Table 1.6 (continued)
Effect on nitrogen form Removal Process Process Reference Treatment process of total Advantages Disadvantages Chapter
Organic N NHflH4' NO, nitrogen entering
process %
Denitriication using methanol No effect No effect 80-90% removal 70-95 Denitrification following a nitrii- rapid; demonstra- cation stage. ted stability
of operation; stabi- lity not linked to clarifier as orga- nisms on media. High degree of nitrogen removal possible.
IU Simultaneous nitri- O3 fication and denitri-
fication (SND). No effect +NO3- 4 2 60-80 Conversion of ammonium to gaseous nitrogen. Rapid nitrogen removal compared with suspended cultures. Low space required for application.
Methanol required. 5 Greater number of unit processes re- quired for nitrifi- catioddenitrifica- tion than in combined systems.
Still only possible on pilot plant scale. Fluctuation in stability. Very sensitke to high BOD, in influent.
5
Table 1.6 (continued)
Effect on nitrogen form
Organic N NH#VH4+ NO; Treatment process
Process Refeyce Removal Process of total advantages disadvantages Chapter nitrogen entering process %
Combined carbon oxidation No effect No effect 80-90% removal nitnbation/denitnhcabon in
suspended-grovdh reactor using endogenous carbon source
Combined carbon oxidation nitnfication/denitnfircabon in wspendedgrowlh reactor using waste water carbon source
No effect No effect 80-90% removal
5-20 No methand required; lesser number of unit processes required; better control of Ma- mentous organisms in activated-sludge pro- cess possible; single basin c a n be used; adaptable to sequen- cing batch reactor: process can be adap- ted to include b o b -
val. cal phosphorus remo-
5-20 No methand required. lesser number of unit processes required. better control of fila- mentous organisms in activated-sludge pro- cess possible. single basn c a n be used, adaptable to sequen- ung batch reactor; process can be adap- ted to include bological phosphorus removal
Denitnficabon occurs at 6 very slow rates, longer detenbon bme and much larger struclures required than methanol-based system, stability of opera- bon linked to clanfmr for biomass retum. dtfflcult to opbrnize nitnficabon and denitnfication separately biomass requires sufficient dtssolved-oxygen level for nitnfication to occuc less nitrogen removal than methand based system
6 Denitnftcabon occurs at slow rates, longer detenbon bme and larger structures required than methanol- based system. stability of operabon linked to danfier for biomass return. difficult to opbmize nitnficatm and denitnfication separately biomass requires sufficient dissolved-oxygen level for nitnfication to occur; less nitrogen removal than methanol-based system
Table 1.6 (continued) -
Effect on nitrogen form
Organic N NH3m(H4+ NO; Treatment process
Removal Process Process Reference of total advantages disadvantages Chapter nitrogen entering process %
Suspenced-growth denitnfi- No effect No effect 80-90% cients using methanol removal
stage followng a nitnhcabon
0 0
Bactenal assmilation No effect 40-70% Slight removal
70-95 Denitnfication rapid. Methanol required stability 6 small structures requir- ed demonstrated clanfier for biomass return stability of operation few limitations in treatment sequence nitnficationldenitnfcation options. exess metha- no1 oxidation step can be easily incorporated each process in system can be separately opbmized. high degree of nitrogen removal possible
of operation linked to
greater number of unit processes required for
than in combined systems
30-70 3
Tabel 1.6 (continued)
Effect on nitrogen form Removal Process Process Reference Treatment process of total Advantages Disadvantages Chapter
Organic N NH3/1vH4+ NO3 nitfugen entering process %
Denitrification using methanol No effect No effect 80-90% removal 70-95 Denitriiication following a nitrlica- rapid; demonstra- cation stage. ted stability
of operation; stabi- lity not linked to clarifier as orga- nisms on media. High degree of nitrogen removal possible.
Simultaneous nitri- fication and denitri-
fication (SND). No effect -:N!13. +NZ 60-80 Conversion of ammonium to gaseous nitrogen. Rapid nitrogen removal compared with suspended cultures. Low space required for application.
Methanol required. 5 Greater number of unit processes re- quired for nitrifi- catioddenitrfi- tion than in combined systems.
Still only possible on pilot plant scale. Fluctuation in stabi- Very sensitive of high BOD, in influent.
5
Table 1.6 (continued)
Effect on .nitrogen form
Organic N NH#VH4* NO; Treatment process
Removal Precess PlrrceSS Reference of total advantages disadvantages Chapter nittugen entering process %
Suspended growth processes
-Nitrification (separate stage) Limited effect -+NO; No effect 70-90
w Iu
-Nitrification (combined oxidation and nitnfication). Limited effect +NO3- No effect 5-20
Good protection Sludge inventory 6 against most toxi- requires careful cants. Stable opera- control when BOD5TTKN tion. Low effluent ratio is low. Stability ammonia possible. of operation linked to
operation of secondary clarifier for biomass return.
Combined treatment No protection against of carbon and toxicants. Only mode- ammonia in a single rate stability linked stage. Inventory to operation of secon- control of mixed- dary clarifier for bio- liquor due to high mass return. Large BOD5TTKN ratio. reactors required in
cold weather.
6
Table 1.6 (continued)
Effect on nittvgen form Truatment prvcess
Organic N NHflH,+ NO3
Removal Pmcess PlVCeSS Retbrunce of total advantages disadvantage Chapter nittvgen entering process %
0 0
Physical and chemi- cal treatment pro-
Air stripping (Air)
cBSS8S.
Noeffect 60-95% No effect removal
50-90 Process can be con- trolled for selected am- monia removals. Most applicable 1 required seasonally in wmbi- nabn with lime system for ~ ~ O S ~ ~ O N S remo- val. Process may be able to meet total nitro- Qen standards. Not sensitive to toxic sub- stances.
Process is senitive to tem- perature. Ammonia solubi- lity increases with lowr temperatures. Air require- ments also vary. Fogging and king OQxr in cold weather. Ammonia readon with sulphur dioxide may caw air pollution problems. Process usually requires lime for pH control. thereby increasing treabnent cost and limsrelated operating and maintenance pr0- M a s . Calbonate scaling of paw and piping. Potential noise and *she- Ik problems.
7
Table 1.6 (continued)
Effect on nitrogen form Treatment process
Organic N NH+H4+ NO,
Renwval Process Prvcess Reference of total advantages disadvantages Chapter nitrogen entering process %
0 P
Break-point Uncertain 90-100% No effect 80-95 With pmper conhol. Bw ammonia nibvgen can chlorination removal be oxidized. Pmcess can be used fobwing other nitrogen removal processes for fine hr- ning of nitrogen remo- Val. Concunenl effluent dsinfeclion. Limited opace require- ment. Not sens~live to toxic substances and temperalure. Low capital costs. Maptable to existing faalii.
May produce high chlorine RMiduals mat am toxic to aquatic organisms. Wastematar contains a variety d chbrine-deman- dng substances which hcmase cos1 d huabnent. Pmcess is sensilive to pH. rvhichalfedsdosagers quirements. Trihabmechane fonnalian may impacl cpalily of water supplies. Mdition of chbri- ne raises effluent TDS. Process may not be& to meettotdnitrogenstan- dads. C~NM contd d pH to avoid.for- malion of nibqlen hich- bMa ga%. w i r e s him skilled opetator.
8
Table 1.6 (continued)
Effect on nitrogen form Removal Pmcess ProCeSS Reference Treatment pmcess of total advantages disadvantages Chapter
Organic N NHflH4* NO; nitrogen entering process %
Ion-Exchange
- Ammonium
- Nitrate
Slight, 80-97% No effect uncertain removal
Slight Slight 75-90 effect eff eci
Can be used where dimak conditions inhibit biological nihifi- cation and where Ptringent effluent standards are required. Produces a relalively low TDS effluent. Produces a redaimable product (aqueous ammonia). P- may be able to meet total nitmgen stan- dards. Ease of product qLIaMy control.
70-95 Organic matter in eflluent 9 from biobgcal tmatment can cause resin binding. Pre-treatment by filtration is usually required to prevent the build up of excessive headbss due to suspended solids accumulation. H i h concentration of other cations will reduce ammo- nia remwal capability. Regeneration recovery may
another unit process (e.9.. as stripping). Hgh capital and operating costs. Rqeneraticm pm- ducts must be disposed of. Requires hghly skillled operator.
require me addtion of
Table 1.6 (continued)
Effect on nitrogen form Treatment process
Organic N NHpH4+ NO;
Removal Process Pmcess Reference of total advantages disadvantages Chapter nitrogen entering process %
Membran processes
Electrodialysis 100% of 30-500/0 30-50% suspended removed removed organic N removed
40-50 High degree of nitrogen Chemical precipitaion of 10 removal. salts with low solubili on Removes all forms of the membrane surface. nihogen. dmng of the membrane
by the residual c d l o i organic matter in waste water emuents usually about 10 per cent of the feed volume. is required to wash h e membrane conti- now.
Table 1.6 (continued)
Effect on nitrogen form Treatment process
Organic N N H f l H 4 + NO3
Removal Process Pnrcess Refemnce of total advantages disadvantages Chapter nitrogen entering process %
H i amount of nitro- Membrane elements in the 10 gen removed.
removed removed removed Removes all forms of be fouled by colloidal nitrogen. matter. Pretmabnent of a
semndary efRuent by chemical clarification and some sort of filtration is usually necesary. Iron and manganese in influent can pmv& decm- ased scaling potential. Regular cleaning of mem- branenecessary.
reverse osmosis unit can Reverse Osmosis 60-90% 60-90% 60-90% 80-90
Precipitation 50-70% Slight Slight effect 20-30 removed eff ecl
Resub in net increase in total dssotved sol& of emuent Large amount of sludge requiring treatment Only organic nitrogen can be NmOVed.
Partly adapted from: EPA (1975), Metcalf and Eddy (1991). WPCF Nutrient Control Manual (1983), Weston (1984).
Table 1.7. The building and running costs of various treatment processes compared with efficiency and reliability of the process. The building and running costs are indicated as expensive, moderate or low.
COST pr. P.E
High
-
Medium
-
Low
Activated sludge Membrane Processes
Trickling filters Ion Exchange
Submerged filters Trickling filters
Rotating Rotating Biological Contactors (RBC) Contactors (RBC)
Membrane processes
Activated sludge Trickling filters Stripping Ion Exchange Precipitation
Activated sludge
(Precipitation) Submerged filters (Stripping)
High I
Medium I
Low
NITROGEN CONTENT IN WASTE WATER
P.E = Personal Equivalent
38
1.8 Some Useful Definitions
the following definitions.
To understand the concept of biological treatment processes, it will be helpful to know
Aerobic processes are biological treatment processes that occur in the presence of oxygen.
Anaerobic processes are biological treatment processes that occur in the absence of
oxygen.
Anoxic denitrification is the process by which nitrate-nitrogen is converted biologically
into nitrogen gas in the absence of oxygen. This process is also known as anaerobic
denitrification.
Biologicalnutrient removalis the term applied to the removal of nitrogen and phosphorus
in the biological treatment processes.
Nitrification is the biological process by which ammonia is converted first to nitrite and
then to nitrate.
Denitrification is the biological process by which nitrate is converted into nitrogen gas.
Substrate is the term for the organic matter or nutrients that are converted during the
biological treatment or that may be limiting in the biological treatment.
Suspendeocgrowlhprocessesarethe bologlcal treatment pmcesses in which the microorganisms
responsible for the conversion of the organic matter or other constituents in the waste
water to gases and cell tissue are maintained in suspension within the liquid.
Atfachecfgmwth p m s s e s are the biological treatment processes in which the microorganisms
responsible for the conversion of the organic matter or other constituents in the waste
water to gases and cell tissue are attached to some inert medium such as rocks, slag,
or specially designed ceramic or plastic materials. Attached-growth treatment processes
are also known as fixed-film processes.
SOD, (Siological oxygen demand): The determination of the biochemical oxygen demand
(BOD) is an empirical test in which standardized laboratory procedures are used to determine
the relative oxygen requirements of waste water, effluents, and polluted waters during
5 days. The test measures the oxygen required for the biochemical degradation of organic
material and the oxygen used to oxidize inorganic material such as sulphides and iron.
It may also be used to oxidize reduced forms of nitrogen unless their oxidation is prevented
by an inhibitor. The method consists of placing a sample in a full, airtight bottle and incubating
the bottle under specified conditions for a specific time. Dissolved oxygen (DO) is measured
initially and after the incubation. The BOD is computed from the difference between initial
39
and final DO.
COD (chemical oxygen demand): The chemical oxygen demand (COD) is a measure
of the oxygen equivalent of the organic matter content of a sample that is susceptible
to oxidation by a strong chemical oxidant. For samples from a specific source COD can
be related empirically to BOD, organic carbon or organic matter content.
Ammonia (NH,) is a gas that is very soluble in water. Ammonia is a base.
Ammonia is produced in nature when any nitrogen-containing organic material decomposes
in the absence of air.
Ammonia is a colourless gas with an irritating odour. Gaseous ammonia has a vapor
pressure of about 10 atmospheres at 25OC, and is readily liquefied, giving a colourless
liquid that boils at minus 33°C.
Ammonium is a weak base, ammonia readily accepts protons from acids and hydronium
ions, forming salts of the ammonium ion (NH4+).
Ammonium-nitrogen or ammonium-N: both terms are used to indicate that the nitrogen
is calculated as 1 mole of ammonium-N equal to 14 g of ammonium-N. It convinient
to calculate from one nitrogen form to another using this term.
Nitrate: (NO,-) Is a substrate for the denitrification process.
Nitrate-N: as for ammonium-N
Nitrite: (NO;) In the gas phase this exists in equilibrium with the colourless dimer N,04.
When it dissolves in water, nitrogen dioxide disproportionates and forms nitric acid.
It is a component of the nitrification process.
Nitrite-N: as for ammonium-N
40
PART B
BIOLOGICAL UNIT PROCESSES FOR THE REMOVAL OF NITROGEN
NITRIFICATION
DENlTRlFlCATlON
ATTACHED GROWTH PROCESSES
TRICKLING FILTER
ROTATING BIOLOGICAL CONTACTER
SUBMERGED FILTERS
SUSPENDED GROWTH PROCESSES
ACTIVATED SLUDGE PROCESSES
This Page Intentionally Left Blank
2. BIOLOGICAL NITRIFICATION AND DENlTRlFlCATlON
2.1 Introduction This chapter aims to give a broad overview of the biological nitrification and denitriification
systems and to compare the different unit processes explained in detail in later chapters.
This should facilitate the understanding of the following chapters 3-6, dealing with the
biological unit processes.
The contents of this chapter may be summarized as follows:
1) Classification of the different nitrification and denitrification unit processes
(section 2.2).
2) The terminology used in the basic waste water treatment (section 2.3).
3) Comparison of the biofilm (attached-growth) and the activated sludge (suspended-
growth) unit processes (section 2.4).
4) Comparison of the nitrification rate for the unit processes described in later
chapters (section 2.5).
The removal of nitrogen by biological nitrification and denitrification is a two-step
process. In the first step (nitrification) ammonia is converted aerobically to nitrate (NO,).
In the second step (denitrification) nitrates are converted to N20 or nitrogen gas (N2)
under anoxic conditions.
Nitrification is an autotrophic process which means that the energy for bacterial growth
is derived from the oxidation of nitrogen compounds, primarily ammonia. In contrast
to heterotrophs, nitrifiers use carbon dioxide as a carbon source rather than organic
carbon for the synthesis of new cells. Nitrifier cell-yield per unit of substrate metabolized
is many times smaller than the cell yield for heterotrophs and denitrifier, see Table 2.3.
As will be described in Chapter 3 the nitrification process is a two-step process involving
two genera of microorganisms, Nitrosomoms and Nifrobacfer. In the first step, ammonium
is converted to nitrite; in the second step, nitrite is converted to nitrate. The conversion
processes are outlined in Section 3.4.
43
Chapter 4 describes how the denitrification can be accomplished biologically under
anoxic conditions. Two types of enzyme systems are involved in the reduction of nitrate:
assimilatory and dissimilatory. In the assimilatory nitrate reduction process, NO, -N
is converted to ammonia nitrogen for the use by the cells in biosynthesis. It occurs when
NO, -N is the only form of nitrogen available. In the dissimilatory nitrate reduction process,
nitrogen gas is formed from nitrate. This latter process is normally called denitrification
of waste water, and demands a carbon source to provide energy for the process.
More than 2000 species of bacteria can perform the dissimilatory denitrification process.
2.2 Classification of Nitrification and Denitrification Unit Processes The nitrification and denitrification unit processes can be divided into two broad classes,
the attached growth systems, and the suspended growth systems.
In the attached-growth (biofilm) process (Chapter 5), the bulk of the biomass is retained
on a medium and it does, therefore, not require a solids separation step for returning
the solids to the nitrification reactor. The media that carry the nitrifying biofilm can be
anything from plastic media to Nitrogen ion-selective zeolites. Trickling towers, Rotating
Biological Contactors (RBC), Upflow Fixed Bed Reactors (UFBR) are the most widely
used for biofilm systems.
Suspended-growth (activated sludge) processes (Chapter 6) take place on suspension
of the biological solids in a mixed liquid. The result is the activated sludge processes,
based on only nitrifying bacteria, or on a combination of oxidative and nitrifying bacteria,
depending on the influent waste water. A subsequent clarification stage is required to
return the microorganisms to the nitrification stage.
The activated sludge and the biofilm systems can be further subdivided into systems
which use different variations of com bined oxidation-nitrification processes and separate
stages of nitrification or denitrification processes. Table 2.1 gives an overview of the
different applications.
Further details of the different biological unit operations for the removal of nitrogen
are outlined in Chapter 5 for the attached-growth systems and in Chapter 6 for the activated
sludge processes.
The demarcation between the biofilm and the activated sludge processes is not always
very clear. For example in the fluidized bed, the medium consists of solid particles covered
with a biofilm, and moves in the reactor. This principle is, therefore, similar in some ways
to the activated sludge process.
44
Table 2.1 Classification of different combined nitriflcationldenitrification and separate stage nitrification or denitriflcation units.
Combined carbon oxidation and nitrification processes (secondary treatment).
Suspended growth processes (activated sludge processes).
Activated sludge system. Single stage
Two-stage
Attached growth processes (Biofilter processes).
Trickling filters, with different filling material.
Rotating Biological Contractors.
Upflow Fixed Bed Reactor (UFBR), applying different media.
Combination of Biofilter and activated sludge process in two stages.
Combined nitrification and denitrification process.
Suspended growth processes.
Activated sludge systems with alternated oxic conditions.
Attached growth processes (biofiiter processes).
Simultaneous nitrification and denitrification applying N-ionselective medla in an upflow Fixed Bed Reactor (UFBR).
Separate stage nitrification processes (tertiary treatment).
Suspended growth processes.
Activated sludge processes.
Attached growth processes (Biofilter processes).
Trickling filter.
Rotating Biological Contactor (RBC)
Upflow Fixed Bed Reactor (UFBR).
Fluidized Bed Reactor.
45
2.3 Terminology Used in Waste Water Treatment. The terminology used in the treatment of waste water is often confusing. Terms
such as primary, secondary and tertiary treatment, in the treatment of municipal waste
water, frequently appear in the literature, and their usage is not always consistent.
The meanings of these terms, as used in Chapters 5 and 6 are therefore outlined
in this section. Figure 2.1 shows a flow diagram of a typical sewage treatment
plant, and indicates the different nitrogen removal steps.
compounds in the waste water takes place.
The latter part of this section will show at which step the removal of the nitrogen
Primary treatment:
Primary treatment removes solid material from the incoming waste water. Large
particles are removed by screens or reduced in size by grinding devices. Inorganic solids
are removed in grit channels and much of the organic suspended solids is removed
by sedimentation.
A typical primary treatment system is shown in Fig. 2.2. The primary treatment system
will remove almost one-half of the suspended solids in the incoming waste water.
The waste water transported to secondary treatment is called the primary effluent.
Secondary treatment.
Secondary treatment usually consists of a biological conversion of dissolved and
colloidal organic compounds into biomass, and its respiration. Some nutrient removal
takes place in secondary treatment units, depending on the ratio of heterotrophs and
nitrifier in the different unit processes. The different unit processes during secondary
treatment are the so-called combined carbon oxidation and nitrification processes. Fig.
2.3 shows the secondary treatment process.
Secondary systems normally produce an excess biomass that is sometimes recycled
into the secondary treatment with the influent.
Primary and secondary treatment can sometimes be accomplished simultaneously
in an oxidation pond or an aerated lagoon, as shown in Fig. 2.4.
In an oxidation pond, the oxygen is supplied from natural sources, and the oxygen
concentration, is therefore low, that is why oxygen rarely penetrates to the bottom of
the pond, and the solids that settle are decomposed anaerobically. In aerated lagoon
systems, oxygen is supplied by mechanical aeration, and the lagoon is, therefore, aerobic.
46
Figure 2.1 Flow diagram of a typical sewage treatment plant.
Figure 2.2 Plan of a primary treatment process.
47
luent to further
I I
Effluent recycle
TO sludge treatment/ 'sludge return
Figure 2.3 Plan of a secondary treatment process.
Tertiary treatment.
The reliability of stable processes has become increasingly important in order to
meet today's effluent standards for the nitrogen content in a waste water. It is, therefore,
often necessary to introduce another treatment step to refine the waste water. Tertiary
nitriiing or denitrifying steps are normally the same processes as described under secondary
treatment; but the concentration of a nitrifying or denitrifying biomass is much higher,
because the influent of organic compounds into a tertiary nitrifying treatment is so low,
that it will not cause a competition between the heterotrophic and nitrifying bacteria,
and thus lowers the nitrification rate. Tertiary nitrifying unit processes have, therefore,
a higher nitrification rate than the combined oxidation and nitrification steps.
2.4 Comparison of the Biofilm and Activated Sludge Unit Processes Biofilm techniques are generally used in small sewage works, serving populations
of less than 20 000. They tend to be higher in capital costs but lower in running costs
than activated sludge plants.
Biofilms oxidize generally more nitrogen than activated sludge per unit of bed volume,
but the final effluent carries more suspended solids. Activated sludge processes usually
require more skilled operators and more frequent maintenance than biofilms, and activated
sludge processes are often difficult to apply, particularly in small communities.
48
(a).
Raw waste water or to stream
Ib).
Raw waste water
Figure 2.4 Oxidation pond and aerated lagoon with simultaneous primary and secondary
treatment
Several experiments have been made to combine the suspended and attached growth
systems as listed in Table 2.2.
The main reasons for the combined cultivation are as follows:
- increase in reactor capacity
- increase in the biomass content in the system without an additional loading of the unit
- achievement of better and more stable nitrification.
process
Table 2.3 shows a comparison of the amount of suspended solids produced and
the yield coefficient in different nitrifying and denitrifying unit processes. Data for the
organic compounds are added to the list for comparison. Nitrifies, both in suspended
49
and attached growth systems have a low yield coefficient and a low sludge production.
Denitrifers have a low sludge production, but much higher yield coefficient. The nitrification
process would, therefore, appear to be vety difficult to initiate, compared with the denitrification
process. As a comparison the heterotrophic bacteria have a high yield coefficient and
ten times greater sludge production than the nitrifier.
Table 2.2 Examples of combination of suspended and attached growth processes for the nitrifying units described in the literature.
Example References
Plastic foam particles freely dispersed in the suspended culture
of activated sludge.
Hegemann (1 983)
Rogella and Payraudeau (1 987)
Rogella and Jarosz (1 987)
Blocks of Trickling Filter packing materials submerged in activated
sludge tanks.
Lang (1981)
Rogelia and Jarosz (1 987)
Rogella et a/. (1 988)
Schiegel (1988)
Rotating Biological Contactor (RBC) partly submerged into activated sludge
(SURFACT process). Guarino eta/. (1980)
Packed-Cage RBC Wanner eta/. (1990)
2.5 Comparison of the Nitrification Rate for Different Unit Processes. In Table 2.4 a comparison is made between the different nitrification rates as a function
of the temperature, from data found in the literature. The results shown are presented
either with the surface rate in g N/ m2 * day or the media volume rate in kg N/ m3 day.
Results show that the submerged filters generally has high nitriiication rates; in particular
the submerged filter named biocarbone, developed by O W in France, is among the
unit processes with the highest nitrification rate. Generally the biofilm unit processes
appear to have a higher nitrification rate than the activated sludge processes, expressed
50
with the above indicated units.
Table 2.3 Comparison of the developed amount of suspended solids and yield
coefficients In the different nltrifylng and denltrifylng unit processes. For comparison, data for heterotrophs are added to the list.
Process Yield coefficient Sludge Production
volatile suspended g vss / m3 sewage
solids (VSS)
Activated sludge
with nitrification 0.6 g VSS / g BOD 120
Trickling filter 0.4 g VSS / g BOD 80
Separate stage nitrification 0.1 g VSS I g NH,' - N 2
Comb. suspended nitldenit 0.5 g VSS I g BOD 100
Separate stage denitri-
fication with suspended unit 0.8 g VSS / g NO,' - N 16
Separate stage denitri-
fication with biofilter unit 0.6 g VSS 19 NO, - N 12
From: EPA (1975)
51
Table 2.4 A comparison of the peak nitrification rate for various units, both attached and suspended growth processes, at different temperatures, a).
Volumetric nitrification rates (kg N/m3 . d) at various temperatures OC
The RBC reactor is indicated as kg N/m2 . d (supemcial nitrification rate)
and results indicated in brackets.
10' 15' 20' 22' 25' Reference
Reactor type
Simultaneous nitrification and denitrification pilot plant b)
Biofilm-Controlled Nitrifying Trickling Filter (BCNTF)
RBC
RBC
Biocarbone (BAF and 1. Krirger)
Linpor (foam cubes in suspension)
Packed bed reactor gravel (5 mm gravel)
Fluidized bed reactor High porosity medium, activated carbon
Activated sludge
3.6 Halling-Srarensen and Hjuler (1 993)
0,s 0,32 0,40 Parker et a/. (1 989)
(1,7-2,1) Gujer and Boller (1989)
(3,6) Antonie (1 974)
>0,75 Rogella and Payraudeau (1 987)
0,32 Rogella and Payraudeau (1 987)
0,21 0,24 0,32 Gasser eta/. (1974)
0,48 Metcatf and Eddy (1991)
0,12 0,19 0,28 0,32 0,38 Wild eta/. (1971) 0,40 030 0,60 Stamberg et a/. (1974)
Partly from Parker et al. (1990)
a) Data are reported for comparative purposes only. If any of these processes are to be applied, pilot plant testing is recommended to verify removal rates.
b) Only reactor type that can perform simultaneous nitrification and denitrification.
52
2.6 Conclusions The following conclusions can be made on the basis of a comparison between the
nitrifying attached and suspended growth processes.
1. The nitrification rate for the attached-growth processes is higher than for the suspended-
growth processes.
2. The attached-growth processes are generally used in small sewage works (less
than 20 000 Person Equivalent (P.E.)), while the suspended-growth processes are
used in large treatment works. Today much effort is being put into the development
of large attached-growth sewage works. The future will, therefore, without any doubt
show more and more use of the biofilm technology for even larger treatment plants.
3. Attached-growth processes normally carry more suspended solids in the effluent
than the suspended-growth processes.
4. Activated sludge processes usually require more skilled operators and more frequent
mainteinance than the attached-growth process.
53
This Page Intentionally Left Blank
3. PROCESS CHEMISTRY AND BIOCHEMISTRY OF
NITRIFICATION
3.1 Introduction The purpose of this chapter is to present a review of the chemistry and
biochemistry of nitrification. An understanding of this subject is important for an
understanding of the factors affecting the performance, design and operation of
nitrification.
Biological processes for the control of nitrogenous residuals in effluents can
be classified in two broad areas: the production of an effluent where nitrogen
(ammonia and organic nitrogen) is converted into nitrate nitrogen: and the reduction
of nitrate into nitrogen gas.
In the first stage, nitrification is carried out by bacteria oxidizing ammonia into
nitrate with the intermediate formation of nitrite. Nitrification must conform to existing
water standards, where reduction of the residual demand on nitrogenous oxygen due
to the presence of ammonia is necessary, or where reduction of ammonia is required
to conform with existing standards.
The second stage, denitrification (for details see Chapter 4), is used following
the nitrification when the total nitrogenous content of the effluent must be reduced.
These conversions are of great importance because ammonia is a highly toxic
metabolic waste of aquatic organisms. Nitrite is somewhat less toxic than ammonia
(as NH,), although nitrite toxicity may occur at concentrations of less than 2.5 ppm for
some species (Westin 1973). Nitrate is considered relatively non-toxic to most aquatic
organisms. As an example, Knepp and Arkin (1985) showed that for channel catfish
(a highly tolerant species) the LD,, value for ammonia was 37.5 ppm, but nitrate
concentrations as high as 400 ppm did not affect feeding activities or growth rates.
3.2 Nitrification The two principal genera of bacteria of importance in biological nitrification
processes are Nitrosomonas and Nitrobacter. But Nitrosospira, Nitrosolobus and
Nitrosovibrio are also nitrifying bacteria.
These groups are classified as autotrophic organisms. They are distinguished
from heterotrophic bacteria in deriving energy from oxidation of inorganic nitrogen
55
compounds, rather than from the oxidation of organic compounds. These organisms
are also special because carbon dioxide is used for the synthesis of biomass rather
than organic carbon. Each group is limited to the oxidation of certain species of
nitrogen compounds. N itrosomonas, Nitrosospira, N itrosolubus and N itrosovibrio can
all oxidize ammonia into nitrite, but cannot complete the oxidation into nitrate. On the
other hand, Nitrobacter is limited to the oxidation of nitrite into nitrate. The apparent
inability of these organic developments has been investigated and there is evidence
that Nitrobacter can also utilize organic carbon as an energy source. This bacterial
species is therefore classified as a facultative autotroph. As complete nitrification is a
sequential reaction, treatment processes must be designed to provide an environment
suitable for the growth of both groups of nitrifying bacteria.
In contrast to many heterotrophs, the growth of nitrifiers is very slow, and the yield of
cells per unit of energy oxidized is low. Like other micro-organisms, nitrifiers can grow
at their maximum growth rate when optimum environmental factors can be obtained
in an environment without any toxic substances.
Two conditions, therefore, must be fulfilled in order to obtain nitrification in a
treatment plant. First, the sludge age has to be sufficiently high to prevent the wash
out of the slow-growing nitrifiers applying active sludge design systems. Second, the
contact time between the bacterial mass and the ammonia must be long enough to
oxidize the ammonia. Table 3.1 compares some characteristics of Nitrosomonas and
Nitrobacter.
In Chapters 5 and 6, different plant designs will be outlined. The different mass
balance equations for different nitrification plants will also be discussed, showing the
relationship between the biomass content and the nitrification efficiency of these
plants.
3.3 The Biochemical Pathway in the Nitrification Process At the biochemical level the nitrification process is more complex than simply
the sequential oxidation by Nitrosomonas of ammonia into nitrite, and the subsequent
oxidation by Nitrobacter, of nitrite to nitrate. Various reaction intermediates and
enzymes are involved in this processes. In soils, streams and treatment plants,
conditions permitting the oxidation of ammonia and nitrite can be created by a variety
of micro-organisms. Table 3.2 show some of the factors influencing the nitrification.
56
Table 3.1 Some characteristics of nitrifying bacteria and biological nitrification.
Nitrosomonas Nitrobacter
Morphology
Cell shape Cell size Motile Gram test Cell weight
Ovoid to rod-shaped 1 x 13 pm may or may not be negative 0.12-0.5 x lo-"' g
Estimated generation time hours 8-36
Ovoid to rod-shaped 0,5 x 1,0 pm may or may not be negative
12-59
Autotroph Obligate Facultative
Dissolved oxygen require- ments to nitrify Strict Aerobe Strict Aerobe
Process NH, + 13 0, + NO,- + H,O + H+ NO, + 0,5 0, + NO, AGO kJ/mole NH,-N -271 -78
Maximum growth rate at 20 OC 03
Nitrogen oxidation rate mg N/g VSS at 20 OC 100
Yield constant mg vss/mg N 0,08
100
0,03
pH-optim um 7,8 + 9,2 8,5 + 9,2 Long-term temperature constant susp. culture, OC-' 0,05
Long-term temperature constant att. culture, "C-' 0,03
Temperature range for process OC 50-350
Reaction Kinetics used in literature first order
Saturation constant, mg N/Iiter 03
Saturation constant,
Monod, zero order
mg O,/liter 1 ,o
0,04
0,03
5O- 3 5 O
Monod, zero order first order
57
Table 3.2 Factors influencing the Nitrification Process and the section considering the this
influence
Influence Section
Temperature 3.8
Dissolved Oxygen 3.9
PH 3.10
Bacterial Population Dynamics 3.12
Inhibitors 3.13
3.4 The Energy and Synthesis Relationship
The overall stoichiometric reactions in the oxidation of ammonia into nitrate can be
summed upas follows:
NH,' + 1,5 0, => 2 H+ + H20 + NO;
NO, + 0,502 => NO,-
(3.1 )
Equations (3. I) and (3.2) serve as energy-yielding reactions for Nitrosomonas and
Nitrobacter, respectively.
58
Equation (3.1) has been estimated by various investigators to yield a loss of free
energy between 58 and 84 kcal per mole of ammonia.
Equation (3.2) has been estimated to release between 15.4 and 20.9 kcal per mole
of nitrite. Thus, Nitrosomonas obtains more energy per mole of nitrogen oxidized than
Nitrobacter.
The overall oxidation of ammonium is obtained by adding equations (3.1) and (3.2),
providing equation (3.3).
NH4+ + 20, => NO; + 2 H+ + H,O (3.3)
Using the empirical formula C,H,NO, for the formation of biomass, the following
reactions can be written to represent growth of the Nitrosomonas and Nitrobacter
respectively:
15 CO, + 13 NH4+ => 10 NO; + 3 C5H7N02 + 23 H+ + 4 H20 (3.4)
5 CO, + NH4+ + 10 NO,- + 2 H,O => 10 NO; + C,H,NO, + H73.5)
Although about 99 per cent of carbon dioxide in solution shown in equations (3.4)
and (3.5) exists in the form of dissolved carbon dioxide, the carbonic acid-bicarbonate
equilibrium system is as follows depending on the pH in the environment.
CO, + H,O <=> H,C03 <=> Ht + H C O i (3.6)
C02 + H,O <=> H+ +HCO; (3.7)
The free acid produced in equations (3.1), (3.4) and (3.5) reacts to produce carbonic
acid according to equations (3.6) and (3.7).
The equations for synthesis-oxidation using representative measurements of yields
and oxygen consumption for Nitrosomonas and Nitrobacter are, according to Haug 8,
McCarty (1 972):
59
N itrosomonas
55 NH,' + 76 0, + 109 HCO, => C,H,NO, + 54 NO; + 57 H20 + 104 H2C03 (3.8)
N itrobacter
400 NO, + NH,' + 4 H2C03 + HCO, + 195 0, => C,H7N02 + 3 H20
+ 400 NO,- (3.9)
Equations (3.8) and (3.9) show that the oxidation of 100 mg NH,' -N produces 14,6
Adding equations (3.8) and (3.9) and simplifying, the overall synthesis and oxidation
mg of Nitrosomonas biomass and 2,O mg of Nitrobacter biomass, respectively.
reaction for the conversion of ammonium into nitrate is:
NH,' + 1,83 0, + 1,98 HCO, => 0,021 C,H,NO, + 1,041 H20 + 0,98 NO, + 1,88 H2C03 (3.10)
The conversion of 100 mg/l of ammonia nitrogen to nitrate-nitrogen according to
equation (3.10) therefore yields about 17 mg/l of total nitrifying biomass. This relatively
low yield has some far reaching consequences in the design of nitrification treatment
plants, as will be seen in later sections.
The oxygen consumption ratios in equation (3.10) are 3.22 mg 0, per mg NH,' -N
oxidized and 1.1 1 mg 0, per mg NO,- -N oxidized, respectively. This gives a total
oxygen need of 4,32 mg 0, per mg NH,' -N oxidized to NO, - N (Gujer and Jenkins
1 974).
60
3.5 Kinetics of the Nitrification Process The aim of this section and the following sections is to consider the number
of environmental factors affecting the rate of growth and nitrification of a nitrifying
biomass. A combined kinetic expression is proposed which accounts for the effect of
ammonia concentration, temperature, pH, organic content, and dissolved oxygen
concentration.
At several points, references are made to data obtained from various types of
nitrification processes. One distinction that needs to be clearly understood in this
Chapter is the difference between combined carbon oxidation-nitrification processes
and the separate stage nitrification process (also called a tertiary nitrifying treatment
process). The combined carbon oxidation-nitrification processes oxidize a high
proportion of influent organics relative to the ammonia nitrogen content. This causes
relatively low populations of nitrifiers to be present in the treatment plant relative to
oxidizers of the total bacterial biomass.
Separate stage nitrification systems, on the other hand, have a relatively low
organic load, relative to the ammonia load. As a result, higher proportions of nitrifiers
are obtained.
A nitrifying activity test was proposed by Tomlinson et a/. (1966) and later by
Painter and Loveless (1981). The test is able to determine the activity of sludge to
oxidize ammonia and ii is therefore suitable to determine the kinetics of the nitrification
in activated sludge.
3.6 The Kinetic Expressions for the Nitrification Process A review of the literature concerning the nitrification process shows diverse
opinions regarding the reaction rate equation for the nitrification process. Several rate
equations have been proposed. Each stems from different assumptions, and different
results have therefore been obtained. A review of these equations is presented in
Table 3.3.
Knowles, Downing and Barrett (1965) and Downing (1968), were among the
first to attempt to quantify nitrifying bacteria in waste water treatment plants. They all
used the Monod Model of population dynamics proposed by Monod in 1942, which is
similar to the Michalis-Menten relationship for enzyme reactions.
Huang and Hopson (1974) reviewed four different reaction rate equations (see
61
Table 3.4) to determine the appropriate equation. From the initial ammonia-nitrogen
concentration and the contact time studies, the nitrification process was shown to
follow a zero-order reaction.
The Monod Model used to describe the kinetics of biological growth of either
Nitrosomonas or Nitrobacter is the standard expression used in formulating the rate
equation:
(3.1 1)
where p = growth rate of micro-organisms, in day-'.
pmax = maximum growth rate of microorganisms, in day-'.
KS,-, = saturation constant = substrate concentration, mgA, at half the
S, = growth limiting substrate concentration, mg/l expressed as NH,' - N.
maximum growth rate.
When the reaction rate is independent of the substrate concentration, the
reaction rate can be considered as a zero order reaction. This results from a high
substrate concentration which leads to a maximum growth rate, indicating that no
diffusional limitations exist.
When the reaction is directly proportional to the substrate concentration then the
reaction can be considered as first order and the rate of reaction would be directly
governed by the ambient ammonia concentration.
The saturation constant KS,-, is temperature dependent, as will be discussed
in section 3.8. As the maximum growth rate of Nitrobacter is considerably higher than
the maximum growth rate of Nitrosomonas, and as the KS,., values for both organisms
are less than 1 mg/l NH,' -N at temperatures below 20' C, nitrite does not accumulate
in large amounts in biological treatment systems under steady-state conditions.
Table 3.7 and Fig. 3.4 presents values for Ks for both nitrifying species as
found under different environmental conditions.
62
Table 3.3 Summary of the different kinetic equations used in the literature to describe the nitrification process.
Order
Rate law Zero First Monod
Integrated rate law
(T, 0
Plot needed to give a straight line
Slope of the straight line
Half-life
ds - dt = - k
[product] versus t
Slope = - k
In [N] versus t
Slope = - k
0,693 t,= 7
Slope = h- "In,
Table 3.4 An overview of the kinetic rate equation used in different studies refered in the literature.
Plant N H ~ + range Process Kinetics References design application
to describe system
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Lab exp.
Trickling filter
Upflow
2,5-67,3
1,642
6,O-60,O
Up to 100 mg/l
Up to 20 mg/l
100- 1 100 mg/l
0-8000 mg/day
submerged filter Trickling - filter Trickling filter
Nitrification
Nitrification
Nitrification
Nitrification
Nitrification
Nitrification
Nitrification
Nitrite oxida- tion
Nitrification
Nitrification kinetic
Nitrification
Nitrification
Nitrification
Nitrification
Nitrification
0. order
0. order
0. order
Monod kinetics
Monod kinetics
Monod kinetics
Michealis- Menten
1. order
0. order
Monod
0. order
1. order
Close to 1. order
0. order
112 order
H u a n g a n d Ho son 1974) Kiff( 1971)
Wild eta/, (1971)
Stratton and McCarty (1 967)
Downing and Hopwood (1 964)
Knowles et a/. (1 965)
Charley et a/. (1980)
Charley et a/. (1980)
Loehr (1975)
C h u rchwel I et a/. (1980)
Wongchongd
Watanabe et a/. ( 1 980)
Balkrishnan and Eckenfelder (1 970)
Huang and McCarty (1 972)
Harkness (1 966)
Harremoes (1 978)
64
a b
t ”
-1 - K m
d 1.order 1 InA
C order
1 Product
Figure 3.1 Graphical representations showing a) Monod kinetics; b) Tranformation
of Michaelis-Menten Kinetics to the Lineweaver-Burk Plot; c) Zero Order kinetics and
d) First Order kinetics.
65
Nitrosomonas and Nitrobacter are both sensitive to their own and each others
substrate. Tables 3.5 and 3.6 show that wide ranges of ammonia and nitrite ion
concentrations can be oxidized by the nitrifiers. Different conditions can account for
the apparent discrepancies. Normal ammonia and nitrite ion concentrations in
domestic waste waters are not in the inhibiting ranges. Substrate and product
inhibition, however, are of significance in the treatment of industrial and agricultural
wastes. Table 3.19 show the ammonium nitrogen and nitrate nitrogen concentration
range for Nitrobacter inhibition as function of pH.
It would be desirable for the process of nitrification to be a reaction having
zero-order kinetics at least to low concentrations (< 5 mg/l) as the rate would be
constant and unaffected by the substrate concentration.
Mateles eta/. (1965) showed that while the Monod Model for microbial growth
was useful for steady-state cultures, its application in predicting the dynamic behaviour
of chemostats has limitations.
3.7 Relationship Between Growth Rate and Oxidation Rate
as follows:
The ammonia oxidation rate can be related to the Nitrosomonas growth rate,
(3.12)
or in the differantiated form of Michaelis-Menten:
(3.13)
66
Table 3.5 Effect of ammonia concentration on nitrification and nitrifying bacteria.
Concentration of EffecVObservation Condition of observa- Reference Ammonia-nitrogen tion method of study
m@
2,5 - 110,O Ammonia oxidation, Film reactor; mixed a zero order reation culture
26,4 - 46,5 Ammonia oxidation, Activated sludge a zero order reaction.
u p to 60 No inhibition.
u p to 10 Rate of ammonia oxida- Submerged filter
lab. scale
tion; a function of ammonia concentration feed. (between first and second orders).
receiving pre-oxygenated
0,063
600
Michaelis constant for Pure culture; Nitrosomonas growth at Warburg respirometer 25 'C.
Michaelis constant Dropping-mercury for Nitrosomonas growth electrode; pure culture at 20 'C.
Michaelis constant for Mixed continuous culture growth of ammonia oxidi- constant obtatined by zers at 23 ' C . computer fit of experi-
mental data with assumed yield coefficient value.
Oxidation possible Poultry waste; Repeated nitrification on a batch scale.
100 - lo00 Ammonia oxidation, a Lab. scale batch studies zero order reaction with mixed culture and
mineral salt media.
800 10,1% oxidation possible Bench scale studies, activated sludge, synthetic waste derived from nitrified poultry waste.
Huang (1973)
Metcalf and Eddy (1 973)
Haug and McCarty (1 972)
Painter (1970)
Loveless and Painter (1 968)
Poduska and Andrews gWi (1 975)
Praksam er aL(l974)
Wong-Chong and Loehr (1 975)
Anthonisen (1974)
67
Table 3.6 Effect of nitrite concentration on nitrification and nitrifying bacteria.
Concentration of EfiecVObservation Condition of observe- Reference Nitrate-nitrogen tion method of study
man
c 10 Limiting
500
140, 160, 280, 700 and 1400
Rate of oxidation may be described by first order rate equations; decrease in rate constant with in- creasing initial concentra- tion explained by Michaelis- Menten kinetics.
1200
1400
Nitrate toxic in the lag phase at all pH values; not so in the lag phase at alkaline pH.
Ammonia oxidizers not completely inhibited
Causes 40% inhibition of Nitrobacter activity
4200 Complete inhibition of Nitrosomonas.
Activated sludge; lab.scale
Batch studies in a marine nitrifying filter system.
Batch and pure culture of N itrosomonas
Mixed culture from an oxidation ditch; poultry waste: respirometric experi- ment.
Measured by decrease in oxygen uptake by bacteria
Tomlinson, Boon and Trotmann (1 966)
Srna and Baggaley (1975)
Pokallus (1 963)
Prakasam eta/. (1974)
Boon and Laudelout (1 962)
Painter (1970)
68
Q, W
Table 3.7 Kinetic constants for nitrifying bacteria.
Organism Max spec. Cellular yield Ks Reference growth rate YOb* dnf P m m d-' g V S S / g N
Nitrosomonas 0,46-1,86 (30°C) 0,06 10 (30°C)
3 3 (25°C)
1,2 (20°C)
1,5 (18°C)
0,46-2,20 0,03-0,13 0,06-5,6
(1 5"-32"C)
Nitrobacter 1,39 (32°C) 0,02 8 (32°C)
5 (25 "C)
0,28-1,44 0,02-0,08 0,07-8,4
(1 5-32°C)
0,5 (30°C) Painter (1 977)
0,3 (20°C)
Marais and Ekema
(1 976)
0,3-1,3 Charley et d(1980)
1,0 (30°C) Painter (1 977)
0,5 (32°C)
0,25 (18°C)
0,25-1,3 Sharma (1 977)
where
pmax = peak Nitrosomonas growth rate, day-',
dSn/dt = peak ammonia oxidation rate, mg NH,' - N
Y,
S,
&,, X,
oxidized /mg VSSI day,
= nitrifying yield coefficient, mg Nitrosomonas grown
(VSS) per mg NH,' -N removed,
= The substrate concentration, mg/l,
= Saturation constant, NH,' -N in mgll,
= nitrifying mass cell concentration in mg/l,
If the substrate concentration S is much higher than K, then equation (3.13)
can be written as:
(3.14)
In equations (3.13) and (3.14) only the effect of ammonia concentration is
considered; in later sections, the effect of temperature, pH, organics and dissolved
oxygen are also discussed.
If the temperature, pH, organics and dissolved oxygen concentration are
unknown, equations (3.13) and (3.14) are proposed. But if the indicated parameters
are known, equation (3.37) will be more precise to use.
The growth rate of organisms can be related to the design of activated sludge
systems by noting the inverse relationship between solids retention time and growth
rate of nitrifiers:
(3.15)
70
where $c = solids retention time, days.
p = growth rate of nitrifying organisms in day -'. The solids retention time can be calculated from systems operating data by
dividing the inventory of microbial mass in the treatment system by the quantity of
biological mass losted daily (EPA 1975).
3.8 The Influence of Temperature on the Nitrification Rate The optimum temperature for the growth of nitrifying bacteria, according to the
literature, is between 28" C and 36" C, although an optimum temperature of up to
42" C has been reported for Nitrobacter by Painter (1970). Growth constants of
nitrifying bacteria are greatly affected by temperature (Table 3.9). Figure 3.2 shows
that the nitrification rate is a function of temperatures between 5" and 35" C. The
maximum growth rate occurs at approximately 30" C. Curve A, which was produced
by Borchardt (1966) indicates that no sharp optimum temperature can be defined and
that there is a plateau of maximum activity between 15" C and 35" C. Below 15" C
however, the nitrification rate drops sharply, and is reduced by 50 per cent at 12" C.
Wild et a/. (1971) found (curve B) that an almost straight-line relationship exists
between the nitrification rate and temperature. Similar temperature dependencies have
been reported in single stage nitrification-denitrification schemes.
Data are also available on the effects of temperature on the oxidation of
ammonia to nitrite by Nitrosomonas (curves E, F, G and H), and of nitrite to nitrate by
Nitrobacter (curves C and D). Both species seem to be similarly influenced by
temperature.
Randall and Buth (1 970), however demonstrated that although both nitrite and
nitrate formation were strongly inhibited at temperatures of 10" C or less, the inhibitory
effect of lowered temperature was greater for Nitrobacter than for Nitrosomonas; this
was evident from the nitrite build-up at low temperatures.
Barrit (1933) found that the thermal death point of a pure culture of Nitrosomo-
nas was between 54" and 58" C. Almost no growth of nitrifying bacteria was found
below 4" C.
Suspended growth cultures are more sensitive to temperature changes than
biofilms (Murphy and Dawson 1972). The dependency on temperature of attached and
71
suspended growths is illustrated below (Fig 3.3).
Downing et a/. (1964) presented results for the relationship between
temperature and saturation concentration Ks,n and temperature and maximum specific
growth rate p. Their results are presented in Fig. 3.4. As can be seen, both the
maximum growth rate, p and the saturation constants, K, for Nitrosomonas and
Nitrobacter are markedly affected by temperature. Further, the maximum growth rate
for Nitrosomonas in activated sludge was found to be considerably less than for
Nitrosomonas in a pure culture.
The literature suggests the following general relationship between the
saturation constant K,,n and temperature t in "C.
K , , ~ = 100,051* t - 1,158
(3.16) Reference: EPA 1975; Nitrosomonas in river water and activated sludge.
K ~ , * = 100,063*t - 1,149
Reference: EPA 1975; Nitrobacter in river water. (3.17)
(3.18)
Reference: Watanabe et a/. 1980 applied to suspended culture of nitrifier at T "C.
72
Nitrification Efficiency y, t A
100 -
80 -
60 -
/ Tnmn O C
0' W I I I I I t L L . ".,my v
c 0 5 10 15 2 0 25 30 3 5 40 4 5
Fig. 3.2 The influence of temperature on the nitrification process, presented in the text as A to H, (Source: EPA 1975).
% of Nitrification Rate t 100
80
60
2 LO
20
Temp "C 0 I I I I I I 1 c
5 10 15 2 0 25 30 35 L O
Fig 3.3 Comparison on the effect of the temperature on suspended growth and attached growth nitrification systems. A) to
D) are attached growth systems and E) is a suspended growth system (Source EPA 1975).
Table 3.8 The influence of temperature on the nitrification process.
Temperature Degree of Circumstances of Reference C inhibiflon % obsevafion
15'-35' 0 Nitrification in Borchardt (1 966) 13' 25 activated sludge 1 20 50 5' 90
30' 27' 17'
0 10 50
Nitrification in activated sludge
Wild eta/. (1971)
26' 7' 5'
Sutton et a/. (1974) 0 21 53
30'
15' 5'
Stratton and McCarty (1 967)
0 Nitrobacter in
river water 60 75
30' 15' 5'
30' 15' 5'
Knowles et a/.( 1965) 0 62 77
Nitrobacter in estuary water
Nitrosomonas in pure culture
Buswell et a/.( 1954) 0 70 83
30'
15' 5'
Stratton and McCarty (1 967)
0 Nitrosomonas in
river water 75 85
30' 15' 5'
Knowles et a/.( 1965) 0 80 90
Nitrosomonas in estuary water
30' 15' 50
Downing (1968) 0 85 93
Nitrosomonas in activated sludge
75
Knowles et a/. (1965) proposed the following two relationships between
temperature and the saturation constant for Nitrosomonas and Nitrobacter, following
the Arrhenius law:
(3.19)
Ks,Nitrabacter - - 0, 405 * e 0 , 1 4 6 ( T - 1 5 ) (3.20)
Neufeld et a/. (1 986) showed that the nitrification rate followed Michaelis-
Menten Kinetics and proposed the following relationship between KM and the
temperature. KM was found to decrease in the temperature range of 22-30 "C in
accordance with the equation:
log(K,) = 1,53-0,032(T) (3.21)
and at temperatures > 30 "C KM was found to follow the expression:
lOg(K,) = -1,88+0,082 ( T ) (3.22)
The relationships between the effect of temperature t in "C and the maximum
growth rate kmax in d -' for nitrifying organisms:
(3.23)
Reference: EPA (1975); Nitrosomonas in river water and pure culture.
76
(3.24)
Reference: EPA (1 975); Nitrosomonas in activated sludge.
2 = 6,5*0,914(e-20) CIINlX
(3.25)
Reference: Faup, G.M et a/. (1982); Nitrosomonas in a UFBR (upflow fluid bed
reactor). Temperature, t, between 9 "C and 20 "C.
- 0,7g*eO,69(C-15) P m x -
Reference: EPA (1 975); Nitrobacter in river water. (3.26)
Table 3.9 Temperature dependence of the maximum growth rates of nitrifiers.
T "C pmax d"
5
10
15
20
25
0.18
0.29
0.47
0.77
1.25
0.13
0.23
0.40
0.73
1.30
Source Knowles et a/. (1965)
77
The literature shows that the relations obtained between the temperature and
Ks," and the temperature and pmax are dependent upon the environment and test
circumstances.
Somewhat differing temperature effects have been found for attached growth
systems and suspended growth systems.
Comparing the suspended-growth and attached-growth nitrification data, one
can conclude that attached-growth systems have an advantage in withstanding low
temperatures (below 15°C) without significant reduction in nitrification rates.
Measurements of nitrification rates for suspended-growth systems, however, are not
normally made on the same basis as those made on attached-growth systems. In
suspended-growth systems, rates are expressed on a per-unit-of-biomass basis
(MLVSS is used). Precise measurements of biomass are normally not possible in
attached-growth systems so other parameters are used, such as reaction rate per unit
surface or volume.
Attached-growth systems can also compensate for colder temperature
conditions by the biofilm growth growing thicker. If rates could be expressed on a unit
biomass basis for both system types, the variation in reaction rates with temperature
might thus be more similar.
Shammas (1 986) showed that the effect of temperature on nitrogen kinetics
fitted the popular modified Arrhenius relationship.
(3.27) where
KS," = maximum growth rate at temperature t (d -').
K20 = maximum rate constant at 20 "C
b = temperature coefficient
78
a KS, HALF SATURATION COEFFICIENTS, mg/ L
6.0-
4.0 -
2.0
1.0 0.8 -
0.6
0.4 -
o*2 1 T I Temperature, o c
0.1 ’ * 12 16 20 24 28 32 8
~c~,,~ MAXIMUM GROWTH RATES, DAY-^ 6.0
b
2.0
1.0 0.8
0.6
0 . 4
0 . 2
0.1
16 20 24 28 32 0 12
Figure 3.4 The influence of temperature on a) KS,” and b) pmax for the nitrification process, (EPA 1975).
79
Shammas (1986) also showed that b varies with the bacterial concentration
calculated as MLVSS. Different values of b is shown in Table 3.10.
b = 0,00044 *X0169
(3.28)
where
X = MLVSS concentration in mg/l.
and b is constant with respect to pH.
The same authors reported that values of the nitrification rate constant KS,"
ranged from 0,0085 d -' at 4 "C and pH = 7 to 0,175 d -' at 33 "C.
The temperature relationship to maximum specific growth by an exponential
expression has been described by several authors (Zanoni 1969; Andersen and
Poulsen (1 976); Jenkins (1 969) and McHarness et a/. (1 975)):
(3.29)
where:
pm and pm,rer are the maximum specific constants at temperature t and t, (0°C)
respectively, and A is a constant for a specific temperature range referred to as
the "temperature coefficient".
All studies mentioned in Table 3.1 1 were conducted under steady-state
Only very few studies were conducted with rapid temperature changes, and
conditions, obtained with long-term temperature conditions.
then only under marine conditions.
80
Table 3.10 Values of b with comparable values from different literature sources. The
highest coefficient for b for ammonia oxidation in an activated sludge medium was
reported by Downing et a/. (1968).
Temperature Condition
coefficient
b
Reference
0,028
0,059
0,121
0,073
0,095
0,059
0,084
0,056
0,120
0,075
Activated sludge
ammonia to nitrate pH 7,O to 8,3
MLVSS = 430 mg/l t = 4 "C to 33 "C MLVSS = 1200 mg/l t = 4 "C to 25 "C
MLVSS = 3200 mg/l t = 4 "C to 25 "C
Pure culture
Ammonia to nitrite
Thames estuary water
Ammonia to nitrite
Nitrite to nitrate
River water
Ammonia to nitrite
Nitrite to nitrate
Activated sludge
Ammonia to nitrite
Single stage activated sludge
Nitrification
Shammas eta/. (1986) I1
Buswell et a/. (1954)
Knowles et a/. (1 965) I1
Stratton et a/. (1 967) I,
Downing et a/. (1 968)
Sutton eta/. (1978)
From Shammas (1 986).
81
500
400
300
2 00
10 0
0
mg/l
1000 2000 3000
Figure 3.5 Variation of maximum nitrification velocity with MLVSS concentration at different temperatures. (From Shammas 1986).
82
Table 3.1 1 Temperature coefficient for nitrifying bacteria.
Process Range Tmf A &,,ref Reference
Nitrogenous phase
in BOD bottle
analysis
Nitrification in
suspended culture
Nitrifying in
treatment process
Nitrosomonas in
fill and draw pilot
plant activated sludge
Nitrosomona in water
from Thames estuary
Nitrobacter in water
from Thames estuary
Nitrosomonas in
activated sludge
Nitrosomonas in
pure culture
Attached separate
culture
10-22' c
5-20' C
10-30' C
5-10' C
8-20' c
6-30' C
8-30' C
10-25' C
10-25' C
5-25' C
20' C 1.097
20'c 1,12
12'C 1,07
10'c 1,19
1,12
15'C 1,099
15' C 1,058
15'C 1.123
15'C 1,103
1,08
0,12
0 3
0,25
1,18
0,47
0,79
0,18
0,47
Zanoni (1 969)
Andersen and
Poulsen (1 976)
McCarty (1 976)
I, I,
Jenkins (1 969)
Knowles eta/., from
(1 965)
Downing &
Hopwood (1 964)
McHarness et
et a/. (1975)
Partly from Ohgaki and Wantawin (1990).
83
3.9 The Influence of Dissolved Oxygen on the Nitrification Rate In engineering calculations, an aeration requirement of 4,6 mg 0, per mg NH,'
-N is just sufficient to be used for the nitrification process.
In almost all treatment systems, oxygen is also required to oxidize other
materials than ammonia present in the waste water. This, therefore, often raises the
total oxygen demand in a nitrifying plant.
Results from a number of studies on the effect of dissolved oxygen con-
centrations on the nitrification efficiency are summarized in Table 3.12. Most studies
were conducted on suspended-growth systems. In the case of attached growth
systems, the oxygen availability to the nitrifying biofilm can be affected by many
parameters.
The concentration of dissolved oxygen (DO) has a significant effect on the
rates of nitrifier growth and nitrification in biological waste treatment systems. The
Monod relationship has been used to model the effect of dissolved oxygen,
considering oxygen to be a growth limiting substrate, as follows:
(3.30)
where: DO = dissolved oxygen, mg/l and
K O,,n = half-saturation constant for oxygen, mgA, in the nitrification process.
While the general effect of DO on kinetics is firmly established, further study
is needed to determine the factors affecting the value of K O,,n. All of the various
estimates are from systems where combined carbon oxidation-nitrification is practiced,
and no measurements have been made on separate stage nitrification systems.
K O,,n values for separate stage nitrification systems may very well be different from
those for combined carbon oxidation-nitrification systems. Most often the operating DO
is 2.0 mg/l or less, in studies (see Table 3.12), therefore a value of K O,,n of
approximately 1,3 mg/l, will give a nitrification (or nitrifier) growth rate (equation 3.30)
of about 60 % of the peak rate, following Downing et a/. (1978).
84
Table 3.12 The influence of dissolved oxygen on the nitrification process.
Dissolved oxygen Observation CircumstanceMethod Reference concentration of observation m@
Below 2
Below 4
Below 3
0,08
> 7,5
Below 1 - 1 5
5 0,l
0,5-0,7
Saturation
1
0,647
Limiting tor Nitrosomonas growth ('1
Limiting for Nitrobacter growth
Degree of nitrate about 10% lower at 2 mgd
Limiting
Critical
Inhibiting
Limiting for growth
Nitrification
Critical (")
Limiting
Limiting
Limiting
Dropping-mercury Painter and Jones method used to measure (1963) oxygen uptake
10-1 batches; water Knowles, from Thames; Downing & determination made Barrett (1 965) from a model
Small-scale plant
Activated sludge
Pure culture of Nitrosocystis oceanus
Activated sludge
Pure culture of Nitrosocystis oceanus
Activated sludge
Batch tests with activated sludge
Pilot plant; activated sludge
Percolating filter receiving sea water marine nitrifiers
British Ministry of Technology (1 965)
Downing & Knowies (1-
Gunderson (1 966)
Wuhnann (1964)
Carlucci & McNally (1 969)
Downing and Knowles (1 966)
Kiff (1972)
Metcalf & Eddy (1973)
Forster (1 974)
u p to 60 No inhibition Submerged filter Haug & McCarty (1 972) no increase in receiving pre-oxygenated rate of ammonia waste water oxidation
(*) Rate of nitrification is the concentration below this value. (") Minimum concentration necessary for nitrification to occur.
85
' 100-
% of Nitrification Rate
0
50-
DO mg/l
0 c 0 0 5 1 0 1 5 2 0 2 5 30
Figure 3.6 The influence of dissolved oxygen on the nitrification rate.
Most mathematical models for biological growth take into account only one sub-
strate, such as the Monod model, since experimental studies are usually performed
with all other nutrients in excess. But Stenstram and Poduska (1980) used a double
substrate-limiting kinetic expression to describe the combined effect of dissolved
oxygen and ammonia-nitrogen on the growth rate, as shown in the following equation.
The equation is a modified form of the Monod single substrate model.
where
p = Specific growth rate (d-')
pmax = Maximum specific growth rate (d-')
SN = Ammonia concentration
DO = Dissolved oxygen concentration
Ks,N = Half saturation constant for ammonia nitrogen
= Half saturation constant for dissolved oxygen
Kd = decay or maintenance coefficient (d-')
(3.31)
86
The double substrate-limiting kinetics is interesting, because substrate diffusion
through biofilms will result in the limitation of either the electron donors or the electron
acceptors in the biochemical reaction.
Typical values of the half saturation constant KO,, are shown in Table 3.7
It would appear, looking at Table 3.7 that the activity of Nitrobacter is suppressed
under low dissolved oxygen concentrations more than that of Nitrosomonas.
Painter (1977) noted that the presence of organic matter can directly inhibit nitrifiers
by virtue of heterotrophs oxidizing the compounds and successfully competing for the
available dissolved oxygen, if this is kept at a fairly low concentration, as the Ks,o
for heterotrophs is generally lower than that for nitrifiers.
3.10 The Influence of pH on the Nitrification Rate In the literature, the optimum pH value for the nitrification process varies between
8 and 9. Figure 3.7 summarizes investigations of pH effects on the nitrification rate.
Usually the nitrification rate decreases, as the pH decreases. By measuring the
nitrification rates Meyerhof (1916) found the pH optimum for Nitrosomonas to be
between 83 and 8,8, and for Nitrobacter to be 8,3 to 9,3.
Hofman eta/. (1973) made similar investigations, and found for both organisms
an optimum pH of 8.3, and that the nitrification rate fell almost to zero at pH 9,6. They
also found that nitrification proceeded with considerable speed until the pH was as low
as 6 5 Hofman eta/. (1973) further reported that the optimum pH for nitrite oxidation
by Nitrobacter was 7,7 and not 8,8 as found by Mayerhof (1915). Wild et a/. (1964)
suggested the optimum pH for nitrification to be 8,4 and that 90 per cent of the
maximum nitrification rate occurs between pH 7,8 and 8,9. Less than 50 per cent of
the optimum rate occurs outside the range of pH 7,O to 9,8. Painter (1972) reported
that the point at which the rate of nitrification decreased was between pH 6,3 and 6,7,
and that between pH 5 and 53, nitrification ceased.
Anthonisen (1974) suggested the following mechanism by which pH affects the
rate of nitrification. His hypothesis is based on the fact that the ammonia/ammonium
and nitrite/nitrous acid equilibria depend on pH. Both "free ammonia" NH, and "free
nitrous acid" HNO, inhibit the nitrifying organisms. When the intracellular pH of a
nitrifying organism is lower than the pH of the extracellular environment, free ammonia
will penetrate the cell membrane, and inhibit the bacteria.
87
30
20
10
4 - ; AFTER MYERHOF I I I I I
- I I
I I I PH
-
, I n - b
OQ--Q - N
0
/ \\ O AFTER ENGEL / o / '! ;\ O I Y
/ /
/
AND ALEXANDER lp I
I I , I
30
20
10
I I I \ I \ I
I I
/ O/ I I
4 ; AFTER MYERHOF I I I I I
I I
I I I , I n -
7.0 8.0 91)
- -
- PH
A 10.0
Figure 3.7 The influence of pH on the nitrification process.
Ionized ammonia NH,', on the other hand, will remain in the extracellular
environment. Similarly, when intracellular pH is higher than that of the extracellular
environment, free nitrous acid penetrates the cell, not the nitrite ions. Anthonisen
proposed, therefore, that the ability of ammonia and nitrous acid to penetrate the
nitrifying organisms was one of the reasons why the nitrification process is less
affected at pH values between 8 and 9.
Equation (3.3) shows that H+ is produced by the oxidation of ammonia and
carbon dioxide. When the biomass synthesis is neglected, it can be calculated that
7,14 mg of alkalinity, as CaCO,, is destroyed per mg of ammonia nitrogen oxidized.
Experimentally determined ratios are presented in Table 3.13. A ratio of 7,l mg
alkalinity (as CaCO,) destroyed per mg of ammonia nitrogen oxidized may be used
theoretically in plant design.
88
As the nitrification process reduces the HCO, level and increases the H,CO,
level, it is obvious that the pH would tend to be decreased. This effect is mediated by
stripping of carbon dioxide from the liquid by aeration, and the pH is therefore often
raised. If the carbon dioxide is not stripped from the liquid, the pH may be depressed
to as low as 6,O. Haug eta/. (1974) calculated that to maintain the pH greater than 6,O
the alkalinity of the waste water must be 10 times higher than the amount of
ammonium nitrified.
It is important to distinguish between long-term and short-term pH effects on the
environment where the nitrification process is to occur.
There is a great difference in the effects that can be observed in the nitrification
process, if pH varies over short (hours, days) or long periods (months, years). Most
investigations referred to in this text have been on a short-term basis. Investigations
of long-term effects have not been described in the literature.
Table 3.13. Alkalinity destruction ratios in experimental studies
System X mg alkalinity destroyed Reference mg NU,+ -N oxidized
Suspended growth 6 4 Mulbager ef a/. (1971)
Suspended growth 6,l Horstkotte et a/.(1973)
Suspended growth 7,1 Newton et a/. (1973)
Attached growth 63 Gasser et a/. (1974)
Attched growth 6,3 to 7,4 Osborn et a/. (1965)
Attached growth 7,3 Haug eta/. (1972)
as CaCO,, the theoretical value is 7.1 From EPA (1975).
The hydrogen ion concentration (pH) has been found to have a strong effect on
the rate of nitrification. There is a wide range in reported pH optima; the almost
universal finding is that, as the pH moves into the acid range, the rate of ammonia
oxidation declines. This has been found to be true for both unacclimatized and
acclimatized cultures, although acclimation tends to moderate pH effects.
89
Downing ef a/. (1966) showed that the effect of pH on nitrification for pH values
less than 7,2 can be estimated from the following relationship:
(3.32)
This expression was developed for combined carbon oxidation-nitrification
systems, but its application to separate stage nitrification systems would appear useful.
For pH levels between 7,2 and 8,0 the rate is assumed constant.
Table 3.14 Effect of pH on the nitrification.
PH Degree of Circum.tcvla. of Refwsnce inhibition % ohmration
8,5 - 8.8 739 9,3
6,7 - 8,O 595 9 2
8,0 - 8,5
8,3 - 8,6
7,2 - 8,2
7,2 - 8,2 6 2
9,6
7.3 - 8,4
7,5 - 8,O
7,O - 8,0
5,5 - 6.0 6 1
4,9 - 7,2
8,4 - 8,5
0 50 50
0 100 100
0
0
0
0 50
50
0
0
0
Pure culture of Nitrosomonas
Pure culture, test. tube scale
Pure culture of Nitrosomonas
Pure culture of Nitrosornonas
Pure culture of Nitrobacter
Pure culture of Nitrosomonas
Batch Culture
Pure culture of Nitrobacter
Pure culture of Nitrosomonas iso- lated from activated sludge.
Submerged filter, mixed but predomi- nantly nitrifying bacteria.
Mixed culture; lab. scale
Two-stage, activated sludge pilot plant.
Mayerhof (1917)
Barritt (1 933)
Buswell et a/. (1 954)
Lees (1 954)
Lees (1954)
Engel & Alexander (1 958)
Engel & Alexander (1 958)
Boon & Landelout (1962)
Loveless & Painter (1 968)
Haug & McCarty (1 972)
Praksam & Loehr (1972)
Rimer & Woodward (1972)
90
Table 3.14 (eontlnued)
8,O - 8.8 0 7,l 50 9.8 50
8,O 0 579 50
7,45 0
73 0
Batch activated sludge; lab. study
Medcalf & Eddy (1973)
Percolating filter Forster (1 974) lab. scale mixed population. Marine nitrifying filter system; batch studies
Sma & Baggaley (1975)
Simultaneous nitri- Halling-Ssrensen & fication and de- nitrification attached growth UFBR.
Hjuler (1 992)
(‘):Adaptation in 10 days, the rate of ammonia oxidation becomes the same as that at pH 745. (+): pH not controlled, nitrification occured at pH 4.9; no improvement between pH 5 and 11.
Because of the effect of pH on the nitrification rate (see Fig. 3.8), it is especially
important that there be sufficient alkalinity in the waste water to balance the acid
produced by nitrification. Addition of alkalinity to the waste water may be necessary.
Boon and Laudelout (1962) developed a kinetic expression for the effect of pH
on the nitrite oxidation by Nitrobacter winogradskyi. They suggested that inhibition of
high nitrite concentration results from non-competitive inhibition of nitrous acid, while
at pH over 7 there is a competitive inhibition of the adsorption of nitrite on the enzyme
sites by OH- -ions.
The rate equations for pH below 7 and pH above 7 are shown separately in
equations (3.33) and (3.34) respectively.
(3.33)
(3.34)
91
where:
S = nitrite concentration.
Ka = equilibrium constant of nitrous acid and nitrite ion dissociation.
Ki = dissociation constant of the enzyme-nitrous acid complex.
K, = basic acid-base dissociation constant of the active enzyme site.
The total rate equation for pH effects was thus determined by combining
equations (3.33) and (3.34) as in equation (3.35).
(3.35)
Results showed that K, and Ki were 0,004 and 8,2 pM of NO,', respectively.
Suzuki et a/. (1974), using the Lineweaver-Burke plot, in the study of the pH
effect on the oxidation of ammonia by Nitrosomonas europaea, found that the value
of the Monod saturation ammonia constant decreased when pH increased. This means
that having pH as the parameter, the plot shows competitive inhibition.
As Nitrosomonas and Nitrobacter are both sensitive to their own substrates of
unionized ammonia and nitrite, and the unionized-ionized nitrogen equilibria depend
on pH, it follows that the pH value is an important factor.
92
% of Maximum Oxidation Rate t - Engel and Alexander (1959)
A Wlld at al. (1964)
o Meyerhof (1916)
W Hoiman and Lees (1953)
Q Meyerhof (1916)
Figure 3.8 The influence of pH on the nitrification rate. A summary of different results
found in the literature. Source: Shammas (1986).
3.1 1 A Kinetic Expression Combining Several Limiting Factors of the Nitrification Process
In previous sections, the effects of ammonia level, temperature, pH, and
dissolved oxygen on the nitrification rate have been presented. In all practical systems,
these parameters influence the nitrification rate simultaneously. Chen (1 970) showed
93
that the combined effect of several limiting factors on biological growth can be
introduced as a product of a Monod-type expression.
Taking this approach for nitrification, the combined kinetic expression for nitrifier
growth would take the following form (EPA 1975):
DO * ( 1 - 0 , 8 3 3 ( 7 , 2 - p H ) ) S p = p-*-* (K,+S) (Ko,+DO)
(3.36)
where: p = maximum nitrifier growth rate at temperature T and pH less than 7,2.
Using specific values for temperature, pH, ammonia and oxygen, from Tables
shown in the EPA (1975), the following expression results for pH less than 7,2 for
Nitrosomonas and is valid for temperatures between 8 "C and 30 "C:
m p = 0 , 4 7 * ( e 0 ~ 0 9 s ' ( t - 1 5 ~ ) * ( 1 - 0 , 8 3 3 ( 7 , 2 - p ~ ) ) * 100,051rt-',' sn * m + 1 , 3
(3.37)
In equation (3.36) the first term in brackets allows for the effect of temperature.
The second term in brackets considers the effect of pH. For pH less than 7,2 the
second quantity in brackets is taken to be unity. The third term in brackets is the
Monod expression for the effect of the ammonia nitrogen concentration. Similarly, the
fourth term in brackets accounts for the effect of DO on the nitrification rate.
Equation (3.37) has been adopted for illustrative use. When other reliable data
become available, equation (3.37) can be modified to suit particular circumstances.
If the ammonia removal rate is defined as in equation (3.36), then equation
(3.38) can be written as follows:
ds, = Pmax *Xn*-*- Sn Do * (1-0,833 (7,Z-pH) dt yn K,,,+S, Ko2+D0
(3.38)
The biggest problem in the analysis of rate data for microbial nitrifying bacteria,
with or without heterotrophic bacteria, is the estimation of nitrifier concentration for
determination of the specific growth rate p, the yield coefficient Y, and the saturation
constant KS,,.
94
3.1 2 Bacterial Population Dynamics Applied in the Nitrification Pro-
cess The kinetics of the growth of nitrifiers have been discussed in the previous
sections. In all practical applications in waste water treatment, nitrifier growth takes
place in waste treatment processes, where other types of biological growth occur. In
no case are there opportunities for pure cultures to develop.
This fact has significant implications in process design for nitrification.
In combined carbon oxidation-nitrification systems as well as in separate stage
nitrification systems, there is sufficient organic matter in the waste water to enable the
growth of heterotrophic bacteria. In this situation, the yield of heterotrophic bacteria
growth is greater than the yield of the autotrophic nitrifying bacteria. Because of this
dominance of the culture, there is the danger that the growth rate of the heterotrophic
organisms will be established at a value exceeding the maximum possible growth rate
of the nitrifying organisms. When this occurs, the slower growing nitrifiers will gradually
diminish in proportion to the total population, and be washed out of the system.
Because waste water is a mixed culture system, a knowledge of the mutual
relationship between nitrifying and heterotrophic bacteria is very important in the
construction of nitrifying waste water plants.
Painter (1 977) showed that the maximum specific growth of nitrifying bacteria,
determined in the treatment process, is significantly different from that observed in a
pure culture.
The reasons for this difference may be explained as follows:
1) Domination of heterotrophic bacteria which suppress nitrifying growth,
because growth conditions, i.e the COD/N ratio, in the treatment plant enable the
growth of heterotrophic bacteria prior to nitrifying bacteria.
2) Because the half saturation constant Ks,o for heterotrophs is generally lower than
that for nitrifiers, heterotrophs will generally compete with the nitrifiers for the
available dissolved oxygen.
3) The toxic constituents of waste water may inhibit nitrification.
95
4) Fluctuation or limitation of nutrients.
5) A genuine difference between isolated strains and those effecting nitrification in the
treatment process.
Especially 1) is an important factor in the construction of nitrifying waste water
systems. Stover eta/. (1 976) have presented experimental results showing the effects
of different COD/N ratios on nitrification, in both the activated sludge process and in
the UFBR, system in both cases applied using non-toxic synthetic media.
The competition for nitrogen by heterotrophs, or inhibition, interferes with the
removal of ammonia and reduces the production of nitrate under the conditions of a
high COD/N loading. Applying a high COD/N loading also favours the development of
a heterotrophic bacteria population and producing a lower nitrifying population.
Christensen and Harremoes (1978) have explanied how it is to be expected
that nitrification in the attached growth treatment process, under a high organic carbon
loading will not occur in the upper part of the trickling filter, nor on the first disks of a
rotating disk unit.
It may be assumed that in the upper layer, the nitrifying population will lose in
the competition with the heterotrophic bacteria, and carbonaceous matter only will be
removed. In the lower part of the trickling filter and at the last disk unit, the ammo-
nium-N loading is now high, compared with the organic loading, and, therefore the he-
terotrophic bacteria will be suppressed by the nitrifying bacteria. Nitrification will
consequently occur there.
A few models have been developed involving the competition between
heterotrophic and nitrifying bacteria (Harremoes, 1982; Wanner and Gujer 1984). All
of these models, developed recently, have predicted that the fraction of nitrifiers in
relation to the heterotrophic population is greater in the inner layer (near the surface
of the media) than in the outer layer of biofilm.
There are many types of competition between two or more microbial
populations. Competition occurs when the component populations are restricted in
either their growth rates or their final population sizes, as a result of a common
dependence on an external factor.
96
Competition can occur in either a closed culture, where growth is ultimately
limited by the availability of a particular growth resource, or in an open culture (as a
waste water plant), where growth is continuously limited. In open culture systems, as
in a waste water plant, it is inevitable that those populations which are the least
competitive, are eliminated from the growth environment. In this case the saturation
constant Ks,", usually becomes the most important factor determining the outcome of
com pet it ive growth.
Figure 3.9 shows different systems with competition between organisms A and
B. Organism B is initially a minor population compared to A.
The dilution rate of organisms, D is used to predict the washout of organisms from a
system plant.
Theoretically, if the growth rate p > D, then ds/dt (the substrate removal per
unit of time) is negative and the growth limiting substrate concentration decreases. The
biomass concentration is increasing under this condition.
If the growth rate p c D, then ds/dt is positive and the growth-limiting substrate
concentration increases, and the biomass concentration decreases.
Finally, if p = D, then ds/dt = 0, and the growth limiting substrate concentration
reaches a constant, steady-state value at the same time as the biomass concentration.
There are two basic cases to consider in assessing whether or not the growth
of population B is more or less competitive than that of the established population A,
where neither of the two organisms are limited by the substrate.
For the new population B to succeed in becomming greater than population
A, dXB/dt from the Monod equation (3.1 1) has to be positive. This can be achieved,
if pB > D, and pertains if either P,.,,,,,~ (the maximum growth rate for organism 8) >
P,,,,,,~ (Fig. 3.9a) or Ks,B c Ks,A (Fig. 3.9b). It must be noted, however, that it is the
combined effect of these which is important, in determining whether or not organism
B is more competitive than organism A. Fig. 3 . 9 ~ illustrates the situation in which
P,,,,~ p,,,,A, but KsB > Ks,A. For this pair of organisms, at any growth-limiting concentration, organism B is the more competitive, sustaining a higher growth rate
than organism A at all substrate concentrations.
Initially, the growth rate of organism B is determined by the steady-state
conditions established by organism A ; that is at a dilution rate D, the growth limiting
substrate concentration sA. Gradually, as the proportion of the two populations begins
to change in favour of population B, s begins to decrease and tend towards sA (see
97
Fig. 3.9a and 3.9b) which is the growth-limiting substrate concentration, which
supports a growth rate of pB = D. At this substrate concentration dSA/dt must be
negative, and accordingly population A is unable to grow at the imposed dilution rate
and must continue to be washed out of the culture vessel.
The opposite situation is that population B does not replace population A, if
pB c D and so dXB/dt is negative, a situation which results if either pmax,B < p,,,ax,A
(Fig. 3.9d) or K, > Ks,A (Fig. 3.9e).
Table 3.15 Comparison of parameters of heterotrophs and autotrophs (nitrifier)
determining bacterial population dynamics (Fruhen et a/. 1991).
Parameter Symbol Value
Heterotrophic bacteria maximum growth rate, d-’ pH,max 4 9 0
Heterotrophic bacteria decay coefficient, d-’ bH 0,15
Heterotrophic yield coefficient, g/g-’ Y, 0,57
Autotrophic bacteria maximum growth rate, d-’ pN,max 0983
Autotrophic bacteria decay coefficient, d-’ bN 0,05
Autotrophic yield coefficient, g/g-’ YN 0,24
The parameters presented in Table 3.15 show that both p,.,,, and K, for the hetero-
trophic population favour heterotrophic growth. Supplying a treatment plant with both
heterotrophs and nitrifier (autotrophic bacteria), it is therefore important to stock the
plant with a high nitrifying biomass X,, so the nitrifying population initially dominates
the plant. A combination of high nitrifier and a limitation of heterotrophic substrate may
be necessary.
To establish condition for a consistent nitrification it is therefore important that
the specific nitrifier growth pn is higher than the maximum heterotrophic growth ph,
assuming pH and DO do not limit the growth of the nitrifier.
This can be expressed in the following terms:
98
I.rmB
I"mA
D
a P
b B c
S1 KSA
KSB
'SA 'SB
fimA 1 I' 'SB KSA
51
B
C d
'SA
KSB
e
f - S
'SA 'SB
Fig 3.9 The various possible Monod relationships between two organisms, A and B,
used to predict the outcome of free competition between them under conditions of
growth limited by the substrate. After Slater and Bull (1978).
99
(3.39)
where: pLn = maximum growth rate of the nitrifying population.
p,, = growth rate of the herterotrophic population.
Reduced DO or pH can act to depress the growth rate of the peak nitrifier
pmax," and cause a wash out situation. A new growth rate pobs will then be the peak
nitrifier growth rate. The Monod Equation for this special condition is presented in EPA 1975:
*(1-0,833(7,2-pH)) KS,,+DO
c(obS=Pmx, n * (3.40)
where: pobs = maximum possible nitrifier growth rate under environmental conditions
of T, pH, DO and S>> K,.
To "correct" the calculations for the competition between the nitrifier and the
heterotrophic bacteria in the application of biological treatment, Lawrence and McCarty
(1968) introduced the concept of a safety factor (SF). A conservative safety factor is
recommended to minimize process variation caused by pH extremes, low DO,
fluctuation of substrate, and toxicants.
The growth rate can be expressed in reciprocal form in terms of a solid
retention time.
(3.41)
where qC = solids retention time in days.
100
1 DOUBLINGTIME 4) =-= [ = P 1112
(3.42)
Equation (3.42) is useful from the standpoint of process design.
The safety factor was defined as the ratio of the minimum retention time for
solids. The safety factor can also be related to the nitrifier growth rate.
(3.43)
where +obs = the minimum retention time for solids in days for nitrification at a given
pH, T and DO.
EPA 1975 proposes that the safety factor should equal or exceed the ratio of
peak load expected in the suspended growth nitrification system.
Today the safety factor approach is rarely used in the literature, but it is
absolutely necessary to use some form of safety factor in designing biological
nitrification plants, because the knowledge of the risk of introducing more species of
bacteria into the same system is still very limited.
Today, therefore, too many treatment plants still show too many differences
in their efficiency of nitrogen removal.
101
3.13 Effect of Inhibitors on Nitrification Nitrifiers are slow-growing organisms and they are accordingly particularly
susceptible to toxicants. Certain heavy metals and organic compounds are toxic to
nitrifiers. The presence of toxic compounds causes a change in the environmental
conditions for the nitrifying population, and they are therefore, a threat to any
nitrification plant.
Tomlinson eta/. (1966), however showed that nitrifiers are capable of adapting
to almost any toxic substances, when the toxic compound is consistently present at
a concentration higher than the concentration of the toxic compound that would cause
sludge discharge of the plant. Most toxic compounds in municipal systems stem from
industrial dumps or urban storm water inflow.
The possibility of a toxic inhibition must be recognized in every design of
nitrification systems. Either implementation of source control programs or inclusion of
toxicity removal processes upstream may be required, particularly in cases where
significant industrial discharges are tributary to the collection system.
It is therefore important to understand the difference between long-term and
short-term toxic inhibition. Figure 3.10 shows the difference in nitrification efficiency,
applying a long-term or a short-term inhibition with a toxic substance. This difference
is brought about because nitrifying bacteria are capable of developing adaptation to
most toxic compunds especially during a long-term contact.
Any inhibition of the nitrification process results in a decrease in the maximum
specific reaction rate of the nitrifying organisms. A change in the maximum specific
reaction rate can be compensated for by a longer solid retention time in a waste water
plant. If we suppose that for a specific plant an SRT (solids retention time) of 8 days
were required for an efficient nitrification and carbonaceous removal in a single
process; and if, after the plant was built, a new waste flow containing an inhibitory
compound were added; and if the maximum specific reaction rate of the nitrifying
organisms was reduced by 40%, it would be necessary to increase the SRT to 8
days/0,40 = 12 days. Such a large increase in SRT might not be possible without
extensive plant modifications, and when carried out, it might harm the heterotrophic
population.
Today, unfortunately only very little is known about the influence of different
groups of toxic substances on nitrifiers. Almost nothing is known about the consequen-
102
ces, when two or more toxic substances are present at the same time. It is, therefore
difficult to predict how a toxic compound or a number of toxic compounds will change
the biomass concentration in a plant. Investigators should in future study this field
carefully, because it would be of benefit to and facilitate the daily maintenance of any
type of nitrifying plant.
Nitrification Efficiency 1 100%
8 5%
10% Time
Fig 3.10 Differences in nitrifying efficiency, comparing long- and short-term effects of
a toxic substance.
The reduction of maximum specific growth rates which results from the effect
of environmental parameters on enzyme reactions can be expressed by different
models of enzyme inhibition.
An enzyme inhibitor is a compound which acts to reduce the rate of an
enzymatically catalysed reaction by binding with either the free enzyme E and/or with
103
the enzyme-substrate complex ES as shown in Table 3.16. Types of enzyme
inhibition can be classified (following Grady and Lim 1980) into five groups for
reversible inhibitors. Reversible inhibitors are inhibitors where the activity of the
enzyme returns to normal, when the inhibitor is removed.
1. Competitive inhibition.
An inhibitor which is classed as competitive competes for the same active sites as the
substrate.
2. Uncompetitive inhibition.
An uncompetitive inhibitor binds with the enzyme-substrate complex to form an
inactive enzyme substrate-inhibitor complex which cannot undergo further reaction to
yield the product.
3. Non-competitive inhibition.
A non-competitive inhibitor can combine with both free enzyme and the enzyme
substrate complex.
4. Substrate inhibition.
When their concentrations are very high, some substrates will bind with the enzyme
substrate complex as well as with the free enzyme.
5. Product inhibition.
The product may bind with the enzyme substrate complex, forming an unreactive
enzyme substrate product complex, ESP.
The mechanisms and inhibition-model of these different types are shown in
Table 3.16 and Fig. 3.11. The figures show the inhibition models for competitive,
uncompetitive and non-competitive inhibition.
Transforming the Michaelis-Menten expressions into one of the linear
equations, i.e. Lineweaver-Burke, makes it easier to quantify the various parameters
that are affected by the inhibitior. A specific pmax and KS," can therefore easily be
distinguished for each condition and type of inhibitor.
Krittiya (1 984) used the Lineweaver-Burke plot to estimate the effect of sodium
104
ion on the nitrite oxidizing bacteria, as shown in Fig. 3.12. Results showed that the
sodium ion inhibiton on the nitrite oxidizing process was categorized as a non-
competitive type and the inhibition constant Fnhib, was 2,0 g/t as Na'.
Visut (1 985) made similar experiments with sodium inhibition on ammonium
oxidizing bacteria and proposed the following expression for the inhibitory effect of
sodium ion on oxidizing bacteria:
(3.44)
where
p = specific growth rate, d-'
S, = ammonium concentration mg/l as N I = inhibitor concentration g/l as Na'
pmax = maximum specific growth rate
K,,., = saturation constant
Kinhib. = inhibition constant
Kd = decay rate, h-'
Visut (1 985) found the following experimental values:
kax = 0,0313 h-', Ks.n = 11,6 / 13,5 mg/l as N, Kinhib, = 6,64 mg/l as Na'
and Kd = 3,l h-'.
Hockenbury and Grady (1977); Beg et a/. (1982) ; Akai et a/. (1983) and
Hassan et a/. (1988) have all used the rate expression for enzyme inhibition in their
studies of effects of inhibitors in the nitrification process.
105
Competitive Uncom pet it ive Noncompetitive
a) l u
0 KC KS
0
a, b and c are Monod plots
- 1 /Kb -1 IK; '"h
d. e and f are Lineweaver - Burk olots
. " 9. h and i are Hanes plots
0 0
j, k and I are Hofstee plots
Fig 3.11 Typical plots for identifying the types of enzyme inhibition. The solid cu
represent the uninhibited cases, the dashed curves the inhibited cases.
(Ohgaki and Wanttawin 1990).
106
Symbol Conc. Na' Correlation in g /I
0 0.137 0.99 X 1.052 0.96
2.630 0.98 1 1N h I/mg 0 5.260 0.98
m g-'
-0 .Xl a0 01 02 Q3 Q4 05 06 0.7 08 0.9 10 1.1 12
Fig 3.12 The Lineweaver-Burke plot for identifying the type of inhibition of sodium ion
concentration for nitrite oxidizing bacteria (Krittiya 1984).
107
Hassan et a/. (1988) evaluated the performance of a packed-bed biological
reactor in the presence of inhibitors, following either complete or partial modes of
competitive, non-competitive, mixed or uncompetitive inhibition. For all types of
inhibition, it was found that an increase in the inlet substrate concentration reduces the
steady-state conversion in the reactor. The increase in the value of the parameter l/q, which indicates the specific action of the inhibitor, increases the conversion for the
partially competitive and non-competitive inhibition mode, while it reduces that for
product inhibition.
Substances inhibitory to nitrifying bacteria or nitrification.
Some research has been carried out by microbiologists on the effect of specific
organic and inorganic compounds on pure cultures of Nitrifiers. Table 3.17 show the
results presented by Blum and Speece (1991) for nitrosomonas toxicity due to organic
compounds for IC,, concentration of less than 20 mg/l.
More compounds have been found to be inhibitory to ammonia oxidation by
Nitrosomonas species than to nitrite oxidation by Nitrobacter species. No explanation
for this has so far been given in the literature.
Most inhibitory compounds in a waste water treatment plant are present in the
range of mg/l and even some in the range of pg/I, and may, therefore, be difficult to
detect analytically when they are present in waste water.
Only a few studies have been made on nitrification inhibition in activated
sludge; the most complete one was made by Tomlinson et a/. (1966). Five of the
compounds included in the list are among the compounds most used by industry. Two
of these, chloroform and phenol, are general inhibitors of bacterial metabolism.
Most of the very potent inhibitors in the nitrification process are sulphur-
containing compounds; they can act as metal-chelating compounds, and thus inhibit
enzymes requiring metals for activation (Dixon et a/. 1964; Downing eta/. 1964).
No reports have been found on inhibition of ammonia oxidation induced by
aliphatic or aromatic amines. Hockenbury and Grady (1977) pointed out that the
inhibitory effect of nitrogen-containing compounds was caused by competition with
ammonia for the active site on an enzyme, although no supporting evidence has been
given in the literature. Likewise, compounds, similar in structure to nitrite, have been
hypothesized to be inhibitory because of their competitive effects, although only few
108
Table 3.16 Different types of inhibition models.
?ypes of Mec hanistn Hate expression Michaelis-Menten form nhibition
- -~
k l k2 PrnS Competitive E + S - E S - E + P P = p'm = Pm
k - 1 (KS( l +I/kl)+S) K', = Ks( 1 +I/KI) where KI = k31k.3 k3
k - 3 E + I - E l
k l k2 PrnS P = -___-
where Kl = k31k.3
Un- E + S - E S - E + P Competitive k - 1 (K, + S ( l + l / k l ) )
k3
k - 3
k l k2 PrnS
k3
k - 3
k4
k- 4
k l k2 k2EOS
E + I - E l
P = -___- competitive k - 1 (Ks +S) (1 +I /k l ) )
where K I = kg/k.g +
Non- E + S - E S - E + P
€ + I - E l k4/k. 4
E S + I - E S I
Substrate E + S - E S - E + P CI = k. I (K, +S + S2/K',)
K', = K,
k3
k- 3 E S + S - S E S
where Kc 8 K'c are the disso- ciation constants for ES and SES respectively
k l k2 prns Prn Product * E + S - E S - E + P P = p'rn = 1 + PlKp
k - 1 (K, + S(l+Plkp))
where Kp = k3/k.3 Ks K' - ___ k3
k.3 E S + P - E S P - 1 +P/Kp
where pm = k2Eo (E, = initial enzyme concentration). K, = (k.l + k2)/kl
This is the simplest mechanism. other mechanlsrns could be hypothesized which would lead to alter- native rate expresstons
109
data have been presented by Hockenbury and Grady (1977).
An investigation conducted by Hockenbury and Grady (1 977) concluded that
p-Nitrobenzaldehyde, p-nitroaniline and n-methylaniline were all inhibitors of nitrite
oxidation by Nitrobacter species when present in a concentration of at least 100 mg/l.
Dodecylamine, aniline and n-methylaniline were potent inhibitors of ammonia oxidation
by Nitrosomonas species, causing 50% inhibition at concentrations of less than 1 mg/l.
Aniline, ethylenediamine, hexamethylenediamine and monoethaniolamine are
commonly used organic substances, known to inhibit ammonia oxidation by
Nitrosomonas species. Ammonia exerts substrate inhibition on its own oxidation, and
the inhibition of ammonia oxidation by aniline, dodecylamine and ethylenediamine is
niether competitive nor non-competitive. Hockenbury and Grady (1 977) proposed that
it is related to substrate inhibition. The inhibitory effect of aniline, dodecylaniline and
ethylenediamine increases as the concentration of ammonia nitrogen in the medium
is increased. The results presented by Hockenburg and Grady (1977) are shown in
Table 3.18. The results are divided into two levels, compound concentrations yielding
50 and 75 YO inhibition of the nitrifying culture.
Table 3.1 9 show the ammonium nitrogen and nitrate nitrogen concentration
range for nitrobacter inhibition as function of pH at 20°C. The results are in
accordance with knowledge of the ionisation of both ammonium nitrogen and nitrate
nitrogen . Neufeld eta/. (1986) presented different equations for the inhibition of phenolic
compounds on the nitrification and discussed the influence of free cyanide and
complexed cyanide compounds on the nitrification kinetic.
Figure 3.13 shows that even small amounts of free cyanide in solution inhibit
the biological rate of nitrification. The relationship of the maximum reaction rate V,,,
and the free cyanide concentration was found to follow the equation:
(3.45)
where [CN-] is the free cyanide concentration in mg/l at pH = 8,O. It is important to
know the actual pH in the waste water environment and correct the [CN-] to pH = 8
using the proposed equation.
110
Table 3.17 Inhibitory effect of organic compounds with an IC, value of less 20 mg/l,
on pure cultures of Nitrifiers.
Organic Compound IC,, Concentration mgA
4-Aminophenol
3-Chlorophenol
2-Aminophenol
2-Bromophenol
2,3-Dichlorophenol
2,3.6-TrichlorophenoI
1,3-DichIoropropene
5-Chloro-1 -pentyne
2,3-Dichlorophenol
1 ,3-Dichloropropene
Chlorobenzene
4-Chlorophenol
2,4-Dichlorophenol
Trichloroethylene
4-Brornophenol
1,l-Dichloroethane
2,3,5,6-Tetrachlorophenol
1,1,2,2-TetrachIoroethane
1.1,2-Trichloroethene
2.2,2-Trichloroethanol
4-Nitrophenol
2-Chlorophenol
3,5-Dichlorophenol
2,3,5-Trichlorophenol
2.4.6-Tribrornophenol
Resorcinol
2,4,6-Trichlorophenol
Pentachloroethane
2,6-Dichlorophenol
1,1,1,2-Tetrachloroethane
1,2,4,5-TetrachIorobenzene
2-Nitrophenol
Benzene
1 ,&Dichloropenthane
1,2,3,4-Tetrachlorobenzene
0,07
0,20
0,27
0,35
0,42
0.48
0,59
0.61
0,67
0,71
0,73
0,79
0,81
0883
0,91
1.30
1940
1,90
2,00
2.60
2,70
3.00
3,90
7,70
7.80
7,q0
7,90
8,10
8,70
9.80
11 ,oo 13,00
13,00
20.00
Source: Blum and Speece (1991)
111
Table 3.18 lnhibitoriy effect of organic and inorganic compounds in pure Nitrobacter
culture on the nitrification process.
Compound Concentration (m@) at approximately 75%
inhibition ~
Acetone' Allyl alcohol Allyl chloride Allyl isothiocyanate Benzothiazole disulfide Carbon disulfide' Chloroform' &resol Di-ally1 ether Dicyanidiam ide Diguanide 2,4-Dinitrophenol Dithio-oxamide Ethanol' Guanidine carbonate H ydrazine 8-Hydroxyquinoline Mercaptobenzothiazole Methylamine hydrochloride Methyl isothiocyanate Methyl thiuronium sulfate Phenol' Potassium thiocyanate Skatole Sodium dimethyl dithiocarbamate Sodium methyl dithiocarbamate Tetramethyl thiuram disulfide Th ioacetam ide Thiourea Trimethylamine
'In the list of industrially significant chemicals.
2 000
180
38 35 18 12,8
100 250
50 460
1,1 2 400
16,5 58 72,5 3,O
0 3 6 3 5,6
300 7
13,6 0 3
30 0,53 0,076
193
1 3
1 550
118
112
Table 3.18 (continued)
Compound Concentration (mgA) at approximately 505% inhibition
~
Dodecylamine Aniline n-Methylaniline Ethylenediamine Napthylethylenediamine-di-HCI 2,2 Bipyridine p-Nitroaniline p-Aminopropiophenone Benzidine-di-HCI p-Phenylazoaniline Hexamethylene diamine p- N it ro benzaldeh yde Triethylamine N in h ydrin Benzocaine Dimethylgloxime Benzylamine Tannic acid Monoethanolamine
Source: Hockenburg and Grady (1977)
< 1 < 1 < 1 15 23 23 31 43 45 72 85 87 127 > 100 > 100 140 > 100 > 150 > 200
Compound Inhibition Concentrations
Phenol 100 mg/l Vitamins:
Riboflavin 50 mg/l Thiamine 5 mg/l
Amino acids: L-Lysine L-Threon ine L-Histidine L-Valine L-Arginine L-Meth ion ine 4 mg/l
2-Chloro-6-trichloromethyl-pyridine 10 mg/l Diethyldithiocarbamate M
Tannin M Tannin derivatives M
Source: Shrama and Ahlert (1976)
Methyl Blue M
113
Table 3.19 Ammonium Nitrogen and Nitrate Nitrogen Concentration Range for Niirobacfer Inhibition as function of pH (T = 20 C).
PH
NH4+ - N NO, - N
Range, mg/l Range, mgf/
210 - 2100 30 - 330
70 - 700 88 - 1050
20 - 210 260 - 3320
7 - 70
2 - 20
1.00
0.5 0
0.10
0.05
0
g NHdg VVS day t
mg/l
Fig 3.13 The influence of [CN-] on the nitrification rate. After Neufeld eta/. (1986).
114
Complexed cyanide was also found to cause a decrease in the maximum
reaction rate for nitrification processes in accordance with the following equation:
0.1
where [CN] is the complexed cyanide concentration in mg/l.
Using thiocyanate, Fig. 3.14 shows that a plot of V,,, versus the thiocyanate
concentration yield a constant reaction rate up to a thiocyanate level of about 236
mg/l. Above this value the reaction rate declined according to the following equation:
'\ -
[SCN] mg/l
log ( V,,) =1,9 1-0,7 7 log [SCN (3.47)
g NH,/g VVS day t
I I I 1 c 1.0 10.0 100.0 1000.0
0 0
Fig 3.14 The influence of thiocyanate on the nitrification rate. After Neufeld et a/.
(1986).
115
Beg and Atiqullah (1 983) conducted experiments with a fixed film reactor and
showed that As3+, Cr6' and F- were reversible non-competitive inhibitors, having
inhibitor constants of 305, 65,3 and 1276 mg/l, respectively. Also, interaction between
the three inhibitors showed that it did not affect the zero-order kinetic of nitrification
with respect to the NH,' - N substrate concentration.
When the concentration of a strong inhibitor was kept constant, and that of a
weaker one was varied, two phenomena were observed (Beg and Atiqullah 1983).
Firstly, for a shock dose at lower concentrations of the stronger inhibitors such as
chromium and arsenic, the degree of inhibition was increased with the increase in the
concentration of the weakest inhibitor, fluoride. This tendency was more pronounced
at lower concentrations of the weaker inhibitor than at higher concentrations. Secondly,
at higher concentrations of the stronger inhibitors, chromium (> 40 mg/l) and arsenic
(> 300 mgll), the degree of inhibition initially decreased to a minimum value, and then
increased with the increase in the concentration of the weaker inhibitor.
Beg and Atiqullah developed the following rate expression for the shock load of As*,
Cra and F-, in pairs:
(3.48)
where; I is the concentration of the stronger inhibitor I,, that of the weaker one in the
pair.
Table 3.20 show a list of inorganic compounds that lead to inhibition of the
nitrification process. It is important to remember that the inhibition of inorganic
compounds is dependent on the actual pH in the environment, because it is often the
free inorganic compound, for instance copper ion, that inhibits the nitrification process.
As an example, the free copper ion concentration increases with decreasing pH.
116
Table 3.20 Inorganic compounds that lead to inhibition of the nitrification process.
Compound Concentration mgL1 Type of Plant References
CN- Toxic in all conc. Coke plant Neufeld et a/.
waste water (1 986)
Fe(CN-),” 80 II I,
SCN- 236 II
As3+ 305
Cr6+
F-
Ag+
Zn2+
cu+
N i+
Hg+
S2-
65,3
1267
5
Toxic in all conc.
Fixed bed
reactor
Plastic media
trickling filter
Fixed bed pilot
pilot plant.
II
Beg and Atiqullah
(1 983)
I,
II
USPHS (1965)
USPHS (1 965)
USPHS (1965)
USPHS (1965)
USPHS (1965)
Hjuler (1992)
(unpublished)
117
1.6
1.4
1.2
1.0
0.8
0.6
0 .L
0.2
mg/l per minute
A o 0
inhibitor Conc. mg/i
0 I 1 1 1 1 1 -
0 80 160 24 0 320 400 480
Figure 3.15 Effects of As3+, Cr6+ and F- on the nitrification rate. After Beg and Hassan
(1 987).
118
4. PROCESS CHEMISTRY AND BIOCHEMISTRY OF
4.1 Introduction The biological process of denitrification involves the reduction of nitrate
nitrogen, NO,, to a gaseous nitrogen species. The gaseous product is primarily
nitrogen gas, N,, but may also be nitrous oxide, N,O, or nitric oxide, NO. Gaseous
nitrogen is not readily available for biological growth, thus denitrification converts
nitrogen to a harmless form which has no significant effect on the environment.
Some confusion has arisen in the terminology used in the literature. The
process has been termed anaerobic denitrification. The principal biochemical pathways,
however, are not anaerobic, but merely minor modifications of aerobic biochemical
pathways. The term anoxic denitrification is therefore preferable, as it describes the
environmental condition involving the absence of oxygen, without implying the nature
of the biochemical pathways.
Denitrification is of interest because:
1. It is a major mechanism for loss of fertilizer nitrogen in agriculture, resulting in a
decreased efficiency of the fertilizer.
2. It is of great potential application in the removal of nitrogen from high-nitrogen waste
materials such as animal residues.
3. Many factors affect the accumulation of denitrification intermediates, such as N,O,
but only very few attempts have been made to develop a unifying explanation of the
different intermediates.
4. Denitrification is the mechanism by which the global nitrogen cycle is balanced.
5. Most ground water resources of the world are facing a major nitrate contamination,
which may result in infant methemoglobi . 6. It is a method for the removal of nitrogen from waste water.
The contribution of waste treatment systems to atmosheric N,O is of some
concern, because N,O is involved in the stratospheric reactions, which result in the
depletion of ozone, but little information is available. It is noteworthy, however, that
fermentation, waste water acclimated to or supplemented with nitrate, released small
quantities of N,O during denitrification, whereas the waste water adapted to or
119
supplied with nitrite, produced none.
Nitrate contamination of ground water resources is becoming an ever
increasing problem. Because of the adverse effects on health associated with nitrate
in drinking water, and the concerns regarding diminishing water quality, the interest in
nitrate removal technologies increases.
The drinking-water standard set by the U.S. Environmental Protection Agency
(EPA), for nitrate is 10 mg/l as nitrate-nitrogen. The European Economic Community
has a standard of 50 mg/l as nitrate (1 1,3 mg/l nitrate-nitrogen).
4.2 Types of Bacteria Accomplishing Denitrification As distinct from nitrification, a relatively broad range of bacteria can accomplish
denitrification. Genera of bacteria that are known to contain denitrifying bacteria include
Pseudomonas, Micrococus, Archromobacter, Thiobacillus, and Bacillus (see Table 4.1).
These bacteria are biochemically and taxonomically very diverse. Most are he-
terotrophs and some utilize one-carbon compounds, whereas others grow auto-
trophically on H, and CO,, or on reduced sulphur compounds. Most of the mentioned
bacteria possess the enzyme reductase necessary to reduce nitrate to gaseous
nitrogen. But some lack the nitrate reductase enzyme and are termed nitrite
dependent; and others lack N,O reductase and thus yield N,O as the terminal product.
Still other organisms possess N,O reductase but cannot produce N,O from nitrate or
nitrite. These different groups of bacteria also accomplish nitrate reduction by what is
known as a process of nitrate dissimilation, whereby nitrate or nitrite replaces oxygen
in the respiratory process of the organism under anoxic conditions. Because of the
ability of these organisms to use either nitrate or oxygen as the terminal electron
acceptor while oxidizing organic matter, these organisms are termed facultative
heterotrophic bacteria.
Surprisingly, most of the organisms known to denitrify are not strict anaerobes,
but rather facultative organisms, which under anoxic conditions use nitrate as a final
electron acceptor. The sludge in combined nitrification and denitrification design
processes is alternatively exposed to aerobic and anaerobic conditions, and because
the denitrifying bacteria are facultative, the change of an oxic environment will provoke
only minor adaptation problems.
120
Table 4.1 Genera of bacteria which are abundant in sewage and capable of performing denitrification.
Genera Abundant in sewage Species within the genera are denitrifiers NO, -+ N2
Only NO,- + NO2-
Ach rorn obacter
Aerobacter
Alcaligenes
Bacillus
Flavobacteriurn 2
!?
Micrococcus
Proteus
Pseudomonas
Van Gils (1964)
Harris ef a/. (1927)
Doelle (1 969), Payne (1 973). Smith et a/. (1972)
Doelle (1 969)
Van Gils (1964), Harris et a/. (1927)
Smith etal. (1972)
Van Gils (1964), Jasewicz and Porges (1 956) Payne (1 973)
Jasewicz and Porges (1 956) Payne (1 973), Porra and Lascelles (1 965)
Harris et a/. (1 927) Payne (1 973)
Jasewicz and Porges (1956) Best and Payne (1965). Fewson and Nicholas (1961),
Payne (1973), Smith etal. (1972) Fry (1955),
Source: Henze Christensen and Harremoes (1 977).
The fact that common sewage bacteria are denitrifiers makes it simple to
create an appropriate environment for the denitrification process. All that is needed is
the presence of nitrate, an electron donor (carbon source) and an anaerobic
environment. A more specialized knowledge of species of bacteria is hardly necessary
in most cases. Exceptions are where a special carbon source, such as methane, is
used, as only very few bacteria can metabolize methane under anaerobic conditions.
Denitrifying bacteria can be identified according to the methods described in
the Standard Methode (1985). Other are listed in Table 4.2.
4.3 Biochemical Pathways Denitrification is a two-step process in which the first step is a conversion of
nitrate into nitrite. The second step carries nitrite through two intermediates to nitrogen
gas. This two-step process is normally termed "dissimilation".
Each step in the denitrification process is catalysed by a separate enzyme system.
Denitrifiers are also capable of an assimilation process whereby nitrate
(through nitrite) is converted into ammonia. Ammonia is then used for the nitrogen
requirements of the bacteria cells. The step or steps, from nitrite to hydroxylamine are
not fully known.
Table 4.2 Methods for the identification of denitrifying bacteria.
Method Reference
Chromatographic techniques Payne (1 973)
Tood and Nuner (1973)
MPN-technique Tood and Nuner (1973)
Measurements of the
enzymatic activity plates Lenhard (1 969)
122
If ammonia is already present, for example in a nitrification plant, assimilation
of nitrate need not occur to satisfy cell requirements.
The transfer of electrons from the carbon source (the electron donor) to nitrate
or nitrite (the electron acceptor) to promote the conversion into nitrogen gas, will be
discussed in detail in Section 4.4. It involves the "electron transport system" of the
denitrifiers and consists of the release of energy from the carbon source for the use
in the growth of the organism. This electron transport system is identical to that used
for respiration by organisms oxidizing organic matter aerobically, except for one
enzyme. Because of this very close relationship, many facultative bacteria can shift
between using nitrate (nitrite) or oxygen rapidly and without difficulty.
Most investigators consider oxygen an inhibitor in the denitrification process.
But some species have been reported to denitrify in systems with oxygen tension still
as high as 0.2 atm. Table 4.3 show the metabolic processes in biological denitrification.
There is also evidence that nitrification and denitrification may occur
simultaneously in soil or when applying special porous media, as for example
clinoptilolite. Though not fully explained, these phenomena may occur in anaerobic
micro-zones in otherwise aerobic systems (Masuda eta/. 1987, 1990; Watanabe 1990;
Halling-S~rrensen and Hjuler 1992; 1993).
Many nitrate-reducing bacteria exhibit both dissimilatory and assimilatory
behaviour. From an engineering point of view the ratio between dissimilated and
assimilated nitrogen is of interest, as it is more desirable to produce nitrogen gas than
to produce organic nitrogen bound in bacteria. Christensen and Harremoes (1977) and
Painter (1970) indicate the yield coefficient for denitrifying bacteria Ydenit. to be
approximately 0,4 mg VSS per mg NO,- - N. If the nitrogen content in the organic
matter is 10'30, then 0,04 mg N is assimilated for every 1 mg NO; - N converted into
nitrogen gas.
An electron transport system for nitrate reduction is shown in Table 4.3,
example 3. The steps from the electron donor to the cytochrome are always identical,
while the final steps depend upon the final electron acceptor (nitrate, nitrite etc.).
Different species of bacteria may have slightly different electron transport
systems, in particular in respect to quinone and cytochrome (Painter 1970).
For each of the steps in the dissimilatory nitrate reduction sequence a
reductase enzyme has been isolated (Mudrack 1971).
123
Table 4.3 Metabolic processes in biological denitrification.
1 : Dissimilatory nitrate reduction (denitrification).
NO, NO, + NO + N2O + N2
2: Assimilatory nitrate reduction (synthesis).
NO,' + NO2- + X + NH,OH + Org. N
3: Possible electron transport system of the first step of denitrification.
e- donor + NAD + FAD + Quinone + Cytochrome + Nitratereductase + NO,-
4.4 Energy and Synthesis Relationship The use of oxygen as the final electron acceptor is more energtically favored than
the use of nitrate. By oxygen respiration the energy yield per mole of glucose is 686
kcal/mole and by nitrate dissimilation the energy yield per mole glucose is only 570
kcal/mole.
The greater free energy released for oxygen favors its use whenever it is available.
Therefore, denitrification must be conducted in an anoxic environment to ensure that
nitate, rather than oxygen, serves as the final electron acceptor.
Methanol, ethanol, acetic acid, have been most frequently used as the electron donor
in experiments, rather than glucose.
Using methanol as an electron donor and neglecting synthesis, denitrification can
be represented as a two-step process as shown in equations (4.1) and (4.2).
First step:
NO,- + 1/3 CH,OH => NO2- + 2/3 H,O (4.1)
Second step:
NO2- + 0,5 CH,OH 0,5 N2 + 0,5 CO, + 0,5 H20 +OH-
(4.2) The overall transformation is obtained by addition of equations (4.1) and (4.2)
=>
yielding equation (4.3).
124
NO3- + 5/6 CH30H => 0,5 N, + 5/6 CO, + 7/6 H,O + OH- (4.3)
Methanol serves as the electron donor in this equation and nitrate as the electron
acceptor. This can be shown by splitting equation (4.3) into the following oxidation-
reduction reactions.
Electron acceptor:
NO; + 6H' + 5 e - => 0,5N, + 3H,O (4.4)
Electron donor:
5/6 CH30H + 5/6 H,O => 5/6 CO, + 5 H+ + 5 e- (4.5)
It is clear from equations (4.4) and (4.5) that nitrate gains electrons and is reduced
to nitrogen gas, which is the electron acceptor. The carbon source, in this example
methanol, loses electrons and is oxidized to carbon dioxide, therefore it is the electron
donor.
As mentioned in Section 3.4, these reactions take place in the context of the
carbonic acid system. Equations (4.4) and (4.5) can be modified to reflect the fact that
the hydroxide (OH-) produced reacts with carbonic acid (carbon dioxide) to produce
hydrogen carbonate alkalinity.
Nitrogen dissimilation and growth in denitrifcation reaction:
Nitrate to nitrite:
NO,- + 1/3 CH30H => NO,- + 1/3 H,O + 1/3 H2C03 (4.Q
Nitrite to nitrogen gas:
NO,- + 0,5 CH,OH + 0,5 H2C03 => 0,5 N, + HC03- + H20(4.7)
Nitrate to nitrogen gas:
NO3- + 5/6 CH30H + 1/6 H,CO, => 0,5 N, + 413 H,O +HC03- (4.8)
125
Synthesis denitrification:
14 CH,OH + 3 NO3- + 4 H2C03 => 3 C5H702N + 20 H20 + 3 HCO, (4.9)
Combined dissimilatory-assimilatory equations for denitrification (after McCarty et
al. 1969):
Overall nitrate removal:
NO,- + 1,08 CH,OH + 0,24 H2C03 => 0,056 C5H7N02 + 0,47 N2
(4.10)
+ 1,68 H,O + HCO,
Overall nitrite removal:
NO2- + 0,53 H2C03 + 0,67 CH,OH => 0,04 C5H7N02 + 1,23 H20 + 0,48 N2 + HCO,
(4.1 1)
Overall deoxygenation:
0, + 0,93 CH,OH + 0,056 NO, => 0,056 C5H7N02 + 1,04 H20 + 0,59 H2C03 + 0,056 HCO,
(4.12)
Equation (4.1 2) is shown since if any oxygen is present, it will be used preferentially
before the denitrification.
The theoretical methanol requirement for nitrate reduction, neglecting synthesis is
1,9 mg methanol per mg nitrate-N (4.1). Including synthesis (equation 4.10) the
requirement is increased to 2,47 mg.
Similarly, calculation of methanol requirements for nitrite reduction and deoxygena-
tion allows a combined expression to be formulated for the methanol requirement.
Cm = 247 * NO3- - N + 133 * NO2- - N + 0,87 * DO (4.1 3)
126
Where
Cm = required methanol concentration mgl.
NO, - N = nitrate-nitrogen concentration removed mg/l.
NO, - N = nitrite-nitrogen concentration removed mg/l.
DO = dissolved oxygen removed mg/l.
The biomass X, rngl can be calculated similarly.
X, = 0,53 * NO, - N + 0,32 * NO, - N + 0,19 * DO (4.1 4)
For instance, for a NO3- value of 25 mg/l of nitrate-N, 0,5 mg/l nitrite-N and 3,O mg/l
dissolved oxygen, the methanol requirement can be calculated to be 64,l mg/l from
equation (4.13). The M/N ratio, which is the mg of methanol per mg of initial nitrate
nitrogen concentration, is therefore 237 (64,l / 25), which is only 4 percent greater
then the requirement for nitrate alone.
Most experimental data is expressed in terms of the C/N ratio, which is the mg of
carbon per mg of C per mg of initial nitrate-nitrogen concentration. The ratio includes
the requirements for nitrite and oxygen, which are usually small relative to the nitrate
requirement.
Values of the C/N ratio required for complete denitrification range from 1,5 to 5.
Table 4.4 show C/N ratio for different types of carbon sources used to perform
denitrification. It has been suggested that column denitrification systems require a
lower C/N ratio than suspended growth systems due to the higher concentration of
biomass maintained in the column systems.
Higher biomass levels produce longer solids retention times and reduce organism
yields due to increased endogenous metabolism. In turn this lower yield would result
in less carbon required for synthesis and reduced C/N ratio.
In general, a C/N ratio of 2 to 3 will enable "complete denitrification" (95 % removal
of nitrate) and this value may be used for design purposes when methanol is used as
the carbon source for denitrification. Fig. 4.1 show the C/N ratio using methanol as
carbon source as a function of the denitrification, in two different studies for submerged
filters. The dotted line is the theoretical C/N ratio needed for total denitrification.
127
Table 4.4 C/N ratio for different types of carbon sources used to perform denitrifi- cation.
Organic matter C/N oprimum Unit
as internal source 3-33 kg BOD/kg N 4-5 kg COD/kg N
in sludge
Methanol
Acetic acid
1,5-2,5 2,9-3,2
kg BOD/kg N kg COD/kg N
2,3-2,7 kg MeOH/kg N 33-4,1 kg COD/kg N
2,9-3,5 3,l-3,7
kg HAc/kg N kg COD/kg N
4.5 Alternative Electron Donors and the C/N Relationship As shown in section 4.3, (equations 4.1 and 4.2), the denitrification process needs
an electron donor to be accomplished.
A variety of compounds that can substitute for methanol as a carbon source have
been evaluated experimentally and described in the literature. Table (4.5) shows the
wide variety of carbon sources which have been used experimentally other than
methanol and internal carbon.
The selection of an electron donor depends upon three factors which will be
discussed in this section: availability of the electron donor, the reaction rate, and costs.
The combination of a high reaction rate and moderate costs is achieved by the use of
methanol.
Denitrification rates achieved with waste water organics, also called the internal
carbon source, are approximately one third of those achieved when methanol is
employed as the electron donor; this is because the availability of the electron donor
is one of the most important factors controlling the activity of the denitrifiers. If the
availability of the electron donor fluctuates, then the performance of the denitrification
will also fluctuate, yielding a lower denitrification rate.
Denitrification reactors must, therefore, be proportionately larger using an internal
carbon source than when methanol is used.
Volatile acids have also been used as a carbon source for denitrification. (Climen-
hage 1982). In studies of nitrate reduction in waste water generated in the manufacture
128
of nylon, is was found that a mixture of C, to C, volatile acids was very effective as
a carbon source for denitrification.
It is also possible to use inorganic compounds as electron donors. Hydrogen and
sodium sulphide have been used in these experiments (Kurt et a/. 1987).
Some of the alternative carbon sources cause greater sludge production than
others. About twice as much sludge is produced per mg of nitrogen reduced when
saccharose is used, than when methanol is employed, because the yield coefficient of
the bacteria using the first carbon source is greater.
On the other hand, acetone, acetate and ethanol produced similar quantities of
sludge to that produced when methanol was employed.
Methanol has certain advantages over carbon sources in waste water. It is free of
contaminants such as nitrogen, and can therefore be used directly in the process
without taking special precautions that must be made for the use of a system with an
internal carbon source. Using a external carbon source produces a consistent quality,
while waste water sources may vary in strength and composition, either daily or
seasonally, which complicates both process control and optimization. Use of waste
water sources will require regular assays of the source to check its purity, and strength
and its biological availability.
f h e disadvantage of using methanol is its cost, and this alone advocates the
necessity of economic comparisons of alternate carbon sources.
Denitrification is considered to be a heterotrophic process, conducted by microorga-
nisms that require a reduced organic substrate for energy and cell synthesis.
Heterotrophic denitrifying microorganisms can use a variety of organic carbon
sources, while most of the published reseach regarding the denitrification of water,
presumes the use of methanol, ethanol and acetic acid.
Figure 4.2 show the denitrification reaction rate as a function of temperature for
different carbon sources. The more easily degradable the carbon source, such as
methanol is, higher is the reaction rate. Heavily degradable endogenous carbon has
a low reaction rate, especially at low temperature.
129
100
50 A
0 0
0
% NO3- - N removal
_ - - -6-o- b-o- - --,-- a-o--- /,-3T;--o 0 oo
I I
/ I
I
I / g CH,OH/g NO, - N
I I 1 1 1 I I c
0 1 2 3 4 5 6
Fig. 4.1 C/N ratio using methanol as carbon source, in two different studies indicated as x and 0, as a function of the denitrification efficiency, for a submerged filter. The dotted line is the theoretical afnount
(Source: Henze and Harremoes 1978)
The stoichiometric relationships for these substrates have been formulated as
follows:
Methanol (Sherrad 1988):
5CH,OH + 6NO,- => 3N, + 5C0, + 7 H 2 0 + 6 0 H - (4.15)
6 N, + 10C0, + 9 H,O + 12 OH- (4.16)
4 N, + 10 CO, + 6 H,O + 8 OH-
Ethanol (similar to equation 4.3)(Richard et a/. 1980):
5 C,H,OH + 12 NO3- =>
Acetic acid (Frick and Ricard 1985):
5 CH,COOH + 8 NO3- => (4.17)
Glycol (Monteith eta/. 1980):
0,5 (CH20H),+ NO, =>
Formaldehyde (Monteith et a/. 1980):
0,5 N, +CO, + H,O + OH- (4.18)
1,25 HCHO + NO, => 0,5 N, + 1,25 CO, + 0,75 H,O + OH- (4.19)
lsopropanol (Montheith et a/. 1980):
0,278 C3H,0H + NO3- => 0,5 N, + 0,833 CO, + 0,611 H,O + OH- (4.20)
Fuse1 oil (as amyl alcohol) (Montheith eta/. 1980):
0,167 C5Hl10H + NO3- => 0,5 N, + 0,833 CO, + 0,5 H,O + OH- (4.21)
Dextrose (Montheith ef a/. 1980):
0,208 C6H1206 + NO,- => 0,5 N, + 1,25 CO, + 0,75 H,O + OH- (4.22)
Gaseous organic substrates, such as methane and carbon monoxide, can also be
used as substrates in denitrification. Among gaseous substrates, methane is one of the
most studied; but some contradictions remain in the literature regarding methane
metabolism. There is evidence that methane can be used as a terminal electron
acceptor by some denitrifiers (Davies 1973).
131
Other investigators have suggested that methane oxidation requires aerobic or
microaerophilic conditions, and that subsequent denitrification may be the result of a symbiotic relationship between two groups of organisms with different trophic
requirements (Yull-Rhee et a/. 1978).
It is likely that both phenomena occur, indicating two possible mechanisms for
methane utilization during denitrification. Fewer studies have been published involving
carbon monoxide, but there is evidence that it can be used as a substrate for
denitrification (Park and Hegeman 1984).
Stoichiometric relationships for methane and carbon monoxide utilization have been
proposed.
Methane (Barrenstein et a/. 1986):
5CH4 + 8 N O i + 8 H + => 5C0, + 4 N 2 + 1 4 H 2 0 (4.23)
Carbon monoxide:
2 NO3- + 5 C O + H,O => N, + 2OH- + 5C0, (4.24)
Denitrification can also be accomplished by autotrophic bacteria, which can use
hydrogen or various reduced-sulphur compounds as energy sources. Under autotrophic
growth conditions, no organic carbon sources are required, rather carbon dioxide or
bicarbonate is used as a carbon source for cell synthesis.
Paracocus denitrificans and Thiobacillus denitrificans can denitrify using hydrogen
and reduced-sulphur compounds, respectively. Both of these bacilli can also grow
heterotrophically, if an organic carbon source is present.
The following stoichiometric relationships for hydrogen and sulphur have been
reported:
Hydrogen (Kurt et al. 1987):
2 NO3- + 5H, => N, + 4H,O + 2OH- (4.25)
5 S2032- + 8 NO; + H,O => 4 N, + 10 SO:- + 2 H'(4.26)
Thiosulphate (Claus and Kutzner 1985):
132
Sulphide (Barrenstein et a/. 1986):
S2- + 8 NO& + 8 H+ => 5 SO:- + 4 N, + 4 H20 (4.27)
The C/N relationship decribes the quantity of organic matter, which is needed per
Organic matter of many kinds (as shown in Table 4.5) can be used for the following
unit of nitrate-nitrogen that is converted to nitrogen gas by denitrification.
three purposes in a denitrification plant.
1) Reduction of nitrate or nitrite into nitrogen gas.
2) Sludge production, i.8. biomass production.
3) Respiration with oxygen.
Knowing the values of the three parameters described, it is possible to quantify the
C/N relationship for a denitrification plant.
If the C/N ratio is smaller than is stoichiometrically needed, the denitrification
process will not proceed or be applied with reduced capacity. If there is less nitrate or
nitrite it will be converted into nitrogen gas.
Monteith et a/. (1980) conducted an experiment in 30 industrial waste water
streams. Twenty-seven of the 30 industrial waste streams were evaluated as external
sources of carbon, added to domestic waste water. Fifty per cent of the waste water
tested supplied a sufficient content of carbon for a constant denitrification of domestic
waste water and exhibited denitrification rates equal to or greater than those observed
using methanol. The C/N ratio found in the described experiments with external
sources of carbon were between 0,7 to 2,6 kg FOC/kg NOT-N removed. If methanol
were used at carbon source an average of C/N ratio was found to be 1,17 kg FOCkg
NO,-N removed. FOC is the amount of fully oxidisable carbon. NOT - N er the total
amount of nitrate and nitrite.
133
Table 4.5 Carbon sources other than methanol and internal carbon source in denitrifying experiments.
Compound Reference
Acetic acid
Acetone
Alanine
Bakery sludge
Bouillon/Casein
Brewery waste
Chemical industry waste
Cherry juice
Citrate
Corn starch
Ethanol
Fish meal
Gelatine
Glucose
Ide et a/. (1 972) Kiff (1972) McCarty (1 969)
McCarty (1 969)
Ide et a/. (1972)
Adams et a/. (1970)
Clayfied (1 974) Edholm et a/.(1970) Ericsson et a/.( 1966)
Wilson and Newton (1973)
Englehart and Haltrich (1 968) Haltrich and Jager (1 963, 1970)
Adams et a/.( 1970)
Ide et a/ (1 972)
Adams et a/.( 1970) Ide et a/. (1972)
Bringmann et a/. (1 959) Finsen and Sampson (1959) McCarty (1 969) McCarty et a/. (1969)
Ludzack and Ettinger (
Ludzack and Ettinger (
Balakrishnan (1968)
962)
962)
Balakrishnan and Eckenfelder (1 969) Barth and Ettiger (1967) Christenson et a/. (1956) Clayfied (1 974)
134
Table 4.5 (continued) Ide et a/. (1972) McCarty (1 969) Schroeder and Busch (1967, 1968) Wuhrmann (1960)
Lactate
Margarine
Methane
Milk solids
Molasses
Nitro-cellulose waste
Peptone
Saccharose
Ide et a/. (1972) Toit and Davis (1973)
Bringmann et a/. (1 959)
Christensen (1972) Harremoes and Christensen (1971) Parker et a/. (1975) Pretorius (1 972)
Aguirre and Gloyna (1 967) Hermann (1 962) Parker et a/. (1 975) Pretorius (1972)
Finsen and Sampson (1959)
Mudrack (1971)
Clayfied (1 974) Ide et a/. (1972)
Das et a/. (1966) Finsen and Sampson (1 959) Klotter (1 969) McCarty (1 969)
Sodium citrate Dawson and Murphy (1972)
Sugary syrup Adams et a/. (1970)
Source: Henze Christensen and Harremoes (1977)
135
10
3
1
0.3
0.1
Detrification rate
=0.05 ‘C’
= 0.06 ‘C-’
= 0.08 *C-’
Temp “C I I I I 1 1 c
10 20 30 0
0 Figure 4.2 The denitrification reaction rate as a function of temperature for different
carbon sources. The more easily degradable the carbon source, such as methanol the
greater is the reaction rate. Heavily degradable endogenous carbon has a low reaction
rate, especially at low temperature. (Source: Henze and Harremoes 1978)
136
4.6 Kinetic Expression for the Denitrification Process Environmental factors also have a significant effect on the kinetic rates of
denitrifier growth and nitrate removal. Temperature, pH, carbon concentration and
substrate concentration are considered below. A combined kinetic expression
considering factors that affect denitrification is proposed.
As in the case of nitrification the Monod Kinetic, equation (4.28), has also been
proposed to explain the rate of conversion of nitrate to nitrogen gas, by several
investigators, for example Henze and Harremoes (1972) and Moore and Schroeder
(1 970).
where:
(4.28)
pD = growth rate for the denitrifier, day-'.
p,,,,,,-, = maximum growth rate for the denitrifier, day-'.
Sdenit = concentration of substrate to be denitrified (nitrate nitrogen) in mg/l.
KD = saturation constant mg/l nitrate nitrogen.
Even though the Monod Kinetics is used by several investigators to explain the
denitrification, the suspended denitrification process with methanol as carbon source
is often described in the literature as zero order with respect to nitrate and methanol.
The equation used in following this approach is presented as (4.36).
Denitrification filters appear to conform very well to the laws of biofilm kinetics.
Because of the low saturation constant, K, < 1 mg NO,- -N / liter, the intrinsic reaction
in the biofilm is zero order. This becomes a half-order reaction in thick biofilms owing
to diffusional resistance in the biofilm. Where the substrate concentration in the bulk
liquid is high enough, the biofilm is penetrated fully, and the overall process becomes
zero-order (Harremoes 1982).
137
4.7 Relationship Between Growth and Removal Rate
organism growth rates by the following relationship:
Using the Monod Kinetic approach, denitrification rates can be related to the
(4.29)
where XD = biomass of the nitrifier bacteria and YD the yield coefficient.
4.8 Kinetic Constants in the Denitrification Process The value of the saturation constant KD is very low. Davies (1973), found the
KD value for suspended growth systems to be 0,08 mg/l nitrate nitrogen without solids
recycling. For attached growth systems the value of KD was found to be 0,06 mg/l
nitrate nitrogen at 25 "C. Using these small KD values in equation (4.29), that is Sdenit
is above 1-2 mg nitrate nitrogen, the denitrification will approach a zero order rate.
Several investigators (Christensen and Harremoes 1972; Stensel et a/. 1973;
Murphy and Dawson 1972; More and Schroeder 1970) have all reported zero order
rates for the denitrification process, when the substrate concentration is above 1-2 mg/l
N. Table 4.6 show kinetic constans for the denitrification process. The low value of the
saturation constant, KD, indicates that the denitrification process can be operated at
near maximum unit removal rates and still give an acceptable nitrogen removal.
Table 4.6 Examples of kinetic constants for the denitrification process, using methanol
as carbon source.
10°C
K, mg/l 12.6
Kd d-' 0.05
Y, gVSS/gCOD 0.17
20°C
9.1
0.04
0.18
Source: Stensel and Bernard (1992)
138
4.9 The Influence of Oxygen on the Denitrification Rate Investigators have reported various results for the influence of oxygen on the
biochemistry of the denitrification process. Dissimilatory nitrate reduction (denitrifica-
tion) is inhibited by oxygen, whereas assimilatory nitrate reduction is unaffected.
Payne (1973) explains that oxygen either represses the formation of the
enzyme nitrate reductase or acts just as an electron acceptor, thereby preventing the
reduction of nitrate.
Beneficial effects of oxygen in the denitrification process have been observed
by Ide eta/. (1972). The activity of denitrifying organisms seems to be enhanced after
exposure to oxygen. This effect could be explained by the presence of haem in the
electron transport system, as some organisms need oxygen in order to synthesize
haem (Porra and Lascelles 1965; Tanaiguchi 1961).
The exact control mechanism exerted by oxygen on denitrifying enzyme
synthesis, has not been clearly demonstrated yet, and may very well vary among
species of denitrifiers.
When using attached cultures, it is especially important to distinguish between
oxygen tension within the micro-environment around the bacteria, and oxygen tension
within the macro environment.
It appears that 1-2 mg Odl does not influence denitrification in filters; but in
suspended cultures the oxygen concentration should be below 0,5 mg O#. Table 4.7
show the oxygen concentration in various denitrifying experiments.
4.10 The Influence of Temperature on the Denitrification Rate Denitrification can be performed in the temperature range 5 "C - 35 "C.
Many of the denitrifying species are adaptive to temperature changes.
It is, therefore, important to realize that there is a difference between long-term and
short-term temperature influences on the denitrification process.
The growth rate of the organism and removal rate of nitrate are both affected
by temperature. To show the effect of temperature on growth and denitrification rates,
the results at 20 "C from the literature are summarized in Fig.4.3. Denitrification pro-
ceeds at a reduced rate, at temperatures as low as 5 "C. Above 20 OC, the data indica-
tes that the denitrification rates are constant. Murphy et a/. (1973) showed that
attached growth systems are less affected by low temperatures than are suspended
growth systems. It is important to distinguish between two types of temperature
139
Table 4.7 Oxygen concentration in denitrification experiments, and literature concerned
with the technical importance of oxygen concentration. ~~
Oxygen concentration in experiments Reference
(mg4
< 0,5
0,5 1 10,O
c 0,5
03
0,2 - 5,O 0,o - 0,2
0,O - 1,5
1,5 - 1,8
0,O - 2,5
0,o - 2,o
0,15 - 0,72
c 0,2
0 - 0,3 c 1.5
Ludzack and Ettinger (1 962)
Ruffer (1 964)
Pasveer (1 965)
Schuster (1 970)
Dholakia et a/. (1 970)
Carlson (1 971)
Matsche (1971)
Smith et a/. (1 972)
Jones (1972)
Haltrich (1 972)
Toit and Davies (1 973)
Christensen (1973)
Drews and Greef (1 973)
Parker et a/. (1 975)
Source: Henze Christensen and Harremoes (1 977).
responses during denitrification, as described in Section 3.8 on the influence of
temperature on nitrification.
The first type of response is an immediate (rapid) temperature response, which
is much smaller than the long-term (slow) temperature response. The second type is
the most interesting one; the former is the one often encountered in laboratory ex-
periments. The long-term temperature response is a mixture of an immediate
temperature response and adaptation of the microorganisms (Henze and Harremoes
1 978).
Very little is known about the relationship between long-term and short-term
temperature dependencies.
140
180
160
1 LO
120
100
80
60
L O
2 0
0
% of denitrification I
I
I I
Symbol Reference I
_ _ _ _ _ _ _ _ _ _ Dawson and Murhy (1973) I
at 20 "C I
I I
- EPA (1975)
Temp "C I I 1 I I 1 -
0 5 10 15 20 25 30
Figure 4.3 Temperature dependence of the denitrification process.
Mathematically, the dependence on temperature can be described by the
following exponential expression:
(4.30)
where:
As the temperature in most cases changes slowly, long-term temperature depen-
Y is the temperature coefficient in Table 4.8.
The expression is valid only within the range from 5 O C to about 35 "C.
dencies are the most important for practical purposes.
denitrification processes are listed.
In table 4.8, the long-term temperature constants k,
The following temperature expression is proposed by Hultmann
&At-20) I.r,,t= Pmax,;YT c*'O
and 8 for various
1971):
(4.31)
According to Table 4.8 the literature shows that the temperature dependency
for attached growth is smaller than that for suspended growth.
4.1 1 The Influence of Carbon Concentration on the Denitrification Rate
The effect of carbon concentration on the rate of denitrification has been
explained with a Monod type of expression by, Stensel et a/. (1 973). Using methanol
as the carbon source, the following expression was employed:
where:
M = methanol concentration, mg/l
KM = saturation constant for methanol, mg/l.
(4.32)
The kinetic value of KM is normally very low, normally in the order of 0,l mg/1
methanol.
142
Table 4.8 The influence of temperature on denitrification rate.
Process
Suspended separate culture
Suspended combined culture
Suspended combined culture
A
P 0
Attached separate culture
Attached culture
Y Temp. range OC Reference Carbon Source kt -1 OC
Methanol 0,05 1,12 10-25
Raw sewage 0,06 1,15
Endogenous 0,08 1,20
Methanol 0.02 1,05
0,03 1,07
5-20
15-25
5-20
18-29
Henze and Harremoes
Mulbager (1971) (1 977)
Henze et at. (1 977)
Bernard (1 975)
Harremoes and Rimer (1 977)
Mechala et a/. (1 970)
Table 4.8 (continued)
Process Carbon Source kl -1 Y OC
Temp. range OC Reference
Suspended
Suspended
Suspended
Suspended
- Suspended P P
Suspended
Attached
Suspended
Suspended
0,06 1,15 11-21
0,05 1,12 10-20
0,05 1,12 10-20
0,05 1,12 5-27
0.05 1,12 10-40
0,03 1,07 6-25
0,03 1,07 6-25
0,04 1,lO 5-25
0,07 1,17 15-24
Hunerberg and Sarfert (1 967)
Mulbager (1971)
Stensel (1971)
Davvsm and Murphy (1 972)
Ide et a/. (1972)
Murphyand Sutton (1974)
Murphyand Sutton (1974)
Sutton et a/. (1 975)
Parker et a/. (1 975)
4.12 The Influence of pH on the Denitrification Rate Denitrification only partially offsets the alkalinity loss caused by nitrification,
as the alkalinity gain per mg of nitrogen is only one-half of the loss caused by
nitrification. This is because the alkalinity gain per mg of nitrogen is only one-half the
loss caused by nitrification.
A value for alkalinity production suitable for engineering calculations would be
3,O mg alkalinity as CaCO, produced per mg nitrogen reduced.
In the design of systems where alternating nitrification and denitrification are
used, a sudden high load of ammonia in the waste water can cause a self-destruction
of the system, because of the high H+ concentration developed during nitrification
(Fig. 4.4) The denitrification will not occur because of the decreased pH, as the
denitrifying organisms can not denitrify under a low pH condition.
40 mg/l
30mg/ l
2 0 mg/l
1 Omg/l
pH in the effluent Amount of nitrified N, mg/l
amount in mg/l of nitrification
Time
Figure 4.4 Self-destruction of a system applying alternated nitrification and denitrifica-
tion due to a high H+ production during nitrification.
145
Table 4.9 presents observations from the literature of the effect of pH on
denitrification rates. It would appear that for most systems the denitrification rate is
depressed below pH 6,O and above pH 8,O. Different studies indicate different pH
values as the optima for denitrification, but most studies show the highest rates of
denitrification occur within the range of pH 7.0 to 7.5.
All results are presumably long-term pH dependence studies, but this is
impossible to determine from the information available.
The influence of pH on denitrification is also dependent upon the duration of
the effect. The short-term effect of a pH change is the most interesting, because a pH
change generally does not vary over a long period.
In Section 4.3 it is shown that denitrification produces alkalinity, which will result in an
increase in the pH value. The magnitude of this increase depends upon the buffering
effect of the sewage, because nitrification, on the other hand, produces acidity.
In a combined nitrification-denitrification process, the pH of the two processes
should thus balance each other out, the result being a constant pH. (Barth eta/. 1968; Halling-Sfarensen and Hjuler 1992).
Timmermann and Van Hauten (1983) determined the growth rate p as a
function of pH in batch reactors at 25 "C. The biomass of the bacteria was measured
as a MLVSS- concentration. Figure 4.5 shows that a maximum growth rate was found
at pH 8 5
According to Hartmann and Laubenberger (1968), a deviation of the pH from
the optimum pH reduces the bacterial activity according to the mechanism of non-
competitive inhibition (see Section 3.13).
Table 4.9 pH variation in denitrification experiments, and pH studies.
p H-interval Reference
7,O - 9,0 7,2 - 7,5 6,5 - 7,5 7,9 - 8,l 7,2 - 8,O
Hermann (1 962)
Johnson and Schroepfer (1 964)
Meiring and Stander (1964)
McCarty (1 969)
Barth and Ettinger (1968)
146
(Table 4.9 continued)
6,5 - 7,5 6,O - 9,0 7,5 - 8,l 6,O - 10,O 7,4 - 9,l 6,O - 8,O 5,5 - 8,5 6,O - 10,O 5,O - 8,O 7.7 - 7.8
Moore (1969)
Renner (1 970)
Hamm (1970)
Edholm et a/. (1970)
Stensel (1971)
Mulbager (1 972)
Kiff (1972)
Ide et a/. (1972)
Clayfield (1 974)
Halling-Ssrensen and
Hjuler (1 992)
Timmermann and Van Hauten (1 983) also showed the methanothitrate-N
ratio as a function of pH. At optimum pH (=8,3 proposed by Timmermann et a1.1983)
the methanol/ nitrate-N ratio was found to be 2,52 g CH,OH / NO,’ - N, (Fig 4.6).
4.1 3 Combined Kinetic Expression for the Denitrification Process As for the nitrification process, a combined expression for the denitrifer
growth pLo and nitrate removal, taking some of the environmental factors into account,
can be formulated.
Removal rates can be related to growth rates through equation (4.34).
(4.33)
(4.34)
Timmermanns and Van Hauten (1983) proposed an equation similar to
(4.34), that also takes the influence of pH and temperature into account:
147
0.20
0.1 5
0.1 0
005
Time 0
0 1 2 3 . 4 5 6 7 8
Figure 4.5 Determination of the growth rate p at different pH values in a batch reactor
at 25 "C. The biomass, X , is measured as a MLVSS concentration. After Timmermann
and Van Hauten (1 983).
(4.35)
Assuming zero order kinetics, the equation proposed by Timmermann eta/.
(1 983) can be rewritten as:
(4.36)
148
ratio (CH,OH/NO, - N)
5 -
L -
3 -
2
,,
PH I I I I -
Figure 4.6 Methanol / nitrate-N ratio as a function of pH for the denitrification process.
After Timmermann and Van Hauten (1983).
4.1 4 Bacterial Population Dynamics for the Denitrification Bacteria The population dynamics of the denitrifying bacteria resemble the dynamics
proposed for the nitrification bacteria, but the growth rate for the denitrifying bacteria
is larger than for the nitrifying bacteria. It is, therefore not difficult for the denitrifying
bacteria to compete with oxidizing bacteria in a combined organic and nitrogen
removal, as is the case for the nitrifying bacteria.
The safety factor SF concept used in Section 3.12 can also be applied to
denitrification. It can be related to nitrate removal rates through the following equation:
4 d SF=- 41n
(4.37)
149
where:
$d = solids retention time for the denitrification process
$,,, = minimum solids retention time for the denitrification process
In the case of denitrification, the safety factor can be related to nitrate
removal rates, using the following two equations:
1 -=pD-Kd Qd
(4.38)
(4.39)
4.1 5 Influence of Toxic Substances on the Denitrification Process The inhibition equation of the denitrification process resembles the equation
proposed for the nitrification process in Section 3.1 3.
As for nitrification, the following overall expression takes both toxic
substances and oxygen inhibition into account:
(4.40)
where f[l] is a term taking the inhibition of toxic substances into account, and f[O,] the
oxygen inhibition, during the denitrification.
150
The major influence of toxic substances on denitrification is the short-term
influence on the growth rate. It is of great importance that the denitrifying population
is capable of dealing with different toxics, because then a long-term influence of the
same toxic will not be as persistent as the short-term influence, since a bacteria
population is very adaptive to all every environmental changes.
Many of the results referred to in the literature show how a short-term response can
influence a population of bacteria, and may, therefore, often appear to be much more
dramatic than a long-term influence, where the bacteria would have had time to adapt.
4.1 6 Conclusion Chapters 3 and 4 summarize the results from many scientists concerning
different factors affecting nitrification and denitrification.
It is often difficult in practice to evaluate the relevance of the different results,
and thus it is also difficult to select the appropriate results for the planning of a
particular biological nitrogen removal unit. The authors therefore recommend
considering as many as possible of the different results mentioned, for the case study
at the planning stage of a particular plant.
For example, Sections 3.8 and 4.12 give an overview of the influence of
temperature on nitrification and denitrification processes, and refer to the results of
numerous investigators. The various equations proposed should be tried in turn to see
how they fit the case study, in order to avoid dimensioning errors in the completed unit.
This approach, in effect, brings a safety factor into the plant design.
151
This Page Intentionally Left Blank
5. ATTACHED GROWTH REACTORS
5.1 Introduction In attached growth systems the waste water is in contact with a microbial film,
attached to the surface of a solid materiavmedium. The surface area for growth of the
biofilm is increased by the use of a porous medium in the reactor. The biological
reactions take place in the biofilm, while suspended bacteria are washed out of the
systems.
When randomly packed reactors, are used and the waste water flows by
gravity as a free surface stream, the reactor is called a trickling filter. The use of
rotating discs, covered with biofilm, partially submerged in waste water is called a
rotating biological contactor (RBC) process, where the biofilm development is controlled
by the rate of rotation.
Other attached-culture systems are the submerged filters with up-flow or down flow application, i.e. Up-flow Fixed Bed Reactors (UFBR) and the Fluidized Beds. Thay
may both have applications under certain conditions such as high-nutrient-containing waste water. Figure 5.1 shows some of the nitrifying attzched growth units in use.
In the trickling filter, the medium is stationary and the waste water is passed
over the biofilm in intermittent doses. In the RBC, the medium moves the biofilm
alternately through water and air. Using the UFBR, the waste water is pumped up-flow
through a fixed medium. The fluidized bed consists of spherical particles coated with
a biofilm fluidized by up-flowing water. A segregation generally occurs: the apparent
density of the particles decreases, as the thickness of the biofilm increases.
Continuous control of the biofilm is not possible with a stationary support
medium. The filters therefore have to be backwashed in order to prevent clogging.
Experience has shown that many kinds of support material can be used, for
example stones, gravel, sand, plastic, asbestos plates, wood, zeolites, and activated
carbon particles.
In addition to the biological reactor, an attached growth system usually includes
both primary and secondary clarification. Recirculation of sludge is normally not
necessary in biofilm reactors, because the amount of biomass is huge compared with
activated sludge systems.
The major work in an attached growth system is to establish a unit con-
figuration, where oxygen can uniformly be supplied during the nitrifying process and
153
where water at the same time can pass through the support media without any
limitations.
Many experiments have been carried out, assuming that the nitrifying rate per
unit surface of biofilm is the same from one reactor to the other, independent of
influent characteristics. Serious errors have, therefore, been made, for example with
respect to scaling up. Results from pilot scale experiments with submerged filters
cannot be scaled up on the assumption that the rate of substrate removal per unit m3
is the same at full scale.
The processes for which the biofilm reactors have been used or proposed for
use in waste water treatment are oxidation of organic matter, nitrification, denitrification
or combinations of these.
5.2 The Biofilm Nitrifier, denitrifier, oxidizer or a combination of these types of bacteria can
attach themselves to different types of medium and grow into dense films of a viscous,
gelatinous matrix called the biofilm. Waste water passes over this film in thin sheets,
with dissolved organics, NH,' or NO, passing into the biofilm due to diffusion
gradients within the film. Suspended particles and colloids cannot penetrate the surface
of the biofilm, but will be decomposed on the surface of the biofilm into soluble
products. Oxygen from the waste water and from air in the void spaces of the medium,
provides oxygen for the aerobic reactions at the surface of the biofilm. Figure 5.3 show
a diagrammatic representation of a biofilm with involved processes.
Waste products from the metabolic processes diffuse outward and are carried
away by the water or air. Growth of the biofilm is restricted to the outward direction
from the solid surface. As the film grows thicker (see Fig. 5.2), concentration gradients
of both oxygen and nutrient develop. Eventually, when the biofilm is of an appropriate
size, both anaerobic and endogenous metabolism occur in the interface of the biofilm.
In a well developed biofilm, the attachment mechanism to the solid medium is
weakened, and the shearing action of the waste water flowing across the film pulls it
from its attachment and washes it away. This process is called sloughing, and is a
function of both the hydraulic and the organic loading rates. But the biofilm is quickly
re-established, and, therefore, sloughing is a beneficial mechanism for development
of new biofilm.
154
Trickling Filter Rotating Disk unit
+- - +
Submerged Filters Fluidized Filter
down - flow up - flow
Figure 5.1 Biofilm reactors used in waste water treatment. a) Trickling filter,
b) Rotating Biological Conductor (RBC), c) Submerged filters with down-flow or up-flow
application, and d) Fluidized filters.
5.3 The Development of a Bacterial Biofilm
bed as follows (Elmaleh and Grasmick 1985):
The successive steps of the development of an aerobic biofilm can be descri-
step 1 - The biofilm is composed of a few aerobic bacteria included in a gelatinous
matrix, i.e the density is low.
155
step 2 - Aerobic micro-organisms grow rapidly, and the density is an increasing
function of the thickness.
step 3 - As oxygen depletion begins to occur in the biofilm, an anaerobic zone
appears near the solid material.
step 4 - Anaerobic and facultative bacteria grow near the solid material as aerobes
decay, causing decreasing density.
step 5 - An equilibrium between the anaerobes and aerobes is reached, which means
that the density is stabilized.
step 6 - Now the equilibrium is maintained, until the substrate concentration is
exhausted in the deeper zone, the anaerobe bacteria will begin to decay, and parts of
the biofilm will finally slough away.
step 7 - The newly developed space will be used by new aerobic bacteria which will
start all over again and build new biofilm.
The rate of nutrient removal in attached-growth systems depends on the flow
rate of the waste water, the organic loading rate, rates of diffusivity of nutrients into the
biofilm, and temperature. The depth of penetration of both oxygen and nutrients is in-
creased at higher loading rates. Oxygen diffusivity is usually the limiting factor. Aerobic
zones of the biofilm are usually limited to a depth of 0,l to 0,2 mm, the remaining
thickness of the biofilm being anaerobic (see Fig. 5.4). Depending on hydrodynamic
conditions, Atkinson and Fowler (1974), found values between 0,07 and 4 mm in
thickness. When the biofilm is mechanically or hydraulically controlled, its thickness
does not exceed 0,2 mm, which is the maximum depth for having a full aerobic biofilm
(Grasmick 1982).
Arvin and Harremoes (1990) reported that the thickness of the biofilm is
controlled by the following factors:
1. Growth of active biomass as a result of influx of the substrate.
2. Decay of active biomass.
156
a b C d e f 9
0 aerobic viable microorganisms
fi anaerobic microorganisms
o g a s e o u s metabol i tes dead microorganisms 7x4 support
Figure 5.2 The different steps in developing a biofilm, shown as transients of a
biofilm.
3. Accumulation of inert organic material from the decay of active bacteria.
4. Accumulation of polymers from the metabolism of the substrate.
5. Deposition/flocculation of suspended particles from the bulk liquid.
6. Erosion of small particles from the surface of the biofilm.
7. Sloughing of large fractions of the biofilm.
At present our ability to predict the thickness of a biofilm is relatively low.
It is also difficult to define and measure the thickness of a biofilm. Experimental
measurements can be made directly, or can involve a procedure called congelation of
the whole reacting medium. Accuracy in the values of a thickness less than 100 pm
is 20%, but on values of about 2 mm, the discrepancy can reach 300 % (Grasmick
1 982).
The number of variables affecting the growth of biomass, and subsequently the
rate of substrate utilization, makes it difficult to describe the systems mathematically.
Masuda eta/. (1 987) reported that oxidizing, nitrifying and denitrifying bacteria
can exist almost uniformly in the entire biofilm. Oxidation of organics, nitrification and
denitrification occur in the same biofilm. Probably the denitrifying bacteria exist in the
most anoxic areas, in the deeper layer of the biofilm.
157
Biofilm Liquid Bulk water Diffusion film I
02 conc.
E 0 -4
cu cu
Bulk Transport Reaction
n Products
Matrix
~
7
Figure 5.3 Diagrammatic presentation of a biofilm with involved processes.
f (3;;n Concentration
-100 0 100
Figure 5.4 Oxygen concentration profile showing that at thickness over 100 pm the
biofilm is anaerobic.
158
5.4 Modelling the Transport and Reactions within a Biofilm In spite of the heterogeneity of the biofilm, it is assumed in most of the models
that the substrate is transported by molecular diffusion and, therefore, that an effective
diffusivity is a characteristic constant of the system. (Atkinson and Fowler 1974;
Harremoes and Riemer 1975; Harremoes, 1975, 1976, 1978,; Arvin and Harremoes
1990; La Motta 1976 ; Williamson and McCarty 1976 ; Grasmick et a/., 1979, 1981;
Rittman and McCarty 1981).
The rate of reaction in a biofilm is based on the concept of the limiting
substrate. If the waste water is aerobic, the limiting substrate will consist of oxygen,
organic carbon and/or ammonia.
The intrinsic reaction rate of a limiting substrate can be described, depending
on the authors, as a Monod-type, first, or zero order equation.
In waste water treatment, it has been shown that the best approximation is the
zero order ( La Motta 1976; Riemer and Harremoes 1978; Grasmick 1982). Depending
on the penetration into the biofilm, the apparent reaction rate will be zero order kinetics
for full penetration, and half-order kinetics for partial penetration (Harremoes 1978).
Table 5.1 presents values for the biofilm kinetics.
Arvin and Harremoes (1990) proposed that the basic feature in the biofilm
model is the kinetics of the processes performed by the active bacteria in the film.
This approach can be used for describing processes other than the nitrification and
denitrification such as aerobic mineralization, sulphate reduction, fermentation, or
methanogenesis.
Bacteria / activity:
described by transforming the monod equation (3.1 1) to:
If pmax and Y can be considered universal, then the bacterial activity can be
where:
k, = the maximum soluble substrate (zero order) utilization rate.
The kinetic characteristic of the biofilm reactor is:
159
1. The diffusion resistance to the movement of the substrate into the biofilm.
2. The products developed in the biofilm.
a substrate can penetrate the biofilm fully or partly (Harremoes 1978).
There is a difference in the performance of a reactor depending upon whether
The diffusion resistance:
Arvin and Harremoes (1990) explain that:
The diffusion resistance affects both the removal rate and the order of reaction.
- A first order reaction in the interior of the biofilm is converted into a first order bulk
reaction at a reduced rate.
- A zero order reaction in the interior of the biofilm remains a zero order bulk reaction,
if the biofilm is fully penetrated, but is converted into a half-order reaction, if the film
is only partly penetrated.
Assuming that the reaction rate for the nitrification and denitrification is zero
order, the following kinetic equations can be used:
The biofilm where the substrate penetrates fully:
dS/dt = k, L (zero order) (5.2)
The biofilm where the substrate penetrates partly:
or:
where:
dS/dt = (2 D k, s)” (half order)
dS/dt = the reaction rate per unit surface of the biofilm.
k, = the intrinsic reaction rate per unit volume of the biofilm.
160
L =the thickness of the biofilm.
D = the diffusion coefficient.
S = the substrate concentration, which can be forms of nitrogen
carbon or oxygen.
The transition from zero order to half order kinetic is governed by the relative
penetration of the substrate and the rate order can be determined using the following
equation.
Where 0 > 1 than the biofilm is fully penetrated by the substrate.
and where 0 < 1 it is partly penetrated.
a half order kinetic.
The biofilm in a trickling filter treating domestic waste water will usually follow
The appearance of non-diffusible matter in the biofilm reactor.
The theories for substrate removal in the biofilm suffer from the fact that only
very little is known about how the biofilm affects non-diffusible matter (Levine 1985;
and 0degaard 1987).
The two main questions concerning non-diffusible matter are:
- How can particulate matter be attached to the biofim surface?
- What is the mechanism for the extracellular degradation of the attached particulate
matter?
The removal of the particulate matter depends on the following aspects (Arvin
and Harremoes 1990):
1. The size and the chemical charge of the particulate.
2. The size, shape and chemical composition of the support media.
3. The surface of the biofilm.
4. The waste water flow through the biofilter.
161
Table 5.1 Values for the biofilm kinetic using zero order or an apparent half-order kinetic coefficient.
Pollution Limiting Conditions Temp. Intrinsic zero Apparent half- Apparent half- Reference substrate OC order rate per order coefficient order coefficient
unit biofilm per unit biofilm per unit reactor
(kg/m3 S lC3) (kgHm%' 1 C5 (kgHmas-' 1 @) volume volume
Milk 02 fixed bed 20 0,12 Grasmick et a/. (1 980)
Beef 02 fixed bed 20 extract
A
a, Methanol methanol rotating 22 ru reactor
0.32 Grasmick eta/. (1980)
0,16 - 0,19 0,38 - 158 Jansen and Kristensen (1 980)
Milk TOC rotating 20 0,27 - 0.59 0,083 - 0,18 Grasmick (1982)
From: Grasmick (1985)
The transport of non-diffusible matter into the biofilm is very slow, compared
to the transport of diffusible matter.
Degradation of particulate matter outside the biofilm is conducted by
extracellular enzymes, released into the waste water by the biofilm, or enzymes
working on the membrane of the biofilm. This conversion of particulate matter into a
soluble product, that is able to diffuse into the biofilm, may be by a special mechanism,
which facilitates the penetration of the biofilm.
The liquid film diffusion. Before any reaction takes place inside the biofilm, the substrate needs to
be transferred from the bulk liquid to the solid phase. The existence of a mass transfer
resistance in liquid-biofilm har been demonstrated. (La Motta 1976; Grasmick 1982)
The flux, J, of substrate into the biofilm follows Fick's first law.
where: S and Ss are the bulk and interfacial concentrations; LB is the thickness of the
boundary diffusion layer. L, can be determined using a method described by Bouwer
and McCarty (1985).
In practice, oxygen is always the rate limiting factor rather than the ammonia
concentration, because the critical ratio between the two concentrations for performing
nitrification is of the order of 0.3-0.4 mg NH, per mg 0, (Gonec 1982). If the
concentration of 0, is, for example, 4 mg/l, then the concentration of NH, has to be
smaller than 1.3 mg/l to be limiting. Table 5.2 presents values for the effective
diffusivity in pure water and in biofilms.
Bacterial population dynamics in the biofilm. If a biofilm has to oxidize carbon matter and nitrify simultaneously, the two
electron donors will compete for the same electron acceptor, oxygen. Both processes
will take place in the aerobic zone of the biofilm. The relative use of the limited electron
163
acceptor resource is determined by the population dynamic of the heterotroph and
nitrifiers in the biofilm. In the aerobic zone, both types of bacteria will grow, but at
different rates, determined by the available substrate and the growth rate of the
different species of bacteria.
At a particular ratio of organic matter to ammonia, the nitrifiers will be
outgrown by the heterotroph, and no nitrification will occur. This effect is similar to the
wash-out of the nitrifiers in an activated sludge plant at too low a sludge age.
If the biofilm is not fully penetrated by oxygen (Fig. 5.4), it will be divided into an
aerobic part adjacent to the bulk liquid and an underlying anaerobic part. Applying
clinoptilolite as a support medium in an upflow fixed bed reactor (UFBR) (See section
5.8.1) it was possible to obtain simultaneous nitrification and denitrification (Halling-
Serrensen and Hjuler 1992).
The biofilm composition is always a "mirror" of the composition of the waste
water applied to the treatment system. Different zones may be developed as a function
of the loading of substrate to the biofilm. According to Kinner (1983) (see Table 5.3)
the most varied biofilm induced by a heavily loaded waste water can have four different
layers, as follows:
1. An outer layer with heterotrophic oxidation of organic carbons, nitrification and
denitrification and sulfide oxidation.
2. A microaerophilic layer with denitrification and fermentation.
3. An anaerobic layer with sulphate respiration and fermentation activity.
4. An anaerobic layer adjacent to the support material with methanogenesis and
fermentation.
If the waste water becomes less heavily loaded, or possibly acquires a different
composition, the biofilm will be built up of layers 1 and 2 only, or consists of layer 1
only.
At steady-state the fraction of organisms fi, in one of the layers is given,
indicating a balance between growth and decay, as:
ra* Yi = bi X, * fi I (5.7)
164
where:
r, = the removal rate per unit surface area of substrate utilized by organism
Yi = the yield constant for organism group i.
bi = the decay constant for the organism group i.
X, = biomass of the whole biofilm.
fi = the fraction of organism group i.
I = the length of the zone with organism group i.
group i.
The product X,*fi*l = ra * Yi/b is derived from equation (5.7) and reflects the
steady-state active biomass of organism group i per unit area of biofilm.
The pH effect in the biofilm.
Nitrification is an acidity-producing process, while denitrification is an
alkalinity producing process as outlined in sections (3.4) and (4.4).
In bulk waters of low alkalinity the result can, therefore, be a significant drop in pH in
the biofilm conducting nitrification. This can lead to an inhibition of the nitrification
because of too low a pH.
In the denitrifying biofilm, the pH can be increased in the rear of the film to
an extent where precipitation of phosphate can occur (Arvin and Christensen 1979).
There is no mention in the literature of the pH in a biofilm conducting simultaneous
nitrification and denitrification.
5.5 A Massbalance Equation for a Biofilm Plant
outlined as follows:
A mass-balance equation for a biofilm plant without recirculation can be
165
Table 5.2. Values for the effective diffusivity in pure water and in biofilms.
Pollution Substrate Conditions Temp. Effective Reference
Pure water
Milk
Beef extract
Wastewater
Glucose a a
Nitrogen compounds
Methanol
GI u cose
Milk
0 2
0 2
0 2
0 2
0 2
0 2
methanol
glucose
TOC
fixed bed
fixed bed
rotating cylinder
pure culture Nitrobacter Nitrobacter + Nitrosomonas
rotating reactor rotating reactor rotating reactor
20
20
20
20
20
22
22
25
15
3 - 9
6 - 1 0
12 - 17
4 - 20
25
20 - 50
1 - 5
6 - 50
Grasmick et a/. (1980)
Grasmick eta/. (1980)
Tomlinson and Snaddon (1 966)
Matson and Charaklis (1 976)
Williamson and McCarty (1976)
Jansen and Kristensen (1 980)
La Motta (1976)
Grasmick (1982)
From: Grasmick (1985).
Predominant Metabolic Reactions Limiting Bacteria process substrate
7- 1 1 I I I
Reactants + products
Aerobic + nitrifying Aerobic respiration CH,O + 0, helerotrophs
co, + H,O
Heterotrophic CH20N.NH4+ + 0 nitrification NO; + CO, + HZ8 CH,O
H,S + 0, + S + H20 Beggiatio~s like sulphur storage filaments
.----- ------ - - - - Nitrate reduction 5 CHZO + 4 NO; + 4 H' -
(CH20jn + H,O - (CH2O),,., + CO, + H, + 2 H'
(CH,O), + H20 - (CH,O)"., + CO, + H, + 2 H+
i Dennritiers Dennrification 2 N2 + 5 COP + 7 H,O 0,
------- --- 2 CH,O + so: s2. + 2 CO, + 2 H,O c'o
Y co2 ;
1 Metanogens Meihanogenesis 2 CH20 - CH, + CO, I , Facultative anaerobs Fermentation (CH-OI.. + H,O - ICH,OI... + CO, + H., + 2 H' c%o
' Aerobic + nltrifying Aerobic r9SPir9tiOn CH,O + 0, - CO, + H20
, t Oennriliers
heterotrophs Heterotrophic CH20N.NH4* + 0, - nitrification
Nitrate reducers Nitrate reduction Denndlication
Facultative anaerobe Fermentation
NO,' + CO, + H,O
5 CH20 + 4 NO; + 4 H+ 2 N, + 5 CO, + 7 H,O (CH,O), + H,O - (CH20)n., + CO, + H, + 2 H+
----- -------___
RBC Plastic media
Aerobic + ni t r l fy i~ AeroMc respiration CH,O + O2 - CO, + H,O
I 6 NH,' + 2 0, -. NO, + 2 Hi + H,O NO-' 0. cb, NO. ?+
I i Aulotropha niirifiers Nitrilication ~.
RBC Plastic media
167
where:
Qinf = the influent flow in I/s.
Cinf = the influent substrate concentration in mg/l.
rx,s = the process reaction rate in kg substrate per kg biomass /m3
day.
V
Q,, = the effluent flow in I/s. C,, = the effluent substrate concentration in mg/l.
= the volume of the reactor in m3.
Because it is difficult to quantify or estimate the biomass concentration X, in
a biofilm plant, it has been suggested in the literature that the term rx,s V * X, be
changed to a term taking into account the volume or the area of growing bacteria.
In the literature the following removal terms have often been used.
where
rv,s = the amount of substrate removal per m3 per day, expressed as a
volumetric reaction rate.
ra,s = the amount of substrate removal per m2 per day (a term often
used for RBC plants), expressed as a surface reaction rate.
Using one of the above terms avoids the necessity of knowing the con-
centration of the biomass X,, but can simply relate the reaction rate to the present
biomass X,, under steady-state conditions, of a specific area or volume. Table 5.4
show the different units which indicate the substrate (nitrogen) removal rate.
Depending upon whether the substrate can fully penetrate the biofilm or not,
the kinetic will follow zero order or % order kinetics. Equation (5.2) or (5.3) must then
be introduced as the kinetic rate.
Applying zero order kinetics the mass-balance for the biofilm plant will be:
168
and for % order kinetics:
Qinf Cinf - (2 D k, * s)' A = Qeff * ceff (5.11)
where:
kB = - cr, *XB (5.12) ymx
The use of this equation requires also the knowledge of the biomass
concentration X,, therefore k, is usually an experimentally found Constant observed
in a special set of conditions , which also makes possible the estimation of biomass.
Table 5.4 The different units which indicate the substrate (nitrogen) removal rate.
nitrification rate term unit
as biomass rx,s kg N/kg biomass per m3*d
as volumetric rate rv,s kg N/ m3 d.
as surface rate 'a,s g N/m2 d.
If recirculation of the waste water is used in a biofilm plant, the following
equation of biomass-balance can be used:
Qinf 'inf + 'x,x " 'B = Qeff * 'eff + 'sedimentation 'sedimentation (5*1 3,
where:
rx,x = the rate of biomass activity per unit of biomass.
The recirculation of water is used to ensure a constant water passage through
the support material.
169
5.6 The Nitrifying Trickling Filters (NTF) The trickling filter was introduced at the end of the last century, and is one of
the first methods used for the removal of nitrogen from waste water. At first the trickling
filters were only a primitive form of land treatment, where sewage was spread at
intervals over sandy ground, allowing the sand to dry between each spreading.
Later, the sand was replaced by stones, but the operational procedures
remained the same, with down flow application of the waste water. Trickling filters were
first introduced in Great Britain. They were circular and supplied with a rotating
distributor at the top of the filter, and measures were taken to secure accurate
aeration.
An underdrain system is designed to carry away the treated waste water and
the sloughed biomass. Several operational modes are available for trickling filters.
Standard-rate filters have low hydraulic loading and do not include provision for
recycling. High-rate filters maintain high hydraulic loading by recirculating portions of the effluent. Filters placed in series increase the effective depth, thus increasing the
efficiency. A great number of possibilities exists for different flow regimes.
Figure 5.1 shows the basic design of a trickling filter. Modern trickling filters
(sometimes called bio-towers) are packed with different types of plastic media , which
allow the filters to be more efficient and also able to treat highly polluted industrial
waste water. Plastic media consists of either vertical-flow or cross-flow substances.
The advantages of using plastic media are a high specific surface in addition
to high void fraction and low weight, reducing the construction costs and high stability
to shock-loads; this again, allows the construction and application of smaller trickling
filters.
Process improvements of trickling filters, using bioflocculation components as
a post-treatment following biological treatments, have produced higher quality effluents
than previously. This improvement makes the trickling filter compatible with the
activated sludge systems, and can produce a high quality effluent, comparable with
that produced using the activated sludge process.
Trickling filters used for nitrification, are employed either to nitrify secondary
effluent, or to combine organic removal with nitrification of the primary effluent. Depen-
ding upon the composition of the influent waste water, a different bacterial population
will be developed.
Reduced removal efficiency of nitrogen can occur in a trickling filter for a
170
variety of reasons. Most important among these is the removal of biofilm by predators
(worms and fly larvae), and incomplete wetting of the media. Depressed nitrification
rates can, however, also result from competition between nitrifiers and heterotrophs for
dissolved oxygen.
Parker and Richards (1986) determined that a soluble BOD, concentration
above 20 mg/l was sufficient to prevent nitrification in a Nitrifiying Trickling Filter (NTF).
A typical removal rate for a conservatively loaded NTF is between 0.20 and
0.39 g N / m2 d. With the development of the BiofiIm-Controlled-Nitrifying-Trickling-
Filter (BCNTF) in the late 1980's the reaction rate in these filters has been increased
significantly.
As a comparison, the normal removal of organics with a trickling filter is in the
order of 2-3 kg BOD /m3 * d. But extreme removal rates of up to 10-20 kg BOD/m3
d have been reported (Audoin et al. 1971).
5.6.1 The Performance of Trickling Filters The performance of trickling filters is affected by many factors, such as the
hydraulic, organic and nitrogen loadings, characteristics of the influent waste water, its
temperature, distribution, distribution frequency, and composition. Other factors are
concentration of bacteria and macroorganisms, oxygen supply, the volume and
geometric shape of the filter medium, and depth of filter.
The trickling filter medium.
The requirements for the trickling filter medium are to present a large surface
for the bacterial population to grow on, and provide a large enough empty space to
secure aeration.
Only by applying plastic-medium is it possible to satisfy these two requirements
simultaneously, because of its low weight. The geometric shape of the packing is also
of importance, not only in relation to the maximum available surface for biological
growth, but for its influence on the hydrodynamics of the filter; this again, influences
the retention time in the filter. Table 5.5 show propoerties of the trickling filter media.
The influence of variation of nitrogen and organic load on a trickling filter.
Significant load variations of ammonium-nitrogen are normal during the course
of the day. The nitrifying trickling filter must, therefore, be designed to be able to treat
171
peak loads, otherwise an ammonium breakthrough must be expected in the effluent.
Trickling filters are specially sensitive to ammonium-nitrogen in the effluent, because
of the very short hydraulic residence time coupled with the down flow system.
As the lower parts of the trickling filter obtain ammonium for only a few hours
each day, it may require a long time to establish a fully developed biofilm at the bottom
of the reactor. During periods of warm weather only the upper part of the trickling filter
may be active, due to the higher reaction rate per unit area. Sudden temperature drops
may, therefore, cause an ammonium breakthrough since the biofilm may not be
developed at the bottom of the reactor. To avoid this, Boller and Gujer (1986) sugge-
sted that two trickling filters should operate in series. Their sequence should be
reversed once every week to obtain a homogeneous distribution of biomass throughout
the reactors.
Easily degradable organics will always be preferred by the bacteria, and the
capacity of the trickling filter, treating waste water with such a composition will,
therefore, be high.
Several investigators have shown that the removal per volume filter at
moderate loads can be described as a linear function of the load per volume.
Thus, the performance of the trickling filter evaluated for removal of organics
and nitrogen would depend on the amount of total organic load applied to the filter,
rather than its concentration or the flow rate.
The oxygen transfer in a nitrifying biofilm in an NTR plant.
By calculating the total oxygen transfer to a nitrifying biofilm, the maximum
removal of nitrogen can be determined for different types of plastic media as shown
in Fig. 5.5.
Cross-flow media are predicted to produce a higher nitrification rate than
vertical flow media of identical surface area, because of fluid disruption at mixing points
in the cross-flow media (Parker et al. 1989).
172
Table 5.5 Properties of trickling filter media.
Medium Nominal size Masdunit volume Specific surface Void space mm kg/m3 area d / m 3 per cent
River rock
Small
Large
25-65
1 00- 1 20
1250- 1 450
800-1 000
Blast-furnace slag
Small 50-80 900-1 200
Large 75- 125 800- 1000 A
-l 0
Plastic
Conventional 600 x 600 x 1200 30- 1 00
High-specific surface 600 x 600 x 1200 30- 1 00
Redwood 1200 x 1200 x 500 150-1 75
55-70
40-50
40-50
00-60
55-70 40-50
45-60 50-60
80-100 94-97
100-200 94-97
40-50 70-80
From: Metcalf and Eddy (1991)
‘(I. 1
A. Vertical Media 6. Cross Flow Media
Figure 5.5 Downward flow pattern in vertical and cross-flow media. After Parker and
Merrill (1 984).
The hydraulic load.
The hydraulic load is a factor affecting the retention time, which is considered
one of the most important factors influencing the performance of the trickling filter.
A high loading rate results in rapid growth of the biomass, and excessive
growth may result in the plugging of pores and subsequent flooding of portions of the
medium. Increasing the hydraulic loading rate, increases sloughing and helps to keep
the bed open.
One of the limitations is the incomplete wetting of the packing at low loads and
percolation at high loads. But other factors can also enhance or slow down the
performance of the trickling. If diffusion in the liquid film somewhere in the filter controls
the reaction rate, an increased flow rate will increase the reaction.
For plastic-packed trickling filters with a high specific surface, this effect will
most likely influence the reaction rate at even normal loadings (for NTF 0.20-0.40 g N
/m2 day). In the literature the influence of the hydraulic load on the wetted area of the
filter has been suggested to be an important factor in this performance. The wetted
area might vary with depth, because of an uneven distribution of biomass in the filter.
174
The relation of the depth and retention time for the trickling filter.
The retention time is considered to be directly proportional to depth, and
therefore using the retention time automatically includes depth. Depth is normally used
as a measure of total available biomass, while retention time is a measure of time of
contact between organisms and substrate. The following equation is generally accepted
in calculating the retention time in a trickling filter:
H t = m s t a n t * - 0"
where:
(5.14)
Q = the flow in Ihour
H = Hydraulic retention time
n = no of recyclings
t = time
As the removal of organic pollutants from liquids takes place mainly through
adsorption and absorption, the time of contact between organisms and substrate is
considerably longer than the retention time of the liquid.
The removal per unit of biofilm surface sometimes increases at higher flow,
which is contrary to the theory, used in most models, that only the flow influences the
retention time. The same is observed when applying the SND mechanism, as shown
in Section 5.8.1.
The influence of temperature on the performance of the trickling filter.
Very little information is available on the relationship between nitrification rate
and temperature (see Fig. 5.6), because most studies of combined carbon oxidation
and nitrification trickling filters have been carried out above 16 "C. Data for lower temperatures can hardly be obtained because of lack of in-
vestigation, and nitrification data obtained at higher temperatures cannot be easily
converted to represent performance at ten degrees lower for example, because
changes in the nitrification rate will reflect changes in the relative growth rates of two
175
different types of organisms in the treatment plant. No information is available on the
influence of temperature on the competition between nitrifiers and heterotrophs.
The interfaces of biomass, water and air also makes the trickling filters
extremely sensitive to variations in temperature. Effluent quality is thus likely to show
drastic seasonal changes, due primarily to changes in ambient air temperatures. The
temperature of the waste water and air also determine the direction of air flow through
the medium. Cool water absorbs heat from the air, and the cooled air sinks towards
the bottom of the filter in a co-current fashion with the water.
Conversely, warm water heats the air, causing it to rise through the underdrain
and up through the medium. Extreme cold may result in icing and destruction of the
biofilm.
The effect of recirculation in a trickling filter.
Recirculation is done to ensure that a constant volume of waste water enters
the plant, to dilute a strong or toxic waste, to increase the surface load, or to prolong
the retention time, so that each "substrate particle" passes through the filter more than
once.
In several investigations, recirculation has been proved to enhance the
efficiency of the plant. The most important factor in determining the extent of
recirculation is to identify which factor controls the reaction rate, because the effect of
recirculation might change the control of the reaction rate from one factor to the other,
for example a process controlled by liquid diffusion might become controlled by biofilm
diffusion or metabolic activity.
The influence of substrate composition on performance of the trickling filter.
With a complex substrate such as domestic sewage, there will most likely exist
different organic and nitrogen compounds which can be difficult to break down. Such
differences in composition of the waste water are very important to take into account
in the calculation of possible efficiency.
176
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
Figure 5.6 The
After Gujer and
Nitrification Rate,
g N / m 2 * d
.L ' //
,+'+
Central Vailey
0 - 0
+ .' /- /'+ +
a Lima
TemD "C I I 1 I *
5 10 15 20 25
effect of temperature on nitrification in a trickling filter.
Boller (1 986).
5.6.2 Equations for Modelling the Nitrifying Trickling Filter (NTF)
by Erkenfelder (1961), and is as follows:
The most commonly used formula for designing a trickling filters was proposed
(5.15)
177
where:
SNH4e = effluent substrate concentration, mg/l
SNH4i = influent substrate concentration, mg/l
D = depth of the medium in meter, m.
Q = hydraulic loading rate m3/m2 min.
k = treatability constant relating to the waste water and the medium
characteristics, min-' .
n = Coefficient relating to the medium characteristics.
The values of the treatability constant k range from 0.01 to 0.1. The average
value for municipal waste on plastic media is of the order of 0.06 at 20 "C
(Germain 1966)
factor kT as follows:
Correction for other temperatures can be made by adjusting the treatability
(5.16)
The treatability factors kT should be determined for each situation from a pilot-
plant analysis of the waste water, and for the selected medium. The coefficient n for
plastic media is 0.5 following Benefield and Randall (1980).
Including recirculation of the waste water into the equation, equation 5.15 can
be modified to:
(5.17)
178
Table 5.6 Typical design criteria for the Trickling filter.
ltem Low-rate filter Intermediate rate filter High rate filter
Hydraulic loading m3/m2 . d 1-4
Depth m 1.5-3.0
Recirculation ratio 0
Filter media Rock, slag etc
Power requirements kw/103 m3 2-4 A
-l (0
Filter flies Many
Dosing intervals Less than 5 min
Effluent Usually fully nitrified
4-1 0
1.25-2.5
0- 1
Rock, slag etc
2-8
Intermediate
15-60 sec
Partially nitri- nitrified
10-40
1 .o-2.0
1.3
Rock, slag Synthetic materials
1-10
Few, larvae are washed away
Less than 15 sec
Nitrified at low loadings
FROM: Metcalf and Eddy (1991)
where:
SNH4,a = the content in the mixture of raw and recycled mixture applied to a
R = the ratio of the recycled flow to the influent flow.
medium.
' N h o + R* sNH40
1 +R 'NH4a = (5.18)
Gujer and Boller (1986) proposed the following line-fit equation for the decline
in the nitrification rate with depth in a trickling filter:
e-k'z 'n,z,t = 'n,z=O,t (5.19)
where:
z = depth in tower in metres.
rn,z,t = nitrification rate at depth, g N/m2 d.
rn,z=o,t - - nitrification rate at the top of the tower,
g NI m2 d.
k = empirical parameter describing the decrease of the rate with depth
(k varies between 0.075 and 0.16).
t = temperature in degrees Celsius.
Gujer and Boller (1986) developed a biofilm model for predicting the surface
nitrification rate as a function of the ammonia concentration in the bulk fluid, that
180
considered mass balance for oxygen and ammonia within the biofilm.
By combining equation 5.18 with the normal monod kinetic equation, Gujer and
Boller developed the following two solutions for the design of NTF's.
Introducing k (the empirical parameter) different from 0:
*(I -e-7=S,,,-Sn+N*ln- snl kr Vn Sn
and k = 0
(5.20)
where:
S, = bulk liquid ammonia nitrogen concentration in mg/l.
Sn,i = influent ammonia-nitrogen concentration in mg/l.
jn,max = maximum nitrification rate at high ammonia levels,
g N/m2 d.
jn(s,t) = nitrification rate at ammonia concentration g N/m2 d.
N = saturation parameter mg/l.
a = specific surface area of the trickling filter media in m2/m3.
(5.21)
V,= hydraulic load on the trickling filter in l/m2 s.
181
Because the oxygen transfer efficiency of different plastic media differs, the
following equation can be used to correct the nitrification rate for this difference:
where:
(5.22)
E = media effectiveness factor. The value of E depends on the media
used, see Table 5.8
j0, max(T) = maximum surface oxygen rate for specific media design
in g N/m2 * d.
The factor 4.3 in equation (5.22) reflects the unit mass of oxygen consumption
per unit mass of ammonia nitrogen oxidized.
Where recirculation is used, a repetitive solution of the above equation is
necessary because recycle effects are included in both the S,,, and V, terms.
The effect of the media on the nitrification rate is not considered in this modelling
approach.
5.6.3 The Application of the Trickling Filter Most trickling filters are used in single stage removal of organics. If the organic
loading is lowered to about 0.16 kg BOD / m3 d, combined oxidation of organics and
nitrification will occur, whereby a part of the influent ammonium will be nitrified (see
Tables 5.7 and 5.8). But single stage nitrifying trickling filters are also becoming
popular in treating secondary or tertiary influents, because the recent efforts in
improving these filters have made the effluent produced of a better quality, so the NTF
is comparable with the activated sludge processes in regard to nitrification efficiency
and amount of the suspended solids in the effluent.
The concentration of ammonium-nitrogen must be less than 25 mg/l to obtain
the best results in a conventional nitrifying trickling filter. The trickling filters are,
182
therefore, often used to treat municipal waste water, where BOD removal has already
been accomplished. Experiments have been made with the new compact plastic media
trickling filters in the treatment of industrial waste water of higher nitrogen con-
centration.
Combined oxidation of organics and nitrification.
Despite much interest in trickling filters, relatively little research has been made
on the simultaneous organic removal and nitrification taking place in a single trickling
filter unit.
The EPA (1975) showed that for rock media trickling filters, organic loading
must be limited to 0.16 kg BOD / m3 * d to attain 75% conversion of ammonium to
nitrate. Nitrification decreased at a higher organic load. At an organic loading of 0.64
kg BOD / m3 d, nitrification of only 10 % of ammonium was obtained. This reduction
in nitrification was attributed to the domination by heterotrophic bacteria of the
microbial biofilm.
The difference between rock and plastic media in loading capacity, as shown
in Table 5.5, was attributed to the higher specific surface area of the plastic, whereby
less competition between the species of bacteria was necessary.
Wanner and Gujer (1984) showed the concentration of ammonium versus different
COD concentrations for a trickling filter. They predicted that most of the organic
removal occurred in the upper reaches of the trickling filter, where heterotrophic
organisms dominated, and nitrifies were absent. Nitrification occurred at the highest
rates in the bottom portions of the tower where concentration of organics was the
lowest, and the autotrophic population could dominate.
Nitrification only occurred in the bottom half of the reactor. The most significant
nitrification occurred in the bottom 1.2 m of the filter. Most combined trickling filters do
not produce nitrate before the soluble BOD concentration is less than about 20 mg/l.
Figure 5.8 show the relationship between nitrification and soluble BOD, levels exposed
to the biofilm for cross-flow media.
Nitrification in a nitrifying trickling filter (NTF).
The NTF is designed to oxidize ammonia in secondary effluents, where most
of the BOD is already removed, so that the NTF can concentrate on the removal of
ammonium-nitrogen. The first demonstration of the system was a pilot scheme in
183
Michigan (Duddles eta/. 1974).
Typical removal rates, for conventially loaded NTF filters are as low as 0.20
to 0.39 g N/ m2 * d as indicated by investigations in the US. Ammonia removal
efficiency for rock and plastic filters at various sites in the US, applying different
amounts of organic loading per unit of surface in kg BOD / 1000 m2 day is shown in
Fig. 5.7.
\ O Rock Media - \ '+ -Plastic Media? '\ - Chino \
\ \ .\
\ -\ O
-* I I I 0 *-
f Removal of ammonia NH,, %
60
40
20
0
kg / 1000 m2 d, BOD,
Figure 5.7 Ammonium removal efficiency for rock and plastic filters at various sites in
the USA, applying different amounts of organic loading per unit of surface in kg BOD
/ 1000 m2 * day. After Parker and Richards (1986).
The EPA (1975) manual showed the removal rate for the NTF to be between
0.83 and 1.50 g N / m2 d. A conventional design practice has been to follow the NTF
with either effluent filtration or clarification.
Recognizing the costs advantages of operation and maintenance of NTF
technology, studies have been undertaken to assess the factors limiting the possible
184
nitrification rates, and to modify the processes of the NTF.
As a result of those studies Parker et a/. (1989) proposed the Biofilm-
Controlled-Nitrifying-Trickling-Filter (BCNTF). The new design incorporated weekly
flooding and backwashing of the BCNTF for predator control, and cross-flowed plastic
media were applied for better oxygen transfer to the biofilm, resulting in a higher
biomass content. The peak nitrification rates obtained for the BCNTF were between
2.3 and 3.2 g N /m2 d (0.32 and 0.44 kg N/ m3 d). The BCNTF process has
therefore, a peak nitrification rate of about 3 times the NTF process.
10 -
8 -
6 -
4 -
2 -
NITRITE AND NITRITE LEVEL AS N, mg/ L
t
0 0 0
1 I b
20 SOLUBLE BODS
ms/ L
Figure 5.8 Relationship between nitrification and soluble BOD, level exposed to the
biofilm for cross-flow media; After (Parker and Richards 1986).
Additional advantages of the BCNTF is the smaller land area needed and that
it can be constructed without disruption of secondary treatment operations. These
changes in design and other improvements have made the BCNTF very competitive
with the nitrifying activated sludge process.
185
Denitrification with a trickling filter.
Trickling filters are also able to conduct denitrification, when part of the filter
has low oxygen concentration, the presence of nitrate and a carbon source that can
act as an electron donor in the denitrification process. Effluent recycling is predicted
to be favourable to the denitrification process. Almost all NTF units can denitrify a part
of the formed nitrate-nitrogen, depending on the circumstances mentioned above.
Stenquist et al. (1974) mentioned an example where a combined trickling filter loaded
with 0.36 kg BOD / m3 d, caused denitrification of 25 % of the ammonium-nitrogen
applied to the plant, and 89 O h of the ammonium-nitrogen applied was nitrified.
5.6.4 Recent Developments in the Technology of the Nitrifying
Trickling Filters (NFT) The development of the Biofilm-Controlled-Nitrifying-Trickling-Filter (BCNTF)
(see Fig. 5.9) is the latest effort to enhance the nitrification rate in nitrifying trickling
filter technology. The BCNTF has a peak nitrification rate of about three times that of
a conventional NTF. The suspended solids (SS) from the effluent from a BCNTF are
almost the same as those found in the influent. If the existing secondary effluent,
therefore, is already of a high quality (i.e. the average effluent SS and BOD are less
than 15 mg/l) it has been shown in the literature that applying BCNTF is less costly
than using a conventional activated sludge process.
Further information about the BCNTF is presented above.
5.6.5 Nitrogen Loading Capacity and Removal Efficiency of the
Different NTF-applications Gulliecks and Cleasby (1986) proposed the curves shown in Figs 5.10A and
5.1 OB as design curves for application of nitrification to municipal secondary effluent,
which has been settled before use of the trickling filter. The filter used for these design
curves contained 6.55 m of vertical-type plastic media with a specific surface area of 88.6 m2/m3. Figure 5.10A is proposed for waste water with a temperature below 10 "C,
and Fig. 5.10B for temperatures between 10 and 14 "C.
The curves correlate the influent ammonia-nitrogen concentration, the applied
hydraulic flow in l/m2 s, with the expected yield in nitrification rate in kg N/m2 * d for
the trickling filter. It is important to note that the maximum range for the influent
186
ammonia-nitrogen and hydraulic flow on the axes of the curves. If the concentrations
in a sample of waste water exceed the values on the axes of curves presented in Figs
5.10A and 5.10B, it is then necessary to use recirculation in order to achieve a mixed
concentration, which is applied to the proposed curves for the use of the curves in
estimations.
TRICKLING FILTER / SOLIDS
CONTACT PROCESS (TF / SC) BlOFlLM CONTROLLED
NITRIFYING TRICKLING
F!LTER (BCNTF)
Figure 5.9 Applying Biofilm-Controlled-Nitrifying-Trickling-Filter (BCNTF) to process
application of a conventional trickling filter. After Parker et a/. (1989).
Figure 5.1 1 shows that there is a great variation of the peak nitrification rate
at different depths in an NFT.
This decline is attributed to the patchy development of the biofilm at greater
depths, caused by the absence of a continuous supply of ammonia to support biofilm
development at such depths. Most peak nitrification rates are, therefore, calculated for
the whole NTF, and not at certain depths in an NTF.
187
Table 5.7 shows the peak nitrifying rate for the main types of nitrifying trickling
filters. The Table indicates that the BCNTF system developed by Parker and co-
workers yields a peak nitrification rate of 2,3 to 3,2 g N / m2 d, which is high
compared with previously developed NTF's.
Stenquist et a/. (1974) reported that up to 25 % of denitrification (complete
nitrogen loss) were found in NTF plants, depending on the design, as indicated in
Section 5.6.1.
A
Applied NH,' - N mg/l.
(including recirculated NH4+ - N)
,075
O l I I I 0
0 0 0 5 1 0 1 5
Applied Hydraulic Load
I / s m2 01 cross Section
(including recycle)
B
Applied NH,' - N mg/l.
(including reclrculated NH,' - N)
25 f I ,075 10-3
\ \ \ ,125 x 10% N/dm
I 1 I C
0 0 0 5 1 0 1 5
Applied Hydraulic Load
I s m2 of cross Section
(including recycle)
Figure 5.10A and 5.108 The predicted removal of kg N / m2 * day of the media
surface, versus the applied hydraulic load and applied ammonia-nitrogen for nitrification
of a municipal secondary clarifier effluent at a waste water temperature below 10 "C (A) and between 10 O C and 14 OC (B). After Gulliecks and Cleasby (1986).
188
1.2
1.0
0.8
0.6
0.4
0.2
0
Nitrification Rate,
g N / m 2 * d
Temperature: 10 “C
0-120 cm (01
120-285 cm (4 a
A
((35-675 L l 2 4 5 4 3 5 cm cm (XI (01
I I 1 1 1 b
0 5 10 15 20 25
Ammonia Nitrogen Conc.,
mg/l
Figure 5.11 The nitrification rate as a function of ammonia concentration at four
different depths in a trickling filter. After Parker et a/. (1989).
189
Table 5.7 The results from different NFT’s as presented in the literature.
Organic loading Trickling filter media Petfomnce total possible Reference
0,16 kg BOD, /m3- d Rock 75% EPA (1 975)
0,64 kg BOD, /m3. d Rock 1 0% EPA (1 975)
0,36 kg BOD, /m3. d Plastic media 89% nitrification 25% denitrification
Stenquist et a/. (1 974)
2,5 kg BOD, /lo00 m2- d Plastic Cross f!ow media 92% Parker et a/. (1986)
2,5 kg BOD, /lo00 m2. d Plastic vertical flow media 60% Parker (1 976)
A 6.3 kg BOD, /lo00 m2. d Plastic media 42% CD 0 Garland Texas, USA
Parker et a/. (1986)
Table 5.8 Comparison of nitrification rates for different NFT plants.
Location Media Temp. "C Nitrification E' Reference
Midland, Vertical flow 13 1,20 0,86 Duddles G.A. et a/. (1974) Mich media
7 0,93 0,74 I
Lima, Ohio Vertical flow 21 1,70 1,Ol Sumpayo E.F.(1973)
Bloom Township Vertical flow 20 1,20 0,88 Baxter and Woodman (1973) Ill media
17 1.10 0,82
Zurich Switz. Cross flow 17-20 1,40 0,65 Richards (1 988) -. (3.9 m tower) plastic media 2 Zurich Switz. Cross flow 13 1,lO 0,39
(6,8 m tower) plastic media
I,
Central Valley Cross flow 18 2,60 0,80 Parker et a/. (1 989) plastic media
= Media Effectiveness factor (E).
5.6.6 Advantages and Disadvantage of the NTF
The following advantages and disadvantages can be listed for the application
of a nitrifying trickling filter.
Advantages:
Their simplicity and low operational cost make the trickling filters an attractive option
for small communities in warmer climates.
The recovery from hydraulic and substrate shock-loads is fast.
* It is possible to obtain a high content of biomass, especially when highly porous
plastic media are used.
Disadvantages:
Trickling filters achieve only with difficulty the high efficiency which is demanded by
recent effluent standards in many countries.
Most trickling filter effluent needs a polishing process, because the concentration of
suspended matter at high loadings is unacceptable for meeting effluent standards.
It is difficult to ensure an effective predator control, so the maximum nitrification rate
can rarely be obtained.
192
5.7 Rotating Biological Contactors (RBC) The Rotating Biological Contactor (RBC) is used for a variety of purposes:
Aerobic degradation of organic material; combined organic removal and nitrification;
and denitrification and nitrification of secondary and tertiary effluent (after filtration).
A rotating biological contactor (RBC) consists of a series of closely spaced
rotating circular discs made of different kinds of materials, for example plastic, wood,
and galvanized plates.
These discs are approximately 40 % immersed in a tank through which waste
water flows continuously. The discs are mounted on a shaft which usually rotates
through the water at a velocity of 1 rpm. A layer of biological growth, depending on the
composition of the waste water, builds up on the wet surface of the discs and forms
a biofilm ranging from 1-2 mm in thickness. The formation of a fully developed biofilm
takes from 1 to 4 weeks. As the discs rotate through the waste water, the ammonium
content is nitrified, and the organic carbon content is oxidized by various microorga-
nisms. Excess growth on the discs is disposed of at the same time. The discarded
biofilm is washed out of the unit and removed during a secondary clarification. As the
biofilm is passed out of the liquid and through the air, oxygen is absorbed to keep the
growth aerobic.
An RBC treatment plant will generally consist of a number of shaft trains, each
operating as a completely mixed, fixed-film biological reactor. Each train is generally
set up in a number of stages, separated by baffles for more efficient treatment and
stability. By doing so, it is possible to achieve a high degree of nitrification. Figure 5.12
shows a flow diagram of the rotating biological contactor process.
5.7.1 The Performance of the RBC The factors affecting nitrification in the RBC process are the same as in other
nitrifying plants, namely, organic concentration, influent nitrogen concentration and
composition, waste water temperature, DO concentration, pH and alkalinity, and
influent flow and load variability.
Most empirical design procedures are based on the assumption that significant
nitrification does not begin in an RBC system until the bulk liquid soluble BOD, has
been reduced to 15 mg/l. In combined carbon oxidation-nitrification units this will
193
typically first be encountered in the third or fourth stage, depending on strength,
organic loading rate and temperature of the influent.
Hydrogen ions are produced during nitrification. In poorly buffered RBC
nitrifying systems, alkaline chemicals such as lime, soda ash, and sodium hydroxide
may have to be added to the waste water in order to maintain a sufficient alkalinity to
prevent a sudden decrease in pH and thereby a decrease of the nitrification rate.
Rotating Biological Contactor
(RBC units) Primary clarifier Final Clarifier
t
To sludge treatment
Figure 5.12 Flow diagram for the rotating biological contactor process.
The media in an RBC.
It must:
1. provide a surface area for the development of a large, fixed, suitable biomass,
2. provide vigorous contact of the biological growth with the waste water,
3. aerate the waste water efficiently,
4. provide a positive means of continuously removing excess biomass, and
5. agitate the mixed liquid to keep the discharged solids in suspension and thoroughly
mix each stage of treatment.
The RBC media must serve several purposes according to Antonie (1976):
Many different materials for RBC media have been used over the years, from
the wooden slats of Poujelet in 1916 to the plastic discs used today. The discks are
usually 2-3 m in diameter and 1.2 cm thick.
Today the discs are made from a high-density polyethylene in alternating flat
194
and corrugated sheets, which are bonded together. This design provides more than
twice the surface area per unit volume than flat sheets.
The rotational speed of an RBC.
The rotation of discs serves a varity of purposes in the RBC process:
1. It provides a contact between the biomass and waste water.
2. Removal of excess biomass.
3. Mixing of the liquid and aeration of the waste water.
If the rotation rate is increased, the effects mentioned above are enhanced to
The optimum rotational speed for the RBC varies depending upon the
In practice, most RBC units are operated at 1.0 to 1.4 rpm.
a point above which further increase is not productive.
composition of the waste water and the disc size of the RBC.
The aeration of an RBC. In some RBC facilities, aeration equipment has been installed, either to drive
the RBC shafts or to provide supplemental aeration. RBC with aeration facilities usually
results in a thinner biofilm on the discs in the first compartments because of the
stripping action of the bubbles, thereby allowing more of the biofilm to remain aerobic.
It appears, however, that the dissolved oxygen in the mixed liquid has little
effect on the transfer of oxygen into the biofilm (see Fig. 5.13). A study of the mass
transfer of oxygen in the biofilm indicates that very little of the oxygen utilized by the
microorganisms in the film comes from the bulk liquid in the RBC tank; it comes from
the atmosphere, when the disc surface is exposed to air. The transfer of oxygen to the
biofilm is better increased by lengthening the exposure time to air or reducing the
thickness of the liquid film on the disc by more efficient emptying than by aerating the
waste water. Figure 5.13 show the relative concentrations of oxygen and substarte for the loading condition and RBC rotational spped as a function of the media.
195
A
D a media
IN ATMOSPHERE IN BULK LIQUID Distance from RBC
ronr.nt,atmn rmc*ntratim
I I
A - - Direction of process
Figure 5.13 The relative concentrations of oxygen and substrate for the loading
condition and RBC rotational speed as a function of location of the media.
The arrangement of multiple RBC units.
Staging of RBC media is recommended to maximize the removal of
ammonium. In secondary treatment applications, three or four stages are generally
provided for each stream. For small installations, four stages can be provided on a
single shaft by installing three inter-stage baffles within the tank, and introducing the
flow parallel to the shaft. Installations requiring two RBC units may be placed in series
with a single baffle in each tank, thus providing four stages. Four or more units can be
placed in series, with each unit becoming a single stage. Various schemes of staging
RBC units are shown in Fig. 5.14.
196
One unit, four stages - Two units in series,
two stages each
Three units in parallel, +Gz- four stages each
Multiple parallel flow streams,
four or more units per flow stream,
single-stage units
Figure 5.14 Various schemes of staging RBC units.
The biomass of the RBC.
If an RBC is supplied with secondary influent, the unit will be divided into four
sections. The first section will not be able to accomplish nitrification, because of a high
content of organic matter and, therefore, no nitrifying population will be able to develop.
Both nitrification and organic oxidation will be carried out in the second section.
The waste water content of ammonium is high and, therefore, the nitrification is relying
upon on the oxygen content in the waste water and on the size of the nitrifying
biomass, developed in relation to the size of the heterotrophic biomass.
In the third section most of the organic load in the waste water is oxidized, so
197
that this section will function as the nitrifying section. As in trickling filter processes,
nitrification will only proceed after the carbon concentration has been substantially
reduced. Only the oxygen content in the waste water will limit the nitrifying rate.
In the fourth section the ammonium content is so low that it is not the oxygen
content that limits the nitrification rate, but the content of ammonia itself.
It is, therefore, important to have full control over the organic content in the different
parts of an RBC plant, and to design at least four or five modules in series, if nitrifica-
tion is required, because as illustrated in Section (3.13) ammonia itself can inhibit the
nitrification rate.
5.7.2 Equations for Modelling the RBC Reactor Matsuo and Yamamoto (1985); Watanabe (1985) and Gujer and Boller (1990)
have all modelled the process of RBC units.
Gujer and Boller (1 990) proposed a model containing two levels: a microscopic
level and a macroscopic level. The microscopic level considered the transport and
reaction processes within the biofilm. The macroscopic level described the system as
a whole. Mixing conditions within the individual compartments, influent and effluent
transport processes, gas exchange processes, exchange of substrate, nutrients and
biomass within the biofilm, and reactions catalyzed by biomass in suspension were
considered as important factors in the performance of the RBC.
Three submodels, a kinetic model, a biofilm-model, and a reactor compartment
model were proposed to take account of the above factors.
The kinetic sub-m odel.
The equations proposed by Gujer and Boller (1990) for the kinetic sub-model
were the same as outlined in Section 5.4 for the biofilm kinetics, depending on either
zero order or half order kinetics.
The biofilm sub-model.
The biofilm sub-model takes the following variables into account: the dissolved
components, the particulate components, the removed biomass, the surface floccula-
tion and the thickness of the biofilm. The different equations used in this sub-model are
shown in Table 5.9 and the relevant constants are given in Table 5.10.
198
The reactor sub-model.
The reactor sub-model takes into consideration the rotation of the RBC and the
design of the reactor compartment.
Variation in the concentration of dissolved oxygen in the depth of the biofilm,
due to rotation of the RBC, depends upon the processes of diffusion and reaction. The
depth of penetration of dissolved components due to molecular diffusion is given by:
Li = (Di’ t)% (5.23)
where:
Li = Depth of penetration of compartment i during
time t in metres.
D,’ = Effective diffusion coefficient within the biofilm,
assumed to be 80% of the value in pure water.
t = time.
199
Table 5.9 The different equations used in the biofilm sub-model for the RBC.
Process Equation Symbols
Transport of dissolved components
N 0 0
Dissolved component
where J = 0 for z = LB
dS, =dTsl +rs, dt dz
z = depth of biofilm; z = 0 at surface and z = LB at support material.
J = flux of component i due to molecular diffusion within the biofilm.
Si (z) = concentration of dissolved com- ponent i at biofilm depth z.
Di = effective diffusion coefficient within the biofilm, assumed to be 80% of the value in pure water.
rs,i =transformation rate of the dissolved component i per unit volume of biofilm.
Table 5.9 (continued)
Process Equation Symbols
Particulate components Xi = concentration of particulate species i within the biofilm.
%o, = sum of all particulate species con- centration.
where
Surface flocculation
Jz,i = flux of particulate species i within the biofilm.
rx,i = rate of production of particulate species i within the biofilm.
, J,Lo,,i = flux of particulate material i floc- culated form from the bulk liquid to the surface of the biofilm.
k,,, = flocculation mass transfer coeffi- cient.
Table 5.9 (continued)
Process Equation Symbols
KSHEAR = 0,lO d-' for primary effluent
= 0,05 -' for secondary and tertiary effluent.
Biofilm thickness
JSHEAR.I = KSHEAR . LB ' '6,i JsHEA,,i = flux of particulate material i, sheared from the surface of the biofilm to the reactor bulk liquid.
KSHEAR,, = Shear rate constant.
XB,i = concentration of particulate material i, at the surface of the biofilm (z=O)
Table 5.10 Different constants proposed for use in the Biofilm sub-model.
Symbol Unit
Heterofrophic organisms:
Maximum growth rate
Saturation coefficient
for COD
for NH,’ -N
for 0,
for HCO,
for NO,’
Denitrification coefficient
Decay rate
Yield coefficient (CODICOD)
Fraction particulate decay product
Nitrogen content of biomass (NICOD)
Decay product
Nifrosomonas
Maximum growth rate
Saturation coefficient
for NH,-N
for HC03-
for 0,
Decay rate
Yield coefficient (CODIN0,- -N)
Nifrobacfer
Maximum growth rate
Fmax. H
Ks
K N H ~ +
KHA
KHNO
DEN
KHO
bH
YH
’1
Ymax,N
KNH4
bN
2,OO d-’
IO,O g/m3
O,I g/m3
O,I g/m3
O , I O g/m3
030 g/m3
0,35 d-’
0,70 g/m3
0,20MoMm3
020 g/m3
0,05 d-’
0,18 glg
0,60 d-’
203
Table 5.10 (continued)
Saturation coefficient
for NH4+-N
for NO,'-N
for 0,
Decay rate
Yield coefficient (COD/NO,--N)
KNH4+ 0,05 g/m3
K N 0 2 030 g/m3
KO2 O,I o g/m3
~ N B 0,09 d-'
YNB 0,06 g/g
Diffusion coefficients within a biofilm correlated for temperature (10 "C) and
reduction to 80% of values in pure water.
Dissolved oxygen 106 . 1 0-6 m2/d
Degradable COD 31 . m2/d
Ammonium 86 . m2/d
Nitrate 85 . m2/d
Nitrite 84 . m2/d
Bicarbonate 53 . m2/d
Rate constants for biofilm surface reactions
Flocculation rate KFLOC 0,lO d-'
Shear rate constants KSHEAR Primary effluent 0,lO d-'
Secondary and tertiary effluent 0,05 d-'
204
Gujer and Boller (1990) demonstrate that it is only at a very slow rotation
speeds and a low concentration of residual pollutants, that the effect of the rotation of
the support material must be considered.
RBC's today are usually compartmentalized; each drum of rotating surface
areas is calculated as an individual reactor compartment.
For the Nth reactor compartment the substrate balance is written as:
(5.24)
where:
VN = Volume of reactor N (bulk water phase) in m3.
Ci,,, = Bulk concentration of dissolved or suspended component i
in reactor N in kg per m3.
Q + R = Influent and recycle flow rate in m3 per hour.
ii,N = Rate of production of component i within the bulk liquid in kg
i/m3 * day.
Ji,N = Flux of component i into the biofilm of reactor N, in kg / m2 day.
AN = Surface area of support material in reactor A in m2.
The entire reactor system can then be modelled as a series of reactors with the
option of recirculation of effluent from the last compartment to the first one. It is also
possible to reverse the flow, either from the first to the last or, alternatively, from the
last to the first reactor.
205
5.7.3 The Application of the RBC Rotating Biological Contactors are popular in small-scale waste water treatment
plants (c 100 P.E.), because of their easy maintenance, low sludge production, and
low power requirements. One of the key costs in the production of an RBS is the
support material. The RBC is therefore, ideal for the treatment of small volumes, which
cannot easily be connected to a central treatment system for economic or geographic
reasons.
The RBC can be used as a combined oxidation and nitrification system for
secondary effluent, or as a tertiary nitrifying system depending on the composition of
the influent waste water. Examples of anaerobic denitrification with an RBC are also
presented in the literature (Hosomi et a/. 1991).
The different functions of the RBC in the nitrogen removal processes are listed
below:
A) Combined oxidation and nitrification with an RBC unit.
Treatment of industrial waste water or small scale waste water treatment plants,
with an RBC treatment plant, will usually provide a unit with combined oxidation of
carbonic material and nitrification of ammonium to nitrate. Because the waste water is
a mixture with high carbon and ammonium contents which act as substrate for both
oxidizing and nitrifying bacteria, the bacteria will compete for the space on the RBC
disks.
B) Nitrification with an RBC unit.
Using the RBC as a tertiary nitrifying treatment plant has been shown to be
highly efficient. The RBC units can, therefore, be expected to be used for final
refinement, in an effort to reach present effluent standards for nitrogen content,
because they can be integrated into existing flow schemes.
C) Denitrification with an RBC unit.
Denitrification in RBC systems may be observed in the following two situations:
1) Addition of oxygen may be reduced, either by nearly complete submersion of the
rotating contactors, or by maintaining an atmosphere poor in oxygen. This situation will
allow denitrification even at low levels of the organic loading, if nitrate is fed to the
reactor, for example via recirculation of nitrate from the effluent waste water.
206
2) Denitrification may occur in the depth of a biofilm, where oxygen has fallen to
insignificant levels. This requires high concentration of organics and nitrate to secure
denitrifying conditions in the depth of a biofilm. Since the recirculation of nitrate results
in dilution of the concentration of carbon compounds, this situation may only occur with
industrial waste water where concentration of carbon compounds is high.
D) Simultaneous nitrification and denitrification (SND) with an RBC unit.
Masuda eta/. (1991) reported the loss of nitrogen in an RBC plant treating the
leachate from a sanitary landfill located at Miyazaki, Japan. The loss of nitrogen was
appreciable during the summer. The RBC was covered by a hood, and during the
summer the oxygen pressure in the hood was 1.8 to 1.9 atm., which was a little less
than in an unhooded RBC. Matsuda and his co-workers measured the production of
nitrogen gas from the biofilm, using a covered RBC, to be able to observe denitrifica-
tion. In order to explain the loss of nitrogen, the authors made the following hypothesis:
Nitrifiers and denitrifiers co-exist in a biofilm; the denitrifiers become active, if the
transfer rate of oxygen to the biofilm decreases sufficiently to result in the formation
of a micro-anaerobic environment.
Halling-Sarensen and Hjuler (1 992: 1993) have observed the same occurrence
in a submerged filter, using clinoptilolite as the media.(See section 5.8.1).
Masuda and his co-workers conducted a series of experiments to discover the
factors which influence simultaneous nitrification and denitrification (SND).
They showed that the highest capacity of the SND in the RBC unit, in midsummer, was
a total conversion of 130 s/m3 NH,' - N to 80 s/m3 gaseous nitrogen N,, and the
remaining 50 g/m3 was converted to NO, - N. The efficiency of SND was accordingly
61.5 %.
5.7.4 Recent Development in the RBC Technology Much effort has been made to enhance both capacity and effluent quality of the
RBC treatment systems because in the near future, full nitrification will have to be
adopted in many treatment plants in order to reach present effluent standards. Using
the RBC as a tertiary treatment step, it can in most cases be integrated into existing
flow schemes.
207
Bolter eta/. (1990) proposed a two-stage nitrifying RBC, including precipitation,
primary clarification and a solid separation step; this was after the use of the first BOD-
removing RBC's using a cloth filter, and finally a nitrifying RBC with the possibility of
reversing the flow so as to obtain a higher utilization of the surface of the biomass
carrier throughout the RBC. (see Fig. 5.15). This design also provides a better
possibility for fluctuations of ammonia in the waste water because of the higher
nitrification potential. This type of RBC plant, with filtration before nitrification and two-
sided loading of waste water, requires about 40% less surface area and volume than
a conventional RBC with one-sided flow, where a nitrification capacity of 1,8-2,9 g N/
m2 * d, is established.
To further enhance the capacity of the RBC, Wanner eta/. (1990) proposed a
packed-cage RBC. This is an RBC which is a combination of suspended and fixed-film
biomass. The discs or groups of discs from a conventional RBC were replaced with a
cage, packed with a random medium. The cage is equipped with tubular aeration and
mixing elements.
The combination of suspended and fixed film biomass should enhance the
capacity for nitrification and lower the the cost, because aeration of the activated
sludge is separated from the rotation of the RBC and, therefore, an external source of
air is avoided, and a large biomass is developed. This combination should make this
design suitable for plants handling the treatment of 500 to 800 P.E.
To improve the effluent quality of the RBC process, Tanaka et a/. (1991)
investigated the behaviour of the fine particles throughout the processes; they found
that an increase in the hydraulic retention time in the RBC reduced the amount of fine
particles and increased the amount of coarse suspended solids, which are easily remo-
ved by clarification.
5.7.5 Nitrogen Loading Capacity and Removal Efficiency of the
RBC-process If nitrification is desired, loading rates should be reduced to 0'03 to 0,08 m3 /
m2 d. This is about one third of the capacity of an RBC when applied only to the
removal of organics.
208
9 fication
- -------a
Back-
wash
water
Figure 5.1 5 The two-stage nitrifying RBC, including precipitation, primary clarification
and a solids separation step after the first BOD removing RBC's with a cloth filter and
lastly an RBC for nitrification. (From Gujer and Boller 1990).
The following mass balance equation can be used to calculate the removal of
ammonium per m2 filter per day as an average for the whole filter.
Figure 5.1 8 show the relationship between nitrification capacity and temperature in an
RBC unit.
where:
Q = loading rate for waste water in m3 /day.
A'= the total disc area in m2.
(5.25)
dNH,+/dt = g NH; - N /(m2 * d) removed.
209
'0, 0
P , ji n 0
0
0 1 I I I I I
! I
0 5 10 15 20 25
Temperature "C
Figur 5.16 The relationship between nitrification capacity and temperature in an RBC unit (partly after EPA (1984) and La COUr Jansen and Henze (1 990)).
Figure 5.16 show the relationship between nitrification capacity and temperature
in an RBC unit, and Table 5.1 1 some examples of nitrifying removal rate for the RBC
using different types of waste water. Table 5.1 1 show the removal rate with different
applications of the RBC.
Table 5.1 1 The removal rate for the RBC using different types of waste water
Wastewater
type
Nitrification
rate
g N /m2 d
maximum minimum
Domestic 1,69
Percolate 2,42
Fertilizer industry 2,36
Leather industry 2,35
Sewage water 133
2,56
2,66
2,67
2,62
1,97
Source La Cour Jansen and Henze (1 990)
21 1
Table 5.12 The removal rate with different applications of the RBC.
Treatment Treatment step Capacify Temp. Process rate nitrification Reference "C gnVd - d
Nitrification
Com bined N itrificat ion/oxidation
Combined nitrification/oxidation
2 Combined n itrif ication/oxidation
tertiary 800 P.E. 1 ooc 1,8-2,9
secondary (combined RBC and less than activated sludge) 100 P.E. 1,04
compact RBC 0,6
RBC with simultaneous - nitrification and denitrification 1 ,o
Boller et a/. (1 990)
Wanner et a/.(1990)
Ahn and Chang (1991)
Matsuda eta/. (1991)
5.7.6 Advantages and Disadvantages of the Nitrifying RBC The following points summarize the major advantages and disadvantages
connected with the use of the RBC in the process of nitrogen removal in treatment
plants.
Advantages:
1. Only a small land area is required.
2. Ability to obtain a high content of biomass per m3 or m2 of disc because of the
highly developed disc units and, therefore, the lower contact time with the waste water.
3. Simple operation of the equipment.
4. Ability to handle shock loads and, therefore, suitable for treatment of highly
concentrated industrial waste water.
5. Ability to achieve a high degree of waste water purification, including nitrification.
6. Good performance even with a low oxygen level in bulk waste water because most
oxygen is absorbed during rotation in the air phase.
7. Good performance of tertiary nitrification and, therefore, a solution to introduction
of full biological removal of nitrogen in existing plants.
8. Using an RBC unit, pumping large amounts of waste water is avoided, because the
water is passed slowly through the basin, where the contactor is rotating.
21 3
Disadvantages:
1. Enclosures are necessary to protect against low temperatures, rain and wind.
2. High capital cost.
3. Upsets can and do occur, because of too great wash-out of biofilm.
5. Most RBC’s are mainly designed for BOD removal, although some nitrification may
occur in some plants.
4. Lack of documented operating experience.
21 4
5.8 Submerged Filters Submerged biological filters (also known as biological aerated filters or contact
aerators) are filters where the fixed material upon which the biofilm develops is
continously submerged in the waste water they treat being treated.
The use of submerged filters has received renewed interest because of the
development of plastic media and other sorts of filter media upon which large
quantities of bacteria can grow. Also it has been shown that submerged biological
filters may be very efficient at nitrification. Dillon and Thomas (1990).
Submerged filters (Fig. 5.1) can be designed both in up-flow or down-flow
modes. In both cases, there is often a combination of both fixed-film and suspended
growth between the filter media. Air is supplied to provide oxygen to the microorga-
nisms, to promote mixing, and to scrub excess biofilm from the filter media to prevent
irregular sloughing and plugging problems.
Because of the large biomass concentration, the contact time is often, low
compared to other treatment systems to achieve the same efficiency.
Only few submerged filters are installed for nitrification. But indications are that
they can be cost effective from both a capital standpoint and an operation and
maintenance standpoint, that they reduce land area requirements, and that used, as
teritiary treatment, have an efficiency of up to 90 percent of nitrogen removal with very
low retention times.
Examples of the use on full size plants, of a biological fixed-film reactor for
combined oxidation and nitrification treatment step for treatment of municipal waste
water, are the systems Biocarbone and Biofor developed by respectively O W and
Degremont. (Dillon and Thomas 1990; Gilles 1990; Mange and Gros 1990; Paffoni et
a/. 1990, and Rogella and Bourbigot 1990).
The Biocarbone process use grain-sized biodagene (expanded schist) as
bedvolume and the Biofor use spherical biotite as bedvolume.
The Biocarbone is a counter-current, granular media, aerobic filter with a water
down-flow and an air up-flow. Its name is related to the earlier use of activeted carbon
as matrix. Biofor is an abbreviation form from BlOlogical Oxygenated Reactor. Biofor
is defined as an aerobic treatment using fixed biomass on a 1-5 mm granualr medium
with an upflowing co-current of injected air and water (Paffoni et a/. 1990).
These two processes primaryly differ in the fluid direction, namely a cocurrent
in the Biofor process and a counter-current in the Biocarbone process.
215
The case study presented involves the use of simultaneous nitrification and
denitrification (SND) with an upflow fixed bed, applying clinoptilolite as matrix. The
combination of nitrification and denitrification in one single reactor has been described
in the literature (Matsuda eta/. 1987; 1991) The development of the process described
has been conducted at the Section of Environmental Chemistry in the Royal Danish
School of Pharmacy in Copenhagen, Denmark. During the summer 1992 the first pilot
plant project was built as tertiary treatment step of slaughterhouse waste water.
5.8.1 Case Study;
Simultaneous Nitrification and Denitrification (SND) as Tertiary Treatment Step, Using a Submerged Biofilter of Clinoptilolite
Introduction The potential for using a simultaneous nitrification and denitrification (SND)
upflow fixed bed reactor (UFBR) as a tertiary treatment step for removing nitrogen from
waste water is examined in this case study. Clinoptilolite (with a grain size of 2.0-4.0
mm) was used as supporting medium for the bacterial growth. As indicated in Chapter
9 on ion-exchange, clinoptilolite is a natural zeolite which selectively sorbs NH,'. Furthermore the media has a porous surface, and has a high specific surface area,
ideal for bacterial growth. The removal of the adsorbed ammonium from the zeolite by
nitrifying bacteria allows regeneration of the zeolite surface and thus enables the same
zeolite to be used repeatedly.
Thus the purpose of this case study is to explain the mechanisms and show
the results of a single-stage simultaneous nitrification and denitrification (SND) reactor
that biologically transforms ammonium-N to nitrogen gas, with ethanol as electron
donor for denitrification.
Laboratory reactors were constructed (Fig. 5.17) of plexiglass tubes and used
in three different runs using clinoptilolite as media. The loads conducted during the 3
different experimental runs, each of the duration of six months, are presented in Table
5.14.
216
Table 5.13 Denitrification rates depending of nature of medium surface in packed-bed coloumns
Nature of
medium surface
Media trade name Specific surface area Lknitrification rates
(cm2/cm3) cJm3 d (at stated temp 'C)
High porosity corrugated sheet modules or dumped media
Kock Flexirings
Envirotech Surface
lntalox Saddelse
Rashig Rings
Filter A and B
Low-porosity media
2,13
3,34
45 (13"C), 53 (15"C), 54 (17"C), 136 (2OoC), 115 (2loC), 61 (23"C),
336 (27°C)
0,89 40 (1 0-23OC)
4,66-8,99 216-417 (20°C)
2,59 192-304 (25°C) 1 00- 1 20 (20°C)
47,8 + 495 (25-35 "C)
200-400 (20 "C)
After: EPA (1975) and Metcalf and Eddy (1979)
Outline of the 3 experimental runs:
Run 1 : Waste water containing Ammonium-N, and COD (Chemical
oxygen demand) in the influent waste water.
Run 2: Waste water containing Ammonium-N, Nitrate-N and COD in the
influent. The Nitrate-N was introduced to the waste water to see if
denitrification could proceed. The clinoptilolite media do not bind nitrate-N.
Run 3: Waste water containing Ammonium-N without applying COD to
the waste water. This should prevent denitrification from proceeding, and
the influent ammonium-N should be recovered as nitrate-N.
A pilot-plant to treat, tertiary stage, industrial waste water using clinoptilolite
as media, were built, at the Island of Fyn in Denmark.
Results and discussion As indicated in Chapter 9 clinoptilolite is a natural zeolite, which selectively
sorbs NH,'. The ionbinding capacity is 1.3 meq/g media (Jsrgensen eta/. 1976). The
efficiency of fresh support matrix, is therefore high until the ion-exchange capacity is
used up. The removal of ammonium from the waste water will thus decline until the
introduced nitrifying biomass becomes sufficient to convert all of the influent
ammonium.
The step of biomass development is critical, because if the developed ratio of
nitrosomonas and nitrobacter is out of balance, breakthrough of nitrite-N (NO,-) will
appear in the waste water and inhibit the development of nitrosemonas. If the biomass
concentrations of the two bacteria species are adjusted, the nitrifying efficiency is
raised. Figure 5.18 show the removal efficiency during the 10 first days of biomass
development, on previously unused clinoptilolite. Because of its ionbinding capacity,
clinoptilolite will bind nearly all ammonium in the first few days. After the capacity is
used up, a breakthrough of ammonium will appear until the concentration of nitrifying
biomass will be able to convert some of the ammonium to nitrate.
When nitrate-N is developed during nitrification and suitable conditions (anoxic
and carbon source) exist, denitrifying bacteria will be developed and nitrate can be
converted to nitrogen gas.
218
Figure 5.19 show the first six days running of a 30 mg/l ammonium influent on
a clinoptilolite reactor. The first three days of treatment, a breakthrough of nitrate was
observed in effluent samples, because of lack of denitrifying bacteria. For first day, 5.2
mg/l of nitrate-N was found. This amount declined the following days, due to the rapid
development of denitrifying bacteria.
f
A
0 0
c)
3
+r
Efficiency
Experimental RUN 1.
Simultanous nitrification and denitrification (SND) of waste water containing
ammonium and a organic source, measured as COD, in the influent waste water, was
conducted during RUN 1.
1 Effluent
T
.- - "I n o Sampling port S T
Synthetic
waste water
Influent $20 crnt
1 I 1
Gas Collection Bottle
Figure 5.17 Laboratory reactor used to conduct experimental Runs 1 to 3.
219
Table 5.14 Loads conducted during Run 1 to 3 applying clinoptilolite as media.
Run 1
NH,' -N mgA 30=> 1000
NO;-N mgA
COD inf. mglL 120=>4OoO
PH 7,7 - 7,8
Flow vhours 0,8 => 5,3
Oxygen conc. mgA 2 - 3
Reactor media clinoptilolite stones
Grain size mm 2 - 5
Void volume liter 8
Bed volume liter 30
Bedlvoid volume 3,75 ratio
Average pore meq4 0 ,u
N-ionbinding 113 capacity
Reactor volume I 33
Reactor high 1,05
Reactor dia- 0,20 meter m. Intervals between samplingports. mrn 250 Number of samp- lingports 4 SND occurred YES
Run2 Run3
30 30=> lo00
up to 180
800 No
up to 70
7,7 - 7,8
0,Q => 3,O
7,2 - 75
1,2 => 5,O
2 - 3 2 - 3
clinoptilolite clinoptilolite
2 - 5 2 - 5
8 8
30 30
3,75 3,75
0,44 0,44
1 3 1 3
36 36
1,15 1,15
0,20 0,20
250 250
4 4 YES NO
220
Removal of ammonium NH4+ - N, %
100
4.0
2.0
of
-
-
- -
-
days
Figure 5.18 The removal efficiency during the first 10 days of biomass development
on previously unused clinoptilolite.
effluent NO,. - N. mg/l t 6'o t
0 1 I I 1 I I I & 1 2 3 4 5 6
Number of days
Figure 5.19 Application, during the first 6 days, of waste water containing 30 mg/l
NH,+ - N, with a reactor of previously unused clinoptilolite. In the first three days, a
breakthrough of nitrate was observed.
22 1
The removal efficiency of the simultaneous nitrification and denitrification
(SND) and the effluent concentration of ammonium-N and nitrate-N were measured
throughout this run. Table 5.14 summarizes the results of RUN 1 where an organic
carbon source was applied to the system in stoichiometrically correct amounts. Table
5.1 5 shows some examples of mass balance for the simultaneous nitrification and
denitrification. The amount of SND is equal to the amount of ammonium-N which in the
reactor is totally converted via nitrate-N to nitrogen gas (NJ.
Figure 5.20 shows the relation between the loading of ammonium-N in kg N/m3
voidvolume day versus the simultaneous nitrification and denitrification (SND) reaction
rate in kg N/m3 * day. The SND reaction rate as bed volume, is calculated in kg N/m3
day, as:
SND(kgMm3 bed volume * dafl=[ N,,d -[ Nn,,j *E +8/30 (5.26) HRT
where: HRT is the hydraulic Retention Time, in hours.
Nintl = NH,' - N
N,, = NH,+- N,, + NO; - N,, NO; - N in,, + NO, - N infl
+ NO; - N
The factor 8/30 is the conversion factor between void volume and bed volume
The maximum amount of SND obtained during RUN 1 was 13.5 kg N/m3 void
volume * day (= 3.6 kg N/m3 bed volume * day) and the efficiency of SND is up to
99 % with a loading of up to 14 kg N/m3 void volume day under the following condi-
tions: temperature 20 OC, DO 2-3 mg/l, pH 7.7-7.8 and stoichiometric addition of
organic compound as ethanol to obtain denitrification.
Because Fig. 5.20 yields a linear relationship between loading and SND the
theoretical maximum amount of SND is not found. The linear equation (5.27) can be
obtained from Fig. 5.20 by linear regression. ? for the linear regression is 0.99.
SND=O.97*[ Nw,,J +0.014 (5.27)
222
Data obtained from several sampling ports along the reactor presented in Fig.
5.17 yield the following equation estimating the simultaneous nitrification and
denitrification, at a specific height "2' along the reactor in kg N / m3 bed volume * day,
knowing the SND at height T.
10
SND kg N I m3 voidvolume * day
-
.
Loading kg N I m3 voidvolume * day _.usJ
,
I I-
/
/' / w I
Figure 5.20 Relation between the loading of ammonium-N in kg N/m3
bedvolume" day versus the simultaneous nitrification and denitrification (SND).
For a different substrate concentration k may be inserted in equation (5.28).
(5.28)
For ammonium concentration in the influent below 100 mg/l NH4- - N ; k=0.080
(s=18.1%). Between 100 and 500 mg/l NH4- - N; k=0.065; (s=29.2%) and between 500
and 1000 mg/l NH4- - N; k=0.034 (s=21.9%). s is the standard deviation.
223
Table 5.15 Mass balance of SND using clinoptilolite as media.
input NH4+ - N
NO, - N
Total N
output
Removed by
SND
Effluent water
NH4+ - N
NO, - N
Total
30,O
02
30,2
*29,5
0,51
0,22
30,2
unit mg/l
100,o 500,O lo00,O
04 0,7 0,4
100,4 500,7 1000,4
*96,8
1,96
1,65
100,4
*499,7 %7,4
0,72 42,3
0,23 0,7
500,7 1000,4
* The removal by SND is found as the difference between influent and
effluent waste water samples.
The relationship between the amount of SND removal in mg/l and the hydraulic
retention time (HRT) is shown in Fig. 5.21, for the different feed concentrations shown
in Table 5.15. From Fig. 5.21 it can be seen that the SND removal does not decrease
by reduction of HRT with the ratio of HRT used during RUN 1. A greater flow through
the reactor increases the daily removal capacity of the clinoptilolite medium, and the
efficiency remains the same. This may be explained by considering that the surface
that the bacterial population can occupy on the media is so large, that the maximum
utility of the surface is not reached during the experiments. Therefore the use of a
lower HRT and thus higher daily loading, yields a larger biomass, and therefore a
higher capacity. Perhaps a higher flow can mechanically wash out the dead and older
bacteria from the reactor and thereby also provide new surface for fast development
of a fresh, new biofilm.
224
Table 5.16 Efficiency of SND during RUN 1, using clinoptilolite as media.
Flow Relsndlon L d n g Temp 00 pH NH,' -N NO; -N SND €Md-- SND LI, tlme kgNhd "C wn Inf Eff Efffd- h f Eff k g N H m y Kg N/ms
void m f l mgn mgn mfl vald- % ~rrpporl volwne % volume media 'day 'day *day
t H W
1.2 6.66
2.0 4,04
3.6 2.22
03 8.88
1.5 5.33
2.0 4.10
23 3.25 ru
5.3 1 ,so 1.5 5.33
2-6 3.03
4,7 1.70
0 8 10.03
2-4 3.38
4,7 1.70
0,108
0,178
0,324
0,270
0,450
0,585
0,738
1,600
2,250
3,960
7.059
2.392
7.100
14,118
20
20
20
20
20
20
20
20
20
20
20
20
20
20
2.0-3.0
2,0-3.0
2,O-3,0
2,O-3,0
2,040
2.0-3.0
2.0-3.0
2.0-3.0
2.0-3,0
2.0-3.0
2.0-3.0
2.0-3.0
2,O-3,0
2.0-3.0
7.7-7.8
7.7-7.8
7,7-7.8
7.7-7.8
7,7-7.0
7,7-7,8
7.7-7,0
7.7-7.8
7,7-7.8
7,7-7.0
7.7-7.8
7,7-7,8
7.7-7.8
7.7-7.8
30,O
30,O
30,O
100.0
100.0
100.0
100.0
100,o
500.0
500,O
500,O
1000,0
1000.0
1000.0
1.49
0.22
031
1.44
1,61
0,70
138
1.96
0.67
0.71
0.72
2.23
5.51
42.3
950
99,3
98,3
98,6
993
98.6
98.0
98.0
99,9
99.9
99.9
998
99.4
95.8
0.10 1.30
0,lO 2.12
0.20 022
0.40 4,63
0,lO 3.95
0.70 625
0.10 1,65
0.40 0.49
0,46 0,49
0.42 0,30
0.70 023
0,12 1,97
0.26 0.66
0.35 0,70
0,098
0.165
0.31 9
0,255
0,425
0,412
0,683
1.550
2,248
3.956
7.055
2.383
7.059
13.516
90,7
92.7
98,s
94,4
70.4
92.5
96.9
99.9
99.9
99,9
99,9
99,6
99.4
957
0,026
0,044
0.085
0,068
0,110
0,182
0,412
0.598
0.598
1,052
1,877
0,634
1.878
3,595
NH4 +-N ' mg/L
t 1100
1000
900
800
700
600
500
400
300
200
100
0
- 0 n - 0 1000 mg/L
-
-
-
-
-0 - - W 500 mg/L
-
-
- a A L - U " , " v 100 mg/L - - HRT, hours
I " r I " I 30 mp/L I m
Figure 5.21 HRT vs. SND at different influent concentrations during RUN 1.
RUN 2. The results from RUN 2 were obtained with waste water containing ammo-
nium-N, nitrate-N and an organic compound source in form of ethanol, applied to the
clinoptilolite reactor. The clinoptilolite medium does not bind nitrate as it does with
ammonium.
The aim of this run was to observe if an addition of both ammonium-N and
nitrate-N would be converted simultaneously to nitrogen gas.
Table 5.17 shows the efficiency of SND during RUN 2. The reactor was thus
able to denitrify both the added amount of nitrate and the amount produced during the
nitrification of the added ammonium.
The efficiency of Run 2 was comparable with Run 1, for the influent
concentrations applied.
The conclusion is that it is possible to perform denitrification of nitrate added
in excess of the nitrate produced by conversion of ammonium by nitrification.
226
Table 5.17 Efficiency of RUN 2.
Flow Retention Loading MI,+ - N NO, - N SND time k g N / m 3 Inf eff Inf eff kgN/m3 (HRT) matrix per day mgfl mgA mgfl mgfl matrix per day
1,2 6,67 0,059 30,O 0,28 30,8 7,59 0,051
1,2 6,67 0,043 30,O 0,26 14,l 0,lO 0,042
3,O 2,67 0,501 30,O 0,88 177,9 105,3 0,244
1,2 6,67 0,501 30,O 1,41 99,0 30,5 0,377
0,9 8,89 0,150 30,O 1,29 70,l 0,lO 0,095
1,2 6,67 0,055 30,O 0,12 27,3 0,20 0,055
Temp; 20” C , DO; 2-3 mg/l in bulk solution, pH = 7.7 - 7.8. Nitrite - N was not detected in the samples.
RUN 3 During RUN 3, ammonium-N was added to the clinoptilolite reactor without any
continuous additon of a carbon source for denitrification. This should make it possible
to recover the nitrate or nitrite produced during the nitrification process. The aim of this
run was to be able to determine the efficiency of the nitrification process alone. For
RUN 1 and RUN 2 the efficiencies of nitrification and denitrification are difficult to
separate.
The clinoptilolite applied was fresh, to ensure that no organic compounds were
left from previous experiments which would lead to uncertainty about to obtained
results.
Nitrosomonas, can however (as the only bacteria in the biofilm) use CO, from
the atmosphere to synthetize biomass (La Cour Jansen and Henze 1990).
Three times, during the 36 days of the test period, a shock-load of organic
compound, in the form of ethanol, was added. On each occasion it resulted in a
sudden development of the SND process.
Figure 5.22 shows the amount of ammonium-N nitrified, the amounts of
produced nitrite-N and nitrate-N. Only between 25 and 30 percent of the loaded
ammonium-N is nitrified. This low yield of nitrification is presumably due to the following
two factors.
1) At low pH it is difficult to obtain sufficient biomass concentration to convert the
amount of applied ammonium-N.
2) Because of the high nitrite-N and nitrate-N concentrations, the nitrification process
can be inhibited by its own products.
At day 13 and 22 respectively 50 mg/l and 100 mg/l of COD (organic
compound) were added to the waste water. At days 33, 34 and 35 1000 mg/l of COD
were added.
Figure 5.22 indicates that the concentration of nitrite-N produced during the
test period, was subject to great fluctuation. At the two first COD additions the
concentration of nitrite-N increased and, therefore, at least some of the added COD
were used to produce Nitrosomonas biomass.
The amount of produced nitrate-N increased when the nitrite concentration had
reached its peak-value. This is natural because Nitrobacter (NO, conversion to NO,)
is not developed until nitrite has been produced. Nitrite, however, both inhibits the
nitrification and acts as a substrate for nitrobacter. The second step of the nitrification
228
At day 13 and 22 respectively 50 mg/l and 100 mg/l of COD (organic
compound) were added to the waste water. At days 33, 34 and 35 1000 mg/l of COD
were added.
Figure 5.22 indicates that the concentration of nitrite-N produced during the
test period, was subject to great fluctuation. At the two first COD additions the
concentration of nitrite-N increased and, therefore, at least some of the added COD
were used to produce Nitrosemonas biomass.
The amount of produced nitrate-N increased when the nitrite concentration had
reached its peak-value. This is natural because Nitrobacter (NO; conversion to NO,)
is not developed until nitrite has been produced. Nitrite, however, both inhibits the
nitrification and acts as a substrate for nitrobacter. The second step of the nitrification
process (see Chapter 3) is therefore difficult to initiate.
A change in the biomass concentrations of both nitrosomonas and nitrobacter
is therefore observed during the period of nitrate production. If no nitrite is produced,
then nitrobacter is not developed due to a lack of the substrate that nitrobacter uses.
On the other hand if the nitrite concentration is low, compared to the nitrate con-
centration, it was observed that both nitrosomonas and nitrobacter occurred in great
amounts.
The applied shock-loads of COD seem, therefore, to have three important
concequences in this investigation:
1) Maintenance of a fast formation of SND, during about 1 day.
2) Initiation of the development of Nitmomonas.
3) Offering a carbon source for the synthesis of nitrobacter as soon as nitrite was
available as substrate.
On days 34, 35 and 36 of the experiments, higher amounts of COD were
added and a more persistent SND was introduced as during RUN 1. Both the amount
of nitrite and nitrate therefore declined rapidly because there was a sufficient carbon
source for the denitrification process.
Kinetics A comparison of the nitrogen removal rate for the following submerged filters;
the Biocarbone, Biofor and the SND processes, are outlined in Table 5.18 The kinetic
rate of the SND process using clinoptilolite as matrix, was about three times higher
than for the Biocarbone and Biofor processes, expressed as kg N / (m3 matrix day).
229
200
150
100
5 0
4 0
COD addition
5
COD addition addition
Figure 5.22 Results obtained during Run 3.
Table 5.18 The nitrification rate for the three submerged processes.
Process Maximum Nitrification rate Reference
kg /Wm3 matrix * day
Biocarbone (OTV) 0.74"
Biofor (Degremont) 0.75"
SND' 1.7-3.4
Rogella et a/, (1 990)
Paffoni et a/. (1990)
Halling-Starensen and
Hjuler (1992; 1993)
* Only laboratory experiments
** Maximum loading 1,OO kg N / m3 matrix per day.
230
Pilot-plant experiments In Vantinge on the island of Fyen in Denmark, a pilot-plant has been built
following the same concept as presented for experimental RUN 1. The only difference
is that the carbon source is not ethanol, but endogeneous carbon from the waste
water. Figure 5.23 shows a photo of the pilot-plant. The pilot-plant consists of 80 m3
bedvolume of clinoptilolite distributed in six connected concrete bassins, with upflow
waste water and air distribution.
The plant is used as the tertiary treatment stage for removal of nitrogen from
slaughterhouse waste water. As secondary treatment stage, an activated sludge
process unit for combined carbon oxidation and nitrification is used. After a secondary
clarifier the waste water is pumped into the SND pilot-plant.
The total SND obtained during the first months of pilot-plant experiments were
of the order of 0.45 kg N /m3 bedvolume day and 1.0 kg N /m3 bedvolume day.
Table 5.19 show the different influent and effluent concentrations found at the
pilot-plant.
Table 5.19 Influent and effluent concentration of important parameters at the SND
pilot-plant.
Parameter NH4'-N NO,-N COD
m g/l m g/l mg/l
Influent activated a20
sludge treatment
Influent SND 450
tertiary treatment
step.
Effluent SND 30
teriary treatment
step.
30
10
3
5500
1200
50
231
Figure 5.23 Photo of the SND pilot-plant at Vantinge on the island of Fyen.
Figure 5.24 show a cross-section of a clinoptilolite stone. The porosity of the
clay stone makes it possible to obtain aerobic and anaerobic conditions simultaneously.
On the surface of the clinoptilolite stone oxygen diffuses into the biofilm and is used
for the nitrification process. Ammonium is also diffusing towards the biofilm on the
clinoptilolite stone. The ion-exchange ability of the stones binds ammonium on the
surface (Jsrgensen 1976; Haralambous eta/, 1992) and nitrifying bacteria converts it
to nitrate. The ion-exchange mechanisms may also play an role in the mechanisms,
but is not totally clear.
The concentration of nitrate is highest at the upper layer of the biofilm which
is most aerobic. Nitrate diffuses to the more anoxic areas in the lower part of the
biofilm, where it is denitrified. Because of the concentration gradient a continuous
diffusion to the center of the stone will take place.
Figure 5.25 is a micro-scope photo of the clinoptilolite stone covered by an
SND bio-film.
232
Air
-
-
7p ;) I
Bulk water Liquid I film
) A
I 02 I
I I I
N HL+ I
N 0,- I
1
I I I
I I
Organics I I I I I
co2 I
I I I
NZ I
Aerobic Anoxic
Figure 5.24 Cross-section of a clinoptilolite stone with aerobic and anoxic biofilm.
Figure 5.25 Clinoptilolite stone covered by SND bio-film, as seen under a micro scope.
233
Conclusions
nitrification and denitrification (SND) on basis of the experiments described:
1) Nitrification and denitrification occur simultaneously with different loadings.
2) The results show that a higher flow through the reactor permits greater daily loading
with the same removal efficiency.
3) If no carbon source was added to the influent, nitrate and nitrite was recovered,
showing that only nitrification occurs.
4) The SND is not able to treat organically bounded nitrogen.
5) The SND is relatively easy to start and fairly trouble free to maintain. In general the
response of the reactor to changes is immediate and steady state conditions were ap-
parently achieved quickly.
6) The faster nitrogen removal for this process, compared with suspended cultures, is
partly due to a higher concentration of microorganisms, but it must be anticipated that
some "additional effect" (i.e. ion-exchange) is needed to explain the high removal rate.
7) The pilot-plant experiments show an SND removal of 0.45 to 1.0 kg N /m3 bed
volume * day, while laboratory columns have shown up to 4 to 5 times higher
efficiency.
8) This study has shown that a simultaneous nitrification and denitrification is a
technologically feasible process for nitrogen control.
The following conclusions can be made concerning the simultaneous
234
6 SUSPENDED-CULTURE REACTORS
6.1 Activated Sludge Unit Processes The activated-sludge process is based upon a suspended-culture system that
has been in use since the beginning of the century. The most common arrangements
for nitrogen removal are the single-stage carbon oxidation and nitrification systems and
the separate stage nitrification system.
The activated-sludge process can be designed with or without recycling of
sludge, and may involve either a completely mixed or a plug-flow process (fig. 6.1).
Other possibilities are the aerated lagoons, contact stabilization and extended
aerations. Many different applications of the activated sludge process are used. Most
of these are presented in Section 6.3.
The return of sludge, containing living or active organisms, is conducted to
increase the available biomass and accelerate the reactions.
Most activated sludge applications are used for oxidation of organic content in
the waste water, but also nitrogen conversion is to some extent possible with a suitable
sludge age of 9-10 days (see Fig. 6.2). The sludge age is important because an
appropriate sludge age makes possible the development of nitrifying bacteria in the
flocs. These flocs will thereby be able, under suitable conditions, to convert ammonium
to nitrogen gas.
The activated-sludge process is normally used for secondary treatment of large
amounts of municipal wastewater, where only little nitrification can be expected.
EPA (1975) indicated that the organic loading should be below 0.16 kg BOD/ m3 day
if nitrification is to be possible simultaneously with the carbon oxidation, due to the
bacterial composition. The performance of nitrification in an activated sludge treatment
plant, is used mostly to treat large quantities of municipal waste water.
In the activated sludge process there are two main biological activities whereby
nitrogen is removed from the waste water:
1) The sludge production: only a minor fraction of nitrogen can be removed by sludge
production.
2) Nitrification and denitrification depending of the oxic conditions.
235
The EPA (1 975) manual gives the following classification between the
combined carbon oxidation and nitrification process and the separate stage nitrification
process. The ability of various activated sludge processes to nitrify has been correlated
to the BOD5TTKN ratio. TKN is the totalkjeldahlnitrogen, which is the organic nitrogen
plus the ammonia nitrogen. For BOD,TTKN ratios between 1 and 3, which roughly
correspond to the values encountered in separate-stage nitrification systems, the
fraction of nitrifying organisms is estimated to vary from 0,21 at a BOD5/TKN ratio of
1 to 0.083 at a ratio of 3. In most conventional activated-sludge processes, the fraction
of nitrifying organisms would therefore be considered less than the 0.083 value. The
EPA (1975) manual indicates that when the BOD5TTKN ratio is greater than 5 the
process can be classified as a combined carbon oxidation and nitrification process,
and, when the ratio is less than 3, it can be classified as a separate-stage nitrification
process (see Table 6.1).
* a Primary
Mixed Secondary reactor clarifier Effluent
-
* secondary * Effluent Plug Flow
b Primary
effluent . Sludge return Sludge underflow
1 L 11
4
Sludge waste
1L Sludge return
Figure 6.1 Diagram of a) completely mixed activated sludge process. b) plug-flow
process.
236
Table 6.1 Relationship between the fraction of nitrifying organisms and the BOD5/TKN
ratio.
BOD,/ TKN ratio N itrifier fraction
03
1
2
3
4
5
6
7
8
9
0,35
0,21
0,12
0,083
0,064
0,054
0,043
0,037
0,033
0,029
Source: EPA (1 975)
6.2 Process Design Several design variations of the completely mixed and plug-flow systems are
used. Some involve minor modifications, such as application of air or waste water, or
different retention times, or reactor shapes. Others involve more drastic differences,
such as sorption and settling prior to the biological processes and the use of pure
oxygen rather than air.
The most commonlyapplied of these design variations are described in Section
6.3. The two main types are the plug-flow and the completely mixed reactors as shown
in Fig. 6.1.
In the following discussion attention is focused on some of the factors affecting
the activated sludge process, i.e the loading criteria, the sludge production, the air
diffusion, control of filamentous organisms and the control of sludge recycling.
Loading criteria. Many parameters have been proposed for the design and control of the
237
activated-sludge process. The two most commonly used parameters are:
1) The food-to-microorganism ratio (F/M).
2) The mean cell-residence time Qc. (sometimes called the Solids Retention time, SRT)
The food-to-microorganism ratio is defined as:
where:
F/M = the food-to-microorganism ratio, d-’.
So = the influent substrate concentration in mg/l (g/m3).
Q = the mean cell-residence time of the aeration tank, day. V = the aeration tank volume.
Q = the influent waste water flow rate, m3/d.
X = the concentration of volatile suspended solids in the aeration tank, mg/l
The relationship between the food-to-microorganism ratio and the specific utilization
rate U is:
(g/m3).
E U = ( F / M ) * - loo
where E = the process efficiency in %.
Substituting the first equation for the food-to-microorganism ratio and [(So - S)/So] for
the efficiency yields the following term:
238
where S = the effluent substrate concentration in mg/l (g/m3).
following relationship, defined on the aeration tank volume:
The mean cell residence time (sludge age) $c can be defined from the
If the definition is based on the total volume of the system, then the mean cell-
residence time $& can be expressed by the following relationship.
where:
$C
$ct v X
*inn.
Xinfl.
Qeff.
Xeff.
= mean cell-residence time based on the aeration tank volume, d.
= mean cell-residence time based on the total system, d.
= aeration tank volume.
= concentration of volatile suspended solids in the aeration tank, mg/l.
= waste sludge flowrate, m3/d.
= concentration of volatile suspended solids in the waste sludge, mg/l (g/m3)
= treated effluent flow rate, m3/d.
= concentration of volatile suspended solids in the treated effluent, mg/l 0
It is recommended that the design of the reactor is based on $c, because
Comparing these parameters, the specific utilization rate, U, can be considered
substantially all of the substrate conversion occurs in the aeration tank.
239
as a measure of the rate at which substrate (nitrogen) is utilized by a unit mass of
organisms, and 9, can be considered as a measure of the average residence time of
the organisms in the system.
The relationship between mean cell-residence time, I$~, the food-to-microorga-
nism ratio F/M, and the specific utilization rate U is:
F E 1 pc M 100
- Y * - * - - kd= YU- kd _ -
where:
Y = the cell yield coefficient.
E = the process efficiency, %.
k, = the endogenous decay coefficient, time-'
It has been found that a mean cell-residence time of more than 9-10 days
results in the production of a stable nitrifying sludge with good settling characteristics.
Sludge production.
It is important to know the quantity of sludge produced per day because it will
affect the design of sludge-handling and disposal facilities necessary for the excess
sludge.
The relationship between the mean cell-residence time (sludge age) and the
nitrification efficiency in per cent, in the activated sludge is presented in Fig. 6.2.
The quantity of sludge produced daily can be estimated from the following:
240
C
2 4 6 8 10 12 14
Figure 6.2 Relationship between the mean cell-recidence time (sludge age) and the
nitrification efficiency in per cent, in an activated sludge (Source: Jargensen 1989).
where:
P = the net waste activated sludge produced each day, measured in VSS, kg/d.
Oxygen requirements for a nitrifying activated sludge plant.
When nitrification has to be considered, the total oxygen requirements can be
found from the following equation.
where:
No = the influent total nitrogen-N in mgA (g/m3).
N = the effluent total nitrogen-N in mg/l (g/m3).
For the activated-sludge process the oxygen utilization rate will always exceed
the rate of natural replenishment. Thus, some artificial means of adding oxygen must
be used. Oxygen is normally supplied by aerating the waste water in the biological
reactor.
The oxygen utilization rate (oxygen consumed by the microorganisms) is a
function of the characteristics of both the waste water and the reactor.
Treatment of ordinary municipal waste water by extended aeration usually
results in an oxygen utilization rate of approximately 10 mg/l * hours. Treatment of the
same waste water by a conventional activated sludge process results in an oxygen
utilization rate of about 30 mg/l hours and up to 100 mg/l hours. The oxygen
addition should be sufficient to match the oxygen utilization rate and still maintain a
small excess in the waste water at all times to ensure aerobic metabolism.
Aeration techniques consist of using air diffusers to inject compressed air into
the biological reactor and/or using mechanical mixers to stir the contents violently
enough to entrain and distribute air through the liquid. It is common practice to use
diffused air in plug-flow systems and mechanical aerators in completely mixed
systems.
Control of filamentous organisms.
The growth of filamentous microorgansims is the most common operational
problem in the activated sludge process. Filamentous organisms in the system result
in poorly settling sludge usually termed "bulking sludge".
In the single-stage activated sludge system it is normal to see a growth of
filamentous organisms because of the low-substrate levels uniformly present in the
reactor.
In some plug-flow reactors, where significant back-mixing occurs, a similar
phenomenon takes place.
When oxygen limits the growth of microorganisms, filamentous organisms may
predominate. In practice the dissolved-oxygen concentration in the aeration tank should
242
be maintained at about 1.5-4 mg/l in all regions of the aeration tank.
Recent research has shown that prevention and control of filamentous
organisms growth can be obtained by using a separate compartment or "selector" as
the initial contact zone, between microorganisms and waste water, in a biological
reactor. In the selector the primary effluent and return activated sludge are combined,
so that the biomass concentration is increased in the initial treatment of the waste
water and therefore the reaction rate of the removal of nitrogen is increased. A selector
can be used in most types of activated sludge.
Return activated-sludge control.
The purpose of the return of activated sludge is to maintain sufficient
concentration of activated sludge in the aeration tank so that the required degree of
treatment can be obtained in the time interval desired.
The return of activated sludge from the final clarifier to the inlet of the aeration
tank is the essential feature of the process.
Sludge production
The excess activated sludge produced each day must be wasted to maintain
a given food-to-microorganism ratio or mean cell residence time. The most common
practice is to waste sludge from the return sludge line because it is more concentrated
and requires smaller waste sludge pumps. The waste sludge is discharged to the
primary tanks, to thickening tanks, or to other sludge-thickening facilities.
Operational problems.
The most common problems encountered in the operation of an activated-
sludge plant are bulking sludge, rising sludge or Nocardia foam.
A bulking sludge is one that has poor settling characteristics and compac-
tability. Two principal types of sludge-bulking problems have been identified. One is
caused by the growth of filamentous organisms or organisms that can grow in a
filamentous form under adverse conditions. The other is caused by bound water, in
which the bacterial cells composing the floc swell through the addition of water to the
extent that their density is reduced and they will not settle.
The main waste water characteristics that can affect sludge bulking includes
fluctuations in flow and strength; pH, temperature, nutrient content, and the nature of
243
the waste components (Eddy and Metcalf 1991). But some design limitations, including
air supply capacity, clarifier design, return sludge-pumping capacity limitations, and
poor mixing of the waste water are also factors that can affect sludge bulking.
Filamentous bulking can also be due to operational causes which include low
dissolved oxygen in the aeration tank, insufficient nutrients, widely varying organic
waste loading, or a low F/M ratio.
More than 20 different types of filamentous organims have been found in
activated sludge plants (Eddy and Metcalf 1991).
In an emergency situation or while the factors provoking bulking are being
investigated, chlorine and hydrogen peroxide may be used to provide temporary help,
but chlorination of a nitrifying sludge will produce a turbid effluent due to dead nitrifying
organisms.
Occasionally sludge that has a good settling characteristics will be observed
to rise or float to the surface after a relatively short settling period. The cause of this
phenomenon is denitrification in which the nitrites and nitrates are converted to
nitrogen gas. Rising sludge can be differentiated from bulking sludge by noting the
presence of small gas bubbles attached to the floating solids.
Rising sludge problems may be overcome by increasing the return activated-
sludge withdrawal rate from the clarifier, to reduce the detention time of the sludge in
the clarifier, or by decreasing the rate of flow of the aeration tank, or by decreasing the
mean cell-residence time (solids retention time) by increasing the size of the sludge-
wasting tank.
The last operational problem to be discussed is the viscous brown foam, that
can cover the aeration basins and secondary clarifiers. This foam has led to many
problems in activated-sludge plants. The foam is associated with a slowgrowing
filamentous organisms of the Nocardia genus.
Reducing the sludge age is the method that has been used most commonly
for Nocardia control, but this prevents nitrification occurring in the plant.
Air diffusers.
Two main type of diffusers exist. Fine-bubble diffusers produce many bubbles
of approximately 2,O to 2,5 mm in diameter, while coarse-bubble diffusers inject fewer
bubbles of a larger (up to 25 mm diameter) size. Both types have advantages and
disadvantages. With respect to oxygen transfer, the fine-bubble diffuser is more
244
efficient because of the larger surface area per volume of air. However, head loss
through the small pores necessitates greater compression of the air and thus greater
energy requirements, and compressed air must be filtered to remove all particulates
that would plug the tiny diffuser openings.
Coarse-bubble diffusers offer less maintenance and lower head loss, but
poorer oxygen transfer efficiencies. A compromise is to locate a mechanical turbine
just above a coarse-bubble diffuser so that the shearing action of the blade at high
rotational speed breaks the large bubbles into smaller ones and disperses them
through the waste water.
Mechanical aerators. Mechanical aerators produce turbulence at the air-water interface, and this
turbulence entrains air into the liquid. Mechanical aerators may have high-speed
impellers that add large quantities of air to relatively small quantities of water.
This aerated water is then mixed with the reactor contents through velocity gradients.
Large impellers driven at slow speed agitate larger quantities of water less violently.
Use of smaller, high speed units is common in extended aeration systems,
while the slow-speed units are more common in conventional activated sludge
systems. Brush-type aerators are used to provide both aeration and momentum to
waste water in the oxidation-ditch variation of the activated sludge process.
6.3 Activated-sludge Process Configurations Two basic activated sludge process configurations have been developed for
single sludge biological nitrification and denitrification. Depending of the anoxic
conditions throughout the plant, more or less denitrification is achieved. The two
arrangements are:
1) The Wuhrmann configuration.
2) The Ludzack-Ettinger configuration.
Both can undergo completely mixed and plug-flow regimes for the respective reactors.
These two configurations are explained in detail below.
245
The Wuhrmann configuration.
The single sludge nitrification-denitrification system in which endogenous
energy release provides the energy source for denitrification was first proposed by
Wuhrmann (1964).
It consists (Fig. 6.3) of two reactors in series, the first aerobic and the second
anoxic. The influent is discharged to the first reactor where aerobic growth of both the
heterotrophic and nitrifying organisms takes place. Provided the sludge age is
sufficiently great and the aerobic fraction of the system is adequately large, nitrification
will be complete in the first reactor. In the second anoxic reactor, the denitrification
takes place. The overflow from the anoxic reactor passes through a settling tank and
the underflow is recycled back to the aerobic reactor. The energy source for the
denitrification process is provided by energy release by the sludge mass due to the
death of organisms. However, the rate of release of energy is low, which implies the
rate of denitrification is low too. Consequently, in order to obtain sufficient denitrifica-
tion, the anoxic fraction of the plant must be large compared with the oxic fraction. This
may cause a breakdown of the nitrification process.
It is usually not possible to remove all the nitrate, particularly if the tempera-
tures are low, below 15°C. Furthermore, in the anoxic reactor, organic nitrogen and
ammonia are released due to dead organisms, some of this combined nitrogen passes
out with the effluent thereby reducing the total nitrogen removal of the system. To
minimize the ammonium content of the effluent, a flash or reaeration reactor may be
placed between the anoxic reactor and the settling tank. In this reactor the ammonium
is then nitrified to nitrate.
Waste flow
Settler
* Effluent
Anoxic reactor
Y Sludge recycle s
Figure 6.3 The Wuhrmann process for the removal of nitrogen.
246
The Ludzack- Ettinger configuration.
This configuration was first proposed in 1962 by Ludzack and Ettinger (Fig
6.4). It is a single sludge nitrification and denitrification process utilizing the bi-
odegradable material in the influent as an energy source for the denitrification process.
It consists of two reactors, only partially separated, in series. The first reactor
is maintained in an anoxic state by stirring without aeration. The second reactor is
aerated and nitrification takes place. As there is only partial separation between the
two reactors a mixing of the nitrified and anoxic waste water is induced, and the nitrate
entering the anoxic reactor is reduced to nitrogen gas. With this type of configuration
a varying denitrification result is obtained, probably due to the lack of control of the
exchange of waste water between the two reactors.
Anoxic reactor
lnfl uent
Sludge recycle s
Figure 6.4 The Ludzack-Ettinger configuration for nitrogen removal.
Since the beginning of the 1960's many improvements of the above two types
of plants for nitrogen removal activated sludge have been proposed.
Some of the most popular are the modified Ludzack-Ettinger process and the
Bardenpho process.
The modified Ludzack-Ettinger configuration (Fig 6.5) completely separates the
anoxic and aerobic reactors, recycling the underflow from the settler to the anoxic
reactor, and providing an additional recycle from the aerobic to the anoxic reactor.
These modifications offer a significant improvement in control over the process
performance. The high influent energy source discharged to the anoxic reactor, also
called the pre-denitrification reactor or primary anoxic reactor, yields a high rate of
247
denitrification. But complete denitrification cannot be achieved because a part of the
total from the aerobic reactor is not recycled to the anoxic reactor but is discharged
directly with the effluent.
Waste flow
Settler
Sludge recycle s
Figure 6.5 The modified Ludzack-Ettinger process.
The Bardenpho configuration (Fig. 6.6) is intended to overcome the incomplete
denitrification. The low concentration of nitrate discharged from the aerobic reactor to
the secondary anoxic reactor will be denitrified to produce a effluent free of nitrate. To
strip the nitrogen bubbles generated in the secondary anoxic reactor attached to the
sludge flocs, a flash aeration is introduced between the secondary anoxic reactor and
the final settling tank.
The flash aeration is also considered necessary to nitrify the ammonia released
during the sludge residence time in the secondary anoxic reactor. In order to reduce
the possibility of flotation of sludge in the settler due to denitrification of residual nitrate,
the sludge accumulation in the settler is kept to a minimum. This is achieved by a very
high recycle rate from the settler, approximately equal to the mean influent flow.
Aerated lagoons, contact stabilization and extended aeration
These three processes cover the extremes in operation between zero and
complete nitrification, by aerated lagoons and extended aeration respectively, with
contact stabilization typically achieving an intermediate degree of nitrification (Gujer
and Jenkins 1974). Aerated lagoons operate essentially as completely mixed, no-
recycle systems, which are distinguished by the fact that their hydraulic retention time
248
Secondary anoxic
t
Sludge recycle s
Figure 6.6 The Bardenpho process.
and mean cell residence times are equal. Such systems commonly have mean cell
residence values of 1 to 5 days and may achieve nitrification at higher values under
appropriate conditions, such as during summer. It is unlikely that aerated lagoons
would be used where nitrification is required at low temperatures because of the large
reactor volume required. One advantage of these lagoons, where they are designed
to nitrify, is that their large volume serves to dilute the incoming waste water, thus
reducing the impact of shock loads on nitrifier growth rate. With the exception of this
reduced impact of transient loads, the design relationships developed for the complete
mixed activated sludge process are directly applicable to the aerated lagoons.
Extended aeration operates at very high mean cell residence values and low
organic loading rates such that nitrification is assured under all conditions. Contact
stabilization differs from the flow sheet of the other processes in that it consists of two
aeration stages. The first is a contact tank at short detention times of 2 to 3 hours,
after which the sludge is separated from the effluent and returned to a second aeration
tank (stabilization tank) with 4 to 6 hours of detention time. The short detention time
in the contact tank limits the nitrification performance of this system (Gujer and Jenkins
1 974).
249
Compressed air Secondary
Primary effluent
&actor
(a) sludge return Sludge waste
Compreased air Secondary clarifier
primary effluent
Reactor
L -------- L,,, Sludge return Sludge waste
Secondary clarifier
Effluent
Influent
( 0 ) Compressed air
Pure oxygen Oxygen return Waste gas
Secondary clarifier
Primary effluent q+-mffluent
Reactor
L-,-------- -+---- sludge return sludge waste
Brush-type aerator
secondary ( 0 ) clarifier
250
Primary effluent
I
secondary clarifier
I
Primary effluent
L,,, -- - - - - - -L-,,, sludge return (omitted in some systems) sludge waste
c D
Figure 6.7. Overview of common applications of the activated-sludge process. (a) step
aeration; influent addition: influent addition at intermidate points provides more uniform
removal throughout the tank. (b) Tapered aeration: air added in proportion to nutrient
exerted. (c) Contact stabilization: biomass adsorbs organics in contact basin and
settles out in secondary clarifier; the thickened sludge is aerated before being returned
to the contact basin. (d) Pure-oxygen activated sludge: oxygen added under presurre
keeps dissolved oxygen level high. (e) Oxidation ditch, plan view. (f) High rate: short
detention time and high food/mass ratio in aerator to maintain culture in log-growth
phase. (9) Extended aeration: long detention time and low food/mass ratio to maintain
culture in endogeneos phase.
25 1
This limited efficiency makes contact stabilization less attractive as a design
alternative for nitrification.
6.4 The Kinetics of the Activated Sludge Process The kinetics of the nitrification process are well-defined for the suspended-
growth systems. From experience, it has been found that the following factors have a
significant effect on the kinetics of the nitrification process.
1) Ammonia and nitrite concentration, 2) COD/total N ratio, 3) Dissolved-oxygen
concentration, 4) Temperature and 5) pH.
The impact of these variables on the nitrification and denitrification processes
and the approach developed to account for them are reported in Chapters 3 and 4.
Table 6.2 shows typical kinetic coefficients for the activated-sludge nitrification process.
The kinetic expression used for analysis of suspended-growth nitrification and
denitrification are summarized in Table 6.3.
6.5 Modification of Activated Sludge Plants for Biological Nitrogen
Removal Today's high standards for nitrogen removal from waste water often demand
modification of existing plants. The approache necessary to convert an existing waste
water treatment plant to a biological nitrogen removal plant is dependent on the site
conditions and on the level of treatment required.
For existing systems that accomplish only removal of organic material, a higher
solid retention time will have to be provided for nitrification to occur. This can be done
by increasing the size of the aeration tank and/or the sludge concentration. This will
need a greater quantity of oxygen.
If the system is already designed for nitrification, additional volume may be
required to provide anoxic zones for denitrification. The anoxic volume in an activated
sludge nitrification-denitrification system may account for 20 to 40% of the total tank
volume. If denitrification is required the oxygen supply must be reduced.
A number of activated-sludge designs have been developed for the combined
removal of nitrogen and phosphorus. Some of these processes were developed
originally for phosphorus removal and later developed into combined phosphorus and
nitrogen removal systems.
252
Table 6.2 Typical coefficient for the different parameters in the nitrifying activated sludge process.
Coefficient Unit Value
Range Typical
Reported at 20 C
Nitrosomonas
Pm
K S
Nitrobacter
hnax
K S
Overall
Pmax
K S
Y
Kd
d-'
NH,+-N mg/l
d-'
NN,'-N mg/l
d-'
NH,+-N mg/l
NH,+-N mg VSS/mg
d-
0,3 - 2,O 0,2 - 2,o
0,4 - 3,O 0,2 - 5,O
0,3 - 3,O 0,2 - 5,O 0,l - 0,3 0,03 - 0,06
0,7
0,6
After: Schroeder (1976); EPA (1975) and Eddy and Metcalf (1991).
The most commonly used processes for combined nitrogen and phosphorus
removal are: 1) the A2/0 process (Hong et a/.1984), 2) the five-stage Bardenpho
process, 3) the UCT process and 4) the VIP process. They are all described in Metcalf
and Eddy (1991). Stensel et a/. showed in Table 6.4 the nitrification rate obtained,
based on both total MLVSS and on calculated Nitrosomonas biomass for the biological
nutrient removal (BNR) and the conventional activated sludge process.
253
Table 6.3 Summary of kinetic expressions used for the analysis of activated-sludge nitrification and denitrification. See also Chapters 3 and 4.
Equation Definition of terms
p = specific growth rate, time” S
Ir = Ir,- Ks+S
_ - ’ - W - k d 4 2
ds/dt = substrate utilzation rate, mass/unit volume.
S = concentration of growth limiting substrate in solution, masdunit volume.
Y = maximum yield coefficient, mass of cell formed per mass of substrate consumed.
K, = maximum rate of substrate utilization.
k = maximum rate of substrate utilizaion.
4 = hydraulic detention time, time.
Q c = d e s i g n m e a n c e l l - residence time, time.
U = substrate utilization rate, time-’.
@cm = minimum mean cell-residence time.
SF = safety factor
So = influent substrate concentration masdunit volume.
X = conc. of microorganisms.
254
Table 6.4 Summary of specific Nitrification Rates and Ammonia, Oxidation Rates in tbe biological nutrient removal process (BNR) and
the conventional activated sludge process.
System SRT Aerobic SRT T Total NH3-N Aerobic Specific Nitrosomonas Ammonia
d d "C Oxidized MLVSS Nitrification vss CMHmIae
mg/1 mg/1 Rate mg/l mgN/mg
mgN/gMLVSS/h Nosannas d
BNR 15
5
2.7
1.5
Conventional 15
15
5
2.7
8.3
2.7
1.5
0.9
15
15
5
2.7
20
20
20
20
20
15
20
20
18.8
23.2
21.6
12.3
21.2
26.6
26.5
27.1
2636
1014
749
446
1348
2177
1284
658
1.783
5.720
7.210
6.895
1.986
1.527
2.580
5.148
122
74
42
14
101
143
72
47
0.834
1.729
2.695
4.382
0.631
0.560
1.107
1.716
Source: Stensel et a/. (1 992)
6.6 Modelling the Activated Sludge Process A mathematical model, Activated Sludge Model No. 1, for the removal of
carbonaceous biodegradable material, nitrification and denitrification was developed
by the IAWPRC Task Group (Henze eta/. 1987) and modified by Wentzel eta/. (1991)
and Dold (1991).
A total of ten dissolved and seven particulate components are used to
characterize the wastewater and the activated sludge. These include:
1) Dissolved oxygen, bicarbonate alkalinity, and soluble phosphorus.
2) Three forms of biomass (Heterotrophs and two types of autotrophs,
all represented in terms of COD)
3) Five forms of nitrogen (particulate and soluble biodegradable organic
nitrogen, ammonia, nitrite and nitrate).
4) Six forms of COD (inert soluble and particulate in feed, two forms of
biodegradable soluble, enmeshed slowly degradable particulate, and inert
particulate COD from endogenous decay).
For a detailed overwiev of the formula matrix the authors recommend
consulting the Activated Sludge Model No 1. (Henze et a/. 1987), because the most
recent attempts at modelling the activated sludge have been made with this model.
6.7 Advantages and Disadvantages of the Separate and Combined
Activated Sludge Process The following gives an overview of some of the advantages and disadvantages
of the activated sludge process, both as A) a separate stage process and B) as a
combined stage process.
A) Separate stage activated sludge process for nitrification.
Advantages:
1) Good protection against most toxicants.
2) Stable operation.
3) Low effluent ammonia concentration possible.
256
Disadvantages:
1) Sludge inventory requires careful control when BODS/TKN ratio is low.
2) Stability of operation linked to operation of secondary ctarifier for biomass return.
3) Greater number of unit processes required than for the combined oxidation and
nitrification unit.
B) Combined carbon oxidation and nitrification activated sludge process for nitrifica-
tion.
Advantages:
1) Combined treatment of carbon and ammonia in a single stage.
2) Low effluent ammonia is possible.
3) Inventory control of mixed-liquor sample due to high BOD5TTKN ratio.
Disadvantages:
1) No protection against toxicants.
2) Only moderate stability of operation.
3) Stability linked to operation of secondary clarifier for biomass return.
4) Large reactors required in cold weather.
257
This Page Intentionally Left Blank
Part C
PHYSICO= CHEMICAL PROCESSES
Air Stripping Breakpoint Chlorination Ion Exchange Membrane Processes Precipitation
This Page Intentionally Left Blank
7. AIR STRIPPING
b Influent, waste water with higt- PH.
7.1 Physico-chemical Principles of Air Stripping
Stripping unit for instance a packed tower
A -
The stripping process is used to remove volatile gases, such as hydrogen
sulfide, hydrogen cyanide and ammonia by blowing air through the waste water. The process is therefore to be considered as a transfer from a liquid phase to a gas phase. The basic principle of this process of nitrogen removal is illustrated in Fig.
7.1.
Air + ammonia out t
The rate at which ammonia can be removed by air stripping is highly
dependent on pH, because the exchange between the two forms, ammonium,
which is the ionic form, and ammonia, which is a highly water-soluble gas, is an acid-base reaction. The ammonia stripping is based on the following reaction:
261
The equilibrium constant for this process is 10-9.25at 18”C, which means that:
By separating H+ in this equation and converting to a logarithmic form, we get:
(7.3)
Knowing the ammonium concentration in an aquatic ecosystem, this
relationship can be used to estimate the toxicity level of the water, see Section 1.4. From equation (7.3) we can see that at pH = 9.25, 50% of the total ammonia-
nitrogen is in the form of ammonia and 50% in the form of ammonium.
Correspondingly the ratio between ammonia and ammonium is 10 at pH 10.25 and
100 at pH 11.25. A graph showing the ratio ammonia to ammonium is given in Fig.
7.2. Consequently it is necessary to adjust the pH to 10 or more before the stripping
process is used. The pK,value, which is the negative logarithm to the equilibrium constant, is dependent on the presence of other ions, or expressed in another way, of the ionic strength of the influent. The ionic strength is defined by the following
expression:
I = 11/2 c z 2 (7.4)
where C =the molar concentration of the considered ions and Z = the charge.
f, from:
On the basis of the ionic strength, it is possible to find the activity coefficient,
0.5 2 2 dl -log f =
d + 1 (7.5)
262
where I = ionic strength, Z = charge and f = activity coefficient. The activity
coefficient, f, is defined as the activity a, divided by the concentration c. The activity is used in the mass equations to replace the concentrations, if the ionic strength is sufficient high to play a significant role, see also below.
99.99
99.9 99.8
99.5
- - - - - -
- -
- - - -
.- 5 40.0 ////////.I m .- s E E
40.0 al m
2
4-
C al
l? 10.0
5.0
1.0
0.1
E l - m
4- H C
2 2
80.0
90.0
95.0
98.0 99.0 99.5
99.8 99.9
aa 00 11.11
6 7 8 9 10 11 12 0.01
Fig. 7.2. Distribution of ammonia and ammonium as function of pH and
temperature.
If the ionic strength plays a role, the concentrations in equation (7.2) are
replaced by activities. As pH is defined from the activity of hydrogen ions, (7.2) will
be changed to the following expression in this case:
[NHs] *aH+
[NHs+]
- - 10-9.25* f
263
Equation (7.3) will be changed correspondingly:
or
As seen from equation (7.8) the ratio ammonia I ammonium is disfavored by increased ionic strength, implying that a higher pH is need to obtain the same stripping effect at higher ionic strength.
Table 7.1 gives the activity coefficients for different ionic charges, calculated
from the equation (7.5).
TABLE 7.1
Activity coefficient f at different ionic strengths
dl f for f for f for I 1 +d z = 1 z = 2 z = 3
0
0.001
0.005
0.01
0.02
0.05
0.1
0.2
0.5
0
0.03
0.07
0.09
0.12
0.18
0.24
0.31
0.41
1 .oo 0.95
0.93
0.90
0.87
0.81
0.76
0.70
0.62
1 .oo 0.82
0.74
0.66
0.57
0.43
0.33
1 .oo 0.64
0.51
0.40
0.28
0.15
0.10
I = ionic strength, Z = charge, f = activity coefficient
Since calcium hydroxide is the cheapest source of hydroxide ions, it is most
264
often used for adjustment of pH before the stripping process. The addition of calcium hydroxide leads to an increased ionic strength. However, the ionic strength
of most waste waters, after addition of sufficient calcium hydroxide to obtain a pH of 10 or above, is only in the order of 0.05-0.1, which implies that the increase of pH
needed to obtain the same stripping effect as for distilled water is approximately
only 0.1.
7.2 Process Variables
As much as 13 g ammonia gas is soluble at room temperature in 100 ml
water. Due to this very high solubility of ammonia in water a large quantity of air is
required to transfer ammonia effectively from the water to the air. In principle there are three different configurations of stripping units, as shown in Fig. 7.3; see
Montgomery (1 985).
out
in
rir in lnfiuentl
+ Air in
out
Air out +
1 Effluent
Figure 7.3. Configuration of air stripping units. From left to right: countercurrent,
cocurrent and cross flow.
The efficiency of the process depends on:
1. pH, according to the considerations mentioned above. Equations (7.2) and
265
(7.3) may be applied and in case where the ionic strength is significant, equations
(7.7) and (7.8) are used. 2. The temperature. The solubility of ammonia decreases with increasing temperature. The efficiency at three temperatures - OOC, 20°C and 40°C - is plotted
versus the pH in Fig. 7.4 and versus the tower height in Fig. 7.5
3. The quantity of air per mS of water treated. At least 3000 m3 of air per
msof water are required (see Fig.7.6).
4. The height of the stripping tower. The relationship between the
efficiency and the quantity of air is plotted for three heights - Figs. 7.5 and 7.6.
5. The specific surface of the packing (m2Im3). Greater specific surface
results in greater efficiency.
6 8 10 12
Fig. 7.4. Stripping efficiency as function of pH at three different temperatures.
266
4
100
90
6 0
30
0
0
Efficiencies (%) 20 OC
4 a Tower depth (m)
Figure 7.5. Effect of water temperature on ammonia stripping. 4 m3 air is used
per liter of waste water. The efficiencies are plotted versus the tower height for
various temperatures.
Figure 7.7 demonstrates the principle of a stripping tower. The waste water
treatment plant at Lake Tahoe, California, includes a stripping process. 10,000 m3 of waste water is treated per 24h at a cost of approximately 8 US cents (1992) per
m3. The capital cost is in the order of 20 US cents per m3 (based on 16%
depreciation and interest per year of the investment).
The cost of stripping is therefore relatively moderate, but the process has two
crucial limitations:
1. It is practically impossible to work at temperatures below 57°C. The
large quantity of air will cause considerable evaporation, which results in h e
267
water in the tower freezing. 2. Deposition of calcium carbonate can reduce the efficiency or even block
the tower.
Due to limitation 1) it will be necessary to use warm air for the stripping during winter in temperate climates, or to install the tower indoors. This makes the process too costly for plants in areas with more than 10,000 inhabitants and limits the application for treatment of bigger volumes to tropical or possibly subtropical
latitudes.
100
80
6 0
40
20
0 1600 3200 4800 6400 8000 9 6 0 0
Figure 7.6. Efficiency as function of m3 of air per m3 of water for three different
tower heights ..... line = 8 m, __ line = 6.7 m, --- line = 4 rn.
A very important shortcoming of some technological solutions is, that they
do not consider a total environmental solution, as they solve one problem but
create a new one. The stripping process is a characteristic example, since the
268
ammonia is removed from the waste water but transferred to the atmosphere, unless recovery of ammonia is carried out. In each specific case it is necessary to
assess whether the air pollution problem created is greater that the water pollution
problem solved. If a significant amount of municipal waste water were be treated by
air stripping, the ammonia removed by air would make a crucial contribution to the
air pollution problem of nitrogenous compounds on a regional basis.
A i r i n --C
h a s t e w a t e r - o u t
Fig. 7.7. The principle of a stripping tower.
269
7.3. Gas Transfer
Both aeration and stripping involve a gas-liquid mass-transfer process in
which the driving force is created by a departure from equilibrium. In other words,
the driving force in the gas phase is a partial pressure gradient, and is a concentration gradient in the liquid phase.
The transfer of a gas can be treated as a four-step process. The first step of a
stripping process involves passage of the dissolved gas from the liquid phase to
the gas-liquid interface. The second step is the passage of the gas through a liquid
film on the liquid side of the interface The gas must then pass through a gas film on
the vapor side of the interface. The gas must in the final step be dispersed
throughout the bulk of the gas. General conditions are such that one of the steps is
rate-limiting and the overall gas-transfer rate can be calculated on the basis of this step. The remaining steps are most often insignificant in the overall process.
In stagnant conditions diffusion of the gas through the bulk solution is
generally the slowest step and an expression for molecular diffusion can be used
to predict the transfer rate.
The diffusion can be calculated by means of Fick's Law:
dc N = - D * A * -
dY (7.9)
where
N A = the cross-sectional area across which diffusion occurs
dcldy = the concentration gradient perpendicular to the cross-sectional area, A
D = diffusion coefficient.
= mass transfer per unit time
If, however, the solution is sufficiently agitated either by natural turbulence or
by mechanical mixing, the rate of transfer through the gas-liquid interface becomes
the controlling factor. For sparingly soluble gases such as oxygen and carbon
dioxide, the resistance of the liquid film controls the rate of gas transfer, while for
highly soluble gases such as ammonia, the transfer rate is controlled by the
resistance of the gas phase.
270
Gas solubility
calculated by Henry's Law:
The equilibrium concentration of a gas in contact with a liquid can be
Ceq = p l H (7.10)
where
Ceq H = Henry's Constant
p
= the equilibrium concentration of the gas in solution as molar fraction
= the partial pressure of the gas in the gas phase.
Henry's Constant is roughly proportional to the temperature; i.e., with increased temperature the solubility of a gas decreases. Figure 7.8 gives the
relation between solubility of ammonia and the temperature. As can be seen, the solubility changes significantly with the temperature; see also Figs. 7.2, 7.4 and 7.5.
Solubility
0 20 40 60 80 100
Temperature ( "C)
Figure 7.8. The solubility of ammonia plotted versus the temperature.
The temperature dependence of Henry's constant may be found by
27 1
use of one of the following two equations: (Srinath and Loehr 1974 and Montgomery 1985)
H = 0.268 exp ( 0.0525* t OC) bar (7.11)
H = 3754 1(1.987* (273 + t°C) + 6.135 bar (7.12)
Henry's Constant is also influenced by the presence of dissolved solids. The
combined effects of dissolved solids and temperature on the solubility of oxygen in
water are expressed by the following equation (Gameson and Robertson, 1955):
475 - 2.65 * C ~ S
33.5 + (T - 273) Ceq = (7.13)
where
Cds T
= the concentration of total dissolved solids expressed in gA = the absolute temperature expressed in K
It must be emphasized that this equation is developed under the conditions
that the pressure is 760 mm Hg and that clean water is in contact with wet air.
In this context it must be stressed that Henry's Law is an ideal law and gives
only approximate values. It is preferable to use solubility data if these are available.
Mass transfer Lewis and Whitrnan (1 924) developed equations for the transfer rate
controlled by the gas-film resistance as well as for the transfer rate controlled by the
liquid-film resistance:
N = KL * A(Ceq-C) = KG A(p- peq) (7.14)
where
N
A = area of cross-section
Ceq
= mass transfer per unit time
= concentration at equilibrium (saturation)
272
p peq KL
KG
DL DG
= partial pressure in the gas phase
= partial pressure at the interface = liquid-film coefficient defined as DLNL
= gas-film coefficient defined as DGNG
= diffusion coefficient in the liquid
= diffusion coefficient in the gas
Figure 7.9 shows a schematic representation of the liquid-gas mass transfer.
The liquid-film-controlled process can be expressed in concentration units by dividing by the volume, V:
(7.15)
KL,a = KL + ( A N ) is termed the overall film coefficient.
The transfer coefficient, KL, is affected by a number of variables. In general, the liquid-film coefficient increases with increasing temperature according to:
KL(t) = KL,~OO 1.028 b20) (7.16)
t =temperature ("C).
For KL,a in a bubble aeration system, the equation becomes
KL,a(t) = KL.a,No 1.02 0-20) (7.17)
The presence of surface-active agents in the waste water has a significant
effect on KL and A N (area to volume ratio). A decrease in surface tension will
decrease the size of the bubbles generated, which will increase A N . In some
instances the increase in A N will exceed the decrease in KL, with the overall effect that the transfer rate increases. Generally, KL,a decreases with increasing
concentration of impurities in water. A coefficient, b, defined as the as the ratio of KL,a for waste water to that for distilled water, is used to account for the influence of
273
the impurities in the waste water on KL,a. Figure 7.10 shows a characteristic change
in the coefficient b, as a function of BOD5 of water.
The liquid film resistance is usually not of importance for ammonia stripping. It is therefore possible to relate the transfer process directly to the gas film resistance, which in practice is performed by empirical relations between the resistance coefficients and the tower packing.
Gas- f i l m
P y G -b
lnterface
L i q u i d - f i l m
Figure 7.9. Schematic representation of interfacial mass transfer.
274
b
1 .o
0.5
0 100 200 7
300
BOD-5 of influent
Figure 7.1 0. A typical BOD5 / b relationship.
7.4. Design of Stripping Tower
Figure 7.1 1 shows the application of the mass conservation principle on a
countercurrent tower. The tower may be either a packed or a spray tower filled with
bubble-cap trays, or of any internal construction to bring about a good gas-liquid
contact. The following relationships are valid (y eel):
P - Y Y = -
1 - Y R -P and
G
l + Y GS = G(l -y) =
(7.18)
(7.19)
275
where G
y p = partial pressure Y
GS Pt = total pressure
= gas stream total moles / h /m2
= mole fraction of diffusing solute
= mole ratio of diffusing solute
= moles / h I m2 of non-diffusing, essentially insoluble gas
c2
L2 c S
Ls y 2
x2 y2
X 2 p2
Figure 7.1 1. Principles of mass conservation applied to countercurrent tower.
Similarly, the following equation is valid for the liquid stream (x eel):
X
x = 1 - x
(7.20)
276
L
1 + x Ls = L(l - x ) = (7.21)
where
L = liquid stream moles/h/m2
X = mole fraction of soluble gas
X = mole ratio of soluble gas
LS = moles /h I m2 of non-volatile solvent
Since the solvent gas (air) and solvent liquid (water) are essentially
unchanged in quantity as they pass through the tower, it is convenient to express
the material balance in terms of these.
The balance in the lower part of the tower (see Fig. 7.1 1) can be expressed
by
Gs(Yi - Y) = Ls(Xi - X) (7.22)
This is the equation of a straight line, the so-called operating line, which has a
slope of LsGsand passes through (Xi,Yi). The operating line also passes through
the point (X2,Y2).
In Fig. 7.12 the operating line is plotted together with the equilibrium
solubility curve, which may be found from Henry’s law and plotted in terms of the
mole ratio.
For a stripping tower, the operating line is always below the equilibrium solubility curve (see Fig. 7.12).
If we consider a packed or spray tower of unit area cross-section, it is
convenient to describe the interfacial surface between the gas and liquid as a
function of the dispersion of the liquid in the thin film over the packing. The following equation is valid:
dS = a*dZ (7.23)
where
S
a
= area of the interface expressed as m2/m2 tower cross-section
= m2 interfacial surfaceIm3 packed volume
277
2 = the height (m) of the tower.
Y
2 Y
Y, I
1 X X
2 X
Figure 7.12. Equilibrium curve (1) and operating line (2) for a stripping process.
The amount of solid in the gas passing the differential section of the tower is
G y mole/hlm*, and the rate of mass transfer to the liquid, d(G * y). This can be
related to the mass transfer coefficient as follows:
d(G *y) = KY a(y-yeq)dZ (7.24)
where Ky =the overall transfer coefficient.
Both G and Y vary from one part of the tower to another, but GS does not. Therefore, it is more convenient to use GS in these expressions:
278
The mass-transfer coefficient for diffusion of one component through a
second (the solvent) includes a term involving the average concentration, Ym, of the
non-diffusing gas along the path of the diffusion. If the concentration of solute varies considerably from one end of the tower to another, the quantity KG*a( 1 -y)m will be much more constant than KG*a alone. Therefore, equation (7.25) will be
transformed to
or KG*a( l -y)m*dZ
- (7.27) dY - (1 - Y b
(1-Y) Y - Y e s G -~
Equation (7.27) may be integrated to obtain, 2, in terms of KG%, but for many
situations the first term on the left-hand side is very close to unity. Since the number
of transferred units Ntog is defined as
(7.28)
then
Z = Ntog Htog (7.29)
Ntog can, as shown, be related to the height of the packing and the height per transfer unit, termed Htog. The height per transfer unit is an experimental
quantity, but it is more convenient to use it than KG*a and other mass-transfer
coefficients in the design of towers. Htog has the dimension of length. The subscript,
tog, is used, as seen in the terms Ntog and Htog to indicate that these terms are
based on an overall driving force y - yeq within the gas phase. These terms,
therefore, represent the vertical distance between the operating line and
equilibrium curve at any liquid concentration on a graph plotted in mole fractions. The quantity (l-y)m is the average concentration of non-diffusing gas at
either end of the diffusion path. (1-y) is the concentration of the main body of the
279
gas and (l-yeq) that at the liquid gas interface.
For all ordinary purposes the arithmetic mean is entirely satisfactory, and
equation (7.28) is changed to
Y i dY 1-Y2
Y2 Y - Y e q 1 - y i
Ntog = J + 1/2 In (7.31)
The calculation of the number of transfer units for dilute mixtures can be
simplified. When the gas mixture is dilute, the second term of equation (7.29)
becomes negligible and the equation may be simplified as follows:
Y i dY Ntog = J
Y2 Y-Yeq
(7.32)
If the equilibrium curve and the operating line in terms of mole fraction are considered as straight lines, it is possible to rewrite equation (7.32) as:
(7.33)
Equation (7.33) demonstrates that one overall gas-transfer unit is obtained when the change in gas composition equals the average of the overall driving forces causing the change. Let us consider the diagram shown in Fig. 7.13. The line (3) is vettically half-way between the operating line (2) and the equilibrium
curve (1). The step CFD, which corresponds to one transfer unit, has been constructed by drawing the horizontal line CEF, so that CE is equal to EF, and
continuing vertically to D.
280
TABLE 7.2
Liquid-film height of transfer unit
L
PL HtL = cp ( - )" SCLO.~
HtL = m, L=kg/h/m2, pL = kg/m/h, ScL = dimensionless (Schmidt number)
paddng cp n Rangeof L -----~-___----__I_ -------I- __-
Raschig rings:
3l8 in.
112 in.
1 in.
1.5 in.
2 in.
3.15 0.46
7.05 0.35
2.30 0.22 1,800-68,000
2.56 10-3 0.22
2.88 * 10-3 0.22
Berl saddles:
112 in.
1 in.
1.5 in.
3-in. partition rings
(stacked staggered)
Spiral rings (stacked
staggered):
3-in. single spiral
3-in. triple spiral
Drip-point grids (continuous flue):
No. 6146
No. 6295
1.43 10-3 0.28
1.26 l o 3 0.28
1.34 * 103 0.28
0.0168 0.09 13,000-63,000
1.95 ' l o 3 0.28 1,800-68,000
2.49 ' l o 3 0.28 13,000-63,000
3.51 l o 3 0.23 15,000-135,000
1.50 l o 3 0.31 11,000-100,000
- __---------I____ __
From the data of Shewood et al. (1940), and Moktad et al. (1943)
YG - YH may be considered as the average driving force for the exchange in
28 1
gas composition yo - yFCOrreSpOnding to this step. As GE is equal to EH and if the operating line is straight DF = 2 GE = GH, and the step CFD corresponds to one
transfer unit. In a similar way the other transfer units are stepped off.
Y
Figure 7.1 3. Graphical determination of transfer units (absorption).
The resistance to mass transfer in absorption and stripping processes in the
case both the gas film and liquid film are contraling factors can be calculated on the basis of the following equation:
(7.34)
where m = the slope of the equilibrium solubility curve (mole fraction in the
gadmole fraction in the liquid).
282
By comparing equation (7.26) with (7.29), Htog can be expressed by the contribution of individual phase resistances, HtG and HtL:
(7.35)
For diluted solutions, the ratio of concentrations of non-diff using substances will be nearly unity, and:
mG
L Htog = HtG + - HtL, (7.36)
where L is the flowrate in kg / h / m2.
Stripping of very insoluble gases such as oxygen, hydrogen or carbon
dioxide, is controlled by resistance to mass transfer in the liquid, for which H~L is a
direct measure. HtL can be found for common packing material from the empirical
expression
(7.37)
where cp and n can be found from Table 7.2 for different packings.
L = the flow rate kg/h/m2
SCL = the dimensionless Schmidt number = ~ L / P L * DL
p~ = the viscosity (kg/m/h)
PL = specific gravity
DL = diffusion coefficient.
In some instances Htog 5 H~G. This almost obtains for the stripping of ammonia from
water into air, but in this case the liquid-foam resistance is still not completely negligible although ammonia is very soluble in water.
It is possible to calculate HtG from empirical data:
(7.38)
283
where a, 0 and y are empirical constants, SCG = the dimensionless Schmidt
number, SCG = p~ / PG'DG, G and L = the gas and liquid flow rates respectively
measured in kg /h / m2. pc is the specific gravity of the gas.
The diameter of the tower is calculated on the basis of the minimum liquid rate for wetting and on the so-called flooding point.
Values of the empirical constants are listed in Table 7.3.
The minimum liquid rate for wetting Iw, can be calculated from the following
equation:
L
dL a Iw = (7.39)
where
dL
a L = See Table 7.2
= the density of the liquid kg/ms
= surface area of the packing m2/ m3
The flooding point has been defined as the gas velocity at which a liquid
layer forms on top of the packing. Based on experimental data, the following equation can be used for the determination of IW at the flooding point:
where dh = the hydraulic diameter of the packing and pL = the viscosity in kg/m*s.
Table 7.3 is based on data of Fellinger and Pigford (1952) and Molstad et al.
The function is shown in Fig. 7.14, where (1 943).
z = IW (1000 do.' is expressed as a function of Q. dh2/3
284
TABLE 7.3 Gas-film height of transfer unit
ClGO
LY
HtG = - SCGO.~
HtG=m, G=kg/h/rn2, L =kg/h/m2, ScG=dirnensionless (Schmidt number)
Raroed
G L Packing a O Y
Raschig rings:
318 in.
1 in.
1.5 in.
2 in.
4 in
Berl saddles:
112 in.
1 in. 1.5 in.
3-in. partition rings
(stacked staggered)
Spiral rings
(stacked staggered):
3-in. single
spiral 3-in. triple
spiral
Drip-point
(continuous flue):
No. 6146
No. 6295
.39 0.45 0.47 900-2,300 2,300-6,800
9.31 0.39 0.58 900-3,600 1,800-2,300
8.53 0.32 0.51 900-2,700 2,300-20,000 26.4 0.38 0.66 900-3,200 2,300-6,800
2.66 0.38 0.40 900-3,200 6,800-20,000
4.06 0.41 0.45 900-3.600 2,300-20,000
1.80 0.40 0.40 5,000-10,000 2,500-20,000
62.8 0.30 0.74 900-3,200 2,300-6,800
0.741 0.30 0.24 900-3,200 6,600-20,000
2.09 0.36 0.40 900-3,600 1,800-20,000 6.14 0.32 0.45 900-4,500 1,800-20,000
1338 0.58 1.06 700-4,100 13,000-20,000
2.17 0.35 0.29 600-3,200 13,000-45,000
21.7 0.38 0.60 900-4,500 2,300-13,000
4.02 0.37 0.39 600-4.500 13,000-30,000
5.40 0.17 0.27 450-4,500 9,000-52,000
285
0.01 0.03 0.1 0.3 1 .o 3 10
Figure 7.14. Plot for determination of flooding point. (1) Grids. (2) Stacked
rings. (3) Random packing of rings.
The flooding point represents the upper limit for the operation of the tower.
Operating conditions of the tower can be improved by increasing the gas flow.
Usually a gas flow of 50-60% of the flow corresponding to the flooding point is used. The diameter of the tower is found by the following procedure:
1.Based on L, G the specific gravity of the liquid and the gas, PL and Fig.
7.14, is found (1 w/dh2" * 103). dh2R is shown in Table 7.4 for different packing
materials.
2 . l ~ and dh must be chosen so, that IW is greater than 0.08 mslmlh for
common packing including raschig rings less than 7.5 cm, and greater than
0.12 m3/h for raschig rings larger than 7.5 cm.
3.Generally, 0.4 mslmlh can be considered as the upper limit for all types of
packing.
286
4.Based on equation (7.39) and the total flows (kg/m2) it is possible to find
the area of cross-section of the absorption stripping tower.
Table 7.4
Characteristic packing data I ___
Dimensions (inch) Poro- dh=hydrau- Gas flow
Diam. Height ness per& m2h3 (-) (m) lddhdn (&)
Packing Thick- Number Surface sity Iic. diam. entry tower
Coke 3 " 1-2 -
1
Brokenstone - 2
,I
1Q-114 -
Grids: " 1 1 1 I4
1 2 1 I4
Jagged grids: I, 4 4 112
2 2 3/8
1 1Q 1 112 3/16
(0
Stacked
Raschig rings:
Stoneware 4 4 3/8
3 3 3/8
3 3 1 14
2 2 1 I4
2 2 311 6
"
49 0.50 0.041 8.3 0.54-0.96
115 0.40 0.014 1.7 0.26
131 0.45 0.014 1.7 0.15-0.26
62.5 0.46 0.029 4.9 0.51-0.60
144 0.40 0.011 1.15 0.13
98.5 0.75 0,019 2.65 1.5-2.4
88.5 0.75 0.019 2.65 1.7-2.5
19.5 0.69 0.089 26.5 2.4-3.6
42.5 0.83 0.041 8.4 2.1-3.3
54.0 0.89 0.033 6.1 2.1-3.0
950 62.5 0.73 0.047 10.2 1.6-2.4
2300 82 0.66 0.032 5.7 1.1-1.5
2300 82 0.76 0.037 7.1 1.7
7400 118 0.67 0.023 3.5 0.86
7400 118 0.72 0.024 3.7 0.89
Metal 2 2 1116 6180 98.5 0.92 0.037 7.1 0.72-0.90
1 1 1116 47600 194 0.86 0.018 2.4 0.57-0.69
1Q 112 1B2 370000 377 0.87 0.009 0.85 0.3
Random
packings of
Raschig rings:
Stoneware 3 3 3/8 1810 65.5 0.72 0.044 9.2 0.67-1.2 " 2 2 114 5820 92 0.74 0.032 5.7 0.54-0.66
2 2 3/16 6000 95 0.79 0.033 6.0 0.63-0.93
287
TABLE 7.4 (continued) ____-__----
Dimensions (inch) Poro- dh=hydrau- Gas flow
Dam. Height ness perm? m2h3 (-) (m) l$dh312 (mh)
Packing Thick- Number Surface sity lic. diarn. entry tower
_ _ _ _ _ - - - ~ - 1112 1112 3/16 14100 125 0.73 0.023 3.5 0.51-0.81
1 1 3/32 46OOO 104 0.80 0.017 2.2 0.42-0.60
314 314 3/32 lOBo00 236 0.74 0.013 1.5 - " 1R 1l2 1116 37oooO 377 0.73 0.006 0.72 0.19
Berl-saddles:
Stoneware 1l2 - 528000 460 0.65 0.0057 0.43 " 81000 258 0.69 0.0107 1.10 not 1
11R - 22900 165 0.72 0.017 2.21 indicated
2 Boo0 120 0.72 0.024 3.72
Partly after G.A. Morris and J. Jackson, Absorption Towers, 1953.
7. 5. Practical Experience
The best results in practice are achieved by use of countercurrent packed
towers; see 0degaard (1988). The water is distributed on the top of the packing with distribution trays or spray nozzles. For a high air to water ratio, a mist
eliminator is necessary at the air outlet. Random packing of Raschig rings or
saddles or grids, made of metal, ceramic, plastic or even impregnated wood, can
be used.
Stripping ponds, see Fig. 7.15, might be used to remove 3040% ammonia,
but higher efficiencies can hardly be expected, even by introduction of agitation of
the pond surface. It might, however, be practical to install stripping ponds as supplement to stripping tower to account for peak loadings.
Figures 7.1 6 and 7.1 7, taken from Fetting (1 989), are constructed to facilitate
1. The operating temperature is selected for determination of Henry's
the design in practice.
constant; see equations 7.1 1 and 7.12.
288
2. The, minimum ratio air to water, A M I can be derived from a simple mass balance
A M I = 1244*p*(1 - ef) I H (7.41)
where p is the total pressure, ef is the required efficiency, i.e., the ratio between the
concentration of ammonia in the effluent and in the influent. It can be recommended to multiply the minimum value of A M I by 1.2 -2.3 in practice.
Figure 7.15. Ammonia stripping pond system. (Drawn by Morten V. Jsrgensen).
3. The stripping factor R is found, based upon the selected A N ratio, s:
R = h *~11244 (7.42)
4. Figure 7.16 gives the number of transfer units, when R and the fraction
removed are known. Figure 7.16. is valid for countercurrent operation, while Fig. 7.1 7 is constructed for single-stage cross-flow operation. Note that this latter figure uses the inverse stripping factor and the fraction remaining.
Due to the growing concern over air pollution problems, including the dry and wet deposition of nitrogen components as an increasing source of nutrients to
289
fresh and marine waters, it is necessary in most cases to combine the stripping unit with an absorption unit. The removed ammonia is absorbed in sulfuric acid for production of ammonium sulfate, which can be used as fertilizer. Figure 7.18 shows a flow chart of the combination of stripping and absorption.
Number of transfer units
Figure 7.16. Number of transfer units for counter current operation as a function
of removal efficiency and stripping factor, R. Reproduced from Fetting (1989).
High efficiency in ammonia removal requires adjustment of pH to about 11 .O before the stripping process. It implies that the pH after the stripping must be readjusted. The pH might drop about 0.2 by the stripping process due to removal of
290
ammonia, but a pH of 6-8 is required for the effluent.
C 0 .rl U 0 a k E
Figure 7.17. Number of transfer units for a single-stage cross-flow operation as a function of the concentration of ammonia remaining in water and of stripping factor, R. fleproduced from Fetting (1989).
The readjustment of pH can be carried out by recarbonization. Carbon dioxide is easily obtained from incineration of bio-gas, sludge or solid waste. Sulfuric acid might also be applied, but it is a less cost-effective alternative, which can only be recommended if there is no easy access to carbon dioxide.
291
+
J
Stripping
unit
r 4
A bsorp-
tion unit Recycled absor- bent liquid 1 77
Ammonium salt low down liquid
Figure 7.18. Process for stripping and recovery of ammonia.
7.6. Application of stripping
The stripping process is used to remove volatile gases such as hydrogen
sulfide, hydrogen cyanide as well as ammonia. The removal of ammonia by
stripping is used in the treatment of municipal waste water, where it has found very
little application due to the problems mentioned in Section 7.2. Generally it can be
concluded that the method is not economic in a temperate climate for large flows of
waste water with relatively small concentrations of ammonia, as is found in municipal waste water. An additional problem is the air pollution caused by the removed ammonia, see Section 7.1. A recovery of ammonia by absorption in acid
is possible, but the value of the recovered ammonia as ammonium sulfate is less
than the costs of the recovery process.
The process has, however, found application at two well-known waste water
treatment plants: at Lake Tahoe and in Pretoria. The flow chart of the latter plant is
shown in Fig. 7.19. The main problem behind this solution is, however, not a pollution problem, but the scarcity of water.
If the concentration of ammonia is higher and the volume of waste water to
292
be treated smaller, the process becomes more favorable. This is for instance the case for the reject water, produced by dewatering of municipal sludge. The
concentration here is 2-5 times higher than in municipal waste water and the
process has therefore found some application for the treatment of this water
particularly where the treatment plant is too small to handle the reject water in addition to the waste water.
Stripping has also been suggested for the treatment of industrial waste
water and for the regeneration of the liquid used for eluting ion exchangers
(Jsrgensen, 1975). In these cases ammonia is removed from relatively small
volumes and is present in high concentrations. As the amount of air needed is
roughly independent of the ammonia concentration, see equation (7.41), the cost
per kg of ammonia removed is much lower at high ammonia concentrations. The
method therefore becomes much more attractive for industrial waste water with
high ammonium concentrations or for recovery of elution liquids, used for
regeneration of ion exchangers. Up to now stripping has not been used widely for
treatment of industrial waste water, but with the growing demand for nitrogen
removal, it is anticipated that the application of the method will increase in the
coming decade.
Typical concentrations in waste water originating from production of ammonia, meat-bone-meal or fish meal are in the order of 500-1000 mg/l or 10-25
times higher than for municipal waste water. Elution liquids after regeneration of
ion exchange columns may contain even higher ammonia concentrations and
have already a high pH ( see also Chapter 8 ).
293
10 9
Figure 7.19. Waste water treatment plant, Pretoria. After mechanical-biological treatment (not shown) there follows 1) an algae pond, 2) aeration 3) lime precipitation 4) sludge drying 5) air stripping of ammonia 6) recarbonization 7)
sand filtration 8) chlorination 9) adsorption on activated carbon 10) a second chlorination.
294
8. BREAKPOINT-CHLORINATION
8.1. Principles of Breakpoint-chlorination
Chlorine can oxidize ammonia according to the following reaction scheme:
CI2 + H20
NH3 + HOCl <=> NHZCI + H2O
NH2CI + HOCI <=> NHClp + HOCl
<* HOCI + HCI
NHC12 + H20
<=>NC13 + H20
Activated carbon is able to adsorb chioramines, and so a combination of chlorination and adsorption on activated carbon can be applied for removal of ammonia.
The most likely reaction for chloramine on activated carbon is a surface oxidation :
C + 2NHC12 + H2O <=> N2 + 4H+ + 4CI- + CO (8.2)
Furthermore, it is important to know that the Clnl NH3-N oxidized mole ratio is 2:1,
for oxidation by this pathway. The mono-chloramine reaction with carbon appears more complex. On fresh carbon the reaction is most probably:
NH2Cl + H20 + C = NH3 + H+ + CI- + CO (8.3)
After this reaction has proceeded to a certain extent, partial oxidation of mono- chloramine is observed, possibly according to the equation:
2NH2CI + CO <=> N2 + H20 +2H+ + 2CI- + C (8.4)
It has been observed that activation of fresh carbon is necessary before
295
mono-chloramine can be oxidized. However, the reaction of chlorine with ammonia or amino compounds
presents a problem in the practice of chlorination of waste water containing such
nitrogen compounds.
Added chlorine (mgA)
Fig. 8.1. Breakpoint chlorination.
Figure 8.1 shows the residual chlorine as a function of the chlorine applied. Between points 1 and 2 in the figure, mono- and di-chloramine are formed. The
oxidation processes with chlorine occurring between points 2 and 3 give a decline
in residual chlorine. Point 3 is called the breakpoint. It corresponds to a stoichiometric ratio of chlorine to ammonium-N of 7.6. It is sufficient to add this
amount of chlorine for ammonium removal, provided that the waste water does not
contain other components, that are oxidized by chlorine. A ratio of chlorine to
ammonium-N of 8-10 is, however, required in most cases in practice. Addition of
chlorine in this interval probably produces free nitrogen gas as the predominant product of oxidation. Fair et al (1968) even propose that the reaction involving the
formation of NOH as an intermediate, followed by the formation of nitric oxide, NO, could explain the observations between points 2 and 3:
296
2NHCI2 + 6H20 = 2NOH + 4H@ + 4CI-
2NOH + HOCl = 2N0 + H30+ + CI-
In total:
2NHC12 + HOCl + 6H20 = 2N0 + 5H30+ + 5CI- (8.7)
Further addition of chlorine beyond the breakpoint gives an increasing
residue of free chlorine. Chlorine doses below the breakpoint requirement can be
used to oxidize ammonia if chlorination is followed by contact with activated carbon
(Bauer and Vernon, 1973).
When accidental overdosing of chlorine has occurred or after an intentional
addition of large quantities of chlorine to accelerate disinfection, it will be desirable
to remove the excess chlorine. This is possible with a reducing agent, such as
sulfur dioxide, sodium hydrogen sulfite or sodium thiosulfate:
SO2 + CI2 + 2H2O = H2S04 + 2HCI (8.8)
NaHS03 + C12 + H20 = NaHS04 + 2HCI (8.9) 2Na2S203 + C12 = Na2S406 + 2NaCI (8.10)
Oxidative degradation by chlorine is limited to a small number of
compounds. Nevertheless, oxidation of these compounds contributes to overall
reduction of BOD, in wastes treated with chlorine. A disadvantage is that
chlorinated organic compounds may be formed in large quantities. A variety of
chlorine compounds is applied in waste water treatments. For these compounds
the available chlorine can be calculated, and is generally expressed as percentage chlorine having the same oxidation ability. Data for the different chlorine-containing
compounds are given in Table 8.1.
It can be seen that the actual percentage of chlorine in chlorine dioxide is
52.5, but the available chlorine is 260%. This is, of course, because the oxidation state of chlorine in chlorine dioxide is +4 which means that five electrons are transferred per chlorine atom, while C12 only transfers one electron per chlorine
atom.
297
Hypochlorite is obtained by the reaction of chlorine with hydroxide in
aqueous solution:
CI, + 2NaOH <=> NaCl + NaOCl + H,O (8.11)
Table 8.1
Actual and available chlorine in pure chlorlne-containing compounds
Mol. Chlorine equiv. Actual chlorine Available Compound mass (moles of CI,) (W chlorine (“10)
~~~ ~~~
(312 71 1 CI 2 0 87 2 c10, 67.5 2.5
NaOCl 74.5 1
CaClOCl 127 1
HOCI 52.5 1
Ca(OCI), 143 2
100 100 81.7 163.4
52.5 260
47.7 95.4
56 56
49.6 99.2
67.7 135.4
Chlorinated lime, also called bleaching powder is formed by reaction of
chlorine with lime:
Ca(OH), + CI, <=> CaCI(0CI) + H,O (8.12)
A higher content of available chlorine is present in calcium hypochlorite,
Ca(OCI),. Chlorine dioxide is generated in situ by the reaction of chlorine with
sodium chlorite:
2NaCI0, + CI, <=> 2C10, + 2NaCI (8.13)
298
8.2. Process Variables
In the removal of ammonia with a dose of chlorine followed by contact with activated carbon, pH determines the major chlorine species. The studies reported herein indicate that a pH value near 4.5 should be avoided, because NHCI,
predominates and thus 10 parts by weight of chlorine are required for each part of
NH3-N oxidized to N,. At a slightly higher pH and using acclimated and activated carbon, the portion of mono-chloramine increases and the chlorine required per unit weight of NH3-N oxidized should approach 7.6 parts, ignoring the chlorine
demand resulting from other substances. However, further testing should be used to verify this conclusion in each case.
Laboratory studies at Blue Plains in Washington (Pressley et al., 1970 and
1973), in which buffered distilled ammonia nitrogen solutions of 20 mg/l
concentrations were subject to breakpoint-chlorination dosages, showed a definite optimum pH for breakpoint in the range of pH 6 to 7. The chlorine dosage at optimum pH levels were found to be 8:l (chlorine to ammonium-N).
The reaction rate has not been measured quantitatively, but it has been noted that the reaction is very rapid (Morris, 1965). The optimum pH for the reaction rate is 8.3, but at pH 6-7 the reaction is completed in 0.2 seconds.
There is no evidence that normal variations in the temperature of waste
water effluents and initial mixing conditions affect the nitrogen removal by this process.
Organic nitrogen is to a certain extent removed by the breakpoint- chlorination according to Brown and Caldwell (1975), while Taras (1 953) has
reported a very slow reduction of amino acids. Nitrate and nitrogen chloride are occasionally found in the effluents from the
breakpoint-chlorination process. An increasing level of pretreatment decreases the amount of chlorine
required to achieve breakpoint, as demonstrated in Table 8.2, where results reported by Pressley et al. (1973) and Brown and Caldwell (1975) are summarized. Increase of total dissolved solid will generally imply a higher chlorine to ammonium-N ratio.
The application of activated carbon for dechlorination is recommended, as it serves several functions other than removal of residual chlorine. Carbon, as demonstrated in Section 8.1, can effectively catalyze the chemical reactions and
299
remove soluble organics through adsorption.
Table 8.2.
Effect of Pretreatment on Ch1orine:ammonium-N breakpoint ratio.
Type of water pH Initial N conc. Final N conc. Breakpoint-ratio mg/l mg/l
Buffered water 6-7 20
Raw waste water 6.5-7.5 15
Secondary effluent 6.5-7.5 1 1.2
Tertiary effluent 6.5-7,5 9.2
Lime clarified raw
waste water, filtered 7.0-7.3 11.2
0.1 8: 1
0.2 9:l - 1O:l
0.1 8:l - 9:l
0.1 8: 1
0.1 9:l
Stasuik et al., (1973) has studied the required contact time for complete dechlorination of both free and combined chlorine. They found that 10 minutes
were sufficient.
8.3. Design of Breakpoint-Chlorination Units
The design of breakpoint-chlorination follows the stoichiometric relations
already presented in Sections 8.1 and 8.2. The amounts of chlorine, and other
chemicals including acids and bases for pH-adjustment and sulfur dioxide for dechlorination, can be calculated from these relations.
The design of the adsorption unit is mainly based on empirical relations:
- A hydraulic application rate of 5 0.1 ml m2 is recommended.
- 50 000 - 100 000 m3 waste water can be dechlorinated per m3 of activated
carbon between two regenerations of activated carbon.
The spatial requirements are low due to the high rate of the chemical
300
reactions involved in the breakpoint chlorination and the various dechlorination
processes.
The TDS (total dissolved solid) increment as a result of break point chlorination can be found from the figures in Table 8.3.
Table 8.3.
Ratios of total dissolved solids (TDS) to ammonium N - removed for the application of different chemicals by break point chlorination
Chemical Addition TDS increase : ammonium-N removed
________________________________________------__--- Chlorine gas 6.2 : 1
Sodium hypochlorite 7.1 :1
Chlorine gas+lime for neutra-
Chlorine gas+sodium hydro-
xide for neutralization of all acidity 14.8:l
lization of all acidity 12.2:l
8.4. Application of Breakpoint-Chlorination for Removal of Nitrogen
Complete removal of the 25-40 mg per liter ammonium-N is far too costly by
this method. Chlorine costs about 38-45 US cents per kg, which means that the
chlorine consumption alone will cost about 14 US cents per m3 waste. When the capital cost and the other operational costs are added the total treatment cost will
be as high as 30-45 US cents per m3, which is considerably more expensive than
other nitrogen removal methods.
It is possible to use chlorine to oxidize ammonium compounds to free nitrogen, but this process involves even higher chlorine consumption and, is
therefore, even more expensive.
The formation of organic chlorine compounds is another crucial
disadvantage of this process, because discharge of these compounds should be
301
avoided due to their high toxicity.
1) By using sufficient chlorine it is possible to obtain a very high efficiency. 2) The low spatial requirement makes it particularly suitable for certain
applications, including addition to an existing facility, where nitrogen removal is
required, but space constraints exist. This means that the method has found application mainly after other
ammonium removal methods, where high efficiencies are required. This is the
case when the waste water is reclaimed, for example in the two plants shown in
Figs, 7.19 and 8.2. As can be seen, it is necessary to use several treatment processes to achieve a sufficient water quality after the treatment. Chlorination and treatment on activated carbon are used as the last treatment to assure good
ammonium removal and sufficient disinfection of the water. An additional
chlorination is even used after the treatment on activated carbon to ensure a
chlorine residue in the water supply system. Note that the solution in Fig. 7.19,
where the major portion of ammonium-nitrogen is removed by stripping before the
residue of ammonium-N is removed by breakpoint-chlorination, is preferable,
because the operating costs become more limited due to the pronounced lower
consumption of chlorine. It should be mentioned in this context that ozonation, which is a disinfection
process widely used for treatment of water and related to chlorination, is able to oxidize amines. It is possible by ozonation to oxidize amines completely to nitrite and nitrate, provided that ozone is used in a ratio to the concentration of amines
slightly above the stoichiometric ratio; see Elmghari-Tabib et al., 1982.
Ozone is also able to oxidize ammonia to nitrate. This process is catalyzed
by the presence of bromide ions; see Haag et al., 1984.
The method has, however, two advantages:
302
. : .:
.
-. .......
Figure 8.2. Production of potable water from waste water in Windhoek, Namibia.
3 03
This Page Intentionally Left Blank
9.1 Principles of Ion Exchange
Ion exchange is a process in which ions on the surface of a solid are
exchanged for ions of a similar charge in a solution with which the solid is in
contact. Ion exchange can be used to remove undesirable ions from waste water.
Cations (positive ions) are exchanged for hydrogen or sodium, and anions
(negative ions) for hydroxide or chloride ions.
The cation exchange on a hydrogen cycle can be illustrated by the following
reaction, using, in this example, the removal of calcium ions, which are one of the
ions (Ca2+ and Mg2+) that cause hardness of water:
H2R + Ca2+ <=> CaR + 2H+ (9.1)
where R represents a cation exchange resin. The anion exchange can be similarly illustrated by the following reactions:
S042- + R(OH), = S04R + 20H- (9.2)
When all the exchange sites have been replaced with calcium or sulfate
ions, the resin must be regenerated. The cation exchanger can be regenerated by
passing a concentrated solution of sodium chloride or a strong acid through the bed, while the anion exchanger, which in this case is of hydroxide form, must be
treated by a solution of hydroxide ions, e.g., sodium hydroxide.
Ion exchange is known to occur with a number of natural solids, such as soil,
humus, metallic minerals and clay.
Clay, and in some instances other natural materials, can be used for demineralization of drinking water. In the context of adsorption, the ability of
aluminum oxide to make a surface ion exchange should be mentioned. The natural
clay mineral, clinoptilolite, can be used for waste water treatment as it has a high
305
selectivity for removal of ammonium ions; see also Section 5.8. Synthetic ion exchange resins consist of a network of compounds of high
molecular weight to which ionic functional groups are attached. The molecules are cross-linked in a three-dimensional matrix and the degree of the cross-linking determines the internal pore structure of the resin. Since ions must diffuse into and out of the resin, ions larger than a given size may be excluded from the interaction through a selection dependent upon the degree of cross-linking. However, the nature of the groups attached to the matrix also determines the ion selectivity and
thereby the equilibrium constant for the ion exchange process. The cation exchangers contain functional groups such as sulfonic R-S03-H - carhxy/ic, R-
COOH - phenolic, R-OH and phosphonic, R-P03H2 (R represents the matrix). It is
possible to distinguish between strongly acidic cation exchangers derived from a strong acid, such as H2S0,, and weakly acidic ones derived from a weak acid,
such as H2CO3. It is also possible to determine a pK-value for the cation
exchangers in the same way as for acids generally. Thus:
R-S03H = R-S03- + H+
[H+] * [R-SO<] = K pK = -bgK (9.3)
[R-SOSH]
Anion exchange resins contain such functional groups as primary amine, R-
NH2, secondary amine, R-RiNH, and tertiary amine R-RI-R~N groups and the
quaternary ammonium group R-R I R2RsN+OH-.
It can be seen that the anion exchanger can be divided into weakly basic and strongly basic ion exchangers derived from quaternary ammonium
compounds. It is also possible to introduce ionic groups onto natural material. This is
done by using cellulose as a matrix, and due to the high porosity of this material it is possible to remove even high molecular weight ions such as proteins and
polypeptides. Preparation of cation exchange resin, using hydrocarbon molecules as a
3 06
matrix, is carried out by polymerization of such organic molecules as styrene and methacrylic acid. The degree of cross-linking in styrene is determined by the amount of divinylbenzene added to the polymerization. This can be illustrated by the example shown Fig. 9.1.
k H = CH2
- . - - - . -CH - CH2 - CH - CH2 - CH - CH2 - CH - - - - -
Figure 9.1. Polymerization of styrene and vinylbenzene to form polystyrene with degree of cross-linking.
It is characteristic that the exchange occurs on a chemical equivalent basis. The capacity of the ion exchanger is therefore usually expressed as equivalents per liter of bed volume.
307
When the ion exchange process is used for reduction of hardness, the
capacity can also be expressed as kg of calcium carbonate per m3 of bed volume. Since the exchange occurs on an equivalent basis, the capacity can be found based either on the number of ions removed or the number of ions released. Also,
the quantity of regenerant required can be calculated from the capacity. However, neither the resin nor the regeneration process can be utilized with 100% efficiency.
100,
75
C in
C
.- 2 ,I
50
25
0
02%
0 25 50 75 100
% in solution
Figure 9.2. Illustration of the preference of an ion exchange resin for a particular ion. The selectivity coefficient at 50% in solution can be found from the diagram to be 821 18 = 4. 6.
308
Figure 9.2 illustrates the preference of an ion exchange resin for a particular ion. The percentage in the resin is plotted against the percentage in solution.
The selectivity coefficient, KAB, is not actually constant, but is dependent upon experimental conditions. A selectivity coefficient of 50% in
solution is often used = a-50%.
If we use concentration and not activity, it will involve, for monocharged ions:
CB = CA
CRA
CRB a-50% = KAB,~o% = - (9.4)
The plot in Fig. 9.2 can be used to read a-50%.
The selectivity of the resin for the exchange of ions is dependent upon the
ionic charge (and the ionic size. An ion exchange resin generally prefers counter ions of high valence. Thus, for a series of typical anions of interest in waste water
treatment one would expect the following order of selectivity:
Similar for a series of cations:
But this is under circumstances where the internal pore structure of the resin
does not exclude the ions mentioned from reaction. Organic ions are often too large to penetrate the matrix of an ion exchange, an effect which is, of course, more
pronounced when the resins considered have a high degree of cross-linking. As most kinds of water and waste water contain several types of ions besides those
which must be removed it is naturally a great advantage to have a resin with a high
selectivity for the ions to be removed during the ion exchange process.
The resin utilization is defined as the ratio of the quantity of ions removed during the actual treatment to the total quantity of ions that could be
309
removed at 100% efficiency; this is the theoretical capacity. The regeneration efficiency is the quantity of ions removed from the resins compared to the quantity
of ions present in the volume of the regenerant used. Weak base resin has a signi-
ficant potential for removing certain organic compounds from water, but the efficiency is highly dependent upon the pH.
It seems reasonable to hypothesize that an adsorption is taking place by the formation of a hydrogen bond between the free amino groups of the resin and
hydroxyl- groups of the organic substance taken up. As pH decreases, so that the
amino groups are converted to their acidic form, the adsorption capacity
significantly decreases.
The exchange reaction between ions in solution and ions attached to the
resin matrix is generally reversible. The exchange can be treated as a simple stoichiometric reaction. For cation exchange the equation is:
An+ + n(R-)B+ = nB+ + (R-)nA"+ (9.5)
The ion exchange reaction is selective, so that the ions attached to the fixed
resin matrix will have preference for one counter ion over another. Therefore the
concentration of different counter ions in the resin will be different from the corresponding concentration ration in the solution.
According to the law of mass action, the equilibrium relationship for reaction
(9.5) will give:
aen aRA
aA * a m KAB =
where aB and a~ are the activity of the ions B+ and A"+ in the solution and
correspondingly aRB and a m are the activities of the resin in B- and A-form,
respectively. Note that the activities are used, which means that the activity
coefficients should be calculated as shown in Section 7.1.
As mentioned above the clay mineral, clinoptilolite, can take up ammonium
ions with a high selectivity. This process is used for the removal of ammonium from
municipal waste water in the U.S.A., where good quality clinoptilolite occurs. Clinoptilolite has less capacity than the synthetic ion exchanger, but its high
310
selectivity for ammonium justifies its use for ammonium removal. The best quality
clinoptilolite has a capacity of 1 eqv. or slightly more per liter. This means that 1 liter
of ion exchange material can remove 14 g ammonium -N from waste water,
provided all the capacity is occupied by ammonium ions. Municipal waste water
contains approximately 28 g (2 eqv.) per m3, which means that 1 m3 of ion
exchange material can treat 500 m3 waste water (which represents a capacity of
500 bed volumes). The practical capacity is, however, considerably less - 150-250
bed volumes - due to the presence of other ions that are taken up by the ion
exchange material, although the selectivity is higher for ammonium that for the
other ions present in the waste water. The concentration of sodium, potassium and
calcium ions might be several eqv. per liter, compared with only 2 meqv. per liter of
ammonium ions.
Clinoptilolite is less resistant to acids or bases than synthetic ion
exchangers. A good elution is obtained by use of sodium hydroxide, but as the material is dissolved by sodium hydroxide a very diluted solution should be used
for elution to minimize the loss of material. A mixture of sodium chloride and lime is
also suggested as alternative elution solution.
The flow rate through the ion exchange column is generally smaller for
clinoptilolite than for synthetic material resin - 10 m/h as against 20-25 mh.
The elution liquid can be recovered by air stripping, as mentioned in Section 7.6. The preconcentration on the ion exchanger makes this process attractive - the sludge problem is diminished and the cost of chemicals is reduced
considerably. For further details about this method of recovery, see Jerrgensen
(1973 and 1975).
Another ion exchanger selective for nitrogen compounds is the above
mentioned cellulose ion exchanger. It has a capacity of about 1 eqv./ I. of which at least 50% is highly selective for proteins and other high molecular nitrogen
organics. It makes the application of this ion exchanger attractive for industrial
waste water with high concentrations of proteins and where recovery of the
proteins is desirable.
A combination of chemical precipitation and ion exchange has developed as
an alternative to the mechanical-biological-chemical treatment method. A flowchart
of such a plant is shown in Fig. 9.3. After the chemical precipitation the waste water
31 1
is treated on two ion exchangers (which, however, could be in one mixed bed column). The first ion exchanger is cellulose-based for removing proteins and reducing BOD,. The nitrogen concentration is here typically reduced from total N
30 mgA to total N 15-20 mgA due to the high selectivity of the cellulose-based ion exchanger for organic nitrogen compounds. The second column could be either clinoptilolite andlor activated alumina. A plant using this process has been in
operation since 1973 in Sweden, giving results comparable with or even better than the generally applied 3 steps treatment (see Table 9.1).
Figure- 8.3. Flowchart of a combination of chemical precipitation and ion exchange. (A) a submersible pump, (B) the settling basin, (C) an intermediate
vessel, where carbon dioxide is added, (D) a carbon dioxide container (50 atm., 25 liters.), (E) a pump feeding the ion exchangers, (F) elution liquid, (G) a hand-pump, (H) a dosing pump.
312
The capital cost and operating costs are approximately the same as for a three-steps plant. However, the plant produces 2-4 times less sludge than the
normal 3 step plant, giving a correspondingly lower sludge treatment cost.
TABLE 9.1
Analysis (mg 1-1) of municipal waste water after chemical precipitation in combination with ion exchange (flowchart see Fig. 9.3)
BOD5 10 - 18
COD 30 - 45
P
N < 1 if clinoptilolite is used, otherwise 10-15 mgll
< 0.1 if activated alumina is used, otherwise 2-4 mgA
9.2. Process Variables
The pH value is crucial for the ion exchange process, as the form of the ion
exchanger is dependent on pH, see equations (9.1) - (9.3), unless the ion
exchanger is a strong acid or base, and as the form of the ions to be taken up is
dependent on pH.
Optimum ammonium exchange by clinoptilolite occurs within an influent pH
range of 4 to 8. If the pH drops below this range, hydrogen ions begin to compete with ammonium for the available ion exchange capacity. As the pH increase above
8, a shift in the ammonia-ammonium equilibrium toward ammonia begins. Consequently, operation outside the pH range 4 to 8 results in a pronounced
decrease of exchange capacity.
The rate of exchange increases with decreasing clinoptilolite size. However,
the improved rate of exchange is accompanied by disadvantage of higher head loss. A suitable flow rate in practice is 8-10 bed volume / hr, providing that the influent is clarified secondary waste water with less than 30 mg / I suspended
matter. Biological growth which occurs is adequately removed by the regeneration.
The break through is determined by the desired concentration of ammonium
313
in the effluent. A typical break through curve is shown in Fig. 9.4. The corresponding utilization of the ion exchange capacity as a function of the column depth is illustrated in Fig. 9.5. The transition zone will have a certain depth, dependent on the flow rate but independent of the depth of the column; see Fig. 9.6, which is reproduced from Jprrgensen et al., (1978).
It implies that a shallow bed utilizes less than a deeper bed, although the flow rate expressed in bed volume per unit of time will mean a smaller actual flow
for the shallow bed. A deeper bed, on the other hand, will mean a higher head
loss. A compromise between the head loss and the utilization has to be found and a bed depth of 1.5-2 meters is recommended in practice; see Koon and Kaufman (1971) and Suhr and Kepple (1974).
0 100 7
200
Number of bed volumes
Figure 9.4. An ammonium break through curve is shown. The ammonium-N concentration in the effluent is plotted versus the number of bed volumes treated.
Furthermore, it is recommended to use two or more columns in series, because this makes it possible to utilize the entire capacity of the first column and
314
let the second column provide the required concentration of ammonium-N in the final effluent. After regeneration of the first column at saturation, the second column
becomes the first and the freshly regenerated column number two in the series. Thereby it becomes possible simultaneously to achieve a low concentration in the effluent and a full utilization of the ion exchange capacity.
Figure 9.5. The concentration of ammoniurn-N in the ion exchanger is shown as
a function of the depth at ammonium breakthrough. The ion exchanger is saturated
up to the transition zone, where the capacity is not used entirely. The depth of the
transition zone is dependent on the flow rate (mlh), but not on the total depth of the column.
Although clinoptilolite prefers ammonium ions to other cations, it is not
absolutely selective and other ions do compete for the available ion exchange capacity. The ion exchange equilibria for the exchange of ammonium versus
sodium, potassium, calcium and magnesium are available in the literature. Figure
315
9.7 gives the selectivity coefficients versus concentration ratios of sodium,
potassium, calcium and magnesium respectively. These curves illustrate that
clinoptilolite is selective for ammonium relative to all the examined ions except for
potassium. It is possible from such curves to predict the ammonium capacity of clinoptilolite in the presence of various concentrations of other cations.
20
1 0
0 0 3 6 9 12 14
Flow rate in m /hr
Figure 9.6 Z-nu, the layer not used is plotted versus the flow rate. Z-nu is
independent of the height of the column, but as shown on the figure dependent
upon the flow rate.
Clinoptilolite is available in different purities, dependent on the geological
formation of this clay mineral. The clinoptilolite from California has a purity of 85- 95%, while a Hungarian type from Tokaj has a purity of only 60-70%. The capacity is roughly proportional to the purity.
Investigations of the latter type of clinoptilolite, see Jsrgensen et al (1975)
and (1978), have demonstrated that a treatment of the clay mineral by sodium
hydroxide or sodium carbonate, before use, will improve the uptake of ammonium.
The results are expressed by use of the following equation:
K = (QIC)*(C, - C)"I(Qo- Q) (9.7)
316
where C is the equilibrium concentration of ammonium ions in solution, meqv A , C, is the total initial concentration of ammonium ions in solution, meqv A , Q is the ammonium ions taken up by clinoptilolite meqv 19, Q, is the total ion exchange
capacity of the sorbent, meqvlg, while K and n are characteristic constants, which can be found by use of a logarithmic plot of equilibria data. The equation can be used for all types of waste waters and clinoptilolites. In each case equilibria data must be used to find K, n and GI,.
10
1 .o K I-
Concentration ratio competing ion: ammonium
Figure 9.7. The selectivity coefficient a-50% of ammonium competing with potassium, sodium, calcium or magnesium versus concentration ratio of competing
ion I ammonium. Note that the graph is double logarithmic.
317
Table 9.2 gives the results of investigations of treated (with sodium
hydroxide or sodium carbonate) and untreated Hungarian clinoptilolite, using
ammonium solutions in distilled water to find the equilibrium data.
The equilibrium curve, resulting from equation (9.7) can be used directly in the design of ion exchange columns as presented in Section 9.4. Note that the
untreated clinoptilolite gives an equilibrium curve quite different from the treated
one and that the treated clinoptilolite will give a far better uptake of ammonium; see Table 9.2.
The regeneration of the ion exchange material is carried out either by
sodium or calcium ions by passing the regenerant through the clinoptilolite in the opposite flow direction of the normal service cycle. Lime-slurry was used for the first
studies of this process. It was, however, found that elution with lime could be
speeded up by the addition of sufficient sodium chloride (0.1 M). Ammonium ions
are converted to ammonia, so it can readily be removed from the regenerant, and the volume of regenerant required for complete regeneration decreases, with
increasing pH of the regeneration liquid. Precipitation of calcium carbonate and
magnesium hydroxide occurs, however, at high pH, which leads to clogging of the
exchanger inlets and outlets.
Table 9.2.
The three parameters in equation (9.7) for untreated and treated clinoptllollte, origlnating from Tokaj, Hungary.
Parameter Untreated Treated
K 1.16
n 1.18
QO 0.65
1.78
1.25
0.76
Two large municipal waste water installations in California and in Virginia
utilize a regenerant with a pH near neutral. The active portion of the regenerant is
a 2 percent sodium chloride solution. A typical elution curve for ammonium with this
type of regenerant is shown Fig. 9.8. It is seen that approximately 25-30 bed
318
volumes are required before the ammonium concentration reaches equilibrium, while 10-20 bed volumes are sufficient at high pH regeneration. If the regenerant is
recovered, see below, the volume is not very critical. Variations in regenerant flow rates of 4-20 bed volumes I hr do not affect regenerant performance. Typical design values are 10 bed volumes / hr.
= 600 F I- € = 300
0 0 10 20
Bed volumes 30
Flgure 9.8. Ammonium elution with 2% sodium chloride. The concentration of ammonium-N, mgA in regenerant is plotted versus the number of bed volumes
used.
The regeneration cycle is usually followed by back washing with 2-4 bed volumes. The back wash water is mixed with the influent to remove the minor
ammonium present. The regenerant may be recovered by either air or steam stripping. When the ammonia is removed by this process, the clinoptilolite is ready
to be used again for next regeneration cycle.
Cellulose ion exchangers are selective to proteins and offer a possibility for
31 9
protein recovery. The capacity is about 1 meqv / g , but with a low bulk weight , the capacity will only be roughly 0.2 eqv I I . As 50% of this capacity is selective for
proteins and proteins have a high equivalent weight, the capacity on a weight to weight basis is still attractive, although it is strongly dependent on the source of
protein, including how much time the proteins have had to decompose before the
treatment. Due to the slow diffusion rate of proteins, the retention time of waste wa-
ter in the cellulose ion exchanger is required to exceed 12-15 minutes.
Regeneration can be carried out by sodium hydroxide, which expand the cellulose
fibers, whereby the proteins are released. To ensure the presence of a sufficiently high sodium concentration and thereby obtain the sodium-form of the ion exchange
material, elution by a mixture of sodium hydroxide and sodium chloride is
recommended. Proteins dissolved in the elution liquid may be recovered for
instance by precipitation; see also Section 1 1 .l.
Figure 9.6 is also valid for this ion exchange process, for which a column
height of about 1.5 m is recommended.
9.3. The Sequential and Continuous Ion Exchange Operation
The sequential adsorption or ion exchange operation is limited to treatment
of solutions where the solute to be removed is adsorbed relatively strongly when
compared with the remainder of the solution. This is often the case when colloidal
substances are removed from aqueous solutions using carbon, as in the production of process water.
The method for dealing with the spent adsorbent or ion exchanger depends
upon the system under consideration. If the material taken up is valuable (e.g.,
proteins), it might be desorbed by contact with a solvent other than water. If the
removed component is volatile (e.g., ammonia), it may be desorbed by reduction of
the partial pressure of the adsorbate over the solid by passing steam or air over the
solid, i.e., air or steam stripping is applied; see also Section 9.2. In the case of
most sequential operations in the context of waste water treatment, the adsorbate is of no value and it is not easily desorbed. The adsorbent may then be regenerated
by burning off the adsorbate, followed by reactivation.
A mathematical treatment of the sequential operation distinguishes between
single-stage operations, multi-stage cocurrent operations and multi-stage
320
countercurrent operations. The mathematical treatment does not distinguish
adsorption from ion exchange - the basic equations are the same. A schematic flowchart for a single-stage operation is shown in Fig. 9.9.
As the amount of ion exchanger is usually very small compared with the
amount of solution treated and since the solutes to be removed are taken up much
more strongly than the other components present, the up take of the latter may be
ignored. Furthermore, the ion exchanger is generally insoluble in the solution. If the
water (see Fig. 9.9) to be treated contains S kg of unadsorbed substance (water)
then the adsorbable solute concentration is reduced from Yo to Y 1 kg of solute per
kg of solvent.
A : adsorbent or ion exchanger Xo : adsorbate conc. = 0 for fresh absorbent 1
S : solvent Yo : adsor- bate conc.
b S Single stage operation
b Y1 : adsorbate conc. after ad-
Figure 9.9 Flowsheet for the single-stage operation. Application of the mass
conservation principle for the component removed from S to A leads to equation
(9.8).
exchange A X1 : adsorabte conc. after adsorption or
If the adsorbent (ion exchanger) added is A kg, then the solid adsorbate
content increases from Xo to X 1 kg of solid per kg of adsorbent. In most cases fresh adsorbent is used so that Xo = 0. The mass balance of the solid removed is given
by the following equation:
321
S(Y0- YI) = A(X1- XO) (9.8)
This equation gives the so-called operating line, shown in Fig. 9.10 together with the equilibrium curve. This could be either Freundlich's or Langmuir's isotherm or equation (9.7), which may be considered a modified Freundlich's
isotherm.
It is presumed in Fig. 9.9. that the solvent and adsorbent can be separated completely after the ion exchange process, which is not always the case in
practice. The presence of solvent in the used ion exchanger may not to interfere
with the further treatment of the adsorbent or it may be possible to remove the
solvent by drying or other processes. It is under all circumstances important to
consider this problem in the application of adsorption and ion exchange processes in practice.
Yo -
Y1 -
xo x1
(1): Operating line (2): Equilibrium curve
Figure 9.1 0. Operating line and equilibrium curve for a single-stage operation.
322
If sufficient time of contact is allowed, so that equilibrium is almost reached,
the final liquid and solid concentration will correspond to a point (see Q, Fig. 9.10),
which is quite close to the equilibrium curve.
The mass balance assumes that the amount of liquid mechanically retained
with the solid after filtration or settling is negligible. This is usually the case. If Freundlich's isotherm can be used, we can, at the final equilibrium
condition, set up the following equation:
YI = k * Xln (9.9)
Since the adsorbent (ion exchanger) normally used contains no initial
adsorbate, that is Xo = 0, then the two equations yield:
A Yo - Y1 - - -
S (Yl/k)lh (9.10)
As can be seen, this permits analytical calculation of the adsorbent solution
ratio for a given change in solution concentration, provided that the constants in the
equation system are known.
However, removal of a given amount of solutes may be accomplished by less adsorbent, if the solution is treated with separate small batches of ion
exchanger rather than a single large batch. This method is the multi-stage cocurrent operation. The savings are greater the larger the number of batches, but
the expense of equipment and even handling costs will increase with the number
of stages. It is therefore rarely economical to use more than two or three stages. A
schematic flowchart and operating diagram for two ideal stages of cocurrent
adsorption are shown in Fig. 9.11. As seen, the same quantity is treated in each stage, but by two different amounts of adsorbent A1 and A2. The mathematical balances are given by the following equations:
S(Y0 - YI) = AI(X1- XO) (9.1 1)
S ( Y 1 - Y2) = A2(X2- XO) (9.12)
323
These two equations provide the operation lines as shown in Fig. 9.1 2.
When Freundlich's expression is used as a description of the adsorption
isotherm and fresh adsorbent is used in each stage, Xo = 0, the two-stage system
can be computed directly:
A1 b
x o
A1 Stage 1
x1
Figure 9.1 1. Flowchart for a two stages cocurrent operation.
A2 b Stage 2
xo
(9.13)
A2
x 2
or
S Y 2
A2 Y 1 - Y2 - - -
S (Y2k) lh
324
(9.14)
A1 +A2 Yo-Y1 Y1- Y2
S Yl’”1 Y l ’ h + 1 = k l /n ( (9.14)
xo x2 x1
1: operating line 1 2: operating line 2 3: equilibrium curve
Figure 9.12 Operating diagram for two stages cocurrent ion exchange or adsorption.
The minimum total adsorbent is found by setting
d(A1 + A2) = o
dY 1
This reduces to: Y1 1 Yo 1
Y2 n Y1 n ( - ) 1 ” 1 - - * ( - ) = I - -
3 25
(9.15)
(9.16)
Equation (9.16) can be solved for Y I, and the adsorbed quantity can be found by equations (9.1 3) and (9.14).
Even greater economy in the use of adsorbent / ion exchanger can be
achieved by a countercurrent operation. Figure 9.13 shows a diagram of this operation and Fig. 9.14 shows the operation line and equilibrium curve for this
case. The operating line can be set up as follows:
S ( Y 0 - Y2) = A(XO - XI) (9.17)
and if Freundlich’s adsorption isotherm can be used and Xo = 0, then a
combination of this equation and (9.1 7), provides the following expression:
An equation for calculating YI can be found by eliminating SIA:
Yo Y1 Y1 - - 1 = (-) 1h ( -4) Y2 Y2 Y2
(9.18)
(9.19)
It is then possible to calculate S/A directly from (9.17).
If Freundlich’s adsorption isotherm cannot be used, it is of course possible to use the diagram for the necessary calculation as shown in Fig. 9.14.
In the continuous operation the water and the adsorbent / ion exchanger are
in contact throughout the entire process without a periodic separation of the two
phases. The operation can either be carried out in strictly continuous steady-state
fashion by movement of the solid as well as the fluid or in a semi-continuous
fashion characterized by moving fluid but stationary solid, the so-called fixed bed adsorption / ion exchange, which is widely used in waste water treatment, including
by the removal of ammonium and proteins from waste waters. It is generally found
more economical to use a stationary bed for waste water treatments due to the relatively high cost of continuously transporting solid particles. Only this case will therefore be treated mathematically.
3 26
Figure 9.1 3. Flowsheet for a two stages countercurrent adsorption.
The design of a fixed bed ion exchanger and the prediction of the length of the cycle requires knowledge of the percentage approach to saturation at the break
point. Figure 9.15 shows an idealized break-through curve.
Let us consider a case where the flow of water through an ion exchange bed
is S kg/h m* - entering with an initial solute concentration of Yo kg solute / kg
solvent. The total, solute free, effluent after a given time is W kg/m* (see Fig. 9.15).
The break-through curve should be steep and the solute concentration in the effluent rises rapidly from close to zero to that of the incoming water. Some low value YB is arbitrarily chosen as the break-point concentration and the column is
considered exhausted when the effluent concentration has risen to some other
arbitrarily chosen concentration of value YE, close to Yo. The critical values are the
quantity of effluent We and WE (see Fig. 9.15).
327
YO
xo x 2 x 1
1: operating line 1 2: operating line 2 3: equilibrium line
Figure 9.1 4. Operating diagram for two stages countercurrent adsorption.
The effluent accumulated during the occurrence of the break-through curve is:
WA = \IvE - WB (9.20)
The adsorption or ion exchange zone, that part of the bed in which the concentration changes from YB to YE, is considered to have a constant height of ZA m. If we use TA for the time required for the adsorption zone to move its own height down the column after the zone has been established, then:
TA = WA I S (9.21)
328
I Y, -
Figure 9.15 Idealized break-through curve.
Correspondingly we call the time required for the ion exchange zone to establish itself and move out the bed, TE, which can then be calculated from:
WE TE = -
S (9.22)
If we call the height of the entire ion exchange bed, Z(m), and, TF, the time required for formation of the ion exchange zone, we get:
TA
TE - TF z A = z (9.23)
329
The quantity of solid removed from the water in the ion exchange zone from
the break-point to exhaustion is U kg solid I m2. This area is areas 1 and 2 in Fig.
9.15. If all the ion exchanger in the zone was saturated with solute, it would
Consequently at the break-point, the zone is still within the column. The
contain Yo WA kg solute I m*.
fractional ability, f, of the adsorbent in the zone still to adsorb is:
(9.24)
U JWE (Yo - Y) dW f = - - - WB = Jl.0 (1 - Y / Y o ) d ( W )
Yo 'WA Yo 'WA 0.0 WA
If f = 0 it means that the ion exchanger in the zone is saturated, and the time
of formation of the zone at the top of the bed, TF, should be the same as the time
required for the zone to travel a distance equal to its zone height, TA. On the other
hand, if f = 1 .O so that the solid in the zone has essentially not taken up anything of
the component considered, the zone formation should be very short.
These limiting conditions are described by:
TF = (1 - f)Ta (9.25)
Equations (9.23) and (9.25) provide:
TA WA z a = z = z (9.26)
TE - (1 -f)TA WE- (1 - f) WA
The ion exchange column is Z m tall of unit cross sectional area, and contains Z * Q kg adsorbent, where Q is the apparent packed density of the solid in
the bed. If the column was in complete equilibrium and saturated at an ion concentration of XT kg I kg solid, the weight of the component taken up would be Z
* Q XT kg. At the break-point the adsorption zone of height, Za, is at the bottom of
the column, but the rest of the column, Z - ZA (m), is substantially saturated. At the
break-point therefore, the removed amount of the considered component is:
(Z -ZA) Q * XT + * Q f * XT
330
(9.27)
The fractional saturation of the column at the break-point is:
(z - ZA) Q XT + h Q f * XT z - (1 -f)*ZA - - (9.28)
Z * Q * X T Z
In the fixed bed of ion exchange, the active zone moves through the solid in
The operating line of the entire tower is:
the flow direction as we have seen.
S(Yo - 0) = A (XT- 0)
or
(9.29)
(9.30)
Since the operating line passes through (0,O) of Fig. 9.16 at any level in the column, the concentration of solute in the water, Y, and the removed component on the solid, X, are then related by the equation:
S * Y = A * X (9.31)
Over the differential height, dZ, the rate of ion exchange is:
SdY = Kt a (Y - Y+)'dZ (9.32)
where Kt = the overall transfer coefficient, a = the outside surface area of the solid particles and Y+ = the equilibrium concentration.
For the entire ion exchange zone:
YE dY za za
YB Y - Y + Ht S/Kt*a - - - N t = / - - - (9.33)
where Nt = the overall number of transfer units in the ion exchange zone.
331
0 0 5 10
mg ammonium / g clinoptilolite
Figure 9.10. Y* is the operating line, Y-eq the equilibrium curve, Y, is
considered as break-point and the bed is saturated at YE.
The success of this analysis hinges upon the constancy of Ktor Ht for the concentration within the adsorption zone. This will of course depends upon the
relative constancy of the resistance to mass transfer in the fluid and within the
pores of the solid. An alternative method to determine Ht will be described below;
see page295
The ion exchange rate can be limited by external diffusion, internal diffusion
or by the actual ion exchange process. The external diffusion controls the transfer
of solute from the water to the boundary layer of fluid immediately adjacent to the external surface of the ion exchanger. The external diffusion is governed by molecular diffusion and in a turbulent flow by eddy diffusion.
The process can be described by the following equation:
Va = ke a(Y - Y+) (9.34)
where Va = the rate of ion exchange; Y = the concentration of the ion in the fluid
332
and Y+ = the concentration of the ion in the fluid in equilibrium with the existing concentration in the ion exchanger. ke is the external mass transfer coefficient.
Internal diffusion processes control the transfer of solid from the exterior of the ion exchanger to the internal surface. This condition is represented by the
following equation:
Va = k i "a 'z ' (Xx-X) (9.35)
where 1 = the interparticle void ratio; X X = the concentration of the ion in the solid
phase that is assumed to be in equilibrium with the coexisting liquid phase at
concentration, Y; X =the actual concentration of ion in the solid phase.
If the internal and the external diffusions occur at comparable rates the
respective mass transfer coefficients, measured individually, may be added (King,
1 965) :
(9.36)
The diffusion coefficient as used in the design of a practical column must be
found in the literature or by determined experimentation. The internal diffusion can
be found by equilibrium experiments by use of equation (9.34).
Ht may be found alternatively by a series of experiments, where the capacity
(expressed as volume of water, which can be treated with a required efficiency, i.e.,
Y, is given) is found for different flow rates. The values found are expressed as a
percentage of the theoretical capacity, which gives the percentage of the total
column "not used," which is equal to Z-nu = (1-f)*Za in Fig. 9.6. Equation (9.33) is
used to find Nt by graphic integration. f is furthermore found by graphic integration
of Y / Yo versus (W -W), corresponding to equation (9.24). Hence Za can be
found, since (1-f*Za) is known. Finally is Ht determined (see also the example in Appendix C2) as a function of the flow rate from equation (9.33): Ht = Za / Nt and
can be used for design of full-scale columns. The method is published in
Jsrgensen et al. (1978) for the exchange ammonium-sodium on clinoptilolite. Ht as function of the flow rate may furthermore be used for an economic optimization of the ion exchange column. Higher flow rate means that the required column volume
333
is reduced but that the utilization of the column is also reduced, which in turn means, that a more frequent regeneration is required to obtain the same effluent quality. A lower flow rate, on the other hand, means that more ion exchange volume is needed, but the frequency of elution is decreased.
The design of an ion exchange column will be exemplified as mentioned
1. Z-nu is determined in laboratory or pilot scale tests as function of the flow, i.e., for various S-values. 2. Hence (1 -f)*Za is known, see equation (9.28)
3. Nt is determined by graphic integration of equation (9.33).
4. f is determined by graphic integration of equation (9.24)
5. Za is determined from a combination of the results in points 2 and 4.
6. Ht as a function of the flow rate is determined by use of the expression
Ht = Nt / Za. Ht as function of the flow rate can now be used for any design.
above in Appendix C2. The steps to be followed may be summarized in 6 points:
9.4. Application of Ion Exchange
Mercer et al. (1970) has reported a successful application of the specific ion exchanger clinoptilolite for removal of ammonium from municipal waste water. Jsrgensen (1 979) reported the possibilities of recovering ammonium (ammonia) from industrial waste water. It is clear from these examinations that recovery of the regenerant by air stripping seems important, because even the neutral regenerant will cause discharge problems. An economic analysis shows, moreover, that the recovery of the regenerant will in most cases more than pay for the cost of the recovery, as the air stripping of small volumes is relatively moderate in costs as discussed in Section 7.6. As already discussed in Section 7.6 the ammonia removed by stripping should be absorbed in sulfuric acid to avoid air pollution by the released ammonia. This implies that an entire chain of processes: ion exchange, recovery of regenerant and recovery of the air stripped ammonia as ammonium sulfate, must be applied. Figure 9.17 shows a flow chart of the described process chain.
Cellulose anion exchangers have been used for removal of azo-dyes from waste water from the textile industry, as reported by Jargensen (1978) and
334
Gangneux et al. (1976). The removal of azo-dyes is required due to the strong color
of the waste water, rather than to remove the nitrogenous compounds in the waste
water, which are unimportant for the nitrogen balance in the receiving water. Proteins can be recovered from slaughterhouses, fish filleting plant, dairies
and other food processing industries by use of a cellulose cation exchanger. The
method has not found a wide application as chemical precipitations of these waste waters are sufficient to produce an effluent comparable to municipal waste water.
The values of the proteins still in solution after this treatment are hardly able to pay
for recovery of the proteins. It can, however, not be excluded that the process will be of increasing interest in the future due to lack of proteins and due to increasing charges imposed by the water authorities on industrial waste water effluents.
Many industries discharge waste water with high concentrations of
ammonium, as referred to in Section 7.6. Ion exchange is, however, not a very
attractive treatment method for removal of high ammonium concentrations,
because the regeneration becomes more frequent and the operation costs are very
dependent on the elution frequency. As air stripping becomes more attractive the
higher the concentration of ammonium is, these types of industrial waste waters are probably better treated by biological methods or by air stripping at least from an economic’s point of view. Ion exchange is an attractive method particularly for
concentrations up to 100 mgA (Haralambous et al, 1992) and for waste water and
drinking water, which do not contain sufficient organics to allow a biological
treatment. Ion exchange has furthermore been applied for removal of ammonium
from water in recycling aquaculture plants. The advantage is here the low
ammonium concentration, which makes it attractive to use ion exchange to
concentrate the waste product, in this case ammonium, several thousand times. However, for all these applications of clinoptilolite, it is necessary to have sufficient
contact time to allow the intracrystalline diffusion to take place; see Jsrgensen
(1979) and Neveu et al. (1985). Longer contact time brings about a reduced
discrepancy between theoretical and practical capacity. A flow rate of 2-6 bedvolumes / h will in most cases correspond to the optimum contact time.
Ion exchange has been used for removal of nitrate from drinking water; see for instance Dore et al., 1986. They used a strong base ion exchanger, regenerated by sodium chloride and were able to remove as much as almost 1 mole of nitrate
per liter of resin at a low sulfate concentration. The selectivity to nitrate is, however,
335
reduced by increased sulfate concentration.
The standards for nitrate in drinking water (see Section 1.4) are exceeded
for many ground water bore holes. The nitrate can, however, be removed by ion
exchange, but as there is no ion exchanger, that is specific for nitrate take up, nitrate removal by this method is associated with high costs. As the ion exchange process has a high efficiency, the nitrate can easily be removed to a concentration
far below the standards. To reduce the costs it is therefore possible to treat a
fraction of the ground water by ion exchange and then to mix the treated and
untreated water afterwards and still obtain a drinking water, that can meet the
standards.
INFLUENT
Sulfuric acid
Ammonium sulfate
Figure 9.17. A flow chart of a combination of ion exchange, air stripping and absorption. The regenerant is recovered by air stripping, while the ammonia from
the air stripping is utilized to produce a fertilizer, ammonium sulfate.
336
10.1. Principles of Membrane Processes
Membrane separation, electrodialysis, reverse osmosis, ultrafiltration and
other such processes are playing an increasingly important role in waste water
treatment . A membrane is defined as a phase that acts as a barrier between other
phases. It can be a solid, a solvent-swollen gel or even a liquid. The applicability of
a membrane for separation depends on differences in its permeability to different
compounds.
Table 10.1 gives a survey of membrane separation processes and their
principal driving forces, applications and their useful ranges.
Figure 10.1 shows the relation between the membrane permeability and the
size of various impurities in waste water. The selection of membrane process is, as
seen from this figure, a questionof which impurities are required to be removed
from the waste water.
Osmosis is defined as a spontaneous transport of a solvent from a dilute
solution to a concentrated solution across a semi-permeable membrane. At a certain pressure - the so-called osmotic pressure - equilibrium is reached. The
osmotlc pressure varies with concentration and temperature, and depends on
the properties of the solution.
For water, the osmotic pressure is given by:
n TI = - R T
v (10.1)
where n = the number of moles of solute
V = the volume of water
R = thegasconstant
T = the absolute temperature
This equation describes an ideal state and is valid only for dilute solutions.
For more concentrated solutions the equation must be modified by the van’t Hoff
337
factor by using an osmotic pressure coefficient:
n
v n = 0 " - RT (10.2)
For most electrolytes the osmotic pressure coefficient is less than unity and will usually decrease with increasing concentrations. This means that equation (10.1) is usually conservative and predicts a higher pressure than is observed. If the pressure is increased above the osmotic pressure on the solution side of the membrane, as shown in Fig. 10.2, the flow is reversed. The solvent will then pass from the solution into the solvent. This is the basic concept of reverse osmosis. Reverse osmosis can be compared with filtration, as it also involves the moving of liquid from a mixture by passing it through a filter.
Table 10.1
Membrane separation processes
Driving Range (Pm) Function of Process force particle size membrane
Electrodialysis Electrical poten- < 0.1 Selective to tial gradient certain ions
Dialysis Concentration < 0.1 Selective to solute
Reverse osmosis Pressure < 0.05 Selective transport of water and small ions
Ultrafiltration Pressure 5 ' 10-3- 10 Selective to mole- cular size and shape
However, one important difference is that the osmotic pressure, which is very small in ordinary filtration, plays an important role in reverse osmosis. Second, a filter cake with low moisture content cannot be obtained in reverse osmosis,
338
because the osmotic pressure of the solution increases with the removal of solvents. Third, the filter separates a mixture on the basis of size, whereas reverse
osmosis membranes work on the basis of other factors. Reverse osmosis has
sometimes also been termed hyper-filtration to be distinguished from ultrafiltration,
where dissolved ions and other inorganic molecules are not separated.
, PROCESSES: Convent. filtration - Micro filtration -
Ultra filtration 7
Reverse osmosis - Suspended solid -
Macromoelcules, colloids - Bacteria -
lons, inorg. molecules - - 4 -3 -2 -1 0 +1 +2
Log (size pm)
Figure 10.1, Membrane processes and particle size.
The relation between the process and the removable particle size; see
Figure 10.1, indicates the possibilities of using membrane processes for nitrogen
removal. Proteins can accordingly be removed from waste water and waste
339
products such as whey by application of ultrafiltration. This has found a wide use
particularly in the dairy industry. Ammonium and nitrate can be removed at least to
a certain extent by use of reverse osmosis. This application has, however, some shortcomings:
1) The osmotic pressure increases to very high levels due to high
concentrations of inorganic ions in the reject. This implies that the
permeation rate decreases and the required size of the equipment
increases. This means high installation costs. The alternative is to accept
smaller concentrations of the reject, which, however, increases the problem
of reject discharge. 2) It is difficult to avoid a certain clogging of the membranes, although
removal of most organics and all suspended matter reduces the problem.
3) The high pressure needed for the process implies high energy costs and therefore high operation costs.
Membrane technology has developed rapidly during the last decades, and it
cannot be excluded that reverse osmosis will find a much wider application in the neart future for nitrogen removal, too.
Osmotic pressure n
A B
}A,.,
Fig. 10.2. A - illustrates equilibrium. An osmotic pressure appears. B - illustrates the principle of reverse osmosis.
340
10.2. Process Variables
Reverse osmosis The permeate flux, F, through a semipermeable membrane is given by:
Ds * CW V F = (AP - IT)
RTd (1 0.3)
where
DS = the diffusion coefficient
Cw V
AP R = thegasconstant
d =thickness of membrane
= the concentration of water = the molar volume of water
= the driving pressure (see Fig. 10.2)
The equation (10.3) indicates that the water flux is inversely proportional to
the thickness of the membrane. These terms can be combined with the coefficient
of water permeation, Wp, and equation (10.3) reduces to:
F = Wp * (AP-IT)
where
Ds CW V wp =
R T d
(1 0.4)
(10.5)
For the solute flux, Fs, the driving force is almost entirely due to the
concentration gradient across the membrane, which leads to the following equation
(Clark, 1962):
dCi AC'i
dx d - FS = Ds- - Ds- (10.6)
where
341
C'i = the concentration of species, if within the membrane AC'i = concentration difference measured across the membrane
This equation can be restated in terms of the concentration of the solution, Ci, on either side of the membrane, incorporating the so-called distribution
coefficient, Kd, which is a constant for the membranes generally used (Lonsdale et
al., 1965):
ACi
d F s = D s * K d * - = Kp ACi (1 0.7)
where Kp is termed the coefficient of permeability.
Wp and Kp are both characteristics of the particular membrane type.
As seen from equations (10.4) and (10.7), the water flux depends on the net
pressure difference, while the solute flux depends only on the concentration.
Therefore, as the feed water pressure increases, water flow through the membranes increases, while the solute flow is approximately constant.
Consequently the amount and quality of purified water increase as the net driving
pressure is increased, but the quality of the water decreases as the feed water
solute concentration increases, with a constant pressure, because of an increase in
osmotic pressure. As ever more water is extracted from the waste water, the solute
concentration becomes higher and the water flux falls. Figures 10.3 and 10.4
illustrate these relations. The water flux as a function of the water recovery and at a
fixed pressure is shown in Fig. 10.3 for two different salinities. The variation of the water quality with recovery is shown in Fig. 10.4. As can be seen, the water quality
decreases with increasing feed salinity and increasing recovery. This is the
problem touched upon as point 1 on p. 339.
The rejection ratio in reverse osmosis, R, is defined on the basis of the following equation:
Ci - CDi R =
Ci (10.8)
where
Ci = the concentration of the species, i, in the concentrated stream (reject)
Cpi = the concentration of i in the permeate (product).
342
1 .o
X 3 = Q fi
Q
fl Q U
L
c
0.5 I-
c -
0 1 0 20 4 0 60 80 100
Percentage water recovery
Figure 10.3
inorganic components (salts). Water flux related to water recovery for two concentrations of
The rejection ratio is also expressed by the following equation:
-1
R = ( 1 + K D * m D ) Wp (AP - n)
(10.9)
where, Cwp, is the water concentration in the permeate. Notice that Kp Cwp and Wp (AP - TT) must be expressed in the same units. As Wp(AP - n) = F is often
expressed as g or kg / cm2 or m2/ sec. Cwp must be expressed as g / cm3 or kg / m3.
The equations given so far are idealized because a good mixing on the
brine side has been assumed, so that there is no concentration polarization. However, in reality salt concentrations build up at the membrane surface and a concentration gradient is established.
The increased concentration of the membrane surface raises the local osmotic pressure, so reducing the driving force. Concentration polarization is
343
defined as the ratio of the salt concentration at the membrane surface to the salt concentration in the stream.
loo0 f
0 20 40 60 80 1 00 Percentage water recovery
Figure 10.4 Quality of product related to water recovery for two concentrations.
When the concentration adjacent to the membrane surface exceeds a critical value, the flux begins to level off with increasing driving pressure, AP. The flux is then controlled by the membrane permeability as well as the concentration polarization. This is illustrated in Fig. 10.5.
The following differential equation describes concentration polarization:
F * Ci dCi - Ds- = Kp * ACi
c w c dx (1 0.10)
where CWC is the water concentration in the reject.
344
If the membrane is impermeable to the solute, it means that Kp = 0, and
equation (1 0.10) can be integrated to give:
Cim F * d
Cia Cwc* Ds - = exp( )
where
Cim = the concentration of i in the fluid at the membrane surface
Cia = the average concentration of i in the reject.
(10.11)
1 .o
Driving pressure (AP)
Figure 10.5. Curve 1 illustrates the relation between flux and driving pressure by
membrane permeability control, and curve 2 shows the same relation by
membrane permeability- and concentration polarization control.
Equation (1 0.12) indicates the usual relation between flux and concentration
of retained substances. The flux decreases, see equation (1 0.12) with increasing concentration of retained substances:
F = k*In(Ce/Ci) (1 0.12)
345
where k is an overall mass transfer coefficient Cs is the concentration of retained species adjacent to the membrane
Ci is the concentration of the species, i, in the concentrated stream (reject) surface
The following description for polarization in turbulent flow has been developed :
Cim 2F St9n - = 1.333 exp ( Cia
) p 0.75 v f
(10.13)
where V = the mean velocity Sc f
p
= the Schmidt number (the definition; see Section 7.4)
= the fanning friction factor = the specific gravity of the solution.
The concentration polarization is seen to be a function of the ratio, average product flow rate to average brine velocity, the fanning friction factor and the Schmidt number. Since F/(p*v) is almost proportional to recovery, polarization is favored by high recovery. However, high recovery can be maintained at low concentration polarization by recirculating the brine. The concentration polarization can be reduced by increasing the friction factor, so promoting turbulence.
The concentrations of the ions in the waste water, the required concentrations in reject and effluent are the dominating variables in membrane processes. They determine the relation between flux and pressure according to the equations given above.
However, the temperature and pH play an important role in the durability of membranes. This is illustrated for cellulose acetate membranes in Figure 10.6.
Cellulose acetate is not recommended for extreme pH values or high temperatures, but it is widely used due to its moderate costs. At extreme conditions other membranes should be chosen; see Section 10.4. The relationships for all types of
346
membranes between on the one side the durability and on the other side pH and temperature are approximately as illustrated for cellulose acetate in Fig. 10. 6.
1 3 6 9 11
Figure 10.6. Hydrolysis rate of cellulose acetate membrane as function of pH at
two different temperatures. Results from Voss et al. (1966).
Ultraf iltration Both ultrafiltration and reverse osmosis depend on pressure as the driving
force and require a membrane that is permeable to some components and
impermeable to others.
The difference between the two processes is that, while ultrafiltration is
usually used to separate solutes above a molecular weight of 500-2000, which
implies a relatively small osmotic pressure, reverse osmosis is used to remove
material of low molecular weight, which causes a high osmotic pressure. The
polarization is, however, usually greater by ultrafiltration than by reverse osmosis,
because the diffusion constant is two or three orders of magnitudes smaller for the
347
macro-molecules than for inorganic ions.
10.3. Design of the Reverse Osmosis Unit
A reverse osmosis plant consists of a series of modules arranged in parallel.
The design data include recovery, pressure, brine, flow rates, product water quality
and flux maintenance procedure.
To be able to design a reverse osmosis unit one must know the feed water
composition, its temperature and osmotic pressure. The capacity requirements of a
plant are usually based on a certain reject flow rate at a given temperature or, in
the case of waste water treatment, on the feed flow rate.
Based on mass balance for the water as well as the solute, the following equations can be set up:
Qi = Qr+Q ( 1 0.1 4)
Qi * Cii = Qr * Ci + Qp Cip (10.15)
where
Qi
Qr
Qp Cif
Ci
Cip
= flow rate of feed stream = flow rate of reject
= flow rate of permeate
= concentration of i in the feed stream
= concentration of i in the reject
= concentration of i in the permeate
The mean concentration of i, Cia, on the one side of the membrane is given by:
Qr Ci + Qi * Cif Cia =
Cr + Cf (1 0.1 6 )
The water quality in the permeate (product) can be expressed by means of
Cia and the average rejection ratio, R ~ v :
Cip = Ga (1- Rav)
348
( 1 0.1 7)
The average salt rejection is given by:
Kp * Cwp ACi
Wp (AP - n) Cit Rav = 1 - (1 0.18)
This equation can be solved most easily by an iteration. If we assume Cip = 0, we have:
Qt * Cit = Qr Ci (1 0.19)
Qt * Cif Cif
Qt ' Qr 2- R' - Cia = -
where
QP R' = -
Qr
Cip is then estimated:
Cit Cip = (1 - Rav)
2 - R'
(10.20)
(1 0.21 )
(10.22)
A better approximation can be obtained by utilizing the value given by
The minimum free energy requirement is determined (Johnson et at., 1966) equation (1 0.22) as next Cip-value, etc.
by means of:
1
AG = -RT I lnaw * dnw 0
(10.23)
where
aw = the chemical activity of water nw = number of moles of water recovered
349
aw can be calculated from:
haw = Icp*G/55.5
where cp is a coefficient.
(10.24)
10.4. Reverse osmosis system
In constructing a system for reverse osmosis many problems have to be
solved :
1. The system must be designed to give a high liquid flux reducing the
concentration potential.
2. The packaging density must be high to reduce pressure vessel cost.
3. Membrane replacement costs must be minimized.
4. The usually fragile membranes must be supported as they have to sustain
a pressure of 20-100 atm.
Table 10.2
Comparison of the various techniques
Packing Useful Water flux
density pH Easeof NaCl at 40 atm.
Modul concept (m2/m3) range cleaning rejection (m3/m2/day)
~~ ~ ~
Plate and frame 450 2-8 fair verygood 0.5
Large tubes 150 2-8 verygood verygood 0.5
Spiral 750 2-8 good verygood 0.5
Hollow fine fibers 7.5-15* 103 0-12+) fair goodlair 0.05-0.2
+) Pdyamide
Four different system designs have been developed to meet the solution to problem 4. These are the plate and frame technique, large tube technique, spiral wound technique and the hollow fine fiber technique.
350
The various techniques are compared in Table 10.2. The most widely used membrane is the cellulose acetate membrane made by the Loeb-Sourirajan
technique. This membrane is asymmetrical and consists of a thin dense skin of
approximately 0.2 p on an approximately 100 p thick porous support.
Polyamide membranes have also been developed. They are considerably
more resistant to high pH-values, but give a smaller flux. During the last two decades there has been an intense research activity in the development of membranes, resulting in several new types. Cellulose acetate-butyrate resin,
cellulose acetate-methacrylate, polyacryl-acid and cellulose nitrate-acetate, are
among the recently developed membrane materials, which are more resistant to
pH and temperature, but do not reduce the initial fluxes. Several natural materials
could also be of use as membranes and extensive laboratory investigations may
hold promise for the application of such natural membranes soon (Kraus et al.,
1 967).
Table 10.3 gives the characteristics of some widely used types of membranes.
Table 10.3.
Characteristics of membrane material Material pH-stability Chlorine Biological Temp. % Ion 8 8
reslstance reslstance rangeOC paratlon
________________________________________---_----__-_---- Polyamide 4-1 1 not good good < 35 >90
Cellulose acetate 2-8 good notgood <30 90
Polyacrylic acid 2-1 1 fair good c 40 >90
butyrate 2-1 0 good fair < 35 90
branes 2-12 fair good <50 >90
Cell. tri-acetate 4-8 fair fair < 30 90
Cell. acetate-
Combined mem-
As mentioned in Section 10.1 it cannot be excluded that new and better
membranes will be developed in the coming decade, which will make the use of
351
reverse osmosis economically attractive for removal of inorganic nitrogen ions, i.e., ammonium and nitrate. This will have particular interest, where production of drinking water quality from municipal waste water will be needed due to problems of water shortage.
10.5. Application of Reverse Osmosis and Ultrafiltration
EPA has for several years performed experiments to determine the feasibility
of membrane techniques in treatment of municipal waste water. The results can be summarized in the following 5 points (EPA, 1969, Feige and Smith, 1974 and Bilstad, 1989):
1. The flux decreased over a period of 20 days and was then stabilized. 2. The quality of the influent was important for the flux. Chemical precipitation seems to be an appropriate pretreatment to use in this context. 3. It is technically feasible to separate nitrogen and other compounds from the waste water. 4. The major problems are concerned with the material-technology. These problems may be solved in the very near future. 5. It is possible to remove impurities on the membranes chemically to obtain
the same flux as for new membranes.
The results obtained by EPA at the Pomona waste water treatment plant are shown in Table 10.4. The shown results were obtained with the spiral technique
used at a pressure of 31 kg / cm2. The waste water was pretreated on activated carbon.
Similar experiments have been performed in Tokyo, using different types of membrane processes. The aim was to find suitable methods to recover waste water. A final report from these experiments is expected soon, but the provisional results have indicated that it is possible to obtain a certain removal of nitrogen compounds by ultrafiltration.
The widest application of membrane processes for removal of nitrogen compounds from waste water or waste has been the use of ultrafiltration to remove water from whey (rich in proteins) and municipal sludge. Whey was previously used as pig feed, but due to the high dilution, the transportation to the farms
352
became uneconomical. It is, however, possible to obtain protein concentrations 4-6
times higher by ultrafiltration, which reduces the transportation cost correspondingly and makes it again profitable to utilize whey as pig feed. The discharge of nitrogenous material by dairies may thereby be reduced correspondingly.
As seen from this review on the application of membrane processes for the removal of nitrogenous material, the present use is limited, but many waste water
engineers and scientists in the field of membrane processes expect a rapid growth
in the use of these technologies in the very near future. It seems therefore appropriate to include the presentation of membrane processes in a review of nitrogen removal techniques.
Table 10.4.
Results obtained by reverse osmosis after pretreatment of municlpal waste water by biological treatment and activated carbon adsorption. 75% of the water was recovered by the process.
Parameter influent Effluent % separation
353
This Page Intentionally Left Blank
11. PRECIPITATION
11 .I Principles of Precipitation
Precipitation in a strictly chemical sense is the transition of a substance from
the dissolved state to the non-dissolved state upon the addition of other (dissolved) reagents that lead to the formation of precipitates.
However, if chemicals causing precipitation are added to water, other
reactions may also take place such as for instance coagulation. Thus, in any practical application of the precipitation process it is often very difficult to distinguish between these reactions. Only on the basis of a detailed knowledge of
the composition of the (waste-) water matrix is it possible to describe the direction into which the process advances, i.e., which reaction is favored or which reaction is suppressed.
Precipitation is accomplished by a reaction between a specific metal ion and an anion, for instance:
cu2+ + co32- <=> CUCOS (11.1)
In surface water, and in the pore water, there is a predominance of the following anions: chloride, sulfate, carbonate, hydrogen carbonate, hydroxide, and under reducing conditions anionic species derived from hydrogen sulfide. The chlorides and sulfates of the common metals are readily soluble, whereas the carbonates, hydroxides and sulphides only dissolve with difficulty.
Hydroxides precipitate in several forms, which may behave quite differently with respect to the effects of co- precipitation or later redissolution. Precipitates may persist in metastable equilibrium with the solution and may slowly convert into the aged forms, thereby becoming more stable and inactive.
The solubility is highly dependent on pH, as the concentration of the precipitating anions hydroxide, carbonate or sulfide, decreases with decreasing pH due to reaction with the hydrogen ion:
OH- + H+ <a H,O (1 1.2)
355
C032- + H+ <* HC03- (1 1.3)
S, + H+ <=> HS- (1 1.4)
HC03- + H+ <=> H2C03 = H2O + CO,(g) (11.5)
HS- + H+ <=> H2S(g) (11.6)
With increasing pH, first carbonates and then hydroxides become the stable
phase for many metal ions. For negative values of the reduction potential, the
sulfide remains the stable phase over a wide pH range for many metal ions.
The various interacting processes, which determine the solubility at different
pH values can conveniently be illustrated in a graphical double logarithmic
representation; see below.
The concentrations of proteolytic species are characterized by the total
alkalinity A, and pH. The total alkalinity is determined by adding an excess of a standard acid (e.g., 0.1 M), boiling off the carbon dioxide formed and back titrating
to a pH of 6. During this process all the carbonate and hydrogen carbonate are
converted to carbon dioxide which is expelled and all the borate is converted to
boric acid. The amount of acid used (i.e., the acid added minus the base used for
back titration) then corresponds to the alkalinity, Al, and the following equation is
valid:
(1 1.7)
where C = the concentration in moles per liter for the indicated species.
In other words the alkalinity is the concentration of hydrogen ions that can be
taken up by proteolytic species present in the sample examined. Obviously, the
higher the alkalinity, the better the solution is able to maintain a given pH value if acid is added. The buffering capacity and the alkalinity are proportional (see e.g.
Stumm and Morgan, 1981).
Each of the proteolytic species in an aquatic system has an equilibrium
constant. If we consider the acid HA and the dissociation process:
HA <a H+ + A- (11.8)
3 56
we have:
(11.9)
where Ka = the equilibrium constant.
It is possible, when the composition of the aquatic system is known, to
calculate both the alkalinity and the buffering capacity, using the expression for the equilibrium constants. However, these expressions are more conveniently used in
logarithmic form. If we consider the expression for Ka for a weak acid, the general
expression (1 1.9), may be used in a logarithmic form:
[A-I
[HA1 pH = pKa + log __ = pKa + log [Am] - log [HA] (11.10)
multiplying both sides of the equation by -1 and using the symbol p for -log and pH for -log H+.
It is often convenient to plot concentrations of HA and A- versus pH in a
logarithmic diagram. If C denotes the total concentrations C=[HA]+[A-1, we have at
low pH:
[HA] = C (11.11)
log[A'] = pH - pKa + logC (11.12)
This means that log[A-] increases linearly with increasing pH, the slope
being +l. The line goes through (log C, pKa) as pH=pKa gives log[A-] = log C.
Correspondingly, at high pH, [A*] = C and
log[HA] = pKa - pH + logC (11.13)
which implies that log[HA] decreases with increasing pH, the slope being -1. This line also goes through (IogC, pKa).
357
At pH = pKa, [A-] = [HA] = C/2 or log [A-] = log [HA] = log C - 0.3
acid-base system.
dissociation of 2H+:
Table 11.1 and Fig. 11.1 show the result of these considerations for a single
Note that for H2A the slope will be -2 at pH>pK2, corresponding to the
H2A = 2H+ + A2- and for A2- the slope will be +2 at pHepK. This is
0, the buffer capacity, is defined as dC I dpH, where C is the added acid or
It can hence be shown that:
demonstrated in Fig. 11.2.
base in moles of hydrogen or hydroxide ions respectively.
0 [A-I [HA1 log ( -) = log ( [H30+] + [OH-] + 1 1
2.3 C (11.14)
At log pH [HA] = C, and only [H30+] plays a role.
PH -4 -2 0 2 4 6 8 10 12 14 16 18
Figure 11.1, H3O? OH- , HA and A- are plotted versus pH for a weak acid with
pKa = 4.64 and C = 0.1 M.
358
Table 11.1
[HA] and [A-] at various pH-values
<< pKa log c pH - pKa + log C
>> pKa -pH + pKa + log C log c
= pKa log C12 = log C-0.3 log C12 = log C-0.3
-
- 0 D - O 4
Figure 11.2. pH - log C diagram for phosphoric acid.
359
[A-I [HA1 0 At higher pH, also = [A-] contributes to -
C 2.3
0
2.3 where [H3O? = [A'], log ( - ) = log ( 2 [H3O? ) = -pH + 0.3 = log (2 [A-1).
At still higher pH, but with values of pH<pKa, log [A-] dominates.
D At pH>pKa, [A-] = C and log [HA] contributes the most to -
2.3
At very high pH, log [OH-] will dominate. These considerations are used in the
construction of Fig. 11.3
Figure 11.3. Buffering capacity of sea water as function of pH.
360
If one is dealing with waters containing very few components, then it is frequently possible to refer to precipitation processes in a stricter sense. Precipitation is a chemical reaction with a relatively fast rate. Thus, in most instances the rate itself is of no direct concern, and there exist no models to
describe the rate aspects of the process. The application of the process depends to
a large degree upon the equilibrium situation that characterizes this process. Thus, the precipitation process is described by use of the equilibrium or the end-point of the reaction for specified boundary conditions. However, it must be pointed out that for most practical applications such equations derived from thermodynamic principles, have to be modified.
If precipitation is defined as the transformation of two or more dissolved components to a non-dissolved substance, the so-called precipitate, then dissolution processes and precipitation processes are similar reactions but of opposite directions. The solubility of a product, or vice versa the degree to which precipitation will control the dissolved species, is determined by the capacity of the solute to accommodate specific ions. This capacity is controlled by:
- the energy of bonding between the ions under consideration - the dielectric characteristics of the solute - the type and number of ions present in the system
The solubility of certain species or the relationship between two or more precipitating partners is furthermore controlled by third partners which lead to so- called side reactions. The solubility is also determined by temperature and ambient pressure.
The mass law describes the solubility and the corresponding precipitation reaction in terms of a solubility product. As seen from the example below the solubility product, K,, describes the equilibrium concentrations of the precipitating ions, in particular the ion to be removed by precipitation.
[Ca2+] * [C032-] = K * [CaCO,] = Ks (11.15)
The stolchiometrlc relationship describes how many atoms, molecules or ions of one reaction partner react with corresponding forms of the other partner. Using the above example:
361
[Ca2+] = [C03*-] = [CaCOdpmc (11.16)
The reaction rate with which the precipitation occurs, or with which a disturbed solubility J precipitation equilibrium is balanced again, is finite. However, in most instances of interest for the practical application of this process, the reaction rate is so large that the available detention time, or reaction time, suffices to reach the equilibrium. It has been indicated above that either in the stage of mixing of chemicals with the waste water stream, or in the transport of the precipitating system from one reactor (mixing reactor) to the next one (in most cases a reactor for the liquid-solid separation) the flow or detention time is large enough. However, there are situations where a change in the stoichiometric parameters in terms of an overdosing of the precipitation causing reagent leads to improved reaction rates and to increased efficiency. Efficiency in this instance is interpreted as reduced remaining concentrations of the ion to be precipitated.
The stoichiometric parameters for practical purposes can also be formulated as a quotient of concentration values. This is indicated in the example below:
(11.17)
On the basis of thermodynamic arguments this quotient should have a value of 1 in this instance shown above. However, practical observations have indicated that increased quotients may increase the reaction rate and the separation of the precipitate. Usually each precipitation reaction is followed in its practical application by a liquid-solid separation step. Depending upon the specific gravity of the solids formed or upon the amount of solids formed, such separation steps can be sedimentation, flotation or filtration.
It has been indicated that the equilibrium concentrations are a function of ambient pressure and temperature. Similarly the reaction rate is strongly affected by these parameters.
One further variable needs be described or defined: the pH value of the precipitating system which is of utmost importance. In aqueous solutions. The role
of the process variables presented about will be treated in more detail in Section 11.2.
The application of precipitation as waste water treatment process involves a combination of three unit operations:
362
1. Addition of chemicals to obtain a precipitation. The process conditions
determine the stoichiometric coefficient and thereby the amount of chemicals
needed to produce a proper precipitation.
2. Mixing and flocculation of the chemicals to produce flocs, which settle or
flotate readily. 3. A separation process, whereby the precipitated components are removed from the water. It might be performed either by sedimentation, flotation,
centrifugation or filtration.
The first operation has been treated in details above, while the two following
processes are presented below.
Colloidal particles often possess an electrical charge, which creates a repelling force and prevents aggregation. Stabilizing ions are adsorbed to an inner
fixed layer, which gives its particles its electrical charge, the latter varying with the
valence and number of adsorbed ions. Ions of an opposite charge are held near
the surface by electrostatic forces. The psi potential is defined as the gradient between the interface of the colloidal particles and the solution, while the zeta
potential is defined as the gradient between the slipping plane and the solution.
The zeta potential is related to the particle charge and to the thickness of the
double layer. It is not possible to measure the psi potential, but the zeta potential can be determined and expressed.
The zeta potential can be used as an expression for the stability. It is
possible to measure it on the basis of the following equation:
4PP
X * E zeta potential = U
where
E = the dielectric constant of the medium
= the viscosity of the medium
X = the thickness of the double layer
U = the electrophoretic mobility.
(11.18)
The zeta potential is determined by measuring the mobility of the colloidal
363
particles across the electrophoresis cell, viewed through a microscope. Several types of zeta meters are commercially available.
La Mer (1 964) distinguished between two types of particle destabilization: coagulation and flocculation.
According to La Mer, coagulation results from compression of the electric
double layer surrounding the colloids, while flocculation refers to a destabilization
by adsorption of large organic polymers with a subsequent formation of bridges between particles and polymers. These definitions of the two terms - coagulation
and flocculation - are not universally accepted, but are useful because they have a practical significance.
Fig. 11.4 is a schematic presentation of destabilization by flocculation.
Figure 11.4 Destabilization by flocculation.
364
Lawler et al. (1983) has presented a mathematical model describing changes in the particle size (PSD) immediately below the solid/liquid interface in
gravity thickening based upon Brownian motion, fluid shear, and differential
sedimentation. Although the model predicted trends for the coagulation and differential sedimentation for changes in time, solids concentration, particle stability,
and the subsidence velocity at the interface, the model was limited because the
subsidence 'velocity cannot be predicted and a simplified approach to the
hydrodynamics of differential sedimentation was used. Several mathematical
models were developed by Babenkov (1983) to describe the relation between the
density and the size of the flocs formed during coagulation. The characteristics of
the final flocs depended on the size and density of the micro-flocs. Aluminum
sulfate and cationic polymers provided efficient coagulation for coal processing
waste waters.
In many cases agitation is used to accelerate the aggregation of colloidal particles. When particles follow a fluid motion they have different velocities, so that
opportunities exist for interparticle contacts. When a contact between particles is
caused by fluid motion the process is sometimes called ortho-kinetic flocculation
(Overbeck, 1962).
The following equation describes the rate of change in the concentration of
particles:
-2h * Ed3 N2 - -
dN
dt 3 -
where h = collision efficiency factor
G = velocity gradient
N d = diameter of particles
t = time.
-
= concentration of particles (numberhol.)
-
G can be calculated (Camp and Stein, 943) and (Camp,
(11.19)
955) from
3 65
(1 1.20)
where
P V = thevolume
m
- = the power input to the fluid
= the viscosity of the fluid
Agitation will not increase the aggregation rate of particles smaller than about 1 p diameter, whereas particles with a diameter of 1 j~ or more will grow as a
result of fluid motion. Since 1 p particles do not settle well, a flocculation tank to allow aggregation must be included in a treatment system which uses
sedimentation tanks at a later stage to separate solids from water. Flocculation tanks are designed to provide interparticle contact by orthokinetic flocculation.
Design data include selection of velocity gradients, reactor configuration, reactor data and detention time necessary to produce sufficient aggregation. It is difficult to
base the design on equations because such parameters as h and P are almost impossible to measure, and even the velocity gradient G can be difficult to
determine. It is therefore necessary to provide information for design based on laboratory and pilot plant experiments. However, the interpretation of such an experiment is only possible using a mathematical description of the orthokinetic
flocculation.
Suspended matter is removed from water by various separation processes,
including sedimentation or settling. The principles of this process will be presented
here, while other possible separation processes will be touched upon in Section 11.3, dealing with the design problems. Precipitates and coagulates might settle. Settling rates depend on the difference in density between the suspended matter and the water, the size and shape of the matter, the viscosity of water, the
turbulence and velocity of the flow field. In addition, the physiological state of the
phytoplankton cells also plays an important role. In most cases it is not difficult to describe the sedimentation itself, but it is far
more difficult to account for the influence of the hydrological flow pattern. Therefore theoretical approaches based upon physical considerations should almost always
366
be accompanied by measurements of sedimentation rates, either directly or indirectly. This latter determination is often carried out by use of tracers, for instance
by use of isotopes.
Removal by settling is most often described as a first order reaction:
(1 1.21)
where m is the concentration of suspended matter and s is the rate of removal by
sedimentation, s is thereby also the ratio between the settling rate, Vs, and the depth D:
VS
D s = - (11.22)
Discrete settling The settling of a discrete non-flocculating particle in a dilute suspension can be
described by means of classical mechanics. Such a particle is not affected by the presence of other particles, and settling is therefore a function only of the properties of the fluid and the characteristics of the particles. As shown in Fig. 11.5 the particle is affected by three forces: (1) Gravity, Fg; (2) the buoyant force, Fb and (3) the frictional force, Ff.
In accordance with Newton's second law of motion, the following equation can be set up:
dvs
dt m - = F g - F b - F f (1 1.23)
where vs = the linear settling velocity of the particles, m = the mass of the particles
and t =time.
The gravity effect is given by:
Fg = p * V * g (1 1.24)
367
where p = the particle density, V = the particle volume and g = the acceleration due
to gravity.
The buoyant force is:
Fb = p e * V * g
where pe = the fluid density.
(11.25)
Figure 11.5. The settling particle is affected by three forces: The gravity, Fg, the
buoyant force, Fb and the frictional force, Ff.
The frictional force is a function of different particle parameters, such as
roughness, size, shape and velocity of the particle, and of the density and viscosity
of the fluid. It can be described by the following relationship:
Cd A pe v t
2 Ff = (11.26)
368
where Cd = Newton's dimensionless drag coefficient and A = the projected particle
area in the direction of the flow. Cd varies with the Reynolds number.
By substituting the equations (11.24), (11.25) and (11.26) in equation (11.23), an expression for the dynamic behavior of the particles is obtained:
dvs
dt 2
Cd A pe vs2 m - = g(p-P,)V - (11.27)
After an initial transient period the acceleration becomes zero and the velocity is
constant. This velocity can be obtained from equation (1 1.27):
If the particles are spherical and the diameter is d, the V/A is equal to 2/3*d and
equation (6.8) becomes:
(11.29)
Newton's drag coefficient Cd is, as mentioned, a function of the Reynolds number
and of the shape of the particle. The relationship between Cd and the Reynolds number for spheres and cylinders is given in Fig. 11.6.
When the Reynolds number is below 1, the relationship between Cd and Re
can be approximated by Cd = 24/Re, where Re = Reynolds number defined as:
d pe * vs
P
wherep = theviscosity
In this case (1 1.29) conforms with Stokes law:
369
(1 1.30)
Figure 11.6. Experimental variation of the drag coefficient with Reynolds number.
After Fair et al. (1968).
From Fig. 11.6 it can be seen that Cd is approximately constant for turbulent flow in the range for Reynolds number between 1000 and 250,000. For this region
the velocity vs is given by:
V, = 1.82 ( ( (p - p,) " d " g) / pe)lI2 (for spheres only) (1 1.31)
Stokes law can be modified to account for non-spherical flocs by use of an "equivalent radius" and shape factor in the formulation:
(1 1.32)
where
VS
9 R
PP
= settling velocity, lengthhime
= acceleration due to gravity, lengthhim$ = equivalent radius (based on a sphere of equivalent volume),length
= density of the cell, mass/length3
370
PW
W
Fs
The shap
= water density, massllengte
= kinematic viscosity
= shape factor
factor has a value c 1.0 and ccounts for all factors, reducin the settling velocity.
Most nitrogen components are unfortunately readily dissolved in water,
which implies that precipitation cannot be used as an easy solution to the problem
of nitrogen removal, in contrast to phosphorus, which is widely removed from
waste water by the use of chemical precipitation. Nitrogen removal by use of precipitation may, however, be carried out by the following two processes:
M P + NH4+ + HPO4- c=> Mg (NH4) PO4 (s) + H+ (11.33)
Dissolved proteins + precipitants = insoluble proteins (11.34)
Process (11.33) looks at the first glance as a very attractive solution, as phosphate and ammonium are precipitated simultaneously. The stoichiometric ratio between the two components in municipal waste water is however not
favorable for the precipitation. The concentration of phosphate is about 10 mg / I or
0.3 mmol I I , while ammonium is normally present as ammonium-N in a concentration of about 30 mg I I or about 2 mrnolA.
This implies that phosphate must be added to assure a proper precipitation.
This makes the process much more expensive, although the product magnesium-
ammonium-phosphate is an appreciated fertilizer. The process has therefore not been used for removal of nitrogen from municipal waste water except in pilot plant experiments in Bari, Italy. It might be more attractive to utilize the process, where
the stoichiometric ratio between N and P is more favorable, but the ratio in most
industrial waste waters is even more unfavorable than in municipal waste water.
The precipitation of proteins by use of various precipitants such as ligno-
sulfonic acid, iron (111) chloride, calcium hydroxide, glucose-tri-sulfate or just pH-
adjustment has been widely used. The precipitation is carried out at the isoelectric point of the proteins, where the destabilization is most easily performed. As waste
371
water often contains a wide range of proteins, it is not possible to adjust the pH for all proteins at the same time, so that precipitation of proteins never can become
100 O h effective.
11.2. Process Variables
It is possible to play on two variables to optimize the application of
precipitation by nitrogen removal: the stoichiometric coefficient and pH. The
composition of the waste water determines the possibilities of finding a good
solution to a particular waste water problem by the use of these two variables.
The optimum pH for precipitation of magnesium-ammonium-phosphate may
be found by use of double logarithmic diagrams, as presented in Section 11.1. The
method is best illustrated by presentation of a concrete case study. Let us consider
a waste water with the concentration of ammonium at 2mmoll I and of phosphate at
0.3 mmol A, corresponding to municipal waste water. Let us furthermore presume
that we use 0.02 moll1 magnesium for the precipitation. What, under these
circumstances, is the optimum pH? Several processes are interacting: the acid-
base reactions of phosphate. Phosphoric acid has three pKa-values: 2.1, 7.2 and 12.3; see also Fig. 11.2. Ammonium has a pKa-value of 9.25. The solubility product
of magnesium-ammonium-phosphate is 10-12.6. Let us also assume that the ionic
strength is too small to have any significant influence on the equilibrium constants.
Figure 11.7. is a double logarithmic diagram of phosphate and ammonium in the actual concentrations. It can be seen on the diagram that the product of
ammonium and phosphate reaches its highest value at about pH = 10.7, which is
the optimum for the precipitation, when the concentrations of free magnesium ions are accounted for.
Figure 11.8 is constructed from 11.7. The product of the phosphate and
ammonium concentrations are plotted versus pH and on the diagram shows,
where the product exceeds 1 O-10.6, corresponding to a magnesium concentration
of 10 mmol / 1. It is possible to obtain an effective precipitation but the stoichiometric
ratio between phosphate and ammonium must of course be 1 :1 to assure that both
components are readily precipitated. It implies, that if 1.7 mmolll phosphate and 10
372
mmol A magnesium are added, an almost complete removal of ammonium and phosphate is possible.
0 2 4 6 0 10 12 14
-2
-4
-6
-8
-10
- 12
-14
Figure 11.7. Double logarithmic plot of ammonium and phosphate in the concentrations found generally in municipal waste water.
It is not possible theoretically to calculate the optimum condition for precipitation of proteins. It is necessary to make laboratory experiments to arrive at the relationship between removal efficiency on the one hand and pH, the amount
and type of precipitant on the other; see for instance Jargensen (1989). Several precipitants in combinations with at least a few polyflocculants must be tested at 3
or more different pH values. The settling rate is observed and used, as will be
373
shown in Section 11.3, to design the sedimentation unit, while plots of the type shown in Figs. 11.9 -10 are used to determine which precipitant to use, in which amount and ht which pH to obtain the best precipitation. As seen in Fig. 11.9 the
obtained COD of the effluent is plotted versus the amount of precipitant added for three different precipitants. BOD5 or the permanganate number or the total nitrogen concentration could of course also be used.
-9 4 A
A T
n r- 0 d
+ * I
CT 0 4
10
12
14 6 8 10 12 13
PH
Figure 11.8.The product of phosphate and ammonium concentration, taken from Fig. 11.7, is plotted versus pH in a double logarithmic diagram.
The effluent quality at optimum dosage is plotted versus pH for each of the selected precipitants to find the optimum pH; see Fig. 11.10. Here the nitrogen concentration of the effluent is used as quality parameter.
Another question to raise is of course: what is the economic optimum? An answer to this question requires that the relationship between the cost of waste water discharge and the effluent quality is known and can be compared with the cost of the added chemicals (precipitant, polyflocculant) and acid (to obtain the
374
right pH). Recovery of proteins might, on the other hand give an income, which
should be included in the financial calculations on the various alternatives. The
experience gained from such calculations shows that the economic optimum very often is close to the technical optimum, i.8. it pays to utilize the technical
possibilities to obtain the best effluent quality with the lowest nitrogen
concentration.
2 3000 5 cc
L 0
0 0 1000
0 50 100 150 200 250 mg precipitant added pr. liter of waste water
Figure 11.9. COD (mgll) of effluent versus dosage of three different precipitants
used on waste water from a brewery are shown. 1. corresponds to the use of iron 111 chloride at pH = 4.2, 2 the use of glucose-tri-sulfate at the same pH and 3 the use of lignosulphqnic acid at pH = 4.5. Note that the initial COD is 3300 rngll.
375
0 ' 3.5 4.0 4.5
Figure 11.10. The figure gives the quality of the effluent obtained by precipitation
with lignosulphonic acid (lower curve) and glucose-tri-sulfate (upper curve) on
brewery waste water. Optimum dosage of chemicals according to Fig. 11.9 is
presumed.
The temperature influences the solubility of mangesium-ammonium-
phosphate and of proteins. The solubility decreases for both with increasing
temperature, but it is hardly possible to regulate the temperature to obtain better conditions for the precipitation.
The settling rate is also dependent on the temperature,too. Various
expressions have been suggested to describe this relationship:
VS,T = VS,Tr TIT^) (11.35)
where VS,T is the settling rate at the absolute temperature T, and VS,Tr is the settling rate at the absolute reference temperature Tr.
376
Tetra Tech (1 980) uses:
157.5 fs(T) =
0.069T2 - 5.3T + 177.6 (11.36)
where T = temperature in "C and fs(T) is a temperature adjustment function.
0.0 18
0.0 16
> 0.0 14
c .- u) : 0.012 v)
> .-
0.010
0.008 0 10 20 30
Figure 11.1 1 .Viscosity, v, plotted versus temperature. A regression analysis will
show the following relationship: p, = 0.178 / (1 + 0.0337 T + 0.00022*T2)
Straskraba and Gnauck (1 985) suggest another method for considering the
influence of temperature on the settling rate. They use the known relations
between viscosity and density of water on the one hand and the temperature on the other; see Figs 11.1 1 and 11.12.
These relations have been used in the construction of a diagram giving the
settling rate versus the temperature, as it is known that the sedimentation rate, U, is
proportional to da - dw / p,. The larger da, the more strongly pronounced is the temperature dependence. da is the density of the particles, dw the density of water
and p, is the viscosity of water.
377
1.000
0.999
$ 0.998 U
>r -c
'Y, 0.997 C a, CI
0.99 6
0.995 0 10 20 30
T ( O C )
Figure 11.12 Density of water, dw, plotted versus temperature. A regression
analysis will show the following relationship: dw = 0.999879 + 6.02602 *
T3.
11.3. Design of Plants for Precipitation of Nitrogen Compounds
As mentioned in Section 11.1, the application of precipitation requires a
three-step plant. Addition of chemicals is the first step. It requires some sort of
automatic dosage equipment, where the amount of chemicals added to the waste water is determined by either pH, the flow or another parameter, that is feasible to measure and, relates to the quality of the influent.
The design of the flocculation tank can be based on a first order process.
378
The number of particles/ volume, N, is transformed into the volume of
particles per unit volume of suspension:
n d3
6 R = N (1 1.37)
where R =the volume of colloidal particles per unit volume of suspension.
Substitution of equation (1 1.37) into equation (1 1.19) gives:
dN h - - = -4-G R N dt n
(1 1.38)
- a first order reaction.
yields:
Integration of this equation for the boundary conditions N = No at t = 0
N
No n
-4hR 6 * t - - In - (1 1.39)
These considerations allow us to apply the equations for a complete mixed
A complete mixed flow (CMF) reactor is generally designed on the basis of
flow reactor in combination with an equation for a first order reaction.
the following equation; see Fig.ll.13:
VdCi
dt = Q CO - QCi + V * r(Ci) (11.40)
where r(Ci) =the reaction rate.
have:
For steady state conditions, provided the reaction is a first order reaction, we
Q * C o - Q C i - k * C i * V = 0 (1 1.41)
where k = the reaction coefficient. Dividing this equation by Q * Ci, gives:
379
co - - 1 - k * t m = O c1
(1 1.42)
where tm = V/Q, the mean residence time in the complete mixed flow reactor.
The equation can also be written as:
c1 1
CO l + k * t r n - - -
or
tl-n = - ( - - 1 ) k C
(1 1.43)
(1 1.44)
However, there are advantages in applying a number of reactors in series. Let us consider m first order CMF-reactors each with volume, V. A mass balance
identical to the one used for equation (1 1.43) gives for the second tank:
c2 1
c1 1 +k*trn - - - (1 1.45)
where C2 = the effluent concentration from tank 2.
The effluent concentration from reactor 2 can also be expressed in terms of
inflow concentration of the first reactor by multiplying equations (1 1.43) and (1 1.45):
a=(-- )2
co 1 +k*trn
(1 1.46)
In a similar way, the effluent concentration, Crn, from the last reactor in a series of first order CMF-reactors may be expressed in terms of the concentration of the inflow to the very first reactor:
(1 1.47)
The total detention time required to achieve a given reaction will therefore
380
be :
If this consideration is used for the flocculation unit the following equation
can be set up:
n’ m No l lm 1 mtm = - ((-) - )
4nGR Nm (11.49)
Figure 11.13. Complete mixed flow reactor. Flow rate Q, volume of tank V,
concentration in tank C1, and the input concentration is Co.
The third step is the separation of the suspended matter and the clear water phase. Several possibilities are available for this step, as mentioned in Section
11 .l. Centrifugation and filtration are, however, rarely used due to their high costs
for the great amount of waste water which must be treated in most cases. The design of these two operations is therefore not included, while settling and flotation
will be covered in this section.
381
Suspended solid in waste water cannot usually be described as discrete particles. If any of the interacting particles have characteristics that might cause agglomeration, growth of individual particles to larger size is a natural consequence. Hence, the greater the tank depth, the greater is the opportunity for contact among particles and so sedimentation depends on the depth as well as on the properties of the fluid and the particles.
At present there is no satisfactory formulation for predicting the effect of flocculation on the settling rate. Thus flocculent settling requires extensive testing to define the characteristics of the waste water in this respect.
Evaluation of the sedimentation characteristics of flocculent settling can be accomplished by placing a quantity of the waste water in a column similar to the one shown in Fig. 11.14. The diameter of the column must be sufficient to ensure that the edge effect is almost eliminated. The suspension is settled and the concentration of the particles is determined from samples withdrawn at the different sampling points. The fraction of the particles removed at each step is used to construct lines showing equal fraction or equal percentage removal, as illustrated in Fig. 11.15. The lines are named iso-concentration lines; the per cent maximum settling path for the indicated per cent removal.
If the tank has an overflow of v l = H4 / 12, (see Fig. 11.15) all particles having a settling velocity 2 v l will be removed from the tank and particles with a velocity v < v l will be removed in proportion to v / v l . The figure shows that the remaining solid between Ra and Rb has settled with an average velocity of v = H’ / t2, and the solid between Rc and Rd has settled with an average velocity of H” /t2.
An approximation for the total overall removal, R, by the chosen overflow is given by:
R = Rc + H’* (Rb - Rc) I t2”vl + H”* (Ra - Rb) / t r v l (1 1.50)
This approximation can be improved by adding more terms and decreasing the interval between the iso-concentration lines.
382
t H1
r r r r
Figure 11.14. Column with four sampling points for settling tests.
Figure
I 1
0 '
D Time
t 2
11.15. The results of a settling test illustrated with iso-concentration lines.
383
Zone settling of flocculated chemicals suspension occurs when the concentration of solids exceeds approximately 0.5 g/ I The particles form a mass, which settles as blanket with a distinct interface between the settling sludge and the clarified effluent. The interface can be observed in a batch settling test. Initially all the suspension is at a uniform concentration and the height of the interface as 20; see Fig. 11.16, which shows the height of the interface plotted versus time. In the region A-B, settling is hindered, but proceeds at a constant rate. The region B-C shows a transition into the compression zone, represented by C-D. The zones are further illustrated in Fig. 1 1.17.
Height 20
D
Time
Figure 11.16. Height of interface in zone settling as function fo the time.
384
Cbrified zone
Discrete settling zone
Hindered settling zone
Transition zone
Compression zone
Figure 11.17. Illustration of the zones in zone settling.
It is possible to design a continuous clarifier based on the batch test. Two areas must be calculated; A l , the area required for clarification, and A2, the area required for thickening. A1 can be calculated from:
A1 = Q/v, (1 1.51)
where v, is the velocity for hindered settling and Q is the rate of flow through the tank. To find A2 it is necessary to find the relationship between settling rate and the concentration of the sludge. The tangent is drawn at different points of the settling curve and the slope of the tangent indicates the settling rate, v; see Fig. 11.1 8. The corresponding concentration in the sludge is calculated from the following equation :
c = W Z O / Z (1 1.52)
385
where Co is the slurry concentration at the start of the settling, 20 is the total height of the clarifier and 2 is shown in Fig. 11.18. By this equation it is possible to
calculate C, the concentration of suspended solid in the sludge layer, as a function of the settling rate. It is now possible to calculate WS, defined as the weight of solid
in sludge produced per minute per m2:
w, = v / (1 /C - lC,) (1 1.53)
where C, isthe required concentration of suspended solid in the layer. W, is
calculated for values of C, and the minimum value is used to determine the area
necessary for thickening. The area per m3 h, A, is found by dividing the sludge concentration Co by W,, where Co is defined above. It means that:
A = CO/W, (11.54)
height I
Z t - = v
Figure 11.18. Sedimentation curve. 20 is total height. Slope of tangent (0 settling rate) is found as Z / t.
386
It is frequently possible to improve the performance in an existing settling tank by making modifications based on the results of a dispersion test. The addition of stream-deflecting baffles, inflow dividing mechanism and velocity dispersion feed wells may decrease short circuiting and increase efficiency.
Fig. 11.19 illustrates the principle of tube settlers. The design incorporates the use of very small diameter tubes in an attempt to apply the shallow depth principle as suggested by Camp (1 946).
Flow through tubes with a diameter of 5-10 cm offers optimum hydraulic conditions and maximum hydraulic stability. Culp et al. (1968) have reported excellent results using tube settlers with a retention time of less than 10 minutes. The retention time can be calculated according to the following equation:
where
L
S YA = vs (-cos D + 1) (11.55)
Q flow rate
A area of tube settler Y A = - =
L = length of tube
S 0
vs = direct settling rate
= distance between the tubes (the diameter of the tubes) = the angle of the tube to the horizontal (see Fig. 11.19)
As can be seen from this equation, Q/A will increase as 0 decreases. It
should therefore be an advantage to place the tubes as near as possible to
horizontal. However, the horizontal settler is not self-cleaning and must be back- washed. Therefore, the steeply inclined 60" tube settler is more commonly used.
Continuous gravity draining of settled solid might be achieved from tubes inclined at angles between 45 and 60".
The clarifier may be designed as a rectangular or circular tank, and may utilize either center or peripheral feed. The tank can be designed for center sludge withdrawal or for withdrawal over the entire tank bottom.
It is very difficult to design a full-scale sedimentation tank based on settling
experiments, as presented above. Severa: important factors influencing particle behavior in a full-scale operation are neglected in such experiments. Tanks are
387
subject to eddies, currents, wind action, resuspension of sludge, etc. A full-scale
clarifier will therefore show a slightly reduced efficiency compared to settling
experiments, but this can be considered by incorporating a safety factor. The choice of an acceptable safety factor requires experience. The practical factor might vary from 1.5 when the tank is very small, baffled and protected from wind, to
3.0 in the case of a large tank, unbaffled and unprotected from wind. Even with the
use of the safety factor, however, perfect performance should not be expected.
o u t l e t
L sludge out
Figure 11.19. Steeply inclined tube settler.
Flotation is used to remove suspended solid from waste water and to concentrate sludge. Thus flotation offers an alternative to sedimentation, especially
when the waste water contains fat and oils.
Either a portion of the waste water or the clarified effluent is pressurized at 3-
6 atm. When the pressurized water is returned to normal atmospheric pressure in a
flotation unit, air bubbles are created. The air bubbles attach themselves to
particles and the air-solute mixture rises to the surface, where it can be skimmed
off, while the clarified liquid is removed from the bottom of the flotation tank.
Fig. 11.20 shows a flotation system with partial recirculation of the effluent.
Generally it is necessary to estimate the flotation characteristics of the waste water by use of a laboratory flotation cell:
388
1 .The rise of the sludge interface must be measured as afunction of time. 2.The retention time must be varied and the corresponding saturation of
pressurized water determined.
3.The effluent quality must be determined as a function of the airlsolids ratio, Based on such results it is possible to scale up appropriately.
Tank
Air Compressor
Figure 11.20. Flotation unit.
11.4. Application of Nitrogen Removal by Precipitation
Nitrogen removal by precipitation of magnesium-ammonium-phospate has not yet found a full scale application, but it cannot be excluded that the process will
be used in the nearest future for industrial waste water of the right composition to
allow an economical removal of phosphorus and nitrogen at the same time.
Schulze-Rettmer (1991) has examined the process in details and finds that it
is an attractive method to use for nitrogen removal, from a technical as well as from
an economic’s point of view. He calculates that the removal of 1 kg ammonium- N by precipitation as magnesium-ammonium-phosphate, using magnesium oxide and phosphoric acid as chemicals, will cost about 5 US. Dollars. The costs are
389
reduced if the waste water contains significant quantities of phosphate and magnesium. The cost of chemicals is estimated to be 70% of the total costs. This implies that a reduction of the ammonia concentration in municipal waste water from 40 mg ammonium-N / I to 5 mg ammonium- N I I will cost about 25 U.S. cents /
m 3 , which is comparable to nitrification and denitrification. The value of the magnesium ammonium phosphate produced by this process can be estimated as 12 US. cents / kg, considering the purity of the product, compared with 25 US.
cents / kg for the usually applied technical quality. The conclusion from this review of the process by Schulze-Rettmer is that the precipitation of ammonium-N as
magnesium-ammonium-phosphate is economically feasible and should be considered as a serious alternative to other nitrogen removal processes.
Precipitations of proteins have, however, been widely used. A discharge fee for waste water related to the concentrations of pollutants has been introduced in many countries, i.e. the fee is found on basis of BOD5, COD, phosphorus and/or nitrogen concentrations in the effluent. This has provoked many industries and in
particular food industries to introduce a waste water treatment, which is able to
reduce the concentrations of BODS, COD, phosphorus and /or nitrogen to the level
of municipal waste water. The industries are thereby able to reduce their discharge costs considerable. It can be shown that the costs of the treatment including depreciation and interest of the treatment plant often are much lower that the discharge costs, which makes it profitable for the industries to introduce treatment of the effluent.
Recovery of proteins gained by precipitation of industrial waste water is, unfortunately, only accomplished in few industries. Some industries deliver free of charge the protein-rich sludge to meat-bone-meal factories, where the sludge is
treated as other waste, which is the raw material for the production. As it is expected that dumping of any solid waste product will be more and more limited in the future, the use of the sludge from treatment of food processing waste water for
production of animal feed will probably become more and more attractive. The general development seems clear for industrial waste water: from no treatment, to treatment due to high discharge fees and finally to recirculation and recovery of waste products.
Figure 11.21 is a flow diagram of the combination of chemical precipitation and ion exchange used in the treatment of waste water from the food industry
390
(Jsrgensen, 1971, 1973, 1976 and 1978). This process allows recovery of fat, grease and proteins. Table 11.2 gives the analytical data obtained when this process was used on waste water from herring filleting after centrifugation of the raw waste water to recover fish oil. Table 1 1.3 gives the analyses of this process for waste water from an abattoir. For comparison Table 11.13 includes the results obtained by using a biological plastic filter.
It can be concluded from these results that the application of chemical precipitation to waste water from the food processing industry is advantageous to
use to reduce the pollution to or almost to the level of municipal waste water. The process is able to reduce the nitrogen concentration of these types of waste water considerably and can therefore be considered as an attractive method for the removal of nitrogen, although the method is most often selected because of its over-all effect of BOD-5, COD, P and N-reduction. The method is simultaneously a practical method for recovery of proteins and it is expected that this feature of the process will become increasingly important in the coming years.
Precipitant I
Screening I
Flocculation s [ u d g e d
1 Recovery of prote ins(+ grease)
Figure 11.21 Recovery of proteins (+ grease).
391
Table 11.2.
Analytical data of waste water from herring filleting
Raw After cen- After chem. After Cel- waste trif uga- precipi- lulose ion- water tion tation exchanger
1. step 2. step 3. step
BOD5 (mgll) 1 1000 5800 2000 1100
Susp. matter (mgA) 400 170 40 2 N (mg4 180 162 60 23
KMn04 (mgll) 8000 4000 1200 600
Table 11.3
Analysis of waste water from an abattoir (mgh)
After chern. After chem.
Raw gical plastic (glucose sul- and ion ex- water filter fate is used) change
After biolo- precipitation precipitation
BOD5 1500 400 600 50
KMn04 950 350 460 60
Total N 140 42 85 15
HN3-N 20 15 18 2
NO3-N 4 5 4 1
P 45 38 39 1.5
392
APPENDIX of
PART B:
DESIGN
EXAMPLES
Determination of kinetic coefficients
This Page Intentionally Left Blank
APPENDIX B 1. DETERMINATION OF KINETIC COEFFICIENTS k,
Ks, hnaxr Yobs AND K, FROM LABORATORY DATA.
Data are derived from a high-strenght bench-scale mixed activated sludge reactor
without recycle, show the following substrate concentrations.
Table Bl.1
Sample no. so S @ Biomass (X)
m gll m gll d mg VSSII
NH,' NH,'
300 7 3.2 128
300 12 2.0 125
300 20 1.6 130
300 30 1 .o 130
300 40 1.1 120
Problem Formulation
Determine the saturation coefficient K, and the constant k for the data presented
in Table B1 .l.
Solution
Set up a table to determine the coefficients K, and k using the following transformation
of the Monod equation (3.1 1).
(B1.l)
395
Table 8 1 . 2
so - s Biomass (X) 9 Biomass (X) $/(So - S) 1 IS
m gll mg VSS I d /I d (ms/l)-’
293 409,6
288 250,O
280 208,O
270 130,O
260 132,O
0.14
0,08
0,05
0,033
0,025
Plot the term X $ / So - S versus 1/S, as shown in figure B1.l
1.4
1.2 1 .
o.2 0 i 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
1 I S
Figure Bl .1 Plot of X $ I So - S versus 1 I S.
396
From equation Bl.1 the y intercept equals (l/K).
l /k = 0.32 d, k = 3,l d-’
From figure B1.l the slope on the curve equals K, / k. Knowing k, K, can be found to
be 24,O mg/l.
Problem Formulation
Determine the coefficient Yobs and the decay rate K, using the following equation.
Solution
Plot the term l / $ versus (So - S) / $ X.
1
0.9 1
? 0.5 i rn
rn
(B1.2)
rn
O J I
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
(so-S) / 4 x
Figure 81.2 119 versus (So - S) / $I X.
397
The y intercept on Figure 81.2 equals (- Kd) = 0,05 d-'.
The value of the slope of the curve on Figure B1.2 equals the yield factor Yobs.
Yobs = 0,35 d-' / 0,70 d-' = 0,5
Determine the value of the coefficient kax using the following equation:
(B1.3)
Using this equation kax is found to be 1,5 d-'.
398
APPENDIX of PART C: DESIGN EXAMPLES
A Stripping Column An Ion Exchange Column A Reverse Osmosis Unit A Sedimentation Tank
This Page Intentionally Left Blank
APPENDIX C 1. DESIGN OF A STRIPPING COLUMN
Problem Formulation
A stripping column for a 90% removal of ammonia from waste water must be
designed. The maximum flow of water is 10 m3 / h. The ammonium concentration is
80 mg /I. The temperature is 18OC.
Solution
Equation (7.39) is used to find the diameter of the tower. Iw should be 0.08 or
above, but for stripping column it is very difficult to obtain such a high IW value and therefore it will be attempted to select IW = 0.04. It implies that a cross sectional
area of 4 m2 should be used according to equation (7.39), giving L = 2500 kg/h/m2, provided that a is about 60 1/ m. It is the case for 4 inch raschig rings (see Table
7.4), which are chosen. The minimum ratio air to water is about 3000, which is
selected. It will correspond to 30 000 m3 / h air or 36 000 kg / h. It gives a flow rate 9
000 kg I h / m2 or 7 500 m / h, corresponding to about 2 m / s, which is fully acceptable see Table 7.4.
The flooding point is found from equation (7.40), using Figure 7.14. Q is found to be:
Q = (2500 / 9000)* 40.0012 = 0.01
which will give a Z value of about 5. As pL is 0,001 kg / m's and dh3I2 is 0,Ol (see Table 7.4), hnr is therefore 0.05 or slightly more than found above. which is
acceptable.
HtG is found from equation (7.38), as the constants are found in Table 7.3:
401
NB rigtigt symbol??a = 1.8
0 = 0.4
y = 0.4
Sc for air at 15O C can be found from the the viscosity of air (0.0648 kg / m h
) , the diffusion coefficient ( 0.0392 m2 I h) and the specific gravity ( 1.2 kg / m3 ) to be 1.37. HtG 'is now found from equation (7.38):
HtG = 2 ' ( 9000 /2500)0.4 41.37 = 3.9 m.
R is found from equation (7.42).
R = H'300011244
Henry's constant i found from (7.11) to be 0.69 bar. Therefore R = 1.66,
which by use of Fig. 7.16 is translated to 3 transfer units, as the fraction 0.9 is removed.
The height of the tower is calculated to be 3.9' 3 = 11. 7 m.
402
APPENDIX C2. DESIGN OF AN ION EXCHANGE COLUMN
Problem Formulation
Figure C2.1 illustrates the equilibrium data for protein uptake by a cellulose
Waste water with a protein concentration of 200 mg / I is considered. The break-point will be considered as the time at which the effluent has a
protein concentration of 20 mg / I and the bed will be considered exhausted when the effluent has a protein concentration of 180 mgA.
Ht = 0.05 m The depth of the ion exchange bed is 0.5 m. Find Za and the saturation in percentage.
ion exchanger.
Solution
The equilibrium data as indicated above are plotted in Fig. C2.1.. Table C2.1
lists the value of Y on the operating line between Ye and YE, and the corresponding value of Y+.
In Table 7.3 l/(Y-Y+) has been computed. Column 4 in the table is based on Fig. C2.2. and column 5 indicates the corresponding values of (W-WB)MIA.
By means of column 6, which shows YIYo, Fig. C2.3 is plotted. The total number of transferred units is found in Table C2.1 to be 4.23. It is now possible by
use of Fig. C2.3. to find f, as: 0.64.
Za = N * Ht = 4.23 *0.05 = 0.21 m
z- (I - f ) B (0.5 - (1 - 0.64) 0.21) 100
Saturation (oh) = ( )I00 = = 85% Z 0.5
403
Figure C2.1. Equilibrium line and operation line.
404
Figure C2.2. (W - We) / WA = f ( Y / Yo ). f can be found to 0.64.
1 Y-Y' -
0.1
OD5
0 20 E
I
4 I 150 2
1
Y - Yx Figure C2.3. Y = f ( )
405
Table C2.1
Theoretlcal column calculations
1 dY W - W e Y
Y - Y' Y - Y ' WA Yo Y Y* -- -
20 10
30 14
40 20
50 24
60 29
70 33
80 39
90 44
100 49
110 53
120 58
130 64
140 68
150 74
160 80
170 85
180 93
0.100
0.063
0.050
0.038
0.032
0.027
0.024
0.022
0.020
0.01 8
0.01 6
0.01 5
0.014
0.013
0.01 3
0.01 2
0.01 0
0
0.8
1.35
1 .81
2.16
2.47
2.72
2.94
3.14
3.32
3.48
3.63
3.74
3.87
4.00
4.12
4.23
0
0.189
0.31 9
0.428
0.51 0
0.584
0.643
0.695
0.742
0.785
0.823
0.858
0.884
0.915
0.946
0.974
1 .ooo
0.1
0.1 5
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
406
APPENDIX C3. DESIGN OF A REVERSE OSMOSIS UNIT
Problem Formulation
It is desired to produce 1000 m3/24h of potable water (500 mgA dissolved salts) from water containing 3000 mgA of dissolved salts, mainly ammonium salts.
Design a reverse osmosis unit for this job. A membrane is available that has shown
ammonium chloride rejection of 0.95 at 45 atm. pressure. The permeability is 2.5 ' 10-5g /cm* /sec/atm.
Solution
2 3000
58.5 ' 103 n = 0.082'298 = 2.5atm.
cp= 1
QP
Qf To be conservative we use: R' = - = 0.9
Therefore the osmotic pressure of the concentrate is approximately 1 On feed or 25
atm.
Q p = 1000m3/24h
Qf = 1110 m3 /24h
Qr = 111 d / 2 4 h
2Cif 2 ' 3000
2 - R' CP = - (1 - Rav) - (1 - 0.95) = 272 mgA
2 - 0.9
407
Qr Cp - Q p * Cp 1110 3000 - 1000 272 Ci = - - = 27577mg/I
Qr 111
Q Qi + Qf Cp 111 *27577+1110*3000 Cia = - - = 5232mgll
Qr + Qf 1222
Cp = Ca(1-Rav) = 5232(1-0.95) = 262mg/I
F = 2.5" 10-5 (45-2.5) = 1.06 10-3 g/cm2/sec
QP 1000 A = - - - = 1092 m2
F 1.06 * lO-3* 104 * 104* 3600 * 24
It is suggested that 1400 m2 be used to allow for compaction and fouling of
membranes.
408
APPENDIX C4. DESIGN OF A SEDIMENTATION
Problem Formulation
Figure C4.1 shows the results of six different batch settling experiments
(taken from Jorgensen, 1971). Find the area per m3 of waste water for the six different precipitants on basis of a sludge concentration of 20 g I I. Co = 1.1. gA for precipitation with sulfuric acid and 1.4 g / I for precipitation with the other
precipitants.
Solution
Figures C4.2. and C4.3 are constructed from Fig. C4.1 using equations (1 1.52) and (1 1.53). The area is found by the use of equation (1 1.54). The results are summarized in Table C4.1.
Table C4.1.
Calculatlons of areas needed per m30f waste water to obtaln a sludge concentration of 20 g / I
Precipitant Ws-min Co Area CS by add. Chemical
k g / h m 2 g / l m2 settl. g / I g A ........................................................ 1 .Sulfuric acid 0.06 1.1 18.2 48 0.02 2.Aluminurn sulfate 0.15 1.4 9.3 72 0.1
3.Glucose trisulfate 0.15 1.4 9.3 90 0.1
4.Sulfite liquor 0.83 1.4 1.7 78 0.1
5.Lignin sulfonic acid 0.83 1.4 1.7 78 0.1
6. 3+10% azoprotein 1.32 1.4 1.1 102 0.1
409
The results show that precipitant number 6 is far the best due to the fast
settling. The example shows furthermore, the importance of the use of
polyflocculants. The more rapid settling implies that the need for settling area is
reduced significantly.
30 60 90 120 150 rnin
rnl t
1 - 15 30 4 5 60 90
min
Figure C4.1. Settling is plotted versus time for precipitation with six different
precipitants. The number used are explained in Table C4.1.
41 0
. - I 2 3 4 5 6 7 8 9 1011 12 I
Figure C4.2. Settling rate in cm / min. is plotted versus the slurry concentration at the transition layer for the six precipitants. The numbers refer to the precipitants explained in Table C4.1.
-3
C I -
1 2 3 4 5 6 7 8 9 1 0 1 1 1 2
Figure C4.3. Weight of solid produced kg I h m2 ,Ws , for different values of C = concentrations of solid in the transition layer. Numbers see Table C4. 1.
41 1
This Page Intentionally Left Blank
References
Adams, Krenkel and Bingham (1970): Investigations into the reduction of high nitrogen concentrations. Proc. 5th Int. Water Poll. Res. Conf. 1-13.
Aguirre and Gloyna (1967): Nitrification and denitrification in a model waste stabili- zation pond. Univ. Tex. Centr. Res. Water Resource. CRWR-19.
Ahn, K.-H. and Chang, J.4. (1991): Performance evaluation of compact RBC- settling tank system. Wat. Sci. Tech. 2 1467-1476.
Akai, D., Miki, 0 and Ohgaki, S (1983): Nitrification model with an Inhibitory Effect of Sea Water. Ecological Modelling, Vol. 19, pp 189/98.
Ambuhl, H. (1 969): Die neueste Entwicklung der Vierwaldstattersees. Inst. Verein. Theor. angew. Limnologie 17 210-230.
Andersen, K.B. and Poulsen, K.M. (1976): M.Sc. Thesis, Department of sanitary
Anderson, J.H. (1 965): Studies on the oxidation of ammonia by Nirrosomonas.
Eng. Technical Univ. of Denmark.
Biochem. J. 688-698.
Anthonisen, A.C. (1974): The effect of free ammona and free nitrous acid on the nitrification process. pH.D. (Eng) Thesis. Cornell Univ, lthaca New York.
Anthonisen, A.C., Loehr, R.C., Prakasan, T.B.S. and Shinath, E.G. (1976): Inhibition of nitrification by ammonia and nitrous acid. J. Wat. Poll. Contr. Fed. 48 835-851.
Antonie, R.L. (1974): Nitrification of activated sludge effluent with the Bio-Surf process. Paper presented at the Annual Conference of the Ohio Water Pollution Control Association, Toledo, Ohio, USA.
Antonie, R.L. (1976): Fixed Biological Surfaces - Wastewater Treatment, CRC Inc.,
Antoniou, Hamilton, Koopman, Jain, Holloway, Lyberatos and Svoronos (1 990):
Cleveland, Ohio, pp. 60-64.
Effect of temperature and pH on the effective maximum specific growth rate of nitri fying bacteria. Wat. Res. 24 97-101.
Arvin, W. and Harremoes, P. (1990): Concepts and models for biofilm reactor performance. Wat. Sci. Tech. 22 171-192.
Atkinson, B. et al. (1967): Kinetics, mass transfer and organism growth in a biological film reactor. Trans. Inst. Chem. Eng. 45 T257-T264.
41 3
Atkinson, B. and Dauod, I.S. (1 968): The analogy between micro-biological "reac- tions" and heterogenous catalysts. Trans. Inst. Chem. Eng. 46 19.
Atkinson, B. and Fowler, H.W. (1974): The significance of microbial film in fermen- ters. Adv. Biochem. Eng. 3 221.
Audic, J.M., Faup, G.M. et al. (1984): Specific activity of Nitrobacter through attachment on granular media. Wat. Res. 18 745-750.
Audoin, L., Barabe, J.P., Brebion, G. and Huriet, B. (1970): The use of plastic material as a medium for trickling filters treating domestic sewage. 5th Inter- national Water Pollution Research Conference . Pergamon Press, Oxford,
Austin, G.T. (1974): The industrially significant organic chemicals. Chem. Eng. 81 127.
Babenkov, E.D. (1983): Size density characteristics of the flocs of a coagulated suspension. Sov. J. Water Chem. (USSR) 3 (1983) 3. 24. Abstr. WRC Info 10 83-2049.
Badger and Bander0 (1955): Introduction to Chemical Engineering. McGraw Hill, New York, 636 pp.
Balakrishnan, S. (1968): Kinetics of biochemical nitrification and denitrification. Thesis, Univ. of Texas.
Balakrishnan, S. and Erkenfelder, W.W. (1 969): Nitrogen relationship in biological treatment processes. Wat. Res. V0l.S p. 177-1 88.
Balakrishnan, S. and Erkenfelder, W.W. (1 970): Nitrogen removal by modified activated sludge process. J. Sanit. Enging. Dh. Am. SOC. ci;. Engrs. V01.96, p 501-512.
Barnard, J.L. (1974): Cut 0 and N without chemicals. Wat. & Wastes Eng. Z, p.33- 36.
Barrenstein, A., Kramer, U. and Obermann, P. (1986): Underground treatment of nitrate rich groundwater by infiltration with treated wastewater or methane rich natural gas. DVGW-Schriftreihe, Wasser, Frankfurt, Germany. Vo1.116. p.99-116.
Barrith, N.W. (1933): The nitrification process in soils and biological filters. Ann.
Barth, E.F. and Ettinger (1 967): Managing continous flow biological denitrification.
appl. Biol. V01.20, p 165-184.
7.th Int. Water Water Conf. Texas Water Pollut. Contr. Assoc. Univ of Texas, Austin.
414
Bauer, R.C. and Vernon, L.S. (1 973): Reaction of chloramines with active carbon. J. Wat. Poll. Contr. Fed. 45 p.2290-2295.
Contr. Fed. 44 p.1916. Beckman, W.J. et al. (1972): Combined carbon oxidation-nitrification. J. Wat. Poll.
Beg, S.A. and Atiqullah, M. (1983): Interactions of noncompetitive inhibitors on the nitrifica tion process. J. Wat. Poll. Contr. Fed. 55 p.1080-1086.
Beg, S.A. and Atiqullah, M. (1983): Synergism and antagonism of arsenic, chro- mium and fluoride on nitrification process. J. Environ. Sci. Health A18 p.633-650.
Beg, S.A. and Hassan, M.M. (1987): Effect of inhibitors on nitrification in a packed- bed biological flow reactor. Wat. Res. 1 p.191-198.
Beg, S.A., Siddiqi, R.H. and Ilias, S. (1982): Inhibition of nitrification by arsenic, chromium and fluoride. J. Wat. Poll. Contr. Fed. 54 p.482-488.
Benefield, L.D. and Randall, C.W. (1980): Biological Process Design for Wastewater Treatment, Prentice-Hall, Englewood Cliffs, .
Benmoussa, H. et al. (1986): Inhibition study of the nitrification by organic com- pounds. Wat. Res. &I p.1465-1470.
Benmoussa, H., Martin, G., Richard and Leprince (1986): Inhibition of nitrification by heavy metal cations. Wat. Res. 20 p.1333-1339.
Best and Payne (1965): Preliminary enzymatic events in asparagine-dependent denitrification by Psudomonas perfectomarinus. J. Bact. V01.89, p.1051- 1054.
Bilstad, T. (1 989): Membranseparasjon av nitrogen i avlprpsvann, in H. 0degaard (ed.), Fjerning av Nitrogen i Avlprpsvann, Tapir Forlag, , pp. 82-94.
Blum, D.J.W. and Speece, R.E. (1991): "A Database of Chemical Toxicity to Environmental Bacteria and its use in lnterspecies Comparisons and Correlations" Res. J. Wat. Poll. Con. Fed. V01.63, p 198.
Bock E. et al. (1988): Growth of Nifrobacfer in the absence of dissolved oxygen. Water Res. 22 245-250.
Bolin, B. and Cook, R.B. (1983): The Major Biochemical Cycles and Their Interac-
Boiler, M. and Gujer, W. (1986): Nitrification in tertiary trickling filters followed by
tion, Scope 21, Wiley & Sons, New York, .
deep-bed filters. Wat. Res. 20 p.1363-1373.
41 5
Boller, M., Gujer, W. and Nyhuis, G. (1990): Tertiart Rotating Biological Contacter.
Boon, A.G. and Burgess, D.R. (1974): Treatment of crude sewage in two high-rate
Wat. Sci. Tech. Vo1.22, p.89-100.
activated-sludge plants operated in series. Wat Pololut. Control 73 p.382.
Boon, B. and Laudelout, H. (1962): Kinetics of nitrate oxidation by Nitrobacter. Biochem. J. 85 p.440.
Borchardt, J.A. (1 966): Nitrification in the activated sludge process. Univ of Michi-
Bringmann, Kuhn and Wagner (1 959): Modellversuche zur biologischen Stickstoff-
gan. Ann Arbor.
Abgassung aus Klarwassen. Gesundtheitsing V01.80- p.364-367.
Brown and Caldwell (1975): Report on Tertiary Treatment Pilot Plant Studies, prepared for the City of Sunnyvale, California, February 1975.
Buswell, A.M., Shiota, T., Lawrence, N. and Van Meter, I. (1954): Laboratory studies in the kinetics of growth of Nitrosomoms with relation to the nitrification phase of the BOD test. Appl. Microbiol. 2 p.21-25.
California State Water Resources Control Board. (1 974): Tentative Water Quality Control Plan, San Fransisco Bay basin.
Camp, T.R. (1946): Sedimentation and design of settling tanks. Trans. Am. SOC. Chem. Eng. jlJ p.895.
Camp, T.R. (1955): Flocculation and flocculation basins. Trans. Am SOC. Chem. Eng. 120 p.l.
Camp. T.R. and Stein, P.C. (1943): Velocity gradients and internal work in fluid motion. J. Boston SOC. Civ. Eng. 30 p.219.
Carlucci, A.F. and McNally, P.M. (1969): Nitrification by marine bacteria in low concentration of substrate and oxygen. Limnol. Oceanogr. Vo1.u. p 736139.
Charley, R.C., Hooper, D.G. and McLee, A.G. (1980): Nitrification kinetics in activated sludge at various temperatures and dissolved oxygen concentra- tions. Wat.Res. 14 p.1387-1396.
Carlson, (1 971): Nitrogen removal and identification for water quality control. OWRR
Chen, C.W. (1970): Concepts and Utillities of Ecological Model. JSED. Proc. ASCE
Project No A040 Univ. of Washigton. Dept Civil. Engng.
96. NO SA5 pp 1085-1097.
41 6
Chen, G.H., Ozaki and Terashima. (1989): Modelling of the simultaneous removal of organic substances and nitrogen in a biofilm. Wat. Sci. Tech. 21 p.791-804.
Chen, S.-K., Juaw, C.-K. and Cheng, S.-S. (1991): Nitrification and denitrification of high-strength ammonium and nitrite waste water with biofilm reactors. Wat. Sci. Tech. 23 p.1417-1425.
Christensen, M.H. and Harremoes, P. (1 972): Biological denitrification in water treatment. Rep. 72-2. Dep. San. Engng. Tech. Univ. of Denmark
Christensen, M.H. and Harremoes, P. (1977): Biological denitrification of sewage: a literature review. Prog. Wat. Tech. €3415) .
Christensen, M.H., Harremods, P. and Roed Jensen 0. (1977): Combined sludge denitrification of sewage utilizing internal carbon sources. Prog. Wat. Tech. V01.8, p 589-599.
Christensen, M.H. and Harremods, P. (1978): Nitrification and denitrification in wastewater treatment. Chapter 15 in R. Mitchell (ed.), Water Pollution Microbiology, Vol. 2, , pp. 391-414.
Christensen, (1 972): Discussion of the paper: The complete treatment of raw sewage with special emphasis on nitrogen removal. Presented: 6.th IAWPR Conf. Israel.
Christensen, (1 973): Rapport over laboratorieforserg med denitrifikation. Rep. 73-1, Dept. San. Engng., Technical Univ of Denmark (in danish).
Christenson, Rex, Webster and Virgil (1 956): Reduction of nitrate-nitrogen by modified activated sludge. U.S. Atomic Energy Comm., TID-7517.
Claus, G. and Kurtzner, H.J. (1985): Physiology and kinetics of autotrophic denitrifi- cation by Thiobacillus denitrificans. Appl. Microbiology and Biotechnology. Vol. 22, p.283-288.
Clayfield, (1 974): Respiration and denitrification studies on laboratory and works activated sludge. Wat.Poll.Contr. Vo1.73, p.51-76.
Clark, W.E. (1 962): Prediction of ultrafiltration membrane performance. Science 138 148.
Climenhage, D.C. and Stelzig, A. (1 973): Biological process for nitrogen-BOD removal at Maitland Works, Du Pont of Canada Limited. Proc. 24th Ind. Waste Conf.,Toronto. Ontario, ,
Culp, G., Hansen, S. and Richardson, G. (1968): High-rate sedimentation in water treatment works. J. Am. Water Works Ass ., Vol. 60. p. 681.
417
Czaran, E. et al.: (1988) Separation of ammonia from waste water using clinoptiloli- te as ion-exchanger. Nuclear and chemical waste management, 4 p.107- 113.
Das, Khan and Dutta, (1966): Removal of nitrogen from the fertilizer factory effluent by biochemical nitrification and denitrification. Technol. Sindri. Vo1.3, p.41- 45.
Davies, (1973): Isolation of bacteria capable of utilizing methane as a hydrogen donor in the process of denitrification. Water Res. Vol.Z, p.575-579.
Davies, T.R. and Pretorious, W.A. (1 975): Denitrification with a bacterial disk unit. Wat. Res. 2 459-463.
Dawson and Murphy (1 972): The temperature dependency of biological denitrifica- tion. Water Res. V01.6, p.71-83.
Dixon, M. et al. (1964): Enzyme, Academic Press, New York, , pp. 313-359.
Dholakia, Stone and Burchfeld, (1 970): Methonal requirements and temperature effects in wastewater denitrification. Water Poll. Contr. Res. Ser. Cincinati USA.
Doelle, (1 969) Bacterial Metabolism. Academic Press. New York
Dold, P.L. (1991): Incorporation of Biological Excess Phosphorus Removal in a General Activated Sludge Model. Paper Dept. of Civil Engr. and Engr. Mechanics, McMaster Univ. Ontario L8S 4L7.
Dore, M., Simon, P., Deguin, A., and Victot, J. (1986): Removal of nitrate in drinking water by ion exchange-impact on the chemical quality of treated water. Wat. Res. 20 p.221-232.
Downing, A.L. (1968): Factors to be considered in the design of activatetd sludge plants, in E.F. Gloyna and W.W. Eckenfelder, Jr. (eds.), Advances in Water Quality Improvement, University of Texas Press, Austin, , pp. 190.
Downing, A.L. and Knowles, G. (1 966): Population dynamics in biological treatment plants, in Proc. 3rd Int. Conf. Wat. Poll. Res., Series 2, , pp. 117-137.
Downing, A.L., Painter, H.A. and Knowles, G. (1964): Nitrification in the activated sludge process. J. Proc. Inst. Sew. Purif. 63 p.130-153.
Downing, A.L., Tomlinson, T.G. and Truesdale, G.A. (1964): Effect of inhibitors on nitrifica tion in the activated sludge process. J. Proc. Inst. Sew. Purif. p.537.
41 8
Downing, A.L. and Hopwood A.P. (1964): Some observations of the Kinetics of Nitrifying Activated Sludge Plants. Schweiziche Aeitschrift fur Hydrologie. Vo1.26, p 271
Duddles, G.A., Richardson, S.E. and Barth, E.F. (1974): Plastic medium trickling filters for biological nitrogen control. J. Wat. Poll. Contr. Fed. 3 937-946.
Edholm, Hultman, Lowen and Sven Nilsson. (1 970): Ndringsaltreduktion. Fortsatte forsak over denitrifikationskinetik. KTH Publ. 70:1, Stockholm, Sverige (in Swedish)
Ekama, G.A., Marais, G.v R. and Siebritz, I.P. (1984): Biological excess phosporous removal, in Theory, Design and Operation of Nutrient Removal Activated Sludge Processes, Chapter 7, Water Research Commission, Pretoria, South Africa.
Elmalah, S. and Grasmick, A (1985): Mathematical models for biological aerobic fluidized bed reactors, p.525. In Developments in Environmental Modelling 7, Ed. Jsrgensen and Gromiec. Elsevier Science Publications, The Netherland.
Elmghari-Tabib, M., Laplanche, A., Venien, F. and martin, G. (1982): Ozonation of amines in aqueous solutions. Wat. Res. 16 223-229.
Engel, M.S. and Alexander, M. (1958): Growth and autotrophic metabolism of Nitrosomoms europaea. J. Bacteriol. 's p.217.
Engelhart and Haltrich (1 968): Vorarbieten fur die Gemeinschaftsklaranlagte der BASF und der stadt Ludwigshafen an Rhein. Chem.lng.Tech. V01.40, p.275- 279.
EPA (1975): Process Design Manual for Nitrogen Control, Office of Technology Transfer, Washington DC.
Ericsson, Gustavson and Westberg, (1 967): Ntiringsaltreduktion vid avloppsverk. KTH Publ. 675 Stockholm. (in Swedish)
Erkenfelder (1961): Trickling filter design and performance. J. San. Eng. Div.
Fair, G.M., Geyer, J.C. and Okun, D.A. (1968): Water and Waste Engineering,
- 87(SA6) 87.
Wiley 81 Sons, New York, .
Faup, G.-M., Leprince, A. and Pannier. (1982): Biological nitrification in an up flow fixed bed reactor (UFBR). Wat. Sci. Tech. 14 795-810.
Feige, W.A. and Smith, J.M. (1974): Wastewater applications with a tubular reverse osmosis unit. Water (April).
41 9
Fellinger and Pigford, R.L. (1 952): Absorption and Extraction, 2nd ed., McGraw-Hill, New York,
Fetting, J. (1989): Removal of ammonium nitrogen from wastewater by air strip ping - a state-of-the-art report, in H. 0degaard (ed.), Fjerning av Nitrogen i Avlapsvann, Tapir Forlag, pp. 58-77.
Finsen and Sampson (1 959): Denitrification of sewage effluents. Wat. Waste Treat. VOl.1, p. 289-300.
Forster, J.R.M. (1974): Studies on nitrification in marine biological filters. Aquacul- ture V01.4, p 387-397.
Fowlie, P.J.A. and Shannon, E.E. (1973): Utilization of industrial wastes and waste by-products for phosphorus removal: an inventory and assessment. Canada-Ontario Agreement on Great Lakes Water Quality, Research Report No.6.
Fox, M.E. (1973): Rapid gas chromatographic method for the determination of residual methanol in sewage. Environ. Sci. Technol. z p.838-840.
Frick, B.R. and Richard, Y. (1985): Experience with biological denitrification in a full scale drinking water treatment. Vom. Wasser, Vol. 64. p. 145-154.
Fry (1 955): The Nitrogen Metabolism of Microorganisms. Methuen 1 .Edn. London.
Fruhen, M., Christan, E., Gujer, W. and Wanner, 0. (1991): Significance of spatial distribution of microbial species in mixed culture biofilms. Wat.Sci. Tech. V01.23, p. 1365-1 374.
Gameson, A.H. and Robertson, H.B. (1955): J. Appl. Chem. 5 p.503.
Gangneux, A., Wattiez, D. and Marechal, E. (1976): Synthese et etude de cellulo- ses echangenses d’ions leur emploi dans I’epuration. Des eaux residuaires de I’industrie textile - Ill. Eur. Polymer J. 12 p.551.
Gasser, J.A., Chen, C.L. and Miele, R.P. (1974): Fixed film nitrification of secondary effluent. Paper presented at the EED-ASCE Specialty Conference, Penn. State Univ., Pennsylvania. USA.
Gee, C.S., Suidan, M.T. and Pfeffer, J.T. (1990): Modelling of nitrification under substrate-inhibition conditions. J. Environ. Eng. 116 p.18-31.
Germain, J.E. (1966): Economical treatment of domestic waste by plastic-media trickling filters. J. Wat. Poll. Contr. Fed. 2 p.192.
Van Gils, (1964): Bacteriology of activated sludge. Delft Holland
420
Grady, C.P.L., Henze, M., Gujer, W., Matsuo, T. and Marais, G.v.R. (1986): A model for single-sludge wastewater treatment systems. Wat. Sci. Tech. 18 47-61.
Grady, C.P.L. and Lim, H.C., (1 980): in P.N. Cheremisinoff (ed.), Biological Wastewater Treatment: Theory and Applications, Marcel Dekker, New York.
Grasmick, A. (1982): Contribution a la modelisation des reacturs a cellules immobili- seessur support granulaire en couche fixe ou fluidisee. Dr.Sc. Thesis, Toulouse.
Grasmick, A., Elmaleh, S. and Ben Aim, R. (1979): Theorie de I’bpuration par filtration biologique immerge. Water Res. Vo1.13, p. 1137.
Grasmick, A., Elmaleh, S. and Ben Aim, R. (1980): Etude experimentale de la filtration biologique immerge. Water Res. V01.2, p 61 3.
Grasmick, A., Chatib, B., Elmaleh, S. and Ben Aim, R. (1981): Epuration hydrocar- bon6e en couche fluidisbe triphasique. Water Res. Vol.a, p.719.
Guarino, C.F. et a/. (1 980): Upgrating activated-sludge plants using rotary biological contactors. Wat. Pollut. Control. Vo1.79, p.255-267.
Gujer, W. and Boller, M. (1984): Operating experience with plastic media tertiary trickling filters for nitrification. Wat. Sci. Tech. 16 p.201-213.
Gujer, W. and Boller, M. (1986): Design of a nitrifying tertiary trickling filter based on theoretical concepts. Wat. Res. 20 p. 1353-1 362.
Gujer, W. and Boller, M. (1989): A mathematical model for rotating biological contractors. Wat. Sci. Tech. 22 53-73.
Gujer, W. and Jenkins. D. (1974): The Contact Stabilization Process, Oxygen and Nitrogen Mass Balances, University of California, Berkeley, Sanitary Engine- ering Research Laboratory Rep. No. 74-2. .
Gujer, W. and Jenkins. D. (1975): A nitrification model for the contact stabilization activated sludge process. Wat. Res. 9 p.561-566.
Gujer, W. and Wanner, 0. (1989): Modelling mixed population biofilms, in W.G. Charackliss and K.C. Marshall (eds.), Biofilms, Wiley & Sons, New York, .
Gulliecks, H.A. and Cleasby, J.L. (1986): Design of trickling filter nitrification tower J.WPCF 58 p.60-67.
Gundersen, K. (1966): The growth and respiration of Nitrocytis Oceanus at different partial pressure of oxygen. J. Gen. Microbial. Vo1.42, p 387-396.
42 1
Gonec, E. (1 982): Nitrification on Rotating Biological Contactors. Department of Environmental Engineering, Technical Univ. of Denmark. (Report 82-42).
Haag, W.R., Hoigne, R. and Bader, H. (1984): Improved ammonia oxidation by ozone in the presence of bromide ion during water treatment. Wat. Res. 18 p.1125-1128.
Halling-Starensen, 8. and Hjuler, H. (1 992): Simultaneous nitrification and denitrifica- tion with an upflow fixed bed reactor applying clinoptilolite as media. Water Treatment I p.77-88.
Halling-Ssrensen, B. and Hjuler, H (1993): The kinetics of the complete removal using a Biological Aerated Filter with clinoptilolite as matrix. (submitted).
Haltrich and Jager, (1 963): Beobachtungenbei der biologishe Reinigung nitrathalti- ger industrieller Abwasser mit denitrifizierenden Belebtsclamm. Gas Wasserfach, Wasser - Abwasser V01.104, p.347. (in german).
Hamm (1970): Simultane N- und P- Elimination. Z. Wasser- und Abwasserforch. V01.3, p. 102-1 07.
Haralambous, A., Maliou, E. and Malamis, M. (1992): The use of zeolite for ammonium uptake. Wat. Sci. Tech. 25 p.139-145.
Harkness, N. (1966): Bacteria in sewage treatment processes. J. Inst. Sew. Purif. - 33 p.542-557.
Harremoes, P. and Henze Christensen (1 971): Denitrification with methan. Vand V01.2, p.7-11
Harremoes, P. and Riemer, M. (1975): Pilot experiments on down filter denitrifica- tion. Conf. on Nitrogen as a Water Pollutant, Copenhagen.
Harremogs, P. (1975): The significans to pore diffusion to filter denitrification. J. Water Pollut. Control Fed. V01.48, p. 2.
Harremoes, P. (1976): The significance of pore diffusion to filter denitrification. J.
Harremoes, P. (1978): Biofilm kinetics. Chapter 4 in R. Mitchell (ed.), Water
Water Poll. Contr. Fed. &3 p.377-388.
Pollution Microbiology, Vol. 2, , pp. 71-109.
Harremoes, P. (1 982): Criteria for nitrification in fixed film reactors. Wat. Sci. Tech. - 14 p. 167-1 87.
Harris, Cockburn and Anderson, (1 927): Biological and physical properties of activated sludge. Waterworks Vol.66, p.24.
422
Harris, R.F. and Sommers, L.E. (1968): Plate dilution frequency technique for assay of microbial ecology. Appl. Microbiol. p.330-334.
Hassan, M.M. and Beg, S.A. (1988): Effect of Enzyme Inhibition on the performan- ce of Packed-bed Biological Reactor - A Theoretical study. Chem.Eng.Tech- nol. Vol.u, p.50-56.
Haug, R.T. and McCarty, P.L.( 1972): Nitrification with submerged filter. J. Wat. Poll.
Hegemann, W. and Englmann, E. (1 983): Belebungsverfahren mit Schaumstoffkbr-
Contr. Fed. 44 2086-21 02.
pern zur Aufkonzentrierung von Biomasse. GWF-Wasser/Abwasser 124 233-239.
Henze, M. (1986): Nitrate vs oxygen utilization rates in wastewater and activated sludge systems. Wat. Sci. Tech. 18 p.115-122.
Henze, M. (1987): Theories for estimation of the fraction of denitrifiers in combined nitrifying-denitrifying treatment plants. Wat. Res. 21 p. 1521 -1 524.
Henze, M. and Bundgaard, E. (1 981): Bemessung von kombinierten Nitrifikations- und Denitrifikationsanlagen. GWF-Wasser/Abwasser 123 p.240-246.
Henze, M., Gujer, W., Grady, C.P.L., Jr., Marais, G.v.R. and Matsuo, T. (1987): Activated Sludge Model No. 1, Scientific and Technical Report No. 1, IAWP- RC, London.
Henze, M. and HarremoBs, P. (1977): Biological denitrification of sewage. A litera- ture review. Prog. Wat. Tech. 8 p.509-555.
Henze, M., Harremoes, P. and Roed Jensen, 0. (1977): Combined sludge denitrifi- cation of sewage utilizing internal carbon sources. Prog. Wat. Tech. S p.589-599.
Hermann (1962): Stabilization pond as a nitrate-reducing reactor. Proc ASCE J. San. Engng. Div. V01.88, p.1-20.
Hjuler. H. (1 992): Personal communication to B. Halling-Sarrensen.
Hockenbury, M.R. and Grady, C.P.L., Jr. (1977): Inhibition of nitrification effects of selected organic compounds. J. Wat. Poll. Contr. Fed. 49 p.768-777.
Hoek, J.P. van der, Ven, P.J.M. van der and Klapwijk, A. (1988): Combined ion exchangdbiological denitrification for nitrate removal from ground water under different process conditions. Wat. Res. 22 p.679-684.
423
Hofman, T. and Lees, H. (1953): The biochemistry of the nitrifying organisms. The respiration and intermediary metabolism of Nitrosornonas. Biochem. J. 54 p.579.
Hopwood, A.P. and Downing, A.L. (1965): Factors affecting the rate of production and properties of activated sludge in plants treating domestic sewage. J. Proc. Inst. Sew. Purif. 64 p.435-448.
Horstkotte, G.A., Niles, D.G. and Parker, D.S. (1974): Full-scale testing a water
Hosomi, M., Yuhei, I., Matsushige, K. and Sudo, R. (1991): Denitrification of landfill
reclamation system. J. Wat. Poll. Contr. Fed. 46 p.181-197.
leachate by the modified RBC. Wat. Sci. Tech. 23 p.1477-1485.
Huang, C.4. (1973): Kinetics and process factors of nitrification on a biological filter reactor. Ph.D. thesis. State Univ. of New York at Buffalo.
Huang, C.4. and Hopson, N.E. (1974): Nitrification rate in biological processes. J. Environ. Eng. Div. 100 p.409-422.
Hultman (1971): Kinetics fo biological nitrogen removal. KTH Publ. 71% Stockholm.
Hunerberg and Sarfert, (1 967): Versuche zur stickstoffelimination aus dem Berliner abwasser. Gas Wasserfach. Wasser - Abwasser, V01.108. p.966-69 and 1 1 97- 1 205.
Ide, Tohya, Suzuki, Matsuo and Osanai (1972): Removal of nitrogen and phosp- horus in waste water treatment. Pachec 72, Session 16-3
Ishiguro, M., Watanabe, Y. and Masuda, S. (1978): Advanced waste-water treat- ment by rotating biological disk unit. J. Jap. Sewage Works Assoc.
Jansen, J. and Kristensen, G.H. (1 980): Fixed film kinetics. Denitrification in fixed films, Technical Univ. of Denmark, Dept of Sanitary Engineering.
Jasewicz and Porges (1956): Biochemical oxidation of dairy wastes VI. Isolation and study of sludge micro-organisms. Sew. Ind. Wastes Vo1.28, p.1130.
Jenkins, D. and Garrison, W.E. (1968): Control of activated sludge by mean cell residence time. J. Wat. Poll. Contr. Fed. 40 p.1905.
Jenkins, D., Richard, M.G. and Daigger, G.T. (1986): Manual on the Causes and Control of Activated Sludge Bulking and Foaming, Water Research Com- mission, Pretoria.
Jenkins, S.H. (1 969): Nitrification. Water Pollution Control p.610-618.
424
Johnson, W.K. and Schroepfer, G.J. (1 964): Nitrogen removal by nitrification and denitrification. J. Wat. Poll. Contr. Fed. 36 p.1015.
Johnson, J.S., Dresner, L. and Kraus, K.A. (1966): "Hyperfiltration" in Principles of Desalination. Acdemic Press, New York.
Jones (1972): A study on partial nutrient removal system for waste water treatment
Jsrgensen, S.E. (1971): How to treat phenolic waste water. Vatten, 27 p.434.
plants. Water Res. Volg, p.1389-1397.
Jsrgensen, S.E. (1971): Precipitation of proteins in waste water. Vatten 27 p.58-72.
Jsrgensen, S.E. (1 973): The combination precipitation-ion exchange for waste water from the food industry. Vatten 29 p.40.
Jsrgensen, S.E. (1973): Industrial waste water treatment by precipitation and ion exchange, in G. Lindner and K. Nyberg (eds.), Environmental Engineering, D. Reidel Publ. Co., Holland, pp.364.
Res. 9 1187. Jargensen, S.E. (1975): Recovery of ammonia from industrial waste water. Wat.
Jsrgensen, S.E. (1976): A eutrophication model for a lake. Ecol. Modelling 3 p.147-165.
Jsrgensen, S.E. (1 976): Reinigung hauslicher Abwasser durch Kombination eines chemischen Fallungs- und Lonenaustausch Verfahrens. Thesis, Karlsruhe University, .
Jsrgensen, S.E. (1 978): The application of cellulose ion exchanger in industrial waste water management. Water Supply and Management 5 p.90.
Jsrgensen, S.E. (1 988): Fundamentals of Ecological Modelling, 2nd ed., Elsevier, Amsterdam.
Jsrgensen, S.E. and Johnsen, I. (1989): Principles of Environmental Science and Technology, Elsevier, Amsterdam.
Jsrgensen, S.E., Jsrgensen, L.A., Kamp Nielsen, L. and Mejer, H.F. (1981): Parameter estimation in eutrophication. Ecol. Modelling 13 11 1-129.
Jsrgensen, S.E., Kamp-Nielsen, L., Christensen, T., Windolf-Nielsen, J. and Westergaard, B. (1985): Validation of a prognosis based upon a eutrophica- tionmodel. Ecol. Modelling.
Jsrgensen, S.E., Libor, O., Graber, K.L. and Barkacs, K. (1976): Ammonia removal by use of clinoptilolite. Wat. Res. lo p.213-224.
425
Jargensen, S.E., Libor, O., Barkacs, K. and Kuna, L. (1979): Equilibrium and capacity data of clinoptilolite. Wat. Res. 13 p.159-165.
Kaufman, W.J. (1974): Chemical Pollution of Ground Waters. JAWWA, 66, No 3 pp 152-1 59
Kiff, R.J. (1972): The ecology of nitrification and denitrification systems in activated sludge. Wat. Pollut. Control Fed. V01.71, p 475-484.
King (1965): Mass transfer during short surface exposures in counter-current flow.
Kinner N. (1983): A study of the micro-organisms inhibiting the RBC Biofilms during
Ind. Eng. Chem. Fundam., 4c: 125.
various operting conditions. Ph.D. Thesis, University of New Hampshire, Durham, New Hampshire
Knowles, G., Downing, A.L. and Barrett, M.J. (1965): Determination of kinetic constants for nitrifying bacteria in mixed culture with the aid of an electronic computer. J. Gen. Microbiol. 38 p.263-278.
Kohn, J.H. and Kaufman, W.J. (1971): Optimization of Ammonia Removal by Ion Exchange Using Clinoptilolite, Environmental Protection Agency Water Pollution Control Research Series No. 17080 DAR 09/71.
Krittiya Lertpocasombut (1 984): Nitrification Model with Inhibition of Sea Water. Master Thesis, Asian lnstitut of Technology, Bangkok.
Kraus, K.A., Shor, A.J. and Johnson, J.S. (1 967): Hyperfiltration studies. Desalina- tion, Vol 2, p. 243.
Kugaprasatham, S., Nagaoka, H. and Ohgaki, S. (1991): Effects of short-term and long-term changes in hydraulic conditions on nitrifying biofilm. Wat. Sci. Tech. 23 1487-1 494.
Kurt, M., Dunn, I.J. and Bourne, J.R. (1987): Biological denitrification of drinking water using autotrophic organisms with hydrogen in a fludizied-bed biofilm reactor. Biotechnology Bioengineering. Vol29, p. 493-501.
Kvalstofomsaetningen i dansk landbrugsjord - en evalueringsrapport, Statens Jordbrugs- og Veterinaervidenskabelige ForskningsrBd, 1985.
La Cour Jansen, J. and Henze H. (1990): Rensningsanlaeg ti1 Nitrifikation, Kapitel 6 i Spildevandsrensning; Biologisk og Kemisk. Polyteknisk Forlag, (in danish).
La Mer, V.K. (1964): Coagulation symposium introduction. J. Coll. Sci. 19 p.291.
La Motta, E.J. (1976): Internal diffusion and reaction in biological films. Environ. Sci. Tech. 10 p.765-769.
426
Lang, H. (1981): Nitrification in biological treatment with the aid of "Bio-2-SCHLA- MM" processing. Wasserwirthschaft 71 p.166-169 (in German).
Lawler, D.F. et al. (1983): Particles in thickening mathematical model. J. Environ. Eng. Div., Proc. Am. SOC. Civ. Eng. 109 2, p.332.
Lawrence, A.W. and McCarty, P.L. (1968): Unified basis for biological treatment design and operation. J. San. En$. Div., Proc. Am. SOC. civil Eng. 96(SA3) p.757.
Lee, Y.W. et al. (1988): Kinetics of low solids bio-denitrificaiton of water supplies. J. Water Poll. Contr. Fed. 60(10) .
Lees, H. (1954): The biochemistry of the nitrifying bacteria. Symp. Autotrophic
Lenhard, (1 969): The determination of nitrotase activity in denitrifying systems.
Microorganism. Cambridge Univ. Press London pp 84-98.
Hydrobiologia Vo1.33, p.186-192.
Lewis, W.K. and Whitman, W.G. (1924): Principles af gas absorption. Ind. Eng. Chem. 16 p.1215.
Lonsdale, H. Merten, U. and Ripley, R.L. (1965): Transport properties of cellulose acatate osmotic membranes. J. Appl. Polymer Sci. 9 p.1341.
Loveless, J.E. and Painter, H.A. (1968): The influence of metal ions and pH value on the growth of a Nitrosomonas strain isolated from activated sludge. J. Gen. Microbiol. 52 p.1-14.
Ludzack, F.J. and Ettinger, M.B. (1 962): Controlling operation to minimize activated
Marais, G.v.R. and Ekema, G.A. (1976): The activated sludge process part I -
sludge effluent nitrogen. J. Wat. Poll. Contr. Fed. 34 p.920-931.
Stady-state behaviour. Water S.A. Vol.2 p. 163-200.
Masuda, S., Watanake, Y. and Ishiguro, M. (1987): Japan Sewage Works Ass. VO1.24, p. 19-31.
Masuda, S., Watanake, Y. and Ishiguro, M. (1991): Biofilm properties and simulta neous nitrification and denitrification in aerobic rotating biological contractor. Wat. Sci. Tech. 23 p.1355-1363.
Mateles, R.I., Ryv, D.Y. and Yasuda, T. (1965): Measurement of unsteady-state growth rates of microorganisms. Nature Vo1.2J&. p 263-265.
Matschb, N.(1971): The elimination of nitrogen in the treatment plant of Vienna- Blumental. Water Res. VoI.6, p. 485-486.
427
Matson, J.V. and Charakilis, W.G. (1 976): Diffusion into Microbial Aggregates, Dept. of civil. Eng., Univ. of Huston.
McCarty, P.J: (1969): Feasibility of the denitrification process for removal of nitrate- nitrogen from agriculture drainage waters. Appendix. Calif.Dept.Wat.Res.- Bull. p 174-3.
McCarty, P.J., Beck and Amant (1969): Biological denitrification of waste waters by addition of organic materials. 24.th Ind. Waste Conf. Purdue Univ. p.1271- 1285.
McCarty, P.J. (1976): Energetics and bacterial growth, in Organic compounds in Aquatic Environments, Marcel Dekker, New York, pp. 635.
McLaren, J.R. et al. (1 973): Factors affecting ammonia removal by clinoptilolite. J. Environ. Eng. Div. (August).
McHarness, D.D., Hang, R.T. and McCarty, P.L. (1975): Field studies of nitrification with submerged filters. J. Wat. Pollut. Control Fed. Vo1.47, p 291-309.
Mechala, Allan and Matyskiela (1 970): A study of nitrification and denitrification. Water Poll. Control Res. Ser. USA
Meiring and Stander, (1964). Extended aeration in mining sanitation. Civ. Engr. Sth. Afr. V01.6, p. 19-27.
Mercer, B.W., Ames, L.L., Touhill, C.J., Van Slyke, W.J. and Dean, R.B. (1972): Ammonia removal from secondary effluents by selective ion exchange. J. Wat. Poll. Contr. Fed. 42(2, pt. 2) R95.
Meyerhof, 0. (191 6): Untersuchungen uber den Atmungsvorgang Nitrifizierenden Bakterien I. Die Atmung des Nitratbildners Pflugers. Arch. Ges. Physiol. 164 p.353.
Meyerhof, 0.: (1 917) Untersuchungen uber den Atmungsvorgang Nitrifizierenden Bakterien IV. Die Atmung des Nitritbildners (und ihre Beeinflussung durch chemische Substanzen). Pflugers Arch. Ges. Physiol. 166 p.240.
Metcalf and Eddy Inc. (1 973): Nitrification and Denitrification Facilities: Wastewater Treatment 33 pp. Technol. Transfer Seminar Publ. USEPA Washington D.C.
Metcalf and Eddy Inc.(l979): Wastewater Engineering, Treatment Disposal and Reuse, 2nd ed., McGraw-Hill, New York.
Metcalf and Eddy Inc. (1 991): Wastewater Engineering, Treatment Disposal and Reuse, 3nd ed., McGraw-Hill, New York.
428
Michaelis, L. and Menten, M.L. (1931): Die Kinetic der Invertinwiekung. Biochem. Z. - 49 p.333-369.
Mitsch, W.J. and Jsrgensen, S.E. (eds.) (1989): Ecological Engineering, An Introduction to Ecotechnology, Wiley & Sons, New York,.
Molstad, M.C., McKinney, J.F. and Abbey, R.G.: (1943) Trans. Am. Inst. Chem.
Monod, J.: (1949) The growth of bacterial cultures. Annu. Rev. Microbiol. 3
Eng. 39 p.605.
p. 371-394.
Montgomery, J.M.: (1985) Water Treatment - Principles and Design, Wiley Inter- science, New York.
Monteith, H.D. et al (1 980): Industrial waste carbon source for biological denitrifica- tion. Prog. Wat. Tech. 2 p.127-141.
Moore, S.F. (1969): An investigation on the effects of residence time on anaerobic bacterial denitrification. Ms.Thesis. Univ. Calif.
Moore, S.F. and Schroeder, E.D.: (1970) An investigation of the effects of residence time of anaerobic bacterial denitrification. Wat. Res. 4 p.685-694.
Morris and Jackson: (1 953) Absorption Towers, McGraw-Hill, New York.
Morris, J.C.(1953): Kinetics Reactions between Aqueous Chlorine and Nitrogen Compounds, 4th Rudolphs Res. Conf., Rutgers University, June 15-1 8,
Mudrack, (1971): Die anwendung der mikrobiellen Denitrifikation zur biologishen reinigung von industrieabwasser. Gas Wasserfach, Wasser - Abwasser. V01.112, p.33. (in german)
Mulbarger, M.C. (1971): Nitrification and denitrification in activated sludge systems. J. Wat. Poll. Contr. Fed. 43 p.2059-2070.
Murphy, K.L and Dawson (1972): The temperature Dependency of Biological Denitrification. Water Res. Vo1.6, pp 71 -83.
Murphy, K.L. and Sutton, P.M.: (1954) Pilot scale studies on biological denitrifica-
Neufeld, R.D., Greenfield, J. and Rieder, B. (1986) Temperature, cyanide and phenolic nitrification inhibition. Water Res. V01.20, p.633-642
tion. Prog. Wat. Tech.
Neveu, A., Gaspard, M., Blanchard, G. and Martin, G. (1985): lntracrystallilne self- diffusion of ions in clinoptilolite ammonia and sodium cations studies. Wat. Res. 19 p.611-618.
429
Newton, D. and Wilson, T.E. (1973): Oxygen Nitrification Process at Tampa. The Center for Research in Waste Resources.
NPO-redega relsen, Miljastyrelsen, 1984.
Ohgaki, S. and Wantawin, C. (1989): Nitrification, chapter 13 in Mathematical Submodels in Water Quality Systems. Edited by S.E. Jargensen and M.J. Gromiec. Elsevier Sciense Publications.
Okey and Albertson (1989): Diffusion's role in regulating rate and marking tempera- ture effects in fixed-film nitrification. J. Wat. Poll. Contr. Fed. 61 4 p.500-519.
Osborn (1 965): Operating Experience with Double Filtration in Johannesburg. J. lnst of Sew. Purif. Part 3. pp. 272-281.
Overbeck, J.T.G.: (1962) Kinetics of flocculation. Chapter VII in H.R. Kruyt (ed.), Colloid Science, Vol. I, Irreversible Systems, Elsevier, Amsterdam.
Painter, H.A.: (1970) A review of literature on inorganic nitrogen metabolism in micro-organisms. Wat Res. 4 p.393-450.
Painter, H.A.: (1 977) Microbial transformations of inorganic nitrogen. Prog. Wat. Tech. 8(4/5) p.3-29.
Painter, H.A. and Jones, K.: (1963) The use of the wide-bore dropping-mercury electrode for the determination of rates of oxygen uptake and of oxidation of ammonia by micro-organisms. J. Appl. Bacteriol. 29 p.471-483.
Painter, H.A. and Loveless, J.E.: (1981) Effect of temperature and pH value on the growth-rate constants of nitrifying bacteria in the activated-sludge process. Wat.Res. 17 p.237-248.
Park, Y. and Hegeman, G. (1984): The oxidation of carbon monoxide by bacteria "Microbial Chemoautotroph" W.R. Stohl and O.H. Tuovien eds. Ohio State Univ. Columbus Ohio.
Parker, D., Aberley and Cadwell, (1 975): Development and implementation of biological denitrification for two large plants. Presented at Conf. on nitrogen as a Water Pollut. Copenhagen.
Parker, D., Lutz, Dahl and Berkopf: (1989) Enhancing reaction rates in nitrifying trickling filters through biofilm control. J. Wat. Poll. Contr. Fed. 61 p.618-626.
Parker, D.S., Lutz, M.B. and Partt, A.M.: (1990) New trickling filter application in the USA. Wat. Sci. Tech. 22 p.215-226.
430
Parker, D.S. and Merrild, D.T.: (1984) Effect of plastic media configuration on trickling filter performance.
Parker, D.S. and Richards T. (1 986): Nitrification in trickling filters. J. Wat. Poll.
Pasveer, (1 965): Beitrag uber Stickstoffbeseitigung aus Abwassern. Munchner
Contr. Fed. 58, p. 896-902.
Beitrage zur Abwasser-, Fisherei- und Flussbiolgie (ed. Liebmann). Bd. 12 p. 197-200.
Payne, W.J. (1 973): Reduction of nitrogenous oxides by microorganisms. Bacteriol. Rev .3J p. 409-452.
Poduska, R.A. and Andrew, J.F. (1975): Dynamics of nitrification in the activated sludge process. J. Wat. Poll. Contr. Fed. 47 p.2599-2619.
Polprasert, C. et al.: (1983) Bacterial die-off kinetics in waste stabilization ponds. J. Wat. Poll. Contr. Fed. 55 p.285-296.
Polprasert, C. and Hoang, L.: (1983) Kinetics of bacteria and bacteriophages in anaerobic filters. J. Wat. Poll. Contr. Fed. 55 p.385-391.
Polprasert, C. and Park, H.S. (1 986): Effluent denitrification with anaerobic filters. Wat. Res. a p.1015-1021.
Pokallus (1 963): Toxicity of nitrite to Nifrosomonas Europea. Ph.D. Thesis. Rutgers Univ. New Brauswick, New Jersey.
Porra and Lascelles (1 965): Haemoprotein and heamsynthesis in facultative photosynthesis and denitrifying bacteria. Biochem. J. Vo1.94, p. 120.
Praksam, T.B.S. and Loehr, R.C. (1 972): Microbial nitrification and denitrification in concentrated wasts. Water Res. Vol.fi, p 859-869.
Praksam, T.B.S., Joo, Y.D., Srinath, E.G. and Loehr, R.C. (1974): Nitrogen removal from a concentrated waste by nitrification and denitrification. Proc. 29th Int. Waste Conf. Purdue Univ. pp 497-509.
Pressley, T.A., Bishop, D.F. and Roan, S.G. (1970): Nitrogen Removal by Break- point Chlorination, report prepared for the Environmental Protection Agency, September.
Pressley, T.A., Bishop, D.F., Pinto, A.P. and Cassel, A.F. (1973):Ammonia-Nitrogen Removal by Breakpoint Chlorination, report prepared for the Environmental Protection Agency, Contract No. 14-12-818, February.
Pretorius (1972): The complete treatment of raw sewage with special emphesis on nitrogen removal. Presented at 6.th Int. Conf. Wat. Poll. Res. Jerusalem.
43 1
Process Design Manual for Nitrogen Control, Office of the Technology Transfer, Environmental Protection Agency (EPA), Washington, D.C., 1975.
Randall C.W. and Buth (1984): Nitrite build-up in activated sludge resulting from temperature effects. J, Water Pollut. Control Fed. Vo1.56. p. 1039.
Reimann, H.: (1 990) The Linpor-process for nitrification and denitrification. Wat. Sci.
Renner (1970): Production of nitric oxide and nitrous oxide during denitrification by
Tech. 22 p.297-298.
Corynebacferium nephridii. J. Bact. Vol.lol. p. 821 -826.
Reverse osmosis renovation of municipal wastewater. Water pollution control research series, Ord-17040EFQ12, EPA, 1969.
Richards, T. (1988): Personal Communication to D.S. Parker. City of Atlanta, Georgia Univ.
Riemer, M. and Harremoes, P.: (1 978) Multi-component diffusion in denitrifying
Rimer, A.E. and Woodward, R.L.: (1 972) Two-stage activated sludge pilot-plant
biofilms. Prog. Wat. Tech. lO(5) p.149-165.
operations at Fitchburg, Massachusetts. J. Wat. Poll. Contr. Fed. 44 p.101.
Ritchard eta/. (1980): Denitrification of water for human comsumption. Prog. Wat. Tech. V01.2, pl73-191.
Rittmann, B.E. and McCarty, P.L.: (1 978) Variable-order model of bacterial-film kinetics. J. Environ. Eng. Div. 104(EE5) p.889.
Rittmann, B.E. and McCarty, P.L. (1981): Substrate flux into biofilms at any thick- ness. J. of the Env. Eng. Div. ASCE.
Robertson, L.A. and Kuenen, J.G. (1984): Aerobic denitrification: a controversy
Rogella, F. and Payaraudeau, M. (1 987): Tertiary nitrification with fixed biomass
revived. Arch. Microbiol. J3J p.351-354.
reactors. Paper presented at the Specialized Conference on Nitrogen Pollution of Water, IAWPRC, Brussels, Belgium.
Rogella, F., Bacquet, G, Payaraudeau, M., Bourbigot, M.M., Sibony, J. and Gilles, P. (1 988): Nutrient removal with biological aerated filters. Paper presented at the WPCF Conference, Dallas, 2-6 oktober 1988.
Rogella, F., Ravarini, Larminat and Coutella: (1990) Large-scale biological nitrate and ammonia removal. J. IWEM 4 p.319.
432
Rozich, A.F. and Castens, D.J.: (1986) Inhibition kinetics of nitrification of conti- nuous-flow reactors. J. Wat. Poll. Contr. Fed. 59 p.220-226.
Ruffer (1964): Nitrifikation und Denitrifikation bei der Abwasserreinigung. Vom. Wasser. Vol.3l- p.134-152.
Rusten, B. and Odegaard, H.: (1983) Nitrogen removal in an anoxidaerobic biodisc system. Proc. EWPCA-IAWPRC seminar on Rotating Biological Discs, Fellbach, BDR.
Sato, H., Leung, S.W., and Schutor, J.L.: (1988) Toxic response of Nitrosomonas europae to copper in inorganic medium and wastewater. Wat. Res. 22 p.1117-1127.
Sawyer, C.N. and McCarty, P.L.: (1978) Chemistry for Environmental Engineering, 3rd ed., McGraw-Hill, New York.
Scavia, D. (1980): Conceptual model of phosphorus cycling. In Nutrient Cycling in the Great Lakes, ed. D Scavia and R. Moll, pp 119-140. Special Report 83, Great lakes Research Division, University of Michigan, Ann Arbor.
Wat. Sci. Tech. V01.20. (4/5) p. 177-187. Schlegel, S. (1 988): The use of submerged biological filters for nitrification.
Schroeder and Busch, (1968): The role of nitrate nitrogen in bio-oxidation. J.WPCF vo1.a p.445-459
Schulze-Rettmer, R. (1991): The simultaneous chemical precipitation of ammonium and phosphate in the form of magnesium-ammonium-phosphate. Wat. Sci. Tech. 23 p.659-667.
wassern. Fortschr. Wasserchem. Grenzg. V01.12, p. 124-1 38. Schuster (1 970): Zum gegenwastingen stand der stickstoffelelimination aus ab-
Semmens, M.J., Booth, A.C. and Tauxe, G.W.: (1978) Clinoptilolite column ammo- nia removal model. J. Environ. Eng. Div. 104(EE2) p.231.
Semmens, M. et al.: (1 980)Modelling ammonium exchange and regeneration on clinoptilolite. Wat. Res. 15 p.655-666.
Shammas, N.K.: (1986) Interaction of temperature, pH and biomass on the intrifica- tion process. J. WPCE a ( 1 ) p. 52-59.
Sharma, 6. and Ahlert, R.C. (1976): Nitrification and nitrogen removal. Wat. Res. 11 p.897-923.
Sharma, B. (1977): Nitrification in a continous stirred tank biofilm reactor. Ph.D. Thesis, Rutgers University, New Brunswick, New Jersey.
433
Sherwood, T.K. and Holloway, F.A.L.: (1940) Trans. Am. Chem. Eng. 36 p.39.
Siegrist, H. and Gujer, W. (1987): Demonstration of mass transfer and pH effects in
SJVF, Statens Jordbrugs- og Veterinservidenskabelige forskningsrdd (1 985)
a nitrifying biofilm. Wat. Res. 21 p.1481-1487.
Kvaelstofomsatningen i dansk landbrug - en evalueringsrapport, (in danish).
Skinner, F.A. and Walker, N.: (1961) Growth of Nitrosornonas europea in batch and continuous culture. Arch. Mikrobiol. 38 p.339.
Slater J.H. and Bull A.T. (1978): Interactions between microbial populations. In Companion to Microbiology, p. 181 (ed. A.T. Bull and P.M. Meadow). Longmans, London.
waste water by columnar denitrification. Envir. Sci. Tech. Vol. g p.260-267. Smith, Masse, Feige and Kamphake, (1972): Nitrogen removal from municipal
Srinath, E.G. and Loehr, R.C.: (1974) Ammonia desorption by diffused aeration. J. Wat. Poll. Contr. Fed. 46 p.1939-1957.
Srna, R. and Baggaley, A. (1975): Kinetic response of perturbed marine nitrification systems. J. Wat. Poll. Contr. Fed. 47 p.472-486
Stamberg, J.B., Hais, A.B., Bishop, D.F. and Heidman, J.A. (1974): Nitrification in
Standard Methods for the Examination of Water and Wastewater, 16th ed., Ameri-
oxygen activated sludge. Unpublished paper, USEPA.
can Public Health Association (APHA)-AWWA-WPCF, Washington, D.C., 1985.
Stasuik, W.N., Hetling, L.J. and Shuster, W.W. (1973): Removal of Ammonia Nitrogen by Breakpoint Chlorination Using an Activated Carbon Catalyst, New York State Department of Environmental Conservation Technical Paper No. 26.
Stensel, (1 971) Biological kinetics of the suspended growth denitrification process. Ph.D. Thesis, Cornell University.
Stensel, Loehr and Lawrence (1 973): Biological Kinetics of suspended growth denitrification. WPCF Vol.45- p.399-410.
Stenstram, M.K. and Poduska, R.A. (1980): The effect of Dissolved Oxygen Concentration on Nitrification. Water Research, Vol. 14. pp 643-649.
Stover, E.L. and Kincannon, D.F. (1976): One - versus two stage nitrification in the activated sludge process. J. Wat. Pollut. Control. Fed. Vela p 645-651.
434
Straskraba, M. and Gnauck, A.H. (1 985): Freshwater ecosystems, Modelling and simulation, in Developments in Environmental Modelling, Elsevier, Am- sterdam.
Stratton, F.E. and McCarty, P.L. (1967): Prediction of nitrification effects on dis- solved oxygen balance of streams. Env. Sci. Technol. Vol.1, p. 405-510.
Stumm, W. and Morgan, J.J. (1981): Aquatic Chemistry, Wiley & Sons, New York.
Suhr, L.G. and Kepple, L.: (1974) Design of a Selective Ion Exchange System for Ammonia Removal, presented at the ASCE Environmental Engineering Division Conference, Pennsylvania State University.
Summary of Design Information on Rotating Biological Contractors, Report from U.S. EPA, 1984.
Sutton, K.L. Murphy, K.L. and Dawson, R.N. (1980): Low-temperature biological denitrification of wastewater. J. Wat. Poll. Contr. Fed. 47 p.122-134.
Sutton, P.M., Murphy, K.L. and Dawson, R.N. (1974): Continous biological denitrifi- cation of wastewater, Water Pollution Control Directorate, Environment Canada, Ottawa, Ontario, Environmental Protection Service Report No. EPS 4- WP-74-6.
Sutton, P.M., Murphy, K.L. and Jank, B.E. (1977): Nitrogen control: A basis for design with activated sludge systems. Prog. Wat. Tech. 8 p.467-481.
Suzuki, I., Dular, U. and Kwok, S.C. (1974): Ammonia or ammonium ions as substrate for oxidation by Nitmomonas Europea cells and extracts. J. Bact. V O l . U , p. 556-558.
Tanaiguchi (1961): Comparative biochemistry of nitrate metabolisme. Z. allg. Mikrobiel Vol.l, p.341-375.
Tanaka, K., Tada, M., Ito, M. and Shimizu, N. (1991): Performance analysis of the RBC process based on particle fraction and improvement of final effluent quality. Wat. Sci. Tech. 23 p.1457-1466.
J A W W A B p.47. Taras, M.J. (1953): Effect of free residual chlorine on nitrogen compounds in water.
Tentative Water Quality Control Plan, San Francisco Bay Basin, California State Water Resources Control Board, 1974.
Tetra Tech. Inc. (1980): Methodology for Evaluation of Multiple Power Plant Cooling System Effects, Volume V. Report EPRl EA - 1111. Lafayette, California.
435
Timmermans, P. and Van Haute, A. (1983): Denitrification with methanol. Funda- mental study of the growth and denitrification capacity of Hyphomicrobium sp. Water Res. V01.17, p.1249-1256.
Toit and Davies (1 973): Denitrification studies with lab. scale continous flow-units. Water Res. Vol.3 p.489-500.
Tood and Nuner (1973): Comparison of two techniques for assesing denitrification in terrestrial ecosystems. Bull. Ecol. Res. Comm. Stock. Vol.z, p.277-278.
Tomlinson, T.G. and Snaddon, D.H.M. (1966): Biological oxidation of sewage by films microorganisms. Air Wat. Poll. Inst. J. 10 p.865-881.
Tomlinson, T.G., Boon, A.G. and Trotman, C.N.A. (1 966): Inhibition of nitrification in the activated sludge process of sewage disposal eller treatment. J. Appl. Bacteriol. 29 p.266-291.
USPHS, Division of Water Supply and Pollution Control (1965) Interation of Heavy Metals and Biological Sewage Treatment Processes.
Visut lam-A-Ram (1985): Effect of Sodium Ion on Nitrification. Master Thesis, Asian Institute of Technology.
Wanner, J. and Grau, P. (1 989): Identification of filamentous microorganisms from activated sludge: A compromise between wishes, need and possibilities. Water Res. 23 p.883-891.
Wanner, J., Miroslav, K. and Grau, P. (1990): An innovative technology for up- grading nutrient removal activated sludge plants. Wat. Sci. Tech. 2(7/8) 9-20.
Wanner, 0. (1 989): Modelling dynamics, in W.G. Charackliss and P. Wilderer (eds.), Structure and Function of Biofilms, Dahlem Konferenzen, Wiley & Sons, Chichester.
Wanner, 0. and Gujer, W. (1986): A multispecies biofilm model biotechnologi and
Wanner, 0. and Gujer, W.A. (1984): Competition in biofilms. Wat. Sci. Tech. 17
bioengineering. 28 p.314-328.
27-44.
Watanabe, Bravo and Nishidome (1 982): Simulation of nitrification and its dynamics in a rotating biological contractor. Wat. Sci. Tech. 14 p.811-832.
Watanabe, Y. and Ishiguro, M. (1978): Denitrification kinetics in a submerged rotating biological disk unit. Prog. Wat. Tech. 10 p.187-195.
436
Watanabe, Y., Ishiguro, M. and Nisaidome, K. (1980): Nitrification kinetics in a rotating biological disk reactor. Prog. Wat. Tech. 2 p.233-251.
Watanabe, Y. (1 985): Mathematica Modelling of nitrification and denitrification in RBC, p.419 in Jsrgensen and Gromiec, Developments in Environmental Modelling 7. Elsevier.
Watanabe, Y., Masuda, S. and Ishiguro, M. (1992): Simultaneous nitrification and denitrification in Micro-aerobic Biofilms. WatSci. Tech. V01.26, p.511-522.
Water Pollution Control Federation (1 983): Nutrient Control, Manual of Practise FD- 7.
Weber, W.J. and Rumer J.C. (1 965): Intraparticle transport of sulfonates alkylben- zenes in a porous solid. Diffusion with nonlinear adsorption. Wat Resources Reseach No 1.
Wentzel, M.C., Ekama, G.A. and Marais, G.v.R. (1991): Kinetics of nitrification denitrification. Biological excess phosphorus removal systems - A review. Wat. Sci. Tech. 23 p.555-565.
Weston, R.F. (1 984): Advanced Waste Treatment Performance Evaluation Sum- mary Report USEPA Contract No 68-03-3019.
Wild, H.E., Sawyer, C.N. and McMohan T.C. (1971): Factors affecting nitrification kinetics. J. Wat. Poll. Contr. Fed. 43 p.1845-1854.
Williams, D.R. (1 978): Denitrifying Pseudomonas aeruginosa: some parameters of growth and active transport. Appl. Environ. Microbiol. 36 p.257-263.
Williamson, K. and McCarty, P.L.(1976): A model of substrate utilization by bacterial films. J. Wat. Poll. Contr. Fed. 48 p.9.
Williamson, K. and McCarty, P.L. (1976): Verification studies of the biofilm model for bacterial substrate utilization. J. Wat. Poll. Contr. Fed. 48 p.231.
Williamson, K.J. and Chung, T.H. (1975): Dual limitation of substrate utilization kinetics within bacterial films, presented at 49th Natl. Meet. Am. Inst. Chem. Eng.,Houston, Texas.
Wilson, T.E, and Newton, D.: (1973) Brewery wastes as a carbon source for denitrification at Tampa, Florida. Proc. 28th Ind. Waste Conf., Purdue University, West Lafayette, Indiana, pp. 138.
Wong-Chong, G.M. and Loehr, R.C. (1975): The kinetics of microbial nitrification. Wat. Res. 9 p. 1099-1 106.
437
Voss, K.D., Burris, F.O. jr. and Riley, R.L. (1966): Kinetic study of hydrolysis of cellulose acetate in the pH range 2-10. J. Appl. Poly. Sci., 10: p.825.
Wuhrmann, K. (1960): Effect of oxygen on biochemical reactions. Proc. 3.th Conf. Biol. Waste Treatm. Manhatten College, N.Y.
Wuhrmann, K. (1 964): Grundlagen fur die Dimensionierung der Beluftung bei Belebt-schlammanlagen. Schweiz. Z. Hydrol. 26 p.310.
Yull-Rhee, G. and Frutts, G.W. (1 978): Wastewater denitrification with one carbon compounds as energy source. J. Water Poll. Contr. Fed. vol. 50, p. 21 11- 21 19.
Zanoni, A.E. (1969): Secondary Effluent Deoxygenation at Different Temperatures. J.Water Pollution Control Fed., Vol. 41. pp. 640-59.
0degaard, H. (1988): OversiM over metoder for fjerning af nitrogen i avlspsvann, in H.0degaard (ed.), Fjerning av Nitrogen i Avlspsvann, Tapir Forlag, pp. 17- 35.
438
Index
abattoir: 392 absorption: 292 acclimatized culture: 89-90 activated alumina: 31 3 activated carbon: 295,300,302,353 activated sludge process: 22,44-45,51- 52,8 1,85,236 active biomass: 156-157, activity coefficient: 262-264 adsorption: 295,300,305,320,328 aerated lagoons: 248-251 aeration tank volume: 238 aerator: 245 air stripping: 26 1 - 194, see also stripping alkalinity: 88-90 amino acids: 16-1 8,299 ammonia load: 61 ammonium: 8,17,19-21,23-24,35,43,55- 57,66-68,81,88,92,105,110,114,154, 18 1-1 84,209,2 16,246,255,261-266,291- 293,298-302,314-322.332,353-372,392 anilines: 1 10 aquaculture: 335 arsenic: 1 16,118 assimilatory reduction: 124 attached growth: 22,27,43-45,53,78, 153-234
break-point: 21,23,34 breakpoint chlorination: 295-303 bromide: 302 bubble aeration system: 273 bubble-diff user: 245 buffering capacity: 358-360 buoyant force: 367
C/N ratio: 127-1 33 capital costs: 301 carbon concentration: 142-1 44 carbon oxidation: 29 carbon source: 229,134- 135 cell residence time: 254 cell shape: 57 cell size: 57 cell-residence time: 239 cellulose ion exchanger: 334-335 cellulose resin: 351 chemical methods: 21 chemical precipitation: 3 12,355-392 chloramine: 295 chlorination: 21,23,34 chlorine: 23 chlorine resistance: 351 chromatographic techniques: 122 chromium: 1 16,118 classification: 43 clay: 305
bacterial assimilation: 30 Bardenpho process: 249 batch culture: 68,8590-91 Berl saddles: 281,285,287-288 biocarbone: 50,52,215,229-230 biochemical pathways: 56-58 biofilm: 43,71,137,153-169 biofilm controlled: 52,186 biofilm reactor: 154-169 biofilm submodel: 198-199 biofilm surface reaction: 204 biofilm theory: 22,154-169 bioflocculation: 170 biofor: 215,229-230 biological removal: 21-22
BODWKN ratio: 236 break through curve: 314,329
BOD5: 39,183- 1 85,274,3 1 3,390,392
cli nopti lolite : 23,24, 2 1 8-232,305,3 1 5- 3 18,332,335 CMF reactor: 379 cocurrent operation: 324-326 COD/N ratio: 96 COD: 40,2 19,375,390, coke plant: 1 17 collision efficiency factor: 365 colloids: 339 combinations of several limiting factors: 93-94, 147- 149 competition: 97-100 complete mixed flow: 379-381 contact stabilization: 248-251 continuous ion exchange: 320 copper: 116 costs: see treatment costs counter current operation: 290,327-331 cross-linking: 307
439
cyanide: 1 10,115 friction factor: 346 frictional force: 367
decay coefficient: 240 decay rate: 105 definition: 39-40 denitrification: 11,21-22,28,29,31,43-
denitrification efficiency: 130 denitrification rate: 119-151 density-temperature relationship: 378 destabilization: 364,371 dialysis: 337-339 diffusion coefficient: 270-27 diffusion, liquid film: 163 diffusion resistance: 160,163,204 dinitrogen oxide: 119-120 discrete settling: 367 disinfection: 302 dissimilation: 122 dissimilatory reduction: 124 dry deposition: 10- 1 1 ecological models: 3 electrodialysis: 24-25,337-339 electron transport: 124 elution liquid: 293 enzymatic activity: 122 enzyme inhibition: 104-107 enzymes: 105,108 eutrophication: 8,12 eutrophication models: 15 extended aeration: 248-251
45,5042.1 19-151,216-234
facultative organisms: 120 fertilizer: 8,12-15 fertilizer industry: 21 1 Ficks law: 270 filamentous organisms: 237,242-244 first order kinetics: ,17,62-65,379 fixed bed reactors: 44-45,117,153, 21 6 fixed-film reactor: 193 flocculation tank: 378 flocculation: 364-367 flooding point: 284,286 flotation: 388-389 fluidized bed: 4432,153,166 fluoride: 1 18 frame: 350 free energy: 57-60
gas transfer: 270-275 generation time: 57 glucose-tri-sulfate: 371,375 grease: 391 ground water: 119 growth rate: 66,70
half order kinetics: 64,160 half saturation constant:95, see also saturation constant health hazard: 3,19 Henry's constant: 271,277 herring filleting: 392 heterotrophic bacteria: 98-100,203 hollow fibers: 350 hydraulic load: 174-1 79.186.188,
hydraulic retention time: 254 hydraulic stability: 387 hydrolysis rate: 347
192,224-225
incineration: 291 inhibition models: 109 inhibition types of: 104 inhibition: 66,68,85,91-92,104,150-151 inhibitors: 58,102-1 18 ion exchange: 21,23-24,234 ion exchange: 293, 305-335 ion-selective: 44 ionic strength: 262-264 irrigation: 10
kinetic constants: 138 kjeldahl nitrogen: 236
Lake Tahoe: 267,292 LD50: 55 leather industry: 21 1 lignosulfonic acid: 376 limiting factor: 13-14,16 Lineweaver-Burk Plot: 65,104,107 linpor: 52 loading criteria: 237 Ludzack-Ettinger configuration: 245,247 -248
440
magnesium-ammonium-phosphate:
mass balance for a biofilm: 165-169 mass balance of SND: 224 matrix: 153 maximum growth rate: 57,62,71,77- 80,98,103 membrane processes: 21,24-26,36, 337-339 methanogenesis: 159,164 methemoglobinemia: 19 Michaelis constant: 67 Michaelis-Menten kinetics:, 65-68,see also Monod kinetics microaerophilic layer microfiltration: 339 model of activated sludge: 256 Model of NTF: 177-179 Monod equation: 97-99 Monod kinetics: 62-70 Monod model: 62 MPN-techniques: 122 multi-stage operation: 320
374,389-390
nitrate; 4,8,17,21,35,55,114,120,123-
nitrification: 2 1 -22,27,29,32,43-4530-
nitrification-rate: 43,50,71-83,188-
nitrifying trickling filter: 170-192 nitrite toxicity: 55 nitrite: 17,23,56,64,66,71,81,92,123-126 nitro-compounds: 1 10- 1 1 Nitrobacter winogradskyi: 91 Nitrobacter: 4335-60,62,66-68,71-
255 nitrogen cycle: 4-9 nitrogen fixation: 11 Nitrosolobus: 55-56 Nitrosomonas: 43,55-60,62,66-68,71-
255 Nitrosospira: 55-56 Nitrosovibrio: 55-56 Nocardia genus: 244
127,149,154,218-232,302,336,392
52, 55-118
191,216-234
73,77-79,90-91,166,203,218-232,253-
73,77-79.90-91,166,203,218-232,253-
operating line: 322-328 organic loading: 154- 157 orthokinetic flocculation: 366 osmotic pressure: 25,337-341,348 oxidation ditch: 251 oxidation pond: 49 oxidation rate: 66 oxygen concentration: 16,84-87 oxygen concentration, influence on nitrification:
oxygen consumption: 60 oxygen profile: 158 oxygen requirements: 241 oxygen transfer: 172,244 oxygen, influence on denitrification: 139 ozonation: 302
84-87
packed bed reactor: 52 packed tower: 277 percolate: 21 1 percolating filter: 85 permeability: 342-345 permeate: 348-350 person equivalent: 53 pH effect in biofilm: 165 pH influence: 87-93 pH optimum: 57,87-93 pH-logC diagrams: 356-360 phenolic compounds: 110 phosphorus: 13-1 5 physical methods: 21 plastic filter: 391 plastic foam particles: 50 plastic media: 117,170,186,391 plate: 350
polyflocculant: 373-375 Pomona: 352 porosity: 217 potable water: 18 precipitation: 2 1,25,37,3 12,355-392 predator control: 192 Pretoria: 292-294 primary treatment: 46 proteins: 17,335,371, 390 public health hazard: 19 pure oxygen activated sludge: 251
plug-flow: 236
44 1
Raschig rings: 217,281,285,287-288 rate of denitrification: 119-151
RBC media: 194-195 recarbonization: 29 1 recirculation: 176-1 79,205 recovery of proteins: 390 reductase: 120 regeneration: 24 rejection ratio: 343,348 residence time: 238 residual chlorine: 296 resin utilization: 309-31 0 retention time: 70-71,222 reverse osmosis: 24,37 reverse osmosis: 337-339,341-353 Reynolds number: 369-371 river water: 18,81 rotating contactor: 22,44-45,50,52,153-
running costs: 38,301
RBC: 193-214
155,166,193-214
safety factor: 100- 102,149,254 saturation constants: 57,70,105, 137,181 see also half saturation constant secondary treatment: 46 sedimentation: 366-388 selectivity coefficient: 308-309,317-318 selectivity: 22,44,308 sequential ion exchange: 320 settling: 366-388 shock load: 192 simultaneous nitrification and denitrification: see SND sludge age: 235,244 sludge production: 50,235,240-241,243 sludge residence time: 248 SND, mechanism: 231-233
solubility of ammonia: 271-273 solubility, temperature dependence: 266 spiral rings: 281 spiral: 350 spray tower: 277 steady state culture: 66 Stokes law: 369 stripping tower: 266-268,277-280
SND: 28,31,45,52, 207,216-234
stripping: 21,23.33 ,261-294 submerged filter: 22,24,50,64,85,90-91,
sulfate: 159,164 support material: 153 surface area: 181-1 83,186-1 88,205 surface of packing: 266 surface rate: 168- 169 suspended growth: 22,30,32,45,53,7 1 , 1 0 1 synthetic ion exchange resin: 306
2 15-234
temperature coefficient: 83 temperature influence: 72-80, 141-
terminology: 46-48 tertiary treatment: 48 thiobacillus denitrificans: 132 thiocyanate: 1 15 toxic constituents: 95,100,117 toxicity: 3,12,117,150-152 transfer coefficient: 273,278-280 transfer units: 279-284,290 treatment costs: 38 trickling filter: 22,45,50-52,64,170-192 trickling filter medium: 171 tube settler: 387-388
142,175-1 77,209-21 0,347
ultrafiltration: 337-339 unit processes: 4330 urban run-off: 10 urea: 17 utilization rate: 239 utilization rate: 254
viscosity-temperature relationship: 377 volatile acids: 128 volumetric rate: 168- 169
Warburg respirator: 67 water recovery: 344 wet deposition: 10-1 1 wetted area: 174 Windhoek: 303 Wuhrmann configuration: 245-246
yield coefficient: 49,57,98,203, 254
442
zeolite: 44,216 zero order kinetics: 57,62-68, 160- 163,168 zeta potential: 363 zone settling: 384-388
443
This Page Intentionally Left Blank