The removal of nitrogen compounds from wastewater (studies in environmental science)

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, including 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


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


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 secondary clarifier as organisms are attached to media.

Combined treatment of carbon and ammonia in a single stage not linked to a secondary clarifier as biomass attached to media

Greater number of 5 unit processes required than for combined carbon oxidation and nitrification.

No protection against 5 toxicants. Only moderate stability of operation. 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; stability not linked to clarifier as organisms 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 required 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


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 process possible; single basin c a n be used; adaptable to sequen- cing batch reactor: process can be adapted to include b o b -

val. cal phosphorus remo-

5-20 No methand required. lesser number of unit processes required. better control of filamentous organisms in activated-sludge process possible. single basn c a n be used, adaptable to sequen- ung batch reactor; process can be adapted 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 operabon 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 required 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; stability not linked to clarifier as organisms on media. High degree of nitrogen removal possible.

Simultaneous nitrification 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 required 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


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


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 chemical treatment pro-

Air stripping (Air)

cBSS8S.

Noeffect 60-95% No effect removal

50-90 Process can be controlled for selected ammonia removals. Most applicable 1 required seasonally in wmbi- nabn with lime system for ~ ~ O S ~ ~ O N S removal. Process may be able to meet total nitro- Qen standards. Not sensitive to toxic substances.

Process is senitive to temperature. Ammonia solubility increases with lowr temperatures. Air requirements 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



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 requirement. 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 chbrine 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


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 standards. 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 ammonia 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



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.



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


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


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 requirements 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 oxidation

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 increasing initial concentration 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 experiment.

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'


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'


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 Nitrosornonas

Pure culture of Nitrobacter


Batch Culture

Pure culture of Nitrobacter

Pure culture of Nitrosomonas isolated 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 denitrification 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 dissociation 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 alternative 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


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 denitrification.

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)


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


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 component 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.



Particulate components Xi = concentration of particulate species i within the biofilm.

%o, = sum of all particulate species concentration.

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 flocculated form from the bulk liquid to the surface of the biofilm.

k,,, = flocculation mass transfer coefficient.



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


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



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 samplingports 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


Part C

PHYSICO= CHEMICAL PROCESSES

Air Stripping Breakpoint Chlorination Ion Exchange Membrane Processes Precipitation


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 absorbent 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


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


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 : adsorbate 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 molecular 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


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


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.


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 backwashed. 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


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


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


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 stabilization 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 "reactions" 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, chromium 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 compounds. 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 concentrations. 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 denitrification 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 clinoptilolite 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 denitrification. 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 denitrification 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 denitrification. 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 performance 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 literature review. Prog. Wat. Tech. 8 p.509-555.

Henze, M., Harremoes, P. and Roed Jensen, 0. (1977): Combined sludge denitrification 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 phosphorus in waste water treatment. Pachec 72, Session 16-3

Ishiguro, M., Watanabe, Y. and Masuda, S. (1978): Advanced waste-water treatment 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 denitrification. 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 temperature 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 thickness. 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 continuous-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 ammonia 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 dissolved 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 denitrification 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


The removal of nitrogen compounds from wastewater (studies in environmental science)

Engineering

rotating biological

policy implications

fjerning av

rotating biological

slaughterhouse

incoming waste

la cour jansen

environmental