EEUU a eC T tr.) Waste Water Treatment Design and optimisation of activated sludge systems Adrianus van Haandel BETOE UR UM Cla aT) COTTE UESHandbook Biological Waste Water Treatment Design and optimisation of activated sludge systems Adrianus van Haandel and Jeroen van der LubbeThis publication has been financially supported by Biothane Systems International — \wiw-biothane.com and Tebodin Consultants and Engineers - www.tebodin.com Published by: Quist Publishing, Leidscnendam, The Netherlands Cover Design: jacky-o.nl Proofreading: E.A. Codling Copyright © 2007 Quist Publishing, AC. van Haandel and J.G.M van der Lubbe ISBN 978-90-77983-22-5 NUR 973 Key words: waste water, sewage, biological teatment, activated sludge system, design, optimisation, nitragen removal, phosphorus removal, aerobic, anaerobic, setting All rights reserved, No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission in writing of Quist Publishing, Brederode 20, 2261 HE Leidschendam, The Netherlandsvii CONTENTS IN BRIEF Chapter 1 Introduction This chapter provides a brief introduction into the field of biological waste water treatment and about the scope of this text. Chapter 2 Organic Material and Bacterial Metabolism The different types of organic material present in waste water are quantified while the main metabolic processes are identified that result in its removal. The appropriate parameter is selected to indicate mass of organic material, for use in mass balance equations. Chapter 3 Organic Material Removal the interaction between the activated sludge system and the organic material in the waste water and its effects on effluent quality, oxygen demand and excess sludge production is evaluated. The importance of the sludge age as an operational parameter and as the crucial design parameter will be highlighted An ideal steady-state activated sludge model is developed that will be used throughout this text as the basis for activated sludge system design and optimisation. The theoretical principles involved in the aeration process are extensively discussed, including methods for the appropriate sizing of aeration equipment (both surface aerators and diffused aeration systems). Practical examples demonstrate how to determine the values of aeration parameters in full-scale activated sludge systems. This includes the use of the oxy gen uptake rate (respirometry ) as a tool for calibration of the steady state model. Finally, different Chapter 4 Nitrogen Removal The stoichiometry and kinetics of nitrification and denite state activated sludge model is extended to include nitrogen removal, An optimisation procedure is developed that will allow the reader to define the optimal configuration of the activated sludge system, as a function of the required effluent nitrogen quality, the influent characteristics and the operational conditions. Recent developments on the subject of nitrogen removal will also be discussed (such as ‘the Sharon and Anammox processes). The performance of these new technologies will be evaluated in several extensive examples, including the effect of their implementation on the performance of the main activated sludge process "Chapter 5 Phosphorus Removal ‘The stoichiometry and kineties of biological- and chemical phosphorus removal processes are presented. ‘The design and performance of various chemical phosphorus removal configurations are discussed (pre-precipitation, simultancous- and postprecipitation). As for biological phosphorus removal, the most popular configurations will be reviewed The steady-state activated sludge model is extended to accommodate both biological and chemical phosphorus removal. An extensive design case, in which the design and performance of several nutrient removal configurations is compared, concludes this chapter.viii Chapter 6 Sludge Settling An extensive mathematical description is given of the activated sludge settling process and of its implications for the design and optimisation of final settlers and sludge thickeners. An optimisation procedure is developed to allow the minimal costs design of (1) the combination aeration tank and settler and (2) the combination of thickener and sludge digester, as a function of the influent composition and the settling characteristics of the sludge The static-point procedure is presented as a method to evaluate the performance of existing final settlers. A significant part of this chapter is devoted to the mechanisms involved in the occurrence of sludge bulking and scum formation, while methods are presented to prevent these problems. Chapter 7 New System Configurations The main focus of this chapter is on Membrane Bioreactors (MBR). The theoretical principles of MBR treatment are explained and a design procedure for the main configurations is presented (submerged and cross-flow MBR), including membrane unit sizing The operation (including membrane fouling and -cleaning) of membrane bioreactors is discussed and the effect of MBR treatment on activated sludge system performance is evaluated. The steady-state activated sludge model is adapted to reflect the changes introduced by MBR treatment, Design data of the main membrane suppliers is presented The recently developed aerobic granular sludge process is also discussed: this includes theoretical principles, reactor configuration and operation Chapter 8 Sludge Treatment and Disposal ‘The kinetics, stoichiometry and operational parameters of both aerobie- and anaerobie sludge digestion are discussed. Models are developed to allow for design optimisation of the paired units sludge thickener and -digester. This mcludes the evaluation of the use of primary settling as a pre-treatment process, Several alternatives for the final disposal of stabilized sludge are discussed, The focus is on the design and performance of sludge drying beds. as this is « cost-effective method to produce a very dry and high quality sludge in regions with a warm climate. Chapter 9 Anaerobie-Acrobic Waste Water Treatment A model for the anaerobic treatment process is presented, including the influence of the main operational and design parameters: retention time, sludge age and temperature. Modern high rate anaerobic treatment systems and combined anaerobic-aerobie treatment systems for secondary and tertiary treatment are discussed. The impact of the digestion of aerobic excess sludge on UASB performance is evaluated. ‘The steady state activated sludge model is extended to include anaerobic pre-treatment. Chapter 10 Integrated Cost-Based Design and Operation In this integration chapter, the optimised cost-based design procedure of the activated sludge system is demonstrated in several extensive worked examples for common secondary- and tertiay treatment configurations. This includes the implications on effluent quality, the dimensions of the main treatment units and treatment costs (both investment- and operational costs). Finally, the application of the steady-state activated sludge model for the optimisation and upgrading of existing wastewater treatment facilities is demonstrated.TABLE OF CONTENTS PREFACE ABOUT THE AUTHORS ACKNOWLEDGEMENTS CONTENTS IN BRIEF LIST OF SYMBOLS AND ABBREVIATIONS 1 INTRODUCTION 1.1 ADVANCES IN SECONDARY WASTE WATER TREATMENT 1.2 TERTIARY WASTE WATER TREATMENT 1.3. TEMPERATURE INFLUENCE ON ACTIVATED SLUDGE DESIGN 1.4 OBJECTIVE OF THE TEXT. 2 ORGANIC MATERIAL AND BACTERIAL METABOLISM 2.1 MEASUREMENT OF ORGANIC MATERIAL 2.1.1 The COD test 2.1.2 The BOD test 2.4.3 The TOC test 2.2 COMPARISON OF MEASUREMENT PARAMETERS 2.3 METABOLISM 2.3.1. Oxidative metabolism 2.3.2 Anoxic respiration 2.3.3 Anaerobic digestion vi vi vi xix3 ORGANIC MATERIAL REMOVAL 3.1_ORGANIC MATERIAL AND ACTIVATED SLUDGE COMPOSITION. 32 33 34 35 3.1.1 3.12 Re Organic material fractions in waste water The composition of organic sludge 3.1.2.1 Active sludge 3.1.2.2 Inactive sludge 3.1.2.3 Definition of sludge fractions Mass balance of the organic material STEADY-STATE MODEL OF THE ACTIVATED SLUDGE SYSTEM 3.21 3.2.2 3.23 3.24 Model development 3.2.1.1 Definition of sludge age _ 3.2.1.2 COD fraction discharged with the e 3.2.1.3 COD fraction in the excess sludge 3.2.1.4 COD fraction oxidised for respiration 3.2.1.5 Model summary Model calibration Model applications 3.2.3.1 Sludge mass and composition 3.2.3.2 Biological reactor volume 3.2.3.3. Sludge producton and nutrient demand 3.2.3.4 Temperature effect 3.2.3.5 F/Mratio Selection and control of the sludge age AERATION 334 3.3.2 Aeration theory 3.3.1.1 Effect of impurities on ki, and DO, 3.3.1.2 Effect of local pressure on DO, 3.3.1.3. Effect of temperature on Ka and DOs 3.3.1.4 Oxygenation capacity for surface aerators 3.3.1.5 Power requirement for diffused aeration Methods to determine the oxygen transfer constant 3.3.2.1 Determination of the standard oxygenation capacity 3.3.2.2 ation of oxygenation capacit ixed liquor 3.3.3.3 Determination of the aeration efficiency GENERAL MODEL OF THE ACTIVATED SLUDGE SYSTEM 3.4.1 3.4.2 3.43 Model development Model calibration Application of the general model CONFIGURATIONS OF THE ACTIVATED SLUDGE SYSTEM 3.51 3.5.2 Conventional activated sludge systems Sequential batch systems 21 21 24 24 25 25 26 31 31 33 33 37 38 40 43 43 48 48 51 52 54 55 57 58 58 59 60 62 63 63 65 68 7” 73 76 7 78 78 80 3.53 Carrousels 3.5.4 Aerated lagoons 82xi 4.1 FUNDAMENTALS OF NITROGEN REMOVAL. 86 Forms and reactions of nitrogenous matter 86 Mass balance of nitrogenous matter 88 Stoichiometrics of reactions with nitrogenous matter “ 41.32 Effectsonalkainty SS 44.3.3. Effects on pH 7 4.2_NITRIFICATION 100 3 soap ae 90) 4.2.2 _ Nitrification in systems with non aerated zones 106 4.2.3 Nitrification potential and nitrification capacity 108 4.3 DENITRIFICATION. = tt 4.3.1 System configurations for denitrification 111 4.3.1.1 Denitrification with an external carbon source 114 ; enitrification with aninternal carbon source 442 4.3.2 Denitrificationkineties tS _ 4.3.2.1 Sludge production in anoxic/aerobic systems 115 jenitrficationrates 43.2.3 Minimum anoxic mass fraction in the pre-D reactor 118 4.3.3 Denitrification capacity 120 4.3.3.1 Denitrification capacity in a pre-D reactor 120 apacily in a post-D reactor 121 4.4 DESIGNING AND OPTIMISING NITROGEN REMOVAL. 124 ity 124 4.4.2 _ Optimised design of nitrogen removal 128 44.2.1 Complete nitrogen removal 120 4.2.2. Incomplete nitrogen removal 132 |_Model enhancem 4.5 NEW DEVELOPMENTS IN NITROGEN REMOVAL SYSTEMS. 134 4.5.1 _ Nitrogen removal over nitrite - Sharon 135 44.3.1 Oxygen consumption 4.5.1.1. Kinetics of high 4.1.2 Reactor configuration and oper: 45.1.3 Required model enhancements 139 Nitrogen removal over nitrite - Babe 146 Anaerobic ammonium oxidation (Anammox) 147 45.3.1 Anammox process characteristics = 150 45.3.2 Reactor design and -configuration 152 454 Effect on the main activated sludge process 157 4.5.5 Side stream nitrogen removal: evaluation and potential 163xii 5 PHOSPHORUS REMOVAL 5.1_ BIOLOGICAL PHOSPHORUS REMOVAL _ 165 5.1.1__ Mechanisms involved in biological phosphorus removal 165 5.1.2 _ Bio-P removal system configurations 169 5.1.3 Model of biological phosphorus removal 172 5.1.3.1 Enhanced cultures. 472 5.1.3.2 Mixed cultures 17 5.1.3.3. Denitrification of bio-P organisms 181 5.1.3.4 Discharge of organic phosphorus with the effluent 183° 5.2_OPTIMISATION OF BIOLOGICAL NUTRIENT REMOVAL 184 5.2.1 Influence of waste water characteristics 184 5.2.2 Improving substrate availability for nutrient removal 185 5.23 Optimisation of operational conditions 188 5.2.4 Resolving operational problems 191 5.3. CHEMICAL PHOSPHORUS REMOVAL 192 5.3.1 Stoichiometrics of chemical phosphorus removal 193, .1.1_ Addition of metal salts 193, 5.3.1.2 Addition offime tg 5.3.1.3 Effect on pH 195 5.32 Chemical phosphorus removal configurations 196 6.3.2.1. Pre-precipitation 198 5.3.22 Simultaneous precipitation 199 5.3.2.3 Post-precipitation 201 5.3. idestream precipi 202 5.33 Design procedure for chemical phosphorus removal 205 5.4 NUTRIENT REMOVAL - DESIGN CASE 208sili 6 SLUDGE SETTLING 221 6.1 METHODS TO DETERMINE SLUDGE SETTLEABILITY 221 6.1.1. Zone settling rate test 221 6.1.2 Altemative parameters for sludge settleability 225 6.1.3 Relationships between different settleability parameters 225 6.1.4 _ Influence of sludge concentration and composition 226 6.2 MODEL FOR SETTLING IN A CONTINUOUS SETTLER 227 6.2.1 Determination of the limiting concentration X, 231 6.2.2 Determination of the critical concentration X- 232 6.2.3 Determination of the minimum concentration Xx 233 6.3 DESIGN OF SECONDARY SETTLERS 234 6.3.1 Optimised design procedure 234 6.3.2 Determination of the critical recirculation rate 238 6.3.3 Optimisation of the reactor - settler system 239 6.3.4 Applicability of the optimised settler design procedure 242 6.3.4.1 US EPA design guidelines 242 6.3.4.2 WRC and modified WRC design guidelines, 243 63.4.3 STORA/STOWA design guidelines 243 6.3.4.4 ATV design guidelines 244 6.3.4.5 Solids flux compared with other design methods 244 6.3.5 Physical design aspects for sludge settlers 248 6.4. SETTLERS UNDER VARIABLE LOADING CONDITIONS 249 6.5 SLUDGE THICKENERS 253 6.6 SLUDGE BULKING AND SCUM FORMATION 255 6.6.1 Control of sludge buking in aerobic systems 258 6.6.2 Control of bulking sludge in anoxic-aerobic systems 260 6.6.3 Non-specific measures to control bulking sludge 262 6.6.4 Causes and control of scum formation 265xiv 7 NEW SYSTEM CONFIGURATIONS 7.1 72 MEMBRANE BIOREACTORS (MBR) 7.1.1 MBR configurations 7.1.1.1 Submerged MBR 7.1.1.2 Cross-flow MBR 7.1.1.3 Comparison of submerged and cross-flow MBR 7.1.2 Theoretical concepts of membrane filtration 7.1.3 MBR design considerations 7.1.3.1 Pretreatment 7.1.3.2. Module configuration 7.1.3.3. Membrane aeration 7.1.3.4 Key data of different membrane types 7.1.3.5 Impact on activated sludge system configuration 7.1.3.8 Integrated MBR design example 7.1.4 MBR operational considerations 7.1.4.1 Operation of submerged membranes 7.1.4.2 Operation of cross-flow membranes 7.1.4.3 Membrane fouling 7.1.44 Membrane cleaning 7.4.5 MBR technology: evaluation and potential AEROBIC GRANULATED SLUDGE 7.2.1 Mechanisms responsible for granule formation 7.2.2 Design and operation of aerobic granulated sludge reactors 7.23 Start-up of aerobic granulated sludge reactors 7.24 Granulated sludge: evaluation and potential 267 268 270 271 273 275 277 279 279 280 281 282 283 2868 296 298 297 297 298 301 302 304 306 308 309xv 8 SLUDGE TREATMENT AND DISPOSAL 311 8.1 EXCESS SLUDGE QUALITY AND QUANTITY 311 8.2 AEROBIC DIGESTION 314 8.2.1 Kinetic model for aerobic sludge digestion 315 8.2.1.1 Variation of the volatile sludge concentration 316 8.2.1.2 Variation of the oxygen uptake rate 316 8.2.1.3. Variation of the nitrate concentration 317 8.2.1.4. Variation of the alkalinity 317 82.1.5 Variation of the BOD 321 Aerobic digestion in the main activated sludge process 321 Aerobic digester design 324 Optimisation of aerobic sludge digestion 329 Operational parameters of the aerobic digester 334 8.3 ANAEROBIC DIGESTION 339 Stoichiometry of anaerobic digestion 342 Configurations used for anaerobic digestion 344 Influence of operational parameters 347 Performance of the high rate anaerobic digester 350 Energy generation in anaerobic sludge digesters 356 Design and optimisation of anaerobic digesters 357 8.4 STABILISED SLUDGE DISPOSAL 359 8.4.1 Natural sludge drying 360 8.4.2 Design and optimisation of natural sludge drying beds 364 84.2.1. Determination of the percolation time 365 84.2.2. Determination of the evaporation time 366 8.4.2.3 Influence of rain on bed productivity 372 8.4.3 Accelerated sludge drying with external energy 373 8.4.3.1 Use of solar energy 374 84.3.2 Use of combustion heat from biogas 3769 ANAFROBIC-AFROBIC WASTE WATER TREATMENT _379 9.1. MODERN ANAEROBIC WASTE WATER TREATMENT PROCESSES — 380 92 93 94 95 9.1.1 Performance of different anaerobic treatment configurations _ 382 94.11 Anaerobicfiter 9.1.1.2 Fluidised and expanded bed systems 383 9.1.1.3 Upflow Anaerobic Sludge Blanket (UASB) reactor 384 9.1.1.4 The RALF system 385 9.1.2 Comparison of different anaerobic sewage treatment systems 386, MODEL FOR THE ANAEROBIC PRE-TREATMENT PROCESS 387 9.2.1 Sludge age as the key parameter for anaerobic treatment 388 9.22 Influence of temperature on anaerobic sewage digestion 390 ANAEROBIC-AEROBIC SYSTEMS FOR SECONDARY TREATMENT 393 9.3.1. Effect of anaerobic pre treatment on secondary post treatment 394 9.3.2 Stabilisation of aerobic sludge in the UASB reactor 396 9.3.3 Optimised design of anaerobic-aerobic treatment systems 398 ANAEROBIC-AEROBIC SYSTEMS FOR TERTIARY TREATMENT 403. 9.41 Anaerobic digestion with reduced methanogenic efficiency 403 9.4.2 Application of new nitrogen removal configurations 405 MODIFICATIONS OF THE UASB REACTOR 406 9.5.1 The EGSB reactor 406 95.2 Two stage anaerobic digestion 40710 INTEGRATED COST-BASED DESIGN AND OPERATION 10.1 PREPARATIONS FOR SYSTEM DESIGN 10.1.1 10.1.2 10.1.3 10.1.4 10.1.5 The basis of design 10.1.1.1 Weste water characteristics 10.1.1.2 Kinetic parameters and settieabilty of the sludge Costing data 10.1.2.1 Investment costs 10.1.2.2 Operational costs 10.1.2.3 Annualised costs Performance objectives Applicable system configurations Limitations and constraints 10.2 OPTIMISED DESIGN PROCEDURE 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5 10.2.6 10.2.7 10.2.8 10.2.9 System A1: Conventional secondary treatment System A2: Secondary treatment with primary clarification System B1: UASB pre-treatment with nitrification System C1: Tertiary treatment with nitrogen removal System C2: Tertiary treatment with N and P removal System comparison Influence of the waste water temperature Influence of the sludge age Comparison of different operational regimes 10.3 OPERATIONAL OPTIMISATION PROCEDURE, 10.4 FINAL REMARKS wwii 409 410 410 411 414 414 414 418 419 420 421 424 426 426 436 439 452 457 464 465 467 469 472 474xviii APPENDICES A1 DETERMINATION OF THE OXYGEN UPTAKE RATE A1.1 DETERMINATION OF THE APPARENT OUR (1.2 CORRECTION FACTORS OF THE APPARENT OUR 12.1 Representativeness of mixed liquor operational conditions 1.2.2 Critical dissolved oxygen concentration 1.2.3 Hydraulic effects 1.2.4 Absorption of atmospheric oxygen 1.2.5 The relaxation effect A2 CALIBRATION OF THE GENERAL MODEL A2.1 CALIBRATION WITH CYCLIC LOADING A2.2 CALIBRATION WITH BATCH LOADING A3 THE NON-IDEAL ACTIVATED SLUDGE PROCESS A4 DETERMINATION OF NITRIFICATION KINETICS ‘A4.1 METHODOLOGY 4.2 APPLICATION EXAMPLE A5 EXTENSIONS TO THE SIMPLIFIED MODEL A5.1 IMPERFECT SOLID-LIQUID SEPARATION IN FINAL SETTLER 5.1.1 Particulate organic nitrogen in the effluent 5.1.2 Particulate organic phosphorus in the effluent AS.1.3 Excess sludge production and composition AS.2 NITRIFIER FRACTION IN THE VOLATILE SLUDGE MASS SUBJECT INDEX REFERENCE LIST 477 479 480 480 480 481 482 482 485 489 490 494 497 501 502 504 509 509 509 511 511 512 515 521aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.xx List of parameters not included in naming convention a recirculation factor (-) a, = annualisation factor (-) cp proportionality constant (DSVI/SSVI3 5) G, = active sludge mass per unit daily applied biodegradable COD DeiDes = denitrification capacity in pre-D respectively post-D anoxic zone f = fraction remaining as endogenous residue F/M feed/mass ratio (mg COD.mg! VSS.d") I investment costs (USS) i interest rate Vesilind constant (1g! TSS) xy gen transfer coefficient (h) denitrification constant (mg Nmg" X,d") thermodynamic constants in the carbonic acid system adsorption rate constant (h”) Henry constant (-) specific utilisation rate of adsorbed organic material (mg COD.mg" VSS.d") specific utilisation rate of easily biodegradable organic material (mg COD I") half saturation constant for nitrification (mg N.I") half saturation constant for oxygen (mg O31") half rate (Monod) constant for S,, (mg COD.mg” VSS) half rate (Monod) constant for S,, (mg COD.I") dissociation constant for water gradient n expected economical lifetime (years) Po = energy dissipation per unit volume (W.m*) r anoxie reeycle factor (bio-P system) R = annualised financing costs (US$.year") Ra minimum sludge age where Nw: =D. (4) Roo = minimum sludge age required to allow for anoxic zones, i.e. When Nye = Ra = minimum sludge age required for establishment of nitrification (d) Ry minimum sludge age where f, = fy, (d) Vo Vesilind constant (m.d") u downward velocity in final settler (m.h) v upward velocity in final settler (mh) v viscosity a alpha factor in aeration (-), gradient (m.m") p beta factor in aeration (-) 6 temperature dependencyaa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.‘Compressed air injection ‘or surface aerators (Secondary sludge) Raw sewage Setted sewage Mixed tiguor ot it influent en & Settler 1 Effluent Pe 1 \ 1 Primary sludge ! 1 t t | Excess studge i \ Figure 1.1 Representation of the basic configuration of the activated sludge process (primary and secondary waste water treatment) The biological reactor or aeration tank is filled with amisture of activated sludge and influent, known as “mixed liquor”. The aeration equipment (either surface aerators or compressors connected to submerged air diffusers) transfers the oxygen necessary for the oxidation of organic material into the reactor, while simultaneously introducing enough turbulence to keep the sludge flocs in suspension. ‘The continuous introduction of new influent results in a continuous discharge of mixed liquor to the final settler (or secondary clarifier), where phase separation of solids and liquid takes place. The liquid leaves the system as treated effluent, whereas the sludge is recirculated to the aeration tank and for that reason is called “return sludge”. A primary settler (or primary clarifier) may be introduced to remove part of the suspended solids present in the influent. This reduces the organic load to the biological reactor, The settled suspended solids (“primary sludge”) are often sent to an anaerobic digester, together with the activated sludge that is discharged from the biological reactor: the excess sludge. In the anaerobic digester, the volatile suspended solids in the excess sludge are, in the absence of oxygen, partly degraded into methane and carbon dioxide. Without the discharge of excess sludge, there would be a continuous growth of sludge in the reactor and consequently, an increase of the sludge concentration in the process In practice the activated shidge concentration must not be allowed exceed a certain maximum value in order to guarantee proper functioning of the final settler (secondary clarifier), For concentrations beyond the maximum, sludge will escape together with the effluent. A constant sludge mass is maintained when the rate of sludge production is equal to the rate of sludge wastage, where this wastage may be unintentional (in the effluent) or intentional (as excess sludge). In practice, excess sludge is discharged either directly from the aeration tank or from the return sludge line, although the first option has some advantages, which will be explained later in this book, The production of excess sludge adds an extra dimension to the activated sludge process apart from the waste water treatment process, there is also a need to find a suitable method for the treatment and final disposal of the produced excess sludge. In practice the sludge is generally submitted to a biological stabilisation process with the objective to reduce the fraction of biodegradable material (biomass and organic material) and as such to avoid pulrification. After removing a large part of the water fraction of the sludge, a solid end product is obtained that may be used in agriculture, disposed into a landfill or sent to an incinerator.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.10 Example In the traditional COD test (open reflux), a mixture of 10 ml of waste water sample, 5 mi of 0.25 N potassium dichromate and 15 mi of sulphuric acid is utlised. What is the highest value of the COD concentration that can be determined? Solution In the initial mixture the available quantity of dichromate = 5.0.25 = 1.25 meq. If the oxidant is entirely used during the COD test, this would mean that 1.25 meq of organic material is utilised, equivalent to 1.25:8 = 10 mg O2 as the equivalent weight of oxygen is 32/4 = 8 gram.eq”, see Eq, (2.1b). As the 10 mg of organic material (expressed as COD) were present in a 10 ml waste water sample, its concentration was 10 mg per 10 ml or 1000 mg." It is concluded it is impossible to determine a COD concentration higher than 1000 mg." because there would be no residual dichromate left. In practice it will be attempted to dilute the ‘sample so that the expected COD concentration is about equal to 500 ma" 2.1.2 The BOD test In the BOD test oxygen is used for the oxidation of organic material and this process requires the presence of micro-organisms. If the waste water does not contain enough of these micro-organisms, they must be added at the beginning of the test, together with mineral nutrients and a buffer to maintain a neutral pH While in the COD test the oxidation of organic material is essentially complete in less than two hours, in the BOD test the oxidation rate is very slow and will take several weeks. As it is impractical to wait such a long time for the result of the test, a standard period of 5 days has been introduced, even though it is well known that this period is not enough for complete oxidation, Because temperature has been found to affect the oxidation rate, a standard temperature of 20°C is used. Hence, unless differently stated, BOD means the BODs,>) Le. the BOD after 5 days of incubation at 20°C. Some organic compounds (especially those with small molecules) can be metabolised immediately by micro-organisms. On the other hand, in most waste waters there are also suspended solids, colloids and macro molecules that need to be reduced by hydrolysis into smaller molecules, before they ean be metabolised. Also some organic compounds have a very low rate of metabolism, thus resulting in a very low consumption of oxygen during the five day test period for BOD. The organic material metabolised during the test is determined by the oxygen consumption and is called the biodegradable material. Organic compounds that cause no measurable oxygen consumption are called non-biodegradable and are therefore not detected with this test. In the case of biodegradable material, the oxidation will not be complete after 5 days of incubation, so that it is not possible to calculate a priori a theoretical value for a solution of a known composition, as was done above for the COD test. This is due to the fact that before the test it is not known which proportion of the organic material metabolised by the micro-organisms will be oxidised (hence contributing to the BOD) and which part will be incorporated in the cell mass. An often-used empirical equation for the consumption of oxygen and hence for the BOD in a solution of biodegradable material is: BOD, = BOD.zar[1 - eX} (23)aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.iW 2.2 COMPARISON OF MEASUREMENT PARAMETERS In this section, we will attempt to relate the parameters used to quantify’ the concentration of organic material (COD and TOC) to the chemical energy contained in the material, To do so, it will be necessary to introduce a basic thermodynamic concept, free energy, defined as the amount of useful energy released during @ chemical reaction, for example during oxidation of organic material, Values of the free energy release of many compounds can be found in thermodynamic tables, Table 2.4 shows that the values of the released free energy, expressed as kI mol", vary enormously for different chemical compounds. However, when the released free energy per unit mass of theoretical COD is calculated, the value is more or less constant for the different compounds. The only substantially different values are those for the first two from the list: oxalic acid 21.6 kI.g' COD and formic acid (18.0 kI.g! COD). If these two compounds are excluded, an average value of 13.7 kJis calculated for all other compounds and none of the individual values deviates more than 10 percent from this average value. ‘Table 2.4 Free energy release / electron transfer upon oxidation of selected compounds Component I 0 mm Iv v GsJmot) | (kJ.gt COD) | (k-g!TOC) | Na) Na Onalie acid 344.4 21.35 14.45 2 1 Formic acid 285.6 18.02 23.98 2 2 Citric acid 72.4 15.04 30.07 6 3 Glucose 2881.2 14.99 40.03 m4 4 Lactic acid 1369.2 14.24 38.01 2 4 Acetic acid 869.4 13.57 36.20 8 4 Glycerine 1625.4 14.49 30.44 4 Vhs Phenol 3036.6 13.52 42.04 28 4, Ethyl. glycol 11802 14.74 49.10 10 5 Benzene 3196.2 13.31 4431 30 5 Acetone 1722 13.44 S116 16 3's Palmitic acid | 9819.6 13.36 51.16 2 3% Cyelohexane | 3784.2 13.10 52.42 36 6 Ethylene 13188 13.73 54.94 12 6 Ethanol 13104 13.65, 54.60 12 6 Methanol 693 14.45, 37.79 6 6 Ethane 1444.8 12.89 60.19 4 7 Methane 802.2 12.52 65.44 g 8 Tieadings. I~ free energy content, IT - energy content per g COD, III — energy content per g TOC, TV = number of electron equivalents per mol compound and V = number of electrons transferred per C-atom,aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 21 Chapter 3 ORGANIC MATERIAL REMOVAL In this chapter, a quantitative model is developed that describes the removal of organic material in the activated sludge system and its consequences for the principal parameters of the process: effluent quality, excess sludge production and oxygen consumption, The validity of the model has been thoroughly tested during experimental work at bench-, pilot- and full-scale processes, treating different waste waters under very diverse operational conditions. In all situations the correlation between the experimental values of the process variables and the values predicted by the model was excellent. The model is applicable to all aerobic suspended sludge treatment systems, which include the different versions of the conventional activated sludge system, sequential batch reactors, carrousels and aerated lagoons In Chapter 9 it is shown that the model remains valid when it is used for sewage that has received anaerobic pre-treatment, while in Chapters 4 and 5 itis demonstrated that the model can be extended to include nitrogen- and phosphorus removal. Most of the concepts presented in Chapter 3 have been developed at the University of Cape ‘Town (UCT) in South Africa and form the backbone of the activated sludge models as developed by the specialist group of the International Water Association (TWA, 1986, 1994 and 2000). The general model presented in this chapter is capable of predicting the variation in space and in time of all measurable parameters related to organic material removal in activated sludge systems with reactors in series and operated under variable flow and load conditions. For the most general (dynamic) application of the model, it is necessary to use a computer program, However, a dynamic model, writien in the form of differential equations, is very suitable for simulating system behaviour but less so for optimised design. Advanced and often costly simulation software is available, but in general requires many kinetic and stoichiometric parameters to be estimated. On the other hand, the main parameters of interest in biological wastewater treatment, ic. effluent quality, sludge production, average oxygen demand and required treatment volume, can be calculated very well with a simplified steady state model. In summary, the main advantages of this steady state model are = Itallows for easy design optimisation of activated sludge systems, = Itcan be extended with nitrogen- and phosphorus removal (Chapiers 4 and 5); = Design of auxiliary systems such as settlers and thickeners (Chapter 6), aerobic- and anaerobic digesters (Chapter 8), primary clarifiers and UASB reactors (Chapter 9) can all be included. This allows for truly optimised design; 3.1 ORGANIC MATERIAL AND ACTIVATED SLUDGE COMPOSITION 3.1.1 Organic material fractions in waste water In this text COD will be used as the parameter for organic material measurement, Some advantages of its use over its altematives BODs, BOD») and TOC have been mentioned already in the previous chapter. In Section 3.2 another important advantage will be presented: the possibility to verify if the organic material mass balance closes, The concentration and composition of the organic material depends on the origin of the waste water. For the purpose of modelling the activated sludge system, it is necessary to divide the influent organic material into four different fractions,aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 28 ‘The composition of the active sludge may differ considerably from one system to the other, depending on the nature of the influent waste water and the operational conditions. In spite of the complex nature of the active sludge mass, in this text it will be considered (for the purpose of modelling) as an equivalent bacterial suspension, To test the validity of this assumption, the predictions generated by the model will be compared to the experimentally observed results. It must be stressed that although bacteria are predominant in the active sludge its actual behaviour may be very different from a pure culture of bacteria, 3.1.2.2 Inactive sludge The inactive sludge is composed of non biodegradable organic material and can be subdivided in two fractions in accordance with its origin: (1) the inert sludge and (2) the endogenous residue. The inert sludge fraction is generated from the accumulation of particulate non-biodegradable organic material present in the influent, This material is flocculated and becomes part of the solid phase, forming the inert fraction The endogenous residue has its origin in the decay of living bacteria cells, a process occurring continuously in the activated sludge system, During the decay provess of the active sludge, part of the microbial mass is oxidised in a process called endogenous respiration. However. only part of the cellular mass is biodegradable: after decay a fraction of the decayed active sludge remains in the activated sludge as a non-biodegradable particulate fraction, The existence of the endogenous residue will be demonstrated and quantified in Chapter 8 ‘Apart from the different organic factions of the sludge there is also an inorganic one ‘This material can be determined after the organic material is removed by means of ignition at 550°C. It should be noted that part of the measured weight is due to the formation of phosphates, sulphates and bicarbonates during the combustion of the organic material Experimentally the value of the formed inorganic residue of the orgame sludge has been determined as approximately 20%. This means that the measured volatile sludge mass (active, inert and endogenous) is accompanied by a fixed solids mass fraction of 1/0.8 -1 = 25% Inadgition to the inorganic fraction formed after ignition of the volatile sludge, inorganic solids may accumulate in the sludge from inert influent material such as silt and clay. Thus in practice the ratio between volatile and total suspended solids tends to be less than 0.80. Depending on the origin of the waste water and the operational conditions (pre-sedimentuation, applied sludge age, quality of the sewer system), the ratio between volatile and total solids for domestic sewage will be in the order of 0.65 to 0.75. For industrial waste waters containing a very low or even no inorganic material, this ratio will be close to the maximum value of 0.80. 3.1.2.3 Definition of sludge fractions Having defined the different sludge fractions, it is convenient to introduce symbols for each Using the letter “X” to indicate generically sludge concentration one has X, = active sludge concentration (mg VSS.I") X, = endogenous sludge concentration (mg VSS.I") X, = inert sludge concentration (mg VSS.I") X, = organic or volatile sludge concentration (mg VSS") X,, = mineral, fixed or inorganic sludge concentration (mg FSS.1") KX, = sludge concentration (mg TSS.I")aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 29 From Eq, (3.13) it can be concluded that the theoretical value of the recovery factor is identical to one, Due to experimental errors and spontaneous fluctuations in the phy iological activity of the sludge, the value of B, will deviate from its theoretical value. However, when the average value of a series of experiments over a period (for example a few weeks) is considered, the deviation between the theoretical and the experimental value of the recovery factor will typically be less than 10 percent. Stated differently, if there is a systematic difference between the theoretical and experimental value of B,, there is good reason to suspect that one or more of the tests used to calculate B, is not being carried out properly or that the activated sludge system is not yet operating under steady state conditions. On the other hand, a closing mass balance (i.e. an experimental B, between 0.9 and 1.1) is clear indication that the system was operating under steady state conditions and that the tests to determine B, were carried out correetly. Henee, the verification of a closing mass balance is powerful indication that the experimental data are reliable Example 3.1 As an example of a mass balance calculation, the experimental data of Dias et al (1981) in Table 3.2 will be analysed. In this experiment a bench scale activated sludge system was utilised and raw sewage was treated at § different sludge ages in five sets of experiments (I to V). Table 3.2 shows the results of the daily analyses, reported per set in column |. The OUR was determined for oxidation of carbonaceous material only (nitrification was inhibited by adding allylthio urea, a toxic compound for nitrifers, but not for heterotrophs) Table 3.2 Experimental results of an activated sludge system (steady state conditions) wot |v | q x, Bo exp. | litre | Vd" | Ld? mgI" 3) j 9 | 10] 1 | 339 1060 4.04 uf] o9 | 42] 16 | 1.20 2235 4.02 m9 | 4s] 16 | 075 1 2538 1.03 ww} 14 | 12 | 16 | 060 155 | 3012 4.00 Vv 20 15 14 0.50 7 2686 0.97 Solution Applying Eq, (3.13) for each of the five sets of experiments, the Bo values can be calculated. As an example, for set | one has: Bo = Ste + (WQForXy + RrrOcVSy, = (127 + (3.33/16)-1.5-1060 + (10/16)-20.3:24)/730 = 1.04 The calculated values for Bo, are in the last column of Table 3.2 It can be noted that in all experiments the B. values tend towards the theoretical value of 1.00. The weighted average taking into consideration all experiments was Bo= 1.02, which means that there is a difference of 2 percent between the experimental and the theoretical value of B.. As this difference is very small, it is concluded that the experimental data were obtained with the aid of correct procedure,aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 33 3.2.1.1 Definition of sludge age Having defined the conditions to formulate the simplified model, the most important operational variable will now be defined: the sludge age R, This parameter indicates the average retention time of the sludge in the system and is defined as the ratio between the sludge mass present in the system and the daily sludge mass discharged from it. Using the model of Fig. 3.3 and assuming hydraulic sludge wasting (i.e. excess sludge discharge from the aeration tank) one has: R, =MX/ME, = VeXVqXd = va (15) Where R, — =sludge age (d) MX, = sludge mass in the system (kg TSS) ME, Sd), Equation (3,15) can also be written in another way a = WR 16) Equation (3.16) expresses that the flow of excess sludge, when discharged from the biological reactor, is a fraction /R, of the reactor volume, ie. over a period of R, days the volume of wasted sludge is equal to the reactor volume. The sludge age is independent of the liquid retention time R,,. This parameter is defined as the ratio between the reactor volume and the influent flow R, = WQ (17) Using the definitions for the sludge age and the liquid retention time, it is now possible to derive expressions to predict the values of the COD fractions mS,.,mS,, and mS,, which is the objective of the simplified model for the activated sludge system 3.2.1.2 COD fraction discharged with the effluent In the ideal activated sludge system, the effluent COD and the COD of the liquid phase of the mixed liquor are both equal to the non-biodegradable dissolved organic material in the influent, Sou: From the definition of S., in Eq, (3.2), one has: MS. = Sw/Su = Sysi/Sa = fas 3.18) Hence the simplified model predicts a constant effluent COD, independent of the sludge age or the liquid retention time and equal to thenon biodegradable, dissolved influent COD fraction 3.2.1.3 COD fraction in the excess sludge The determination of this fraction is more complicated and requires derivation of expressions for the three fractions that compose the organic sludge: inert sludge, active sludge and endogenous residue,aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 37 3.2.1.4 COD fraction oxidised for respiration Oxygen is consumed for both exogenous and endogenous respiration. The oxygen uptake rate (OUR) due to exogenous respiration ,, is determined from Fig. 3.3, where it is shown that upon metabolism of 1 gram of COD, there will be a production of active sludge of Y gram of VSS with a COD value of f.y"Y gram COD. Hence the remaining fraction of (I-f°Y) gram COD will be oxidised and for that oxidation, by definition, an oxy gen mass of (I-f4"Y) gram (Op is required. Hence the exogenous consumption rate can be expressed as: Ox = (1- fa Ytu = (far) S/R G39) ‘The OUR for endogenous respiration Ow. is calculated fiom the oxidation rate of decayed activated sludge, which is the difference between the decay rate and the production rate of endogenous residue re = (dX ata AX ata = byXa- fb X= (1 - bby Xe G40) Where 1, = oxidation rate of decayed active sludge Again using the proportionality constant £,, the endogenous respiration rate can be calculated Om = feo = fer - fyb Xs Gl) The total OUR for the oxidation of organic material is equal to the sum of the values for exogenous and for endogenous respiration: 0. = Ont Oo Using Eqs. (3.3 and 3,29) to substitute for Sy, and X, leads to: 0, = (1= fre fig): (1 = far ¥ + fal = f)dy CSRs (3.42) The influent COD fraction that is oxidised in the activated sludge system is now expressed as: mS, =MOJMS,= (VrO.V(QiSs) = = tas fap) TC = fos ¥) + fev (l = DbyCr] G43)aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 41 Naturally, the procedure presented above is only valid when the behaviour of the activated sludge system approaches ideality: i.e. when the concentration of suspended solids in the effluent is very low. An example of the determination of the f,, and fy, values is presented in Fig. 3.4 The collected data refer to an experiment conducted with raw sewage from the city of Campina Grande by Dias et al (1981), which was discussed in Example 3.1. This data set was complemented by Van Haandel and Catunda (1985 and 1989). Table 3.2 deals specifically with the data presented by Dias et al (1981). In this table it can be seen that the value of the recovery factor B, deviates less than ten percent from the theoretical value of one, so that the data is considered acceptable The experimental values of mS,., mSq, and mS,, were calculated using Eq, (3.14), as indicated in Table 3.3 while Eq. 3.15 was used to calculate the sludge age R,. In Fig. 3.4 the measured values of mS,e, mSyy and mS, are shown as a function of the sludge age Ry. In so far as the non biodegradable and dissolved influent COD fraction is concerned, Fig. 3.42 shows that the ratio of effluent and influent soluble COD oscillates around 0. 14 so that this value is accepted as the ‘best” value for fy. Fraction in the eftluent Oxidized fraction Fraction in tho oxcoss sludge 1 1 ‘Van HaanceiCatnda (1985) 98) van HaanceiCatinds (1989) 02 oO ‘Sludge age (days) ‘Sludge age (days) Sludge age (days) Figure 3.4 Experimental results of the COD fractions mS,,mS,, and mS, compared to model calculated values for different f,, and f,,, values Once the f,, value has been established, the f, value is determined as follows: (1) With the aid of Eq. (3.38), calculate as a function of the sludge age theoretical values of mS, for different fi, values: (2) Plot the theoretical mS. curves as a function of R, for the chosen fy values: (3) Similarly, using Eq. (3.43), calculate and plot theoretical curves of mS, as a function of Ry for the same series of fy values. (4) By comparing the theoretical curves of mSy and mS, and the experimental results, the fy value that gives the closest correlation between experimental and theoretical results is selected as the “best” value for the sewage under consideration. In the case of Fig, 3.4, theoretical curves were generated for values of f,y ranging from 0.00 to 0.12, The figures 3.4 b and e show clearly that the value fy = 0.06 results in the closest correlation between the theoretical and experimental values. In Fig. 3.4 there is a close correspondence between theory and practice over the entire sludge age range from 2 to 30 days.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 45 Faw sewage Settted sewage 1 O14 0.02 208 08 20°C fav = > 3 os 1 5 1 3 fot 02 + 7 0 2 4 6 & 10 12 14 18 18 20 4 6 8 10 12 14 16 18 20 ‘Sludge age (4) Sludge age (d) Figure 3.8 Active sludge fraction as a function of the sludge age for raw and settled sewage Example 3.2 An activated sludge system treats raw sewage (frs = 0.14 and fyp = 0.10) andis operated at a sludge age of 20 days. Itis required to increase the organic load by 50 percent from 10 to 15 ton COD per day without increasing the sludge mass in the system. What vill be the new maximum sludge age? How much does the fraction of influent COD wasted as excess sludge change? How much will the oxygen consumption increase? Solution From Fig. 3.7a itis determines that mX, (the organic sludge mass per unit mass of daily applied influent COD) for the gven sludge age of 20 days is equal to 3.6 mg VSS.mg" COD.d", Hence the maximum sludge mass in the system is: MX, = mX, MSs = 3.610 = 36 ton VSS. After the load increase, the total sludge mass is not to increase, so that the sludge age must be feduced, For the new sludge age the mX, valueis given as: mX, = MXJ/MSy = 36/15 = 2.4 mg VSS.mg'COD.d" Again using Fig. 3.7, it is noted that for mX, = 2.4 the sludge age is 12 days. Hence it is concluded that the sludge age must be reduced from 20 to 12 days, due to the increase of the organic load. The change of the active sludge fraction can be evaluated with the aid of Fig. 3.8: for Rs = 20 days one has fay = 0.33 while for R= 12 days the value of fay = 0.45aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 49 os 0.0128 0.05 @ 04 0.04 0.01 9 = sa o a a * 9 Q z © | sane = 03 0.03 "m f 0.0075 7, 8 5 E = z a g Blas 2 = 02 02 E fo.008 E wf z a € = € o4 0.01 0.0025 o o 2 4 6 6 10 12 14 16 18 20 Sludge age (d) Figure 3.10 Typical profile of volatile sludge production and nitrogen / phosphorus demand as function of the sludge age for raw (fyp= 0.10) and settled sewage (fap = 0.02) Once the values of mN; and mP; have been established, it is simple matter to calculate the corresponding minimum required nutrient concentrations in the influent. For nitrogen one has: mN, = MN/MSq = (QeN)/(QiSu) = N/Sy or N Fyne, Sy= fye[(1 = fn fap) [CL + EY RAYCUR, + fg /fen |S 61) For phosphorus one calculates: mP, =MPJMS, =(Q,P)(Q,'S,) = P/Seor Py =mPySq= fpMEySy= fy [(-forlag (ED ER.YCIR + fopfes Ss (3.62) Where N= influent nitrogen concentration required for excess sludge production P, = influent phosphorus concentration required for excess sludge production In the case of domestic waste water, the concentrations of the nutrients will be much higher than the minimum values required for sludge production. However, in industrial waste waters, especially those of vegetable origin, the nutrient concentrations are low and additions to the influent may be necessary to have a properly functioning system.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Material Removal 57 Chapter 3- Organi In general surface aeration equipment is significantly less expensive than diffused aeration. On the other hand, the oxy gen transfer efficiency in kg O>kWhr' is lower at well. A disadvantage of surface aerators is that operation in a reactor with a depth of more than 4 metres is difficult. Some suppliers equip the surface aerators with draft tubes, extending from below the turbine blades to near the bottom of the reactor, which induces an upward (or downward) flow. This way is it possible to operate surface aerators up to depths of 6 metres. An advantage of surface aerators is that they are particularly suited for circuit systems such as carrousels, as a flow direction may be induced. Thus the need for propulsors to circulate the mixed liquor can be dispensed with 3.3.1 Aeration theory Aeration theory is based on Henry's law: at equilibrium, the partial pressure of a component in the gas phase is proportional to the concentration of this component dissolved in the liquid phase. In the case of aeration, the liquid phase is the mixed liquor and the gas phase is air, whereas the component in question is oxygen. Hence, equilibrium exists if: DO, Poo G67) Where DO, = saturation concentration of dissolved oxygen in the mixed liquor (mol I” or mg.I") ky = Henry's constant Poo = partial pressure of oxygen in air = 0.21 atm at atmospheric pressure Note that the value of ki is dependent on temperature, In principle, by using Eq. (3.67) one can calculate the equilibrium oxygen concentration in the mixed liquor for the existing atmospheric pressure and the temperature at hand. Fortunately, standard tables are available where the equilibrium dissolved oxygen concentration is listed as a function of temperature, pressure and salinity. In biological treatment systems under normal circumstances, the dissolved oxygen concentration in the mixed liquor will be less than the saturation value, because oxy gen is being consumed by the micro-organisms. Under those conditions there is a natural tendency of atmospheric oxygen to be transferred to the mixed liquor. According to Fick’s law, the transfer rate is proportional to the difference that exists between the saturation concentration and the actual dissolved oxygen concentration in the mixed liquor (dDO//at) = ky,(DO, - DO) (3.68) Where: dDO/dt = wansfer rate of atmospheric oxy gen (mg O21" h") DO, = saturation dissolved oxygen concentration in the mixed liquor (mg 02.1") DO, — = dissolved oxygen concentration in the mixed liquor (mg O21") ky = oxygen transfer coefficient (h") ‘The value of the transfer constant depends on the aeration system, the geometry of the reactor, operational conditions (temperature, atmospheric pressure) and impurities in the mixed liquor. In practice, the concept of oxygenation capacity is often used The oxygenation capacity of an aerator is the maximum oxygen transfer rate under standard operational conditions. The oxygen transfer is measured in pure water without oxygen (DO, = 0) under atmospheric pressure (1.013 bar or 760 mm Hg) and at a temperature of 20°C. The oxygenation capacity (in mg O>.".h”) is given as:aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 61 When OC, is given, it is easy to calculate the required aeration capacity with the previously determined ratio between OC/OC,, For surface aerators, the standard oxy gen transfer capacity OC, is expressed by the equipment suppliers in kg O>.kWh'! Pay =MO/OC, (3.78) Where: P,, = required motor power (kW) OC, — = actual oxygen transfer capacity (kg OkWh™) MO, — = oxygen consumption (kg O2.h"') Example 3. if a mechenical aerator has an oxygenation capacity of 2.0 kg O2.kWh"' under standard conditions, what will the oxygenation capacity be in an activated sludge system (DO, = 2 mg O21") at an altitude of 800 m in winter (T = 18°C) and summer (T = 28°C). Calculate the required aeration capacity (i.e. the required motor power), assuming that the daily oxygen consumption (MO,) is equal to 2500 kg O2.d" Solution At 800 m the atmospheric pressure is 919 mbar or 690 mm Hg (interpolation of Table 3.7). Using Eq. (3.77) the ratio between actual and standard oxygenation capacity is calculated as: (690 - 16)51.60.99.1 OCS OC, = 0.8-(1.012) 97-{ —___________- 29.1 = 0.49 at 18°C and (760- 16)(31.6 + 18) Pm = MOY(24.OC.-(OCJOC,)) = 2500/(24-0.49:2.0) = 106 kW (690 - 30)51.6:09:9.1 OCY OC, = 0.8(1.012) %*.; —___- gy9.1 (780-30) (31.6 + 28) 43 at 28°C Pm = MOy(24-0Cs(OC/OC,)) = 2500/(24-0.43-2,0) = 121 kW In both cases the oxygenation capacity under process conditions is well below that under standard conditions, Furthermore the oxygenation capacity decreases with increasing temperature for the chosen 0 value of 1.012, which is assumed to be representative for mecharical aerators. It is concluded that when the temperature rises, the increase in the value of the ka constant does not compensate for the negative effect of the reduction of the oxygen saturation concentration. As the oxygen demand tends be higher with increasing temperature as well (due to a higher decay rate of active sludge), one would design the aeration capacity to be installed for the highest temperature.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 65 Table 39 Experimental results from the test to determine the oxygenation capacity Time DO, DO.-DO; In(00,-DO)) Kia min. mg O: mg 02." &) min.” oO 1.0 77 2.04 oO 1 22 65 1.87 017 2 3.0 57 1.74 0.15 4 42 45 1.50 0.13 6 52 35 4.25 0.13 8 6.0 27 0.99 0.13 12 7.0 17 0.50 012 16 7 oT -0.30 0.15 32 86 0.1 -2.30 0.13 Solution At 22°C the dissolved oxygen saturation concentration is 87 mg Oz!" In Table 3.9 column 3, ‘the deficit (DO.- DO), as well as the natural log of the deficit (column 4) have been calculated. In Column 5 the values of the transfer constant has been calculated as [In(DO.-DO)J/t using Eq. (3.77). The experimental results suggest a value of 0.13 min’ for Kis at 22°C. Hence, the oxygenation capacity is ka DO, = 0.13°8.7 = 1.13 mg Oal".min" at 22°C or (0.13/(1.012)°)9. 1.16 mg OaI"".min” at 20°C, Therefore, in the 5000 m” reactor the maximum transfer rate of oxygen at 20°C would be 5000:1.16 g O2 min’ or 347 kg O2.h*. As the measured power consumption during the test was equal to 150 kWh, the maximum transfer efficiency is 347/150 = 2.31 kg Oz.kWh", more than specified by the supplier (22 kg O2: kWh"), 3.3.2.2 Determination of oxygenation capacity in mixed liquor If an activated sludge system is already in operation, the procedure outlined above to determine the oxygen transfer coefficient cannot be applied, because of the oxygen consumption by the micro-organisms. However, if the system operates under steady-state conditions the k,, value can still be determined since the dissolved oxygen concentration in the liquid phase does not change because the oxygen transfer rate is equal to the oxygen consumption rate. Hence: (4DOyat) =0 = (ADOVA),- O,=kyxDO-DO) - 0, (3.88a) Where: dDOydt = rate of change of the dissolved oxygen concentration (dDO/dt), = aeration rate OF jotal oxygen uptake rate Rearranging Eq, (3.882) leads tok, =O,/(DO,-DO,) (3.88b)aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 7 Table 3.12 Process kinetics and production rates in the general activated sludge model Par, | Equation Description No. Tu | =KmcXeSi/(Ste + Kw) Utilisation of easily biodegr. material G.95) (= wKyp-SiX.) __| Adsorption of slowly biodegr. material | 3.96) Tu | =KayXvSwSi+ Ky Xo) __| Hydrolysis of sored material G97) | =Yate Growth of active sludge 6.98) = Yu'(Kas'Xa'St/(Ste+ Kis) ta | =-biXs Decay of active sludge (3.95) Tita | = thie fos Production of easily biodegr material | (3.102) tage | =ot Production of slowly biodegr. material | (3.103) Tye | = To ti Production of adsorbed biodegr. material | G.104) Ta | =te= T= Yivha “Ta Production rate of setive sludge G.100) Tw |=fta Production rate of endogenous residue | (3.105) 0. [=O =fe Yr tf Dety | Oxygen consumption rate @.101) 3.4.3 Application of the general model The most important practical applications of the general activated sludge model are = The determination of the sludge age required to obtain an substantially complete removal of the biodegradable organic material, ie. approaching an ideal activated sludge system - The determination of the O, as a function of time and space (and hence the oxygenation capacity to be installed) in systems with non-ideal behaviour, which may for instance be caused by a series or plug-flow configuration and/or a variation in input flow and -load. In Table 3.13 the values of the kinetic constants are presented, together with the temperature dependencies as determined by Dold, Ekama and Marais (1980). Computer simulations using these values show that the utilisation of organic material in an activated sludge system is nearly complete, even at very short sludge ages. For temperatures >18°C the required minimum sludge age is only 1.5 to 2 days (Van Haandel and Marais, 1981). In practice, and certainly when an activated sludge process is designed for nutrient removal, the sladge age will be much higher than the minimum required for organic material removal. In general, at such a short sludge age the organic material concentration in the effluent of an activated sludge system is higher than the calculated value from the general model. This is not due to a failure in the kinetic expression, but to the inability of the settler to separate the free bacteria (i.e, not attached to the sludge flocs), which are abundant at short sludge ages In regions with a low temperature, the required sludge age for complete removal of organic material is considerably higher and when the process is operated at a short sludge age, part of the wasted sludge will actually be stored organic material that has not yet been metabolised, In this case the activated sludge process functions also as a bioflocculation process Some activated sludge processes are explicitly designed for this and part of the removal of the influent organic material is achieved in the anaerobic sludge digesteraa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 3 - Organic Material Removal 81 For example, the aeration time in SBR systems for nutrient removal is only 20 to 25% of the total cycle time, the remainder being occupied with denitrification and settling, As the total oxygen consumption in the SBR is necessarily equal to that of a conventional system with continuous aeration, the required oxygen transfer rate and therefore also the installed aeration capacity in an SBR will be about 3 o 4 times larger. Another disadvantage that is often attributed to SBR systems is the inflexibility in dealing with flow variations, as the SBR only receives influent during a minor part of the total cycle time. This can be resolved in several ways: for small applications a buffer feed tank can be constructed (as tank volume is much cheaper than reactor volume) while for larger applications a series of parallel SBR reactors can be constructed, which are operated out of sequence with each other. Furthermore at high flow rates a reduction of the cycle time may be temporarily implemented. A recent modification of the SBR jis the granulated sludge bed reactor (Chapter 7). It has been demonstrated that under certain conditions a granular activated sludge may develop in an SBR type of reactor. As the granular sludge has excellent settling characteristics, the sludge concentration can be increased and the settling time reduced. Both result in reduced reactor volume as compared to conventional activated sludge treatment. To cultivate aerobic sludge granules, the preferential growth of substrate accumulating organisms is required, such as bio-P organisms, which will be discussed in Chapter 5 3.5.3 Carrousels The carrousel system is also known as cireulation- of circuit system, to indicate the main difference from a completely mixed or plug-flow system: in a carrousel the mixed liquor is recirculated at relative high speed (e.g. 20 cms") through a long curved reactor of limited Width. The layout of a modem carrousel often resembles a car racetrack, To remain in this analogy, during the time period of a single hydraulic residence time, a specific volume of mixed liquor will cover many “laps” The first of these systems was the so-called oxidation ditch (a design that has become obsolete), which was developed by Pasveer in Holland as a treatment unit to be used by small communities (500 to 1000 inhabitants). In the original and simplest version, all treatment processes were cartied out in a single physical unit. The biological reactor had an ellipsoidal form and consisted of an excavated ditch in which the sludge circulated. This circulation was induced by asurface aerator, which also introduced the oxygen needed for the treatment of the waste water (Fig 3.18a), Using a time controller, the aerator was switched off at regular intervals (normally at night, when the sewage flow is small) and after settling the clarified supernatant was discharged. Hence the operation was similar to a sequencing batch reactor. Any incoming sewage during the settling period was accumulated in the sewage network, As the sludge age was very long (25 to 50 days), the sludge in the system was already well stabilised and could be applied directly on sludge drying beds without any further treatment. After the successful introduction, several new versions of the oxidation ditch were developed to handle larger waste water flows. The first modification was the introduction of an auxiliary ditch where the sludge was settled before discharge, so that the system could be operated continuously (Fig 3.18b). Another alternative was to have divisions in the ditch, so that parts of it could be used as a settler (ig. 3.18c). However, the real breakthrough of the circulation system came when the oxidation ditch was converted in the late seventies of the last century into a system with a dedicated settler: the Carrousel system (Fig. 3.18d), with several thousand units in operation throughout the world. While the first carrousels were all equipped with surface aerators, for the dual purpose of moving the mixed liquor through the reactor and for aeration, newer carrousels may be equipped with dedicated propulsors and diffused aeration systems, which allows them to operate at increased liquid depth (5 - 6m)aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.aa You have either reached a page that is unavailable for viewing or reached your viewing limit for this book.Chapter 4 - Nitrogen Removal 85 Chapter 4 NITROGEN REMOVAL In an activated sludge plant designed for tertiary treatment, the objective is to remove nutrients. suspended solids and organic matter. During the last few decades, the importance of nutrient removal has increased as a result of the necessity to avoid eutrophication of water bodies receiving untreated waste water and the effluent of waste water treatment plants. For thi reason, many new waste Water treatment plants are now designed for tertiary treatment. ‘Apart from the important repercussions on the effluent quality, tertiary treatment also has a beneficial influence on the performance of the waste water treaiment process itself. This is particularly noticeable in the case of nitrogen removal. The development of nitrification in an activated sludge process is practically unavoidable when the sewage reaches temperatures of 22 to 24°C, which will be the case for at least part of the year in tropical and subtropical regions. The formed nitrate can be used by most micro-organisms in the activated sludge as a substitute to dissolved oxygen. In an anoxic environment, characterised by the presence of nitrate and the absence of dissolved oxy gen, the nitrate ion can be reduced by organic matter to nitrogen gas: this process is called denituification If the biological reactor is kept completely aerobic, the nitrified mixed liquor will flow to the settler, where an adequate environment is established for denitrification as soon as the oxygen is consumed: this will take only a few minutes. As a result of the development of denitrification in the settler, nitrogen bubbles will be formed and these will rise to the liquid surface. While rising, the bubbles may attach to sludge flocs, causing the flocs to be lifted to the liquid surface. Thus a floating layer will be formed, composed of sludge with attached nitrogen bubbles, Some of the nitrogen gas will escape, but enough will remain in the layer to keep it floating, which will eventually lead to sludge loss due to unintentional discharge with the effluent. Of course, the presence of sludge in the effluent will lead to unacceptable high concentrations of suspended solids and organic matter. Another negative aspect of floating layers of denitrifying sludge end the discharge of sludge with the effluent refers to the operational stability of the activated sludge process. The sludge loss reduces the sludge age and sludge mass in the system. The remaining sludge mass may be insufficient to metabolise the influent organic matter, resulting in a reduction of secondary treatment efficiency, Furthermore the reduction of the sludge age may lead to wash out of the nitrifiers and hence interruption of the nitrification process. Ifno nitrate is produced, denitrification cesses as well and the problem of floating sludge will disappear Thus favourable conditions are established for a recuperation of the secondary treatment efficiency First, the fast growing heterotrophs will be re-established and efficient organic matter removal will be resumed, Subsequently the slower growing mitrifiers will also retum in the sludge mass in sufficient quantities to nitrify the applied ammonium load. Nitrate will be formed and a new cycle of operational instability will be initiated In contrast, when nitrification and denitrification are controlled and occur as planned in the activated sludge process itself, before the mixed liquor reaches the final settler, this unit will behave as a normal liquid-solid separator. This allows an effluent containing very low concentrations of suspended solids, organic matter and nitrogen to be produced. Nitrogen removal also has important economic consequences. In the nitrification process both oxygen and alkelinity are consumed, while in the denitrification process part of this consumption may be recovered. In the case of municipal sewage, the oxygen demand for nitrification is about '/s,q of the total demand,86 From stoichiometrics, it can be calculated that “Is or 62 percent of the oxygen demand for nitrification may be recovered in the denitrification process. Thus in activated sludge processes with complete biological nitrogen removal (nitrification + denitrification), oxy gen consumption will be ‘/g 1/s= */s or about 20 percent lower than in comparable processes with nitrification only, Since aeration isthe main part of the operational costs in an activated sludge process, a reduction of 20 percent of the oxygen demand is very significant in economic terms. The alkalinity consumption by nitrification may result in a reduction of pH. The magnitude of this reduction depends on the initial alkalinity and the oxidised ammonium concentration, In many cases the pH tends to become unacceptably low and alkalinity addition (for instance in the form of lime: Ca(OH)s) will be necessary. During denitrification, half of the alkalinity consumption for nitrification is recovered. Thus the alkalinity demand is smaller and in many cases after the introduction of denitrification, addition of alkalinity is no longer necessary 4.1 FUNDAMENTALS OF NITROGEN REMOVAL 4.1.1 Forms and reactions of nitrogenous matter Nitrogenous matter in waste waters is mainly composed of ammonium nitrogen (which can be present as gaseous (NH,) and ionic form (NH,")) and organic nitrogen (urea, amino acids and other organie compounds with an amino group), Sometimes waste waters contain traces of oxidised forms of nitrogen, mainly nitrite (NO=) and nitrate (NOs). Different to organic matter, nitrogenous matter can be defined quantitatively and unequivocally by one parameter the nitrogen concentration in its different forms, In practice, spectrophotometric tests and specific ion electrodes are used to determine the concentrations of ammonium, nitrate and nitrite, Organic nitrogen can be determined after its conversion to ammonium nitrogen by chemical digestion. The sum of the organic and ammonium concentrations is called Total Kjeldahl Nitrogen, TKN. Molecular nitrogen gas (N.) = ~+ Boundary of activated Denitrification ' sludge system Nitrate | (N,) Ne! >; 1 1 Nitrification t {Nitrogen in the effluent Ammonium | (N) Nec p |_| (dissolved) ™% > 1 N=NyetNye ‘Ammorification 1 SENS N Ng Assimilation | Organic nitrogen (N.) N, i 9 geny (No) pl 1 Nitrogen in excess sludge (N) (solid) Figure 4.1 Schematic representation of the forms of nitrogenous material and their reactions in the activated sludge processChapter 4 - Nitrogen Removal 87 In the activated sludge process several reactions may oceur, that change the form of nitrogenous matter. Figure 4.1 shows the different possibilities: (a) ammonification or the inverse: ammonium assimilation by the orgenisms, (b) nitrification and (¢) denitrification (a) Ammonification/assimilation Ammonification is the conversion of organi¢ nitrogen into ammonium, whereas the inverse process, the conversion of ammonium into organic nitrogen, is called bacterial anabolism or assimilation, Considering that the pH in mixed liquor is typically near the neutral point of pH= 7, ammonium will be present predominantly in its ionic form (NH,°) and the following reaction equation may be written: RNH, +H,0 +11 > ROH + NH" 4.) (b) Nitrification Nitiification is the biological oxidation of ammonium, with nitrate as the end product. The reaction is mediated by specific bacteria and is a two-step process: in the first step ammonium is oxidised to nitrite by bacterial species such as Nitrosomonas spp. The complementary step, oxidation of nitrite to nitrate, is mediated by species such as Nitrobacter_spp. Both Nitrosomonas and Nitrobacter can only develop biochemical activity in an environment containing dissolved oxygen. The two steps can be written as: NH, +%50, —NOy+H,0+2H" NOs +02 NOs NH +20, —NO; +H,0+2H™ (4.2) (©) Denitrification Denitrification is the biological reduction of nitrate to molecular nitrogen, with organic matter used as a reductor. For organic matter with a general structural formula C,H,O,, the half reactions of this redox process can be expressed as (1) Oxidation reaction: C\HjOz+ 2x -2) H3O0 x CO + 4s +y = 22) H+ Gx ty - 22 e (2) Reduction reaction: e+ 95H +¥,NOe + MyN+/sH.0. (3) Overall redox reaction: CHO, + (Ax ty - 22/5 H+ ax +y - 2z)/5 NOx x CO2 + (2x + 3y - Z)/5 H:0 + (4x +y -22)/10 Ny 43) The TKN concentration in municipal sewage typically is in the range of 40 to 60 mg N-I", i.e, a fraction in the range of 0.06 to 0.12 of the influent COD. About 75 percent of the total TKN concentration is ammonium nitrogen and the remaining 25 percent is predominantly organic nitrogen.88 In the activated sludge process, organic nitrogen is converted rapidly and almost quantitatively to ammonium nitrogen (ammonification). If nitiification occurs and the oxygenation capacity is sufficient, the oxidation of ammonium nitrogen will be almost complete, If afier nitrification the formed nitrate is removed by denitrification, the total nitrogen concentration in the effluent is in general smaller than 5 to 10 mg NI, with a nitrogen removal efficiency of 90 percent or more. 4.1.2 Mass balance of nitrogenous matter Fig. 4.1 shows that nitrogen may leave the activated sludge process in one of the following forms - As solid matter in the excess sludge (N;). Depending on the liquid-solid separation efficiency of the secondary settler, part of the nitrogen in the excess sludge may actually leave with the effluent suspended solids (Nocp). However, as Nocp << Ni, for mass balance purposes Nj is not corrected for the value of Nos (the consequences of this choice are discussed in Section 4.4.2.3), - As dissolved matter in the effluent: ammonium (N,.), nitrate/nitrite (Nj) and soluble organic nitrogen (Nos, oF actually Now): - As gaseous material (molecular nitrogen) to the atmosphere (N>). In Fig. 4.1 the possibility of ammonium volatilisation is not considered because this process can only have importance when the pH approaches a value of 9 or more and a significant fraction of the ammonium is present in the unionised form. In practice such a situation can only develop under very special conditions Although the quantity of particulate nitrogen leaving with the effluent may be ignored for mass balance purposes, in those cases where strict nitrogen limits apply, the contribution of Newp to the total nitrogen effluent concentration may in fact be quite significant. The volatile suspended solids concentration in the effluent of a well designed final setiler is typically between 5 - 10 mg VSS.I". resulting in an particulate organic nitrogen concentration between 05 and 1,0 mg N.I", Using the concepts developed in mass balance calculations for organic material (refer to Section 3.2), the nitvogen recovery factor can be defined as the ratio of the nitrogen mass flaxes leaving and entering the activated sludge process: B, = (MN\+ MN, + MN/MNg (44) Where: B, recovery facior for nitrogenous material (-) MN, flux of nitrogenous matter in the excess sludge (kg Nd") MN. flux of nitrogenous matter in the effluent (kg N.d") MNa flux of denitrified mitrogen (kg N.d*) MN flux of nitrogenous matter in the influent (kg N.d") Eq. (4.4) is only useful, when the different fluxes are formulated in terms of measurable parameters, so that the B, value can be determined experimentally and compared to its theoretical value of one, For the nitrogen flux in the excess sludge, an expression was already derived in the previous chapter (Section 3.2.3.3) MN, = =fyMX/R, (4.5)Chapter 4 - Nitrogen Removal 89 ‘The fluxes in the influent and the effluent are easily calculated as’ MN; =Q(NatNat+Nu) =QeNa 4.6) MNe = Qi(Noet+ Newt Nne) = QiNee (47) ‘Where: N = total nitrogen concentration (mg N.I") N, = ammonium nitrogen concentration (mg N.I") Ng = organic mitrogen concentration (mg N.I") Na = nitrate nitrogen concentration (mg NI) The indices “i” and “e” refer to influent and effluent respectively. In Eqs. (4.6 and 4.7) the nitrite concentration is assumed to be insignificant, which is usually justified in practice. If this is not the case, then this indicates a process disturbance that should be remedied Inorder to calculate the denitrified nitrogen flux, the process configuration must be taken into consideration, When the objective of the process is nitrogen removal, there will be anoxic zones where denitrification takes place. The flux of removed nitrogen is calculated as the product of the flow passing through the anoxic reactor and the decrease of the nitrate nitrogen concentration in it, Hence: MNa = QvANu 48) Where: MN = flux of denitrified nitrogen in anoxic reactor “k” (kg N.d!) Q. = flow rate to reactor “k” (m*.d") = influent flow plus return sludge and possible other recycle flows AN, = nitrate-N concentration difference between inlet and outlet in anoxic reactor “1 Ifa system has a total aumber of “k” anoxic reactors, the total denitrification nitrogen flux can be expressed as K K MNe = =MNa = 2(Qr Nu) 4.9) kl Now, using the expressions of Eqs. (4.5, 4.6, 4.7 and 4.9): K Bu = (fy MXR, + QrNict E QeANu MQ Ni) (4.10) k=1 In Eq, (4.10) all parameters on the right hand side are measurable so that it is possible to calculate the nitrogen recovery factor based on experimental data.ol) Example 4.1 ‘As an example of the application of mass balance concept for nitrogen removal, the experimental data obtained during the operation of a nitrogen removal pilot plant are discussed (Ekama et al, 1976). The process was composed of five reactors each with volume 5 m° and a settler. The first reactor (receiving all the influent) was kept anoxic, whereas the other four were aerated, The average temperature was 21.6°C. Fig. 4.2 shows the flow scheme of the process, Excess sludge (1,4 m' tntuent —|v,=5m'|_|v,=5m'|_|vy=sm|_|vy-sm'|_|vz-5m’ Eftuent | Sattler 4am] | (roxie}| | (Aerobic)/ | (Rerobic)| | (Aerobic | (Aerobic) sae nid’ Return sludge (120 m'.d") Figure 4.2 Flow scheme of the pilot process from Example 4.1 Table 4.1 shows the average values of the experimental determinations that were carried out over a period of 18 days, after a steady state had been established. Fig. 4.1 and Table 4.1 contain all information necessary to perform the mass balance calculations as shown below. Table 4.1 Avorage values of procoss paramoters in the pilot oxporimont of Example 4.1 Parameter Infiuent Reactor 1 2 3 4 5 [ cop | mgt 47 25 19 18 18 18 TKN | mgN7 45.1 92 48 33 26 20 NH¢ | mg NI 32.8 94 | 34 06 0.3 02 NO; | mgNJ* <0.3 12 5.8 92 97 98 vss_| mgvss.i" 25 | 2447 | 2466 | 2406 | 247 OUR | mg O2I".h" 50 | 593 | 365 | 234 | 193 Note: (1) Average of the five reactors. Solution The following calculation procedure is out to determine the nitrogen recovery factor Bn (1) Calculate MNs (Ea. (4.3)) Nu = Qi(No + Nai+ Na) = 40(45.1 + 0.3) = 1816 g N.d* (2) Calculate MN (2) Calculate the sludge age (Eq. (3.15) Re = Wiq = 25/1.4 = 18 daysChapter 4 - Nitrogen Removal 1 (b) Calculate MIN (Eq. (4.5) MINi= fi Ve XIRs (3) Calculate MN. 1 25:2469/18 = 343 gN.d* MINie = Qi (Noe + Nae + Nne) = 40:(1.9 + 8.7) = 424 g N.d™ (4) Calculate MNg The data indicates that denitrification took place in the first reactor and in the settler (in these two the nitrate concentration decreased). The nitrate concentration entering into the first reactor Nno is calculated as the weighted average of the concentrations in the influent- and recycle flows. Assuming the nitrate concentration in the retum sludge flow is equal to the effluent nitrate concentration, one has: Nno = (QiNoi + QrNpeV(Q+ Q,) = (40 .0.3+ 120:8.7)(40 + 120) = 6.6 mg N.* As the nitrate concentration in the flow leaving the first reactor was 1.2mgN.I" (Table 4.1), the decrease equals AN; = 66 - 1.2= 5.4 mg NI”. Hence the flux of removed nitrogen in the first teactor was: MNgy = (Q) + Qy)-ANns = (40 + 120):5.4 = 864g Nd? Similarly the flux of nitrogen removed in the settler is calculated as: MNea = (Qi + Qi ANna = (40 + 120)-(9.8 - 8.7) = 176 gN.d* Where: MNea = denitrified nitrogen in the settler (g N.d'") ANng = decrease of the nitrate concentration in the settler (mg N.I") Now the total flux of nitrogen removed by denitrification can be calculated as: MINe = MNe: + MNe = 864+ 176 = 1040 g N.I" Having calculated all the relevant flows, the nitrogen recovery factor can now be determined with the aid of Eq. (4.1): Bn = (MI + MNte + MNgYMINy_ = (343 + 424 + 1040)/1816 = 1807/1816 = 0.995 In the example there is a difference of only 0.5% between the experimentally determined values of the nitrogen fluxes entering and leaving the pilot plant. This shows that the experimental procedures to obtain the data have been correct and for this reason the data is attributed a high degree of reliability. Itis interesting to note that once one has established that the nitrogen mass. balance closes, itis also possible to determine the recovery factor for the organic material. To do this, first the three fractions mS., mS,, and mS, must be calculated In the case of the above example one has: MS, = S/Sj= 18/477 = 0.038 MSxy = fay MEV = fey Xy-Ri/(Re Se) = 1.5.2469-0.625/(18-477) = 0.2702 The value of mS. is calculated as the sum of the oxygen consumption for organic matter and the equivalent oxygen recovered in the denitrification process: mS, = MSJMSt= (MO. + MOec)/MSt In the above expression, the oxygen consumption for the oxidation of organic matter is the difference between the total consumption and the consumption for nitrification. The total consumption in the four aerobic reactors (with a total volume of 20 m°) is: MO; — = Vy(OUR + OUR» + OUR + OURs) 5:(59.3 + 36.5 + 23.4+ 19.3)24 = 16,620 g 02d" In order to calculate the oxygen consumption for ritrfication, first the flux of nitrified ammonium is determined as the difference between the TKN flux in the influent and the fluxes leaving the system in the effluent or the excess sludge. Knowing there is an oxygen consumption of 4.57 mg Q2 per mg N nitrfied, one has in the case of the example MO, 4.57-(MNs~ MNos~ MNi~ MNeo ~ MN) 457-(1,816 - 40-0.3 -343- 40-19) = 63299 O.d* Hence the oxygen consumption for oxidation of organic matter is: MO, = MO,-MO, = 16,620 ~ 6329 = 10,291 g Oa" The equivalent oxygen recovered in the denitrification process is easily determined as: MOzg = 2.86-MNg= 2.86-1040 = 2974 g On." Now, the fraction of influent COD that is oxidised in the activated sludge is determined as: MS, =(MOc+MOzg/MSs = (10,291 + 2974)/(40-477) = 13,210/19,080 = 0.695 Finally, the recovery factor for organic matter can be calculated as: B, mS, + MSyy + mS_ = 0.038 + 0.270 + 0.695 = 1.003 It can be concluded that the mass balance for organic material also closes: the experimentally determined recovery factor of 1.003 is practically equal to the theoretical value of 1.000. In practice it should be expected that the recovery factors By and B, deviate more from the theoretical value than in the above example. The main reason is that most activated sludge processes are not completely in steady state conditions.Chapter 4 - Nitrogen Removal 93 4.1.3 Stoichiometrics of reactions with nitrogenous matter 4.1.3.1 Oxygen consumption Only nitrification and denitrification are of interest when calculating the oxygen consumption by nitrogenous matter. Fig, 43 schematically shows the electron transfer in the nitrification- and the denitrification processes, Denitrification 5 electrons per N-atom = 2,86 mg O,mgN”* No, NO3 1 1 2 3 4 Component NH," Oxidation number -3 9-2 A 1 1 Nitrification: 8 electrons per N-atom --&--- = 457 mg O,.mgN Figure 4.3. Variation of the nitrogen oxidation number in the processes of nitrification and denitrification. In the nitrification process, the oxidation number of the nitrogen atom in ammonium increases from -3 to +5 by the transfer of 8 electrons to the electron acceptor (oxidant): ie oxygen These electrons are accepted by two molecules (four atoms) of oxygen (thereby changing its oxidation number from 0 to -2). Hence, for the nitrification of | mol of ammonium nitrogen (14 g N), there is a demand for nwo moles (64 g) of oxygen, so that the stoichiometric oxygen consumption can be calculated as 64/14 or 457 mg O2mg N™ Inthe denitrification process, nitrate (oxidation number +5) is reduced by organic matter to molecular nitrogen (oxidation number 0), so that 5 electrons are transferred per nitrogen atom, Hence, of the 8 electrons released by nitrogen in the nitrification process, 5 electrons are recovered when nitrate is reduced to nitrogen. Thus, in oxidimetric terms, the nitrate has an oxidation capacity of ‘/, of the oxygen used in the production of the nitrate by nitrification. In other words, a fraction of °/s or 0.625 of the oxygen consumption in the nitrification process can be recovered as “equivalent oxygen” in the process, ic. 0.625-4.57 = 2.86 mg O»mg N". It can be concluded that there is anet oxy gen consumption of 4.57 - 2.86 = 1.71 mg Oymg N" during complete biological removal of nitrogen, Example 4.2. ‘Awaste water contains a COD concentration of 600 mg," and a TKN concentration of 60mg." it is assumed that 10 percent of the influent COD is discharged with the effluent, 30 percent leaves the system in the excess sludge and the TKN concentration in the effluentis 3 mg N.I" Determine the fraction of the oxygen consumption necessary for nitrogenous matter in the cases of (a) nitification and (b) nitrification and denitrification.4 Solution (1) Calculate the oxygen consumption for the removal of organic matter: Oc = (1- MSe- MSy)Sq = (1 = 0.1 - 0.3)'600 = 360 mg O27 (2) Calculate the nitrogen concentration (expressed as mg N.I" of influent), leaving the system together with the excess sludge: Ni =f, mB, Sp= 0.4-(0.3/1.5)600 = 12 mg NI* Calculate the nitrified TKN concentration: No = Nai Noi Nae ~ Nop -Ni=60 + 0-3-12= 45 mg NI" (4) Calculate the oxygen consumption for nitrification: On = 4.87 Ne = 206 mg Oz!" (5) Calculate the equivalent oxygen recovered in the denitrification process Oey = 2.86-Ne = 129 mg O27 Itis concluded that in the case of nitrification, the total oxygen consumption is Or= Os + On = 360 + 206 = 566 mg N.I" of which a fraction O,/O, = 208/566 = 0.36 is consumed for the oxidation of ammonium. In the case of nitrification followed by denitrification, the oxygen consumption decreases to O:= O. + On - Oaq= 360 + 206 - 129 = 437 mg O2.I"' and the fraction of the oxygen consumed by the nitrogenous material would decrease to (437/360)/437 = 0.18. In Example 4.2 the inclusion of denitrification in the process contiguration reduces oxygen consumption from 566 to 437 mg Oz", or a reduction of 23 percent. 4.1.3.2 Effects on alkalinity The effect of ammonification, nitnfication and denitrification on the (carbonate) alkalinity of the mixed liquor can be calculated from simple stoichiometric relationships using the reaction equation of the three proce 's (Eqs. 4.1, 4.2 and 4.3), It can be noted that in all equations the hydrogen ion is involved: in the ammonification provess and the denitrification process there isa consumption of | mol H' per mol N, whereas during nitrification there is a release of 2 moles H' per mol N, Knowing that the formation of I mol of H’ (mineral acidity) is equivalent to the consumption of | mol of alkalinity or % mol of CaCOs (50 g CaCO3), the following alkalinity changes are calculated = Ammonification process: production of 50 g CaCO3per mol N, - Nitrification process: consumption of 2:50 = 100 g CaCOs per mol N, = Denitrification process: production of 50 g CaCO per mol N. The alkalinity changes may be expressed as (AAIK/AN)m = 50/14 = 3.57 mg CaCOsmg N? (41a) (AAIK/AN), 100/14 =-7.14 mg CaCO3.mg Nt (4.11b) (AAIK/AN), 50/14 = 3.57 mg CaCO;mg N! (4.1) Where (AAIK/AN) = alkalinity variation in mg CaCO; per mg N Indices am, n and d refer to ammonification, nitrification and denitrification respectivelyChapter 4 - Nitrogen Removal 95, In the case of municipal sewage, the alkalinity effect of ammonification is usually very small, as the following analysis shows, The concentration of ammonified nitrogen in the activated sludge process is given by the difference between the organic nitrogen in the influent and the sum of the organic nitrogen concentrations in the effluent and the excess sludge (see Fig. 4.1) so that: AN =Na-No-N (4.12) AN. = ammonified nitrogen concentration in the activated sludge process (mg N.I" influent). Normally in the case of municipal sewage, the organic nitrogen concentration in the effluent is very small, only | or 2mg NI", whereas the values of the N,, and N; concentrations are both approximately equal to 25 percent of the influent TKN concentration. Hence the variation of the organic nitrogen concentration in the activated sludge process is very small, Consequently the associated alkalinity change will also be small and can be expressed as Alka = 3.57-ANen (Nox ~ Noe Ni) (4.13) The alkalinity change due to nitrification is calculated from the concentration of nitrified ammonium. This concentrion is equal to the difference of the influent TKN concentration (Ni) and the sum of the TKN concentrations in the effluent (Nj.,) and the excess sludge (N)): Ne =Nu-Nu-Ni (4.14) Where N, = influent ammonium concentration, nitrified in the system (mg NI") ‘The effect of nitrification on alkalinity can be expressed as: AAI, = -7.144 7.14-(Njs - Nix “ND (4.15) ‘The alkalinity change due to denitrification depends on the removed nitrate concentration, This, concentration can be calculated as ANg = Na +Ne- Now Hence the alkalinity change due to denitrification can be expressed as AMAlky =3.57-ANY=3.57(Nu +Nui~ Nie= Ni= Now) 4.16) ‘The total alkalinity change in the activated sludge process, due to the reactions of nitrogenous matter, will be equal to the sum of the effects calculated for ammonification, nitrification and denitrification. AAIK, = AAlKam + AAIK, + AAIKa B.57+(Nei ~ Noe - Ni) - 7.14-(Nig - Neo - Ni) + 3.57-(Nai + Nis = Nae Ni--Noe) (4.17)96 Knowing that the TKN concentration (N,) is the sum of the concentrations of organic (N) and ammonium nitrogen (N,), Eq. (4.17) can be simplified to AAlk, ==3.57-(Ny-Nas-Nuc+ Nox) = 3.57(AN, - AN,) (418) Where: AN, = variation of the ammonium concentration in the activated shudge process AN, weave the nitrate concentration in the activated sludge process (mg N.I") All parameters on the right hand side of Eq. (4.18) can be determined experimentally by standard tests. Hence it is possible to calculate the stoichiometric alkalinity change due to the combined effect of ammonification, nitrificetion and denitrification in the activated sludge process. On the other hand, itis also possible to measure the alkalinity change directly In Fig. 4.4 the calculated (according to Eq. 4.18) and the experimentally determined values of alkalinity change in different activated sludge processes have been compared. The data in Fig. 4.4 refers to systems with « different extent of nitrogen removal: (a) without nitrification, (b) with nitrification but without denitrification and (c) with both nitrification and denitrification. Fig. 4.4 shows that in all cases there is an excellent correlation between the calculated and the experimentally observed elkalinity changes, for a very large range of changes (AATk between -600 and +100 ppm CaCOs) and for very diverse operational conditions, Thus, the conclusion is justified that the alkalinity change in an activated sludge process is due to the stoichiometric effets of the reactions with nitrogenous material, ammonification, nitrification and denitrification 8 20 3.57-(AN, - AN, + 10) (427) Alk, = influent alkalinity (mg CaCO.) Alk, = effluent alkalinity (mg CaCO3.") Example 4.4 What would be the minimum required alkalinity of the sewage in the system of Example 4.1 to ensure a stable and neutral pH value? Solution In Table 4.1 it can be noted that Nai = 32.8; Nae = 0.0, Nni= 0.3 and Nre = 8.7 mg NI". Hence by using Eq, (4.27) the minimum influent alkalinity is cslculated as: Alki > 3.57 (8.7 - 0.3 - (0.0 - 32.8) + 10) = 183 ppm CaCOs In practice the neutral alkalinity of waste waters may be less than the value required to maintain a stable pH in the activated sludge process. This is a particular risk when nitrification without subsequent denitrification occurs in the process. In such cases it is necessary to increase the influent alkalinity, which is usually done by lime addition, Without the addition of alkalinity, the behaviour of the activated sludge process will be irregular; there will be periods with nitrification and the consequential decrease of alkalinity and pH, until a pH value is established that is inhibitory for nitrification. When nitrification ceases, alkalinity automatically increases and pH rises, so that once again favourable conditions for nitrification are established and a new cycle of instability is initiated. If denitrification is included as a treatment slep, the decrease of alkalinity will be smaller and often there will be no need for lime addition, because the denitrification process produces alkalinity100 4.2 NITRIFICATION Nitrification is a two-step biological process, but only the first step - oxidation of ammonium to nitrite - is of importance for the nitrifier kinetics in an activated sludge system. When the nitrifying population is well established in the activated sludge process the second step, oxidation of nitrite to nitrate, is always so fast that it can be considered as instantaneous for all practical purposes. Consequently, the nitrite concentration in the effluent of activated sludge processes is in general very small. 4.2.1 Nitrification kinetics Downing et al (1964) were the first to show that the growth of Nitrosomonas in the oxidation process of ammonium can be described by the Monod equation (X/dt) = (dX,/dt), + (dX,/d) with (AXYdg =WXp= HaeXaNM(N, +K,) and EX fAq > -byX (4.28) Where: Xn = Nitrosomonas concentration (mg VSS.1') (dX,/dt) = rate of change of the Nitrosomonas concentration (mg VSS.I'.d"!) net growth rate of Nitrosomonas (dX,/d), = Nitrosomonas growth rate (mg VSSI.d") (dX,/dt)y = Nitrosomonas decay rate (mg VSS.I".d") nm = specific growth rate for Nitrosomonas (d"') Ha maximum speeifie growth rate for Nitrosomonas (d") by decay rate for Nitrosomonas (d"') Ka = Monod half saturation constant (mg N.I") In the Monod equation, the parameter jt represents the growih rate of the micro-organisms per time unit, For example, a value of 1 = 0.6 d! means that the daily rate of micro-organism synthesis is equal to 60 percent of the mass initially present, Equation (4.28) shows that the value depends on the substrate concentration N,. At high N, concentration (saturation) the maximum growth rate ji is attained. The constant K,,is equal to the substrate concentration for Which jt = % 41, and for that reason is called the “hall” saturation constant. The basic equation of Downing et al (1964) can be used to calculate the residual ammonium concentration in a completely mixed. steady state activated sludge process. Under these conditions, there is no variation of the mass of Nitrosomonas in the system: the net growth rate (defined as the growth rate minus the decay rate) is equal to the discharge rate due to abstraction of excess sludge Hene (dX,/dt) = 0 = (dX,/dt), + (AX./dt)a+ (dX/dt). (4.29) The rate of change of the Nitrosomonas concentration due to the discharge of excess sludge (EX pit), can be expressed as (refer also to Section 3.2.3.3) (dXqdt).= - X/ReChapter 4 - Nitrogen Removal 11 Now, using Eqs. (4.28 and 4.30) in Eq. (4.29) one has: (dXJ/dt) Har Xa NV(Na + Ky) - by Xa- XW/Ry Han’ No/(Na + Kn) - ba VRe (4.30) Or, by rearranging Ny = Kyr(by + VR,)ifHin = (Py + URI (431) Equation (4.31) gives the ammonium concentration in the mixed liquor of a completely mixed activated sludge process and hence also in the effluent of this process. This residual is indicative of the efficiency of the nitrification process, depends on the numerical values of the three kinetic parameters’ jt, Ky and b, and one operational variable the sludge age R,. It is interesting to note that the residual ammonium concentration does not depend on the initial concentration. Furthermore the residual ammonium concentration can never be superior to the ammonium concentration that is available for nitrification, the nitrification potential Np, that will be discussed in Section 4.2.3. Therefore this condition defines the minimum sludge age for nitrification as concentration, which Ny =Ke(bo+ VRaV(Hn ~~ VR) oF Ra = (+ K YN, [Pm = bed + KN] (4.32) When the activated sludge system is operated at the minimum sludge age for nitrification, then for domestic sewage the ammonium concentration available for nitrification will always be much greater than the half saturation value K,, In that case the ratio K,/N, « | and Eq, (4.32) can be simplified to Re = Mm - ba) (4.33) Equation (4.33) expresses that nitrification will not develop if the sludge age is shorter than a minimum value of Ry = I/(jty ~b,), because the rate of Nitrosomonas discharge in the exe sludge will exceed the net growth rate, When the sludge age R, is higher than the minimum value, nitrification will develop and its efficiency will depend on the sludge age and the kinetic constants Kp, ply and by. After Downing’s work, many research workers have carried out experimental investigations to determine the kinetic parameters for nitrification in the activated sludge process Table 4.2 a, b and ¢ show experimental values of jt, b, and K,, obtained by different researchers for Nitrosomonas. It can be observed that the numerical values obtained by different authors have a very large spread. This may be attributed partially to differences in the experimental methods, but certainly the fact that different waste waters have been used must have had an influence. Thus it can be concluded that the value of the kinetic parameters of the nitrifiers depends on the origin of the waste water. Ideally these values should be determined for each specific design case.102 Table 4.2a Values of p1,of Nitrosomonas as determined by various authors Hair @") TCO Mae") | Reference 0.33 15 0.66 Bamard (1991) 0.47 15 045 Kayser (1991) 0.33 20 033 Downing et al (1964) 0.33 - 0.65 20 033-065 | Ekama et al (1976) 0.34 - 0.40 12 086-101 | Gujeret al (1974) 0.45 15 073 Eckenfelder (1991) 0.40 -0.50 14 080-100 | Gujer (1977) 0.50 20 050 Lawrence et al (1973) 0.33 25 026 Sutton et al (1979) 057 16 076 Gujeret al (1974) 0.94 29 033 Lijklema (1973) 1.08 - 1.44 23 0.76- 1.02 | Poduska et al (1974) Table 4.2b Values of b, of Nitrosomonas as determined by various authors byr (-1) TCO Dyao(d) Reference 0.0 20 00 Downing et al (1964) 0.0 15 00 Downing et al 1964) 0.0 10 00 Gujer (1979) 0.04 20 004 Ekama et al (1976) 0.12 29 0.09 Lijklema (1973) 0.12 B oul Poduska et al (1974) Table 4.2c Values of K, of Nitrosomonas as determined by various authors Kar(mg.t) TCO, Kao (mgt) Reference 0.0 2B 0.04 Poduska et al (1974) 0.2 15 Ol Downing et al (1964) 0.2 20 02 Downing et al (1964) 0.2 10 06 Guyer (1977) 05 4 10 Ekama et al (1976) 10 20 10 Ekama et al (1976) 10 20 10 Lijklema (1973)Chapter 4 - Nitrogen Removal 103 (@) Influence of 1, (in d') (b) Influence of b, (in d') (c) Influence of K,,(in mg.! ') Be 4 $8 10 Sludge age (4) Sludge age (4) Sludge age (d) Figure 4.6 Residual ammonium concentration as a function of different values of the Kinetic parameters tq, by and Ky In order to be able to compare the data collected at different temperatures, all values have been corrected to a standard value at 20°C, using the temperature dependencies as determined by Ekama and Marais (1976). ptye= fty2o'l 123; bay = bygo'l 04 and Kyr= Kygotl 123. In order to evaluate the influence of the numerical values of the kinetic parameters for nitrification on the efficiency of the process, the following procedure has been followed: the values in Table 4.2 suggest average values at 20°C of p= 0.4 d'"; by = 0.04 d" and K,= 0.5 mg NI". The influence of the values of these parameters on the residual ammonium concentration is shown in Figs. 4.6a, b and ¢. Figure 4.6a shows the residual ammonium concentration N, as a function of the sludge age for average b,, and K,, values (b,, = 0.04 d and K, = 0.5 mg NI") and for two values of jt. one extremely high (jt, = 0.8 d) and the other extremely low (jin = 0.2 d"}), Hence, the difference between the curves for the residual ammonium concentration in Fig. 4.6a reflects the influence of the different j1,, values (the curves were calculated using Eq. 4.31). Similarly, in Fig 4.6b, the influence of the value of the decay rate b,, is analysed for average values of the otherkinetie parameters: jt, = 0.4 d" and K,, = 05 mg N.I". The residual ammonium concentration N, is calculated as a function of the sludge age for a very high value of the decay rate (b, = 0.1 d"!) and without decay rate at all (b, = 0.0 d"), The difference between the two curves is due exclusively to the variation of the bj value, Finally, in Fig. 4.6c the influence of the K, value on the residual ammonium concentration is evaluated, For average values of the other two parameters (jtq = 0.4 d" and by = 0.04 d") curves were drawn for N, as a function of R, for K, = 2 mg N. I" (very high value) and K,, = 0.00 mg N. I" (very low value). From Fig. 4.6 the following conclusions can be drawn - The influence of jy on the residual ammonium concentration - and hence on nitrification efficiency - ismuch more pronounced than that of the other two parameters b,, and Ky; - Forsludge ages of more than 50 percent beyond the minimum sludge age for nitrification Rew the residual ammonium concentration is so low that for practical purposes nitrification may be considered to be complete. As the minimum sludge age for nitrification depends mainly on the value of pp, it is necessary to analyse why such large differences in the values of jig are reported in Table 4.2a.104 The numerical values of the parameters b, and K, are of minor importance. The factors influencing the jy, value can be divided in Wo categories, (1) factors related to the origin of the waste water and (2) factors related to the operational conditions in the activated sludge process In so far as the origin of the waste water is concerned, there are several compounds that are known to inhibit nitrification. There are clear indications that the jy, Value depends on the fraction of industrial waste in municipal waste water. In the case of a small industrial contribution, the ty, value is determined in the range of 0.5 to 0.7 d at 20°C, but this value decreases to 0.25 to 0.3 d" or even lower when the proportion of industrial wastewater in the total influent is higher. Wilson and Marais (1976) measured an jim value of 0.17 d" for a predominantly industrial waste. In the case of purely industrial waste waters, the jty, may be very small: a research project at CETREL in Brazil, where petrochemical wastes are processed, showed a j1,, value of less than 0.1 d” at a temperature of 26°C, which is equivalent to fly < 0.05 d° at 20°C. The dominant influence of the origin of the waste water on the 1, value indicates that this parameter should be seen as a sewage characteristic rather than a kinetic constant, In so far as operational conditions are concemed, the following factors have been shown to influence mitrification and particularly the }1,, value: temperature, dissolved oxygen concentration and pil (a) Temperature There is a strong influence of temperature on the jtq, Value as the experimental results obtained by several authors show. Often a simplified Arhenius equation is used to describe the influence of temperature, ic. Har =H 67 (4.33a) Where 0 = Arhenius temperature dependency coefficient Table 4.3 shows the experimental results of some authors. The @ value ranges from 1.11 to 1.13; which means that the 41, value increases by 11 to 13 percent per °C of temperature increase. Hence, the jg, value doubles for every 6 to 7°C C of temperature increase Table 4.3 Temperature dependency of the maximum specific growth rate of Temp. factor (0) Temperature interval (°C) Reference 1.116 19-21 Gujer (1977) 1.123 15-20 Downing et al (1964) 1.123 14-20 Ekama et al (1976) 1.130 20-30 Lijklema (1973) The influence of the temperature on the growth rate of the mitrifiers has an important repercussion on the activated sludge process. In regions with a moderate climate, waste water temperatures in winter are in the range of 8 to 14°C, resulting in a low value of fly. For a medium value of jty of 0.4 d"' at 20°C, one would expect values 0.2 d” at 14°C and 0.1 dat 8°C. From Eq, (4.32), itis calculated that the minimum sludge age for nitrification in this case will be in the range of 6 to 14 days. Therefore in Europe, it is common that activated sludge processes for nutrient removal are operated at a sludge age of more than 15 days,Chapter 4 - Nitrogen Removal 105 In contrast, in tropical regions water and sewage temperatures are high. For example. in Campina Grande in North East Brazil (a.ka. the Queen of the Borborema Heights), the average temperature is 26°C during summer. If again it is assumed that p1,, = 0.4 d"! at 20°C, the In Value at sewage temperature is calculated as ut, = 0.8 d" at 26°C so that the minimum sludge age for nitrification is now only Ry, = 1.25 days. In practice, the activated sludge process will be almost invariably operated at a longer sludge age, so that nitrification will develop if enough oxygenation capacity is available. (b) Dissolved oxygen concentration The influence of the dissolved oxygen concentration on nitrification kinetics has been the object of several studies. Several authors have proposed a Monod type equation to incorporate the influence of the dissolved oxygen concentration (Stenstrom and Poduska, 1980), In the IWA models no. 1 and 2, this approach has also been followed. Both ammonium and dissolved oxygen are considered substrates and the maximum specific growth rate is expressed as = HeNV(N,HK,)DOMDO+K,) (434) Where: DO = dissolved oxygen concentration (mg O21") K, = half saturation constant (mg O21) The value attributed to K, varies considerably between different authors and values ranging from 0.3 to2.0mg Ost" have heen published. This wide range may be due to the fact that itis only possible to determine the dissolved oxygen concentration in the bulk of the liquid phases. In the sludge flocs, where consumption occurs, the dissolved oxygen concentration is lower than in the bulk, The oxygen consumption creates a concentration gradient trom the floc surface (where the dissolved oxygen concentration is considered to be equal to the bulk concentration) to the centre, Fig. 4.7 schematically shows the dissolved oxygen concentration profile in a sludge floc as a function of the distance to its centre (a spherical floc is assumed), 1 Floc diameter I | 208 foo) —-j SuerettiealD0. ! Gres 00g Subcrtcal co By & = Fou Distanceto| 2 — Ekama etal (1978) floccentre Downing et al (1965) ae Malan et al (1966) == Gavwjer et al (1975 Figure 4.7 Figure 4.8 Dissolved oxygen (DO) concentration Influence of the pH on the gradient as a function of distance from the nitrification rate according to EPA floc surface (1976)106 Depending on the existing bulk dissolved oxygen concentration and the rates of dissolved oxygen transport and -consumption within the floc, anoxic micro regions may develop in the floc centre, where no dissolved oxy gen is present and where, as a consequence, no nitrification will take place, Instead denitrification may develop. This phenomenon is called simultaneous denitrification and is often observed in circulation systems such as the carrousel, which essentially is a completely mixed system (for all components except oxygen) in which the mixed liquor is subjected to an oxygen gradient over the length of the reactor, The minimum bulk dissolved oxygen concentration that is required to maintain the centre of the flocs in an aerobic state depends on several factors such as floc size, stirring intensity, temperature and the oxygen uptake rate As these factors may differ significantly between different active sludge processes, the required minimum dissolved oxygen concentration will vary as well. In general a bulk dissolved oxygen concentration of 2 mg 0,1" should be sufficient to prevent oxygen limitation in the nitrification process (© Mixed liquor pH Several authors have found approximately constant ply, Values over plT range from 7 to 8.5. For pH values below this range, the value of pi,, decreases rapidly, as shown in Fig. 4.8. In practice, many waste waters (e.g. municipal waste water) have a pl value between 7 and 8, In the activated sludge process this value tends to decrease, because of the consumption of alkalinity due to nitrification and an increase of acidity due to the production of CO, from the oxidation of organic matter. For this reason, unless the influent contains a high alkalinity, the mixed liquor pH will be less than 8. Hence, only the pH lower limit of mixed liquor is of practical importance, As discussed in Section 4.1.2, a pH below 7 can be avoided by maintaining the alkalinity above a minimum value of 35 ppm CaCO3 4.2.2 Nitrification in systems with non aerated zones In activated sludge processes designed for biological nitrogen removal. part of the reactor volume is not aerated in order to allow for denitrification. ‘The presence of these anoxic zones influences the nitrification efficiency, because the nitrifiers can only grow in an aerobic environment If it is assumed that the decay of the nitrifiers is not affected by the presence or absence of dissolved oxygen, the effect of anoxic zones on nitrification can be evaluated as follows: in a steady state system the total nitrifier mass MX, has a constant value and can be expressed as: aMX,/dt = = (MX, /d), + @MX,/d)y + (AMX,/D), (435) Indices g, d and e refer to growth, decay and excess sludge discharge respectively. Since the nitrifiers only grow in an aerobic environment one has AMX, = (1-£) Vr(dXi/dte = (fo Vert Xn, (436) f, =anoxie sludge mass fraction V, =biological reactor volume (aerobic plus anoxic zones) By substituting Eq (4.36) in Fq.(4.35) and using Eqs (4.28 and 4.29) one has: Ny =Kerlba+ 1/RS)/[(-f)'Ha ba V/Rs (437)Chapter 4 - Nitrogen Removal 107 The expression for the residual ammonium concentration in a process containing anoxic and aerobic zones (Eq. 4.37) is very similar to the one derived by Dowaing for the completely aerobie process (Eq. 4.30). When the two equations are compared, it can be noted that the presence of the anoxic sludge mass fraction f, has the effect of a reduction of the t., value by a factor (1 - f,) te. Be = (1L- fH (438) Where wt’, = apparent maximum nitrifier growth rate in systems with non aerated zones (d') Fig. 4.9 shows the residual ammonium concentration as a function of the anoxic sludge ma fraction for three different j1,, values: 0.2 d" (low), 0.4 d" (normal) and 0.8 d" (high). It can be noted that for each of these cases there isa maximum anoxie sludge mass fraction above which nitr 3quation 4.37 can also be written explicitly in terms of the anoxic sludge mass fraction f= 1-(FK Y/N (+ RO /he (439) When a certain nitrification efficiency is to be maintained and therefore a maximum residual ammonium concentration is specified, there is a consequential maximum to the sludge ma fraction that can be placed in an anoxic environment. This maximum anoxic mass fraction fy, can be calculated from Eq. 4.39 by substituting N, with the specified effluent residual ammonium concentration Na fication does not oceur fy = LAF KYNyOa+ VR (4.40) The maximum anoxic sludge mass fraction not only depends on the specified residual ammonium concentration, but also on the sludge age and the Kinetic constants for nitrification. 10 1 Tb, = 004d Nag=2mai" Fe IR=10d § ]igzos mg, 2 1k, =0.5mg.i S| b,=004d" Ee 1 Bos § ' |g 2 |. osd o £ ! z “ g 6 ! 206 2 t S 03 3 3 ga i os E ' E 3? ' E02 3 1 5 G 1 = é, 1965 0.825 ° 0 01 02 030405 0607080910 ‘0 7 20 30 Anoxic sludge mass fraction (f,) Sludge age (d) Figure 4.9 Figure 4.10 Residual ammonium concentration as a Maximum anoxic sludge mass fraction fa, function of the anoxic sludge mass as a function of the sludge age for different fraction for different values of fm ‘values of fim108 The values of Ky and b, have relatively little influence on the value of f,, and when no information is available default values may adopted, such as by = 0.04-(1.03)"™ and Ky = 0.5-(1.123)™. In contrast, the influence of }1,, on the maximum anoxic sludge mass fraction is considerable. In Fig. 4.10, f,, values are shown plotted as a function of the sludge age for jt values between 0.2 and 0.8 d"! ‘The numeric value of fy, is of great practical importance: the extent of denitrification tends to increase as the anoxic sludge mass fraction is enlarged. Hence in principle, to maximise the nitrogen removal capacity of a system, the anoxic sludge mass fraction should be maximised, However, apart from the maximum set by the need for efficient nitrification, there are two other factors that may influence the value of f,. (1) the removal efficiency of organic matter and (2) the sludge settleability When the anexic sludge mass fraction is very large, there is the possibility that the metabolism of organic matter in the process becomes incomplete because the rate of ‘metabolism in an anoxic environment is lower than in an aerobic environment (refer to Section 4.3). In such a case, the organic matter may still be removed efficiently from the liquid phase, but the sludge production will increase, because part of the stored organic matter will not be metabolised but will instead be discharged as excess sludge. Furthermore in processes with a high anoxic sludge mass fraction, sludge settleability may be poor and development of filamentous or bulking sludge (refer to Chapter 6) may be frequent, possibly because of the presence of non-metabolised organic matter in the sludge Thus there is an upper limit to the anoxic sludge mass fraction, independent of the maximum value set by the requirements for efficient nitrification. Presently, there are full-scale plants with an anoxic sludge mass fraction of fifty percent that operate satisfactorily, but there is little information about the possibility to increase the anoxic siudge mass fraction beyond this point. In the Netherlands for example, the anoxic mass fraction in activated sludge system designed for nitrogen removal seldom exceeds forty percent. Based on the results of a pilot plant study (Arkley et al, 1982), the water research commission of South Africa (1984) suggests a maximum value of f,, =0.6. This value is indicated in Fig. 4.10 as well ‘There may yet be another limitation to the value of the anoxic sludge mass fraction: as fy increases, the volume of the aerobic reactors decreases and consequently the OUR increases. Hence, to maintain the floes in an aerobic environment (i.c. to prevent anoxic conditions within the sludge floc), operation at higher bulk dissolved oxygen concentration is required The higher dissolved oxygen concentration in tun leads to an increased energy requirement for aeration. Due to the increase of aeration costs, an increase of fy, may become unattractive from the point of view of economies. 4.2.3 Ni ication potential and nitrification capacity The nitrification potential is defined as the TKN concentration in the influent that can be mitrified, i.e. is available for nitrification. This concentration can be expressed as: Ny =Nu-Noe-Ni aly Where N, — =nitrification potential (mg NI") Ny, =influent TKN concentration (mg N.I') Ni =nitrogen concentration required for sludge production (mg N.I") Ne organic nitrogen in the effluent (mg N.I")Chapter 4 - Nitrogen Removal 109 (@) By = 0.3 0" () Hp = 0.6 d" 50 4 50 08 49 z os Pax 3zZ +3 z 04 S20 z z 02 49 0 RaRem 10 20 so 30 Sludge age (d) Sludge age (4) Figure 4.11 Values of N.,Npy Ni and Naas a function of the sludge age for different values of Hm The values Ruy (fy = 0) and Ryo (fm ™ fay) are indicated as wel To reduce the model complexity. the Nw. fraction is the particulate organic nitrogen. In fact, the particulate organic nitrogen fraction is part of the nitrogen present in the produced excess sludge Nj, in this case leaving with the effluent because of imperfect solid-liquid separation in the final settler, The consequences of this modelling decision are small but will be discussed nonetheless in Section 4.4.2.3. The value of N;has been determined previously with Eq, (3.61). Using this expression in Eq. (4.41) one has: issumed to contain both the soluble and Np = Nui- Noo firl(] = fw~ fi) (1 + fbi RYCIR. + faplfev Si, (4.42) The nitvifivation capacity is defined as the influent TKN concentration that is effect nitrified in the activated sludge process, Hence the nitrification capacity is the diff between the nitrification potential and the effluent ammonium concentration Ne =Np-New (4.43) Where N, = nitrification capacity (mg N.I') Using Fq, (4.37) for N, and Eq. (4.42) forN, the following equation is derived = Nu- Now [C1 = fas fap + POY RI CUR SF faplfos} Su “Kebat UR) [Hor ~ £0 ~ (a+ VRS) (44) Fig. 4.11 shows the values of Ny. N, and f,, as function of the sludge age for the following conditions. = Composition and concentration of the influent organic matter (used to calculate Nj) te =0.1 and Sy = 500 mg CODI": + Nitrification kinetic parameters: 1, K,= 1.0 mg N17; b, = 0.04 d"; Nyy: * Influent nitrogen concentration: N,, = 03d" (Fig 4.11.a) and 0.6 d"' ig. 4.11.b). mg NI! and fing Omg NI" and Nu o=N,=50 me NI!110 4.3, DENITRIFICATION ‘The necessary conditions for the denitrification process to develop in an activated sludge process can be summarised as (1) Presence of a facultative bacterial mas (2) Presence of nitrate and absence of DO in the mixed liquor (i.e. an anoxie environment); (3) Suitable environmental conditions for bacterial growth: (4) Presence of an electron donor (nitrate reductor). (1) Presence of a facultative bacterial mass Facultative bacteria are characterised by the fact that they can use both oxygen and nitrate as an oxidant for organic matter. A large fraction of the bacterial mass that develops in an activated sludge process is facultative. It has been established experimentally that activated sludge generated under aerobic conditions will use nitrate immediately when it is placed in an anoxic environment (Heidman, 1979), The rate of nitrate utilisation continues without change, as long as the anoxic condition and the availability of organic matter persist. However, anoxic oxidation of organic matter occurs at a lower rate than aerobic oxidation under otherwise comparable conditions. (2) Presence of nitrate and absence of dissolved oxygen in the mixed liquor The presence of dissolved oxygen in mixed liquor inhibits the development of denitrification Itis difficult to quantify this influence b ration gradients of dissolved oxygen will develop in the flocs so that the micro-environment in a floc may be very different from the bulk of the liquid phase (see also Fig. 4.7). In effect, efficient (though irregular) nitrate removal has been observed in aerobic activated sludge processes (Pasveer, 1965 and Maatsche, 1971), mainly those of the carrousel type. ‘This can be explained only if it is accepted that anoxi¢ micro regions are formed within the flocs, In general it hes been observed that a dissolved oxygen concentration of more than 0.2 to 0.5 mg O21" reduces the rate of denitrification significantly. In general, nitrogen in waste ‘water is present in the form of ammonium or organic nitrogen. Thus, the necessity to have nitrate present in an anoxic environment normally implies the need for nitrification as a prerequisite for denitrification. The magnitude of the nitrate concentration has little influence ‘on the denitrification rate: when the nitrate concentration is higher than 0.5 mg N.11, the denitrification rate will be independent of the nitrate concentration “ause cones (3) Suitable conditions for bacterial growth ‘Temperature and mixed liquor pH are among the most important environmental conditions for bacterial growth. The denitrification rate increases with temperature until an optimum is reached at 40°C, At higher temperatures the denitrification rate is quickly reduced, The influence of temperature on denitrification kineties is discussed in more detail in Section 4.3.2. Conceming the influence of pH, it has been established that there is a maximum denitrification rate for the pH range of 7 to 8.5, whereas for pH values smaller than 6 and larger than 8.5 there is a sharp decrease in the denitrification activity. It is very unlikely that a pH > 85 isestablished in an activated sludge process. On the other hand, a low pH valuelike pH <6 is not only inhibitory for denitrification, but also for nitrification, rendering nitrogen removal practically impossible. For municipal waste water, in order to maintain the pH in the optimal range of 7 ] reactor 's" recycle Figure 4.13 Configuration of three widely used designs for biological nitrogen removal: pre-D (a), post-D (b) and Bardenpho (c) However, the pre-D system has one important disadvantage: complete nitrate removal is not possible. A fraction of the nitrate generated in the aerobic reactor is discharged direetly from the settler without passing through an anoxic reactor. The maximum nitrate removal efficiene of the pre-D system depends on the recirculation rates from the aerobic reactor and from the settler to the anoxic reactor, However, predenitrification designs are still applied, mainly when the following conditions apply - Complete nitrogen removal is not required; = The COD concentration in the influent is insufficient to remove alll nitrate, ie. the (Ny ratio isunfavourable. If so, a pre-D system can be the optimal configuration, assuming the addition of an external carbon source is not an option; - For relatively small systems where energy consumption is not an important criterion: the recireulation flow from the acrobic reactor to the anoxic reactor ean be increased to such an extent (for instance 8 to 10 times the influent flow) so that enough nitrate is returned to the pre-D reactor for denitrification in order to meet the nitrogen discharge limits. Barnard (1973) proposed the Bardenpho system, thus combining the advantage of the post-D system (feasibility of complete denitrification) with that of the pre-D system (high-rate denitrification), Figure 4.13c shows the Bardenpho system. It is composed of four reactors, the second and the fourth being aerobic and the first and the third anoxic, Nitrification takes place in the second reactor. Inthe Bardenpho process both pre- and post denitrification are applied. In the first reactor a large part of the nitrate is removed. The remaining nitrate is reduced in the third reactor and a mixed liquor, substantially free of nitrate, passes to a (optional) fourth reactor, from where it flows to the settler. The function of the fourth reactor isto provide a short period of re-aeration (the fourth reactor is much smaller than the other ones). This ensures that the sludge does not remain excessively long in an anoxic environment: without the re-aeration reactor, the sludge would be continuously in an anoxic environment from the third reactor through the settler and back to the first reactor.14 Re-aeration also removes nitrogen bubbles formed in the post-D reactor, which might otherwise cause problems in the seitler due to aggregation to sludge floes, resulting in flotation of the sludge blanket. As an alternative for the fourth aerobic reactor, a cascade can be placed between the post-D reactor and the secondary clarifier, if the hydraulic profile permits this The feasibility to produce an effluent with a very low total dissolved nitrogen concentration has made the Bardenpho configuration a very popular design. When the single sludge system (and particularly the Bardenpho system) is compared to the three sludge system several important advantages of the former become apparent: - Inthe single sludge system there is no cost for the addition of organic material. In contrast the costs of adding organic material to the three sludge system are considerable as the following evaluation shows. For an assumed per capita contribution of nitrogen in the sewage of 10 g Nhab'.d" and an estimated requirement for sludge production of 2 ¢ N.hab".d" (ie. twenty percent of the influent TKN). the nitrification potential is 8 g N.hab".d". If the consumption of external organic material is 2.5 g CH;OH.g N™ (Christensen et al, 1977), the daily per capita methanol consumption for denitrification is 2.5-8 = 20 g. This quantity amounts to about 10 litre hab! year“! with a cost comparable to that of aeration: USS 3 to 5 per capita and per annu - In the single sludge system part of the oxygen used for nitrification can be recovered as “equivalent” oxygen for the oxidation of organic material. In Section 4.1.3, it was shown that the use of nitrate for the oxidation of organic material reduces oxygen consumption by some twenty percent. For complete denitrification, the nitrate mass to be denitrified equals 8 g N.hab".d' Knowing that | mg N is equivalent 10 2.86 mg O,, it can be calculated that denitrification reduces the oxy gen demand by 8-2.86 = 23 ¢ Orhab’'.d". If it is further assumed that the energy consumption of the aerators is 1 Wh.g'O>. the application of denitrification reduces the required power by 23 Wh.hab".d"! or 23/2: Whab"; - The reduction of | W.hab" in power consumption is very significant in economic terms, because aeration is the largest item of the operational costs for waste water treatment plants. On an annual basis the reduction of energy consumption amounts to 8.7 KW.hab", which at an assumed price of 0.10 US$. kWh" results in a cost reduction of almost US$ 1 per capita.year": - In the single sludge system, the alkalinity produced during denitrification can be used in the process. In Section 4.1.3, it was demonstrated that in the activated sludge process there is an alkalinity consumption of 7.14 ppm CaCO;.mg N" in the nitrification process and a production of 3.57 ppm CaCQ.mg N” during the denitrification process. Hence in single sludge systems half of the alkalinity consumed during nitrification can be recovered when denitrification is complete. In the three-sludge system, nitrification and denitrification develop sequentially in the second and the third part of the system respectively. Thus, the recovery of alkalinity by denitrification in the last part of the system cannot be used to balance the consumption of alkalinity due to nitrification in the second part. For this reason, in the three-sludge system there is usually a need for alkalinity addition (e.g. lime), whereas the alkalinity of most municipal waste waters ishigh enough to operate a nitrogen removing single sludge system without alkalinity addition, - In the three-sludge process, it is very difficult to exactly match the dosage of organic material with the nitrate concentration so that neither organic material nor nitrate are present in the final effluent. In practice it will be required that a small aerobic reactor is added afier the third reactor, where excess organic material is removed biologically, thereby complicating even more the already complex configuration of the three sludge system,Chapter 4 - Nitrogen Removal 15 + Forthe biological excess removal of phosphorus it is necessary to create a truly anaerobic zone, characterised by the absence of both dissolved oxygen and nitrate. Such an angerobie reactor is only feasible in a single sludge system with a pre-D reactor. Thus in the three sludge system biological phosphorus removal is not possible. which reduces its applicability in practice. There is one advantage that the three-sludge process may have compered with the single sludge system: in a single sludge system nitrification oecurs in the aerobic part of the system In asystem with a large anoxic sludge fraction (which in practice will usually be required), the sludge age needs to be relatively high and hence a large treatment system is required. Thus itis possible that the reactor volume of the single sludge process is larger than the volume of the three reactors of the three sludge system together. However. this possible advantage will certainly not compensate for the very serious disadvantages inherent to the three sludge system as discussed above. For that reason, only the single sludge system will be considered further. 4.3.2 Denitrification kinetics Marais and his group of research workers developed an empirical model for the kinetics of denitrification. This model is an extension to the model for the removal of organic material presented in the previous chapter. 4.3.2.1 Sludge production in anoxic/aerobic systems When the data published on sludge production in aerobic/anoxic systems is compared, it can be concluded that sludge production is not affected by the presence of anoxic zones and is equivalent to that of a purely aerobic system. In this context the data collected by Sutton et al (1979) during their experiments are possibly the most illustrative. In Fig 4.14 these data are presented. The organic sludge mass per unit mass of daily applied COD (mX,) is plotted as function of the sludge age for different anoxic sludge mass fractions, temperatures and sludge ages (both pre-D and post-D systems), 19 10 10 ret) (4.46) tp,/top = denitrification rate associated with the utilisation of easily/slowly biodegradable material K,/K2 =denitrification constant for easily/slowly biodegradable organic material (mg Nmg"X,.d") t, _ =duration of the primary phase (d)118 In the post-D reactor, denitrification is only associated with the utilisation of slowly biodegradable material. Since this concentration in a post-D reactor will be smaller than in a corresponding pre-D reactor, the denitrification rate will also be smaller. The rate of nitrate removal in a post-D reactor can be expressed as ty =dN/dt= KyX,, @ar The kinetic expressions for denitrification in Eqs. (4.46 and 4.47) are all zero order equations: in a steady state situation, the denitrification rate does not change with time as the active biomass can be considered to be constant. Therefore the above expressions can be used to calculate nitrate removal in anoxic reactors, independent of its hydraulic regime, and may also be applied to completely and partially mixed reactors. Van Haandel et al (1981) calculated the values of the denitvification rate constants Ky, K and K; from the experimental results obtained by several authors, all using municipal waste water as influent, From the data obtained by Stern et al (1974), Wilson et al (1976), Marsden et al (1974), Van Haandel et al (1981), Nichols (1981) in South Africa; Sutton et al (1969) in Canada, Heide (1975) in the Netherlands and Heidman (1979) in the United States, the following average values were calculated for the range of temperatures from 12 to 26°C: K, =0721.27 Kz =0.10-1.08"?) K3 =0.08-1.037" (448) Research that has not yet been published, using municipal waste water from Campina Grande (Brazil), shows that the formulas in Eq. 4.48 remain valid for temperatures up to 28°C. In all cases the data were obtained with waste waters containing only minor industrial contributions ILis possible that in waste waters with a significant or predominant industrial contribution the constants have different values due to the presence of toxie materials. 43.2.3. Minimum anoxic mass fraction in the pre-D reactor In the previous section it was shown that the denitrification rate in the pre-D reactor is high, as long as easily biodegradable organic material is present. As the objective of the anoxic reactors is to remove nitrate, it is important that the denitrification rate is kept as high as possible. Therefore it is necessary that the retention time in the pre-D reactor is sufficiently long to guarantee complete utilisation of the easily biodegradable material, To determine the minimum retention time, the removal rate of the easily biodegradable material is compared with the feeding rate to the pre-D reactor. The feeding rate of easily biodegradable material can be expressed as: Taos = Sto/Ri = Gp'Se/Ri (4.50) Where: Tim = feeding rate of easily biodegradable material to the pre-D reactor R, =hydraulic retention time in the pre-D reactor = V\/Q, (a) V;_ = volume of the pre-D reactorChapter 4 - Nitrogen Removal 19 The utilisation rate of easily biodegradable material is proportional to the associated denitrification rate rp, = Ky, (Eq. 4.46). In the process of utilisation, a fraction of (1- f.y"Y) is oxidised. As stoichiometrically | mg NO,-N equals 2.86 mg 03, the utilisation rate of easily biodegradable material can be expressed as. tps = (1 ~ fey YV2.86% 5 = Lantus (451) Where fa, = (1 -f..°¥/2.86 = denitrification constant 1p, = denitrification rate due to the utilisation of easily biodegradable material yy = utilisation rate of easily biodegradable material Now the minimum required retention time in the pre-D reactor to remove the easily biodegradable organic material can be calculated by the condition fis = (452) Using Eqs. (4.50 and 4.51) in Eq, (4.52) one has: fay Sw/Ruau= toefan KrXefis (453) Rein = minimum retention time required for complete utilisation of the material in the pre-D reactor (d) sily biodegradable Substituting for X, from Eq, (3.29) and rearranging RauilRy =farki/(R-C) (54) The minimum retention time RysniS ass reactor. Since Rnin= VisiQ; one has sociated to a minimum sludge mass fraction in the pre-D RaniR = (Vini¢ W/V Qd) = (Vani! V> Where: ‘min OF Fonin = Lenfy/(KiCx) (4.55) Vain = minimum pre-D reactor volume required for complete utili biodegradable material fig = minimum sludge mass fraction in the pre-D reactor For “normal” values of f, K, and C,, the minimum frection fig, is always very small and the anoxic sludge mass fraction in a full-scale activated sludge process will be invariably much larger than fyi. It can be concluded that the utilisation of easily biodegradable material ean be considered complete in the pre-D reactor, provided that sufficient nitrate is available.120 4.3.3 Denitri ation capacity In practice, the most important parameter in a nitrogen removing activated sludge process is the amount of nitrate that can be removed per litre of influent, This parameter is called the denittification capacity and is determined from Eqs. (4.46, 4.47 and 4.55) as shown below. 43.3.1 Denitrification capacity in a pre-D reactor If the volume of a pre-D reactor is insufficient for complete removal of the easily biodegradable material, the removed nitrate mass can be expressed as: MNe = pV) = (KitK)XyVi (Vi fev ¥/2.86MSi5i = futrfab’StQ or Np; = MNps/Qi = far'fi Sts (4.59) Where: MNp, = removed nitrate mass per time unit, associated to the utilisation of easily biodegradable material (MSya) No, =nitrate removal in mg N per litre of influent through utilisation S,, ‘The mass of removed nitrate per time unit due to the utilisation of slowly biodegradable material in a pre-D reactor can be calculated as: MNp, ~K2Xa'Vi (4.60) Now, using the same procedure as above, the removed nitrate concentration (expressed as mg N11" influent) due to the utilisation of slowly biodegradable material, Nop is: ‘Cefn Sii ol)Chapter 4 - Nitrogen Removal 21 ‘The denitrification capacity of the pre-D reactor is the sum of the values of Np, and Np». From. Egs.(4.59 and 4.61) one has, Da sha )'Spi for fa > fain (4.62) 4.3.3.2 Denitrification capacity in a post-D reactor Cefs Si (463) D.3 = denitrification capacity of a post-D reactor (mg N.1" influent) fs = sludge mass fraction in the post-D reactor In Fig. 4.17 the denitrification capacities of a pre-D and a post-D reactor are plotted as a function of the anoxic sludge mass fraction. The values of Do) and Do; have been calculated for a sludge age of 10 days and under the following conditions: Sy = 400 mg COD.1"; 20°C; £3, = 0.24. The ratio Do/Syis also indicated (on the right hand scale). It can be observed that the denitrification capacity depends on the following factors: = Concentration and composition of the influent organic material, i. €. the values of the fractions fy, fp and fp. ~ Sludge age: the value of C, = Y'R,/(I+by-R,) increases at higher sludge age and thus the value of D, will be higher as well; ~ Temperature: the values of the denitrification rate constants K» and K, inerease at higher temperatures, resulting in an increase of D.. On the other hand, the value of the decay constant b, will be higher as well, which reduces the overall temperature effect; = Size of the anoxic sludge mass fractions: when f, and f increase in size, so do the denitrification capacities Dy, and Ds In practice, the values of fx and fs are limited by the requirement to maintain efficient nitrification and good sludge settlesbility Pre-D configuration Post-D contliguration » os ao oors = «= 2 L 0.050 3 20 0050 § 20 "2 Ez Q 2 SP 2 : Es Ez = Zo b ‘0 0025 & 10 0025 2 2 3 ° ° ° ° oo of 02 03 04 oo of 02 03 04 Anoxic mass fraction Anoxic mass fraction Figure 4.17 Denitrification capacity as a function of the anoxic sludge mass fraction for a sludge age of 10 days in a pre-D and a post-D anoxic reactor122 Determine the denitrification capacity of the activated sludge process of Example 4.1, assuming fa» = 0.20. Solution ‘The composition of the organic material can be calculated from the influent and effluent COD and the concentration of volatile sludge X, in Table 4.1. With Sy = 477 mg COD1I”, Se = 18 mg COD.I" and X, = 2,469 mg VSS.1", the following values are calculated: i = Se/Sp= 18/477 = 0.04 MX = MX/MSu= VeX(QrSt) For the applied sludge age R. = 18 days anda temperature of 21.6°C the values of C, and b, are calculated as’ Cc =Y-Rel(1 + by Rs) = 0.45-18/(1 + 0.24-(1.04)'*18) = 1.45 mg VSS.mg" COD.d™ by = 0.24 (1.04)'°= 026d" Using the data from Table 4.1 and Eq, (3.69): mX, = 25-2469/(40:477) = 3.23 mg VSS.mg" COD.d" = (1+ fro fro): (1 + f Bir Re)Cr+ foo Rees = (1-004 -fop) (1 + 0.2:0.26-18)-1.45 + frp'18/1.5) = (0,96 - fag):2.78 + fop'0.12, OF frp = 0.06 Now, knowing the total non-biodegradable COD fraction the biodegradable COD concentration is calculated as; Sou = (1- fhe fro) Sy = (1 - 0.04 - 0.06) 477 = 430 mg COD." Sis =fe'Sy = 86mg COD.I" Using Eq, (4.48), the values of the denitrification rate constants are calculated as: Ky =0.1.(1.08)" =0.1(1.08)"© =0.113mgN.mg" Xd? Ke = 0.08 (1.03)' = 0.08.(1.03)'° = 0.084 mgNmg" X..d" The anoxic sludge mass fraction fx: = Vi/V; = 5/25 =0.2. This is much larger than the minimum anoxic sludge mass fraction required for the removal of easily biodegradable organic material fin = 0.114 fas/lKr-C,) = 0.12: 0.25((0.72 (1.2)"° 1.45) = 0.02 Hence, Eq.(4,62) can be applied: Da = (0.114 faxtK2Crfia) So: (0.114 0.20 + 0.113 1.45 0.2)-(1 - 0.04 - 0.06) 477 = 24.3 mg N.I*Chapter 4 - Nitrogen Removal 123 Example 4.6 Verity if the experimentally observed nitrate removal in the previous example corresponds to the calculated denitrification capacity. Estimate the denitrification capacity, assuming (a) the lasttwo reactors are anoxic (post-D configuration) and (b) the first end fourth reactor are anoxic (Bardenpho configuraton), Solution In Example 4.1 the daily removed nitrate mass was calculated as 864 g N.d". As the influent flow is 40 m°.d", the experimentally observed ritrate removal is 864/40 = 21.6 mg NI". This value corresponds to 88 percent of the calculated value of the denitification capacity of 24.6 mg N.I". The denitrification for the post-D configuration is equal to Des = ke Crh Sei = 08-(1.03)'°-1.45-0.4-430 = 21 mg NI" For the Bardenpho configuration: Do = Der + Des = 24.5 + 21/2 = 35 mg NI* When comparing the denitrification capacity in the post-D reactor (21/2 = 10.5 mg N.I" per reactor) with the value calculated in the pre-D reactor in example 4.5 (24.3 ma NI"), itis concluded that under the specified conditions, the pre-D reactor removes more than twice the amount of nitrate of the post-D reactor. Itis interesting to compare the denitrification capacity of the Bardenpho configuration with the nitrfication capacity. N- can be calculated with the aid of Eq, (4.43) Ne = Ny Noi Ni- Nw = 45.1 + 0.3 -349/40-1.9 = 34.9mgN.I° The denitrification capacity (35 mg NI”) is marginally larger than the nitrification capacity in the Bardenpho configuration (34.9 mg N1*). Thus in principle it is possible to produce an effluent with a very low nitrate concentration. However, to do so, it will be required to introduce a recirculation flow from the aerobic- to the pre-D reactor. The model presented above for the calculation of the extent of denitrification in anoxic reactors is based on experimental observations of denitrification in single sludge activated sludge processes, The only way to verify the validity of the model is to compare experimental values with the theoretical model values, Unfortunately, the majority of the data presented in literature cannot be used to check the validity of the model, because some elements to determine the denitrification capacity are missing, like the sludge age, the temperature, or the anoxic sludge mass fraction. Furthermore, the anoxic reactors often did not contain nitrate in their effluent, ic. the availability of nitrate was restricted and more nitrate could have been removed in the anoxic reactor. In such a reactor, the removed nitrate mass will always be inferior to the denitrification capacity. However, in all cases where it was possible to verify the validity of the model, aclose correlation was found between the predicted theoretical nitrate removal values and the observed experimental values, This was demonstrated in pre-D and post-D reactors on sy stems operating under the most diverse conditions124 - Size of the process: up to 60,000 m*: - Applied sludge age from 3 t0 35 days: - Temperature from & to 28°C; - Anoxic sludge mass fraction from 10 to 82 percent; - Pre-D, post-D and Bardenpho configurations, = Influent COD values between 220 and 850 mg COD.1" = Recirculation factors of nitrified mixed liquor “2” and retum sludge “s” of 0.2 to 6 times the size of the influent flow; - Origin of the waste water: municipal sewage {rom South Africa, United § the Netherlands and Brazil fates, Canada, ‘The data show that the model adequately describes nitrogen removal in single sludge activated sludge process. On the other hand, analy'sis of the data also shows the limits of model validity. ‘The denitrification capacity depends on factors that vary from one waste water to another: - The concentration and composition of the influent organic material; - The denitrification rate constant K> (in the case of a large proportion of industrial waste water being present in the influent) Because of the variability of several factors determining the denitrification model, ideally the values of the model parameters should be determined experimentally for each waste water Hence it is important to have a simple and reliable calibration method to determine these factors. In Chapter 3 a procedure was described for the determination of the parameters defining the composition of the organic material. In Appendix 4 a simple experimental method will be presented to determine the value of the denitrification constants K» and Ks 4.4 DESIGNING AND OPTIMISING NITROGEN REMOVAL. 4.4.1 Calculation of nitrogen removal capacity The concepts of nitrification and denitrification ‘ities are very convenient to describe nitrogen removel in the activated sludge process, as demonstrated in the following example Consider the nitrogen removal in an activated sludge process characterised by the following parameters: pa - - fy =0.10 ~ by =004d7 . fy = 0.06 z mg N.I" * - f,=0.25 . O.1 mg Nmg'X,d" - a 1 ‘The nitrification and denitrification capacities can be calculated as functions of the sludge age using Eq (4.44) forN. and Eqs. (457 and 4.58) for Da, for two valves of the maximum growth rate jl, (0.3 and 0.6 d"). To calculate D,;, it is necessary to first determine the maximum allowable anoxic sludge mass fraction f,, as a function of sludge age. using Eq. (4.40). In Fig 4.18 the resulting curves of N., D,: and f,, are shown It is convenient to introduce yet another parameter related to the nitrification capacity the available nitrate in the anoxic reactor N,,. In a pre-D system, the available nitrate is the fraction of the nitrification capacity that is recirculated to the anoxic reactor, Having recirculation factors “a” from the aerobic reactors and “s” from the settler, the flow entering into the aerobic reactor is (a+s+1)-Q; of which a fraction (a+s\(a+s+) is recirculated to the anoxic reactor. The remaining fraction 1/(a+stl) is discharged from the system.Chapter 4 - Nitrogen Removal 125 Hence for a nitrate production of N. in the aerobic reactor, a nitrate concentration of (at3)(ats4l)-N¢ is effectively available for denitrification in the pre-D anoxic reactor Nw = (ats(ats+l)Ne (4.64) Where Nyy ~ available nitrate in the pre-D reactor (mg N.1" influent) The value of Nyy is indicated in Fig. 4.18 as a function of the sludge age for recirculation factors a=4 and $=, 1. fOr Nay = (441)/(44/H)N.= 0.833Ne, The value of Nyy represents the maximum nitrogen concentration that can be removed in a pre-D activated sludge process. Fig. 4.18 is a useful illustration that demonstrates the utility of the concept of nitrification and denitrification capacity. With increasing sludge age the following situations can be observed 3d"): I, =0.3d" 12 60 0.12 01080 0.10 0.089 240 0.08 Be 8 of 8 006 FZ 20 0.06 2 22 2 Fe z 004 S20 oo z 002 * 10 0.02 0 ° © O48 12 16 20 28 Rnd 8 1218 2024 28 Sludge age (d) Sludge age (d) Figure 4.18 Nitrification- and denitrification capacity in a pre-D configuration as a function of the sludge age for the maximum anoxic mass fraction (1) When R, Ry, nitrification is possible. However, it is not yet possible to comply with the condition that N,,< Nya. the specified effluent ammonium concentration. The reduction of Nyc 0 a Value < Nyg is possible when the applied sludge age is higher than Ryq, This sludge age Ryn can be calculated from the condition that fy = 0 i. fn = 1-1 +KyNeg(/Rant ba)/tn OF Ram = Vlpto/{l + Ka/Nadd ~ bal (4.65) For the data in this example Rox = 1/(0.3/(1 + 1/2) - 0.04] = 6.25 days126 (3) For sludge ages longer than R,q, it is possible to attain the specified residual ammonium concentration N,q and to include an anoxic zone in the system as well, Using Eq, (4.40) to determine the anoxic sludge mass fraction, the denitrification capacity can be calculated from Eqs. (4.58 or 4.62). The nitrification capacity is calculated with the aid of Eq.(4-44) fa (AK YN) OA UR 6”) Da =(Ki+K9)CSiVilVe= Ki KIC faSsi (Lar find (462) (4) For some particular sludge age R,o, the feasible anoxic sludge mass fraction f, will be equal to the maximum value allowed fngs. For the given operating conditions and for frase = 066 the value of R,, is calculated as fn = +K G/N): Rao Du), OF Mfpn*( = faaMC1 + KfNoa) = bal (4.66) For the example considered: Ry = 1/[0.3-C1 - 0.6)1 + 1/2) - 0.04] = 25 days. (5) When the sludge age R, > Ro, then both nitrification capacity and the denitrification capacity increase marginally with sludge age. D,, will increase slightly more than N, Using the values of N, and Da, the effluent nitrogen (ammonium and nitrate) concentration can be calculated as a function of the sludge age @R Row Now it is possible to include an anoxic reactor. At increasing sludge age, the maximum allowed anoxic sludge mass fraction increases as well and so does the denitrification capacity For a particular sludge age R, = Ry, the value of Dg; will be equal to N,, so that: Da = Now OF (fafa + KaCy-fa) Sti = Nota + sa +s + 1) (4.67) ‘The value of R,;can be graphically determined in Fig. 4.18 and is equal to 11 days (this value can also be caleulated by trial and error using Eq. 4.67). In the range Ruy Nay and the anoxic reactor is underloaded, even when the maximum recirculation factors (a+s)= 5 are applied. All nitrate recirculated to the anoxic reactor will be removed, The effluent nitrate conceatration will be equal to the fraction of the nitrification capacity that is discharged directly from the aerobic reactor to the effluent, without passing through the anoxic reactor: N,. = N/(a+s+l), It is assumed here that no denitrification will take place in the settler. The ammonium concentration will be constant at Nyy=_2 mg NI" (as fq will still be equal to fig). In this range of sludge ages, the nitrogen removal efficiency could be increased by taking part of the pre-D reactor and using it to create a post-D reactor R>R, Now the anoxic sludge mass fraction is limited by the condition that it may not exceed a maximum value: f, No. the effluent nitrate concentration will still be given as Noe = NU(atsH) In Fig. 4.19 the concentration of the different nitrogen fractions Ny, Nye, Nand Nj is shown as function of the sludge age for the conditions specified in this example and in Fig. 4.18. For this, particular example, it can be observed in Fig. 4.19 that for a sludge age of |] days almostall of the influent nitrogen concentration (N,) of 50 mg NI” 1s removed. ‘The effluent nitrogen concentration (Nj) is equal to Nt Noe: The value of Nye= Ny(atts+1) is 38,0/6 = 63 mg NI" 90 Ne=2.0+6.3=83 mgNI" Atthe selected sludge age of 11 days the nitrogen concentration that is removed with the excess sludge N; equals 100meN 1". Hence, the denitrified nitrogen concentration Naisequal 10 Nui Naj~ Nye = 50 = 2-63 - 10,0 = 31.7 mg N.I", which is equal to the denitrification capacity for R,= 11 dan If it is desired to reduce the effluent nitrogen concentration below the minimum value that can be obtained in the current pre-D configuration (in the example N,, = 8.3 mg NI" for R, = Ra = 11 days), it will be necessary to increase the sludge age and modify the reactor configuration of the proc \sorming it from a pre-D system to a Bardenpho system. The optimisation of the Bardenpho system will be discussed in the next section128 50 0.10 Nitrogen = in excess sludge Ny 240 0.08 > & € 2 230 006 § E g'{ | Removed nitrogen Ny + 5 3 (by denitrification) 2 § 20) 2 0.04 2 B 8 5 = £ £ & E > |2 £10 0.02 z tRy Nitrate Noe 0 0 0 10 20 30 Sludge age Figure 4.19 jon of the nitrogen in the influent over the different nitrogen fractions Calculated as a function of the sludge age (for p1,,= 0.3 d"!) 4.4.2 Optimised de: in of nitrogen removal The objectives of design optimisation for nitrogen removal activated sludge processes are: - To produce an effluent with minimum total nitrogen concentration; - To carry out this nitrogen removal at minimum construction and operational costs Before starting with the optimisation procedure, it is necessary to remember that there are several consiraints to single sludge activated sludge processes (1) The anoxic sludge mass fraction is limited by two independent criteria: (a) the nitrification efficiency must be high, which implies a certain minimum for the sludge mass fraction in the aerobic zones and a comesponding maximum for the anoxic sludge fraction and (b) the mechanic properties and the kinetic constants can be adversely effected by an excessive anoxic sludge mass fraction. With regard to the second imposition, it is suggested that the anoxic sludge mass fraction should not be larger than sixty percent. (2) The value of the recirculation factor “a” for the mixed liquor from the aerobic reactor to the pre-D reactor and the value of the return sludge factor “s” have upper constraints from 4 viewpoint of energy efficiency, As the headloss in the “a” recirculation is always smaller and since the nitrate concentration in the aerobic reactor at least as high as in the settler, the “a”-recyele is always more cost-efficient than the “s” recycle to introduce nitrate into the pre-D reactor. For that reason, the value of the sludge recycle factor “ will be dictated by the requirements for efficient settling only. In practice this often means an Je with a value of 0.5 < a Foo 2 8 E : z # 3 007s], |2 : £ £ 31 compete nitogen | 2 3 a ‘enol Foo | € 1; Zz : 2 j toe |[# : geo i = i : EY re i oo Li s 1 2 2 Hw 0 8S 6 Ow ‘Sludge age (days) Sludge age (days) Figure 4.20 Figure 4.21 Maximum TKN/COD ratio in the Values of the ratios (NUS»), and (N/S»)1 asa function of the sludge age (for a=4dands=1) influent allowing complete nitrate removal as a function of R, for different values of “a” Figure 4,20 shows as an example of the (N,/S,)oratio as a function of the sludge age for values of the “a” recirculation factor ranging from | to 10. The value of fy, is indicated as well. The following parameter values were used to construct Fig. 4.20 -T 20° It can be observed in Fig. 4.20 that the required sludge age for the complete removal of nitrogen tends to increase when the recirculation factor “a” decreases. The choice of the optimal recirculation factor then becomes a question of economics, where the costs for pumping of nitrified mixed liquor to the pre-D reactor (energy and pumps) are compared to the construction and operational costs asa function of the sludge age. In practice the optimal value will almost invariably be less than a = 4, There are two exceptions (1) Recirculation systems such as carousels do not have dedicated nitrification and denitrification reactors, but consist of one or several large “racetrack” loops. They are often designed with surface aerators, double acting as propulsors, allowing higher recirculation factors of up to 10 - 20 to be applied without much additional cost: Industrial systems treating waste water with a high N\/S, ratio and a high Ny concentration, requiring the addition of an external carbon source. Often it is preferred to add this external carbon source to the pre-D zone, to prevent accidental overdosing in the post-D zone with the consequential risk of exceeding the effluent COD limit (2)132 44.2.2 Incomplete nitrogen removal For high N,/Sq ratios or other unfavourable conditions (low temperature, toxic compounds, ow easily biodegradable COD concentration), complete denitrification might not be possible, unless the applied sludge age is so high that the system will be unacceptably large or alternatively, the addition of an external carbon source is required, Should the addition of an external carbon source not be considered as a viable option due to prohibitive costs, the maximum nitrogen removal efficiency may be determined as follows: in the pre-D reactor of the Bardenpho system, nitrate is being introduced with the “a” reeyele from the nitrification zone and with the “s” recycle from the settler so that Day = (luis + Ke Cefa)Sy= (a +5 + NG + SNe (473) Where N,. =effluent nitrate concentration (equal to the concentration in the return sludge) The effluent nitrate concentration is given by the difference between the nitrate concentration in the aerobic reactor and the nitrate removal in the post-D reactor. If Eq(4.73) is valid, the nitrate concentration in the pre-D reactor will be zero and the concentration in the aerobic reactor will be a factor /(ats+) of N, (the influent is diluted (1+a+s) times before it reaches the aerobic reactor). The decrease of nitrate concentration in the post-D reactor is Dey/(s+l), so that: Ne =Nd(at+st 1)-Desl(s +1) (4.74) Now, by substituting for Dg from Eq,(4.63) and knowing that fs = fy, - fq: one has Nee =NU(ats+ 1) -kyCe(fn= fa Swl(s + 1) (475) By using Eq (4.75) in Eq.(4.73) one calculates that (NJSy)-@tsMatst 1) ~ farts- KC, FQ = $$$ (4.76) In Eq (4.76) the calculated value of fy is the one that leads to the highest nitrate removal in the Bardenpho system, if complete denitrification is not possible, The residual nitrate concentration is given by Eq. (4.75). In Bq, (4.76). as the NJS1. ratio increases. so does the value of fy; (While at the same time the value for f,3 decreases), until f= fy (and f= 0), For this limiting value of N/Sy. the entire anoxic sludge mass fraction is placed in the pre-D reactor. In other words, the Bardenpho configuration ceases to be advantageous and it is therefore changed into a pre-D system. The value of the ratio NUS, for which this occurs is obtained by substituting f, = fin Eq. 4.76). (NISwi = Carfa +KeCrtgat s+ Dats) (477) Where (NJS.):= limiting ratio for the applicability of the Bardenpho process The (Ny/S_); ratios corresponding to the (NYS), values can be calculated using a similar expression to the one in Fq (472). In Fig. 4.21 the (NUS); curve is plotted as a function of sludge age for the same conditions specified in Fig 4.20, but with the “a” reeycle fixed at a value of four. Now, basically three different situations can be distinguished:Chapter 4 - Nitrogen Removal 133 (1) Zone A: low Ny/Seratio ie, Ny/St <(Ny/Su)o. In this particular case the proportion between nitrogenous material and biodegradable organic material is favourable and complete nitrate removal is possible; (2) Zone B: average Ny/Sj ratio, ie., (Ny/So < (Nw/Si) < (Ne/S,)i. In this case, complete denitrification 1s not possible, but the minimum effluent nitrogen concentration is obtained in a Bardenpho configuration; (3) Zone C: high Ny/Sy ratio, ie. (NYS,) > (Ne/S,); In this case, the proportion between nitrogenous and biodegradable organic material is unfavourable for nitrate removal. The lowest total nitrogen concentration in the effluent is obtained in a pre-D system. In general, for regions with a wamn climate the ratio between the TKN and the COD concentration in raw sewage is such that complete nitrogen removal is feasible, even at relatively short sludge ages (5 to 10 days). A large industrial waste water fraction may lead to a low pi, value and the consequential need to inerease the sludge age On the other hand some industrial waste waters (especially those of vegetable origin) have a low N,/S, ratio so that complete N removal is relatively easy. Primary and/or anaerobic pretreatment of the raw sewage tends to have a negative effect on the nitrogen removal process. as in such systems more organic than nitrogen material is removed. The Ny/Sy ratio i resulting in a longer sludge age or incomplete nitrogen removal. An analysis of the factors that affect the required sludge age for complete nitrogen removal reveals that the maximum specific growth rate of nitrifiers 4, is the most important one. As it is known that this value tends to vary considerable depending on the origin of the waste water, it is important to determine its value experimentally whenever possible. In Appendix 4 the method used to determine this parameter is explained 4.4.2.3. Model enhancements The most accurate model may not always the best model, as it should be fit for purpose. Although added complexity may make a model much more true to reality, it will also be increasingly cumbersome to use, Calibration is a well known problem, as is interpretation of the results, while it is also difficult to avoid errors as the number of equations grows. When the simplified activated sludge model was developed, il was with the specific objective of being used as a tool for the design and optimisation of activated sludge systems, not as a scientific tool, Therefore some assumptions and simplifications have been implemented to facilitate the required calculations at the expense of a small decrease in accuracy. Notwithstanding, the model is complex as it stands. Two model simplifications regarding the removal of nitrogen are the following: = Nis modelled as a single fraction containing both soluble and particulate nitrogen. This overestimates the amount of nitrogen discharged with the excess sludge: - Nitvifiers are not included as a separate sludge fraction, which is perfectly justified for municipal waste water with a relatively high TKN/COD ratio, However, this is not the case for some industrial waste waters and for the new nitrogen removal configurations that will be discussed in the subsequent sections The effect of these simplifications on the model results is discussed in Appendix AS, where the model enhancements that can be applied to enhance the simplified model (if required) will be discussed,134 4.5 NEW DEVELOPMENTS IN NITROGEN REMOVAL SYSTEMS. ‘As can be observed from Fig, 4.11, there is a significant nitrogen demand for sludge production. In the case of sewage treatment, the influent nitrogen concentration required for sludge production is in the order of 2 to 3% of the influent COD concentration. ‘Thus, for an N,/Syratio of 0.1 mg N.ng" COD the nitrogen demand for sludge production amounts to 20 to 30% of the influent nitrogen concentration, If the sludge age of the activated sludge process is not very long, the excess sludge must be stabilised by anaerobic or aerobic digestion, before it can be dewatered (Chapter 8). During the digestion process, the nitrogen associated with the mineralised sludge is released to the water phase as ammonium. After phase separation, the liquid phase of the digester effluent is returned to the activated sludge process, thus increasing the nitrogen load that has to be removed. As during sludge stabilisation typically some 30% of the volatile suspended solids are digested, this represents 0.3-(20 to 30%) = 6 to 9% of the influent nitrogen concentration. When a primary settler is present this percentage will be even higher. Using Eq. 4.78, the amount of nitrogen released in the digester can be calculated: MN =fir(Rap'fay + Rav (-fye))"MEy (478) Where Ry conversion fraction of primary sludge and the active secondary excess sludge Ry, = conversion fraction of the inactive secondary sludge The nitrogen concentration in the liquid phase of the digester is high, typically in the order of 500 - 1500 mg NI", while the reject water temperature is relatively high at 25 - 35°C. Due to the ammonification of organic nitrogen to ammonium in the digester, the amount of generated alkalinity is approximately equivalent to the molar ammonium concentration. Due to the high nitrogen concentration it can be advantageous to treat the reject water in a separate side stream process, especially if the main activated sludge process is heavily loaded or receives waste water with @ high N,/S, ratio, In the Netherlands, the foundation of applied water research (STOWA) funded several studics during the period 1994-1998 with the objective to develop and test new biological and/or physical-chemical methods for the treatment of sludge digestion reject water Since then, several promising processes have been developed further and some have been implemented at full-scale. Three of those processes, Sharon, Babe and Anammox will be discussed in the following sections. The Sharon and Babe configurations basically use the same principles as traditional nitrogen removal systems: ammoniumis oxidised by autotrophie organisms and the oxidised form of nitrogen is reduced to molecular nitrogen by heterotrophic bacteria in an anoxic environment. However, the sludge age and temperature are controlled to prevent the second step in the nitrification process, i.e. conversion of nitrite to nitrate. This reduces oxygen consumption and simultancously less organic material is required for denitrification In the Anammox configuration anew biomass type is used consisting of bacteria that use nitrite nitrogen to oxidise ammonium. In this process both forms of nitrogen are converted into molecular nitrogen. Thus the Anammox system has the significant advantage that there is no need for organic material to reduce the oxidised nitrogen. The Anammox concept has a very large potential for application to post-treatment of effluents from anaerobic pre-treatment systems (which will be discussed in Chapter 9). Anaerobic-aerobie treatment is. very advantageous in regions with a warm climate, but in the process the Ny/S, ratio increases to the point where conventional denitrification will become impossible.Chapter 4 - Nitrogen Removal 135 In principle, part or all of the anaerobic effluent could be subjected to oxidation of the ammonia to nitrite, Subsequently the nitrite-rich stream is contacted with the remainder of the anaerobic effluent Thus nitrogen removal without the need for organic material may be effected, However, some form of post treatment will be necessary to meet the effluent limits 4.5.1 Nitrogen removal over nitrite - Sharon As can be observed from Fig. 4.3, nitrogen removal in the activated sludge process is possible by (1) ammonium oxidation to nitrate and subsequent denitrification of the nitrate to molecular nitrogen or (2) ammonium oxidation to nitrite and denitrification to Ny, The nitrification reactions can be written as: NH, +1402 > NOs +H:0 +2 H” (ammonium oxidation) (4.79) Nor+% b NO; (nitrite oxidation) (4.79b) For the denitrification process, assuming methanol is used (one of the cheaper commercial carbon sources available), the following (catabolic) reaction equations can be written NOs+%CH,OH+H" > %N)+%COz+1%4H,0 (4.80a) NOs +%/6CH;OH+H" > %4No+°/¢CO:+2%/6H2O0 (4.80b) From these equations it can be concluded that nitrogen removal over nittite is attractive, Not only because il requires less oxygen (oxidation to nitrite requires only */;" of the demand for oxidation to nitrate), but also because denitrification of nitrite requires only */;"" of the organic material required for nitrate removal. However, these reduction percentages are based on the catabolic reactions only, while the anabolic (cell mass formation and maintenance) reactions are ignored: in practice the reduction in organic material consumption will be less. Furthermore, at temperatures less than about 20°C, the oxidation of nitrite normally proceeds at ahigher rate than the oxidation of ammonium, so that in activated sludge systems operated in countries with a cold or moderate climate it is impossible to limit the ammonium oxidation process to nitrite generation only. On the other hand, at higher temperatures the rate of nitrite oxidation tends to be less than the rate of ammonium oxidation, especially at increased pH and under those circumstances it may be feasible to end the nitrification process at the nitrite step (i. nitritation) and not allow the generation of nitrate. As an example in Fig. 4.22, typical p,, values for both ammonium- and nitrite oxidisers are plotted as a function of the temperature. The corresponding minimum (aerobic) sludge age can be calculated from Eq, (4.33) as Rey = 1/(jln -ba) and is also shown in Fig. 4.22. In the temperature range of practical interest to most activated sludge processes (10 = 20°C), the growth rate of the ammonium oxidisers is either lower than, or practically equal to, the growth rate of the nitrite oxidisers. However, above 20°C the difference in maximum. growth rate and hence in required minimum aerobic sludge age becomes more significant as the temperature increases. As anacrobic sludge digesters normally operate in the range of 25 10 35°C, the sludge age can indeed be used as a control parameter to induce nitrogen removal to nitrite. This is exactly the philosophy behind the Sharon process, which is an acronym for “Single reactor for High activity Ammonium Removal Over Nitrite”. The applied aerobic sludge age depends on the operational temperature as shown in Fig. 4.22136 Ammonium oxidizers oxidizers: Nitrite oxidizers Maximum growth rate (4) imum aerobic sludge age (d) 0 10 20 30 0 10 20 30 Temperature (*C) Temperature (°C) Figure 4.22 Typical profiles of maximum growth rate and minimum requii sludge age for ammonium. and nitrite oxidisers as a function of the temperature: derived from Veldhuizen et al, 1997 and Jetten et al, 2000 Atsuch a short sludge age and considering the relatively small size of the reject water flow, the reactor can be operated as a chemostat” ie. a reactor without final settler, in which the sludge age (aerobic + anoxic) is equal to the hydraulic retention time. The basic process configuration of a Sharon process, represented in Fig. 4.24, is v simple and consists of a single completely mixed reactor that is operated at a temperature between 30 - 35°C and is subjected to alternating aerobic and anoxic conditions. The aerobic sludge age is controlled at a value low enough to prevent growth of the nitrite oxidisers, so the conversion of ammonium will proceed to nitrite only. The excess sludge produced is discharged with the effluent. Micto-organisms will predominantly be present as (clusters of) free bacteria rather than conglomerated into sludge flocs As a result, the biomass concentration in the Sharon effluent will be high and consequently the effluent will contain a considerable amount of organic nitrogen (refer also to Example 4.7). ‘The combination of a short residence time and a high ammonium influent concentration allows application of high volumetric nitrogen loading rates: full-scale reactors have demonstrated ammonium conversion rates between 0.4 to 0.8 kg Nand" at removal percentages of 80 to 95%, However, under these conditions other factors may become limiting such as substrate inhibition and/or -toxicity, the oxygen transfer- and diffusion rates and an excessive reactor temperature. A negative side-effect from the selection on fast-growing “feast” micro-organisms, which are adapted to abundant supply of substrate, is that these organisms have a lower substrate affinity than slow-growing “famine” organisms. Jetten et al (2000) established that the K,, value of the fast-growing ammonium oxidisers (in their research project identified as Nitrosomonas eutropha) ranges between 20 to 60 mg NH,NI" at temperatures between 30 to 35°C. As discussed in Table 4.3c, the K,, value of conventional mitrifiers is between 2 and 4 mg N.I" for the same temperature range. The combination of a high K,, value and short aerobic sludge age precludes a low effluent ammonium concentrationChapter 4 - Nitrogen Removal 137 100 nao =24507;byy= 0.80" | 2 nn = 2.0.6"; by = 0.84" 3 TP Se: ‘etvated nal Eton = 00 | “ae rs eae 2 systen = TL Retun stutge 2. ei & Ste L__ rpictener 2 ‘Thicvened 5 Tee se E 40 water E < = ‘Sharon Pason geste 5 a digest = Rojpt i Doestes water Sidge sege ; wating o 1 2 3 | Dewatered Aerobie sludge age (days) at Figure 4.25 Figure 4.24 Influence of the K, value on the effluent Sharon configuration for nitrogen ammonium concentration removal As shown in Eq, (4.31), the effluent ammonium concentration is independent of the influent concentration (except in the event of substrate inhibition). Effluent ammonium concentrations between 10 - 50 mg NI" have been reported for the full-scale Sharon reactors. This corresponds well with the range of values predicted by Eq, (4.31) for the values of the kinetic constants at 30-35°C, as can be seen in Fig. 4.23. Although it is possible to lower the ammonium nitrogen effluent concentration in a Sharon, this requires a significant increase in the applied acrobie sludge age (and thus in reactor volume). Asa result, nitrite oxidisers will be re-established in the reactor, converting the produced nitrite into nitrate at the expense of additional oxygen- and COD demand. Residual ammonium concentrations at the order of magnitude indicated above are not considered a problem in those cases where complete ammonium removal isnot required. For instance, when sludge digestion reject water is treated, the objective is the efficient and effective removal of the bulk of the nitrogen load, instead of obtaining a low effluent ammonium concentration: the effluent of the Sharon will be returned to the activated sludge process where any residual ammonium and nitrite can be removed 4.5.1.1 Kinetics of high rate ammonium oxidation As disenssed above, the observed maximum growth rate ly of the high-growth ammonium oidisers variant has been established at around 2 - 3 d"' in the temperature range of interest (30 - 35°C), Jetten et al (2000) reported data on the loss in nitrification rate resulting from the interruption of the feed to a lab-scale Sharon for a prolonged period of time at different temperatures. When itis assumed that the decrease in both maximum ammonium consumption rate and -nitrite production rate is directly proportional to the decrease in the active ammonium oxidising biomass concentration, the value of b, can be estimated using dX,/dt = -b,:X,. or X Xerexp(-b,.+t). From Jetten’s data, the value of b, is estimated at 0.23 d", which is much greater than the extrapolation of the value for conventional nitrifiers at 20 °C: b, =0.04+1.0472” = 0.07 d" at 35°C.138 ‘The difference might be attributed to predation, which more easily occurs in Sharon reactors where the bacteria are free in suspension instead of concentrated (and protected) in flocs. The (overall) nitrifier yield, Y,, has been determined previously in this chapter to be 0.16 g VSS.g"N oxidised: ic. for the complete oxidation of ammonium to nitrate, ‘Thus far, in literature no values have been reported for the yield of the two individual groups of nittifiers, An estimate may be made based on the following reasoning: the complete oxidation of one molecule of NH,” to NOs will deliver 8 electrons to the micro-organisms. The first step, oxidation of NH," to NOZ supplies 6 electrons, while the second step, oxidation of NOx to NOs, supplies only 2 electrons. Assuming the value of the yield is proportional to the number of electrons transferred, Y,, = 0.12 g VSS.g"N and Y,, = 0.04 g VSS.g™N. 45.1.2. Reactor configuration and operation In general, the Sharon reactor will be either constructed as a single tank with altemating aerobic and anoxic periods or as a series of two tanks, the first one aerobic and the second anoxic. The main advantage of the single tank concept is simplicity of construction. Furthermore the alkalinity produced during denitrification is immediately available to compensate for the acid produced in the aerobic phase. On the other hand, if the reactor is fed in the anoxic phase, part of the influent ammonium will be discharged directly with the effluent. However. the effect of this “short-circuiting” on the effluent ammonium concentration is limited, due to the dilution of the influent in the tank. A second advantage is that adjustment of the anoxic and aerobic sludge mass fractions is simple by manipulation of duration of the aeration period. For the configuration with two tanks, installation of a recirculation pump will be necessary to return the produced alkalinity in the denitrification tank to the nitrification tank The advantage of this configuration is that the influent will always be nitrified first and therefore the ammonium effluent concentration will be lower than in the other configuration. On the other hand, methanol consumption will be slightly higher as the easily biodegradable influent COD will be metabolised in the aerobic zone. A typical process cycle of the single tank Sharon consists of an aerobic period of 80 minutes followed by an anoxic period of 40 minutes (Ten Have, 2004). It follows that in the two tank configuration the aerobic tank will be twice as large as the anoxic tank. In the process of ammonium oxidation, regardless of whether oxidation proceeds to nitrite or nitrate, the removal of | mole of ammonium results in an acid production of 2 moles of Hor the equivalent consumption of 2 moles of bicarbonate (HCO;). Itis known that a pH value outside of the range between 6.5 and 8.5 will result in severe inhibition of the nitrification process, cither by free nitrous acid or free ammonia The same applies to the Sharon process, where it has been determined by Tetten et al (2000) that the degree of conversion of ammonium into nitrite is highest for pH values between 7.1 and 7.8. At a pH value below 7.1 the degree of conversion quickly decreases: while at a pH value of 6.8 about 50% of the influent ammonium is still converted: this ceases completely whea the pH is lower than 6.4. At pH values higher than 7.8, the free ammonia concentration slowly becomes an inhibiting factor. Even though sludge digestion reject water stoichiometrically contains 1 mole of bicarbonate per mole of ammonium generated during the mineralisation process in the digester, this is not enough to absorb the acid produced (in the form of H* ions) from complete nitrification ‘To compensate for the decrease in pH, either an inorganic base ean be added, e.g Ca(OH), or NaOH. or altematively the produced nitrite/itrate can be denitrified, which will generate alkalinityChapter 4 - Nitrogen Removal 139 In practice, denitrification with methanol as a carbon source is often the preferred solution, as it is cheaper than the addition of an inorganic base and because it will simultaneously reduce COD demand in the mainstream activated sludge system to which the treated reject water is retumed. Depending on the alkalinity present in the reject water, partial or full denitrification may be required. NaOH dosing equipment may be installed (or rented) as a back-up system or for pH control during reactor start-up 45.1.3 Required model enhancements Nitrogen removal over nitrite is easily incorporated in the ideal model, requiring only four changes to be introduced: (1) Inclusion of nitrifiers in the sludge: this subject is discussed in Appendix AS (2) Reduction of the oxygen demand for nitrification by 25% MO, = 0.75-4.57-MN, = 3.42-MNe G81) (3) Reduction of equivalent oxygen demand by 40% Mo, a = 0.6-2.86 = 1.72-MD, (482) (4) Increase in the value of f4, to reflect the decrease in COD demand for reduction of nitrite far = (I-foy°¥i)/(0.6-2.86) = 0.189 (483) Example 4.7 The management of a large waste water treatment plant (275,000 m*.d" and 1,650,000 P.E.) is considering side stream treatment of the sludge digestion reject water. The reject water production is 1831 m*.d", containing an average ammonium concentration of 665 mg N.I" Design a Sharon to treat this reject water stream. Minimum requirements are 95% removal of the ammonium load and a maximum effluent nitrite/nitrate concentration of 100 mg N.I" Compare the Sharon design to that of a conventional SBR in regard to effluent nitrogen composition, required treatment volume, oxygen demand, methanol- and caustic consumption and excess sludge production. Use the data in Tables 4.4 and 4.5 and apply the model enhancements in Appendix 5: i. include nitrifiers as a fraction in the volatile sludge and correct Nj for the loss of particulate organic nitrogen with the effluent. Table 4.4 Cycle times used in Example 4.7 Conventional SBR: Duration Sharon (stand alone) Duration Anoxic fill (pre-D) 10h | Aerobic: 80 min Anoxic react (pre-D): Osh Anoxic: ‘40 min. Aerobic react: 25h Anoxic react (post) 15h _ | Total cycle tme: 20h Decant & discharge (post-D) O6h Total cycle time: 60h140 Table 4.5 Kinetic- and stoichiometric parameters of Example 4.7 Par. Value UoM Par. Value) UoM Su 800 |mgCODI |Yao/Yn | 0.12/016 |mgVSS.mg™N fos /fnp__| 0.340.075 | (-) Himao/ bso | 205/019 |d" fab 0.2 () Kao! Kn 20/32 |mgN.I7 Re sar 8 d Hina Br 1.28/0.06 |" it 30 [°C Yn 045 |mgVSS.mg* COD Xeser 5.0 kgTSs.m* | K2/Ks 0.26/0.11_|mgN.mg" VSS.d" Xe aR 25 mgTSsi" |b, 036/057 | a" Aly 2380 |mgCaCOs.!" |f, 04 mg N.mg" VSS Noes 10 mg NIT oc 15 kg O2.kWh™ f 0.2 () fy 0.8 mg VSS.mg' TSS. Note: (1) Values of all kinetic parameters are given at the operational temperature of 30°C (2) Decay rate of high growth heterotrophic bacteria in a Sharon system Solution (a) Conventional nitrogen removal (SBR reactor) With some minor modifications, the theory presented in this chapter can be applied to a SBR reactor. The major change is the deletion of the “a’- and’‘s"- recirculation streams, required to return respectively nitrate and biomass to the activated sludge system in systems with a final settler. This change influences the values of Nat and Naw, as in the SBR configuration all nitrate producedin the aerobic phase will be available in the subsequent anoxic phase. As there are no dedicated anoxic and aerobic reactors in an SBR, the value off. and for is determined by the relative duration of the anoxic- and aerobic periods. In this example, the value of fy is (1.0 + 0.5)/6.0 = 0.25 and that of foe is 2.5/6.0 = 0.42. As nitrate is available in surplus during the decant- and discharge phase, denitrification using Spp as a COD source will continue. Therefore this process cycle is included in the post-D phase and fxs = 2.0/6.0 = 0.33. The bulk of the nitrate will be removed through the addition of methanol in the post-D react phase As the methanol will be completely metabolised in this phase, it will therefore not be available for metabolisation in the pre-D and the aerobic phase. However, in both phases the endogenous respiration is increased as a result of the increased active sludge concentration. This cycle set-up maximises the use of reject water COD for denitrification On the other hand, an influent buffer is required to store reject water produced during the react- and discharge phases of the SBR. As there are 4 process cycles a day, the required buffer volume equals (5/6)1831/4 or 381 m*. This volume could be reduced significantly to (0.6/6)-1831/4 or 46 m° if the feed flow to the SBR is only interrupted during the decant- and discharge phases. Additional methanol will then be required to compensate for the increased aerobic metabolisation of Svs. To remove 95% of the influent ammonium load, a maximum Nze concentration of 33.3 mg NI" ig allowed. Although the required sludge age is very short at 30°C, in practice other considerations (such as sludge settleability) require operation at a higher sludge age. Hence the (conservative) selection of Rs = 8 days.