A unified basis for the design of nitrogen removal activated sludge process — The Braunschweig exercise

e>

Pergamon

War. Sci.

te«

Vol. 38. No. I, pp. 227-236.1998 .

IAWQ Ie 1998Publishedby ElsevierScience Ltd.

Printed in Great Britain. AllrighlJ reserved

PIT: S0273-1223(98)00408-9

0273-1223198 SI9'oo + 0-00

A UNIFIED BASIS FOR THE DESIGN OF

NITROGEN REMOVAL ACTIVATED SLUDGE PROCESS - THE BRAUNSCHWEIG EXERCISE D. Orhon, O. Hanhan, E. Gorgun and S. Sozen Environmental Engineering Department, IstanbulTechnical University. 1.T.U.lnpaat Fakidtesi, 80626 Maslak; Istanbul; Turkey

ABSTRACf In this study a unified approach is defined for the design of single sludge pre-denitrification systems. based upon recently developed multi-component modelling of activated sludge. The "Braunschweig exercise" is adopted as a reference point, mainly for the critical review of different design methods proposed. Emphasis is placed upon the significance of process stoichiometry when using different parameters for substrate and biomass. Relationships proposed for the assessment of excess sludge production and oxygen requirements are evaluated in comparison with the unified approach. A new design methodology is defined and numerically tested with the basic data generated for the Braunschweig exercise.
KEYWORDS Activated sludge; endogenous decay model; nitrogen removal; pre-denitrification system design .; process kinetics; single sludge systems. INTRODUCfION Stringent effluent requirements necessitate expert approach in designing activated sludge systems for nitrogen removal. "The Braunschweig exercise" is perhaps the most striking example of recent efforts towards the improvement of system design supported by the mechanistic understanding of related process kinetics and stoichiometry (Kayser, 1991). This exercise is a request from interested scientists to calculate anoxic and aerobic volumes, oxygen requirement and excess sludge generation of a nitrogen removing activated sludge system, for the main purpose of comparing different design approaches to be displayed. Scientists from Austria, Denmark, Germany, South Africa, U.K. and U.S.A. have responded to the request and submitted their "homework". The basic data provided for design reflect primary effluent characteristics of an existing wastewater treatment plant. The statistical distribution of the daily BODs load was given as correlated with Corresponding BODs concentrations using a nominal daily flow rate of 32000 m 3 d-I. The magnitude of other significant polluting parameters in the primary effluent were defined, in relation to BODS as follows: SSIBODs = 0.80; TKNIBODS = 0.25; TPIBODS = 0.05. The correlation experimentally measured between

227

228

D. ORHON et al.

=

COD and BODs was plotted and could be evaluated with the following relationship: COD 30 + 2.25 BODs- The ratio between the maximum hourly and the daily average TKN loads was given as 1.81. 20°C and IO°C were indicated as summer and winter wastewater temperatures respectively. The effluent limitations were set to secure an effluent inorganic nitrogen concentration at 30 % of the incoming TKN concentration, at wastewater flow and loads 20 % higher than values associated with the existing plant. This study is an attempt to repeat and revive this very useful exercise with the main objective of defining a unified approach for the design of single sludge pre-denitrification systems, based upon recently developed multi-component modelling of activated sludge. The same basic data was adopted as a reference point, mainly for the critical review of different design methods proposed. Emphasis was placed upon the significance of process stoichiometry when using different parameters for substrate and biomass. Relationships proposed for the assessment of excess sludge production and oxygen requirements were evaluated in comparison with the unified approach. A new design methodology was defined and numerically tested with the basic data generated for the Braunschweig exercise. CONCEPTUAL BASIS A rational process evaluation requires mass balance relationships between major parameters and relevant microbial kinetics. Consistent derivations of such expressions have to rely upon multi-component models recently developed and proposed for activated sludge systems (Henze et al., 1987). The matrix structure of such a model based on the concept of endogenous decay is outlined in Table I (Orhon and Artan, 1994). In contrast to their traditional two-component counterparts, these models include, as illustrated in Table I, a number of substrate and biomass fractions as separate process components, each involving different stoichiometric and kinetic expressions. The most striking feature of the new modelling concept, as displayed in Table I, is the adoption of COD as a model component, a turning point in ending the empirical and often misleading guesswork for substrate utilization and oxygen requirement calculations based on BODs. Experimental fractionation of COD can now be used to identify biodegradable and inert components and biodegradable COD fraction, (CS>, sets an electron equivalence between substrate, active biomass and oxygen. BODs however is worse than an "iceberg", reflecting only a variable and very poorly defined portion of the organic substrate sustaining biological growth. Table 1. Process stoichiometry and kinetics in endogenous decay model for nitrogen removal Component

2

3

4

Xp

XH

Process

s

6

7

8

So

SALK

Process Rate

s,

_ tX/J

Aerobic Growth of Heterotrophs

·IXB

YII

·i XB

Anoxic Growth of Heterotrophs

_ l- YH

.....!.:.!!...-_l.rIJ 14~2.86

2.86Yu

-~-tx s,

Orowthof Autotrophs Aerobic Decay of Heterotrophs Anoxic Decay of Heterotrophs

J.Iii--XH Ks+Ss Ss

14

YH

14

1'\~

SNII

Il... - - X...

1

K~8NH

fA

I

-I

- (I-!Bx)

-I

~ !ll. 14

I~cayof

b"x...

• (1.fBx)

I"utotrophs COD

cell

cell

cell

COD

COD

COD

NH3-N

XH

Ks+Ss

molarunits

The Braunschweig exercise

229

The modelling approach also indicates that the heterotrophic yield coefficient, YH' is a parameter of capital importance in setting the necessary stoichiometric relationships for the design and operation of activated sludge systems. On the basis of the energetics of aerobic growth, this coefficient assumes a constant value when expressed in terms of g cell COD/g COD (eo.e biomassze,e substrate). Ekama and Marais (1984) reported a YH value of 0.66 g cell COD/g COD, experimentally determined to characterize domestic wastewaters. In the Task Group Model, an almost identical default value of 0.67 g cell COD/g COD was later advocated (Henze, et al., 1987). Theoretical evaluations based on the energy balance of domestic sewage yielded similar YH levels of 0.63-0.64 g cell COOl g COD (Orhon and Artan, 1994 ; Orhon et al., 1996). A recent experimental confirmation was also provided by Ubay Cokgor (1997), characterizing Istanbul domestic sewage with a YH range of 0.64-0.68 g cell COOl g COD. A very important point to note is that the above growth stoichiometry expressed in terms of electron equivalence (g cell COD/g COO) does not allow any latitude for the arbitrary selection of the heterotrophic yield coefficient in any other units; in fact, when "VSS/COD" or "VSSIBODs" units are selected for biomass/substrate as in the Braunschweig exercise, the corresponding heterotrophic yield coefficients must be defined as follows, using the basic stoichiometry for units conversion:

YH

Y~ (gVSS/gCOD)=T

(1)

x

(2) where,

=

fx COD equivalent of VSS ( 1.42 g cell COOl g VSS) f the ratio of BODs to the total biodegradable substrate

=

A UNIFIED DESIGN PROCEDURE BASED ON MODELLING Nitrogen removal in activated sludge systems may be achieved by means of different process configurations; therefore, a design exercise should not be initiated without clear specification of the selected flow scheme. Although the problem statement of the Braunschweig homework did not include this vital information, a pre-denitrification system was adopted in this study, to allow comparison with most of the solutions provided. A unified procedure for the design of the predenitrification process, based upon the modelling approach outlined in Table 2, is defined and presented in Figure I. This section will only elaborate in detail, the basic design data, sludge production and oxygen requirement steps of the proposed procedure. Detailed evaluation of the entire procedure is given elsewhere (Hanhan, 1994). Basic desi~n data A complete description of the basic data is a prerequisite for a reliable process design and should be considered as the decisive factor differentiating rational design from intuitive and empirical approaches. The required basic data should be evaluated in two major groups: wastewater characterization and kinetic and stoichiometric information.

Wastewater Characterization. Different wastewater characteristics used in this study and by the participants to the Braunschweig exercise are summarized in Table 2 (Barnard, 1991; Eckenfelder, 1991; la Cour Jansen, 1991; von der Emde, 1991). One of 'the suggested designs was left out mainly because it involved a totally empirical approach for a step feeding flow scheme (Boon, 1991). The model framework adopted and described for the evaluation dictates that COD be used at least as the organic substrate parameter. Table 2 however shows that BODS was generally preferred against COD, except in the South African design (Barnard, 1991). In this study, the 80 percentile BODs load, the nominal sewage flow of 32.000 m 3/day and the given BODs/COD relationship were used to calculate the influent BODs and the total COD, (C T 1), concentrations as 220 mgll and 525 mgll respectively. As no specific information was provided, the results

D. ORHON et al.

230

of extensive studies related to the COD fractionation of Istanbul domestic sewage were used. defining inert fractions as 15%, thus yielding an influent biodegradable COD,(C SI)' concentration of 450 mg/I (Orhon et al., 1994; 1997). The influent VSS concentration is an important parameter that gives. when coupled with the accompanying SS concentration. the magnitude of the fixed solids built-up in the reactor; the alkalinity of the influent stream is equally important for the pH balance of the nitrification-denitrification processes. As the problem statement of the Braunschweig exercise did not include any relevant information. a VSS/SS ratio of 80%. and an alkalinity concentration of 300 mgll Caco 3 were similarly adopted from the Istanbul domestic sewage studies.

II.

inpul DC desllllldata

12.leloct ellA

NO

14. leloct VoN (~0.5)

I'.

calculale total9x

...-----110. Ill.

",Ioclll· ~,

calQI\ate

s..o

NO

114. ",1m MLVSS. calculate v

116.

c:aJcuIU realdual alkallnlly

a-RESULTS Figure I. Proposed design procedure for predenitrification system.

The Braunschweig exercise

231

Table 2. Influent wastewater characterization data Parameter

This Study

Q m3d·1 .... I S.. mgBODs.r kg BODs.lfl Cyl,mgCOD.r l kgCOD.d· 1 c,..mgCon.rll kgCOD.lf Ss.. mgCOD.rl SII. mgCOD.rl XII, mgCOD.rl CTXNh mgTKN.rl kgTKN.lf' VSS,mgl·J SS,mgr l Alkalinity mgCaC0Jrl

S. Africa

U.S.A

(Barnard, 1991)

( Eckenfelder, 1991)

36.000 300 9000 70S 21150

30.000

32.000 220 7000 525

Denmark (1a CaurJansen. 1991) 32.000

ISO 5500

Germany (von derEmde, 1991)

40.000 200 8000

450

20250 52 25 SO 55

1760 140 175 300

25 37.5 1125

37.5 1125

44

SO 2000

1400

Kinetic and Stoichiometric Information. Kinetic and stoichiometric coefficients selected by different contributors to the design exercise are summarized in Table 3, together with the default values of the Task Group Model (ASMI) (Henze et al., 1987) and the ones used in this study, again characterizing Istanbul domestic sewage. Kinetics of microbial growth and substrate hydrolysis were overlooked with the assumption that all biodegradable substrate would be consumed at sludge ages sustaining nitrification and denitrification. Table 3. Kinetic and stoichiometric data used for the design Coefficient

YH YA

day·1 IlA bA day" bH datI

Ktm TJ fEX

mgl'

ASMI (Henze et aI. 1987) 0.67(11 0.24(2) 0.3

Denmark:

Germany

This Study

S.Africa

U.S.A

0.65(1) 0.20(2) 0.2 0.05 0.18 I 0.6

0.45(31

0.55(4)

0.6(6)

0.05 0.18

0.15 (5) 0.276 0.033 0.067

0.288 0.056

0.24 I 0.8 0.2 0.2 Ii)gcellCOD/gCOD (2) gccllCOD/gN

0.5 0.2 0.2 tl) gYSS/gCOD «) gYSSlgBOD (l)gYSSlgN

0.1 (6) gSSlgBOD

A very significant issue to be noted is the wide variation in Y H, both in terms of units and the arbitrary values selected. This variation may be explained by the fact that this coefficient requires correct assessment of the total biodegradable substrate and the ratio represented by BODs (f = g BODslg biodegradable COD), even if COD is not used as an organic substrate parameter; in other words, an accurate process evaluation based on BODs is only possible when it is supported and coupled with COD fractionation defining biodegradable COD for each specific case. In fact, the assumption of a biodegradable COD fraction of 85 %, a ratio characterizing Istanbul domestic sewage and supported by the existing experimental evidence (Henze, 1992), corresponds to f = 0.5 g BODslg COD, again a generally acceptable value for domestic sewage (Orhon et al.• 1997). Conversion of the proposed yield coefficients, using f = 0.5 in equations 1 and 2, yields YH values of 0.67, 0.41 and 0.36 g cell COOl g COD for the South African, the U.S. and the German design approaches respectively. Assuming a YH range of 0.6 - 0.7 g cell COD I g COD theoretically acceptable on the basis of electron balance, the corresponding ranges in other units selected may be computed as shown in Table 4 using the following conversion factor commonly accepted on the basis of process stoichiometry and derived from the problem statement: f=0.5 gBODslg.biodegradable COD, f 5=O.8 gVSS/gSS, fx=J.48 gVSS/gceIlCOD. The existing discrepancies from the given ranges inevitably reflect in the design steps and lead to wrong and misleading results for excess sludge generation and oxygen requirement.

232

D. ORRON et al.

Table 4. Range of theoretical yield coefficient

Generation of excess

Unit

Range

gVSSlgCOD gVSSlBOD, gSS/BOD,

0.41-0.47 0.8\-0.95 1.0\·1.18

slud~e

A rational evaluation of excess sludge should rely on a separate and consistent consideration of solid components ascertained by means of volatile suspended solids (VSS) and total suspended solids (SS) measurements. Microbial activities generate active biomass (XH) through growth and endogenous solids or residues (Xp) as end products of endogenous decay. There may be inert particulate organics (XU) in the influent stream that also accumulate in the biological reactor. In this context, the excess sludge is likely to have three major organic components defined interchangeably by VSS or cell COD measurements using the stoichiometric constant f x = 1.42 g cell COD/g VSS. with the provision that all influent particulate biodegradable organics (X S) are practically consumed at the selected operating conditions: Px (VSS)=Pm +PXE +PXJ

(3)

There may be solids of inorganic origin or fixed solids (X F \) in the influent accumulating in the biological reactor similarly to their organic counterparts; such solids. also generated to a much smaller extent through mineralization during endogenous decay. are quantified as: Xp

-x, (SS)-X r (VSS)

(4)

Consequently. the total excess sludge generation may be expressed as follows, in terms of its organic (VSS) and inorganic (SS - VSS) components: Pxr(SS) = Pm + PXE + PXJ + PXF

(5)

This expression assumes that sludge generation due to autotrophic activities is practically negligible. as commonly accepted in the design exercise. Equation 5 may be kinetically defined and conveniently rearranged to give the amount of sludge generated per unit volume of sewage treated, using relevant mass balance expressions derived from the model presented in Table I :

(6) where,

=

yIH 0.45 g VSS I g COD and c= 1-(1-11) . VoN

Different formulations proposed by the contributors for the computation of excess sludge production and arranged to allow comparative evaluation with equation 6. are given in Table 5. It may be stated that the South African expression is exactly the same as equation 6. except for the fixed solids term and the correction factor "c" that should be used to character ize. as previously defined, single sludge predenitrification system. All the other expressions suffer from the arbitrary value chosen for the heterotrophic yield coefficient. In the U.S. approach. the concept of endogenous solids generation coupled with a single biomass parameter and an overall endogenous decay coefficient which presumably covers this fraction is quite debatable. In the German practice the units specified for Yb (g SS/g BOOS> seems to be a misnomer. firstly because a YH' expressed in terms of the SS parameter involving a significant fixed solid fraction of influent origin. cannot be used to reflect the stoichiometry of microbial growth. but at best an observed yield coefficient which then should be a function of the sludge age. and secondly it creates a

The Braunschweig exercise

233

duplication with the fixed solids term "0.6 SS I". Furthermore the latter represents an unrealistically high value, at least with respect to the data presented in Table 3. This issue is elaborated in detail elsewhere (Sozen et al., 1997). Table 5. Kinetic interpretationof different approachesfor the assessment of sludge production PXFi'Q S.Afrita (kgYSS.If\)

USA

Pxv'Q

-

XII

-

.

(kgYSS.If\)

Denmark (kgSS. d·l )

Germany (kgSS.If\)

Qxy~en

0.68S.

Pxw'Q y IH4\-buylH4IElx/(I+1>H Ox) 0.4S 4'- 0.18 0.4S 4\ Elxl (1+{).18Elx) YH (S.-S) • buX T 6H1 (1+tEx bu Ox) O.SS (S.-S) -0.067 2.S 6H I (I + 0.2 0.067 Ox) Pxr/Q-y.., SI 0.8 S\ y_ SI - bH y_ S. ex I (1+bu Ox) 0.6 S•• 0.OS6 0.6S\ ex 1(1+{).OS6 Ox)

PXIlIY'Q (EXbu Y'H 41 exl (1+bu Ox)

0.2 0.18 0.4S 4' Elxl (1+{).18 Ox) (ex bu6HXr l (1+Cz bu Ox) 0.2 0.067 6H 2.SI (1+{).2 0.067 ex>

c..bH y_s. ex I (1+buex> 0.1 0.OS6 0.6 S\6xI(1+{).OS66x)

reQuirement

In multi-component activated sludge models. oxygen requirement relates to growth and endogenous decay processes of heterotrophic and autotrophic biomass. The heterotrophic oxygen requirement, ORH, is based upon the general principles of the stoichiometry of heterotrophic activities under aerobic conditions and it may be defined by the statement that the fraction of substrate removed which is not converted into biomass is equivalent to oxygen utilized. When both substrate and biomass parameters are defined in terms of COD. as in the model adopted for this study, the followingexpression reflects this stoichiometric relationship:

(7) Assuming again that all biodegradable COD is consumed and using the corresponding mass balance expressions. the above equation may be kineticallydefined as,

(8)

Oxygen equivalence of nitrogen transfonnations requires identification of three important parameters. The first parameter. NX. is the amount of nitrogen incorporated into biomass per unit volume of wastewater treated. Neglecting autotrophic biomass for practical reasons, Nx may be defined as:

(9) where iXB =nitrogen incorporated into biomass. The second parameter, Nox • reflects the amount of ammonia nitrogen oxidized or nitrate nitrogen formed per unit volume of wastewater treated: (10)

The third parameter, ND, defines the concentrationof nitrate nitrogen removed: (11)

234

D. ORHONet al.

In denitrification, the electron acceptor requirement of heterotrophs is partially satisfied by nitrate nitrogen: UR IID =2.86N o Q

(12)

Similarly, the oxygen utilization rate in nitrification may be expressed as: ORA --=(4.57-

Q

YA )Nox l+bA OX(

(13)

Neglecting again the biomass and endogenous decay of autotrophs, the above equation may be simplified into: UR... Q=4.57N ox

(14)

This way, the total oxygen requirement of the system may be expressed as; ORr 1 --=-(ORH+OR A -ORHD)

Q

(15)

Q

These expressions clearly indicate the electron equivalence and the direct balance between substrate, electron acceptor and biomass when both organic matter and the yield coefficient, Y H are measured and expressed in terms of COD. When substrate is defined as BODs and the yield coefficient involves VSS or SS, as is the case in the Braunschweig exercise, the above expressions can only be accepted and manipulated for the calculation of the oxygen requirement, if appropriate conversion factors are introduced. The definition of such factors necessitate correct information on the magnitude of related parameters (BODs/COD ratio. VSS/SS ratio) to be used consistently in each of the calculation steps. Table 6 outlines the different oxygen requirement expressions adopted by the contributors. To visualize the importance of the consistency of conversion factors, the oxygen requirement equation in the German approach for example, should be expressed as; (16)

Table 6. Kinetic interpretation of different approaches for the assessment of oxygen requirement S.Afiica

(kg.crl)

USA (kg.d")

Denmark (kg.crl)

Germany (kg.crl )

ORuIQ ORwiQ 2.8SNo.a (I - fx Y111l Ca. + fx(l -fex)bH Y1HCalllx l (I+bH9x) (1-1.42 0.4S) Ca. +1.42(1..Q.2)O.18 0.4SCa.llxl(1+0.18llx) 3(Nox--SIIO) (I·fx fylll (SI_- S) I f + fx(l-t£x) bH9HXTI (l+fE bH ex> 0.65(S._-S) + 1.42(I..Q.2) 0.067 9H 2.SI (I+0.20.067ex) 2.67No_ O.SS._+O.16WCT 0.5SI_+O.1 9H 4 2.9Nop (I - fx tify.) S.I f+ (bH fx fs Y.ex SI)I (1+bH 9x) O.SS. + ( 0.24 0.6 &x s.) I (I +O.OS6 &x)

ORAIQ 4.4SNox-

4.33Nox4.6Nox4.6Nox

Setting the equivalence of the first term of this expression with the one given in Table 6, is only possible with a BODS / COD ratio of 0.83 rather than 0.44, a value derived from the problem statements, for the fs ratio of 0.8 gVSS/gSS accepted for design. Similarly, the equivalence of the endogenous decay terms implies the following equality; (17)

which is only satisfied for b == 0.23/day.

The Braunschweig exercise Desj~n

235

results

In this study, evaluation of the exercise defined by Kayser (1991) on the basis of the unified approach based upon the multi-component modelling and the related design algorithm developed is presented in summary in Table 7, together with the results of other contributors. The discrepancy of different results is self evident on the basis of preceding discussions. Table 7. Comparative evaluation of design results Parameter

This Study 16 0.37 23 201.3 323 444.7

S.Africa

USA

Germany

20.5

S6

14.S

0.12 22.2 131.3 896.7 1093.3

0.S3

0.45

19

19.8 201.6 434.4 636.7

50.4 170.4

44S.6

Denmark

0.46 IS.87

IOS.6 496 694.4

CONCLUSIONS In the light of evaluations summarized in the preceding sections, the following points may be underlined as the concluding remarks related to the proposed unified approach for the design of nitrogen removal activated sludge systems: (a) Basic process calculations have to rely upon the electron balance between substrate, biomass and the electron acceptor. This balance is best defined with the COD parameter, coupled with a reliable COD fractionation defining the inert and biodegradable COD fractions. When BODs is selected as the substrate parameter, the significance of biodegradable COD is further increased in assessing the fraction of biodegradable organics represented by BODs. (b) The heterotrophic yield coefficient, YH is the main parameter for basic process calculations. On the basis of related scientific evidence a constant value may be associated to Y H when expressed in terms of e-.e biomass/e.e substrate. This leaves no grounds for arbitrary selection of YH values when other unit systems are preferred. (c) The recent multi-component models provide the required mechanistic clarification and support of the rational approach that should be adopted for the design of nitrogen removal activated sludge systems. NOMENCLATURE Subscript I concentrations in the influent, Subscript e concentrations in the effluent, bH, bA Endogenous decay rates for heterotrophs and autotrophs, T-I CN Biodegradable TKN, MN.L-3 Cs Total biodegradable substrate MCOD.L-3 CT Total substrate, MCOD.L-3 CTlrn Total Kjeldahl Nitrogen, MN.L-3 f Ratio of BODs to total biodegradable substrate, MBODs.(MCOD)-1 Inert fraction of biomass fEX COD equivalent ofVSS, McellCOD.(MVSS)-1 fx VSS to S5 ratio fs iXB, iXE nitrogen contents of active biomass and inert biomass, MN.(McellCOD)-1 Half saturation constants for carbon removal, MCOD.V 3, and ammonia oxidation, MN.L-3 Ks, KNH PXT Total organic sludge production, McellCOD.L-3 PT Total sludge production, MSS.L-3 Q Flowrate, L3.T-I

236

R

D. ORHON t!tal.

Total sludge recycle ratio

Rs, R I Sludge and internal recycle ratios

S SALK SI SNH SNO So Ss Xp

Total substrate MBOD5.L-3 Alkalinity, MCaC0 3.L-3 Inert soluble substrate MCOD.L-3 Ammonia nitrogen, MN.V 3

Oxidized nitrogen, MN.L-3 Oxygen, M02·V3 Readily biodegradable soluble substrate MCOD.V3 Fixed solids, MSS.L-3 Xl Inert particulate substrate, MceIICOD.LX H, X A Heterotrophic and autotrophic biomass, MceIICOD.L-3 Inert particulate product, MceIICOD.L-3 Xp XR Recycled biomass concentration, McellCOD.V3 V, VD Total and anoxic volumes, L3 Y H' Y A Heterotrophic, McellCOD.(MCOD)"1 , autotrophic, MceIICOD.(MN)"1 yield coefficients 11 Correction factor for anoxic conditions ~H' ~A Maximum specific growth rates for heterotrophs and autotrophs 9H Hydraulic detention time, T 9x, 9XA Total and aerobic sludge ages, T

REFERENCES Barnard, S. L. (1991). Design of nitrification/denitrification process. In: Dt!!ign For Nitrogen Removal and Guarantees lor Aeration. R.Kayser (ed .), Vertlffentlichungen des Instituts fllr Sledlungswasserwirt-schaft, 1U Braunschweig Heft SO E. 9-26. Boon.A. G. and Anderson, L. J. (1991). Nitrogen removal by the activated sludgeprocess. In: Design For Nitrogen Removal and Guarantees for Aeration, R. Kayser (ed.), VerOlfentlichungen des Instituts fOr Siedlungswasserwirtschaft, 1U Braunschweig HeftSO E. 27-64. Eckenfelder, W. W. Jr., (1991). Design example for nitrogen removal. In: Duign For Nitrogen Removal and Guarantu! lor Aerasion , R.Kayscr (ed.), Verllffendlictungen des Instituts filrSiedlungswasserwirt- schaft, 1U Braunschweig Heft so E, 65-74. Ekama, G. A. and Marais, G. v. R. (1984). Nature of Municipal Wastewaters. In Theory Design andOperation of Nutrient Removal Acttvated Sludge Process, Chapter2, Pretoria. SouthAfrica.WaterResearch Commission. 2.1-2.8. Hanhan, O. (1994) Biological nutrient removal and design alternatives in activated sludge systems. Master thesis, Ystsnbul Technical University (in Turkish). Henze, M., Grady. C. P. L. Jr., GujerW., Marais. G. v, R. andMatsuo, T.(1987). Activated SludgeModel No.1 IA WPRCSci. and Tech. ReportNo.1 , IA WPRC,London. Henze. M. (1992). Characterization of Wastewater for Modelling of Activated SludgeProcesses. Wat.Sci. Techn; 25(6), I-IS. Kayser. R. (1991). Design example for nitrogen removal. In: De!ign For Nitrogen Removal and Guarantees lor Aeration, R. Kayser(ed.), Verllffentlichungen des Instituts fUr Sicdlungswasserwirtschaft, TU Braunschweig Heft SO E, 1-7. la Cour Jansen, S. (1991). Danish design practice for nitrogen removal. In: De!ign For Nitrogen Removal and Guarantees lor Aeration, R. Kayser (ed.), Verllffentlichungen des Instituts flIrSiedlungswasserwirt schaft, 1U Braunschweig Heft SO E, 75-90. Orhon,D. and Artan, N. (1994). Modelling 01 Activatt!d Sludgt!Systt!m!. Technomic Press,Lancaster, PA. Orhon, D.• SOzen. S. and Ubay, E. (1994). Assessment of nitrification denitrification potential of Istanbul domestic wastewater. Wat. Sci. Teen; 30(6), 21·30. Orhon,D., SOzen, S. and Artan, N. (1996). The effect of heterotrophic yield on the assessment of the correction factorfor anoxic growth. Wat. Sci. Tt!ch.. 34(5·6),67·74. Orhon D•• Ales. E.• Sllzen. S. and Ubay Cokgllr. E. (1997). Characterization and COD Fractionation of Domestic Wastewaters. Envir. Pollut., 95(2). 191·204. SOzen, S•• Orhon, D.• Wilderer. P.• Anan. N. (1997) Evaluation of German Stsndarts A131 on the Basis of Process Kinetics for Carbonand Nitrogen Removal. Wat. Res., (submitted for publication). Ubay Cokgllr, E. (1997). Respirometric Evaluation of Process Kinetics and Stoichiometry for Aerobic Systems. Ph.D. Thesis. Istanbul Technical University. von der Emde, W. (1991). Design of wastewater treatment plants for nitrogen removal according to the ATV design guideline A13I. In: Design For Nitrogen Removal and Guarantee! lor At!ration, Verllffent·lichungen des Instituts fIlr Siedlungswasserwirtsehaft, 1U Braunschweig Heft SO E, 91-106.