Documente Academic
Documente Profesional
Documente Cultură
EXECUTIVE SUMMARY
In an existing community of about 110 000 people, each household has an existing septic tank- soak away system. However, environmental impact assessments have indicated that groundwater pollution is evident. Accordingly it is proposed to provide an aerobic activated sludge system for the community. There are two basic design considerations. The first is to connect the sewerage reticulation system into the septic tanks to be treated in an activated sludge system or to bypass the septic tank system completely. In this report an investigation is made into the two proposed activated sludge systems. The objective of this investigation is to determine which of these systems is the most economically viable solution. It should however be noted that the systems are only designed to accomplish total COD removal. A short overview is given of three different activated sludge model calibration protocols as background to wastewater design. These protocols are the BIOMATH protocol, STOWA protocol and the HSG guidelines. The objective of the protocol overview is to emphasize the importance of water characterization for any activated sludge system design. This report then covers the wastewater characterization for raw and settled sewage produced by the community. The calculation procedurefor the required volume of the two proposed activated sludge reactors are also included. It was found that the total COD present in the raw influent was 800 mg/L and in the settled wastewater it was 400 mg/L. This has a considerable effect on the required volume, oxygen required and overall capital cost of an activated sludge system. The sludge production was also found to be dramatically different between the two proposed systems. A summary of these results are given in the table below: Raw wastewater Settled wastewater 13638.1 4305.1 m3 ZAR 6,637,204.13 ZAR 2,338,587.78 6676.4 3509.4 kg/d 3636.8 6908.0 kg/d
It is finally recommended that an activated sludge plant without a primary settling tank be provided for the community. The initial capital costs are large, but over the operational lifetime of the activated sludge system the operational costs will be much lower than a settled wastewater facility. Another reason for this recommendation is the fact that an environmental impact assessment has already indicated that groundwater pollution is evident. The decommissioning of the septic tank system will ensure that ground water pollution is not a problem in the future.
LIST OF ABBREVIATIONS
ASM Bh COD dXt FSA fus fup fcv f fi M( ) MLVSS MLSS Nti Nobsi Nobpi Nousi Noupi OrgN Q Qw Rs Rhn Activated Sludge Model Endogenous Respiration Rate Chemical Oxygen Demand Sludge production concentration Free and Saline Ammonia Unbiodegradable Soluble COD fraction Unbiodegradable Particulate COD fraction COD/VSS ratio Endogenous Residue MLVSS/MLSS Ratio Mass Mixed Liquor Volatile Suspended Solids Mixed Liquor Suspended Solids Total Influent TKN Organic Biodegradable Soluble Influent TKN Organic Biodegradable Particulate Influent TKN Organic Unbiodegradable Soluble Influent TKN Organic Unbiodegradable Particulate Influent TKN Organic Nitrogen TKN Flow rate Waste flow rate Sludge Age Nominal hydraulic retention time
Sti Sbi Sui Supi Susi Sbsi Sbpi TKN TSS VSS Vp WWTP Xa Xe Xi Xv Xt Yh
Total Influent COD concentration Biodegradable Influent COD concentration Unbiodegradable Influent COD Concentration Unbiodegradable Particulate Influent COD concentration Unbiodegradable Soluble Influent COD concentration Biodegradable Soluble Influent COD concentration Biodegradable Particulate Influent COD concentration Total Kjeldahl Nitrogen Total Suspended Solids Volatile Suspended Solids Volume of the process reactor Waste Water Treatment Plant Active Volatile Solids concentration Endogenous Residue Volatile Solids concentration Inert Volatile Solids concentration Total Volatile Suspended Solids concentration Total Suspended Solids concentration Yield Coefficient
TABLE OF CONTENTS
LIST OF FIGURES ................................ ................................ ........... Error! Bookmark not defined. LIST OF TABLES................................ ................................ ............. Error! Bookmark not defined. 1. INTRODUCTION ................................ ................................ ................................ ........................ 5 2. CALIBRATION PROTOCOLS OF ACTIVATED SLUDGE MODELS ................................ ........ 5 2.1. BIOMATH CALIBRATION PROTOCOL................................ ................................ .............. 6 2.2. STOWA CALIBRATION PROTOCOL ................................ ................................ .................. 8 2.3. HOCHSCHULGRUPPE GUIDELINES ................................ ................................ ................ 10 2.4. SUMMARY OF PROTOCOLS ................................ ................................ ................................ 11 3. WASTEWATER CHARACTERIZATION ................................ ................................ .................. 12 3.1. WASTEWATER CHARACTERIZATION METHODOLOGY ................................ ............ 12 3.3. WASTEWATER CHARACTERIZATION SUMMARY ................................ .......................... 18 4. ACTIVATED SLUDGE SYSTEM DESIGN ................................ ................................ ............... 22 4.1. DESIGN METHODOLOGY ................................ ................................ ................................ . 24 4.2. ACTIVATED SLUDGE DESIGN RESULTS ................................ ................................ ....... 25 5. CONCLUSION AND RECOMMENDATIONS ................................ ................................ .......... 29 6. REFERENCES ................................ ................................ ................................ ............................ 31 APPENDIX A: WASTEWATER EXCEL CALCULATIONS ................................ ............................ I APPENDIX B: ACTIVATED SLUDGE EXCEL CALCULATIONS ................................ ............... II
1. INTRODUCTION
In an existing community of about 110 000 people, each household has an existing septic tank soak away system producing about 600 m3/h. However, environmental impact assessments have indicated that groundwater pollution is evident. Accordingly it is proposed to provide an aerobic activated sludge system for the community. It is therefore considered to either connect the sewerage reticulation system into the septic tanks to be treated in an activated sludge system or to bypass the septic tank system completely. In this report an investigation is made into the two proposed activated sludge systems. The objective of this investigation is to determine which of these systems isthe most economically viable solution. It should however be noted that the systems are only designed to accomplish total COD removal. The removal of the nitrogen and phosphates does not fall within the scope of this report.This report covers the wastewater characterization of the sewerage produced by the community as well as the estimation of the required volume for the two proposed activated sludge reactors. A short overview is given of three different activated sludge model calibration protocols. The objective of the protocol overview is to emphasize the importance of water characterization for any activated sludge system design. A finalrecommendationis then made for the most economically viable solution, whichis mainlybased upon the cost-to-volume relationship of the reactor.
calibration (Melcer, et al. 2003). Only the BIOMATH protocol, STOWA protocol and HSG guidelines will be briefly explained in this section. Firstly one needs a definition of what an activated sludge model entails. Activated sludge models are the compact and elegant summary of the state-of-the-art understanding of activated sludge processes (Grkan Sin, et al. 2005). Thus the protocols provide a framework where the ASMs may be calibrated to make use of dynamic models for a wastewater treatment plant (WWTP). The calibration procedure is complex and consists of steps such as lab-scale experiments for the characterization of wastewater and determining the stoichiometry of the biological processes active in the WWTP (Grkan Sin, et al. 2005). These complex steps have always been implemented by numerous methods of unsystematic calibration applications to full-scale modelling. The basic steps in these unsystematic calibrations are the identification of the influent and effluent variables, and the transfer of collected results from the lab-scale experiments to the characterization of the wastewater used in the calibration The waste . water characterization and basic stoichiometry are covered by a later sectionin this report. These basic steps continue in using the wastewater characterization variable to construct dynamic influent data to be implemented to the plant model. Dynamic data refers to data that covers a range of differences in the data due to varying different factors such as time, temperature, sludge age, etc. After the dynamic influent data has been created, the data is then applied to the ASM and calibrated to obtain a satisfactory fit to the measured variables in the WWTP such as the MLSS. Calibration is normally performed manually, this thus means that the modeller changeseach parameter one at a time until a satisfactory fit is found and this way of calibration is based upon expert knowledge only (Grkan Sin, et al. 2005). Thus far every calibration modelfollows its own approach and procedures with influent lab experiment characterization, stiochiometric estimations, model hydraulics and settling characterization (Grkan Sin, et al. 2005). It can thus be deducted that by using these unstructured and unsystematic procedures it is not possible to compare result from different models. This is where the introduced protocols may develop a framework for a more uniform calibration method. These calibration protocols have the main objective of aiding the modellers and creating a standard from where the models may be analysed and compared (Grkan Sin, et al. 2005).
performance data is obtained (Grkan Sin, et al. 2005). It is possible to perform quality checks on the obtained data through the application of mass-balances for the COD, Nitrogen, etc. This obtained data then leads to the selection of the different method for steps 3, 4 and 5. The selection of the models for the hydraulic and aeration model, settling characterization and the biological processes are of utmost importance for an accurate study of the proposed WWTP. In the selection process these models are calibrated individually from the data collected in the previous steps. The models are then incorporated together in the full scale model and calibrated once again for steady state conditions.The objective of the steady state conditions is then to obtain a satisfactory fit to the parameters of the WWTP. These parameters are the sludge production and the oxygen consumption within the activated sludge process. The estimation of the sludge production and the oxygen consumption are discussed in a later section of this report. The calibration of the selected parameters is performed in conjunction with a sensitivity analysis on the parameters that affect the sludge production over a long period such as bh and Yh (Grkan Sin, et al. 2005). After the calibrations of the steady state conditions have been performed the dynamic calibration of the full-scale model is performed. The most important parameter is once again calibrated using a sensitivity analysis. After completion of the dynamic calibration the model is then validated. This is performed by collecting data from the WWTP under different operation conditions and if the calibrated model delivers satisfactory predictions, the model is accepted for modelling. If the predictions are not found to be satisfactory, the whole calibrationprocess is repeated (Grkan Sin, et al. 2005).
Stage I
Stage III
12. Evaluations
ASM. In the next step the choice and description of the required processes can be made. In the third stage, the data relevant to the calibration and application of the ASM are identified and defined. The data obtained is then validated for different flows and process volumes. The verification is once again performed with the use of mass balance checks for data consistency.If the obtained datais not consistent, the data should be corrected before any further calibration steps are taken (Grkan Sin, et al. 2005). The fourth step consists of defining the model structure, through selecting the appropriate models for the WWTP hydraulics, settlers and aeration. The fifth step is the wastewater characterization and quantification of the influent and effluent as described by this report. The STOWA protocol bases its wastewater characterization on the biological oxygen demand tests and presents a detailed discussion on the use of BOD tests (Grkan Sin, et al. 2005). The sixth step is where the ASM is calibrated according to the obtained results and a preliminary calibration is performed to compare the simulated results with the actual WWTP data. If there are large differences between the simulated data and measured data there must be additional data validation through mass balances or the data should be obtained again. If the verification is satisfactory a detailed calibration is performed on the ASM and follows a manual calibration procedurefor finding the best fit to the WWTP data (Grkan Sin. et al. 2005). After successful calibration of the ASM the data should be validated against the WWTP data and can then be used for its study.
1. Formulation of objectives
2. Process description
5. Characterization of flows
6. Calibration
8. Validation
9. Study
7. Detailed characterization
10
1. Definition of objectives
3. Data quality control 4. Evaluation of o el structure an ex eri ental esign 5. Data collection for si ulation stu y
3. WASTEWATER CHARACTERIZATION
Wastewater characterization greatly affects the biological treatment of waste in activated sludge system designs. These flows can be characterized chemically and physically. Chemical characterization consists of determining the different fractions of carbonaceous and nitrogenous elements in the wastewater. These fractions are then further divided into biodegradable and unbiodegradable groups that can be classified as soluble or particulate. The physical characterization consists of distinguishing between dissolved, suspended and settleable fractions. The following sections describe the basic methodology and results for characterising the wastewater. Data for the wastewater was obtained through water batch testing and is provided in this report. The test result data used for characterization ispresented in Table 1.
tank or treating raw wastewater directly without any primary sedimentation.The first step is to calculate what concentration of the COD and TKN is removed by primary sedimentation within the septic tank. The assumption is made that 100% of settleable solids are removed within the primary settling tank (septic tank). The COD of settled solidsis the difference between the raw unsettled COD concentration and the effluent (supernatant) from the septic tank. The calculation is similar for the TKN and FSA. The calculations are represented in the Table 2.
Settled solids
COD TKN FSA = = = Raw COD - Septic COD = Raw TKN - Septic TKN = Raw FSA - Septic FSA = 400 mg COD/L 12 mg N/L 0 mg N/L
Next the suspended solids are calculated. The suspended material is the solids retained by the 0.45m filter membrane. This is basically the difference between the septic tank effluent and the filteredraw sewage for the COD, TKN and the FSA. The calculations are represented in Table 3.
Suspended solids
COD TKN FSA = = = Septic COD - Filtrate COD = Septic TKN - Filtrate TKN = Septic FSA - Filtrate FSA = 184 mg COD/L 2.1 mg N/L 0 mg N/L
Now the different concentrations for settleable, suspended and soluble materials can be determined from the above results. The flow that has passed through the 0.45m filter membrane represents the solublefraction of wastewater. A comprehensive summary of all the settleable, suspended and soluble material concentrations is represented by Table 4. The soluble fraction is represented by the single soluble portion and the particulate portion is represented by the suspended and settleable materials in the wastewater.
13
Septic COD
Raw TKN
Septic TKN
Raw FSA
Septic FSA
The next classification comprises of distinguishingbiodegradable and unbiodegradable materials that are soluble, suspended and settleable. The results from the aerobic laboratory-scale activated sludge system are used to calculate the unbiodegradable soluble COD (S ). It is assumed that all the usi biodegradable COD, soluble and particulate, are degraded within the laboratory-scale biological reactor. This assumption allows us to determine the Susi concentration directly from the laboratory results. Thus from the laboratory results, the Susi component of the wastewater can be calculated as a fraction of the total COD (Sti). This fraction is represented by the symbol fus. Another fraction of importance is the unbiodegradable soluble organic nitrogen (fnu). This fraction is used to determine the unbiodegradable Organic Nitrogen as described later in the next section. The calculations for Susi,fus and fnuare shown in Table 5.
14
Unbiodegradable
Susi =
Raw sewage Laboratory Effluent COD = Septic tank effluent Laboratory Effluent COD = Raw sewage Susi / Sti = (Eff TKN - Eff FSA)/Influent TKN = Septic tank effluent Susi / Sti = (Eff TKN - Eff FSA)/Influent TKN =
56 mg COD/L
Susi
56 mg COD/L
fus fnu
= =
fus fnu
= =
Now the different components of the wastewater classification can be determined. Assumptions are made for the unbiodegradable particulate fractions (fup ). Ekama & Marais ([s.a.])prescribe values of fup = 0.15 for unsettled conditions and fup = 0.04 for settled conditions. All the necessary calculations for the raw sewage and the septic tank effluent are presented in Table 6 and 7 respectively. The equations are presented by Ekama & Marais ([s.a.]). The value of the nitrogen fraction of the influent biodegradable volatile particulate material is prescribed by Ekama & Marais ([s.a.]) as fn = 0.1 mgN/mgXii. The COD/VSS ratio of the volatile solids is a constant value where fcv = 1.48 mgCOD/mgVSS (Ekama, et al. 2008).
15
Sti * ( 1 - fus - fup ) = Sti - Sbi = Soluble COD - Susi = Sbi - Sbsi = Sti * fus = Sti * fup =
TKN
Nti FSA OrgN Unbiodegradable OrgN Biodegradable OrgN Noupi Nousi Nobsi Nobpi = = = = = = = TKN = 70 mg N/L 52.5 mg N/L 17.5 mg N/L
Nti - FSA =
Noupi + Nousi = 10.19 mg N/L Nobsi + Nobpi = 7.31 mg N/L fn * Supi / fcv = fn * Nti = 8.11 mg N/L 2.08 mg N/L 1.32 mg N/L 5.99 mg N/L
16
Sti * ( 1 - fus - fup ) = Sti - Sbi = Soluble COD - Susi = Sbi - Sbsi = Sti * fus = Sti * fup =
56 mg COD/L 16 mg COD/L
TKN
Nti FSA OrgN Unbiodegradable OrgN Biodegradable OrgN Noupi Nousi Nobsi Nobpi = = = = = = = TKN = 58 mg N/L 52.5 mg N/L 5.5 mg N/L 3.16 mg N/L 2.34 mg N/L 1.08 mg N/L 2.08 mg N/L 1.32 mg N/L 1.02 mg N/L
17
% % % % mg/L % %
18
% % % % mg/L % %
All these wastewater characteristics are not however required in the next section to determine the reactor characteristics for COD removal, but is presented in this report to demonstrate the methodology required for basic wastewater characterization.
19
20
21
Figure 6: Activated sludge system with a primary settling tank (Stellenbosch University, 2005)
22
Figure 7: Activated sludge system without a primary settling tank (Stellenbosch University, 2005)
In the reactor the biodegradable organic material is utilized by the heterophic organisms to reproduce and maintain their population(Ekama, et al. 2008). These organisms form part of the volatile suspended solids (VSS) in the reactor that is called the sludge. There are three different sludge fractions that are applicable to determine the required volume forCOD removal in activated sludge reactor. These components are the active sludge (Xa), endogenous residue (Xe) and the inert sludge (Xi) (Stellenbosch University, 2005). The active sludge represents the living organisms in the reactor that are actively utilizing the substrate. The organisms also use oxygen (O2) throughout the whole process. Oxygen is utilized firstly as anelectronacceptor to oxidize the substrate in orderto create new cell material and secondly by the respiratory functions of the organisms (Stellenbosch University, 2005). In Figures 6 and 7 this oxygen demand is represented as a required addition to the reactor contents.The second fraction of the sludge is the endogenous sludge. The organisms in the reactor die and leave behind unbiodegradable organics that is called endogenous residue. This then also becomes part of the VSS mass in the reactor. Normally there are additional particles in the influent that are 23
biologically inactive. These particles do not take part in any reactions, but form part of the produced sludge and is called the inert sludge. It is very important to have an accurate estimation of the wastewater characteristics. The characterization for COD removal is however elementary, but it is very important to determine the influent COD and the different fractions of biodegradable and unbiodegradable COD. The unbiodegradable particulate COD concentration has a strong effect on the sludge mass accumulation within the reactor (Ekama, et al. 2008). The systems under investigation are analysed and designed by assuming steady state (Q = 600 m3/h) conditions and that 100% of the settleable solids in the influent does settle in the primary settling tank. In the next section the basic activated sludge COD removal system design for the steady state is described and applied to the two activated sludge systems in Figures 6 and 7.
1. Supi, Susi, Xii 2. M(Sti) 3. M(Xa), M(Xe), M(Xi), M(Xv), M(Oc), M(Xt) 4. Vp 5. Rhn 6. Ste
The values for Supi , Susi and Xii can be calculated from the total influent COD (Sti ) and the
fractions fus and fupis as determined by the wastewater characterization. The second step is calculating the mass of the total influent COD ( M(Sti) ). From M(S ti) the other mass fractions for the active sludge, endogenous sludge and inert sludge can be determined. The total volatile suspended solids massM(Xv) is then represented by the sum of M(Xa), M(Xe) and M(Xi). The total suspended solids can then be determined from M(Xv) through the MLSS/MLVSS ratio (fi). The carbonaceous oxygen can also be determined from the known relationships of fus, fup and Sti. All the above characteristics are dependant on the sludge age (Rs), yield coefficient (Yh), endogenous respiration rate (b h), endogenous residue (f) and the COD/VSS ratio for volatile solids (fcv). All these values are constant except for bh which depends on the reactor temperature. The only chosen parameter is the sludge age. All of the
24
above characteristics are heavily dependant on the sludge age. In example, a long sludge age will lead to large sludge masses in the reactor and will directly lead to large reactor volumes. The reactor volume calculation is determined by the total mass of solids (MXt ) in the reactor and a specified value for the MLSS concentration allowed in the reactor. With the required reactor volume known it is possible to calculate the nominal hydraulic retention time (Rhn). The nominal hydraulic retention time is however irrelevant to the design of an activated sludge system (Ekama, et al. 2008). Finally the effluent COD concentration (Ste) is determined. The COD in the effluent mainly consists of soluble unbiodegradable organics (COD). The soluble biodegradable COD is completely utilized within a short period. The biodegradable and unbiodegradable particulate is however absorbed and forms part of the sludge in the system (Ekama, et al. 2008). The sequence of calculations briefly explained above is best illustrated by Table 10 and 11 in the next section. The relevant formulas are also presented in the tables and explain the calculation procedure.
The values in orange are the required input values for the Excel spreadsheet.
Raw wastewater
Input from wastewater characterization
Sti = fus = fup = fi = T= fcv = Rs = 800 mg COD/L 0.07 0.15 0.750 17 mgCOD/mgCOD mgCOD/mgCOD MLVSS/MLSS C
(given) =
25
Q= MLSS = Yh = bh = f=
14400 m3/d 4000 mg TSS/L 0.45 mgVSS/mgCOD 0.2203 /d 0.2 mgVSS/mgVSS = 1.029
(given) =
6676.4 kg O/d
Vp = Cost =
M(Xt) / MLSS = M(Xt) / Xt 13638.1 m3 1000 * 770 * (Vp/1000)^0.761 = ZAR 6,637,204.13 Total price of two reactors if the volume is larger than 10 000 m2 Vp / Rs = Vp / Q = Mass wasted per day = M(Xt) / Rs = Filtered effluent = Susi =
3 909.2 m /d
26
Settled wastewater
Input from wastewater characterization
Sti = fus = fup = fi = T= fcv = Rs = Q= MLSS = Yh = bh = f= 400 mg COD/L 0.14 0.04 0.850 17 mgCOD/mgCOD mgCOD/mgCOD MLVSS/MLSS C
(given) =
1.48 mgCOD/mgVSS 15 days 14400 m3/d 4000 mg TSS/L 0.45 mgVSS/mgCOD 0.2203 /d 0.2 mgVSS/mgVSS = 1.029
(given) =
3509.4 kg O/d
Vp = Cost =
M(Xt) / MLSS = M(Xt) / Xt 4305.1 m3 1000 * 770 * (Vp/1000)^0.761 = ZAR 2,338,587.78 Total price of two reactors if 27
the volume is larger than 10 000 m2 Qw = Rhn = Vp / Rs = Vp / Q = Mass wasted per day = M(Xt) / Rs = Filtered effluent = Susi =
3 287.0 m /d
0.299 days
M(dXt) = Ste =
Total system mass wasted per day = M(dXt) + Settled solids in primary settling tank*Q = 6908.0 kg / d
The above tables represent the basic calculations of the sludge masses and required reactor volumes. The raw wastewater required volume is more than 10 000 m3 and thus the required volume is divided between two reactors. This implies that the total capital cost for the required reactors for raw wastewater are for two reactors. In the case of settled wastewater there is a primary settler present. This causes that a large amount of sludge is trapped in the primary settling tank (septic tank). Thus the actual total mass of sludge wasted per day is the settled sludge in the primary settling tank together with the sludge wasted by the reactor. This calculation is shown above.In Table 12 the criteria for design selection is given for the two different proposals. The last two columns on the right of Table 12 indicate how much the values change from raw wastewater to settled wastewater.
Raw wastewater Settled wastewater 13638.1 4305.1 m3 ZAR 6,637,204.13 ZAR 2,338,587.78 6676.4 3509.4 kg/d 3636.8 6908.0 kg/d 909.2 287.0 m3 mg 56.0 56.0 COD/L
28
costs of the system. The treatment of raw waste water however does have its advantages through the fact that the overall sludge production is much less than with the settled wastewater system (47.4% lower). The sludge that is produced from the secondary settling tank is less, thus thesludge extraction required is less.The process of stabilizing the wasteis also accomplished much easier due to the fact that the energy content of the waste is lower due to degradation of the sludge by the micro-organisms. The last two points contribute to far cheaper operational costs. It is finally recommended that an activated sludge plant without a primary settling tank be provided for the community. The initial capital costs are large, but over the operational lifetime of the activated sludge system the operational costs are much lower than a settled wastewater facility. Another reason for this recommendation is the fact that an environmental impact assessment has already indicated that groundwater pollution is evident. The decommissioning of the septic tank system will ensure that ground water pollution is not a problem in the future. The recommendation presented in this report only serves as a preliminary recommendation. Further studies and investigations should be made on the impact of the proposed activated sludge system. Other site specific factors should also be considered for the proposed activated sludge system . Theseare the sufficiency of existing sludge treatment facilities, sludge disposal, the future growth of the community and the available finances and adequate management services.
30
6. REFERENCES
Ekama, G.A. Marais, G.v.R. [s.a.]Nature of municipal wastewaters. Cape Town:University of Cape Town. Ekama, G.A. Wentzel, M.C. 2008.Biological Wastewater Treatment: Principles, Modelling and Design, Chapter 4: Organic Material Removal. Cambridge University Press. Grkan Sin. et al. 2005. A critical comparison of systematic calibration protocols for activated sludge models: A SWOT analysis. Ghent University, Belgium. Hulsbeek, J.J.W. et al. 2002. A practical protocol for dynamic modelling of activated sludge systems. Water Science Technology 45 (6), 127-136. Langergraber, G. et al. 2004. A guideline for simulation studies of wastewater treatment plants. Water Science Technology 50 (7), 131-138. Melcer, H. 2003. Methods for wastewater characterization in activated sludge modelling. Water Environment Research Foundation, Alexandria, USA. Vanrolleghem, P.A. 2003. A comprehensive model calibration procedure for activated sludge models. 76th Annual Technical Exhibition and conference, Los Angeles.
31
II