İSTANBUL TECHNICAL UNIVERSITY  INSTITUTE OF SCIENCE AND TECHNOLOGY M.Sc. THESIS JUNE 2013 AN ANAEROBIC TREATMENT MODEL APPROACH FOR INDUSTRIAL WASTEWATER TREATMENT: CASE STUDY FOR FULL SCALE ANADOLU EFES WASTEWATER TREATMENT PLANT ANAEROBIC MODELING Deniz ALBAYRAK Department of Environmental Engineering Environmental Biotechnology Programme Anabilim Dalı : Herhangi Mühendislik, Bilim Programı : Herhangi Program JUNE 2013 İSTANBUL TECHNICAL UNIVERSITY  INSTITUTE OF SCIENCE AND TECHNOLOGY AN ANAEROBIC TREATMENT MODEL APPROACH FOR INDUSTRIAL WASTEWATER TREATMENT: CASE STUDY FOR FULL SCALE ANADOLU EFES WASTEWATER TREATMENT PLANT ANAEROBIC MODELING M.Sc. THESIS Deniz ALBAYRAK (501091816) Department of Environmental Engineering Environmental Biotechnology Programme Anabilim Dalı : Herhangi Mühendislik, Bilim Programı : Herhangi Program Thesis Advisor: Assoc. Prof. Dr. H.Güçlü İNSEL İSTANBUL TEKNİK ÜNİVERSİTESİ  FEN BİLİMLERİ ENSTİTÜSÜ ENDÜSTRİYEL ATIKSULAR İÇİN ANAEROBİK ARITMA MODEL YAKLAŞIMI: ÖRNEK ÇALIŞMA ANADOLU EFES ATIKSU ARITMA TESİSİNİN MODELLEMESİ YÜKSEK LİSANS TEZİ Deniz ALBAYRAK (501091816) Çevre Mühendisliği Anabilim Dalı Çevre Biyoteknolojisi Programı Anabilim Dalı : Herhangi Mühendislik, Bilim Programı : Herhangi Program Tez Danışmanı: Doç. Dr. H.Güçlü İNSEL HAZİRAN 2013 v Deniz Albayrak , a M.Sc. student of ITU Institute of Science and Technology s tudent ID 501091816, success full y defend ed the thesis entitled “AN ANAEROBIC TREATMENT MODEL APPROACH FOR INDUSTRIAL WASTEWATER TREATMENT: CASE STUDY FOR FULL SCALE ANADOLU EFES WASTEWATER TREATMENT PLANT ANAEROBIC MODELING”, which she prepared after fulfi ll ing the requireme nts specified in the asso ciate d legisl ati ons, befo re the jur y whos e sign atures are below. Thesis Advisor : Assoc. Prof. Dr Hayrettin Güçlü İNSEL .......................... İstanbul Technical University Jury Members : Prof. Dr. Orhan İNCE ............................. İstanbul Technical University Prof. Dr. Neşe TÜFEKÇİ .............................. İstanbul University Date of Submission : 03 May 2013 Date of Defense :09 June 2013 vi vii FOREWORD First of all, I would like to deeply thanks to my advisor Assoc. Prof. Dr. H.Güçlü İnsel. I always felt confident under his supervision and support. He answered my questi ons and solved pr oblem with a great pati e nce. It was a bi g chan ce for m e to work with him . I would like to sa y man y thanks for endless support and beli eving in in me scope of this stud y. I would like to thank to Anadolu Efes Lüleburgaz Factory and Mr. Şamil Ateş for sharing their wast e wat er treatm en t plant’s exist ing information and ope rat ional data also thank to their cooper ati on for this stud y . Finall y, I would also ind icate m y gr ati tude to m y fami l y for m y whole lif e that the y provided me, for all kindl y supports and fo r bein g with me alwa ys. I dedicate this thesis to my fami l y Ma y 2013 Deniz ALBAYR AK Environmental Engin eer vii i i x TABLE OF CONTENTS Page FOREWORD ............................................................................................................ vii TABLE OF CONTENTS .......................................................................................... ix ABBREVIATIONS ................................................................................................... xi LIST OF TABLES .................................................................................................. xiii LIST OF FIGURES ................................................................................................. xv LIST OF SYMBOLS ............................................................................................. xvii SUMMARY ............................................................................................................. xix ÖZET ........................................................................................................................ xxi 1. INTRODUCTION .................................................................................................. 1 1.1 Aim of Stud y ................................ ................................ ................................ ...... 1 1.2 Scope of the Stud y ................................ ................................ .............................. 2 2. LITERATURE ....................................................................................................... 3 2.1 Introdu cti on ................................ ................................ ................................ ........ 3 2.2 The Anae robic Tr eat ment Process ................................ ................................ ..... 3 2.2.1 Anaerobi c de gr adat ion of organic pol ymers ................................ ............... 6 2.2.2 Working with COD balance ................................ ................................ ........ 8 2.2.3 Anaerobi c re actor sys tem ................................ ................................ .......... 10 2.2.4 Anaerobi c proc ess kineti cs ................................ ................................ ........ 22 2.3 Modeling ................................ ................................ ................................ .......... 23 2.3.1 Sim ple models and prin cipl e kinetics ................................ ....................... 24 2.4 Brewer y Wastew ater ................................ ................................ ........................ 40 2.4.1 Brewer y wastew ate r char acte risati on ................................ ........................ 40 2.4.2 Brewer y wastew ate r treatm ent methods ................................ ................... 42 3. THE DEVELOPED MODEL STRUCTURE .................................................... 49 3.1 Introdu cti on of t he Model ................................ ................................ ................ 49 3.2 The Conceptual Of The Developm ent Model ................................ .................. 50 3.3 The Processes and Components of The Model ................................ ................ 52 3.4 The Kineti c and Stoi chiom etr y of the Model ................................ ................... 53 4. MATERIAL AND METHODS ........................................................................... 57 4.1 Con ceptual Approa c h ................................ ................................ ....................... 57 4.2 Efes Pilsen Lüleburgaz Plant ................................ ................................ ............ 57 4.2.1 Wastewater tre atm e nt plant ................................ ................................ ....... 58 4.3 Model Impl em entation Using Op erati onal Data ................................ .............. 62 4.3.1 Sim ulation ................................ ................................ ................................ . 62 4.3.2 The Wastewater C OD Fr ac ti ons Fo r Simul ati on ................................ ...... 63 4.3.3 Stead y State Sim ulation ................................ ................................ ............ 63 4.3.4 Dyn ami c sim ulation ................................ ................................ .................. 65 5. RESULTS AND DISCUSSION .......................................................................... 69 5.1 Stead y State Sim ulation Result s ................................ ................................ ....... 69 5.2 Dyn ami c Sim ulation Result s ................................ ................................ ............ 71 x 6. CONCLUSION ..................................................................................................... 81 CURRICULUM VITAE .......................................................................................... 85 x i ABBREVIATIONS ASM : Acti vated slud ge model ADM : Anae robi c Di gesti on M odell ing AD : Anae robic Dgesti on IWA : I nternati onal Water Ass ociation IC : Inor ganic C arbon IN : Inor ganic Nitro gen BOD : Bioch emi cal Ox ygen Demand, mg/l COD : C hemi cal Ox ygen Dem ans, mg/l F/M : Food/M icroor ganism s rati o, d - 1 HRT : Hydraul ic Retention Ti me, h IR : Internal Re c ycle ML(V)SS : Mix ed Liquor (Volati le ) Suspended Soli ds, mg/l STD : S tandart Deviation SRT : S ludge Retention Time, d TKN : Total Kjeldahl Nit ro gen , mg/l TSS : Total Suspended Soli ds, mg/l T-P : Total Phosphorus , mg/l LCFA : Lon g Ch ain Fatt y Acid SCFA : S hort Ch a in Fatt y Acid CSTR : Compl etel y Sti rred Tan k Reactor EGSB : Ex panded Granula r Sludge Bed R eactor AnWT : Anae robic Wastewate r Treatm ent WWTP : Wastewater Treatm ent Plant VFA : Volati le Fatt y Acid, m g/l x ii x iii LIST OF TABLES Page Table 2.1: Applicati on of anae robic te chnol ogy to industrial wastewater………... . 13 Table 2.2: Kineti c param eters of main substrat es/i ntermediate produ cts in the anaerobic conversion process…………………………………….. . ..... . . . 23 Table 2.3: Overview of some anae robic di gesti on models ……………………….. . 26 Table 2.4: The brew er y wastewate r gen eral ch ar a cteri s ation………….... .... .......... . 41 Table 2.5: The COD fractions of brewery wastewater…………………………….. 4 2 Table 2.6: Wastewater treatment unit operations and processes…………………. .. 43 Table 2.7: Anaerobic treatment as compared to aerobic treatment………………. .. 48 Table 3.1: Biochemi cal rate coe fficients (V i , j ) and processes for the dev elop ed model. Stoi chiom etr y fo r Gujer Matrix for Anaerobic Model ……… ... . 5 4 Table 3.2: Kineti c rate equati ons (p j ) of the devel oped model …………………... . 55 Table 4.1: Avera ge annu al wastewat er ch ara cteriz ati on of the plant …………….. .. 58 Table 4.2: The wastew ater cha ra cterisatio n of the inl et EGSB rea ctor …………… . 62 Table 4.3 : The COD frac ti ons of brewer y wa stew ater f or simulation…………….. . 63 Table 4.4: Influ ent waste water ch ar acteris ati on used in stead y state sim ulation ….. 6 4 Table 4.5: Kinetic and stoichiometric values of th e literature model s…………… ... 64 Table 5.1: Kinetic and stoichiometric values that determined for the model…… .... 69 Table 5.2: S tead y state s im ulation result «………………………………………... 7 0 x iv x v LIST OF FIGURES Page Figure 2.1 : Fate of carbo n and ener g y in aerobi c and anaerobi c waste water T reatm ent ..................................................................................................... 5 Figure 2.2: COD bal anc e of an anaer obi c react or ................................ ............................. . 9 Figure 2.3: R elative load ing capacit y of dif fren et AnW T systems ......................... . 12 Figure 2.4: Anae robic contac t proces s, equi pped with flocculat or or degasi fi er unit ......... 14 Figure 2.5: S che mat ic repres ent ati on of UAS react or .............. ........... ................................ 1 6 Figure 2.6: EGSB and IC reactor of the major an aerobic s ystem ................. .......... .. 2 0 Figure 2.7: P roposed rea cti on scheme for ana ero b ic digesti on of domesti c s ewa ge s ludge based on Gujer and Zehnder .... ................................................ . .. 29 Figure 2.8: Matrix of stoichiometric coef ficients (v j , i ), yields (Y i ) and cons ervati ves f or ana erobic di gesti on model of Siegrist .. .................................. .... ...... 30 Figure 2.9: COD flux for a particul ate compos it e material........................... ...... .... . 3 2 Figure 2.10: The anaerobi c model as implemented including biochemi c a l p rocesses.... ....... ...................................................... .............................. 33 Figure 2.11: Bioch emi ca l rate coe fficients (V i , j ) and kinetic rate equ ati ons (p j ) for solub le components of ADM1. i=1 - 12, j=1 - 19 ............ ........... ..... ....... 3 4 Figure 2.12: Bioch emi ca l rate coe fficients (V i , j ) and kinetic rate equ ati ons (p j ) for particulate components of ADM1. i =13 - 24, j=1 - 19 ................ ............ 3 5 Figure 2.13: Matrix for gas tr ansfe r . ..................... .................................................... . 3 6 Figure 2.14: Ana erobic digesti on proc esses sch e me of Gujer and Zehnde r (1983) .. 3 8 Figure 2.15: Ana erobic digesti on proc esses sc h e me of Universit y of Cape Town Anaerobi c Digesti on Model No 1 (UCTADM1).. .. ................ ..............3 8 Figure 2.16: Gen eric ad vanta ge and disadv anta ges of conventi onal and non - conventi onal waste water tr eatm ent technolo gies.............................. ... 4 4 Figure 3.1: The pro cesse s scheme of the develope d model...................... ................ 5 2 Figure 3.2: Bio gas stripp ing proc esses and bio gas flow calculat ed with a p ressure control loop......................................................... .................... 5 3 Figure 4.1: The flow di a gr am of the wastewate r tr eatm ent plant of Efes Pil sen.. .... 59 Figure 4.2: The EGS B reactor of the plant ........... ....................... ............................. 6 2 Figure 4.3: The inl et flo w of the pla nt ................ . . ......................... ........................... 6 6 Figure 4.4: TSS influent and TSS effluent values of the EGSB rea ctor........ ........... 6 6 Figure 4.5: The COD inl et values of of the EGSB reacto r ....................... .... ..... ..... 67 Figure 4.6 : Th e COD ou tl et values of the EGSB reactor............. ......... ......... ........... 6 7 Figure 4.7: Th e OLR val ues of the EGSB reactor.. ..................... ............................. 68 Figure 4.8: The pH values of the EGSB reactor . ... ........................ ........................... 68 Figure 5.1: Dynami c si mul ati on result s of the TSS concentrati ons......................... 7 3 Figure 5.2 : D yn ami c si mul ati on result s of the pH values .............. ................ .......... 7 4 Figure 5.3: Dynami c si mul ati on result s of the ef fluent COD..... . ......................... ... 7 5 Figure 5.4: D ynami c si mul ati on result s of theVFA..................... .............. .............. 7 6 Figure 5.5: D ynami c si mul ati on result s of the Bi ogas............. .... ................ ............. 77 Figure 5.6: T he Biom ass compos it ion of the syste m..... ...................... ..................... 78 Figure 5.7: The OLR an d F/M rate of the plant.... ..................... ............................... 79 x vi x vii LIST OF SYMBOLS YOHO : Ordinar y h eterotrophic organism s yield co effi cie nt YAHO : Aceto gens yi eld coef fic ient YMHO : Acetocl asti c methano ge n organism s yield co effi c ient YMAO : H ydro genotrophi c methanogen or ganism s yield coefficient —OHO : Max im um growth rate of ordinar y het erotrophic organism s —AHO : Max im um growth rate of propionate de gradin g organism s —MHO : Max im um growth rate of aceto clastic methano gen or ganism s —MAO : Max im um growth rate of autot rophic methano gen esis or ganism s kd_OHO : E ndogenous de ca y co ef ficient of ordina r y hetero trophic organism s kd_AHO : Endogenous de ca y co ef ficient of propionate de gr ading or ganism s kd_MHO : Endogenous de ca y co ef ficient of acetoclasti c me thanogen or ganism s kd_MAO : Endogenous de ca y co ef ficient of auto trophic met hanogen or g. kh : Max im um h ydrol ysis ra te KA : Half satur ati on const ant for ac etate de grad ati on KPR : Half satur ati on const ant for propionate de grad ati on KF : Half satur ati on const ant for ana erobic ferment ati on. KH2 : Half satur a ti on const ant for uptake of h ydro gen KAB : R ate coe fficient fo r the base to acid rea cti on fEX : Fraction of particulate inert COD generated in biomass decay KLa_CH4 : Mol eculer dif fusion coe fficient of methan e KLa_CO2 : Mol eculer dif fusion coe fficient of c arbondiox ide KLa_H2 : Mol eculer dif fusion coe fficient of h ydro gen He_CH4 : Henry’s law coefficient for methane He_CO2 : Henry’s law coefficient for karbondioxide He_H2 : Henry’s law coefficient for hydrogen MG_CH4 : Mol eculer wei ght of me thane MG_CO2 : Mol eculer wei ght of car bondiox ide MG_H2 : Mol eculer wei ght of h ydrogen pH OHO HH : High pH inhibi ti on level for X OHO pH OHO LL : Lo w pH inhibi ti on level for X OHO pH AHO HH : High pH inhibi ti on level for X A HO pH AHO LL : Lo w pH inhibi ti on level for X A HO pH MHO HH : High pH inhibi ti on level for X M HO pH MHO LL : Lo w pH inhibi ti on level for X M HO pH MAO HH : High pH inhibi ti on level for X M A O pH MAO LL : Low pH inhibi ti on level for X M A O XS : S lowl y biode gr eadabl e COD XSBİN : Influent slowl y biode g r eadable COD XI : P articulate In ert COD XP : P articulate inert microb ial products SF : Fe rmentable COD conc entrati on SF_IN : Influent fermentable C OD conc entrati on SAC : Acetat e conc entrati on SHAC : Aceti c acid con centrati on SPR : P ropionate conc entrati o n SHPR : P ropionic acid con centr ati on SH2 : H ydro gen gas conc entr ati on SCH4 : Methane gas conc entrat ion x viii SIC : Inor ganic ca rbon conc e ntrati on SN : Nitrogen con centrati on S_AN : Anion conce ntr at ion S_CAT : C ati on concentr ati on SI : S olubl e In ert COD XOHO : Ord inar y heterotrophic biom ass concentrati on XAHO : Aceto gen esis or ganism s concentr ati on XMHO : Acetocl asti c methano ge n organism s concentr ati on XMAO : H ydro genotrophi c methanogen or ganism s Qin : Influent flow rate QIR : Internal rec yc le flow rat e VAN : Anae robi c re actor volu me iN_BM : Nitrogen fracti on of bio mass iN_XP : Nitrogen fracti on of X P iN_XS : Nitrogen fracti on of X S ICAC : Inor ganic ca rbon conte nt of acetat e ICBM : Inor ganic ca rbon conte nt of biom ass ICCH4 : Inor ganic ca rbon conte nt of methane ICCO2 : Inor ganic ca rbon conte nt of karbondiox ide ICPR : Inor ganic ca rbon conte nt of propionate ICSF : Inor ganic ca rbon conte nt of fermentabl e COD ICXP : Inor ganic ca rbon conte nt of x ix AN ANAEROBIC TREATMENT MODEL APPROACH FOR INDUSTRIAL WASTEWATER TREATMENT: CASE STUDY FOR FULL SCALE ANADOLU EFES WASTEWATER TREATMENT ANAEROBIC MODELING SUMMARY Ever yda y, a new wastew ater tre atm ent plant has been buil t and investm ents are mad e accordin g to regulations . But also, correct and ef ficientl y ope rati on of the treat ment plant is a growin g probl em with these investm e nts. Therefo re, some tool s such as models of the tre atement plants are need ed fo r th e supportin g and cont rollin g of the operati on condit ions. These models can formul at e operati onal and control strategi es f or the s ystem. Good stra tegies will reduc e operati ng costs , improve proces s stabil it y and enhan ce tre atm ent efficienc y. In thi s thesis, a new model was developed fo r ana erobic waste water tr eatm ent plants and sim ulated for brew e r y anae robic waste wate r treat ment plant under st ead y - state and d ynami c condit ions. In thi s model, a sin gle hydrol ysis pro cess and end product were includ ed. This en d product was chosen to be the ideal carbo h ydr ate as “glucose”. Based on subst rat e of glucose , th e model includes all bio chemi cal (h yd rol ysis , acido gen esis , aceto gen esis and me thanogenesis ) pro cess es, ph ysic o chemi cal (liqui d - li quid process and liquid - gas t ransfe r) proc esses and a process to describe the strippi ng of biogas components . Also , in thi s model but yrat e and but yri c acid are ne glected and we consi der ed acet ate and propionate as VFA concentrati on for brewer y wastew ater. The result s of mod el were evaluated with the ex ist operati n g co ndit ions of the pl ant and we could have main idea about the real op era ti on condit ions what must be . Also unique paramet er were found for brewer y wastew ater usin g EGSB rea ctor. After thi s work that we can sa y; this model is sui table for dir ectl y int e grati on to ASM modeling. It is not onl y t o use sludge stabil iz ati on also it can be used for w astewate r anaerobi c tre atm ent pro c esses. This stud y is suc ce sfull y implement ed in the d ynami c sim ulation for brewer y wastewate r. With this stud y i n the future studi es, the stoi chiom etr y and other COD fra cti ons (Bu , Va etc. ) can be ex tende d and it can be d eveloped for other industries’s anaero bic wastewate r tre atm ent pla nt xx xx i ENDÜSTRİYEL ATIKSULAR İÇİN ANAEROBİK ARITMA MODEL YAKLAŞIMI: ÖRNEK ÇALIŞMA ANADOLU EFES ATIKSU ARITMA TESİSİNİN MODELLEMESİ ÖZET Her geçen gün, yönetmeliklere bağlı olarak yeni bir atıksu arıtma tesisi yatırımı yapılmaktadır. Ayrıca, bu tesislerin doğru ve verimli işletilmesi de bu yatırımlarla giderek büyüyen bir so r undur . Bu ned enle , işletme koşullarına yardımcı olmak ve kontrol etmek için arıtma tesisi n i n modelleri gibi yardımcı araçlara ihtiyaç vardır Anaerobik ar ıtma prosesi, yüzyılı aşkın bir süredir yük sek kirlil i ğe sahip evsel ve endüstriyel atıksuların arıtılmasında kullanılan bir prosestir. Bu proses, organik maddeleri net bir enerji kaynağı olabilecek metan gazına dönüştürür. Anaerobik arıtma prosesinin d inamik matematik modeller i, bu sis teml er içindeki karmaşık ekosi stem anlayışını geliştirmek ve giriş atıksu ile çalışma koşullarında ki değişikliklere sistemin cevabını tahmin etm ek için geliştirilmiştir . Bu modeller, sistem için işletmeyi ve kontrol stratejilerini formüle edebilir. İyi stratejiler, işletme maliyetlerini azaltacak, proses kararlılığını geliştirecek ve arıtma verimliliğini arttıracaktır. Bir modelin diğer potansiyel kullanım yerleri, yeni bir reaktör tasarımının değerlendirilmesinde, kötü çalışma sistemlerinin tanımlanmasında, performan s ve işletme maliyetleri tahminlerinde ve yüklerin artması ile prosesdeki değişikliklerin karar verilmesi ve kontrol edilmesinde kullanılabilir. Öncelikle her bir hücreyi, her su molekülünü ve prosesin her detayını açıklayan bir model asla geliştirilemez. Modeller, gerçek durumu anlamak ve on unla ilgil i çalışmak için gerçeği tanımlayacak şekilde basitleştirilmesi için kullanılır . Anaerobik arıtmanın veya anaerobik çürütmenin optimizasyonu ve işletmenin, beslemenin değişmesi ve koşullarının değişen bir fonk si yon u olarak değerlendirilmesi önemli hedeflerdir ve bunlar uygun modeller kullanılarak takip edileblir. Bu modeller, kararlı durum modelleri olabilir. (i) İstenen bir sistem performansı için, alıkonma süresi, reaktör hacmi, gaz üretimi ve kompozisyonu tahmin etmek (ii) Çeşitli parametreler için sistem performansının duyarlılığını araştırmak (iii) Simülasyon sonuçları ve tesis performanının çapraz kontolünü sağlamak ve (iv) Çürütme prosesinin, yukarı veya aşağı akışlı AAT’nin işletilmesinde dizaynı nasıl etkilediğini belirlemekde yardımcı olmaktadır. Sistemin, zaman bazlı olarak, atıksu debisi veya kompozisyonu, sıcaklık, inhibisyon, pH, v.b işletme parametrelerindeki ani veya sürekli olabilecek değişikliklere karşı xx ii nasıl bir tepki gösterdiğini tahmin etmek için daha karmaşı k dinamik modeller, tesis mo dell emesine ente gr e edil ebil ir. Karbonhidrat bazlı atıksuların anaerobik arıtılması için dinamik modeller 1980 li yıllardan itibaren gelişmiş ve bu modelleme çalışmaları Uluslararası Su Birliği (IWA) tarafından Anaerobik Çürütme Modeli (ADM No 1)’nin yayınlanması ile sonuçlanmıştır. Bu tez de, anaerobik atıksu arıtma tesisleri için yeni bir model geliştirilmiş ve tam ölçekli bira endüstrisi anaerobik atıksu arıtma tesisi için uygulanmıştır. Çalışmanın amacı ilk önce anae rob ik bir model oluşturmak, modeli kararlı durum da kali bre etmek ve tam ölçekli sistemde uygulamak için pro ses i ve şartları araştırmaktır. Bu model oluşturulurken, literatürde daha önce oluşturulmuş ve çalışılmış olan modeller araştırılmış, incelenmiş ve bu anaerobik çürütme modellerindeki prosesler dikkate alınarak anaerobik arıtma ve bira endüstrisi için en uygun hale indirgenerek hazırlanmıştır. Bu modelde, ADM1 deki gibi , a tıksu içerisindeki yağlar, karbonhidratlar ve proteinler gibi büyük organik moleküllerin a yrı ayrı ölçümünün çok mümkün olmaması ve zor olmasından ötürü atıksu içerisindeki organik kirliliğin karbonhidrat bazlı olduğu kabul edilmiştir. Böylece bu organik maddelerin ayrı ayrı hidroliz i sonucu oluşak üç ayrı hidroliz ürününe artık ihtiyaç duyulmamıştır. Böylece karbonhidrat bazlı atıksuyun tek bir hidroliz prosesi ve son ürünü kabul edilmiştir. S on ürün olarak, ideal bir karbonhidrat olan “glukoz” kabul edilmiştir. Ayrıca, sistemde glukoz bozulma şartlarında bile birikmez. Gujer and Zehnde r an aer obik reaksiyon şemasında, sabit bir miktardaki hidroliz son ürünü ara ürün kısa zincirli yağ asitlerine (propiyonat, bütirat v.b) ve kalanıda direk olarak asetata dönüşür. Aynı zamanda, Ian R Ramsa y in çalışmasına göre, bira endüstrisi atıksuyunda, belli b ir zamanda a setik ve propiyonik asitin karşılaştırılması iyi bir şekilde tahmin edilebilir ancak bütrik asitin tahmin edil mesi i y i değildir. Bu nedenle bu modelde, bira atıksuyu için, bütirat ve bütrik asit ihmal edilmiş ve UYA olarak as etat ve propi yon at dikkate alınmıştır. Hidroliz adımından sonra glukoz asetat ve propiyanata dönüşür. Bu adıma asidojenik porosesi denir. Üçüncü adımda, propiyonat asetat ve hidrojene dönüşür. Daha sonra bu asetat karbondioksit ve metana dönüşür. Bu biyogazın üretilmeye başladığı anlamın gelmektedir. Son adımda ise, sistemde üretilen karbondioksit ve hidrojen metana dönüşür. Tüm bu reaksiyonlar biyokimyasal reaksiyonlardır (hidroliz, asidojenik, acetat oluşumu ve methan oluşumu prosesleri) ve model bu reaksiyonları içermektedir. Aynı zamanda, fiziko kimyasal (sıvı - sıvı geçiş prosesi ve sıvı - gaz transferi) proseslerini ve biyogaz bileşenlerinin sıyrılması diye tanımlanan biyogaz oluşumunu gösteren prosesi de içermektedir. Prosesler hidroliz ya da organizma gruplarının büyüme pr o sesleri olar ak formü le edil di . Tüm organizma gruplarının içsel solunuma tabi olduğu kabul edilerek, her bir grup için içsel solunum kütle kaybıda modelde gösterilmiştir. Proseslerin reaksiyon kinetiği, hidroliz için birinci derece kinetik reaksiyonu, diğ er kalan dört mikroorganizma büyüme prosesi için ise Monod tip kinetik reaksiyonu kabul edilmiştir. xx iii Bira endüstrisi, yüksek miktarlarda organik kirletici içeren büyük miktarlarda atıksu oluşturmktadır. Bu atıksuların en az maliyetle ve yöentmeliklere uygu n olraka arıtılması gerekmektedir. Her litre bira için yaklaşık 4 - 10 litre su, mayalama, yıkama ve soğutma proseslerinde kullanılmakta ve 3 - 9 litre atıksu oluşmaktadır. Genellikle bira atıksuyu, yüksek toplam biyokimyasal oksijen ihtiyacı (TBOİ) ve topl am kimyasal oksijen ihtiyacı (TKOİ) ve toplam askıda katı madde olarak karakte riz e edil ebil ir. Bira atıksuyunun toplam KOİ içeriği biyolojik olarak bozunabilen ve bozunamayan olarak ayrılmaktadır. Atıksuyun, biyoljik olarak bozunabilen kısmı yüksek olup topl a m KOİ nin %96 ‘sıdır. Atıksu arıtma tesisinin modellemesi çalışması kapsamında, tam ölçekli tesis olarak , Anadolu Efes Lüleburgaz, biyolojik atıksu arıtma tesisi seçilmiştir. Anadolu Efes 79 % ‘luk pazar payı ile Türkiye’de sektöründe lider konumundadır. An adolu Efes biyolojik atıksu arıtma tesisi aerobik ve anaerobik arıtma proseslerini içermektedir. Tesisde, anaerobik arıtma için “G enleşmeli G ranül Ç amur ” ( EGS B ) tip anaerobik reaktör kullanılmaktadır. Bu reaktör 337 m 3 ’lük bir kapasiteye s a hipt ir. Böylece;  İlk olarak tesisin 2010 yılı için atıksu karakteristiği analiz sonuçları kullanılmıştır.  Literatürdeki modeller kullanılarak ve uygulanabilir halde basitleştirilerek bira endüstrisi atıksuyu anaerobik arıtma prosesi için yeni bir model geliştirilmiş.  Dah a sonra, tesis geliştirilen model ile kararlı durumda (yıllık ortalama atıksu sonuçlarına göre) modellenmiştir. Bu kararlı durum modellemesi sonuçlarına göre, bira endüstrisi için kinetik ve stokiyometrik değerler belirlenmiştir. Daha sonra elde edilen değerler ile, tesis bir yıllık süre için günlük veriler ile simule edilerek sonuçlar elde edilmiştir . Model sonuçları, tesisin mevcut işletme şartları ile karşılaştırılmış ve gerçek işletme koşullarının ne olması ile ilgili ana bir fikir elde edilmiştir. . Bu çalışma sonucunda, modelin tam ölçekli bir tesisde dinamik simülasyonun başarılı bir şekilde yapıldığını söyleyebiliriz. Genleşmeli granül çamur reaktörü kullanılarak, bira endüstrisi atıksuyuna özgün parametreler bulunmuştur ve bunl ar daha sonraki çalışmalar için kullanılabilir. Bu sonuçlar tesisin ayrıca enerji verimliliği ve işletme koşullarının optimizasyonu içinde kullanılabilir. Yağlar, karbonhidratlar, proteinler gibi detaylı bir atıksu karakterizasyonuna gerek duyulmamıştır. Bu çalışmadan sonra, bu modelin ASM modellerine direk olarak ente gr as yonun uygun olduğu söyleyebiliriz. Ayrıca bu model sadece çamur stabilizasyonunda kullanmak için değil, atıksu anaerobik arıtma proses leri içinde kullanılabilir. Bu çalışma ile gelecek çalışmalarda, stoki yometr isi ve diğer KOİ fraksiyonları (Bu, Va v.b) genişletilebilir ve diğer endüstrilerin anaerobik arıtma tesisleri için geliştirilebilir. 1 1. INTRODUCTION 1.1 Aim of Study Anaerobic proc esses hav e been used fo r the tr eat ment of conc entrated muni cipal and indus trial wastewate rs for well over a centur y. These pro cesses conve rt or ganic materials int o methan e, a fu el that can yi eld a net ene r g y gain fro m process operati ons. Becaus e of recent adv anc es in treat ment technolog y and kn owledge of process microbiolo g y, appli ca ti ons are now ex tensive for treatm ent of dil ute indus trial wastewate rs as well . Dyn ami c mathematical models of the anaerobic treatm ent process have been developed to improv e un derstandin g of th e compl ex ecos ystem withi n these s yst ems and to predict the res ponse of the system to chan ges in influent and operati n g condit ions. Mathematica l models can serve as a tool to formul ate opera ti onal and control strate gies for t he s ystem. Good strate gies will reduc e oper ati ng costs , improve process stabil it y, and enhanc e tre atm en t efficienc y and throu gh put. Other potential uses of a model include assessment of new reactor desi gns, di agnosi s of poorl y pe rformi n g s yste ms, performanc e and op erati n g cost predictions for loading increas es from pro cessi ng plant up gr ades, and as soft s ensors in decisi on support s ystems for plant operati on. Dynami c mathem ati cal models for anae robic treatm ent of carboh yd rate - based wastewate r have been dev eloped since the earl y 1980s and these modelin g efforts culm inated with the pu bli cati on of An ae robic Digesti on Model No. 1 (ADM No. 1) by the Intern ati onal Water Association (Batst one et al. 2002). In thi s stud y a model was developed for an indus trial wastewate r and it is appli ed to a full - scal e plant tre ati ng br ewe r y wastewate r. The objecti ves of the st ud y were to first to buil d an ana ero bic model, investi gat e t he proc ess and require ments for cali brati n g the model at stead y state and to appl y t o a full - scale s ystem. 2 1.2 Scope of the Study Ever yda y , a new wastew ater tre atm ent plant has been buil t and investm ents are mad e accordin g to regulations . But also, correct and ef ficientl y ope rati on of the treatm ent plant is a growin g probl em with these investm en ts. To provide the disca r ge st andart is not a big problem actual problem is to provide these disc h ar g e stan darts with corre ct and effi cientl y operati on with less en e rg y consum pti on. Ther e fore, some tool s are need ed for th e supporting and controll ing of the oper ati on condit ions. The scope of thi s thesis, developi ng a new m odel and to sim ulate the brewe r y ana erobi c wastewat er tr eatm ent plant und er ste ad y - state and d yn ami c condit ions . The result s are ev aluate d with the ex ist operati ng condit ions and unique param eter subsets were found fo r brewe r y wastew ater usin g EGSB rea ctor. To use t his newl y buil t model wi ll enable to check the dyn ami c loadings on the overall s ystems performan ce . 3 2. LITERATURE 2.1 Introduction The opti mi sati on of the anae robic treatm ent or AD and the assess ment of its operati on as a fun cti on of var yin g feed or operati n g condit ions are importa nt objecti ves and can be pu rsued b y usin g approp ria te digesti on models. The se models can be of st ead y - stat e mode l (v) to esti mate retention time, rea ctor volum e, gas producti on and compos it ion for a requested s ystem pe rformanc e, (vi) to investi gate the s ensit ivi t y of the s ystem pe rform ance to va rious param ete rs, (vii) to provide cross - checkin g of simul ati on result s an d plant performan ce, and (viii) to determi ne how the digesti on proc ess can af fe ct the design of upst rea m or downstream WW TP operati ons. More compl ex dynami c mode ls could be in tegr ated in plantwide modelli ng, predicting on a time basis how the s ystem wil l react to sudden or progressi ve chan ges in oper ati ng paramete rs of feedstock flow rate and compos it ion, temperatur e, inhi bit ion, pH, etc. The number of models pres ented in liter ature is ex tensive, and often of ver y sp ecific nature. The most freque ntl y us ed m odel, ADM1 developed by the IW A , forms a good basis and is often used in expanded models, as proposed by, e.g. Sötemann et a l . . Sim pler models for digesti on h av e be en pr oposed b y, e. g. Bal a, Siegrist and others. 2.2 The Anaerobic Treatment Process The fermentation proc e ss in which or ganic material is degr aded and biogas (compos ed of mainl y m ethane and ca rbon diox ide) is produc ed, is refer red to as anaerobi c di gesti on. An aerobic digesti on proc e sses occu r in man y pla ces wh ere organic mate rial is avail a ble and redox potential is low (zero ox ygen). Anaerobic tr eatm ent itself is ver y ef fe cti ve in removi ng biode grad abl e organic compounds leavin g minerali sed compounds like NH 4 + , PO 4 3 - , S 2 - in the solut ion. Anaerobic tre atm ent can be conducted in te cnica ll y plain s yst ems, and the pro cess 4 can be appli ed at an y scale and at alm ost an y pla ce. Mor eove r the amount of ex cess sludge produc ed is ver y small and well stabil ised, even hav in g a ma rket value wh en the so - call ed granular anaerobic slud ge is prod uced in the biore actor. Moreover, useful ener g y in the fo rm of biogas is produ c ed inst ead of hi gh - gr ad e ener g y consum ed. Accepti n g t hat ana erobic digesti on in fact mer el y removes or gani c poll utants, there are virtuall y few if an y serious dr awbacks left, even not wit h respect to the rate of st art - up of the system. Fi gure 3.1 s hows the fate of carbon and ene r g y in both aerobic and an aerobic wastew ater tr ea tm ent (AnW T) assum ing that the ox i dati on of 1 kg COD requires 1 kWh of aer ati on ener g y. In contr ast to anaerobi c treatm ent, aerobi c tratm ent is gener all y ch ar act erised by hi gh oper ati onal costs (ener g y) , while a ver y large fra cti on of the waste is converted to another t ype of waste (sludg e). Aerobi c treatm ent in a conventi onal acti vated sludge pro cess yields about 50% (or more) ne w sludge from the COD converted, which requir es further treatm ent, e. g. ana erobic digesti on, befor e it is reused, dispo sed off or incinerated . The carbon/en er g y f low principles of aerobic and anaerobic bioconve rs ion largel y affe ct the set up of the correspondi n g wast e water tre atm ent tecnolog y. Man y different t ypes of or gani call y poll uted wastewate rs, even those that were previous l y beli ved not to be suit able for An W T, ar e no w treated b y ana erobic high - r ate conversion pro cesses. 5 Figure 2.1: Fate of carbo n and ener g y in aerobi c (above) and ana erobic (b e low) wast ewat er treat ment (Bi ologi ca l ww treat ment , IWA Publ ishing) . Anal ysin g the reasons wh y the sele cti on for AnW T was made, the followi ng striki ng advanta ges o f AnW T over conventi onal aerobi c treatm ent s ystem can be given;  R educti on of ex cess sludge producti on up to 90%.  Up to 90% reducti on in space requi rement when using ex panded sl udge bed s ystems.  High appli cable COD loading rates rea chin g 20 - 3 5 kg COD per m 3 of rea c tor per da y, requirin g small e r rea ctor volum es.  No use of fossil fuels for tre atm ent, savin g abo ut 1kW h/kgC OD remov ed, dependin g on aerati on ef ficienc y.  P roducti on of a bout 13. 5 MJ CH 4 energ y/k gC O D removed, givi ng 1.5 kW h electricit y (assum in g 40 % electri c conve rsion eff icienc y)  R apid start up (< 1 week ), using gr anular anae rob ic sludge as seed mat erial .  No or ver y litt le use of chemi cals.  P lain technolog y with hi gh tre a tm ent efficienci es.  Anaerobic sludg e can be stored un fed, rea ctors can be ope rated dur ing agricultural campai gns o nl y (e. g. 4 mont hs per ye ar in the su gar industr y).  Ex cess sludge has a mark et value. 6 High rate s ystems facil it ate water rec yc li ng in fact orie s ( towards closed lo ops). 2.2.1 Anaerobic degradation of organic polymers The ana erobic de grad ati on pathwa y of or ganic ma tt er is a multi step process of series and parall el rea cti ons. This process of or ganic matter de gr adati on proc eeds in four successi ve sta ges , namel y;  H yd rol ysis  Acidogenesis  Aceto gen esis  Methano genesis The anae robic ecos ys tem is the result of compl ex interacti ons among microor ganism s of sever al diffe rent spe cies. Th e major groupin gs of bacteria and reacti on the y medi ate ar e ;  Ferment ati ve bac t eria  H yd ro gen - p roducin g ace togenic bacte ria  H yd ro gen - consum ing ac etogenic ba cteria  C arbon diox ide - reducin g methanogens  Aceti clastic methano gen s The digesti on process ma y be subdi vided into the foll owing four ph ases. 1. H yd rol ysis , wh ere enz ymes ex creted b y f erment ati ve bacte ria (so - called ‘exo - enzymes’ ) convert complex, undissolved material into less complex, diss olved compounds which can pass through th e cell wall s and membra nes of the ferm entative bacte ria. 2. Acidogenesis , wh ere t he diss olved compou nds prese nt in cell s of fermentative bact eria ar e conve rted int o a nu mber of sim ple compo unds which are then ex creted. The compounds produced during thi s phase include volatil e fatt y acids (VF As), alcohols, lacti c acid , CO 2 , H 2 , NH 3 and H 2 S, as well as new cell mat erial. 3. Aceto gen esis (intermedi ar y acid producti on) where di gesti on produ cts are converted int o acetate, hyd ro gen (H 2 ) and CO 2 , as well as new cell materi al. 7 4. Methano genesis , whe re acetate, h ydro gen plus carbonate, fo rmate or methanol are converte d int o meth ane, CO 2 and new cell material. 2.2.1.1 Acidogenesis During the acido gen esis step, the hydrol ysis pro ducts (ami no acids, sim ple suga rs, LCF As), which are rel ati vel y small solub le compounds , are diffused insi de the bacterial cell s through the cell membrane and subs equentl y fermented or anaerobi call y ox idi z ed. Acidogenesis is a ver y comm on reacti on and is performed b y a large group of hydrol yt ic and nonh yd rol yti c microor ganism s. About 1% of all know bacteria are (facult ati ve) ferment ers. The acidi ficati on products co nsis t of a variet y of small organic compounds , mainl y VFAs, i.e. acetate and hige r organic acids such as propionate and but yrate, as well as H 2 , C O 2 , some lacti c acids, ethanol and amm onia. 2.2.1.2 Acedogenesis The short chain fatt y ac ids (SC FA), other than acetat e, which ar e produ ced in the acido genesis st ep are fur ther conv erted to acetate , hyd ro gen gas and carb on diox ide by the acetogenic bacter ia. The most important aceto genic subsr ates are propionate and but yr ate, ke y - int erm ediates in the ana erobic digesti on p roc ess. But also lactate, ethanol, methanol and even H 2 and CO 2 are (homo) aceto genic all y co nverted to acetat e. 2.2.1.3 Methanogenesis Methano genic bact eria accompl ish the final stage in the overall anaerobic conversion of or gani c matte r to met hane and carbon diox ide. Durin g thi s fou rth and l ast sta ge of anaerobi c de gr adati on of or ganic matte r, a gro up of methano genic ar chea both reduce the carbon dioxi de using hydro gen as electron donor and dec arbox yl at e acetat e to from CH 4 . It is onl y i n thi s sta ge when the influ ent COD is conv erted to a gas eous fo rm that autom ati call y leav es the reacto r s ystem. Meth ano gens are obli gate anaerob es, with a ver y narro w subst rat e spe ctrum. Some can onl y use ce rtain determi ned subst rates su ch as acet ate, meth yl ami nes, methanol, formate and H 2 /C O 2 or CO. For en ginee rin g purposes, methano gens are classified int o two ma jor groups: the acetate convertin g or aceti clastic methano gens and the hydro gen uti li sing or hydro genotrophic metha nogens. Gener all y, abo ut 70% of the produced methane origi nates from acet ate as the main precursor. The rest mainl y origin ates from H 2 and 8 C O 2 . The growth rat e of the aceti clastic m ethanogens is ver y low, result ing in doubli ng times of sever al da ys or even more. Th e ex tremel y low growth rat es ex plain wh y anae robi c reactors require a ver y lon g s tart - up time with unadapted seed material and wh y hi gh sl udge conc entrati ons are pursued. H ydro genotrop hic bacte ria have a much high er max im um growth rate th an the ac etoclasti c bact eria with doubli ng times of 4 to 12 hours . 2.2.2 Working with COD balance Like an y biol o gical s yst em an an ae robic tre atm ent proc ess must be mo nit ored for relevant par ameters, and measurements must be evaluated for ad equate op erati on and control. The reason fo r t his is that in contr ast to aerobic s ystems th ere is no COD destruction in an anaerobic reactor. During anaerobic treatment the COD is only ‘re - arranged’. Complex organic compounds are broken down in more simple int ermediates and eventu all y miner ali sed to CH4 and CO2. All COD that entered the s ystem ends up in th e en d - product CH4, minus t he COD that is incorpor ated in th e new bact erial mass. Sinc e a per fe ct mass balanc e can be mad e b y onl y using the COD as a paramete r, the COD is therefore ge nerall y t aken as a contr ol tool to operate an ana erobic s yst em: COD i n =C OD ou t For practi cal purposes thi s equati on shoul d be ex panded to the various outl ets of th e an a erobic rea ctor as depict e d in Fi gur e 2.2. For iden ti f yin g the fate of COD in an ana ero bic rea ctor detailed anal yses of the gaseous, liqui d and solid o utl ets should be perfor med (Table 16.8). B y diff erenti ati n g the COD fr acti ons of gas, liquid and soli ds, the missi ng paramete rs can be esti mated from the more easil y measu ra ble parameters . Based on the basic influent char acte risti cs, i.e. flow rate and COD concentrati ons, and in for mation on the biodegradabil it y of the COD, the ex pected CH4 producti o n rate ca n be easil y esti mated f rom equati on; CH 4 + 2 O 2 →C O 2 + 2H 2 O which means that 22 . 4 m 3 CH4 (STP ) requires 2 mol es of O 2 (COD), which equals 64 kg COD. Th ere fore, t heoreti call y, 1 kg COD can be conv erted in 0.35 m 3 CH4. 9 Figure 2.2: COD balan c e of an an ae robic re actor. Simil arl y, th e theor eti cal COD equ ivalent fo r 1 kg ‘bacterial VSS’, with an estimated compos it ion of C 5 H 7 O 2 N, can be calculated as 1.4 2 kgCOD/k gVSS . Having both the final pr oducts CH4 and newl y gro wn bacteri a ex presse d as COD, the balanc e can be made if influent and ef fluent are properl y m e asured. Often ‘gaps’ in the COD balance occur which can be attributed mostly to the ‘loss of electrons’ when these are ch annell ed to ox idised anions like SO4 2 and NO 3 - , Therefo re, in this cas e, fo r closi ng the CO D balanc e eit her all reduced gases shoul d b e taken int o account or the concentrati on of electron acc eptors nee ds to be measured. It shou ld be reali sed that soluble COD containing gases like H 2 S , will be present in the ef fluent. In thi s ex ampl e, or ganic COD is converted int o inorgani c COD of which a pH dependent fra cti on will end in the biogas while the remainde r will sta y in the ef fluent. Another frequ entl y cit ed cause for a COD gap is the entrapment or accu mul ati on of COD in the sludge bed, someti mes drasti call y changin g the stochiom etri c value of 1.42 kgCOD/k gVSS . The latter is particularl y tr ue during the tr eatm ent of fat or LCF A - cont aini ng waste water. Wit h these subst rates, COD removal effic iencies ar e gen erall y ve r y hi gh, but low CH 4 producti on rates lead to huge gaps in th e balanc e. In thi s ex ampl e, th e COD gap indi c ates sever e long term ope rati onal pro blems. The accumul ati n g solids will deteriorat e the SMA of t he sludge , finall y result in g in a compl ete fail in g of the anaerobic process. 10 2.2.3 Anaerobic reactor system Anaerobic reactors ar e i n use sin ce the 19th ce ntur y, when Mou ras an d Cameron developed the autom ati c scaven ge r and the sept ic tank to reduce the amount s of soli ds in the sewe ra ge s ystem. Althou gh at a ver y poor rate, the first anae robic stabil isatio n processes occurr ed in the tanks that were desi gned fo r int ercepti ng the black - wat er soli ds. The first anae robic reactor was develop ed in 1905 when Karl Imhof f desi gned th e Imh off tank, in which soli ds sedim ents are stabil ised i n a singl e tank. The actu al controll ed digesti on of entrapp e d soli ds in a separate re actor was developed b y the Ruhrv e rband, Essen - R eli n ghaus en in German y. In the same de cad es, Bus well started to adopt the same technolo g y for t rea ti ng liquid wastes and indus trial wastewater. All these s yste ms can be chara cterised as low rate s ystems sinc e no sp eci al features wer e includ ed in the desi gn to au gment th e anaerobi c catabolic cap a cit y. Th e process feasibi li t y of these s yst ems was ver y much dependent on the growth rate of the an aerobic co nsortia. As a result , rea c tors wer e ver y bi g and ver y fra gil e in operati on. In the final decad es of the 19th centur y also some first trials of upw a rd flow fix ed film rea ct ors were perfo rmed, but it was too earl y to make thes e s yste ms successful McCart y ( 2001). Also the ana erobi c pond can be regar ded as a low loa ded ana erobic tr eatm ent s ystem. Ana erobic ponds are often const ructed in conjun cti on with facult ati ve an d maturati on ponds. The appli ed loading rate to ana erobic ponds ran ges betw een 0.025 - 0.5 kgCOD/m 3 .d, while using pond depths of 4 m. The big disadv anta ges of an aerobic ponds ar e probl e ms related to odour as these s yste ms easil y becom e overl oaded. Also the loss of ener g y rich CH4 to the atm osphere is a reco gnis ed disadvanta ge. 2.2.3.1 High rate anaerobic systems One of th e major su cces ses in the de v elopm ent of ana erobic wastew ater treatm ent was the int roducti on of high - r ate reactors in which biom ass retention and liquid retention are unc oupled. Contrar y to aerobic pr ocesses, in an ana erobic or anox ic (denit rificati on) pro cess, the max imum permiss ibl e load is not governe d by the max im um rate at which a necessar y reactant ca n be suppl ied (e.g. ox ygen durin g aerobic pro cesses), but by th e amount of viable anaerobi c biocatal ysts or the anaerobi c bacteri a whic h are in full contact with the wastewater cons t it uents. In anaerobi c high – rate s yst ems, high sludg e con ce ntrati ons are obtained by ph ysical retention and or immobi li sati on of anaerobic slu dge. Hi gh biom ass conc entrati ons 11 enable the appli cati on of high COD loading rate s, while maintaining long SRTs at relativel y short HRTs. Di ffer ent hi gh – rat e s ystems wer e dev eloped ov er th e last three decad es includin g the an aerobic conta ct pro cess (ACP ), anae robic filters, t he UAS B, FB and EGSB reacto rs and the baffled rea ctor s. To enable an an aero bic rea ctor s ystem t o accomm odate high or gani c loading rate s for treati n g a spe cific wastewate r, the foll owing condit ions shoul d be met:  High retention of viable sludg e in the reactor under op erati onal condit ions. The high er the amount of sludge retained, the higher will be t he loading potential of the syst em. Therefo re, it is necessar y to cult ivate a well sett leable or immobi liz ed biom ass, and that the slud ge will not deteriorate in thi s respect.  S ufficient contact betwe en viable bacte rial biom ass and waste wate r. In the case whe re pa rt of the sludge r etained in the reactor remains deprived of subst rate, thi s sludge is o f littl e if an y value.  High reacti on rates and absence of serious transpo rt l im it ati ons. It is clear that the kinetics of the degrad ati on processes are a fact o r of gr eat im portance. It is essential that metaboli c end products can easil y escape from the aggr e ga te. The siz e of the biofil ms shoul d remain rel ati vel y small and the accessi bil it y of the organism s ins ide the biofil m should be high.  The viable biom ass sh o uld be suffi cientl y ad apted and/or ac cli matiz ed . For an y waste wate r subj ect e d to treatm ent, the slud ge shoul d be enabl ed to adapt to the specific ch ara cteris ti cs of the conc ernin g wastewater.  P revalenc e of favourabl e environmental condit ions for all requi red organism s insi de the reactor und er all imposed operati onal condit ions, focusing on the rate limi ti ng steps. It sh ould be emphasiz ed here that thi s condit ion doesn't mean that the circumst a nces shoul d be sim il ar at an y loc ati on withi n the reacto r and at an y inst a nt. As a matter of fact even the contrar y is tr ue. Rega rdin g the fact that a lar ge variet y of dif fer en t organism s ar e invol ved in the degradati on of more compl ex compounds , the ex ist ence of micro - niches withi n the s ystem is an absol ute pre - requis it e. Onl y in thi s way can the required flou rishing gro wth of the requir ed ve r y dif fer ent or ganism s be achieved. It shoul d be no ti ced that particula rl y in the int erior of biofil ms and 12 gr anules, the conc entrati on of subst rates and metaboli tes are low enou gh to all ow even the ver y en der gonic ac etogenic rea cti ons to proceed, e.g. the ox idation of propionate at the ver y low h ydro gen concentr ati ons. The ana erobic contact process b y Sch roepf er et al. (1955) inde ed turned out to be reasonabl y suc cessful fo r the tre atm ent of high e r stren gth indus trial wastewaters. Wit h a few ex cepti ons, hardl y an y at that time would thi nk that anae robic treatm ent ever could become feasi ble for low strength wastewaters. Re gardin g the problems ex perienced with the vari ous versions of t he anae r obic contact pro cess, onl y ver y few even beli eved an aerobic treatm ent could become appli cable for tre ati ng medium strength waste wate r. However in the sixt ies and seventi es the sit uati on changed rapidl y, and in the nineti es the anae robic treatm e nt c oncept even was shown feasibl e for ver y low stren gth wastewate rs at low ambi ent temperatures. These unforese en developm ents can be att ributed to supe rior meth ods of slud ge retention, based on sludge immobi li z ati on. Figu re 2.3 illust rates the developm ent o f high rate reactor s ystems and the impact of improved sludg e ret enti on and enh anced contac t on the appli cable or ga nic loadin g rates. Whil e the first trials of Buswell did not rea ch loading rates of 1 kgCOD/m 3 .d, modern AnW T s ystems are sold on th e ma rk et with guar anteed loa ding rates ex ceeding 40 kgCO D/m 3 .d. Figure 2.3: Relative load ing capacit y of dif fer ent AnW T systems. Max im um appli ed loading rates und er full scale condit ions rea ch about 45 kgCOD/m 3 .d appl yin g en hanced contact in EGS B t ype s ystems . A t prese nt, most appli ca ti ons of AnW T can be found as end - of - the - pipe treatm ent technolog y for food pro cessi ng wastew aters and agro - indus trial wastewa ter. Table 13 2.1 lists the various indus trial sectors whe re the s urve yed 2,266 rea ctors ar e inst all ed. It sho uld be noti ced that the number of ana erobic appli cati ons in the non - f ood sector is rapidl y growin g. Co mm on ex ampl es are th e pape r mill s and the chemi cal wastewate rs, such as tho se containing formaldeh yd e, benz aldeh ydes, ter e phthalates, etc. Flores et al .( 2006). The latter is surprisi n g si nce it is particularl y di ffi cult for the chemi cal indus tries to enter with anae robic technolo g y, owing to the gener al prejudi ces against biol ogical treatm ent and ana ero bic treatm ent in particula r . Table 2.1: Applicati on of anaerobi c technolo g y to indus trial wastewate r. Indust rial Se ctor Tyep e of wastewat er Nr . of reactors % Agro food indus tr y Sugar, pot ato, starch, ye ast, pecti n, cit ric acid, canner y, confecti ona r y, fruit , ve getables, dair y, bake r y . 816 36 Beve ra ge Bee r, malti ng, so ft drink s, wine, fruit jui ces, coff ee . 657 29 Alcohol disti ll er y Can jui ce, cane mol ass es, beet mol asses, gr ape win e, gr ain, fruit . 227 10 Pulp and pap er indus tr y Rec ycl e paper , mechanical pul p, NSSC, sul phi t e pul p, st raw, bagas se . 249 11 Mis cell aneous Chemi cal, pharmac euti cal, sludge liquor, landfil l lea chate, acid mine water, m unicipal sewa ge. 317 14 Onl y ver y recentl y, hi gh - rate AnW T s ystems we re developed for tre ati ng cold and ver y low str en gth waste waters. In addit ion to muni cipal sewa ge , man y indus trial wastewate rs ar e dischar ged at low temperatures, e.g. beer and malte r y wastewaters. Full scale result s so far show that an y of the cit ed wastewat ers are an aerobic all y treated usin g comm on se ed materials, illust rati ng the robustness and fle x ibi li t y of the anaerobi c proc ess. 14 2.2.3.2 Single stage anaerobic reactors 2.2.3.2.1 The anaerobic contact process Empl oyi ng ext er nal sett ler s and sl udge ret ur n proce sses ar e known as the anaer obi c cont act process (ACP). Figure 2.4 . Figure 2.4: Ana erobic contact process, equipped with flocculator or a de gasifier unit to enhance sl udge s edim entati on in the seconda r y cla rifier . The va rious versions of t he first gener ati on of 'high rate an aerobic tre atm e nt s ystems for medium stren gth wastewaters wer e not ve r y success ful. In practi c e, the main difficult y appe ared to be the sepa rati on of the sl udge from th e tre ated water. Th ese difficulti es can be mainly due to the fact that a too int ensive agit ati on in the bio - reacto r was consi der ed necessa r y. The idea was t hat the more int ensiv e th e mix ing, the bett er would beco me the contact betwe e n sludge and wastewat er Various methods for sludge sep arati on have be en tested and/or empl o yed in th e differ ent versions of the ACP . These methods include vac uum degasifi cati on in conjunction with sedim entation, the addit ion of or ganic po l ymers and inor ganic fl occulants, centrifu gati on and ev en aerati on (in orde r to sto p digesti on). Howev er, t he result s were usu all y unsatisfact or y. At pres ent, with the cur rent knowled ge on anae ro bic digesti on technolo gies, a more gentl e and int erm it tent mode of mix ing is appli ed. Wit h such an approach, the sludge will acquire and keep ex cell ent sedim entation properties, and the an aerobic cont act proc ess can cert ainl y mak e a valuable contribut ion t o environmental protecti on and ener g y recove r y, partic ularl y with wastewate rs containin g high fracti ons of suspend e d solids and semi liquid wastes. If well design ed, modern ACP ma y re ach or ganic lo ading rates of 10 kgCOD /m 3.d. 2.2.3.2.2 Anaerobic filters (AF) The modern version of upflow anae robic filter (UAF) was developed in the USA by Young and McCa rt y (19 64, 1982) in the late sixties. The sludge retention of the UAF 15 is based on: the att achment of a biofil m to the soli d (stati onar y) carrie r material, t he sedim entation and entra pment of sludge particl es betwe en the int erstic es of the packin g mate rial, fo rma ti on of ver y well sett li ng slud ge aggre gat es. Initiall y, a suit able ca r rier mat erial for the s ystems was ha rd to find Youn g ( 1991 ). Various t ypes of s ynthetic pa cki ng hav e been inve sti gat ed and natural mate rial s such as gr ave l, cok e and bambo o segments as well . It turned out that the shape , siz e and weight of the packin g ma terial ar e important aspec ts. Problems with UAF s ys tems in particular generall y o ccu r durin g long - te rm operati on. The m ajor disadvanta ge of the UAF packing m aterial are important aspects. conc ept is the difficul t y of maintaining t he required contact betw een sludg e and wastew ater, be cause cloggin g of the 'bed ' ea sil y occurs. This pa rtic ularl y is th e case for pa rtl y solub le wast ewate rs. Thes e clo ggin g problems obvious l y can be overcome (at least partl y) b y appl yin g a prim ar y sett ler a nd/or a pre - ac idi ficati on step Se yf ried ( 19 88). H o wever, thi s would requir e the const ru cti on and operati on of addit ional unit s. Moreover, apart from the higher costs , it would not compl etel y eli mi nate the problem of short - circuit ing (cloggi ng of the bed) f lows, leadin g to disappoint ing treatm ent efficienci es. The ex perien ces with the s ystem ce rta inl y ar e rath er sati sfa ctor y, appl yi n g modest to relativel y hi gh loading rates up to 10 kgCOD/m 3 .d. The UA F s ys tem will remain att ract ive for tre atm ent of main l y soluble types of wastew ater, pa rticularl y when the proc ess of sludg e granul ati on will no t proceed s ati sfac tor y. On the other hand, lon g term problems related to s yste m cloggin g and the stabil i t y of filter mate rial ca used a decli ne in the number of inst all ed full scale AF s ystems. Various modes of operati on and filt er mater ial were investi gated but fu ll - scal e appli cati on is rather disappoint ing. The limi ti n g factor is the appli cable low organ ic loading rate owin g to the limi ted amount of biomass that can be reta ined in su ch a system as it is prim a ril y based on att a ch ment of biom ass to the surfac e of the packin g mate rial. 2.2.3.2.3 Anaerobic sludge bed reactors (ASBR) The an aerobic slud ge be d rea ctors (ASBR ) undo ubtedl y are b y far th e m ost popular AnW T s ystems so far. The slud ge retention i n such a rea ctor is bas ed on the formation of easil y set tl ing sludge aggr e gates (flocs or gr anules), and on the appli cati on of an int ernal gas - li quid - soli ds separ ati on s yst em (GLS S devic e). B y far 16 the best known ex ampl e of thi s concept is the upflow an aerobic slud ge bed reactor (UASB), which was dev eloped in the Neth erland s in the earl y s eventi es Lett in ga et al. ( 1976, 1980 ). In view of its prosp ects, and the fact that alm ost 90% of the newl y inst all ed high - r ate reacto rs are slud ge bed s ystem s. At the start of 2007, about 1,750 full - scale UAS B in stallati ons have been put int o o perati on. Most of these full scale reacto rs are us ed for tr eati ng agro - indus trial wastewate r, but its appli cati on fo r wastewate r from chemi c al indus tries and se wa ge is increasin g. Fi gure 2. 5 shows a schematic repres entation of a UASB reactor. Figure 2.5: S chematic re presentation of a UASB reactor . 1. S im il ar to the UAF s yste m the wastew ater mov es in an upwa rd mode throu gh the rea ctor. How eve r, contrar y to the AF s ys tem gener all y no packi ng material is present in therea ctor vessel. The sludge bed reactor con c ept is based on the following ideas: Anae robic sludge has or acquires good sedim entation properties, provid ed me chanic al mi x ing in the reactor remai ns gentl e and the proc ess is operated co rre ctl y. Fo r that reason, but also beca use it reduces the investm ent and maintenance costs , mechanical mix ing is not appli ed in UAS B reactor s. Becaus e of the ex cell ent sett li ng characte risti cs of the sludge, hi gh super fic ial liquid velociti es can be appli ed without an y risk of consi derabl e sludg e wash - out. 17 2. The requir ed goo d cont act betwe en the sludg e and wastew ater in UAS B - s ystems gen erall y is acco mpl ished (i) by feedin g t he wastew ater as unifor ml y as possi ble over the bott om of the reactor, or (ii) as a result of the agit ati on caused b y the producti on of biogas. 3. P articular l y with low s trength waste water, rea ctors with a high hei ght - diameter rati o are used reachin g hei ghts of 20 - 25 m. A low surface area will facil it ate the feedin g of the s ystem, whe reas the ac cumul ati ng bio ga s producti on over the height of the tow er reacto r will cause a turbulentfl ow. Also the incr eased upfl ow velocit y result s in a bett er contact between the sludge and the poll utants. Wit h wastewaters containing biodegrad a ble addit ionall y achiev ed by appl yin g a liquid recirc ulation flow. As a result , a more c ompl etel y mix ed flow patt ern is acquire d and strati ficati on of the subst rate and int erm ediate products ov er the height of the rea ctor is minim ised, thereb y mi nim isi ng potential inhi bit ion. 4. The washout of sludg e aggre gat es is prevented by sep arati n g the pro du ced biogas using a gas coll e cti on dome inst all ed at the top of the reactor. In thi s wa y a zone with relatively litt le turbulence is crea ted in the uppermost part of the reacto r, consequentl y the reactor is equippe d with an in - buil t secondar y clarifier. T he gas coll ecti on dome acts like a thr ee phase GLS S . The GLS S devi ce const it utes an essential pa rt of a UASB re actor and serv e s to:  C oll ect, separate and disc har ge the produced bio ga s.  R educe liquid turbulenc es in the sett ler compar tm ent for enhanc ement of sludge settl in g, result in g from the gas produ cti on.  R emove sludge particl e s by a mechanism of sedim entation, flocculation and/or entrapm ent in a sludge bl anket (if present i n the sett ler).  Lim it the ex pansion of the sludge bed in the di gest er compa rtment.  Aaccompl ish some poli shing of the waste wate r with respect to suspend ed matter. Some resea rche rs and pr acti ti oners su ggest repla cing th e GLS S device by a packed bed in the uppe r pa rt of the rea ctor. This so - call e d upflow h ybrid reacto r is a mer ge 18 between t he UASB and t he UAF rea ctors. Specifi c ch emi cal wastewat ers show bett er treatm ent effici encies fo r all compounds using hyb rid s ystems compared to UAS B reacto r. The most know n disadvanta ge of h yb rid rea ctors is the dete riora ti on of the filter secti on after pro longed periods of ope rati on. Hybrid reactors are also advanta geous for achie ving enh anc ed ef fluent poli shing as coll oidal matter is entrapped at the top part of the syst em . 2.2.3.2.4 Anaerobic expanded and fluidized bed systems (EGSB and FB) Ex panded be d and fluidi z ed bed s ystems ar e re garded as the second gen erati on of sludge bed rea ctors achieving ex treme or ganic l oading rates (ex ceedin g 30 to 40 kgCOD/m 3 .d). The FB process is based on the occurren ce of bacte rial att a chment to mobi le carri er particles, which consi st, for ex ampl e, of fine sand (0.1 - 0.3 mm), basalt , pumi ce, or plasti c . The FB s ystem can be regard ed as an adv anced ana erobi c technolog y Li and Sutt on ( 1981 ); Heij nen ( 1983, 1988), that ma y rea ch loading rat es of 50 - 60 kgCO D/m 3 .d. Modern FB s y stems like the Anaflux s ystem Holst et al. ( 1997), rel y on bed ex pansion rathe r than on bed fluidi z ati on. As bed ex pansion all ows a much wider di stribut ion of prevail ing biofil ms, the s ystem is much more eas y to oper ate. The An aflux reactor uses a tripl e p hase sep ar ator at the top of the reacto r, more or less sim il ar to the GLS S device in UAS B and EGSB rea ctors. When the biofil m la ye r att ach e d to the media becomes ex cessi vel y over develope d, and the concernin g (lighte r) aggr eg ates then tend to accu mul ate i n the separato r device , the material is periodicall y ex tracted from the rea cto r by an ex ternal pump in which it is subj ected to the appl icat ion of sufficient shea r to remove pa rt of the biof il m. Then both the media and detached biom ass ar e retu rned to the rea ctor, an d the free biom ass is then all owed to be rinsed out from the s ys tem. The s ystem is app li cable to wastewate rs wit h a suspe nded solids concentrati on <500 mg/l . The EGSB s ystem emp l o ys granular slud ge, which is ch ara cterised b y go od sett l ing chara cterist ics and a hi gh methanogenic acti vit y . H igh liquid velocit ies, togeth er with the lifting acti on of gas evolv ed in the bed, leads to a sli ght ex pansion of the sludge - b ed. And as a result of that, an ex cel l ent contact betw een sl udge and wastewate r pre v ail s in the system, le adin g to signi ficantl y hi ghe r loadin g potentials compared to conv enti onal UAS B inst all ati ons. The EGSB s ysems rel y on a compl ete rete nti on of the granula r sludg e. Ex cell ent r esult s have been obtained wi th modern 19 full - scale EG S B inst a ll at ions using various kinds of wastewat ers, rea ching or ganic loading rates of up to 40 - 45 kgCOD/m 3 .d. Int ere sti ngl y, b y appl yin g EG S B reactor s ystem sev eral othe r t yp es of waste wate rs can be treat ed which cannot be treat ed using conv enti onal UASB s ystems such as:  W astewaters containin g biodegr adable compound s.  C old (even < 10 o C ) and dil ute (COD << 1 g/l) wastewaters, i.e. wh en speci fic gas producti on is ver y l ow and biogas mix ing is absent Rebac et al. ( 19 98). EGSB reactors are charact erised b y an im prove d h ydrauli c mix ing, independent from the bio gas p roducti on .  W astewaters con tainin g long chain fatt y acids Rinz ema ( 1988) . At low upflow velociti es (UAS B), LC FAs tend to abs orb to the sludg e and form inaccessi ble fatt y clum ps. At hi gh upflow velociti es (EGS B) th e subst rate is int roduced at a low er concentrati on and is mor e evenl y dist ributed to the biom ass .  W astewaters with foami n g problems in UAS B s ystems. A special version of the EGSB - conce pt is the so - c alled Internal Circulation (IC® ) reac to r Vell inga et al. ( 1 986). In thi s t yp e of rea ctor, the produc ed bio gas is separ ated from the liquid halfwa y the reactor b y means of a gas/l iqui d separator device and conve yed upw ards throu gh a pipe to a degasi fier unit or ex pansion device. Here, the separat ed bio gas is rem oved from th e s ystem, where as the slud ge - wat er mix ture drops back to the bott om of the reactor via anoth e r pipe. In fact, the lifting forces of the coll ected bi o gas are used to bring about a recirculation of liqui d an d gr anula r sludge ov er the lowe r par t of the rea ctor, which res ult s in improved contact betwe en sludg e and wastew ater. The ex tent of liquid/ sludge re c irculation depends on the gas prod ucti on. Th e most common EGSB s ystems ar e pr esented in Figu re 2.6. 20 Figure 2.6 : EGS B and IC reactor of the major an aerobic s ystem 21 2.2.3.2.5 Other anaerobic high rate systems W here ACP , UAS B an d EGSB rea ctors ar e base d on a mix ed to compl etel y mix ed reacto r content, various designs have been tes ted which empl o y sta gi ng of the various phases of anae ro bic treatm e nt Lier et al. ( 2001). An ex treme ex ampl e is the two stage pr ocess wh er e the acidi ficati on step is compl etel y sepa rated from the methanogenic step . Alth ough some l ar ge r scale appli cati ons were m ade on domesti c sewa ge, the reactor is not further dev eloped. Th e major problem is the hyd rod yn ami c limi tation givi ng const r aint s to the achiev able SRT in the system, since the superficial liquid velocit y in a baffled s ystem is s ubst anti all y high er than i n a single step sludg e bed reactor. As a lo gic result s, most of the sludg e will mov e with the liquid through the vario us compartm ents and then has to be sepa rated after the last compartm ent in a sett ler and then return ed to the head of the reacto r. Ver y int eresti ng possi bil it ies ma y ex ist for anae robic sequen c ing ba tch reactor (ASBR ) which consi sts of a set of ana er obic rea ctors ope rated in a batch mode using a 'fill and dra w ' method. A certain amoun t of the raw wastewat er i s suppl ied to the an aerob ic rea ctor, after the supe rnatant liquid of a previous batch has been dischar ged. The n a 'gentl e ' t ype of mix ing of the reactor cont ents is started i n order to enable the s ett led viable sludge to conta ct the was tewater and to eli mi nate the biodegradabl e or gani cs. After a sufficient period of react ion time, the sludge is al lowed to sett le and the supernatant solut ion is discharged. The nex t cycl e is then started. Granulation proce eds well in an ASBR on dil ute wastewa ters, also at lower ambi e nt temperatures Banik et al. ( 1997). ASBR s yst ems were s hown to be of particu lar int e r est for LC FA containing wastewate rs (Alves et al. , 2001). More recentl y anae robic membrane biorea ctors (AMBR ) a re int ensiv el y resea rched Liao et al. ( 2006 ), J eison and Li er ( 2006 ). Memb rane te chnolog y can be consi dered an int eresti n g opti on in those cases wh ere establ ished technologi es ma y fail . This likel y is th e case when ex treme condit ions prev ail , such as hi gh tempe r atures and high sali nit y, or waste waters with ref racto r y and/or tox ic compounds. Full - scale ex periences have demon strated that under those condit ions sludge immobili z ati on by gr anule formation does not develop successfull y, ne gati vel y aff ecti ng sludg e retention. The requirem ents of wastew ater tr eat ment under ex treme co ndit ions is ex pected to become more and more comm on, following the curr ent trend of closi ng indus trial process water cycl es. Under such co ndit ions, MBR s ys tems are ver y 22 effe cti ve in the retention of specifi call y required micro - or ganism s which are need ed f or the remov al of accu mul ati ng ref ra ctor y com pou nds in closed cyc le indus tr ial processes. At p resent onl y a few full scale AM BR s ystems are in oper ati on. 2.2.3.2.6 Acidifying and hydrolytic reactors Ex cept for well stirred tank reactors no specific reactor con cepts have bee n developed for acido gen e sis so far. The pro cess of acido genesis gener all y proc eeds sufficientl y fast in a sti rred tank reactor and in practi ce ther e gener all y do es not ex ist an y real ne ed fo r a compl ete acidogenesis . Mor eov er, now ada ys it is full y understood that joi nt acidi ficati on with methanogenesis is be nef icial fo r granul e formation Verstra ete et al. ( 1996). Furthermo re, it is incr eas ingl y accept ed that the presenc e of higher concent rati ons of acidi f yin g or ganism s in the feed of the met hanogenic reacto r is quit e detrimental for the granular m ethano genic slu d ge pr es ent in that reacto r. The latter means that the sludge retention of an acidogeni c rea ctor needs to be improved . 2.2.4 Anaerobic process kinetics Bact erial conve rsion rate s, including anae robic pr ocesses, ar e gene rall y de scribed as appl yin g Monod kinetics for subst rate conversion . Anaerobic conve rsion kinetics, including all kinetic parameters, have be en rec e nt l y and ex tensivel y rev iewed by Batst one et al. (2002) who presented a unified anae robic digesti on model, denomi nated as ADM1 i n analo g y with the ASM1 for acti vat ed sludge. ADM1 model also makes use of the COD balance fo r de scribin g the flow of electrons du ring the ana e robic conv ersion pro cess. Striki ng are th e la r ge va riati ons in the cit ed assessed kinetic paramete rs for the spe c i fic conversion rea cti o ns, see Table 2.2, after Batst one et al. (2000). This means that process configu ra ti on, ex act prevail ing microbial flor a, and actu al o pe rati on of the s ystem lar gel y dete rmine the appli cable kineti c pa ram eters . 23 Table 2.2: Kineti c param eters of main subs trat es/i ntermediate produ cts in the anaerobic conve r sion process (after Batst o ne et al., 2000) . Sustrate Uptake rate kg/k g VSS d µ max 1/d Y kgVSS /kg K S kg/m 3 K d 1/d Hyd ro gen 2 - 65 0 . 012 - 12 0 . 014 - 0 . 183 0 . 00002 - 0 . 0006 0 . 009 Acetate 3 - 18 0 . 0 5 - 1 . 4 0 . 014 - 0 . 076 0 . 011 - 0 . 930 0 . 004 - 0 . 036 Propionate 0 . 16 - 0 . 31 0 . 004 - 0 . 016 0 . 025 - 0 . 05 0 . 06 - 1 . 15 0 . 01 - 0 . 04 But yr ate 5 - 14 0 . 35 - 0 . 90 0 . 066 0 . 012 - 0 . 30 0 . 027 Valerat e 15 - 19 0 . 86 - 1 . 20 0 . 058 - 0 . 063 0 . 062 - 0 . 36 0 . 01 - 0 . 03 LCF A 1 . 4 - 37 0 . 10 - 1 . 65 0 . 045 - 0 . 064 0 . 06 - 2 . 0 0 . 01 - 0 . 20 Amino acids 36 - 107 2 . 36 - 16 0 . 06 - 0 . 15 0 . 05 - 1 . 4 0 . 01 - 3 . 2 Monosaccha rides 29 - 125 0 . 41 - 21 . 3 0 . 01 - 0 . 17 0 . 022 - 0 . 63 0 . 02 - 3 . 2 2.3 Modeling A model can be defin ed as a purpos eful representation or des cript ion (often sim pl ified) of a s ystem of int erest We ntz el and Ekama ( 1997) . One neve r develops a model that de scribes eve r y sin gle or ganism , ev er y mol ecul e of water o r ever y detail of the process. Models are used as sim pli ficati on of reali t y in such a wa y that the y describe that pa rt of reali t y that is rele v ant to unde rstand and to deal wit h . The most prominent adva nta ges of the us e of mo dels in wastewate r tre atm ent are:  gett in g insi ght into plant performan ce  evaluatin g possi ble scena rios for up gradin g  evaluatin g new plant d esign  supporting mana gement dec isi ons  developi ng new control s chemes  providi ng ope rator tr aini ng 24 The opti mi sati on of the AD and the assessment of its operati on as a functi on of var yin g feed or oper ati ng condit ions ar e importa nt objecti ves and can be pursued b y using approp riate di gest i on models. These models can be of st ead y - stat e mode l  to esti mate retention time, rea ctor volum e, gas producti on and compos it ion for a requested s ystem pe rformanc e,  to investi gate the s ensit ivi t y of the s ystem pe rform ance to va rious param ete rs,  to provide cross - checkin g of simul ati on result s an d plant performan ce, and  to determi ne how the digesti on proc ess can af fe ct the design of upst rea m or downstream WW TP operati ons. More compl ex d ynami c models could be int egrated in plant - wide modelli ng, predicting on a time basis how the s ystem wil l react to sudden or progressi ve changes in operating parameters of feedstock flow rate and composition, temperatur e, inhi bit ion, pH, etc. 2.3.1 Simple models and principle kinetics Most ini ti al models were based on a sin gle rate - l im it ing step, which itself ma y be dependent on various condit ions such as wastewater char acte risti cs, hy dr auli c loading and temperatu re. Some models consi dered acetogenic methano gen esis as the rate - li mi ti ng step, wher e as others consi dered the conversion of fatt y aci ds , or the hydrol ysis of biode grad able suspended soli ds . Pavlost athi s and Gosset t studi ed, developed and ev aluate d a comprehensive kin eti c model capable of predictin g digeste r perform ance when fed biol ogic al sludge. Preli mi nar y conversion mechani sms such as cell death, l ysis , and h ydrol ysis responsi ble for render ing viable biol ogical slud ge or ga nism s to avail able subst rate wer e studi ed in depth. The result s of thi s stud y indi cate that hydrol ysis o f the dead, particulate biom as s prim ar y consi sti ng of protein is t he slowest step and ther e fore kin eti call y cont rols t he ove rall process of AD of biol ogi cal sludge. This rate control by h yd rol ysis was co nfirmed by several authors, includin g,e. g. Hid erani et al. , who used anaerobi c respir ometr y to determi ne di gesti on kinetics. The developed models are sim ple but do not very accur atel y describ e the digeste r behaviour. 25 Table 2 . 3 reviews th e ke y AD models that have been dev eloped so far. Some models have assum ed various forms of the kinetics, the bacterial gro u ps, occurrin g processes, rate - li mi ti ng s teps and possi ble inhibi tion. 26 Table 2.3: Overview of some anae robic di gesti on models . Model Kineti cs Bact erial Groups Processes Lim it ing Step In cluded Inhibi ti on Batst one et. al First order Disi ntegr ati on pH Fi rst order Hyd rol ysis Hyd rol ysis pH Monod Sugar - de gr adin g acido ge ns Acidogenesis pH Monod Amino acid de gradin g ac idogens Acidogenesis pH Monod Propionate util isi ng aceto gens Aceto gen esis pH,H 2 Monod But yr ate and valer ate uti li sing aceto gens Acet o gen esis pH,H 2 Monod Acetoclasti c meth ano gen s Methano genesis pH,fre e NH 3 Monod Hyd ro genotrophic meth a nogens Methano genesis pH,fre e NH 3 Siegrist et al. Mathematical Biogas strippi ng First order Hyd rol ysis pH, free NH 3 , H 2 , acetat e Monod Fermant ati on pH,fre e NH 3 , H 2 , acetat e Monod Anaerobic ox idation of LCFA pH,fre e NH 3 , H 2 , 27 acetat e Monod Anaerobic ox idation of propionate pH,fre e NH 3 , H 2 , acetat e Monod Acetoclasti c meth ano gen s Acetot rophic methano ge nesis pH,fre e NH 3 , H 2 , ace tat e Monod Hyd ro genotrophic meth a nogens Hyd ro gentrophic methanogenesis pH,fre e NH 3 , H 2 , acetat e Sötemann et al. First order, monod, contoi s Acidogenic ba cteria Hyd rol ysis Hyd rol ysis Monod Glucose - uti li sing acidogens Acidogenesis H 2 Monod Propiona te util isi ng acid ogens Acidogenesis H 2 Monod Aceto genic bacte ria Aceto gen esis pH, H 2 Monod Acetoclasti c methano gens on aceti c acid Acetoclasti c meth ano gen esis pH, H 2 Monod Hyd ro genotrophic meth anogens on H 2 H yd ro genotrophic methanogenenis pH, H 2 28 2.3.1.1 The Model of Siegrist et al. The aim of thi s stud y is to develop an ex tended and improved ver sion of a mathematical model to describe the dyna mi c behavior of mesophil ic and thermophi li c digesti on. The approa ch of the model is simil ar to the IW A acti vate d sludge models wher e the ph ysical, biol ogic al, an d chemi cal process es of the model are descri b ed in a stoi chiom etric matrix . The ana erobic mod el is based on the reacti on sch eme des cribe d by Guje r and Ze hnder. (Fi gure 2. 7 ) The model all ows the variati on of di gest ed sludge and bio gas compos it ion, the gas producti on due to tempe r ature and load va riati on, and the indus trial waste addit ion to be sim ulated. Special emphasis is given to gas strippi ng, acet ate and propionate degradati on, h ydrol ysis of particulate de gr adable chemi cal ox ygen dema nd (COD), and inhi bit ion due to pH, free amm onia, h ydro gen and ac etate. The dyn ami c mathemati cal model presented her e describes the anae robi c digesti on process in a compl etel y s ti rred tank reactor (CS T R ). It includes a proces s to describe the strippi ng of bio gas components , the hyd rol ys is of particulate COD, six subst rate degradati on pro cesses to geth er wit h their six inherent biomass deca y ph en omena. The de ca y process es pro duce ine rt and de gr adabl e particul ate or ganic m a tt er. Fou r addit ional chemi cal proc esses, describin g the aci d/base equil ibrium of CO 2 /HCO 3 , NH 4 +/NH 3 , aceti c acid/ a cetate, and propioni c aci d/propionate, respe cti vely, account for pH prediction. Th e model includes an inhi bi ti on term due to free amm onia for aceto trophic methano gen esis and propionate de gr adati on to describe NH3 inhi bit ion in muni cipal sewa ge slu dge tre atm ent at inc re ase d fr e e amm onia conc entr ati on, e. g. at basic pH . 29 Figure 2.7: P roposed rea cti on scheme for ana erob ic digesti on o f domesti c sewa ge slud ge based on Gu jer and Zehnd er. 2.3.1.1.1 Mathemetical modeling of anaerobic sludge digestion The dyn ami c mathemati cal model presented her e describes the anae robi c digesti on process in a compl etel y s ti rred tank reactor (CS T R ). I t includes a process to describe the strippi ng of bio gas components, the hyd rol ys is of particulate COD, six subst rate degradati on pro cesses (Figure 2. 7 and Fi gure 2. 8 ) togethe r with their six inherent biom ass deca y phenom e na. The de ca y process es pro duce i ne rt and de gr adabl e particul ate or ganic m a tt er. Fou r addit ional chemi cal proc esses, describ - in g the ac i d/base equil ibrium of CO 2 /HCO 3 - , NH 4 + / NH 3 , aceti c acid/ a cetate and pr opioni c aci d/propionate, respec ti vel y, ac count for pH pr ediction. Thus, the rea ctor state i s define d by 23 state variables ; 15 solub le (Si) and 8 particulat e components (Xi) whi ch are sho wed below ( Fi gure 2. 8 ) as matrix . The biogas compos it ion i s described by the partial pres sures of metha ne (p C H4 ), car bon diox ide (p C O2 ) and hyd ro gen (p H2 ) . 30 Figure 2.8: Matrix of stoichiometric coef ficients (v j , i ), yields (Y i ) and cons ervati ves (i THOD, i , i N, i , i l, i , i c , i ) of the processes defin e d for an aerobi c digesti on model of Siegrist. 31 2.3.1.2 The IWA AD Model No.1 (ADM1) The ADM1 model, ini ti all y de v eloped b y th e IW A - ADM Test Group was present ed in book form. This boo k presents th e outcome of the stud y und ertakin g and is the result of 4 years of coll aborati ve work b y a nu mber of int ernati onal ex perts from various fields of an aero bic proc ess technolo g y. The approa ch provides a unifie d basis for AD modelli ng and promot es th e incr e ased appli c ati on of mod ell ing and sim ulation as a tool for rese arch, desi gn, ope rat ion and opti mi sati on of anae robic processes. The ADM1 m odel was developed on the basis of the ex tensive but often disparate work in mode ll ing and sim ulation of AD s ystems ove r the pr evious 20 yea rs. In developi n g t he ADM1, the Task Group tried to establi s h comm on nomenclature, unit s an d model structure, co nsis tent with ex isting anae robic modelli ng literature and the popular - acti vated s ludge models. Outputs from the model include comm on process variables such as gas flowand compo sit ion, pH, separat e or gani c acids, and amm onium . The struc ture encou ra ges speci fic ex tensions or modi ficati ons wh er e requ ir ed, but sti ll mai ntaining a comm on pla tform. The model structure is presented in a readil y appli cabl e matrix format for implementation in man y av ail able diff ere nti al equati on solvers. The ADM1 includes biochemi cal as well as ph ysi cochemi c al proc esses. Th e biochemi cal part in cludes all three ov erall biol ogical (cell ular ) steps, i.e. acido genesis , ac eto genes is of both VF A and LC F As, and methanogenesis as well as an ex tracell ular (partl y non - biol ogic al) disi nte gr ati on step and an ex tracell ular hydrol ysis ste p. Th e ph ysicoch emi cal equati o ns describe ion associ ati on and diss ociation, and gas –li quid transfer. The bioch emi cal part of the model uses the following basis :  All biochemi cal ex tracell ular steps ar e assum ed of first order.  S ubst rate uptake us e M onod t ype kinetics as the basis fo r all int ra cell u lar biochemi cal reacti ons.  Biom ass growth is impli cit in substrate uptake.  Death of biom ass is rep re sented by first - ord er kine ti cs.  Inhibi ti on by pH, h ydro gen and free amm onia is included. The ph ysi cochemi c al fac tor s taken int o account ar e: 32  li quid –li quid reacti ons;  gas –li quid ex chan ges. The model has been su c cessfull y t ested on a ran ge of syst ems from full - scale waste sludge di gesti on to labor ator y - scal e thermophi li c high - r ate UAS B rea ctors . Various modi ficati ons hav e be e n developed with the ADM1 as a basis . The se ex tended models were revi ewed by Batst one et al. and amongst th e most promi sing expansions, the reader is referred to Sötemann et al., Zaher et al. and Blumensaat and Kell er. You can see at th e below the COD fl ux (Figu re 2. 9 ) and ana erobi c model as implemented includi ng biochemi cal proc esses ( Fi gur e 2. 10 ). Figure 2.9: C OD flux for a particul ate compos it e . 33 Figure 2.10: The an ae ro bic model as impl e mente d including b ioch emi ca l proc esse s . (1) acidogenesis from s uga rs; (2) acido genesis from ami no acids; (3 ) ac etogenesis from LC FA; (4) ac eto ge nesis from propionate; (5) aceto gen esis from but yrat e and valerate; (6) aceti clastic methanogenesis and (7 ) hydro genotrophic metha noge n esis . All ex tracell ular steps were assum ed to be first order, which is an empric al functi on, refle cti ng the cumul ati ve effe ct of a multi step process Eastm en and Fer gus an ( 1981). Cell ular kinetics are des cribed b y thre e ex pressi ons (uptake, grow t h and dec a y; se e Figu re 2. 11 and Fi gur e 2. 12 ) . 34 Figure 2.11: Bioch emi ca l rate coe fficients (Vi,j) and kinetic rate equ ati ons (p j ) for soluble compone nts of ADM1 . i=1 - 12, j= 1 - 19 35 Figure 2.12: Bioch emi ca l rate coe fficients (V i , j ) and kinetic rate equ ati ons (p j ) f or particul ate compo nents of ADM1 . i=13 - 24, j=1 - 1 9 36 Figure 2.13: Matrix for gas tr ansfe r. Liqui d phas e yi eld coef ficients (V İ,J ) and rate equati ons (p j ) for liquid - gas transfe r. 2.3.1.3 The model of S|temann et. al The approach of Sötemann et al. is very comprehensive. As an alternative to chara cterisi n g the s ewa ge sludg e feed int o carbo h ydr ates, prot eins and li pids, as is done in ADM1 , it is cha racte rised in term s of tot al COD, its particulate nonbiodegradable COD fracti on, the short chain fatt y acid (SC FA) CO D and the CHON content of the particulate or ganics, i.e. X, Y, Z and A in C X H Y O Z N A . Havin g thus char acte rised the sludge in te rm s of measur able pa ramete rs, the mo del all ows COD, C and N mass balances to be set up over the AD system. The in teracti ons between the biol o gical processes and weak acid/ base ch emi str y ar e pr edicted fo r stable stead y - state operation of ADs. The model of Söt emann et al. is a stead y - stat e model, vali dated onl y fo r condit ions of stead y fl ow and load. The mod el equati ons can howev er be transf ormed to predict the digesti on under d yn ami c operati n g condit ions. All kinetic and stoi chio metric const ants in the model, ex cept those for hydrol ysis , were obt ained from the l it erature so that model cali brati on is reduced to determi ning the non - biode grad able particulate COD fracti on of the sewa ge slud ge, th e associated const ants of th e h ydrol ys is kinetics and the s e wa ge slud g e CHON co mpos it ion. Various fo rmulati ons for the h yd rol ysis rat e of sewa ge slud ge particulate biodegr adable or gani cs wer e ev aluated and s urfac emediated reacti on (Contoi s) kinetics were select ed s im il ar to that used by Dold et al. and ASM1 for slowl y biodeg r adable or ganics i n acti vated slud ge s yste ms. Vali dati on of the model under ste ad y - state condit ions vali dates onl y its sto ichiometr y and the s ystem rat e - li mi ti ng pro cess, whi ch is hyd rol ysis . Howev er, the model, which includes the influe nce of hi gh h ydro ge n par ti al pressure on the acido gen esis 37 and aceto genesis proc e sses, shows the ex pected sensit ivi t y to a di gester upset (alt hough comm onl y un noti ced due to the s ystem inerti a) ini ti ated b y temporar y inhi bit ion of the acetocla sti c methano gens, whi ch is the us ual cause in pr a cti se. The model demons trates that even a brief inhi bit ion of thi s organism group causes an irreversible fail ure of the digeste r (pH <6.6). 2.3.1.3.1 Conceptual model In the literatur e there is consi derable va riati on in conceptual schemes fo r describin g the biol ogic al proc esses of AD with sew age sludge as influent, from simple two stage reacti on s chemes including onl y h ydrol ysis /acido gen esis and methano gen esis Kiel y et al.( 1997) to the most comm onl y used six step rea cti on scheme as pr oposed b y (Gujer and Zehnde r, 198 3). In the rea cti on scheme of Guje r and Zehnd er (1983) (Fi g 2. 14 ), the hyd rol ysis process acts sepa ratel y on three main groups of compl ex organi cs, viz . proteins, carboh ydr ates and lipids. These compl ex pol ymeric materi als ar e h yd ro l ysed b y ex tracell ular enz ymes t o solub le products th at are small enou gh to all ow their trans port across the ce ll membrane. The products of the separate hydrol ysis processes are ami no ac ids, sugars and fatt y acids respecti vel y. These relativel y sim ple, sol uble compounds are ferme nted (acido genesis ) or ana erobicall y ox i dised to short chain fatt y acids (S CFAs) (acet ate), alcoho l s, CO2, hydro gen and amm onia. A portion of the hyd rol ys is products are also co nverted to int ermediat e products (propionate, but yrat e, et c.), which are then conv erted to acetate, h yd ro ge n gas and CO2 through a proc ess call ed aceto gen esis . Lastl y, methano gen esis occurs b y hydro gen redu cti on with CO2 (h ydro genotrophic methanogenesis ) and fr om acetate cleava ge (as etoclasti c m ethano genesis ). At the model of Sötemann et. al, The Gujer and Zehnder (1983 ) rea c ti on scheme formed the basis for the AD model developed but with four main modificati ons (Fig. 2. 15 ) 38 Figure 2.14: Ana erobic digesti o n proc esses sch e me of Gujer and Zehnd er (19 83) . Figure 2.15: Anaerobic digesti on proc esses sc h e me of Universit y of Cape Town. 39 Anaerobic Di gesti on M odel No 1 (UCTADM1 ) including (i) th e effe ct of hi gh hydro gen partial pr essur e on acido gen esis and (i i) COD, carbon and nit rogen m ass balances with a gene ric CHON sludge compos it ion. The main modi ficati ons at the model of Sötemann et. al;  R ecognisi n g that ca rboh yd r ate, protein and lipid measure ments on sew a ge sludges are unli kel y to be routinel y av ail able and indeed are di fficult to do, the hydrol ysis of the t hre e separ ate or ganic mat eri als w as modi fied to a single hydrol y sis pro cess acti n g on a gene ric or ganic material repres enti ng se wa ge sludge (C X H Y O Z N A ) Mc Cart y ( 1974). This sim pli fi cati on is not unre asona ble since the end products of hydrol y sis and subseque nt acidogen esis of the three organic groups are essent iall y the same, nam el y S CFAs. In thi s approach, t he C, H, O and N contents of sewa ge slud ges are ne eded to dete rmine the X, Y, Z and A values in C X H Y O Z N A .  W it h the proposed singl e h ydrol ysis proc ess, r e cognit ion of thre e sep ar ate hydrol ys is products was no longe r nec es sar y. Accordin gl y, a sin gle hydrol ysis proc ess and end prod uct wer e included. This end product was chosen to be the idealised carbohydrate “glucose” for a number of reasons: The subsequent biological processes on “glucose” are well established  In an y event, the end pr od ucts of h yd rol ysis and acido gen esis in the sch e me of Gu jer and Zehnder (1983) at Fi g. 2. 14 ar e the same as in the revis ed scheme at Fi g. 2 . 15 , the net result is the sam e in both schemes. In orde r to maintain the COD, C, H, O and N balan ces, water and ca rbon diox ide are taken up from the bulk li quid to gene rate th e gluc ose from the se wa ge slud ge (Fi g. 2 . 14 ), and amm onia is released.  As a consequ enc e of acc epti ng a single h ydro l ysis process, sep arat e an aer obic ox idation of fatt y acids d oes not need to be includ ed.  In the rea cti on s cheme of Gujer and Zehnde r (19 83) at Fi gure 2. 14 , a fix ed proportion of h ydrol ysis end products are conve rted to int er mediate SC FA (propionate, but yra te, etc.) and the balan ce directl y to acetate. As an alt ernati ve, the influen ce of the h ydro gen partial pressure ( p H2 ) on acido genesis of glu cose t o acetat e and propionate as proposed b y Soon et al. (1991) was included in t he revised sch eme. This provides a bett er des cript ion 40 of AD behaviour under fail ure condit ions. In thi s revised sch eme, gene rati on of but yr ate and hi gh er SCFAs was not consi de re d. 2.4 Brewery Wastewater During th e last two dec a des the brewin g indus tr y has shown incre asing aw areness for enviro nmental prote cti on and the need of sust aina ble producti on proc esses. The bee r brewin g pro cess often gen erat es lar ge amount s of wast ewate r effluent and soli d wastes that must be dispo sed off or tre ated in the least costl y and saf est wa y for to meet the disc h ar ge regul ati ons. It is widel y esti m ated that fo r ev er y on e liter of beer that is brewed, close to ten liters of water is used; most l y for the brewin g, rinsing, and cooli ng process es. Despit e dischar gin g lar ge volum es of highl y poll uti ng effluents throu gh o ut the year, the bre wing indus tr y co nsti tut e an important ec onomi c se gment of an y countr y. In fact, be er is the fifth most consum ed beve ra ge in the world behind te a, carbonates , milk a nd coffe e . Be er br ewing invol ves two mai n steps, i.e., bre wing and pa ck a gin g of the finished product. The byproducts (e. g., spent grains from mash ing, ye ast surplus, etc) gen erated from these steps ar e resp onsi ble for poll uti on when mix ed with effl uents . In addit ion, cleanin g of tanks, bott les, machines, and floors produces h igh qu anti ti es of poll uted wate r . It is esti mated that for the produ cti on of 1 L of beer, 3 –10 L of waste ef f luent is gene rated dep en ding on the producti on and specific water usa ge . In other wo rds, ver y lar ge quanti ti es of water are consum ed during the bee r brewin g pro cess. Sim il arl y and bec ause of volum inous water usa ge, t he brew er y indus tr y discha r ges lar ge volum es of highl y poll uti ng ef fluents throu ghout the ye a r . 2.4.1 Brewery wastewater characterisation Bre wer y wastewat er t ypi call y has a hi gh chemi c a l ox ygen demand (COD ) from all the organic compon ents (sugars, solub le starch, ethanol , volatil e fatt y aci ds, etc) . It usually has temperatures ranging from 25 °C to 38 °C, but occasionally reaching much high er temp eratur es. The pH level s can ran ge betwe en 2 and 1 2 and are influenced b y the amoun t and t yp e of chemi cals used in cleaning and sanit iz ing (e.g. , causti c soda, phosphoric acid, nit ric acid, et c.). Nitrogen and phosphoru s levels are mainl y dep endent on the handli ng of raw mate rial and the amount of yeast p resent in the effluent. 41 Bee r indus tr y is mainl y compos ed of biod e grad a ble waste wate r contami nated with organic subst anc es ther ef ore it is treated b y biol ogic all y. Th e wast ewate r producti on is 3 - 9 liter/l it er product b eer. Th e wat er consum pti on is 4 - 10 li ter/l it er beer. The main sources of the or ganic po ll uti on in the wastewater are as follows; (EC, 2006)  Yeast and surplus ye ast  Grains  R esidue  W aste stum  P rocess tanks empt yin g and washing  The fil ter materi al (Kiese lguhr)  R eturn and residual b eer  S ti cker a nd other aux il iary mate rials used in packi ng. The gene ral bre wer y was tewater char acte risati on is showed Table 2.4 (EC, 2006) Table 2.4: The brew er y wastewate r gen eral ch ar a cterisatio n (EC, 2006) . Parameter Value Ran ge Unit COD 1800 - 3000 mg/l BOD 1000 - 1500 mg/l TSS 90 - 700 mg/l Total Nit rogen 30 - 100 mg/l Total Phosphorous 30 - 100 mg/l Heav y Metals LOW Low PH 3 - 13 Temperatur e 13 - 49 °C 42 2.4.1.1 Brewery wastewater COD fractions The tot al COD in the wastewate r is usu all y made up biod e grad able and non - biode gr adable fra cti ons. For th e br ewe r y wastew ater the tot al biode grada ble fracti on consi sted of 96% of th e tot al COD. The readil y biodegr adable COD frac ti on (S S ) was det ermined as the 9% of the tot al COD, readil y hydrol ysable frac ti on (S H ) consi sted of 78% an d the slowl y hydrol ysabl e portion (X S ) was determ ined as 9% of the tot al COD. ( Karlikanovait e et al , 2012) Table 2.5: The COD fracti ons of brew er y wast e water ( A. Karlikanov ait e et al , 2012) . C T I Total In fluent COD C S I Total Biode gr adable CO D C II Total I n ert COD S S I Readil y Biode gr adable COD S H I Rapidl y Hyd rol yz able COD X S I Slowl y Hyd rol yz able COD S S I Solubl e Ine rt COD X II Particulate In ert COD %9 %78 %9 %2 %2 2.4.2 Brewery wastewater treatment methods Bre wer y plants produce large quanti ti es of wastewater wi th high conc e ntrati on of organic poll utants and nutrients, which is cha rac teriz ed b y lar ge variati o ns in these paramete rs. In particular, brewe r y ef fiuents ar e ge nerall y cha ra cteriz ed b y high tot al biochemi cal ox ygen dem and (T BODS ), tot al che mi cal ox ygen d emand (T COD) and tot al sus pended solids. Most large br ewe ries require some de gr ee of wastewate r pr etre atm ent. In cases where the brew er y does not discha r ge to the m unicipal sew er, then pri mar y and secondar y t reatm ent of t he efflu ent is required. However, if th e bre wer y is permitt ed to dischar ge int o a muni c ipal sewer, Pretr eatm ent ma y be requir ed to meet muni cipal regulations and/or to less en the load on the munici pal treatm ent plant . 43 P retreatm ent is me ant to alt er the ph ysic al, ch emi cal, and/or biol o gical pr operties of feed water, T hus impro ving the per forman ce of upst ream proc esses. Therefo re, pretre atm ent is done by ph ysic al, ch emi cal, or biol ogic al methods, or b y a combi nati on of all th ese methods . Table 2.6 lists the unit oper ati ons included withi n eac h cate gor y . Table 2.6: W astewater te reatm ent uni t oper ati ons and process es. P h ysical uni t oper ati ons Screenin g Comm inut ion Flow equali z ati on Sedim entation Flot ati on Chemi cal uni t operati ons Chemi cal precipi tation Adsorpti on Disi nfecti on Chlorinati on Ot her chemi c al applicati ons Biol ogic al uni t operati on s Acti vated sludge pro cess es Aerated la goons Trickling fil te rs Rotating biol ogic al conta ctors Pond stabil iz ati on Anaerobic di gesti on Biol ogic al nut rient remo val Figu re 2.16 is a summ a r y of the gene ric ad vant ages and disadvant a ges of various wastewate r tr eatm ent pr ocesses as shown in liter ature . Th ese cha ract eristi cs ( Fi gure 2.16 ) gene rall y rel ate to t he cost of const ructi on and ease of ope rati on. Generall y, the compl ex it y and cost of wastewate r treatm ent te c hnologies incr ease with the quali t y of the effluent produc ed . In fact, the wat er man agement and wast e disp osal in the brewe r y indus tr y ar e co nsidered as si gnific ant cost factors and importan t aspects in the operati ons of a brew e r y plant . 44 Figure 2.16 : Generi c adv anta ges and disadv anta ge s of conventi onal and no n - conventi onal wast ewat er tre atm ent technolo gies . 45 2.4.2.1 Physical treatment methods Among the first treatm ent methods used are ph ysical unit operati ons, in which ph ysical for ces are appli e d to remove con taminant s. Ph ysical methods rem ove coa rse soli d matter, rather than diss olved poll utants. It ma y be a passi ve pro ces s, such as sedim entation to all ow suspended poll utants to sett le out or float to the top naturall y. In gen eral, these meth ods have yielded l it tl e success; most often result ing in incomplete contaminant removal and/or sep arati on . For ex ampl e, sedimenta ti on has been found to be unsat isfactor y ev enwith the addit ion of coa gulants and other addit ives . 2.4.2.2 Chemical treatment methods Differ ent chemi cals can be added to the brew er y wastew ater to alt er the water chemi str y . Chemi cal pretreatm ent ma y invol ve pH adjust ment or coa gu lation and fl occulation. Th e acidi ty or alkalini t y of was tewater af fects both wastewate r treatm ent and the enviro nment. Low pH indi cates increasin g acidi t y while a high pH indi cates incre asing alkal ini t y. Th e pH of wastew ater ne eds to remain bet ween 6 and 9 to protect or ganism s. Waste CO 2 may be us ed to neutrali z e causti c effl uents from clean - i n - pl aces (C IP ) s ys tems and bott le washers . The waste CO2 can also be used as a che ap acidi f yin g agent for decr easin g the pH of alkaline wastew aters before the anaerobi c reactor, thus replacin g the conv enti ona ll y used acids. Neutr ali zati on with H 2 S O 4 and HCl acids is usuall y not recomm end e d becaus e of their co rros ive nature and sulfate and chloride dischar ge limi tations , which ma y add to the cost of effluent treatm ent ope rati ons . C oagul ati on and flocculat ion are ph ysicoch emi cal process es comm onl y used for the removal of coll oidal material or color from water and wastewate r. In water an d wastewate r tre atm ent, coa gulation impli es the step wher e particles are destabil iz ed by a coa gulant, and thi s ma y in clude th e fo rmati on of small aggre gat es b y Browni an mot ion (perikineti c co a gulation). On the othe r ha nd, the subsequent process in which larger aggr e gates (f locs) are form ed by the acti on of shear is then kno wn as flocculation . Aft er s mall particles hav e for med lar ge r aggre gat es, coll oidal material can then be more easil y removed b y ph ysical separati on pr o cesses su c h as sedimentati on, flotati on, and filt rati on. 46 2.4.2.3 Biological treatment methods Biol ogic al waste treatm e nt processes pla y a centr al role in the way so ciety m ana ge their wastewate rs. It is based on the acti vit y of a wide range of microorganism s, con vertin g the biodegr ad able organic poll utants in the wastewaters. In fac t, brewer y effluents having both chemi cal (wit h ver y hi gh or ganic content) and microbial contaminants are gen er all y tre ated by biol o gi cal methods. Therefo re, after the brewe r y wastewate r ha s under gone ph ysical and chemi c al pr etreat ments, the wastewate r can then und er go biol ogical tre atm ent. Compared to ph ysico ch emi cal or chemi cal methods , biol ogic al methods have thr ee advanta ges : 1 the treatm ent technolo g y is mature, 2 high ef ficien c y in C OD and BOD removal, ran ging from 80 to 90%, and 3 low investm ent cost. Howeve r, thou gh biol ogic al treatm ent proc esses ar e pa rticularl y eff ecti ve fo r wastewate r treatm ent, the y require a high en e rg y input . Biol ogic al treatm ent of wastewate r can be eit he r aer obi c (wit h air/ox yg en suppl y) or an aerobi c (wit hout ox ygen ). 2.4.2.3.1 Aerobic treatment methods Aerobic biol o gic al tre atm ent is perfo rmed in t he pres enc e of ox ygen by aerobi c microor ganism s (prin cipall y bacte ria) th at meta boli z e the or ganic matter in th e wastewate r, ther eb y pro ducing mor e microo r ga nism s and inorganic en d - products (principall y CO2, NH3 and H2O). Aerobic tr ea tm ent uti liz es biol ogical treatm ent processes, in which micr oorganism s convert non - sett leable soli ds to sett leable soli ds. Sedim entatio n t ypicall y foll ows; all owing the s ett le - able soli ds to sett le an d separ ate out. Three opti ons include: 1 Acti vated slud ge proc es s : In th e acti vated slud ge pro cess, the waste w ater flows int o an aerated an d agit ated tank th at is prim ed with acti vated sludge. This compl ex mix ture containing ba cteria, fun gi, protoz oans, and other microor ganism s is coll ec ti vel y refe rred to as the b iom ass. In thi s pro cess, t he suspension of aerobic microorganism s in the aerati on tank is mix ed vigorousl y b y aerati on devices, whi ch also suppl y ox yge n to th e biol o gical suspension. 47 2 Attached growth (biofil m) proc ess: The second t yp e of aerobic biol ogi cal treatm ent s ystem is call e d “att ach ed growth (bio fil m ) process ” and deals wit h microor ganism s that ar e fix ed in place on a so li d surfac e. This “att ach ed gro wth t yp e ” ae robic biol ogic al treatm ent pro cess creat es an envi ronment that supports the growth of microor ganism s that pre fer to rem ain att ach ed to a soli d material. 3 Trickling filter process: In the tricklin g filter process, the wast ewa te r is spra yed over the surfa ce of a bed of rough soli ds (such as grav el, rock or plastic) and is allowed to “trickle down” through the microorganism cover ed media. A variati on of a trickli ng filtrati on proc ess is the biofil trati on tower or otherwise known as th e biot ower. The biot owe r is packed with plasti c or redwood media containing the att ached mic robial gro wth. 4 R otating biol ogic al con tactor pro cess: The rot ati ng biol o gical contact or process consist s of a seri es of plasti c disks att ache d to a comm on sha ft. 5 La goons: Th ese are slow, cheap, and relativel y in e fficient, but can be used for various t yp es of waste w ater. The y rel y on the int eracti on of sunl i ght, algae, microor ganism s, and ox ygen (someti mes ae rated ). 2.4.2.3.2 Anaerobic treatment methods Anaerobi c wastewate r tr e atm ent is the biol o gical treatm ent o f wastewate r without the use of air or elemental ox ygen. Anae robic tre at ment is chara cteriz ed b y biol ogical conversion of or ganic compounds by ana erobic microor ganism s int o biogas, whi ch can be used as a fuel; mainl y methane 55 –75 vol% and carbon diox ide 25 –40 vol% with traces of hyd ro gen s ulfide . In br ewe ries, dire ct uti liz ati on of biogas in a boil er is usuall y the pref err ed solut ion. The reason for thi s is that investm ent costs for a combi ned heat and po we r unit (CHP) are highe r and more ex tensive biogas treatm ent is required. In the conte x t of decreasing fossil fuel reserv es, anae robic wastewate r treatm ent makes a br ewer y more indep end ent from ex ternal fuel suppl y. Furthermo re, it contribut es to a more s ustainable brewin g proc ess. 1 Upflow an aerobic sludge blanket: On e of th e most popular an ae ro bic processes is the Upflow Anaerobi c Sludge Bl an ket (UASB ). In th e UAS B reacto r, the wastew ater enters a vertic al tank at the bott om. The waste w ater passes upw ards throu gh a dense bed of an aerobic sludg e wh ere the 48 mi croor ganism s in the sludge com e int o contact with wastewater subst rat es . This sludge is most l y of a granular nature (1 –4 m m) having supe rior sett li ng chara cterist ics (i.e., at a rate of more than 50 m h − 1 ). The organi c materials in the solut ion are att acked by the microbes, which release bio gas. As the bio gas rises, it car ries some of the granula r microbial blanket. At the top of t he UAS B rea ctor, a so call ed three - ph ase separ ator separates the biom ass fro m the biogas and wastewat er . Th e thr ee - ph ase sepa rator is also known as the gas –li quid –soli d - separ at or . 2 Flui diz ed bed reactor: In a Flui diz ed Bed Re actor (FBR ), wastewate r flows in through the bott om of th e rea ctor, and up throu gh a media (usuall y sand or acti vated carbon ) that is coloni z ed by an acti ve bacterial biom ass. The me dia provides a growth area for the biofil m. This media is “fluidized” by the upward flow of wastew a ter int o the vessel, with the lowest densit y pa rtic les (those with highest biom ass ) moving to the top . Table 2.7 pres ents a gen eral compa rison betwee n anaerobi c and aerobic biol ogical treatm ent s yst ems such as acti vated slud ge. Table 2.7: Anaerobic tr e atm ent as compar ed to aerobic treatm ent . Aerobic s ystems Anaerobic s yst ems Ener g y co nsum pti on High Low Ener g y produ cti on No Yes Biosol ids producti on High Low COD removal (% ) 90 –98 70 – 85 Nutrients (N/P ) removal High Low Space requi rement High Low Disconti nuous operati on Difficult Eas y 49 3. THE DEVELOPED MODEL STRUCTURE 3.1 Introduction of the Model Anaerobic treatm ent of wastewate r is ver y well suit ed for indus tries dischar gin g highl y concent rated (o ver approx im atel y 1,50 0 mg COD/l ) wastewa ters, with nit rogen con centr ati ons that are not too high. The food and food processi ng indus tr y, beer bre weries, soft dri nk producing factories and paper producin g or processi ng factories, and som e ch emi cal indus tries all dischar ge waste wate rs of thi s t ype . wastewate rs. Ther efor e, anaerobi c tre atm ent pro cesses hav e becom e vi a ble for the treatm ent of hi gh stren gt h indus trial waste wate r. High or ganic loadin g rates and low sludg e producti on ar e among the man y advanta ges ana erobic pro cesses ex hibi t over other biol ogic al unit ope rati ons. But th e one feature emer gin g as a major dr ive r for the increas ed appli cati on of anae robic processes is the en er g y producti on. Not onl y does thi s technolog y hav e a posit ive net ener g y produ cti on but the biogas produc ed can also replace fossil fuel sources and therefor e has a di rect posit ive effe ct on gr eenh ouse gas reducti on. This will most certainl y ensure the on go ing, and likel y drasti call y increased, popularit y of anaerobi c digesti on pro cesses for waste tre atm ent in the future. But wh y is th ere a need for a gen eric model? Several benefits are ex p ected from the dev elopm ent of thi s first gen eric model of ana erob ic digesti on:  increas ed model appli cati on for full - scale pla nt design, operati on and opti mi z ati on;  further dev elopm ent work on process opti mi z ation and control, aim ed at direct impl ementati on in full - scale plants;  comm on basis for furthe r model dev elopm ent an d vali dati on studi es to m ake outcomes  more compar able and co mpatibl e;  assi sti ng technolo g y tran sfer from rese arch to ind ustr y. Man y of the abov e point s relate to practi cal, indus trial appl icati ons. Inde ed, thi s is one of the ar eas wh ere most benefits from the appli cati on of a gener ali sed proc ess 50 model can be gained. Wh il e man y dif fer ent an aero bic models have been de vised over the ye ars (and indeed for m the basis of the ADM1 ). These all mod els devi sed and dev eloped for the anae robic slud ge (biosoli ds) digesti on proc esses. Bu t anaerobic process es are not onl y AD proc es ses. Also, anaerobi c wastewat er tr e atm ent is the bi g part of the biol o gical treatm ent methods. Toda y, a lot of indus tries are using ana erobic wa stewater tre atm ent plant because of the removal ef ficienc y of the BOD/C OD par amete rs. The desi gn also depen ds on the volum etric organic loadi ngs known to be lum pe d paramete r and it does not reflect the dyn ami c condit ions of the syst em. In thi s stud y based on these information a mo del was dev eloped and sim pli fied accordin g to other devise d model for the important indus tr y that br ewe r y waste water treatm ent. This model has som e adv anta ges the y are ;  To use not onl y fo r slu dge stabil iz ati on a lso to use wast ewate r an aero bic treatm ent proc esses.  It is sui table for di rectl y i ntegr ati on to ASM modeli ng.  It is sui table for removi n g int er fac e betw een of ASM and ADM.  It can be used for oth er food indus tr y wast ewat er an aerobi c tre atm ent with some smal l modificati ons.  C OD based influent wastewater char acte risti cs can easil y be appli ed in as influent wastewat er ch ar acteriz ati on. 3.2 The Conceptual Of The Development Model C ompl ex compos it e particulate waste is assum ed to be homogeneous, which disi ntegrat es t o carboh yd rate, protein and lipid pa rticu late susbs trate. Disi ntegr ati on and hydrol ysis are ex tra cell ular biol ogic al and non biol ogical processes mediati ng the breakdown and solu bil isatio n af compl ex organi c material to solub le subst rates. The subst rates a re compl ex compos it e particulates and pa rticulate car boh ydr ates, proteins and lipids. Th ese subst rates ar e also products from disi ntegr ati on of compos it e particulates. Other products of disi ntegrati on ar e inerts parti culate and inert solub le mate rial. The produ cts from (enz ymatic) de gradati on of carbohdrates, 51 proteins and lipids are monos accha rides, ami no acids and long ch ain fatt y acids, respecti vel y. Recognisi n g th at ca rboh yd r ate, protein and lipid measurements in the was tewater ar e unli kel y to be routine l y avail able and ind eed are difficult to do it. The hydrol ysis of the three sepa rate or gan ic materials was modi fied to a single h ydrol ysi s process acti ng on a gen eric or ga nic m aterial rep resenti n g the wastewat er. Wit h the propose d single h ydrol ysis pro ces s , reco gnit ion of thr ee s eparat e h ydrol ys is produ cts was no longer ne ces sar y. Accor dingl y, a single h ydrol ysis process and end pro d uct wer e included. This end product was chosen to be the ideal carbohydrate as “glucose” for a number of reasons: The subseque nt biological processes on “glucose” are well established and the acidogenic/fermentation process acting on “glucose” to convert it to SCFAs is unli kel y ever to be rat e limi ti ng. Accordin gl y , in model appli cati on accumul a tion of ‘glucose’ will not occur, e ven un der fail ure condit ions. This impli es that the “glucose” acts merely as an intermediate compound, which is acidified to S C FAs as soon as it is produced. In the rea cti on scheme of Gujer and Zehnd er (198 3) at Fi gure 2.14 a fix ed proportion of h ydrol ysi s end produc ts are conve rted to inte r mediate SCFA (propionat e , but yr ate, etc.) and the balan ce directl y to acetate. Also accordin g to stud y of Ramsa y and Pratap ( 2005) prediction s of aceti c and propionic acid compa re well at certain time but but yric acid i s not predicted in th e br ewer y wastewate r . Th ere fore in thi s stud y we negle cted the but yr at e and but yric acid and we consi der ed acetate and propionate as VFA con centr ati on . After the h ydrol ysis step the glucose is conve rte d to acetate and propionate. This step is named as acido genesi s step. In the thi rd step, propionate is converted to acetate an d hydrojen. Then, thi s ac etate is converted to carb ondiox ide and methane. Th at mean s the biogas sta rted to pr oduce. In the l ast step, the produc ed carbondiox ide and h ydr ojen in the s yst em is converted to m ethane. All these reacti ons are bioch emi ca l reacti ons are includ ed in model. Also, two phase (aqueous - gas ) chemi c al rea cti ons included in the model. The process es wer e formul ated eit her as hyd rol ysis or or ganism gro ups growth pro ces ses . All four or ganism groups wer e acc epted to be subj ect to endogenous respirati o n and so an endogenous mass loss process was included in the model for each group. 52 3.3 The Processes and Components of The Model Based on Anae robic di ge sti on an d Anaerobic tr eat ment process, the model includes 1) Biocemi c al process es  H yd rol ysis pro cess  Acidogenesis (f ermentati on)  Aceto gen esis (ana erobic ox idationof organic acids)  Methano genesis 2) P h ysico ch emi cal proc es ses  Liqui d - li quid process es (Modelli ng of acid - b as e reacti ons)  Liqui d - gas tr ansfe r. Also the model is included inhi bit ion modelling fo r each bio chemi cal pro ce ss. The proc esses schem e of the model is shown simply at Fi gure 3.1 S F 1 X S 2 S p r 3 S AC 4 C H 4 CO 2 5 H 2 Figure 3.1: The pro cesse s scheme of the develope d model . Accordin g to this scheme the processes in the mod el are; 1) Hydrolysis W astewater Glucose ( S F ) 2) Acidogenesis 3C 6 H 12 O 6 → 4CH 3 C H 2 C OOH + 2CH 3 C OOH + 2CO 2 + 2 H 2 O Hydr ol yse d to 53 3) Acetogenesis 4CH 3 C H 2 C OOH + 12H 2 O →4CH 3 C OOH + 4HCO 3 - +4 H + + 12H 2 4) Acetoclastic methanogenesis 6CH 3 COOH → 6CH 4 + 6CO 2 5) Autotrophic methanogenesis 8CO 2 + 32 H 2 →8 CH4 + 8H 2 O 3.4 The Kinetic and Stoichiometry of the Model The dyn ami c mathemati cal model prese nted her e describes the anae robi c digesti on process in a compl etel y s ti rred tank reactor (CS T R ). It includes a process to describe the strippi ng of bio gas components (Figu re 3.2), the hydrol ysis of particu late COD, four subst rat e de gradati o n process es to ge the r wit h their fou r inhe rent bio mass dec a y phenomena. The deca y processes produ ce p a rtic ulate inert microbial pro ducts . The stoi chiom etr y fo r Gujer Matrix for anaerobic mo del is shown Table 3.1 Four addit ional chemi cal process es, describi ng the acid/ base equil ibrium of CO 2 /HCO 3 - NH4 + /NH3, aceti c acid/ a cetat e, and propioni c acid/ propionate, r especti vel y account fo r pH prediction . The biogas compos it ion i s described by the partial pres sures of metha ne (p C H4 ), carbon diox ide (p C O2 ) and hydro gen (p H2 ) . The re a cti on kinetics of the processes ar e first order for h ydrol ysis and of Monod t ype fo r the rem aini ng four process es of microor ganism growth . Th e r ate equati ons are shown Table 3.2. Figure 3.2 : Bio gas Strippi ng Proc esse s and Bio gas Flow Calculat ed with a Pressure . 54 Table 3.1: Bioch emi ca l rat e coeffi c i en t s (V i , j ) an d proc esses for th e devel o p ed mod el. St oi ch i omet ry for Gu j er Mat ri x for An a erob i c Mod el . Compounds i XS SIC X P Processes j 1 )An a erob i c Hyd ro l ysi s of X S - 1 1 Z 1 2)An a erob i c growt h on - Aci d ogen esi s Z 2 1 3)Gro wt h on prop i on a t e - Ac et o gen esi s Z 3 1 4)Ac et oc la st i c Met ha n ogen esi s Z 4 1 5)Au t o t rop hi c Met ha n ogen esi s - Z 5 1 6) Lysi s of X O HO Z 6 N 6 f E X - 1 7) Lysi s of X A HO Z 7 N 7 f E X - 1 8) Lysi s of X M HO Z 8 N 8 f E X - 1 9) Lysi s of X MAO Z 9 N 9 f E X - 1 COD IC 1 1 1 1 0 1 0 1 1 1 1 1 0 1 0 55 Table 3.2: R ate equati ons (p j ) of the develop ed model . NO PROCES S ES RATE 1 Anaerobic h yd rol ysis of X S 2 Anaerobic growth on - Acidogenesis (G rowth of ) 3 G rowth on propionata - A ceto genesis (Growth of ) 4 Acetoclasti c M ethano gen esis (Growth ) 5 Autotrophic Methano gen esis (Growth of ) 6 Lys is of X O HO 7 Lys is of X AHO 8 Lys is of X M HO 9 Lys is of X M AO 56 57 4. MATERIAL AND METHODS 4.1 Conceptual Approach W it hin the scope of the work of wastew ater tr eat ment plant modeling , An adolu Efes Lüleburgaz, biological treatment plant was selected. Anadolu Efes is the sector leader of Turke y with a market sha re of 79 %. Ana dolu Efes biol ogi cal waste water treatm ent is includi ng aer obic and ana erobic tr eat ment processes.  First , the influent wast e water char acteriz ati on of the plant for the yea r of 2010 anal ysis result s wer e used .  Using th e literatu re mod els, a new mod el was developed for brew er y waste water an ae robic tre atm en t process.  Then, the facil it y was modelled with develope d model under stead y - s tate (based on annuall y ave ra ge result s) condit ons. Acco rding to stead y s tate condit ions, kinetic and stoi chiom etric values are dete rmined for brew er y indus tr y. Wit h these values the dynami c condit ion s (dail y basis dynami cs ) are modelled for one ye ar.  Finally, the calibrated model results were compared with plant’s operation condit ions. These evaulation can be used for the ener g y ef ficienc y and opti mi z aton operati onal conditi ons of the plant. 4.2 Efes Pilsen LülebXrgaz Plant Anadolu Efes is a comp an y th at produ ces and m arket beer, malt and non - alcoholic bevera ges in wide ran ge of geo gr aph y includin g Turke y , Russi a, Centr al Asia and Middle East. It is fou nded in 1969 with two brewe ries in Turke y with tot al producti on cap aci t y of 0,3 mill ion hl per yea r. Anadolu Efes become th e leade r in Turkish bee r mar k et in a short period of time . Anadolu Efes is the significant regional pla ye r with 18 brewe ries, 7 malte ries 1 hops proc essi ng fac il it y in 6 countries. It is leade r of Turke y with 83 % market share in thi s sector with 5 brewe ries, 2 malt and 1 hop processi n g plant loc ated in Turke y. It has 10 , 4 mhl beer producti on capa cit y . 58 Efes Pil sen L ulebu r gaz Plant has started product ion in 1998. Producti on capa cit y is for filli ng 677 , 000 hl, m alt producti on 633 , 000 hl. It has tot al 94 , 710 m 2 area with 23 , 735 m 2 closed malt and beer produ cti on area. 4.2.1 Wastewater treatment plant The wastew ater from the p ro ducti on of the plant and int ernal domesti c wastewater of the plant is treated in t he wastew ater tr eatm ent plant that i n cluded ph ys ical and biol ogical process es. Als o producti on waste from the proc ess which is was te ye ast is mix ed with wastewater and go es throu gh waswater tre atm ent plant. The flow diagr am of the plant is sh own Fi gure 4 . 1 . Accordin g to data that belongs to treatm ent plant is evaulated and based on one year operati on data aver a ge dail y wastew ater ch ara c teriz ati on is found. It is shown in Ta ble 4.1. Tablo 4.1: Avera ge annu al wastewat er ch ara cteri s ati on of the Efes Pil sen wastew ater tre atm ent plant . Parameter Value Unit Avera ge Flow rate, Q 577 m 3 /da y BOD 5 2 , 654 mg/l COD 4 , 740 mg/l TSS 953 mg/l Temperatur e 30 o C pH 7 . 25 - TN 38 mg/l TP 15 mg/l 59 Figure 4.1: Th e Flow dia gr am of the wastewate r tr eatm ent plant of Efes Pil sen. (Görgün et al , 2007) . 60 The capa cit y of the trea tm ent plant is approx imatel y 1500 m 3 /da y. The treatm ent plant is consi st of;  S creen Chann el  Liftin g Tank  P rimar y Clarifi cati on  Balan ce Tank  C ondit ioni ng Tank  EGSB Ana erobic Re acto r  S elector Tank  Aerati on Tank  S econdar y Clarifi cati on  S ludge Thickenin g Tank  Decant er 4.2.1.1 The anaerobic treatment reactor of the plant In the pl ant EGSB t ype anaerobi c rea ctor is use d for the an aerobic t reat ment. This reacto r has 337 m 3 capa c it y. The len gth of the tan k is 13.2 m and diameter is 5.7 m. The wast ewate r ci rculati on betwe en condit ioni n g tank and a na erobic t ank i s provided 2 pcs , each 7 . 5 kw and 91 m 3 /h capacit y pumps . An ex panded granul ar sl udge bed (E GSB) reacto r is a variant of the UAS B reactor. The dist inguishi ng featur e is that a faster rate of upward - flow velocit y is designed fo r the wastewate r passi ng t hrough the slud ge bed. The increas ed flux permits part ial ex pansion (flui diz ati on) of the granular sludg e bed, improvin g wast ew ater - slud ge contact as well as enhan cing s e gre gati on of smal l inacti ve suspended par ti cle from the sludg e bed. Hi gh rate mix ing condit ions pr ovide high gr anulation propert y to sludge and or ganic loadi ng rate of the reactor can be incr eas ed to 30 kgCO D/m 3 . The increas ed flow velocit y is eit her accompl ished by uti li z ing tall react ors, or by incorporati n g an effluent rec ycl e (or both). 61 The COD removal effic ienc y of the EGSB t yp e rea ctor f or beer waste water was obtained between %85 - %93 at low (15 O C ) and medium ( 35 O C ) temperature. Th e higher h yd rauli c loadin g rate is more suit able at low temperature for bee r wastewate r that is found by ex perime ntall y (Connau ghton et al., 2006) . The inl et wastewate r charact erisati on of the EGSB reactor is shown Table 4.2. Befo re int rodu cti on of wastew ater in a EGS B reactor, the condit ions of the wastewate r are opti mi sed for anae robic biom a ss in a condit ioni ng ta nk. In the condit ioni ng tank pH an d temperatu r e are co rre ct ed and micronutri ents ca n be added to the wast ewate r if requ ired. Aft er condit ioni ng the waste water is pumpe d int o the reacto r, and evenl y dist ributed over the rea ctor bott om through an influent dist ributi on s ystem. Th e wastew ater passes th ro ugh a dens e bed of anae robic gr anular biom ass, t ypical l y consi sti ng of self form ing black pell ets with a diameter of 2 - 4 mm. The granul ar bi omass is characte riz ed by a hi gh sett li ng velocit y of >50 m/h and a densit y of 40 - 80 kg Dry Solids/m³. In top of the reactor, the biom ass, produced biogas and tre ated wate r are sepe rated b y a th ree phase sep ar ator (set tl er). The treated wate r is discharged to the condit ioni ng tank where part of it is mixed with the feed water and recircula ted to the rea ctor. The rec ycli n g rate from the reacto r to condit ioni ng tank is 2 . 4 - 3 of the plant. Produced biogas is also conve yed to the gas chamber of the condit ioni ng tank be fore furth er uti li z ati on. By recircula ti on of the treated wat er and removi ng of degassi n g CO 2 with the bioga s strea m from the condit ioni ng tank the required causti c use for wast ewater neut rali z ati on is reduced b y rec ycli n g of alkalini t y. Since both the condit ioni ng tank and the EGS B reacto r ar e operated gas tight, th e s ys tem is odour fre e and gas holder for gas stor age is not necessa ril y required. Ad dit ionall y an emer genc y flare is alw a ys lo cated close to the anaerobi c re actor. EGSB Technolo g y is 100% natur al and it ma ke use of th e natu rall y occu ring anaerobi c microb es to deal wit h wastes. EGSB Re actor is the mai n compon ent of this s ystem. It includes a ta nk with a water inl et dist ributi on s ystem at the bott om. Bact erial sludg e blanket is developed insi de the reacto r. As the efflue nt passes through the bio - film, the biodegradabl e matter contained in it is degra ded . This brings down the BO D an d COD signific antl y. 62 Figure 4.2: The EGS B reactor of the plant. Table 4.2: The wastewat er cha ra cterisatio n of the inl et EGSB rea ctor. Parameter Unit Value Q m 3 /da y 435 COD mg/l t 3 , 750 TSS mg/l t 943 4.3 Model Implementation Using Operational Data 4.3.1 Simulation The model was cali br at ed and verified b y Aqu asim ( www.e awa g. ch ), a pro gr am designed m ainl y fo r co nducti ng res ear ch with tool s for paramete r esti mation and sensit ivi t y anal ysis of the mo del. The program AQUAS IM was desi gn ed for the identificati on and sim ula ti on of aquati c s ystems i n the labor ator y, in t echn ical plants and in nature. It p erforms the four tasks of  sim ulation,  identi abil it y anal ysis ,  paramete r esti mation,  uncertaint y an al ysi s . 63 A simulation run usually starts with findin g a stead y st ate b y means of forw ard int egr ati on. An int egrati on period 5 times the HRT is necessar y to find an eff ecti ve stead y state withl ess th an 1% devi ati on. After rea ching st ead y - stat e condit ions, dynami c sim ulation data are calculat ed with the aid of forward int egrati on on the basis of a feed variati on file. In stead y state calcul ati on, the aver a ge biom ass compos it ion was tried to be obtained prior to d yna mi c sim ulation. The pro gr am offe rs a free de finiti on of the bioki neti c model, flow sc heme, and process control strat e gie s, graphic support of the sim ulation, and ex peri mental data as well as communi cati o n with spreadshe et pro gr ams. 4.3.2 The Wastewater COD Fractions For Simulation For the br ewer y waste water , t he readil y biod egradabl e COD fracti on (S S ) was determi ned as the 9% of the tot al COD, readil y hyd rol ysabl e fra c ti on (S H ) consi sted of 78% and the slowl y hyd rol ysabl e portion (X S ) was determi ned as 9% of the tot al COD Karlikanovait e et al ( 2012) . Accordin g to thi s stud y the br ewer y wastewate r COD fracti ons consi dere d as Table 4.3 fo r the sim ulation. Table 4.3: C OD fracti on s of brew er y wast ewate r for sim ulation C OD Fracti on Abbreviation Fracti on of CT Total COD C T %100 Solubl e COD S T % 89 Particulate COD X T % 11 Ferment able COD S F %87 Solubl e Ine rt COD S I %2 Slowl y Biode gradabl e COD X S %9 Particulate Ine rt COD X I %2 4.3.3 Steady State Simulation Firstl y the plant was modelled in stead y stat e condit ions. For the stead y - state condit ions model implementati on, the avera ge influent data of the plant presented in the Table 4.3 for 2010 were us ed at oper ati ng condit ions and the model cali brated. 64 For the an ae robic mode li ng of the plant the sto ichiometric and kinetic values ar e selected ac cordin g to between des cribed li teratu re valu es (T able 4.4) . The EGSB reacto r was regarded as a CSTR s ystem be cau se the eff ects of the hi gh re circulation rate Brito and Melo (199 5). Therefo re the EGSB reactor is acc epted as CSTR s ystem in this stud y . When sim ulating the mo del a c cordin g t o stead y s tate condit ions the result s give the model’s, kinetic and stoi chiom etric values fo r brew er y an aerobic wastewate r treatm ent. Th e de fault va lues for model was ini ti all y used, then th e pa ram eters wer e tuned indivi duall y to fit the simulation result on ex perimantal data. Table 4.4: Influ ent wast ewater cha ract eriz ati on used in stead y - state sim ulation P arameters Abbreviation Unit Measurement Sludge ret enti on time SRT d 20 In fluent flowr ate Q m 3 /da y 435 In fluent total COD COD 0 mg/l 3 , 750 TSS TSS 0 m g/l 943 pH 7.3 Table 4.5: Kineti c and stoichiometric values of th e literature models Parameters S IEGR IS T BATSTONE SAM SOON ADM1 (RANGE) k H (1/da y) 0 . 25 0 . 25 0 . 381 0 . 041 - 1 . 94 k d _ OHO (1/da y) 0 . 8 0 . 02 0 . 041 0 . 01 - 3 . 20 k d _ AHO (1/da y) 0 . 06 0 . 02 0 . 015 0 . 0 1 - 0 . 04 k d _M HO (1/da y) 0 . 05 0 . 02 0 . 037 0 . 012 - 0 . 036 k d _M AO (1/da y) 0 . 3 0 . 02 0 . 01 0 . 009 - 0 . 30 µ OHO (1/da y) 4 3 0 . 8 0 . 41 - 21 . 25 µ AHO (1/d a y) 0 . 60 0 . 52 1 . 15 0 . 02 - 1 . 07 µ M HO (1/da y) 0 . 37 0 . 4 4 . 39 0 . 1 - 0 . 474 µ M AO (1/da y) 2 2 . 1 1 . 2 2 - 2 . 6 K F (gCOD/ L) 0 . 05 0 . 5 0 . 15 0 . 023 - 0 . 630 65 K PR (gCOD/ L) 0 . 02 0 . 3 0 . 01 0 . 020 - 1 . 146 K A ( gC OD/ L) 0 . 04 0 . 15 0 . 000832 0 . 028 - 0 . 930 K H2 (gCOD/ L) 0 . 001 2 . 5*10 - 5 0 . 0025 0 . 000088 - 10 - 6 K I _ H2 (gC OD/ L) 10 - 6 3 . 5*10 - 6 4.3.4 Dynamic simulation For the dyn ami c sim ulation model implementation, dail y plant result s for influent wastewate r char acte riz ati on and the other oper ati ng paramet ers wer e use d and th e model cali brated. The an nuall y result are used an d cali brated. Acco rdin g to modelin g result the ex ist operati on condit ions are evaulated . These eva ulation can be used for the ener g y efficien c y and opti miz aton operati onal condit ions of the plant. This model can be used as a support ing tool fo r the oper ati on of the plant. Th e efflu ent data of the plant could be well charact eriz ed with the model . The pl ant’s EGSB anaerobic reactor’s daily resu lt s of year 20 10 were us ed for th e sim ulation. The flow, C OD inl et and outl et TSS influent and ef fluent OLR and pH values ar e shown below figures. 66 Figure 4.3: The inl et flo w of the plan t . Figure 4.4: TSS inf luent and TSS ef f luent values of the EGSB rea ctor. 67 Figure 4.5 Th e COD inl et values of the EGS B re actor Figure 4.6: The COD ou tl et values of the EGSB reactor 68 Figure 4.7: The OLR val ues of the EGSB reactor Figure 4.8: The pH values of the EGSB re acto r 69 5. RESULTS AND DISCUSSION 5.1 Steady State Simulation Results With treatment plant’s annual average data the new developed model is calibrated and some kinetic and stoi chiom etric values are determi ned fo r the brewe r y wastewate r ( T able 5 . 1 ) . When determi ned thes e values the diff er ent valu es ar e used several tim es and the mo st suit able ones are verifi ed for the brewe r y wast e water. Also the stead y state s im ulation result of the model to com pa rin g wastewate r treatment plant’s data is shown Table 5 .2. Table 5.1: Kineti c and stoichiometric values that are det ermined fo r the m odel. Consi derati ons S ymbol Unit Value Default Valu e (ADM) Hyd rol ysis rate const ant k h yd d - 1 1.7 0.041 - 1.94 Max im um growth rate of or dinar y heterotrophic or ganism s µ OHO d - 1 4 0.41 - 21.25 Max .specific growth rate of aceto gen esis or ganism s µ AHO d - 1 0.6 0.02 - 1.07 Max .specific growth rate of acetocl asti c methano gen organism s µ M HO d - 1 0.37 0.1 - 0.474 Max .specific growth rate of hydro genotrophic metha nogen organism s µ M AO d - 1 2 2 - 2.6 Yield of ord inar y heterot rophic biom ass Y OHO g cell C OD/ gCOD subst rate 0.12 0.01 - 0.17 Yield of acetogenesis bi omass Y AHO g cell C OD/ gCOD subst rate 0.08 0.025 - 0.050 Yield of a cetoclasti c met hanogen biom ass Y M HO g cell C OD/ gCOD subst rate 0.07 0.014 - 0.076 Yield of h ydro genotroph ic methanogen biom ass Y M AO g cell C OD/ gCOD subst rate 0.05 0.014 - 0.060 Endogen ous dec a y coe ffi cient for ordinar y hete rotrophic organism s k d _ OHO d - 1 0.0 8 0.01 - 3.20 Endogen ous dec a y coe ffi cient for k d _ AHO d - 1 0.0 8 0.01 - 0.04 70 aceto gen esis or ganism s Endogen ous dec a y coe ffi cient for acetocl asti c methano gen organism s k d _M HO d - 1 0. 18 0.012 - 0.036 Endogen ous dec a y coe ffi cient for hydro genotrophic metha nogen organism s k d _M AO d - 1 0. 18 0.009 - 0.30 Half saturati on const ant for anaerobi c fe rmentation K F g COD/ L 0. 0 25 0.023 - 0.0630 Half saturati on const ant for propionate de gr adati on K PR g COD/ L 0.0 25 0.020 - 1.146 Half saturati on const ant for acetat e de gr adati on K A g COD/ L 0. 42 0.028 - 0.930 Half saturati on const ant for uptake of h ydro gen K H2 g COD/ L 2.5 x 1 0 - 5 0.000088 - 10 - 6 Rate coef ficient base / aci d conversion K AB M - 1 d - 1 10 8 10 8 High pH inhibi ti on level for X OHO pH OHO HH 8.5 5.5 - 8.5 High pH inhibi ti on level for X AHO pH AHO HH 8 .5 5.5 - 8.5 High pH inhi bit ion level for X M HO pH M HO HH 8.0 6.7 - 8.5 High pH inhi bit ion level for X M AO pH M AO HH 8.0 6 - 6.7 Low pH inhibi ti on level for X OHO pH OHO LL 5. 5 4 - 4.5 Low pH inhibi ti on level for X AHO pH AHO LL 5 .5 4 - 6 Low pH inhi bit ion level for X M HO pH M HO LL 6.0 5.8 - 6 Low pH inhi bit ion level for X M AO pH M AO LL 6.0 5 - 5.8 Table 5.2 : Stead y State Sim ulation Results P arameters Unit Effluent of EGS B Model Result s COD mg/l t 957 835 TSS mg/l t 781 785 Biogas m 3 /da y 805 784 pH - 7 . 38 7 . 35 71 Accordin g to stead y sta te sim ulation result s; we can sa y the an aerobi c r eactor ’s removal efficien c y is app rox im atel y 76%. Th e CO D removal ef ficienc y o f t he EGSB t ype reactor for brew er y wastewate r was obtained between %85 - %93 at low (15 O C ) and medium ( 35 O C ) temperature. According to these value, the plant’s anaerobic reacto r is workin g l ess removal ef ficienc y. Also the EGSB rea ctor is gen erall y can achieve ex treme or ganic loading rates (ex ceedin g 30 to 40 kgCOD/m 3 .d). But in the plant the ave ra ge OLR is 5 kg COD/ m 3 .d . According to thi s value we can sa y that anaerobi c re actor is work ing less capacit y. 5.2 Dynamic Simulation Results In th e d ynami c sim ulatio n, dail y op erati onal data s in 2010 wer e used for modeling. The result s of the sim ulation was evaluated wi th the plant data togeth er in the dynami c sim ulation fi gur es below. Sim ulation result shown in Figure 5.1, fo r most da ys of the yea r, effl uent TSS concentr ati ons and th e model result s are close to each othe r. From the result s, TSS removal for the facil it y is nearl y % 17 . How ev er, some differ enc es ap pear in the figur e from the model result because of the plant operati on condit ions (shock loads, anaerob i c reactor oper ati on condit ions etc. ) . Also we can sa y with thi s resu lt the TSS removal efficie nc y ver y low. In the reacto r there is accumul ati on of the soli ds therefor e the removal ef f icienc y of TSS ver y low. pH sim ulation result s shown in Figure 5.2, pH values under operati n g co ndit ions is sli gthl y more than the model result s in period of March - M a y and end of the year . In operati on al con dit ions HC L and NaOH ch emi cal s are added for th e adju sti ng pH value ther efor e thi s ef fe cts the ope rati onal pH values and som e diff eren ces app ea r from the model result s. Simulation result s shown in Figu re 5.3 for effluent COD concent rati ons . Th e efflu ent concentr ati on of the syst em is equ a l to the sum of X T and S T . X T is equal to sum of the tot al biom as conc entrati on produc ed in th e s yst em, particulate ine rt COD X I and slowl y biode gr adabl e C OD Xs. S T is equal to su m VFA con centrati ons, fermen table COD S F and solub le inert COD. From begini n g of the year unti ll April 2010, the effluent COD con centrat ions are close to mod el result s. Since Ap ril there are some 72 chan ges from model res ult because of the plant operati on condit ions in the summ er time , wastewate r cha ract erisati on and some adjus ments for the ana erobic reactor. Simulation result s shown in Fi gur e 5.4 fo r the VFA con centrati on. VFA concentr ati on is represe nted as the sum of the acetat e, aceti c aci d, pro pionate and propioni c acid. As show n i n the fi g u re S PR is quit e small er than S AC in the model alread y acet ate is consi de red in the s ystem . Simulation result s shown in Figure 5.5 for the tot al biogas flowrate of the EGSB plant. Gener all y the bio gas concentr ati on of the plant is si milar to mode l result s. Good prediction of gas producti on is imporatant because it provide s a direct indi cati on of the removal efficien c y of the process. Accordin g to stead y st ate sim ulation result the approx im atel y av e ra ge 2 . 14 m 3 bio gas / m 3 reactor is pr oduced. As sho wn table 5.1 the growth rates, yi eld of biom ass, deca y coe ffic ients, half saturati on coef ficients and the pH values ar e cali brated and verified with the literature. T h ese values are in the ran ge of the literature values. Th ese values are suit able for the b r ewer y wastewate r and can be us ed for it. . 73 Figure 5.1: D ynami c si mul ati on result s of the TSS concentrati ons 74 Figure 5.2: D ynami c si mul ati on result s of the pH values. 75 Figure 5.3: D ynami c si mul ati on result s of the ef fluent COD . 76 Figure 5.4: D ynami c si mul ati on result s of the VFA . 77 Figure 5.5: D ynami c si mul ati on result s of the Bi ogas . 78 In th e Fi gur e 5.6 the biom ass compos it ion of the system is shown. The most abundant biomass are Pa rticulate In ert COD (X I ) and Ordina r y heterotroph ic bio mass (X OHO ) that each on e has 33% of the total biom ass. Figure 5.6: The b iom ass compos it ion of the syste m . In the Fi gur e 5.7 the OLR and F/M rate are sho w n of the syst em. Acco rding to these result s; the max OLR of the sytem is 15 kgCOD/ m 3 .da y. Even thi s OLR value is less than EGSB reactor’s literature value. Therefore the EGSB reactor is not working with efficient capacit y. Also thi s result s show when OLR is incr easin g F/M rate is incresin g. The s ystem is verified wit h thi s result . Accordin g to all thes e result s we can mainl y sa y th at, the model is succesfull y implemented in the d yna mi c sim ulation for the Ef es Pil sen plant. Acco rdin g to O LR of the an ae robic reacto r the reactor is not workin g with sufficient capa cit y. Also in the anae robic reactor the r e i s biom ass accumul ati on therefor e some adjust m ents must be made for takin g out t he sludg e. In the syst em the TSS removal ef ficie nc y is not enough. Based on the model also the prim a r y clarific ati on proses, the operati onal condit ions of the plant must be chec k ed and reme died 79 Figure 5.7: The OLR an d F/M rate of the plant . 80 81 6. CONCLUSION The anae robic model is a powerful tool to describe the d ynami c be havior of mesophil ic or thermophi li c digesti on or tr eatm ent The number of models presented in liter at ure is ex tensive, and often of ver y sp ecific nature. The most freque ntl y us ed m odel, ADM1 developed by the IW A , forms a good basis and is often used in expanded models, as proposed by, e.g. Sötemann et a l.. Sim pler models for digesti on h av e be en pr oposed b y, e. g. Bal a, Siegrist and others. Gene rall y all thes e models are us ed for do mesti c digesti on sludge. This stud y sho wed that a full scale indus trial anerobic tr eatm ent pla nt can be sim ulated with a newl y developed anae robic mode l. This model has some advan ta ges the y are;  This anaerobi c model is developed based on lit eratur e models. This model can be used for modelin g of ana ero bic wastewat er treatm ent plants and also industry’s waste water treatment plants.  This model is sui table for directl y int e grati on to ASM modeli ng .  To use not onl y fo r slu dge stabil iz ati on also to use wast ewate r an aero bic treatm ent proc esses.  It is succ essfull y implemented in the d ynami c sim ulation.  It is sui table for removi n g int er fac e betw een of ASM and ADM.  The cali br ated par ameter s c an be used for br ewe r y wastewate r.  The cali br ated model can be used s ystem optim iz ati on.  C OD based influent wastewater char acte risti cs can easil y be appli ed in as influent wastewat er ch ar acteriz ati on.  H yd rol ysis pro cess is the rate lim it ing step  No detai led ch ara cteriz ati on is required like lipids, carboh ydr ates etc.  The proc ess simpl y runs on propionate and acetat e 82 In the future studi es thi s stud y can be developed for other indus tries. The wastewate r chara cteriz ati on can be different for other indus tri e s there fore dif fer ent stoi chiom etr y can result s. Wit h ex periments studi es thi s stoichiometr y. Also obtaining for mor e detail cali brati on res ult s of the modelli n g, it can be measured VFA con centr ati ons with ex periments and other COD fra cti ons can be ex tende d. 83 REFERENCES Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A., Sanders, W.T.M., Siegrist, H. and Vavilin, V.A., 2002 . The IW A Ana erobic Digesti on Model No 1 (ADM1), IW A Publi shing, London. Biological Wastewater Treatment, 2008. Principl es, Modelli ng and Design, IW A Publi shing, London. Blumensaat, F., Keller, J., 2004 . Modelli ng of two stage ana erobic di ges ti on using the IW A Ana erobic Di ge sti on Model No. 1 (ADM1), Water Research, 39 (2005), 171 - 1 83. Brito, A.G., Melo, L.F., 1995 . A Simpl ified Analys is of Reacti on and Mass Transfer in UAS B and EGSB Re actors, Univ ersit y of Minho, Environmental Technology,18, 35 - 44. Cronin, C., Lo K.V., 1997. Anaerobic Tr eatm e nt Of Brew er y Wastewa ter Using UAS B Re act ors Seed ed Wit h Acti vated S ludge, Bioresource Technology 64 (1998), 33 - 38. G|rgün, E., İnsel, G., Çokg|r, E., 2007. Efes Pilsen Lüleburgaz Tesisleri Atıksu Arıtma Tesisi Performansının İncelenmesi ve İyileştirilmesi Projesi Sonuç Raporu, İo Çevre Çözümleri, İTÜ. Hansruedi Siegrist, Dea Vogt, Jaime, L., Garcia-Heras and Willi Gujer, 2002 . Mathematical Model for Meso and Thermophil i c Ana erobic Sewa ge Sludge Di gesti on, Environ. Sci. Technology, 36, 1113 - 1123. Henze, M., Gujer, W., Mino, T., and van Loosdsrecht, M.C.M., 2000. Acti vated Sludge Models: ASM1, ASM2, ASM2d and ASM3, Scientific and Technical Report No:9 , IWA Publ ishi ng, London. Henze, M., Gujer, W., Mino, T., Matsuo, T., Wentzel, M.C., Marais, G.v.R., and van Loosdsrecht, M.C.M., 19 99 . Acti vated sl ud ge model No.2d, ASM2d, Wat.Sci.Tech. , 39(1), 165 - 182. Ian, R. Ramsay, and Pratap, C. Pullammanappallil, 2005. Full Scale Applicati on of a D ynami c Model for High - R ate An ae robic Wastewate r Tre atm ent S ystems, Journal of Environmental Engineering, 131 (7), 103 0 - 1036 . İnce, O., Germirli, F., KasaSgil, B., Anderson, 2010. Ex perimental Det erminati on of the In ert Solubl e COD Fracti on of a Brew er y W astewat er under Anaerobic Condit ions, Environmental Technology, 19 (4), 437 - 442 . Kato, M.T., Rebac, S., and Lettingo, G., 199 8 . Anaerobi c Tre atm e nt of Lo w Strength Brewer y Wast ewater in Ex panded Granular Slud ge Bed Reactor, Applied Biochemistry and Biotechnology , 99 (76), 273 - 2289 . Lise Appels, Jan Baeyens, Jan Degreve, Raf Dewil, 2008. P ricipl es and potential of the anae r obic di gesti on of waste - a cti vated sludge , Progress in Energy and Combustion Science 34 (2008), 755 - 7 81 . Orhon, D., Germirli, F., Karahan, Ö., 2009, Industrial Wastewater Tr eatm ent b y Acti vated Sludge, IW A P ubli shing, London. 84 Öktem, Y., Tüfekçi, N., 2006, T reatm ent of Br ewer y Wastew ater B y P il ot Scale Upflow Ana erobic S ludge Blank et Rea ctor In Mesophili c Temperatur e, Journal of Scientific&Industrial Research , 65, 248 - 251 S|temann, S.:., RistoZ, N.E., :entzel., MC., and Ekama, G.A., 200 4 , A stead y state model for anaerobi c digesti on of sewa ge sludges, Water SA, 31 (4), 378 - 4738. S|temann, S.:., RensbXrg, P., RistoZ, N.E., :entzel., M.C.,LoeZenthal, RE., and Ekama, GA., 2005, Int e grated chemi c al/ ph ysical and biol ogi cal processes modelli n g Par t 2 - Anaerobi c diges ti on of sewa ge slud ges, Water SA, 31 (4), 378 - 47 38. Simate, G.S., Cluett, J., Iyuke, S.E., Musapatika, E.T., Ndlovu, S., Walubita, L.F., Alvare, A.E., 2011 , The tr ea t ment of brew er y wastew ater for reuse: State of the art , Desalination, 273 (2011), 235 - 247 . 85 CURRICULUM VITAE Candidate¶s fXll name: Deniz Albayrak Place and date of birth: Kelkit- GÜMÜŞHANE 00111 e-mail: denizal@yahoo.com Universities and Colleges attended: : Trakya University-Environmental Engineering Atak|y CXmhXriyet High School Work Experience: PWT Wasser und Abwassertechnik GmbH (2011-Present) Process Enginerr ERA Environmental Technologies (2008-2011) Project Engineer TAYPA and TAYSÖRME Te[tile ComSany (2005-2006) Customer Representative ECESOY Tekstil San. Tic. Ltd. Şti. (2004-2005) Environmental Engineer