Biyoreaktör verimlerine akım şartlarının etkisinin araştırılması

Abstract

ÖZET BİYOREAKTÖR VERİMLERİNE AKIM ŞARTLARININ ETKİSİNİN ARAŞTIRILMASI Atıksuda kirliliği meydana getiren bileşenlerden organik maddelerin, kendilerini ayrıştıracak mikroor ganizmalarla yeterli bir süre temas etmeleri gerek tiğinden, biyolojik reaktörlerde arıtma verimiyle akım şartları' arasında sıkı bir ilişki vardır. Bu çalışmada' aktif çamur reaktörleri akım modellerine göre incelenmiş, piston akımlı reaktörler için yeni bir model geliştirilmiş, yüksek kirlilikte lif levha endüstrisi atıksularının orijinleri, özellikleri`, - aktif' çamur metoduna göre arıtım esasları ve kinetik katsayıları tesbit edilmiş ve deneysel çalışma ile elde edilen sonuçlar incelenen akım modellerine uygulanmıştır. Çalışmanın birinci bölümünde konuya genel bir giriş yapılmış, biyoreaktörlere ait akım modelleriyle ilgili literatürde mevcut temel çalışmalar hakkında özet bilgi verilmiş ve kısaca çalışmanın maksadı belirtilmiştir. İkinci bölümde esas olarak aktif çamur reaktörleri tanıtılmıştır o Biyoreaktörlerde meydana gelen olayların mikrobiyolojik esaslarını tanımlayan ifadeler özetlenmiş ve kütle korunum denklemleri vasıta sıyla tam karışımlı reaktörleri karakterize eden denklemlerin birinci mertebe ve Monod kinetiği için el de edilişi ayrı ayrı gösterilmiştir. Seri bağlı tam karışımlı reaktörler birinci mertebe kinetiği ve Monod kinetiği için ayrı ayrı incelenmiş, (X /X), (© /©) ve r arasındaki bağıntı grafik olarak göste rilmiştir,, Birinci mertebe ve Monod kinetiğine göre piston akımlı reaktör denklemleri çıkarılmıştır» Pis ton akımlı reaktörler için genel bir kabul gören Lawrence-McCarty modeli üzerinde özellikle durulmuş tur. Üçüncü bölümde piston akımlı reaktörler için geliştirilen yeni modelin dayandığı temel esaslar belirtildikten sonra, kütle korunum denklemleri Xkullanılarak reaktör çıkışında biyokütle ve organik madde konsantrasyonlarını veren yeni ifadeler geliştirilmiştir. Elde edilen yeni denklemlerin, bilgisayar tekniklerinin de kullanıldığı çözüm tekniği verilmiş, ayrıca, yeni modele göre çamur yaşını hesaplamaya yarayan yeni bir ifade elde edilmiştir. Geliştirilen modelle Lawrence-McCarty modeli mukayese edilmiş, yeni modele göre hidrolik bekletme süresinin ve reaktör hacminin hesap esasları ve bilgisayar programı verilmiştir. Dördüncü bölümde lif levha endüstrisinin özel likleri, atıksularınm kaynakları, miktarları, karak- tersitikleri ve arıtılması ile ilgili daha önce yapıl mış çalışmalar özetlenmiştir. Ayrıca SEKA Bolu Mües- sesesi'nde yapılan çalışmanın maksadı açıklanmış, de ney tesisatı- tanıtılmış, deney sonuçları değerlendirilmiş ve fabrika atıksularını karakterize edebilecek kinetik katsayılar hesaplanmıştır» Beşinci bölümde lif levha atıksuları üzerinde yapılan deneysel çalışma ile elde edilen sonuçlar kullanılarak, bu atıksuların tam karışımlı ve piston akımlı reaktörlerde arıtılması haline ait nümerik uygulama yapılmıştır. Nümerik uygulamada özellikle tam karışımlı hal ile Lawrence-McCarty modeli ve yeni mo delle dizayn esasları üzerinde durulmuş ve bu üç mo delden elde edilen sonuçlar mukaeyese edilerek yorum lanmış tır o Altıncı bölümde bu çalışmada elde edilen sonuç lar Özetlenmiştir. XI

SUMMARY INVESTIGATION OF THE EFFECT OF FLOW CONDITIONS ON THE BIOLOGICAL REACTOR EFFICIENCIES It is necessary for a satisfactory removal of organics in polluted water that microorganisms should be in conctact with substrate under favourable environmental conditions for a sufficient period of time. Biological treatment is based on the biochemi cal reactions. These reactions can be only accomp lished by contacting the organic compounds with microorganisms for a sufficient duration,. Therefore, tehere is a strong relation between the' efficiency of biological treatment and flow regimes ' in the reactor, In this studj, two extreme conditions de fining the flow regime in biological reactor, have been evaluated and a new model has been devaloped. Then, the new model was compared with the model by Lawrence and McCarty. In addition, the origin and characteristics of the highly polluted wastes- water from a hardboard industry have been analyzed by activated sludge process as well as its treati- bility and kinetics. Later, considering the results obtained from the experimental studies, a numerical application has been presented. In the first chapter, a general introduction to the studied subject has been presented. Then a literature review on the works related to the ideal plug flow models has been given and the aim of the study has been explained. In the second chapter, the flow regimes ac cording to activated sludge systems have been introduced, the expressions defining the microbiolo gical principals of reactions occured in bioreactors have been summarized and the equations wich cha racterize continious flow - stirred tank reactors have been obtained by using mass balance principle for the first order reaction model and Monod kinetics. XllContinuous, flow-stirred tank reactors in series has been investigated for the first order and monod kinetics one by and the besic equations have been presented. The continuous flow - stirred reactors with and without sludge recirculation have been summarized. The relationships between (6 /0), (X /X) and r have been demonstrated, since recircu lation ratio and the concentration of VSS in re turn sludge are accepted as control parameters, by using this graphs it is possible to define the recirculation ratios for both continuous flow - stirred tank and plug flow reactors realistically. Some activated sludge modifications have been designed as the plug flow model. In this study the plug flow models which are generally accepted in the literature have also been investigated and the. basic equations of three models have- been compared. Methods by Sundstrom and Klei and Methods by Lawrence and McCarty which are employed the Monod equation were considered after the basic equ ation related to the first order kinetics for plug- flow reactors had been obtained. The Methods by Sundstrom and Klei gives the basic equations depen ding on the length of the reactor. The Method by Lawrence and McCarty is an appoximate solution technic which gives rather reasonable results espe cially at high values of 0 /0 to ratios. In the third chapter after giving the princip les of the new model develd-ped in this study, derivation of the relationships for the estimation of the concentration of biomass and that of orga- nics in the effluent have been shown by using two mass balance equations. Since the biomass concentration increase towards to the outlet of the reactor according to the new model, a new defini tion for the sludge retention time age has been used. Basic equations of the new developed model have been given below. Y(S -S)fX i + k.e d XIIIKB[(it-k.e)-(x/xB)] s1 d g s` ö 5S - - - - - - - - - -^- - - - - ^ - #o««»««»o»»o»»»»o»oooooo»o« /^y (y.-k )0-l+(X /Xa) ma g s.(...? X.).9 Q - 1`1 ^ ( `I) c X n o X. (1+r-r - - ) n X n These equations depend on the inflow and outflow biomass and organics concentrations unlike- the classical expressions, and by their very, general structure they explaines the equations of continu ous flow stirred reactors without recirculation. However, the parameter of S has been entered to the equations as a new unknown parameter in ad dition to X and S. The solution of these equa tion can only be obtained by employing computers due to two equations and there unknowns. The technics for solution of these equations have been given in chapter 3. 3. and a computer program listing in Appendics 3.1. developed for this purpo se. This program estimates the sludge age in addition to X and S in the outflow of the reactor. The principle of the new model is based on the assumption which defines the plug flow reactor as a combination of theorotically infinite but pratically finite number of continuous - flow stirred reactors. The equations (1) and (2) consider only a part of the reactor and the solution is obtained by applying these two equations to each part of the reactor. Namely the input values of X and S for any finite part of the reactor art the ofitput values of the previous part. However, it is necessary to decide what number of continuous-flow stirred reactors are employed to be able to ac complish plug flow conditions. Computer pogram has been developed for this purpose and the effects or the number of the reactor in the results have been monitored for a series of parameters. The XlVnumber of continuous flow stirred tank reactors which may characterize the plug flow conditions have been determined by using dimensionless para meters» When a 50 continuous flow stirred tank reactors are used the results differ mostly 3 % from 500 00 continuous flow stirred tank reactors. In practice 50 or 100 number of continuous-flow stirred tank reactors are consider to characterize the plug flow reactor, and the effects of number of the reactors on the results have been mostly 3 % and 1,5 % respectively. In this study 1000 number of continuous-flow stirred tank reactor have been selected» The new model developed in this study has been compared with the model by Lawrence and McCarty which is a very respective study in lite rature. The same parameters have been used in this comparison. As a result of this comparison it has been shown that the new model has a. general vali dity not only for the the range which requires in the model by Lawrence and McCarty but also for the most general cases in which the value of (0 /0) greather than 20. Both of the models have the same results for these ratios between 5 and 20. In the case (9 /0) less than 5 the Law rence-McCarty ' s model is out of the acceptable ranges'. Even in Lawrence-McCarty ' s models the concentration of pollutants may be higher than that of the raw wastewater shown that the steady state results may not be obtained by using the Lawrence-McCarty * s model for the high strength was tewaters. Leside this the results obtained from new model are within the acceptable range. Accoording to the new model, some computers programs listing in appendics 3.5 and 3.6 establis hed to estimate the detention time in plug reac tors and the. volume of the reactor. In the fourth chapter some informations on the production methods of hardboard industry have been given and the origins, characterictics and quantiny of wastewater as well as the previus studies have been summarized. Then the aim of the experimental study which is carried out in SEKA Bolu Factory has been explanined, the experimental XVinstallation has been introduced, the values of ki netics coefficients which characterized the wastewa ter from hardboard industry have been determined,, Experimental studies have been carried out both in pilot and laboratory scale model reactors. Kinetics coefficients were estimated from the studies consi dering the continuous flow and recirculation models. The respiration activity of wastewaters from the hardboard industry is determined is as 39 mg/lt.h0 In the experimental research, the samples taken from the main collector of the factory for the period of three months have been examined and COD has been determined as 6680- 500 mg/lt. Kinetic- coefficients have been predicted as k=1.9 day. By comparing with municipal wastewater it has been found that `k` coefficient is low and K coeffici- «.. s ent is very high0 Other kinetxcs parameters have been predicted as Ks = 2885 mgCOD/lt, Y = 0.40 and,kd = 0o24 day-1» Therefore, this characteristics have shown that there are some difficulties on biodegradab ility of wastewater from hardboard industry and longer hyd raulic detention times required. The larger value of `kj` coefficient is interpreted as the presence of colloidal and organic matter which are not decomposed easily by the microorganisms. It has also shown that to provide a COD removal rate appoximately 80 %, it is required to choose the F/M value less then û`75 day and the sludge age greather than 30 days. In the fifth chapter, numerical aplication to the treatment of wastewater from hardboard in dustry has been made by using the results obtained from the experimental study. In the numerical app lication, the desing principles have been establis hed by considering especially Lawrence-McCarty model, and the new model, developed in this thesis for the continuous, flow stirred tank models,, Then, evaluation has been made by comparing the results obtained from these three models. The desing prin ciples have been given in the case of continuous- flow stirred reactors and for this purpose some graphs have been prepared. In this application, the new model has given smaller value for the volume required than the Lawrence-McCarty ' s models for a contunuous flow stirred tank. XVIIn the sixth chapter, the results achived in this study have been given. XVÎÎ