Aktif çamurda çözünmüş kalıcı ürün oluşumu modeli

Abstract

Bu çalışmada aktif çamur sistemlerinde çözünmüş ürün oluşumu mekanizmasını da içeren bir matematik model geliş tirilmiştir. Birinci bölümde, aktif çamur sistemlerinin tasarım ve işletilmesinde model kullanmanın önemi vurgulanmış, kon- vansiyonel modellerin özellikle çıkış KOI konsantrasyonunun bulunmasında yetersiz kalması nedeniyle yeni bir model ge liştirilmesi gereği ortaya konmuştur. Çalışmanın amaç ve kapsamı açıklanmıştır. ikinci bölümde, aerobik arıtmanın temel prensipleri ele alınmış ve aktif çamur kinetiği ve stokiometrisi hak kında daha önce yapılmış çalışmalar değerlendirilmiştir. Bu bölümde özellikle aktif çamurda çözünmüş ürün oluşumu süreci üzerinde durulmuş, ürünlerin yapıları ayrıştırıla- bilirlikleri araştırılmış ve mikrobiyoloji literatüründe ürün oluşumu için verilen kinetik ifadelerin aktif çamura uygulanabilirliği tartışılmıştır. Üçüncü bölümde, aktif çamur modelleri tarihsel ge lişimi içinde ele alınmış ve her modelin vurguladığı ve ihmal ettiği konular açıklanmıştır. Dördüncü bölümde,. aktif çamurda substrat ve biyokit- le ölçümünde kullanılan parametrelerin anlamı üzerinde durulmuş ve bunların model değişkenleri olan büyüklükleri ne derece yansıttığı tartışılmıştır. Aktif. çamur modelle rinde kullanılan kinetik katsayıların hesaplanma yöntemleri ele alınmış ve bulunan sabitlerin ölçüm yöntemlerinden nasıl etkilendiği araştırılmıştır. Beşinci bölümde, bu çalışmada geliştirilen "çözünmüş kalıcı ürün oluşumu" modeli anlatılmış ve çeşitli tipteki reaktörler için belirlenen model denklemlerinin çözümü verilmiştir. Altıncı bölümde, model yardımı ile elde edilen so nuçların bir değerlendirmesi yapılmış ve son bölümde çalış mada elde edilen sonuçlar özetlenmiştir.

The activated sludge process, which dates back to the dawn of wastewater treatment, still possesses a central place in wastewater engineering. The design of activated sludge systems, like other engineering systems, call for the utili zation of models. Design of an appropriate activated sludge system for a given wastewater/treatment goal combination, or assessment of probable behaviors of a number of design alter natives under various operational, conditions are also realized through the use of mathematical models. The mathematical model used can be an empirical model obtained by statistical evaluation of data from physical models such as bench or pilot scale reactors, as well as a mechanistic model based upon the fundamental theoretical information concerning the prominent phenomena within the system. Generally speaking, mechanistic models have an obvious advantage over the empirical models, in that the latter may be expected to predict the system behavior with some accuracy, only within the range of conditions under which the data for model construction were collected. Due to restrictions on time and money, it is practi cally impossible to experiment and evaluate all possible solutions with physical models. This necessitates the use of mechanistic mathematical models which afford a quick and inexpensive means of predicting system behavior and assessing various design alternatives under a wide range of operational conditions, and thus singling out a few alternatives which appear most suitable. These selected alternatives can be subjected to further experimentation with physical models. - xiv Predominantly mechanistic models, assuming a direct relationship betwen growth and substrate removal, and employing the concept of rate limiting substrate have been used exten sively since the early 70's. The common conceptual structure underlying these models is that growth is exclusively supported by the original organic substrate in the influent and that the residual organic material in the effluent is the original substrate itself. This conceptualization leads to the prac tical result that the rate limiting substrate concentration in the effluent from a competely mixed reactor is independent of the influent concentration. However, numerous observations on pilot and full scale plants have clearly shown that the influent substrate concentrations do influence the effluent quality, both measured as COD or TOC. Difficulties associated with the measurement of biomass and the organic content of wastewaters have traditionally caused much inconvenience and confusion perhaps to the extent of causing some hindrance in the development of the activated sludge theory. The routinely used collective analytical para meters (e.g. SS, VSS, BOD, COD, TOC) do not correspond to the system variables (i.e. biomass and substrate) of the Monod- Herbert type models. This lack of correspondence is one of the important factors which obscure the validity of model results. BOD, COD, and TOC are the three extensively employed collective parameters for.the measurement of the organic content of wastes. Among these, the COD is the only parameter which enables making materials balance among substrate, bio mass and oxygen consumption in terms of electron equivalents. Another merit of the COD is its being both cheap and quick. With those advantages, COD has an ever increasing importance in academic studies, in the design and operation of treatment plants, and as a control parameter in quality control activities by the authorities. The adoption of COD, by the authorities, as a quality limiting parameter for the secondary effluents make utilization of COD orientated models obligatory i.e. models in which all phenomena having a significant effect on the variation of COD within the system are represented by suitable rate equations. A mechanistic mathematical model has been developed xn rnis study, which is designed to predict the soluble CUD concentration in the activated sludge plant effluent. In the course of model development, the system components and phe nomena associated with organic matter removal and their inter actions have been revised with specific emphasis on soluble inert product generation. - xv - Soluble inert product generation by the activated sludge systems, although evidenced or implied by the data from almost all relevant experimental studies, has been some what ignored and definitely not regarded as a phenomenon worth being represented in activated sludge models. However, as clearly reported by several workers, particularly in the last two decades, the organic substances in the activated sludge effluents are totally different from those in the influent, and are mainly metabolic products of the biomass. This fact, to a great extent explains the inability of a whole genus of models, having as their backbone the rate limiting substrate concept, in simulating activated sludge system behavior. For theoretical accuracy and consequentially for better prediction of activated sludge behavior, it seems necessary to incorporate inert product generation into the models. This necessity becomes more pronounced with increasing adoption of COD as a control parameter. The model developed incorporates six system components and three processes. The components are soluble substrate (Ss), active biomass (X^). particular substrate (Xg), soluble inert product (Sp), particular inert product (Xp), and oxygen (Sq ). The processes are growth, decay (loss of activity) and hydrolysis. The substrate removal and inert product formation processes are expressed in terms of associated processes employing suitable stoichiometric coefficients. Methods have been proposed for the measurement of model components and parameters. In the model, distinction has been made between soluble and particular substrate. According to the so called bisub- strate models, the soluble substrates can readily be utilized for growth whereas the particular substrate has to undergo a prior hydrolysis step before it becomes available for growth. The utilization of soluble substrate is expressed in Monod kinetics, with S denoting the bulk concentration. Utilization of particular substrate is expressed following the above men tioned hydrolysis hypothesis of the IAWPRC Task Group. Loss of activity and biomass are modeled in accordance with the death-regeneration concept instead of that the endogenous respiration. Activity loss of biomass leads to conversion to particular substrate and then to soluble sub strate via hydrolysis. The rate of hydrolysis is expressed by a saturation type expression. Oxygen consumption occurs only during growth. Oxygen consumed by regrowth on secondary substrate corresponds to - xvx - endogenous oxygen consumption of the endogenous respiration approach. The release into the bulk solution of numerous organic substances with varying biodegradabilities, by the micro organisms, is represented in the model by a hydrolysis rate expression. Although quite a many of the hydrolysis products are biodegradable, some are practically non biodegradable within the retention times normally employed in secondary treatment plants. Thus, soluble inert product formation is assumed in the model to be directly proportional to hydrolysis. Alternatively, a different version of the model has been constructed, with the soluble inert product generation being proportional to growth, and consequences of two different assumptions on the model results has been evaluated. Sensitivity analyses of the model parameters have been made. Several sets of simulations have been made for different reactor types, and their efficiencies defined according to various criteria have been compared. The results are sum marized in Chapter VII. The greater part of the soluble organic materials in the effluent from an activated sludge plant are microbial products,<*and hence they differ in character from the organic materials in the influent. Since the product formation has a pronounced effect on the effluent substrate concentration, Monad-Herbert models which. merely represent the substrate removal mechanism can only inaccurately predict the substrate removal efficiencies. The soluble microbial products show a wide spectrum of structures and molecular weights. Intermediary products of energy metabolism are low molecular weight compounds, and this type of products formation is not effective in continuous systems, not considering delays due to causes such as bio- synthetic restrictions. Most of the hydrolysis products are also readily biodegradable. Products which are not degraded within the range of detention times normally employed in activated sludge systems are considered to be biologically inert, and designated as inert soluble product (Sp). The Sp production is modeled in two alternate approaches, viz. as being hydrolysis related and growth related. An activated sludge model is formulated, taking into account the soluble inert product formation, and the model is employed to assess the relative significance of variables - xvii - that directly affect the effluent soluble COD concentration. It is found that the Sp formation is proportional to the active biomass in the reactor and is therefore related with influent soluble substrate concentration (Sgo) an^ biological solids retention time 9q, which in turn determine the active biomass concentration. The inert soluble product formation model explains the dependence of the effluent COD concentration on the influent COD concentration in a completely mixed reactor. Theoretically, the effluent COD concentration is independent of the influent concentration, however, since the inert product formation is influenced by the influent substrate concentration, this in turn influences the effluent soluble COD concentration. For every influent substrate concentration there 0is an optimum biological solids retention time which results in a minimum effluent soluble COD. This implies that the sludge age should not be held unnecessarily long, provided that the settling properties are also given due consideration. A completely mixed reactor can be more efficient than a plug flow reactor, when efficiency is defined in terms of soluble COD removal. However, a plug flow reactor is more efficient when the oxygen demand of the wasted sludge. is also taken into account.