Search

A stochastic model is employed to account for the dynamic behaviour of floc breakage. More specifically, the evolution of the particle-size distribution over time is presented. The model is compared to experimental data obtained by exposing kaolin—iron(III) flocs to hydrodynamic stress generated by a turbine impeller. The agreement between the model and experimental data appears to be satisfactory.

A combined model consisting of an active region and a stagnant region is used in representing the air flow characteristics in a continuous flow system. The active region is represented by the one dimensional axial dispersion model while the stagnant region is considered to be a dead volume having no fluid interchange with the active region. The mathematical model is fitted to experimental data obtained by subjecting the system to tracer step inputs. In the process of model fitting a dimensionless parameter of Pe. (Peclet number of the active region) is estimated. The effects of the internal geometry and structure of the flow system on the flow patterns are analyzed and substantiated by experimental data. Two possible mechanisms explaining the formation of stagnant regions have been proposed — that of reduced length, and that of reduced cross-section. For the present investigation, the reduced length mechanism is found to explain the experimental data better. On utilise un modele combine, forme d'une region active et d'une region stagnante, pour representer les caracteristiques de l'ecoulement de l'air dans un systeme d'ecoulement continu. Le modele de dispersion axiale a une dimension represente la region active, tandis qu'on considere la region stagnante comme un volume inerte dans lequel il ne se fait aucun echange mutuel de fluide avec la region active. On adapte le modele mathematique aux resultats experimentaux obtenus en introduisant un indicateur (ou traceur) par etapes. Dans le processus d'adaptation du modele, on evalue un parametre Pea sans dimensions (nombre de Peclet de la region active). On analyse les effets de la geometrie interne et de la structure du systeme d'ecoulement sur l'allure de l'ecoulement et les verifie experimentalement. On a propose deux mecanismes possibles pour expliquer la formation des regions stagnantes, a savoir le mecanisme de longueur reduite et celui d'une coupe tranversale; dans le cas du present travail, on trouve que le mecanisme de longueur reduite explique mieux les resultats experimentaux.

CORN dust was gasified with steam in a 5 cm (2 in.) I.D. bench scale fluid-bed reactor. Experiments were conducted over a temperature range of 867 K to 1033 K (1100 °F-1400 °F) to examine the influence of tempera-ture on the gas composition, yield, heating value and the energy recovery. The gas yield was found to be strongly temperature dependent and increased linearly from 0.13 mVkg (2.0 SCF/lb) at a 867 K to 0.73 mVkg (11.7 SCF/lb) at 1033 K. The produced gas heating value ranged from 9.4 MJ/m3 (251 Btu/SCF) at 867 K to 11.5 MJ/m3 (308 Btu/SCF) at 1033 K with maximum value of 11.6 MJ/m3 (311 Btu/SCF) at 1002 K. The principal components of the product gas were H2, C02 and CO which comprised over 90 percent of the gas. The remain-ing 10 percent consisted of CH4, C2H4 and C2H6. The energy recovery increased linearly from 8 to 55 percent over the temperature range studied.

The devolatilization reactions of feedlot manure were studied by a thermogravimetric analyzer. The effects of size fraction, heating rate, and purge gas rate on the thermogram characteristics, such as the total weight percent devolatilized and the temperature at which the maximum rate of weight loss occurred, were analyzed statistically. The total weight percent devolatilized was mainly influenced by the size fraction, while the temperature at which the maximum rate of devolatilization occurred was mainly affected by the heating rate. The devolatilization data were fitted to a multireaction model, and the kinetic parameters were determined. The mean value of the activation energy, E, and the standard deviation of the activation energy, sigma, were analyzed statistically to assess their dependence on the heating rate, size fraction, and purge gas rate. E was not affected by any of these variables, while sigma was affected by the heating rate and the purge gas rate.

A geometrical and microscopic mixing index defined by the contact number was used to study the transverse mixing in a Kenics motionless mixer. The mixing index, a measure of transverse mixing, increased exponentially as the number of helices in the motionless mixer increased. The helices in the mixer also had the significant effect of reducing the void fraction of the mixture. The relationship between the coordination number and the compaction in the mixture through the mixer was studied. The mean coordination number indicates that the packing of these mixtures are between the cubic and hexagonal packings.

A modified dynamic method is introduced to determine the oxygen transfer coefficient, KL a, in aerobic fermentation systems which are not mechanically agitated. The dissolved oxygen concentration is measured continuously following a step down or a step up in aeration rate. The response curve is analyzed to obtain the value of KLa Experiments were carried out at several different air flow rates using mixed culture in concurrent tower fermentors with motionless mixers. The effect of sieve trays and Koch motionless mixers on oxygen transfer was investigated using a 3 in. diameter column. The values of KL aobtained at the bottom of each column were found to be higher than those obtained at the top. Comparison of the results showed that the values ofKL a were higher when the Koch mixers were used than when the sieve trays were employed. The oxygen uptake rate by the organisms rX, is also calculated by using the KL a values obtained. They compare favorably withthe experimentally measured values.

In this work dispersive mixing and chemical reactions are treated simultaneously by resorting to the theory of stochastic processes. A fluidized-bed reactor is modeled by discretizing it into ideally stirred tanks of various sizes corresponding to bubble, cloud, and emulsion phases. All parameters in the model are correlated with known or experimentally obtainable quantities. Examples using a complex chemical reaction are given to demonstrate the applicability of the approach.

Wind velocities are often nonstationary with an almost cyclical repetition of strong gusts followed by relative calmness. It was demonstrated that the effects of nonstationary wind velocity on soil erosion are significant, and that a transient model is necessary to estimate rates of wind erosion. /Author/

A stochastic approach, namely, a Markov chain, has been adopted to simulate the dynamics of complex reactions in a flow chemical reactor without either solving directly differential equations governing the performance of the reactor or obtaining a closed form solution to such equations. All calculations have been carried out iteratively with the aid of a computer. The results are in good agreement with the known results obtained from the deterministic approach. The present technique can be applied to a variety of flow reactors with complex chemical reactions, which cannot be handled easily by deterministic approaches.

A transient analysis of a fluid-solid pyrolysis system in which a solid is pytolyzed to form a fluid product and a solid product is presented. The analysis, based on utilization of the so-called volume reaction model, takes into account the simultaneous heat and mass transfer phenomena in the particle being pyrolyzed. Variations of physical parameters with respect to temperature and/or solid concentration are not neglected. Effects of the Lewis number on the solid conversion, the fluid and solid concentration distributions, and the temperature profile in the particle are examined. The results are graphically presented and interpreted. On presente une analyse transitoire d'un systeme de pyrolyse impliquant un fluide et un solide et dans lequel un solide est transforme par pyrolyse en deux produits, un fluide et un solide. Ladite analyse, qui est basee sur l'utilisation du modele a soi-disant reaction en volume, tient compte des phenomenes de transfert simultane de chaleur et de masse dans la particule soumise a la pyrolyse; on y considere egalement les variations des parametres physiques par rapport a la temperature et/ou a la concentration du solide. On examine les effets du nombre de Lewis sur la transformation du solide, les distributions des concentrations du fluide et du solide qui en resultent et du profil des temperatures dans la particule. On presente les resultats sous forme graphique et les interprete.

Mixing in a conventional solids mixer is governed mainly by two basic mechanisms, diffusion and convection. The characteristics of these processes have been investigated separately. Model mixers were constructed using these processes, and a new mixing system was synthesized by a proper combination of the two. Performances of the elementary model mixers, as well as the synthesized mixing system, were simulated on the computer, and the results of simulation were experimentally verified. The method of synthesis, which appears to be effective, can be extended to the design of other practical mixing systems.

The optimum degree of primary treatment in an activated sludge process is investigated in this study. In this investigation, a so-called parallel biochemical oxidation kinetic model, which takes into account the low metabolic rate and masking effect of suspended wastes, is employed to characterize the kinetics in the aeration chamber. The dynamic programming technique is used to determine the optimal primary clarifier detention time. The analysis shows that the optimal primary clarifier detention time depends on the concentration and masking characteristics of inlet suspended wastes, as well as the concentration of inlet soluble substrate.

A fairly general model of the biochemical oxidation, which takes into account the activity of microorganisms, is presented. Parameters of the model have been determined by fitting the model to available experimental data through the use of a straightforward gradient technique.

Free-radical polymerization in a flow reactor represented by the tanks-in-series model with backflow was considered. Conversions and molecular weight distributions were computed as functions of the backflow parameter, and the results were compared with the conversion and molecular weight distribution from a CSTR and those from a plug-flow reactor. Backflow was found to be undesirable for the polymerization mechanism under investigation. Values of the degree of segregation for the tanks-in-series model were calculated by using Zwietering's approach as a function of backflow.

Interaction of macromixing and micromixing in microbiological flow reactors based on a kinetic model of the Michaelis-Menten type has been treated by considering two general types of flow systems: the system of n continuous stirred-tank reactors in series (n-CSTR's in series) and the series combination of a plug-flow reactor and a continuous stirred-tank reactor. Systematic schemes to describe the micromixing conditions of a reactor system are presented. Additional micromixing states are considered besides those proposed by Kramers and those proposed by Zwietering if the systems are employed as empirical models of a flow system. When the number of stirred-reactor units in the system of n-CSTR's in series is small, micromixing has a significant effect on the growth processes. As the number of reactor units increases, the micromixing effect on the growth processes decreases. The effect of micromixing is also important for a system described by a series combination of a plug-flow reactor and a continuous stirred reactor. In addition, the exit concentrations from the system having the same residence time distribution and the same degree of segregation may be different from each other. Information obtained in this study is pertinent in the design of biological flow reactors and sewage treatment systems.

Mechanical and dynamical characteristics of semifluidized beds of single-size particles in solid-liquid system were investigated. A semifluidized bed is a type of fluidized bed in which the bed expansion is partially restricted. The emphasis was placed on the study of packed bed formation and pressure drop increase when the semifluidized beds were formed by the compression of fluidized beds. The data showed good agreement with the theoretical and semiempirical equations based on a simple model of fluidized beds. The results of this investigation would also contribute to the understanding of fluidized beds.

Distributed parameter systems under load disturbances regulatory control by feedforward feedback and state measures

Measurements of diffusion coefficients of liquids have been difficult and time consuming. A laser interferometric system with which diffusion coefficients of liquids may be measured with greater ease and accuracy than by the use of previously reported systems was developed. The system developed was tested by measuring the diffusion coefficients of benzene in cyclohexane. A helium—neon gas laser was used as a light source. Its use permitted accurate computation of instantaneous profiles of concentration vs position in the diffusion cell of the system. These profiles were derived from photographs of interference fringe lines taken during experimental runs.

An optimal waste discharge policy along a stream is investigated based on a program of minimizing the total cost benefit function on a regional basis. The system consists of a stream along which are located several inputs of waste and intakes of water. The cost benefit function comprises of the sum of the cost of waste treatment and the cost of water treatment minus the benefits derived from maintaining a certain water quality in the stream. The biochemical oxygen demand and the dissolved oxygen are considered to be the primary indicators of water quality. A steady state dispersion model is employed to predict the BOD and DO concentrations along the stream. In order to optimize the system or minimize the cost benefit function within certain constraints on water quality, a search scheme is proposed. The feasibility of obtaining a solution by a sequential minimization technique, employing a discrete search, is studied for cases with two and three decision variables. The analysis procedure can be used to examine existing water quality programs and to study the influence of alternate policies and constraints.

A process governing the transport rate of cattle manure as measured by their COD (chemical oxygen demand) through a wet porous stratum is investigated under simulated conditions. A mathematical model of the system is presented. The model simulates diffusion and biological reaction processes taking place simultaneously in the system. Experimental observations have been made to determine the diffusion coefficient and the biological reaction rate constant of the cattle manure solution. Values of approximately 6.76 × 10−6 cm2 s−1 for the diffusion coefficient and 3.05 × 10−2 day−1 for the reaction rate constant at the temperature of 25 ± 2°C were obtained. For the system with known diffusion coefficient and reaction rate constant, the analytical expressions obtained here can be used to approximately evaluate or predict the rates that animal wastes, under conditions similar to those in this study, will contaminate surface water, soil or ground water.

Common practice for design of biological wastewater treatment systems involves evaluation of certain fixed structure systems at several alternative conditions. This study considered optimization of both the structure and the design of the system using the structural parameter system synthesis technique. The system was composed of a trickling filter, an activated sludge aeration vessel, and a secondary clarifier. Results of optimal synthesis indicated that the activated sludge system was the most cost effective system for the specific set of process parameters employed. The same conclusion was reached by optimizing several systems with different structures. La pratique courante, dans le cas de la conception des systemes pour le traitement biologique des eaux usagees, comprend l'evaluation de certains systemes a structure fixe dans plusieurs alternatives. Dans le travail present, on a etudie l'optimisation de la structure et de la conception du systeme en employant la methode de synthese d'un systeme a parametres structuraux. Le systeme comprenait tin filtre a degouttement, un recipient pour l'aeration de la boue activee et un clarificateur secondaire. Les resultats d'une synthese optimale ont indique que le systeme a boue activee etait le plus efficace au point de vue du cou't, dans le cas des parametres specifiques qu'on a employes. On en est venu a la me'me conclusion en optimisant plusieurs systemes de structures differentes.

Two progression models which can be employed to represent flow systems with varying degree and mode of mixing along the direction of flow such as a shallow fluidized bed, radial flow reactors, and short tubular reactors, etc. are proposed. Residence time distributions are derived from the standpoint of stochastic processes. The use of progression models is illustrated with a second order, isothermal reaction. On propose deux modeles de progression qu'on peut utiliser pour representer des systemes d'ecoulement qui possedent des degres et modes differents de melange suivant le plan de direction de l'ecoulement, tels que, par exemple, un lit fluidise et peu profond, des reacteurs a ecoulement radial et des reacteurs tubulaires de faible longueur. On deduit les distributions du temps de sejour moyen dans le reacteur, en se basant sur les methodes conjecturales et on illustre l'emploi des modeles de progression au moyen d'une reaction isothermique de second ordre.

For autocatalytic reactions such as those associated with many biochemical processes, the rate of a reaction is influenced by the concentrations of some of the products as well as those of the reactants. By use of the so-called diffusion (axial dispersion) model in conjunction with the Monod growth kinetics, it is shown that partial (incomplete) fluid mixing is often desirable for this class of reactions. Under sterile feed conditions, the optimal Peclet number appears to range between 0 and 2 with the exact value depending on the values of the kinetic constants and initial substrate concentration. By varying this single parameter, the performances of such reactors under various degrees of fluid mixing can be readily investigated using the axial dispersion model. The optimal design may be physically realized by the use of a tower fermentor with controlled axial dispersion. Dans des reactions auto-catalytiques comme celles qui se produisent dans nombre de procedes biochimiques, la vitesse de reaction est influencee par les concentrations de certains produits et des composes qui reagissent. En employant le soi-disant modele de diffusion (dispersion axiale) conjointement avec la cinetique de croissance de Monod, on voit qu'il est souvent desirable d'utiliser ce melange partiel ou incomplet de fluide pour cette categorie de reaction. Dans des conditions steriles d'alimentation, le nombre optimal de Peclet semble varier entre 0 et 2, la valeur exacte dependant des valeurs des constantes cinetiques et de la concentration initiale du materiel de base. En variant ce seul parametre, on peut facilement etudier, au moyen du modele de dispersion axiale, les rendements de reacteurs de ce genre lorsque le melange fluide se fait a des degres differents. On peut realiser physiquement le modele optimal en utilisant un dispositif a fermentation muni de tour et ou il y a dispersion axiale controlee.

In this work, a mathematical model which can be used to describe butch growth in fermentations with two liquid phases present is developed for systems in which the growth limiting substrate is dissolved in the dispersed phase. The model takes into account the drop size distribution, the rate of adsorption of cells on the drop surface, the rate of desorption of cells from the drop surface, substrate transport between phases, phase equilibrium, and growth kinetics. The model also considers the effect, of coalescence and redispersion of oil drops in the system. It is assumed that the composition of the dispersed phase is such that substrate utilization from it causes little or no change in the interfacial area. A discrete uniform distribution and a discrete normal distribution which is obtained from an experimental distribution curve are used as drop size distributions. Simulation results are obtained for a wide range of parameter values using the IBM S/360 Continuous System Modeling Program.

Unter Voraussetzung konstanter Warmestromdichte an der Wand werden die Temperaturprofile und Warmeubergangsparameter einer laminaren MHD-Stromung in der Einlaufpartie eines flachen Kanals berechnet. Geschwindigkeit und Temperatur werden im Eintrittsquerschnitt als konstant vorausgesetzt. Eine Kennzahl β wird eingefuhrt, die ein Kriterium fur den Einfluss der Zahigkeitskrafte ist und ahnlich definiert ist wie die Eckertsche Kennzahl. Numerische Ergebnisse werden angegeben fur den Wert 1 der Prandtlzahl und die Werte 4 und 10 der Hartmannzahl. Die genannte Kennzahl β wird dabei zwischen −1 und + 1, eine das elektrische Feld kennzeichnende Grosse zwischen 0,5 und 1,0 variiert.

A systematic procedure was developed for selecting the type of reactors or a reactor system configuration for polymerization reactions. Two different mechanisms were investigated, and the “best” reactor system to give the desired quality of the product was determined using a systems synthesis technique. The behavior of the system in the neighborhood of the optimal solution was explored, and the effect of variation in the rate constants and the initial concentrations of the catalyst and the monomer on the optimal reactor system was examined. Recycle streams were introduced and their effect on the system performance was investigated, and finally the applicability of the systems synthesis technique to other polymer reactor design problems was discussed.

The laminar boundary layers on a moving continuous flat surface in non-Newtonian fluids characterized by the power law model are investigated using exact and approximate methods. Both pseudoplastic and dilatant fluids are considered. Numerical solutions of the boundary-layer equations are obtained for values of the parameter n in the power law model ranging from 0.1 to 2.0. An integral solution of the momentum equation, which can be used to obtain values of the dimensionless shearing stress that are in good agreement with the exact values, is developed. An integral solution to the energy equation is also presented.

Eine Optimalisierungsmethode fur einen vielstufigen Prozess mit Ruckkopplung wird vorgeschlagen. Die Methode liefert speziell die Rechenvorschriften im Spezialfall eines eindimensionalen Prozesses. Als Beispiel wird ein Problem der Wasserverteilung behandelt.

The optimal startup policy of a jacketed tubular reactor, in which a first-order, reversible, exothermic reaction takes place, is presented. A distributed maximum principle is presented for determining weak necessary conditions for optimality of a diffusional distributed parameter system. A numerical technique is developed for practical implementation of the distributed maximum principle. This involves the sequential solution of the state and adjoint equations, in conjunction with a functional gradient technique for iteratively improving the control function.

Γ-ray attenuation method was applied to determine the axial fluidizing bulk density profile for an air-solid aggregative fluidized bed. Except for beds fluidized at very high air velocities two distinct density zones were noted. The density of the bed in the lower portion of the column was relatively constant for a particular set of operating conditions and did not vary with bed height. The density in the top portion of the bed fell rapidly with bed height. The average density of the lower zone was correlated as a function of the operating variables. The density in the top zone was represented by a one-tail normal distribution curve. A somewhat better correlation was obtained by a two-parameter equation at high air velocities. The effects of operational variables, such as air velocity, static bed height, particle size, and particle-size distribution on these correlations were investigated. Friction-factor correlations based on two different definitions of bed height were also obtained.

This study is to present an analysis of two-dimensional, steady-state, laminar, compressible flow of an electrical conducting fluid inside a flat duct with a transversely applied magnetic field. The boundary-layer equations for the compressible MHD flow are solved by a new approach based on Morris and Smith's work on non-MHD flow. The compressible boundarylayer equations including the continuity, momentum, and energy equations are reduced to ordinary differential equations which are solved simultaneously by numerical integration. No restrictions are imposed on the Prandtl number. A very general form of viscosity variation with respect to temperature is used in the development of the general equations. A sample problem with the constant wall temperature, the Mach number of 0.8, and the Prandtl number of 0.01 is solved numerically. Investigations include the growth of the dynamic boundary and thermal boundary layers, the development of the velocity and temperature profiles, and the skin friction and heat transfer. The results show that the skin friction and heat transfer are increased with the increase of magnetic field strength. To obtain an estimate of the validity of the method, it is applied to the solution of incompressible flow problems for which some reliable solutions are available.

An analytical solution is developed to predict the BOD and DO profiles in a stream having several outfalls of waste and intakes of water. The model also incorporates continuous addition and removal of BOD and DO along the stream. A steady state one dimensional dispersion model is used to describe the transport of BOD and DO in a stream. Changes in stream properties along the length are considered by dividing the stream into several segments and changing the parameters of the BOD and DO equations as needed in each segment. By applying continuity equations at the boundaries of these segments and equilibrium conditions at either ends of the infinitely long stream, the arbitrary constants in the solutions to the differential equations for BOD and DO can be determined. Some applications of the results in studying water pollution problems are illustrated.

The laminar, isothermal entrance region flow of the Bingham fluid in a circular pipe is studied at first by using the momentum integral method and the boundary-layer equation for the Bingham fluid. In addition to the velocity boundary layer, the existence of the shear stress boundary layer is considered. The solution is valid only near the entry because of limitation of the boundary-layer model. The Campbell-Slattery method originally devised for the Newtonian fluid is also used to analyze the Bingham entrance region flow. The results are compared with those obtained through the use of the momentum integral method and those obtained through the use of the variational method by other investigators. While the results obtained in this work appear to be generally reasonable and valid, the results by the other investigators have been found to be somewhat erroneous.

This paper reviews research in the area of polymerization systems engineering. Polymerization systems engineering is divided into six subareas: thermodynamics, modeling and simulation, optimal reactor design, dynamic optimization, stability analysis, and optimal synthesis. Basic concepts of polymerization are presented, and research in these subareas is reviewed. It is concluded that much remains to be done.

Evolutionary operation (EVOP) was used to experimentally investigate the optimum steady state operating conditions for a step aeration activated sludge waste treatment process. A laboratory scale two tank step aeration activated sludge unit with fixed total volume, total influent flow rate, recycle flow rate, and sludge wasting rate was employed. The volume ratio and flow rate ratio which minimized effluent chemical oxygen demand were determined. The results indicate that EVOP is a useful technique for improving the performance of biological processes.

The problem of estimating missing values in water quality data using linear interpolation and harmonic analysis is studied to see which one of these two methods yields better estimates for the missing values. The data used in this study consisted of midnight values of dissolved oxygen from the Ohio River collected over a period of one year at Stratton station. Various hypothetical cases of missing data are considered and the two methods of supplementing missing values are evaluated using statistical tests. The results indicate that when the percentage of missed data points exceeded ten percent of the total number in the original sample, harmonic analysis usually yielded better estimates for both the regularly and irregularly missed cases. For data that exhibit cyclic variation, examples of which are dissolved oxygen concentration and water temperature, harmonic analysis as a data generation technique appears to be superior to linear interpolation.

A consecutive, first-order, irreversible, biochemical reaction, , taking place in a series of N reactors with product recycle is considered. A discrete version of the maximum principle is used to derive general equations necessary for maximizing the production of (1) the final product, C, by choosing the temperature or the pH value in each reactor, and (2) the intermediate product, B, by choosing the reactor volume. A numerical computation for a series of three reactors with recycle is illustrated. The effects of varying the recycle rates on the optimal state and decision variables are also presented.

A model of highway air pollution is developed by considering macroscopic material balance of the polluted air mass over a segment of a highway that passes through an urban area. Problems of traffic...

Die Untersuchung befasst sich mit der Warmeubertragung von den Wanden eines flachwandigen Kanals auf eine elektrisch leitende Flussigkeit bei erzwungener Laminarstromung und in Gegenwart eines quergerichteten Magnetfeldes. Betrachtet wird der Fall konstanter Wandtemperatur mit variierender innerer Warmeentwicklung durch viskose und elektrische Energiedissipation. Die massgebende Differentialgleichung wird durch eine Differenzengleichung ersetzt und mit der elektronischen Rechenmaschine gelost. Als Resultat wird die Nusseltzahl angegeben, fur die Prandtlzahl 1, die Hartmannzahlen 0, 4, 10 und die Graetzzahlen von 10 bis 10 000, wobei die Kennzahlen fur Zahigkeit und elektrische Feldstarke als Parameter auftreten.

A modified transient version of the Streeter-Phelps model along with the energy balance equation is employed to analyze the effects of waste heat discharge from power plants on stream water quality. Analysis is also made to examine the effects of the upstream water quality and stream velocity on the downstream DO concentration level. The resulting coupled nonlinear hyperbolic partial differential equations representing the energy, BOD and DO concentrations are solved by the method of characteristics and simulated on a digital computer. Final numerical results indicate that the allowable quantity of thermal discharge does heavily depend on the upstream quality.

The kinetic behavior of heterogeneous microbial populations of sewage origin was studied in a single-stage, isothermal, continuous flow, completely mixed aeration tank. A series of experiments were carried out at various dilutions rates using glucose as the limiting substrate. The cell dry weight and substrate concentration in terms of chemical oxygen demand (COD) were continuously monitored. The results indicate that reproducible steady-state conditions can generally be obtained; however, multiple steady states were observed at dilution rates near washout. At low dilution rates (below about 0.1 hr−1) the contribution of microorganism decay became appreciable. Using the multiresponse data of cell dry weight and COD, the parameter values in various existing growth models were estimated. The analyses of variance and residuals revealed that models proposed by Moser, Monod, and Contois, each with a decay term added, were significantly better than the other models which were tested.