Heterogeneous modeling, identification and simulation of activated sludge processes.

Author:Mustafa, Ibrahim H.
Position::Report
 
FREE EXCERPT

INTRODUCTION

Activated sludge is a complex dynamic process and simulation of such systems must necessarily account for a large number of reactions between a large numbers of components. There is a need for simulation models that describe the dynamic behavior of such important process process. Simulation models of the activated sludge process are believed to be useful tools for research, process, optimization and troubleshooting at full-scale treatment plants, in addition to serving a teaching and design assistance tools. However, the application of the models in most treatment plants is limited due to a lack of advanced input parameters values required by the models. To improve the operating efficiencies of current wastewater treatment plants, both municipal and industrial engineers have looked at automatic process control. This work is an extension of our previous work Ibrahim Mustafa et al.(1).

Figure 1 shows a schematic diagram of the activated sludge process where aeration basins (reactors) are typically open tanks containing equipment to provide aeration and to provide sufficient mixing energy to keep the MLSS in suspension. The depth is mainly determined by energy transfer/mixing characteristics and usually ranges from 3-7.5 m (2). A single piece of equipment such as diffused air, mechanical surface aerator, or jet aerator is used in many cases to provide aeration and keep the solids in suspension. Auxiliary mechanical mixers are used when the aeration does not provide sufficient mixing energy.

[FIGURE 1 OMITTED]

The secondary clarifier performs two functions in the activated sludge process. The first function, clarification, is the separation of MLSS from the treated wastewater to produce a clarified effluent that meets the effluent suspended solids goal. The other is the thickening of sludge to be recycled to the bioreactor. Since both functions are affected by clarifier depth, the design depth must be selected to provide an adequate volume for both functions (1), (3). For instance, the volume must be sufficient to store the solids during periods of high flow.

The objectives of this study are:

* To build a process model considering mass transfer limitations and simulates an Egyptian plant: Helwan wastewater treatment plant that exists in the south of Cairo and has a capacity of 350,000 [m.sup.3] [day.sup.-1] and average removal efficiency of 85% for substrate and 62% for ammonia. To be more sure of the simulation results, the model validation was performed for Zenine wastewater treatment plant that exists in the west of Cairo and has a capacity of 330000 [m.sup.3] [day.sup.-1] and average removal efficiency of 87.6% for substrate

* To adjust the model kinetic parameters of the biochemical reactions of the three growth processes: Carbon oxidation, nitrification and denitrification under the effect of mass transfer conditions to be prepared for the simulation purpose

* To Study the effect of the operating conditions such as flow rate, recycle ratio and feed substrate concentrations on the removal efficiency of both substrate and ammonia

Activated sludge process model development: The key to successful modeling of the activated sludge process is the appropriate assumptions to achieve a compromise between complexity and utility. The present study is concerned with the general derivation of a dynamic model of the activated sludge process in the bioreactor. The bioreactor (aeration basin) model describes the removal of organic matter, nitrification and denitrification. The bioreactor model is based on the previous biofloc model which depends on activated sludge model number one, ASM1, by considering both the mass transfer limitations and biochemical process reactions (3). The simulation model considers the four main components: BOD readily biodegradable Substrate (S), ammonia (H), Nitrate (Z) and oxygen (C). The assumptions in the biofloc model are considered here in the process model in addition to the following assumptions:

* The power input used in the bioreactor is assumed to be 80% of the maximum value to realize complete mixing in the reactor

* The effluent biomass concentration is be neglected,

* The consumption of substrate, ammonia and oxygen in the settler is neglected

* Constant average volumetric flow rate of the influent is considered

* Constant average recycle ratio and wastage ratio is considered

* The details of the process model developed appear in the schematic diagram shown in Fig. 1

Derivation of the process model: The derivation of mass balances on the settler and the bioreactor is considered. The derivation of mass balance equations of substrate is considered by applying mass balance on the settler in order to get the biomass concentration exiting which is recycled to the bioreactor; then mass balance equations of the bioreactor will be derived.

Applying a component mass balance on biomass for the settler gives:

Q (1 + R) X = Q (R + W) [X.sub.r] + [Q.sub.e][X.sub.e] (1)

Hence, the effluent biomass concentration is neglected, [X.sub.e] = 0.0.

Then:

[X.sub.r] = ([1 + R/R + W]) X (2)

By performing mass balance on the reactor, the following equations are obtained:

Mass balance on substrate (S): If substrate consumption in the settling tank is neglected and there is substantial decrease in the water content of the settled sludge related to that measured so that:

[S.sub.before settler] = [S.sub.after settler]

Applying a component mass balance on substrate for the bioreactor gives:

Substrate Inflow = Outflow + Net growth + Accumulation:

Q [S.sub.f] + RQ [S.sub.b] = Q(1 + R) [S.sub.b] + [K.sub.gs] [A.sub.tf] ([S.sub.b] - [S.sub.s]) + V [d[S.sub.b]/dt] (3)

[V/Q] [d[S.sub.b]/dt] = [S.sub.f] - [S.sub.b] - [[K.sub.gs][A.sub.tf]/Q] ([s.sub.b] - [S.sub.s]) (4)

Mass balance on ammonia (H): Ammonia nitrogen can be removed from wastewater by volatilization of gaseous ammonia. Gas stripping is most effective when contaminated wastewater is exposed to free air. Hence, this process is considered by adding a factor of ammonia stripping [G.sub.f] to the first order differential equation of ammonia.

Applying a component mass balance on ammonia for the bioreactor gives:

[NH.sub.3] accumulated = [NH.sub.3] Inflow-[NH.sub.3] Outflow-[NH.sub.3]

Volatilized by air stripping:

[V/Q] [d[H.sub.b]/dt] = [H.sub.f] - [H.sub.b] - [G.sub.f] [[K.sub.gh][A.sub.tf]/Q] ([H.sub.b] - [H.sub.s] (5)

Mass balance on nitrate (Z): Applying a component mass balance on nitrate for the bioreactor gives:

Nitrate Inflow = Nitrate Outflow + Net growth of Nitrate + Nitrate Accumulation

[V/Q] [d[Z.sub.b]/dt] = [Z.sub.b] - [Z.sub.b] - [[K.sub.gz][A.sub.tf]/Q] ([Z.sub.b] - [Z.sub.s]) (6)

Mass balance on oxygen (C): Applying a component mass balance on oxygen for the bioreactor gives:

Oxygen Inflow = Oxygen Outflow + Net growth of Oxygen + Oxygen Accumulation

[V/Q] [d[C.sub.b]/dt] = [c.sub.f] + [V/Q] [K.sub.1]a(c* - [C.sub.b]) - [C.sub.b] - [[K.sub.gt][A.sub.tf]/Q] ([C.sub.b] - [C.sub.s]) (7)

Mass balance on biomass (X): Applying a component mass balance on biomass for the bioreactor gives:

Input = Output + Rate of Reaction + Accumulation

Hence:

[QX.sub.f] + RQ ([1 + R/R + W]) x = Q(1 + R) X - pr.sub.x] V + V[dx/dt] (8)

Hence:

[V/Q] [dX/dt] = [X.sub.f] + [V/Q] [r.sub.x] + [W(1 + R)/(W + R)] X (9)

From task group (4) the rate of reaction of heterotrophic and autotrophic biomass can be obtained as follows:

[r.sub.x] = [[mu].sub.H] ([[S.sub.b]/[K.sub.s] + [S.sub.b]]) ([[C.sub.b]/[K.sub.oH] + [S.sub.Cb]]) + [[eta].sub.g[[mu].sub.H] ([[S.sub.b]/[K.sub.s] + [S.sub.b]]) ([[K.sub.c]/[K.sub.c] + [C.sub.b]]) ([[Z.sub.b]/[K.sub.z] + [S.sub.b]]) + [[mu].sub.A] ([[H.sub.b]/[K.sub.h] + [H.sub.b]]) ([[C.sub.b]/[K.sub.cA] + [C.sub.b]]) - [b.sub.H] - [b.sub.A] (10)

These dynamic model equations are first order differential equations and can be solved by the following technique.

Solution technique: The initial value problems given by the above equations are solved by Gear's method for multiple ordinary differential equations (1985). A computer program was written for this purpose. A large number of data points have been taken to improve the accuracy of the results when the parameters in the model were being estimated.

Results and Discussion: Selecting suitable kinetic and stoichiometric parameters is considered by using Helwan WWTP data. It covers also the model testing by carrying out the simulation on Helwan WWTP on substrate BOD and ammonia concentrations. Zenine WWTP was used for the testing and validation of the process model through predicting the response of substrate only as will be shown.

Parameters Evaluation: Shieh and Mulcahy (5) used an experimental procedure for the determination of intrinsic kinetic coefficients. The experimental apparatus (rotating disk biofilm reactor) provides a relatively simple, yet rigorous means for examination of both intrinsic and mass transfer limited kinetics. It allows for direct measurement of intrinsic kinetic coefficients and biological parameters relevant to a given reaction. The intrinsic kinetic coefficient of biological denitrification measured in their study is [K.sub.z] = 2.875 mg NO3-[N.sub.2] [day.sup.-1] and nitrate-nitrogen effective diffusivity is De = 0.815X[10.sup.-5] [cm.sup.2]...

To continue reading

FREE SIGN UP