JPH05169089A - Nitrification and denitrification process simulator - Google Patents

Abstract

PURPOSE:To provide a simulator for supporting operation by estimating the optimum value (optimum manipulated variable) of respective control factors in consideration of such status that an operator must decide the manipulated variables of the respective control factors in a trial-and-error method because the number of the control factor is made plenty and the control factors interfere with each other. CONSTITUTION:In the case of constructing a process model part 12 having a processing process model part 14 and a controlling process model part 15, a simultaneous nonlinear differential equations showing the network of principal reaction of a processing process are replaced with a linear equation to construct the processing process model part 14. Repeated calculation is executed by the process model part 12 to perform numerical convergent calculation by a stationary analysis part 19. Thereby a stationary solution is obtained. Furthermore optimization of the operating conditions is enabled by changing the operating conditions by an arithmetric part 21 for optimizing the operating conditions and repeatedly calculating the stationary solution.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control device for an activated sludge process in sewage treatment, and more particularly to a simulator for obtaining an optimum value of a control factor for a nitrification denitrification process.

[0002]

2. Description of the Related Art Nitrogen compounds contained in wastewater [ammonia nitrogen (NH ₄ -N), nitrite nitrogen (NO ₂
N), organic nitrogen (ORN-N)] is discharged untreated from a sewage treatment facility into a river, the sea, a lake or the like, and it is known that various environmental problems occur. Therefore, various biological or physicochemical methods have been proposed as methods for removing nitrogen from the wastewater. Of these, the biological nitrification and denitrification process that applies the metabolic activity of microorganisms has become the mainstream of sewage treatment because it can be treated in large quantities at low cost and is relatively easy to maintain and manage. .. An overview of the biological nitrification denitrification process is shown in FIG. As shown in the figure, after the inflowing water is subjected to nitrification and denitrification treatment in the aeration tank 1, the final settling tank 2 finally sets the SS component
Remove. The sludge settled in the final settling tank 2 is discharged as excess sludge, and a part of the discharged sludge is returned to the aeration tank 1 as activated sludge. Here, the aeration tank 1 is divided into a denitrification tank (anoxic state) 3 and a nitrification tank (aerobic state) 4,
In that the nitrification liquid is circulated from the outlet 1b to the inlet 1a,
Different from the conventional standard activated sludge method.

FIG. 11 shows a control system of this circulation type nitrification denitrification process. The blower 5 aerates the nitrification tank 4 of the aeration tank 1, and a control valve 6 for adjusting the air flow rate.
Is controlled by the nitrification tank dissolved oxygen concentration control unit (DOC) 7. The amount of sludge discharged from the final settling tank 2 by the surplus sludge pump 8 is determined by the average sludge retention time control unit (SRT) 9
Controlled by. Reference numeral 10 is a returning sludge pump for returning sludge to the aeration tank 1, and 11 is a circulation pump for circulating treated water in the aeration tank 1.

In controlling the circulation pump 11, the nitrification tank dissolved oxygen (DO) concentration, sludge average residence time (SRT), circulating water amount (circulating ratio; circulating ratio to inflowing water amount), A / O are important controllable factors. Ratio (volume ratio of denitrification tank and nitrification tank) and the like. At present, these four items are all controlled to be constant. For example, DO = 2 (mg / l),
SRT = 10 (days), R2 (circulation ratio) = 2, A / O =
It is controlled 1: 1. Generally, each management target value is determined by the experience of the operator.

[0005]

The above-mentioned four control factors DO, SRT, R2 and A / O ratio respectively affect the nitrification reaction rate and the denitrification reaction rate. For example, in the case of the DO concentration, the nitrification reaction rate increases as the DO concentration increases, but the DO concentration in the circulating water and the DO concentration in the denitrification tank also increase, so the denitrification rate decreases. Therefore, when the DO concentration is increased, whether the TN removal rate finally increases or decreases depends on the increase range of the DO concentration and the D to nitrification denitrification.
It cannot be determined unconditionally because it changes depending on the strength of the O concentration limiting rate. Further, increasing the DO concentration may cause other control (operation) factors (SRT, R2, A / O ratio) to deviate from appropriate values.

[0006] As described above, this type of process is the same as the conventional B
Compared to the standard activated sludge method whose main purpose is to remove OD (biochemical oxygen demand), the number of control factors is large, and since these factors interfere with each other, each manipulated variable is determined by trial and error by the operator. The current situation is to decide.

In view of the above circumstances, the present invention provides an optimum value of various control factors (optimum operation) in controlling an activated sludge process in sewage treatment, particularly a treatment process in which anaerobic / aerobic treatment is combined to perform nitrification denitrification. The purpose is to provide a simulator that estimates driving amount and supports driving.

[0008]

In order to achieve the above object, the present invention is directed to a treatment process in which anaerobic / aerobic treatment is combined to perform nitrification / denitrification, and an optimum value of a control factor of this treatment process is set. As a nitrification / denitrification process simulator for determining, the following is provided.

(1) A process model part having a processing process model part and a control process model part, and the two model parts mutually perform calculations. The treatment process model section is a simulation of the treatment process by replacing the simultaneous nonlinear differential equations representing the main reaction network of the treatment process with a linear form. The control process model section imitates the processing process model section and the control algorithm.

(2) A steady solution of the simulated output of the treatment process is obtained from the set values of the treatment process state variable and the control factor, and the process model section is made to repeatedly perform calculations, and numerical results are converged using the calculation results. A steady analysis unit that performs calculation and uses the converged value as the steady solution.

(3) A control factor optimizing arithmetic unit for optimizing the control factor by repeatedly operating the steady state analyzing unit while changing the set value of the control factor.

[0012]

As described above, control factors interfere with each other in this type of processing process. Therefore, in the present invention, in simulating the processing process, a method of performing calculation using numerical convergence calculation is adopted. That is, when the main reaction network of the treatment process is mathematically modeled by an appropriate method, it is represented by simultaneous nonlinear differential equations. Since a steady solution cannot be obtained with this mathematical model, the simultaneous nonlinear differential equations described above are replaced with a linear form, repeated calculations are performed, and the numerical values are converged to obtain analysis results.

In this way, the process model is constructed,
This model is used to perform control factor optimization calculations. That is, the optimum manipulated variable of the control factor is obtained by causing the process model to repeatedly calculate while changing the setting value of the control factor given to the process model. For optimization calculation,
Well-known methods can be adopted.

[0014]

EXAMPLE An operation support system for an activated sludge process according to an example of the present invention will be described below. In terms of its function, this system is roughly divided into a process model section 12 and an analysis section 13, as shown in FIG. The process model unit 12 includes a processing process model unit 14 and a control process model unit 15. The processing process model unit 6 is a model of the actual processing process 8 by a mathematical model, and analyzes using the state variables and dynamic parameters given from the state variable storage unit 16 and the dynamic parameter storage unit 17. To do. The control process model unit 7 is a model of a control algorithm, and analyzes using the control factor target value provided from the control factor target value storage unit 18. The analysis unit 2 includes a steady state analysis unit 19 and an optimization calculation unit 20. The optimization calculation unit 20 is a driving condition optimization unit 2 that optimizes driving conditions.
1 and a dynamic parameter optimization unit 22 that optimizes dynamic parameters. Hereinafter, each component of this driving support system will be described in detail.

1. Process Model Section In determining the mathematical model in the treatment process model section 14, a state variable and a network of main reactions of the circulating nitrification denitrification method were assumed. As shown in FIG. 2, the organic substances existing in the inflowing water and in the reaction tank are roughly classified into soluble organic substances and floating organic substances. Soluble organics soluble BOD
(S) and soluble organic nitrogen (ORN). The floating organic matter is assumed to be composed of BOD-assimilating bacteria (X) and nitrifying bacteria (XN), and is converted into soluble BOD and soluble organic nitrogen by hydrolysis.

Further, in view of the fact that most of activated sludge bacteria are facultative anaerobic bacteria, denitrifying bacteria are treated with BO.
Identical to D-assimilating bacteria, depending on the degree of DO concentration rate-controlling, BO
D removal and denitrification reaction shall proceed. In addition, when the soluble BOD component is decomposed by BOD-assimilating bacteria,
It is assumed that the soluble organic nitrogen component (ORN) is also decomposed at the same time to ammonia nitrogen (NH ₄ —N). This deamination reaction is not accompanied by oxygen consumption. Ammonia nitrogen (produced by inflow and deamination reaction) is oxidized to nitric nitrogen (NO ₃ -N) by nitrifying bacteria under aerobic conditions, and at the same time, BOD-assimilating bacteria (denitrifying bacteria) and nitrifying bacteria grow. In this case, it shall be used as a nitrogen source. In addition,
In this model, a hydraulic model in which N perfect mixing tanks are connected in series is assumed as a reaction tank.

In the above model, Table 1 shows the rate-limiting items and their formulas in consideration of the main reactions of BOD removal, nitrification and denitrification reactions.

[0018]

[Table 1]

As shown in this table, the soluble BOD removal and nitrification reaction are rate-determined by DO in the form of the Monod equation,
It was assumed that the denitrification reaction is rate-controlled in the form of C-Monod for DO. The C-Monod equation is (1-x / (k +
x)). Here, x is a rate-determining factor and k is a saturation constant. The pH was calculated from the relational expression with alkalinity. In addition, in the nitrification reaction and denitrification reaction, the pH changes as the alkalinity is consumed and produced.
The effect of pH was taken into account using the relational expression described by g et al. This relational expression is represented by (1-a (b-pH)). Here, a and b are constants. Further, the temperature dependence of the BOD removal rate coefficient, the maximum specific nitrification rate coefficient, and the maximum specific denitrification rate coefficient was set according to the Arrhenius equation.
This equation is represented by a ₂₀ · θ ^(T-20) . Where a ₂₀ , θ
Is a constant.

The control process model unit 15 builds a model for each controllable factor. Here, the control items include nitrification rate control of the nitrification tank, DO control, circulation ratio and A / O ratio control of the nitrification solution, pH control of each circuit of the denitrification tank and nitrification tank, and SRT control.

2. Analysis unit 2.1. Steady state analysis part Steady state analysis is important as a tool to determine the steady state characteristics of state variables including control factors, and as an optimization tool and a means for determining the initial value of dynamic simulation. As can be seen from FIG. 2 and Table 1, the process model is a simultaneous non-linear differential equation, and a steady solution cannot be obtained algebraically. Therefore, we decided to transform the model by the following method and obtain a steady solution by numerical convergence calculation.

A) The differential term of the nonlinear simultaneous equations is set to 0 by the steady analysis.

B) Solve the equation for each variable.

C) When solving the equation in b), the variables in the denominator of the hyperbolic function (Monod equation), which is a non-linear element, are
The true value is converged by leaving it as it is and repeating the calculation.

Specifically, the calculation is performed according to the following procedure.

A) Enter the water temperature of the reaction tank (denitrification tank and nitrification tank), the inflow load, and the initial value of each state variable. Further, the initial value of the control target value for DO control, SRT control, etc. is input.

B) Each control target value is compared with a calculated value obtained by repeated calculation, and each manipulated variable is obtained by PID or fuzzy calculation so as to eliminate the deviation.

C) Compare the previous variable value with the current value,
Convergence occurs when the change falls within the allowable range.

2.2. Optimizer The function of the optimizer is roughly divided into two. One is optimization of operating conditions, and the other is optimization of kinetic parameters.

(1) Optimization of driving operation conditions Evaluation criteria and constraint conditions are set, and the optimum operation amount satisfying them is determined. For example, the evaluation criterion is "maximize the TN removal rate", and the SIMPLEX method can be used as a method of searching for an optimal solution that satisfies this. The parameters desired to be optimized can be arbitrarily selected.

(2) Optimization of kinetic parameters Similar to (1), evaluation criteria and constraints are set, and optimum kinetic parameters satisfying these criteria are determined. Generally, the optimum values of the dynamic parameters fluctuate due to long-term load fluctuations, water temperature fluctuations, and the like. Therefore, in order to stably operate the system as a driving support system for a long period of time, it is necessary to periodically review the kinetic parameters.

For example, the nitrate nitrogen concentration in the denitrification tank, the ammonia nitrogen concentration in the nitrification tank, the BOD concentration, the MLSS concentration,
And measured values such as nitrifying bacteria concentration, and corresponding steady-state analysis is the output computed value and evaluation criteria that the sum of squares to minimize the deviation, denitrification rate constant (K _D), nitrification reaction rate constant (K It is possible to obtain optimal solutions of kinetic parameters such as _N ) and BOD removal rate constant (K _L ).

3. Operation procedure Figure 3 shows the outline of the operation procedure of this system. First, the process model unit 12 is set with initial values of each state variable necessary for steady state analysis and optimization, initial values such as dynamic parameters and control target values (S1). Next, the process model unit 12 starts repeated calculation based on this initial setting value, and the steady analysis unit 19 continues the calculation until the convergence condition is satisfied (S2, 3).

When the convergence condition is satisfied (S3; Yes),
The steady-state analysis unit 19 sets the calculation result at that time as a steady-state solution, and 2 of the evaluation reference values for determining whether or not the solution is the optimum solution.
Determine the item. The optimization unit 20 uses the evaluation reference value output from the steady state analysis unit 19 to determine whether the current steady state solution is the optimum value (or has converged to the optimum value) (S4). If the optimal solution has not yet been obtained,
Each manipulated variable is assigned a predetermined algorithm (eg SIMPL
(EX method), and the calculation of steady state analysis / optimization is repeated again (S5). In the case of optimizing the dynamic parameters, the calculation is performed in the same procedure.

4. Specific Example (1) Steady State Analysis FIGS. 4 to 8 show one example of the steady state analysis result. In this example, the water temperature is 20 ° C., the nitrification rate at the outlet of the nitrification tank is 90%, and the A / O ratio is 1:
3, when the reaction tank residence time is 6 hours. From these results, the T-N removal rate, the nitrifying bacteria concentration (XN), the MLSS concentration (X) and the air flow rate (G
The relationship with s) becomes clear. In the case of T-N, the removal rate is improved by operating the SRT for a long time and decreasing the DO concentration, but it can be seen that when it is 15 days or more, the required air flow rate increases accordingly.

(2) Optimization From the results of steady-state analysis, it is considered that the required air flow rate (which serves as a guideline for electric energy) in addition to the T-N removal rate is also necessary as an evaluation criterion for optimization. As an example, the MLSS concentration at the outlet of the nitrification tank is 1,000 mg /
The optimum combination of the DO concentration set value and the circulation rate when operated at 1, 2,000 mg / l and 3,000 mg / l was examined. The result is shown in FIG. However, the water temperature in the reaction tank was 20 ° C., the A / O ratio was 1: 3, and the residence time in the reaction tank was 6 hours. MLSS concentration is 1,000 mg / l
It is better to decrease the DO setting value and increase the circulation ratio (R) as the operation increases from TN to 3,000 mg / l.
It is advantageous in improving the removal rate. This is MLS
When the S concentration is low, it is shown that increasing the DO setting value to improve the nitrification rate is more effective in removing T-N than increasing the circulation ratio to increase the NO ₃ -N amount to the denitrification tank. There is. The reason why the circulation ratio is kept low is that an increase in the DO set value causes an increase in the amount of DO flowing into the denitrification (anaerobic) tank due to circulation, which suppresses the denitrification reaction. On the other hand, it is possible to suppress the MLSS concentration (nitrifying bacteria concentration) by lengthening the SRT.

[0037]

As described above, according to the present invention, the simultaneous nonlinear differential equations representing the networks of the main reactions of the treatment process are replaced with the linear form to construct the model of the process, and the iterative calculation is performed on this model. A steady solution of the process simulation output is obtained by performing the numerical convergence calculation together, and the steady solution is repeatedly calculated while changing the setting value of the control factor given to the process model, thereby controlling based on an arbitrary operating condition, etc. Find the optimum manipulated variable of the factor. As a result, even a plurality of operation control parameters that interfere with each other can be quantitatively determined not to be determined by trial and error, but to satisfy arbitrary standard criteria and constraint conditions.

Therefore, this device can be operated, for example, so as to output an optimum solution at regular intervals, and can be used to support the operation of the nitrification / denitrification process, which brings about the following various effects.

(1) Since more appropriate operation of the nitrification / denitration process is possible, the quality of treated water is improved and the influence on the environment of the discharge destination can be suppressed to the maximum.

(2) Efficient operation is possible, which is advantageous in terms of operating cost.

(3) Since the operation can be performed easily and accurately without requiring skill, it is possible to efficiently use human resources, which contributes to labor saving.

[Brief description of drawings]

FIG. 1 is a functional block diagram of an operation support system for an activated sludge process according to an embodiment of the present invention.

FIG. 2 is a diagram showing a network of state variables and main reactions.

FIG. 3 is a flowchart showing the operation of the system of FIG.

FIG. 4 is a graph showing the results of steady state analysis.

FIG. 5 is a graph showing a steady-state analysis result.

FIG. 6 is a graph showing the results of steady state analysis.

FIG. 7 is a graph showing a steady-state analysis result.

FIG. 8 is a graph showing the results of steady state analysis.

FIG. 9 is a graph showing an optimization calculation result.

FIG. 10 is an explanatory diagram showing an outline of a circulation type nitrification / denitration process.

FIG. 11 is an explanatory diagram showing an outline of a control unit of a circulation type nitrification / denitration process.

[Explanation of symbols]

 12 ... Process model part 13 ... Analysis part 14 ... Processing process model part 15 ... Control process model part 19 ... Steady-state analysis part 20 ... Optimization calculation part 21 ... Operating condition optimization part 22 ... Kinetic parameter optimization part

Claims (2)

[Claims] 1. A device for a treatment process in which anaerobic / aerobic treatment is combined to perform nitrification / denitrification, and an optimum value of a control factor of the treatment process is obtained. A system of nonlinear differential equations representing a main reaction network of the treatment process. Has a process model part that simulates the process and a control process model part that simulates the control algorithm. The process model part performs mutual calculation in both model parts, and the process process state variables and control factors. Is to obtain a steady solution of the simulated output of the processing process from the set value of, and the process model section is repeatedly calculated, and numerical convergence calculation is performed using the calculation result, and this converged value is used as the steady solution. By controlling the steady state analysis unit and the steady state analysis unit repeatedly while changing the set values of the control factors, A nitrification denitrification process simulator, comprising: a control factor optimization calculation unit that performs a control factor optimization calculation. 2. The nitrification denitrification process simulator according to claim 1, wherein the process model unit is repeatedly operated while changing the kinetic parameters used for constructing the treatment process model unit. The nitrification denitrification process simulator according to claim 1, further comprising a dynamic parameter optimization calculation unit that performs a parameter optimization calculation.