JPS649077B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention uses an aeration tank as the main equipment of a treatment facility, such as a sewage treatment plant that employs the so-called activated sludge method. This invention relates to a method for simulating sewage treatment facilities, which is used in electronic computer simulation for operating sewage treatment facilities.

A sewage treatment facility using the activated sludge method is configured as shown in Figure 1. In other words, the main equipment in the activated sludge method is an aeration tank 1 and a settling tank 2. In the aeration tank 1, the substrate in the inflowing sewage is converted into sludge by the action of microorganisms, and then in the settling tank 2, solid-liquid separation is performed by sedimentation. , the supernatant liquid is treated water 3
The water is discharged into rivers as water. In FIG. 1, 4 indicates a return sludge pipe, 5 indicates an excess sludge pipe, 6 indicates an inflow sewage pipe, and 7 indicates an aeration air pipe. The operating characteristics of this treatment facility are that the time-varying characteristics of the inflowing sewage have large daily fluctuations and that the quality of the treated water is maintained below a predetermined value. The operation method of this treatment facility is largely dependent on the experience of operating engineers, but on the other hand, computer simulations are often used to examine operating conditions. In the case of the latter, that is, simulation, a mathematical model of the aeration tank and settling tank is required, but the main element is the aeration tank, and the mathematical model of this aeration tank must be highly accurate.

The main reaction factor in the aeration tank is the mixed state of activated sludge containing microorganisms and substrate, and the indicators usually used are biochemical oxygen demand (substrate concentration) BOD, suspended solids concentration (sludge concentration) SS, and dissolved oxygen. Concentration DO 3
It is one. As a model of the mixing state, a tank array model is generally widely used, and when the number of stages is set to the minimum one, it becomes a complete mixing model, and when the number of stages is set to infinity, it becomes an extrusion flow model. The actual mixing state is considered to be an intermediate state between these two extremes, and a multi-stage complete mixing model is sometimes used. Various proposals have been made to express the sludge growth process, but here we will use the Monod as a microbial chemical reaction equation.
Based on the equation, the complete mixing model and extrusion flow model can be expressed by the following equations.

() Complete mixture model () Extrusion flow model Here, Xc, Xp, Sc, and Sp are the sludge concentration (SS) and substrate concentration (BOD), respectively, and the subscripts C and P represent the values for the complete mixing model () and the extrusion flow model (), respectively. X _I and S _I are the input sludge concentration (SS) and substrate concentration (BOD), respectively, μ^ is the maximum growth rate,
K _s is the saturation constant, Y is the yield coefficient, K _d is the autooxidation rate,
T _R is the aeration tank residence time, and t is an arbitrary time.

In addition, the following equation is generally used as a dissolved oxygen (DO) model.

Here, C is the dissolved oxygen concentration, C _s is the saturated dissolved oxygen concentration, C _I is the input value of the dissolved oxygen concentration, K _La is the overall oxygen transfer coefficient, F is the aeration air volume, r _r is the oxygen consumption rate, and K ₁ to _K4 are coefficients. S and X become S _c and X _c and Sp and Xp, respectively, in the above perfect mixing model ( ) and extrusion flow model ( ).

The settling tank model calculates the sludge concentration drawn from the settling tank as a function of the flow rate flowing from the aeration tank and the sludge concentration, and is obtained by modeling physical phenomena or statistical modeling.

As mentioned above, each coefficient (μ^, K _s , Y, K _d , K ₁
~K ₄ ) has not been previously performed using a time-varying model, and only simulations using rough calculations have been performed. For this reason, in the past, it was not possible to obtain highly accurate index values through simulation, and it was not possible to sufficiently examine treatment methods corresponding to various operating states of the target system.

The present invention was made against the background of the above, and in simulating a sewage treatment facility in which an aeration tank is the main equipment of the treatment facility,
The purpose of this paper is to provide a simulation method for sewage treatment facilities that takes into account the characteristics of inflowing sewage that vary greatly from day to day and makes it possible to obtain rational and highly accurate index values.

In other words, the present invention is characterized by dividing the mathematical model of the aeration tank into a plurality of ranges of inflow sewage flow rate, taking into account the characteristics of inflow sewage that vary greatly from day to day.
The aim is to create an unsteady model by making individual mixed models correspond to these divided modes of mutually different amounts of water and connecting them in time series.

Hereinafter, one embodiment of the present invention will be described in detail.

When trying to construct a mathematical model of an aeration tank that takes into account the characteristics of inflowing sewage, 8 in equations (1) to (3) above
An unsteady model in which the coefficients (μ^, K _s , Y, K _d , K ₁ to _{K 4} ) are time-varying can be considered, but this is currently impossible to realize both theoretically and experimentally. Therefore,
We decided to divide the inflow sewage flow rate into two regions, high water flow and low water flow, and consider that each region has a different mixing state, so we assigned separate mixing models to each region and connected these in chronological order.

As a basic mixing model, the actual mixing state of the aeration tank will be expressed by a model in which the above-mentioned two extreme state models, the single-stage complete mixing model, and the extrusion flow model are combined by weight coefficients.

In this case, the concentration Z is expressed by the following equation.

Z=α・Z _c +(1−α) Z _p (4) Here, the concentration Z is X or S. Also, α
is a weighting coefficient, and the weighting coefficient α(t) at any time t is expressed by the following equation.

α(t)＝α _H when Q(t)≧Q _M α _L when Q(t)＜Q _M ……(5) Here, Q(t) is the inflow sewage flow rate, and Q _M is the use of the flowing sewage flow rate. This is the median value of the actual data group Q.

As described above, the 8 coefficients in equations (1) to (3) are kept constant, and a simple method is used to process these time changes by introducing a weighting coefficient and making this weighting coefficient binary in relation to the flow rate. This means that we have obtained a non-stationary model. The advantage of this model is that when a multistage perfect mixing model is used, the number of stages will be different when the flow rate is high and when the flow rate is low, and when connected in time series, the initial value at each stage will be different. However, since it is a single stage, this initial value calculation is not necessary, and the calculation time is shortened.

Next, in order to construct the above-mentioned model, a total of 10 values must be determined, including the eight coefficients of equations (1) to (3) and two weighting coefficients.

FIG. 2 is a flowchart showing the procedure for determining these coefficients, and the following description will be made along this flowchart.

The first is a combination of weighting factors. In this case, if equations (1) to (3) were used as they were, it would be necessary to determine the value of the 8 coefficients in advance, but according to the actual data in the aeration tank, the growth term of the sludge concentration and the The difference in the oxidation term is 1 compared to the sludge concentration.
Taking note of the fact that the value is more than an order of magnitude smaller, when determining the weighting coefficient, these terms are omitted and the data of only the sludge concentration is used to determine the weighting coefficient. In this case, the following equations are used as the complete mixing model and extrusion flow model.

() Complete mixing model dX _c / dt = − (X _c − X _I ) / T _R ... (1a) () Extrusion flow model X _p = X _I (t - T _R ) ... (2a) Since it takes a value between 0 and 1, 0
If the range from 1 to 1 is divided into N equal parts, (N+1) points are obtained. The combination of the two weighting coefficients α _H and α _L is (N
+1) ² pieces, and for this total number, the case where the calculated value using formulas (1a) and (2a) and the measured value best match is selected by the enumeration method. The resulting combination of two coefficients is the optimal coefficient combination α ^* _H and α ^* _L .

Next, the four coefficients related to substrate concentration (BOD) and sludge concentration (SS) are determined using equations (1) and (2) above, but if M values are considered for one coefficient, In the exhaustive enumeration method, the number of cases is M ⁴ , which is too many, so we will decide one by one. In this case, since the order must be determined, a sensitivity analysis is performed between the four coefficients and the index indicating the individual degree of fitness, and the coefficients are determined in descending order of sensitivity. In this way, 4 coefficients can be determined using 4M cases. Here, the index I ₁ indicating the goodness of fit is defined as shown in the following equation.

I ₁ = _D 〓 〓 {β(X _A −X _C ／X _A ) ² +(1−β)・(S _A −S _C ／S _A ) ²
}...(6) Here, X and S are the sludge concentration (SS) and substrate concentration (BOD), respectively, and the subscripts A and C indicate the measured value and calculated value, respectively. β is the weighting coefficient,
D is the number of discrete times of data used here.

Subsequently, the four coefficients of equation (3) are determined. this is
This is performed using the same handling as in the case of equations (1) and (2), but in this case, the following equation is used as the index I ₂ indicating the goodness of fit.

I ₂ = _D 〓 (C _A −C _C /C _A ) ...(7) Here, C is the dissolved oxygen concentration, and the subscripts A and C
are the measured value and calculated value, respectively.

The algorithm for determining the coefficients described above is illustrated in the flow chart shown in FIG.

In this way, once the 10 coefficients are determined,
This means that we have obtained an unsteady aeration tank model that can calculate three indicators: substrate concentration (BOD), sludge concentration (SS), and dissolved oxygen concentration (DO).

Using the unsteady model of the aeration tank that takes into account large fluctuations in the sewage inflow characteristics of the sewage treatment plant, which was materialized as described above, when the inflow sewage 6 and return sludge 4 in the system shown in Fig. 1 are If the series characteristics and the time series data of the aeration air volume 7 are given, the aeration tank 1
The above three index values (BOD, SS, DO) at the exit of
are calculated sequentially, and shown in Figure 3 a, b, and c, respectively.
MLSS (mixed liquid suspended solids concentration (SS) in aeration tank), DO
(dissolved oxygen concentration), BOD (biochemical oxygen demand)
As shown in the example of the actually measured value (solid line) of the time series data and the calculated value (broken line) according to this embodiment, highly accurate time series data can be obtained. In addition, in Figures 3 a to c, it is clearly seen that the calculated values of MLSS fit well with the measured values, whereas the degree of fit between the calculated values and the measured values for DO and BOD appears to be somewhat low. However, this is because DO and BOD have significantly smaller values than MLSS, and it can be seen that the trends in change are in good agreement. In this case, if the model of the settling tank 2 is used, the time-series characteristics of the returned sludge 4 can be calculated.
Variable factors in this target system include inflow sewage characteristics, operating motors, regulating valves, and inflow flow rate distribution ratios, but numerical experiments can be attempted on various combinations of these variable factors that are not possible with actual equipment. , it is possible to quantitatively understand the influence of each factor on each index.

In this way, by performing simulations using the above-mentioned unsteady aeration tank model, it is possible to obtain time-series data of the three indicators (BOD, SS, DO) with high accuracy, so it is possible to obtain time-series data of the three indicators (BOD, SS, DO) with high accuracy. By performing trial calculations on various combinations of operating motors, regulating valves, inflow flow rate distribution ratios, etc., and comparing and examining the results, it is possible to learn about all possible operating conditions. This makes it possible to consider in advance how to deal with conditions that may pose an index problem during the operation of the installed treatment facility, allowing stable operation over a long period of time.
In addition, it is possible to easily change the numerical ratings of the equipment and equipment that are actually installed, so it is possible to perform trial calculations on the operation of equipment and equipment that exceed the ratings, and use the results to modify the equipment and equipment. It becomes possible to revise the ratings, that is, to design new equipment and equipment. Overall, this simulation is an extremely useful tool for comprehensively understanding the operation of activated sludge method sewage treatment facilities without interfering with the operation of the actual equipment.

It should be noted that the present invention is not limited to the embodiments described above and shown in the drawings, but can be implemented with various modifications without changing the gist thereof.

For example, in the above embodiment, the weighting coefficient α(t) in equation (5) was set to two values, but this was changed to the k value (k≧
3; k is a positive integer) In other words, it is also possible to divide the range of inflow sewage flow rate into three or more parts and assign different weighting coefficients to each part. In this case, 0
If the range of ~1 is divided into N equal parts, the number of cases is (N+1) ^k
Therefore, as k becomes larger, the simulation cost increases accordingly, and in practical terms, restrictions are placed on this cost, so it is not a good idea to make it too large.

In addition, many variations of Monod's equation as a microbial chemical reaction equation in the sludge growth process have been published in addition to those shown in equations (1) and (2), and it is also possible to use these other equations. be. In this case, the number of coefficients to be determined may change, but as a result of sensitivity analysis,
The procedure of determining coefficients in descending order of sensitivity to the index (formula (6)) is exactly the same.

Furthermore, as an index indicating the degree of conformity, in addition to those shown in equations (6) and (7), the absolute value of the difference between the measured value and the calculated value may be used.

As described in detail above, according to the present invention, it is possible to provide a simulation method for a sewage treatment facility that makes it possible to obtain a rational and highly accurate index value in consideration of the characteristics of inflow sewage that vary greatly from day to day. can.

[Brief explanation of drawings]

FIG. 1 is a schematic configuration diagram showing a schematic configuration of an example of an activated sludge method sewage treatment facility to which an embodiment of the method of the present invention is applied, FIG. 2 is a flowchart showing the coefficient determination algorithm in the same embodiment, and FIG. Figures a to c are diagrams showing the relationship between simulation results and actually measured values using the same example. 1... Aeration tank, 2... Sedimentation tank, 3... Treated water (discharge), 4... Return sludge pipe, 5... Excess sludge pipe, 6... Inflow sewage pipe, 7... For aeration air duct.

Claims (1)

[Claims] 1. In a method for determining at least one concentration index among sludge concentration, substrate concentration, and dissolved oxygen concentration in an aeration tank in a sewage treatment facility, the characteristics of inflow sewage that vary greatly over time are The flow rate range is divided into a high water flow area and a low water flow area, and at predetermined intervals, it is determined whether the inflow sewage flow rate is a high water flow area or a low water flow area, and if it is in a high water flow area, single-stage complete mixing is performed. The concentration index is determined using a model, and in the case of the low water flow region, the concentration index is determined using an extrusion flow model, and then, using an unsteady model in which these concentration indexes are combined by a weighting coefficient, A method for simulating a sewage treatment facility, the method comprising performing a simulation operation. 2. In the sewage treatment facility simulation method described in claim 1, when determining the weighting coefficient, data of only the sludge concentration measured at the aeration tank outlet side is used for the single-stage complete mixing model and the extrusion flow model. A method for simulating a sewage treatment facility, characterized in that the compatibility between calculated values and measured data is determined using an enumeration method to determine the best fit to the measured data. 3. In the sewage treatment facility simulation method described in claim 2, each coefficient is A method for simulating a sewage treatment facility, characterized in that a sensitivity analysis is performed to determine coefficients in descending order of sensitivity.