JP2017209642A - Method of adjusting return sludge amount and excess sludge amount from final sedimentation place - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology for kinetically predicting MLSS concentration simply and adjusting return sludge amount and excess sludge amount from a final sedimentation place based on the prediction.SOLUTION: A water treatment system has a reaction tank and a final sedimentation place. MLSS concentration at a terminal of a reaction tank during a prescribed period having time length at a level that MLSS concentration of the terminal of the reaction tank is not deviated from a prescribed range is predicted. Then return sludge amount and excess sludge amount from the final sedimentation place in the prescribed period is determined based on a prediction value of the MLSS concentration and the water treatment system is operated to have the return sludge amount and the excess sludge amount in the prescribed period.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for adjusting the amount of returned sludge and excess sludge from the final sedimentation basin of a sewage treatment plant.

  In operation management at the sewage treatment plant, the amount of return sludge and excess sludge are routinely manipulated to determine the amount of withdrawal from the final sedimentation basin. For example, in order to secure the activated sludge in the reaction tank, the return sludge amount is adjusted and the return sludge concentration is confirmed. In order to prevent sludge carryover in the final sedimentation basin, the amount of excess sludge is adjusted and the sludge interface is confirmed. In order to adjust the load on the sludge treatment facility, the amount of excess sludge is adjusted and the concentration of excess sludge is confirmed.

  In consideration of such a plurality of factors, adjusting the amount of returned sludge and the amount of excess sludge largely depends on the experience of skilled workers.

  On the other hand, it has been conventionally studied to realize stable operation management by maintaining the MLSS (activated sludge suspended matter) concentration in the reaction tank within an appropriate range.

  For example, in Patent Document 1, the amount of returned sludge and the amount of excess sludge from the final sedimentation basin are adjusted so that the measured value of the MLSS concentration follows the target value.

  In Patent Document 2, the future behavior is predicted, and the amount of returned sludge and the amount of surplus sludge from the final sedimentation basin are adjusted so that the predicted value of MLSS concentration and the predicted value of returned sludge concentration follow the target values.

JP-A-9-168791 JP-B-1-47238

  However, in the adjustment method described in Patent Document 1, since the MLSS concentration is confirmed afterwards, the amount of returned sludge and excess sludge are determined and adjusted, so the response when the MLSS concentration fluctuates is delayed. In many cases, it was difficult to maintain the MLSS concentration within an appropriate range.

  On the other hand, the adjustment method described in Patent Document 2 increases the possibility of avoiding a delay in responding to the variation in the MLSS concentration. However, it is necessary to maintain the instantaneous values of both the MLSS concentration and the returned sludge concentration in an appropriate range at the same time, which complicates the algorithm. In addition, advanced knowledge is required for prior experiments for setting various parameters. As a result, it is difficult to utilize at the operation management site.

  The present invention solves the above-mentioned problems, and dynamically and simply predicts the MLSS concentration, and based on this, an object is to provide a technique for adjusting the amount of returned sludge and excess sludge from the final sedimentation basin. And

  The present invention that solves the above problems is a method for adjusting the amount of returned sludge and excess sludge in a water treatment system including a reaction tank and a final sedimentation basin. The MLSS concentration at the end of the reaction vessel is predicted for a predetermined period having a length of time that does not deviate from the predetermined range, and the final value in the predetermined period is determined based on the predicted value of the MLSS concentration. The amount of returned sludge and the amount of excess sludge from the settling basin are determined, and the operation is performed so that the amount of returned sludge and the amount of excess sludge are the same during the predetermined period.

  In the above invention, preferably, an average of predicted values of the MLSS concentration in the predetermined period does not deviate from a predetermined range.

  Because operations are performed based on forecasts, delays in response to fluctuations can be avoided. In addition, since the operation is based on long-term prediction (predetermined period), it is possible to avoid complication of the operation management site due to the complexity of the algorithm compared to the operation based on the instantaneous value prediction.

  In the above invention, preferably, the prediction of the MLSS concentration at the end of the reaction tank in the predetermined period is based on a prediction model, and the prediction model is returned to the reaction tank model for predicting the MLSS concentration at the end of the reaction tank. It consists of a final sedimentation basin model that predicts the sludge concentration, and the MLSS concentration prediction value at the end of the reaction tank in the reaction tank model is input to the final sedimentation basin model, and the return sludge concentration prediction value from the final sedimentation basin model is The output sludge concentration predicted value in the return final sedimentation basin model is input to the reaction tank model, the MLSS concentration predicted value is output from the reaction tank model, and input and output are repeated in the predetermined period.

  Thereby, long-term prediction of MLSS concentration can be performed.

  In the above invention, preferably, in the reaction tank model, MLSS concentration at the end of the reaction tank is predicted based on the sludge residence time distribution in the reaction tank, and in the final sedimentation tank model, the sludge residence time in the final sedimentation tank. Based on the distribution, the return sludge concentration is predicted.

  Thereby, MLSS concentration can be easily and accurately predicted.

  In the above invention, preferably, the MLSS concentration at the end of the reaction vessel is measured, and an error is evaluated based on the MLSS concentration predicted value and the measured value, and the time length of the predetermined period is changed based on the error evaluation. To do.

  In the above invention, more preferably, in the error evaluation, when it is determined that the deviation has deviated from the allowable range, a period having a time length from the start of the present predetermined period to the determination of the deviation from the allowable range is set as the next predetermined period. And MLSS concentration at the end of the reaction tank in the next predetermined period is predicted.

  In the above invention, more preferably, in the error evaluation, when it is determined that the allowable range is not deviated in a predetermined period, a period longer than the present predetermined period is set as the next predetermined period, and the reaction tank in the next predetermined period is set. Predict terminal MLSS concentration.

  Such operations can further reduce the burden on the operation management site while ensuring stability.

  In order to solve the above problems, the present invention is a water treatment system including a reaction tank, a final sedimentation basin, and a control device. The control device sets a predetermined period of time that does not deviate from the predetermined range of the MLSS concentration at the end of the reaction tank, predicts the MLSS concentration at the end of the reaction tank in the predetermined period, and The amount of returned sludge and surplus sludge from the final sedimentation basin during the predetermined period is determined based on the predicted value of the final sedimentation basin so that the amount of return sludge and excess sludge during the predetermined period is determined. Manipulate.

  In order to solve the above problems, the present invention is a program for operating a water treatment system including a reaction vessel, a final sedimentation basin, and a control device. A process for setting a predetermined period having a time length such that the MLSS concentration at the end of the reaction tank does not deviate from a predetermined range; a process for predicting the MLSS concentration at the end of the reaction tank in the predetermined period; Based on the predicted value, a process for determining the amount of return sludge and surplus sludge from the final sedimentation basin during the predetermined period, and the adjustment valve for the final sedimentation basin so as to obtain the amount of return sludge and surplus sludge during the predetermined period And processing for outputting an operation command to the control device.

  According to the present invention, MLSS concentration can be dynamically predicted easily. By adjusting the amount of return sludge and excess sludge from the final sedimentation basin based on the dynamic prediction of MLSS concentration, the burden on the operation management site can be reduced.

System overview Long-term forecast conceptual diagram Functional block diagram of control device Prediction model conceptual diagram Predictive model verification

~ System overview ~
FIG. 1 shows a basic configuration of a water treatment system. The water treatment system has a reaction tank 10 and a final sedimentation tank 20.

  Sewage containing organic matter flows into the reaction tank 10 from the outside. In the reaction tank 10, the microorganisms in the tank are agitated and aerated. Thereafter, it flows out from the end of the reaction tank 10 to the final sedimentation tank 20. Solid-liquid separation is performed by precipitation in the final sedimentation tank 20. The supernatant is discharged through a discharge pipe.

  The sludge settled in the final sedimentation basin 20 is withdrawn, and a part thereof is returned to the reaction tank 10 via a pump. The rest is carried out of the system as excess sludge through a pump.

  At this time, the return sludge amount and the excess sludge amount are adjusted by operating the adjustment valve 21.

~System operation~
The operation of the system in this embodiment will be described.

  In the present embodiment, first, dynamic prediction of the MLSS concentration at the end of the reaction tank 10 in a predetermined period (for example, two weeks) is performed. FIG. 2 is a conceptual diagram of long-term prediction.

  The dynamic prediction is preferably based on a prediction model (for example, a prediction model described later), but may be calculated based on past data. The dynamic prediction obtained from the prediction model may be corrected by past data.

  At this time, the predicted value is set so as not to deviate from a predetermined range (allowable range). In addition, it is preferable that all the predicted values in the predetermined period are maintained in the allowable range, but even if some predicted values deviate instantaneously, if the average of the predicted values is maintained in the allowable range, Some deviation may be allowed.

  Next, the return sludge amount and surplus sludge amount from the final sedimentation basin are determined based on the solids balance so as to follow the dynamic prediction, and the return sludge amount and surplus sludge amount are adjusted.

  In this embodiment, since operation based on prediction is performed, it is possible to avoid a delay in dealing with fluctuations. In addition, since the operation is based on long-term prediction (for example, two weeks), it is possible to avoid the algorithm from becoming complicated as compared to the operation based on the instantaneous value prediction.

  Thereby, the burden at the operation management site can be reduced.

  Additional operations of the system in this embodiment will be further described.

  This embodiment is characterized in that operation is performed based on long-term prediction (for example, two weeks) of the MLSS concentration at the end of the reaction tank 10. When the operation for a predetermined period is completed, a long-term prediction is newly made and the operation for the predetermined period is started. Repeat this.

  At this time, the MLSS concentration at the end of the reaction vessel 10 may be measured via the sensor 11 and error evaluation may be performed based on the predicted value and the measured value to confirm the prediction accuracy.

  The prediction model described later is accurate, and it is premised that the MLSS concentration at the end of the reaction vessel 10 is maintained within an appropriate range for a predetermined period. However, if the error exceeds an allowable range due to unexpected factors, Suspend operation before the end (for example, 2 weeks). Then, the period from the current predetermined period start time to the allowable range deviation determination time is set as the next predetermined period (for example, 10 days). A long-term prediction for the next predetermined period is newly made, and operation for the predetermined period (10 days) is started.

  Thus, the predetermined period may be changed based on the error evaluation.

  On the other hand, if the error is within an allowable range in the predetermined period (the MLSS concentration at the end of the reaction tank 10 is maintained in an appropriate range), a period (for example, 18 days) longer than the current predetermined period (for example, 2 weeks) The next predetermined period may be set, a long-term prediction for the next predetermined period may be newly performed, and the operation for the predetermined period (18 days) may be started.

  Further, for example, after the operation for two weeks is repeated several times (for example, five times), if the error is within the allowable range, the operation for 18 days may be started.

  Such operations can further reduce the burden on the operation management site while ensuring stability.

~Control device~
Furthermore, a control device 30 may be provided and the above operation may be performed based on the determination of the control device 30. FIG. 3 is a functional block diagram of the control device 30.

  The control device 30 includes an operation period setting unit 31, an MLSS concentration prediction unit 32, a return sludge amount / surplus sludge amount adjustment unit 33, and an error evaluation unit 34.

  The operation period setting unit 31 sets an operation period (= prediction period). The MLSS concentration prediction unit 32 performs dynamic prediction of the MLSS concentration at the end of the reaction tank 10 during the operation period.

  The return sludge amount / surplus sludge amount adjustment unit 33 determines the return sludge amount and the excess sludge amount from the final sedimentation basin so as to follow the dynamic prediction, and outputs a control signal to the adjustment valve 21.

  The error evaluation unit 34 inputs the measured value of the MLSS concentration at the end of the reaction vessel 10 from the sensor 11 and performs error evaluation. Further, the operation period is changed as appropriate based on the error evaluation result. The operation period setting unit 31 resets the operation period based on a command from the error evaluation unit 34.

  The control device 30 is composed of a CPU, a storage device, and the like, and a part or all of them are operated by a program.

~ Outline of prediction model ~
In this embodiment, in order to perform the operation based on the long-term prediction of the MLSS concentration at the end of the reaction tank 10, it is necessary to ensure the prediction accuracy. Furthermore, the prediction model needs to be simple in order to avoid the complexity of the algorithm and to operate the operation management site.

  FIG. 4 is a conceptual diagram of the prediction model. The present application prediction model is constructed by combining the reaction tank model and the final sedimentation basin model. MLSS concentration prediction results are output in the reactor model, MLSS concentration prediction results in the reactor model are input to the final sedimentation basin model, return sludge concentration prediction results are output in the final sedimentation basin model, and the final sedimentation basin model The return sludge concentration prediction result is input to the reaction tank model, and the MLSS concentration prediction result is output again from the reaction tank model.

  Thus, input and output are repeated based on the solid balance in a predetermined prediction period. This enables long-term prediction of MLSS concentration.

  In the reaction tank model, the relationship between the inflow solids amount and the outflow solids amount in the reaction tank is mathematically expressed through the sludge residence time distribution in the reaction tank.

  In the final sedimentation basin model, the relationship between the inflow solids amount and the outflow solids amount in the final sedimentation basin is formulated through the sludge residence time distribution in the final sedimentation basin.

  The sludge residence time distribution in the reaction tank or the final sedimentation basin may be set based on the measurement result by performing a tracer test. Further, it may be assumed to be a normal distribution or set based on past research results.

  The inventor confirmed that even if the sludge residence time distribution is assumed to be a uniform distribution (the longest and shortest are set), the error from the measured value is within an allowable range and sufficient prediction accuracy can be obtained. ing. Further, assuming a uniform distribution, the parameters to be set can be limited, and a simpler model can be obtained. Water quality analysis and prior experiments for parameter setting are not required.

  Furthermore, you may reflect the influence of the solid increase from the outside using the solid increase rate in a reaction tank.

  As described above, in the present embodiment, long-term prediction of the MLSS concentration can be performed easily and accurately. As a result, it is possible to adjust the amount of return sludge and excess sludge from the final sedimentation basin based on long-term prediction of MLSS concentration.

~ Prediction model verification ~
The inventor conducted a verification test on the prediction model (uniform distribution). The forecast period was set at 30 days. FIG. 5 shows the verification test results.

  The numerical value obtained by dividing the absolute value of the difference between the predicted value of MLSS concentration and the actual measurement value by the actual measurement value was taken as the error rate. The difference between the predicted value of MLSS concentration and the measured value was small (average error rate: 7.3%), and it was confirmed that the prediction accuracy of the model was high. It was also confirmed that the prediction accuracy did not decrease over time.

  From the above verification results, it was shown that the operation of this embodiment for adjusting the return sludge amount and the excess sludge amount from the final sedimentation basin based on the long-term prediction of the MLSS concentration is highly practical.

DESCRIPTION OF SYMBOLS 10 Reaction tank 11 Sensor 20 Final sedimentation basin 21 Control valve 30 Control apparatus 31 Operation period setting part 32 MLSS density | concentration estimation part 33 Return sludge quantity and excess sludge quantity adjustment part 34 Error evaluation part

Claims (9)

In a water treatment system equipped with a reaction tank and final sedimentation basin,
Predicting the MLSS concentration at the end of the reactor in a predetermined period of time that does not deviate from the predetermined range.
Based on the predicted value of the MLSS concentration, determine the amount of return sludge and the amount of excess sludge from the final sedimentation basin during the predetermined period,
The return sludge amount / surplus sludge amount adjustment method, wherein the return sludge amount and the excess sludge amount are operated during the predetermined period. The return sludge amount / surplus sludge amount adjusting method according to claim 1, wherein an average of the predicted values of the MLSS concentration in the predetermined period does not deviate from a predetermined range. The prediction of the MLSS concentration at the end of the reactor in the predetermined period is based on a prediction model,
The prediction model consists of a reaction tank model that predicts the MLSS concentration at the end of the reaction tank and a final sedimentation basin model that predicts the return sludge concentration.
MLSS concentration predicted value at the end of the reactor in the reactor model is input to the final sedimentation basin model,
The return sludge concentration prediction value is output from the final sedimentation basin model,
The predicted sludge concentration in the return final sedimentation basin model is input to the reactor model,
MLSS concentration predicted value is output from the reactor model,
The input and output are repeated during the predetermined period. The return sludge amount / surplus sludge amount adjusting method according to claim 1 or 2. In the reactor model, the MLSS concentration at the end of the reactor is predicted based on the sludge residence time distribution in the reactor,
The return sludge amount / surplus sludge amount adjusting method according to claim 3, wherein the final sedimentation basin model predicts the return sludge concentration based on sludge residence time distribution in the final sedimentation basin. Measure the MLSS concentration at the end of the reactor,
Perform error evaluation based on the MLSS concentration predicted value and the measured value,
The return sludge amount / surplus sludge amount adjusting method according to claim 1 or 2, wherein the time length of the predetermined period is changed based on the error evaluation. In the error evaluation, if it is determined that the tolerance has deviated,
A period having a length of time from the start of the predetermined period at the present time to the determination point of deviation from the allowable range is set as the next predetermined period,
The method for adjusting the amount of returned sludge and excess sludge according to claim 5, wherein the MLSS concentration at the end of the reaction tank in the next predetermined period is predicted. In the error evaluation, when it is determined that the allowable range is not deviated in a predetermined period,
Set a period longer than the current period as the next period,
The method for adjusting the amount of returned sludge and excess sludge according to claim 5, wherein the MLSS concentration at the end of the reaction tank in the next predetermined period is predicted. Equipped with a reaction vessel, final sedimentation basin and control device,
The control device
Set a predetermined period of time so that the MLSS concentration at the end of the reactor does not deviate from the predetermined range,
Predicting the MLSS concentration at the end of the reactor during the predetermined period,
Based on the predicted value of the MLSS concentration, determine the amount of return sludge and the amount of excess sludge from the final sedimentation basin during the predetermined period,
A water treatment system, wherein a regulating valve of a final sedimentation basin is operated so that the amount of returned sludge and the amount of surplus sludge are obtained during the predetermined period. A program for operating a water treatment system comprising a reaction tank, a final sedimentation basin, and a control device,
A process for setting a predetermined period of time that does not deviate from the predetermined range of the MLSS concentration at the end of the reaction vessel;
A process for predicting the MLSS concentration at the end of the reaction vessel in the predetermined period;
Based on the predicted value of the MLSS concentration, a process for determining the amount of return sludge and the amount of excess sludge from the final sedimentation basin during the predetermined period;
A water treatment system operation program that causes the control device to execute a process of outputting an operation command to a regulating valve of a final sedimentation basin so that the amount of returned sludge and the amount of surplus sludge during the predetermined period.