JP2021159870A - Water treatment system, operation and management support system for water treatment system, and operation method for water treatment system - Google Patents

Abstract

To provide a water treatment system, an operation and management support system for a water treatment system, and an operation method for a water treatment system, which is capable of properly evaluating an operating condition in a water treatment system equipped with a reverse osmosis membrane device and optimizing the operating condition of the water treatment system.SOLUTION: A water treatment system 1000 includes a reverse osmosis membrane device 10 which treats feed water by reverse osmosis to obtain concentrated water and permeate water, collects reverse osmosis membrane operation information including at least one of the following: water quality information of the feed water, discharge pressure of the pressure pump that pressurizes the feed water supplied to the reverse osmosis membrane device, flow rate of the permeate water, flow rate of the concentrated water, and flux information of the reverse osmosis membrane. Predicting the pressure loss and/or the quality of the permeate water of the reverse osmosis membrane device based on the learned model obtained by machine learning using the collected reverse osmosis membrane operation information, producing optimization information including operating conditions and/or maintenance timing information of the reverse osmosis membrane device based on the results of the prediction, and controlling the reverse osmosis membrane device based on the produced optimization information. The water treatment system is equipped with an operation management support system 30 which controls the reverse osmosis membrane device based on the produced optimization information.SELECTED DRAWING: Figure 1

Description

The present invention relates to a water treatment system, and more particularly to a water treatment system using a reverse osmosis membrane device, an operation management support system for the water treatment system, and an operation method for the water treatment system.

A water treatment system using a reverse osmosis membrane is used for the production of industrial pure water, desalination of seawater, and recovery and reuse of industrial wastewater. In a water treatment system using such a reverse osmosis membrane, the temperature of the water to be treated rises, the desalination rate decreases due to membrane deterioration, scale occurs as the concentration progresses, and membrane fouling progresses. It is known that various operation troubles such as an increase in filtration differential pressure occur, and various measures have been taken to optimize the operation conditions.

For example, Japanese Patent Application Laid-Open No. 8-180311 (Patent Document 1) proposes a technique for determining the state of a reverse osmosis membrane using an index called a water permeability coefficient and determining the timing for membrane cleaning or membrane replacement. There is.

Japanese Unexamined Patent Publication No. 2013-161336 (Patent Document 2) relates to a plurality of prediction models, operation record data of plant equipment, data on current operation status, meteorological observation data, and weather forecast in the operation of a water treatment plant. A technique for predicting the monitored amount of plant equipment using data has been proposed.

Japanese Unexamined Patent Publication No. 8-180311 Japanese Unexamined Patent Publication No. 2013-161336

In the technique described in Patent Document 1, in order to evaluate a reverse osmosis membrane, a water permeation coefficient based on three measurement results of permeation water flow rate, supply water pressure, and supply water side membrane surface osmotic pressure is calculated. There is. However, when evaluating the performance of the reverse osmosis membrane module, factors other than the permeated water flow rate, the supply water pressure, and the supply water side membrane surface osmotic pressure may have a combined effect, so this is a more appropriate performance evaluation method. Still has room for consideration.

The technique described in Patent Document 2 mainly aims at automatic monitoring at a sewage treatment plant, and it seems that it is meaningful to use meteorological observation data and weather information. However, even if Patent Document 2 is applied to the performance evaluation of the reverse osmosis membrane, optimum monitoring cannot always be performed.

In view of the above problems, the present invention is a water treatment system and a water treatment system capable of appropriately evaluating the operating condition in a water treatment system provided with a reverse osmosis membrane device and optimizing the operating conditions of the water treatment system. Provides operation methods for operation management support systems and water treatment systems.

As a result of diligent studies, the present inventor created an artificial intelligence model using the operation information of a water treatment system equipped with a reverse osmosis membrane device, and used the prepared artificial intelligence model to reduce the pressure loss of the reverse osmosis membrane and / Alternatively, it was found that it is effective to predict the quality of permeated water and optimize the operating conditions of the water treatment system based on the prediction results.

In one aspect, the present invention completed based on the above findings supplies the supplied water to the back-penetration membrane device for obtaining concentrated water and permeated water by treating the feed water with the back-penetration membrane, and the water quality information of the supply water and the back-penetration membrane device. Back-penetration membrane operation that collects and collects back-penetration membrane operation information including at least one of the discharge pressure of the pressurizing pump that pressurizes the supply water, the flow rate of permeated water, the flow rate of concentrated water, and the flux information of the back-penetration membrane. Based on the trained model obtained by machine learning using information, the pressure loss and / or the quality of permeated water of the back-penetrating membrane device is predicted, and based on the prediction result, the operating conditions and / or the operating conditions of the back-penetrating membrane device and / Alternatively, it is a water treatment system including an operation management support system that creates optimization information including maintenance timing information and controls the back-penetrating membrane device based on the created optimization information.

In one embodiment of the water treatment system according to the present invention, the operation management support system creates discharge pressure control information and / or permeated water that controls the discharge pressure of the pressurizing pump as operating conditions based on the prediction result. Optimized information for permeated water production amount control information for controlling the amount of water is produced, and the discharge pressure of the pressurizing pump and / or the amount of permeated water produced is controlled based on the produced optimization information.

In another embodiment of the water treatment system according to the present invention, the operation management support system performs cleaning including control information of the membrane cleaning timing and cleaning time of the reverse osmosis membrane device as maintenance timing information based on the prediction result. Optimization information about the timing control information and the replacement timing control information including the control information of the membrane replacement timing of the reverse osmosis membrane device is created, and the optimization is performed based on the prepared cleaning timing control information and replacement timing control information. A warning signal is issued when the membrane cleaning timing or membrane replacement timing of the reverse osmosis membrane device is reached.

In still another embodiment, the water treatment system according to the present invention is arranged on the upstream side of the reverse osmosis membrane device, and is pretreated to obtain supply water for pretreating the water to be treated and supplying it to the reverse osmosis membrane device. A device is further provided, and the operation management support system creates pretreatment device operation control information for optimizing the operation conditions of the pretreatment device, and controls the pretreatment device based on the created pretreatment device operation control information. do.

In still another embodiment of the water treatment system according to the present invention, the pretreatment apparatus includes a coagulant addition means for adding a coagulant to the water to be treated, and the operation management support system predicts the water quality of the water to be treated. Based on the result, the optimization information of the coagulant injection rate to be injected into the water to be treated is prepared, and the coagulant injection rate is controlled based on the prepared optimization information.

In still another embodiment of the water treatment system according to the present invention, the reverse osmosis membrane apparatus includes a plurality of reverse osmosis membrane banks connected in series with each other, an instrument capable of measuring the operating status of the reverse osmosis membrane banks, and the like. The optimization information is created and created so that the operation management support system has the optimum cleaning timing, cleaning time, or replacement timing for each of the multiple reverse osmosis membrane banks. The reverse osmosis membrane device is controlled based on the chemical information.

In another aspect, the present invention is an operation management support system for a water treatment system including a back-penetration membrane device for obtaining concentrated water and permeated water by back-penetrating the supply water, and water quality information of the supply water. Acquisition of back-penetration membrane operation information including at least one of the discharge pressure of the pressurizing pump that pressurizes the supply water supplied to the back-penetration membrane, the flow rate of permeated water, the flow rate of concentrated water, and the flux information of the back-penetration membrane. A learning unit that creates a trained model of a water treatment system using the back-penetrating membrane operation information, and a prediction that predicts the pressure loss of the back-penetrating membrane and / or the water quality of permeated water using the trained model. For the optimization information creation unit and the optimization information that create the operation optimization information of the back-penetrating membrane device including the operating conditions and / or maintenance timing information of the back-penetrating membrane device based on the prediction results of the unit and the prediction unit. Based on this, it is an operation management support system for a water treatment system including a control unit that outputs a control signal for controlling the water treatment system.

In yet another aspect of the present invention, in a method of operating a water treatment system including a back-penetration membrane device for obtaining concentrated water and permeated water by back-penetrating the supply water, water quality information of the supply water and the back-penetration membrane device Reverse permeable membrane operation information is collected and collected, including at least one of the discharge pressure of the pressurizing pump that pressurizes the supply water supplied to the water supply, the flow rate of permeated water, the flow rate of concentrated water, and the flux information of the back permeable membrane. Based on the trained model obtained by machine learning using the permeable membrane operation information, the pressure loss and / or the water quality of the permeated water of the reverse permeable membrane device is predicted, and the operation of the reverse permeable membrane device is performed based on the prediction result. It is an operation method of a water treatment system including creating optimization information including condition and / or maintenance timing information and controlling the water treatment system based on the optimization information.

According to the present invention, the operation of a water treatment system and a water treatment system capable of appropriately evaluating the operating condition in a water treatment system equipped with a reverse osmosis membrane device and optimizing the operating conditions of the water treatment system. It is possible to provide a method of operating a management support system and a water treatment system.

It is the schematic explaining an example of the water treatment system which concerns on embodiment of this invention. It is a schematic diagram which shows the structural example of the operation management support system of the water treatment system which concerns on embodiment of this invention. It is a flow chart which shows an example of the operation method of the water treatment system which concerns on embodiment of this invention. It is the schematic which shows another apparatus mode of the water treatment system which concerns on embodiment of this invention. It is a graph showing the correlation between the predicted value (sim) and the measured value (obs) when the total pressure loss of the reverse osmosis membrane apparatus is predicted using the trained model according to the embodiment of the present invention. It is a graph which shows the comparison result of the predicted value (sim) and the measured value (obs) when the total pressure loss of the reverse osmosis membrane apparatus is predicted using the trained model which concerns on embodiment of this invention. It is a graph which showed the correlation of the predicted value (sim) and the measured value (obs) at the time of predicting the conductivity of permeated water using the trained model which concerns on embodiment of this invention. It is a graph which shows the comparison result of the predicted value (sim) and the measured value (obs) when the conductivity of permeated water is predicted using the trained model which concerns on embodiment of this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The embodiments shown below exemplify devices and methods for embodying the technical idea of the present invention, and the technical idea of the present invention specifies the structure, arrangement, etc. of components as follows. It is not something to do.

(Water treatment system)
As shown in FIG. 1, the water treatment system 1000 according to the embodiment of the present invention comprises a reverse osmosis membrane device 10 that obtains permeable water by treating the supplied water with a reverse osmosis membrane, and pretreating the water to be treated. It includes a pretreatment device 20 that generates water to be supplied to the reverse osmosis membrane device 10, and an operation management support system 30 that supports the operation management of the reverse osmosis membrane device 10 and the pretreatment device 20.

The reverse osmosis membrane device 10 includes a reverse osmosis membrane bank 3 that treats the supplied water with a reverse osmosis membrane. The reverse osmosis membrane bank 3 includes a plurality of reverse osmosis membrane modules containing a plurality of reverse osmosis membrane elements. Although one reverse osmosis membrane bank 3 is shown in the example of FIG. 1, typically, a plurality of reverse osmosis membrane banks 3 can be arranged in series with each other with respect to the flow of supply water. For example, as shown in FIG. 4, by arranging the reverse osmosis membrane banks 3 in multiple stages in series with the flow of the supply water, the concentrated water discharged from the reverse osmosis membrane bank 3 on the upstream side can be effectively used. A large amount of permeated water can be obtained as a whole.

In addition, the configuration in which the concentrated water of the first-stage reverse osmosis membrane bank (also referred to as the first bank and the first stage) is treated by the second-stage reverse osmosis membrane bank (also referred to as the second bank and the second stage) in the subsequent stage is 2 The bank configuration and the configuration in which the concentrated water in the second bank is further treated by the reverse osmosis membrane bank in the third stage are referred to as a three-bank configuration.

As shown in FIG. 1, a pressurizing pump 2 that pressurizes the supply water of the reverse osmosis membrane bank 3 is connected to the upstream side of the reverse osmosis membrane bank 3. A pressure regulating valve 6 and a flushing valve 7 are connected to a pipe for discharging concentrated water from the reverse osmosis membrane bank 3. In order to supply the cleaning tank 4 for accommodating the cleaning chemical solution and the cleaning chemical solution in the cleaning tank 4 to the reverse osmosis membrane to perform cleaning on the pipes on the concentrated water outlet side and the permeated water outlet side of the reverse osmosis membrane bank 3. Is connected to the pump 5.

In order to measure the amount of concentrated water discharged from the reverse osmosis membrane bank 3, a concentrated water flow meter (FI2) 16 is arranged in a pipe on the outlet side of the concentrated water of the reverse osmosis membrane bank 3. In order to measure the amount of permeated water obtained from the reverse osmosis membrane bank 3, a permeated water flow meter (FI1) 17 is arranged in a pipe on the permeated water outlet side of the reverse osmosis membrane bank 3. Further, a measuring meter (CIR) 15 capable of measuring the water quality of the permeated water including the conductivity, the water temperature, etc. is arranged in the pipe through which the permeated water flows.

A pH meter 11 for measuring the pH of the supply water is arranged in the pipe for supplying the supply water to the reverse osmosis membrane bank 3. A thermometer (TIC) 12 for measuring the temperature of the supply water flowing into the reverse osmosis membrane bank 3 is arranged in the pipe on the upstream side of the reverse osmosis membrane bank 3. An inlet pressure gauge (PI1) 13 and an outlet pressure gauge (PI2) 14 are arranged on the inlet side and outlet side pipes of the reverse osmosis membrane bank 3, respectively, and reverse osmosis is caused by the output pressure difference between the pressure gauges 13 and 14, respectively. The pressure loss of the membrane bank 3 can be measured.

The pretreatment device 20 includes a booster pump 21 for pumping raw water to be treated, a chemical supply unit (adding means such as a coagulant) 22 for supplying chemicals such as a coagulant to the water to be treated, and water to be treated. It is provided with a pretreatment means 23 for performing pretreatment for removing impurities such as turbidity contained in the water, and a thermometer (TIC) 24 for measuring the temperature of the water to be treated. The pretreatment means 23 includes, for example, a screen device, a sand filtration device, a coagulation sedimentation device, an aerodynamic levitation device, a membrane separation device, and the like.

(Operation management support system)
The operation management support system 30 according to the embodiment of the present invention is connected to the reverse osmosis membrane device 10 and the pretreatment device 20, and is a general-purpose or dedicated information processing that sends a predetermined operation command based on a predetermined control strategy. The device is available.

For example, as shown in FIG. 2, the operation management support system 30 according to the embodiment of the present invention can control the operation of each device (reverse osmosis membrane device 10, pretreatment device 20) included in the water treatment system 1000. A control device 100, a storage device 110 capable of storing information necessary for various controls, an input device 120, an output device 130, and a communication means 140 can be provided.

A part of the operation management support system 30 does not necessarily have to be arranged in close proximity to the reverse osmosis membrane device 10 and the pretreatment device 20, as shown in FIG. 1, for example, the reverse osmosis through the network 40. It may be connected to the membrane device 10 and the pretreatment device 20. Then, the operation management support system 30 may be configured to be remotely controlled (online control) or automatically operated from another remote support center 1002 (FIG. 2) or the like connected via the network 40.

The operation management support system 30 is communicably connected to the remote support center 1002 or the server 50 via the network 40, and is connected to the trained model created by the operation management support system 30 or the server 50 or another water treatment system 1001. The recorded information may be configured so that it can be transmitted and received to and shared with each other via the communication means 140.

The control device 100 includes an acquisition unit 101, a learning unit 102, a prediction unit 103, an optimization information creation unit 104, a control unit 105, and a warning unit 106. The storage device 110 includes a driving information storage means 111, a learned model storage means 112, a prediction result storage means 113, and an optimized information storage means 114.

The acquisition unit 101 acquires, for example, real-time reverse osmosis membrane operation information in the water treatment system 1000 having the configuration of FIG. 1 and reverse osmosis membrane operation information for the past several decades, and stores the reverse osmosis membrane operation information in the operation information storage means 111. The reverse osmosis membrane operation information includes water quality information of supplied water and permeated water, installation conditions of the reverse osmosis membrane device 10, reverse osmosis membrane device information including operation conditions and operation results, maintenance information of the reverse osmosis membrane device 10, and the like. Is included.

The water quality information of the supply water includes information on the water temperature, pH, silica, metals such as sodium and aluminum, and organic substances of the supply water. The water quality information of the permeated water includes information on the conductivity of the permeated water, the measurement result of metals such as sodium and aluminum in the permeated water, and the like. The installation conditions of the reverse osmosis membrane device 10 include information on the device configuration of the reverse osmosis membrane device 10. For example, the number of stages of the reverse osmosis membrane bank 3 introduced in the water treatment system 1000 according to the present embodiment, the connection relationship when there are a plurality of reverse osmosis membrane banks 3, and the reverse osmosis membrane accommodated in the reverse osmosis membrane bank 3. The number of modules and reverse osmosis membrane elements, the flux of the reverse osmosis membrane, and the characteristic information required to evaluate the characteristics of the reverse osmosis membrane (reverse osmosis membrane dimensions, model number, membrane material, effective filtration area, design inflow flow rate). , Design permeation water flow rate, design permeation flow flux, etc.) are included as information on the device configuration of the reverse osmosis membrane device 10. The operating conditions of the reverse osmosis membrane device 10 include the supply flow rate of the supply water, the discharge pressure of the pressurizing pump 2 for pressurizing the supply water, and the like. The operation results of the reverse osmosis membrane device 10 include various measurement results of the permeated water flow rate, the permeated flux, the concentrated water flow rate, and the permeated water integrated water amount obtained by the reverse osmosis membrane treatment. These reverse osmosis membrane operation information is stored in the operation information storage means 111.

The acquisition unit 101 further obtains measurement results of various instruments (thermometers 12, 24, pH meters 11, pressure meters 13, 14, flow meters 16, 17, water quality meters 15) obtained during operation of the water treatment system 1000. By acquiring it, the operating status of the water treatment system 1000 is monitored.

The learning unit 102 creates a learned model (artificial intelligence model) of the water treatment system by machine learning using a predetermined machine learning algorithm using the reverse osmosis membrane operation information. Examples of machine learning algorithms include PCR method (principal component regression method), PLS method (partial minimum square method), SVR method (support vector regression method), ARIMA, neural network (ANN and RNN) method, and random forest method. An appropriate tool can be appropriately selected and used from various analysis tools using the decision tree method or the like.

The trained model created by the learning unit 102 can include any of a fully connected neural network model, a random forest model, a clustering model, or an RNN model. Above all, in the present embodiment, the learning unit 102 creates a nonlinear model using a neural network method or the like rather than a linear model such as the PCR method, so that the pressure loss and / or transmission of the reverse osmosis apparatus as the objective variable This is preferable because it improves the accuracy of predicting the quality of water.

The parameters of the machine learning model, for example, the number of layers in the fully connected neural network method and the number of neurons in each layer are appropriately changed according to the number of explanatory variables, and are not limited to the following, for example. The number of neurons in each layer can be 1 to 50, further 1 to 30, and the number of layers can be 1 layer, further 1 to 4 layers. The trained model created in this way is configured to receive a predetermined explanatory variable and output a predetermined objective variable. The trained model is stored in the trained model storage means 112.

The prediction unit 103 predicts the operation result of the water treatment system using the trained model. In the present embodiment, the prediction unit 103 includes at least predicting the pressure loss of the reverse osmosis membrane device and / or the water quality of the permeated water. The water quality of the permeated water predicted by the prediction unit 103 includes, for example, the conductivity of the permeated water, the concentration of metals such as sodium (Na), and the like.

According to the present embodiment, predetermined variables as explanatory variables, for example, water quality information of supply water, discharge pressure of a pressurizing pump for pressurizing supply water to be supplied to a reverse osmosis membrane, flow rate of permeated water, flow rate of concentrated water, By using the information on the flux of the reverse osmosis membrane, the pressure loss of the reverse osmosis membrane device and / or the quality of the permeated water can be predicted with high accuracy. The operating conditions in the water treatment system can be evaluated more appropriately. Based on this evaluation result, optimization information of the water treatment system 1000 can be created, and based on this optimization information, at least the operating conditions and maintenance timing information of the water treatment system 1000 can be optimized.

For example, the prediction unit 103 uses (1) the water temperature [° C.] of the supply water flowing into the reverse osmosis membrane bank 3 and (2) the pH [-] of the supply water flowing into the reverse osmosis membrane bank 3 as explanatory variables. 3) Pressure of supply water flowing into reverse osmosis membrane bank 3 [MPa], (4) Flow rate of supply water flowing into reverse osmosis membrane bank 3 [m ³ / h] (5) Replacement of reverse osmosis membrane bank 3 or After cleaning, the cumulative amount of permeated water obtained from the reverse osmosis membrane device 10 [m ³ ], (6) concentrated water flow rate [m ³ / h], and (7) reverse osmosis membrane Using the flux ratio [-], (8) the differential pressure (pressure loss) [MPa] and / or (9) the conductivity of permeated water [μS / cm] of the reverse osmosis membrane bank 3 should be output as objective variables. By executing the calculation using the trained model constructed in, the pressure loss of the reverse osmosis membrane bank 3 and / or the water quality of the permeated water can be predicted.

The prediction unit 103 further adds water quality information such as water quality (concentration of metals such as silica, sodium, calcium, aluminum, organic substances, etc.) of the supply water flowing into the reverse osmosis membrane bank 3 as an explanatory variable to obtain a differential pressure. The effect is that it is possible to more accurately predict the rise in calcium and the occurrence of operational troubles (decrease in recovery rate, formation of scale, etc.). For example, as explanatory variables, in addition to the above-mentioned (1) to (7), (10) silica concentration of supply water, (11) sodium concentration of supply water, (12) calcium concentration of supply water, (13) supply water At least one of the concentration of metals such as aluminum in (14) and the concentration of organic substances in the feed water can be further taken into consideration.

The "reverse osmosis membrane flux ratio" used in the above explanatory variables is represented by the ratio of the measured flux value of the reverse osmosis membrane to the standard flux value (measured flux value / standard flux value). By using this flux ratio as an explanatory variable, the water treatment system to be predicted differs in the configuration (type and number of reverse osmosis membranes) of the reverse osmosis membrane device of the water treatment system operation information acquired by the acquisition unit 101. Even in some cases, predictability can be maintained properly.

The optimization information preparation unit 104 diagnoses the operating state of the reverse osmosis membrane device 10 based on the prediction result of the prediction unit 103, and prepares the optimization information of the water treatment system 1000. Here, the "optimization information" refers to a control condition for optimizing the water treatment system 1000 so that the operation cost can be reduced. In the present embodiment, maintenance timing information and / or operation information of the reverse osmosis membrane device is used. At least include. The maintenance timing information includes cleaning timing control information for controlling the membrane cleaning timing and cleaning time of the reverse osmosis membrane device 10, and replacement timing control information for controlling the membrane replacement timing of the reverse osmosis membrane device 10. .. The operating conditions of the reverse osmosis membrane device 10 include at least discharge pressure control information for controlling the discharge pressure of the pressurizing pump and / or permeated water production amount control information for controlling the water production amount of the permeated water.

In the current operation of the water treatment system 1000, the skill level of the operator engaged in the operation of the water treatment system 1000 is based on the measurement results by various measuring devices arranged in the water treatment system 1000 and the design values of the operating conditions. And the operating conditions have been changed depending on the feeling for many years. However, the causes of troubles in reverse osmosis membrane devices are multiple, and in particular, there are various causes such as a decrease in recovery rate, contamination of permeated water with impurities, or an increase in pressure loss due to membrane fouling of the reverse osmosis membrane and generation of scale. Factors often overlap. Therefore, at present, operation management with a relatively large margin is performed, such as replacing the reverse osmosis membrane module in advance before equipment trouble occurs, and it cannot always be said that the operation is efficient. In addition, even if an expert refers to the driving result, it may not always be possible to make an appropriate judgment from the huge amount of data.

According to the operation management support system 30 of the water treatment system 1000 and the water treatment system 1000 according to the embodiment of the present invention, the optimization information creation unit 104 operates the water treatment system 1000 using the trained model by machine learning. Automatically create optimization information for condition and / or maintenance timing information. For example, regarding the cleaning timing, the cleaning cycle is on the horizontal axis, the operation cost (cleaning cost) is on the vertical axis, and the optimization information creation unit 104 creates a correlation graph based on the prediction result of the prediction unit 103, and the operation cost. The cleaning cycle, which is the lowest point of the above, can be automatically determined, and if necessary, this cleaning cycle can be updated in real time based on the actual operation result. As a result, the maintenance timing information of the reverse osmosis membrane device 10 can be optimized regardless of the skill level of the expert. As a result, the maintenance timing of the reverse osmosis membrane device 10 can be further optimized while suppressing device troubles, and the operating cost can be reduced.

The optimization information can include, for example, the following control information.
・ Cleaning timing control information for controlling the membrane cleaning timing of the reverse osmosis membrane device 10.
-Replacement timing control information for controlling the membrane replacement timing of the reverse osmosis membrane device 10.
Discharge pressure control information regarding control of the discharge pressure of the pressurizing pump that pressurizes the supply water of the reverse osmosis membrane device 10.
-When the reverse osmosis membrane bank 3 is multi-stage, the permeated water production amount control information regarding the control of the permeated water production amount of each stage, and the control information for controlling the membrane cleaning timing and replacement timing of each stage, respectively.

(Washing timing control information)
When the cleaning cycle of the reverse osmosis membrane device 10 is lengthened, the treatment time for obtaining permeated water can be maintained for a long time and the operating time can be increased, but the result is that a load is applied to one cleaning treatment. On the contrary, when the cleaning cycle is shortened, the treatment time for obtaining the permeated water is shortened and the operating time is shortened, while the one-time cleaning treatment is facilitated. The optimization information creation unit 104 comprehensively considers the prediction result of the prediction unit 103, the water quality of the supply water, and the operating conditions, and creates cleaning timing control information including control information for the optimum cleaning cycle and cleaning time. .. This makes it possible to select the optimum membrane cleaning timing according to the water quality of the supply water and the operating conditions. Further, as compared with the conventional case, the deterioration of the reverse osmosis membrane module in the reverse osmosis membrane bank 3 due to the supply of unnecessary cleaning chemicals and the work time required for the cleaning work can be reduced, so that the cost related to the cleaning treatment can be reduced. ..

(Replacement timing control information)
The optimization information preparation unit 104 comprehensively considers the prediction result of the prediction unit 103, the water quality of the supply water, and the operating conditions, and prepares the membrane exchange timing control information including the control information that becomes the optimum membrane exchange cycle. As a result, the optimum membrane replacement timing can be selected according to the water quality of the supply water and the operating conditions, and the life of the reverse osmosis membrane can be extended. Further, in the reverse osmosis membrane apparatus 10, among a plurality of reverse osmosis membrane banks connected in series, a specific reverse osmosis membrane bank, or among a plurality of reverse osmosis membrane banks connected in parallel, a specific reverse osmosis membrane bank It may be exhausted. The optimization information creation unit 104 provides not only information on the exchange frequency for exchanging a plurality of reverse osmosis membrane banks and all reverse osmosis membrane modules based on the prediction result of the prediction unit 103, but also individual reverse osmosis membrane banks and the individual reverse osmosis membrane banks. Control information for optimizing each replacement timing of the reverse osmosis membrane module, that is, replacement with optimized replacement locations and replacement sequences of the first reverse osmosis membrane bank 3a and the second reverse osmosis membrane bank 3b, respectively. The position order control information can be further produced. The exchange position order control information includes the operating conditions of the first reverse osmosis membrane bank 3a and the second reverse osmosis membrane bank 3b (water quality, pressure, amount of water produced, etc.) and / or the first reverse osmosis membrane bank 3a. And by making predictions from the measurement results of instruments such as pressure gauges that can measure the operating status of the second reverse osmosis membrane bank 3b, if there are multiple reverse osmosis membrane banks or modules, the replacement location and replacement order can be determined. Can be optimized. As a result, the replacement of reverse osmosis membrane banks and modules that have not actually reached the replacement life is reduced to extend the life of the reverse osmosis membrane, and the replacement frequency of reverse osmosis membrane banks and modules that need to be replaced is increased. Can be optimized.

(Discharge pressure control information)
Considering the energy consumption of the entire water treatment system of FIG. 1, the pressurizing pump 2 that pressurizes the reverse osmosis membrane module has the largest energy consumption. Further, the higher the concentration of the supplied water, the higher the energy consumption of the pressurizing pump 2 per unit water production amount. The optimization information creation unit 104 can realize energy saving of the pressure pump by optimizing the discharge pressure of the pressure pump 2 based on the prediction result of the prediction unit 103.

(Permeated water production amount control information)
The water treatment system 1000 can improve the recovery rate of permeated water by using a plurality of reverse osmosis membrane banks. It is necessary to increase the pressure of the pressurizing pump 2 when treating the supplied water with high conductivity due to fluctuations in water quality, but reduce the pressurization of the pressurizing pump when treating the supplied water with low conductivity. May be possible. Further, when the reverse osmosis membrane banks are connected in series in multiple stages, the inflow water concentration of the reverse osmosis membrane banks in the subsequent stage becomes high, so that membrane fouling and scale are likely to occur. As a result, the energy consumption per unit water production of the reverse osmosis membrane bank in the subsequent stage is higher than that in the reverse osmosis membrane bank in the previous stage, and the membrane cleaning cycle of the reverse osmosis membrane bank in the latter stage is also shortened, so that the cleaning cost is reduced. May be higher. The optimization information preparation unit 104 determines the energy per unit water production amount of the reverse osmosis membrane bank based on the prediction result of the water quality of the supply water and the pressure loss and / or the water quality of the permeated water of each reverse osmosis membrane bank of the prediction unit 103. For water treatment by reducing the discharge pressure of the pressurizing pump 2 or adjusting the amount of supplied water to optimize the amount of permeated water obtained from each stage of the reverse osmosis membrane so that the amount of consumption can be kept low. Such energy consumption can be reduced.

The optimization information production unit 104 can be further configured to optimize the replacement timing control information, the maintenance control information, the discharge pressure control information, and the permeated water production amount control information. For example, according to the request of the operator, the priority order for optimizing each control information can be set for the above four optimization information via the input device 120 or the like. Then, the optimization information creation unit 104 creates optimization information according to the priority order, so that it is flexible according to the operator's request, the water quality of the supplied water and the permeated water, the characteristics of the reverse osmosis membrane module, and the like. Can drive. The produced optimization information is stored in the optimization information storage means 114. The control unit 105 controls the operation of the water treatment system 1000 based on the produced optimization information.

The warning unit 106 sends a warning signal for outputting a predetermined warning signal to the operator of the water treatment system 1000 via the output device 130. For example, it is configured to output a warning signal at the optimum cleaning timing and replacement timing based on the cleaning timing control information and replacement timing control information produced by the optimization information creation unit 104. By providing the warning unit 106, the operator of the water treatment system 1000 can know the appropriate cleaning timing and replacement timing at an early stage, so that preparations necessary for cleaning and replacement can be performed.

(Predictive control of pretreatment equipment)
The operation management support system 30 can also control the operation conditions of the pretreatment device 20 so that the operation conditions of the pretreatment device 20 are optimized according to the operation status of the reverse osmosis membrane device 10. In that case, the acquisition unit 101 further acquires the pretreatment device operation control information including the water quality information of the water to be treated and the installation condition, operation condition, and operation result of the pretreatment device 20. The learning unit 102 creates a learned model by machine learning using the pretreatment device operation control information and the reverse osmosis membrane operation information. The prediction unit 103 supplies the operating conditions of the chemical supply unit 22 and the pretreatment means 23 of the pretreatment device 20, for example, the chemical supply unit 22 to the water to be treated, using the trained model created by the learning unit 102. The injection rate of the chemical (ratio of the injection amount to the treated water amount), the flow rate of the treated water supplied to the pretreatment means 23, and the like are predicted. The optimization information creation unit 104 further creates pretreatment device operation control information for optimizing the operation conditions of the pretreatment device 20.

For example, when agglomerated sand filtration or agglomerated membrane filtration is used as the pretreatment means 23, the predictor 103 predicts the water quality of the water to be treated, and the optimization information preparation unit 104 adds the coagulant injection rate to the water to be treated. By creating the optimization information of, the coagulant injection rate can be optimized and the water quality of the supplied water can be stabilized. Therefore, membrane fouling and scale generation of the reverse osmosis membrane bank 3 due to water quality fluctuation etc. Can be suppressed. As a result, according to the water treatment system according to the embodiment of the present invention, the number of times of washing and the number of times of replacement of the reverse osmosis membrane bank 3 can be reduced, and efficient water treatment can be performed.

(How to operate the water treatment system)
FIG. 3 shows an example of the operation flow of the operation method of the water treatment system using the operation management support system 30 of the water treatment system 1000 shown in FIG. 1 or 2. As shown in step S1 of FIG. 3, the acquisition unit 101 included in the operation management support system 30 includes the water quality information of the supply water and the permeated water, the installation condition of the reverse osmosis membrane device, the operation condition, and the operation result. The reverse osmosis membrane operation information including at least one of the device information and the maintenance information of the reverse osmosis membrane device is acquired, and the water treatment system operation information is stored in the operation information storage means 111. The acquisition unit 101 further obtains each instrument (thermometer 12, 24, pH meter 11, pressure meter 13, 14, flow meter 16, 17, water quality measuring meter 15) obtained by monitoring the water treatment system 1000 in FIG. 1 during operation. ) Real-time measurement results can be obtained.

In step S2, the learning unit 102 creates a trained model of the water treatment system. The trained model is stored in the trained model storage means 112. Subsequently, in step S3, the prediction unit 103 at least predicts the pressure loss of the reverse osmosis membrane device 10 and / or the water quality of the permeated water, for example, the conductivity of the permeated water, using the reverse osmosis membrane operation information and the learned model. do.

In step S4, the optimization information preparation unit 104 diagnoses the operating state of the reverse osmosis membrane device based on the prediction result of the pressure loss of the reverse osmosis membrane device and / or the water quality of the permeated water, and maintains the reverse osmosis membrane device. Produce optimization information for the water treatment system 1000, including timing. The produced optimization information is stored in the optimization information storage means 114. The control unit 105 controls the reverse osmosis membrane device 10 based on the optimization information produced by the optimization information production unit 104.

According to the water treatment system 1000 and the operation management support system 30 of the water treatment system according to the embodiment of the present invention, a learned model is created using the reverse osmosis membrane operation information of the reverse osmosis membrane device 10, and the prepared learning is performed. The reverse osmosis membrane device 10 is provided by predicting the pressure loss and / or water quality (conductivity) of the reverse osmosis membrane device 10 using the completed model and creating optimization information of the water treatment system based on the prediction result. The operating condition in the water treatment system 1000 can be appropriately evaluated, and the operating conditions of the water treatment system 1000 can be optimized.

FIG. 4 shows another configuration example of the reverse osmosis membrane device 10 when the reverse osmosis membrane banks of FIG. 1 are connected in series in multiple stages. In the example of FIG. 4, an example composed of two stages of the reverse osmosis membrane banks 3a and 3b is shown, but as described above, two or more reverse osmosis membrane banks 3a and 3b may be arranged. Of course. In the example of FIG. 4, for the sake of simplicity, the description of the cleaning tank 4 and the pump 5 used at the time of cleaning is omitted.

The concentrated water in the first reverse osmosis membrane bank 3a is supplied to the second reverse osmosis membrane bank 3b. The second reverse osmosis membrane bank 3b receives the concentrated water of the first reverse osmosis membrane bank 3a and produces concentrated water and permeated water. The permeated water of the first reverse osmosis membrane bank 3a and the permeated water of the second reverse osmosis membrane bank 3b are mixed and sent to the outside of the water treatment system 1000.

In order to measure the amount of concentrated water discharged from the reverse osmosis membrane bank 3b, a flow meter (FI3) 16b is arranged in a pipe on the outlet side of the concentrated water of the reverse osmosis membrane bank 3b. In order to measure the amount of permeated water discharged from the reverse osmosis membrane bank 3b, a flow meter (FI4) 17b is arranged in the pipe on the permeated water outlet side of the reverse osmosis membrane bank 3b. A pressure gauge (PI2) 14a is connected to the pipe on the inlet side of the reverse osmosis membrane bank 3b, and a pressure gauge (PI4) 14b is placed on the pipe on the outlet side of the reverse osmosis membrane bank 3b. The pressure loss of the reverse osmosis membrane bank 3b is caused by the pressure difference between the pressure gauge 13b that measures the pressure of the inflow water flowing into 3b and the pressure gauge 14b that measures the pressure of the concentrated water treated by the reverse osmosis membrane bank 3b. It can be measured.

When creating the trained model of the water treatment system shown in FIG. 4, for example, the following variables can be used as explanatory variables.
(A) Supply water quality (b) Discharge pressure of the pressurizing pump 2 [MPa]
(C) First stage (reverse osmosis membrane bank 3a) permeated water flow rate [m ³ / h]
(D) Second stage (reverse osmosis membrane bank 3b) permeated water flow rate [m ³ / h]
(E) Total flow rate of permeated water [m ³ / h]
(F) After replacing any of the reverse osmosis membrane banks 3a and 3b, the cumulative amount of permeated water obtained in the reverse osmosis membrane banks 3a and 3b [m ³ / h] before the next replacement.
(G) Cumulative amount of permeated water obtained from the reverse osmosis membrane banks 3a and 3b after washing any of the reverse osmosis membrane banks 3a and 3b before the next washing [m ³ / h]
(H) Concentrated water flow rate discharged from the water treatment system 1000 [m ³ / h]
(I) Water temperature of supply water [° C]
(J) pH [-] of supply water flowing into the reverse osmosis membrane banks 3a and 3b, respectively.
(K) First stage (reverse osmosis membrane bank 3a) flux ratio [-]
(L) Second stage (reverse osmosis membrane bank 3b) flux ratio [-]

As the objective variable, for example, the following variables can be used.
(A) First stage (reverse osmosis membrane bank 3a) pressure loss [MPa]
(B) Second stage (reverse osmosis membrane bank 3b) pressure loss [MPa]
(C) Overall pressure loss [MPa]
(D) Permeable water conductivity [μS / cm]

Predicted value (sim) and measured value of total pressure loss when prediction is performed using a neural network with the above-mentioned (a) to (l) as explanatory variables and (A) to (D) as objective variables. An example of a graph showing the correlation with (obs) is shown in FIG. The correlation coefficient between the predicted value and the measured value is 0.987, and it can be seen that an appropriate prediction is performed. Further, FIG. 6 shows a result of comparison using another expression method with the operation time as the horizontal axis using the prediction result and the actual measurement result of FIG. Also in FIG. 6, the predicted value (sim) and the measured value (Obs) are almost the same, and it can be seen that the prediction can be performed with relatively high accuracy.

Similarly, when the above-mentioned (a) to (l) are used as explanatory variables and (A) to (D) are used as objective variables and prediction is performed using a neural network, the predicted value (sim) of the permeated water conductivity is used. ) And the measured value (obs) are shown in FIG. 7 as an example of a graph showing the correlation. The correlation coefficient between the predicted value and the measured value is 0.966, and it can be seen that an appropriate prediction is performed. FIG. 8 shows the results of comparison using another expression method with the operation time as the horizontal axis using the prediction results and the actual measurement results of FIG. 7. Also in FIG. 8, it can be seen that the permeation water conductivity of the reverse osmosis membrane device can be predicted with relatively high accuracy.

As described above, according to the water treatment system 1000 according to the modified example of the embodiment of the present invention, the pressure loss of the reverse osmosis membrane banks 3a and 3b and / / based on the learned model produced by the operation management support system 30. Alternatively, since the conductivity of the permeated water can be appropriately predicted, the operating condition in the water treatment system 1000 provided with the reverse osmosis membrane device 10 can be appropriately evaluated, and the operating conditions of the water treatment system 1000 can be optimized. ..

For example, the second reverse osmosis membrane bank 3b for treating the concentrated water of the first reverse osmosis membrane bank 3a has a higher concentration than the supply water passed through the first reverse osmosis membrane bank 3a. Since troubles such as membrane blockage are more likely to occur than the first reverse osmosis membrane bank 3a, it is necessary to perform cleaning and replacement more frequently. Furthermore, although the cause is unknown depending on the characteristics of the device configuration of the water treatment system 1000, only a specific reverse osmosis membrane bank unintentionally increases the pressure loss as compared with other reverse osmosis membrane banks. As a result, the film life may be shortened or many troubles may occur. According to the water treatment system 1000 according to the present embodiment, a plurality of reverse osmosis membrane banks 3a and 3b are predicted by predicting the pressure loss and the conductivity of the permeated water for each of the reverse osmosis membrane banks 3a and 3b. The optimum cleaning and replacement frequency can be adjusted according to the individual characteristics of the membrane bank.

Thus, although the present invention has been described by the above embodiments, the statements and drawings that form part of this disclosure should not be understood to limit the invention and are based on the above disclosure. Those skilled in the art can carry out various aspects.

For example, in the present embodiment, a method of predicting the pressure loss and / or the conductivity of permeated water of the reverse osmosis membrane module using the above-mentioned trained model has been described, but other than this, existing operation information is used. Various predictions can be made.

For example, when creating a trained model of the water treatment system of the present invention, an abnormal increase in the differential pressure of the reverse osmosis membrane banks 3a and 3b due to fluctuations in raw water quality is predicted based on past operation information and current operation results. Based on the prediction result, optimization information that optimizes maintenance information including information on cleaning timing and replacement timing of the reverse osmosis membrane banks 3a and 3b is produced, and based on the optimization information, FIG. 1 or FIG. It may be configured to operate the water treatment apparatus of 4. As described above, the present invention can be modified and embodied at the implementation stage without departing from the gist thereof.

2 ... Pressurizing pumps 3, 3a, 3b ... Reverse osmotic membrane bank 4 ... Cleaning tank 5 ... Cleaning pump 6 ... Pressure adjusting valve 7 ... Flushing valve 10 ... Reverse osmotic membrane device 11 ... pH meter 12 ... Thermometers 13, 14 , 14a, 14b ... Pressure meter 15 ... Water quality measuring meters 16, 17, 17a, 17b ... Flow meter 20 ... Pretreatment device 21 ... Booster pump 22 ... Chemical supply unit 23 ... Pretreatment means 24 ... Thermometer 30 ... Operation management support System 40 ... Network 50 ... Server 100 ... Control device 101 ... Acquisition unit 102 ... Learning unit 103 ... Prediction unit 104 ... Optimization information production unit 105 ... Control unit 106 ... Warning unit 110 ... Storage device 111 ... Operation information storage means 112 ... Trained model storage means 113 ... Prediction result storage means 114 ... Optimized information storage means
120 ... Input device 130 ... Output device 140 ... Communication means 1000, 1001 ... Water treatment system 1002 ... Remote support center

Claims (8)

A reverse osmosis membrane device that obtains concentrated water and permeated water by treating the supplied water with a reverse osmosis membrane.
At least one of the water quality information of the supply water, the discharge pressure of the pressurizing pump that pressurizes the supply water supplied to the reverse osmosis membrane device, the flow rate of the permeated water, the flow rate of the concentrated water, and the flux information of the reverse osmosis membrane. Based on a learned model obtained by collecting reverse osmosis membrane operation information including the above and using the collected reverse osmosis membrane operation information, the pressure loss of the reverse osmosis membrane device and / or the permeated water The water quality is predicted, optimization information including operating conditions and / or maintenance timing information of the reverse osmosis membrane device is prepared based on the prediction result, and the reverse osmosis membrane is prepared based on the prepared optimization information. A water treatment system equipped with an operation management support system that controls the equipment. Based on the result of the prediction, the operation management support system controls the discharge pressure control information for controlling the discharge pressure of the pressurizing pump and / or the permeated water production amount for controlling the permeated water production amount as the operation conditions. Claim 1 is characterized in that the optimization information for the water amount control information is produced, and the discharge pressure of the pressurizing pump and / or the amount of water produced in the permeated water is controlled based on the produced optimization information. The water treatment system described in. Based on the result of the prediction, the operation management support system includes cleaning timing control information including control information of the membrane cleaning timing and cleaning time of the reverse osmosis membrane device as the maintenance timing information, and the reverse osmosis membrane device. The optimization information about the replacement timing control information including the control information of the membrane replacement timing is prepared, and the optimum membrane of the reverse osmosis membrane device is optimized based on the prepared cleaning timing control information and the replacement timing control information. The water treatment system according to claim 1 or 2, wherein a warning signal is issued when the cleaning timing or the membrane replacement timing comes. Further provided is a pretreatment device which is arranged on the upstream side of the reverse osmosis membrane device and obtains the supply water for pretreating the water to be treated and supplying the reverse osmosis membrane device.
The operation management support system creates pretreatment device operation control information for optimizing the operation conditions of the pretreatment device, and controls the pretreatment device based on the created pretreatment device operation control information. The water treatment system according to any one of claims 1 to 3. The pretreatment apparatus includes a coagulant addition means for adding a coagulant to the water to be treated.
The operation management support system creates optimization information of the coagulant injection rate to be injected into the water to be treated based on the prediction result of the water quality of the water to be treated, and the optimization of the prepared coagulant injection rate. The water treatment system according to claim 4, wherein the coagulant injection rate is controlled based on the information. The reverse osmosis membrane device
With multiple reverse osmosis membrane banks connected in series with each other,
An instrument that can measure the operating status of the reverse osmosis membrane bank,
With
The optimization information is created so that the operation management support system has the optimum cleaning timing, cleaning time, or replacement timing for each of the plurality of reverse osmosis membrane banks. The water treatment system according to any one of claims 1 to 5, wherein the operation of the reverse osmosis membrane device is controlled based on the information. It is an operation management support system of a water treatment system equipped with a reverse osmosis membrane device that obtains concentrated water and permeated water by reverse osmosis membrane treatment of supplied water.
At least information on the quality of the supply water, the discharge pressure of the pressurizing pump that pressurizes the supply water supplied to the reverse osmosis membrane device, the flow rate of the permeated water, the flow rate of the concentrated water, and the flux information of the reverse osmosis membrane. An acquisition unit that acquires reverse osmosis membrane operation information including any of them,
A learning unit that creates a trained model of the water treatment system using the reverse osmosis membrane operation information, and
Using the trained model, a prediction unit for predicting the pressure loss of the reverse osmosis membrane and / or the water quality of the permeated water, and
An optimization information creation unit that creates optimization information for the reverse osmosis membrane device, including operating conditions and / or maintenance timing information for the reverse osmosis membrane device, based on the prediction results of the prediction unit.
A control unit that outputs a control signal for controlling the water treatment system based on the optimization information, and a control unit.
Operation management support system for water treatment system. In the operation method of a water treatment system provided with a reverse osmosis membrane device for obtaining concentrated water and permeated water by reverse osmosis membrane treatment of supplied water.
At least one of the water quality information of the supply water, the discharge pressure of the pressurizing pump that pressurizes the supply water supplied to the reverse osmosis membrane, the flow rate of the permeated water, the flow rate of the concentrated water, and the flux information of the reverse osmosis membrane. Collect reverse osmosis membrane operation information including
Based on the learned model obtained by machine learning using the collected reverse osmosis membrane operation information, the pressure loss of the reverse osmosis membrane and / or the water quality of the permeated water is predicted.
Based on the prediction results, optimization information including operating conditions and / or maintenance timing information of the reverse osmosis membrane device is created.
A method of operating a water treatment system, which comprises controlling the water treatment system based on the optimization information.

Images