WO2018070548A1

WO2018070548A1 - Filtering device

Info

Publication number: WO2018070548A1
Application number: PCT/JP2017/038095
Authority: WO
Inventors: 眞佐夫山田; 勇輝宮里
Original assignee: 株式会社日建
Priority date: 2016-10-14
Filing date: 2017-10-12
Publication date: 2018-04-19
Also published as: JP6917635B2; CN211514099U; JPWO2018070548A1

Abstract

The purpose of the present invention is to solve a problem in which a washing function is degraded by the washing of a tubular-type filter membrane. When a liquid to be filtered is filtered by being pressure-fed into a UF membrane tube 15 in a tubular-type membrane separation device 10, a sponge ball is made to flow inside the UF membrane tube 15 by the liquid to be filtered, so that deposits attached to the inner surface of the UF membrane tube 15 are cleaned out by the sponge ball. To ensure that a passageway 125 for the liquid to be filtered moving inside the tubular-type membrane separation device 10 is longer than the length from the end on one side of the tubular-type membrane separation device 10 to the end on the other side, the passageway 125 is curved inside the tubular-type membrane separation device 10. A reception tool 18 is provided for receiving the sponge ball, which has moved inside the UF membrane tube 15 and come out from a liquid outlet 118 of the tubular-type membrane separation device 10, while allowing a concentrate to pass through. The reception tool 18 has a detachable cap for enabling the received sponge ball 17 to be taken out and replaced.

Description

Filtration processing equipment

The present invention, for example, filtration of emulsion waste liquid discharged from factories, filtration of viruses and endotoxins during production of pharmaceuticals and medical water, filtration of heat-sensitive substances such as proteins and enzymes, etc. The present invention relates to a filtration apparatus that separates liquid. Specifically, the present invention relates to a filtration apparatus that filters a liquid to be filtered by a cross-flow method using a tubular filtration membrane and separates it into a treated liquid and a concentrated liquid.

As this type of filtration apparatus, for example, there is one described in Patent Document 1. In this filtration apparatus, the liquid to be filtered is filtered into the treated liquid (membrane permeated water) and the concentrated liquid by pumping the liquid to be filtered (raw water) into the tubular membrane module. A specific configuration will be described with reference to FIG.
First, referring to FIG. 29A, raw water introduction and

discharge ports

210 and 212 are provided at the left and right ends of the pressure vessel 216. The inside of the pressure vessel 216 is divided into three regions by

partition members

222 and 224. A plurality of tubular membranes 218 are provided in parallel in the central region. The left and right regions are configured in a merge chamber 230 in which concentrated liquids that have passed through the plurality of tubular membranes 218 are merged. The raw water introduced from one of the inlet /

outlet ports

210 and 212 enters the merge chamber 230 and then is supplied to each tubular membrane 218, and the concentrated liquid that has passed through each tubular membrane 218 is received by the merge chamber 230. After joining, it is discharged from the other of the

introduction discharge ports

210 and 212.
The membrane permeated water (treated solution) that passed through the membrane and the concentrated solution that did not pass through the membrane while the liquid to be filtered fed to the end of each of the plurality of tubular membranes 218 moved to the other end. And separated. The membrane permeated water (treated liquid) is discharged from the discharge port 214.
As a result of such a filtration treatment, a contaminant or the like of the liquid to be filtered adheres to the inner surface of the plurality of tubular membranes 218. As a method of cleaning this deposit, in the filtration apparatus of Patent Document 1, a wiper body composed of a sponge ball 220 for cleaning is placed in a plurality of tubular membranes 218 one by one, and the tubular membrane 218 is placed. The inflow direction of the raw water introduced into the inside is switched, and the sponge balls 220 are reciprocated in the tubular membranes 218, whereby the inside of the membranes is wiped and cleaned with the sponge balls.
In the filtration apparatus described in Patent Document 1, a cleaning material retaining mechanism 222 is provided at both ends of each tubular membrane 218 (see FIG. 29B), and the sponge that has moved through the tubular membrane 218 has been moved. The ball 220 is received and prevented from falling off the end of the tubular membrane 218. The cleaning material retaining mechanism 222 is configured so that the liquid to be filtered can pass even when the sponge ball 220 is received.

JP-A-6-304455

The filtration apparatus described in Patent Document 1 has a drawback in that the contaminants and the like adhere to the sponge ball itself as a wiping body by reciprocating in the tubular membrane module, and the cleaning function is lowered.
The present invention has been conceived in view of the actual situation, and an object thereof is to provide a filtration apparatus capable of solving the problem that the cleaning function is reduced due to contamination of the wiping body itself.

The specific contents of the embodiment corresponding to the means for solving the problem are shown below in parentheses.
The present invention is a filtration apparatus that separates a liquid to be filtered and a concentrated liquid by filtering the liquid to be filtered by a cross-flow method using a tubular filtration membrane (for example, a UF membrane tube 15),
The tubular filtration membrane is built-in, and a filtration processing unit (for example, the tubular membrane separation device 10) for flowing and filtering the liquid to be filtered in the tubular filtration membrane;
Cleaning means (for example, a membrane cleaning tank 8, a device control panel 23, three-

way valves

36, 37, and 38, a receiving mechanism 18 and a sponge ball 17) for cleaning deposits adhering to the inner surface of the tubular filtration membrane by filtration treatment; With
The cleaning means moves a wiping body (for example, sponge ball 17) in contact with the inner surface of the tubular filtration membrane to wipe off deposits adhering to the inner surface of the tubular filtration membrane,
It further includes an exchange mechanism (for example, a receiving mechanism 18 and a detachable cap 20) for moving the inside of the tubular filtration membrane to exchange the wiping body that has come out of the filtration processing unit.
Moreover, the said filtration process part has the entrance / exit (for example, the liquid in / out 118 of FIG. 2) of the filtration object liquid formed in the 1st place and the 2nd place, and the filtration object which approached from one said entrance / exit The movement path (for example, movement path 125) until the liquid reaches the other entrance / exit is configured to be one (for example, FIG. 2),
The wiping body is moved by a liquid cross flow and wipes off deposits adhering to the inner surface of the tubular filtration membrane.
The exchange mechanism receives the wiping body that moves through the tubular filtration membrane and comes out of the entrance and exit, and also allows the concentrated liquid to pass therethrough (for example, the receiving mechanism 1). )
The wiping body receiving mechanism may have a take-out mechanism (for example, a detachable cap 20) for taking out the received wiping body and making it replaceable.
In addition, the movement path (for example, movement path 125) may be bent in the filtration processing unit such that the movement route (for example, movement path 125) is longer than the entire length of the filtration processing unit (for example, FIG. 2).
Further, the cleaning means is a cleaning liquid cross flow means (for example, the membrane cleaning tank 8, cleaning process switching process in FIG. 24 (a), FIG. 24 (a), for flowing a cleaning liquid along the tubular filtration membrane. b) a process flow path switching process).
Further, a treated liquid storage tank (for example, a membrane cleaning tank 8) for collecting and storing the treated liquid is further provided,
The cleaning liquid crossflow means has a processed liquid use means (for example, S79, motor valve MV6) for using the processed liquid stored in the processed liquid storage tank as the cleaning liquid. May be.
Also, a process for switching from a filtration process (for example, FIGS. 1 and 10) for filtering the liquid to be filtered by the filtration processing unit to a cleaning process (for example, FIGS. 12 and 13) for cleaning by flowing the cleaning liquid. Switching means (for example, the process switching process of FIG. 22B, the process switching process of FIG. 23B);
When switching from the filtration step to the washing step, a flushing means (for example, a water pushing step flow path control process in FIG. 23C) for flushing the remaining filtration target liquid with the washing liquid; May be further provided.
Moreover, you may further provide the time adjustment means (for example, S182, S231) for adjusting the execution time of the filtration process which filters the filtration object liquid by the said filtration process part, and the washing | cleaning process by the said washing | cleaning means.
Further, a concentration circulation tank (for example, the concentration circulation tank 2) for storing the filtration target liquid and receiving and storing the concentrated liquid after the filtration target liquid is supplied to the filtration processing unit and filtered.
The filtration target liquid in the concentration circulation tank is concentrated by the filtration by the filtration processing part as the filtration target liquid returns from the concentration circulation tank to the concentration circulation tank via the filtration processing part. In response to being reduced to a predetermined storage amount (for example, L1 level), a stock solution addition means (for example, raw water tank 1, raw solution water supply pump 27, pH adjustment) for adding the stock solution into the concentration circulation tank is added. Tank 5, pH adjustment water pump 30, S162, S164, S30, S31 to S34),
When the number of times of addition by the stock solution addition means reaches a predetermined number (for example, the number of times of concentration NK set in advance by S61) (for example, YES by S163), it is stored in the concentration circulation tank. Concentrated liquid extracting means (for example, S167, S251 to S254) for extracting the concentrated liquid may be further provided.
Further, machine learning means for performing machine learning for acquiring knowledge adapted to each filtration environment (for example, filtration environments A, B,... In FIG. 26) classified according to the type of the filtration target liquid ( For example, an artificial intelligence server 55) is further provided,
The machine learning means inputs data that can specify the filtration efficiency of the liquid to be filtered by filtration in the filtration processing unit as a state (for example, state data s in FIG. 26) with respect to the filtration environment, and also adds the filtration efficiency to the filtration efficiency. Reinforcement learning means (for example, performing reinforcement learning for improving the filtration efficiency by repeating the input and output by outputting the affecting control as an action on the filtration environment (for example, action data a in FIG. 26). The reinforcement learning process in FIG. 26B may be included.

According to the present invention, the wiping body that passes through one tubular filtration membrane provided in one filtration processing section can be replaced outside the filtration processing section, and the wiping body itself is cleaned along with dirt. The problem of reduced functionality can be solved.

FIG. 1 is an overall configuration diagram of the filtration apparatus in the state of the concentration step 1.
FIG. 2 is a schematic view of a tubular membrane separator.
FIG. 3 is an exploded perspective view of the main part of the tubular membrane separator.
FIG. 4 is a diagram showing the internal structure near the end of the tubular membrane separator.
FIG. 5 is a cross-sectional view of an essential part for explaining the sealing function of the tubular membrane separator.
6A is a cross-sectional view of the end cap on the input / discharge side, and FIG. 6B is a cross-sectional view of the end cap on the return side.
FIG. 7A is a cross-sectional view showing a holding state of a plurality of pipes in the housing, and FIG. 7B is a view showing a holding state of the plurality of pipes in a state where the housing is removed.
FIG. 8 is a diagram showing the principle of filtration with a UF membrane tube and the cleaning action of the inner surface of the UF membrane tube with a sponge ball.
FIG. 9 is a view showing a receiving mechanism for holding a sponge ball with a strainer.
FIG. 10 is a configuration diagram of the filtration apparatus in the state of the concentration step 2.
FIG. 11 is a configuration diagram of the filtration apparatus in the state of the water pushing step.
FIG. 12 is a configuration diagram of the filtration apparatus in the state of the cleaning process 1.
FIG. 13 is a configuration diagram of the filtration apparatus in the state of the cleaning process 2.
FIG. 14 is a block diagram showing a control circuit of the pH adjustment control panel and the apparatus control panel.
(A) of FIG. 15 is a flowchart which shows the main routine of pH adjustment control processing, (b) is a flowchart which shows the subroutine program of each tank abnormality check process.
(A) of FIG. 16 is a flowchart which shows the subroutine program of each inter-tank liquid movement process, (b) is a flowchart which shows the subroutine program of pH adjustment process.
FIG. 17A is a flowchart showing a main routine of the apparatus control process, and FIG. 17B is a flowchart showing a subroutine program of the operation set value input process.
FIG. 18A is a flowchart showing a continuation of the subroutine program for the operation set value input process, and FIG. 18B is a flowchart showing the subroutine program for the abnormal set value input process.
FIG. 19 is a flowchart showing a subroutine program for the abnormality check process.
FIG. 20 is a flowchart showing the continuation of the subroutine program for the abnormality check process.
FIG. 21A is a flowchart showing the continuation of the subroutine program for the flow path switching process, and FIG. 21B is a flowchart showing the continuation of the subroutine program for the stock solution supply process.
22A is a flowchart showing the continuation of the subroutine program for the concentration process, FIG. 22B is a flowchart showing the continuation of the subroutine program for the process switching process, and FIG. It is a flowchart which shows a subroutine program.
FIG. 23A is a flowchart showing the continuation of the subroutine program for the water pushing process, FIG. 23B is a flowchart showing the continuation of the subroutine program for the process switching process, and FIG. It is a flowchart which shows the continuation of the subroutine program of a process.
24A is a flowchart showing a continuation of the subroutine program for the cleaning process, FIG. 24B is a flowchart showing a continuation of the subroutine program for the process switching process, and FIG. 24C is a flowchart of the cleaning process flow path control process. It is a flowchart which shows the continuation of a subroutine program.
FIG. 25A is a flowchart showing a continuation of the subroutine program for the concentrate discharge process, and FIG. 25B is a flowchart showing a continuation of the subroutine program for the monitor display process.
FIG. 26 is a diagram showing an overall system in an embodiment to which artificial intelligence is applied.
FIG. 27A is a block diagram showing the control circuit of the artificial intelligence server, FIG. 27B is a diagram for explaining the principle of reinforcement learning by the agent, and FIG. 27C shows the main routine of control by the artificial intelligence server. It is a flowchart to show.
FIG. 28A is a flowchart showing a subroutine program for filtering environment classification processing, and FIG. 28B is a flowchart showing a subroutine program for reinforcement learning processing.
FIG. 29 shows a conventional example, (a) is a longitudinal sectional view of a tubular membrane module, and (b) is a longitudinal sectional view showing a cleaning material retaining mechanism.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The filtration apparatus shown in the present embodiment, for example, filtration of emulsion waste liquid discharged from various factories, filtration of viruses and endotoxins during the production of pharmaceuticals and medical water, filtration of heat-sensitive substances such as proteins and enzymes, etc. And is used when separating into a treated liquid and a concentrated liquid. For example, washing wastewater, cutting wastewater, polishing wastewater, compressor drain wastewater, car wash wastewater, high frequency coolant wastewater, die casting wastewater, rolling coolant wastewater, wastewater from food factories, etc. are the main applications. More specifically, it is used when, for example, a filtration target liquid is filtered and separated into treated water and concentrated liquid by tubular membrane separation using a UF membrane tube. The filtration apparatus according to the present invention is a wide concept including not only a single apparatus but also a system in which a plurality of apparatuses and computers are linked via a network, for example, as shown in FIG.
With reference to FIG. 1, the whole structure of a filtration apparatus is demonstrated. The filtration device stores a raw water tank 1 for storing a stock solution such as a factory waste liquid, a pH adjustment tank 5 for adjusting the pH of the stock solution (hereinafter simply referred to as “pH”), and a filtered and concentrated concentrate. Accepts the treated water filtered by the membrane cleaning tank 8 and the tubular membrane separator 10, storing the cleaning water for cleaning the adhering matter adhering to the UF membrane tube 15 of the concentration circulation tank 2 and the tubular membrane separator 10. The treated liquid relay tank 11 that relays the concentrated liquid, the concentrated liquid relay tank (also referred to as a waste liquid relay tank) 3 that receives and relays the concentrated liquid stored in the concentrated circulation tank 2, and the concentrated liquid stored in the concentrated liquid relay tank 3 A concentrated water tank 4 is provided for receiving and storing the water. In FIG. 1, the emulsion separator is shown surrounded by a two-dot chain line.
First, a stock solution such as a factory waste liquid is supplied to the raw water tank 1. The raw water tank 1 is provided with a level sensor LS4, and the amount of stock solution stored in the raw water tank 1 is detected. If the level sensor LS4 detects that the raw water tank 1 is full, the supply of the raw solution is stopped.
By operating the raw liquid feed pump 27, the raw liquid stored in the raw water tank 1 is supplied to the pH adjustment tank 5 through the pipe 80. The pH adjustment tank 5 is provided with a level sensor LS6, and the stock amount of the stock solution in the pH adjustment tank 5 is detected. When the pH adjustment tank 5 is filled with the stock solution, the operation of the pump 27 is stopped. The pH adjustment tank 5 is provided with a pH sensor 24. This pH sensor 24 detects the pH of the stock solution stored in the pH adjustment tank 5. When the pH sensor 24 detects that the stock solution is alkaline, the acid injection pump 29 is operated to inject the acid in the chemical injection tank 7 into the pH adjustment tank 5. Conversely, when the pH sensor 24 detects that the stock solution is acidic, the alkali injection pump 28 operates to inject the alkali in the chemical injection tank 6 into the PH adjustment tank 5. Thereby, the stock solution is neutralized in the pH adjusting tank 5 and adjusted to a neutral state.
Since PVDF (polyvinylidene fluoride) is used as the material of the UF membrane tube 15, a stock solution in the range of pH 2 to pH 12 can be filtered. In the case of an acidic undiluted solution having a pH of less than 2, there is a problem that the metal pipe corrodes and the undiluted solution leaks. In the case of an alkaline stock solution exceeding pH 12, there is a disadvantage that the UF membrane tube 15 deteriorates. By neutralizing the undiluted solution in the pH adjusting tank 5, the above inconvenience is prevented. Further, since the material of the UF membrane tube 15 is PVDF (polyvinylidene fluoride), it is possible to filter the stock solution up to an upper limit of 60 ° C., and there is an advantage that the heat resistance to the stock solution is high. The PH adjusting tank 5 is provided with a stirring propeller 50. By rotating the stirring propeller 50 by a stirring motor 35, the stock solution in the pH adjusting tank 5 is stirred and neutralized uniformly.
A pipe 81 connecting the pH adjustment tank 5 and the concentration circulation tank 2 is provided with a pH adjustment water pump 30. By driving the pH adjustment water pump 30, the stock solution neutralized in the pH adjustment tank 5 is supplied to the concentration circulation tank 2 through the pipe 81. The concentration circulation tank 2 is provided with a level sensor LS1, and the stock amount of the stock solution in the concentration circulation tank 2 is detected. When the stock solution in the concentration circulation tank 2 reaches the operation start level, the operation of the pH adjustment water pump 30 is stopped, and the circulation pump 33 is operated so that the stock solution (concentration liquid) in the concentration circulation tank 2 is tubular membrane separation. Supplied to the apparatus 10. At this time, in FIG. 1, the three-way valve 36 whose flow path is switched by the motor valve MV3 is a B → C flow path. As a result, the undiluted solution (concentrated liquid) in the concentration circulation tank 2 flows in the direction of B → C at the three-way valve 36 via the pipe 82, the motor valve MV 1, the circulation pump 33, and the pipe 83, and via the pipe 84. A plurality (six in the figure) of tubular membrane separators 10 are supplied.
The tubular membrane separation device 10 filters the stock solution (concentrated solution) and separates it into treated water and concentrated solution. The treated water is taken out from the treated water take-out pipe 21 and supplied to the treated water relay tank 11 via the motor valve MV5 and the pipe 91. The treated water stored in the treated water relay tank 11 is purified by activated

carbon

12 and 13 by operating the treated water delivery pump 34 and then discharged to the outside. Instead of or in addition to purification with activated carbon, ions contained in the treated water (electrolyte solution) may be removed using ion exchange. Ion exchange refers to ions (for example, ammonia ions NH) contained in the electrolyte solution in contact with a certain substance. ₄ (+) Is taken in and ion species are exchanged by releasing other types of ions instead. A substance exhibiting an ion exchange action is called an ion exchanger. As this ion exchanger, for example, an ion exchange resin is used.
In addition, you may control so that the treated water by which the undiluted | stock solution (concentrated liquid) was filtered by the tubular type | formula membrane separator 10 may be supplied to the membrane washing tank 8 instead of supplying to the treated water relay tank 11. FIG. Specifically, the process water is supplied to the membrane cleaning tank 8 via the pipe 90 by controlling the motor valve MV5 to close and opening the motor valve MV6. In this way, there is an advantage that treated water can be stored in the membrane cleaning tank 8 and used as cleaning water even in a place where fresh water such as tap water, industrial water, and agricultural water is difficult to obtain (for example, a desert area).
On the other hand, the concentrated solution filtered and concentrated by the tubular membrane separator 10 is returned to the concentration circulation tank 2 via the three-way valve 38. Specifically, in FIG. 1, the flow path of the three-way valve 37 switched by the motor valve MV4 is changed from C to A. A three-way valve 38 switched by the motor valve MV9 is a C → B flow path. As a result, the concentrated solution filtered and concentrated by the tubular membrane separator 10 passes through the three-way valve 37, the pipe 87, the three-way valve 38, the pipe 89 and through the flow rate adjustment valve 48 to the concentration circulation tank 2. Reduced.
By operating the circulation pump 33 with the flow rate adjusting valve 48 adjusted, the stock solution (concentrated liquid) can be pumped into the tubular membrane separator 10 while maintaining an appropriate pressure. Thus, filtration into the UF membrane tube 15 becomes possible. In order to maintain the appropriate pressure, pressure sensors PS1 and PS2 are provided. Adjusting the flow rate adjusting valve 48 based on the detected values of the pressure sensor PS1 provided on the upper side of the tubular membrane separator 10 and the pressure sensor PS2 provided on the lower side of the tubular membrane separator 10 Thus, the UF membrane tube 15 is filtered with an appropriate pressure. In the case of the filtration step shown in FIGS. 1 and 10, the motor valve MV2 is closed.
This filtration device is controlled by a pH adjustment control panel 22 and a device control panel 23. The control territory of the pH adjustment control panel 22 is in the range indicated by the two-dot chain line in FIG. 1 and mainly performs pH adjustment. On the other hand, the control territory of the device control panel 23 is in the range indicated by the two-dot chain line in FIG. 1, and mainly controls the filtration by the tubular membrane separator 10 and the cleaning of the tubular membrane separator. The pH adjustment control panel 22 and the apparatus control panel 23 are configured to be able to communicate with each other. A raw solution supply signal is transmitted from the apparatus control panel 23 to the pH adjustment control panel 22, and the pH adjustment control panel 22 transmits the apparatus control panel. A batch abnormality signal and a raw water supply signal are transmitted to 23. The apparatus control panel 23 outputs a raw water supply signal, a collective abnormality signal, an operation signal, and an abnormality signal.
In FIG. 1, FL1, FL2, and FL3 are flow meters, 26 is a temperature sensor, 25 is a photoelectric sensor, and 40, 41, 42, 43, and 44 are check valves.
A motor valve MV8 is provided in a pipe connecting the pipe 97 and the concentration circulation tank 2. This motor valve MV8 prevents the UF membrane tube 15 in the tubular membrane separator 10 from becoming a negative pressure. Details will be described later.
At both ends of the tubular membrane separator 10, receiving mechanisms 18 that hold the sponge balls 17 are provided. The sponge ball 17 is composed of a spherical sponge and wipes and removes impurities adhering to the inner surface of the UF membrane tube 15. The sponge ball 17 is a foamed urethane material. In the case of a urethane material compared to a rubber material, there are advantages that resistance to oil is high and expansion can be suppressed.
Next, the tubular membrane separation apparatus 10 will be described in detail with reference to FIGS. First, an outline of the tubular membrane separation apparatus 10 will be described with reference to FIG. In the tubular membrane separation apparatus 10, a plurality of (for example, 18) UF membrane tubes 15 are provided in a housing 14. End caps 101 are provided at the left and right ends of the housing 14. A one-dot chain line in FIG. 2 indicates a movement path 125 of the liquid to be filtered (raw solution) flowing in the UF membrane tube 15. The UF membrane tube 15 is inserted into the pipe 106 (see FIG. 3). The left and right end caps 101 are formed with return paths 117 that cause the flow of the liquid to be filtered received from a certain UF membrane tube 15 to U-turn and flow into other UF membrane tubes 15. This will be described later with reference to FIG. Further, a liquid inlet / outlet port 118 is formed at two positions on the left end cap 101 in FIG.
The stock solution is supplied to one liquid inlet / outlet port 118 via the receiving mechanism 18 and the valve 10, and sent into the UF membrane tube 15 through the inlet / outlet path 110 a in the end cap 101. The fed stock solution passes through the UF membrane tube 15 and reaches the other end cap 101, where it makes a U-turn through the return path 117 and is fed into the next UF membrane tube 15. The stock solution that has passed through the UF membrane tube 15 and has reached the other end cap 101 passes through the return path 117 again to make a U-turn and is sent into the next UF membrane tube 15. The concentrated liquid that has passed through all the UF membrane tubes 15 by repeating this U-turn is discharged from the other liquid inlet / outlet 118 through the inlet / outlet path 110a. Such a moving path 125 of the stock solution in the tubular membrane separation apparatus 10 is shown by a one-dot chain line in FIG. The stock solution moves along the movement path 125 that is bent and reciprocates in the housing 14 a plurality of times, and is filtered by the UF membrane tube 15 to be separated into treated water and concentrate. The treated water is discharged from the treated water take-out pipe 21 provided in the housing 14 and sent to the treated water relay tank 11. The concentrated liquid is discharged from the liquid inlet / outlet 118 and is returned to the concentrated circulation tank 2.
As shown in FIG. 2, since the movement path 125 is bent in the tubular membrane separator 10, the path through which the stock solution is filtered by the UF membrane tube 15 is defined as the full length of the tubular membrane separator 10 (one end cap). The length from the outer end surface of 101 to the outer end surface of the other end cap 101 can be made longer. As a result, there is an advantage that efficient filtration can be performed. The means for lengthening the movement path 125 is not limited to the reciprocation of the movement path 125 shown in FIG. Any shape may be used.
Next, the internal structure of the tubular membrane separator 10 will be described mainly with reference to FIG. FIG. 3 shows a state in which the end cap 101 provided at one end of the housing 14 is removed and various components in the housing 14 are pulled out. A pipe holding plate 105 and a disk-shaped metal plate 103 are accommodated in the housing 14. A rubber plate 104 for sealing is attached to the surface of the metal plate 103 on the pipe holding plate 105 side. As described above, a plurality of UF membrane tubes 15 are provided in the housing 14, and each of the UF membrane tubes 15 is inserted into a metal pipe 106. The UF membrane tube 15 is formed in a tube shape by winding a belt-shaped nonwoven fabric in a spiral shape and heat-welding the tube, and the inner surface of the tube is coated with a UF membrane.
A pipe holding plate 105 holds the end of each pipe 106 into which the UF membrane tube 15 is inserted. A plurality of pipe insertion holes are formed in the pipe holding plate 105, and the end portions of the pipes 106 are inserted into the insertion holes. Each pipe 106 in the inserted state is welded and fixed to the pipe holding plate 105. Thereby, a plurality of (18) pipes 106 are held by the pipe holding plate 105.
Insertion holes 109 corresponding to the number of pipes 106 are formed in the rubber plate 104 and the metal plate 103 so that the insertion joints 108 can be inserted into the insertion holes 109 from the metal plate 103 side. The insertion joint 108 is made of a material having a sealing function, such as rubber, in a mushroom shape, and has an insertion portion 108a and an umbrella portion 108b having a saw-shaped cross section. The insertion portion 108a of the insertion joint 108 is inserted from the opposite side (the metal plate 103 side) in a state where each pipe 106 is aligned with the insertion hole 109, whereby the saw-like insertion portion 108a becomes the UF membrane tube 15 in the pipe 106. It will be in the state inserted in. In this state, the umbrella part 108b comes into contact with the metal plate 103 (see FIGS. 4 and 5). When the insertion portion 108a is inserted into the UF membrane tube 15, it can be easily inserted by inserting the UF membrane tube 15 in a state where it is slightly pulled out from the pipe 106. The saw-shaped insertion portion 108a inserted into the UF membrane tube 15 has an effect of preventing the UF membrane tube 15 from coming off. That is, the end portion of the UF membrane tube 15 can be sandwiched and held by the insertion portion 108a of the insertion joint 108 and the end portion of the pipe 106. The insertion portion 108a is inserted into both ends of the pipe 106 (see FIG. 5), and the ends of the UF membrane tube 15 are held at both ends of the pipe 106. As a result, it is possible to prevent the UF membrane tube 15 near the end of the pipe 106 from being displaced toward the center of the pipe 106, and the concentrated liquid (raw solution) leaks outside through the small hole 120 of the pipe 106 from the displaced position. Inconvenience can be prevented. A fixed holding member (fixed holding means) for fixing and holding the end portion of the UF membrane tube 15 to the end portion of the pipe 106 is configured by the insertion portion 108 a of the insertion joint 108. An O-ring groove 119 is formed on the outer periphery of the pipe holding plate 105, and a sealing O-ring 111 is fitted in the groove 119.
Each pipe 106 is provided with a large number of small holes 120 in one or several rows along the longitudinal direction. The treated water is discharged into the housing 14 through the small hole 120. The treated water discharged into the housing 14 is discharged out of the tubular membrane separation apparatus 10 from the treated water extraction pipe 21 (see FIG. 2).
Positioning

protrusions

112a and 112b are formed on the pipe holding plate 105. The metal plate 103 and the rubber plate 104 are formed with positioning insertion holes 113a and 113b. Further, positioning holes 114 a and 114 b are formed in the end cap 101. By inserting the

protrusions

112a and 112b into the

positioning holes

114a and 114b in a state where the

protrusions

112a and 112b are inserted into the positioning insertion holes 113a and 113b, the pipe holding plate 105, the metal plate 103 with the rubber plate 104, and the end cap 101 are positioned relative to each other. Is done.
Further, a stud bolt 107 is provided at the center of the pipe holding plate 105. An insertion hole 115 for inserting the stud bolt 107 is formed in the end cap 101. An insertion hole 150 for inserting the stud bolt 107 is also formed in the rubber plate 104 and the metal plate 103. After the stud bolt 107 is inserted into both the insertion holes 115 and 150, the end cap 101, the metal plate 103 with the rubber plate 104, and the pipe holding plate 105 are positioned by the

positioning protrusions

112a and 112b. In this state, the nut 102 is screwed and screwed into the stud bolt 107, whereby the end cap 101, the metal plate 103 with the rubber plate 104, and the pipe holding plate 105 are fastened and fixed in close contact with each other. The state will be described with reference to FIG.
FIG. 4 is a longitudinal sectional view of the vicinity of the end of the tubular membrane separator 10. The plurality of pipes 106 are inserted to such a depth that the end portions reach the end surface of the pipe holding plate 105. In this state, the ends of the pipe 106 and the UF membrane tube 15 are brought into pressure contact with the rubber plate 104 by the tightening force of the stud bolt 107 and the nut 102. An insertion joint 108 is inserted into the UF membrane tube 15 pressed against the rubber plate 104.
A plurality of recesses 110 are formed by recessing the periphery of the inlet / outlet portion of the stock solution to the plurality of return paths 117 (see FIGS. 2 and 6) formed in the end cap 101 (see FIG. 3). The umbrella portion 108 b of the insertion joint 108 enters the recess 110. In this state, the UF membrane tube 15 and the return path 117 are in communication with each other via the insertion joint 108.
Further, as shown in FIG. 5, a recess 110 is also formed in the inlet / outlet path 110 a following the liquid inlet / outlet 118, and the umbrella portion 108 b of the insertion joint 108 enters the recess 110. In this state, the UF membrane tube 15 and the inlet / outlet path 110a are in communication with each other via the insertion joint 108.
With reference to FIG. 5, the sealing function of the tubular membrane separator 10 will be described. When the nut 102 is screwed onto the stud bolt 107 and tightened, the end portion of the pipe 106 is pressed against the rubber plate 104, and the umbrella portion 108 b of the insertion joint 108 is connected to the concave portion 110 of the metal plate 103 and the end cap 101. And press. As a result, the stock solution pumped into the UF membrane tube 15 does not leak outside. The insertion joint 108 constitutes a seal member (seal means) that prevents the liquid to be filtered from leaking out of the UF membrane tube 15. Further, the treated water filtered by the UF membrane tube 15 and coming out of the housing 14 through the small hole 120 is prevented from leaking out of the housing 14 by the O-ring 111. The O-ring 111 constitutes a seal member (seal means) for preventing leakage of treated water.
Next, the end cap 101 will be described. First, the end cap 101 on the inlet / outlet side where the liquid inlet / outlet 118 is formed will be described with reference to FIG. In the end cap 101, liquid inlets / outlets 118 are formed at two locations, and an inlet / outlet path 110 a that continues to each liquid inlet / outlet 118 is formed. As a result, the stock solution pumped from one liquid inlet / outlet port 118 is sent into the UF membrane tube 15 through one inlet / outlet route 110a, and the concentrated solution coming out from the other UF membrane tube 15 passes through the other inlet / outlet route 110a. It is configured to be discharged from the other liquid inlet / outlet port 118 via. In addition, U-shaped return paths 117 are formed in eight places on the end cap 101 on the input / discharge side. In the figure, reference numeral 110 denotes a recess into which the umbrella part 108b of the insertion joint 108 enters.
As shown in FIG. 6B, the return end cap 101 in which the liquid inlet / outlet 118 is not formed has nine return paths 117, but the liquid inlet / outlet 118 is not formed. The end cap 101 in which the return path 117 and the entry / exit path 110a are formed is manufactured by casting.
In such a configuration, the stock solution pumped from the liquid inlet / outlet 118 of the inlet / outlet end cap 101 (FIG. 6A) is sent into the UF membrane tube 15 from the insertion joint 108, and the return side end cap 101 diagram is shown. 6 (b) is reached. The stock solution then makes a U-turn through a return path 117 formed in the return-side end cap 101, is sent into the next UF membrane tube 15, and returns to the inlet / outlet-side end cap 101. Therefore, the undiluted solution makes a U-turn through a return path 117 formed in the end cap 101 on the inlet / outlet side, is sent into the next UF membrane tube 15 and moves again to the end cap 101 on the return side. The U-turn is repeated a plurality of times (17 times), and the concentrated liquid passing through all the UF membrane tubes 15 is discharged from the other liquid inlet / outlet 118. The end cap 101 constitutes a filtration target liquid return member (filtration target liquid return means) that returns the filtration target liquid received from a certain UF membrane tube 15 and sends it to another UF membrane tube 15.
Next, intermediate portions in the longitudinal direction of the plurality of pipes 106 are positioned by being held by the holding plate 116. This will be described with reference to FIGS. 7 (a) and 7 (b). Six holding plates 116 are provided on the pipe 106 located on the outer periphery of the 18 pipes 106. One holding plate 116 is provided across three pipes 106. A holding plate 116 is welded and fixed to the central pipe 106 of the three pipes 106. Of the three pipes, the pipes 106 at both ends are sandwiched between the end portions of the holding plates 116 located on both sides of the pipe 106.
In this way, the mid-longitudinal positions of all the pipes located on the outer periphery are held and positioned. These holding plates 116 are provided at a plurality of positions at predetermined intervals in the longitudinal direction of the pipe 106. As a result, there is an advantage that each pipe 106 can be held as straight as possible and the UF membrane tube 15 can be easily inserted into the pipe 106. The holding plate 116 constitutes a holding member (holding means) that holds a part of the pipe in the longitudinal direction.
The filtration function by the UF membrane tube 15 will be described with reference to FIG. The material of the UF membrane tube 15 is PVDF (polyvinylidene fluoride), which has an advantage of high heat resistance and a wide pH range of the liquid to be filtered. The diameter of the UF membrane tube 15 is about 15 mm. This about 15 mm is merely an example, and for example, a diameter in the range of 5 mm to 26 mm can be used. The UF membrane tube 15 is a porous membrane having a pore diameter of approximately 0.01 to 0.001 μm. By feeding a stock solution in which impurities 16 such as oil, polymer, turbidity and the like are mixed with low molecules, ions, water, etc., into the UF membrane tube 15 at an appropriate pressure, low molecules, ions, water, etc. are transferred to the UF membrane. It passes through the tube 15 and is discharged into the housing 14 from a large number of small holes 120 (see FIG. 3) formed in the pipe 106, and is taken out from the treated water outlet pipe 21 as a permeated liquid (treated water). On the other hand, impurities 16 such as oil, polymer, and turbidity in the stock solution are discharged from the inlet / outlet 118 in a concentrated state without passing through the UF membrane tube 15 and are reduced into the concentrated circulation layer 2.
By filtering the stock solution with the UF membrane tube 15, the impurities 16 adhere to the inner surface of the UF membrane tube 15. If the impurities 16 adhere, the filtration performance deteriorates, so it is necessary to remove the impurities. A sponge ball 17 is placed in the UF membrane tube 15 in order to remove impurities attached to the inner surface of the UF membrane tube 15. The sponge ball 17 has a size slightly larger than the diameter of the UF membrane tube 15 (about 15 mm). As a result, the sponge ball 17 inserted into the UF membrane tube 15 is in a state where its outer peripheral surface is in contact with the inner surface of the UF membrane tube 15 (see FIG. 8). The sponge ball 17 is moved in the UF membrane tube 15 by the stock solution fed into the UF membrane tube 15, and impurities adhering to the inner surface of the UF membrane tube 15 are wiped off and removed. The sponge ball 17 that has moved through the UF membrane tube 15 and has been pushed to the other end of the tubular membrane separator 10 is received and held by a receiving mechanism 18 (see FIGS. 2 and 9). The sponge ball 17 constitutes a wiping body that moves in the UF membrane tube 15 and wipes impurities adhering to the inner surface of the UF membrane tube 15. In addition, as a wiping body, it is not limited to the sponge ball | bowl 17, For example, a pig (bullet), a brush, etc. may be sufficient.
Next, the receiving mechanism 18 will be described based on FIG. The receiving mechanism 18 has a cylindrical portion 121, and an M-shaped strainer 19 in a plan view is provided in the cylindrical portion 121. Left and

right pipes

122 and 123 are connected to the tubular portion 121 by pipe joints 124, respectively. One pipe 122 is connected to the liquid inlet / outlet port 118 of the tubular membrane separator 10 (see FIG. 2), and the other pipe 123 is connected to the pipe 97 (see FIG. 1). Further, a screw groove is formed in the upper part of the cylindrical portion 121, and the detachable cap 20 is screwed into the screw groove.
By sending the stock solution into one liquid inlet / outlet 118, the sponge ball 17 moves through the UF membrane tube 15 of the tubular membrane separator 10 and is discharged from the other liquid inlet / outlet 118. The sponge ball 17 reaches the strainer 19 of the cylindrical portion 121 through the pipe 122 and is received by the strainer 19. The strainer 19 is formed with a stitch 19a so as to receive the sponge ball 17 but allow the concentrated liquid to pass therethrough. The concentrated liquid is reduced to the concentrated reduction tank 2 through the stitches 19a. The strainer 19 constitutes a receiving member that allows the liquid to be filtered to pass through but receives the sponge ball 17 (wiping body). Also,
The sponge ball 17 held by the receiving mechanism 18 at the other end of the tubular membrane separation device 10 is moved again in the UF membrane tube 15 by switching the flow of the stock solution (concentrated solution) in the reverse direction, so that the tubular type It reaches the receiving mechanism 18 on one end side of the membrane separation apparatus 10. In this way, by switching the flow of the stock solution (concentrated solution), the sponge ball 17 is reciprocated in the UF membrane tube 15 to remove the impurities 16.
By causing the sponge ball 17 to reciprocate a plurality of times in the UF membrane tube 15, impurities adhere to the sponge ball 17 itself, resulting in a disadvantage that the cleaning effect decreases. In order to eliminate the disadvantage, the dirty sponge ball 17 is configured to be replaceable. First, the valve 100 is provided in the pipe 122 (see FIG. 9) connected to the receiving mechanism 18 that receives the sponge ball 17, and the valve 100 is closed. This prevents the concentrate from flowing from the tubular membrane separator 10 to the receiving mechanism 18. In this state, the removable cap 20 of the receiving mechanism 18 is rotated and opened. The detachable cap 20 constitutes a take-out mechanism for taking out the received wipe and making it replaceable. Although the receiving mechanism 18 is provided at a lower position than the tubular membrane separation apparatus 10, since the valve 100 is closed, the concentrated liquid does not spout even when the detachable cap 20 is opened. Since the receiving mechanism 18 is provided at a higher position than the pipe 123 (see FIG. 9) on the side where the valve 100 is not provided, the concentrated liquid is supplied to the pipe 123 even when the detachable cap 20 is opened. Will not flow backwards. When the removable cap 20 is opened, the sponge ball 17 received by the strainer 19 can be taken out and replaced with a new one. After the replacement, the detachable cap 20 is screwed into the cylindrical portion 121 to cover it, and the valve 100 is opened. In addition, what is necessary is just to make it provide a valve also in the piping 123, when the receiving mechanism 18 is provided in the low position with respect to the piping 123 (refer FIG. 9). The valve 100 constitutes an ejection prevention mechanism (ejection prevention means) that prevents ejection of the liquid to be filtered when the detachable cap 20 is opened.
FIG. 1 shows a concentration step 1 in which a stock solution (concentrated solution) flows in the forward direction with respect to the tubular membrane separator 10, and the stock solution (concentrated solution) flows in the reverse direction with respect to the tubular membrane separator 10. Flowing concentration step 2 is shown in FIG. Referring to FIG. 10, in the concentration step 2, the three-way valve 36 is switched to the flow path C → A. The three-way valve 37 is switched to the B → C flow path. As a result, the concentrated liquid sent out by the circulation pump 33 is pressure-fed from the reverse direction to the tubular membrane separator 10 through the

pipes

83 and 85, the B → C of the three-way valve 37, the pipe 86, and the pipe 97. The concentrated liquid concentrated in the tubular membrane separation apparatus 10 is returned to the concentration circulation tank 2 through the pipe 84, the C → A of the three-way valve 36, the pipe 88, the C → B of the three-way valve 38, and the pipe 89.
After the concentration process 1 shown in FIG. 1 and the concentration process 2 shown in FIG. 10 are repeatedly performed a predetermined number of times, the filtration device moves to the water pushing process. In this water pushing step, the concentrated liquid remaining in the tubular membrane separation apparatus 10 and the piping is pushed out by the washing liquid and is reduced to the concentration circulation tank 2.
This water pushing step is performed for a short time of about 10 seconds and will be described with reference to FIG. The cleaning liquid stored in the membrane cleaning tank 8 is a case where fresh water such as tap water is used (fresh water cleaning mode in S78), and a case where treated water separated by the tubular membrane separator 10 is used (processing water cleaning mode in S79). ) When fresh water is used, the motor valve 10 is opened and fresh water is supplied to the membrane cleaning tank 8 through the pipe 92. The membrane cleaning tank 8 is provided with a level sensor LS2, which can detect four levels of a full water level HH2, an operation start level H2, an operation stop level L2, and a drought level LL2. On the other hand, when cleaning is performed using treated water instead of fresh water, it is necessary to supply the treated water separated by the tubular membrane separation apparatus 10 to the membrane cleaning tank 8. For this purpose, the motor valve MV5 and the closed motor valve MV6 are opened, and control is performed so that the treated water is supplied to the membrane cleaning tank 8 via the pipe 90 and the motor valve MV6. In both the fresh water cleaning mode and the treated water cleaning mode, fresh water or treated water is replenished until the membrane cleaning tank 8 reaches the operation start level H2.
A cleaning heater 49 is provided in the membrane cleaning tank 8, and the cleaning liquid is warmed to a temperature suitable for cleaning the UF membrane tube 15 (for example, 40 to 50 ° C.). Therefore, the membrane cleaning tank 8 is made of a stainless steel material having excellent heat resistance. In the case of cleaning with a normal temperature cleaning liquid without providing the cleaning heater 49, the film cleaning tank 8 may be made of a resin material. Further, by operating the cleaning agent injection pump 32, the alkaline cleaning agent is supplied to the membrane cleaning tank 8. This alkaline cleaning agent can further improve the cleaning efficiency. Further, depending on the type of concentrate (concentrated solution), the use of an acidic cleaning agent can improve the cleaning efficiency. In such a case, an acidic detergent is used instead of the alkaline detergent.
In the water pushing step, the circulation pump 33 is operated with the motor valve MV1 closed and the motor valve MV2 opened. When the concentration step performed immediately before the water pushing step is the concentration step 1 (see FIG. 1), the cleaning liquid is caused to flow in the reverse direction with respect to the tubular membrane separation apparatus 10. On the other hand, when the concentration process performed immediately before the water pushing process is the concentration process 2 (FIG. 10), the washing liquid is flowed forward with respect to the tubular membrane separation device 10 to perform the water pushing. FIG. 11 shows a water pushing process in which the cleaning liquid is flowed in the forward direction to push out the residual concentrated liquid. The three-way valve 36 is switched to the B → C flow path and the three-way valve 37 is switched to the C → A flow path. As a result, the cleaning liquid in the membrane cleaning tank 8 is supplied to the UF membrane tube 15 from one end of the tubular membrane separator 10 through the pipe 95, the circulation pump 33, B → C of the three-way valve 36, and the pipe 84. As a result, the sponge ball held in the receiving mechanism 18 on one end side of the tubular membrane separation apparatus 10 is pushed into the UF membrane tube 15 and moves in the UF membrane tube 15 to move into the UF membrane tube 15. Wipe off any adhering material that adheres to the surface. The extruded concentrated liquid is returned to the concentration reduction tank 2 through the pipe 86, C → A of the three-way valve 37, pipe 87, C → B of the three-way valve 38, and pipe 89. After the concentrate remaining in the tubular membrane separation apparatus 10 and the piping is pushed out into the concentration reduction tank 2 and the water pushing process is completed, the process proceeds to the washing process.
In the cleaning process, the cleaning liquid stored in the membrane cleaning tank 8 is supplied to the tubular membrane separator 10 to clean the UF membrane tube 15. The cleaning process includes a cleaning process 1 in which the cleaning liquid stored in the membrane cleaning tank 8 is washed in a forward direction with respect to the tubular membrane separation apparatus 10 and a cleaning liquid in a reverse direction with respect to the tubular membrane separation apparatus 10. There is a cleaning step 2 in which it is washed and washed. In the cleaning process 1, as shown in FIG. 12, the three-way valve 38 is switched to the C → A flow path, and the cleaning water is controlled to be returned to the membrane cleaning tank 8. As a result, after the cleaning liquid in the membrane cleaning tank 8 passes through the tubular membrane separator 10 via the pipe 95, B → C of the three-way valve 36, and the pipe 84, C → A of the three-way valve 37, pipe 87, three-way The valve 38 is returned to the membrane cleaning tank 8 through C → A and the scrap 96. After the cleaning process 1 is executed for a predetermined time, the process is switched to the cleaning process 2.
In the cleaning process 2, referring to FIG. 13, the three-way valve 36 is switched to the C → A flow path, and the three-way valve 37 is switched to the B → C flow path. As a result, the cleaning liquid in the membrane cleaning tank 8 passes through the pipe 95, the circulation pump 33, the

pipes

83 and 85, the flow path B → C of the three-way valve 37, and the pipe 86 from the opposite direction with respect to the tubular membrane separator 10. Supplied. The cleaning liquid that has passed through the UF membrane tube 15 in the tubular membrane separator 10 is reduced to the membrane cleaning tank 8 via the pipe 84, the C → A of the three-way valve 36, the pipe 88, the C → A of the three-way valve 38, and the pipe 96. Is done. After repeatedly performing this washing process 1 (see FIG. 12) and washing process 2 (see FIG. 13), the process moves to the concentration process again.
By repeatedly performing the concentration step, the water pushing step, and the washing step as described above, the concentrated solution is gradually concentrated by filtration by the tubular membrane separation device 10, and the amount of the concentrated solution stored in the concentration circulation tank 2 is reduced. Decrease. The concentrated liquid in the concentration circulation tank 2 has decreased to a predetermined amount (L1 level).
Is detected by the level sensor LS1, the pH adjustment water pump 30 is operated to supply the stock solution in the pH adjustment tank 5 to the concentration circulation tank 2. If the level sensor 1 detects that the concentrate in the concentration circulation tank 2 has reached the operation start level H1 by adding the stock solution, the pH adjustment water pump 30 is stopped and the addition of the stock solution is stopped. . In this state, the above-described concentration step, water pushing step, and washing step are repeatedly executed. Then, when the concentrated liquid in the concentration circulation tank 2 is reduced to a predetermined amount (L1 level) again, the stock solution in the pH adjustment tank 5 is again used.
Add to the concentration circulation tank 2 and supply. When the number of times of addition of the stock solution (concentration number) reaches a predetermined number, the motor valve MV7 is opened and the concentrate in the concentration circulation tank 2 is discharged to the concentrate relay tank 3 through the pipe 93. When the level sensor LS1 detects that the concentrate in the concentration circulation tank 2 has reached the discharge completion level LL, the motor valve MV7 is closed and the discharge of the concentrate is completed. On the other hand, in the concentrate relay tank 3, if the level sensor LS 3 detects that the amount of concentrate stored has increased, the concentrate feed pump 31 is operated to pass the concentrate in the concentrate relay tank 3 through the pipe 94. It sends out to the concentration storage tank 4. The concentration storage tank 4 is provided with a level sensor LS5. If the level sensor LS5 detects that the concentrated liquid in the concentrated storage tank 4 is full, the concentrated liquid in the concentrated storage tank 4 is taken out. Work is done.
A process from the start of the first concentration process to the discharge of the concentrated liquid in the concentration circulation tank 2 where the concentration is completed is referred to as one batch. By performing this one batch of processes a predetermined number of times, the filtering operation by the filtering device is completed and the filtering device is stopped.
The turbidity of the treated water taken out from the treated water take-out pipe 21 is detected by the photoelectric sensor 25. If the turbidity is equal to or higher than a predetermined value, it is determined that the filtration error has occurred, and an error notification process is performed.
Further, the temperature of the cleaning liquid pushed out by the circulation pump 33 is detected by the temperature sensor 26. Based on the detected value, the cleaning heater 49 is controlled so that the temperature of the cleaning liquid is maintained at a predetermined temperature (40 to 50 ° C.).
Next, a control circuit for the pH adjustment control panel 22 and the apparatus control panel 23 will be described with reference to FIG. The pH adjustment control panel 22 is provided with a pH adjustment control panel microcomputer 60. The pH adjustment control panel microcomputer 60 includes a CPU (Central Processing Unit) 66 as a control center, a ROM (Read Only Memory) 67 in which control programs and data are stored, and a RAM that functions as a work area for the CPU 66. (Random Access Memory) 68 and the like are provided.
On the other hand, the device control board 23 is provided with a device control board microcomputer 63. The device control panel microcomputer 63 includes a CPU 69 serving as a control center, a ROM 70 storing control programs and data, a RAM 71 functioning as a work area for the CPU 69, an EEPROM (Electrically Erasable Programmable Read-Only Memory) 72, and the like. Is provided.
The pH adjustment control panel microcomputer 60 receives detection signals from the pH sensor 24 and the level sensors LS4 to LS6. In addition, an operation signal from the input operation unit 61 such as a keyboard, a mouse, or a touch panel is input to the pH adjustment control panel microcomputer 60.
The pH adjustment control panel microcomputer 60 outputs pump control signals for the raw solution water pump 27, the alkali injection pump 28, the acid injection pump 29, and the pH adjustment water pump 30. A control signal for driving the agitation motor 30 is output. In addition, monitor display signals such as various error displays and the operation state of the filtration device are output to the display unit 62 to the operator.
The device control panel microcomputer 63 receives the sensor signals of the level sensors LS1 to LS3. Each flow rate detection signal of the flow meters FL1 to FL3 is input. Further, sensor signals of the photoelectric sensor 25, the temperature sensor 26, and the pressure sensors PS1 and PS2 are input. In addition, an operation signal is input from the input operation unit 64 such as a keyboard, a mouse, or a touch panel.
A heater control signal is output from the apparatus control panel microcomputer 63 to the cleaning heater 49. Further, each pump control signal is output to the concentrated liquid feed pump 31, the cleaning agent injection pump 32, and the circulation pump 33. Further, each valve control signal is output to the motor valves MV1 to MV10. Further, a display control signal for performing an error display and a monitor display of the operation state of the filtration device to the operator is output to the display unit 65.
The pH adjustment control panel microcomputer 60 and the apparatus control panel microcomputer 63 can transmit and receive signals to and from each other. The apparatus control panel microcomputer 63 outputs a stock solution supply signal for supplying the stock solution to the concentrated storage tank 2 to the pH adjustment control panel microcomputer 60. Receiving this, the pH adjustment control panel microcomputer 60 starts the control to supply the stock solution to the concentration and reduction tank 2 as described above, and sends a raw water supply signal to the device control panel microcomputer 63 that the supply of the stock solution has started. Send back. Further, the pH adjustment control panel microcomputer 60 transmits a collective abnormality signal to the apparatus control panel microcomputer 63 when an abnormality occurs in the control territory.
Next, based on FIG. 15 and FIG. 16, the control operation by the pH adjustment control panel microcomputer 60 will be described. The programs of the flowcharts shown in FIGS. 15 and 16 are stored in the ROM 67. First, the main routine of the pH adjustment control process will be described with reference to FIG. Each tank abnormality check process is performed in step (hereinafter simply referred to as “S”) 1. This is to check abnormalities in the storage state of the raw water tank 1 and the concentrated storage tank 4. Next, each tank liquid movement process is performed by S2. Next, pH adjustment processing is performed by S3. This pH adjustment treatment is a treatment for adjusting the pH by neutralizing the stock solution in the pH adjustment tank 5 with an alkali or an acid.
With reference to FIG. 15B, a flowchart of the subroutine program of each tank abnormality check process described above will be described. By S7, it is determined whether or not the value of the level sensor LS4 has abnormally decreased. When the amount of raw water stored in the raw water tank 1 is abnormally reduced and the water is in a drought state, a determination of YES is made in S7, the control advances to S8, and the process of turning on the raw water tank drought error flag is turned on Is made. On the other hand, when it is determined that the value of the level sensor LS4 is not abnormally lowered, the control proceeds to S9, and control is performed to turn off the raw water tank drought error flag.
Next, in S10, it is determined whether or not the value of the level sensor LS4 has risen abnormally. If it is determined that the temperature has risen abnormally, the control proceeds to S11, where control is performed to turn on the raw water tank full error flag. On the other hand, when it is determined that the level sensor LS4 has not risen abnormally, the process proceeds to S12, and processing for turning off the raw water tank full error flag is performed.
In S13, it is determined whether or not the value of the level sensor LS6 has abnormally decreased. If the stock solution in the pH adjustment tank 5 has dropped to an abnormal level, a determination of YES is made in S13, control proceeds to S14, and a process of turning on the pH adjustment tank drought error flag is performed. On the other hand, when the value of the level sensor 6 is not abnormally lowered, the control proceeds to S15, and the pH adjustment tank drought error flag is turned OFF. Next, in S16, it is determined whether or not the value of the level sensor LS6 has risen abnormally. If the stock solution stored in the pH adjustment tank 5 has risen to an abnormal level, a determination of YES is made in S16, the control proceeds to S17, and the pH adjustment tank full error flag is turned ON. On the other hand, if it is determined that the value of the level sensor 6 has not risen abnormally, the pH adjustment tank full error flag is turned OFF in S18.
Next, the control proceeds to S19, where it is determined whether or not the level sensor 5 has abnormally increased. When the concentrated liquid stored in the concentrated storage tank 4 is full and the level sensor LS5 detects that it is full, a determination of YES is made in S19, the control proceeds to S20, and the concentrated storage tank full water error flag is turned ON. . On the other hand, when the value of the level sensor 5 has not increased abnormally, the control proceeds to S21, and control for turning off the concentrated storage tank full error flag is performed.
Next, control proceeds to S22, where it is determined whether any error flag is ON. If it is not ON, this subroutine program returns and control proceeds to S2.
On the other hand, if any one of the error flags is ON, the control advances to S23, and a collective abnormality signal is transmitted to the apparatus control panel microcomputer 63, and the response to the ON error flag is made in S24. Control is performed to display the abnormality display by the display unit 62, and the process returns.
Next, a flowchart of a subroutine program for each inter-tank liquid transfer process shown in S2 will be described with reference to FIG. It is determined by S30 whether the raw water supply signal has been received. If the raw water supply signal is transmitted from the apparatus control panel microcomputer 63, a determination of YES is made in S30 and the control advances to S31, the pH adjustment water pump 30 is operated and the raw water in the pH adjustment tank 5 is concentrated and circulated. 2 is supplied. Next, in S32, a raw water supply signal indicating that the raw water supply is started is returned to the apparatus control panel microcomputer 63. Next, in S33, it is determined whether or not the value of the level sensor LS6 has decreased to the supply level. When the stock solution in the pH adjustment tank 5 is reduced to a level that requires replenishment, a determination of YES is made in S33 and the control advances to S34, and the stock solution water pump 27 is operated to remove the stock solution in the raw water tank 1. The pH adjustment tank 5 is replenished.
The flowchart of the subroutine program for the pH adjustment process shown in S3 will be described with reference to FIG. It is determined through S40 whether or not a pH adjustment value has been input. When the operator operates the input operation unit 61 to input a pH adjustment value in the pH adjustment tank 5, a process of updating the pH adjustment value to the input new pH adjustment value is performed in S41. Next, in S42, a process of comparing the current PH sensor value PS with the pH adjustment value PT is performed, and in S43, it is determined whether PS exceeds PT. If PS exceeds PT, control proceeds to S44, where the alkali injection pump 28 is operated to inject alkali into the pH adjusting tank 5. On the other hand, if PS does not exceed PT, the control proceeds to S45, where the acid injection pump 29 is operated to inject the acid into the pH adjustment tank 5. By this pH adjustment process, the stock solution in the pH adjustment tank 5 is controlled to have the pH adjustment value input by the operator.
Next, a flowchart of a control program for the apparatus control panel microcomputer 63 will be described with reference to FIGS. These flowcharts are stored in the ROM 70. First, the main routine of the apparatus control process will be described with reference to FIG. Various operation value setting input processes necessary for the operation of the filtration device are performed in S50, and various abnormal value setting input processes that are criteria for determining whether or not an abnormality has occurred are performed in S51, and an abnormality has occurred in S52. An abnormality check process is performed. In S53, each motor valve is controlled to switch the flow rate, and in S54, a monitor display process is performed.
Next, a flowchart of a subroutine program for the operation set value input process shown in S50 will be described with reference to FIGS. 17 (b) and 18 (a). This operation set value input process is a process for the operator to input and set various values necessary for operating the filtration device by the input operation unit 64. In S60, it is determined whether or not the number of times of concentration has been input. If there is not, it is determined in S62 whether or not the number of batches has been input. In S66, it is determined whether or not there has been an input of the discharge relay tank level. If not, it is determined whether or not an automatic cleaning time has been input in S67. If not, it is determined whether or not a water pushing time has been input in S70. If not, it is determined whether or not the reverse operation time has been input in S72. If not, it is determined in S74 whether or not the washing mode switching operation has been performed in S76. If not, it is determined whether or not a chemical cleaning setting has been input in S80. If not, the film cleaning is performed in S82. Is determined whether or not an input of the tank Shimizu replenishing time, not be determined whether or not the input of the heat exchanger operating temperature by S84 when, in the absence this subroutine program returns.
As described above, if the operator operates the input operation unit 64 to input the number of times the stock solution is added to the concentration circulation tank 2 (concentration number), a YES determination is made in S60, and the control advances to S61 for storage. The process of updating the number of times of concentration performed to the newly input number of times of concentration NK is performed. When the number of batches for how many steps of one batch are executed is input by the input operation unit 64, a determination of YES is made in S62, the control advances to S63, and the stored batch number is newly input. The process of updating to the batch count VC is performed.
When an operation for inputting and setting the level of the concentrate stored in the concentration circulation tank 2 is performed by the input operation unit 64, a determination of YES is made in S64, the control proceeds to S65, and the stored concentration circulation Processing to update the tank level to the newly input levels HH1, H1, L1, and LL1 is performed. HH1 is a full water abnormality level, H1 is an operation start level, L1 is an operation stop level, and LL1 is a drainage completion level.
If the operator operates the input operation unit 64 to input the level of the waste liquid relay tank (concentrated liquid relay tank) 3, a determination of YES is made in S66 and the control advances to S68, and the stored waste liquid relay tank level. Is updated to the newly input levels HH2, H2, L2, and LL2. This HH2 is a level of abnormally high water, H2 is a level to close the motor valve MV7 and prevent the concentrate from entering the concentration relay tank 3, and L2 starts the operation of the concentrate feed pump 31. LL2 is a level at which the concentrated water pump 31 is stopped.
When the automatic cleaning time is input by the input operation unit 64, a determination of YES is made in S67 and the control advances to S69, and the stored automatic cleaning time is updated to the newly input automatic cleaning time. Is made. The automatic cleaning time is a time for repeatedly executing the above-described cleaning process 1 (see FIG. 12) and cleaning process 2 (see FIG. 13). If the water pushing time is input by the input operation unit 64, a determination of YES is made in S70 and the control advances to S71, and the stored water pushing time is updated to the newly inputted water pushing time. .
When the reverse operation time is input by the input operation unit 64, a determination of YES is made in S72, the control proceeds to S73, and the stored reverse operation time is updated to the newly input reverse operation time. Made. The reverse operation time is the time from the start of the concentration step 1 to the end of the concentration step 1, the time from the start of the concentration step 2 to the end of the concentration step 2, and the start of the washing step 1 It is the time from the start to the end of the cleaning step 1 and the time from the start of the cleaning step 2 to the end of the cleaning step 2. When the waste liquid discharge time for discharging the concentrate stored in the concentration circulation tank 2 to the concentration relay tank 5 is input by the input operation unit 64, a determination of YES is made in S74, and the control proceeds to S75. Processing is performed to update the stored waste liquid discharge time to the newly input waste liquid discharge time.
If the input operation unit 64 is operated to determine which cleaning mode is to be used, that is, cleaning with clean water or cleaning with the processing liquid, YES is determined in S76 and the control advances to S77. Then, it is determined whether or not the current mode is the treated water cleaning mode. If the current mode is the treated water cleaning mode, the control proceeds to S78, and processing for switching to the fresh water cleaning mode is performed. On the other hand, if the current mode is not the treated water cleaning mode, that is, if it is the fresh water cleaning mode, the control proceeds to S79 and a process of switching to the treated water cleaning mode is performed.
If there is a chemical cleaning setting input operation through the input operation unit 64, a determination of YES is made in S80 and the control advances to S81, where the chemical injection time corresponding to the set batch number is updated to the input value. The When the membrane cleaning tank fresh water replenishment time for supplying fresh water to the membrane cleaning tank 8 is input by the input operation unit 64, a determination of YES is made in S82 and the process proceeds to S83 to store the stored membrane cleaning tank fresh water supply. Processing is performed to update the time to the newly input membrane cleaning tank fresh water replenishment time.
If the heat exchange operation temperature, which is the temperature of the cleaning liquid in the membrane cleaning tank 8, is input from the input operation unit 64, a determination of YES is made in S84 and the control advances to S85, and the stored heat exchange operation temperature is set. Processing to update the newly input heat exchange operation temperature is performed. Then, this subroutine program returns.
Next, a subroutine program for the abnormal value setting input process shown in S51 described above will be described with reference to FIG. In S90, it is determined whether or not there has been a processing flow rate lowering input. If not, it is determined in S92 whether or not a circulating flow rate has been input. If not, it is determined in S94 whether or not a pressure has been input. In S96, it is determined whether or not a concentration / discharge error has been input. If not, this subroutine program returns.
When the input of the treatment flow rate reduction for determining whether or not the flow rate of the treated water filtered by the tubular membrane separation apparatus 10 is abnormal is performed by the input operation unit 64, the control proceeds to S91 and stored. The processing flow reduction value that is being updated is updated to the newly inputted processing flow reduction value SR. When the circulation flow rate which is the circulation amount of the concentrate circulated by the circulation pump 33 is input by the input operation unit 64, the control proceeds to S93, and the stored circulation flow rate is changed to the newly input circulation flow rate. Processing to update the upper limit value JRU and the lower limit value JRL is performed. When the pressure of the concentrated liquid pumped to the tubular membrane separation apparatus 10 in the concentration step is input by the input operation unit 64, the control proceeds to S95, and the stored pressure is the upper limit of the newly input pressure. Processing for updating to PU and lower limit PL is performed. If the concentration discharge error time for determining an abnormality is input by the input operation unit 64 when the time for discharging the concentrate stored in the concentration circulation tank 2 exceeds a predetermined value, the control proceeds to S97. The stored concentration discharge error time is updated to the newly input concentration discharge error time NHT. Then, this subroutine program returns.
Next, a subroutine program for the abnormality check process shown in S52 will be described with reference to FIGS. First, in S105, the current value of each sensor is read and stored. Specifically, the value of the level sensor LS1 is stored as LS1, the value of the level sensor LS2 is stored as LS2, the value of the level sensor LS3 is stored as LS3, the value of the flow meter FL1 is stored as F1, and the flow rate The value of the meter FL2 is stored as F2, the value of the flow meter FL3 is stored as F3, the value of the photoelectric sensor 25 is stored as PS, the value of the temperature sensor 26 is stored as T, and the value of the pressure sensor PS1 is PS1 And the value of the pressure sensor PS2 is stored as PS2.
Next, it is determined by S106 whether the process currently performed by the filtration apparatus is a concentration process. If it is not the concentration step, this subroutine program returns. However, if it is the concentration step, the control advances to S107, and it is determined whether or not the value F2 of the flow meter FL2 is less than the processing flow rate drop value SR. If the determination result is YES, the processing flow rate decrease error flag is turned ON in S108. On the other hand, if the determination result in S107 is NO, a process for turning off the processing flow rate decrease error flag is performed in S109. Next, the control proceeds to S110, where it is determined whether or not the value F1 of the flow meter FL1 exceeds the upper limit JRU of the circulation flow rate. If exceeded, the operation flow rate upper limit error flag is turned ON in S111. Thereafter, the control proceeds to S116. On the other hand, if it is determined in S110 that it has not exceeded, control proceeds to S112, and after the operating flow rate upper limit error flag is turned OFF, control proceeds to S113. In S113, it is determined whether or not the value F1 of the flow meter FL1 is less than the lower limit JRL of the circulation flow rate. If it is less than S114, the operation flow rate lower limit error flag is turned on in S114, and if it is determined NO in S113, the operation flow rate lower limit error flag is turned off in S115.
In S116, it is determined whether or not the value PS1 of the pressure sensor PS1 exceeds the pressure upper limit PU. If it has exceeded, the process proceeds to S112 after the operation pressure upper limit error flag is turned on in S117. On the other hand, if PS1 does not exceed PU, a process for turning off the operating pressure upper limit error flag is performed in S118. Next, in S119, it is determined whether or not the value PS1 of the pressure sensor PS1 is below the lower limit PL of the pressure. If it is lower, the operating pressure lower limit error flag is turned ON in S120. On the other hand, if not below, the operation pressure lower limit error flag is turned OFF in S121.
Next, the control proceeds to S122, where it is determined whether or not the waste liquid drain flag has been switched from OFF to ON. This waste liquid discharge flag is a flag for discharging the concentrated liquid into the concentrated circulation tank 2 to the concentrated liquid relay tank 3, and is turned ON by S127 described later and turned OFF by S129 described later. Further, if the number of times of concentration (the number of times of addition) reaches the input number of times of concentration NK (see S61), when the number of batches has not reached VC, the waste liquid discharge flag is turned ON in S167 described later. In S122, a determination of YES is made only at the moment when the waste liquid discharge flag is switched from OFF to ON. If YES is determined in S122, the control proceeds to S123, and a process of setting the waste liquid discharge timer HHP is performed. Further, after the control for opening the motor valve MV7 is performed in S124, the control proceeds to S131.
If NO is determined in S122, the control proceeds to S125, and it is determined whether or not the waste liquid discharge flag is ON. If it is not ON, the process proceeds to S130, the waste liquid discharge timer HHP is cleared, and control is performed to close the motor valve MV7. Then, the process proceeds to S131.
On the other hand, if it is determined in S125 that the waste liquid discharge flag is ON, the control proceeds to S126, and it is determined whether or not the waste liquid discharge timer HHT exceeds the concentration discharge error time NHT.
Is done. If exceeded, control is performed to turn on the waste liquid discharge flag in S127, and then control proceeds to S128. When the waste liquid discharge timer HHT does not exceed the concentration discharge error time NHT, the control proceeds to S128, and it is determined whether or not the level sensor LS1 is at the discharge completion level (LL1). When the discharge completion level is not reached, the control proceeds to S131. However, when the discharge completion level is reached, the control proceeds to S12, and after the control for switching the waste liquid discharge flag to OFF is performed, the control proceeds to S131. move on.
In S131, it is determined whether or not the value of the level sensor LS3 exceeds the level HH2 of the full water abnormality. If it has exceeded, control proceeds to S132, the discharge P error flag is turned ON, and the process of turning ON the concentrate relay tank full error flag is performed, followed by S134.
On the other hand, if NO is determined in S131, the control proceeds to S133, in which the waste liquid P error flag is turned OFF and the concentrate relay tank full error flag is turned OFF. Next, the control proceeds to S134.
In S134, it is determined whether or not the value of the level sensor LS1 exceeds the full water abnormality level HH1. If it has exceeded, control proceeds to S135, and control is performed to turn on the concentration / circulation tank full water error flag. On the other hand, when the value of LS1 does not exceed HH1, the control proceeds to S136, and processing for turning off the concentration circulation tank full error flag is performed. Next, in S137, it is determined whether or not the value of the level sensor LS2 exceeds the full water level ML of the membrane cleaning tank 8. If exceeded, control proceeds to S138, and after processing for turning on the membrane cleaning tank full water error flag is performed, control proceeds to S143. On the other hand, when LS2 does not exceed the full water level ML, the control proceeds to S139, and control for turning off the membrane cleaning tank full error flag is performed.
Next, the control advances to S140, and it is determined whether or not the value of the level sensor LS2 is below the drought level KL of the membrane cleaning tank 8. If it falls below, the control advances to S141, and a process for turning on the membrane cleaning tank drought error flag is performed. On the other hand, when the value of LS2 is not lower than the drought level KL, the control proceeds to S142, and control for turning off the membrane cleaning tank drought error flag is performed. Next, in S143, it is determined whether or not any error flag is ON. If any error flag is ON, emergency stop control is performed in S144.
Next, a subroutine program for the flow path switching process shown in S53 will be described with reference to FIG. By S150, the stock solution replenishment process for supplying the stock solution to the concentration circulation tank is performed. Next, in S151, a concentration process is performed in which the stock solution stored in the concentration circulation tank 2 is filtered and concentrated. Next, a water pushing process is performed by S152. Next, a cleaning process for cleaning the UF membrane tube 15 of the tubular membrane separation apparatus 10 using the cleaning liquid stored in the membrane cleaning tank 8 is performed in S153. Next, a concentrated liquid discharge process for discharging the concentrated liquid stored in the concentrated circulation tank 2 is performed through S154.
Next, a flowchart of a subroutine program of the stock solution replenishment process shown in S150 will be described with reference to FIG. By S160, it is determined whether or not the operation flag indicating that the filtration apparatus is operating is ON. The operation flag becomes NO when the operator turns on the operation start switch with the power of the filtration apparatus turned on, and is turned off in S168. If the operation flag is not ON, this subroutine program returns. If the operation flag is ON, the control proceeds to S161, and it is determined whether or not the value of the level sensor LS1 is equal to or lower than the discharge completion level LL1 of the concentration circulation tank 2. If it is not less than LL1, the control advances to S162, and it is determined whether or not the value of the level sensor LS1 is at the operation stop level L1. If it is not L1, this subroutine program returns.
When this stock solution replenishment process is executed in a state where the concentrate in the concentration circulation tank 2 is discharged, the value of the level sensor LS1 is equal to or lower than the discharge completion level LL1, and therefore the control proceeds to S169. A process of transmitting a stock solution supply signal for replenishing the stock solution by the initial replenishment amount to the pH adjustment control panel 22 is performed. This “initial replenishment amount” is a replenishment amount at which the stock solution reaches the operation start level H1 in the concentration circulation tank 2 in a state where the concentrate is discharged.
If the concentrate replenishment process is executed in a state where the concentration process has progressed and the concentrated liquid in the concentration circulation tank 2 has been filtered to become thicker and the volume has been reduced to the operation stop level L1, a determination of YES is made in S162. The control advances to S163, and it is determined whether or not the concentration counter is NK. This concentration counter is a counter that counts the number of times of addition to add the stock solution into the concentration circulation tank 2. NK is the number of times of concentration (number of times of addition) preset by S61. If the current number of times of concentration (the number of times of addition) has not yet reached the preset number of times of enrichment NK, the control proceeds to S164, and the stock solution supply signal for replenishing the stock solution to the operation start level H1 in the concentration circulation tank 2 Is transmitted to the pH adjustment control panel 22. Thereby, it determines with YES by S30, a stock solution is supplied to the concentration circulation tank 2 by S31, and the stock solution is added.
While the stock solution is being added, the device control microcomputer 63 stops and controls the circulation pump 33. Thereby, supply of the stock solution (concentrated solution) to the tubular membrane separation apparatus 10 is stopped. When the circulation pump 33 is stopped, a back flow of the stock solution (concentrated liquid) may occur in the circulation pump 33. When a backflow occurs, the UF membrane tube 15 in the tubular membrane separation apparatus 10 located above the plurality of tubular membrane separation apparatuses 10 becomes negative pressure, and the treated water flows back to the UF membrane tube 15. There is a disadvantage that the flattened shape is crushed. In order to prevent this inconvenience, the apparatus control microcomputer 63 opens the motor valve MV8 while the circulation pump 33 is stopped. Then, the UF membrane tube 15 in the tubular membrane separator 10 is in communication with the upper part of the concentration circulation tank 2. The space between the liquid level of the stock solution (concentrated liquid) stored in the concentration circulation tank 2 and the ceiling portion of the concentration circulation tank 2 communicates with the outside air. As a result, the UF membrane tube 15 in the tubular membrane separator 10 is in communication with the outside air, and the UF membrane tube 15 is prevented from becoming negative pressure, and the UF membrane tube 15 is crushed and flattened. Is prevented. Next, the control advances to S165, and a process of updating the concentration number counter by 1 is performed.
On the other hand, if it is determined in S163 that the current concentration counter has reached NK, control proceeds to S166, where it is determined whether the batch counter has reached VC. This VC is the number of batches input and set in advance in S63. If the number of batches at the current time has not yet reached VC, the control advances to S167, and a process of turning on a waste liquid discharge flag for discharging the concentrated liquid in the concentration circulation tank 2 is performed. On the other hand, if the number of batches at the current time reaches VC, the control advances to S168, and processing for switching the operation flag to OFF is performed.
As described above, in this embodiment, the stock solution is added each time the stock solution in the concentrate circulation tank 2 decreases and reaches a predetermined amount (L1 level) as the stock solution is concentrated. Instead of this, the stock solution may always be supplied to the concentration circulation tank 2 little by little as the concentration proceeds. In the case of the former pouring method, the temperature of the concentrate gradually rises as the concentration progresses, so that there is an advantage that the filtration performance is improved. Inconvenience arises. In the case of the latter constant supply method, the temperature of the concentrate does not rise so much and the improvement of the filtration performance cannot be expected so much, but there is an advantage that the operation stop at the time of addition does not occur.
Next, the flowchart of the subroutine program for the concentration process shown in S151 will be described with reference to FIG. A process switching process is performed in S175. Next, in S176, a concentration process flow rate control process for switching the flow rate of the three-way valve or the like is performed.
Next, a flowchart of the subroutine program for the process switching process shown in S175 will be described with reference to FIG. In S180, it is determined whether or not the cleaning process flag is ON. This cleaning process flag is a flag indicating that the cleaning process is being executed. If the cleaning process flag is ON, this subroutine program returns. If the cleaning process flag is not ON, the control advances to S181 to determine whether or not the water pushing process flag is ON. The water pushing process flag is a flag indicating that the water pushing process is being executed. If the water pushing process flag is not ON, the control advances to S184, and it is determined whether or not the concentration process flag is ON. The concentration process flag is a flag indicating that the concentration process is being executed. When the concentration process flag is not ON, the control proceeds to S185, and it is determined whether or not the value of the level sensor LS1 has reached the operation start level H1 in the concentration circulation tank 2. If not reached, this subroutine program returns. However, if the stock solution is replenished and the value of the level sensor LS1 reaches the operation start level H1, the control advances to S186, and the process of turning on the concentration process flag is performed. Done.
If the concentration process flag is turned ON, a determination of YES is made in S184, the control proceeds to S182, and it is determined whether or not one unit concentration time has elapsed. This 1-unit concentration time refers to the water pushing step (see FIG. 11) after the concentration step 1 (see FIG. 1) and the concentration step 2 (see FIG. 10) are repeated a plurality of times after the concentration step is started. This is the time until the concentration process is completed for transfer. This 1 unit concentration time is set in advance so that it can be adjusted to a desired time. If one unit concentration time has not yet elapsed since the concentration step was started, the control proceeds to S183, where it is determined whether the reverse operation time has elapsed. The reverse operation time is the time from the start of the concentration step 1 until the end of the concentration step 1 to start the concentration step 2, and the transition from the start of the concentration step 2 to the concentration step 1 The time until the concentration step 2 is completed. This reverse operation time is the reverse operation time input and set in advance in S73.
If the reverse operation time has not yet elapsed, this subroutine program returns. On the other hand, when the reverse operation time has elapsed, a determination of YES is made in S183, the control proceeds to S187, and it is determined whether or not the reverse direction flag is ON. The reverse direction flag is a flag indicating that the concentrated liquid or the cleaning liquid is flowing in the reverse direction with respect to the tubular membrane separator 10. If the current direction is flowing in the reverse direction, the reverse direction flag is ON. Therefore, a determination of YES is made in S187, the control proceeds to S188, and the reverse direction flag is turned OFF. On the other hand, when the reverse direction flag is OFF, the control advances to S192, and processing for turning the reverse direction flag ON is performed. This subroutine program returns soon.
When one unit concentration time has elapsed, a determination of YES is made in S182, the control proceeds to S189, and it is determined whether or not the concentration number counter has reached NK. When the number of additions of the stock solution to the concentration circulation tank 2 has reached the preset concentration number NK, the control advances to S190, and processing for turning off the concentration process flag is performed. On the other hand, if the concentration counter has not reached NK, the control proceeds to S191, where the concentration process flag is turned OFF and the water pushing process flag is turned ON.
As a result, YES is determined in S181, and the subroutine program for process switching processing returns. The subroutine program for this process switching process also returns when the cleaning process flag is ON.
Next, a subroutine program for the concentration process flow path control process shown in S176 will be described with reference to FIG. In S195, it is determined whether or not the concentration process flag is ON. If it is not ON, this subroutine program returns. If it is ON, control proceeds to S196, where it is determined whether the reverse flag is OFF. If the reverse flag is OFF, the control proceeds to S197, the motor valve MV1 is opened, the motor valve MV3 is switched to the B → C flow path, and the motor valve MV4 is switched to the C → A flow path, After the motor valve MV9 is switched to the flow path C → B and the motor valve MV2 is closed, this subroutine program returns. By the control of S197, the concentrated solution (stock solution) flows in the forward direction as shown in FIG. 1, and the concentration step 1 is executed.
On the other hand, if the reverse direction flag is ON, NO is determined in S196 and the control advances to S198, the motor valve MV1 is opened, the motor valve MV3 is switched to the flow path C → A, and the motor valve MV4 is opened. Is switched to the flow path B → C, the motor valve MV9 is switched to the flow path C → B, and the motor valve MV2 is closed. As a result, the concentrated solution (stock solution) flows in the reverse direction as shown in FIG. 10, and the concentration step 2 is executed.
Next, the flowchart of the subroutine program for the water pushing process shown in S152 will be described with reference to FIG. A process switching process is performed in S205, and a water pushing process flow path control process is performed in S206.
Next, a flowchart of the subroutine program for the process switching process shown in S205 will be described with reference to FIG. By S210, it is determined whether or not the water pushing process flag is ON. If the water pushing process flag is not ON, this subroutine program returns. If it is ON, the control advances to S211 to determine whether or not the water pushing time has elapsed. This water pushing time is the water pushing time set in advance by S71. When the water pushing process proceeds and the water pushing process is executed for this water pushing time, a determination of YES is made in S211 and the control advances to S212, turning off the water pushing process flag and turning on the washing process flag. Switching control is performed.
Next, a flowchart of a subroutine program for the water pushing process flow path control process shown in S206 will be described with reference to FIG. In S215, it is determined whether or not the water pushing process flag is ON. When the water pushing process flag is OFF, this subroutine program returns. However, when the water pushing process flag is ON, the control advances to S216, and it is determined whether or not the reverse direction flag is ON. The If the concentration step 2 in which the stock solution (concentrate) is flowed in the reverse direction was executed in the concentration step that was executed immediately before the water pushing step, a YES determination is made in S216, and the control advances to S217, and the motor Control for closing valve MV1, opening motor valve MV2, switching motor valve MV3 to the flow path B → C, switching motor valve MV4 to the flow path C → A, and switching motor valve MV9 to the flow path C → B Is done. As a result, as shown in FIG. 11, the water is pushed by flowing the washing water in the forward direction.
On the other hand, if the concentration process performed immediately before the water pushing process is the concentration process 1 in which the stock solution (concentrated liquid) is flowed in the forward direction, the reverse flag is OFF, so that NO is determined in S216. The control proceeds to S218, the motor valve MV1 is closed, the motor valve MV2 is opened, the motor valve MV3 is switched to the flow path C → A, the motor valve MV4 is switched to the flow path B → C, and the motor valve Control for switching the MV 9 to the flow path C → B is performed.
As a result, when the concentration process performed immediately before the water pushing process is the concentration process 2 in which the concentrated liquid is flowed in the reverse direction, the cleaning liquid is flowed in the forward direction in the water pushing process to be performed next. The flow path control is performed (S217), and the sponge ball held in the receiving mechanism 18 on one end side of the tubular membrane separation apparatus 10 is pushed out into the UF membrane tube 15, and the sponge ball 17 is passed through the inside of the UF membrane tube 15. With this, the water is pushed while being washed. When the concentration process performed immediately before the water pushing process is the concentration process 1 in which the stock solution (concentrated liquid) flows in the forward direction, the reverse direction flag is OFF, so the cleaning liquid is tubular. A water push is applied to the membrane separation device 10 in the opposite direction (S218), and the sponge ball 17 held by the receiving mechanism 18 on the other single side of the tubular membrane separation device 10 is placed in the UF membrane tube 15. The water is pushed while being pushed out to clean the inside of the UF membrane tube 15.
Next, a flowchart of a subroutine program for the cleaning process shown in S153 will be described with reference to FIG. A process switching process is performed in S225, and a cleaning process flow path control process is performed in S226.
Next, a flowchart of a subroutine program for the process switching process shown in S225 will be described with reference to FIG. In S230, it is determined whether or not the cleaning process flag is ON. If it is not ON, this subroutine program returns. If the cleaning process flag is ON, the control advances to S231 to determine whether or not one unit cleaning time has elapsed. The one unit cleaning time is a time period from when the cleaning process 1 (see FIG. 12) and the cleaning process 2 (see FIG. 13) are executed a plurality of times until the cleaning process is completed. This 1 unit cleaning time is set in advance so that it can be adjusted to a desired time. If one unit cleaning time has not yet elapsed since the start of the cleaning process, NO is determined in S231, and the control advances to S234 to determine whether the reverse operation time has elapsed. This reverse operation time is the reverse operation time input and set in advance in S73. When the reverse operation time has not elapsed, the subroutine program returns. However, when the reverse operation time has elapsed, the control proceeds to S235, and it is determined whether or not the reverse direction flag is ON. When the reverse flag is ON, control to switch the reverse flag OFF is performed at S236, and when the reverse flag is not ON, control to switch the reverse flag to ON is performed at S237. Is called.
Next, a subroutine program for the cleaning process flow path control process shown in S226 will be described with reference to FIG. In S240, it is determined whether or not the cleaning process flag is ON. If it is not ON, this subroutine program returns. If the cleaning process flag is ON, the control advances to S241 to determine whether or not the reverse direction flag is OFF. Since this reverse direction flag is not switched in the above-described water pushing process (see FIG. 23), the reverse direction flag at the end of the concentration process is carried over as it is in the cleaning process and is determined in S241. As a result, at the start of the cleaning process, control is performed so that the cleaning liquid flows in the same direction as the flow direction of the stock solution (concentrated liquid) that has been executed at the end of the concentration process. If it reduces, control which starts washing | cleaning by flowing a washing | cleaning liquid in the direction opposite to the direction of the washing | cleaning liquid currently performed by the water pushing process will be performed. Specifically, when the reverse direction flag is ON, the control proceeds to S243, the motor valve MV1 is closed, the motor valve MV3 is switched to the flow path C → A, the motor valve MV2 is opened, and the motor valve MV4 is set to B Control is performed to switch to the flow path of C and switch the motor valve MV9 to the flow path of C to A. As a result, as shown in FIG. 13, a cleaning process 2 is performed in which the cleaning liquid is flowed in the reverse direction with respect to the tubular membrane separation apparatus 10.
If the reverse flag is OFF, the control proceeds to S242, the motor valve MV1 is closed, the motor valve MV3 is switched to the flow path B → C, the motor valve MV2 is opened, and the motor valve MV4 is changed from C → A. Control is performed to switch to the flow path and switch the motor valve MV9 to the flow path C → A. As a result, as shown in FIG. 12, the cleaning is performed by flowing the cleaning liquid in the forward direction with respect to the tubular membrane separator 10. Since the reverse flag is switched when the reverse rotation time elapses (S234 to S237), the cleaning process is switched from the cleaning process 1 to 2 or from the cleaning process 2 to 1.
Next, the flowchart of the subroutine program of the concentrate discharging process shown in S154 will be described with reference to FIG. In S250, it is determined whether or not the waste liquid discharge flag is ON. If it is not ON, this subroutine program returns. If it is ON, the control advances to S251, and control for opening the motor valve MV7 is performed. As a result, the concentrate stored in the concentration circulation tank 2 is discharged to the concentrate relay tank 3 via the motor valve MV7. Next, in S252, it is determined whether or not the value of the level sensor LS1 has reached the discharge completion level LL1 in the concentration circulation tank 2. If it is not yet LL1, this subroutine program returns, but at the stage when LL1 is reached, a determination of YES is made in S252, control proceeds to S253, the waste liquid discharge flag is turned OFF, and batch is performed in S254. This subroutine program returns after the process of updating the number counter by 1 is performed.
Next, the flowchart of the subroutine program for the monitor display process shown in S54 will be described with reference to FIG. In S260, it is determined whether or not a monitor display operation has been performed. If not, the subroutine program returns. If the operator operates the input operation unit 64 to input a monitor display operation, a determination of YES is made in S260 and the control advances to S261, and a menu for displaying various display items to be displayed on the display unit 65 is displayed. Display is executed. If the operator who sees it selects the menu to be displayed by the input operation unit 64, a determination of YES is made in S262, the control advances to S263, and the selected item is displayed on the display unit 65.
Specific examples of the monitor display include, for example, automatic cleaning time, water pushing time, reverse operation time, waste liquid discharge time, full water abnormality level HH1, operation start level H1, operation stop level L1, discharge completion level LL1 in the concentration circulation tank 2 , Current water level, full water abnormality level HH2 in concentrate relay tank 3, level H1 closing MV7, level L2 starting operation of concentrate feed pump 31, level LL2 stopping concentrate feed pump 31, current water level, Various input operation setting values such as the number of times of concentration (the number of times of pouring) and the number of times of batch are displayed on the monitor. Furthermore, error display corresponding to various error flags that are turned on may be performed.
Next, an embodiment in which optimum filtration control is performed using machine learning by artificial intelligence will be described with reference to FIGS. Referring to FIG. 26, each of the device control panels 23 of the large number of filtration processing devices is connected to the artificial intelligence server 55 via the Internet 60. Various data generated during the operation of the filtration device is transmitted from each device control panel 23 to the artificial intelligence server 55, and the artificial intelligence server 5 performs machine learning based on the transmitted various data, and the learning result is obtained. The reflected control command is returned to each device control panel 23. Such a service may be executed by the artificial intelligence server 55 as a cloud service. The artificial intelligence server 55 uses a general Neumann computer, but may also use a neural network processor (NNP). A large number of “artificial neurons” modeled on real neurons are mounted on the NNP chip, and each neuron cooperates in a network. Further, a quantum computer employing a “quantum annealing method” may be used. Thereby, the time required for the optimization calculation in machine learning can be greatly shortened. “Artificial intelligence” is a broad concept that includes software agents. Further, machine learning described later may be performed in combination with deep learning.
A learning database 56 and a device control panel database 57 are connected to the artificial intelligence server 55. The learning database 56 includes various filtering environments corresponding to the environment in reinforcement learning, state data s indicating the state of the filtering environment, and action data a as an action performed by the artificial intelligence server 55 on the filtering environment. It is stored for each filtering environment.
The filtration environment is classified according to the type of waste liquid to be filtered, such as waste liquid from a food factory, waste liquid from a mechanical grinding factory, waste liquid from a petrochemical plant factory, waste liquid from a chemical factory, and the like. For example, even if it is a filtration environment which filters the waste liquid from a food factory, you may classify | categorize a filtration environment further subdivided for every kind of the foodstuff made into object. Further, if the artificial intelligence stores the waste liquid component as knowledge for each classified filtration environment, a more useful service can be provided.
The state data s is data for specifying the current state in the filtration environment based on the data transmitted from each device control panel 23. For example, the time required for the value of the level sensor LS1 to decrease from the operation start level H1 to the operation stop level L1, error information based on ON error flag data, and the like. The shorter the time for the level sensor LS1 to decrease from H1 to L1, the more efficiently the concentrate in the concentration circulation tank 2 is concentrated. The shorter the time, the higher the reward r is given to the artificial intelligence server 55. This is given to the agent engine 56 for reinforcement learning. On the other hand, based on the error information that has occurred, the lower the error occurrence frequency, the higher the reward r is given to the reinforcement learning agent engine 56.
Each device control panel 23 transmits the time required for the level sensor LS1 to decrease from H1 to L1 to the artificial intelligence server 55 via the Internet 60. In addition, the device control panel 23 transmits error flag data that is ON among the above-described error flags to the artificial intelligence server 55 via the Internet 60.
Examples of the action data a instructed to the device control panel 23 to which the artificial intelligence server 55 corresponds include, for example, the number of times of concentration (number of times of addition) NK, automatic cleaning time, reverse operation time, heat exchange operation temperature, concentrated liquid circulation flow rate, cleaning liquid. Circulation flow rate, concentration pressure, etc.
The device control panel database 57 records the filtration environment in which the device control panel 23 is currently performing the filtration process in association with the control panel ID assigned to each device control panel 23.
Next, the control circuit of the artificial intelligence server 55 will be described with reference to FIG. The artificial intelligence server 55 is provided with a CPU 161 as a control center, a ROM 163 storing a control program and control data, and a RAM 162 functioning as a work area for the CPU 161. In addition, a bus 164 for transferring data control signals and an interface unit 165 for transmitting / receiving data to / from an external device are provided. First, a communication unit 166 for transmitting and receiving signals and data via the Internet 60, a display unit 167 for displaying various information to the operator, and an input for the operator to perform an input operation on the artificial intelligence server 55 An operation unit 168 and the like are provided.
Next, the basic principle of reinforcement learning will be described based on FIG. The reinforcement learning agent engine 56 stored in the artificial intelligence server 55 exchanges information with the filtration environment S. The filtering environment S can be modeled by a discrete set of states S = {s | sεS}. When the reinforcement learning agent engine 56 performs the action a on such an environment S, a reward r for the action a is obtained. The machine learning agent engine 56 acquires data on the state s of the filtration environment S, determines an action a based on the state s, and executes the action a on the filtration environment S.
A control program for such an artificial intelligence server 55 will be described. FIG. 27C shows a main routine of the artificial intelligence server. A filtering environment classification process is performed through S270. Next, reinforcement learning processing is performed in S271. The filtration environment classification process determines which filtration environment the filtration processing apparatus currently performing the filtration process belongs to.
Specifically, this will be described with reference to FIG. In S275, it is determined whether or not filtration environment specifying data has been transmitted from the apparatus control panel 23. The filtration environment specifying data is data for specifying the waste liquid to be filtered, such as food factory waste liquid, chemical factory waste liquid, and the like, and it is considered that the operator manually inputs the input operation unit 64 and transmits it to the artificial intelligence server 55. . If the filtration environment specifying data is received, the control proceeds to S276, where it is determined whether or not there is a corresponding filtration environment among the filtration environments already stored in the learning database 56. If there is a corresponding filtration environment, the control advances to S279, and processing for storing the corresponding filtration environment in association with the transmitted ID of the apparatus control panel is performed. Next, in S280, the action data a corresponding to the filtration environment is calculated from the learning database 56 and returned to the apparatus control panel 23. In response to this, the apparatus control panel 23 controls the filtration apparatus in accordance with the received action data a (for example, the number of times of concentration, automatic cleaning time, reverse operation time, etc.). The state data s as the control result is transmitted to the artificial intelligence server 55 via the Internet 60 again.
On the other hand, if the filtration environment specifying data transmitted from the device control panel 23 does not apply to the filtration environment stored in the learning database 56, the control advances to S281 to create a new filtration environment and create the device control panel. Is stored in the device control panel database 57 in association with the ID of the device ID. Next, the process proceeds to S282, and processing for requesting the initial value of the action data a to the apparatus control panel 23 is performed. In response to this, the device control panel 23 displays the fact on the display unit 65, and the operator who sees it inputs the action data a that seems to be suitable for the filtration environment from the input operation unit 64 and passes through the Internet 60. And transmitted to the artificial intelligence server 50.
The artificial intelligence server 55 that has received the determination makes a YES determination in S277, and a process of storing the new filtering environment and the initial value of the action data a in the learning database 56 is performed in S278.
The initial value of action data a for a new filtration environment may be set by analogy on the artificial intelligence server 55 side. At that time, a rough component or the like of the liquid to be filtered to be newly filtered is input from the input operation unit 64 and transmitted to the artificial intelligence server 55 via the Internet 60. In response to this, the artificial intelligence server 55 determines the initial value of the action data a based on the components of the filtration target liquid (waste liquid) for each filtration environment already stored in the learning database 56. For example, a regression method in supervised learning may be used. Regression is an algorithm for obtaining a reasonable output value predicted from an input. In order to obtain an appropriate output value for unknown data, it is considered that a target is output based on a function with input data, and the problem of obtaining the function is a regression problem.
Next, a flowchart of a subroutine program for reinforcement learning processing shown in S271 will be described with reference to FIG. It is determined whether or not the status data s has been received in S283. If not received, this subroutine program returns.
The device control board 23 transmits the status data s and its own control board ID to the artificial intelligence server 55 via the Internet 60. If it is received, the control proceeds to S284, the device control board database 57 is searched, the filtration environment corresponding to the received control board ID is determined, the learning database 56 is searched, and the state corresponding to the filtration environment A process for updating the data s is performed. This update can be done in multiple types. A possible method is to update the average of the state data s as an average. For example, when the number of device control panels 23 that have transmitted the state data s as the filtration environment A is N and the state data that is newly transmitted this time is ss, the updated state data is (s × N + ss) / (N + 1).
As a second conceivable method, there are two types of the above-described average as a whole and state data unique to the device control panel 23 that has been newly transmitted, and each uses weighted data. The average state data as a whole can be calculated by the above-described formula. The unique state data is data obtained by tabulating state data transmitted for each control panel ID. When the overall average state data is sz, the specific state data is sk, the weight of the total average state data is wg, and the weight of the specific state data is wd, the artificial intelligence server 55 makes a response to the specific device control board 23. When calculating the action data a, w1 × sz + w2 × sk is used as the state data. However, w1 + w2 = 1.
In this way, it is possible to feed back to the device control panel 23 the action data a having specific validity, including the state data unique to each filtration processing device.
After updating the state data s, the control proceeds to S285, and a process for calculating the reward r is performed. The reception of the state data s in S283 is data transmitted from the device control board 23 as a result of feeding back the previous action data a to the device control board 23, and the reward is based on the state data s that is the result data. r is calculated. For example, if the time required for the level sensor LS1 to decrease from the driving start level H1 to the driving stop level L1 is shortened, a higher reward r is calculated, and the smaller the error that has occurred, the higher the reward r is. calculate.
Next, in S286, Q (a) = r _a As described above, the process of calculating the action a having the highest Q value is performed. If the value of action a is defined as Q (a) and the correct value of Q (a) (hereinafter referred to as Q value) is obtained by the learning process, the action that maximizes the Q value becomes the learning result. At first, in order not to know how much reward can be obtained by performing action a, the value of Q (a) is initialized to 0 for all actions a. Next, possible a is performed in order, and the reward r at that time _a To get. And for each a Q (a) = r _a A having the highest Q value is obtained. The action data a stored in the learning database 56 is updated to the new action data a calculated in S286, and the action data a is transmitted to the corresponding device control panel 23 and fed back.
The following invention is disclosed by embodiment described above.
(1) In the wastewater treatment facility described in Japanese Patent Application Laid-Open No. 2014-14745, for example, an ultrafiltration membrane having a fine hole that allows water to pass through but blocks contaminants such as oil is formed into a cylindrical shape. Water was filtered and separated from pollutants by pumping and flowing industrial wastewater into the cylinder. Then, when the equipment is stopped, the ultrafiltration membrane is washed with water. However, in an area where it is difficult to secure water (for example, a desert area), it is difficult to secure sufficient water for cleaning, and there is a disadvantage that sufficient cleaning cannot be performed. Moreover, when it wash | cleans using a tap water in the area where water supply is complete, the problem that a water charge is added according to the usage fee for the tap water arises.
The filtration processing device conceived in view of the actual situation is a filtration processing device that uses a filtration membrane (for example, a UF membrane tube 15) to filter a liquid to be filtered by a cross flow method and separates it into a processed liquid and a concentrated liquid. Because
A filtration processing unit (for example, a tubular membrane separation device 10) for flowing and filtering the liquid to be filtered along the filtration membrane;
Cleaning means (for example, the membrane cleaning tank 8, the device control panel 23, the three-

way valves

36, 37, 38, the receiving mechanism 18, the sponge ball 17) for cleaning the deposits attached to the filtration membrane by the filtration treatment;
A treated liquid storage tank (for example, a membrane cleaning tank 8) for collecting and storing the treated liquid,
The cleaning means is a cleaning liquid crossflow means (for example, the membrane cleaning tank 8, the process switching process in FIG. 24A, the cleaning process in FIG. 24B) for flowing a cleaning liquid along the filtration membrane. Including flow path switching processing)
The cleaning liquid crossflow means includes processed liquid use means (for example, S79, motor valve MV6) for using the processed liquid stored in the processed liquid storage tank as the cleaning liquid.
With such a configuration, it is possible to wash the deposits attached to the filtration membrane by effectively using the treatment liquid generated by filtration.
(2) In the wastewater treatment facility described in Japanese Patent Application Laid-Open No. 2014-14745, a cleaning liquid tank for storing water (or cleaning liquid) is provided, and water (or cleaning liquid) is washed through the ultrafiltration membrane when the facility is stopped. Yes. The cleaning liquid that has passed through the ultrafiltration membrane is returned to the cleaning liquid tank again.
However, at the start of cleaning, industrial wastewater still remains in the ultrafiltration membrane, and the industrial wastewater is pushed out into water (or cleaning liquid) and returned to the cleaning liquid tank. As a result, there is a drawback that the industrial wastewater is mixed with the water (or cleaning liquid) in the cleaning liquid tank to become dirty water (or cleaning liquid).
If the industrial wastewater pushed out at the start of cleaning is discarded without being returned to the cleaning liquid tank in order to eliminate such drawbacks, there is a disadvantage that the disposal fee becomes high. That is, when disposing of industrial waste, a fee is determined according to the type and amount of waste, and the disposal fee increases as the amount of waste increases.
The filtration processing device conceived in view of the actual situation is a filtration processing device that uses a filtration membrane (for example, a UF membrane tube 15) to filter a liquid to be filtered by a cross flow method and separates it into a processed liquid and a concentrated liquid. Because
A concentration circulation tank (for example, concentration circulation tank 2) for storing the filtration target liquid and receiving and storing the concentrated liquid after the filtration target liquid is supplied to the filtration membrane and filtered;
A filtration processing unit (for example, a tubular membrane separation device 10) for flowing and filtering the liquid to be filtered along the filtration membrane;
Cleaning means (for example, a membrane cleaning tank 8, a device control panel 23, three-

way valves

36, 37, 38, a receiving mechanism 18, a sponge ball for cleaning the deposits adhered to the filtration membrane by a filtration treatment) 17)
In order to switch from a filtration process (for example, FIGS. 1 and 10) for filtering the liquid to be filtered by the filtration processing unit to a cleaning process (for example, FIGS. 12 and 13) for cleaning the filtration membrane by flowing the cleaning liquid. Process switching means (for example, the process switching process in FIG. 22B, the process switching process in FIG. 23B),
When switching from the filtration step to the washing step, a flushing means (for example, a water pushing step flow path control process in FIG. 23C) for flushing the remaining filtration target liquid with the washing liquid;
And a filtration target liquid reduction means (for example, S217, S218, motor valves MV3, MV4, MV9) for reducing the filtration target liquid washed away by the flushing means to the concentration circulation tank.
According to such a configuration, the liquid to be filtered pushed away by the flushing means is returned to the concentration circulation tank and further concentration is possible, and the waste amount of the liquid to be filtered can be reduced as much as possible.
Moreover, you may further provide the flushing time setting means (for example, S70, S71) for changing and setting the flushing time by the flushing means.
Furthermore, a cleaning liquid storage tank (for example, a film cleaning tank 8) for storing the cleaning liquid is further provided,
The cleaning means returns the cleaning liquid after the filtration membrane cleaning to the cleaning liquid storage tank (for example, S242, S243, motor valve MV9) in the cleaning step executed after the flushing by the flushing means is completed. .
(3) In a filtration apparatus that uses a filtration membrane to filter a liquid to be filtered by a crossflow method and separates it into a treated liquid and a concentrated liquid, filtration proceeds through various steps.
An object of the present invention is to provide a filtration apparatus capable of automatically switching each process necessary for performing a good filtration process and omitting an artificial operation as much as possible.
The filtration apparatus devised in view of the actual situation is subject to filtration by a cross-flow method using a filtration membrane (eg, UF membrane tube 15) provided in a filtration membrane unit (eg, tubular membrane separation device 10). A filtration apparatus for filtering a liquid to separate a processed liquid and a concentrated liquid,
Concentration reduction which reduces the concentrate which separated by supplying the filtration object liquid stored in the concentration circulation tank (for example, the concentration circulation tank 2) to the said filtration membrane unit and filtering with the said filtration membrane to a concentration circulation tank Concentration reduction means for executing the process (for example, S197, S198, motor valves MV1, MV2, MV3, MV4, MV9),
A flushing means for supplying a cleaning liquid to the filtration membrane unit and flushing the liquid to be filtered remaining in the filtration membrane unit to reduce it to the concentration circulation tank (for example, FIG. 23C) Water pushing process flow path control process),
Cleaning means (for example, a membrane cleaning tank 8, an apparatus control panel 23, three-

way valves

36, 37, 38, a receiving mechanism 18, a sponge ball) that executes a cleaning process for cleaning the deposits attached to the filtration membrane with a cleaning liquid 17)
And a process switching control means (for example, a flow path switching process in FIG. 21A) for performing a process switching control for automatically switching the processes to be executed in the order of the concentration reduction process, the flushing process, and the cleaning process. ,
The process switching control means returns to the concentration reduction process again after the cleaning process is completed and repeatedly executes the process switching control (for example, S231 to S233).
(4) In the wastewater treatment facility described in Japanese Patent Application Laid-Open No. 2014-14745, washing is performed by flowing water through the ultrafiltration membrane when the facility is stopped. However, there is a drawback that it is difficult to perform a filtration process suitable for the type of liquid to be filtered.
The filtration processing device conceived in view of the actual situation is a filtration processing device that uses a filtration membrane (for example, a UF membrane tube 15) to filter a liquid to be filtered by a cross flow method and separates it into a processed liquid and a concentrated liquid. Because
A filtration processing unit (for example, a tubular membrane separation device 10) for flowing and filtering the liquid to be filtered along the filtration membrane;
Cleaning means (for example, the membrane cleaning tank 8, the device control panel 23, the three-

way valves

36, 37, 38, the receiving mechanism 18, the sponge ball 17) for cleaning the deposits attached to the filtration membrane by the filtration treatment;
Machine learning means for performing machine learning for acquiring knowledge adapted to each filtration environment (for example, filtration environments A, B... In FIG. 26) classified according to the type of the liquid to be filtered (for example, An artificial intelligence server 55),
The machine learning means inputs data capable of specifying the filtration efficiency of the liquid to be filtered by the filtration in the filtration processing unit as a state with respect to the filtration environment (for example, state data s in FIG. 26).
At the same time, control that affects the filtration efficiency is output as an action on the filtration environment (for example, action data a in FIG. 26), and reinforcement learning is performed to improve the filtration efficiency by repeating the input and output. Reinforcement learning means (for example, the reinforcement learning process in FIG. 26B) is included.
With such a configuration, filtration processing reflecting the result of reinforcement learning becomes possible, and filtration processing performance can be improved.
(5) In the wastewater treatment facility described in Japanese Patent Application Laid-Open No. 2014-14745, washing is performed by flowing water through the ultrafiltration membrane. When the pump for supplying the stock solution to the ultrafiltration membrane is stopped in order to stop the equipment, a back flow of the stock solution (concentrated solution) may occur in the pump. When backflow occurs, the ultrafiltration membrane becomes negative pressure. If this ultrafiltration membrane is composed of, for example, a tube having no shape-retaining force, the disadvantage is that the treated water flows backward due to the negative pressure and the tube of the ultrafiltration membrane is crushed and becomes flat. Arise.
An object of the present invention devised in view of such circumstances is to eliminate the disadvantage that the inside of the filtration membrane tube becomes negative pressure when the supply of the liquid to be filtered to the filtration membrane tube is stopped.
The present invention is a filtration apparatus for filtering a liquid to be filtered using a filtration membrane tube (for example, a UF membrane tube 15) to separate it into a treated liquid and a concentrated liquid,
A feeding means (for example, a circulation pump 33) for feeding a liquid to be filtered into the filtration membrane tube;
Communication means (for example, a motor valve MV8) is provided that allows the inside of the filtration membrane tube to communicate with the outside air when the feeding of the stock solution by the feeding means is stopped.
Thereby, when the supply of the filtration target liquid to the filtration membrane tube is stopped, the disadvantage that the inside of the filtration membrane tube becomes negative pressure can be solved.
The characteristic points and modifications of the embodiment described above will be described below.
In the water pushing step, the filtration target liquid may be washed away in a state where the remaining filtration target liquid and the cleaning liquid are partitioned by a wiping body (for example, a sponge ball). That is, when the liquid to be filtered remaining in the filtration processing unit is washed away by the flushing means, the liquid to be filtered is separated in a state where the remaining liquid to be filtered and the cleaning liquid are separated by the wiping body. Configure to flush away. Thereby, after preventing that the filtration object liquid and the washing | cleaning liquid which remain | survive in the filtration process part are mixed as much as possible, the filtration object liquid can be washed away, and the filtration object liquid can be washed away effectively.
In the above-described embodiment, PVDF (polyvinylidene fluoride) is used as the material of the UF membrane (ultrafiltration membrane), but other materials may be used. In the above-described embodiment, a UF membrane (ultrafiltration membrane) is adopted as an example of the filtration membrane, but the filtration membrane is not limited to this. For example, an MF membrane (microfiltration membrane) may be used. In addition, separation and concentration at the ion level may be made possible by using an RO membrane (reverse osmosis) and an NF membrane (nanofiltration membrane). As for the pore size of the membrane, the RO membrane has a salt removal rate of about 99 to 99.8%, the NF membrane has a salt removal rate of about 40 to 97%, and the UF membrane has a membrane pore size of about 0.001 to 0. The MF membrane has a pore diameter of about 0.01 to 10 μm.
In the above-described embodiment, the urethane type is used as the material of the sponge ball. However, the present invention is not limited to this, and a vinyl type, a rubber type, or a polyethylene type may be used.
In the above-described embodiment, the automatic cleaning time (S69), the water pushing time (S71), the reverse operation time (S73), the heat exchange operation temperature (S85), the processing flow rate drop value SR (S91), which are input as set values, The upper limit JRU and lower limit JRL (S93) of the circulation flow rate, the concentration discharge error time NHT (S97), and the like were fixed values that do not vary. However, these values are controlled so as to vary to optimum values according to the progress of the concentration process (for example, the current number of times NK, the elapsed time of the concentration process, etc.) and / or the progress of the cleaning process. May be. If such variation control (dynamic optimization control) is performed by using the machine learning (for example, reinforcement learning) by the artificial intelligence server 55 described above, it can be realized with minimal human labor.
You may control so that the filtration apparatus shown by the above-mentioned embodiment can be monitored remotely. For example, an Ethernet unit is built in a filtration apparatus installed in the field, and the following can be performed by way of a PLC (Programmable Logic Controller).
a Filtration device sends an operation record by e-mail.
b Send e-mail when the filtration device is abnormal.
c The filtering device that received the mail from the PC resets the abnormality.
d The filter device that has received the mail from the PC starts / stops operation.
The stock solution water pump 27 shown in the above-described embodiment uses a mechanical seal pump. For example, a seal is provided at a location where the pump shaft passes through the casing, and the inside and outside of the casing are shut off to prevent leakage of internal liquid and intrusion of air or liquid from the outside. A magnet pump (a kind of sealless pump) may be used instead of the mechanical seal pump. This magnetic pump stops the penetration of the power transmission shaft from the outside to the inside of the pump in order to prevent liquid leakage from the seal that cannot be avoided due to the shaft seal mechanism of the sealed pumps. The power is transmitted by a permanent magnet or an electromagnet with a gap. Therefore, there is no leakage because there is no shaft seal. This is a feature of the magnet pump. This non-leakage quality creates reliability and safety. In addition, pumps other than the raw liquid feed pump 27 (pH adjustment water pump 30, alkali injection pump 28, acid injection pump 29, circulation pump 33, treated water delivery pump 34, cleaning agent injection pump 32, and concentrate feed pump 31) are also available. A magnet pump may be used instead of the mechanical seal pump.
As mentioned above, although embodiment of this invention was explained in full detail, this invention is not limited to these embodiment, A various change is possible in the range which does not deviate from the meaning of this invention.

DESCRIPTION OF SYMBOLS 1 Raw water tank 2 Concentration circulation tank 3 Concentrated liquid relay tank 4 Concentration storage tank 8 Membrane washing tank 10 Tubular type membrane separator 14 Housing 15 UF membrane tube 17 Sponge ball 18 Holding cylinder 19 Strainer 23 Apparatus control board 55 Artificial intelligence server 56 Learning Database 108 Liquid inlet / outlet

Claims

A filtration processing apparatus that separates a liquid to be filtered and a concentrated liquid by filtering the liquid to be filtered by a cross flow method using a tubular filtration membrane,
The tubular filtration membrane is built-in, and a filtration processing unit for flowing and filtering the liquid to be filtered in the tubular filtration membrane;
Cleaning means for cleaning deposits adhering to the inner surface of the tubular filtration membrane by filtration treatment,
The cleaning means moves the wiping body that is in contact with the inner surface of the tubular filtration membrane to wipe off the adhering matter adhering to the inner surface of the tubular filtration membrane,
The filtration apparatus further comprising an exchange mechanism for exchanging the wiping body that has moved through the tubular filtration membrane and has come out of the filtration section.
The filtration processing unit has an inlet / outlet for the liquid to be filtered formed in the first place and the second place, and the movement of the liquid to be filtered that has entered from one of the inlet / outlet until the other inlet / outlet is reached. It is configured to have one route,
The wiping body is moved by a liquid cross flow and wipes off deposits adhering to the inner surface of the tubular filtration membrane.
The exchange mechanism has a wiping body receiving mechanism that moves through the tubular filtration membrane and receives the wiping body that has come out from the entrance, and allows the passage of the concentrate.
2. The filtration apparatus according to claim 1, wherein the wiping body receiving mechanism has a take-out mechanism for taking out the received wiping body and making it replaceable.
3. The filtration apparatus according to claim 2, wherein the movement path is bent in the filtration processing section so that the movement path is longer than the entire length of the filtration processing section.
4. The filtration apparatus according to claim 1, wherein the cleaning means includes a cleaning liquid cross-flow means for flowing a cleaning liquid along the tubular filtration membrane.
Further comprising a treated liquid storage tank for collecting and storing the treated liquid;
The filtration processing apparatus according to claim 4, wherein the cleaning liquid crossflow means includes a processed liquid using means for using the processed liquid stored in the processed liquid storage tank as the cleaning liquid. .
A process switching means for switching from a filtration process of filtering the liquid to be filtered by the filtration processing unit to a cleaning process of flowing and washing the cleaning liquid;
The filtration processing apparatus according to claim 4, further comprising a flushing means for flushing the remaining liquid to be filtered with the washing liquid when the filtration process is switched to the washing process.
The filtration process according to any one of claims 1 to 6, further comprising a time adjusting unit for adjusting an execution time of a filtration step of filtering the liquid to be filtered by the filtration processing unit and a cleaning step of the cleaning unit. apparatus.
A concentration circulation tank for storing the filtrate to be filtered and receiving and storing the filtrate after the filtrate to be filtered is supplied to the filtration processing unit and filtered;
The filtration target liquid in the concentration circulation tank is concentrated by the filtration by the filtration processing part as the filtration target liquid returns from the concentration circulation tank to the concentration circulation tank via the filtration processing part. And a stock solution adding means for adding a stock solution into the concentration circulation tank in response to being reduced to a predetermined storage amount,
2. The apparatus according to claim 1, further comprising: a concentrate extracting means for extracting the concentrate stored in the concentration circulation tank when the number of times of addition by the stock solution adding means has reached a predetermined number. The filtration apparatus according to any one of ~ 7.
Machine learning means for performing machine learning to acquire knowledge adapted to each filtration environment classified according to the type of liquid to be filtered;
The machine learning means inputs data capable of specifying the filtration efficiency of the liquid to be filtered by filtration in the filtration processing unit as a state for the filtration environment, and performs control that affects the filtration efficiency as an action for the filtration environment. 9. The filtration processing apparatus according to claim 1, further comprising reinforcement learning means for performing reinforcement learning for outputting and repeating the input and output to improve the filtration efficiency.