JP3311158B2 - Operation control device for desalination plant - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an operation control device for a fresh water plant using a reverse osmosis membrane module.

[0002]

2. Description of the Related Art Conventionally, a desalination plant using a reverse osmosis membrane module is known. In controlling the operation of this desalination plant, there is a demand for an efficient operation such as performing an appropriate operation according to the conditions of the raw water and the state of the membrane, and measuring the life of the membrane.

[0003] However, conventionally, the film is periodically cleaned without monitoring the contamination of the film, so that the film is cleaned more than necessary so that the film is not contaminated, and the cleaning cannot be performed when the cleaning is required. There was something.

For example, if the cleaning cycle comes before the membrane is clean, unnecessary use of chemicals, unnecessary work on the operator, and a sudden change in pressure during cleaning adversely affect the membrane. give.

On the other hand, if the film becomes dirty before the cleaning cycle, the film is clogged due to the dirt and requires more pressure than necessary. The membrane cannot be used without dropping, or pressure is applied to another membrane module, causing deterioration of the membrane.

In addition, since the contamination of the membrane is not uniform throughout the plant but differs for each series or membrane module, an appropriate cleaning time is different, and the above-described problem occurs if the cleaning is performed at the same cleaning cycle. . In addition, the conditions of raw water vary depending on the season and the conditions of the surrounding sea area, and the time when the membrane is easily contaminated and the time when the fouling is likely to be different.Therefore, even in the same membrane module, the appropriate cleaning cycle differs each time, and as described above. Problems occur.

[0007] In the conventional periodic cleaning, these problems occur, so that it is not possible to extend the life of an expensive film and perform an efficient operation.

[0008]

In a fresh water generator having a reverse osmosis membrane module, the amount of permeated salt increases due to an increase in water temperature, and the amount of permeated salt increases even when the membrane is deteriorated. Conventionally, attempts have been made to monitor the amount of permeated salt to diagnose membrane fouling.However, whether the increase in permeated salt is due to an increase in water temperature or deterioration of the membrane cannot be detected early. In some cases, the replacement of the membrane may be delayed, or, on the contrary, the membrane which is still usable may be replaced. In addition, it is considered that the membrane is periodically replaced for fear of an accident due to deterioration of the membrane. However, a usable membrane may be replaced.
As described above, conventional monitoring of the amount of permeated salt and regular replacement of the membrane cannot make the most of the expensive membrane.

In a fresh water generator having a reverse osmosis membrane module, the amount of permeated water tends to increase as the water temperature rises, and is proportional to the supply water pressure. Conventionally, in order to obtain a constant amount of water from a fresh water generator, the supply water pressure is reduced when the water temperature rises, or the number of operating membrane modules is changed about once a year. Since the capacity of the fresh water generator is remarkably increased as the water temperature rises, if this is used to change the number of membrane modules to be operated, the operation cost can be reduced. In addition, when the number of membrane modules to be operated is changed, it is more efficient to selectively halt and clean the membrane in which dirt or deterioration has progressed. However, in the conventional operation, since the number of membrane modules is not largely changed, efficient operation cannot be performed.

[0010] The present invention has been made in consideration of such points, and according to the contamination and deterioration of the membrane and the conditions of the raw water,
It is an object of the present invention to provide an operation control device of a fresh water plant using a reverse osmosis membrane module that can perform cleaning of a membrane at an appropriate time and enable automatic operation with an appropriate number of membrane modules.

[0011]

According to the first aspect of the present invention,
It has a reverse osmosis membrane module that makes permeate through feed water and makes non-permeate water concentrated water, a cleaning device that periodically cleans the reverse osmosis membrane module, and measurement means that measures a state value of a plant. In the operation control device of the desalination plant, an abnormality notification unit that notifies a plant abnormality,
Based on the feedwater pressure measured by the measuring means, the feedwater side membrane surface osmotic pressure of the reverse osmosis membrane module, and the permeate flow rate, the water permeability coefficient of the reverse osmosis process is obtained from the following equation,

(Equation 2) When the membrane state value obtained by subtracting the initial value from the water permeability coefficient is equal to or less than the first reference value, it is diagnosed that the membrane is dirty, and the cleaning device is activated. An operation control device for a desalination plant, comprising: a membrane state diagnosis unit that diagnoses and activates the abnormality notification unit.

According to a second aspect of the present invention, the membrane condition diagnosing means diagnoses the membrane contamination if the value obtained by subtracting the initial value from the water permeability coefficient is equal to or less than the first reference value within a predetermined time. 2. The operation control device for a fresh water plant according to claim 1, further comprising fuzzy inference means for not diagnosing the deterioration of the membrane if the return is made within a predetermined time even if the return is within the second reference value.

According to a third aspect of the present invention, the fuzzy inference means further diagnoses the flow of the membrane based on the supply water pressure and the concentration of the permeated water, and when the input number of reverse osmosis membrane modules fluctuates. 3. The operation control device for a fresh water plant according to claim 2, wherein the membrane is not diagnosed.

According to a fourth aspect of the present invention, in the membrane condition diagnosis means, the permeated water concentration, the supplied water concentration and the supplied water flow rate measured by the measuring means are further inputted, and the permeated water concentration, the supplied water concentration, the permeated water concentration are measured. The concentration difference between the membranes is determined based on the flow rate and the supply water flow rate, the amount of the permeated salt is determined from the permeated water concentration and the supplied water concentration, and the membrane state value is determined by the salt permeability coefficient obtained from the intermembrane concentration and the permeated salt content. The operation control device for a desalination plant according to claim 1, wherein

The invention according to claim 5 is characterized in that the membrane condition diagnostic means corrects the water permeability coefficient and the salt permeability coefficient based on the input number of operating modules of the reverse osmosis membrane module. It is an operation control device of a desalination plant.

According to a sixth aspect of the present invention, the film state diagnostic means can learn about the causal relationship for obtaining the film state value from the plant state value using the plant state value as an input signal and the film state value as a teacher signal. The operation control device for a fresh water plant according to any one of claims 1 to 5, further comprising a membrane state learning unit including a neural network.

According to a seventh aspect of the present invention, a plurality of reverse osmosis membrane module devices are provided, and the order of cleaning is determined in the order of the reverse osmosis membrane modules diagnosed by the membrane condition diagnostic means as having membrane contamination. In the case where the membrane modules are concentrated in the same period, the membrane cleaning planning means for dispersing the cleaning time so that the other membrane modules are cleaned earlier by a predetermined time based on the membrane module with the latest cleaning order among them is further provided. The operation control device for a manufacturing plant according to claim 1, wherein:

According to the invention of claim 8, a plurality of reverse osmosis membrane module devices are provided, and the module number determining means for determining an appropriate number of operating modules from the supply water pressure and the water temperature of the supply water; 2. The operation of the desalination plant according to claim 1, further comprising module operation planning means for determining which one of the membrane modules is to be operated on the basis of the above-mentioned signal and the past operation history of each module. It is a control device.

[0019]

According to the first aspect of the present invention, in the membrane condition diagnosis means, the water permeability coefficient of the reverse osmosis process is obtained based on the feed water pressure, the osmotic pressure on the side of the feed water of the reverse osmosis membrane module, and the permeate flow rate. If the value obtained by subtracting the initial value from the water permeability coefficient is equal to or less than the first reference value, the film is diagnosed as being soiled, and if the value is equal to or greater than the second reference value, the film is diagnosed as being deteriorated. When the membrane condition diagnostic means diagnoses the contamination of the membrane, when the reverse osmosis membrane module is cleaned by the cleaning device and the deterioration of the membrane is diagnosed,
The abnormality notification unit notifies the abnormality of the plant.

According to the second aspect of the present invention, by providing the fuzzy inference means, even if the water permeability coefficient fluctuates due to the condition of the plant, if the water recovery coefficient is restored within a predetermined short time, the cleaning device or the abnormality notification means is provided. Will not be activated unnecessarily.

According to the third aspect of the present invention, the fuzzy inference means further diagnoses the contamination of the membrane based on the supply water pressure and the permeate concentration. Do not diagnose dirt.

According to the fourth aspect of the present invention, the membrane condition diagnostic means determines the transmembrane concentration and the permeated salt content, and corrects the water permeability coefficient according to the actual conditions of the plant based on the transmembrane concentration and the permeated salt content. can do.

According to the fifth aspect of the present invention, the water permeability coefficient is further corrected in accordance with the actual situation by correcting the water permeability coefficient and the salt permeability coefficient based on the number of operating modules in the membrane condition diagnostic means. be able to.

According to the sixth aspect of the present invention, the film state learning means of the film state diagnosis means uses the plant state value as an input signal, and uses the film state value as a teacher signal to convert the plant state value by a neural network. By learning the causal relationship with the value, the film state value can be obtained with high accuracy.

According to the seventh aspect of the present invention, when the reverse osmosis membrane modules diagnosed as having membrane contamination by the membrane condition diagnosis means are concentrated at the same time, the membrane cleaning planning means can execute the same.
The concentration of the cleaning time can be dispersed by advancing the cleaning time of the other film modules based on the film module having the latest cleaning order as a reference.

According to the present invention, the number of operating modules is determined from the supply water pressure and the supply water temperature by the module number determining means, and the module operation planning means is determined based on the determined number of operating modules. Determines which membrane module to operate.

[0027]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an embodiment of an operation control device of a fresh water plant using a reverse osmosis membrane module according to the present invention.

In FIG. 1, the fresh water producing process is constituted by connecting a high pressure pump 1 and a plurality of reverse osmosis membrane modules 2 having a membrane 2a, and a washing device 3 is connected to the reverse osmosis membrane module 2. I have. The plurality of reverse osmosis membrane modules 2 are connected in parallel.

A sampling pump 5 is provided in the supply water inflow path 4 of the high pressure pump 1 to the reverse osmosis membrane module 2, and the raw water passes through the sampling pipe 6 to measure the state value of the plant, for example, SDI total 7, OR
It is sent to P meter 8 and pH meter 9. Further, a water temperature gauge 10, a supply water pressure gauge 11, a supply water concentration meter 1
2. A supply water flow meter 13 is provided.

The reverse osmosis membrane module 2 has a permeated water outflow passage 14 for permeated water passing through the membrane 2a, and a concentrated water outflow passage 15 for concentrated water which does not permeate the membrane 2a and is concentrated in salt.

A sampling pump 16 is installed in the permeated water outflow passage 14, and the permeated water is supplied to the sampling pipe 1.
7 and sent to a permeate concentration meter 18. Further, a permeated water flow meter 19 is installed in the permeated water outflow passage 14,
On the other hand, a concentrated water pressure gauge 20 is installed in the concentrated water outflow passage 15.

The operation control device 21 of the reverse osmosis fresh water plant according to the present invention comprises a high pressure pump 1, a reverse osmosis membrane module 2
To achieve the desired driving,
As shown in FIG. 1 which will be described in detail below, the operation control device 21 of the reverse osmosis freshwater plant uses a fuzzy inference unit 22 to diagnose the dirt and deterioration of the reverse osmosis membrane.
3 and a film state display means 24 for displaying dirt and deterioration of the film
A cleaning confirmation unit 25 for confirming to the operator the automatic cleaning start of the membrane from the diagnosis result, and a cleaning device 3 for automatically cleaning the membrane.
And a film cleaning means 26 for operating the device.

The diagnosis result from the film state diagnosis means 23 is notified to the operator by the abnormality notification means 27. Further, the relationship between the input process value and the diagnosis result of the contamination or deterioration of the film is learned by the film state learning means 28 using a neural network, and the accuracy of the film state diagnosis is improved.

The cleaning order of each membrane module is determined by the membrane cleaning planning unit 29 based on the stain and deterioration of the diagnosed film output from the membrane state diagnosis unit 23, the past cleaning history of the film, and the cleaning capability of the cleaning device. The cleaning time is planned, and the appropriate number of operating membrane modules is determined by the module number determining means 30 from the supply water temperature and the target permeated water amount input by the operator.

The determined number of modules is displayed by the appropriate number of modules display means 31, and the module number change confirmation means 32 confirms the operator to automatically change the number of modules. After the confirmation of the number of modules, the number of modules is automatically changed by the number-of-modules changing means 33, and the number of modules output from the number-of-modules determining means 30, the operation history of each module in the past, and the operation policy input by the operator are used. In the operation planning means 34, an appropriate operation module is selected.

The comprehensive planning means 35 makes a comprehensive judgment based on the results output from the membrane cleaning planning means 29 and the module operation planning means 34 and the operation policy inputted by the operator. Perform a pause. The exchange of signals with the operator is performed via the man-machine interface means 36.

On the other hand, an SDI meter 7, an ORP meter 8, a pH meter 9, a water temperature meter 10, a supply water pressure gauge 11, a supply water concentration meter 12,
Supply water flow 13, permeate concentration meter 18, permeate flow system 1
9 and the signal from the concentrated water pressure gauge 20 are input to the operation control device 21, and the weather information is simultaneously input to the operation control device 21 through the man-machine interface means 36.

The operation control device 21 of the reverse osmosis fresh water plant outputs information on the state of contamination or deterioration of the membrane, abnormal information of the device, and the appropriate number of modules via the man-machine interface means 36. A washing start command and an operation start / stop command are output to the reverse osmosis membrane module 2.

Next, the operation of this embodiment having the above configuration will be described.

First, the feed water pressure gauge 1 installed in the plant
1, the supply water pressure Pf, the supply water temperature Temp, and the permeate flow rate Qp measured by the permeate flow meter 19 and the water thermometer 10, respectively, are input to the membrane condition diagnostic means 23. The water permeability coefficient As under the use condition indicating the water capacity is the transport equation of the reverse osmosis process (1).
It is obtained using the formula.

In equation (1), the reverse osmotic pressure difference ΔP between membranes and the osmotic pressure difference Δπ between membranes are expressed by the following equations (2), (3) and (3), respectively.
And (4). In the equation (4), the feedwater-side membrane surface concentration Cm is determined by the feedwater concentration Cf and the recovery rate RR.
However, when the operation is being performed while the permeated water amount is constant, the value is substantially constant. Therefore, the value is calculated here assuming that it is constant. The recovery rate refers to the ratio of the amount of permeated water to the amount of supplied water.

[0042] As: Water permeability coefficient under use conditions Qp: Permeate flow rate ΔP: Reverse osmotic pressure difference between membranes Δπ: Osmotic pressure difference between membranes F: Unit membrane area ΔP = Pf−Pp (2) Pf: Supply water pressure Pp: Permeate water pressure (approximately 0) Δπ = πm−πp (3) πm: Supply water side membrane surface osmotic pressure πp: Permeate water osmotic pressure (approximately 0) πm = Kπ × Temp × Cm (4) Kπ: Parameter Temp: Water temperature Cm: Supply water side membrane surface concentration Water permeation coefficient As under use conditions is supply water pressure Pf, water temperature T
emp, the water permeability coefficient Ano under standard conditions
rm is converted by equation (5). The temperature correction function number (6) and the pressure correction function (7) are obtained by conducting an experiment for each installation location of the reverse osmosis membrane module.

[0043] Anorm: water permeability coefficient under standard conditions At: A value temperature correction coefficient Ap; A value pressure correction coefficient At = f (Temp) ... (6) Ap = f (Temp) ... (7) Under standard conditions Is compared with the initial water permeability coefficient A0 before the operation of the membrane based on equation (8),
From equation (8), the film state value S is the film contamination reference value (first reference value).
If it has decreased below α, it can be determined that the film is dirty. From equation (8), the value is the film deterioration reference value (second reference value).
If it has increased beyond β, it is diagnosed that there is a risk of deterioration (that is, a risk of breakage).

S = Anorm−A 0 (t / t ₀ ) ^−m (8) S: Membrane state value A 0: Initial water permeability coefficient (standard condition) m: Parameter t: Current time t 0: Start use of membrane Time Next, the permeate concentration pressure difference ΔPm, which is the difference between the permeate pressure and the concentrate water pressure, is determined using equation (9). The permeation pressure difference ΔPm obtained by the equation (9) is compared with the reference value by the reference equation (10), and if it is increased, it is determined that the membrane is contaminated. The standard expression (10) of the permeation concentration pressure difference ΔPm is obtained by conducting an experiment for each installation location of the reverse osmosis membrane module.

ΔPm = Pf−Pb (9) Pb: concentrated water pressure ΔPm = f (Qp) (10) Further, in the membrane state display means 24, the membrane state as a result of the membrane state diagnosis means 23 is obtained. Value S, water permeability coefficient Anor in standard condition
m, permeate flow rate Qp, supply water pressure f, and water temperature Tem
A process value such as p is also displayed, and the operator is notified of the state of contamination or deterioration of the film and the progress thereof.

Further, the cleaning confirmation means 25 notifies the operator that it is determined that the film should be cleaned based on the diagnosis result of the film state diagnosis means 23. If the operator determines that it is appropriate to start the automatic cleaning of the membrane, the operator inputs the fact to the operation control device 21 through the man-machine interface means 36, and thus the confirmation of the start of the cleaning is obtained from the operator.

Next, the membrane cleaning means 26 outputs a cleaning command to the cleaning device 3 for the membrane module 2 which has requested the cleaning, after confirmation of the operator through the cleaning planning means 29 and the comprehensive planning means 35 described later. At the same time, the abnormality notification means 27
Then, based on the result of the film state diagnosis means 23, the operator is notified of the deterioration of the film, and the inspection and replacement are prompted.

Next, the function of the film condition diagnosis means 23 will be described in detail below.

The membrane state diagnosis means 23 does not obtain the membrane state value S merely by comparing the water permeability coefficient Anorm converted to the standard condition and the initial value as described above, but uses the fuzzy inference means 22 Diagnosis of film contamination and deterioration can be performed in consideration of non-linear factors such as film characteristics, site characteristics, and operator intuition. In this case, the fuzzy inference means 22 makes a diagnosis according to the following rules, for example. Rules can be added.

S ← Fuzzy inference Rule: 1) If the current water permeability coefficient Anorm is smaller than the water permeability coefficient A0 in the initial state, the membrane is dirty.

2) If the supply water pressure Qf greatly increases in a short period of time during the constant permeate flow rate operation, the membrane is dirty.

3) In a state where the membrane is not contaminated during the constant permeate flow rate operation, if the water temperature Temp increases, the permeate flow rate Qp per unit membrane F and the pressure increases, so the supply water pressure Pf becomes Decrease.

Therefore, if the supply water pressure Qf increases when the water temperature Temp rises, the film is dirty.

4) The current water permeability coefficient As under the standard conditions is compared with the initial water permeability coefficient Ao before the operation of the membrane based on equation (8). Even if it drops below, it is not determined that the film is dirty for a short period of time. According to the equation (8), even if the value is abnormally increased to the film deterioration reference value β, there is no fear of deterioration in a short time.

5) When the permeated water flow rate Qp greatly decreases in a short period of time during the constant permeated water flow rate operation, or when the permeated water concentration Cp
If it rises sharply, the membrane may be ruptured.

Further, the membrane condition diagnosis means 23 does not obtain the membrane condition S only by comparing the water permeability coefficient Anorm converted into the standard condition and the initial value, but also obtains the permeated water concentration C, the supplied water concentration Cf, and the supplied water concentration. The film state value S may be corrected from the flow rate Qf so that the state of the film can be expressed more accurately. First, the salt permeability coefficient Bs is determined by equation (11) which is an equation of the reverse osmosis process.

Since the salt permeability coefficient Bs under the use condition changes with the supply water pressure Pf and the water temperature Temp, the salt permeability coefficient Bnorm under the standard condition is converted by the equation (12). in this case,
The temperature correction function (13) and the pressure correction function (14) are obtained by performing experiments for each installation location. The concentration difference ΔC between the films in the equation (11) is obtained using the equations (15) and (16).

[0058]

(Equation 3) Next, the initial salt permeability coefficient Bo before operation of the membrane is compared with the salt permeability coefficient Bnorm converted to the standard state, and the membrane state value S is corrected based on the degree of increase in salt permeability, and the diagnosis accuracy of membrane contamination and deterioration is evaluated. Improve. The comparison formula (17) is obtained by an experiment for each setting location.

S ′ = S × f (Bnorm, Bo) (17) Bo: Initial salt permeability coefficient before operation of the membrane Next, the corrected membrane state value S ′ obtained by the equation (17) Diagnosis of contamination and deterioration of the film is performed based on

In a desalination plant in which a plurality of reverse osmosis membrane modules 2 are arranged in parallel, the number of operating reverse osmosis membrane modules (the number of operating modules) is proportional to the fresh water production capacity. Therefore, the membrane state diagnosis means 23 may input the number of operation modules moj and diagnose the contamination or deterioration of the membrane in consideration of the change in fresh water production capacity. In this case, equation (1)
Expressions (1) 'and (11)' are used instead of expression (11) to improve the accuracy of diagnosis of dirt and deterioration.

[0061] moj: number of modules Further, in the membrane condition diagnostic means 23, the feed water SDI indicating the concentration of suspended solids in the feed water, the feed water PRP indicating the oxidizing ability to promote deterioration, and the amount of suspended solids and dissolved salts in the feed water. The pH of the feed water that affects the sedimentation on the membrane and the weather information that greatly affects the quality of the raw water are further input, and taking into account the acceleration of membrane fouling and deterioration due to these raw water conditions, the equation (17) is used. The film state value S ′ corrected in step (1) may be further corrected by equation (18) to improve the accuracy of the film contamination and deterioration diagnosis.

S ″ = S ′ × f (SDI, ORP, pH) (18) In the membrane condition diagnostic means 23, the conditions of the raw water (number of operating modules moj, supply water pressure Pf, water temperature Temp, permeate water) Flow rate Qp, permeate water concentration Cp, supply water concentration Cf, supply water flow rate Qf, supply water SDI, supply water ORP, supply water p
(H, weather information), the fuzzy inference means 22 may be used to diagnose the contamination or deterioration of the film in consideration of non-linear factors such as film characteristics, site characteristics, and operator intuition. An example of the rule in this case is shown below. The aforementioned rules 1) to 5) are also used here. Rules can be added.

S ← Fuzzy inference Rule: 6) In a state where the permeate flow rate is constant and the membrane is not contaminated, the feed water pressure Pf and the permeate concentration Cp are
They are in inverse proportion.

Therefore, if the permeate concentration Cp also increases when the supply water pressure Pf increases, the membrane is contaminated. Conversely, if the permeate concentration Cp also decreases when the supply water pressure Pf decreases, the membrane is contaminated.

7) When the number of modules moj is changed,
Since there is a change in the supply water pressure Pf and the permeate concentration Cp accompanying this, diagnosis of membrane contamination is not performed.

Further, the causal relationship between the condition of the raw water and the membrane state value is determined by the membrane state diagnostic unit 23.
May be learned by In this case, the membrane state diagnostic means 2
Raw water conditions (number of operating modules mo)
d, supply water pressure Pf, water temperature Temp, permeate flow Qp,
The time series data of the permeated water concentration Cp, the supplied water concentration Cf, the supplied water flow rate Qf, the supplied water SDI, the supplied water ORP, the supplied water pH, and the weather information) are input to the neural network of the membrane state learning means 28 for a certain period of time. The input is repeatedly performed at regular intervals. When the input is accumulated, the film state diagnostic means 23
The causal relationship between the condition of the raw water and the state of the membrane is learned by a back propagation method in the neural network of the membrane state learning means 28 using the membrane state value S output as the teacher signal as a teacher signal. By repeatedly performing this learning, it becomes possible to grasp the contamination and deterioration factors of the film. The learning result in the film state learning means 28 is notified to the operator and is also reflected in the film state diagnosis means 23 to improve the accuracy of the film state diagnosis.

As described above, after diagnosing the contamination and deterioration of the film state by the film state diagnosing means 23, the diagnosis result is inputted to the film cleaning planning means 29. Membrane cleaning planning means 29
Then, based on the film state value S indicating the degree of contamination of each film input from the film state diagnosis means 23, the past cleaning history of the film, and the cleaning performance of the cleaning device 3, the range in which the film contamination is to be cleaned is determined. Depending on the capacity of the cleaning device and the progress of membrane contamination, the cleaning time of each reverse osmosis membrane module should be set in order from the reverse osmosis membrane module that arrives earlier, and if the dirt is the same degree, so that the previous cleaning is performed in the order of oldness. To plan. When the cleaning of each reverse osmosis membrane module is concentrated, the cleaning time is determined so as to perform the pre-cleaning with respect to the membrane state set value based on the permeated water concentration Cp and the membrane cleaning range.

In this case, first, from the permeate concentration Cp, the number of operating membrane modules moj, and the permeate flow Qp, the following (19)
The film cleaning range Δ is determined as shown in the equation (20).

If C1 ≦ Cp → ΔS = Sh if C1> Cp → ΔS = Si (19) C1: membrane washing range switching concentration set value Cp: permeated water concentration Sh: membrane washing range Sl: membrane washing range d2 = Qp / moj if C2 ≦ d2 → ΔS = Shif C2> d2 → ΔS = Sl (20) d2: Permeation flow rate per module C2: Membrane washing range switching concentration setting value It is determined as follows. Based on the film cleaning range ΔS, the cleaning time of the membrane module is determined as follows.

As shown in FIG. 2, from the pond of the reverse osmosis membrane module (i-1) having the lowest actual membrane state, the membrane washing range Δ
The film state S is divided for each S (ΔS> 0), and the cleaning order is k, (k-1), (k-2),
... Determine the number 1 (k: number of operating membrane modules) and the cleaning order.

At this time, as shown in FIG. 2, there is a case where the m-layer membrane state values are concentrated in a certain membrane cleaning range (No. 1 pond,
No. 2 pond, No. 3 pond: m = 3, among them, the order of cleaning 1 of each osmosis membrane module is in descending order of operation continuation time (in the case of FIG. 2, the operation continuation length is No. 2 pond> Pond No. 1> Pond No. 3), and the membrane module (3
As shown in FIG. 3, the film state value of each membrane module is corrected by the film cleaning range ΔS as shown in FIG.

Si ′ = Si + (ml) · ΔS (21) Si ′: corrected virtual film state value of i-th pond Si: actual film state value of i-th pond m: concentrated film module Number l: Cleaning order within the membrane cleaning range If there is no membrane module whose actual membrane state value has reached the membrane state set value, if the virtual membrane state obtained by calculation has reached the membrane state set value, A cleaning request is output to the membrane module. However, at this time, the film state lower limit value SL at the start of cleaning is set for the film state value at the start of cleaning, and a cleaning request is output if the actual film state value is equal to or higher than this SL.

When there is a membrane module whose actual membrane state or actual operation continuation time has reached the membrane state set value and the operation continuation time set value, a cleaning request is output to the membrane module with the highest priority. However, when there are a plurality of membrane modules that have reached, the cleaning requests are output in order from the membrane module with the largest actual membrane state value.

As described above, the equations (1) and (6) show that the fresh water production capacity changes with the change of the supply water pressure Pf and the supply water temperature Temp. , An appropriate number of operating membrane modules is determined according to changes in the supply water pressure Pf and the water temperature Temp.

Next, in the appropriate module number display means 31,
The appropriate number of modules calculated by the module number determining means 30 is displayed on the man-machine interface means 36 to notify the operator. Next, the module number change confirmation means 32
Then, the operator confirms the automatic change of the number of modules by the cleaning confirmation means 25 in the same manner. At the same time, the module number change means 33 outputs an operation and stop command of the reverse osmosis membrane module 2 after the operator is confirmed by the module number change confirmation means 32, and automatically changes the number of modules.

The signal from the module number determining means 30 is also input to the module operation planning means 34, and the module operation planning means 34 obtains the number of modules input from the module number determining means 30 and the past operation history of each module. From the operating policy input by the operator, which reverse osmosis membrane module is to be operated and stopped according to the operating policy. If the operating policy from the operator is to use the same reverse osmosis membrane module intensively and replace it after the membrane has deteriorated, select the most frequently used membrane module and operate it. And suspend the membrane module that is not frequently used. If the policy is “average use of membranes”, the operation is started from the membrane module that has been operating for the longest time among the membranes used in the past, and stopped from the one that has been operating for the longest time. .

The principle of the operation sequence determination method is the same as the principle used in the membrane cleaning planning means 29.

Next, the comprehensive planning means 35 determines which reverse osmosis membrane module is to be operated according to the operating policy based on the results output from the membrane cleaning planning means 29, the module operation planning means 34, and the operating policy input by the operator. Plan, do or pause. If it is a higher priority to operate the reverse osmosis membrane module at an appropriate time, reestablish a module operation plan in line with the membrane cleaning. Conversely,
If the priority of the operation policy is to make the number of operation modules appropriate, the cleaning plan of the reverse osmosis membrane module is re-established so as to clean the reverse osmosis membrane module to be stopped.
However, if there are more membranes to be cleaned than the number of membrane modules to be stopped, the timing for changing the number of membrane modules is postponed.

The principle of the method of determining the cleaning and operation sequence is the same as that of the membrane cleaning planning means 29.

Next, a signal from the comprehensive planning means 35 is input to the module number changing means 33 and the membrane cleaning means 26,
Reverse osmosis membrane module 2 by module number changing means 33
The number of operation modules is adjusted, and the cleaning device 3 is operated by the signal from the membrane cleaning means 26 to clean the reverse osmosis membrane module 2.

In general, in the method of injecting by determining the injection rate target value at a fixed ratio with respect to the ammonia concentration, excessive chlorine may be injected to supply tap water having a strong chlorine odor,
There is a danger of supplying incompletely disinfected tap water due to insufficient injection. However, the operation control device of the reverse osmosis freshwater plant according to the present invention can cope with the fact that the consumption of chlorine with respect to the measured ammonia concentration varies depending on the season, the weather, and the state of maintenance. Can be done,
Chlorine can be injected without excess and deficiency, and the residual chlorine concentration of the treated water is stabilized. This is useful for supplying tap water in which the residual chlorine concentration is always in a certain range. In the operation control device of the reverse osmosis membrane fresh water plant of the present invention, the injection of chlorine according to the ammonia concentration of the inflow raw water is automatically controlled, so that the labor of the operator can be reduced.

[0082]

As described above, according to the first aspect of the present invention, the membrane condition diagnostic means diagnoses dirt and deterioration of the reverse osmosis membrane module, and performs cleaning and inspection of the reverse osmosis membrane module at an appropriate time. Therefore, it is possible to extend the life of the membrane and to continuously produce a target amount of high-quality permeated water. Further, since abnormality detection of the reverse osmosis membrane module can be performed at an early stage, fatal fouling of the membrane can be prevented, and the life of the membrane can be extended. Further, since an abnormality of the reverse osmosis membrane module can be notified to the operator, it contributes to grasping the process state of the operator, and efficient and safe operation can be performed.

According to the second aspect of the present invention, even if the water permeability coefficient fluctuates, if it is restored within a predetermined short time, the cleaning device or the abnormality notifying means is not unnecessarily activated. For this reason, unnecessary false information or malfunction can be prevented.

According to the third aspect of the present invention, since the diagnosis of the film contamination is not performed when the module which causes a disturbance in the diagnosis of the film contamination is not changed, unnecessary false alarms or malfunctions can be prevented.

According to the fourth aspect of the present invention, the water permeability coefficient can be corrected according to the actual situation by the transmembrane concentration and the permeated salt content. For this reason, highly accurate operation control can be performed.

According to the fifth aspect of the present invention, since the water permeability coefficient and the salt permeability coefficient are corrected based on the number of operation modules, operation control according to the actual situation can be further performed.

According to the sixth aspect of the present invention, the causal relationship between the plant state value and the membrane state value by a neural network using the plant state value as an input signal and the membrane state value as a teacher signal based on the membrane state learning state. By learning, the film state value can be obtained with high accuracy in accordance with the actual situation of the plant, and thereby accurate operation control can be performed.

According to the seventh aspect of the present invention, the reverse osmosis membrane module to be cleaned can be prevented from being concentrated at the same time by the membrane cleaning planning means. For this reason, it is possible to disperse the amount of cleaning water used by the cleaning device and to perform an appropriate cleaning action.

According to the invention, the number of operating modules is determined from the supply water pressure and the supply water temperature by the module number determining means, and the membrane module to be operated is determined by the module operation planning means. Thereby, a plurality of membrane modules can be used in a well-balanced manner.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing an embodiment of an operation control device for a fresh water plant according to the present invention.

FIG. 2 is a view for explaining the operation of a cleaning planning means for a membrane module.

FIG. 3 is a diagram for explaining the operation of the cleaning planning means for the membrane module.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 High-pressure pump 2 Reverse osmosis membrane module 3 Cleaning device 21 Operation control device 22 Fuzzy inference means 23 Membrane state diagnosis means 24 Membrane state display means 25 Wash confirmation means 26 Membrane cleaning means 27 Abnormality notification means 28 Membrane state learning means 29 Membrane cleaning plan Means 30 Module number determination means 31 Appropriate module number display means 32 Module number change confirmation means 33 Module number change means 34 Module operation planning means 35 Overall planning means 36 Man-machine interface means

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Ryosuke Miura 1 Toshiba-cho, Fuchu-shi, Tokyo Inside the Fuchu Plant, Toshiba Corporation (72) Koji Yamashita 1-1-1, Shibaura, Minato-ku, Tokyo Shares (72) Inventor Hirokazu Kondo 1 Toshiba-cho, Fuchu-shi, Tokyo Toshiba Fuchu Plant (56) References JP-A-08-1158 (JP, A) JP-A-2-174920 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) C02F 1/44 B01D 61/12 B01D 65/02

Claims (8)

(57) [Claims] 1. A reverse osmosis membrane module that uses permeate water and non-permeate water as concentrated water through feed water, a cleaning device that periodically cleans the reverse osmosis membrane module, and a measurement that measures a state value of a plant. Means for controlling the operation of a desalination plant, comprising: an abnormality notification means for notifying a plant abnormality; a supply water pressure measured by the measurement means; a supply water side membrane surface osmotic pressure of a reverse osmosis membrane module; Based on the flow rate, the water permeability coefficient of the reverse osmosis process is obtained from the following equation. The membrane state value obtained by subtracting the initial value from this water permeability coefficient is
A film condition diagnosis means for diagnosing that the film is contaminated when the value is equal to or less than the first reference value and operating the cleaning device, and diagnosing film deterioration and operating the abnormality notification means when the value is equal to or more than the second reference value; An operation control device for a desalination plant, comprising: The membrane condition diagnosis means does not diagnose the membrane contamination if the value obtained by subtracting the initial value from the water permeability coefficient is equal to or less than the first reference value within a predetermined time. 2. The method according to claim 1, further comprising fuzzy inference means for not diagnosing the deterioration of the film if the return is made within a predetermined time even if it is equal to or more than the reference value.
An operation control device for a desalination plant as described in the above. 3. The fuzzy inference means further diagnoses the flow of the membrane based on the feed water pressure and the permeate concentration.
3. The membrane diagnosis is not performed when the input number of reverse osmosis membrane modules changes.
An operation control device for a desalination plant as described in the above. 4. The permeated water concentration, the supplied water concentration and the supplied water flow rate measured by the measuring means are further inputted to the membrane state diagnosis means, and the permeated water concentration, the supplied water concentration, the permeated water flow rate and the supplied water flow rate are inputted. In addition to calculating the concentration difference between membranes based on the concentration of permeated water and the concentration of feed water, the amount of permeated salt is calculated, and the membrane state value is corrected by the salt permeation coefficient obtained from the concentration between the membrane and the amount of permeated salt. The operation control device for a fresh water plant according to claim 1. 5. The operation control of a desalination plant according to claim 4, wherein the membrane condition diagnosis means corrects the water permeability coefficient and the salt permeability coefficient based on the input number of operation modules of the reverse osmosis membrane module. apparatus. 6. A membrane state diagnosis means for a causal relation for obtaining a membrane state value from a plant state value, wherein a membrane network having a built-in neural network capable of learning using the plant state value as an input signal and the membrane state value as a teacher signal. The operation control device for a desalination plant according to any one of claims 1 to 5, further comprising a state learning unit. 7. A plurality of reverse osmosis membrane module devices are provided, and the order of cleaning is determined in the order of the reverse osmosis membrane modules that have been diagnosed by the membrane condition diagnosis means as having membrane contamination, and the plurality of membrane modules to be cleaned are the same. In the case where the concentration is concentrated on the timing, the cleaning module is further provided with a membrane cleaning planning means for dispersing the cleaning timing so that another membrane module is cleaned earlier by a predetermined time based on the membrane module having the latest cleaning order among them. The operation control device for a manufacturing plant according to claim 1. 8. A plurality of reverse osmosis membrane module devices are provided, a module number determining means for determining an appropriate number of operating modules from a supply water pressure and a supply water temperature, and a signal from the module number determination means and The operation control device for a fresh water plant according to claim 1, further comprising: module operation planning means for determining which membrane module is to be operated based on the operation history of each module.