WO2012081746A1

WO2012081746A1 - System and method for membrane fouling diagnosis in a water-treatment process using a kalman filter algorithm

Info

Publication number: WO2012081746A1
Application number: PCT/KR2010/009020
Authority: WO
Inventors: 김준하; 이영근; 이윤석
Original assignee: 광주과학기술원
Priority date: 2010-12-16
Filing date: 2010-12-16
Publication date: 2012-06-21

Abstract

Disclosed are a system and method for membrane fouling diagnosis in a seawater desalination process using a Kalman filter algorithm. In a preferred embodiment of the present invention, the system for membrane fouling diagnosis in a water-treatment process using a Kalman filter algorithm may comprise: a plant unit wherein inflow water is output as product water via an actual reverse-osmosis water-treatment process; a process-modelling unit for running a reverse-osmosis simulation process which simulates the plant unit, using input values that are the same as input values corresponding to the inflow water flowing into the plant unit; and a Kalman filter unit which compares difference values between the actual output value output via the plant unit and the simulation output value output via the process modelling unit, and which applies the difference values to a Kalman filter algorithm and thereby performs real-time estimation of the filter resistance coefficient in the process modelling unit.

Description

Membrane Contamination Diagnosis System and Method for Water Treatment Process using Kalman Filter Algorithm

The present invention relates to a water treatment process, and more particularly, to a real-time membrane contamination diagnosis system and method in a seawater desalination process using the Kalman filter algorithm.

In general, a series of water treatment processes to remove high-purity drinking water, living water, and industrial water by removing dissolved substances, including salts, from seawater (sea water), which are difficult to use directly for domestic or industrial water, are called desalination or seawater desalination. The facilities used to produce seawater as freshwater are called seawater desalination plants or seawater desalination plants.

Desalination methods are largely classified according to their basic principles. The reverse osmosis method is used to heat seawater using a heat source and condense the generated steam to obtain fresh water and reverse osmosis to produce fresh water by passing the semi-permeable membrane through osmosis. It is a representative method of seawater desalination.

In existing real seawater desalination plants, operation and maintenance did not use optimized technology, but simply operated and maintained the system based on experience. However, as the size of the reverse osmosis desalination plant increases, the introduction of the automation concept is greatly required.

Currently, seawater desalination plants determine membrane fouling levels by reducing production flow rates or increasing inflow pressures to determine when to clean and replace membranes. However, since this method is dependent on the influent condition and the surrounding environment, it is difficult to determine the degree of contamination of the membrane itself.

Therefore, there is a problem that does not optimize maintenance, such as cleaning and replacing the membrane at an inappropriate time, resulting in unnecessary operating and maintenance costs.

The present invention has been made to solve the above problems, and an object thereof is to provide a real-time membrane fouling diagnosis system and method in the seawater desalination process using the Kalman filter algorithm.

The objects of the present invention are not limited to the above-mentioned objects, and other objects which are not mentioned will be clearly understood by those skilled in the art from the following description.

In order to achieve the above object, the membrane fouling diagnosis system of the water treatment process using the Kalman filter algorithm according to a preferred embodiment of the present invention, the plant unit for outputting the influent to the production water through the reverse osmosis membrane water treatment process, and the plant unit A process model unit for performing a reverse osmosis membrane simulation process that uses the same input value as the input value corresponding to the inflow water, and simulates the plant unit, a real output value output through the plant unit, and a simulation output value output through the process model unit. It may include a Kalman filter unit for comparing the difference value with each other and applying the difference value to the Kalman filter algorithm to estimate the film resistance coefficient in the process model unit in real time.

Here, the water treatment process may include a seawater desalination process.

In addition, the Kalman filter unit compares each output value consisting of the flow rate and concentration of the production water of the measured value output through the plant unit and the simulation value output through the process model unit, and if there is a difference between the Kalman filter The film resistance coefficient value is estimated in real time in the direction of minimizing this using an algorithm.

In addition, the membrane resistance coefficient estimated by the Kalman filter unit is input to the process model unit and updated, and a diagnosis unit estimates the updated membrane resistance coefficient to diagnose the degree of contamination of the membrane in real time.

In order to achieve the above object, the membrane contamination diagnosis method of the water treatment process using the Kalman filter algorithm according to a preferred embodiment of the present invention, (a) the input value corresponding to the influent flowing into the plant portion performing the actual reverse osmosis membrane water treatment process Inputting the same input value as the input to the process model part for performing the reverse osmosis membrane simulation process simulating the plant part, and (b) the actual output value output through the plant part from the Kalman filter part and the simulation output value output through the process model part. Comparing the difference values of each other and applying the difference values to the Kalman filter algorithm; and (c) the film resistance in the process model unit in a direction that the Kalman filter algorithm minimizes the difference between the measured output value and the simulated output value. Estimating the coefficients in real time, and (d) using the estimated film resistance coefficient. It may include the step of diagnosing the degree of real-time.

Here, in the step (b), the Kalman filter unit compares each output value composed of the flow rate and concentration of the production water of the measured value output through the plant unit and the simulation value output through the process model unit, and the If there is a difference value, the film resistance coefficient value is estimated in real time in the direction of minimizing it using the Kalman filter algorithm.

In addition, the film resistance coefficient estimated in the step (C) is characterized in that the input to the process model unit is updated.

In addition, the step (d) is characterized in that the degree of contamination of the membrane in real time diagnosis by estimating the updated membrane resistance coefficient.

Specific details of other embodiments are included in the detailed description and drawings.

According to the present invention constituted as described above, after constructing a process model that well simulates the reverse osmosis membrane seawater desalination plant, applying a Kalman filter algorithm to estimate the membrane resistance coefficient (membrane resistance) in the process model in real time In this way, the degree of membrane contamination can be effectively diagnosed, and it is expected that the membrane can be used for maintenance such as cleaning or replacing the membrane at an appropriate time. As a result, this can result in a reduction in operational maintenance costs.

That is, in the present invention, the Kalman filter algorithm is applied in the seawater desalination process to diagnose the membrane fouling degree in real time through the numerical membrane fouling index so that the driver can easily determine the membrane fouling degree in real time. Therefore, it is possible to reduce production costs through efficient operation and maintenance using optimized maintenance and management technology.

The effects of the present invention are not limited to the above-mentioned effects, and other effects not mentioned will be clearly understood by those skilled in the art from the description of the claims.

1 is a schematic configuration diagram of a membrane fouling diagnosis system of the seawater desalination process using the Kalman filter algorithm according to an embodiment of the present invention;

2 is a flowchart illustrating a method for estimating a membrane resistance coefficient in real time for diagnosing membrane fouling in a seawater desalination process using a Kalman filter algorithm according to an exemplary embodiment of the present invention.

Figure 3 is a graph comparing the experimental results estimated by using the mathematical Kalman filter the membrane fouling state with time in the present invention and the conventional seawater desalination process.

Advantages and features of the present invention, and a configuration and method for achieving them will be apparent with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only the embodiments are to make the disclosure of the present invention complete, and common knowledge in the art to which the present invention pertains. It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims. For reference, in describing the present invention, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted.

Prior to the description, the water treatment process described below exemplifies a seawater desalination process as an example, but it is revealed in advance that not only the seawater desalination process may include all water treatment processes such as groundwater and wastewater.

Hereinafter, a membrane fouling diagnosis system and method of a water treatment process using a Kalman filter algorithm according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic configuration diagram of a membrane fouling diagnosis system of the seawater desalination process using the Kalman filter algorithm according to a preferred embodiment of the present invention.

As shown in FIG. 1, the membrane contamination diagnosis system 1 of the seawater desalination process according to the preferred embodiment of the present invention includes a plant unit 10, a process model unit 20, a Kalman filter unit 30, and a diagnosis unit. 40 and the like.

Plant unit 10 refers to the actual seawater desalination plant to produce the production water (freshwater) by performing a seawater desalination water treatment process through the reverse osmosis membrane of the influent (sea water). Here, the input and output before and after the plant unit 10 means an inflow water input value and a production water output value. Since the plant part 10 can be understood by a known technique, detailed description thereof will be omitted.

The process model unit 20 performs a reverse osmosis simulation process as a process model well replicating the plant unit 10. The process model unit 20 uses the same input value as the input value corresponding to the inflow water flowing into the plant unit 10 as an input. That is, the input value of the plant section 10 is used as the input value of the process model section 20 at the same time.

The Kalman filter unit 30 refers to a Kalman filter algorithm for estimating the film resistance coefficient R _m .

The Kalman filter unit 30 compares the difference between the actual output value output through the plant unit 10 and the simulation output value output through the process model unit 20, and applies the difference value to the Kalman filter algorithm. The film resistance coefficient in the model unit 20 is estimated in real time. That is, the Kalman filter unit 30 compares each output value composed of the flow rate and concentration of the production water of the actual value output through the actual plant unit 10 and the simulation value output through the process model unit 20 with each other. If there is a difference, the Kalman filter algorithm estimates the value of the film resistance coefficient in real time. In addition, the film resistance coefficient estimated by the Kalman filter unit 30 is input to the process model unit 20 and updated.

The diagnosis unit 40 diagnoses the contamination information of the membrane in real time by estimating the updated membrane resistance coefficient. As membrane fouling increases with time, the number of production decreases under the same operating conditions and the membrane resistance coefficient increases. At this time, the production water is also affected by the temperature and concentration of the seawater, but the membrane resistance coefficient excludes all these effects and expresses the pollution phenomenon of the membrane itself.

Hereinafter, the Kalman filter algorithm applied to the membrane fouling diagnosis system of the seawater desalination process of the present invention will be described in detail.

The basic filter definition is to pass what you want and remove what you don't need. As such, the Kalman filter is an algorithm mainly used to remove noise in the data to obtain a desired signal or information. The basic concept of the Kalman filter is to estimate the optimal value representing the current state through recursive data processing based on the accumulated past and present data. That is, the Kalman filter predicts the state value to be estimated by minimizing possible errors by predicting the state of the linear system by mathematical method. In practice, since it is nonlinear and not Gaussian noise, the Extended Kalman filter (EFK) is used to linearize the nonlinear. These Kalman filters are mainly used in spacecraft, missiles, signal processing and stocks. In the present invention, the Kalman filter algorithm based on the model is applied to estimate the membrane contamination which cannot be measured in the reverse osmosis membrane desalination process.

First, prior to explaining the Kalman filter algorithm, the reverse osmosis membrane desalination process model used in the present invention is as follows. First, the production water flux produced through the membrane membrane is expressed as follows.

In the above formula, v, Δp, Δπ, R _m means the production water flux, trans-membrane pressure, osmotic pressure, membrane resistance (membrane resistance), respectively. And x and t represent the position and operating time in the membrane channel. Intra-channel membrane differential pressure is calculated by considering the influence of friction in the rectangular channel and the spacer in the membrane channel as shown below.

In the above formula, H, k _f , μ denote channel height, spacer friction coefficient and viscosity, respectively. In addition, in the case of reverse osmosis is calculated through the following formula configured by the experiment.

In the above equation, c _w means the concentration of the membrane surface and T _a means the absolute temperature of the influent. The crossflow velocity in the membrane channel is then calculated by the overall mass balance in the channel as shown below.

In the above equation, u means crossflow velocity. In addition, the average concentration (c) in the membrane channel is calculated based on the solute mass balance of the membrane channel, and the membrane surface concentration (c _w ) is calculated in consideration of concentration polarization.

In the above equation, B is the solute permeability coefficient, D is the diffusion coefficient, r_jIs the membrane removal rate. Finally, the following temperature correction coefficient (TCF) was considered to consider the temperature.

In the reverse osmosis membrane seawater desalination process model, membrane resistance (membrane resistance) is a characteristic value of the membrane, the production number will vary depending on the membrane resistance coefficient. The membrane resistance coefficient is independent of the temperature and concentration of the influent and independently represents the resistance of the membrane itself. Therefore, as the membrane fouling progresses with time, the membrane resistance coefficient increases. In this respect, membrane contamination can be diagnosed by estimating the real-time membrane resistance coefficient.

The main equation of the Kalman filter algorithm for estimating membrane fouling in the present invention is as follows.

The parameter representing the membrane fouling to be estimated in the above formula is

to be. this

Is the membrane resistance (R _m ) value in the reverse osmosis membrane seawater desalination process model. In the above algorithm, the value of the evaluation function (J) is minimized to estimate the optimal film resistance coefficient. In other words, the film resistance coefficient reflecting the current membrane fouling condition is estimated by finding a value that minimizes the mean square error. For this purpose, the process noise covariance (Q) and the measured noise covariance (R) are assumed to be Gaussian noise, and the algorithm progresses as the Kalman gain (K) and the estimated error covariance (P) in the Kalman filter algorithm are updated over time. . In this algorithm, f is the reverse osmosis membrane desalination process model mentioned above, the input values are seawater conditions and operating conditions, and the membrane resistance coefficient, and the output values are the concentration and flow rate of the produced water. The algorithm is to find the film resistance coefficient that minimizes the error in the concentration and flow rate of the produced water calculated by the simulation of the process model unit and the value of the flow rate and concentration of the produced water measured through the actual plant part. As a result, the membrane resistance coefficient can be modeled to represent the current state of the membrane itself regardless of other influent or operating conditions.

2 is a flowchart illustrating a method for estimating a membrane resistance coefficient in real time for diagnosing membrane contamination in a seawater desalination process using a Kalman filter algorithm according to an exemplary embodiment of the present invention.

As shown in FIG. 2, first, an input osmosis membrane desalination process model in which an input value identical to an input value corresponding to an inflow water flowing into the plant portion 10 performing the actual reverse osmosis membrane seawater desalination process simulates the plant portion 10. It is input to the unit 20.

Next, the difference between the actual output value output from the Kalman filter unit 30 through the plant unit 10 and the simulation output value output through the process model unit 20 is compared with each other, and the difference value is compared to the Kalman filter algorithm. Apply.

Next, the Kalman filter algorithm estimates the film resistance coefficient in the process model unit 20 in real time in order to minimize the difference between the measured output value and the simulated output value. Specifically, the Kalman filter unit 30 outputs each output value including the flow rate and concentration of the production water of the measured value output through the plant unit 10 and the simulation value output through the process model unit 20. Compare and, if there is a difference value, real-time estimation of the film resistance coefficient value in the direction of minimizing it using the Kalman filter algorithm, the estimated film resistance coefficient is input to the process model unit 20 is updated.

As a result, the estimated membrane resistance coefficient serves as a membrane fouling diagnostic indicator that informs the membrane of the current contamination status in real time.

In the graph shown in FIG. 3, the line ① represents the film resistance coefficient R _m , and the line ② represents the operating pressure. The operating pressure (line ②) in the conventional seawater desalination process is not independent of seasonal changes in seawater temperature and concentration, and shows a value reflecting this, whereas the membrane resistance coefficient in the seawater desalination process of the present invention (line ①). ) Is a value that excludes the effects of influent conditions in the model. Here, sometimes it is decreasing and increasing up and down, which may happen because all the parameter values are not the same instantaneous value because the data used is a daily average value, and it is increasing as a whole, but at some point flush Due to the phenomenon, the film resistance coefficient seems to decrease. If the actual operation data is estimated in real time, it is expected to show more accurate estimates.

Although the embodiments of the present invention have been described above with reference to the accompanying drawings, those skilled in the art to which the present invention pertains may be embodied in other specific forms without changing the technical spirit or essential features of the present invention. It will be understood that. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. The scope of the present invention is shown by the following claims rather than the detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present invention. do.

The present invention establishes a process model that well simulates a reverse osmosis membrane seawater desalination plant, and estimates the membrane resistance coefficient in real time through a Kalman filter algorithm, and provides a membrane contamination that informs the current contamination state of the membrane in real time. We propose a seawater desalination process to be used as a diagnostic indicator. The water treatment process using the Kalman filter algorithm can be widely used in various water treatment process industries such as wastewater, groundwater, as well as seawater desalination process.

Claims

Plant unit for outputting the influent to the production water through the actual reverse osmosis membrane water treatment process;

A process model unit which uses an input value corresponding to the inflow water flowing into the plant portion and performs a reverse osmosis membrane simulation process that simulates the plant portion; And

A Kalman filter comparing the difference between the actual output value output through the plant unit and the simulation output value output through the process model unit, and applying the difference value to the Kalman filter algorithm to estimate the film resistance coefficient in the process model unit in real time. Membrane contamination diagnosis system of the water treatment process using the Kalman filter algorithm.
The method of claim 1, wherein the Kalman filter unit compares each output value consisting of the flow rate and concentration of the production water of the measured value output through the plant unit and the simulation value output through the process model unit, the difference value is The membrane contamination diagnosis system of the water treatment process using the Kalman filter algorithm, characterized in that to estimate the value of the film resistance coefficient in real time in the direction of minimizing it using the Kalman filter algorithm.
2. The Kalman method of claim 1, wherein the film resistance coefficient estimated by the Kalman filter unit is input to the process model unit and updated, and a diagnosis unit estimates the updated membrane resistance coefficient in real time to diagnose the degree of contamination of the membrane. Membrane contamination diagnosis system of water treatment process using filter algorithm.
The membrane fouling diagnosis system of claim 1, wherein the water treatment process comprises a Kalman filter algorithm comprising a seawater desalination process.
(a) inputting an input value equal to an input value corresponding to an inflow water flowing into a plant portion performing the actual reverse osmosis membrane water treatment process, into a process model portion performing a reverse osmosis membrane simulation process simulating the plant portion;

(b) comparing the difference value between the actual output value output through the plant unit and the simulation output value output through the process model unit in the Kalman filter unit, and applying the difference value to the Kalman filter algorithm;

(c) the Kalman filter algorithm estimating a film resistance coefficient in the process model unit in a direction to minimize the difference between the measured output value and the simulated output value; And

(D) membrane fouling diagnostic method of the water treatment process using a Kalman filter algorithm comprising the step of real-time diagnosis of the membrane fouling degree through the estimated membrane resistance coefficient.
The method of claim 5, wherein the step (b) comprises: outputting each output value including a flow rate and a concentration of the production water of the measured value outputted through the plant unit and the simulation value outputted through the process model unit; And comparing and comparing the Kalman filter algorithm with a Kalman filter algorithm to estimate the membrane resistance coefficient value in real time.
7. The method of claim 6, wherein the film resistance coefficient estimated in step (C) is input to the process model unit and updated.
8. The method of claim 7, wherein the step (d) comprises diagnosing the contamination of the membrane in real time by estimating the updated membrane resistance coefficient.
The method of claim 5, wherein the water treatment process comprises a seawater desalination process.