Batch reverse osmosis desalination apparatus and method based on scaling prediction

Abstract

A batch reverse osmosis desalination device based on scaling occurrence prediction includes a scaling characteristic data collection unit that collects scaling characteristic data of influent water and concentrated water in a batch reverse osmosis process, a scaling index derivation unit that evaluates the possibility of scaling occurrence based on the scaling characteristic data and calculates a scaling index, a scaling model unit that inputs the scaling index as input data into a deep learning-based scaling prediction model to predict the time of scaling occurrence, and a flushing processing unit that adjusts the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and proceeds to inject a scaling inhibitor into a reverse osmosis membrane module through a batch reverse osmosis line.

Description

Technology Field

[0001] The present invention relates to batch reverse osmosis (Batch-RO) desalination technology, and more specifically, to a batch reverse osmosis desalination apparatus and method that provides a technology for predicting the occurrence of scaling by utilizing real-time sensor data in a batch reverse osmosis process, and optimizes flushing and the injection of inhibitors based on the scaling prediction results. Background Technology

[0003] Batch Reverse Osmosis (Batch-RO) processes play a crucial role in desalination technology and are attracting attention as a method capable of achieving high recovery rates through the recirculation of concentrated water. Reverse osmosis is a process that removes dissolved salts and impurities using semipermeable membranes and is utilized in various fields, including seawater desalination, industrial water treatment, and water purification systems. While batch-RO processes offer the advantage of reducing energy consumption compared to continuous-cycle processes, they present a problem in that the risk of scaling increases due to the high concentration of the concentrated water.

[0004] In conventional batch reverse osmosis processes, methods such as adding antiscalants at regular intervals or performing periodic flushing are applied to prevent scaling. However, the scaling occurrence cycle is inconsistent depending on changes in the operating environment, and simple cycle-based antiscalant addition and flushing methods are not optimized, leading to reduced operational efficiency. Furthermore, scaling prediction methods utilizing empirical models have limitations in effectively suppressing scaling because they struggle to reflect real-time fluctuations.

[0006] Korean Registered Patent No. 10-1633314 (June 20, 2016) provides a reverse osmosis desalination apparatus and method that prevents efficiency degradation in a reverse osmosis process with intermittent operation, reduces unnecessary energy consumption during repeated operation and shutdown, and enables variable and intermittent operation of the reverse osmosis process according to the inflow load or water demand during the osmotic desalination process.

[0007] The reverse osmosis desalination device includes an influent tank for storing influent water drawn from a water source, a high-pressure pump for pressurizing and supplying influent water to the downstream end, an influent water control valve for controlling the flow rate of influent water supplied by the high-pressure pump, a reverse osmosis membrane module for receiving influent water and filtering it through a reverse osmosis process to produce filtered water, an energy recovery device for recovering pressure from the concentrated water resulting from the influent water filtration treatment of the reverse osmosis membrane module and transmitting it to the influent water control valve to mix with the influent water, a concentrated water branch valve for branching the concentrated water from the reverse osmosis membrane module to the energy recovery device and the influent water control valve, and a reverse osmosis circulation pump for pumping influent water to the reverse osmosis membrane module. Prior art literature

[0009] Korean Registered Patent No. 10-1633314 (June 20, 2016) The problem to be solved

[0010] One embodiment of the present invention provides a technology for predicting the occurrence of scaling in a batch reverse osmosis process by utilizing real-time sensor data, and aims to provide a batch reverse osmosis desalination apparatus and method based on scaling occurrence prediction that can optimize flushing and the addition of an inhibitor based on the scaling prediction results.

[0011] One embodiment of the present invention aims to provide a batch reverse osmosis desalination device and method based on scaling occurrence prediction that can maximize energy efficiency through an energy recovery device that recovers the pressure of concentrated water generated in a reverse osmosis membrane module and mixes it with influent water.

[0012] One embodiment of the present invention aims to provide a batch reverse osmosis desalination apparatus and method based on scaling occurrence prediction, which can operate the reverse osmosis process variably and intermittently according to water demand and inflow load.

[0013] One embodiment of the present invention aims to provide a batch reverse osmosis desalination device and method based on scaling occurrence prediction, which can efficiently recycle energy by branching the flow of concentrated water and delivering it to an energy recovery device and an influent water control valve. means of solving the problem

[0015] Among the embodiments, the batch reverse osmosis desalination device based on scaling occurrence prediction includes a scaling characteristic data collection unit that collects scaling characteristic data of influent water and concentrated water in a batch reverse osmosis process; a scaling index derivation unit that evaluates the possibility of scaling occurrence based on the scaling characteristic data and calculates a scaling index; a scaling model unit that inputs the scaling index as input data to a deep learning-based scaling prediction model to predict the time of scaling occurrence; and a flushing processing unit that adjusts the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and proceeds to inject a scaling inhibitor into a reverse osmosis membrane module through a batch reverse osmosis line.

[0016] The scaling characteristic data collection unit can measure at least one physical or chemical process data among TDS (Total Dissolved Solids), pH, flow rate, pressure, and temperature as scaling characteristic data through a first sensor module connected to a storage tank storing the influent water in the batch reverse osmosis process and a second sensor module connected to a reverse osmosis membrane module storing the concentrated water. The scaling characteristic data collection unit can periodically measure the scaling characteristic data according to a measurement cycle, and the measurement cycle can be dynamically adjusted based on the rate of change in flow rate or TDS of the influent water or the concentrated water. The scaling characteristic data collection unit can predict a point in time when the probability of scaling occurrence exceeds a specific standard through the deep learning-based scaling prediction model, and can automatically adjust the measurement cycle or measurement event according to the predicted point in time.

[0017] The above-described scaling index derivation unit can generate normalized data by performing missing value correction and normalization on the above-described scaling characteristic data, and derive the above-described scaling index by calculating the Supersaturation Index (SI), Scaling Potential Index (SPI), Langelier Saturation Index (LSI), and Stiff & Davis Saturation Index (S&DSI) for the above-described normalized data. The above-described scaling index derivation unit can finally calculate the above-described scaling index by applying weights based on the importance of each of the above-described Supersaturation Index, the above-described Scaling Potential Index, the above-described Langelier Saturation Index, and the above-described Stiff & Davis Saturation Index.

[0018] The scaling model unit can predict the timing of scaling occurrence by implementing the scaling prediction model as a Long Short-Term Memory (LSTM) algorithm that learns the temporal change of the scaling index. The scaling model unit can perform learning optimization by adjusting the weights of the loss function based on a pattern in which the frequency of scaling occurrence decreases with the injection of the inhibitor.

[0019] The flushing treatment unit can determine the flushing flow rate, flushing duration, and flushing pressure intensity for the injection of the scaling inhibitor by adjusting the flushing conditions according to the time of scaling occurrence. The flushing treatment unit can monitor the injection of the scaling inhibitor in real time to dynamically adjust the amount of the scaling inhibitor injected, and can adjust the flushing pressure intensity to prevent excessive injection while maintaining the minimum effective concentration of the scaling inhibitor.

[0020] Among other embodiments, a batch reverse osmosis desalination method based on a prediction of scaling occurrence performed in a batch reverse osmosis desalination device based on a prediction of scaling occurrence comprises: a scaling characteristic data collection step for collecting scaling characteristic data of influent water and concentrated water in a batch reverse osmosis process; a scaling index derivation step for evaluating the possibility of scaling occurrence based on the scaling characteristic data and calculating a scaling index; a scaling model step for predicting the time of scaling occurrence by inputting the scaling index as input data into a deep learning-based scaling prediction model; and a flushing treatment step for adjusting the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and introducing a scaling inhibitor into a reverse osmosis membrane module through a batch reverse osmosis line.

[0021] Among other embodiments, a batch reverse osmosis desalination device capable of optimizing the amount of an anti-scaling agent injected in real time comprises: a scaling characteristic data collection unit that measures at least one of the TDS, pH, flow rate, pressure, and temperature of the influent and concentrated water in a batch reverse osmosis process as scaling characteristic data; a scaling index derivation unit that calculates a scaling index by calculating a supersaturation index, a scaling possibility index, a Langelier saturation index, and a Stiff & Davis saturation index for the scaling characteristic data; a scaling model unit that predicts the time of scaling occurrence through a scaling prediction model based on an LSTM (Long Short-Term Memory) algorithm that learns the temporal change of the scaling index; and a flushing processing unit that adjusts the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and injects an anti-scaling agent into a reverse osmosis membrane module through a batch reverse osmosis line. Effects of the invention

[0023] The disclosed technology may have the following effects. However, this does not mean that a specific embodiment must include all of the following effects or only the following effects; therefore, the scope of the rights of the disclosed technology should not be understood as being limited by this.

[0024] A batch reverse osmosis desalination apparatus and method based on scaling occurrence prediction according to one embodiment of the present invention can predict the occurrence of scaling by utilizing real-time sensor data, and maximize the efficiency of the reverse osmosis process by optimizing flushing and the injection of an inhibitor based on the predicted results.

[0025] A batch reverse osmosis desalination apparatus and method based on scaling occurrence prediction according to one embodiment of the present invention can maximize energy efficiency by recovering the pressure of the concentrated water generated in the reverse osmosis membrane module and combining it with the influent water.

[0026] A batch reverse osmosis desalination apparatus and method based on scaling occurrence prediction according to one embodiment of the present invention can minimize energy consumption by operating the reverse osmosis process variably and intermittently in accordance with water demand and inflow load.

[0027] A batch reverse osmosis desalination device and method based on scaling occurrence prediction according to one embodiment of the present invention can efficiently recycle energy by diverting the flow of concentrated water and supplying it to an energy recovery device and an influent water control valve. Brief explanation of the drawing

[0029] FIG. 1 is a drawing illustrating a batch reverse osmosis desalination system (100) based on scaling occurrence prediction according to one embodiment of the present invention. FIG. 2 is a diagram illustrating the system configuration of a batch reverse osmosis desalination device (130) based on the scaling occurrence prediction of FIG. 1. FIG. 3 illustrates a batch reverse osmosis desalination device (130) based on the scaling occurrence prediction of FIG. 1, and is a diagram showing the device configuration including a sensor unit for collecting data during batch desalination operation, a valve, a free piston, and an additional Batch-RO Flushing location. Figure 4 is a diagram illustrating the functional configuration of a batch reverse osmosis desalination device based on scaling occurrence prediction of Figure 1. Figure 5 is a flowchart illustrating the functional configuration of a batch reverse osmosis desalination device based on scaling occurrence prediction of Figure 1. Figure 6 is a diagram illustrating a change in the flushing method performed in the batch reverse osmosis process of the batch reverse osmosis desalination device based on scaling occurrence prediction of Figure 1. Figure 6a shows a conventional flushing step, and Figure 6b shows an improved flushing step utilizing a concentrated water circulation line. Specific details for implementing the invention

[0030] The description of the present invention is merely an example for structural or functional explanation, and therefore the scope of the present invention should not be interpreted as being limited by the examples described in the text. That is, since the examples are subject to various modifications and may take various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical concept. Furthermore, the objectives or effects presented in the present invention do not imply that a specific example must include all of them or only such effects; therefore, the scope of the present invention should not be understood as being limited by them.

[0031] Meanwhile, the meaning of the terms described in this application should be understood as follows.

[0032] Terms such as "first," "second," etc., are intended to distinguish one component from another, and the scope of rights shall not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0033] When it is stated that one component is "connected" to another component, it should be understood that it may be directly connected to that other component, or that there may be other components in between. Conversely, when it is stated that one component is "directly connected" to another component, it should be understood that there are no other components in between. Meanwhile, other expressions describing the relationships between components, such as "between" and "exactly between," or "adjacent to" and "directly adjacent to," should be interpreted in the same way.

[0034] A singular expression should be understood to include a plural expression unless the context clearly indicates otherwise, and terms such as "include" or "have" are intended to specify the existence of the implemented features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood not to preclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0035] In each step, identifiers (e.g., a, b, c, etc.) are used for convenience of explanation and do not describe the order of the steps; the steps may occur differently from the specified order unless a specific order is clearly indicated in the context. That is, the steps may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

[0036] The present invention may be implemented as computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. Additionally, the computer-readable recording medium may be distributed across networked computer systems, so that computer-readable code can be stored and executed in a distributed manner.

[0037] Unless otherwise defined, all terms used herein have the same meaning as generally understood by those skilled in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having meanings consistent with the context of the relevant technology and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined in this application.

[0039] FIG. 1 is a drawing illustrating a batch reverse osmosis desalination system (100) based on scaling occurrence prediction according to one embodiment of the present invention.

[0040] Referring to FIG. 1, a batch reverse osmosis desalination system (100) based on scaling occurrence prediction may include a user terminal (110), a batch reverse osmosis desalination device (130) based on scaling occurrence prediction, and a database (150).

[0042] The user terminal (110) may correspond to a computing device that is connected to a batch reverse osmosis desalination device (130) based on scaling occurrence prediction and can transmit and receive data or check analysis results in real time. To this end, the user terminal (110) can install a dedicated application or software to check the scaling occurrence prediction results and operation efficiency in real time and provide analysis results regarding the reverse osmosis desalination process to the user.

[0043] The user terminal (110) may be composed of a single or multiple terminals, and if composed of multiple terminals, it may include a first user terminal, a second user terminal, ..., the nth user terminal (n is a natural number). For example, the user terminal (110) may be implemented as a smartphone, laptop, or computer capable of operating in connection with the scaling occurrence prediction-based batch reverse osmosis desalination device (130), and may also be implemented as various devices such as a tablet PC. Additionally, the user terminal (110) may be connected to the scaling occurrence prediction-based batch reverse osmosis desalination device (130) via a wired network or a wireless network such as Bluetooth, Wi-Fi, or LTE.

[0044] The batch reverse osmosis desalination device (130) based on scaling occurrence prediction is a device for predicting scaling occurrence and optimizing the reverse osmosis process, and can analyze the results of the scaling occurrence prediction and perform process optimization in real time based on this. In addition, the batch reverse osmosis desalination device (130) based on scaling occurrence prediction can transmit and receive data with a user terminal (110) via a wired or wireless network or store and manage analysis results by linking with a database (150).

[0045] The database (150) may correspond to a storage device that stores various information required during the operation of the batch reverse osmosis desalination device (130) based on scaling occurrence prediction. For example, the database (150) stores scaling occurrence prediction results, operational efficiency records, and process performance data, and the scaling occurrence prediction results are used to evaluate the predicted scaling state, and the operational efficiency records are used as data for process optimization and performance tracking. In addition, the database (150) can provide data required for the batch reverse osmosis desalination device (130) based on scaling occurrence prediction by analyzing process performance or tracking patterns based on the stored data.

[0046] In FIG. 1, the database (150) is depicted as a separate device from the batch reverse osmosis desalination device (130) based on scaling occurrence prediction, but it is not necessarily limited thereto and can be implemented by being included in the batch reverse osmosis desalination device (130) based on scaling occurrence prediction. In addition, if implemented as a separate device, real-time data synchronization is possible through a network, and results can be provided to a user terminal (110).

[0048] FIG. 2 is a diagram illustrating the system configuration of a batch reverse osmosis desalination device (130) based on the scaling occurrence prediction of FIG. 1.

[0049] Referring to FIG. 2, a batch reverse osmosis desalination device (130) based on scaling occurrence prediction may include a processor (210), memory (230), user input / output unit (250), network input / output unit (270), and communication port unit (290).

[0051] The processor (210) processes data for predicting scaling occurrence and optimizing the reverse osmosis process, and controls the operation of the reverse osmosis desalination process. The processor (210) is linked with the memory (230) to manage data related to the results of scaling occurrence prediction, operational efficiency analysis, and process optimization, and can perform control and adjustment of the reverse osmosis desalination device (130). The processor (210) controls the overall operation of the batch reverse osmosis desalination device (130) based on scaling occurrence prediction, and controls the data flow by being electrically connected to the memory (230), user input / output unit (250), and network input / output unit (270). The processor (210) can be implemented as a central processing unit (CPU) or a graphics processing unit (GPU).

[0052] The memory (230) may correspond to a storage device that stores data and commands required during the operation of the batch reverse osmosis desalination device (130) based on scaling occurrence prediction. For example, the memory (230) may store scaling occurrence prediction results, reverse osmosis process performance data, and operational efficiency records, and may execute scaling prediction analysis algorithms and process optimization algorithms. The memory (230) may be implemented as non-volatile memory (e.g., SSD or HDD) and volatile memory (e.g., RAM).

[0053] The user input / output unit (250) may correspond to a device for receiving input from a user and outputting scaling prediction results and process performance evaluation results. For example, the user input / output unit (250) includes input and output devices such as a touchscreen, monitor, and keyboard, and may provide process performance evaluation results to the user in a visual or auditory manner. The user input / output unit (250) may support a remote input / output environment when the scaling occurrence prediction-based batch reverse osmosis desalination device (130) operates as an independent server or is connected to a user terminal (110) via remote access.

[0054] The network input / output unit (270) provides a communication environment for connecting the batch reverse osmosis desalination device (130) based on scaling occurrence prediction to a user terminal (110) and a database (150). The network input / output unit (270) supports wired or wireless networks and may include Wi-Fi, Bluetooth, and wireless communication functions of 4G or higher. In addition, it supports network connections such as LAN (Local Area Network) and WAN (Wide Area Network) to perform data transmission and reception.

[0055] The communication port section (290) provides a port mapping function that performs data routing during the process of transmitting and receiving data through a network. For example, the communication port section (290) assigns a unique source port to the user terminal (110) to distinguish communication sessions between the user terminal (110) and the batch reverse osmosis desalination device (130) based on scaling occurrence prediction and prevent data collisions.

[0057] FIG. 3 illustrates a batch reverse osmosis desalination device (130) based on the scaling occurrence prediction of FIG. 1, and is a diagram showing the device configuration including a sensor unit for collecting data during batch desalination operation, a valve, a free piston, and an additional Batch-RO Flushing location.

[0058] Referring to FIG. 3, a batch reverse osmosis desalination device (100) based on scaling occurrence prediction may include a pretreatment unit (310), a sensor unit (320), a high-pressure pump (330), a pressure exchange vessel (340), a free piston (350), a reverse osmosis membrane (360), a pressure gauge (370), a filtered water tank (380), a concentrated water tank (390), and a Batch-RO Flushing system (a).

[0060] The pretreatment unit (310) is responsible for the initial purification and removal of impurities from the influent water. The influent water is connected to the sensor unit (320) for physical and chemical treatment, and major components that can cause scaling are removed. The pretreated influent water is transferred to the high-pressure pump (330) and flows into the pressure exchange vessel (340).

[0061] The sensor unit (320) is a sensor system that monitors the operating status of all components in a batch reverse osmosis desalination device (100) based on scaling occurrence prediction in real time. The sensor unit (320) measures the TDS, pH, concentration, and temperature of the raw water of the influent water and optimizes the operation of the high-pressure pump (330) and the reverse osmosis membrane (360). The sensor unit (320) plays an important role in predicting scaling occurrence and optimizing the process through real-time data.

[0062] The high-pressure pump (330) is a device that supplies the influent water treated in the pretreatment unit (310) to the pressure exchange vessel (340) at high pressure. The pressure of the high-pressure pump (330) is adjusted based on data provided by the sensor unit (320) and maintains an optimal pressure so that it can flow into the reverse osmosis membrane (360) through the pressure exchange vessel (340).

[0063] The pressure exchange vessel (340) is a device that recovers the pressure of the concentrated water and mixes it with the influent water. After recovering the pressure, the concentrated water is combined with the influent water and supplied appropriately to the reverse osmosis membrane (360). The process of recovering the pressure of the concentrated water in the pressure exchange vessel (340), combining it with the influent water, and supplying it to the reverse osmosis membrane (360) plays an important role in maximizing energy efficiency and supporting the prediction of scaling occurrence and process optimization.

[0064] The free piston (350) serves to convert and regulate the pressure of the concentrate within the pressure exchange vessel (340). The free piston maximizes energy efficiency and optimizes the mixing of the influent and concentrate. Additionally, the free piston plays a crucial role in the cleaning process of the Batch-RO Flushing system, contributing to the prevention of scaling and the improvement of process efficiency.

[0065] The reverse osmosis membrane (360) is a key device for removing salt and impurities from the influent water. The reverse osmosis process takes place in the reverse osmosis membrane (360), and filtered water is produced. The filtered water is transferred to a filtered water tank (380) and stored. The Batch-RO Flushing system (a) washes the reverse osmosis membrane (360) and prevents scaling.

[0066] The pressure gauge (370) is a device that measures the pressure of the high-pressure pump (330) and the reverse osmosis membrane (360) and monitors the process status in real time. The measured pressure serves as an important variable for predicting the occurrence of scaling and is used to optimize the process and adjust energy efficiency.

[0067] The filtered water tank (380) is a tank that stores filtered water generated from the reverse osmosis membrane (360) and supplies filtered water according to water demand. The filtered water tank (380) stores the filtered water generated from the reverse osmosis membrane (360) for a certain period of time and serves to maintain the stability of the system.

[0068] The concentrated water tank (390) is a tank that stores the concentrated water generated from the reverse osmosis membrane (360). The concentrated water is transferred to a pressure exchange vessel (340) and used in the process of energy recovery and process efficiency improvement. The concentrated water is subsequently reused or treated to contribute to preventing scaling and optimizing energy efficiency.

[0069] The Batch-RO Flushing System (a) plays an important role in cleaning the reverse osmosis membrane (360), preventing scaling, and optimizing the process. The Batch-RO Flushing System (a) uses concentrated water to maintain the long-term efficiency of the reverse osmosis process and prevents scaling.

[0071] The influent water undergoes initial purification through a pretreatment unit (310), and impurities and major components that can cause scaling contained in the influent water are removed. The pretreated influent water is transferred to a pressure exchange vessel (340) via a high-pressure pump (330) based on real-time status detection by a sensor unit (320). The influent water transferred to the pressure exchange vessel (340) is supplied to a reverse osmosis membrane (360) to start the reverse osmosis process. In the reverse osmosis membrane (360), the influent water is separated into filtered water and concentrated water. The filtered water is transferred to a filtered water tank (380) and stored. The concentrated water is returned to the pressure exchange vessel (340), and the sensor unit (320) detects the status of the concentrated water in real time to provide feedback for process optimization.

[0072] The concentrated water is delivered to the high-pressure pump (330) and flows into the pressure exchange vessel (340) according to the pressure measured by the pressure gauge (370). The free piston (350) converts and regulates the pressure of the concentrated water within the pressure exchange vessel (340), then mixes it with the influent water to increase the energy efficiency of the process. The Batch-RO Flushing System (a) periodically cleans the reverse osmosis membrane (360) to support the prevention of scaling and the optimization of process efficiency. The sensor unit (320) measures the chemical characteristics, temperature, pH, ion concentration, etc. of the influent water in real time and provides important data for predicting the occurrence of scaling and optimizing the process. In addition, the pressure gauge (370) monitors the pressure of the high-pressure pump (330) and the reverse osmosis membrane (360) to provide the timing for adding the anti-scaling agent and accurate judgment criteria.

[0074] Figure 4 is a diagram illustrating the functional configuration of a batch reverse osmosis desalination device based on scaling occurrence prediction of Figure 1.

[0075] Referring to FIG. 4, a batch reverse osmosis desalination device (130) based on scaling occurrence prediction may include a scaling characteristic data collection unit (410), a scaling index derivation unit (420), a scaling model unit (430), a flushing processing unit (440), and a control unit (450).

[0076] The batch reverse osmosis desalination device (130) based on scaling occurrence prediction does not necessarily have to include all of the above functional components simultaneously; depending on each embodiment, some of the components may be omitted, or some or all of the components may be selectively included. Additionally, the batch reverse osmosis desalination device (130) based on scaling occurrence prediction may be implemented as an independent module that selectively includes some of the above components, and may perform the scaling occurrence prediction-based batch reverse osmosis desalination function according to the present invention through interoperability between the modules. The operation of each component is described in detail below.

[0078] The scaling characteristic data collection unit (410) can perform the role of collecting scaling characteristic data of the influent water and concentrate in a batch reverse osmosis process. The scaling characteristic data collection unit (410) can provide data necessary to analyze the possibility of scaling occurring within the batch reverse osmosis process and to establish a control strategy to prevent it. The scaling characteristic data collection unit (410) includes a first sensor module (411) and a second sensor module (412), and can measure various physical and chemical process data by detecting the state of the influent water and concentrate in real time.

[0079] The first sensor module (411) is connected to a storage tank that stores influent water and can measure the TDS (Total Dissolved Solids), pH, flow rate, pressure, and temperature of the influent water. The second sensor module (412) is connected to a reverse osmosis membrane module that stores concentrated water and can detect the TDS, pH, flow rate, pressure, and temperature of the concentrated water.

[0080] The scaling characteristic data collection unit (410) can analyze scaling characteristics based on data measured by the first sensor module (411) and the second sensor module (412). The measurement data of the influent water and the concentrated water constitute scaling characteristic data and can be transmitted to the scaling index derivation unit (420) and the scaling model unit (430). The transmitted data can be used as basic data for calculating the scaling index and can be used as input values ​​for real-time process optimization and scaling prediction.

[0081] The scaling characteristic data collection unit (410) can measure the TDS concentration of the influent water before the batch reverse osmosis process begins. The measured TDS concentration can be compared with a reference value to analyze whether it exceeds it, and if the TDS concentration of the influent water is detected to be above a specific standard, additional preliminary treatment can be performed to minimize the risk of scaling.

[0082] The scaling characteristic data collection unit (410) can measure the pH value of the concentrated water in real time, and if the measured value deviates from a set range, it can determine that there is a high probability of scaling occurring. For example, if the pH value of the concentrated water decreases rapidly to 7.0 or lower, the scaling characteristic data collection unit (410) records the data and can provide the data to the scaling index derivation unit (420) so that additional analysis can be performed.

[0083] The scaling characteristic data collection unit (410) can periodically measure data according to the measurement cycle, and the measurement cycle can be dynamically adjusted based on the rate of change in flow rate and the rate of change in TDS of the influent or concentrated water.

[0084] If the flow rate of the influent water deviates from a certain range, the scaling characteristic data collection unit (410) can obtain more detailed data by shortening the measurement cycle. Conversely, if the rate of change of the flow rate of the influent water and concentrated water remains low, the scaling characteristic data collection unit (410) can reduce unnecessary data collection by extending the measurement cycle.

[0085] The scaling characteristic data collection unit (410) can measure data at 10-minute intervals under normal conditions. However, if the TDS concentration of the influent changes rapidly or the flow rate of the concentrated water increases, the measurement cycle may be changed to 1-minute intervals. Conversely, if no change in the process is detected for more than 1 hour, the scaling characteristic data collection unit (410) can adjust the measurement cycle to 20-minute intervals to reduce the system computational burden.

[0086] The scaling characteristic data collection unit (410) can automatically adjust the measurement cycle or measurement event when the probability of scaling occurring is predicted to be above a certain standard by utilizing a deep learning-based scaling prediction model.

[0087] The scaling characteristic data collection unit (410) collects scaling data in real time, and the scaling model unit (430) can analyze the collected data to detect a pattern in which the probability of scaling occurring increases. If it is predicted that the probability of scaling occurring increases at a specific point in time, the scaling characteristic data collection unit (410) can perform additional data collection by adjusting the measurement cycle.

[0088] The scaling characteristic data collection unit (410) can analyze patterns in which the likelihood of scaling occurring increases under specific conditions based on past data. For example, if the TDS growth rate rises sharply in a specific temperature range, the scaling characteristic data collection unit (410) can be set to collect data at shorter intervals than the existing measurement cycle. Conversely, in sections where the likelihood of scaling is low, the number of measurement events can be reduced to decrease the overall system load.

[0090] The scaling index derivation unit (420) can evaluate the possibility of scaling occurring based on the scaling characteristic data provided by the scaling characteristic data collection unit (410) and calculate a scaling index. The scaling index derivation unit (420) can be used to predict the occurrence of scaling in advance in a batch reverse osmosis process to establish control strategies for process optimization and scaling prevention. The scaling index derivation unit (420) can evaluate the possibility of scaling occurring in real time by performing quantitative analysis and can determine scaling prevention measures based on the evaluated data.

[0091] The scaling index derivation unit (420) can correct missing values ​​in the scaling characteristic data and normalize the data to increase the reliability of the analysis. The normalized data can be used to calculate various scaling-related indicators such as the Supersaturation Index (SI), Scaling Potential Index (SPI), Langelier Saturation Index (LSI), and Stiff & Davis Saturation Index (S&DSI).

[0092] The supersaturation index (SI) indicates whether a specific ionic component is in a supersaturated state in water; if the value is positive, there is a high possibility of scaling, and if it is negative, the possibility of corrosion may increase. The supersaturation index (SI) is defined as shown in Equation 1 below.

[0093] [Mathematical Formula 1]

[0094]

[0095] Here, IAP represents the actual solubility product (Ion Activity Product), and K sp represents the equilibrium solubility product constant.

[0096] If the supersaturation index (SI) is greater than 0, it means that the water is likely to form scale, and if it is less than 0, it indicates that the water is closer to a corrosive environment.

[0097] The Scalability Index (SPI) is an index that quantitatively evaluates the likelihood of scaling formation occurring under specific process conditions, and if the SPI value exceeds a certain standard, it can be determined that there is a high likelihood of accelerated scaling. The Scalability Index (SPI) is defined as shown in Equation 2 below.

[0098] [Mathematical Formula 2]

[0099]

[0100] Here, f is temperature (T), pH, calcium ion concentration [Ca 2+ ]), bicarbonate ion concentration ([HCO3 - It is a correction function that takes into account ]). The higher the Scalability Index (SPI) value, the greater the likelihood of scaling formation.

[0101] The Langelier Saturation Index (LSI) is an index that evaluates whether water has a tendency toward sedimentation or corrosion based on the chemical composition and temperature of the influent water, and if it exceeds a certain range, the likelihood of scaling occurring may increase. The Langelier Saturation Index (LSI) is defined as shown in Equation 3 below.

[0102] [Mathematical Formula 3]

[0103]

[0104] Here, pH represents the measured pH value of the water quality, and pH s represents the pH value at the saturated state of water.

[0105] If the LSI value is greater than 0, there is a possibility of scaling, and if it is less than 0, there is a possibility of corrosion.

[0106] The Stiff & Davis Saturation Index (S&DSI) is an index that evaluates the likelihood of scaling occurring in environments with high salt concentrations and can be utilized in high-concentration ionic environments such as seawater desalination processes. The Stiff & Davis Saturation Index (S&DSI) is defined as shown in Equation 4 below.

[0107] [Mathematical Formula 4]

[0108]

[0109] Here, LSI stands for Langelier Saturation Index, an index used to evaluate the scaling or corrosive tendency of water. I represents Ionic Strength, indicating the overall strength of ions within an aqueous solution. ] is the sulfate ion concentration and can be expressed in mg / L or mol / L units. A, B, and C are experimentally determined coefficients and are used to more precisely evaluate the possibility of scaling occurring in desalination systems.

[0110] The scaling index derivation unit (420) can evaluate the possibility of scaling by calculating the supersaturation index (SI), scaling possibility index (SPI), Langelier saturation index (LSI), and Stiff & Davis saturation index (S&DSI) based on data collected in real time from a batch reverse osmosis process.

[0111] For example, the scaling index derivation unit (420) can measure the TDS concentration of the influent water as 1500 mg / L, the pH value as 7.8, and the temperature as 30°C, and detect the TDS concentration of the concentrated water as 4500 mg / L and the sulfate ion concentration as 250 mg / L. The scaling index derivation unit (420) can calculate individual scaling indices using real-time measurement data.

[0112] When the scaling index derivation unit (420) calculates the supersaturation index (SI) and the value is evaluated to be 2.3, it means that the ionic components of the influent water have reached a supersaturated state, and there is a possibility that the formation of scaling will be accelerated. In addition, the scaling index derivation unit (420) can analyze the permeate flow rate (LMH, Litre per Square Meter per Hour) to evaluate the impact of scaling on the performance degradation of the membrane. For example, if the permeate flow rate shows a tendency to decrease from an initial 30 LMH to 24 LMH, it indicates that scaling is occurring gradually, and the process operator needs to consider immediate countermeasures.

[0113] Additionally, when the scaling index derivation unit (420) calculates the scaling possibility index (SPI) and the value is 85, it indicates that there is a high probability that scaling formation will proceed rapidly. The scaling index derivation unit (420) can continuously monitor changes in permeate flow rate and determine that the scaling progress rate is accelerating if it decreases below a specific threshold value. The scaling index derivation unit (420) can analyze whether it is necessary to shorten the existing flushing cycle and perform additional cleaning work in conjunction with the flushing processing unit (440). For example, if the permeate flow rate decreases from 30 LMH to 20 LMH or less, the flushing processing unit (440) can increase the flushing pressure or extend the flushing duration to maximize the scaling removal effect.

[0114] When the scaling index derivation unit (420) calculates the Langelier saturation index (LSI) and the value is evaluated as 1.7, it means that the water has a tendency to precipitate and there is a high possibility of scaling occurring. The scaling index derivation unit (420) can prevent the formation of scaling by transmitting a signal to the control unit (450) to adjust the pH value of the influent water, or by adding a weakly acidic solution if necessary to adjust the LSI value to 1.0 or less.

[0115] When the scaling index derivation unit (420) calculates the Stiff & Davis saturation index (S&DSI) and the value is evaluated as 2.5, it indicates that there is a very high possibility of scaling occurring in a high salt concentration environment. The scaling index derivation unit (420) can work in conjunction with the flushing treatment unit (440) to shorten the flushing cycle from the existing 60 minutes to 30 minutes and adjust the cleaning pressure to maximize the scaling inhibition effect.

[0116] The scaling index derivation unit (420) can calculate a comprehensive scaling index by integrating individual scaling indices and can optimize by applying weights that reflect the importance of each index. The final scaling index can be defined as shown in Equation 5 below.

[0117] [Mathematical Formula 5]

[0118]

[0119] Here, SI final w1 represents the final calculated scaling index. w1, w2, w3, and w4 represent weights for the Supersaturation Index (SI), Scalability Index (SPI), Langelier Saturation Index (LSI), and Stiff & Davis Saturation Index (S&DSI), and can be dynamically adjusted according to environmental conditions and process variables.

[0120] For example, the scaling index derivation unit (420) can more accurately reflect the risk of scaling due to high salt concentration by increasing the weight w4 of the Stiff & Davis saturation index (S&DSI) when the TDS concentration of the influent water remains high at 35,000 mg / L or higher during the process of evaluating the scaling index in a specific seawater desalination process.

[0121] The scaling index derivation unit (420) is used in a desalination process using groundwater as raw water where the water temperature is 15 If maintained below, the weight w3 of the Langelier saturation index (LSI) can be set relatively high. The scaling index derivation unit (420) can increase the weight of the Langelier saturation index (LSI) to precisely evaluate the possibility of water precipitation in a low-temperature environment, and can apply preliminary measures using a pH control device according to the evaluation results.

[0122] The scaling index derivation unit (420) can generate a warning signal to allow the process operator to take immediate action by adjusting the weight in real time when the scaling possibility index (SPI) exceeds a specific reference value. For example, if the scaling possibility index (SPI) rises above 80, there is a high probability that scale will form on the reverse osmosis membrane, so the scaling index derivation unit (420) can automatically adjust the process conditions by shortening the flushing cycle.

[0124] The scaling model unit (430) can predict the timing of scaling occurrence in a batch reverse osmosis process by utilizing the scaling index provided by the scaling index derivation unit (420) as input data. The scaling model unit (430) can support the evaluation of the possibility of scaling formation proceeding under specific operating conditions in advance and the implementation of countermeasures to prevent scaling by applying a deep learning-based scaling prediction model.

[0125] The scaling model unit (430) can implement the scaling prediction model using a Long Short-Term Memory (LSTM) algorithm that learns the temporal change of the scaling index. The LSTM algorithm has characteristics that are advantageous for learning the long-term dependency of time series data, and can more accurately predict the time of future scaling occurrence by considering the fluctuation pattern of the scaling index and the permeate flow rate per membrane area (LMH) together.

[0126] For example, the scaling model unit (430) can predict the likelihood of scaling occurring in real time by reflecting variables such as changes in chemical composition of the influent water (pH, TDS (Total Dissolved Solids), temperature), operating conditions of the batch reverse osmosis process, and the amount of scaling inhibitor added. Additionally, the final scaling index and permeate flow rate per membrane area (LMH) that change over time can be considered as input variables for an LSTM model for predicting the occurrence of scaling. In a specific batch reverse osmosis process, the scaling model unit (430) can utilize data on changes in the scaling index and permeate flow rate per membrane area (LMH) measured over the past 30 days as training data. If the scaling model unit (430) evaluates the likelihood of scaling occurring under specific conditions to be 80% or higher within the next 24 hours based on the learned pattern, it can provide the time when scaling is expected to occur to the process operator as a warning signal to support the performance of preventive measures.

[0127] The scaling model unit (430) can analyze the change in the frequency of scaling after the injection of a scaling inhibitor or flushing, and reflect the analyzed pattern in the model learning optimization. The scaling model unit (430) can learn the effect of the change in permeate flow rate per membrane area (LMH) after the injection of a scaling inhibitor on the reduction in the frequency of scaling, and reflect this to adjust the weights of the loss function of the LSTM model.

[0128] For example, if it is confirmed that in a specific reverse osmosis process, the frequency of scaling decreases by more than 50% whenever the amount of anti-scaling agent added increases by 10%, the scaling model unit (430) can learn the pattern of change in the frequency of scaling and adjust the loss function weights of the scaling prediction model. The scaling model unit (430) can dynamically adjust the sensitivity of the scaling prediction by reflecting the correlation between the increase in the amount of anti-scaling agent added and the decrease in the scaling rate. In addition, the scaling model unit (430) can more precisely predict the time when scaling is likely to occur by adjusting the loss function weights and minimize the occurrence of unnecessary scaling warnings. As the loss function weights are optimized, the scaling model unit (430) can more effectively identify the time when scaling is likely to occur and support the decision-making of the process operator.

[0129] The scaling model unit (430) can support the stable operation of a batch reverse osmosis process by performing real-time scaling occurrence prediction. The scaling model unit (430) can provide information to enable the process operator to take immediate response measures based on the scaling occurrence prediction results.

[0130] For example, the scaling model unit (430) can provide a warning signal to the process operator if the probability of scaling occurring within the next 6 hours is evaluated to be 85% or higher through an LSTM-based prediction model. Additionally, if the scaling model unit (430) determines that adjustment of the flushing cycle is necessary as a measure to prevent scaling, it can transmit the relevant information to the control unit (450) to support changes in process conditions.

[0131] The scaling model unit (430) can optimize the timing and amount of the inhibitor injection by analyzing real-time data. For example, if a pattern is observed in which the frequency of scaling decreases after the injection of the scaling inhibitor in a specific process, the scaling model unit (430) can calculate the optimal amount of the inhibitor injection by learning the correlation between the amount of the inhibitor injection and the scaling reduction rate.

[0132] The scaling model unit (430) can support the process operator in identifying the possibility of scaling in advance and optimizing preventive measures through real-time scaling occurrence prediction. Additionally, the scaling model unit (430) can learn the effect of adding a scaling inhibitor through loss function weight adjustment and perform model optimization to perform scaling occurrence prediction more precisely.

[0133] The scaling model unit (430) can improve the accuracy of scaling prediction, thereby supporting the process operator to more precisely adjust the timing and amount of the injection of the scaling inhibitor. Optimized injection of the inhibitor using the scaling model unit (430) can contribute to improving economic efficiency by increasing the operational efficiency of the batch reverse osmosis process and reducing the use of unnecessary inhibitors.

[0135] The flushing treatment unit (440) can adjust flushing conditions based on the timing of scaling occurrence in the batch reverse osmosis process and inject a scaling inhibitor into the reverse osmosis membrane module through the batch reverse osmosis line (a). The flushing treatment unit (440) can analyze the scaling index in real time and, when specific scaling occurrence conditions are met, dynamically adjust the flushing flow rate, flushing duration, and flushing pressure intensity to maintain an optimal scaling prevention effect.

[0136] The flushing treatment unit (440) can apply an optimal operation method to prevent scaling of the reverse osmosis membrane module by adjusting the flushing conditions according to the timing of scaling occurrence. When the likelihood of scaling occurrence is predicted to be above a certain threshold, the flushing treatment unit (440) can provide a more active cleaning effect by adjusting the existing flushing flow rate and duration.

[0137] For example, in a specific batch reverse osmosis process, if the scaling index derivation unit (420) predicts a probability of scaling occurring of 90% or more, the flushing treatment unit (440) can adjust the existing set flushing duration from 60 seconds to 120 seconds and increase the flushing pressure by 10% compared to the existing one to effectively remove sediment inside the reverse osmosis membrane.

[0138] Additionally, if the scaling index is maintained below a certain standard, the flushing processing unit (440) may extend the flushing interval to reduce unnecessary flushing. For example, if the predicted scaling occurrence value is lowered to 50% or less, the flushing interval may be adjusted from the existing 30 minutes to 45 minutes, and the flushing pressure may be reduced by 5% to optimize energy consumption.

[0139] The flushing treatment unit (440) monitors the injection of the scaling inhibitor in real time and can dynamically adjust the amount of inhibitor injected according to process conditions. If the concentration of the inhibitor drops below a certain level, the scaling prevention effect may decrease, and conversely, excessive injection of the inhibitor may lead to increased operating costs and environmental burden. The flushing treatment unit (440) can adjust the flushing pressure intensity to prevent excessive injection while maintaining the minimum effective concentration of the scaling inhibitor.

[0140] For example, if flushing is performed with the initial amount of the anti-scaling agent set to 1.5 mg / L but the rate of scaling occurrence remains high, the flushing treatment unit (440) can automatically adjust the anti-scaling agent concentration to 2.0 mg / L and increase the flushing pressure by 10% compared to the existing level to maximize membrane cleaning efficiency. On the other hand, if the predicted value of scaling occurrence decreases and the anti-scaling effect is maintained, the flushing treatment unit (440) can reduce the amount of the anti-scaling agent to 1.2 mg / L and extend the flushing cycle by 15 minutes compared to the existing level to minimize the use of unnecessary chemicals.

[0141] The flushing treatment unit (440) can adjust the flushing conditions and the amount of inhibitor injected in real time to maintain the performance of the reverse osmosis membrane module for a long period and improve the energy efficiency and economic feasibility of the process. The flushing treatment unit (440) can analyze the scaling occurrence pattern to continuously update the optimal flushing strategy and apply an adaptive operation method based on real-time data.

[0142] For example, if the scaling rate does not increase even when the flushing cycle is extended from 30 minutes to 45 minutes in a specific process, the flushing treatment unit (440) can subsequently extend the flushing cycle to 60 minutes to reduce operating costs. On the other hand, if the frequency of scaling increases faster than expected in a specific environment, the flushing treatment unit (440) can immediately adjust the flushing flow rate and duration to prevent contamination of the reverse osmosis membrane module.

[0143] The flushing treatment unit (440) optimizes the flushing strategy based on real-time data analysis of the batch reverse osmosis process and adjusts the flushing conditions to minimize scaling and reduce maintenance costs of the reverse osmosis membrane.

[0145] The control unit (450) controls the overall operation of the batch reverse osmosis desalination device (130) based on the prediction of scaling occurrence and can manage the control flow or data flow between the scaling characteristic data collection unit (410), the scaling index derivation unit (420), the scaling model unit (430), and the flushing processing unit (440).

[0147] Figure 5 is a flowchart illustrating the functional configuration of a batch reverse osmosis desalination device based on scaling occurrence prediction of Figure 1.

[0148] Figure 5 consists of a scaling characteristic data collection step (S510), a scaling index derivation step (S530), a scaling model step (S550), and a flushing processing step (S570).

[0150] The scaling characteristic data collection step (S510) is a process of measuring and collecting scaling characteristic data of the influent water and concentrate in a batch reverse osmosis process. The scaling characteristic data collection unit (410) includes a first sensor module (411) and a second sensor module (412), and can measure physical and chemical data such as TDS (Total Dissolved Solids), pH, flow rate, pressure, and temperature of the influent water and concentrate.

[0151] For example, the scaling characteristic data collection unit (410) can store and analyze the data in real time when the TDS of the influent is measured to be 1,500 mg / L and the TDS of the concentrated water is measured to be 3,500 mg / L. In addition, the scaling characteristic data collection unit (410) can maintain the pH of the influent at 7.8 or higher and the temperature of the concentrated water at 45 If it exceeds, it is determined that the environment increases the likelihood of scaling, and the measurement cycle can be shortened.

[0152] The scaling index derivation step (S530) is a process of evaluating the possibility of scaling occurring based on the data provided in the scaling characteristic data collection step (S510) and calculating the final scaling index. The scaling index derivation unit (420) can derive the final scaling index by normalizing the scaling characteristic data and calculating the oversaturation index (SI), scaling possibility index (SPI), Langelier saturation index (LSI), and Stiff & Davis saturation index (S&DSI).

[0153] For example, when the scaling index derivation unit (420) measures the pH of the influent water as 7.5, the TDS as 1,800 mg / L, and the TDS of the concentrated water as 4,000 mg / L, the supersaturation index (SI) can be calculated as 2.1. If the scaling index derivation unit (420) derives a supersaturation index of 2.1 or higher, it means that the water has reached a supersaturated state and can be evaluated as having a high probability of scaling occurring. In addition, when the scaling index derivation unit (420) calculates the scaling probability index (SPI), if the SPI value is calculated as 80 or higher, it can be determined that a more active response is required during the flushing treatment stage because there is a possibility that scaling will proceed rapidly.

[0154] The scaling model step (S550) is a process of predicting the timing of scaling occurrence by utilizing the scaling index and permeate flow rate per membrane area (LMH) calculated in the scaling index derivation step (S530) as input data. The scaling model unit (430) can apply a deep learning-based scaling prediction model to analyze the fluctuation patterns of the scaling index and permeate flow rate per membrane area (LMH) that change over time, and determine the possibility of scaling formation proceeding. In addition, the scaling model unit (430) can learn the temporal change of the scaling index by utilizing a Long Short-Term Memory (LSTM) algorithm and simultaneously reflect the change in permeate flow rate per membrane area (LMH) to predict the timing of scaling occurrence more precisely.

[0155] For example, the scaling model unit (430) can learn the scaling index measured over the past 30 days and predict that there is an 85% or higher probability of scaling occurring within the next 24 hours. The scaling model unit (430) can provide a warning signal to the process operator and send a notification so that a scaling inhibitor can be injected in advance. In addition, if a pattern is detected in which the probability of scaling occurring increases rapidly at a specific point in time, the scaling model unit (430) can be configured to suppress scaling formation by adjusting the flushing cycle in real time.

[0156] The flushing treatment step (S570) is a process of adjusting the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence predicted in the scaling model step (S550) and injecting a scaling inhibitor into the reverse osmosis membrane module through the batch reverse osmosis line. The flushing treatment unit (440) can determine the flushing flow rate, flushing duration, and flushing pressure intensity to apply an optimal inhibitor injection strategy.

[0157] For example, if the probability of scaling occurring is predicted to be 90% or higher in the scaling model stage (S550), the flushing processing unit (440) can shorten the existing flushing cycle (60 minutes) to 30 minutes.

[0158] Additionally, the scaling treatment unit (440) can monitor the effect of the scaling inhibitor in real time and automatically adjust the amount of inhibitor injected so that it is not excessive. For example, the scaling treatment unit (440) can stop additional injection when the inhibitor concentration is 20 mg / L or higher, and conversely, when the inhibitor concentration decreases to 15 mg / L or lower, it can proceed with additional injection to maintain the minimum effective concentration.

[0160] Figure 6 is a diagram illustrating the change in the flushing method performed in the batch reverse osmosis process in the batch reverse osmosis desalination device based on the scaling occurrence prediction of Figure 1.

[0161] Figure 6a shows a conventional flushing step, and Figure 6b shows an improved flushing step utilizing a concentrated water circulation line.

[0163] FIG. 6a illustrates the flushing process performed by the flushing treatment unit (440) in the conventional flushing method. In the conventional flushing method, flushing is performed by supplying the influent water to the reverse osmosis membrane (360) through the high-pressure pump (330), and then discharging the concentrated water to the concentrated water tank (390) after passing through the pressure exchange vessel (340).

[0164] In the existing flushing method, the flushing flow rate and flushing duration are maintained at a fixed level according to preset values, and flushing is performed at a constant interval regardless of the possibility of scaling occurring.

[0165] For example, in the case where the flushing cycle is set to 60 minutes in the conventional flushing method, the flushing treatment unit (440) can be set to operate the high-pressure pump (330) every 60 minutes to supply the influent water to the reverse osmosis membrane (360) and discharge the concentrated water to the concentrated water tank (390). However, the conventional flushing method may have reduced operational efficiency due to the large amount of water used for flushing, and there may be limitations in optimizing the flushing effect.

[0167] FIG. 6b illustrates a method in which a flushing treatment unit (440) improves flushing efficiency by utilizing a concentrated water circulation line. In the method utilizing a concentrated water circulation line, a scaling model unit (430) analyzes the scaling index and the probability of scaling occurring, and the flushing treatment unit (440) can dynamically adjust the flushing cycle and flushing conditions based on the scaling prediction results. Additionally, the flushing effect can be maximized by reusing the concentrated water in the circulation line.

[0168] For example, if the scaling model unit (430) predicts that the probability of scaling occurring is 85% or higher, the flushing treatment unit (440) may apply a method of recirculating concentrated water by shortening the flushing cycle to 30 minutes and adjusting the flushing flow rate, unlike the existing flushing method. The flushing treatment unit (440) may optimize the flushing flow rate by circulating the concentrated water to the pressure exchange vessel (340) and control it so as to maintain the same cleaning effect while reducing the amount of wastewater discharged.

[0169] Additionally, the scaling treatment unit (440) can monitor the concentration of the scaling inhibitor in real time and dynamically adjust the amount of inhibitor injected to maintain a minimum effective concentration during the flushing process. For example, when the flushing treatment unit (440) measures the TDS concentration of the concentrated water, if the TDS concentration exceeds 5,000 mg / L, the flushing treatment unit (440) can adjust the flushing pressure to prevent excessive injection of the inhibitor. Conversely, if the TDS concentration decreases below a reference value, the flushing treatment unit (440) can control the injection of additional inhibitor to maintain an appropriate concentration.

[0171] The difference between the conventional flushing method (Fig. 6a) and the improved flushing method using a concentrated water circulation line (Fig. 6b) can be seen in the difference in the flushing execution method and the scaling prevention effect. In the conventional flushing method, the flushing cycle and flushing flow rate are maintained at pre-set fixed values, and flushing is performed at a constant interval. On the other hand, in the flushing method using a concentrated water circulation line, the scaling model unit (430) predicts the possibility of scaling occurring in real time, and the flushing processing unit (440) can dynamically adjust the flushing cycle and flushing flow rate according to process conditions.

[0172] The flushing treatment unit (440) with the concentrated water recirculation method can reduce wastewater discharge compared to the conventional method while maximizing the effect of preventing scaling. The flushing method with the concentrated water recirculation method can contribute to reducing the amount of water used in the batch reverse osmosis process, extending the lifespan of the reverse osmosis membrane, and effectively suppressing the occurrence of scaling.

[0174] Although the present invention has been described above with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as described in the following claims. Explanation of the symbols

[0176] 100: Batch reverse osmosis desalination system based on scaling occurrence prediction 110: User terminal 130: Batch Reverse Osmosis Desalination Device Based on Scaling Occurrence Prediction 150: Database 210: Processor 230: Memory 250: User I / O 270: Network I / O Section 290: Communication port section 310: Preprocessing unit 320: Sensor section 330: High-pressure pump 340: Pressure Exchange Vessel 350: Free Piston 360: Reverse osmosis membrane 370: Pressure gauge 380: Filtered water tank 390: Concentrated water tank a: Batch reverse osmosis line 410: Scaling characteristic data collection unit 411: 1st sensor module 412: Second sensor module 420: Scaling Index Derivation Section 430: Scaling Model Section 440: Flushing Process Unit 450: Control unit

Claims

Claim 1 A batch reverse osmosis desalination device based on scaling occurrence prediction, comprising: a scaling characteristic data collection unit that collects scaling characteristic data of influent water and concentrated water in a batch reverse osmosis process; a scaling index derivation unit that evaluates the possibility of scaling occurrence based on the scaling characteristic data and calculates a scaling index; a scaling model unit that inputs the scaling index as input data into a deep learning-based scaling prediction model to predict the time of scaling occurrence; and a flushing treatment unit that adjusts the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and proceeds with the injection of a scaling inhibitor into a reverse osmosis membrane module through a batch reverse osmosis line. Claim 2 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 1, wherein the scaling characteristic data collection unit measures at least one physical or chemical process data among TDS (Total Dissolved Solids), pH, flow rate, pressure, and temperature as scaling characteristic data through a first sensor module connected to a storage tank storing the influent water in the batch reverse osmosis process and a second sensor module connected to a reverse osmosis membrane module storing the concentrated water. Claim 3 A batch reverse osmosis desalination device based on scaling occurrence prediction, characterized in that, in paragraph 2, the scaling characteristic data collection unit periodically measures the scaling characteristic data according to a measurement cycle, and the measurement cycle is dynamically adjusted based on the rate of change in flow rate or the rate of change in TDS of the influent or the concentrated water. Claim 4 In paragraph 2, the scaling characteristic data collection unit predicts a point in time when the probability of scaling occurrence is above a specific standard through the deep learning-based scaling prediction model, and automatically adjusts the measurement cycle or measurement event according to the predicted point in time, characterized by a scaling occurrence prediction-based batch reverse osmosis desalination device. Claim 5 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 1, wherein the scaling index derivation unit generates normalized data by performing missing value correction and normalization on the scaling characteristic data, and derives the scaling index by calculating the Supersaturation Index (SI), Scaling Potential Index (SPI), Langelier Saturation Index (LSI), and Stiff & Davis Saturation Index (S&DSI) on the normalized data. Claim 6 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 5, wherein the scaling index derivation unit finally calculates the scaling index by applying weights based on the importance of each of the supersaturation index, the scaling possibility index, the Langelier saturation index, and the Stiff & Davis saturation index. Claim 7 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 1, wherein the scaling model unit implements the scaling prediction model using a Long Short-Term Memory (LSTM) algorithm that learns the temporal change of the scaling index to predict the time of scaling occurrence. Claim 8 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 7, characterized in that the scaling model unit performs learning optimization by adjusting the weights of the loss function based on a pattern in which the frequency of scaling occurrence decreases according to the injection of the inhibitor. Claim 9 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 1, wherein the flushing treatment unit adjusts the flushing conditions according to the time of scaling occurrence to determine the flushing flow rate, flushing duration, and flushing pressure intensity for the injection of the scaling inhibitor. Claim 10 A batch reverse osmosis desalination device based on scaling occurrence prediction according to claim 9, wherein the flushing treatment unit monitors the injection of the scaling inhibitor in real time to dynamically adjust the injection amount of the scaling inhibitor and adjusts the flushing pressure intensity to prevent excessive injection while maintaining the minimum effective concentration of the scaling inhibitor. Claim 11 A batch reverse osmosis desalination method based on scaling occurrence prediction performed in a batch reverse osmosis desalination device based on scaling occurrence prediction, comprising: a scaling characteristic data collection step for collecting scaling characteristic data of influent water and concentrated water in a batch reverse osmosis process; a scaling index derivation step for evaluating the possibility of scaling occurrence based on the scaling characteristic data and calculating a scaling index; a scaling model step for inputting the scaling index as input data into a deep learning-based scaling prediction model to predict the time of scaling occurrence; and a flushing treatment step for adjusting the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and proceeding to inject a scaling inhibitor into a reverse osmosis membrane module through a batch reverse osmosis line. Claim 12 A batch reverse osmosis desalination device capable of optimizing the amount of an anti-scaling agent injected in real time, comprising: a scaling characteristic data collection unit that measures at least one of the TDS, pH, flow rate, pressure, and temperature of the influent and concentrated water in a batch reverse osmosis process as scaling characteristic data; a scaling index derivation unit that calculates a scaling index by calculating a supersaturation index, a scaling possibility index, a Langelier saturation index, and a Stiff & Davis saturation index, respectively, for the scaling characteristic data; a scaling model unit that predicts the time of scaling occurrence through a scaling prediction model based on an LSTM (Long Short-Term Memory) algorithm that learns temporal changes of the scaling index; and a flushing treatment unit that adjusts the flushing conditions of the batch reverse osmosis process according to the time of scaling occurrence and injects an anti-scaling agent into a reverse osmosis membrane module through a batch reverse osmosis line.

Images