Water electrolysis system

Abstract

To detect abnormality of a water electrolysis stack while reducing efficiency deterioration of water electrolysis and deterioration of the water electrolysis stack.SOLUTION: The water electrolysis system includes a plurality of water electrolysis stacks, a power supply that supplies a direct current to each of the plurality of water electrolysis stacks, and a detection unit that detects an abnormality of each of the plurality of water electrolysis stacks. The detection unit generates a plurality of characteristic models by calculating, for each of the plurality of water electrolysis stacks, a characteristic model approximating a response characteristic of an output voltage with respect to a change in an input current of the water electrolysis stack for each predetermined period based on measured values of a current and a voltage of the water electrolysis stack, and estimates an abnormality of the water electrolysis stack based on a characteristic difference between at least two characteristic models among the plurality of characteristic models.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a water electrolysis system. [Background technology]

[0002] Conventionally, water electrolysis systems that produce gases such as hydrogen by water electrolysis using a water electrolysis stack have been known, as disclosed in, for example, Patent Documents 1 to 3. The water electrolysis stack may experience an abnormality called a micro-short circuit.

[0003] For example, Patent Document 1 describes a water electrolysis system that applies a voltage for diagnosing an abnormality to a single cell of the water electrolysis stack when the water electrolysis stack is stopped, calculates the resistance of the single cell from the voltage applied to the single cell and the current flowing through the single cell, and detects whether or not a micro-short circuit has occurred in the single cell based on the calculated resistance. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-196906 [Patent Document 2] Japanese Patent Application Publication No. 2023-157554 [Patent Document 3] Japanese Patent Publication No. 2023-028092 Summary of the Invention [Problem to be solved by the invention]

[0005] In the water electrolysis system described in Patent Document 1, it is necessary to stop the operation of water electrolysis when detecting the occurrence of a micro-short circuit. As a result, the water electrolysis system described in Patent Document 1 has a problem of reducing the efficiency of water electrolysis. Furthermore, when detecting the occurrence of a micro-short circuit periodically, the water electrolysis operation is repeatedly stopped, which also has a problem of accelerating the deterioration of the water electrolysis stack.

[0006] In consideration of the above circumstances, an object of at least one aspect of the present disclosure is to detect an abnormality in a water electrolysis stack while reducing a decrease in efficiency of water electrolysis and deterioration of the water electrolysis stack. [Means for solving the problem]

[0007] In order to solve the above problems, a water electrolysis system according to one embodiment of the present disclosure includes a plurality of water electrolysis stacks, a power supply that supplies DC current to each of the plurality of water electrolysis stacks, and a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks. The detection unit generates a plurality of characteristic models by calculating, for each of the plurality of water electrolysis stacks, a characteristic model that approximates a response characteristic of the output voltage of the water electrolysis stack to a change in input current, based on measured values ​​of the current and voltage of the water electrolysis stack, and estimates an abnormality in the water electrolysis stack based on a characteristic difference between at least two of the plurality of characteristic models.

[0008] A water electrolysis system according to another aspect of the present disclosure includes a plurality of water electrolysis stacks, a power supply that supplies direct current to each of the plurality of water electrolysis stacks, and a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks, wherein the detection unit estimates an abnormality in each of the plurality of water electrolysis stacks based on multiple sets of measurement values ​​of the current and voltage of the water electrolysis stack during water electrolysis operation. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a configuration diagram of a water electrolysis system according to an embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of a water electrolysis stack and a supply device. [Figure 3] FIG. 1 is a diagram showing an electric circuit model of a water electrolysis stack. [Figure 4] 1 is a graph showing the change over time in output voltage relative to input current of a water electrolysis stack. [Figure 5] 3 is a flowchart showing the operation of the water electrolysis system according to the embodiment. [Figure 6] FIG. 10 is a diagram for explaining fitting using measured values ​​of current and voltage of a water electrolysis stack. [Figure 7] FIG. 10 is a diagram for explaining a characteristic model. [Figure 8] FIG. 10 is a diagram for explaining estimation of an abnormality in the water electrolysis stack. [Figure 9] FIG. 10 is a diagram illustrating an example of notification of an abnormality in the water electrolysis stack. DETAILED DESCRIPTION OF THE INVENTION

[0010] The embodiments for carrying out the present disclosure will be described with reference to the drawings. Note that the dimensions and scale of each element in each drawing may differ from those of the actual product. Furthermore, the embodiment described below is an exemplary embodiment that may be envisioned when carrying out the present disclosure. Therefore, the scope of the present disclosure is not limited to the embodiment exemplified below.

[0011] 1. Embodiment 1-1. Overview of the water electrolysis system 100 Fig. 1 is a configuration diagram of a water electrolysis system 100 according to an embodiment. The water electrolysis system 100 produces hydrogen and oxygen through water electrolysis using renewable energy or power supplied from a power transmission and distribution system. As shown in Fig. 1 , the water electrolysis system 100 includes multiple water electrolysis stacks 10, a power source 20, a supply device 40, a temperature sensor 30, a control device 50, and a display device 60.

[0012] Each of the water electrolysis stacks 10 is a structure that generates hydrogen and oxygen by water electrolysis. The water electrolysis stack 10 includes at least one cell of, for example, an alkaline water electrolysis type, a polymer electrolyte membrane (PEM) type, or an anion exchange membrane (AEM) type. An example of the configuration of the anion exchange membrane type will be described in detail later with reference to FIG. 2.

[0013] The power supply 20 is a device that supplies DC current to each of the multiple water electrolysis stacks 10. In the example shown in FIG. 1 , the power supply 20 supplies DC current to each of the multiple water electrolysis stacks 10 using power generated by renewable energy such as wind power or solar power. Specifically, the power supply 20 has multiple DC power supplies 21 corresponding to the multiple water electrolysis stacks 10. Each DC power supply 21 converts the power from AC to DC as necessary and supplies the DC current to the corresponding water electrolysis stack 10. Note that the power supply 20 may also use power generated by energy other than renewable energy.

[0014] The temperature sensor 30 is a sensor that detects the temperature of at least one water electrolysis stack 10. The temperature sensor 30 is, for example, a thermistor or a thermocouple that is installed in the water electrolysis stack 10. Note that a plurality of temperature sensors 30 may be provided for each water electrolysis stack 10.

[0015] The supply device 40 is a mechanism that supplies the electrolytic solution to each of the water electrolysis stacks 10 and collects hydrogen and oxygen generated by water electrolysis together with excess electrolytic solution from each of the water electrolysis stacks 10. An example configuration of the supply device 40 will be described later with reference to Fig. 2. Note that the supply device 40 may be provided individually for each water electrolysis stack 10.

[0016] The display device 60 is a display device that displays various types of information, and includes various display panels such as a liquid crystal panel and an organic EL panel. In this embodiment, the display device 60 is used to notify information related to an abnormality in the water electrolysis stack 10, as will be described later. The display device 60 is provided as needed and may be omitted.

[0017] The control device 50 is a computer system that controls the operation of the power supplies 20, and controls the current flowing through each water electrolysis stack 10 based on the input voltage and output voltage of each water electrolysis stack 10. In the example shown in Fig. 1 , the control device 50 controls not only the operation of the power supplies 20, but also the operation of the supply device 40 and the display device 70. Note that the element that controls the operation of the supply device 40 may be separate from the control device 50.

[0018] In this embodiment, the control device 50 has a function of detecting an abnormality in each water electrolysis stack 10 based on the input voltage and output voltage of each water electrolysis stack 10.

[0019] The control device 50 includes a storage device 51 and a processing device 52 .

[0020] The storage device 51 is a device that stores programs executed by the processing device 52 and data processed by the processing device 52. The storage device 51 includes, for example, a hard disk drive or a semiconductor memory. Note that part or all of the storage device 51 may be provided in an external storage device or server outside the control device 50.

[0021] The storage device 51 stores characteristic data Da-1 to Da-N and abnormality information Db.

[0022] Each of the characteristic data Da-1 to Da-N is information indicating a characteristic model that approximates the response characteristic of the output voltage to a change in the input current of the water electrolysis stack 10. N is a natural number equal to or greater than 2. N indicates a chronological order, and the characteristic data Da-1 to Da-N indicate characteristic models for different time periods. Hereinafter, the characteristic data Da-1 to Da-N may be referred to as characteristic data Da without distinction. Although not shown, the characteristic data Da-1 to Da-N are stored in the storage device 51 individually for each water electrolysis stack 10.

[0023] The abnormality information Db is information related to an abnormality in the water electrolysis stack 10, and includes, for example, information indicating the presence or absence of an abnormality in the water electrolysis stack 10 and information indicating the degree of the abnormality in the water electrolysis stack 10. An abnormal state in the water electrolysis stack 10 refers to an abnormal state in the water electrolysis stack 10 that is different from the normal state, and includes, for example, a state in which the degree of abnormality in the water electrolysis stack 10, such as a micro-short circuit, has reached a predetermined level or higher, as well as a state in which the deterioration level of the water electrolysis stack 10 has reached a predetermined level or higher. The degree of abnormality in the water electrolysis stack 10 is represented, for example, by the characteristic difference between the characteristic models indicated by at least two of the characteristic data Da among the characteristic data Da-1 to Da-N. Although not shown, the abnormality information Db is stored in the storage device 51 for each water electrolysis stack 10. The abnormality information Db may include information indicating the degree of abnormality in the water electrolysis stack 10, or may include information indicating the degree of past abnormalities in the water electrolysis stack 10. The abnormality information Db may also include information related to the abnormality in the water electrolysis stack 10, such as information about the past operation history of the water electrolysis stack 10, such as the current and voltage.

[0024] The processing device 52 is a device having a function of controlling each part of the control device 50 and a function of processing various data. The processing device 52 includes at least one processor such as a CPU (Central Processing Unit). Note that some or all of the functions of the processing device 52 may be realized by hardware such as a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). The processing device 52 may also be configured integrally with the storage device 51.

[0025] The processing device 52 executes a program stored in the storage device 51 to function as a detection unit 52a, a notification unit 52b, and an output control unit 52c.

[0026] The detection unit 52a detects an abnormality in each of the water electrolysis stacks 10 based on the measured values ​​of the current and voltage of the water electrolysis stack 10. Specifically, the detection unit 52a first calculates a characteristic model for each of the water electrolysis stacks 10 for each predetermined period based on the measured values ​​of the current and voltage of the water electrolysis stack 10. As a result, multiple characteristic models are generated for each predetermined period. Information indicating the generated characteristic models is stored in the storage device 51. The predetermined period may be a fixed period or a variable period.

[0027] The detector 52a estimates an abnormality in each of the water electrolysis stacks 10 based on the characteristic difference between the characteristic models indicated by at least two pieces of characteristic data Da among the characteristic data Da-1 to Da-N stored in the storage device 51. Information related to the estimated abnormality is stored in the storage device 51 as abnormality information Db.

[0028] The notification unit 52b notifies information about abnormalities in the plurality of water electrolysis stacks 10 based on the abnormality information Db, which is the detection result of the detection unit 52a. In this embodiment, the notification unit 52b displays information about abnormalities in the plurality of water electrolysis stacks 10 on the display device 60. The timing of notification by the notification unit 52b is not particularly limited, and may be, for example, each time the abnormality information Db is generated, when a user gives an instruction, or when the level of abnormality in at least one water electrolysis stack 10 reaches a predetermined level or higher.

[0029] The output control unit 52c controls the current flowing through each of the water electrolysis stacks 10 based on the measured current and voltage values ​​of the water electrolysis stack 10. The output control unit 52c may adjust or stop the current flowing through each of the water electrolysis stacks 10 based on the abnormality information Db. For example, as the level of the abnormality increases, the output control unit 52c may reduce the current flowing through the water electrolysis stack 10 or stop the current flowing through a water electrolysis stack 10 whose level of abnormality has reached a predetermined level or higher.

[0030] 1-2. Water electrolysis stack and supply device FIG. 2 is a diagram illustrating an example configuration of the water electrolysis stack 10 and the supply device 40. Note that FIG. 2 illustrates an example in which the water electrolysis stack 10 is an anion exchange membrane type, but the type and configuration of the water electrolysis stack 10 are not limited to the example illustrated in FIG. 2 and may be of various known types and configurations. Also, while FIG. 2 illustrates the water electrolysis stack 10 having one cell, the number of cells is not limited thereto and the water electrolysis stack 10 may be configured with multiple cells stacked one upon the other. Furthermore, the configuration of the water electrolysis stack 10 is not limited to the example illustrated in FIG. 2 and may include, for example, a mechanism having only the function of collecting hydrogen generated by water electrolysis together with excess electrolyte, instead of the first supply mechanism 41 described below.

[0031] 2, the water electrolysis stack 10 includes a cathode 11, an anode 12, an anion exchange membrane 13, a first flow path member 14, a second flow path member 15, a first gasket 16, and a second gasket 17. These members form an anion exchange membrane cell, and perform water electrolysis using an electrolyte.

[0032] The electrolyte is, for example, pure water or an alkaline solution. Preferably, the electrolyte is an aqueous solution of potassium hydroxide (KOH), obtained by dissolving potassium hydroxide at a concentration of 1 M (mol / L) in pure water with a purity of about 20 μS / cm. The electrolyte may contain metal ions such as sodium (Na), magnesium (Mg), or calcium (Ca) as impurities.

[0033] A laminate of the cathode 11, the anode 12, and the anion exchange membrane 13 constitutes a membrane electrode assembly (MEA). Here, the cathode 11, the anion exchange membrane 13, and the anode 12 are laminated in this order. That is, the anion exchange membrane 13 is disposed between the cathode 11 and the anode 12. In the example shown in FIG. 2, the cathode 11 is disposed on the left side of the anion exchange membrane 13 in FIG. 2, and the anode 12 is disposed on the right side in FIG. 2.

[0034] The anion exchange membrane 13 is also called an AEM (Anion Exchange Membrane) and is an anion exchange membrane that exchanges hydroxide ions (OH - The anion exchange membrane 13 is an electrolyte membrane that selectively allows the permeation of anions such as cations, etc. The material constituting the anion exchange membrane 13 is not particularly limited, but examples thereof include polymer materials such as polyacrylonitrile (PAN), polysulfone (PS), polyvinyl alcohol (PVA), and polyethylene oxide (PEO).

[0035] The anion exchange membrane 13 has a first surface F1 and a second surface F2. The first surface F1 and the second surface F2 face in opposite directions. In the example shown in FIG. 2, the first surface F1 faces left in FIG. 2. On the other hand, the second surface F2 faces right in FIG. 2. The thickness of the anion exchange membrane 13 is, for example, about 30 μm.

[0036] The cathode 11 is laminated on the first face F1. In the cathode 11, hydrogen gas is generated by, for example, the following first reaction.

[0037] 4H2O+4e - →2H2+4OH - As shown in FIG. 2, the cathode 11 has a catalyst layer 11a and a diffusion layer 11b. These are stacked in this order from right to left in FIG. 2: catalyst layer 11a, diffusion layer 11b. Therefore, the catalyst layer 11a is disposed between the first surface F1 and the diffusion layer 11b. The catalyst layer 11a is a membrane that promotes the first reaction described above, and is in close contact with the first surface F1. The catalyst layer 11a is made of, for example, carbon supporting a metal material such as platinum (Pt). The material that makes up the catalyst layer 11a is not particularly limited and may be any material, such as nickel (Ni).

[0038] The diffusion layer 11b is an element for efficiently separating and discharging the hydrogen gas generated by the first reaction. The diffusion layer 11b is made of, for example, a nickel (Ni) foam, a porous film of a carbon-based material, or conductive fibers. The diffusion layer 11b is also conductive and functions as a path for electrons exchanged with the catalyst layer 11a.

[0039] The anode 12 is laminated on the second face F2. In the anode 12, oxygen gas is generated, for example, by the following second reaction.

[0040] 4OH - →O2+2H2O+4e - As shown in FIG. 2, the anode 12 has a catalyst layer 12a and a diffusion layer 12b. These are stacked from left to right in FIG. 2, in the order of catalyst layer 12a and diffusion layer 12b. Therefore, the catalyst layer 12a is disposed between the second surface F2 and the diffusion layer 12b. The catalyst layer 12a is a thin film that promotes the second reaction and is in close contact with the second surface F2. The catalyst layer 12a is made of a metal material such as iridium (Ir), iron (Fe), or nickel (Ni). The material that makes up the catalyst layer 12a is not particularly limited and may be made of, for example, an oxide of the metal material.

[0041] The diffusion layer 12b is an element for efficiently separating and discharging the oxygen gas generated by the second reaction. The diffusion layer 12b is made of, for example, a nickel (Ni) foam, a porous membrane of a carbon-based material, or conductive fibers. The diffusion layer 12b also functions as an element for efficiently supplying the electrolyte to the catalyst layer 12a. Furthermore, the diffusion layer 12b is conductive and functions as a path for electrons exchanged with the catalyst layer 12a.

[0042] The above-described stack of the cathode 11, anode 12, and anion exchange membrane 13 is disposed between a first flow path member 14 and a second flow path member 15. Here, the first flow path member 14 and the second flow path member 15 are fixed to each other with fasteners such as screws or bolts, so that the stack is sandwiched between the first flow path member 14 and the second flow path member 15.

[0043] The first flow path member 14 is a plate-shaped structure made of a conductive material such as metal, and is also called a separator. The first flow path member 14 faces the first surface F1 across the cathode 11. That is, the cathode 11 is located between the first surface F1 and the first flow path member 14.

[0044] A first flow path 14a is provided on the plate surface facing the first surface F1 of the pair of plate surfaces of the first flow path member 14. The first flow path 14a is a flow path defined by the cathode 11 and the first flow path member 14. Therefore, for example, hydrogen gas generated at the cathode 11 by the first reaction flows into the first flow path 14a.

[0045] The second flow path member 15 is a plate-shaped structure made of a conductive material such as metal, and is also called a separator. The second flow path member 15 faces the second surface F2 across the anode 12. That is, the anode 12 is located between the second surface F2 and the second flow path member 15.

[0046] A second flow path 15a is provided on the plate surface of the pair of plate surfaces of the second flow path member 15 that faces the second surface F2. The second flow path 15a is a flow path defined by the anode 12 and the second flow path member 15. Therefore, for example, oxygen gas generated at the anode 12 by the second reaction flows into the second flow path 15a. The second flow path 15a also functions as a path for circulating the electrolytic solution.

[0047] The first gasket 16 is a frame-shaped sealing member that seals the cathode 11 in the space between the first flow path member 14 and the anion exchange membrane 13. Similarly, the second gasket 17 is a frame-shaped sealing member that seals the anode 12 in the space between the second flow path member 15 and the anion exchange membrane 13. Each of the first gasket 16 and the second gasket 17 is made of an elastic material such as vinyl methyl silicone rubber (VMQ), fluororubber (FKM), ethylene propylene diene rubber (EPDM), or polytetrafluoroethylene (PTFE).

[0048] A DC power supply 21 is electrically connected to the first flow path member 14 and the second flow path member 15. This allows a predetermined current to be supplied from the DC power supply 21 to the water electrolysis stack 10. The current density during water electrolysis is, for example, 1 A / cm. 2 The DC power supply 21 may be electrically connected to the cathode 11 and the anode 12.

[0049] A supply device 40 is connected to each of the first flow path 14a of the first flow path member 14 and the second flow path 15a of the second flow path member 15. The supply device 40 has a first supply mechanism 41 and a second supply mechanism .

[0050] The first supply mechanism 41 is connected to the first flow path 14a and supplies the electrolytic solution to the first flow path 14a and recovers hydrogen produced by water electrolysis together with excess electrolytic solution from the first flow path 14a. The first supply mechanism 41 has a first container 41a and a first pump 41b.

[0051] The first container 41a is a container that stores the electrolytic solution. The first pump 41b is a liquid transfer mechanism that supplies the electrolytic solution stored in the first container 41a to the water electrolysis stack 10. Specifically, the first pump 41b supplies the electrolytic solution in the first container 41a to the first flow path 14a via a first supply pipe 41c. The first supply pipe 41c is a conduit for supplying a remover solution (described below) to the water electrolysis stack 10.

[0052] When the power supply 20 supplies power to the water electrolysis stack 10, hydrogen gas is generated at the cathode 11 due to the first reaction using the electrolytic solution supplied from the anode 12. The hydrogen gas generated by the first reaction and excess electrolytic solution are discharged from the first discharge pipe 41d to the first container 41a. The first discharge pipe 41d is a conduit for discharging the hydrogen gas generated by the first reaction and excess electrolytic solution from the first flow path 14a. The first container 41a functions as a gas-liquid separation tank that separates the hydrogen gas and the electrolytic solution.

[0053] The second supply mechanism 42 is connected to the second flow path 15a and supplies the electrolytic solution to the second flow path 15a and recovers oxygen produced by water electrolysis together with excess electrolytic solution from the second flow path 15a. The second supply mechanism 42 has a second container 42a and a second pump 42b.

[0054] The second container 42a is a container that stores the electrolytic solution. The second pump 42b is a liquid transfer mechanism that supplies the electrolytic solution stored in the second container 42a to the water electrolysis stack 10. Specifically, the second pump 42b supplies the electrolytic solution in the second container 42a to the second flow path 15a via the second supply pipe 42c. The second supply pipe 42c is a pipeline for supplying the electrolytic solution to the water electrolysis stack 10.

[0055] When the power supply 20 supplies power to the water electrolysis stack 10, oxygen gas is generated at the anode 12 due to the second reaction using the electrolytic solution. The oxygen gas generated by the second reaction and excess electrolytic solution are discharged to the second container 42a via the second discharge pipe 42d. The second discharge pipe 42d is a conduit for discharging the oxygen gas generated by the second reaction and excess electrolytic solution from the second flow path 15a. The second container 42a functions as a gas-liquid separation tank that separates the oxygen gas from the electrolytic solution.

[0056] 1-3. Characteristic model of water electrolysis stack 3 is a diagram showing an electric circuit model of the water electrolysis stack 10. In the electric circuit model, as shown in FIG. 3, the anode 12 is connected in parallel with a resistor R a and capacitor C a The anion exchange membrane 13 has a resistance R m The cathode 11 is connected to a voltage source V ref In this electrical circuit model, the resistance R a and capacitor C a and resistor R m and the voltage source V ref are connected in series. Therefore, if the input current i at time t is i(t), the output voltage V of the water electrolysis stack 10 at time t is cell V cell (t), and the voltage V of the anode 12 at time (t) act V act (t), and the voltage V of the anion exchange membrane 13 at time (t) m V m (t), the output voltage V cell (t) satisfies the relationship of the following formula: This relationship is an electric circuit equation of the water electrolysis stack 10.

number

[0057] In this formula, K1, K2, K3, and K4 are correction coefficients such as temperature coefficients, and C0, R0, and R m0 , V ref0Each of these is a reference value.

[0058] Fig. 4 is a graph showing the change over time in the output voltage versus the input current of the water electrolysis stack 10. In Fig. 4, the vertical axis represents the output voltage or input current, and the horizontal axis represents time. In Fig. 4, the input current is represented by a dashed line, and the output voltage is represented by a solid line.

[0059] Since the water electrolysis stack 10 is represented by the electric circuit model described above, when the input current changes, the output voltage gradually increases, lagging behind the change in input current, even if the input current increases sharply, as shown in Fig. 4. This steady-state value of the output voltage increases as the degree of deterioration of the water electrolysis stack 10 increases. Therefore, the degree of deterioration of the water electrolysis stack 10 can be estimated based on this change in the output voltage.

[0060] Furthermore, an abnormality called a microshort circuit may occur in the water electrolysis stack 10. A "microshort circuit" refers to a state in which a conductive material, such as conductive fibers, constituting the diffusion layers 11b and 12b (described below) penetrates the anion exchange membrane 13, electrically connecting the cathode 11 and the anode 12. When the anion exchange membrane 13 is thinned to improve water electrolysis efficiency, the probability of such a microshort circuit occurring increases. When a microshort circuit occurs, a current flows through the diffusion layers 11b and 12b during water electrolysis, causing the diffusion layers 11b and 12b to generate heat, which in turn accelerates the burnout of the anion exchange membrane 13 located near the diffusion layers 11b and 12b. As a result, a pinhole is formed in the anion exchange membrane 13. The formation of a pinhole in the anion exchange membrane 13 can lead to the problem of water or gas leaking from one side of the cathode 11 to the other side of the anode 12.

[0061] When an abnormality such as a micro-short occurs in the water electrolysis stack 10, the resistance value of the water electrolysis stack 10 decreases. Therefore, the degree of the abnormality such as a micro-short in the water electrolysis stack 10 can be estimated based on the resistance value of the water electrolysis stack 10.

[0062] However, because the response characteristic of the output voltage to a change in input current of the water electrolysis stack 10 is a static characteristic of the water electrolysis stack 10, directly measuring this response characteristic requires temporarily halting the water electrolysis operation. Temporarily halting the water electrolysis operation can result in a decrease in the efficiency of the water electrolysis. Furthermore, repeated halts in the water electrolysis operation can accelerate the deterioration of the water electrolysis stack 10. These problems also arise when measuring the resistance of the water electrolysis stack 10 while the water electrolysis operation is halted, as in Patent Document 1.

[0063] Therefore, the water electrolysis system 100 periodically calculates a characteristic model that approximates the response characteristic of the output voltage to changes in the input current of the water electrolysis stack 10 based on measured values ​​of the current and voltage of the water electrolysis stack 10, and estimates an abnormality in the water electrolysis stack 10 based on the characteristic model. This makes it possible to obtain the response characteristic of the output voltage to changes in the input current of the water electrolysis stack 10 even during water electrolysis operation. As a result, it is possible to detect an abnormality in the water electrolysis stack 10 while reducing a decrease in water electrolysis efficiency and deterioration of the water electrolysis stack 10. If the level of the abnormality in the water electrolysis stack 10 reaches a predetermined level or higher, the water electrolysis stack 10 is replaced with a new water electrolysis stack 10.

[0064] 1-4. Operation of the water electrolysis system 5 is a flowchart showing the operation of the water electrolysis system 100 according to an embodiment. In the water electrolysis system 100, first, in step ST1, the detector 52a measures the current and voltage of each water electrolysis stack 10. This obtains the measured values ​​of the current and voltage of each water electrolysis stack 10. In this embodiment, as described below, in step ST1, the detector 52a obtains multiple sets of measured values ​​of the current and voltage of each water electrolysis stack 10.

[0065] After step ST1, in step ST2, the detection unit 52a calculates a characteristic model for each of the plurality of water electrolysis stacks 10 based on the measured values ​​of the current and voltage of the water electrolysis stack 10.

[0066] After step ST2, in step ST3, the detection unit 52a stores information indicating the calculated characteristic model for each of the plurality of water electrolysis stacks 10 as characteristic data Da in the storage device 51. When step ST3 is executed for the nth time, the information indicating the calculated model is stored in the storage device 51 as characteristic data Da-n, where n is a natural number between 1 and N.

[0067] After step ST3, in step ST4, the detection unit 52a estimates an abnormality in each of the water electrolysis stacks 10 based on the characteristic difference between the characteristic models indicated by the characteristic data Da-1 to Da-N stored in the storage device 51. Information related to the estimated abnormality is stored as abnormality information Db in the storage device 51. When step ST4 is executed for the first time, the detection unit 52a does not estimate an abnormality in the water electrolysis stack 10, or estimates an abnormality in the water electrolysis stack 10 based on the characteristic difference between a preset characteristic model and the characteristic model indicated by the characteristic data Da-1.

[0068] After step ST4, in step ST5, the notification unit 52b determines whether to issue a notification based on the anomaly information Db. This determination is made based on, for example, whether a user operation has been performed to issue a notification command, or whether the level of abnormality in at least one of the water electrolysis stacks 10 has reached a predetermined level or higher. Therefore, for example, in step ST5, the notification unit 52b determines to issue a notification based on the anomaly information Db when a user instruction has been received, or when the level of abnormality in at least one of the water electrolysis stacks 10 has reached a predetermined level or higher. Note that step ST5 may be omitted. In this case, a notification based on the anomaly information Db is issued each time the anomaly information Db is generated.

[0069] If a notification based on the abnormality information Db is to be performed (step ST5: YES), in step ST6, the notification unit 52b notifies information related to the abnormality in each water electrolysis stack 10 based on the abnormality information Db. In the present embodiment, the notification unit 52b causes the display device 60 to display information related to the abnormalities in the plurality of water electrolysis stacks 10.

[0070] If a notification based on the abnormality information Db is not made (step ST5: NO), or after step ST6, in step ST7, the detection unit 52a determines whether a predetermined time has elapsed. The predetermined time is, for example, the length of time since the last execution of any of steps ST1 to ST6, or a time at a preset time interval. The length of the predetermined time is not particularly limited, but is, for example, between 1 second and 30 minutes.

[0071] If the predetermined time has not elapsed (step ST7: NO), the detection unit 52a proceeds to step ST5. Note that step ST7 may be repeatedly executed until the predetermined time has elapsed, without going through steps ST5 and ST6.

[0072] If the predetermined time has elapsed (step ST7: YES), in step ST8, the detection unit 52a determines whether to end the process. This determination is made based on, for example, whether the user has performed an operation to instruct the process to end.

[0073] If the process is not to be terminated (step ST8: NO), the detection unit 52a returns to step ST1. As a result, multiple characteristic models are generated for each predetermined period, an abnormality in the water electrolysis stack 10 is estimated for each predetermined period, and information about the abnormality is reported as necessary.

[0074] If the process is to be ended (step ST8: YES), the detection unit 52a ends the process. In this way, the water electrolysis system 100 detects an abnormality in the water electrolysis stack 10. Specific examples of steps ST1 to ST4 and ST6 will be described below.

[0075] FIG. 6 is a diagram illustrating fitting using the measured values ​​of the current and voltage of the water electrolysis stack 10. In step ST1, the detection unit 52a measures the current and voltage of each water electrolysis stack 10 multiple times. This results in multiple sets of measured values ​​of the current and voltage of each water electrolysis stack 10. In the example shown in FIG. 6, 10 sets of measured values ​​are obtained. The number of sets of measured values ​​of the current and voltage of each water electrolysis stack 10 is not limited to 10, and may be 2 to 9 or 11 or more. However, from the viewpoint of achieving both high accuracy of the characteristic model and a short calculation time, the number of sets of measured values ​​of the current and voltage of each water electrolysis stack 10 is preferably 10 to 80, and more preferably 40 to 80.

[0076] Multiple sets of measurement values ​​of the current and voltage of each water electrolysis stack 10 are obtained by measurements at a predetermined sampling period. This sampling period is set so that at least two sets of measurement values ​​are different from each other, and is not particularly limited, but is, for example, 0.01 seconds to 5 seconds. Note that step ST1 is performed during water electrolysis operation. During water electrolysis operation, the current from the power source 20 to each water electrolysis stack 10 may fluctuate, and therefore the multiple sets of measurement values ​​may differ from each other. Since at least two sets of measurement values ​​are different from each other, a characteristic model can be suitably calculated in step ST2, as described below.

[0077] In step ST2, the detection unit 52a calculates a characteristic model of each water electrolysis stack 10 based on the plurality of sets of measured values ​​of the current and voltage of each water electrolysis stack 10 and the detection results of the temperature sensor 30, using an estimation model representing the electric circuit equation of the water electrolysis stack 10. The estimation model calculates a characteristic model of each water electrolysis stack 10 based on the plurality of sets of measured values ​​of the current and voltage of each water electrolysis stack 10 and the detection results of the temperature sensor 30. a , capacitor C a , resistance R m and the voltage source V ref The following is an arithmetic expression for generating a voltage value from a combination of the detection result of the temperature sensor 30 and the current value, using the above coefficients.

[0078] Specifically, in step ST2, the detection unit 52a sets each coefficient of the estimation model for each water electrolysis stack 10 by a nonlinear least squares method or the like so as to reduce the difference between the voltage (solid line in the figure) obtained when a pair of the measured current value and the detection result of the temperature sensor 30 is applied to the estimation model, and the measured voltage. a , capacitor C a , resistance R m and the voltage source V ref In step ST3, information indicating the set values ​​is stored as characteristic data Da in the storage device 51. Note that the characteristic data Da may be information indicating an estimation model to which the set coefficients are applied.

[0079] In this way, the control device 50 calculates a characteristic model for each of the plurality of water electrolysis stacks 10 by fitting the plurality of sets of measurement values ​​using an estimation model representing the electric circuit equation of the water electrolysis stack 10. This makes it possible to suitably generate a characteristic model even when the water electrolysis stack 10 is in water electrolysis operation. Note that the equation used for the estimation model is not limited to the above-described electric circuit equation of the water electrolysis stack 10, and may be an equation representing a similar function such as a quadratic function.

[0080] In the present embodiment, when setting the coefficients of the estimation model, the detection unit 52a adjusts the correction coefficients K1, K2, K3, and K4 based on the detection result of the temperature sensor 30. In this manner, the control device 50 adjusts the estimation model based on the detection result of the temperature sensor 30. This allows the coefficients of the estimation model to be set with high accuracy even if the temperature of the water electrolysis stack 10 fluctuates. Therefore, the characteristic model can be calculated with high accuracy even if the temperature of the water electrolysis stack 10 fluctuates. Note that the detection result of the temperature sensor 30 may not be used to set the coefficients of the estimation model, and the correction coefficients K1, K2, K3, and K4 may be fixed values. In this case, the coefficients of the estimation model may be set, for example, when the temperature range of the water electrolysis stack 10 is within the applicable range of the correction coefficients K1, K2, K3, and K4.

[0081] Fig. 7 is a diagram illustrating a characteristic model. By using an equation in which the coefficients set in step ST2 are applied to the estimation model described above, it is possible to obtain a characteristic model that shows the response characteristics of the output voltage when the input current to the water electrolysis stack 10 changes as shown by the dashed line in Fig. 7, as indicated by the solid line or dashed-dotted line in Fig. 7. The condition of the input current is not particularly limited and can be set arbitrarily in advance.

[0082] 7 shows multiple characteristic models calculated for each predetermined period. The characteristic models shift so that the output voltage increases as the deterioration of the water electrolysis stack 10 progresses. However, within a short period of time, such as a few minutes, the degree of abnormality and deterioration of the water electrolysis stack 10 hardly progresses. Therefore, the differences between the multiple characteristic models during that period can be attributed to calculation errors of the characteristic models.

[0083] Therefore, in step ST4, in order to reduce the influence of such calculation errors, the results of statistical processing of multiple characteristic models are appropriately used for estimating the deterioration level of the water electrolysis stack 10 every several minutes. FIG. 7 illustrates an example in which an intermediate characteristic model, indicated by a two-dot chain line in the figure, is used among the multiple characteristic models within the period. Note that the statistical processing is not limited to processing using the intermediate characteristic model, and may be, for example, processing to average the multiple characteristic models within the period. Furthermore, the statistical processing may be performed as needed or may be omitted. The statistical processing may be omitted before the first period has elapsed, for example, if the frequency of execution of step ST4 is equal to the frequency of execution of step ST2, as shown in FIG. 5 above.

[0084] By using the transition of the results of such statistical processing, the degree of abnormality and the progress of deterioration of the water electrolysis stack 10 can be grasped with high accuracy. Furthermore, the conditions for undesirable measurement values ​​can be determined from the relationship between the results of such statistical processing and multiple characteristic models during the period in which the statistical processing was performed. The conditions can be obtained, for example, by machine learning the relationship over the operation period of water electrolysis. By using the conditions in calculating the characteristic model, the calculation accuracy of the characteristic model can be improved. To use the conditions in calculating the characteristic model, for example, the conditions can be incorporated as weights into the above-mentioned estimation model.

[0085] In this way, the control device 50 can improve the accuracy of the characteristic model by correcting the estimation model based on the results of machine learning of the relationship between a plurality of characteristic models.

[0086] Fig. 8 is a diagram for explaining estimation of an abnormality in the water electrolysis stack 10. Fig. 8 shows a first characteristic model MD-1 and a second characteristic model MD-2.

[0087] The first characteristic model MD-1 is a characteristic model that shows the first result of the above-mentioned statistical processing. Note that the first characteristic model MD-1 may be a characteristic model based on one of the plurality of characteristic models, but if the above-mentioned statistical processing is not performed, it is preferably the first characteristic model, i.e., the characteristic model shown by the characteristic data Da-1.

[0088] In this way, the first characteristic model MD-1 is the first characteristic model among the plurality of characteristic models or is a characteristic model based on the first characteristic model, thereby enabling the deterioration level of the water electrolysis stack 10 to be suitably estimated.

[0089] The second characteristic model MD-2 is a characteristic model that indicates any result from the second or later time in the above-mentioned statistical processing. Note that the second characteristic model MD-2 may be any characteristic model based on a characteristic model that is later than the first characteristic model MD-1 among the multiple characteristic models, and if the above-mentioned statistical processing is not performed, the second characteristic model MD-2 is any characteristic model from the second or later time, i.e., a characteristic model indicated by any characteristic data Da among the characteristic data Da-2 to D1-N, and is preferably the latest characteristic model.

[0090] In step ST4, the detection unit 52a estimates an abnormality in the water electrolysis stack 10 based on the characteristic difference between the first characteristic model MD-1 and the second characteristic model MD-2.

[0091] Specifically, when the difference between the voltage indicated by the first characteristic model MD-1 and the voltage indicated by the second characteristic model MD-2 is ΔV, the deterioration degree D [%] of the water electrolysis stack 10 is expressed, for example, as D = ΔV × (cell voltage operating range) × 100. The cell voltage operating range is the voltage range of the water electrolysis stack 10 during water splitting operation.

[0092] The degree of abnormality such as a micro-short circuit in the water electrolysis stack 10 can be calculated by, for example, the resistance R a and resistor R m and the resistance R based on the second characteristic model MD-2 a and resistor R m It is expressed as the difference from the sum of

[0093] Furthermore, for example, when characteristic curves showing the response characteristics of the output voltage to changes in the input current of the water electrolysis stack 10 are drawn for each of the first characteristic model MD-1 and the second characteristic model MD-2, if the shape of the characteristic curve based on the second characteristic model MD-2 is extremely different from the shape of the characteristic curve based on the first characteristic model MD-1, it is estimated that there is an abnormality, such as a micro-short circuit, in the water electrolysis stack 10.

[0094] In this way, since the characteristic difference between the first characteristic model MD-1 and the second characteristic model MD-2 is the difference between the first characteristic model MD-1 and the second characteristic model MD-2, an abnormality in the water electrolysis stack 10 can be suitably estimated.

[0095] 9 is a diagram showing an example of notification of an abnormality in the water electrolysis stack 10. In step ST6, the notification unit 52b causes the display device 60 to display an image G based on the abnormality information Db, as shown in FIG.

[0096] The image G includes an image Ga-1 showing the state of the water electrolysis stack 10-1, an image Ga-2 showing the state of the water electrolysis stack 10-2, and an image Ga-3 showing the state of the water electrolysis stack 10-3. Hereinafter, the images Ga-1, Ga-2, and Ga-3 may be referred to as image Ga without distinction.

[0097] In the example shown in Fig. 9, the image Ga includes information indicating the resistance value of the water electrolysis stack 10, information indicating the degree of deterioration of the water electrolysis stack 10, and information indicating the presence or absence of an abnormality in the water electrolysis stack 10. The display mode of the image Ga is not limited to the example shown in Fig. 9 and may be any desired mode. The information displayed in the image Ga is not limited to the example shown in Fig. 9 and may include, for example, information related to the abnormality in the water electrolysis stack 10, such as information on the past operation history of the water electrolysis stack 10, such as the current and voltage.

[0098] As described above, the notification unit 52b notifies information about abnormalities in multiple water electrolysis stacks 10 based on the detection results of the detection unit 52a, thereby making it possible to notify the operator of information about the abnormality in each water electrolysis stack 10.

[0099] In the water electrolysis system 100 described above, the degree of deterioration of the water electrolysis stack 10 is estimated based on the characteristic difference ΔV between at least two of the multiple characteristic models for each predetermined period, making it possible to detect an abnormality in the water electrolysis stack 10 even when the water electrolysis stack 10 is in operation. Furthermore, even if an abnormality in the water electrolysis stack 10 is detected periodically, there is no need to repeatedly stop the water electrolysis operation, which has the advantage of not accelerating the deterioration of the water electrolysis stack 10. As described above, it is possible to detect an abnormality in the water electrolysis stack 10 while reducing the decrease in water electrolysis efficiency and the deterioration of the water electrolysis stack 10.

[0100] 2. Variations Specific modified embodiments that can be added to each of the embodiments exemplified above are exemplified below. Two or more embodiments arbitrarily selected from the following examples may be appropriately combined within a range that does not contradict each other.

[0101] 2-1. Variation 1 In the above-described embodiment, the processing device 52 of the control device 50 functions as the detection unit 52a, but the present invention is not limited to this. For example, a computer system separate from the control device 50 may function as the detection unit 52a. The computer system may be communicatively connected to the control device 50 and acquire measured values ​​of the voltages and currents of the multiple water electrolysis stacks 10 from the control device 50.

[0102] 2-2. Variation 2 In the above-described embodiment, the water electrolysis system 100 includes three water electrolysis stacks 10, but the present invention is not limited to this. The number of water electrolysis stacks 10 included in the water electrolysis system 100 may be two or four or more.

[0103] 2-3. Variation 3 In the above-described embodiment, the estimation model is corrected based on the detection result of the temperature sensor 30. However, the present invention is not limited to this example. For example, the correction may not be performed. In this case, the temperature sensor 30 may be omitted.

[0104] 2-4. Variation 4 In the above-described embodiment, the statistical processing of a plurality of characteristic models is performed in step ST4, but the present invention is not limited to this. For example, the statistical processing may be performed in step ST2. In this case, each of the characteristic data Da-1 to Da-N may be information indicating the result of the statistical processing for each predetermined period.

[0105] 2-5. Variation 5 In the above-described embodiment, the notification unit 52b notifies of an abnormality in the water electrolysis stack 10 by displaying an indication on the display device 60, but the present invention is not limited to this. For example, the notification unit 52b may notify of an abnormality in the water electrolysis stack 10 by sound, such as an alarm sound. Alternatively, the notification unit 52b may notify of an abnormality in the water electrolysis stack 10 by using a combination of multiple notification methods.

[0106] 2-6. Variation 5 In the above-described embodiment, a notification of an abnormality in the water electrolysis stack 10 is given based on the detection result of the detection unit 52a, but the present invention is not limited to this, and the detection result of the detection unit 52a may be used in any manner. For example, the detection result of the detection unit 52a may be used for power allocation to the multiple water electrolysis stacks 10 by the output control unit 52c.

[0107] 3. Notes From the above-described exemplary embodiments, the following configurations can be understood, for example.

[0108] (Supplementary Note 1) A first aspect, which is a preferred example of a water electrolysis system according to the present disclosure, comprises a plurality of water electrolysis stacks, a power supply that supplies DC current to each of the plurality of water electrolysis stacks, and a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks. The detection unit generates a plurality of characteristic models for each of the plurality of water electrolysis stacks by calculating, at predetermined intervals, a characteristic model that approximates the response characteristic of the output voltage of the water electrolysis stack relative to changes in input current based on measured values ​​of the current and voltage of the water electrolysis stack, and estimates an abnormality in the water electrolysis stack based on a characteristic difference between at least two of the plurality of characteristic models.

[0109] In the above-described aspect, an abnormality in the water electrolysis stack is estimated based on the characteristic difference between at least two of the plurality of characteristic models for each predetermined period, so that an abnormality in the water electrolysis stack can be detected even when the water electrolysis stack is in operation. Furthermore, detecting an abnormality in the water electrolysis stack periodically does not require repeatedly stopping the water electrolysis operation, which has the advantage of not accelerating deterioration of the water electrolysis stack. As a result, it is possible to detect an abnormality in the water electrolysis stack while reducing a decrease in water electrolysis efficiency and deterioration of the water electrolysis stack.

[0110] (Supplementary Note 2) In a second aspect which is a preferred example of the first aspect, the measurement values ​​are multiple sets of measurement values ​​of the current and voltage of the water electrolysis stack, and the detection unit calculates a characteristic model for each of the multiple water electrolysis stacks by fitting the multiple sets of measurement values ​​using an estimation model representing an electric circuit equation of the water electrolysis stack or an equation similar thereto. In this aspect, the characteristic model can be preferably generated even when the water electrolysis stack is in water electrolysis operation.

[0111] (Supplementary Note 3) In a third aspect, which is a preferred example of the second aspect, the device further includes a temperature sensor that detects the temperature of at least one of the plurality of water electrolysis stacks, and the detection unit corrects the estimation model based on the detection result of the temperature sensor. In this aspect, the characteristic model can be calculated with high accuracy even if the temperature of the water electrolysis stack fluctuates.

[0112] (Supplementary Note 4) In a fourth aspect which is a preferred example of the second or third aspect, the detection unit The estimation model is corrected based on the results of machine learning of the relationship between the plurality of characteristic models. In the above aspect, the accuracy of the characteristic model can be improved.

[0113] (Supplementary Note 5) In a fifth aspect which is a preferred example of any of the first to fourth aspects, the characteristic difference is a difference between a first characteristic model based on one of the plurality of characteristic models and a second characteristic model based on a characteristic model subsequent to the first characteristic model. In this aspect, an abnormality in the water electrolysis stack can be suitably estimated.

[0114] (Supplementary Note 6) In a sixth aspect which is a preferred example of the fifth aspect, the first characteristic model is a first characteristic model among the plurality of characteristic models, or a characteristic model based on the first characteristic model. In this aspect, an abnormality in the water electrolysis stack can be suitably estimated.

[0115] (Supplementary Note 7) In a seventh aspect which is a preferred example of any of the first to sixth aspects, the water electrolysis system further comprises a notification unit which notifies an operator of information relating to an abnormality in the plurality of water electrolysis stacks based on a detection result from the detection unit. In the above aspect, the information relating to the abnormality in each water electrolysis stack can be notified to an operator.

[0116] (Supplementary Note 8) An eighth aspect of a preferred example of a water electrolysis system according to the present disclosure includes a plurality of water electrolysis stacks, a power supply that supplies DC current to each of the plurality of water electrolysis stacks, and a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks, wherein the detection unit estimates an abnormality in each of the plurality of water electrolysis stacks based on multiple sets of measured values ​​of the current and voltage of the water electrolysis stack during water electrolysis operation.

[0117] In the above-described aspect, an abnormality is estimated based on multiple sets of measurement values ​​of the current and voltage of the water electrolysis stack during water electrolysis operation, thereby making it possible to detect an abnormality in the water electrolysis stack while reducing a decrease in water electrolysis efficiency and deterioration of the water electrolysis stack. Here, a characteristic model can be calculated by fitting the multiple sets of measurement values ​​with an estimation model representing an electric circuit equation of the water electrolysis stack or an equation similar thereto. Therefore, an abnormality in the water electrolysis stack can be estimated based on the characteristic model even during water electrolysis operation. [Explanation of symbols]

[0118] 10...water electrolysis stack, 11...cathode, 11a...catalyst layer, 11b...diffusion layer, 12...anode, 12a...catalyst layer, 12b...diffusion layer, 13...anion exchange membrane, 14...first flow path member, 14a...first flow path, 15...second flow path member, 15a...second flow path, 16...first gasket, 17...second gasket, 20...power source, 21...DC power source, 30...temperature sensor, 40...supply device, 41...first supply mechanism, 41a...first container, 41b...first pump, 41c...first supply pipe, 41d...first discharge pipe, 42...second supply machine structure, 42a...second container, 42b...second pump, 42c...second supply pipe, 42d...second discharge pipe, 50...control device, 51...storage device, 52...processing device, 52a...detection unit, 52b...alarm unit, 52c...output control unit, 100...water electrolysis system, Ca...condenser, Da...characteristic data, Da-1 to Da-n...characteristic data, Db...abnormality information, F1...first surface, F2...second surface, K1...correction coefficient, K2...correction coefficient, K3...correction coefficient, K4...correction coefficient, MD-1...first characteristic model, MD-2...second characteristic model, R a …Resistance, R m ...resistance, ST1...step, ST2...step, ST3...step, ST4...step, ST5...step, ST6...step, ST7...step, ST8...step, V act …voltage, V cell …output voltage, V m …voltage, V ref …voltage source, i…input current, ΔV…characteristic difference.

Claims

1. a plurality of water electrolysis stacks; a power supply that supplies a direct current to each of the plurality of water electrolysis stacks; a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks, The detection unit For each of the plurality of water electrolysis stacks, generating a plurality of characteristic models by calculating, for each predetermined period, a characteristic model that approximates the response characteristics of the output voltage to a change in the input current of the water electrolysis stack based on the measured values ​​of the current and voltage of the water electrolysis stack; estimating an abnormality in the water electrolysis stack based on a characteristic difference between at least two of the plurality of characteristic models; Water electrolysis system.

2. the measured values ​​are a plurality of sets of measured values ​​of current and voltage of the water electrolysis stack, The detection unit calculating a characteristic model for each of the plurality of water electrolysis stacks by fitting the plurality of sets of measurement values ​​using an estimation model representing an electric circuit equation of the water electrolysis stack or an equation similar thereto; The water electrolysis system according to claim 1 .

3. a temperature sensor for detecting a temperature of at least one of the plurality of water electrolysis stacks; the detection unit corrects the estimation model based on the detection result of the temperature sensor. The water electrolysis system according to claim 2 .

4. The detection unit correcting the estimation model based on a result of machine learning of the relationship between the plurality of characteristic models; The water electrolysis system according to claim 2 .

5. the characteristic difference is a difference between a first characteristic model based on one of the plurality of characteristic models and a second characteristic model based on a characteristic model later than the first characteristic model; The water electrolysis system according to claim 1 .

6. the first characteristic model is a first characteristic model among the plurality of characteristic models, or a characteristic model based on the first characteristic model; The water electrolysis system according to claim 5 .

7. a notification unit that notifies information regarding abnormalities in the plurality of water electrolysis stacks based on the detection results of the detection unit. The water electrolysis system according to claim 1 .

8. a plurality of water electrolysis stacks; a power supply that supplies a direct current to each of the plurality of water electrolysis stacks; a detection unit that detects an abnormality in each of the plurality of water electrolysis stacks, The detection unit For each of the plurality of water electrolysis stacks, an abnormality in the water electrolysis stack is estimated based on a plurality of sets of measured values ​​of current and voltage of the water electrolysis stack during water electrolysis operation; Water electrolysis system.

Images