Method, device, and computer program of monitoring state of multicell type electrolytic tank

Abstract

To provide an improved method, device, computer program and system, for monitoring the state of a multicell type electrolytic tank.SOLUTION: An example of the present invention relates to a method, a device, a computer program and a system of monitoring the state of a multicell type electrolytic tank. The method includes step (110) of acquiring sensor information which at least includes information on the measured temperature of the individual cell of the multicell type electrolytic tank. The method includes step (130) of predicting an anticipated temperature value of each cell of the multicell type electrolytic tank for each cell of the multicell type electrolytic tank according to the sensor information. The method includes step (170) of providing a notice when the deviation of the anticipated temperature value of a certain cell from the measured temperature of the cell, coincides with the temperature deviation condition.SELECTED DRAWING: Figure 1a

Description

[Technical field]

[0001] Examples of the present invention relate to methods, apparatus, computer programs and systems for monitoring the condition of a multi-cell electrolyser. [Background technology]

[0002] An electrolyser is a device that uses electricity to drive otherwise non-naturally occurring chemical reactions, such as splitting compounds, such as water, into their constituent elements by electrolysis. For example, hydrogen electrolysers and chlor-alkali electrolysers are important tools for the production of hydrogen and chlorine / sodium, respectively. For large-scale applications, electrolysers often use many individual electrolyser cells, which are often arranged in a horizontal stack.

[0003] The temperature in an individual electrolyser cell is an important metric since it is considered to be important for the overall efficiency of the electrolyser. In addition, the temperature can indicate problems in the individual cells and can cause problems in the cells, which can be related to a loss of efficiency or can be caused by cell degradation and can lead to cell degradation. Furthermore, in the case of hydrogen electrolyser cells, leaks can cause external flameless combustion, which is difficult to detect. However, due to the corrosive environment in the electrolyser cells, the temperature of the cells is often not measured or is measured only inadequately. Summary of the Invention [Problem to be solved by the invention]

[0004] There may be a need for improved concepts for monitoring the condition of multi-cell electrolyzers. [Means for solving the problem]

[0005] The subject matter of the independent claims meets this need.

[0006] Various examples of the invention are based on the discovery that during normal operation it is usually possible to predict the temperature of the individual cells of a multi-cell electrolyser using, for example, knowledge of the physical and chemical properties of the cells and information about other process parameters such as the stack current and the individual cell voltages. Whereas in the case of normal operation of healthy cells the temperature prediction is accurate to begin with, the temperature prediction can be used to identify faults and degradation in the cells by comparing the predicted temperature with the measured temperature of the cells. This measured temperature can be obtained, for example, using a fiber optic temperature sensor. Deviations between the predicted and measured temperatures can be used to identify anomalies. In this case, a notification is provided to alert an operator of the multi-cell electrolyser of the anomaly. When the electrolyser is considered to be operating normally, the measured and predicted values ​​can be provided in a graphical format to allow the operator to visually check that the electrolyser is operating as expected. The proposed concept applies to various types of electrolyser, including electrolyser with monopolar cells and electrolyser with bipolar cells, as well as electrolyser with cells arranged in a series configuration and electrolyser with cells arranged in parallel, for example in a parallel stack.

[0007] Some aspects of the invention relate to a method for monitoring the condition of a multi-cell electrolyser. The method comprises obtaining sensor information, the sensor information including at least information about measured temperatures of individual cells of the multi-cell electrolyser. The method comprises predicting, for each cell of the multi-cell electrolyser, an expected temperature value for said cell of the multi-cell electrolyser based on the sensor information. The method comprises providing a notification when a deviation between the expected temperature of a cell and the measured temperature of this cell corresponds to a temperature deviation condition. In this way, an operator of the electrolyser is notified when the electrolyser is not operating as expected, for example due to cell deterioration or due to an event such as a flameless combustion caused by a hydrogen leak.

[0008] There are different ways to predict the temperature of the cells. One option is to derive the predicted temperature from the cell voltage of the respective cell and from the stack current by a model, such as an energy balance based model. For example, the sensor information may include information about the stack current of the multi-cell electrolyzer and about the individual cell voltages of the cells of the multi-cell electrolyzer. The predicted temperature value of a cell can be predicted based on the cell voltage of this cell and based on the stack current of the multi-cell electrolyzer. In particular, the predicted temperature value of a cell can be predicted based on an energy balance based on the cell voltage of this cell and based on the stack current of the multi-cell electrolyzer. In this way, a model based on the physical and chemical properties of the cell is used to predict the temperature of the cell.

[0009] Instead, the temperatures of other cells can be used as a reference. For example, the expected temperature value of a cell can be predicted based on the measured temperatures of one or more adjacent cells of a multi-cell electrolyser. Adjacent cells are expected to operate in a similar manner, such that any deviation indicates cell degradation or failure.

[0010] Alternatively, or in addition, the temperature of the cell can be predicted using an average value across other cells (e.g. across all cells of the electrolyser, or across a selected group of cells). For example, the predicted temperature value of a cell can be predicted based on the average or mean value of the measured temperatures of multiple cells of a multi-cell electrolyser. This can reduce the influence of a single temperature on the predicted temperature value.

[0011] In some cases, the expected temperature value of a cell can be predicted based on one or more reference cells of a multi-cell electrolyzer. In this way, the expected temperature can be determined using (only) healthy (e.g. new) cells, which facilitates identifying cells whose deviation from the reference indicates their age.

[0012] Generally, the one or more reference cells can be selected by a user, for example, based on which cell was last replaced. Alternatively, the one or more reference cells can be selected using an automated process. For example, the method can further include selecting the one or more reference cells based on sensor information. In this manner, no user interaction is required to select the one or more reference cells.

[0013] To avoid false alarms in the event of occasional abnormal measurements or to detect the occurrence of small deviations over time, instead of (or in addition to) using the temperature measurements directly, an integral or derivative of the temperature can be calculated and compared to the integral or derivative of the predicted temperature. For example, the predicted temperature of a cell can correspond to the integral or derivative of the predicted temperature of the cell with respect to time, voltage, or current. The deviation of the predicted temperature of a cell from the measured temperature of the cell can be determined as the difference between the integral or derivative of the predicted temperature of the cell and the corresponding integral or derivative of the measured temperature of the cell.

[0014] A major concern when operating hydrogen electrolysers is the occurrence of so-called flameless combustion, which is difficult to detect as it is invisible. However, flameless combustion can be detected by detecting a spike in the measured temperature of the cells. The method thus includes the step of providing a notification of a suspected flame if the deviation between the expected temperature value of a cell and the measured temperature of this cell exhibits a spike.

[0015] Another important parameter monitored in a multi-cell electrolyser is the cell voltage of the individual cells, i.e. the potential difference between the cathode and anode of the cell. Similar to the temperature monitoring, the expected voltage value can be used to compare with the measured cell voltage value to determine whether the cell is behaving as expected. For example, the sensor information can include information about the individual cell voltages of the cells of the multi-cell electrolyser. The notification can be provided if the deviation between the expected temperature value of a cell and the measured temperature value of this cell matches a temperature deviation condition and if the deviation between the cell voltage of this cell and the expected voltage value matches a voltage deviation condition. In addition to the temperature deviation, a voltage deviation can also indicate cell degradation or failure.

[0016] There are various ways of obtaining the expected voltage values. In some cases, the voltage values ​​can be estimated from the stack current and / or other parameters. For example, the sensor information can include information about the stack current of the multi-cell electrolyser and about the individual cell voltages of the cells of the multi-cell electrolyser. The method can include predicting, for each cell of the multi-cell electrolyser, expected voltage values ​​of the cells of the multi-cell electrolyser based on the sensor information. Alternatively, the expected voltage values ​​can be user defined. In other words, the method can include obtaining expected voltage values ​​of the cells of the multi-cell electrolyser as input.

[0017] As outlined above, a voltage deviation consistent with a voltage deviation condition may also indicate cell degradation or failure. Thus, the method may include providing a notification when a deviation between an expected voltage value of a cell and a measured voltage of the cell is consistent with a voltage deviation condition. This may be done separately from (e.g., in addition to) a temperature deviation based notification.

[0018] Generally, the individual cell voltages depend on the state of the cell and the current being used to power the cell. To further assess the state of the cell, the response of the cell to a load change, and in particular the time it takes to react to the load change, can be monitored. Thus, the expected voltage value can include an expected voltage and an expected time to reach the expected voltage in response to a load change.

[0019] The cell conditions change over time because the electrolysis performed by the cells causes degradation of the membranes of the individual cells. The resulting cell voltage depends on the cell conditions and can be modeled using a multi-parameter regression equation. For example, the expected voltage value can be based on a multi-parameter regression equation. The multi-parameter regression equation can include a first parameter that models an undesired reaction at the electrodes, a second parameter that models at least one of membrane ruptures, pinholes, and degradation, and a third parameter that models electrode degradation. The method can include determining the deviation between an expected voltage value of a cell and a measured voltage of the cell. The method can include identifying the contribution of the first, second, and third parameters to the deviation using a multi-parameter regression model. The equation has three parameters, one independent of the current, one linearly dependent on the current, and one logarithmically dependent on the current. By varying the current, the effect of different parameters on the expected voltage of the cell can be identified, which improves insight into whether and when a cell needs to be replaced or maintained.

[0020] As explained above in relation to the temperature values, the predicted and measured voltage values ​​also do not necessarily correspond to a single voltage. Instead, the respective voltages can be integrated or a derivative can be calculated. For example, the predicted voltage value of a cell can correspond to the predicted voltage integrated or differentiated with respect to time, temperature or current. The deviation between the predicted voltage value of a cell and the measured voltage of this cell can be determined as the difference between the corresponding integrated or differentiated predicted voltage of this cell and the measured voltage of this cell. This can help to avoid false alarms in the case of occasional abnormal measurements or to detect the occurrence of small deviations over time.

[0021] Another important factor is the evolution of the cell voltages over time. For example, the method can include, for each cell of a multi-cell electrolyser and / or for one or more reference cells, determining a previous voltage value of the cell based on the sensor information and comparing the previous voltage value with the current voltage measurement contained in the sensor information. In this way, the effect of degradation of each cell can be measured.

[0022] In some cases, notifications need to be provided to alert an operator of critical or potentially critical cell conditions or incidents. In addition to notifications, the measurements and / or calculations can be visualized, allowing the electrolyzer operator to draw his or her own conclusions. For example, the method can further include providing a visual representation of at least one of the sensor information, the predicted temperature value of each cell, the predicted voltage value of each cell, and the previous voltage value of each cell.

[0023] As outlined above, one option for measuring the temperature of the cells is to use a fiber optic temperature sensor. Thus, information about the measured temperatures of the cells of a multi-cell electrolyzer can be obtained from the fiber optic temperature sensor. With the fiber optic temperature sensor, the temperature of the cells can be measured using a single (or a few) fiber optic cables, which can reduce the effort for measuring the temperature.

[0024] Alternatively, information about the measured temperatures of the cells of a multi-cell electrolyser can be obtained from a camera-based temperature sensor, although this method may require a greater amount of computation since the temperatures must be derived from the image data.

[0025] One aspect of the invention relates to a computer program having a program code for performing the above method, when the computer program is run on a computer, on a processor or on a programmable hardware component.

[0026] Some aspects of the invention relate to a corresponding apparatus for monitoring the condition of a multi-cell electrolyzer, the apparatus comprising an interface circuit for acquiring sensor information from one or more sensors, and a processing circuit configured to carry out the above method.

[0027] Some examples of apparatus and / or methods will now be described, by way of example only, and with reference to the accompanying drawings in which: [Brief description of the drawings]

[0028] [Figure 1] Figure 1a shows a flow chart of an example of a method for monitoring the condition of a multi-cell electrolyser, and Figure 1b shows a schematic diagram of an example of an apparatus for monitoring the condition of a multi-cell electrolyser, and a system including such an apparatus. [Diagram 2] 2a and 2b are flow charts of examples of measurement and processing of process values ​​of cells of a multi-cell electrolyser. [Diagram 3] FIG. 1 is a schematic diagram of an example of an electrolyzer cell surrounded by a fiber optic cable. [Figure 4] FIG. 1 is a schematic diagram of an example electrolyzer cell having interwoven fiber optic cables. [Diagram 5] FIG. 1 is a schematic diagram of an example of a data processing system. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0029] Detailed Description Some examples will now be described in more detail with reference to the enclosed drawings. However, other possible examples are not limited to the features of these embodiments described in detail. Other examples may include modifications of features and equivalents, and alternatives of features. Furthermore, the terms used in this specification to describe specific examples should not limit further possible examples.

[0030] Throughout the description of the figures, the same or similar reference numbers refer to the same or similar elements and / or features, which may be the same or may be implemented in modified form while providing the same or similar functionality. The thickness of lines, layer thicknesses, and / or area sizes in the figures may be exaggerated for clarity.

[0031] When combining two elements A and B using "or", this should be understood as disclosing all possible combinations, i.e. A only, B only, as well as A and B, unless expressly specified otherwise in individual cases. Other phrases for the same combinations may be used: "at least one of A and B" or "A and / or B". This applies equally to combinations of more than two elements.

[0032] Where singular forms such as "a," "an," and "above" are used, and where the use of only a single element is not mandated, either explicitly or implicitly, additional examples may use multiple elements to achieve the same function. Where a function is described below as being achieved using multiple elements, additional examples may achieve the same function using a single element or a single processing entity. It will be further understood that "include," "including," "comprises," and / or "comprising," when used, describe specified features, numbers, steps, operations, processes, elements, components, and / or groups thereof but do not exclude the presence or addition of one or more other features, numbers, steps, operations, processes, elements, components, and / or groups thereof.

[0033] Various examples of the present invention relate to methods, devices and computer programs for monitoring the condition of a multi-cell electrolyzer. Optionally, the device may be part of a system that may include one or more sensors or processing entities for providing sensor information. In some examples, the system may include the multi-cell electrolyzer itself.

[0034] Figure 1a shows a flow chart of an example of a method for monitoring the condition of a multi-cell electrolyser 100 (shown in Figure 1b). The method comprises a step 110 of obtaining sensor information. The sensor information comprises at least information about the measured temperatures of the individual cells of the multi-cell electrolyser 100. The method comprises a step 130 of predicting, for each cell of the multi-cell electrolyser, an expected temperature value for said cell of the multi-cell electrolyser based on the sensor information. The method comprises a step 170 of providing a notification when a deviation between the expected temperature value of a cell and the measured temperature of this cell corresponds to a temperature deviation condition.

[0035] In the following, a short introduction is given of the various components contained or referred to in this specification. In Fig. 1b, a schematic diagram of an apparatus for monitoring the condition of a multi-cell electrolyzer and a system comprising such an apparatus is shown. The apparatus 50 comprises an interface circuit 52 for acquiring information from one or more sensors and a processing circuit 54 configured to execute the method of Fig. 1a. Optionally, the apparatus 50 further comprises a memory circuit 56 for temporarily or permanently storing information such as reference values, previous measurements, etc. The processing circuit 54 is coupled to the interface circuit 52 and to the optional memory circuit 56. The processing circuit 54, together with the interface circuit 52 (e.g. for exchanging information with the temperature processing entity 10 or the fiber optic temperature sensor 20, the cell voltage measurement entity 30, and / or the stack current measurement entity 40) and / or the optional memory circuit 56 (for storing information), provides the functionality of the apparatus 50. For example, the apparatus 50 can provide its functionality by the processing circuit 54 executing machine-readable instructions, which can be stored in the memory circuit 56.

[0036] The above method and device 50 and corresponding computer program are used to monitor the condition of the multi-cell electrolyzer 100. In the following, the monitoring of the condition of the multi-cell electrolyzer focuses on monitoring the temperature of the individual cells of the electrolyzer and subsequently the cell voltage of the individual cells of the electrolyzer. The temperature and the cell voltage of the individual cells both depend on the load of the electrolyzer, which in turn depends on the stack current supplied to the cells of the electrolyzer, i.e. the current supplied to the entire stack of cells. The sensor information 110 therefore comprises at least information on the measured temperature of the individual cells. In addition, the sensor information 110 may comprise information on the stack current of the multi-cell electrolyzer and on the individual cell voltages of the cells of the multi-cell electrolyzer, which information may also be used to monitor the multi-cell electrolyzer. Components of the sensor information 110 may be obtained (e.g. received, requested) from different sensors or processing entities of the system. For example, information on the measured temperature may be obtained from a temperature sensor (such as the fiber optic temperature sensor 20 shown in FIG. 1b). Information about the individual cell voltages may be obtained from a processing circuit 30 that performs or processes the cell voltage measurements. Information about the stack current may be obtained from a processing circuit 40 that performs or processes the stack current measurements. Typically, the processing entity 40 that performs or processes the stack current measurements is part of the control system of the electrolyser 100.

[0037] While stack currents and differential cell voltages are fairly easy to capture, capturing individual temperatures can require more effort. In the following, two different methods are described.

[0038] In a first method, a fiber optic temperature sensor is used to provide information about the temperature of the individual cells. In other words, information about the measured temperatures of the cells of a multi-cell electrolyser can be obtained from the fiber optic temperature sensor. This method is illustrated in Figs. 1b, 3, 4 and 5 and described below. In the following, it is assumed that the measurements of the fiber optic temperature sensor are processed by a processing entity 10 and provided to the device 50. However, the processing performed by the processing entity 10 can instead be performed by the device 50 or by the fiber optic temperature sensor 20. For example, the processing entity 10 can be realized similarly to the device 50 and includes an interface for communicating with the fiber optic temperature sensor and the device 50, a processing circuit for performing calculations and a memory circuit for storing information such as mappings.

[0039] Generally, a fiber optic temperature sensor may be used to provide temperature sensor information including temperature values ​​measured at multiple sections of a fiber optic cable 25 (shown in FIG. 1b) used by the fiber optic temperature sensor. The temperature sensor data may be used to calculate at least one temperature for each cell of a multi-cell electrolyzer based on the temperature values ​​measured at the multiple sections.

[0040] A first method is based on using optical fiber temperature measurements. In general, such optical fiber temperature measurements are based on the effect of temperature on the optical fiber cable 25. For example, depending on the temperature, the reflection or scattering of light or sound may be different in different sections of the optical fiber cable. Various techniques exist for realizing such optical fiber temperature sensors. For example, the optical fiber temperature sensor may be a fiber Bragg grating sensor, i.e. a sensor based on detecting the temperature-dependent change in the Bragg wavelength. Alternatively, the optical fiber temperature sensor may be a Raman scattering-based temperature sensor, which is based on the temperature-dependent change of the inelastic scattering on optical phonons. Alternatively, the optical fiber temperature sensor may be an interferometric optical fiber temperature sensor (intrinsic or extrinsic), i.e. a temperature sensor based on the temperature-dependent change of the length of an optical resonator. Alternatively, the optical fiber temperature sensor may be a Brillouin scattering-based distributed temperature sensor (based on the temperature-dependent scattering of acoustic phonons). However, the proposed concept is not limited to a specific realization of an optical fiber temperature sensor.

[0041] Fiber optic temperature sensors have in common that they allow the measurement of temperature in user-defined or system predefined intervals. For example, in one example of a fiber optic temperature sensor used to realize the proposed concept, the fiber optic temperature sensor provides measurements in 1 meter increments, i.e., a separate temperature value for every meter in a multi-kilometer length of fiber optic cable. In some examples, temperature value measurements are taken at specific points every meter, while in other examples, an average temperature value is measured over a 1 meter stretch or a specific stretch of the fiber optic cable. These temperature values ​​can then be used to calculate the temperature of the individual cells, depending on how the fiber optic cable is arranged in the electrolyzer cells. Thus, the process begins with obtaining temperature sensor information having temperature values ​​from the fiber optic temperature sensor. The temperature sensor information includes a plurality of temperature values, i.e., one temperature value per interval. For example, the multiple intervals can be equidistant intervals (e.g., the 1 meter intervals mentioned above), for example, defined by the fiber optic temperature sensor 20. Instead, the sections can be custom sections that are adjusted to the geometry of the cells of the multi-cell electrolyzer and to the pattern used to place the fiber optic cable 25 in the cells (or in the cells). The temperature values ​​can then be used to calculate the temperature for the individual cells. For this purpose, information about which temperature value in which cell of the electrolyzer is to be obtained can be used. In the following, this information is represented as a correspondence between the sections of the fiber optic cable and the cells of the multi-cell electrolyzer. For example, the sections can be associated with the cells of the multi-cell electrolyzer. For example, a section (or a subset of the sections, in case a section is between two cells or is not located at a cell) can be associated with one of the cells. Thus, each cell can have at least one (preferably multiple) sections associated with it, and this correspondence can then be used to calculate the temperature of the cell. This correspondence can be used to perform a selection as to which temperature value is used to calculate the temperature of at least one of the cells.In other words, one or more temperature values ​​can be selected for each cell of the multi-cell electrolyser based on the mapping of the zones to the cells of the multi-cell electrolyser. The selected temperature values ​​can be used to calculate at least one temperature of the cell. For example, in some cases a single temperature can be calculated for each cell based on the zones and thus the temperature values ​​associated with the cell, e.g. using the average or median of these temperature values. Alternatively, multiple temperatures can be calculated for each cell based on the zones and thus the temperature values ​​associated with the cell. Calculating multiple temperatures for each cell is preferred in order to detect flameless combustion in different sections of the cell.

[0042] The correspondence between sections and cells depends on the way in which the fiber optic cable is arranged in the respective cell or in the cell. In general, two general arrangement options can be distinguished - these are either attached to the outside of the cell (for example attached to the outer shell or surface of the cell or between the cell and the insulation) or integrated in the cell. For example, the fiber optic cable can be attached to the outer shell or surface of a multi-cell electrolyser. This method is more suitable for retrofitting existing electrolysers or for use with third party electrolysers. Alternatively, the fiber optic cable can be integrated in the cell, in the cell wall, in the heating pipes or in the cooling pipes of the multi-cell electrolyser. This requires more effort. However, the closer proximity allows the measurements to be more accurate than measurements from outside the cell. In both cases, the fiber optic cable can cover several cells of the multi-cell electrolyser. For example, a single fiber optic cable may be sufficient to measure the temperature in all the cells of a multi-cell electrolyser. Alternatively, several fiber optic cables can be used, with several sections running through the several fiber optic cables. In this case, the complexity of the correspondence between intervals and cells may increase slightly.

[0043] The correspondence between the sections and the cells depends on the geometry of the electrolysis cell and the pattern used to place the fiber optic cable in the cell or cells. Figures 3 and 4 show two options for placing the fiber optic cable. Figure 3 (and Figure 1b) shows a fiber optic cable surrounding the cells of the electrolysis cell. In other words, the fiber optic cable can be placed in a pattern that surrounds each cell in a continuous manner around the periphery of the cells (e.g., between the cell wall and the insulation). For example, as shown in Figure 1b, the cells of the electrolysis cell can be arranged in a horizontal stack (or alternatively, a vertical stack). The fiber optic cable can be placed in a pattern that substantially surrounds each individual cell of the stack, for example, such that each cell is surrounded (e.g., surrounded, enveloped) at least once. In some cases, each cell can be surrounded multiple times, e.g., multiple small areas per cell, to determine the average temperature over different positions per cell, or a two-dimensional temperature profile of the cell can be determined. For example, a spiral or snail pattern (e.g., having mostly straight lines or mostly curved lines (not shown)) can be used. The spiral or snail pattern can be repeated multiple times per cell, with the same (or similar) pattern being used for the other cells. Alternatively or additionally, the serpentine / zigzag pattern can be repeated multiple times per cell, with the same (or similar) pattern being used for the other cells. For example, if the diameter of the electrolyzer is, for example, less than 1 meter (e.g., less than the measurement accuracy of the fiber optic temperature sensor), or if it is necessary to measure the temperature of different areas on the exterior of the electrolyzer, wiring can be added, where, for example, one portion of an individual electrolyzer cell can be surrounded or covered with a circle of wiring two or three times (e.g., by zigzagging up and down over 20 cm, or by using a spiral / snail pattern) before moving to the next spot on the same cell that is further from the exterior wall.In a more general sense, the fiber optic cable is arranged in a pattern (i.e., around a cell or in a serpentine pattern) that includes sub-patterns, such as a spiral / snail sub-pattern or a serpentine / zigzag sub-pattern, that are repeated multiple times within the pattern. Each of the sub-patterns may be arranged in exactly one of the cells (i.e., a sub-pattern may not cover multiple cells). Each cell may be covered by multiple repetitions of the sub-pattern, for example, by repeating a series of sub-patterns around the periphery of each cell. Thus, the correspondence of sections to cells may be based on such circular patterns, or combinations of circular and sub-pattern repetitions, for example, based on knowledge of which sections are to be placed in which cells of an electrolytic cell. This may be calculated using three-dimensional computer-aided design drawings, or may be determined manually using exterior markings on the fiber optic cable.

[0044] Instead, a weave or serpentine pattern can be used, as shown in Figures 4 and 5. In other words, the fiber optic cable can be arranged in a serpentine pattern, whereby the fiber optic cable repeatedly runs along a series of cells of a multi-cell electrolyzer, turns around, runs backward along the series of cells, turns around again, and runs along the series of cells again. With a serpentine pattern, the fiber optic cable passes through each cell multiple times. Depending on the size of the cell and the size of the section, a temperature value can be associated with the cell each time the fiber optic cable passes through the cell (or for a subset of the number of passes). As with the example shown in relation to the circular pattern, the spiral / weave pattern can be combined with a smaller spiral / snail pattern, or a smaller serpentine / zigzag pattern, to cover a portion of an individual electrolyzer cell multiple times. Thus, again, the association of sections with cells can be based on such a meander or weave pattern, e.g., based on knowledge of which sections are to be placed in which cells of the electrolyzer. This can be calculated using three-dimensional computer aided design drawings, or can be determined manually using exterior markings on the fiber optic cable.

[0045] As outlined in relation to the circular pattern, in some examples, multiple temperature values ​​can be obtained for each cell, e.g., using multiple sub-patterns, to generate a two-dimensional temperature profile for each cell. For example, the method can include generating a two-dimensional temperature profile for each cell based on the temperature values ​​and based on a correspondence between the multiple zones and the cells of the multi-cellular electrolyzer. In this case, the correspondence between the multiple zones and the cells can include a correspondence between the multiple zones and a plurality of points on each cell of the multi-cellular electrolyzer, e.g., a plurality of points in a two-dimensional grid of points on each cell of the electrolyzer.

[0046] In a second method, information about the measured temperatures of the cells of the multi-cell electrolyser may be obtained from a camera-based temperature sensor, for example an infrared camera sensor. In this case, the pixels of the (infrared) image data provided by the camera-based temperature sensor may be associated with the cells of the multi-cell electrolyser, for example by the processing entity 10 or the device 50. The sensor information may thus comprise the (infrared) image data, or cell temperatures derived from the infrared image data.

[0047] The sensor information includes information on the measured temperature of the individual cells and, optionally, on the stack current and / or the individual cell currents, which is then used to predict the temperature values ​​of the cells of the multi-cell electrolyser. In particular, a separate temperature value is predicted for each cell of the multiple cells of the electrolyser. This prediction is performed using different methods. For example, the physical and chemical properties of the cells can be used together with the cell voltage and the stack current to predict the temperature of the cell. In other words, the expected temperature value of the cell can be predicted based on the cell voltage of the cell (based on the sensor information) and based on the stack current of the multi-cell electrolyser. In particular, this can be done by examining the energy balance of the cell. In other words, the expected temperature value of the cell can be predicted based on the energy balance based on the cell voltage of the cell and based on the stack current of the multi-cell electrolyser. For example, electrochemical and thermal models can be used to calculate the hydrogen prediction and the oxygen prediction, the temperature rise caused by the electrolysis, to predict the temperature of the individual cells. In some cases, the prediction of the expected temperature value of the cell based on the cell voltage of the cell and based on the stack current of a multi-cell electrolyser can take into account the (known or measured) cell conditions with respect to undesired reactions at the electrodes, membrane ruptures, pinholes, degradation, and electrode degradation.

[0048] Alternatively or additionally, the temperature of a cell can be predicted using the temperature of other cells. For example, since a multi-cell electrolyser is driven by a single stack current, all cells are expected to have similar temperatures, with small variations due to ageing or external influences. The predicted temperature can therefore be predicted based on the (measured) temperatures of one or more other cells. For example, the predicted temperature value of a cell can be predicted based on the measured temperatures of one or more adjacent cells of the multi-cell electrolyser, i.e. one or more cells that are located directly adjacent to the cell whose temperature is to be predicted. For example, a single cell can be used to predict the temperature of the adjacent cell. Alternatively, if two adjacent cells are available, both can be used. Alternatively, one or more cells that are known (or assumed) to be healthy can be used as reference cells, for example because they have been recently replaced or because they have been shown to have a "healthy" temperature (or a "healthy" cell voltage). In other words, the predicted temperature value of a cell can be predicted based on one or more reference cells of the multi-cell electrolyser. The one or more reference cells may be user-defined reference cells, or may be determined automatically based on the age of the cell (using the most recently added cell as the reference cell), or based on sensor information determined for the cell. For example, as further shown in FIG. 1a, the method may further comprise a step 120 of selecting one or more reference cells based on the sensor information, e.g., by a selection criterion. For example, after selecting one or more reference cells, the method may proceed (directly) to a step 135 of determining the deviation between the expected temperature value and the measured temperature value, a step 150 of determining the deviation between the expected voltage value and the measured voltage value, and / or a step 165 of comparing a previous voltage value with a current voltage measurement. For example, the selection criterion may relate to the difference between the measured temperature (or measured voltage) and a user-defined reference temperature (or user-defined reference voltage), the difference between the measured temperature (or measured voltage) and the average temperature (or average voltage) over all cells, or the difference between the measured temperature (or measured voltage) and the minimum measured temperature (or minimum measured voltage) over all cells.In either case, the prediction of the temperature of the cell may not be based on a single other temperature measurement, but on multiple other temperature measurements, if available. For example, the expected temperature value of the cell may be predicted based on the average or mean of the measured temperatures of multiple cells of a multi-cell electrolyser (e.g. of all the cells, of all the other cells, of a reference cell, or of adjacent cells).

[0049] In the previous description, it was assumed that the "temperature value" corresponds to a temperature, for example, a measured or predicted temperature. In some cases, the temperature can be integrated over time (or current, or voltage) to filter out spurious fluctuations in the measurements, and the respective integrals can be compared with each other. In other words, the predicted temperature value of a cell can correspond to the integral of the predicted temperature of the cell over time, and the deviation of the predicted temperature value of a cell from the measured temperature of the cell is determined as the difference between the integral of the predicted temperature of the cell and the corresponding integral of the measured temperature of the cell. In this way, small differences in temperature that persist over a long period of time can be identified, while spurious events (such as invalid measurements) are smoothed out (averaged). Similarly, the first or second derivative of each temperature curve can be taken to identify trend differences between different cells. In other words, the predicted temperature value of a cell can correspond to a derivative (e.g., a first or second derivative) of the predicted temperature of the cell, and the deviation between the predicted temperature of a cell and the measured temperature of the cell is determined to be the difference between the derivative of the predicted temperature of the cell and the corresponding derivative of the measured temperature of the cell. An example of such a comparison is described in connection with the "Fifth Model" described in connection with FIG. 5.

[0050] The proposed concept provides a notification if the deviation between the predicted temperature value of a cell and the measured temperature of the cell matches a temperature deviation condition. As the basis for the step 170 of providing a notification, the method includes a step 135 of determining the deviation between the predicted temperature value of a cell and the measured temperature of the cell to evaluate the temperature deviation condition. The step of determining the deviation includes comparing the respective temperature values, e.g., temperature or the integral or derivative of temperature with respect to time (or current or voltage). The result of the comparison is then evaluated for the temperature deviation condition. In some cases, the temperature deviation condition can be a threshold and a notification is provided if the threshold is exceeded or not (depending on the temperature deviation condition). In some cases, the temperature deviation condition can include a range of acceptable deviations and / or one or more ranges of deviations that lead to providing a notification. If the temperature deviation condition is met, a notification is provided. For example, the step of providing a notification can include providing a visual indication by a display device and / or the step of providing a notification can include providing a notification packet to a control system or to one or more client devices, the notification packet including the notification. For example, the notification may include an alert intended for an operator / user of the electrolyzer.

[0051] Below are some examples of deviations, temperature deviation conditions, and other triggers for providing notifications. For example, at least some of the following examples are described in more detail in relation to FIG. 5, where various "models" are introduced, which serve as the basis for notifications and / or visual representations of different values. In the context of the present invention, models are mathematical functions or computational tasks that receive one or more input values ​​and include calculations (such as prediction of temperature (or voltage) values, and optionally comparison and evaluation of deviation conditions). In other words, the models introduced in relation to FIG. 5 are algorithms, computational tasks, or mathematical functions for processing signal information.

[0052] For example, in the first model, the individual cell voltages, the individual cell temperatures and the stack current are measured and plotted on a graph for manual evaluation by an operator of the electrolyser. Thus, the method may include a step 175 of providing a visual representation of at least one of the sensor information, the predicted temperature value of each cell, the predicted voltage value of each cell and the previous voltage value of each cell.

[0053] The second and third models compare individual cell voltages to each other, to, for example, one or more adjacent cells, to one or more reference cells, or to the average or mean temperature of the adjacent cells, the reference cells, or all of the cells.

[0054] In the 11th model, H 2 To detect leaks (and possible unseen flameless combustion), any measured high temperature spikes are 2 A flame may be assumed and an alarm may be generated. In other words, the method may include a step 170 of providing notification of a suspected flame if the deviation between the expected temperature value of a cell and the measured temperature of this cell exhibits a spike. In this context, a spike is a change in the measured temperature over a predefined time interval that is greater than a threshold value, for example, greater than 10° C. per minute.

[0055] In addition to the cell temperature value, the cell voltage is another main indicator of the health of the cells. Thus, an expected voltage value can be determined for each cell and the measured cell voltage values ​​can be compared with the expected voltage value to determine the voltage deviation. This expected voltage value can be determined in a similar manner to the determination of the expected temperature value. Thus, the method can include a step 140 of predicting an expected voltage value of the multi-cell electrolyzer for each cell of the multi-cell electrolyzer based on sensor information. For example, the expected voltage value can be determined based on the stack current of the electrolyzer and based on the above-mentioned formula V=A+B·I+C·log(I), where A represents undesired reactions at the electrodes, B represents membrane ruptures, pinholes, deterioration and C represents electrode deterioration. Alternatively or in addition, the expected voltage value can be determined based on the voltage values ​​of one or more other cells, for example the voltage values ​​of one or more adjacent cells, the voltage values ​​of one or more reference cells, the voltage values ​​of one or more cells with a similar age or the voltage values ​​of all the cells. An average voltage value or a median voltage value between two or more other cells can be used for this purpose. As an alternative to determining the voltages of individual cells, the voltage values ​​can be user defined. In other words, the method can include a step 145 of obtaining as input (by an operator / user of the electrolyzer) expected voltage values ​​of the cells of the multi-cell electrolyzer. As with the temperature values, the voltage values ​​can correspond to a single voltage or to the integral or derivative of the voltage with respect to time, current or temperature. For example, the expected voltage value of a cell can correspond to the integral or derivative of the expected voltage of this cell with respect to time (or current or temperature). In other words, the expected voltage value of a cell can correspond to the integral or derivative of the expected voltage of this cell with respect to time, temperature or current. The deviation between the expected voltage value of a cell and the measured voltage of this cell can thus be determined as the difference between the integral or derivative of the expected voltage of this cell and the corresponding integral or derivative of the measured voltage of this cell.

[0056] For example, a notification can also be provided in step 170 if the deviation of a cell voltage from the expected voltage value of this cell corresponds to a voltage deviation condition. For example, in the fourth model, a rapid drop of an individual cell voltage below a predefined threshold can be detected, thus indicating a membrane problem, and a notification can be provided if a rapid drop is detected (relative to the expected voltage value of this cell or to a previously measured voltage value). In this context, a rapid drop in voltage is a change in the measured voltage over a predefined time interval that is greater than a threshold, for example, greater than 1 V per minute. Such a rapid drop in an individual cell voltage can be determined relative to the expected voltage value or relative to a previously measured voltage of this cell or of a reference cell. Thus, as further shown in FIG. 1a, the method can include, for each cell of the multi-cell electrolyzer and / or for one or more reference cells, a step 160 of determining a previous voltage value of the cell based on the sensor information, and a step 165 of comparing the previous voltage value with a current voltage measurement (of the same cell or of any cell in case a previous voltage value is determined for the reference cell) contained in the sensor information. For example, a notification may be provided upon detecting a rapid drop between a current voltage value and an expected voltage value, or between a current voltage value and a previous voltage value (e.g., according to a predetermined time interval).

[0057] The fifth and sixth models use the integral and derivative of each voltage with respect to time and current, respectively. In other words, the deviation of a cell's expected voltage value from its measured voltage can be determined as the difference between the integral or derivative of the expected voltage of the cell and the corresponding integral or derivative of the measured voltage of the cell. For example, a notification can be provided if the voltage deviation of the integral is greater than a predefined average (tolerable) voltage deviation over a predefined period, e.g. 10 mV per day. In this way, small deviations over a period of time are visible and individual deviations do not cause false alarms. With respect to the differential voltage, anomalies in the voltage deviation between different cells or between a cell and a reference cell can be detected and a notification can be provided. This can be used to detect anomalies in the dynamic behavior between the cells.

[0058] Electrolyzers are dynamic systems with different states and phases, such as a shutdown phase and a startup phase. Shutdown can be defined as 0% current and startup can be defined as a current rising from 0% to 20% in less than 60 minutes. The time it takes for an individual cell to reach a voltage plateau (i.e., steady state) in response to such a condition is of particular interest to identify the state (and especially age-related effects) of the individual cells. Thus, in some examples, the expected voltage value can include an expected voltage and an expected time to reach the expected voltage in response to a load change. As described in relation to the seventh model, the time it takes a single cell to reach a complete startup or complete shutdown can be determined, which time provides an indication of the cell efficiency (longer time equates to higher efficiency). Once a startup or shutdown is detected, it can be determined for each individual cell the time t it takes for the cell to reach a voltage level that is a predefined voltage level while measuring the voltage. This voltage level can be defined as a simple threshold, with the first derivative being zero, or the second derivative being zero. This time t can be compared to an expected time to reach an expected voltage in response to a load change. The expected time can be derived from the time one or more reference cells take to make the same transition, or the expected time can be user-defined. If the deviation between the time it takes a cell to reach an expected voltage in response to a load change and the expected time matches a corresponding voltage deviation criterion, a notification can be provided. Details of this process are described in conjunction with the ninth model, which is described in conjunction with FIG. 5.

[0059] In many cases it is useful to know how each cell degrades over time. For this purpose, the method may also include a step 160 for each cell of the multi-cell electrolyser, determining a previous voltage value for that cell based on the sensor information and comparing the previous voltage value with the current voltage measurement contained in the sensor information. This difference may be used to measure degradation over time, as described in relation to the twelfth model described in relation to figure 5. Similarly, the difference between the current voltage measurement and a reference voltage, either user / operator defined or derived from a reference cell, may be calculated to determine the deviation of the individual cells from the desired reference voltage.

[0060] As outlined above, the cell voltage depends primarily on two factors - the cell condition and the stack current. This is reflected by the multi-parameter regression formula / equation mentioned above: V=A+B·I+C·log(I), where V represents the cell voltage, A represents the undesired reactions at the electrodes, B represents membrane ruptures, pinholes, and degradation, and C represents electrode degradation. Once the parameters A, B, and C are known (or at least estimated), this voltage can be used to calculate the expected voltage values ​​for different cells. In other words, the expected voltage values ​​can be based on a multi-parameter regression formula that includes a first parameter that models the undesired reactions at the electrodes, a second parameter that models at least one of membrane ruptures, pinholes, and degradation, and a third parameter that models electrode degradation. As is clear from this formula / equation, one of these parameters is independent of the current I, while one is linearly dependent and one is logarithmically dependent. By varying the current, the individual contributions of these parameters can be determined. For example, the method can include a step 150 of determining the deviation between the expected voltage of a cell and the measured voltage of this cell, and identifying the contributions of the first, second and third parameters to this deviation using the multi-parameter regression equation described above, for example by varying the current / load and calculating the contributions of the first, second and third parameters to this deviation based on the response of the cell to the change in current. The result of such parameterization and monitoring it over time is that the changes in the parameters can be observed, making it possible to ascertain the type of deterioration or failure occurring in the cells of the electrolyser. More details of this process are given in relation to the eighth model described in relation to figure 5.

[0061] In some examples, both temperature deviation and voltage deviation can be used as a common trigger to trigger providing a notification: if the deviation between the predicted temperature value of a cell and the measured temperature of the cell matches a temperature deviation condition, and if the cell voltage of the cell and the predicted voltage value match a voltage deviation condition, a notification can be provided.

[0062] Although the present invention is mainly concerned with hydrogen and chlor-alkali electrolysers, the proposed concept is not limited to a specific type of electrolyser. For example, the proposed concept can be used for different types of (hydrogen) electrolysers, such as PEM (Proton Exchange Membrane) electrolysers, alkaline electrolysers, solid oxide electrolysis-based electrolysers, or anion exchange membrane water electrolysis-based electrolysers. The proposed concept applies to various types of electrolysers, including electrolysers with monopolar cells and electrolysers with bipolar cells, as well as electrolysers in which the cells are arranged in a series configuration and electrolysers in which the cells are arranged in parallel, e.g. in a parallel stack. Monopolar electrolysers, also known as "tank-type" electrolysers, use a porous separator (e.g. a membrane) to keep the positive and negative electrodes separated. These electrodes are linked in parallel with each other and placed in a single electrolytic bath to form a cell. Multi-cell electrolysers can be built up by connecting the cells in series. Bipolar electrolysers use metal sheets or bipoles to connect adjacent cells in series. For example, one side of a bipole can be coated with a negative electrode electrocatalyst while the other side can be coated with the positive electrode electrocatalyst of the next cell. In the case of a bipolar cell, the total voltage is a combination of the cell voltages, resulting in a module that operates at higher voltages and lower currents than monopolar, tank-type designs. To form a larger multi-cell electrolyzer, multiple modules (each with multiple cells) are connected in parallel to increase the current. In general, an electrolyzer uses electricity to drive chemical reactions that would otherwise not occur naturally. In the context of the present invention, an electrolyzer is a device that splits compounds such as water into their constituent elements such as hydrogen and oxygen by electrolysis.

[0063] The interface circuitry 12 may represent inputs and / or outputs for receiving and / or transmitting information within a module, between modules, or between modules of different entities, and this information may be digital (bit) values ​​according to a specified code. For example, the interface circuitry 12 may include interface circuits configured to receive and / or transmit information.

[0064] For example, the processing circuitry 14 can be implemented using any means for processing, such as one or more processing units, one or more processing devices, processors, computers, or programmable hardware components operable by appropriately adapted software. In other words, the described functionality of the processing circuitry 14 can also be implemented in software, which is thus executed on one or more programmable hardware components. Such hardware components can include general-purpose processors, digital signal processors (DSPs), microcontrollers, etc.

[0065] For example, storage device 16 may comprise at least one member of the group of computer readable media such as a magnetic or optical storage medium, e.g., a hard disk drive, a flash memory, a floppy disk, a Random Access Memory (RAM), a Programmable Read Only Memory (PROM), an Erasable Programmable Read Only Memory (EPROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), or a network storage device.

[0066] Further details and aspects of the systems, and methods, apparatus and computer systems for monitoring the condition of a multi-cell electrolyser are described in relation to the proposed concepts or one or more of the examples described above or below (e.g. Figures 2a-5). The systems, and methods, apparatus and computer systems for monitoring the condition of a multi-cell electrolyser may comprise one or more additional optional features corresponding to one or more of the proposed concepts or one or more of the examples described above or below.

[0067] Various examples of the present invention relate to performance monitoring of individual electrolyser cells based on V (cell voltage), C (stack current) and T (temperature).

[0068] The present invention relates to the use of chlor-alkali (of various kinds), H 2 The present invention relates to monitoring electrolytic cells, such as electrolyzers or other types of electrolytic cells. In one implementation, the following hardware can be used to capture process-related data (current, individual cell voltages, and individual cell outer surface temperatures). For example, current measurements can be performed by a local control system (in the electrolyzer) (current is typically measured and / or controlled by the control system). Alternatively, a separate system for measuring stack current can be used. A cell voltage measurement system can be used to accurately measure individual cell voltages. A temperature measurement device based on optical fiber temperature measurements can be used to measure the outer surface temperatures of individual electrolyzer cells with sufficient accuracy. The optical fiber cable of the optical fiber temperature sensor can be arranged, e.g. wrapped, in various ways, e.g. vertically, horizontally, etc. Alternatively, the optical fiber-based temperature measurement device can be replaced by an infrared camera or some other temperature measurement device that measures the individual cell temperatures. An additional alternative is to measure the cell internal temperature. However, this is complicated due to the corrosive environment inside the cells.

[0069] The captured data may be received by a monitoring server on which the electrolyser performance monitoring system runs, as shown for example in Figures 2a, 2b, 3, 4 and 5.

[0070] 2a and 2b show a flow chart of an example of a flow for measuring and processing process values ​​of cells of a multi-cell electrolyzer. The flow includes data acquisition 210 (of temperature measurements) by fiber optic cables or other means, and conversion of the data to temperature values ​​220. The flow further includes acquisition of individual cell voltages 230, and stack current acquisition 240. In block 250, the temperature, cell voltage, and stack current data are processed. For example, as shown in FIG. 2b, the process 250 can include steps of comparing individual cell temperatures and voltages to modeled cell temperatures and voltages (250a), comparing individual cell temperatures to optimal cell temperatures (250b), monitoring individual cell temperatures and voltages over time (250c), and checking for high temperature spikes that may indicate a flame (250d). Appropriate follow-up actions can then be taken (step 260).

[0071] In Fig. 3, a schematic diagram of an example of electrolyzer cells of a multi-cell electrolyzer 300 is shown, which are surrounded by a fiber optic cable 25 used to measure the individual cell temperatures. Each of the electrolyzer cells comprises a cathode and an anode (shown in different patterns), and the cell voltage (V1...Vn) is measured between the cathode and anode of each cell. In addition, a current measurement probe 40a is used to measure the stack current of the multi-cell electrolyzer. The current measurement probe 40a can be part of the control system of the electrolyzer. In Fig. 4, the fiber optic cable 25 is arranged in a woven or serpentine pattern instead of surrounding the cells. In Fig. 4, a schematic diagram of an example of an electrolyzer cell with a woven fiber optic cable is shown.

[0072] The temperature values, cell voltages and stack current can be evaluated by one or more processing units or entities. Figure 5 shows a schematic diagram of an example of a data processing system comprising such processing units or entities. In Figure 5, the process values ​​of the multi-cell electrolyser 500 are captured by processing units 10, 30 and 40, where processing unit 10 converts the measurements (measured by the optical fiber cable 25) into temperature values, processing unit 30 captures the individual cell voltages and processing unit 40 captures the stack current. The temperatures, voltages and stack currents are processed by a processing entity 50 (e.g. a personal computer or a server) according to the concepts presented in connection with Figures 1a and / or 1b and / or according to the concepts or models below.

[0073] Monitoring electrolyser performance can be beneficial for a variety of reasons, such as overall safety (to prevent or detect flames or to prevent explosions), energy costs (which are lowest at optimal operating temperatures, whereas a malfunctioning cell is expensive to keep operating), and to make an overall assessment of whether or when maintenance is (or is not) required. These objectives can be met, inter alia, using individual electrolyser cell-based monitoring.

[0074] Algorithms and functions run on the monitoring entity (e.g., a monitoring server) that are able to monitor, on a continuous basis, the performance, maintenance needs, and safety of the individual electrolyzer cells. Various examples of the present invention are based on a combination of electrochemical models for the electrolyzer (based on physical properties and literature) and continuously measured values ​​of cell voltage, temperature, and stack current.

[0075] The electrochemical process in an individual cell can essentially be described by the following equation, which also includes coefficients describing certain performance indicators: V = A + B I + C log(I) where A is the undesired reaction at the electrode, B represents membrane ruptures, pinholes, and degradation, and C represents electrode degradation. This equation can be expanded with a heat balance, where the cell temperature is one of the major components.

[0076] By measuring cell voltages, stack currents and cell temperatures, a comparison of the measured values ​​with modeled values ​​(e.g. expected temperature / voltage values) can be made to detect anomalies. However, in more detailed forms, specific differences in the measured individual cell voltages or temperatures and / or the overall performance of the cells can also be determined. Based on the measured data and the models described below, the proposed concept allows the individual cell performance and failure mechanisms of specific cells to be derived.

[0077] In the following, several models are described, one or more of which may be used by the entire system. For example, one or more of the following models may be used as part of the method or computer program of FIG. 1a and / or calculated by the device of FIG. 1b. In the context of the present invention, the models are the tasks of the computer, and they may have different kinds of outputs, such as a graphical representation of the calculation or notification. For example, in a first model, the individual cell voltages, the individual cell temperatures and the stack current may be measured and plotted in a graphical form for manual evaluation by the electrolyzer operator. In a second model, the individual cell temperatures may be compared with each other. Cells with relatively large deviations from what is defined as healthy are alarmed (i.e. an alarm notification is provided). This is achieved by creating and updating a table (or other type of data structure) of each cell and its temperature. In a third model, the individual cell temperatures may be compared with adjacent cells. In general, a malfunctioning cell will produce a temperature different from that expected. The results may be presented in a tabular or graphical form, making it easy to detect suspect cells.

[0078] A fourth model detects a rapid drop in individual cell voltage below a predefined threshold, which would therefore indicate a membrane problem and, in combination with an increase in cell temperature, would trigger an alarm. Such behavior can be monitored and a follow-up alarm can be generated if such behavior occurs.

[0079] In the fifth model, a number of cells are defined as reference cells that are known to behave in a healthy mode. The voltage can be measured / monitored over a certain period of time (e.g., 1, 2 days). The voltage can be integrated over time and the integral value can be specified as the reference value for a certain time frame. The integral value can now be used as the reference value. For each individual cell, the cell voltage can be measured and integrated over a similar time frame, e.g., 1-2 days. The integral value of the individual cell can then be compared to the reference value. An alarm can be generated if the difference is greater than a predefined average (tolerable) deviation over a predefined period of time, e.g., 10 mV per day. In this way, small deviations over long periods of time are visible and individual one-time deviations do not generate false alarms. To check the value of the reference cell, a user-defined reference value can be entered into the system and a calculation method similar to that described above can be used. This can then be used to check the calculated reference value in a similar manner as described above. The advantage of choosing a limited set of metrics is that a better overall indication of the health of the individual cells compared to the possible health of the cells is achieved compared to integrating all measurements. As with voltage, the same can be done for temperature, and further integrals of parameters other than time can be considered, for example the integral of voltage by current. This model detects anomalies over time.

[0080] In the sixth model, as an alternative to integrating over time in line with the fifth model, the derivatives of the voltage and temperature changes can be used as a function of the stack current (or derivatives can be extracted using other parameters), especially in the case of load changes (current changes), and individual cells are compared with a reference cell. An alarm can be issued when an anomaly is found. This can be used to detect anomalies in the dynamic behavior between the cells.

[0081] In the seventh model, the time taken for a single cell to reach a complete start-up or complete shutdown is determined, providing an indication of the cell efficiency (longer times equate to higher efficiency). Individual times can be compared to define poorly behaving cells, averagely behaving cells and well behaving cells. This can be achieved by measuring the individual cell voltages of each electrolyser cell and collecting a control signal that controls the current supplied to the electrolyser as a criterion for start-up or shutdown operation detection. Shutdown can be defined as 0% current and start-up can be defined as a current that rises from 0% to 20% in less than 60 minutes. Once a start-up or shutdown is detected, it can be determined for each individual cell how long it takes the cell to reach a voltage level t while measuring the voltage. This voltage level can be defined as a simple threshold, with the first derivative being 0 or the second derivative being 0. The collected points can be integrated in time for a selected time frame, which can be derived for each single cell, so that a weighted average value gives an indication of the average time to reach a given voltage level. For each cell, the results can be plotted and the cells can be grouped into groups, such as poorly behaving cells (linked to low t values), averagely behaving cells (linked to average t values), and highly behaving cells (linked to highest t values). A well-behaving reference cell (e.g. a brand new cell) can be used to compare the relative results to an ideal situation.

[0082] In the eighth model, healthy cells can be modeled in the form of start-up, full load, shutdown and load variation zones (where the current changes during normal operation) using a multi-parameter regression equation, with fault-specific parameters assigned to the regression equation (model). Individual cells can then also be modeled by regression, and individual parameters can be compared to indicate specific faults, the measured inputs being cell voltage, stack current and cell temperature. To verify whether an individual cell is working properly or has started to age, each cell can be fitted to a current-voltage curve constructed for a number of predefined parameters (see model 7). This curve can be constructed such that the characteristics of a particular cell are present in the form of coefficients. By placing thresholds on one or more characteristics, conclusions can be drawn about the state of the cell with reference to predefined reference values. Furthermore, individual parameters of individual cells can be compared to indicate differences in cell performance. This fitting process can be done in four zones - start-up zone (current 0-20%), full load zone (20-100%), shutdown zone (current is set to 0) and load variation zone (current is varied). A similar approach can be taken across the electrolyser unit to determine electrolyser characteristics that are useful during operation.

[0083] In the ninth model, several cells known to behave in a healthy mode can be defined as reference cells. In this way, the cell that performs "best" can be set as a reference for other cells. The average value of these reference cells can be calculated and assigned as the calculated reference value for a certain instant. This reference value can be used as the calculated reference value for time t. The user / operator can define a theoretical reference voltage by inputting a reference voltage into the system. This reference voltage can be used as a user input to detect deviations between a first reference value (calculated based on the measured cell voltage of the reference cell) and a second reference value (entered by the user) over time. For each individual cell, the cell voltage can be measured within a similar time frame as the reference cell measurement. Both results can be compared and an alarm can be generated if the difference is greater than a certain average deviation, for example 10 mV. To verify the quality of the reference calculation, the calculated reference value can be compared with a user-defined reference value after each calculation of the reference value. If the deviation is above a certain level, a new set of reference cells can be assigned to the algorithm. For start-up and shut-down, the algorithm can be adapted to include process conditions to prevent false alarms from occurring.

[0084] In the tenth model, the energy balance can be used to predict the cell temperature based on stack current and cell voltage as inputs. This cell temperature can be compared to the measured cell outer surface temperature. Any anomalies can be flagged with an alarm.

[0085] In the 11th model, H 2 To detect leaks (and possible unseen flameless combustion), any measured high temperature spikes are 2 It can be considered a flame and an alarm can be issued.

[0086] The 12th model allows for the analysis of degradation over time compared to the initial model settings and for alarms to be generated. The expected situation can be modelled using the equation introduced above, V=A+B / I+C·log(I). This equation can be separated into its individual parts (e.g. A, B, C) to identify the causes of degradation.

[0087] In the case of a hydrogen production plant, the cell temperatures are useful for prediction and monitoring in order to balance efficiency and overall cell life. A thirteenth model can be created based on the provided equations, which predict individual cell temperatures and cell voltages. By comparing the predicted voltages and temperatures with the measured values, an optimization routine can be used to further iteratively improve or optimize the electrolyzer operating temperature, thus improving running costs (operating expenses) versus production yield.

[0088] By measuring the individual cell voltages and cell temperatures, comparisons can be made on the performance of individual electrolyzer cells based on continuously measured data with stack current as input. By measuring cell voltage, stack current and cell temperature, three key parameters for modeling the cell performance are available as measurements, which can therefore be used to validate predictions against measurements, providing an accurate method for detecting cell problems. Based on these models and the measured data, the cause of anomalies, e.g. undesired reactions in the membrane, electrodes or electrode degradation, can be identified.

[0089] Cell temperature is an important factor in the overall cell performance, especially for hydrogen production. If modeling is used to predict the cell temperature, a validation can be performed between the model and the individual cell temperatures and monitored over time. Deviations can indicate cell degradation or undesirable operating temperatures and can be monitored online and appropriate follow-up can be taken. If the cell temperature is not modeled, the measured cell temperature can be used to improve the cell performance. Furthermore, the cell temperature can be monitored over time. Changes in the cell temperature of individual cells are an indication of a cell problem and can be flagged for follow-up action. In case of a leak, flameless combustion can occur. Temperature spikes can be measured in such cases. An alarm can be generated to inform the authorities to take the necessary action, for example shutting down the electrolyzer.

[0090] In the previous examples, it was assumed that the individual cell temperatures were measured using fiber optic temperature sensors. Other methods exist for measuring the cell's outer surface temperature, for example, by a camera that measures the surface temperature. Multiple cameras, or cameras incorporating a moving or rotating mechanism, can be used, as cameras generally have a limited viewing angle. The data captured by a camera is in a different format compared to a fiber optic cable. Thus, the exact location of a particular measurement point requires some data processing to be accurately determined. Furthermore, measurements may include a much larger amount of data, which can be fed into sophisticated algorithms that detect changes in temperature in more detail. These details can be processed and compared, for example, to a calculated value. In order to detect hydrogen flames, such a camera may need the right characteristics to do so.

[0091] In some examples, the robot can be used to capture similar and / or other electrolytic cell data, such as temperature measurements by devices located on the robot, e.g., infrared cameras or temperature probes, or image capture devices, to detect, for example, specific coloring of the tubes through which the final product is released from the electrolytic cell. Additional data can be input into the described (and highlighted) models or new models for further processing.

[0092] Various examples of the present invention use individual cell temperature measurements using fiber optic cables or other methods, individual cell voltage measurements, and stack current measurements. A processing unit can be used to convert the measured temperatures into a format that can be used to compare with the individual cell temperatures. A processing unit with a model to compare measurements between cell temperatures can be used to detect anomalies. A processing unit can be used that includes models to predict individual cell voltages and temperatures and models to compare them with measurements. A processing unit can be used that further includes one or more of the above units. A processing unit can be used to detect flames based on temperature spikes. For example, the above-mentioned processing units can be implemented as a single processing unit or as separate processing units. Optionally, other means of detecting cell temperatures can be used, such as image-based temperature measurements. Optionally, other means of detecting other relevant cell performance parameters can be used, such as cell temperature captured by the robot, images of transparent pipelines containing the produced products. In case of problems occurring inside the electrolyzer, such pipelines can change color.

[0093] Aspects and features described with respect to a particular one of the previous examples may be combined with one or more of the additional examples, replacing the same or similar features in such additional examples, or these features may be additionally introduced into such additional examples.

[0094] Examples of the invention may further be or relate to a (computer) program, which includes program code for performing one or more of the above methods when the program is executed on a computer, a processor, or other programmable hardware component. Thus, the different method steps, operations, or processes of the above methods may also be performed by a programmed computer, a processor, or other programmable hardware component. Examples of the invention may also cover program storage devices, such as digital data storage media, which are machine-readable, processor-readable, or computer-readable and encode and / or contain machine-executable, processor-executable, or computer-executable programs and instructions. The program storage device may include or be, for example, a digital storage device, a magnetic storage medium, such as a magnetic disk and a magnetic tape, a hard disk drive, or an optically readable digital data storage medium. Other examples may include a computer, processor, control unit, (Field) Programmable Logic Array ((F)PGA), (Field) Programmable Gate Array ((F)PGA), Graphics Processor Unit (GPU), Application-Specific Integrated Circuit (ASIC), Integrated Circuit (IC) or System-on-a-Chip (SoC) system programmed to perform the steps of the methods described above.

[0095] It is further understood that the disclosure of several steps, processes, operations, or functions disclosed in the detailed description or claims should not be construed as implying that these operations are necessarily order dependent unless expressly stated in individual cases or required for technical reasons. Thus, the foregoing description does not limit the performance of several steps or functions to a particular order. Furthermore, in other examples, a single step, function, process, or operation may include several sub-steps, sub-functions, sub-processes, or sub-operations, and / or a single step, function, process, or operation may be divided into several sub-steps, sub-functions, sub-processes, or sub-operations.

[0096] Where some aspects are described in terms of an apparatus or system, these aspects should also be understood as descriptions of a corresponding method. For example, certain blocks, devices, or functional aspects of such an apparatus or system may correspond to features such as method steps of a corresponding method. Thus, aspects described in terms of a method should also be understood as descriptions of corresponding blocks, corresponding elements, properties, or functional features of a corresponding apparatus or corresponding system.

[0097] Hereby, the following claims are included in the detailed description, with each claim standing alone as an independent example. Also, in the claims, where a dependent claim refers to a specific combination with one or more other claims, other examples may include combinations of the dependent claim with the subject matter of any other dependent or independent claim. Such combinations are hereby expressly suggested, unless it is expressly stated that a particular combination is not intended in a particular instance. Moreover, the features of a claim should also be included in any other independent claim, even if that claim is not directly defined as dependent on any other independent claim.

Claims

1. 1. A method for monitoring the condition of a multi-cell electrolyzer, comprising the steps of: acquiring (110) sensor information, the sensor information including at least information about measured temperatures of individual cells of the multi-cell electrolyser; - predicting (130) for each of the cells of the multi-cell electrolytic cell based on the sensor information, an expected temperature value for that cell of the multi-cell electrolytic cell; providing (170) a notification if a deviation between the predicted temperature value for the cell and the measured temperature for that cell matches a temperature deviation condition; The method includes:

2. 2. The method of claim 1, wherein the sensor information comprises information on a stack current of the multi-cell electrolyzer and information on individual cell voltages of the cells of the multi-cell electrolyzer, and wherein the expected temperature value of the cell is predicted based on the cell voltage of the cell and based on the stack current of the multi-cell electrolyzer.

3. 3. The method according to claim 1 or 2, wherein the expected temperature value of the cell is predicted based on measured temperatures of one or more cells adjacent to the cell in the multi-cell electrolyser.

4. 4. The method according to any one of claims 1 to 3, wherein the expected temperature value of the cell is predicted based on an average value of the measured temperatures of the cells of the multi-cell electrolyser.

5. The method according to any of the preceding claims, wherein the expected temperature value of the cell is predicted based on one or more reference cells of the multi-cell electrolyser.

6. The method of claim 5 , further comprising the step of selecting (120) the one or more reference cells based on the sensor information.

7. 7. The method according to claim 1, wherein the predicted temperature value of the cell corresponds to an integral or derivative of the predicted temperature of the cell with respect to time, voltage or current, and the deviation between the predicted temperature value of the cell and the measured temperature of the cell is determined as the difference between the integral or derivative of the predicted temperature of the cell and a corresponding integral or derivative of the measured temperature of the cell.

8. The method of any of claims 1 to 7, further comprising the step of providing (170) notification of a suspected flame if a deviation between the predicted temperature value of the cell and the measured temperature of that cell exhibits a spike.

9. 9. The method according to any of the preceding claims, wherein the sensor information comprises information on the stack current of the multi-cell electrolyser and information on individual cell voltages of the cells of the multi-cell electrolyser, the method comprising the step of predicting (140) for each of the cells of the multi-cell electrolyser based on the sensor information, or obtaining as input the expected voltage values ​​of the cells of the multi-cell electrolyser.

10. The method of claim 9 , wherein the expected voltage values ​​include an expected voltage and an expected time to reach the expected voltage in response to a load change.

11. 11. The method of claim 9 or 10, wherein the predicted voltage value is based on a multi-parameter regression equation, the multi-parameter regression equation including a first parameter that models an undesired reaction at an electrode, a second parameter that models at least one of membrane ruptures, pinholes, and degradation, and a third parameter that models degradation of the electrode, the method including the steps of determining (150) a deviation between the predicted voltage value of the cell and a measured voltage of the cell, and using the multi-parameter regression equation to identify the contribution of the first parameter, the second parameter, and the third parameter to the deviation.

12. 12. The method according to claim 9, wherein the predicted voltage value of the cell corresponds to an integral or derivative of the predicted voltage of the cell with respect to time, temperature or current, and the deviation between the predicted voltage value of the cell and the measured voltage of the cell is determined as the difference between the integral or derivative of the predicted voltage of the cell and a corresponding integral or derivative of the measured voltage of the cell.

13. 13. The method according to any of claims 9 to 12, wherein the method comprises, for each cell of the multi-cell electrolyser and / or for one or more reference cells, determining (160) a previous voltage value for that cell based on the sensor information, and comparing the previous voltage value with a current voltage measurement contained in the sensor information.

14. The method according to any of the preceding claims, wherein information about the measured temperatures of the cells of the multi-cell electrolyser is obtained from a fibre optic temperature sensor or from a camera-based temperature sensor.

15. A computer program having a program code for performing the method according to any of claims 1 to 14, when the computer program is run on a computer, a processor or a programmable hardware component.

16. 1. An apparatus (50) for monitoring the condition of a multi-cell electrolytic cell, comprising: an interface circuit (52) for acquiring sensor information from one or more sensors; A processing circuit (54) configured to carry out the method according to any one of claims 1 to 14; An apparatus comprising:

Images