Method for operating an electrolysis cell and electrolyser having a plurality of electrolysis cells
Patent Information
- Application Number
- PCT/EP2026/055893
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-03-26
- Filing Date
- 2026-03-04
- Publication Date
- 2026-10-01
Smart Images
Figure EP2026055893_01102026_PF_FP_ABST
Abstract
Description
[0001] 2024 PF00867
[0002] 1
[0003] Description
[0004] Method for operating an electrolysis cell and electrolyzer with a plurality of electrolysis cells
[0005] The invention relates to a method for operating an electrolysis cell comprising an electrolyzer for producing hydrogen and oxygen as product gases. The invention further relates to such an electrolyzer with a plurality of electrolysis cells.
[0006] Hydrogen is currently produced, for example, using proton exchange membrane (PEM) electrolysis or alkaline electrolysis. These electrolyzers use electrical energy to produce hydrogen and oxygen from the supplied water.
[0007] An electrolyzer typically comprises a large number of electrolysis cells arranged adjacent to one another. Water is split into hydrogen and oxygen in these cells via electrolysis. In a so-called PEM electrolyzer, demineralized water is typically supplied as the reactant at the anode and split into hydrogen and oxygen across a proton-exchange membrane (PEM). The water is oxidized to oxygen at the anode. The protons produced in this process pass through the proton-exchange membrane. Hydrogen is produced at the cathode. The water is usually pumped from one side into the anode compartment and / or cathode compartment. Besides PEM electrolysis, water electrolyzers are also known that are designed and operated for alkaline electrolysis (AEL).Furthermore, anion exchange membrane (AEM) electrolysis is a known form of electrochemical decomposition of water to produce hydrogen and oxygen as product gases. 2024 PF00867.
[0008] 2
[0009] This electrolysis process takes place in the so-called electrolysis stack, which consists of several electrolysis cells. Water is introduced as a reactant into the electrolysis stack, which is under DC voltage. After passing through the electrolysis cells, two fluid streams emerge, consisting of water and gas bubbles (oxygen O2 and hydrogen H2, respectively). The separation of the water and gas phases in the fluid streams is carried out in gas separators.
[0010] In practice, small amounts of hydrogen are present in the oxygen gas stream, and small amounts of oxygen are present in the hydrogen gas stream. The quantity of each foreign gas depends on the electrolysis cell design and also varies under the influence of current density, catalyst composition, aging, and, in the case of a PEM electrolysis system, the membrane material. It is inherent to the system that the gas stream of one product gas always contains very small amounts of the other. In subsequent process steps, even trace amounts of oxygen are typically removed from the hydrogen in gas purification steps, sometimes involving complex and costly processes, especially when a particularly high product gas quality is required, as is the case, for example, when using the hydrogen in fuel cells.
[0011] For example, in an electrolysis plant for gas purification, the product gas streams from the electrolyzer can be fed to a respective catalytically activated recombinator, in which a catalyst recombines the hydrogen with the oxygen to form water (DeOxo unit). For this to work, the gas stream must first be heated to at least 80 °C so that the conversion rates of the recombinator are sufficiently high and the required gas purity is achieved. However, the process equipment used for this is expensive and, due to its energy consumption, reduces the overall system efficiency of the electrolysis plant. Therefore, the purity and quality of the gas streams are already being carefully considered.
[0012] Attention must be paid to the product gas streams that initially arise in the electrolyzer and are discharged from the electrolyzer, in order to keep the costs and effort for the subsequent cleaning steps within a reasonable range, in addition to operational safety aspects.
[0013] The purity and quality of the two product gas streams originally produced in the electrolyzer depend on many parameters and can change during the operation of an electrolysis plant. A problematic and particularly safety-relevant issue arises when the concentration of oxygen in hydrogen increases, and conversely, when the concentration of hydrogen in oxygen increases. If a certain concentration limit is exceeded, especially in the respective gas separator (vessel) immediately downstream of the electrolysis unit, the produced oxygen gas can no longer be used for other purposes. If the proportion of hydrogen in the oxygen product gas continues to rise, a flammable or even explosive mixture can form. This creates a potentially hazardous operating condition in the gas separator (vessel), which must be avoided at all costs for safety reasons.This also applies to the hydrogen side.
[0014] Therefore, reliable and continuous monitoring of the product gas quality during electrolysis plant operation is essential. This is particularly important on the oxygen side of the electrolyzer, specifically monitoring the concentration of hydrogen as an extraneous gas in the oxygen produced during electrolysis. Monitoring and appropriate operational management represent a crucial safety precaution for detecting critical operating conditions and implementing safety measures, including the temporary shutdown of the electrolysis plant.
[0015] It can therefore be observed that in water electrolyzers of all types as well as in fuel cells, the undesirable 2024 PF00867
[0016] 4
[0017] Gas transfer occurs between the two half-cells. A membrane, or in the case of alkaline electrolysis, a separator, separates the anodic half-cell—where oxygen is produced—from the cathodic half-cell, where hydrogen is formed. However, the product gases can diffuse through the membrane or separator into the other half-cell, potentially forming highly explosive and flammable gas mixtures in either the anode or cathode compartment.
[0018] The diffusion of hydrogen from the cathode compartment to the anode compartment across the membrane is of particular importance, as the diffusion rate of hydrogen is approximately twice that of oxygen. Furthermore, diffusion also results in a loss of valuable product gas, thus reducing the system's efficiency. For these reasons, it is crucial to understand the diffusion processes and limit them to a level that remains tolerable from an operational safety perspective. During full-load operation of a properly functioning electrolyzer, the hydrogen diffusion rate is so low compared to gas production that the extraneous gas is diluted to a safe concentration.However, the actual gas transfer and thus the proportion of foreign gas is the result of different cell components of an electrolysis cell, which age differently during operation and change their properties, thus having a respective influence and contribution on the gas permeation through the membrane.
[0019] The invention is therefore based on the objective of enabling improved operation of an electrolysis cell with regard to safety and efficiency, whereby more accurate predictions of efficiency-reducing aging processes should be possible. A further objective is the provision of an electrolyzer comprising a plurality of electrolysis cells. 2024 PF00867
[0020] 5
[0021] The object of the invention is achieved by a method for operating an electrolysis cell in which hydrogen and oxygen are produced as product gases, wherein the electrolysis cell has an anode compartment and a cathode compartment which are separated by an electrode-coated membrane, wherein the electrolysis cell is supplied with an electrolysis current, so that oxygen product gas is formed in the anode compartment and hydrogen product gas is formed in the cathode compartment, wherein an extraneous gas fraction of hydrogen is transferred from the cathode compartment through the membrane into the anode compartment, characterized in that the current density is varied and the hydrogen permeation rate through the membrane is determined as a function of the current density and as a function of water transport from the anode compartment through the membrane into the cathode compartment.
[0022] The invention is based on the understanding that effective monitoring of aging processes, the calculation of efficiencies, and a sufficiently good understanding for the further development of components require the targeted evaluation of measured values from experimental and field systems. One difficulty lies in the fact that only integral values, resulting from a complex interplay of various phenomena and components, can be measured.
[0023] The first factor influencing the foreign gas concentration is the effective diffusivity of hydrogen through the membrane, which depends not only on the polymer structure but also on the water content and other parameters. With increasing current density, the supersaturation of hydrogen in the cathode compartment also plays a significant role. This is determined by the mass transport properties of the cathode. These properties depend, among other things, on the ionomer content, the thickness of the electrode layer, and especially the layer and surface structure of the electrode. 2024 PF00867
[0024] Furthermore, the transfer water, due to electro-osmotic flow, also known as "drag," can influence the local concentrations and thus the effective net transport of hydrogen through the membrane. It has been shown that simply measuring the concentration of hydrogen in the oxygen in the anode compartment does not allow for a satisfactory and sufficient direct quantification of the contribution of the individual influencing components and phenomena to the measured value. The invention recognizes that this knowledge is crucial for conducting a precise analysis of aging processes ("predictive maintenance") and operational lifetime predictions (service life) of electrolysis cells and electrolyzers, and for planning service measures.This is also essential for the development and characterization of new electrolysis cell components, particularly the design, material composition, and structure of the membrane-electrode arrangement. Building on this, the design of the overall electrolyzer system with a large number of electrolysis cells can be carried out, as well as the determination and monitoring of the efficiency over the operating period from commissioning (BoL: Beg-of-Life) to decommissioning (EoL: End-of-Life).
[0025] The invention recognizes that existing operational management and service concepts for electrolysis cells and electrolyzers, with regard to the determination, monitoring, prediction, evaluation, and, if necessary, remediation of critical operating conditions, are both technically complex and highly inaccurate, resulting in significant economic disadvantages. For example, the aging condition is often overestimated, and a large safety margin is incorporated, leading to premature shutdown or replacement of electrolysis cell components. Monitoring the concentration of extraneous gases, particularly hydrogen in oxygen, is already an integral part of the safety concept of known electrolyzers. Thus, over the entire operating period, the concentration of extraneous gases is monitored. 2024 PF00867
[0026] 7
[0027] The system ensures that a fixed safety margin to the critical value of 4% hydrogen in oxygen is maintained. This simple approach guarantees that no flammable mixture is formed. However, it only provides limited information about the aging of various components and their influence on gas permeation, and this information is only considered in the evaluation.
[0028] Since measuring the concentration of extraneous gases alone does not allow for a direct conclusion regarding the contribution of individual phenomena, the invention incorporates further state-measurement variables, thereby increasing the informative value of the findings. This is achieved by incorporating and combining these measurements with a model-based description of physical processes, from which a measure or operational characteristic value for the operational degradation of the membrane or the membrane-electrode assembly (MEA) of an electrolysis cell is derived. In this method, the hydrogen permeation rate through the membrane is determined as a function of the current density, and for the first time, the effect of water transport from the anode compartment through the membrane to the cathode compartment is considered "in situ" during characterization.The simple approach used so far – for example, in a PEM electrolysis cell – completely disregarded this effect and led to a systematic underestimation of the transport resistance of the cathode, and also incorrectly attributed possible temporal changes in the electro-osmotic flux to the cathode, thus resulting in an insufficiently accurate indicator or measure of degradation.
[0029] In addition to considering the transfer water through the electrode-coated membrane, it is also possible to consider and measure the pressure in the electrolysis cell as an operating parameter, so that the operating concept within the scope of the invention can also be extended to an advantageous differential pressure operation, in which a differential pressure between the anode compartment and the cathode compartment is set and measured. 2024 PF00867
[0030] 8
[0031] A differential pressure, in turn, influences the material or mass flows accordingly and therefore the result for the effective foreign gas input through the membrane of hydrogen from the cathode compartment to the anode compartment or of oxygen from the anode compartment to the cathode compartment.
[0032] From the hydrogen concentration measured in the anode compartment at a specific operating point of the electrolysis cell and the oxygen production rate of the electrolysis process, the hydrogen permeation rate through the membrane can be calculated as a function of the current density. When this quantity is plotted, it typically results in a value that increases with current density, up to approximately 2 A / cm². 2 approximately linear progression.
[0033] The y-intercept corresponds to the permeation rate at a current density of 0 and is directly determined by the effective diffusivity of hydrogen in the membrane and its thickness. The slope of the graph is explained by the increased hydrogen concentration at the cathode electrode, often referred to as supersaturation. According to this model, the slope of the permeation rate versus current density is determined by the relationship between the gas permeability of the membrane, which in turn is determined by the membrane thickness and diffusivity, and the electrode's tendency to supersaturate. The electrode's tendency to supersaturate is characterized by its transport properties. Therefore, according to general understanding, the slope is interpreted as a property of the electrode.
[0034] Based on this, the invention proposes an operating method that implements a significantly more reliable and advanced condition diagnosis for an electrolysis cell compared to the prior art, by advantageously also determining and taking into account the water transport resulting from the electro-osmotic flow through the membrane. This effect occurs in a PEM electrolysis cell with diffusion-driven gas permeation of hydrogen from the cathode compartment through the membrane. 2024 PF00867
[0035] The membrane moves towards the anode compartment, so that the net transfer of hydrogen in the anode compartment is actually lower. This is now being implemented for the first time in the new operating concept for the process of operating an electrolysis cell and is taken into account via an extended physical model, as explained below:
[0036] To quantify the influence of the transfer water on the transport of hydrogen across the membrane, a balance equation for hydrogen is set up that takes diffusion and convection into account:
[0037] clC[_[2. ^H2 ^H2 = ~DH2 ^H20,drag 0)
[0038]
[0039] ax c H2O
[0040] In equation (I), the quantity n is denoted. H2 the molar hydrogen flux, D H2 the effective diffusivity of hydrogen, c H2 the local hydrogen concentration, x the coordinate through the membrane and n H20 , dragthe molar transfer water flow. The molar transfer water flow can be determined directly from the change in the fill level of the gas-liquid separators downstream of the electrolysis cells on the anode and cathode sides, respectively, during the operation of an electrolysis cell, an electrolyzer comprising a multitude of electrolysis cells, or an electrolysis plant. In the equation (I) above, c H20 The concentration of water is described.
[0041] An analytical solution to this inhomogeneous first-order ordinary differential equation leads to the following result:
[0042] ^H20,drag ^H2,supsat ' exp ^Mem ^H2,an DH2 ' c H2O ^H2 — - (1 _ ^H20,drag
[0043]
[0044] ^-H20,drag exp DH2 ' c H2O c H2,supsat represents the supersaturated hydrogen concentration at the cathode electrode. This value depends on the source. 2024 PF00867
[0045] 10
[0046] The electrode's distance from the surface and cannot be measured directly. c H2 an This is the hydrogen concentration on the anode side. This can be calculated from the measured hydrogen-to-foreign gas concentration using thermodynamic equations (saturation vapor pressure of water, Dalton equation, and gas solubility). Mem is the thickness of the membrane. The effective diffusivity D H2 It cannot be measured directly during electrolysis operation. While data exists for certain membranes, it is only an approximation for real electrolysis cells and can change during operation.
[0047] The equation (II) above describes the rate of hydrogen transport, i.e., the hydrogen permeation rate n H2 through the membrane, taking into account the transfer water and the resulting local concentration distribution.
[0048] In a preferred embodiment of the method of the invention, a characteristic curve is determined for a specific water transfer coefficient, which describes the hydrogen permeation rate through the membrane as a function of the current density at this water transfer coefficient.
[0049] Thus, in the operating method of the invention, a characteristic curve field is advantageously used that is parameterized via the transfer water coefficient. The transfer water coefficient is a component-independent characteristic value of the electrode-coated membrane. Any possible change over time due to start-up and degradation processes can, for example, be easily and advantageously determined by repeatedly evaluating the fill levels in the gas-liquid separators at a specific operating flow rate. Therefore, at a specific operating point of the electrolysis cell, the net hydrogen permeation rate for a given, current transfer water coefficient can be determined based on the current density. 2024 PF00867
[0050] 11
[0051] Model calculations and initial practical investigations have shown that the transfer water coefficient, determined by the membrane, alters the slope of the approximate straight line along which the hydrogen permeation rate increases linearly as a function of current density during operation of the electrolysis cell. The invention recognizes and advantageously utilizes the fact that, consequently, the hydrogen permeation rate is not solely a function of the ratio of the permeability of the membrane and the electrode. This finding is of particular importance for comparing the design and structure of different membrane-electrode arrangements and for interpreting and utilizing measurement data over the lifetime of an electrolysis cell or electrolyzer with the aim of analyzing the aging of various components and, in particular, quantifying it more effectively than previously used simple assumptions.
[0052] To evaluate and operationally utilize measurement data and the calculated rates for hydrogen permeation in such a way that the two properties membrane permeability, which is characterized by D H2 , as well as the transport property of the electrode, which is characterized by c H2 , supsat not be corrupted by the transfer water, the equation ( II ) derived above should be integrated into the data evaluation and the operating procedure of the electrolysis cell.
[0053] In a particularly preferred embodiment of the method, this is achieved by selecting data points from which the effective diffusivity for characterizing hydrogen permeation and the supersaturated hydrogen concentration in the cathode arranged in the cathode space for characterizing the transport property of the cathode are determined.
[0054] It is advantageously possible to easily calculate the effective diffusivity D from multiple data points. H2 as well as the supersaturated hydrogen concentration c H2 , supsat using the2024 PF00867
[0055] 12
[0056] Equation (II) is to be determined. This is preferably carried out by determining the hydrogen permeation rate n. H2 measured at different current densities j or read out based on a stored characteristic curve for a specific transfer water coefficient.
[0057] Since equation (II) is implicit and nonlinear, a first approach would be to use a global optimization algorithm. However, such an algorithm-based approach would be quite computationally intensive, and the results could be influenced by the given initial values.
[0058] Therefore, when implementing the method of the invention, it is particularly advantageous to use a simplified solution approach and to carry it out in the method. This can also be easily implemented in a diagnostic and monitoring device of an electrolyzer, which is equipped with appropriate sensors, input and output interfaces, as well as memory and processors.
[0059] For automated data evaluation of operational and condition data on a large scale, equation ( II ) is simplified.
[0060] A Taylor approximation of the exponential function, followed by the neglect of two terms that turn out to be insignificant compared to the third summand, leads to the following relationship:
[0061] c H2O c H2,cath n H2 ■ V - + - 5 - n H20,drag / ^H2 ' C H20 \ f^H2 ' C H2O ' C H2,cath '
[0062]
[0063] ~ \2 - TWK ■ d Mem - k cath ) l TWK ■ d Mem ) ' j 1}
[0064] This form allows the values for the left-hand term in equation (III) to be plotted, i.e., n H2 • >. Cfl2 ° - 1- > CH2 ' cath above the recip- ^ t H2O, drag 22024 PF00867
[0065] 13
[0066] Roken current density -. Subsequently, a linear regression is performed, which always has a unique solution.
[0067] c H2 cath The hydrogen concentration in the cathode compartment outside the cathode electrode can be calculated directly from the pressure and temperature. Therefore, in the method, the pressure and temperature in the cathode compartment and the anode compartment are preferably also measured using appropriate sensors, with which the measuring and diagnostic system is advantageously equipped.
[0068] This is k cath a characteristic coefficient that is introduced to describe the substance transport within the cathode and to calculate c H2 supsat via the following approach. The approximate application is based on the physically plausible assumption of neglecting the hydrogen permeation rate through the membrane compared to the hydrogen production rate:
[0069] kcath ' (j-H2,supsat <-H2,cath) > p
[0070]
[0071] 0^)
[0072] One obtains an approximately linear relationship with the current density, the slope of which additionally depends on the transfer water coefficient TWK (see equation III).
[0073] TWK is the transfer water coefficient – the transfer water flow is proportional to TWK and the operating current – which can be determined directly from the electrolyzer's system data, and F is the Faraday constant. Unless measurement data is available for arbitrarily low current densities, the results obtained with the linearized equation will still contain errors. Nevertheless, these serve as good starting points and approximations for optimization using the nonlinear function.
[0074] This method thus enables the determination of the relevant mass transport properties in a particularly advantageous way.-2024 PF00867
[0075] 14
[0076] The membrane and electrode (MEA: membrane-electrode assembly) are able to function without being distorted or obscured by the effect of the transfer water. The technical implementation of the invention in the operating method for an electrolysis cell or an electrolyzer with a plurality of electrolysis cells is therefore particularly advantageous.
[0077] In a particularly preferred embodiment of the method, therefore, especially by applying equation ( II ), the supersaturated hydrogen concentration (c ) is H 2,supsat) in the cathode arranged in the cathode from the hydrogen concentration ( o H 2, cath ) in the cathode space outside the cathode, where the hydrogen concentration ( c H 2, cath ) in the cathode space outside the electrode is determined from the pressure and temperature in the cathode space.
[0078] The relationship between the supersaturated hydrogen concentration (c H2, supsat ) in the cathode arranged in the cathode compartment and the hydrogen concentration ( c H The gas solubility (2, cath) in the cathode compartment outside the electrode is described by equation (IV). In addition, the gas solubility under the specific state variables pressure and temperature can be taken into account. Equation (IV) then includes two unknown quantities k cath and D H2 l which describe the transport properties of the cathode and membrane and can be determined by comparison with the measurement data, preferably by linear regression.
[0079] In this way, the approximately linear relationship found according to equation (IV) is implemented in the operating procedure to determine the supersaturated hydrogen concentration (c). H 2, supsat ) to determine from measured values in the cathode space. The coefficient k cath characterizes – as explained above – the substance transport within the cathode electrode. Depending on the operating mode of the electrolysis cell or electrolyzer, a differential pressure operation may also be implemented, in which a pressure difference is set between the anode and cathode compartments during electrolysis. A pressure difference 2024 PF00867
[0080] 15
[0081] This leads to changes in substance transport compared to constant-pressure operation of the electrolysis cell. This effect can, for example, be accounted for in the linear approach of equation (IV) as a differential pressure-dependent correction via the characteristic coefficient k. cath can be determined and taken into account without making fundamental changes to the model. Thus, for example, k cath = k cath (Jp) « k cath(p = 0) + a • Ap, a differential pressure-dependent correction term is taken into account. Additionally, the pressures in the anode and cathode compartments are considered by c H2 an and c H2 cath calculated from the solubility and the respective pressure using the Henry and Dalton equation.
[0082] In a further preferred embodiment of the method, this is carried out in an electrolyzer with a plurality of axially stacked electrolysis cells, wherein the operation of the electrolyzer is monitored and an aging state of the electrolysis cells is diagnosed.
[0083] By implementing the method and applying it in the operation of an electrolyzer, the aging of the membrane electrode arrangement in si tu of electrolysis cells can be determined and reliable lifetime predictions can be obtained.
[0084] In general, the focus of previous operating concepts involving gas permeation across the membrane has been on avoiding potentially hazardous gas mixtures during operation to prevent the risk of explosion. The development of new cell concepts is primarily based on testing different variants over a limited trial period. The precise processes within the cell, especially after extended operation, remain largely unknown. This also introduces uncertainty into the calculation of actual overall efficiencies.
[0085] The presented operating procedure goes a step further in terms of understanding the actual processes in the electrolysis cell and increases the added value and quality of gewon-2024 PF00867
[0086] 16
[0087] This allows for the application of new measurement data to the operational management of an electrolysis cell or electrolyzer. Long-term tests and field data can thus be interpreted more effectively with regard to the degradation of different components (keyword: "predictive maintenance"). This improves the understanding and knowledge of actual losses and overall system efficiency, and serves the purpose of more targeted development and optimization of future, more stable cell components and service planning.
[0088] According to another aspect of the invention, an electrolyzer with a diagnostic system is proposed, which includes a measuring device and a data processing device, wherein the diagnostic system is set up to carry out a method.
[0089] The diagnostic system can be part of a higher-level control system or the process engineering unit (PTU) of a complex electrolysis system or an electrolyzer.
[0090] In a particularly preferred embodiment of the electrolyzer, the data processing device has a memory into which a characteristic curve can be read or written, wherein the characteristic curve represents the hydrogen permeation rate n at an operating point. H2 described by the membrane as a function of the current density at a water transfer coefficient (TWK).
[0091] The data processing unit is designed and configured so that a characteristic curve for a specific transfer water coefficient (TWK) can be read into or out of the storage system. It is also possible to modify and update an existing characteristic curve, in particular to overwrite it and replace it with a new one. Updates and adjustments are made based on comparison with actual operational measurement data, e.g., if the predicted characteristic curve for the gas-2024 PF00867 is not met at an operating point with a set current density.
[0092] 17
[0093] Permeation deviates from the measured value beyond a permissible tolerance range – derived, for example, from the measurement uncertainty. In this way, the characteristic curve array for an electrolysis cell or electrolyzer can be iteratively improved. A characteristic curve array is a set of characteristic curves characterized by the transfer water coefficient (TWK). Advantageously, a characteristic curve array with a large number of characteristic curves for a given value of the transfer water coefficient is stored in the memory of the data processing device.
[0094] In a further preferred embodiment of the electrolyzer, it is designed as a PEM electrolyzer with a proton-conducting membrane or as an AEM electrolyzer with an anion-conducting membrane.
[0095] Preferably, the electrolyzer can be operated under differential pressure, using a differential pressure control device connected to the data processing unit for data exchange. In differential pressure operation, the differential pressure between the hydrogen product gas and the oxygen product gas is regulated such that, for example, a maximum pressure difference across the membrane is not exceeded or a predetermined differential pressure is set. The characteristic curve is therefore extended or modified by a value for differential pressure (Ap > 0) in addition to the transfer water coefficient at constant pressure (Ap = 0). This can be implemented simply, for example, by a differential pressure-dependent correction function, such that k cath = k cath (Ap) « k cath (Ap = 0) + a • Ap is taken into account as a differential pressure-dependent correction term.
[0096] Differential pressure control in a PEM-based or AEM-based electrolysis system protects the membrane in particular, as the pressure difference between the oxygen side and the hydrogen side is maintained at a permissible setpoint to achieve the highest possible system efficiency and corresponding hydrogen yield while simultaneously [2024 PF00867]
[0097] 18
[0098] Operational reliability. Advantageously, the differential pressure can still be regulated in the invention via existing control valves and control devices for operational management. The pressure levels can therefore differ on the hydrogen and oxygen sides, as long as a permissible differential pressure is observed with respect to the diaphragm, to which regulation is directed. Electrolyzers are generally designed and well-suited for operation in differential pressure mode. For example, the hydrogen side can be operated at a high pressure while the oxygen side simultaneously vents to the atmosphere without pressure. Alternatively, both the hydrogen and oxygen sides can operate at a higher pressure relative to atmospheric pressure.
[0099] Further advantages, features, and details of the invention will become apparent from the following description of preferred embodiments and from the drawings. The features and combinations of features mentioned above in the description, as well as those mentioned below in the figure description and / or shown individually in the figures, can be used not only in the combinations specified, but also in other combinations or individually, without departing from the scope of the invention.
[0100] Exemplary embodiments of the invention are explained in more detail with reference to the drawing. The drawing shows, schematically and in a highly simplified form:
[0101] FIG 1 shows the basic structure of an electrochemical cell, which is exemplified as a PEM electrolysis cell;
[0102] FIG 2 shows a diagram of the hydrogen permeation rate as a function of the current density of an electrolysis cell in a qualitative progression; 2024 PF00867
[0103] 19
[0104] FIG 3 shows in a diagram the hydrogen permeation rate as a function of the current density for different transfer water coefficients;
[0105] FIG 4 shows an electrolyzer with a circuit on the oxygen side and with a diagnostic system.
[0106] The same reference symbols have the same meaning in the figures.
[0107] Figure 1 schematically shows the structure of an electrochemical cell 1, which is configured as an example electrolysis cell 1 for PEM electrolysis. The electrochemical cell 1 is part of an electrolyzer (not shown in detail here) for splitting water (H₂O) by direct current (I) at a current density j to produce hydrogen and oxygen.
[0108] The electrochemical cell 1 comprises an electrolyte consisting of a proton-conducting membrane 7 (proton exchange membrane, PEM), on which electrodes 11a and 11b are located on both sides. The unit consisting of membrane 7 and electrodes 11a and 11b is commonly referred to as a membrane electrode assembly (MEA). Electrode 11b is designated as the cathode, and electrode 11a as the anode. The electrolysis cell 1 is composed of an anodic half-cell and a cathodic half-cell, which are spatially and functionally separated by the common electrode-coated membrane 7. Thus, an anode compartment 3 is formed in the anodic half-cell and a cathodic half-cell in the cathodic half-cell, into which the respective electrode 11a and 11b are placed. Electrodes 11a, 11b comprise a catalyst material which is contained in a thin catalytically active layer of approximately 1-1.5 mg / cm². 2applied to membrane 7. The catalyst material is, for example, iridium on the anode side, which catalytically accelerates the conversion reaction, and 2024 PF00867
[0109] 20
[0110] which is very expensive. Therefore, many efforts are being made to manage with a lower amount of catalyst material at electrodes 11a and 11b.
[0111] Each of the electrodes 11a and 11b thus configured has a gas diffusion layer 13 present: an anodic gas diffusion layer 13 at the anode electrode 11a and a cathodic gas diffusion layer 1 at the cathode electrode 11b. The gas diffusion layers 13 are each contacted by a so-called bipolar plate 15. In the assembled state of an electrolysis stack, comprising a plurality of individual electrolysis cells 1 connected in series, these plates spatially separate the electrolysis cells 1 from one another. During operation, the electrolysis cell 1 is supplied with fully demineralized water H₂O as the reactant, which is converted at the anode 11a into oxygen gas O₂ and protons H₂. + is broken down. The protons H +The molecules migrate through the electrolyte membrane 7 towards the cathode 11b. On the cathode side, they recombine in the cathode compartment 5 at the cathode 11b, which is coated with a catalyst material, to form hydrogen gas H₂. Thus, hydrogen H₂ and oxygen O₂ are produced. Due to electroosmotic effects, a flow of water H₂O from the anode compartment 3 across the membrane 7 into the cathode compartment 5 is also observed. This results in the flow of transfer water or "drag water" into the cathode compartment 5. This transfer water also transports a portion of hydrogen H₂ from the anode compartment 3 back into the cathode compartment 5 (convection), for example, by several water molecules hydrating a hydrogen molecule H₂ at the molecular level and carrying it along.Due to this effect of the transfer water, in the present example of PEM electrolysis, the effective (net) concentration-driven extraneous gas input of hydrogen H2 by diffusion from the cathode compartment 5 via the membrane 7 into the anode compartment 3 is reduced, i.e., the effective extraneous gas concentration of hydrogen H2 in oxygen O2 in the anode compartment is lower due to this transport phenomenon. 2024 PF00867.
[0112] 21
[0113] In another embodiment, the electrochemical cell 1 can be designed as a galvanic cell or fuel cell for generating electricity.
[0114] The gas diffusion layer 13 is a planar component and a functional layer of significant importance for the electrolysis cell 1, performing various tasks during its operation. The gas diffusion layer 13 ensures optimal water distribution and the removal of the product gases hydrogen (H₂) and oxygen (O₂). In the case of a galvanic cell, the gas diffusion layers 13 serve to supply reactants to the respective electrodes. A crucial aspect here is that the gas diffusion layer 13 must be sufficiently permeable to the gaseous products or reactants to allow their removal. This requires a certain degree of porosity to enable and facilitate this transport. In an electrolysis cell 1, the gas diffusion layer 13 also acts as a current distributor. For these reasons, the gas diffusion layer 13 is made of an electrically conductive, porous material.It is important to achieve a uniform, i.e., homogeneous, current distribution so that the current density j at electrodes 11a, 11b is as homogeneous as possible across the catalytically active area during operation, and for a longer operating time. This reduces or prevents degradation of the electrolysis cell 1. Uniform and multiple contacting of electrodes 11a, 11b also has a positive effect on the material required for the catalyst coating. Likewise, a fine-pored or microporous structure at the contact surface of the gas diffusion layer 13 with the respective electrode 11a, 11b is desirable.
[0115] During operation of electrolysis cell 1, it is supplied with a direct current I. Hydrogen (H₂) is produced in cathode compartment 5 and oxygen (O₂) in anode compartment 3, respectively, as the product gases. Anode compartment 3 and cathode compartment 5 are coated with an electrode. 2024 PF00867
[0116] 22
[0117] Membrane 7 is separated. A foreign gas fraction of hydrogen (H₂) is transferred from the cathode compartment 5 through the membrane 7 into the anode compartment 3. To determine the operational degradation state of the electrolysis cell 1 and, in particular, the sensitive electrode-coated membrane 7, the current density j and the hydrogen permeation rate n are varied periodically in one or more measurement cycles. H2 through the membrane 7 as a function of the current density j and as a function of the water transport n H2O,drag from the cathode compartment 5 through the membrane 7 into the anode compartment 3. This can be carried out during initial commissioning for acceptance testing and / or periodically and regularly, for example, along equivalent full-load hours of the electrolysis cell 1.
[0118] FIG 2 shows the hydrogen permeation rate n in a diagram. H2as a function of the current density j of an electrolysis cell 1 or also of an electrolyzer 10a (see FIG. 4) in the qualitative course. Here, the electro-osmotic effect due to the transfer water is neglected, i.e., for the water transport through the membrane 7 from the anode compartment 3 to the cathode compartment 5, the following relationship applies: H2O drag = 0 or the transfer water coefficient is assumed to be TWK = 0. This corresponds to the conventional, highly simplified, and idealized model for determining the hydrogen permeation rate n. H2 Since the transfer water coefficient TWK is not zero in practice for PEM or AEM electrolysis, this assumption systematically overestimates the proportion of hydrogen (H2) in the anode compartment 3 during operation. This leads to incorrect or inaccurate conclusions regarding the aging and condition of the electrode-coated membrane 7.
[0119] In FIG. 2, the current density j is plotted on the x-axis and the hydrogen permeation rate n on the y-axis. H2 The discrete data points 17 exhibit a nearly linear relationship, so that a connecting line is plotted as the characteristic curve 9 by linear regression. 2024 PF00867
[0120] 23
[0121] The y-intercept therefore corresponds to the permeation rate n. H2 at a current density of j = 0 and is directly determined by the effective diffusivity D H2The hydrogen permeation rate is determined by the hydrogen content in membrane 7 and the membrane thickness, i.e., by the material and diffusive transport properties of the component itself. The linear increase in the graph is explained by the increased concentration of hydrogen (H₂) at the cathode electrode 11b, often referred to as supersaturation. According to current understanding, this arises because the gas formed in the electrode can only be rapidly transported away at a physically limited rate. Diffusion out of the cathode electrode 11b into the open-pore cathode compartment 5 also requires a certain concentration gradient towards the bulk concentration. This increased local concentration of hydrogen (H₂) in turn increases the driving force for diffusion through membrane 7 and the gas transfer into the anode compartment 3. Thus, according to this simplified model, the slope of the hydrogen permeation rate n H2The current density j is determined by the ratio of the gas permeability of the membrane to the permeability of the electrode.
[0122] In contrast, FIG 3 shows the hydrogen permeation rate n in a corresponding diagram. H2 as a function of the current density j for different transfer water coefficients TWK, where a characteristic curve 9 for TWK = 0 and, for comparison, two characteristic curves 9 for TWK ≠ 0 are shown.
[0123] To quantify the influence of the transfer water on the transport of hydrogen (H₂) through membrane 7 more precisely, a balance equation for hydrogen (H₂) was applied and implemented in the operating procedure of electrolysis cell 1. This equation takes into account both the diffusion of hydrogen (H₂) and the convection of hydrogen (H₂) dissolved in the transfer water. In the improved procedure, a characteristic curve (9) was determined for a specific transfer water coefficient (TWK), which represents the hydrogen permeation rate (n). H2 through the membrane 7 as a function of the 2024 PF00867
[0124] 24
[0125] Current density j at this water transfer coefficient TWK is described. In this way, a field of characteristic curves from a multitude of characteristic curves 9 can be provided for an electrode-coated membrane 7 and applied for the operation and monitoring of the transport properties of the electrolysis cell 1 over its lifetime. In the diagram of FIG. 3, discrete data points are selected from which the effective diffusivity D is determined. H2 to characterize hydrogen permeation and the supersaturated hydrogen concentration c H2,supsat The transport properties of the electrode can be determined in the cathode compartment 5. A continuous characteristic curve for a specific transfer water coefficient TWK can be generated from this using linear regression.
[0126] The data show that the transfer water coefficient (TWK), determined by the membrane, alters the slope. Consequently, it is not solely a function of the ratio of the respective gas permeability of the membrane and electrode. This finding is of particular importance for comparing and characterizing different membrane-electrode configurations and for interpreting measurement data over the lifetime of an electrolyzer, with the aim of analyzing the aging of various components and carrying out more precise service planning. The operating procedure of the invention and the underlying methodology take a further step in understanding the actual processes in the electrolysis cell 1 and increase the added value of operating data. Long-term tests and field data can thus be interpreted more effectively with regard to the aging of different components (keyword: "predictive maintenance").This improves the understanding and knowledge of actual losses and overall efficiency of the system and serves the purpose of more targeted development and optimization of future long-term stable components of an electrolysis cell 1.2024 PF00867.
[0127] 25
[0128] FIG 4 shows an electrolysis plant 10 with a circuit on the oxygen side and with a diagnostic system 21, which is set up to carry out the process during the operating time of the electrolyzer 10.
[0129] The electrolysis plant 10 is configured for the electrolysis of water (H₂O) as the reactant. The electrolysis plant 10 has only one circuit on the oxygen side. This is a simple embodiment of an electrolysis plant 10 to implement the invention. The electrolysis plant 10 has an electrolyzer 10a with an electrolysis cell stack 30, comprising a plurality of electrolysis cells 1 (not shown in detail) stacked axially. An anodic half-cell and the cathodic half-cell of each electrolysis cell 1j are separated by a membrane 7 (shown only schematically), thus forming an anode compartment 3 and a cathode compartment 5.
[0130] In this simple circuit, the electrolysis cell stack 30 is supplied with water for the electrolysis reaction, with the water simultaneously serving to cool the electrolysis cells 1. The product gas of the electrolysis, oxygen O2, is fed together with excess water H2O in a phase mixture into the gas separator 25 for oxygen O2. Phase separation takes place in the gas separator 25, and the gaseous oxygen O2 is separated from the liquid water H2O and removed from the circuit via the outlet 27 for oxygen O2. To maintain water circulation in the circuit, the circulation pump 41 is provided. Consumed water H2O is replenished by supplying demineralized water (DI water) via the feed line 29 and a controllable solenoid valve 31b. Although it is demineralized water, any minor impurities could accumulate in the circuit.To counteract this effect, a solenoid valve 31 is temporarily opened and some of the water H2O is discharged from the circuit via the drain line 33. 2024 PF00867.
[0131] 26
[0132] In the embodiment shown in Figure 4, there is no recirculation on the hydrogen soap of the electrolyzer 10a. The produced hydrogen H₂ is simply discharged via the product gas line 35 and is available for further use, for example, compression. A pressure relief valve is typically arranged in the product gas line 35, although this is not shown in detail in the embodiments. This valve serves to discharge the hydrogen H₂ at a certain overpressure, which is highly desirable in most applications for further processing of the hydrogen H₂. Since liquid water H₂O is generally produced on the hydrogen side during PEM electrolysis, a condensate line 37 is also provided, which opens when a certain amount of water has accumulated in order to discharge it. This can be implemented, for example, with a float switch.Figure 4 indicates that this water is discarded. However, it is also possible for this water to be reused for electrolysis by being recycled back into the process on the oxygen side. This recycling is generally an economically viable approach for large electrolysis plants 10. In another possible embodiment and modification of the electrolysis plant 10 shown in Figure 4, a process water cycle can also be located on the hydrogen side of the electrolysis plant 10, similar to the oxygen side shown in Figure 4.
[0133] Furthermore, a conductivity sensor 39a and a conductivity sensor 39b for measuring the specific conductivity are installed in the electrolysis plant 10. These conductivity sensors 39a and 39b are used to infer ion concentrations in the process water via a specific correlation, for example, a fluoride concentration as a degradation product of the membrane material 7.
[0134] It is particularly advantageous to place the conductivity sensors 39a, 39b at locations where the pressure 2024 PF00867
[0135] 27
[0136] The pressure within the system is as high as possible, since degassing of dissolved hydrogen (H2) or oxygen (O2) is particularly low or unlikely here. Gas bubbles would disrupt the precise measurement of conductivity and distort the result. Therefore, the conductivity sensors 39a and 39b are positioned at a geodetically low point in the electrolysis plant 10 to take advantage of hydrostatic pressure and thus effectively counteract degassing.
[0137] The specific conductivities of the fluid streams, particularly the water, are measured using conductivity sensors 39a and 39b to determine the fluoride release from the membrane over the operating time. The temporal profile of the fluoride concentration is determined, with the specific conductivity serving as a possible measure of the operational degradation of the proton-permeable membrane due to fluoride release.
[0138] In addition to the specific conductivities of the fluid streams, volumetric flow rates are quantified and balanced to determine the fluoride release rate. In principle, a volumetric flow sensor could be installed on each outgoing water stream to measure these flow rates. However, this would be very disadvantageous, as it involves considerable effort in an electrolysis plant 10, particularly regarding the cost of the flow sensors and the calibration effort, coupled with a relatively high susceptibility to errors and inaccuracies.
[0139] The invention pursues a more precise approach and proposes a highly advantageous method in which the degradation and thus the aging state of the electrode-coated membrane 7 can be determined in-situ, or at least periodically determined and updated, via a diagnostic system 20. For this purpose, the diagnostic system 20 comprises a measuring device 19 equipped with sensors and a data processing device 21. The sensors, in particular, record physical measurements from the electrolysis cells 1, which can be processed in the data processing system 21. 2024 PF00867
[0140] 28
[0141] In this way, characteristic curves 9 or a field of characteristic curves can be determined for electrolysis cells 1 by varying the current density j and the hydrogen permeation rate n. H2 through the membrane 7 as a function of the current density j and as a function of the water transport n H2O,dragThe temperature is determined from the anode compartment 3 through the membrane 7 into the cathode compartment 5 (see FIG. 1). Numerous operational sensor data from the electrolyzer 10a can be acquired in this way and read into the diagnostic system 20 via I / O interfaces for processing. A sensor calibration function is also integrated into the diagnostic system. For an electrolyzer 10a with a multitude of axially stacked electrolysis cells 1, the operation of the electrolyzer 10a can thus be monitored, the aging state of the electrolysis cells 1 can be diagnosed, and an adjustment of the operating characteristic curve 9 can be determined. It is also possible to provide a valid lifetime prediction during the development and design of a new electrolysis cell 1 and to define an initial characteristic curve 9 at the start of operation (BoL) during an acceptance measurement.Over the service life of the electrolyzer 10a until decommissioning (EoL) or a service measure, an assessment can be carried out based on regular measurement cycles and, if necessary, the characteristic curve 9 can be updated and adapted to the measured state of aging and thereby, for example, iteratively optimized with regard to the predictive quality of the remaining service life.
[0142] For this purpose, the data processing unit 21 of the electrolyzer 10a is equipped with a memory 23, i.e., the cell characteristic curves 9 can be read in or out, wherein a selected characteristic curve 9 at an operating point determines the hydrogen permeation rate n H2The operation and diagnostic procedure described and outputs the current density j through the membrane 7 at a given water transfer coefficient TWK. Implementing this procedure is particularly advantageous in an electrolysis system 10 that uses a PEM electrolyzer 10a2024 PF00867.
[0143] 29
[0144] The electrolyzer can be equipped with a proton-conducting membrane 7 (PEM electrolyzer) or with an anion-conducting membrane 7 (AEM electrolyzer). Combinations of different electrolysis types are also possible, and the operating procedure is applicable to them.
Claims
2024 PF00867 30 Patent claims 1. A method for operating an electrolysis cell (1) in which hydrogen (H2) and oxygen (O2) are produced as product gases, wherein the electrolysis cell (1) has an anode compartment (3) and a cathode compartment (5) separated by an electrode-coated membrane (7), wherein the electrolysis cell (1) is supplied with an electrolysis current (I) such that oxygen product gas is formed in the anode compartment (3) and hydrogen product gas is formed in the cathode compartment (5), wherein a foreign gas fraction of hydrogen (H2) is transferred from the cathode compartment (5) through the membrane (7) into the anode compartment (3), characterized in that the current density (j) is varied and the hydrogen permeation rate (n) H2 ) through the membrane ( 7 ) as a function of the current density ( j ) and as a function of the water transport (n ) H2O,drag) from the anode compartment ( 3 ) through the membrane ( 7 ) into the cathode compartment ( 5 ).
2. Method according to claim 1, wherein a characteristic curve ( 9 ) is determined for a specific transfer water coefficient (TWK) which determines the hydrogen permeation rate (n ). H2 ) through the membrane ( 7 ) as a function of the current density ( j ) at this water transfer coefficient (TWK).
3. A method according to claim 1 or 2, wherein data points are selected from which the effective diffusivity (D) is determined. H2 ) to characterize hydrogen permeation and the supersaturated hydrogen concentration ( c H2,supsat ) in the cathode ( 11b ) arranged in the cathode space ( 5 ) to characterize the transport property of the cathode ( 11b ).
4. The method of claim 3, wherein the supersaturated hydrogen concentration (c H2,supsat) in the cathode (11b) arranged in the cathode compartment (5) approximately from the hydrogen concentration (c) H2,cath ) in the cathode space ( 5 ) outside the cathode ( 11b ) is determined, where the hydrogen concentration2024 PF00867 31 tion ( c H 2, cath) in the cathode space ( 5 ) outside the cathode ( 11b ) is determined from the pressure and temperature in the cathode space ( 5 ).
5. A method according to one of the preceding claims, which is carried out in an electrolyzer (10a) with a plurality of axially stacked electrolysis cells (1), wherein the operation of the electrolyzer (10a) is monitored and an aging state of the electrolysis cells (1) is diagnosed.
6. Electrolyzer (10a) with a diagnostic system (20), wherein the diagnostic system (20) comprises a measuring device (19) and a data processing device (21), and wherein the diagnostic system (20) is configured to carry out the method according to one of the preceding claims.
7. Electrolyzer (10a) according to claim 6, wherein the data processing device (21) has a memory (23) into which a characteristic curve (9) can be read in or out, wherein the characteristic curve (9) at an operating point represents the hydrogen permeation rate n H2 ) through the membrane ( 7 ) as a function of the current density ( j ) at a water transfer coefficient ( TWK).
8. Electrolyzer according to claim 7, characterized by a configuration as a PEM electrolyzer with a proton-conducting membrane ( 7 ) or as an AEM electrolyzer with an anion-conducting membrane ( 7 ).