A health state prediction method for a hydrogen production device by electrolysis of water
By decomposing voltage and current data, a two-dimensional evaluation system for electrochemical performance and physical structural integrity was established, solving the problem of identifying microstructural damage in water electrolysis hydrogen production devices under dynamic operating conditions, and realizing accurate health status prediction and proactive maintenance of water electrolysis hydrogen production devices.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA UNIV OF PETROLEUM (EAST CHINA)
- Filing Date
- 2026-03-13
- Publication Date
- 2026-05-29
AI Technical Summary
Existing technologies are unable to effectively predict sudden short circuits or short-circuit faults caused by physical structural damage under dynamic operating conditions. Traditional voltage signal assessment methods cannot identify microstructural damage, leading to safety hazards.
By collecting operational data from a water electrolysis hydrogen production device, decomposing voltage data into overpotential components, and combining current data to calculate microstructural damage parameters, a two-dimensional evaluation system of electrochemical performance and physical structural integrity is established. A health status prediction is achieved using mapping functions and weighted fusion methods.
It enables early warning of microstructural damage before the macroscopic voltage deteriorates significantly, accurately predicts sudden failures of water electrolysis hydrogen production units, provides comprehensive and accurate health status assessment, and avoids misjudgment of status and potential risks.
Smart Images
Figure CN122105524A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of electrochemical hydrogen production technology, and particularly relates to a method for predicting the health status of a water electrolysis hydrogen production device. Background Technology
[0002] Against the backdrop of large-scale utilization of renewable energy, water electrolysis for hydrogen production has gradually become a key technology for absorbing fluctuating power sources such as wind and solar due to its advantages such as high energy conversion efficiency, zero carbon emissions, and suitability for long-term energy storage. However, under the dynamic operating conditions of wind and solar coupling, the electrolyzer needs to frequently cope with fluctuations in input power, making its operating conditions far more complex than traditional steady-state conditions. Furthermore, the frequent potential jumps, damp-heat cycles, and the coupling effect of mechanical stress significantly accelerate the performance degradation and microstructural damage of membrane electrode assemblies and other key components.
[0003] Currently, the industrial assessment of the health status of water electrolysis hydrogen production units mainly relies on the monitoring of macroscopic electrochemical signals, particularly by monitoring the rate of change of the total voltage of the electrolyzer over time to estimate the extent of performance degradation. However, this assessment method based on a single voltage signal reveals significant limitations under actual dynamic operating conditions. On the one hand, voltage signals are the result of the superposition of multiple physicochemical processes, which can be misleading. For example, in proton exchange membrane (PEM) water electrolysis, the PEM thins under the attack of chemical free radicals. Although this is physically structural damage, it initially reduces ohmic resistance, thus appearing as performance improvement or a delayed voltage rise on the voltage curve, masking the potential risk of structural degradation. On the other hand, damage to microstructures is often insidious and sudden, especially phenomena such as physical peeling at the interface between the catalyst layer and the PEM, microcrack propagation, and catalyst dissolution, which are difficult to identify on voltage signals. Furthermore, experimental studies have shown a misalignment between the decline in structural health and the degradation of electrochemical performance. When a significant inflection point appears in the voltage signal, irreversible physical damage has often accumulated inside the unit, bringing it close to the brink of failure. This asynchrony between macroscopic performance and microscopic structural evolution is known as the "performance-structure gap." Therefore, existing evaluation methods cannot quantify this gap, making it difficult to effectively warn of sudden short circuits or short-circuit faults caused by interface decoupling or membrane perforation under dynamic operating conditions, thus posing a threat to the safe operation of the system. Summary of the Invention
[0004] To address the difficulty in effectively predicting sudden short circuits or short-circuit faults caused by physical structural damage under dynamic operating conditions in the existing technology, this invention provides a health status prediction method for a water electrolysis hydrogen production device, comprising the following steps: Step S1: Collect operating data of the water electrolysis hydrogen production device, wherein the operating data includes at least voltage data, current data and operating time; Step S2: Decompose the voltage data into overpotential components, and calculate the electrochemical performance score based on the maximum voltage decay threshold and the overpotential components; Step S3: Based on the current data and running time, calculate the cumulative damage amount of the microstructure damage parameters, and convert the cumulative damage amount into a physical structure integrity score through a mapping function. Step S4: The electrochemical performance score and the physical structure integrity score are weighted and fused to obtain the health status of the electrochemical hydrogen production device.
[0005] Based on the above scheme, step S2, based on the electrochemical performance prediction sub-model, obtains the electrochemical performance score by calculating the proportion of the overpotential component to the maximum voltage decay threshold. The electrochemical performance prediction sub-model is as follows: , in, The overpotential components include activation overpotential, ohmic overpotential, and mass transfer overpotential, w k These are the weighting coefficients corresponding to the overpotential components. This is the maximum voltage decay threshold.
[0006] Furthermore, step S3 specifically includes: S31: Calculate the cumulative damage based on the physical-chemical coupled dynamics model, which represents the influence of different operating conditions on damage evolution through the operating condition modulation coefficient and the stress term driving structural damage. S32: By introducing a shape coefficient, the cumulative damage amount is converted into a health sub-score of the microstructural damage parameter, and the physical structural integrity score is obtained based on the contribution of the microstructural damage parameter to the physical structural integrity.
[0007] Based on the above scheme, in step S32, the accumulated damage is transformed into a health sub-score of the microstructural damage parameter by introducing a shape coefficient, specifically using the following nonlinear mapping function: , Among them, s k (t) represents the health sub-score corresponding to the k-th microstructural damage parameter. Damage tolerance factor This is the failure sensitivity factor.
[0008] Preferably, step S3 further includes failure determination, specifically including: When the cumulative damage amount of any microstructural damage parameter reaches or exceeds the corresponding physical failure threshold, the health sub-score of the damage parameter is set to 0, and the physical structure of the water electrolysis hydrogen production device is determined to be faulty; the physical failure threshold includes the cumulative membrane thickness reduction threshold, the cumulative interfacial gap increase threshold, and the cumulative catalyst dissolution threshold.
[0009] Specifically, the microstructure damage parameters include at least the interfacial gap between the proton exchange membrane and the catalyst layer, the thickness of the proton exchange membrane, the amount of anolyte catalyst dissolved, the amount of cathode catalyst dissolved, and the passivation degree of the porous diffusion layer.
[0010] Specifically, the health status in step S4 is: , in, For electrochemical performance score, Score for physical structural integrity. , These are the preset weights for the electrochemical performance score and the physical structure integrity score, respectively. .
[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention can effectively identify "performance-structure gaps", that is, provide early warning of fatal damage to the microstructure before the macroscopic voltage deteriorates significantly, thereby achieving accurate prediction and proactive maintenance of sudden failures of water electrolysis hydrogen production devices under different operating conditions; 2. This invention constructs a dual-dimensional evaluation system that integrates electrochemical performance and physical structural integrity. By decoupling the analysis of macroscopic voltage decay and microstructural damage, it solves the problem of misjudgment of state caused by relying solely on voltage signals in traditional methods, and can more comprehensively and accurately reflect the true health level of the water electrolysis hydrogen production device. 3. This invention introduces damage tolerance factor and failure sensitivity factor to establish a normalized mapping function, thereby achieving a unified evaluation of physical quantities with different dimensions. This not only shields the influence of initial minor damage on the score, but also simulates the suddenness of structural failure, thus realizing the integrity assessment of the physical structure. 4. The device health status index is synthesized by combining the electrochemical performance score and the structural integrity score through preset weighting coefficients. It can not only quantify the current performance level, but also reveal the potential risks of the internal structure through voltage appearance, so as to accurately capture the "performance-structure gap" and scientifically define the end of the life. Attached Figure Description
[0012] Figure 1 This is a flowchart illustrating the health status prediction of the water electrolysis hydrogen production device of the present invention. Figure 2This is a schematic diagram of the aging process of the PEMWE device under multiple operating conditions according to the present invention; Figure 3 This is a schematic diagram of the operating conditions for this invention. Figure 4 This is a schematic diagram of the voltage evolution curves under multiple operating conditions according to the present invention; Figure 5 This is a schematic diagram showing the comparison of health status prediction under multiple operating conditions over 500 hours according to the present invention. Figure 6 This is a radar image showing the physical structural integrity of the present invention under multiple operating conditions over 500 hours. Figure 7 This is a schematic diagram illustrating the changes in health status scores under multiple operating conditions over 50,000 hours according to the present invention. Figure 8 This is a schematic diagram showing the changes in electrochemical performance and physical structural integrity under multiple operating conditions over 50,000 hours. Figure 9 This is a radar image showing the physical structural integrity of the present invention under multiple operating conditions for 50,000 hours. Detailed Implementation
[0013] The invention will be further described below with reference to specific embodiments.
[0014] Example 1 This embodiment provides a health status prediction method for a water electrolysis hydrogen production device, applicable to the health status assessment and failure early warning of proton exchange membrane water electrolysis (PEMWE) under various dynamic operating conditions. This method solves the technical problems of traditional single-voltage assessment being unable to identify latent structural damage and predict sudden failures by integrating macroscopic electrochemical performance and microstructural integrity. Figure 1 As shown, the method includes the following steps: First, data acquisition and initialization are performed through step S1, specifically as follows: Set initial parameters, including the maximum allowable voltage degradation value, physical failure threshold, damage tolerance factor, and failure sensitivity factor; Collect the current operating data of the water electrolysis hydrogen production unit. This operating data includes at least voltage data U, current data i, and operating time t.
[0015] Secondly, electrochemical performance and physical structure integrity are scored based on the collected data in steps S2 and S3: Step S2: Decompose the voltage data into overpotential components, and calculate the electrochemical performance score based on the maximum voltage decay threshold and the overpotential components; In this embodiment, step S2 specifically includes: Step S21: Decouple the voltage data into activation overpotential, ohmic overpotential, and mass transfer overpotential components using polarization curve decomposition technology: Specifically, based on the current-voltage relationship, the formula for calculating the total voltage of water electrolysis is as follows: , in, V cell This represents the total voltage (electrolytic cell voltage U). E rev Indicates reversible potential. η kin Indicates activation overpotential, η ohm Indicates the ohmic overpotential. η mass It represents the mass transfer overpotential.
[0016] In step S21, the voltage data U is decomposed into activation, ohmic, and mass transfer overpotentials through impedance decoupling. Among them, the reversible potential is related to the Gibbs free energy of the reaction and is affected by temperature. Under standard atmospheric pressure, it is calculated using the following standard reversible potential formula: , In the formula, T is the thermodynamic temperature of the test system.
[0017] Step S21 decomposes the total voltage into overpotential components from different physical sources, which can quantify the contribution of each decay mechanism to macroscopic performance and avoid the confusion of a single voltage signal.
[0018] Step S22: Based on the electrochemical performance prediction sub-model, the electrochemical performance score reflecting the current electrochemical efficiency is obtained by calculating the proportion of the overpotential component to the maximum voltage decay threshold.
[0019] The electrochemical performance prediction sub-model is as follows: , in, To score electrochemical performance, The overpotential components include activation overpotential, ohmic overpotential, and mass transfer overpotential. w k These are the weighting coefficients corresponding to the overpotential components. This is the maximum voltage decay threshold.
[0020] The electrochemical performance prediction sub-model normalizes the electrochemical efficiency performance score by using the maximum voltage decay threshold. The electrochemical performance score is a phenomenological index directly determined by the measured voltage data input. As the electrolyzer ages, the voltage required to reach a specific current density increases, thus the electrochemical performance score decreases.
[0021] like Figure 2 and Figure 3 As shown, membrane electrode assemblies undergo aging under various operating conditions, such as proton exchange membrane (PEM) degradation, catalyst dissolution, and passivation of the porous transport layer (PTL). These damages lead to a decline in device performance. This invention achieves a quantitative assessment of hidden structural damage through step S3, solving the problem of identifying the "performance-structural gap".
[0022] Step S3: Calculate the cumulative damage amount of the microstructure damage parameters under current data and running time, and convert the cumulative damage amount into a physical structure integrity score through a mapping function.
[0023] The microstructural damage parameters in this embodiment include at least the interfacial gap P between the proton exchange membrane and the catalyst layer. gap Film thickness loss P thin Anode catalyst solubility P diss,Ir Cathode catalyst solubility P diss,Pt and the passivation degree P of the porous diffusion layer PTL Step S3 specifically includes: Step S31: Calculate the cumulative damage based on the physical-chemical coupled kinetic model: , in, For changes in microstructural damage parameters, This reflects the rapid structural rearrangement that occurred during the early stages of operation of the membrane electrode assembly. As a pre-factor, The linear loss reflects the loss during stable operation, where v k For P k The rate of degradation.
[0024] According to this embodiment, the change in cumulative damage is defined as: , where C k is the operating condition modulation coefficient, used to reflect the acceleration or inhibition factor of different operating conditions on a specific degradation path. f(t) is the physical / chemical stress term that drives structural damage, used to distinguish the evolution law of different damage types, as shown in Table 1 and Table 2.
[0025] Table 1 Definition of Modulation Coefficient under Operating Conditions
[0026] Table 2 Definition of f(t)
[0027] According to step S31 above, the cumulative damage amount is not directly equivalent to the physical structural integrity score SOH. structTherefore, step S32 is executed.
[0028] S32: Health sub-score that converts cumulative damage into microstructural damage parameters by introducing a shape factor. The shape factor includes damage tolerance factor a and failure sensitivity factor b; In step S32, the cumulative damage is transformed into a health sub-score of the microstructural damage parameters by introducing a shape coefficient, specifically using the following nonlinear mapping function: , Among them, s k (t) represents the health sub-score corresponding to the k-th microstructural damage parameter. Representing the damage tolerance factor, it can mask the effects of early minor injuries; the higher the value, the higher the SOH. struct The flatter the curve is in the initial stage; It is the failure sensitivity factor, representing the rate of change when entering the failure stage. The smaller the value, the lower the SOH. struct The curve shows a more pronounced sharp decline in the later stages.
[0029] Furthermore, a sub-model for predicting physical structural integrity is derived based on the contribution of microstructural damage parameters to the physical structural integrity: , in, Score for physical structural integrity. , , , , The health sub-scores for each component (interfacial gap, film thickness loss, porous diffusion layer passivation, anode catalyst dissolution, and cathode catalyst dissolution) are respectively. , , , , These are the contribution weights of each component to the physical structure integrity score.
[0030] In this embodiment, step S3 further includes failure determination, used to simulate voltage surges when the MEA physical structure fails: A preset physical failure threshold is set, which includes the cumulative reduction in film thickness, the cumulative increase in interfacial gap, and the cumulative dissolution of catalyst.
[0031] When the cumulative membrane thickness reduction threshold is reached, it is determined that membrane perforation has occurred, leading to a short circuit in the device; where the cumulative membrane thickness reduction is: In this embodiment, an N117 membrane is used, and the cumulative membrane thickness reduction threshold is set to 180 μm. Membrane perforation occurs when the membrane thickness is ≥ 180 μm, indicating a short circuit in the PEMWE. At this point, s thin =0.
[0032] When the cumulative increase in interfacial gap or the cumulative dissolution of catalyst threshold is reached, it is determined that the interfacial contact has been broken or the catalyst has been depleted, causing the device to disconnect; whereby the cumulative increase in interfacial gap is: In this embodiment, the threshold for the cumulative increase in interface gap is 20 μm. When the thickness is ≥ 20 μm, the three-phase interface is determined to be completely disconnected, and a PEMWE circuit is broken. s gap = 0.
[0033] Cumulative catalyst dissolution A diss,Ir / Pt In this embodiment, the cumulative catalyst dissolution threshold is 90%. A diss,Ir / Pt When the catalyst is ≥ 90%, it is considered depleted, and a PEMWE circuit breaks down. s diss,Ir / Pt = 0.
[0034] When the cumulative damage of any microstructural damage parameter reaches or exceeds the corresponding physical failure threshold, the health sub-score of the damage parameter is set to 0, and the physical structure of the water electrolysis hydrogen production device is determined to be faulty. Step S4: The electrochemical performance score and the physical structure integrity score are weighted and fused to obtain the health status of the electrochemical hydrogen production device.
[0035] Specifically, the health status prediction model in step S4 is as follows: , Among them, SOH MEA For a healthy state, For electrochemical performance score, Score for physical structural integrity. , These are the preset weights for the electrochemical performance score and the physical structure integrity score, respectively.
[0036] In this embodiment, considering that the industry places greater emphasis on electrochemical performance, but structural risks cannot be ignored, therefore a configuration is set... .
[0037] Example 2 This embodiment uses the method and steps in Embodiment 1 to test the electrolytic hydrogen water production device. The parameter settings and results are as follows: In the electrochemical performance prediction sub-model, the maximum voltage decay threshold is set to 500 mV, assuming that each overpotential component contributes consistently to the electrochemical performance score. w kin = w ohm = w mass = 1; The damage tolerance factor in the nonlinear mapping function used to calculate the health sub-score is set to: a gap = a thin = a diss,Ir = a diss,Pt = 2, a PTL = 1; Failure sensitivity factor is set to: b gap = b thin = b diss,Ir = b diss,Pt = 0.5, b PTL = 1; In the physical structural integrity prediction sub-model, it is assumed that each component contributes equally to the physical structural integrity score. = = = = = 1; set in the health status prediction model w perf = 0.6, w struct = 0.4.
[0038] First, a 500-hour prediction was performed on the PEMWE device. The voltage evolution curves of the device under various operating conditions are shown below. Figure 4 As shown, health status prediction is as follows: Figure 5 As shown, from Figure 5 As can be seen from (a), the health status of the device declines with time, and the degree of device health degradation varies under different operating conditions. Figure 5 (b) reflects the comparison of the electrochemical performance and physical structural integrity of the device after 500 hours of operation, with performance scores varying under different operating conditions. The physical structural integrity under various operating conditions is shown below. Figure 6 As shown, the film thickness loss is not significantly different under different operating conditions, while the integrity of other physical structures differs significantly under different operating conditions.
[0039] like Figure 7As shown, a longer-term prediction (50,000 hours) of the PEMWE device reveals that a sudden change in health status occurs at a certain point in time, and the timing of this change varies depending on the operating conditions. Furthermore, the changes in health status, electrochemical performance, and physical structural integrity corresponding to the long-term prediction are as follows: Figure 8 As shown, the changes in electrochemical performance and physical structural integrity differ significantly under different operating conditions; the physical structural integrity under multiple operating conditions is as follows: Figure 9 As shown, each physical structure undergoes significant changes during long-term operation. It should be noted that although the physical structure integrity is not zero under constant current conditions, according to the present invention, in order to provide early warning before the macroscopic voltage deteriorates significantly, the component is judged to be faulty when the failure state determination threshold is reached, the physical structure integrity score is 0, and the health status index is 0.
[0040] According to the present invention, the health status index of the water electrolysis hydrogen production device is composed of both electrochemical performance score and structural integrity score. It can not only quantify the current performance level, but also reveal potential risks in the internal structure through voltage appearance, and more comprehensively and accurately reflect the true health level of the water electrolysis hydrogen production device.
[0041] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
[0042] While the specific embodiments of the present invention have been described above, they are not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.
Claims
1. A method for predicting the health status of a water electrolysis hydrogen production device, characterized in that, Includes the following steps: Step S1: Collect operating data of the water electrolysis hydrogen production device, wherein the operating data includes at least voltage data, current data and operating time; Step S2: Decompose the voltage data into overpotential components, and calculate the electrochemical performance score based on the maximum voltage decay threshold and the overpotential components; Step S3: Based on the current data and running time, calculate the cumulative damage amount of the microstructure damage parameters, and convert the cumulative damage amount into a physical structure integrity score through a mapping function. Step S4: The electrochemical performance score and the physical structure integrity score are weighted and fused to obtain the health status of the electrochemical hydrogen production device.
2. The method for predicting the health status of a water electrolysis hydrogen production device according to claim 1, characterized in that, Step S2, based on an electrochemical performance prediction sub-model, obtains the electrochemical performance score by calculating the proportion of the overpotential component to the maximum voltage decay threshold. The electrochemical performance prediction sub-model is as follows: ， in, The overpotential components include activation overpotential, ohmic overpotential, and mass transfer overpotential, w k These are the weighting coefficients corresponding to the overpotential components. This is the maximum voltage decay threshold.
3. The health status prediction method for the water electrolysis hydrogen production device according to claim 1, characterized in that, Step S3 specifically includes: S31: Calculate the cumulative damage based on the physical-chemical coupled dynamics model, which represents the influence of different operating conditions on damage evolution through the operating condition modulation coefficient and the stress term driving structural damage. S32: By introducing a shape coefficient, the cumulative damage amount is converted into a health sub-score of the microstructural damage parameter, and the physical structural integrity score is obtained based on the contribution of the microstructural damage parameter to the physical structural integrity.
4. The method for predicting the health status of a water electrolysis hydrogen production device according to claim 3, characterized in that, In step S32, the cumulative damage is converted into a health sub-score of the microstructural damage parameter by introducing a shape coefficient, specifically using the following nonlinear mapping function: ， Among them, s k (t) represents the health sub-score corresponding to the k-th microstructural damage parameter. Damage tolerance factor This is the failure sensitivity factor.
5. The method for predicting the health status of a water electrolysis hydrogen production device according to claim 3, characterized in that, Step S3 further includes failure determination, specifically including: When the cumulative damage amount of any microstructural damage parameter reaches or exceeds the corresponding physical failure threshold, the health sub-score of the damage parameter is set to 0, and the physical structure of the water electrolysis hydrogen production device is determined to be faulty; the physical failure threshold includes the cumulative membrane thickness reduction threshold, the cumulative interfacial gap increase threshold, and the cumulative catalyst dissolution threshold.
6. The method for predicting the health status of a water electrolysis hydrogen production device according to claim 1, characterized in that, The microstructure damage parameters include at least the interfacial gap between the proton exchange membrane and the catalyst layer, the thickness of the proton exchange membrane, the amount of anode catalyst dissolved, the amount of cathode catalyst dissolved, and the passivation degree of the porous diffusion layer.
7. The method for predicting the health status of a water electrolysis hydrogen production device according to claim 1, characterized in that, The health status in step S4 is: ， in, For electrochemical performance score, Score for physical structural integrity. , These are the preset weights for the electrochemical performance score and the physical structure integrity score, respectively. .