A method for qualitatively and quantitatively evaluating leakage frequency of vehicle-mounted hydrogen supply system

Abstract

The present application relates to a kind of vehicle-mounted hydrogen supply system leakage frequency qualitative and quantitative evaluation method.The method described in the present application identifies the safety risk unit of vehicle-mounted hydrogen system, the frequency of leakage of vehicle-mounted hydrogen system is qualitatively and quantitatively evaluated by establishing matching fault tree, the occurrence probability of each bottom event of fault tree is determined using the method of subjective and objective combination, the leakage probability of entire vehicle-mounted hydrogen system is quickly and accurately estimated, which lays a foundation for subsequent calculation and analysis of vehicle-mounted hydrogen system risk expression, and greatly helps to quickly and accurately quantify the risk of hydrogen leakage of vehicle-mounted hydrogen system.The present application has important significance for preventing the occurrence of accident after leakage of vehicle-mounted hydrogen system and formulating emergency treatment measures for accident, and provides reference basis for safety design of vehicle-mounted hydrogen system, also has positive effect on safety in the promotion of hydrogen fuel cell vehicle.For formulating relevant safety specifications, industry standards and the like, it also has important significance.

Description

Technical Field

[0001] This invention relates to the field of hydrogen leakage risk assessment, specifically a qualitative and quantitative assessment method for the leakage frequency of an on-board hydrogen supply system. Background Technology

[0002] With the massive consumption of non-renewable energy sources such as coal, oil, and natural gas, the advantages of hydrogen energy—its renewability, pollution-free nature, high energy efficiency, and high safety—have been amplified, and countries are increasingly emphasizing this potential future primary energy source. While hydrogen energy offers numerous advantages, as a high-density energy carrier, its use often faces many safety challenges. Over the past few decades, several serious accidents have occurred during hydrogen use. The factors leading to these catastrophic events can be categorized as follows: mechanical or material failure, corrosion, overpressure, embrittlement of storage and transportation facilities at low temperatures, vapor explosions caused by liquid expansion, ruptures caused by shock waves and impacts from nearby explosive fragments, and human error. Not all hydrogen safety accidents cause fatal injuries; often, the damage is within acceptable limits. Therefore, timely safety risk assessments of hydrogen systems are crucial. Safety risk assessments mainly consist of frequency assessment and consequence assessment. Since current data on the failure probability of onboard hydrogen systems is still incomplete, reasonable frequency estimation and qualitative and quantitative analysis of onboard hydrogen systems are essential.

[0003] In domestic and international research, hydrogen system risk assessment methods are mainly divided into two types: Rapid Risk Ranking (RRR) and Quantitative Risk Assessment (QRA). RRR is a qualitative assessment, while QRA is a quantitative assessment.

[0004] Risk Response Ratio (RR) is an empirical qualitative risk assessment method. It compares the results obtained after expert analysis and discussion with a risk matrix to obtain the corresponding risk level, which can quickly identify the main hazards. The main operational steps of the RR method in hydrogen system risk assessment are as follows: First, the risk of the target is divided into 5 levels using hazard occurrence probability analysis. Then, the target risk is also divided into 5 levels using consequence analysis. This results in a 5×5 risk matrix. By comparing this matrix with the risk matrix obtained through expert discussion, three risk levels—high, medium, and low—can be determined.

[0005] The RRR method is widely used due to its simplicity and speed, but because of its strong subjectivity and reliance on expert experience, it can only be applied to rough qualitative analysis. For a more in-depth analysis and evaluation of existing risks, quantitative analysis methods are still necessary. Quantitative risk assessment is currently the most widely used method for hydrogen safety assessment, but it still has the following problems: there is a lack of statistically significant failure and leakage frequency data for hydrogen equipment, and the applicability of general leakage frequencies still needs further research; directly using them will lead to significant errors.

[0006] The QRA (Quality Assurance Response) method, based on the accident consequence calculation method, comprehensively considers the frequency and consequences of accident scenarios, quantifies risks, and determines safe distances using the acceptable range of quantified risk. The basic process includes: hazard analysis of the assessment unit, failure scenario analysis, accident consequence analysis, individual risk calculation, and safe distance determination. Hydrogen infrastructure and its application in transportation are relatively new fields; therefore, available data on hydrogen-related failures and accidents are limited, which is the main reason why reliable QRAs for hydrogen remain challenging.

[0007] Despite the difficulties in implementing QRA (Quality, Reliability, and Availability) hydrogen infrastructure, many papers have discussed it to some extent and attempted to improve the field. Some methods only consider the leakage problem of a single component to represent the leakage situation of the entire onboard hydrogen system: [Ke Sun, Zhiyong Li. Quantitative risk analysis of life safety and financial loss for road accident of fuel cell vehicle[J]. International Journal of Hydrogen Energy,2019,44(17)] analyzed the typical accident development process of hydrogen fuel cell vehicles in road collision accidents, used the event tree method to analyze the main hydrogen consequences, and fixed the cause of hydrogen leakage to the TPRD valve; [Mohammad Dadashzadeh, Sergii Kashkarov, Dmitriy Makarov, Vladimir Molkov. Risk assessment methodology for onboard hydrogen storage[J]. International Journal of Hydrogen Energy,2018,43(12)] only considered the risk assessment problem after the hydrogen storage tank leaked, without considering the safety problems caused by other components of the entire onboard hydrogen system. Some methods also employ fault tree analysis to qualitatively and quantitatively analyze hydrogen system leaks; [Yang Taihua, Qin Jing, Li Zhixiang. Safety Risk Assessment of New Energy Vehicle Hydrogen Refueling Stations Based on Fault Tree and Improved Grey Relational Analysis [J]. Modern Electronics Technology, 2021, 44(2): 115-121] The improved grey relational analysis method semi-quantitatively analyzes the safety risks of hydrogen systems; [Zhou Yangning, Chen Min. Fault Tree Analysis of Hydrogen-Oxygen Fuel Cell Modules [J]. Marine Electric Technology, 2017, 37(08): 17-21] Analyzes The composition of the hydrogen system is analyzed from the components to determine the probability of hydrogen leakage. However, the probability of each bottom event is determined through multiple rounds of reliability allocation and prediction. However, the method for determining the probability of each bottom event is still unclear and is mainly based on estimation. [Wang Wenzheng, Chen Liang, He Huizong, et al. Quantitative analysis of LNG storage tank leakage using triangular fuzzy fault tree [J]. Safety, Health and Environment, 2021, 21(10):38-44] Fuzzy fault tree was used to analyze the leakage problem, but it cannot avoid the drawback of being too subjective. Summary of the Invention

[0008] To address the shortcomings of existing technologies, the present invention aims to provide a qualitative and quantitative assessment method for the leakage frequency of on-board hydrogen supply systems. This method effectively utilizes data from a database and employs an expert judgment method based on the analytic hierarchy process (AHP) to determine the fuzzy failure rate. It effectively combines subjective and objective data to determine the basic probability of bottom-level events in the fault tree, making the failure probabilities of these bottom-level events more convincing.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0010] A method for qualitative and quantitative assessment of leakage frequency in an on-board hydrogen supply system, characterized by comprising the following steps:

[0011] Step 1: Perform structural analysis on the on-board hydrogen system to identify components that may leak, and determine hazardous areas and risk units;

[0012] Step 2: Based on the hazardous locations and risk units identified in Step 1, establish a fault tree for the on-board hydrogen system leak.

[0013] Step 3: Perform a qualitative analysis on the vehicle hydrogen system leakage fault tree established in Step 2, and find all the minimum cut sets or minimum path sets of the above fault tree.

[0014] Step 4: Quantitatively analyze the on-board hydrogen system leakage fault tree established in Step 2, and use expert judgment to determine the subjective failure rate f of all basic events. n This refers to the fuzzy failure rate; the objective failure rate f of all basic events is obtained by querying the database and acquiring failure experiment data. m The databases mentioned above include the OREDA database or the HyRAM+ database, and the failure test data mentioned above are from the manufacturers of on-board hydrogen supply system components.

[0015] Step 5, failure rate f n This refers to the fuzzy failure rate; the objective failure rate f of all basic events is obtained by querying the database and acquiring failure experiment data. m The databases mentioned above include the OREDA database or the HyRAM+ database, and the failure test data mentioned above are from the manufacturers of on-board hydrogen supply system components.

[0016] Step 5, calculate the subjective failure rate f for each basic event obtained in Step 4. n and objective failure rate f m Each is assigned a weight to ensure that the sum of the subjective failure rate weight and the objective failure rate weight for each basic event is 1. Then, the failure rate f for each basic event is determined according to the following formula; finally, the failure rates of all basic events are obtained.

[0017] f = f m ·wm +f n ·w n ;

[0018] In the above formula, W n and W m These are the subjective failure rate weights and the objective failure rate weights, respectively.

[0019] Step 6: Using the fault tree analysis software AutoFTA, perform top event probability analysis, probability importance analysis, and critical importance analysis on the failure rates of all basic events obtained in Step 5 to obtain the probability of the top event, namely, leakage of the on-board hydrogen system.

[0020] If the failure rate of the basic events is known, the probability, probability importance, and critical importance of the top event can be directly calculated using the AutoFTA software described above. The probability of the top event is the probability of hydrogen system leakage; the probability importance indicates the degree to which the change in the probability of the basic event causes a change in the probability of the top event, i.e., the contribution of the change in the probability of the basic event to the change in the probability of the top event; the critical importance indicates the rate of change in the probability of the i-th basic event causing a change in the probability of the event in the term. It is a standard for measuring the importance of each basic event from both sensitivity and probability perspectives. Critical importance can reflect the influence of the probability of each event on the top event when the probabilities of the basic events are not equal.

[0021] Based on the above plan,

[0022] The hazardous parts and risk units mentioned in step 1 include: hydrogen storage cylinders, hydrogen supply pipelines, one-way valves, solenoid valves, overflow valves, pressure reducing valves, sensors, joint connections, and accidental collisions.

[0023] Based on the above scheme, the subjective failure rate f of all basic events described in step 4 is obtained. n The specific steps are as follows:

[0024] Step 4-1: Consult n experts to obtain fuzzy linguistic values ​​for the probability of a certain basic event occurring from each expert. The fuzzy linguistic values ​​include "very small", "small", "relatively small", "medium", "relatively large", "large", and "very large".

[0025] Step 4-2: Based on the fuzzy language values ​​obtained in Step 3-1, establish the corresponding membership function: f VS (x), f S (x), f RS (x), f M (x), f RL (x), f L (x), f VL (x); The upper and lower limits of the α-cut corresponding to each of the above membership functions are:

[0026] Very small: VS α =[vs1,vs2], vs1=0, vs2=0.2-0.1α;

[0027] Small: S α =[s1,s2], s1=0.1α+0.1, s2=0.3-0.1α;

[0028] Smaller: RS α =[rs1,rs2], rs1=0.1α+0.2, rs2=0.5-0.1α;

[0029] Chinese: M α =[m1,m2], m1=0.1α+0.4, m2=0.6-0.1α;

[0030] Larger: RL α =[rl1,rl2], rl1=0.1α+0.5, rl2=0.8-0.1α;

[0031] Large: L α =[l1,l2], l1=0.1α+0.7, l2=0.9-0.1α;

[0032] Very large: VL α =[vl1,vl2], vl1=0.1α+0.8, vl2=1;

[0033] Step 4-3: Use the analytic hierarchy process (AHP) to assign weights w to the fuzzy linguistic values ​​of each expert regarding the probability of a certain basic event. i i takes values ​​from 1 to n; the membership function f corresponds to each expert opinion. i (x), where i takes values ​​from 1 to n; the comprehensive fuzzy number of a certain basic event is obtained according to the following formula:

[0034] f W =max[w1·f1(x)∧w2·f2(x)∧w3·f3(x)∧w4·f4(x)∧…∧w n ·f n (x)];

[0035] Step 4-4: Convert the comprehensive fuzzy number obtained in step 4-3 into a fuzzy probability value (FPS) based on the left and right fuzzy sorting method.

[0036] FPS = [FPS] R (w)+1-FPS L (w)] / 2;

[0037] In the above formula, FPS L(w) represents the fuzzy number, the left fuzzy probability value, and FPS. R (w) represents the right fuzzy probability value of the fuzzy number;

[0038] Step 4-5: Transform the fuzzy probability value obtained in step 4-4 into the fuzzy failure rate FFR, which is the subjective failure rate f of a certain basic event. n ;

[0039]

[0040]

[0041] Steps 4-6: Repeat steps 4-1 to 4-5 to obtain the subjective failure rate of all basic events.

[0042] The qualitative and quantitative assessment method for leakage frequency of an on-board hydrogen supply system described in this invention has the following advantages:

[0043] (1) Fault tree analysis has clear logic and can analyze the possible causes of leakage in the vehicle hydrogen system more quickly. It is more targeted to the structure of the vehicle hydrogen system and its unique leakage causes. The basic event is also more replaceable, and the fault tree can be flexibly changed to adapt to different vehicle hydrogen system structures.

[0044] (2) The basic probability of the bottom event in the fault tree is determined by combining subjective and objective methods, effectively utilizing data from similar databases. An expert judgment method based on the analytic hierarchy process (AHP) is also used to determine the fuzzy failure rate. The effective combination of these two types of data makes the failure probability of the bottom event more convincing. By using fault tree analysis software, the leakage frequency of the on-board hydrogen system can be quickly assessed qualitatively and quantitatively, laying the foundation for subsequent risk assessment of the on-board hydrogen system. Attached Figure Description

[0045] The present invention includes the following figures:

[0046] Figure 1 Flowchart of the evaluation method described in this invention;

[0047] Figure 2 Schematic diagram of the vehicle-mounted hydrogen system described in this invention;

[0048] Figure 3 The vehicle-mounted hydrogen system leakage fault tree described in this invention;

[0049] Figure 4 The hierarchical model of expert estimation capability described in this invention;

[0050] Figure 5 The membership function and fuzzy number relationship diagram described in this invention. Detailed Implementation

[0051] The present invention will be further described in detail below with reference to the accompanying drawings.

[0052] First, a structural analysis of the on-board hydrogen system is conducted to analyze the components that make up the system, identify potential leak points, and determine hazardous areas and risk units.

[0053] The on-board hydrogen supply system consists of a hydrogen storage module, a hydrogen filling module, and a hydrogen supply module. The hydrogen storage module mainly consists of a hydrogen storage cylinder and a combination valve at the cylinder opening; the hydrogen filling module mainly consists of a hydrogen filling port and a hydrogen filling pipeline; and the hydrogen supply module mainly consists of a valve and a hydrogen supply pipeline.

[0054] The hydrogen storage cylinder is also equipped with accessories such as a combination cylinder valve, a pressure relief valve, and sensors. The combination cylinder valve consists of a cylinder safety valve, a cylinder solenoid valve, and a manual shut-off valve. Hydrogen gas flows from the storage cylinder to the fuel cell through the hydrogen supply system pipeline, which is also an important component of the on-board hydrogen supply system.

[0055] Pipelines can be subdivided by function into hydrogen supply pipelines, hydrogen filling pipelines, and venting pipelines. To stably supply hydrogen to fuel cells at suitable pressure, temperature, and flow rate, the on-board hydrogen supply system relies heavily on numerous valves along the pipeline, including solenoid valves, pressure reducing valves, pressure regulating valves, overflow valves, check valves, and filter valves. Connectors are the components that link the pipelines to the valves. On-board hydrogen supply systems often use compression fittings, which are simple in structure, easy to install, and offer good sealing. Their performance is assessed through gripping ability, airtightness, and vibration resistance.

[0056] The main causes of hydrogen leakage are hydrogen embrittlement and poor system sealing. With the development of composite material cylinders and continuous improvement in fiber winding technology, the possibility of hydrogen leakage from the storage cylinder itself is low. On-board hydrogen supply systems operate under certain pressure and impact / vibration conditions, therefore, the possibility of hydrogen leakage at the connection points of the supply pipeline, valves, and joints is relatively higher. If fatigue cracks exist in the supply pipeline, especially under high pressure, these cracks can propagate due to hydrogen embrittlement during long-term contact with hydrogen, eventually forming through cracks or through holes, leading to hydrogen leakage. Hydrogen can also leak from the seals due to issues such as aging of valve sealing rings. Joints on the supply pipeline may loosen or detach due to the impact and vibration experienced by the on-board hydrogen supply system, allowing hydrogen to leak outwards from the connection.

[0057] Secondly, based on the identified hazardous locations and risk units, a fault tree for the on-board hydrogen system leak is established:

[0058] The risks associated with on-board hydrogen systems are typically based on leaks. The top event is selected as the primary hazardous object. The fault tree, with on-board hydrogen system leaks as the top event, is mainly divided into two parts: pre-existing defects in the on-board hydrogen system and leaks occurring during use. Each part has corresponding failure events. Both natural aging of equipment and damage caused by human error and the environment are considered. Events leading to on-board hydrogen system leaks are considered from both temporal and spatial perspectives. The impact of on-board hydrogen system safety valves and human-induced protection measures on system safety is analyzed. The ease of obtaining the probability of basic events is also considered, without delving into more detailed causes of failure. The bottom events of the fault tree are determined through statistical analysis of historical leak failure data and leak accidents.

[0059] Secondly, a qualitative analysis of the fault tree for the on-board hydrogen system leakage is performed, including but not limited to minimum cut set analysis, minimum path set analysis, and structural importance analysis:

[0060] The main task of qualitative fault tree analysis is to find all minimal cut sets of the fault tree. Minimal cut sets are special sets in the cut set combination, meaning that, under the condition that no other cut sets occur, the top event only occurs if all basic events in the cut set occur simultaneously; if no basic event occurs in the cut set, the top event does not occur. Methods for finding minimal cut sets include the upward method and the downward method. The basic principle of the downward method is to start from the top event and proceed downwards, replacing the output events of each "gate" with their input events. For "OR gates," the input events are written vertically; for "AND gates," they are listed horizontally, until all "gate" events have been replaced with basic events. The resulting vertically arranged items are the cut sets of the fault tree. Further simplification and absorption using Boolean algebra rules of set operations yields all minimal cut sets. The basic principle of the upward method is to start from the bottom event and proceed upwards, using Boolean algebra rules of set operations for simplification and absorption, thus obtaining all minimal cut sets. By analyzing minimal cut sets, vulnerabilities in the security system were identified. Corresponding preventative measures were then developed to ensure the reliability of these measures while minimizing the occurrence of unexpected events, thus comprehensively controlling accidents and improving system security. The minimal path set was derived by utilizing the duality of minimal path sets and minimal cut sets. As defined, the more minimal path sets in a fault tree, the safer the system. Considering objective conditions and economic factors, the most effective and economical solution for controlling accidents was selected. Controlling fundamental events with high structural importance can also effectively reduce the probability of accidents. Qualitative analysis of the established fault tree can also be performed using commercial fault tree analysis software such as AutoFTA. This qualitative analysis reveals that, apart from uncontrollable factors such as natural disasters, strictly controlling the probability of events such as quality control system failures, untimely maintenance, substandard maintenance processes, non-compliant repairs, and safety valve failures will significantly reduce the likelihood of accidents. Events such as valve leakage have a significant impact on the occurrence of the final event. In terms of safety protection, prevention should be carried out layer by layer according to the importance coefficient of the structure. The inspection and maintenance of the vehicle hydrogen system should be strengthened. Attention should also be paid to the quality of hydrogen system components and the specifications during design and processing to prevent corrosion, aging or leakage due to its own process problems. The timeliness and correctness of the control system should also be checked. Finally, in order to prevent users from making mistakes, appropriate theoretical knowledge and operation training should be provided to users.

[0061] Finally, a quantitative analysis was conducted on the fault tree of the on-board hydrogen system leakage, including but not limited to determination of bottom event probabilities, analysis of top event probabilities, probability importance analysis, and critical importance analysis.

[0062] The probability of basic events is determined by a combination of data support and expert scoring. The quantitative analysis is based on a large amount of statistical data. Since there is currently no complete and reliable database of failure rates for each basic event in on-board hydrogen systems, the objective failure rates of some basic events are determined by sorting out data from multiple relevant failure databases, such as the Purple Book compiled by the Netherlands Organization for Applied Scientific Research (TNO), the OREDA database released by DNV in Norway, and the HyRAM+ database of Sandia National Laboratories in the United States, as well as a large number of relevant literature and data from component manufacturers.

[0063] The subjective basic event probability is obtained by using expert judgment to obtain fuzzy probabilities. Expert judgment is a common method for determining the probability of an event. It involves selecting experts from different fields to form an evaluation group, designing the evaluation content into a clearly defined evaluation opinion questionnaire, and anonymously consulting experts through written communication or other means. Experts fill out the evaluation opinion questionnaire based on their experience, provide anonymous feedback, and finally synthesize the opinions of all experts to obtain the final evaluation result. The Analytic Hierarchy Process (AHP) is used to calculate and analyze the weighting of each expert's score. A hierarchical model is established to consider factors affecting experts' estimation ability, such as education, professional title, years of service, and familiarity with hydrogen systems. At least five experts are consulted. An evaluation matrix is ​​established based on the characteristics of the participating experts, and the AHP is used to weight the experts' estimates. The following explanation uses consulting five experts as an example:

[0064] When assessing failure probabilities, experts typically don't provide precise numbers directly. Instead, they use fuzzy numerical values ​​to represent magnitude and employ defuzzification to convert these fuzzy numbers into fuzzy failure rates. In this method, experts can choose from fuzzy opinions ranging from Very Small (VS), Small (S), Relatively Small (RS), Medium (M), Relatively Large (RL), Large (L), and Very Large (VL), corresponding to membership functions f. VS (x), f S (x), f RS (x), f M (x), f RL (x), f L (x), f VL (x) are respectively:

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072] The α-cut method of fuzzy sets is used to process expert opinions. The upper and lower limits of the cut corresponding to the membership function of fuzzy language are:

[0073] Very small: VS α =[vs1,vs2], vs1=0, vs2=0.2-0.1α;

[0074] Small: S α =[s1,s2], s1=0.1α+0.1, s2=0.3-0.1α;

[0075] Smaller: RS α =[rs1,rs2], rs1=0.1α+0.2, rs2=0.5-0.1α;

[0076] Chinese: M α =[m1,m2], m1=0.1α+0.4, m2=0.6-0.1α;

[0077] Larger: RL α =[rl1,rl2], rl1=0.1α+0.5, rl2=0.8-0.1α;

[0078] Large: L α =[l1,l2], l1=0.1α+0.7, l2=0.9-0.1α;

[0079] Very large: VL α =[vl1,vl2], vl1=0.1α+0.8, vl2=1.

[0080] If the expert opinion weights calculated using the analytic hierarchy process are w1, w2, w3, w4, and w5, and their respective opinions on a basic event are very small, small, relatively small, medium, and relatively large, then the comprehensive fuzzy number can be obtained:

[0081] f W =max[w1·f VS (x)∧w2·f S (x)∧w3·f RS (x)∧w4·f M (x)∧w5·f RL (x)]

[0082] ={[w1·0+w2(0.1α+0.1)+w3(0.1α+0.2)+w4(0.1α+0.4)+w5(0.1α+0.5)],[w1 (0.2-0.1α)+w2(0.3-0.1α)+w3(0.5-0.1α)+w4(0.6-0.1α)+w5(0.8-0.1α)]}

[0083] After simplifying the above formula with constant weights, we can obtain the expression f for α. W =[m1α+n1,n2-0.1α], where m and n are constants. The fuzzy set W is W = [x1,x2] = [m1α+n1,n2-0.1α], then we have:

[0084]

[0085]

[0086] The membership function for W is:

[0087]

[0088] Based on the left-right fuzzy sorting method, the comprehensive fuzzy number is transformed into a fuzzy probability value (FPS). This method defines the maximum and minimum fuzzy sets as follows:

[0089]

[0090]

[0091] The right fuzziness probability values ​​and the left fuzziness probability values ​​of the fuzzy number are respectively:

[0092]

[0093]

[0094] The fuzzy probability value FPS is then:

[0095] FPS = [FPS] R (w)+1-FPS L (w)] / 2

[0096] Then convert the fuzzy probability value FPS into the fuzzy failure rate FFR:

[0097]

[0098]

[0099] The fuzzy failure rate (FFR) can be used as the subjective failure rate of a certain basic event. This process can be repeated to calculate the subjective failure rate for all basic events.

[0100] The subjective failure rate and objective failure rate of a basic event are initially assigned a weight of 0.5 each to determine the overall failure probability of that basic event. The weights of these subjective and objective failure rates will be determined based on the accuracy of the objective failure rate: if the objective failure rate data is predominantly from experimental data, its weight should be larger; conversely, if the objective failure rate data is predominantly from databases, its weight should be smaller. The total weight coefficient is 1. Each basic event is determined by its objective failure rate f. m Multiplied by the objective failure rate weight w m and subjective failure rate f n Multiplied by the subjective failure rate weight w n The results are summed to obtain the final failure rate f for each basic event, as shown in the following formula:

[0101] f = f m ·w m +f n ·w n

[0102] Subsequently, fault tree analysis software can be used to perform calculations such as top event probability analysis, probability importance analysis, and critical importance analysis for quantitative analysis.

[0103] The probability of the top event is calculated using the minimum cut set, as shown in the following formula:

[0104]

[0105]

[0106] In the formula: Q(T) represents the probability of the top event occurring, P(K) i Q represents the probability of a minimal cut set occurring. i =P(X) i ) represents the basic event X i The probability of occurrence.

[0107] Probability importance represents the degree to which a change in the probability of a basic event causes a change in the probability of a top event; that is, the contribution of a change in the probability of a basic event to a change in the probability of a top event. The formula for calculating probability importance is as follows:

[0108]

[0109] In the formula: I g (i) represents the probability importance of the i-th event.

[0110] Critical importance analysis represents the rate of change in the probability of the top event caused by the rate of change in the probability of the i-th basic event. It measures the importance of each basic event from both sensitivity and probability perspectives. Critical importance can reflect the influence of the probability of each event on the top event when the probabilities of the basic events are unequal. The formula for calculating critical importance is as follows:

[0111]

[0112] The final probability of a leak in the onboard hydrogen system during the top event is calculated to be: 1.11 × 10⁻⁶. -2 Although safety measures such as safety valves and timely maintenance are in place, the high probability of leakage in components like the TPRD valve, safety valves, and filters, as indicated in the data, leads to a high overall probability of leakage in the hydrogen system. High-probability events include leaks in various valves and malfunctions that directly cause leaks in components. Preventing natural damage is difficult and has a low probability; therefore, prevention of human-caused damage should be prioritized, especially protecting the hydrogen system from external impacts. A reasonable protective framework can be designed externally to the system. Special attention should be paid to checking the sealing of valves and connections, as well as the quality of components, especially the safety valves protecting against overpressure in the hydrogen system. Low-probability events are those with a low probability of failure; even if such events occur, the failure of the system's safety measures would require the failure of these events to lead to a hydrogen system failure.

[0113] This method starts with the structure of the on-board hydrogen system and analyzes in detail the failure rate of each component that may leak. Furthermore, the fault tree content can be flexibly modified according to different system structures, making it more targeted and better suited to the leakage probability analysis of a specific system. Because the on-board hydrogen system analyzed in the example has a vent needle valve with a high leakage probability, the calculated hydrogen system leakage probability is too high. The on-board hydrogen system should be improved promptly, further optimizing its structure or replacing the valve with one that has better sealing performance. Current research and analysis of hydrogen system leakage probability mainly focuses on large systems such as hydrogen refueling stations, with relatively little data analysis on on-board hydrogen systems. Due to the difficulty in obtaining leakage probabilities, research often only considers the leakage probability of specific equipment, resulting in an underestimation of the calculated hydrogen system leakage probability. Some components that may increase the leakage probability are not considered. This method's fault tree is built based on the specific on-board hydrogen system structure, providing greater completeness. While fuzzy fault tree methods, after collecting expert opinions, rely solely on mathematical formulas for calculations, resulting in fuzzy failure rates, they remain essentially semi-qualitative analysis methods. They excel at assessing the magnitude of component failure rates but cannot accurately represent the true failure rate. Furthermore, current research data on the safety of on-board hydrogen systems is limited, and leakage probability data for system-specific components is lacking. Some component leakage probabilities utilize data from other petrochemical industries, resulting in slight discrepancies. Moreover, leakage failure data requires substantial experimental and application data support; however, as hydrogen energy technology advances, the database will become increasingly comprehensive. This method, in determining the probability of each basic event, combines subjective and objective data. Subjective data supplements the currently lacking objective data, while objective data is further refined using real-world experimental and application data to correct subjective data, bringing the final result closer to the true value. This approach reduces the significant uncertainty inherent in purely subjective fuzzy fault tree methods and addresses shortcomings such as untimely database updates, incomplete data, and poor correlation. When combined with fault tree analysis software, it can not only perform rapid qualitative and quantitative analysis of on-board hydrogen systems and analyze the weaknesses of hydrogen systems, but also quickly calculate the probability of hydrogen system leakage, providing a data basis for frequency analysis for on-board hydrogen system safety risk assessment, and making the on-board hydrogen system safety risk assessment faster.

[0114] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. A method for qualitative and quantitative assessment of leakage frequency in an on-board hydrogen supply system, characterized in that, Includes the following steps: Step 1: Perform structural analysis on the on-board hydrogen system to identify components that may leak, and determine hazardous areas and risk units; Step 2: Based on the hazardous locations and risk units identified in Step 1, establish a fault tree for the on-board hydrogen system leak. Step 3: Perform a qualitative analysis on the vehicle hydrogen system leakage fault tree established in Step 2, and find all the minimum cut sets or minimum path sets of the above fault tree. Step 4: Quantitatively analyze the on-board hydrogen system leakage fault tree established in Step 2, and use expert judgment to determine the subjective failure rate f of all basic events. n This refers to the fuzzy failure rate; the objective failure rate f of all basic events is obtained by querying the database and acquiring failure experiment data. m The databases mentioned above include the OREDA database or the HyRAM+ database, and the failure test data mentioned above are from the manufacturers of on-board hydrogen supply system components. Step 5, calculate the subjective failure rate f for each basic event obtained in Step 4. n and objective failure rate f m Each is assigned a weight to ensure that the sum of the subjective failure rate weight and the objective failure rate weight for each basic event is 1. Then, the failure rate f for each basic event is determined according to the following formula; finally, the failure rates of all basic events are obtained. f=f m ·w m +f n ·w n ; In the above formula, W n and W m These are the subjective failure rate weights and the objective failure rate weights, respectively. Step 6: Using the fault tree analysis software AutoFTA, perform top event probability analysis, probability importance analysis, and critical importance analysis on the failure rates of all basic events obtained in Step 5 to obtain the probability of the top event, namely, leakage of the on-board hydrogen system.

2. The qualitative and quantitative assessment method for leakage frequency of an on-board hydrogen supply system as described in claim 1, characterized in that, The hazardous parts and risk units mentioned in step 1 include: hydrogen storage cylinders, hydrogen supply pipelines, one-way valves, solenoid valves, overflow valves, pressure reducing valves, sensors, joint connections, and accidental collisions.

3. The qualitative and quantitative assessment method for leakage frequency of an on-board hydrogen supply system as described in claim 1, characterized in that, Calculate the subjective failure rate f of all the basic events described in step 4. n The specific steps are as follows: Step 4-1: Consult n experts to obtain fuzzy linguistic values ​​for the probability of a certain basic event occurring from each expert. The fuzzy linguistic values ​​include "very small", "small", "relatively small", "medium", "relatively large", "large", and "very large". Step 4-2: Based on the fuzzy language values ​​obtained in Step 3-1, establish the corresponding membership function: f VS (x), f S (x), f RS (x), f M (x), f RL (x), f L (x), f VL (x); The upper and lower limits of the α-cut corresponding to each of the above membership functions are: Very small: VS α =[vs1,vs2], vs1=0, vs2=0.2-0.1α; Small: S α = [s1, s2], s1 = 0.1α + 0.1, s2 = 0.3 - 0.1α; Smaller: RS α =[rs1,rs2], rs1=0.1α+0.2, rs2=0.5-0.1α; Chinese: M α =[m1,m2], m1=0.1α+0.4, m2=0.6-0.1α; Larger: RL α =[rl1,rl2], rl1=0.1α+0.5, rl2=0.8-0.1α; Large: L α = [l1, l2], l1 = 0.1α + 0.7, l2 = 0.9 - 0.1α; Very large: VL α =[vl1,vl2], vl1=0.1α+0.8, vl2=1; Step 4-3: Use the analytic hierarchy process (AHP) to assign weights w to the fuzzy linguistic values ​​of each expert regarding the probability of a certain basic event. i i takes values ​​from 1 to n; the membership function f corresponds to each expert opinion. i (x), where i takes values ​​from 1 to n; the comprehensive fuzzy number of a certain basic event is obtained according to the following formula: f W =max[w1·f1(x)∧w2·f2(x)∧w3·f3(x)∧w4·f4(x)∧…∧w n ·f n (x)]; Step 4-4: Convert the comprehensive fuzzy number obtained in step 4-3 into a fuzzy probability value (FPS) based on the left and right fuzzy sorting method. FPS=[FPS R (w)+1-FPS L (w)] / 2; In the above formula, FPS L (w) represents the fuzzy number, the left fuzzy probability value, and FPS. R (w) represents the right fuzzy probability value of the fuzzy number; Step 4-5: Transform the fuzzy probability value obtained in step 4-4 into the fuzzy failure rate FFR, which is the subjective failure rate f of a certain basic event. n ; Steps 4-6: Repeat steps 4-1 to 4-5 to obtain the subjective failure rate of all basic events.

Images