Fire risk analysis method and system for ship carrying new energy automobile

Abstract

The invention discloses a fire risk analysis method and system for a ship carrying new energy automobile. The method comprises the steps of assuming a new energy automobile loading condition and a fire scene, determining a life and property loss assumption, establishing a risk matrix, and proposing a fault failure mode hazard source identification guide word. Dividing loading, transportation and unloading nodes and operation behaviors according to a ship transportation vehicle flow, and constructing an event tree-fault tree model; the event tree starts from the fire of the new energy automobile at the roll-on-roll-off place, and layer-by-layer analysis is performed according to the type of the place, early detection and the like; the fault tree and the event tree are subjected to corresponding hierarchical association and subdivision; the accident probability is determined through an initial event probability calculation module and a hierarchical probability quantification system, the result is quantified through property loss grading, and a risk value is calculated; setting risk balance, determining risk control measures, and recommending an optimal risk control scheme for the customer through cost-benefit analysis. The problem that risk control input and output are difficult to balance for a long time in the industry is solved.

Description

Technical Field

[0001] This invention belongs to the field of ship transportation safety technology, and more specifically, relates to a method and system for analyzing fire risks of new energy vehicles transported by ships. Background Technology

[0002] With the rapid development of the global new energy vehicle industry, the demand for roll-on / roll-off (Ro-Ro) shipping of new energy vehicles continues to rise. The power batteries used in these vehicles face numerous challenges in the marine transportation environment, including fire safety hazards. Furthermore, the limitations of existing technologies in fire risk assessment create a serious contradiction with actual needs.

[0003] At the level of fire risk assessment, the industry currently largely uses the assessment model of traditional gasoline-powered vehicles, neglecting the battery characteristics of new energy vehicles and the special scenarios of ship transportation. This contradiction directly leads to the inability of traditional methods to accurately identify key factors of fire risk in new energy vehicles, and to effectively predict battery thermal runaway and analyze fire spread paths. For example, traditional methods lack quantitative assessment tools for the actual fire-fighting effects of different fire-fighting systems in different compartments of a ship, making it difficult for shipping companies to predict in advance the possibility of fire breaking through fire compartments and spreading to other compartments. Once a fire occurs, due to the lack of accurate assessment as support, emergency response and handling are often delayed and inefficient, missing the best fire-fighting opportunity, exacerbating the losses caused by the fire, and posing a serious threat to the lives of people on board and the safety of ship navigation.

[0004] Regarding the formulation of prevention and control measures, the lack of a scientific and comprehensive risk assessment system has put shipping companies in a dilemma when developing risk control measures. Some companies, in an effort to mitigate fire risks, adopt excessive protective measures, such as adding large amounts of expensive fire-fighting equipment and conducting frequent, unnecessary inspections and maintenance. This leads to a significant increase in operating costs, weakens the company's competitiveness in the market, compresses profit margins, and may even force them to reduce their business scale due to excessive costs. On the other hand, other companies, due to inadequate prevention and control measures, face huge economic losses in the event of a fire, including cargo damage, ship repairs, and compensation for personal injury or death. Furthermore, the social impact of the accident may damage the company's reputation, lead to customer loss, hinder subsequent business expansion, and seriously affect the company's sustainable development.

[0005] In terms of cost-benefit balance, existing risk assessment methods cannot comprehensively weigh the input costs and long-term benefits of risk control measures from the perspective of the entire ship lifecycle. Enterprises either blindly invest in risk prevention and control, resulting in wasted resources, or suffer high costs from fire accidents due to insufficient investment. This imbalance not only restricts the economic benefits of individual enterprises but also negatively impacts the healthy development of the entire shipping industry. Against the backdrop of continuously growing global demand for new energy vehicle transportation, if this contradiction cannot be effectively resolved, the shipping industry will struggle to meet the safe and efficient transportation requirements of the new energy vehicle industry, hindering the globalization of new energy vehicles and consequently affecting the development process of the global new energy industry.

[0006] Therefore, developing a fire risk analysis method applicable to ships carrying new energy vehicles, accurately identifying potential risks, effectively controlling disasters, and optimizing the cost and benefits of risk control has become an important issue that urgently needs to be addressed in the shipping industry. Summary of the Invention

[0007] This invention aims to change this situation by focusing on classifying and assessing the fire risk levels of ships carrying new energy vehicles and calculating the consequences of shipping under different risk levels. Through scientific and systematic classification and assessment, it clarifies the specific losses corresponding to different risk levels, such as the scale of casualties, the extent of property damage, and the condition of the ship, providing shipping companies with a clear understanding of the risk consequences.

[0008] In view of the above-mentioned defects or improvement needs of the existing technology, as a first aspect of the present invention, the present invention provides a method for fire risk analysis of new energy vehicles transported by ships, including:

[0009] S1. Assumptions are made regarding the loading conditions and the fire scenario;

[0010] S2. Determine the assumptions regarding loss of life and property damage;

[0011] S3. Propose guiding words for hazard identification and establish a risk matrix for the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, divide the nodes including loading, transportation and unloading and clarify the specific operation behavior of each node;

[0012] S4. Construct a model consisting of an event tree and a fault tree; the event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event, and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions; the fault tree is associated with the corresponding layer of the event tree, and is subdivided into multiple subtrees according to different scenarios and conditions, and the fire risk development process is systematically analyzed.

[0013] S5. Determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system; quantify the consequences of an accident by using a property loss classification method and calculate the risk value using potential property loss; set risk criteria, determine risk control measures, screen and combine risk control options, and conduct benefit and cost analysis on each risk control option; finally, evaluate the cost-effectiveness based on the risk criteria and recommend the most cost-effective risk control measures to the client.

[0014] Furthermore, the assumption method for making assumptions about the loading conditions and fire scenario in S1 includes:

[0015] Loading assumptions:

[0016] Assuming the ship meets current regulations and standards and is not carrying dangerous goods;

[0017] Assume two loading scenarios: the roll-on / roll-off space is entirely loaded with new energy vehicles, and the roll-on / roll-off space is mixed with new energy vehicles and fuel vehicles.

[0018] Fire scenario assumptions:

[0019] Only the fires of new energy vehicles, the mutual influence between fires of new energy vehicles and fires of fuel vehicles, and the impact of fires of vehicles and cargo on new energy vehicles are considered. The conventional fires of fuel vehicles and cargo themselves and their impact are not assessed.

[0020] Based on the development of fire hazards in new energy vehicles and the effectiveness of active and passive fire suppression systems on ships, a tiered fire scenario is constructed, including:

[0021] Fire Scenario A: Assume that a single vehicle fire is successfully extinguished, that is, the thermal runaway or fire of a single new energy vehicle battery is detected and the correct decision is made in time, and the crew successfully suppresses or extinguishes the fire within the vehicle's range.

[0022] Fire Scenario B: Assume that no fewer than 3 vehicles are successfully extinguished, that is, if a single new energy vehicle fire is not successfully suppressed or extinguished, it ignites no fewer than 2 adjacent vehicles, which are then successfully extinguished or suppressed by the fire extinguishing system within the range of no fewer than 3 vehicles.

[0023] Fire Scenario C: Assume that the fire was not extinguished but the fire was contained within the location of the fire. That is, a single new energy vehicle fire was not successfully extinguished or suppressed, igniting surrounding vehicles, but the fire extinguishing system failed to extinguish or suppress the fire. The entire floor of vehicles was burned or the battery experienced thermal runaway, but the fire was contained within the location of the fire.

[0024] Fire Scenario D: Assume that firefighting has failed and the fire is not contained within the location of origin;

[0025] Fire Scenario E: This scenario specifically refers to the assumption that manual firefighting on the open deck has failed.

[0026] Furthermore, the assumptions made in S2 regarding the determination of loss of life and property loss are as follows:

[0027] Assuming that the loss of life caused by a fire in a ship carrying a new energy vehicle is the same as that caused by a fire in a fuel vehicle, that is, the ship carrying a new energy vehicle does not pose an additional risk of loss to people.

[0028] Assuming that the consequences of a fire involving a ship carrying new energy vehicles only consider property damage, the extent of which depends on the fire's progression within the premises and the effectiveness of the ship's active and passive fire suppression systems, as detailed below:

[0029] Let S be the set of fire scenarios, S = {S1, S2} s ,…,S n}, where {A,B,C,D,E}∈S; for each scenario s∈S, there is a corresponding set of vehicles V. s The vehicle number is i, and the value of each vehicle is v. i (The value of new energy vehicles is V) ev The value of a gasoline-powered car is V f The loss coefficient is δ. i (0<δ i <1), the hull structure maintenance cost corresponding to scenario s is C. s The indirect losses caused by the impact on the lease term are C. t (For scenario A, Ct = 0; for scenarios D and E, the additional calculation of this indirect loss is not considered for now), the total loss value of the ship is V. ship Therefore, the method for calculating property loss is as follows:

[0030]

[0031] Where V represents the specific loss assuming a fire occurs.

[0032] Furthermore, the specific method for proposing hazard identification keywords and establishing a risk matrix for new energy vehicle failure modes in S3 is as follows:

[0033] Let the set of hazard sources be: Ω={ω1,ω2,…,ω n}, where ω i Indicates the i-th hazard source;

[0034] The set of operation nodes is: Π={π1,π2,…,π m}, where π j This represents the j-th operation node;

[0035] The risk matrix R is a two-dimensional matrix. ij R represents the risk level of the i-th hazard source at the j-th operation node. ij∈{1,2,…,k}, where k is the number of risk levels;

[0036] P ij S represents the probability that the i-th hazard will occur at the j-th operation node. ij The severity of its consequences;

[0037] For the guiding words in hazard identification, let the set of guiding words be G = {g1, g2, ..., g...} l}, the introductory word g s The weight is ω s ,and

[0038] r is This indicates that the i-th hazard source is indicated by the introductory word g. s The score

[0039] The comprehensive score for hazard sources is calculated as follows:

[0040]

[0041] In the formula, T i This represents the overall score of the i-th hazard source, reflecting the potential hazard level of the i-th hazard source;

[0042] The probability P of the i-th hazard source at the j-th operation node ij and severity of consequences S ij Combined score T of hazard sources i Related, while also considering the influence factor α of the characteristics of the operating nodes. ij and β ij Then we have:

[0043] P ij =α ij T i

[0044] S ij =β ij T i

[0045] Based on probability P ij and severity of consequences S ij Determine the risk level R in the risk matrix ij This transformation can be achieved using a mapping function f:

[0046] R ij =f(P ij ,S ij )

[0047] The likelihood and severity of consequences are divided into different ranges, and the risk level is determined based on the combination of these ranges:

[0048]

[0049] The overall risk of the entire transportation process can be obtained by weighted summation of the elements in the risk matrix, where the weight of each operation node is q. j ,and but:

[0050]

[0051] In the formula, R total This indicates the overall risk of the entire transportation process.

[0052] Furthermore, the specific method for constructing the model consisting of an event tree and a fault tree in S4 is as follows:

[0053] The top event in the fault tree that is related to the initial event of the event tree is defined as T, and its occurrence probability is P(T), which is calculated by combining the basic events in the fault tree through logic gates.

[0054] Define the parameters related to the event tree: Let the number of levels in the event tree be L, and the event set of the l-th level be... Where n l It is the number of events in the l-th layer;

[0055] Event e ij The probability of occurrence is denoted as P(e ij );

[0056] The transition probability from the 1st level to the (1+1)th level is denoted as P(e (l+1)k |e lj ), indicating that the event e is at level l. lj Under the condition that it occurs, the event e at level l+1 (l+1)k The probability of occurrence;

[0057] The method for calculating the probability of events in an event tree is as follows:

[0058] The first level event probability P(e) of the event tree 1j The probability of events in subsequent layers is given directly based on experience or statistical data; the probability of events in subsequent layers is calculated using the conditional probability formula.

[0059]

[0060] This indicates that event e at level l+1 (l+1)k The probability of occurrence is equal to the probability of all possible events e at level l. lj Under what conditions does e occur (l+1)k The probability of occurrence is weighted sum, with weights equal to the probability of the l-th level event e. lj The probability of occurrence;

[0061] Constructing an overall risk model:

[0062] Each terminal event e in the event tree lk For a specific accident scenario, the probability of its occurrence is:

[0063]

[0064] Where k1,k2,…,k L It is the event index path from the first level to the Lth level;

[0065] Let each terminal event e LK The corresponding severity of the consequences is C(e) LK If the system risk R represented by the entire model is:

[0066]

[0067] In other words, system risk equals the sum of the products of the probability of occurrence of all end events and the severity of their corresponding consequences. The probability of the impact of basic events on intermediate and top events is calculated by fault tree, and the probability of occurrence of different accident scenarios is calculated by event tree based on the transition probability of events at each level. Finally, the overall risk of the system is obtained by combining the results.

[0068] Furthermore, the specific calculation method for the top event T related to the initial event of the event tree in the fault tree, and its occurrence probability P(T), is as follows:

[0069] Define the parameters related to the fault tree: The basic event set of the fault tree is B = {b1, b2, ..., b}. m}, basic event b i The probability of occurrence is denoted as P(b) i );

[0070] The set of intermediate events is M = {m1, m2, ..., m} s The top event is T;

[0071] Logic gates are represented by ⊙; for intermediate events m i It is composed of basic events or other intermediate events combined through logic gates;

[0072] The method for calculating the event probability in a fault tree is as follows:

[0073] In the AND gate case: if intermediate event m i From basic event b i1 ,b i2 ,…,b it By connecting to the door, then:

[0074]

[0075] This is because the AND gate requires that the output event occurs only when all input events occur, and this formula is derived from the multiplication principle of the probabilities of independent events;

[0076] In the OR gate case: if the intermediate event mi is connected by the basic events bi1, bi2, ..., bit through an OR gate, then:

[0077]

[0078] The principle is to first calculate the probability that none of the input events will occur, and then subtract this probability from 1 to get the probability that at least one input event will occur.

[0079] For the top event T, by recursively applying the above rules, the probabilities of intermediate events are calculated step by step starting from the basic events, and finally the probability P(T) of the top event is obtained.

[0080] Furthermore, in S4, the consequences of the accident are quantified using a property damage classification method, and the risk value is calculated using potential property damage. The specific method is as follows:

[0081] Let the set of accident scenarios be A = {A1, A2, ..., A...} n}, where A i Let P(A) represent the i-th accident scenario, and let P(A) represent the probability of each accident scenario occurring. i This probability is calculated using a model that combines event trees and fault trees;

[0082] V(A i Let be the property loss value under the i-th accident scenario, determined according to the property loss classification method; let PLP represent the potential property loss, i.e., the expected value of property loss per year.

[0083] Potential property damage PLP is the sum of the products of the probability of occurrence of each accident scenario and the corresponding property damage value under all accident scenarios. The calculation formula is as follows:

[0084]

[0085] By using a weighted summation method, the expected value of annual property loss for ships carrying new energy vehicles is obtained, thereby quantifying the overall property loss risk value.

[0086] Furthermore, in step S4, risk criteria are set, risk control measures are determined, and risk control options are selected and combined. Benefit and cost analyses are then performed on each risk control option. The specific method is as follows:

[0087] Risk criteria are defined as follows: This assessment considers only property damage and adopts the following principles, with the lowest cost / benefit ratio taking precedence:

[0088] ΔC<ΔB

[0089] In the formula, ΔC is the input cost of RCO; ΔB is the relevant benefit obtained after reducing risk;

[0090] The costs ΔC and ΔB are allocated over the entire lifespan of the vessel, calculated using the following formula:

[0091]

[0092] In the formula, A is the initial cost of implementing RCO, T is the ship's life cycle, r is the depreciation rate, and X is the cost or benefit of RCO in a specific year.

[0093] As a second aspect of the present invention, a fire risk analysis system for ship-borne new energy vehicles is provided, comprising:

[0094] The loading condition and fire scenario assumption unit is used to make assumptions about the loading condition and fire scenario;

[0095] The assumption unit for loss of life and property loss is used to determine the assumptions for loss of life and property loss.

[0096] The risk identification preparation unit is used to establish a risk matrix and propose hazard identification keywords for the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, it divides the process into nodes including loading, transportation and unloading and clarifies the specific operation behavior of each node.

[0097] The risk analysis model building unit is used to construct a model consisting of an event tree and a fault tree. The event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions. The fault tree is associated with the corresponding layer of the event tree and is subdivided into multiple subtrees according to different scenarios and conditions to systematically analyze the development process of fire risk.

[0098] The risk assessment and control decision-making unit is used to determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system; quantify the consequences of an accident using a property loss classification method; calculate the risk value using potential property loss; set risk criteria; determine risk control measures; screen and combine risk control options; conduct benefit and cost analysis on each risk control option; evaluate cost-effectiveness based on risk criteria; and finally recommend the most cost-effective risk control measures to the client.

[0099] As a third aspect of the present invention, a computer-readable storage medium is also provided, on which a computer program is stored, wherein the computer program is executed by a processor in any of the steps of the above-described method for fire risk analysis of new energy vehicles transported by ships.

[0100] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0101] 1. This invention provides a method for analyzing the fire risk of new energy vehicles transported by ship. By scientifically setting analytical premises, it lays the foundation for accurate assessment. Specifically, it first assumes that the ship complies with regulations and does not carry dangerous goods. Two loading scenarios are established: one entirely loaded with new energy vehicles, and the other a mix of new energy vehicles and gasoline-powered vehicles. This constructs graded fire scenarios covering different severity levels, such as successful extinguishing of a single vehicle fire and multiple vehicle fires. Simultaneously, it assumes that personal loss will not increase due to fires involving new energy vehicles, and sets corresponding property damage levels based on the fire scenarios. These assumptions clarify the analytical boundaries and assessment standards, making subsequent risk analysis more targeted, avoiding interference with the accuracy of the assessment due to complex actual conditions, and ensuring that the analysis results accurately reflect the fire risk status of new energy vehicles during ship transportation.

[0102] 2. This invention provides a method for analyzing the fire risk of new energy vehicles transported by ships. Through innovative risk analysis models and quantitative methods, it achieves in-depth analysis of fire risk. An initial fire probability model based on the characteristics of new energy vehicles is established, and a fire probability ratio coefficient is introduced to calculate the specific fire probability. A hierarchical probability quantification system is constructed, including multiple modules such as location type classification and early detection failure analysis, refining the risk probability calculation from different dimensions. Using a model composed of event trees and fault trees, with a new energy vehicle fire as the initial event, different situations are analyzed layer by layer, structuring and visualizing the complex fire risk development process. Based on this, the consequences of accidents are quantified through property damage classification and potential property damage calculation, thereby comprehensively and accurately assessing fire risk. Compared with traditional assessment methods, this method can capture more potential risk factors and improve the accuracy and reliability of the assessment.

[0103] 3. This invention provides a method for analyzing the fire risk of new energy vehicles transported by ships. Through a systematic cost-benefit analysis process, it achieves a balance between risk control and economic benefits. After setting risk criteria, multiple risk control measures are identified and combined to form options. Each option undergoes benefit and cost analysis, evaluating benefits from aspects such as reducing accident probability and minimizing property damage, while comprehensively considering equipment investment, labor costs, and other factors to calculate costs. Based on the risk criteria, the cost-benefit of each option is compared, and the optimal risk control measures are recommended to the client. This avoids increased costs due to excessive prevention and control, while also preventing significant losses due to insufficient prevention and control. It helps shipping companies to rationally plan resources and improve operational efficiency while ensuring transportation safety, thus solving the long-standing industry problem of balancing risk control input and output. Attached Figure Description

[0104] Figure 1This is a flowchart of a fire risk analysis method for transporting new energy vehicles by ship, according to an embodiment of the present invention.

[0105] Figure 2 This is a sample metadata for calculating risk quantification in an embodiment of the present invention;

[0106] Figure 3 This is a schematic diagram of a system unit in an embodiment of the present invention. Detailed Implementation

[0107] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0108] Example 1

[0109] Please refer to Figure 1 This embodiment 1 provides a method for analyzing the fire risk of new energy vehicles transported by ships, including:

[0110] S1. Assumptions are made regarding the loading conditions and the fire scenario;

[0111] S2. Determine the assumptions regarding loss of life and property damage;

[0112] S3. Establish a risk matrix and propose guiding words for hazard identification based on the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, divide the process into nodes including loading, transportation and unloading, and clarify the specific operation behavior of each node;

[0113] S4. Construct a model consisting of an event tree and a fault tree; the event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event, and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions; the fault tree is associated with the corresponding layer of the event tree, and is subdivided into multiple subtrees according to different scenarios and conditions, and the fire risk development process is systematically analyzed.

[0114] S5. Determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system; quantify the consequences of an accident by using a property loss classification method and calculate the risk value using potential property loss; set risk criteria, determine risk control measures, screen and combine risk control options, and conduct benefit and cost analysis on each risk control option; finally, evaluate the cost-effectiveness based on the risk criteria and recommend the most cost-effective risk control measures to the client.

[0115] Next, this embodiment 1 will provide a detailed explanation of each step.

[0116] (1) Assumptions are made regarding the loading conditions and the fire scenario.

[0117] The loading conditions are assumed to be as follows:

[0118] It is assumed that the ship to be analyzed meets the current relevant laws and regulations on shipping and does not carry dangerous goods.

[0119] Two typical loading scenarios are set: first, all vehicles in the ship's roll-on / roll-off (Ro-Ro) space are new energy vehicles; second, a mix of new energy vehicles and gasoline vehicles are loaded in the Ro-Ro space.

[0120] By setting these two loading scenarios, we can cover common new energy vehicle loading modes on ships, providing a basis for subsequent analysis.

[0121] Assumptions are made regarding the fire scenario, specifically as follows:

[0122] Regarding fire scenarios, the assessment focuses only on fires involving new energy vehicles, the mutual influence between fires involving new energy vehicles and fires involving fuel vehicles, and the impact of fires involving vehicles and cargo on new energy vehicles, excluding assessments of conventional fires involving fuel vehicles and cargo themselves and their impact.

[0123] Based on the development of fire hazards in new energy vehicles and the effectiveness of active and passive fire suppression systems on ships, a graded fire scenario is constructed:

[0124] Fire Scenario A: Assuming a single vehicle fire is successfully extinguished, that is, when a single new energy vehicle experiences battery thermal runaway or catches fire, it can be detected in time by a fixed automatic fire detection and fire alarm system, or discovered by personnel, and the crew can quickly make the correct decision and successfully suppress or extinguish the fire within the range of the new energy vehicle.

[0125] Fire Scenario B: Assume that at least 3 vehicles are successfully extinguished. That is, after a single new energy vehicle catches fire, it fails to be successfully contained or extinguished, and then ignites at least 2 adjacent vehicles, but is eventually successfully extinguished or contained within the range of at least 3 vehicles by the fire extinguishing system.

[0126] Fire Scenario C: Assume that the fire was contained within the place of origin, but the fire was not extinguished. That is, after a single new energy vehicle caught fire, the fire extinguishing system failed to extinguish or suppress the fire, resulting in the fire affecting the entire floor of vehicles or thermal runaway of the battery. However, the fire was contained within the place of origin and did not spread to other areas.

[0127] Fire Scenario D: Assume that firefighting has failed and the fire is not contained within the location of origin. At this point, the fire has spread to other decks, exceeding the ship's own firefighting capabilities.

[0128] Fire Scenario E: This scenario specifically refers to the failure of manual firefighting on an open deck. Considering that the open vehicle deck is usually located on the top deck of a ship, the impact on the ship's damage is relatively small.

[0129] (2) Determine the assumptions of loss of life and loss of property

[0130] Assumption of loss of life: It is assumed that the loss of life caused by a fire in a ship carrying a new energy vehicle is the same as that caused by a fire in a fuel vehicle, that is, the ship carrying a new energy vehicle will not bring additional risk of loss of life to people.

[0131] Property Loss Assumptions: It is assumed that the consequences of a fire involving a ship carrying new energy vehicles will only consider property loss. The extent of property loss will be determined by the development of the fire within the premises and whether the ship's active and passive fire suppression systems function, including the following assumptions:

[0132] Fire Scenario A: At least one new energy vehicle is completely destroyed, and at least four surrounding new energy vehicles are damaged but do not cause battery thermal runaway. The ship's hull structure above the vehicles is slightly damaged and can be restored with simple repairs, without affecting the ship's charter period.

[0133] Fire Scenario B: No fewer than 3 new energy vehicles are completely destroyed, and no fewer than 8 surrounding new energy vehicles are damaged but have not caused battery thermal runaway, or no fewer than 4 fuel vehicles are damaged. The ship's hull structure above the vehicles is damaged, including hydraulic pipelines and electrical wiring, which will affect the ship's charter period.

[0134] Fire Scenario C: Due to the failure to effectively control the fire, it spread throughout the entire deck, damaging all vehicles and triggering battery thermal runaway. The vessel requires repairs, incurring significant costs and severely impacting the charter period. Considering the limited heat resistance of the premises' fireproof structure, and the possibility that the fire could spread to other vehicle decks without external assistance, a comprehensive assessment deems no less than 50% of the total loss of the vessel as the loss.

[0135] Fire Scenario D: Since the fire has spread to other decks and exceeds the ship's own firefighting capabilities, it is assumed that the ship is a total loss.

[0136] Fire Scenario E: The open-air vehicle deck is usually located on the top deck of the ship and has a relatively small impact on the ship's losses. It is assumed that the loss is no less than 10% of the total loss of the ship.

[0137] Furthermore, let S be the set of fire scenarios, S = {S1, S2} s ,…,S n}, where {A,B,C,D,E}∈S; for each scenario s∈S, there is a corresponding set of vehicles V. s The vehicle number is i, and the value of each vehicle is v. i (The value of new energy vehicles is V) ev The value of a gasoline-powered car is V fThe loss coefficient is δ. i ((0<δ i <1), the hull structure maintenance cost corresponding to scenario s is C. s The indirect losses caused by the impact on the lease term are C. t (For scenario A, Ct = 0; for scenarios D and E, the additional calculation of this indirect loss is not considered for now), the total loss value of the ship is V. ship Therefore, the method for calculating property loss is as follows:

[0138]

[0139] Where V represents the specific loss assuming a fire occurs.

[0140] (3) Establish a risk matrix and identify hazards and divide operational nodes.

[0141] Since the transportation of new energy vehicles by ships is still in its early stages, there is currently no detailed statistical data on dangerous accidents. Therefore, this embodiment is supported by statistical data from domestic and foreign reports on fire accidents involving passenger roll-on / roll-off ships and new energy vehicles.

[0142] A risk matrix was established by referencing the International Maritime Organization (IMO) circular "Revised Guidelines on the Use of Integrated Safety Assessment (FSA) in the Development of IMO Codes" (MSC-MEPC.2 / Circ.12). Furthermore, a series of hazard identification keywords were proposed for failure modes of new energy vehicles, including thermal runaway, explosion, fire, fire spread, fixed failure, and vehicle collision.

[0143] Based on the operational procedures for ship transport vehicles, the process is divided into three stages: loading, transportation, and unloading, encompassing the entire process from vehicle loading onto the ship to vehicle unloading. Each stage includes the following actions:

[0144] Loading: The process of loading cargo onto a ship is the same as loading vehicles onto the ship: the vehicle arrives at the dock; the driver starts the vehicle and drives into the vehicle compartment; the vehicle goes up the ramp to each deck and stops in the appropriate position; the vehicle is turned off, the driver gets out of the vehicle, and a designated person secures and ties the vehicle.

[0145] Transportation: Vehicle transportation refers to the process of transporting vehicles while the ship is sailing. During the voyage, compared to the loading and unloading process, vehicles are parked for an extended period of time. The parking time varies depending on the ship's route.

[0146] Unloading: Unloading a ship is the process of vehicles disembarking: the driver gets on the vehicle and is untied; the vehicle is started and driven; the vehicle goes up the ramp to the freeboard deck or upper deck; the vehicle leaves the ship and arrives at the dock.

[0147] In a preferred embodiment, the specific method for proposing hazard identification keywords and establishing a risk matrix for new energy vehicle failure modes is as follows:

[0148] Let the set of hazard sources be: Ω={ω1,ω2,…,ω n}, where ω i Indicates the i-th hazard source;

[0149] The set of operation nodes is: Π={π1,π2,…,π m}, where π j This represents the j-th operation node;

[0150] The risk matrix R is a two-dimensional matrix. ij R represents the risk level of the i-th hazard source at the j-th operation node. ij ∈{1,2,…,k}, where k is the number of risk levels;

[0151] P ij S represents the probability that the i-th hazard will occur at the j-th operation node. ij The severity of its consequences;

[0152] For the guiding words in hazard identification, let the set of guiding words be G = {g1, g2, ..., g...} l}, the introductory word g s The weight is ω s ,and

[0153] r is This indicates that the i-th hazard source is indicated by the introductory word g. s The score

[0154] The comprehensive score for hazard sources is calculated as follows:

[0155]

[0156] In the formula, T i This represents the overall score of the i-th hazard source, reflecting the potential hazard level of the i-th hazard source;

[0157] The probability P of the i-th hazard source at the j-th operation node ij and severity of consequences S ij Combined score T of hazard sources i Related, while also considering the influence factor α of the characteristics of the operating nodes. ij and β ij Then we have:

[0158] P ij =α ij T i

[0159] S ij =β ij T i

[0160] Based on probability P ij and severity of consequences S ij Determine the risk level R in the risk matrix ij This transformation can be achieved using a mapping function f:

[0161] R ij =f(P ij ,S ij )

[0162] The likelihood and severity of consequences are divided into different ranges, and the risk level is determined based on the combination of these ranges:

[0163]

[0164] The overall risk of the entire transportation process can be obtained by weighted summation of the elements in the risk matrix, where the weight of each operation node is q. j ,and but:

[0165]

[0166] In the formula, R total This indicates the overall risk of the entire transportation process.

[0167] (4) Construct a model consisting of an event tree and a fault tree.

[0168] In a preferred embodiment, system risk is quantified using a model composed of event trees and fault trees, as detailed below:

[0169] The top event in the fault tree that is related to the initial event of the event tree is defined as T, and its occurrence probability is P(T), which is calculated by combining the basic events in the fault tree through logic gates.

[0170] Define the parameters related to the fault tree: The basic event set of the fault tree is B = {b1, b2, ..., b}. m}, basic event b i The probability of occurrence is denoted as P(b) i );

[0171] The set of intermediate events is M = {m1, m2, ..., m} s The top event is T;

[0172] Logic gates are represented by ⊙; for intermediate events m i It is composed of basic events or other intermediate events combined through logic gates;

[0173] The method for calculating the event probability in a fault tree is as follows:

[0174] In the AND gate case: if intermediate event m i From basic event b i1 ,b i2 ,…,b it By connecting to the door, then:

[0175]

[0176] This is because the AND gate requires that the output event occurs only when all input events occur, and this formula is derived from the multiplication principle of the probabilities of independent events;

[0177] In the OR gate case: if the intermediate event mi is connected by the basic events bi1, bi2, ..., bit through an OR gate, then:

[0178]

[0179] The principle is to first calculate the probability that none of the input events will occur, and then subtract this probability from 1 to get the probability that at least one input event will occur.

[0180] For the top event T, by recursively applying the above rules, the probabilities of intermediate events are calculated step by step starting from the basic events, and finally the probability P(T) of the top event is obtained.

[0181] Define the parameters related to the event tree: Let the number of levels in the event tree be L, and the event set of the l-th level be... Where n l It is the number of events in the l-th layer;

[0182] Event e ij The probability of occurrence is denoted as P(e ij );

[0183] The transition probability from the 1st level to the (1+1)th level is denoted as P(e (l+1)k |e lj ), indicating that the event e is at level l. lj Under the condition that it occurs, the event e at level l+1 (l+1)k The probability of occurrence;

[0184] The method for calculating the probability of events in an event tree is as follows:

[0185] The first level event probability P(e) of the event tree 1j The probability of events in subsequent layers is given directly based on experience or statistical data; the probability of events in subsequent layers is calculated using the conditional probability formula.

[0186]

[0187] This indicates that event e at level l+1 (l+1)k The probability of occurrence is equal to the probability of all possible events e at level l. lj Under what conditions does e occur (l+1)k The probability of occurrence is weighted sum, with weights equal to the probability of the l-th level event e. lj The probability of occurrence;

[0188] Constructing an overall risk model:

[0189] Each terminal event e in the event tree lk For a specific accident scenario, the probability of its occurrence is:

[0190]

[0191] Where k1,k2,…,k L It is the event index path from the first level to the Lth level;

[0192] Let each terminal event e LK The corresponding severity of the consequences is C(e) LK If the system risk R represented by the entire model is:

[0193]

[0194] In other words, system risk equals the sum of the products of the probability of occurrence of all end events and the severity of their corresponding consequences. The probability of the impact of basic events on intermediate and top events is calculated by fault tree, and the probability of occurrence of different accident scenarios is calculated by event tree based on the transition probability of events at each level. Finally, the overall risk of the system is obtained by combining the results.

[0195] In some embodiments, corresponding designs and applications are carried out for specific scenarios. Specifically, in a more specific embodiment, the developed fire risk analysis model consists of one event tree (fire event tree) and four fault trees (early detection fault tree, decision fault tree, manual fire extinguishing fault tree, and system fire extinguishing fault tree).

[0196] Event tree construction: Taking the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event, the analysis is carried out step by step according to different levels:

[0197] Level 1 "Land Type": Considering the differences in fixed fire suppression systems equipped with different structural types of roll-on / roll-off spaces, it is divided into four categories: "Closed Space - Carbon Dioxide", "Closed Space - Pressurized Water Mist", "Open Space" and "Open Deck".

[0198] The second level, "early detection," refers to the detection by fixed automatic fire detection and alarm systems or the discovery by personnel (crew or passengers) of thermal runaway of the battery or fire in a new energy vehicle. It is divided into two scenarios: "successful early detection" and "failed early detection." Early detection is more beneficial for crew decision-making, but it is also crucial to the success of manual firefighting. Successful early detection means that the initial fire may be extinguished or suppressed, while failed early detection means that the fire has already developed, directly leading to the inability to conduct manual firefighting. Even fixed fire suppression systems may fail to extinguish the fire.

[0199] The third level, "Decision-Making," refers to the crew's (including the captain's) decisions regarding fire handling, including on-site emergency response, fire suppression using fixed fire extinguishing systems, and personnel evacuation. It is divided into two scenarios: "Correct Decision" and "Incorrect Decision." In the case of a "Correct Decision," manual firefighting may succeed or fail, and fire suppression using fixed fire extinguishing systems may succeed or fail. However, an "Incorrect Decision" will directly lead to the failure of manual firefighting; the system cannot completely extinguish the fire, but can only suppress it, which may succeed or fail.

[0200] Level 4, "Manual Firefighting," refers to the crew's emergency response to an initial fire. It is divided into two scenarios: "Successful Manual Firefighting" and "Failed Manual Firefighting." If "Successful Manual Firefighting," the incident ends; if "Failed Manual Firefighting," the process proceeds to Level 5.

[0201] Level 5, "System Fire Extinguishing": This refers to the evacuation of on-site personnel and the activation of the fixed fire suppression system. It is divided into two basic scenarios: "System Fire Extinguishing or Suppression Successful" and "System Fire Extinguishing or Suppression Failed." "System Fire Extinguishing Successful" means the fire is completely extinguished, while "System Suppression Successful" means the fire is not completely extinguished but is contained within the affected vehicle and adjacent vehicles, maintaining stable combustion until it is extinguished. This depends on the correctness of the decision made at Level 3. If the decision is correct, it may result in either "System Fire Extinguishing or Suppression Successful" or "System Fire Extinguishing or Suppression Failed." If "System Fire Extinguishing or Suppression Successful," the event ends. However, an "Incorrect Decision" will directly lead to system fire extinguishing failure, with the consequence being either "System Suppression Successful" or "System Suppression Failed." Regardless of the success of suppression, the process will proceed to Level 6. Open decks do not have this stage because they lack a fixed fire suppression system.

[0202] Level 6, "Fire Limitation": This refers to a situation where a fixed fire suppression system has failed to extinguish a fire (including a fire that has been suppressed but is now burning stably), and the fire continues to spread, but is limited by the fire-resistant structure of the premises. It is divided into two scenarios: "Fire Limitation Successful" and "Fire Limitation Failed." "Fire Limitation Failed" means that the actual fire temperature load exceeds the standard temperature rise curve or lasts for more than one hour, or that the compartment's sealing is poor, such as an open door or the presence of openings, which will cause fire smoke to spread to other areas. "Fire Limitation Successful" ends the event; "Fire Limitation Failed" proceeds to Level 7. Open decks, lacking structural separation, do not have this node.

[0203] The fault tree includes:

[0204] Early detection fault tree: If early detection fails, it is associated with the second level of the event tree; depending on whether there is a fixed automatic fire detection and fire alarm system in the roll-on / roll-off space, it is divided into three subtrees: "early detection fault tree for closed spaces", "early detection fault tree for open spaces" and "early detection fault tree for open decks".

[0205] Decision Fault Tree: Decision errors are associated with the third level of the event tree;

[0206] Manual fire suppression fault tree: Initial fire suppression failure is associated with the 4th level of the event tree; based on the location type, it is divided into 3 subtrees: "closed location manual fire suppression fault tree", "open location manual fire suppression fault tree" and "open deck manual fire suppression fault tree".

[0207] System fire suppression fault tree: When the system fails to extinguish or suppress the fire, it is associated with the 5th level of the event tree. Based on the differences in location type and the type of fixed fire suppression system installed, and considering the correctness of the overall decision, it is divided into 6 subtrees: "CO2 closed location - correct decision - system fire suppression or suppression failure", "CO2 closed location - incorrect decision - system fire suppression or suppression failure", "Pressurized water mist closed location - correct decision - system fire suppression or suppression failure", "Pressurized water mist closed location - incorrect decision - system fire suppression or suppression failure", "Open location - correct decision - system fire suppression or suppression failure", and "Open location - incorrect decision - system fire suppression or suppression failure".

[0208] (5) Risk Analysis Model Construction

[0209] In a preferred embodiment, a specific risk model for a specific scenario is constructed, including:

[0210] Let f0 be the baseline fire frequency in ship roll-on / roll-off (Ro-Ro) spaces; fire data related to gasoline-powered vehicles, obtained from publicly available domestic and international vehicle fire statistics, are used to determine the proportional relationship. Let k be the fire probability ratio coefficient between new energy vehicles and gasoline-powered vehicles. Then, the fire probability specifically for new energy vehicles in ship Ro-Ro spaces is:

[0211] f ev =k×f0

[0212] A hierarchical probability quantification system is established, which specifically includes: a location type classification module, an early detection failure analysis module, a decision logic quantification module, a manual fire extinguishing efficiency module, a system fire extinguishing efficiency module, and a fire limitation assessment module;

[0213] Location type classification module: Set the proportions of "closed location - carbon dioxide", "closed location - pressurized water mist", "open location" and "open deck" as p1, p2, p3, p4 respectively, and p1+p2+p3+p4=1;

[0214] Early detection failure analysis module: Calculates the early detection failure probability q1 for closed premises, the early detection failure probability q2 for open premises, and the early detection failure probability q3 for open decks using the fault tree quantization method.

[0215] Decision Logic Quantification Module: Obtains the probability r of decision error based on statistical data from accident cases;

[0216] Manual fire extinguishing efficiency module and system fire extinguishing efficiency module: Combining practical experience in fire extinguishing of new energy vehicles and physical tests, a "early detection-decision-fire extinguishing" linkage model is constructed according to different types of locations;

[0217] Let the probability of artificial fire suppression be s for "closed space - carbon dioxide". 11 "Closed space - pressurized water mist" is s 12 "Open space" is s 13 And "open deck" is s 14 ;

[0218] Let the system fire extinguishing probability be s for “closed space - carbon dioxide”. 21 "Closed space - pressure water mist protection" is s 22 "Open space" is s 23 And "open deck" is 1;

[0219] Fire Limitation Assessment Module: Calculates the fire limitation success rate within the premises based on accident case statistics. Let the fire limitation success rates be t1 for "closed premises - carbon dioxide", t2 for "closed premises - pressurized water mist", t3 for "open premises", and 1 for "open deck" with no fire limitation nodes.

[0220] The probability of occurrence for each accident scenario is calculated using the following formula:

[0221] P1 = f ev ×p1×q1×r×s 11 ×s 21 ×t1

[0222] P2 = f ev ×p2×q1×r×s 12 ×s 22 ×t2

[0223] P3 = f ev ×p3×q2×r×s 13 ×s 23 ×t3

[0224] P4 = f ev ×p4×q3×r×s 14 ×1×1

[0225] In the formula, P1, P2, P3 and P4 correspond to the scene, the closed space-pressure water mist, the open space and the open deck scene, respectively.

[0226] Furthermore, this embodiment further quantifies how to quantify the consequences of an accident through a property damage classification method and calculate the risk value using potential property damage, as follows:

[0227] Let the set of accident scenarios be A = {A1, A2, ..., A...} n}, where A i Let P(A) represent the i-th accident scenario, and let P(A) represent the probability of each accident scenario occurring. i This probability is calculated using a model that combines event trees and fault trees;

[0228] V(A i ) represents the property damage value under the i-th accident scenario, determined according to the property damage classification method;

[0229] PLP represents potential property loss, which is the expected value of property loss per year.

[0230] Potential property damage PLP is the sum of the products of the probability of occurrence of each accident scenario and the corresponding property damage value under all accident scenarios. The calculation formula is as follows:

[0231]

[0232] By using a weighted summation method, the expected value of annual property loss for ships carrying new energy vehicles is obtained, thereby quantifying the overall property loss risk value.

[0233] In a preferred embodiment, an asset value benchmark model is constructed based on actual conditions, encompassing a ship asset calculation module, a vehicle asset hierarchical calculation module, and a vehicle cargo additional asset calculation module. A fire loss scenario-based model defines five levels of loss scenarios corresponding one-to-one with five fire scenarios, and associates them with disposal costs, establishing a dynamic mapping relationship between fire development and property loss. In scenario-based loss prediction, the battery thermal runaway risk unique to new energy vehicles is incorporated into the ship loss assessment system. A thermal propagation coupling algorithm is introduced, and a threshold for the heat resistance time of the fire-resistant structure is set. When the fire duration exceeds the threshold, a loss coefficient of 50% total loss is automatically triggered.

[0234] Table 1. Fire scenarios and their corresponding Level 5 loss scenarios.

[0235]

[0236] Furthermore, in step S4, risk criteria are set, risk control measures are determined, and risk control options are selected and combined. Benefit and cost analyses are then performed on each risk control option. The specific method is as follows:

[0237] Risk criteria are defined as follows: This assessment considers only property damage and adopts the following principles, with the lowest cost / benefit ratio taking precedence:

[0238] ΔC<ΔB

[0239] In the formula, ΔC is the input cost of RCO; ΔB is the relevant benefit obtained after reducing risk;

[0240] The costs ΔC and ΔB are allocated over the entire lifespan of the vessel, calculated using the following formula:

[0241]

[0242] In the formula, A is the initial cost of implementing RCO, T is the ship's life cycle, r is the depreciation rate, and X is the cost or benefit of RCO in a specific year;

[0243] Preferably, in some preferred embodiments, Risk Control Options (RCOs) are further proposed. Possible risk control measures are identified, and measures that can reduce the frequency of accidents and mitigate their consequences are selected, forming a list containing all control measures. Subsequently, the risk control measures are screened and combined to determine a certain number of applicable risk control options.

[0244] Preferably, in some preferred embodiments, a Risk Control Option (RCO) benefit analysis is further proposed. Each RCO is analyzed individually, incorporating expert judgment and literature review, to assess the impact of each RCO implementation on each node of the fault tree. The impact of each RCO is calculated separately (assuming no other RCOs are implemented). Using a ship without RCOs as a reference, the risk value is recalculated using an event tree to obtain the PLP after each RCO implementation, and then the risk reduction over the ship's entire life cycle is calculated. The benefits of RCOs are calculated based on the reduction in property damage, excluding benefits such as reduced loss of life, reduced environmental damage, reduced liability to third parties, extended average ship lifespan due to compensation, and reduced insurance premiums.

[0245] Preferably, in some preferred embodiments, a Risk Control Option (RCO) cost analysis is further proposed. The cost over the entire lifecycle is calculated for each RCO, taking into account the periodic replacement of certain equipment.

[0246] Preferably, in some preferred embodiments, a cost-benefit analysis is further proposed. Based on the established risk criterion ΔC < ΔB, the cost-benefit of each RCO is evaluated. For a specific RCO, if the cost incurred is less than the benefit, the RCO is considered cost-effective; conversely, if the cost incurred is greater than the benefit, the RCO is considered cost-ineffective.

[0247] Ultimately, considering cost-effectiveness, we recommend risk control measures to customers to ensure the overall safety of ships transporting new energy vehicles.

[0248] Example 2

[0249] Please refer to Figure 3 This embodiment 2 provides a fire risk analysis system for new energy vehicles transported by ships, including:

[0250] The loading condition and fire scenario assumption unit is used to make assumptions about the loading condition and fire scenario;

[0251] The assumption unit for loss of life and property loss is used to determine the assumptions for loss of life and property loss.

[0252] The risk identification preparation unit is used to establish a risk matrix and propose hazard identification keywords for the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, it divides the process into nodes including loading, transportation and unloading and clarifies the specific operation behavior of each node.

[0253] The risk analysis model building unit is used to construct a model consisting of an event tree and a fault tree. The event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions. The fault tree is associated with the corresponding layer of the event tree and is subdivided into multiple subtrees according to different scenarios and conditions to systematically analyze the development process of fire risk.

[0254] The risk assessment and control decision-making unit is used to determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system; quantify the consequences of an accident using a property loss classification method; calculate the risk value using potential property loss; set risk criteria; determine risk control measures; screen and combine risk control options; conduct benefit and cost analysis on each risk control option; evaluate cost-effectiveness based on risk criteria; and finally recommend the most cost-effective risk control measures to the client.

[0255] Example 3

[0256] This embodiment 3 also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement any step of the fire risk analysis method for new energy vehicles transported on ships.

[0257] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0258] For a description of the computer-readable storage medium provided in this application, please refer to the above method embodiments; further details will not be repeated here.

[0259] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for analyzing the fire risk of new energy vehicles transported by ships, characterized in that, include: S1. Assumptions are made regarding the loading conditions and the fire scenario; S2. Determine the assumptions regarding loss of life and property damage; S3. Propose guiding words for hazard identification and establish a risk matrix for the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, divide the nodes including loading, transportation and unloading and clarify the specific operation behavior of each node; S4. Construct a model consisting of an event tree and a fault tree; the event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event, and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions; the fault tree is associated with the corresponding layer of the event tree, and is subdivided into multiple subtrees according to different scenarios and conditions, and the fire risk development process is systematically analyzed. S5. Determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system; quantify the consequences of an accident by using a property loss classification method and calculate the risk value using potential property loss. Establish risk criteria, determine risk control measures, screen and combine risk control options, and conduct benefit and cost analysis on each risk control option; finally, evaluate cost-effectiveness based on risk criteria, and recommend the most cost-effective risk control measures to the client.

2. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, The assumption methods for making assumptions about loading conditions and fire scenarios in S1 include: Loading assumptions: Assuming the ship meets current regulations and standards and is not carrying dangerous goods; Assume two loading scenarios: the roll-on / roll-off space is entirely loaded with new energy vehicles, and the roll-on / roll-off space is mixed with new energy vehicles and fuel vehicles. Fire scenario assumptions: Only the fires of new energy vehicles, the mutual influence between fires of new energy vehicles and fires of fuel vehicles, and the impact of fires of vehicles and cargo on new energy vehicles are considered. The conventional fires of fuel vehicles and cargo themselves and their impact are not assessed. Based on the development of fire hazards in new energy vehicles and the effectiveness of active and passive fire suppression systems on ships, a tiered fire scenario is constructed, including: Fire Scenario A: Assume that a single vehicle fire is successfully extinguished, that is, the thermal runaway or fire of a single new energy vehicle battery is detected and the correct decision is made in time, and the crew successfully suppresses or extinguishes the fire within the vehicle's range. Fire Scenario B: Assume that no fewer than 3 vehicles are successfully extinguished, that is, if a single new energy vehicle fire is not successfully suppressed or extinguished, it ignites no fewer than 2 adjacent vehicles, which are then successfully extinguished or suppressed by the fire extinguishing system within the range of no fewer than 3 vehicles. Fire Scenario C: Assume that the fire was not extinguished but the fire was contained within the location of the fire. That is, a single new energy vehicle fire was not successfully extinguished or suppressed, igniting surrounding vehicles, but the fire extinguishing system failed to extinguish or suppress the fire. The entire floor of vehicles was burned or the battery experienced thermal runaway, but the fire was contained within the location of the fire. Fire Scenario D: Assume that firefighting has failed and the fire is not contained within the location of origin; Fire Scenario E: This scenario specifically refers to the assumption that manual firefighting on the open deck has failed.

3. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, The specific assumptions made in S2 regarding the determination of loss of life and property loss are as follows: Assuming that the loss of life caused by a fire in a ship carrying a new energy vehicle is the same as that caused by a fire in a fuel vehicle, that is, the ship carrying a new energy vehicle does not pose an additional risk of loss to people. Assuming that the consequences of a fire involving a ship carrying new energy vehicles only consider property damage, the extent of which depends on the fire's progression within the premises and the effectiveness of the ship's active and passive fire suppression systems, as detailed below: Let S be the set of fire scenarios, S = {S1, S2} s ,…,S n }, where {A,B,C,D,E}∈S; for each scenario s∈S, there is a corresponding set of vehicles V. s The vehicle number is i, and the value of each vehicle is v. i The loss coefficient is δ i (0<δ i <1), the hull structure maintenance cost corresponding to scenario s is C. s The indirect losses caused by the impact on the lease term are C. t The total loss value of the ship is V ship Therefore, the method for calculating property loss is as follows: Where V represents the specific loss assuming a fire occurs.

4. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, The specific method for proposing hazard identification keywords and establishing a risk matrix for new energy vehicle failure modes in S3 is as follows: Let the set of hazard sources be: Ω={ω1,ω2,…,ω n }, where ω i Indicates the i-th hazard source; The set of operation nodes is: Π={π1,π2,…,π m }, where π j This represents the j-th operation node; The risk matrix R is a two-dimensional matrix. ij R represents the risk level of the i-th hazard source at the j-th operation node. ij ∈{1,2,…,k}, where k is the number of risk levels; P ij S represents the probability that the i-th hazard will occur at the j-th operation node. ij The severity of its consequences; For the guiding words in hazard identification, let the set of guiding words be G = {g1, g2, ..., g...} l }, the introductory word g s The weight is ω s ,and r is This indicates that the i-th hazard source is indicated by the introductory word g. s The score The comprehensive score for hazard sources is calculated as follows: In the formula, T i This represents the overall score of the i-th hazard source, reflecting the potential hazard level of the i-th hazard source; The probability P of the i-th hazard source at the j-th operation node ij and severity of consequences S ij Combined score T of hazard sources i Related, while also considering the influence factor α of the characteristics of the operating nodes. ij and β ij Then we have: P ij =a ij T i S ij =b ij T i Based on probability P ij and severity of consequences S ij Determine the risk level R in the risk matrix ij This transformation can be achieved using a mapping function f: R ij =f(P ij ,S ij ) The likelihood and severity of consequences are divided into different ranges, and the risk level is determined based on the combination of these ranges: The overall risk of the entire transportation process can be obtained by weighted summation of the elements in the risk matrix, where the weight of each operation node is q. j ,and but: In the formula, R total This indicates the overall risk of the entire transportation process.

5. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, The specific method for constructing the model consisting of an event tree and a fault tree in S4 is as follows: The top event in the fault tree that is related to the initial event of the event tree is defined as T, and its occurrence probability is P(T), which is calculated by combining the basic events in the fault tree through logic gates. Define the parameters related to the event tree: Let the number of levels in the event tree be L, and the event set of the l-th level be... Where n l It is the number of events in the l-th layer; Event e ij The probability of occurrence is denoted as P(e ij ); The transition probability from the 1st level to the (1+1)th level is denoted as P(e (l+1)k |e lj ), indicating that the event e is at level l. lj Under the condition that it occurs, the event e at level l+1 (l+1)k The probability of occurrence; The method for calculating the probability of events in an event tree is as follows: The first level event probability P(e) of the event tree 1j The probability of events in subsequent layers is given directly based on experience or statistical data; the probability of events in subsequent layers is calculated using the conditional probability formula. This indicates that event e at level l+1 (l+1)k The probability of occurrence is equal to the probability of all possible events e at level l. lj Under what conditions does e occur (l+1)k The probability of occurrence is weighted sum, with weights equal to the probability of the l-th level event e. lj The probability of occurrence; Constructing an overall risk model: Each terminal event e in the event tree lk For a specific accident scenario, the probability of its occurrence is: Where k1,k2,…,k L It is the event index path from the first level to the Lth level; Let each terminal event e LK The corresponding severity of the consequences is C(e) LK If the system risk R represented by the entire model is: In other words, system risk equals the sum of the products of the probability of all end events occurring and the severity of their corresponding consequences.

6. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, The top event T in the fault tree that is related to the initial event of the event tree, and the specific calculation method for its occurrence probability P(T) is as follows: Define the parameters related to the fault tree: The basic event set of the fault tree is B = {b1, b2, ..., b}. m }, basic event b i The probability of occurrence is denoted as P(b) i ); The set of intermediate events is M = {m1, m2, ..., m} s The top event is T; Logic gates are represented by ⊙; for intermediate events m i It is composed of basic events or other intermediate events combined through logic gates; The method for calculating the event probability in a fault tree is as follows: In the AND gate case: if intermediate event m i From basic event b i1 ,b i2 ,…,b it By connecting to the door, then: This is because the AND gate requires that the output event occurs only when all input events occur, and this formula is derived from the multiplication principle of the probabilities of independent events; In the OR gate case: if the intermediate event mi is connected by the basic events bi1, bi2, ..., bit through an OR gate, then: The principle is to first calculate the probability that none of the input events will occur, and then subtract this probability from 1 to get the probability that at least one input event will occur. For the top event T, by recursively applying the above rules, the probabilities of intermediate events are calculated step by step starting from the basic events, and finally the probability P(T) of the top event is obtained.

7. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, In S4, the consequences of an accident are quantified using a property damage classification method, and the risk value is calculated using potential property damage. The specific method is as follows: Let the set of accident scenarios be A = {A1, A2, ..., A...} n }, where A i Let P(A) represent the i-th accident scenario, and let P(A) represent the probability of each accident scenario occurring. i This probability is calculated using a model that combines event trees and fault trees; V(A i ) represents the property damage value under the i-th accident scenario, determined according to the property damage classification method; PLP represents potential property loss, which is the expected value of property loss per year. Potential property damage PLP is the sum of the products of the probability of occurrence of each accident scenario and the corresponding property damage value under all accident scenarios. The calculation formula is as follows: By using a weighted summation method, the expected value of annual property loss for ships carrying new energy vehicles is obtained, thereby quantifying the overall property loss risk value.

8. The method for fire risk analysis of new energy vehicles transported by ship according to claim 1, characterized in that, In step S4, risk criteria are set, risk control measures are determined, and risk control options are screened and combined. Benefit and cost analyses are then performed on each risk control option. The specific method is as follows: Risk criteria are defined as follows: This assessment considers only property damage and adopts the following principles, with the lowest cost / benefit ratio taking precedence: ΔC<ΔB In the formula, ΔC is the input cost of RCO; ΔB is the relevant benefit obtained after reducing risk; The costs ΔC and ΔB are allocated over the entire lifespan of the vessel, calculated using the following formula: In the formula, A is the initial cost of implementing RCO, T is the ship's life cycle, r is the depreciation rate, and X is the cost or benefit of RCO in a specific year.

9. A fire risk analysis system for ship-borne new energy vehicles, characterized in that, include: The loading condition and fire scenario assumption unit is used to make assumptions about the loading condition and fire scenario; The assumption unit for loss of life and property loss is used to determine the assumptions for loss of life and property loss. The risk identification preparation unit is used to establish a risk matrix and propose hazard identification keywords for the failure modes of new energy vehicles; according to the operation process of ship transport vehicles, it divides the process into nodes including loading, transportation and unloading and clarifies the specific operation behavior of each node. The risk analysis model building unit is used to construct a model consisting of an event tree and a fault tree. The event tree takes the fire of a new energy vehicle in a roll-on / roll-off facility as the initial event and analyzes different situations layer by layer according to the facility type, early detection, decision-making, manual fire fighting, systematic fire fighting, and fire restrictions. The fault tree is associated with the corresponding layer of the event tree and is subdivided into multiple subtrees according to different scenarios and conditions to systematically analyze the development process of fire risk. The risk assessment and control decision-making unit is used to determine the probability of an accident by establishing an initial event probability calculation module and a hierarchical probability quantification system. Quantify the consequences of an accident using a property damage classification method; Risk values ​​are calculated using potential property damage. Establish risk criteria, determine risk control measures, screen and combine risk control options, and conduct benefit and cost analysis on each risk control option; evaluate cost-effectiveness based on risk criteria, and finally recommend the most cost-effective risk control measures to the client.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor according to any one of claims 1-8, which describes a method for analyzing the fire risk of new energy vehicles transported by ships.

Images