A Resilience Assessment Method for Storage Tank Facilities Defense System under Multi-Hazard Coupling

Abstract

The present invention discloses a method for evaluating the resilience of a storage tank facility defense system under the coupling of multiple disasters, including obtaining basic information; analyzing the performance degradation of safety barriers and the vulnerability of equipment under natural disasters, identifying initial accident scenarios and the possibility of accident expansion; constructing a dynamic Bayesian network to obtain the node probability values of each equipment at different time points; evaluating the system function according to the possible economic losses of the tank farm system at different time nodes; evaluating the system resilience according to the system function, and drawing a time-varying curve of the system resilience. Due to the cascading effect of multi-disaster coupling accidents, the performance of the original defense system may be reduced due to natural disasters, resulting in a decline in system resilience and making it difficult to resist the further expansion of accidents. By using the method of the present invention, the system resilience evaluation at different stages of accident evolution under the coupling of multiple disasters can be realized, providing a theoretical basis and decision-making support for the prevention and control of multi-disaster coupling accidents and the optimization of the defense system.

Description

Technical Field

[0001] The present invention relates to the technical field of chemical process safety, and specifically to a method for evaluating the resilience of a storage tank facility defense system under the coupled action of natural disasters and industrial accidents. Background Art

[0002] A safe and stable energy system is an important foundation for the development of the national economy. As an important part of the national energy system, the resilience of the defense system of major oil and gas storage and transportation infrastructure to disasters is an important guarantee for the effective operation of the system and oil and gas supply. Major oil and gas storage and transportation infrastructure mainly includes three categories: oil and gas storage areas, long-distance pipelines, and oil and gas stations, and shows the characteristics of large-scale, large-sized, and centralized installations, resulting in a large concentration of hazardous chemicals in the installations. Once an accident occurs, it is extremely easy to affect the surrounding installations, leading to the multi-level expansion of the accident. China has a vast territory, complex and changeable meteorological conditions and geographical environment, and is affected by various natural disasters, and the disasters show a trend of frequent occurrence. Once a natural disaster acts on major oil and gas storage and transportation infrastructure, it is extremely easy to cause multi-source initial accidents, damage the surrounding safety barriers, lead to the rapid evolution of the accident, and form a multi-disaster coupling scenario. Storage tanks are typical equipment and facilities for storing hazardous substances in major oil and gas storage and transportation infrastructure. Therefore, evaluating the resilience of the storage tank facility defense system under a multi-disaster coupling scenario can provide theoretical guidance for improving the oil and gas storage and transportation safety system, and has great practical significance.

[0003] Resilience assessment has received extensive attention from the academic community in recent years. Compared with risk assessment for a single accident type, resilience assessment focuses on the resistance and recovery ability of the system to different accident types. At present, domestic and foreign scholars use the index system method and the time-based resilience index method to evaluate the resilience of the power grid system under the action of natural disasters, and propose improvement strategies for the subsequent defense system. However, the storage tank facility system is different from the power grid system. Once it is impacted by natural disasters, it will not only cause damage to equipment, but also the hazardous substances leaked after equipment damage may trigger a fire accident and further evolve and expand, resulting in a multi-disaster coupling scenario, which is a great challenge to the system resilience. Existing resilience assessment methods are difficult to evaluate the system resilience under the evolution process of this coupled accident, and the impact of multi-disaster disturbances changing with time on the system function is the focus of the resilience assessment method. Summary of the Invention

[0004] Aiming at the problems existing in the existing resilience assessment technology for storage tank facilities, the present invention provides a method for evaluating the resilience of a storage tank facility defense system under multi-disaster coupling.

[0005] This method takes into account the harmful effects of natural disasters or technological disasters on the system and the system's own resistance ability during different time stages of the multi-hazard coupling accident evolution process. Starting from the system function characterization function, it analyzes the system resilience, which can provide theoretical support for improving the defense system and disaster prevention and mitigation. The specific implementation steps are as follows:

[0006] Step (1): Collect the basic information for resilience assessment under the multi-hazard coupling scenario;

[0007] Step (2): According to the vulnerability model of storage tank equipment, calculate the failure probability of each storage tank facility under the action of specific natural disasters. Given the initial accident unit identification threshold, identify the most likely initial accident unit based on the failure probability and the initial accident unit identification threshold. Then, evaluate the performance degradation of the safety barrier under natural disasters, and mark the moment when the initial accident unit is affected by natural disasters as the starting point t of resilience assessment 0 ;

[0008] Step (3): Predict the possible initial accident scenarios according to the ignition time, mark the moment of the initial accident as t 1 , evaluate the intensity of the escalation vector generated by the initial accident, calculate the actual received escalation vector intensity of each storage tank facility under different safety barrier states, determine the secondary unit after comparing with the accident escalation threshold, and calculate the failure time of the secondary unit. After the moment of t 1 , superimpose the failure time of the secondary unit, and record this moment as t 2 , and calculate the escalation probability of the secondary unit. During this process, the synergistic effect between accidents is represented by the superposition of escalation vectors; identify the secondary accident scenario, substitute the secondary accident into the initial accident, repeat the previous process, and calculate the failure time of the tertiary unit. After the moment of t 2 , superimpose the failure time of the tertiary unit, denoted as t 3 , and so on, until all equipment units are involved in the accident chain or there is not enough escalation vector to cause accident expansion;

[0009] Step (4): Set natural disasters, safety barriers, and chemical equipment as nodes of the dynamic Bayesian network, connect the nodes according to the accident expansion sequence, and construct the dynamic Bayesian network to obtain the node probability values of each equipment at different time points. Among them, the node probability is determined as the accident probability of the equipment at a specific time point;

[0010] Step (5): Calculate the possible economic losses of the tank farm system at a specific time point, calculate the system performance at each time point, and then determine the system resilience based on the system performance. The system resilience refers to the ability of the system to resist or mitigate disaster disturbances and maintain the stability of the system function state. Among them, the calculation method of the system performance is:

[0011]

[0012] In the formula, represents the performance of the system at time t under the influence of the destructive event e j ; n represents the number of storage tanks covered by the tank farm system; v i and u i represent the equipment value of a single storage tank equipment i and the value of the stored hazardous substance respectively; P(i∣e j ,t) represents the failure probability of the storage tank equipment i at time t under the influence of the destructive event e j ; is the possible economic loss;

[0013] The determination method of the system resilience is as follows:

[0014]

[0015] In the formula, R(t∣e j ) represents the resilience of the system at time t under the influence of the destructive event e j ; represents the resilience of the system at the lowest moment t j of the system performance under the influence of the destructive event e d ; represents the initial performance level of the system.

[0016] Compared with the prior art, the method provided by the present invention has at least the following beneficial effects:

[0017] (1) Based on the quantitative analysis of the damage effect of multi-hazard coupled accidents on storage tank facilities, this method considers potential accident evolution scenarios, analyzes the system performance changes from different stages of accident expansion, and can more clearly describe the system resilience behavior.

[0018] (2) This method fully considers the impact of the performance degradation of safety barriers under the multi-hazard coupling effect on system resilience. By using Bayesian network to model the accident evolution process and calculate the equipment accident probability, it can solve the complex link expansion problems involving natural disasters, safety barriers, and multi-source initial accidents.

[0019] (3) This method analyzes the resistance ability of the system to multi-hazard coupled accidents based on the concept of resilience, considering both the resistance ability of the equipment itself to natural disasters and industrial accidents and the ability of safety barriers to prevent or delay accident expansion, which is more in line with the actual engineering situation.

[0020] (4) This method can identify key equipment units related to the improvement of system resilience during the design and operation stages of the storage tank facility defense system, provide a theoretical basis for the layout of safety barriers, and effectively improve the system resilience level. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the resilience evaluation process of the storage tank facility defense system under multi-hazard coupling provided by the embodiments of the present invention.

[0022] Figure 2 It is a schematic diagram of the Bayesian network for accident evolution of the storage tank facility defense system under multi-hazard coupling provided by the embodiments of the present invention.

[0023] Figure 3 It is a schematic diagram of the resilience curve of the storage tank facility defense system under multi-hazard coupling provided by the embodiments of the present invention. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0025] A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling based on a dynamic Bayesian network and a system performance function provided by the present invention combines an equipment vulnerability model under natural disasters with an accident expansion probability model under the action of a defense system, evaluates the system function through the possible economic losses at a certain accident stage, and further converts it into system resilience through a function. In this process, the damage of natural disasters to the defense system is considered, and the directed graph topological structure of the dynamic Bayesian network is used to describe the accident evolution process of multi-hazard coupling. On the one hand, the possible high-order accident expansion is considered, and on the other hand, the change in the system's resistance to disasters during the multi-hazard coupling process is considered. Please refer to Figures 1 - 3 , and the method specifically includes the following steps:

[0026] Step (1): Collect the basic information for resilience evaluation in a multi-hazard coupling scenario;

[0027] In some embodiments of the present invention, the basic information includes information on natural disaster characteristics, safety barrier characteristics, storage tank facility characteristics, meteorological conditions, and environmental characteristics.

[0028] Step (2): According to the existing vulnerability model of storage tank equipment under natural disasters, calculate the failure probability of each storage tank facility under the action of specific natural disasters. Given the initial accident unit identification threshold, when the equipment failure probability obtained through the equipment vulnerability model is greater than the initial accident unit identification threshold, the equipment is identified as a possible initial accident unit. And evaluate the performance degradation of the safety barrier under natural disasters through the existing safety barrier performance correction model to support the subsequent accident expansion probability prediction. And mark the moment when the initial accident unit is affected by natural disasters as the starting point t of resilience assessment 0 ;

[0029] Among them, the vulnerability model of storage tank equipment under natural disasters adopts the existing model. For example, Huang Kongxing et al. disclosed the vulnerability model of chemical equipment under natural disasters in "Review on Quantitative Risk Assessment and Prevention and Control System of Na-Tech Events in Chemical Industrial Parks [J]. Chemical Industry and Engineering Progress, 2019, 38(7): 3482-3494". The safety barrier performance correction model is an existing model. For example, Misuri et al. disclosed the safety barrier performance correction model in "Assessment of safety barrier performance in the mitigation of dominoscenarios caused by Natech events [J]. Reliability Engineering and System Safety, 2021, 205: 107278".

[0030] In some embodiments of the present invention, the action of natural disasters will cause Figure 1 the storage tank facilities T1 and T2 in to fail, and the storage tank facilities T1 and T2 are used as the initial accident units. And the performance of the safety barrier related to the storage tank facility T3 degrades under natural disasters, while the safety barrier related to the storage tank facility T4 maintains its original performance. Mark the moment when the initial accident unit is affected by natural disasters as the starting point t of resilience assessment 0 。

[0031] In some embodiments of the present invention, a given initial accident unit identification threshold is 0.1, that is, when the equipment failure probability obtained through the equipment vulnerability model is greater than 0.1, the equipment is identified as a possible initial accident unit.

[0032] Step (3): Predict possible initial accident scenarios according to the ignition time, mark the initial accident occurrence time as t 1 , evaluate the intensity of the escalation vector generated by the initial accident. Calculate the actual escalation vector intensity received by each storage tank facility in different safety barrier states, compare it with the accident escalation threshold to determine potential secondary units, and calculate the failure time of the secondary units, at t 1After the moment, superimpose the failure time of the secondary unit, and record this moment as \(t\). 2 , calculate the upgrade probability of the secondary unit. During this process, the synergistic effect between accidents can be represented by the superposition of upgrade vectors; identify the secondary accident scenario, substitute the secondary accident into the initial accident, repeat the previous process, and calculate the failure time of the tertiary unit. After time \(t\) 2 , superimpose the failure time of the tertiary unit, and record it as \(t\) 3 , and so on, until all storage tank facilities are involved in the accident chain, or there are not enough upgrade vectors to trigger accident expansion;

[0033] Among them, when the intensity of the upgrade vector received by the equipment is higher than the corresponding accident upgrade threshold, the equipment is identified as a potential secondary unit, and it may cause accident expansion due to the action of the initial accident, that is, a secondary accident occurs. The equipment that may cause accident expansion due to the accident of the secondary unit is a tertiary unit, that is, the equipment whose received upgrade vector intensity is higher than the accident upgrade threshold under the action of the secondary unit accident.

[0034] Among them, the assessment of the intensity of the upgrade vector generated by the initial accident can be carried out through quantitative consequence analysis software (such as DNV Phast) or by using theoretical formulas (Quantitative Risk Assessment Guidelines for Petrochemical Plants [M]. China Petrochemical Press, 2007, by Qingdao Safety Engineering Research Institute of China National Petroleum and Chemical Corporation).

[0035] For the intensity of the upgrade vector actually received by each storage tank facility, when the safety barrier fails, the intensity of the upgrade vector received by the target storage tank does not decrease; when the safety barrier is normally enabled, the intensity of the upgrade vector actually received by the target storage tank is reduced due to the action of the safety barrier, and the reduced intensity of the upgrade vector can be calculated through the existing model (Landucci et al., Quantitative assessment of safety barrier performance in the prevention of domino scenarios triggered by fire [J]. Reliability engineering & system safety, 2015, 143: 30 - 43).

[0036] The failure time of the secondary unit is calculated based on the existing model. Landucci et al. disclosed the relevant model in "Quantitative assessment of safety barrier performance in the prevention of domino scenarios triggered by fire[J]. Reliability engineering & system safety, 2015, 143: 30 - 43".

[0037] The upgrade probability of the secondary unit is calculated by the existing Probit model. For example, Landucci et al. in "Quantitative assessment of safety barrier performance in the prevention of domino scenarios triggered by fire[J]. Reliability engineering & system safety, 2015, 143: 30 - 43" disclosed the Probit model.

[0038] In some embodiments of the present invention, for atmospheric storage tanks, the accident upgrade threshold is taken as 15 kW / m 2 ; for pressure storage tanks, the accident upgrade threshold is taken as 45 kW / m 2 .

[0039] Specifically, in some embodiments of the present invention, the initial accident occurrence time is marked as t 1 , and the initial accident is a pool fire occurring in storage tank facilities T1 and T2. The intensity of the upgrade vector generated by the initial accident can be evaluated through consequence simulation software or empirical formulas. Combining the safety barrier state, calculate the actual upgrade vector intensity received by each device, determine that the secondary unit is T3, and calculate the failure time of T3 under the synergistic effect of the initial accident. After the time t 1 , push back the unit failure time, obtain the time t 2 , and calculate the upgrade probability of T3. Substitute T3 into the initial accident, repeat the above process, and calculate the failure time of T4, and obtain the time t 3 .

[0040] Step (4): Set natural disasters, safety barriers, and storage tank equipment as nodes of the dynamic Bayesian network, connect each node according to the accident expansion sequence, and construct a dynamic Bayesian network to obtain the node probability values of each device at different time points. Among them, it is determined that the node probability is the accident probability of the corresponding storage tank at a specific time point;

[0041] In some embodiments of the present invention, the constructed dynamic Bayesian network is shown in Figure (2), and the accident probability values of each storage tank facility node can be obtained. The nodes T1' and T2' are auxiliary nodes used to calculate the fire probabilities of the storage tanks T1 and T2. Figure 2 SB3 and SB4 in Figure 2 respectively represent the safety barriers related to the storage tanks T3 and T4. The arcs connecting the nodes represent the causal relationships between the nodes. The node from which the arc starts is called the parent node, and the node to which the arc points is called the child node. The number on the arc represents the order of the arc, which can be used to represent the time delay effect, that is, the current state of the child node depends on the state of the parent node in the previous time period. For example, the first-order arc from the storage tank equipment T1 to the auxiliary node T1' can represent the causal relationship between the failure of the storage tank equipment T1 and the occurrence of a fire accident in the storage tank equipment T1. Due to the influence of the ignition time, whether the fire accident represented by T1' occurs depends on whether the storage tank equipment T1 failed in the previous time period. The first-order arc from T1' to T3 represents whether the accident of the storage tank equipment T3 escalates, which depends on the state of the fire accident of T1 before the failure time of T3. The second-order arc between T1' and T4 can represent the difference between the domino expansion levels. As a third-level device, T4 has a probability of accident escalation only after the accident of T3 escalates. Therefore, the arc between T3 and T4 is a first-order arc, and the arc between T1 and T4 is a second-order arc.

[0042] Step (5): Calculate the possible economic loss of the tank farm system at a specific time point t. Calculate the system performance at each time point through formula (1). After completing the system performance calculation, obtain the system resilience value through formula (2) and connect the points to draw the system resilience curve. Among them, the system resilience refers to the ability of the system to resist or mitigate disaster disturbances and maintain the stability of the system function state. The setting of safety barriers can effectively prevent the spread of industrial accidents, but may lead to a decrease in the prevention and control ability due to the action of natural disasters, resulting in a decrease in system resilience.

[0043] In some embodiments of the present invention, the time-varying curve of the system resilience drawn is shown in Figure (3).

[0044]

[0045] In the formula, represents the performance of the system at time t under the influence of the destructive event e j ; n represents the number of storage tanks covered by the tank farm system; v i and u i respectively represent the equipment value of a single storage tank equipment i and the value of the stored hazardous substance; P(i∣e j ,t) represents the failure probability of the storage tank equipment i at time t under the influence of the destructive event e j ; is the possible economic loss.

[0046]

[0047] Wherein, R(t∣e j ) represents the resilience of the system at time t under the influence of the disruptive event e j . represents the resilience of the system at the moment of the lowest system performance (set as t j ) under the influence of the disruptive event e d . represents the initial performance level of the system, and t s represents the moment before the disruptive event. It is generally considered that the system maintains its complete performance and is usually set to 1.

[0048] Based on the vulnerability models of equipment and the calculation models of upgrade probabilities for different types of natural disasters, and considering the performance degradation of different safety barriers under natural disasters, the system resilience under different multi-hazard coupling scenarios can be quantitatively evaluated.

[0049] The present invention is based on the existing equipment failure probability models under natural disasters or technical disasters (such as the vulnerability model of storage tank equipment under natural disasters and the Probit model for calculating the accident upgrade probability in a fire scenario mentioned above), considers the degradation of safety barriers under natural disasters, obtains the failure probability of storage tank facilities under multi-hazard coupling based on a dynamic Bayesian network, obtains the possible economic losses of the system, characterizes the system state at different accident stages through system functions, and then evaluates the system resilience, in order to provide a reference for the disaster prevention and mitigation work of oil and gas storage and transportation systems under multi-hazard coupling accidents.

[0050] The above is one of the specific implementation cases of the present invention. If the changes made according to the concept of the present invention do not exceed the spirit covered by the specification and the drawings in terms of their functional effects, they should still fall within the protection scope of the present invention.

Claims

1. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling, where a multi-hazard coupling accident refers to an industrial accident chain triggered by natural disasters. It is characterized in that: It includes the following steps: Step (1): Collect the basic information for resilience evaluation under multi-hazard coupling scenarios. Step (2): According to the vulnerability model of storage tank equipment, calculate the failure probability of each storage tank facility under the action of specific natural disasters. Given the initial accident unit identification threshold, identify the most likely initial accident unit based on the failure probability and the initial accident unit identification threshold. Then, evaluate the performance degradation of safety barriers under natural disasters, and mark the moment when the initial accident unit is affected by natural disasters as the starting point t of resilience assessment 0 ; Step (3): Predict possible initial accident scenarios based on the ignition time, and mark the initial accident occurrence time as t 1 , evaluate the intensity of the escalation vector generated by the initial accident, calculate the actual intensity of the escalation vector received by each storage tank facility under different safety barrier states, determine the secondary units after comparing with the accident escalation threshold, and calculate the failure time of the secondary units. After t 1 moment, superimpose the failure time of the secondary units, and record this moment as t 2 , and calculate the escalation probability of the secondary units. During this process, the synergistic effect between accidents is represented by the superposition of escalation vectors; Identify the secondary accident scenario, substitute the secondary accident into the initial accident, repeat the previous process, and calculate the failure time of the tertiary unit. After time t 2 add the failure time of the tertiary unit, denoted as t 3 , and so on until all equipment units are involved in the accident chain or there are not enough upgrade vectors to trigger accident expansion; Step (4): Set natural disasters, safety barriers, and chemical equipment as nodes of a dynamic Bayesian network, connect the nodes according to the accident expansion sequence, and construct a dynamic Bayesian network to obtain the node probability values of each device at different time points. Among them, the node probability is recognized as the accident probability of the device at a specific time point. Step (5): Calculate the possible economic losses of the tank farm system at a specific time point, calculate the system performance at each time point, and then determine the system resilience based on the system performance. The system resilience refers to the ability of the system to resist or mitigate disaster disturbances and maintain the stability of the system function state. Among them, the calculation method of the system performance is: In the formula, represents the performance of the system at time t under the influence of the disruptive event e j ; n represents the number of storage tanks covered by the tank farm system; v i and u i represent the equipment value of a single storage tank equipment i and the value of the stored hazardous substance respectively; P(i∣e j ,t) represents the failure probability of the storage tank equipment i at time t under the influence of the disruptive event e j ; is the possible economic loss. The determination method of the system resilience is: Wherein, R(t∣e j ) represents the resilience of the system at time t under the influence of the disruptive event e j ; represents the resilience of the system at the lowest moment t j of the system performance under the influence of the disruptive event e d ; represents the initial performance level of the system.

2. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: The basic information in step (1) includes natural disaster characteristics, safety barrier characteristics, storage tank facility characteristics, meteorological conditions, and environmental characteristics.

3. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: Identifying the most likely initial accident unit based on the failure probability and the initial accident unit identification threshold in step (2) means that when the failure probability of a device is greater than the initial accident unit identification threshold, then the device is identified as a possible initial accident unit.

4. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: The safety barrier in step (2) is an important setting for industrial accident prevention and control, including an active protection system and a passive protection system.

5. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 4. It is characterized in that: The active protection system includes a sprinkler device, and the passive protection system includes a fireproof coating.

6. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: In step (2), the performance degradation of the safety barrier under natural disasters is evaluated through an existing safety barrier performance correction model.

7. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: In step (3), the intensity of the escalation vector generated by the initial accident is evaluated through quantitative consequence analysis software.

8. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: In step (3), each device is considered to be in a safe state before the time point related to the calculated failure time, and it is only considered to have the probability of an accident after the time point.

9. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to claim 1. It is characterized in that: In step (3), the upgrade probability of the secondary unit is calculated using the Probit model.

10. A method for evaluating the resilience of a storage tank facility defense system under multi-hazard coupling according to any one of claims 1-9, characterized in that after obtaining the system resilience value in step (5), connecting each point can draw the time-varying curve of the system resilience.

Images