Chemical fire domino effect simulation method and system based on dynamic fire source update of FDS

Abstract

The application discloses a chemical fire domino effect simulation method and system based on FDS dynamic fire source updating and belongs to the technical field of fire numerical simulation. The method comprises the following steps: establishing a basic model of a storage tank area, simulating a fire by using FDS, and monitoring thermal radiation of a target storage tank. If the storage tank reaches a failure condition, a failure time is recorded and the storage tank is set as a new fire source, a fire scene is updated, and simulation is repeated until there is no new failure. Finally, fire scenes and failure data at each stage are output, and a whole-process simulation result of the domino effect is formed. The application solves the technical problems of physical process distortion and inaccurate thermal radiation field calculation caused by fixed fire source combination and starting time in the prior art.

Description

Technical Field

[0001] This invention belongs to the field of fire numerical simulation technology, specifically relating to a method and system for simulating the domino effect of chemical fires based on FDS dynamic fire source updates. Background Technology

[0002] Tank farms, as core storage units in the chemical industry, often concentrate large quantities of flammable and explosive hazardous chemicals. If a tank catches fire unexpectedly, the resulting massive heat radiation can severely threaten the safety of adjacent tanks, causing them to fail and triggering new fires. This can ultimately create a "domino effect," a chain reaction of accidents escalating from a single incident to multiple tanks. This complex and serious accident propagation path is a key focus of chemical safety prevention and control.

[0003] Accurately predicting the domino effect of a fire is crucial for developing effective emergency plans and disaster prevention measures. Currently, numerical simulations of tank fires primarily utilize tools such as Fire Dynamics Simulators (FDS). However, most existing simulation methods focus on single fire scenarios or pre-defined, fixed combinations of fire sources. The common practice is to initially set all potentially ignitable tanks as fire sources and ignite them simultaneously, based on experience or worst-case assumptions, and then run the simulation. This approach has significant technical limitations: 1. Timing Distortion: The essence of the domino effect is the sequential process of an accident expanding over time. Simulating the simultaneous ignition of subsequent burning tanks at the initial moment completely ignores the order and time interval of the fire sources, resulting in severe distortion of the simulation results in the time dimension, failing to reflect the real physical process.

[0004] 2. Inaccurate Calculation of Thermal Radiation Field: Thermal radiation is a key driving force behind the domino effect. As a fire expands and new ignition sources emerge, complex interactions occur between these sources, including flame entrainment, smoke plume obstruction, flame deflection, and the migration of local hotspots. Current technologies simply perform algebraic superposition of the thermal radiation from multiple ignition sources, failing to capture these complex fluid dynamics and radiative heat transfer phenomena. This leads to severely inaccurate assessments of the intensity of thermal radiation received by the target storage tank.

[0005] 3. Inability to reflect dynamic response: The failure of the target storage tank depends not only on the intensity of thermal radiation but also on its heating history. Existing static simulation methods cannot dynamically adjust the subsequent ignition source state based on the simulation results of the previous stage, and therefore cannot simulate the cumulative heat dose response of the target storage tank under a dynamically changing thermal radiation field in a real accident.

[0006] In summary, there is an urgent need for a simulation method for the domino effect of tank fires that can take into account the dynamic timing of fire sources and more realistically calculate the coupled thermal radiation field of multiple fire sources, in order to overcome the shortcomings of existing technologies. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for simulating the domino effect of chemical fires based on FDS dynamic fire source updates. This method can solve the technical problems of physical process distortion and inaccurate thermal radiation field calculation caused by fixed fire source combinations and start-up times in existing domino effect simulation methods for storage tank fires.

[0008] In a first aspect, the present invention provides a method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates, comprising the following steps: S1: Obtain basic information about the target tank area. The basic information includes at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information. S2: Establish an FDS numerical simulation model based on basic information, set an initial fire source in the model, and arrange thermal radiation monitoring points on the candidate target tanks. Run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank.

[0009] S3: Based on the thermal radiation response data and the preset accident propagation criteria, determine whether each candidate target tank has reached the failure condition. If it has, record the failure time of the candidate target tank. S4: Select the candidate target tank that has reached the failure condition as a new fire source, and configure the fire source activation parameters of the new fire source in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire source. S5: Based on the new fire scenario, repeat S2 to S4 until no candidate target tanks meet the failure conditions. S6: Outputs fire scene data and tank failure event data for each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

[0010] As an alternative implementation method, thermal radiation monitoring points are arranged on the fire-facing side of the tank wall of the candidate target tank, the hot spot area of ​​the tank wall, and the location where hot spots may migrate. Each time S2 is repeated, based on the flame pattern, smoke plume flow direction and thermal radiation response of each thermal radiation monitoring point under the current fire scenario, multiple spatially adjacent high thermal radiation monitoring points are reselected from the preset thermal radiation monitoring point array as a hot spot monitoring point group.

[0011] As an alternative implementation method, the thermal radiation response data consists of time history data of thermal radiation intensity at multiple measuring points on the tank wall of each candidate target storage tank under the current fire scenario. Before determining whether the failure condition has been met based on the thermal radiation response data, the time history data of thermal radiation intensity at each measuring point is preprocessed. The preprocessing includes removing transient data during the fire ignition stage, fire source switching stage, and flame fluctuation stage, and extracting representative thermal radiation intensity based on the quasi-steady-state data of the remaining time period. The representative thermal radiation intensity is used as the input for the accident expansion criterion.

[0012] As an alternative implementation, the accident propagation criterion is any one or a combination of the following: thermal radiation intensity threshold criterion, cumulative heat dose threshold criterion, and time to failure model criterion. When using the cumulative heat dose threshold criterion, the cumulative heat dose of the candidate target storage tank is calculated by piecewise integration to adapt to the fluctuations in heat radiation intensity caused by the appearance of new fire source storage tanks.

[0013] As an alternative implementation method, the ignition parameters of the new ignition source are configured through the time-varying function built into FDS. The time-varying function is used to control the new ignition source to remain in an extinguished state before the failure time and to switch to a combustion state when the failure time is reached.

[0014] As an optional implementation method, the full-process simulation results include the combination of fire sources at each stage of the fire, thermal radiation response data of candidate target tanks, representative thermal radiation intensity, cumulative heat dose, tank failure time, and fire domino effect propagation path; it also includes data on the changes in the growth rate of representative thermal radiation intensity and heat dose of candidate target tanks before and after the addition of new fire sources. The change data is used to characterize the effects of new fire sources on candidate target tanks, such as promoting heating, smoke plume obscuring, flame deflection, or hot spot migration.

[0015] As an alternative implementation, the method further includes a probabilistic post-processing step: calculating the accident escalation probability of each candidate target tank using a probabilistic unit model based on the failure time of each candidate target tank in the tank failure event data.

[0016] As an alternative implementation method, a Bayesian network is constructed based on the fire domino effect propagation path diagram, and the accident escalation probability is used as the conditional probability of the corresponding node in the Bayesian network to calculate the propagation probability of the entire domino chain and identify key expansion nodes.

[0017] Secondly, the present invention provides a chemical fire domino effect simulation system based on FDS dynamic fire source updating, comprising: The data acquisition module is configured to acquire basic information about the target tank area, including at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information. The model building and running module is configured to: build an FDS numerical simulation model based on basic information, set an initial fire source in the model, arrange thermal radiation monitoring points on candidate target tanks, and run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank; The data processing and judgment module is configured to: determine whether each candidate target tank has reached the failure condition based on the thermal radiation response data and the preset accident extension criteria; if it has, record the failure time of the candidate target tank. The dynamic update module is configured to: take candidate target tanks that have reached the failure condition as new fire sources, and configure the fire source activation parameters of the new fire sources in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire sources. The repeated execution module is configured to: repeatedly execute the model building and running module to the dynamic update module based on the new fire scenario, until no candidate target tank reaches the failure condition; The output module is configured to output fire scene data and tank failure event data at each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

[0018] Thirdly, the present invention provides an electronic device including a memory and a processor, and computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the method described in the first aspect.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention proposes a method and system for simulating the domino effect in chemical fires based on FDS dynamic ignition source updates. It abandons the traditional static simulation method of "simultaneous ignition" and, by dynamically updating the ignition source, ensures that subsequent storage tanks only begin to burn at their actual failure moment. This fully and accurately reproduces the physical sequence of the accident's gradual expansion over time, providing a data foundation that conforms to physical laws for accident analysis and emergency decision-making.

[0020] Furthermore, this invention dynamically couples the added fire source as part of the model, enabling the FDS solver to naturally calculate the interactions between multiple fire sources, such as flame entrainment, plume obstruction, and flame deflection. This calculation method based on first principles (fluid dynamics and radiative heat transfer) can more accurately predict the actual thermal radiation load of the target storage tank in dynamic scenarios compared to traditional empirical formulas or algebraic superposition methods.

[0021] Furthermore, this invention uses piecewise integration or dynamic thresholding to evaluate tank failure, fully considering the heating history (cumulative effect of heat dose) of the target tank under variable thermal radiation fields, avoiding misjudgments caused by sudden changes or overestimation of thermal radiation in static simulation, and improving the reliability and scientific nature of accident expansion prediction.

[0022] Furthermore, this invention not only outputs accurate physical simulation results, but also integrates a probabilistic post-processing module, which can quantitatively assess the probability of the domino effect occurring, providing strong technical support for quantitative risk assessment, fire separation distance design, and optimal allocation of emergency rescue resources in chemical industrial parks. Attached Figure Description

[0023] Figure 1 This is an overall flowchart of the chemical fire domino effect simulation method based on FDS dynamic fire source update disclosed in the embodiments of the present invention; Figure 2 This is a schematic diagram of the tank area layout and the FDS numerical simulation model of the initial fire domino effect disclosed in the embodiments of the present invention, wherein (a) is a schematic diagram of the tank area layout and (b) is a schematic diagram of the FDS numerical simulation model of the initial fire domino effect. Figure 3 This is a schematic diagram of the domino effect expansion of a storage tank fire with dynamic fire source updates disclosed in an embodiment of the present invention; Figure 4 This is a schematic diagram of the extended chain Bayesian network structure and probabilistic reasoning of the domino effect in a storage tank fire disclosed in an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0025] The technical solutions disclosed in the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0026] Example 1 like Figures 1 to 4 As shown, this embodiment provides a method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates, including the following steps: S1: Obtain basic information about the target tank area. The basic information includes at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information. S2: Establish an FDS numerical simulation model based on basic information, set an initial fire source in the model, and arrange thermal radiation monitoring points on the candidate target tanks. Run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank. S3: Based on the thermal radiation response data and the preset accident propagation criteria, determine whether each candidate target tank has reached the failure condition. If it has, record the failure time of the candidate target tank. S4: Select the candidate target tank that has reached the failure condition as a new fire source, and configure the fire source activation parameters of the new fire source in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire source. S5: Based on the new fire scenario, repeat S2 to S4 until no candidate target tanks meet the failure conditions. S6: Outputs fire scene data and tank failure event data for each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

[0027] This invention discloses a domino effect simulation method for chemical fires based on FDS dynamic ignition source updates. The method primarily uses FDS fire numerical simulation, obtaining the thermal radiation response of candidate target tanks at different fire stages through FDS. Based on representative thermal radiation input and accident propagation criteria, the propagation time of candidate target tanks is determined. Tanks meeting the propagation conditions are dynamically updated into the FDS model as new ignition sources, allowing the ignition source combination, ignition time, and fire scenario to change progressively with the accident evolution process. After obtaining the full-process FDS simulation results, a Probit model and Bayesian network can be used for probabilistic post-processing of the propagation time, accident escalation probability, and domino propagation path. This probabilistic post-processing does not alter the FDS dynamic ignition source update process but is used for probabilistic interpretation of the simulation results and critical path identification. The specific steps of this invention are as follows: S1: Obtain basic information about the target tank farm.

[0028] This step is part of the simulation preparation, designed to provide complete data input for subsequent modeling. The basic information specifically includes: Tank layout information includes the tank number, three-dimensional coordinates of the center point, tank spacing, arrangement, and geometric parameters of ancillary facilities such as dikes. In this embodiment, a tank area consisting of six atmospheric pressure vertical tanks is used. The six tanks are arranged in two rows and three columns. The top row consists of tanks 1 (T1), 2 (T2), and 3 (T3), and the bottom row consists of tanks 4 (T4), 5 (T5), and 6 (T6). According to... Figure 2 For the layout shown, enter the coordinates from T1 to T6.

[0029] Tank structural parameters include the diameter, height, wall thickness, volume, and material of each tank. In this embodiment, all tanks have a diameter of 20 meters, a height of 15 meters, and a volume of 4712 cubic meters.

[0030] Storage medium parameters include the name, density, boiling point, flash point, heat of combustion, and mass loss rate of the medium. In this embodiment, tanks 1, 2, and 3 store benzene, while tanks 4, 5, and 6 store ethanol, with a fill rate of 70%. The heat of combustion of benzene is taken as 40,200 kJ / kg, and that of ethanol as 29,700 kJ / kg. The mass loss rate can be found in the data provided by Babrauskas in the literature "Estimating large pool fire burning rates".

[0031] Meteorological conditions: including ambient temperature, ambient pressure, wind speed, and wind direction. In this embodiment, the ambient temperature is set at 25°C and the northwest wind speed is 3 m / s.

[0032] Initial fire scenario: Specifies the storage tank where the fire initially ignites and its fire type. In this embodiment, a full-surface fire is specified in storage tank T1, and the remaining storage tanks are designated as candidate target tanks.

[0033] The above basic information is used to determine the geometric parameters, ignition source parameters, and boundary conditions of the FDS model.

[0034] S2: Establish and run the initial FDS fire numerical simulation model.

[0035] Based on the basic information obtained from S1, an initial fire numerical simulation model is established in the FDS software. This is the starting point of the recursive simulation process of this invention.

[0036] Specifically, the establishment of the model includes: Define a computational domain and ensure that its boundaries are far enough away from the fire source to eliminate the influence of boundary conditions on the simulation results.

[0037] Mesh generation is performed. Before the formal simulation, mesh sensitivity analysis and validity verification are necessary. The mesh size is usually selected based on the characteristic diameter of the flame. Mesh sensitivity analysis can be conducted by comprehensively considering changes in the mesh size and comparing the changes in flame morphology, heat release rate, temperature field, and thermal radiation intensity under different mesh conditions. Validity verification can be performed by comparing the simulated flame height, heat release rate, or thermal radiation intensity with theoretical models, empirical formulas, experimental data, or results from existing literature to determine whether the model parameters meet the computational accuracy requirements. It is crucial to ensure that the selected mesh size captures the key flame structure while controlling computational costs. In this embodiment, the initial mesh size is set to 1 meter.

[0038] Establish a geometric model of the storage tank and define the tank wall and internal liquid level.

[0039] The initial ignition source is set up by using T1 as the initial ignition source tank, with its liquid surface area designated as the combustion surface, and according to the formula... Calculate its heat release rate. Among them, The combustion efficiency of the fire is taken as 0.7. The mass combustion rate of fuel; Heat of combustion; To determine the combustion area, the remaining storage tanks are not initially set as ignition sources, but rather as candidate target storage tanks to receive thermal radiation from the ignition source storage tanks.

[0040] Set environmental conditions and wind field boundaries. Specifically, in the MISC entry (a named list entry for setting global calculation parameters) of the FDS input file, set the ambient temperature parameter TMPA and the background pressure parameter P_INF. Set TMPA to the ambient temperature obtained by S1 and set P_INF to the ambient pressure of the simulated operating condition location. In this embodiment, TMPA is set to 25.0℃ and P_INF is set to 101325Pa. Establish the correspondence between the FDS model coordinate axes and the actual geographical location, where the positive x-axis corresponds to east and the positive y-axis corresponds to north. In the WIND entry (a named list entry for setting wind field parameters such as wind speed and wind direction), set the wind speed parameter SPEED and the wind direction parameter DIRECTION. In this embodiment, SPEED is set to 3.0m / s and DIRECTION is set to 315° to form a wind field blowing from northwest to southeast.

[0041] For the computational domain boundaries, VENT entries (named list entries used to apply boundary conditions to specified surfaces or computational domain boundaries) are set on the four lateral outer boundaries and the top outer boundary, and their surface type SURF_ID is set to OPEN (open boundary). Taking a single-grid computational domain as an example, the MB (Mesh Boundary) is specified as XMIN, XMAX, YMIN, YMAX, and ZMAX, where XMIN, XMAX, YMIN, YMAX, and ZMAX represent the minimum x-coordinate boundary, maximum x-coordinate boundary, minimum y-coordinate boundary, maximum y-coordinate boundary, and maximum z-coordinate boundary, respectively. The SURF_ID of the above boundaries is set to OPEN, allowing air, smoke, and fire plumes to flow into or out of the computational domain according to the local flow state at the boundary. The MB is specified as ZMIN, which means that the bottom boundary corresponding to the minimum z-coordinate boundary is set as the solid boundary of the tank area ground, and is not set to OPEN. The above ambient temperature, background pressure, wind speed, wind direction, and outer boundary conditions remain consistent in the initial fire scenario and in the fire scenarios after each dynamic fire source update.

[0042] Thermal radiation monitoring points should be deployed on candidate target tanks T2-T6. These monitoring points should not be located at a single point on the fire-facing side of the tanks, but rather in "critical heat-affected areas" that may be subjected to high-temperature radiation. For example, in the initial stage, thermal radiation mainly acts on the side of tanks T2 and T4 facing T1; therefore, monitoring points should be densely deployed in the upper part of the fire-facing tank walls of T2 and T4 (the area with the most concentrated thermal radiation), such as... Figure 2 As shown. Simultaneously, basic measuring points should also be set up at T3, T5, and T6 to monitor potential thermal radiation levels.

[0043] After completing the above settings, submit and run the initial fire scenario FDS simulation. The simulation will continue until the preset total simulation duration is reached or no new physical changes occur. The FDS simulation output includes time history data of thermal radiation measurement points, flame morphology, plume diffusion direction, flame tilt, distribution of heated areas in the target tank, and images of the fire scene at key moments. These output results are used to determine the representative thermal radiation input and accident propagation time of the candidate target tank.

[0044] S3: Extract thermal radiation response data of candidate target storage tanks.

[0045] During the FDS simulation, data from each thermal radiation monitoring point is output at preset time intervals (e.g., every 1 second or 5 seconds). In this embodiment, we focus on extracting data from the output data to assess tank failure after the simulation has progressed to a certain stage. Specifically, in the current simulation stage, the data on the change of thermal radiation intensity over time at each measuring point on all candidate target tanks (i.e., tanks that have not yet caught fire) is read from the FDS output data.

[0046] One of the key aspects of this invention lies in addressing the hotspot migration problem. When multiple fire sources occur, the highest thermal radiation area on the candidate target tank wall may shift. Therefore, in each round of judgment, it is not advisable to use a fixed single measuring point as the sole criterion. When extracting data, several spatially adjacent measuring points with the highest thermal radiation intensity on the tank surface (e.g., the first four measuring points with the highest values) should be identified as the hotspot measuring point group. Simultaneously, to eliminate violent fluctuations or occasional turbulent pulsations during the initial stage of flame ignition, the average value within a quasi-steady-state period (e.g., the last 60 seconds of the current stage) is typically selected as the thermal radiation intensity value used for failure judgment in this stage. To exclude transient data from the fire source ignition stage, fire source switching stage, and flame fluctuation stage, a sliding time window is used to perform stability analysis on the total heat release rate time history data of the current stage, and the period that continuously meets the preset stability conditions is determined as the quasi-steady-state period. According to the preset time window length and sliding step size, the total heat release rate time history data output by the FDS is divided into multiple sliding time windows, and the coefficient of variation of the total heat release rate within each time window and the relative change rate of the average value of adjacent time windows are calculated. , ,in, and The first j The average and standard deviation of the total heat release rate within a time window, wherein the total heat release rate data is obtained from the FDS output. When the coefficient of variation and the relative rate of change are not greater than the corresponding stability threshold within a consecutive preset number of time windows, and the average total heat release rate reaches a preset combustion level, the corresponding time period is determined as a quasi-steady-state time period; the remaining time periods are regarded as the ignition, switching, or strong fluctuation stages of the fire source.

[0047] After determining the quasi-steady-state period, the results are verified based on the output of the thermal radiation monitoring points on the tank walls of the candidate target storage tanks. Multiple monitoring points with high thermal radiation intensity and spatial proximity are selected to form a hotspot monitoring group. The average thermal radiation intensity of this hotspot monitoring group during the quasi-steady-state period is used as the representative thermal radiation intensity of the current fire stage. In this embodiment, during the initial stage of T1 burning alone, the quasi-steady-state period is 80 to 200 seconds, and the representative thermal radiation intensities for T2 and T4 are extracted to be 24.06 kW / m² and 23.28 kW / m², respectively.

[0048] The coefficient of variation is the ratio of the standard deviation to the average value of the total heat release rate within the sliding time window; the relative change rate of the average value of adjacent time windows is the ratio of the absolute value of the difference between the average values ​​of two adjacent time windows to the average value of the previous time window. In this embodiment, the sliding time window length is 30 seconds, the sliding step size is 5 seconds, and both the first and second preset thresholds are 5%. A quasi-steady-state period is defined as two consecutive time windows meeting the stability condition. The above parameters can be adjusted according to the fire source size and data sampling interval.

[0049] When a new storage tank catches fire, the intensity of the fire source changes, or the FDS fire scenario is updated, the accident evolution process is divided into a new thermal exposure stage. The hotspot measurement point group and representative thermal radiation input are dynamically verified using the flame shape, flame tilt direction, smoke plume diffusion direction, and target tank wall thermal radiation distribution output by the FDS. When the flame tilt direction is towards the candidate target tank, or when the area affected by the flame and high-temperature smoke plume migrates towards the candidate target tank, the hotspot measurement point group is preferentially selected from the corresponding heated surface. When smoke plume obstruction or flame deflection causes a decrease in thermal radiation in the original hotspot area, the hotspot measurement point group is re-determined based on the current tank wall thermal radiation distribution, and the representative thermal radiation input is re-extracted.

[0050] S4: Determine whether the candidate target tank has met the failure conditions and determine the failure time.

[0051] Based on the representative thermal radiation input extracted in S3, and combined with a preset accident extension criterion, a failure determination is made for each candidate target storage tank. The criterion can take various forms: Thermal radiation threshold criterion: When the thermal radiation intensity exceeds a certain critical value (such as 15kW / m²) and continues for a certain period of time, it is judged as a failure.

[0052] Cumulative heat dose criterion: This is one of the preferred criteria of this invention. It takes into account the cumulative effect of thermal radiation. For the first... i The candidate target storage tanks, in t Cumulative heat dose at time It can be calculated using the following formula: ; in, It is the intensity of thermal radiation it receives at time τ. This is the moment when the current heat exposure phase begins, and α is the heat radiation dose index (usually taken as 1.128). When When the failure threshold is exceeded, the storage tank is deemed to have failed. Furthermore, since the thermal radiation intensity is time-varying, the above formula can be discretized as: ; in, I i,k For the first k The first stage of the fire i The representative thermal radiation intensity received by each candidate target storage tank. It is the first k Duration of each fire stage n This refers to the number of stages of a fire.

[0053] Time expires ( ttf Criteria: The dynamic failure time can be calculated based on the representative thermal radiation input received by the candidate target tank, the tank volume or structural dimensional parameters, and the tank protection status parameters. In this embodiment, an empirical or semi-empirical model, such as the failure model for an atmospheric pressure tank, is used. ; in, Q The representative thermal radiation intensity (kW / m²) received by the target storage tank. V Let be the volume of the storage tank (m³). Q =24.06 and V Substituting 4712 into the model, the failure time of tank T2 was calculated. ttf 2 is approximately 473 seconds. This means that 473 seconds after the start of the fire in T1, the T2 storage tank will fail.

[0054] During the determination process, the candidate storage tank with the earliest failure time will be identified as the new ignition source for the first round of expansion. In this embodiment, the failure time of T2 is 473 seconds, which is earlier than that of T4 (496 seconds), therefore T2 is selected as the new ignition source.

[0055] S5: Dynamically update the fire source in the FDS model.

[0056] This is the core step of the invention. Based on the determination result of S4, the storage tank that meets the failure conditions (T2 in this embodiment) is taken as a new ignition source and dynamically updated in the FDS model. The combustion surface, ignition source intensity, combustion parameters, and start-up time are dynamically updated in the FDS model. The dynamic ignition source update includes updating the combustion surface of the new ignition source, updating the ignition source intensity, updating the ignition source start-up time, updating the ignition source combination, and updating the fire scenario.

[0057] The specific operation is as follows: New ignition sources can be controlled through the ignition source time function, combustion surface activation function, or heat release rate time function in FDS, enabling them to initiate combustion at the corresponding expansion time. For example, in the FDS input file, an ignition source is defined for the liquid surface area of ​​tank T2. Its heat release rate is set to a time-dependent variable using a built-in time control function in FDS (e.g., the RAMP function). This function is set as follows: from t=0 to t=473s, the heat release rate of T2 is 0 (i.e., no combustion); at t=473s, the heat release rate linearly increases from 0 to the heat release rate corresponding to full surface combustion within a very short time (e.g., within 1 second), and then remains constant. In this way, the new ignition source is activated at the actual failure time, thereby dynamically updating the current ignition source combination from [T1] to [T1,T2].

[0058] After each new ignition source is added, the simulation results of the current fire stage are compared with those of the previous fire stage to analyze the impact of the new ignition source on the thermal exposure state of the still-functioning storage tanks. For the first... i The candidate target storage tank, the first k The representative change in thermal radiation intensity of each fire stage relative to the previous fire stage is: ,in, and These represent the representative thermal radiation intensities for the current fire stage and the previous fire stage, respectively. The change in the heat dose growth rate of the candidate target storage tank is as follows: Where α is the heat dose index, which is taken as 1.128 in this embodiment. When the changes in the representative heat radiation intensity and the heat dose growth rate are positive, it indicates that the newly added fire source generally promotes the thermal exposure of the candidate target tank; when the changes are negative, it is determined whether they are caused by smoke plume obstruction, flame deflection, or hot spot migration, based on the flame morphology, plume distribution, and tank wall heat radiation field output by the FDS. Since there are nonlinear coupling effects such as air entrainment, smoke plume obstruction, and radiation feedback among multiple fire sources, the changes are used to characterize the marginal impact after the addition of the new fire source and are not used as the linear decomposition result of the independent radiation contribution ratio of each fire source.

[0059] This step embodies the core idea of ​​the invention: new ignition sources are not pre-programmed to burn simultaneously at the initial moment, but are dynamically updated into the FDS model step by step based on the FDS thermal radiation response and accident propagation judgment results of the previous stage, so that the FDS fire scenario changes step by step with the domino chain evolution process. In specific implementation, the dynamic ignition source update can also be understood as writing back the information of new ignition sources and the ignition time to the FDS model.

[0060] S6: Recursive simulation and iterative determination.

[0061] After the fire source update is completed in S5, the updated fire scenario (i.e., T1 and T2 burning simultaneously) is used as the new initial state to continue running or restart the FDS simulation. Then, return to S3 to execute a new round of the "simulation-extraction-determination-update" cycle.

[0062] In this cycle, the focus is on tanks T3, T4, T5, and T6 that have not yet failed. At this point, the ignition sources become T1 and T2, and their coupled thermal radiation fields will act on these candidate tanks. An important phenomenon is that the newly added T2 ignition source may alter the local flow field and flue gas distribution. For example, simulation results may show that, due to the obstruction of the T2 ignition source plume, the thermal radiation values ​​at some measuring points of tank T3 are actually lower than when T1 is burning alone. In this invention, after the ignition source combination is updated, the hot spot measuring point group for the current stage is re-identified from the preset thermal radiation monitoring point array, and the heat dose increment for this stage is calculated using the new representative thermal radiation intensity; the cumulative heat dose already formed in the previous stage for the candidate target tanks is retained and is not reset to zero due to changes in hot spot location or a decrease in thermal radiation in the current stage. Therefore, a decrease in thermal radiation only indicates a decrease in the current stage's heat dose growth rate, thus ensuring the continuity and comparability of accident expansion determination at each stage. This result also indicates that multi-ignition source thermal radiation should not be processed using a static algebraic superposition method.

[0063] In this embodiment, during the second stage of simulation, the failure time of T4 is determined to be approximately 497 seconds, earlier than T5 and T3. Therefore, at 497 seconds, T4 is added to the model as the second new fire source, the fire source combination is updated to [T1, T2, T4], and the next round of simulation begins.

[0064] Repeat the above process. Under the operating conditions given in this embodiment, the entire domino expansion path is as follows: 1(0s)→2(473s)→4(497s)→5(873s)→3(978s)→6(1089s); In each iteration, the "critical heating areas" of each candidate tank need to be re-identified and the monitoring points updated based on the new combination of ignition sources and flow fields. For example, when ignition source T4 appears, the fire-facing side of tank T5 becomes the side facing T4 and T2, so the monitoring points in this area need to be rearranged or adjusted instead of using the monitoring points from the initial stage.

[0065] S7: Output the simulation results of the entire process.

[0066] When a simulation round is completed and no new candidate tanks are found to have met the failure criteria, the entire recursive process ends. At this point, all simulation results are output.

[0067] The output results include, but are not limited to: Time series data: heat release rate time history of all storage tanks (including initial ignition sources and new ignition sources), and heat radiation intensity time history data of key measuring points of all candidate storage tanks.

[0068] Event List: The failure (i.e., fire) times of each storage tank form a domino-extended event list.

[0069] Dynamic visualization results: Flame morphology, temperature field, and thermal radiation field distribution maps at each stage can be synthesized into a dynamic video demonstrating the fire spread process. For example... Figure 3 The diagram illustrates the expansion process from a single tank fire (T1) to a three-tank fire (T1, T2, and T4).

[0070] Extended path diagram: Displays the extended path of an accident in the form of a directed graph, clearly indicating the direction and time of the accident's propagation from the initial tank to other tanks.

[0071] S8 (optional): Probability post-processing steps.

[0072] To further enhance the application value of this invention, probabilistic post-processing can be performed after obtaining deterministic physical simulation results.

[0073] In this embodiment, the failure times of each storage tank obtained in S7 (such as 473 seconds for T2, 497 seconds for T4, etc.) are used as input and substituted into the classic Probit model to calculate the conditional probability of failure for each storage tank. For example, for the first... i The failure probability of a storage tank. Through Probit variables calculate: ; ; in, Let be the Probit variable for the i-th target storage tank. The input variables characterizing the effects of heat exposure are a and b, which are model parameters, and Φ is the cumulative distribution function of the standard normal distribution.

[0074] Then, as Figure 4 As shown, a Bayesian network is constructed based on the expansion path (T1→T2→T4→T5→T3→T6) determined by the physical simulation. The conditional probability table for each tank node (e.g., T2) is calculated by its parent node (e.g., T1) using a Probit model. Through the inference capabilities of the Bayesian network, we can: 1. Calculate the chain propagation probability: Calculate the joint probability of the entire domino path (e.g., T1→T2→T4) occurring.

[0075] 2. Posterior probabilistic inference: Given some observed evidence (e.g., it is known that T2 has caught fire), update the failure probability of downstream tanks (e.g., T4 and T5).

[0076] 3. Identify key nodes: Through sensitivity analysis, identify the tank nodes that have the greatest impact on the probability of the entire domino chain propagation, providing decision support for developing targeted fire prevention measures.

[0077] Example 2 This embodiment provides a chemical fire domino effect simulation system based on FDS dynamic fire source updates, including: The data acquisition module is configured to acquire basic information about the target tank area, including at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information. The model building and running module is configured to: build an FDS numerical simulation model based on basic information, set an initial fire source in the model, arrange thermal radiation monitoring points on candidate target tanks, and run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank; The data processing and judgment module is configured to: determine whether each candidate target tank has reached the failure condition based on the thermal radiation response data and the preset accident extension criteria; if it has, record the failure time of the candidate target tank. The dynamic update module is configured to: take candidate target tanks that have reached the failure condition as new fire sources, and configure the fire source activation parameters of the new fire sources in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire sources. The repeated execution module is configured to: repeatedly execute the model building and running module to the dynamic update module based on the new fire scenario, until no candidate target tank reaches the failure condition; The output module is configured to output fire scene data and tank failure event data at each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

[0078] It should be noted that the above modules correspond to the steps in Embodiment 1, and the examples and application scenarios implemented by the above modules and their corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1. It should also be noted that the above modules can be executed in a computer system as part of the system.

[0079] In further embodiments, the following is also provided: An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the method described in Embodiment 1. For brevity, further details are omitted here.

[0080] It should be understood that in this embodiment, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0081] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for simulating the domino effect in chemical fires based on FDS dynamic fire source updates, characterized in that, Includes the following steps: S1: Obtain basic information about the target tank area, including at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information; S2: Based on the aforementioned basic information, establish an FDS numerical simulation model, set an initial fire source in the model, and arrange thermal radiation monitoring points on the candidate target tanks. Run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank. S3: Based on the thermal radiation response data and the preset accident expansion criteria, determine whether each candidate target tank has reached the failure condition. If it has, record the failure time of the candidate target tank. S4: The candidate target storage tank that has reached the failure condition is taken as a new fire source. The fire source activation parameters of the new fire source are configured in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire source. S5: Based on the new fire scenario, repeat S2 to S4 until no candidate target tank reaches the failure condition. S6: Outputs fire scene data and tank failure event data for each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

2. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 1, characterized in that, The thermal radiation monitoring points are arranged on the fire-facing side of the tank wall of the candidate target tank, the hot spot area of ​​the tank wall, and the location where hot spots may migrate. Each time S2 is repeated, based on the flame shape, smoke plume flow direction and thermal radiation response of each thermal radiation monitoring point under the current fire scenario, multiple spatially adjacent high thermal radiation monitoring points are reselected from the preset thermal radiation monitoring point array as a hot spot monitoring point group.

3. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 1, characterized in that, The thermal radiation response data refers to the time history data of thermal radiation intensity at multiple measuring points on the tank wall of each candidate target storage tank under the current fire scenario. Before determining whether the failure condition has been met based on the thermal radiation response data, the time history data of thermal radiation intensity at each measuring point is preprocessed. The preprocessing includes removing transient data during the fire ignition stage, fire source switching stage, and flame fluctuation stage, and extracting representative thermal radiation intensity based on the quasi-steady-state data of the remaining time period. The representative thermal radiation intensity is used as the input of the accident expansion criterion.

4. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 1, characterized in that, The accident propagation criterion is any one or a combination of the following: thermal radiation intensity threshold criterion, cumulative heat dose threshold criterion, and time-to-failure model criterion; When the cumulative heat dose threshold criterion is used, the cumulative heat dose of the candidate target storage tank is calculated by piecewise integration to adapt to the fluctuations in heat radiation intensity caused by the appearance of the new fire source storage tank.

5. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 1, characterized in that, The ignition parameters of the newly added ignition source are configured through a time-varying function built into FDS. The time-varying function is used to control the newly added ignition source to remain in an extinguished state before the failure time and switch to a combustion state when the failure time is reached.

6. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 1, characterized in that, The full-process simulation results include the fire source combination at each stage of the fire, the thermal radiation response data of the candidate target tank, the representative thermal radiation intensity, the cumulative heat dose, the tank failure time, and the fire domino effect propagation path; it also includes the change data of the growth rate of the representative thermal radiation intensity and heat dose of the candidate target tank before and after the addition of a new fire source. The change data is used to characterize the effects of the new fire source on the candidate target tank, such as promoting heating, smoke plume obscuring, flame deflection, or hot spot migration.

7. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 6, characterized in that, The method further includes a probabilistic post-processing step: based on the failure time of each candidate target tank in the tank failure event data, the probability of accident escalation for each candidate target tank is calculated using a probabilistic unit model.

8. The method for simulating the domino effect of chemical fires based on FDS dynamic fire source updates as described in claim 7, characterized in that, A Bayesian network is constructed based on the fire domino effect propagation path diagram, and the accident escalation probability is used as the conditional probability of the corresponding node in the Bayesian network to calculate the propagation probability of the entire domino chain and identify key expansion nodes.

9. A chemical fire domino effect simulation system based on FDS dynamic fire source updates, characterized in that, include: The data acquisition module is configured to acquire basic information about the target tank area, including at least tank layout parameters, tank structural parameters, storage medium physical property parameters, and initial fire scenario information. The model building and running module is configured to: establish an FDS numerical simulation model based on the basic information, set an initial fire source in the model, arrange thermal radiation monitoring points on the candidate target tanks, and run the FDS simulation under the current fire scenario to obtain thermal radiation response data of each candidate target tank. The data processing and judgment module is configured to: determine whether each candidate target tank has reached the failure condition based on the thermal radiation response data and the preset accident extension criteria; if it has, record the failure time of the candidate target tank. The dynamic update module is configured to: take the candidate target storage tank that has reached the failure condition as a new fire source, and configure the fire source activation parameters of the new fire source in the FDS numerical simulation model according to the failure time, so as to update the current fire scenario to a new fire scenario that includes the new fire source. The repeated execution module is configured to: repeatedly execute the model construction and running module to the dynamic update module based on the new fire scenario, until no candidate target tank reaches the failure condition; The output module is configured to output fire scene data and tank failure event data at each stage of the entire simulation process, forming the simulation results of the entire process of the domino effect extension of the tank fire.

10. An electronic device, characterized in that, It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the method according to any one of claims 1-8.

Images