Alpha-olefin fischer-tropsch synthesis device safety early warning and emergency treatment system

Abstract

The application discloses an alpha-olefin Fischer-Tropsch synthesis device safety early warning and emergency treatment system, which comprises a data acquisition layer, a safety boundary evaluation layer, an intelligent early warning layer and an emergency treatment layer. The data acquisition layer is used for collecting multi-parameter data of a reactor, a separation tower and a compressor in real time. The safety boundary evaluation layer is used for calculating quantitative safety margins of each key equipment according to the multi-parameter data. The intelligent early warning layer is used for performing multi-parameter correlation analysis on the multiple quantitative safety margins, calculating a comprehensive risk index and triggering a graded early warning signal. The emergency treatment layer comprises an emergency scheme library and an automatic disposal unit, and is used for automatically matching and executing a standardized disposal process when the highest grade early warning is triggered. A safety interlocking system is independently arranged and used for executing an emergency shutdown operation when a parameter reaches an interlocking threshold. The application realizes the transformation from passive response to active defense by combining quantitative safety margin evaluation, multi-parameter correlation analysis, graded early warning, automatic emergency disposal and independent safety interlocking, and improves the safety operation level of the alpha-olefin Fischer-Tropsch synthesis device.

Description

Technical Field

[0001] This invention belongs to the field of Fischer-Tropsch synthesis unit management technology, specifically relating to a safety early warning and emergency response system for α-olefin Fischer-Tropsch synthesis units. Background Technology

[0002] Fischer-Tropsch synthesis is a chemical process that uses syngas (mainly composed of carbon monoxide and hydrogen) as raw material to convert it into hydrocarbon compounds under the action of a catalyst. It is the core technology of coal indirect liquefaction and natural gas to petroleum production. α-Olefins, as an important high-value-added product of Fischer-Tropsch synthesis, have broad application prospects in plastics, lubricants, surfactants, and other fields. α-Olefin Fischer-Tropsch synthesis units typically include core equipment such as fixed-bed or slurry-bed reactors, product separation towers, and circulating gas compressors, making them a typical complex chemical production system. Because the Fischer-Tropsch synthesis reaction is a strongly exothermic reaction, insufficient heat removal can easily lead to bed "runaway" phenomena, resulting in catalyst sintering and deactivation, or even reactor overheating and overpressure explosions. Simultaneously, abnormal operating conditions such as separation tower flooding, compressor mechanical failures, and feed interruptions also seriously threaten the safe operation of the unit.

[0003] Currently, the safety control of α-olefin Fischer-Tropsch synthesis units mainly adopts a technical solution combining distributed control systems (DCS) and instrumented safety systems (SIS). Chinese patent CN110681328A discloses an ESD interlock control method and control system for a Fischer-Tropsch reactor. This system achieves graded temperature interlock protection by setting a first temperature threshold to control the venting of purge gas and a second temperature threshold to control the feed cutoff and replenishment of inert gas. Another existing technology connects a nitrogen storage tank to the inlet pipeline of the Fischer-Tropsch synthesis reactor, introducing nitrogen and venting tail gas to cool the reactor during runaway temperatures. Furthermore, some newly built units have introduced advanced process control systems (APC), which implement multi-variable automatic adjustment control based on DCS. When the controlled variable reaches a set threshold, an audible and visual alarm is triggered, and abnormal information is pushed out. Simultaneously, Fischer-Tropsch synthesis units are typically equipped with instrumented safety systems that execute emergency shutdown operations when the reaction temperature or pressure reaches a preset hard threshold.

[0004] However, the existing technologies still have the following shortcomings: First, the safety judgment relies on a single criterion, mainly depending on whether individual parameters such as temperature and pressure exceed fixed thresholds to trigger alarms or interlocks. They lack comprehensive correlation analysis of multiple parameters such as the rate of change of reaction temperature, the rate of change of pressure, and the flow rate of cooling media, making it difficult to accurately reflect the overall risk level of the device. Second, the early warning mechanism is passive, triggering alarms only when parameters reach thresholds, which is a reactive response mode and cannot provide proactive, tiered early warnings based on the decay trend of safety margins before parameters enter dangerous areas. Third, there is a lack of an automated emergency response layer between early warning and interlock shutdown. Alarm information in existing systems requires operator judgment before manual operation or waiting for interlock action, creating a "handling window" between early warning triggering and hard interlock action, making it difficult to promptly curb the evolution of risks. Fourth, there is a lack of clear independence between the safety interlock system and the process control system. When the process control system fails, the safety interlock function may be lost simultaneously, failing to guarantee the reliability of emergency shutdown. To address the aforementioned issues, there is an urgent need for a safety early warning and emergency response system for α-olefin Fischer-Tropsch synthesis units that can combine quantitative assessment of safety margins, multi-parameter correlation analysis, tiered early warning, automatic emergency response, and independent safety interlocks. Summary of the Invention

[0005] In view of this, the present invention proposes a safety early warning and emergency response system for α-olefin Fischer-Tropsch synthesis apparatus.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a safety early warning and emergency response system for α-olefin Fischer-Tropsch synthesis units, comprising: The data acquisition layer is used to collect multi-parameter data of key equipment in the α-olefin Fischer-Tropsch synthesis unit in real time, including reactors, separation towers and compressors. The multi-parameter data includes at least temperature, pressure, pressure difference, vibration and flow rate. The safety boundary assessment layer communicates with the data acquisition layer and is used to calculate the quantitative safety margin of each key device's current operating condition relative to the preset risk boundary based on multi-parameter data. The safety margin is expressed as a percentage to represent the degree to which the actual parameters deviate from the safety limit. The intelligent early warning layer communicates with the safety boundary assessment layer and is used to perform multi-parameter correlation analysis on multiple quantitative safety margins. The multi-parameter correlation analysis includes the combined effects of temperature change rate, pressure change rate and cooling medium flow rate. Based on this, a comprehensive risk index reflecting the overall risk level of the α-olefin Fischer-Tropsch synthesis unit is calculated, and a graded early warning signal containing at least three levels is triggered according to the degree of decrease of the comprehensive risk index or any quantitative safety margin. The emergency response layer is connected to the intelligent early warning layer. It includes an emergency plan library and an automatic response unit. The emergency plan library contains standardized response procedures that correspond one-to-one with various preset abnormal working conditions. When the intelligent early warning layer triggers the highest level early warning signal, the automatic response unit automatically matches the corresponding standardized response procedure according to the current working condition and automatically executes some or all of the emergency operations in the standardized response procedure within a preset time. The safety interlocking system is independent of the basic control system and is set up relatively independently from the data acquisition layer and emergency processing layer. It has a preset safety integrity level and is used to independently execute emergency shutdown operations when the operating parameters of critical equipment reach the preset interlocking threshold.

[0007] Preferably, in the multi-parameter correlation analysis, the comprehensive risk index is generated based on a comprehensive judgment of at least the following three dimensions: temperature change rate dimension, when the temperature rise rate exceeds a first preset rate threshold, it is judged as a high-risk contribution; pressure change rate dimension, when the pressure rise rate exceeds a second preset rate threshold, it is judged as a high-risk contribution; cooling capacity dimension, when the cooling medium flow rate is lower than a preset flow rate threshold or the cooling medium temperature is higher than a preset temperature threshold, it is judged as a high-risk contribution; the comprehensive risk index is the weighted comprehensive result of the risk contributions of the above three dimensions, the weight of each dimension is determined by the analytic hierarchy process based on expert experience, and each preset threshold is determined based on the design parameters of the α-olefin Fischer-Tropsch synthesis unit, catalyst characteristics, and historical operating data.

[0008] Preferably, in the safety boundary assessment layer, the quantified safety margin of the reactor is calculated as a percentage of the actual reactor temperature relative to the reactor runaway temperature boundary, which is dynamically determined based on the type of catalyst and reaction conditions suitable for α-olefin production. The dynamic determination of the runaway temperature boundary is achieved by using a mechanistic model based on the kinetics and transport process of the α-olefin Fischer-Tropsch synthesis reaction, and determining the runaway temperature boundary by performing parameter sensitivity analysis on the hot spot temperature of the reactor bed; or by using a regression model or machine learning model based on historical operating data, and calculating the runaway temperature boundary in real time based on the current catalyst parameters and reaction conditions; or by combining the above two methods. When the quantified safety margin of the reactor is lower than the first safety threshold, the system determines that the reactor is in a high-risk runaway state. The quantified safety margin of the separation tower is calculated as a percentage of the actual flooding factor relative to the maximum allowable flooding factor, and the actual flooding factor is calculated in real time based on the gas-liquid phase load and fluid properties in the tower. The quantified safety margin of the compressor is calculated as a percentage of the actual vibration amplitude relative to the maximum allowable vibration amplitude.

[0009] Preferably, the graded early warning signals include three levels: yellow, orange, and red. A yellow warning corresponds to a decrease in the safety margin to a first preset range, which triggers system prompts and increases the monitoring frequency. An orange warning corresponds to a decrease in the safety margin to a second preset range, which triggers an audible and visual alarm and pushes warning information containing recommended preventive measures to the operator. A red warning corresponds to a decrease in the safety margin to a third preset range or a comprehensive risk index exceeding a preset high-risk threshold, which triggers an emergency alarm and automatically initiates emergency response operations at the emergency handling layer. The safety margin ranges corresponding to the three levels decrease sequentially and do not overlap.

[0010] Preferably, the intelligent early warning layer also includes an early warning accuracy guarantee unit. This unit trains an early warning model using a long short-term memory network based on historical operating data of the α-olefin Fischer-Tropsch synthesis unit. The early warning model is used to identify the difference between actual risk and normal parameter fluctuations during multi-parameter correlation analysis, thereby reducing the false alarm rate generated by fixed-threshold alarm logic. The training of the early warning model includes the following steps: a data preprocessing step, which imputes missing values ​​and normalizes the time-series data of historical operating parameters, and constructs time-series samples according to a preset time window; and a model construction step, which constructs a time-series prediction model based on a multi-layer long short-term memory network. The model consists of an input layer, at least two hidden layers, and an output layer. The model training step involves training the Long Short-Term Memory (LSTM) network using a training set and adjusting hyperparameters using a validation set. The early warning and discrimination step involves inputting real-time collected multi-parameter data into the trained LTM network model to obtain predicted values ​​and confidence intervals for each parameter. When the actual observed value exceeds the predicted confidence interval and the deviation exceeds a preset threshold, it is determined to be a real risk; otherwise, it is determined to be a normal parameter fluctuation. The early warning model uses an incremental learning strategy for online updates, periodically or continuously incrementally training the model with newly added labeled operating data to adapt to long-term changes in the operating conditions of the α-olefin Fischer-Tropsch synthesis unit.

[0011] Preferably, the emergency response plan library includes at least a first response plan for overheating of the α-olefin Fischer-Tropsch synthesis reactor, a second response plan for flooding of the separation tower, a third response plan for compressor failure, and a fourth response plan for feed interruption. Each response plan includes triggering conditions, automatic operation sequence, expected effect parameters, and manual confirmation steps. The automatic response unit automatically matches the corresponding option from the first to the fourth response plan according to the type of abnormal condition currently triggered, and first executes the preceding emergency operation in the automatic operation sequence, with a total response time of less than 5 seconds.

[0012] Preferably, in the first treatment scheme, when the rate of temperature rise in the α-olefin Fischer-Tropsch synthesis reactor exceeds a preset rate threshold, the automatic treatment unit sequentially performs the following emergency operations: cutting off the reactor feed, increasing the cooling medium supply to the maximum, opening the pressure relief channel, and stopping the reactor heating source; after the emergency operations are completed, the system monitors the temperature change trend within a preset time and decides whether to continue with the subsequent treatment steps based on the monitoring results.

[0013] Preferably, the emergency response layer also includes an intelligent plan push unit. When the intelligent early warning layer triggers a yellow or orange warning, the intelligent plan push unit selects the most matching emergency plan from the emergency plan library based on the currently detected abnormal operating conditions and parameter change trends, and pushes it to the operator interface. The push content includes a description of the abnormal phenomenon, recommended handling steps, key operating parameters, and precautions, so as to provide the operator with decision-making reference and operation guidance before the red warning is triggered.

[0014] Preferably, the emergency response layer also includes a human-machine collaborative response unit, which is used to push the subsequent operation steps in the standardized response process to the operator terminal after the automatic response unit has completed the emergency operation. The subsequent steps can only be executed after the operator confirms. The human-machine collaborative response unit is also used to pause the automatic execution process at key decision points and wait for the operator's manual decision input, and to record all operation data and parameter change data from the triggering of the warning to the completion of the response, for post-event traceability and analysis.

[0015] Preferably, the safety interlock system has a preset safety integrity level of SIL-3, and the interlock threshold is higher than the warning threshold corresponding to the highest level warning signal triggered by the intelligent warning layer. The interlock logic includes at least the following: interlock shutdown when the temperature of the α-olefin Fischer-Tropsch synthesis reactor exceeds the first hard threshold, interlock shutdown when the reactor pressure exceeds the second hard threshold, and interlock shutdown when the cooling medium flow rate is lower than the third hard threshold. The safety interlock system is independently configured with measuring instruments and actuators and does not depend on the normal operation of the data acquisition layer and the basic control system.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a complete closed-loop protection system from data acquisition → safety margin quantification → multi-parameter correlation early warning → automatic emergency response → independent safety interlock by setting up a data acquisition layer, a safety boundary assessment layer, an intelligent early warning layer, an emergency response layer, and a safety interlock system. This realizes the safety management upgrade of the α-olefin Fischer-Tropsch synthesis unit from "passive response" to "active defense".

[0017] 2. This invention generates a comprehensive risk index by calculating the quantitative safety margin and performing multi-parameter correlation analysis, which eliminates the limitations of traditional single-parameter threshold judgment and realizes a comprehensive assessment and graded early warning of the overall risk level of the device, significantly improving the accuracy and timeliness of early warning.

[0018] 3. By setting up an emergency response plan library and an automatic handling unit, this invention can trigger automated intervention measures before the hard interlock threshold is reached, with a total response time of less than 5 seconds. This provides valuable response time for operators and effectively reduces the probability of major accidents such as overheating and flooding.

[0019] 4. By setting up an independent SIL-3 safety interlock system, this invention forms a three-level progressive protection architecture of "early warning - response - interlock". Even in extreme cases where the early warning and response systems fail, the independent safety interlock system can still reliably perform emergency shutdown operations to ensure the safety of the device.

[0020] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0021] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following figures are provided for illustration: Figure 1 This is a schematic diagram of the overall system architecture in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the classification of graded early warning signals and their trigger response relationship in an embodiment of the present invention; Figure 3 This is a schematic diagram of the handling process from early warning triggering to safety interlock action in an embodiment of the present invention. Detailed Implementation

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0023] like Figure 1 As shown, the present invention provides a safety early warning and emergency response system for an α-olefin Fischer-Tropsch synthesis unit, comprising: The data acquisition layer is used to collect multi-parameter data of key equipment in the α-olefin Fischer-Tropsch synthesis unit in real time, including reactors, separation towers and compressors. The multi-parameter data includes at least temperature, pressure, pressure difference, vibration and flow rate. The safety boundary assessment layer communicates with the data acquisition layer and is used to calculate the quantitative safety margin of each key device's current operating condition relative to the preset risk boundary based on multi-parameter data. The safety margin is expressed as a percentage to represent the degree to which the actual parameters deviate from the safety limit. The intelligent early warning layer communicates with the safety boundary assessment layer and is used to perform multi-parameter correlation analysis on multiple quantitative safety margins. The multi-parameter correlation analysis includes the combined effects of temperature change rate, pressure change rate and cooling medium flow rate. Based on this, a comprehensive risk index reflecting the overall risk level of the α-olefin Fischer-Tropsch synthesis unit is calculated, and a graded early warning signal containing at least three levels is triggered according to the degree of decrease of the comprehensive risk index or any quantitative safety margin. The emergency response layer is connected to the intelligent early warning layer. It includes an emergency plan library and an automatic response unit. The emergency plan library contains standardized response procedures that correspond one-to-one with various preset abnormal working conditions. When the intelligent early warning layer triggers the highest level early warning signal, the automatic response unit automatically matches the corresponding standardized response procedure according to the current working condition and automatically executes some or all of the emergency operations in the standardized response procedure within a preset time. The safety interlocking system is independent of the basic control system and is set up relatively independently from the data acquisition layer and emergency processing layer. It has a preset safety integrity level and is used to independently execute emergency shutdown operations when the operating parameters of critical equipment reach the preset interlocking threshold.

[0024] In this embodiment, the α-olefin Fischer-Tropsch synthesis unit refers to a chemical production unit based on the Fischer-Tropsch synthesis reaction and aimed at producing α-olefins. Its core equipment includes a fixed-bed or slurry-bed reactor, a product separation tower, and a circulating gas compressor.

[0025] In this embodiment, multi-parameter data is the basic information for measuring the safe operating status of the device. Temperature, pressure, pressure difference, vibration, and flow rate reflect the thermodynamic state of the reaction, the mechanical health of the equipment, and the material flow state, respectively. These parameters are conventional data collection parameters in the field of safety monitoring of Fischer-Tropsch synthesis devices.

[0026] In this embodiment, the function of the safety boundary assessment layer is to transform the collected real-time parameters into a safety margin index with physical meaning. This index is expressed as a percentage, which facilitates horizontal comparison between parameters of different dimensions and subsequent comprehensive evaluation.

[0027] In this embodiment, the intelligent early warning layer fully considers the interactive influence between the rate of temperature change, the rate of pressure change, and the flow rate of cooling medium by performing coupled analysis of multiple safety margins, thus realizing a technical upgrade from single parameter threshold judgment to multi-parameter collaborative risk assessment.

[0028] In this embodiment, the emergency response layer is the system's active intervention module. Its core function is to automatically activate when the warning signal reaches the highest level, and to perform emergency operations without waiting for operator instructions, effectively shortening the emergency response time.

[0029] In this embodiment, the safety interlock system is set up independently of the basic control system, ensuring that the safety interlock system can still operate reliably and perform emergency stop operations to ensure the safety of the device in the event of failure or malfunction of the basic control system.

[0030] In a preferred embodiment, in the multi-parameter correlation analysis, the comprehensive risk index is generated based on a comprehensive judgment of at least the following three dimensions: temperature change rate dimension, when the temperature rise rate exceeds a first preset rate threshold, it is determined to be a high-risk contribution; pressure change rate dimension, when the pressure rise rate exceeds a second preset rate threshold, it is determined to be a high-risk contribution; cooling capacity dimension, when the cooling medium flow rate is lower than a preset flow rate threshold or the cooling medium temperature is higher than a preset temperature threshold, it is determined to be a high-risk contribution; the comprehensive risk index is the weighted comprehensive result of the risk contributions of the above three dimensions, the weight of each dimension is determined by the analytic hierarchy process based on expert experience, and each preset threshold is determined based on the design parameters of the α-olefin Fischer-Tropsch synthesis unit, catalyst characteristics, and historical operating data.

[0031] In this embodiment, the rate of temperature change is an important criterion for determining whether the reaction is approaching runaway temperature. In the Fischer-Tropsch synthesis, a strongly exothermic reaction, when the rate of temperature rise exceeds a certain limit, it indicates that the rate of heat generation has significantly exceeded the rate of heat removal, which is a high-risk condition.

[0032] In this embodiment, the rate of pressure change reflects the accumulation or blockage of gaseous materials in the reactor. A sudden increase in pressure often indicates that the reaction is out of control or the flow channel is blocked, which also constitutes a high-risk factor.

[0033] In this embodiment, the cooling capacity dimension reflects the device's ability to remove reaction heat. Too low a cooling medium flow rate or too high a cooling medium temperature will lead to a decrease in heat exchange capacity and make it impossible to effectively control the reaction temperature.

[0034] In this embodiment, the weighted aggregation method of the comprehensive risk index is determined by the analytic hierarchy process (AHP). This method can transform the qualitative judgment of experts into quantitative weights, ensuring that the contribution of each dimension to the final risk index matches its actual impact.

[0035] Specifically, the formula for calculating the comprehensive risk index is as follows: in, This represents the overall risk index. The risk contribution value represents the temperature change rate dimension (it is 1 when the temperature rise rate exceeds the first preset rate threshold, and 0 otherwise). This represents the risk contribution value in the pressure change rate dimension (it is 1 when the pressure rise rate exceeds the second preset rate threshold, and 0 otherwise). The risk contribution value represents the cooling capacity dimension (it is 1 when the cooling medium flow rate is lower than the preset flow rate threshold or the cooling medium temperature is higher than the preset temperature threshold, and 0 otherwise). , , These are the weight coefficients for the corresponding dimensions.

[0036] In this embodiment, each preset threshold is determined based on the device design parameters, catalyst characteristics, and historical operating data. The design parameters include the reactor heat exchange area, the maximum flow rate of the cooling medium, and the design pressure, etc. The catalyst characteristics include catalyst activity, selectivity, and thermal stability, etc.

[0037] In a preferred embodiment, in the safety boundary assessment layer, the quantified safety margin of the reactor is calculated as a percentage of the actual reactor temperature relative to the reactor runaway temperature boundary, which is dynamically determined based on the type of catalyst and reaction conditions suitable for α-olefin production. The dynamic determination of the runaway temperature boundary is achieved using a mechanistic model based on the kinetics and transport processes of the α-olefin Fischer-Tropsch synthesis reaction, through parameter sensitivity analysis of the reactor bed hotspot temperature; alternatively, a regression model or machine learning model based on historical operating data is used to calculate the runaway temperature boundary in real time based on current catalyst parameters and reaction conditions; or a combination of both methods is employed. When the quantified safety margin of the reactor is lower than a first safety threshold, the system determines that the reactor is in a high-risk runaway state. The quantified safety margin of the separation tower is calculated as a percentage of the actual flooding factor relative to the maximum permissible flooding factor, which is calculated in real time based on the gas-liquid phase load and fluid properties within the tower. The quantified safety margin of the compressor is calculated as a percentage of the actual vibration amplitude relative to the maximum permissible vibration amplitude.

[0038] In this embodiment, the fly-through temperature boundary is a key control parameter for the safe operation of the Fischer-Tropsch synthesis reactor. Its physical meaning is that when the hot spot temperature of the reactor bed exceeds this boundary value, the rate of heat generation will exceed the rate of heat removal, and the temperature will enter an uncontrollable self-accelerating rise state.

[0039] In this embodiment, the flyby boundary temperature is not a fixed value, but changes dynamically with the type of catalyst and reaction conditions. The active components, loading, pore structure, and particle size of different catalysts all affect the reaction kinetic parameters, thereby changing the flyby boundary temperature.

[0040] In this embodiment, the method for determining the fly-temperature boundary based on the mechanism model specifically adopts a one-dimensional or two-dimensional pseudo-homogeneous fixed-bed reactor mathematical model, coupled with the Fischer-Tropsch synthesis reaction kinetic equation, energy conservation equation, and mass conservation equation. By performing parameter sensitivity analysis on the inlet temperature and cooling medium temperature in the model, the critical temperature value for the system to transition from a steady state to an unstable state is determined.

[0041] Specifically, the formula for calculating the quantitative safety margin of the reactor is as follows: in, This indicates the quantitative safety margin of the reactor. Indicates the temperature at the fly-temperature boundary. This indicates the actual temperature of the reactor. When... When the temperature approaches 0, it indicates that the actual temperature is close to the runaway temperature boundary and the safety margin is insufficient.

[0042] In this embodiment, the first safety threshold is a preset safety limit value, for example, it can be set to 20%. When the quantified safety margin is lower than this threshold, it indicates that the reactor is in a high-risk state of runaway temperature and intervention measures need to be taken immediately.

[0043] In this embodiment, the flooding factor of the separation tower is an important parameter for measuring the gas-liquid processing capacity of the tower equipment. When the actual flooding factor is close to or exceeds the maximum allowable flooding factor, flooding will occur in the tower, causing a sharp drop in separation efficiency. In severe cases, it can cause drastic fluctuations in tower pressure and damage to the equipment.

[0044] Specifically, the actual flooding factor of the separation tower is calculated in real time based on the gas-liquid phase load and fluid properties within the tower. The calculation formula is as follows: in, Indicates the actual flooding factor. This indicates the actual apparent flow rate of the gas inside the tower. This represents the flooding gas velocity under the corresponding operating conditions. Those skilled in the art can calculate this velocity using conventional flooding gas velocity correlations based on tray structure and physical property parameters. .

[0045] In this embodiment, the vibration amplitude of the compressor is a key indicator reflecting the mechanical health status of the compressor. When the vibration amplitude exceeds the maximum allowable value, it indicates that there is a risk of dynamic imbalance of the rotating parts, bearing wear or seal failure.

[0046] In a preferred embodiment, refer to Figure 2The tiered early warning signals include three levels: yellow, orange, and red. A yellow warning corresponds to a decrease in the safety margin to the first preset range, triggering system prompts and increased monitoring frequency. An orange warning corresponds to a decrease in the safety margin to the second preset range, triggering audible and visual alarms and sending warning information containing recommended preventative measures to the operator. A red warning corresponds to a decrease in the safety margin to the third preset range or a comprehensive risk index exceeding a preset high-risk threshold, triggering an emergency alarm and automatically initiating emergency response operations at the emergency handling layer. The safety margin ranges corresponding to the three levels decrease sequentially and do not overlap.

[0047] In this embodiment, the three-level warning signals constitute a complete response chain from information prompts to automatic handling. The triggering conditions of each level are strictly corresponding to the severity of the risk, avoiding over-response or under-response.

[0048] In this embodiment, the yellow alert is the lowest level of alert. Its main function is to increase the operator's vigilance and attention, while also increasing the system's own monitoring frequency, such as increasing the data collection frequency from once per minute to once every 10 seconds, so as to track parameter change trends more intensively.

[0049] In this embodiment, the orange alert is a medium-level alert, indicating that the device operating conditions have deviated from the normal range and operator intervention is required. The audible and visual alarms are used to attract the operator's attention in the control room environment, and the pushed alert information includes recommended preventative measures, such as suggesting reducing the device load or adjusting the cooling medium flow rate.

[0050] In this embodiment, the red alert is the highest level of alert, indicating that the device has entered a dangerous state or the overall risk level has exceeded the acceptable limit, requiring the system to immediately perform automatic handling operations, without waiting for operator confirmation.

[0051] In a preferred embodiment, the intelligent early warning layer further includes an early warning accuracy guarantee unit. This unit trains an early warning model using a long short-term memory network based on historical operating data of the α-olefin Fischer-Tropsch synthesis unit. The early warning model is used to identify the difference between actual risk and normal parameter fluctuations during multi-parameter correlation analysis, thereby reducing the false alarm rate generated by fixed-threshold alarm logic. The training of the early warning model includes the following steps: a data preprocessing step, where missing value imputation and normalization are performed on the time series data of historical operating parameters, and time series samples are constructed according to a preset time window; and a model construction step, where a time series prediction model is constructed based on a multi-layer long short-term memory network. The Long Short-Term Memory (LSTM) network consists of an input layer, at least two hidden layers, and an output layer. The model training step involves training the LTM network using a training set and adjusting hyperparameters using a validation set. The early warning and discrimination step involves inputting real-time collected multi-parameter data into the trained LTM network model to obtain predicted values ​​and confidence intervals for each parameter. When the actual observed value exceeds the predicted confidence interval and the deviation exceeds a preset threshold, it is determined to be a real risk; otherwise, it is determined to be a normal fluctuation in parameters. The early warning model uses an incremental learning strategy for online updates, periodically or continuously incrementally training the model with newly added labeled operating data to adapt to long-term changes in the operating conditions of the α-olefin Fischer-Tropsch synthesis unit.

[0052] In this embodiment, the purpose of setting up the early warning accuracy guarantee unit is to solve the inherent technical defect of high false alarm rate in traditional fixed threshold alarm logic. In the operation of chemical plants, it is normal for parameters to fluctuate due to noise, environmental disturbances, etc., but these fluctuations do not pose an actual risk. Fixed threshold alarms often generate a large number of invalid alarms because they cannot distinguish the nature of the fluctuations.

[0053] In this embodiment, the Long Short-Term Memory Network is a variant of a recurrent neural network with a gating mechanism, specifically designed to process and predict long-range dependencies in time series data, making it suitable for data types with temporal evolution characteristics, such as the operating parameters of Fischer-Tropsch synthesis devices.

[0054] In this embodiment, in the data preprocessing step, missing value imputation can be performed using linear interpolation or neighbor-value filling methods, and normalization processing can be performed using the minimum-maximum normalization method to uniformly map each parameter to the [0,1] interval. The preset time window can be set to, for example, 5 minutes, 10 minutes or 30 minutes to construct time series samples.

[0055] In this embodiment, during the model construction step, the number of hidden layers in the multilayer long short-term memory network is determined based on the amount of training data and the complexity of the sequence, typically using 2 to 3 layers, with each layer having a corresponding number of memory units.

[0056] In this embodiment, the predicted confidence interval in the early warning judgment step provides the range within which each parameter should fall under normal operating conditions. When the actual observed value exceeds this confidence interval and the deviation exceeds a preset threshold, it indicates that the parameter change has exceeded the normal fluctuation range and is a real risk signal; conversely, if the actual observed value touches the fixed alarm threshold but is still within the predicted confidence interval, it is determined to be a normal parameter fluctuation and no early warning is triggered.

[0057] In this embodiment, the incremental learning strategy enables the early warning model to continuously evolve as the unit's operating data accumulates. During long-term operation, the α-olefin Fischer-Tropsch synthesis unit experiences operational drift, including catalyst activity decay and changes in feedstock properties. The incremental update mechanism ensures that the early warning model can automatically adapt to these changes, maintaining the accuracy of its warnings.

[0058] Specifically, when the early warning model is updated online, the formula for calculating the loss function is: in, Represents the total loss function. Indicates the number of samples. The model represents the Predicted value at time, express The actual observed value at time [time]. Indicates model parameters, This represents the regularization coefficient, used to control model complexity and prevent overfitting.

[0059] In a preferred embodiment, the emergency response plan library includes at least a first response plan corresponding to overheating of the α-olefin Fischer-Tropsch synthesis reactor, a second response plan corresponding to flooding of the separation tower, a third response plan corresponding to compressor failure, and a fourth response plan corresponding to feed interruption. Each response plan includes triggering conditions, an automatic operation sequence, expected effect parameters, and manual confirmation steps. The automatic response unit automatically matches the corresponding option among the first to fourth response plans according to the type of abnormal condition currently triggered, and first executes the preceding emergency operation in the automatic operation sequence, with a total response time of less than 5 seconds.

[0060] In this embodiment, reactor runaway temperature is one of the most serious safety risks in α-olefin Fischer-Tropsch synthesis units. When the reaction temperature gets out of control, the temperature can rise rapidly in a very short time, causing serious consequences such as catalyst sintering, product cracking, or even reactor damage.

[0061] In this embodiment, flooding in the separation tower can lead to failure of gas-liquid mass and heat transfer within the tower, resulting in simultaneous deterioration of product quality at both the top and bottom of the tower, and in severe cases, causing tower vibration and damage to internal components.

[0062] In this embodiment, a compressor failure may cause the circulating gas in the reaction system to be interrupted, which in turn may trigger a chain reaction risk within seconds that the heat of reaction cannot be carried away.

[0063] In this embodiment, feed interruption can lead to an imbalance in the material ratio within the reactor. If not handled properly, this could result in secondary risks such as catalyst reduction or oxidation, or localized overheating.

[0064] In this embodiment, each disposal plan includes complete emergency operation logic, including trigger condition determination, execution of automatic operation sequence, verification of expected effect parameters, and manual confirmation steps, to ensure the integrity and traceability of the operation.

[0065] In this embodiment, the technical effect of a total response time of less than 5 seconds is achieved by pre-setting key emergency operation instructions in the control system and directly calling and executing them when the highest level of warning is triggered. This avoids operator judgment time, instruction transmission time, and manual operation time, and significantly improves the response speed in emergency situations.

[0066] In a preferred embodiment, in the first treatment scheme, when the rate of temperature rise in the α-olefin Fischer-Tropsch synthesis reactor exceeds a preset rate threshold, the automatic treatment unit sequentially performs the following emergency operations: cutting off the reactor feed, increasing the cooling medium supply to the maximum, opening the pressure relief channel, and stopping the reactor heating source; after the emergency operations are completed, the system monitors the temperature change trend within a preset time and decides whether to continue to perform subsequent treatment steps based on the monitoring results.

[0067] In this embodiment, cutting off the reactor feed is the primary operation of the fly-through treatment, the purpose of which is to stop the continuous input of fresh reactants and reduce the amount of heat generated from the source.

[0068] In this embodiment, the cooling medium supply is increased to the maximum to maximize the heat removal capacity of the heat exchanger, quickly remove the accumulated reaction heat, and prevent the temperature from rising further.

[0069] In this embodiment, the pressure relief channel is opened to release pressure when the reactor pressure rises synchronously, preventing the equipment from being damaged due to overpressure.

[0070] In this embodiment, stopping the reactor heating source is used to prevent external heating from further aggravating the temperature rise during the reactor start-up phase or under low load conditions.

[0071] In this embodiment, after the emergency operation is completed, the system automatically enters the monitoring and verification phase. The temperature change trend within a preset time (e.g., 30 seconds to 2 minutes) is the direct basis for judging the treatment effect. If the temperature tends to decrease and the rate of decrease meets expectations, the treatment is deemed effective, and subsequent stable operation can proceed. If the temperature continues to rise, subsequent treatment steps need to be performed, such as reactor isolation or emergency discharge.

[0072] In a preferred embodiment, the emergency response layer further includes an intelligent plan push unit. When the intelligent early warning layer triggers a yellow or orange warning, the intelligent plan push unit selects the most matching emergency plan from the emergency plan library based on the currently detected abnormal operating condition type and parameter change trend, and pushes it to the operator interface. The push content includes a description of the abnormal phenomenon, recommended handling steps, key operating parameters, and precautions, so as to provide the operator with decision-making reference and operation guidance before the red warning is triggered.

[0073] In this embodiment, the function of the intelligent plan push unit is to provide operators with response references in advance before the warning level reaches the highest level, so as to achieve the effect of early warning and auxiliary decision-making.

[0074] In this embodiment, the criteria for matching degree screening include the degree of matching between the current abnormal working condition type and the triggering conditions set in the contingency plan, as well as the degree of consistency between the current parameter change trend and the expected effect parameters of the contingency plan.

[0075] In this embodiment, the description of abnormal phenomena in the push content is used to help operators quickly understand the nature of the current working condition; the recommended handling steps are used to provide standardized operation guidance; the key operation parameters are used to prompt operators to pay attention to the key control indicators; and the precautions are used to warn operators of potential operational risks.

[0076] In this embodiment, by pushing out contingency plans for the yellow and orange warning stages, the operator has a full understanding of the type of abnormal operating condition and the corresponding response methods before the red warning is triggered. When the highest level warning is triggered, the operator can cooperate with the automatic handling unit to complete the emergency response with greater confidence.

[0077] In a preferred embodiment, the emergency response layer further includes a human-machine collaborative response unit, which is used to push the subsequent operation steps in the standardized response process to the operator terminal after the automatic response unit has completed the emergency operation. The subsequent steps can only be executed after the operator confirms. The human-machine collaborative response unit is also used to pause the automatic execution process at key decision points and wait for the operator's manual decision input, and to record all operation data and parameter change data from the triggering of the warning to the completion of the response, for post-event traceability and analysis.

[0078] In this embodiment, the essence of the human-machine collaborative processing unit is to retain the space and authority for human judgment while ensuring rapid response.

[0079] In this embodiment, the automated handling unit executes the preliminary emergency operations in the standardized handling process. These operations are characterized by high time urgency, high standardization, and clear execution paths. After the emergency operation is completed, the system status has transitioned from a critical state to a controllable state, and subsequent operation steps can be confirmed and executed by the operator to achieve more flexible judgment and decision-making.

[0080] In this embodiment, the key decision point refers to the location point in the process where there are multiple optional paths or where the specific state of the device needs to be comprehensively judged to determine the subsequent steps. For example, when determining whether to discharge the reactor, it is necessary to make a comprehensive judgment based on multi-dimensional information such as the current temperature drop rate, catalyst bed temperature distribution, and pressure drop changes. At this time, pausing the automatic execution process and waiting for manual input can effectively avoid misjudgment caused by insufficient information in the automated system.

[0081] In this embodiment, the full-process recording function provides a complete data foundation for post-accident tracing and cause analysis. The recorded content includes the warning trigger time, the change sequence of each warning level, the actual execution time of automatic handling operations, parameter response data, and operator confirmation and input records. The above data is of great value for the continuous improvement of the device safety management system.

[0082] In a preferred embodiment, refer to Figure 3 The safety interlock system has a preset safety integrity level of SIL-3, and the interlock threshold is higher than the warning threshold corresponding to the highest level warning signal triggered by the intelligent warning layer. The interlock logic includes at least the following: interlock shutdown when the temperature of the α-olefin Fischer-Tropsch synthesis reactor exceeds the first hard threshold, interlock shutdown when the reactor pressure exceeds the second hard threshold, and interlock shutdown when the cooling medium flow rate is lower than the third hard threshold. The safety interlock system is independently configured with measuring instruments and actuators and does not depend on the normal operation of the data acquisition layer and the basic control system.

[0083] In this embodiment, the safety integrity level is SIL-3, which is the safety instrumented system reliability level classified according to international standards. The SIL-3 level requires that the average probability of failure of the system in the required mode be between 10 and 10. -4 Up to 10 -3 The values ​​between these represent a relatively high level of safety and reliability.

[0084] In this embodiment, the setting logic of the interlock threshold being higher than the warning threshold is as follows: the warning system issues a warning signal when the parameter deviates from the normal range but has not yet reached the dangerous level, while the safety interlock system only performs the stop operation when the parameter has entered the real dangerous area, forming a two-level protection barrier between the two.

[0085] In this embodiment, the first hard threshold, the second hard threshold, and the third hard threshold are the final protection limits for triggering interlocking shutdown. Their values ​​are determined by process safety analysis and are usually based on equipment design limits and critical conditions for runaway reaction.

[0086] In this embodiment, the safety interlock system is independently configured with measuring instruments and actuators, which means that the temperature transmitters, pressure transmitters, flow transmitters and corresponding shut-off valves used in it are all independent of the basic control system. Even if the basic control system stops, fails or communication is interrupted, the safety interlock system can still monitor the interlock conditions and execute the interlock actions normally, ensuring the reliability and independence of the safety protection function.

[0087] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. A safety early warning and emergency response system for an α-olefin Fischer-Tropsch synthesis unit, characterized in that, include: The data acquisition layer is used to collect multi-parameter data of key equipment in the α-olefin Fischer-Tropsch synthesis unit in real time, including reactors, separation towers and compressors. The multi-parameter data includes at least temperature, pressure, pressure difference, vibration and flow rate. The safety boundary assessment layer communicates with the data acquisition layer and is used to calculate the quantitative safety margin of each key device's current operating condition relative to the preset risk boundary based on multi-parameter data. The safety margin is expressed as a percentage to represent the degree to which the actual parameters deviate from the safety limit. The intelligent early warning layer communicates with the safety boundary assessment layer and is used to perform multi-parameter correlation analysis on multiple quantitative safety margins. The multi-parameter correlation analysis includes the combined effects of temperature change rate, pressure change rate and cooling medium flow rate. Based on this, a comprehensive risk index reflecting the overall risk level of the α-olefin Fischer-Tropsch synthesis unit is calculated, and a graded early warning signal containing at least three levels is triggered according to the degree of decrease of the comprehensive risk index or any quantitative safety margin. The emergency response layer is connected to the intelligent early warning layer. It includes an emergency plan library and an automatic response unit. The emergency plan library contains standardized response procedures that correspond one-to-one with various preset abnormal working conditions. When the intelligent early warning layer triggers the highest level early warning signal, the automatic response unit automatically matches the corresponding standardized response procedure according to the current working condition and automatically executes some or all of the emergency operations in the standardized response procedure within a preset time. The safety interlocking system is independent of the basic control system and is set up relatively independently from the data acquisition layer and emergency handling layer. It has a preset safety integrity level and is used to independently execute emergency shutdown operations when the operating parameters of critical equipment reach the preset interlocking threshold.

2. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, In multi-parameter correlation analysis, the comprehensive risk index is generated based on a comprehensive judgment of at least the following three dimensions: temperature change rate dimension, when the temperature rise rate exceeds the first preset rate threshold, it is judged as a high-risk contribution; In terms of the rate of pressure change, when the rate of pressure increase exceeds the second preset rate threshold, it is determined to be a high-risk contribution. In terms of cooling capacity, a high-risk contribution is identified when the cooling medium flow rate is lower than the preset flow rate threshold or the cooling medium temperature is higher than the preset temperature threshold. The comprehensive risk index is a weighted composite result of the risk contributions of the above three dimensions. The weights of each dimension are determined by the analytic hierarchy process based on expert experience, and the preset thresholds are determined based on the design parameters, catalyst characteristics and historical operating data of the α-olefin Fischer-Tropsch synthesis unit.

3. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, In the safety boundary assessment layer, the quantified safety margin of the reactor is calculated as a percentage of the actual reactor temperature relative to the reactor runaway temperature boundary. The runaway temperature boundary is dynamically determined based on the type of catalyst and reaction conditions applicable to α-olefin production. The dynamic determination of the runaway temperature boundary adopts a mechanistic model based on the kinetics and transport process of the α-olefin Fischer-Tropsch synthesis reaction, and determines the runaway temperature boundary by performing parameter sensitivity analysis on the hot spot temperature of the reactor bed; or it adopts a regression model or machine learning model based on historical operating data, and calculates the runaway temperature boundary in real time according to the current catalyst parameters and reaction conditions. When the quantified safety margin of the reactor is lower than the first safety threshold, the system determines that the reactor is in a high-risk runaway state. The quantitative safety margin of the separation tower is calculated as a percentage of the actual flooding factor relative to the maximum permissible flooding factor. The actual flooding factor is calculated in real time based on the gas-liquid phase load and fluid properties in the tower. The quantitative safety margin of the compressor is calculated as a percentage of the actual vibration amplitude relative to the maximum permissible vibration amplitude.

4. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The tiered early warning signals include three levels: yellow, orange, and red. A yellow warning corresponds to a decrease in the safety margin to the first preset range, triggering system prompts and increased monitoring frequency. An orange warning corresponds to a decrease in the safety margin to the second preset range, triggering audible and visual alarms and sending warning information containing recommended preventative measures to the operator. A red warning corresponds to a decrease in the safety margin to the third preset range or a comprehensive risk index exceeding a preset high-risk threshold, triggering an emergency alarm and automatically initiating emergency response operations at the emergency handling layer. The safety margin ranges corresponding to the three levels decrease sequentially and do not overlap.

5. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The intelligent early warning layer also includes an early warning accuracy guarantee unit. This unit trains an early warning model using a long short-term memory network based on historical operating data from the α-olefin Fischer-Tropsch synthesis unit. The early warning model is used to identify the difference between actual risk and normal parameter fluctuations during multi-parameter correlation analysis, thereby reducing the false alarm rate generated by fixed-threshold alarm logic. The training of the early warning model includes the following steps: a data preprocessing step, which imputes missing values ​​and normalizes the time-series data of historical operating parameters, and constructs time-series samples according to a preset time window; and a model construction step, which constructs a time-series prediction model based on a multi-layer long short-term memory network. The long short-term memory network contains… The model consists of an input layer, at least two hidden layers, and an output layer. The model training steps involve training the Long Short-Term Memory (LSTM) network using a training set and adjusting hyperparameters using a validation set. The early warning and discrimination steps involve inputting real-time multi-parameter data into the trained LSM model to obtain predicted values ​​and confidence intervals for each parameter. When the actual observed value exceeds the predicted confidence interval and the deviation exceeds a preset threshold, it is determined to be a real risk; otherwise, it is determined to be a normal parameter fluctuation. The early warning model uses an incremental learning strategy for online updates, periodically or continuously incrementally training the model with newly added labeled operating data to adapt to long-term changes in the operating conditions of the α-olefin Fischer-Tropsch synthesis unit.

6. The safety early warning and emergency response system for an α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The emergency response plan library includes at least a first response plan for overheating of the α-olefin Fischer-Tropsch synthesis reactor, a second response plan for flooding of the separation tower, a third response plan for compressor failure, and a fourth response plan for feed interruption. Each response plan includes triggering conditions, automatic operation sequence, expected effect parameters, and manual confirmation steps. The automatic response unit automatically matches the corresponding option from the first to the fourth response plan based on the type of abnormal condition currently triggered, and first executes the preceding emergency operation in the automatic operation sequence.

7. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 6, characterized in that, In the first treatment plan, when the temperature rise rate of the α-olefin Fischer-Tropsch synthesis reactor exceeds the preset rate threshold, the automatic treatment unit sequentially performs the following emergency operations: cut off the reactor feed, increase the cooling medium supply to the maximum, open the pressure relief channel, and stop the reactor heating source; After the emergency operation is completed, the temperature change trend is monitored within a preset time, and a decision is made on whether to continue with subsequent handling steps based on the monitoring results.

8. The safety early warning and emergency response system for the α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The emergency response layer also includes an intelligent plan push unit. When the intelligent early warning layer triggers a yellow or orange warning, the intelligent plan push unit selects the most matching emergency plan from the emergency plan library and pushes it to the operator interface based on the currently detected abnormal working condition type and parameter change trend. The push notification includes a description of the abnormal phenomenon, recommended handling steps, key operating parameters, and precautions, in order to provide operators with decision-making references and operational guidance before a red alert is triggered.

9. The safety early warning and emergency response system for an α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The emergency response layer also includes a human-machine collaborative response unit, which pushes the subsequent operation steps in the standardized response process to the operator terminal after the automatic response unit has completed the emergency operation. The subsequent steps can only be executed after the operator confirms. The human-machine collaborative response unit is also used to pause the automatic execution process at key decision points and wait for the operator's manual decision input. It also records all operation data and parameter change data from the triggering of the warning to the completion of the response, for post-event traceability and analysis.

10. The safety early warning and emergency response system for an α-olefin Fischer-Tropsch synthesis unit according to claim 1, characterized in that, The safety interlock system has a preset safety integrity level of SIL-3. The interlock threshold is higher than the warning threshold corresponding to the highest level warning signal triggered by the intelligent warning layer. The interlock logic includes at least the following: interlock shutdown when the temperature of the α-olefin Fischer-Tropsch synthesis reactor exceeds the first hard threshold, interlock shutdown when the reactor pressure exceeds the second hard threshold, and interlock shutdown when the cooling medium flow rate is lower than the third hard threshold. The safety interlock system is independently configured with measuring instruments and actuators and does not depend on the normal operation of the data acquisition layer and the basic control system.

Images