A method and system for optimizing reaction parameters for hydrogen production from natural gas

Abstract

The application discloses a kind of natural gas hydrogen production reaction parameter optimization method and system, it is related to reaction parameter optimization technical field;Including the following steps: determining core process parameters in key unit parameters, obtaining historical operation data, according to the three-dimensional structure classification storage of raw material attribute, reaction process, product index, combing the key correlation law between parameters, form parameter optimization benchmark library;According to parameter correlation law, construct parameter prediction model, generate the final parameter set of next batch reaction based on the historical operation data in benchmark parameter library;Its technical points are: according to the three-dimensional structure classification storage of raw material attribute, reaction process, product index, form parameter optimization benchmark library, and obtain the historical operation data of recent 1 year, contain different working conditions, effectively avoid the single benchmark parameter under fixed working condition in prior art, lead to the case that conversion rate fluctuation is big, can be conveniently subsequent more accurately optimize parameter, with good use prospect.

Description

Technical Field

[0001] This invention relates to the field of reaction parameter optimization technology, specifically to a method and system for optimizing reaction parameters in natural gas to hydrogen production. Background Technology

[0002] Natural gas to hydrogen production is a mainstream hydrogen production technology in the industrial field. It refers to a chemical process that uses natural gas as raw material and, under the support of catalysts and specific process conditions, produces high-purity hydrogen through multiple reactions and separations. Its core process includes three key steps: steam reforming, shift reaction, and adsorption separation. First, under the action of a nickel-based catalyst, methane and superheated steam undergo a steam reforming reaction to produce a mixture containing hydrogen, carbon monoxide, and carbon dioxide. Then, the mixture enters the shift reaction unit, where carbon monoxide is further converted into hydrogen and carbon dioxide. Finally, impurities are removed through pressure swing adsorption technology to obtain pure hydrogen.

[0003] An existing invention patent application with publication number CN115159458B, entitled "A Natural Gas Hydrogen Production System and Method," describes a system including a steam reformer, a steam shift unit, and an adsorption tower. The steam reformer includes a desulfurizer and a converter, with the converter connected to the steam shift unit. The steam shift unit includes a boiler feedwater preheater, a converted gas water cooler, and a converted gas-water separator. The converted gas-water separator is connected to the adsorption tower. A burner is located at the bottom of the converter, and the heat required for methane conversion is provided by burning the fuel-gas mixture at the bottom burner. This invention enables standardized production, forming a standardized series of products, facilitating equipment management for users, ensuring universal spare parts, and reducing the operating costs of the equipment.

[0004] However, the existing solutions have certain drawbacks. Existing patents generally use a single benchmark parameter under fixed working conditions, which does not cover actual fluctuating working conditions such as seasonal changes and supplier switching. This results in the parameter reference being out of touch with the actual site conditions, leading to high fluctuations in conversion rates. Moreover, existing solutions often rely on engineers' experience to set initial parameters. When working conditions fluctuate, parameter mismatch leads to a decrease in conversion rates. In summary, existing natural gas hydrogen production systems and methods do not meet market demands. Therefore, we propose a method and system for optimizing reaction parameters in natural gas hydrogen production. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for optimizing reaction parameters in natural gas-to-hydrogen production includes the following steps: Identify the core process parameters among the key unit parameters, acquire historical operating data, classify and store them according to the three-dimensional structure of raw material properties, reaction process, and product indicators, sort out the key correlation rules between parameters, and form a parameter optimization benchmark library; Based on the correlation patterns of parameters, a parameter prediction model is constructed, and based on historical operating data in the benchmark parameter library, the final parameter set for the next batch of reactions is generated. The final parameter set of the next batch reaction is applied, and a monitoring and control system is used to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; Analyze real-time data in the real-time dataset, establish a threshold judgment system, screen and identify abnormal parameters through the threshold judgment system, calculate the parameter coupling influence degree, and locate the root cause of the deviation. Based on the root causes of the deviation, the core parameters of the reaction are adjusted in a stepwise manner. During the adjustment process, real-time monitoring data is called in simultaneously to verify the matching between the adjustment measures and the root causes of the deviation.

[0006] Preferably, the steps for forming a parameter optimization benchmark library are as follows: Integrate key unit parameters of natural gas to hydrogen production process and establish a historical data benchmark framework; Refine and anchor the core process parameters in the key unit parameters; Acquire continuous historical operating data for the past year, which includes continuous reaction data under different raw material properties, different loads, and different ambient temperatures. A database table is constructed based on a three-dimensional structure of raw material properties, reaction process, and product indicators; Obtain correlation data between steam replenishment and hydrogen-to-carbon ratio, and correlation data between fuel gas flow rate and methane steam conversion temperature, and form a core correlation data comparison table to clarify the linear correlation between flow rate and temperature.

[0007] Preferably, the key units of the natural gas to hydrogen process include feedstock pretreatment, methane steam reforming reaction, and product adsorption separation. The feedstock pretreatment parameters include feedstock pressurization parameters, feedstock preheating parameters, and desulfurization parameters. The methane steam reforming reaction parameters include methane steam reforming temperature reference parameters, water-to-carbon ratio reference parameters, and fuel gas flow rate reference parameters. The product adsorption separation parameters include adsorption tower configuration parameters and initial adsorption cycle setting parameters.

[0008] Preferably, the steps for constructing a parameter prediction model based on parameter correlation patterns include: Determine the core logic of the input and output of the parameter prediction model, and clarify the core framework for constructing the parameter prediction model; Core indicators are selected based on core logic, and these core indicators are used as the objects of deviation quantification. A unified deviation calculation method was developed for the core indicators, and the optimal value standard for each indicator was clearly defined. Based on the deviation calculation method, the corresponding influencing factors are marked according to the deviation direction, and a correlation mechanism between deviation and influencing factors is established. Regression analysis is conducted based on historical operating data to obtain the preset formulas for core parameters, and a model is constructed and trained to obtain a parameter prediction model. The initial parameter set for the next batch of reaction is predicted using a parameter prediction model, and the parameter value range is limited by embedded process constraints to form the final parameter set for the next batch of reaction.

[0009] Preferably, the steps for using a monitoring and control system to achieve real-time acquisition of parameters at each monitoring point include: Analyze the data from the final parameter set of the next batch of reaction, analyze the monitoring nodes of raw material pretreatment, reaction process and product separation, and clarify the monitoring coverage; Based on the data required to be collected by the monitoring nodes, select and deploy monitoring equipment; The system acquires real-time data from monitoring equipment, integrates and correlates the acquired data in real time, constructs correlation curves between raw materials, reactions, and products, forms a complete data storage chain, and obtains a real-time data set. Implement latency control for the entire data storage chain.

[0010] Preferably, the steps for analyzing real-time data in a real-time dataset and establishing a threshold determination system include: Based on the safety and performance thresholds of the steam reforming process, a method combining single-parameter initial screening and multi-parameter verification is used to retrieve the optimal thresholds from real-time full-process related data and parameter optimization benchmark library. Based on process characteristics, equipment limits, and product requirements, a threshold determination system combining safety thresholds and performance thresholds is constructed. Based on the characteristics of the process equipment and safety protection requirements, and according to the equipment tolerance limit and catalyst protection mechanism, the core safety boundary is defined and the safety threshold is obtained. By anchoring reaction efficiency and product quality targets, matching the requirements of subsequent processes, defining performance compliance boundaries, and obtaining performance thresholds.

[0011] Preferably, the steps for identifying the root cause of positioning deviation include: The real-time data is compared one by one with the threshold system, parameters that are out of range are marked, a list of abnormal parameters is generated, and suspicious indicators are identified. For suspicious indicators, call the multi-parameter correlation curve of the data system to clarify the range of correlation parameters and obtain coupled data; By comparing the correlation logic analysis to determine the synchronicity of parameter changes, isolated fluctuations are eliminated, and related factors are identified. By combining the process mechanism to verify the rationality of the correlation, eliminating non-correlated interference, clarifying the type of deviation, the core cause and the related parameters, a deviation root cause determination report is formed.

[0012] Preferably, the steps for implementing stepwise adjustments to the core reaction parameters include: The data in the deviation root cause determination report are analyzed, and the deviations are divided into reaction parameter deviations and auxiliary system deviations; For each type of deviation, a plan is developed. For deviations of reaction parameters, a plan is developed by clarifying the correction law in conjunction with the reaction mechanism. For deviations of auxiliary systems, a plan is developed by determining the adjustment path based on the operating characteristics of the equipment. The scheme adjusts the parameters step by step according to the preset minimum gradient within the safety and performance thresholds.

[0013] Preferably, it also includes periodically reviewing parameter adjustment data and core product indicators, updating the benchmark parameter library, and revising the preset calculation formula based on the newly added operating data, and calibrating the threshold judgment system in combination with the fluctuation range of raw material properties.

[0014] A natural gas-to-hydrogen reaction parameter optimization system includes a data analysis module, a model analysis module, a data acquisition module, an anomaly analysis module, and a parameter adjustment module. Data analysis module: Identify the core process parameters among the key unit parameters, acquire historical operating data, classify and store it according to the three-dimensional structure of raw material properties, reaction process, and product indicators, sort out the key correlation patterns between parameters, and form a parameter optimization benchmark library; Model analysis module: Based on the correlation patterns of parameters, constructs a parameter prediction model and generates the final parameter set for the next batch of reactions based on historical operating data in the benchmark parameter library; Data acquisition module: The next batch of reaction parameters is applied, and a monitoring and control system is used to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; Anomaly Analysis Module: Analyzes real-time data in the real-time dataset, establishes a threshold judgment system, screens and identifies abnormal parameters through the threshold judgment system, calculates the parameter coupling influence degree, and locates the root cause of deviation. Parameter adjustment module: Based on the root cause of deviation, the core parameters of the reaction are adjusted in a stepwise manner. During the adjustment process, real-time monitoring data is called up simultaneously to verify the matching between the adjustment measures and the root cause of deviation. Calibration and Adjustment Module: Periodically reviews parameter adjustment data and core product indicators, updates the benchmark parameter library, and corrects preset calculation formulas based on newly added operating data, and calibrates the threshold judgment system in combination with the fluctuation range of raw material properties.

[0015] This invention provides a method and system for optimizing reaction parameters in natural gas-to-hydrogen production, which has the following beneficial effects: This invention is based on a parameter optimization benchmark library formed by classifying and storing parameters according to the three-dimensional structure of raw material properties, reaction process, and product indicators. It also obtains historical operating data from the past year, including different operating conditions. This effectively avoids the situation in existing technologies where a single benchmark parameter under fixed operating conditions leads to large fluctuations in conversion rate. It facilitates more accurate parameter optimization in the future and has good performance.

[0016] This invention, based on historical operating data in a benchmark parameter library, uses a constructed model to comprehensively analyze and process the data, generating the final parameter set for the next batch of reactions. Compared with existing methods that rely on engineers' experience to set initial parameters, this invention has a higher accuracy rate in parameter setting. Furthermore, by analyzing real-time data, it identifies the root causes of deviations and further adjusts the real-time data, thereby greatly optimizing the parameters of the natural gas to hydrogen production reaction. This effectively improves the conversion rate, demonstrates good performance, and shows promising application prospects.

[0017] This invention implements a step-by-step adjustment of the core reaction parameters, simultaneously calling real-time monitoring data during the adjustment process to verify the matching of adjustment measures with the root causes of deviations. Compared with the fixed parameter adjustment method of the prior art, its adjustment scheme is more flexible and has a wider range of adaptability. It solves the defects of the existing schemes, such as single benchmark, empirical parameter preset, monitoring lag, and rough adjustment. It achieves accurate matching and continuous optimization of reaction parameters under different operating conditions, greatly improves the hydrogen purity compliance rate, has good use effect, and has good application prospects. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method for optimizing reaction parameters in natural gas to hydrogen production according to the present invention; Figure 2 This is a structural block diagram of a natural gas-to-hydrogen reaction parameter optimization system according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention is based on patent documents of this category in the background art and is mainly applied to the optimization and adjustment of parameters when using this technology.

[0021] Example 1: Please see Figure 1 This embodiment provides a method for optimizing reaction parameters in natural gas to hydrogen production, including the following steps: S1. Determine the core process parameters among the key unit parameters, obtain historical operating data, classify and store them according to the three-dimensional structure of raw material properties, reaction process, and product indicators, sort out the key correlation rules between parameters, and form a parameter optimization benchmark library. This step mainly involves integrating the parameters and optimization indicators of mature steam reforming processes and establishing a historical data benchmark framework. The core parameters of the steam reforming process include key control values ​​for feedstock pressurization, preheating, and desulfurization, temperature and water-to-carbon ratio parameters for methane steam reforming, and operating parameters for multi-adsorption tower rotation separation, which are used for subsequent parameter optimization.

[0022] The steps to create a parameter optimization benchmark library are as follows: Integrate key unit parameters of natural gas to hydrogen production process and establish a historical data benchmark framework; Obtain the natural gas to hydrogen production process, clarify the core framework of the process, and integrate the key unit parameters of the existing steam reforming process, which generally include raw material pretreatment, steam reforming reaction, product separation and other links.

[0023] Refine and anchor the core process parameters in the key unit parameters; The key units of the natural gas to hydrogen process include feedstock pretreatment, methane steam reforming, and product adsorption and separation. The parameters of feedstock pretreatment include feedstock pressurization parameters, feedstock preheating parameters, and desulfurization parameters. The parameters of methane steam reforming include the methane steam reforming temperature baseline parameter, the water-to-carbon ratio baseline parameter, and the fuel gas flow rate baseline parameter. The parameters of product adsorption and separation include the adsorption tower configuration parameters and the initial setting parameters of the adsorption cycle.

[0024] Under normal circumstances, natural gas undergoes three-stage pressurization via a reciprocating compressor, with the final outlet pressure stabilized at 1.6 MPa, meaning the feedstock pressurization parameter is 1.6 MPa. The pressurized natural gas then enters a shell-and-tube heat exchanger, where it is preheated using the high-temperature tail gas from the reaction as a heat source. The preheating temperature is controlled between 370-390℃, with temperature fluctuations not exceeding ±5℃, meaning the feedstock preheating parameter is 370-390℃. A fixed-bed desulfurizer filled with zinc oxide desulfurizer is used; the desulfurizer particle size is typically 3-5 mm, and the bed height to diameter ratio is generally 4:1. The methane steam reformer adopts a tubular structure. The tube is filled with a nickel-based catalyst, and the reaction temperature is set at 850℃, which is the baseline parameter for methane steam reforming. The water-to-carbon ratio is set at 3-4. The fuel gas flow rate is set at 80 cubic meters per hour to maintain the thermal balance of the methane steam reforming reactor. The adsorption tower configuration is set to 4-6 adsorption towers operating in parallel. The adsorption towers are filled with molecular sieve adsorbent, and the adsorbent filling amount is calculated based on 1.2 cubic meters of adsorbent for every 100 cubic meters of gas processed per hour. The initial adsorption cycle is set at 120 seconds to ensure continuous hydrogen production.

[0025] The parameters mentioned above are all commonly used parameters for an existing process. Different processes may have different parameters, but the processing principle and subsequent processing steps are the same.

[0026] Acquire continuous historical operating data for the past year, which includes continuous reaction data under different raw material properties, different loads, and different ambient temperatures. The historical operating data for the past year is at least from the last 30 consecutive reactions, including operating conditions with different raw material properties, different loads, and different ambient temperatures, to ensure comprehensive data and avoid any missing data.

[0027] A database table is constructed based on a three-dimensional structure of raw material properties, reaction process, and product indicators; The specific fields of the database table constructed by the three-dimensional structure are raw material attributes, reaction process, and product indicators. The raw material attributes are methane purity, raw material gas flow rate, raw material gas pressure, raw material gas temperature, sulfur content, and humidity. The reaction process is methane steam reforming inlet temperature, methane steam reforming bed temperature, system pressure, water-to-carbon ratio, steam makeup amount, and fuel gas flow rate. The product indicators are hydrogen purity, hydrogen production, methane conversion rate, carbon monoxide conversion rate, and hydrogen-to-carbon ratio.

[0028] Obtain correlation data between steam replenishment and hydrogen-to-carbon ratio, and correlation data between fuel gas flow rate and methane steam conversion temperature, and form a core correlation data comparison table to clarify the linear correlation between flow rate and temperature.

[0029] Two types of core correlation data were extracted to form a comparison table, which showed that the relationship between steam replenishment and hydrogen-carbon ratio was positively correlated, and there was a significant positive correlation between fuel gas flow rate and methane steam conversion temperature. As the fuel gas flow rate increased, the reactor bed temperature increased accordingly, and the two showed an approximately linear growth relationship.

[0030] This invention is based on a parameter optimization benchmark library formed by classifying and storing parameters according to the three-dimensional structure of raw material properties, reaction process, and product indicators. It also obtains historical operating data from the past year, including different operating conditions. This effectively avoids the situation in existing technologies where a single benchmark parameter under fixed operating conditions leads to large fluctuations in conversion rate. It facilitates more accurate parameter optimization in the future and has good performance.

[0031] S2. Based on the parameter correlation rules, construct a parameter prediction model and generate the final parameter set for the next batch of reactions based on historical operating data in the benchmark parameter library. Based on the correlation patterns of parameters, the steps to construct a parameter prediction model include: Determine the core logic of the input and output of the parameter prediction model, and clarify the core framework for constructing the parameter prediction model; Core indicators are selected based on core logic, and these core indicators are used as the objects of deviation quantification. Six core indicators that play a decisive role in reaction efficiency, product quality, and system safety were selected as the objects of deviation quantification. The six core indicators are hydrogen purity, methane conversion rate, system pressure, methane steam reforming bed temperature, hydrogen-to-carbon ratio, and adsorption tower pressure fluctuation range. These are core indicators in general processes. For other processes, they can be screened separately. All six core indicators are key control indicators marked in the parameter optimization benchmark library.

[0032] A unified deviation calculation method was developed for the core indicators, and the optimal value standard for each indicator was clearly defined. Deviation value = Actual value of the previous batch - Optimal value of the parameter optimization benchmark library. The optimal value of the parameter optimization benchmark library is the optimal statistical value of the corresponding indicator in the historical 30 batches of data.

[0033] If the parameter optimization benchmark library contains 50 sets of data, then the optimal value of the parameter optimization benchmark library is the optimal statistical value of the corresponding indicator in the historical 50 batches of data.

[0034] Based on the deviation calculation method, the corresponding influencing factors are marked according to the deviation direction, and a correlation mechanism between deviation and influencing factors is established. Based on the calculation results, the deviation direction is marked and associated with the corresponding influencing factors. For example, regarding hydrogen purity deviation, if the purity of the previous batch of hydrogen was 99.93%, and the optimal value in the parameter optimization benchmark library is 99.97%, then the hydrogen purity deviation value = 99.93% - 99.97% = -0.04%, resulting in a negative number, i.e., a negative deviation. A negative deviation indicates that the purity does not meet the standard. In this case, combined with the analysis of the product separation process characteristics, the purity deviation is mainly related to the adsorption time, and the adsorption time needs to be adjusted. The calculated adsorption time deviation t1 = +5 seconds, where + indicates that the adsorption time needs to be extended to improve the purity. Correspondingly, if the result is positive, the calculated adsorption time deviation t1 = -5 seconds, where - indicates that the adsorption time needs to be extended to improve the purity. The choice of 5 seconds is because the absolute value of hydrogen purity deviation is generally less than 0.1%, and the adjusted adsorption time deviation is always 5 seconds. It can be adjusted according to the process situation. Adjusting the set value is mainly for ease of adjustment and to reduce the amount of data for calculation and analysis.

[0035] Regression analysis is conducted based on historical operating data to obtain the preset formulas for core parameters, and a model is constructed and trained to obtain a parameter prediction model. Based on the parameter correlation patterns of the steam reforming process and the regression analysis results of historical data, we construct generation formulas for three core preset parameters: methane steam reforming temperature, steam replenishment rate, and adsorption cycle. Process constraints are embedded in the formulas to ensure that the calculation results are within a safe and efficient range.

[0036] The preset temperature for methane steam reforming is calculated as 850 - 0.6 × F, where 850 is the optimal temperature baseline for the methane steam reforming process, F is the deviation of the previous batch of fuel gas flow rate (i.e., the actual fuel gas flow rate of the previous batch - 80), and 80 is the optimal flow rate for temperature stability and energy consumption balance in historical data. When F is positive, it indicates that the previous batch of fuel gas flow rate was too high and the temperature was too high, requiring a lower preset temperature; when F is negative, it indicates that the flow rate was too low and the temperature was too low, requiring a higher preset temperature. 0.6 is a correction coefficient, obtained through regression analysis of historical data. The calculated result must be limited to the safe and efficient range of 830-870℃. If the calculated value is <830℃, then 830℃ is used; if the calculated value is >870℃, then 870℃ is used.

[0037] The preset value for steam replenishment is 8.0 mol / s + 0.3 × t2, where 8.0 is the baseline value for steam replenishment, t2 is the deviation of the adsorption time of the previous batch. When t2 is positive, it indicates that the adsorption time of the previous batch was too long, and the steam replenishment can be appropriately reduced; when it is negative, it indicates that the adsorption time was too short, and the steam replenishment needs to be increased to improve the hydrogen-carbon ratio and indirectly assist in purity optimization. 0.3 is the correction coefficient, which is obtained through orthogonal experiments. For every 1 second deviation in adsorption time, the steam replenishment needs to be adjusted in the same direction by 0.3 mol / s to maintain a stable hydrogen-carbon ratio.

[0038] The preset adsorption cycle value is calculated as 120 seconds - 0.5 × purity deviation. 120 seconds is the baseline value for the adsorption cycle. The purity deviation is the difference between the purity of the previous batch of hydrogen and the baseline value of 99.97%, calculated as actual purity - 99.97%. When the purity deviation is negative, it indicates that the purity does not meet the standard, and the adsorption cycle needs to be shortened to improve the adsorption efficiency. When the purity deviation is positive, it indicates that the purity exceeds the standard, and the cycle can be extended to save energy. 0.5 is a correction coefficient, which is obtained through regression analysis of adsorption experimental data. Specifically, for every 1% deviation in hydrogen purity, the adsorption cycle needs to be adjusted in the opposite direction by 0.5 seconds.

[0039] The initial parameter set for the next batch of reaction is predicted using a parameter prediction model, and the parameter value range is limited by embedded process constraints to form the final parameter set for the next batch of reaction.

[0040] S3. Apply the final parameter set of the next batch reaction and use a monitoring and control system to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; The steps for real-time acquisition of parameters at each monitoring point using a monitoring and control system include: Analyze the data from the final parameter set of the next batch of reaction, analyze the monitoring nodes in the raw material pretreatment, reaction process and product separation stages, and clarify the monitoring coverage; according to the monitoring node layout plan, configure sensors such as temperature, pressure, flow and component analyzers to ensure that the data acquisition accuracy and response speed meet the control requirements; set the data acquisition frequency to once per second to ensure the integrity of process dynamic capture; Based on the data required to be collected by the monitoring nodes, select and deploy monitoring equipment; An online sulfur analyzer is deployed for raw material pretreatment to detect the sulfur content of the natural gas after desulfurization, strictly matching the requirements of mature desulfurization processes. A gas chromatograph is configured to detect the methane purity in the natural gas, providing a basis for adjusting the water-carbon ratio in the steam reforming process. A humidity sensor is installed to detect the humidity of the raw gas, assisting in judging the deviation of the actual water-carbon ratio.

[0041] The reaction process involves deploying thermocouples at different axial positions in the methane steam reformer to detect bed temperature and monitor in real time whether the temperature is within a safe and efficient range; installing pressure sensors to detect system pressure in real time and ensure that the pressure is maintained within the range of 1.6-1.8 MPa; and deploying steam flow meters and feed gas flow meters to synchronously collect flow data and calculate the real-time water-to-carbon ratio.

[0042] The product separation process is equipped with an online hydrogen purity meter to detect hydrogen purity and ensure that the purity meets the standards; hydrogen flow meters and carbon monoxide flow meters are set up to record hydrogen production and carbon monoxide production simultaneously and calculate methane conversion rate; pressure sensors are installed in the adsorption towers to detect the pressure of each tower and monitor whether the pressure fluctuation meets the requirements.

[0043] The system acquires real-time data from monitoring equipment, integrates and correlates the acquired data in real time, constructs correlation curves between raw materials, reactions, and products, forms a complete data storage chain, and obtains a real-time data set. A data acquisition and monitoring control system is adopted to realize the real-time correlation and time series storage of parameters at each monitoring point, construct multi-dimensional correlation curves, and perform delay control on the complete data storage link to ensure that the data transmission delay is less than 1 second, so that the complete parameter change chain can be traced when deviation occurs.

[0044] S4. Analyze the real-time data in the real-time dataset, establish a threshold judgment system, screen and identify abnormal parameters through the threshold judgment system, calculate the parameter coupling influence degree, and locate the root cause of the deviation. The steps for analyzing real-time data in a real-time dataset and establishing a threshold determination system include: Based on the safety and performance thresholds of the steam reforming process, a method combining single-parameter initial screening and multi-parameter verification is used to retrieve the optimal thresholds from real-time full-process related data and parameter optimization benchmark library. The core benchmark for deviation analysis is defined as the safety and performance thresholds of existing mature steam reforming processes. A two-dimensional analysis method combining single-parameter anomaly screening with multi-parameter linkage verification is adopted to complete the basic preparation work. The correlation data of the entire process of raw materials, reaction and products in the real-time monitoring system are retrieved, and the optimal threshold data of the corresponding operating conditions in the benchmark parameter library are called simultaneously to ensure the real-time and correlation of the analysis data.

[0045] Based on process characteristics, equipment limits, and product requirements, a threshold determination system combining safety thresholds and performance thresholds is constructed. Based on the characteristics of steam reforming processes, the limits of equipment materials, and product quality requirements, a dual threshold system including safety thresholds and performance thresholds is constructed as the core basis for deviation judgment.

[0046] Safety thresholds cover key indicators such as the upper limit of reactor wall temperature, the maximum pressure limit of the system, and the warning line for the concentration of combustible components in hydrogen, ensuring that operation does not exceed the safety boundaries of equipment and process. For example, the methane vapor conversion temperature is lower than the reactor shell material's tolerance limit of 870℃, the methane vapor conversion temperature is higher than the catalyst coking temperature of 830℃, the sulfur content after desulfurization is lower than the content of nickel-based catalyst poisoning by 0.1ppm, the system pressure is lower than the upper limit of the adsorption tower's design pressure of 1.8MPa, and the operating pressure fluctuation of the adsorption tower is lower than the tower body stress change pressure ±0.05MPa.

[0047] For example, the performance thresholds are methane conversion rate ≥90%, hydrogen-to-carbon ratio ≥2.0, hydrogen purity ≥99.97%, adsorption tower pressure fluctuation ≤±0.02MPa, and steam reforming bed temperature fluctuation ≤±5℃. These data can be adjusted according to different processes.

[0048] Based on the characteristics of the process equipment and safety protection requirements, and according to the equipment tolerance limit and catalyst protection mechanism, the core safety boundary is defined and the safety threshold is obtained. By anchoring reaction efficiency and product quality targets, matching the requirements of subsequent processes, defining performance compliance boundaries, and obtaining performance thresholds.

[0049] The steps to address the root causes of positioning deviations include: Compare real-time data one by one with the threshold system, mark parameters that are out of range, generate a list of abnormal parameters, and lock down suspicious indicators; Using second-level / minute-level data collected by the real-time monitoring system as input, threshold comparisons are performed on key parameters throughout the entire process one by one to identify abnormal data points that exceed safety and performance thresholds, and a list of abnormal parameters containing parameters exceeding limits, the magnitude of exceeding limits, and time series characteristics is generated. For suspicious indicators, call the multi-parameter correlation curve of the data system to clarify the range of correlation parameters and obtain coupled data; For suspicious indicators, the multi-parameter correlation curves stored in the data acquisition and monitoring control system are called up to analyze the coupling relationship between abnormal parameters and other related parameters.

[0050] By comparing the correlation logic analysis to determine the synchronicity of parameter changes, isolated fluctuations are eliminated, and correlation factors are identified. By combining the process mechanism to verify the rationality of the correlation, eliminating non-correlated interference, clarifying the type of deviation, the core cause and the related parameters, a deviation root cause determination report is formed.

[0051] Based on the results of multi-parameter linkage analysis, combined with the core process mechanisms such as steam reforming and adsorption separation, the root cause of the deviation is finally determined. For example, the rationality of the correlation is verified by comparing with the process mechanism, non-correlated interference factors are eliminated, and a list of parameter anomalies containing parameters exceeding the limit, the magnitude of the exceedance, and the time sequence characteristics is generated to clarify the type of deviation, the core cause, and the related parameters.

[0052] This invention, based on historical operating data in a benchmark parameter library, uses a constructed model to comprehensively analyze and process the data, generating the final parameter set for the next batch of reactions. Compared with existing methods that rely on engineers' experience to set initial parameters, this invention has a higher accuracy rate in parameter setting. Furthermore, by analyzing real-time data, it identifies the root causes of deviations and further adjusts the real-time data, thereby greatly optimizing the parameters of the natural gas to hydrogen production reaction. This effectively improves the conversion rate, demonstrates good performance, and shows promising application prospects.

[0053] S5. Based on the root cause of the deviation, implement stepwise adjustments to the core parameters of the reaction. During the adjustment process, call real-time monitoring data simultaneously to verify the matching between the adjustment measures and the root cause of the deviation.

[0054] Primarily based on the established parameter safety and performance thresholds of existing technologies, this approach combines single-parameter anomaly screening with multi-parameter linkage analysis to accurately pinpoint the root cause of deviations, providing a clear direction for parameter adjustments.

[0055] The scheme adjusts the parameters step by step according to the preset minimum gradient within the safety and performance thresholds.

[0056] The steps for implementing stepwise adjustments to the core reaction parameters include: The data in the deviation root cause determination report are analyzed, and the deviations are divided into reaction parameter deviations and auxiliary system deviations; For each type of deviation, a plan is developed. For deviations of reaction parameters, a plan is developed by clarifying the correction law in conjunction with the reaction mechanism. For deviations of auxiliary systems, a plan is developed by determining the adjustment path based on the operating characteristics of the equipment. The scheme adjusts the parameters step by step according to the preset minimum gradient within the safety and performance thresholds.

[0057] For example, a deviation scenario during desulfurization system adjustment is when the online sulfur analyzer detects sulfur content >0.1ppm twice consecutively.

[0058] Immediately switch to the standby desulfurization bed, close the original bed valves and record the regeneration start time; reduce the methane vapor conversion temperature to 830℃; when the sulfur content is ≤0.1ppm for 3 consecutive times, raise the temperature back to the baseline value at 5℃ / min, ensuring that the methane conversion rate is ≥85% during this period; regenerate the desulfurizing agent by purging with nitrogen.

[0059] S6. Periodically review parameter adjustment data and core product indicators, update the benchmark parameter library, and revise the preset calculation formula based on the newly added operating data, and calibrate the threshold judgment system in combination with the fluctuation range of raw material properties.

[0060] Specifically, the system takes parameter adjustment records and product index data as input, processes the data through statistical analysis, regression modeling, and process adaptation, corrects the preset parameter formulas, calibrates the benchmark thresholds, and finally synchronizes them to the system to achieve model adaptation to fluctuations in operating conditions.

[0061] For example, when revising the model formula, the system uses linear regression to correct the preset formula based on the latest data for different operating conditions, ensuring that the formula prediction error is ≤0.1. At the same time, the system periodically calls extreme operating condition samples from the historical database to verify the accuracy of the corrected model's response under boundary conditions, ensuring that it still has reliable guidance under scenarios such as sudden changes in raw gas composition and load fluctuations.

[0062] Different seasons bring different external environments, different humidity levels in raw materials, and different levels of methane purity. Therefore, the threshold range needs to be adjusted accordingly. For example, when the purity is low in winter, the lower limit of temperature is adjusted from 830℃ to 835℃, and the fuel gas flow rate benchmark is adjusted to 82 cubic meters per hour in winter to compensate for the temperature requirements.

[0063] This invention implements a step-by-step adjustment of the core reaction parameters, simultaneously calling real-time monitoring data during the adjustment process to verify the matching of adjustment measures with the root causes of deviations. Compared with the fixed parameter adjustment method of the prior art, its adjustment scheme is more flexible and has a wider range of adaptability. It solves the defects of the existing schemes, such as single benchmark, empirical parameter preset, monitoring lag, and rough adjustment. It achieves accurate matching and continuous optimization of reaction parameters under different operating conditions, greatly improves the hydrogen purity compliance rate, has good use effect, and has good application prospects.

[0064] After each adjustment, the data from the entire adjustment is summarized to form an iterative optimization report. The report includes a comparative analysis of parameters before and after adjustment, trends in product index changes, and system response timeliness, which is used to evaluate the effectiveness of the adjustment. Based on the report's conclusions, the system automatically identifies highly sensitive parameters and marks them as optimization priorities, incorporating them into the next round of calibration plan. At the same time, iterative data is synchronously updated to the model training library to enhance its adaptability to seasonal working conditions. All operation records are archived in real time to ensure complete traceability.

[0065] Example 2 Based on Example 1, this example describes the following: A natural gas-to-hydrogen reaction parameter optimization system includes a data analysis module, a model analysis module, a data acquisition module, an anomaly analysis module, and a parameter adjustment module. Data analysis module: Identify the core process parameters among the key unit parameters, acquire historical operating data, classify and store it according to the three-dimensional structure of raw material properties, reaction process, and product indicators, sort out the key correlation patterns between parameters, and form a parameter optimization benchmark library; Model analysis module: Based on the correlation patterns of parameters, constructs a parameter prediction model and generates the final parameter set for the next batch of reactions based on historical operating data in the benchmark parameter library; Data acquisition module: The next batch of reaction parameters is applied, and a monitoring and control system is used to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; Anomaly Analysis Module: Analyzes real-time data in the real-time dataset, establishes a threshold judgment system, screens and identifies abnormal parameters through the threshold judgment system, calculates the parameter coupling influence degree, and locates the root cause of deviation. Parameter adjustment module: Based on the root cause of the deviation, the core parameters of the reaction are adjusted in a stepwise manner. During the adjustment process, real-time monitoring data is called up simultaneously to verify the matching of the adjustment measures with the root cause of the deviation.

[0066] Calibration and Adjustment Module: Periodically reviews parameter adjustment data and core product indicators, updates the benchmark parameter library, and corrects preset calculation formulas based on newly added operating data, and calibrates the threshold judgment system in combination with the fluctuation range of raw material properties.

[0067] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0068] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing reaction parameters in natural gas to hydrogen production, characterized in that, Includes the following steps: Identify the core process parameters among the key unit parameters, acquire historical operating data, and classify and store them according to the three-dimensional structure of raw material properties, reaction process, and product indicators. Summarize the key correlation patterns between parameters and form a parameter optimization benchmark library. Based on the correlation patterns of parameters, a parameter prediction model is constructed, and based on historical operating data in the benchmark parameter library, the final parameter set for the next batch of reactions is generated. The final parameter set of the next batch reaction is applied, and a monitoring and control system is used to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; Analyze real-time data in the real-time dataset, establish a threshold judgment system, screen and identify abnormal parameters through the threshold judgment system, calculate the parameter coupling influence degree, and locate the root cause of the deviation. The steps to address the root causes of positioning deviations include: Compare real-time data one by one with the threshold system, mark parameters that are out of range, generate a list of abnormal parameters, and lock down suspicious indicators; For suspicious indicators, call the multi-parameter correlation curve of the data system to clarify the range of correlation parameters and obtain coupled data; The system retrieves multi-parameter correlation curves stored in the data acquisition and monitoring control system to analyze the coupling relationship between abnormal parameters and other correlated parameters. By comparing the correlation logic analysis to determine the synchronicity of parameter changes, isolated fluctuations are eliminated, and correlation factors are identified. Verify the rationality of the correlation by combining the process mechanism, eliminate non-correlated interference, clarify the type of deviation, the core cause and the related parameters, and form a deviation root cause determination report; Based on the root causes of the deviation, the core parameters of the reaction are adjusted in a stepwise manner. During the adjustment process, real-time monitoring data is called in simultaneously to verify the matching between the adjustment measures and the root causes of the deviation.

2. The method for optimizing natural gas-to-hydrogen reaction parameters according to claim 1, characterized in that: The steps to create a parameter optimization benchmark library are as follows: Integrate key unit parameters of natural gas to hydrogen production process and establish a historical data benchmark framework; Refine and anchor the core process parameters in the key unit parameters; Acquire continuous historical operating data for the past year, which includes continuous reaction data under different raw material properties, different loads, and different ambient temperatures. A database table is constructed based on a three-dimensional structure of raw material properties, reaction process, and product indicators; Obtain correlation data between steam replenishment and hydrogen-to-carbon ratio, and correlation data between fuel gas flow rate and methane steam conversion temperature, and form a core correlation data comparison table to clarify the linear correlation between flow rate and temperature.

3. The method for optimizing reaction parameters of natural gas to hydrogen production according to claim 2, characterized in that: The key units of the natural gas to hydrogen process include feedstock pretreatment, methane steam reforming, and product adsorption and separation. The parameters of feedstock pretreatment include feedstock pressurization parameters, feedstock preheating parameters, and desulfurization parameters. The parameters of methane steam reforming include the methane steam reforming temperature baseline parameter, the water-to-carbon ratio baseline parameter, and the fuel gas flow rate baseline parameter. The parameters of product adsorption and separation include the adsorption tower configuration parameters and the initial setting parameters of the adsorption cycle.

4. The method for optimizing reaction parameters of natural gas to hydrogen production according to claim 1, characterized in that: Based on the correlation patterns of parameters, the steps to construct a parameter prediction model include: Determine the core logic of the input and output of the parameter prediction model, and clarify the core framework for constructing the parameter prediction model; Core indicators are selected based on core logic, and these core indicators are used as the objects of deviation quantification. A unified deviation calculation method was developed for the core indicators, and the optimal value standard for each indicator was clearly defined. Based on the deviation calculation method, the corresponding influencing factors are marked according to the deviation direction, and a correlation mechanism between deviation and influencing factors is established. Regression analysis is conducted based on historical operating data to obtain the preset formulas for core parameters, and a model is constructed and trained to obtain a parameter prediction model. The initial parameter set for the next batch of reaction is predicted using a parameter prediction model, and the parameter value range is limited by embedded process constraints to form the final parameter set for the next batch of reaction.

5. The method for optimizing natural gas-to-hydrogen reaction parameters according to claim 1, characterized in that: The steps for using a monitoring and control system to achieve real-time acquisition of parameters at each monitoring point include: Analyze the data from the final parameter set of the next batch of reaction, analyze the monitoring nodes of raw material pretreatment, reaction process and product separation, and clarify the monitoring coverage; Based on the data required to be collected by the monitoring nodes, select and deploy monitoring equipment; The system acquires real-time data from monitoring equipment, integrates and correlates the acquired data in real time, constructs correlation curves between raw materials, reactions, and products, forms a complete data storage chain, and obtains a real-time data set. Implement latency control for the entire data storage chain.

6. The method for optimizing reaction parameters of natural gas to hydrogen production according to claim 1, characterized in that: The steps for analyzing real-time data in a real-time dataset and establishing a threshold determination system include: Based on the safety and performance thresholds of the steam reforming process, a method combining single-parameter initial screening and multi-parameter verification is used to retrieve the optimal thresholds from real-time full-process related data and parameter optimization benchmark library. Based on process characteristics, equipment limits, and product requirements, a threshold determination system combining safety thresholds and performance thresholds is constructed. Based on the characteristics of the process equipment and safety protection requirements, and according to the equipment tolerance limit and catalyst protection mechanism, the core safety boundary is defined and the safety threshold is obtained. By anchoring reaction efficiency and product quality targets, matching the requirements of subsequent processes, defining performance compliance boundaries, and obtaining performance thresholds.

7. The method for optimizing reaction parameters of natural gas to hydrogen production according to claim 6, characterized in that: The steps for implementing stepwise adjustments to the core reaction parameters include: The data in the deviation root cause determination report are analyzed, and the deviations are divided into reaction parameter deviations and auxiliary system deviations; For each type of deviation, a plan is developed. For deviations of reaction parameters, a plan is developed by clarifying the correction law in conjunction with the reaction mechanism. For deviations of auxiliary systems, a plan is developed by determining the adjustment path based on the operating characteristics of the equipment. The scheme adjusts the parameters step by step according to the preset minimum gradient within the safety and performance thresholds.

8. The method for optimizing reaction parameters of natural gas to hydrogen production according to claim 1, characterized in that: It also includes periodically reviewing parameter adjustment data and core product indicators, updating the benchmark parameter library, and revising preset calculation formulas based on newly added operating data, and calibrating the threshold judgment system in conjunction with the fluctuation range of raw material properties.

9. A natural gas-to-hydrogen reaction parameter optimization system, using the method described in claim 1, characterized in that, include: Data analysis module: Identify the core process parameters among the key unit parameters, acquire historical operating data, classify and store it according to the three-dimensional structure of raw material properties, reaction process, and product indicators, sort out the key correlation patterns between parameters, and form a parameter optimization benchmark library; Model analysis module: Based on the correlation patterns of parameters, constructs a parameter prediction model and generates the final parameter set for the next batch of reactions based on historical operating data in the benchmark parameter library; Data acquisition module: The next batch of reaction parameters is applied, and a monitoring and control system is used to realize the real-time acquisition of parameters at each monitoring point to obtain a real-time data set; Anomaly Analysis Module: Analyzes real-time data in the real-time dataset, establishes a threshold judgment system, screens and identifies abnormal parameters through the threshold judgment system, calculates the parameter coupling influence degree, and locates the root cause of deviation. Parameter adjustment module: Based on the root cause of deviation, the core parameters of the reaction are adjusted in a stepwise manner. During the adjustment process, real-time monitoring data is called up simultaneously to verify the matching between the adjustment measures and the root cause of deviation. Calibration and Adjustment Module: Periodically reviews parameter adjustment data and core product indicators, updates the benchmark parameter library, and corrects preset calculation formulas based on newly added operating data, and calibrates the threshold judgment system in combination with the fluctuation range of raw material properties.

Images