Simulation method and device for combustion process of hydrogen-doped natural gas

Abstract

The invention provides a simulation method and device for the combustion process of hydrogen-doped natural gas, and belongs to the technical field of hydrogen-doped natural gas simulation. According to the method, a basic model for simulating the combustion process of the hydrogen-doped natural gas is established; generating a zero-dimensional model and a one-dimensional model based on the basic model and the instantiation mode; performing primary simulation on the combustion process of the hydrogen-doped natural gas by using a zero-dimensional model, and screening a plurality of target combustion parameters from the combustion parameter set according to a primary simulation result; secondary simulation is carried out based on the simulation input parameters and the one-dimensional model, and the optimal combustor configuration is determined according to the secondary simulation result; the adjusted zero-dimensional model is used for conducting three-time simulation on the combustion process of the hydrogen-doped natural gas; calculating a mechanism simplified reference value according to results of the three times of simulation; and simplifying the GRI3.0 mechanism based on the mechanism simplification reference value, and generating a mechanism description of the hydrogen-doped natural gas combustion process. The simulation method and device for the combustion process of the hydrogen-doped natural gas are used for rapidly and accurately describing the combustion mechanism of the hydrogen-doped natural gas.

Description

Technical Field

[0001] This application relates to the field of hydrogen-blended natural gas simulation technology, and in particular to a simulation method and apparatus for the combustion process of hydrogen-blended natural gas. Background Technology

[0002] As the global energy structure shifts towards cleaner energy, hydrogen-blended natural gas, combining the economic advantages of natural gas with the low-carbon properties of hydrogen, has become an important alternative fuel in industrial combustion and vehicle power. Its combustion process involves the synergistic oxidation reaction of methane and hydrogen. Accurate simulations are needed to understand the impact of parameters such as the hydrogen blending ratio and stoichiometry on combustion stability and pollutant emissions, providing support for burner design and operating condition optimization. This is currently a key research focus in the field of clean energy utilization.

[0003] Currently, simulations of hydrogen-blended natural gas combustion processes primarily rely on two technical approaches. One approach uses multi-dimensional models based on detailed chemical reaction mechanisms, such as GRI 3.0, which can accurately depict spatial parameter distributions but demands high computational resources and struggles to efficiently cover multiple operating conditions. The other approach employs models with simplified logic. These models often have fixed core settings, such as calculating heat loss in the combustion system using a uniform standard and describing the fuel-air mixing process in a single manner, without tailoring adjustments for different simulation objectives. Whether used for initial screening of basic combustion conditions or in-depth verification of the burner's structural design rationality, the core computational logic of the model remains consistent, resulting in a mismatch between the accuracy of simulation results and the demands of real-world applications. This lack of reliable data support for determining optimal burner configurations and simplifying mechanisms hinders the engineering application of hydrogen-blended natural gas combustion technology. Summary of the Invention

[0004] In view of this, this application provides a simulation method and apparatus for the combustion process of hydrogen-blended natural gas, so as to improve the versatility and accuracy of the simulation and quickly and accurately describe the combustion mechanism of hydrogen-blended natural gas.

[0005] Specifically, this application is implemented through the following technical solution:

[0006] The first aspect of this application provides a simulation method for the combustion process of hydrogen-blended natural gas, the method comprising:

[0007] Establish a basic model for simulating the combustion process of hydrogen-blended natural gas;

[0008] The instantiation method of the basic model corresponding to the simulation target is determined, and a zero-dimensional model and a one-dimensional model are generated based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process;

[0009] The zero-dimensional model is used to perform a simulation of the combustion process of hydrogen-blended natural gas. Based on the results of the simulation, multiple target combustion parameters are selected from the set of combustion parameters. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model.

[0010] The simulation input parameters are determined by the multiple target combustion parameters. A secondary simulation of the hydrogen-blended natural gas combustion process is performed based on the simulation input parameters and the one-dimensional model. The optimal burner configuration is determined based on the results of the secondary simulation.

[0011] Based on the optimal burner configuration, the zero-dimensional model was adjusted, and the hydrogen-blended natural gas combustion process was simulated three times using the adjusted zero-dimensional model.

[0012] Based on the results of the three simulations, the computer simplified the reference values.

[0013] Based on the simplified reference values ​​of the aforementioned mechanism, the GRI3.0 mechanism is simplified to generate a mechanistic description of the hydrogen-blended natural gas combustion process.

[0014] A second aspect of this application provides a simulation device for the combustion process of hydrogen-blended natural gas, the device comprising a construction module, a simulation module, and a calculation module; wherein,

[0015] The building module is used to establish a basic model for simulating the combustion process of hydrogen-blended natural gas;

[0016] The construction module is further configured to determine the instantiation method of the basic model corresponding to the simulation target, and generate a zero-dimensional model and a one-dimensional model based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process;

[0017] The simulation module is used to perform a single simulation of the hydrogen-blended natural gas combustion process using the zero-dimensional model, and to select multiple target combustion parameters from the combustion parameter set based on the results of the single simulation. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model.

[0018] The simulation module is also used to determine the simulation input parameters based on the multiple target combustion parameters, perform a secondary simulation of the hydrogen-blended natural gas combustion process based on the simulation input parameters and the one-dimensional model, and determine the optimal burner configuration based on the results of the secondary simulation.

[0019] The simulation module is also used to adjust the zero-dimensional model based on the optimal burner configuration, and to perform three simulations of the hydrogen-blended natural gas combustion process using the adjusted zero-dimensional model.

[0020] The calculation module is used to calculate a simplified reference value based on the results of the three simulations.

[0021] The calculation module is also used to simplify the GRI3.0 mechanism based on the simplified reference value and generate a mechanistic description of the hydrogen-blended natural gas combustion process.

[0022] The simulation method and apparatus for hydrogen-blended natural gas combustion provided in this application are based on the instantiation of a basic model for all burner models. For a hydrogen-blended natural gas combustion process, two instantiations with different levels of precision are performed. The first instantiated zero-dimensional model is used for the simulation of the combustion mechanism, achieving rapid simulation. The second instantiated one-dimensional model is used to solve for the optimal burner configuration that matches the current combustion scenario, and then used for the burner structure in the zero-dimensional model simulation. Firstly, this application obtains the zero-dimensional model for simulation through the instantiation of the basic model, reducing the complex modeling process. The combustion simulation model is obtained by instantiating the model based on the pre-established model template. On the other hand, in order to improve the simulation accuracy of the zero-dimensional model that inherits the characteristics of the basic model, the simulation of the optimal configuration obtained from the more accurate one-dimensional model instantiated from the same basic model can be processed in parallel with the zero-dimensional model without interfering with each other, thus improving the simulation efficiency. At the same time, the optimal configuration obtained from the one-dimensional model simulation directly intervenes in the burner structure in the zero-dimensional model, improving the accuracy of the combustion process simulation. Finally, based on the most accurate combustion process simulation calculation, the mechanism characterization is simplified. The accurate combustion process characterization can be obtained quickly through simulation without the need for complex mathematical formula derivation or reliance on experience to make inaccurate mechanism simplifications. This ensures the adaptability of the combustion process simulation to actual working conditions and achieves a concise and efficient mechanism description, providing reliable technical support for the design optimization and reaction kinetic simulation of hydrogen-blended natural gas combustion systems. Attached Figure Description

[0023] Figure 1 A flowchart of an embodiment of the simulation method for the combustion process of hydrogen-blended natural gas provided in this application;

[0024] Figure 2 This is a schematic diagram of the second embodiment of the simulation device for the combustion process of hydrogen-blended natural gas provided in this application. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0029] Figure 1 This is a flowchart of an embodiment of the simulation method for the combustion process of hydrogen-blended natural gas provided in this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0030] S101. Establish a basic model for simulating the combustion process of hydrogen-blended natural gas.

[0031] Specifically, hydrogen-blended natural gas refers to a new type of clean fuel formed by physically mixing hydrogen with traditional natural gas (mainly methane CH4, and also containing small amounts of hydrocarbons such as ethane and propane) in a certain volume or mass ratio. The basic model refers to a general theoretical framework that integrates the core physical and chemical laws of the combustion process. It is based on the equations of mass conservation, energy conservation, and momentum conservation, and incorporates the detailed chemical reaction mechanism of GRI3.0. The detailed chemical reaction mechanism of GRI3.0 covers 325 elementary reactions and 53 chemical components. At the same time, it defines the thermodynamic properties of combustion gases (CH4, H2, and air). Thermodynamic properties can include correlations between specific heat capacity, thermal conductivity, and other properties and temperature changes, forming an expandable general model framework that provides a theoretical basis for the subsequent generation of zero-dimensional and one-dimensional models.

[0032] Furthermore, for different application scenarios, the characteristics and core requirements of the corresponding combustion environment should be clearly defined first. Industrial combustion scenarios (such as kilns and boilers) need to consider large-volume combustion spaces and continuous and stable energy supply requirements, and the model should focus on high-temperature heat release efficiency and pollutant control; vehicle power scenarios need to be adapted to dynamic operating conditions (such as acceleration and deceleration), and the model should strengthen the description of combustion response speed and transient stability; household gas scenarios focus on miniaturized burners, low noise and safe operation, and the model should optimize the mixing and heat loss calculation in small spaces. Based on these scenario differences, scenario adaptation interfaces are reserved in the basic model, allowing adjustment of parameters such as reaction domain volume, initial pressure range, and boundary condition type, ensuring that the model can flexibly adapt to the simulation requirements of different scenarios. For different general burner configurations, it is necessary to sort out the structural commonalities and differences of various configurations and transform them into geometric and flow parameters that the model can recognize. The core of the premixed burner configuration is the design of the premixing chamber for fuel and air, and the model needs to include the mixing chamber volume and flow guiding structure parameters to describe the premixing uniformity. The diffusion burner configuration focuses on the injection and diffusion paths of fuel and air, and the model needs to include parameters such as nozzle diameter, injection angle, and airflow velocity distribution. Some premixed burner configurations combine both features, and the model needs to distinguish the axial length of the premixing section and the diffusion section, as well as the segmented boundary conditions, to simulate the segmented combustion process. By standardizing these configuration parameters and establishing a configuration parameter library in the basic model, the basic parameters of different configurations can be directly called, reducing the workload of repetitive modeling. For different combustible characteristics, it is necessary to focus on the changes in the ratio of hydrogen to methane in hydrogen-blended natural gas, as well as the impact of possible trace impurities (such as ethane and propane) on combustion. When the hydrogen blending ratio is different, the reactivity, calorific value, and flame propagation speed of the fuel will change significantly. The model needs to embed the reaction pathway of the co-oxidation of hydrogen and methane to ensure accurate description of the chemical reaction kinetics under different hydrogen blending ratios. For cases containing trace impurities, the degree of interference of impurities on the main combustion reaction needs to be assessed. If the impurity content is extremely low and has a weak impact on combustion characteristics, it can be simplified in the basic model, retaining only its potential impact on pollutant formation. If the impurity content is high, the oxidation reaction mechanism of the impurities needs to be supplemented to avoid simulation deviations caused by ignoring impurities. Simultaneously, the model needs to define the correlation between the thermodynamic properties of the combustible and changes in temperature and component ratio, such as the dynamic calculation logic of parameters like specific heat capacity, thermal conductivity, and viscosity, to ensure that it can reflect the changes in the physical properties of different combustibles at different combustion stages.

[0033] Furthermore, the GRI 3.0 detailed chemical reaction mechanism, developed by the General Research Institute (GRI) in the United States, is a classic chemical reaction model used to accurately describe the combustion process of methane and natural gas. It is also a detailed mechanism widely used in the current field of hydrogen-blended natural gas and pure natural gas combustion simulation. It includes 325 elementary reactions and 53 chemical components, fully covering the key reaction pathways of fuel oxidation such as methane (CH4) and hydrogen (H2). It includes both main combustion reactions (such as CH4+OH→CH3+H2O, H2+OH→H2O+H) and side reactions such as the generation and conversion of nitrogen oxides (NOx) and the evolution of intermediate products (such as H, O, OH free radicals), which can comprehensively characterize the chemical kinetics of the combustion process.

[0034] The establishment of the basic model unifies the core theoretical basis of simulation, avoiding simulation deviations caused by inconsistencies in the basic laws of subsequent models of different dimensions. At the same time, by reserving parameter interfaces (such as component ratios and boundary condition input ports), it provides the possibility for instantiation operations adapted to different simulation targets, ensuring the compatibility and accuracy of subsequent model generation.

[0035] S102. Determine the instantiation method of the basic model corresponding to the simulation target, and generate a zero-dimensional model and a one-dimensional model based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process.

[0036] Specifically, the zero-dimensional model, based on the assumption of ignoring spatial gradients, does not distinguish parameter differences at different locations within the combustion system, focusing only on the dynamic changes of the combustion reaction in the time domain. It outputs globally averaged combustion characteristic parameters by solving conservation equations in the time dimension, and its simulation accuracy is suitable for rapid screening of basic operating conditions. The one-dimensional model, on the other hand, is based on the assumption of retaining a single spatial dimension (usually the burner axis) and ignoring gradients in other spatial dimensions. Building upon the time-domain reaction calculation, it supplements the description of parameter distribution along the axial direction, outputting parameters that combine temporal dynamics and axial spatial characteristics through conservation equations coupling time and axial space. Its simulation accuracy is more suitable for burner structural design verification scenarios. While the assumptions and simulation accuracies of the zero-dimensional and one-dimensional models differ, they are both generated from the same basic model through differentiated instantiation, and the core physicochemical laws and reaction mechanism framework remain consistent. The simulation target refers to the required simulation accuracy; the simulation accuracy differs between the zero-dimensional and one-dimensional models. The instantiation method is a concrete operational approach that transforms the general basic model into an adaptive target.

[0037] Furthermore, the core simulation objective of zero-dimensional models is to match specific simulation accuracy requirements. These objectives directly limit the range of assumptions and rules used in the model. From the perspective of simulation accuracy requirements, zero-dimensional models do not need to precisely characterize the spatial parameter distribution; they only need to ensure that the calculation accuracy of macroscopic combustion characteristic parameters (such as ignition delay time, adiabatic flame temperature, and peak heat release rate) is within a reasonable range required for basic screening, meeting the accuracy standard for quickly determining the feasibility of operating conditions. There is no need to pursue higher spatial accuracy in higher-dimensional models. The purpose of zero-dimensional models is to rapidly simulate the combustion process of hydrogen-blended natural gas, covering a large number of basic operating conditions with relatively low computational cost, and screening core parameters for subsequent high-precision simulations. Based on simplification assumptions such as ignoring spatial gradients and unifying heat loss rates, the zero-dimensional model eliminates the need to solve transport equations in the spatial dimension. It can complete calculations solely through time-domain mass and energy conservation equations, significantly reducing the simulation time for single-condition scenarios. This allows for rapid simulation of various combinations of combustion parameters, including different hydrogen doping ratios, stoichiometric ratios, and initial temperatures, outputting macroscopic combustion characteristic parameters for each condition. The results obtained from rapid simulations can preliminarily determine the combustion feasibility and basic performance of different conditions. For example, it can screen for target combustion parameters that are reasonable in terms of ignition delay time, peak heat release rate that matches the equipment's tolerance range, and total pollutant generation that meets preliminary emission requirements. Conditions that clearly do not meet the requirements can be eliminated, preventing subsequent higher-precision simulations, such as those using one-dimensional models, from wasting resources on invalid conditions. This improves the overall efficiency of the simulation process and provides preliminary data support for combustion system condition optimization and structural design.

[0038] The core objective of one-dimensional model simulation is to meet higher simulation accuracy requirements. From the perspective of simulation accuracy requirements, one-dimensional model needs to further accurately characterize the spatial distribution characteristics of parameters based on the macroscopic parameter calculations of zero-dimensional model. This ensures the calculation accuracy of spatially related parameters such as flame length, axial temperature gradient, and pollutant concentration distribution along the axial direction, meeting the accuracy requirements for spatial adaptability in burner structural design verification. Its accuracy level is higher than that of zero-dimensional model, and it can more closely resemble the spatial evolution law of actual combustion system. The purpose of one-dimensional model is to simulate both time-domain combustion reaction and axial spatial distribution, providing accurate spatial parameter support for burner configuration design. At the same time, it can accept the target combustion parameters selected by zero-dimensional model, realizing the accurate correlation between parameters and structure. Based on the assumption of preserving the axial spatial dimension, the one-dimensional model can solve the equations for axial momentum, energy, and component transport, outputting spatial characteristic parameters such as flame length, axial position of the flame face, axial temperature gradient, and axial distribution of pollutant concentration. These parameters directly reflect the spatial evolution of the combustion process within the burner and are the core basis for judging whether the burner configuration (such as axial length, cross-sectional diameter, and inlet velocity distribution) is suitable. In practical applications, the one-dimensional model uses the target combustion parameters selected by the zero-dimensional model as input, combined with different burner configuration parameters (such as axial length, wall material, and secondary air position) to conduct simulations. By comparing the spatial characteristic parameters corresponding to different configurations, the optimal burner configuration is selected, where the flame length matches the burner volume, the temperature gradient meets material tolerance requirements, and pollutant emissions meet standards, avoiding the problem of parameter-structure disconnect in traditional empirical design. Simultaneously, the output spatial parameters can also feed back into the zero-dimensional model's tertiary simulations, providing real structural data for the zero-dimensional model to adjust burner configuration parameters and correct heat loss assumptions, thus improving the reliability and engineering adaptability of the overall simulation system.

[0039] By using goal-oriented instantiation, a differentiated match between accuracy and efficiency is achieved. The zero-dimensional model can quickly complete multi-condition calculations, solving the problem of slow computation caused by excessive accuracy in traditional models. The one-dimensional model supplements the spatial characteristic description, avoiding the defect that a single zero-dimensional model cannot support structural design, and laying the model foundation for subsequent phased simulations.

[0040] Furthermore, determining the instantiation method of the basic model corresponding to the simulation target, and the specific implementation steps for generating the zero-dimensional model and the one-dimensional model based on the basic model and the instantiation method include:

[0041] (1) Determine a set of assumption rules with a computational load less than a preset value based on the simulation accuracy requirements of the simulation target;

[0042] Specifically, the simulation objectives include at least rapid screening of basic operating conditions and verification of burner structural design. Based on the accuracy requirements of the simulation objectives, a preset computational threshold is established, and a set of multiple assumption rules is constructed. Each assumption rule clearly defines its impact on accuracy and the proportion of computational cost reduction. The assumption rules include at least spatial dimension rules, physical process rules, and thermal boundary rules. Control dimensions and deviation boundaries are determined according to the simulation accuracy requirements. An initial library of candidate assumption rules is constructed, and the bidirectional impact of rules on accuracy and computational cost is quantified based on this library. Candidate assumption rules are then screened based on the stated simulation accuracy requirements.

[0043] Furthermore, the simulation objectives are clearly defined to cover the combustion stages. The combustion stages of hydrogen-blended natural gas mainly include the ignition stage, flame propagation stage, stable combustion stage, and combustion decay stage. The core physicochemical processes of different stages are different. The ignition stage focuses on the generation of free radicals and the initiation of the reaction, the flame propagation stage focuses on the movement of the flame surface and the diffusion of components, the stable combustion stage focuses on the balance between heat release and pollutant generation, and the combustion decay stage involves fuel consumption and temperature drop. Based on the combustion stage selection range defined by the simulation objective, if the objective focuses only on the stable combustion stage, complex rules related to ignition and combustion decay (such as dynamic distribution rules of ignition sources) can be eliminated. If the objective covers the entire combustion stage, rules that describe the transition characteristics of each stage (such as dynamic adjustment rules of the flame surface from initiation to stabilization) should be retained. This initially defines the basic boundary for rule selection, avoiding the inclusion of redundant rules unrelated to the target stage. For the clearly defined accuracy requirements of the simulation objective, the allowable deviation range of core control parameters (macroscopic parameters such as ignition delay time and peak heat release rate, and spatial parameters such as flame length and axial temperature gradient) is broken down. Each candidate hypothesis rule is associated with the accuracy requirements. For objectives with high accuracy requirements for macroscopic parameters, simplified rules that significantly affect macroscopic parameters should be excluded. For objectives with accuracy requirements for spatial parameters, necessary rules related to spatial description (such as axial gradient retention rules) should be retained, and rules that ignore spatial dimensions should be prohibited. By establishing a rule-accuracy impact correspondence table, the deviation contribution of each rule to the core parameters is quantified, ensuring that the cumulative deviation of the selected rules does not exceed the target accuracy limit, providing an accuracy constraint basis for rule selection.

[0044] Furthermore, for candidate rules that initially meet the accuracy constraints, their computational cost reduction effect is quantified one by one. For example, ignoring the radial gradient rule can eliminate the radial transport equation solution module, reducing computational cost; the minor intermediate product quasi-steady-state approximation rule can transform differential equations into algebraic equations, reducing the component solution cost; ignoring the radiative heat transfer rule can omit radiative heat transfer calculation terms, reducing thermal boundary computational cost. By statistically analyzing the computational cost reduction ratio of each rule and combining the necessity of the rule in the accuracy constraints, an accuracy-efficiency dimension evaluation matrix is ​​constructed. Rules with high accuracy compatibility and high computational cost reduction are prioritized for retention, rules with high accuracy compatibility and medium computational cost reduction are temporarily stored, and rules with low accuracy compatibility and any computational cost reduction are eliminated.

[0045] Furthermore, based on the above analysis of boundaries, accuracy, and efficiency, multiple rounds of iterative optimization of rule combinations are conducted. First, core rules that require both necessary accuracy and the highest efficiency are selected (such as the axial gradient preservation rule, which requires necessary accuracy and reduces some computational load). Second, auxiliary rules that are compatible with accuracy and have relatively high efficiency are added (such as the quasi-steady-state approximation rule for minor intermediate products). Finally, it is verified whether the cumulative deviation of the combined rules still meets the accuracy requirements, and the reduction ratio of the total computational load is calculated. If multiple rule combinations meet the accuracy requirements, the combination with the highest reduction ratio of the total computational load is selected as the optimal solution. If a combination meets the accuracy requirements but the computational load is still too high, within the allowable range of accuracy deviation, some rules can be replaced with alternative rules that slightly reduce accuracy but significantly improve efficiency, until the unique optimal rule combination that meets the accuracy requirements and reduces the computational load the most is found, thus completing the determination of the rule range corresponding to the simulation target.

[0046] Furthermore, based on the simulation objectives of the zero-dimensional model and the one-dimensional model, the key points of accuracy control for each are broken down. The zero-dimensional model aims to quickly screen basic combustion conditions, and accuracy control focuses on macroscopic combustion characteristics. Core parameters include ignition delay time, adiabatic flame temperature, peak heat release rate, and total pollutant generation. It is necessary to ensure that the relative deviations of the simulation results of these parameters from experimental data or detailed model baseline values ​​do not exceed the upper limit of the effectiveness of supporting condition screening, without needing to focus on accuracy requirements related to spatial distribution. The one-dimensional model aims to support the spatial structure design of the burner, and accuracy control needs to take into account both macroscopic characteristics and spatial distribution characteristics. The deviation of macroscopic parameters needs to be more stringent than that of the zero-dimensional model to ensure the reliability of basic combustion performance. At the same time, new accuracy requirements for spatial parameters are added, including flame length, axial temperature gradient, axial position of the flame surface, and axial distribution of pollutant concentration. It is necessary to ensure that the deviations of the simulation results of these spatial parameters from the actual structural adaptation requirements are within an acceptable range to avoid burner design failure due to inaccurate spatial description.

[0047] Furthermore, based on the physicochemical nature of hydrogen-blended natural gas combustion simulation, an initial library of multiple candidate hypothesis rules is constructed from four dimensions: spatial simplification, neglect of physical processes, setting of thermal boundary conditions, and component treatment methods. Each rule needs to be clarified through literature review, pre-simulation experiments, or theoretical analysis to determine its specific impact on accuracy and computational load, forming quantifiable evaluation indicators. Spatial dimension rules cover rules such as neglecting global spatial gradients, retaining single axial gradients, neglecting radial gradients, and retaining multi-dimensional gradients. It is necessary to clarify the simulation deviation changes of spatial parameters (such as temperature and concentration differences at different locations) under different rules, as well as the increase or decrease in computational load due to deleting or retaining spatial dimension calculation modules. For example, neglecting spatial gradients can significantly reduce the computational load generated by solving spatial transport equations, but it will result in spatial parameters being undesirable, only outputting macroscopic parameters with global averages. The impact of this rule on the indirect deviation of macroscopic parameters needs to be quantified. Physical process rules include rules such as neglecting the influence of turbulent fluctuations on combustion, neglecting radiative heat transfer, and neglecting the mixing delay of fuel and air. It is necessary to analyze the physical processes neglected by each rule in the simulation. The weighting of actual combustion factors, such as turbulent fluctuations affecting flame propagation rate and mixing uniformity, should be carefully considered. Ignoring this process can cause deviations in flame stability parameters (such as flameout threshold conditions) and reduce computational load caused by coupled turbulence models. The relationship between the deviation magnitude and the computational load reduction ratio needs to be clearly defined. Thermal boundary rules include those using a constant heat loss rate across the entire domain, setting dynamic heat loss in segments according to burner wall material and structure, ignoring wall heat loss, and assuming adiabatic combustion. These rules need to be combined with the actual heat dissipation characteristics of the burner to quantify the simulation deviations of parameters such as temperature field and heat release rate under different thermal boundary rules. For example, a constant heat loss rate can simplify the thermal boundary calculation logic and reduce computational load, but it cannot reflect the uneven heat dissipation caused by differences in wall temperature in different regions (such as the burner inlet section and core combustion section). The degree of influence of this deviation on core parameters such as adiabatic flame temperature and combustion efficiency needs to be clearly defined.

[0048] Furthermore, based on accuracy requirements and the quantified impact of rules, rules with negligible spatial dimensions and simplified physical processes are selected from the candidate library. Examples include ignoring global spatial gradients, using constant heat loss rates, and ignoring turbulent fluctuations. For each selected rule, its total deviation from macroscopic parameters is accumulated to ensure the cumulative deviation does not exceed the accuracy limit of the zero-dimensional model. Simultaneously, the reduction ratio of total computational load is calculated to determine if the simplified computational load meets the efficiency requirements for rapid screening of multiple operating conditions. If the cumulative deviation does not exceed the upper limit but the computational load is still too high, lower-deviation simplification rules can be added (e.g., using a quasi-steady-state approximation for minor intermediate products). If the deviation exceeds the upper limit after adding rules, the rules with the greatest impact on accuracy must be removed and recombined until a rule combination with "deviation ≤ accuracy upper limit, computational load ≤ efficiency requirements" is found. The selected hypothetical rules are then used as the hypothetical rules for the zero-dimensional model.

[0049] Furthermore, while preserving the core spatial description capabilities, it is necessary to balance accuracy and computational load, prioritizing the retention of essential rules related to spatial dimensions (such as "retaining axial gradients and ignoring radial gradients") to ensure accurate description of spatial parameters. Then, simplified rules with less impact should be selected from physical processes, thermal boundaries, and component processing rules. For example, if radiative heat transfer accounts for a very low proportion in actual combustion, ignoring radiative heat transfer can be retained, but segmented dynamic heat loss must be retained to reflect the heat dissipation differences in different areas of the burner. Similarly, the cumulative deviation of rules on macroscopic parameters and spatial parameters needs to be accumulated to ensure that the deviations of both types of parameters do not exceed the accuracy requirements of the one-dimensional model. At the same time, the computational load should be calculated. If the computational load is too high, simplified rules with minimal impact on accuracy (such as removing minor pollutant components with extremely low content) can be appropriately added without affecting the core function of spatial description. Through multiple rounds of iterative adjustments, a rule combination that meets the requirements of spatial accuracy, better macroscopic accuracy, and controllable computational load can be formed. The selected assumption rules are then used as the assumption rules for the one-dimensional model.

[0050] (2) Divide the simulation accuracy levels according to the simulation accuracy requirements;

[0051] Specifically, based on the difference in simulation objectives between zero-dimensional and one-dimensional models, two levels of accuracy are defined, and the core control indicators for each level are clarified. Level 1 accuracy (adapting to zero-dimensional models): with "rapid screening of basic working conditions" as the simulation objective, the deviation of macroscopic parameters (ignition delay time, peak heat release rate) is controlled to be ≤10%, and the computational load is ≤5 minutes / working condition. Level 2 accuracy (adapting to one-dimensional models): with "support structure design" as the simulation objective, the deviation of spatial parameters (flame length, axial temperature gradient) is controlled to be ≤8%, the deviation of macroscopic parameters is controlled to be ≤5%, and the computational load is ≤30 minutes / working condition.

[0052] (3) Determine matching hypothesis rules from the set of hypothesis rules for each simulation accuracy level;

[0053] Specifically, based on the correspondence between accuracy requirements and rule impact, rules are selected and combined from a preset rule set to ensure that the model, after rule matching, can both meet accuracy indicators and control computational load. Zero-dimensional models are adapted for Level 1 accuracy, with the simulation objective of quickly selecting basic operating conditions. Control parameters are combustion characteristics (ignition delay time, peak heat release rate, etc.), with a relatively high allowable upper limit for deviation. The rule selection priority is "computational load reduction rate > accuracy impact," prioritizing rules that significantly reduce computational load and have minimal impact on macroscopic parameter deviations. One-dimensional models are adapted for Level 2 accuracy, with the simulation objective of supporting burner spatial design. Control parameters include macroscopic characteristics (lower deviation upper limit) and spatial characteristics (flame length, axial temperature gradient, etc.). The rule selection priority is "spatial description completeness > macroscopic accuracy impact > computational load reduction rate," prioritizing rules that accurately describe spatial parameters while also considering macroscopic accuracy and computational efficiency.

[0054] Furthermore, the rule of ignoring the global spatial gradient is prioritized. This rule removes all spatial dimension calculation modules and retains only the time domain reaction solution, which can significantly reduce the amount of computation and has only a small impact on the deviation of macroscopic parameters, which is consistent with the characteristic that the first-level accuracy does not require spatial description. The constant heat loss rate rule is added to replace the dynamic heat loss calculation that changes with space / temperature with a fixed value, which further reduces the amount of computation and has a small impact on the deviation of macroscopic temperature and heat release rate. It also does not require additional input of burner structural parameters, which is suitable for the zero-dimensional model without spatial parameter input. The rule of ignoring turbulence fluctuation is added. The impact of turbulence fluctuation on macroscopic parameters (such as ignition delay time) can be ignored in the basic working condition screening, but it can reduce the amount of computation of coupled turbulence model, which is in line with the principle of efficiency first. The deviations of the selected rules are superimposed to confirm that they do not exceed the upper limit of the first-level accuracy. After the verification is passed, the selected rule combination is used as the assumption rule of the zero-dimensional model.

[0055] Furthermore, the rule of retaining axial gradient and ignoring radial gradient is mandatory. This rule retains the axial spatial dimension calculation (outputting flame length and axial temperature distribution), meeting the core requirements of one-dimensional model spatial description. Although the computational load is higher than ignoring the global space, the impact on spatial parameter deviation is minimal, making it a fundamental rule for second-level accuracy. A segmented dynamic heat loss rule is added, setting the heat loss rate according to the burner's axial segments (e.g., setting separate rates for the inlet, combustion, and outlet segments). This accurately reflects the heat dissipation differences in different regions, avoiding axial temperature calculation distortion caused by constant heat loss, and adapting to the structural design's requirements for thermal boundary accuracy. A rule of retaining core components and eliminating trace impurity components is selected. Core components such as CH4, H2, OH, and NOx are retained, while trace impurity components are eliminated, reducing computational load and balancing efficiency and accuracy. High-deviation rules are eliminated, such as those excluding constant-volume combustion assumptions and ignoring flow coupling. Deviations of the selected assumption rules are accumulated, and after meeting the second-level accuracy requirements, the selected assumption rules are used as the assumption rules for the one-dimensional model.

[0056] (4) Simplify the basic model according to the assumption rules to generate the zero-dimensional model and the one-dimensional model. The number of assumption rules in the zero-dimensional model is less than that in the one-dimensional model.

[0057] Specifically, the basic model is simplified based on the corresponding matching assumption rules of the zero-dimensional model and the one-dimensional model. The assumption rules based on the zero-dimensional model simplify the basic model by deleting spatial dimension-related equations (such as axial transport equations), retaining the time-domain mass and energy conservation equations, embedding the simplified chemical reaction mechanism (only retaining the core elementary reactions), and outputting macroscopic combustion parameters without spatial distribution. The assumption rules based on the one-dimensional model simplify the basic model by retaining the axial momentum, energy, and component transport equations, coupling flow and chemical reaction, and outputting parameters such as temperature and component concentration distributed along the axis. Furthermore, the zero-dimensional model has fewer assumption rules than the one-dimensional model. Since the one-dimensional model needs to retain axial spatially related rules to describe structural adaptability, it has a higher computational load but better accuracy. The two can form a complementary relationship of efficiency priority versus accuracy priority.

[0058] Furthermore, assumption rules can be matched based on the combustion participants or combustion stages. Combustion participants include fuel, oxidant, intermediate products, combustion products, and burner structure; rules need to be set according to the model's accuracy requirements. The zero-dimensional model (Level 1 accuracy) prioritizes efficiency, employing a homogeneous mixing assumption for fuel and oxidant, ignoring differences in the mixing process, assuming complete mixing at the initial moment and uniform component concentrations across the entire domain, eliminating the need to calculate hybrid dynamics to reduce computational load; for intermediate products, it adopts a rule of retaining core intermediate products and merging minor products, tracking only H, O, and OH. Key free radicals are addressed by treating trace amounts of minor intermediate products as quasi-steady-state approximations, reducing the number of component equations to be solved. For the burner structure, a wall heat loss equivalence rule is adopted, transforming wall heat dissipation into a globally uniform heat loss coefficient, ignoring local heat dissipation differences between materials and structures, and simplifying thermal boundary calculations. The one-dimensional model (second-order accuracy) prioritizes accuracy while retaining spatial descriptive capabilities. An "axial stratified mixing assumption" is adopted for fuel and oxidant, describing the mixing gradient of component concentrations along the axial direction through axial transport equations, adapting to the actual mixing process within the flow channel to ensure spatial parameter accuracy. For intermediate products, a full-spectrum intermediate product tracking and quasi-steady-state treatment rule is adopted, retaining the dynamic concentration calculation of all intermediate products in the flame core region and using a quasi-steady-state approximation in the downstream reaction plateau region, balancing accuracy and computational load. For the burner structure, a "wall characteristic segmented mapping" rule is adopted, inputting parameters such as thermal conductivity and roughness of the wall material in corresponding regions along the axial direction, calculating local heat loss through heat conduction equations, reflecting the spatial influence of structural differences on the temperature field.

[0059] Furthermore, the combustion process includes ignition, flame propagation, heat release, pollutant generation, and flow coupling, requiring rules to be set based on the model's functional positioning. Zero-dimensional models, lacking spatial dimensions and focusing on rapid macroscopic screening, employ a uniform ignition assumption in the ignition stage, igniting differences in ignition source locations and assuming that ignition conditions are met simultaneously across the entire domain, simplifying ignition dynamics with a single ignition delay time. The flame propagation stage uses an equivalent rule for overall combustion rate, replacing the flame surface spatial propagation process with the average reaction rate across the entire domain, indirectly reflecting the intensity of combustion through the peak heat release rate. The pollutant generation stage uses a simplified calculation rule for total generation, ignoring NOx. The generated spatial distribution, calculated solely based on global temperature and residence time, meets the emission assessment requirements for basic operating conditions. The one-dimensional model must support burner structural design and accurately describe spatial correlation processes. The ignition stage employs axial ignition source positioning rules, setting specific ignition locations and simulating ignition through local temperature jumps, tracking the flame surface's axial propagation trajectory to ensure accurate spatial parameters such as flame length and ignition delay distance. The flame propagation stage uses laminar flame velocity coupled with axial flow rules, combining axial flow velocity and local flame propagation velocity to calculate changes in the axial position of the flame surface, reflecting the interaction between flow and combustion to support inlet velocity design. The pollutant generation stage uses a zoned generation model rule, calculating thermal NOx in the high-temperature zone and fuel-rich NOx in the fuel-rich zone, outputting the NOx concentration distribution curve along the axial direction to provide a basis for denitrification structure design.

[0060] Furthermore, the zero-dimensional model's global gradient-free rule can be extended to time-domain single-node computation, where all parameters change only with time and have no spatial coordinates; the flow-ignoring rule can be extended to isochoric / isobaric simplification, selecting assumptions based on burner type to avoid solving momentum equations; the simplified mechanism rule can be extended to core reaction path locking, retaining only the core reaction of H2 / CH4 oxidation. The one-dimensional model's axial gradient retention rule can be extended to axial multi-node discretization, dividing the burner into dozens of computational nodes, with each node independently solving for parameters; the flow coupling rule can be extended to axial momentum conservation plus frictional drag correction, considering the influence of wall friction on flow velocity to approximate reality; the detailed mechanism tailoring rule can be extended to spatially differentiated mechanisms, using detailed mechanisms in the flame zone and simplified mechanisms downstream; in terms of rule interrelationships, the one-dimensional model rules are an upgrade in precision from the zero-dimensional model rules, sharing the basic chemical reaction mechanism framework and macroscopic conservation equations. The homogeneous mixing rule of the zero-dimensional model is a simplified special case of the one-dimensional model's axial stratified mixing rule (when the axial mixing gradient approaches zero). (One-dimensional model degenerates into zero-dimensional model); It is complementary. The zero-dimensional model rules cover a large number of basic operating conditions through efficient calculation. The output target combustion parameters provide focused input for the one-dimensional model. The one-dimensional model rules obtain burner configuration parameters through accurate spatial analysis, which feed back into the parameter adjustment of the three simulations of the zero-dimensional model; It is also constraining. The macroscopic parameter deviation of the zero-dimensional model needs to be controlled within the acceptable range of the one-dimensional model to avoid the one-dimensional model from conducting analysis based on incorrect parameters. The spatial parameter accuracy of the one-dimensional model needs to verify the rationality of the zero-dimensional model. For example, the peak value of the heat release rate predicted by the zero-dimensional model needs to be consistent with the average value of the axial peak value of the one-dimensional model to ensure the logical coherence and reliable results of the entire simulation process.

[0061] S103. Perform a simulation of the hydrogen-blended natural gas combustion process using the zero-dimensional model. Based on the results of the simulation, select multiple target combustion parameters from the combustion parameter set. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model.

[0062] Specifically, the combustion parameter set includes different hydrogen doping ratios, different stoichiometric ratios, and different initial temperatures. For each set of parameters, a zero-dimensional model is set as input (e.g., 20% hydrogen doping corresponds to a volume ratio of 80% CH4 + 20% H2, and a stoichiometric ratio of 0.8 corresponds to an oxidant excess of 25%). The zero-dimensional model is run to solve the conservation equations and output combustion characteristic parameters such as ignition delay time, adiabatic flame temperature, and total NOx formation. A scoring standard is set to score each set of parameters, and the top 10 parameters are selected as target combustion parameters.

[0063] By leveraging the efficiency of zero-dimensional models, multiple sets of basic parameters can be quickly selected, parameters that do not meet the requirements of the working conditions can be eliminated, and subsequent simulations can focus on the core target parameters, significantly reducing the number of working conditions in the secondary simulation of one-dimensional models and improving the overall simulation efficiency.

[0064] Furthermore, the specific steps for performing a single simulation of the hydrogen-blended natural gas combustion process using the zero-dimensional model, and selecting multiple target combustion parameters from the combustion parameter set based on the results of the single simulation, include:

[0065] (1) Based on multiple combustion parameters in the combustion parameter set, set the input conditions for the zero-dimensional model;

[0066] Specifically, the input conditions for the zero-dimensional model are set based on multiple combustion parameters from the combustion parameter set. This set covers key control parameters for hydrogen-blended natural gas combustion, including the hydrogen blending ratio (the volume percentage of hydrogen in the fuel mixture), stoichiometric ratio (the ratio of the actual fuel / oxidizer ratio to the theoretical complete combustion ratio), initial temperature (the initial temperature after the fuel and air mixture), and initial pressure (the initial operating pressure of the combustion system). For each combination of combustion parameters (such as a specific combination of hydrogen blending ratio and stoichiometric ratio), it needs to be converted into an input form recognizable by the zero-dimensional model. For example, the hydrogen blending ratio directly determines the component concentration ratio of methane and hydrogen in the model, the stoichiometric ratio corresponds to the calculation of the oxidizer (air) supply, and the initial temperature and pressure are used as thermodynamic boundary condition inputs to ensure that the model accurately reflects the initial combustion environment under these parameters.

[0067] (2) Input the input conditions into the zero-dimensional model and run the zero-dimensional model to perform simulation calculations on the combustion process of hydrogen-blended natural gas under each set of combustion parameters, and obtain the output combustion characteristic parameters;

[0068] Specifically, the input conditions are imported into the zero-dimensional model for simulation. The combustion process of hydrogen-blended natural gas is calculated for each set of combustion parameters, and the corresponding combustion characteristic parameters are output. The zero-dimensional model, based on the homogeneous mixing assumption and a simplified chemical reaction mechanism, solves the mass and energy conservation equations in the time domain. The calculation process covers the entire cycle from fuel and oxidant mixing and ignition to the completion of the combustion reaction. The output combustion characteristic parameters focus on macroscopic combustion performance, including ignition delay time (the time interval from initial conditions to the violent occurrence of the combustion reaction), adiabatic flame temperature (the theoretical maximum combustion temperature under no heat loss conditions), peak heat release rate (the maximum heat released per unit time during combustion), and total pollutant formation (such as the total emissions of NOx and CO). These parameters directly reflect the combustion stability, energy efficiency, and emission levels under the given set of combustion parameters. For details on the specific simulation process of the zero-dimensional model, please refer to the descriptions in relevant technical documents; they will not be elaborated here.

[0069] (3) Scoring the combustion characteristic parameters corresponding to each set of combustion parameters, and filtering multiple target combustion parameters based on the scores.

[0070] Specifically, the combustion characteristic parameters corresponding to each set of combustion parameters are scored, and multiple target combustion parameters are selected based on the scores. The scoring needs to be based on multi-dimensional evaluation standards set according to actual operating conditions. For example, the ignition delay time needs to be controlled within a range that ensures ignition reliability without being too short to cause detonation; the adiabatic flame temperature needs to match the burner material's tolerance limit; the peak heat release rate needs to meet combustion efficiency requirements; and the total amount of pollutants generated needs to comply with emission regulations. Weights are assigned to each indicator (e.g., emission indicators have a higher weight than efficiency indicators). The characteristic parameters of each set of parameters are compared with the standards and quantified and scored. Finally, the parameter combinations with the highest scores are selected as target combustion parameters. These target parameters must simultaneously meet multiple core indicators (e.g., reasonable ignition delay time, compliant pollutant emissions, and flame temperature within the material's tolerance range) to provide focused input parameters for subsequent secondary simulations of the one-dimensional model, reducing unnecessary calculations and improving overall simulation efficiency.

[0071] The core function of the zero-dimensional model lies in achieving efficient computation based on the assumption of ignoring spatial gradients. It can quickly cover multiple combinations of parameters such as hydrogen doping ratio and stoichiometry in the combustion parameter set. Through a single simulation, it can screen out target combustion parameters with basic combustion feasibility, significantly reducing the amount of unnecessary calculations in subsequent high-precision simulations. At the same time, by simplifying the modeling process, it avoids the work of repeatedly building complex models, laying an efficient screening foundation for the overall simulation process. The core function of the one-dimensional model is to achieve higher precision computation based on the assumption of retaining the axial spatial dimension. It focuses on the adaptability of the burner configuration to the combustion scenario. Through a second simulation, it outputs spatial characteristic parameters such as flame length and axial temperature gradient, accurately determining the optimal burner configuration that matches the target combustion parameters. This provides a quantitative basis for burner structural design, and its accuracy advantage can make up for the shortcomings of the zero-dimensional model in spatial parameter description.

[0072] Furthermore, the target combustion parameters selected by the zero-dimensional model provide focused simulation input for the one-dimensional model. This eliminates the need for the one-dimensional model to perform calculations on all parameters in the combustion parameter set, allowing it to focus solely on configuration matching simulations for the core target parameters. This significantly reduces the computational burden on the one-dimensional model, enabling it to focus more efficiently on configuration optimization and improving the solution efficiency for the optimal burner configuration. The optimal burner configuration solved by the one-dimensional model can then feed back into the parameter adjustment of the zero-dimensional model, replacing the simplified burner structural parameters in the zero-dimensional model with the actual optimal configuration parameters. Simultaneously, the heat loss assumptions of the zero-dimensional model are corrected by incorporating the wall heat dissipation characteristics of the optimal configuration, freeing the zero-dimensional model from the limitations of ideal assumptions and significantly improving simulation accuracy. This provides support for outputting high-precision dynamic characteristic parameters in the subsequent three simulations. In this way, the efficient selection of the zero-dimensional model reduces the burden on the one-dimensional model, while the accurate optimization of the one-dimensional model improves the efficiency of the zero-dimensional model. This avoids the problem of either low efficiency or insufficient accuracy of a single model, and through mutual collaboration, the overall simulation process achieves both high efficiency and high accuracy.

[0073] S104. Determine the simulation input parameters based on the multiple target combustion parameters, perform a secondary simulation of the hydrogen-blended natural gas combustion process based on the simulation input parameters and the one-dimensional model, and determine the optimal burner configuration based on the results of the secondary simulation.

[0074] Specifically, the target combustion parameters are converted into one-dimensional model inputs, and the burner configuration parameters are supplemented simultaneously. The mixing mode is determined according to the target parameters (e.g., the premixing mode corresponds to the premixing one-dimensional sub-model, and the diffusion mode corresponds to the diffusion sub-model). Different sub-models output different flame spatial characteristic parameters (the premixing model outputs the flame propagation rate, and the diffusion model outputs the flame length). For each set of target combustion parameters and configuration parameters, the simulation outputs the flame surface position, axial temperature gradient, NOx concentration peak, etc., and the judgment criteria are set (flame length ≤ 80% of the combustion chamber, NOx ≤ 150ppm, flame surface centered) to select the optimal configuration (e.g., axial length 1.5m, cross-sectional diameter 80mm, suitable for 20% hydrogen doping).

[0075] Using target combustion parameters as input, the spatial characteristics simulation of a one-dimensional model accurately correlates the matching relationship between combustion parameters and burner structure, avoiding the cost waste of traditional trial-and-error structural design. At the same time, by matching different hybrid scenarios through sub-models, it ensures that the optimal configuration is adapted to the actual combustion mode and improves the engineering applicability of the configuration.

[0076] Furthermore, the specific implementation steps include:

[0077] (1) The multiple target combustion parameters are used as input parameters for multiple simulations. The combustion scenario is determined based on the target combustion parameters of each simulation. The combustion scenario is used to describe the mixing process of combustion gases under the target combustion parameters.

[0078] Specifically, the selected target combustion parameters are used as input parameters for multiple simulations. For each set of target combustion parameters, a corresponding combustion scenario is defined. The combustion scenario needs to describe in detail the mixing mode of fuel and oxidizer (e.g., premixed combustion, where fuel and air are fully mixed before entering the combustion chamber; diffusion combustion, where fuel and air are mixed and burned in the combustion zone; or partial premixed combustion, where some fuel is mixed in advance and the rest diffuses in the combustion zone), mixing uniformity (e.g., high uniformity corresponds to the premixed scenario, and stratified mixing corresponds to the diffusion scenario), and initial flow state (e.g., laminar or turbulent inlet conditions). These characteristics directly determine the spatial evolution law of the combustion process.

[0079] (2) Based on the target one-dimensional sub-model matching the combustion scenario, the target combustion parameters are used as the input of the target one-dimensional sub-model. Different burner configuration parameters are input to simulate the combustion process, and flame space feature parameters corresponding to the target combustion parameters and the burner configuration parameters are obtained.

[0080] Optionally, each of the target one-dimensional sub-models corresponds to a different hydrogen-doped natural gas and air mixing mode, the target one-dimensional sub-model is used to describe the changes in the combustion flame, and the flame spatial characteristic parameter type of each target one-dimensional sub-model is different.

[0081] Specifically, flame spatial characteristic parameters are quantitative indicators centered around axial spatial distribution and core flame characteristics. They directly reflect the spatial evolution of the combustion process and the adaptability of the burner. These parameters include flame morphology and position parameters (flame length, axial position of the flame face, and flame initiation position), temperature and heat distribution parameters (axial temperature distribution curve, temperature gradient, and axial distribution of heat release rate), component and pollutant distribution parameters (axial distribution of component concentration, axial distribution of NOx concentration, and axial distribution of CO concentration), and flow and stability related parameters (axial velocity distribution, flame elongation, and flame stability coefficient). Together, these parameters constitute a spatial portrait of the combustion state, directly serving the determination of the optimal burner configuration. For example, by adjusting the axial dimensions of the burner through flame length, optimizing the cooling structure through NOx distribution, and matching the inlet design through velocity distribution, a configuration scheme with high combustion efficiency, low pollutant content, and good stability can ultimately be achieved.

[0082] Furthermore, the sub-models of the one-dimensional model can include a one-dimensional diffusion flame model, a counter-flushing flame model, and a one-dimensional premixed / semi-mixed model. The one-dimensional diffusion flame model is designed for diffusion combustion scenarios where fuel and oxidizer mix and react simultaneously in the combustion zone. Its core assumption is that fuel and oxidizer flow in axial stratification without premixing. The diffusion rates of CH4, H2, and air (O2, N2) are calculated using axial component transport equations. Combined with chemical reaction kinetics, the diffusion flame surface formed in the axial direction is simulated. The flame surface position is determined by the diffusion rate ratio of fuel to oxidizer, and the combustion reaction only occurs in a thin region near the flame surface. When solving the model, the axial inlet velocity of fuel and oxidizer, initial concentration (such as hydrogen doping ratio), axial segment size of the burner, and wall heat loss parameters must be input. The outputs are flame length (axial distance from the fuel inlet to the end of the flame surface), axial temperature gradient (highest temperature at the flame surface, gradually decreasing towards both sides), and NOx concentration distribution along the axial direction (NOx concentration in the high-temperature flame surface region). Spatial characteristic parameters such as the highest generation rate are used to adapt the configuration design of hydrogen-blended natural gas diffusion burners, such as jet burners in industrial kilns and some household gas stoves. In these scenarios, fuel and air usually enter the combustion chamber through different channels without a premixing process. It is necessary to optimize configuration parameters such as the fuel injection angle at the burner outlet and the air inlet position through modeling to avoid the flame being too long and touching the wall (leading to local overheating) or too short (incomplete combustion), while controlling the local high-temperature NOx that is easily generated in diffusion combustion.

[0083] The counter-current flame model simulates the flame stretching effect during combustion by assuming two opposing airflows. Its core structure involves fuel and oxidizer airflows being injected axially in opposite directions, forming a stable flame surface at their intersection. This model requires coupling flow and chemical reaction equations, focusing on calculating the stretching ratio (the degree of flame surface deformation caused by the airflow velocity gradient) at the flame surface, and its impact on flame propagation speed and ignition delay time. When the stretching ratio exceeds a critical value, the flame will extinguish due to excessive deformation. The model can output this critical stretching ratio as a flame stability index. Input parameters, in addition to the hydrogen doping ratio and stoichiometric ratio, include the inlet velocities, pressures, and counter-current distances (axial spacing between the inlets of the two airflows). Output parameters include the axial position of the flame surface, flame propagation speed at different stretching ratios, the critical stretching ratio for extinguishing the flame, and the axial temperature / ... Component concentration distribution; used to evaluate the flame stability of hydrogen-blended natural gas combustion, especially suitable for high-flow-rate combustion scenarios (such as gas turbine combustors and aero-engine combustors). In these scenarios, high airflow velocities can easily cause flame stretching. It is necessary to determine the maximum allowable inlet velocity of the combustor (based on the critical stretching rate of flameout) under a certain hydrogen blending ratio through modeling, or to optimize the airflow guiding structure of the combustor (such as adding guide vanes) to reduce the impact of stretching effect on flame stability and avoid flameout or combustion oscillation.

[0084] The one-dimensional premixed / semi-mixed model is further divided into two subtypes: premixed and semi-mixed, to adapt to combustion scenarios with different degrees of mixing. The one-dimensional premixed model assumes that the fuel (hydrogen-blended natural gas) and oxidant are completely mixed before entering the burner. The model ignores the mixing process and only calculates the combustion evolution of the premixed gas along the axial direction. By solving the axial energy equation and component equation, it simulates the flame propagation, temperature rise, and component consumption process of the premixed gas from the inlet to the outlet. The core outputs are flame propagation speed (the speed at which the flame moves along the axial direction in the premixed gas), flame length (the axial distance required for complete combustion of the premixed gas), and axial temperature uniformity parameters. The one-dimensional semi-mixed model is for semi-premixed scenarios where some fuel is premixed with air and the remaining fuel diffuses and supplements in the combustion zone. The model combines the calculation logic of premixing and diffusion. The first axial section is calculated in premixing mode (combustion of the mixed part), and the second axial section is calculated in diffusion mode (supplementary combustion of the unmixed part). The input is the premixing ratio (the proportion of premixed fuel), and the outputs are the segmented flame surface position, the heat release rate of each segment, and the amount of pollutants generated. The premixed sub-model is suitable for fully premixed hydrogen-blended burner designs, such as premixed burners in residential wall-hung boilers and industrial low-NOx burners. In these scenarios, the model needs to optimize the burner's axial length (to match the premixed gas flame length) or adjust the premixed gas inlet velocity (to match the flame propagation speed and avoid backfire or flameout) to achieve low NOx combustion (uniform premixed combustion temperature distribution, reducing localized high temperatures). The semi-mixed sub-model is suitable for semi-premixed burner designs, such as some gas water heaters and commercial cooktops. In these scenarios, full premixing is prone to backfire due to velocity fluctuations, requiring the retention of some diffusion combustion to supplement stability. The model can be used to optimize the premixing ratio and burner segment structure (such as the axial length distribution of the front premixed combustion zone and the rear diffusion supplementation zone) to balance combustion efficiency and flame stability.

[0085] By selecting the corresponding target one-dimensional sub-model based on the actual combustion scenario, the flame spatial feature parameters corresponding to the actual combustion scenario are output.

[0086] (3) Determine the optimal burner configuration based on the flame space characteristic parameters.

[0087] Specifically, considering the engineering requirements of hydrogen-blended natural gas combustion, a quantitative evaluation system is established by setting judgment criteria for flame spatial characteristic parameters from five dimensions: flame morphology adaptability, temperature safety, combustion completeness, emission compliance, and stability. For each group of burner configuration parameters (such as axial length, cross-sectional diameter, inlet velocity, wall material, and secondary air location), the flame spatial characteristic parameters are checked one by one. If a configuration fails to meet the criteria, it is directly eliminated to avoid subsequent invalid evaluations. For configurations that pass the initial screening, weights are assigned to each judgment dimension, and scores are calculated based on the closeness of the characteristic parameters to the standards. The configuration with the highest comprehensive score is selected as the candidate optimal solution. If multiple configurations have similar scores, economic indicators (such as manufacturing cost and operating energy consumption) and engineering feasibility (such as whether the cross-sectional diameter is compatible with existing pipe sizes and whether the inlet velocity is easily achieved) are further compared to determine the unique optimal configuration. Meanwhile, for dimensions with lower scores, the direction of configuration optimization is deduced in reverse. For example, if a certain configuration causes "excessive outlet temperature due to the flame face position being too far back", the axial length of the burner can be shortened or the fuel inlet position can be adjusted to move the flame face forward.

[0088] Determining the optimal burner configuration by using flame space characteristic parameters avoids the high cost of trial-and-error physical experiments. By directly linking the configuration with combustion effects through quantitative parameters, the optimization direction can be accurately identified. This ensures that the configuration is adapted to the combustion characteristics of hydrogen-blended natural gas (such as high flame propagation speed and low ignition energy), avoiding problems such as poor stability and excessive emissions caused by traditional configurations being unsuitable for hydrogen blending. It also provides real configuration parameters (such as wall heat loss and operating pressure) for subsequent zero-dimensional model adjustments (three simulations), making the calculation of simplified mechanism reference values ​​closer to the actual combustion environment.

[0089] S105. Based on the optimal burner configuration, adjust the zero-dimensional model and use the adjusted zero-dimensional model to perform three simulations of the hydrogen-blended natural gas combustion process.

[0090] Specifically, based on the optimal configuration, the zero-dimensional model is modified, and the burner model corresponding to the zero-dimensional model simulation is used. The uniform heat loss over the entire domain is changed to a dynamic heat loss model. The initial pressure is fixed as the working pressure of the optimal configuration, and the fuel component ratio is calibrated as the target parameter. The simulation is repeated 3 times for the key range of the optimal configuration (hydrogen doping 18%-22%, stoichiometry 0.7-0.8), and the modified conservation equation is solved to output dynamic combustion parameters (ignition delay time 8-10ms), reaction mechanism parameters (peak OH radical concentration), and energy characteristic parameters (heat release rate fluctuation ≤5%).

[0091] By feeding back configuration parameters to adjust the zero-dimensional model, the deviation between the ideal assumptions of the original zero-dimensional model and the actual configuration is eliminated (such as the overestimation of temperature due to neglecting wall heat dissipation), making the results of the three simulations closer to the actual combustion process; repeated simulations ensure data stability and provide high-precision reaction kinetic data for subsequent mechanism simplification.

[0092] Furthermore, based on the optimal burner configuration, the zero-dimensional model was adjusted, and three simulations of the hydrogen-blended natural gas combustion process were performed using the adjusted zero-dimensional model; including:

[0093] (1) Adjust the burner configuration parameters in the zero-dimensional model based on the optimal burner configuration;

[0094] Specifically, based on the physical characteristics of the optimal burner configuration, the basic input parameters of the zero-dimensional model are corrected to eliminate the deviation between the initial simplification and the actual structure. The key geometric parameters of the optimal configuration (such as the total axial length, cross-sectional diameter, and inlet channel size) are transformed into equivalent parameters of the zero-dimensional model. For example, the reaction domain volume of the model is adjusted according to the actual burner volume, and the equivalent flow rate of the initial mixed gas is corrected according to the inlet cross-sectional size (although the zero-dimensional model has no spatial dimension, the reaction residence time needs to be correlated with the volume and flow rate). At the same time, the material properties of the optimal configuration (such as wall thermal conductivity and high temperature resistance limit) are input into the model as the basic data for subsequent heat loss correction. Combined with the working environment of the optimal configuration (such as the actual operating pressure and the inlet gas premixing method), the initial pressure setting and mixing assumption boundary of the zero-dimensional model are adjusted. For example, if the optimal configuration is a low-pressure burner (such as a household stove), the initial pressure of the model needs to be corrected from the default high pressure value to the actual working pressure. If the optimal configuration adopts a partial premixing method, the premixing ratio coefficient needs to be adjusted in the model to make the initial component distribution closer to the actual mixing state.

[0095] (2) Based on the wall heat dissipation characteristics of the optimal burner configuration, the global uniform heat loss rate of the zero-dimensional model is modified into a dynamic heat loss model related to the wall temperature.

[0096] Specifically, replacing the original zero-dimensional model's assumption of a uniform global heat loss rate, a dynamic heat loss calculation logic is established based on the wall heat dissipation characteristics of the optimal burner configuration. The wall temperature at different combustion stages is calculated using the material's thermal conductivity and axial temperature distribution (taken from the one-dimensional model output): In the initial combustion stage (low temperature stage), the wall temperature is close to ambient temperature, and heat loss is mainly convective; in the middle combustion stage (high temperature stage), the wall temperature increases with flame heating, and the proportion of radiative heat dissipation increases; in the final combustion stage (temperature decrease stage), the wall temperature decreases slowly, and the heat loss rate decreases with decreasing temperature difference. The heat loss rate is changed from a fixed value to a function related to the wall temperature and flame temperature difference. For example, a combined formula of convective and radiative heat dissipation is used, where the convective term is proportional to (flame temperature - wall temperature), and the radiative term is proportional to (flame temperature - wall temperature). 4 - Wall temperature 4 The coefficient is directly proportional to the wall roughness and surface area of ​​the optimal configuration. This model enables the zero-dimensional model to reflect the heat dissipation changes of the burner wall from a cold state to a hot state and then to a steady state, avoiding temperature calculation errors caused by a constant heat loss rate.

[0097] (3) Based on the adjusted zero-dimensional model, three simulations of the hydrogen-blended natural gas combustion process were performed to solve the corrected mass conservation and energy conservation equations and output dynamic characteristic parameters, including dynamic combustion parameters, reaction mechanism parameters and energy characteristic parameters.

[0098] Specifically, based on the adjusted zero-dimensional model, the dynamic evolution of hydrogen-blended natural gas combustion is simulated by solving the modified conservation equations, outputting dynamic characteristic parameters. These dynamic combustion parameters include the ignition delay time (a more precise ignition timing after correcting for heat loss), the heat release rate curve (reflecting the dynamic fluctuations in combustion intensity, such as whether multiple exothermic peaks occur), the combustion duration (the time from ignition to the point where the fuel is almost completely consumed), and the concentration-time curves of key components (such as the consumption rates of H2 and CH4, and the generation and destruction patterns of OH radicals). These parameters can reveal the dynamic stability of the combustion process (such as whether there is a flameout and re-ignition phenomenon). The reaction mechanism parameters focus on the dynamic changes of the chemical reaction pathway, outputting the rate constant time series of the core elementary reactions (such as H+O2→OH+O). The data support for simplifying and optimizing chemical reaction mechanisms includes the reaction rate as a function of temperature, the sensitivity coefficient of key free radicals (the degree to which a change in the concentration of a certain free radical affects the overall reaction rate), and the proportion of reaction pathways under different hydrogen doping ratios (such as the evolution of the competitive relationship between hydrogen oxidation and methane oxidation over time). Energy characteristic parameters cover the dynamic temperature curve of the combustion process (a more realistic temperature rise and fall trend after correcting for heat loss), pressure changes over time (pressure peak and pressure rise rate under constant volume combustion, or pressure fluctuation under constant pressure combustion), and energy conversion efficiency (the change of the ratio of effective heat output to fuel chemical energy over time). These parameters are directly related to the actual energy efficiency and operational safety of the burner (such as whether there is an overpressure risk).

[0099] Three simulations were conducted to correct the model by feeding back the optimal configuration. On the one hand, the introduction of the dynamic heat loss model and configuration parameters freed the zero-dimensional model from the limitations of pure theoretical simplification, and the output dynamic parameters were closer to the actual operating state of the burner. On the other hand, the combination of reaction mechanism parameters and dynamic combustion parameters can provide a precise basis for the subsequent dynamic control strategy of the burner (such as adjusting the fuel supply according to the fluctuation of heat release rate) and the engineering tailoring of chemical reaction mechanism, ultimately improving the scientific nature of the entire process from design to operation of the hydrogen-blended natural gas combustion system.

[0100] S106. Based on the results of the three simulations, the computer simplifies the reference values.

[0101] Specifically, the time histories of component concentrations (such as the changes in H and OH radical concentrations over time), temperature, and heat release rate from three simulations are extracted. The sensitivity coefficient is calculated using perturbation analysis (the minimum cumulative percentage of 90% of the basic reaction rate constant is taken as the critical value for perturbation of ±5%), the contribution threshold is calculated using reaction flow analysis (the minimum cumulative percentage of 95%), and the relative change rate is calculated as the quasi-steady-state judgment value (1% of the maximum change rate in the steady-state stage). All three are used together as a simplified reference value for the mechanism.

[0102] Based on high-precision simulation data under optimal configuration, the calculated reference values ​​are closer to the actual combustion mechanism, avoiding the simplification deviation caused by traditional reference values ​​calculated based on ideal operating conditions. This provides a quantitative basis for the accurate simplification of the GRI3.0 mechanism, ensuring that the simplified mechanism retains the core reaction while reducing redundant calculations.

[0103] Furthermore, based on the results of the three simulations, the specific steps for simplifying the reference values ​​by computer include:

[0104] (1) Obtain the dynamic characteristic parameters of the three simulation outputs, including component concentration time history data, temperature time history data and heat release rate time history data;

[0105] Specifically, core dynamic characteristic parameters are extracted from the outputs of the three simulations. These include time-history data of component concentrations, covering reactants and products such as CH4, H2, O2, CO2, and H2O, as well as the concentration-time variation curves of key free radicals such as OH, H, and O; temperature time-history data, representing the temperature rise and fall curves throughout the combustion process, reflecting the dynamic changes from the initial temperature to the peak flame temperature and then to the cooling phase; and heat release rate time-history data, representing the fluctuation curve of heat released per unit time over time, reflecting the dynamic characteristics of combustion intensity. These data are preprocessed to remove abnormal fluctuations (such as numerical jumps) and aligned by time step to ensure consistency in the time dimension of subsequent analyses. For example, all time-history data are interpolated to the same time interval (e.g., 0.1 ms / step) to facilitate cross-parameter correlation analysis.

[0106] (2) Based on the time history data of the component concentration, the sensitivity coefficient of each elementary reaction to the specified component concentration in the detailed mechanism of GRI3.0 is calculated by perturbation analysis, the absolute value distribution of the sensitivity coefficients of all elementary reactions is statistically analyzed, and the sensitivity screening value is determined according to the absolute value distribution.

[0107] Specifically, based on the time history data of component concentrations, the influence of each elementary reaction on the concentration of a specified component in the detailed mechanism of GRI3.0 is quantified using perturbation analysis. This is done for each elementary reaction (e.g., H+O2⇌). The reaction rate constant of OH+O is slightly perturbed, and the zero-dimensional model is rerun to calculate the time-history changes in the concentration of a specified component (such as OH radicals that determine ignition delay and O atoms that affect NOx formation). By comparing the concentration deviations before and after the perturbation, the sensitivity coefficient of the elementary reaction to the specified component is calculated. The sensitivity coefficient = relative change in concentration / relative perturbation of the reaction rate constant. The absolute value distribution of the sensitivity coefficients of all elementary reactions is statistically analyzed to determine their numerical patterns. Generally, most reactions have a weak effect on the component concentration, with small absolute values ​​of sensitivity coefficients, while a few reactions are core influencing reactions with significantly higher absolute values. Based on the distribution characteristics, a sensitivity screening value is determined, and the 90th percentile of the absolute value distribution of the sensitivity coefficient is selected as the threshold. That is, 90% of the elementary reactions have absolute values ​​of sensitivity coefficients below this value. Reactions exceeding this value are judged to have a significant effect on the concentration of the specified component and should be retained in the mechanism simplification.

[0108] (3) Based on the time history data of the component concentration, calculate the net reaction flow contribution of each elementary reaction to the target component through reaction flow analysis, and determine the reaction flow screening value based on the net reaction flow contribution;

[0109] Specifically, reaction flow analysis quantifies the contribution of each elementary reaction to the formation and consumption of target components (such as key combustion products CO2 and pollutant NOx). Based on the time history data of component concentration, the net reaction flow of each elementary reaction at different times is calculated, i.e., the difference between the formation rate and the consumption rate. The total net reaction flow of the entire combustion process is accumulated to obtain the total contribution of each elementary reaction to the target component (positive values ​​indicate promotion of formation, and negative values ​​indicate promotion of consumption). The net reaction flow contribution of most elementary reactions is close to zero, i.e., the difference from zero is less than a preset threshold, indicating that they have no substantial effect on the formation or consumption of the target component. Only a few reactions have a dominant contribution (such as the contribution of key steps in the methane oxidation chain to CO2). The minimum contribution of the reactions with the highest absolute value of total contribution is selected as the threshold. That is, reactions with a value lower than this value have a negligible impact on the net formation / consumption of the target component and can be eliminated in mechanism simplification.

[0110] (4) Based on the time history data of the component concentration, extract the change curve of the concentration of each intermediate product over time, calculate the ratio of the concentration change rate to the concentration itself at each time point as the relative change rate, and determine the quasi-steady-state component judgment value based on the relative change rate.

[0111] Specifically, for intermediate products (such as non-core free radicals or intermediate species like C2H6 and CH3) in the time-history data of component concentrations, their concentration change curves over time are extracted. The ratio of the concentration change rate (first derivative) at each moment to the concentration itself at that moment is calculated to obtain the relative change rate (a dimensionless parameter reflecting the intensity of concentration change). The relative change rate of intermediate products typically exhibits rapid fluctuations in the early stages of the reaction, a gradual plateauing in the middle stages, and near-zero values ​​at the end. The core of the quasi-steady-state assumption is that when the relative change rate is sufficiently small, the concentration of intermediate products can be approximated as a constant (solved by replacing differential equations with algebraic equations). Therefore, a judgment value is determined based on the distribution of the relative change rate: the maximum absolute value of the relative change rate during the entire combustion process is selected as a reference, and 1 / 10 (or a smaller proportion) of it is taken as a threshold. When the absolute value of the relative change rate of a certain intermediate product is consistently lower than this threshold, it can be determined as a quasi-steady-state component, and the quasi-steady-state approximation is used in the mechanism simplification.

[0112] (5) The sensitivity screening value, the reaction flow screening value and the quasi-steady-state component determination value are used together as the simplified reference value of the mechanism.

[0113] Specifically, sensitivity screening values, reaction flow screening values, and quasi-steady-state component determination values ​​are integrated into a complete mechanism simplification reference system. Sensitivity screening values ​​are used to retain elementary reactions that have a significant impact on the concentration of key components, reaction flow screening values ​​are used to eliminate reactions that contribute little to the net generation / consumption of target components, and quasi-steady-state component determination values ​​are used to identify intermediate products that can be approximated. The three work together to ensure that the simplified mechanism retains the core reaction pathways and dynamic characteristics of key components, while significantly reducing the number of reactions and equations to be solved. This achieves the goal of mechanism simplification with controllable accuracy loss and a significant reduction in computational load, providing efficient and reliable chemical reaction model support for the engineering application of hydrogen-blended natural gas combustion simulation.

[0114] S107. Based on the simplified reference value of the mechanism, simplify the GRI3.0 mechanism to generate a mechanistic description of the hydrogen-blended natural gas combustion process.

[0115] Specifically, by employing a three-tiered screening logic—retaining core reactions, eliminating secondary reactions, and simplifying intermediate product processing—the scale of the mechanism is significantly reduced while ensuring the accuracy of key combustion characteristics (ignition delay, pollutant generation, and heat release dynamics). The absolute values ​​of the sensitivity coefficients of each elementary reaction are compared with the sensitivity screening values, retaining all reactions whose absolute values ​​are greater than or equal to the screening values. These reactions are the core driving reactions affecting the concentration changes of key components in hydrogen-blended natural gas combustion (e.g., OH and H radicals determine ignition delay; NO and NO2 determine pollutant emissions; CH4 and H2 determine fuel consumption rate), such as H + O2 ⇌ OH+O (extremely sensitive to flame propagation speed), CH4+OH⇌ CH3 + H2O (the key initiation step of methane oxidation), N2 + O⇌ Reactions such as NO+N (the core reaction for thermal NOx formation) must retain their complete reaction formulas and rate constant parameters. Reactions retained after sensitivity screening are further verified using reaction flow screening values. Their net reaction flow contribution to target components (such as CO2 and H2O combustion completeness indicators; NOx emission indicators) is calculated. If the absolute value of a reaction's contribution is less than the reaction flow screening value, even if its sensitivity is high (possibly due to its influence on minor components), it is still considered a non-core reaction and removed. For example, some C2-C3 hydrocarbon decomposition reactions (such as C2H6⇌) Although 2CH3 is somewhat sensitive to ignition delay, its net contribution to overall heat release and pollutant generation is weak in scenarios with a high proportion of hydrogen doping, and it can be eliminated.

[0116] Furthermore, for the numerous intermediate products in GRI3.0 (such as CH3O, C2H5, HCO, etc.), the relative change rate of each intermediate product during the entire combustion process is calculated. If its absolute value is consistently less than or equal to the quasi-steady-state component determination value, it is determined to be a quasi-steady-state component. The concentration change of these components is slow, and their formation rate can be approximated as equal to their consumption rate. Therefore, it is not necessary to solve their concentration dynamics through differential equations; instead, their instantaneous concentration can be directly calculated using algebraic equations (based on the rate equilibrium of relevant elementary reactions). For example, the relative change rate of HCO (formaldehyde group) is extremely low in most hydrogen-blended combustion scenarios, and a quasi-steady-state approximation can be used. The correlation between its concentration and components such as OH and CH2O can be derived through the equation "HCO formation reaction rate = consumption reaction rate". For intermediate products with a relative change rate greater than the quasi-steady-state component determination value (such as free radicals such as OH, H, O, etc.), their concentration dynamics directly affect the combustion reaction process. Therefore, it is necessary to retain the solution of their differential equations to ensure the calculation accuracy of key characteristics such as ignition delay and flame propagation. By integrating the screened core reaction with the simplified intermediate products, a simplified mechanism suitable for the combustion of hydrogen-blended natural gas is formed.

[0117] The simplified mechanism focuses on the core reaction pathway of hydrogen-blended natural gas combustion (the key steps of synergistic oxidation of hydrogen and methane and NOx generation). It can accurately reproduce the core characteristics of dynamic parameters in the three simulations, while reducing the amount of computation. It can be directly used for rapid simulation of zero-dimensional / one-dimensional models. At the same time, it provides efficient chemical reaction model support for higher-dimensional (such as two-dimensional) burner numerical simulation, achieving a balance between accuracy and efficiency.

[0118] Furthermore, based on the simplified reference value, the specific implementation steps for generating a mechanistic description of the hydrogen-blended natural gas combustion process to simplify the GRI3.0 mechanism include:

[0119] (1) Based on the detailed chemical reaction mechanism of GRI3.0, all elementary reactions and related chemical components involved in the reaction were identified;

[0120] Specifically, based on the detailed chemical reaction mechanism of GRI3.0, we first sort out the 53 chemical components and 325 elementary reactions it contains, clarifying the type of each component and the core framework of the reaction network. The components cover fuels (CH4, H2, etc.), oxidants (O2, N2), combustion products (CO2, H2O), intermediate products (CH3, OH, H radicals and C2-C3 hydrocarbons), and pollutants (NO, NO2, etc.). The elementary reactions include complete pathways such as chain initiation (e.g., CH4 decomposition), chain propagation (e.g., H+O2→OH+O), chain termination (e.g., OH+NO2→HNO3), and NOx generation / reduction.

[0121] (2) Calculate the sensitivity coefficient of each elementary reaction to the component concentration, compare the sensitivity coefficient with the sensitivity screening value, and retain the elementary reactions whose sensitivity coefficient is greater than or equal to the sensitivity screening value;

[0122] Specifically, based on the component concentration time history data output from three simulations, the perturbation analysis method was used to calculate the sensitivity coefficients of 325 elementary reactions to the concentrations of key components (such as OH which determines ignition delay, H which affects combustion rate, and NO which is associated with NOx emissions). That is, the relative change in the concentration of the target component when the rate constant of a certain elementary reaction is perturbed by 1%. The calculated sensitivity coefficients were compared with the preset sensitivity screening values, and all elementary reactions with sensitivity coefficients greater than or equal to the screening values ​​were retained. These reactions are the key driving forces for the core characteristics of combustion (such as ignition, flame propagation, and pollutant generation), such as H2+OH→H2O+H (the core step of hydrogen oxidation), CH4+OH→CH3+H2O (the dominant reaction for methane consumption), and N+O2→NO+O (the key chain reaction for NO generation). Small changes in their reaction rates can significantly affect the combustion process, so they must be retained.

[0123] (3) For the retained elementary reactions, the contribution ratio of each elementary reaction to the generation or consumption of key components is calculated by reaction flow analysis, and elementary reactions with a contribution ratio less than the reaction flow screening value are eliminated.

[0124] Specifically, for the elementary reactions retained after sensitivity screening, their actual role is further evaluated through reaction flow analysis: the net contribution of each reaction to the generation or consumption of key components (such as major products like CO2 and H2O, and pollutants like NOx) throughout the entire combustion cycle is calculated, and its proportion of the total reaction flow is statistically analyzed. If the contribution of a reaction is less than the reaction flow screening value, it indicates that its actual impact on the generation / consumption of the target component is weak (even if the sensitivity is high, it may only act on a secondary pathway), and it is then eliminated. For example, although some C2H4 decomposition reactions are somewhat sensitive to ignition delay, in scenarios with a high proportion of hydrogen doping, their contribution to the overall heat release and CO2 generation is less than 1%, and they can be safely eliminated to simplify the reaction network.

[0125] (4) Based on the concentration evolution law of intermediate products in actual combustion mechanism, components with concentration change rate less than the quasi-steady-state component judgment value are screened to generate a simplified mechanism description.

[0126] Specifically, for intermediate products in the mechanism (such as CH3O, HCO, C2H5, etc.), the concentration change rate (first derivative) over the entire period is calculated based on their concentration time history data and compared with the quasi-steady-state component judgment value. If the concentration change rate of a certain intermediate product is consistently less than the judgment value, it indicates that its formation and consumption rates are approximately in equilibrium, and its concentration dynamics are stable, thus it can be judged as a quasi-steady-state component. For such components, it is not necessary to solve for their concentration changes through differential equations; instead, the instantaneous concentration is directly calculated using algebraic equations (based on the rate equilibrium of related reactions). At the same time, key intermediate products with concentration change rates greater than the judgment value (such as OH, H, and O free radicals) are retained, as their concentration dynamics directly affect the reaction process and require accurate solutions through differential equations.

[0127] Furthermore, the above screening results are integrated to form a simplified mechanism description, which mainly includes a simplified component list: retaining fuels (CH4, H2), oxidants (O2, N2), major products (CO2, H2O), key free radicals (OH, H, O), pollutants (NO, NO2), and non-quasi-steady-state intermediates (such as CH3), while eliminating minor hydrocarbons and trace intermediates, reducing the number of components to 25-30; core reaction network: retaining the elementary reactions (approximately 80-120 steps) screened by both sensitivity and reaction flow, covering the core pathway of hydrogen / methane oxidation and key reactions for NOx generation, and fully retaining the Arrhenius rate parameters of each reaction (ensuring temperature / pressure correlation); quasi-steady-state treatment rules: clearly labeling quasi-steady-state components and their algebraic equations, limiting applicable operating conditions (such as hydrogen doping ratio of 5%-30%, atmospheric pressure), ensuring clear boundaries of the simplified mechanism's accuracy.

[0128] Optionally, the combustion parameter set includes at least a hydrogen doping ratio, wherein the hydrogen doping ratio is inversely proportional to the ignition delay time, directly proportional to the adiabatic flame temperature, and directly proportional to the peak heat release rate.

[0129] The simulation input parameters include at least the hydrogen doping ratio, which is inversely proportional to the flame length, directly proportional to the flame propagation rate, and directly proportional to the upper limit of flame elongation tolerance.

[0130] Specifically, the hydrogen blending ratio, as a core parameter in the combustion parameter set, directly influences key indicators such as ignition delay, flame morphology, and stability by affecting combustion chemical reaction kinetics and hydrodynamic characteristics, forming clear quantitative laws. In zero-dimensional model simulations, the hydrogen blending ratio significantly impacts combustion characteristic parameters by altering fuel reactivity and energy density; the hydrogen blending ratio is inversely proportional to the ignition delay time. Hydrogen's chemical reactivity is far higher than methane's (H2 has a lower ignition activation energy). After hydrogen blending, the initial generation rate of free radicals such as H and OH in the combustion system accelerates, shortening the time interval from initial conditions to the violent occurrence of the combustion reaction. For example, when the hydrogen blending ratio increases from 0% to 30%, the ignition delay time can be shortened by 40%-60%, especially under low-temperature initial conditions (such as 300-600K), where this effect is more significant, thus improving the ignition reliability of the combustion system; the hydrogen blending ratio is directly proportional to the adiabatic flame temperature. The calorific value of hydrogen (approximately 120 MJ / kg) is higher than that of methane (approximately 55 MJ / kg), and the H2O molecules produced by the combustion of H2 have a higher heat capacity than the CO2 molecules produced by the combustion of CH4. Under conditions of no heat loss, a higher hydrogen blending ratio leads to an increase in the total heat released by combustion, resulting in a rise in the adiabatic flame temperature. For example, when the hydrogen blending ratio increases from 10% to 50%, the adiabatic flame temperature can rise from approximately 2000 K to 2200 K, requiring matching the high-temperature resistance limit of the burner wall material (e.g., ceramic materials need to withstand localized high temperatures ≥2200 K). The hydrogen blending ratio is directly proportional to the peak heat release rate. The combustion reaction rate of hydrogen (especially chain reactions such as H + O2 → OH + O) is much faster than that of methane. With a high hydrogen blending ratio, the fuel consumption rate increases, and more heat is released per unit time, leading to a higher peak heat release rate and an earlier occurrence time. For example, the peak heat release rate when hydrogen is added at 30% is about 1.5 times that of pure methane. Combustion oscillations caused by excessive local heat release need to be avoided through burner configuration design (such as optimizing the inlet flow rate).

[0131] The above patterns provide a clear basis for parameter optimization of hydrogen-blended natural gas combustion systems. For low-hydrogen-blended scenarios, the ignition delay is relatively long and the flame propagation is slow, requiring the flame residence time to be extended through burner configuration (such as increasing the axial length) and the ignition energy input to be enhanced. For high-hydrogen-blended scenarios, the flame temperature is high and the propagation is fast, requiring the optimization of the burner's cooling structure (such as adding water-cooled walls), shortening the axial length, and increasing the inlet flow velocity to match the flame propagation rate. At the same time, its high tensile strength tolerance characteristics can be used to simplify the airflow control design. For the full range of hydrogen blending ratios, it is necessary to use co-simulation of zero-dimensional and one-dimensional models to transform the correlation between the hydrogen blending ratio and parameters such as ignition delay and flame length into quantitative design indicators (such as "for every 10% increase in hydrogen blending, the axial length of the burner is shortened by 8%), to ensure that the combustion system can achieve efficient, stable, and low-emission operation under different hydrogen blending ratios.

[0132] The simulation method for hydrogen-blended natural gas combustion provided in this embodiment has multi-dimensional technical value and can effectively solve the pain points of traditional simulation methods in terms of efficiency, accuracy and engineering adaptability. In terms of efficiency improvement, the zero-dimensional model quickly completes parameter screening for a large number of basic operating conditions, avoiding the direct use of computationally intensive multi-dimensional models, reducing redundant calculations, allowing simulations to focus on core parameters, and significantly reducing the overall process time. The basic models are differentiated and instantiated according to different simulation objectives. The zero-dimensional model focuses on efficiency adaptation to basic parameter screening, while the one-dimensional model retains spatial dimensions and is further subdivided into sub-models adapted to different combustion scenarios. Simultaneously, the heat loss assumptions of the zero-dimensional model are corrected by combining the optimal burner configuration, making the simulation results more closely resemble the actual combustion process, with controllable deviations. Secondary simulations accurately correlate combustion parameters with burner configurations, determining the optimal configuration based on flame spatial characteristic parameters, avoiding the high cost of trial and error in physical experiments, and providing a quantitative basis for burner design. The mechanism simplification process relies on high-precision data from tertiary simulations to simplify the GRI3.0 mechanism, reducing computational complexity while retaining core reaction characteristics, and adapting to simulation needs of different dimensions. Furthermore, the solution clearly outlines the correlation between hydrogen blending ratios and combustion characteristics and configuration parameters, providing scientific support for the optimized design of combustion systems under different hydrogen blending scenarios and facilitating the engineering implementation of hydrogen-blended natural gas combustion technology.

[0133] Corresponding to the aforementioned embodiment of a simulation method for hydrogen-blended natural gas combustion process, this application also provides an embodiment of a simulation device for hydrogen-blended natural gas combustion process.

[0134] Figure 2 This is a schematic diagram of the second embodiment of the simulation device for the combustion process of hydrogen-blended natural gas provided in this application. Please refer to... Figure 2 The apparatus provided in this embodiment includes a construction module 210, a simulation module 220, and a calculation module 230; wherein,

[0135] The construction module 210 is used to establish a basic model for simulating the combustion process of hydrogen-blended natural gas.

[0136] The construction module 210 is further configured to determine the instantiation method of the basic model corresponding to the simulation target, and generate a zero-dimensional model and a one-dimensional model based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process;

[0137] The simulation module 220 is used to perform a single simulation of the hydrogen-blended natural gas combustion process using the zero-dimensional model, and to select multiple target combustion parameters from the combustion parameter set based on the results of the single simulation. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model.

[0138] The simulation module 220 is also used to determine the simulation input parameters based on the multiple target combustion parameters, perform a secondary simulation of the hydrogen-blended natural gas combustion process based on the simulation input parameters and the one-dimensional model, and determine the optimal burner configuration based on the results of the secondary simulation.

[0139] The simulation module 220 is also used to adjust the zero-dimensional model based on the optimal burner configuration, and to perform three simulations of the hydrogen-blended natural gas combustion process using the adjusted zero-dimensional model.

[0140] The calculation module 230 is used to calculate a simplified reference value based on the results of the three simulations.

[0141] The calculation module 230 is also used to simplify the GRI3.0 mechanism based on the simplified reference value and generate a mechanistic description of the hydrogen-blended natural gas combustion process.

[0142] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0143] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0144] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0145] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A simulation method for the combustion process of hydrogen-blended natural gas, characterized in that, The method includes: Establish a basic model for simulating the combustion process of hydrogen-blended natural gas; The instantiation method of the basic model corresponding to the simulation target is determined, and a zero-dimensional model and a one-dimensional model are generated based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process; The zero-dimensional model is used to perform a simulation of the combustion process of hydrogen-blended natural gas. Based on the results of the simulation, multiple target combustion parameters are selected from the set of combustion parameters. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model. The simulation input parameters are determined by the multiple target combustion parameters. A secondary simulation of the hydrogen-blended natural gas combustion process is performed based on the simulation input parameters and the one-dimensional model. The optimal burner configuration is determined based on the results of the secondary simulation. Based on the optimal burner configuration, the zero-dimensional model was adjusted, and the hydrogen-blended natural gas combustion process was simulated three times using the adjusted zero-dimensional model. Based on the results of the three simulations, the computer simplified the reference values. Based on the simplified reference values ​​of the aforementioned mechanism, the GRI3.0 mechanism is simplified to generate a mechanistic description of the hydrogen-blended natural gas combustion process.

2. The method according to claim 1, characterized in that, The process of determining the instantiation method of the base model corresponding to the simulation target, and generating a zero-dimensional model and a one-dimensional model based on the base model and the instantiation method, includes: Based on the simulation accuracy requirements of the simulation target, determine a set of assumption rules with a computational load less than a preset value; The simulation accuracy levels are divided according to the aforementioned simulation accuracy requirements; For each simulation accuracy level, a matching hypothesis rule is determined from the set of hypothesis rules; The base model is simplified according to the assumption rules to generate the zero-dimensional model and the one-dimensional model, wherein the number of assumption rules in the zero-dimensional model is less than that in the one-dimensional model.

3. The method according to claim 1, characterized in that, The zero-dimensional model is used to perform a single simulation of the hydrogen-blended natural gas combustion process. Based on the results of this single simulation, multiple target combustion parameters are selected from the set of combustion parameters, including: Based on multiple combustion parameters in the combustion parameter set, the input conditions of the zero-dimensional model are set; The input conditions are input into the zero-dimensional model and the zero-dimensional model is run. The combustion process of hydrogen-blended natural gas under each set of combustion parameters is simulated and calculated to obtain the output combustion characteristic parameters. The combustion characteristic parameters corresponding to each set of combustion parameters are scored, and multiple target combustion parameters are selected based on the scores.

4. The method according to claim 1, characterized in that, The process of determining simulation input parameters using the multiple target combustion parameters, performing a secondary simulation of the hydrogen-blended natural gas combustion process based on the simulation input parameters and the one-dimensional model, and determining the optimal burner configuration based on the results of the secondary simulation includes: The multiple target combustion parameters are used as input parameters for multiple simulations. The combustion scenario is determined based on the target combustion parameters of each simulation. The combustion scenario is used to describe the mixing process of combustion gases under the target combustion parameters. Based on the target one-dimensional sub-model matching the combustion scenario, the target combustion parameters are used as the input of the target one-dimensional sub-model. Different burner configuration parameters are input to simulate the combustion process, and flame spatial feature parameters corresponding to the target combustion parameters and the burner configuration parameters are obtained. The optimal burner configuration is determined based on the flame space characteristic parameters.

5. The method according to claim 4, characterized in that, Each of the target one-dimensional sub-models corresponds to a different hydrogen-doped natural gas and air mixing mode. The target one-dimensional sub-model is used to describe the changes in the combustion flame, and the flame spatial characteristic parameter types of each target one-dimensional sub-model are different.

6. The method according to claim 1, characterized in that, The process of adjusting the zero-dimensional model based on the optimal burner configuration and performing three simulations of the hydrogen-blended natural gas combustion process using the adjusted zero-dimensional model includes: Adjust the burner configuration parameters in the zero-dimensional model based on the optimal burner configuration; Based on the wall heat dissipation characteristics of the optimal burner configuration, the global uniform heat loss rate of the zero-dimensional model is modified into a dynamic heat loss model related to the wall temperature. Based on the adjusted zero-dimensional model, three simulations of the hydrogen-blended natural gas combustion process were performed to solve the corrected mass and energy conservation equations and output dynamic characteristic parameters, including dynamic combustion parameters, reaction mechanism parameters, and energy characteristic parameters.

7. The method according to claim 1, characterized in that, The simplified reference value calculated based on the results of the three simulations includes: The dynamic characteristic parameters output from the three simulations are obtained, including time history data of component concentration, time history data of temperature, and time history data of heat release rate. Based on the time history data of the component concentration, the sensitivity coefficient of each elementary reaction to the specified component concentration in the detailed mechanism of GRI3.0 is calculated by perturbation analysis. The absolute value distribution of the sensitivity coefficients of all elementary reactions is statistically analyzed, and the sensitivity screening value is determined according to the absolute value distribution. Based on the time history data of the component concentration, the net reaction flow contribution of each elementary reaction step to the target component is calculated by reaction flow analysis, and the reaction flow screening value is determined based on the net reaction flow contribution. Based on the time history data of the component concentration, the change curve of the concentration of each intermediate product over time is extracted, and the ratio of the concentration change rate to the concentration itself at each time point is calculated as the relative change rate. Based on the relative change rate, the quasi-steady-state component determination value is determined. The sensitivity screening value, the reactive flow screening value, and the quasi-steady-state component determination value are used together as the simplified reference value for the mechanism.

8. The method according to claim 1, characterized in that, The simplified GRI3.0 mechanism based on the aforementioned simplified reference value generates a mechanistic description of the hydrogen-blended natural gas combustion process, including: Based on the detailed chemical reaction mechanism of GRI3.0, all elementary reactions and related chemical components involved in the reaction were identified; Calculate the sensitivity coefficient of each elementary reaction to the component concentration, compare the sensitivity coefficient with the sensitivity screening value, and retain the elementary reactions whose sensitivity coefficient is greater than or equal to the sensitivity screening value; For the retained elementary reactions, the contribution percentage of each elementary reaction to the generation or consumption of key components is calculated by reaction flow analysis, and elementary reactions with a contribution percentage less than the reaction flow screening value are eliminated. Based on the concentration evolution of intermediate products in actual combustion mechanisms, components with concentration change rates less than the quasi-steady-state component determination values ​​are screened to generate a simplified mechanism description.

9. The method according to claim 1, characterized in that, The combustion parameter set includes at least a hydrogen doping ratio, which is inversely proportional to the ignition delay time, directly proportional to the adiabatic flame temperature, and directly proportional to the peak heat release rate. The simulation input parameters include at least the hydrogen doping ratio, which is inversely proportional to the flame length, directly proportional to the flame propagation rate, and directly proportional to the upper limit of flame elongation tolerance.

10. A simulation device for the combustion process of hydrogen-blended natural gas, characterized in that, The device includes a construction module, a simulation module, and a calculation module; wherein... The building module is used to establish a basic model for simulating the combustion process of hydrogen-blended natural gas; The construction module is further configured to determine the instantiation method of the basic model corresponding to the simulation target, and generate a zero-dimensional model and a one-dimensional model based on the basic model and the instantiation method; wherein, the simulation accuracy of the zero-dimensional model is lower than that of the one-dimensional model, and both the zero-dimensional model and the one-dimensional model are used to simulate the hydrogen-blended natural gas combustion process; The simulation module is used to perform a single simulation of the hydrogen-blended natural gas combustion process using the zero-dimensional model, and to select multiple target combustion parameters from the combustion parameter set based on the results of the single simulation. The combustion parameters are used to control the parameters of the combustion gas input to the zero-dimensional model. The simulation module is also used to determine the simulation input parameters based on the multiple target combustion parameters, perform a secondary simulation of the hydrogen-blended natural gas combustion process based on the simulation input parameters and the one-dimensional model, and determine the optimal burner configuration based on the results of the secondary simulation. The simulation module is also used to adjust the zero-dimensional model based on the optimal burner configuration, and to perform three simulations of the hydrogen-blended natural gas combustion process using the adjusted zero-dimensional model. The calculation module is used to calculate a simplified reference value based on the results of the three simulations. The calculation module is also used to simplify the GRI3.0 mechanism based on the simplified reference value and generate a mechanistic description of the hydrogen-blended natural gas combustion process.

Images