Multi-scale coupling modeling method for adiabatic axial fixed bed reactor

Abstract

The invention discloses a multi-scale coupling modeling method for an adiabatic axial fixed bed reactor, and relates to the field of chemical process modeling and reaction engineering.The modeling method comprises the following steps that S1, a three-level basic structure unit is constructed; s2, establishing a physical-chemical coupling calculation model; s3, embedding wide modulus parameter dynamic adaptation logic; s4, acquiring micro-scale reaction kinetic parameters, mesoscale transfer characteristic parameters and macro-scale equivalent performance parameters; s5, establishing a diffusion-reaction coupling equation of a single catalyst particle, and obtaining a comprehensive reaction-transfer parameter of the particle size; s6, establishing a macroscopic coupling model, describing a flow process in the reactor, and realizing cross-scale real-time coupling calculation of flow-diffusion-reaction-heat transfer; and S7, based on a cross-scale real-time coupling calculation result, extracting target performance parameters of the whole reactor. According to the method, the prediction precision of the model is improved, reactor hotspot control and safety production are effectively served, and the stability of industrial production is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of chemical process modeling and reaction engineering technology, and in particular to a multi-scale coupled modeling method for an adiabatic axial fixed-bed reactor. Background Technology

[0002] Adiabatic axial fixed-bed reactors, with their simple structure and stable operation, are widely used in catalytic reaction processes across various industrial sectors, including chemical, energy, and environmental protection. These reactors typically operate under adiabatic conditions, and their internal physicochemical processes span multiple levels, from molecular-scale reactions and diffusion within catalyst micropores to mesoscale particle agglomeration transport, and finally to macroscopic reactor-scale flow and heat transfer, forming a typical and complex multi-scale coupled system. In actual industrial production, the operating conditions of the reaction system are not static. Adjustments to the production load and fluctuations in feedstock quality can lead to significant changes in reaction rate, diffusion efficiency, and heat transfer, with the corresponding reaction rate modulus, diffusion modulus, and heat transfer modulus potentially varying over an extremely wide range. This wide modulus range presents a significant challenge to the accurate modeling and simulation prediction of the reactor, as traditional modeling methods often struggle to maintain sufficient accuracy and stability under such dynamic and broad operating conditions.

[0003] Various modeling methods have been developed to address the multi-scale characteristics of adiabatic axially fixed-bed reactors. These methods typically aim to construct a model framework from micro to macro scales, approximating the overall behavior of the reactor by describing the physicochemical processes at different scales. When addressing multi-scale coupling problems, a common approach is to pre-set or calibrate a fixed set of model parameters for a specific or narrow modulus operating range, and then unidirectionally transfer the parameters calculated at the microscale to the macroscale model. This approach may be effective under relatively fixed operating conditions.

[0004] However, existing technical solutions have significant drawbacks and shortcomings. First, due to the lack of a dynamic adaptation mechanism for a wide modulus range, when actual operating conditions cause the modulus to exceed the preset range, the model parameters cannot accurately match the changed physicochemical processes, leading to a severe decrease in model prediction accuracy. In the low modulus range, the regularity of parameter changes may be ignored; in the high modulus range, numerical oscillations or result distortions may occur due to the lack of parameter adaptation logic. Second, in the transition region where the modulus range changes, model parameters often exhibit discontinuous jumps, which not only disrupts the consistency of cross-scale parameter transfer but may also make the numerical solution process difficult to converge, i.e., numerical instability exists. Furthermore, the coupling between existing macroscopic and microscopic models is mostly a simple one-way parameter transfer, which cannot achieve bidirectional real-time collaborative solution. It is difficult to capture and reflect the dynamic influence of reaction heat accumulation on microscopic reaction dynamics under adiabatic conditions, as well as the immediate feedback of microscopic reaction changes on macroscopic flow and temperature fields, i.e., scale coupling is not real-time.

[0005] Therefore, it is necessary to improve upon the shortcomings of existing technologies in order to solve the above problems. Summary of the Invention

[0006] This invention overcomes the shortcomings of the prior art and provides a multi-scale coupled modeling method for adiabatic axial fixed-bed reactors.

[0007] To achieve the above objectives, the technical solution adopted by this invention is as follows: This invention provides a multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor, comprising the following steps:

[0008] S1. Based on the multi-level structural characteristics of the catalyst and the requirement for a wide modulus range of the reaction system, a three-level basic structural unit is constructed, consisting of microscale pore units, mesoscale aggregation units, and macroscale particle units.

[0009] S2. For the three-level basic structural units, physical-chemical coupling calculation models including diffusion, adsorption, desorption, chemical reaction and adiabatic heat transfer processes are established respectively.

[0010] S3. Embed wide-modulus parameter dynamic adaptation logic in the physical-chemical coupling calculation model to dynamically match and switch the corresponding parameter function according to the real-time modulus range, so as to maintain the continuity and accuracy of parameters within the wide modulus range.

[0011] S4. Based on the physical-chemical coupling calculation model, perform numerical simulation to obtain microscale reaction kinetic parameters, mesoscale transport characteristic parameters, and macroscale equivalent performance parameters.

[0012] S5. Establish the diffusion-reaction coupling equation for a single catalyst particle, using mesoscale transport characteristic parameters and microscale reaction kinetic parameters as inputs, and obtain the comprehensive reaction-transport parameters at the particle scale through numerical solution.

[0013] S6. Establish a macroscopic coupling model based on the structural parameters of the reactor, use computational fluid dynamics to describe the flow process inside the reactor, and combine wide modulus parameter dynamic adaptation logic with particle-scale integrated reaction-transfer parameters to achieve cross-scale real-time coupled calculation of flow-diffusion-reaction-heat transfer.

[0014] S7. Based on the cross-scale real-time coupled calculation results, extract the overall target performance parameters of the reactor.

[0015] In a preferred embodiment of the present invention, in step S3, the dynamic adaptation logic for the wide modulus parameter includes the following steps:

[0016] S31. Divide the reaction rate modulus, diffusion modulus, and heat transfer modulus into low, medium, and high modulus ranges, respectively;

[0017] S32. For different modulus ranges, construct specific functional relationships for the reaction rate constant, effective diffusion coefficient, and effective thermal conductivity;

[0018] S33. Based on the real-time calculated modulus value, dynamically switch the corresponding dedicated function to achieve a smooth transition of parameters within a wide modulus range.

[0019] In a preferred embodiment of the present invention, the reaction rate modulus range is divided into: low modulus range [0.01, 0.5], medium modulus range (0.5, 10], and high modulus range (10, 100];

[0020] The diffusion modulus range is divided into: low modulus range [0.1, 1], medium modulus range (1, 10], and high modulus range (10, 50].

[0021] The heat transfer modulus range is divided into: low modulus range [0.05, 1], medium modulus range (1, 15], and high modulus range (15, 80).

[0022] In a preferred embodiment of the present invention, the effective diffusion coefficient D of the microscale pore unit is... eff,micro According to the Bosanquet formula:

[0023] ; where D K D is the Knudsen diffusion coefficient; mol is the molecular diffusion coefficient.

[0024] In a preferred embodiment of the present invention, the effective diffusion coefficient D of the mesoscale agglomeration unit is... eff,meso Based on the modified Bruggeman relation:

[0025] ;

[0026] Among them, D bulk ε is the bulk gas diffusion coefficient; meso τ is the porosity of the aggregate; meso The degree of curvature.

[0027] In a preferred embodiment of the present invention, in step S5, the diffusion-reaction coupling equation of the single catalyst particle is in two-dimensional form in spherical coordinates, including radial and axial dimensions; the source term of the equation is composed of microscale reaction kinetics and mesoscale transfer parameters.

[0028] In a preferred embodiment of the present invention, in step S6, the macroscopic coupling model and the particle-scale model are solved collaboratively through a bidirectional real-time coupling mechanism, including the following steps:

[0029] S61. Within each time step, the local flow velocity, temperature, and component concentration of the macroscopic coupling model are used as the boundary conditions of the particle-scale model.

[0030] S62. Use the reaction rate, heat generation rate and effective physical property parameters calculated by the particle-scale model as source terms for the macroscopic model.

[0031] S63. Parameter transfer and collaborative solving are completed within each time step, and data interaction is encapsulated in a standardized format.

[0032] In a preferred embodiment of the present invention, in step S4, the numerical simulation employs the finite volume method for spatial discretization, and the time progression uses a second-order implicit Crank-Nicolson scheme; the linear system solution uses the GMRES iterative algorithm combined with ILU preprocessing, and the convergence residual threshold is set to 1×10⁻⁶. -6 .

[0033] In a preferred embodiment of the present invention, in step S7, the target performance parameters include: reactant conversion rate, product selectivity, bed temperature distribution, carbon deposition, and reactor pressure drop.

[0034] In a preferred embodiment of the present invention, the method further includes: establishing a reaction mechanism network comprising a main reaction, side reactions, and a carbon deposition reaction; the kinetic parameters of each reaction pathway independently depend on the local temperature and the concentration of key components.

[0035] This invention addresses the shortcomings of the prior art and has the following beneficial effects:

[0036] (1) This invention provides a multi-scale coupling modeling method for adiabatic axial fixed-bed reactors. By constructing a three-level structure of microscale pore units, mesoscale agglomeration units and macroscale particle units, and embedding wide-modulus parameter adaptation logic in the model, the structural division can cover the entire scale process from molecular reaction and diffusion in catalyst micropores to macroscopic reactor flow and heat transfer. The adaptation logic realizes dynamic parameter switching through modulus interval division and dedicated functions, thereby realizing dynamic matching of reaction transfer parameters in different modulus intervals, ensuring the continuity and accuracy of parameters in a wide modulus range. Compared with traditional methods that lack dynamic adaptation mechanism and are prone to parameter discontinuity or insufficient accuracy when switching modulus intervals, and cannot cope with changing working conditions, this invention effectively solves the technical bottleneck of traditional models being unable to adapt to a wide modulus range, and improves the applicability of the model in industrial scale-up and optimization.

[0037] (2) In this invention, a physicochemical coupling model is established for each structural unit, including diffusion, adsorption, desorption, reaction mechanism and adiabatic heat transfer process, and a two-way real-time coupling at the macroscopic and particle scale is realized. It integrates the interaction of multiple physicochemical processes. The two-way coupling mechanism allows the macroscopic flow field and temperature boundary conditions to be transferred to the microscopic model, while the microscopic reaction rate and heat source term are fed back to the macroscopic model. This can accurately reflect the dynamic influence of the accumulation of reaction heat under adiabatic conditions on the microscopic reaction dynamics and macroscopic flow field, and avoid the prediction deviation caused by the lack of reaction heat feedback. Compared with the existing methods, which are mostly unidirectional transmissions and cannot solve in real time, it is difficult to capture the dynamic changes of reaction heat. This invention improves the prediction accuracy of the model, effectively serves the control of reactor hot spots and safe production, and ensures the stability of industrial production.

[0038] (3) This invention supports networked dynamics of main reaction, side reaction and carbon deposition reaction, and outputs standardized target performance parameters. It enables the parameters of each reaction path to be updated in conjunction with temperature and component concentration field. The standardized output format facilitates integration with mainstream process simulation or factory system, so as to scientifically reflect the complex reaction path in real industrial process, provide comprehensive reactor performance data, and provide full-link data support for industrial design and operation optimization, realize seamless connection between model and production system, and thus improve the efficiency and reliability of industrial production. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1This is a schematic diagram of the multi-scale structural layering principle;

[0041] Figure 2 This is a schematic diagram of reaction-transfer-thermal coupling and parameter adaptation;

[0042] Figure 3 This is a diagram showing the percentage of influence of each scale parameter in the model of this invention;

[0043] Figure 4 This is a schematic diagram of the two-way real-time coupling mechanism between the macroscopic and particle-scale models in this invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0046] This invention provides a multi-scale coupled modeling method for adiabatic axial fixed-bed reactors. By constructing a three-level structure of microscale pore units, mesoscale agglomeration units, and macroscale particle units, a fully coupled computational model of the physical-chemical process is established. Furthermore, a wide-modulus parameter dynamic adaptation logic is embedded, ultimately achieving bidirectional real-time collaborative solving between the macroscale reactor scale and the particle scale. The technical solution of this invention is described in detail below with reference to specific embodiments.

[0047] It should be noted that, based on the multi-level pore structure characteristics of supported metal oxide catalysts (such as Co-Mo / Al2O3 or Ni-W / γ-Al2O3) commonly used in adiabatic axial fixed-bed reactors in industrial applications, a three-level basic structural unit is established.

[0048] Preferably, such as Figure 1 As shown, the geometric size of the microscale channel unit ranges from 10 nm to 1 μm, and it contains a microporous structure with an average pore size of 5-50 nm. Catalytic active sites are uniformly distributed on the pore walls, with an active site density of 1.2 × 10⁻⁶. 19 sites / m 2The mesoscale aggregates have a geometric size of 1 μm-100 μm and are formed by the physical aggregation of 50-200 microscale pore units through van der Waals forces and capillary bridging. Their internal pore structure includes mesopores and macropores, and the porosity ε of the aggregates is [not specified]. meso The tortuosity τ is 0.35-0.45. meso The value is 3.2-4.8. Macroscale particle units correspond to cylindrical or cloverleaf-shaped catalyst particles used in actual industrial applications, with geometric dimensions of 5 mm-10 mm and a packing porosity ε. bulk The particle density ρ is 0.38-0.42. p 1850 kg / m 3 Specific heat capacity C p It is 920 J / (kg·K).

[0049] Specifically, such as Figure 2 and Figure 3 As shown, the influence ratios of parameters at various levels, from microscopic to mesoscopic to macroscopic, are scientifically allocated. Within the microscale porous unit, a coupled model is established, incorporating molecular diffusion, surface adsorption, desorption, chemical reactions, and local adiabatic heat transfer. The molecular diffusion process is described using a modified Fick's diffusion law, with an effective diffusion coefficient D. eff,micro The Knudsen diffusion coefficient D K With molecular diffusion coefficient D mol Obtained using the Bosanquet formula:

[0050] ;

[0051] Wherein, the Knudsen diffusion coefficient d p Where is the pore size, R is the universal gas constant, T is the local temperature, M is the molar mass of the component, and D is the molecular diffusion coefficient. mol Calculations were performed using Chapman-Enskog theory. The adsorption and desorption processes were modeled using the Langmuir isotherm kinetic model, and the adsorption rate... Desorption rate Where θ is the surface coverage and k ads With k des All are temperature-dependent functions, in the form of Pre-exponential factor A and activation energy E a The results were obtained by fitting in-situ infrared spectroscopy experiments.

[0052] Surface chemical reactions follow the Arrhenius equation; for example, the rate expression for the main reaction from A to B is: Where n is the reaction order, θ c This is a correction term for carbon deposit coverage. The local energy conservation equation is:

[0053] ;

[0054] Where, ρ s C is the density of the solid skeleton. p,s For solid phase heat, k s Let ΔH be the thermal conductivity of the solid phase. r The enthalpy change is the main reaction. The sign of the derivative with respect to time is used to describe the rate at which a physical quantity changes with time.

[0055] Furthermore, in mesoscale agglomeration units, mass transfer employs a porous media effective diffusion model, with an effective diffusion coefficient D. eff,meso Modified by Bruggeman relation:

[0056] ;

[0057] Among them, D bulk Let be the bulk gas diffusion coefficient. Heat conduction follows Fourier's law of thermal conductivity for porous media, with an effective thermal conductivity k. eff,meso Calculated using the Wakao model:

[0058] ;

[0059] Where k g k is the thermal conductivity of the gas. s Let J be the thermal conductivity of the solid framework. Introducing an interfacial resistance term at the agglomerate boundary, the mass flux J satisfies:

[0060] ;

[0061] Where k int C is the interfacial mass transfer coefficient. ext With C int These represent the external and internal concentrations, respectively.

[0062] Furthermore, in macroscale particle units, using a two-dimensional diffusion-reaction coupling equation in spherical coordinates (considering the radial r and axial z directions), the macroscale component mass conservation equation is:

[0063] ;

[0064] The macroscopic energy conservation equation is:

[0065] ;

[0066] Where D eff,macro With k eff,macro Determined by the mesoscale simulation output, ν ij r is the stoichiometric coefficient. j Let be the rate of the j-th reaction.

[0067] In this implementation, to achieve parameter continuity over a wide modulus range, the present invention embeds dynamic parameter adaptation logic. The reaction rate modulus is defined. diffusion modulus Heat transfer modulus , where L is the characteristic length (pore size for microscale, agglomerate diameter for mesoscale, and particle radius for macroscale), k0 is the reference reaction rate constant, and h is the convective heat transfer coefficient.

[0068] It should be noted that the boundaries of each modulus interval were determined through global sensitivity analysis (using the Sobol index method): the reaction rate modulus has a low interval of [0.01, 0.5], a middle interval of (0.5, 10], and a high interval of (10, 100]; the diffusion modulus has a low interval of [0.1, 1], a middle interval of (1, 10], and a high interval of (10, 50]; and the heat transfer modulus has a low interval of [0.05, 1], a middle interval of (1, 15], and a high interval of (15, 80).

[0069] For different regions, specific parameter functions are constructed. For example, the reaction rate constant k(T) adopts the original Arrhenius form in the low-modulus region; and a Thiele modulus correction factor η(Φ) is introduced in the high-modulus region. r ), Cubic spline interpolation is used in the middle interval to ensure the continuity of the first derivative. Effective diffusion coefficient D eff In the Knudsen-dominated area, D K Dominated by D, in the molecular diffusion-dominant region mol The primary region uses the explicit solution of the Bosanquet formula, while the intermediate region employs the secondary region. The effective thermal conductivity k... eff Based on solid-phase continuity, a parallel-series hybrid model is adopted:

[0070] Series: ;

[0071] in parallel: ;

[0072] In actual calculations, the average of the two is taken, and the weight is determined by the pore connectivity index.

[0073] In this implementation, the numerical solution of each level of the model is spatially discretized using the finite volume method. The microscale model uses unstructured triangular elements, with a total of 12,500 elements and a time step Δt = 1 × 10⁻⁶. -4 The mesoscale model uses quadrilateral elements, with 8200 elements, and Δt = 5 × 10⁻⁶. -3The macroscale particle model uses a structured mesh with 50 radial layers and 100 axial layers, with Δt = 0.01 s. All time-progression methods employ a second-order implicit Crank-Nicolson scheme. The linear system solution uses the GMRES iterative algorithm with ILU preprocessing, and the convergence residual threshold is set to 1×10⁻⁶. -6 .

[0074] Furthermore, the microscale simulation output includes: the activation energy of the main reaction, E. a,main = 85.2 kJ / mol, pre-exponential factor A main = 2.1×10 6 s -1 The reaction order n = 1.2; the side reaction E a,side = 112.5 kJ / mol, A side = 8.7 × 10 8 s -1 The rate constant of the carbon deposition reaction, k coke = 1.5×10 3 exp(-75000 / RT) mol / (m 2 ·s), carbon deposit coverage correction term θ c = 1-exp(-α Γ coke ), where Γ coke For carbon deposition per unit area, α = 0.85 m 2 / g.

[0075] Mesoscale simulation output includes: D eff,meso = 1.8 × 10 -6 m 2 / s (H2 at 300 K, 2.0 MPa), k eff,meso = 0.42 W / (m·K), pressure drop factor β = 1.35×10 9 Pa·s / m 2 .

[0076] The macroscale particle model outputs the equivalent reaction rate r. eq = ηkC A Equivalent heat transfer coefficient h eq = k eff,macro / R p , where R p Where is the particle radius.

[0077] Furthermore, after obtaining the comprehensive parameters at the particle scale, a single-particle diffusion-reaction coupling model is established. This model is based on spherical coordinates, including radial and axial dimensions, and the source term is composed of microscale dynamics and mesoscale transfer parameters. For example, the main reaction source term... Energy source item The internal temperature and concentration fields of the particles are solved using a fully coupled iterative method, with 5-8 internal iterations performed per time step until the residuals of both concentration and temperature are less than 1×10⁻⁶. -5 .

[0078] Preferably, a macroscopic coupling model is established based on the macroscopic structural parameters of the reactor. Taking an industrial hydrodesulfurization reactor as an example, the inner diameter is 2.4 m, the bed height is 8.0 m, and the catalyst bulk density is 720 kg / m³. 3 The inlet is a tapering section (contraction angle 15°), and the outlet is a straight pipe section. The macroscopic model governing equations are as follows:

[0079] Conservation of mass: ;

[0080] Conservation of momentum (modified Ergun equation): ;

[0081] Component transport: ;

[0082] Energy conservation (adiabatic boundary): ;

[0083] Where ε is the bed porosity, d p Y is the equivalent diameter of the particle. i Here, represents the mass fraction, and represents the net generation rate, fed back in real-time by the particle-scale model. An adiabatic boundary condition is set on the reactor wall: ∇T=0.

[0084] In the implementation method, such as Figure 4 As shown, the macroscopic model and the granular-scale model are solved collaboratively through a two-way real-time coupling mechanism. The specific process is as follows:

[0085] 1) At the macroscopic time step t n For each computational unit (representing a local bed micro-element), the local flow velocity u, temperature T, and component concentration C are extracted as boundary conditions for the particle model.

[0086] 2) Use the particle-scale solver to calculate the reaction rate r under these local conditions. local Heat generation rate q local and effective physical properties D eff,local k eff,local ; will r local With q local As a source term injected into the macroscopic components and energy equations, D eff,local With k eff,local Used to update local transfer coefficients;

[0087] 3) After completing the macroscopic full-field solution, proceed to the next time step t. n+1Data interaction is encapsulated in HDF5 format, and each coupling loop takes approximately 180 ms (on an Intel Xeon Gold 6248R 2.4 GHz x 32 core server).

[0088] In a preferred embodiment of the present invention, the reaction mechanism network comprises three pathways: the main reaction (thiophene hydrodesulfurization to butane and H2S), the side reaction (over-hydrogenation of olefins to alkanes), and the coking reaction (condensation of heavy aromatics to coke). The kinetics of each pathway independently depend on local temperature (T) and temperature (T). Carbon deposit coverage correction item β=1.2m 2 / g,Γ coke This is obtained from the integral carbon deposition rate. All reaction heat effects are included in the energy equation, ΔH. main = -125 kJ / mol, ΔH side = -85 kJ / mol, ΔH coke = -320 kJ / mol.

[0089] As another preferred implementation, the macroscopic model uses a structured mesh with 220 layers axially and 60 layers radially; the particle-scale model uses unstructured quadrilateral elements with 10,800 elements; the time step is initially 0.01 s and is adaptively adjusted according to the CFL condition, with a maximum of 0.8 s.

[0090] To verify the effectiveness of the present invention, the following embodiments and comparative examples were designed.

[0091] Example 1: A full-condition simulation of a 2.4 m inner diameter adiabatic fixed-bed hydrogenation reactor was conducted using the multi-scale coupled modeling method described in this invention. Operating conditions: inlet temperature 320 ℃, pressure 3.5 MPa, H2 / oil ratio 300 Nm. 3 / m 3 Liquid hourly space velocity 1.2 h⁻¹ -1 After the reactor reached steady state, the dynamic carbon deposition process was simulated for another 120 hours.

[0092] Comparative Example 1: The traditional single-scale homogeneous model is adopted, which assumes that the catalyst particles are homogeneous media, the reaction rate adopts apparent kinetics (fitted by laboratory small-scale tests), the heat and mass transfer coefficients are taken as constants, there is no cross-scale coupling, and no carbon deposition kinetics.

[0093] Comparative Example 2: A two-level coupling model (macro-micro scale only) is adopted, ignoring mesoscale aggregation structures. Microscale parameters are directly pushed up to the particle scale, without wide modulus adaptation logic, and parameter switching uses a hard threshold.

[0094] The simulation results are shown in Table 1.

[0095] Table 1: Performance simulation results of Example 1 and Comparative Examples 1-2

[0096] Performance indicators Example 1 Comparative Example 1 Comparative Example 2 Export thiophene conversion rate (%) 99.2 98.5 98.9 Butane selectivity (%) 94.7 91.2 93.1 Maximum bed temperature (°C) 382.5 375.0 379.8 Hotspot location (m from entrance) 2.1 2.8 2.4 Pressure drop (kPa) after 120 hours of operation 86.3 62.1 74.5 Carbon deposits (wt%) 4.8 3.2 4.1 Numerical stability (maximum residual) <![CDATA[8.7×10 -7 ]]> <![CDATA[1.2×10 -5 ]]> <![CDATA[3.5×10 -6 ]]> Single-step coupling time (s) 182 28 95

[0097] As shown in Table 1, the proposed method significantly outperforms the comparative method in terms of prediction accuracy, hotspot detection capability, carbon deposition evolution, and pressure drop growth. Especially under adiabatic conditions, due to the accurate characterization of the thermo-chemical-fluid positive feedback effect, the proposed method can indeed predict hotspots that appear earlier and have higher temperatures, which is in high agreement with the actual operating data of industrial plants (hotspot temperatures measured on-site were 380-385℃, and the location was 2.0-2.3 m).

[0098] Among them, Comparative Example 1 underestimated the temperature gradient because it ignored the internal transmission resistance; Comparative Example 2, although it partially considered the microstructure, showed numerical oscillations when the modulus crossed due to the lack of mesoscale transition and smooth parameter adaptation.

[0099] In summary, this invention, through the physical correspondence of three-level structural units, continuous adaptation of wide modulus parameters, and bidirectional real-time data flow at the macro-particle scale, enables a unified description of the entire scale, from nanoscale channel reactions to meter-scale bed flows. Under adiabatic boundary conditions, the proposed method can stably, continuously, and accurately predict the entire lifecycle behavior of reactors under dynamic operating conditions, providing a high-fidelity digital twin foundation for the design optimization, safety monitoring, and intelligent control of industrial reactors.

[0100] The above description is based on the preferred embodiments of the present invention. It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered exemplary and non-limiting in all respects. The scope of the invention is defined by the appended claims rather than the foregoing description, and all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0101] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor, characterized in that, Includes the following steps: S1. Based on the multi-level structural characteristics of the catalyst and the requirement for a wide modulus range of the reaction system, a three-level basic structural unit is constructed, consisting of microscale pore units, mesoscale aggregation units, and macroscale particle units. S2. For the three-level basic structural units, physical-chemical coupling calculation models including diffusion, adsorption, desorption, chemical reaction and adiabatic heat transfer processes are established respectively. S3. Embed wide-modulus parameter dynamic adaptation logic in the physical-chemical coupling calculation model to dynamically match and switch the corresponding parameter function according to the real-time modulus range, so as to maintain the continuity and accuracy of parameters within the wide modulus range. S4. Based on the physical-chemical coupling calculation model, perform numerical simulation to obtain microscale reaction kinetic parameters, mesoscale transport characteristic parameters, and macroscale equivalent performance parameters. S5. Establish the diffusion-reaction coupling equation for a single catalyst particle, using mesoscale transport characteristic parameters and microscale reaction kinetic parameters as inputs, and obtain the comprehensive reaction-transport parameters at the particle scale through numerical solution. S6. Establish a macroscopic coupling model based on the structural parameters of the reactor, use computational fluid dynamics to describe the flow process inside the reactor, and combine wide modulus parameter dynamic adaptation logic with particle-scale integrated reaction-transfer parameters to achieve cross-scale real-time coupled calculation of flow-diffusion-reaction-heat transfer. S7. Based on the cross-scale real-time coupled calculation results, extract the overall target performance parameters of the reactor.

2. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: In step S3, the dynamic adaptation logic for the wide modulus parameter includes the following steps: S31. Divide the reaction rate modulus, diffusion modulus, and heat transfer modulus into low, medium, and high modulus ranges, respectively; S32. For different modulus ranges, construct specific functional relationships for the reaction rate constant, effective diffusion coefficient, and effective thermal conductivity; S33. Based on the real-time calculated modulus value, dynamically switch the corresponding dedicated function to achieve a smooth transition of parameters within a wide modulus range.

3. The multi-scale coupled modeling method for an adiabatic axial fixed-bed reactor according to claim 2, characterized in that: The reaction rate modulus range is divided into: low modulus range [0.01, 0.5], medium modulus range (0.5, 10], and high modulus range (10, 100]. The diffusion modulus range is divided into: low modulus range [0.1, 1], medium modulus range (1, 10], and high modulus range (10, 50]. The heat transfer modulus range is divided into: low modulus range [0.05, 1], medium modulus range (1, 15], and high modulus range (15, 80).

4. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: The effective diffusion coefficient D of the microscale pore unit eff,micro According to the Bosanquet formula: ; where D K D is the Knudsen diffusion coefficient; mol is the molecular diffusion coefficient.

5. The multi-scale coupled modeling method for an adiabatic axially fixed bed reactor according to claim 1, characterized in that: The effective diffusion coefficient D of the mesoscale agglomeration unit eff,meso Based on the modified Bruggeman relation: ； Among them, D bulk ε is the bulk gas diffusion coefficient; meso τ is the porosity of the aggregate; meso The degree of curvature.

6. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: In step S5, the diffusion-reaction coupling equation of the single catalyst particle is in two-dimensional form in spherical coordinates, including radial and axial dimensions; the source term of the equation is composed of microscale reaction kinetics and mesoscale transfer parameters.

7. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: In step S6, the macroscopic coupling model and the particle-scale model are solved collaboratively through a bidirectional real-time coupling mechanism, including the following steps: S61. Within each time step, the local flow velocity, temperature, and component concentration of the macroscopic coupling model are used as the boundary conditions of the particle-scale model. S62. Use the reaction rate, heat generation rate and effective physical property parameters calculated by the particle-scale model as source terms for the macroscopic model. S63. Parameter transfer and collaborative solving are completed within each time step, and data interaction is encapsulated in a standardized format.

8. The multi-scale coupled modeling method for an adiabatic axial fixed-bed reactor according to claim 1, characterized in that: In step S4, the numerical simulation employs the finite volume method for spatial discretization, and the time progression uses a second-order implicit Crank-Nicolson scheme; the linear system solution uses the GMRES iterative algorithm combined with ILU preprocessing, and the convergence residual threshold is set to 1×10⁻⁶. -6 .

9. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: In step S7, the target performance parameters include: reactant conversion rate, product selectivity, bed temperature distribution, carbon deposition, and reactor pressure drop.

10. The multi-scale coupled modeling method for an adiabatic axially fixed-bed reactor according to claim 1, characterized in that: The method further includes: establishing a reaction mechanism network that includes the main reaction, side reactions, and coking reaction; the kinetic parameters of each reaction pathway depend independently on the local temperature and the concentration of key components.

Images