A Numerical Simulation Method and System for Full-Cycle Hydrogen Storage in Aquifers Using a Dual-Porosity Model

By using a dual porosity model and intelligent injection-production control, the problems of distorted seepage description and lack of dynamic control in aquifer hydrogen storage have been solved, enabling high-precision hydrogen storage design and safe operation.

CN122509030BActive Publication Date: 2026-11-10CHINA UNIV OF MINING & TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202610994721.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-07-06
Publication Date
2026-11-10
Estimated Expiration
2046-07-06

AI Technical Summary

Technical Problem

Existing numerical simulation techniques for hydrogen storage in aquifers suffer from problems such as distorted descriptions of seepage, oversimplification of hydrogen properties, lack of adsorption models, and lack of dynamic injection and production control, leading to safety hazards and low efficiency in the design and operation of hydrogen storage facilities.

Method used

A dual-porosity model, combined with a multi-component fluid model and an adsorption-desorption model, is used to accurately characterize the fracture-matrix two-phase flow. Considering the actual physical properties of hydrogen under high pressure, an intelligent injection-production control strategy is set up, and numerical simulation is performed using the fully implicit finite difference method.

Benefits of technology

It improves the accuracy of gas injection front prediction, reduces hydrogen recovery error, enhances the safety and operational efficiency of hydrogen storage facilities, and is applicable to various reservoir types and gas injection schemes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122509030B_ABST
    Figure CN122509030B_ABST
Patent Text Reader

Abstract

This invention relates to a method and system for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model. The method includes: obtaining reservoir geological boundary conditions; constructing a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions; introducing an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model; defining a dual-porosity relative permeability curve; using the dual-porosity relative permeability curve to set an intelligent control scheme for injection and production wells to constrain the dual-porosity integrated model; and calculating the integrated model parameters using a fully implicit finite difference method to obtain numerical simulation results; evaluating reservoir performance based on the numerical simulation results to obtain reservoir performance results. This invention effectively improves the simulation accuracy and safety prediction capability of the hydrogen injection and production process in aquifers.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of underground energy storage technology, and in particular to a method and system for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model. Background Technology

[0002] Underground Hydrogen Storage in Aquifers (UHSA) is a key technology for large-scale underground energy storage. Its core principle is to inject hydrogen into the pores and fractures of deep saline aquifers for storage and extraction during peak periods. Compared to hydrogen storage in salt caverns and depleted oil and gas reservoirs, aquifers are widely distributed, have large capacities, and diverse resource endowments. They are well-suited to the intermittent nature of renewable energy generation, making them an ideal carrier for cross-seasonal peak-shaving energy storage and of great significance for energy structure transformation. However, due to the complexity of reservoir geology and the unique physicochemical properties of hydrogen, UHSA projects face numerous technical challenges, as follows: One challenge is the seepage control in a dual-medium pore structure. Sandstone aquifers, as primary target reservoirs, generally exhibit a dual-medium characteristic: matrix pores with pore sizes <1μm and high permeability. The fluid diffuses slowly; the permeability of the fracture system reaches... The two are rapid fluid transport channels, and their seepage characteristics differ by 3 to 6 orders of magnitude. This extreme heterogeneity leads to a "fast-slow differentiation" in the transport path after hydrogen injection, posing a great challenge to reserve assessment, rate control, and leak prevention. Secondly, there are safety and efficiency issues arising from the unique physical properties of hydrogen. Hydrogen has the smallest molecular weight (2 g / mol) and a density of 0.0899 kg / m³ under standard conditions. 3 Viscosity This is significantly different from conventional reservoir fluids. Its low viscosity and density give it strong breakthrough capabilities, making it prone to flow along high-permeability channels, increasing the risk of caprock breakthrough and leakage. Furthermore, its interfacial tension with formation water is only 72 mN / m, making gas-water separation difficult and exacerbating gas phase retention losses, thus affecting recovery efficiency. In addition, the significant change in hydrogen compressibility under high pressure further increases the difficulty of pressure control. Third, the dynamic coupling complexity of multiphase flow in dual-porosity media. Hydrogen injection requires the displacement of formation water, resulting in gas-water two-phase migration. In the dual-media system, the fracture system responds quickly and has small reserves, serving as the primary initial migration pathway; the matrix pores respond slowly and have large reserves, requiring pressure differentials to drive fluid exchange for full utilization. This dynamic exchange is influenced by multiple factors such as pressure gradients and interfacial tension, making the gas-water distribution pattern difficult to predict and increasing the difficulty of dynamic control of the hydrogen storage reservoir. Fourthly, hydrogen loss is caused by geochemical and biological reactions. In the underground environment, hydrogen readily reacts with formation minerals, fluids, and microorganisms, resulting in irreversible loss: at the chemical level, it undergoes reduction reactions with iron oxides, etc., and at the biological level, it is metabolized into methane by hydrogen-trophic microorganisms. Studies have shown that the annual hydrogen loss rate reaches 10% to 40%, which not only reduces effective reserves but may also alter reservoir properties, complicate the composition of produced gas, and affect engineering economics and safety. Numerical simulation is a core tool for UHSA engineering design, risk assessment, and operational optimization, and its accuracy directly determines the rationality of hydrogen storage facility design and operational reliability. However, current numerical simulation technology has many shortcomings, making it difficult to meet the requirements of high-precision, full-cycle, and multi-physics coupled simulations. Specific deficiencies are as follows: First, the single-porosity model leads to distorted seepage descriptions. Traditional simulators (such as Eclipse and CMG-GEM) and most studies use homogeneous single-porosity models, ignoring fracture-matrix differences and fluid exchange dynamics, which are seriously inconsistent with actual reservoirs. Studies of existing technical solutions show that this model has a prediction error of over 30% for the gas injection front of fractured aquifers and a pressure distribution calculation deviation of over 20%, resulting in distortions in parameters such as reservoir capacity assessment and well network layout, and creating potential safety hazards. Secondly, the oversimplification of hydrogen properties negatively impacts accuracy. Under high pressure (>150 bar), the thermodynamic properties of hydrogen deviate from those of an ideal gas, with the deviation factor Z deviating by as much as 15%, and its viscosity and density exhibiting nonlinear changes with temperature and pressure. However, most studies treat it as an ideal gas, using constant property parameters, and fail to adequately consider the risk of cross-flow due to its low viscosity (only 1 / 10 that of natural gas). Existing technical solutions confirm that this simplification significantly increases the errors in pressure evolution and gas migration simulations, failing to provide reliable support for pressure control and leak prevention. Third, the lack of an adsorption model leads to biased gas storage capacity assessments. Existing technologies primarily focus on free hydrogen, neglecting its physical adsorption in coal-bearing strata and shale interlayers. In reality, hydrogen can be adsorbed on the microporous surfaces of organic matter; adsorbed gas accounts for 5%–15% of the total gas storage in coal-bearing interlayer reservoirs, and exceeds 20% in some organic-rich reservoirs. The lack of an adsorption model results in inaccurate storage capacity assessments, affecting engineering economics evaluation and scale design. Fourth, the lack of dynamic injection-production control exacerbates operational risks. Most simulations use fixed injection-production rates and fail to establish a dynamic feedback mechanism for reservoir pressure. In actual engineering, fixed-rate injection may lead to pressure exceeding the caprock rupture pressure, causing leakage; excessively rapid gas production may result in excessively low bottomhole pressure, causing formation water backflow. This static model is disconnected from dynamic operational requirements, making it difficult to ensure the long-term safe and efficient operation of the hydrogen storage facility.

[0003] In summary, UHSA technology faces multiple complex challenges. Existing numerical simulation techniques have significant shortcomings in key areas such as reservoir characterization, physical property description, reservoir mechanism coverage, and injection-production control, making it difficult to meet engineering requirements. There is an urgent need to develop numerical simulation methods that are adaptable to dual media, accurately characterize hydrogen properties and multi-field coupling, and integrate dynamic control, providing core support for UHSA engineering design optimization and safe operation. Summary of the Invention

[0004] To address the problems existing in the prior art, the purpose of this invention is to provide a numerical simulation method and system for the entire lifecycle of hydrogen storage in aquifers using a dual-porosity model. This method accurately describes the fracture-matrix two-phase flow and fluid exchange process, controls the prediction error of the injection front to within 5%, precisely calculates the PVT characteristics and adsorption-desorption behavior of hydrogen under high pressure conditions, considers actual gas effects, and can also predict the evolution of reservoir performance under long-term cyclic injection and production, guide engineering optimization design, and further provide intelligent injection and production control strategies to avoid overpressure and gas breakthrough risks.

[0005] To achieve the above objectives, the present invention provides the following solution: A numerical simulation method for the entire lifecycle of hydrogen storage in aquifers using a dual-porosity model includes: Obtain the reservoir geological boundary conditions, construct a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions, and introduce an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model. At the same time, define the dual-porosity relative permeability curve. The intelligent control scheme for injection-production wells is set using the dual-porosity relative permeability curve to constrain the dual-porosity integrated model. The parameters of the integrated model are calculated using the fully implicit finite difference method to obtain numerical simulation results. The performance of the storage facility is evaluated based on the numerical simulation results, and the storage facility performance results are obtained.

[0006] Optionally, constructing the dual-porosity geological model includes: By constructing the three-dimensional structure of the top plate barrier, the aquifer reservoir and the bottom plate barrier in sequence through the reservoir geological boundary conditions, a prototype of the dual porosity structure is formed, and an initial three-dimensional geological model is obtained. An ellipsoid is embedded in the initial three-dimensional geological model to simulate a hydrogen storage cavity, while retaining the internal boundary, thus forming a three-dimensional model containing the hydrogen storage cavity; The three-dimensional model of the hydrogen storage cavity is divided into free tetrahedral meshes, and the differentiating size parameters are customized according to the regional accuracy requirements. Physical property parameters are assigned to the matrix and fracture zones in the aquifer reservoir. The physical property parameters are assigned to the corresponding computational domain to establish a dual-medium seepage structure at the grid level, and the fluid exchange channels between the matrix and fracture zones are established through the shape factor to generate the dual-porosity geological model.

[0007] Optionally, the shape factor includes: ; Where σ is the shape factor, n is the number of crack groups, and L is the characteristic length of the matrix block.

[0008] Optionally, constructing the multi-component fluid model includes: Using the reservoir temperature and pressure conditions and pore structure of the aforementioned dual-porosity geological model as boundary constraints, a component identification mechanism is established using a four-component system of H2-CO2-H2O-N2 through hydrogen-related parameters and circulation parameters. A multi-component fluid control model is established based on the component identification mechanism, and a hydrogen PVT characteristic calculation model is introduced. At the same time, a phase equilibrium constant is defined, and the correlation between the phase equilibrium constant and temperature and pressure is determined. Furthermore, the Van Laar model is introduced to correct the ideal solution error, thus generating the multi-component fluid model.

[0009] Optionally, establishing the multi-component fluid control model includes: ; Where t is time, ε p Porosity, ρ i For component i, density u i Q is the flow rate. m,i For the mass source term, k is the permeability, and μ is the permeability. i Where ρ is the dynamic viscosity, p is the pore pressure, and g is the acceleration due to gravity. Here, z is the Hamiltonian operator, and z is the vertical depth.

[0010] Optionally, the hydrogen PVT characteristic calculation model includes: ; ; Where P is the system pressure, R is the universal gas constant, T is the thermodynamic temperature, V is the molar volume, b is the repulsion parameter, a(T) is the temperature-dependent attraction parameter, and ω is the eccentricity factor. For temperature comparison, This is a temperature correction function. This is the temperature-dependent attractive force correction factor.

[0011] Optionally, the adsorption-desorption model includes: Under simulated reservoir temperature and pressure conditions, core samples were pretreated and placed in a high-pressure adsorption instrument. Target multi-component gas was introduced, and the adsorption amount was measured at different pressure points. Experimental fitting parameters were obtained through data fitting. Substituting the experimental fitting parameters into the extended Langmuir model, the adsorption-desorption model is obtained: ; Among them, V ads,i P represents the adsorption capacity of component i. i Let b be the partial pressure of component i. i Let V be the Langmuir constant, n be the number of components, and V be the number of components. L,i Let P be the Langmuir volume of component i. j Let b be the partial pressure of component j. j Let be the Langmuir pressure constant of component j.

[0012] Optionally, the feature is that defining the dual-porosity relative permeability curve includes: The wettability characteristics of the matrix and the crack partition are obtained, and the matrix partition is matched with the strong hydrophilic Corey model and the crack partition is matched with the weak wettability X-shaped curve model respectively. Steady-state or unsteady-state methods were used to conduct core seepage experiments. After saturating the matrix core with formation water, hydrogen was injected to obtain pressure difference and flow rate data at different water contents. The relative permeability was calculated and the parameters of the strongly hydrophilic Corey model were fitted. At the same time, a wettability correction coefficient was introduced into the weakly wetted X-shaped curve model. The relative permeability curve of the dual porosity is generated by fitting the strongly hydrophilic Corey model and the weakly wetting X-shaped curve model with a wettability correction coefficient.

[0013] Optionally, the intelligent control scheme for the injection-production well includes: Set safety thresholds for the maximum gas injection rate and caprock fracture pressure, and collect the average reservoir pressure. When the average reservoir pressure exceeds the safety threshold, reduce the gas injection rate to 50%Q. injmax,x And issue a warning; among them, Q injmax The maximum gas injection rate is given by x, where x is the pressure monitoring point number. The rate is maintained when the average pressure of the reservoir is not greater than the safety threshold of the caprock fracture pressure. The well shut-in threshold is set using the bound water saturation in the dual-porosity relative permeability curve. The formation water saturated vapor pressure is collected. When the formation water saturated vapor pressure is greater than the well shut-in threshold, the rate is maintained. When the formation water saturated vapor pressure is not greater than the well shut-in threshold, the well is shut in. The well is restarted after the formation water saturated vapor pressure recovers to the target value.

[0014] To achieve the above objectives, the present invention also provides a full-cycle numerical simulation system for hydrogen storage in aquifers using a dual-porosity model, comprising: The model building module is used to obtain reservoir geological boundary conditions, construct a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions, and introduce an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model. At the same time, the dual-porosity relative permeability curve is defined. The numerical solution module is used to set up an intelligent control scheme for injection and production wells using the dual-porosity relative permeability curve, so as to constrain the dual-porosity integrated model, and to calculate the integrated model parameters using the fully implicit finite difference method to obtain numerical simulation results. The performance evaluation module is used to evaluate the storage performance based on the numerical simulation results and obtain the storage performance results.

[0015] The beneficial effects of this invention are as follows: Significantly improved prediction accuracy: This invention accurately characterizes the fracture-matrix fluid exchange through a dual-porosity model, reducing the prediction error of the gas injection front from 30% in the single-porosity model to less than 3%, and the prediction error of hydrogen recovery rate from 22% to less than 3%.

[0016] More realistic physical property description: This invention adopts the PR equation of state to consider the actual gas effect under high pressure, and the Z factor calculation error is less than 2%; the introduction of the adsorption model improves the accuracy of reserve prediction of coal-bearing strata by 15%.

[0017] Significantly improved safety: The intelligent injection-production control strategy of this invention can monitor reservoir pressure in real time and automatically adjust the gas injection rate when the pressure approaches the caprock fracture threshold, thus avoiding safety accidents. Examples show that the maximum pressure is controlled within 85% of the caprock fracture pressure.

[0018] Wide applicability: This invention can simulate various reservoir types such as sandstone, carbonate rock, and coal-bearing strata, and supports various injection schemes such as pure hydrogen, hydrogen-nitrogen mixture, and hydrogen-carbon dioxide mixture. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the dual porosity-dual permeability model according to an embodiment of the present invention; Figure 2This is a flowchart of the full-cycle numerical simulation method for hydrogen storage in aquifers using a dual-porosity model, as described in this embodiment of the invention. Figure 3 This is a diagram illustrating the evolution of the reservoir pressure field according to an embodiment of the present invention. Figure 4 This is a comparison chart of prediction errors between the single-porosity model and the dual-porosity model in this invention. Figure 5 This is a schematic diagram illustrating the effect of the intelligent injection and extraction control strategy in an embodiment of the present invention. Detailed Implementation

[0021] 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.

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] like Figure 2 As shown in the figure, this embodiment discloses a full-cycle numerical simulation method for hydrogen storage in aquifers using a dual-porosity model, including: obtaining reservoir geological boundary conditions; constructing a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions; introducing an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model; defining a dual-porosity relative permeability curve; setting an intelligent control scheme for injection and production wells using the dual-porosity relative permeability curve to constrain the dual-porosity integrated model; calculating the integrated model parameters using the fully implicit finite difference method to obtain numerical simulation results; and evaluating reservoir performance based on the numerical simulation results to obtain reservoir performance results.

[0024] Specifically, this embodiment discloses a full-cycle numerical simulation method for hydrogen storage in aquifers using a dual-porosity model. The steps proceed sequentially according to the logical order of "geological modeling → fluid model → adsorption model → seepage parameters → injection-production control → numerical solution and evaluation." The output parameters of the preceding steps directly serve as input constraints for subsequent steps: The dual-porosity geological model provides the grid structure and physical property parameters (porosity, permeability, shape factor, as shown in Table 1), defining the domain of action for the fluid and adsorption models; the multi-component fluid model and adsorption-desorption model determine the fluid components, PVT characteristics, and adsorption amount, collectively constituting the source and sink terms of the governing equations; the dual-porosity relative permeability curve provides the gas-water two-phase seepage capacity, constraining the fluid distribution ratio between the matrix and fractures; the intelligent control scheme for injection-production wells uses the reservoir pressure field output by the above models as a real-time feedback signal to dynamically adjust the injection-production rate; finally, the numerical solution module integrates all models, outputting a three-dimensional pressure field, saturation field, and component concentration field, and conducts reservoir performance evaluation based on the results. Specifically, the following steps are included: like Figure 1 As shown, a dual-porosity geological model is constructed, and a three-dimensional corner mesh model containing a matrix system and a fracture system is established, wherein the two systems are coupled through a shape factor σ. Table 1

[0025] A multi-component fluid model was established, using a four-component system of H2-CO2-H2O-N2; Introduce an adsorption-desorption model; Define the dual-porosity relative permeability curves: the matrix system uses the strongly hydrophilic Corey model, and the fracture system uses the weakly wetting X-curve to reflect the difference in wettability. The matrix ingress pressure P is used as the reference value. entry,m =3-8 kPa, crack entry pressure P entry,f =0.3-1.5 kPa; A smart control scheme for injection and production wells is implemented, with hydrogen injection wells employing dual constraints of maximum gas injection rate and maximum bottom hole pressure: reservoir average pressure P avg > 0.9×P cap When the caprock rupture pressure is reached, the gas injection rate is automatically reduced to 50% to avoid the risk of caprock rupture. Before proceeding to the numerical solution stage, the outputs of the aforementioned sub-models need to be integrated into a unified set of equations for a two-phase flow with dual porosity, serving as the core governing equations of the fully implicit finite difference method. This set of equations is not derived independently, but rather is formed by coupling each term of the preceding models according to physical mechanisms. The specific composition is as follows: (1) The gas-water two-phase mass conservation equations for the matrix and fracture systems are derived from the multi-component fluid model. This model establishes a component continuity equation for the four-component system of H2-CO2-H2O-N2, and introduces the actual gas PVT characteristics through the PR equation of state, and corrects the non-ideality of the liquid phase through the Van Laar model, thereby providing the fluid property parameters such as density, viscosity, and compressibility of each component as a function of temperature and pressure for the equation set.

[0026] (2) The fluid exchange term between the matrix and fractures is derived from the dual-porosity geological model. This model uses the shape factor σ (σ=4n / L) 2 It quantitatively characterizes the mass transfer rate from the matrix block to the fracture system, couples two independent continuity equations into a unified dual-porosity system, and enables the equation set to simultaneously describe the rapid seepage channel characteristics of the fracture system and the slow diffusion storage characteristics of the matrix system.

[0027] (3) The relative permeability in the Darcy flow term of the gas-water two-phase flow is derived from the relative permeability curve of the two-porosity system. The curve obtained by the strong hydrophilic Corey model is used for the matrix system, and the weak wetting X-type curve with the introduction of the wettability correction coefficient is used for the fracture system. Both are assigned to the Darcy equation of the corresponding computational domain, so that the equation set can accurately capture the significant difference between the matrix and the fracture in the gas-water two-phase flow capacity.

[0028] (4) The source and sink terms of adsorbed hydrogen originate from the adsorption-desorption model. The adsorption amounts V of each component calculated using the extended Langmuir model are... aea The mass conservation equation is embedded in the form of source and sink terms, so that the equation system can take into account the contribution of adsorbed hydrogen storage to the total gas storage, and avoid omitting 5% to 20% of the adsorbed gas in coal-bearing interlayer storage.

[0029] Numerical solution and performance evaluation: The fully implicit finite difference method is used to solve the equations for a two-phase flow with dual porosity: pressure convergence criterion ≤ 50 kPa, saturation convergence criterion ≤ 0.001. Output: Spatiotemporal distribution of three-dimensional pressure field, hydrogen saturation field, and component concentration field; Key indicators include working gas volume, bottom gas volume, hydrogen recovery rate, and reservoir capacity utilization rate; Multi-cycle performance evolution curves (pressure, temperature, and composition changes).

[0030] More specifically, this embodiment discloses a full-cycle numerical simulation method for aquifer hydrogen storage using a dual-porosity model. This method can fully capture the stress concentration phenomenon and complexity of the reservoir, provide accurate spatial stress distribution and hydrogen migration path analysis, and effectively reveal the mechanical behavior and potential safety risks of aquifer hydrogen storage reservoirs under multidimensional stress, providing a key reference for the feasibility and safety of underground hydrogen storage technology. Specifically, it includes the following steps: The core objective of constructing a dual-porosity geological model is to build a three-dimensional geological framework that can accurately characterize the dual media of reservoir matrix and fractures, providing fundamental support for subsequent thermal-fluid-solid multi-field coupled simulations.

[0031] Specifically, the implementation process includes defining reservoir geological boundary conditions, constructing an initial three-dimensional geological model, model simplification and optimization, mesh generation and quality verification, assigning physical property parameters and coupling with fluid exchange. The operations for each step are as follows: First, clarify the geological boundary conditions of the reservoir. Through geological survey reports, seismic exploration data, and borehole sampling analysis, determine the spatial distribution range of the target aquifer reservoir, clarify the top and bottom plate elevations, lateral extension boundaries, lithological combinations, and fracture development, and analyze key parameters such as the distribution patterns, orientation, and dip angle of the matrix and fractures. Simultaneously, collect data on the lithology and density of the overlying strata to determine the top load boundary, and combine this with hydrogeological data to clarify hydrological boundaries such as initial water pressure and groundwater flow direction.

[0032] Second, an initial three-dimensional geological model was constructed. Using the geometry module of COMSOL Multiphysics software, the three-dimensional structure of the top plate interlayer, aquifer reservoir (including matrix and fracture partitions), and bottom plate interlayer was constructed sequentially according to the geological boundary conditions, forming a preliminary biporosity structure. The model size was set to an 800m×800m×600m cube region to ensure complete capture of the pressure, temperature, and displacement field changes during hydrogen storage.

[0033] Third, the initial model was simplified and optimized. To reduce computational load, an ellipsoid was embedded in the initial model to simulate the hydrogen storage cavity. Boolean difference operations were used to preserve the internal boundaries, resulting in a three-dimensional model containing the hydrogen storage cavity. The model was then simplified to a quarter-formation model using its symmetry, significantly reducing computational costs while maintaining accuracy.

[0034] Fourth, mesh generation and quality verification. Free tetrahedral meshing is used, with customized size parameters based on regional accuracy requirements: the mesh is finer around the hydrogen storage cavity and in crack areas (maximum element size 4m, minimum 0.16m, growth rate 1.2, etc.); the mesh size is increased in matrix areas far from the cavity (maximum element size 125m, minimum 22m, growth rate 1.4, etc.). After generation, the mesh accuracy is checked to ensure there are no negative volume meshes; otherwise, the mesh is regenerated.

[0035] Fifth, the assignment of physical property parameters and coupling with fluid exchange. Based on a qualified mesh model, the matrix system (porosity φ) is coupled with fluid exchange. m Penetration rate k m (etc.) and fracture systems (porosity φ) f Penetration rate k fPhysical property parameters (e.g., matrix and fracture) are assigned based on measured data from indoor core experiments, clearly demonstrating a 6-8 order of magnitude difference in permeability between fractures and the matrix. Based on this difference, physical property parameters for both matrix and fracture elements are assigned to their respective computational domains within the mesh model: matrix system physical property parameters (low porosity and low permeability) are assigned storage functionality, while fracture system physical property parameters (high porosity and high permeability) are assigned flow channel functionality. This establishes a dual-medium seepage structure at the mesh level, laying the physical property foundation for subsequent multi-field coupled calculations. After assigning physical property parameters to both the matrix and fracture systems, a fluid exchange channel needs to be established between the two systems. The assigned physical property parameters are combined with the mesh geometry, and fluid exchange coupling is achieved through the shape factor σ. The calculation of the shape factor σ depends on the characteristic length L of the matrix block determined by the fourth step of mesh generation and the number of fracture groups n determined by the geological survey. It reflects the organic combination of the mesh geometry (result of the fourth step) and physical property parameters (result of the fifth step). Its value directly controls the mass transfer rate of fluid between the matrix and fractures and has a decisive influence on the simulation accuracy of the biporosity model. It is calculated using the classic publicly available formula of biporosity seepage mechanics: ; Where σ is the shape factor (m -2 ), where n is the number of fracture groups and L is the characteristic length of the matrix block (m). n is determined through geological survey, L is obtained by combining core data, and σ is calculated by substituting into the formula to complete the coupling setup.

[0036] The construction and parameter setting of the multi-component fluid model aims to build a multi-component fluid system that conforms to actual engineering conditions, accurately characterize the physicochemical properties of hydrogen and formation fluids, and provide a foundation for simulating hydrogen migration and phase changes. Specifically, the implementation process includes: using the reservoir temperature and pressure conditions (depth, geothermal gradient) and pore structure determined by the first step of the dual-porosity geological model as boundary constraints, and proceeding in the logical order of "parameter acquisition → component determination → governing equation establishment → PVT characteristic configuration → phase equilibrium setting". That is, firstly, the basic physical properties and cyclic load parameters of hydrogen are obtained (to provide physical property input for subsequent equations), then the fluid component system is determined (to clarify the simulation object), based on which the multi-component seepage governing equation is established (to describe the mass conservation and Darcy flow of the fluid in the geological model grid), then the hydrogen PVT characteristic model is configured (to make parameters such as density and viscosity dynamically change with temperature and pressure), and finally the phase equilibrium module is set (to capture gas-liquid phase transitions). The outputs of each sub-step are passed sequentially to jointly complete the construction of the fluid model. The connection between this model and the first-step dual-porosity geological model is reflected in the following: the fluid model's seepage equations directly operate on the matrix and fracture meshes provided by the geological model; the porosity, permeability, and other parameters of the geological model serve as coefficient inputs to the fluid equations, controlling the fluid transport rate in the two media and the mass exchange between the matrix and fractures. The coupling of these two models constitutes a complete dual-porosity multi-component seepage model. Specifically, this includes obtaining hydrogen and circulation parameters, determining the fluid component system, building the multi-component fluid model, configuring the hydrogen PVT characteristic model, and setting the phase equilibrium module. The operations for each step are as follows: First, obtain the relevant parameters and cycle parameters for hydrogen. In the COMSOL global definition module, define the molar mass of hydrogen (0.002016 kg / mol) and the dynamic viscosity (8.915 × 10⁻⁶ kg / mol). -6 Basic physical property parameters such as Pa·s (refer to the IUPAC database). Cyclic loads from 0 to 3600 days (maximum 1.5 × 10⁻⁶ Pa·s) are defined in the component module using a piecewise cubic interpolation function. 7 Pa, minimum 7×10 6 Pa) and cycling temperature (maximum 60°C, minimum 45°C).

[0037] Second, the fluid composition system and basic properties were determined. A four-component system of H2-CO2-H2O-N2 was adopted, covering injected hydrogen, formation water, and associated inert gases. Each component was added to the fluid property module, and its molar mass and critical temperature T were determined. c Critical pressure P c Based on fundamental properties (refer to authoritative physical property manuals), establish a component identification mechanism.

[0038] Third, a multi-component fluid model is constructed. Based on Darcy's law, a seepage model is built, and combined with the law of conservation of mass, the governing equations for the multi-component fluid (classical publicly available equations in seepage mechanics) are established: ; Where t is time (s); ε p Porosity (%); ρ i Density of component i (kg / m³) 3 ); u i Q is the flow velocity (m / s); m,i For mass source items (kg / (m 3 ·s); k is the permeability (m 2 ); μ i ρ is the dynamic viscosity (Pa·s); p is the pore pressure (Pa); g is the acceleration due to gravity (m / s²). 2 Seepage parameters were set separately for the matrix and the fracture system to clarify their differences.

[0039] Fourth, configure the hydrogen PVT characteristic calculation model. The publicly available Peng-Robinson (PR) equation of state is used to describe the hydrogen PVT characteristics, applicable to the temperature and pressure range of T∈[25, 200]℃ and P∈[3.5, 700]bar. Select the PR equation in the software and input the hydrogen critical temperature T. c, H2 Critical pressure P c, H2 By setting parameters such as density and viscosity, the correlation equations are derived. The core expression of the PR equation is: ; Where P is the system pressure (Pa); R is the universal gas constant; T is the thermodynamic temperature (K); and V is the molar volume (m³). 3 ( / mol); b is the repulsion parameter; a(T) is the temperature-dependent attraction parameter. The parameters are calculated using the critical parameters for hydrogen, ensuring a Z-factor error ≤ 2%.

[0040] Fifth, set up the phase equilibrium calculation module. Define the phase equilibrium constant K. i (Molar fraction ratio between gas and liquid phases), combined with the PR equation and the criterion for equal chemical potential, establish the phase equilibrium constant K. i The relationship with temperature and pressure. The Van Laar model (a publicly available classical activity coefficient model) is introduced to correct for errors in the ideal solution; the expression is: ; In the formula, γ1 and γ2 are activity coefficients; A 12 A 21 x1 and x2 are model parameters (experimental fit); x1 and x2 are mole fractions. Accurately captures phase changes such as hydrogen dissolution and precipitation.

[0041] The adsorption-desorption model is introduced and its parameters are configured. This step is related to the two models mentioned above as follows: First, the adsorption-desorption model operates on the matrix grid units (matrix regions corresponding to organic matter interlayers) established by the first-step dual-porosity geological model. The matrix porosity, grid size, and other parameters provided by the geological model determine the domain of action of the adsorption model and the spatial distribution of adsorption. Second, the dynamic change in adsorption with pressure is coupled into the seepage control equations of the second-step multi-component fluid model through mass source and sink terms. That is, when the pressure increases, the increase in adsorbed gas acts as a negative source term (fluid is captured by the solid phase), and when the pressure decreases, the increase in desorbed gas acts as a positive source term (solid phase releases hydrogen into the fluid), thus achieving closed-loop coupling with the migration process of free hydrogen and jointly ensuring mass conservation. Therefore, the adsorption-desorption model is not an isolated link, but rather uses the geological model as the spatial carrier and the fluid model as the equation framework to incorporate adsorbed hydrogen storage into a complete dual-porosity multi-component simulation system. Specifically, for aquifer reservoirs containing organic matter interlayers, it accurately characterizes the hydrogen adsorption-desorption effect to avoid underestimating the storage capacity. Specifically, the implementation process includes selecting an adsorption model, obtaining model parameters, assigning parameter values, and model coupling. The operations for each step are as follows: First, select a suitable adsorption model. Use the extended Langmuir adsorption model (an extension of the classic model, with publicly available theoretical sources). This model is suitable for multi-component fluid competitive adsorption scenarios and can accurately describe the competitive adsorption behavior of hydrogen with CO2, N2, etc., on the matrix surface. The competitive adsorption coefficient reflects the actual reservoir adsorption behavior.

[0042] Second, indoor adsorption experiments were conducted to obtain parameters. Under simulated reservoir temperature and pressure conditions, core samples were pretreated and placed in a high-pressure adsorption instrument. Target multi-component gas was introduced, and the adsorption amount was measured at different pressure points. Key parameters of the extended Langmuir model (maximum adsorption amount and Langmuir pressure) were obtained through data fitting.

[0043] Third, complete the model parameter assignment. Substitute the experimental fitting parameters into the standard formula of the extended Langmuir model: ; Among them, V ads,i P represents the adsorption capacity of component i. i Let b be the partial pressure of component i. i Let n be the Langmuir constant, n be the number of components, and P be the number of components. j Let b be the partial pressure of component j. j The Langmuir pressure constant (MPa) of component j -1 ).

[0044] Fourth, model coupling is achieved. The adsorption / desorption module is linked to the geological-fluid model within the software, with the scope of action limited to matrix mesh elements (fracture adsorption effects are negligible). Dynamic equilibrium conditions are set: adsorption dominates when pressure increases, leading to a greater adsorption amount; desorption dominates when pressure decreases, with desorbed hydrogen participating in fluid transport, thus achieving coordinated simulation of adsorption-desorption, fluid flow, and stress changes.

[0045] The definition of the dual-porosity relative permeability curve is related to other steps as follows: First, the relative permeability curves are defined separately for the matrix system and fracture system established in the first step of the geological model. The dual-grid structure of the geological model necessitates the configuration of two sets of differentiated curves, rather than a single curve. Second, the relative permeability curve is a key coefficient term in the seepage control equation of the second step multi-component fluid model, determining the effective seepage capacity ratio of the gas phase and water phase in their respective media at a given water saturation, directly affecting the solution results of the fluid equation. Third, the bound water saturation and residual gas saturation parameters determined in this step are also mutually constrained with the scope of the adsorption-desorption model in the third step and the start-stop threshold of the injection-production well control scheme, jointly ensuring the reasonable distribution of gas and water saturation during the numerical solution process. The differentiated relative permeability curves characterize the differences in wettability and seepage capacity between the matrix and fracture systems, providing core parameters for gas-water two-phase seepage simulation. Specifically, the implementation process includes determining wettability characteristics, selecting a model, obtaining parameters, generating curves, and verifying rationality. The operations of each step are as follows: First, determine the wettability characteristics. Through contact angle measurement experiments and the Amott method, determine the wettability type between the matrix (mainly strongly hydrophilic) and the crack (weakly hydrophilic or neutral), and clarify the determining role of wettability in the morphology of the relative permeability curve; Second, a suitable relative permeability model was selected. The strongly hydrophilic Corey model (a classic publicly available model) was used for the matrix system, and the weakly wetting X-curve model (a publicly available weakly wetting medium model) was used for the fracture system, to match the seepage characteristics of the two types of media respectively; Third, obtain model parameters. Conduct core seepage experiments using steady-state or unsteady-state methods. After saturating the matrix core with formation water, inject hydrogen gas and record pressure differences and flow rates at different water cuts. Calculate the relative permeability and fit the Corey model parameters. The Corey model expression is: ; Where, k rw k rg The relative permeability of the aqueous and gas phases; k rw,max k rg,max S represents the maximum relative permeability. w Water saturation; S wr S represents the bound water saturation. gr n represents the residual gas saturation.w n g The Corey index is used. The matrix (S) is obtained through fitting. wrm S grm etc.) and cracks (S wrf S grf (etc.) System parameters, specifying matrix bound water saturation of 20%-40%, residual gas saturation of 10%-25%, and corresponding crack saturation of 5%-15% and 3%-10%; Fourth, the Corey model parameters (n) obtained from the third step of fitting are... w n g Substitute the saturation and exponent at equal endpoints into the model to generate a relative permeability curve. Input two types of system parameters into the software's seepage parameter module to generate a curve based on the selected model. The X-shaped curve of the fracture system needs to be adjusted by introducing a wettability correction coefficient (0.6-0.9). Fifth, complete the curve rationality verification. Check the curve trend's compatibility with wettability, its continuity within the saturation range, and the rationality of the boundary saturation values ​​(S). w =S wr k rw =0, etc.) and symmetry, monotonicity, etc. If they do not meet the requirements, adjust the parameters or replace the model.

[0046] The intelligent control scheme setting for injection and production wells, and its relationship with other steps, are as follows: First, the pressure safety threshold (P) of the injection and production control scheme. safe = 0.9×P frac The first pressure is directly taken from the caprock fracture pressure determined by the geological model in the first step. The geological model's description of the reservoir boundary and caprock mechanical properties is a prerequisite for setting reasonable pressure constraints. The second pressure is the reservoir average pressure P monitored by the control logic. avg and bottom hole pressure P bh All pressure field results are derived from the real-time solution of the multi-component fluid model on a dual-porosity grid in the second step. The fluid model serves as the "sensing layer" for the control strategy. Third, the minimum bottom-hole pressure threshold for the gas production well must also refer to the bound water saturation determined in the relative permeability curve in the fourth step to prevent excessively low bottom-hole pressure from causing gas-water interface instability. Fourth, the dynamic adjustment of the injection-production rate is fed back to the fluid model and adsorption-desorption model as boundary conditions, triggering multi-field coupling solutions in the next time step, forming a closed-loop iteration of "simulation → monitoring → regulation → re-simulation." This provides safe and reliable operational data for the final performance evaluation, constructs a dynamic feedback intelligent control strategy, ensures the safe operation of the hydrogen storage facility, and guarantees stable and efficient injection and production. Specifically, the implementation process includes setting dual constraints for hydrogen injection wells and gas production wells, coupling control logic, and strategy verification and optimization. The operations for each step are as follows: First, a dual-constraint control mechanism is established for hydrogen injection wells. This mechanism employs a constraint of "maximum gas injection rate + maximum bottom hole pressure," with the maximum gas injection rate Q set.injmax (Based on equipment and reservoir capacity, a safety margin of 10%-20% should be reserved); the caprock fracture pressure P should be obtained through experiments. frac Set a safety threshold P safe =0.9P frac Set up at least 3 pressure monitoring points to collect P data in real time. avg Control logic: P avg >P safe Reduce the injection rate to 50%Q injmax And issue a warning; P avg ≤P safe Maintain the rate or adjust it as planned; Second, a dual-constraint control mechanism is established for the gas production wells. This mechanism employs a constraint of "maximum gas production rate + minimum bottom hole pressure," with the maximum gas production rate Q set. prodmax (Used as a rate limit constraint for gas production wells in numerical simulations to prevent excessively rapid extraction from causing excessive formation pressure drop, aquifer water intrusion, or hydrogen storage loss. Together with the minimum bottom hole pressure constraint, it forms a dual control mechanism to ensure the safe and stable operation of the hydrogen storage system.) (Reserving a safety margin); Obtaining the formation water saturated vapor pressure P vap Set the shut-in threshold P stop =P vap +0.5~1MPa. Monitoring points were set up at the bottom of the well to collect P data. bh Control logic: P bh >P stop Maintenance rate; P bh ≤P stop When the well is shut off, the pressure returns to P. stop Restart after +1~2MPa; Third, establish the coupling relationship between the control logic and the geological-fluid model. Deeply couple the injection-production control module with the geological and fluid models to construct a closed-loop system of "monitoring-feedback-regulation": monitoring data is input to the control module, and when the threshold is exceeded, regulation commands are generated to adjust the injection-production rate. The regulation parameters act in reverse on the model, and the new field distribution data is fed back to the control module to form a cycle. A sampling interval of 1-6 hours is set to ensure timely response. Fourth, conduct verification and optimization of control strategies. Simulate normal injection and production, peak gas injection and other operating conditions to verify the strategy's effectiveness in suppressing pressure fluctuations, ensuring efficiency and mitigating risks, and optimize parameters such as the safety factor γ and pressure margin ΔP to achieve safe and efficient operation.

[0047] Specifically, the implementation process includes configuring solution parameters, setting time strategies, running simulations, outputting results, and performance evaluation. The operations for each step are as follows: 1. Configure numerical solution parameters. The fully implicit finite difference method (a publicly available classical method) is used for the solution. In COMSOL, the parameters are configured as follows: linear solvers are selected as the PARDISO method (steady-state) and the MUMPS method (transient); nonlinear solvers use the Newton method with a maximum of 25 iterations; the transient solution uses an implicit backward difference formula. The convergence criterion is set as: pressure ε. p ≤50kPa, saturation ε s ≤0.001, displacement ε u ≤1×10 -4 m; 2. Set the simulation time strategy. Adopt a "variable time step" mode: Δt1 = 1 day in the initial injection and production phase (first 100 days), Δt2 = 5 days in the middle phase (100-1000 days), and Δt3 = 10 days in the later phase (1000-3600 days); set an automatic adjustment mechanism (reduce the step size by 50% when the pressure change exceeds 500 kPa, and increase it by 20% when it is less than 50 kPa). Set 3-5 complete injection and production cycles (1-year cycle), and achieve automatic switching through the stage control module; 3. Run the numerical simulation. Activate each model in sequence, check the integrity of the parameters, and then start the solver: first calculate the initial ground stress in steady state, then calculate the thermal-fluid-solid multi-field coupling data in transient state, monitor the convergence status in real time, and solve non-convergence problems by adjusting the mesh, number of iterations, etc. 4. Output simulation results. Output three-dimensional spatial pressure field, temperature field, and other spatiotemporal evolution data, key parameters of the injection and production stages, and well operation parameters. Calculate the hydrogen storage tank volume shrinkage rate using the following formula: ; Among them, V before The initial volume (m 3 ); V after Volume (m) after simulation 3 Generate cloud maps and contour maps to visually display the dynamic changes of the reservoir; 5. Conduct performance evaluation of hydrogen storage facilities. Quantitatively evaluate performance from three dimensions: storage performance (hydrogen storage capacity, recovery rate, etc.), operational efficiency (injection and production efficiency, energy loss, etc.), and safety and stability (caprock displacement, maximum principal stress, etc.). Based on the results, identify risk points and propose engineering recommendations such as adjusting cavity dimensions and optimizing injection and production parameters, providing data support for actual engineering construction.

[0048] like Figure 3 As shown, Figure 3 This is a reservoir pressure field evolution diagram according to an embodiment of the present invention. The diagram illustrates the spatiotemporal evolution of the three-dimensional pressure field of the matrix system and fracture system within the aquifer hydrogen reservoir during a complete injection-production cycle. In the initial stage of hydrogen injection, high-pressure hydrogen preferentially and rapidly expands along the fracture system, causing the fracture pressure to quickly rise to the target pressure range, while the matrix pressure, due to its low permeability (10...),... -10 ~10-9 The response lag (on the order of md) is approximately several days to weeks. As injection continues, the pressure difference between the matrix and fractures drives fluid to infiltrate into the matrix through exchange channels coupled with shape factors. The matrix pressure gradually increases and approaches the fracture pressure, forming a quasi-steady-state distribution. The hydrogen extraction phase exhibits the opposite process: the fracture pressure decreases first, and the matrix replenishes fluid to the fracture, maintaining a certain production pressure difference. From the spatial distribution of the pressure field, the pressure gradient is largest around the hydrogen storage cavity and gradually decreases towards the surrounding aquifer. The pressure below the caprock never exceeds the caprock rupture pressure threshold (set as 90% of the maximum injection pressure), verifying the effective control of overpressure risk by the intelligent control scheme. This figure visually demonstrates that the dual-porosity model can accurately capture the fracture-matrix pressure difference that the single-porosity model cannot characterize, which is a key manifestation of the core technological advantage of this invention.

[0049] like Figure 4 As shown, Figure 4 This figure compares the prediction errors of the monopore and dual-porosity models in this invention. By comparing the simulation errors of the two models in three core indicators—gas injection front location prediction, reservoir mean pressure evolution, and hydrogen recovery rate—the figure quantitatively demonstrates the accuracy improvement of the dual-porosity model compared to the traditional monopore model. Regarding gas injection front prediction, the monopore model treats the matrix and fractures as homogeneous media, ignoring the permeability difference of 6-8 orders of magnitude between them, leading to severe distortion in gas migration path prediction, with an error exceeding 30%. The dual-porosity model of this invention, by independently describing the seepage characteristics of the matrix and fractures and introducing a shape factor coupling exchange term, controls the front prediction error to within 3%. Regarding reservoir pressure prediction, the monopore model, by ignoring the fracture-matrix pressure hysteresis effect, has a pressure distribution calculation deviation exceeding 20%. The pressure field prediction deviation of the model of this invention is reduced to within 5%, and the convergence criterion satisfies a pressure error ≤ 50 kPa. In terms of hydrogen recovery prediction, the single-porosity model systematically underestimates the amount of hydrogen retained in the matrix, with a recovery prediction error of up to 22%. The model of this invention reduces this error to less than 3%, significantly improving the reliability of reservoir performance evaluation and providing more accurate data support for engineering optimization design.

[0050] like Figure 5 As shown, Figure 5 This diagram illustrates the effect of the intelligent injection-production control strategy according to an embodiment of the present invention. The diagram shows the real-time regulation effect of the intelligent control scheme on reservoir pressure and production rate during multi-cycle injection-production processes, intuitively demonstrating the effectiveness of the dynamic safety control mechanism of the present invention. During the hydrogen injection phase, the system continuously collects the average reservoir pressure P. a ᵥ g When P a ᵥ gWhen the pressure approaches the safety threshold of the caprock fracture pressure, the control system automatically reduces the gas injection rate to 50% of the maximum injection rate and triggers an early warning. As shown in the figure, the gas injection rate curve exhibits a step-like automatic reduction, with the reservoir pressure increase slowing significantly during the corresponding period, consistently remaining within the safe range of 85% of the caprock fracture pressure. During the hydrogen production phase, the system uses the formation water saturated vapor pressure corresponding to the bound water saturation level as the shut-in threshold. When the monitored value falls below the threshold, the well is automatically shut off, and gas production is restarted after the pressure recovers to the target value. The periodic characteristics of shut-in-pressure recovery-restart are clearly visible in the figure, effectively avoiding the risk of sudden pressure drops at the bottom of the well due to formation water backflow and excessive gas extraction. Overall, after regulation by the intelligent control strategy, the reservoir pressure fluctuation amplitude during multi-cycle operations is effectively suppressed, and the hydrogen recovery rate remains within the preset target range. The reservoir operation safety and production efficiency are significantly better than the fixed-rate injection-production scheme, verifying the engineering practical value of the intelligent control scheme for injection-production wells of this invention.

[0051] This embodiment also discloses a full-cycle numerical simulation system for hydrogen storage in aquifers using a dual-porosity model, comprising: a model construction module for acquiring reservoir geological boundary conditions, constructing a dual-porosity geological model and a multi-component fluid model based on these conditions, and introducing an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model, while defining a dual-porosity relative permeability curve; a numerical solution module for setting an intelligent control scheme for injection and production wells using the dual-porosity relative permeability curve to constrain the dual-porosity integrated model, and calculating the integrated model parameters using a fully implicit finite difference method to obtain numerical simulation results; and a performance evaluation module for evaluating reservoir performance based on the numerical simulation results to obtain reservoir performance results.

[0052] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A numerical simulation method for the entire lifecycle of hydrogen storage in aquifers using a dual-porosity model, characterized in that, include: Obtain the reservoir geological boundary conditions, construct a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions, and introduce an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model. At the same time, define the dual-porosity relative permeability curve. Constructing the dual-porosity geological model includes: By constructing the three-dimensional structure of the top plate barrier, the aquifer reservoir and the bottom plate barrier in sequence through the reservoir geological boundary conditions, a prototype of the dual porosity structure is formed, and an initial three-dimensional geological model is obtained. An ellipsoid is embedded in the initial three-dimensional geological model to simulate a hydrogen storage cavity, while retaining the internal boundary, thus forming a three-dimensional model containing the hydrogen storage cavity; The three-dimensional model of the hydrogen storage cavity is divided into free tetrahedral meshes, and the differentiating size parameters are customized according to the regional accuracy requirements. Physical property parameters are assigned to the matrix and fracture zones in the aquifer reservoir. The physical property parameters are assigned to the corresponding computational domain to establish a dual-medium seepage structure at the grid level, and the matrix and fracture partition fluid exchange channels are established through the shape factor to generate the dual-porosity geological model. Constructing the multi-component fluid model includes: Using the reservoir temperature and pressure conditions and pore structure of the aforementioned dual-porosity geological model as boundary constraints, a component identification mechanism is established using a four-component system of H2-CO2-H2O-N2 through hydrogen-related parameters and circulation parameters. A multi-component fluid control model is established based on the component identification mechanism, and a hydrogen PVT characteristic calculation model is introduced. At the same time, a phase equilibrium constant is defined, and the correlation between the phase equilibrium constant and temperature and pressure is determined. Furthermore, the Van Laar model is introduced to correct the ideal solution error, and the multi-component fluid model is generated. The adsorption-desorption model includes: Under simulated reservoir temperature and pressure conditions, core samples were pretreated and placed in a high-pressure adsorption instrument. Target multi-component gas was introduced, and the adsorption amount was measured at different pressure points. Experimental fitting parameters were obtained through data fitting. Substituting the experimental fitting parameters into the extended Langmuir model, the adsorption-desorption model is obtained: ; Among them, V ads,i P represents the adsorption capacity of component i. i Let b be the partial pressure of component i. i Let V be the Langmuir constant, n be the number of components, and V be the number of components. L,i Let P be the Langmuir volume of component i. j Let b be the partial pressure of component j. j Let J be the Langmuir pressure constant of component j; The intelligent control scheme for injection-production wells is set using the dual-porosity relative permeability curve to constrain the dual-porosity integrated model. The parameters of the integrated model are calculated using the fully implicit finite difference method to obtain numerical simulation results. The performance of the storage facility is evaluated based on the numerical simulation results, and the storage facility performance results are obtained.

2. The method for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model according to claim 1, characterized in that, The shape factor includes: ; Where σ is the shape factor, n is the number of crack groups, and L is the characteristic length of the matrix block.

3. The method for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model according to claim 1, characterized in that, Establishing the multi-component fluid control model includes: ; Where t is time, ε p Porosity, ρ i For component i, density u i Q is the flow rate. m,i For the mass source term, k is the permeability, and μ is the permeability. i Where ρ is the dynamic viscosity, p is the pore pressure, and g is the acceleration due to gravity. Here, z is the Hamiltonian operator, and z is the vertical depth.

4. The numerical simulation method for hydrogen storage in aquifers throughout the entire lifecycle using a dual-porosity model according to claim 1, characterized in that, The hydrogen PVT characteristic calculation model includes: ; ; Where P is the system pressure, R is the universal gas constant, T is the thermodynamic temperature, V is the molar volume, b is the repulsion parameter, a(T) is the temperature-dependent attraction parameter, and ω is the eccentricity factor. For temperature comparison, This is a temperature correction function. This is the temperature-dependent attractive force correction factor.

5. The method for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model according to claim 1, characterized in that, The definition of the dual-porosity relative permeability curve includes: The wettability characteristics of the matrix and the crack partition are obtained, and the matrix partition is matched with the strong hydrophilic Corey model and the crack partition is matched with the weak wettability X-shaped curve model respectively. Steady-state or unsteady-state methods were used to conduct core seepage experiments. After saturating the matrix core with formation water, hydrogen was injected to obtain pressure difference and flow rate data at different water contents. The relative permeability was calculated and the parameters of the strongly hydrophilic Corey model were fitted. At the same time, a wettability correction coefficient was introduced into the weakly wetted X-shaped curve model. The relative permeability curve of the dual porosity is generated by fitting the strongly hydrophilic Corey model and the weakly wetting X-shaped curve model with a wettability correction coefficient.

6. The method for full-cycle numerical simulation of hydrogen storage in aquifers using a dual-porosity model according to claim 1, characterized in that, The intelligent control scheme for the injection and production wells includes: Set safety thresholds for the maximum gas injection rate and caprock fracture pressure, and collect the average reservoir pressure. When the average reservoir pressure exceeds the safety threshold, reduce the gas injection rate to 50%Q. injmax,x And issue a warning; among them, Q injmax The maximum gas injection rate is given by x, where x is the pressure monitoring point number. The rate is maintained when the average pressure of the reservoir is not greater than the safety threshold of the caprock fracture pressure. The well shut-in threshold is set using the bound water saturation in the dual-porosity relative permeability curve. The formation water saturated vapor pressure is collected. When the formation water saturated vapor pressure is greater than the well shut-in threshold, the rate is maintained. When the formation water saturated vapor pressure is not greater than the well shut-in threshold, the well is shut in. The well is restarted after the formation water saturated vapor pressure recovers to the target value.

7. A numerical simulation system for the entire lifecycle of hydrogen storage in aquifers using a dual-porosity model implemented according to any one of claims 1-6, characterized in that, include: The model building module is used to obtain reservoir geological boundary conditions, construct a dual-porosity geological model and a multi-component fluid model based on the reservoir geological boundary conditions, and introduce an adsorption-desorption model to incorporate adsorbed hydrogen storage into the dual-porosity multi-component simulation system to generate a dual-porosity integrated model. At the same time, the dual-porosity relative permeability curve is defined. The numerical solution module is used to set up an intelligent control scheme for injection and production wells using the dual-porosity relative permeability curve, so as to constrain the dual-porosity integrated model, and to calculate the integrated model parameters using the fully implicit finite difference method to obtain numerical simulation results. The performance evaluation module is used to evaluate the storage performance based on the numerical simulation results and obtain the storage performance results.

Citation Information

Patent Citations

  • Underground salt cavern hydrogen storage reservoir site selection coupling numerical simulation method

    CN116415317A

  • Simulation calculation method for studying influence of pore structure deformation on underground aquifer hydrogen storage efficiency

    CN121328091A