A system and method for joint inversion and quantification of water pressure, temperature and stress in a geological body

Abstract

The present application relates to geophysical exploration technology, disclose a kind of geologic body water pressure, temperature and stress joint inversion quantification system and method;The method first fuses well point monitoring and surface area geophysics and other multi-source heterogeneous data, form physical consistent constraint;Based on three-dimensional geological model dynamic configuration and update the nonlinear coupling relationship of permeability, thermal conductivity and other stress, temperature and strain history;Further construct heat-water-force coupling forward model, using the hybrid algorithm of fusion bayesian inference and gradient optimization, simultaneously inverted to three field distribution and its posteriori uncertainty;Finally, risk quantification and output engineering decision;Corresponding system includes multi-source heterogeneous data fusion, dynamic coupling coefficient configuration, adaptive inversion solution and decision support and visualization four modules;The present application significantly improves the inversion accuracy and reliability under complex geological conditions such as fracture zone, thermal shock zone, provides high confidence decision basis for shale gas fracturing, geothermal development and other engineering.

Description

Technical Field

[0001] This invention belongs to the field of geophysical exploration and geotechnical engineering monitoring technology, specifically relating to a joint inversion and quantification system and method for water pressure, temperature and stress in geological bodies. Background Technology

[0002] In deep resource extraction, geothermal energy development, and the construction of major underground engineering projects, the water pressure field, temperature field, and stress field within a geological body constitute a coupled and dynamically evolving thermo-hydraulic-mechanical system. Accurately understanding the three-dimensional distribution and evolution of this system is a key scientific issue and engineering foundation for assessing engineering safety, optimizing resource recovery, and providing early warning of geological hazards. However, the internal structure of geological bodies is complex and cannot be directly observed. How to utilize limited, multi-source, heterogeneous field monitoring data to accurately reconstruct the true state of the three physical fields of water pressure, temperature, and stress through joint inversion is a core challenge currently facing the fields of geophysical exploration and geotechnical engineering monitoring.

[0003] Existing geological inversion techniques generally employ simplified physical models, often ignoring or linearly approximating the complex coupling relationships between multiple fields. This makes it difficult to accurately depict the nonlinear effects of temperature on fluid properties and stress on fracture permeability under high temperature and high pressure conditions. At the data level, existing methods primarily rely on sparse borehole data, lacking deep fusion with large-scale geophysical data such as InSAR deformation and seismic imaging based on a unified physical mechanism. This limits the improvement of spatial resolution and the effectiveness of global constraints. Regarding inversion methods, most algorithms are limited to deterministic optimization, easily getting trapped in local optima and lacking the ability to quantify the posterior uncertainty of model parameters, making it difficult to assess the reliability of the inversion results. Furthermore, existing system architectures are typically rigid, with coupling parameters and boundary conditions unable to be dynamically adjusted according to actual geological structures and engineering disturbance stages. They also lack adaptive optimization mechanisms for the inversion process, making it difficult to balance computational efficiency and inversion accuracy when dealing with highly heterogeneous geological bodies.

[0004] The aforementioned shortcomings collectively result in significant discrepancies between the inversion results and the actual physical conditions when existing technologies are applied in deep and complex geological environments (such as fractured zones, active fault zones, and areas disturbed by long-term mining), leading to insufficient predictive reliability. This makes it difficult for existing technologies to meet the refined and intelligent decision-making needs of high-risk projects such as efficient fracturing of shale gas, safe operation of enhanced geothermal systems, and stability control of surrounding rock in deep tunnels.

[0005] Therefore, there is an urgent need to develop a method and system for the joint inversion and quantification of water pressure, temperature and stress that can deeply integrate multi-source heterogeneous data, embed dynamic coupling physical mechanisms, have adaptive inversion solution capabilities, and perform strict uncertainty quantification on the results, so as to improve the accuracy of understanding the multi-physical field state of geological bodies and the reliability of engineering applications. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a system and method for the joint inversion and quantification of water pressure, temperature and stress in geological bodies.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] This application provides a joint inversion and quantification system for water pressure, temperature, and stress in geological bodies, including:

[0009] The multi-source heterogeneous data fusion module is used to acquire and fuse observation data from point sensors and area exploration equipment to generate a spatiotemporally correlated multidimensional observation dataset.

[0010] The dynamic coupling coefficient configuration module is used to dynamically establish and update the dependencies between permeability, thermal conductivity and fluid properties and stress, temperature and deformation history based on the stratigraphic structure attributes and engineering conditions represented by the three-dimensional geological model.

[0011] The adaptive inversion solution module is used to construct a thermal-hydraulic-mechanical coupled forward model embedding the aforementioned dependencies, and based on the aforementioned multidimensional observation dataset, it uses a hybrid inversion algorithm that integrates Bayesian inference and deterministic optimization to simultaneously invert the distribution and uncertainties of the water pressure field, temperature field, and stress field within the geological body.

[0012] The decision support and visualization module is used to perform quantitative analysis on the field distribution and uncertainty obtained from the inversion, generate engineering assessments or decision recommendations, and provide visualization output.

[0013] Optionally, the multi-source heterogeneous data fusion module is specifically used for:

[0014] Spatiotemporal registration and gridding alignment are performed on observation data from different sources and formats;

[0015] Based on physical mechanisms, a transformation relationship is established to map inverted field variables to various types of observation data;

[0016] Based on the error characteristics of the observation data, an observation error covariance matrix is ​​constructed for weighting in the hybrid inversion algorithm.

[0017] Optionally, the dynamic coupling coefficient configuration module is specifically used for:

[0018] Based on the lithological zoning and structural features identified in the three-dimensional geological model, an initial set of physical property response parameters is assigned to different regions.

[0019] During the iterative process of the adaptive inversion solution module, the permeability and thermal conductivity of each region are dynamically updated based on the stress field, temperature field and strain history obtained from the current inversion.

[0020] Optionally, in the adaptive inversion solution module:

[0021] The thermo-hydraulic-mechanical coupled forward model is constructed based on the mass, energy, and momentum conservation equations.

[0022] The hybrid inversion algorithm aims to maximize the posterior probability. Its objective function includes observation data fitting terms weighted by the observation error covariance matrix and prior constraint terms for model parameters.

[0023] Optionally, the adaptive inversion solution module is further configured to:

[0024] During the inversion process, the data fitting residuals and convergence status are monitored.

[0025] When a poor fit occurs in a specific geological structure region, the constraint strength of the model parameters in that region or the mesh resolution of the local model is automatically adjusted.

[0026] Secondly, this application provides a method for the joint inversion and quantification of water pressure, temperature, and stress in a geological body, including:

[0027] Acquire and fuse multi-source heterogeneous observation data to form a spatiotemporal dataset for joint inversion constraints;

[0028] Based on a three-dimensional geological model and engineering information, a coupling relationship model between water pressure, temperature and stress field is established and dynamically updated.

[0029] A forward numerical model embedded in the coupling relationship model is constructed, and a hybrid inversion algorithm is used to simultaneously solve for the optimal estimates of the water pressure field, temperature field and stress field and their posterior uncertainties, with the spatiotemporal dataset as constraints.

[0030] Engineering risk analysis and decision support are based on the inversion results.

[0031] Optionally, acquiring and fusing multi-source heterogeneous observation data includes:

[0032] The point-measured time-series data are processed to construct the initial field sequence;

[0033] The area exploration data is resampled to a unified grid that matches the three-dimensional geological model;

[0034] Correlating different observation types with core field variables through physical constitutive or empirical relations;

[0035] Quantify and utilize the errors and spatial correlations of observation data.

[0036] Optionally, establishing and dynamically updating the coupling relationship model includes:

[0037] Based on the differences in lithological zoning and structural features in the three-dimensional geological model, an initial set of coupling coefficients is assigned;

[0038] In the inversion iteration, the permeability and thermal conductivity are corrected in real time through a preset function based on the updated state of the field variables.

[0039] Optionally, the use of a hybrid inversion algorithm includes:

[0040] A gradient optimization algorithm is used for fast deterministic optimization.

[0041] The Markov chain Monte Carlo sampling method is used to probabilistically explore the neighborhood of the optimal solution in order to quantify the uncertainty of parameters.

[0042] Optionally, the engineering risk analysis and decision support based on the inversion results includes at least one of the following applications:

[0043] Calculate the probability of rock mass failure and assess stability risk;

[0044] Identify the dominant fluid transport path or the extent of thermal shock impact;

[0045] Provide suggestions for segment cluster selection and pumping procedure optimization for hydraulic fracturing projects;

[0046] Provide suggestions for optimizing injection and production well networks and circulation strategies for geothermal engineering.

[0047] Compared with the prior art, this application has the following beneficial effects:

[0048] By establishing a transformation operator based on physical mechanisms, a deep fusion of physical consistency between point measurement data and surface geophysical data was achieved. Optimal weighting was performed using an error covariance matrix incorporating spatial correlation, effectively overcoming the limitations of traditional methods that rely on sparse point measurement data. This provides a high-confidence, high-spatial-coverage full-space constraint for inversion, significantly improving the global accuracy and resolution of multi-physics field reconstruction. Furthermore, by introducing a dynamic coupling coefficient configuration mechanism based on a three-dimensional geological model, the nonlinear dependence of physical property parameters on stress, temperature, and strain history can be initialized and updated online according to lithological zoning and structural characteristics. This enables the inversion model to realistically depict the thermo-hydraulic-mechanical coupling process under strongly heterogeneous and nonlinear geological conditions. This significantly improves the physical accuracy of the inversion results and their applicability in complex geological areas. Employing a hybrid inversion algorithm framework integrating Bayesian inference and deterministic optimization, it efficiently seeks the optimal solution using gradient optimization while rigorously quantifying the posterior probability distribution of model parameters through Markov chain Monte Carlo sampling. This not only yields more reliable optimal estimates but also provides comprehensive uncertainty information for the inversion results (hydraulic pressure field, temperature field, and stress field) for the first time, offering crucial reliability assessment for high-risk engineering decisions. Through the system's built-in inversion process monitoring and adaptive strategy adjustment logic, it can dynamically optimize local parameter constraints and grid resolution based on real-time data fitting, enabling the system to possess intelligent adaptive capabilities. Finally, through quantitative risk analysis and 3D visualization, it directly outputs decision-making recommendations for specific projects such as shale gas fracturing and geothermal development, achieving a closed loop from data inversion to intelligent decision-making, significantly enhancing the technology's practicality and engineering value. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the overall technical solution architecture of the present invention.

[0050] Figure 2 This is a schematic diagram of the principle framework of the unified THM forward model and hybrid inversion algorithm that embeds dynamic coupling constitutive relations.

[0051] Figure 3 It is a logical flowchart of the fusion of multi-source heterogeneous observation data and the construction of multi-dimensional observation datasets.

[0052] Figure 4 This is a schematic diagram of the dynamic coupling coefficient configuration and update mechanism based on lithology and engineering disturbance.

[0053] Figure 5 This is a data flow diagram illustrating the quantitative analysis of inversion results and support for engineering decisions. Detailed Implementation

[0054] 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.

[0055] Furthermore, in this invention, an element referred to as fixed to or disposed on another element may be directly disposed on the other element, or there may be an intermediate element. When an element is considered to be connected to another element, it may be directly connected to the other element, or there may be an intermediate element present simultaneously. The terms vertical, horizontal, left, right, and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0056] The core of the water pressure-temperature-stress joint inversion quantification system and method for geological bodies provided by this invention lies in constructing a complete technical system from multi-source data acquisition and fusion, dynamic coupled physical modeling, adaptive intelligent inversion to visualization decision support. The entire scheme uses a unified three-dimensional geological model as its spatial framework. The system first integrates time-series data acquired by various field measuring instruments (such as water pressure gauges, temperature sensors, and strain gauges) with large-scale surface geophysical data such as InSAR deformation fields, seismic wave velocity imaging, and resistivity cross-sections through a multi-source heterogeneous data fusion module. Based on rigorous physical mechanism and error statistical analysis, this integration forms a spatiotemporally registered, physically consistent, and reliability-weighted multidimensional observation dataset, providing high-confidence global constraints for subsequent inversion. Subsequently, the dynamic coupling coefficient configuration module, based on the lithological units and fault zones defined in the 3D geological model and combined with the phase information of the engineering implementation, configures nonlinear response relationships of physical property parameters for different geological regions, which can be dynamically updated during the inversion solution process. The key parameters are the functional dependence of permeability on stress and temperature, the response law of thermal conductivity to strain history, and the changes of fluid properties with temperature and pressure, thereby ensuring that the physical model can realistically depict the thermal-hydraulic-mechanical coupling process under strong heterogeneity. On this basis, the adaptive inversion solution module constructs a thermal-hydraulic-mechanical coupling forward numerical model embedding the above dynamic constitutive relationship, and uses a hybrid inversion algorithm that integrates Bayesian inference and deterministic optimization to solve it. This algorithm not only efficiently seeks optimal estimates for the water pressure field, temperature field, and stress field, but also quantifies the posterior uncertainty of each field variable through rigorous probability sampling. Finally, the decision support and visualization module performs in-depth interpretation of the inversion results, calculating quantitative risk indicators such as rock mass failure probability and dominant fluid migration path, directly generating engineering decision recommendations for shale gas fracturing design, geothermal reservoir stimulation, or tunnel stability assessment, and presenting them intuitively and interactively through 3D visualization. The implementation process of this system is detailed below through two typical engineering examples.

[0057] Example 1

[0058] This embodiment applies the invention to the monitoring and optimization of hydraulic fracturing processes in deep shale gas reservoirs. The goal is to accurately invert the three-dimensional dynamic changes in reservoir internal water pressure, temperature, and stress field induced by fluid injection during fracturing, thereby assessing the effectiveness of the fracturing network, predicting production capacity, and providing early warning of the risk of induced microseismic events.

[0059] During implementation, the multi-source heterogeneous data fusion module is initiated first. The raw data received by this module includes high-frequency time-series point measurement data collected by array-type fiber optic barometers, distributed temperature sensors, and strain gauges deployed in the target well and adjacent monitoring wells; time-series surface deformation field data acquired by synthetic aperture radar satellites covering the entire block; three-dimensional seismic velocity variation data obtained by tomographic processing of microseismic and active source data collected by ground and well seismic arrays; and multi-period resistivity cross-sectional data obtained by magnetotelluric methods. These data constitute a typical multi-source heterogeneous observation system.

[0060] To transform these data into effective constraints usable for joint inversion, the fusion module performs a series of standardization processes. For point-based time-series data, an adaptive filtering algorithm based on wavelet transform and empirical mode decomposition is first used for denoising. Then, a preliminary, spatiotemporally continuous field sequence is generated using Kriging interpolation combined with physical constraints. For surface-domain data such as InSAR deformation fields and seismic velocity imaging volumes, a pre-constructed three-dimensional geological structure model is used as the reference grid, and gridded resampling technology is used to ensure strict spatial alignment. The key to achieving data fusion lies in establishing a transformation operator based on physical mechanisms. This operator is constructed based on Darcy's flow theory, thermoelasticity, and the empirical relationships between resistivity, porosity, fluid saturation, and temperature. Its core function is to predict simulated observation values ​​that are completely consistent with the physical dimensions of various actual observation data by forward physical simulation of the core field variables to be inverted. For example, substituting the predicted stress field and pore pressure field into the pore elastic constitutive equation allows for the calculation of the predicted surface deformation field; substituting the predicted temperature field, pressure field, and fracture field into Archie's formula and its modified form allows for the calculation of the predicted resistivity distribution. Through this step, observational data from different sources and of different types are unified to a common inversion objective. Finally, based on the calibration accuracy, spatial resolution, and spatial correlation analysis between data points of various observation instruments, the module constructs a large observational error covariance matrix that may contain off-diagonal elements. This matrix is ​​used to assign reasonable statistical weights to observational data of different reliability and information content in the subsequent inversion objective function; thus, a physically consistent, spatiotemporally registered, and fully error-informed multidimensional observational dataset is constructed.

[0061] With the data ready, the dynamic coupling coefficient configuration module begins operation. Based on a 3D geological model including lithological unit divisions and major fault zone identification, and combined with the specific design parameters of the fracturing project, this module assigns an initial set of coupling coefficients to each computational grid cell. These coefficients parameterize the constitutive relationship of the thermo-hydraulic-mechanical coupling, specifically including the permeability sensitivity coefficient to stress changes. Permeability sensitivity coefficient to temperature changes Correction factor for thermal conductivity on strain history variation The coefficients for fluid density and viscosity as a function of temperature and pressure are also included. These coefficients are assigned values ​​that fully account for geological heterogeneity.

[0062] More importantly, these coupling relationships are dynamically updated during the inversion solution iteration; this module will update the stress field estimated in the current iteration step. and temperature field Dynamically update the penetration rate of each unit. The update follows the following functional relationship:

[0063]

[0064] in, This represents the permeability, effective stress, and temperature of the element under reference conditions. This function quantifies the combined effect of increased effective stress leading to fracture closure and thus reduced permeability, as well as the influence of temperature changes on permeability; simultaneously, it incorporates the element's plastic strain increment calculated from the strain field history. Update thermal conductivity :

[0065]

[0066] This reflects how the generation and evolution of fractures alter the heat transfer capacity of rocks; fluid properties are also updated in real time through nonlinear equations of state; this dynamic configuration mechanism enables the physical model upon which the inversion depends to adaptively characterize the strong nonlinear changes in reservoir properties during fracturing.

[0067] The core numerical inversion calculation is undertaken by the adaptive inversion solution module. This module first constructs a unified thermo-hydraulic-mechanical forward numerical model embedding the aforementioned dynamically coupled constitutive relations. This model is spatially discretized based on the finite element method, and its governing equations fully consider the conservation of fluid mass, thermal energy, and momentum of the soil skeleton, while coupling pore pressure effects, thermal stress effects, thermal convection, and nonlinear variations of physical parameters. This model constitutes the forward simulation operator. .

[0068] The inversion problem is solved within a Bayesian probabilistic framework; the model parameter vector to be inverted... It includes the initial state or changes of three field variables—water pressure, temperature, and stress—for each three-dimensional mesh element, as well as the coupling coefficients of each element. etc.; objective function Defined as the negative log-posterior probability density:

[0069]

[0070] in, This is the multidimensional observation dataset constructed as described above; It is its error covariance matrix; It is a vector of prior estimates of the mean values ​​of model parameters given based on geological models and engineering experience; It is the prior covariance matrix of the model parameters.

[0071] To efficiently and reliably solve this high-dimensional nonlinear inversion problem, this module employs a hybrid inversion algorithm. The solution process consists of two stages: First, a gradient descent optimization algorithm based on the adjoint state method is used to deterministically optimize the objective function, quickly locating the high-probability region where the predicted data best fits the observed data. In each iteration of gradient descent, the dynamic coupling coefficient configuration module participates in updating the constitutive relation. When the gradient optimization converges to a local optimum neighborhood, the algorithm enters the second stage: initiating Markov chain Monte Carlo sampling. This sampling starts from the current optimal solution and randomly explores the parameter space, generating a series of samples that follow a posterior probability distribution. These samples not only provide optimal estimates of the water pressure field, temperature field, and stress field, but more importantly, they provide the posterior probability distribution of each field variable at each spatial location, thus achieving rigorous quantification of the uncertainty of the inversion results.

[0072] This module also incorporates inversion process monitoring and strategy adjustment logic. This logic monitors the descent rate of the objective function and the spatial distribution characteristics of the data fitting residuals in real time. For example, when the system detects that the fitting residuals of InSAR deformation data remain persistently high and the optimization process stalls near a known natural fracture development zone, it will automatically trigger strategy adjustment: relaxing the key coupling coefficients in that area. and The prior variance constraint, or the local adaptive refinement of the computational grid for the region; this adaptive mechanism ensures that the inversion accuracy of key heterogeneous geological structures is significantly enhanced while maintaining overall computational efficiency.

[0073] Finally, the decision support and visualization module performs in-depth processing and presentation of the three-dimensional multiphysics field distribution and its uncertainty information obtained from the inversion. This module first performs quantitative risk analysis: based on the posterior probability distribution of the stress field, it calculates the shear failure probability of each unit rock mass using the Mohr-Coulomb criterion; combined with the distribution of the water pressure field, it identifies the fluid overpressure region and potential fluid dominant migration paths; and considering the changes in the temperature field, it assesses the range of thermal shock caused by cold water injection to the reservoir and its impact on the long-term conductivity of fractures. Subsequently, the module automatically generates a special decision recommendation report for shale gas fracturing engineering. The report is specific and actionable. For example, it identifies reservoir clusters with insufficient fracturing coverage based on inversion results, recommending priority targets for the next stage of fracturing operations; it optimizes pumping parameters for subsequent fracturing stages based on the stress shadow zone distribution predicted by inversion; it delineates areas with high risk of inducing microseismic events based on the calculated rock mass failure probability map, and proposes real-time pumping warning thresholds accordingly; all inversion and analysis results are rendered with high quality through the integrated 3D visualization engine of this module, supporting rotation, scaling, profile cutting, and dynamic evolution demonstrations, providing engineers with intuitive and comprehensive decision-making support.

[0074] Through the systematic implementation of this embodiment, high-resolution and high-precision dynamic characterization of reservoir multi-physics fields during shale gas fracturing has been successfully achieved, along with rigorous reliability assessment. This provides key technical support for the refinement of fracturing design, real-time control of construction risks, and optimization of production enhancement effects.

[0075] Example 2

[0076] This embodiment applies the invention to the long-term operation monitoring and optimization of a hot dry rock enhanced geothermal system. The goal is to understand the spatiotemporal evolution of the temperature field, pressure field, and induced stress field within the reservoir under multi-year periodic injection-production cycles, in order to maintain system heat recovery efficiency, assess induced seismic risks, and optimize well network layout and operation strategies.

[0077] In this scenario, the multi-source heterogeneous data fusion module needs to process observational data spanning several years. Point data mainly includes time-series data of wellbore temperature and pressure profiles collected by fiber optic thermo-and-pressure measurement systems permanently installed in injection and production wells, as well as microseismic event catalogs and waveform data continuously recorded by downhole seismograph arrays. Area data includes InSAR surface deformation monitoring data from multiple periods and multiple satellite platforms, as well as resistivity time-lapse data obtained from periodically conducted magnetotelluric exploration. The data fusion process needs to pay special attention to the separation of long-term trends and short-term periodic fluctuations, and utilize the microseismic source mechanism to unconstrain the direction of the local stress field. When constructing the physical transformation operator, it is necessary to fully consider the supercritical properties of fluids under high temperature and high pressure conditions and their special physical models affecting resistivity and seismic wave velocity.

[0078] In this embodiment, the dynamic coupling coefficient configuration module needs to characterize the special rock mass and fluid behavior under high temperature and cyclic loading. The initial coupling coefficient will be assigned mainly based on previous indoor experimental results such as high temperature and high pressure triaxial tests of core samples. During the dynamic update process of inversion iteration, the permeability sensitivity coefficient to temperature is... The potential thermal deterioration effect caused by high temperatures must be considered, as its functional relationship may be non-monotonic. The correction factor for thermal conductivity with respect to strain is needed. This requires introducing cumulative damage variables related to temperature history and stress cycle number to simulate the long-term evolution of cracks due to fatigue effects under repeated opening and closing cyclic loading.

[0079] The forward numerical model constructed by the adaptive inversion solution module needs to be able to simulate the complete cycle from cold water injection and reservoir heating to hot water production, as well as the gradual movement of the thermal front over many years of operation. Therefore, the model parameters to be inverted may be extended to background fracture permeability fields, long-term heat exchange coefficients, etc. Due to the long data time span and large inversion parameter space, the Markov chain Monte Carlo sampling part in the hybrid inversion algorithm needs to use parallel multi-chain and other techniques to accelerate the calculation. The inversion process monitoring logic will pay special attention to the long-term deviation between the production well temperature decrease rate and the model prediction value. If the deviation persists, the system may automatically trigger the correction of the large-scale fracture channel model of the reservoir or the reassessment of the inter-well connectivity structure.

[0080] In this embodiment, the decision support and visualization module focuses on the long-term operational optimization and risk management of the system. Based on the temperature field evolution history and trends obtained through inversion, the module can predict the thermal power decay curve of production wells in the coming years. Combining the inverted pressure and stress field results, the system can analyze the effective stress changes caused by injection-production imbalance and calculate and plot the probability risk map of induced earthquakes. Ultimately, the optimization decision recommendations generated by the module will be more strategic, such as: suggesting adjusting the flow ratio of injection wells and production wells to maintain reservoir pressure more evenly; planning periodic reservoir intensification operation time windows to restore fracture conductivity; or proposing new drilling target areas based on the assessment of the reservoir's remaining thermal potential. All analyses and recommendations are based on inversion results with accompanying confidence intervals or probability assessments, providing quantitative scientific decision support for the long-term safe, efficient, and robust operation of enhanced geothermal projects.

[0081] 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 in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, 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.

[0082] 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 joint inversion and quantification system for water pressure, temperature, and stress in a geological body, characterized in that, include: The multi-source heterogeneous data fusion module is used to acquire and fuse observation data from point sensors and area exploration equipment to generate a spatiotemporally correlated multidimensional observation dataset. The dynamic coupling coefficient configuration module is used to dynamically establish and update the dependencies between permeability, thermal conductivity and fluid properties and stress, temperature and deformation history based on the stratigraphic structure attributes and engineering conditions represented by the three-dimensional geological model. The adaptive inversion solution module is used to construct a thermal-hydraulic-mechanical coupled forward model embedding the aforementioned dependencies, and based on the aforementioned multidimensional observation dataset, it uses a hybrid inversion algorithm that integrates Bayesian inference and deterministic optimization to simultaneously invert the distribution and uncertainties of the water pressure field, temperature field, and stress field within the geological body. The decision support and visualization module is used to perform quantitative analysis on the field distribution and uncertainty obtained from the inversion, generate engineering assessments or decision recommendations, and provide visualization output.

2. The system according to claim 1, characterized in that, The multi-source heterogeneous data fusion module is specifically used for: Spatiotemporal registration and gridding alignment are performed on observation data from different sources and formats; Based on physical mechanisms, a transformation relationship is established to map inverted field variables to various types of observation data; Based on the error characteristics of the observation data, an observation error covariance matrix is ​​constructed for weighting in the hybrid inversion algorithm.

3. The system according to claim 1, characterized in that, The dynamic coupling coefficient configuration module is specifically used for: Based on the lithological zoning and structural features identified in the three-dimensional geological model, an initial set of physical property response parameters is assigned to different regions. During the iterative process of the adaptive inversion solution module, the permeability and thermal conductivity of each region are dynamically updated based on the stress field, temperature field and strain history obtained from the current inversion.

4. The system according to claim 1, characterized in that, In the adaptive inversion solution module: The thermo-hydraulic-mechanical coupled forward model is constructed based on the mass, energy, and momentum conservation equations. The hybrid inversion algorithm aims to maximize the posterior probability. Its objective function includes observation data fitting terms weighted by the observation error covariance matrix and prior constraint terms for model parameters.

5. The system according to claim 4, characterized in that, The adaptive inversion solution module is also used for: During the inversion process, the data fitting residuals and convergence status are monitored. When a poor fit occurs in a specific geological structure region, the constraint strength of the model parameters in that region or the mesh resolution of the local model is automatically adjusted.

6. A method for joint inversion and quantification of water pressure, temperature, and stress in a geological body, characterized in that, include: Acquire and fuse multi-source heterogeneous observation data to form a spatiotemporal dataset for joint inversion constraints; Based on a three-dimensional geological model and engineering information, a coupling relationship model between water pressure, temperature and stress field is established and dynamically updated. A forward numerical model embedded in the coupling relationship model is constructed, and a hybrid inversion algorithm is used to simultaneously solve for the optimal estimates of the water pressure field, temperature field and stress field and their posterior uncertainties, with the spatiotemporal dataset as constraints. Engineering risk analysis and decision support are based on the inversion results.

7. The method according to claim 6, characterized in that, The acquisition and fusion of multi-source heterogeneous observation data includes: The point-measured time-series data are processed to construct the initial field sequence; The area exploration data is resampled to a unified grid that matches the three-dimensional geological model; Correlating different observation types with core field variables through physical constitutive or empirical relations; Quantify and utilize the errors and spatial correlations of observation data.

8. The method according to claim 6, characterized in that, The establishment and dynamic updating of the coupling relationship model includes: Based on the differences in lithological zoning and structural features in the three-dimensional geological model, an initial set of coupling coefficients is assigned; In the inversion iteration, the permeability and thermal conductivity are corrected in real time through a preset function based on the updated state of the field variables.

9. The method according to claim 6, characterized in that, The hybrid inversion algorithm includes: A gradient optimization algorithm is used for fast deterministic optimization. The Markov chain Monte Carlo sampling method is used to probabilistically explore the neighborhood of the optimal solution in order to quantify the uncertainty of parameters.

10. The method according to claim 6, characterized in that, The engineering risk analysis and decision support based on the inversion results includes at least one of the following applications: Calculate the probability of rock mass failure and assess stability risk; Identify the dominant fluid transport path or the extent of thermal shock impact; Provide suggestions for segment cluster selection and pumping procedure optimization for hydraulic fracturing projects; Provide suggestions for optimizing injection and production well networks and circulation strategies for geothermal engineering.

Images