Research On Simulation And Inversion Of Magnetotelluric For Hot Dry Rock Exploration And Its Application

At present,the energy crisis and environmental pollution are the key problems that restrict the sustainable development of society.The research of renewable new energy with large reserves and environment-friendly has been widely valued by governments and research institutions of various countries.As a new type of geothermal resources,the Hot Dry Rcok(HDR)type geothermal resources is characterized by high temperature rock mass without water or with little water.It is a new and stable clean energy with broad prospects for development.Its development system is called Enhanced Geothermal Systems(EGSs).Geophysical method is one of the main means of HDR exploration.However,the burial depth of HDR is large(3km-10km),and its physical properties are often only caused by temperature,which is a new challenge to the selection of geophysical methods and data processing and interpretation in HDR exploration.Magnetotelluric(MT)is an electromagnetic sounding method with natural electromagnetic field as its source.This method has great exploration depth and is sensitive to low resistivity body.It can not only detect tectonic background related to thermal anomaly,but also directly detect low resistivity anomaly caused by high temperature rock.It shows great development potential and wide application in HDR exploration.Aiming at the research of magnetotelluric method in HDR exploration,this paper studies and implements 2D MT forward based on nodal finite element method and Gauss-Newton inversion.The unstructured triangular mesh generation method based on Delaunay algorithm is used to fit the geometric shape of complex geological bodies.In finite element analysis,by establishing the mapping relationship between regular reference triangle and irregular local triangle,the stiffness matrix and mass matrix of any element can be obtained by mapping the stiffness matrix and mass matrix of the reference model,which improves the programming efficiency.The parallel method about frequency point and matrix calculation is used to improve the efficiency of forward and inverse algorithm.In the aspect of 3D MT forward and inversion,because the nodal finite element method can not satisfy the divergence condition when it is extended to three-dimensional model,the pseudo-solution appears.In this paper,the 3D MT forward algorithm based on edge finite element method is studied and implemented.By decomposing the total field into primary field(background field)and secondary field(scattering field),the calculation accuracy is improved.The first field is calculated analytically by isotropic horizontal layered medium,and the second field is solved by edge finite element method based on unstructured tetrahedral mesh.On the basis of forward algorithm,3D MT inversion algorithm is realized by using Gauss-Newton method.The calculation of sensitivity matrix is combined with forward calculation by"quasi-forward"method,which reduces the amount of calculation.The validity of the forward and inverse algorithm is proved by the verification calculation of the 3D standard model.Temperature is the most important indicator of HDR.In this paper,3D forward simulation of steady and unsteady temperature fields based on nodal finite element method is studied and realized.According to the main forms of geothermal resources in HDR,the conceptual models of steady-state HDR and unstable HDR are designed,and their temperature distribution is calculated.In the steady-state HDR model,because of the influence of temperature on resistivity,the resistivity interface between sedimentary layer and basement in the basin weakens,and the HDR mass is shown as a low resistivity body extending deep.The unsteady HDR model simulates the heat dissipation process of the magma sac with a radius of 1.5 km.The results show that the temperature of the magma sac tends to be stable after 0.5 Ma,which indicates that the isolated magma sac before Quaternary could not form HDR without special heat source.Gonghe Basin in Qinghai Province is rich in geothermal resources and is an important demonstration research area for HDR exploration and development in China.Based on the data processing and interpretation of MT in Gonghe Basin,the genetic mechanism of HDR in Gonghe Basin is studied.The 2D electrical structure in Gonghe basin can be obtained by processing and inversion of measured MT data;the inversion of gravity data constrained by electrical structure can provide underground density structure;combined with regional aeromagnetic data,it is recognized that the Langshan-Wuwei-Gonghe fault zone is the heat-controlling structure of the Gonghe thermal anomaly,and this fault zone is a hidden deep and large fault.Based on geological information,the conceptual model of geothermal system of Gonghe HDR is given:HDR thermal reservoir is high-temperature granite;geothermal caprock is Paleogene,Neogene and Quaternary sedimentary strata;geothermal source is thermal energy in deep crust;Langshan-Wuwei-Gonghe fault zone is the channel of upward thermal energy transmission.Heat flow value is an important parameter to analyze the genesis mechanism of geothermal field.In this paper,Bayesian inversion of geothermal field constrained by prior model is studied and implemented.Geothermal field of Gonghe Basin was obtained by inversion of temperature logging data of DR3 borehole and underground heat flow was calculated from geothermal field.The results show that the maximum heat flux of the Gonghe Basin basement exceeds 110 mW/m~2.Based on the criterion that the temperature is greater than 150?and the depth is less than 5 km,the HDR target area can be delineated near Gonghe with 10 km width along the MT line.The heat flux in the target area exceeds 100 mW/m~2,of which 75%comes from the heat conduction from deep heat source and 25%comes from radioactive heating of granite itself.The type of HDR thermal reservoir in Gonghe Basin is heat conduction type,and the thermal recharge mainly comes from the thermal conduction by deep and large faults,while the thermal generation of granite is a secondary factor.HDR thermal reservoir is located near the upward passage of deep heat flow,with high heat flow value,stable thermal recharge,huge development potential and broad prospects for sustainable utilization.In summary,this paper systematically studies MT forward and inversion methods based on unstructured finite element method,and temperature field forward and inversion methods.Through the improvement of the algorithm,the operation efficiency is improved while maintaining the accuracy.After studying the genetic mechanism of HDR based on MT exploration in Gonghe Basin,we have gained a clear understanding of the geothermal elements such as the geothermal types and the composition of heat sources of HDR in Gonghe Basin.In the process of research,MT method shows its good practicability in HDR exploration.This study has practical significance for improving and perfecting the level of exploration and interpretation of HDR.