| Massive tunnels are being constructed in the field of transportation, mine and water conservancy and hydropower in China. These tunnels generally have the remarkable characteristics of large buried depth, long tunnel line and complicated geological condition, which lead to a lot of geological hazards such as water inrush in tunnel construction period. Among the tunnel serious accidents in the world, the water inrush hazard is in the front rank from the aspect of death toll and number of occurrence. The water inrush hazard has become one of the bottleneck problems that constrain the construction of underground engineering such as tunnels. Therefore, advanced detection of the water-bearing geological structures and real-time monitoring and prediction of the water inrush hazard in tunnel construction period has become the important scientific and engineering problem to be studied and solved urgently.For the above problems, the solving methods with the electrical resistivity method and Induced Polarization (IP) method as the guide is proposed. Taking 3D resistivity forward modeling and inversion theory and water volume estimation method as the starting point, the technical system of 3D tomography and water volume estimation for tunnel water-bearing geological structures prediction is established. And the system and method of water inrush hazard real-time monitoring and prediction based on Electrical Resistivity Tomography (ERT) is put forward. And the research results have got successful application in practical engineering and physical model test. The main research work focuses on the following.1) For 3D resistivity forward modeling problem, Finite Element Method (FEM) is used to solve 3D point source electric field. And the optimization scheme of computation efficiency for solving the large linear systems based on Preconditioned Conjugate Gradient algorithm and Cholesky decomposition algorithm is proposed and rapid and accurate forward modeling of 3D resistivity detection is realized. Based on the 3D resistivity forward modeling method, equivalent resistivity method is used for 3D polarizability forward modeling.2) For 3D resistivity inversion and tomography problem, the linear inversion method, nonlinear inversion method and combined inversion method are studied respectively. Least square method as a linear-inversion method based on smooth constraint and its optimization scheme of computation efficiency are put forward. Least square method for inversion is efficient in computation but has a strong dependence on the initial model. The improved Genetic Algorithm (GA) as nonlinear inversion method based on smooth constraint is proposed. And the mutation direction control technology is studied for solving the problem of low searching efficiency that hinders the usage of GA method in 3D resistivity inversion. By this method, inversion result dependence on the initial model is basically eliminated, but it is quite time-consuming. The integrated inversion method combined with linear and nonlinear inversion methods is presented, by which inversion result dependence on the initial model is basically eliminated and the inversion speed is efficiently improved by the premise of guarantee of inversion effect. The 3D resistivity inversion program 3D E-Imager is compiled, building a general software platform for 3D resistivity inversion and forward modeling.3) 3D tomography method for tunnel water-bearing geological structure advanced detection based on direct current (DC) resistivity method is proposed. The types and characteristics of water-bearing structures are analyzed and generalized. The basic problems of tunnel advanced detection are solved (e.g., solution of point source field in 3D whole space, measurement mode, recognition and removal of the measurement disturbance).The response characters of observation data are summarized on the foundation of further study on the forward modeling of advanced detection of water-bearing structures. And the inversion object region of advanced detection is determined.3D tomography and location method for water-bearing geological structure advanced detection based on 3D resistivity inversion method is put forward. The feasibility and effectiveness of 3D tomography for water-bearing structures (e.g., water-bearing faults, water-bearing caves) has been proved by massive inversion examples. And the tomography characters for different kinds of water-bearing structures are analyzed and summarized.4) For the problem of water volume estimation in water-bearing structure, the water volume estimation method based on half-decay time difference of two-current induced polarization is proposed. The time domain two-current induced polarization equipment is developed. It is found that the relation between difference of half-decay time and water volume is linear positive correlation by physical model test. And the method of water volume estimation is established on the basis of above discovery, providing a feasible way for solving the problem of water volume estimation in practical engineering.5) Taking 3D resistivity inversion technique and water volume estimation method with half-decay time difference as the solving way, the technical system of 3D tomography and water volume estimation for tunnel water-bearing geological structures prediction is established combined with the geological analysis method and other prediction technique. And this technical system is introduced into geological forecast work in tunnel engineering (e.g., Jigongling Tunnel in Fanba Highway,4# diversion tunnel in Jinping II Hydropower Station). The prediction result is basically identical with the practical excavation result, by which the feasibility and effectiveness of 3D tomography and water volume estimation technique for water-bearing geological structures is verified in the aspect of engineering application.6) For the problem of water inrush geological hazard real-time monitoring and prediction in tunnel construction, the system and method of water inrush hazard real-time monitoring and prediction based on ERT method is put forward. The geo-electrical models of tunnel water inrush are established according to related data, and numerical forward modeling is performed. Taking rapid inversion and tomography method for 3D resistivity as key technology, the real-time monitoring system with ERT is designed. This system has advantages of automatic monitoring and comprehensive judgment with multiple parameters. And this system is used in tunnel water gushing model test and mining area water bursting model test for real-time monitoring work. The intuitive dynamic images of important precursory process is obtained, e.g., seepage development, rock fracture, water inrush channel formation. And the precursory information for water inrush is captured, providing an important reference for water inrush hazard prediction.7) In the end, the thought and framework of two-stage prediction and early-warning system for tunnel water inrush is proposed based on the DC resistivity and IP method, pointing out the direction for future study work of this paper and laying a solid foundation for prediction and early-warning of water inrush hazards in practical engineering.
|
PDF Full Text Request