Study On Fluid Immigration Mechanisms And Numerical Simulation Of Coalbed Methane Reservoirs

Coalbed methane(CBM)as a type of unconventional resources is becoming an important addition to conventional natural gas supply.To date,commercial development of CBM resources has been undertaken in several basins in China.However,mechanisms of fluid storage and transport at in situ conditions are still far from clarification due to the complexity in pore structure and flow processes and more efforts are needed to bring them in sight.The accuracy and robustness of numerical simulation,which is a technique for CBM reservoir evaluations and performance predictions depends heavily on the accuracy in description of transport mechanism.As such,it is of practical significance to apply multiple techniques to investigate the fluid transport mechanisms in CBM reservoirs and subsequently to improve mathematical models for numerical simulation.This thesis starts with the study of methane adsorption capacity in coals.A large number of adsorption isotherm data were acquired from previous literature and an empirical model for predicting methane sorption capacity in coals was developed using alternating conditional expectation(ACE).The controlling effects of varying varying geological factors on methane adsorption capacity were analyzed quantitatively.Laboratory measurement of adsorption/desorption isotherms was also conducted and underlying mechanisms were discussed that results in adsorption/desorption hysteresis.A modified Langmuir type isotherm model was proposed using perturbation theory and validated using numerous experimental data.Second,pore-scale network modeling of gas diffusion at the presence of moisture in coal matrix was developed based on the percolation theory.Simulation results indicate that pore structure,moisture content and gas pressure have ineligible effects on gas diffusion coefficient.A empirical model for estimating diffusion coefficient was developed based on the simulation results.As for the cleat system,a mathematical modeling was proposed to describe the cleat aperture change with gas pressure based on linear elastic deformation theory.This model was then incorporated with Darcy-flow pore-scale network modeling technique to simulate permeability dynamics for heterogeneous cleat network.Sensitivity analysis reveals that cleat heterogeneity exerts profound effect on permeability evolution trend.Therefore,a semi-empirical permeability evolution model was proposed for heterogeneous cleat network,which was further validated with laboratory test data.Third,with the above study addressed,an improved dual-porosity/single-permeability model was developed that considers the pore structure and fluid storage/transport characteristics for CBM reservoirs at different ranks.The model was subsequently solved with fully implicit finite difference method and a corresponding simulator was developed.Forth.a step-wise well placement procedure that incorporates stochastic optimization algorithm with numerical reservoir simulation was proposed.The performance of different optimization algorithms was compared.The proposed step-wise optimization procedure was demonstrated to significantly outperform direct optimization procedure in terms of convergence speed,robustness and optimal target function.Finally,numerical reservoir simulation was applied in an actual CBM reservoir for history matching,well performance prediction and optimization of well placement.The geological factors that controls well drainage characteristics were analyzed based on history matching results.The development performance using optimized well placement was demonstrated to significantly outperform the actual well placement.