Modeling and simulation of methane production from hydrate reserves

This dissertation describes the computational fluid dynamics (CFD) simulations of methane production from hydrate reserves.;The dissociation of methane hydrate in porous media has been a critical research area due to methane hydrate's promising potential as a source of natural gas. After methane hydrate dissociates in the porous media, the gas and water from dissociation flow in the porous media due to the pressure gradient. In the experimental part of this thesis, the gas and water permeabilities under different overburden pressures (OBP) were measured for sands of different size. These permeabilities were used as input in the simulation of hydrate dissociation process in porous media by depressurization method. It was found that OBP has little effect on the gas permeability while the water permeability is very sensitive to the OBP.;Numerical algorithm based on coordinate transformation and method of lines was used to solve the moving front model for hydrate dissociation. The numerical algorithm was validated by comparison with available analytical solution in the literature. The water phase movement after hydrate dissociation was included in the two phase flow moving front model. It was found that the assumption of whether the water phase is stationary affects the gas production at the well significantly. The assumption of stationary water phase overpredicts the location of the moving front and underpredicts the gas production at the well.;In a single vertical production well in cylindrical coordinate system, it was found that thermal stimulation in a single vertical well is not effective due to the small surface area for heating. Horizontal well can increase the gas production by thermal stimulation due to the increase of heating surface area. It was found that overheating the reservoir can increase the rate of hydrate dissociation but decrease the ratio of energy output to energy input because most of the heat is used by heating the reservoir. We proposed that the optimal well temperature is around the initial reservoir temperature.;Comparison between two phase moving front model with the reaction model and available experimental data in the literature show that the experimental work is dissociation reaction limiting. Thus, reaction model is more appropriate for experimental work due to the available heat from the surrounding air bath.;Comparison between 1D and 2D axisymmetric model with experimental data show that 2D model takes more time to reach steady state due to the no-slip assumption at the wall. Parametric studies show that the reaction model is sensitive to parameters such as the intrinsic reaction rate and the activation energy.