Experiments involving the dissociation of artificial methane-hydrate-bearing sediments were performed using X-ray computed tomography (X-CT, 40 s scanning speed at 2 min intervals) to directly observe dissociation behavior in the sediments and the gas and water flows generated by dissociation. Dissociation by depressurization was performed using a backpressure regulator, and showed that the temperature reduction induced by depressurization depends on the phase equilibrium state of methane hydrate, and that preferential dissociation occurs along the periphery of the core. This behavior is caused by heat flux from the outside of the core, and this controls the dissociation rate. A heat exchanger was installed at one end of the core to simulate thermal stimulation, and propagation of a clear and unidirectional dissociation front was observed. Depending on the heating temperature, the dissociation rate was less than that observed for depressurization. Hot water was also injected at a constant rate from the bottom of the core, and CT images showed the movement of distinct accumulations of dissociated gas being pushed by the hot water. The gas production rate increased immediately after the gas accumulation reached the opposite end of the core where the gas and water flow out.

Hot water injection is a production technique proposed to gas recovery from methane hydrate reservoirs. However, from a practical point of view, the injected water experiences a drop in temperature and re-formation of hydrates may occur in the reservoir. In this work, we proposed a model expressing permeability hysteresis in the processes between hydrate growth and dissociation, and studied hydrate dissociation behavior during hot water injection. The model of permeability hysteresis was incorporated into the simulator MH21-HYDRES (MH21 Hydrate Reservoir Simulator), where the decrease in permeability with hydrate saturation during hydrate growth process was assumed to be much larger than the decrease during hydrate dissociation process. Laboratory hydrate dissociation experiments were carried out for comparison. In each experiment, we injected hot water at a constant rate into a sand-packed core bearing hydrates, and the histories of injection pressure, core temperature, and gas/water production rates were measured. Numerical simulations for the core experiments showed the re-formation of hydrates led to the increase in injection pressure during hot water injection. The simulated tendencies of pressure increase varied markedly by considering permeability hysteresis. Since the experimental pressure increases could not be reproduced without the permeability hysteresis model, the influence of permeability hysteresis should be considered to apply hot water injection to hydrate reservoirs.