(74b) Pore-Scale Study on the Effects of Hydrate Distribution Morphology on Dissociation Process Utilizing a Coupled Multiphase Hydrodynamic and Thermal Lattice-Boltzmann Model

Natural gas hydrate, plentifully distributed in ocean floor sediments and permafrost regions, is considered a promising unconventional energy resource. Because of the high energy storage capacity, the amount of CH₄ reserved in hydrate is larger than other gas resources. Thus, recovering the CH₄ from hydrate has attracted more and more research efforts from academia to industry. Although many hydrate deposits have been explored, high risks for geological hazards, low gas recovery, and expensive investment prevent the commercial gas production. To realize long-term production, much work remains in elucidating the hydrate dissociation mechanism in porous mediums. Hydrate dissociation in porous medium is complex, including coupled processes of multiphase flow, conjugate heat transfer, mass transfer, etc. Furthermore, hydrate distribution morphology can significantly affect the dissociation and flow. To elucidate the effects of hydrate distribution morphologies on dissociation under production condition, a pore-scale study based a coupled multiphase hydrodynamic and thermal Lattice-Boltzmann model (LBM) is carried out. To the best of our knowledge, the LB method has not been applied for hydrate dissociation study with considering the whole physicochemical processes, previously.

To describe the physicochemical process of hydrate dissociation, the developed LB model is consist of three sub-models. For the hydrodynamic process, the color-gradient LB model is developed. To solve the numerical difficulties in simulating multiphase flow with high density ratio, an enhanced equilibrium distribution function, which added the density gradient into the momentum flux tensor, is introduced into the multi-relaxation-time (MRT) collision operator. To eliminate the spurious current at the phase interfaces, 8^th order accuracy schemes for the gradient computations are applied into the single phase and multi-phase collision operators. For the heat transfer process, the conjugate heat LB model which solves the conservation form of the energy equation is developed. By introducing a conjugate heat source at the phase (gas, water, sand) interfaces, it can conserve the conductive and advective flux simultaneously, resulting an accurate description for the temperature field evolutions in the heterogeneous heat capacitance system. For the mass transfer process, the hydrate dissociation rate is calculated from the Kim‐Bishnoi model combined with the Kamath model, and the hydrate nodes updating process due to the dissociation is described by the VOP method. To validate our model, several benchmark cases with analytical or numerical solutions are applied. The layered Couette/Poiseuille flows with high density ratio fluids are simulated to verify the hydrodynamic process. The heat transfer between two flowing fluids with different thermophysical properties is simulated to verify the conjugate heat transfer process. The natural convection flow in a cavity is simulated to validate the coupling process of flow and heat transfer. The simulation results showed excellent agreements with the analytical or numerical solutions, proving the reliability of our model.

In this work, three hydrate distribution morphologies and two sandy sediments are applied to construct six hydrate-bearing sediments. Then the hydrate dissociation processes under the coupled thermal stimulation and depressurization approaches are simulated with the developed LB model. The selected hydrate distribution morphologies are based on the grain-scale experiments, including the pore-filling, grain-coating, and dispersed cases. The sandy sediments including a synthetic homogeneous sediment and a real heterogeneous sediment consider the effects of heterogeneity into the dissociation. Because hydrate dissociation rate is influenced by the temperature and pressure fields, different subsurface structures caused by hydrate morphologies can obviously affect hydrodynamic and heat transfer processes, resulting in distinct dissociation evolutions. From the simulation results, we found that with the identical initial hydrate saturations and production conditions, the grain-coating case performs an obvious dissociation lag compared with the pore-filling and dispersed cases. In addition, the dissociation lag can become more obvious by increasing the heterogeneity of porous mediums.

The main differences of heat transfer evolutions are caused by the distribution type of hydrate and grains. In the hydrate-bearing sediments, the specific heat capacities of grains (sand, clay, silt) are larger than the hydrate. Therefore, the grain-coating morphology inhibited the heat conduction by significantly reducing the surface area of grains. In addition, the endothermal heat could create cold zones around the grains, further suppressing the heat conduction from the ambient heat source. On the contrary, hydrate is wrapped around by the grains for the pore-filling and dispersed morphologies, improving the heat conduction rate from the ambient heat source to the hydrate surface. Thus, the average temperatures of the pore-filling and dispersed cases are always larger than that in the grain-coating case during the dissociation.

The subsurface structure of hydrate-bearing sediments can affect the pressure fields evolutions by influencing the hydrodynamic processes. To quantitatively describe the subsurface structure evolutions during the dissociation, we applied the automated algorithm, which is based on the city-block distance function and watershed segmentation method, to extract the pore and pore throat radii during the dissociation. We found that the hydrate layers in the grain-coating morphology can significantly inhibit the fluid flowing capability by reducing the average pore throat size, and the confined fluids can cause larger average pressure than that in the pore-filling and dispersed cases. We also applied the water percolation numerical experiment to obtain the connectivity evolutions for three cases. The results showed that during the dissociation, the connectivity in the pore-filling and dispersed cases increases gradually and smoothly, whereas it increases with some jumps due to the sudden improvement of pore throat sizes in the grain-coating case. Based on the connectivity evolution characteristics of three cases, we found that the gas deliverability from the inlet to the outlet could get more inhibitions in the grain-coating case.

This pore-scale study not only provides intuitive descriptions for dissociation evolutions of different hydrate distribution morphology cases in the porous mediums, but also can give some suggestions for the large-scale productions. From this pore-scale study, we found that hydrate distribution morphologies could significantly affect the subsurface structures, which could further influence the heat transfer, fluids transport, and gas deliverability, resulting in different hydrate dissociation processes under the same production conditions. Some hydrate distribution morphologies, such as the grain-coating case, could enhance the self-preservation phenomenon in the dissociation process and inhibit the gas deliverability. To improve the production performances for such distribution morphology hydrate reservoirs, more pressure reductions and ambient heat sources are required.