Visualization and interpretation of methane hydrate growth and dissociation in synthetic porous media
MetadataShow full item record
Sedimentary natural gas hydrates might play an important role in the future energy mix and the impact of methane release from natural gas hydrates can be significant also with respect to climate change. Vast amounts of methane gas are trapped in the subsurface permafrost and in submarine environments below sea floors. To understand the mechanisms of formation and dissociation on pore scale is fundamental to explain phenomena on core scale and making hydrate simulators on field scale. The three scales combined, will provide valuable data and understanding for field planning and testing. The experimental work presented in this thesis is visual observations conducted in a 2D silicon wafer micro-model with a one-to-one representation of the Berea sandstone pore network using a vertical pore depth of 25µm. In this thesis the overall objective was to characterize hydrate formation and dissociation on pore scale by direct visualization. The second objective was to identify the limiting factors for hydrate formation and growth in the micromodel and compare these to bulk properties. Hydrate formation attempts has been done both with vacuumed and non-vacuumed water. This to mimic the oxygen and nitrogen levels in reservoir and surface conditions respectively and tracing the three-phase-equilibrium line in both cases. No significant difference was found. Thin film interference was used to determine the thickness of the water and hydrate film. This turned out to be limited by wavelength of the visual light spectrum. Water films were found to be thinner than 140nm and the hydrate films to be thicker than 1500nm. The experimental setup was modified to enable live view monitoring of the micro-model and fluid flow through all four ports. A new cooling-chamber was constructed allowing a more efficient temperature control and quicker heating/cooling of the still water surrounding the micro-model. During fixed pressure and variable temperature experiments, it was found that even at conditions well within the hydrate equilibrium line, growth did not necessarily commence spontaneously. However, a disturbance in the fluids by pressure regulation was a key factor. Therefore, the degree of which turbulence and flow promotes hydrate formation in a porous medium by high flowrates has been studied. This was done at 145, 130 80 and 55 bars of differential pressure and temperature range from 0.6-11°C, with vacuumed and non-vacuumed distilled water. 21 formation runs with high flow rate through the model to initiate the visual growth were done and in 21 of 21 attempts visual hydrate formation occurred within 4 minutes. This method of initiating hydrate growth turned out to be very effective making the stochastic nature of nucleation insignificant. In every instance the hydrate was observed to grow as a black film inwards in the gas phase that subsequently turned more transparent upon further growth. Due to the difference in refractive index of hydrate and gas the hydrate film appears as black while re-soaking in water makes it more transparent. Growth into the water phase has not been observed. After flow, hydrate grew evenly from thin water films retaining on flat surfaces inside the micro- model, while growth from a static situation led to a hydrate film advancing through the gas phase from the interface to the center. Dissociation experiments with temperature change and fixed pressure and vice versa were conducted. The methane hydrate was observed to appear as a layer within the micro-model. The dissociation was observed mainly to occur along the edge of the hydrate film upon pressure reduction. Contact with a free gas phase seemed to be a necessity for quick heat and mass transport leading to melting. Immobile water surrounding the hydrate films acted as a hindrance for dissociation. When heating the system at constant pressure, methane released from the hydrate were more evenly spread and occurred at the hydrate film edge as well as within the hydrate layer. This may be explained by heat transport through the matrix and silicon wafer.