A Visual Micromodel was used to study the potential for the underground sequestration of carbon dioxide in methane hydrate reservoirs. Phase behaviour of reservoir fluids in porous media have been extensively studied earlier using micromodels (Sohrabi et al., 2000), while its potential application in gas hydrate studies has been demonstrated by Tohidi et al.
(2001). Pore scale studies were aimed towards providing insight into gas hydrate growth from dissolved gas (CO2-water) and gas hydrate distribution / cementing characteristics of grains in THF-, CO2- and CH4-water systems. Later Anderson et al. (2001) produced visual information on phase distribution in porous media, and hydrate-grain cementation for CH4-water and CH4-CO2-water systems. Here we present the results of experiments conducted using the Medium Pressure Micromodel for studying the effect of CO2 injection on already-existing methane hydrates in porous media.
Two micromodel rigs are currently in operation: a medium pressure setup (1204 psia) and a high-pressure set-up (5947 psia). The central glass micromodels consist of an etched glass base-plate topped with a sealed glass cover plate (Figure 6). Either a geometrically designed network of pores, tubes or reproductions of actual thin sections of real sediments can be used to construct the micromodels by etching with hydrofluoric acid. The cover plate has an inlet and an outlet, which allows fluids to be pumped through the enclosed pore network using small-volume piston vessels or a precision Quizix pump (Figure 7).
In both set-ups the glass micromodels are mounted in a vessel that exerts an overburden pressure and are surrounded by coolant jackets controlled by temperature-controlled baths. Temperature is measured by a probe mounted in the overburden cell, and transducers measure pressure on the model inlet and outlet lines. Temperature can be kept stable to within + 0.05 oC. Temperatures and pressures are monitored by means of a PRT and Quartzdyne pressure transducers (accuracy of + 1.16 psia for 0-20000 psia), respectively.
Magnifying cameras are mounted above the models, with illumination being provided by cold light sources. Because the micromodel pore structure is only one pore thickness deep it is possible to clearly observe phase changes and fluid flow behaviour. Digital video footage is recorded during the experiment, and the pictures presented here are ones recorded either from video footage or camera clippings.
Existing equipment has been employed in the study of a wide variety of hydrate systems pertaining to various scenarios, from hydrates in sub-sea sediments to flow assurance. The technique provides novel visual information on the mechanisms of clathrate growth, micro-scale dissociation and phase distribution, with respect to pressure, temperature, wettability and fluid composition (presence of inhibitors, liquid hydrocarbons, free and dissolved gas).
Piston vessel (oulet)
Coolant Micromodel wm
■Temperature [ probe Insulated jacket
Piston vessel (inlet)
Piston vessel (inlet)
Figure 7. Glass micromodel: experimental set-up.
In all experiments the liquid (water) phase was dyed with methyl blue. Hydrates and gas exclude this dye, thus increasing the contrast between the phases, while it is not known to have any measurable effect on clathrate stability.
Glass micromodels are used for the visual observation of the CO2 trapping mechanism. The tests were conducted in the simple methane hydrate - water system, thereby mainly simulating reservoir conditions where excess free water exists in naturally-occurring hydrate-rich sediments.
Formation of small methane hydrate crystals occurs from free CH4 gas, which is then followed by the injection of CO2 into the methane hydrate-water system. Figures 8-A through D are representative micromodel images of the phase distribution prior to and after CO2 injection. This experiment was conducted under temperature and pressure conditions where only simple CH4 hydrates are stable, but not simple CO2 hydrates, and CO2 is in the liquid state. Temperature was kept constant throughout the experiment.
Another important aspect of the subsurface storage of carbon dioxide is that favourable temperatures and pressures are not the only adequate conditions for the formation of gas hydrates in sub-sea sediments. In case of CO2 subsurface injection, be it permafrost / seafloor storage or sequestration coupled with additional energy recovery (such as CO2 injection into methane hydrate reservoirs), hydrate-forming gas concentration (carbon dioxide) should exceed its solubility in water in equilibrium with gas hydrates. This maximum concentration is designated by the solubility of CO2 in the presence of gas hydrates at the given system pressure and temperature. If the CO2 concentration in the water is higher than the solubility limit, excessive CO2 will precipitate from solution producing gas hydrates, thereby forming new gas hydrate cells on the gas hydrate surface.
Methane hydrates were formed from dissolved gas in the region of the hydrate stability zone at 10.8 oC. Figure 8-A shows tiny methane hydrate crystals, whose morphology changes over the next 45 hours in the encircled areas. During this time, CO2 surrounding the CH4 hydrate crystals seems to have displaced methane in the hydrate lattice, however its release in the system is hindered by available dissolved/excess CO2 and by the existing excess water. Once methane is released from the structure it immediately mixes with the dissolved CO2. Since the thermodynamic conditions are favourable for clathrate hydrate formation the mixed CH4-CO2 gas hydrates forms. The time lapsed in this process is very small, about 8 to 9 hours, when we carefully analyze the morphology change in the encircled sections of Figure 8-C over Figure 8-B. Mixed hydrate formation occurred around the locations where methane hydrates were already present before CO2 injection. CO2 hydrates were formed from the gas dissolved in the water phase. The solubility of CO2 played an important role in the formation of CO2 hydrate from the water phase, which resulted in a concentration gradient and the subsequent diffusion across CO2 hydrate-water interface. The re-formation of the mixed-gas hydrates was nearly complete after 89 hours (Figure D). However under real reservoir conditions the heat released from CO2 and mixed CH4-CO2 hydrate formation could result in local temperature increases and the dissociation of methane hydrates. This possible phenomenon was not investigated in this phase of the work. Nevertheless, the glass micromodel proved to be an invaluable facility for generating micro-scale data to better understand gas hydrate formation and dissociation mechanisms.
Was this article helpful?