Effects of salinity on hydrate stability and implications for storage of CO2 in natural gas hydrate reservoirs

AbstractThe win-win situation of CO2 storage in natural gas hydrate reservoirs is attractive for several reasons in addition to the associated natural gas production. Since both pure CO2 and pure methane form structure I hydrate there is no expected volume change by replacing the in situ methane with CO2, and there is not net production of associated water which requires extra handling. The geo-mechanical implication of the first of these may be a very important issue since hydrates in unconsolidated sediments are the most promising targets for exploitation of natural gas. The stability of CO2 stored in the form of hydrate is probably one of the safest options today, even though also this option relates to safety of sealing cap-rock or clay layer. The stability of hydrates in a reservoir depends on many factors, including the interactions between minerals, surrounding fluids and hydrate. The natural level of salinity increases with depth in a reservoir. In addition formation of hydrate will lead to increased salinity of the fluids surrounding the formed hydrate. This may lead to liquid pockets of residual aqueous solution with increased salinity as well as very non-uniform hydrate. The latter due to the fact that hydrate composition and stability relates to properties of surrounding fluids. In the work presented here methane hydrates were formed in several sandstone cores. The cores were all partially saturated with brine of different salinities in order to identify the effect salinity has on the fill fraction, the amount of methane per available structural site in hydrates. The results indicate that salinities lower than regular sea water composition has no significant impact on the fill fraction of methane hydrate in porous media. When the salinity surpasses regular sea water composition there is a significant drop in fill fraction. The methane hydrate fill fraction is dominated by total brine salinity rather than brine distribution in the core