Master of Science

thesisGlobal warming enhanced by anthropogenic CO2 is an ever-increasing concern to scientists, policy makers, and the public at large. One plausible method of mitigating growing emissions is to inject CO2 into subsurface aquifers, which have large storage potential, thus helping diminish the accumulation of CO2 in the atmosphere. The earth stores and releases large amounts of CO2 naturally. Studying known natural sources and leaks of CO2 can help to better understand deep-saline CO2 storage. This study focuses on the effects of barometric pressure on surface CO2 leakage through a fault from subsurface CO2 storage. We measured the natural release of CO2 at the Little Grand Wash Fault near the Crystal Geyser, a cold-water geyser located near Green River, Utah. We observed that barometric pressure affects CO2 flux in two different manners, an immediate direct and inverse effect and a longer, 21.5-hour delayed indirect relationship. Two one-dimensional simulation models were developed using the fluid flow equation for vertical flow. Both models simulated gaseous phase CO2 flow from a fully saturated reservoir to the surface. The first model was run to identify a range of permeabilities that resulted in the mean observed surface flux. The second model was run to understand the indirect and time-lagged influence of barometric pressure on the CO2 fault. The observed barometric pressure was propagated through the subsurface and directly added to or subtracted from the reservoir pressure and was able to mimic the resulting change in observed CO2 flux. Reservoir pressures included a range from minimum pressure necessary to induce flow to the surface, through hydrostatic pressure to artesian conditions established by water infiltration that recharges the CO2 reservoir from two elevations much higher than the seepage site. Even though barometric pressure flux ranges can be accounted for in the near subsurface, the simulation results from both models imply that daily barometric pressure fluctuations do not greatly affect the leakage rates of deeply-stored CO2. Conversely, the 21.5-hour trend can be shown to directly affect the surface seepage of CO2 in a partially confined reservoir