• journal_title：AAPG Bulletin
• Contributor：Shuiquan Li ; Ye Zhang ; Xu Zhang ; Changan Du
• Publisher：American Association of Petroleum Geologists (AAPG)
• Date：2012-
• Format：text/html
• Language：en
• Identifier：10.1306/07261110178
• journal_abbrev：AAPG Bulletin
• issn：0149-1423
• volume：96
• issue：4
• firstpage：635
• section：Geohorizons

The Moxa arch anticline is a regional-scale northwest-trending uplift in western Wyoming where geologic storage of acid gas (carbon dioxide, methane, nitrogen, hydrogen sulfide, ethane) is under consideration. Nugget Sandstone, a deep saline aquifer at depths exceeding 17,000 ft (5180 m), is a candidate formation. This study builds three-dimensional local- to regional-scale geologic and fluid-flow models for the Nugget Sandstone and its neighboring formations. Geologic and engineering characterization data were assembled and screened for accuracy. Using geostatistical simulations (first, sequential indicator simulation of facies, then the sequential Gaussian simulation of porosity [ϕ]), the data were integrated to create a regional-scale geologic model from which a local-scale simulation model surrounding the proposed injection site was extracted. Using this model, acid gas injection was simulated for 50 yr, followed by 1950 yr of monitoring. A sensitivity analysis was conducted, exploring the impact of geologic and engineering variables on model predictions. Results suggest that, at the simulation time scale, low dissolved solids in formation water, large gas-phase relative permeability (k_rg) hysteresis, and low vertical-to-horizontal intrinsic permeability (k) anisotropy all contribute to enhanced storage of acid gas in both residual (trapped) and dissolved forms. The parameter that exerts the largest control on gas storage is relative permeability hysteresis. However, given parameter uncertainty, the total predicted gas storage varies significantly. Prediction uncertainty increases in the order of dissolved gas, trapped gas, and mobile gas. In comparison, petrophysical uncertainty, as represented by multiple ϕ realizations, has limited impact on prediction, although future work is needed to expand the uncertainty analysis by developing alternative facies models for the storage formations.