Leakage of stored bulk phase CO₂ is one risk for sequestration in deep saline aquifers. As the less dense CO₂ migrates upward within a storage formation or in layers above the formation, the security of its storage depends upon the trapping mechanisms that counteract the migration. The trapping mechanism motivating this research is local capillary trapping (LCT), which occurs during buoyancy-driven migration of bulk phase CO₂ within a saline aquifer with spatially heterogeneous petrophysical properties. When a CO₂ plume rising by buoyancy encounters a region where capillary entry pressure is locally larger than average, CO₂ accumulates beneath the region. One benefit of LCT, applied specifically to CO₂ sequestration and storage, is that saturation of stored CO₂ phase is larger than the saturation for other permanent trapping mechanisms. Another potential benefit is security: CO₂ that occupies local capillary traps remains there, even if the overlying formation that provides primary containment were to be compromised and allow leakage. Most work on LCT has involved numerical simulation (Saadatpoor 2010, Ganesh 2012); the research work presented here is a step toward understanding local capillary trapping at the bench scale. An apparatus and set of fluids are described which allow examining the extent of local capillary trapping, i.e. buoyant nonwetting phase immobilization beneath small-scale capillary barriers, which can be expected in typical heterogeneous storage formation. The bench scale environment analogous to CO₂ and brine in a saline aquifer is created in a quasi-two dimensional experimental apparatus with dimension of 63 cm by 63 cm by 5 cm, which allows for observation of plume migration with physically representative properties but at experimentally convenient ambient conditions. A surrogate fluid pair is developed to mimic the density, viscosity and interfacial tension relationship found at pressure and temperature typical of storage aquifers. Porous media heterogeneity, pressure boundary conditions, migration modes of uprising nonwetting phase, and presence of fracture/breach in the capillary barrier are studied in series of experiments for their influences on LCT. A variety of heterogeneous porous media made of a range of sizes of loosely packed silica beads are used to validate and test the persistence of local capillary trapping mechanism. By adjusting the boundary conditions (fluid levels in reservoirs attached to top and to bottom ports of the apparatus), the capillary pressure gradient across the domain was manipulated. Experiments were conducted with and without the presence of fracture/potential leakage pathway in the capillary seal. The trapped buoyant phase remained secure beneath the local capillary barriers, as long as the effective capillary pressure exerted by the trapped phase (proportional to column height of the phase) is smaller than the capillary entry pressure of the barrier. The local capillary trapping mechanism remained persistent even under forced imbibition, in which a significantly higher hydraulic potential gradient, and therefore a larger gradient in capillary pressure, was applied to the system. The column height of buoyant fluid that remained beneath the local capillary barrier was smaller by a factor corresponding to the increase in capillary pressure gradient. Mimicking a breach of the caprock by opening valves at the top of the apparatus allowed buoyant mobile phase held beneath the valves to escape, but buoyant phase held in local traps at saturations above residual, and therefore potentially mobile, was undisturbed. This work provides systematic validation of a novel concept, namely the long-term security of CO₂ that fills local (small-scale) capillary traps in heterogeneous storage formations. Results from this work reveal the first ever unequivocal experimental evidence on persistence of local capillary trapping mechanism. Attempts to quantify the nonwetting phase saturation and extent of LCT persistence serve as the initial steps to potentially reduce the risks associated with long-term storage security. / text

Coal beds are one of the most promising reservoirs for geologic carbon dioxide (CO₂) sequestration, as CO₂ can strongly adsorb onto organic matter and displace methane; however, little is known about the long-term fate of CO₂ sequestered in coal beds. The "2800' sand" of the Olla oil field is a coal-bearing, oil and gas-producing reservoir of the Paleocene–Eocene Wilcox Group in north-central Louisiana. In the 1980s, this field, specifically the 2800' sand, was flooded with CO₂ in an enhanced oil recovery (EOR) project, with 9.0×10⁷m³ of CO₂ remaining in the 2800' sand after injection ceased. This study utilized isotopic and geochemical tracers from co-produced natural gas, oil and brine from reservoirs located stratigraphically above, below and within the 2800' sand to determine the fate of the remaining EOR-CO₂, examining the possibilities of CO₂ migration, dissolution, mineral trapping, gas-phase trapping, and sorption to coal beds, while also testing a previous hypothesis that EOR-CO₂ may have been converted by microbes (CO₂-reducing methanogens) into methane, creating a microbial "hotspot". Reservoirs stratigraphically-comparable to the 2800' sand, but located in adjacent oil fields across a 90-km transect were sampled to investigate regional trends in gas composition, brine chemistry and microbial activity. The source field for the EOR-CO₂, the Black Lake Field, was also sampled to establish the δ¹³C-CO₂ value of the injected gas (0.9‰ +/- 0.9‰). Four samples collected from the Olla 2800' sand produced CO₂-rich gas with δ¹³C-CO₂ values (average 9.9‰) much lower than average (pre-injection) conditions (+15.9‰, average of sands located stratigraphically below the 2800' sand in the Olla Field) and at much higher CO₂ concentrations (24.9 mole %) than average (7.6 mole %, average of sands located stratigraphically below the 2800' sand in the Olla Field), suggesting the presence of EOR-CO₂ and gas-phase trapping as a major storage mechanism. Using δ¹³C values of CO₂ and dissolved organic carbon (DIC), CO₂ dissolution was also shown to be a major storage mechanism for 3 of the 4 samples from the Olla 2800' sand. Minor storage mechanisms were shown to be migration, which only affected 2 samples (from 1 well), and some EOR-CO₂ conversion to microbial methane for 3 of the 4 Olla 2800' sand samples. Since methanogenesis was not shown to be a major storage mechanism for the EOR-CO₂ in the Olla Field (CO₂ injection did not stimulate methanogenesis), samples were examined from adjacent oil fields to determine the cause of the Olla microbial "hot-spot". Microbial methane was found in all oil fields sampled, but indicators of methanogenesis (e.g. alkalinity, high δ¹³C-DIC values) were the greatest in the Olla Field, and the environmental conditions (salinity, pH, temperature) were most ideal for microbial CO₂ reduction in the Olla field, compared to adjacent fields.