ASSESSMENT OF THE EFFECTS OF CO2 INJECTION ON A QUARTZ-RICH SANDSTONE RESERVOIR THROUGH BATCH REACTION EXPERIMENTS: CASE STUDY OF THE FARNSWORTH, TEXAS CARBON SEQUESTRATION SITE

The Farnsworth, Texas hydrocarbon field has been an ongoing site of CO₂ injection since 2010, totaling around 1 million tonnes of CO₂ thus far. The destination for the CO₂ is the Morrow B Sandstone, which is composed predominantly of quartz with lesser amounts of albite, chlorite, calcite, and clay minerals. A key concern about the CO₂ injection process is how the mineralogy, porosity, and pore water composition of the Morrow B will change. This concern has been addressed by several numerical modeling studies, which predict substantial changes in mineralogy and pore water composition but generally negligible changes in porosity. However, because of the implications of these changes for CO₂ storage efficacy and risk assessment, it is important to validate the theoretical model predictions where possible through laboratory experiments. To that end, batch reaction experiments were conducted to simulate conditions in the Morrow B Sandstone near an injection well and at further distances from an injection well where the CO₂ has been significantly diluted by the formation water. The experiments were conducted by immersing thin sections of a coarse- and a fine-grained facies of the Morrow B Sandstone in variably CO₂-charged samples of formation water in hydrothermal reaction vessels maintained at a temperature of 75° C. An initial experimental run was conducted for 61 days and a second run was conducted for 73 days using the same initial fluid composition as in the first experimental run. Changes in mineralogy in the thin sections were determined by SEM using the Tescan Integrated Mineral Analyzer (TIMA), and changes in formation water composition were determined by ICP-AES.

In each of the experiments, a thin film of dolomite and silica grains precipitated on the surface of the thin sections, accompanied by sharp decreases in Ca, Mg, and Sr in the formation water. This is consistent with numerical model predictions that dolomite should be the main mineral sink for injected CO₂ and that the formation water is supersaturated with respect to silica. Also consistent with numerical modeling results are that the native reservoir minerals, albite, muscovite, and illite dissolved and calcite precipitated in the lower carbon fluid representing further distances from the CO₂ injection wells.