Long term reactivity of CO2 in a low salinity reservoir-seal complex

Abstract

An understanding of the long-term reactivity of different rock types to injected CO2 is needed for sequestration site assessment. Relative to saline aquifer studies, the long term reactivity of CO2 in low salinity aquifers has received little attention. Currently in Australia, the Surat Basin is being appraised for its large-scale CO2 storage potential within low salinity aquifers. Sixteen core samples from the Precipice Sandstone and Evergreen Formation – the notional target reservoir and seal complex – were characterized for mineral content; helium, mercury-injection and micro CT porosities; air permeability; and, imaged with SEM-EDS. Samples consisted of quartz rich reservoir sandstones, feldspar and clay rich or calcite cemented sandstones (secondary reservoir), mudstones (sealing complex), and oolitic ironstones (sealing complex) derived from braided river, fluvial-deltaic, and restricted marine shoal depositional environments, respectively. The reservoir sandstone samples characterized here had measured total porosity that ranged from 11 to 23% with pore throats mainly between 90 and 100 μm, and core air permeability from 558 to 3397 mD. In the Precipice Sandstone reservoir sample μCT plugs, 98% of the pore space was connected with calculated vertical permeability 145–4611 mD and horizontal 4291–8200 mD. Feldspar and clay rich sandstone and mudstone samples from the overlying Evergreen Formation had porosity that ranged between 0.2 and 22.9%, with a wide range of pore throat sizes from ~0.005 to 30 μm, and permeability from 0.2 to 28.1 mD, respectively. Ironstone and mudstone samples from the Westgrove Ironstone Member (Evergreen Formation) had porosity from 0.7 to 9.7% and a low permeability of 0.04 mD.Kinetic geochemical CO2 reactivity models made predictions over two time-scales: 30 or 1000 years. Selected models also accounted for the potential presence of 10 ppm SO2 gas. The Precipice Sandstone quartz-rich reservoir sandstones had consistently small amounts of reactive minerals and mineral trapping or scaling of the reservoir was not predicted over 30 years, with the pH approximately 4.5 after 30 years. Samples from the Evergreen Formation included feldspar and clay rich sandstones and mudstones, several contained variable amounts of carbonate cement. Their response to CO2 was more influenced by mineral content than rock type. Plagioclase feldspars and Fe-rich chlorite were the main silicate minerals that reacted to produce siderite and ankerite mineral trapping up to 2.57 kg/m3 CO2. In the very unlikely event that CO2 rich fluids migrated upwards as far as the Westgrove Ironstone Member, chlorite is predicted to alter to siderite. This study indicates that the Precipice Sandstone reservoir in the study region has a low likelihood of mineral scaling which is favorable to avoid CO2 injectivity issues. Mineral trapping as ankerite and siderite could be expected to trap CO2 in the chlorite and plagioclase rich Evergreen Formation seal lithologies. Further work is suggested on validating long term predictions with observation data from natural analogue studies.