Reactive transport simulation to assess geochemical impact of impurities on CO2 injection into siliciclastic reservoir at the Otway site, Australia

It is generally recognized that the removal of impurities from CO2 obtained from coal-fired flue gas carbon capture systems can significantly increase the cost of Carbon Capture and Storage (CCS). During the period from September to December 2014, the CO2CRC in collaboration with Callide Oxyfuel Services Pty Ltd conducted a series of CO2 injection tests into the Paaratte formation (Otway Basin, Victoria, Australia) utilizing product CO2 (near food grade) and product CO2 with added impurities (nominal SO2: 67 ppmv, O2: 6150 ppmv and NOx: 9 ppmv) from the Callide Oxyfuel Project. The purpose of the injection testing was to assess the geochemical effect of the CO2 injected with water on the siliciclastic reservoir. The geochemical test, which was designed and conducted by CO2CRC, consists of water production, Test 1 (injection of CO2-saturated water without impurities), and Test 2 (injection of CO2-saturated water with impurities). In this paper, 2-D radial reactive transport simulations of injection tests were carried out using TOUGHREACT, including: (1) simulation of the pre-injection period from Otway Stage 2B Test in 2011 to the start of the geochemical test; (2) simulations of the pure CO2 and the impure CO2 cases; (3) long-term simulation of the impure CO2 case; and (4) simulations of injection of CO2-saturated water with various impurity concentrations. The numerical model was able to reproduce the observed geochemical changes during the geochemical test. The results indicate that there were negligible differences between the pure CO2 and the impure CO2 cases for both changes in water chemistry and mineralogy, if SO2 or O2 was less than 1000 ppmv. Based on this evaluation of both field test results and simulations, by extrapolation it is concluded that co-injection of impurity gases (up to 1000 ppmv SO2 and 1000 ppmv O2) with CO2 would have insignificant impacts on the reservoir at the Otway site. In the long-term simulation, the distribution area of residual gas in the reservoir was small and the pH of the formation water was near neutral after 1000 years. Moreover, it was predicted that the water disturbed by CO2 impurities would come close to the initial geochemical conditions in the long-term, and that it is highly probable that the long-term impact after 1000 years would become much less compared to the short-term impact in the test period. The method and model presented here was suitable for the Otway site test and would be applicable to other CO2 storage sites with similar conditions.