A comparison of three methods for monitoring CO2 migration in soil and shallow subsurface in the Ressacada Pilot site, Southern Brazil

Abstract

In a joint R&D project under the full sponsorship of PETROBRAS, the Brazilian National Oil Company, the first CO2 monitoring field lab was started-up in Brazil in 2011.The site chosen, the Ressacada Farm, in the Southern region of the country, offered an excellent opportunity to run controlled CO2 release experiments in soil and shallow subsurface (< 3 m depth). This paper focuses on the presentation and comparison of the results obtained using electrical imaging, CO2 flux measurements and geochemical analysis of the groundwater to monitor CO2 migration in both saturated and unsaturated sand-rich sediments and soil. In 2013 a controlled release campaign was run, covering an area of approximately 6,300 m2. Commercial food-grade gaseous carbon dioxide was continuously injected at 3 m depth for 12 days. The average injection rate was 90 g/day, totaling ca. 32kg of gas being released. The low injection rate avoided fracturing of the unconsolidated sediments composing the bulk of the local soil matrix. Monitoring techniques deployed during 30 consecutive days, including background characterization, injection and post-injection periods, were: (1) 3D electrical imaging using a Wenner array, (2) soil CO2 flux measurements using accumulation chambers, (3) water sampling and analysis, (4)3D (tridimensional) and 4D (time-lapsed) electrical imaging covering depth levels to approximately 10 m below the surface. Water geochemical monitoring consisted of the analyses of several chemical parameters, as well as acidity and electrical conductivity in five multi-level wells (2m; 4m and 6 m depth) installed in the vicinity of the CO2 injection well. Comparison of pre- And post-injection electrical imaging shows changes in resistivity values consistent with CO2migration pathways. A pronounced increase in resistivity values occurred, from 1,500 ohm.m to 2,000 ohm.m, in the vicinity of the injection well. The accumulation chamber assessment show significant changes in the CO2 flux during the release experiment: maximum values detected were ca. 270 mmol/m2/s(during injection) as compared to background values of c.a. 34mmol/m2/s. The pH showed variations after CO2 injection in two monitoring wells at 2m, 4m and 6m depth. After the CO2 injection ceased, the lowest pH measured was 4.1, which represents a decrease of 0.5 relative to the background values. Slight variations in the oxidation-reduction potential (Eh) were observed near the CO2 injection well. There was a decreasing trend of this potential, especially in a monitoring well at 6m depth, ranging from 308mV to 229mV, between the background and the injection scenarios. Ppb level increments were detected in the measurements carried out for the major cations (Ca, Mg, Na, and P) and trace elements (Ag, Al, As, B, Ba, Cd, Pb, Cu, Cr, Ni, Mn, S, V, and Zn). Electrical conductivity and alkalinity, however, remained constant throughout the experiment, with values around 40 μS.cm-1 and 2.5 mgCaCO2.L-1, respectively. The response to CO2 injection was not uniformly observed by the different methods deployed on site. The highest percentage change in resistivity values near the injection well occurred 5 days after the injection had started. However the highest percentage changes in the CO2 flux values occurred 9 days after the injection, 4 days after the observed changes in resistivity values. This delay is probably due to the migration time of the gas from 0.5m depth to the surface.