Surface gas measurements and related studies for the characterization and monitoring of geological CO2 storage sites; experiences at Weyburn and in Salah.

Abstract

Preliminary baseline soil gas data collected in the summer and autumn of 2001 above the Phase 1A injection area of the EnCana Enhanced Oil Recovery project at the Weyburn oilfield in south Saskatchewan was presented at GHGT-6 in Kyoto. Data can now be presented for all three years of the study with conclusions, the predominant one being that the major controls on soil gas levels are seasonal and meteorological with no indications of leakage from depth. In the autumns of 2002 and 2003 further in situ monitoring of CO2, CO2 flux, O2, CH4, radon (222Rn) and thoron (220Rn) was carried out. Soil gas samples were also collected for laboratory analysis of helium, permanent gases, sulphur species and light hydrocarbons. All sampling was repeated over the same 360 point sampling grid and more detailed profiles for both follow-up years. Marked changes in CO2 levels (especially flux) for each of the three-year datasets indicate changes in surface conditions, rather than CO2 from a deeper source. The radon and thoron data was found to be similar for the three years but appears to vary in response to drift composition, and seasonal effects, rather than migration from a deep source. In 2003 further work was agreed in addition to the main grid and profile data. A control area was sampled for the same suite of gases, 10km to the northwest of the oil field. It included similar topography, land use and drift composition to the main sampling grid. There were 35 sample locations on a 7 x 5 point grid with 100m spacing and two additional sites. Early conclusions indicate that the soil gas results in the control area are very similar to those from the main grid, vindicating control site selection and further supporting a lack of deeply sourced CO2 over the injection area. Along with the control site, five zones of possible CO2 leakage were also surveyed and sampled. Two cross a river lineament that may be associated with deep faulting, two were abandoned oil well sites and one site overlays a deep salt dissolution feature. (Unfortunately CO2 flux and gamma measurements were not carried out at these sites.) A northeast/southwest trending lineament, just north of the main grid, was sampled along two profiles perpendicular to the feature, with an increased density of sampling over the feature. The feature generally followed an incised river valley and anomalous CO2 was only detected on the valley floor, where it would be expected as there was lush vegetation in this zone. There were no coincident anomalies for other gases. Soils around two abandoned wells were also sampled. A 16-site grid was surveyed around each well. One well had been completely abandoned and the other was suspended due to failed casing. Such boreholes represent possible points of weakness that may be routes for CO2 migration. The well with failed casing had weakly anomalous CO2 locally to the south, again unmatched for other gases. The fully abandoned well had background CO2 values. Two perpendicular profiles of 10 sites at 25m spacing were sampled for soil gas over the mapped centre of the dissolution feature. Background values were obtained. In 2003 two vertical profiles were performed both indicating an increase in CO2 to a depth maximum of 1.80m; this increase is matched by a corresponding decrease only in O2, indicating biological respiration. Radon concentration indicated no anomalies. Portable gamma spectrometric data was collected in 2003 over the west-centre area of the grid, the profiles and over the control grid. The composition of soils from both areas was found to be very similar.