Improving the Ease and Accuracy of Environmental Monitoring at CCS Sites

The ability to quickly and accurately attribute the source of CO2 found in the near-surface above CCS sites is important for leakage detection, public engagement, leakage quantification and environmental protection at these sites. Comparison of background or baseline CO2 concentrations in the vadose zone pre- and post- injection is thought to be the best way of identifying signal from noise. However, historical research at the Weyburn-Midale Monitoring and Storage Project (WMP) shows that leakage attribution using the popular but complex ‘baseline comparison” technique will not be accurate for MVA; the background monitoring area at the WMP produced CO2 concentrations consistently lower than in the on-site monitoring grid. Detailed analysis of direct and indirect monitoring parameters, surface conditions, and weather parameters indicated no leakage at the project site; however, if a baseline comparison technique had been strictly applied, leakage would have been inferred at WMP.  Thus, research at the WMP shows that baseline measurements alone are not sufficient for attribution, and that additional complex analysis and comparison with a multitude of methods and a number of parameters is required for this approach.

Findings from a number of US DOE demonstration projects and at the Kerr Farm near the WMP, indicate that the use of simple geochemical ratios may be easier and more accurate than currently-accepted baseline techniques. One of these methods, known as a ‘process-based’ technique, can promptly attribute the source of near-surface CO₂ and has been demonstrated at Cranfield and ZERT and at the Kerr Farm and is currently being implemented and/or tested at the Otway project (Australia), Aquistore (Canada), CO₂ Field Lab (Norway), Lac Rousse (France) and the U.S. DOE Bell Creek demonstration site (Montana, USA). The method is also being adapted in Japan for use in offshore sediments. The process-based technique is a low-cost technique that does not require large data sets or prolonged baseline, provides instant data reduction and graphically-displayed results, and thus is simple enough for the general public to implement. Because this method does not rely on prolonged baseline data sets, it creates the potential for a paradigm shift in near-surface monitoring.