Groundwater baseline sampling programs designed to identify potential leakage from unconventional gas plays in Alberta, Canada

Abstract:

With the rapid expansion of natural gas exploitation from unconventional reservoirs including coalbed methane and shale gas plays, there is significant public concern about potential future contamination of shallow potable groundwater with stray gases, formation waters or chemicals used during hydraulic fracturing. In order to enable a scientifically sound assessment of potential future deterioration of freshwater resources in shallow aquifers, it is essential to first establish and understand the current baseline of groundwater quality including its dissolved or free gases. Since 2006, we have conducted monitoring programs determining the chemical and isotopic compositions of water, its dissolved constituents, and of gases obtained from shallow groundwater and formation fluids collected from coalbed methane and shale gas plays in Western Canada. For groundwater samples, we placed special emphasis on determining the sources of dissolved and free gases using isotope techniques to assess whether gases produced from shale gas plays or potentially leaking from the intermediate zone are isotopically distinct from those in shallow aquifers. Methane and ethane in free gas samples obtained from shallow aquifers (n = 24) were found to have mean δ¹³C values of -72.4 ‰ and -50.2 ‰, respectively. These values are markedly different from the much higher δ¹³C values of methane and ethane in deeper portions of the Western Canadian Sedimentary Basin and in shale gas plays. Therefore, it appears highly feasible to identify potential gas leakage from unconventional gas plays provided that baseline data for shallow groundwater have been determined. Repeat baseline sampling of free gas from selected wells revealed a comparatively low variability of δ¹³C values of methane and ethane of usually < 2 ‰ over periods of several years, suggesting that it is not necessary to conduct baseline analyses more than three times. Also, δ¹³C values of methane in free gas samples and dissolved gas samples typically agreed within 1 ‰ suggesting that both sampling approaches are suitable for determining baseline carbon isotope ratios in case that methane occurs in shallow aquifers.