Assessing Connectivity Between an Overlying Aquifer and a Coal Seam Gas Resource Using Methane Isotopes, Dissolved Organic Carbon and Tritium

Coal seam gas (CSG) production can have an impact on groundwater quality and quantity in adjacent or overlying aquifers. To assess this impact we need to determine the background groundwater chemistry and to map geological pathways of hydraulic connectivity between aquifers. In south-east Queensland (Qld), Australia, a globally important CSG exploration and production province, we mapped hydraulic connectivity between the Walloon Coal Measures (WCM, the target formation for gas production) and the overlying Condamine River Alluvial Aquifer (CRAA), using groundwater methane (CH4) concentration and isotopic composition (δ13C-CH4), groundwater tritium (3H) and dissolved organic carbon (DOC) concentration. A continuous mobile CH4 survey adjacent to CSG developments was used to determine the source signature of CH4 derived from the WCM. Trends in groundwater δ13C-CH4 versus CH4 concentration, in association with DOC concentration and 3H analysis, identify locations where CH4 in the groundwater of the CRAA most likely originates from the WCM. The methodology is widely applicable in unconventional gas development regions worldwide for providing an early indicator of geological pathways of hydraulic connectivity.


Data Analysis and Quality Control.
The mobile Picarro G2301 was calibrated before, during and after the field survey using 6 different laboratory primary standards at RHUL (Table S1). All of these had an offset that fell within the precision range of the instrument as specified by the manufacturer. Two calibrations during the survey were done using a Southern Ocean Air standard FB03727 provided by CSIRO.
The offset was larger with these calibrations, possibly due to the effect of out-of-lab conditions. week and was 0.00035 ppm. Each sample was analysed for 4 minutes on the Picarro G1301 and the concentration value was taken as the mean of the last 2 minutes 20 seconds of the measurement. Analysis of δ 13 C-CH 4 was carried out using a modified gas chromatography isotope ratio mass spectrometry (GC-IRMS) system (Trace Gas and Isoprime mass spectrometer, Isoprime Ltd.) with 0.05‰ repeatability. 58 All measurements were made in triplicate.
The ambient air samples from the January and March field campaigns were analysed in March. The air was passed into the Picarro G2132-i via a Teflon tube with a simple open/close valve system that allowed either outside air or the sample in the Tedlar bag. Each sample was analysed for 10 minutes, which allowed for stabilization of the readings. Values were taken as the mean over a 2-minute period once the readings were stable. In between each sample, ambient air was run through the Picarro for at least 5 minutes so as to leave clear markers between each bag sample and to avoid contamination from previous bags.
To calibrate the Picarro G2132-i we measured the 94 Tedlar bag samples on both the G2132-i and the GC-IRMS system at RHUL. 58 A cross-plot of the isotopic values obtained on the GC-IRMS versus the Picarro G2132-i is shown in Figure S1. The 1:1 line is shown in the graph.
At lower δ 13 C-CH 4 values the Picarro G2132-i readings match the GC-IRMS readings. However, as the isotopic values became more enriched in 13 C the Picarro G2132-i readings drifted upwards away from the 1:1 line. We performed a linear correction on the isotopic values obtained by the Picarro G2132-i and found that this drift was within ±1‰ bounds of the corrected 1:1 line, resulting in a total drift of 2‰ ( Figure S2). This drift of 2‰ on the Picarro G2132-i can be clearly seen when there were numerous samples run with very similar δ 13 C-CH 4 values. In Figure S1 it is clear that when the isotopic value on the GC-IRMS is -47‰ the drift over 24 hours on the Picarro G2132-i is between -46‰ and -48‰. This falls within the drift of < 2‰ over 24 hours quoted by the manufacturer.
A linear correction was also obtained for the Picarro G2132-i [CH 4 ] against a NOAA standard calibrated Picarro G1301 at RHUL using the 94 ambient air samples ( Figure S3). These values were found to have a total precision of 0.002 ppm. This value falls within the precision quoted in the manufacturer's specifications.
The δ 13 C-DIC samples were analysed at the Australian Nuclear Science and Technology The water samples were distilled and enriched electrolytically prior to analysis by liquid scintillation. Tritium activity was measured by counting beta decay in a liquid scintillation counter (LSC) and had a combined standard uncertainty of ± 0.03 TU. Only samples that measured above the quantification limit (> 0.04 TU) were considered to have measurable 3 H activity.