Deriving background concentrations of contaminants of potential concern in groundwater: an example from the Ranger uranium mine, Australia

Authors: Weaver, TR; Kovacsy, D; Windle, DE; Gemson, W; Hausmann, N; Shibata, M; Staggs, D; Iles, M Download Paper Open access courtesy of:

DOI https://doi.org/10.36487/ACG_repo/2215_50

Cite As:
Weaver, TR, Kovacsy, D, Windle, DE, Gemson, W, Hausmann, N, Shibata, M, Staggs, D & Iles, M 2022, 'Deriving background concentrations of contaminants of potential concern in groundwater: an example from the Ranger uranium mine, Australia', in AB Fourie, M Tibbett & G Boggs (eds), Mine Closure 2022: Proceedings of the 15th International Conference on Mine Closure, Australian Centre for Geomechanics, Perth, pp. 691-704, https://doi.org/10.36487/ACG_repo/2215_50



Abstract:
In the absence of pre-mining groundwater quality data, ERM interrogated nearly 220,000 data points in the Ranger uranium mine groundwater database to establish a background dataset for analytes for each hydrolithologic unit (HLU) at the site. These datasets were used to identify if analytes in groundwater were operations related, and therefore contaminants of potential concern (COPCs), and to develop background threshold values (BTVs) to inform groundwater monitoring programs and mine closure activities. The assessment assumes COPC concentrations from monitored areas comprise both operations-derived and background concentrations. The approach applied allows a site-specific background dataset to be extracted from a dataset obtained from impacted areas at a site; it relies on a weight-of-evidence approach consistent with that described in guidance from the US Navy, Interstate Technology and Regulatory Council and the US Environmental Protection Agency. In this assessment, if analyte concentrations were not related to mining activities (i.e. derived only from background conditions), the analyte was not considered to be a COPC. BTVs were then developed for the background datasets to support decision-making during closure at the site. Ranger data was compiled and reviewed so as to meet data quality standards. Iterative population partitioning was used to identify breakpoints in each HLU analyte–specific dataset using quantile-quantile (QQ) plots, independent of site-qualifying information. The breakpoint was refined using multiple lines of evidence, including temporal concentration trends, covariance with known site sources, expected source composition, and spatial patterns of impacts. For each analyte within a given HLU, 95/95 upper tolerance limits were used as BTVs for each background dataset. This approach allows background concentrations to be developed for sites where insufficient up-hydraulic gradient background or pre-development data are available, thus capitalising on a site’s operational database. This is particularly relevant in mineralised areas where background (naturally occurring) concentrations of metals or other inorganic constituents may be higher than guideline or default trigger values and concentrations.

Keywords: background threshold values, contaminants of potential concern, quantile-quantile plots, groundwater monitoring programs, mine closure

References: