Improving probabilistic predictions of post-closure groundwater solute loads for Ranger uranium mine

Authors: Sigda, J; Askar, A; Jones, T; Pickens, J; Paulka, S; Harvey, I; Stockdale, R Download Paper Open access courtesy of:

DOI https://doi.org/10.36487/ACG_repo/2415_40

Cite As:
Sigda, J, Askar, A, Jones, T, Pickens, J, Paulka, S, Harvey, I & Stockdale, R 2024, 'Improving probabilistic predictions of post-closure groundwater solute loads for Ranger uranium mine', in AB Fourie, M Tibbett & G Boggs (eds), Mine Closure 2024: Proceedings of the 17th International Conference on Mine Closure, Australian Centre for Geomechanics, Perth, pp. 555-568, https://doi.org/10.36487/ACG_repo/2415_40



Abstract:
Groundwater transport of mine-derived solutes to surface waters and the resulting concentrations after closure of the Ranger mine in the Northern Territory, Australia, are important to stakeholders and Energy Resources of Australia (ERA). In 2020 at ERA’s request and in consultation with stakeholders, INTERA Incorporated developed source terms for 20 constituents of potential concern (COPCs) and conducted a sitewide groundwater uncertainty analysis (GW UA) to estimate the predictive uncertainty in peak COPC loads from groundwater to surface waters. History matching to thousands of head observations constrained uncertainty in hydraulic parameters. Available data informed uncertain future COPC concentrations from each source. The GW UA’s many equally probable predictions of COPC loads became inputs to a surface water model that predicted COPC concentrations in receptor creeks, all supporting ERA’s Pit 3 backfill application. In 2023 INTERA updated source terms and ran a Pit 3-specific UA to support an updated pit backfilling application. This Pit 3 UA focused on pit COPC sources, especially expressed tailings porewater, called pit tailings flux (PTF), which was the primary driver for peak total magnesium loading in the 2021 GW UA. After ERA revised its reclamation plan to reduce the PTF volume, INTERA changed the UA workflow to allow uncertainty in PTF volume and location that had been fixed in 2021. Source concentration probability distribution functions were updated to incorporate new information. The Pit 3 UA applied a Bayesian approach to produce hundreds of equally likely predictions of COPC loads to nearby Magela Creek. Stakeholder input was received throughout implementation. The 2023 results showed a much smaller mean and variance in peak total magnesium loading to Magela Creek than the 2021 results. New information helped constrain the previous input uncertainty, leading to a decrease in predictive uncertainty in the peak load compared to 2021 even though the PTF volume and extent were treated as uncertain.

Keywords: groundwater flow and solute transport model, uncertainty, mine closure, tailings, waste rock, probabilistic prediction

References: