Pereira, N & Prabhu, SS 2026, 'Three-dimensional dam breach assessment of a closed 
tailings storage facility using the material point method', in AB Fourie, G Boggs, J Heyes & M Tibbett (eds), Mine Closure 2026: Proceedings of the 19th International Conference on Mine Closure, Australian Centre for Geomechanics, Perth, pp. 1-16, https://doi.org/10.36487/ACG_repo/2615_49
(https://papers.acg.uwa.edu.au/p/2615_49_Prabhu/)
Abstract: Dam breach assessments (DBAs) for tailings storage facilities (TSFs) have historically been undertaken using hydraulic modelling approaches developed primarily for operational conditions. However, closure significantly alters key assumptions related to geometry, material behaviour, saturation, and credible failure mechanisms. Recognising these limitations, recent guidance from organisations such as the Canadian Dam Association (CDA) and the US Society on Dams (USSD) encourages the use of advanced numerical models that more realistically represent site‑specific conditions, particularly for closed tailings facilities. For mine closure planning, reducing unnecessary conservatism in DBA outcomes is critical to enable risk‑informed allocation of closure and post‑closure resources.
This paper presents a three‑dimensional DBA undertaken for a closed TSF where complex geometry and proximity of critical infrastructure necessitated a fully 3D approach. The assessment was developed directly from a site‑specific failure modes and effects analysis (FMEA) updated for closure conditions with all credible failure modes superimposed as a conservative bounding construct rather than a credible compound event. Initial DBAs for the site, based on conventional hydraulic modelling with rheological assumptions, predicted extensive inundation of nearby office infrastructure. These outcomes were reassessed using a material point method (MPM) model, which explicitly incorporates geotechnical material behaviour and large‑deformation failure mechanics.
The 3D MPM results demonstrated that, under worst‑case breach scenarios, inundation of the relevant infrastructure was not predicted when mean material properties were adopted. To further strengthen confidence in the findings, a statistical sensitivity analysis was undertaken by varying key dam parameters by one and two standard deviations. The results showed that even under these reduced-strength cases, predicted inundation remained limited in extent. The study highlights the value of three‑dimensional, physically based modelling for DBAs at closed TSFs, demonstrating how advanced numerical methods can refine consequence estimates and support more defensible, risk‑informed mine closure decision‑making.

Keywords: material point method, dam breach assessment, DBA, MPM