Impact of draw strategy on wet muck spill hazard severity at the Deep Ore Zone mine

Abstract

A wet muck spill is a sudden inflow of wet material into an underground mine excavation, typically from an extraction level drawpoint. Wet muck spills pose a major risk to personnel safety, assets (equipment and infrastructure) and production. A key factor that determines the extent and impact of a wet muck spill is its severity, which can be described by its volume, flow velocity and runout distance. Evidence from historical wet muck spill events has shown that the volume and runout distance of a spill can exceed 3,000 m3 and 200 m, respectively. Wet muck spill severity assessments are an essential component of a mine’s risk management plan, serving to guide the delineation of potential wet muck impact zones and the development of effective mitigation strategies. The potential for the occurrence of a severe spill event increases as the amount of fines and water exposed at a drawpoint increase. One strategy to mitigate the risk of a severe spill event is to limit draw at high-risk drawpoints or close them entirely. However, this approach might not be as effective for a mature cave with high quantities of rapidly percolating fine material and stored water, where many drawpoints are likely prone to severe spills. In such conditions, understanding how draw strategy influences spill severity can guide risk-informed draw optimisation, with the goal of minimising severe spill events. A practical approach towards a better understanding of how draw strategy contributes to severe spills must include a systematic review of historical spill data. In this study, statistical methods were used to analyse spill data from the Deep Ore Zone (DOZ) cave mine in Papua, Indonesia, with the aim of identifying trends between the probability of a high-volume spill (defined as spill volume > 500 m3) and draw strategy. Draw strategy was quantified by two draw-related variables: draw rate and differential draw index. Spill probability was calculated based on the frequency of high-volume spill observations at different ranges of these variables. Results from this study showed that high-volume spills were more likely to happen under high draw rates and non-uniform draw conditions. The probability of a high-volume spill was highest when both draw rate and differential draw index were in their upper ranges, and much less when either of these variables were in their lower ranges.