Natural hazards and mineral commodity supply: Quantifying risk of earthquake disruption to South American copper supply

Abstract

Mineral resources, and their mining and enrichment operations, are not equally distributed across Earth. The concentration of mineral supply in certain regions, owing to the geology or geography of the mineral resource, raises the level of risk related to supply disruption. Where mineral production coincides with areas prone to natural hazards, supply may be especially at risk. However, the level of risk that natural hazards pose to mineral supply has yet to be quantified on a global or regional scale. Using copper in South America as a case study, this paper offers methods for quantifying (i) the coincidence of mineral production and seismic hazards, and (ii) the Expected Annual Disruption (EAD) of the mineral supply from earthquakes. The first of these methods indicates that, of the 101 copper producing facilities in South America considered, 76 are located within an area of high seismic hazard, taken here as the area with>85% chance of exceeding Modified Mercalli Intensity VI earthquake shaking in 50 years. Collectively, the 76 facilities comprise 82%, 87%, and 91% of the 2015 South American mine production, smelter capacity, and refinery capacity, respectively. For each of the 101 facilities, the second method calculates the EAD using a full earthquake shaking hazard forecast at the location, the annualized copper production of the facility, and models of the vulnerability of that production to shaking. The EADs are summed by country, here within South America, as a demonstration of how supply risk could eventually be quantified globally. Consideration of two illustrative vulnerability models shows that future work is needed to determine percentages of disruption to mineral production for different levels of earthquake shaking. Ultimately, the methods presented herein could be applied to other mineral commodities and/or adapted for other natural hazards, and the resulting EADs could be summed. Results from these methods could be used to focus more detailed risk assessments where the risk is highest.