Geostatistical modelling of mine tailings and comparative analysis of sampling methodologies: A case study of the Otanmäki ilmenite tailings storage facility project

Abstract

Increasing concerns about the availability of Critical Raw Materials (CRMs) have shifted focus to extractive waste sites as potential sources for secondary resources. Resource assessment requires robust investigation and resource definition, necessitating representative sampling and use of an appropriate modelling method. In this study, spatial sampling sensitivity was tested at the Otanmäki tailings storage facility (TSF) by conducting multiple models at different scales, utilizing both previously collected and new data. Additionally, three tailings sampling methodologies were evaluated to ensure sample representativity and minimize contamination: a tube sampler with a valve, a flow-through blade, and an auger drill.

Results indicated that both the auger drill and the tube sampler perform well and provide similar results under low moisture conditions. However, the tube sampler was estimated to be more accurate in varying conditions, despite it having a smaller sample volume and being more time-consuming to use. According to the modelling results, there were minimal changes in grade and tonnage when increasing sampling density or changes in model scale. Thus, the existing resource estimation from the Otanmäki TSF that was based on a 100 × 100 m grid could be considered as having a high-confidence in relation to the G-category in the United Nations Framework Classification for Resources (UNFC) system with effective investigation of tailings sites requires a tailored approach considering factors such as ore deposit type, mineralization style, and information on ore processing and tailings disposal practices. This work provides insights into tailings resource modelling, estimation confidence, sample representativity, and optimal sampling strategies, aiding in the resource classification of tailings sites.