Assessment of resource potential from mine tailings using geostatistical modeling for compositions: A methodology and application to Katherine Mine site, Arizona, USA

The mining industry, in most cases, targets a specific valuable commodity that is present in small quantities within large volumes of extracted material. After milling and processing, most of the extracted material and the effluents are stored as waste (tailings) in impoundments, such as dams or waste dumps, or are backfilled into underground mines. In time, tailing materials may become an issue of environmental and health concern due to the hazardous elements, ions, and oxides contained within the waste material. In addition, handling and storage of such waste in dams may pose the risk of dam failure with catastrophic consequences to nature and nearby communities. On the other hand, tailings may offer potential as secondary sources of critical elements (CEs), including rare earth elements (REEs), which may have been overlooked during primary production and processing. Therefore, treating mine tailings as a resource has economic and environmental benefits by reducing the waste from new and historical mine sites through remining. One of the critical steps for taking advantage of these benefits is to spatially quantify the resources and the pollutants, which require the application of adequate data analysis and modeling methods, often to compositional geochemical data. Utilizing adequate methods is especially important for correctly quantifying resource potential, as the quantities will often be at low concentrations.

This work presents quantification of resource potential (Au, Ag, Cu, Zn, Pb) and elements of environmental concern (Hg and As) from the tailings of a historic mine site, Katherine Mine, AZ, USA. Data reported by the U.S. Bureau of Mines (USBM) after extensive field campaigns in the 1990s, including sampling from tailing impoundment and surrounding areas for geochemical characterization and geophysical surveys, were used. First, compositional data (CoDa) analysis was employed to explore associations of sampling locations, geochemical parts, and the clustering of samples. Next, sequential Gaussian simulation (SGSIM) was applied to samples that showed a genetic link to tailing material after isometric log-ratio transformation (ilr) and mix/max autocorrelation factor (MAF) transformation for spatial modeling and uncertainty evaluation. Geostatistical results revealed spatial variability of concentrations within the tailing area. Uncertainty evaluation based on realizations indicated that Cu (14.27–20.01 t), Zn (44.23–76.23 t), and Pb (22.56–38.28 t) are the most abundant elements within a 5 %–95 % interval, followed by Ag and Au (~5.3 and 0.18 t, at 50th percentile), respectively. Of the elements of health concern, As was found to be ~4.8 t (50th percentile) in the tailing area. The work also showed that ~0.51 t As, 0.005 t Hg, 0.020 t of Au, and 0.62 t of Ag were carried to Lake Mohave by an ephemeral stream called Katherine Wash, which transects the tailings.