Thin layering in tailings deposits and its implications for CPT-based state parameter estimation

Abstract

The cone penetration test (CPT) is used to characterize tailings and infer the state parameter (ψ) using methods calibrated on uniform specimens. This work explores whether subaerially deposited tailings resemble uniform specimens. We quantified layering in tube samples from three active gold tailings storage facilities and visually appraised a 3 m profile at a fourth. Pronounced thin layering was found in every sample and in the exposed profile; the thickest apparently uniform layer was ~15 cm, far below the >0.6 m thickness recommended for ψ estimates. The variability of particle size distribution within a single 25 cm sample frequently matched or exceeded that measured across two entire impoundments of different commodities. At the observed layer thicknesses the cone cannot develop an isolated-layer response, so several layers jointly govern its readings. Thin layering is therefore a clear departure from the uniformity assumption underpinning widely used CPT interpretation methods, and it also obstructs the very testing that would be needed to characterize the resulting error. We conclude that, for subaerially deposited tailings, the confidence currently placed in CPT-based ψ estimates exceeds what the evidence supports. We argue for more space for the characterization of specimens that preserve in situ layering.