Using the NorSand Constitutive Model as a Tool in Assessing the Static and Dynamic Stability of Tailings Slopes Containing Contractive Material- Presenter & WINNER of the YOUNG PROFESSIONAL FORUM AWARD -KATY O’BRIEN

Download Publication (PDF)

SANCOLD 2021

By K O’Brien¹, T O’Brien² and B Robinson³

1.& 2. ARQ Geotech (Pty) Ltd 3. BRT

The ICOLD Tailings Dam Safety Bulletin 2020 Draft, advocates that if tailings material is classified as contractive, then post-peak/residual strengths must be considered in stability analyses. Should the target Factor of Safety (generally 1.1) not be achieved for this condition, either re-design is required or a rigorous and thorough assessment to evaluate potential for strain softening and liquefaction under various triggering mechanisms must be undertaken. The avenue for rigorous triggering analysis afforded by the ICOLD Bulletin, is explored in this article. A case study based on an existing Tailings Storage Facility (TSF) is presented and discussed, with the aim of demonstrating that despite being contractive, the stress state of tailings under various potential triggering events does not reach critical state or result in liquefaction. The advanced modelling techniques, including 2- and 3-dimensional finite element modelling, along with the use of the critical state constitutive NorSand material model, demonstrates and validates the patent stability of the facility. The approach provides an engineering method to rationally analyse tailings slopes containing contractive material. The findings of the analyses are also corroborated by real life TSF behaviour.

1.      Background

Until the 1950’s, at least in South Africa, tailings dams were created without any theoretical analysis or quantified measure of stability. Thereafter, from the late 1950’s, simplistic analytical models provided a “mostly adequate” measure or simulation of stability, with a margin for error (the factor of safety) providing a degree of perceived comfort and confidence in the actual practice.

Persistent failures led to an increase in the desired factor of safety (from 1,3 to 1,5) towards the end of the 20^th century. Nevertheless, failures continued with unacceptable frequency, prompting that the input parameters to the prevailing simplistic analytical model be revised to consider both peak and residual strengths, the latter being mandatory wherever a saturated contractive material occurs. This is the current routine approach advocated in the latest ICOLD (2020) bulletin and guidelines.

In many parts of the world this residual strength assumption is used, indicating in particular that upstream dam development should not be pursued in light of the FoS frequently not complying with requirements. In practice, particularly in South Africa where most tailings dams are upstream structures, accepting such an approach would be significantly and detrimentally consequential to the mining industry, the economy and hence SA society at large.