Improving governance will not be sufficient to avoid dam failures

Abstract

The last two decades have seen five large tailings dam failures, which developed suddenly and statically with damages of US$ billions and many deaths. These are company-threatening events, far outside societal tolerance, with investors questioning the situation. While media attribute these failures to mining companies, the underlying cause is a failure in engineering education: engineers of record, and their reviewers, lacked an adequate understanding of soil behaviour. The initiative to improve tailings stewardship by the International Council on Mining & Metals, with its focus on process and governance, will not achieve its aims unless this shortfall in understanding of soil behaviour is addressed. Critical state theory quantifies how and why void ratio controls soil behaviour, and was necessary to understand the Fundoa, Cadia and Brumadinho liquefactions (the rapid drained to undrained transition in particular); this critical state framework must become a ‘core competence’ for tailings dam engineers of record. Little additional cost will arise from doing this, with the largest change in practice being adoption of finite-element analysis for stability assessment. The biggest challenge is education, with engineers needing to familiarise themselves with the largely untaught critical state theory (there are public-domain resources for this, as given in the Appendix to this paper).