Ryan Singh, Jiri Herza, James Thorp, Michael Ashley
Performance-based risk-informed decision making is an underlying principle of the Global Industry
Standard on Tailings Management (GISTM). While owners make significant efforts to align with this
principle, commonly used risk assessment and management practices in the mining industry have largely been based on the HSE principles, which consider more frequent, lower consequence incidents.
As a result, the existing risk assessment frameworks do not provide the owners with a comprehensive understanding of the risk profiles of their tailings storage facilities (TSFs). Without the understanding of a facility’s risk profile, the owners cannot appreciate how changes to their facility, processes and operational activities may impact the risk profile. A large step-change in thinking is therefore required in risk assessment practices for the owner to align their TSF management with GISTM requirements.
Beyond risk assessments, the mining industry has other valuable concepts to manage the safety of their tailings management practices, such as Critical Controls, however, commonly used risk assessment and management practices do not incorporate these concepts.
This paper explores commonly used risk assessment practices and the concepts of Critical Controls. It proposes how these concepts can be linked, with Critical Controls being embedded in the risk assessment process. The outcomes of linking these concepts result in an estimation of the effectiveness of the Critical Controls and how they can be improved to demonstrably reduce the risk presented by a TSF. A case study has been included to demonstrate the benefits of linking risk assessment with Critical Controls and how owners can readily identify deficiencies and efficiently manage the risk profiles of their facilities.
— OR —
Mark Pearse, Mark Foster, Peter Hill, Sam Banzi, Muhammad Hameed, Benson Liu
Determining which risk control measures are required is one of the top issues for dam owners as they contend with limited resources generally and capex in particular. The key issue addressed in this paper is how a dam owner can both identify the control measures that they should implement and demonstrate that they are acting reasonably and responsibly. The Framework developed in this paper provides a practical and transparent way to address the relevant matters that are required to be considered under common law, work, health and safety (WHS) legislation and the NSW Dams Safety legislation for determining whether a risk control measure is reasonably practicable. It provides dam owners with a transparent and defensible way of both identifying the controls and demonstrating that they are acting in a reasonable and responsible manner.
Chris Nielsen, Irene Buckman
As individuals, we are concerned about how a risk affects us and the things we value
personally. We may be willing to live with a risk if it secures us certain benefits and if the
risk is kept low and clearly controlled. We are less tolerant of risks over which we have little
ANCOLD’s risk assessment guideline (2003) identifies an individual risk threshold as being
one where “the dam safety risk to an individual should be close to the average background
risk of the population”. This is a principle of equity, where “all individuals have
unconditional rights to certain levels of protection” (HSE, 2001). The definition of
population at risk applied to Queensland’s referable dams (DNRME, 2018), being
individuals within a residence or workplace and typically not participating in any risky
activities such as driving a vehicle or walking through flooded waters, provides further
justification of this right.
In practice addressing societal risk tolerances and duty of care considerations may result in
individual risks being substantially lower than the thresholds. This may not always be the
case and, irrespective, should not distort the purpose of the individual risk tolerance test;
the principle of equity that drives individual risk tolerability has foundations in our societal
values and is easily and widely understood as a core value. This should be succinctly
described when justifying expenditure on risky infrastructure such as dams.
This poster describes aspects to consider when selecting a threshold individual risk
tolerance. Subject to site-specific considerations of the particular age group of individuals
most at risk, the wider benefit of the dam to society and ALARP, a single threshold
individual risk tolerance of less than 10-5 per annum (or 1 in 100,000 years) would appear
The aspects described are elaborated in the revised Guidelines on Safety Standards for
Referable Dams, soon to be published on the Queensland Government website (RDMW,
Hench Wang, Edward Funnell, Albert Shen, Matt Scorah, Peter Hill
The use of simulation models to assess dam failure consequences has progressively advanced in Australia over the past few years. For example, it is now common for HEC-LifeSim to be used to estimate potential loss of life from the failure of large dams with large populations at risk downstream. Since its introduction to Australia, numerous presentations and papers have been provided by USACE and industry professionals that highlight the benefits of using HEC-LifeSim Version 1.0.1 for a range of different case studies.
This paper identifies some of the new features in the latest version of HEC-LifeSim that can improve the robustness and defensibility of the potential loss of life estimates for dambreak consequence assessments. The techniques that have been used to overcome these challenges are also discussed using some case studies.
The first case study demonstrates the sensitivity of the model performance and potential loss of life to changes in version and number of iterations used to simulate the life loss. This is done by comparing the differences in simulation run time and life loss between the previous and new versions of HEC-LifeSim for an example model. The second case study presents an example application of both versions of HEC-LifeSim to compare the results between one version and the other for a different dam and the final case study illustrates an improved method for interrogating the available outputs from HEC-LifeSim to provide the user with more information that otherwise could not be obtained from the default outputs.
Zara Bostock, Helena Sutherland
Ewen Maddock Dam is located approximately 12.0 km west of Caloundra, in the Sunshine Coast area of Southern Queensland. The dam is a homogeneous earthfill embankment dam 10.5 m high and 724 m long. The dam was originally built between 1973 and 1976 and later upgraded in 1982 to raise the ogee spillway crest by 2.44 m to the current Full Supply Level (FSL) of 25.38 m AHD.
Seqwater is undertaking a staged upgrade of Ewen Maddock Dam to address deficiencies identified during the Acceptable Flood Capacity (AFC) Review (GHD, 2010). The consequence category assigned to Ewen Maddock Dam is ‘Extreme’ with a downstream Population at Risk greater than 1000.
Stage 1 construction was completed in 2012 to manage the seepage underneath the dam to reduce the risk of piping and improve embankment stability. Stage 2A involved retrofitting a filter in the existing embankment and raising the dam 1.61 m to 30.11 m AHD using a reinforced concrete parapet wall. Stage 2B involves spillway upgrade works and was split from 2A due to approval constraints.
Stage 2A construction was completed in April 2021, navigating various project and dam safety challenges. This paper presents some practical ways dam safety and risk was managed on the ground from the perspective of both the designer and owner.
David Reid, Andy Fourie, Riccardo Fanni, Cristina Vulpe, Alexandra Halliday
Recent failures of a number of tailings storage facilities (TSFs) has highlighted the need for better
governance and operational management of these structures. One means to improve their safety is clearly better and more focussed monitoring. Significant efforts are underway in this area, with a number of technologies being deployed. In particular, the monitoring of deformations through a variety of means (direct, satellite inferred) is increasingly being applied. While deformation monitoring to warn against failure has a long history in geotechnical engineering, some aspects of the rapid triggering and resulting flow of some TSFs may not be amenable to deformation monitoring, in the sense that actionable warning of an impending failure is not assured.
To examine this issue, a series of numerical models of an idealised TSF are carried out. This idealised TSF is brought to failure by means of a rising phreatic surface – often referred to as the constant shear drained (CSD) stress path. Deformations of the outer slope and crest of the numerical model – i.e. those that could be monitored for a real TSF – are tracked and analyses for the models carried out. It is seen that under CSD loading distinct deformation patterns indicative of impending failure are not always clear. Rather, minimal deformations and indeed swelling of the crest is seen leading to failure. The importance of recognising the minimal pre-failure deformation patterns that may manifest with a rising phreatic surface is noted.