James Stuart, Michael Hughes
Several recent rain events in Australia have resulted in impoundment flood levels where there was a surprising variability between the Annual Exceedance Probability (AEP) of the flood level and that of the rainfall. The issue was highlighted during the Queensland Flood Commission of Inquiry (QFCI, 2011) by the Queensland Dam Safety Regulator who suggested there may be a problem with design hydrology after a dam safety event that saw impoundment levels of around 1:9000 AEP with a 1:200 AEP catchment rainfall at North Pine Dam, north of Brisbane in 2011. Wide disparities have occurred at Wivenhoe Dam west of Brisbane, at Callide Dam, west of Gladstone and at other locations.
This paper examines the Generalised Short Duration Method (GSDM) (BoM, 2003) and the Revised Generalised Tropical Storm Method (GTSMR) (BoM, 2003) typically used for dam flood capacity assessments in an attempt to explain the variability outlined above and whether it is, in part, exacerbated by the methods themselves.
It finds that processes of generalising rainfall depth, intensity, temporal and spatial characteristics are working together with adopted hydrological methods to contribute to such variability, that in the worst case could lead to PMF levels in dams with much less rainfall than the associated PMP would infer.
Moreover, two key assumptions; that of AEP neutrality (AEP of rainfall is equal to that of the flood) and frequency of PMP based on catchment area, which are the foundations stones of our understanding of flood frequency for large structures, are found to be untested or simply interim advice. This leads to the conclusion that the likelihood of floods in the range 2000 year AEP to PMF may continue to show surprising variability, potentially of an order of magnitude or more, compared to the rainfall AEP.
There is a need for a review of these methods and potentially provision of interim guidance as these methods are currently being used in dam upgrade programs throughout Australia and are also the basis for emergency planning. The identification of these issues concerns current methods and are independent to any discussion on climate change.Prior to commencing, it is worth defining two terms that re-occur throughout the document:
Annual Exceedance Probability (AEP): The probability that a given rainfall total accumulated over a given duration will be exceeded in any one year. AEP Neutrality is the theory that assumes the probability of the rainfall can be transferred to the resulting flood.
Average Variability Method (AVM): Technique for estimating design temporal pattern of average variability to ensure AEP Neutrality in transition from PMP to PMP design flood
Andrew Balme, Dan Forster, Tim Logan
The MW7.8 Kaikōura earthquake on 14 November 2016, ruptured over 20 faults during the initial shaking,which lasted nearly two minutes. A complex series of fault ruptures propagated northeast for nearly 180 km from the initial rupture location. Instrumentation from dams across New Zealand shows that whilst most dams did not suffer physical damage, piezometric responses were measured in dams and their foundations. Earthquake related changes in seepage regimes are not unusual and depend on the characteristics of the ground motions,and site specific characteristics that influence how a dam and its foundation respond to ground motions. The ability to measure a piezometric response in a dam or foundation is heavily influenced by the instrumentation network and method of monitoring. Data collected during events such as the Kaikōura earthquake provides valuable information for both characterising performance of a dam during the event, and assisting future analysis such as failure mode assessments. Careful consideration must be given to the scope of installed instrumentation and the frequency of monitoring in order to provide these benefits,and the robustness of the system to ensure it adequately survives the event.
Mark Stephen Rynhoud, David Johns and Len Murray
The Hamata tailings storage facility at the Hidden Valley mine is being constructed in a remote, high rainfall, tropical environment in a mountainous region of Papua New Guinea. Implementation of the design hasrequired adapting the design in response to various challenges encountered on the site during the ongoing construction period, based on observations by the designers and site monitoring data which is continuously collected and compared against design assumptions. This paper describes some of the design and construction modifications which have been implemented since construction of the tailings facility started and provides a case history of some of the challenges facing designers and construction crews when mining in remote, tropical conditions.
Although the total tailings dam failure frequency peaked in 1960s through 1980s, the failure rate of significant tailings dams has not dropped. The significant tailings dam failures the mining industry experienced in the recent history include: Merriespruit, South Africa, 1994; Los Frails, Spain, 1998; Kolontár, Hungry, 2010; Mount Polley, Canada, 2014; and Samarco, Brazil, 2015. The dam failures may be due to inadequate design, poor construction and inappropriate operations.This paper discusses the lessons learned and some recommendations and good practices to reduce the tailings dam failure risks. It addresses existing issues and provides some recommendations in risk based design, water management-integrity of facilities and water balance modelling, loading rates, tailings farming, adequate governance and roles and responsibilities of designers and nominated engineer.
Dr Andy Hughes
Tailings dams continue to undergo failures at an unacceptable rate compared to water storage dams, including failures at operations owned by high profile mining companies.Tailings dams have often a different form and method of construction than water storage dams in that tailings dams continue to be raised over time as part of the mine operations and rise to considerable heights. These failures are often the result of a combination of design, construction and operations actions that are controlled by humans and must be better coordinated and managed in the future. The consequence of failure can be widespread flows of tailings and water over the landscape and water courses. This can have extreme consequences in terms of life loss, environmental damage, social license to operate, company value, and mining industry sustainability. Therefore,it is necessary that the mining industry strive for zero failures of tailings facilities. Any additional technology and information that enables an owner of a tailings dam to be more certain of its condition and thereby reduce the risk of failure is of tremendous value to reliable tailings and mine water management.The Willowstick method uses low voltage, low amperage, and alternating electrical current to directly energise the groundwater by way of electrodes placed in wells or in contact with seepage or leaks. This approach has been successfully used to identify water flow paths through, under and around tailings dam in plan and elevation.The Willowstick technology provides additional information to supplement the geological, geotechnical and hydrological, evaluations analyses and designs, and to further improve tailings dam safety by more robust designs if necessary. This paper, using several tailings dam case studies, illustrates the procedure, findings, and the benefits of the Willowstick methodology. The findings of many Willowstick surveys range from tailings dams where the methodology has confirmed the design evaluations, to tailings dams where new groundwater and leakage flow paths were identified. In the latter case, the dam designers were able to update the designs, based on the new information,to mitigate the identified risks and to improve the overall safety of the tailings dams in accordance with the goal of zero failure.