Jiri Herza and John Phillips
The design of dams for mining projects requires processes and technology that are unfamiliar to many mine owners and managers. Dam designers rely on ANCOLD assessments of Consequence Category, commonly leading to a High rating for mining dams due to a combination of potential loss of life, impact on environment and damage to assets such as mine voids, process plants, workshops, offices, roads, railways etc.
From this High Consequence Category the relevant annual exceedance probabilities for design parameters and loading conditions such as earthquakes and floods are selected.
Mining companies have sophisticated methods available for assessing risk, yet for their assets they often adopt an order of magnitude lower security for earthquake and floods even though the consequences in terms of lives at risk and impact on project are similar.
The discrepancies in the design standards lead to situations where extreme dam loads are adopted to prevent damage and loss of life in assets that theoretically would have already collapsed under much lower loads.
One difference may be that some mining dams exist in an environment which is controlled by a single entity. Unlike other dams, failure of these mining dams would therefore impact only individuals and assets which fall under the responsibility of the same entity.
This paper discusses the discrepancies between the design of mining dams and the design of other mine infrastructure. The paper considers the impact of discrepancies on the overall risk to the mine and compares the degree of protection offered by a factor of safety and the influence of reliability of design input parameters, alternate load paths and design redundancy.
Keywords: Dams, tailings dams, mining, acceptable risk, factors of safety
Michael McKay and Francisco Lopez
Mt Bold Dam impounds the largest reservoir in South Australia. The dam wall comprises 19 concrete monoliths, 11 forming a central arch section and 8 forming gravity sections on the left and right abutments. The upstream face of the arch section is vertical, but the top portion overhangs on the reservoir side. The dam was originally constructed in the 1930s, and was raised by 4.3 m in the 1960s. In this upgrade the gravity abutments were raised using mass concrete blocks and the arch non-overflow crest was raised with hollow, reinforced concrete portals. On the spillway section a pier and gate system was installed on top of a hollow ogee section. The maximum height of the dam in its current configuration is 58 m.
GHD has been conducting a staged safety review of Mt Bold Dam since 2011. This included a detailed finite element nonlinear, time-history seismic analysis of the dam-foundation-reservoir system. The analysis was carried out using finite element techniques and included a detailed 3D model of all major components of the dam and different domains of the foundation rock. The nonlinearity of the model was included by explicitly incorporating contact elements at the dam-foundation interface, at the monolith contraction joints, and at some identified unbonded horizontal concrete lift joints within the dam wall. The seismic analysis was conducted for three different accelerograms corresponding to Maximum Design Earthquakes (MDEs) with 1 in 10,000 Annual Exceedance Probability (AEP).
This paper explains the purpose of the study, the adopted methodology and material properties, the results of the modelling phases, and the anticipated seismic behaviour and damage on the main components of the dam resulting from the MDEs. Finally, a conclusion is made in regards to whether or not Mt Bold Dam passes the adopted performance criteria for seismic loading.
Keywords: Arch, gravity, seismic, nonlinear, damage prediction.
Russell Cuerel, Richard Priman, Michel Raymond, Ian Hanks
Following significant flood events across Queensland over the last five years causing significant damage in South East Queensland, Bundaberg Burnett region, St. George in the south west and more recently in Central Queensland in the Callide Valley, there has been renewed interest in finding solutions to flooding issues.
Increasing the available flood storage within a catchment is a well-known method of improving flood mitigation outcomes for developed areas. In many basins/catchments, potential flood storage development options (new storages or augmentations to existing storages) can be identified by reviewing previous water supply investigations and flood studies and by scanning topographic mapping. From such site identification there will often be numerous combinations of possible flood storage development options to consider because of the number of tributaries which may contribute to major flood events.
This paper outlines a methodology to screen, within a relatively short timeframe and at relatively low cost, a large number of identified flood storage development options and combination development scenarios and shortlist for more detailed analysis. The screening process is heavily reliant on hydrologic assessments to rapidly short-list scenarios for assessment and then relies on traditional engineering and economic assessments to do the fine tuning of the analysis.
Keywords: flooding, damages, impacts, flood storage, flood mitigation, dams, benefit-cost ratio.
P C Blersch, W van Wyk , R Steenkamp
Construction of the partially completed Calueque Dam on the Cunene River in Angola was abandoned in 1976 due to the hostilities in Angola. In 1988 the dam was bombed, causing significant damage to the bridge deck, other structures and equipment. Work to complete and rehabilitate the dam commenced in late 2012 and included major earthworks, extensive concrete repairs and refurbishment and installation of mechanical equipment, including ten spillway radial gates and two outlet gates with lifting equipment, emergency gates and cranes, including electrical and control systems. A number of challenges were encountered in planning and executing the project but were overcome largely as a result of detailed historical project information having been retained well beyond the norm and through the involvement of a key member of the original project team in the current project.
Keywords: Dam rehabilitation, radial gates, zoned earthfill embankment
Simon Lang, Chriselyn Meneses, Kelly Maslin, Mark Arnold
It is now common practice for dam owners in Australia to take a risk based approach to managing the safety of their large dams. Some dam owners are also using risk based approaches to manage other significant assets. For example, Melbourne Water manage the safety of their retarding basins in a manner similar to their water supply dams.
Assessing the risks posed by retarding basins using methods developed for larger dams can raise challenges. For example, the Graham (1999) approach to estimating potential loss of life (PLL) is generally applied when estimating the consequences of dam failure. However, Graham (1999) may not be the most suitable model for estimating PLL downstream of structures with relatively low heights and storage volumes (e.g. retarding basins), given the characteristics of the case histories used to develop the method.
In this paper six potential methods for estimating PLL are tested on four retarding basins in Melbourne. The methods are Graham (1999), the new Reclamation Consequence Estimating Methodology (RCEM), the UK risk assessment for reservoir safety (RARS) method, a spreadsheet application of HEC-FIA 3.0, and empirical methods developed by Jonkman (2007) and Jonkman et al. (2009). Results from the methods are compared, and comment is made about which is most suitable.
Keywords: potential loss of life, dam safety, risk analysis, retarding basins.
Maz Mahzari and Chi-Fai Wan
Upgrading of an existing dam often faces challenges in both static and seismic safety assessment. The use of new hydrological and seismological data and improved design methods often mean more severe loading which outdates the original design and demands expensive upgrade works. Establishing the design criteria for checking the structural adequacy of an existing dam for multiple unusual load events occurring within a relatively short time frame presents another challenge.
A probabilistic approach is presented to rigorously address the effects of multiple load events while maintaining a consistent risk of failure for the structure. This is based on a probabilistic conditional combination where probability of each event is defined and used to develop a joint probability distribution. For instance if an earthquake occurs following a severe flood, the seismic hazard curve of the site can be used to adjust the seismic loading with shorter average recurrence interval to be used in conjunction with the pre-earthquake flood when assessing the structural adequacy of the dam. With this method of adjustment, the design can benefit from the choice of a reduced seismic design loading and hence a more cost effective design solution.
The proposed method is straightforward and can be effectively used in most engineering practices, including the design of hydraulic structures such as dams.
Keywords: Dams, Seismic Hazard, Post-earthquake, Risk analysis