Steven E Pells, Philip J N Pells, William L. Peirson; Kurt Douglas and Robin Fell
The method of Annandale (1995) is widely used by Australian practitioners for the assessment of erosion in unlined spillways. This method is based on comparison to various case studies, where the geology at each site is characterised using the Kirsten index (a rock mass index previously developed to assess the rippability of rock), and the hydraulic conditions are characterised using the unit stream power dissipation. In this paper, the historical development of this comparative design technique is traced and is critically reviewed against the original geotechnical and hydraulic data, and against a new, independent, dataset gained from unlined spillways in fractured rock in Australia, South Africa and the USA. It is shown that, while erosion can be usefully correlated against rock-mass indices and hydraulic indices, this ‘comparative’ design technique has been promoted beyond its reach – the data do not support the inference of an erosion ‘threshold’ as presented by Annandale (1995). It is argued that this type of analysis should be used only as an initial ‘first indication of erosion potential’, as originally proposed by van Schalkwyk (1994b).
Keywords: scour; erosion; spillways.
T. I. Mote, M.L. So, N. Vitharana, and M. Taylor
This paper explores the sensitivity of selection of earthquake design magnitude to liquefaction triggering in Australia for ground motions typically used for dams. The low seismicity of Australia creates a situation where liquefaction triggering is marginal at design hazard levels and this low level of seismic hazard makes the liquefaction trigger analysis very sensitive to the derivation of the seismic inputs. A methodology is presented that couples the probability of liquefaction triggering with the distribution of earthquake contribution to the hazard from the magnitude-distance deaggregation. The results show that for the “typical” soil profile and input ground motions approximately equivalent to the maximum design earthquake for Australia, the probability of liquefaction triggering varies significantly with the design magnitude selected. Using the maximum credible earthquake or mean magnitude may provide significantly different liquefaction triggering implications. Combining the probability of liquefaction triggering with the contribution of varying magnitudes to calculate liquefaction probability is a useful method to understanding the sensitivity of liquefaction to design magnitude.
Keywords: Liquefaction Assessment, Design Magnitude, Probability of Liquefaction, Magnitude-distance deaggregation, Australia
R. Nathan, P. Jordan, M. Scorah, S. Lang, G. Kuczera, M. Schaefer, E. Weinmann
This paper describes the development and application of two largely independent methods to estimate the annual exceedance probability (AEP) of Probable Maximum Precipitation (PMP). One method is based on the Stochastic Storm Transposition (SST) approach, which combines the “arrival” and “transposition” probabilities of an extreme storm using the total probability theorem. The second method – termed “Stochastic Storm Regression”(SSR) – combines frequency curves of point rainfalls with regression estimates of areal rainfalls; the regression relationship is derived using local and transposed storms, and the final exceedance probabilities are derived using the total probability theorem. The methods are used to derive at-site estimates for two large catchments (with areas of 3550 km2 and 15280 km2) located in inland southern Australia. In addition, the SST approach is used to derive regional estimates for standardised catchments within the Inland GSAM region. Careful attention is given to the uncertainty and sensitivity of the estimates to underlying assumptions, and the results are compared to existing AR&R recommendations.
Keywords: Annual exceedance probability, Probable Maximum Precipitation.
Vicki-Ann Dimas, Wayne Peck, Gary Gibson and Russell Cuthbertson
Globally, reservoir triggered seismicity (RTS) is a phenomenon sometimes observed in newly constructed large dams worldwide, for over 50 years now. Over 95 sites have been identified to have caused RTS by the infilling of water reservoirs upon completion of their constructions worldwide. In Australia, there are seven confirmed sites with observed RTS phenomenon that are summarized by temporal and spatial means.
With almost 40 years of seismic monitoring, primarily within eastern Australia, several of Australia’s largest dams have monitored and recorded many RTS events. At present, twelve dams are 100 metres and above in height as possible candidates, with seven of these actually causing RTS and a disputed possible eighth dam.
Important factors of RTS are reservoir characteristics (depth of the water column and reservoir volume), geological and tectonic features (how active nearby faults are and how close to the next cycle of stress release they are temporally) and ground water pore pressure (decrease in pore volume under compaction of weight of reservoir and diffusion of reservoir water through porous rock beneath). RTS is an adjustment process often delayed for several years after infilling of reservoir before eventually subsiding within 10 to 30 years, when seismic activity then returns to its prior state of stress.
Generally there are two type of RTS events, either a major fault near the reservoir most likely leading to an earthquake exceeding magnitude 5.0 to 6.0, or more commonly, a series of small shallow earthquakes.
Seismic monitoring of all dams (except for Ord River) are presented with spatial and temporal series of maps and cross sections, showing the largest earthquake, build-up and decay of RTS events.
Keywords: Seismic monitoring, reservoir triggered seismicity (RTS), earthquake cycle
Maree Dalakis, Dr Saman de Silva, Siraj Perera and Dr Gamini Adikari
This paper describes the results of a statistical and qualitative analysis on historical dam safety incidents in Victoria, the first study of its kind conducted in the State. The study investigates trends arising from qualitative dam safety incident data collected by the Department of Environment, Land, Water and Planning since the year 1996. The reported incidents are categorised based on their severity and statistical trends are identified in relation to the types of incidents common to regulated and unregulated dams, as well as common responses to incidents, including their post-incident operation. The geographical distribution of incidents across the State is also analysed to determine the effects of seismicity on dam safety incident rates. Furthermore, the unique Victorian conditions of sustained drought and subsequent flooding and their impact on incident rates are investigated through the combined analysis of geographical incident distribution and streamflow data. The incident data is further assessed according to the frequency of visual inspection and reporting of the structures in order to gauge the relative influence of these practices, and dam regulation in general, on mitigating incident risk in dams. An understanding of dam safety incident trends and the impact of inspection and reporting practices is increasingly important given the increasing expectation for dam owners to properly operate and maintain their assets with minimal resources and finances.
Keywords: dam, safety, incident, historical, failure.
Peyman Bozorgmehr, Sarah McComber, David Harrigan, Erik F R Bollaert
Boondooma Dam is a concrete-faced rockfill dam with an unlined, uncontrolled spillway chute. The Acceptable Flood Capacity of Boondooma Dam is 1:60,000 AEP (equal to the Dam Crest Flood (DCF) and has a maximum inflow of 14,330 m3/s.
Significant rainfall events during 2010/11 and 2013 subjected the spillway to moderate discharges over the crest which caused significant scour to the spillway chute.
Following these events, a 3D physical hydraulic model was constructed at a 1:80 scale to investigate repair options. Originally the spillway chute was modelled using a mobile bed set up which showed that that future scour could occur. However, the model could not determine the rate and characteristics of this damage.
In order to determine how future scour may occur, the 3D model was modified using laser survey mapping of the spillway chute after each flood event. Using milled aluminium and concrete capping the model was able to accurately portray the damage profile sustained by the spillway in the 2010/11 and 2013 flood events.
Transient pressure, static pressure, water elevation, velocity and jet measurements of the model were used in a Comprehensive Scour Model to help inform how damage to the chute may progress in future flood events.
Keywords: Boondooma Dam, flood damage, 3D physical hydraulic modelling, comprehensive scour assessment