David Dole and Brian Haisman
The $75 million remedial works at Hume Dam on the Murray River near Albury have been of national significance. The rehabilitation program associated with the structure itself and with its appurtenant works is now in the final steps of construction. The authors summarise this program with an emphasis not on technical details, but on decision processes. Equity in this dam is, in effect, held in equal parts by three State governments and by the Commonwealth government. At the same time, in response to the national water reform agenda, the governments have agreed upon new cost sharing arrangements that more nearly reflect the value of services to each government. The particular problems of decision-making within this evolving inter-State environment are discussed.
Lessons from experiments with application of risk analysis are discussed. Finally, the matter of adequacy of the structure for extreme floods is still under consideration. Hume Dam will presently pass the Design Flood developed in accord with Book VI (1999) of Australian Rainfall and Runoff, and the Dam Crest Flood has been estimated to have an annual exceedance probability of 1:110,000. _ Retrofitting a spillway to pass the estimated Probable Maximum Flood will double the cost of remedial works and is estimated at 10 times the cost of similar capacity built into original construction. The authors discuss the public policy elements of this pending decision.
— OR —
P.W. Heinrichs and R. Fell
Ben Boyd Dam, a 29 m high earthfill embankment built in 1978, has had an unusual history. In 1979, a number of seeps developed during first filling with water 5 m below FSL indicating unexpectedly high pressures. Investigations concluded the coarse filter permeability was very low due to excess fines. Remedial works in 1982 included a drainage filter beyond the toe and a new stability berm above. New piezometers were installed, including several in the blanket filters in the existing dam. These later indicated up to 10.5 m head in isolated areas within the filter. Pump out tests partially lowered the water level in the standpipes but in 1995 the water level rose by 4 m back to its previous high level. All this during a period of relatively low rainfall. Stability analyses were carried out and further investigations in 1999 concluded that apart from general leakage from the foundation abutment into the filters, the rise in pressures was due to leakage from a riser hole from one of the nearby foundation piezometers. A potential for piping along the piezometer tubes within the dam was also identified.
This situation was managed without resort to costly capital works, because it was concluded that the pressures from the vertical riser were not a potential failure mode, and potential piping failure would be adequately handled by the existing chimney drain, intersecting the piezometer tubes trench. Any potential piping failure would also give warning signs which increased frequency of monitoring (now in place) would pick up in time to allow lowering of the storage.
Mark Locke, Buddhima Indraratna, Phillip Cummins and Gamini Adikari
ABSTRACT: Australia has a large number of older embankment dams, which have been in service and performed adequately for over 50 years. However, current industry practice in embankment dam design predicts that the granular filters within these dams may not be adequate. This may require refurbishment of the dam by retro-fitting a new filter to ensure the continued safety of the structure. This paper outlines the potential problems with older embankment dam designs, and the reasons for constructing a new filter. Potential problems may include inadequate or non-existent filters, risk of failure due to earthquake, piping, or excessive foundation seepage. Design methods for granular filters are described briefly, concentrating on whether an existing filter is adequate, and the potential improvement by constructing a new filter. Construction issues for placing filters on existing dams are also discussed.
A new analytical method, developed to describe the time dependent erosion and filtration within embankment dams, is described briefly. The model predicts particle erosion, transport and retention based on fundamental fluid mechanics and geotechnical concepts. The application of this model to the design of filters for new and existing dams will be described. The predictions of such analytical modelling can give a designer a significantly clearer picture of the purpose of a granular filter, the extent of core erosion that can be expected, and the effect of retrofitting a new filter to an existing dam.
I. R. Forster
Lyell Dam is a concrete-faced rockfill dam, located on the Coxs River, near Lithgow, NSW. The dam forms part of the Coxs River Water Supply Scheme, which supplies water to Delta Electricity’s Wallerawang and Mount Piper Power Stations. In 1994, the spillway capacity of the dam was upgraded, and the storage augmented with the addition of two 40 m long by 3.5 m high inflatable rubber dams to the spillway crest. An automatic deflation system, controlled by a programmable logic controller, was installed to provide a staged bag deflation sequence during flooding, and hence minimise the downstream impact of rubber dam operation.
Although the rubber dams and control system initially operated as designed, more recently, two uncontrolled bag deflations have occurred, which have caused flooding downstream and loss of significant storage volumes. In the first incident, a spontaneous uncontrolled deflation of the rubber dams released about 1600 ML, before the bags re-inflated automatically. An investigation revealed that the incident was most likely the result of design deficiencies in the control system. Recommendations were made for improvements to the system.
During the most recent deflation, one of the rubber dams failed by spontaneous rupture, and approximately 6000 ML of water was released from the dam. The Dam Safety Emergency Plan was activated to ensure persons at risk downstream were notified of the impending flood wave. A post- failure inspection of the ruptured bag suggested that the likely cause of failure was a manufacturing defect, which allowed air to penetrate the layers of rubber forming the bag. The rupture most likely occurred when the resulting air pocket expanded on exposure to the sun.
The paper examines the two deflation incidents in detail, and analyses the emergency response to the second incident.
B. S. Sherman
Many large Australian dams currently lack selective withdrawal capabilities and release water mainly from deep within the hypolimnion. Deep-water releases coupled with the strong thermal stratification typical of Australian reservoirs results in discharge temperatures 10 °C or more colder than would normally be expected. Cold water pollution has impacted more than 1000 km of river habitat in Australia where it is known to impair spawning, feeding and survival of many native fishes.
This report reviews alternative approaches for the mitigation of cold water pollution below dams. The underlying theory and practical limitations of operation as well as field experience (including cost) with each of the methods are discussed. Two methods in particular, suface pumps and submerged curtains, appear to offer cost-effective alternatives to the expensive retrofitting of dams with multi-level outlet structures (estimated to cost $5-35m per dam for major dams in NSW). These methods are predicted to be capable of increasing discharge temperatures by 4-10 ° throughout the range of irrigation releases without any redirection of flows, i.e. hydropower releases can be maintained at present levels. This holds the promise of restoring more desirable temperatures over hundreds of kilometres of river.
Glen Hobbs and Danny Azavedo
Recent years have seen a growing awareness and understanding of the factors that contribute to the reliability of spillway gates and the incorporation of reliability data into overall dam risk studies.
The study of a number of spillway gate failures shows that no one component or incident leads to gate failure, but rather a combination of factors have resulted in gate failure. A rigorous reliability assessment should consider all factors, not only the equipment condition and performance but the complete system, from the receipt of data through to the actuation of the gates. It should take into account issues such as human factors, poor design, maintenance history and policy. Unfortunately one of the main hindrances to quantifying gate reliability is the lack of information on spillway gate equipment and system performance and failures.
This paper considers a number of gate failures, then looks at some of the tools of reliability assessment and the role of human factors in gate reliability.
The paper then discusses a recent study of four gated dams. For this study a systems approach was adopted and human factors were considered. The results compare favourably with other similar critical structures, and show that for these well designed and maintained structures human factors are the limiting criteria in multiple gate operations. The study also shows that the probability of opening all the spillway gates at a dam improves with time (2-4 hours) during the flood operation, and it is considered that time based reliability provides a more meaningful and useful assessment of overall spillway gate reliability.