Internal erosion and piping within embankment dams may initiate in cracks caused by differential settlement or desiccation, in cracks caused by hydraulic fracture and in very poorly compacted layers of soil. It generally cannot occur unless one of these defects is present because backwards erosion, the other mechanism for internal erosion, will not occur in embankments under normal gradients and will not occur in cohesive soils unless gradients are exceptionally high.
As a result it is very unlikely that it will be possible to detect initiation of erosion with piezometers, and the most likely successful method is seepage observation and monitoring. However the time from the first detection of increased seepage to breach of the dam may be very short-a matter of hours in some situations.
Thoughtfully positioned and read piezometers are more likely to be successful in identifying the critical gradients which may lead to the onset of backwards erosion in cohesionless soils in the foundation of dams.
Piezometers are more useful in establishing the pore pressures for use in analysis of stability, but in most cases where stability is marginal undrained strength analysis is required and the pore pressures and effective strengths alone are not sufficient to assess stability. In a number of cases differential settlements, and acceleration of settlements have proven valuable in detecting the on-set of instability and the conditions in which internal erosion and piping to initiate. Once these conditions are recognised more detailed survey monitoring and borehole inclinometers can be valuable in better defining the geometry of instability.
— OR —
Lawrie Schmitt and Angus Paton
As the owner of most of the large dams in South Australia the South Australian Water Corporation (SA Water) is responsible for the safety of these structures and their designed function of water supply and flood control. In order to meet these responsibilities SA Water monitors the performance of the structures using engineering deformation surveys and various forms of instrumentation. This paper outlines the instrumentation and survey monitoring undertaken at SA Water large dams and discusses the issues arising.
John Bosler and Francisco Lopez
The ANCOLD “Guidelines for the Design of Dams for Earthquake” were published in August 1998. The guidelines contain a brief outline of the performance requirements and recommend, in general terms, a method of analysis for intake towers.
Over the last three decades there has been considerable research on the seismic performance of intake towers as they move into their inelastic range. In the years following the publication of the ANCOLD guidelines, some of the findings from this research have been incorporated into revised design procedures issued by the US Army Corps of Engineers. These procedures, if embraced by ANCOLD and the local dam engineering community, are likely to have a significant impact on how the structural adequacy of existing towers under seismic loading are assessed.
Rocking behaviour in which the tower becomes unstable as a transient condition has long been recognised as acceptable under certain conditions. Attempts to prevent tower rocking by measures such as retrofitting tensioned ground anchors may, in some situations, be of limited value in improving the seismic performance of a tower and could result in an increase in bending moments in the tower stem. Guidance is now available on the amount of rocking behaviour that is tolerable.
For seismic events greater than the Operating Basis Earthquake most towers will start to exhibit inelastic behaviour. Specific guidance is also now available on the length of time during an earthquake that bending moments in excess of the elastic capacity can be tolerated, the amount by which these moments can exceed the nominal bending moment capacity and the vertical extent of the tower stem that can be stressed beyond its elastic limit.
The paper discusses the different approaches taken by ANCOLD and the Corps of Engineers. Key differences in outcomes are highlighted using a worked example for a typical reinforced concrete tower and the ANCOLD approach is found to be generally, but not always, more conservative. The paper concludes with recommendations for dealing with these differences.
For many years most emergency management agencies in Australia have used a framework called Prevention, Preparedness, Response and Recovery (PPRR). This approach has worked very well in the past and has been incorporated into the more recent framework of Emergency Risk Management.
While Emergency Management Agencies use practice sessions in the form of Desktop/Tabletop Exercises and Field Exercises as part of Preparedness (the 2nd P in PPRR) these activities can suffer from a lack of engagement with the community.
State Water Corporation, a dam owner in NSW, has installed warning systems to trigger plans written by the SES to warn affected residents of possible dam failure. Although the systems are maintained and tested regularly in a technical sense, the next logical step is to encourage the affected communities to understand their role in the event of evacuation.
A joint exercise involving the NSW State Emergency Service (SES), State Water Corporation and the community, was conducted in a town in the Namoi valley in 2005 and has provided an opportunity to explore this concept. State Water Corporation is now confident that not only will the technical side of the warning system work but that residents should be more aware of their role and that of the SES and State Water Corporation.
Other benefits from the exercise are: the opportunity for improving general flood awareness in the community; the SES identifying community representatives; fine tuning procedures between and within the SES and State Water Corporation; allaying fears within the community about what is required of them in a dam failure; and demonstrating the dam owner’s duty of care to affected residents.
Nerida Bartlett, David Scriven, Peter Richardson
The failure of a number of consecutive wet seasons has resulted in storage levels in Eungella Dam being at dangerously low levels such that supply could be exhausted by June 2007. Eungella Dam supplies bulk water to the Bowen Basin coal fields as well as the Collinsville power station and the Collinsville township.
The Collinsville township, power station and coal mine as well as the Newlands mines take water from the Bowen River Weir which is supplied from Eungella Dam some 95 kilometres upstream. Transmission losses of the order of 25 to 50% have been experienced for releases from Eungella Dam to Bowen River Weir.
The Eungella Dam catchment area is 142 square kilometres. Significant flows occur in the Bowen River downstream of Eungella Dam, the catchment area above Bowen River Weir being 4,520 square kilometres. The topography in the surrounding area (near Collinsville) is not suitable for dam construction.
The opportunity existed for the construction of an offstream storage adjacent to the Bowen River Weir so that the downstream flows could be captured reducing the demand on Eungella Dam thus making more water available for upstream users.
A 5,200 ML offstream storage, associated pump station and rising main was designed, constructed and filled within a period of 12 months.
Foundations at the site are highly permeable sands. Marginally suitable clay for a seal was in short supply as was suitable rock for slope protection. A fixed price budget had been set by the contributing customers.
This paper describes the hydrology, site conditions, design and construction of the project.
Janice H. Green and Jeanette Meighen
The Probable Maximum Precipitation (PMP) is defined as ‘the theoretical greatest depth of
precipitation that is physically possible over a particular catchment’. The PMP depths provided by
the Bureau of Meteorology are described as ‘operational estimates of the PMP’ as they represent the best estimate of the PMP depth that can be made, based on the relatively small number of large events that have been observed and our limited knowledge of the causative mechanisms of extreme rainfalls.
Nevertheless, the magnitudes of the PMP depths provided by the Bureau are often met with scepticism concerning their accuracy when compared to large rainfall events which have been observed within catchments and which are, typically, only 20% to 25% of the PMP estimates. The recent increases in the PMP depths, resulting from the revision of the Generalised Tropical Storm Method (GTSMR), have served only to entrench this cynicism.
However, analyses of the magnitudes of the storms in the databases adopted for deriving PMP depths show that these observed storms constituted up to 76% of the corresponding GTSMR PMP depths and up to 80% of the Generalised Southeast Australia Method PMPs for the storm location. Further, comparisons of the PMP depths to large storms observed in similar climatic regions around the world indicate that the PMPs are not outliers.
The results of these analyses are presented for a range of catchment locations and sizes and storm durations and demonstrate that the PMP estimates provided by the Bureau of Meteorology are reasonable and are not unduly large.