P Maisano, M Taylor , M Barker and A Parsons
South Para Dam, completed in 1958, is located on the South Para River, 38 km north of Adelaide. The embankment is 45 m high and comprises compacted crushed phyllite with rockfill toes. The 13 m high rock fill toes are protected with three-stage filters but the remaining 32 m of embankment height has no downstream filter protection.
The South Australian Water Corporation (SA Water), the owner and operator of the dam, is considering modifications to the dam, to augment its flood mitigation role. The proposed works, while not affecting the full supply level, involve a modification to the spillway crest and raising of the embankment crest to accommodate increased flood levels. SA Water therefore commissioned a dam safety review to assess the need for any piping or overtopping protection that may be required. This was followed by concept designs to ensure that flood mitigation work is compatible with any required dam safety upgrade work.
The results of a detailed dam failure risk analysis using event trees showed that the Societal Risk for the existing dam needed to be reduced, and that the proposed spillway modifications increased the Societal Risk due to the increased risk of piping failure with higher flood levels.
The risk analysis showed that eliminating the overtopping modes of failure by raising the dam crest is not sufficient in itself to achieve the required reduction in risk. The provision of filter protection to reduce the risk of piping failure is required, but it was shown that it is not necessary to provide full height filters as the provision of filters only above full supply level would be sufficient to achieve the required reduction in risk.
The recommended upgrade works, in addition to the proposed spillway modification for flood mitigation purposes, consist of filter protection and a weighting fill above the top berm (4.4 m below FSL) to facilitate connection to a possible full height filter in the future, and a parapet wall to provide overtopping protection.The resulting cost saving compared with the installation of full height filters is in excess of $2 Million.
The Ross River Dam was first commissioned in 1974 and raised in 1976. The 8200 m long
embankment was not fitted with chimney filters and has suffered extensive desiccation cracking since it was raised. A significant component of the dam upgrade is the retrofitting of filter zones to ensure the embankment meets current dam safety guidelines.
This paper describes the process of investigation of the existing desiccation cracks and the use of Hole Erosion Tests (HET) and No Erosion Filter (NEF) tests to validate the design of the retrofitted filter. A significant challenge in the design is to provide a cost effective solution given the 7500 m length of embankment requiring treatment. Assessment of flow rates within cracks and expected piping erosion along the cracks was used to assess the required drainage capacity. This assessment of expected flow capacity allowed the deletion of the coarse filter in the design reducing the filter requirement from a triple filter to a single fine filter. Results of this assessment were incorporated into the Risk Assessment based design validation process.
Changes to the Regulatory and legal environment have resulted in an increased focus on the
importance of proficient management of dams. Operation and maintenance manuals are now a
Regulatory requirement in Tasmania for all but very low hazard dams and are also required to ensure that dams are managed efficiently and safely. To meet these requirements Hydro Tasmania has developed the ‘Smart’ operations and maintenance manual.
Hydro Tasmania has a large portfolio of dams and as a result requires a large number of operations and maintenance manuals. This would result in an overwhelming array of information that is subject to evolving change if the traditional approach to the manual was adopted. To overcome this burden, a controlled electronic manual was developed to enable:
• Critical operation and maintenance information to be collated with minimal effort;
• Electronic hyperlinks to key existing operation and maintenance documents, reference
materials, and portals into operational data bases; and
• A means of updating and controlling information that is subject to change.
This paper will discuss how Hydro Tasmania developed its user-friendly operation and maintenance manuals in an innovative, unique and controlled manner to ensure prudent management of dams and to comply with Regulatory change.
The Ross River Dam, designed in the early seventies, does not meet current dam safety criteria for overtopping and piping within the embankment or the foundation. The dam comprises a 40m long concrete overflow spillway flanked by a central core rockfill embankment of 130 m on the right bank and 170m on the left bank with a 7620 m long left bank earth fill embankment, which has no internal filter zones for piping protection. The embankment was extensively assessed and treated forfoundation deficiencies in 1982, and further assessed in 20002002 for appropriate upgrade options.
This paper describes the process of validation of the detailed design using Risk Based Design Criteria. This process included data mining for historical performance and original design intention,
comparison of the original design against current and historical investigations and assessment of the upgrades using the large volume of data available from previous work. A design team comprising specialist hydrologists, hydrogeologists, geologists, geotechnical and dams engineers worked within a risk assessment framework at all stages of the design to ensure the design was validated using the design Validation Model. This process incorporated assessment of crest level based on flood risk and wave overtopping, review of 2D and 3D seepage models to assess piping and foundation erosion potential, assessment of fissured soils within the embankment foundation for structural stability and evaluation of spillway model testing for potential spillway failure modes.
David S. Bowles, Loren R. Anderson, Terry F. Glover, Sanjay S. Chauhan, Ronn S. Rose
A risk assessment was performed for the Sacramento District of the U.S. Army Corps of Engineers to explore the justification for imposing an operating restriction on Lake Success to reduce the probability and consequences of an Earthquake-induced dam failure. The potential for both a sudden overtopping failure and a delayed “seepage erosion through cracks” failure were considered.
The risk assessment focused on the seismic performance of the dam, the potential life loss and economic consequences of Earthquake-induced dam failure, and the estimated residual risk and degree of risk-based justification for the Existing operating regime, a range of Potential Operating Restrictions, and an Indicative Improved Warning and Evacuation System. Risk assessment inputswere supported by seismic deformation analyses under various Earthquake loadings and pool elevations, dam break-inundation modelling, and reservoir simulation.
Evaluations against tolerable risk guidelines from the USBR, ANCOLD, and the UK HSE, together with insights into the relationship between pool elevation and dam failure risk, provided important inputs for the decision to implement an operating restriction.
Ian Cordery, Peter S. Cloke
Scientists advocate more hydrological monitoring but in most regions publicly funded monitoring is in
steady decline. The lack of measured data at dam sites means there are many designs for new dams and remedial work that are insufficiently supported by factual information. Unfortunately data –free modelling exercises will usually produce favourable results – favourable to the modeller’s purposes, but not necessarily favourable to the determination of physical reality or truth. In these days of the popularity of modelling it is common to find decisions being made based on model studies for which little or no local data were available for model calibration or verification. How can the ‘large dam’ fraternity encourage (ensure) more data use? Causes of lack of data are many. For example governments fund data collection but others need the data, and data collection is a long-term activity that produces few benefits in the short term. Some years ago it was shown that hydrological data collection and archiving provided benefits to the community of at least nine times the costs of the data.
The real costs of comprehensive data collection are not large but examples will be given of the huge
costs, mainly due to the need to allow for uncertainty, that result from unavailability of data. Those
who understand this problem need to explain it to their communities, politicians and CEOs in a clear,
unmistakably persuasive manner, and to demand an increase in data collection. If we do not, no one