R.J. Nathan, P.I. Hill, and H. Griffith
Recently released ANCOLD and IEAust flood guidelines include provision for the estimation of the Probable Maximum Flood (PMF) as well as the Probable Maximum Precipitation Design Flood (PMP DF). This paper examines the theoretical justification for derivation of these two different types of floods, and discusses how they may be used to characterise the hydrologic risk relevant to dam spillways. Recent experience has indicated that there is some confusion in the industry about the different uses of these estimates, and thus one objective of the paper is to clarify the different concepts involved and to provide an illustration of the differences between flood estimates for the two methods. Examples are provided to illustrate how the different estimates may be derived, and the practical implications for risk analysis are discussed.
— OR —
Now showing 1-12 of 26 2965:
M.A. Foster, R. Fell, R. Davidson, C.F. Wan
The probability of failure of embankment dams by internal erosion and piping can be estimated using historic performance, and event tree methods. Event tree methods are preferred for all except preliminary assessments, because they can better model the characteristics of the dam. This paper provides guidance on how to estimate the conditional probabilities within the event tree based on an understanding of the process involved, the historic performance of dams, and experience in recent risk analyses. This includes methods for representing the reservoir water level; assessing the likelihood that piping initiates; assessing the likelihood that erosion will be controlled by the filters or transition zones; and assessing the likelihood of development of a pipe and breach. The paper will be useful for those carrying out risk analyses and will also have relevance to those who are assessing piping of dams using traditional methods.
M.B.Barker and B.A. Vivian
Tumut Pond and Island Bend Dams are owned and operated by the Snowy Mountains Hydro Electric Authority. These dams, which are gated, have recently had significant electrical supply and control system upgrades. Subsequent reliability analyses performed for the gates provided unexpected results which highlighted issues concerning common mode failures and common cause failures associated with the mechanical systems. A further unexpected outcome of the analyses was the minor affect of human error and response to the emergency operating conditions of the gates in the event of electrical supply failure due to the over-riding mechanical system failures. This outcome was of benefit to the owners who had some concern that centralization of operation and consequent reduction in operating personnel would have an adverse effect on the reliability of the gates. The operation of the automatic control system is an ongoing issue for Island Bend where hunting of the gate operation is yet to be resolved. The preparation of the fault trees, development of failure probabilities and outcomes of the analyses are discussed in the paper which highlights some of the difficulties in design and operation of spillway gates, particularly where human response time is limited and automatic control is essential.
The Forth River dams consisting of Cethana (110m), Devils Gate (84m) and Paloona (43m), were constructed between 1964 – 1971. The Population at Risk (PAR) downstream of this cascade system is significant in the event of hypothetical dam failure.
By 1990 a Generalised Method had been fully developed for estimating extreme rainfalls for South East Australia. Using these extreme rainfall estimates, flood estimates were updated for all dams owned by Hydro Tasmania. These estimates indicated that the spillway capacity of the three Forth River dams no longer complied with current practice.
The risk position of these Forth River dams did not comply with the ANCOLD risk based criteria, indicating that some level of upgrade should be considered to reduce the risk associated with flooding. Given the risk position, considerable priority was placed on resolving this issue.
Due to physical constraints within the Cethana Dam site area, it was difficult to upgrade to a “Standards Based” level of upgrade without very high expenditure and imposing additional risk arising from major dam modifications. Instead the ALARP (As Low As Reasonably Practicable) principle was adopted to determine an appropriate level of upgrade, which did not preclude upgrading to a higher standard, should this be necessary at some time in the future.
The spillway upgrade for the three Forth River dams was approved in 1999 and detailed design has commenced with completion of construction planned for 2003. This paper will include discussions on the decision making process, communicating complex dam safety issues to senior management and some interesting details of the design.
Lelio Mejia, Murray Gillon, Jim Walker, Tom Newson
This paper describes the criteria for developing seismic loads for the safety evaluation of dams of two New Zealand owners. The criteria were constrained to satisfy the requirements of the NZSOLD Dam Safety Guidelines and to be consistent with international practice in countries with levels of seismicity and socio-economic development similar to New Zealand. In selecting the criteria, dam seismic load standards from several countries were surveyed and summarized. The selected criteria follow a standards-based approach to the seismic safety evaluation of dams. Guidelines for the use of deterministic and probabilistic procedures to develop seismic loads were formulated as a function of the Potential Impact Classification of a dam. In addition to the traditional deterministic definition of evaluation earthquakes, the selected criteria allow the use of a probabilistic definition in cases where the deterministic definition yields very low probability evaluation events.
Allan Crichton, Jon Williams, Anthony Ford
Wivenhoe Dam was constructed in the early 1980’s and is the largest source of water to the southeast Queensland region. The dam also provides significant flood mitigation benefits to the large communities in the Brisbane valley including the cities of Ipswich and Brisbane. Changes in the methods of determining the probable maximum flood, which is the design flood for the dam, have meant that the dam was not capable of passing the significantly larger design flood event. The feasibility study undertaken to assess the most appropriate method of upgrading the flood passing capacity identified more than 240 options. These options were short-listed and the capital cost and consequences costs determined for each of the short listed options.
This paper describes the process used to identify the options to upgrade the flood passing capacity and the method used to assess the consequences costs, which are primarily the costs of flood damage resulting from each of the options. The consequences costs for each option are the costs associated with changing the flow conditions in the river downstream of the dam. For example the option to upgrade the dam to pass the design flood is a benefit to the community however if this benefit is achieved by installing large gates or a fuse plug that operates frequently the scheme may increase the costs to the community.
The preferred option initiates at the lowest probability of occurrence of all those analysed — average return period of 10,000 years. There will be opportunities during the Environmental Impact Assessment process to test acceptability of this initiation level — a more frequent occurrence would be a lower capital cost solution. The EIA process may require a solution with a higher initiating level. Informal talks with the regulator have indicated a preference for the less frequent initiation level.