Uncertainty inevitable in dam design

Risk and Uncertainty in Dam Safety by CEA Technologies' Dam Safety Interest Group. Review by Dragutin Jovanovic, chief engineer for the Ministry of Water Resources, Ethiopia

The Chi-Chi earthquake in Taiwan, in September 1999, destroyed the right wing of the Shih Kang Dam, the breach resulted in severe water shortage in Tai-Chung.  Photo: Tim Little/BC Hydro.

The Chi-Chi earthquake in Taiwan, in September 1999, destroyed the right wing of the Shih Kang Dam, the breach resulted in severe water shortage in Tai-Chung. Photo: Tim Little/BC Hydro.

The Teton Dam in Idaho, USA, failed abruptly on 5 June 1976 with the eventual loss of 14 lives and an estimated US$1 billion cost.  Photo: US Bureau of Reclamation
Although the three basic methods of risk analysis proposed in this book were originally developed for mechanical and electrical engineering, and the aerospace, petrochemical and nuclear industries, they have only relatively recently begun to be applied to the design of dams. However, many years ago in the USA and other developed countries, mathematical calculations of dam breaches and, for dams with large storage capacity, breach simulation and assessment of the consequences on the downstream valley were performed on physical models in hydraulic laboratories.

For the engineer intensively engaged in design or construction of dams, it would probably be most useful to concentrate on the second and particularly third method of risk analysis described in the book, since the first method, Failure Modes and Effects Analysis, deals with individual component failure modes and is not so relevant to dam design.

The second method, Event Tree Analysis (ETA) can be applied when the sequence of events is known. It is used to identify possible outcomes and their probabilities following the initial event.

The basic question in this analysis is: what happens in the case of an earthquake of a certain magnitude or what happens in the case of a big flood wave? For example, if the initial event is a big flood wave, the subsequent events will be: gate closed but inoperable, loss of reservoir control, dam failure through overtopping.

The third method, Fault Tree Analysis (FTA) should be applied when the system is large and complex. It requires examining different initiating events such as earthquakes, flooding, wind, wave action and loss of off-site power.

This method is multidisciplinary and incorporates many aspects outside the traditional domain of engineering. It suggests that the design should include the following models: reservoir operation, formation of breach in dam, characteristics of flood caused by the breach, state and magnitude of losses in the downstream affected areas, including the size of inundation for different intensities of flood.

It also suggests establishing an early warning system to be operated by dam maintenance and operating engineers. An interesting, empirically based equation of the relationship between warning time and population at risk is described in the book. To ensure impartiality, it is desirable that independent analysis is undertaken by an expert well-acquainted with the design documents, operational and maintenance procedures and technical specifications.

Numerous destroyed dams testify that uncertainty in dam safety is a reality; risk always exists, even when the dam is designed by the most experienced engineer. Even after thorough investigation prior to construction, a small, undiscovered weakness in the fundament could cause complete destruction of the dam. This is exemplified in the book by two relatively recent case studies of arch dams.

Although Malpasset dam in France was designed by a top international designer, the dam collapsed instantaneously in December 1959 when its support moved due to undetected weakness in the ground and over 450 lives were lost.

The French government established a commission and after inspection of all documents and the field site, its report revealed that design and construction were performed in accordance with common technical rules and practice and that enough investigation work, including core drillings, was performed. It concluded that the cause was uncertainty and that it would not have been possible to discover the weaknesses in the fundament.

At the 262m high Vaiont dam in Italy, geological investigation had not revealed the potential danger of an underground layer in danger of sliding when submerged. In 1962 a landslide of about 240 million m3 of material slid into the reservoir. So much water was squeezed out of the reservoir that the dam overflowed and an enormous flood wave wiped out villages along the valley, taking 2600 lives.

In conclusion, this book would be very useful for establishing internationally recognised standards, and even laws, for dam design, including early warning systems.

Contact: CEA Technologies
Web: www.ceatech.ca



Click a keyword to see more stories on that topic, view related news, or find more related items.


You need to be logged in to make a comment. Don't have an account? Set one up right now in seconds!

© Faversham House Group Ltd 2005. edie news articles may be copied or forwarded for individual use only. No other reproduction or distribution is permitted without prior written consent.