H. MONTAZERI - Department of Civil engineering, Sharif University of Technology, Tehran, Iran
A. ABRISHAMCHI - Department of Civil Engineering, Sharif University of Technology, Tehran, Iran

Risk Analysis is a principal part of risk management in design and operation of hydro systems. For the first time in ICOLD congress in 2000, risk and risk analysis were discussed as pivotal topics in dam engineering; however, their application in aviation, nuclear, and marine industries goes back to the middle of the past century and in dam engineering to the early 1980s. A conventional method of evaluating failure probability is the use of probability theory in event tree. Due to the complexity and uncertainty of the failure causing events in dams, assigning probabilities to these events is a matter of engineering judgment. In order to make better use of engineering judgments in combination with engineering science, many attempts have been made in the field of dam risk analysis. A flexibleapproach used to delineate uncertainties in many fields including risk analysis is the fuzzy logic. In addition to its benefits in dealing with uncertainties, fuzzy logic makes possible the use of ambiguous, imprecise, qualitative and linguistic information as good as quantitative variables. Also human, social and environmental consequences are quantifiable and comparison of results with acceptable risks becomes easier and more meaningful. Also, the analyst can evaluate both the input data and the risk of an event directly, using linguistic uncertainties common for stating the risk. This paper, after enumerating difficulties and deficits of conventional methods of dam risk analysis due to floods, describes the fuzzy approach to this analysis and finally presents a simple event tree and fuzzy system for risk analysis of 1000 years flood in Karkhe dam.

Dam risk analysis, Uncertainties, Probability theory, Event tree, Fuzzy logic