A practical probabilistic approach for simulating life loss in an urban area associated with a dam-break flood

Abstract

Dam safety simulation is crucial to design and improve emergency plans to mitigate the flood disaster risk associated with break events, particularly in the areas located downstream. Applying deterministic estimation of the peak flow has some limitations for deciding in terms of uncertainty. In this study, we proposed an integrated approach based on Monte Carlo Simulation (MCS), HEC-RAS 2D, and HEC-LifeSim models to simulate dam break to deal with the flood risk with probability rather than analyzing a few dam break scenarios. This approach was applied to simulate a dam break event and its impact on Life Loss (LL) according to 12 alternatives in Sidi Yahya Zaer town, in Morocco. The results showed that for dam-break peak flows associated with Exceedance probabilities of 1%, 5%, 10%, 50%, 90%, and 95% the mean of the simulated LL are ranging from 2 to 334 people at night time and from 2 to 173 people at day time. They exponentially depend on dam-break peak flows. Furthermore, the comparison of the LL results with those of the historic events showed that a fatality of 125 people that was occurred by the Buffalo Creek dam break in the USA (1972) is possible in SYZ town with a probability of 9.2%. Globally, the proposed methodology is valuable to evaluate the dam-break consequences associated with a probability. Also, it can be reproduced to evaluate the vulnerability of the future urbanized areas not only to flood disasters but to other natural and industrial disasters.