Analysis of human risks due to dam-break floods—part 1: a new model based on Bayesian networks

Abstract

Dam breaks have catastrophic consequences for human lives. This paper presents a new human risk analysis model (HURAM) using Bayesian networks for estimating human risks due to dam-break floods. A Bayesian network is constructed according to a logic structure of loss-of-life mechanisms. The nodes (parameters) and the arcs (inter-relationships) of the network are quantified with historical data, existing models and physical analyses. A dataset of 343 dam-failure cases with records of fatality is compiled for this purpose. Comparison between two existing models and the new model is made to test the new model. Finally, sensitivity analysis is conducted to identify the important parameters that lead to loss of life. The new model is able to take into account a large number of important parameters and their inter-relationships in a systematic structure; include the uncertainties of these parameters and their inter-relationships; incorporate information derived from physical analysis, empirical models and historical data; and update the predictions when information in specific cases is available. The application of this model to the study of human risks in a specific dam-break case is presented in a companion paper.