A Bayesian method for dam failure risk analysis using causal loop diagrams and word frequency analysis

• Published in: Natural hazards (Dordrecht) Vol. 119; no. 3; pp. 2159 - 2177
• Main Authors: Wang, Fang, Li, Hongen, Sheng, Jinbao, Yuan, Li, Pan, Yuxuan, Zhao, Jianguo
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01-12-2023; Springer Nature B.V
• Subjects: Analysis; Bayesian analysis; Bayesian theory; Civil Engineering; Dam failure; Dam safety; Dam stability; Dams; Decision making; Earth and Environmental Science; Earth dams; Earth Sciences; Environmental Management; Failure analysis; Frequency analysis; Geophysics/Geodesy; Geotechnical Engineering & Applied Earth Sciences; History; Hydrogeology; Machine learning; Mathematical models; Modelling; Natural Hazards; Original Paper; Overtopping; Probability theory; Risk analysis; Risk assessment; Risk factors; Seepage; Statistical analysis; Structural stability; China; Dam failure statistics; Word frequency analysis; Bayesian network; Nonlinear causal analysis; Earthen dam risk
• ISSN: 0921-030X; 1573-0840
• DOI: 10.1007/s11069-023-06196-3

Earthen dams are exposed to complex environments where their safety is often affected by multiple uncertain risks. A Bayesian network (BN) is often used to analyze the dam failure risk, which is an effective tool for this issue as its excellent ability in representing uncertainty and reasoning. The validity of the BN model is strongly dependent on the quality of the sample data, making convincing modeling rationale a challenge. There has been a lack of systematic analysis of the dam failure data of China, resulting in limited exploration of the potential associations between risk factors. In this paper, we established a comprehensive database containing various dam failure cases in China. Herein, historical dam failure statistics are used to develop BN models for risk analysis of earthen dams in China. In order to unleash the value of the historical data, we established a Bayesian network through the Causal Loop Diagrams (CLD) based on the nonlinear causal analysis. We determined the conditional probabilities using Word Frequency Analysis (WFA). By comparing with the Bayesian Learning results, the modeling method of BN proposed in our study has apparent advantages. According to the BN model established in this paper, the probabilities of dam failure due to seepage damage, overtopping, and structural instability are estimated to be 22.1%, 58.1%, and 7.9%, respectively. In addition, we presented a demonstration of the inference process for the dam failure path, which will offer valuable insights to dam safety practitioners during their decision-making process.