Assessing the effects of extreme rainfall patterns and their impact on dam safety: a case study on Indian dam failures

This study investigates the impacts of extreme rainfall variations on dam safety, focusing on six large Dam Failure (DF) events in India: Tigra, Khadakwasla, Pagara, Machu-2, Koyana, and Kaddem. Daily gridded rainfall data is obtained from the India Meteorological Department, and the Inverse Distance Weighted interpolation method is used to get location-specific daily rainfall data. The severity of extreme rainfall events on dam safety is highlighted by computing the average rainfall (AR) and accumulated rainfall (ACR) for 5-, 10-, and 15-day prior to the date of DF. Shockingly, the magnitude of 15-day ACR prior to DF exceeds 50% of the normal annual rainfall at most of the study locations. This unexpected situation may put tremendous pressure on the dams and eventually lead to their failure. Further, the Probable Maximum Precipitation (PMP) is computed at each dam location using the Annual Maximum Daily Precipitation (AMDP) time series across 121 years. Next, the Efficiency Factor (EF) is calculated to check the severity of rainfall prior to the DF. The annual EF values are computed, and the maximum EF value over 121 years indicates the maximum rainy day in that time horizon. The value of EF above 0.85 poses a threat to dams, and approaching 1.0 (almost equal to PMP) could result in DF. This study established a robust correlation between dam failures and heavy rainfall preceding them. Some dams, like Machu-2, Kaddem, and Pagara, experienced clear rainfall peaks on the day of the collapse, indicating heavy rainfall over 5 days (5-day ACR) as the primary cause. Others, such as Tigra and Khadakwasla exhibited continuous moderate rainfall (ACR) for 5 to 10 days is the principal cause of failure. These findings are of significant relevance to professionals in the field of dam engineering, offering a comprehensive understanding of how extreme rainfall events can impact dam failures and providing valuable insights into rainfall patterns and their implications for dam safety. Most importantly, the dam owners will be notified at least 5 days prior to the catastrophe (dam failure), which is sufficient to take suitable measures for safe reservoir operations.