Back Analysis of a Rapid Drawdown Failure at Sparmos Dam Using UAV-Based 3D Failure Geometries

This study investigates the use of high-resolution 3D geometric data collected by unmanned aerial vehicles (UAVs) for estimating material properties in dam stability assessments following failure events. The analysis focuses on the Sparmos Dam in Greece, which collapsed in 2016 due to under seepage and at a different location due to rapid drawdown. UAVs provided timely, high-quality 3D data that allowed for the estimation of failure geometry. Using this data, 2D and 3D rapid drawdown analyses were conducted, applying the Duncan three-stage method to assess stability under rapid drawdown conditions. The comparison between the critical sliding blocks from the simulations and the failure surface captured by UAVs indicated that material properties such as undrained shear strength and overconsolidation ratio could be constrained effectively. This study demonstrates that UAV-collected data can provide valuable insights into failure mechanisms and serve as a rapid, reliable tool for post-disaster response, offering an understanding of dam stability before even sampling and laboratory testing results can be available.