A Probabilistic Modeling Approach of Internal Erosion for Earthen Embankment Dams
« All Publications

Earthen embankment dams often fail due to internal erosion.

These failures can be analyzed in a deterministic manner using physical processes, coupling hydraulics and soil erosion (e.g., WinDAM, DLBreach, EMBREA, AREBA) to help understand and mitigate risks. A key model input parameter is the soil erodibility, which can vary over orders of magnitude for different dams and spatially within the same dam. Other model input parameters, such as soil critical shear stress, headwater elevation, internal erosion conduit elevation and initial conduit size also have ranges of possible values. One approach to assess the effects of the variability of model inputs is to develop a simulation based upon best estimates, and then systematically vary the inputs as part of a sensitivity analysis (e.g., Preston et al. 2017). However, interpretation of these results can be difficult and, as the practice moves toward a risk-informed decision making (RIDM), it is preferable to develop probabilistic approaches that can provide statistical distributions of results.

This current work implements WinDAM in a Monte-Carlo approach to enable probabilistic simulations to be performed considering multiple random inputs. This was achieved by developing Python scripts that modify inputs to create thousands of WinDAM input files with key input parameters selected from defined random distributions. Each of these inputs were assigned probability distributions based on available data, additional analyses, and engineering judgment. The Python scripts were then used to collate, post-process, and plot the results, including time-series traces (e.g., breach outflow rate) of thousands of simulations and statistical summaries of key outputs including peak outflow rate and time to critical outflow rates. These outputs can then be used as inputs into a quantitative risk analysis (QRA) to enable the dam owners to evaluate dam safety and consequences and develop and quantify the likelihood of success of mitigation strategies.

The current paper focuses on internal erosion as the potential failure mode (PFM), but the method can also be used to assess PFMs related to seismic-induced deformations leading to overtopping and/or cracking.

Publication Summary

  • Geosyntec Authors: Al Preston, Jai Panthail, Lucas de Melo, Glenn Rix
  • All Authors: Al Preston, Jai Panthail, Lucas de Melo, Glenn Rix
  • Title: United States Society on Dams (USSD) 2022 Annual Conference
  • Event or Publication: Event
  • Practice Areas: Dams; Dam Safety
  • Citation: United States Society on Dams (USSD) 2022 Annual Conference and Exhibition at the Town and Country Resort in San Diego, California on April 11 through 14, 2022.
  • Date: Tuesday, April 12 at 4:05 p.m. PDT
  • Location: San Diego, California
  • Publication Type: Platform Presentation