Skip to main content
ARS Home » Research » Publications at this Location » Publication #181573

Title: EARTHEN EMBANKMENT BREACH RESEARCH: PHYSICAL AND SIMPLIFIED NUMERICAL MODELING OF BREACH WIDENING

Author
item Hunt, Sherry
item Hanson, Gregory
item Temple, Darrel

Submitted to: State Dam Safety Officials Association Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 8/1/2005
Publication Date: 10/1/2005
Citation: Hunt, S., Hanson, G.J., Temple, D.M. 2005. Earthen embankment breach research: Physical and simplified numerical modeling of breach widening. In: Dam Safety 2005. Proceedings of the Association State Dam Safety Officials, September 23-27, 2005, Orlando, Florida (New Orleans, LA). 2005 CDROM.

Interpretive Summary: With many earthen embankment dams reaching the end of their planned service life, the increased possibility of one of these structures failing is generating concern within the dam safety community. The ages of these embankments along with characteristic changes within the watershed can trigger embankment failures through flood events that cause overtopping and/or piping in the embankments. Understanding how a breach develops over time can improve prediction of flood magnitudes and timing that can be used to develop warning systems and emergency action plans. The research discussed in this document provides a summary of four large-scale embankment failure tests conducted at the USDA-ARS Hydraulic Engineering Research Unit. Soil properties including water content at compaction and soil type have been shown to significantly influence how a soil erodes and the rate at which it erodes. These properties directly affect breach timing, formation, and geometry, and researchers hope to gain a better understanding of how these variables influence and characterize breach embankment failure processes through the development of a research tool known as SIMBA. SIMBA is a research tool developed to evaluate embankment failure processes. The primary erosion process of interest during testing was breach widening. The measured results from the four physical tests are compared to results obtained from SIMBA. SIMBA does a reasonable job of predicting the breach outflow and the breach width over time, with slight differences attributed to the simplified approach taken with the development of SIMBA. More attention is expected to be given to these differences as the computational routines are refined. This research may lead to the development of future warning systems for dam failures and may impact breach modeling and dam safety.

Technical Abstract: With many earthen embankment dams reaching the end of their planned service life, the increased possibility of one of these structures failing is generating concern within the dam safety community. Over time, sediment pools fill, and land use and topography change due to increasing development. These transformations in the environment can trigger embankment failure through flood events that cause overtopping and/or piping in the embankments. Understanding how a breach develops over time can improve prediction of flood magnitudes and timing that can be used to develop warning systems and emergency action plans. Research using large-scale physical models of earthen embankments is on-going at the USDA-ARS Hydraulic Engineering Research Unit. Soil properties including water content at compaction and compaction energy have been shown to significantly influence how a soil erodes and the rate at which it erodes. These properties directly affect breach timing, formation, and geometry. Soil properties that show promise for characterizing breach embankment failure processes and used for performance prediction during the breach widening stage are discussed in this document. Large-scale embankment failure tests were conducted to examine the mechanics of breach widening during the reservoir draw down stage of a breach. Data from these laboratory experiments provide a basis for developing numerical models for describing the breach geometry as a breach widens over time. The objective of this paper is to evaluate the time rate of breach widening for four large-scale earthen embankment tests and compare the physical results from these tests to simulated results obtained from a simplified numerical model, SIMBA. SIMBA is a research tool developed to evaluate embankment failure processes. For a simplified approach, SIMBA does a reasonable job predicting the breach outflow and the breach width as it develops over time. Although SIMBA predicts the breach outflow fairly well for the early stages of the tests, SIMBA does not account for changes in tailwater that may result in differences between the predicted and measured values. Additionally, widening is observed as an episodic process while SIMBA simulates this process in a continuous fashion which may account for the differences in the predicted and measured breach outflow and breach width. Since SIMBA is a research tool, more attention is expected to be given to these issues as the computational routines are refined.