Erosion Test Database Reassessment with Application to Engineered Soils

Identifying and predicting soil erodibility characteristics is a complex process critical for the effective risk management of facilities and structures susceptible to erosion such as riverbanks, bridges, levees, and dams. This information is necessary to assess whether systems are at risk of surface erosion due to flowing water. For U.S. Army Corps of Engineers (USACE) infrastructure, this information is relevant for the prediction of overtopping breach of levees and dams. Additionally, erodibility measurements of soil are used to predict scour of dam spillways, coastal shoreline movement, and riverbank migration. Reliable and economical ways to determine these erosion properties are necessary for quantitative risk assessments to appropriately manage the safety of these structures and ensure their reliable operation.
Erosion Database compiled by Texas A&M University as part of the National Cooperative Highway Research Program was assessed in this investigation. The assessment was performed to explore the possibility of reducing uncertainty in erodibility estimates by investigating a subset of the data that was reported as compacted or engineered fill, which is relevant to breach and scour prediction for levees and dams.
Much of the uncertainty was found to due to inherent variability in soil characteristics, since the standard correlations used to predict erosion properties are derived from a database of tests mainly conducted on natural riverine sediments (e.g., intended to estimate bridge scour). These correlations may not be relevant for assessing much of the USACE built infrastructure, which are comprised of engineered fills compacted in place as opposed to naturally sedimented soil deposits.
While the database assessment concluded that subdividing the database into compacted soils and engineered fills did not appear to reduce the uncertainty significantly, this may be due to the reduced number of datapoints, recommendations for further research and development to improve predictions are presented. Further research to explicitly explore the effects of compaction and aging in the field on soil erodibility, consideration of the liquidity index as a predictor of erosion as it is linked to both the void ratio and the strength, in addition to the plasticity index, and the variation between methods among others, to address research consistent with USACE practices and interests.