H193-0006
Failure Probability Analysis of Earthen Levees Affected by Bioerosion

Wednesday, 16 December 2020
Poster
Matteo Balistrocchi1, Giovanni Moretti1, Roberto Ranzi2 and Stefano Orlandini1, (1)University of Modena and Reggio Emilia, Modena, Italy, (2)University of Brescia, Brescia, Italy
Abstract:
Bioerosion caused by burrowing mammals is an emergent threat to earthen levee stability. In fact, many levee failures were found to be associated with the existence of dens. Dens facilitate the triggering of failure mechanisms caused by seepage and inner erosion. Bearing in mind the amount of data and the computational burden required by numerical analyses, probabilistic methods that are capable to incorporate the natural variability along with the epistemic uncertainty of floods are developed in this study. To this aim, a bivariate approach is needed to provide a comprehensive view of the failure mechanism, since the extent of seepage fronts depends on riverflow stages and their duration. The assessment of the probability of failure due to seepage based on copula functions is illustrated. In addition, mammal dens were incorporated, to assess the corresponding increase in failure probability. The developed methods are applied to a real world earthen levee located near to Ponte Bacchello on Secchia River, northern Italy, where a disastrous flood event occurred in 2014 due to a den located 2 m below the levee top. For this river a 42˗year time series of hourly flow discharges is available, making it possible to fit the copula function relating the peak discharge to the duration. The extents of saturation fronts assessed by using Darcy law for this elevation are shown in the attached figure (soil hydraulic conductivity of 1.88×10-6 m s-1 and porosity of 0.404). A linear combination of Clayton and Gumbel copulas was found to be suitable to fit the data. The method uses a limit state function to split the population into safety and failure regions, by comparing the maximum extent of the seepage front to the distance between the den end and the riverside levee talus (den entrance on the landside). The probability of failure due to seepage only, in a period of 100 years, is estimated to be negligible in the absence of dens, whereas it increases up to 58.3% (return period of 115 years) when the den-end-to-riverside-talus distance is 1.5 m. This event is highly probable in 100 years (failure probability of 99.2%, return period of 15 years) when the distance is 1.0 m. By assuming a soil hydraulic conductivity of 1.88×10-5 m s-1, the 100˗year failure probability is larger than 99.9% (return period of 2˗3 years), when the distance varies between 1.5 and 2.0 m.