Impact of Flood Loading and Initial Conditions on Transient Seepage Analyses

Abstract

Levee seepage analysis for design often uses steady-state analyses, which produce conservative results. Transient seepage analyses are appealing because they can account for time-dependent flood loadings. Uncoupled transient seepage analyses were used in this investigation to examine their accuracy, the impact of initial conditions, and different boundary condition assumptions. Remote monitoring results from a Mississippi River levee embankment collected over multiple flood events were used in this investigation. A 2015 Mississippi River flood was used to quantify the impact of different initial pore water pressure distributions. Four different initial conditions were assumed; results of these analyses showed that initial conditions greater than hydrostatic resulted in greater progression of the phreatic surface. Uncoupled models were shown to reasonably match the conditions measured in the field. A survey of flood hydrographs from river gages across the United States was used to investigate flood boundary conditions. The results showed that durations of extreme events could be hundreds of days long. It was also shown that typically flooding occurs when embankment pore water pressures are at yearly maximums.