Experimental Investigations of Levee Breach Flows

Two individual but related problems involving levee breach flow are studied. The first one involves experimental and dynamic modeling of steady flow through a levee breach, and the second one involves experimental investigation of flood management by an engineered levee breach. Both studies are conducted in the Hydraulics Laboratory, the University of South Carolina. In the first problem, an idealized levee breach on a trapezoidal embankment was studied. Flow data including water depths, surface velocities, and flow discharges were collected from 28 experiments considering different cases of breach width and downstream water depth. Ultrasonic sensors were used to record the water depth in the channel and the breach. The water surface velocity was measured using a particle tracking velocity method. A one-dimensional theoretical model was developed to predict levee breach flow. This model is based on the conservation of mass and momentum in two control volumes. For known upstream flow discharge, water depth, and the downstream stage, the model predicts the breach discharge and the average flow depth along the breach. Two model coefficients were calibrated by using the experimental data, namely, the interfacial shear force coefficient between the two control volumes and the separation shear force coefficient in the channel downstream of the breach. The proposed model shows satisfactory performance against an additional set of experimental data. In the second problem, flood management using engineered levee breach was studied. Laboratory experiments were conducted to investigate the effects of the inflow hydrograph characteristics, breach opening, tailwater depth, and floodplain storage volume on flood attenuation. The inflow hydrograph was controlled by a computer-controlled valve allowing for various peak discharges, hydrograph shapes and flood durations in the channel. Two approaches were used to generate the hydrograph. In the first approach three different inflow hydrographs having the same volume of water, but different durations were released. In the second approach, three different inflow hydrographs that had different volumes of water, but different durations were released. Three different breach widths - 0.2, 0.4, and 0.8 m - were considered by instantaneously opening a gate. Three different areas of flood basin (9, 6 m2 , and open floodplain) were tested for each hydrograph and breach width. Both active and passive flood management scenarios were considered. In the active flood management scenario, the breach was created instantaneously as the flood wave approached breach location. In the passive management scenario, the floodgate was already open, and the floodplain had the same water depth as the main channel prior to the arrival of the flood wave. The engineered levee breach provided a reduction of flood stage both upstream and downstream of the breach by modifying the flood wave. Dimensional analysis has been done to express the reduction of flood stage as a function of the breach width, channel width, distance of the target location from the breach centerline, hydrograph characteristics, and area of the basin