Sediment Flushout from Pond of River Diversion Barrages by Gate Operation

Construction of river diversion barrages produces a shallow reservoir, called the pond, which is used more often for flow balancing between the inflows of the river and the outflows of the off-taking canal. However, deposition of sediment in the pond due to the relatively low velocities reduces the pond capacity. The current study investigates the effectiveness of gate operation and variations of other parameters in flushing out these sediment mounds, or shoals, from the barrage pond. Data from laboratory experiments on a scaled model of a prototype barrage is used to train and test different Artificial Neural Network (ANN) models of the system. The models map the relationship between flushing efficiency of a sediment shoal from the upstream of a barrage, and the parameters river discharge, barrage pond depth and area of gate opening, position of the sediment shoal with respect to the barrage. The ANN models are then used to study the effect of different parameters on the sediment flushing efficiency. Apart from river discharge and net area of gate opening, upstream pond depth is also found to have a significant effect on flushing efficiency, with a general trend of decrease in efficiency with increase in pond depth becoming apparent. Efficacy of different gate opening pattern is also tested, with the 'inverted arch' gate opening pattern proving to be the most efficient when compared with 'arch' and 'uniform' gate opening.