Calculating sediment trapping efficiency for Reservoirs in series

Abstract

Trapping Efficiency (TE) is defined as the proportion of inflowing sediment that is accumulated in a reservoir. Accurately predicting the trapping efficiency (TE) of sediment in reservoirs is critical to estimate their useful life and inform catchment models for sediment budget estimation. A daily formulation of the Churchill sediment trapping efficiency equation previously implemented in the eWater Source modelling framework, enables the user to account for the accumulation of sediment and hence the subsequent sediment yield exiting a reservoir. Where reservoirs are in series, the particle size of sediment passing through an upstream storage is generally finer than the ‘local’ sediment. Therefore, the application of the Churchill equation to each reservoir in isolation, not taking into account the different sources of sediment, may result in an under prediction of the sediment yield for a basin. Churchill (1948) proposed a method to account for sediment that had already passed through an upstream reservoir. The original daily Churchill equation was applied to all storages and compared to a modified Churchill equation, applied to upstream reservoirs for three scenarios across the Great Barrier Reef (GBR), Namely: 1. The Pioneer River in the Mackay Whitsunday region which contains three small weirs (<10,000 ML); 2. The Burnett River in the Southern GBR which contains three large upstream storages that flow into a reservoir in the lower reaches of the catchment and; 3. The proposed Hell’s Gate Dam to be constructed on the Burdekin River upstream of the Burdekin Falls Dam. The results showed negligible differences for the small storages and indicated that the application of TE models for such storages is probably redundant. Larger differences in sediment loads were calculated with the application of both TE models for the larger storages which demonstrate its usefulness to incorporate into future modelling. The scenarios also demonstrate the usefulness of models to quickly forecast the influence of proposed new reservoirs (or the modification of existing reservoirs) on sediment loads. This paper will describe the modified Churchill approach and provide estimates of sediment yield with and without the modification and recommend where the curves are appropriate to be used in the GBR. The approach will also provide a method to improve estimates of sediment yield from basins where reservoirs are in series.