A hybrid approach to evaluate sedimentation in large dams

Abstract

When completed, the Renaissance Dam on the Blue Nile River will be the largest hydropower dam in Africa. The consequent reduction of sediment fluxes to the lower Blue Nile by the Grand Ethiopian Renaissance Dam (GERD) is expected to impact agricultural production, river channel morphology, and downstream reservoirs’ management. Due to its vicinity, Roseires Dam is expected to be the most affected. This study assesses sedimentation in the GERD and its impact on Roseires sedimentation. A novel approach was used, combining flow velocities from a 2D hydrodynamic model (Delft3D) with empirical trap efficiency (TE) formulas by Churchill and Tan. The operational rules of both dams were considered in the TE calculation. Results were verified using Roseires bathymetric surveys. The Blue Nile sediment load was estimated using historical discharge and sediment concentration data (1981–2022) and a data-driven model (ANN) to fill in missing sediment records. Churchill and Tan formulas showed different responses to reservoir capacity and inflow variations, with long-term TE estimates of 92 % (Tan) and 97 % (Churchill). GERD's annual storage loss was calculated at 0.28 % (189 Mm³/year). Roseires's annual storage loss was estimated at 0.26 % before the GERD and is projected to drop to 0.01 % after GERD completion.