The Port of Ngqura, situated in the Algoa Bay of South Africa, was commissioned in 2009 together with a sediment bypass system that is meant to intercept sediment being transported naturally eastwards towards the entrance of the port. The sea state in the Algoa Bay is dominated by waves generated in the Southern Ocean and flow from the Agulhas Current as it flows westwards along the southern coastline of South Africa. This sea state results in waves with an average significant wave height of more than 2 m over all seasons of the year. The sediment bypass system got blocked by rock fragments and stones migrating into the sediment trap created to accommodate eduction pumps sucking the fluidised sediment onshore for pumping downstream, to the right of the eastern breakwater. This resulted in the need for regular dredging in order to keep the entrance channel into the port open. The resulting sediment transport that necessitated the dredging operation was studied numerically by using the Delft3D software code. Delft3D Flow with its morphology module was coupled with Delft3D SWAN (Simulating WAves Nearshore) in stationary mode where data from National Centre for Environmental Prediction (NCEP) averaged over a 3 hour period was used as input for wave and wind data. Two nested grids were used to compute the wave propagations using SWAN where the larger grid took input of significant wave height, peak wave period and wave direction from NCEP, and it had a grid resolution of about 1000 m. The smaller inner grid (which had a resolution of about 500 m) got its boundary inputs from the calculated solution of the larger grid. All the wave conditions for SWAN were implemented with a directional spreading of 25 degrees with the JONSWAP shape. Thin dams were used to model the breakwaters of ports in the model and small islands. The influence of the Agulhas current was approximated by a current with a magnitude of 0.2 m/s and a direction of 250 degrees. Boundary conditions input into the Delft3D Flow model were water level computed using an in-house code that took current, wind and water level as input to calculate the water level at the right boundary node. The water level solution from Delft3D-Flow was used as input to the SWAN models. The bed and suspended sediment transport terms were computed over a period of 6 months, and compared to discern dominant processes responsible for the migrating stones, rock fragments and sediment filling up the bypass system sediment trap.
Papers presented at the 13th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Portoroz, Slovenia on 17-19 July 2017 .