A large African dam modeled in EPFL laboratory
The Kariba Dam on Zambezi River
The dam creating Lake Kariba and its spillway have been modeled in EPFL’s Hydraulic Constructions Laboratory. The school has been commissioned to study the erosion caused by the high-velocity jets issuing from the spillway gates during flood. A few years from now, if no action is taken , the stability of the dam might be endangered.
The great Lake Kariba, on the border between Zambia and Zimbabwe, is contained by an imposing 128-meter-high arch dam, built between 1955 and 1959. The lake is twice the volume of Lake Geneva, and drains a basin equivalent to 16 times the surface area of Switzerland.
The nature of the climate of this tropical zone obliges the owner of the dam, the Zambezi River Authority, to regularly open the enormous spillway gates during floods. This operation, however, has consequences: in the bed of the Zambezi river, a huge scour hole has been eroded by high-velocity jets leaving the six spillway openings – each 9 meters by 9 – with a rate of 9 million liters per second. “This scour hole is now 70 to 80 meters deeper than the river bed”, explains Anton Schleiss, Director of EPFL’s Hydraulic Constructions Laboratory.
World-renowned for its competence in the study of the erosion of rock formations by high-velocity jets, the laboratory was commissioned by the Zambezi River Authority to assess the gravity of the situation. “However, no sensor would be able to resist the violence of the water flowing into this river”, explains Michael Pfister, scientific employee at the LCH. “This is why it was necessary to proceed with creating a model of the dam and the river below it.”
The physical model was designed and developed in about three months by teams from EPFL and some external providers. It’s 65 times smaller than the actual dam. “Mathematical similarity laws enable us to extrapolate the measurements that we take on our physical model, to determine what happens under real conditions”, explains Anton Schleiss. And this, “even though certain parameters such as the viscosity of the water don’t change with respect to the scale”, adds Michael Pfister. “This is the essence of the art of modeling: to choose the right scale in order to get usable results, while keeping the hydraulic model to a reasonable size.”
The systematic measurements, taken using pressure sensors installed in the bottom of the model, were began last week. The model dam is powered by the laboratory pumps, working in a closed circuit, that enable the circulation of more than 300 liters per second.
The scientists will now try to modify certain parameters, in order to identify practical solutions and mitigation measures which should avoid the problems that the owners of the dam will certainly face if the erosion continues at the same rate. “It will probably mean enlarging the scour hole in downstream direction, using underwater mining methods, to allow a more efficient dissipation of these enormous amount of energy”, believes Anton Schleiss. However, only the results of the measurements and simulations performed on the model after its transformation will allow us to evaluate whether this approach is appropriate or not.