Monitoring sediment transport by means of optic-acoustic system to optimize the operation of a hydropower plant

The Cheves hydroelectric power plant (HPP) is located in the Andes Mountains in Peru near Lima. The operation of the Statkraft asset is heavily influenced by sediment transport during the rainy season (February to June), both by bed load and suspended load. To avoid sediments from filling up the reservoir, sediment routing during rainy season is applied. However, during the routing, high water velocities through the gates are causing flab abrasion resulting in high maintenance costs. In addition, the water being transferred to the head race tunnel is of high sediment concentration and causes severe erosion of the turbine blades, resulting in low efficiency. This occurs despite desanders are operating continuously during rainy season.

To optimize the operation of the power plant, both during power production and during sediment routing, a sediment monitoring instrument was installed at four locations on the HP system. The positions were chosen to provide accurate information on sediment inflow to all major units of HP, such as the desilting units, reservoir, forebay, headrace tunnel, and turbine outlets. Both acoustic (e.g., 8 MHz, acoustic backscatter) and optical (e.g., optical backscatter) instruments will be installed to accurately estimate suspended sediment concentration (SSC). The AoBS, manufactured by Sequoia AS, provides both acoustic-optical measurements and a proprietary method for combining the measurements and determining the SSC in mg/l. To validate SSC, a pump sampler was installed at each location to sample once a day in dry season and twice a day in the wet season. The water samples were analyzed using both the typical filter method and the laser diffraction method.

The initial results (in the dry season) confirmed that the combination of optical and acoustic methods provided the most accurate results, although the Sequoia AS method appears to underestimate by 30-50%. Another method based on a weighted summation of both results is under development and promises better results for the existing data. Based on these results, four indicators have been developed: i) intensity of sediment inflow in the turbine, ii) coarseness of particles at all positions (optics vs acoustics scattering index), iii) sediment discharge, iv) desander performance (sediment input from the desanders). Further testing needs to be conducted in the wet season to validate the indicators as well as the method of combined acoustic-optical sediment data collection.