Michael Müller
Reservoir sedimentation and the resulting storage losses impact reliability, efficiency, safety and thus the sustainability of the hydropower schemes. Beside traditional storage hydropower plants, also pumped-storage facilities are affected. Flexible turbine and pump operations between two reservoirs allow demand-depending electricity production and absorption. Thus, they play a dominant role in peak load energy production as well as grid regulation. The storage volumes are influenced by continuous in- and outflow cycles. Their impact on sediment settling processes has not been addressed in research yet. Inspired by the cyclic, bidirectional water exchange of pumped-storage plants, a novel method for sediment removal from reservoirs was investigated in the present study. After high sediment yield into the reservoir, fine sediment should remain in suspension in front of the water intake due to the pumped-storage induced turbulence. Thus, the settling process may be delayed and sediment be evacuated by the power or flushing tunnels. In prototype measurements and laboratory experiments the influence of pumped-storage operations on flow patterns as well as the settling rates and sediment balance in two interconnected reservoirs has been studied. The two approaches were completed by numerical simulations. In the lower reservoir of a pumped-storage hydropower plant in the Swiss Alps, flow velocities in front of the intake/outlet were measured by Acoustic Doppler Current Profilers (ADCP). The devices with independent energy supply were implemented on the reservoir bottom and sampled 2D flow velocity profiles over several weeks. The measurements showed only local influence of pumping (outflowing water) near the intake, whereas turbine operations (inflowing water) induce large eddy flow fields in the reservoir. Depending on lake topography, patterns with backflow can appear. A frequency analysis of the discharge and flow velocity series indicated a corresponding main period between the flow velocity profiles and the in- and outflow cycles of 1 day. The ANSYS-CFD simulated flow fields corresponded to the in situ flow patterns. The computed turbulent kinetic energy input due to turbine operation was some 25 times higher than the natural input by wind-forcing. In the upstream part of the power shaft of the same hydropower scheme, a turbidity probe was installed for monitoring reasons. Over a period of eight months, sediment concentration of the operated water was continuously measured. An autonomous remote data acquisition and transfer system may be helpful for real-time monitoring by the hydropower operators. The measurements showed seasonal change of sediment concentration in the power system. In winter, high reservoir levels and ice-cover reduced sediment content, whereas in spring, snowmelt and low reservoir levels increased sediment yield. Short-term variations of sediment concentration up to 16% correlated with the in- and outflow cycles, especially for low reservoir filling. During the sampling period, about 45’000 t of fine sediment were moved between the two lakes. However, the sediment balance due to pumped-storage operation remained equilibrated. Fine sediment was kept in suspension in front of the intake/outlet structure by the operation induced flow patterns. Pumped-storage facilities would only increase sedimentation in case of unequal sediment concentrations in the two reservoirs. The impact of magnitude and frequency of the in- and outflow cycles on flow field and fine sediment settling behavior was studied in laboratory experiments. Between two interconnected rectangular basins, sediment-laden water was pumped back and forth according to regular in- and outflow cycles. In both volumes, turbidity was continuously measured. In the main basin, Ultrasonic Velocity Profilers (UVP) allowed 2D flow velocity measurements and the analysis of the developing inflowing jet. In a numerical model, time for achieving steady state conditions in the reservoir was defined. This discharge dependent duration indicated the frequency of the initial in- and outflow cycles. A reference test in stagnant water conditions defined the sediment settling curve without operation, corresponding to a basis of comparison for the investigated test configurations. During five in- and outflow cycles of various magnitude and frequency, the evolution of suspended sediment ratio SSR was measured and compared to the reference case. Sediment influx and release (influx sediment rate ISR and evacuated sediment rate ESR) and thus the sediment balance SB of the system could be defined. In a first stage, 60% of the initially suspended sediment settled independently from magnitude of the in- and outflow cycles. High cycle frequency applied during this phase led to considerably higher SSR by the end of the test sequence. Low discharges increased particle suspension by 10 to 40% compared to conditions without operation, whereas high discharges increased concentration by 50 to 80%, depending on cycle frequency. With an intake/outlet structure close to the bottom or to the reservoir surface, 20% higher sediment concentrations could be achieved than in case of intermediate position. Sediment balance is only marginally impacted by in- and outflow cycles. In situ as well as laboratory investigations highlight the fact that pumped-storage operations influence the flow patterns in reservoirs, especially in the near intake area. The induced turbulence keeps fine sediment in suspension, leading to an extensive sediment exchange between the storage volumes. The sediment balance remains unaffected, as long as the sediment concentrations in the reservoirs are similar.
EPFL2012
