Modelling pumped-storage effects on thermal structure, ice cover and water quality of lakes and reservoirs in a changing climate
New renewable electricity is nowadays often generated by photovoltaics and wind. Yet, their intermittent nature calls for energy storage, which is today still provided to ~95% by pumped-storage (PS) hydropower plants. However, PS is known to affect abiotic and biotic characteristics of the two connected water bodies. Thus, a two-dimensional laterally-averaged hydrodynamic and water quality model was set up to assess the impacts on water quality and temperature in a first step. Then, this analysis was extended to evaluate the additional effect of climate change, which also modifies abiotic and biotic characteristics. For this purpose, 150-years long synthetic time series of meteorological conditions for current (1998-2012) and future climate (2078-2092) were generated with a weather generator. To assess the robustness of projected impacts on the ice cover of the upper reservoir (Sihlsee) three one-dimensional, vertical hydrodynamic models were additionally set up.
To attribute effects to either the PS flows or to water withdrawal from the hypolimnion, two reference scenarios were defined: one with deep-water withdrawal (NoPS) and another one with surface outflow (QNat). While the hypolimnetic temperature differs by <1 °C comparing present PS and NoPS, the temperature differences can reach up to ~10 °C for the comparison with QNat. Thus, the effects at Sihlsee are mainly shaped by the deep-water withdrawal. The hypolimnetic warming will be further amplified by climate change with an increase, e.g. for September, of < 0.6 °C for QNat compared to ~2.5 °C for the extended PS operation.
The major hypolimnetic temperature dissimilarities between QNat and all other PS scenarios further result in different dynamics of stratification, oxygen depletion and nutrient release from sediments. Climate change, in case of QNat, further strengthens and prolongs summer stratification, which consequently leads to declining dissolved oxygen and increasing phosphate concentrations in the hypolimnion. This effect can partly be counteracted by deep-water withdrawal and PS operation.
While the comparison of present with both reference scenarios yields no major differences for the ice-covered period, the ice thickness is projected to decline by ~30% along with an earlier ice-off by ~1 month for the extended PS scenario. Climate change generally results in an additional thinning and shortening of the ice-covered period. This trend is supported by the results of the one-dimensional models. For the extended PS operation, future climate could even result in a complete disappearance of the ice cover at Sihlsee, which is not supported by all models, as the two-dimensional model shows different impacts for segments closer to the PS intake/outlet.
Overall, the results highlight the importance of the PS intake/outlet position, and of a clear definition of the reference state for the environmental impact assessment. This seems to be crucial for existing reservoirs, as the choice of the reference state could result in a different assessment of the ecological effects and subsequently requirements and costs for compensation measures. Moreover, the findings show that future climate should be considered for the planning of new and recommissioning of existing PS hydropower plants, which will be in place for many decades.
EPFL_TH9211.pdf
openaccess
86.23 MB
Adobe PDF
505cd80e783a78daddafa5c214e344f7