Impact of climate change on water availability and its propagation through the Western U.S. power grid

Climate change is expected to affect the availability of water for electricity generation, yet the propagation of climate impacts across a large and diverse power grid remains unexplored. In this study, we evaluate how projected changes in water availability affect electricity generation at hydroelectric and thermal power plants and how the coincident impacts propagate locally and throughout the interconnected power grid of western United States. We also evaluate whether the prospect of climate-driven change could affect regional power dependencies. Hydrologic simulations derived from three Global Circulation Models (CCSM4, INMCM4, and GFDL-CM3), two radiative scenarios (RCP4.5 and RCP8.5) and the VIC hydrology model are used to force a large-scale, distributed water management model (MOSART-WM), which translates water availability into power generation constraints at hydropower plants and water-dependent thermoelectric plants. Power system dynamics are evaluated using the production cost model PLEXOS. We find that the interregional connections across the contemporary Western U.S. electricity infrastructure play an essential role in managing variations in regional generation due to hydrological variability. Projected WECC-scale changes in mean annual precipitation ranging from −3.8% to +17% are moderated to −6% to +4% in mean annual production cost changes. Climate change impacts on water availability in the Northwest drive future changes in other regions’ generation and in regional power flows. Northwest total generation influences interannual variability in other regions’ net generation, explaining about 40%, 50%, and 35% of the variability in Southwest, Rockies, and Southern California regions respectively. The propagation of Northwest climate change impact throughout the grid is exacerbated by the occurrence of dry years in Northern California. Generation from the Desert Southwest emerges as a critical resource to compensate for variations in water availability, and generation, in these regions. Though the regional power flow directions seem insensitive to long-term variations in water availability, our analysis highlights the need to consider other compounding regional factors, such as changes in Southern California's net load and changes in regional fuel prices.