Designing Robust Power Systems Through Transdisciplinary Approaches in Energy Systems and Climate Modeling

Ensuring electricity reliability while transitioning to renewable power sources despite an increasing number of record-setting weather events is imperative to public health, economic prosperity, and national security. Predicting the impact of climate change on electricity generation and planning capacity expansion efforts accordingly is needed to ensure electricity reliability in the near-term future. Yet, there is a gulf between energy systems and climate modeling practices in spatial and temporal resolutions and challenges in managing the uncertainty of combined models. Here we use the WUS-D3 dynamically downscaled GCM dataset in a robust decision-making approach to present a framework for identifying capacity expansion pathways that will ensure electricity reliability despite uncertainty. We generate capacity expansion pathways via an optimization function designed to minimize hours of electricity supply and demand imbalances in a mixed integer linear programming model under differing levels of renewable generator penetrations. To test the resource adequacy of the created future power systems, a stochastic, Monte Carlo-based simulation model quantifies hours of electricity demand and supply imbalances. With data at a 9km hourly resolution we analyze changes in demand, net demand, and surplus available capacity (SAC) of electricity generation in the Western Electricity Coordinating Council (WECC) region. Regression models were created to forecast demand based on temperature and hydropower based on runoff. Solar capacity factors, wind capacity factors, and derates of thermal generation units were calculated based on temperature, humidity, solar radiation, and wind speeds. SAC is a key metric that combines these factors to measure the health of the electricity grid by quantifying the unused generation capacity available on the grid, based on projected demand and renewable energy generation at any given time. We utilize 14 GCMs in the WUS-D3 dataset to account for uncertainty in climate model projections by incorporating each GCM as a potential future climate reality in our modeling.

Initial results indicate that areas of WECC will experience continual declines in summertime SAC between 2025 and 2063 without changes to the region’s electricity generation portfolio. The most precipitous losses are projected to occur between 2035 and 2045. In California, the Desert Southwest, and the Central Southwest, low to negative SAC is most common in June through September, aligning with high demand because of high temperatures. SAC concerns in the Northwest are more common between September and January, linked to the period of lowest hydropower availability. The September overlap between high demand in the south and lower hydropower availability in the north, is likely to become increasingly problematic across the region. Negative SAC indicates that demand on the grid is greater than available power supply and often leads to blackouts. If California is unable to import power from neighboring areas, SAC values may reach as low as -8 to -10 GW between September and November by 2055 and -11 GW between June and August before 2050. The Northwest could see similar problems with summertime SAC values as low as -4 GW. We aim to prove that these impacts can be mitigated via planning approaches that combine energy systems modeling and climate datasets under robust decision-making which accounts for uncertainty in our understanding of future climate scenarios.