A Nationwide Multi-Location Multi-Resource Stochastic Programming Based Energy Planning Framework
The global increase in energy consumption and demand has forced many countries to transition into including more diverse energy sources in their electricity market. To efficiently utilize the available fuel resources, all energy sources must be optimized simultaneously. However, the inherent variability in variable renewable energy generators makes deterministic models ineffective. On the other hand, comprehensive stochastic models, including all sources of generation across a nation, can become computationally intractable. This work proposes a comprehensive national energy planning framework from a policymaker's perspective, which is generalizable to any country, region, or any group of countries in energy trade agreements. Given its relative land area and energy consumption globally, the United States is selected as a case study. A two-stage stochastic programming approach is adopted, and a scenario-based Benders decomposition modeling approach is employed to achieve computational efficiency for the large-scale model. Data is obtained from the U.S. Energy Information Administration's online data collection. Scenarios for the uncertain parameters are developed using a k-means clustering algorithm. Various cases are compared from a financial perspective to inform policymaking by identifying implementable cost-reduction strategies. The findings underscore the importance of promoting resource coordination and demonstrate the impact of increasing interchange and renewable energy capacity on overall gains. Moreover, the study highlights the value of stochastic optimization modeling compared to deterministic modeling to make energy planning decisions.