The Renewable Efficient Frontier: Geographic Optimization of Solar and Wind for Practical Utility Planning

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Methodology combining mean-variance optimization with renewable energy data can be adapted to other regions or integrated with more detailed grid models.

🏢実務担当者:Utility planners can directly apply the geographic optimization framework to screen and rank potential renewable sites for portfolio construction.

🏛政策担当者:Findings on cost and volatility trade-offs inform renewable energy siting policies and transmission expansion planning.

Electric utilities pursuing decarbonization increasingly rely on solar and wind resources which exhibit inherently intermittent generation driven by weather conditions. Managing this intermittency poses a central challenge for grid planning and operations, as utilities must avoid both shortfalls in generation (load shedding) and excess supply (curtailment). This study develops a replicable, and utility-oriented framework for selecting solar and wind generation sites that reduce the aggregate intermittency. In close collaboration with Salt River Project (SRP), I analyze nine prospective renewable sites in the western United States. Using 36 years of hourly reanalysis weather observations, I simulate hourly generation using the System Advisor Model (SAM) to analyze diurnal and seasonal complementarities between solar and wind. I apply a mean–variance (MV) optimization framework drawing on Markowitz and Freund to determine portfolios that balance expected generation and volatility. The model identifies both system performance and the composition of e!cient portfolios across prospective sites. I begin by isolating diversification benefits using average capacity factor as expected return and portfolio variance as a proxy for intermittency risk. Later, a cost-adjusted model weights output by site-specific cost estimates, allowing assessment of how economic tradeo”s alter optimal site selection. Finally, I extend the cost-adjusted model by restricting it to late-afternoon hours in the summertime, when Arizona electrical demand is high but solar production wanes. I find that solar sites within Arizona are highly correlated and yield limited diversification benefits on their own. However, substantial reductions in portfolio volatility emerge when geographically dispersed wind resources are added. Once cost differences are incorporated, portfolio composition shifts toward sites closer to SRP’s service territory, as they become increasingly competitive relative to more distant resources with higher transmission costs. Nevertheless, cost-adjusted portfolios continue to leverage some out-of-state wind for smoothing generation, but favor lower-cost regional sites over more capital-intensive alternatives.

• Zenodo https://zenodo.org/records/21520945first seen 2026-07-24 04:13:44