Quantifying the Impact of Energy System Model Resolution on Siting, Cost, Reliability, and Emissions

TL;DR

This work tackles structural uncertainty in energy system models by quantifying how spatial, temporal, and operational resolution affect siting, cost, reliability, and emissions. Using a state-of-the-art GenX-based pipeline with a high-resolution baseline (HRB) and a two-phase capacity expansion plus operations approach, the authors compare many downscaled cases under BAU and carbon-penalty scenarios. They show that finer spatial resolution yields more realistic siting and improved emissions and reliability, while low spatial resolution substantially misplaces VRE and underestimates bottlenecks; temporal coarsening also biases results but to a lesser extent. The study demonstrates that there is often no diminishing return to increasing resolution for multiple metrics and argues for balanced, computationally efficient granularity across spatial, temporal, and operational dimensions to support robust policy decisions, aided by advances such as Benders decomposition in GenX.

Abstract

Energy systems models, critical for power sector decision support, incur non-linear memory and runtime penalties when scaling up under typical formulations. Even hardware improvements cannot make large models tractable, requiring omission of detail which affects siting, cost, and emission outputs to an unknown degree. Recent algorithmic innovations have enabled large scale, high resolution modeling. Newly tractable, granular systems can be compared with coarse ones for better understanding of inaccuracies from low resolution. Here we use a state of the art model to quantify the impact of resolution on results salient to policymakers and planners, affording confidence in decision quality. We find more realistic siting in recommendations from high resolution energy systems models, improving emissions, reliability, and price outcomes. Errors are generally stronger from low spatial resolution. When models have low resolution in multiple dimensions, errors are introduced by the coarser of temporal or spatial resolution. We see no diminishing returns in accuracy for several key metrics when increasing resolution. We recommend using computationally efficient techniques to maximize granularity and allocating resolution without leaving any aspect (spatial, temporal, operational) of systems unduly coarse.

Figures (21)

• Figure 1: Graphical methodology. High resolution (HR) cases are run at 26-zone, 52-week resolution with unit commitment (UC) constraints active.
• Figure 2: Region aggregations used. Name abbreviations are listed in Table \ref{['region_glossary']} in the SI.
• Figure 3: Transmission topology under two spatial resolutions. When increasing spatial resolution, some intraregional capacity becomes interregional. Black lines are backbone transmission capacity. Yellow circle represents an example renewable resource investment site connecting via a spurline to an urban area (red polygon.) Real systems include tens of thousands of such sites for wind and hundreds of thousands for solar.
• Figure 4: Optimal installed capacity by region (\ref{['total_r']}) and technology (\ref{['total_t']}) for the 26-zone 52-week case for both the business-as-usual (BAU ) and carbon penalty (CP ) cases. Carbon penalty increases total capacity 36.3%, from 1580 to 2154 GW, increases solar capacity 67.7%, and onshore wind capacity 124.2%. Carbon penalty increases fraction of buildout on the east coast. Aggregations listed in \ref{['total_r']} are copied here from Fig \ref{['geography_7']} and are listed in SI Table \ref{['region_glossary']}.
• Figure 5: Difference in installed capacity by technology relative to the high resolution baseline (HRB). Zonal resolutions are shown in \ref{['diff_zone_co2']}, \ref{['diff_zone_ref']}, spatial and temporal resolutions in \ref{['diff_temp_co2']}, \ref{['diff_temp_ref']}. Increasing spatial granularity generally increases solar buildout and decreases onshore wind. Trends are stronger for spatial resolution and the carbon penalty cases. Omission of unit commitment (UC) decreases renewable energy buildout.