As electric companies plan for the future electric grid, they rely on complex models to understand which technologies, like solar, wind, and energy storage, will best fit future grid needs over time. These models help answer critical questions: How much storage should be built? Where should it go? And how will it operate day-to-day alongside other resources?
But electric power system planning models—such as capacity expansion, production cost, and resource adequacy tools—can be data-intensive and computationally demanding. To keep modeling practical, planners often must simplify, trading modeling detail for efficiency. However, balancing modeling detail with tractability can be nuanced, as simplifications may influence modeled results.
For energy storage technologies, achieving this balance is important, as accurately modeling the state of charge (SoC) – the amount of energy available at a given moment to meet load demands in future periods – is essential to understanding their value in a resource portfolio. In practice, planners use a range of methods to feasibly model energy storage in long-term resource plans (see Energy Storage in Long-Term Resource Planning: A Review of Modeling Approaches and Utility Practices).
These modeling choices are critical as they directly affect how much storage is built, the type of storage, and where it’s located. Ultimately, planners and modelers must carefully weigh the tradeoffs between enhancing modeling strategies and increasing model complexity within the current suite of long-term modeling tools.
Recent work by EPRI’s Resource Planning for Electric Power Systems Program explores how two key modeling dimensions—temporal resolution (how time and chronology are represented) and spatial resolution (the level of detail in which the transmission network is represented) —shape energy storage modeling outcomes. The findings highlight how simplifications in these areas can lead to inaccurate evaluations that may underestimate or overestimate the value of energy storage in planning studies.
Assessing Temporal and Spatial Modeling Choices for Energy Storage in Long-Term Resource Planning
To better understand how simplifications impact long-term planning model results, EPRI research explores the trade-offs between different spatio-temporal resolutions (see the research brief on ”Assessing Temporal and Spatial Modeling Choices for Energy Storage in Long-Term Resource Planning”). The analysis focuses on how modeling choices impact the representation and planning of energy storage technologies of varying duration. Key highlights:
Modeling storage resources can be complex due to various storage technology types with different operational characteristics.
The timing and quantity of storage capacity installations over time are influenced by both temporal resolution and chronology. The figure below illustrates how different temporal modeling strategies lead to significantly different expansion results for new storage capacity across low-carbon resource portfolios.
Figure 1. Comparison of new storage capacity (MW) across three low-carbon resource portfolios, identified using different temporal resolution modeling strategies. Partial chronology simplifies the hourly load into a set of blocks per month based on load duration curves. Sampled chronology selects the most representative sample days, weeks, or months in the year while maintaining the chronological order of the simulation periods. Fitted chronology approximates the hourly load series with a step function using a weighted least-squares technique, and simulation periods preserve their chronology since only continuous intervals are aggregated into steps.
Higher temporal resolution tends to select more new, longer-duration storage capacity, which can discharge energy for extended periods. On the other hand, models with lower temporal resolution tend to select greater amounts of shorter-duration storage. The figure below illustrates how these temporal representation strategies influence new storage capacity, comparing 4-hour and 23-hour storage technologies.
Figure 2. Comparison of storage deployment by capacity (MW) under different temporal resolutions. The darker line, “Fitted – 8 blocks per day”, represents a fitted chronology approach that divides the day into eight 4-hour blocks. The grey line, “Fitted – 24 blocks per day”, represents a fitted chronology approach that uses 24 blocks per day, corresponding to all 24 hours.
Including detailed modeling of the transmission network can help mitigate future congestion, allowing planners to pinpoint deployment timing and locations where storage can be most effective. The figure below illustrates how new storage capacity varies across a resolved nodal representation under two different temporal modeling approaches, highlighting the differences in near-term new storage capacity between spatial representations, and long-term differences in new storage capacity between temporal representations.
Figure 3. Comparison of storage deployment by capacity (MW) under different spatial and temporal resolutions. In the regional approach, the transmission network is represented as a single region where all resources are connected to a single node, following a copperplate model. In the nodal approach, the full transmission network is modeled at the nodal level, explicitly representing system nodes and grid constraints.
Resources and Detailed Technical Reports
EPRI research also supports resource planners seeking guidance on modeling the characteristics of storage technologies. This includes cost and performance estimates for selected mechanical and thermal long-duration energy storage (LDES) technologies and evaluations of installed costs and performance attributes for both commercial and emerging battery technologies.
For more technical insights, full research publications referenced in this article can be found below:
Assessing Temporal and Spatial Modeling Choices for Energy Storage in Long-Term Resource Planning (3002028963) and companion research brief.
Energy Storage in Long-Term Resource Planning: A Review of Modeling Approaches and Utility Practices (3002028378).
EPRI’s Resource Planning for Electric Power Systems Program supports long-term planning with objective electric sector data, outlooks, methods, and state-of-the-art modeling. Research featured in this post is by Karen Tapia-Ahumada and Sean Ericson. Article written by Ryan Fulleman and Karen Tapia-Ahumada.