Share this
As the energy sector rapidly evolves, long-duration energy storage (LDES) is emerging as a key solution for addressing the complexities of integrating renewable energy into the grid. (Long-duration energy storage includes pumped hydropower and batteries with more than 10-hour durations.)
Utilities must understand how to incorporate LDES into their integrated resource planning process; however, it can be challenging to model LDES and quantify the benefits.
We’ll delve into how modeling can be used to evaluate long-duration energy storage effectively, ensuring that future energy needs are met reliably and sustainably.
Why Is It Difficult to Model Long-Duration Energy Storage (LDES)?
Most capacity expansion models use aggregate or typical days to be able to optimize integrated resource planning over multiple years considering several technology types. This is because model run times can be extensive.
The primary value of LDES lies in covering multi-day reliability events that you can only “see” when you’re considering every day of the year in your optimization.
What Are the Main Objectives of Resource Planning?
- Keep prices low: To minimize revenue requirements and cost risk.
- Keep the lights on: To ensure resource adequacy and grid resiliency.
- Increase sustainability: To assist in decarbonization and meet sustainability goals.
What Are the Challenges of Long-Term Integrated Resource Planning?
What makes the capacity expansion problem difficult is the need to look over a long-term horizon.
We’re trying to capture long-term capital costs, but we also see that there are tightening environmental regulations. Often, a 2030 or 2035 target will be to achieve an 80% reduction in greenhouse gas emissions, and then by 2050 the goal will be to come close to net zero emissions. You don’t want too short of a horizon to factor in these regulatory changes.
Also, with the Inflation Reduction Act, there are numerous tax credits for renewable and storage projects, as well as other types of technologies, but most will expire in the mid 2030s. You’ll likely want to take advantage of those before then.
In addition, you’ll want to consider new technologies, especially as they become commercially viable. You don’t want to box yourself in with near-term decisions when new technologies may emerge to help you meet your long-term goals more cost effectively.
Finally, LDES can help you meet demand during multi-day renewable droughts. It’s important to be able to capture ancillary services and reserves so that you have a resource and operating plan that allows you to meet expected weather conditions and prepare if weather conditions change significantly. You can often achieve this by looking at a Monte Carlo analysis and creating a plan that allows you to meet your reliability targets, commonly expressed in terms of a loss of load expectation (LOLE), for instance, a one-day in ten-year rule.
Your Model Wish List
In an ideal world, what do you want your capacity expansion model to be capable of?
You want it to be able to take a long-term view, look at capital recovery, understand changes that could happen on your system, and account for your project finance parameters.
Also, you want it to be able to capture ancillary services that will let you maintain reliability and resiliency during extreme weather events.
You want it to do all that and still be able to maintain some daily resolution so that you can capture some of those renewable droughts and give credit for resources for peak times or for long-duration energy storage.
Finally, you want to be able to look at your reliability metrics to incorporate some type of a Monte Carlo analysis into the resource planning process.
Given the magnitude and complexity of this endeavor, it’s too big of a problem to solve at the push of a button. We suggest breaking up the problem into four phases.
Breaking Up the Resource Adequacy Problem
Resource Adequacy Analysis
First, the resource adequacy analysis will evaluate the necessary planning reserve margins to meet the reliability targets of your current portfolio.
It's a common practice as part of that resource adequacy analysis to determine what the effective load-carrying capability (ELCC) of the various resources are and to come up with curves for the ELCC. Generally, you’ll see that if you were to increase the amount of wind, solar, or short-term storage, the marginal benefit of adding those from a load-carrying capability will decline. You want to be able to capture those curves and then feed them into a capacity expansion model.
Long-Term Optimization
For many years, models have focused on a capacity expansion problem, which we’ve labeled as the long-term optimization. The long-term expansion plan will look over multiple years at renewable or decarbonization targets, account for when the tax credits would expire, and capture various technologies.
Then we'll take that long-term expansion plan and lock in most or all of that. In particular, it's usually safe to lock in your renewable build-outs and some of your short-term storage requirements.
The chart above shows what wind shapes might be for each day of a month. The red line shows a typical day aggregation of what the 28 to 31 days might look like. That typical day aggregation, while accurately capturing the capacity factor and the range of outputs, can “hide” multi-day events where the wind output will be consistently low.
Mid-Term Optimization
The mid-term optimization refers to a one or two-year optimization. It lets you look at reliability events and add capacity, such as long-duration energy storage, to meet that demand.
Sometimes we have reliability droughts that are effectively hidden from our long-term optimization view that we can see when we look at it on a daily basis. If we identify additional peaking and LDES resources that are necessary, we may go back and put those in and redo our long-term optimization.
To give an example of midterm optimization’s value, the chart below is an example of where we ran a year with full daily and hourly granularity (8,760 hours). One of the candidate resources was a 100-hour storage technology. You can see when tracking its state of charge over time how it was used seasonally, in particular with deep discharges over specific summer and winter days.
Long Duration Energy Storage Levels
Short-Term Optimization
Then, there’s the production-cost problem, which we call the short-term optimization.
We can inform that production-cost model of not just the expansion plan that we derived but also the allocation of limited resources over the year. There may still be some shortfalls that reveal themselves because of other additional detail, such as the number of annual hours represented or resource outages, that we put into that short-term optimization.
This means it may be necessary to adjust the plan and either add some additional capacity or change what our allocation of those annual limits would be.
Why EnCompass?
Currently, about 29 utilities, five state commissions, and several consultants and intervenors employ the EnCompass model. While EnCompass provides a robust computation solving these complex problems, the software takes advantage of more current software development technologies to provide an easier-to-use, easier-to-understand framework. This allows the user to spend more time analyzing the problem instead of wrestling with modeling intricacies.
In EnCompass you have options to model many different scenarios and to analyze and defend data from one cohesive system. You can perform a full optimization of all resources within a market or utility control center enforcing transmission limits and a complete set of operating costs and real-world constraints.
This is invaluable for:
- Integrated Resource Planning
- Valuation and Risk Assessment
- Power Market Forecasting
- Economic Transmission Analysis (LMP)
Want to learn more? Ask our team a question or request a demo.
About the author: Norm Richardson built the EnCompass model, offered by Yes Energy. His past software and data experience includes leading a group of product managers and consultants at Ventyx (previously NewEnergy Associates, now Hitachi Energy) focused on power modeling software and databases, price forecasting, integrated resource planning, and economic transmission analysis.
