To Build a Zero-Carbon Grid, We First Need to Model It Accurately

The following is a contributed article by Scott Burger, analytics lead at Form Energy, Marco Ferrara, co-founder and senior vice president of analytics and business development at Form Energy, Roderick Go, technical manager at E3, and Arne Olson, senior partner at E3.

Electricity is deeply entwined in our daily lives, but most people only think about it when the bills arrive — or when the power goes out. Rolling blackouts in August left millions of Californians without power during a sweltering heatwave, bringing reliable electricity to top of mind. Renewable energy was not to blame for the blackouts, according to the state's root cause analysis and the views of informed observers. However, the events did shine a spotlight on a key limitation of renewable energy sources: that they can only generate when the resource is available. Additional resources will be needed to ensure reliable, around-the-clock electricity systems.

Energy storage can complement wind and solar energy by charging during times when wind and sunlight are plentiful and discharging during times when they are scarce. California already expects over 2,000 MW of lithium-ion batteries to be installed by the summer of 2021, and as much as 10,000 MW by 2030, providing a significant capacity boost and enabling the state to reduce its reliance on gas generation.

However, lithium-ion batteries are themselves limited by their short duration (typically 2-4 hours). While they can make a significant contribution to California's capacity needs, they may not be available during multi-day weather events with low wind and solar production. Recent studies by E3 and others have concluded that  California will require up to 35 GW of "firm" capacity — generation that can operate whenever needed — through 2050 on a deeply-decarbonized grid. Similar conclusions are reached for New England and the Pacific Northwest. Today, this need is filled by natural gas generation, but the search is on for cost-effective, lower-carbon alternatives.

Multi-day energy storage systems such as "green" hydrogen or Form Energy's aqueous air battery can function as alternative sources of firm capacity for highly renewable and deeply decarbonized grids. These systems are quite different from today's battery technologies, unlocking the potential to cost-effectively store excess wind and solar production for days or even weeks and eliminating the need for fossil generation. Multi-day load flexibility may be able to play a similar role by relying on storage capability that is inherent in existing supply chains.

New tools are needed to plan low cost, reliable, clean grids with long-duration energy storage
Unfortunately, many of the planning models used by utilities and their regulators today are not well-suited to consider the role of long-duration energy storage and multi-day load flexibility as a complement to variable renewable generation.

The industry has already improved upon yesterday's planning tools that use simple heuristic "screening curves" to evaluate the tradeoffs between "baseload" and "peaker" plants to meet peak demands. Newer tools in use today are designed to answer questions about integrating and valuing relatively modest amounts of renewables on "typical operational days" and "net load" peak reliability, as shown in the chart below. However, looking forward, the industry needs another quantum leap in modeling to capture the full operational value of long-duration storage and very high renewable penetrations while ensuring reliability under a wide range of complex, interconnected system conditions.

While reliability models based on loss-of-load probability have been adapted to appropriately consider the role of renewable energy in maintaining reliability, optimizing the use of long duration storage introduces significant additional complexity that is well beyond what current models can handle.

Power system planners need new tools to efficiently plan cost-effective, reliable, low carbon grids with long-duration energy storage. What do these new tools need to do? While we don't have all the answers, some key themes are emerging from research and practice.

First, planning models need to better capture the hour-to-hour dynamics of the grid over contiguous days, considering all potential operating conditions rather than a snapshot of a few hours. Form Energy's research shows that planning investments over a time horizon that captures the hour-to-hour dynamics of demand and renewable energy supply for a full year can have a significant impact on utility portfolio costs and reliability.

Temporally granular planning captures the value of emerging technologies like multi-day storage that absorb energy in one period and discharge this energy in another hour, day, week, or season. This longer-horizon arbitrage can reduce renewable curtailment and capacity needs but is not possible to capture in incumbent planning models that do not model all hour-to-hour grid dynamics.

Second, planning models need to better capture key system constraints such as local reliability challenges with greater fidelity. E3's research and Form Energy's research highlight how meeting demand in certain constrained pockets of the grid requires a clear understanding of the hour-to-hour demand in those pockets over periods from days to weeks. These demand pockets are often challenging to access with new transmission and are commonly supplied by polluting fossil fuel-fired power plants, driving environmental justice concerns.

Long duration storage requires proper planning
Best practice modeling techniques are particularly important when planning for firm, low carbon technologies like multi-day storage. Because many of today's planning models don't capture the full hour-to-hour dynamics of the grid across days or weeks or extreme weather events during the portfolio design stage, they are unable to realize the full value of technologies that arbitrage energy production and prices across days or weeks and that manage the risks posed by extreme weather.

As Form Energy's experience and research shows, some models exclude long-duration storage from the portfolio purely because of the model design, even when long-duration storage could add value to the portfolio. Long duration, multi-day storage is garnering increased attention, investment, and procurements, increasing the importance of the tools necessary to accurately plan for it.

Fortunately, the California Energy Commission recognizes the challenge and recently awarded grants to two teams, including one consisting of E3, Form Energy and the University of California, San Diego, to analyze the role of long duration storage in the California grid. One of the outcomes of this effort will be new, publicly-available tools that can more accurately model California's decarbonizing grid and the role of long duration storage within it, including both bulk grid and localized values. The goal is for these tools to be used in subsequent planning processes in California and to provide a template for other states and regions.

Widespread adoption of improved planning techniques holds the potential to spur innovation in firm, low carbon power sources; lower customer costs; and improve reliability. However, it will require utilities and their regulators nationwide to update antiquated tools that are embedded in historically long-running regulatory processes and practices. The stakes are high. As the California blackouts demonstrate, there is little room for error.