Four years ago, researchers at the National Renewable Energy Laboratory (NREL) won Department of Energy (DOE) ARPA-E funding to invent a new long-duration thermal energy storage technology able to discharge heat or power for 100 hours.
The goal was to devise long-duration storage with a storage cost as low as pumped hydro, with the target of a $0.05/kWh Levelized Cost of Storage (LCOS). To achieve this, the research team chose sand as the storage media and devised a way to heat it by pouring it through a series of heating elements like in a huge toaster.
Sand is a favored thermal energy storage media as it has very high thermal stability allowing it to cycle between ambient air temperature and over 1000°C. The wide temperature range increases energy storage density and system efficiency. Sand is widely available and cheap at about $30 a ton. In an insulated silo, such as the NREL team proposes, it would lose only 1% of the heat a day.
The ETES long-duration thermal energy storage in sand thermal energy storage demo. Because the storage media – sand – is cheap and durable, adding additional storage duration is relatively easy, once the power conversion infrastructure is built—similar to pumped hydro. Batteries, by comparison, would have to be placed in series to reach these long durations and be subject to degradation because each battery has a short lifecycle and would need replacing over that timescale. IMAGE ©NREL
Like pumped hydro, which cycles water for decades, a sand-based storage system could cycle the sand for decades once built. Just as the water in pumped hydro can be reused again and again for decades, so can sand be heated and cooled again and again, so this thermal battery would also last decades.
Sand does not degrade at these very high temperatures. The ability to withstand multiple heating and cooling cycles means sand-based thermal storage combines a durable storage media and a very long plant lifetime.
The team’s latest paper, “Modeling electrical particle thermal energy storage systems for long-duration, grid-electricity storage applications,” is due out in the October 2024 issue of Applied Energy.
“The purpose of our paper is to understand the influence of component design and performance parameters as well as operations on the overall technical performance and design of the system,” explained NREL Researcher Jeffrey Gifford, lead author of the paper.
Electrical resistance heaters heat the particles to charge this thermal battery to 1200°C. The heated sand is stored for up to several days in insulated industrial silos. The particles transfer heat to air through a novel pressurized fluidized-bed heat exchanger the team designed to discharge the stored heat; the team has been issued a patent for this novel heat exchanger (U.S. Patent No. 11,740,025).
The sand-based long-duration thermal energy storage (demonstration scale schematic) IMAGE ©NREL
The hot air can be used directly in industrial processes or drive an air-Brayton combined cycle to generate electricity, similar to modern, high-efficiency natural gas combined cycle power plants.
Round trip efficiency of the stored heat exceeds 95% for multiple days, while for electricity-in to electricity-out, the team’s modeling of their electro-thermal energy storage (ETES) system predicts a round-trip efficiency of 50-52% Gifford explains the difference between two metrics:
“The 95% storage efficiency was just focusing on the thermal energy left in the silos,” Gifford noted. “While accurate, people are often confused by this when comparing to the round-trip efficiency (RTE) of other energy storage technologies, which is the most often cited term. In our paper, we calculated a RTE of 50-52%.”
With success in that initial four-year step of modeling, engineering, and testing each new component individually, in July 2024, the NREL team has begun the next five-year step toward commercialization—building and getting all the components working together in a pilot-scale demonstration plant.
The DOE’s Office of Clean Energy Demonstration will support their project with $4 million, and the resulting particle-based thermal energy storage plant will be built, demonstrated, and displayed at the NREL campus in Colorado.
“The purpose of our demonstration is so stakeholders can visit to see how it does,” said Zhiwen Ma, Thermal Systems R&D head at NREL, who leads the project.
“For our demonstration purpose, we will only put in 100 kilowatts and 10 hours to demonstrate the integration between the heat exchanger and the power generator. This will be just 17 meters tall, nothing near a commercial unit. At commercial size – a silo for a hundred hours of thermal storage could be up to 70 meters tall and 20 meters in diameter and store and discharge hundreds of megawatts.”
Once the demo is built, the next step would be to partner with a business to make a first commercial project. Ma is confident.
“A lot of people have interest in our system,” he said.
“So we think the demonstration we build here on the campus can prove to the customer and gain their confidence to invest. This thermal storage has no site restrictions compared with pumped hydro. It can be deployed anywhere. People can see that energy storage is critical. The beauty of sensible storage is that you can go from a few hundred kilowatts to a few hundred megawatts; depending on the customer’s need, we can scale this thermal storage. So there’s a very broad range of heat and or power capacity. ”
Many industries require heat, and its application as a direct heat source would be the largest and most straightforward market, replacing industrial heat that currently relies heavily on fossil energy. The funding goal is to reduce the LCOS to 5 cents/kWh. The team estimates a 7 cents/kWh LCOS for a stand-alone build-from-scratch system.
However, Ma said that if the ETES plant used the existing infrastructure in a former coal plant by repurposing its assets, it would meet that 5 cents/kWh LCOS goal.
In this case, the ETES could be hosted in a former coal power plant, taking advantage of funding in the IRA that incentivizes conversions of fossil energy to renewable energy so that a retired coal plant could keep selling power to the grid but not by burning coal. Instead, renewable energy would provide the electricity to charge the thermal storage, like a battery, but can discharge for longer durations than a battery.
Instead of burning coal to run the thermal power block, the former coal plant site would regularly store sand heated by electricity from curtailed wind or daytime solar. The heat stored in the sand would then be converted back into electricity through the thermal power block and delivered to the grid for hours or days when needed, for example during periods of low solar or wind availability.
This would be a win-win. As grids add more renewables, the demand for energy storage increases. And as coal and gas plants retire, the opportunities for reusing their thermal assets for thermal energy storage grow, helping a resilient and renewable grid.
