As renewable penetration grows, the need for energy storage grows as well. However, existing storage technologies may not be able to meet that need.
A Department of Energy (DOE) grant program aims to address that problem by trading efficiency for low cost.
Existing storage technologies fall short in terms of duration. Lithium-ion batteries, which have captured about 90% of the energy storage market, can economically run for about four hours. Pumped hydro storage, which accounts for the most energy storage today in terms of capacity, can generate power by releasing water from storage reservoirs for up to 20 hours under certain conditions.
But as renewable energy penetration reaches and surpasses 50%, the need for even longer durations will grow.
The sunny forecast for storage demand
The need for longer durations is particularly applicable to areas where solar power dominates the resource mix.
In a future where 70% of power is generated by solar panels, it is easy to imagine a scenario where a couple of cloudy days in a row could create a gap in meeting customer demand, Michael Jacobs, senior energy analyst at the Union of Concerned Scientists, told Utility Dive.
Filling that gap is the aim of a program at the DOE that is providing funding for technologies that can extend energy storage durations to up to 100 hours.
Last month, the DOE's Advanced Research Projects Agency-Energy (ARPA-E) awarded just over $28 million to 10 projects that aim to push the limits of energy storage duration. ARPA-E's Duration Addition to electricitY Storage (DAYS) program aims to push the duration of energy storage systems out to 100 hours.
One hundred hours, just a little more than four days, is an exponential leap from current durations but the role of ARPA-E is to focus on early stage technologies that are not yet commercial or quite ready for the private sector.
"Wind and solar will clearly be the cheapest forms of electric energy in the future," Paul Albertus, the director of the DAYS program, told Utility Dive. So, "it is pretty clear that over the next 10 years or so" the need for longer duration energy storage is going to grow, he said.
Some pumped hydro storage could get to the 100-hour mark, but the cases are limited and most max out at around 20 hours, Joseph Manser, a fellow at ARPA-E who helped set up the DAYS program, told Utility Dive. Lithium-ion batteries could be arrayed to discharge for 100 hours, but that would be very expensive.
What the DAYS program is looking for is not tweaking existing technology or even a technological breakthrough but for technologies that "trade off costs in unique ways," Manser said.
"We are working on systems with low energy costs," Albertus said. Energy storage systems with very long durations have a different type of cost structure, he said. "They cycle less often, so you no longer get daily revenues," and capital costs and operation and maintenance tend to be lower. Capital costs for very long duration energy storage systems generally fall in a range of $5/kWh to $30/kWh compared with an "optimistic" $150/kWh cost for a lithium-ion system, Albertus said.
"It comes back to the fact that grid is built on plants that can run forever, given enough fuel. Until they are not there anymore, that is your long term storage."
Senior research analyst, Navigant Research
Flow batteries are perhaps the most viable of the technologies in today's market, but they face also face efficiency challenges. Flow batteries typically have a round trip efficiency of about 75% compared with 90% to 99% for lithium-ion batteries. When durations get out to around eight to 10 hours, though, their levelized costs beat lithium-ion, Alex Eller, senior research analyst at Navigant Research, told Utility Dive. The problem is there is not necessarily a real incentive for something with that long a duration in today's market, Eller said.
"It comes back to the fact that grid is built on plants that can run forever, given enough fuel. Until they are not there anymore, that is your long term storage," Eller said.
Taking the heat
Of the 10 projects that won DAYS awards, thermal systems were the most prevalent.
"There are some compelling reasons why thermal storage is viable," Manser said. The materials are cheap, which makes storing energy as heat cheap.
Four of the awards went to systems that used some form of heat transfer, including projects by Echogen Power Systems, Antora Energy, the National Renewable Energy Laboratory and Michigan State University. A fifth project by Brayton Energy combines thermal storage and a gas turbine that uses molten salt as a storage medium.
Three of the awards went to firms researching some form of flow batteries, which create electricity via the interchange of electrolytes. Form Energy is working on a chemistry that uses sulfur in a water-based solution. Primus Power is working with the Columbia Electrochemical Energy Center to develop a new approach to zinc bromine flow batteries. And United Technologies is focusing its research on chemistries that use inexpensive and readily available sulfur-manganese based materials.
Quidnet Energy won a grant to develop a technology that pumps water into rocks under ground. When the pressurized water is released, it is used to produce electricity. And a team from the University of Tennessee, Knoxville won a grant to develop an electrolyzer-fuel cell combination.
For the thermal storage projects the challenge is working against the second law of thermodynamics. In essence, energy is lost when energy is transferred. "It is a real challenge," Tim Grejtak, an energy storage analyst at Lux Research, told Utility Dive.
Preparing for that day, ARPA-E has approached the DAYS program with a strategy of "giving up efficiency to get capital costs down," Albertus said.
"[I]t would be okay if long duration storage systems are less efficient, as long as they are cheap to build."
Senior energy analyst, Union of Concerned Scientists
"Efficiencies matter less as [you] get closer to 100% renewable because of the amount of surplus renewable energy on the grid," Jacobs said. The least efficient plants are only used to meet peak demand. "The idea that the last units are less efficient goes with the territory." In that sense, Jacobs said, "it would be okay if long duration storage systems are less efficient, as long as they are cheap to build."
Typically, successful ARPA-E projects can take anywhere from seven to 10 years before becoming commercial. There could be a second phase of DAYS projects, depending on what Congress does with appropriations, said Albertus. For the current projects, the next phase would be to build a small system for field demonstration.
"Three years from now, we might see what's going on with these technologies," he said.