Storage is renewable energy's greatest challenge -- this low-cost sulfur battery may help

"It has become increasingly clear that in order for renewable energy to become the main part, if not all, of our electricity generation system, it needs to match the output of the demand that we have as a society," says senior author Yet-Ming Chiang of MIT's Department of Materials Science and Engineering. "We think that this work helps move us in the right direction and creates more hope that this is possible, but we need to push it ahead very quickly because we don't have a lot of time."

One of the criticisms of renewable energy is its variability. For example, there are times when a cloud goes in front of the sun or when the wind dies down, and so being able to store energy for those down times is essential for uninterrupted energy flow. At the moment, the coupling of energy storage to renewable generation is in its infancy -- it does happen, but of the total amount of solar and wind energy generated, a very small percentage is actually stored, with the cost of energy storage being one of the greatest barriers.

Under the former Secretary of Energy Steven Chu, the Department of Energy's Joint Center for Energy Storage Research set a goal for "5-5-5" (meaning 5 times reduction in cost, 5 times increase in energy density, accomplished in 5 years) for grid storage. In response, Chiang's group focused on the first part of the problem, examining how to create a storage unit with a low cost-per-stored-energy metric (US dollars per kilowatt hour, $/kWh), based on the cost of the cathode, anode, and electrolytes of a battery. Current chemical costs often hover between $10 and $100/kWh as battery materials often need to be mined and shipped from around the globe.

Chiang and his colleagues were particularly interested in the potential of sulfur -- an abundant nonmetal that is a product of natural gas use -- as a core component of a lightweight and inexpensive storage battery. All batteries are made up of a positive anode, a negative cathode, and an electrolyte to carry the charge, and the research group wanted to explore how sulfur could be the cathode and water could be the electrolyte.

"We went on a search for a positive electrode that would also have exceptionally low cost that we could use with sulfur as the negative electrode," Chiang says. "Through an accidental laboratory discovery, we figured out that it could actually be oxygen, and therefore air. We needed to add one other component, which was a charge carrier to go back and forth between the sulfur and air electrode, and that turned out to be sodium." The total chemical cost of this battery is about $1/kWh.

Once the researchers decided on the components, they then needed to decide what the rest of the battery was going to look like. Since all of the chemical components of the battery are dissolved in water, they decided on a flow battery architecture in which, through a set-up of pumps and tubes, electrical charge causes the components of the battery to flow past each other, generating chemical reactions that help it capture electrons. One complication to this approach is that the amount of electrical charge that can be stored depends on the amount of liquid in the anode and cathode. This means that the battery needs to take up more space than what is traditionally used, but the cost of the materials offsets that drawback.

"We hope to get the community thinking more about long-duration storage, which we'll need more of as we reach higher penetration of renewables onto the energy grid," Chiang says. "For example, there are seasonal variations, and we'll have to figure out how to deal with that. Up until now, electrochemical storage is not the first thing that people think about to accommodate that seasonal variation, just because the cost of it is so high."

The researchers plan to continue working to make their storage battery more efficient, drive down costs of the battery architecture, and increase its lifespan -- it can currently operate up to 1,500 hours, but that's far from the 5-20 year lifespan it would require in practice. They are also considering how best to scale their prototype and where to commercially test their product.