Like most grid-scale battery systems, EnerVault’s battery will be as large as shipping containers. Its first demonstration system, which is still being built, will supply 250 kilowatts for four hours. Credit: EnerVault.
Flow batteries have emerged as one of most promising ways to store many hours of energy on the electricity grid. To make costs more competitive, startup EnerVault is pursuing a novel chemistry and unique mechanical design.
The Silicon Valley-based company plans to install its first demonstration this year which will be connected to a solar array in California, says Bret Adams, EnerVault’s director of business development. During the course of this year, it plans to test a number of its systems for different uses to measure performance and costs. About half of the $9.5 million expense of EnerVault’s first installation and the product development are being funded by the 2009 stimulus fund.
EnerVault is one of many companies developing alternatives to more common grid-level batteries. Flow batteries are well suited to renewable energy because they can store many hours of energy and the storage capacity can be expanded independently from the power rating. Experts say there are dozens of flow batteries installed on the grid, either smoothing the flow of wind and solar farms or providing power during peak times to lower electricity bills. But as with all grid storage technologies other than pumped hydro and compressed air storage, the costs need to come down. (See, Ambri’s Better Grid Battery.)
Flow batteries use two big tanks of liquid electrolytes, which are circulated several times through a vessel where an electrochemical reaction takes place across a membrane. When connected to a load, a current is produced when electrons move from negative electrolyte to the postive. During recharge, a current is applied to reverse the reaction. Flow batteries are generally considered safe, an important issue for grid-scale batteries where thermal runaway of conventional batteries has caused fires at least two cases.
Founded in 2008, EnerVault has developed a new electrolyte pumping system to improve the efficiency of the charge and discharge cycle. It’s also using an iron chromium chemistry, materials that are one sixth the cost of the vanadium used in some flow batteries.
In a typical flow battery, electrolytes flow into a “stack” where the reaction takes place and then are pumped back into the holding tank. As an electrolyte flows out of the stack and back into holding tank, the state of charge decreases, which reduces the amount of material available to produce electricity, Adams explains. In the EnverVault design, electrolytes flow through through a “cascade” of cell stacks, each holding electrolyte with a progressively lower state of charge. By optimizing the membrane for a particular state of charge, the EnerVault battery can increase the energy density compared to a conventional flow battery, he says.
Like many new grid-storage companies, EnerVault is still seeking where its multi-hour storage systems are best suited. Multi-megawatt-hour batteries could provide solar or wind power at peak times to earn energy project owners more money. Or they could be used to expand the capacity of substations, allowing utilities to defer upgrades to meet peak demand. Some companies hope to provide local storage to campuses or businesses that have on-site generation.
Still, grid-scale batteries of all kinds have years to go before a commercial demand emerges. On a simple cost basis, they compete with the price of power from a natural gas peaker plant which, in the U.S., is much lower the price of a battery. And for the most part, utility regulations are geared more towards construction of power plants rather than energy storage.
It’s too early to project the actual cost and reliability of EnerVault’s battery but Adams says using low-cost electrolyte material and a low-pressure flow system is one way to compete with other technologies. “We started with readily available commodities—iron and chromium. Even if we were massively successful, we’d be using a very small potential piece of that supply,” he says.
This post was updated on Monday March 25 with clarifications.