The two fastest growing sources of renewable energy, wind and solar, are intermittent—they don’t always generate power when you need them to. The obvious solution is to add storage, like batteries, to allow you to shift some of the electricity to when demand is highest. Elon Musk is betting a Gigafactory that consumers are going to be interested in doing this, while California has mandated that 1.3 gigawatts of storage be added to the power grid before the decade is out.

But there are a number of different types of storage, each of which has distinctive properties: how fast electrons can be shuffled in and out, how easy it is to expand the storage capacity, and so on. All of these have different costs, and figuring out what storage is most economically viable is a serious challenge.

Three academics from MIT have decided to take up that challenge. They’ve tried to calculate when it makes economic sense to add storage to renewable projects in three different locations in the US. Their analysis indicates the finances are close right now, but only for things other than batteries.