A California start-up named Advanced Rail Energy Storage (ARES) has a clever idea for storing electrical power at the most extreme scales, e.g. those of the power grid itself. It's a battery of sorts, but a battery that doesn't resemble anything we'd normally associate with the term. Rather than chemicals, ARES stores power as gravity.
The key components of ARES storage scheme include a really big hill and a few railroad cars. Energy to be stored in the system is first used to pull the rail cars via electric locomotives to the top of the hill, where it persists as potential energy. So long as the cars are at the top of the hill, the initial energy expended to get them up there remains trapped within the system. To recover that energy, the cars are simply lowered down the hill, turning built-in motor-generators in the process. This power is collected and then returned to the grid.
So, energy goes in and then it comes back out. Up and down. Earlier this month, ARES won approval from the Bureau of Land Management for a lease in southern Nevada featuring some train tracks on a hill and connectivity to the local power provider, which in turn will provide ARES with a means to connect to the greater western US power grid.
A fair question to ask is what's the point? We're not actually generating any power here and are most likely losing it, at least to some extent. But generation isn't the goal. What ARES is after is grid stabilization at very large scales (up to 50 megawatts with the Nevada project)—as the grid becomes more diversified, smoothing its peaks and valleys becomes a more and more vital task. Sometimes it's cloudy, and sometimes the wind doesn't blow.
So, if on some day the wind was really blowing and wind farms were really kicking the power out, ARES locomotives would automatically kick in, pulling several rail cars to the top of the hill. Should the grid dip, the cars would then be lowered back down the hill, returning about as much power to the grid as they removed.
The 50 megawatts of the Nevada facility is not huge—enough to power 15,000 or so homes for an hour—but ARES imagines large regional facilities capable of storing 2 to 3 gigawatts. The scalability of a such a system seems intuitive enough: more tracks, more cars, more grid.
Of course, one might wonder what advantage there is to doing this with train cars vs. literally any other thing that can be raised or powered, such as water (as in pumped-storage hydropower). ARES answer is efficiency: the system is able to recover 80 percent of the power that it takes in. That might not seem too impressive, but it's better than most car batteries, at least. And then part of the answer is the aforementioned scalability. This is a limiting factor in water-based storage schemes in that it's challenging to find workable sites and water supplies.
Once ARES' final environmental compliance reports are in, construction of the Nevada facility should take about eight months.