Grid-scale Storage Summary
As discussed in my previous post, GW of grid-scale storage are needed to work with wind and solar generation.
As EV drivers, we think batteries first, but we know how expensive batteries are. They have high efficiency, fast response, but limited lifetimes. California currently has 50K+ BEVs and 40K PEVs. If we assume that each car could supply an average of 10kWh at 10kW, 1 hour, we would have just short of 1 GW. Realistically few owners would be willing to subject their cars to 10kWh of daily cycling, but when their batteries are retired from their vehicles, we might have .6 GW of available grid storage.
Nationally, just several years ago, 99% of grid-scale storage was pumped hydro, where you pump water uphill to store energy and run it downhill to release it. You need two reservoirs close together separated by substantial altitude difference. There had been no new pumped hydro projects initiated for 20 years.
The efficiency of pumped storage is typically 80%, but new proposals claim 90+% is achievable.
The simplest way to use hydro storage like a battery is to modulate the flow of water through a dam, higher flow when demand and the price is high, lower or no flow off-peak when the price is low. The efficiency of this strategy is 100% and the only cost is the impact of variable flow on downstream uses of the water, so this has been common practice at many dams for a century or more.
Pumped hydro projects have two versions: Open Loop, and Closed Loop. Open Loop operates in contact with an existing water source, such as a reservoir and dam. To add pumped storage you provide a lower reservoir, pump, and pipe back up to the original higher reservoir. The cost is low, water is available, and the additional evaporative loss can be minimal. Closed Loop operates without an associated water source. This means two reservoirs have to be built, water purchased, and evaporative loss is now an additional concern. The addition of open water to an area that did not have it can lead to ecological concerns.
With the dramatic growth of wind generation in the midwest, and growth of solar in California and other states, FERC has this Fall 2014 map of new pumped storage projects, including almost 10 GW for California:
In addition, other technologies are now being considered for new grid-scale storage.
CAES, compressed air energy storage, uses large underground caverns such as abandoned mines. It has suffered from a relatively low efficiency, because the compression usually is isothermal, not adiabatic, so heat is lost between the compression and subsequent expansion. New technology promises to reduce this loss.
A 20MW flywheel project became operational in 2014 in PA. It is used for short-term, fast-response frequency regulation:
In the bottom image each blue dot is the top of a flywheel unit, .1MW each. They are installed below ground level for safety in case one explodes.