February 04, 2017
Utility Battery Projects Driven By Price Drops
For large electric power storage projects the cost of batteries has plummeted.
2008, when battery prices were 10 times higher than they are today.
This advance is timely as photovoltaic electric power prices have dropped so far that in SoCal PV is causing a growing drop in mid-day demand and therefore a much bigger spike in evening demand. Therefore there's a growing need for a cheaper way to store power generated in mid day and deliver it in the evening. You can see how much solar power output surges each day in California by clicking on some of the Daily Renewables Watch links at the Cal ISO site (the organization that manages California's electric grid).
The growing supply of wind and especially solar power is causing a huge decrease in demand for non-renewable energy in mid-day in California. This is known as the renewable duck curve and it will place a limit on the growth of solar power unless battery costs fall far enough to enable use of solar power during evenings.
Randall Parker, 2017 February 04 12:40 PM
Or they could just abolish net metering, and let the people who generate excess power the utility doesn't want eat it themselves, instead of forcing the utilities to find some way to handle it.
But for the grid, Lithium is not needed: a stadium-size lead battery in the city would be very cheap and also easy to recycle.
Tesla is marketing relatively expensive Lithium batteries even for residential storage, but this is because it produces this kind of batteries at its factory.
The cost figures that Tesla is claiming are below anything I can recall seeing for Sadoway, whose batteries were supposedly made from dirt. Or maybe I have this confused.
ISTR a zinc-air effort which came out at $160/kWh or thereabouts. Not sure if it has delivered yet.
There are so many gaps here. Batteries which can back things up until gas-fired turbines can fire up only need to run for 15-25 minutes. Batteries which can store noontime's solar energy until evening only need to work for a few hours.  Batteries which can back up unreliable RE until the average comeback need to last for most of a WEEK... about 20x the cost of noon/evening shifts so need to be about 1/20 of the cost. Nothing I have ever seen can get close to the cost figures for that. Sane people would write RE/batteries off. Apparently, no one is sane.
Flow batteries would probably be more practical still for the grid. But most practical of all is, as I said, repealing net metering, so the utilities don't need to find some way to handle power they don't want. It's not like they have a crying need for intermittently available power supplied at the whim of their customers.
Utility scale batteries are a solution looking for a problem.
Utility-scale batteries are a viable form of spinning reserve, able to respond in sub-cycle timescales to plant outages or other disruptions. 15 minutes of battery power is enough time to cold-start many types of peaking/reserve generators and get them on-line, and the savings in fuel helps pay for the batteries.
We'd be better off making batteries do double or triple duty, of course. Making EV chargers into interruptible loads co-opts them as reserves too, spinning reserves if they can act fast enough. But time-shifting output from flaky supplies? Holding power for a week requires battery costs in the <$10/kWh range. Tesla's PowerPack is currently over $400/kWh. The cost of materials alone won't let Li-ion fall below about $80/kWh.
I still think that railroad cars full of rocks,run up the side of a hill would provide a cheap,viable source of energy storage.
Railroad cars full of anything are very expensive compared to water.
The big users of energy, like petroleum refining, and the chemical industries, could use all this solar power.