http://www.caiso.com/docs/2003/09/08/2003090815135425649.pdf
The poster isn't saying they're producing energy, just that they must be ready to do so at any time.
The cheapest storage battery is simple lead acid figure a quarter per watthour installed. Or a KWh is $250. But with a 100% charge/discharge cycle the battery will be dead and need replacing in about 10 or so cycles. To get up to 1000 or so cycles (which is only 3 yrs daily cycles) means you only get to use about 10% of the capacity. Lets round up because rectifier/inverter gear, and buildings, and operators and their stuff, are not free. So you can guess about $3/watt of storage as an absolute best case, but probably more realistically a turn key battery storage facility would cost more like $4 and would depreciate fully in about half a decade.
Hydro turnkey is about $1/watt plus or minus massive corruption (what is the dollar value of the hetch hetchy valley of yosemite national park, etc?) Coal plants sell turnkey for a bit over $2/watt, natgas is arguably the most expensive around $6/watt. All of those last like 50 years, so divide those costs by about 10 to compare with a battery bank that only lasts at best 5 years.
Spinning capacity (well, sorta, in case of natgas) is around 10 times cheaper per KWH than batteries. You have to remember that power companies don't really care about pushing an agenda, more or less. There is no conspiracy, they just want to sell KWH. If they could install a battery bank instead of a natgas peaking plant, and keep huge profits, they most certainly would.
There are some interesting math problems too. If each stored utility grade battery costs $2500/KHW and the total overall worldwide battery industry is about $50B, that means if we abandon all other forms of battery use in the world and get rid of all laptops, cellphones, etc, we could build nothing but lead acid batteries at a pitiful rate of ... drum roll ... 20 megawatt hours of utility grade storage per year. Now since the batteries are scrap in 5 years, that means if in a Manhattan style worldwide project we focus the entire industry on utility grade storage, we can never store more than 100 megawatt-hours worldwide. Which if you assume a daily charge discharge cycle is about 10 MW continuous or about the capacity of ONE small gas turbine system. So its not as simple as going down to "batteries plus" and picking up a battery large enough to UPS a nuke plant.
http://juneauempire.com/state/2011-04-17/fairbanks-battery-b...
http://www.gvea.com/energy/bess
They chose Ni-Cad (presumably for good reason).
Edit: Some numbers:
It's about 5 megawatt hours (varies somewhat based on draw, could stretch it to 6.5 MW-h for 15 minute runtime, less at higher draws).
Cost $35 million, so ~$7000 per KW-h of capacity.
Planning authorized in 1993, online in 2003.
Expected battery life of 20 years (maybe 30).
So a project that didn't seem to take up the entire manufacturing capacity of the battery industry managed to bring up 0.5 megawatt hours per year, with a lifetime of 20 years, half of your prediction of capacity production capacity (but at much higher cost than you started from).
Instead we keep hearing the old "Wind and solar will never work because base load" garbage.
http://library.abb.com/GLOBAL/SCOT/scot271.nsf/VerityDisplay...
"An economically and ecologically more viable alternative to ‘spinning reserve’ – gas turbines kept running in case of an emergency – is battery back-up."