Rates vs. quantities - more unit confusion

It's widely accepted (though not universally) in the renewable energy/clean tech/green community that one of THE major problems in the widespread adoption of such renewable sources of electricity as solar and wind is their intermittent character. It's further believed that the ability to store the energy from these sources will enable their intermittency to be smoothed out, thus making them a reliable source of energy and enabling them to become much more easily integrated into the grid, possibly even a source of baseload power.

I read today that my state, California, has, through our Public Utilities Commission, set a goal of "1.3 gigawatts of energy storage by 2020." My state is certainly at the leading edge of sustainability with AB32, the Global Warming Solutions Act of 2006, the Cap and Trade Program, and many others. But I worry about people who make laws write regulations and yet aren't able to distinguish between rates and quantities.

Here's a link to the Assigned Commissioner's Ruling on this. Throughout the document, Carla Peterman discusses storage in megawatts. But a "watt" is a rate of energy utilization, or rate of performing work. A megawatt is a million joules per second. Certainly, the rate at which a storage system can deliver energy is important, but the key is the quantity that can be stored. This would be measured in watt hours, kilowatt hours, megawatt hours, gigawatt hours, terawatt hours, etc. Or, equivalently, in joules, megajoules, etc. Saying "we need 1.3 gigawatts of storage" is analogous to saying San Francisco is 80 miles per hour away from Los Angeles.

A typical gasoline pump will pump, conservatively, around five gallons per minute. Each gallon of gasoline has a chemical potential energy through oxidation of about 132 megajoules. Thus, when you fill your tank, you're delivering energy at the rate of 132*10^6 joules/gallon * 5 gallons per minute/60 seconds per minute or 11 megawatts. About 120 people filling their tanks are delivering energy at about the 1.3 gigawatts mentioned by the PUC. But this doesn't tell you a thing about how many miles these 120 drivers can travel. To know that, you need to know the capacity of the 120 tanks in gallons (along with the rate of fuel consumption of the vehicles).

To provide a bit of orientation as to the rates being discussed, in 2011, California generated or imported a total of about 292,454 gigawatt hours of electricity. This is a rate of about 33.36 gigawatts so the storage being discussed could deliver a bit under 4% of the average California rate of electricity usage.  Of course, the planned storage is divided between different utilities and geographical locations and would be deployed locally. But key to the discussion is FOR HOW LONG? A second? A minute? An hour? A day? Nothing in the document tells us.

Addendum, June 15, 2013: It's bad blog form to edit a posted blog without saying so. In reviewing the above, I realize that, while it's quite true that capacity is a fundamental metric of the viability and practicality of storage, rate is also important. I alluded to that above but I want to make it clear that I realize that vast capacity is not relevant if the rate at which it can be delivered isn't matched to the demand present in the area served.

This merits a bit of analysis of what sort of capacity might "match" a delivery rate of 1.3 gigawatts. For a starting point, let's take a look at the 4% calculated above. And I'll arbitrarily assume that we'd like to be able to "even out" the output from intermittent sources over a 24 hour period with a reserve capacity (no wind, cloud covered sky, whatever) for three days. In three days, on average, California might use (3/365)*29,2454 or about 2,400 gigawatt hours of electrical energy. 4% of this is 96 gigawatt hours.

Arguably, the storage method with the best combination of capacity, dispatchability, delivery rate, and efficiency is pumped hydro storage. If we assume a round trip efficiency of 75%, we'll need to store 128 gigawatt hours of energy. Hoover Dam delivered, at its 1984 peak, 10.348 terawatt hours so, in an average three day period in 1984, it delivered 10348*(3/365) = 85 gigawatt hours. So we're talking about something like one and a half Lake Mead/Hoover Dam storage schemes.

By the way, this goes to show just what an amazing resource gasoline is. 3032 gasoline pumps can deliver the energy equivalent to the entirety of California's average electrical consumption.

Update: The Blenheim-Gilboa Power Station in New York can deliver electricity at the rate of 1.6 gigawatts and, per a comment in a post on pumped hyrdo storage at one of my favorite sites, Do the Math, can deliver this for 16 hours for a total of 25.6 gigawatt hours. Another site says 1,000 megawatts for eight hours, or 8 gigawatt hours. In any case, storage of the magnitude specified is certainly achievable.

Also, as seen here, it's clear that pumped hydro plants are typically rated in rate in watts rather than capacity and, in fact, it's difficult to find the energy capacity in megawatt hours or gigawatt hours. In a comment in the article linked above, Ben K. say that this is because generating plants are rated in this way. But I don't see that this is a valid reason. A generating plant generates continuously as long as coal, natural gas, uranium, etc. is delivered - the amount of source material is not the issue. In a storage facility, the amount of source material (compressed air, battery chemical potential energy, water, flywheel rotational energy, capacitor electrical charge energy, whatever) is THE issue. After all, we look at battery storage capacity in terms of amp-hours which, at a fixed voltage, is a measure of quantity of energy.