PG&E to compress air to store wind power

Despite all the talk about needed breakthroughs in batteries, Pacific Gas & Electric is pursuing a less high-tech approach to store wind power: underground compressed air.

The utility on Wednesday said that it is seeking $25 million in smart-grid stimulus funds to build an underground compressed-air storage facility that would be able to deliver as much electricity as a medium-size power plant for about 10 hours.

(Credit: PG&E)

PG&E said the project is part of its smart-grid initiative and would take about five years to develop and build but, in a company blog post, didn't offer any other details on the proposal.

With compressed-air energy storage (CAES), air is compressed and then pumped in natural underground reservoirs. The air is released later and converted into electricity.

There are currently two compressed-air energy storage facilities in operation--one in Alabama and one in Germany--but the technique has been getting more attention because it is a relatively cheap approach to storage.

Utilities are starting to use flywheels to smooth out fluctuations on the grid or truck-size batteries to provide backup power for a couple of hours for a single substation.

CAES is well-suited to an intermittent source of energy like wind because a large amount of energy can be stored for many hours. PG&E's proposal calls for storing 300 megawatts worth of power for 10 hours, while most utility storage batteries being tested are 1 or 2 megawatts for shorter periods.

PG&E said that it plans to use wind turbines to compress the air during off-peak times and then draw from the reservoir during peak times. Shifting the energy from off-peak to peak times, such as the middle of the day, makes it more valuable as utilities pay more for energy at peak times. A wind farm in Iowa has been working on CAES storage for a few years to take advantage of peak pricing for wind.

PG&E quoted a Princeton University study on CAES that concluded that "CAES appears to have many of the characteristics necessary to transform wind into a mainstay of global electricity generation."

Industry executives say that the most cost-effective utility storage is pumped hydro, where water is pumped uphill and released at peak times to make electricity.

This technique, which has been around for decades, is tough to beat on cost. But like compressed-air storage, it requires that utilities find the suitable geography.

Updated at 12:45 p.m. PT to clarify the capacity and energy storage of the facility.

[Perry_Clease]

Sounds like the makings of a cheesy science fiction movie. The compressed air cracks the fault lines. The eastern part of the United State sinks into the Atlantic Ocean and California becomes an island. :)

Seriously this is probably an excellent idea. Ofttimes I am driving by the wind turbine farms near Altamont, or Tehachapi when there is a strong wind blowing, but only a few turbines turning.

[Hunnter2k3]

This is actually one of my concerns with the idea.
Compressed air under the ground doesn't sound that safe to me. (especially when it is coming OUT...)

And i wonder if it would actually be more efficient to have containers used for this instead.
You can punch the ground and displace it pretty easily, punch a metal container and you're likely going to need a plaster.
Metal containers will almost certainly be able to contain significantly higher amounts of air in a smaller space, therefore more air in the same space as the "cavern" with more containers.
You could probably turn those large generic storage containers in to air storage with some welding and reinforcement.

[timber2005]

Metal containers can contain a higher pressure (and therefore volume) in a smaller dimensions...
However, large pressure vessels cannot be made in large sizes to store the same pressures.
AKA, a small tank can store high pressures, but make it larger and it can't store to the same pressure.

What you need to think about in your fear of air leaking out, is that these cavers are sometimes many stories below the surface, and very little ground water can even penetrate to those levels. Most water comes up from below the caverns. If water cannot penetrate down, air won't be able to penetrate up, let alone explode out.

[Mr_MechaLizard]

Arent there more efficient ways to store power?
I suppose the cheapness of building and running it out weigh the cost of low efficiency?
Is it meant to say 'truck size batteries' ? (can i have one?)

[contentcreator--2008]

Energy storage is measured in megawatt-hours not megawatts (which is appropriate only for power, ie energy rate of use)

[contentcreator--2008]

Just to amplify --- this is a pretty important point and distinction between different systems.

A battery system might be 1 megawatt (in/out) but only 0.5 megawatt-hours, ie 1/2 hr runtime

A compressed air system might be 30 megawatt (in/out) and 300 megawatt-hours, ie a 10 hour runtime

[mlamonica]

Yes, I understand the difference between watts and watt-hours. The facility will have a capacity of 300 megawatts and be able to deliver that for 10 hours. I've updated to be more clear there. As reader below notes, that equates to total storage of 3,000 megawatt-hours.

[contentcreator--2008]

OK, thanks, I had expected it to go the other way; it's quite a bit of energy then.

Fun with Wolfram Alpha : 3000 MW-hr ~= 2.6 kilotons of TNT
Wikipedia: yield of Hiroshima bomb: ~15 kilotons

Hmmm.

[contentcreator--2008]

OK, thanks, it's quite a bit of energy then:

Wolfram Alpha: 3000 MW-hr ~= 2.6 kiloton of TNT
Wikipedia: yield of Hiroshima bomb: ~15 kiloton

Hmmm.

[TommyKPdx]

I do not understand why writers about the environment almost invariably get their units wrong when it comes to electricity. Anyway, according to http://www.bizjournals.com/sanfrancisco/stories/2009/08/24/daily31.html:

"PG&E said its project would store the equivalent power of that produced by a 300 megawatt power plant for up to 10 hours."

So, it's a 3,000 MWh storage facility. Please update the story accordingly.

[NocturnalCT]

It's still not quite right.

"CAES is well-suited to an intermittent source of energy like wind because a large amount of ENERGY can be stored for many hours."

This would indicate that there is energy leakage. As in "we can store 20 MWh worth of compressed air but it self-empties at about 0.1 MWh per day" or something like that. Assuming the system in the story doesn't leak nor degrade the quoted passage makes little sense. It should simply say 'because a large amount of energy can be stored.' Then the other side of the story is how quickly this energy can be released. It makes little sense to store many MWh worth of power if it trickles out a few KW at a time. Divide capacity (energy) by flow rate and you get the amount it takes to drain the 'battery'.

In other words it's a combination of the size of the bucket (energy) and the hole that's poked into it (flow rate). Both are important. I'm sure Martin understands this based on his other comments but the article doesn't demonstrate it in the quoted passage.

I wonder what type of conversion efficiencies they get generating electricity to compress air and then converting that air into electricity again. Ideally the wind mill drives a compressor directly but maybe a large compressor is more efficient than a bunch of small ones.

Anyway, this is interesting stuff. Thanks for posting and sorry about the nit picking.

[moviegeek65]

This technology isn't really green because it's taking off-peak electricity from the grid to produce compressed air, wouldn't it be more cost efficient to just build more wind turbines?

[TommyKPdx]

The problem with wind turbines is that sometimes the wind doesn't blow very much. At that point it doesn't matter how many wind turbines you have, you can't meet the demand. The compressed air system provides the backup for those times. In principle if you had enough storage capacity and enough turbines, you could build an entire grid from wind turbines and energy storage alone. When there was more supply than demand (e.g. at night), you'd store the excess energy. When there was more demand than supply (e.g. on hot windless days), you'd supplement the wind-generated energy with the stored energy.

Even if all power generation is by fossil fuel, however, having storage available reduces the need to build more power stations as demand increases, because peak demand can be met by a combination of immediate generation and stored energy. (You store the energy when demand is low.) This in itself is a 'greener' solution than no storage at all. When coupled with renewable generation, it becomes even greener, even if there is still fossil fuel generation on the grid (as there will be for a long time to come).

[Seaspray0]

@tommykpdx. True, you can't always count on the wind blowing... but there are areas where that hardly ever happens (i.e. 90% wind flow all the time). If you find 3 locations in 3 different areas of the country like this, then the chances of all 3 not having the wind blow is extremely small. Planners are aware that they will need multiple locations accross the country to minimize the effects of "no wind" in some of the locations at any given time. When you scale wind power accross the entire country, the effects of the wind not blowing becomes somewhat insignificant to maintaining reliability.

[timber2005]

@SeaSpray0 The power grids do not cross certain boundaries, and therefore a nationwide grid won't work.
There are 3 independant power grids in the US.

[hawkeyeaz1]

And if you store heat from the sun, (if necessary) you can heat the air as it is pulled out and increase the volume more.

[solar_thermal]

While I feel that large long term energy storage is the single biggest obstacle for renewable energy, I have my doubts about this one. I looked at some of the projects already in existence and they seem to be claiming only 50-55% efficiency. I don?t know what can be done to improve this but this is horrible. Pumped hydro was mentioned and some of the new systems claim to be able to achieve over 80%. It can be very expensive to build new projects but can be piggybacked on existing dams. I know of one project in Southern California and Lake Oroville Dam uses this technology also. Also this technology has been used for over 100 years. Like I said before I feel long term energy storage is very important but throwing away 50% of your already expensive electricity does not seem like the best way to go.

[carlhage]

Compressed air has been around for a while and is proposed as a low-cost large-scale storage. The real problem with efficiency is that when the air is compressed, it gets hot, and that energy is dissipated into the air (lost)-- at least in all the plants currently built. When the air is expanded, it cools and needs to be reheated in the gas turbine. In effect it's a giant heat pump. There has been plans to try to store the heat extracted after compression, then use that to reheat the expanded air, but according to the Wikipedia page on CAES no unit has been built with this. All units are attached to gas turbines.

The NaS batteries available with higher efficiency are around $300/kWh. I don't know how much the total cost would be over the grant, but $25M/3,000,000kWh is only $8/kWh. (I saw a figure of $800/kW, so a 300MW plant could be $250M ($25/kWh), or around 10X cheaper than batteries.

[esbuck]

Compressed air storage is at present inefficient because the compressors need to be cooled. Much higher efficiency is possible if that waste heat is saved and used to reheat the compressed air. See U.S.Patent number 5,832,728, which has expired. (I proposed compressed air storge to PG&E a few years ago, but they weren't interested, the not-invented-here syndrome, I suppose) It would seem simpler to use the turbines to compress air directly, instead of generating electricity to run the compressors, but, hey, if you are an electric utility.... (like if you are a hammer salesman, everything looks like a nail and you deny the existence of screws) Anyway, it's a good idea, and you don't need an elevated reservoir. If they have any old coal-fired plants they would like to retire for environmental reasons, why not keep the switch gear, turbines, etc. and power them with compressed air from a wind farm?

[solar_thermal]

Esbuck I am not quite sure how to take your post. You say you proposed this to PG&E before? Well PG&E does not really have a large research and development department so unless you have it pretty much ready to go then I would not expect much more.
Ok about making the air directly. Yes it would be more efficient than making electricity, then air and then electricity again but less efficient than just making electricity. Much of what you propose would be very difficult to enact. As far as using an old coal plant here are a couple of problems. First off rarely would you find one close enough to pipe air from a wind farm. Turbine would be a steam one and not exactly what you would need for compressed air. Really about the only thing you could use would be the switchgear and transformers.

[david_harro]

I find this very strange............ to regenerate electricity from this system, the air needs to be reheated by burning natural gas. That the drawing and the information in this article doesn't mention that????

It uses 1/3 of the gas to produce the same amount of electricity a gas power plant uses for the same output. In the case of wind power, the electricity is no longer green. In the case of off peak its 4/3 the amount of gas to provide the electricity when we need it most (standard fossil fuel production + 1/3 gas to return it to electricity). Just because it might be cost effective , it doesnt mean that its good for the environment.