Underground 'Wind Mines' Could Keep Datacenters Powered 109
Nerval's Lobster writes "Major IT vendors have been including custom-built wind- and solar-power farms in their datacenter construction plans. But while wind and solar power may be clean, they're often unreliable, especially by the standards of datacenters that need a way to keep operating through any unexpected surges or drops in power. How about saving the wind that generates the power? That might work, according to researchers at the federal Bonneville Power Administration (BPA), and U.S. Department of Energy's Pacific Northwest National Laboratory. A study published in February (PDF) outlined the potential benefit of pumping pressurized air into caverns deep underground as a way to store wind energy, then letting it out whenever demand spikes, or the wind drops, and the above-ground facilities need help spinning enough turbines to keep power levels steady. The technique, called Compressed Air Energy Storage (CAES) isn't new: existing CAES plants in Alabama and Huntorf, Germany (built in 1991 and 1978, respectively) store compressed air in underground salt caverns hollowed out by solution mining (pumping salt-saturated water out of concentrations of salt buried far underground and replacing it with fresh water). But implementing such a technique for datacenters might take a little work. The BPA and the Pacific Northwest National Laboratory have already identified, and are evaluating, sites in the Pacific Northwest that would be suitable for CAES underground reservoirs; the first, which could be located in Washington's Columbia Hills could—via existing CAES technology—store enough compressed air to generate a steady 207MW for 40 days of continuous usage, ultimately delivering 400 additional hours without adding any compressed air."
A bit confused. (Score:2)
Wouldn't it take a buttload more power to move the air down, and then back up, than it would generate?
Re: (Score:2)
Quite many buttloads, I'd presume.
Re:A bit confused. (Score:5, Funny)
Quite many buttloads, I'd presume.
Fortunately a buttload of pressurized air doesn't actually have much force behind it.
Source: personal experience.
Re: A bit confused. (Score:1)
Just need to eat more Mexican food
Re: (Score:3)
Except the warm surface air you pumped down underground will cool, PV==nRT resulting in less volume and less pressure, so while the air may be under great pressure when you send it down however it will loose pressure and there for potential energy over time.
entropy is a bitch.
Re: (Score:3)
Ultimately I can even see the expulsion of the air in the reservoir passing through heat collection systems in the datacenter to cause ex
Re: (Score:2)
Easiest solution is to install an intercooler that will chill the incoming air before compression, or reheat the outgoing air when it is being released. A big heatpipe sunk into a significant thermal sink (the ground) would do this fine.
Re: (Score:2)
Re:A bit confused. (Score:4, Insightful)
Wouldn't it take a buttload more power to move the air down, and then back up, than it would generate?
I think it presumes excess power during some periods. High winds, excess hydro power, what ever.
They do this at Grand Coulee Dam [wikipedia.org] already by pumping water up-hill to an additional reservoir in periods of excess runoff when they would otherwise have to open the spillways just to get rid of the excess.
Pumping water uphill is probably far more efficient than compressing air.
Re: (Score:3)
another example:
http://en.wikipedia.org/wiki/Salina_Pumped_Storage_Project [wikipedia.org]
I believe this is where Google has a datacenter that consumes hydroelectric power.
Re: (Score:2)
There's an entire category of these [wikipedia.org] on the Wikipedia. One of them isn't too far from where my parents live. It's a great way for handling variable demand when most of your nearby plants run a constant baseload.
Re: (Score:2)
The difference is pumped water doesn't have to deal with PV=nRT. Most are between 70-80% efficient.
Re: (Score:1)
sure. not to mention the atrocious efficiency of wind power. Just pump water to a reservoir instead and let it out when you need it.
Re: (Score:2)
Nope. (Score:3)
Where's the water going to go?
Re: (Score:2)
Re: (Score:2)
What are you more likely to have handy, a hill or a massive abandoned salt mine?
Re: (Score:2)
Re: (Score:2)
Not just studies but it's been working for decades -
https://en.wikipedia.org/wiki/Dinorwig_Power_Station [wikipedia.org]
Re:A bit confused. (Score:4, Insightful)
sure. not to mention the atrocious efficiency of wind power. Just pump water to a reservoir instead and let it out when you need it.
I think that the article is trying to be clever, but missing an important point in doing so. Energy is generated from the compressed air using a more conventional turbine/generator setup... not a wind farm. This system is just an ENORMOUS UPS.
Re:A bit confused. (Score:4, Insightful)
Wouldn't it take a buttload more power to move the air down, and then back up, than it would generate?
Any flavor of energy storage is going to introduce some sort of conversion losses: battery banks aren't 100% efficient to charge or discharge, flywheels suffer from friction losses, pumped-water hydro suffers from inefficiency in the pumping uphill and the conversion to electricity downhill backup generators suffer from the fact that small heat engines generally get lousy efficiency compared to big ones(and need to be kept supplied with diesel, which doesn't help you 'green' cred).
The advantage to pressurised air is that(in geologically suitable locations) you can build in fairly large amounts of storage without anything obtrusive on the surface, and at comparatively low cost(compared to buying and keeping fresh huge banks of batteries, say).
Re: (Score:2)
Now that is a heroically inefficient process, at least with current technology...
Re: (Score:3)
Re: (Score:3)
Nah it goes down due to gravity see...
Seriously though, energy efficiency is not the problem when you are producing far more energy than demand. As long as it makes sense and pays the cost/maintenance on the pumps/storage site, it's a workable idea.
Re: (Score:2)
it's a storage medium (Score:2)
Wouldn't it take a buttload more power to move the air down, and then back up, than it would generate?
it doesn't generate power. it's for storing the power.
it's just a method you could use if for some reason you couldn't use some artificial lake or other efficient method for storing the energy. and if you happen to have a suitable mine handily available.
as long as you're constantly burning coal plants in the the area nearby that's near enough for transmission losses though it gives no actual benefit, could just feed the extra power to the grid and then take the coal(or whatever, but coal and gas is more on/
Re: (Score:2)
Works for cars. Again, slashdot the home of the "I can't think of it, so it must be impossible" or in this case "My (broken) common sense doesn't understand this, so the piles of professionals working on this must have gotten it wrong."
We feed our Sys Admins a steady diet of Chipotle (Score:5, Funny)
burritos. Then we have them run over to the turbines and release any stored energy. The department of energy calls it Flatulence Assisted Regeneration Technology. Doesn't work quite as well as pumping the earth's crust up like a baloon though.
Re: (Score:2)
Or you could use rivers and dams for storage (Score:1)
Oh, I'm sorry, I forgot that that form of clean energy fell out of favor with the hippie set because it keeps the salmon from spawning or some shit.
Re: (Score:2)
Re: (Score:2)
Not to mention the "turbines are ugly" crowd. Yeah - well so are coal power and nuclear power. Suck it up - or live in the stone age MOFO!
Re: (Score:2)
Ugly, check. In our area, most of them are built upon previously protected natural areas that were pretty idyllic before.
Noisy, check. The biggest ones can be heard and felt from more than two kilometers downwind.
Pollutes, check. Each of the turbines is going through 20 gallons or so of grease a month, a lot of which is being released in main wind direction and contaminating the ground.
Crazy subsidies. Here, we pay the producer of wind energy about 12c/kwh, while we are paying less than 4c/kwh for the hydro
Re: (Score:1)
There aren't many locations where pumped storage is feasible - you need a big hill with a lake on top and either another lake or a river at the bottom. When it works its amazing. There is a pumped storage facility in Massachusetts that can dial up 1,000 MWe in about 15 minutes.
On the other hand, the Columbia River (where the Pacific Northwest Lab and Bonneville are located is an amazing pumped storage facility itself. Lots of wind around the Columbia Gorge and many dams with big reservoirs on the river.
Re: (Score:1)
The reason existing dams span rivers was to harvest energy from the river flowing down elevation.
Re: (Score:3)
So just to be clear, you don't care about salmon? Because to me, they are very tasty, and I want them to not be extinct because then I cannot eat them.
Re:Or you could use rivers and dams for storage (Score:4, Informative)
Re: (Score:2)
If the (wind|water|nuke) energy is just going to go unharvested/unused due to low demand during high or peak production, it's much less of an issue.
If you lose 20% of something that you were throwing away anyway, you're still ahead of the game.
I did see they are trying out (or by now using) heat exchangers to help lessen the thermodynamic losses, storing the heat as well as the air.
One plan even used some natural gas heaters during the generation phase.
Re: (Score:2)
They should be using it to help provide additional cooling to the data center. So, if you have to switch over to power generated by the release of compressed air, you should be able to divert the lower temperature air to the data center to lower your cooling need and decrease the overall power draw. Of course, this all sounds good, but the actual implemention is surely fraught with issues.
Re: (Score:3)
If you use an air-water mixture (Score:4, Interesting)
Re: (Score:2)
I'm not quite sure I follow, they say this is grid scale, is it like pumped storage hydro on steroids?
Re: (Score:1)
207 megawatt-hours times 24 hours/day times 40 days represents the storage of over 7 petajoules of energy. By comparison, a one megaton hydrogen bomb releases about 4 petajoules. That means this chamber will be continuously containing the equivalent energy of a city-leveling nuclear blast. It better be really big, because it's going to be really stressed. It's not like salt is nature's pre-made engineered compressed gas cylinder.
And what happens if the salt chamber ruptures and the air finds its way to
Re: (Score:2)
Old underground Mines aren't that big, (or that common), but there would clearly be some local potential for blow-out.
The first thing that would happen is the compressed air would start dewatering the mine. (Almost all abandoned or un-pumped mines fill with water in short order). That might force a lot of polluted water back into the aquifers.
Gives quite a new meaning (Score:2)
"wind mines"? (Score:3)
Calling a chamber to store energy as pressurized air a "wind mine" is like calling the fuel tank in my car a "gasoline well".
Re: (Score:2)
The media likes to keep words down to four characters and no more than two syllables of possible.
And the air comes out cold (Score:2)
OK, this could be really interesting. One of the problems of compressed air storage is that the air is very cold when it is decompressed. You generally need to burn a little natural gas to warm it up a bit before hitting the turbines.
Now we put this at a data center that has enormous cooling requirements....
This could turn out very well.
Re: (Score:2)
Bring on the earthquakes (Score:3)
207MW * 40 days =
207MW * 3456000 s =
715392000MJ =
7.15392*10^11J
This is roughly equivalent to 170 megatons of TNT. 1.7 times the size of the maximum theoretical yield of the Tsar Bomba.
Probably more than enough to start an earthquake in an area that is susceptible (such as the pacific northwest).
Re: (Score:2)
You can't compare static air-pressure contained in an underground mine with an atomic explosion.
Air doesn't explode. Even if a breach is formed in the mine, the air would escape (violently), but it doesn't blow up.
Besides, There was a reason the world went to underground nuclear warhead testing. The ground could contain it.
Re: (Score:2)
Energy is energy. The difference between an explosive detonation and compressed air being released from a ruptured container is the amount of time it takes for it to escape. An atomic warhead lets it all out within milliseconds (the historic Trinity test showed a 300 meter fireball at 25 milliseconds, and that was estimated to be 84 terajoules.) A ruptured pressurized tank would let the energy out much more slowly than a detonation, but still very fast. And we're talking about petajoules of energy.
Let's
Re: (Score:2)
Let's say there was a crack that led to the surface, and the air started escaping. It would quickly erode the walls of the crack, opening a progressively larger hole.
No. It wouldn't.
And if it did, the larger hole would have LOWER velocity, not higher. You don't have constant pressure from an inexhaustible supply at work here.
You have a finite supply, and any significant leak would quickly drop the pressure in the entire reservoir.
It takes thousands of years for wind to carve rock. Hundreds of thousands of years. The initial (natural) pressure of these reservoirs is something around 1196 PSI. (Source is the PDF in the second link in the story). The injected air is
Re: (Score:3)
You had some decent logic up to that last part about earthquakes where you just started making shit up.
Why Datacenters? (Score:2)
Underground 'Wind Mines' Could Keep Electrical Grid Powered
Or is this electricity somehow only usable by data centers? In the same vein, couldn't you also store excess production from nuclear plants or coal generation this way?
Is electricity no longer fungible?
Re: (Score:2)
It wouldn't appeal to Slashdot it you couldn't some how tie computers into it.
But Data Centers can be constructed where power is cheap. (In fact that is the principal site selection consideration). Cities: not so much. There really aren't enough mines left laying around for this to be a practical solution.
Famous last words (Score:3)
"Hey look, Bob, I found a new cave! All we have to do is move those rocks out of the way and..."
since you can not wrap your head around this (Score:2)
you are all thinking about generating electricity and after that making compressed air with the electricity from the wind mill.
instead, generate compressed air directly at the wind mill and store it for later use. you can also use this stored compressed air for your car or bus or whatever.
when you need electricity you run it through a compressed air driven generator.
don't forget that adding heat to the pipe will making more energy available as well.
Or stored water. (Score:3)
The Bath County Pumped Storage Station [wikipedia.org] in Virginia is a pumped storage hydroelectric power plant with a generation capacity of 3,003 MW:
Water is released from the upper reservoir during periods of high demand and is used to generate electricity. What makes this different from other hydroelectric dams is that during times of low demand, power is taken from coal, nuclear, and other power plants and is used to pump water from the lower to the upper reservoir. Although this plant uses more power than it generates, it allows these other plants to operate at close to peak efficiency for an overall cost savings.
Volcanoes (Score:2)
The Columbia Hills are part of the Columbia River Basalt Group, which were flood basalts, not volcanic eruptions, contrary to what TFA states. I was always under the impression that basalt was porous but almost entirely lacking in permeability, too. The fact that CAES is utilized in hollowed out salt caverns elsewhere would suggest to me that you'd need quite a void to make use of compressed air, and you'd need to frack the crap out of basalt to obtain similar volumes.
Not that there isn't permeability, bu
Re: (Score:2)
Boardman is a town mentioned in TFA, south of the Columbia Hills, I should add.
Widow Mines OP (Score:2)
Re: (Score:2)
If the power is just going to go to waste anyways it would defiantly be worthwhile. But compressed air energy storage is pretty inefficient. If you could do it, pumped water storage is much more efficient, but even then your looking at some significant mechanical losses. Its always better to directly use the power if possible but again, if the wind generators are going to just be sitting idle otherwise its always better to use them to do some form of generation.
it's for the hippies. out of sight = doesn't exist! so it's like magic for storing energy.
what about Congress-powered datacenters? (Score:2)
The inexhaustible supply of hot air blasting out of Washington, D.C. these days could probably power half the Third World. (Hint, also looking at you, Executive Branch)
Complex problem, many answers (Score:5, Interesting)
Using compressed air as a storage medium has a number of problems:
1) Low energy density. Air is very compressible. While that's what makes it usable as power storage in the first place, the amount of energy that can be saved in a gallon-sized container as compressed air pales next to a gallon of gasoline.
2) Power loss through thermal contraction.. When you compress air, it heats up. As that heat energy escapes, it effectively takes away a good chunk of the energy you consumed compressing it in the first place. Once the compressed air has cooled, the effective pressure drops. The more air you compress into the same space (See Energy Density above) the worse this effect is.
3) Power loss through leakage Even when available, caves are terrible places to store compressed air. Even if you seal the cave somehow, you then have to deal with seismic shifting, creating leaks and causing your beautiful, N-redundant power source to leak into an underwater stream bed. Lastly, even when sealed properly, air will *still* leak out. Ever wonder why your bike/car tires need to be aired up every so often even when they *aren't leaking?
Air molecules are wily little critters
4) turbine inefficiency The higher the pressure of compressed air, the more the above problems manifest themselves. However, the lower the pressure of compressed air, the less efficient it is to convert the compressed air back to electricity!
Good solutions address the multiple facets of the problem. For example, much of the cost of running a data center is spent on cooling. It might be preferable to store "coolth" in a stone or liquid cooling chamber under the facility than to try to store compressed air. Compressed air can be turned into electricity, which is more flexible, but is also more lossy for the reasons listed above. A combination of technologies will be needed to provide the best answer for redundancy and efficiency.
Re:Complex problem, many answers (Score:4, Informative)
Strike point 1 – we are talking about storage for grid generated by wind turbines. These things are out in the middle of nowhere. Size is not a major factor that would affect the storage cost. Even if built on prime corn land that is going for a thousand dollars an acre you can just burry these things and plow over the top.
You run these at relatively low pressure. That diminishes the impact of points 2 & 4.
That leaves 3. “Wind Alley” – from Texas to the Upper Mid-West is geologically stable.
This solution is only good for a very few select sites. You need wind and cheap caves. You can do the same with water – pumping it uphill, but the Wind Alley tends to lack water and hills.
UM? (Score:2)
Won't this inevitably lead to a situation where we are forced to purchase wind-generated energy just to get the wind farmers to release enough breathable air for humanity to survive? I, for one, say "no thank you" to eternal debt slavery to the alternative energy consortium.
Lazy Acronyms (Score:2)
...So you're telling me, they've invented a way to store wind in caves, and they called the acronym CAES?
Come on guys, figure out something to slap a V in there.
I don't get it (Score:2)
Perhaps I'm missing something, but if the goal was to take wind currently moving through a wind power center, and store that potential energy somewhere for later use when it wasn't windy... haven't we solved that issue many decades ago, with a technology called "batteries"?
Re: (Score:1)
There's this thing called money?
There might even be some in your wallet, but ...since you had to ask, maybe not.
And then there is Blenheim-Gilboa (Score:2)
Blenheim-Gilboa Pumped Storage Power Project [nypa.gov]
I love this place. It is set in the beautiful countryside of the Catskills
Earth Farts? (Score:2)
Old idea, sort of works, but ... (Score:2)
If solar is a peak load loser to start with... (Score:1)
Isn't using it to compress air in confined spaces or pump liquids uphill to recover some single-digit percentage of energy input, an bigger loser? Or is this one of those, we lose a little on each transaction but we make up for it in volume, er, things?
How about setting aside some tiny corner of one of the cave systems to store an itty bitty bit of nuclear waste, for as long as it takes to develop the technology to use the rest of the energy stored within it? For this privilege I am sure the nuclear power
very ineffecient (Score:2)
compressed air is a horribly inefficient way to store and transfer energy, industrial systems get about 15%
Re: (Score:1)
In industry, they're using it to transfer all of the energy they payed for through compressed air. It's the primary source. Excess waste energy cost the exact same if you use it or not, but you get more energy out of it if you can store it for later in stead of no energy at all.
With a giant "free" salt mine tank and with a small power requirement (10's MW) , the cost of storing the energy might be cheaper even with all of its inefficiencies than the cost of using batteries or buying it off the grid. The "
ENORMOUS UPS (Score:1)