Ten to 15% of the server power, depending who you ask, goes to the in-rack fans on a typical server - that adds up to kW's per rack. We have even measured a few kW swing in rack power consumption due just to variable speed board fans reacting to cold aisle temperature resets, so while it's peanuts in the context of a 20 MW datacenter, it still adds up to real money to be saved. And more importantly, since most big facilities are limited by their peak MW feed, a kW not spent on fans can be spent on revenue-generating servers. Small, high speed fans are inherently inefficient compared to larger fans (we're talking 20% versus 65%) and there are a LOT of of those little monsters chewing up power in a server.
That said, eliminating fans is not rocket science, and moves to (massive) passive heatsinking, direct or near water cooling, rack-scale heatpipes, etc are far more likely to beat a Rube Goldberg'esque approach such as this. But I'm still heading over to craigslist to see if there is a good little lathe/mill going for cheap so I can try to build one...
I'm not sure where they're coming up with that 500 W/sf number, even for 2011. The numbers I've been seeing and designing to are more like 200-300 W/sf. Above that, things get awfully tricky to cool using just air. Take a rack at 500 W/sf. The rack footprint with access aisles (included in the industry standard definition of W/sf is the access space) is at least double the rack footprint plus space for pdus and support, so call it 20 square feet. So you need to move out 10 kW with air. Assuming a supply air temperature of 70F (if you're supplying lower, you're pissing away energy and need to get modern equipment or update your education*) and a temperature rise of 30F across the computer, you need to move (forgive the conversion to English units, my mental math doesn't convert)
8000 Watts = 27,300 btu/hr
(34,100 btu/hr) / 1.08 / 30 = 1050 CFM per rack
And that is assuming that you have an airtight seal between the hot and cold aisles (which you should be pretty close to if you're going that dense)(or you should fire your mechanical designer). Unless you increase the footprint or get crazy and supply from above and below (possible but costly), you're supplying that air through a cross-section of 4sf at a diffuser velocity of 250 fpm. That's not a datacenter, it's a windtunnel.
Now, I am going off a bit half cocked here. You could indeed get up to 500 W/sf if the hardware is designed to be cooled, probably by water but possibly by high delta-t air. And the potential for entirely passive cooling of that level of heat is there. I'm just a little sensitive to people tossing out numbers that are patently silly. If they'd at least use an intelligent fan supply algorithm on the recircs (and please god, don't use those constant flow CRACs) it wouldn't waste quite as much power.
As a side note, if google is spending 0.5 Watts to cool per 1 Watt of cpu heat (a power usage effectiveness of 1.5) in their new Portland facility, they need to put some of their eggheads** to work trolling the current datacenter efficient design literature. I know their solar cell powered campus is sexy, but they could save an order of magnitude more energy for a tenth the cost by getting their datacenter up to snuff (the obvious freecooling, the barely-justifiable-but-sexy adsorption chillers, and the esoteric thermal stack harvesting to offset fan power for starters)(a rigid hot aisle/cold aisle design is, of course, assumed just to get a 300W/sf datacenter that won't cook the top slots with recirc).
* OK, if you can make cold air for free via free cooling, go for it but I expect you to do the work on the fan controls to actually harvest a few kW's for the added humidity control complexity.
** Based on a presentation I saw at ASHRAE, Google does have at least one sharp tack in the drawer, so I suspect this is a matter of a Harper's 'expert' making up numbers.
Actually, authors have found via the Baen free library that giving away their books in e format resulted in increased sales. Not only did they see the obvious benefit of giving away book one of a series resulting in an increase in the sales of books two, three, four... But the sales of the free book also increased. Check out this interesting summary. Hopefully this doesn't really come as a surprise to Slashdotters, who are some of the loudest preachers of this type of market behavior.
And, as you file your own tax returns this year, I'll bet you carefully record each internet transaction from out of state, ensuring that you pay full taxes even though it would have been easy to avoid it?... When an individual figures out ways to avoid paying taxes - or paying as little as possible - it's considered frugal. When a corporation does it, it's evil?
Actually, it is trivial to record major internet transactions* from out of state (just file the email invoice and count 'em up at the end of the year) and yes, I do pay state sales taxes on them. To not do so is not called frugal, it's called fraud and it will inevitably lead to the government having to add more regulators to deal with the blatant dishonesty you seem to consider simply "frugal."
*I'm not talking about a $20 purchase of Leak Frogs from Woot.com, but if it's over $100 you don't have an excuse not to note it, only the motive of wanting to cheat on your taxes so honest folks like me help pay for your roads.
Could someone run the numbers on how long it'd take to have them register *every* possible domains up to 50 characters in length? I don't think its in the 'heat-death-of-the-universe' range, but it is looming near 'bigger than Bill O'Reilly's' ego range. (Something like 36 acceptable charactors would mean 36^36*.025 seconds a domain/31536000 seconds per year = Dictionary Attack.)
It is currently serving as a stand for my XP based T40. Vista Business has been a disaster for me in terms of billable hours lost troubleshooting stuff that just works in XP (hibernate resume and wireless)(yes, compared to Vista XP's wireless is rock solid). The wireless networking has a nasty intermittent failure that takes reboot/reinstall driver/manually configure wireless network/sacrafice a goat to resolve, temporarily. As the T61p is a recent purchase, I have little more than Firefox and Putty on it on it, so at least it is relatively painless to wipe it (again, but last time I foolishly reinstalled Vista) and install XP. The only good thing I can say about Vista is that it's remote desktop works as well as XP's.
Based on this experience, I have pointed my parents firmly at a Mac Mini for Christmas and am planning to clear out the few MS stocks in my retirement account. Vista is not a WindowsME level screwup, we're talking more like the love child of Microsoft Bob and Clippy on the aggravation scale.
Putting it in a cave is not a win. The ground water cooling is great, but by putting it in a cave they are severely limiting their access to air-side economization, which is bringing in filtered outside air directly when it is cooler than the rack's exhaust temperature. And if you're doing it right, the rack's exhaust temperature is 95F+ (when we do a hot aisle, we make it a hot aisle dangit). As a professional in the efficient datacenter arena, the mention that they are still using chillers at all when they have a source of ground water is actually pretty disappointing. Sure it takes some custom work, but if you have a high flow enough source of 60F or lower water, you're done. Of course, it has to be very high flow - 13,650 gpm for a 30MW datacenter (a size we're seeing more and more - those 20kW+ blade racks are killers). Also, bringing the water up to the datacenter is way easier than bringing the datacenter to the water - it's called a pump, wiki it.
The comment about "no air conditioning will be required outside of the containers" is a bit stupid. There is nothing located outside of the containers, so of course there is no air conditioning required there. I don't air condition the parking lot just outside a traditional datacenter either.
Short summary: Datacenter in a mine = security win, HVAC fail.
I don't have much sympathy for him - his Buffy and Angel runs show the guy knows how to play the game. Even if Firefly was killed a few seasons early, Buffy was killed a few season late. He's not really a rube walking into Fox's spinning blades of ineptness.
Although to be fair, most of the work I've seen already done in this vein (years ago) did have Lawrence Berkeley National Labs involved, so I guess that includes the feds.
Being personally biased towards stuff I've had a hand in, I think the self benchmarking guide at http://hightech.lbl.gov/documents/DATA_CENTERS/Sel f_benchmarking_guide-2.pdf and the design source book for datacenters at http://hightech.lbl.gov/documents/DATA_CENTERS/06_ DataCenters-PGE.pdf are far more interesting than a watered down Congress-critter report. I'm happy to see they mentioned free cooling, an amazing "duh" approach that is inexplicably ignored far too often (get an engineer to design it and the filtration and humidity control in the winter issues are trivial to handle, plus you end up with a net improvement in redundancy). But I'm annoyed at their ASHRAE boiler plate mention of hot aisle/cold aisle - a design approach that can be a Big Deal as far as saving energy goes as long as the fan control is done properly.
As a closing note, benchmarking has shown that typically about half the power going into a datacenter goes to keeping it cool. So listen to us mere mechanical engineers if you want to save a buck or million on the power bill...
The refrigerant-based approach is an efficiency disaster. Any serious datacenter is cooled with chilled water, or it is using double the watt/ton of cooling it should be. A datacenter that is not using water based cooling and some form of freecooling, which saves money even in climates like Phoenix or Atlanta in a 24/7 flat-load datacenter situation, should be sued for false advertising if they claim to be "efficient." There is benchmarking data available on this - a closed-loop dx system is an energy disaster, no matter how slick it looks. And don't even bring up "Well, they're just like a bunch of little chillers" until you find me a 10 ton* centrifugal compressor or an 80 ton rack (or 150 ton if you don't want to be locked into using a Turbocor compressor based unit).
The chilled water coil on the back of a rack is an excellent solution, and a good approach to dealing with the 12 kW racks that datacenter clients are starting to throw at us HVAC engineers. It is also something that has been offered before by IBM and others, and can be custom fabbed up if you have a large, controlled layout facility (co-los are too chaotic to pull off a custom solution in my experience).
As mentioned elsewhere, this is hardly cutting edge. Checkout the Datacenter Design Sourcebook here:http://www.pge.com/docs/pdfs/biz/rebates/high tech/06_DataCenters-PGE.pdf (I'm biased towards this since I was involced in its creation and it summarizes my thinking as of last year and little has changed, other than a bit more data on the impact of speed controlled server fans impacting the operational approach to hot/cold aisles).
*1 ton = 12,000 btu/hr, I think in Burmese units thank you. And on a similar tangent, you'll pry Fed standard 209E cleanroom ratings from my cold, gowned hands.
(Reposting since I'm an HVAC engineer and somehow missed getting the link in correctly the first time)
The refrigerant-based approach is an efficiency disaster. Any serious datacenter is cooled with chilled water, or it is using double the watt/ton of cooling it should be. A datacenter that is not using water based cooling and some form of freecooling, which saves money even in climates like Phoenix or Atlanta in a 24/7 flat-load datacenter situation, should be sued for false advertising if they claim to be "efficient." There is benchmarking data available on this - a closed-loop dx system is an energy disaster, no matter how slick it looks. And don't even bring up "Well, they're just like a bunch of little chillers" until you find me a 10 ton* centrifugal compressor or an 80 ton rack (or 150 ton if you don't want to be locked into using a Turbocor compressor based unit).
The chilled water coil on the back of a rack is an excellent solution, and a good approach to dealing with the 12 kW racks that datacenter clients are starting to throw at us HVAC engineers. It is also something that has been offered before by IBM and others, and can be custom fabbed up if you have a large, controlled layout facility (co-los are too chaotic to pull off a custom solution in my experience).
As mentioned elsewhere, this is hardly cutting edge. Checkout the Datacenter Design Sourcebook here (I'm biased towards this since it summarizes my thinking as of last year and little has changed, other than a bit more data on the impact of speed controlled server fans impacting the operational approach to hot/cold aisles).
*1 ton = 12,000 btu/hr, I think in Burmese units thank you. And on a similar tangent, you'll pry Fed standard 209E cleanroom ratings from my cold, gowned hands.
I would suggest you simply spec a roofing with appropriate radiative characteristics, aka a 'cool roof'. That's basically a no-brainer for commercial buildings at this point - do you want a 120F roof or a 100F roof on the peak load cooling day?
The 500 pound gorilla in the corner is that in a typical Silicon Valley datacenter only 50-60% of the power goes to the computers while the other half goes to the support equipment. It does not have to be this way, and things are changing. I have not yet walked into a datacenter that could not cut its total power usage by at least 25% (albeit, in some cases the design damage is done and the simple payback required to make it work would stretch to 4-5 years)(I'm looking at you, datacenters with dozens of 20-30 ton air-cooled compressors on the roof).
On the gross kWh/yr side, the vast majority of datacenters are unable to use outside air directly for cooling. A 24 hour a day load and they can't 'open the windows' to cool it at night (with appropriate filtration and redundant humidity control lockouts of course)? Come on people! It would even improve reliability (even 70F outdoor air could hold a well configured hot aisle/cold aisle datacenter). But that doesn't help trimming peak load, to do that you have to get the airflow right.
Efficiency in datacenters starts with just a basic understanding of airflow. You want it very hot behind the racks; you want that hot air to go directly back to your cooling unit not get recirc'd to a rack intake. And you have to have airflow controlled based on the cold aisle temperature to harvest energy savings (fan energy wastage is ridiculous in these things)(oh, and watch out for those server fans that ramp up if you push the cold aisle temp too high - not efficient to provoke a rack of those guys to start screaming).
You have to know hot aisle / cold aisle to properly design and operate an efficient datacenter, even if that exact configuration is not applicable. Period.
Of course, its not "that simple," but to the design engineers it certainly should be pretty straightforward work. The information is out there and more is in the pipeline. A good start on the basics of efficient datacenters is available here (full disclosure, I was associated with producing that report, so I am not impartial)(but don't blame me for the blurry graphics - I did not create the pdf!).
And for god's sake people, quit keeping these places at 55-60F - I'm freezing my butt off and you're making a mockery of your own 'tight humidity control' (70-90% RH at the server intakes, but a good 45% +/- 2% at the air handler return).
In case someone hasn't already posted the relevant wikipedia link, here is the actual debate by people who have studied history over whether the bombing of Hiroshima and Nagasaki were necessary. There is nothing patently obvious about the issue.
You'd think this, unfortunately datacenter design is still mired in "reliability is the only priority." Efficiency just gets lip service. You'd be amazed at the number of 20000 sf datacenters using air cooled Libert systems (about 1.2-1.5 kW/ton cooling) rather than a water cooled chiller system (0.6-1 kW/ton cooling)(yes, including tower, fans, and pumps) that pays off in less than a year.
I think we have different points. My main point is I want verifiable facts. Scientists who do not stand up and voice their disagreements are useless, and history has suggested they are also in error (where's my cold fusion and don't even get me started on that Fermat joker). The vast majority of people voicing doubts about global warming are uninformed people who are not aware of all the currently known facts tossing out FUD. The people I have found who offer scientific opposition with a factual basis they are willing to share and defend are listed in that wikipedia article.
He's not the only one. In fact, back in 2001, some 18,000 scientists signed a petition against global warming, including some 2,000+ climate scientists. On the other side, less than 50 climate scientists sided with the IPCC 2000 report.
This is interesting - could you cite it? Although I can't help but notice that this petition is prior to the last half-decade of intensive study that has only supported global warming and found that opposing theories to not match the observed facts, for example the discovery of the incorrectly tabulated satellite data (lousy source but I don't have time to find the tech journals online - feel free to do so yourself if you're interested in more than a really simple summary of the facts) suddenly made climatic models look significantly better (they were closer to correct than the 'observed' data was). Without the citation, I have no choice but to dismiss the petition you mentioned as at best a false recollection and at worst a lie at some level (which would include if the petition asked scientists something like, "It is a 100% certain fact that man is causing massive global warming, yes or no?"). As an example of what I am looking for, saying that the National Academies of Scientists for the G8 nations plus Brazil, China and India, who comprise far more than 50 climate scientists, support global warming is a fact with citation. I would further say that these Academies contain reputable professionals. I do not have the time to cite the 1,000s of papers detailing facts that support this position, but I trust the thousands of scientists who have.
When I heard a climate scientist at a seminar back in college around 2000, he stated that they did not really know what was going on with global warming - could be man caused, could be a natural variation, could be some sort of pervasive observational error (heat islands, atmospheric effects, misinterpretation of secondary data sources like ice cores and geographical sources, etc.). Now, after years of research and collecting facts, he supports the current theory of global warming.
Arguing that consensus doesn't mean anything is cute, but I am talking about facts. Gravity, evolution, relativity theory and global warming have quite a consensus because one scientist has not come forward with facts that defeat them. OK, maybe gravity but take pity on my non-physicist education - its part of the reason I tend to believe the massive consensus of reputable professionals in their fields. Particularly when massive dollars from industry and the potential of a Nobel prize tempts the first person to defeat global warming theory with facts rather than spin.
And opposing global warming because the climate modeling tools being developed to better understand it are not perfect - I honestly don't know where to begin on that one.
The facts, thousands of them, point to global warming. Are you honestly saying we should toss all those facts and the theory developed to explain them based on the un-reviewed, un-researched, and untested opinion of a single person? What is your competing theory that better explains all the research discovered information? If you do not have a theory that is equally well shown to explain all the facts d
No, I am claiming that people claiming that global warming is "a debate" do not have any factual basis or professional ability to make that claim, and offering them an easy way to refute me. If you changed the wiki to state 12,000,000 skeptics, I would be able to quickly determine it was false. Actually, how would you come up with 12,000,000 names - cat in a phone book?
Try this experiment: Make up a name and affiliation and slip it into the wiki. See how long it lasts.
I did not mention ulcers by accident, but rather to point out that there is a great incentive for scientists to buck even the most solidly accepted 'fact.' Challenging consensus is very difficult, but highly valued - it won a Nobel prize. The current climate is not like the dark ages where the reward for shattering a universally held delusion was being branded a heretic.
According to the article "Global Warming Skeptics," there are only 12 scientists who disagree with global warming. From the discussion here, clearly there must be more disagreement. I'm sure it's not just a bunch of hacks making stuff up (this is slashdot, home of scientific minded folk), so if you folks could go over to the Wiki and list who your reputable sources for questioning the thousands of scientists who have been trying (and failing) to poke holes in global warming for the last 10 years are it would be helpful. Because from the looks of that article, the creationists have better scientific footing than folks arguing against human influenced global warming. And while consensus does not have a causative relationship with fact, it does, given enough time, seem to correlate frequently in the area of modern science (even ulcers were figured out eventually).
Slashdot Mirror
Ten to 15% of the server power, depending who you ask, goes to the in-rack fans on a typical server - that adds up to kW's per rack. We have even measured a few kW swing in rack power consumption due just to variable speed board fans reacting to cold aisle temperature resets, so while it's peanuts in the context of a 20 MW datacenter, it still adds up to real money to be saved. And more importantly, since most big facilities are limited by their peak MW feed, a kW not spent on fans can be spent on revenue-generating servers. Small, high speed fans are inherently inefficient compared to larger fans (we're talking 20% versus 65%) and there are a LOT of of those little monsters chewing up power in a server.
That said, eliminating fans is not rocket science, and moves to (massive) passive heatsinking, direct or near water cooling, rack-scale heatpipes, etc are far more likely to beat a Rube Goldberg'esque approach such as this. But I'm still heading over to craigslist to see if there is a good little lathe/mill going for cheap so I can try to build one...
Doh! I hate it when I accidentally post anon... I'm the measurement worshiper above. In god we trust, all others bring data...
8000 Watts = 27,300 btu/hr
(34,100 btu/hr) / 1.08 / 30 = 1050 CFM per rack
And that is assuming that you have an airtight seal between the hot and cold aisles (which you should be pretty close to if you're going that dense)(or you should fire your mechanical designer). Unless you increase the footprint or get crazy and supply from above and below (possible but costly), you're supplying that air through a cross-section of 4sf at a diffuser velocity of 250 fpm. That's not a datacenter, it's a windtunnel.
Now, I am going off a bit half cocked here. You could indeed get up to 500 W/sf if the hardware is designed to be cooled, probably by water but possibly by high delta-t air. And the potential for entirely passive cooling of that level of heat is there. I'm just a little sensitive to people tossing out numbers that are patently silly. If they'd at least use an intelligent fan supply algorithm on the recircs (and please god, don't use those constant flow CRACs) it wouldn't waste quite as much power.
As a side note, if google is spending 0.5 Watts to cool per 1 Watt of cpu heat (a power usage effectiveness of 1.5) in their new Portland facility, they need to put some of their eggheads** to work trolling the current datacenter efficient design literature. I know their solar cell powered campus is sexy, but they could save an order of magnitude more energy for a tenth the cost by getting their datacenter up to snuff (the obvious freecooling, the barely-justifiable-but-sexy adsorption chillers, and the esoteric thermal stack harvesting to offset fan power for starters)(a rigid hot aisle/cold aisle design is, of course, assumed just to get a 300W/sf datacenter that won't cook the top slots with recirc).
* OK, if you can make cold air for free via free cooling, go for it but I expect you to do the work on the fan controls to actually harvest a few kW's for the added humidity control complexity. ** Based on a presentation I saw at ASHRAE, Google does have at least one sharp tack in the drawer, so I suspect this is a matter of a Harper's 'expert' making up numbers.
Actually, authors have found via the Baen free library that giving away their books in e format resulted in increased sales. Not only did they see the obvious benefit of giving away book one of a series resulting in an increase in the sales of books two, three, four... But the sales of the free book also increased. Check out this interesting summary. Hopefully this doesn't really come as a surprise to Slashdotters, who are some of the loudest preachers of this type of market behavior.
And, as you file your own tax returns this year, I'll bet you carefully record each internet transaction from out of state, ensuring that you pay full taxes even though it would have been easy to avoid it? ...
When an individual figures out ways to avoid paying taxes - or paying as little as possible - it's considered frugal. When a corporation does it, it's evil?
Actually, it is trivial to record major internet transactions* from out of state (just file the email invoice and count 'em up at the end of the year) and yes, I do pay state sales taxes on them. To not do so is not called frugal, it's called fraud and it will inevitably lead to the government having to add more regulators to deal with the blatant dishonesty you seem to consider simply "frugal."
*I'm not talking about a $20 purchase of Leak Frogs from Woot.com, but if it's over $100 you don't have an excuse not to note it, only the motive of wanting to cheat on your taxes so honest folks like me help pay for your roads.
Could someone run the numbers on how long it'd take to have them register *every* possible domains up to 50 characters in length? I don't think its in the 'heat-death-of-the-universe' range, but it is looming near 'bigger than Bill O'Reilly's' ego range. (Something like 36 acceptable charactors would mean 36^36*.025 seconds a domain/31536000 seconds per year = Dictionary Attack.)
It is currently serving as a stand for my XP based T40. Vista Business has been a disaster for me in terms of billable hours lost troubleshooting stuff that just works in XP (hibernate resume and wireless)(yes, compared to Vista XP's wireless is rock solid). The wireless networking has a nasty intermittent failure that takes reboot/reinstall driver/manually configure wireless network/sacrafice a goat to resolve, temporarily. As the T61p is a recent purchase, I have little more than Firefox and Putty on it on it, so at least it is relatively painless to wipe it (again, but last time I foolishly reinstalled Vista) and install XP. The only good thing I can say about Vista is that it's remote desktop works as well as XP's. Based on this experience, I have pointed my parents firmly at a Mac Mini for Christmas and am planning to clear out the few MS stocks in my retirement account. Vista is not a WindowsME level screwup, we're talking more like the love child of Microsoft Bob and Clippy on the aggravation scale.
Putting it in a cave is not a win. The ground water cooling is great, but by putting it in a cave they are severely limiting their access to air-side economization, which is bringing in filtered outside air directly when it is cooler than the rack's exhaust temperature. And if you're doing it right, the rack's exhaust temperature is 95F+ (when we do a hot aisle, we make it a hot aisle dangit). As a professional in the efficient datacenter arena, the mention that they are still using chillers at all when they have a source of ground water is actually pretty disappointing. Sure it takes some custom work, but if you have a high flow enough source of 60F or lower water, you're done. Of course, it has to be very high flow - 13,650 gpm for a 30MW datacenter (a size we're seeing more and more - those 20kW+ blade racks are killers). Also, bringing the water up to the datacenter is way easier than bringing the datacenter to the water - it's called a pump, wiki it. The comment about "no air conditioning will be required outside of the containers" is a bit stupid. There is nothing located outside of the containers, so of course there is no air conditioning required there. I don't air condition the parking lot just outside a traditional datacenter either. Short summary: Datacenter in a mine = security win, HVAC fail.
Battered writer syndrome?
I don't have much sympathy for him - his Buffy and Angel runs show the guy knows how to play the game. Even if Firefly was killed a few seasons early, Buffy was killed a few season late. He's not really a rube walking into Fox's spinning blades of ineptness.
Credit should be given for past accomplishments. After all, this may be even better than It/Ginger/Segway/A-#@$$#^@-Self-Balancing-Scooter.
Although to be fair, most of the work I've seen already done in this vein (years ago) did have Lawrence Berkeley National Labs involved, so I guess that includes the feds. Being personally biased towards stuff I've had a hand in, I think the self benchmarking guide at http://hightech.lbl.gov/documents/DATA_CENTERS/Sel f_benchmarking_guide-2.pdf and the design source book for datacenters at http://hightech.lbl.gov/documents/DATA_CENTERS/06_ DataCenters-PGE.pdf are far more interesting than a watered down Congress-critter report. I'm happy to see they mentioned free cooling, an amazing "duh" approach that is inexplicably ignored far too often (get an engineer to design it and the filtration and humidity control in the winter issues are trivial to handle, plus you end up with a net improvement in redundancy). But I'm annoyed at their ASHRAE boiler plate mention of hot aisle/cold aisle - a design approach that can be a Big Deal as far as saving energy goes as long as the fan control is done properly.
As a closing note, benchmarking has shown that typically about half the power going into a datacenter goes to keeping it cool. So listen to us mere mechanical engineers if you want to save a buck or million on the power bill...
The refrigerant-based approach is an efficiency disaster. Any serious datacenter is cooled with chilled water, or it is using double the watt/ton of cooling it should be. A datacenter that is not using water based cooling and some form of freecooling, which saves money even in climates like Phoenix or Atlanta in a 24/7 flat-load datacenter situation, should be sued for false advertising if they claim to be "efficient." There is benchmarking data available on this - a closed-loop dx system is an energy disaster, no matter how slick it looks. And don't even bring up "Well, they're just like a bunch of little chillers" until you find me a 10 ton* centrifugal compressor or an 80 ton rack (or 150 ton if you don't want to be locked into using a Turbocor compressor based unit).
The chilled water coil on the back of a rack is an excellent solution, and a good approach to dealing with the 12 kW racks that datacenter clients are starting to throw at us HVAC engineers. It is also something that has been offered before by IBM and others, and can be custom fabbed up if you have a large, controlled layout facility (co-los are too chaotic to pull off a custom solution in my experience).
As mentioned elsewhere, this is hardly cutting edge. Checkout the Datacenter Design Sourcebook here:http://www.pge.com/docs/pdfs/biz/rebates/high tech/06_DataCenters-PGE.pdf (I'm biased towards this since I was involced in its creation and it summarizes my thinking as of last year and little has changed, other than a bit more data on the impact of speed controlled server fans impacting the operational approach to hot/cold aisles).
*1 ton = 12,000 btu/hr, I think in Burmese units thank you. And on a similar tangent, you'll pry Fed standard 209E cleanroom ratings from my cold, gowned hands.
(Reposting since I'm an HVAC engineer and somehow missed getting the link in correctly the first time)
The refrigerant-based approach is an efficiency disaster. Any serious datacenter is cooled with chilled water, or it is using double the watt/ton of cooling it should be. A datacenter that is not using water based cooling and some form of freecooling, which saves money even in climates like Phoenix or Atlanta in a 24/7 flat-load datacenter situation, should be sued for false advertising if they claim to be "efficient." There is benchmarking data available on this - a closed-loop dx system is an energy disaster, no matter how slick it looks. And don't even bring up "Well, they're just like a bunch of little chillers" until you find me a 10 ton* centrifugal compressor or an 80 ton rack (or 150 ton if you don't want to be locked into using a Turbocor compressor based unit).
The chilled water coil on the back of a rack is an excellent solution, and a good approach to dealing with the 12 kW racks that datacenter clients are starting to throw at us HVAC engineers. It is also something that has been offered before by IBM and others, and can be custom fabbed up if you have a large, controlled layout facility (co-los are too chaotic to pull off a custom solution in my experience).
As mentioned elsewhere, this is hardly cutting edge. Checkout the Datacenter Design Sourcebook here (I'm biased towards this since it summarizes my thinking as of last year and little has changed, other than a bit more data on the impact of speed controlled server fans impacting the operational approach to hot/cold aisles).
*1 ton = 12,000 btu/hr, I think in Burmese units thank you. And on a similar tangent, you'll pry Fed standard 209E cleanroom ratings from my cold, gowned hands.
I would suggest you simply spec a roofing with appropriate radiative characteristics, aka a 'cool roof'. That's basically a no-brainer for commercial buildings at this point - do you want a 120F roof or a 100F roof on the peak load cooling day?
The 500 pound gorilla in the corner is that in a typical Silicon Valley datacenter only 50-60% of the power goes to the computers while the other half goes to the support equipment. It does not have to be this way, and things are changing. I have not yet walked into a datacenter that could not cut its total power usage by at least 25% (albeit, in some cases the design damage is done and the simple payback required to make it work would stretch to 4-5 years)(I'm looking at you, datacenters with dozens of 20-30 ton air-cooled compressors on the roof).
On the gross kWh/yr side, the vast majority of datacenters are unable to use outside air directly for cooling. A 24 hour a day load and they can't 'open the windows' to cool it at night (with appropriate filtration and redundant humidity control lockouts of course)? Come on people! It would even improve reliability (even 70F outdoor air could hold a well configured hot aisle/cold aisle datacenter). But that doesn't help trimming peak load, to do that you have to get the airflow right.
Efficiency in datacenters starts with just a basic understanding of airflow. You want it very hot behind the racks; you want that hot air to go directly back to your cooling unit not get recirc'd to a rack intake. And you have to have airflow controlled based on the cold aisle temperature to harvest energy savings (fan energy wastage is ridiculous in these things)(oh, and watch out for those server fans that ramp up if you push the cold aisle temp too high - not efficient to provoke a rack of those guys to start screaming).
You have to know hot aisle / cold aisle to properly design and operate an efficient datacenter, even if that exact configuration is not applicable. Period.
Of course, its not "that simple," but to the design engineers it certainly should be pretty straightforward work. The information is out there and more is in the pipeline. A good start on the basics of efficient datacenters is available here (full disclosure, I was associated with producing that report, so I am not impartial)(but don't blame me for the blurry graphics - I did not create the pdf!).
And for god's sake people, quit keeping these places at 55-60F - I'm freezing my butt off and you're making a mockery of your own 'tight humidity control' (70-90% RH at the server intakes, but a good 45% +/- 2% at the air handler return).
In case someone hasn't already posted the relevant wikipedia link, here is the actual debate by people who have studied history over whether the bombing of Hiroshima and Nagasaki were necessary. There is nothing patently obvious about the issue.
You'd think this, unfortunately datacenter design is still mired in "reliability is the only priority." Efficiency just gets lip service. You'd be amazed at the number of 20000 sf datacenters using air cooled Libert systems (about 1.2-1.5 kW/ton cooling) rather than a water cooled chiller system (0.6-1 kW/ton cooling)(yes, including tower, fans, and pumps) that pays off in less than a year.
42.
I think we have different points. My main point is I want verifiable facts. Scientists who do not stand up and voice their disagreements are useless, and history has suggested they are also in error (where's my cold fusion and don't even get me started on that Fermat joker). The vast majority of people voicing doubts about global warming are uninformed people who are not aware of all the currently known facts tossing out FUD. The people I have found who offer scientific opposition with a factual basis they are willing to share and defend are listed in that wikipedia article. He's not the only one. In fact, back in 2001, some 18,000 scientists signed a petition against global warming, including some 2,000+ climate scientists. On the other side, less than 50 climate scientists sided with the IPCC 2000 report. This is interesting - could you cite it? Although I can't help but notice that this petition is prior to the last half-decade of intensive study that has only supported global warming and found that opposing theories to not match the observed facts, for example the discovery of the incorrectly tabulated satellite data (lousy source but I don't have time to find the tech journals online - feel free to do so yourself if you're interested in more than a really simple summary of the facts) suddenly made climatic models look significantly better (they were closer to correct than the 'observed' data was). Without the citation, I have no choice but to dismiss the petition you mentioned as at best a false recollection and at worst a lie at some level (which would include if the petition asked scientists something like, "It is a 100% certain fact that man is causing massive global warming, yes or no?"). As an example of what I am looking for, saying that the National Academies of Scientists for the G8 nations plus Brazil, China and India, who comprise far more than 50 climate scientists, support global warming is a fact with citation. I would further say that these Academies contain reputable professionals. I do not have the time to cite the 1,000s of papers detailing facts that support this position, but I trust the thousands of scientists who have. When I heard a climate scientist at a seminar back in college around 2000, he stated that they did not really know what was going on with global warming - could be man caused, could be a natural variation, could be some sort of pervasive observational error (heat islands, atmospheric effects, misinterpretation of secondary data sources like ice cores and geographical sources, etc.). Now, after years of research and collecting facts, he supports the current theory of global warming. Arguing that consensus doesn't mean anything is cute, but I am talking about facts. Gravity, evolution, relativity theory and global warming have quite a consensus because one scientist has not come forward with facts that defeat them. OK, maybe gravity but take pity on my non-physicist education - its part of the reason I tend to believe the massive consensus of reputable professionals in their fields. Particularly when massive dollars from industry and the potential of a Nobel prize tempts the first person to defeat global warming theory with facts rather than spin. And opposing global warming because the climate modeling tools being developed to better understand it are not perfect - I honestly don't know where to begin on that one. The facts, thousands of them, point to global warming. Are you honestly saying we should toss all those facts and the theory developed to explain them based on the un-reviewed, un-researched, and untested opinion of a single person? What is your competing theory that better explains all the research discovered information? If you do not have a theory that is equally well shown to explain all the facts d
No, I am claiming that people claiming that global warming is "a debate" do not have any factual basis or professional ability to make that claim, and offering them an easy way to refute me. If you changed the wiki to state 12,000,000 skeptics, I would be able to quickly determine it was false. Actually, how would you come up with 12,000,000 names - cat in a phone book? Try this experiment: Make up a name and affiliation and slip it into the wiki. See how long it lasts.
I did not mention ulcers by accident, but rather to point out that there is a great incentive for scientists to buck even the most solidly accepted 'fact.' Challenging consensus is very difficult, but highly valued - it won a Nobel prize. The current climate is not like the dark ages where the reward for shattering a universally held delusion was being branded a heretic.
According to the article "Global Warming Skeptics," there are only 12 scientists who disagree with global warming. From the discussion here, clearly there must be more disagreement. I'm sure it's not just a bunch of hacks making stuff up (this is slashdot, home of scientific minded folk), so if you folks could go over to the Wiki and list who your reputable sources for questioning the thousands of scientists who have been trying (and failing) to poke holes in global warming for the last 10 years are it would be helpful. Because from the looks of that article, the creationists have better scientific footing than folks arguing against human influenced global warming. And while consensus does not have a causative relationship with fact, it does, given enough time, seem to correlate frequently in the area of modern science (even ulcers were figured out eventually).