Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Data Centers Expected to Pollute More Than Airlines by 2020

Posted by timothy on Sat May 03, 2008 10:25 PM
from the it's-all-that-smut-in-the-wires dept.
Dionysius, God of Wine and Leaf, writes with a link to a New York Times story on a source of pollution that doesn't leave contrails: "The world's data centers are projected to surpass the airline industry as a greenhouse gas polluter by 2020, according to a new study by McKinsey & Co. ... [C]omputer servers are used at only 6 percent of their capacity on average, while data center facilities as a whole are used at 56 percent of peak performance." Data centers, though, might have more options for going green than airlines do, given present technology.
+ -
story

Related Stories

[+] Hardware: Green buildings, Green Server Farms? 263 comments
mstansberry writes "Has IT evolved to the point where it can consider energy efficiency without sacrificing uptime or performance? According to an interview with APC's Richard Sawyer, the answer is yes. The green buildings movement, spearheaded by the USGBC and other organizations has some people thinking about computing infrastructure's impact on the environment. Is it an IT issue or something from C-level executives?"
[+] Ask Slashdot: Renewable Energy for the Data Center? 55 comments
rohar asks: "The ISP/Carrier/Colo company I work for has just announced a new 'green' program. Although this is a step forward, they don't have a comprehensive environmental sustainability plan. I have been leading an open renewable energy project and I think we have 2 novel ideas for scalable and reliable renewable electrical power, the Solar Ammonia Absorption Convection Tower and the Compressed Air Wind Electrical Generation System. Do you have new ideas (Solar PV has been done, for example) for renewable power generation and conservation for the data center and other areas of industry?"
This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • More Options? (Score:5, Interesting)

    by Anonymous Coward on Saturday May 03 2008, @10:26PM (#23288574)
    Hardly.

    Most datacenters are contracted out. The companies hiring the datacenters do so based on price. And clean fuels have an enormous amount of catching up to do if they ever want to compete with coal. But let's say that a carbon tax is applied. Then these datacenter contractors will simply move their operations to somewhere that doesn't have these taxes. Heck, why do you think there are so many datacenters in the US?

    But what if the companies hiring these datacenter contractors decide that they want to be green? Then these datacenter contractors will simply do some half-assed unproven carbon-offset like dumping iron into the oceans or planting trees in a place that can't support them (cheap real estate like tundra or desert wins here--especially if it is done in the 'future' while the offset company is preparing its sites).

    The only real solution is the one that applies to the entire electricity grid. Either you need to massively subsidize renewable fuels or slightly subsidize nuclear power to deal with your entire electrical grid carbon problem. You have to do subsidies because you are competing with the energy prices with places like China.
    • Re: (Score:3, Informative)

      It's very expensive to move a datacenter. It's not just the building and server hardware, but local infrastructure, too. The biggest datacenters are in California for a reason.

      Therefore, the carbon tax need only be enough that taking the premium on greener energy tech is cheeper than taking the tax + moving and rebuilding infrastructure.

      • Re:More Options? (Score:5, Insightful)

        by WaltBusterkeys (1156557) * on Sunday May 04 2008, @12:18AM (#23289138)

        It's not just the building and server hardware, but local infrastructure, too.
        More importantly, it's where the big network connections intersect. A big data center in the middle of nowhere (with only 1 route to the outside world) is slow and vulnerable to backhoes. A data center near a major network interconnect (think west side of NY, or One Wilshire in LA [crgwest.com]) is somewhere useful -- data is close to the major lines and can be routed redundantly.

        Until they move the large cross-Pacific network connections to the Hoover Dam, it's going to make sense to keep data centers near network lines.
    • What percentage of the power consumption of running a data center is cooling? If they were to build a data center in a really cold environment, I wonder if they could pump the resulting heat under the ground in the immediate area, warming it up enough to plant trees...

      Although the other thing typical of tundra environments is the lack of sunlight, which may be more of a problem than the cold.

    • I think you have made some interesting points.

      However, I think the major issue is this: The people who design datacenters are some of the smartest people in the world. They've certainly thought about the issues. They know the cost of electricity.

      They know that Intel is delivering 45 nanometer CPU designs. They know that Intel is working on 32 nanometer CPUs, and that there will eventually be 22 nanometer processors, for delivery in 8 years. Each new processor architecture uses less power. So, the problem will solve itself, to some degree.

      The article in the New York Times is ignorant, meant for ignorant readers who don't know any better. Maybe someone took money; maybe the NYT article is really a public relations stunt, a way for McKinsey & Company [wikipedia.org] to attract as clients managers who have little technical experience.

      A lot of people who talk about being "green", are people who are green in the sense of having little experience.
      • Each new architecture does use less power, but that's not the way things are heading...
        New CPUs may have much better performance/watt, but the overall performance is increased too, thus the amount of actual power used stays the same or even increases.
        There's also increasingly bloated software, all this managed high level language code etc, which uses far more energy to do the same work. And modern powerful servers which sit idle for the most part.
        You could easily make lower performing servers using modern techniques, and reduce power consumption hugely... Modern embedded processors are faster than high end server processors from a few years ago, and yet use a small fraction of the power, but they wouldn't be good running modern bloated apps in high level languages.
        • Re: (Score:3, Interesting)

          Or, you could install virtualization on that super-fast machine end consolidate all your servers onto it.
          Modern datacenters sell either tiles (to place a rack), rackspace (for a few servers) or virtual computing power.

          The cost of each is reflected in the price so smart customers will move away from discrete hardware and towards virtual servers.
          That way you can literally run hundreds of low-power servers on one high-power machine.
          Low-power servers are nice, but they're not failure-resistant and the sheer num
        • Re: (Score:3, Insightful)

          When a practically brand new data center needs to be shut down for a weekend after 5 years because the power needs are inadequate and need to upgraded something wasn't planned right.

          I my experience designing buildings, including call centers, data centers, and server rooms, the specifications for the equipment to be used typcially isn't finalized until after the construction of the building is complete.
          To guess that evident continuing improvements in computer efficiencies will cause your data center to u

    • Re: (Score:3, Interesting)

      "But let's say that a carbon tax is applied."

      Why apply a carbon tax, when most coal fired power plants are located in locations where an algae based carbon recovery system (and combo pollution scrubbing system) can create not only vast amounts of vegetable oil, but even larger quantities of vegetable matter that can be feed to livestock, or combusted for energy, or converted to ethanol.

      If you required all coal fired plants to use an algae based carbon recovery system, you would instantly create a massive s
    • Nuclear Power is subsidized by the U.S. Government in an interesting way.

      In order for the first and any subsequent private nuclear reactors to even be built,
      the Congress passed a law [wikipedia.org] capping the amount nuclear reactor operators could be held liable. The operators are required to obtain $300 million per plant in insurance. If claims go beyond that, the industry is on the hook to provide a pool of money to pay claims beyond that $300 million. They are not required to provide this money until an accident oc
    • Re: (Score:3, Informative)

      "You have to do subsidies because you are competing with the energy prices with places like China."

      didn't see that the first go round.. That isn't really true, China has a notorious reputation for producing sub standard products. It's a reputation well earned, it's true that producing a higher quality product costs more, and even companies known for their quality are sub contracting parts of their businesses to china to stay cost competitive.. but you don't have to compete with china for cost of energy.

      w
  • Excellent (Score:5, Funny)

    by Plazmid (1132467) on Saturday May 03 2008, @10:33PM (#23288614)
    Excellent, the faster this planet's resources are used up the faster we start using other planets resources.
    • Vespene gas, anyone?
        • Re: (Score:3, Informative)

          And hyrdocarbon is an organic compound isn't it?
          So is plastic, but there's nothing about it that requires life to have been present. Same with hydrocarbons. Just because the most common source for hydrocarbon chains in our biosphere is organic material doesn't mean that the radically different conditions of an environment like Titan couldn't produce such chains out of raw materials.
          • Re: (Score:3, Informative)

            And hyrdocarbon is an organic compound isn't it?

            So is plastic

            Plastic was originally made from carbohydrates, specifically cellulose, and thus plants such as trees. Kodak [si.edu] the camera company used a method of making Cellulose acetate, a type of plastic, in 1908. If you're old enough you may recall Cellophane [wikipedia.org], the plastic wrap for sandwiches and such, it got it's name from what it was made from, cellulose. Today there's renewed interest in bioplastic [wikipedia.org].

            Falcon

      • Re: (Score:3, Interesting)

        Not to mention I read somewhere that the biggest contributor to global warming isn't the cars,planes,or as the article suggests future datacenters,but the billions of cows we have bred farting their little brains out releasing pure methane while they have more and more land clear cut to feed them. It is a shame we haven't figured out how to capture all those cow farts as there would be a fix to the energy crisis right there!

        But seriously from what I read (I believe on New Scientist but I can't seem to f

  • by Anonymous Coward on Saturday May 03 2008, @10:36PM (#23288634)
    Aiso.net is a smallish hosting provider utilizing ACTUAL SOLAR to power their datacenter,

    NONE OF THIS CARBON TRADING MALARKY. And they're super flexible because they're not huge yet.

    Located in San Diego I believe. Phil, their big tech cheese, is VERY generous with his time.

    Vote with your feet, clean with your wallet, live by your choices.
  • by pembo13 (770295) on Saturday May 03 2008, @10:39PM (#23288644) Homepage
    Data centers need electricity, not jet fuel. There are many semi-environmental ways to generate electricity. At some point companies will do that purely out of cost saving.
    • Re: (Score:3, Interesting)

      It's not just the generation of electricity that is at issue. It is how that electricity is used also at issue.

      2500 servers all converting from AC to DC = sizable loss of juice. Poorly designed data center rackspace using 10-30% (straight from my ass) more A/C than they would with efficient installations. I'm talking about force air systems that are misused etc.

      Installing passive heat exchange systems will also help when they become available.

      The point is that there are MANY things that can be done to cut d
    • Re: (Score:3, Interesting)

      They could relocate the Data Centers to North Dakota and Minnesota, where by 2020 there will be plenty of wind generated electricity, and the cool climate makes air conditioning unnecessary for 8 months of the year.
  • false economies (Score:5, Insightful)

    by timmarhy (659436) on Saturday May 03 2008, @10:39PM (#23288650)
    I love it when they trot out these old war horses.

    let me ask you this - what resources would be consumed if we DIDN'T use computers for these jobs? how many forests would we cut down to store the data in the worlds data centers?

    i think people who write this kind of dribble lack any perspective. computers are energy savers, not wasters.

    • Re:false economies (Score:4, Insightful)

      by jd (1658) <imipak AT yahoo DOT com> on Saturday May 03 2008, @11:16PM (#23288820) Homepage Journal
      Well, I agree that the statistics are useless in and of themselves, but to be fair, how many data centres are actually doing useful work? Point-to-point streaming of broadcasts, for example, is a horrible waste of CPU power and bandwidth, but it is the dominant method used by webcam services. OS overheads are often unnecessarily high, due to the running of excess services or inefficient code. Server rooms are often run far too hot and cooling methods are often inefficent.

      If we measure greenhouse gas production, not as an absolute but as a percentage relative to what is actually required to do the useful component of the work, my guess would be that data centres do not work out to be that green.

  • by Ralph Spoilsport (673134) * on Saturday May 03 2008, @10:43PM (#23288670) Journal
    given that there isn't going to be much of an airline industry in 2020. By then, fuel will be so expensive, air travel will revert to what it was prior to the 1970s: something the rich did.

    RS

    • Re: (Score:3, Insightful)

      wrong. by 2020 EVERYTHING will be too expensive due to poor economic policy based on non science and fear mongering.
    • by Chabil Ha' (875116) on Saturday May 03 2008, @11:39PM (#23288938)
      Or we could go back to trying to do nuclear powered aircraft [tripod.com]. This image [tripod.com] depicts a single prototype engine--its resting place is in southern Idaho.
      • by Watson Ladd (955755) on Sunday May 04 2008, @08:42AM (#23291050)
        It was originally designed to deliver nuclear weapons. But then they realized the radioactive fallout from the exhaust would be much more damaging. The air passed through the core and so was exposed to neutron radiation, making it chock-full of nasty isotopes. The xenon from the reactor would also exit in the exhaust stream, adding to the radioactivity hazard. There are good reasons why that technology died.
    • Re: (Score:3, Interesting)

      Or we're going to switch to biofuels. Several universities have tested aviation engines with a biofuel kerosene replacement. There isn't enough biofuel to replace oil for vehicles, but aviation uses only around 3-5 percent of total oil consumption, so there's some room to play with their fuel supply. No, aviation isn't going anywhere, it's just going to change.
        • Biofuels only exist because Iowa chooses the president of the united states.

          You are either woefully underinformed to the point where you are completely unqualified to contribute to this conversation, stupid, or an astroturfer.

          There are currently two biofuel technologies which are far superior to any topsoil-based biofuel. One of them is Butanol. The other is Algae-based biofuels which can include Ethanol and Biodiesel (mostly the latter.) You can also make biodiesel out of animal fat, and Tyson chicken is building a test plant to do this in Germany.

          The USDOE did a test project in which they determined that it is possible to capture around 80% of the CO2 output of coal or oil-burning plants and feed it to algae in inexpensive raceway ponds. The water in these ponds is approximately one foot deep and is circulated by paddlewheel - a job best done using PV solar. The water needs the most circulation during the periods of most intense sunlight. You could also tent the pools and use them for distillation; the process can be done with fresh or salt water, so it can also provide desalination.

          Butanol is made from a bacteria first used to produce the ingredients for TNT. This bacteria produces ethanol, butanol, and acetone, all of which can be burned in a typical gasoline-powered car. In fact, Butanol is a direct, 1:1 replacement for gasoline, and it is the most voluminous product of the reaction - which can consume any organic matter.

          There are also numerous other options for producing biofuels which should be considered. For example, we currently use extremely inefficient methods for processing sewage. By using a system of ponds which are filled from below, and which utilize a subaquatic plastic tent to capture methane gas using this efficient and attractive (since it is cheap and mostly invisible) method. Methane can be used most places in which we use propane or natural gas, and most especially for cooking. Just to prove the simplicity of the concept, consider that you can get cooking gas by raising pigs, shoveling their shit into a hole, and running a hose from the (covered) hole to a BBQ burner. This scheme also fixes heavy metals.

          It is true that biofuels based on topsoil are retarded. In fact, our current large-scale methods of agriculture are simply unsustainable. The crop waste must be returned to the soil, not burned as we commonly do today! Otherwise, the soil will be depleted over time, no matter what you do to it. It will simply be depleted of more specific things.

  • by ericferris (1087061) on Saturday May 03 2008, @10:43PM (#23288676) Homepage
    I went to a seminar on building new data centers. There we a part about location of new data center. The favorite places in Europe were France and Germany, because of cheap power generated by non-polluting nuclear power plant.

    I am aware of the end-of-life problem surrounding nuclear power, but you got to admit that if your goal is to avoid burning stuff, you cannot get any better than this. Especially in crowded, not-so-sunny Europe, where you cannot even make a "what if we paved the desert with solar cells" hypothesis.
    • Re: (Score:3, Insightful)

      I am aware of the end-of-life problem surrounding nuclear power, but you got to admit that if your goal is to avoid burning stuff, you cannot get any better than this. Especially in crowded, not-so-sunny Europe, where you cannot even make a "what if we paved the desert with solar cells" hypothesis.

      Why not? Africa isn't too far south of Europe. It's not any further than the Eastern USA is from the deserts of the USA, mostly in the southwest. The reason that doesn't matter is because we have a national p

        • Transmission loss over long distances is only a problem with AC. Transmitting electricity as DC at high voltages reduces the loss. Here's a page on using DC in Data centers: Edison's Revenge: Will DC power rise again? [computerworld.com].

          Falcon
          • Re: (Score:3, Informative)

            1. what makes you think it's more economical now? the price of cabling sure hasn't gone down. 2. your example ignores the fact you'd be stringing the line under the mediterannean. under sea cables are a whole lot more expensive. 3. even if you crossed at the most narrow point, it's still greater than 7000km (you best case) from any fesible site in africa into northern europe. remember you are proposing a solar site. that means costal area's are OUT.
    • Re: (Score:3, Informative)

      I went to a seminar on building new data centers. There we a part about location of new data center. The favorite places in Europe were France and Germany, because of cheap power generated by non-polluting nuclear power plant.

      Ah but nuclear power is polluting. Nuclear power pollutes from the ground to the ground, cradle to cradle.

      I am aware of the end-of-life problem surrounding nuclear power, but you got to admit that if your goal is to avoid burning stuff, you cannot get any better than this.

      It's

    • by blind biker (1066130) on Sunday May 04 2008, @05:06AM (#23290164) Journal
      France is predominantly nuclear, but Germany? That can't be right. The green party in Germany got all nuclear plants shot down (or scheduled to shut down) and the main energy source now is... wait for it... the much more polluting COAL! Good job, german green party: did you know that a coal plant will discharge into the environmnent twice as much radioactive pollutants per kilowatt-hour, than a nuclear plant? And we haven't even mentioned various other toxic materials, principally sulphur. And of course, CO2, in droves.

  • by FooAtWFU (699187) on Saturday May 03 2008, @10:52PM (#23288710) Homepage
    Anyone else remember when "pollution" was stuff like sulfuric acid, low-level ozone, toxic chemicals, and stuff like that? Carbon di-oxy-ide, who'da thunk, eh?
      • by evanbd (210358) on Saturday May 03 2008, @11:52PM (#23289000)

        Some of us define pollution as "anything that causes severe enough damage to our environment to make life difficult for us humans." And guess what, low-level ozone, ozone layer depleting compounds, acid rain precursors, CO2, volatile hydrocarbons, fertilizer runoff, and a variety of other things all count under that definition.

        I can be really selfish and even somewhat short-sighted and still come to the conclusion that there is a problem on a massive scale. I have no particular need for us to not create any CO2, but it should be obvious to anyone who bothers to look at the data and the studies that we can't continue on our current pace.

          • Re: (Score:3, Informative)

            you're an idiot. 25 billion tons of CO2 annually and you're suggesting it does nothing. minor greenhouse gas my ass.
          • Re: (Score:3, Insightful)

            Anthropogenic CO2 emissions are pumping enough "depleted" carbon into the atmosphere that the people who do carbon dating have to correct for it, since the atmospheric C-14 ratio is lower now than it was 50 years ago.

            If the people doing archaeological dating have to worry about it, I'd say it's major.
          • by statemachine (840641) on Sunday May 04 2008, @01:55AM (#23289618)
            Ah, another anthropogenic climate change denier. Well, this article [newscientist.com] and the US Supreme Court [msn.com] disagree with you.
  • by xzvf (924443) on Saturday May 03 2008, @10:59PM (#23288752)
    At the late 19th century steam engines were well established technology for shipping, trains and factories but they were very inefficient. Somewhere in the range of 15%. By the early 20th century steam power was at least twice as efficient (maybe more). Today most servers in data centers run around 15% utilization, doubling the utilization will slow the increased need for power. Virtualization, efficient parallel programming, thin client and network centric computing all have potential to double the efficiency of data centers. What would really be a breakthrough is a hybrid plane. Maybe with wireless power from space.
  • by ScentCone (795499) on Saturday May 03 2008, @11:03PM (#23288768)
    At least when it comes to my customers, the stuff that lives in datacenters is there - at least in part - to support distributed workers. In droves, they are shifting towards working from home, avoiding a lot of transportation-intensive face time, and learning to take advantage of not having to have their same back-office systems humming in a closet in a rented office where nobody shows up any more, except to reset the router so they can go back home and get some damn work done.

    Some newly used rack space in datacenters actually offsets other daily fuel burning - sometimes a lot of it.
  • by JRHelgeson (576325) on Saturday May 03 2008, @11:25PM (#23288864) Homepage Journal
    People that make such sweeping claims as this crap just light my fuse. They want to complain, and it seems their only point is to offer compromised solutions... Its like they fell like they're being helpful by getting in the way. If people would just start thinking realistically about these problems and allow the building of Nuclear Power plants, this problem would be solved. But it seems that these people don't want solutions, they want to complain about something. All they can do is point to a NEAR catastrophe, which was a mere accident at 3 mile island 30 years ago. Give. Me. A. Break!

    You get more radiation from eating a BANANA than you do from living next door to a nuclear power plant. And while on the subject, I used to think that these people were simply "NIMBY's", the age old Not In My Back Yard type of folks. But these people aren't NIMBY's, These people are BANANAS! Build Almost Nothing Anywhere Near Anything. They are flat out anti-progress and they do it in the nicest way "we're trying to help".

    I say BULLSHIT! You have three choices: Nuclear Power, Agrarian Society, Global Warming. Pick one.

  • Flat out wrong (Score:4, Insightful)

    by Rix (54095) on Sunday May 04 2008, @03:01AM (#23289794)
    Data centres emit absolutely no carbon. Zero.

    Electricity generation *can*, but it doesn't need to. The simple fact is that we can generate electricity without any carbon emissions with hydroelectric where available and nuclear where not. There's no justifiable reason to attribute carbon emissions from a coal fired plant to it's clients; alternatives are available, but regulators have dropped the ball in allowing coal to be used.
  • No Internet by 2050 (Score:3, Interesting)

    by bxwatso (1059160) on Sunday May 04 2008, @09:08AM (#23291212)
    Assuming population growth in this country does not stop, and assuming the USA can't get electricity from non-carbon sources, we will have to revert to a pre-industrial society to achieve Obama's carbon plan.

    There is no technology in existance that can provide all of the USA's electricity without carbon, except for nuclear. Things like wind and solar can only provide about 10-15% of the USA's current demand because they only work when the sun shines and the wind blows.

    Anyway, 80% emission reductions by 2050 would require that the USA give up a bunch of things, like cars, air conditioning, TV, hair dryers, air planes, buses, and computers. That is because the presidential candidate likes to toss out pleasant figures like 80 by 50 without consideration of reality.

    Population growth makes 80 by 50 impossible without a transforming technology like a nuclear powered economy with hydrogen transportation and storage of energy. It's not impossible to achieve, but politicians only like to talk about happy, fuzzy goals absent concrete plans to achieve them or admiting that they are extremely expensive.

    • Re: (Score:3, Insightful)

      For those of you who are keeping score on who's talking the talk and who's walking the walk I offer this:

      A tale of two houses [snopes.com]

      For a long time Bush has been downplaying or denying the effects of global warming. But behind the American People's backs he went ahead and built a geo-disaster proof bunker in 2001.

      I need to change my pants.