Seawater Cooled Data Centers


Remember Mauritius from High School geography? We didn’t either, but apparently it’s a small island nation east of the southern tip of Africa. It seems they are trying to develop an industry in eco-friendly data centers. The plan is to use a pipeline to gather cold water from the ocean, run it two miles to the island, and use it as inexpensive cooling. Because rooms packed with servers generate copious amounts of heat it’s easy to see how this can reduce the cost of maintaining a data center.

The thing that struck us here is, how eco-friendly is this? The article mentions that this technology is fairly mature and is already used in several places. With that in mind, isn’t this just another way to raise the temperature of the oceans, or does the environmental savings of not using electricity or gas to produce the cooling offset this?

[Thanks Vesanies]

89 thoughts on “Seawater Cooled Data Centers

  1. This is fundamentally a good idea. 40 degree chilled water can be used in hydronic systems to great effect. Moving the water two miles And insulating it well enough to stay cold would be quite a challenge. Piping does carry a (seemingly) small friction coefficient that doesn’t always come into design challenges until you are talking about long lengths or small pipes. And that is in a perfectly insulated system. 58 degree chilled water much less attractive to use in a conventional precision cooling environ.

    To those who say that this has no effect on the thermal capacity of the ocean: you are wrong. “taking” heat out of a system will always have an effect on that system no mater how immeasurable it is on a small scale. Returning more energy to thAt system (hotter water) also.

    I like the Idea of using the heated water in preheat applications much better (with perfect insulAtion again)
    iaaaPE (mechanical)

  2. Nothing new about this sort of thing. The Sydney Opera House does the same thing with it’s Air Conditioning. Also, the electricity supplier in Christchurch, New Zealand used to use waste heat in the public swimming pool across the road heating the water.

  3. You get nothing for free, and cooling is definitely “something”. Now you can pay for it with coal which you burn to create electricity, and the coal will pollute and waste a bunch of energy as heat, and the transmission grid will waste energy via heat and the chilling system will waste a bunch of energy as heat.. Or if you are luck you can do like these guys and use 90% less electricity and harness the temperature differential in the deep trench off the island to be your thermal sink.

    If you do not understand why this it a good idea then you do not understand basic physics, please review the first law of thermodynamics and try again. If that doesn’t help then think about the electrical cooling system as taking a loan from a loan shark (IE burning fossil fuels), to cover the loan to the bank (to power the grid), to pay your credit card (to cool your CPU)…. Rather then just paying the credit card directly. Why go through all the middle men and lose 90+% of the efficiency?

  4. They’ve been using ocean assisted A/C & solar hot water heating in Hawaii for years.

    In places like Vegas they do ground loops where you run water/coolant though a closed loop deep enough into the ground where the temperature is about 55F all year. It used to be common to use evaporative cooling (because it’s so dry here) but the Feds take our water now give it to California.

    For anybody having trouble understanding please consider the free online MIT lectures. It’s very easy to learn the basics.

  5. Everything in large scale more effective, the problem is large amount of $ for building such system, do you think every resident of building will be OK with paying couple $K instantly even if it will save them 100X more over long run

  6. Since the 1970’s my company has specialised in “Sub-sand sea water abstraction”. We install filters, some up to 70 metres long under sandy beaches and suck! Result, clean filtered sea water, with a massive filter bed, constantly washed by waves and tidal action. Nearly all coastal sea water aquariums and research labs in UK and Ireland use our system. In recent years we have installed these intakes in Greece for exactly this purpose. Cool sea water in for air conditioning. The warmed water goes to the pools and spa and the surplus back to the sea. Win win, and very environmentally friendly.

    Compare this with conventional air conditioning which generates heat! No contest.

  7. Warming even a small lake with a datacenter would be like warming a pool with a matchstick.

    Anyway, assuming there effectively would be some measurable water warming, it could be counteracted by putting windowblind-like sliding reflective screens above water level to screen the sunlight during day.

  8. Maybe we could offset the extra energy put into the ocean by warming stuff like houses with AC units that work in reverse. Taking relatively warm surface water chilling it and pumping the cooler water back in at a lower depth (matching the temperature).

    This would give us possibly twice as much heat energy as we put in in electricity.

    Here in the UK we could do with a better way of warming our houses. My gas bill is draining all my money.

  9. You guys do realize that the amount of heat generated by even the most ambitious data center or office building complex will be like 1% of the heat output of an operating nuclear reactor? Comparing the heat output of one of these projects to a power plant is like apples and oranges.

  10. kevin, I’m sorry I had to resort to such a juvenile line, I had a much larger post ready to go when I remembered a few of my own rules, “be quiet” and “keep it simple”. my biggest question to all the global warming types that are commenting on the damage this would do the the ocean is the amount of solar energy that hits the ocean directly, every hour of direct sunlight is worth about a KW of electricity for every 100 sq feet. and they think MAN is warming the ocean when the sun is keeping it from freezing?

  11. Just think of the Ocean as a heatsink. It’s really not that big a deal. Even more beneficial would be to pump the hot exhaust water to buildings. Either for hot water, or through radiators to heat them when it’s cold. There’s plenty of other ways to use the energy that has been dumped in that water from the datacenter.

  12. James, I’d really like to see some numbers run. Either by you, or the people making this cooling system. I just see the local ecosystem being totally destroyed if this is put to use.
    Though, operating costs will be lower, so that’s a plus, right?!

  13. Live will flourish near the warm water release, as long as it is not toxic. Anything but the sun heating the entire ocean is pure science fiction. A silly but honest rough estimation for heat getting in the oceans would be: Sun 1,000,000,000 kWh, vs a large data center at 1 or 2 million kWh. If you have 500 solar panels, and 10000 toaster oven’s in that datacenter, the amount of heat would not be that much more than the heat energy provided by the sun in the 500 solar panel’s square area of ocean. If that would heat the ocean up, the panama canal would have killed all oceanic life.
    You cannot create or destroy energy, you can only change its form. “Heat” is not a problem anywhere on earth, its the co2 that doesn’t let light escape.

  14. yah, it’s that whole “flourishing life” that I DONT like about this. The local ecosystem is going to change. once scenario is algea growing all over the surface, blocking sunlight to everything underneath. I’m not saying it’ll happen, or that it’s likely.
    I’m saying that I want to know the full impact of this project before being in favor or against it

  15. Sayeth the Owen:
    “Using gas or coal to power your cooling system puts carbon into the atmosphere, which has a very long-term effect on the world’s temperature.”

    that’s man made global warming speak to me, what say you?

  16. @aztraph
    The point is that

    > Just heating the ocean does nothing to the long-term temperature of the ocean and the world.


    > Whether or not there’s an environmental benefit, it’s hard to argue with a 90% decrease in electricity consumption.

    There /is/ a scientific consensus that the global warming of the last fifty years is attributable to human activity, but that is absolutely irrelevant in the context of this post.

  17. On the Big Island of Hawai’i we have an existing cold-water infrastructure developed for OTEC. Recently, someone proposed just such a cold-water cooled datacenter here. However, we have very high electricity costs ($.40+ per KwH); I am skeptical that the savings in cooling costs can outweigh the increased cost of operation. I would guess that other islands have similarly expensive energy-generation issues. I’d also love someone with a knowledge of datacenter costs to confirm / refute my assumption that operating costs are more significant than cooling costs.

    (And, of course, OTEC is an awesome potential energy source for islands such as Mauritius and Hawai’i.)

  18. OWEN!
    “There /is/ a scientific consensus” is like trident gum saying 10 million teeth can’t be wrong, oh yes they can. a consensus is like saying it’s unanimous, they can all still be wrong but at least they agree about it. and no it isn’t relevant to the article, but it is relevant to YOUR post, I’m arguing with YOU, not the article, it’s a really good idea that shows a lot of thought and ingenuity. being basically a water source heat pump, it can be fine tuned for the application to maximize efficiency. this means less power and less cost in it’s operation, that is always a good idea.

    again I apologize for the rant, but to ANYONE who buys into the MAN-MADE global warming MYTH, i have some nice beach front property in Nevada for you to by, bit of a walk to get to the beach but believe me, It’s worth it.

  19. With all the talk of heating / cooling the ocean, I’m surprised that nobody has yet mentioned that a 300 baud modem would probably triple the bandwidth in that area of the world, so having this system in place for a data center is sort of a moot point.

  20. With all the talk of heating / cooling the ocean, I’m surprised that nobody has yet mentioned that this would have a net COOLING effect on the ocean as a whole, not increasing it. Seems counter-intuitive, but true.
    The effect would be so negligible as to be unobservable, but the net effect would be to lower, not increase, the oceans overall heat energy.
    At any localized level the effect is neutral, as it is designed to be.

    Understand basic freaking physics and thermodynamics, then comment. Fucking hippies.

  21. hello there, here in Polynesia they do implement this Sea water air conditioning system which made me end up on this post after one night of internet search. From what i know they will go between 600till900m deep to get water around 4-5 degres and drop it at 20 degres around 50m deep where the sea is at same temp… ok I am not a specialist but if you get 300000 L (i have no ideas about the real volume c’se it looks likes they hide it) at 4 degres and you put this volume back but at 20 degres, it cannot be very friendly for the ecosystem no? did anybody heard about methane gas which is very much reactiv to a small raise of ocean temperature, which could be responsible for the bermuda triangle disapearing boats.. For the ones saying it will not affect us, can you please give me the reference where i can find a proper study because they do start to implement this new technology a bit everywhere on the planet, I need your eyes, i have spend one night on this search, apart of a chinese research which have conclude that this is still under study, I do not see anywhere a correct specialist study on the subject but looks like they do not need any autorization and public information to start implement it all around the world… Isabelle

  22. when you say “run numbers” and want to do texts to see diferent effects of this, what types do you use? im trying to complete a science fair project on the posible negative effects of this system and need help with numbers, not just these great theories! thanks!

  23. Okay…so, late to the party, but all of the heat/hatred FOR this idea seems to have missed something.

    You have a cup of water. The bottom half is cold, the top half is hot. There is no exchange from top to bottom. 25% of the bottom is taken and heated to the top half’s temperature.

    Now it is 75% hot, 25% cold…

    Applied to this that would mean a sudden displacement of a ridiculous amount of water, but I think most of you know where I’m going with this. We’re subtracting from cold water and adding to hot water, in other words we’re still adding heat into the ocean regardless of if we’re adding it to a place that is already ‘hot’.

    The change isn’t that we’re heating a particular part of the ocean, but adding more heat directly to the ocean. Cold water is subtracted, hot water is added, period.

    Our atmosphere, the higher you go the cooler it gets…anyway, I missed the party. Oh well.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.