• Riskable@programming.dev
    link
    fedilink
    English
    arrow-up
    9
    ·
    1 day ago

    For those wondering how TF a data center that is not even online yet could be using so much water:

    • Soil compaction
    • Dust suppression

    That’s it. For the scale of that project, that’s all it would take to use 30 million gallons.

    When they’re done, they also need to flush miles of pipes which could also use a few million gallons but I don’t think they’re at that phase yet.

    This amount of water would be used no matter what buildings they were constructing in that amount of space. Meaning: This article is pretty misleading clickbait (because a lot of people hate data centers lately, the headline will generate clicks).

    The alternative is to have loads of data centers instead of one big one. That’s more expensive, so they build a single big one.

    If you don’t like data centers, it makes sense to build a few really, really big ones like this rather than lots of smaller ones. Because data centers are necessary and important aspects of modern living. They’re not going to just go away. There’s nothing that could replace them.

    • Courtney (she/her/they) @lemmy.blahaj.zone
      link
      fedilink
      English
      arrow-up
      11
      ·
      24 hours ago

      The circlejerk for Ai simps is on Twitter. You seem to be lost.

      AI DATACENTERS DO NOT NEED TO BE BUILT AND COMPANIES SHOULD BE EXPANDING THEIR CURRENT PROPERTIES, NOT RUINING ENTIRE ECOSYSTEMS FOR GIANT BUILDINGS MEANT TO HOUSE ELECTRONICS

  • givesomefucks@lemmy.world
    link
    fedilink
    English
    arrow-up
    7
    ·
    edit-2
    1 day ago

    The thing is, we could harvest the heat and use it to generate energy instead of burning thru a natural resource were short of. (Edit: talking about potable water, not resources for energy generation)

    It would just be expensive, and capitalism always takes the cheapest route regardless of anything else.

    Obviously it’s not going to be much energy generation, but “waste heat” is only a thing if you’re wasting it. And at this scale it’s not insignificant, we shouldn’t be wasting it.

    • Blue_Morpho@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      ·
      1 day ago

      Not really. Despite the CPU running super hot, the water in a water cooling loop is a couple degrees C above ambient. Carnot’s law, which provides the theoretical maximum energy extraction not accounting for any real world loss, means you can’t get significant energy from a couple of degrees C temp difference.

      • bluGill@fedia.io
        link
        fedilink
        arrow-up
        5
        ·
        1 day ago

        You can get just enough energy for a science fair demonstration. Which scaled up to a data center size is a lot of energy. What the science fair misses is how much energy goes into making the system - You can generate what looks like a lot of power until you realize that the generators and such needed more energy to make than you will get back.

        • Blue_Morpho@lemmy.world
          link
          fedilink
          English
          arrow-up
          6
          ·
          20 hours ago

          Heat pumps work by using a high power pump to compress the working fluid from its gas state to liquid. This compression heats the gas to 30 C over ambient. That high temp is then cooled through coils to air temp (30C on a hot day). In the winter it runs in reverse. It takes expands gas which in the case of a heat pump fluid like R-410a boils at -48C, runs that -48C gas through the coils which picks up heat from the outside at -1C (48C hotter than the gas). It then compresses that fluid to 20C which is then released into the house. That’s why heat pumps can freeze up in the winter. The cold cycle is so far below freezing that ice will form on the coils. So heat pumps switch to AC mode every now and then in the winter to warm up their coils to melt the ice off (while turning off the fan in the house). The need for a large temperature difference is why heat pumps don’t work if it gets below -20C.

          So a heat pump moves heat but requires a large temp difference which comes from the electric compressor. You also can’t extract significant work from that heat difference once you factor the energy input of the compressor. Otherwise heat pumps would have a device to power part of itself from the heat that it is moving.

          • notfromhere@lemmy.ml
            link
            fedilink
            English
            arrow-up
            1
            ·
            15 hours ago

            So couldn’t we use a heat pump on the few degree above ambient waste heat to then do real work with it?

            • Blue_Morpho@lemmy.world
              link
              fedilink
              English
              arrow-up
              2
              ·
              edit-2
              2 hours ago

              Carnot is 1 - T(coldside)/T(hotside) in Kelvin.

              For a heat pump dumping heat a couple C above ambient (typical of heat pumps which is why people coming from furnaces complain that the air coming out isnt hot) that’s n = 1 - 298/300 = 1-0.9933 = 0.67%

              That means you’d need a machine that is more than 99.33% efficient to get any work out of that heat difference. For comparison, an engine losses 15-30% of energy just from the friction of polished metal cylinders and cams gliding close to another polished metal wall with a layer of oil in-between. The metal isn’t even touching.

              Actionlab just did a video on Carnot a few days ago.

              https://youtu.be/lGbrQJO3E_4

  • rumschlumpel@feddit.org
    link
    fedilink
    English
    arrow-up
    32
    ·
    1 day ago

    The company, which is owned by the private equity firm Blackstone, touts a “closed‑loop” cooling system, which it says does not consume water for cooling. Like a laptop or cellphone, the chips housed in data centers can easily overheat — generally requiring a lot of water to cool them.

    The company said its water consumption was so high last year because of temporary construction-related activities, such as concrete work, dust control and site preparation.

    I’ve never heard of water shortages due to construction needing so much. WTH are they doing on those sites if what Blackrock says is true?