Waste heat from data centers can boost air temperatures in downwind neighborhoods by as much as 4 degrees Fahrenheit, researchers at Arizona State University report in a new study conducted in the Phoenix metro area, the hottest in the U.S.
Once again I am shocked that you always have these big ass heat exchangers on these data centers but no talk of even trying to use some of the waste heat to offset the power use.
They are wasteful on purpose. they could have closed circuit cooling systems where they condense the water from the vapor and reuse it. But they are a giant middle finger to all of us.
Pretty sure most modern data centers are using closed loop cooling or refrigeration. Not evaporative cooling.
You can’t condense the water either that would defeat the point of evaporating it in the first place. Closed loop liquid cooling does not involve boiling or evaporation. You are just pumping a liquid around a circuit. It’s not just water either it’s more like a car antifreeze.
Got any info to back that up? Here’s a screenshot from page 39 of the US Data Center Energy Usage Report, which shows the use of closed loop systems (which they call dry cooling) as one of the smallest percentages of cooling types used. Pretty sure you’ve got it completely backwards on the types of cooling used, and I know for a fact the massive Amazon data center out in Oregon uses evaporative, because you can’t drink the water there as a result.
Why would they call closed loop liquid cooling “dry cooling”? Unless they explicitly said that then I don’t believe you to be honest.
Evaporative cooling wouldn’t make the water undrinkable unless something has gone very wrong. So I don’t think what you are saying about Oregan is true either.
This is also only representative of the USA, not worldwide. I get that much of the world doesn’t have many datacenters compared to the USA, but you at least have to include China and the EU.
The laws of thermodynamics? Can’t create or destroy energy and overall entropy increases over time. A closed loop (or any cooling system) just moves heat away from the hot thing. So yes, they can be used as much as any other cooling system but it won’t stop the issue of “generating lots of heat”. That heat still needs to go somewhere. Dumping it into the atmosphere might be the best option if there’s nothing in the area that needs heat. Should probably build them next to steel plants or something like that. Then a closed loop would be better.
I never said it fixed the issue of generating heat. Heat isn’t really a major problem as far as I am concerned. I thought we were talking about water use.
? OP was claiming that the majority of data centers use closed loop/refrigeration systems and I was pointing out that US data shows the vast majority use evaporative cooling. I’m not sure what the point you’re trying to make is in regards to those two statements. I’m not disputing the accuracy of what you’re saying, just unsure of where you’re going with it.
Thought your request to back it up was in response to the parent comment saying that condensing the water defeats the purpose rather than the first paragraph.
Some people making datacenter looks into way to recycle the extra heat, some uses it to heat local area (willingly). But all of this costs more than just, dumping it out, I guess.
Okay wiz, I’ll bite. While I’m working 60 hours a week, raising awareness online, and attending City council meetings, what else would you like me to fucking do? Mail them a pipe of dynamite?
Attending city council meetings and raising awareness are things you should be doing in the BEST of times. That’s called participation in democracy. If you think that’s all that’s called for right now in the US, I don’t know what to tell you.
If you plug the numbers into the Carnot equation, it looks like the maximum theoretical efficiency of a thermoelectric generator or heat engine operating at that temperature gradient is about 0.75%. And, I could be wrong, but my assumption would be any attempt to reclaim that energy would slow its exchange and potentially bottleneck a cooling system to some extent.
The issue is you are using 4 degrees as the delta of temp, when that is the ambient area temp change. The very localized heat being generated on site (and already nicely conveyed in heat management systems) is going to be a lot more then 4. Also why would you be always using a thermometric generator? They are not know to be efficient.
Once again I am shocked that you always have these big ass heat exchangers on these data centers but no talk of even trying to use some of the waste heat to offset the power use.
They are wasteful on purpose. they could have closed circuit cooling systems where they condense the water from the vapor and reuse it. But they are a giant middle finger to all of us.
Water is cheaper to waste than electricity unfortunately.
Pretty sure most modern data centers are using closed loop cooling or refrigeration. Not evaporative cooling.
You can’t condense the water either that would defeat the point of evaporating it in the first place. Closed loop liquid cooling does not involve boiling or evaporation. You are just pumping a liquid around a circuit. It’s not just water either it’s more like a car antifreeze.
Got any info to back that up? Here’s a screenshot from page 39 of the US Data Center Energy Usage Report, which shows the use of closed loop systems (which they call dry cooling) as one of the smallest percentages of cooling types used. Pretty sure you’ve got it completely backwards on the types of cooling used, and I know for a fact the massive Amazon data center out in Oregon uses evaporative, because you can’t drink the water there as a result.
Why would they call closed loop liquid cooling “dry cooling”? Unless they explicitly said that then I don’t believe you to be honest.
Evaporative cooling wouldn’t make the water undrinkable unless something has gone very wrong. So I don’t think what you are saying about Oregan is true either.
This is also only representative of the USA, not worldwide. I get that much of the world doesn’t have many datacenters compared to the USA, but you at least have to include China and the EU.
The laws of thermodynamics? Can’t create or destroy energy and overall entropy increases over time. A closed loop (or any cooling system) just moves heat away from the hot thing. So yes, they can be used as much as any other cooling system but it won’t stop the issue of “generating lots of heat”. That heat still needs to go somewhere. Dumping it into the atmosphere might be the best option if there’s nothing in the area that needs heat. Should probably build them next to steel plants or something like that. Then a closed loop would be better.
I never said it fixed the issue of generating heat. Heat isn’t really a major problem as far as I am concerned. I thought we were talking about water use.
Fair enough, I thought we were talking about the heat lol.
? OP was claiming that the majority of data centers use closed loop/refrigeration systems and I was pointing out that US data shows the vast majority use evaporative cooling. I’m not sure what the point you’re trying to make is in regards to those two statements. I’m not disputing the accuracy of what you’re saying, just unsure of where you’re going with it.
Thought your request to back it up was in response to the parent comment saying that condensing the water defeats the purpose rather than the first paragraph.
Nah, evaporation removes multiple times more heat than regular air cooling. It’s because water has high specific heat
They could boil seawater, condense the steam into drinking water
its more expensive since you will need to deal with corrosion and heat degrading the equipment overtime.
Nebius is doing it in Finland
Some people making datacenter looks into way to recycle the extra heat, some uses it to heat local area (willingly). But all of this costs more than just, dumping it out, I guess.
Typical waste in the USA. I believe Sweden or Finland pump the heat out for residential use.
I don’t think the main problem of Phoenix is the lack of heating
You need hot water in Phoenix, don’t you?
Use it to heat water for home use and a desert does become cool at night.
my crazy ass idea since 2015:
let’s install mini data centers on residential buildings:
cons:
Phoenix. What are you going to do with it in Phoenix? 24/7 saunas?
Americans simply don’t give a shit.
No, whiny clown about to reply, I don’t care if you’re “one of the good ones.” You STILL don’t give a shit.
Hey don’t lump me in with those whiny clowns! I’m a whiny clown in a clown suit. Damn foreigners making extremely generalized assumptions about us /s
Okay wiz, I’ll bite. While I’m working 60 hours a week, raising awareness online, and attending City council meetings, what else would you like me to fucking do? Mail them a pipe of dynamite?
Everyone has a job. You aren’t special.
Attending city council meetings and raising awareness are things you should be doing in the BEST of times. That’s called participation in democracy. If you think that’s all that’s called for right now in the US, I don’t know what to tell you.
Waste heat is high entropy alredy. You can’t extract any meaningful energy out of it.
Ah yes, famously we have never been able to do anything with heat energy…
It’s not my fault you don’t understand physics and think that all heat=free energy. You can’t extract shit without big temperature difference.
Yeah if only you had enough excess heat in one area to increase the local temperature of an already very hot place by 4 degrees.
If you plug the numbers into the Carnot equation, it looks like the maximum theoretical efficiency of a thermoelectric generator or heat engine operating at that temperature gradient is about 0.75%. And, I could be wrong, but my assumption would be any attempt to reclaim that energy would slow its exchange and potentially bottleneck a cooling system to some extent.
The issue is you are using 4 degrees as the delta of temp, when that is the ambient area temp change. The very localized heat being generated on site (and already nicely conveyed in heat management systems) is going to be a lot more then 4. Also why would you be always using a thermometric generator? They are not know to be efficient.
4 degrees Fahrenheit is nothing. For all practical energy production puropses it’s worthless