[throwaway2037]

    > Thermoelectric power plants: 41% of total US water use

I did some research on this statistic. More than 90% of this is returned to source. So if the power plant is drawing water from a river (or aquifer), it is returned after being used to generate electricity (create steam, turn turbines).

For example, if you have a 1GW coal powerplant on a river, it might draw 4 billion liters of water per day, but will only evaporate about 5-10% of that amount. The rest is returned to the river. Of course, this is much less exciting (click-baity) than writing "41% of US water use is for thermoelectric power plants".

[yreg]

Is it different with data centers?

[consp]

Depends on if they use evaporative cooling or not. But the water returned is warmer which can harm the ecosystem quite a bit, both for power plants and data centers.

[AnthonyMouse]

Water which is evaporated is also returned to the ecosystem. It falls back out as rain.

The heat issue is also widely misunderstood. The state of California gets on the order of a PWh of heat every day from the sun. A million GWh. Even a large nuclear plant doesn't make a dent in that. The issue is localized to the immediate area at the outlet, because the water temperature directly at the outlet is going to be higher than it is at the inlet, before it diffuses back into the background. Naturally the way to mitigate this is to pull more water through the cooling system, all of which you're still returning, but then you raise the temperature of twice as much water by half as much temperature, which is what you want since the temperature delta is the concern. Of course, doing this will cause people to accuse you of using more water, even though you're neither consuming the water nor modifying the amount of heat being transferred to the environment.

In addition to that, water has extremely different levels of scarcity depending on what region you're in. Parts of California and Texas have arid climates and extremely limited economical means to source additional fresh water. Areas like the Midwest get dramatically more rainfall and as a result fresh water is not meaningfully scarce there.

[lesostep]

>> Water which is evaporated is also returned to the ecosystem. It falls back out as rain.

No, not necessary to the same one. Every body of water has a catchment area, and through the bottom parts of some lakes/rivers water sips into the earth, filling underground water levels.

Now, let's say you life on a coast with a regular winds in the ocean direction.

You basin fills with water evaporated by you, your neighbors catchments and – sometimes – ocean (wind doesn't always blow in one direction). But the large percentage of the water you evaporate goes into the ocean. Let's say 90 percent falls again on your catchment area, 10 percent falls in ocean but is regained through neighbors basins evaporation.

Now, if you increase your evaporation, you could easily drain a lot of water from your land.

[AnthonyMouse]

> Now, if you increase your evaporation, you could easily drain a lot of water from your land.

At which point we're back to the quadrillion watt hours of sunlight you get a day, which evaporates such a large amount of water that you have to be doing something pretty dramatic to move the needle. Which thermal power plants and data centers don't, but agriculture? It's pumping trillions of gallons of water out of the ground and into fields in that sun.

[yreg]

Why can't they condense the vapor on site? Or simply use closed cooling loops? Would it be that much more expensive?

[AnthonyMouse]

The premise of evaporative cooling is that the phase change absorbs an enormous amount of heat and carries it away. The heat is returned when it condenses back into water, so doing it on-site would defeat the purpose.

Hypothetically you could recondense it somewhere else, i.e. get a large land area and then capture the rainfall. Probably the big win there would be to take the water which is currently routed into the ocean from storm drains and recover it instead of doing that with it.

[nate_meurer]

> Water which is evaporated is also returned to the ecosystem. It falls back out as rain.

This is about as useful as saying "the water stays on Earth". The ecological damage entirely depends on the nature of the water source.

The most poignant situation is non-rechargeable aquifers. When water is extracted, the mineral structures collapse and the storage capacity is forever destroyed.

And as the other reply stated, evaporation (as from evaporative cooling) exports water out of a catchment basin so that it can't be used downstream, just as if you pumped it away.

[AnthonyMouse]

> This is about as useful as saying "the water stays on Earth".

That is actually an important distinction. There are many things for which this is not the case. If you fail to recover the helium extracted with natural gas deposits then it doesn't stay on Earth. If birds eat your crops, you can't undo it. If you burn coal then it turns into CO2 and heat and it's thermodynamically impossible to reverse the process without inputting more useful energy than can be recovered from burning it to begin with (which is why carbon capture is a sham and we instead need to stop burning fossil fuels).

Whereas you can evaporate and recondense the same water an unlimited number of times and whether it's a problem depends not so much on what you're doing as where you got the water to begin with.

[nate_meurer]

That's a good point, and good examples. I'd compare the destruction of aquifers (some of them anyway) as akin to those things.

[throwaway2037]

I did some research on this matter. There are (at least) two uses for water by data centers. (1) If the data center is in a very low humidity environment, the data center is artificially humidified (is this a real English word?) to reduce the effects of static electricty. (2) Almost all data centers outside of cold regions (northern Europe, etc.) use water for evaporative cooling.

[dredmorbius]

humidified

Yes, a real English word.

Machines that humidify air are called ... humidifiers.

(There are also dehumidifiers.)

[tesnorindian]

Most rivers drain into Oceans while some do drain into inland lakes.