Urban Country Boy said:

---

Coates said:

---

Principal Uncertainty said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

No, they are not. Closed loop chilling for the building envelope, but future servers will have liquid immersion for cooling with plate and frame heat exchangers directly to an evaporative cooling tower. They are in the prototype stage now, but will be the future of compute.

---

Liquid immersion has been the 'future' for at least a decade and is way past the prototype stage, this has not and will likely not be widespread adopted anytime soon.

ETA that if a user went liquid immersion there is absolutely no water needed, it would use a specialized fluid.

---

This is being used now. I would have liked to see the reaction of the data guys when they were told, "We have an idea. We are going to submerge your servers to cool them".

---

So what are the current standards for data centers and are all new DC's being built for AI following them? Closed loop vs open loop? Immersion?

So the OP uses some kind of AI mashup of a midwest farm and field plopped into the AZ desert to be against a data center. Yea, that's convincing. NOT.

That land is pretty but there are literally millions of acres of desert in AZ/NM/TX that is barely to graze 1 cow per 300 acres. A DC is not destroying the countryside.

Plus, as mentioned by others, if you have a cell phone, are on Texags/Facebook/TikTok/ etc.... you [and I] are part of the reason we need more Data Centers. Just stupid NIMBYism, which I'm sick of. I live way in the boonies, and am not concerned about any noisy DC being built withing 50 miles of me.

As far as water goes, it is most likely less water that is currently being used for flood irrigation on alfalfa fields all along the Rio Grande from ABQ south to Socorro.
ANd there is a huge solar farm being installed just south of Los Lunas NM next to I-25 on some pretty barren land that wouldn't support 2 cows per section otherwise. Perfect use for that desert land.

Martels Hammer said:

---

Sulphur Springs TX is getting, I am told, the largest pending build of a new center. And their city budget, by memory, will go from $6M to something like $400M.

So I would assume the people in Sulphur will be pretty happy.

Edit

I "hear" the city is saying the residents will not longer have to pay for garbage etc. I will believe that when I see it but come on, free garbage people.

---

A city budget going from 6 to 400 mm and its several thousand residents will benefit from….

Free trash pick up?

And that's supposed to be exciting?

I can look into a land lease deals for data centers outside ABQ at the Zorro Ranch. Only stipulation if there can be no digging.

insulator_king said:

---

So the OP uses some kind of AI mashup of a midwest farm and field plopped into the AZ desert to be against a data center. Yea, that's convincing. NOT.

That land is pretty but there are literally millions of acres of desert in AZ/NM/TX that is barely to graze 1 cow per 300 acres. A DC is not destroying the countryside.

Plus, as mentioned by others, if you have a cell phone, are on Texags/Facebook/TikTok/ etc.... you [and I] are part of the reason we need more Data Centers. Just stupid NIMBYism, which I'm sick of. I live way in the boonies, and am not concerned about any noisy DC being built withing 50 miles of me.

As far as water goes, it is most likely less water that is currently being used for flood irrigation on alfalfa fields all along the Rio Grande from ABQ south to Socorro.
ANd there is a huge solar farm being installed just south of Los Lunas NM next to I-25 on some pretty barren land that wouldn't support 2 cows per section otherwise. Perfect use for that desert land.

---

Any additional load on an already failing water source seems monumentally stupid. Your example may actually not apply here but DC's elsewhere in the West are short-sighted from a resource standpoint. Article from today on the CO River on the brink of failure:

https://www.wsj.com/us-news/climate-environment/the-colorado-river-is-on-the-brink-of-disaster-628516be

Quote:

---

The Colorado River is running dangerously low, and the seven Western states that rely on it can't agree on how to share what's left.

To make up for shortfalls, states pull water from the reservoirs, which serve as storage banks for the system. Drawing down reserves has taken a toll. Lakes Mead and Powell, the biggest reservoirs in the basin, have fallen about 75% from peak volumes.

---

flown-the-coop said:

---

Urban Country Boy said:

---

agracer said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

but require more electricity.

---

Please explain.

---

Water has to be just not cooled but circulated. In open loop, one way flow and moderate to no cooling of the fresh water.

Just my guess on the need for more electron vs H2O juice.

---

I am not sure what you are trying to say.

BigRobSA said:

---

BrazosDog02 said:

---

Don't blame me. I'm unapologetically completely anti progress. If I could push a button and send my county back to 1948, ship out everyone that moved here that's not an original founding family, I'd smash that button so fast I'd probably break every bone from my fingers to my shoulder….which would be a pain since there aren't hospitals where I am in 1948. But I'd roll those dice anyway.

---

---

This is my fantasy. It can be gay!

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

agracer said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

but require more electricity.

---

Please explain.

---

Water has to be just not cooled but circulated. In open loop, one way flow and moderate to no cooling of the fresh water.

Just my guess on the need for more electron vs H2O juice.

---

I am not sure what you are trying to say.

---

Me either.

I was trying to explain why closed loop would use more electricity..

YouBet said:

---

Urban Country Boy said:

---

Coates said:

---

Principal Uncertainty said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

No, they are not. Closed loop chilling for the building envelope, but future servers will have liquid immersion for cooling with plate and frame heat exchangers directly to an evaporative cooling tower. They are in the prototype stage now, but will be the future of compute.

---

Liquid immersion has been the 'future' for at least a decade and is way past the prototype stage, this has not and will likely not be widespread adopted anytime soon.

ETA that if a user went liquid immersion there is absolutely no water needed, it would use a specialized fluid.

---

This is being used now. I would have liked to see the reaction of the data guys when they were told, "We have an idea. We are going to submerge your servers to cool them".

---

So what are the current standards for data centers and are all new DC's being built for AI following them? Closed loop vs open loop? Immersion?

---

Data centers are not all being built for AI. Microsoft, Amazon, etc. are still the big owners. Immersion is still new. But closed loop is how things have gone. There is a reduction in evaporative cooling.

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

agracer said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

but require more electricity.

---

Please explain.

---

Water has to be just not cooled but circulated. In open loop, one way flow and moderate to no cooling of the fresh water.

Just my guess on the need for more electron vs H2O juice.

---

I am not sure what you are trying to say.

---

Me either.

I was trying to explain why closed loop would use more electricity..

---

Why does it?

Edit: You are correct. But it is tiny compared to the overall use. And the overall less use of water is huge.

My assumption would be that you have to chill the warm water back down before using it whereas open loop bring on fresh water that requires no cooling or less effort to cool.

I would also assume more pumps and similar equipment to ensure the closed loop system is properly circulating (well, recirculating).

I could be wrong, but it made sense in my hewd.

"Closed loop" is not universal. Hence the need to know additional detail to understand how the system works.

Evaporative cooling systems (not closed loop) is not efficient for water, but it's efficient for getting rid of heat. Closed loop, this reverses (to some extent depending on the details), but there are several ways where the system can be cooled with large scale fans or chillers. The bad part is these are most needed when demand is highest, aka it's hot outside.

BrazosDog02 said:

---

BigRobSA said:

---

BrazosDog02 said:

---

Don't blame me. I'm unapologetically completely anti progress. If I could push a button and send my county back to 1948, ship out everyone that moved here that's not an original founding family, I'd smash that button so fast I'd probably break every bone from my fingers to my shoulder….which would be a pain since there aren't hospitals where I am in 1948. But I'd roll those dice anyway.

---

---

This is my fantasy. It can be gay!

---

Embracing the gay, on F16?!

Respect.

Urban Country Boy said:

---

Coates said:

---

Principal Uncertainty said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

No, they are not. Closed loop chilling for the building envelope, but future servers will have liquid immersion for cooling with plate and frame heat exchangers directly to an evaporative cooling tower. They are in the prototype stage now, but will be the future of compute.

---

Liquid immersion has been the 'future' for at least a decade and is way past the prototype stage, this has not and will likely not be widespread adopted anytime soon.

ETA that if a user went liquid immersion there is absolutely no water needed, it would use a specialized fluid.

---

This is being used now. I would have liked to see the reaction of the data guys when they were told, "We have an idea. We are going to submerge your servers to cool them".

---

Used now, sure, but extremely rare. I have been in hundreds of data centers and can count on one hand the sites that have liquid immersion.

All of the new data centers use CDUs, heat exchangers, containment or a mix of all.

Urban Country Boy said:

---

YouBet said:

---

Urban Country Boy said:

---

Coates said:

---

Principal Uncertainty said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

No, they are not. Closed loop chilling for the building envelope, but future servers will have liquid immersion for cooling with plate and frame heat exchangers directly to an evaporative cooling tower. They are in the prototype stage now, but will be the future of compute.

---

Liquid immersion has been the 'future' for at least a decade and is way past the prototype stage, this has not and will likely not be widespread adopted anytime soon.

ETA that if a user went liquid immersion there is absolutely no water needed, it would use a specialized fluid.

---

This is being used now. I would have liked to see the reaction of the data guys when they were told, "We have an idea. We are going to submerge your servers to cool them".

---

So what are the current standards for data centers and are all new DC's being built for AI following them? Closed loop vs open loop? Immersion?

---

Data centers are not all being built for AI. Microsoft, Amazon, etc. are still the big owners. Immersion is still new. But closed loop is how things have gone. There is a reduction in evaporative cooling.

---

I imagine the current data centers, like one in rural Georgia I saw a report on, that are costing locals' electricity bills to all but double, have less water coming in their faucets, and light and noise pollution to the point that they live in constant light, even at night, inside their homes, will be the "older" technology ones in the near future.

It sucks for the folk that will have to live with those consequences for as long as they live there, but for people living in the places where the newer tech doesn't provide those problems, at least what they live through will be the first step towards improvements to where, hopefully, data centers are nothing more than light and noise pollution for locals, but nothing more in terms of bills and utilities.

flown-the-coop said:

---

My assumption would be that you have to chill the warm water back down before using it whereas open loop bring on fresh water that requires no cooling or less effort to cool.

I would also assume more pumps and similar equipment to ensure the closed loop system is properly circulating (well, recirculating).

I could be wrong, but it made sense in my hewd.

---

In an open loop there is condenser water from the evaporative cooling towers that transfer heat from the water cooled chillers. Even in this system no water is just a once through and dumped. It is evaporated to make the cooling "state change". It uses a lot of water. A 50 story high rise on a hot day in Houston uses about 200+ gallons a minute to make up for the evaporation because the use cooling towers. Who knew their office building uses more water than a closed loop data center?

Especially if it China thats putting them in!!!!

Coates said:

---

Urban Country Boy said:

---

Coates said:

---

Principal Uncertainty said:

---

jpb1999 said:

---

Principal Uncertainty said:

---

TAMUallen said:

---

Data centers are going to be such an issue for electricity grids

---

Most maga-size data centers are installing their own generation and not even connecting to any grid. But many mid-size ones are, so point is partially valid. Also, with the new high-powered severs being liquid cooled, they will just use a heat exchanger to a cooling tower. So, no refrigerated chiller and fin-fan coolers. The cooling towers will use much less electricity (for the cooling, not the chip power), so it's by far the cheaper way to go. But evaporative cooling does consume water, so in dry place like west Texas where cheap fuel gas exists wil, indeed, need to manage the water consumption.

---

Most are going to a closed loop system with little water use.

---

No, they are not. Closed loop chilling for the building envelope, but future servers will have liquid immersion for cooling with plate and frame heat exchangers directly to an evaporative cooling tower. They are in the prototype stage now, but will be the future of compute.

---

Liquid immersion has been the 'future' for at least a decade and is way past the prototype stage, this has not and will likely not be widespread adopted anytime soon.

ETA that if a user went liquid immersion there is absolutely no water needed, it would use a specialized fluid.

---

This is being used now. I would have liked to see the reaction of the data guys when they were told, "We have an idea. We are going to submerge your servers to cool them".

---

Used now, sure, but extremely rare. I have been in hundreds of data centers and can count on one hand the sites that have liquid immersion.

All of the new data centers use CDUs, heat exchangers, containment or a mix of all.

---

Agree. Direct to chip is our standard now.

Bear in mind Houston also has at least one if not more central chilling plants downtown. From the plant to and through the buildings is "closed loop" and the chillers at the plant are designed to recapture evaporated water if memory serves and / or just chill the looped watery using other methods.

For instance, I did learn they also have "ice batteries" where the chilling system is used during off peak time to freeze the coolant that is then used during the peak the next day. I don't see why data centers would not do the same if it pencils out.

flown-the-coop said:

---

Bear in mind Houston also has at least one if not more central chilling plants downtown. From the plant to and through the buildings is "closed loop" and the chillers at the plant are designed to recapture evaporated water if memory serves and / or just chill the looped watery using other methods.

For instance, I did learn they also have "ice batteries" where the chilling system is used during off peak time to freeze the coolant that is then used during the peak the next day. I don't see why data centers would not do the same if it pencils out.

---

Ok, the water lost to evaporative cooling is not recoverable. The "Ice" system was gone almost 30 years ago. It failed.

Houston does still have a central chilled water system that some buildings are on. But it is evaporative cooling. It just moves it from one place to another.

flown-the-coop said:

---

Bear in mind Houston also has at least one if not more central chilling plants downtown. From the plant to and through the buildings is "closed loop" and the chillers at the plant are designed to recapture evaporated water if memory serves and / or just chill the looped watery using other methods.

For instance, I did learn they also have "ice batteries" where the chilling system is used during off peak time to freeze the coolant that is then used during the peak the next day. I don't see why data centers would not do the same if it pencils out.

---

You got me going down a rabbit hole and, a la Google, I got this:

Quote:

---

Future data centers will significantly reduce, and in some cases entirely eliminate, external noise and light pollution due to a massive industry-wide transition toward advanced liquid cooling, strict zoning regulations, and dark-sky compliance. While legacy data centers have heavily disrupted nearby communities with stadium-like outdoor lighting and the constant 96-decibel drone of massive air-chilling fans, the explosive demand for high-density AI infrastructure is forcing a technological shift

---

It looks like the future ones are going to be much better in every way.

Here's to the future. As long as it doesn't increase our electricity bills, we should be good to go.

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

Edit: This no disrespect on your comments. What is your year and degree?

Urban Country Boy said:

---

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

---

Most refer to it as condensation.

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

---

Most refer to it as condensation.

---

I take from this you have no idea what you are talking about. Condensation is from the cooling coils. Evaporative cooling is from cooling towers, and is not recovered.

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

---

Most refer to it as condensation.

---

I take from this you have no idea what you are talking about. Condensation is from the cooling coils. Evaporative cooling is from cooling towers, and is not recovered.

---

I take it you have never had distilled water, spirits, beer, etc. Have you ever seen it rain and are you familiar with the water cycle?

Evaporative cooling towers work the same as all other methods of heating and cooling and its driven by heat exchange. Most of the heat exchange is from evaporation but only about 1% of water is evaporated per pass and the vapor plume can be equipped to condense that vapor back to recovered water via various methods.

My degrees are in accounting and information systems. But they are decidedly irrelevant to a basic understanding of the physical states of H2O.

BrazosDog02 said:

---

Queso1 said:

---

This reminds me of my law school's building. It was constructed decades ago a certain way in order to hold computers. The massive computers needed to process whatever data weighed tens of thousands of pounds. I'm now posting with a phone with exponentially more processing power.

I think these things will be obsolete in 10 years and nobody will clean them up. Just another blight on our country's landscape.

---

Don't blame me. I'm unapologetically completely anti progress. If I could push a button and send my county back to 1948, ship out everyone that moved here that's not an original founding family, I'd smash that button so fast I'd probably break every bone from my fingers to my shoulder….which would be a pain since there aren't hospitals where I am in 1948. But I'd roll those dice anyway.

---

Uh, relative to past years, the late 40's probably witness more societal changing tech growth than any other era. The application of nuke power, mass and personal transport, and 1st gen computers is huge by any standard.

Sorry but using 1948 for this example is a huge logic fail. However I do agree with some of your overall argument.

People need to understand this is greater than the nuclear arms race.

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

---

Most refer to it as condensation.

---

I take from this you have no idea what you are talking about. Condensation is from the cooling coils. Evaporative cooling is from cooling towers, and is not recovered.

---

I take it you have never had distilled water, spirits, beer, etc. Have you ever seen it rain and are you familiar with the water cycle?

Evaporative cooling towers work the same as all other methods of heating and cooling and its driven by heat exchange. Most of the heat exchange is from evaporation but only about 1% of water is evaporated per pass and the vapor plume can be equipped to condense that vapor back to recovered water via various methods.

My degrees are in accounting and information systems. But they are decidedly irrelevant to a basic understanding of the physical states of H2O.

---

You have no idea what you are talking about. This is not beer. ME '88. No, not 1 percent is evaporated. Reference my high rise example of 80 gallons a minute needed for make up.

You have no idea how cooling towers work. I do not disparage my fellow Ags but you need to sit this one out.

It's way more relatable to the impact of telegraph / phone / TV / Internet.

It's an information aggregator and coagulator but it is not the messiah folks paint it to be.

Nuclear bombs could kinetically destroy the world via push of button. Unless you put Ai in charge of your nukes, it will never itself be more dangerous.

HTH.

This whole discussion makes me wonder what would have been said during the high growth era of the industrial age.

tunefx said:

---

This whole discussion makes me wonder what would have been said during the high growth era of the industrial age.

---

Everyone always hates progress.

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

Urban Country Boy said:

---

flown-the-coop said:

---

CenTrio uses it and it's not 30 years old.

Central chillers is a big deal in many places, Houston being one of if not the biggest and the first.

https://www.centrioenergy.com/case-studies/houston-ice-battery/

And yes, several systems out there to recover 10%-30% of evaporated water, which is only a very small portion of the water utilized in the cooling towers.

Keep your mind open, you will learn more.

---

How do you recover evaporated water?

---

Most refer to it as condensation.

---

I take from this you have no idea what you are talking about. Condensation is from the cooling coils. Evaporative cooling is from cooling towers, and is not recovered.

---

I take it you have never had distilled water, spirits, beer, etc. Have you ever seen it rain and are you familiar with the water cycle?

Evaporative cooling towers work the same as all other methods of heating and cooling and its driven by heat exchange. Most of the heat exchange is from evaporation but only about 1% of water is evaporated per pass and the vapor plume can be equipped to condense that vapor back to recovered water via various methods.

My degrees are in accounting and information systems. But they are decidedly irrelevant to a basic understanding of the physical states of H2O.

---

You have no idea what you are talking about. This is not beer. ME '88. No, not 1 percent is evaporated. Reference my high rise example of 80 gallons a minute needed for make up.

You have no idea how cooling towers work. I do not disparage my fellow Ags but you need to sit this one out.

---

Tell us how they work and cite your sources. Maybe get the slide rule out and show us the calcs.

I notice you dropped your ice battery misinformation after that was cited. I don't think I need to clutter the thread up about cooling towers and how evaporative cooling works, but to think all water is flash steamed in evaporative cooling and lost permanently to the atmosphere never to become water again is frankly bizarre, but who knows what they taught in ME back in the mid 80s.

Fact remains only ~1% of water is evaporated during each pass and 10%-30% of that can be recovered. Rather than trying to shout down at someone who may have been familiar but who went back to check the facts, and looked into any new developments (I was not aware that they have figured out an approach to "ice battery" backups and is in use and not some failed technology from 30 years ago.

https://spxcooling.com/news/how-to-conserve-water-in-evaporative-cooling-towers/

Enjoy some reading and again an open mind is needed to learn. Spouting an engineering degree from 40 years ago is not open minded.

And yet the Fermi data center going in east of Amarillo plans to use 2.5 million gallons per day, with the potential to scale up to 10 million per day.

https://www.newschannel10.com/2025/10/17/david-prescott-breaks-down-water-agreement-with-fermi-america/

And the one they are trying to force through in Leon County is already talking about the deep wells they will need to drill.





So yes, tell us more about how they don't need water...

Interesting story on the news this morning that the anti-data center movement is being heavily funded by the Chicoms.