> The European Energy Efficiency Directive (EED) obliges data centers with an installed capacity of at least 500 kilowatts (kW) to report data including their energy and water consumption. In practice, few do so, despite the directive being in effect for three years.
I do not understand how that is even possible.
If any law is supposed to have any practical impact on the society, it needs to be enforced. Otherwise it is a guideline at best, not a law.
I think what would help is for these companies to each receive strict and hefty fines for non-disclosing the electricity and water consumption data they are obliged to disclose.
I assume that after that they would report all the required data very quickly.
It's probably because directives have to be read into law at the country level.
(Directives are binding on states. They transpose them into national law, possibly adding additional clauses as it goes past (EU directives set a minimum standard, not a maximum) at which point they're binding on people.)
Nobody ever said the EU was efficient. Look for a deadline for that transposition on the original directive.
You make a very good point. However, in this case, the deadline is two years in the past. From Article 12:
> 1. By 15 May 2024 and every year thereafter, Member States shall require owners and operators of data centres in their territory with a power demand of the installed information technology (IT) of at least 500kW, to make the information set out in Annex VII publicly available, except for information subject to Union and national law protecting trade and business secrets and confidentiality.
> 2. Paragraph 1 shall not apply to data centres used for, or providing their services exclusively with the final aim of, defence and civil protection.
Cool, thanks for looking that up. This gets us closer to where the fault lies. Has the Netherlands implemented the directive yet?
Chatbot says "partially", and "Article 12, concerning data centres, has already been implemented. The relevant Dutch decree was published in May 2024 and entered into force on 7 May 2024." and a link to https://wetgevingskalender.overheid.nl/Regeling/WGK026248
But I neither read Dutch, nor trust the chatbot to do so, so all I can do is report the link.
European Union has procedures for infringement or non-conformity in the reception of directives, but in practice, the fines are relatively small compared to a country budget (or how much money a country gains from taxes by not applying the directive) [1]
It's probably because that 500kw number is too arbitrary. I work in a single warehouse with robotics based logistics that uses 500kw. Nobody is asking us how much water we use.
Reality is if a regulation is too broad it tends to get overwhelmed and dies when it requires 100 people to enforce.
Even a basic in house data center can push 100kw with a few servers.
National governments kind of suck at enforcing these. In some cases, like the notorious Ireland Apple tax shelter, they may not want to enforce the rules because they feel they are benefiting from the evasion.
This kind of basic regulatory work is pretty thankless even when it's important. But I suspect the toxic politics of data centers will kick up some enforcement action sooner or later.
> In some cases, like the notorious Ireland Apple tax shelter, they may not want to enforce the rules because they feel they are benefiting from the evasion.
I thought it turned out the EU had no business in national tax affairs.
The line between a tax break and illegal state aid is exceedingly blurry. (It swung back and forth as it went through the courts, but the final ruling was "illegal state aid")
Extra-national laws are a real minefield, though. Ireland is in a great position without having too do very much of its own to have got there other than sign the right agreements.
The EU is in effect much closer to a confederation than a federation. A lot of the problematic aspects of how the EU is set up are attempts to work around the limitations imposed by not having sufficient public support to reform it into a more federal structure. To some that's a feature, of course, to others it's a significant nuisance.
And the practical effect of this is you get a race to the bottom: countries compete on how lax they are going to be in enforcing GDPR so as to attract American tech companies.
The Data Protection Commission in Ireland are remarkably lax about enforcing it because the government quite likes the tax revenue that Google, Apple and Meta brings them.
It's not a race to the bottom any more than competition lowering consumer prices is a race to the bottom. It's one of the few checks on governments just demanding more and more tax money for little return.
Perhaps this article is one step in the process of forcing disclosure. The Dutch parliament has recently (April 2026) clarified its law implementing the EED.
My account is banned, so for whomever having [showdead] on:
European directive is not a law, it's a directive.
It's up to the national governments to create applicable laws that are to the letter and to the spirit of a European directive. Then national bodies will start applying the law and govern its implementation/adherence.
An by "up to" I mean they are obliged... eventually.
And I wouldn't be so sure on being 'eventually' obliged, someone has to enforce it for that to happen and it's not looking like anyone has profit in it
The purpose of these laws is to give the state the option to crack down on anyone they feel like at any time. They are not meant to be obeyed, and that is obvious to anyone with eyes.
Playing the devil’s advocate: don’t “freedom of information” laws mostly apply to government agencies? If so, why would a private company provide this information?
I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry. Elecricity is a different issue, and governments should consider reporting requirements or something like that.
Being angry with datacenter providers is not the solution. And they also need to be built somewhere, and that place is probably best where electricity is abundant and/or cheap.
> If so, why would a private company provide this information?
Because the law says they have to.
Water usage in terms of total volume is meaningless and certainly an overblown concern on a global scale. In fact, it is not a concern at all on a global scale, but it can be a local problem. If a datacenter uses water normally intended for households and the infrastructure is not designed for this, it is definitely a problem. And that's a good reason for having to report water usage, not just to punish them, but also to plan distribution accordingly.
That isn't a good reason for them to do so. Laws need to be enforced to be effective. If no one is going to make them disclose the info, then the smart move it to just not do it, as they are.
My private company complies with laws that make sense to us and also with laws that we find annoying. We do that because we prefer the efficiencies and benefits of living in a high-trust society, and part of that is behaving that way ourselves.
“Smart” in a selfish way, but not in a bigger-picture kind of way.
I’d rather spend tax money on bridges (or tax rebates) than on bankrolling a muscular cadre of enforcers on the assumption that laws are meaningless and everyone will cheat without a cop breathing down their neck.
mate the reason why you can't just build a house without a permit is because the government needs to account how you're using existing public infrastructure, not because they just wanna throw bureaucracy at you (though it might be the case most of the times)
that's all there is to it, accounting for existing infrastructure, and somehow 500kwh is enough to invoke a reason to do so
> the water usage is heavily overblown compared to the agriculture industry
Data centers tend to concentrate their water usage within urban areas, causing a strain on local resources. The value of their output is much less tangible. Their water consumption scales up around tenfold if you consider the indirect usage from generating the electricity.
But on the other hand; yes, it would also help enormously if people ate less meat.
By “strain on local resources” are you just stating the fact that they use resources? Or are you trying to imply that it causes problems for those resources?
Companies don’t just buy land, connect to water and power, and then the utilities get caught off guard when the water system is suddenly strained by water going somewhere. There are processes for approving this before connection.
If a new project is causing a “strain on resources” then the utility planners have failed that community.
> If a new project is causing a “strain on resources” then the utility planners have failed that community.
Call it an echo chamber but when it comes to data centers that's the preponderance of news that I read. These two stories are about the county next door to mine. The first story highlights how the builders of the data-center continuously broke zoning laws. The second story from a month later describes how the board of commissioners let all that slide and approved construction to continue.
Those guys were fined a "record" $1.1 million, and:
> The state is mandating DataOne apply for and obtain the required permits or cease operations, but is giving the company 45 days to do so, according to a NJDEP spokesperson. During that time, regulators are allowing the facility to continue running the generators.
I agree that utility planners have their work cut out to plan for data centers. It's a different paradigm than what they're used to. I think they are being caught off-guard, but it's a matter of process and infrastructure. They typically plan around averaged or linear trends, but data centers can have volatile usage spikes, and supply can depend on e.g. seasonal rainfall which is challenging to predict.
Households with yards will have far more volatile water usage spikes related to seasonal rainfalls than a fully built out datacenter will.
A recent datacenter proposal near me was going to be built on a ~35 acre site. The entire site projected to use ~300 gallons of water a day when fully operational. That's including toilets, irrigation, etc. A household on a quarter acre lot in that area often uses 10-15,000 gallons of water in a month. 35 acres cut into quarter-acre-ish lots means a bit under 140 households. We'll cut it down and say 100 after streets and what not. That lot being a neighborhood like the others nearby would probably use at least a million gallons of water in a month, about 33,300+ gallons a day.
> A typical data center uses 300,000 gallons of water each day (equivalent to the demands of about 1,000 households), but large data centers can use an estimated 5 million gallons of water each day
Only if allowed to. Most data centres I've used have no connection to water from the data centre halls other than to the toilets. Using open loop water cooling is a cost cutting measure it's a choice whether or not to allow.
> we demonstrate that data centers are an overwhelmingly urban phenomenon, with 97.5% of facilities situated in metropolitan and micropolitan statistical areas
“Metropolitan area” is doing a lot of work. Every metropolitan area I’ve lived or worked in has had industrial areas, too. Those industrial areas have buildings that use a lot of electricity, water, or natural gas and produce a lot of noise.
This isn’t some unusual outcome. It’s like one of those maps that purports to show one thing but really just shows population density.
Most data centres in Australia (even the announced biggest one in the Southern Hemisphere) are close to a water processing plant in an industrial zone in Sydney. The same plant that supplies 85% of Sydney with drinking water. At least there's an upgrade.
> But on the other hand; yes, it would also help enormously if people ate less meat.
At first you had me, but then I realised it's not true if the price is building more data centres! That was just a trick to make the climate worse! Grr /s
Essentially, telling people to go vegan and get rid of their golf courses is going to piss them off, hence the reaction people have to being told that datacenters are more important than their comfort.
I'm personally willing to change my life to help the planet and nature in general to the point it doesn't affect my wellbeing (as in health and safety). For me, personally, the problem is not changing per se, but people finding no amount of change enough, always moving the goalposts with "but, what if" clauses.
The problem with datacenter water use is the state of the water leaving the data center. Agriculture and meat industry can reuse the water multiple times from cleaner to dirtier processes and return it to the nature with reasonable processing.
The additives used by datacenters in some cases can't be filtered or processed out practically or in economically viable ways, so the water returns back as unusable, esp. if they get water from a body of water and return it with added chemicals the process gets very damaging.
> The additives used by datacenters in some cases can't be filtered or processed out practically or in economically viable ways, so the water returns back as unusable, esp. if they get water from a body of water and return it with added chemicals the process gets very damaging.
Why would any water be returned?
Datacenters (and literally every industrial process that requires cooling) use water for evaporative cooling. The water evaporates, then returns as rain elsewhere.
The only water being "returned" I can think of would be from the intermittent or continuous blowdown systems to maintain TDS levels; but blowdown is a small fraction of the water consumed and is typically returned back via the city sewer system.
Roughly 80% of the water evaporates into the atmosphere. The remaining 20% is flushed out of the system to remove mineral build-ups and sent to local treatment plants.
It's complicated, and depends on the hardness of the water entering the system from the water supply as well, since blowdown is ultimately to control the level of dissolved solids in the water so it does not contrate enough to start coming out of solution. Rarely if ever is true demineralized water used in open loop cooling systems; it's simply too expensive to produce to then "throw away" into evaporation.
The previously stated 80/20 isn't a bad general estimate, but it varies depending on the specific site. 10%-25% of makeup water being lost to blowdown is a reasonable range.
As for papers about it, I was only able to find a few in my brief search, but they don't land on a single number since it depends on the ratio of dissolved minerals/solids in the circulating cooling-tower water compared with the incoming makeup water, which varies due to acceptability for the process and due to the incoming water quality.
This short PDF on "Mechanical Systems Water Efficiency Management" from the EPA goes over the concepts pretty well though
This document from the US DOD (UFM 3-230-13, Industrial Water Treatment Operation and Maintenance (2026), Chapter 5) says that most cooling towers operate at 3-10 CoC, with 5-7 CoC identified as a cost-effective range. Where CoC is Cycles of concentration described in the previous PDF, and a higher CoC means less blowdown and better water efficiency, with the trade-off of a greater risk of scaling, corrosion, and biological/water-chemistry problems.
We're talking about water that's released in general use, right? I would expect a closed cooling loop to have some biocides but have almost none of that leave the system.
And one datacenter had a weird bacteria that one time.
Yes, a single hamburger takes 620 gallons(!) of water to produce. Eat plants. Plenty of great plant protein alternatives available, plus they are healthier and don't depend on killing living beings.
Actually your premise is already wrong, although to be fair I'm not sure what bearing that has on the conclusion.
Here in France at least, companies can also be subject to FOIA requests for information pertaining to contracts they have with public entities. E.g. if a city publishes a tender to run tye local bike share program, then the company that gets the contract may have to directly answer (some) requests about that program.
But like I said, not sure if this applies to data centers which are, iiuc, mostly private ventures.
Not sure why you think your finger-in-the-wind wind understanding of future AI compute demand should trump the projections of companies who actually operate in this area?
But if you do end up being correct, then the data centers in question will just sit partially or fully idle, using no / a tiny fraction of the electricity and water they could use (which as others have said on this forum, water usage concerns are massively overblown in most data center builds anyway)
That's a perfectly valid concern, but isn't specific to data centers or their construction. That's like being mad at a factory that produces pens because you don't like what people are writing with them.
Not sure why you think the interests of companies who operate in this area, should trump environmental concerns for the people who live next to where they plan to build the data centers? I for one prefer quality of life for actual people over speculative profits for tech billionaires.
I'm also not sure why you think it's a good thing to produce a bunch of electronics using conflict minerals and let them sit idle? Surely you must see the negative "externalities" of this?
And as we all know, investments of vast amounts of capital by companies can never be contrary to the well-being of humans... See Purdue, Monsanto, Union Carbide, etc.
These are resources acquired from the governments in most places. You have to be able to plan and expand capacity which requires public funds.
I live right next to a waste valorisation plant[1]. They have to exactly specify what amount of water they use and what amount of water they then send out to the waste water treatment facility. It also has to be clean to a very specific standard before it leaves which is also checked.
We have datacenters in Switzerland too, but very little outrage or debate because there are strict rules in place in most places. They can build all they want but it costs significantly more to meet environmental requirements.
The valorisation plant next to my home has water cooling tower/fans significantly oversized just to make sure the noise is keep under the required level.
> I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry.
Apples and oranges comparison. We eat the direct products of water usage in the agriculture industry. It keeps people alive. It’s fallacious to compare that with water usage that goes to serving ads.
> And they also need to be built somewhere, and that place is probably best where electricity is abundant and/or cheap.
This is a Dutch website and here in the Netherlands we have demand issues. In these same areas where issues exist, data centres are being planned and built. There are waiting lists now for normal businesses who want a new connection to the net, or connections with more capacity. These are lists waiting for capacity to become available: The demand on the current net is already too high.
This is something that certain AI advocates are refusing to acknowledge. The concerns are legitimate and can only be alleviated with open access to information by companies and the state. Then it can be clear that the project is doable or not, and let the people decide, as they are the ones that will have to live near it.
> don’t “freedom of information” laws mostly apply to government agencies?
As others (including TFA) have noted, this is false because there's an EU directive requiring "private companies" to disclose this.
In any case, the "it's a private company" defense is as tired now as it was when it was about blocking people from social media, or debanking etc. Just because the legal system hasn't yet caught up to regulating companies in a certain way, doesn't mean that it's not in everyone's best interest (except perhaps the owners of the companies in question) to have such regulation.
> I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry.
The important distinction being that agriculture provides food, while data centers provide AI slop and advertising.
People keep comparing datacenter water use to agricultural water use, and it doesn't make sense to me. I can't tell if it's a bad-faith comparison or if people really know that little about how water is used in these industries.
The vast majority of agricultural water use is either tapped directly from groundwater, drawn directly from surface sources, or rain. The vast majority of datacenter water use is processed water from a municipal water utility. In other words, it is meaningless to compare the two, because only one is competing with residential use, and so only one is worth caring about.
That's not to detract from serious concerns about e.g. overutilization of the Colorado River, or the idiotic practice of irrigating deserts, but it is to say that comparing datacenters to agriculture makes very little sense.
Golf courses are a better comparison. Data centers use a fraction of the water that it takes to keep the grass green for a hobby enjoyed by a handful of people.
On a biological level, there'd be no problem providing a vegetarian diet for everyone. A healthful vegetarian diet is not hard to achieve, at least no harder than a healthful non-vegetarian diet.
From an infrastructure point of view, providing everyone vegetarian food sources overnight would not be possible, but that's not a realistic scenario. With a gradual transition, the transition would reduce overall resource use on basically all fronts, so there's no reason to expect extraordinary problems.
Of course, this is all theoretical unless people care enough to change their lifestyles.
Of course the reason that people don't care enough to change their lifestyles is because
> A healthful vegetarian diet is not hard to achieve, at least no harder than a healthful non-vegetarian diet.
Might be true but isn't sufficient
It's still way easier to have an omnivore diet than a strictly vegetarian one. Partly that's due to how society is currently structured, but people still ate meat long before society was structured this way. Because it was easier.
It doesn't make sense to me either, but in a different way...
We need the food, we'll die without food, telling people to eat less of whatever food, so a few "tech bros" can build a new datacenter is outright evil.
edit: considering this is HN and the downvotes, people here prefer datacenters over food
People here generally don't believe that the water usage in DCs is enough to affect agriculture or that the areas so short on water are the ones we should be growing food in anyways.
The downvotes are probably about reasoning about your your conclusions instead of the things that led you to them. Though it'd be nice if that had been discussion instead of silent downvotes.
If the numbers would make them look good they would publish their numbers.
> Being angry with datacenter providers is not the solution.
Belittling criticisms that way is childish and doesn't address a single one. Literally nobody on the whole planet, at least in public discourse, is using "I am angry" as an argument in any of this. Yet it's the first thing that is always brought up. It's just a cheap, thought-terminating stereotype.
You are totally right to point this out, and I apologize for the whiplash. Let's clarify exactly how overblown data centre water usages is compared to the agriculture industry.
In data centres all that water is so that some wimpy cunt gets to hit return after spending no more than 5 minutes in an half baked prompt so he can insert the answer into his pitch deck, or PR review — and now with pacing the frontier models, he actually doesn't need to insert it, it's inserted automatically wherever it might fit. When it comes to agriculture, that is to produce things like food. This requires obviously much more water, which means that's what we really should be cutting down. Agents don't need food so the solution is just to replace humans with agents.
If you have any other stubborn question resisting clarification just hit me up, and I'll explain it to you immediately and clearly.
> it's probably because it makes sense economically no?
Only because the government forces the cost of the water to be $0 for those who got in early, a policy that's causing extreme issues ecologically and causing the collapse of tons of environments and screwing over many millions downstream.
If we actually allowed the real price today it wouldn't be anywhere near economically sane to do such a thing.
Hmm. I live near some water intensive cultivations (olives, almonds, not USA) in a relatively dry area, but there's dam water at very cheap prices and obviously, for these products it's required, because otherwise you wouldn't have olive oils being sold at current market prices (already quite higher than just some 6 years ago), they would be astronomically higher. But this is to serve global markets. People in the region have done olive oil since forever without dam water, just not at this industrial scale.
But compared to meat and the whole chain of production required for that it certainly pales many moons below. Well, everyone loves their bacon.
> but there's dam water at very cheap prices and obviously, for these products it's required
So yeah, an example of the government coming in and ensuring cheap water for this crop production at the expense of those downstream. You agree, if the government didn't artificially and massively change the hydrology of the area, this industrial level wouldn't really be economically possible. This can make sense and not be terribly destructive depending on the region.
I and the poster above were probably talking about the American Southwest. In this case, its even more extreme. Old land owners got the rights to flood their fields from the river water and dig massive canals to turn extremely arid deserts green. They get the right to have that water for free, forever, because the government gave them the land at one point in time in the past and allowed for that taking based on deeply flawed hydrological models. They use that for agricultural products sold worldwide, including growing alfalfa for Saudi Arabia in the deserts of Arizona[0]. The government won't go in and actually start putting restrictions or calling back their grants to that water allocations, and its resulting in environmental catastrophe for many parts of the region.
That does seem a bit extreme - in dam water you still pay, just much less because the water is also not treated in the same way as regular consumption water, storage is part of a second effect of electricity-gen damns many times, it enables agriculture and other things for communities, etc.
They refuse it because if they would say "1 million liters a year!" the knee-jerk reaction will be completely predictable even though that's just tiny compared to a small swimming pool or industrial or agricultural site.
The debate is completely based on emotion and anger right now, so sharing information will only give the emotional/angry folks more mud to throw with. The majority of people has no clue about how to put things like energy use and water use in perspective, and they're also not interested in doing that. Because they have decided to green-wash their AI-fears as environmental fears... which is a big fat joke because datacenters aren't significant polluters or consumers compared to almost all other industries.
Look, I get it, many people are afraid of the societal impacts of AI, or distrust it in another way. But let's be honest and focus on _that_ instead of naively acting as if our anti-datacenter sentiment comes from environmental concerns (the last few decades have shown we don't GAF about the environment).
>Look, I get it, many people are afraid of the societal impacts of AI, or distrust it in another way.
In the Netherlands it is not just about AI datacenters, even before the rise of AI pushback against Meta's planned datacenter made them cancel it. There has been a lot of criticism of the municipality that allowed Microsoft to expand its datacenters.
The argument is that this is a small country that already has to make tradeoffs to get to its desired energy mix (wind turbines near cities on land) and then these datacenters are built that consume the power that is generated by sacrificing the little remaining open space.
In the last few years there is actually a electricity shortage in some regions so "we can hook up this 200 MW datacenter but not your house" is a bad look.
It is completely irrelevant why they choose to refuse to provide this information.
From what I understand, disclosing this data is not optional. It is a law. And if they have willfully chosen to operate in a country where such a law applies, they have to obey that law.
If they are taking reasonable steps to mitigate consumption and pollution then they should be able to clearly explain that to the public. The fact they won't suggests they aren't.
Just because people haven't cared about the environment before doesn't mean they shouldn't start now. And just because there are worse existing polluters doesn't mean we shouldn't resist new ones.
A new golf course used to be seen as a great thing to have in your back yard and now we know their pesticide use increases the risk of Parkinson's. People are waking up and caring more. We need to have a sustainable society, not a profit obsessed one.
HN has never been friendly to comments like this, have you ever tried explaining something clearly to an emotionally triggered crowd? There are many examples, masks, vaccines, gun laws...
Why is the crowd emotionally triggered? I posit it's because of all the secrecy and bad faith. Data center builders and their supporters in government brought this opposition on themselves.
I posit it's because people say it's killing and stealing art and jobs and destroying farmland (it's not). In my neck of the woods, people show up to the actual meetings and loudly protest and have zero intention of listening.
It is doing those things too, but first and foremost on most people's minds is the issue of power bills, water availability and cleanliness, other local pollution of varying kinds, local taxes, and eminent domain for private profit.
I’m European and I’m extremely sure that when 90% of people hear "x m^3" they don’t think of it as "1000*x litres". And the majority of people likely can’t recall on the spot that 1 l = 1 dm^3 (a unit nobody uses), even if they remember that you multiply by 1000 to go from m^3 to dm^3. In any case most people don’t do conversions like that unless someone specifically asks them to.
Volumes are in general very unintuitive to people due to the cubic scaling. It’s further confused by the fact that the widely used decilitre, centilitre, and millilitre do scale linearly, unlike cubic metres. So 1 l = 1 dm^3 but 1 dl = 100 cm^3…
I don't know where do you live but even my small nephews and nieces understand that a 1 meter cube of water is A LOT OF WATER, because mentally, we have lots of 2 liters bottles to compare it to
Most of it comes from concerns about water and electricity rates sky rocketing and also simply not being available when humans need it. At least at the local level.
But they should be made to report this. AND their use of water should be regulated - they don't need to use any water at all, it's just cheaper for them - and they should be contributing enough to mitigate negative effects on energy prices.
The problem isn't data centres. The problem is not pricing in and regulating externalities, whether caused by data centres or other things.
It would also quickly show us who actually cares about the environmental effects and who are using it as an excuse.
One cubic metre of water is 1000 liters. A typical 25m swimming pool ("short course" / half of olympic size) is 20-25m wide. That's at least 500 m^2 surface area. At an average depth of only 1m (very conservative even for a recreational pool), that is 500,000 liters just to cycle the water once a year. Estimates I can find suggests 3-7 times as much as that on a yearly basis for a typical recreational pool.
So unless you're talking about a tiny private residential pool you're at least an order of magnitude off.
As someone who has been maintaining a pool for several years, you really hope you don't have to cycle the full amount of water once a year. You're really doing something wrong or incredibly unlucky if that's the case. That's at least true for the climate I live in. If I lived in a different climate, I'd probably forego having a pool.
It was incredibly painful when I had to drain the pool for deep maintenance a few years ago, and I'm not looking forward to having to do it again in a little while.
Also, most residential pool owners don't have half olympic pools in their yard. They're usually much smaller than that.
> Also, most residential pool owners don't have half olympic pools in their yard. They're usually much smaller than that.
Hence my caveat at the end. When talking about swimming pools, tiny residential pools is certainly not what springs to mind for me.
Most pools near me are in fitness centres etc. and are 25m pools.
And I certainly wouldn't go swimming in one of those if the water wasn't cycled regularly. Thankfully the data suggest they are.
If we're talking tiny little residential pools, then using a swimming pool worth of water isn't very shocking. Unless I'm way off, a 10-15m shower every day would use a similar amount of water in that case.
> I certainly wouldn't go swimming in one of the water wasn't cycled regularly
Once again speaking as someone who has maintained pools for a while, if the water was properly monitored and maintained I probably wouldn't mind. I'd absolutely take a pool that's actively properly maintained but hasn't had a full replacement in a few years over one that's poorly maintained but was fully drained and refilled a couple of months ago.
From experience while my son was swimming competitively, they had to drain most of the pools he swam in completely more than once a year because of various contamination, in addition to the ongoing cycling...
I can't imagine the state they'd be in after a year.
Once again, if its monitored and maintained properly, sure why not? Partial cycling when things that can't be filtered get too high (like TDS or CYA for an outdoor pool) but otherwise make sure sanitization is happening properly, pH and other metrics stay balanced, and good filtering is happening, I don't see a problem.
> more than once a year because of various contamination
Sounds like that falls into the "unlucky" category.
If its that you don't like thinking about tons of other bodies have touched that water you're swimming in, remember that dinosaurs pooped in much of the water you're drinking today.
> Partial cycling when things that can't be filtered get too high (like TDS or CYA for an outdoor pool)
Partial cycling in a high traffic environment seems like it is clearly ending up being a lot of water when every number I can find suggests it is generally expected to add up to several times the total pool volume per year.
> Sounds like that falls into the "unlucky" category.
To me it feels like it falls in the "expected" category somewhere that is dealing with hundreds of swimmers including small children every day.
Keeping in mind this was across dozens of pools - he swam competitively, not a one-off or rare occurrence in a pool or two.
It seems like we're mixing up terms here and miscommunicating. I thought we were talking about actually replacing the water here (as in, down the drain, fresh water from the tap), not just filtering as in recycling water through filtration systems. The total volume of the water should go through filters every few hours. I'm 100% not for the idea of only "partial filtering".
> To me it feels like it falls in the "expected" category somewhere that is dealing with hundreds of swimmers including small children every day.
Nah, that is definitely not normal to have to completely replace the water multiple times in a year because of water contamination events happening. The water can usually be sanitized. It might require a day's worth of hyperchloronation, but not often an entire drain and fill. That will happen often, probably several times a year. Usually only if your CYA is too high will it require you to actually drain a significant volume of water, which I agreed was one of the reasons to at least do a partial drain and really only applies to outdoor pools. Doing a hyperchloronation will likely increase your TDS, which won't go down until you drain those contaminates. But usually you can get those numbers back in an acceptable range without having to drain and fill the whole pool.
If the pool was properly maintained, what state do you expect the water to be in after a year without a total drain and fill? Do you have experience actually maintaining a pool?
> It seems like we're mixing up terms here and miscommunicating. I thought we were talking about actually replacing the water here (as in, down the drain, fresh water from the tap)
We are. I changed it while you were writing.
Everything I'm seeing suggests typical loss to backwash of the filters for a short course pool is in the 5k - 10k liters of water a day range, possibly less if you invest in expensive additional filters. 5k a day is ~1.8 million liters a year.
To get below a full replacement every year for a small short-course pool, you need to lose less than 1370 liters a day, or ~1.4 m^3
And besides that: news sites love comparing water usage to things like bathtubs, the average household or swimming pools. The headline would likely be "datacenter uses 100 swimming pools of water per year!" and not "datacenter uses 6 million liters of water per year!"
The people who don’t care about environmental impact and resource use are the same people pushing for more and faster datacenter deployment. The Republican Party is the one getting rid of environmental protection rules, it’s the party most financed by fossil fuel companies, and the party that was/is against solar and windmill development. The party that gutted all the regulatory agencies and is colluding with the AI industry.
Maybe I'm being naive, but the policies needed here feel quite simple.
First, since data centers likely need some kind of treated (i.e. demineralized) water, and they want to use local municipal supplies, then they can do so only if they pay full market rate for that water. I would push for them to pay 1% to 5% above market rate to help offset other costs, but market rate at minimum. At no point should homeowners and/or residential customers subsidize this usage.
Second, commercial users (including data centers) should only get water produced by treatment plants in excess of residential needs. Plants can only treat so much, pipes can only carry so much, and priority goes to residential first.
Third, ground water is off limits for any commercial/industrial evaporative cooling uses. If data centers want to cool that way, then they need a different source of water.
If they pay market rate, the clearing price will go up making things more expensive for everyone. If they go off-grid they make the machines needed to produce water or electricity or whatever more expensive. If they borrow money, they make government bond yields go up and cause governments to start thinking about cutting pensions and stuff, not to mention increased mortgage rates for whoever wants to buy a home or is on a variable rate mortgage. There's no world in which this AI stuff doesn't increase costs for the rest of us.
Well, even if you're right, try proposing those bills and then reckon with the deafening wails of the "free-market-big-government-taxation-is-theft" crowd. And the politicians that are in the pockets of BigCo that will say that kind of stuff.
Policies are simple a lot of times. Politics against the interest of big money isn't.
The main thing is evaporative cooling is incredibly wasteful. The water should be returned to the system, if it's going to be used, and not sent into the atmosphere. This is especially important in regions with limited access to water, which, ironically, is where many data centers seem to be located.
I think it would be helpful for people in this thread to understand the differences between EU Regulations and Directives :
• EU Regulation: A binding legal act that applies immediately and identically in all EU countries as soon as it takes effect, without needing national laws.
• EU Directive: A legislative act that sets a mandatory goal that all EU countries must achieve, but allows each nation to pass its own laws to implement it.
It takes time for directives to take effect, unless a member state's national law already implements them.
EU is in such a weird position. They are realizing that software, robotics, and generally automation through AI is the industrial future, but they seem to be paralyzed to actually start the build-up that will feed that frontier industry and economy.
Assuming you are right, the "build-up" of more physical datacenters isn't really the difficult part, training frontier models is. The open-weight Chinese models are several months behind frontier models at this stage and as long as you're not interested in burning cash to stay in the lead you can just wait it out. Many signs (upcoming ridiculous IPO figures and benchmarks) point to the current LLM paradigm reaching a plateau.
I think that currently the EU is focusing on strengthening itself from a military perspective due to US policy shifts, so the EU just doesn't have the funds to divert to AI without impacting day to day life of people, thus prioritizing safety and perhaps energy production, as electricity is considerably more expensive here than in the US or China. If you ask the people who vote, they'll put safety, grocery and gas prices way above EU's LLM capabilities.
> so the EU just doesn't have the funds to divert to AI without impacting day to day life of people
The US Government is not a major funder of frontier labs (who are almost exclusively VC and private money funded)...
> as electricity is considerably more expensive here than in the US or China
Yes, there are government policy reasons behind why this is the case. Some would argue that these policy decisions are a net negative for the EU.
> If you ask the people who vote, they'll put safety, grocery and gas prices way above EU's LLM capabilities.
Fair point, these are good things to focus on. However, EU citizens are going to love it when there are no frontier-tech opportunities left in their countries, with everything in this area having to be imported from the US and China (note that just because Chinese firms make open weight models available now, this could easily change very quickly in the future for a variety of reasons).
> just because Chinese firms make open weight models available now, this could easily change very quickly in the future for a variety of reasons
China is not providing "open weight" models just because they think it's good for humanity. It is a competitive strategy to stay relevant at the forefront of the industry. The reality is that China is severely compute-constrained mostly due to Nvidia chip and ASML embargo. Nearly all available compute capacity has to go toward training, leaving very little for inference. It is why they release their models openly allowing third parties in EU and USA to host inference for them.
This will change when China reaches a critical scale of domestic AI chip production, which is inevitable.
The problem is that EU has no money. France is virtually broke. Germany and eastern/central Europe is hemorrhaging its most valuable industries to cheaper Chinese competition. The whole bloc is carving out $100 billion military aid packages for Ukraine basically every year with no end in sight. And now EU is getting into a trade war with China despite having little leverage. Energy costs were bad even before the Ukraine war and Iran, and now they're at the boiling point.
It's why the USA and Chinese vultures are swarming. It's a bad situation.
Yes, I'm sure that EU doing its best to ignore building any technology frontier capabilities in its population will lead to a good solution for this lack of VC funding... ;-)
No shortage of visionaries in EU leadership these days :-)
The EU already has several hyperscaler data centers and many more colo-style data centers. Plenty of compute power exists in EU; this new generation is about purely inference capacity.
In the US, the political backlash against data centers is also about the inability of government to be transparent about these buildouts (eg. City councils signing NDAs and building commissions quietly approving development plans without community input), the inability for utilities/regulators to actually hold data center developers to promises made (usually regarding resource/ utility usage), more complaints about affordability (cost of utility bills rising), additional hate against the unlikable tech CEOs (who aren't charismatic and whom have a long list of previous violations), fear of what AI will bring (the promised white collar revolution, promises of "infinite AI immigrants" in an age where immigration has been demonized), etc.
Only a few of these are about the "factory" itself. Lots of them are about the people trying to retain some control of their communities. Some of them are about the speed of the buildout (eg. compromises in water use / plumbing and noise / pollution created by the ad hoc on site electrical generation).
The data center CEOs are starting to admit that their NDA-blitzscale strategy may have backfired because communities are pissed and the NDAs prevented anyone from making a positive case for the buildouts while the rage was boiling over.
Love this framing, most industrial warehouses easily use 500kw in power (the EU level for self reporting water consumption), but nobody is drumming up a water war on them.
For whatever reason people have clung onto this idea that the "Gotcha" to slow down AI is to worry about water consumption, and it's an idiotic gotcha.
I don't know which data center it is, some remote location of Canada, that has 100 million litres of fuel trucked up each month for it's generators. I don't think that has a long term plan behind it because 100m liters of anything is a lot, and with the price of fuel now it's definitely not a small line item.
Btw I heard this second hand a friend in the mining industry.
> According to the latest figures from Statistics Netherlands (CBS), data centers used 5.1 billion kilowatt-hours of electricity in the Netherlands in 2024. That is 4.6 percent of the country’s total electricity consumption, a near doubling in five years. Grid operator TenneT projects that data centers will consume 10 to 15 percent of the Netherlands’ electricity by 2030.
Yeah, so 2030 sounds like a long way off. It's 3 years away. We are not going to be meeting any climate targets.
Smart move. Why would they? Maybe there's laws that say they have to, but if no one is going to enforce said laws, then they maintain a competitive advantage by "refusing" to say.
Can someone explain the data center water usage? Data center don't drink water. It cools down the machine.
What's preventing them getting cool water from a lake and draining to it again to cool down? Where is water going after usage since there is literally zero water consumption?
>What's preventing them getting cool water from a lake and draining to it again to cool down?
Some do, but most don't since the requirements for using cooling water are relatively strict due to the pipes, equipment, and heat exchangers the water flows through. The water used needs to be free of any biological contaminants (such as bacteria and algae), maintain proper pH levels, and are often then treated with other chemicals such as biocides.
Lake water certainly can be treated to the level needed and used, it's just often cheaper to buy domestic water from the local utility than to construct and operate a small scale water treatment system along side the industrial process.
> Where is water going after usage since there is literally zero water consumption?
Most of the water is "consumed" via open loop evaporative cooling systems, so the water is never really "used" as in destroyed or anything. The majority of it is evaporated and returns to the land as rain and the rest is returned via sanitary drains, to city sewer systems.
Evaporation losses will move the water to a different locality. DC's not only cool servers, but control humidity by adding or removing water to the air.
Warm water fed back into the ecosystem will (relatively) quickly cause unwanted changes to the lake's biosphere.
Unless it's properly filtered, returned water will be contaminated by material in the cooling system.
It puts local demand on the water system competing with human need.
Didn't you know? Once you heat up water that's it, you can't use that water ever again. It's even worse if you make it evaporate, not just useless but GONE.
Now try again with local and short term impacts on the whole ecosystem instead of saying "Ackchyually water never leave the atmosphere so it's unlimited"
It cools down the machines by evaporating, and is then released into the atmosphere. That also means the water needs to be clean, pretty much drinking water, because otherwise the evaporation would deposit minerals and stuff.
Turning water into steam uses much more energy than heating water, which means it's a much more effective means of cooling.
Imagine fretting about water and electricity usage in europe, where we'll never ever have a datacenter build-up because electricity prices are too high anyway.
This figure is about the water consumption only. The estimate about the electricity consumption, in that same article, is much higher than that.
The main problem is, in my opinion, that those are just estimates. If we are supposed to accommodate the current and potential future needs, we need real data. Not estimates.
Clause Opus 5.5 reports that based on the 5.1 billion kWh used it's about 9.2 billion liters in Europe (cited from the Water Usage Effectiveness WUE industry-average). About 1.8 liters per kWh.
In the US, the rhetorical trick is to include the 3.4T* gallons of water utilized for power generation along with cooling (160B) to make this number super scary when water was utilized for power generation because it's a good cheap relatively clean power source so the datacenters were built near hydroelectric plants.
We can't do much about the 160B, but we could do a lot to build more power generation and storage. As just one example, in west Texas, natural gas is flared into the sky en masse from depleted oil wells because methane is a more powerful greenhouse gas than CO2. Some sites could be tapped to hit hundreds of megawatts to a gigawatt of power generation. And any power you don't need can be exported to the Texas power grid so it can make money the day you turn it on. And if you're willing to spend a little more, you can sequester the CO2 and make the power generation mostly green.
Cue crickets from Silicon Valley on this one because it's neither nuclear nor renewable, but the crypto people love it so I would guess they'll eventually come around. That they haven't already is one facet of my disenchantment with tech culture. Their absurd belief in Star Trek Voyager-like scenarios playing out with AI in the next decade the other major one.
*Which is slightly more than the 3.3T gallons of water Americans used to water their lawns. So, ya know, where's the outrage?
Less than they do if there were people living on that same land, and way less than they use in China. Get over it, data centers and progress is the future.
From the article:
> The European Energy Efficiency Directive (EED) obliges data centers with an installed capacity of at least 500 kilowatts (kW) to report data including their energy and water consumption. In practice, few do so, despite the directive being in effect for three years.
I do not understand how that is even possible.
If any law is supposed to have any practical impact on the society, it needs to be enforced. Otherwise it is a guideline at best, not a law.
I think what would help is for these companies to each receive strict and hefty fines for non-disclosing the electricity and water consumption data they are obliged to disclose.
I assume that after that they would report all the required data very quickly.
It's probably because directives have to be read into law at the country level.
(Directives are binding on states. They transpose them into national law, possibly adding additional clauses as it goes past (EU directives set a minimum standard, not a maximum) at which point they're binding on people.)
Nobody ever said the EU was efficient. Look for a deadline for that transposition on the original directive.
You make a very good point. However, in this case, the deadline is two years in the past. From Article 12:
> 1. By 15 May 2024 and every year thereafter, Member States shall require owners and operators of data centres in their territory with a power demand of the installed information technology (IT) of at least 500kW, to make the information set out in Annex VII publicly available, except for information subject to Union and national law protecting trade and business secrets and confidentiality.
> 2. Paragraph 1 shall not apply to data centres used for, or providing their services exclusively with the final aim of, defence and civil protection.
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ%3AJOL...
Cool, thanks for looking that up. This gets us closer to where the fault lies. Has the Netherlands implemented the directive yet?
Chatbot says "partially", and "Article 12, concerning data centres, has already been implemented. The relevant Dutch decree was published in May 2024 and entered into force on 7 May 2024." and a link to https://wetgevingskalender.overheid.nl/Regeling/WGK026248
But I neither read Dutch, nor trust the chatbot to do so, so all I can do is report the link.
I'm no lawyer but based on the legal text in your link I'd say the law went into effect two years ago at least.
The Dutch newspaper participating in this article (https://www.trouw.nl/duurzaamheid-economie/nederland-weet-am...) states that a quarter of the statistics are public, but also that within Europe, that's one of the more transparent numbers.
It looks like nobody is really enforcing these laws anywhere.
Yes it says it has been implemented, both the Raad van State and the king signed it, it’s been in effect since then.
This doesn't seem like the type of comment it makes sense to create a throwaway for.
maybe they work for a datacenter
So maybe your proposed hefty fines actually need to be imposed on the states which haven't implemented the directive?
European Union has procedures for infringement or non-conformity in the reception of directives, but in practice, the fines are relatively small compared to a country budget (or how much money a country gains from taxes by not applying the directive) [1]
[1] https://commission.europa.eu/law/application-eu-law/implemen...
It's probably because that 500kw number is too arbitrary. I work in a single warehouse with robotics based logistics that uses 500kw. Nobody is asking us how much water we use.
Reality is if a regulation is too broad it tends to get overwhelmed and dies when it requires 100 people to enforce.
Even a basic in house data center can push 100kw with a few servers.
National governments kind of suck at enforcing these. In some cases, like the notorious Ireland Apple tax shelter, they may not want to enforce the rules because they feel they are benefiting from the evasion.
This kind of basic regulatory work is pretty thankless even when it's important. But I suspect the toxic politics of data centers will kick up some enforcement action sooner or later.
(Why is it so toxic? Well, the explicit promise from AI companies of mass unemployment combined with higher electricity prices, and maybe even destroy the world https://www.theguardian.com/technology/2026/sep/09/anthropic... is why.)
> In some cases, like the notorious Ireland Apple tax shelter, they may not want to enforce the rules because they feel they are benefiting from the evasion.
I thought it turned out the EU had no business in national tax affairs.
The line between a tax break and illegal state aid is exceedingly blurry. (It swung back and forth as it went through the courts, but the final ruling was "illegal state aid")
Extra-national laws are a real minefield, though. Ireland is in a great position without having too do very much of its own to have got there other than sign the right agreements.
Because it is an EU-DIRECTIVE.
Directives are not the same as regulations .
Directives are goals that must be achieved by adapting existing national laws or introducing new ones, and that takes time.
Sadly, the EU is also lax when it comes to enforcing regulations. See GDPR.
GDPR enforcement belongs to individual country information commissioners, unfortunately.
The EU is not as federal as the US and has basically no "federales" law enforcement of its own.
The EU is in effect much closer to a confederation than a federation. A lot of the problematic aspects of how the EU is set up are attempts to work around the limitations imposed by not having sufficient public support to reform it into a more federal structure. To some that's a feature, of course, to others it's a significant nuisance.
And the practical effect of this is you get a race to the bottom: countries compete on how lax they are going to be in enforcing GDPR so as to attract American tech companies.
The Data Protection Commission in Ireland are remarkably lax about enforcing it because the government quite likes the tax revenue that Google, Apple and Meta brings them.
It's not a race to the bottom any more than competition lowering consumer prices is a race to the bottom. It's one of the few checks on governments just demanding more and more tax money for little return.
No as much lax as lacking resources to deal with widespread violations. They probably focus on the most egregious cases.
Isn't GDPR a directive as well?
No, the clue is in the name, "General Data Protection Regulation". If it were a directive, it would be called GDPD.
Perhaps this article is one step in the process of forcing disclosure. The Dutch parliament has recently (April 2026) clarified its law implementing the EED.
My account is banned, so for whomever having [showdead] on:
European directive is not a law, it's a directive.
It's up to the national governments to create applicable laws that are to the letter and to the spirit of a European directive. Then national bodies will start applying the law and govern its implementation/adherence.
An by "up to" I mean they are obliged... eventually.
It is not, I can clearly see it without it on.
And I wouldn't be so sure on being 'eventually' obliged, someone has to enforce it for that to happen and it's not looking like anyone has profit in it
The purpose of these laws is to give the state the option to crack down on anyone they feel like at any time. They are not meant to be obeyed, and that is obvious to anyone with eyes.
Playing the devil’s advocate: don’t “freedom of information” laws mostly apply to government agencies? If so, why would a private company provide this information?
I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry. Elecricity is a different issue, and governments should consider reporting requirements or something like that.
Being angry with datacenter providers is not the solution. And they also need to be built somewhere, and that place is probably best where electricity is abundant and/or cheap.
> If so, why would a private company provide this information?
Because the law says they have to.
Water usage in terms of total volume is meaningless and certainly an overblown concern on a global scale. In fact, it is not a concern at all on a global scale, but it can be a local problem. If a datacenter uses water normally intended for households and the infrastructure is not designed for this, it is definitely a problem. And that's a good reason for having to report water usage, not just to punish them, but also to plan distribution accordingly.
> Because the law says they have to.
That isn't a good reason for them to do so. Laws need to be enforced to be effective. If no one is going to make them disclose the info, then the smart move it to just not do it, as they are.
My private company complies with laws that make sense to us and also with laws that we find annoying. We do that because we prefer the efficiencies and benefits of living in a high-trust society, and part of that is behaving that way ourselves.
“Smart” in a selfish way, but not in a bigger-picture kind of way.
I’d rather spend tax money on bridges (or tax rebates) than on bankrolling a muscular cadre of enforcers on the assumption that laws are meaningless and everyone will cheat without a cop breathing down their neck.
mate the reason why you can't just build a house without a permit is because the government needs to account how you're using existing public infrastructure, not because they just wanna throw bureaucracy at you (though it might be the case most of the times)
that's all there is to it, accounting for existing infrastructure, and somehow 500kwh is enough to invoke a reason to do so
> the water usage is heavily overblown compared to the agriculture industry
Data centers tend to concentrate their water usage within urban areas, causing a strain on local resources. The value of their output is much less tangible. Their water consumption scales up around tenfold if you consider the indirect usage from generating the electricity.
But on the other hand; yes, it would also help enormously if people ate less meat.
> causing a strain on local resources
By “strain on local resources” are you just stating the fact that they use resources? Or are you trying to imply that it causes problems for those resources?
Companies don’t just buy land, connect to water and power, and then the utilities get caught off guard when the water system is suddenly strained by water going somewhere. There are processes for approving this before connection.
If a new project is causing a “strain on resources” then the utility planners have failed that community.
> If a new project is causing a “strain on resources” then the utility planners have failed that community.
Call it an echo chamber but when it comes to data centers that's the preponderance of news that I read. These two stories are about the county next door to mine. The first story highlights how the builders of the data-center continuously broke zoning laws. The second story from a month later describes how the board of commissioners let all that slide and approved construction to continue.
https://www.cpr.org/2026/08/07/weld-county-ai-data-center/ https://www.cpr.org/2026/09/10/weld-county-officials-approve...
But it's not just my state:
https://www.theguardian.com/environment/2026/sep/25/new-jers...
Those guys were fined a "record" $1.1 million, and:
> The state is mandating DataOne apply for and obtain the required permits or cease operations, but is giving the company 45 days to do so, according to a NJDEP spokesperson. During that time, regulators are allowing the facility to continue running the generators.
That fine is a slap in the wrist.
You say that, but over and over we see things like
> You can't build an apartment complex here, because our water/sewer system is already over capacity
/ignores that statement and builds the apartment complex because "omg, we need more housing" arguments are used to not pay attention to consequences
/water and sewer system can't keep up and all kinds of expensive problems occur
I agree that utility planners have their work cut out to plan for data centers. It's a different paradigm than what they're used to. I think they are being caught off-guard, but it's a matter of process and infrastructure. They typically plan around averaged or linear trends, but data centers can have volatile usage spikes, and supply can depend on e.g. seasonal rainfall which is challenging to predict.
Households with yards will have far more volatile water usage spikes related to seasonal rainfalls than a fully built out datacenter will.
A recent datacenter proposal near me was going to be built on a ~35 acre site. The entire site projected to use ~300 gallons of water a day when fully operational. That's including toilets, irrigation, etc. A household on a quarter acre lot in that area often uses 10-15,000 gallons of water in a month. 35 acres cut into quarter-acre-ish lots means a bit under 140 households. We'll cut it down and say 100 after streets and what not. That lot being a neighborhood like the others nearby would probably use at least a million gallons of water in a month, about 33,300+ gallons a day.
> A typical data center uses 300,000 gallons of water each day (equivalent to the demands of about 1,000 households), but large data centers can use an estimated 5 million gallons of water each day
https://www.brookings.edu/articles/ai-data-centers-and-water...
Only if allowed to. Most data centres I've used have no connection to water from the data centre halls other than to the toilets. Using open loop water cooling is a cost cutting measure it's a choice whether or not to allow.
Mmm most datacenters i went to were in the middle of nowhere, sharing land with farms (and oil/gas mining)
> we demonstrate that data centers are an overwhelmingly urban phenomenon, with 97.5% of facilities situated in metropolitan and micropolitan statistical areas
https://www.nature.com/articles/s44284-026-00487-z
“Metropolitan area” is doing a lot of work. Every metropolitan area I’ve lived or worked in has had industrial areas, too. Those industrial areas have buildings that use a lot of electricity, water, or natural gas and produce a lot of noise.
This isn’t some unusual outcome. It’s like one of those maps that purports to show one thing but really just shows population density.
Most data centres in Australia (even the announced biggest one in the Southern Hemisphere) are close to a water processing plant in an industrial zone in Sydney. The same plant that supplies 85% of Sydney with drinking water. At least there's an upgrade.
The point is that they create a heavy and volatile local water demand on those metropolitan areas.
Industry uses resources and they pay for those resources at significant cost. It’s not a surprise when the pipes are enabled.
Also using N as the denominator, instead of weighting by floor area?
> But on the other hand; yes, it would also help enormously if people ate less meat.
At first you had me, but then I realised it's not true if the price is building more data centres! That was just a trick to make the climate worse! Grr /s
Essentially, telling people to go vegan and get rid of their golf courses is going to piss them off, hence the reaction people have to being told that datacenters are more important than their comfort.
It's not black and white as it looks.
I'm personally willing to change my life to help the planet and nature in general to the point it doesn't affect my wellbeing (as in health and safety). For me, personally, the problem is not changing per se, but people finding no amount of change enough, always moving the goalposts with "but, what if" clauses.
The problem with datacenter water use is the state of the water leaving the data center. Agriculture and meat industry can reuse the water multiple times from cleaner to dirtier processes and return it to the nature with reasonable processing.
The additives used by datacenters in some cases can't be filtered or processed out practically or in economically viable ways, so the water returns back as unusable, esp. if they get water from a body of water and return it with added chemicals the process gets very damaging.
> The additives used by datacenters in some cases can't be filtered or processed out practically or in economically viable ways, so the water returns back as unusable, esp. if they get water from a body of water and return it with added chemicals the process gets very damaging.
Why would any water be returned? Datacenters (and literally every industrial process that requires cooling) use water for evaporative cooling. The water evaporates, then returns as rain elsewhere.
The only water being "returned" I can think of would be from the intermittent or continuous blowdown systems to maintain TDS levels; but blowdown is a small fraction of the water consumed and is typically returned back via the city sewer system.
Roughly 80% of the water evaporates into the atmosphere. The remaining 20% is flushed out of the system to remove mineral build-ups and sent to local treatment plants.
I have yet to see the standardization in that claim. Are there some reliable papers on this topic?
It's complicated, and depends on the hardness of the water entering the system from the water supply as well, since blowdown is ultimately to control the level of dissolved solids in the water so it does not contrate enough to start coming out of solution. Rarely if ever is true demineralized water used in open loop cooling systems; it's simply too expensive to produce to then "throw away" into evaporation.
The previously stated 80/20 isn't a bad general estimate, but it varies depending on the specific site. 10%-25% of makeup water being lost to blowdown is a reasonable range.
As for papers about it, I was only able to find a few in my brief search, but they don't land on a single number since it depends on the ratio of dissolved minerals/solids in the circulating cooling-tower water compared with the incoming makeup water, which varies due to acceptability for the process and due to the incoming water quality.
This short PDF on "Mechanical Systems Water Efficiency Management" from the EPA goes over the concepts pretty well though
- https://www.epa.gov/sites/default/files/2017-12/documents/ws...
This document from the US DOD (UFM 3-230-13, Industrial Water Treatment Operation and Maintenance (2026), Chapter 5) says that most cooling towers operate at 3-10 CoC, with 5-7 CoC identified as a cost-effective range. Where CoC is Cycles of concentration described in the previous PDF, and a higher CoC means less blowdown and better water efficiency, with the trade-off of a greater risk of scaling, corrosion, and biological/water-chemistry problems.
- https://www.wbdg.org/FFC/DOD/UFC/ufm_3_230_13_2026.pdf
Veolia also has a "Open Recirculating Cooling Systems" handbook that shows some good examples with real numbers in table 31-1.
- https://www.watertechnologies.com/handbook/chapter-31-open-r...
Additives like what?
We're talking about water that's released in general use, right? I would expect a closed cooling loop to have some biocides but have almost none of that leave the system.
And one datacenter had a weird bacteria that one time.
Yes, a single hamburger takes 620 gallons(!) of water to produce. Eat plants. Plenty of great plant protein alternatives available, plus they are healthier and don't depend on killing living beings.
Actually your premise is already wrong, although to be fair I'm not sure what bearing that has on the conclusion.
Here in France at least, companies can also be subject to FOIA requests for information pertaining to contracts they have with public entities. E.g. if a city publishes a tender to run tye local bike share program, then the company that gets the contract may have to directly answer (some) requests about that program.
But like I said, not sure if this applies to data centers which are, iiuc, mostly private ventures.
>And they also need to be built somewhere, and that place is probably best where electricity is abundant and/or cheap.
They don't need to be built at the scale they're being built at. That's the problem.
Not sure why you think your finger-in-the-wind wind understanding of future AI compute demand should trump the projections of companies who actually operate in this area?
But if you do end up being correct, then the data centers in question will just sit partially or fully idle, using no / a tiny fraction of the electricity and water they could use (which as others have said on this forum, water usage concerns are massively overblown in most data center builds anyway)
My biggest problem is that this is surveillance state infrastructure
That's a perfectly valid concern, but isn't specific to data centers or their construction. That's like being mad at a factory that produces pens because you don't like what people are writing with them.
Not sure why you think the interests of companies who operate in this area, should trump environmental concerns for the people who live next to where they plan to build the data centers? I for one prefer quality of life for actual people over speculative profits for tech billionaires.
I'm also not sure why you think it's a good thing to produce a bunch of electronics using conflict minerals and let them sit idle? Surely you must see the negative "externalities" of this?
> They don't need to be built at the scale they're being built at.
According to whom and by what metric? The companies investing vast amounts of capital into them seem to disagree.
Yes and I disagree with them. Ok?
And as we all know, investments of vast amounts of capital by companies can never be contrary to the well-being of humans... See Purdue, Monsanto, Union Carbide, etc.
These are resources acquired from the governments in most places. You have to be able to plan and expand capacity which requires public funds.
I live right next to a waste valorisation plant[1]. They have to exactly specify what amount of water they use and what amount of water they then send out to the waste water treatment facility. It also has to be clean to a very specific standard before it leaves which is also checked.
We have datacenters in Switzerland too, but very little outrage or debate because there are strict rules in place in most places. They can build all they want but it costs significantly more to meet environmental requirements.
The valorisation plant next to my home has water cooling tower/fans significantly oversized just to make sure the noise is keep under the required level.
[1] https://stefan.schueller.net/posts/kva-winterthur/
> where electricity is abundant and/or cheap
Definitely not the Netherlands then
> I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry.
Apples and oranges comparison. We eat the direct products of water usage in the agriculture industry. It keeps people alive. It’s fallacious to compare that with water usage that goes to serving ads.
> And they also need to be built somewhere, and that place is probably best where electricity is abundant and/or cheap.
This is a Dutch website and here in the Netherlands we have demand issues. In these same areas where issues exist, data centres are being planned and built. There are waiting lists now for normal businesses who want a new connection to the net, or connections with more capacity. These are lists waiting for capacity to become available: The demand on the current net is already too high.
This is something that certain AI advocates are refusing to acknowledge. The concerns are legitimate and can only be alleviated with open access to information by companies and the state. Then it can be clear that the project is doable or not, and let the people decide, as they are the ones that will have to live near it.
> water usage is heavily overblown compared to the agriculture industry
Are you suggesting we should stop using water to grow food for humans?
We should absolutely stop using water to grow alfalfa in the desert that we then ship to foreign countries so they can feed it to their cattle.
> don’t “freedom of information” laws mostly apply to government agencies?
As others (including TFA) have noted, this is false because there's an EU directive requiring "private companies" to disclose this.
In any case, the "it's a private company" defense is as tired now as it was when it was about blocking people from social media, or debanking etc. Just because the legal system hasn't yet caught up to regulating companies in a certain way, doesn't mean that it's not in everyone's best interest (except perhaps the owners of the companies in question) to have such regulation.
> I personally believe at the very least the water usage is heavily overblown compared to the agriculture industry.
The important distinction being that agriculture provides food, while data centers provide AI slop and advertising.
People keep comparing datacenter water use to agricultural water use, and it doesn't make sense to me. I can't tell if it's a bad-faith comparison or if people really know that little about how water is used in these industries.
The vast majority of agricultural water use is either tapped directly from groundwater, drawn directly from surface sources, or rain. The vast majority of datacenter water use is processed water from a municipal water utility. In other words, it is meaningless to compare the two, because only one is competing with residential use, and so only one is worth caring about.
That's not to detract from serious concerns about e.g. overutilization of the Colorado River, or the idiotic practice of irrigating deserts, but it is to say that comparing datacenters to agriculture makes very little sense.
Golf courses are a better comparison. Data centers use a fraction of the water that it takes to keep the grass green for a hobby enjoyed by a handful of people.
Bringing up agriculture is definitely a bad faith attack.
People need to eat, it's non negotiable. We do not need data centres.
People need to eat, but they don't need to eat meat.
It's true we're not obligate carnivores, but we do need a lot of the nutrients that meat also provides and is an easy source of
Providing a fully vegetarian diet for everyone is kind of non-starter.
On a biological level, there'd be no problem providing a vegetarian diet for everyone. A healthful vegetarian diet is not hard to achieve, at least no harder than a healthful non-vegetarian diet.
From an infrastructure point of view, providing everyone vegetarian food sources overnight would not be possible, but that's not a realistic scenario. With a gradual transition, the transition would reduce overall resource use on basically all fronts, so there's no reason to expect extraordinary problems.
Of course, this is all theoretical unless people care enough to change their lifestyles.
Of course the reason that people don't care enough to change their lifestyles is because
> A healthful vegetarian diet is not hard to achieve, at least no harder than a healthful non-vegetarian diet.
Might be true but isn't sufficient
It's still way easier to have an omnivore diet than a strictly vegetarian one. Partly that's due to how society is currently structured, but people still ate meat long before society was structured this way. Because it was easier.
It doesn't make sense to me either, but in a different way...
We need the food, we'll die without food, telling people to eat less of whatever food, so a few "tech bros" can build a new datacenter is outright evil.
edit: considering this is HN and the downvotes, people here prefer datacenters over food
People here generally don't believe that the water usage in DCs is enough to affect agriculture or that the areas so short on water are the ones we should be growing food in anyways.
The downvotes are probably about reasoning about your your conclusions instead of the things that led you to them. Though it'd be nice if that had been discussion instead of silent downvotes.
If the numbers would make them look good they would publish their numbers.
> Being angry with datacenter providers is not the solution.
Belittling criticisms that way is childish and doesn't address a single one. Literally nobody on the whole planet, at least in public discourse, is using "I am angry" as an argument in any of this. Yet it's the first thing that is always brought up. It's just a cheap, thought-terminating stereotype.
You are totally right to point this out, and I apologize for the whiplash. Let's clarify exactly how overblown data centre water usages is compared to the agriculture industry.
In data centres all that water is so that some wimpy cunt gets to hit return after spending no more than 5 minutes in an half baked prompt so he can insert the answer into his pitch deck, or PR review — and now with pacing the frontier models, he actually doesn't need to insert it, it's inserted automatically wherever it might fit. When it comes to agriculture, that is to produce things like food. This requires obviously much more water, which means that's what we really should be cutting down. Agents don't need food so the solution is just to replace humans with agents.
If you have any other stubborn question resisting clarification just hit me up, and I'll explain it to you immediately and clearly.
Agriculture is incredibly wasteful because it's allowed to be. Water intensive crops are grown in dry climates with restrictions on everyone else.
And so what, it's probably because it makes sense economically no? It seems tokens are wasteful too, yet here we are.
Meat too is wasteful, but yet again, here we are. All these are direct result of choices made by the majority...
> it's probably because it makes sense economically no?
Only because the government forces the cost of the water to be $0 for those who got in early, a policy that's causing extreme issues ecologically and causing the collapse of tons of environments and screwing over many millions downstream.
If we actually allowed the real price today it wouldn't be anywhere near economically sane to do such a thing.
Hmm. I live near some water intensive cultivations (olives, almonds, not USA) in a relatively dry area, but there's dam water at very cheap prices and obviously, for these products it's required, because otherwise you wouldn't have olive oils being sold at current market prices (already quite higher than just some 6 years ago), they would be astronomically higher. But this is to serve global markets. People in the region have done olive oil since forever without dam water, just not at this industrial scale.
But compared to meat and the whole chain of production required for that it certainly pales many moons below. Well, everyone loves their bacon.
> but there's dam water at very cheap prices and obviously, for these products it's required
So yeah, an example of the government coming in and ensuring cheap water for this crop production at the expense of those downstream. You agree, if the government didn't artificially and massively change the hydrology of the area, this industrial level wouldn't really be economically possible. This can make sense and not be terribly destructive depending on the region.
I and the poster above were probably talking about the American Southwest. In this case, its even more extreme. Old land owners got the rights to flood their fields from the river water and dig massive canals to turn extremely arid deserts green. They get the right to have that water for free, forever, because the government gave them the land at one point in time in the past and allowed for that taking based on deeply flawed hydrological models. They use that for agricultural products sold worldwide, including growing alfalfa for Saudi Arabia in the deserts of Arizona[0]. The government won't go in and actually start putting restrictions or calling back their grants to that water allocations, and its resulting in environmental catastrophe for many parts of the region.
[0] https://www.pbs.org/newshour/politics/in-drought-stricken-ar...
That does seem a bit extreme - in dam water you still pay, just much less because the water is also not treated in the same way as regular consumption water, storage is part of a second effect of electricity-gen damns many times, it enables agriculture and other things for communities, etc.
They refuse it because if they would say "1 million liters a year!" the knee-jerk reaction will be completely predictable even though that's just tiny compared to a small swimming pool or industrial or agricultural site.
The debate is completely based on emotion and anger right now, so sharing information will only give the emotional/angry folks more mud to throw with. The majority of people has no clue about how to put things like energy use and water use in perspective, and they're also not interested in doing that. Because they have decided to green-wash their AI-fears as environmental fears... which is a big fat joke because datacenters aren't significant polluters or consumers compared to almost all other industries.
Look, I get it, many people are afraid of the societal impacts of AI, or distrust it in another way. But let's be honest and focus on _that_ instead of naively acting as if our anti-datacenter sentiment comes from environmental concerns (the last few decades have shown we don't GAF about the environment).
>Look, I get it, many people are afraid of the societal impacts of AI, or distrust it in another way.
In the Netherlands it is not just about AI datacenters, even before the rise of AI pushback against Meta's planned datacenter made them cancel it. There has been a lot of criticism of the municipality that allowed Microsoft to expand its datacenters. The argument is that this is a small country that already has to make tradeoffs to get to its desired energy mix (wind turbines near cities on land) and then these datacenters are built that consume the power that is generated by sacrificing the little remaining open space. In the last few years there is actually a electricity shortage in some regions so "we can hook up this 200 MW datacenter but not your house" is a bad look.
> They refuse it because ...
It is completely irrelevant why they choose to refuse to provide this information.
From what I understand, disclosing this data is not optional. It is a law. And if they have willfully chosen to operate in a country where such a law applies, they have to obey that law.
> From what I understand, disclosing this data is not optional. It is a law.
That is not corrective. There is an EU directive, but it would have to be adopted as local law to become actual law.
The datacenters are following the laws in their jurisdiction.
If they are taking reasonable steps to mitigate consumption and pollution then they should be able to clearly explain that to the public. The fact they won't suggests they aren't.
Just because people haven't cared about the environment before doesn't mean they shouldn't start now. And just because there are worse existing polluters doesn't mean we shouldn't resist new ones.
A new golf course used to be seen as a great thing to have in your back yard and now we know their pesticide use increases the risk of Parkinson's. People are waking up and caring more. We need to have a sustainable society, not a profit obsessed one.
HN has never been friendly to comments like this, have you ever tried explaining something clearly to an emotionally triggered crowd? There are many examples, masks, vaccines, gun laws...
Why is the crowd emotionally triggered? I posit it's because of all the secrecy and bad faith. Data center builders and their supporters in government brought this opposition on themselves.
I posit it's because people say it's killing and stealing art and jobs and destroying farmland (it's not). In my neck of the woods, people show up to the actual meetings and loudly protest and have zero intention of listening.
It is doing those things too, but first and foremost on most people's minds is the issue of power bills, water availability and cleanliness, other local pollution of varying kinds, local taxes, and eminent domain for private profit.
Well, water usage is usually metered in cubic meters. A thousand m^3 per year sounds much less than a million litres.
Perhaps for people in the US, but this would be elementary school arithmetic in Europe.
I’m European and I’m extremely sure that when 90% of people hear "x m^3" they don’t think of it as "1000*x litres". And the majority of people likely can’t recall on the spot that 1 l = 1 dm^3 (a unit nobody uses), even if they remember that you multiply by 1000 to go from m^3 to dm^3. In any case most people don’t do conversions like that unless someone specifically asks them to.
Volumes are in general very unintuitive to people due to the cubic scaling. It’s further confused by the fact that the widely used decilitre, centilitre, and millilitre do scale linearly, unlike cubic metres. So 1 l = 1 dm^3 but 1 dl = 100 cm^3…
I don't know where do you live but even my small nephews and nieces understand that a 1 meter cube of water is A LOT OF WATER, because mentally, we have lots of 2 liters bottles to compare it to
Most of it comes from concerns about water and electricity rates sky rocketing and also simply not being available when humans need it. At least at the local level.
I'm all for data centres.
But they should be made to report this. AND their use of water should be regulated - they don't need to use any water at all, it's just cheaper for them - and they should be contributing enough to mitigate negative effects on energy prices.
The problem isn't data centres. The problem is not pricing in and regulating externalities, whether caused by data centres or other things.
It would also quickly show us who actually cares about the environmental effects and who are using it as an excuse.
A swimming pool requires maybe 60k liters a year...
One cubic metre of water is 1000 liters. A typical 25m swimming pool ("short course" / half of olympic size) is 20-25m wide. That's at least 500 m^2 surface area. At an average depth of only 1m (very conservative even for a recreational pool), that is 500,000 liters just to cycle the water once a year. Estimates I can find suggests 3-7 times as much as that on a yearly basis for a typical recreational pool.
So unless you're talking about a tiny private residential pool you're at least an order of magnitude off.
Well, OP said "small swimming pool". I guess it's ambiguous.
As someone who has been maintaining a pool for several years, you really hope you don't have to cycle the full amount of water once a year. You're really doing something wrong or incredibly unlucky if that's the case. That's at least true for the climate I live in. If I lived in a different climate, I'd probably forego having a pool.
It was incredibly painful when I had to drain the pool for deep maintenance a few years ago, and I'm not looking forward to having to do it again in a little while.
Also, most residential pool owners don't have half olympic pools in their yard. They're usually much smaller than that.
> Also, most residential pool owners don't have half olympic pools in their yard. They're usually much smaller than that.
Hence my caveat at the end. When talking about swimming pools, tiny residential pools is certainly not what springs to mind for me.
Most pools near me are in fitness centres etc. and are 25m pools.
And I certainly wouldn't go swimming in one of those if the water wasn't cycled regularly. Thankfully the data suggest they are.
If we're talking tiny little residential pools, then using a swimming pool worth of water isn't very shocking. Unless I'm way off, a 10-15m shower every day would use a similar amount of water in that case.
> I certainly wouldn't go swimming in one of the water wasn't cycled regularly
Once again speaking as someone who has maintained pools for a while, if the water was properly monitored and maintained I probably wouldn't mind. I'd absolutely take a pool that's actively properly maintained but hasn't had a full replacement in a few years over one that's poorly maintained but was fully drained and refilled a couple of months ago.
In a public pool?
From experience while my son was swimming competitively, they had to drain most of the pools he swam in completely more than once a year because of various contamination, in addition to the ongoing cycling...
I can't imagine the state they'd be in after a year.
Once again, if its monitored and maintained properly, sure why not? Partial cycling when things that can't be filtered get too high (like TDS or CYA for an outdoor pool) but otherwise make sure sanitization is happening properly, pH and other metrics stay balanced, and good filtering is happening, I don't see a problem.
> more than once a year because of various contamination
Sounds like that falls into the "unlucky" category.
If its that you don't like thinking about tons of other bodies have touched that water you're swimming in, remember that dinosaurs pooped in much of the water you're drinking today.
> Partial cycling when things that can't be filtered get too high (like TDS or CYA for an outdoor pool)
Partial cycling in a high traffic environment seems like it is clearly ending up being a lot of water when every number I can find suggests it is generally expected to add up to several times the total pool volume per year.
> Sounds like that falls into the "unlucky" category.
To me it feels like it falls in the "expected" category somewhere that is dealing with hundreds of swimmers including small children every day.
Keeping in mind this was across dozens of pools - he swam competitively, not a one-off or rare occurrence in a pool or two.
> Partial filtering
It seems like we're mixing up terms here and miscommunicating. I thought we were talking about actually replacing the water here (as in, down the drain, fresh water from the tap), not just filtering as in recycling water through filtration systems. The total volume of the water should go through filters every few hours. I'm 100% not for the idea of only "partial filtering".
> To me it feels like it falls in the "expected" category somewhere that is dealing with hundreds of swimmers including small children every day.
Nah, that is definitely not normal to have to completely replace the water multiple times in a year because of water contamination events happening. The water can usually be sanitized. It might require a day's worth of hyperchloronation, but not often an entire drain and fill. That will happen often, probably several times a year. Usually only if your CYA is too high will it require you to actually drain a significant volume of water, which I agreed was one of the reasons to at least do a partial drain and really only applies to outdoor pools. Doing a hyperchloronation will likely increase your TDS, which won't go down until you drain those contaminates. But usually you can get those numbers back in an acceptable range without having to drain and fill the whole pool.
If the pool was properly maintained, what state do you expect the water to be in after a year without a total drain and fill? Do you have experience actually maintaining a pool?
> It seems like we're mixing up terms here and miscommunicating. I thought we were talking about actually replacing the water here (as in, down the drain, fresh water from the tap)
We are. I changed it while you were writing.
Everything I'm seeing suggests typical loss to backwash of the filters for a short course pool is in the 5k - 10k liters of water a day range, possibly less if you invest in expensive additional filters. 5k a day is ~1.8 million liters a year.
To get below a full replacement every year for a small short-course pool, you need to lose less than 1370 liters a day, or ~1.4 m^3
And besides that: news sites love comparing water usage to things like bathtubs, the average household or swimming pools. The headline would likely be "datacenter uses 100 swimming pools of water per year!" and not "datacenter uses 6 million liters of water per year!"
Now GP comes off as knee jerk.
The people who don’t care about environmental impact and resource use are the same people pushing for more and faster datacenter deployment. The Republican Party is the one getting rid of environmental protection rules, it’s the party most financed by fossil fuel companies, and the party that was/is against solar and windmill development. The party that gutted all the regulatory agencies and is colluding with the AI industry.
There is a very obvious pattern here
Maybe I'm being naive, but the policies needed here feel quite simple.
First, since data centers likely need some kind of treated (i.e. demineralized) water, and they want to use local municipal supplies, then they can do so only if they pay full market rate for that water. I would push for them to pay 1% to 5% above market rate to help offset other costs, but market rate at minimum. At no point should homeowners and/or residential customers subsidize this usage.
Second, commercial users (including data centers) should only get water produced by treatment plants in excess of residential needs. Plants can only treat so much, pipes can only carry so much, and priority goes to residential first.
Third, ground water is off limits for any commercial/industrial evaporative cooling uses. If data centers want to cool that way, then they need a different source of water.
Did I miss anything?
If they pay market rate, the clearing price will go up making things more expensive for everyone. If they go off-grid they make the machines needed to produce water or electricity or whatever more expensive. If they borrow money, they make government bond yields go up and cause governments to start thinking about cutting pensions and stuff, not to mention increased mortgage rates for whoever wants to buy a home or is on a variable rate mortgage. There's no world in which this AI stuff doesn't increase costs for the rest of us.
Well, even if you're right, try proposing those bills and then reckon with the deafening wails of the "free-market-big-government-taxation-is-theft" crowd. And the politicians that are in the pockets of BigCo that will say that kind of stuff.
Policies are simple a lot of times. Politics against the interest of big money isn't.
They need to pay for the infrastructure upgrades to water and electric when it inevitably exceeds capacity
The main thing is evaporative cooling is incredibly wasteful. The water should be returned to the system, if it's going to be used, and not sent into the atmosphere. This is especially important in regions with limited access to water, which, ironically, is where many data centers seem to be located.
I think it would be helpful for people in this thread to understand the differences between EU Regulations and Directives :
• EU Regulation: A binding legal act that applies immediately and identically in all EU countries as soon as it takes effect, without needing national laws.
• EU Directive: A legislative act that sets a mandatory goal that all EU countries must achieve, but allows each nation to pass its own laws to implement it.
It takes time for directives to take effect, unless a member state's national law already implements them.
EU is in such a weird position. They are realizing that software, robotics, and generally automation through AI is the industrial future, but they seem to be paralyzed to actually start the build-up that will feed that frontier industry and economy.
Assuming you are right, the "build-up" of more physical datacenters isn't really the difficult part, training frontier models is. The open-weight Chinese models are several months behind frontier models at this stage and as long as you're not interested in burning cash to stay in the lead you can just wait it out. Many signs (upcoming ridiculous IPO figures and benchmarks) point to the current LLM paradigm reaching a plateau.
I think that currently the EU is focusing on strengthening itself from a military perspective due to US policy shifts, so the EU just doesn't have the funds to divert to AI without impacting day to day life of people, thus prioritizing safety and perhaps energy production, as electricity is considerably more expensive here than in the US or China. If you ask the people who vote, they'll put safety, grocery and gas prices way above EU's LLM capabilities.
> so the EU just doesn't have the funds to divert to AI without impacting day to day life of people
The US Government is not a major funder of frontier labs (who are almost exclusively VC and private money funded)...
> as electricity is considerably more expensive here than in the US or China
Yes, there are government policy reasons behind why this is the case. Some would argue that these policy decisions are a net negative for the EU.
> If you ask the people who vote, they'll put safety, grocery and gas prices way above EU's LLM capabilities.
Fair point, these are good things to focus on. However, EU citizens are going to love it when there are no frontier-tech opportunities left in their countries, with everything in this area having to be imported from the US and China (note that just because Chinese firms make open weight models available now, this could easily change very quickly in the future for a variety of reasons).
> just because Chinese firms make open weight models available now, this could easily change very quickly in the future for a variety of reasons
China is not providing "open weight" models just because they think it's good for humanity. It is a competitive strategy to stay relevant at the forefront of the industry. The reality is that China is severely compute-constrained mostly due to Nvidia chip and ASML embargo. Nearly all available compute capacity has to go toward training, leaving very little for inference. It is why they release their models openly allowing third parties in EU and USA to host inference for them.
This will change when China reaches a critical scale of domestic AI chip production, which is inevitable.
Yer this is one such 'reason' for China stopping the release of open weight models in the future, there are many others...
> The US Government is not a major funder of frontier labs (who are almost exclusively VC and private money funded)...
There isn't a lot of VC funding available in Europe compared to the US, so without government intervention catching up will be very tricky.
The problem is that EU has no money. France is virtually broke. Germany and eastern/central Europe is hemorrhaging its most valuable industries to cheaper Chinese competition. The whole bloc is carving out $100 billion military aid packages for Ukraine basically every year with no end in sight. And now EU is getting into a trade war with China despite having little leverage. Energy costs were bad even before the Ukraine war and Iran, and now they're at the boiling point.
It's why the USA and Chinese vultures are swarming. It's a bad situation.
France isn't "broke", its debt-to-GDP ratio is actually lower than the US right now and it's not in a situation where it can't pay its loans.
There is currently no real trade war EU-China; Chinese EVs are streaming into the EU while the US is blocking them completely for one example.
While Germany and more specifically Volkswagen is having some issues, EU industrial production actually had growth last year.
The $100 billion aid packages to Ukraine are also a bit overblown based on public figures https://www.kielinstitut.de/media/news/ukraine-support-track....
I feel like your comment pulls on some real threads, then blows things out of proportion.
Yes, I'm sure that EU doing its best to ignore building any technology frontier capabilities in its population will lead to a good solution for this lack of VC funding... ;-)
No shortage of visionaries in EU leadership these days :-)
I just learned yesterday that the default model on Mistral monthly plan is GLM 5.3 (an open weight Chinese model)
It's kind of sad that the only European LLM lab of significance is not even setting their model as a default one any longer.
Realizing something is not the same as acting on it and Europe is falling further and further behind.
I am not sure why people believe that data centers are the future. I did not down vote you, btw.
"Data center" is really just a proxy for industrial compute capacity and it's a resource like any other. It's a factory.
If the EU does not build their own factories, then they will be reliant on factories built elsewhere.
The EU already has several hyperscaler data centers and many more colo-style data centers. Plenty of compute power exists in EU; this new generation is about purely inference capacity.
In the US, the political backlash against data centers is also about the inability of government to be transparent about these buildouts (eg. City councils signing NDAs and building commissions quietly approving development plans without community input), the inability for utilities/regulators to actually hold data center developers to promises made (usually regarding resource/ utility usage), more complaints about affordability (cost of utility bills rising), additional hate against the unlikable tech CEOs (who aren't charismatic and whom have a long list of previous violations), fear of what AI will bring (the promised white collar revolution, promises of "infinite AI immigrants" in an age where immigration has been demonized), etc.
Only a few of these are about the "factory" itself. Lots of them are about the people trying to retain some control of their communities. Some of them are about the speed of the buildout (eg. compromises in water use / plumbing and noise / pollution created by the ad hoc on site electrical generation).
The data center CEOs are starting to admit that their NDA-blitzscale strategy may have backfired because communities are pissed and the NDAs prevented anyone from making a positive case for the buildouts while the rage was boiling over.
Love this framing, most industrial warehouses easily use 500kw in power (the EU level for self reporting water consumption), but nobody is drumming up a water war on them.
For whatever reason people have clung onto this idea that the "Gotcha" to slow down AI is to worry about water consumption, and it's an idiotic gotcha.
I don't know which data center it is, some remote location of Canada, that has 100 million litres of fuel trucked up each month for it's generators. I don't think that has a long term plan behind it because 100m liters of anything is a lot, and with the price of fuel now it's definitely not a small line item.
Btw I heard this second hand a friend in the mining industry.
Wise.
Continuing a conversation is often just a slow-drip Streisand effect.
FTA:
> According to the latest figures from Statistics Netherlands (CBS), data centers used 5.1 billion kilowatt-hours of electricity in the Netherlands in 2024. That is 4.6 percent of the country’s total electricity consumption, a near doubling in five years. Grid operator TenneT projects that data centers will consume 10 to 15 percent of the Netherlands’ electricity by 2030.
Yeah, so 2030 sounds like a long way off. It's 3 years away. We are not going to be meeting any climate targets.
Smart move. Why would they? Maybe there's laws that say they have to, but if no one is going to enforce said laws, then they maintain a competitive advantage by "refusing" to say.
I live in a country with unlimited water supply.
Can someone explain the data center water usage? Data center don't drink water. It cools down the machine.
What's preventing them getting cool water from a lake and draining to it again to cool down? Where is water going after usage since there is literally zero water consumption?
>What's preventing them getting cool water from a lake and draining to it again to cool down?
Some do, but most don't since the requirements for using cooling water are relatively strict due to the pipes, equipment, and heat exchangers the water flows through. The water used needs to be free of any biological contaminants (such as bacteria and algae), maintain proper pH levels, and are often then treated with other chemicals such as biocides.
Lake water certainly can be treated to the level needed and used, it's just often cheaper to buy domestic water from the local utility than to construct and operate a small scale water treatment system along side the industrial process.
> Where is water going after usage since there is literally zero water consumption?
Most of the water is "consumed" via open loop evaporative cooling systems, so the water is never really "used" as in destroyed or anything. The majority of it is evaporated and returns to the land as rain and the rest is returned via sanitary drains, to city sewer systems.
Evaporation losses will move the water to a different locality. DC's not only cool servers, but control humidity by adding or removing water to the air.
Warm water fed back into the ecosystem will (relatively) quickly cause unwanted changes to the lake's biosphere.
Unless it's properly filtered, returned water will be contaminated by material in the cooling system.
It puts local demand on the water system competing with human need.
Didn't you know? Once you heat up water that's it, you can't use that water ever again. It's even worse if you make it evaporate, not just useless but GONE.
Now try again with local and short term impacts on the whole ecosystem instead of saying "Ackchyually water never leave the atmosphere so it's unlimited"
Some data centers use evaporative cooling instead of a closed loop.
If you don’t cool the water down before putting it back, you’re just creating heat pollution.
It cools down the machines by evaporating, and is then released into the atmosphere. That also means the water needs to be clean, pretty much drinking water, because otherwise the evaporation would deposit minerals and stuff.
Turning water into steam uses much more energy than heating water, which means it's a much more effective means of cooling.
What you describe is the type of cooling used in certain power plants. That heats the lake though, which has its own environmental effects https://en.wikipedia.org/wiki/Thermal_pollution
Some smaller data centers have a cool solution where they deposit extra heat into the municipal hot water grid, which is pretty neat https://www.techradar.com/pro/finally-some-good-usage-this-d...
Imagine fretting about water and electricity usage in europe, where we'll never ever have a datacenter build-up because electricity prices are too high anyway.
Yet, at the bottom of the article:
> In the Netherlands, it is estimated that they account for about 0.1 percent of total consumption.
If that's true, what's the problem?
This figure is about the water consumption only. The estimate about the electricity consumption, in that same article, is much higher than that.
The main problem is, in my opinion, that those are just estimates. If we are supposed to accommodate the current and potential future needs, we need real data. Not estimates.
Duh, check their utility bills?
Clause Opus 5.5 reports that based on the 5.1 billion kWh used it's about 9.2 billion liters in Europe (cited from the Water Usage Effectiveness WUE industry-average). About 1.8 liters per kWh.
In the US, the rhetorical trick is to include the 3.4T* gallons of water utilized for power generation along with cooling (160B) to make this number super scary when water was utilized for power generation because it's a good cheap relatively clean power source so the datacenters were built near hydroelectric plants.
We can't do much about the 160B, but we could do a lot to build more power generation and storage. As just one example, in west Texas, natural gas is flared into the sky en masse from depleted oil wells because methane is a more powerful greenhouse gas than CO2. Some sites could be tapped to hit hundreds of megawatts to a gigawatt of power generation. And any power you don't need can be exported to the Texas power grid so it can make money the day you turn it on. And if you're willing to spend a little more, you can sequester the CO2 and make the power generation mostly green.
Cue crickets from Silicon Valley on this one because it's neither nuclear nor renewable, but the crypto people love it so I would guess they'll eventually come around. That they haven't already is one facet of my disenchantment with tech culture. Their absurd belief in Star Trek Voyager-like scenarios playing out with AI in the next decade the other major one.
*Which is slightly more than the 3.3T gallons of water Americans used to water their lawns. So, ya know, where's the outrage?
North Dakota made it expensive to flare, so companies had to capture the gas or face production limits.
You already know for sure what the reason is, don't you?
the article spells out other data points, so why not query AI to estimate what they fail to disclose?
I wouldn't either. Nor would I if I was a manufacturer of widgets. Competitive secret. You can draw a correlation from resource use to output.
How much electricity does a Dr. Pepper bottling plant use? Oh they don't publish that either? Curious.
what about how much BS?
Less than they do if there were people living on that same land, and way less than they use in China. Get over it, data centers and progress is the future.