Since the article doesn't make this clear, let me explain.
The cheapest (and worst) option is to take in water, use it for cooling and then dixcharge it. Why is this bad? Because DCs don't want to corrode their pipes with untreated water so they add coolant and additives to it, which pollute the water. This is bad. But nobody really does this because you don't want to keep adding additives to water you're constantly discharging.
The next step up are varying degrees of what's called "closed loop" cooling. That is, the DC has treated water in a closed loop that isn't discharged. There's a heat exchange system with external water. This btw is the system that's used in nuclear reactors although nuclear reactors will be far more stringent than DCs are. Best practice for this is one of Google's DCs in Scandanavia that uses ocean water for heat exchange. There are limits to this but there's only so much Arctic Ocean water a DC can meaningfully heat. It is potentially disruptive though and that needs to be considered.
Even so pipes will need to be cleaned. There is debris that builds up and in cases like this you can still get bacterial outbreaks. This is another reason to use additives like chlorine. But again, you don't want to discharge chlorine into bodies of water.
I'm reminded of water management in the Yukon. The Yukon for over a century have been gold fields. If you look at the tech required to extract a tiny amount of gold from a large amount of earth, it's kind of fascinating but it boils down to using a lot of water and having the denser gold sink and get trapped.
So gold miners take in water from rivers, wash rocks with it and then have historically just discharged it back into the rivers. This tended to be heavy in silt that would go into waterways and could create problems. The water was also dirty. So the Yukon authorities have gotten increasingly stricter with water management. Now water has to go through a series of settling ponds so the discharged water is clean/clear.
I kinda think we need similar levels of strict water management for DCs. No discharged coolants and clean water. Figure out how to get that. If that makes your DC more expensive then that's a "you" problem.
> But again, you don't want to discharge chlorine into bodies of water.
Maybe but there are citywide chlorine flushes of their water supply. some worse than others
Erin Brockovich tracks this
As a former microbiologist, this news is important but not too concerning. It’s good that we found detected this and can respond accordingly. It doesn’t mean indicate an immediate, critical issue.
reminds me the meme that every 'thinking...' you see in ChatGPT is just another gallon of poisoned groundwater
Is it really that expensive to not do stuff like this?
I guess 'turning the entirety of the American public against data centers' is not something they factor into the cost analysis.
This is one of the data centres that went to the extra expense of building a closed loop cooling system that would, supposedly, not waste water on a continuous basis. Apparently, even these are not so clean to set up. Governments are going to need to start paying more attention to the commissioning process apparently.
--------
"Meta said that it's supporting its general contractor, Fortis, which stopped discharging and began hauling wastewater offsite"
--------
Governments should also watch where this wastewater is being hauled to, and likely just dumped.
This closed loop system needs to be filled and flushed, which is the operation discussed in the article. The bacterium was described as a possible airborne hazard if used for irrigation, it's not a regulated substance though so dump and dilute might be perfectly safe.
Might be. Might not be. To put it another way, would you want this waste water diluted and dumped in a lake you swim in or a river you fish in without proper study and regulation?
Apparently, other people's quality of life and health are of little to no concern in comparison to company profits, after they've paid off the correct politicians.
After all, the executives of the company (the important people) know not to swim in that lake (their swimming pools are clean) nor drink the water (as they can drink Perrier bottled water).
> To put it another way, would you want this waste water diluted and dumped in a lake you swim in or a river you fish in without proper study and regulation?
Are you going to apply the same standard to every house with a swimming pool, every municipal storm drain that collects rain water having passed over the ground in nature where there are untold strains of bacteria and every other system handling thousands of gallons of unfiltered water?
Otherwise it's an isolated demand for rigor.
You can tell that something has an ulterior motive when the rule or its enforcement special cases the doing of a generic thing only when it's being done by the people being targeted.
Not sure why you got downloaded, but you’re right. Polluting public (and even private!) spaces is philosophically speaking a property rights violation, which is a core libertarian/free market value.
So why the fuck are we in the habit of giving companies the benefit of the doubt on this? Companies always follow the financial incentive. There is rarely a financial incentive to not pollute and always a financial incentive to spend less money on costly processes that slow things down, so they’ll pollute every chance they can. It’s just a side effect of how capitalism works.
So yes, if you actually care about your property (including public property in your town!), you absolutely need to push for more oversight. Companies have absolutely zero incentive to do it themselves, as evidenced in this scenario where the town “caught them in the act” so to speak.
And I’m not saying this company was doing anything deliberately malicious, but it takes the town being on top of their wastewater management processes and doing a solid root cause investigation to even find out this was happening. That doesn’t happen unless people care. A company has no incentive to do it themselves.
> why the fuck are we in the habit of giving companies the benefit of the doubt on this
Because companies pay for lobbyist, give donations, and other perks to politicians. Simple as that ... And companies a a-political, they simply give money to both sides, so who is in power has less effect.
While you the normal voter, only matter a few times, and with increased irrelevant as how powerful your vote is.
The fact that companies can now vote in local elections (in some states, you can guess the states colors), tells you how deep the corruption has gone.
> began hauling wastewater offsite
They towed it outside the environment.
Testing for an extremely
rare bacterium is not the bare minimum. In fact even the water treatment plants rarely do it. They admit in the article we don’t even know whether or not the bacteria originated from the water supplied by the city that entered the pipes in the first place.
DCs should be responsible for their output but this seems to be a super edge case.
Not do what? Not discharge the water with bacterium? But the data center claims that their independent testing shows that they didn’t even discharge the bacterium. It seems that neither the city nor Meta knew where the bacterium came from.
I kinda think the reaction and reporting on this is already being colored by the objection to the datacenters, as opposed to the other way around. And it's not obvious exactly where the problem came from in the first place: it doesn't look like this is some super-obvious, common and easy to prevent issue that they just decided to cheap out on, nor is it obviously a datacenter-specific issue.
Is this a common problem? I would assume not, since it lead to a suspense, and that is probably very expensive and something you aim to avoid.
Yeah, unfortunately.
Effluent and wastewater companies have been getting greedy. If one suddenly 100x's your cost, you're fucked until you build onsite treatment or find a way to ship it out.
By definition, externalizing a cost is less expensive than internalizing it; the only recourse for the rest of us is forcing them to properly internalize their costs.
it's incredibly hard to identify microbes like these, this is not a simple task, especially if you're not aware they exist or that they can occur in this way.. if they did, which i'm skeptical of
Wait until they get nuclear power plants for their data centers.
Profit over safety all day long
'turning the entirety of the American public against data centers' is a blatant public opinion campaign ran by people and entities who didn't give a damn about data centers up until 3 years ago.
Well, it turns out people started noticing local community impacts when new data center construction absolutely exploded 2-3 years ago.
Not just exploded, but went full on "move fast and break things". Nobody complained about the data centers that have no more impact on the local environment that a warehouse. But in all this rush, they're cutting corners everywhere and putting so much strain on the electric grid. Many also have massive onsite natural gas turbines, some big enough to power a small city, built by the contractor with the lowest bidder, half a mile from people's houses. You can hear and FEEL them from inside a house.
you have numbers? I asked a chatbot and the numbers I got were underwhelming.
I feel sorry for you. How about you do your own research?
Move fast and break things isn't a mantra that arose from first principles, it's literally all they know how to do
Move fast and break things is back.
Interesting I’ve been waiting for something like this on our rivers. Hopefully.
Oh, great. A group of men in black shirts and jeans are here to fix the issue
A bit meta - the names in this article made me chuckle:
Goat Systems, Cowboy State Daily, Cupriavidus gilardii, Frank Strong the Board's Manager and the Crow Creek and Dry Creek facilities.
This is gold for a comedy sketch :)
Often life doesn't just imitate a movie, but imitates a goofy movie.
This is why datacenters in central texas are desperate to build anywhere in the edwards aquifer... so they can get "free" water from natural springs (already stressed by draught) and dump the effluent into city wastewater systems.
> after tracing a rare bacterium in the city's reclaimed water to Goat Systems LLC, the entity Meta uses to build its Cheyenne campus
Hey where’s that person from yesterday who argued with me over the 1m vs 1cm hole in the boat?
Everyone saying stop talking about data center water use is missing the entire point as this article shows.
Data center water use is a fairly separate topic from what this article covers. Related of course but the conversation on USE centers around actual volume use, not contaminants.
It’s not a separate topic though, as this article shows. A closed loop system can still mess up the water supply, because (in this case) it wasn’t fully closed.
Discharge, is part of "water usage". Arguing otherwise is embarrassing.
According to the article this is a closed loop cooling system, once it’s up and running it doesn’t use any water. They run water through it during installation and that’s the discharge that they found bacteria in.
You can have a “closed loop system” but you need to shed heat somewhere and that is either by air cooling drawing tons of electricity, or evaporation that draws tons of water.
I’m pretty sure that “closed loop” in this context would mean that the water is recirculated and not evaporated.
From the links in the article, it looks like this DC runs giant heat exchangers to dump the heat to air.
> air cooling drawing tons of electricity
Reference? What's the efficiency of the systems like used here?
Air cooled chillers are generally used in a closed loop system. A coefficient of performance (COP) of 4 is reasonable to expect, which means 1 watt of electricity to remove 4 watts of heat.
Not sure if you're being purposely intellectually dishonest or just unaware, but the common thread on the water usage argument is literally about quantity of usage and rising prices for local areas. So as I said, somewhat related, but still two completely separate discussion topics.