Go back

How much water does AI consume?

8m 59s

How much water does AI consume?

This podcast episode investigates the water consumption of AI systems, addressing a controversial claim that AI could use 4.2-6.6 trillion liters of fresh water annually by 2027. The analysis reveals this figure is inaccurate. It originated from an academic paper that estimated water *withdrawal* (total water taken from a source), not *consumption* (water not returned, e.g., through evaporation). Furthermore, the paper's underlying electricity use estimate was flawed, as it only considered servers produced in a single year, not the cumulative total in operation. Researcher Alex de Vries provides a more grounded estimate, suggesting AI systems are currently consuming water at a rate of approximately 750 billion liters per year, which surpasses global bottled water consumption. About 90% of this occurs offsite at power plants, with only 10% used directly for cooling data centers. While this is a substantial amount with potential local impacts, especially in water-scarce regions, the full environmental effect is hard to quantify. Future growth is constrained by the ability of tech companies to secure sufficient power and infrastructure, making precise long-term predictions challenging. The episode underscores the importance of accurate metrics and the complex, localized nature of AI's resource footprint.

Transcription

1523 Words, 8651 Characters

English
This BBC Podcast is supported by ads outside the UK. Welcome to the interface, the show that decodes the tech that's rewiring your week and your world. On this week's episode, we'll look at the way that algorithms could change how much you're paying for your groceries, how even astronauts have issues with Microsoft Outlook, and whether the next trend in tech is less tech. Listen on BBC.com or wherever you get your podcasts. Hello and thanks for downloading the BOR or less podcast with a program that looks at the numbers and the news and in life and an AI water consumption. I'm Charlotte McDonald. When you sit in front of an AI chat bot and start typing away, the responses appear on your screen like magic. Information apparently spinging out of fresh air. The truth is of course, very different. When we send our queries off, these are dealt with by a vast network of data centers which were into action to come up with the answers. These data centers contain servers, which themselves contain processing chips, those run on electricity from the power grid, and as they operate, they generate lots of heat and need to be called to prevent overheating. Both electricity generation and cooling data centers use water. With AI expanding rapidly, some people have worried about how AI's water use might escalate in the future. One striking figure came in a book called Empire of AI, written by the US journalist Karen Howe. According to her, surging AI demand could consume 1.1 trillion to 1.7 trillion gallons of fresh water globally a year by 2027 or half the water annually consumed in the UK. That's between 4.2 and 6.6 trillion litres of fresh water, which sounds like a big number. But should we trust it? Nathan Gauer has been looking into this one. Hi Nathan. Hi Charlotte. Nathan, 46 trillion litres of fresh water sounds like a lot. It definitely is, but there's all sorts of problems here. So, house claim is about the amount of water that could be consumed. This has a specific meaning. It's water that's taken out of a system or source, but not returned. This happens through evaporation when it's used for cooling, whether that's on-site at data centers or offsite at the power plants where electricity is generated. They also evaporate water when they call the generators. Okay, so this claim seems to be saying that trillions of litres of water are going to get consumed. They're taken out of the water system and then not returned. Yes, but the claim is just not right. Howe got her figures from a paper published by academics from the University of California Riverside. Except those figures of 4 to 6 trillion litres aren't for water consumption. They're for something different called water withdrawal. That is the total amount of water that gets taken out of a water system or source. Now, some of this water will be returned, but some of it won't. It will be consumed. So, basically, the consumption figure is a subset of the wider withdrawal figure. Okay, so this also stuck the wrong label on the 4 to 6 trillion litres figure. It should be for a withdrawal, not consumption. So, what is the actual consumption figure? So, the paper says between 380 and 600 billion litres could be consumed in 2027. That's about 10% of the original 4 to 6 trillion figure. This mistake was pointed out by an American substacker, Andy Masley, and Karen Howe has since issued a correction. Right, so that's that'll now. It will be really great if it was. So, Karen Howe read those figures in the academic paper and misinterpreted them, mistaking the larger figure for withdrawal as the figure for consumption. But I've discovered that those figures in the paper were actually wrong in the first place. How's that? The paper says it takes an estimate for global AI electricity use in 2027, and extrapolates from that to a figure for global AI water use in the same year. But that electricity estimate comes from work by another researcher, Alex DeVries-Gau. I spoke to Alex, and it turns out that he wasn't estimating total global AI electricity use in 2027. Instead, he made an estimate for global AI electricity use by those AI servers that might be produced in 2027 alone. Right, so it's not counting all the AI servers and data centers built in previous years, most of which will presumably still be running. Exactly. So, they've basically underestimated electricity consumption, and that throws off the water estimates. Then, on top of all this, an author comes along and misinterpretes these flawed figures. This sounds like a mess. It's a bit of a mess. So, let's try and salvage something. That researcher, Alex DeVries-Gau, has also made estimates for global AI water consumption. To do that, you need to know how our electricity global AI systems might demand. Now, the tech giants don't publish these numbers, so Alex came up with a novel strategy. Just to get to the power demand, you actually need to take a deep dive into the supply chain of AI hardware. What I did was looking at how many AI chips could have been produced in the past years, how many AI server modules could have been made with that, and ultimately, how many AI servers have been made with that. When you get to that point, you still need some assumptions to figure out how much our server is going to be utilized. What I did there was I tried to look at the biggest buyers of AI server equipment, which is large tech companies like Microsoft, Google, etc. Kind of examined how do these companies data centers typically perform in terms of water intensity, and I used those numbers, ultimately, to translate my power demand estimate into an indirect water consumption estimate for AI server hardware. And then on top of that, you still need to include the water that's actually being consumed in the data center itself, the direct water consumption. Using this method, Alex estimated that AI systems at the end of 2025 were consuming water at a rate of 750 billion litres per year. Now, I know what you're going to ask Charlotte is 750 billion litres a big number, or his Alex. This is absolutely a big number. This is exceeding the level of global bottled water consumption, which is at about 446 billion litres. That in itself is a significant amount of consumption, but is it really a problem? Well, it could certainly be, it could cause a lot of problems if this consumption is concentrated in a single location where water scarcity is already a potential problem. But we just don't know at this time. So, that really makes it hard to make that translation, is this a problem or not? But at the same time, it also makes it impossible to say this is not a problem at all. It's a pretty huge number. It's going to have an effect on local freshwater supply, for sure. And we just don't really know where, and we just don't really know how much this is going to hurt in which locations. For Alex's estimate for water consumption, only about 10% of that is happening on-site at data centres, which typically use drinking water from local supplies. The other 90% is happening offsite at power stations, which is water from sources like rivers and lakes. Also, Alex has only made estimates for water consumption. He thinks that's the most important metric when thinking about overall water scarcity. But I've spoken to another researcher who thinks that water withdrawal matters equally if not more, and argues that even if water is eventually returned to a system or source, increased demand can still have important consequences. What about future predictions? Here's Alex again. One of the big bottlenecks that is starting to appear, can these tech companies even supply sufficient power to power their data centres? They're getting a lot of equipment. Where are they going to be able to actually use all that equipment? I can make statements based on, let's say, the current supply chain capacity for producing AI hardware, which I know this year is probably going to be at least similar to the previous year, 2025, which means that the cumulative power demand of AI systems is still going to be rising. This is still going to be adding on top of the production of the past three years. But again, I can't say anything about whether it's going to be possible to find a home for all this equipment. Well, thank you Nathan, and thanks to Alex DeVries Gow, as well as Professor Shao Le Ren, who also helped with this episode. That's all we have time for this week, but if you have any more questions or comments, please email us on more or less at bbc.co.uk. We'll be back next week, and until then, goodbye. Listen on bbc.com or wherever you get your podcasts.

Podcast Summary

Key Points:

  1. AI systems consume significant water for cooling data centers and electricity generation, raising concerns about environmental impact.
  2. A widely cited claim of AI consuming 4.2-6.6 trillion liters of water annually by 2027 is flawed due to misinterpretation of data (confusing water withdrawal with consumption) and reliance on underestimated electricity use projections.
  3. Researcher Alex de Vries estimates current AI water consumption at about 750 billion liters per year, exceeding global bottled water use, with most consumption occurring offsite at power plants.
  4. The actual impact is uncertain and location-dependent, concentrated in areas with water scarcity, but precise future predictions are difficult due to supply chain and infrastructure constraints.

Summary:

6 trillion liters of fresh water annually by 2027. The analysis reveals this figure is inaccurate. , through evaporation).

Furthermore, the paper's underlying electricity use estimate was flawed, as it only considered servers produced in a single year, not the cumulative total in operation. Researcher Alex de Vries provides a more grounded estimate, suggesting AI systems are currently consuming water at a rate of approximately 750 billion liters per year, which surpasses global bottled water consumption. About 90% of this occurs offsite at power plants, with only 10% used directly for cooling data centers.

While this is a substantial amount with potential local impacts, especially in water-scarce regions, the full environmental effect is hard to quantify. Future growth is constrained by the ability of tech companies to secure sufficient power and infrastructure, making precise long-term predictions challenging. The episode underscores the importance of accurate metrics and the complex, localized nature of AI's resource footprint.

FAQs

AI uses water primarily for cooling data centers and generating electricity at power plants. This includes both direct water consumption at data centers and indirect consumption at power generation facilities.

Karen Howe misinterpreted academic figures, confusing water withdrawal (total water taken from sources) with water consumption (water not returned to the system). She also used flawed electricity estimates that only considered new AI servers, not all existing ones.

Alex DeVries estimates that AI systems were consuming water at a rate of about 750 billion liters per year by the end of 2025. This exceeds global bottled water consumption, which is around 446 billion liters annually.

The impact is hard to assess because water consumption is often concentrated in specific locations, and data on local water scarcity and usage patterns is limited. It's unclear where and how much this consumption will affect freshwater supplies.

Water withdrawal refers to the total amount of water taken from a source, some of which may be returned. Water consumption specifically means water that is taken and not returned, often lost through evaporation during cooling or power generation.

Researchers like Alex DeVries use indirect methods, such as analyzing AI hardware supply chains, estimating server production and utilization, and applying water intensity metrics from major tech companies' data centers to translate power demand into water use.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.