Greenpeace report How Clean is Your Cloud? I saw mentioned in 3T magazine news is actually quite interesting reading. This year’s report provides a look at the energy choices some of the largest and fastest growing IT companies. The report analyzes the 14 IT companies and the electricity supply chain in more than 80 data center cases.
The report contains also lots of interesting background information on both IT and telecom energy consumption. I recommend checking it out. Here are some points picked from How Clean is Your Cloud? report:
Facebook, Amazon, Apple, Microsoft, Google, and Yahoo – these global brands and a host of other IT companies are rapidly and fundamentally transforming the way in which we work, communicate, watch movies or TV, listen to music, and share pictures through “the cloud.”
The growth and scale of investment in the cloud is truly mind-blowing, with estimates of a 50-fold increase in the amount of digital information by 2020 and nearly half a trillion in investment in the coming year, all to create and feed our desire for ubiquitous access to infinite information from our computers, phones and other mobile devices, instantly.
The engine that drives the cloud is the data center. Data centers are the factories of the 21st century information age, containing thousands of computers that store and manage our rapidly growing collection of data for consumption at a moment’s notice. Given the energy-intensive nature of maintaining the cloud, access to significant amounts of electricity is a key factor in decisions about where to build these data centers. Industry leaders estimate nearly $450bn US dollars is being spent annually on new data center space.
Since electricity plays a critical role in the cost structure of companies that use the cloud, there have been dramatic strides made in improving the energy efficiency design of the facilities and the thousands of computers that go inside. However, despite significant improvements in efficiency, the exponential growth in cloud computing far outstrips these energy savings.
How much energy is required to power the ever-expanding online world? What percentage of global greenhouse gas (GHG) emissions is attributable to the IT sector? Answers to these questions are very difficult to obtain with any degree of precision, partially due to the sector’s explosive growth, a wide range of devices and energy sources, and rapidly changing technology and business models. The estimates of the IT sector’s carbon footprint performed to date have varied widely in their methodology and scope. One of the most recognized estimates of the IT sector’s footprint was conducted as part of the 2008 SMART 2020 study, which established that the sector is responsible for 2% of global GHG emissions.
The combined electricity demand of the internet/cloud (data centers and telecommunications network) globally in 2007 was approximately 623bn kWh (if the cloud were a country, it would have the fifth largest electricity demand in the world). Based on current projections, the demand for electricity will more than triple to 1,973bn kWh (an amount greater than combined total demand of France, Germany, Canada and Brazil).
The report indicates that, due to the economic downturn and continued energy efficiency and performance improvements, global energy demand from data centers from 2005-2010 increased by 56%. Estimates of data center electricity demand come in at 31GW globally, with an increase of 19% in 2012 alone. At the same time global electricity consumption is otherwise essentially flat due to the global recession is still a staggering rate of growth.
Given the scale of predicted growth, the source of electricity must be factored into a meaningful definition of “green IT”. Energy efficiency alone will, at best, slow the growth of the sector’s footprint. The replacement of dirty sources of electricity with clean renewable sources is still the crucial missing link in the sector’s sustainability efforts according to the report.
The global telecoms sector is also growing rapidly. Rapid growth in use of smart phones and broadband mobile connections mean mobile data traffic in 2011 was eight times the size of the entire internet in 2000. It is estimated that global mobile data traffic grew 133% in 2011, with 597 petabytes of data sent by mobiles every month. In 2011, it is estimated that 6 billion people or 86.7% of the entire global population have mobile telephone subscriptions. By the end of 2012, the number of mobile connected devices is expected to exceed the global population. Electronic devices and the rapidly growing cloud that supports our demand for greater online access are clearly a significant force in driving global energy demand.
What about telecoms in the developing and newly industrialized countries? The report has some details from India (by the way it is expected that India will pass China to become the world’s largest mobile market in terms of subscriptions in 2012). Much of the growth in the Indian telecom sector is from India’s rural and semi-urban areas. By 2012, India is likely to have 200 million rural telecom connections at a penetration rate of 25%. Out of the existing 400,000 mobile towers, over 70% exist in rural and semi-urban areas where either grid-connected electricity is not available or the electricity supply is irregular. As a result, mobile towers and, increasingly, grid-connected towers in these areas rely on diesel generators to power their network operations. The consumption of diesel by the telecoms sector currently stands at a staggering 3bn liters annually, second only to the railways in India.
What is the case on other developing and newly industrialized countries? I don’t actually know.
NOTE: Please note that that many figures given on the report are just estimates based on quite little actual data, so they might be somewhat off the actual figures. Given the source of the report I would quess that if the figures are off, they are most probably off to direction so that the environmental effect looks bigger than it actually is.
648 Comments
Tomi Engdahl says:
Akshat Rathi / Bloomberg:
Google says it has stopped purchasing carbon offsets and is no longer maintaining operational carbon neutrality, but still aims to reach net-zero carbon by 2030
Google Is No Longer Claiming to Be Carbon Neutral
https://www.bloomberg.com/news/articles/2024-07-08/google-is-no-longer-claiming-to-be-carbon-neutral
The tech giant, which has seen its planet-warming emissions rise because of artificial intelligence, has stopped buying cheap offsets behind the neutrality claim. The company now aims to reach net-zero carbon by 2030.
Google has ended its mass purchase of cheap carbon offsets and thus stopped claiming that its operations are carbon neutral, according to the tech giant’s latest environmental report. The company now aims to reach net-zero carbon emissions by 2030.
The Alphabet Inc. unit has claimed that it’s been carbon neutral in its operations since 2007. The status was based on purchasing carbon offsets to match the volume of emissions that were generated from its buildings, data centers and business travel. But in its latest report, the company states: “Starting in 2023, we’re no longer maintaining operational carbon neutrality.”
Tomi Engdahl says:
https://etn.fi/index.php/13-news/16424-datakeskuksista-kasvaa-500-miljardin-jaettibisnes-saehkoensaannista-tulee-ongelma
Palvelinkeskusteollisuus on kovassa kasvussa tekoälyn käytön lisääntyessä. Stocklytics.comin mukaan palvelinkeskusten liikevaihto ylittää 500 miljardia dollaria vuoteen 2027 mennessä.
Analyytikko Edith Readsin mukaan kasvuun on monia syitä. Useampi monipilviratkaisujen käyttöönotto edistää datakeskusten laajentumista. Myös 5G:tä tukevat verkkopäivitykset myötävaikuttavat tähän kasvuun. Lisäksi hyperskaalaajien laajentamisaloitteilla on merkittävä rooli.
Tomi Engdahl says:
Amazon Vies for Nuclear-Powered Data Center The deal has become a flash point over energy fairness
https://spectrum.ieee.org/amazon-data-center-nuclear-power
When Amazon Web Services paid US $650 million in March for another data center to add to its armada, the tech giant thought it was buying a steady supply of nuclear energy to power it, too. The Susquehanna Steam Electric Station outside of Berick, Pennsylvania, which generates 2.5 gigawatts of nuclear energy, sits adjacent to the humming data center and had been directly powering it since the center opened in 2023.
After striking the deal, Amazon wanted to change the terms of its original agreement to buy 180 megawatts of additional power directly from the nuclear plant. Susquehanna agreed to sell it.
Tomi Engdahl says:
Camilla Hodgson / Financial Times:
FOI docs: water consumption by data centers in Virginia, home to world’s largest cluster of data centers, reached 1.85B+ gallons in 2023, up from 1.13B in 2019
US tech groups’ water consumption soars in ‘data centre alley’
Usage has increased by almost two-thirds since 2019 in the world’s largest concentration of computing infrastructure
https://www.ft.com/content/1d468bd2-6712-4cdd-ac71-21e0ace2d048
Tomi Engdahl says:
Washington Post:
Inside Equinix’s 114,300-square-foot data center in Northern Virginia, a region that is home to the largest concentration of data centers in the world — We toured a facility in Northern Virginia to see how it works and to understand why water use and energy consumption are such a concern.
https://www.washingtonpost.com/dc-md-va/interactive/2024/data-centers-tour-northern-virginia/?pwapi_token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJyZWFzb24iOiJnaWZ0IiwibmJmIjoxNzI2ODkxMjAwLCJpc3MiOiJzdWJzY3JpcHRpb25zIiwiZXhwIjoxNzI4MjczNTk5LCJpYXQiOjE3MjY4OTEyMDAsImp0aSI6ImM4NmMzNzdmLWZlODEtNDE3Yi04YTgwLTk0NWNkMzU0ZTgxNiIsInVybCI6Imh0dHBzOi8vd3d3Lndhc2hpbmd0b25wb3N0LmNvbS9kYy1tZC12YS9pbnRlcmFjdGl2ZS8yMDI0L2RhdGEtY2VudGVycy10b3VyLW5vcnRoZXJuLXZpcmdpbmlhLyJ9.1a1rELwzZD77EzAfU62E12POwGahohlSeCAxvOgqYwg&itid=gfta
Tomi Engdahl says:
https://etn.fi/index.php/13-news/16674-google-kuluttaa-enemmaen-saehkoeae-kuin-moni-valtio
Tomi Engdahl says:
Datakeskuksien sähköntarve voi johtaa energiapulaan. Tekoäly kasvattaa sähkönkulutusta räjähdysmäisesti, Gartner arvioi.
Raju ennnuste: Sähkönkulutus räjähtää käsiin, datakeskuksia uhkaa energiapula
7.12.202417:30
Sähkö
Tekoäly
Digitalous
Energia
Datakeskuksien sähköntarve voi johtaa energiapulaan. Tekoäly kasvattaa sähkönkulutusta räjähdysmäisesti, Gartner arvioi.
https://www.uusisuomi.fi/uutiset/us/6dcf9eef-678a-46f6-b24d-87548d17b0de?utm_term=Autofeed&utm_medium=Social&utm_source=Facebook&fbclid=IwZXh0bgNhZW0CMTEAAR1Bei2Hg6MBhnmFNjq4ioselTyBfgSNOAEcmdImnse1955O6ru6HSx7rfs_aem_10JXPpXf99lP7o9pFg_fhw#Echobox=1733586081
40 prosenttia datakeskuksista tulee olemaan toimintarajoitteisia vuoteen 2027 mennessä, kertoo konsulttiyhtiö Gartner. Tilanne johtuu tekoälyn kasvattamasta sähkönkulutuksesta.
Gartner arvioi, että tekoälyyn optimisoitujen datakeskuksien tarvitsema teho on 500 terawattituntia vuodessa vuonna 2027. Datakeskukset tarvitsivat vain 195 terawattituntia vuodessa viime vuonna. Gartnerin analyytikko Bob Johnson kertoo, että datakeskuksien vaatima sähköntarve tulee ylittämään sähkölaitosten kyvyn laajentaa kapasiteettiaan riittävän nopeasti.
Päivän politiikan aiheet kätevässä paketissa – Tilaa Uuden Suomen maksuton uutiskirje
”Tämä puolestaan uhkaa häiritä energian saatavuutta ja johtaa energiapulaan, mikä rajoittaa datakeskusten kasvua vuodesta 2026 alkaen”, Johnson kertoo Gartnerin analyysissa.
Gartner Predicts Power Shortages Will Restrict 40% of AI Data Centers By 2027
Sydney, Australia, November 12, 2024
Rapid Growth in Energy Consumption For GenAI Will Exceed Power Utilities’ Capacity
AI and generative AI (GenAI) are driving rapid increases in electricity consumption, with data center forecasts over the next two years reaching as high as 160% growth, according to Gartner, Inc. As a result, Gartner predicts 40% of existing AI data centers will be operationally constrained by power availability by 2027.
https://www.gartner.com/en/newsroom/press-releases/2024-11-12-gartner-predicts-power-shortages-will-restrict-40-percent-of-ai-data-centers-by-20270
Tomi Engdahl says:
Dina Bass / Bloomberg:
Microsoft unveils a data center design that uses zero water for cooling, down from the 125M+ liters per year a typical center uses, rolling out since August
https://www.bloomberg.com/news/articles/2024-12-09/microsoft-unveils-zero-water-data-centers-to-reduce-ai-climate-impact
Tomi Engdahl says:
Data centers accounted for about 1.5 percent of global electricity consumption in 2024, an amount expected to double by 2030 because of AI use
Data Centers Will Use Twice as Much Energy by 2030—Driven by AI
Data centers accounted for about 1.5 percent of global electricity consumption in 2024, an amount expected to double by 2030 because of AI use
https://www.scientificamerican.com/article/ai-will-drive-doubling-of-data-center-energy-demand-by-2030/?utm_campaign=sprinklr&utm_medium=social&utm_source=facebook
Tomi Engdahl says:
Microsoft announced that it is buying 3.7 million metric tons of carbon removal credits from CO280, a project developer that works with pulp and paper mills.
The purchase covers 12 years of emissions from CO280’s first carbon capture project at a mill on the Gulf Coast. CO280 expects the facility to start capturing CO2 in 2028.
Microsoft’s purchase should help it meet its goal of becoming a carbon-negative company by 2030 — that is, removing more carbon than its business produces.
Read more from Tim De Chant on Microsoft’s carbon removal deal here: https://tcrn.ch/3YvCuir
#TechCrunch #technews #climate #enterprise #Microsoft
Tomi Engdahl says:
https://hackaday.com/2025/05/22/recovering-water-from-cooling-tower-plumes-with-plume-abatement/
Tomi Engdahl says:
Gartner: sähkö ei riitä AI-datakeskuksiin
https://etn.fi/index.php/13-news/18189-gartner-saehkoe-ei-riitae-ai-datakeskuksiin
Tekoälyn nopea kasvu johtaa datakeskusten sähkönkulutuksen jyrkkään nousuun, varoittaa Gartner tuoreessa analyysissaan. Yhtiön mukaan datakeskukset kuluttavat vuonna 2025 yhteensä 448 terawattituntia sähköä, mutta määrä lähes kaksinkertaistuu vuoteen 2030 mennessä 980 terawattituntiin. Kasvua kiihdyttää ennen kaikkea AI-laskenta, joka lisää energiantarvetta huomattavasti nopeammin kuin perinteinen IT-kuorma.
Gartnerin tutkimusjohtaja Linglan Wangin mukaan erityisesti tekoälyyn optimoidut palvelimet muodostuvat keskeiseksi haasteeksi energiankulutuksen hallinnassa. Niiden sähkön käyttö viisinkertaistuu vuoteen 2030 mennessä 93 terawattitunnista peräti 432 terawattituntiin. Vuonna 2030 AI-palvelimet vievät jo 44 prosenttia kaikesta datakeskusten kuluttamasta sähköstä ja vastaavat 64 prosentista kulutuksen lisäyksestä.
Alueellisesti kasvu painottuu Yhdysvaltoihin ja Kiinaan, jotka muodostavat kaksi kolmasosaa maailman datakeskussähkön kysynnästä. Gartnerin mukaan Kiina on jonkin verran paremmassa asemassa energiatehokkaampien palvelinten ja suunnitelmallisemman kapasiteettirakentamisen ansiosta. Yhdysvalloissa datakeskusten osuus kokonaiskulutuksesta kohoaa 4 prosentista 7,8 prosenttiin vuoteen 2030 mennessä. Euroopassa nousu on maltillisempi, mutta silti selvästi havaittava: 2,7 prosentista 5 prosenttiin.
Kasvava sähkön tarve pakottaa datakeskustoimijat etsimään uusia ratkaisuja energiansaannin turvaamiseksi. Gartner arvioi, että nykyinen malli, jossa valtaosa varavoimasta perustuu fossiilisiin polttoaineisiin, ei ole pitkällä aikavälillä kestävällä pohjalla. Vaihtoehtoiset ”puhdasta sähköä” tarjoavat teknologiat kuten vihreä vety, geoterminen energia ja pienet modulaariset ydinreaktorit ovat vasta nousemassa, mutta Gartner ennustaa niiden tulevan datakeskusten mikroverkoissa käyttökelpoisiksi tämän vuosikymmenen loppupuolella.
Tomi Engdahl says:
Nearly 7,000 of the world’s 8,808 data centers are built in the wrong climate, analysis find — vast majority located outside optimal temperature range for cooling, 600 in locations considered too hot
News
By Luke James published yesterday
Most facilities sit outside the temperature range recommended for efficient operation, as AI growth pushes data centers into hotter regions.
https://www.tomshardware.com/tech-industry/nearly-7000-of-the-worlds-data-centers-are-built-in-the-wrong-climate
Tomi Engdahl says:
Toxic Neighbor
Amazon Data Center Linked to Cluster of Rare Cancers
“The historical precedent here is Flint, Michigan.”
https://futurism.com/artificial-intelligence/amazon-data-center-oregon
For the hundreds of communities who’ve been saddled with data centers in recent years, the bulky fixtures are sources of unbearable noise, soaring energy prices, and plenty of electrical fires.
Add another grim possibility to that list: debilitating rare cancers.
Doherty’s story began when he noticed a rise in bizarre medical conditions among the county’s 45,000 residents, linked to toxins in the local water. Working with the county health office, the rancher-turned-official began a survey of 70 wells throughout his jurisdiction — 68 of which, his testing found, violated the federal limit for nitrates in drinking water.
But the spike in cancer-causing pollution wasn’t just the fault of local farms, as Doherty expected. It had its roots in a 10,000 square foot data center by the commerce giant Amazon, which first went online in Morrow County in 2011.
Basically, the allegations go like this: industrial megafarms operating in the area are responsible for churning out millions of gallons of wastewater, laden with nitrates from fertilizers. All that waste has to go somewhere, which is one way of saying it mostly ends up in the ground.
Amazon’s hulking data center, thirsty for water to cool its blazing hot computer chips, supercharged this process, adding millions of gallons of wastewater a year to the heavy volume of farm runoff, which Morrow County was already struggling to keep up with. Soon even the deepest reaches of the local aquifer were tainted, according to RS, as huge volumes of data center and agricultural wastewater saturated the water table.
Tomi Engdahl says:
https://www.facebook.com/share/p/17oPJ4NPPw/
Someone asked a question that sounds genius at first:
If data centers need so much cooling, why don’t we just put them in Antarctica?
Freezing air.
Natural cooling.
Problem solved.
Except not really.
Because cooling is only one part of what a data center needs.
These buildings also need massive amounts of reliable electricity.
High-speed fibre connections.
Roads.
Workers.
Repairs.
Spare parts.
Backup systems.
Security.
And fast connection to the people actually using the internet.
Antarctica has the cold.
But it does not have the rest.
A data center in the middle of nowhere is not useful just because it is chilly.
It still has to be powered.
Connected.
Maintained.
Protected.
And close enough to users that everything does not feel slow.
That is why cold-climate data centers exist in places like the Nordics, not on the bottom of the world.
The real answer is simple:
Data centers do not just need cold air.
They need an entire civilization around them.
Tomi Engdahl says:
Data center carbon emissions reached an 286 million tonnes in 2025 — exceeding projections by 57%.
The rapid expansion of power-hungry data centers is creating an unprecedented environmental footprint that far exceeds earlier estimates. A comprehensive study by Allianz Trade reveals that these facilities emitted 286 million tonnes of carbon dioxide in 2025, largely driven by the massive computing needs of artificial intelligence. AI workloads currently account for 15% to 20% of data center electricity consumption, a figure projected to surge to 40% by 2030.
Without aggressive grid decarbonization, emissions are expected to more than double by the end of the decade, escalating annual climate damages to $154 billion.
This digital surge is also placing immense pressure on global infrastructure and natural resources.
Data centers are projected to consume up to 1.8 trillion liters of water annually by 2030, a volume comparable to the entire yearly consumption of Switzerland. The environmental impact is heavily concentrated geographically, with nearly 70% of global data center emissions originating in the United States and China.
To curb this impact, experts emphasize that energy grid decarbonization must become a structural priority, as the same computational workload can produce over twenty times more emissions depending entirely on whether it runs on a fossil-fueled or a clean-energy grid.
source: Allianz Trade. (2026). Code, carbon, kilowatts: AI’s hidden toll and the race to green the grid. Allianz Trade Research.
https://www.facebook.com/share/p/1EHXj4cx2n/
Tomi Engdahl says:
https://www.facebook.com/share/p/191hWMQyep/
Report: It may be impossible to make AI data centers pay their fair share.
As tech companies race to build massive data centers across the United States, utilities are being forced to answer a difficult question: who should pay for the new electricity infrastructure needed to power them?
The answer is not as simple as making data centers pay their “fair share.”
Data centers require enormous amounts of electricity, sometimes enough to strain local grids. Supplying that power can require new substations, transmission lines, and power generation. Some costs are easy to assign. If a data center needs a dedicated power line from a nearby substation, it can be billed directly.
But many upgrades become part of the shared electric grid.
If a utility expands a substation, strengthens transmission lines, or secures more electricity for the region, those costs may be spread across many customers, including households and small businesses.
There is another complication. Many data centers can automatically reduce their power use during moments of peak demand. Since some electricity costs are calculated based on how much customers use during those peak moments, a data center could consume huge amounts of electricity overall while still lowering its share of certain grid costs.
That makes “fair share” hard to define.
Ratepayers also face another risk. Utilities often build infrastructure based on projected demand. If a planned data center is delayed, canceled, or uses less electricity than expected, the cost of those upgrades may still remain in the system for years.
The debate is no longer just about AI.
It is about whether ordinary customers could end up helping pay for the power grid built to support it.
Learn more:
Read the report by theodore J Kury, Director of Energy Studies at University of Florida titled “It may be almost impossible to make data centers pay their ‘fair share’ of electricity costs.” The Conversation.
Tomi Engdahl says:
https://www.facebook.com/share/p/1EKsDsseEs/
Data center boom is now sparking a national call for action as more than 230 environmental groups urge Congress to pause new construction until stronger rules are in place. Their concern is that fast-growing AI infrastructure is putting pressure on energy grids, water supplies, and local communities.
These groups say data centers are expanding faster than regulations can keep up. Many facilities use large amounts of electricity and water for cooling, which can affect local utility costs and strain already stressed infrastructure.
The debate shows a bigger challenge in the AI era. Innovation is moving quickly, but the physical systems behind it need land, power, and water. The question now is how to support technology without ignoring its impact on communities and the environment.
#fblifestyle #Polarbear #DataCenters #ArtificialIntelligence #Environment
Tomi Engdahl says:
https://www.facebook.com/share/p/1M5c5RyN42/
A new report suggests the rapid expansion of artificial intelligence could significantly increase the environmental footprint of the world’s data centers.
According to Allianz Trade, global data centers could produce about 286 million metric tons of carbon dioxide in 2025, roughly 57% higher than previous estimates published by the International Energy Agency. The report attributes much of the increase to the growing demand for AI, which requires enormous amounts of computing power and electricity.
AI is already becoming one of the largest users of data center resources.
The report estimates that AI currently accounts for 15–20% of global data center electricity consumption, and that share could rise to around 40% by 2030 as more businesses and consumers adopt AI-powered technologies.
Researchers warn that the long-term climate impact will depend largely on how data centers are powered.
If electricity grids continue relying heavily on fossil fuels, the report projects that emissions from data centers could more than double by the end of the decade, resulting in an estimated US$154 billion in annual climate-related damages.
Water use is also becoming a growing concern.
Many large data centers rely on water-based cooling systems to keep thousands of servers operating efficiently. According to the report, global data center water consumption could reach 1.3 to 1.8 trillion liters per year by 2030, an amount comparable to Switzerland’s annual water use.
The report also emphasizes that not all data centers have the same environmental footprint. Facilities powered primarily by renewable energy generate substantially fewer emissions than those dependent on coal, oil, or natural gas, making the transition to cleaner electricity an important factor as AI infrastructure continues to expand.
Tomi Engdahl says:
steel mills produce much more CO₂ than data centers.The global steel industry is responsible for about 7% to 8% of all human-made CO₂ emissions. By contrast, data centers are responsible for a significantly smaller fraction, though their power needs are growing rapidly.
Tomi Engdahl says:
The concrete industry is responsible for about 8% of global carbon dioxide emissions, producing 2.5 billion tonnes annually. Over 90% of this footprint stems from manufacturing clinker (the binder in traditional cement). About one-third of clinker emissions come from burning fossil fuels for kiln energy, while two-thirds are unavoidable process emissions released when limestone is burned (calcined).
https://www.ecocemglobal.com/act/cement-has-a-problem/
Tomi Engdahl says:
https://www.facebook.com/share/p/1JYHhxBVHg/
Amazon’s data centers consumed 2.5 billion gallons of water in a year.
That is enough water to fill roughly 3,800 Olympic-size swimming pools.
The company operates more than 900 data centers across over 50 countries, supporting services ranging from online shopping and cloud storage to artificial intelligence.
Data centers contain thousands of powerful servers that generate enormous amounts of heat. Many facilities use water-based cooling systems to remove that heat and prevent equipment from failing.
Amazon said its total data center water use fell by 2 percent compared with 2024, even as the company continued expanding its global operations.
The company attributed the decrease to more efficient cooling systems and said it only uses water-based cooling when necessary, particularly during hotter periods.
But the disclosure comes as communities and lawmakers demand greater transparency over how much water data centers consume.
Water use can vary dramatically depending on a facility’s location, climate, cooling technology, and the source of its electricity. In drought-prone areas, even a relatively efficient data center can place additional pressure on local supplies.
Utah recently passed what is believed to be the first U.S. law requiring certain new data centers to publicly disclose their annual water use.
Elsewhere, residents have raised concerns about falling water pressure, higher utility bills, and competition for limited supplies near rapidly expanding data center developments.
Amazon argues that the industry’s water consumption is often exaggerated and says its new figures show a more measured reality.
Still, public support remains limited.
A recent national survey found that only one in three Americans approved of new data center construction, while just 14 percent said they would accept one being built near their home.
Learn more:
“Amazon Says Its Data Centers Used 2.5 Billion Gallons of Water in 2025.” WSJ
Tomi Engdahl says:
https://people.com/amazon-data-centers-used-2-5-billion-gallons-water-2025-11996145
https://www.datacenterdynamics.com/en/news/amazon-data-centers-used-25bn-gallons-of-water-in-2025/
Tomi Engdahl says:
Water changes form but never truly disappears. Through the Earth’s Water Cycle, the exact same water is continuously recycled. While tiny amounts break apart in the upper atmosphere and escape into space, the total volume on Earth remains remarkably constant
Tomi Engdahl says:
Nationally, golf courses consume significantly more water than data centers.According to data from the Golf Course Superintendents Association of America, U.S. golf courses consume roughly 2.08 billion gallons of water per day for irrigation. By comparison, the entire U.S. data center industry directly consumes an estimated 449 million gallons per day for server cooling. This means golf courses out-water data centers nationally by a margin of nearly 4 to 1.
Tomi Engdahl says:
Aaron Feledy data centers that use evaporative cooling convert water from liquid form to gas form (visible steam and/or water absorbed by dry air).
When water evaporates, it absorbs energy from its surroundings to break the intermolecular bonds holding its molecules in liquid form.
This energy is called the latent heat of vaporization.
When hot air passes over water, the heat energy in the air is absorbed by the water, turning it into vapor.
Consequently, the temperature of the air drops significantly.
There are Direct Evaporative Cooling, Indirect Evaporative Cooling, Cooling Towers (Chilled Water Loop).
Consumed Water (~80% to 85%): This is the water that physically evaporates and is released into the atmosphere as water vapor. It does not return to the municipal water supply immediately.
.Aaron Feledy In closed loop system the water is only warmed up a handful of degrees as it passes through the data center. Then water goes to industrial cooler/chiller system (or long tube system in cold sea/lake water) that cools down the water that is sent back to data center.
Tomi Engdahl says:
Data centers can introduce several types of pollutants into water systems. These pollutants fall into three main categories.
Biocides: chemicals to prevent the growth of bacteria, mold, and algae in the cooling system, and the most common pollutant in data center water systems.
Corrosion inhibitors: chemicals that prevent corrosion within cooling systems
Heavy metals: metals, such as copper zinc, and lead, that can enter the water in trace amounts when metal components degrade. Any other questions?
Tomi Engdahl says:
Additionally, fine salt dust that drifts in the atmosphere from evaporative cooling towers can settle on nearby soils and surface waters, potentially harming vegetation and increasing salinity in aquatic systems. This concern is especially pressing given that salt pollution is already a growing issue in certain water supply reservoirs, including the Occoquan.
Tomi Engdahl says:
Data Centers use a closed loop system (similar to car radiators) and thus the water literally cycles through the Data Center.
Any water that DOES evaporate, goes into the air and joins the water cycle. Comes down as rain, goes into rivers, runs to the ocean, evaporates back into rain, etc
Tomi Engdahl says:
The United States agricultural sector withdraws more freshwater in just two days (~236 billion gallons) than the entire U.S. data center industry consumes in an entire year.
Tomi Engdahl says:
Data center in USA use both evaporative cooling (especially older ones) and closed loop cooling systems (new installation).
Average WUE water consumption:
USA ~0.55 L/kWh (Ranges from 0.1 to 1.8+)
Europe ~0.31 L/kWh
Finland ~0.01 L/kWh
Virtually Zero Direct Water Use. Finland’s naturally cold sub-Arctic climate allows data centers to use ambient air-cooling (direct or indirect) year-round. Cold outdoor air is used to cool the servers, meaning they rarely, if ever, need to evaporate water on-site.
The heat from servers is also used to heat buildings on winter time. Google data center heats a very big part of city of Hamina. Microsoft and two other data centers heats up considerable part of Espoo.
Tomi Engdahl says:
The problem is getting the water back to the area it came from, after it condenses and rains.
Tomi Engdahl says:
Then WHY are we still building them.
The better question would be why are the people building polluting and water consuming data centers to wrong locations?
When it is possible to build data centers with very little problems. Google has a data center outside USA that is powered from renewable power, uses almost no water and sends waste heat to heat many buildings during winter time.
Tomi Engdahl says:
maybe the one in China that run off coal. data centers aren’t a issue they are only like 0.04% of industrial energy usage, data centers are more profitable when they run efficiently. more they can process well paying less to do the processing the better.
Tomi Engdahl says:
when you evaporate water to atmosphere, it will sooner or later form clouds that will rain it back down. All plants evaporate much more water.
The Scale Difference: U.S. agriculture evaporates 100 times more water annually than the entire U.S. data center industry.
Tomi Engdahl says:
Data centers are among the most water-efficient industrial facilities ever built. Nationwide, data centers consume less than 0.05 percent of total U.S. freshwater withdrawals, a tiny fraction compared to agriculture, thermoelectric power generation, and manufacturing. Furthermore, unlike traditional factories or refineries, data centers do not produce chemical runoff or industrial wastewater. Their primary output is heat managed through sealed equipment loops with no external discharge.” — Goldwater Institute
Tomi Engdahl says:
Water Recycling & Conservation
Eliminating Potable Water Use: Modern data centers are increasingly shifting away from using drinking-grade water. Instead, they are designed to use recycled wastewater (greywater) or industrial water that isn’t fit for human consumption.
Closed-Loop Systems: Rather than continuously evaporating water into the air, newer facilities use closed-loop liquid cooling. This cycles the same water repeatedly through a sealed system, reducing active water consumption to near zero.
District Heating: In colder regions, some data centers capture the hot water/air generated by servers and feed it directly into municipal district heating systems, recycling the thermal energy to heat local homes.
Tomi Engdahl says:
Spreading data centers out (de-clustering) and focusing on water recycling are two of the most effective ways to solve their localized resource strain. Spreading facilities out to regions with colder climates (like Finland) or areas with abundant, underutilized renewable energy distributes the resource load, making it much easier for local environments to absorb.
Tomi Engdahl says:
The chemicals used in closed-loop systems are the exact same standard HVAC chemicals used in hospital, airport, and high-rise office building air conditioning systems.
Glycol (Propylene or Ethylene): Used as an antifreeze. Propylene glycol is non-toxic (often used in food processing), while ethylene glycol is toxic but highly biodegradable.
Corrosion Inhibitors & Biocides: Used in tiny concentrations to prevent rust and algae growth.
Tomi Engdahl says:
https://www.facebook.com/share/p/1Bo5GVQfax/
AI is transforming the world, but its growing demand for water is raising new environmental concerns. Estimates suggest AI-related data centers consumed around 264 billion gallons of water in 2025, much of it used to cool powerful servers running advanced artificial intelligence systems. At the same time, nearly 63% of the United States experienced drought conditions, highlighting the challenge of expanding digital infrastructure while protecting limited water resources.
Researchers and environmental experts say future AI growth will require more efficient cooling technologies, water recycling, and smarter data center designs to reduce environmental impact. Companies are already investing in air cooling, recycled water, and liquid cooling systems to improve sustainability.
As AI becomes part of everyday life, every breakthrough reminds us that innovation also depends on protecting the natural resources that make it possible.
Sources: Barchart (2026), U.S. Drought Monitor, Mordor Intelligence, The Guardian.
Tomi Engdahl says:
Todd Wolfe AI can be done without consuming a lot of water. Google, Microsoft, Meta and AWS data centers use very little amount of water on many countries outside USA. Evaporative cooling only kicks in on the hottest few days per year.
Tomi Engdahl says:
Crop irrigation is inherently consumptive because water is lost directly to plant transpiration and soil evaporation.
Paul Avery Agriculture completely dwarfs tech infrastructure when comparing total water consumption in the United States. In fact, California’s avocado orchards alone consume roughly 11 to 14 times more water annually than all U.S. data centers combined.
Crop irrigation is inherently consumptive because water is lost directly to plant transpiration and soil evaporation.
However, the core issue is not the national volume, but localized regional stress, as both industries heavily compete for the exact same water basins in arid states like California, Arizona, and Texas.
Tomi Engdahl says:
Direct data center water consumption represents roughly 0.3% to 0.4% of the total daily water withdrawals in the United States. Even when combining direct and indirect water (like power plants), the data center industry consumes under 1% of the nation’s total freshwater footprint.
Tomi Engdahl says:
https://www.facebook.com/share/p/193qHRbZJh/
The rapid expansion of AI infrastructure has focused attention on the energy demands of large scale data centers. Researchers continue studying electricity consumption, cooling technologies, carbon emissions, and sustainable computing strategies.
Advances in processor efficiency, liquid cooling, renewable energy integration, and optimized software are helping reduce the environmental footprint of high performance computing. Sustainable AI remains an active field of research.
For technology enthusiasts, the story highlights one of the most important engineering challenges facing the AI industry.
Tomi Engdahl says:
The future of AI also includes energy challenges https://bit.ly/4fjdfYn
Tomi Engdahl says:
https://www.facebook.com/share/p/195zMzQ5Rb/
A national crowdsourced mapping project has uncovered a terrifying pattern: artificial intelligence data centers are aggressively expanding directly inside America’s most severe drought zones.
For decades, environmental crusader Erin Brockovich fought over contaminated groundwater.
Now, her latest initiative is exposing how the physical infrastructure of the AI boom is quietly threatening municipal water tables across the country.
The numbers are staggering.
When thousands of citizen-reported data center locations were overlaid onto federal drought maps, a glaring concentration appeared across parched regions in the South and Southwest.
The timing represents a critical infrastructure crisis.
More than 60 percent of the United States is currently battling drought conditions, yet hyperscale computing campuses are demanding millions of gallons of potable water every single day just to keep their server racks from melting down.
The concerns are not abstract.
According to environmental models from researchers at UC Riverside, generating a single 100-word AI response consumes roughly one plastic water bottle’s worth of cooling water and grid electricity.
And the impact is already fueling widespread civic backlash.
In community after community, families report identical patterns of corporate secrecy: industrial projects announced only after permits are locked in, local officials signing strict non-disclosure agreements, and developers ignoring neighborhood questions about shrinking aquifers.
We are witnessing a dangerous structural trade-off where the digital cloud is physically drinking our most critical survival resource dry.
When the massive cooling demands of artificial intelligence directly threaten municipal drinking aquifers during severe droughts, should local communities have the absolute legal right to veto high-density data centers?
(Full source and data linked in the comments below!
Source & Geospatial Investigation Verification:
Reported via the national crowdsourced data center mapping project initiated by Erin Brockovich, cross-referenced against federal drought tracking metrics compiled by the U.S. Drought Monitor and regional investigation data published by Newsweek.