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.
1,268 Comments
Tomi Engdahl says:
“Unless governments and regulators insist on a rapid phaseout, we will face a new tidal wave of forever chemicals.” https://trib.al/unhEULZ
You Call That Slop?
Data Centers Are Causing a Huge Surge in Incredibly Toxic Chemicals Polluting the World
“Data centers create problems that don’t stop at municipal boundaries, whether it’s pollution or straining the grid and driving up everyone’s electric bills.”
https://futurism.com/science-energy/data-centers-surge-toxic-chemical-spill-pollution-pfas?fbclid=IwdGRjcAUXo5hjbGNrBRejh3Bkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEeYuUYIMcG6ZhwTvZ9QoOp451zWJl9gDVCuXGiJ9mGILPmXyDAlfcSjFkHu5E_aem_K-tbmlTnJWit4pvTJo_HgQ
Data centers don’t just undergird the avalanche of AI slop polluting our social media feeds — they’re spewing an awful lot of physical slop as well.
On top of the noxious air pollution, as well as the occasionally bacteria-laden wastewater discharge, data centers are responsible for a rising tide of forever chemicals, artificial substances which never degrade naturally.
According to new reporting by The Guardian, nearly all of the top ten global producers of forever chemicals, otherwise known as polyfluoroalkyl substances, or PFAS, plan to ramp up production quotas on the back of the data center boom.
There’s a clear market demand for this: PFAS play a huge role in data center construction, where they’re used in cooling systems, AI chips, circuit boards, and fire suppression systems — not to mention the coating on thousands of miles of cables used to connect these components together.
‘Tidal wave’ of Pfas being launched to satisfy AI industry, campaigners warn
Big manufacturers planning to produce more of the forever chemicals to meet demand from datacentres
https://www.theguardian.com/environment/2026/sep/14/pfas-firms-tidal-wave-forever-chemicals-ai-industry-demand-datacentres?fbclid=IwVERDUAUXo_RwZG9mBWV4dG4DYWVtAjEwAHNydGMGYXBwX2lkDDM1MDY4NTUzMTcyOAABHtbKdH2P375HIBQE4Ms39nMfzryGC8oxqYuQukN7PhPEha2IQkDFdr7Wdmu-_aem_Bk6RiorjHxLhlaKo6bp_-A
Pfas companies are launching a “tidal wave” of forever chemicals production to meet demand from the AI industry, a campaign organisation has warned.
Despite alarm over the chemicals, which have been linked to cancer, a survey of the 10 biggest manufacturers found most had plans to increase production.
Many frame their investments around the “AI revolution”, because of the role Pfas play in semiconductor production and next-generation datacentre cooling systems.
But opponents warn that the chemicals, which never degrade in nature, are a ticking timebomb for human and planetary health. They say increased production risks undermining nascent moves to restrict Pfas in Europe, the UK and parts of the US.
“These companies’ expansion plans show that unless governments and regulators insist on a rapid phaseout, we will face a new tidal wave of forever chemicals,” said Anne-Sofie Bäckar, the executive director of ChemSec, the Swedish chemicals watchdog that carried out the research.
Tomi Engdahl says:
https://www.facebook.com/share/p/19WeHebior/
The cloud is going nuclear, and that’s a good thing. Data center operators, facing the need for clean, firm, 24/7 power, are investing heavily in nuclear energy. Amazon, Google, and Microsoft have all announced deals or investments in nuclear projects, including small modular reactors (SMRs). Amazon is backing X‑energy, a developer of advanced SMRs, and Microsoft has signed a deal with Constellation Energy to power its data centers with nuclear. The total investment exceeds $1 billion. Nuclear offers what solar and wind cannot: continuous, carbon‑free baseload power, unaffected by weather. For data centers that must run 24/7, nuclear is the perfect complement to renewables. The new generation of SMRs is designed to be safer, cheaper, and faster to deploy. The data center that hosts your cloud storage could soon be powered by a reactor the size of a shipping container. The cloud is becoming the nuclear industry’s biggest customer, and that’s accelerating the clean energy transition. #nucleardatacenter #SMR #cleanpower #datacenterenergy #nuclearrenaissance
Tomi Engdahl says:
https://www.facebook.com/share/p/19FqbBkLct/
The explosive growth of AI is creating an environmental challenge that extends far beyond the computers themselves.
AI models require increasingly powerful servers, and those servers operate inside data centers that consume large quantities of electricity. Where that electricity comes from can have a major effect on associated greenhouse-gas emissions.
Researchers are also examining the water and land footprints associated with AI infrastructure. A 2026 review in Nature Reviews Clean Technology identified growing concerns over greenhouse-gas emissions, pressure on power grids and freshwater resources as AI data centers expand.
The United Nations University has similarly warned that AI’s environmental footprint includes energy, water and land, with global data-center electricity demand associated with AI projected to grow substantially by 2030.
Importantly, AI’s environmental impact is not predetermined. More efficient chips, renewable electricity, advanced cooling, better data-center design and improved grid planning can significantly reduce the footprint.
The central question is becoming whether the AI industry can grow fast enough to meet demand while decoupling that growth from rising environmental damage.
#AI #DataCenters #ClimateChange #CleanEnergy #Sustainability
Tomi Engdahl says:
https://www.facebook.com/share/p/1EXAUVwrrM/
The data center that uses no water at all. Microsoft’s new data center region in Sweden is the first in the industry to use zero‑water cooling. Instead of evaporative cooling or water chillers, the facility uses outside air and direct expansion cooling, eliminating water consumption entirely. This is critical in a world facing increasing water scarcity. The data centers are also powered by 100% renewable energy, sourced from Swedish wind and hydro, and Microsoft has committed to matching its consumption on an hourly basis. The facility is a model for sustainable data center design, proving that large‑scale computing can operate with minimal environmental impact. Sweden’s cool climate makes zero‑water cooling feasible, and Microsoft is now applying the lessons learned to other regions. The server that stores your files is now cooled by the Swedish breeze, not by water. The data center that once consumed millions of gallons of water now consumes none. The future of computing is water‑free. #zerowatercooling #microsoftsweden #datacenter #cleancomputing #sustainableinfrastructure
Tomi Engdahl says:
The Environmental Protection Agency announced that it will roll back emissions limits on coal- and gas-fired power plants — one of the Trump administration’s most significant moves yet to undo the climate policies enacted under Presidents Joe Biden and Barack Obama.
The change would overturn a 2024 rule requiring power plants to use carbon sequestration equipment to control pollution; it was one of the biggest climate change rules implemented by the Biden administration’s EPA.
The agency also proposed eliminating all other greenhouse gas emissions rules, arguing that under the Clean Air Act, it doesn’t have the authority to regulate those emissions as they pertain to climate change.
Read more: nbcnews.to/3UIdUfr
Tomi Engdahl says:
https://www.facebook.com/share/p/1DCTPbhqEe/
The water used to cool a data center may be only the visible part of its true water footprint.
When companies discuss water consumption, attention often centers on what is used directly inside the facility. But data centers require enormous amounts of electricity—and many power plants also consume water while producing that electricity.
A Lawrence Berkeley National Laboratory report estimated that U.S. data centers directly consumed about 17 billion gallons of water in 2023. The electricity supplying them was associated with roughly 211 billion gallons of additional water consumption—more than 12 times the on-site amount.
That hidden footprint varies dramatically with the source of electricity. Wind and solar photovoltaic power generally consume very little water during operation. Nuclear, coal and natural-gas power plants can have substantially larger water requirements, particularly when they depend on water-intensive cooling systems.
This matters in water-stressed regions such as Texas and New Mexico, where AI infrastructure is expanding alongside existing agricultural, residential and industrial demand.
One prominent example is Project Jupiter, a huge AI data-center campus under construction in southern New Mexico. Oracle is its tenant, and the project is associated with OpenAI’s Stargate infrastructure. Oracle says a closed-loop cooling system will require an initial fill but very little water afterward. That means the national 12-to-1 estimate should not be applied automatically to this individual project.
However, low on-site water use does not necessarily mean a low total footprint. The water associated with producing a facility’s electricity must also be considered—and that number depends on exactly how its power is generated.
For communities deciding whether they have enough water to support farms, homes and new AI facilities, partial reporting is not enough. Understanding the full impact requires counting both the water used inside the building and the less-visible water used to keep its electricity flowing.
Tomi Engdahl says:
https://www.facebook.com/share/p/1ScffmcJcz/
Europe is increasingly treating data-center waste heat as a potential energy resource rather than simply something to discharge into the environment. Under the EU Energy Efficiency Directive, data centers with more than 1 MW of rated energy input are required to utilize waste heat or other recovery applications unless technical or economic feasibility prevents it.
That distinction is important: the rules do not simply require every data center to pipe its heat directly into nearby homes. They require consideration and utilization where feasible, including connections to district heating systems.
The idea is becoming increasingly relevant as AI and cloud computing push electricity consumption in data centers higher. Much of that electricity eventually becomes heat, creating a potentially valuable secondary energy stream.
Finland is already demonstrating what this can look like in practice, with Microsoft’s data-center projects designed to feed recovered heat into district heating networks.
The broader concept is simple: if communities need heat while data centers need cooling, connecting the two could turn an unavoidable byproduct into useful infrastructure.
#DataCenters #WasteHeat #EnergyEfficiency #AIInfrastructure #CleanEnergy
Tomi Engdahl says:
https://www.scrippsnews.com/the-price-of-data/how-much-water-do-data-centers-really-use-heres-what-we-learned-from-public-records?fbclid=IwdGRjcAUYYzVjbGNrBRhi_HBkb2YFZXh0bgNhZW0CMTEAc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEexqYuQ93ZnKFxXkuiDLKbrTUP2K6voABtG8v84yi3-5ae5V4ZE3RWQZlAXTI_aem_uCGqZJ9OL95R4h4D0THECA
Tomi Engdahl says:
The AI land panic is wildly out of proportion.
By 2030, the land footprint of every data center in America, plus the land around the buildings, is expected to be about the size of Rhode Island.
Meanwhile, corn grown for ethanol occupies an area about the size of New York State—all for a fuel that is probably dirtier than the gasoline it replaces.
Read more: https://humanprogress.org/the-ai-land-and-water-panic-is-wrong/
Tomi Engdahl says:
https://www.foxnews.com/politics/8-everyday-things-would-quit-working-without-data-centers
Tomi Engdahl says:
To put that single AI-generated Facebook post into physical perspective:
Energy (0.0046 kWh): Equivalent to running an efficient LED lightbulb for about 30 minutes, or charging a smartphone from 0% to about 40–50%.
Water (12 mL): Equivalent to less than one tablespoon of water (or roughly 1/20th of a standard 250mL drinking glass).
Tomi Engdahl says:
https://www.facebook.com/share/p/1FHr46Dqej/
People lived centuries without data centers, but now both are competing for the same rivers.
Every time you use artificial intelligence, massive data centers are working hard in the background. Because these powerful machines run incredibly hot, facilities use millions of gallons of fresh water to keep them cool. Surprisingly, nearly half of these buildings are located in areas already struggling with severe water shortages today.
#photography #ArtificialIntelligence #WaterConservation
References:
Lincoln Institute of Land Policy:The growing competition between data centers and local clean water sources.
Fortune Magazine:How AI data centers are consuming local water supplies across America.
Reuters News:Environmental impact of artificial intelligence and water usage in tech facilities.
Tomi Engdahl says:
https://www.facebook.com/share/p/19iJ78Q7kd/
China just powered a desalination plant from nuclear energy at commercial scale — connecting the Tianwan Nuclear Power Plant’s fourth reactor to a dedicated 100,000 cubic metre per day seawater desalination facility, making it the largest nuclear-powered desalination plant on earth and providing drinking water to 500,000 Jiangsu Province residents from Yellow Sea seawater with zero fossil fuel at any stage.
The Tianwan Nuclear Desalination Project Phase 1, developed by China National Nuclear Corporation and Jiangsu Zhongde Desalination, uses low-pressure steam from the Tianwan-4 reactor’s secondary circuit — heat that would otherwise be rejected through cooling towers — to drive multi-effect distillation desalination units that evaporate seawater through a series of decreasing-pressure chambers, condensing fresh water from the vapour in each stage and producing 100,000 cubic metres of drinking water daily at a marginal energy cost approaching zero because the driving heat is waste from power generation.
The nuclear desalination economics are transformative for coastal water security. Conventional reverse osmosis desalination requires 3 to 5 kilowatt-hours of electricity per cubic metre — a significant operating cost at commercial scale. Multi-effect distillation driven by nuclear waste heat requires only 0.3 kilowatt-hours of electricity per cubic metre for pumping and instrumentation, reducing operating cost by 85% compared to electrically driven desalination.
China’s coastal provinces — home to 600 million people — face water stress that groundwater overextraction has converted from seasonal scarcity to structural deficit. Nuclear desalination provides an inexhaustible coastal water supply from the Yellow and East China Seas using energy already generated for the grid.
China waters its cities from the sea using reactor waste heat. The Yellow Sea becomes a freshwater source.
Source: China National Nuclear Corporation (CNNC) & Jiangsu Zhongde Desalination, 2024
Tomi Engdahl says:
If you are reading this, thank the data centers you are depending on, for always being there.
If you think data centers are only for AI, think again.
If you use a cellphone or a computer you are using data centers. If you are using social media, you are using data centers, along with data centers for every app that you use.
If you are employed, your employers is dependent on multiple data centers.
Your grocery store is dependent on data centers. Along with every single business, educational, healthcare, entertainment and public utility service you depend on.
Your local, county, state and federal government are all dependent on data centers.
We are as much dependent on data centers as we are dependent on our phones and electricity itself.
The United Stated is over 3m square miles. Data Centers today occupy 300 square miles. There has got to be room and proper locations for data centers.
Unless you are completely living off the grid, you are dependent upon data centers every day.
Tomi Engdahl says:
and, physically, we are far more far more dependent on water than AI. Emotionally, we are as dependent on peace and quiet. Financially, we’re dependent on affordable energy for our homes. That would include electricity, natural gas and propane. These are listed as power sources for data centers. Their backup power is listed as gasoline and diesel – with limited pollution control.
we’re also dependent on open land for agriculture.
Tomi Engdahl says:
Our phones and personal computers absolutely can use AI. The important distinction is where the AI computation happens.
Some AI tasks run directly on the device, while many content-generation and larger AI models run in cloud data centers. So when we use an AI service through a phone, the phone may simply be the interface while the heavy computation happens on remote servers.
That’s actually part of the point of our post: the environmental footprint isn’t determined simply by whether someone is holding a phone or computer. It depends on where the computing happens, how much computing is required, and how the infrastructure is powered and cooled.
Tomi Engdahl says:
Exactly—the analogy helps explain an important point. A phone’s internal cooling system doesn’t normally require you to refill water because it relies on passive heat dissipation or other non-water cooling methods.
Data centers are different because they operate thousands of high-power servers continuously and generate vastly more heat. Some facilities use air cooling or closed-loop liquid cooling, while others use evaporative systems that require ongoing water replacement.
So it’s not quite the same cooling principle at every scale—but the key point stands: water use is not inherent to computing itself; it depends on how the facility is designed and cooled.
Tomi Engdahl says:
Of course we should be cognizant of environmental concerns with the data centers. But, don’t let China convince you data centers are all bad while they have thousands planned. They are playing us with social media.
There’s a whole litany of pop-up foreign enemy agent profiles across social media posting all kinds of divisive sh*t to cause controversy and discord in the U.S. It’s Psyops 101, and I hope our intel services are doing the same to their countries in spades.
If Americans hadn’t completely lost the ability to critically think, it wouldn’t be effective. But…here we are.
Tomi Engdahl says:
can depend on the case by case how much. I have to disagree with most of your claims. I have worked near one big data center for years and live few hundred meters from a smaller one.
Those have not been very noisy.
They have not taken the water.
I still see wild animals, even rare ones here.
For bees I don’t have data.
Those data centers were built based on European building practices. Your situation in USA could be different.
Tomi Engdahl says:
Microsoft’s Finland data-center project demonstrates how computing infrastructure can become part of a city’s heating system. Microsoft says waste heat from its Finnish data centers will eventually provide heat for around 250,000 people through district heating networks serving Espoo, Kauniainen and Kirkkonummi.
The system uses heat pumps to raise the temperature of heat recovered from the data-center cooling system before it is transferred into the municipal heating network. Microsoft says the system is scheduled to begin operations in 2027.
The important idea is not that data centers produce “free” heat — the recovery system requires infrastructure and energy — but that waste heat which would otherwise be rejected can be put to productive use.
As AI increases demand for high-powered computing facilities, this approach could become increasingly valuable in colder regions where district heating is already widespread.
Instead of treating data-center heat as waste, Finland is showing how it could become another part of the energy system.
#Microsoft #DataCenters #WasteHeat #Finland #EnergyInnovation
Tomi Engdahl says:
The transition toward closed-loop cooling systems in Europe and the United States reflects a shift away from traditional open evaporative towers, driven by regulatory demands, extreme rack heat densities from AI workloads, and public pushback over freshwater consumption.
Unlike open evaporative systems—which continuously consume and evaporate millions of gallons of municipal water—closed-loop systems recirculate fluid through sealed pipe networks with zero continuous water consumption.
By deploying closed-loop architecture, hyperscalers and colocation operators across North America and Europe are effectively trading higher electrical power consumption from local energy grids in exchange for eliminating their operational impact on local water supplies.
Tomi Engdahl says:
Protecting water resources is an undeniable priority, but emotional appeals often obscure the actual mechanics of water consumption. When individual behaviors are policed while massive industrial drains go unregulated, public policy becomes performative rather than effective.
If the goal is genuine water preservation, policy must focus on where water actually goes. Broad environmental rhetoric frequently directs public outrage toward consumer habits or emerging technology while ignoring the largest industrial users.
Treating water as a critical resource requires moving past symbolic consumer bans, data center rage and addressing structural inefficiency.
Protecting life-sustaining resources requires targeting the vast majority of water loss—starting with outdated agricultural infrastructure and legal frameworks—rather than relying on consumer guilt.
Tomi Engdahl says:
Meta’s flagship European node is its massive campus in Luleå, northern Sweden, located just south of the Arctic Circle. Commissioned in 2013 and expanded across three mega-halls, the Luleå facility serves as a blueprint for Nordic data center design, relying almost entirely on local sub-arctic air and hydro-generated grid power.
Tomi Engdahl says:
Meta’s flagship European node is its massive campus in Luleå, northern Sweden, located just south of the Arctic Circle. Commissioned in 2013 and expanded across three mega-halls, the Luleå facility serves as a blueprint for Nordic data center design, relying almost entirely on local sub-arctic air and hydro-generated grid power.
Placing a massive 100+\text{ MW} electrical load directly adjacent to major regional hydroelectric dams creates a clean power sink, minimizing transmission losses across long-distance high-voltage lines.
The campus operates without energy-intensive compressor chillers or continuous open evaporative towers. Direct freshwater draw for cooling is virtually non-existent during normal operation.
Like similar installations across the Nordic region, the sub-arctic location makes waste heat a valuable secondary resource:
Office & Space Heating: Exhaust air from the servers is captured and routed directly into heating systems for the campus’s administrative buildings and support infrastructure.
Regional Integration: Meta works with local Swedish energy partners to evaluate expanding low-grade exhaust heat recovery into nearby municipal forestry, greenhouse agriculture, and local district heating lines.
Tomi Engdahl says:
Europeans build data centers that do not consume water and use “clean energy” to old industrial sites originally built for tree based industrial operations (timber,paper,cardboard).
Less water used, less trees cut and less damage to nature than earlier industry at this location.
Tomi Engdahl says:
https://www.facebook.com/share/p/18fPyUAeRM/
Over 5,000 gallons of diesel spilled from a New Jersey data center into a local creek, reigniting calls for a statewide pause on new data-center construction.
On September 11, 2026, an estimated 5,000 gallons of diesel were released from a storage tank at Equinix’s data-center facility on Hartz Way in Secaucus, New Jersey. Local reports put the total closer to 5,500 gallons. The fuel entered Anderson Creek, a tributary connected to the Hackensack River and part of the Meadowlands ecosystem.
Local officials and environmental representatives said they were told the incident involved a computer or software problem that continued calling for fuel even though the tank was full. However, the New Jersey Department of Environmental Protection said it has not yet determined the specific cause of the discharge. Equinix said it is investigating the root cause.
Cleanup crews installed protective booms and began removing contaminated water and fuel from the creek. NJDEP and the U.S. Coast Guard reported that they had not observed impacts to the Hackensack River or wildlife, and there are no drinking-water intakes in the area. Environmental advocates have cautioned that it is still too early to know whether there will be longer-term ecological effects.
The spill has also intensified New Jersey’s growing debate over data-center development. Ben Dziobek of the Climate Revolution Action Network renewed the group’s call for a statewide moratorium on new data centers, arguing that environmental risks can extend beyond individual municipal boundaries.
The incident does not by itself establish that data centers inevitably cause environmental disasters.
But it does highlight something often overlooked in the AI boom: behind the servers are massive physical facilities, backup generators, fuel systems, electricity infrastructure and other equipment that can create environmental risks when something goes wrong.
As AI infrastructure expands, communities are increasingly asking how these facilities can be built while protecting the land, water and people around them.
Tomi Engdahl says:
https://www.facebook.com/share/p/1EtMd8tTMU/
U.S. data centers consumed an estimated 17 billion gallons of water directly for cooling in 2023, according to Lawrence Berkeley National Laboratory. That is a modeled national estimate rather than a facility-by-facility measurement.
The number becomes even larger when indirect water use is considered: electricity generation associated with data centers consumed an estimated 211 billion additional gallons.
As AI workloads expand, cooling is becoming a major infrastructure issue. The impact is especially important in water-stressed regions, where a single large facility can place substantial pressure on local supplies.
#ArtificialIntelligence #DataCenters #WaterCrisis #TechInfrastructure #Sustainability
Tomi Engdahl says:
https://idahonews.com/news/local/meta-opens-new-ai-data-center-in-kuna-hopes-to-become-water-positive
Tomi Engdahl says:
https://www.facebook.com/share/p/1dH4qmi28Q/
Dublin has become one of the world’s clearest examples of how rapidly data centers can reshape a local electricity system. Carbon Brief has reported that data centers account for around 79% of electricity demand in Dublin, based on analysis from the Öko-Institut.
The figure is striking because data centers operate continuously, creating a concentrated demand for electricity in an area that also has to serve homes, businesses and other industries.
Ireland’s unusually high concentration of data centers has already created challenges for grid planning. The country’s electricity system has needed additional generation, transmission capacity and measures to manage connections for large new users.
However, the claim that Irish families are paying 40% more specifically because of data centers should be treated cautiously. Electricity prices are affected by wholesale fuel costs, network charges, taxes, supply conditions and other factors, so a single cause cannot explain household prices.
Dublin’s experience nevertheless highlights a growing global question: how should rapidly expanding AI and cloud infrastructure share limited electricity capacity with the communities around it?
#Dublin #DataCenters #AIInfrastructure #ElectricityDemand #Energy
Tomi Engdahl says:
https://www.facebook.com/share/p/1Dkkgxz7DY/
Hidden water footprint of AI data centers is becoming a growing concern as technology companies expand infrastructure in water-stressed regions.
While companies often report the water used directly for cooling servers, researchers point out that the electricity powering data centers can create an additional water demand. Power plants that generate electricity may use significant amounts of water, depending on the energy source and cooling systems involved.
A Lawrence Berkeley National Laboratory report estimated that U.S. data centers directly consumed billions of gallons of water in 2023, while the electricity supporting them was linked to a much larger indirect water footprint.
The impact varies based on how electricity is produced. Renewable sources such as wind and solar generally require less operational water use, while some traditional power sources can have higher water demands.
As AI infrastructure grows in areas like Texas and New Mexico, communities are increasingly discussing how to balance technological expansion with water availability, energy needs and local resource management.
#fblifestyle #Polarbear #AI #Environment #Technology
Tomi Engdahl says:
https://www.facebook.com/share/1PJfDpg7Da/
Referring to the US as a hypothetical “laggard state” behind China’s achievement of AGI, Jacob Stokes — now the deputy director of a prominent DC think tank — writes that the country could “try to sabotage the leading state’s AGI systems, either through physical infiltration or offensive cyber operations; that is, cyberattack.”
“Cyberattacks to sabotage AGI might not provoke large-scale retaliation,” Stokes considers, given that “cyberattacks are already to some extent considered normal state behavior below [the] threshold of war.”
But waging an unprecedented cyberwar is just one of many scenarios he outlines. As Stokes observes, AI doesn’t just live online; it also requires immense computing power from physical facilities to run.
“The final option,” therefore, “would be the most dangerous and carry the most escalation risk: kinetic attacks, meaning destroying things with missiles and bombs.”
https://trib.al/HjaOn7c
Tomi Engdahl says:
https://www.facebook.com/share/p/1cDwpv6mWa/
Texas asked. Most filed nothing.
In 2024 one Google facility in Texas got through ten thousand gallons of water. Another, in Iowa, got through a billion. Same firm, same year. The Texas site cools itself with air; the Iowa site does not.
Across the country these facilities take roughly 449 million gallons daily, which works out at some 0.14 percent of everything America uses. Golf courses drink more than four times as much. The bigger figure sits upstream: the power stations feeding those servers were estimated to have used, in 2023, some 211 billion gallons, twelve times what the cooling itself required.
Where it hurts is regional. Something like 40 percent of these facilities stand in places already short of water, Arizona included, and over fifty municipalities have now frozen new construction. When Texas checked, 83 percent of the state’s 341 data centers had filed nothing at all.
Sources: Lawrence Berkeley National Laboratory, Google Environmental Report 2026, EESI, University of Georgia
Tomi Engdahl says:
https://www.facebook.com/share/p/1Cc44z9XTD/
The rapid expansion of AI infrastructure is creating a difficult sustainability challenge for the world’s largest technology companies. Recent environmental reports from Microsoft, Google and Amazon show that electricity consumption and overall emissions have risen substantially as their data-center footprints expand.
Amazon reported that its greenhouse-gas emissions increased by more than 16% in 2025, with electricity use rising 34%. Google also reported record environmental impacts as its AI infrastructure expanded, while Microsoft’s electricity-related emissions increased sharply.
The comparison with national emissions illustrates the scale of these corporations, but exact totals depend on what categories of emissions are included and how companies account for purchased electricity and renewable-energy contracts.
Buying renewable-energy certificates or signing power-purchase agreements can support clean-energy development, but these mechanisms do not necessarily mean every data center is physically running on renewable electricity every hour.
That distinction is becoming increasingly important as AI demand grows. The technology companies are simultaneously investing in cleaner power while consuming dramatically more electricity.
The central challenge is whether efficiency improvements can keep pace with the sheer scale of AI expansion.
#AI #DataCenters #ClimateChange #TechIndustry #CleanEnergy
Tomi Engdahl says:
Right… but those companies provide vast amounts of services we all use and often rely on.
- Google handles ~91% of all web searches worldwide.
- Google Chrome accounts for ~69% of global web browsing.
- Android runs ~68% of the world’s smartphones.
- Gmail has ~3 billion users; Google Maps has 2+ billion monthly users.
- Google has 13 different products with 1+ billion users; five exceed 3 billion.
- Windows runs on more than 1.6 billion active devices each month.
- Microsoft 365 has 450+ million paid commercial seats.
- Amazon AWS alone provides ~28% of global cloud infrastructure; Microsoft Azure ~20% and Google Cloud ~15%.
- Combined, Amazon, Microsoft and Google control ~63% of the global cloud-infrastructure market, the computing backbone behind huge numbers of websites, apps and businesses.
Tomi Engdahl says:
https://www.facebook.com/share/p/1BT7PczXXS/
Every server inside a data center uses electricity.
And much of that electricity eventually becomes heat.
Normally, that heat has to be removed from the facility. But Europe is increasingly looking at it differently: not simply as waste, but as a potential energy resource.
Under the EU’s Energy Efficiency Directive, data centers with a total rated energy input above 1 MW are required to utilize their waste heat or another waste-heat recovery application unless doing so is shown to be technically or economically infeasible. The rules specifically call for consideration of connections to district-heating networks. (EUR-Lex)
That doesn’t mean every data center will automatically pipe hot air or water straight into people’s homes.
The economics and engineering matter.
A nearby district-heating network can make recovery much more practical. In other locations, there may be no suitable heat demand or infrastructure to use the heat efficiently.
Finland is already putting the idea into practice.
Microsoft is working with energy company Fortum on a major project in southern Finland designed to recover heat from its data centers and feed it into district-heating networks serving Espoo, Kauniainen and Kirkkonummi. Microsoft says the system is expected to begin operating in 2027. (Source)
When a data center needs cooling and a community needs heat, the two systems can potentially work together.
Instead of producing heat and simply releasing it into the environment, some of that energy can be captured, upgraded with heat pumps and delivered through existing heating networks.
As AI and cloud computing continue driving demand for data-center infrastructure, the amount of waste heat available will also grow.
The question isn’t simply how much computing power we can build.
It’s what we do with everything that computing produces along the way.
If we’re going to build the infrastructure of the digital future, perhaps we should make sure as little of its useful energy as possible goes to waste.
Tomi Engdahl says:
Scott Jackson computer servers are a core part of a data center.
Cadmium chloride is not an additive deliberately poured into data center cooling loops, but cadmium has emerged as a heavily regulated heavy metal in data center water discharge permits due to the risk of environmental contamination. As hyperscale facilities consume millions of gallons of water daily for evaporative and liquid cooling, the wastewater they cycle out must be tested for toxic elements,
including cadmium, copper, zinc, and residual chlorine before being released back into natural waterways or municipal sewer systems.
In regions like Virginia, some Amazon data centers have been granted permits by the Department of Environmental Quality to discharge hundreds of thousands of gallons of treated wastewater per day directly into natural creeks.
While cadmium is an unwanted byproduct, data center operators do actively treat cooling water with a variety of intense chemical agents to maximize efficiency.
Chemicals like
isothiazolinons
and glutaraldehye are used to sterilize water towers and prevent dangerous bacterial growth like Legionella.
Phosphates and molybdates protect the extensive metal piping but risk causing toxic algae blooms, eutrophication, if leaked into local lakes.
Ethylene glycol or propylene glycol are used in closed loop systems to manage heat transfer but can deplete oxygen in water and suffocate aquatic wildlife if a valve leaks.
Tomi Engdahl says:
how you define most efficient cooling? The options on USA are between most energy efficient huge water consumer to no water needed for running but needs more energy to run (which takes lots of water in many parts in USA).
Tomi Engdahl says:
https://www.facebook.com/share/p/1FgKYnvmiX/
CLEAN WATER IS OLDER THAN ANY TECHNOLOGY WE HAVE EVER BUILT.
And no matter how advanced our technology becomes, communities still depend on it.
AI may feel entirely digital, but the machines making it possible are very physical.
They occupy huge buildings, consume enormous amounts of electricity and generate tremendous amounts of heat. Keeping those servers operating safely can require significant amounts of water, particularly in facilities that rely on evaporative cooling. (Lincoln Institute of Land Policy)
But there isn’t one water footprint for every data center.
Cooling technology makes a major difference.
Some facilities rely heavily on evaporative cooling, while others are turning to closed-loop systems, direct liquid cooling, immersion cooling, reclaimed water or even waterless cooling designs.
Where a facility is built matters just as much.
A data center drawing water from a municipal system, river, reservoir or aquifer creates its greatest direct pressure in the community and watershed where that water is being withdrawn — not somewhere else where someone happens to use an AI service.
That’s why communities considering major data-center projects have to look beyond the building itself.
How much water will it consume?
Where will that water come from?
What happens during drought?
How much wastewater will be produced?
Who pays for upgrades to water infrastructure?
And what happens to the resource over the long term?
These questions can be particularly important for Indigenous Nations and communities whose relationships with water extend beyond its role as a utility. Water can be connected to food systems, fisheries, cultural practices, spiritual traditions and legally protected water rights. Indigenous water rights and cultural interests are already recognized as important considerations in federal resource planning. (Legal Information Institute)
The numbers can also become enormous.
The Lincoln Institute of Land Policy reports that some of the largest data centers can require millions of gallons of water per day, although actual consumption varies substantially by facility and cooling design. (Lincoln Institute of Land Policy)
Cornell researchers have estimated that continued growth in AI computing could result in hundreds of billions of gallons of additional annual water consumption in the United States by 2030, while finding that better siting, cooling efficiency and other measures could substantially reduce that impact. (Cornell Chronicle)
That doesn’t mean data centers shouldn’t exist.
It means their environmental footprint needs to be understood before communities commit to them.
The AI revolution is often described as something happening “in the cloud.”
But there is nothing weightless about the infrastructure behind it.
It needs land.
It needs electricity.
It needs cooling.
And in some cases, it needs millions of gallons of water.
Tomi Engdahl says:
https://www.facebook.com/share/p/1C8VUEXcFH/
A major energy bill just passed the U.S. House with rare bipartisan support, and it could shape how communities pay for the power demands of AI.
On September 16, the House voted 417 to 3 to pass the Ratepayer Protection Act, a bill aimed at preventing households and small businesses from being charged for new grid infrastructure built mainly to serve large data centers.
AI and data facilities can require enormous amounts of electricity, sometimes pushing utilities to build new power generation, transmission lines, and other upgrades. The concern is that those costs could be spread across existing customers through higher utility rates.
The bill would require state utility regulators and unregulated utilities to consider standards for “large-load” customers, including data centers with demand of 100 megawatts or more. Those standards would be designed to make large customers cover the full incremental cost of generation, transmission, or distribution upgrades needed to serve them.
Supporters say the measure keeps electricity costs tied to the companies creating the demand while leaving decisions largely with states. Critics say the bill may be limited because it only requires states to consider the standards, not adopt them. Some states are already developing large-load tariffs and other rules for data centers.
For Native Nations, rural communities, and conservation-focused communities, the growth of energy-intensive infrastructure raises practical questions about land use, water, wildlife, cultural sites, consultation, local jobs, and who pays for new infrastructure. Data centers are often discussed as technology projects, but their footprint can reach into power systems, watersheds, and communities far from the buildings themselves.
The bill now heads to the Senate, where its future remains uncertain.
How should communities balance new technology infrastructure with fair utility costs, local consultation, and protection of land and water?
#EnergyNews #Conservation #IndigenousCommunities #DataCenters
Sources: Utility Dive, CBS News, House, Executive Gov, Thehill, AI Weekly
Tomi Engdahl says:
Data centers are not defined by the exact chips inside them; they are defined by centralization, physical security, connectivity, and land footprint.
The physical structures will endure because software demand always expands to consume available hardware capacity. Ultra-efficient chips will change how servers are cooled and powered inside the building, but the physical centralized computing campus is here to stay.
Data centers began in the 1940s, evolving alongside computing history through a continuous, pendulum-like shift between centralization and decentralization.
Every wave of computing follows a predictable cycle: a breakthrough creates a resource too massive for local devices (forcing centralization), followed by hardware miniaturization that brings computing power back to the individual (decentralization).
Tomi Engdahl says:
data centers began in the 1940s and have been a temporary solution since then?
Nothing is so permanent in tech than a temporary solution!
Tomi Engdahl says:
European electronics waste (e-waste) and circular economy regulations are among the strictest in the world. Rather than focusing solely on recycling, the European Union’s framework covers the entire product lifecycle—from raw material restrictions and eco-design to repairability, consumer rights, and post-consumer waste management.
Tomi Engdahl says:
The Restriction of Hazardous Substances (RoHS) Directive (2011/65/EU) restricts the use of specific hazardous materials in electrical and electronic equipment sold within the EU.
Tomi Engdahl says:
Used it? How? Are they making clouds? Where is it going?
The older evaporative cooling data centers evaporate the water to steam typucallt in cooling towers. That steam rises up to the sky. In around one week that steam becomes clouds on the sky. The water then rains back down to some other area.
Tomi Engdahl says:
The environment seems to be of little concern now!
this applies especially what seems to happen in USA at the moment according to my news feed.
Because I don’t live in USA and work very close to data center projects there, I can’t be 100% sure is of that picture is entirely right.
The environment seems to be taken much more seriously in Europe. They are building and have built data centers in Finland, which have reduced water use and CO2 exhaust compared to situation before them. Paper mills closed and converted to data centers, use lots of CO2 free energy (solar,wind,hydroelectricity,nuclear) and reuse of waste heat for district heating (much less CO2 producing coal,oil,natural gas,wood for heating).
Tomi Engdahl says:
The headline’s statistics regarding 945 TWh and Japan’s energy equivalent are directly sourced from the IEA. However, framing data centers as a primary driver of global deforestation mixes real grid capacity challenges with sensationalized land-use claims.
This graphic combines accurate energy projections with hyperbolic framing around land and water impacts.
1. The IEA Electricity Projection (~945 TWh)
Status: Factually Accurate.
2. “Water pull roughly doubles before 2030″
Status: Plausible Range, but Context-Dependent.
The Reality: Industry research (such as reports from Rystad Energy and environmental modeling groups) indicates direct water consumption for cooling could double or even triple by 2030 if facilities rely on legacy evaporative towers.
The Shift to Dry Cooling: Because water availability is becoming a hard constraint in regions like Texas, Arizona, and Southern Europe, newer hyperscale facilities are increasingly required to deploy dry-cooling systems or closed-loop liquid-to-chip systems that reduce daily water evaporation to near zero (though at a slight energy penalty).
3. “Forests being cleared now are the first instalment”
Status: Hyperbolic / Misleading.
Land Use Reality: Modern computing campuses are high-density industrial buildings. While site acquisition can lead to localized land clearing or agricultural conversion, data centers consume a tiny fraction of global land compared to agriculture, forestry, or sprawling suburban housing.
Tomi Engdahl says:
it’s not just antifreeze it’s got all kinds of nastiness in it to protect it and yes Direct To Chip Cooling the cooling plates are clamped to target chips the rest power supplies,networking and memory must still be cooled by ambient air .
Then there’s Immersion cooling the servers are fully immersed in dielectric fluid
AI learning clusters are whole different beasts
Massive generators and maintenance
Flushing etc.
Leaks happen and flush and fill (with all that nastiness going into ground water and secondary water sources )
Data center water consumption globally for a year was 1.2 TRILLION gallons of water.
A Google data center in Ohio consumed over 1 billion gallons of water in a year !
And offsite use of water to run datacenters due to the the electricity they pull 12 times greater amount than on site
(Lawrence Berkeley National Laboratory)
Now factor in chipset manufacturing and lifespan ultra purified water and PFCS
WITH limited lifespan and obsolete rates
Look at the dumping controversy .
It’s absolutely insane with over 5000 data centers in the USA that we continue building
Unregulated and under regulated environment destroying monsters constantly consuming and polluting from conception to death!
I’ll pass on Death by Data Center thanks
Tomi Engdahl says:
Sadly there is a lot of mistruths being shared around water consumption. The old way of cooling with evaporation simply doesn’t cut the mustard in 2026.
Chilled water loops get filled once (like a swimming pool) and get reused with corrosion inhibitors. Designed to minimise the use of water from external sources.
Since we’re complaining about chemical damage, check any of the CSB videos on YouTube about chemical plants that did damage through leaks, mismanagement and outright negligence. The damage caused by these far exceeds anything.
Whilst I acknowledge an Equinix DC recently had a diesel spill. They have already committed to the cost of cleaning up and restoring the damaged land as a responsible DC owner.
Tomi Engdahl says:
Europe and China are racing to replace use of fossil fuels with a selection of technologies.
When comparing how Europe, China, and the United States are positioning themselves to replace fossil fuels, the race isn’t just about deploying technology—it is an economic, industrial, and geopolitical competition.
Each region has adopted a distinctly different strategy shaped by its local resources, regulatory environment, and market forces.
Unlike the EU—which relies on binding targets, strict carbon pricing, and heavy regulation—the US approach relies primarily on market dynamics and federal tax credits.
Both Europe and China face severe energy security risks due to a lack of domestic oil and natural gas. China imports over 70% of its oil, and Europe was forced to pivot rapidly away from Russian gas.
By contrast, the US is the world’s largest producer of oil and natural gas. This local abundance acts as a double-edged sword.
Summary
China is winning on sheer scale, supply chain control, and rapid manufacturing deployment.
Europe leads in regulatory frameworks, carbon pricing systems, and energy efficiency.
The USA is leveraging private capital, tax incentives, and technology innovation, but faces headwinds from its vast domestic fossil fuel reserves and grid transmission delays.
Tomi Engdahl says:
AI data centers use lots of electricity but not necessarily lot of water.
The European supercomputer LUMI, one of the fastest and greenest of its kind in the world, operates as a closed-loop system that consumes practically zero water through evaporation.
Instead of drawing millions of gallons of freshwater to evaporate in cooling towers, LUMI uses a closed-loop system coupled with waste-heat recovery.
Powered entirely by 100% renewable energy, with its waste heat providing roughly 20% of the district heating for the city of Kajaani.
Google’s facility in Hamina, Finland, is a foundational hub for the company’s global cloud and AI infrastructure. It also uses practically zero clean water and runs mostly on renevable energy. Heat recovered from the server halls is fed directly into Hamina’s municipal district heating network, supplying roughly 80% of the town’s annual heating needs. The Hamina site also uses a seawater cooling system for heat not used for city heating needs.