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.
1,309 Comments
Tomi Engdahl says:
https://www.facebook.com/share/p/1L4zLpbBmN/
Ten years ago, data centers barely registered on Ireland’s power grid. They used about five percent of the country’s electricity. That number is now twenty-three percent. To put that in perspective, every single home in Ireland—both urban and rural combined—uses twenty-eight percent. A handful of massive tech facilities are now drawing almost identical power to the entire residential population. The surge is being driven directly by global cloud computing and the massive power requirements of artificial intelligence. Ireland has accidentally become the primary case study for what happens when infinite digital scaling meets a finite electrical grid. Regulators are now stepping in. New data centers are required to build their own flexible on-site power generation. They must also source at least eighty percent of their energy from new renewable projects. It is a strict pivot. The grid simply cannot keep up with the servers. Data from the International Energy Agency suggests these facilities could consume a third of Ireland’s total power by late 2026.
Tomi Engdahl says:
The core physical claim that “glycol transfers heat better than water” is false. Pure water is actually the gold standard for heat transfer fluids. Glycol is considerably worse for heat transfer.
Specific Heat Capacity: Pure water has an exceptionally high specific heat capacity (\approx 4.18 kJ/kg K). A standard 50/50 water-glycol mix drops that capacity down to around 3.45 kJ/kg K
Thermal Conductivity: Adding glycol reduces the overall thermal conductivity of the solution and increases fluid viscosity, requiring pumps to work slightly harder to move the same amount of heat.
Compared to pure water, a water-glycol mixture actually penalizes thermal transfer efficiency by roughly 12% to 20%.
Why Do Closed-Loop Systems Use Glycol?
If glycol is thermally inferior to water, why do engineers mix it into closed cooling loops?
They do it for freeze protection and anti-corrosion, not heat absorption.
In commercial closed-loop data center cooling systems, two main types of industrial glycols are used: Propylene Glycol (PG) and Ethylene Glycol (EG). They are mixed with deionized water (typically at a concentration of 20% to 50% glycol) along with specialized inhibitor packages.
Formulated industrial glycol coolants contain chemical inhibitors that prevent scale buildup, pipe corrosion, and biological growth inside the closed plumbing.
Tomi Engdahl says:
“Meanwhile, there has been little discussion about refilling reservoirs or expanding desalination efforts. The topic of drought is being ignored, along with concerns about toxic water associated with data centers in Georgia and Texas.”
There are two distinct but closely linked critical issues in the data center debate: long-term water security (desalination and reservoir expansion) and localized toxic water contamination.
While expanding reservoirs or building coastal desalination plants sounds like an obvious fix for data center water draw, operators avoid these solutions due to financial and geographical limits.
Desalination is only viable near coastlines. However, hyperscale clusters are frequently built inland.
Building a modern industrial desalination plant or dedicated reservoir costs hundreds of millions of dollars and requires a decade of environmental reviews. Desalination produces toxic, hyper-saline brine waste that presents its own major disposal and environmental regulation challenges.
For tech operators, it is far cheaper and faster to redesign the facility to use dry radiators or closed-loop dielectric liquid cooling (0 L/kWh on-site water consumption) than to build water infrastructure.
The concern surrounding “toxic water” and data centers in states like Georgia, Texas, and Wisconsin is real, but it spans two distinct stages of the data center lifecycle.
Much of the recent high-profile outcry in rural Georgia (such as near Meta or QTS construction sites in Morgan and Newton counties) stems from heavy site preparation.
For legacy facilities that still run open evaporative cooling towers, the water discharged isn’t pure H₂O:
Concentrated Chemicals: To prevent algae growth and pipe scaling in humid climates (like Georgia and Texas), operators treat cooling water with anti-scaling chemicals, biocide additives, and corrosion inhibitors.
Cooling Tower “Blowdown”: As water evaporates, dissolved heavy metals, minerals, and chemical additives concentrate in the basin. This wastewater is periodically flushed into local sewers or surface waters.
PFAS & Fire Suppression: Many data centers use fluorinated compounds (PFAS) in specialized chip manufacturing, transformer equipment, and clean-agent fire suppression systems. If poorly managed, runoff or industrial discharge can introduce persistent chemicals into surrounding watersheds.
The debate over water is shifting away from simple “volume consumed” to local environmental protection:
Desalination is rarely the answer because facilities are often inland, making 0\text{-water} closed-loop dry cooling a much cheaper engineering fix.
Toxic water concerns are real, but they are primarily driven by construction-stage site disturbance/aquifer disruption and chemical blowdown wastewater from legacy evaporative cooling towers.
Tomi Engdahl says:
Robert Beliveau
“The environmental effects of data centers—including their potential impact on cattle ranchers—also need to be investigated. Ranchers have raised concerns that noise pollution from data centers may be affecting cattle health, including weight loss and stillbirths.”
The claim that ranchers are raising concerns about data center noise affecting cattle health seems to be grounded in real, ongoing agricultural disputes.
Formal university-led veterinary studies specifically targeting data center low-frequency noise seem to be still a work in progress. It seems that continuous low-frequency acoustic stress can induce elevated cortisol in livestock, which directly contributes to weight loss, behavioral disruption, and reproductive failure.
Inadequate Decibel Metrics (dBA vs. dBC): Most municipal noise ordinances use the A-weighted decibel scale (dBA), which filters out low-frequency sound because human ears are less sensitive to low pitches. However, the heavy bass hum of a data center chiller yard is better measured on the C-weighted scale (dBC) or low-frequency spectrum. A site can easily pass local dBA noise limits while still emitting a disruptive, low-frequency vibration across nearby pastures.
Tomi Engdahl says:
Academic researchers and university engineering departments have studied—and continue to study—data center thermal management and waste heat recovery extensively.
The issue is not a lack of research. Thousands of peer-reviewed mechanical, electrical, and thermal engineering papers are published every year on this exact problem. The underlying reason data center heat remains a challenge comes down to fundamental physics, economics, and real-world deployment barriers.
What University Researchers Have Already Solved
Academic institutions (like MIT, Georgia Tech, ETH Zürich, Stanford, and Rice University) have partnered with industry to invent many of the cooling technologies used today:
Direct-to-Chip Liquid Cooling: Researchers pioneered routing non-conductive dielectric liquids or pure water through micro-channel cold plates bolted directly onto CPUs and GPUs.
Two-Phase Immersion Cooling: University labs helped develop dielectric fluids that boil at low temperatures (30^\circ\text{C}\text{–}50^\circ\text{C}) to remove chip heat via vaporization.
Thermoelectric & ORC Power Generation: Engineering groups (such as recent research at Rice University) have developed Organic Rankine Cycles (ORC) boosted by solar thermal collectors to turn low-grade server exhaust heat back into usable electricity.
District Heating Coupling: European universities (such as KTH in Sweden and LUT in Finland) have spent over a decade modeling 4th and 5th Generation District Heating (5GDH) loops to pipe warm server water directly into residential heating networks.
Tomi Engdahl says:
University professors and graduate researchers have spent decades analyzing and improving data center thermal efficiency. Their work is the primary reason modern facilities operate with a Power Usage Effectiveness (PUE) near \approx 1.1\text{–}1.2—meaning only 10\%\text{–}20\% of facility energy goes to cooling overhead.
The remaining issue is not a lack of academic research, but the physical reality that reclaiming huge volumes of lukewarm (40 °C) heat is thermodynamically difficult and expensive to monetize without adjacent municipal district infrastructure.
Tomi Engdahl says:
Tom Villeneuve Myth 1: Data centers are completely automated and don’t create local jobs.
The Myth: People often assume that once a massive data center is built, it runs entirely on its own with a handful of remote engineers, offering little to no economic benefit to the host community. Myth 2: Data centers deplete local water supplies and leave communities dry.
The Myth: It is frequently stated that all data centers consume massive quantities of municipal drinking water continuously through traditional cooling towers.Myth 3: Data centers automatically make residential electricity bills skyrocket.
The Myth: A common belief is that data centers place an uncompensated burden on the local grid, forcing everyday residents to shoulder higher energy costs. Myth 4: Data centers are dead-weight energy hogs that offer zero grid value.
The Myth: Data centers are viewed solely as massive power sinks that drain energy resources without giving anything back. Myth 5: Data centers are whisper-quiet and invisible neighbors.
The Myth: Because data centers look like standard commercial warehouses, people often assume they have little to no environmental presence outside their walls.
Tomi Engdahl says:
water is used to cool the data centres. They can use open loop water that constantly pumps cold water in and hot water out removing the heat. They usually also use closed loop systems with coolant like in a car, this pumps water around a closed loop. They do need to periodically empty the closed loop and flush the system with cleaning chemicals to stop salt/ mineral build up.
The water volume is not an issue, it’s way over-hyped in the media because AI is the new technology. The chemicals are a global issue, although data centres are a small part of this issue.
Tomi Engdahl says:
So how many gallons of water do they use a day? Im curious because toilets alone use 3.5-4.1 billion gallons of water just here in America per day. Thats not including sinks, showers, washing machines and dishwashers. You can add another 15 billion gallons onto that. Agriculture uses the most at 118 Billion gallons of water a day. Again this is all just in America. With that information, Data Centers dont even use 1 billion gallons of water per day.
Stop stressing because you cant comprehend numbers. Food is important, housing is important, clean water in important. But we cant get that without education, and right now I guarantee you cant find a human smarter than AI. Welcome to the future!!
Tomi Engdahl says:
there are other ways to cool the data centers. They are already found. You can save 98-99% or more water compared to legacy coolin designs.
And those are already in use in most new data centers – especially outside USA.
The data centers less than 1% of water consuming problem has been solved by engineers.
Now put the effort to how the rest over 99% of water is used, and is there is possible to reduce it.
Tomi Engdahl says:
“there’s fresh water shortages happening everywhere on the planet.”
Your claim is partially false. Not everywhere but in quite many places.
The claim reflects a fundamental confusion between the physics of water as a resource and the logistics of local access.
“Nope you’ll never convince me that we have enough water…not when it’s a non renewable resource”
Whether water is considered “renewable” comes down to a distinction between Earth’s total, closed system and regional human consumption.
From a physical and chemical standpoint, water on Earth is a closed, continuously renewed system driven by solar energy.
The reason people feel water is “non-renewable” is that human society often withdraws fresh water faster than the natural cycle can replenish it in specific locations.
The world is not “running out” of water molecules; it is facing a distribution, pollution, and management crisis.
Humanity relies heavily on tiny fraction (1.2%) of surface water and shallow groundwater.
When pollution, urban expansion, or drought contaminates or strains local rivers, the local supply of usable fresh water drops—even though the global water molecule count remains unchanged.
Tomi Engdahl says:
Hot Commodity
Data Centers Causing Huge Temperature Spikes for Miles Around Them, Study Suggests
Not very cool.
https://futurism.com/artificial-intelligence/data-centers-temperature-spikes?fbclid=IwVERDUAVEHnhleHRuA2FlbQIxMABwZG9mBXNydGMGYXBwX2lkDDM1MDY4NTUzMTcyOAABHlu3mzvtvnm3Xl22kRU-ZqLa1KtVZN4UeqMAqIyKHHqLyIUCwTCeX16XZxRR_aem_D19-PBZ6YGbtRBA1ZRovZA
Tomi Engdahl says:
US restarts three shut-down nuclear reactors to feed AI data centers — 2.25 GW coming online with backing from Microsoft and Google
https://www.tomshardware.com/tech-industry/big-tech/us-takes-unprecedented-step-of-recommissioning-nuclear-plants-due-to-ai-power-demand-2-25-gw-currently-on-tap-from-refurbished-reactors?fbclid=IwdGRjcAVEtb9jbGNrBUS1nmV4dG4DYWVtAjExAHBkb2YFc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEefA1GdQBE0UwhZ7TSm7k-LmpznO25G1juA6v2vUsTnAqAYK2ZvnV-HBHuxJA_aem_8l_KLG0eQz1DOg9xkNUwsQ
The current and expected power draw of AI data centers has been the subject of strong discussion, but today, there appears to be some good news. Three U.S. nuclear reactors are being restarted after many years offline, expected to bring 2.25 GW total to the grid. According to a report, the Palisades Nuclear Plant in Michigan, the Crane Clean Energy Center on Three Mile Island (Pennsylvania), and the Duane Arnold Energy Center in Iowa are getting a fresh set of proverbial tires to get rolling again.
Tomi Engdahl says:
If you stand outside a steel mill, foundry, auto plant, or coal facility, you are exposed to direct smokestack emissions, chemical odors, heavy metals, or toxic dust.
If you stand outside a data center, the local air is clean—there are no chemical emissions or smokestack pollutants – unless the data center is powered on it’s own power station or running on backup generators (or leaking something it is not supposed to). Physically, a data center is essentially a sterile, air-conditioned warehouse filled with computer racks. It has no smokestacks, no chemical manufacturing processes, and no toxic industrial run-off during normal operation.
However, its environmental footprint is pushed upstream: it acts like a massive “energy vacuum,” causing power plants (sometimes coal or gas plants) elsewhere to run continuously to supply its power grid.
Coal plants are the single largest point sources of industrial air pollution and greenhouse gases globally. Steelmaking is responsible for roughly 7–9% of total global direct CO_2 emissions release carcinogenic compounds (benzene, toluene), toxic dust, and heavy metals directly into surrounding communities. Metal foundries release severe localized chemical and particulate hazards. Modern car plants use substantial natural gas for curing ovens and paint booths. The biggest local environmental concern is VOCs (Volatile Organic Compounds) and solvent vapors from paint shops, along with toxic chemical wastewater from body-washing and anti-corrosion pre-treatment.
Tomi Engdahl says:
https://www.facebook.com/share/p/19y5b6ZaDL/
AI DATA CENTERS ARE EXPANDING, BUT WHAT IS THE ENVIRONMENTAL PRICE?
Artificial intelligence may operate in the digital world, but the infrastructure powering it has a very real impact on the environment. AI data centers need vast amounts of electricity, land, and water to keep their systems running. Some large facilities can use up to 5 million gallons of water per day, raising concerns in communities already dealing with water shortages.
Environmental advocate Erin Brockovich has drawn attention to concerns shared by people living near these expanding facilities. Residents have reported issues ranging from disappearing wildlife and changes in water conditions to reduced water pressure, constant noise, and higher electricity bills. These concerns have fueled calls for greater transparency and stronger public scrutiny before major data center projects receive approval.
As AI infrastructure spreads across the United States, more communities are questioning how these developments could affect their local environment and essential resources. Brockovich has also launched a public reporting initiative to document concerns from residents about data centers and their potential effects on water supplies, energy use, and surrounding communities.
AI may be shaping the next era of technology, but how do we ensure its growth doesn’t come at the expense of our water, energy, and environment?
Tomi Engdahl says:
https://www.facebook.com/share/p/1EwSzuMUtz/
The AI boom is driving data center developers to secure their own energy sources instead of relying on the electric grid, often through new gas-powered plants.
More than half of these proposed gas projects plan to use older and dirtier technology than typical gas plants. These include repurposed jet engines and generators historically used only for backup power.
Now, developers want to run these equipment for thousands of hours each year, increasing greenhouses gases and air pollution.
Tomi Engdahl says:
“AI pollutes 10x more than heavy industry”
The claim is factually false and off by several orders of magnitude.
Steel Mills & Foundries: According to the World Steel Association and the International Energy Agency (IEA), steel production alone accounts for ~7% to 9% of all global CO_2 emissions.
Coal-Fired Electricity: Coal power is responsible for roughly 30% to 35% of total global carbon emissions.
AI & Data Centers: Entire global data centers (including cloud computing, enterprise servers, and AI) account for roughly 1% to 2% of global electricity demand and under 1% of global greenhouse gas emissions. AI specifically makes up about 15% to 20% of that data center energy footprint today.
Tomi Engdahl says:
Where the Challenge Lies for AI:
Intermittency vs. Continuous Load: Solar generates power when the sun shines; wind generates power when the wind blows. However, an AI data center runs at a near 100% load 24 hours a day, 365 days a year.
The Storage Gap: Because grid-scale battery storage is not yet cheap or large enough to power massive gigawatt-scale computing centers through multi-day lulls, tech firms are increasingly turning to nuclear power (restarting reactors or funding Small Modular Reactors) and natural gas alongside renewables to ensure uninterrupted, zero-carbon baseload power.
Tomi Engdahl says:
https://www.facebook.com/share/p/19gdsjFGBN/
Endeavour and Trevi Systems announced a partnership on October 22, 2026, to use waste heat from AI data centers to produce clean water.
The companies aim to produce up to 200,000 gallons (757,000 litres) of water per day for every megawatt of computing capacity by treating brackish groundwater, which contains more salt than freshwater.
The system will combine Endeavour’s cooling technology with Trevi’s heat-driven desalination process, using heat at around 60 to 80°C to separate water from salts.
At the announced target, a 100-megawatt facility could produce up to 20 million gallons per day. For oilfield wastewater, the target is about 50,000 gallons per megawatt per day.
The combined system has not yet demonstrated these output figures at an Endeavour data center. Actual production will depend on the installation and the water being treated.
Source: Endeavour, Trevi Systems and Data Center Dynamics, October 2026.
Tomi Engdahl says:
A Data Center is nothing more then a building to house computers. These computer are not limited with AI. When you order something cool from Amazon, your request is processed in a Data Center. Data centers can or do have their own power supplies. Those that do not have their own power can be fitted with one. BUT this has nothing to do with the amount of power available to the Data Center. Nor does it have anything to do with the amount of water needed, a closed system by the way, or any of the other weird claims associated with them.
Tomi Engdahl says:
https://www.facebook.com/share/p/18mKT8sw8Z/
The heat coming off AI computers could help produce fresh water.
Endeavour and Trevi Systems have announced a partnership to use data-center waste heat for desalination. The companies say the system could produce up to 200,000 US gallons of clean water per day from brackish groundwater for every megawatt of computing capacity.
The process uses forward osmosis. A specialized solution draws water through a membrane, separating it from salts and contaminants. Heat at around 60–80°C then helps regenerate that solution so it can be reused.
The aim is to give heat that would otherwise be discarded a useful second job. The companies also plan to concentrate the remaining brine into solids, reducing liquid waste.
The headline figure is a company projection for the proposed integration. Endeavour plans the first deployments at its own US data centers, where performance at that scale still needs to be demonstrated.
Trevi has already tested its technology in other water-treatment projects. Connecting it to computing infrastructure is the next step.
Could producing fresh water become a valuable second purpose for AI data centers?
Tomi Engdahl says:
https://www.facebook.com/share/p/1E4oD8yjWD/
AI NEEDS DATA CENTERS. BUT HOW MUCH WATER WILL THEY CONSUME?
The artificial intelligence boom is transforming our world. Behind every powerful AI model, cloud service, and digital platform are massive data centers packed with computer servers that generate enormous amounts of heat.
Keeping those machines cool can require something far more essential than computing power: water.
Many data centers use evaporative cooling systems, which remove heat but consume water as it evaporates. And the water footprint doesn’t end there. Generating the electricity that powers these facilities can also require water.
The actual amount varies depending on the cooling technology, local climate, energy source, and size of the facility. Not every data center has the same water demands, but the issue deserves serious attention as AI infrastructure expands.
The concern is especially important in drought-prone regions, where communities are already struggling with limited water supplies.
Families need water to drink. Farmers need it to grow food. Rivers, wetlands, and wildlife depend on it to survive.
When large technology facilities draw from the same limited resources, communities must ask an important question: How do we make technological progress without putting essential water supplies under greater pressure?
There are ways to reduce the impact.
Companies can use recycled wastewater, adopt more water-efficient cooling technologies, carefully consider where new facilities are built, and disclose how much water their operations consume.
Some alternatives can increase electricity demand, so the goal should be to reduce the overall environmental impact—not simply shift the problem from water to energy.
AI has enormous potential to advance medicine, science, education, and countless other fields. But its infrastructure also carries environmental responsibilities.
Because artificial intelligence may be the technology of the future, but clean water is a necessity of life today.
Progress should benefit humanity without putting the natural resources we all depend on at risk.
Tomi Engdahl says:
Suomi on kyllä pitkällä datakeskusten hukkalämmön hyödyntämisessä, mutta hukkalämpöä ei muuteta sähköksi, vaan sitä käytetään suoraan kaukolämmössä. Hukkalämpö menee kaukolämpöön.
Suomi on kyllä datakeskusten hukkalämmön talteenoton ja hyödyntämisen kärkimaailmaa, mutta kyse on lämmityksestä eikä sähköntuotannosta.
Sähkön tuottamista hukkalämmöstö on tutkittu Suomessa ammattikorkeakouluissa ja yliopistoissa (kuten Lappeenrannan LUT-yliopistossa ja Vaasan yliopistossa), mutta kokeilut ja laskelmat ovat vahvistaneet, ettei se ole käytännössä kannattavaa. Ennen kaikkea sitä ei käytetä missään kaupallisessa mittakaavassa.
Syy siihen, miksi tutkimuksista huolimatta tälle tielle ei ole lähdetty käytännössä, liittyy suoraan fysiikkaan ja talouteen.
Akatemiassa ja teollisuudessa on tarkasteltu erikoisprosesseja, kuten ORC-prosessia (Organic Rankine Cycle), jossa sähköä pyritään tuottamaan matalassa lämpötilassa kiehuvien orgaanisten nesteiden ja turbiinin avulla.
Tulokset ja fysiikan rajoitteet matalalämpöiselle (30–60 °C) hukkalämmölle olivat kuitenkin selkeät: Erittäin huono hyötysuhde (Carnot-hyötysuhde): 40-asteisesta hukkalämmöstä saatava sähköntuotannon hyötysuhde jää laitteistoineen helposti alle 5–8 prosentin.
Ainoastaan muutaman prosentin hyötysuhteen saavuttaminen vaatisi valtavan kokoisia laitteistoja, lämmönvaihtimia ja turbiineja. Laitteiston hankinta- ja ylläpitokustannukset olisivat moninkertaiset verrattuna siihen pikkuruisuuteen sähköä, mitä sillä saataisiin aikaan.
Tomi Engdahl says:
Paul Pimental Google’s largest AI data center in Europe: Google’s data center in Hamina, Finland, uses virtually zero fresh/potable water for cooling.
Hamina facility’s annual freshwater consumption is essentially negligible—around 0.1 to 0.4 million gallons per year (~380 to 1,500 m^3/year)—which is strictly used for general facility maintenance, restrooms, and drinking water. The Google Hamina facility’s freshwater usage is equivalent to a typical office building with about 50 to 150 on-site staff.
A single medium-sized residential apartment building in Hamina consumes up to double the municipal freshwater used annually by Google’s entire data center complex.
Tomi Engdahl says:
https://www.facebook.com/share/p/19uwgsjTtf/
Artificial intelligence systems, powered by vast networks of data centers, have developed a substantial environmental footprint that matches striking comparisons to everyday human activities and major urban centers.
Research estimates indicate that in 2025 the electricity required to train and operate these AI models could generate between roughly 33 and 80 million metric tons of carbon dioxide emissions.
The upper end of this range exceeds the annual greenhouse gas output of New York City, which stands near 50 million tons.
This carbon burden arises primarily from the intensive computational demands of large language models and related technologies, whose servers draw power equivalent in scale to that of an entire mid-sized nation.
At the same time, the water requirements associated with cooling these facilities and generating the necessary electricity reach an estimated 312 to 765 billion liters per year.
That volume sits in the same range as, and in higher scenarios surpasses, the total amount of bottled water consumed by people worldwide in a single year, approximately 446 billion liters.
Both figures reflect not only direct on-site use but also the indirect water embedded in power production. Significant uncertainty remains because major technology companies provide limited public data specific to AI workloads, making precise measurement difficult.
Nevertheless, the available evidence underscores how rapidly expanding AI infrastructure is placing measurable pressure on global energy systems and freshwater resources, highlighting the need for greater transparency and efficiency improvements as the technology continues its swift growth across industries and daily life.
Source: de Vries-Gao, A. The carbon and water footprints of data centers and what this could mean for artificial intelligence. “Patterns”.
Tomi Engdahl says:
https://www.facebook.com/share/p/1A9WwGj8jQ/
AI data centers generate substantial low-grade waste heat from intensive computing operations.
A new partnership between data center developer Endeavour and desalination specialist Trevi Systems aims to capture that heat through Endeavour’s ThermalWorks closed-loop, waterless cooling platform and use it to drive Trevi’s forward-osmosis process.
In this system, a specialized draw solution pulls clean water through a membrane from brackish groundwater or oilfield-produced water; the draw solution is then regenerated using 60–80°C waste heat, concentrating salts into solids rather than discharging brine.
The companies project production of up to 200,000 gallons of fresh water per day for every megawatt of computing capacity from brackish sources, or about 50,000 gallons per megawatt from produced water.
A 100-megawatt campus could therefore function as a 20-million-gallon-per-day water plant powered almost entirely by heat already generated by the servers.
Initial deployments are planned at Endeavour’s U.S. facilities, with potential to transform data centers from net water consumers into significant producers while addressing brine disposal challenges associated with conventional desalination.
Source: Endeavour and Trevi Systems press release, PR Newswire, October 22, 2026.
Tomi Engdahl says:
https://www.facebook.com/share/p/1DpFXN5vQz/
Groundwater loss is becoming one of the most important water stories in the United States.
A new study published in PNAS used NASA’s GRACE and GRACE Follow-On satellite missions to track changes in terrestrial water storage from April 2002 to November 2025. These satellites measure tiny shifts in Earth’s gravity, which can reveal where water is being gained or lost across groundwater, soil moisture, rivers, lakes, snow, and ice.
The findings show that groundwater depletion is especially serious in parts of the Southwest, including agricultural regions such as California’s Central Valley. In these areas, pumping for irrigation can exceed the rate at which aquifers naturally refill.
Aquifers store water in pores and cracks within rock and sediment. When water tables fall, farmers may need deeper and more expensive wells, making irrigation harder to sustain. Researchers warn that continued depletion threatens the long-term future of farming, rural economies, and food production tied to these regions.
For Indigenous Nations and communities across the Southwest, water is also connected to homelands, food systems, cultural practices, and future planning. Groundwater decline can affect not only farms and towns, but also the broader relationships people maintain with land, water, and seasonal cycles.
The study identified 40 regions worldwide where water storage changes were significantly associated with human activities such as pumping, irrigation, and reservoir construction. Related global research has also found that some aquifers have recovered or improved after changes in policy or water management, showing that groundwater trends are not the same everywhere.
The findings do not mean every U.S. aquifer is shrinking or that any one supply has a known date when it will run dry. They do show where losses are concentrated and add detail to a long-standing warning: when water is pumped faster than it returns, the communities and economies depending on it face growing risk.
How is groundwater being discussed or managed in your region?
#WaterIsLife #IndigenousNews #Conservation #Groundwater
Sources: The Boston Globe, Headtopics, 1Weather
Tomi Engdahl says:
https://www.facebook.com/share/p/1FaAZdeBWA/
AI may feel digital, but the infrastructure powering it depends on enormous physical resources, including electricity, land and, in some cases, water.
Many data centers use evaporative cooling systems that consume water as heat is removed from powerful computer equipment. Their water footprint can also extend to the electricity supply, since some power plants require substantial amounts of water during generation.
The amount varies significantly depending on the facility, cooling technology, climate and energy source. Some newer data centers are adopting air cooling, closed-loop systems and other approaches designed to reduce water use.
As demand for AI computing continues to grow, the question is becoming harder to ignore: how can the industry expand while protecting the clean water that communities already depend on?
#fblifestyle #Polarbear #AI #DataCenters #Water
Tomi Engdahl says:
https://www.facebook.com/share/p/19wSggFvnG/
What if the heat from AI could make clean water instead of waste?
Two US companies, Endeavour and Trevi Systems, say data center heat can power desalination. They claim up to 200,000 gallons of fresh water a day for every megawatt of computing. A 100 megawatt campus could reach 20 million gallons a day.
This matters because data centers are blamed for using too much water.
Researchers at the University of Texas at Austin found desalination can meet data center water needs with a fraction of their energy. These numbers are company claims and need proof. Imagine a thirsty village getting clean water from wasted heat.
Could waste become our next water source?
Sources: Data Center Dynamics, Interesting Engineering, Anthropocene Magazine on the University of Texas at Austin study in Water Research
Tomi Engdahl says:
“It sounds like an airplane engine but it’s continuous, it’s incredibly loud.” https://trib.al/GOzzzWI
Tomi Engdahl says:
Learn more about the many ways data centers are affecting communities: https://www.wri.org/insights/us-data-center-growth-impacts
Tomi Engdahl says:
There is a stark contrast in cooling technology adoption between Europe (where I am) and the United States, driven by climate differences, municipal water pricing, and regulatory frameworks.
In Europe Closed-Loop / Dry-Cooling Adoption is High (~70% – 85%+).
In USA Evaporative / Open Cooling Towers is mainstream Dominant (~60% – 70%+).
Globally, only an estimated 15% to 25% of existing data centers operate on fully closed-loop, zero-evaporation heat rejection systems. The remaining 75%+ rely on evaporative cooling or open cooling towers for at least part of the year, where water is permanently consumed.
Tomi Engdahl says:
I am not concerned at the water at all where I live because of high natural freshwater reserves (“Land of a Thousand Lakes”) and an exceptionally efficient average data center WUE (~0.07 L/kWh).
Traditional Evaporative Towers in USA typically use 1.50 to 2.50+ L/kWh
Tomi Engdahl says:
When you have shortage of water, in addition to data centers water use, you should also look at other water consumption that does not grow food to you:
1 Year of Exported Alfalfa or 1 Year of Golf Course Turf Irrigation each consume roughly 15 to 30 times more water than direct annual US data center cooling.
1 Year of Exported Almonds uses more water than all US data centers directly consume in several decades.
Tomi Engdahl says:
Anonymous to make some small difference, get rid of data centers that consume lots of water. To make bigger impact, also get rid of many fossil fuel power power station – less water use, less CO2 and less pollution. Maybe get rid of those nuclear power stations with huge cooling tower. Get rid of golf courses.
Think if it makes sense to grow a lot of water consuming (very much more than data centers) almonds and alfa-alfa that you export abroad. Is this right use for your water resources when you say water should be used to grow your food.
Tomi Engdahl says:
agriculture uses at least 100 times more water than data centers. And it is not all food.
Globally, agriculture accounts for roughly 70% of all freshwater withdrawals, and when you factor in livestock feed production and irrigation, the scale dwarfs nearly every other industrial sector.
Agriculture accounts for roughly 80% to 90% of all consumptive water use in the United States.
Direct Domestic Human Food 15-20%
Domestic Meat & Dairy (Animal Feed) 45-50%
Agricultural Exports (Food & Feed) 15% – 20%
Non-Food Industrial Uses 10% – 15%
Tomi Engdahl says:
evaporative cooling was considered to be the “gold standard” round 2005 to 2024 in USA. Evaporative cooling works well on hot environment, is inexpensive, is energy effifient – but needs lots of water. For evaporative cooling you need clean water because with dirty water the equipment will go bad abd there is danger of spreading pollution (chemicals in water, bacteria)
Tomi Engdahl says:
Dewet Scheepers the big issue, from what i know is: when open loop cooling is used for a data centre it uses potable water
That water is taken out of the environment, rivers, lakes, dams etc, the water evaporates into the atmosphere and then moves to another location, it doesn’t necessarily rain back down in the same environment. Hence it removes water from the local environments.
This causes a reduction in potable water for people, animals and even agriculture.
Desalination plant are a wonderful idea to solve this issue. But increasing the Desalination plants power to meet the demand is often an increase in burning fossil fuels, renewable energy solves this.
But that requires more solar/ wind/ nuclear etc. Which requires more mining which in turn requires burning more diesel which increases co2.
Its a bit of a loop of doom.
But we will get there
Tomi Engdahl says:
That “loop of doom” hits the nail on the head. It is the ultimate systemic bottleneck of modern industrial engineering: you can never truly create resource independence; you can only trade one type of thermodynamic or geographic constraint for another.
Tomi Engdahl says:
That “loop of doom” hits the nail on the head. It is the ultimate systemic bottleneck of modern industrial engineering: you can never truly create resource independence; you can only trade one type of thermodynamic or geographic constraint for another.
The Potable Water Trap: Historically, many facilities tapped straight into municipal potable water supplies because it’s treated, clean, and prevents biological fouling or mineral scaling in the pipes. That’s what creates the direct conflict with local human and agricultural use.
The Shift to Reclaimed Water: To mitigate this, many modern facilities are forced to use reclaimed or recycled wastewater (greywater) rather than pristine drinking water. However, while that stops them from drinking up the local municipal taps, the evaporation loss from the local watershed still happens. The water is still gone from the local environment.
Tomi Engdahl says:
Desalination plants are brilliant for coastal cities (like in parts of the Middle East or Australia), but they also introduce their own massive hurdle.
The Geography Problem: Most data centers aren’t built on ocean beaches.
The Energy Penalty: Reverse osmosis and thermal desalination require high energy inputs. If that energy comes from the grid, it drives up electricity demand even further.
That is the classic plot twist of desalination: you fix a water scarcity problem on land, and you create a chemical and ecological disaster on the coast.
For every gallon of fresh water a desalination plant produces, it generates about 1 to 1.5 gallons of hypersaline brine. Multiply that by millions of gallons a day, and you have a massive waste disposal headache.
Engineers have tried to design around the brine problem, but every fix brings its own bottleneck.
Tomi Engdahl says:
“The numbers continue to sort of boggle our minds.” https://trib.al/lH460xk
Double The Fall
New Data Centers in the US Set to Wipe Out 15 Years of Greenhouse Gas Reductions
“The numbers continue to sort of boggle our minds.”
https://futurism.com/artificial-intelligence/data-centers-us-greenhouse-gas-reductions-emissions?fbclid=IwdGRjcAVUndxjbGNrBVSdzWV4dG4DYWVtAjExAHBkb2YFc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEf_9fCafPSdkg8sagfrHd-YnL4ykf7q0NKGuAAY_EQ2hNLdPtN5rQ9eb8jKOo_aem_a00Ltqg5B_k0laZYx5vxeA
Nearly 15 years worth of hard-won pollution reduction efforts are set to be undone in the United States, as a wave of data centers powered by gas plants enters construction.
That’s according to a new analysis by the Sierra Club, first reported by Bloomberg, which notes that the US energy grid is expected to belch out 2.1 billion metric tons of carbon dioxide by 2030, assuming the country’s currently-planned AI developments enter operation as scheduled.
Tomi Engdahl says:
That’s a massive rise compared to even 2025 — itself a near-record year for US carbon emissions. If the AI industry and its data center developers get their way, the analysis found, emissions will likely skyrocket by 42 percent over the five-year period, an expected increase of 618 million tons compared to 2025.
“The numbers continue to sort of boggle our minds,” the Sierra Club’s principal advisor for climate and energy Jeremy Fisher told Bloomberg.
Due to major bottlenecks in the electrical grid, many data center developers have opted for on- or near-site power plants, which are overwhelmingly fueled by noxious fossil fuels. Projects like Elon Musk’s 1.2 gigawatt data center near Memphis are already choking out residents with fleets of methane generators, which legal experts say flaunt all kinds of local permitting and federal environmental regulations.
https://futurism.com/artificial-intelligence/data-centers-us-greenhouse-gas-reductions-emissions?fbclid=IwdGRjcAVUno5jbGNrBVSdzWV4dG4DYWVtAjExAHBkb2YFc3J0YwZhcHBfaWQMMzUwNjg1NTMxNzI4AAEf_9fCafPSdkg8sagfrHd-YnL4ykf7q0NKGuAAY_EQ2hNLdPtN5rQ9eb8jKOo_aem_a00Ltqg5B_k0laZYx5vxeA
Tomi Engdahl says:
Try dual culture. Livestock under raised panels. Or veggies grown between rows of raised panels. Gives the stock and grass shade, increases dew capture, lusher fodder or better crop yield, and free weeding around your panels from the grazing of the stock. It’s a big thing here in europe. Don’t believe the scaremongering, diversify.
Tomi Engdahl says:
Double The Fall
New Data Centers in the US Set to Wipe Out 15 Years of Greenhouse Gas Reductions
“The numbers continue to sort of boggle our minds.”
https://futurism.com/artificial-intelligence/data-centers-us-greenhouse-gas-reductions-emissions?fbclid=IwVERDUAVVea9leHRuA2FlbQIxMABwZG9mBXNydGMGYXBwX2lkDDM1MDY4NTUzMTcyOAABHyLxjtj3osy5sWQjcvFcg0NI7nkA_sQ-XnJY01HbuSkTbfefaEHJ8fzo7xdp_aem_W1ASNjY1L7i9oPKQ6CgzhQ
Tomi Engdahl says:
1) there is no shortage of fresh water on earth. Locally, yes there can be a shortfall, but water can move across terrain by transportation, pipelines, and more. Again, there is no shortage of fresh water globally.
2) Cooling systems for these centers are closed loop. They take in water one time and treat it with coolant and anti freeze to keep it at the perfect temperature for their cooling towers. They dont continually consume water over its operation.
3) A single multistory office tower uses on average 30x more fresh water than the largest data center on record. This is for restrooms, showers, and drinking fountains. These towers use the same cooling systems that data centers do but for exponentially larger square footage.
4) Farms are not closed loop and generally rely on rainfall for the bulk of their fresh water needs. In some cases elaborate irrigation is used. Either way, however, the water goes into the ground and is consumed by plants and then turned into plant material. Once it reaches maturity, it is harvested and consumed. This water is not returned into the soil or back into the atmosphere for reuse. Therefore crops are in need for continual use of fresh water supplies. Meaning that the amount of water necessary FAR exceeds that of any standard closed loop reused cooling system for any office tower or data center.
Know your facts first.
Tomi Engdahl says:
https://www.facebook.com/share/p/1LadeQzPcB/
US power-sector emissions are going to increase 42% by 2030, as planned data centers switched on.
A new analysis by the Sierra Club, reported by Bloomberg, finds that if planned AI data centers are built, the US power sector would emit about 2.1 billion metric tons of carbon dioxide in 2030.
That is 618 million tons more than in 2025, roughly a 42% jump, enough to undo 15 years’ worth of emissions cuts from US electricity. The biggest driver is “bring-your-own-power” plants built on site for data centers, mostly burning natural gas, which alone could emit more than 460 million tons a year, about as much as all of Mexico.
US electricity emissions peaked in 2007 and have fallen since, as coal plants closed and wind and solar grew. The Sierra Club’s own experts say the numbers “continue to sort of boggle our minds,” and their estimate is several times higher than what they projected just 18 months ago.
Other groups project smaller rises, and some energy analysts say many announced projects may never get built, so this is one advocacy group’s projection, not a certainty.
source: Roston, E., & Raimonde, O. (2026, October 29). New data centers projected to emit more carbon than all of Mexico. Bloomberg.
Tomi Engdahl says:
https://www.theguardian.com/business/2026/nov/03/uk-drax-power-station-ai-datacentre?CMP=fb_gu&utm_medium=Social_img&utm_source=Facebook#Echobox=1793707425
Tomi Engdahl says:
Huge 1GW data centers from OpenAI, Meta, SpaceX, Amazon, Google and others were a key issue in the midterms, largely due to concerns over power and water use.
CNBC’s Katie Tarasov visited data centers and spoke to researchers to find out the truth about how data centers use water.
Find the full report: cnb.cx/3Wev5Ga
Video
https://www.facebook.com/share/v/1NU6Hoz9yS/
Tomi Engdahl says:
Closed-Loop Cooling (PCs, Industrial Systems): Using pure distilled water in a cooling loop without corrosion inhibitors (like glycols or specialized biocides/inhibitors) will eventually strip metal from radiators, water blocks, and pumps, leading to metallic contamination, sludge buildup, and component failure.