||| BY MICHAEL RIORDAN, SPECIAL TO THE SEATTLE TIMES |||
According to a recent series of articles in The Seattle Times, co-published with ProPublica, businesses, farmers and residents in central Washington are already beginning to feel the pinch on their power supplies and likely facing substantial rate hikes next year. And that is happening before the full impact of artificial intelligence on information services, which some have suggested could increase data-center electricity demands tenfold.
These data centers — of which there are now over 100 in the state — require oodles of electricity for two major reasons: to power the racks of computer servers inside and to keep their sophisticated microprocessors from overheating. When thousands of these chips are sequestered in an enclosed space, the heat produced can halt their operation; it must be removed by power-guzzling cooling fans or liquids to keep them functioning.
The largest of these facilities, known as hyperscale data centers, draw over 100 megawatts (100 MW, or 100,000 kilowatts) — roughly 10% of the capacity of a large modern power plant. Already, there are several such centers in central Washington. Many other, smaller data centers cluster around Seattle and Tacoma.
All told, data centers in the Pacific Northwest currently require average electrical power of around 1,300 MW, according to energy analyst Massoud Jourabchi of the Greenway Research Group. That’s almost 6% of the average Pacific Northwest regional load, but it will grow rapidly during the next five years. In the local Seattle/Tacoma area, data centers already consume over 10% of available power.
Data-center operators like Microsoft need to secure cheap, reliable power they can use with minimal interruption. That’s why they located many of the larger centers in Grant and Douglas counties, close to Columbia River dams, and contracted with the local public utilities to supply hydropower 24 hours a day. But according to a Times article, the utilities are now having to purchase extra “unspecified power” from the grid to meet the needs of all their customers. Usually generated by out-of-state suppliers, often using fossil fuels, this power comes at a higher cost that must be passed on to its consumers.
One might fault these public utility managers for selling the crown jewels of their domains to voracious outsiders, although the lure of lucrative local tax payments probably influenced their thinking. But they likely did not anticipate the explosive growth of the data-center power needs, which have recently been doubling every four to five years.
A 2024 study by the Northwest Power and Conservation Council projects that NW regional data-center power requirements could double again by 2029. This estimate is extremely uncertain, however. Only an added 2,000 MW is needed for such “tech loads” in one low-end scenario, but 6,500 MW may be required in extreme cases. Much of this huge uncertainty arises from the unknown impact of artificial intelligence, which is only now being felt as Amazon, Google, Meta, Microsoft and others implement AI technologies in their data centers.
But one thing is obvious from this analysis: Data-center power requirements will dominate electrical load growth over the coming years, substantially more than the expected demand from electric vehicles. The next generation of graphics microprocessors soon to be installed in servers to enable AI applications (for example, Nvidia’s Blackwell chipset) will draw over a kilowatt apiece and require nearly as much power to cool them. Such demands exceed the average power needs of a typical Washington house, which will draw 1 to 2 kilowatts during normal weather conditions. (For a 30-day billing period, that translates to 720 to 1,440 kilowatt-hours of electrical consumption.)
The next generation of graphics microprocessors soon to be installed in servers to enable AI applications (for example, Nvidia’s Blackwell chipset) will draw over a kilowatt apiece and require nearly as much power to cool them, writes the author. Pictured are Nvidia H100 chips inside a server room in Navi Mumbai, India, on March 14. (Dhiraj Singh / Bloomberg)
Industry executives are struggling to accommodate these anticipated demands without requiring new fossil-fueled power plants, given their promises — and government mandates — to reduce greenhouse gas emissions. In Pennsylvania, for example, Microsoft just announced its intention to lease the full 835 MW output of a resurrected Three Mile Island nuclear reactor, enough to power 700,000 area homes. This nuclear power will be fed to the grid to supply its expanding fleet of AI-enabled data centers.
Elsewhere, utilities are delaying plans to retire nuclear and coal-fired power plants or converting the latter to natural gas in an attempt to meet anticipated data-center needs. Other utilities and companies are contemplating the possibility of building geothermal power plants, which might be an option for the Pacific Northwest, or hoping to commission small modular nuclear reactors.
But it will be difficult to bring these new power supplies online fast enough to meet the exploding data-center demands. Such a lapse could lead to more frequent near-term power shortages that will have to be met by purchasing costly grid power — as the Grant County public utility has done.
During the severe January 2024 cold snap, for example, Pacific Northwest utilities had to import an average 4,900 MW of power from systems in California and the Rocky Mountain states, at unit costs far above the baseline hydropower prices. Such extra charges feed into the costs per kilowatt-hour paid by ordinary ratepayers, currently growing by 5% to 6% annually in the Puget Sound area.
And as voracious data-center electricity demands exceed the limits of available supplies, the ruthless law of supply and demand will inevitably kick in and increase our rates still further. The only other option would be rolling blackouts, which nobody wants.
If not met by renewable energy sources in the coming years, these demands may also endanger the near-term goals of Washington’s 2019 Clean Energy Transformation Act. Hydropower is already tapped out, and the low-hanging fruit of energy efficiency has largely been picked. And highly variable solar and wind power require long-term storage to provide the steady 24-hour supplies favored by data centers.
Power shortages and consequent rate increases therefore look increasingly likely if artificial intelligence is allowed to proceed unchecked and unregulated. Realizing the full benefits of this technological revolution — which was recognized this year by the Nobel Prizes in physics and chemistry — will require electricity that does not yet exist and will be difficult if not impossible for utilities to supply. This threatened energy shortage may prove to be the Achilles’ heel of the promising AI industry.
ORIGINALLY PUBLISHED IN THE SEATTLE TIMES
Michael Riordan; is a physics historian and co-author of “The Solar Home Book” and “Crystal Fire: The Invention of the Transistor and Birth of the Information Age.”
**If you are reading theOrcasonian for free, thank your fellow islanders. If you would like to support theOrcasonian CLICK HERE to set your modestly-priced, voluntary subscription. Otherwise, no worries; we’re happy to share with you.**
The promise of AI: making our electricity much more expensive; eating up more and more energy and materials to manufacture the computers, chips, cooling systems, buildings, etc.; adding more massive buildings to the concrete corridors of WA state; inventing 10’s of thousands of deadly chemicals for a whole new generation of terrorists; using students and children’s photos to create child sex abuse materials; scamming people with deepfakes and phishing; creating ever more sophisticated malware; a resurgence of uranium mining to supply the new nuclear power plants required to power the AI to do all these things (the Pinyon Plain uranium mine in Arizona opened in January 2024, and at least five other mines in the U.S. are being reactivated in Texas, Utah, Wyoming, and Arizona, where I’m not so sure the generations of people already poisoned by uranium mining will be happy about this)… What promise, indeed.
Meanwhile, ecosystems are crashing, and everyone’s too busy talking to Chat GPT to notice.
“Let the bast***’s freeze to death in the dark” was scrawled on a whiteboard 50 years ago when the Alvin Vogel nuclear plant was put on the shelf after the Three Mile Island accident. Public opinion killed a bright future in the power industry, so I turned to the petro-chemical field, but public opinion rose against that too as I aged in place and opportunities faded.
What was I to do in the twilight of my career when those my age were being “eased out” of the workforce? I had it! Inspiration in my late fifties, I redirected my career into wastewater treatment for a public agency that did not discriminate against age. I figured with so many people so full of sh*t I would be set for life — and I was right, I was able to retire at 68 and live the dream on Orcas.
“We” have over 50 years of lost time to make up as “we” endure the consequences of our caprice. that Dr. Riordan so eloquently describes.
Student body left, student body right — guess its just part of the human condition.
Thank you Michael for sharing your observations, and congratulations again on your recent significant recognition to science.
“Student body left, student body right — guess its just part of the human condition.” It’s ironic how public opinion shifts when people’s eyes are opened and they start to see beyond the corporate message and start realizing the bigger picture. You’re only fooling yourself if you think that public opinion alone has ever been responsible for stopping a for-profit driven corporate industry that was on the wrong side of history.
“Let the bast***’s freeze to death in the dark.” You can believe that if you want… but I don’t. Nobody froze to death, and nobody had to go in the dark after the Alvin Vogel nuclear plant and Three Mile island were “put on the shelf”.
Nuclear power? Petro chemical? “Public opinion killed a bright (meaning ‘your’) future in the power industry.” And which, accident free, no risk, human and environmentally healthy, green nuclear and petro chemical companies did you work for? It’s not all about you Phil… get over it.
The more things change… the more they stay the same. As we shift from one power resource paradigm to another we forget that everything has a caveat, that every very form of refuge has a price, and that there is no free ride. And as we’re once again expected to fall in line and buy into another expensive, polluting, environmentally unfriendly corporate driven short-term response that hosts a plethora of negative long-term economical, environmental, health, and safety consequences the public is expected to accept that we need more energy with the anticipation being that we will eventually normalize the threats to, and the resulting implications to our health and safety, our environment, the ecological functions that support life on this planet, the many wars being fought for resources, and the genocides that are occurring around the world in order to secure our future resource needs, (“our interests”). What’s wrong with this picture?
Are we forgetting that nuclear power plants are expensive… that they are tax-payer subsidized to build, maintain and operate, that they are powered by and produce some of the most toxic substances known to mankind, and that they also produce enormous amounts of greenhouse gases during their construction, and during their fueling up and shutdown phases?
Because of the enormous amounts of electricity that will be needed to power these future AI data centers, and with nuclear power now being labeled a “Green energy” because of their ability to produce enormous amounts of low carbon emission electricity, (forgetting for a moment that CO2 is only one of a handful of greenhouse gas emissions, that it doesn’t come close to being the most prolific one, that on the scale of ecological overshoot it is a low player on the list of climate change threats, and that even if we were to stop burning all fossil fuels now the amount of CO2 already in the atmosphere will cause Earth’s climate to continue warming for decades, triggering heat waves, droughts, rising sea levels, and extreme weather), companies like Microsoft will begin to define our future. As a result of this new need for electrical power, older nuclear power plants that were either shutdown due to being too expensive to operate, or had design failures, operating failures, or were past their operational lifetimes, are now being re-energized to supplant the power needs of these mega-industry AI data centers.
Are we really going to believe that in this time of ecological overshoot and climate change that there’s something logical or ethical about embracing a future based on shifting and increased power needs that hold such enormous and increasing negative economic, environmental, ecological, and health implications for mankind? I say no. We all know that the answers don’t matter when you’re presented with the right questions, and I feel that the current limited corporate narrative that’s being offered to the public is one that poses all the wrong questions. I feel that we are simply, once again, being offered an incomplete narrative, and that we are being led further down the spiraling abyss of a corporate created, government abetted systemic crisis.
They call it “climate change”.” They call it “ecological overshoot.” But knowing what we know now, and realizing that the corporate industries behind this, (and the governments that have abetted them), have known for decades about the risks related to their actions, and knowing now that their models have even predicted the negative implications that are now arising from their profit motivated efforts… I can only agree with Elisabeth when, as she stated once before, “It’s a Ponzi scheme”. Knowing what we know now and what science is revealing about the future that awaits us, isn’t it more apt that we begin to refer to this for what it really is… a worldwide genocide?
If you havn’t seen these, the following are interesting articles relative to the issue at hand–
CounterPunch (Richard Heinberg) 6/08/23– Capturing Carbon With Machines Is a Failure—So Why Are We Subsidizing It?
https://www.counterpunch.org/2023/06/08/capturing-carbon-with-machines-is-a-failure-so-why-are-we-subsidizing-it-2/
Utility Dive (Brian Martucci) 10/22/24– With Palisades and Three Mile Island units set to restart, could more retired reactors follow?
https://www.utilitydive.com/news/palisades-three-mile-island-duane-arnold-nuclear-reactor-restart-holtec-nextera-constellation-nrc/730393/?utm_source=Sailthru&utm_medium=email&utm_campaign=Issue:%202024-10-22%20Utility%20Dive%20Newsletter%20%5Bissue:67056%5D&utm_term=Utility%20Dive
Counterpunch (Joshua Frank) 8/22/24– Of Leaks and Lies: A Looming Nuclear Catastrophe Threatens the Pacific Northwest
https://www.counterpunch.org/2024/08/22/of-leaks-and-lies/
Good comments, Michael “MJ” Johnson! While we have differed in the past, I totally support you on this one!
Almost any new technology is a double-edged sword that cuts both ways. This is especially true of artificial intelligence.
Often called machine learning, AI-enabled technologies have been making major, important inroads into the sciences for over two decades, as the November issue of National Geographic Magazine makes clear:
https://www.nationalgeographic.com/magazine
This year, two Nobel Prizes — in physics and chemistry — are being awarded for originating and employing machine learning in these disciplines. In chemistry, UW Professor David Baker was cited for his research using machine learning to elucidate protein folding and predict the structures of useful new proteins. In my own discipline, particle physics, machine learning has been used since the mid-1990s to decipher the increasingly complex patterns of proton-collision debris at Fermilab and CERN.
But yes, there are downsides to artificial intelligence, and they are big ones. In her dystopian vision of the AI-enabled future, Elisabeth Robson outlines some of the worst possible instances by profiteering and evil actors. One has to realize, however, that these are not inherent characteristics of the new technology; they are the result of its HUMAN abuse. And it is therefore the responsibility of our political representatives to recognize these possibilities and enact measures to counter them. Which I admit will not be an easy task, given the power and complexity of AI technologies.
To take an immediate example that is the focus of my Seattle Times essay, the power demands of AI-enabled data centers can unfortunately spill over into our own electricity bills if these demands cause system-wide power shortages in times of grid stress. Our governor and legislators must build strong guardrails against any such possibility. But so far they seem to be encouraging the data-center industries by giving it unwarranted tax breaks to lure them to the state. This has to STOP!
Somebody has to step up and blunt the hurtful, damaging edge of AI technology if we are to realize its promise.
And since we’re on the topic of new technologies that devour electricity, let’s take a look at cryptocurrency. Which wastes a fabulous amount of power:
https://ccaf.io/cbnsi/cbeci
“In her dystopian vision of the AI-enabled future, Elisabeth Robson outlines some of the worst possible instances by profiteering and evil actors.”
The things I listed are all things that are happening RIGHT NOW with AI, not in some imagined future. Just FYI.
The fair thing to do, I would think, is to say, “If you’re gonna build a data center, you need to find your own power source and it has to be 100% carbon free. Your data farms are disturbing the delicate balance between supply and demand on our system. It shouldn’t be our responsibility to fix it, or our rate payers responsibility to shoulder the cost.”
Chris, there is no such thing as 100% carbon free power. We all know this, right? Even our supposedly “carbon free hydropower dams” are made out of massive amounts of concrete, made from materials that were mined from the Earth, requiring huge amounts of energy and CO2 emissions to make, dams requiring huge amounts of energy and CO2 emissions to build, causing methane release (80x CO2 warming) from the reservoirs behind the dams, and destroying the rivers that are now dammed.
Even if AI data centers had their own what people believe to be “carbon free” power (but isn’t due to materials, supply chain, installation, maintenance, etc. issues), that doesn’t alleviate the harms that data centers and AI can do to the world in other ways.
And it would prove Richard York et al’s 2019 and Tim Garrett’s points that new power/energy is *added* to total energy used, rather than *replacing* other forms of energy. Growth in energy leads to growth in development which demands more energy and materials and grows the economy which then requires more energy to maintain that growth and so the insane cycle of more more more continues.
At what point do we stop this insanity of infinite growth on a finite planet? Unfortunately, most people (like 99.9999%) have normalized it, and so we expect new industries like AI data centers and pretend that it’s totally fine to create a whole new power-hungry industry just at the point when climate change and other symptoms of ecological overshoot are becoming so utterly obvious and devastating that it should be beyond unethical to even consider creating such an industry.
As I’ve said before, we humans tend to do things because we *can*, never asking if we *should*.
What troubles me most about the AI data-center explosion, is the fact that Washington state is encouraging it by giving the industry tax breaks:
https://www.propublica.org/article/data-centers-clean-energy-washington-state
When I questioned our state Representative Alex Ramel about this during my research on the Seattle Times essay, he replied with a question: “Do you want to see these data centers built in Idaho?” It’s clear that he doesn’t grasp the AI problem facing us, despite his strong support of renewable energy. Having written extensively about that subject — including a bestseller, “The Solar Home Book” — I can see no realistic way that renewable energy can hope to keep up with the looming demands of AI-enabled data centers unless they are regulated and limited in their power consumption. It cannot be left to the market.
In the same vein, I am disappointed in Governor Inslee, who just endorsed the development of small modular nuclear reactors in WA state as a way to meet the looming demands with “renewable” resources. Anyone who takes a close look at these reactors, which typically generate only 100 MW, and reviews the history of nuclear power in general realizes that these power plants are a decade away from producing the additional power needed to cope with AI demands. But according to the projections I have seen, those demands are going to be exploding in the next five years — again, if left to the electricity marketplace.
That means undue short-term pressure on our electricity supply and having to import large amounts of power from producers outside the Pacific Northwest in times of grid stress — as happened during the January cold snap that Chris and Chom Graecen have explicitly shown above. Unless the large power users (e.g., the data centers) are made to pay their representative shares of the costs of this added power, some of those costs will be borne by residential customers in the form of increased per-kWh electrical rates. Think of it as an AI tax. And as the ProPublica/Seattle Times article above reports, much of the imported electricity will be generated with fossil fuels, effectively lowering the states renewables percentage. That’s called Double Whammy.
I submit that our state legislators and Governor Inslee need to stop dancing around the difficult questions and face up to the troubling AI music.
A collision is coming between generation and consumption ending up, I suspect, that except for what we are guaranteed under BPA contract (new one coming up soon) and the Northwest Power Act, our own consumption will be either limited or (knowing us) subject to surcharge as BPA goes into a congested market to furnish us excess power say in a winter cold snap, or more often as our primary energy form becomes electric. (Sorry for the long sentence!)
If there is rationing, AI (which is likely to be deemed critical infrastructure) will not be rationed, we will.
The Small Modular Reactor projects are financially infeasible and will be for some time … until, I suspect, rates become high enough to justify their investment. The fact that private as well as government money is going in that direction indicates a belief that at some point the lines on the graph will intersect.
Sadly, the AI & data center race is not about what the society needs or what our earth’s fragile ecological balance can bear. It’s mostly about technological, economic and political advantages and leverage in order to maintain global dominance.
Green tech, like renewable energy, used to be how the U.S. & the West maintained an economic edge but China has won decidedly on clean energy technology, from solar, wind, battery and even EVs.
So now AI is seen as the frontier that the US needs to dominate & cannot lose to China.Setting the climate goals aside, the AI’s massive appetite could even be viewed as a boon to the clean energy race as it suddenly injects new life to the US’s struggling nuclear energy industry. The government is going all in to support AI (CHIPS act) and the nuclear industry (IRA), of course.
What’s more important: geopolitics (democratic values & economic advantage) vs avoiding ecological disasters?
If history offers any guide, geopolitics (or “national security”) tends to trump everything else. This is maddening and will not end well.
A case in point: the number of climate-related academic articles that discuss risks of societal collapse has risen sharply in recent years. We are in for a wild ride!
Wow! I can only add, as an aside, that reading this article and the above comments reminds me of one of the main benefits for me of living in this community–the wealth of informed, balanced and intelligent thought-sharing available to all of us. I learn so much. Not always a pleasant perspective, but always thought-provoking and appreciated. Thank you all, and thank you Lin for making this forum available.
Just to add to the existing voices here (some good comments above), Amazon announced today that they plan to build four small “SMR” reactors near an existing energy generation complex near Richland, Washington and the Columbia River. The investment into newer and safer nuclear tech is a positive thing, but at the same time, it’s a bit crazy that Amazon is so hungry for cheap electricity that they are going to start building nuclear power plants. While Amazon might be pursuing low-carbon or no-carbon energy sources, their business model is built on shipping the maximum amount of “stuff” to the maximum amount of people. The energy supply is only one part of the puzzle. Reducing consumption of junk, and the associated transport emissions and raw materials required, is just as important.
Columbia Riverkeeper made some good comments about this horrific news about amazon building nuclear power plants.
https://www.youtube.com/watch?v=RySTjlFPNhw
I disagree, of course, with her comments about solar, wind, and batteries, but otherwise her comments are excellent.
Elizabeth — so where did humans go off the rails? Was it the invention of the steam engine? The plow? Harnessing fire?
Agriculture, for sure; the plow came later and made it faster and easier.
But of course, humans were exterminating the megafauna long before agriculture.
I think it was when we developed a shoulder capable of throwing a spear.
Been downhill ever since for the rest of the living planet.
We do what all species with an advantage and few predators do: take everything we can, overshoot our carrying capacity, and (soon) suffer collapse as a result. Unfortunately, we are global, and thus our deadly impact is also global. I think it’s unlikely any large land mammals will survive our assault once all is said and done.
Ms. Robson: Homo sapiens sapiens is only a well developed primate, the planet’s current apex predator. Its character traits are inherent and may be immutable with its fate dependent on its adaptability. Its impacts are called evolution, the planet will adapt and continue to evolve with or without it.
My own experience, noted previously, leads me to reflect on 50 years of collective dithering only to return to the same place with no solutions other than solar panels, windmills and Don Quixote. The well-meaning can speak of sustainable alternatives, sunshine and lollipops but the earth will prevail.
Phil, totally agree about evolution. I was just trying to say that our evolutionary traits mean we will not last long as a species on this planet (compared to, say, sharks at 420 million years and counting!) and unfortunately, like the asteroid, we’re taking out many other species as we rampage over the Earth. We’re also the only species currently or recently alive to leave an ever-worsening legacy of global toxic pollution behind us for those species who do make it past the wildlife holocaust we are committing to deal with. Whether the web of life remains intact enough for a few to make it past the juggernaut for future biodiversity to flourish will remain to non-humans to discover.
“Its impacts are called evolution, the planet will adapt and continue to evolve with or without it.”
“The well-meaning can speak of sustainable alternatives, sunshine and lollipops but the earth will prevail.”
That’s all well and good for the Earth… but I want to live.