Do Data Centers Have a Smaller Effect on Electricity Prices Than We Think? (yahoo.com) 52
"According to a new study by researchers at Harvard University and the University of Pennsylvania, data centers are probably having only a small effect on residential electricity prices," reports the Washington Post.
"The researchers calculated that these operations were responsible for just 4% of the utility rate increases over the last five years..." There are certainly some areas — like in Pennsylvania, Virginia, Ohio and other states covered by grid operator PJM — where data centers have skyrocketed demand, pushing some costs onto residential customers. But multiple studies by different researchers and organizations have failed to find a strong connection between data centers and the overall trajectory of U.S. electricity prices in recent years. Another analysis earlier this year by researchers at the Massachusetts Institute of Technology found that, between 2010 and 2024, data centers increased average electricity prices for residential customers by just 2.1%. Last year, research from the Brattle Group and Lawrence Berkeley National Laboratory found that data centers were actually decreasing costs in certain regions. And another study by the Electric Power Research Institute found that a doubling of data center capacity between 2019 and 2024 resulted in a 4 percent drop in home electricity prices...
The new Harvard and Penn study found a slightly higher impact on industrial electricity prices — the price paid by manufacturers and large facilities. For those customers, data centers were responsible for 15% of the price increase between 2020 and 2025. But for ordinary Americans... the changes simply were not substantial [according to Wharton School professor Serguei Netessine, one of the study's authors]. "So far we have no evidence that data centers increase prices all that much," Netessine said. Instead, he said, prices are rising due to more quotidian factors: rising gas prices, an aging electricity grid, and wildfires and hurricanes requiring improvements to the grid and recovery after disaster...
If the number of data centers continue to boom, experts say, residential prices could be affected more down the line. Data center electricity demand is expected to increase steeply over the next few years, requiring new power generation — wind, solar or natural gas — to be put onto the grid. And residents have issues with data centers that go beyond the price of electricity. The new developments have been singled out for loud noise, polluting power generation, and for disturbing residential neighborhoods and natural landscapes. According to the Pew Research Center, more than half of Americans say that data centers are mostly bad for the environment. Others feel that the AI race — which could ultimately leave many Americans out of work — is being forced upon them.
"I think generally what we're seeing is a broader range of reasons the public has opposed data centers," [said Ryan Hledik, principal at the Brattle Group]. "Concern about AI leading to loss of jobs — all of that is getting packaged together."
"The researchers calculated that these operations were responsible for just 4% of the utility rate increases over the last five years..." There are certainly some areas — like in Pennsylvania, Virginia, Ohio and other states covered by grid operator PJM — where data centers have skyrocketed demand, pushing some costs onto residential customers. But multiple studies by different researchers and organizations have failed to find a strong connection between data centers and the overall trajectory of U.S. electricity prices in recent years. Another analysis earlier this year by researchers at the Massachusetts Institute of Technology found that, between 2010 and 2024, data centers increased average electricity prices for residential customers by just 2.1%. Last year, research from the Brattle Group and Lawrence Berkeley National Laboratory found that data centers were actually decreasing costs in certain regions. And another study by the Electric Power Research Institute found that a doubling of data center capacity between 2019 and 2024 resulted in a 4 percent drop in home electricity prices...
The new Harvard and Penn study found a slightly higher impact on industrial electricity prices — the price paid by manufacturers and large facilities. For those customers, data centers were responsible for 15% of the price increase between 2020 and 2025. But for ordinary Americans... the changes simply were not substantial [according to Wharton School professor Serguei Netessine, one of the study's authors]. "So far we have no evidence that data centers increase prices all that much," Netessine said. Instead, he said, prices are rising due to more quotidian factors: rising gas prices, an aging electricity grid, and wildfires and hurricanes requiring improvements to the grid and recovery after disaster...
If the number of data centers continue to boom, experts say, residential prices could be affected more down the line. Data center electricity demand is expected to increase steeply over the next few years, requiring new power generation — wind, solar or natural gas — to be put onto the grid. And residents have issues with data centers that go beyond the price of electricity. The new developments have been singled out for loud noise, polluting power generation, and for disturbing residential neighborhoods and natural landscapes. According to the Pew Research Center, more than half of Americans say that data centers are mostly bad for the environment. Others feel that the AI race — which could ultimately leave many Americans out of work — is being forced upon them.
"I think generally what we're seeing is a broader range of reasons the public has opposed data centers," [said Ryan Hledik, principal at the Brattle Group]. "Concern about AI leading to loss of jobs — all of that is getting packaged together."
As the old saying goes, (Score:2, Interesting)
When Bill Gates walks into a bar, the average person in the room becomes a millionaire.
Dibs on Betteridge! No, heck no. (Score:1)
What we think about electricity prices is a purely subjective metric, but the objective metric is profit and that has no limit, so the masters-of-the-universe building more AI will obviously be driven to use as much electricity as they can get without limit. And to heck with the price effects on the natural idiots fighting against artificial so-called intelligence. Figures don't lie, but liars figure. Now with AI support!
Weak joke, but the recent one I liked was that Ukraine decided the election behind the
stop with pretend questioning of AI in the news (Score:3)
Responding to the article. The post-election CYA don't regulate me news pieces by the AI industry are not helping.
The government wants a huge push into AI with little regulations simply due to it being a national defense priority to be ahead of large unfriendly countries.
Corporations want AI because it can replace the 25 year business as usual stagnation policy of reducing expenses and stock buybacks which result in no new products that make more than 20% of the top-line corporate revenue.
We have had the p
Re: (Score:1)
"Do Data Centers Have a Smaller Effect on Electricity Prices Than We Think?" asks sockpuppet rag for billionaire who has personally invested $18 billion in data centers.
Re: (Score:2)
Gas turbines (Score:3)
I wonder if they're including all of the data centers who are running their own gas turbines for power. This would obviously bring down any average effects on the local power economy.
Re:Gas turbines (Score:4, Interesting)
Re: (Score:2)
We have a similar situation. A very old power plant that was supposed to be downsized as a secondary baseload source is now running at capacity again, because of the power requirements from a new DC. They closed the local coal mine some time ago, so they need to ship in coal. They are working on refurbishing more of the generating units to ramp up the extra power required, as the DC brings compute power online.
Re:Gas turbines (Score:4, Interesting)
Re: (Score:1)
Y'know, we used to use that NatGas for heating, cooking hot water... miss those days.
But, let's go ahead and use all that NatGas for generating power once we convert everything to electric-only... drilling for oil that we don't use anymore gets us more NatGas.
Of course, the prices are gonna go up... someone's gotta make money.
Even if your area has a huge solar array, you're still gonna pay.
(And, you'll pay dearly)
There is always something missing in these stories (Score:2)
Even with CBS, BBC, etc I have noticed the news is always lacking context to point it looks intentional like this article.
Others here have also covered it: There hey engine gas turbines, the restarted video plans. These are designed to reduce electricity usage. But they also have other problems like noise and pollution in residential areas.
Even if the electricity was not as much of an issue add we think, there is still the water, the false job promises, and gpu noise that causes health problems and even
Simple solution... (Score:3)
There is a very simple solution, which is to legislate that electricity price increases due to data center demand are borne solely by the data centers that create the demand. Then the data centres, as they balloon the electricity costs up, can decide whether or not they want to continue to operate as a power sink where they are, fund their own wind farms or other generation, or move to where there is more spare capacity.
It's not like calculating the exact impact is in any way hard. There is nothing opaque about the numbers.
Re:Simple solution... (Score:4, Informative)
The trick is to make sure that nobody cheats. If one jurisdiction offers more favorable terms than everywhere else, then all the data centers go there and ruin the environment. This is the ideal use of federal or even supranational regulation, to make sure the playing field is level.
Re: (Score:2)
Re: (Score:3)
What are you trying to accomplish with this? You claim it's a simple solution, but solution to what? The enemy here is data centers, not utilities so why burden utilities with a solution?
If data centers should bear all the costs of their demand, why not require that they provide their own service? Then regulate environmental impact of their generation. Your "solution" is a Rube Goldberg that relies on non-corruption, exactly opposite of the "simple" you claim. The purpose of a central utility is to get
Re: (Score:2)
Exactly... you want to plop a DC (data center, not the other place) here, also build some power generation system.
The locals don't pay for a single thing, no tax breaks, no hand-outs... Anthropic (or whoever) foots the entire bill themselves.
Not to mention, the entire community should have input into the decision... not just “But the plans were on display ‘On display? I eventually had to go down to the cellar to find them.’ ‘That’s the display department.’ ‘With a f
Re: (Score:2)
Why do you keep saying that the fundamental economic realities which economic systems are developed to handle are functions of Capitalism?
Nuke it from Orbit? (Score:3)
Re: (Score:2)
To extend your reference, that's not "the only way to be sure."
Building nuclear power plants takes a very long time. And it should, to ensure they are safe. Renewables with shorter spin-up times, like wind and solar, deserve consideration.
Re: (Score:2)
Restarting nuclear power plants should also take a long time, if not quite as long as building them, to be safe. It will be interesting to see how that goes.
Only a small effect? (Score:5, Insightful)
So... 4%, which is double the target inflation rate and is solely due to datacenters... is "small"? That's the increase the researchers indicate is due to datacenters, meaning that's on top of inflation factors.
No thanks. Even if the 4% is spread over the 5 years, nearly 1% a year is also too damned much.
If a new subdivision is built and real people live there creating demands on services, that's understandable. If new industrial or commercial developments that create jobs and products is built and that creates demands on services, that's understandable. Yes, there are datacenters that provide things we apparently want such as hosting DoorDash or Amazon or porn sites. But the current rush is about AI, which is being shoved where it isn't asked-for, like Red Dye #3.
If the results of this study was a tenth of a percent per year or less, maybe it'd be "small". This isn't.
Re:Only a small effect? (Score:5, Informative)
4%, which is double the target inflation rate
You're conflating two different things. It's not that data centers increase residential electric prices by 4%, it's that they cause 4% of the observed increase, which was about 50% over 15 years. That's about a 2.7% per year increase, so data centers cause about a 0.11% per year increase -- which you say might be "small".
Re: (Score:2, Flamebait)
So... 4%, which is double the target inflation rate and is solely due to datacenters... is "small"?
4% of the increases due to data centers leaves 96% of the increases to other factors. OF THE INCREASES. Got it? But you got your rage-boner, so I guess it's OK, for you.
Re: (Score:2)
When a utility asks the California Pubic Utilities Commission for a 4% increase in rates, that is considered a significant rate increase.
Re: Only a small effect? (Score:2)
Another major part of the problem is PJM/Dominion Powers forecasts for growth. These two orgs house more data centers than the rest of the world combined.
As recently as 2022 their 10 year forecast showed basically no growth in demand above baseline. Utterly incompetent.
So they're also playing a massive game of catchup while still approving every new GW demand request of DCs.
Good Fast Cheap pick two as the addage goes.
Re: Only a small effect? (Score:2)
Was off a bit, it's only 1/3 of the worlds data centers
Re: (Score:3)
Electricity prices are only part of the utility bill. Locally the utility district is installing a new transmission line because the current one will be overloaded when the next data center goes on-line.
Emergency effect (Score:4, Insightful)
Say the town uses 150 MW.
The 4 nearby generator is capable of 200 MW. If those fail they have to import from the more expensive place at double normal prices.
Before data center, they import once a year.
After data center they import 10 times, all in January/February
That may only be 4% more, but it hits all together .It's like a 40% price raise for that time period.
Re: (Score:2)
Re: (Score:1)
That may only be 4% more, but it hits all together .It's like a 40% price raise for that time period.
Do you think if "they" offer to pay that 40%, their stock price will go up by 4000%?
Now you know why you're debating your home electricity costs vs. cats grooving to the latest Beethoven rap.
consider the source (Score:2, Insightful)
Re: (Score:1)
Re: (Score:2)
coming from the Washington Post, i’m gonna chock this up to “fake news” and move along.
The actual source is the study and two of the three researchers are at The Wharton School at Penn, the alma mater of the current President, who *loves* data centers...
The local power company doesn't charge data center (Score:2)
More specifically, researchers at ... (Score:4, Interesting)
"According to a new study by researchers at Harvard University and the University of Pennsylvania, ...
According to TFStudy, one researcher is from Harvard and two are from The Wharton School at Penn, the alma mater of the current President, who *loves* data centers. Not necessarily implying anything about their less unfavorable conclusions, just throwing that out there.
Well I found some bullshit in TFS first thing (Score:5, Interesting)
TL;DR: No.
I picked one thing at psuedorandom to check up on:
This isn't random because when I see "average" I also see red.
What does the study [mit.edu] actually conclude?
2.7%, not 2.1%, and we pay for it whether it's up front, or on the back end. I also liked this part:
Enjoy, Texans! Though perhaps that's something they can address now?
Rather than go on to check on other "citations" used in this story, I'll let someone else do that (or maybe do it in a separate comment, if inspiration strikes) and instead close this comment with this paragraph from the fine citation:
Re: (Score:2)
Also, future increases is what the concerns were always about. Backward looking studies, that are already two years behind, isn't telling us shit.
Not the point (Score:2)
Economies of scale and their status as public utilities means prices should be reduced by new rollouts. If that's not happening, someone is stealing from your community.
Betteridge's law of headlines (Score:1)
No (Score:1)
Perhaps We Should Not Group AI with other Data (Score:2)