Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/ai-data-centers-are-forcing-dirty-%E2%80%98peaker%E2%80%99-power-plants-back-into-service.24256/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2030770
            [XFI] => 1060170
        )

    [wordpress] => /var/www/html
)

AI data centers are forcing dirty ‘peaker’ power plants back into service

To quote myself, and my answer to your questions:



I wish I was trying to be humorous.
Exactly. Decide what you are going to do and live with the consequences. If you want datacenters, build them. If you don't, then other countries can build them and they can control datacenters.

The US decided not to build nuclear plants and not to dam rivers over the last 30 years. Thats fine.... now that leads to other decisions.
 
None other than the socialist Robert Reich (former US Sec of Labor) and constant critic of capitalism (at when he doesn't personally benefit) is one of the most famous NIMBY people going on record for opposing more dense or more affordable housing in Berkeley, because it would negatively impact the character of the community.
We have NIMBY “protectors” of all political persuasions who keep us from building - Reich’s protecting us from student noise pollution, ranchers and farmers protecting us from transmission lines and high speed rail, Native Americans and allies protecting us from pipelines, and the current administration protecting us (plus whales and birds) from windmills (though seemingly offshore oil drilling is no threat whatsoever).
 
Glass half full - at least AI seems to be marginally more useful than cryptomining that is *still* growing in energy use (and environmental impact).


Emptying the glass..

I personally think humans won't get serious about pollution until it's impacts are truly catastrophic. (Increases in cancer rates and chronic disease for humans and dying animal/plant species and food chain impacts aren't enough for people to generally notice).

AI power demands may accelerate that timetable a little bit..
 
The US decided not to build nuclear plants and not to dam rivers over the last 30 years. Thats fine.... now that leads to other decisions.
The end of damming rivers started in the 1960s, the complete until very recently end of licencing new nuclear power plants began with breaking out the Nuclear Regulatory Commission from the Atomic Energy Commission in 1975.

Another gambit that prevented, damaged, or destroyed projects like NH's Seabrook was too many states changing their laws so a utility had to borrow up front the money for the construction of new plants instead of charging their customers as they progressed.

So half a century of energy poverty, then you have to factor in the deliberate, burn your boats destruction with explosives coal baseline power plants with limited replacements of natural gas fired ones, including some plants converting their boilers.

Thus your question:
What does the electricity use trend look like in the US for the past 25 years?
can't easily be answered as big consumers of electricity adopted to the regime risk of US. E.g. 1980 to the last few years primary aluminum production went from 33% to 2% of the world's output, 4.65 million to 785,000 metric tonnes, per an AI aided search.

Unless you build very expensive storage systems, renewables are a hindrance to operating a grid, providing neither baseline or peaking power. A great deal of this has simply come out of grid reserves, I'm very happy to be in the Southwest Power Pool which has the highest of all the nation's reserves, currently somewhere around the traditional 17% if I remember a list of them all correctly, although that's going down.
 
Last edited:
Thus your question:can't easily be answered as big consumers of electricity adopted to the regime risk of US. E.g. 1980 to the last few years primary aluminum production went from 33% to 2% of the world's output, 4.65 million to 785,000 metric tonnes, per an AI aided search.

Unless you build very expensive storage systems, renewables are a hindrance to operating a grid, providing neither baseline or peaking power. A great deal of this has simply come out of grid reserves, I'm very happy to be in the Southwest Power Pool which has the highest of all the nation's reserves, currently somewhere around the traditional 17% if I remember a list of them all correctly, although that's going down.
I agree with your overall sentinment/post, but renewables not being able to baseline or handle peak power ignores how far grid-scale battery technology has come along, especially LFP based batteries from China. True you can't easily go 100% renewable because of the weather, but peaker plants exist for the same reason - because you need some level of redundancy even when you're depleting finite resources to provide power.

Grid sized batteries can also help double (or more) the available energy from fossil fuel and nuclear sources too as you could run a power plant at 100% longer while storing the energy for high demand times.

Take a look at some of the 'megapack' projects from Tesla (and others) and the economics - especially in Australia and the US. The cost per MwH of storage is about 10% what it was 10-12 years ago, and that's with the batteries lasting longer (both calendar aging and ability to handle higher daily cycle counts).

On the overall cost argument -- in the US at least, fossil fuels are heavily subsidized by the US government (see the US military defense budget for details). If there were truly zero subsidies for both fossil fuels and renewables, I think renewable sources (including batteries -- which can be recycled) would compete well.

OT I wish we'd take nuclear more seriously as a power source. It feels like there's a lot more safety and overall efficiency that can be engineered than even the advanced stuff we have today.
 
I agree with your overall sentinment/post, but renewables not being able to baseline or handle peak power ignores how far grid-scale battery technology has come along, especially LFP based batteries from China. True you can't easily go 100% renewable because of the weather, but peaker plants exist for the same reason - because you need some level of redundancy even when you're depleting finite resources to provide power.
Actually, peaker plants exist so you can build base-load generation to some calculated level, like the 95th percentile load, rather than the 100th percentile load. For generation agencies and companies, this reduces the cost of the base load plants, and saves fuel costs for fossil fuel plants because peaker plants can quickly started and shut off, while base load plants usually run at one speed (redline) and can't be easily or quickly shutdown and restarted.
Grid sized batteries can also help double (or more) the available energy from fossil fuel and nuclear sources too as you could run a power plant at 100% longer while storing the energy for high demand times.
Valid point, but base load plants generally run only at one level. What you describe is the theory behind pumped hydropower.
On the overall cost argument -- in the US at least, fossil fuels are heavily subsidized by the US government (see the US military defense budget for details). If there were truly zero subsidies for both fossil fuels and renewables, I think renewable sources (including batteries -- which can be recycled) would compete well.
How are fossil fuels subsidized in the defense budget? (I know the answer, and I think you're making a political statement, not a factual argument.)

Actually, the USG subsidizes all (most?) mineral extraction with what's called a "depletion allowance". Last I looked, lithium has a higher depletion allowance than oil and gas. :)

And don't ask me to defend US tax codes. I'm starting on my income tax preparations for 2025, and nothing I'm thinking should be shared with polite company, not even you guys.
OT I wish we'd take nuclear more seriously as a power source. It feels like there's a lot more safety and overall efficiency that can be engineered than even the advanced stuff we have today.
Agreed.
 
How are fossil fuels subsidized in the defense budget? (I know the answer, and I think you're making a political statement, not a factual argument.)
There’s an argument often made that the US has had an “oil for security policy” with most of the oil producers in the Middle East that extends as far back as the 1940s (Saudi Arabia was extended Lend/Lease support). Some of the most visible elements included the US-supported 1956 Iranian coup that reinstated the Shah (and lowered oil prices) and the 50 years of the US military policing the Persian Gulf to make sure the Straits of Hormuz stay open to tanker traffic. The Fifth Fleet, reformed in 1995, and operating out of Bahrain, is essentially a fossil fuel continuity/pricing insurance policy.
 
Actually, peaker plants exist so you can build base-load generation to some calculated level, like the 95th percentile load, rather than the 100th percentile load.
You’re right - utility scale batteries (and pumped hydropower) can actually do more than peakers because they can absorb excess production and deliver back on a time shifted basis - not just additional peak capacity. So they are a better match for non-dispatch-able renewables. Batteries also offer far faster response times within their discharge range, vs peakers, which give them strong economic advantages in many situations. And in California, the levelized cost of energy/capacity for 4‑hour lithium‑ion batteries beats new simple‑cycle gas turbines for peaking capacity.
 
How are fossil fuels subsidized in the defense budget? (I know the answer, and I think you're making a political statement, not a factual argument.)

Actually, the USG subsidizes all (most?) mineral extraction with what's called a "depletion allowance". Last I looked, lithium has a higher depletion allowance than oil and gas. :)

Defense examples specifically (resource subsidies via military budget):

(Chose CNN specifically for this topic because I know you love CNN :-P )
+ The US Army Corp of Engineers does prospecting and executes solutions that help extract resources or support resource efforts later

My comment was "US Government" though, not specifically via DoD budget. I was really thinking of the usual -- massive tax subsidies (direct or indirect such as depreciation), tax credits, and other incentives. Take a look into " End Oil and Gas Tax Subsidies Act (H.R. 383). ", an AI summary of some things it's going after:

Key Provisions of H.R. 383 (2025)
  • Intangible Drilling Costs (IDCs): Repeals the deduction for intangible drilling and development costs of oil and gas wells.
  • Percentage Depletion: Eliminates the percentage depletion allowance.
  • Geological and Geophysical Expenditures: Extends the amortization period for these costs from 24 months to 7 years.
  • Marginal Wells & Enhanced Oil Recovery: Removes tax credits for producing oil and gas from marginal wells and for tertiary injectant expenses used in enhanced oil recovery.
  • Passive Loss Limitations: Repeals the exception to passive loss limitations for working interests in oil and gas property.
...

Also -- there are billions of government subsidies for closing old oil wells, costs that should have been burdened by the businesses building and selling the oil: https://en.wikipedia.org/wiki/Orphaned_wells_in_the_United_States

P.S. To tie this back to the purpose of this forum (Semiconductors), I think low cost energy is very important in enabling the Semi industry; but how we achieve that can inform the long term sustainability of energy supplies for this industry. I think no subsidies (i.e. level playing field across all technologies) could be a better solution than what we have today.
 
There’s an argument often made that the US has had an “oil for security policy” with most of the oil producers in the Middle East that extends as far back as the 1940s (Saudi Arabia was extended Lend/Lease support). Some of the most visible elements included the US-supported 1956 Iranian coup that reinstated the Shah (and lowered oil prices) and the 50 years of the US military policing the Persian Gulf to make sure the Straits of Hormuz stay open to tanker traffic. The Fifth Fleet, reformed in 1995, and operating out of Bahrain, is essentially a fossil fuel continuity/pricing insurance policy.
Very funny. And what were the energy alternatives 70 years ago? Or even 30 years ago? Of course there weren't any, or only available at far higher cost and lower reliability.

I don't like fossil fuels any more than you or Xebec do, maybe even less because I live in a major oil and gas production state and extraction is disgusting, but I'm also a realist. Billions of people are alive today and fed because fossil fuels were commercialized. The 21st century economies we know and enjoy wouldn't exist without them.
 
You’re right - utility scale batteries (and pumped hydropower) can actually do more than peakers because they can absorb excess production and deliver back on a time shifted basis - not just additional peak capacity. So they are a better match for non-dispatch-able renewables. Batteries also offer far faster response times within their discharge range, vs peakers, which give them strong economic advantages in many situations. And in California, the levelized cost of energy/capacity for 4‑hour lithium‑ion batteries beats new simple‑cycle gas turbines for peaking capacity.
If only this was the full story. The fundamental problem for the grid is that batteries take at least four hours to recharge to get another four hours of discharge. In the event of long lasting peak events with little wind or sun it's lights out.
 
Last edited:
Defense examples specifically (resource subsidies via military budget):

(Chose CNN specifically for this topic because I know you love CNN :-P )
+ The US Army Corp of Engineers does prospecting and executes solutions that help extract resources or support resource efforts later

My comment was "US Government" though, not specifically via DoD budget. I was really thinking of the usual -- massive tax subsidies (direct or indirect such as depreciation), tax credits, and other incentives. Take a look into " End Oil and Gas Tax Subsidies Act (H.R. 383). ", an AI summary of some things it's going after:

Key Provisions of H.R. 383 (2025)
  • Intangible Drilling Costs (IDCs): Repeals the deduction for intangible drilling and development costs of oil and gas wells.
  • Percentage Depletion: Eliminates the percentage depletion allowance.
  • Geological and Geophysical Expenditures: Extends the amortization period for these costs from 24 months to 7 years.
  • Marginal Wells & Enhanced Oil Recovery: Removes tax credits for producing oil and gas from marginal wells and for tertiary injectant expenses used in enhanced oil recovery.
  • Passive Loss Limitations: Repeals the exception to passive loss limitations for working interests in oil and gas property.
...

Also -- there are billions of government subsidies for closing old oil wells, costs that should have been burdened by the businesses building and selling the oil: https://en.wikipedia.org/wiki/Orphaned_wells_in_the_United_States

P.S. To tie this back to the purpose of this forum (Semiconductors), I think low cost energy is very important in enabling the Semi industry; but how we achieve that can inform the long term sustainability of energy supplies for this industry. I think no subsidies (i.e. level playing field across all technologies) could be a better solution than what we have today.
More politics. And I dislike subsidies too. And after 70 years of trying we haven't solved the nuclear waste problem either. We are a mess, but it is a mess that works for the time being.
 
The fundamental problem for the grid is that batteries take at least fours to recharge to get another four hours of discharge. In the event of long lasting peak events with little wind or sun it's lights out.
An obvious answer from the boosters of renewables would be to build even bigger battery systems.

But our biggest problem is requiring storage systems of any type in the first place, which generally isn't happening in the US.
 
More politics. And I dislike subsidies too. And after 70 years of trying we haven't solved the nuclear waste problem either. We are a mess, but it is a mess that works for the time being.
I mean by definition anything the government does is politically motivated or "politics"? :)

Agreed it's a joke we haven't dealt with nuclear waste, and we are a mess that works for the time being.
 
Very funny. And what were the energy alternatives 70 years ago? Or even 30 years ago? Of course there weren't any, or only available at far higher cost and lower reliability.

I don't like fossil fuels any more than you or Xebec do, maybe even less because I live in a major oil and gas production state and extraction is disgusting, but I'm also a realist. Billions of people are alive today and fed because fossil fuels were commercialized. The 21st century economies we know and enjoy wouldn't exist without them.

I don't think any of us are saying switch off fossil fuels tonight and see what happens.. or that they weren't important to "get here".

But they are a slippery slope for society - if we run out before we get to sustainable energy types, it's not going to be pretty. OTOH if we pollute so much that our health is permanently damaged (see the declining T rate of men over the generations as one possible example of environmental factors) - we've also waited too long to transition.

..

Separately, there's a theory that the 'great filter' could be fossil fuels. A society develops so far on fossil fuels and then if they run out before 'the next step' then their society collapses and they fail to reach the stars. Or that fossil fuels are actually required as a step to make it to the stars. (i.e. pyromania is a pre-requisite for developing energy :) ).
 
Very funny. And what were the energy alternatives 70 years ago? Or even 30 years ago? Of course there weren't any, or only available at far higher cost and lower reliability.
It doesn’t really matter when the subsidies occurred - subsidies are subsidies and strongly affect long term usage. The dual oil shocks of the 1970s show how much our trajectory of usage in the US would have been different had the US not wielded military power and influence to tamp down the cost of imported oil, prior to that point in time. But then again, the whole history of oil, coal and gas is full of government manipulation of the market for various policy purposes, often to keep prices low, but not always.

 
If only this was the full story. The fundamental problem for the grid is that batteries take at least four hours to recharge to get another four hours of discharge. In the event of long lasting peak events with little wind or sun it's lights out.
You’re right and you’re wrong - utility scale batteries aren’t the best fix for extended grid outages for the reasons you highlight. But neither are peakers, since the most significant cause of extended grid outages is damage to distribution and transmission equipment/lines. Mitigating those requires localized generation and/or batteries with renewables.

Texas and California have shown that using batteries with renewables (mostly solar) to reshape supply works and can actually improve grid stability. One only has to look at the CAISO or ERCOT daily data to see that their biggest challenge is a 4-6 hour peak period, after the sun stops producing.
 
Back
Top