WP: Amid explosive demand, America is running out of power

[China]

Amid explosive demand, America is running out of power

 

Vast swaths of the United States are at risk of running short of power as electricity-hungry data centers and clean-technology factories proliferate around the country, leaving utilities and regulators grasping for credible plans to expand the nation’s creaking power grid.

 

In Georgia, demand for industrial power is surging to record highs, with the projection of electricity use for the next decade now 17 times what it was only recently. Arizona Public Service, the largest utility in that state, is also struggling to keep up, projecting it will be out of transmission capacity before the end of the decade absent major upgrades.

 

Northern Virginia needs the equivalent of several large nuclear power plants to serve all the new data centers planned and under construction. Texas, where electricity shortages are already routine on hot summer days, faces the same dilemma.

 

The soaring demand is touching off a scramble to try to squeeze more juice out of an aging power grid while pushing commercial customers to go to extraordinary lengths to lock down energy sources, such as building their own power plants.

 

“When you look at the numbers, it is staggering,” said Jason Shaw, chairman of the Georgia Public Service Commission, which regulates electricity. “It makes you scratch your head and wonder how we ended up in this situation. How were the projections that far off? This has created a challenge like we have never seen before.”

 

A major factor behind the skyrocketing demand is the rapid innovation in artificial intelligence, which is driving the construction of large warehouses of computing infrastructure that require exponentially more power than traditional data centers. AI is also part of a huge scale-up of cloud computing. Tech firms like Amazon, Apple, Google, Meta and Microsoft are scouring the nation for sites for new data centers, and many lesser-known firms are also on the hunt.

The proliferation of crypto-mining, in which currencies like bitcoin are transacted and minted, is also driving data center growth. It is all putting new pressures on an overtaxed grid — the network of transmission lines and power stations that move electricity around the country. Bottlenecks are mounting, leaving both new generators of energy, particularly clean energy, and large consumers facing growing wait times for hookups.

 

Click on the link for the full article

[CousinsCowgirl84]

Get you some of those solar panels. 

 

Actually, with the interest rate the way it is now it doesn’t pay to install them (in Virginia).

 

Break even is like 25 years now… when I got them it was like 11 years… .9 percent interest rate vs 4-5 percent, if not more. 😬

Edited March 7 by CousinsCowgirl84

[Riggo-toni]

40% of America's population lives in the "megalopolis" from DC up to Boston, where there simply isn't enough wind or sun (at least with current or developing technology) to make much of a dent in supply. Natural gas, which fracking has made abundant, must be used, at least as a transition to greener energy sources.  Solar and wind are practical in places like Texas, but sadly there the fossil fuels lobby will hold them back.

[CousinsCowgirl84]

3 hours ago, Riggo-toni said:

40% of America's population lives in the "megalopolis" from DC up to Boston, where there simply isn't enough wind or sun (at least with current or developing technology) to make much of a dent in supply. Natural gas, which fracking has made abundant, must be used, at least as a transition to greener energy sources.  Solar and wind are practical in places like Texas, but sadly there the fossil fuels lobby will hold them back.

Seems to me most of the problem is in Texas and California where solar should be fine and dandy.

 

fyi, solar and wind are practically in the NE as well.

[Larry]

[Destino]

Build some nuclear power plants. Name them after European nations. 

[EmirOfShmo]

Dude I talked to who works for Dominion power said electrical engineers are making bank & jumping jobs due to demand. 

[Jabbyrwock]

39 minutes ago, Destino said:

Build some nuclear power plants.

 

Correct

 

40 minutes ago, Destino said:

Name them after European nations. 

 

Incorrect

[PeterMP]

Nuclear power plants take years and even decades to build, and it isn't just regulation.  There is like one company that makes many of the parts, and they're backed up.

 

And unlike other things building more and as the technology has matured the cost of building one has gone up and not down.

 

https://energy.mit.edu/news/building-nuclear-power-plants/

 

In 2006, there was an application to build a nuclear power plants on site in GA where there already was one.  As of 2024 after nearly 2 decades of time and price over runs, the company that was originally building them going bankrupt and despite the US government putting up over $8 billion on loan guarantees back in 2010, one of them became operational in July 2023 and the other just in March of this year.

 

 

 

Saying build more nuclear power plants is a fine solution if you are planning on having an energy crisis in 2055 and don't care what the final costs will be. (Unless you have some plan also to change how they are built.)

 

 

Edited March 8 by PeterMP

[PeterMP]

We have to recognize that energy is not like other commodities/resources.  Not matter how much we produce, demand will just increase as people will just build bigger/less efficient things.  Yes, you have to have incentives to keep supply high, but there is no way to generate a supply that will out compete demand (given our current system) and generate low prices.  The key is to price as a function of the amount of demand and not per a unit price and maybe as a matter of function.  Energy prices likely should include a societal value judgement.

[Destino]

7 hours ago, PeterMP said:

Saying build more nuclear power plants is a fine solution if you are planning on having an energy crisis in 2055 and don't care what the final costs will be. (Unless you have some plan also to change how they are built.)

I see no reason to think we’ll need less energy in 2055 or 2085.  On the contrary, people want everything to be electric and data centers are only going to get bigger and hungrier for power as we advance into an AI dominated era. Doesn’t it make sense to plan for a future? Anyone know how SMR development is going?

 

[FrFan]

Woke up cancel vultures 😆

 

Edited March 8 by FrFan

[PeterMP]

2 hours ago, Destino said:

I see no reason to think we’ll need less energy in 2055 or 2085.  On the contrary, people want everything to be electric and data centers are only going to get bigger and hungrier for power as we advance into an AI dominated era. Doesn’t it make sense to plan for a future? Anyone know how SMR development is going?

 

 

Yes, it makes sense to plan for the future.  But planning for the future rarely means building things that are essentially the last centuries technology, especially when they take decades to build, take decades more to pay off, and you also have a need in the short term that they aren't going to be able to fill.  People saying building more nuclear power plants are ignoring that.   We also only have so much money to put to energy production.  Every dollar you put into a nuclear power plant is a dollar that you aren't putting into something that can be used in 5 years, and thinking about building nuclear power plants makes it harder to get investment for other things.  You can't ignore the costs that are making it questionable current nuclear power plants are ever going to actually pay off, and if they do it is going to take 50 years or so.  But if you've put the money into build them, they are going to run.  People looking at investing in other types of energy are going to see the investment (which is going to take huge amounts of federal money in terms of things like loan guarantees) and that's going to make it less likely they will invest in things that we can use to get more electricity in the nearer future.

 

SMR development appears as if it is close to the point where we could build them and bring them on the grid.  But the long term economics of SMR are still very much unclear.  Acting like SMR is a real solution is just making things up.

 

At some point, inefficient energy production is as costly as to little energy production.  Especially when there are other solutions.  For example, tax data centers based on the amount of power they use, and force them to be more efficient in their energy consumption.  There's no reason that a greater number of data centers in 50 years or so can't use less power than current ones if we put effort, research, and money into making them more efficient.  If we simply let them run and do what they want, they will remain inefficient and will consume all of the power put out by as many nuclear power plants as you can build.  Same things with cars.  We know how to make cars that are much more efficient than the average car on the road today.  People just chose not to buy them.  We are never going to out build people's/companies' willingness to be energy inefficient.  Putting money into building nuclear power plants so that people/company can be energy inefficient doesn't make sense.

 

(This isn't to say that we shouldn't consider building more nuclear power.  But looking at where we are and saying well just build more nuclear doesn't address the short term problem and actually might be making it worse.  Nuclear needs to be part of the mix for at least the foreseeable future.  And we made a mistake through the 1980s and 1990s of doing no building of nuclear power plants.  But that mistake has been made.  It isn't at all clear if the correction for that mistake is to now try to make up for it.)

[Destino]

Efficiency is a goal, but I’m not at all convinced that it’s a solution. Perhaps all the data centers of the future will use less power, but maybe they won’t. Taxing them into oblivion doesn’t really change incentives too much as the tech sector already is driven towards increases efficiency. I guess having some relocate would technically reduce energy consumption, but not in the way we want. The energy sectors serves the economy, not the other way around. At least it does in my way of thinking. 
 

I hear what you’re saying about needs five years out, but I just don’t agree that infrastructure planning should operate solely on such a short timescale. We can plan for the future and we should. A future of growth.

 

 

[Riggo-toni]

What about SMRs? Apparently Japan is developing plans to produce them at scale, and potentially producing hydrogen as a by-product.

[PeterMP]

5 hours ago, Destino said:

Efficiency is a goal, but I’m not at all convinced that it’s a solution. Perhaps all the data centers of the future will use less power, but maybe they won’t. Taxing them into oblivion doesn’t really change incentives too much as the tech sector already is driven towards increases efficiency. I guess having some relocate would technically reduce energy consumption, but not in the way we want. The energy sectors serves the economy, not the other way around. At least it does in my way of thinking. 
 

I hear what you’re saying about needs five years out, but I just don’t agree that infrastructure planning should operate solely on such a short timescale. We can plan for the future and we should. A future of growth.

 

 

 

I'm not at all sure the tech industry is already driven towards increases in efficiency.  Certainly, their CEO pay doesn't seem to be consistent with that.  It seems to me that Reddit could have afforded to pay more in taxes for the electricity to support their servers/data, or put more money into research driving long-term increases in efficiency, and paid their CEO a little less.  II'll bet Reddit's budget on R&D on how to make a more efficient server or store data more efficiently last year was 0.

 

(I think essentially every industry today is driven by making sure upper management and stock holders make money.)

 

The article mentioned bit coining mining.  The bit coin mining doesn't seem to be driven to increase efficiency.  If you think somebody like Sam Bankman was about driving things to be more efficient long term and making the investments and doing the research required, I think you're wrong.

 

In terms of taxing anything into oblivion, that wouldn't be the objective.  The objective would be to make sure they are working to be efficient. Let's not be ridiculous and take points to the extreme.  Even smoking hasn't been taxed into oblivion.

 

But I'm also not sure of the actual economic advantage to having the computers here in terms of data centers.  In some cases, there are advantages to being close to the computers, but those tend to be pretty small and in rare cases.  The number of workers directly employed by them I suspect is going to be small.  From my perspective, I don't think massive subsidies to the energy industry that can be used to support data centers, especially if there are essentially no controls on how much and how they use energy, make much sense.

 

Money is money.  Spending money to subsidize energy inefficiency is money spent.  

Edited March 8 by PeterMP

[FrFan]

Advanced Nuclear Power Reactors

[Destino]

The motivations toward efficiency already in place are simple. It reduces costs. Something most businesses are constantly looking to do. This is true despite what CEO salaries might be, which is an entirely different topic. Taxing software for not being efficient enough sounds insane to me in an era where AI development is considered a major focus of technology research. Why would we want that?

 

Could Reddit be a bit more efficient. Maybe? Ultimately I care little. Major advances in efficiency tend to come from the hardware side of things. I could see rewards being offered for improving efficiency, but simply raising taxes on energy feels more like a sin tax than anything else. I don’t see how that creates jobs or improves the quality of life of any American.

 

Chasing jobs out of the US by asking questions like if we really needs data centers, seems to be entirely wrong minded to me. One that takes too much for granted. We need every good job. We need more good jobs. We need to plan to create more of them so our people can thrive. 
 

The future will see increased energy consumption. We have every reason to produce more, and more cleanly, than ever before. 

[PeterMP]

4 hours ago, Destino said:

The motivations toward efficiency already in place are simple. It reduces costs. Something most businesses are constantly looking to do. This is true despite what CEO salaries might be, which is an entirely different topic. Taxing software for not being efficient enough sounds insane to me in an era where AI development is considered a major focus of technology research. Why would we want that?

 

Could Reddit be a bit more efficient. Maybe? Ultimately I care little. Major advances in efficiency tend to come from the hardware side of things. I could see rewards being offered for improving efficiency, but simply raising taxes on energy feels more like a sin tax than anything else. I don’t see how that creates jobs or improves the quality of life of any American.

 

Chasing jobs out of the US by asking questions like if we really needs data centers, seems to be entirely wrong minded to me. One that takes too much for granted. We need every good job. We need more good jobs. We need to plan to create more of them so our people can thrive. 
 

The future will see increased energy consumption. We have every reason to produce more, and more cleanly, than ever before. 

 

Increasing efficiency doesn't allow one to instantly decrease costs and save money in most cases.  Increases in efficiency almost always require some up front costs and money is saved in the long run.  And R&D to make new technology to increase efficiency is an even more extreme in that costs are paid up front in hope of saving/making money much later.

 

And then the amount you are paying for something like electricity matters in terms of how long it takes to make back the original investment and so whether it makes sense to do make the initial investment.  Completely making up some numbers, let's say I'm buying a million units of power a year (whatever measure you want to use), and I'm pay 0.35 per a unit of power.  I can install some new equipment that is going to cost up front  $10,000,000 dollars that will cut my energy use in (so 500,000 units).  Well the pay out there is going to be 50+ years.  No company is making that investment.

 

Now, take the same scenario and say I'm paying $3.5 per a unit of energy.  That same piece of equipment now pays out in under 6 years.  There is a much better chance that companies will make that investment.

 

And because the time scale matters the focus on long term savings vs. short term profit making for CEOs and stock holders matters.  If you're primary focus is short term pay outs, then you aren't going to make the investments needed to be reasonably maximally efficient over the long term.

 

So yes, CEO salaries and stock buy backs do matter.  Those are taking money that could be put into equipment and R&D that could increase efficiency but only pay out over the long term.  Companies that are interested in short term maximizing CEO profits and stock prices are almost certainly not investing in equipment, infrastructure, and R&D to maximize efficiency and long term saving/profits.

And there is a feed forward effect because if companies using things are more aggressive about buying equipment/tools that is more efficient, then companies that are developing things will be more aggressive about doing R&D to make things that are more efficient.

 

It isn't a sin tax.  It is here is the true costs of running your business tax.  If your data center is going to require building a new nuclear power plant that is going to be tied to billions of dollar of US government subsidies and then more costs to store the waste indefinitely, then you aren't really paying the true costs of running your business.  

 

We don't need every good job at any costs.  Spending billions of dollars to create a handful of new good jobs doesn't make sense.  Even for jobs, there has to be a cost/benefit analysis.  How much are we really paying to create that job vs. how many jobs and to what benefit are we as the larger public getting.

 

We will need more energy in the future.  That doesn't mean that it makes sense to build (many) more nuclear power plants given the costs and how they will impact the building of other types of energy producing plants.  I'm not arguing that we won't need more energy.  I'm arguing we should try to ensure that companies that are using energy are doing so efficiently and take into account the company's societal value to try to minimize how much energy we need in the future and so how much we have to pay to build the power plants that we'll need in the future.  And not just say, hay let's spend billions and billions of dollars to throw up some nuclear power plants so that people/companies can waste money/energy in the future. 

Edited March 9 by PeterMP

[Corcaigh]

8 hours ago, Riggo-toni said:

What about SMRs? Apparently Japan is developing plans to produce them at scale, and potentially producing hydrogen as a by-product.

I think we are still a few years away from commercial

SMR offerings and until then it’s hard to evaluate the economics. Because of complexity, nuclear reactors are currently about 50% higher than wind/solar and natural gas.
 

Reducing the size of a reactor might reduce some costs but would increase others such as security and waste management given there would be more of them and spread out geographically.

 

 

[Destino]

It’s a sin tax and attempts to dress it up with this modern rhetorical trick of “true cost” doesn’t change it. Your plan is simply to make electricity more expensive so that people will try harder to use less. Just like every sin tax it’s justified by claiming the extra tax money will pay for societal costs of the sin you’re hoping to discourage. I’ve seen the exact same argument made for eating meat, owning cars, and entering a city in a car. Raising the cost of energy sounds great when framed against CEO pay, less great when bankrupts the local bakery. Worse still when jobs flow over the border or across the sea while CEO pay continues to rise unbothered, as it always does. 
 

And in the end, you still need more energy. You’re still left with a world where transportation is switching to electric, where AIs are on the rise, and more of everything we do switches from fossil fuels to electricity. All while populations rise. 
 

I think nuclear is part of the solution, and the price tag doesn’t scare me. I’m an old fashioned democrat, massive terrifyingly expensive infrastructure projects are my idea of heaven. Let’s build 12. To start. With union labor, of course.

[Corcaigh]

I’d like to see a diverse approach:

(1) more renewable, including home solar

(2) better energy efficiency - my utilities are a third now in a properly constructed and insulated home compared to the ****ty construction late-1980s McMansion I used to live in. And the square footage is higher in our new home.

(3) increase nuclear

(4) more reservoirs that can serve to smooth out the peaks and troughs of rainfall with climate change, and also to act as pump storage for when renewables are producing more than we can use.

 

Edited March 9 by Corcaigh

[PeterMP]

4 hours ago, Destino said:

Raising the cost of energy sounds great when framed against CEO pay, less great when bankrupts the local bakery. Worse still when jobs flow over the border or across the sea while CEO pay continues to rise unbothered, as it always does. 

 

Odd to see you call yourself an old fashioned Democrat because that is some first class, let's ignore facts, data, and what the other person is actually saying, Republican fear mongering there.  OMG if we don't subsidize the billionaires companies so they can become trillionaires, they'll take all the jobs, we'll hurt small businesses, and we'll be left with nothing.  (And there's nothing we can do about the billionaires becoming trillionaires anyway.)

 

Fundamentally, companies don't strive to be efficient.  They strive to make money for stock holders and management (and given the current situation in a lot of cases in the short term).  If you supply them with cheap energy, you reduce their incentive to be efficient, and if you are subsidizing energy production you are subsidizing their energy inefficiency.  As we have been building them, nuclear power plants don't get off the ground without huge government subsidizes.  Which actually hurts the building of other types of energy companies and makes it more likely in the near to mid-term we'll have problems.  Those facts are indisputable.

Edited March 9 by PeterMP

[Renegade7]

I underestimated the threat of us not being able to supply enough power as quickly as its happening.

 

If we gonna have hard conversations, like why are people growing food in the desert, then we should be honest about limiting data centers devoted to crypto...

 

I'm not against a requirement, even if it costs subsidies, to put solar panels in as many single gamely houses as possible..if it not more.  If businesses are in every means neccesary mode to find power, maybe what we need is a grid where power plants and always first resort for power but supplementing entities that can get their own (focusing on critical infrastructure, not

data centers)

[Destino]

6 hours ago, PeterMP said:

 

Odd to see you call yourself an old fashioned Democrat because that is some first class, let's ignore facts, data, and what the other person is actually saying, Republican fear mongering there.  OMG if we don't subsidize the billionaires companies so they can become trillionaires, they'll take all the jobs, we'll hurt small businesses, and we'll be left with nothing.  (And there's nothing we can do about the billionaires becoming trillionaires anyway.)

 

Is it odd to see a democrat be willing to spend on big infrastructure projects that create lots of middle class jobs and keep energy costs low? That’s practically the platform. That’s what won us the working class. And facts? You’ve offered unestablished social costs and sending jobs away because we don’t really need them. With promises of making CEOs pay for it, that rings about as true as Mexico paying for a border wall. When have democrats ever balked at infrastructure spending? 
 

Want to hit CEOs in the pocketbook? Tax Capital Gains exactly like income. Raise taxes on the top tax bracket. Fund the IRS. Want to hit companies and help workers? Expand unions. 
 

 

 

 

6 hours ago, PeterMP said:

 

Fundamentally, companies don't strive to be efficient.  They strive to make money for stock holders and management (and given the current situation in a lot of cases in the short term).  If you supply them with cheap energy, you reduce their incentive to be efficient, and if you are subsidizing energy production you are subsidizing their energy inefficiency.  As we have been building them, nuclear power plants don't get off the ground without huge government subsidizes.  Which actually hurts the building of other types of energy companies and makes it more likely in the near to mid-term we'll have problems.  Those facts are indisputable.

Tech has grown more efficient over time though. Significantly so. Everything from light bulbs to computer processors to washing machines have grown more efficient over time without having to raise costs.

 

Should we stop subsidizing green energy too? It wouldn’t have gotten off the ground had we done that. Infrastructure spending grows the economy and improves quality of life. It can also shift us away from fossil fuels. Sending jobs away and making our lives more expensive is not what we elect democrats to do. 
 

Nuclear is part of the energy future, expensive as it may be. So is energy research.