“In 2022, PJM stopped processing new applications for power plant connections after it was overloaded with more than 2,000 requests from renewable power projects, each of which required engineering studies before they could connect to the grid”
If all the actors here were acting in good faith, I would agree. This is not a problem of well-sited and competently researched energy generation plans simply being held up by nothing but red tape and a rubber stamp.
There is certainly some of that, but it’s not the whole story. Operating the grid has historically taken a lot of “gentleman’s agreements” and inertia from the proper overengineering of overbuilding from generation’s past to function. Those social constructs have largely broken down by now.
When you don’t build stuff for a generation (no pun intended) you lose the ability to.
Interesting times ahead indeed.
As a result, grid operators and lawmakers in the west have collectively forgotten how to deal with rapid growth of electricity demand.
[1] https://www.eia.gov/energyexplained/electricity/use-of-elect...
[2] https://www.statista.com/statistics/383650/consumption-of-el...
https://www.opb.org/article/2025/06/05/oregon-data-centers-c...
This is just restricting industry because they don't want to build the infrastructure to support it. Which, fair enough. At this point the fight about industrialising vs de-industrialising has been fought out. But exactly why there is this big round of chip sanctions on China when the US doesn't want to build the power plants to use them domestically will quickly become a baffler. China can build coal plants at a rate of 2/week, plus solar panels, nuclear plants and what have you. I bet they're willing to run all these data centres.
I don't see why residents in Oregon would like to compete with big tech dollars for utility services. It is a losing endouver when you are far from the printing presses.
So I guess "fair" is adjusted for access to capital?
Anyone using that energy to make a profit - that is, run a business - pays market rate from the start.
1) The explicit idea there is to push costs onto businesses that are higher than the market equilibrium.
2) That pressures businesses to leave Oregon and less energy infrastructure will be built.
To be fair, I don't think that is a bug - Oregon is probably doing it on purpose. But that isn't fair, it is just anti-industry/anti-business policy. It is entirely possible (probable, even) that on net the people of Oregon will be worse off after they've been given a dose of "fairness" since there is good reason to believe that in the long term more capital investment is better for the residents. They'll be getting a generous amount of welfare, but they'll need it because investment in power production would be down.
Also, if that is the policy there is an interesting debate question of why the "fair" price shouldn't be $0 under the use threshold. Without market signalling it is just picking a random number anyway, may as well pick $0. But that level of welfare might do so much economic damage that people wouldn't stomach it, revealing some problematic aspects to the entire approach.
Of course that means that they have a perverse incentive to increase them as much as possible and tell then
* "emissions" is a catch all word/concept that includes efficiency,all cost's, sustainability, scalability, local air polution, and CO², CO, etc, while serving a political purpose that has been embraced by every single person in China
Ultimately, the previous pricing tiers seem to have been determined by political means. This presents a contradiction for proponents of "fair share" pricing. If the previous political process resulted in an "unfair" outcome, then why is the new politically determined outcome "more fair"? What does "fair share" really mean here? Wouldn't the previous outcome suggest that the political process has issues with creating "fair" outcomes?
I get the impression that Oregon Public Broadcasting would dismiss or even demonize market based metrics as "unfair". There also seems to be a vague sense that tech bros and cryptocurrency users have become class enemies for some political persuasions.
[citation needed]
The article did go on...
> According to Oregon CUB, large industrial users, like data centers, that have connected to Portland General Electric’s system pay about 8 cents per kilowatt hour, or kWH, which is the unit of energy used when 1,000 watts of power is used in an hour. Residential customers in the same PGE system pay close to 20 cents per kilowatt hour
But that's a disingenuous comparison. Data centers are cheaper to serve because there's ~one massive line going to one place, power use is generally more fixed and predictable, and they might be paying less because they can reduce power use during heat waves.
In Poland/EU during summer we have electricity surplus, not deficit.
Oregon was very slow do adapt solar due to poor regulation and other reasons. It even says so in the article.
https://www.opb.org/article/2024/08/26/fast-growing-energy-d...
Most of the distribution costs occur on the other side of a substation due to efficiency losses with long distance transmission etc. A data center located next to a power plant has some advantages, but still needs power when that power plant is offline.
Total electric production is stagnant over the same period https://www.iea.org/countries/united-states/electricity
If you want to grow value in an economy, how do you do it without growing energy supply and reducing energy cost? Intensive growth (the economist's answer to infinite growth on a finite planet) can only do so much for an economy as we continually add more valuable things to spend energy inputs on.
Ultimately the process of taking an input and making it more valuable is an application of energy.
Take LED lighting, an absolutely enormous energy save producing better lighting than incandescent bulbs.
> There’s no value
That's not a reasonable interpretation. But even so, your example's a good one:
100w bulb lasts hundreds of hours
5w led puts out comparable light with 20x less energy AND lasts tens of thousands of hours
Seems a slam dunk in the LED's favour right?
But it's still the same problem, you fall out of intensive growth and back to extensive growth as soon as you want more lighting. So are we saying there's a bar on how many lights we should ever want on the planet?
most of the increase in value comes out of getting more output out of the same input, which is what the application of intelligence is supposed to do. In qualitative terms this means dematerialisation. Your pocket supercomputer is more valuable than something that used to fill a stadium not because it applies more energy, but the opposite.
You used to drive your steel car with dinosaur fuel to the video rental store where you picked up plastic boxes to put into your single purpose device, now you send information through a fiber-optic cable, calculate the energy differential for that. Real civilizational progress is doing what you used to do with matter with light instead.
Now we're just making more stuff, at some point you make so much more stuff that you overwhelm the benefits of reduced inputs to make each thing.
One of those curves has a finite limit (you can't input less than zero stuff to make a thing). The other is unbounded, i desire more stuff.
It isn't. Your demand and your attention are finite too, in fact they're so finite that competition for it is already extreme. There is so much more for you to consume out there that prices in virtual goods have largely been driven to zero. And of course while there's always a new thing, old things continuously have to go away for you to make room. What you are trying to say is that your interests are open-ended, but certainly not infinite or even increasing over any period of time.
That's exactly why energy consumption has stagnated or been falling already in rich places. Virtualisation has driven down the energy cost of your consumption far in excess of additional stuff you can pile on, and that's only going to accelerate.
AI data centers might temporarily raise energy consumption as existing infrastructure runs in parallel, but if you take the AI promises at face value it will eventually depress energy usage as it dematerialises billions of workers and infrastructure. A "dark factory" is a very early version of that.
It's not like a datacenter just takes power with no oversight, they have to be given a tie-in to the grid of a certain size first by the power company.
The area of solar panels needed to power data centers is .. maybe 100x the area of the data center?
Only a tiny percentage of the USA is covered by data centers. Maybe 100 million square meters? That would be something like 0.001%.
Then covering 0.1% of the USA with solar panels could power all data centers.
Why do we stall human progress and not get started on this end state tech yesterday? It is the key to a functioning singularity.
And the kicker? It’s easy. We already have all necessary parts available off shelf. And it’s green, ain’t no pollution like space pollution.