Getting the sense that a few of these companies can't get out of their own way..
Public opinion is crucial to getting projects developed and there are developers out there not securing air permits for their BTM gas.
newjerseymonitor.com/2026/09…
Interesting commentary from Lotus' CEO re: contracting a CCGT with a hyperscaler:
15-month negotiation, first 6 were strictly on physics of how it would work. Very different from renewables vPPAs.
Shocked at the risks. 99.9% availability, which a standalone plant can't deliver.
Data center load contributed roughly $10/MWh to the cost of wholesale power for the first half of 2026.
Per PJM IMM
On top of this, they just got their CPCN (major permit) revoked by the PSC.
"American Transmission Co. says it’s the first time they’ve seen the PSC revoke a completeness determination in recent decades, if ever."
wpr.org/news/psc-denies-tran…
Big news from Wisconsin
A hyperscaler has sued v large load tariff demands (specifically high collaterals)
- As per new rules customers below A-/A3 credit rating must post huge collateral, in this case >$7B in security + >$100M in annual costs
- Oracle (1 GW Stargate in WI) argues few companies qualify, waiver flexibility was removed, costs are disproportionate, rule would push $$ out of WI
- Reminder that such tariff provisions are being developed for large loads with customer protection in mind, but are certainly unprecedented
- Big risks is utilities build out infrastructure for demand that never energizes, leaving existing ratepayers holding the bag
- Oracle is willing to put up letter of credit for $700M (10%)
1
3
515
Expecting this to be a more common occurence.
Anecdotal stories from the industry referencing over $1 bn in collateral needed for even 300 MW projects.
I'm sure some banks will step in with LCs, but easy to imagine this disqualifies tons of projects.
Big news from Wisconsin
A hyperscaler has sued v large load tariff demands (specifically high collaterals)
- As per new rules customers below A-/A3 credit rating must post huge collateral, in this case >$7B in security + >$100M in annual costs
- Oracle (1 GW Stargate in WI) argues few companies qualify, waiver flexibility was removed, costs are disproportionate, rule would push $$ out of WI
- Reminder that such tariff provisions are being developed for large loads with customer protection in mind, but are certainly unprecedented
- Big risks is utilities build out infrastructure for demand that never energizes, leaving existing ratepayers holding the bag
- Oracle is willing to put up letter of credit for $700M (10%)
3
327
Part of the reason why some of the hyperscalers are moving to off-grid for the first 5 years.
They know any price increase will be blamed on them even if they’re not the true culprit.
Easiest way to prove they aren’t increasing rates is by not connecting to the grid at all.
An unsavory truth for “Just make the data centers pay” - electricity rates in many jurisdictions would go up even w/ zero demand from data centers
Much of the coal fleet is approaching or already at end-of-life. Poles & wires are very old. Gotta run just to stay in place
1
216
Exactly this. Take the LCOE of a project and see if you can reasonable realize that price in the market (through a PPA or merchant pricing).
Solar will have lower LCOE but harder time realizing premium pricing. Gas/other baseload the opposite.
LCOE is the price an asset needs to pencil
That doesn’t often equal the value it creates based on the time weighted generation into the system.
1
177
Discussions about the grid are still focused on meeting peaks.
In ERCOT, those peaks have mostly been handled. Even in winter storms, larger peaks are being met by more batteries.
The emerging constraint isn't the peak itself.
It's the surrounding hours, where net load could soon exceed thermal capacity.
This is an interesting perspective that runs counter to my last tweet.
But this goes back to the discussion that people have been having this past week.
If data centers are not willing to pay for these grid costs, then who ends up paying?
It's clearly the ratepayer.
There is a theory that data centers are just waiting to sign a cheque to fund huge grid projects we could not do otherwise (due to no social consensus BTW), if only we just give them some guarantees on speed/reliability.
Are they really so price inelastic? I'm curious because it does fly in the face of every project I work on, where I find sophisticated operators keenly sensitive to the sticker price and the details of how it amounted to that. I obv need to request intros to these firms with unlimited $$$!! 😅
Don’t take my word for it btw, see this from the CEO of Fervo, an enhanced geothermal startup. I daresay he has more exposure to detailed price negotiations for large loads power offtake than most of us.
We are not going to get realistic solutions by assuming a willingness to pay that does not actually exist and which does not seem to have been benchmarked to anything.
A catch all for 'data centers' anyway does not make much sense given how varied the business models are around inference vs training vs cloud, smaller vs larger operators, co-located tenants vs single facility etc etc. All these players have very different cost sensitivities which makes a coordinated investment even harder.
Dumb question here.. but what's actually stopping hyperscalers from funding transmission?
I'm sure they're willing to.
Isn't it the same issues transmission buildout always runs into? NIMBY, permitting, etc.
I'm not sure that capital and willingness is the issue here.
Yes, generational infrastructure building opportunity with price inelastic demand willing to underwrite or fund things for broader system needs or impact beyond their operational needs for the sake of speed & social license to operate
Instead, many are arguing to build nothing or promoting a narrow set of solutions
Anyone who was invested in growing the electricity system massively could’ve dreamed for years ago in terms of momentum, political attention, ecosystem commitments is here in abundance yet people are fighting over the near-term impacts vs keeping the long term vision in mind 🤦🏾♂️🤷🏾♂️
Every net zero pathway involved >2x’ing the electricity system and massive amounts of transmission infrastructure investment
You see this on the power plant development side too. Here's a quote from the NRG earnings call two days ago.
People who have been in the industry for more than a decade remember the speculative build in the 2000s.
They're just not interested in taking that risk.
i think the more honest answer is that it's hard to build "normal" plants because there's no access to large turbines. and there's no access to large turbines because OEMs are nervous about the durability of demand.
GE ramped capacity and bought alstom at the 2014–15 peak during the post shale boom, expecting 120+ GW/year of sustained demand. within 3 years, global orders had halved. this was a heavy hit on the power division, w/ mass layoffs, CEO leaving, and split into GE Vernova, etc. this is a famous industrial case study L
that scar tissue still shapes the industry today. they are expanding production, but obviously cautious.
Anybody else find it ironic that this is the final bullet of the energy plan?
“Return PJM to Market Fundamentals”
Full plan here:
energy.gov/documents/stateme…
1
1
181
A lot of discussion on whether data centers are willing to curtail given their high opportunity costs.
Seems like different regions are moving towards it being a requirement rather than a choice.
Texas had SB6 and now PJM is considering mandatory interruptibility.
Great commentary from Constellation and great find from @ShanuMathew93:
Gas plants want long term fixed pricing. But it’s hard to do because of uncertainty around fuel expense.
This isn’t zero marginal cost renewables where you can easily sign a 15-20 year PPA.
Replying to @ShanuMathew93
"But the other thing they want is they want certainty on pricing for a 20-year deal. And that's pretty hard to do with the gas plant because you don't know the input price of the natural gas. And so it's always going to be indexed to a certain extent. And it's very difficult to do because you don't know what future environmental policies might mean to the cost of CO2. So typically, when you're looking at a long-term deal at a gas plant, you're saying, hey, it's going to be indexed to something on the supply side or we need a long-term supply arrangement. You need to deal with the firmness of that supply arrangement and then you have to have a conversation with the customer around what happens if a carbon tax or a regional program comes into play or something else happens."
Following up on an interesting discussion over the weekend.
PJM's capacity market, the "Base Residual Auction", was designed to be a residual market.
Here's what PJM's CEO said on the topic.
“The capacity market was not intended to incentivize long-term capital deployment.” – PJM CEO
With PJM’s auction results dropping tomorrow, it's worth asking:
What is the purpose of PJM's capacity market? If not to bring new projects online?
Everyone thinks PJM’s capacity market is supposed to incentivize new generation.
It’s not. PJM’s own CEO has said so.
Details below:
balancingauthority.substack.…





















