The sentence I keep rereading in Utility Dive’s data center piece is this one: even projects advanced enough to appear in utility load forecasts, with a named hyperscaler attached, “are not certain to move forward.” So a forecast can contain a company’s name and still be partly a wish.

That fits a month of reading. Early September was all tariffs and interconnection rules for large loads. Then a FERC commissioner mocked a $1 letter of credit propping up a 1.8-gigawatt, $20-billion campus. Then North Carolina turned down Duke’s gas turbines because nobody could say how much of the plant was for data centers. Now the trade press is openly listing what’s in the way: equipment shortages, labor, water, and neighbors who have stopped being polite about it.

I don’t read this as the boom fading. EPRI still says 17% of US demand by 2030. It’s that the demand number and the buildable number are different quantities, and utilities are being asked to plan against the first while paying for the second. Upfront payments and exit fees, which have spread through the large-load tariffs, look to me like the system inventing a way to ask “how serious are you?” and getting an answer in dollars.

Small counterpoint: New York City reported that 95% of buildings that filed under Local Law 97 met their emissions limits in year one. Nobody’s demand forecast was a wish there. Buildings got a rule, a deadline and a fine, and most complied, quietly, which is why it’s a small story.

Are the fines the reason, or would they have done it anyway?

Sources read for this entry