Texas cuts the cost of holding a power queue position

The Public Utility Commission of Texas voted Friday to make a place in the state's power queue cheaper to hold, scrapping a $50,000-per-megawatt non-refundable fee proposed in March, halving the financial security required of loads with peak demand over 75 MW from $100,000 per megawatt to $50,000, and replacing the proposal's size-tiered study charges with a flat $100,000 for every large load customer. A 100 MW campus that would have paid $5 million in fees and posted $10 million in security now posts $5 million and pays no interconnection fee at all.

That $10 million swing in upfront capital arrives in the middle of the state's open question about electricity: roughly 90% of ERCOT's 474 GW interconnection queue is data centers, and the state is auditing that queue while holding a pause on new data center development. The commission has softened the terms of entry in the same season it set out to establish how much of the queue is real; none of this shortens the line or adds supply, but it reprices the act of waiting.

The March proposal drew a pointed objection on comparative grounds: DLA Piper's blog post that month argued the rules could result in significant upfront capital commitments, and that their financial thresholds ran higher than those imposed by other major US grid operators, where study deposits and interconnection fees for load customers are typically measured in the tens of thousands of dollars rather than the millions. The commission removed the fee that most plainly produced the millions and kept a study charge, settling on a flat $100,000 for all large load customers in place of the proposal's tiers, and wrote that it may amend the amount as more data relating to study costs becomes available. Even after the retreat, that study fee sits above the tens-of-thousands band DLA Piper described for peer operators.

The commission also preserved a distinction worth watching: the eliminated interconnection fee was assessed per megawatt of contracted peak demand, while the security posting is measured against requested peak demand. With the fee gone, the deposit is the only charge that still scales with what a developer asks for rather than with what it has committed to.

The change with the longest tail is temporal rather than financial. Under the proposal, the interconnecting distribution or transmission service provider had to notify ERCOT when a large load customer missed a scheduled energization milestone by six months, with that notice due no less than 30 days after the miss. The adopted rule sets the trigger at 24 months and specifies that the period applies to the energization schedule as a whole rather than to each milestone on its own, so a campus with a dozen milestones gets one clock and the utility cannot start the countdown against a single slipped date while the rest of the schedule holds.

The deposit mechanics move the same way: the proposal would have returned 20% of the posted security to the customer and applied the remaining 80% to outstanding amounts owed, whereas the adopted rule applies the security to what is owed, then returns the balance, and requires the utility to make that return within 60 days of notifying ERCOT. The money stays the developer's unless the developer owes, making the posting behave more like offset collateral than a forfeitable penalty, and the recoverable share no longer starts at 20%.

Texas did not make interconnection faster or cheaper to supply; it made the queue cheaper to occupy. Power rights have become a distinct asset class, as this publication has argued, with queue positions, retrofit rights and transformer slots trading before the electrons do. A rule that lowers the carrying cost of a position without adding a mechanism to separate a real project from a reservation improves the economics of holding and worsens the information available to anyone pricing what is held, and the commission appears to have decided that the December audit, not the tariff, is the right instrument for the sorting.

Read generously, the rule is a state declining to charge for the thing it is trying to attract, and the audit is the correction. Read less generously, the commission has made waiting cheaper in a market where the size of the queue makes the distinction consequential. Both readings land on the same wager: that Texas can count 474 GW of requests accurately enough in December to make the fee schedule beside the point.

The fee schedule is now settled enough to underwrite against; the queue is not. The commission's note that it may revisit the $100,000 study fee once it knows what study costs leaves one number open, and the December filing leaves the larger one open. A developer weighing whether to carry a position into that audit has the arithmetic in front of it: $50,000 per megawatt posted, no interconnection fee, and 24 months before anyone must tell ERCOT that the capacity is going unused.