Texas Pauses 49.8 GW of Data Center Grid Connections; the Real Question Is What Gets Built When the Audit Ends

BloombergNEF reports that Gov. Abbott's August 3 directive to audit all data centers in ERCOT's interconnection queue affects approximately 49.8 GW of projects, nearly one-fifth of the entire U.S. data center development pipeline.^[1] The pause could cost developers as much as $15 billion by the first quarter of 2027 if AI-compute capacity dominates the delayed projects.^[1] But the audit itself, while overdue, does not address the core scandal: Texas ratepayers may be forced to fund generation and transmission upgrades for loads that exist only as redacted special contracts or abandoned pipeline announcements.

Here is what Abbott's audit is supposed to unearth. ERCOT's queue contains approximately 474 GW of pending requests, roughly 90% of them data centers.^[7] That is more than five times Texas' record peak electricity demand.^[7] That number alone tells you the queue is not a forecast of actual build; it is a wish list from which many projects will vanish. The question the audit must answer, and the one the state's regulated utilities must face in every rate case going forward, is whether any of these loads have binding contracts backed by collateral and high minimum-take ratchets, or whether they are merely "announced" projects shopping across multiple utilities and the same interconnection request duplicated in three different regional queues to hedge bets. Grid Strategies and others have documented that U.S. load forecasts for large-load interconnections have quintupled in recent years against prior vintages, a revision so large it screams of double-counting or baseless speculation.

The political and financial stakes are stark. If the audit clears, say, 100 GW for development over the next decade, who pays for the generation, transmission, and grid reinforcement those projects will trigger? Under Texas's current tariff structure, especially for special contracts negotiated in secret between utilities and hyperscalers, the answer is: ratepayers absorb the cost, while the contract's price, term, and minimum-take obligation remain sealed. A utility can build a $2 billion gas plant or transmission upgrade and recover its cost from the entire customer base, even if the hyperscaler's load never reaches the contracted level or the contract ends early. The ratchet, if it exists at all, may allow the customer to pay for only 60% of unused generation capacity, leaving residents and small businesses to cover the rest. This is not a hypothetical: it is the architecture Harvard's Electricity Law Initiative documented in its analysis of data-center tariffs across multiple states, where existing rate structures extract monopoly rents from the public to serve big tech.

Texas does have a new large-load interconnection process designed to isolate risk and shift responsibility to the customer class driving the load.^[9] The audit's credibility will hinge on whether the PUCT uses it to require cost-isolation, collateral, and long-term minimum-take commitments of 80%+ of contracted capacity, or whether it rubber-stamps special contracts with weak ratchets and socialized stranded-cost risk. Abbott's concern for "Texans' safety and quality of life" is sound, but it will ring hollow if the audit approves new capacity without requiring developers to bring their own generation, accept flexible or curtailable grid connections, or commit to additionality (i.e., prove the load is new and not migrated from elsewhere in ERCOT).

The audit also creates a window. Between now and the end of the process, the PUCT can propose a protective tariff for all large loads, modeled on Virginia's GS-5 or Ohio's AEP framework: a standing customer class with long minimum terms tied to asset life, high demand ratchets, collateral covering unamortized investment, and full cost responsibility for dedicated upgrades. Such a tariff would transform the rate-case dynamic: instead of utilities proposing capex justified by inflated load forecasts, developers would bid into a known framework and ratepayers would know they are not carrying the risk. The state legislature, reconvening in January 2027, could codify the tariff and prohibit special contracts unless they meet it.

The danger is that the audit becomes a rubber stamp. If ERCOT and the PUCT emerge to report that the queue is "legitimate" and projects can proceed, but do so without enforcing cost-isolation, collateral, and additionality, then the pause will have bought time for the utilities' lobbyists to rewrite the rules, not for the public to reclaim them. Ratepayers should demand that Abbott's audit be paired with a concrete protective tariff, filed and opened to public comment, before a single new data center gets grid approval.