Louisiana Builds 2.88 GW Gas Plant for Data Center, but Who Bears the Stranded-Capacity Risk?

ONE Nuclear Energy has executed a binding letter of intent for Project Cayman, a 2.88-GW natural gas plant co-located with a data center campus in Ascension Parish, Louisiana, near the RiverPlex MegaPark^[1]. The company frames it as alignment with Entergy's regional transmission buildout. But the press releases omit the essential question: what is actually contracted, and who eats the cost if the data center load never materializes?

The secrecy begins immediately. No announcement discloses the data center operator's name, the contracted megawatts of demand, the term of the power purchase agreement, the minimum-take or demand ratchet, or even ONE Nuclear's capital cost^[7]. This opacity is not incidental; it is the mechanism. When a special contract between a utility (or contracted private developer) and a hyperscaler stays confidential while the generation and transmission assets are built into rate base, ratepayers bear the stranded-capacity risk that the developer and utility do not. The Harvard Electricity Law Initiative's examination of data-center tariffs found this structure across the sector: utilities extract public capital for assets nominally for one customer, then allocate cost to all ratepayers if the load underperforms or never arrives.

The critical facts to extract from Entergy and ONE Nuclear's filings: (1) Is the 2.88 GW contracted to a named customer with a long-term, high minimum-take ratchet (e.g., 80, 90% of contracted capacity over 10+ years)? If not, any shortfall in realized demand becomes a system cost. (2) What is the term of the power contract relative to the 40-year life of the gas plant? A 10, 15 year deal on a decade-long asset leaves 25+ years of unrecovered investment assigned to the rate base. (3) Are Entergy's transmission upgrades supporting Project Cayman cost-isolated to a large-load customer class (as in Virginia's GS-5 or Ohio's recent AEP protocols), or socialized across all rate classes? (4) Is there a collateral requirement and exit fee covering unamortized investment if the data center cancels or relocates? Without these, the cost stays with residential and small-commercial customers.

The data-center pipeline itself warrants skepticism. Project Cayman exists in a context of inflated load forecasts. Utilities nationwide have nearly quintupled their five-year data-center demand projections against 2022 baselines, and independent audits have found the announced capacity figures systematically double-count identical projects shopped across multiple utility territories. Entergy will need to disclose in its next integrated resource plan how much of this 2.88-GW need is contracted (binding with collateral) versus announced or speculative. Historical realization rates matter: if Entergy's own prior forecasts overestimated by 20, 30%, this capacity could sit underused for years.

There is a buildable alternative that Entergy has not disclosed considering: flexible or curtailable interconnection. Independent research from Duke and Princeton found that the US grid can absorb 76, 126 GW of new data-center load if the load agrees to curtail during system peaks, sparing ratepayers roughly $764M per GW in firm capacity costs. ONE Nuclear could offer Project Cayman a three to five-year faster energization path via bring-your-own-capacity and a flexible grid connection, protecting both the data center's timeline and ratepayers' wallets. The fact that this option is not mentioned in any press release is itself evidence.

The window to intervene is now. Entergy will file this project in rate cases and transmission-upgrade dockets over the next 12, 24 months. Ratepayers and consumer advocates should demand, by name: (1) disclosure of the data center customer, the contracted MW, and the minimum-take ratchet (redact the price if you must, not the risk allocation); (2) a separate large-load customer class with cost isolation and 100% responsibility for dedicated generation and transmission; (3) evidence that a flexible-load or bring-your-own-capacity option was studied and rejected, and on what basis; (4) collateral and exit fees covering unamortized investment. Without these, Louisiana is socializing the risk of a bet on a single customer's load growth, and that bet is not transparent.