Energy Transfer Just Delayed a New Mexico Pipeline. Here's Why I'm Not Worried Yet.

Energy Transfer postponed the Green Chile Project's planned in-service date by six months, from August 2026 to February 2027, following regulatory setbacks, land-rights disputes, and loss of expedited federal review status. The 18-mile pipeline was intended to provide up to 400 million cubic feet per day of gas for Bloom Energy fuel cells supporting Oracle and OpenAI's Project Jupiter AI data-center campus. The delay is limited relative to Energy Transfer's 140,000-mile network and more than 32 billion cubic feet per day of gas throughput, suggesting modest company-level financial impact.

Analysis

The economic exposure is asymmetric: ET's downside is likely limited to deferred project-level EBITDA and incremental legal/carrying costs, while BE faces a potentially more meaningful risk if its contracted deployment depends on the delayed gas interconnect. The key diligence item is whether BE has alternative fuel-delivery, grid-interconnection, or temporary generation solutions; without one, revenue recognition and working-capital conversion tied to the campus could slip by multiple quarters. ET's diversified fee base should absorb a single-project delay, but repeated permitting friction would raise the market's required return for future growth capital rather than impair current distributable cash flow.

For ORCL, the relevant read-through is not gas supply itself but whether power availability becomes the binding constraint on AI-capacity commissioning. A delay at one site does not change aggregate cloud demand, but it can shift workload deployment toward regions with existing utility capacity, benefiting power-rich data-center markets and disadvantaging suppliers concentrated in constrained Southwest buildouts. Over 6-18 months, this reinforces scarcity value for dispatchable power and gas infrastructure, although the near-term equity impact is too small to alter ET or ORCL estimates absent evidence that the campus is on the critical path for contracted OCI capacity.

Consensus is likely correct that ET is insulated, but may underappreciate the option value lost from delayed AI-linked infrastructure: permitting risk reduces the premium investors assign to new-project backlog, particularly if management must accept lower returns or greater customer concessions to complete it. Conversely, a rapid legal resolution could create a modest positive catalyst for ET because the market appears unlikely to capitalize the project today. The thesis is falsified by material ET guidance cuts, evidence of cost overruns or contract termination, or BE disclosing that deployment timing and backlog conversion are unaffected.

Key Decisions for Investors

• No standalone ET trade on this development; maintain existing exposure only if ET's next guidance confirms unchanged growth-capex budget, distribution coverage, and project-return targets. Reassess if management identifies more than immaterial EBITDA deferral or incremental capital requirements.
• Use BE as the higher-beta watch item: avoid adding ahead of confirmation of a fuel-supply workaround or revised commissioning schedule. A disclosed deployment deferral would create downside to backlog-to-revenue assumptions over the next 1-3 quarters; confirmed alternative supply would remove the near-term overhang.
• For AI-power exposure, prefer a diversified basket such as long VST or CEG versus BE rather than extrapolating a site-specific constraint into broad AI demand weakness. The structural catalyst is multi-year scarcity of firm power, while the key risk is lower wholesale power pricing, regulatory intervention, or materially slower data-center load additions.
• Set an event alert for permitting or land-rights resolution within the next 3-6 months. A definitive path to construction is a better ET entry catalyst than the current delay, while a further schedule extension or customer contract amendment would warrant reducing any ET position sized for growth-project upside.