Hyperion

• 01

  5.5/ 10

  Power delivery (~5 GW via 10 gas plants + 240 mi 500-kV + storage by 2030)

  Meaningful but qualified

  Credible as a phased utility-led buildout; not yet substantiated as a full 2030 delivery claim — Hyperion has the right institutional actors but still needs public proof of final execution.
  5.5/ 10
• 02

  5.0/ 10

  Timeline (full ~5 GW operational by 2030)

  Preliminary to meaningful

  Phased operations before 2030 are plausible; full ~5 GW completion by 2030 is plausible but unproven absent phase-by-phase energization, permit, and commissioning schedules.
  5.0/ 10
• 03

  4.0/ 10

  Water sustainability (water-positive at multi-GW scale)

  Meaningful but qualified

  Water-positive is an outcome claim. 5 GW IT screens at 13–38 MGD against 2014 Richland Parish withdrawals of 41.73 MGD — material until a basin-specific water balance is public.
  4.0/ 10
• 04

  4.0/ 10

  Renewable / clean-energy matching (100% despite gas-heavy supply)

  Preliminary

  Directionally possible but under-evidenced. At Hyperion scale, the credible proof is executed procurement, matching methodology, deliverability, and emissions accounting — not a corporate pledge.
  4.0/ 10
• 05

  5.0/ 10

  Economic & financing viability (~$27B-plus all-in)

  Preliminary to meaningful

  Strategically rational for Meta even if a stand-alone infrastructure model looks weak; not substantiated as economically robust at full scale on the public record.
  5.0/ 10
• 06

  4.5/ 10

  Community & political durability

  Meaningful but qualified

  Strong state support; visible social-license exposure from reporting on scale, incentives, the Morris land-deal allegations, and construction-phase impacts. Not likely to cancel near term — can affect timing, conditions, and incentives.
  4.5/ 10
• 07

  6.0/ 10

  Ratepayer protection (Louisiana ratepayers insulated)

  Meaningful but qualified

  Strongest process claim. 'Meta-funded' is not the same as fully insulated — enforceable LPSC tariff, service, collateral, overrun, and stranded-cost protections are not yet binding in public.
  6.0/ 10
• 08

  4.5/ 10

  Workforce & local capacity (5,000+ peak, 500+ operating)

  Preliminary

  Plausible as a regional construction mobilization; local-benefit capture is weaker because the record does not yet show enforceable Richland Parish jobs, wages, supplier-spend, and clawback commitments.
  4.5/ 10

1. 01

   Gas generation creates both reliability and credibility tension

   New gas-fired plants make the reliability story more credible than a pure renewables-only claim. They also make the clean-energy narrative harder to prove. A gas-backed campus can still procure clean energy, but the proof burden shifts to matching, deliverability, additionality, and emissions accounting.

2. 02

   Water is under-specified for the scale

   At 5 GW, even efficient cooling can be locally significant. The missing documents are not peripheral; they are the evidence needed to score the water claim above weak: source, volume, permit, seasonal reliability, discharge, and verified restoration accounting against historical Richland Parish withdrawals.

3. 03

   The LPSC process can reduce or amplify risk

   A clear, enforceable LPSC structure can protect non-participating ratepayers and improve project legitimacy. An opaque structure can do the opposite by turning Hyperion into a public debate over whether Louisiana households and businesses are subsidizing Big Tech's load growth.

4. 04

   Construction logistics are not a footnote

   Flat land helps. Poor drainage, floodplain/wetland proximity, heavy-haul needs, and rural road capacity complicate delivery. Heavy transformers, gas-generation components, 500-kV equipment, and multi-year truck traffic can become schedule and community risks.

5. 05

   Incentives and ethics can reframe the project

   The reported ~$3.3 billion tax-break figure, if verified in full project-specific terms, is politically material. The reported Sen. John 'Jay' Morris land-deal issue should be treated as alleged or reported concerns rather than a legal conclusion — but it is material because it can change the public story from 'jobs and investment' to 'public concessions and insider benefit.'

6. 06

   AI demand uncertainty affects financing

   A large AI campus is not just a utility load; it is a bet on sustained model-training and inference demand. Hyperion's feasibility improves if Meta fills the campus with high-value workloads. It weakens if AI hardware cycles, model-efficiency gains, utilization, or monetization underperform after the capital stack is committed.

7. 07

   Cost escalation and financing fragility

   The shift from a $10B announcement to a ~$27B-plus all-in narrative is not fatal, but it raises the proof bar for capex control, incentives, financing terms, and utility economics. An illustrative model at $27B capex, 8.75% WACC, and $3B/yr net operating value produces ~–$4.6B NPV and 5.9% IRR; a one-year slip worsens NPV by another ~$2.1B.

Project

Meta Hyperion (internally Project Sucre) — flagship AI data center

Stage tag

Under Construction (phased)

Location

Richland Parish, Louisiana (Holly Ridge / Rayville)

Sponsor

Meta Platforms · financing involvement reported from Blue Owl Capital

Original announcement (Dec 2024)

~$10B · 4M sq ft · over 2 GW

Expanded scope (2026)

~3,650+ acres · ~5 GW compute · ~$27B-plus all-in with power

Power plan

10 gas-fired plants/units (3 original combined-cycle + 7 added) totaling >5.2 GW

Transmission

~240 miles of 500-kV transmission + battery storage

Utility partner

Entergy Louisiana (LPSC oversight) · new 500-kV substation at Sterlington · new line near Sarepta to near Mount Olive

First tranche capacity

2,133 MW summer baseload already in LPSC planning

General contractors

Mortenson · DPR Construction · Turner Construction

Construction impacts

5,000+ peak construction workers · 500+ permanent jobs · 1,000+ indirect

Scoring arithmetic

Claim mean 4.8125 → 4.8 · controlling claim 5.5 · cap 7.0 (does not bind)

Evidence as of

May 29, 2026

Companion reports — Power & Grid Path, Permitting & Litigation, Community Impact, and Schedule Forensics — are produced on request.