Meta Will Fund 10 Gas Plants and 240 Miles of Transmission Lines for a Single AI Campus. The Bill Is $11 Billion.
On March 27, Entergy Louisiana announced that Meta will fund the construction of seven new natural gas power plants for its Hyperion AI data center campus in Richland Parish, Louisiana. Combined with three plants that regulators approved in August 2025, Meta now has 10 gas-fired plants in its pipeline, delivering 7.5 gigawatts of capacity, enough to power more than 5 million homes and representing a more than 30% increase to Louisiana’s entire grid capacity.
The 10 plants are estimated to cost nearly $11 billion. Meta will also fund up to 2.5 gigawatts of renewable energy capacity including battery storage, 240 miles of new transmission lines connecting South Louisiana to North Louisiana and Arkansas, battery energy storage systems, and nuclear power uprates at existing Entergy facilities. Meta CEO Mark Zuckerberg has said Hyperion would cover “a significant part of the footprint of Manhattan.”
Entergy stock rose 7% on the announcement, reaching a market cap of approximately $50 billion, up nearly 125% over the past two years.
The Deal Structure Matters
The original article correctly identified this as a pivotal moment where hyperscalers become quasi-utilities. The details confirm that framing.
Meta is not just buying electricity. It’s financing the entire power generation, transmission, and storage infrastructure for a single campus. Tom’s Hardware reported that the deal is structured so Meta “pays its full cost of service,” and Entergy projects the agreement will deliver more than $2 billion in customer savings over 20 years. Meta declined to disclose its total spend.
This structure is significant for the ratepayer risk debate. The contractual terms are 15 years. Critics contend ratepayers could be stuck with the bill if Meta no longer requires the power after that span. Entergy argues the opposite: Meta financing the plants protects ratepayers. The Louisiana Public Service Commission approved the first three plants in 2025 despite pushback from environmental groups. The seven new plants require fresh LPSC approval.
How Hyperion Got This Big
Meta initially announced a $10 billion investment in December 2024 for a 2,250-acre campus. Fortune reported in February that Meta quietly acquired an additional 1,400 acres. In October 2025, Meta entered a joint venture with Blue Owl Capital to finance, build, and operate the campus with up to $27 billion in total development costs. The site is now the size of 2,700 football fields.
The expansion from 2.3 GW (three plants) to 7.5 GW (ten plants) in under a year reflects the speed at which AI compute demand is outstripping power availability. The original article’s framing that “Meta’s load forecasts keep climbing as AI training demands explode” is confirmed by the trajectory: a 3x increase in power commitment in roughly 12 months.
Why Gas and What Comes Next
The original article correctly identified the logic: gas plants can be built and deployed faster than nuclear or renewable-plus-storage alternatives for 24/7 AI workloads. Entergy’s plan mixes peaker and baseload units. Meta’s pledge to fund 2.5 GW of renewables, battery storage, and nuclear uprates shows the company is hedging across multiple generation sources, but fossil fuels carry the near-term load because they’re the only technology that can deliver gigawatt-scale 24/7 power within the construction timeline.
The nuclear MOU the original mentioned is confirmed: Meta is funding uprates at existing Entergy nuclear facilities. Small modular reactors remain further out. The renewable commitment is meaningful at 2.5 GW but represents roughly a third of the gas capacity, not a replacement for it.
The Broader Pattern
This deal is consistent with a pattern documented across this article series. Meta’s $115-135 billion capex guidance for 2026, its commitment to $600 billion in data center construction through 2028, and its Ratepayer Protection Pledge signed at Trump’s State of the Union all point to the same conclusion: AI infrastructure now includes power generation as a core competency, not a procurement problem.
The original article cited Crusoe Energy’s deal for 29 GE Vernova aero-derivative turbines for more than 1 GW of AI capacity. This fits the pattern. Tom’s Hardware described a broader trend of data center developers building private natural gas “shadow grid” power plants to sidestep strained public grids. Meta’s Hyperion is the largest example, but the model is replicating across the hyperscaler landscape.
The Investment Implications
The original article’s thesis that the power ecosystem benefits from AI buildouts is confirmed and strengthened by this deal. 10 gas plants at $11 billion means turbine manufacturers, construction firms, and utilities with dedicated-build expertise all see direct demand.
Entergy’s 125% stock appreciation over two years and 7% jump on this announcement quantify what a hyperscaler-dedicated power deal is worth to a utility. Entergy Louisiana’s president said the agreement delivers “meaningful benefits to customers” while “keeping energy rates affordable,” but the LPSC has not yet approved the seven new plants. Regulatory approval is not guaranteed, especially given the scale: adding 30%+ to a state’s grid capacity for a single private customer is unprecedented.
The risks the original article underweighted: environmental opposition (which challenged the first three plants), the 15-year contract duration (what happens when Meta’s AI training needs evolve beyond this campus?), and political scrutiny (Louisiana regulators face pressure from both sides). The Lens noted that the first three plants were approved with costs borne by Entergy ratepayers, while the new seven are financed by Meta. This distinction matters: the financing structure directly affects who bears the risk if demand projections change.
For gas turbine manufacturers, the demand signal is unambiguous. For utilities positioned as dedicated builders for hyperscalers, the revenue visibility extends years. For investors, the key question is not whether AI needs power (it does) but how the regulatory and political landscape will shape who pays for it and who captures the margin.
If you liked this, you’ll love the daily dose we send to our Substack crew. It’s chill, it’s sharp, and it actually helps. Hop on here.
Read the full article here