As Data Center Hate Grows, Plans for Gas to Power them are Uncertain

Above, new PBS reporting on how AI Data Centers have become a campaign issue.

Below, the numbers of new gas powered plants needed to power data centers are daunting, but are they real?

Axios:

Stunning stat: The amount of U.S. capacity somewhere in the development pipeline doubled in the first half of 2026, per Global Energy Monitor’s analysis.

• A whopping 189 gigawatts is planned for on-site data centers through the first half of 2026, roughly doubling (!) the pipeline at year-end 2025.

Driving the news: Overall, the U.S. pipeline has soared to 378 gigawatts, per the nonprofit that closely tracks projects internationally.

• More concretely, projects under construction climbed 76% in the first half of 2026 to 52 GW, around twice what China is currently building.
• A gigawatt can power roughly 750,000 U.S. homes.

Reality check: If all the 378 GW on the drawing board become reality, the U.S. fleet would grow by two-thirds. 

• But the report tracks everything from construction to projects in much earlier stages that may or may not ultimately happen.
• “Pre-construction” and “announced” projects together account for about 86% of the U.S. pipeline. 

State of play: The report highlights big questions. For instance, two-thirds of the capacity in development globally hasn’t yet named a turbine or engine manufacturer.

Global Energy Monitor:

Projects under construction in the U.S. rose 76% in the first half of the year, reaching 52 GW, of which 16.9 GW is for projects intended to directly power data centers. The U.S. is now building twice as much gas-fired capacity as China (24 GW) and has nearly three times its gas-fired power capacity in development

While the growth in gas power proposals is being driven largely by data centers, more than three-quarters of the global pipeline is still in early-stage development. In the first six months of the year, roughly 45 GW of announced and pre-construction capacity had its planned start year delayed. 

Two-thirds of gas-fired capacity in development globally, and more than half of projects tied to data centers, do not have a named turbine/engine manufacturer. Nearly one-quarter of projects earmarked for data centers do not have a named start year.

Turbine supply constraints, financing uncertainty, local data center moratoriums, and mounting public opposition leave the true scale of the gas power buildout uncertain.

Jenny Martos, project manager for the Global Oil and Gas Plant Tracker at Global Energy Monitor, said,

This frenzy of gas proposals  is running headlong into the hurdles of an already tight gas market. It is nearly impossible nowadays to guess what is a pie in the sky proposal, and what has a real chance of getting built. The projects that eventually clear those hurdles are paying top dollar for turbines, locking in emissions, and pushing up electricity prices.” 

Michael Thomas on X:

Amazon recently began construction on one of the largest solar and storage projects in America—a 1.2 GW / 7.2 GWh project in Eastern Oregon.

So why is the company turning to renewables instead of gas in the Pacific Northwest?

Over the last few weeks, Cleanview’s team reviewed hundreds of permits, corporate filings, and other public records to find an answer to that question.

We learned that the company had no other option.

Between 2013 and 2024, Amazon built 40 data centers in Eastern Oregon with a combined power capacity of 3.3 GW, becoming by far the region’s largest consumer of electricity.

Amazon’s primary utility in the region was never built to serve this kind of demand. The company’s data centers now likely use 50 times more electricity than homes and farms in the area.

Over the last few years, utilities haven’t been able to keep up with this demand growth. In multiple complaints to local regulators, Amazon has said that utilities aren’t delivering enough power to its data centers.

Elsewhere in the country, hyperscalers facing grid delays like this have built their own natural gas power plants.

Earlier this month, I was the first to report that Amazon is turning to this strategy in Texas with a 7.65 GW off-grid gas project.

But in Oregon, public policy prevents such a strategy: HB 2021 effectively bars companies from building new fossil fuel power plants, whether they’re a utility or hyperscaler.

It was in this context that Amazon decided to buy and build Sunstone Solar and Storage.

After its original developer went bankrupt, Amazon outbid a utility to buy the project with its shovel-ready permits and interconnection agreement.

In both Texas and Oregon, Amazon has an insatiable demand for electricity. The difference in how it gets that electricity comes down to policy.