Texas is quickly turning into a major hub for AI data centers. However, using large natural gas plants to power these complexes is causing concern over emissions and energy use.

Texas: The New AI Powerhouse

If you drive through West Texas, you’ll notice empty stretches of land suddenly buzzing with plans for gigantic data center campuses. Google, Microsoft, and other tech giants are racing to build AI infrastructure here, betting on Texas’s abundant natural gas supplies to fuel their ambitions.

Google’s Goodnight campus in Armstrong County is a prime example. It’s set to include a private natural gas plant alongside wind power, aiming to generate more than 900 megawatts from gas turbines alone. According to the Texas state air permit application, this gas plant could emit over 4.5 million tons of greenhouse gases yearly — that’s roughly equivalent to adding nearly a million gas-powered cars on the road.

Microsoft is also striking deals to secure natural gas power for its AI data centers. Partnering with Chevron and investment firm Engine No. 1, it’s linked to plans for a $7 billion gas-fired power plant near West Texas, designed to deliver 2,500 megawatts of electricity. The data center campus it will serve was initially developed for Oracle and OpenAI but now falls under Microsoft’s wing.

Why Natural Gas?

Developers choose natural gas since Texas has plenty of it and it can supply the constant, large amounts of power AI data centers need.

These centers run around the clock, processing complex AI models that consume as much electricity as whole cities.

Mastheads

But connecting to the main electric grid is a slow process with long wait times. To beat that, many companies build private power plants — what’s known as behind-the-meter power. It means they generate electricity onsite, sidestepping grid delays and rising consumer costs.

This approach is driving a boom in natural gas plant building throughout Texas. Chevron, ExxonMobil, and Diamondback Energy are all jumping in, planning gas-fired plants specifically to power data centers. The Electric Reliability Council of Texas (ERCOT) projects about 24 gigawatts of new data center power demand by 2031 — roughly the equivalent of adding another Houston metro area's power usage to the grid.

Environmental and Community Concerns

The emissions from these gas plants are significant and can’t be overlooked. Google’s Goodnight campus alone would exceed the pollution of a typical coal plant. This makes people wonder about how “green” Big Tech’s AI push really is — especially as companies publicly commit to renewable energy goals.

Michael Thomas from Cleanview points out that while Google highlights its renewable efforts, its move toward private fossil fuel power suggests a shift as AI heats up. The race to build AI infrastructure might be pushing companies to prioritize speed and reliability over sustainability.

People in local communities are concerned as well. The vast power needs and industrial scale of these projects strain local resources. Water usage, emissions, and the sheer footprint of infrastructure raise concerns about environmental justice and quality of life in these often remote Texas towns.

Looking Ahead: The AI Energy Puzzle

Big Tech isn’t slowing down its AI plans. The $40 billion AI investment Google announced in Texas signals more data centers and, likely, more fossil fuel power plants. Microsoft’s move to rent and power a large Texas data center campus reinforces this trend.

But these private gas plants haven’t fully connected to the Texas grid yet. The hope is that eventually, they’ll connect to wider renewable sources. For now, though, the energy landscape is shifting toward a patchwork of private grids powered largely by natural gas.

This brings up a key question: Can AI’s rapid growth fit with climate goals when it depends so much on carbon-heavy power? The answer may shape not just Texas’s energy future, but the global AI race itself.

Related Articles

Texas could soon lead the world in AI data centers — and the natural gas plants powering them could leave a lasting mark on emissions and energy infrastructure.

This article was created with AI assistance.