Hyperscalers are beginning to show interest in pairing gas with carbon capture and sequestration, says Julio Friedmann, chief scientist at consultancy Carbon Direct. Data center companies are trying to meet urgent power needs while keeping their climate commitments, and gas-plus-CC would not only be firm and low-carbon, but also relatively quick-to-deploy, he said, speaking on a recent episode of Catalyst.
“We’ve learned the top five things that the hyperscalers and people building data centers want [are] in this order: Speed, speed, speed, cost and carbon,” Friedmann told host Shayle Kann. “All of those are important, but speed is the thing that is driving a lot of this.”
Gas-plus-CCS has never deployed for utility-scale power generation; the world’s first few plants are slated to start operations this decade. And of those plants, no hyperscalers have announced involvement so far — but Friedmann anticipates they will soon. For example, he has “every expectation” that the planned 2.23-gigawatt Entergy gas project near Shreveport, Louisiana will use CCS. (Local news organization the Shreveport Times reported in November that the plant will likely serve a massive Meta data center.)
“We are getting a lot of inbound on this exact topic and we are talking to, if not all of the hyperscalers, almost all of the hyperscalers,” he said.
In particular, he outlined a two-part approach for speed: build capture-ready gas-fired plants, and then integrate CCS in parallel or shortly after. “You can build a gas plant fast and then you have another couple of years to build the CCS retrofit and over that time you’re generating power and you’re generating money,” he said.”
Listen to Julio Friedmann’s whole interview on Catalyst:
In a recent Carbon Direct white paper, Friedmann and his co-authors estimated that gas-fired generation can be designed and built in 18 months. Integrating CCS adds 18 to 36 months, if all three stages of planning — carbon capture, transportation, and sequestration — happen simultaneously. For hyperscalers willing to pay a premium for clean, firm power, it could have advantages over the alternatives, Friedmann explained.
“They can build a natural-gas-with-CCS plant in roughly half the time which they could build a nuclear plant,” he said.
But recent challenges in the gas market — a recent string of project cancellations and equipment backlogs, for instance — may complicate this picture. And there’s uncertainty around both the U.S. tax credit for carbon capture (45Q), and the permitting of carbon dioxide wells and pipeline infrastructure. Friedmann also acknowledged that building capture-ready gas-fired generation comes with a risk of not following through on actually deploying CCS.
“We have to move from capture-ready to capture-committed,” he said.
Carbon capture has long been used in industrial-gas production and enhanced oil recovery, but CCS failed to scale on coal-fired power plants due to cost and technical challenges. That said, Friedmann thinks the economics of CCS for gas are surprisingly attractive. Adding CCS to gas-fired generation costs much less per megawatt hour than adding it to coal, despite gas’s lower flue concentrations of CO2. The reason is that coal emits more carbon dioxide per unit of energy than gas does. In short, the LCOE of gas-plus-CCS ends up in a competitive range, if gas itself remains cheap.
“In the U.S., assuming $3 per million BTU for natural gas on a new build — all costs: that means cost of capture, compression, transportation, storage and permitting — you’re looking at $70 to $100 a megawatt hour,” he said. “In many markets, that’s competitive with wind, solar and batteries. In many markets that just beats nuclear cold.”
He added that retrofitting an existing plant is even cheaper, closer to $40 to $70 per megawatt-hour. That’s a range comparable to the LCOE of renewables-plus-storage, according to Lazard.
Many projects are already betting that these economics will work out. The 742-megawatt Net Zero Teesside Power project in Yorkshire, England — a collaboration between BP, Equinor, and TotalEnergies — aims to be the world’s first, with the goal to begin operations in 2028. There’s also Exxon’s 1.5-GW grid-islanded project in a to-be-announced location, built exclusively for a data center by 2029; and NET Power’s 370MW plant in Texas is planning to begin operations in 2029.
But Friedmann offered another reason for the industry to explore CCS, regardless of any qualms: Especially In light of load growth, utilities are expanding their gas operations, CCS or no CCS.
“We are building the gas plants anyway. They’re being built,” he said. “So it’s not a question of do you do A versus B. We’re already doing A; we’re just emitting uncontrolled [CO2].”
