Powering the AI Boom: How Liquid Carbon Dioxide Could Revolutionize Data Center Energy Efficiency

Executive Overview

The explosive global expansion of artificial intelligence, cloud computing, and digital infrastructure has created an unprecedented energy crisis. As technology giants race to secure reliable power for massive new data center builds, they are increasingly turning to on-site fossil-fuel generation—specifically, natural gas turbines—to bypass congested public electrical grids. While these power sources keep servers humming, they come with a catastrophic environmental cost. Many of these installations rely on simple-cycle gas turbines, which are notoriously inefficient and spew staggering volumes of greenhouse gases into the atmosphere.

Enter American Supercritical, a newly emerged startup breaking cover with $8 million in seed funding. The company aims to tackle this environmental bottleneck head-on through a novel approach: retrofitting dirty, inefficient simple-cycle gas turbines with liquid carbon dioxide power units. By harnessing the unique physical properties of "supercritical" carbon dioxide ($textsCO_2$), American Supercritical claims it can boost turbine efficiency by up to 50 percent without burning an extra molecule of natural gas or consuming additional water.

While the technology does not eliminate the underlying fossil fuel emissions from the gas turbines themselves, it acts as a crucial bridging mechanism. By dramatically squeezing more usable electricity out of waste exhaust heat, the startup’s modular units offer a glimpse into how engineers are attempting to wane the tech sector off its dirtiest power habits—even as the global race for AI dominance threatens to strain planetary boundaries to their absolute limits.

Detailed Chronology: From Lab Theory to Startup Emergence

The underlying science of utilizing supercritical carbon dioxide for power generation is not a sudden overnight breakthrough; rather, it is the culmination of decades of quiet, persistent research.

Decades in the National Labs

For more than 50 years, engineers and physicists within United States national laboratories have studied the thermodynamic potential of supercritical CO₂. When carbon dioxide is pressurized and held at a precise temperature threshold, it enters a "supercritical" state. In this unique phase, it achieves the density of a liquid while retaining the viscosity and transport properties of a gas. This dual personality allows $textsCO_2$ to transfer thermal energy far more efficiently through drastically smaller equipment compared to traditional steam-based systems.

Despite its theoretical promise, the concept spent decades trapped in academic papers and lab-scale experiments. The barriers to commercialization were formidable, encompassing everything from material science limitations under high-pressure conditions to the prohibitive costs of manufacturing specialized components.

The Turning Point: Component Innovation

Over the past twenty years, the tide began to turn, propelled by crucial technological breakthroughs in related engineering fields. According to Doug Hofer, an adviser to American Supercritical and a veteran turbine engineer with two decades of experience at General Electric, the most vital innovation was the development of more compact, affordable heat exchangers. These specialized devices—critical for both heating and cooling the pressurized $textsCO_2$ loop—finally made practical, scalable deployment economically feasible.

Stepping Out of Stealth

The culmination of this decades-long evolution arrived on a Wednesday when American Supercritical officially stepped out of stealth mode, announcing an $8 million funding round. Founded by Simon Shuham and Matthew Carlson—who spent over ten years researching supercritical CO₂—the company was formed to bridge the gap between heavy industrial power demands and advanced thermal physics.

Rather than building massive, centralized utility infrastructure from scratch, American Supercritical focused its sights on the immediate pain point: the booming market for localized, off-grid or semi-independent data center power generation. By packaging $textsCO_2$ technology into modular, deployable units, the company positioned itself at the intersection of two skyrocketing trends: the exponential energy demands of AI and the global push for industrial decarbonization.

Supporting Context & Metrics: The Scale of the Data Center Energy Crisis

To understand why American Supercritical’s technology is drawing intense interest from investors and engineers alike, one must examine the staggering mathematics of the modern data center boom.

The $7 Trillion Build-Out

According to forecasts by McKinsey & Company, the global data center build-out is projected to attract a staggering $7 trillion in capital investment by 2030. This unprecedented expansion is driven almost entirely by the computational demands of generative AI models, cloud storage, and automated enterprise infrastructure.

However, tech companies are running into a massive physical bottleneck: the electrical grid. Traditional utility interconnection queues are backlogged for years, forcing hyperscalers and developers to take power generation into their own hands. Increasingly, this means building on-site natural gas power plants to ensure uninterrupted uptime.

Simple-Cycle vs. Combined-Cycle Efficiency

The method by which these localized power plants generate electricity is a central driver of the current environmental crisis. Most large-scale, utility-grade natural gas plants utilize a combined-cycle process:

1. First Stage: Gas turbines burn compressed air and natural gas to spin generators and create primary electricity.
2. Second Stage: A secondary recovery engine captures the scorching exhaust heat, routing it to boil water and create high-pressure steam, which drives an additional steam turbine.

This dual-stage approach achieves an impressive thermal efficiency rate of roughly 60 to 65 percent.

By contrast, data centers operating under tight deployment timelines have frequently opted for simple-cycle turbines. These systems omit the steam-generation component entirely. Without a secondary mechanism to harvest waste heat, simple-cycle turbines are radically inefficient, converting only about 35 percent of the energy from natural gas into usable electricity. The remaining 65 percent escapes directly into the atmosphere as high-temperature exhaust, laden with greenhouse gases.

A Recipe for Climate Catastrophe

The sheer scale of these simple-cycle installations makes them an environmental ticking time bomb. Consider a massive data center power plant currently being constructed in Texas by Amazon. Permitted to run exclusively on simple-cycle turbines, this single facility is authorized to emit over 33 million tons of greenhouse gases annually—a carbon footprint that surpasses the yearly total emissions of numerous sovereign nations.

Furthermore, traditional steam-based combined-cycle plants require immense volumes of water for cooling and steam creation, drawing intense regulatory scrutiny and public backlash in drought-prone regions where data centers are frequently built.

Official Statements & Industry Perspectives

The path to commercializing breakthrough energy technologies is notoriously fraught with structural hurdles. Industry experts and company founders offer a clear-eyed view of both the immense opportunities and the formidable challenges ahead.

Building Miniature Combined-Cycle Plants

Simon Shuham, cofounder of American Supercritical, describes the company’s core innovation in straightforward terms:

"We’re essentially building miniature combined-cycle plants."

Instead of routing simple-cycle turbine exhaust through water boilers to create steam, American Supercritical’s retrofitted units capture that thermal energy directly via a closed-loop system filled with pressurized, supercritical carbon dioxide. The $textsCO_2$ absorbs the heat and drives an additional power-generation turbine with zero additional greenhouse gas emissions and no added water consumption.

Crucially, because the $textsCO_2$ operates in a closed loop, the working fluid does not need to be continually replenished. Matthew Carlson, cofounder and long-time $textsCO_2$ researcher, compares the mechanism to domestic refrigeration systems that continuously circulate a cooling fluid:

"We just need to get something hot, and then we can convert that into electricity."

The Innovator’s Dilemma

Despite the clear thermodynamic advantages, breaking into the entrenched energy market is notoriously difficult. Doug Hofer points to classic corporate inertia—often referred to as "the innovator’s dilemma":

"It’s the innovator’s dilemma, right? Getting this [technology] out there needs to have an outside influence. Larger companies are satisfied with the performance of traditional power plants and unwilling to invest in different technologies—and turbine makers are not interested in funding potential competitors."

Subith Vasu, a professor of engineering at the University of Central Florida who directs a lab at the Center for Advanced Turbomachinery and Energy Research, echoes these cautionary sentiments. While acknowledging the promise of supercritical CO₂, Vasu notes that commercial success requires a mature ecosystem:

"Whenever you are venturing into new technologies, you need to have all the associated components, the supply chain, in place. Technical challenges also abound, and cost has been a huge issue."

Future Outlook: Beyond Gas Turbines and Toward Modular Nuclear

While American Supercritical is launching its commercial journey with a relatively modest 10-megawatt unit designed to increase turbine efficiency by up to 50 percent, the long-term vision for the technology extends far beyond natural gas.

Alleviating the Grid Bottleneck

In the immediate term, the company hopes to capitalize on the tech industry’s desperate willingness to pay for power. By retrofitting existing fossil-fuel assets to squeeze more energy out of the same amount of fuel, American Supercritical aims to give hyperscale operators the breathing room they need to expand operations without tripping local emission caps or overwhelming regional grids.

"There’s this incredible willingness to pay [for power], but they’re bottlenecked," says Shuham. "The goal is to come in here, retrofit their existing systems, and give them the ability to build additional data centers—or get rid of some of the gas turbines."

Global Precedents and Multi-Industry Versatility

The deployment of supercritical CO₂ power systems is gaining international momentum. Earlier this year, China officially commissioned the world’s first geothermal heating plant utilizing supercritical CO₂, proving that the technology is viable outside of laboratory conditions.

Matthew Carlson notes that his career has spanned $textsCO_2$ applications across nuclear, geothermal, and solar energy sectors. Looking ahead, American Supercritical’s founders believe their modular power units can easily be adapted to work with small modular reactors (SMRs)—advanced nuclear fission systems that have garnered substantial bipartisan policy support and venture capital investment. In fact, the startup has already disclosed that it has signed an initial deployment agreement with an unnamed partner in the advanced energy sector.

The Lingering Shadow of Fossil Fuels

Despite the engineering elegance of capturing waste heat with liquid carbon dioxide, industry analysts emphasize a sobering reality: retrofitting inefficient turbines does not equal true zero-carbon energy.

Natural gas extraction, transportation, and combustion inherently warm the planet. Even if American Supercritical succeeds in boosting simple-cycle efficiency by 50 percent, data centers remain an enormous, rapidly expanding source of industrial pollution at a historical juncture when global climate frameworks dictate a rapid, absolute reduction in fossil fuel consumption.

Ultimately, $textsCO_2$ retrofits represent a vital bridge, not a final destination. As the artificial intelligence revolution continues to reshape global energy landscapes, technologies that squeeze maximum efficiency out of imperfect systems will play a critical role in staving off immediate blackouts—buying precious time for the broader energy transition to catch up with our digital future.