“Supercritical” refers to the state of the CO2 that circulates through the system, pushing against the blades and turning the turbine. Carbon dioxide is a gas at room temperature and forms “dry ice” when it freezes. At high temperatures and pressures, it is dense like a liquid but expands like a gas to fill the available space. Smaller pilot projects suggest that supercritical CO2 turbines could be petite enough to fit on a conference table and generate electricity at higher efficiency than steam-powered turbines.
The San Antonio plant, developed in partnership with Southwest Research Institute as part of a $119 million project with the U.S. Department of Energy, will be the largest carbon dioxide turbine ever built. It will generate 10 megawatts — enough to power 10,000 U.S. homes — and operate at 700 degrees Celsius. That’s a hundred times more electricity than the 100-kilowatt CO2 plant now operating at Sandia National Laboratories in New Mexico, and 200 degrees hotter.

The purpose of the new facility is to demonstrate that smaller CO2 turbines can scale up to industrial-size plants. “We’re hoping to prove that we can operate at scale with good safety and control systems,” says engineer Douglas Hofer, who developed the turbine at GE Global Research headquarters in Niskayuna, New York.
The team will be looking closely at half a dozen control variables that determine how efficiently the turbine is being run — things like the temperature of the CO2 as it enters the turbine, pressure at key valves and the status of materials at key joints and seals.
One important goal of the project is to work out precisely how to start the turbine. Just as a turn of the ignition key on a car triggers a series of events — a squirt from the fuel injector, an ignition spark and so forth — starting a gas turbine is a multistep operation. Of course, ramping up a CO2 turbine from room temperature to high temperatures and pressures puts a great deal more stress on components than anything that happens in a car engine. “It’s a thermodynamic power cycle, so the whole thing starts out cold,” Hofer says. “You have to warm it up in a controlled way, avoiding stress on various components. You have to get the right amount of CO2 at the right temperature. When in this sequence do you start rotating the turbine? How do you avoid having liquid CO2 condense, or form dry ice? We’ll be looking at all these things.”

Once operators have the turbine up and running, the next issue is how to balance the load on the grid, matching output moment by moment to fluctuating demand for electricity. Both conventional and CO2 turbines can adjust their outputs by raising or lowering the temperature of the gas (steam or CO2) that enters the turbine. Operators of CO2 turbines can also adjust the amount of CO2 that turns the turbines. “Because we have the option of changing the amount of CO2 in the loop, which changes the circulation rate, we can get a lower power output with the same turbine inlet temperature,” says Hofer. “That allows us to operate at a wide range of outputs at high efficiency levels.”
Hofer expects that GE will start getting data from the new plant in 2020 and finish in 2022. Beyond that, the company may look to applying the lessons learned to concentrated solar plants and next-generation nuclear plants, which use similar closed-loop thermodynamic systems.
Although the GE team is two years into the six-year program, research for the CO2 turbine goes back to early work on components as part of a Department of Energy contract begun in 2012. “Our journey,” says Hofer, “has been longer than two years.”
