Steam turbines still produce most of the world's power, but supercritical carbon dioxide promises to be much cheaper, and 10% more efficient as a medium than water, using 10X smaller turbines. A US$155-million pilot plant is now complete in San Antonio.
Ribbons were cut at the Supercritical Transformational Electric Power (STEP) pilot plant in Texas on October 27 as it was declared "mechanically complete" by project partners Southwest Research Institute (SwRI), GTI Energy, GE Vernova, and the U.S. Department of Energy.
The device in the image above is the world's first supercritical carbon dioxide turbine. Roughly the size of a desk, is a 10-megawatt turbine capable of powering around 10,000 homes. Ten megawatts is pretty small potatoes in the energy business, but to do it with a turbine this tiny? That could prove to be a revolutionary feat.
Carbon dioxide goes supercritical when the temperature and pressure are above about 31 °C (88 °F) and 74 bar (1,070 psi), respectively. At this point, it stops acting like a gas or a liquid, and instead starts acting something like a gas with the density of a liquid. Past this point, relatively small changes in temperature can cause significant changes in density.
Water can of course go supercritical too – it just takes a lot more energy, requiring a temperature and pressure over 373 °C (703 °F) and 220 bar (3,191 psi).
The properties of this supercritical CO2 fluid make it ideal for energy extraction in a closed-loop system, and back in 2016, General Electric announced it would start building a pilot plant to prove the idea in a commercially relevant installation, expecting to achieve 10 MW at an extraction efficiency of 50% – around 10% better than current steam turbines, which operate in the mid-40s – using a turbine about one-tenth the size.
Such a turbine could significantly reduce the capital cost of setting up any power generator reliant on heat and turbines; not only will the smaller turbines be cheaper, but they're so much more compact that you'll need less land for a given power plant. It'd also produce more power from a given heat source, and by default reduce the per-unit carbon emissions even of coal and gas-based generators.
And on top of that, it'll be much faster to start up and switch on. Running at around 700 °C (1,292 °F), GE's prototypes take about two minutes to start generating, where steam turbines take at least half an hour. So these CO2 turbines will be much quicker to respond to load demands, making them even more useful in a renewables-based grid.
Once proven, the tech could scale up to utility-relevant sizes and start displacing steam turbines in power plants. That includes fossil-fueled plants, as well as certain nuclear, geothermal, concentrated solar, and other installations.
There's still work to go before the pilot plant is commissioned, but the consortium expects it'll start operating in 2024.
“STEP will undoubtedly change the way we think about power generation,” said SwRI President and CEO Adam Hamilton, in a press release. “It’s exciting to officially launch this pilot plant, which is home to potentially revolutionary technology developed right here at SwRI.”
Source: Southwest Research Institute
What is the preferred fuel for heating the CO2 to supercritical conditions that enables grid export within 2 minutes?
Assuming that the CO2 turbine is a 10% efficiency improvement on steam, what is the operating cost benefit for using CO2 vs steam?
Smaller size does reduce capital costs, so that's valid. Supercritical CO2 is corrosive to many materials, so that might increase capital costs. As a new technology, there might be some long-term problems presently undiscovered, so that's a risk of unknown value. My guess is that the unknowns will slow adoption. Investors will wait to see whether the pilot plant keeps shutting down "for technical reasons".
To DP, the plant will probably contain very little CO2, so sourcing, leaks, etc, will be a minor concern. The low volume would explain the fast start-up time too. The fast start-up is probably assuming natural gas as fuel, since solid fuels or even liquid fuels take longer for the fire to really get going.
A quick check shows that solar thermal (using molten salt) is only 550C, so supercritical CO2 wouldn't work for that, at least without redesigning the solar heating part.
Gas leaks in this system won't be any greater in number or frequency than the average refrigerant system. They'll use the same kinds of seals on joints and shafts. Who can say what any "preferred" means of heating the CO2 will be for any given location? What is available for the cheapest cost is what will be used at each plant.
As to the operating cost benefit, unless my math is wrong, a 10% increase in efficiency would translate into a 10% cost saving.
This idea s really intriguing to me, using CO2 as the working medium to transfer the heat energy of an evaporator coil to a turbine! Way cool (literally!) of an idea. With the lower operating temps of the system, not as much fuel or other heat source, such as a boiling cup in a solar panel farm, would be needed! The lower Delta P of the system should make it much safer and longer lived.
Crazy how most every method of energy creation still relies on steam to turn heat into electricity. This is a cool new improvement
"...The average efficiency of electricity production from coal in both public electricity-only and public CHP plants is 37% in the OECD (averaged over the 2001 to 2005 period)..." The CHP plants increase the average efficiency of the fleet. Take them out and the average efficiency would be a few points lower.
No doubt these are all Rankine plants using steam turbines. In the largest plants, those steam turbines will be about 90% efficient.
How does it go from 90% through the turbine to 37% or less for the overall plant? I doubt that most of us could explain it.