A former scientist at Sandia National Lab is bringing the technology to market

Temperature / Pressure
ranges for supercritical CO2
Image: S. Wright

Carbon Dioxide is good for a lot of things like adding fizz to your soft drinking in its gaseous form and keeping food cold in its solid form as “dry ice.”

What many people do not know is that the gas is also useful in its lesser known liquid form. You won’t see it in nature since it takes five atmospheres of pressure at slightly higher than room temperature, 31C, to take it to a super state.


Then it can be used to push turbine blades and thus generators connected to them to make power. (Note to readers: Convert Kelvin to Centigrade using this tool.)

Because of its physical properties as a liquid, it has become a target fluid of opportunity to run turbines and thus make electricity. Steven Wright, Ph.D., who recently retired from Sandia National Laboratory (SNL), has set up a consulting company called Critical Energy LLC to bring this technology to a commercial level.

The objective of using supercritical CO2 (S-CO2) in a Brayton-Cycle turbine is to make it much more efficient in the transfer of heat. Wright points out that a steam turbine is about 33% efficient, but that an S-CO2 turbine could be as high as 48% efficient, a significant increase.

A closed loop supercritical CO2 system has the density of a liquid, but many of the properties of a gas. A turbine running on it, “is basically a jet engine running on a hot liquid,” says Wright.

“There is a tremendous amount of scientific and industrial interest in S-CO2 for power generation. All heat sources are involved including solar, geothermal, fossil fuel, biomnass, and nuclear.”

Barber-Nichols S-CO2 turbine wheel
Photo: Sandia National Laboratory

Size matters

It is that last area that has Wright focused on small modular reactors or SMRs. Because developers of reactors in the range of 45-200 MW have promised customers that these compact power units can be delivered to a site on a truck, or by rail, they need turbines of similar scale.

A set of S-CO2 turbines and compressors is is about 3-4% the size of a conventional steam generator of similar power rating. The heat exchangers, taken together with all other components, brings in a configuration that is 30-40% smaller than for similar steam systems.

The footprint of a S-CO2 turbine, compressor and heat exchangers for a 50 MW system would be scaled accordingly, but would still have a smaller footprint than a comparable steam system.  It would be more than a couple of shipping containers, but less than a whole barge of them.

In January 2012 Wright wrote about his work and the fabrication of prototype parts by the firm of Barber-Nichols in Arvada, Colo.

Q&A with the principal investigator

This blog talked with Wright by phone this week. Here’s what he had to say.

Q: The paper mentions a reactor outlet temperature of 500-700C. Is there an ideal temperature coming off the reactor? Also, what is the temperature of the return loop?

A: For a fast reactor with an outlet temperature of 550C, the return loop is 500C. The very small temperature differential means you don’t have to push a lot of heat getting the return loop back up to the right temperature again. The temperature differentials for the CO2 in the secondary loop would be about 150C.

“The small temperature differential makes it more efficient”

Wright claims that as a practical matter, at any reactor outlet temperature above 400-450C, CO2 exceed steam for efficiency.

“These temperatures are consistent with sodium-cooled or other liquid metal cooled and gas cooled fast reactors. The temperatures are not so high that you get into materials issues for containing the heat. The upper boundary is about 650C.”

Conceptual drawing of sodium-cooled
fast reactor: Image: Idaho National Laboratory

In terms of achieving commercial success, Wright is focused on temperatures starting at 500C because he knows he can get stainless steel fabricated from commercial sources that will operated in his design.

Also, he has found through his work that he can get commercial bearings and seals to work well in this temperature range.

Q: Your prototype at Sandia ran at very high speeds. How do you plan to get the power transferred from the turbine to a generator?

A: We are planning for a turbine speed of 36,000 RPM which can be stepped down in a 10:1 gearbox to 3600 RPM to produce power at 60 Hz.

Q: Recognizing that a S-CO2 turbine is not a combustion unit, like one in a jet engine, what are the key technical challenges to scaling up to commercial size?

A: The key challenge will be to build a 100-200 MW unit to show that everything works. My focus is to get the cost and design of the heat recuperators to meet commercial needs. They must be small, compact, and affordable.

Our plan for a 10 MW prototype has two of them using very small channels. In our lab prototype, we have two of them that are operating which were fabricated for us by Heatric.

Recuperator conceptual diagram
Image: Science Direct

The recuperators are used to increase the overall efficiency of the system. The recuperator transfers some of the waste heat in the return loop  preheating the CO2 before it contacts the primary loop again.

Since the CO2 is pre-heated, less energy is needed to then heat it up to the turbine inlet temperature. By recovering some of the energy usually lost as waste heat, the recuperator can make a supercritical gas turbine significantly more efficient.

The recuperators on the S-CO2 systems work much the same way as a gas fired turbine except they are heating a noncombustible gas instead of air to be mixed with fuel.

Q: You indicated that a first-of-a-kind system producing a 10 MW system could be $20-30 million. Do you have a cost estimate for an “Nth of a kind at 50 MW?”

A: Our target is a cost of $1 a watt. If we can find the right industrial partners, we could produce the first units for non-nuclear power applications in three-to-five years.

Separately, Wright said that to use the turbines with small modular reactors (SMRs), they would have to be evaluated as safety-related equipment by the NRC or any nuclear safety agency elsewhere.

He wants to see a revenue stream from non-nuclear applications before spending money on the cost of a regulatory process that covers equipment for nuclear power stations.

Asked about licensing and partnering opportunities, Wright refers inquiries to the Technology Transfer Office at Sandia.  See also this Sandia press release.  For additional technical information, go to the S-CO2 Power Cycle web site.