A new twist on an old technology
Every industry moves at its own pace. The nimble home appliance sector can overhaul a toaster in months, while oil and gas operators can take years to alter equipment.
The electric utility industry is known for being particularly conservative and slow to adopt new technologies. After all, they have to keep the lights on. So when Physics Today heard that GE researchers were working in the lab on a new kind of alternating current (AC)/direct current (DC) power converter, we wanted to know: how much maturation time will this bench-top research need before it can be installed in a real life electrical substation?
As renewable energy power generation is erected in remote locations, high-voltage DC power lines will be built out to transport the energy to cities. DC lines are much more efficient than AC lines at transporting power long distances, but local grids are all based on AC. Every time a high-voltage DC power line links into the local grid, the utility will need to install AC/DC power conversion equipment.
Two classes of conversion device are currently in use: thyristors and voltage-source converters. Both are based on so-called power semiconductor technology. Thyristors are usually more cost-effective, with better efficiency and higher power capabilities than voltage-source converters. They’re preferred when a DC line is transporting gigawatts of power over long distances, but you need lots of them to do the job.
Voltage-source converters offer more functionality, and are handy when you have only a small space to work with, such for use on the ocean floor or an offshore wind farm. If a new technology emerges that is more affordable, more efficient, and more reliable than either thyristors or voltage-source converters, the owner of that technology could do very well. And this is why GE is looking into a new AC/DC power converter.
“We know we can build it. But can it last 20 years in the grid?” asks Timothy Sommerer, a physicist at GE Global Research Center in Niskayuna, New York. The new idea Sommerer and his colleagues are working on takes an old technology—gas tubes used to produce x-rays—and transforms it into something rugged enough for the grid.
The gas tubes are filled with helium and hydrogen gas. When the tube is turned on, the helium and hydrogen atoms ionize to form a plasma that conducts electrical current. The plasma can be made dense enough to conduct 500 amps through a single tube.
Sommerer and his colleagues believe that the gas tubes have the potential to be more efficient than thyristors or voltage source converters. What's more, they produce cleaner waveforms, which require less accessory equipment to smooth out for the grid. Gas tubes can also act as a switch or circuit breaker, which thyristors can’t do. Gas tubes could potentially eliminate the need to include additional circuit breakers in the system.
But the technology has a weak link: the cathode, or negative electrode. The cathode is constantly bombarded with plasma ions that eventually nick it, creating a bump. The bump acts like a lightning rod, attracting more ions that steadily erode it, faster and faster, until it fails.
To avoid that outcome, the team has been working with cathodes made of liquid gallium. Gallium melts at 29.8°C; it conducts just as well when in the liquid state. But it can’t be nicked: The liquid metal will simply flow and heal itself, preventing the runaway erosion that besets solid cathodes. Any liquid metal with a low vapor pressure can be used, including several alloys of lead or bismuth.
At this stage of the R&D, the cathode is decidedly low tech, just a dish with a few millimeters of liquid gallium. But Sommerer has ideas.
“We’ve been in contact with fusion guys who work with tokamaks,” he says. The inside of a gas tube filled with a dense plasma isn’t all that different from a tokamak. The plasma fusion physicists have a technique where they pump liquid metals through metal foams. The liquid metal wicks through the foam like water through a sponge and forms a very flat, thin layer due to surface tension. That could be the high-tech version of a super-durable cathode, if this technology ends up on the market.
It will probably be a year or two yet before Sommerer and his colleagues reach the commercial prototype stage, though. GE uses a system of technology readiness levels (TRLs) to rate the development stage of a new tech. The TRL system originated in NASA but is now common in industry. The scale ranges from 1 to 9. A TRL of 1 means a new technology is just a concept, while a TRL of 9 means it is tested and proven rugged in the field.
Right now, the gas tube power converter is between TRL 2 and TRL 3. This means that they know it works in the lab, and they’re starting on active research and design to see if the technology could be commercially viable. They want to get a good read on its potential, and its risks, before their $4 million grant from ARPA-E runs out.
“GE excels at moving things from one industrial space to another,” says Danielle Merfeld, GE’s global technology director for electrical technologies and systems. Because the new AC/DC power converter technology is based on gas tubes that have been used for decades to produce x rays for medical imaging, “we know how to make these tubes, how they age, how they fail. But all our experience is in a medical setting. If they’re sitting outside in an electrical substation,” she says, GE doesn’t know how they’ll hold up.
Once GE has a proof of concept device, they can start testing different components and trying them out in realistic situations. Eventually, if everything works out, they will build a commercial prototype (TRL 6). Then it’s time to find out how the new technology performs under stress. GE calls these "jugular tests." They’re field tests that push the technology to the limits of its capabilities, in order to observe what really happens when the voltage surges, the grid frequencies go haywire, or the freak-of-nature event occurs.
“Freak-of-nature instances happen all the time,” Merfeld says. So they test and test and test. The most critical test for the gas tube AC/DC power converters, though, will be, not a freak-of-nature, but a lifespan test. GE has to find some way to run the tubes in an accelerated life mode for a year or two, to simulate 20 years of real wear and tear on the cathode. And they need to see if the gas tube can really work as a circuit breaker if there’s an emergency fault in the system that pumps 10 000 amps through it.
If the power converter survives the jugular gauntlet, it’s time to partner up with a utility and do the ultimate test: installing it in a commercial project. Utilities are, understandably, reluctant to be the first to try out new technology, so usually these partnerships are brokered by the Department of Energy. DOE can sweeten the deal for the utility by offering first access to the knowledge gained during the project, or extra money or loan guarantees for the installation of the new equipment and associated facility upgrades, or even by giving the utility preference in another project.
So, to answer our original question—two years developing a proof of concept, another couple of years developing a commercial prototype, then a year or two of jugular testing—means that it takes at least six years to get from concept to commercial test. Permitting a new power line can take longer than that. For now, the team is focusing on basic science, modeling the plasma inside the tube, and working on the gallium cathode.