Facebook

The Gyrotron: From Fusion Plasma to the Rock Beneath Our Feet

A Machine With Two Callings

Thomas Lee Abshier, ND | 10 October 2026

Few devices span two frontiers of energy research the way the gyrotron does. It was built to heat the plasma inside fusion reactors to temperatures hotter than the core of the Sun. Now a handful of engineers are pointing the same beam down a borehole to vaporize granite, hoping to reach the deep heat of the Earth’s crust. Both uses rest on the same physics: a beam of electrons spinning in a strong magnetic field can be made to give up its energy as a narrow, intense beam of millimeter-wave radiation. Following that physics from the vacuum tube to the plasma to the rock face shows how a laboratory instrument becomes a candidate for a new energy industry, and where the remaining hard questions lie.

How a Gyrotron Works

Electrons in a magnetic field

A free electron in a uniform magnetic field moves in a circle around the field lines. The rate of that circling, the cyclotron frequency, depends only on the field strength and the electron’s mass. It comes to about 28 gigahertz per tesla of field. A gyrotron producing 170 GHz therefore needs a field near 6.7 tesla, more than a hundred thousand times the Earth’s field. That field is supplied by a superconducting magnet cooled to a few degrees above absolute zero.

The electron gun and the hollow beam

At one end of the tube, an electron gun boils electrons off a heated cathode and accelerates them through tens of thousands of volts. The gun is shaped so that the electrons leave with a large share of their motion going sideways to the field. The result is a hollow, ring-shaped beam of electrons, each one corkscrewing along the field lines.

Bunching: the key trick

A beam of electrons circling at random phases radiates almost nothing useful, because their individual emissions cancel. The gyrotron works because of a subtle effect from Einstein’s relativity. At these voltages the electrons travel at a sizable fraction of the speed of light, and their effective mass depends on their energy. An electron that loses a little energy to the wave becomes slightly lighter and circles slightly faster. An electron that gains energy becomes heavier and circles slower. Inside a resonant cavity tuned just above the cyclotron frequency, this sorts the electrons into phase-aligned bunches that circle together. Once bunched, they radiate coherently and pour their sideways energy into a single electromagnetic mode of the cavity.

This is why the gyrotron is called a cyclotron resonance maser. The name borrows from the laser’s cousin, but the mechanism is classical. No atomic energy levels are involved. Instead, bunching is driven by relativity and the magnetic field.

Why gyrotrons can be so powerful

Ordinary microwave tubes, such as the magnetron in a kitchen oven, need internal structures about the size of the wavelength they produce. At millimeter wavelengths those structures would be tiny and would melt at high power. The gyrotron sets its frequency with the magnetic field, not with the size of its parts. Its cavity can therefore be many wavelengths across and operate in a high-order mode, spreading the heat load over a large wall area. That is what lets a single tube deliver a megawatt continuously.

Getting the beam out

The cavity’s complex mode is converted inside the tube into a clean, pencil-like beam. It leaves through a window made of synthetic diamond, one of the few materials that can pass a megawatt of millimeter waves without cracking. The spent electrons strike a collector. In modern tubes the collector is held at a reduced voltage so that part of the electrons’ leftover energy is recovered. That recovery raises the overall wall-plug efficiency to roughly half. The output then travels through corrugated metal waveguides or along mirror lines to wherever it is needed.

The Gyrotron in Fusion

Heating electrons where you choose

A magnetically confined fusion plasma must reach well over 100 million degrees. Gyrotron beams heat it by electron cyclotron resonance heating. The beam passes through the plasma largely untouched until it reaches the place where the local electron cyclotron frequency matches the beam frequency. There, it is absorbed strongly and locally. In a tokamak the magnetic field weakens with distance from the central axis. Choosing the beam frequency and steering the launching mirrors therefore lets operators deposit power at a chosen radius, often to within a few centimeters.

More than heating

That precision makes the gyrotron a control tool as well as a heater. Its beams are used to:

Break down the initial gas to start the plasma
Drive electric current in the plasma by pushing electrons in a preferred direction
Suppress magnetic instabilities by depositing power exactly on the unstable surface before the instability grows

ITER as the benchmark

ITER, the international fusion experiment under construction in France, sets the current standard. Its electron cyclotron system calls for 24 gyrotrons at 170 GHz. Europe is responsible for six of them, and the Russian, Japanese, and Indian agencies supply the other eighteen. The requirement is about one megawatt per tube for long pulses, and the development tubes have met it. ITER-India reported holding 1 MW at 170 GHz for 1,000 seconds at about 50 percent RF efficiency. Europe’s industrial prototype delivered 0.9 MW in 180-second pulses and above 1 MW in 40-second pulses, limited by the test stand’s cooling, and ran at half a megawatt for 1,600 seconds. These long-pulse, megawatt-class tubes are what made the drilling idea practical.

From Plasma to Granite

Why drill deep

The Earth’s interior is enormously hot. Estimates place the inner core near 5,000–6,000 °C. Over most of the continents, temperature in the crust rises about 25–30 °C per kilometer of depth. That heat comes mainly from the slow radioactive decay of uranium, thorium, and potassium in the rock, plus heat left over from the planet’s formation. It is not, as is sometimes said, produced by fission reactions; natural fission reactors did operate briefly about two billion years ago at Oklo in Gabon, but that is a geological curiosity, not the Earth’s heat engine.

At depths of roughly 10 to 20 kilometers in ordinary rock, temperatures pass about 374 °C. Water pumped down at the pressures found there exceeds its critical point of 374 °C and about 22 megapascals. In that state it is neither liquid nor gas, and it carries several times more usable energy per unit of flow than ordinary geothermal hot water. A “superhot” well could therefore produce many times the power of a conventional geothermal well. It could also be sited almost anywhere, not just in rare volcanic regions like Iceland.

Why deep drilling is so hard

The obstacle is the drilling itself. The deepest hole ever made, the Kola Superdeep Borehole in Russia, reached about 12.3 kilometers after roughly two decades of work and was stopped largely by heat. Several problems compound with depth:

Bit wear: hot, hard basement rock such as granite destroys drill bits quickly.
Tripping: every bit replacement requires pulling the entire string of pipe out of the hole and running it back in, which can take days at great depth.
Hole stability: deep holes become unstable and must be lined with steel casing and cement.

Drilling cost therefore rises much faster than depth. One published estimate puts two 3-kilometer wells at about $6 million and two 6-kilometer wells at about $27 million.

The millimeter-wave answer

The approach now being commercialized grew out of about a decade of research at MIT. Before 2025, millimeter-wave drilling had been shown only in the laboratory, where MIT’s early system drilled a hole just a few centimeters deep. The company Quaise Energy took the idea forward. Conventional drilling is used through the softer upper layers. Below that, the drill string is replaced with a waveguide, essentially a hollow metal pipe, that carries the gyrotron’s beam to the bottom of the hole.

At the rock face, the beam does three things:

It ablates the rock. The beam melts and then vaporizes the rock. A stream of nitrogen gas sweeps the vapor back up the hole, where it condenses into fine glassy fibers and dust.
It removes the need for bits. Nothing touches the rock, so there is nothing to wear out and no reason to trip pipe.
It may line the hole as it goes. By tuning the beam, part of the wall can be left melted, forming a glassy, obsidian-like lining. If that lining holds, it could do some of the work of steel casing.

A proportion check

The figures Quaise has given publicly make for an instructive calculation.

Quaise’s target: one meter per hour with one megawatt, for a hole about 8.5 inches across. That is one megawatt-hour, or 3,600 megajoules, per meter of hole. A meter of 8.5-inch hole in granite weighs about 100 kilograms. The target therefore budgets roughly 36 megajoules for every kilogram of rock removed.

The laboratory reference: 100 kilowatts making a 4-inch hole at one meter per hour. That works out to about 16 megajoules per kilogram.

Heating silicate rock past its boiling point and vaporizing it takes on the order of ten to twenty megajoules per kilogram. The laboratory figure sits in that range. The commercial target leaves roughly a factor of two of margin for losses along a long waveguide, which matches what the company’s engineers have described.

At $75 per megawatt-hour, the electricity for one kilometer of hole costs about $75,000, and for seven kilometers about $525,000. Those are small sums beside the cost of the rig, the compressors, and the time. The beam’s energy cost is not the main difficulty. The difficulty is everything around it.

Rock type matters

Basalt drills faster than granite, even though the two melt at similar temperatures (roughly 1,200–1,500 °C). The difference lies largely in quartz. Quartz absorbs millimeter waves poorly and reflects part of the beam, so quartz-rich granite needs more energy before it reaches the vaporization point. Coarse-grained granites with large quartz crystals are slower still. The beam does not “see” all rock alike, and drilling recipes will have to be tuned to the formation.

Where Things Stand in 2026

The technology has moved quickly. In April 2025, Quaise tested a system on a full-scale oil and gas rig owned by Nabors Industries, creating what it called the first hybrid rig combining conventional and millimeter-wave drilling. In July 2025, it drilled 118 meters into granite at a quarry in central Texas. The plan for 2026 was to integrate a megawatt-scale system onto the Nabors rig as the company’s first commercial prototype.

In July 2026, Quaise raised $134 million in the first close of its Series B to fund Project Obsidian, billed as the first commercial superhot geothermal power plant, and to push its drilling toward depths beyond 5 kilometers. The company reported it was nearing one kilometer of depth at its Texas site, which would be the deepest penetration by any non-contact drilling method.

The Open Questions

The physics of vaporizing rock with a millimeter-wave beam is settled. Whether it pays at full depth is not. Four questions remain.

Water. Groundwater flowing into the hole absorbs beam energy as it boils. It may also scatter the beam. One proposal is to melt material into the inflow paths to seal them, but sealing has not yet been shown at depth.

The glass liner. The economics depend heavily on whether the vitrified wall can hold a deep hole open under heat and pressure, reducing the steel and cement required. Only deep field wells can settle this.

The waveguide over kilometers. Transmission losses, alignment, and keeping the vapor clear of the beam path all get harder as the waveguide lengthens. The step from one kilometer to five or ten is the real test.

Cost of electricity. A power source succeeds only if its levelized cost competes. Quaise’s own map estimates about $115 per megawatt-hour at its Texas site and as low as $68 elsewhere. That is in the range of new nuclear and above onshore wind. These are projections, not results.

Conclusion

The gyrotron is a fine example of how basic physics travels. A relativistic subtlety in how fast electrons circle a magnetic field line became the most precise heating tool in fusion research. The same tube, with its strict frequency requirements relaxed, is now being asked simply to deliver as much heat as possible down a narrow hole. Fusion still needs years before it powers a grid. The gyrotron may reach the grid first by a different road: unlocking heat that has been under our feet all along.

References

Quaise Energy, “Looking back on 2025.” https://www.quaise.com/news/looking-back-on-2025
ASME Mechanical Engineering, “Digging into Nuclear Fusion Technology to Tap Geothermal,” October 2025. https://magazine.asme.org/issues/october-2025/news-digging-into-nuclear-fusion-technology-to-tap-geothermal
Las Vegas Sun, “Quaise Energy raises $134 million in initial close,” July 7, 2026. https://lasvegassun.com/news/2026/jul/07/quaise-energy-raises-134-million-in-initial-close-/
ThinkGeoEnergy, “Quaise Energy achieves 100 meters of drilling using millimeter wave technology.” https://www.thinkgeoenergy.com/quaise-energy-achieves-100-meters-of-drilling-using-millimeter-wave-technology/
Fusion for Energy, “The EU design of the gyrotron for ITER is pre-validated by F4E.” https://fusionforenergy.europa.eu/news/the-eu-design-of-the-gyrotron-for-iter-is-pre-validated-by-f4e/
S.L. Rao et al., “Commissioning of MW class Gyrotron Test Facility at ITER-India & Demonstration of ITER relevant RF performance (1MW for 1000 s at 170 GHz),” EPJ Web of Conferences 313, 04002 (2024). https://doi.org/10.1051/epjconf/202431304002
T. Rzesnicki et al., “European 1 MW, 170 GHz CW Gyrotron Prototype for ITER — long-pulse operation at KIT,” IRMMW-THz 2022. https://doi.org/10.1109/IRMMW-THz50927.2022.9896073
Real Engineering, video report on Quaise Energy’s gyrotron drilling (source of the cost and rate-of-penetration figures).