Realta Fusion Tries to Turn Fusion's Biggest Leak Into Its Greatest Asset

The leak that became the point

Fusion reactors have always had a nuisance: not all the hot plasma stays trapped. Some of it escapes — through the ends of a magnetic mirror, or across the edge of a tokamak’s confinement field. Every major fusion program treats this as waste heat to be managed, sometimes captured in a blanket, often just lost.

Realta Fusion is flipping that assumption. At its WHAM device in partnership with the University of Wisconsin–Madison, the company used an electrostatic deceleration grid to extract electrical current directly from charged particles exiting the plasma — without converting their energy to heat first. The demonstration, announced on June 19, pulled roughly 100 volts at several amperes. That lights a handful of bulbs. It does not light a city.

But the idea behind the demonstration, if it scales, changes the math for every fusion reactor that follows.

How direct conversion works — and why it matters

A conventional fusion power plant would generate thermal energy from neutron collisions, boil water, spin a turbine, and feed the grid through an electromagnetic generator. This is the standard thermodynamic loop, and it works for fission and fossil fuels because the heat source is already dense and continuous.

Fusion is different. In a deuterium–tritium reaction, about 80 percent of the energy comes out as neutral neutrons and 20 percent as charged alpha particles. The neutrons must be absorbed in a blanket to produce heat. The alphas, which carry kinetic energy and electric charge, are usually just thermalized into the plasma — a process that actually helps sustain the reaction but contributes nothing to net electricity.

Direct energy conversion skips the thermal intermediate. An electrostatic field — typically a series of graded-grid decelerators — slows the charged particles and redirects their kinetic energy into an electrical current. No moving parts. No steam. No Carnot limit constraining the efficiency.

Realta’s WHAM experiment installed a three-layer mesh grid at the exhaust of a magnetic mirror device and demonstrated that particles leaking through could be collected as multi-ampere current at approximately 100 volts. The result was published in a joint UW–Madison / Realta collaboration, with diagrams attributed to D. Endrizzi and colleagues.

The physics is textbook. The demonstration is the novelty.

Why the magnetic mirror matters

Realta is not building a tokamak. Magnetic mirrors predate the tokamak by decades and have largely been abandoned as a confinement concept because they leak plasma at both ends — exactly the behavior that makes direct conversion possible. Tokamaks trap plasma toroidally with no open field lines; mirrors hold it between two magnetic “knots” with natural exhaust ports.

That leak is the design’s strength here. The particles escaping the mirror ends carry a useful fraction of the plasma energy, and they exit with directional velocity — ideal for a decelerator grid. Tokamak-edge particles, by contrast, are slower, isotropic, and harder to harness efficiently without a substantial divertor architecture.

There is a second advantage Realta flags. Magnetic mirrors naturally expel helium ash and impurities more readily than tokamaks. Maintaining plasma purity over long burn periods is one of the persistent operational problems for confinement designs, and an open-field-line geometry sidesteps part of it. Whether that advantage holds at reactor-relevant densities remains an open question.

The gap between proof and plant

Let us be precise about what was demonstrated and what was not.

The WHAM device uses deuterium only — no tritium, no fusion reactions producing alpha particles at MeV energies. The extracted electricity came largely from externally heated plasma, not from self-generated fusion energy. Realta has stated this explicitly: the experiment showed the component technology, not a power plant.

Scaling from a laboratory mirror device to a grid-connected reactor involves several unsolved steps:

• The particle energies in a D–T reactor will be roughly an order of magnitude higher than in WHAM’s deuterium-only plasmas. Decelerator grids must withstand substantially higher voltages and particle fluxes without arcing or erosion.
• The beam must remain collimated enough for the graded-grid design to work. Plasma turbulence at reactor scale could spread the particle angles beyond what a static decelerator can capture efficiently.
• The system must operate continuously, not in short pulses. Grid materials under sustained ion bombardment — tungsten, molybdenum, or novel composites — have not been tested at relevant fluences.
• Integration with the neutron blanket and the reactor’s power balance must be solved. A 20 percent charged-particle fraction is useful, but only if the remaining 80 percent can still be converted to heat efficiently enough to close the plant’s thermodynamic loop.

Realta’s stated trajectory moves toward kilowatt-scale demonstration first, then megawatt. That sequence is reasonable — the company is not claiming a reactor today. But the jump from a few hundred volts in a university lab to multi-megawatt direct-conversion output in an industrial device is the same kind of leap that has stalled other fusion concepts before they reached engineering relevance.

Where this fits in the fusion race

Several other fusion companies have explored direct conversion. Helion Energy, for instance, uses pulsed magnetic compression and has discussed direct-energy capture in its field-reversed configuration design, though it has not publicly demonstrated the technology at scale. General Fusion’s magnetized target approach is built around a thermodynamic cycle, not direct conversion. Commonwealth Fusion Systems, TAE Technologies, and the major national programs — ITER, SPARC, JT-60SA — all rely on the standard heat-to-turbine route.

Realta’s claim to novelty, then, is narrower than the headlines suggest: it is the first US commercial fusion company to demonstrate direct energy conversion with actual plasma in a magnetic mirror device. That is a meaningful first. It is not a first in the broader sense.

What makes this story worth watching is not the demo itself but the strategic implication. Direct conversion reduces the fusion gain (Q) required for a plant to produce net electricity, because it recovers energy that would otherwise be lost or downgraded through a thermal cycle. If the technology scales, it gives smaller, cheaper mirror-based reactors a path to positive balance that tokamak designs must earn through higher performance alone.

That shifts the competitive frame. Tokamaks are betting on confinement excellence. Mirrors are betting on exhaust efficiency. Neither approach is proven at reactor scale. The question is which risk profile the market rewards first.

What to watch next

Three milestones will determine whether Realta’s approach is a credible branch of the fusion portfolio or a promising side experiment:

• A kilowatt-scale demonstration with deuterium–tritium or a high-energy deuterium fuel cycle, showing that the decelerator survives and performs under fusion-relevant particle fluxes.
• Publication of the conversion efficiency under those conditions — not just that current was extracted, but what fraction of the charged-particle energy was recovered.
• An integrated reactor concept that combines the mirror confinement, the direct-conversion exhaust, and a neutron blanket in a coherent power-balance model, published in a peer-reviewed form rather than a press release.

Until then, the WHAM result is best understood as a component proof: a single subsystem shown to work in plasma. That is real progress. It is not a power plant.

Fusion has survived decades of overpromising. What it has rarely seen is a credible reframe of its own waste stream as a resource. Realta is attempting that reframe. The demonstration was small. The idea is worth tracking.