In a Shanghai test cell, a 200 kilowatt hour flywheel spun through 100 straight charge cycles with zero faults, so its maker moved it to higher speeds
Inside a test cell in Shanghai, a steel cylinder spins inside a sealed vacuum chamber.
It never touches the walls around it.
Magnets hold it suspended in open air.
In September, Qnetic reported that Pulsar had completed its first 100 continuous charge and discharge cycles with no faults and no measurable loss of capacity.
No flame, no chemistry, no electrolyte sloshing inside.
Just a rotor gathering speed until it holds enough motion to light a neighborhood for an hour.
Now it has done something no version of this machine had done before, and the next question is why that matters.
Why does spinning steel beat a chemical battery at holding a charge?
A flywheel stores energy the way a child’s top does, just enormously scaled up. Electricity accelerates a heavy rotor to tremendous speed, and that spinning mass holds energy as pure motion rather than as a chemical state locked inside electrodes. When the grid needs power back, the rotor turns a generator and slows down, giving up its motion as current flows outward.
That matters because lithium ion cells wear out through a chemical degradation process every time they charge and discharge.
So the company behind the machine, a Sacramento based startup called Qnetic, built its case on a different kind of durability. A rotor riding on magnetic bearings with no physical contact does not grind against anything the way a battery’s internal chemistry grinds against itself, and degradation, co founder and chief executive Michael Pratt told reporters, is the weakness it set out to design around from the start.
Picture the machine itself, caged in steel and silence
The prototype is called Pulsar. It is built to hold 200 kilowatt hours of energy, enough to run a few dozen homes for several hours, with a path to 500 kilowatt hours on the same platform.
Inside, a carbon fiber rotor spins in a cylindrical steel housing, floating on magnetic bearings borrowed conceptually from the physics that lets a maglev train glide above its rails. That contactless spinning design is the whole point, since friction is what kills spinning machinery over time.
Deployed at full scale, a flywheel like this gets buried in the ground with only a domed hatch exposed at the surface, an arrangement engineers compare to eggs standing upright in a carton. Pratt has also been candid about the risk built into high speed rotating steel, acknowledging the possibility of a rapid unscheduled disassembly if a rotor were ever pushed past its rated limits.
What the hundred cycle run actually proved, and when it happened
In September, Qnetic reported that Pulsar had completed its first 100 continuous charge and discharge cycles with no faults and no measurable loss of capacity. The company disclosed the milestone before moving the prototype into integrated low speed testing at its technology center in Shanghai.
Pulsar’s mechanical assembly had only been finished in June, after a run of structural checks meant to catch problems before the machine ever spun under its own power.
A separate test pushed the motor alone past 11,000 revolutions per minute, producing 130 kilowatts at 97.3 percent efficiency. None of those numbers yet describe a flywheel running at full operating speed. Pratt described the staged approach plainly in a company release, saying the sequence is meant to reduce technical risk before the system reaches full speed or a customer site. Qnetic is also sharing results with the Electric Power Research Institute, an independent body whose review will carry more weight with utilities than internal test data alone.
Where the hundred cycles still fall short of a grid scale claim
Even strong early numbers invite a fair question about what they do not show. Independent reporting on the milestone has noted that Qnetic has not disclosed how many kilowatt hours moved during each cycle, how long each cycle lasted, or at what power level the test actually ran.
That gap matters because flywheels have a track record elsewhere that is real but narrow. An earlier generation of grid connected units has logged years of frequency regulation service in New York and Pennsylvania, at smaller scale and shorter duration than Pulsar is aiming for.
Meanwhile, the roadmap depends on building rotors far larger than anyone has manufactured before. One harder problem ahead involves scaling up composite rotor production, since the filament winding process used to build these parts has rarely been attempted much beyond roughly eight feet across.
What comes next, and why the industry is watching this spot
If Pulsar clears full speed testing without incident, the next stop on Qnetic’s own schedule is a pilot installation back in Sacramento, targeted for the first half of 2027. That would be the first chance to see the technology near a real grid connection rather than inside a shielded test cell.
Utilities have reason to care regardless of how this company fares. The broader storage market is still searching for options beyond lithium for jobs lasting longer than a few hours, much like the thinking behind a coal plant battery now storing solar power for the evening peak in New Zealand.
Other labs are chasing the same gap with entirely different materials, including ceramic heat research out of Barcelona on a suspended chip built to widen temperature differences roughly a hundredfold. For now, the honest summary is narrower than the headlines around it.
A 200 kilowatt hour machine survived one hundred demanding cycles without breaking, a real and measured result, even if the harder proof still lies ahead in a test cell where the steel keeps spinning in silence.
Hugo is an engineer with strong technical expertise. Multilingual from an early age, his writing combines technical clarity with a strong interest in science and energy.
