The Power Ceiling Breaks

Electric propulsion cleared a thirty-year limit three times in five months — and not one of those advances was a thruster advance

The Ecliptic · Issue No. 004 · August 10, 2026

20 kW

Hall thruster fired in orbit in late March — after three decades stuck below about 5 kW in flight

120 kW

lithium-fed magnetoplasmadynamic thruster fired on the ground on 24 February

~10 W/kg

vehicle-level specific power of the 20 kW demonstration: two tonnes of spacecraft, 40 metres of solar wingspan

20 kWe

fission reactor now manifested for Mars in 2028 — because sunlight cannot carry the next step

What actually happened

Three ceilings, five months, one cause

Electric propulsion spent thirty years stuck below roughly 5 kW per thruster in flight. In the first five months of 2026 that limit broke three separate times — in orbit, on the ground, and in policy.

Late March, in orbit. K2 Space fired a 20 kW Hall thruster on a flying spacecraft — a fourfold step over the flight-proven class.
24 February, on the ground. JPL fired a lithium-fed magnetoplasmadynamic thruster at 120 kW, an order of magnitude beyond anything flown.
April, in policy. The White House directed NASA to field a space reactor of at least 100 kWe for the 2030s.
Read together, they look like a propulsion breakthrough. Our desk's finding is that none of the three is a plasma-physics advance.

The diagnosis

It was never the thruster. It was the bus.

Vehicle mass implied by holding vehicle-level specific power at the ~10 W/kg demonstrated in flight. The 2-tonne, 20 kW point is flown hardware; the ~12-tonne, 120 kW point is the report's derived implication for a solar-fed vehicle near Earth, not a built vehicle. Hatch = implied, not flown.

Build the power, mass and thermal budget from the hardware that actually flew and the picture inverts. The thruster was not the scarce part; the electricity to feed it was. K2's 20 kW came from a two-tonne spacecraft carrying a forty-metre solar wingspan — about ten watts per kilogram at the vehicle level. Hold that specific power fixed and the arithmetic for the next step is brutal.

The binding constraint migrated from the thruster to the spacecraft — array area, power conditioning, and heat rejection.
At 60–70% thruster efficiency, every kilowatt delivered is a third of a kilowatt of waste heat that has to be radiated. Above roughly 50 kW, radiators plausibly bind before arrays do.
That makes vehicle-level watts per kilogram the one honest figure of merit — not thruster power, which is the number everyone quotes.

The consequence

Which is why a reactor is manifested for 2028

If a solar-fed 120 kW vehicle runs to roughly twelve tonnes near Earth — and considerably more at Mars, where the sunlight is weaker — then sunlight stops being the answer somewhere between the two demonstrations. That is the whole reason a fission reactor now has a launch date.

A 20 kWe reactor is manifested for Mars in 2028; the directed ≥100 kWe class follows in the 2030s.
The gate on both is not reactor physics but qualified HALEU fuel form — the same downstream fuel-cycle constraint our desk identified before this hardware existed.
Strategic read: electric propulsion has quietly become a power-generation and heat-rejection industry wearing a propulsion label — and the capital is flowing to the wrong half of it.