Entry 125 showed that the corridor beyond GEO is largely unwatched. This sweep asks how the keeper moves through it. Capture is not a launch burn; it is a chase that can last years, against a target whose gravity is so weak that solar radiation pressure matters and whose surface may be a cohesionless rubble pile. The engine choice shapes the timeline, the mass budget, and the business model.

Chemical: fast and thirsty

Chemical propulsion is the default for proximity operations. DART used twelve Aerojet MR-103G hydrazine thrusters for trajectory-correction maneuvers and attitude control, plus a NEXT-C ion engine as an experimental secondary propulsion. Hera, launched in 2024 for the Didymos post-DART reconnaissance, is a 1,214 kg hydrazine spacecraft; its largest deep-space maneuver consumed 123 kg of hydrazine. OSIRIS-REx used hydrazine for everything — four 200 N main thrusters, six 22 N trajectory-correction thrusters, and dozens of smaller reaction-control thrusters. Hydrazine monopropellant delivers an Isp of roughly 180–285 seconds; storable bipropellants reach 300–320 seconds; cryogenic LOX/LH2 can reach 420–450 seconds but is harder to keep for multi-year missions. Chemical is reliable, fast, and expensive in propellant mass. For the multi-kilometer-per-second ΔV of a heliocentric chase, it is usually the wrong answer unless the timeline is urgent.

Solar electric propulsion: the long answer

The missions that have actually chased small bodies across the solar system mostly use solar electric propulsion. Hayabusa2 flew to Ryugu on four microwave-discharge ion engines, three operating at a time, each delivering 10 mN of thrust at 3,000 seconds Isp, with 66 kg of xenon. Over 25,590 hours of powered flight the spacecraft gained roughly 2.2 km/s ΔV and about 1 meganewton-second of total impulse. Psyche, the first NASA science mission to use Hall thrusters beyond lunar orbit, carries four SPT-140 Hall thrusters, 922 kg of xenon, and 21 kW of solar array power at 1 AU. Dawn, which orbited Vesta and Ceres, used three NSTAR ion thrusters and 450 kg of xenon to deliver ~11 km/s ΔV over ten years.

The closest analog to a keeper tug is the Asteroid Redirect Mission concept: a 40 kW-class SEP spacecraft with four 10-kW Hall thrusters, 50 kW of solar arrays, up to 12 tonnes of xenon, a dry mass under 4 tonnes, and a target return mass of up to 1,000 tonnes. The Keck Institute’s 2012 precursor study used similar numbers: 40 kW, 3,000 s Isp, 5.5 tonnes dry, 13 tonnes xenon, returning 250–1,300 tonnes of asteroid in 6–10 years. That architecture is not exotic; it is the only one that has been studied in detail for moving large rocks.

Nuclear, sails, and tethers

Nuclear electric propulsion projects system masses of roughly 10–20 kg per kilowatt of electric power, with multi-megawatt reactors enabling very high ΔV independent of solar distance — but no integrated MWe-class NEP has flown, and the radiator area alone dominates the mass budget. Nuclear thermal propulsion offers 850–1,000 s Isp with high thrust, but long-term liquid-hydrogen storage at 20 K is unsolved for multi-year missions. Solar sails are propellantless and patient: NEA Scout planned an 86-square-meter sail with 0.06 mm/s² characteristic acceleration for a 2–2.5-year cruise to a tiny asteroid flyby. Electrodynamic tethers work in planetary magnetospheres but are useless for deep-space asteroid operations. For the next decade, the realistic choice is SEP.

Capture mechanisms

Reaching the rock is only half the problem; holding it is the other. The Asteroid Redirect Mission Option A proposed a 15-meter inflatable capture bag with drawstring cinch lines to enclose an entire ~7-meter rubble pile. Option B proposed microspine grippers and an anchor drill to capture a boulder from a larger asteroid. Philae tried harpoons and failed because the surface was stronger than expected. Microspine grippers, inspired by gecko feet, can conform to convex and concave surface features. For contactless options, the gravity tractor uses the spacecraft’s own mass to tug the asteroid, while the ion-beam shepherd directs a plasma stream at the surface to impart momentum. For a small captured minimoon with surface gravity measured in micrometers per second squared, almost any of these can work — and almost any of them can fail if the surface cohesion or rotation state is mischaracterized.

Keeper math

For a captured minimoon of 1–10 tonnes, a 40 kW-class SEP tug is directly scalable from the ARM/KISS studies; the smaller target means either lower propellant load or shorter flight time. A hybrid architecture is most plausible: SEP for the heliocentric chase and return, chemical or cold-gas thrusters for proximity operations, despin, and final docking. The capture itself requires negligible ΔV — escape velocities are centimeters per second — but matching the target’s rotation and irregular gravity field requires autonomy and patience.

Power is the binding constraint. A 40 kW SEP tug needs roughly 50 kW of solar array area at Earth, scaling down with solar distance. For cislunar operations the arrays are manageable; for main-belt targets they become enormous. If the captured body is C-type and water-bearing, it can repay the propellant debt: 100 tonnes of carbonaceous asteroid might contain 10–20 tonnes of water, which electrolyzes into hydrogen and oxygen for chemical propulsion or life support. The tug then becomes a reusable logistics vehicle, and the desktop platform can refuel from the rock it helped park.

Recalled

  • The Forever War (Joe Haldeman, 1974). Haldeman’s interstellar war is fought with high-Isp drives and relativistic time dilation; soldiers depart Earth, age months, and return to find decades have passed. The engine technology is never the point of the novel, but it is the axis on which everything turns: the length of the chase, the mass of the ship, the shape of society left behind. A keeper tug is the same transaction at a smaller scale — years of low-thrust spiral, kilograms of xenon converted into kilometers per second, a captured rock delivered to a cislunar orbit where the economics may have shifted while it was en route. Haldeman’s soldiers pay in lost time; the keeper pays in propellant, power, and patience. The unifying truth is that propulsion is never just engineering; it is the contract between the mission and the calendar.

What this changes

  • The capture architecture is logged as SEP-primary, chemical-secondary: a 40 kW-class Hall/ion tug is the nearest-term scalable option for moving 1–1,000 tonne rocks, with chemical thrusters reserved for proximity and despin.
  • Capture mechanism is flagged as a separate risk from propulsion: the engine gets you there; a bag, microspines, harpoons, or a gravity tractor determines whether you come back with the rock.
  • ISRU closes the loop: a water-bearing C-type capture can repay propellant costs via electrolysis, turning the tug into a reusable cislunar tanker. This reinforces entry 123’s finding that C-type chemistry is valuable even if most minimoons are not C-type.
  • The desktop platform’s propulsion is decoupled from the tug’s: stationkeeping and reboost can use small SEP or electrodynamic tethers, while the heavy interplanetary ΔV belongs to a dedicated tug.
  • Timeline realism is updated: capture and return are multi-year operations. Any keeper business plan that assumes rapid retrieval is using chemical assumptions for a job that electric propulsion does better but slower.