Entry 129 wondered about boiling captured volatiles to cool the computer. This sweep asks the practical question underneath that speculation: how do you actually get the volatiles out? The answer is that a captured C-type body is less like an ice cube and more like a wet stone. The water is chemically bound in phyllosilicates; freeing it takes heat, engineering, and time.

What is in the rock

The sample-return missions set the bounds. Ryugu contains roughly 6.8 wt% water equivalent, mostly in serpentine and saponite phyllosilicates that make up ~90% of its volume. Bennu has ~80 vol% Mg-rich phyllosilicates, 4.5–4.7 wt% carbon, 0.23–0.25 wt% nitrogen, and hydrogen content around 0.86–0.98 wt%. The water is not free ice waiting to be melted; it is OH and H₂O locked in mineral structures. That distinction dominates the energy budget. Melting and vaporizing ice costs about 2.9 MJ/kg. Dehydrating serpentine requires heating the silicate matrix to roughly 600–800 °C and supplying the bond-breaking enthalpy. The rock must be cooked, not just warmed.

Optical mining

TransAstra’s approach is the most developed concept. Enclose the asteroid in a thin-film inflatable bag, focus concentrated sunlight through non-imaging optics into a cavity, and heat the surface until volatiles escape and thermal shock spalls off fragments, exposing fresh material. The water vapor is cryopumped into passively cooled storage bags. Lab demonstrations at White Sands and elsewhere have used 8–12 kW solar concentrators on CI-like simulants in vacuum chambers and successfully trapped water. A 10-meter solar collector test on synthetic carbonaceous material showed the host rock fracturing as volatiles escaped. The method is physically plausible; the engineering details are where the risk lives.

A 2025 Acta Astronautica paper on Nectar simulant tests reports that beam profile, mineral morphology, and organic vapor deposition strongly affect excavation and water production. Organics released during heating can redeposit on the surface and slow the process. Broader, more uniform beams work better. The surface temperatures in optical-mining models can reach ~1000 K. These are not gentle processes; they are controlled pyrolysis in a bag.

Energy and throughput

No published study gives a clean kg-water-per-kW-per-day figure for asteroid optical mining. The closest benchmark is lunar ice extraction: the Aqua Factorem NIAC study estimated that 800 kW thermal could extract 2,450 tonnes of water per year from regolith containing 5 wt% ice — roughly 8 tonnes of water per installed thermal kilowatt-year. Asteroid phyllosilicates will be worse because 90% of the mass is silicate that must be heated to dehydration temperature. A rough scaling suggests 5–10 MJ per kilogram of water for a carbonaceous body, several times the lunar-ice benchmark. A 10-meter concentrator at 1 AU collects on the order of 100 kW of solar energy; at that scale, daily water production might be tens to a few hundred kilograms, depending on efficiency and asteroid size.

From water to propellant and products

Extracted water can be electrolyzed into hydrogen and oxygen. Commercial electrolysis uses roughly 50–55 kWh per kilogram of hydrogen; the thermodynamic minimum is ~39.4 kWh/kg. One kilogram of water yields ~0.11 kg hydrogen and 0.89 kg oxygen. If carbon or CO₂ is available, the Sabatier reaction can produce methane and regenerate water, closing a methane/oxygen propellant loop. Oxygen can also be extracted directly from regolith by molten regolith electrolysis at 1600–1800 °C, a path Blue Origin is pursuing with its Blue Alchemist system, which has produced silicon, iron, aluminum, and oxygen from lunar simulant.

Other products matter too. A 500-tonne carbonaceous asteroid, per the Keck Institute study, might contain roughly 83 tonnes iron, 6 tonnes nickel, 1 tonne cobalt, and 1 tonne platinum-group elements in residue, plus ~200 tonnes of volatile and carbon-rich compounds. Sintered regolith can reach compressive strengths comparable to concrete and serve as radiation shielding or construction material. The Mond process can extract iron and nickel as gaseous carbonyls at moderate temperatures. The chemistry is rich, but each step adds mass, power, and complexity.

State of the art

TransAstra’s APIS family scales from Mini Bee (250 kg tech demo) through Honey Bee (~5 tonne spacecraft to extract ~100 tonnes of water from a 10-meter asteroid) to Queen Bee for 40-meter bodies. NASA’s PRIME-1 and VIPER are lunar demonstrators, not asteroid, but they test drill-based volatile extraction at the Moon’s south pole. MOXIE on Mars proved in-situ oxygen production from CO₂. ESA’s PROSPECT package aims to do the same for lunar volatiles. The asteroid-specific hardware is at lower technology-readiness level than the lunar analogues, but the physics is the same: heat the material, collect the volatiles, process them.

Keeper math

For a 1–10 tonne captured minimoon at 7 wt% water, the total available water is 70–700 kg. That is enough to cool the desktop for months in an open-cycle system, or to produce a few hundred kilograms of propellant, but it is not a vast resource. The extraction plant must therefore be very small — perhaps under 100 kg — to achieve mass payback quickly. A TransAstra Honey Bee-scale system is far too large. The keeper needs a miniature optical-mining cavity or a beneficiation approach that avoids heating the entire rock matrix.

Time scale is another constraint. If a compact plant extracts 1–10 kg of water per day, the first tonne takes 100–1,000 days. For a small volatile-rich capture, ISRU is a long-lead activity, not a quick refuel stop. And there are contamination risks: vented water, CO₂, sulfur compounds, and dust create a localized plume around the processing platform that can coat radiators, optics, and solar arrays. Entry 122’s radiator problem and entry 129’s open-cycle cooling idea both run into this: the heat sink and the contamination source are the same object.

Recalled

  • The Martian (Andy Weir, 2011). Weir’s Mark Watney survives by treating Mars as a chemistry set: hydrazine into hydrogen, hydrogen into water, water into potatoes, Martian soil into growth medium, every calculation checked against mass and energy budgets. The book’s hero is not bravery but competence — the willingness to cook dirt and trust the numbers. The ISRU sweep is the same temperament applied to a captured asteroid. The rock is not a gift; it is a set of reagents that must be heated, cracked, electrolyzed, and stored without killing yourself. Watney’s lesson for the keeper: the most valuable resource is not water or carbon; it is the ability to model the process before committing the mass and power.

What this changes

  • ISRU is logged as a long-lead, high-temperature process, not a simple ice-melt. The binding energy of phyllosilicate-bound water dominates the power budget.
  • Optical mining is flagged as the leading extraction concept but with no published asteroid throughput numbers; any keeper business plan using it must treat energy-per-kg-water as an uncertainty to be measured, not assumed.
  • Plant scale is a hard constraint for small captures: a 1–10 tonne minimoon yields only 70–700 kg water, so the extraction hardware must be miniaturized, or the keeper must target larger bodies.
  • Product hierarchy is clarified: water for cooling/propellant/life support is the easiest value; oxygen via electrolysis is next; metals and PGEs require more aggressive processing; sintered regolith for shielding is the simplest structural product.
  • Contamination is added to the risk list: volatile extraction creates dust and gas plumes that threaten radiators, optics, and solar arrays, and may have no clear regulatory framework in cislunar space.