A Propellant Depot Is a Cryogenic Storage and Transfer System

An orbital propellant depot is useful only if it can receive, preserve, measure, and deliver enough flight-ready propellant on the customer’s schedule. Launching a tank of cryogenic liquid solves none of those verbs by itself. Heat leak creates vapor and pressure, low gravity changes fluid position, transfer hardware must be chilled and connected, and every loss reduces the mass the customer can actually burn.

This article concerns cryogenic liquids such as oxygen and hydrogen. Storable chemical propellants and electric-propulsion fluids have different storage and transfer problems.

Heat leak turns inventory into a pressure problem

Cryogenic propellants remain liquid only below their boiling temperatures at the relevant pressure. Sunlight is not the only heat source: radiation from Earth, warm spacecraft hardware, tank supports, plumbing, cables, and transfer connections all create paths into the fluid.

As liquid vaporizes, tank pressure rises. Venting can control pressure but discards inventory. NASA’s current Zero Boil-Off Tank research overview describes alternatives that mix the fluid and use passive or active cooling to condense vapor and enable reduced- or zero-boiloff storage. It also reports that microgravity experiments exposed non-intuitive boiling, ullage motion, cavitation, and pressure behavior that simple Earth-gravity assumptions did not predict.

“Zero boiloff” is therefore a system operating point, not a property of insulation. Cryocoolers consume power and reject heat, pumps add hardware and failure modes, and pressure control must work across fill levels, disturbances, and non-condensable gases. For short storage, accepting some vent loss may be lighter than active cooling. For a long loiter or scarce propellant, refrigeration may repay its mass and power cost.

Inventory is hard to measure in low gravity

A terrestrial tank gauge often assumes gravity places liquid below vapor. In orbit, surface tension, acceleration, slosh, and tank geometry determine where the liquid sits. Pressure alone does not uniquely reveal mass when temperature and two-phase conditions vary.

The depot needs a gauging method with a stated accuracy across its operating envelope. NASA’s Radio Frequency Mass Gauge work uses a tank’s electromagnetic response to estimate cryogenic quantity; NASA reports that the method verified propellant levels on the 2024 IM-1 lunar mission. Other methods may infer inventory from thermal response, bookkeeping, or settled conditions, but each needs calibration and uncertainty.

That uncertainty becomes reserve mass. A depot that cannot confidently distinguish deliverable liquid from vapor, trapped residuals, or inaccessible liquid must hold extra propellant or risk short-loading a customer.

Transfer is a thermal transient

Warm pipes and couplers initially boil part of the incoming liquid. Chilldown consumes propellant or refrigeration before stable liquid transfer begins. The system must acquire liquid at the source tank outlet, prevent unacceptable vapor ingestion, control pressures in both tanks, measure transferred mass, and leave the customer within its temperature and pressure limits.

Low gravity provides no passive guarantee that liquid covers the outlet. Settling thrust, surface-tension devices, pumps, or other liquid-acquisition methods can establish flow, each with penalties. Settling expends propellant and affects docking loads; pumps require power and add machinery; passive devices have flow and geometry limits.

The connection is another subsystem. It must align, seal, tolerate thermal contraction, disconnect without hazardous leakage, and repeat the cycle. NASA stated in June 2026 that in-orbit cryogenic refueling between two spacecraft had not yet been performed. Its developmental cryocoupler testing addressed automated mating, misalignment, repeated connection, flow, and temperature differences. Ground launch-pad couplers are not automatically suitable for this job.

Measure useful throughput, not tank capacity

Nameplate capacity can hide an underperforming depot. Track propellant from launch vehicle separation to customer departure:

delivered mass = received mass
               - storage loss
               - conditioning and chilldown loss
               - transfer and disconnect loss
               - inaccessible residuals
               - required depot reserve

Then divide delivered mass by time and by the resources that matter to the architecture. A useful performance sheet includes:

Metric	Why it matters
Net delivered mass per cycle	What the customer can load, after all losses
Delivery uncertainty	Extra reserve needed to guarantee the load
Maximum and minimum loiter	Whether launch delays can be absorbed
Turnaround time	How quickly another tanker or customer can be served
Average and peak power	Solar-array, battery, and thermal-system sizing
Loss per day and per transfer	Separates storage performance from handling performance
Compatible vehicles and fluids	Whether interfaces, cleanliness, pressure, and temperature match
Abort and safing capacity	How the depot handles a leak, stuck valve, failed cooler, or missed docking

A high-throughput depot may accept more daily loss if propellant moves quickly. A strategic reserve needs very low loss and long-lived reliability. One design does not optimize both automatically.

Demonstrations must close the integrated loop

NASA describes Cryogenic Fluid Management as the technologies needed to store, transfer, and measure ultra-cold fluids. Its May 2026 LOXSAT plan grouped boiloff reduction, transfer, pressure maintenance, and gauging into an orbital liquid-oxygen demonstration. That integration is important because success in an isolated component test does not prove the sequence works as a service.

A representative demonstration should receive or condition fluid, loiter through thermal cycles, determine inventory independently, acquire liquid, chill the path, transfer a measured amount, disconnect, and reconcile the mass balance. It should repeat the operation at different fill levels and after credible delays or component faults.

The hardest depot requirement may be schedule tolerance. A delayed tanker can leave the depot empty; a delayed customer can extend storage beyond the planned thermal case. Launch cadence, rendezvous opportunity, maintenance, and spare capacity therefore belong in the same model as insulation and pumps.

Calling a depot a gas station is a useful analogy only at the service boundary. Underneath, it is a long-duration cryogenic spacecraft whose product is a verified mass of conditioned propellant at a particular interface and time.