Research Note: Orbital Refueling Moves the Bottleneck
Published May 2026
Question
Orbital refueling is one of the clearest examples of a lunar-cost trade. It can reduce the amount a single vehicle must launch with, but it adds depot, tanker, transfer, storage, boiloff, rendezvous, and schedule dependencies. CosmosExplore should model it as a logistics system, not as a yes/no feature.
Source-Backed Data Points
- NASA's Artemis III page describes a sequence in which reusable tankers carry propellant to a storage depot, and the uncrewed Starship human landing system fills its tanks before traveling to near-rectilinear halo orbit. Source: NASA Artemis III.
- NASA reported that a Starship test objective tied to Artemis operations involved transfer of thousands of pounds of cryogenic propellant between internal tanks during coast phase. Source: NASA Artemis mission progresses with SpaceX Starship test flight.
- GAO found that SpaceX must complete complex technical work for Artemis III, including the ability to store and transfer propellant while in orbit. Source: GAO-24-106256.
Reading
Refueling can be a cost reducer and a risk concentrator at the same time. The model has to ask how many launches are required, what launch cadence is needed, how long propellant waits, what transfer losses are assumed, what abort modes exist, and how failures propagate through the chain.
The right unit is not just kilograms of propellant. It is kilograms delivered to the right orbit, at the right temperature, with enough schedule confidence to support crew operations. If any one of those words becomes uncertain, the cost model needs a penalty or an unknown field.
Model Rule
CosmosExplore will represent refueling with explicit tanker count, transfer efficiency, storage duration, boiloff assumption, and depot fixed cost fields before it treats refueling as a mature cost reducer. Until those fields are sourced, refueling remains a scenario, not a baseline.