Analysys Mason projects the in-orbit servicing, assembly, and manufacturing market at $15 billion by 2031. The refueling segment alone — propellant depots, space tugs, and transfer vehicles — is forecast to grow from $2.7 billion in 2024 to $8 billion by 2034. Three years ago these numbers did not exist in any credible forecast. They exist now because the critical constraint has shifted: no longer can the Pentagon afford to treat fuel as a launch-day fixed cost.
The U.S. Space Force is moving from studying on-orbit refueling to funding operational demonstrations — two confirmed missions in 2027, a $25.5 million Astroscale contract, a $54.5 million Starfish Space contract, and a SpaceWERX challenge for multi-orbit propellant warehouses.
Why it matters
Every satellite today launches with a lifetime fuel supply. On-orbit refueling turns satellites into reusable assets — changing constellation economics, orbital slot value, and the strategic logic of space warfare.
What to watch
Astroscale Provisioner (summer 2026 launch), Northrop Grumman Elixir refueling payload, Orbit Fab RAFTI interface adoption, Starfish Space US-Otter 1 tug, Quantum Space fuel depot
In-orbit refueling market projection
Compounding at 11–12% annually, driven by defence demand for sustained satellite maneuverability. · GMI Insights, 2024
Why the Pentagon Is Buying Fuel Stations
A geostationary communications satellite costs $300 million to $600 million to build and launch. Its operational life is defined not by its electronics but by its propellant tank — typically 15 years of station-keeping, after which a perfectly functional spacecraft is drifted into a graveyard orbit and replaced. The U.S. Space Force operates dozens of such assets. The math of disposing and rebuilding versus refueling is not close.
In September 2023, Space Systems Command awarded Astroscale U.S. a $25.5 million contract to build the Astroscale Prototype Servicer for Refueling (APS-R), since renamed Provisioner. The 300-kilogram spacecraft is designed as a standard ESPA-class satellite, carries hydrazine, and will dock with client vehicles in geostationary orbit to transfer fuel. It committed another $12 million of its own money. Southwest Research Institute was selected to build the bus.
The bet has expanded since. In February 2026, the Space Force awarded Starfish Space a $54.5 million contract under the APFIT program for a dedicated Otter servicing vehicle, scheduled for 2028 delivery. In June 2026, Quantum Space secured a Department of War OECIF contract for a fuel depot based on its Ranger platform. And the Defense Innovation Unit's RAPIDS initiative is funding both Orbit Fab's RAFTI refueling interface and the broader mission architecture.
Summer 2026 will see the first live test: Astroscale's Provisioner will launch, rendezvous with an AFRL Tetra-5 spacecraft, transfer hydrazine, then return to an Orbit Fab propellant depot to refuel itself — and repeat the operation. It will be the first on-orbit refueling of a U.S. Department of Defense asset.
The Interface War Already Has Two Winners
The Space Force has approved two competing refueling port standards: Orbit Fab's RAFTI (Rapidly Attachable Fluid Transfer Interface) and Northrop Grumman's PRM (Passive Refueling Module). Both are now designated as acceptable commercial solutions for military satellites. This dual-standard approach is deliberate — the service wants to avoid vendor lock-in while keeping enough commonality for industry to invest.
Orbit Fab moves fastest. Its RAFTI port is already flying on the Tetra-5 demonstration satellites. The company has the first commercial fuel depot in orbit (Tanker-001 Tenzing, launched 2021) and a $25.5 million DIU contract for the first in-space fuel sale in geostationary orbit. In March 2026, Orbit Fab announced the RADICAL project with Airbus Defence and Space, funded by the UK Space Agency and ESA, to integrate RAFTI into future European telecom satellites. The company's RAVEN shuttle and NEST depot, introduced in 2026, expand the architecture from single-port refueling to a depot-shuttle-client ecosystem.
Northrop Grumman takes a complementary track. Its SpaceLogistics subsidiary has already demonstrated the model: MEV-1 and MEV-2 have provided nearly a decade of combined on-orbit mission extension services for Intelsat. The company's Elixir refueling payload program, funded by a $70 million Space Force contract, will demonstrate PRM-based refueling through a vehicle hosted on an ESPAStar platform. The Mission Robotic Vehicle (MRV), targeting summer 2026 launch, carries robotic arms developed with the Naval Research Laboratory and NASA for modular on-orbit servicing.
The dual-standard approach creates a predictable short-term inefficiency — satellite integrators must support both RAFTI and PRM — but it accelerates the long-term goal: refueling as a service, bought by the government rather than integrated bespoke per mission.
The Economics of Not Replacing Satellites
The market numbers tell a consistent story across multiple analysts. GMI Insights valued the in-orbit refueling and life-extension segment at $2.7 billion in 2024, projecting $8 billion by 2034 — an 11–12% compound annual growth rate. Analysys Mason puts the broader ISAM market at $15 billion by 2031. Both projections assume one thing that has not yet happened: the first successful commercial refueling of a government satellite, demonstrating that the model works operationally, not just on paper.
Astroscale's Provisioner mission is that demonstration. The spacecraft will prove three interdependent capabilities: rendezvous and docking with a client not designed for servicing, hydrazine transfer in geostationary orbit, and the depot-shuttle-client logistics loop. Each of these has been demonstrated individually in prior missions (MEV proved docking, Orbit Fab proved depot operations), but never as an integrated chain.
The financial inflection point comes when the Space Force shifts from buying bespoke refueling demonstrations to contracting refueling as a service. Col. Scott Carstetter, SSC's director of servicing, mobility and logistics, told reporters in May 2026 that the upcoming demonstrations are explicitly meant to inform a service-based acquisition model — "how we can actually use them operationally with programs and operational clients that we have in need today."
The same logic applies to commercial satellite operators. A geostationary telecom operator facing a $400 million replacement cost and a 36-month build cycle will pay a significant premium for a refueling service that extends the existing asset by five to seven years. The premium is calculable. The alternative is not.
The strategic dimension reinforces the economic one. Satellite constellations are force multipliers — communications, reconnaissance, navigation — and each one is only as resilient as its fuel supply. The Space Force's stated doctrine of dynamic space operations requires the ability to maneuver satellites on demand without counting down to a fuel depletion that ends the mission. On-orbit refueling eliminates that countdown. It is the difference between a navy that can stay at sea and one that must return to port after every patrol.
Space Tugs: The Delivery Trucks of the Orbital Economy
Propellant depots are warehouses. Space tugs are the delivery trucks that move fuel and components between them. The distinction matters because the logistics architecture is not settled — the Space Force is deliberately keeping both models in play.
Starfish Space's Otter is the pure tug: an autonomous vehicle designed to dock with any satellite regardless of refueling interface, then maneuver it to a new orbit, graveyard, or servicing position. The $54.5 million APFIT award is for the second Otter acquired by SSC, following a 2024 STRATFI award and an SDA contract for end-of-life disposal services. The Otter does not carry its own fuel for transfer — it moves the client.
Quantum Space's Ranger fuel depot takes the opposite approach: a stationary orbital warehouse that stores propellant and transfers it to visiting spacecraft. The June 2026 DoW OECIF contract funds a depot demonstration on the highly maneuverable Ranger platform, explicitly for national security space operations.
Provisioner splits the difference — it is both a servicer and a shuttle, carrying fuel to clients and returning to a depot for replenishment. This hybrid model may prove to be the most practical near-term architecture: it requires one depot (stationary, simple) and multiple servicers (mobile, complex) rather than a full mesh network.
The SpaceWERX challenge launched in June 2026 is designed to explore all three architectures simultaneously. It calls for multi-orbit propellant warehouses, autonomous orbital transfer vehicles, and distribution network mechanics — proprietary orbital modeling to compute optimal shuttle routes. Two fully funded pathfinder missions are already manifested for early 2027 on a ULA Vulcan Centaur rocket (USSF-23).
Signals to Track
1. Provisioner launch date. Summer 2026 is the target. Any slip pushes the entire refueling-as-service timeline by at least a year — there is no backup mission with equivalent scope.
2. Interface standard convergence. If the market consolidates around one of RAFTI or PRM within 18 months of the first successful refueling, depot economics improve dramatically. If both persist, the logistics chain stays fragmented.
3. Commercial operator adoption. The first commercial GEO operator to sign a refueling service contract — not a demonstration, a paid service — will signal whether the model works outside defence procurement.
4. China's Shijian-25 program. China performed its first on-orbit refueling in GEO in June 2025. The U.S. program is two years behind on demonstrated capability, which compresses the timeline for operational deployment.
