Static Mass Station vs. Reconfigurable Station Economics

Summary

Avalon turns orbital infrastructure from a series of fixed capital projects into a reusable production platform. Standard Elogium blocks are joined in orbit and rebuilt around each tenant's equipment.

• The problem. A fixed station commits its mass to one configuration before launch. It costs about $7–8M per m³ to build, every expansion costs about the same again, and its floor costs about $8.3M per m³ per year. That is 43 to 290 times what Earth-scale production earns per m³.
• The innovation. Novo Navis weaves standard structural blocks at production scale and joins them into pressure structure in orbit with carbon nanotube welding. Bays are built around the equipment and rebuilt when a tenant's process changes, and every tenant draws on one shared power, cooling and life-support backbone.
• The economic consequence, including the core platform. Per rented m³ per year, with the core platform, expansion, capital and operating cost all included, against a fixed station's $6.6M:

• Cheaper launch helps Avalon most. Launch is most of Avalon's expansion cost and a small slice of a fixed station's. Falling launch prices cut Avalon's expansion cost by about 90%, and a fixed station's by about 5%.
• Shared power costs a fraction of today's price. Avalon's shared solar power costs about $4.80 per kWh today and $2.70 at $1,000/kg launch. The ISS's price to commercial users is $42 per kWh.

In this document, Avalon (A.C.S.) is the station and Novo Navis is the company that builds and operates it.

The production gap

A fixed-module station must charge about $8.3M per m³ per year, 43 to 290 times what an Earth-scale production floor earns.

No commercial station publishes floor pricing, so the price is rebuilt from cost.

• Cost basis: Starlab's own estimate of $2.8–3.3B to design, build and launch about 400 m³ of pressurized volume, or $7–8M per m³.
• Break-even: paying off $3.05B over 30 years at a 15% return, plus operating costs of 7.5% of build cost a year, spread over 105 rented m³ (150 m³ usable, 70% rented), comes to about $6.6M per m³ per year.
• Price: a 15–25% margin gives about $8.3M per m³ per year, with a range of $4–14.5M.

A fixed-station floor costs more each year than everything production equipment on it could earn. That is why today's stations sell research space, not production space.

Fixed stations can't get cheaper

A fixed station's cost per m³ is the same at every size, because each expansion is another module bought at the original cost.

• The cost per m³ never falls. Doubling a fixed station doubles its cost, so its floor price can't drop as it grows.
• The size limit never lifts. Every module is capped by the hatch it connects through (1.3 m on the ISS) or by its rocket's fairing (5.2 m for Falcon 9, about 9 m for Starship). Production equipment bigger than that has nowhere to go.
• Starship is the cheapest fixed option, and it is still fixed. Each ship is a 9 m-wide container whose cost repeats with every ship added.

Stranded mass

Every kilogram launched into a fixed module is locked into one shape, one hatch size and one purpose until it is thrown away.

• Fixed modules carry 137–311 kg of mass per m³ of volume: Destiny 138, Columbus 137, Harmony 189, Haven-1 311. None of it can be reshaped when a tenant leaves, a process changes or a bigger machine arrives.
• When the station ends, all of its mass ends with it. The ISS weighs 430 t. NASA is paying SpaceX $843M just for the vehicle that will push it into the ocean. At today's launch price of about $3,250/kg, that mass took about $1.4B in launch alone to put up.
• Short-life stations strand mass sooner. Haven-1 is designed for about 3 years. Every kilogram it launches is written off on that schedule.
• Cheaper launch doesn't rescue stranded mass. It makes the next module cheaper. It does nothing for mass already flown, which keeps its shape, its limits and its cost until it is deorbited.

Capital has to earn back the launch it rode on

When launch gets cheaper, a fixed station can match a newer rival's price only by writing off capital it has already flown.

It works like gasoline already in the tank: a lower price at the pump next week doesn't refund what you paid. A fixed operator can still cut its price. The difference comes out of the return on its capital.

• Old capital competes with new. A module flown today carries $445k–1.0M of launch per m³. A rival flying the same module at $1,000/kg carries $137k–311k. To match the rival's price, the first operator writes off the difference.
• Each module earns only from its own tenants. A fixed module's capital is paid back by whoever can use that shape, until it is deorbited. When a tenant's needs change, the module has no other use.
• Avalon's capital earns across every configuration. Blocks are re-emplaced when needs change, so each kilogram's capital keeps earning through every tenant it serves. In the two-tenant example below, 79% of the second tenant's structure cost is avoided by reusing the first tenant's blocks.
• Avalon buys each expansion at that day's launch price. Its average cost per m³ falls as launch prices fall, rather than being set by the price of its first launch.

The depreciation mismatch

On Earth, a fab building lasts 30–50 years while the tools inside it turn over every 3–5 years. Owners depreciate the building and the tools separately, and the building outlives many tool cycles.

A fixed station fuses the building and the tool envelope into one module. A tenant who signs a 3-year lease for a purpose-built module leaves the operator holding 27 years of hull that may suit no one else. Avalon separates the two the way an industrial landlord does: Elogium structure is paid off over 30 years across 6–10 successive tenant tool cycles, and each cycle's bay is rebuilt from the same blocks.

Reconfigurable mass outclasses fixed mass

Avalon adds enclosed volume for $102k–564k per m³, 13 to 75 times less than a fixed module's $7.6M, and reuses that structure across tenants.

What an expansion costs

Three measures, kept separate

The ranges run from 80 kg of kit per m³ enclosed (20 m radius) to 160 kg (10 m radius).

What the first platform costs

Before the first expansion, Avalon needs a core: power, thermal control, life support, docking, robotics and welding systems. That core is inferred at $500M–1B, or $114M–227M a year once paid off and operated. Spread across rented floor, it costs $162k–645k per rented m³ in year one. By year five it falls to $14k–55k as one loom's output fills in around it.

Worked example: two tenants, same structure

• Tenant A runs a crystallization suite in a 100 m³ bay, built from 16 t of blocks.
• Tenant A leaves. Tenant B needs 120 m³, including headroom for a 12.7 m silicon crystal puller.

Why it holds

• Cost per m³ falls with size. A bigger enclosure needs less wall per m³ inside, because volume grows faster than surface. Doubling the radius halves the kit mass per m³.
• No size limit. Bays are built around the equipment, so Earth-scale tools go up unchanged. They aren't shrunk to fit a hatch and requalified, which takes about 4 years for a new production line.
• Nothing is stranded. Blocks are moved and re-emplaced when a process changes, so the mass keeps earning through every configuration.
• The station feeds itself. Avalon makes methane, recycles water, grows biomass and produces Elogium feedstock on orbit. Every kilogram it makes is one a fixed station must buy as resupply at $20,000–70,000/kg.

Shared utilities: power, cooling and life support

Avalon runs one power, cooling and life-support backbone for its own production and shares it with every tenant. Its power costs about $4.80 per kWh today, against the ISS's price to commercial users of $42 per kWh.

A fixed station sizes its utilities to itself: Starlab's entire 35 kW can't run one 50 kW crystal furnace. Avalon adds arrays and radiators with each expansion, mounted on the same structure, so power grows with floor space.

• The biorefinery is the station's battery. Solar arrays are the primary power source. The anaerobic biorefinery turns biomass into methane, which fuel cells burn through each eclipse in place of batteries, and which also serves as propellant.
• Waste is shared too. Furnace waste heat warms the biorefinery's digesters. CO₂ from crew and processes feeds the biomass.
• One plant beats many small ones. A single large power, thermal and life-support plant shared across tenants costs less per kW than each fixed station running its own small system, the way an industrial park does on Earth.

What Novo Navis built to make this real

Two Novo Navis technologies turn reconfigurable mass from an idea into a pressurized production floor: a scalable 3D carbon fiber weaving process, and carbon nanotube welding.

Scalable 3D-woven carbon fiber (Elogium)

• Repeatable structural units. The process produces solid, void-free blocks of a fixed size, 14 × 14 × 36 inches. A repeatable unit is what makes a building standard possible: every block is interchangeable, and every bay is assembled from the same parts.
• Production scale. A single loom load yields about 250 tonnes of material a year. That supplies structure by the tonne, not part by part.
• Machinable aerospace structure. The blocks are dense enough to machine like metal, so cleats, interfaces and hardware are cut into them directly. They are 2.4 times thicker than conventional 3D-woven preforms.
• Proven origin. The process was developed under an Air Force SBIR Phase I award.

Carbon nanotube welding

• Blocks become pressure structure. Novo Navis's carbon nanotube welding joins composite structures on orbit. Individually launched blocks become continuous, sealed walls, which turns modular construction into pressurized modules.
• Bays can be added and changed. Because joints are made on orbit, the station's shape isn't fixed at launch. Bays are built around the equipment that goes in them and rebuilt when a tenant's process changes.

Together, the weave supplies standard structural units at production scale, and the weld makes them airtight in orbit. That is what lets Avalon grow without the hatch and fairing limits of every fixed station.

The moat

Novo Navis holds a two-layer technical moat, and a competitor would need both layers to build this way.

• The weave can't be bought. The loom solves the problem that keeps 3D-woven composites small everywhere else. The matrix resin is a proprietary in-house blend held as a trade secret. Together they produce void-free, machinable blocks at production scale.
• The weld can't be bought. Carbon nanotube welding of composite structures is a Novo Navis process, also held as a trade secret. Without it, woven blocks are building material. With it, they are a pressure hull assembled in orbit.
• The pair is the moat. Blocks without the weld can't hold pressure, and a weld without the blocks has no standard structure to join. A competitor would have to independently solve both, then prove them in orbit.
• SBIR Phase III rights. The Elogium SBIR award carries Phase III rights. Federal agencies can award Novo Navis follow-on work that extends or completes the SBIR effort without a new competition.

Industry by industry

At today's launch price, Avalon floor costs less per unit than the ground price for ZBLAN fiber, GaN wafers and SiC substrates. At $1,000/kg it does for six of nine products. A fixed station's floor costs 3 to 2,200 times the ground price for all nine.

Each row runs the same Earth-scale production equipment on each floor. Figures are floor cost per unit of product, the share of production cost set by the station, per rented m³ with capital and operating cost included. Avalon figures are the midpoint of the 80–160 kg/m³ range.

Total landed cost: floor, power and shipping

A producer pays for floor, power, and shipping up and down. At today's launch price, ZBLAN fiber and GaN wafers come in below their ground price with all three included. Biologics come in at a fraction of a percent of the drug's selling value.

• First customers: ZBLAN fiber, GaN wafers and biologic crystallization. Their product is valuable enough per kilogram to carry floor, power and shipping today.
• Next: SiC substrates and SiC epi. They need about a 3× orbital premium today, and less as launch and return costs fall.
• Silicon, metal powder, separators and cell therapy need return costs to fall, or the product to be used in orbit, before they close. For silicon, shipping alone is 98% of the landed cost.

What return-shipping price does to the ranking

If the price of bringing product down drops from $47,000/kg to $2,000/kg, SiC moves to about 2× the ground price and silicon from 60× to 8×. ZBLAN and GaN barely move, because their product weighs almost nothing per unit of value.

Each cell is the landed cost as a multiple of the ground price, at today's launch price.

ZBLAN and GaN depend on high value per kilogram, not on cheaper return. SiC and silicon depend on both cheaper return and cheaper floor. The base case uses $47,000/kg throughout.

What Avalon earns as launch costs fall

Falling launch prices cut Avalon's floor cost by about 90% and a fixed station's by about 5%, so every drop widens Avalon's lead.

A fixed station's cost is mostly hardware, and launch is a small slice of it. Avalon's expansion cost is mostly launch, so it falls nearly in step with launch prices.

Avalon's expansion floor costs less than a Starship used as a station at every launch price above. A ship is also capped at about 9 m wide, so equipment like a 12.7 m crystal puller can't go in it at any price.

A ship is also sold whole. A GaN producer needs about 15 m³ and a ZBLAN producer about 35 m³, and neither can finance a 350 m³ rented ship alone. Putting several industrial tenants in one open hull means one tenant's leak, vibration or delay stops all of them. Avalon leases right-sized bays, from about 12 m³ to 500 m³ and up, on a shared utility grid. Each tenant rents only what it needs and expands bay by bay.

What Avalon can charge, and what it keeps

Avalon leases floor the way an industrial landlord does: base rent per m³, with power, cooling and shipping metered and billed separately at cost. Base rent is set so that it costs the client 20% of what their production earns on the ground. Biologics are the exception, priced on product value at about 1% of the drug value each m³ produces.

ZBLAN and biologics are profitable anchor tenants at today's launch price. No industry in the table can pay a fixed station's price for production. Falling launch costs open industry after industry to Avalon, and none of them to fixed stations.

From loom output to profit

Capacity becomes revenue in six steps: weave, kit, launch, assemble and weld, commission, rent. One loom makes 250 t of structure a year: 1,560–3,125 m³ of new enclosed volume, carried by 15–19 Falcon 9 launches or about 3 Starship launches.

Profit here is operating profit after the capital charge: base rent, minus 22.7% a year of the build cost of all capacity built (rented or not), minus the same charge on the core platform. Taxes, financing costs and working capital are not included. Power is billed at cost and nets to zero.

Two pricing cases at $1,000/kg launch. GaN-class rent is $133k per m³ per year. The blended rent of $250k assumes early bays go to the highest-paying tenants first: ZBLAN, biologics and federal anchor tenants.

• Light build (80 kg/m³, $500M core) passes a fully rented Starlab-class station's profit of about $173M a year in year 3 at GaN-class rent, and in year 2 at blended rent.
• Heavy build (160 kg/m³, $1B core) turns profitable in year 3 at blended rent and recovers its cumulative losses in year 4.
• A fixed station's profit is capped by what it launched once. Avalon's grows with every loom-year of new floor.

Assumptions for this section:

• Avalon costs: (Elogium $100 + kit $75 + assembly and welding $100 + launch) per kg × 80–160 kg per m³ enclosed. Paid off over 30 years at 15%, plus operating cost of 7.5% of build cost a year: 22.7% a year in all. 75% of volume usable.
• Occupancy: 30% of usable volume rented in year 1, 50% in year 2, and 70% from year 3.
• Core platform: $500M–1B, on the same 22.7% annual charge.
• Starlab-class costs: $3.05B including about $150M of launch, which scales with launch price. 150 of 400 m³ usable, 70% rented. Profit is a 20% margin on 105 rented m³ at $8.25M.
• Starship used as a station: about $90M of ship hardware plus 100 t lifted at the stated launch price. 1,000 m³, 50% usable, 70% rented.
• Base rent: each product's ground revenue per m³, from the equipment space and output in the assumptions table. Biologics: 2 kg of drug substance per m³ a year at $25M/kg or more.

Fuel, food and material for future expeditions

Avalon also produces fuel, food and structural material in orbit, so it supplies the expeditions that come after it as well as its tenants.

• Fuel. Avalon's biorefinery makes methane on orbit. Every kilogram made there is a kilogram of propellant that never has to be lifted out of Earth's gravity.
• Food. The same biomass system that feeds the biorefinery grows food for crews on the station and for crews passing through.
• Material. Biomass is also the precursor to Elogium, so the station grows the raw material for its own expansion and for structures built further out.

Each kilogram made in orbit is worth at least what it would cost to deliver: $1,000–3,250/kg to reach orbit, and $20,000–70,000/kg delivered to a station. A fixed station must buy all three from Earth at those prices. Avalon makes them, which makes it a supply depot as well as a production floor.

The cost and profit figures elsewhere in this document count none of these savings, apart from methane as eclipse storage. Every Elogium kilogram in the model is made on Earth and launched. Fuel, food and material made in orbit add to those figures; the case doesn't depend on them.

The case is clear. Unlike fixed stations, Avalon gets cheaper as it grows, gains from every drop in launch price, never strands its mass, and makes the fuel, food and material the next expeditions will need.

Assumptions and sources

Every comparison in this document uses one basis: cost per rented m³ per year, capital and operating cost included, for both Avalon and the fixed station.

Assumptions

Power and shipping assumptions

Asset lives

Sources

• Voyager Technologies 10-K, FY2025: Starlab cost estimate
• Starlab one-pager: volume and power
• NASA commercial and marketing pricing policy: ISS prices
• NASA OIG IG-22-005: ISS cost
• Vast Haven-1 lab and Haven-1: volume, mass, life
• Vast large docking adapter
• Destiny module and NASA BEAM fact sheet: module mass and cost
• Thales Alenia Axiom modules contract
• NASA picks SpaceX to build ISS deorbit vehicle: $843M, 430 t
• SpaceX Dragon and Falcon Heavy: launch and cargo
• ALD VIGA atomizer brochure: atomizer size and output
• ISPE, cell therapy cost of goods: floor and cost per patient
• SiC epitaxial wafer market report: epi wafer prices
• Rosendahl Nextrom draw tower: fiber tower size
• Brückner separator film lines: line width and output