Launch and mission economics

Space economics is dominated by fixed costs, low volumes and long lead times. The unit economics behave differently from most industries as a result, and intuitions carried over from manufacturing or software mislead.

Where launch cost goes

A launch’s cost divides into categories with very different behaviour.

The propellant figure is worth dwelling on because it is so frequently misapplied. Propellant for a Falcon-class vehicle costs a few hundred thousand dollars against a launch price in the tens of millions. This is used to argue that launch is intrinsically cheap and current prices are markup.

That inference is wrong. The cost is in building a vehicle that survives the loads, in a workforce that can integrate and operate it, and in a range and recovery infrastructure that exists whether or not it flies. Propellant being cheap tells you only that chemical energy is cheap, which was never the constraint.

The reuse arithmetic

From Reusable launch systems, the cost per flight of a reusable vehicle is

with $V$ the vehicle build cost, $N$ flights per vehicle, $R$ recovery and refurbishment cost per flight, and $F$ amortised fixed recovery infrastructure.

Reuse beats expendable when

Two conclusions follow that are easy to state and often missed.

Refurbishment cost dominates, not flight count. If $R$ is 10% of $V$, the scheme is ahead after two flights. If $R$ approaches $V$, no value of $N$ helps. The Space Shuttle failed on this term, not on durability: the orbiter flew repeatedly, and returning it to flight status cost close to what new expendable hardware would have.

Cadence must cover the fixed term. Recovery vessels, landing pads and inspection facilities cost the same at five flights a year as at a hundred. At low cadence $F$ dominates and reuse loses on its own, independent of how cheap refurbishment is.

Satellite economics

The relevant metric depends on what the satellite sells.

For communications the transformation came from frequency reuse rather than from cheaper satellites. From Communications satellites, moving from a single wide beam to a hundred spot beams in a four-colour scheme multiplies capacity roughly twenty-five-fold from the same spectrum. A satellite costing somewhat more than its predecessor delivered two orders of magnitude more capacity, and cost per Gbit/s fell accordingly.

This is the general pattern in space economics: the large improvements come from architectural change rather than from manufacturing efficiency.

Constellation economics

A constellation’s economics are dominated by a term that single-satellite businesses do not have: continuous replacement.

With $t$ satellites and a design life $L$, steady-state replacement is $t/L$ per year. For 1584 satellites on a five-year life, that is about 317 per year, or six per week, sustained for as long as the service exists.

The user terminal line is frequently the most underestimated. A phased array terminal that must track fast-moving satellites and hand over between them is expensive, and consumer broadband economics generally require subsidising it and recovering the cost over a subscriber’s lifetime. That makes churn a critical variable in a way that has nothing to do with space.

The break-even condition is demanding: revenue must cover operating costs, continuous replacement, and the return on the initial deployment capital. That requires a very large subscriber base, and whether one exists at the required scale remains the central open question of the LEO broadband business.

Insurance

Space insurance is a small market with unusual dynamics, and it prices risk in a way that reveals what the industry actually believes.

Rates move sharply with recent loss history because the market is small enough that a few large claims exhaust a year’s premium across all underwriters. A single constellation failure or a launch vehicle anomaly can reprice the whole market.

Two structural features are worth noting.

Reliability is priced, and it is priced on demonstrated flights. A new vehicle pays a high rate until it has accumulated a record, which is a real barrier to entry that has nothing to do with engineering quality.

Large constellations often self-insure. With hundreds of satellites, the loss of one is a statistical event absorbed within normal operations rather than a claimable catastrophe. This has removed a substantial share of demand from the insurance market and concentrated remaining demand in high-value single satellites, which makes the market smaller and more volatile.

What actually reduces cost

Ranked by demonstrated effect rather than by attention received:

The consistent lesson is that cost reductions in space have come from changing what is built and how it is bought, rather than from making the same thing more cheaply.

References

• Jones, H. W. “The Recent Large Reduction in Space Launch Cost”, ICES-2018-81, 2018.
• Wertz, J. R. and Larson, W. J. (eds.) Reducing Space Mission Cost, Microcosm Press, 1996. Dated but structurally sound.
• Pielke, R. and Byerly, R. “Shuttle programme lifetime cost”, Nature, 472:38, 2011.
• Del Portillo, I., Cameron, B. G. and Crawley, E. F. “A technical comparison of three low earth orbit satellite constellation systems to provide global broadband”, Acta Astronautica, 159:123, 2019.
• Bryce Tech, State of the Satellite Industry Report, annual.