The Drop In Space Development Costs
If Starship is compared to previously proposed and built reusable transport vehicles, the difference in launch costs are impressive. However, the decrease in development costs are astounding.
The systems for comparison were the Space Shuttle [1], the Boeing Heavy Lift Launch Vehicle (HLLV) proposed in 1980, and the Rockwell HLLV proposed in 1980. The sources of data were the Payload Starship Report [1], the GAO 1993 Shuttle Cost Estimates [2], and the 1981 NRC report on Space Based Solar Power (SBSP) [3]. This last report included budget estimates for a SBSP satellite system, launch systems, and other support costs. The rocket system specs and budgets were extracted. The budgets in all reports were normalized to 2024 dollars. More details on the sources data and normalization is in the comments.
The estimated Starship R&D development budget is half of the Shuttle development budget, while it will carry 4 to 6 times more cargo to LEO. Additionally, Shuttle development was low by contemporary expectations because it reused a lot of hardware developed during the Apollo program. A clean sheet design would have been much more expensive. Comparing Starship to the HLLV budgets from 1980, SpaceX is developing a similar capability vehicle for less than 10% of the projected budget for those systems.
A 90% reduction in development costs reflects fundamental changes in how space vehicles are developed.
Many things have changed since 1980. Simulation is able to accurately predict stresses in structures and propulsion systems. SpaceX has more than a decade of operations experience with Falcon. Avionics and controls are much more advanced. Telemetry provides firehoses of data in early testing and operations. Additive manufacturing produces complex shapes faster and more capable. It also eliminates failures and complexity from fasteners and fittings that were required when using traditional manufacturing. And critically, there have been massive improvements in engines.
This is another data point highlighting that an inflection in the space economy is happening. The technologies that accumulated over the past 40 years are now combining to provide massive productivity improvements.
These same trends are available to move old ideas off the drawing board into production. Space based solar power (SBSP), which has been technically possible since the 1980s, is one of those technologies. Since SBSP is fundamentally an electronics design problem, there are even more trends to help it become competitive.
References
[1] Payload Research Starship Report - https://payloadspace.com/starship-report/
[2] 1993 GAO Space Shuttle Cost Estimates - https://www.gao.gov/assets/nsiad-93-115.pdf
[3] Committee of Satellite Power Systems, Electric Power From Orbit: A Critique of a Satellite Power System, Volume 2 - https://nss.org/wp-content/uploads/2017/07/1981-NRC-Electric-Power-From-Orbit-2-Appendices.pdf
Other Reading
- Committee of Satellite Power Systems, Electric Power From Orbit: A Critique of a Satellite Power System, Volume 1 - https://nss.org/wp-content/uploads/2017/07/1981-NRC-Electric-Power-From-Orbit-1-Report.pdf
Methodology notes:
1) All costs were normalized to 2024 dollars.
2) Published shuttle cargo costs cover a large range, including one report that placed cargo at $250,000/kg. For the purposes of this analysis, the costs were calculated by amortizing the shuttle unit costs over the design life and adding the launch costs, then dividing by the cargo capacity.
3) 1980 HLLV numbers were based on the final adjusted costs made by The Aerospace Corporation in "Electric Power From Orbit: A Critique of a Satellite Power System". Those number were usually 20-30% more than the original contractor estimates
