Source review: Shotwell 2017 refurb statement

Summary

Claims reviewed

Claim 1 — Refurb cost "substantially less than half" of new-build

Quote: "refurbishing a recycled Falcon 9 booster cost 'substantially less than half the build' of a new rocket" Verdict: Consistent (with caveat)

This is the load-bearing public anchor for q1 calc's refurb-rate assumptions. "Less than half" is consistent with my early-era 30% assumption. It does not, however, tighten the bound — "<50%" doesn't tell us whether the true 2017 figure was 40%, 30%, 20%, or 10%.

Caveat: the statement is from April 2017, one week after the first booster reuse. It reflects the first-of-kind refurb cost when SpaceX was still developing the process. By 2026, after 34 reuses on a single booster (B1076), the rate has clearly compressed. SpaceX has not publicly updated this number.

Claim 2 — Implicit: refurb cost compresses over time

Verdict: Merits investigation

The Falcon 9 lifecycle from 2017 (single reuse, "<50%") to 2026 (34 reuses) implies the per-cycle refurb rate compressed substantially. My calc extrapolates to 15% mid-era and 8% late-era. Without an updated SpaceX statement, these are conjectures consistent with the lifecycle pattern but not directly evidenced.

A subsequent search-for or interview with current SpaceX leadership (Shotwell still President as of 2026) could update this anchor materially. Until then, the late-era 8% figure should be treated as a working estimate, not a hard number.

Cross-reference

• Anchors the refurb-rate variable in [pass-02-calc.py] across all three operational scenarios
• The "first booster reuse just happened" context is important — Shotwell was making this claim 9 years before the 34-reuse milestone
• For source-review breadth, additional refurb data points would help: a 2020/2022/2024 SpaceX statement on refurb cost, or third-party analysis from launch-industry trade press

What we'd want next

• An updated public statement from SpaceX (Shotwell or Musk) on current refurb cost % — Codex's source list mentioned a 2020 Musk "rapid reusability" tweet but didn't quantify refurb.
• Bennett-type external TEA on Falcon 9 economics with current data

Shotwell — April 2017 statement on Falcon 9 refurb cost

The claim (Gwynne Shotwell, SpaceX President):

refurbishing a recycled Falcon 9 booster cost "substantially less than half the build" of a new rocket

Date/venue: April 5, 2017, 33rd Space Symposium, Colorado Springs.

Context: Made one week after the first-ever booster reuse in spaceflight history. This was SpaceX's first public quantification of refurb economics.

Reuse cadence target at the time: refly within 24 hours (Musk goal); first recovery-to-relaunch took ~1 year in practice.

Relevance for q1

Anchors the EARLY-era refurbishment rate assumption in pass-02-calc.py. "Substantially less than half" implies <50% of new-build cost. My calc used 30% for early-era, 15% mid, 8% late — Shotwell's 2017 statement is consistent with the 30% early-era anchor (it's less restrictive — could be 30%, could be 40%; matches "<50%").

What's missing: an updated public statement on current refurb rates in the Falcon 9 mature era. The 34-reuse milestone (B1076, March 2026) implies the rate has compressed significantly but SpaceX hasn't publicly quantified it. Treat my 15% mid / 8% late assumptions as conjectures consistent with the lifecycle pattern.