Public figure: George Sowers

Quotes reviewed

Quote 1

Statement: "The cost of a luxury hotel in Las Vegas" — framing the $4 billion initial investment Source: 21st Century Technology interview 2021 → sources/figure-sowers/extract.md (4b-luxury-hotel) Verdict: Different conclusion — different scope, not directly comparable Why: Sowers' $4B is for the Commercial Lunar Propellant Architecture — a fully-robotic tent-sublimation ISRU plant producing ~2,450 t/yr propellant. Architecture is fundamentally different from our q5 net-positive-export manufacturing base (which includes habitat for crew, manufacturing complement, electronics fab, mobility). Our BAU $150-400B is not directly comparable. Per claim q5.c13 (revised after Codex critique): φ-rescaling from 20 to 534 does not give a defensible numeric bridge — the two architectures cannot be reconciled by simple scaling. Sowers' figure is the published lower bound for "propellant-only commercial architecture"; ours is for "net-positive-export full base." Severity: high (frequently cited as "lunar base cost" without scope-disclaimer)

Quote 2

Statement: "$2.5 billion for development, fabrication and delivery to the lunar surface" (capture tent ISRU system) Source: Sowers 2021 → sources/figure-sowers/extract.md (25b-capture-tent) Verdict: Consistent within Sowers' own framework Why: $2.5B for the ISRU plant alone, with the remaining $1.5B for the surrounding architecture, is internally consistent with Sowers' tent-sublimation φ ≈ 534 case. Useful as a primary-source breakdown of where the $4B is allocated. Severity: medium (per-subsystem allocation is informative for comparing against our calc-baseline ISRU figure)

Quote 3

Statement: "$550M of revenue per year generating a 9% return-on-investment and $2B in profit over the life of the project of 15 years" Source: Sowers 2021 → sources/figure-sowers/extract.md (550m-revenue-9-pct-roi) Verdict: Out of scope for q5 (demand/revenue side, not capex) Why: Revenue and ROI are q6 demand-side anchors; q5 is buildup capex. Useful only as a sibling check that Sowers' architecture has a published business case (not just a capex figure). Severity: low

Quote 4

Statement: "On the Moon surface itself, the cost of propellant would be $500 per kilogram" vs "$36,000 to the lunar surface" from Earth-launched Source: Sowers 2021 → sources/figure-sowers/extract.md (500-per-kg-moon; 36k-per-kg-from-earth) Verdict: Consistent with q3 ISRU TRL and q1 transport-cost framing Why: 72× cost compression for lunar-made vs Earth-launched propellant is consistent with the IE regime's mass-compression mechanism. Direct support for the q5 IE regime's lunar-substitution direction (though for propellant specifically, not manufacturing-complement). Severity: low

Quote 5

Statement: "Lunar operations designed to use robots exclusively (no human presence required)" Source: Sowers 2021 → sources/figure-sowers/extract.md (robotic-only) Verdict: Different conclusion — q5 includes M6 crewed milestone Why: Sowers' architecture deliberately avoids crewed operations to suppress capex. Our q5 calc includes M6 (12-month sustained crewed occupation) as a load-bearing milestone for the net-positive-export endpoint. This is a scope difference: Sowers chooses propellant-only architecture to avoid the crewed-base capex; our q5 includes that capex as part of the manufacturing-base definition. The robotic-only path is a credible alternative that materially reduces capex by removing M3/M6 mass budget. Severity: medium (architectural choice; could be a legitimate alternative q5 framing — robotic-only manufacturing base — that we have not separately costed)

Quote 6

Statement: "A 10-year timeline for cislunar infrastructure realization" Source: Sowers → sources/figure-sowers/extract.md (10-year-timeline) Verdict: Consistent with our IE regime cadence; between BAU and IE Why: 10-year buildup is between our BAU (20-25 yr) and IE (5 yr) cadence brackets. Sowers' figure assumes commercial-launch (not SLS-class) transport, which is the assumption that compresses the timeline relative to BAU. Severity: low

Quote 7

Statement: "In some areas of the lunar surface water represents 30% of the material by weight" Source: Sowers → sources/figure-sowers/extract.md (water-30pct) Verdict: Consistent with PSR water-content evidence Why: 30% water by mass in PSR regions is the optimistic end of the published range. Consistent with q3 ISRU mass-throughput assumptions. Severity: low

Cross-reference

• Sowers is the highest-credibility tier-B figure on commercial lunar-ice-mining economics. His $4B figure is the lowest published-credibility capex anchor for ANY lunar-export architecture, and his New Space 2021 paper is the underlying tier-S primary source (sowers-2021-ice-mining).
• The $4B anchor is for propellant-only architecture; our BAU $150-400B is for net-positive-export manufacturing base. These are not in conflict — they are different scopes.
• Sowers' robotic-only choice is a credible alternative q5 framing that we have not separately costed.
• Cross-references jones-2020-breakeven (Tier S; 35-yr SLS-class pessimistic-mirror analysis) and sowers-2018-clpa (Tier S; the underlying CLPA architecture study).
• Codex anti-hallucination check: all quotes trace verbatim to figure-sowers/extract.md anchors.

Abstract

George Sowers, former ULA Chief Scientist and now at the Colorado School of Mines Space Resources program, is the canonical credentialed public figure arguing the commercial case for lunar ice mining. His central claim: $4 billion initial investment — "the cost of a luxury hotel in Las Vegas" — is sufficient to complete a lunar propellant production facility that yields a 9% ROI, $550M annual revenue, and $2B profit over 15 years. The architecture is fully robotic (no humans required at the mining site). Propellant cost from Moon surface: $500/kg, versus the $36,000/kg current cost to deliver Earth-launched propellant to the lunar surface. He publishes extensively and presents the Sowers 2021 business case and the underlying Commercial Lunar Propellant Architecture (CLPA, 2018).

Key statements

• 4b-luxury-hotel: "The cost of a luxury hotel in Las Vegas" — Sowers' framing of the $4 billion initial investment needed
• 25b-capture-tent: "$2.5 billion for development, fabrication and delivery to the lunar surface" (capture tent ISRU system)
• 550m-revenue-9-pct-roi: "$550M of revenue per year generating a 9% return-on-investment and $2B in profit over the life of the project of 15 years"
• 500-per-kg-moon: "On the Moon surface itself, the cost of propellant would be $500 per kilogram"
• 36k-per-kg-from-earth: "Today a kilogram of propellant costs $4,000 to low-Earth orbit, $16,000 to geosynchronous-Earth orbit where most of our communication satellites reside, and $36,000 to the lunar surface"
• robotic-only: "Lunar operations designed to use robots exclusively (no human presence required)"
• 10-year-timeline: "A 10-year timeline for cislunar infrastructure realization"
• water-30pct: "In some areas of the lunar surface water represents 30% of the material by weight"

Reviewer notes

Tier-B figure with strong primary-source backing (Sowers 2021, CLPA 2018, both Tier S/A separately indexed). Load-bearing for q5: Sowers' $4B figure is the lowest published-credibility capex anchor for net-positive-export lunar manufacturing. It is structurally consistent with the tent-sublimation φ ≈ 534 branch of Metzger 2023. Cross-references jones-2020-breakeven (which uses SLS-class transport and arrives at 35-yr / 7-Mars-mission breakeven — the pessimistic mirror image). Notably absent in Sowers' framing: explicit treatment of finance cost ξ; explicit replacement schedule for the tent / processing hardware. To be reviewed in sub-pass 4.