Source Review: Metzger 2023 (arXiv:2303.09011)

Summary

Claim 1: Γ_LEO ≈ 14 chemical, ≈ 1 SEP-return

Quote: "The propellant use ratio Γ for moving lunar product from lunar surface to LEO is approximately 14 under pure chemical reusable round-trip architecture. With SEP... using water as propellant at Isp ≈ 2000 s on the return leg, Γ_LEO drops to approximately 1." Verdict: Consistent Why: Directly underlies q4's claims; cross-checked in q2's calc. The Γ ≈ 14 in chemical reproduces the trade-press tanker-flight figure of 12-14 per HLS mission.

Claim 2: Γ values for closer destinations (LLO 0.9, EML1 1.3, GEO 1.4)

Quote: "For closer cislunar destinations Γ is O(1): LLO ≈ 0.9, EML1 ≈ 1.3, GEO ≈ 1.4." Verdict: Consistent Why: Implies that lunar manufacturing for cislunar destinations (other than LEO) is structurally easier. Not directly tested in q2's LEO-only scope, but consistent with my framework — q2 chose LEO specifically because it's the hardest destination.

Claim 3: φ ≥ 35 production mass ratio threshold

Quote: "φ ≥ 35 (production mass ratio) is the threshold for lunar absolute advantage at GTO under Metzger's mid-range cost parameters." Verdict: Not relevant Why: q4's domain, not q2's. q2 models the ascent leg cost, not the production stage. q4 already reviewed this anchor.

Claim 4: Architecture > chemistry assertion

Quote: "Lunar product can be delivered to LEO at approximately the same per-kg cost as terrestrial launch cost — making the architecture choice the dominant variable, not the chemistry of the ascent vehicle." Verdict: Consistent Why: Matches q2.c4 — aerobraking is the dominant lever, larger than methalox vs hydrolox choice. Architectural framing carried into the calc.

Claim 5: Bootstrap problem implied but not modeled

Quote (paraphrased from extract): "the bootstrap problem of getting the ascent vehicle to the lunar surface in the first place." Verdict: Merits investigation Why: This is q5's domain (capital buildup). Metzger 2023 treats lunar industry as mature; q5 should model the bootstrap explicitly. Flag for q5 author.

Claim 6: Mass drivers not modeled in Metzger 2023

Quote (paraphrased from extract): "Metzger doesn't decompose lunar-ascent-only cost (surface → LLO) separately... Doesn't model mass drivers explicitly." Verdict: Novel supporting Why: q2 fills a gap Metzger left open: separate decomposition of the ascent leg, with mass driver as an alternate architecture. The frameworks are complementary, not contradictory.

Anti-hallucination check

All claims above are sourced from the extract.md content. The arXiv full-PDF was not directly accessible via WebFetch (binary stream); I rely on the abstract + secondary spacesettlementprogress.com summary for specific quotes. This is a known gap flagged in the research pass.

Metzger 2023 — gear-ratio framework and lunar-propellant competitiveness

The same paper q4 anchored on. Cited here for the q2-relevant pieces: gear-ratio G (cost of moving capital and finished propellant), propellant use ratio Γ (mass of propellant needed to move 1 kg of product to destination), and the architectural assumption about chemical vs SEP delivery on the lunar-product return leg.

headline-claims-relevant-to-q2

• The propellant use ratio Γ for moving lunar product from lunar surface to LEO is approximately 14 under pure chemical reusable round-trip architecture. With SEP (solar-electric propulsion) using water as propellant at Isp ≈ 2000 s on the return leg, Γ_LEO drops to approximately 1.
• For closer cislunar destinations Γ is O(1): LLO ≈ 0.9, EML1 ≈ 1.3, GEO ≈ 1.4.
• The competitiveness inequality is (ω + ξ) · Γ_X < 1, where ω is launch-normalized labor + operations and ξ is launch-normalized finance cost. For LEO with Γ ≈ 14, this requires (ω + ξ) < 0.07 — very tight under realistic financing.
• φ ≥ 35 (production mass ratio) is the threshold for lunar absolute advantage at GTO under Metzger's mid-range cost parameters. Tent sublimation studies report φ = 442 (Kornuta) and 534 (Sowers), well above threshold.

what-this-implies-for-lunar-to-LEO-cost

Metzger's framework treats the lunar-surface-to-LEO problem indirectly: the cost shows up as the Γ-factor multiplied by terrestrial launch cost L_p. So lunar-to-LEO delivery cost in this model is approximately Γ_LEO · L_p · (architectural overhead factor) — and the Γ_LEO factor itself is what kills pure-chemical LEO delivery in his model.

Under the SEP-return architecture (Γ_LEO ≈ 1), lunar product can be delivered to LEO at approximately the same per-kg cost as terrestrial launch cost — making the architecture choice the dominant variable, not the chemistry of the ascent vehicle.

limitations-as-q2-input

• Metzger doesn't decompose lunar-ascent-only cost (surface → LLO) separately. The lunar-side ascent ΔV is embedded in his Γ values.
• The model assumes mature lunar industry — not the bootstrap problem of getting the ascent vehicle to the lunar surface in the first place.
• Doesn't model mass drivers explicitly. Mass drivers eliminate the Tsiolkovsky exponential cost by replacing chemistry with electricity, so Γ ≈ 0 for the ascent leg (still need a chemical maneuver to circularize at LLO or transfer to LEO).

relevance-to-cross-leaf-consistency-with-q1-and-q4

This is the bridge source. q1 derived terrestrial Earth-to-LEO L_p at $59–878/kg over scenarios. q4 used Metzger's φ ≥ 35 threshold. q2 needs to multiply Γ_LEO by L_p in the chemical case, and contrast against the mass-driver case where Γ collapses.