CALC-137

The 1979 launch cost estimate of USD 27.0 per kg to low Earth orbit, in 2024 dollars, beside the 2025 average cost per kg

REGISTERED 2026-10-14 (ON-482). Drafted for the lessons records set (commit f8cfb1ce), checked by an independent reader (FIX, all applied, round 2 PASS), landed by SBSP-coord. verified: VER-2353 CONFIRMED 2026-10-14.

Basis: USD 2024 per kg, moved with US CPI-U annual values. The 1979 figure is a cost per flight of one planned vehicle to low Earth orbit (LEO). The 2025 figure is an average cost over all launches, expressed per kg of LEO-equivalent payload.

Why it is needed. The two figures are in different price years, 1977 dollars and 2024 dollars. They can be set side by side only in one price year.

Inputs.

What each figure is. - USD 27.0 per kg (1977 dollars) is a planning estimate for one vehicle that the programme planned. It comes from the November 1979 cost data book of NASA's Johnson Space Center, which was the programme's only stated transport cost basis. The estimate is parametric, from a proprietary cost model (CLM-1024). The 1978 reference system describes the Earth-to-LEO cargo vehicle as "Vertical take-off, winged 2-stage", with a dedicated electric orbit transfer vehicle for the trip on to geostationary orbit (GEO) (SRC-085 Figure 7, PDF p.28, printed p.16). The cost per flight spreads the vehicle's hardware over 300 flights and adds propellant, refurbishment, ground operations, tooling and staff. It assumes 100 per cent reliability and no reserve vehicles, and it leaves out launch facilities (CLM-1024). It is the cost for a fleet flying 11,606 flights to build 60 satellites (CLM-1028). Destination: LEO. - USD 3,868 per kg (2024 dollars) is an average over all launches in 2025. It is a cost, not a price: the paper warns that a near-monopoly supplier has an incentive to charge more than cost. It averages all vehicles and is not the cost of any one vehicle. Payload to other orbits is converted to a LEO-equivalent mass with engineering coefficients given in the dataset notes, not in the paper, and some costs are imputed (CLM-1389). The paper gives no cost to GEO. Destination: launches to all orbits, expressed as LEO-equivalent mass.

The two figures therefore differ in kind. One is a forecast cost for a single planned vehicle at a high, steady flight rate. The other is an estimated average over every vehicle that flew in one year, built partly from imputed costs.

Working (script output, transcribed).

factor 1977->2024 = 313.689 / 60.608 = 5.1757 27.0 x 5.1757 = 139.74 USD2024/kg ratio 3,868 / 139.74 = 27.68 net 374 t: 27.005 USD1977/kg x 5.1757 = 139.77 USD2024/kg; 3,868 / that = 27.7 gross 424 t: 23.821 USD1977/kg x 5.1757 = 123.29 USD2024/kg; 3,868 / that = 31.4 360 t payload: 28.056 USD1977/kg -> 145.21 USD2024/kg; 3,868 / that = 26.6 53.9 USD1977/kg Earth to GEO x 5.1757 = 278.97 USD2024/kg; 3,868 / that = 13.9

Result. USD 27.0 per kg to LEO in 1977 dollars is about USD 140 per kg in 2024 dollars (139.7). The 2025 average cost over all launches, USD 3,868 per kg to LEO-equivalent in 2024 dollars, is about 28 times that (27.7; 26.6 to 31.4 across the three payload figures). The 1979 plan's average transport cost per kg of satellite from Earth to GEO, about USD 279 in 2024 dollars, is still about one fourteenth of the 2025 average to LEO-equivalent alone. It leaves out development, launch facilities and some support costs (CLM-1028).

What the comparison can show. After US consumer price inflation, the 1979 plan assumed a launch cost per kg far below the average cost per kg that launches actually achieved in 2025.

What it cannot show. - That the 1979 figure could have been reached. It is a model estimate for a reference design, at 100 per cent reliability. - How today's cheapest vehicle compares. An average over all vehicles includes costly ones, so one vehicle's cost can sit well below it. - What a buyer would pay. Both figures are costs, not prices.

Where this fails. - Deflator. CPI-U is a consumer price index, not a launch or aerospace cost index. No US producer price index or GDP deflator is held, so the effect of the index choice over 47 years was not measured. The data book's own 1977 to 1979 factor is 1.15 (SRC-532 PDF p.111). CPI-U gives 1.197 for the same two years (72.575 / 60.608). The page does not say which index or forecast gives its 1.15. - The 1977 index value is our 12-month mean, not a BLS annual average. CALC-124 uses the same value. - Payload. The record takes the 374 t net payload as delivered. The 424 t gross payload gives 31.4 times, and the 360 t that Battelle uses elsewhere (CLM-1024) gives 26.6 times. - The 1979 model's cost relationships were not disclosed to the reviewer (CLM-1024). The figure cannot be checked below the level of the data book's line items. - The 2025 figure rests partly on imputed costs and on LEO-equivalent coefficients that are outside the paper (CLM-1389). - Rounding. The record's 27.0 is 27.005 before rounding; this changes the result by 0.02 per cent.

Cite this