What Is the True Cost Floor for Payload to LEO?

A First-Principles Analysis Using Fuel Economics and the Aviation Markup Analogy

1. Establishing the Transport Cost Hierarchy#

Before analyzing rockets, it's useful to understand what mature transport industries charge per kilogram across thousands of miles:

The assumption that LEO launch will always cost more than air freight per kg seems safe — the energy requirement is fundamentally different. But how much more?

2. Raw Fuel Cost: The Theoretical Minimum#

Falcon 9 Propellant Economics#

A Falcon 9 Full Thrust carries approximately:

• RP-1 (kerosene): ~186,000 kg at ~$2.30/kg = $428,000
• Liquid oxygen (LOX): ~312,000 kg at ~$0.27/kg = $84,000
• Total propellant cost: ~$512,000

Payload to LEO (expendable): ~22,800 kg

Fuel cost per kg of payload = $512,000 ÷ 22,800 ≈ $22.50/kg

This is the absolute thermodynamic floor — the cost of the chemical energy alone, ignoring the rocket, the people, the pad, insurance, and everything else. It represents just 0.8% of the current $2,700/kg price, confirming that fuel is essentially irrelevant to current launch costs.

Key Insight: The Mass Ratio Problem#

The reason fuel cost per kg of payload is so much higher for rockets than aircraft is the tyranny of the rocket equation:

• A 747-400F burns ~70,000 kg of fuel to carry ~113,000 kg of payload — a fuel-to-payload ratio of 0.62:1
• A Falcon 9 burns ~498,000 kg of propellant to deliver ~22,800 kg — a ratio of 21.8:1

You need 35× more fuel per kg of payload for orbit than for a transatlantic flight. This is physics, not engineering — achieving orbital velocity (~9.4 km/s with gravity and drag losses) simply requires exponentially more energy than cruising at Mach 0.85.

3. The Aviation Markup: From Fuel Cost to Freight Price#

The most instructive comparison is between the fuel cost of moving cargo by air and the actual price charged.

NYC → London by 747-400F#

• Distance: ~3,450 miles (~3,000 nautical miles)
• Fuel burned: ~70,000 kg of Jet-A
• Jet-A cost (airline wholesale): ~$2.34/gallon (BTS, July 2025)
• Fuel density: ~3.04 kg/gallon
• Fuel cost per kg: $2.34 ÷ 3.04 = ~$0.77/kg of fuel
• Total fuel cost: 70,000 kg ÷ 3.04 = ~23,000 gallons × $2.34 = ~$53,800
• Maximum payload: ~113,000 kg
• Fuel cost per kg of payload: $53,800 ÷ 113,000 = ~$0.48/kg

Actual Air Freight Price: NYC → London#

General cargo air freight on high-volume transatlantic lanes runs approximately $3.50–5.00/kg (2025 rates, all-in with fuel surcharges).

Using $4.00/kg as a representative midpoint:

Markup = $4.00 ÷ $0.48 ≈ 8.3×

This 8.3× markup covers: aircraft purchase/lease amortization, crew salaries, maintenance and inspections, ground handling, airport fees, navigation charges, insurance, regulatory compliance, corporate overhead, and profit margin. Aviation is a mature, hyper-competitive industry with thin margins — this ratio represents what "efficient" looks like after 70+ years of optimization.

4. Applying the Aviation Markup to Rockets#

If we naively apply the aviation 8.3× markup to rocket fuel costs:

$22.50/kg (fuel) × 8.3 = ~$187/kg

This would imply a long-run theoretical "mature industry" price of roughly $150–200/kg to LEO.

Why This Is Almost Certainly Too Optimistic#

The aviation markup is generous when applied to rocketry for several structural reasons:

1. Extreme operating environment: Rocket engines operate at combustion temperatures of ~3,670K and chamber pressures of ~100 atm. Jet engines operate at ~1,700K and ~40 atm. This means far faster wear, more frequent inspection, and shorter component lifetimes. Maintenance costs scale non-linearly with operating severity.

2. Turnaround time: A 747 flies 2–4 sectors per day, accumulating 4,000+ flight hours per year. A Falcon 9 booster currently flies ~10–15 times per year (improving to perhaps 20–30). Each rocket "flight hour" must amortize a much larger share of the vehicle cost.

3. Utilization fraction: A 747 spends ~90% of its operational day either flying or preparing to fly. A rocket booster spends >95% of its time being inspected, transported, or waiting. This idle capital is expensive.

4. Insurance and risk: Aviation hull-loss rates are ~0.2 per million flights. Rocket failure rates are ~1–3 per 100 flights (even for mature vehicles). Insurance alone adds significant cost per flight that has no aviation equivalent.

5. Range infrastructure: Launch requires specialized pads, range safety, exclusion zones, and regulatory overhead with no aviation parallel.

A more realistic "mature rocketry" markup would be 15–25×, accounting for these structural differences:

$22.50 × 15 = $338/kg (optimistic) $22.50 × 25 = $563/kg (more realistic)

5. Cross-Check: Starship Projections#

SpaceX's Starship targets ~100–150 tonnes to LEO with full reusability. SpaceX has publicly targeted costs of $200–500/kg near-term, with aspirations toward $50–100/kg at high flight rates.

Using Starship's methane/LOX propellant (CH₄ at ~$8.8/kg, LOX at ~$0.27/kg), with ~4,600 tonnes of propellant per full stack:

• Propellant cost: ~$1.5–2 million per launch
• At 100,000 kg payload: ~$15–20/kg in fuel alone
• At a 15–25× mature markup: $225–500/kg

This aligns closely with SpaceX's own near-term cost targets, suggesting the markup analysis is reasonable. The aspiration of $50/kg would require either:

• A markup of only ~3×, which no transport industry has ever achieved, OR
• A dramatic increase in payload mass fraction, OR
• A fundamental change in vehicle architecture (e.g., space elevator, orbital ring)

6. What This Means for Space Datacenter Economics#

Revised Probability Assessment#

The original analysis assumed a potential 20-year price of $50/kg. This fuel-based analysis suggests that $200–500/kg is a more realistic mature-industry floor under standard progress assumptions, with $100/kg as an aggressive but not impossible stretch goal at very high flight rates.

At $200–500/kg, launching the components for a 100 MW space datacenter (estimated 2–5 million kg total system mass including radiators, solar arrays, compute, and structure) would cost:

• At $200/kg: $400M – $1B in launch costs alone
• At $500/kg: $1B – $2.5B in launch costs alone

These figures are on top of the hardware cost, and the hardware cannot be readily maintained or upgraded. A comparable terrestrial datacenter costs $1–3B total and lasts decades with continuous upgrades.

Updated Probability Table#

The fuel-economics analysis actually lowers the 20-year probability slightly from the original estimate, because it demonstrates that the $50/kg aspiration requires a transport-industry markup ratio (fuel cost to total price) that no industry operating in extreme environments has ever achieved.

7. Summary of Key Numbers#

The bottom line: fuel is cheap, but the physics of orbital velocity creates a mass ratio that makes everything else expensive. The aviation analogy shows that even highly mature, hyper-competitive transport industries charge 8× their fuel cost. Rocketry's harsher operating environment pushes that multiplier higher. A "mature rocketry industry" price of $200–500/kg is a defensible long-run estimate — still far too expensive to make space datacenters economically compelling against continuously improving terrestrial alternatives.