The radiator tax
In Vacuum, the Only Way Out for Heat Is Radiation
No conduction, no convection, and an area that scales with the fourth power of temperature.
Every installation in this book that generates power also has to get rid of the waste heat, and on an airless body there is exactly one mechanism available. A 300 K surface radiating to space emits about 459 W/m², so rejecting a megawatt of waste heat takes on the order of 2,000 square meters of panel. That area is structure, mass, deployment mechanism, and a micrometeoroid target. It sizes reactors, it sizes orbital compute, and it is one of the load-bearing constraints on the beamed-power receiver architecture.
Titan has an atmosphere at 94 K, 5.374 kg/m³, and it moves. Observed
The arithmetic
A 200-Kelvin Temperature Difference Into Dense Cold Gas
Convective flux is the heat transfer coefficient times the temperature difference, and both terms are generous here.
| Route | Flux | Basis |
|---|---|---|
| Vacuum radiator, 300 K to space | 459 W/m² | Stefan-Boltzmann at unit emissivity |
| Titan convective, 300 K into 94 K | ~4,000 W/m² | Coefficient ~20 W/m²·K across ΔT = 200 K |
| Area for the same duty | ~1 : 8.7 | Roughly an order of magnitude smaller |
The temperature difference is the part that is not in dispute. Nothing on Earth rejects heat into a 94 K sink, and the difference term alone is more than twice what a terrestrial plant works with.
The coefficient nobody has measured
20 W per Square Meter-Kelvin Is Plausible and Unverified
A convective heat transfer coefficient depends on wind speed, surface geometry and the local flow regime, none of which has been characterized on Titan. The value used here is a plausible figure for forced convection in a dense gas, not a measurement. If Titan’s near-surface atmosphere is more stagnant than assumed, the coefficient falls, the reactor gets its radiator back, and the first-article manifest grows. This derivation was made for this analysis and has not been through an independent derivation pass.
Dragonfly carries a meteorology package and will fly in that boundary layer from 2034, which makes near-surface wind at the dune fields one of the more industrially consequential things it will report.
What it changes
If It Holds, the Atmosphere Is Plant Equipment
The same pattern as arrival and mobility — the air is not something the architecture works around, it is something the architecture uses.
The 100 kWe fission plant in the Titan first article is listed at 12 to 15 tons, lighter than the Jovian equivalent, and the radiator is the reason. The same saving applies to the electrolysis plant, the hydrocarbon processing train and any compute sited on the surface. Titan is cold, dense and windy, and all three of those work in the plant’s favor.
Aerocapture takes the arrival cost, rotorcraft take the transport network, the atmospheric column takes the shielding, and convection takes the radiator. Four line items that every other destination pays for, absorbed by the same atmosphere. Modeled
Record
Provenance and Data Sources
- Confidence
- Modeled — the temperature difference is measured, the coefficient is assumed
- Derivation
- Vacuum: σT&sup4; at 300 K = 459 W/m². Titan: hΔT at h = 20 W/m²·K, ΔT = 200 K
- Status
- Derived in this analysis; never independently reviewed. Flagged as such in the source analysis
- Weakest link
- Near-surface wind and flow regime at Titan, unmeasured until Dragonfly
- Reviewed
- Q3 2026