The sorting rule

The rule

A radiator does not know where it is.

The flux advantage passes through at full strength to processes whose product is energy or whose bottleneck is collector mass, and does not pass through at all to processes whose bottleneck is heat rejection.

The reason is that a radiator sheds heat at the Stefan–Boltzmann rate, which depends on the radiator’s own temperature and nothing else: 459 watts per square meter at 300 kelvin, 1,452 at 400, 3,544 at 500. Those numbers are identical at Mercury and at Neptune.

The sink term is where intuition goes wrong. A 40-kelvin lunar crater floor returns 0.15 watts per square meter and a 100-kelvin Mercurian one returns 5.7, and against a 300-kelvin radiator’s 459 both are noise. The difference between those two floors is worth about one percent of radiator area for any warm system. It matters enormously for cryogenic systems and not at all for anything else.

So a thermal system’s mass is set by its radiator, its radiator by its waste heat, and its waste heat by how much power it processes — not by where it sits. Three consequences follow, and each one changes an architecture.

Photovoltaics do not work here

A 30-percent panel runs at 487 kelvin.

A flat panel converting 30 percent of incident light radiates the rest from both faces. At 1 AU that balance settles at 303 kelvin, which is why terrestrial-heritage panels work in Earth orbit. At Mercury’s semi-major axis the same panel runs at 487 kelvin, and at perihelion 546.

Triple-junction cells are typically qualified to around 150 Celsius and degrade sharply above it, and the feedback runs the wrong way: efficiency falls with temperature, which raises the waste fraction, which raises the temperature. MESSENGER solved this by tilting its arrays away from the Sun and running at a third of normal-incidence flux, which is a fine spacecraft solution and a hopeless industrial one, because it discards the entire reason for being at Mercury.

Mercury’s power architecture is solar-thermal: a concentrator, a high-temperature receiver, a Brayton or Stirling cycle at 40 to 50 percent, and reject heat leaving at 500 to 700 kelvin, where a radiator is 4 to 15 times more compact per watt than at 400. High input flux is an asset to that architecture rather than a liability, which is the opposite of the photovoltaic case and is why the polar ring is built out of heliostat fields rather than solar panels.

Concentration cannot beat brightness

The same peak flux at both locations.

A natural follow-on is that Mercury must be the best place in the solar system to do direct solar metallurgy, because the concentrated flux is higher. It is not higher.

The maximum concentration ratio achievable with any optic is set by the angular size of the source, and the Sun is 6.7 times larger in solid angle at Mercury: the achievable concentration falls by exactly the factor the flux rises. Peak achievable flux is 6.3 × 10⁷ watts per square meter at both locations, and the stagnation limit at both is the photosphere’s 5,800 kelvin. What is conserved is brightness, and no passive optic anywhere exceeds it.

Direct solar metallurgy at Mercury survives this correction intact, but its justification changes. Mercury buys collector mass, not temperature. The furnace you can build there is the same furnace you could build in Earth orbit; it just costs a sixth of the mirror to feed it.

Compute is worth about two

And the site has a hostile sky.

A one-megawatt compute installation is a power plant plus a radiator. The plant shrinks by the full flux ratio; the radiator does not move at all. The system runs 283 watts of compute per kilogram at Mercury against 143 at 1 AU — a factor of two, not 6.7 — and the advantage shrinks further as designs mature, because the plant term falls toward zero while the radiator term does not.

There is a penalty on the other side. A radiator at 0.39 AU has a hostile sky: it must be edge-on or shaded, and any pointing error admits 9,000 watts per square meter onto a surface designed to shed 1,452. Mercury makes radiator pointing safety-critical in a way it is nowhere else. Add the 100-kelvin floors, too warm for the superconducting-logic case, and Earth’s market three seconds from cislunar rather than eight minutes from here.

What the 6.7× buys

Activity by activity.

The rule in one line: the flux advantage survives conversion into light and dies in conversion into heat. Mercury’s economy is the set of activities whose product is photons or whose input is direct heat, and it is a poor location for everything else. That is not a limitation on the argument. It is the argument, because the two things Mercury is uniquely good at are the two things the rest of the solar system most needs and cannot make locally.

The result that outlives Mercury

A bound that holds everywhere.

One result generalizes past this destination and is worth more than Mercury is.

Any system that rejects waste heat at temperature T has a specific power bounded by its radiator, and that bound is independent of where in the solar system it sits. Location buys collector area. It never buys radiator area.

It is the same constraint the outer-system pages hit from the other direction, where the binding number past Jupiter is watts per kilogram and the reactor that supplies them is radiator-bound too. The sorting rule and the specific-power wall are one constraint seen from two distances, which is why a beam pointed at Callisto attacks both at once.