The payback
One hundred seconds, then 47 minutes.
One square meter of thin film at 10 grams per square meter, launched to escape at 4.25 kilometers per second, costs 90.3 kilojoules of kinetic energy, or 181 from the grid at 50 percent launcher efficiency. In orbit at 0.39 AU it intercepts 9,083 watts and delivers 1,817 at a 20-percent chain. The launch energy is repaid in 100 seconds.
Add a generous 500 megajoules per kilogram for a fully in-situ film line — mining, beneficiation, reduction, rolling, metallizing and quality loss — and total energy payback is 47 minutes. Terrestrial silicon photovoltaics repay in one to two years, a ratio of roughly 10,000 to one. Even if every assumption here is wrong by a factor of 10 in the unfavorable direction, the payback is eight hours.
That says nothing about capital payback, which depends on the mass and reliability of the factory, the launcher, the robots and the spares. What it says is that energy is not the constraint on the growth rate of a Mercurian collector industry, and that everywhere else in the solar system it is at least a contributor.
Placement, not launch cost
The Moon is only six times worse at launching.
The same calculation on the Moon gives 208 seconds and a 5.1-hour total — only about six times worse, because both bodies have negligible escape energy against solar flux integrated over any useful lifetime. Launch cost is not what distinguishes Mercury. Placement is.
A square meter of film delivered to 0.39 AU is worth 6.7 square meters of the same film at 1 AU, permanently. The question is not where film is cheapest to make but where it is worth the most once deployed, and those have different answers.
Which raises the obvious objection: make the film on the Moon, where the industry already exists, and ship it inward. Two answers. The velocity cost is against it — lunar surface to a 0.39 AU heliocentric orbit is lunar escape plus about 9.6 kilometers per second of heliocentric change, against 4.25 total from Mercury’s surface. And more decisively, the film flies itself.
The film is its own transfer stage
Lightness 0.153, and 500 m/s a day.
A 10-gram-per-square-meter reflective film is a solar sail. Its lightness number — the ratio of solar radiation pressure to solar gravity on the same object, which is what decides whether a sail can do useful work — is 0.153, giving 6.1 millimeters per second squared of radiation-pressure acceleration at 0.39 AU.
That is about 500 meters per second per day, propellant-free. The film reshapes its own orbit into position after launch and holds station afterward using the same pressure. No other cargo in the architecture has this property, and it is what makes a distributed collector field of thousands of separately launched panels a sane thing to propose.
It also means the collector is deployed rather than assembled. There is no construction phase in orbit, no crew, and no tug fleet; there is a launcher throwing panels and a fleet of panels sailing themselves to their stations.
The launcher is electromagnetic
The planet contributes 3 m/s.
Mercury’s equatorial rotation speed is three meters per second, against Earth’s 465, so the planet gives a launcher nothing. And with the oxidizer scarce and no methane route, a chemical launcher would be running on polar water that had to be mined and split. The launcher is a mass driver.
| Body | Escape velocity | Track at 30 g | Track at 10 g |
|---|---|---|---|
| Ceres | 0.51 km/s | 0.61 km | 1.8 km |
| The Moon | 2.38 km/s | 10.8 km | 32 km |
| Mercury | 4.25 km/s | 30.7 km | 92 km |
The ranking follows escape velocity exactly. A hardened bulk payload tolerating 1,000 g needs 920 meters, so the long track is a consequence of launching something fragile rather than of launching from Mercury. Thin film is fragile.
This is a scaling problem rather than a physics problem, and the lunar mass driver is the direct precedent — that page owns the technology, and Mercury inherits it at three times the track length. What Mercury adds is a benign construction environment: no atmosphere, no weather, no seismicity worth the name, and no seasons, so a 30-kilometer structure surveyed once stays aligned.
Power for it is undramatic. A program launching 1.87 million tons a year needs about 533 megawatts of beam power, which is 1,600 tons of solar-thermal plant and a quarter of a square kilometer of concentrator. The power plant that drives the launcher is a quarter square kilometer of mirror. The launcher is the hard part.
What a terawatt costs
In film, in tons, and as a hole in the ground.
At 0.39 AU with a 20-percent chain, one terawatt delivered needs 550 square kilometers of collector at 10 grams per square meter: 5,505 tons of film, from about 27,500 tons of processed regolith, which is a hole 24 meters on a side. Every quantity scales linearly, so the ladder reads at any rung.
| Delivered power | Collector area | Film mass | Regolith processed | Excavation |
|---|---|---|---|---|
| 1 TW | 550 km² | 5,505 t | 27,500 t | 24 m cube |
| 20 TW (all current human primary energy) | 11,010 km² | 110,100 t | 550,000 t | 65 m cube |
| 1,000 TW | 550,500 km² | 5.5 Mt | 27.5 Mt | 239 m cube |
An inner-system collector matching everything the human species currently consumes is an excavation 65 meters on a side. Against Terafab’s stated ambition of 10 million tons a year to orbit, the film for it is about four days of one facility’s output — a figure that should be read with all the skepticism corporate roadmaps deserve, and that survives being wrong by two orders of magnitude.
The launch energy is repaid in 50 seconds at every rung, because both terms scale together. That scale invariance is what makes the industry autocatalytic rather than merely large.
This site does not use a logarithmic civilizational classification scale, and the reason is arithmetic rather than taste. Current human primary energy is about 2 × 10¹³ watts. The most aggressive credible figure for a 2075 network is around 10¹⁵, which would already be the steepest sustained energy growth in human history. The first threshold on that classification scale sits at 10¹⁶ — 500 times current consumption and 10 times an optimistic 2075 network. Describing that gap as a step to be crossed makes a 500-fold difference sound like a milestone. The honest unit is watts.
The real constraint
Throughput, and institutional persistence.
So the constraint is not energy, materials, launch, or planetary mass. It is how fast film can be manufactured and how long an institution stays pointed at one thing.
A collector at world-primary-energy scale is not a physics achievement. It is a 30-year procurement that nobody cancels, which is a much harder thing and a much less glamorous one. What makes that plausible at Mercury rather than anywhere else is that the constituency is a defense constituency, and defense constituencies are the only ones that have historically held a capital program together for 30 years.
One caution against over-reading. Nothing here says the collector must be at Mercury; it could be assembled from lunar or asteroidal film and flown inward. The claim is narrower and firmer: the collector must be inside about 0.4 AU or it is 6.7 times larger, and Mercury is the only body of any size inside that radius.