Space, Materials, and Real Abundance

Volume VII — Planetary and Cosmic

Nine movements, one gravity well.

THE PLATE

A watercolour of a young woman with an open book, flowers and leaves around her. — Plate VII.09The Weight of the Bottle.The bottle holds one kilogram. On this bench it is worth a tenth of a penny. Four hundred kilometres above her head, the same kilogram is worth three thousand and fifty-nine dollars — and nothing about the water has changed.

THE LETTER

You have been told, more than once, that space will solve the materials question. A single metallic asteroid, the sentence usually runs, contains more platinum than has ever been mined on Earth, and therefore scarcity is a temporary condition of a species that has not yet gone to fetch what is waiting.

The sentence is true in its first half and it does not follow in its second, and the distance between those two facts is the most expensive misunderstanding in the whole of resource economics. This chapter closes it with arithmetic.

We are going to do something unfashionable with the most romantic subject in this book: we are going to cost it. Not dismiss it — cost it. Every figure in what follows is computed in lib/verify/VII_09.py and printed there with its inputs, so that you can disagree with an assumption rather than with a conclusion. That is the whole purpose of putting the numbers where you can reach them.

What you will find is that the case for space resources is far stronger than the enthusiasts argue and points somewhere completely different from where they are pointing. There is a genuine three-billion-to-one asymmetry out there and it is real and it is exploitable. It simply does not run in the direction of Earth.

Chapter II.01 established the finding this chapter extends: scarcity is positional, not absolute. Almost no shortage is a claim about how much of something exists; nearly all of them are claims about where it is, what form it is in, and whose it is. Space is the largest possible test of that finding, because space is the one place where the quantity really is effectively unlimited and the position really is the entire cost. If the positional theory of scarcity is right anywhere, it is right here — and if it is wrong here, we should want to know.

It is right here. It is right here in a way that will surprise you, and the surprise is in the fourth movement.

— The Editors

DISCOVERY

What is already working

Begin where the evidence is strongest, which is not with a proposal but with hardware that has flown.

Two samples have come back. On 6 December 2020 the Japanese spacecraft Hayabusa2 delivered 5.4 grams of the asteroid Ryugu to a paddock in South Australia. On 24 September 2023 the American spacecraft OSIRIS-REx delivered 121.6 grams of the asteroid Bennu to the Utah desert, comfortably over the sixty grams the mission had been required to bring. Both worked. Both returned material that had not been altered by an atmosphere or a terrestrial biosphere, and both are now producing chemistry nobody could have done any other way — hydrated clays, carbon at a few percent by mass, amino acids, the ordinary constituents of a wet early solar system.

Those two missions are also, and this is the more useful fact, the only prices ever actually paid for asteroid material delivered to Earth. OSIRIS-REx cost $1.16 billion and returned 121.6 grams, which is $9.54 million per gram, or $9.54 billion per kilogram. Hayabusa2 cost about $150 million for 5.4 grams, which is $27.78 million per gram. Hold those two figures. Everything that follows is measured against them, and they are the reason this chapter is arithmetic rather than prophecy.

The price of getting up has genuinely collapsed, and by a mechanism worth naming precisely. Harry Jones's survey for the International Conference on Environmental Systems puts the cost of a kilogram to low Earth orbit at about $18,500 in the Shuttle era and about $2,720 by 2018 — a fall of 6.8 times over thirty-seven years, which is a compound decline of 5.05 percent a year and a halving time of 13.4 years. The Space Shuttle, on its own program accounts, costs $56,296 per kilogram. A Falcon 9 at list price costs $3,059 per kilogram.

The mechanism was not a new propellant or a new engine cycle. It was not throwing the vehicle away. A booster that costs $30 million to build and flies twenty-four times, with a million dollars of work between flights, amortises to $2.25 million a flight — a saving of 13.3 times against expending it. The cheapest thing ever discovered in spaceflight turned out to be reuse, which is the same finding as Chapter II.01's circulation multiplier arriving in a different industry.

Closure is already outperforming extraction, in orbit, today. The International Space Station's environmental control system recovers 98 percent of the water that passes through it. Put that into II.01's multiplier, M = 1/(1 − p), and a kilogram of water launched to the station does the work of 50.0 kilograms. Seven people at three kilograms a day would need 7,670 kilograms of water a year open-loop; at 98 percent closure they need 153 kilograms of make-up. At $3,059 a kilogram that single engineering choice avoids $23.0 million of launch a year — more value per year than any asteroid mining proposal has ever put on paper, on a system that is flying now, and nobody calls it mining.

Power has been beamed, in space, and it worked. The Caltech Space Solar Power Demonstrator flew MAPLE in January 2023 and transmitted power wirelessly between elements in orbit and, at very low levels, to a receiver on the ground — the first time either had been done. It sits on a much older result: William Brown's microwave work at Raytheon reached a measured 54 percent DC-to-DC link efficiency in 1975, and the Goldstone demonstration that year captured 82.5 percent at the rectenna.

And the commons has begun to be tended. Astroscale's ELSA-d and then ADRAS-J rendezvoused with genuinely uncooperative objects — a discarded upper stage tumbling with no docking fixture — and held station close enough to inspect it. The European Space Agency contracted ClearSpace-1, the first commercial removal of a piece of debris, for €86 million. That is an expensive first of a kind, and it is a first of a kind, which is the point.

Five things working, on five different principles: sample return, reuse, closure, beamed power, and removal. Not one of them is extraction. Hold that observation; the arithmetic is about to explain it.

THE ARITHMETIC

What works, what does not, and where the line sits

First, the asymmetry, which is real and enormous.

The specific energy required to escape a body is ½v². For Earth, with an escape velocity of 11.186 km/s, that is 62.6 megajoules per kilogram. For the Moon, at 2.38 km/s, it is 2.83 megajoules per kilogram — Earth costs 22.1 times as much. For Bennu, whose escape velocity is about twenty centimetres a second, it is 0.02 joules per kilogram, and Earth costs three billion times as much.

That ratio is the whole case for space resources, and it is not a small case. A kilogram of anything already out there sits at the top of the well we spend our entire launch industry climbing.

Second, the budget, for a real body, with real numbers.

Take Bennu, because we have actually been there. Its Shoemaker–Helin delta-v from low Earth orbit is 5.10 km/s for rendezvous. The return is almost free: OSIRIS-REx's departure burn, as flown, was 0.28 km/s, and Earth capture is aerobraking, which costs nothing. Allow 0.22 km/s of midcourse margin and the return leg is 0.50 km/s — the outbound leg is 10.2 times the return. The budget is lopsided in our favour, which is why the answer below is so unwelcome.

Now the rocket equation, at a flight-proven storable specific impulse of 300 seconds, giving an exhaust velocity of 2.942 km/s:

  outbound mass ratio   exp(5.10 / 2.942)  =  5.661
  return   mass ratio   exp(0.50 / 2.942)  =  1.1852

Third, mass returned per unit of propellant. Per kilogram brought home you burn 0.1852 kg of propellant at the asteroid — which is 5.40 kilograms returned per kilogram of propellant, and read alone it looks like a business.

It is not, because that propellant had to be carried out there first, and carrying it costs the outbound mass ratio:

  LEO mass per kg returned  =  0.1852 × 5.661  =  1.049 kg

So: 1.049 kilograms in low Earth orbit per kilogram returned from Bennu — propellant only, no vehicle, no capital. And that kilogram in orbit took 15.17 kilograms of propellant on the pad to put there. Sea freight from Chile to Rotterdam burns about 0.0096 kg of fuel per kilogram of cargo. Asteroid return is 1,580 times more propellant-intensive per kilogram than a bulk carrier.

Fourth, the break-even price, which is the number this chapter exists to give.

  P*  =  1.049 kg  ×  $3,059/kg  =  $3,208 per kilogram

Three thousand two hundred and eight dollars a kilogram, and that is the floor. Zero hardware, zero development, zero refining, zero capital: nothing but propellant. Give the return vehicle a dry mass of twenty percent of its payload and it becomes $7,313. Amortise $500 million of system capital over a hundred tonnes of lifetime return and the all-in figure is $12,313 per kilogram.

Against $3,208, here is the entire periodic table of commerce:

  material        USD/kg      / P*     verdict
  ---------------------------------------------------
  iridium        144,678    45.1x      clears
  gold           112,528    35.1x      clears
  platinum        41,796    13.0x      clears
  silver           1,286     0.40x     misses by 2.5x
  cobalt           33.00     0.010x    misses by 97x
  nickel           15.00     0.005x    misses by 214x
  copper            9.50     0.003x    misses by 338x
  water             0.001              misses by 3,207,908x

Four materials clear, and all four are precious metals. Every structural material misses by two to three orders of magnitude. That is the honest negative and it is severe: at every launch cost ever achieved, and at the floor of the most optimistic accounting, essentially no material is worth returning to Earth.

And now the cut, which is the part nobody costs.

Everything above treats the price of platinum as a constant. It is not. It is a function of its scarcity, which is precisely the thing the mission exists to abolish.

World platinum mine supply is about 180 tonnes a year. At $41,796 a kilogram the entire world platinum market is $7.52 billion a year — which is 1.83 launches of the SLS/Orion vehicle you would need to go and get more of it. Short-run demand for platinum is inelastic, at roughly −0.40, because autocatalyst demand is technically determined. Marginal revenue is P × (1 + 1/ε), which at that elasticity is −1.50 times price: negative. For an inelastic good, additional supply lowers total revenue.

Work it through. To book one billion dollars at today's price you must land 23.9 tonnes of platinum, which is 13.3 percent of world supply, which moves the price by −33.2 percent. You realise $668 million, not a billion — and you destroy $2.50 billion of value in the incumbent industry. $3.74 of value destroyed for every dollar earned.

Then the grade. Platinum in a carbonaceous chondrite runs about 1 ppm; in an iron meteorite about 10 ppm; the Merensky Reef of the Bushveld runs about 6 g/t. The best asteroid is 1.67 times the grade of a mine that already exists on Earth — not a hundred times. To land 23.9 tonnes of platinum at 10 ppm and perfect recovery you must process 2,392,577 tonnes of metal: 100,000 kilograms handled for every kilogram returned.

So the only materials that clear the transport floor are the ones whose price is their scarcity — and the mission that succeeds destroys its own revenue. The asteroid-mining business was never selling material. It was selling position, and it was proposing to sell the one commodity whose position it would abolish in the act of arriving. That is II.01's finding turned inside out and pointed at the sky: where scarcity is positional, the return on ending it accrues to everyone except the person who ended it.

Fifth: what would actually have to change.

Break-even scales linearly with launch price: P* = 1.049 × C_LEO. So:

  material   needs C_LEO      cut required     years at −5.05%/yr
  -------------------------------------------------------------
  silver        $1,226.42          2x                  18
  cobalt           $31.47         97x                  88
  nickel           $14.30        214x                 104
  copper            $9.06        338x                 112

At Starship's advertised $100 per kilogram — a vehicle that has not yet flown an operational payload at that price — P is $105 per kilogram and nickel still misses by 7.0 times. At the aspirational $10 per kilogram, P is $10.49 and nickel finally clears by 1.43 times while copper still misses. Base-metal return from asteroids requires roughly ten dollars a kilogram to orbit — a 306-fold cut on the best price ever charged — and on the measured trend that arrives in about 112 years.

That is the quantitative answer, and it is not a refusal. It is a date.

DREAM

What becomes ordinary

In the economy that has taken this arithmetic seriously, the first thing that is different is that nobody argues about asteroid mining any more, because the industry stopped calling itself that.

The commodity is propellant, and it is priced the way propellant has always been priced — against the delivered alternative. A kilogram of water at a municipal tap is worth a tenth of a penny. The same kilogram in low Earth orbit is worth $3,059, in a lunar halo orbit $9,717, in geostationary orbit $11,517. Nobody finds this strange, any more than anybody finds it strange that gravel is cheaper at the quarry. A depot publishes a delivered price the way a fuel terminal publishes a rack price, and a satellite operator compares it to the launch it would otherwise buy, and that is the whole transaction.

The arithmetic it sits on is visible to everyone. A five-tonne satellite going from low orbit to geostationary needs 13.8 tonnes of propellant — 2.76 times its own mass. The payload was always the minority of the stack. Propellant is the majority, propellant is water with a power supply, and water is ten percent of a hydrated chondrite by mass and 5.6 percent of the Cabeus regolith. Ten tonnes of rock for one of water; and that tonne of water is worth $3.06 million where it sits.

The second difference is that closure is a reported number. Every crewed system, every depot, every station carries its own p — the fraction of each consumable that comes round again — on the standing pack beside its mass budget, because the multiplier 1/(1 − p) is doing more work than any launch contract. At 85 percent it is 6.67 times. At 93 percent, 14.3 times. At 98 percent, 50.0 times. An engineer who raises closure from 93 to 98 percent has done more for the programme's mass budget than a factor of three in launch price would, and everybody in the building can say so with the number.

The third is that orbital shells are addressed by name, priced, and cleaned. A satellite carries a removal bond the way a mine carries a rehabilitation bond, and the bond is released when the object is gone. Operators file for a shell, not for "orbit", because the shells have different carrying capacities and everybody knows it.

And the fourth, the quietest: the question asked of any space-resource proposal is no longer how much is out there. It is where is it, what does moving it cost, and who is the buyer standing at that altitude. That question is faster, it is answerable from data that already exists, and it is right more often — which is the same sentence Chapter II.01 wrote about shortages on the ground.

DESIGN

The structure that gets there

The design is a sequence, and the sequence is ordered by what the arithmetic says is decidable now.

Stage one — sell position, not material, and sell it at the top of the well.

The first product is propellant and consumables delivered in orbit, priced against the launch cost to that orbit. This is the only space-resource proposition in the chapter with a customer, a benchmark price and a margin. The test it must pass is a single inequality:

        delivered price at the depot
        -----------------------------   <   1
        launch cost to the same orbit

That is all. The buyer does not have to believe anything about the future of space; they have to compare two prices for the same kilogram in the same place.

Stage two — raise p before you raise supply. Every percentage point of closure is worth more than a percentage point of launch-cost reduction once p is above about 0.9, because the multiplier is convex. Fund closure research before extraction research. This is not an ethical preference. It is where the derivative is.

Stage three — treat the shell as the asset and price it. Orbital slots and debris are a commons in Ostrom's exact sense: subtractable, hard to exclude from, and with a carrying capacity nobody can state precisely. So state it as a bracket and act on the bracket. Kessler's condition is that fragment generation exceeds removal:

   0.5 · k · n*²  · F   =   n* / τ        ->      n*  =  2 / (k · F · τ)

with k = σ v̄ T / V. For the 200–2,000 km shell, volume 1.269 × 10¹² km³, at a mean relative velocity of 10.0 km/s, using F = 2,900 catalogued fragments per catastrophic break-up — the mean of the 3,500 from Fengyun-1C and the 2,300 from Iridium–Cosmos, both measured — and a mean residence time of fifty years:

  assuming 1.00 m objects   n* =   4,412     rate  2.564 /yr
  assuming 0.30 m objects   n* = 196,111     rate  0.058 /yr
  geometric mean            n* =  29,417
  catalogue today                40,500

The bracket is 44.4 times wide and the catalogue sits inside it. On the central estimate we are already 1.38 times over carrying capacity. That is the finding, and the honest part of it is the width: nobody can tell you where the limit is to better than a factor of forty, and every operator is pricing as though it sits at the top of the range.

And the commons is not one commons. Capacity scales as 1/τ, and residence time varies by three orders of magnitude with altitude. A hundred-kilometre shell at 400 km, where objects decay in about a year, holds 66,796. The same shell at 1,000 km, where they persist a millennium, holds 79. They are governed by one rule. They are not one resource, and the position — altitude — is the scarcity, exactly as II.01 said.

Governance. Two conditions, and they are the same two that hold any classification honest. The carrying-capacity estimate is published with its assumptions, so a disagreement is about an identifiable term — σ, F or τ — rather than about intent. And whoever sets the fee is not whoever collects it. The fee funds removal; the estimate is made by a body that does not receive the revenue.

Where this design fails, plainly. It fails if launch prices fall fast enough that delivered propellant never beats launched propellant — the depot's whole margin is the launch price it undercuts, and a cheaper rocket is the depot's competitor, not its enabler. It fails if the shell is congested by a single actor with sovereign backing, because a fee is a price and a price only governs a party that can be charged. And it fails if σ is at the high end of the bracket, in which case the shell above 800 km was already lost while we were computing it.

DESTINY

How it holds when nobody is pushing

Three things make this self-sustaining, and they are not enthusiasm.

The depot has a benchmark, and the benchmark is public. A price that can be compared to another published price does not need advocacy. Nobody campaigns for diesel. The delivered-in-orbit price either undercuts the launch price to that altitude or it does not, and the customer can check in an afternoon.

The bond survives the operator. A removal bond posted at licence and released on verified disposal does not depend on anyone remembering why it exists. It is on the balance sheet as a restricted asset, the auditors ask about it annually, and it is released or forfeited by an event rather than by a judgement.

Closure pays the person who improves it. If p is on the standing pack, every point of it shows up in a mass budget, and a mass budget is money. The engineer who raised closure from 93 to 98 percent has a fifty-fold multiplier with their name on it.

Now the honest part — here is where this fails. It fails when a mission is sized to be impressive rather than measurable, which in this field means sized to the press release rather than to the propellant budget. It fails when the break-even is computed against a launch price that has not been charged yet, and Starship's advertised hundred dollars a kilogram is the standing temptation: a business plan whose break-even requires a 30.6-fold cut in a price somebody else controls is a bet on that person, not a plan. It fails when the carrying capacity is asserted as a number rather than a bracket, because a single number invites a single counter-number and the argument never reaches the assumption underneath. And it fails, most often, when somebody tries to win this philosophically before winning it numerically. The order matters. Number first, then the instrument, then the story.

DELIGHT

What it feels like

There is a particular pleasure in holding a bottle of water and knowing exactly what it would be worth four hundred kilometres up. It is not greed and it is not awe. It is the specific satisfaction of a unit conversion that changes what you can see — three million to one, for the same molecule, purely on account of where it is standing.

And there is a better one behind it. Once you have run this arithmetic you stop finding space either magical or disappointing. It becomes a place with freight costs, like Rotterdam, and a place with freight costs is a place you can plan for. The romance does not leave. It relocates into the detail: into the fact that the departure burn from Bennu was twenty-eight hundredths of a kilometre a second and the ocean of air did the rest of the braking for free. Into the fact that a station is quietly running its water round fifty times and calling it plumbing.

Best of all is the moment somebody in a meeting says we should mine an asteroid, and instead of arguing you ask what they would sell, and to whom, and at what altitude — and the conversation gets better. Not smaller. Better. That is what a number is for.

OPERATIONALIZE THIS

At the level of finance

The structure: a delivered-in-orbit offtake agreement with an attached disposal bond.

This is a take-or-pay supply contract of a shape a commodities treasurer already knows, priced against a public benchmark, with an environmental surety bolted on in the way a mining rehabilitation bond is bolted on. Nothing here is novel finance. That is the point — novel finance is how a physically marginal project gets funded.

The mechanics.

The commodity. Propellant, water, or consumables, delivered to a named orbit. Never "asteroid resources." The contract names an altitude, because altitude is the product.
The benchmark. The published launch price per kilogram to that orbit — $3,059 to low Earth orbit, $3,493 to geostationary by electric tug. The contract price is struck as a discount to the benchmark and floats with it. A floating strike against a falling benchmark is the honest structure here, because the benchmark is falling at 5.05 percent a year.
Volume. Take-or-pay on a minimum tonnage, with the minimum set so that the supplier's capital is amortised at a price below benchmark. At $500 million of capital, 10,000 tonnes of lifetime delivery amortises to $50 per kilogram; 1,000 tonnes to $500; 100 tonnes to $5,000, which is already above the low Earth orbit benchmark. Throughput, not geology, is the variable that decides this business.
The disposal bond. Posted at licence, sized at the expected removal cost, released on verified disposal. The only removal ever contracted is ClearSpace-1 at €86 million; a fleet-scale target of $5 million per object is 10 percent of a $50 million satellite. Restricted cash, not a provision.
The orbital-use fee. Rao, Burgess and Kaffine compute an optimal fee of $14,900 per satellite-year in 2020, rising to $235,000 by 2040 — an implied 14.8 percent a year — and find that charging it raises industry value from $600 billion to $3,000 billion, a factor of 5.0. At 2020's level the fee is 0.21 percent of the annual cost of a $50 million satellite over a seven-year life. It is not a burden. Its absence is what makes the shell worth nothing.

The balance-sheet treatment. The depot's tankage and the tug are long-lived assets and depreciate normally. The disposal bond is restricted cash, not a provision, because a provision can be released by an estimate and restricted cash cannot. The offtake, if it is genuinely take-or-pay, is an onerous-contract test for the buyer every reporting period, which is precisely the discipline that keeps the minimum volume honest.

The counterparty. A national agency first, as anchor tenant — the model that built commercial cargo and commercial lunar delivery — because an agency can sign a multi-year minimum volume against an appropriated budget and a commercial operator cannot. Only once two deliveries have been made against an agency offtake should the structure be taken to a commercial satellite operator, because now you are presenting a delivered price history rather than a projection.

The number that decides it. One inequality, on the front page:

      delivered price at the depot, USD/kg
      ------------------------------------   <   1
      launch price to that orbit, USD/kg

If that holds, this is not a space venture. It is a logistics contract with an unusually high freight rate, and it should be presented as one. If it does not hold, no amount of geology will make it hold, because the buyer is comparing two prices for the same kilogram in the same place and geology is not one of the terms.

The first ninety days.

Day	Action	Artifact
1–15	Name the orbit and pull the published launch price to it	The benchmark, sourced and dated
16–30	Compute the delivered price at three throughput levels	The throughput table
31–45	Identify the anchor tenant and their appropriated line	The counterparty memo
46–60	Draft the offtake with a floating strike	The floating-strike term sheet
61–75	Size the disposal bond against the shell's residence time	The bond schedule
76–90	Publish the carrying-capacity bracket with its assumptions	The bracket, with σ, F and τ named

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

Where in our own operation are we already paying for position rather than material, and what would we learn if we priced it that way on purpose?
What is the highest closure rate anyone here has ever achieved on any consumable, in any system — and what made it possible?
Which of our assets is genuinely valuable because of where it sits rather than what it is made of, and who was it who put it there?

Dream — what becomes possible

If every consumable we use carried its own circulation rate on the standing pack, which number would improve first, and who would improve it?
Imagine the delivered price of something we buy falling by half every thirteen years. What would we build that we currently do not?
If we priced the congestion in a shared resource we already use, what would become possible that is currently crowded out?

Design — what we build

What is the smallest contract we could write that names a place rather than a thing?
Which of our published estimates should be a bracket rather than a number, and what would we have to disclose to make that credible?
Who would set the fee, and who would collect it, and are they the same people today?

Destiny — how it holds

What would keep our version of the disposal bond in place after everyone in this room has moved on?
What is the benchmark our price floats against, and what happens to us if it falls faster than we assumed?
What is the first sign we would see that our carrying-capacity assumption was at the wrong end of the bracket, and who would notice it first?

WORKS CITED

Adilov, N., Alexander, P. J. and Cunningham, B. M. (2015). "An Economic Analysis of Earth Orbit Pollution." Environmental and Resource Economics, 60(1), 81–98.

Brown, W. C. (1984). "The History of Power Transmission by Radio Waves." IEEE Transactions on Microwave Theory and Techniques, 32(9), 1230–1242.

Colaprete, A. et al. (2010). "Detection of Water in the LCROSS Ejecta Plume." Science, 330(6003), 463–468.

Cooperrider, D. L. and Whitney, D. (2005). Appreciative Inquiry: A Positive Revolution in Change. Berrett-Koehler.

European Space Agency, Space Debris Office (2024). ESA's Annual Space Environment Report. ESA/ESOC, Darmstadt.

Hardin, G. (1968). "The Tragedy of the Commons." Science, 162(3859), 1243–1248.

Jones, H. W. (2018). "The Recent Large Reduction in Space Launch Cost." 48th International Conference on Environmental Systems, ICES-2018-81.

Kessler, D. J. and Cour-Palais, B. G. (1978). "Collision Frequency of Artificial Satellites: The Creation of a Debris Belt." Journal of Geophysical Research, 83(A6), 2637–2646.

Lauretta, D. S. et al. (2017). "OSIRIS-REx: Sample Return from Asteroid (101955) Bennu." Space Science Reviews, 212, 925–984.

Lewis, J. S. (1996). Mining the Sky: Untold Riches from the Asteroids, Comets, and Planets. Addison-Wesley.

National Aeronautics and Space Administration, Office of Inspector General (2021). NASA's Management of the Space Launch System Stages. IG-22-003.

National Aeronautics and Space Administration, Office of Technology, Policy, and Strategy (2024). Space-Based Solar Power. Washington, DC.

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

Rao, A., Burgess, M. G. and Kaffine, D. (2020). "Orbital-Use Fees Could More Than Quadruple the Value of the Space Industry." Proceedings of the National Academy of Sciences, 117(23), 12756–12762.

Tsuda, Y. et al. (2020). "Hayabusa2 Mission Status." Acta Astronautica, 171, 42–54.

United States Geological Survey (2024). Mineral Commodity Summaries 2024. Reston, VA.

Watanabe, S. et al. (2019). "Hayabusa2 Arrives at the Carbonaceous Asteroid 162173 Ryugu." Science, 364(6437), 268–272.

Note on figures. Every figure in this chapter is computed in lib/verify/VII_09.py, which prints its inputs, its intermediate terms and its results with units and sources. Metal prices are entered at their 2025 trading ranges, rounded, and marked as inputs; every conclusion is stated as a ratio to the break-even so that the argument survives a different price year. Launch prices are published list prices and program accounts. The collision arithmetic is stated as a bracket rather than a number, and the bracket's width is part of the finding. Chapter II.01 establishes scarcity as positional; this chapter is that finding applied where position is the entire cost.