I recently read Platinum Moon by Bill White, in just 4 days, it's just that much of a page-turner. Clark Lindsey has written an extensive review, which is mostly positive but has this little dig at the end:
A kilogram of pure platinum today sells for something like $53,000/kg. On the Moon even rich PGM ore would have have to be extensively refined to get anywhere close to that purity. A kilogram of raw ore would be worth a tiny fraction of that.
Until there are fully reusable vehicles flying frequently enough to LEO to bring costs there down to the low hundreds of dollars per kg, it's difficult to see how space mining can even begin to be viable.
My first reaction is to suggest that obviously high purity enrichment of platinum should be done on the Moon, and only "pure platinum" returned to the Earth - but I should first point out that Platinum Moon made the realistic argument that lunar platinum would be worth a lot more than market value in the form of commemorative coins and other trinkets - at least initially.
Extraction of oxygen from lunar regolith is a critical part of the plot in Platinum Moon and it's not unreasonable to expect a platinum enrichment facility to be capable of doing it as a side process. Similarly, extraction of aluminum from lunar regolith is an easy process which would also be available. Spinning that aluminum into tanks is simple manufacturing that could be done in-situ, and during lunar night the oxygen would naturally liquefy to make filling easier.
The landing vehicles in Platinum Moon use the fuel RP-1, a form of kerosene which is often approximated as dodecane in chemical formulas - this means it has 12 carbon atoms and 26 hydrogen atoms per molecule. On the Moon, carbon is about as rare as hydrogen, and the biggest deposits are in cold traps at the lunar poles. A simpler hydrocarbon, with higher performance, is methane with just 1 carbon atom and 4 hydrogen atoms. Liquid hydrogen could also be used as a fuel for a reusable lander. It seems inevitable that a complete propellant production plant - both fuel and oxidizers - would be an early infrastructure goal of an operational lunar platinum mine, but it would likely be separated from the mining and refining sites; requiring suborbital hops to refuel.
Estimating the size of a propellant production facility is difficult at this time. Current NASA estimates for a "pilot demonstration" plant to produce oxygen from lunar regolith are in the ~300kg range, producing up to 500kg/year. A similar sized plant in a cold trap could be expected to produce thousands of times as much, and would more likely be bound by the availability of tanks; which I imagine being transported from the distant platinum enrichment facility.. whether local production of tanks is more efficient depends on the flight rates.
Sticking with the "gateway" architecture presented in the book, lunar production of both fuel and oxidizer is game changing. Launching fuel for Earth departure stages from stockpiles on the lunar surface, via EML1, is cheaper than launching from Earth's deep gravity well. Storage of cryogenic propellants in the cold traps of the Moon until needed brings just-in-time delivery economics to spaceflight. This would allow the launch of larger processing plants and more sophisticated mining vehicles that can increase the production in a virtuous cycle.
If one could obtain pure platinum from the surface of the Moon, would it be profitable to return it to Earth at current market prices? I've previously shown that the SpaceX Dragon has a downmass of 3000kg at $28,330/kg, returning pure platinum from LEO at a profit of $24,670/kg. With the architecture described, cislunar transport is essentially free, but to make a profit, the initial costs of the architecture have to be amortized over every kg returned.
If the architecture costs low billions to setup then around 100,000kg needs to be returned over several years. For comparison, only 239,000kg of platinum was sold in 2006. The initial premium for lunar platinum would quickly fall to market levels, but would the market value of platinum fall significantly thereafter? I have had long arguments over whether or not platinum is an elastic market that is significantly effected by the opening of a new mine.. it's simply not clear what the demand for platinum is because the supply is so low at present, but it should be clear to see from the growing price of platinum over the years that more and more demand is out there for a very limited supply.