Thread by @DrPhiltill

More detail on a non-intuitive fact from the Space Resources Roundtable talk this week: to out-compete Earth-launched rocket fuel, the rocket fuel made from lunar or asteroids resources does *not* have to be cheaper. In fact… /1

2/…it can be a LOT more expensive than Earth-launched rocket fuel and still be competitive. Economists say it like this: it doesn’t need an absolute advantage; it only needs a comparative advantage. Here is the example I thought would be interesting enough to tweet…

3/ To launch humans to the Moon (or Mars since the delta-v is about the same), the SpaceX starship will need about 8 more launches to refuel before leaving Earth orbit. Let’s say the cost per launch gets down to $10M each. So one Moon (or Mars?) mission will cost $90M.

4/ There will be some maximum rate of launches. If Elon wants to settle Mars, the rate will need to be very high. Going slow would be costlier since Mars needs to become as self-sustaining as possible as quickly as possible. Let’s say for a rough example the launch rate is 1/day.

5/ (If they launch every day year-round, the cost is usually much higher due to non-alignment of the planets, but is still do-able for less than a factor of 10 increased cost and acceptably longer transit times, but let’s ignore that since this is just an example of the concept.)

6/ So Elon would not do these launches unless they were providing more value than the money he puts in to do them. What is the value of his life-goal to start civilization on another planet? It is probably infinite return to him. But let’s be utterly conservative & say only 40%.

7/ So if Elon is getting only 40% return of value for his $90M expense for every load of settlers to Mars, that equates to $126M value, or $36M net gain in value, occurring once every 9 days. His profit is thus $4M per day. Now compare using lunar water…

8/ If Starship is refueled using rocket propellant made from lunar (or asteroid) resources, then every Starship launch can carry settlers. Let’s say the lunar propellant costs 20% more than Earth-launched. So each load of travelers requires the same cost as 8 more launches +20%.

9/ That means each launch costs $106M, an increase of $16M compared to Starship bringing all the propellant. But now, Elon gets $126M in value every day, rather than once every 9 days. This yields net profit of $20M per day, which is 5x higher…despite paying 20% more.

10/ In general, the lunar-derived propellant can be as much more expensive than Earth-launched propellant as the value customer attaches to the missions. If Elon sees that settling Mars is worth 200% more than the cost, then the break-even point for lunar water is +200%.

11/ But this is an academic question because as the rest of my talk showed, lunar-derived rocket fuel will quickly become cheaper than Earth-launched. It might take a few years, on the order of 1 to 10 years, for that to happen due to experience curve and economies of scale.

12/ I gave several examples of how it becomes cheaper than Earth-launched. These examples use the following economic factors. 1) Hardware fabrication cost is reduced by optimizing reliability. This depends on launch cost. As launch cost drops, hardware fabrication drops faster.

13/ Second, there is a well-known “experience curve” in industry. The more you have produced, the lower the costs become. Economists have extensive data quantifying this. Wright’s Law is a formula to describe this experience.

14/ Third, as the lunar business grows, it can gain economies of scale. Economists have documented this with extensive data, too.

15/ So here is the first example how these factors will affect lunar rocket fuel production. This first case uses a lunar water mining system that I worked on for a NASA grant in 2019-2020, “Aqua Factorem”.

16/ This case starts in year 1 as cheaper than terrestrial water all the way from the Moon to Geostationary Transfer Orbit, where it can be used to boost telecommunication satellites to their final orbits in GEO. We showed in the NASA study that this business case closes. But…

17/ …it gets even better because by year 14 it is now cheaper all the way to Low Earth Orbit, even though launch costs are dropping as fast as optimists say they will. And this was despite the unfair assumption that the lunar mining had to use much more expensive launch systems!

18/ So let’s get rid of the unfair assumption and use the same projected Starship launch cost for lunar mining. The optimization of reliability shifts because of this, too. The result: even by year 1, lunar water is now cheaper all the way down to Low Earth Orbit.

19/ That is actually too optimistic. There must be a few years of getting the bugs worked out, I am sure. So let’s be super conservative and assume the analysis team’s cost estimates (hardware AND ops costs) were off by a whopping factor of 5…

20/ Actually, I skipped a step. Before we increase costs by a factor of 5, let’s look at the effect of using solar electric thrusters. The case for lunar rocket fuel gets even better.

21/ Now let’s bump the costs a factor of 5 to see what happens if the analysts grotesquely underestimated lunar mining costs. If you only look at year 1, it seems that lunar water cannot compete, not even in low lunar orbit! But lunar water inevitably wins to LEO by year 11.

22/ (In that prior plot, ignore everything after year 10 when the costs start to plummet dramatically. I included an additional factor starting in year 10 in that plot, which I will explain in a moment.)

23/ So let’s try another study besides Aqua Factorem. How about the well-known study by Charania & DePasquale? Skeptics of lunar resources sometimes point to that study to say lunar resources are not competitive anywhere off the Moon. (See: trs.jpl.nasa.gov/bitstream/hand…)

24/ I had trouble interpreting parts of that study but did my best to replicate it, and sure enough, it shows that lunar water is not viable off the Moon. But this uses unfair launch costs.

25/ So let’s make just one change to that study. Let’s assume Starship is used to launch the mining hardware to the Moon. Suddenly, lunar water is now out-competing Earth-launched water in successively larger regions of space. But it gets better…

26/ Now let’s use electric propulsion to move the lunar water to its point of sale. (Let’s also add an interest cost for these longer transit times, for completeness, though it has little effect.) Lunar now beats Earth in GTO by year 10. But I think it gets much better…

27/ …because I believe the ops cost estimate is too high at $46M per year for the starter system that produces only 69 t of water per year. I believe that can be reduced a factor or 3 (or more!) if it is a commercial rather than NASA-led operation (no reliability premium).

28/ But even if you take it at face value, lunar rocket fuel is only a small % higher than Earth-launched fuel all the way down to Low Earth Orbit starting in year 1, so by comparative advantage (see start of 🧵) it still wins over Earth-launched fuel everywhere in space.

29/ And so finally, coming back to that extra factor after Year 10 in this plot. Here, I was saying, what if these ultra low transportation costs reach the tipping point to enable new industries in space? (Which will certainly happen.) So…

30/ Let’s assume lunar metal-making enables fabrication of large structures in space rather than launching it all from Earth. And lunar propellant giving cheap boosts helps Space Based Solar Power over the tipping point for broader economic viability.

31/ But it needn’t be SBSP that plays this role. It could be any additional in-space activity that uses rocket fuel. The question is, what happens when lunar mining business doubles due to other products besides water, and when boost services triple above Elon settling Mars?

32/ So the economies of scope and experience curve factors reduce the costs to maintain and use capital assets and skills even further. The result, per the equations based on extensive industry data, is a giant reduction in the cost of rocket fuel in-space.

33/ Conclusion: I do believe human civilization is on the cusp of moving beyond Earth to establish a bigger, more capable, more exciting world. When we can stop relying solely on Earth for all our economic resources, we have reached the turning point. We are very close. /end🧵.