Over the next decade or so, Elon Musk proposes building 1,000 Starships, and flying each one three times a day (1,000 times per year)...for at least ten years.

His goal: to first move 100 megatons (100 million tons) of cargo to low Earth orbit (each Starship can carry about 100 tons to LEO), and then relaunch all 1,000 ships loaded with 100,000 passengers (each Starship can carry about 100 passengers comfortably). There in orbit, the passenger ships will presumably pick up the cargo, and the whole lot will move en masse to colonize Mars.

SpaceX Starship MK1 in Texas

SpaceX Starship MK1 in Texas. Image source: SpaceX.

It's an ambitious plan. CNET actually uses the word "miraculous" -- and they're not wrong. To help visualize what Musk wants to do, 100 megatons is approximately the weight of the 13,000-mile-long Great Wall of China ... times two.  

Is this even possible? And if so, how much would these insanely grand designs of Mr. Musk cost his space company, SpaceX, to implement?  

Tallying up the costs

The answers may surprise you.

Speaking at the U.S. Air Force's annual "pitch day" in Los Angeles back in November, Musk told Space and Missile Systems Center commander Lt. Gen. John Thompson that the fuel needed to launch Starship to orbit atop a SuperHeavy booster, and bring it back to Earth, will only cost about $900,000 per launch. Adding in other operational costs (repairing damage from rough landings, paying salaries, and so on), the incremental cost of each launch will still be only twice that: "If you consider operational costs, maybe it'll be like $2 million," Musk told the general.  

How much to build the rocket?

So much for operating costs. What about development costs? (Starship hasn't actually flown yet, after all.) What about the cost to build each Starship rocket?

In an interview with CNN Business last September, Musk said that when all's said and done, developing Starship will cost "probably" $2 billion or $3 billion. That's a pretty hefty sum, but amortized across 10 million flights (1,000 rockets flying 1,000 times a year over 10 years), it dwindles to insignificance -- adding only about $250 to the cost of each flight.  

Likewise construction costs. SpaceX puts the "list price" of its Falcon 9 rocket at $54 million, exclusive of the costs to fuel and launch it.  

If you assume that Starship, with four times Falcon's payload, costs roughly four times as much to build, this implies that building a single Starship might cost roughly $216 million -- about the cost of a Boeing 767 airplane. A thousand Starships would accordingly cost $216 billion.

In fact, they might cost a bit less than that. Instead of the expensive aluminum-lithium alloy that Falcon is built from, Musk says Starship will be made of steel that costs SpaceX only "$3 a kilogram." But even assuming roughly proportional construction costs, $216 billion isn't so bad. Amortized over 10 million launches, again the cost shrinks to form a very manageable number: just an extra $21,600 per launch.  

Amortizing both development and construction costs over 10 million launches and adding them to Starship's operating cost thus raises the total cost of each Starship launch to about...$2.022 million.

Small numbers, a big number, and one very big problem

And that's probably the most important takeaway here: Amortized over a sufficiently large number of launches, the cost of Starship essentially shrinks to its operating cost: $2 million. Inventing and even building the thing doesn't add much cost at all!

Thus Musk's plan for colonizing Mars actually does look affordable -- except for one big problem. Yes, you can turn big numbers into small numbers by amortizing them over 10 million launches. But that still leaves those two Great Walls of China that Musk says we need to move -- and the 10 million launches SpaceX must conduct to move them into orbit.

Even at a low, low $2.022 million per launch, times 10 million launches, this project will ultimately cost SpaceX the staggering sum of more than $20.22 trillion.

Which is a very big number.

It might even become an impossibly big number for Musk to surmount, when you consider that he's been promising space fans that a ticket to Mars aboard one of his Starships will cost about $500,000. Even if every one of the 100,000 colonists Musk plans to recruit ponies up and pays full freight, that's still going to end up raising only $50 billion in ticket revenue...which still leaves Musk $20.17 trillion short of covering his costs.  

And that's OK.

Not Mars, but Earth

Why? Because even if the U.S. government, or 100,000 of its citizens, won't willingly foot a $20 trillion bill to make Mr. Musk's Mars dream a reality, the prospect of $2 million-a-shot space launches still has a lot of potential for non-Mars-colonization-related purposes.

Assume Starship works as promised. Each rocket can launch three times a day year-round, and can last for 20 to 30 years (pretty much like an airplane). Assume further that each Starship can be built for $200 million or so (also like an airplane). Even with a fuel bill significantly more than any airplane's, I suspect a lot of people will be able to find a lot of uses for Starship right here on Earth (or at least nearby).  

Just to start with, NASA's cost of resupplying the International Space Station with just a few tons of cargo a year is approaching $1 billion a mission. If Starship can deliver 100 tons of supplies for just a couple million dollars, I suspect NASA will find that prospect mightily attractive.

And here's another thing: Launching 100 passengers into space for $2 million works out to $20,000 a ticket for space tourists. That's better than a 90% discount to the prices that Virgin Galactic (NYSE:SPCE) and Jeff Bezos's Blue Origin want to charge.

Elon Musk also plans to use Starship for point-to-point travel around the globe as speeds faster than any airplane on Earth has ever dreamed of. Currently, Tripadvisor is quoting first-class tickets from New York to Shanghai at around $10,000 for a 24-hour flight. Might well-heeled travelers pay twice that price for a chance to cut the travel time down to less than one hour? I think they might.  

And then there's the obvious: Using Starship to transport satellites to orbit.

Currently, "expendable" rocket launchers such as United Launch Alliance and Arianespace charge tens or even hundreds of millions of dollars to lift just a few tons of cargo to orbit. Even SpaceX itself charges $62 million for a Falcon 9 launch carrying just 23 tons to LEO. Dropping that cost to just $2 million for 100 tons of cargo will mean an order-of-magnitude improvement in launch cost per ton -- actually, a couple orders of magnitude. By my calculations, moving a ton of cargo to orbit with Starship should cost about 1/125th the price of moving the same ton of cargo to orbit with Falcon 9.  

If SpaceX gets Starship built, and proves it can fly, this is going to change space launch forever. Whether or not we ever colonize Mars, Starship will give birth to an entire new space industry based right here on Earth.