[nixos]

His biggest (current) problem regarding SpaceX is that satellites aren't mass-produced.

If it was, buying a spot on a Falcon or on a Delta would be simple:

TCOFalcon = cost of launch + self-insurance-markup.

TCODelta = cost of launch + self-insurance-markup.

Falcon costs $1233/lb

Delta IV costs $8694/lb.

To compare apples to apples, to launch 50,000 lbs (a full Falcon FT 50,000 lb), a Falcon would cost 61,650,000, and a Delta IV would cost 434,700,000).

A Delta IV failed once out of 33, and the Falcon 9 failed three out of 29.

Therefore, TotalCostOfLaunch = CostOfLaunch+costOfSatellite * failureRate.

For TCODelta = TCOFalcon

CostOfLaunchDelta+costOfSatellitefailureRateDelta = CostOfLaunchFalcon+costOfSatellitefailureRateFalcon.

Plugging in Numbers,

434,700,000+c * (1/33) = 61,650,000+c * 3/29

c/33-2 * c/28 = 61,650,000 - 434,700,000

c = 5,100,126,428

Any satellite worth less than five billion (!! That's an _insanely_ expensive single mission) would be cheaper to launch on a Falcon, despite its failure rate

The only problem is that you have to wait for a new satellite

[ethbro]

I see it as somewhat of a catch-22, which I think is the point of dogfooding the boosters with SpaceX's own launch demand.

Nobody builds commodity satellites because there's no cost efficient way to launch them (microsats aside), and nobody tries to pioneer more cost efficient but lower reliability launch systems because there's no proven demand.

Your reasoning shows the edge under current economics. But I think the real money will be made once we get to "Well, I could build a second satellite for lower unit cost and have two in orbit." Because when demand shifts to that, suddenly anyone without a cost-efficient launch system to offer gets priced out of that chunk of the market.

[nickik]

There were only 2 failures. There was one partial failure, but the primary mission was successful and the secondary mission was partially successful.

I also think that the Falcon 9 has changed far more during its live time, compared to the Delta 4, so you would expect some more failures.