Many believe that reusable vehicles will ultimately give us low cost access to space. The Shuttle was supposed to do this. Spacex is rapidly approaching this goal with the first stage. I assume that once the 1st stage is recovered they will soon try to recover 2nd stages, capsules, solar panels, trunks, payload fairings, all the way down to minutiae like GSE cables and dust covers. The eventual goal being airline like reuse. This will eventually lead to a beautifully refined optimized jewel like perfectly executed complex space launch system. Every component will be reusable, repairable, refurbishable .
Right now if a vacuum Merlin has a bad day you loose the whole mission. In the airline industry engine failures are a regular occurrence, but since they have robust intact abort capability and procedures we don't hear about it. If a rocket is going to achieve airline like operations it will need this capability as well. Once the value of the reusable vehicle is as great or greater than the payload intact abort gets even more compelling.
Some large helicopter systems may require 60 hrs of maintenance for every flight hour.
How much maintenance will a reusable space craft require?
Amortizing the cost of developing the reusable space craft will need to be recovered, fixed and variable operations costs, rebuild and overhaul costs etc.... etc... etc... all of this will need to be covered. Suppose spacex does 100 flights a year and the rocket is 100% reusable and propellant is free.... They currently have ~4000 employees this means with all other costs at zero the standing army at 100 launches a year will cost $4M a launch if everything else was free.
Even with this best case scenario this is a long path and with chemical propulsion the gravity well is deep.
I can buy any number of Rolex watches (lowest cost on on the page is $31K) . I expect it to be a fine example of swiss craftsmanship. If it ever breaks or misbehaves it will be delicately repaired by a skilled artisan. It will be a family heirloom for generations to come. If I go on amazon I can find digital watches starting at $3. By my estimation the $3 digital watch probably keeps time as well as the very best all mechanical Rolex. (For an apples to apples comparison the lowest cost mechanical watch I could find is $20.00)
No one is going to repair a $3 watch, you just throw it out and get another. Even if you think throwing it out is wasteful, is it really more wasteful than the standing army of rolex dealers and repair centers that stand behind the use forever rolex?
With the early microcomputer , all of the chips, and boards were replaceable. If a ram chip died it would be individually replaced. Now the fabrication is so cheap dead motherboards go to scrap. No one seriously considers repairing them. (Amazon lists 28,000 motherboards under $25.00)
What if space craft were so cheap that if one failed it wasn't a big deal? You just launched another one...My guess is that if there is a production failure on the $3 digitial watch they do no rework.
For some kinds of cargo this is not acceptable (Humans for instance)
My friend Joe Caroll has a saying to reduce the cost of space we need reusable rockets, or reusable rocket factories. Without reuse things just get simpler. At some point if you can build them cheap enough in an automated way, expendable is cheaper than reusable.(How may of you gray hairs remember TV repair shops?)
A rocket is a complex item with first, second, third stages, different engines,tanks, fittings etc...Each one optimized, a reusable rocket factory will need a flight rate high enough to justify this factory. If your only building 10 a year does it make sense to build a fully automated factory? How about 640? You will use a lot more automation building 640 than you will building 10.
This brings me to my favorite rocket that almost was, OTRAG See here, here and here. Each 1 Ton rocket had 64 identical Common Rocket Propulsion Units (CRPU) . Each otrag CRPU had one moving element a single ganged on-off throttle valve.
Also the increased drag and many of the square/cube scaling issues that are bad for Otrag are much worse for a nano-otrag.