The debacle of Starliner’s record, IMHO, is not helpful to the notion of selling it.
The other huge fly in the ointment is that, contra what NASA specified, Starliner is not launch vehicle agnostic (It’d take some work to put it on another LV), plus, it’d requite a man-rated LV. Right now, that Atlas 5, of which there are, what, 5 left?
So, to me, the real question is, could anyone make a profit from buying Starliner and using it for the four or five possible ISS missions? If ISS is decommissioned as planned (and it looks like it will be), that’s all there will be. Without a sure replacement for ISS in LEO, that’s the ballgame, because Starliner would be rather useless for the proposed Gateway lunar station (That I doubt will ever exist).
When I first heard that Boeing was considering selling Starliner, the old saying “Buying a pig in a poke” came to mind. It still does.
If a rocket has a capsule with an abort system on it, it is “human rated.” I can’t imagine what ULA could do to Vulcan to make it more reliable, given that it’s meant to carry billion-dollar satellites.
It’s not so much reliability as compatibility. For example, if Starliner initiates an abort you would want Vulcan to stop thrusting to reduce the chances of recontact. I doubt that capability is wired into Vulcan at present.
I would assume that it has failure-onset detection and everything else they did to Atlas to carry Starliner, because Vulcan was supposed to be the Atlas replacement.
Sorry, but I disagree; I wasn’t talking about (and should have made clear) sane or rational definitions, I was talking about NASA human rating. (the same NASA that never human-rated Shuttle, for example, and now pretends SLS meets the human-rating standards they apply to others).
There are all sorts of NASA “human rating” requirements (except for their own projects), even for a vehicle designed to get that rating. (Look at all the absurd hoops they made SpaceX jump through to human-rate the F9, and for that matter, why FH still isn’t human rated.) Also, though Vulcan is indeed designed to human rate, I think ULA said they don’t currently plan to do so, as it’s a long and expensive process (and would include having to human-rate the BE4 engines, etc).
While Starliner was designed to notionally launcher agnostic, it isn’t as simple as putting it on a different rocket and lighting the fuse. There has to be wind tunnel testing to ensure the configuration will be stable. Look at the interface between the Starliner and the Atlas V. They had to leave that gap as a result of wind tunnel testing results. There’s also vibration and other testing that must be performed. Going with a Falcon 9 would be easier because it’s already human rated. It still would need other tests before being accepted. Vulcan would be a likely choice but getting everything approved through NASA’s safety requirements. It’s a different argument as to whether all of these requirements are reasonable.
“At the time of writing, there are only 15 Atlas V missions outstanding. One is for ViaSat, eight for Amazon’s Project Kuiper, and the remaining six for Boeing’s Starliner missions to the ISS. Many of these boosters have already been delivered to Cape Canaveral and are in storage, ready for use. The timetable could see Atlas V still flying as late as 2030.
Recently, further delays were announced to Starliner’s schedule. Following the issues experienced on the CFT-1 mission, Boeing requires more time before flying the spacecraft again. It is conceivable that the ISS could be decommissioned before Boeing has had time to fly all of its expected missions to the Station. This situation could find ULA with several spare Atlas Vs.”
The issues with Starliner appear to me to be fundamental, supply-chain driven and aren’t really Boeing’s problem to fix.
Any buyer would have to be aware that righting this ship is going to involve expensive mods made largely by third party suppliers (*cough* Rocketdyne *cough*) and at a price THEY will charge any potential buyer even after its been purchased from Boeing.
And that’s not even addressing the launcher problem.
The best business case I can make for Starliner is to entirely avoid the LV and subsystems problems via using it for one-way unmanned missions to ISS after the ISS deorbit.
Reusability can not only reduce costs, it can dramatically improve reliability. The SpaceX design philosophy pushed as much as possible into the reusable capsule, including the thrusters and abort system. The latter gave some commenters on Spacenews the vapors. In contrast, the service module mostly is an aerodynamic stabilizer, a cargo space, and a simple solar panel holder, all basically simple passive systems.
In contrast Starliner has far too many vital and complex systems in their expendable service module, so it’s very hard for them to learn from design mistakes.
The Dragon capsules don’t have a service module, they have a trunk. The trunk carries solar panels for electricity and radiators to get rid of waste heat. They have fixed fins for stabilization during an abort. That’s about it. Only the Cargo Dragon trunks ever carry unpressurized cargo, the Crew Dragons never do. I believe that’s the case because in the event of an abort, you wouldn’t want the extra mass of unpressurized cargo going along with the ride. You could add explosive bolts to blow the cargo attachment to the trunk, but that adds another possible failure mode and would require testing to verify the design.
To the best of my knowledge, each Crew Dragon gets a new heat shield after every flight. It many also get new parachutes but I’m not sure about that. Of course, it also gets a new trunk. There may be other things that get replaced that I don’t know about.
Starliner has to have a new Service Module built. That module includes 4 abort rocket engines and many smaller attitude control rocket engines, along with a complex set of propellant tanks and valves. It also needs a new heat shield and perhaps new parachutes. My guess is it’ll also need a new set of airbags. Refurbishing a Starliner for another flight is a much more expensive proposition than what’s needed to a Crew Dragon. SpaceX has flown Crew Dragons many times now, so they’re well along the learning curve for the process. There has only been three Starliner flights, each separated by about 2 years. The Boeing personnel aren’t nearly as far along on the learning curve.
There was no one willing to continue investment in this scheme. So it appears the SABRE HTOL space-plane is all but dead.
I can remember Henry Spencer way back when saying the company was solving the wrong problem in rocket design, as LOX is relatively cheap to obtain. And mfg. it in quantity (on the ground, which is key) will be front and center as part of Starship operations as Eric Berger points out…
How plausible are the trade offs of hypersonic boost or first stage, whatever? Some LOX against the weight and complexity of two power plants, or in this case, a chimera that surely compromises both modes of operation and a trajectory that minimizes the path through the atmosphere against a much longer, oblique path. That’s a mighty narrow needle to thread, or really, series of needles.
As far a Starliner is concerned, is the revenue stream enough to pay for the work that still needs to be done? It would take very deep pockets and a lot of guts to pay as much as a dollar for the thing at this point. Is it even likely that Boeing could fly it again before ’26, let alone a new owner?
Only partially tongue-in-cheek I have suggested elsewhere that Impulse Space might be able to transform the problematical Starliner into a real spacecraft.
1. Starliner’s present shortcomings are heavily thruster-related. Impulse Space makes thrusters.
2. Starliner’s thruster problems are heavily concentrated in its non-recoverable service module. A service module bears a more-than-passing resemblance to a space tug. Impulse Space also makes entire space tugs. In addition to crafting a service module that actually works, Impulse Space might even be able to produce one that is also recoverable – either separately or while still attached to Starliner.
3. Given the close ties Impulse Space has to both its founder’s former employer, SpaceX, and to commercial LEO space station start-up Vast, an Impulse Space-revamped Starliner – modified to launch on Falcon 9 – might be just the ticket to share crew transport duties with SpaceX’s Crew Dragon once Vast starts lofting modules of its newly-proposed Haven-2 station. Given Haven-2’s planned size and crew capacity, SpaceX might otherwise have to build additional Crew Dragons to handle that chore until Haven-2 is built out enough to require a Starship-based crew rotation solution. At which point, at least one of the other, smaller, Commercial LEO Destination stations might also be on-orbit and sized more appropriately for continued use of Crew Dragon and Starliner to transfer crews.
“Developing Starliner has cost over $5B, with both NASA and Boeing sharing the expenses.”
And?
That’s just the sunk cost fallacy being applied to someone else’s project. The main benefit the author brings up is government wanting something in addition to SpaceX but that doesn’t make Starliner special or provide any competitive advantage over future competition. Limited lifespan combined with uncertain costs to finish along with the other issues people raised above.
The debacle of Starliner’s record, IMHO, is not helpful to the notion of selling it.
The other huge fly in the ointment is that, contra what NASA specified, Starliner is not launch vehicle agnostic (It’d take some work to put it on another LV), plus, it’d requite a man-rated LV. Right now, that Atlas 5, of which there are, what, 5 left?
So, to me, the real question is, could anyone make a profit from buying Starliner and using it for the four or five possible ISS missions? If ISS is decommissioned as planned (and it looks like it will be), that’s all there will be. Without a sure replacement for ISS in LEO, that’s the ballgame, because Starliner would be rather useless for the proposed Gateway lunar station (That I doubt will ever exist).
When I first heard that Boeing was considering selling Starliner, the old saying “Buying a pig in a poke” came to mind. It still does.
If a rocket has a capsule with an abort system on it, it is “human rated.” I can’t imagine what ULA could do to Vulcan to make it more reliable, given that it’s meant to carry billion-dollar satellites.
It’s not so much reliability as compatibility. For example, if Starliner initiates an abort you would want Vulcan to stop thrusting to reduce the chances of recontact. I doubt that capability is wired into Vulcan at present.
I would assume that it has failure-onset detection and everything else they did to Atlas to carry Starliner, because Vulcan was supposed to be the Atlas replacement.
Sorry, but I disagree; I wasn’t talking about (and should have made clear) sane or rational definitions, I was talking about NASA human rating. (the same NASA that never human-rated Shuttle, for example, and now pretends SLS meets the human-rating standards they apply to others).
There are all sorts of NASA “human rating” requirements (except for their own projects), even for a vehicle designed to get that rating. (Look at all the absurd hoops they made SpaceX jump through to human-rate the F9, and for that matter, why FH still isn’t human rated.) Also, though Vulcan is indeed designed to human rate, I think ULA said they don’t currently plan to do so, as it’s a long and expensive process (and would include having to human-rate the BE4 engines, etc).
In a sane world, you’d be right. 🙂
While Starliner was designed to notionally launcher agnostic, it isn’t as simple as putting it on a different rocket and lighting the fuse. There has to be wind tunnel testing to ensure the configuration will be stable. Look at the interface between the Starliner and the Atlas V. They had to leave that gap as a result of wind tunnel testing results. There’s also vibration and other testing that must be performed. Going with a Falcon 9 would be easier because it’s already human rated. It still would need other tests before being accepted. Vulcan would be a likely choice but getting everything approved through NASA’s safety requirements. It’s a different argument as to whether all of these requirements are reasonable.
https://www.nasaspaceflight.com/2024/10/atlas-v-history/
“At the time of writing, there are only 15 Atlas V missions outstanding. One is for ViaSat, eight for Amazon’s Project Kuiper, and the remaining six for Boeing’s Starliner missions to the ISS. Many of these boosters have already been delivered to Cape Canaveral and are in storage, ready for use. The timetable could see Atlas V still flying as late as 2030.
Recently, further delays were announced to Starliner’s schedule. Following the issues experienced on the CFT-1 mission, Boeing requires more time before flying the spacecraft again. It is conceivable that the ISS could be decommissioned before Boeing has had time to fly all of its expected missions to the Station. This situation could find ULA with several spare Atlas Vs.”
If Ortberg thinks he can solve Boeing’s problems by selling the losers off, there won’t be a boeing left.
The only viable response.
The only viable response.
The issues with Starliner appear to me to be fundamental, supply-chain driven and aren’t really Boeing’s problem to fix.
Any buyer would have to be aware that righting this ship is going to involve expensive mods made largely by third party suppliers (*cough* Rocketdyne *cough*) and at a price THEY will charge any potential buyer even after its been purchased from Boeing.
And that’s not even addressing the launcher problem.
“Pig in a poke” sounds about right.
The best business case I can make for Starliner is to entirely avoid the LV and subsystems problems via using it for one-way unmanned missions to ISS after the ISS deorbit.
At this point it’s a systems integration disaster…and who is the systems integrator?
Good point that wasn’t addressed at the link
Reusability can not only reduce costs, it can dramatically improve reliability. The SpaceX design philosophy pushed as much as possible into the reusable capsule, including the thrusters and abort system. The latter gave some commenters on Spacenews the vapors. In contrast, the service module mostly is an aerodynamic stabilizer, a cargo space, and a simple solar panel holder, all basically simple passive systems.
In contrast Starliner has far too many vital and complex systems in their expendable service module, so it’s very hard for them to learn from design mistakes.
The Dragon capsules don’t have a service module, they have a trunk. The trunk carries solar panels for electricity and radiators to get rid of waste heat. They have fixed fins for stabilization during an abort. That’s about it. Only the Cargo Dragon trunks ever carry unpressurized cargo, the Crew Dragons never do. I believe that’s the case because in the event of an abort, you wouldn’t want the extra mass of unpressurized cargo going along with the ride. You could add explosive bolts to blow the cargo attachment to the trunk, but that adds another possible failure mode and would require testing to verify the design.
To the best of my knowledge, each Crew Dragon gets a new heat shield after every flight. It many also get new parachutes but I’m not sure about that. Of course, it also gets a new trunk. There may be other things that get replaced that I don’t know about.
Starliner has to have a new Service Module built. That module includes 4 abort rocket engines and many smaller attitude control rocket engines, along with a complex set of propellant tanks and valves. It also needs a new heat shield and perhaps new parachutes. My guess is it’ll also need a new set of airbags. Refurbishing a Starliner for another flight is a much more expensive proposition than what’s needed to a Crew Dragon. SpaceX has flown Crew Dragons many times now, so they’re well along the learning curve for the process. There has only been three Starliner flights, each separated by about 2 years. The Boeing personnel aren’t nearly as far along on the learning curve.
So you agree with me.
Slightly off topic but relevant to failing business cases.
Reaction Engines Officially Goes Belly Up
There was no one willing to continue investment in this scheme. So it appears the SABRE HTOL space-plane is all but dead.
I can remember Henry Spencer way back when saying the company was solving the wrong problem in rocket design, as LOX is relatively cheap to obtain. And mfg. it in quantity (on the ground, which is key) will be front and center as part of Starship operations as Eric Berger points out…
How plausible are the trade offs of hypersonic boost or first stage, whatever? Some LOX against the weight and complexity of two power plants, or in this case, a chimera that surely compromises both modes of operation and a trajectory that minimizes the path through the atmosphere against a much longer, oblique path. That’s a mighty narrow needle to thread, or really, series of needles.
As far a Starliner is concerned, is the revenue stream enough to pay for the work that still needs to be done? It would take very deep pockets and a lot of guts to pay as much as a dollar for the thing at this point. Is it even likely that Boeing could fly it again before ’26, let alone a new owner?
Only partially tongue-in-cheek I have suggested elsewhere that Impulse Space might be able to transform the problematical Starliner into a real spacecraft.
1. Starliner’s present shortcomings are heavily thruster-related. Impulse Space makes thrusters.
2. Starliner’s thruster problems are heavily concentrated in its non-recoverable service module. A service module bears a more-than-passing resemblance to a space tug. Impulse Space also makes entire space tugs. In addition to crafting a service module that actually works, Impulse Space might even be able to produce one that is also recoverable – either separately or while still attached to Starliner.
3. Given the close ties Impulse Space has to both its founder’s former employer, SpaceX, and to commercial LEO space station start-up Vast, an Impulse Space-revamped Starliner – modified to launch on Falcon 9 – might be just the ticket to share crew transport duties with SpaceX’s Crew Dragon once Vast starts lofting modules of its newly-proposed Haven-2 station. Given Haven-2’s planned size and crew capacity, SpaceX might otherwise have to build additional Crew Dragons to handle that chore until Haven-2 is built out enough to require a Starship-based crew rotation solution. At which point, at least one of the other, smaller, Commercial LEO Destination stations might also be on-orbit and sized more appropriately for continued use of Crew Dragon and Starliner to transfer crews.
“Developing Starliner has cost over $5B, with both NASA and Boeing sharing the expenses.”
And?
That’s just the sunk cost fallacy being applied to someone else’s project. The main benefit the author brings up is government wanting something in addition to SpaceX but that doesn’t make Starliner special or provide any competitive advantage over future competition. Limited lifespan combined with uncertain costs to finish along with the other issues people raised above.