I don't think we disagree. I'm not a fan of CLPS. Using small, inexperienced, resource constrained companies increases the risks of failure... NASA is insulated against cost overruns, unless they bankrupt the contractor, in which case NASA loses the whole mission. That happened with Masten.
This is a naive opinion. The reality is that big companies are just as prone as small companies to technical failures with huge cost and schedule implications — see repeated Boeing Starliner flight and test problems. And the reality is that big companies are just as prone to terminating agreements and pulling out of projects — see Boeing pulling out of XS-1 at DARPA when they became resource-constrained in the wake of 737 MAX costs. (Not picking on Boeing — these examples just came to mind.) No organization is promised mission success because of its size. Success comes down to the quality of mission formulation, the vigilance of mission management during development, and a certain degree of luck. These things will be good or bad regardless of company size. Good procurements literally ask for and compare the resumes of team leadership. Bad procurements rely on lazy impressions of organizational experience.
I started the COTS program on which this lunar lander effort is partly modeled and served as its first program exec. We had one failure (RpK terminated because of fundraising issues), one middling success (OSC, now NG, Cygnus/Antares), and one spectacular success (SpaceX Dragon/F9). There was no big versus small pattern of success/failure. That program provided NASA’s first successful launch and space transportation developments in decades and now serves as the model, more or less, for all new human space flight development programs since. The underlying key to that program’s success was being humble enough to understand that we don’t know enough to pick the single best performer for any particular procurement. Rather, we need to place several bets. This lunar lander program is following the same strategy and will see similar a spread of results.
They will be seen as NASA missions so failures will do some reputational damage to NASA
These lunar lander missions are the planetary science equivalent of the Explorer missions from the astronomy and heliophysics side of the house. The Explorer Program had two spacecraft failures in 1999. I doubt anyone can bring up their names from memory or find a press article on those failures. What is remembered are bigger missions with more press going in and more expensive failures coming out. Examples like MCO/MPL or HST optics. And even those failures arguably had no long-term implications for the “reputation” of their parent programs or NASA science (or NASA) in general. And then there are the human space flight failures with loss of life that do loom large in NASA history. So there’s failures, Failures, and FAILURES. They’re not all the same, and we shouldn’t paint them all with the same brush. NASA, including planetary science, needs pockets of experimentation where new concepts can be tried, risks taken, and failures accepted and learned from.
and science payloads that NASA funded will be lost.
Like the lower-ends of the Explorer Program for astrophysics and heliophysics, for the most part, these are planetary science payloads that would never have flown (and planetary science researchers that never would have been flight-trained) without this program. Either you take some risks at the low-end to get more instruments flown, more data back, and more researchers flight experience. Or you forgo these kinds of riskier, lower-end programs and either take new instrument and new PI risks on more expensive programs like Discovery and New Frontiers or just accept a smaller pipeline of new instruments and PIs.
The lander on this mission, NOVA-C, is methalox. So not only do they have to land on the moon, they also have to manage boil-off issues. Somebody may correct me on this, but I think it will be the longest duration flight of any cryogenic rocket stage.
NASA stored LCH4 on orbit with zero boiloff for four months on CRS-16. See Robotic Refueling Mission 3. LCH4 with boiloff for days is comparatively straightforward.
Centaurs have demonstrated LOX (and LH2) storage and firings after ~10 hours with about 2% loss. There are several types of diesel submarine that rely on multi-week storage of liquid oxygen for underwater operation and propulsion. One can buy LOX dewars online. The devil is always in the details, but multi-day storage of LOX is hardly witchcraft.
Again, NASA needs programs like this where it can prove out new technologies and subsystems and have some failures along the way. A new propulsion system like this could not be tried on a critical-path Mars lander mission, but it can be tried in this place-some-bets lunar lander program.
I don't see a commercial market for these landers. I think they will be dependent on NASA funding for a long time. If the landers perform well, then maybe the companies can win other NASA and Pentagon contracts.
There’s no doubt that these lunar lander missions would not exist without NASA. But these companies have non-NASA customers for these missions, and they are involved in other NASA and non-NASA work besides these lunar lander missions. Astrobotic’s Griffin 1 has 14 non-NASA payloads, and Masten terrestrial flight testing continues under Astrobotic. DSS works Orion avionics. Draper goes back to Apollo. Firefly has three SLV launches under their belt. Intuitive Machines has the same deep pocket backer as Axiom, IM-1 has six non-NASA payloads, and they do terrestrial drones. Tyvak does smallsats. Big firms like LockMart, SpaceX, Sierra Nevada, and Blue Origin have also qualified for awards under this program. Etc.
That is true. The way the contracts are written could limit NASA's risk if the company fails to deliver. But just about every contract, including a cost-plus contract, is also going to have performance-based awards. Look at it like this: suppose NASA signs a contract with Lockheed-Martin that is cost-plus, but it reaches orbit and then fails. Although NASA paid the money to build it (and any overruns that occurred), NASA will not pay any award fees they would have had to pay for successful operations. Not exactly potato/potahto when comparing it to a services-based contract, but not completely different either.
No, they’re very different.
Award fees on a cost-plus contract are a few or handful of percent of the total contract value, whatever that total value turns out to be. The government is on the hook to pay for 95-97% of the work done on a cost-plus contract, even if that work ends up costing 200% (or whatever overrun) of the original contract value. The government may be able to withhold 3-5% of the original contract value for poor performance, but costs will still double (or whatever) overall under cost-plus
Fixed-price contracts are milestone-driven, and unmet milestones on a fixed-price contract represent large chunks of the total contract. If the contractor fails to complete, say, 50% of the milestones (because of overruns, technical issues, etc.) on a fixed-price contract, then the government only loses ~50% of the original contract value and can reapply that funding to a different performer. There is no overrun, and unused budget can be reapplied.
The type of contract used is really a function of whether the government is willing to risk cost increases (cost-plus) or project loss (fixed-cost). The way to manage those contract risks programmatically is via multiple performers. If the government relies on only one performer, the government is held over a barrel on cost (cost-plus) or project (fixed-cost). But with multiple performers, the government can credibly threaten to terminate a contract of either type when there is a big overrun (cost-plus) or milestone failures (fixed-price) and still have a successful program overall. Multiple performers hold cost-plus performers’ feet to the fire on cost (in ways that award fees never have), and they provide a off-ramp for the program when fixed-price performers are failing.
(The type of contract is also a function of what performers can be attracted to do the work. No one will develop a megawatt fusion drive under fixed-price, to use an extreme example.)
That said, I've heard that NASA has been providing a lot of under the table support to these projects to reduce the risk of them failing. For instance, having experienced NASA personnel help the contractor's inexperienced personnel. So although a services-based contract is supposed to put more risk burden on the contractor, NASA may have recognized that letting the contractor fail is not a good option.
The reality here is that NASA provides a lot of support to contractors of all types and sizes on NASA missions because contractors can’t afford the carrying costs of all the domain expertise that NASA maintains. You’re not going to find, say, lunar terrain and regolith experts in most aerospace firms. NASA has to provide that expertise. LockMart couldn’t pull off Mars landers without specialized JPL expertise, and JPL has found out that it can’t do an MSR lander without LockMart’s (or some other firm’s) engineering experience and cost structure. It’s why every Explorer, Discovery, and New Frontiers proposal is a partnership between a government lab and an aerospace contractor, with a (usually university) PI riding herd on science prioritization and instrument teams.