Starship II: A Reprise
Et tu, Starship?
I wrote about Starship’s first orbital flight test just a week and a half ago, and that was recieved very well. I wrote from the heart — day of launch, experiences fresh and with a feeling of hope after watching what was a spectacular launch (at least, visually).
But as fate would have it, we always learn more as time moves on. I knew from day one that the flight was by no means a technical success. I concluded that first post with the following:
If you had told anyone just two years ago that SLS would beat Starship, fly perfectly, and become the component of the Artemis program that contributed to delays the least, you’d have been laughed out of the room. And yet, here we are — of America’s two new and flown super-heavy launch vehicles, one has sent a deep-space crew-capable spacecraft to the Moon and back, while the other totaled its launchpad, struggled and failed the whole way up, and finally exploded not even a third of the way to space — nevermind orbit — carrying nothing more than the hopes and dreams of the Artemis generation.
That was an understatement.
As time has gone on, I’ve reflected a lot on how Starship’s first flight has been recieved and what actually happened. To put it bluntly, it’s not good. Compared to Artemis I? Well, the Artemis hype train has been going ever since that launch. Yet it feels like most of the people I know want to forget Starship’s test flight ever happened. Or if they don’t, they want to remember the cool launch, the awesome photos we got, but not the implications of what actually happened.
When I showed up at the press site for the Falcon Heavy launch I’m currently covering, the #1 question people asked when they saw someone roll up is “how many cameras did you lose in Texas?”. It really felt like nobody had anything good to say for the launch, beyond just the experience of being there. I said much of the same.
We all shared the same feeling of anxiety for what would come next for Starship, while not a mile behind us, SLS’s Mobile Launcher-1 was lit up and clean, seemingly more than ready to support another launch when the time came. While at pad 39A for Falcon Heavy, I looked at the Starship pad under construction there in a much different light, as if it had lost its shine. The Starship pad wasn’t the pillar of the grand future to come anymore. Nor was it the towering behemoth it appeared to be at Starbase. No, it was now an overbuilt grave — condemned before it could even be finished, a reminder that the arrogance that enabled the flight at Starbase would no longer fly. Not here.
All that a look over to pad B down the road revealed was its three towers, each standing much taller than the Starship pad, invariably blinking as they stood ready to protect another super-heavy rocket. They made the Starship pad feel small, and really, so did everything else. It may have been the biggest and baddest in Texas, but Florida’s established spaceport brought some much needed context to the table.
So, what changed? Or what did we learn that wasn’t there on day one?
I think it is first most critical to set expectations for what the Starship system’s contractual obligations to NASA, and by extension the U.S. taxpayer, are. Context is critical, and if you’ll bare with me, there’s a lot needed here to understand the scale of the problem we are dealing with.
Starship has to land on the moon three times under the NextSTEP-2 contract. The first two, under Appendix H Option A, is the “Human Landing System” contract that was awarded to SpaceX for the landing currently scheduled for Artemis III, as well as an uncrewed landing demo before that mission. The second, under App. H Option B, is a demonstration of an evolved “Sustaining HLS” scheduled for Artemis IV.
We’ll get back to NextSTEP and sustainable landers in a moment, but the most important takeaway right now is three landings.
Each landing requires several launches to carry out. In my previous article, I named “at least sixteen times” as the number of landings per mission. That is the number that was revealed during a certain company’s lawsuit over the initial App. H contract award, in which they insisted that the number (released by the government) made SpaceX’s program “immensely complex and high risk” in some flashy infographics. A lot of folks openly joked about it then, but looking back, maybe they were right.
There’s been some doubt expressed about those figures since then, so to shed some light: it’s only worse now. I have to be careful with specifics as to not put any sources at undue risk for what is a personal blog post, but know that there are three general concepts of operations being studied.
- Concept 1 involves 16 tankers, a fuel depot, and an HLS. Both the depot and HLS are completely filled in low earth orbit, 8 tankers each, and then the depot fills the HLS in a higher orbit.
- Concept 2 also involves 16 tankers, a depot, and an HLS. The depot is filled twice, with HLS only being filled by the depot. This presumably helps with boiloff.
- Concept 3 involves 16 tankers, two depots, and an HLS. A depot is placed in LEO and a higher-energy orbit, and both are filled accordingly. HLS fills from the LEO depot first, then the high-energy depot, then continues to the Moon.
You will have to take my word on the numbers here — while these are direct from program, I can only say so much for a personal blog post.
Throughout every mission type, the ships must contend with boiloff — on the order of about a ton per day, while each tanker only carries about 150 tons each. That doesn’t sound like much, but if you add up the losses, every 5 months of time-in-space per ship is a launch wasted. Factor in enough time to get 18 ships off the ground, and if you can pull that off in one year — that’s still at least two launches effectively wasted. We will assume 18 launches accounts for expected losses.
To get to the point where Starship can even start doing landing demos, there are 3 missions that still must be completed — the orbital flight test, a propellant-transfer test, and a long duration flight test. For the sake of illustration we will assume all launches go perfectly — akin to what SpaceX internally likes to call a “green light schedule.” (if all the traffic lights were green, how fast could you get there?)
This puts us at 4 launches for the initial demos, and 54 launches for the three landings required. That means that at minimum, for Starship to complete its contractual obligations by Artemis IV, the rocket must fly fifty-eight times with zero failures by 2028. That’s not zero failures in the “nothing blew up” sense, but in the sense that every test must go flawlessly — redesigns and lessons learned must fit within the schedule. The clock is ticking, and one thing’s for sure, those lights sure aren’t all green.
In summary, for SpaceX to meet its contractual obligations, we need 4 Apollo programs’ worth of super-heavy lift out the door by, well let’s be kind, the end of the decade.
Looking at the schedule of Artemis III alone here, have a look at the “Orbital Launch Test” milestone. SpaceX said they would have that milestone — literally step 1 — complete in Q2 of fiscal year 2022.
We are now in Q3 of FY23, 15 months behind on step one and counting. How long will it take for even one Starship to complete the test?
It really doesn’t get any easier from here, unfortunately. Musk did an after-launch report on Twitter Spaces, where he admitted a lot about the flight that really sent home just how bad it all actually went. I want you to keep the baseline expectations we just went over in mind as I go through just some of the faults he mentioned (with the assistance of Michael Sheetz’ reporting, linked here).
- Three engines did not light before liftoff due to faults detected at startup. (er, “chose not to start,” in Elon’s words, as if a human decided.)
- Due to the engine failures at startup, the rocket experienced a “powerslide” and drifted several body diameters away from the center of the launchpad. Had the failures happened in a different location on the booster, it would have drifted directly into the tower.
- Communications with another engine were completely lost at T+ 27 seconds, and the heatshields of engines 17 through 20 were knocked out due to an explosion around the same time.
- Thrust vector control over the vehicle was completely lost at T+ 85 seconds — Starship was essentially uncontrolled for another two and a half minutes after.
- At least 7 engines completely failed throughout the flight, with Musk stating that the “rock tornado” the rocket formed at liftoff did not damage the engines or heatshields in any material way.
- Musk stated the damage to the launchpad was “certainly not expected” with the complete destruction of the concrete under the launch mount being “actually quite small” in terms of pad damage.
- The rocket’s automated flight termination system took 40 seconds to terminate the flight from the command being sent. In other words, even though the vehicle was in an unsafe-to-public state where it should have been terminated instantly, it kept flying for nearly half a minute longer than it should have. Musk says the FTS must be re-qualified, adding that “it took way too long to rupture the tanks.”
- SpaceX expects to spend $2 billion on Starship this year.
- As a cherry on top, Musk’s expectation for the next flight is pad readiness in “6 to 8 weeks,” with the goal being “making it to staging and hopefully succeed(ing).” He also re-iterated that he does not believe Starship HLS will be the long-lead item for Artemis III.
“We will be the first thing to really be ready.”
I have to ask, do you see this system successfully launching another 40 times before Artemis III’s SLS and Orion are ready? You know, the third SLS that’s well into production at Michoud, and the Orion whose pressure vessel is currently being outfitted at KSC right now, ahead of a flight just under three years from now?
Since he pulled from my last post, I’ll pull from his — Ben Kellie, CEO of The Launch Company and a SpaceX vet himself, wrote the following:
Worse, their newest contract with NASA stipulates landing humans on the Moon by 2024, which is next year. Keep in mind, that Starship has nothing inside of it except the tanks, valves, and wires needed to make it fly. There is no payload bay. There are no seats. There is no life support system. We don’t know if the re-entry tiles will work. Honestly, if this was any company other than SpaceX I would declare them toast. Companies have folded over much less, especially in this industry.
SpaceX has three years to get Starship orbital, figure out how to not expend the pad in doing so, solve cryogenic refueling in space, prove re-entry and reuse of ships, prove the long-duration capability of the ships, develop an entire lander and all of the extra systems Kellie mentions, execute a successful test campaign with 18 launches, and then do it all again to have HLS ready for Artemis III’s arrival, if they are to meet Elon’s claim that they’ll be the first ready.
I told you SpaceX likes their green-light schedules.
I think Kellie’s article is a very grounded take on the whole situation. While he shows that yes, even in his own experience, SpaceX can and has done great things, this test is far more “crappy” than “scrappy.” And while the program may some day succeed, I have a lot of concerns for the now very real implications of how this test went in terms of the Artemis program.
Where SpaceX had room to fail and iterate with landing their boosters on barges, when it comes to Artemis and Starship, play time is over. Starship, and in turn Artemis, explicitly rely on the technologies SpaceX is testing, and crappy tests only lead to further delays for a program that has to keep a schedule. Remember that OIG article I mentioned earlier? Even in 2021, it was noted that “as a contingency plan, [NASA] Exploration Systems Development officials raised the possibility of using Artemis III to complete an additional fly-by of the Moon rather than a lunar landing mission should necessary systems not be available in time.”
To put it to you blunt, Artemis III will more than likely not land on the Moon. While I believe NASA has their own reasons they can’t make that official yet, I think as an industry we’ve at least got to start accepting that there is a problem, that problem now has real and significant implications for the Artemis program, and solutions must be found. NASA knows this too, which is why in 2022, the agency put out the Sustaining Lunar Development contract, under NextSTEP-2 Appendix P. (I told you we’d get back to NextSTEP.)
This is actually something I wrote about for Space Scout almost the day it happened, and if you want to read my day-of reporting without my opinions, you can do so here: NASA moves to procure a second crewed lunar lander — Space Scout.
Otherwise, here’s the elevator pitch: SLD is about acquiring a second lander to be developed alongside Starship. That lander should be ready by 2028, anyone but SpaceX can bid, and that lander should be focused on being a sustainable solution for transportation, not a one-off. A risk reduction study, App. N, was completed with Blue Origin, Dynetics, Lockheed, and Northrop all showing interest.
The thinking behind SLD is sound. At best, if Starship is a success, we have two redundant ways to reach the Moon’s surface, one of them with significant capabilities. However, if Starship can’t be trusted, SLD offers a second chance and decouples the Artemis program’s success from Starship’s.
So who do we turn to for a second lander? Well, we know of one company that’s incredibly interested in putting humans on the Moon. They practically went down kicking and screaming when SpaceX won their current contract. They’ve been quiet, but they’ve assembled a world-class team and have access to much of the same resources as SpaceX.
Yes. Blue Origin’s National Team.
Now perhaps I am biased. I’ve seen inside Blue’s facilities, and while I can’t tell you what I did see, I can tell you everyone who shares that experience with me (including NASA’s exploration officials!) share much of the same final takeaway — that Blue’s doing pretty damn well.
But I’ve also paid close attention to the company’s efforts to secure a lander contract over the years. A lot of folks will look back to the phrase “immensely complex and high risk” with humor and contempt, practically a meme from Blue’s lawsuits against the government for choosing SpaceX.
I ask you now, were they wrong?
I am not going to pretend that a Blue Origin team would have performed perfectly. But perhaps it would have struck a finer balance, the line between crappy and scrappy Kellie talked about in his article. Blue has the nearly unchecked resources and power that SpaceX has, but so far, the company has been much smarter about it. Instead of going about HLS and SLD by itself, it assembled the National Team, today containing Lockheed, Draper, Boeing, Astrobotic, and Honeybee. The National Team’s lander wasn’t nearly as flashy as Starship, but it didn’t have to be. It fit the bill, it needed much less new development to be successful, and it had options where proposed solutions had risk.
(While we’re here, Dynetics’ lander deserves a shout too — I don’t know as much about it, but it brings its own simpler take to the table as well.)
Design is based on requirements. There’s no justification for designing something one bit “better” than the requirements dictate.
This is something that I find is difficult to explain to the aerospace community, but is very easy to explain to the industry. To anyone who hasn’t really dealt with these types of programs, a system as flashy and capable on paper as Starship seems incredible, especially at its bid price. But I find many of the industry veterans I’ve talked to take a much more wary view towards such concepts, especially when they far exceed what was asked.
Sometimes, over-delivering is great. But one should be incredibly wary of confusing “nice to have” with “program requirement.” I feel many have missed this in their coverage of the Starship program, where Starship is often only discussed in context of its incredible claimed capabilities rather than its concerning real status. All of these incredible things that Starship aims to be capable of become liabilities as they are embedded into the design, resulting in a stackup of things that really were never asked for in the original contract now holding up the entire program.
This truly is why Starship’s “immensely complex” concept of operations is a risk in the first place. Do not get me wrong — it would be incredible to have two dozen super-heavy lift rockets flying every year in support of the Artemis program. You know how much I loved the actual in-person experience of the first flight. But pragmatically, it’s time to acknowledge that things can in fact be ahead of their time today and not just in hindsight, and that’s what we’re seeing with Starship. While I still have hope for the rocket as a LEO mass mover, and I have hope for SpaceX as a whole, it’s time to genuinely reconsider whether we can count on it to carry our first crew back to the Moon in fifty years.
It is truly rare that there is this much momentum in the spaceflight industry, and we must not take it for granted. The money is on the table to return to the Moon now. It’s time to make the responsible choice, pick someone new for Sustaining Lunar Development, and start reviewing the timelines for Artemis III and IV.
Quite honestly, this is already in progress. NASA is critically aware of these risks, and it’s why the SLD contract exists even after HLS was significantly underfunded. Rumor has it that Starship’s situation within the Artemis program is already in hot water, but we will have to wait and see. No matter what, if we are to be competent in our return to the Moon, tough decisions must be made.
I leave you with a few lines from Gene Kranz.
Mission Control was behind in virtually every area, and the flight and test procedures changed daily. Nothing we did had any shelf life. Not one of us stood up and said, “Dammit, stop!”
I don’t know what Thompson’s committee will find as the cause, but I know what I find. We are the cause! We were not ready! We did not do our job. We were rolling the dice, hoping that things would come together by launch day, when in our hearts we knew it would take a miracle. We were pushing the schedule and betting that the Cape would slip before we did.
From this day forward, Flight Control will be known by two words: “Tough” and “Competent”. Tough means we are forever accountable for what we do or what we fail to do. We will never again compromise our responsibilities. Every time we walk into Mission Control we will know what we stand for. Competent means we will never take anything for granted. We will never be found short in our knowledge and in our skills.
P.S. — This came after I wrote a majority of this post, but the FAA is now being sued by a bunch of environmental groups over it allowing SpaceX to proceed with operations at Starbase without an environnmental impact statement. The case has a strong claim in that the FAA failed to accurately assess alternatives for development at Starbase, including entirely neglecting to assess use of the (now) existing pad at 39A or a notional 49 instead of Starbase, in turn violating NEPA. I am not an environmental law expert, but that doesn’t look good.
P.P.S — Covering Falcon Heavy was a blast, and honestly left me with some hope for what SpaceX can still do, even though the vehicle itself was quite the “frankenrocket” as Max Evans called it. A shoutout to both Dereks (Newsome and Wise), Kyle, Madison, Sean, Max, and everyone else who made the detour to ViaSat-3 Americas an incredible journey. I hope to see you all again soon.
I’m a spaceflight reporter, photojournalist, and managing editor of Space Scout. I’ve attended over 20 orbital launches so far. If you’d like to keep up with me, you can follow me here: https://twitter.com/lavie154