Starship Launch 13 Exposed Its Fatal Achilles Heel

Delays are more than an annoyance - they are potentially mission-ending.

Photo by Nicolas Thomas on Unsplash

I have already covered how Starship’s flight test 13 wasn’t the huge leap forward SpaceX needed. But this launch demonstrated something crucial that the media has barely even noticed, let alone discussed: a major flaw with Starship’s promised future. A fundamental problem so catastrophic, we will look back and wonder why anyone thought this launch vehicle would ever work. It isn’t just me who has spotted this, though; industry insiders, including ex-Space Shuttle engineers, have too. The issue at hand is boil-off and delayed launches, and while that might sound trivial, the impact is devastating.

Delays

Starship is no stranger to delays. An entire upper stage exploded before it even reached the launch pad, the FAA have grounded it numerous times for months on end, and SpaceX has had to conduct considerable design overhauls to fix Starship’s stark underperformance, delaying launches by over six months, just to name a few. Quite frankly, Starship makes England’s woeful train system look like the pinnacle of punctuality.

Launch 13 demonstrated that, even when all of these critical issues have been dealt with and the rocket is ready and waiting to go, major delays can still happen. The flight was originally scheduled for July 16th, but the launch was aborted immediately after ignition, as four of the Super Heavy Booster’s engines failed to ignite, thanks to LOX (liquid oxygen) fuelling issues, leaving the rocket with insufficient lift. SpaceX had to replace these engines, before testing them on July 22nd, and then attempted another launch on July 23rd. But, it was too cloudy that day, so SpaceX delayed the launch by another 24 hours. To be clear, Starship can launch through clouds, but SpaceX wanted to be able to observe the rocket during the launch. So, while this incident caused an actual delay of eight days, we can ignore this 24-hour delay, as it wasn’t caused by technical issues, meaning four failed engines caused a week-long launch delay!

Why is this a problem?

Well, for pretty much the entire history of space launches, delays like these haven’t been an issue at all. You see, most missions rely on a single launch, and the few that have relied on multiple launches (such as the ISS) are explicitly designed to cope with extensive launch delays.

But this is simply not the case for Starship.

To reach the Moon or Mars (which is Starship’s intended purpose), Starship needs to be entirely refuelled in LEO (Low Earth Orbit), and thanks to the kind of propellant Starship uses, such refuelling is a frantic race against time, making delays like this insanely costly and potentially mission-ending.

Let me explain why.

Mission Profile

In order to land on the Moon or Mars, SpaceX plans to use three different variants of Starship. The first to launch is the depot variant. This will be placed into LEO and will be used to store propellant in orbit. Then, there is the tanker variety, which will launch into orbit, dock with the depot Starship, transfer propellant to it, and then land back on Earth. Starship supposedly has a 100-ton payload to LEO, meaning the tanker variety can only transfer 100 tons of propellant per launch to the depot. This is a problem, as our final variant, the Moon/Mars-bound Starship, has a propellant capacity of 1,600 tons and requires its tanks to be fully topped up to make the journey.

So, you might think that such a mission would require at least 17 launches (i.e., 16 tanker launches and one outbound Starship). Indeed, on the rare occasion SpaceX addresses this refuelling process, they hint that it will require a little over a dozen launches.

But, this is neglecting one major factor: boil-off.

Boil-Off

Ultimately, the propellant Starship uses is a mixture of cryogenic liquid oxygen (LOX) and liquid methane. Both are normally gases, but storing them as a gas takes up far too much room. However, when placed under immense pressure and cryogenically chilled to -183°C (-297.3°F) for oxygen and -162 °C (-259.6°F) for methane, they turn into liquids, which makes them energy dense enough to be used as rocket fuel.

It’s no mystery as to why Starship uses cryogenic LOX and methane — together, they are an incredibly efficient and powerful rocket fuel mixture. There is just one problem, though: they aren’t exactly what you would call shelf-stable.

If the fuels’ temperature creeps above that cryogenic mark, then they start to turn back into gas (boil) and increase the already extreme pressure in their tanks. This is called boil-off, and to prevent these tanks from reaching unsafe pressures and risking catastrophic explosive failure, fuel has to be vented out. In other words, over time, these cryogenic fuels evaporate away.

This is where I think Musk has made a rather rookie mistake. He seems to believe that space is cold, and so it makes sense to store cryogenic fuels and AI data centres there. But that isn’t true. In space, you are exposed to the raw, unfiltered radiation of the Sun.

Take the ISS; it operates in the same orbit as the Starship depot, and its exterior facing the Sun routinely reaches temperatures of 121°C (250°F), which is hotter than a slow cooker (crockpot) turned all the way up.

So, Musk is effectively taking this highly temperature-sensitive, and arguably most mission-critical, component, and storing it at a temperature hot enough to make a hotpot in record time.

Now, yes, insulation can slow down boil-off, and the square cube law is in Starship’s favour here. But, even if SpaceX stuffed the Starship depot with all the insulation in the world, it wouldn’t stop this loss of fuel. That is why, despite trying to find a way to make zero-boil-off cryogenic fuel tanks for a while now, NASA has found it simply impossible to store cryogenic fuels in space for more than a few days.

So, can you see why Starship having even small launch delays can quickly become a catastrophic problem?

What This Means

Boil-off alone has a disastrous impact on a Starship lunar/Mars mission. In a previous article, I very optimistically assumed the Starship depot will only have a 1% per day boil-off rate, that Starship could transfer 100 tons of cryogenic propellant per refuelling mission (while still being a fully reusable rocket), and that these refuelling missions would happen once a week. I calculated that it would take over two years and 110 refuelling missions for the depot to reach 89% of its required capacity, as at that point it would reach equilibrium, where the fuel being lost to boil-off per week equals the amount it is being refuelled per week.

Those were only ever rough numbers. They were supposed to demonstrate the scope of the problem, not lay it out in detail.

But aerospace industry insiders, including one who worked on the Space Shuttle program, reached out to me personally with far more solid figures, which also included the effect of launch delays.

Like me, they found that Starship would experience at least a 1% per day boil-off rate but noted that it could easily be 2% or higher. They believed I was being far too generous with a 100-ton payload to LEO, and if you have been reading my articles for a while, you will know I completely agree with them. So they instead assumed a 50-ton payload for the tanker Starships, which is still wildly optimistic considering the current Starship struggles to complete a suborbital flight with less than 70% of that payload. They also assumed that all four of Starship’s launch pads were operational and that there were enough Starship and Super Heavy boosters built for them to all run at maximum launch capacity, equivalent to a launch every 1.75 days (one launch per pad per week).

With their more accurate assumptions, they found that it would take just over two and a half months (80.5 days) and 46 launches to fully refuel the Starship depot. That is nearly three times the number of launches SpaceX has “suggested” it would take.

But, as Flight 13 showed, delays happen, and often. The chances of SpaceX perfectly executing 47 back-to-back launches (46 refuelling flights plus one outgoing Starship) at this insanely fast pace is basically impossible.

So, they assumed a launch cadence rate 25% below their theoretical limit of one launch every 1.75 days. That is equivalent to one of the four launch pads having a launch delay similar to Flight 13 each week. This is still a very optimistically high rate, not just because Starship has experienced repeated technical delays, such as with Flight 13, but also because factors like weather can cause delays, and nowhere in the world has more than 80 straight days of perfect launch weather.

With this reduced launch rate, the rate of boil-off increases, as there is more time between refuelling, meaning more refuelling missions are needed. With this 25% launch cadence reduction, they found that it would take four months (119 days) and a whopping 68 refuelling missions to fill up the Starship depot.

I feel I need to put that figure into perspective.

If we accept these calculations, then a Starship lunar mission will require at least 69 launches (68 tanker missions plus one outgoing Starship) to carry 50 tons of payload to the Moon. Now, I have calculated that a realistic launch cost for Starship is $70 million (read more here), putting the launch costs of this mission alone at $4.83 billion. That is equivalent to $96,600 per kg to the Moon.

By comparison, NASA’s SLS Block 1B can take 38 tons to the Moon in a single launch, at a cost of roughly $2.5 billion per launch, which works out to $65,800 per kg to the Moon.

So, hang on. NASA’s “super expensive” rocket can take a payload to the Moon at a 30% lower cost per kg than Musk’s rocket and carry nearly as much payload in a single launch to the Moon as Starship can take with nearly 70 launches! What’s more, we know SLS works. It has already carried astronauts to lunar orbit and back, meanwhile Starship is nowhere near being a fully reusable rocket capable of taking 50 tons to LEO, let alone one that can do it over and over again for four months.

Again, to be clear, even the big wigs at NASA are struggling with this conundrum. Jared Isaacman, SpaceX investor, private SpaceX customer, Musk glazer and now NASA administrator, has publicly stated he doesn’t know how many refuelling missions it will take to refuel the Starship depot. In truth, no one can estimate the number, because you can’t know how badly the tanker launches will be delayed. Will there be technical delays? Or weather delays? Without a crystal ball, you won’t know, and therefore, won’t know how many refuelling missions are needed, or even when the very narrow window will open for the outbound Starship to launch, enter orbit, refuel, and then jet off to its destination.

The Operational Issues

Somehow, NASA overlooked these glaring issues when it awarded SpaceX the multi-billion-dollar HLS contract to use a modified Starship as a lunar lander for its Artemis IV and beyond missions. You see, astronauts won’t take off in Starship. Instead, they will take off in an Orion spacecraft on top of a NASA SLS, and then rendezvous with the HLS Starship in lunar orbit, before using the HLS to land. But it takes months to prepare such an SLS launch, weeks to ensure the rocket and spacecraft are ready to launch, and days for Orion to travel to the Moon.

In other words, NASA needs to know months in advance the exact date the HLS Starship will be waiting for them in lunar orbit. If they launch too early, then they will have no way of reaching the lunar surface and will have to simply return home, rendering the entire multi-billion-dollar mission useless. If they launch too late, then there will be an HLS Starship waiting for them in lunar orbit, but all of its fuel will have boiled off, making it a useless lunar lander, which again makes it impossible for them to reach the lunar surface. This is another scenario where they would be forced to return home empty-handed.

For Artemis IV and beyond to be a success, NASA has to somehow prep and launch SLS to rendezvous with Starship HLS at exactly the right time. But how can they do that when not even SpaceX themselves know when Starship will get there?

I will put it bluntly: using Starship in this way is utterly moronic. No one who understands even basic engineering or risk mitigation would approve it for this purpose.

But, oh wait, it gets worse! This all assumes that the Starship depot doesn’t explode during refuelling, like Starships have while being refuelled on Earth, or becomes inoperable thanks to, say, I don’t know, cryogenic fuel putting too much temperature stress on critical parts and breaking them? There is a reason orbital cryogenic fuel transfer has never been attempted: it is incredibly dangerous and has the potential to be wildly unreliable.

This is all amplified by the number of times refuelling has to happen. If there is even a small chance of refuelling causing a mission-ending explosion or critical damage that renders the depot unusable, the cumulative likelihood across 60+ refuelling missions can quickly add up to become more likely than not (which I have discussed before; read more here). So, there is a notable chance that a Starship lunar mission will end in a catastrophic failure in LEO rather than meeting NASA astronauts around the Moon.

Conclusion

For the life of me, I can’t figure out why NASA chose this utterly foolish mission architecture when the SLS is already a proven launch vehicle and can most likely complete Artemis missions significantly cheaper than Starship with far less risk and pollution. I also can’t believe that the media at large isn’t openly criticising SpaceX for thinking this is a good idea. I know Starship looks like a big and shiny futuristic thing, and Musk uses all the relevant buzzwords, but it isn’t hard to apply a little critical thinking to his propaganda. Truthfully, all it takes is a little curiosity, like looking at test flight 13 and going, “I wonder how such a delay would impact a Starship mission to the Moon or Mars.” Because it is basic questions like these which make Starship look like more of a fools’ errand than a revolution.

---

Thanks for reading! Everything expressed in this article is my opinion, and should not be taken as financial advice or accusations. Don’t forget to check out my YouTubechannel for more from me, or Subscribe. Oh, and don’t forget to hit the share button below to get the word out!