I think the chances of the SpaceX mission around the Moon going ahead on schedule in 2018 is tiny. But on the remote chance it does, I would not fly on that mission, if you paid me a billion dollars. The problem is that they have to rely on many innovations working just right that are hardly tested. Their current Dragon spacecraft is only rated for re-entry from LEO (Low Earth Orbit) and not for the much faster re-entry from a trip around the Moon. That’s why they plan to use the larger Dragon 2 which has its first flight in 2018. If we go by Elon Musk’s suggested timeline, this flight around the Moon would probably be its second crewed flight, and third flight ever.

They need a new rocket to launch it as well. They plan to use the block 5 Falcon 9 “full thrust” to launch the first unmanned Dragon 2 to the ISS as well as their first crewed flight to the ISS. But this version of the Falcon 9, which hasn’t flown yet, is not going to be powerful enough to launch a Dragon 2 around the Moon. So they have to depend on the Falcon Heavy to launch their second crewed mission, which is another rocket that they hope will fly for the first time in late 2017.

So, going by his sketched out timeline, it’s probably also going to be the first flight of a Dragon 2 on a Falcon Heavy, and one of the early Falcon Heavy flights, at most one year after its first launch And they send it around the Moon! With no possibility to abort once they leave low Earth orbit.

For SpaceX fans, just to say, I'm quite critical in this article but it has an up beat ending. It's not my aim to discourage SpaceX in their human spaceflight ambitions :). Rather the aim is to encourage them to do it safely :). They can do that while at the same time continuing to make a profit with every launch. That's the beauty of their approach.

SpaceX Dragon 2 which will fly for the first time in 2018. This spaceship is a major upgrade from the existing Dragon with a thick enough aeroshell to handle the much higher speed of re-entry into the Earth’s atmosphere from the Moon with plenty of margin to spare

The existing Dragon is over-engineered, so its heatshield could withstand a re-entry from the Moon, but it's not rated for that.

Their unmanned rockets have blown up once each year for the last two years (Sept 1, 2016 and June 28, 2015). It was for a different reason reason each time, and they fixed the problems. If one of the early flights of the Falcon Heavy blows up, then that will delay things a lot and surely lead to questions of passenger safety for this mission.

They also load the fuel after the astronauts are on board - an unusual procedure never done before with passengers. This is a concern that was raised by Tom Stafford, command pilot for Apollo 10, and retired Air Force general, and other veterans of NASA's Gemini, Apollo and space shuttle programs. The rocket can explode while the fuel is being loaded, as happened to one of their rockets last year (the explosion on Sept 1, 2016), where the rocket exploded as a result of a tank buckling as the fuel was loaded, creating voids where liquid oxygen could pool.

Tom Stafford, former commander of Apollo 10, who has concerns about SpaceX's policy of loading the fuel after the astronauts.

In his letter, written on December 2015, summarizing the conclusions of NASA International Space Advisory Committee he wrote:

"There is a unanimous, and strong, feeling by the committee that scheduling the crew to be on board the Dragon spacecraft prior to loading oxidizer into the rocket is contrary to booster safety criteria that has been in place for over 50 years, both in this comttry and internationally. Historically, neither the crew nor any other personnel have ever been allowed in or near the booster dming fueling. Only after the booster is fully fueled and stabilized are the few essential people allowed near it."

He repeated this in a conference call in 2016 after the SpaceX accident, saying

"It was unanimous ... Everybody there, and particularly the people who had experience over the years, said nobody is ever near the pad when they fuel a booster,"

If the fuel is loaded before the passengers, you have eliminated one major potential risk. However, this is an essential part of their rocket design because they use fuel that has to be kept very cold. The oxygen is cooled to -207 °C before loading it onto the rocket, by passing it through a nitrogen bath in a partial vacuum. This increases its density by 8%. (See details on Spaceflight 101). They can’t keep the fuel this cold for long after it is loaded. This makes it impossible to load the fuel hours in advance before the crew.

SpaceX have a launch escape system that should let passengers escape if the rocket explodes on the launchpad with them on board. It is triggered automatically in the case of an explosion like that - but it is itself one more thing that could go wrong, and so far has never been tested with humans on board, just once with dummies.

As Doug Messier put it:

"There was talk after the Falcon 9 accident about whether the Dragon cargo ship could have been saved if it had the abort system planned for human-rated Dragon V2. The general consensus was that the capsule could have been rocketed away and parachuted to safety. However, the discussion misses a key point."

"An abort system is like an inflatable aircraft slide: an essential safety feature that you never, ever want to actually use. It’s an option of last resort when all else has failed and there’s no other way to save the lives of the crew."

If the first crewed flight to the ISS blows up, even if the crew survive in the escape capsule, that would surely again lead to questions of passenger safety and delay things, as launch escape systems are meant as a final backup and not expected to be triggered.

But if nothing happens during the unmanned missions, it doesn't show that the rocket is safe. As Richard Feynman wrote about the Challenger disaster:

"The argument that the same risk was flown before without failure is often accepted as an argument for the safety of accepting it again. Because of this, obvious weaknesses are accepted again and again, sometimes without a sufficiently serious attempt to remedy them..."

In this case, loading the fuel after the passengers is potentially a risky approach, and perhaps his remark is therefore relevant to it. Even if they get several launches without incident using this procedure, it doesn't prove that it is now safe. It’s surely something that they need to keep a careful eye on.

They are also depending on a life support system working for a week in space which has only ever been tested for that long on the ground. There have been many problems with the ISS life support in zero g that never showed up in ground testing. And there is no way to abort the mission back to Earth. This is the worst thing about it for me. If they have an Apollo 13 style failure of life support on the way out, then they will have to make do with whatever they have in their spaceship to try to fix it. And Apollo 13 had the lunar module as a “lifeboat” which they won’t have.

The environment control and life support is one of the most complex systems on a spacecraft anyway, and there have been concerns raised about their working methods, see Doug Messier’s review for a perspective on it. Also, the June 28, 2015 failure was due to a fault strut, which failed at 2,000 lbs force though rated to 10,000 lbs force. They have stopped using that strut and added additional quality controls.

So it’s not impossible that a vital part of their life support system fails in some way. If the carbon dioxide scrubbers stop working, for instance, the carbon dioxide build up would kill the crew on those time scales. Apollo 13 were able to use the attached lunar module as a “lifeboat”, but they will have no lifeboat.

There is no problem with life support going wrong in LEO, not minor things like the CO2 scrubbers not working, as you can just abort back to Earth within a few hours of noticing the problem. But on this mission it could easily be several days before you can return. Such a tragedy would unfold very slowly. And there would be absolutely nothing that anyone on Earth could do except give them advice.

Wayne Hale, former manager of the Space Shuttle Program is quoted by Space.com as saying:

"Even with today's technology, it's still an extraordinarily difficult, extraordinarily dangerous task to undertake, period — I don't care who you are,"

If they do this, I wouldn’t fly on it if you paid me a billion dollars for the ride. But I expect it will be delayed and delayed, as happens so often with SpaceX.

Also the FAA will need to approve it for safety, and it doesn’t seem very safe, to do this so soon, at least, not as they have outlined it. I don’t think it is “bluffing” but it is hugely optimistic, that they will be able to achieve so many ground breaking innovations so quickly, and that nothing will go wrong with any of them, and that they will all be completed on timetable and finally all be passed by the FAA as safe for flight by the end of 2018. They often claim that they will be able to do things many years before they actually do. For instance they claimed the Falcon Heavy would be ready some years ago (first they said 2011, then late 2013 to 2014) and it is still not ready.

The Trump administration are also planning a somewhat similarly risky mission - pushing back the first humans to use the Space Launch System (SLS) to the maiden flight of their rocket, with a journey around the Moon. See Gerard Black’s Human flight around the Moon: a worthy goal, but using the wrong vehicles. That’s not a fun ride by paying tourists, but an early flight with test pilots on board. These are people doing a job they know is risky and which they have chosen as a career, testing spaceships. However this also is risky, and it could be a personal disaster for the astronauts of course, and a huge set back for NASA’s human spaceflight plans if that first flight is a disaster and they all die.


It is easy to say "take the risk" until the first crash. It sounds romantic and exciting. If they succeed then there will be those who cheer them on and say it’s brilliant. But sooner or later, encouraged by that success, they send more such flights. Then one of them crashes, and then they will once more get just about everyone saying they pushed too quickly too soon. Here is a "future fake news" story to dramatize it, maybe help it seem more real, as something like this could actually happen.

It’s so different from an uncrewed spaceflight. With uncrewed spaceflight this “gamble on success” has worked for SpaceX. Though they have had at least one occasion where a single failed flight would have been the end of their company (in 2008, Elon Musk says if the fourth SpaceX test flight had failed like the first three, that would have meant bankruptcy and the end of both SpaceX and Tesla). 

So, SpaceX are used to taking huge and spectacular risks and it has worked for them for unmanned flight, though it could easily have been there downfall if they had been less lucky. But I don’t think it is the right way to do human spaceflight, to do spectacular risky flights like this when you don’t have to.

I happen to think that many who are now -saying "good for SpaceX, go for it, rush to the Moon in 2018" may well be saying "Why didn't you go more slowly and do more testing" if they do send humans around the Moon in 2018 and all seven die of carbon dioxide poisoning or whatever. Due to some flaw they'd have noticed and fixed if they'd done a 2 week flight in LEO first.

This is for human spaceflight especially. If they want to go on doing spectacular and risky unmanned firsts, then no problem. But with human spaceflight, I think it is especially important to reduce risks. It's risky enough as it is.

I say this as someone who is keen on humans returning to the Moon. But - humans safely on the Moon!


Timing is everything. This is the fast timeline SpaceX suggest:

  • Unmanned Dragon 2 with cargo to ISS end of 2017
  • First crewed Dragon 2 to ISS mid 2018
  • Crewed mission around the Moon end 2018

I think that is far too fast and unsafe, as many commentators have also said. Instead, why not be ambitious and aim for a superb safety record? Something like this:

  • Unmanned Dragon 2 with cargo to ISS end of 2017. Freeze the Dragon 2 design and the Falcon 9 full thrust block 5 (or whatever version they are using by then). Only fixes after that, this is the “release” no longer “betas” to use a software analogy.
  • Continued launch of Dragon 2 with cargo to ISS through 2018 - 2019. Each of these missions earns SpaceX good dollars.
  • First crewed mission to ISS 2019-2020
  • Continue with more crewed missions to ISS early 2020s
  • First tourist missions to LEO, with two week long stays, in early 2020s. This tests the life support system and the Dragon 2 generally for other issues that may occur in long space flights. They could end the mission, or start it, with a visit to the ISS, some arrangement to take some cargo to the ISS or to return experiments to Earth in exchange for the astronauts there giving the tourists a guided tour of the ISS.
  • Early 2020s: Freeze the Falcon Heavy and start doing unmanned flights to the ISS with Falcon Heavy + Dragon 2 - lots of cargo.
  • First tourist trip around the Moon, middle 2020s using Dragon 2 + Falcon Heavy. This coincides with the running down of the ISS and the beginning of the ESA village on the Moon which would be built at around that time.

Remember that SpaceX earn good dollars for every one of those missions. There is no empty hardware flying into space. I think this will put them in a much stronger position, lead to safer human flight and it gets further in the end, even if it seems a little slower, to start with. Of course it could easily be that this slower timetable is what would actually happen anyway.

It’s certainly possible to have reasonably safe human spaceflight. The Russian Soyuz flies several times a year, and hasn’t had a fatal accident since Soyuz 11 in 1971. Their launch escape system has only been used once, for mission 45 in 1983, which is also the only time to date that a launch escape vehicle has been used in any mission with crew on board. Everyone survived. They are now on mission 132. That makes it 111 missions in a row so far without a fatality and only minor issues since 1983. With a somewhat slower approach, but still progressive and exciting, hopefully SpaceX can achieve a safety record approaching that of the Soyuz.

Although Elon Musk so far takes many risks in his unmanned space ventures, he pays a great deal of attention to his safety record for cars (even though he personally drives racing cars and is no stranger to taking personal risks in cars himself). So I think he will in practice want to develop a spacecraft with the best safety record he can achieve, which will surely need a longer timeline with more prior flights with test pilots before the first passengers and with less ambitious first flights with passengers.

Also, as more companies develop the ability to go into space, wouldn't billionaires choose the company with the best safety record, so that they are most likely to survive the journey, even if it costs a little more? If SpaceX have a crash that kills all the passengers and Blue Origin or Virgin Galactica don't, or indeed, some other competitor such as Skylon in the mid 2020s or whenever, surely everyone would fly with those companies instead? At least surely they'd do that once they have the capability to get to LEO or wherever it is SpaceX offers to take you? Even if they cost a little more? SpaceX would then have its human spaceflight program relegated to offering flights for passengers to risky destinations that nobody else flies to.

For all those reasons, I wouldn't be surprised if in practice he continues with human spaceflight with a much higher safety consciousness than he seems to have in these advance announcements.

It's especially an issue for paying passengers. I'm sure the Apollo 8 crew would never have taken a paying passenger on board, no matter how much he or she paid. A first test flight of a risky procedure is just not the right place for a paying passenger. And actually the FAA has regulations about such things, which may perhaps force at least a somewhat slower timeline.

FAA regulation

If you build a new passenger planes there are many regulations that govern what you can and can't do to protect your passengers. It would make no difference if they were to sign a waiver to absolve you of your responsibilities for their safety. You still have to comply with those regulations. The regulations are nowhere near as strict for spaceflight as for airplanes. However they do still have many detailed requirements for life support, and other matters for spaceflight. 

These cover all flights including test flights, so they have requirements to protect the test pilots too. They can also request additional tests until they are sure the vehicle is safe to fly.

They have additional stronger requirements for paying passengers over those for test pilots. They can't be involved in an "envelope expansion flight". Quoting from their 2006 regulations

" The FAA notes that someone paying to fly, whether as a passenger or at the controls, is a space flight participant rather than an employee (page 75619

... "A space flight participant would not be allowed on an envelope expansion flight, that is, a space flight participant would not be allowed to be carried during first time flight testing in a different operational environment than what was tested (page 75623)."

So, what about a mission returning from the Moon with the first ever fast re-entry of the Dragon 2 into the Earth's atmosphere, and first test of life support for that length of time? Would that be an "envelope expansion flight"? If so, I'd expect the FAA to require additional flights.


First, how many previous flights would there be? The announcement from SpaceX says that they will do it after human crewed spaceflights to the ISS - so it’s not going to be their first manned flight. So, it would depend on how many of those flights they do first.

“Once operational Crew Dragon missions are underway for NASA, SpaceX will launch the private mission on a journey to circumnavigate the moon and return to Earth.”

SpaceX to Send Privately Crewed Dragon Spacecraft Beyond the Moon Next Year

But according to this report, it’s very soon, first crewed flight to the ISS in the second quarter of 2018, and then flight around the Moon later that year SpaceX to fly two tourists around Moon in 2018 - BBC News.

Elon Musk gave more details about his timeline to Universe Today:

“Musk replied that SpaceX hopes to launch the first uncrewed Dragon 2 test flight to the ISS by the end of this year on the firm’s Falcon 9 rocket – almost identical to the rocket that just launched on Feb. 19 from pad 39A.

“That would be followed by crewed launch to the ISS around mid-2018 and the private Moonshot by the end of 2018.

“’The timeline is we expect to launch a human rated Dragon 2 on Falcon 9 by the end of this year, but without people on board just for the test flight to the space station’ Musk told Universe Today.”

“’Then about 6 months later we would fly with a NASA crew to the space station on Falcon 9/Dragon 2.’”

“And then about 6 months after that, assuming the schedule holds by end of next year, is when we would do the lunar orbit mission.””

So in that time frame it seems that there is only one unmanned flight of the Dragon 2, then one crewed to the ISS then next one, around the Moon. He says that the mission is risky himself - as quoted on the BBC site, the passengers

"are entering this with their eyes open, knowing that there is some risk here. They're certainly not naive, and we'll do everything we can to minimise that risk, but it's not zero."

There are many issues with SpaceX for human spaceflight and with this mission particularly.

  • There will be no lifeboat - all except the first Apollo mission to the Moon (Apollo 8) had the lunar module attached which acted as a lifeboat in an emergency on the journey out and saved the crew of Apollo 13. It’s a multi-day voyage without a lifeboat.
  • There's the risk of explosion before they launch, with humans on board the spacecraft, as they load fuel after the passengers as described already

Then there are some concerns about their working methods. These criticisms are by Doug Messier, managing director of Parabolic Arc. in his article Are SpaceX’s 60 to 80 Hour Work Weeks Really Such a Good Idea? It's obviously one person's point of view on the matter but I think it deserves some attention.

  • They have had issues with quality control, including an explosion at launch due to a strut that passed quality control and shouldn't.
  • They have a history of unsafe practices such as engineers standing on components they shouldn't stand on.
  • They have engineers working 80 hours a week.
  • They continually change their design treating spaceship hardware much as you treat software.

Also he doesn't say which rocket he will use. But the Falcon 9 Full Thrust can transfer only 8.3 tons to GTO and 4.02 tons to Mars so it seems to be beyond its capability to send a Dragon 2 at 6.4 tons plus crew and supplies to TLI. I can't find payload figures for the Falcon 9 Full Thrust to TLI - if you know do say. The block 5 will have improved performance, and they don't say what this will do to the payloads, however it would require a major increase in performance to send the loaded Dragon 2 to TLI.

So, it seems likely that they will use the Falcon Heavy, which has a payload to TLI of 16 tons. If so, it will be one of the first few flights of the Falcon Heavy, which is yet to launch, and the first ever crewed launch on the Falcon Heavy.

This is all assuming they fly in a Dragon 2. He doesn't actually say he will use it for the lunar mission but they couldn't really use anything else. The re-entry speed from a spaceship returning from the Moon is much higher than for re-entry from LEO. For instance Apollo 8, a similar mission, had a re-entry speed of 24,696 mph or about 11.04 km / sec. By comparison typical LEO re-entry speeds are around 17,000 mph or about 7.6 km / sec

Apollo 8 reentry, December 27, 1968 photographed from a US air Force KC-135A flown at 40,000 ft altitude

The Dragon heat-shield is not rated for a return from the Moon. So they have to fly the Dragon 2, which hasn’t flown yet. It will have a heat shield rated for this, but it has a dry mass (without crew or payload) of 6.4 tons and its heat shield is of course not yet tested.

In a little more detail then on the Dragon, then Garrett Reisman in a statement before the House of Representatives in February 2015 wrote: 

"Designed in partnership with NASA and fabricated by SpaceX, Crew Dragon’s heat shield is made of PICA-X, a high-performance improvement on NASA’s original phenolic impregnated carbon ablator (PICA). PICA-X is designed to withstand heat rates from a lunar return mission, which far exceed the requirements for a low-Earth orbit mission" 

The SpaceX website says about the Dragon 2:

"Dragon v2 represents a leap forward in spacecraft technology from its Version 1 predecessor. Additional upgrades include a SpaceX-designed and built ISS docking adapter, impact attenuating landing legs, and a more advanced version of the PICA-X (Phenolic Impregnated Carbon Ablator-X) heat shield for improved durability and performance. Dragon v2’s robust thermal protection system is capable of lunar missions, in addition to flights to and from Earth orbit."

So, it seems clear they will indeed use a Dragon 2, so unless there is a massive performance boost with the Falcon 9 block 5, they will have to launch it on a Falcon Heavy.

The Dragon 2 has a capacity of seven crew.  So they will probably have five trained crew as well as the passengers. But it's still a dangerous mission, and I think there is a high risk of an explosion with them all dying on launch, or the life support failing and them all dying on the way around the Moon or indeed during the re-entry into Earth's atmosphere.

So first, there's the matter of the rockets exploding. That has happened each year for the last two years (Sept 1, 2016 and June 28, 2015). Surely they should have a few years of non exploding rockets before they send paying passengers, at least of whatever type of rocket they send the passengers on? And should they not settle down on a stable rocket first? 

They do have a launch system abort.

Crewed Dragon Pad Abort Test in 2015. Elon Musk says it would have saved the crew in the event of the 2016 launch pad explosion,

The Space Shuttle didn't have that - and it would have saved Challenger (which broke apart 73 seconds into its flight) if they'd had one. However that just underlines the need for a launch abort system. The idea is to first get your spacecraft safe enough so that they are not likely to explode on the launch pad. You certainly don't want the system to be triggered routinely.

Then on the life support, the problem is that it doesn't take much of a change in the atmosphere inside the cabin to kill the crew. A level of just 1% of carbon dioxide is enough to be hazardous to humans. That's not a lot, in a small spaceship. That was the main issue for the Apollo 13 astronauts. They had plenty of oxygen, and that wasn't the problem. But they would have all died of carbon dioxide poisoning long before they could get back to Earth, if they hadn't found a way to improvise a CO2 scrubber.

It would be the same for any issue that develops slowly over a period of hours or days is more hazardous in a multi-day mission. Another example of a problem of that nature would be a breach of the spacecraft hull, or damage of essential equipment through a micrometeorite or space debris. Or a failure of a part that works just fine for a few hours but then it fails, if it is a part that’s essential to life support.

It's different with test pilots. The Apollo crew were trained test pilots and jet fighter pilots, used to taking high risks. They knew what they were doing, they knew what they themselves were capable of from long experience in dangerous situations in flight, and they knew the spacecraft they were flying inside out too. If they all died, as happened with Apollo 1- that's very different from an Apollo 1 disaster with paying customers aboard, or an Apollo 13 type disaster that doesn't have a happy ending, again with paying customers. NASA never sent paying passengers with Apollo, but they did send an astronaut who was not a test pilot. But they did this once only, on Apollo 17. Also Apollo did a step by step approach which turned up many problems that needed to be fixed.

So, Apollo 1, the Apollo 10 incident which would have killed the astronauts if they had attempted to complete the landing, and the Apollo 13 incident, all underline this need to go slowly with human spaceflight, the step by step approach. Of course we have learnt a lot since then. Apollo 1 isn't going to happen again. But the likes of Apollo 10 and Apollo 13 could easily, because all the hardware is new, and we have no recent experience of human missions outside of LEO.

There's bound to be an element of risk in space flight. But the attitude so far of the US, Russia and China has been to reduce that risk as much as possible, on the basis that it is hard enough anyway as it is. They all use a much slower step by step approach as a way to reduce risk. 

China for instance has sent several taikonauts to LEO, and has actually built a space laboratory there as well, the Tiangong 2 (meant for testing, not a permanent space station).

But it doesn't have any plans in the near future to send its astronauts to go around the Moon. They do have a heavy lift vehicle, the Long March 5, successfully launched in August 2016, capable of sending 8 tons on a lunar transfer orbit, similar in capacity to the Delta IV heavy.

Back in the 1960s, NASA could have built a big rocket and sent the astronauts to the Moon at an early stage, before Apollo 8. They could have attempted a landing on the Moon even, at a very early stage. As soon as they had the Saturn V, they knew how to do it in theory, and could have got something together that might have worked, but the technology wasn’t tested yet. They would have had a small chance of beating the Russians by a huge margin, but almost certainly everyone would have died. 

They showed the value of the step by step approach as they turned up numerous problems at every stage of the process which would have surely doomed the entire crew if they hadn’t done this approach (right up to Apollo 10 which did everything Apollo 11 did except land and turned up a problem that would have doomed the crew if they had tried to land).

And even with that careful step by step approach, they had all the crew die in the Apollo 1 explosion before they fully understood issues with oxygen atmospheres, and Apollo 13 came close to disaster and was only saved because of the lunar module which they could use as a “lifeboat” to supplement the systems in the command module.

Now of course the Apollo and Gemini programs had to test many things we don't need to test. At the beginning they didn't know even if humans could survive a few hours of zero g (hence all the monkey tests). So I'm not suggesting that SpaceX has to redo all the stages of Gemini or even Apollo. Rather I'm just making the general point that a step by step approach is what saved Apollo from what could have been far worse disasters. There are many steps they had to do which we can skip. But this is just too few steps too quickly in my view.

The risk must surely be far higher than for the Space Shuttle. And surely it's more dangerous than Virgin Galactic who are at least doing many tests with test pilots first before sending the first passenger on a sub orbital hop. And of course Virgin Galactic also had one of their test pilots die in an accident. They aren’t attempting anything as dangerous and spectacular as this. Virgin Galactic sneaks in just one more SpaceShipTwo glide test to cap off 2016

We can do simulations of these missions in software - but that didn't stop the Virgin Galactic crash, or the SpaceX several explosions and failures. Our spacecraft are also much more complex than in the Apollo era - and a simulation is only good as far as it accurately matches what you built and launched. And simulations always involve approximations and sometimes those turn out to be significant and it leaves out something the authors of the software thought didn't matter which leads to a crash.


SpaceX are honest about the risks - but I don't know if their paying passengers will really evaluate that properly. They are bound to have agreements to sign but that doesn't mean they truly understand it.

They may treat it more like a scary fairground ride. Or, they may treat it like a Star Trek movie. The crew keep running huge risks, and at just the last second or minute they find a solution. But this is reality, not fiction. Yes, Apollo 13 found a solution, it could almost have been a movie script the way it worked out. But the two Space Shuttle crew didn’t even know that they were about to die, and a new Apollo 13 could easily not have the same happy ending.

Also, the risk may well be much higher than they think. I wouldn't be surprised if it is as high as 25%?? If they do it as quickly as planned. Of course there are far too many intangibles to evaluate properly with a never tested spacecraft.

Also, if the Falcon Heavy doesn’t crash and the Dragon 2 works fine too, with two successful tests of each one first (say) - two successful launches don’t mean that a spacecraft is safe. If it has a 50/50 chance of failure with each launch, you could have two successful launches in a row with a probability of 25%.

The problem isn't so much, probabilities that they can predict. but the unknown, the mistakes, the flawed component, the failure of integration of something properly, the life support critical component, the onboard fire or explosion etc etc. So much has to go exactly right in the first ever manned mission after only one unmanned mission test. And SpaceX have a track record of things often going wrong, pushing the envelope and taking huge risks and gambling on them succeeding. 

So, yes, this really could happen. It's not like Star Trek or a scary fairground ride.


I have issues with Virgin Galactic too. They are following a slower more responsible approach to the testing, but they are somewhat playing down the risks and encouraging celebrities to fly with them who have no history of doing risky things like base jumping and can’t possibly really truly evaluate the risks, seems to me. Even with a VG suborbital hop, I think the risk will surely be higher than for the Space Shuttle and a lot higher than for the Soyuz missions to the ISS. Unless they do many tests before they fly. The big problem is the high cost of each test. If planes cost so much to fly each time, and could only be tested a few times before the first paying passengers go on board, then it would be just as hard to make a safe plane. I don’t think planes are intrinsically safer than spacecraft. It’s just that, at present anyway, it’s so much easier to test them.

Their announcement is here SpaceX to Send Privately Crewed Dragon Spacecraft Beyond the Moon Next Year - and more about it here: SpaceX reveals plan to send two private citizens around the moon in 2018

The Trump administration are also planning a similar risky mission - pushing back the first humans to use the Space Launch System (SLS) to the maiden flight of their rocket, with a journey around the Moon

Quoting from Human flight around the Moon: a worthy goal, but using the wrong vehicles, this is about the SLS flight but much of this also applies to the SpaceX idea:

“The most serious drawback to the proposed plan, however, is the added risk to the first crew. Flying crew on the first flight of a new launch vehicle is not without precedent. The first flight of the Space Shuttle had a crew. However, the shuttle only orbited the Earth and could quickly return if a problem developed. The Orion capsule, however, would embark on an flight around the Moon lasting more than a week. If a problem occurred at the wrong time, it would take days to get back to Earth.”

“And it’s not just that it is the first flight of the SLS. The Orion capsule has not been adequately flight-tested, either. The Orion capsule did make a single, short uncrewed flight in December 2014, when a Delta IV Heavy rocket boosted the capsule to a high Earth orbit for a five-hour mission. This flight was successful, but the life support system was not installed and the capsule only attained a velocity about 80% of that of a return from the Moon—not enough to fully test the heat shield. Not only that, but this flight lacked the service module that is the critical second component of the Orion spacecraft. The service module, which will be supplied by the European Space Agency, has never been flown and there are no plans to do so prior to the EM-1 mission.”

So, I have concerns with humans on the SLS too. Each mission is so very expensive, half a billion dollars per mission, that they won’t be able to do much testing. It’s not quite the same though because they won’t be doing commercial flights. Rather, it’s more like an early flight with test pilots on board. These are people doing a job they know is risky and which they have chosen as a career, testing spaceships. But it’s risky and could be a huge set back if that first flight is a disaster and they all die.


Everything is in the timing. So far this article has been quite critical of SpaceX's human spaceflight plans. However my reason for writing this is to encourage safe human flight.

So the first thing is to realize, that every time they send a capsule to the ISS with cargo, they make a profit, using a spaceship that could theoretically take humans on board. They have a brilliant plan there, to use the same cabin for humans as they use for payloads to the ISS that require a pressurized cabin. There’s a lot of potential in that idea for helping them to achieve safe human spaceflight quickly.

But they lose a lot of that advantage by continually changing their designs. The Falcon 9 failed in 2015 after nineteen successful flights, or 13 flights if you count the v1.1 as a new launch vehicle. That’s a great track record for an unmanned rocket, but the standards for human spaceflight are far higher. And they probably could have achieved a much better track record if they didn't continually change the design. This philosophy is very unusual for a space launch company, as Doug Messier comments in his article SpaceX’s Philosophy: Reliability Through Continual Upgrades.

 As we saw, the Russians are way ahead of this with the Soyuz spacecraft. They are now on mission 132. That makes it 111 missions in a row so far without a fatality and only minor issues since 1983. I can understand that SpaceX are keen to send humans to space as soon as possible. And they are not required to do more than one unmanned flight first. But they have a great design now, the Dragon 2. This can send far more cargo to the ISS in one go, it’s commercially valuable for cargo, but they will do only one unmanned flight before the first crewed flight.

So, supposing that SpaceX aim for a safety record approaching that of the Soyuz, what could they do differently?

  • Have not just one but several unmanned Dragon 2 flights before the first crewed flight

Freeze it, not just for one flight before the first crew, but long before. There is no financial reason why they have to send humans early on, as they get a profit from every unmanned test of the Dragon 2. There is no competition yet apart from the Soyuz TMA. The SLS is hardly competition with a cost of half a billion dollars per launch, they aren't going to be doing crewed missions to the ISS.

The Space Shuttle was designed in such a way that they had to have humans on board even for the first mission. But SpaceX don’t, so why not exploit that as an advantage?

In this case they would freeze the combination of the Dragon 2 and Falcon block 5 “Full thrust”. To take a software analogy, once you sign it off for human flight, it’s the “release” now and you only do “bug fixes” from then on. You no longer treat it like beta software that you constantly update. So treat spaceships for humans as release builds rather than as beta tester builds.

Later, before they go to the Moon, they need to launch humans on a Dragon 2 + Falcon Heavy. So, why not use freeze that combination also for several unmanned launches before sending humans on board?

Then my second suggestion is to do with life support testing in space, and other problems that may develop slowly:

  • Do two week “shake down” missions to Low Earth Orbit (LEO) first before the first mission around the Moon.

I’m sure many would pay good money for that. Two weeks instead of one, so that you have a generous safety margin for the later Moon trip when it comes. As a bonus, they could end each mission with a one day visit to the ISS perhaps. If Virgin Galactica hope to make money from selling sub-orbital hops, surely SpaceX can make it from selling a week in weightlessness in LEO. This approach is far far safer. If anything goes wrong they can be back on Earth in hours.

I say this as someone who is keen on humans returning to the Moon. But - humans safely on the Moon!

See also my answer to What are your opinions on the SpaceX ‘tourist flight around-the-Moon’ in 2018? This also has more details and cites for the various facts and figures here.

I say this as someone who is keen on humans returning to the Moon. But - humans safely on the Moon!

For more about the SpaceX plans see SpaceX’s Private Lunar Mission in Work for Last Two Years; Other Opportunities on Horizon

It is easy to find news stories gushing with praise for this mission, they fill the Google news search results. But if you look around there are a fair few that sound a more cautious note.

Other critical articles about their mission:

This originates as my answers to two Quora questions: What are your opinions on the SpaceX ‘tourist flight around-the-Moon’ in 2018? and Does SpaceX have the capability to send people around the moon?


You may be interested in my Moon First books. I have many criticisms of the idea of "colonizing Mars" but I'm keen on humans back to the Moon! I wish I could be more enthusiastic about this Elon Musk idea for sending humans to go around the Moon as it would draw attention to the Moon again.

"MOON FIRST Why Humans on Mars Right Now Are Bad for Science", available on kindle, and also to read for free online.

Case For Moon First: Gateway to Entire Solar System - Open Ended Exploration, Planetary Protection at its Heart - kindle edition or Read it online on my website (free).

Facebook group

I've made a new facebook group which you can join to discuss this and other visions for human exploration with planetary protection and biological reversibility as core principles. Case for Moon for Humans - Open Ended with Planetary Protection at its Core

See also

Robert Walker's posts - on Quora

And on Science20

Robert Walker's posts on Science20

Kindle bookshelf - for my author's page

And I have many other booklets on my kindle bookshelf

My kindle books author's page on amazon


NOTE - Sorry I can't reply to comments on this article right now. Some kind of glitch. I keep getting an code 500 internal server message when I try.