Leaving People Behind
In order to increase the number of people in space, we cannot limit ourselves to the number of seats in spacecraft docked at the space station.
News from the International Space Station has recently been dominated by the story of two astronauts “stranded” there by a faulty spacecraft, with much discussion around if the term applies at all to the situation.
Sunita Williams and Butch Wilmore launched in a Boeing Starliner for its first manned test flight intended to last 8 days, and now have spent 75 days in space. The cause of the problem is in the reaction control engines of the spacecraft. Helium pressurant leaks were detected during the approach to the station, and stopping these leaks required taking some of the RCS engines offline, but the spacecraft was still able to dock successfully. The spacecraft still has enough helium on board to complete its mission, but there are concerns about a faulty RCS system causing recontact with the station after undocking, or failing to perform a deorbit burn.
NASA is not bringing the astronauts home yet because they are safe where they are, and they have plenty of time to assess the status of the Starliner. At time of writing, no decision has been publicly announced as to whether the astronauts will come home on Starliner, or wait on the station to be returned in a later SpaceX Dragon flight.
This has spurred me to discuss something I’ve been meaning to talk about for a while. At present, every person staying in space has a seat in a spacecraft attached to the station that can bring them home. This coming into doubt is the cause of the current issues. But it would be fine, from a point of view of health and safety at least, for the two astronauts to return to Earth another way. So do we actually need one seat per crew member at all times? Doing so places a hard limit on how many people can be in space at once - and that is a figure that we want to see increase in the future.
I will mostly not be talking about immediate emergency situations here - more about the need to have some way to bring people home no matter what happens, how has this problem been approached in the past, and how it might be approached in the future.
A Brief History of Rescue Missions
In the rush of the early space race, there was not much scope for a rescuing astronauts. There were launch abort systems, and some capsules had ejection seats - Vostok astronauts notably had to eject before landing, which the Soviets kept secret in order to qualify for altitude records. But aside from this any problems in space required astronauts to make their own way home if they could - for Gemini 7 and Apollo 13 this succeeded, but for Soyuz 1 and Soyuz 11 it did not an the crews were killed.
It was understood that this was not ideal by both sides - when planning their own lunar landings, the Soviets considered flying an unmanned LK lunar lander to the target site first so that the single cosmonaut visiting the surface would have a backup spacecraft to leave if the one he came in could not get him back into lunar orbit. Meanwhile, NASA contractors did detailed studies on how to recover astronauts from any future Moon base that might follow Apollo.
But the first operational capability was prepared during the three missions to the Skylab station. The rescue mission would use the next Saturn-1B/Apollo that was due to launch (or the backup hardware if the problem occurred on the final flight). A conversion kit would be applied to the capsule to give it 5 seats, and two astronauts would launch aboard it and dock to Skylab’s second docking port (circled below).
This was a contingency in the case the Apollo capsule the astronauts had traveled to the station in was incapable of returning, and not suitable for an immediate emergency. At the start of the mission it would have taken 48 days to prepare a rescue mission, whereas towards the end it could have taken as few as 9 days, as the rocket and capsule would be further along their preparations for the next mission.
When in 1984 NASA committed to building Space Station Freedom, it had to contend with the problem that the Shuttle could not stay in space for more than a couple of weeks, and could not as designed extend its life being docked to s station. To this end, there were various proposals for what they termed an Assured Crew Return Vehicle (ACRV) that could return the entire station crew if there was an emergency or an interruption in Shuttle flights. Unlike Skylab rescue, this would be attached permanently to the station and available for use at short notice. When the Freedom program was merged with Mir-2 to form the International Space Station, the Soyuz capsule became available for this role and there was little case for the US building its own ACRV anymore, so the project was cancelled.
After the loss of Columbia in 2003 due to a damaged heat shield, The Columbia Accident Investigation Board had looked into whether or not a rescue could have been flown for STS-107 if the extend of the damage had been known - Atlantis was preparing for its next mission and if a heroic effort had been made to accelerate this, it could have safely been launched 5 days before Columbia ran out of consumables. A reduced crew of 4 on Atlantis would then rendezvous with Columbia and evacuate the crew via EVA and return home.
Based on the recommendation of CAIB, the remaining Shuttle flights would always have the next Shuttle being prepared for flight standing by to fly a rescue mission known as Contingency Crew Support, timed such that there would not have to be such a rush to launch it as there would’ve been with a hypothetical Columbia rescue,
Most of the remaining Shuttle missions were to the ISS, which substantially eased the situation. Each Shuttle flight performed a backflip next to the station that was filmed so as to allow its heat shield to be inspected. If an orbiter had been found to be unfit to return, it could be abandoned and the crew supported by the station until the contingency Shuttle arrived. This was to happen in about 40 days, which is longer than the Shuttle can maintain its crew but ISS is better supplied, removing a bit of time pressure.
There was one post-Columbia mission that required a more timely rescue, as it was the final Hubble servicing mission and Hubble is in a very different orbit from the ISS. This rescue, STS-400, would have to be ready in 3 days. Thankfully it was not needed.
The final flight of the Shuttle program STS-135 could not have such a contingency flight available at all, but was a flight to the station and thus the crew could wait it out there until bought back in Soyuz capsules had that orbiter not been able to land. This is perhaps the contingency most similar to that currently faced by the Starliner crew.
Be Wary Of Analogies
It is important when examining this topic to avoid reasoning by analogy - for instance, I often hear talk of “lifeboats”. Every schoolboy knows the importance of lifeboats on ships from the story of the RMS Titanic - but space stations do not sink. Even if the analogy is taken seriously, the lifeboats on the Titanic were picked up by RMS Carpathia in a couple of hours and could not have bought the survivors to a safe port on their own. In contrast to such simple vehicles designed just to keep people alive until rescue comes, the “lifeboats” of the ISS provide complete transport all the way home.
So maybe a space “lifeboat” need just be a module that floats around, keeps its occupants alive, and loudly broadcasts a signal to attract rescuers. Perhaps. But in many situations, its far from clear that the people in such a module are better off being detached from the station. The reason that lifeboats have to be deployed from sinking ships is that they will end up being dragged down as well if they do not. So when I say “space stations do not sink” I’m not just being flippant. Its key to why the analogy is flawed.
We can look at some historical accidents to gain some better perspective. In February 1997 there was a fire aboard the Mir space station caused by a chemical oxygen generator. The fire was extinguished although the crew had to use respirators for some time. In July of that year a more serious accident occurred when an out of control Progress spacecraft crashed into the Spektr module, causing it to depressurise and requiring the crew to seal it off. Neither incident required evacuation although it was available in both instances.
So what are the actual requirements? In the case of a fire, depressurisation, loss of attitude control etc. we want the people on board to be kept safe until such a time as they can return home (and ideally that should be done as quickly as possible). Thus far no space station incident has required an immediate evacuation to Earth - the only fatalities in any space station program have been the crew of Salyut 1, who died when their Soyuz capsule depressurised after leaving the station (contrary to some reports, this occurred when the capsule split apart ready for reentry, not when it undocked from the station).
There are cases where rapid evacuation to Earth would be helpful - e.g. medical issues that cannot be addressed on the station. But that almost certainly would not for the entire crew and thus wouldn’t need to stick to one-seat-per-person. Severe damage to the station, such as a debris strike, may require the entire crew to return - but what are the chances that the impact is so bad that there are no safe parts of the station at all, and yet all the return craft are left intact?
No Time For Caution
Future space stations, for the sake of comfort and health, are likely to spin to produce artificial gravity. Does this complicate any potential rescue?
ABOVE, a space startup, proposes the Voyager station below as part of its long term vision. The crew can be seen transported by Starships, but the gravity ring they occupy is hemmed with what appear to be Sierra Nevada Dreamchaser spacecraft, to support a potential evacuation.
It seems to me this design ignores the financial depreciation and the physical damage of the LEO environment on these very expensive bits of hardware that will be doing nothing most of the time. Why do this? Aside from following the one-seat-per-crew paradigm, there may be a concern that if the station spin becomes unstable somehow that rescue vehicles may not be able to dock with the central hub. Perhaps that is the use case for evacuating the station rather than finding a safe haven within it? Not necessarily.
The docking of Soyuz T-13 with the disabled and out of control Salyut 7 shows it is possible - although I can’t find a source for the rotation rate of the station, if it were even measured, but most descriptions of the missions suggest it was fairly low. It was also described as rotating around its long axis, making the cosmonauts job a lot easier as the docking part lay along this axis. If this axis of rotation was at a large angle from the direction the spacecraft needed to dock from, it might have been trickier. Note that such a situation is depicted in the movie Salyut 7, but that film is fairly liberal with the truth in many ways.
There have been other instances of attaching to spinning objects in space. In this video of a satellite retrieval on STS-51-I in 1985, I eyeball the target as spinning at around 2 revolutions per minute. Future space habitats may spin a bit faster, at 4-6 rpm, but not that much faster. A crew member at the end of the Shuttle’s arm was able to grab the satellite and bring it into the cargo bay for repair.
So mounting a rescue on a slowly spinning object is not out of the question. And if needs be, the spin rate of a station can be reduced by using a ‘yo-yo de-spin’ device, whereby weights are released on the ends of tethers, which by conservation of angular momentum causes the rotation rate to slow. These are in routine use on satellites and could be scaled up.
Increasing the Numbers
Understanding that its almost invariably the best move to stay with a station if anything goes wrong either there or with the transportation system, we can then use the prior contingency plans to figure out pushing beyond one seat per person would look like. Let’s say you wanted a much larger station than the ISS - one with 100 people on board. Let’s also say that, like the late Shuttle missions, you wanted to be able to retrieve your people in about 40 days.
If it was decided to evacuate then you would have to fly 100 seats worth of Dragon capsules in that time. At 4 seats per capsule, that means you need to fly once every 1.6 days to have this contingency. If you modified the Dragon for 7 seats as originally intended, it takes the figure to a flight every 2.8 days. The Falcon rocket seems to be already capable of this cadence, but for every such flight to carry a Dragon might need a much larger fleet of capsules. This seems like it would be expensive and difficult.
Obviously, if you instead used Starship, the problem becomes rather trivial. If your crew taxi can take 100 people then you only need to fly a crew Starship every 40 days, which is not that much faster than the current cadence of test flights. But with Starship in hand, you probably want to push for even larger crews. Below I’ve done a breakdown of how many people could be kept in space given the above constraints.
Note I’ve used a log scale on the y-axis here, simply to fit all the curves in, but they are straight lines on a linear graph. The points are shown for monthly, bimonthly, weekly, biweekly, daily and twice daily flights. Launching two Starships a week, close to the current Falcon 9 rate, would permit over 1000 people to stay in space with the above evacuation criterion.
Conclusion
There have been calls in the US to formally establish some sort of rescue service, either separately or as part of the US Space Force. This seems to me an eminently sensible idea, and I think that there should be international cooperation on this. Having more launch sites with a rescue vehicle ready to launch means that an emergency rescue can on average be performed more quickly as the time from signalling distress to overflying a spaceport that has help available will be shorter. Rescue missions could perhaps be designed to use overpowered boosters allowing the vehicle to perform sharp “dog leg” manoeuvres, increasing the coverage of each spaceport.
At a certain scale though, complete evacuation won’t be needed though. As the population and size of a station increases, its capability to deal with medical emergencies will undoubtedly increase as well. Fires and depressurisation events will be of smaller relative scale and with more people and equipment will be easier to deal with. Also, the premise of evacuating people within a certain time frame is the need to bring them home - and at some point a very large habitat becomes a home in itself.
Until they, let’s hope to see some new ideas in this field to help support more people living and working in space.
This is the end of the article. This one is free for all readers, but some of my work is for paid subscribers only. Upgrading gets you full access to articles like these:
An analysis of Japan’s contribution to the Artemis program and what the UK can learn from it.
The history and current state of the art of partial gravity research, important for building human colonies on Mars or the Moon.
How the threat of ASAT weapons and satellite constellation technology could lead to a militarisation of space.
…and many more, including my regular monthly Mass Value Report where I apply a unique analytical lens I developed in this paper to the launch and space industries.
To subscribe costs about the same as a cup of coffee every month. Would you buy me a coffee to hear a complete set of my insights each month?
Keep reading with a 7-day free trial
Subscribe to Planetocracy to keep reading this post and get 7 days of free access to the full post archives.