Human Exposure To Vacuum: Part 1

The topic of explosive decompression generates a lot of nonsense, particularly in science fiction films and television series, but also scattered across the internet generally. We actually know quite a lot about what would happen if a human being were exposed to the vacuum of space—and it turns out not to be like the movies.

For this first part, I’m going to write a bit about basic physics and physiology, and discuss what that can tell us about the accuracy (or otherwise) of the common SF tropes we see in the movies. In the second part, I’ll move on to the evidence we have from actual vacuum exposures and explosive decompressions.

Exploding, Outland
John Ratzenberger explodes in Outland

Will people explode when exposed to vacuum?
Absolutely not.

Liquid pressures aren’t an issue—when decompressed, the liquids in our tissues will expand only very, very slightly before their ambient pressure drops to zero. Gas pressures are the problem. Our bodies contain various gas cavities which are at the same pressure as the surrounding atmosphere. Most spacecraft and spacesuits aren’t pressurized to one atmosphere, but we can take that as the worst-case scenario for explosive decompression. So at the moment tissue pressures fall to zero, these gas cavities will press outwards against the surround tissue with one atmosphere of pressure—that’s 100 kilopascals, which is 100,000 newtons per square metre.

But skin and soft tissue is strong. Here’s film from Arthur C. Clarke’s World of Strange Powers (1985). The scientists are reproducing a traditional “hook-hanging” rite carried out at Kataragama, Sri Lanka:

The volunteer weighs 55 kilograms, and hangs from six slim hooks. With a generous allowance of 30 square centimetres for the total suspension area, that comes out to pressures of 180,000 newtons per square metre on the soft tissues the hooks supports. That’s almost twice our worse-case limit, and the skin doesn’t even stretch very far.

So no exploding.

Troy Garity freezes in "Sunshine"
Troy Garity freezes (and later shatters) in Sunshine

Will people freeze solid as soon as they are exposed to space?
Absolutely not.

Vacuum is a good insulator. At cool ambient temperatures, our bodies lose heat mainly by conduction and convection, which is why air temperature and wind speed are so important to the way we dress outdoors. In the absence of air, our skin will cool by radiation—the loss of energy at infrared wavelengths emitted by our warm bodies. Depending on skin temperature and clothing, we radiate at anything from 100 to several hundred watts. So that’s how fast we’ll lose heat.

Now, we’re made mainly of water, and water has a high specific heat capacity, around 4000 J/kg/ºC—which means a kilogram of water needs to lose 4000 joules to fall in temperature by one degree Celsius. So an 80-kilogram bag of water (that’s approximately me), is going to need to lose over 300,000 joules of energy before its temperature falls by one degree. (That’s neglecting the continuing metabolic production of energy in the meantime.) If I’m radiating at a generous 500 watts, and producing no internal energy, and receiving no energy from sunlight, that’s ten minutes of vacuum exposure before my temperature falls by just one degree Celsius.

There may be some local difficulties, though. We also cool by evaporation, which becomes significant when it’s hot enough to cause sweating. Water will evaporate from any moist surfaces exposed to vacuum, and it will take energy with it as it does so, driving down the temperature of the tissue it evaporates from. So on exposure to vacuum the eyes, nasal cavity, and probably mouth and respiratory tract are going to start cooling by evaporation. How cold they get will depend on how quickly water moves out of the tissues to replace what is lost from the surface.

But no-one is going to turn to ice crystals and shatter.

Breath-holding detail, 2001
Keir Dullea holds his breath in 2001: A Space Odyssey

Should people hold their breath if about to be exposed to vacuum?
Not a good idea.

With the tissues equilibrated to an ambient zero pressure, the cardiovascular system will continue to work as usual—all its pressures are relative pressures (what engineers call “gauge” pressures). A blood pressure of 120/80 is telling you that the systolic pressure is 120 millimetres of mercury (mmHg)  above ambient, and the diastolic pressure 80 mmHg above ambient. That’s true at sea level with an ambient pressure of one atmosphere, at twenty metres down in the ocean with an ambient pressure of three atmospheres, on top of Everest with an ambient pressure of a third of an atmosphere, or in vacuum at zero atmospheres.

Trouble is, respiratory gas exchange works on absolute pressures. No matter what your tissue ambient pressure is, you still need to breathe a partial pressure of 160 mmHg of oxygen (21% of an atmosphere) to get a normal concentration of oxygen into your blood. This is fine when you’re breathing at one atmosphere of ambient pressure. It’s even fine on top of Everest—if you mix supplementary oxygen into a third of an atmosphere of breathing gas, you can easily get normal oxygenation while still balancing the pressure inside your lungs against the ambient pressure outside your body and in your tissues. There’s no resulting pressure gradient, and nothing gets squashed or stretched.

But in a vacuum, the pressure in your lungs (necessary for gas exchange) is not balanced by any external pressure. Holding air in your chest is going to cause pressure outwards, stretching the lungs; and inwards, compressing the heart and large blood vessels in the middle of your chest. And notice that even a standard 160 mmHg pressure of oxygen is a large pressure, exceeding the normal pressures of arterial blood. It’s enough pressure to squash your heart, which is not going to have a good effect on its ability to pump. This is why people can make themselves faint while trying too hard to blow up a balloon—the high pressure inside their chest interferes with the flow of blood through the heart. So trying to hold on to a lungful of oxygen in vacuum will make your blood pressure crash, and you’ll almost certainly pass out.

And that 160 mmHg is the smallest plausible pressure someone might find themselves trying to hold when suddenly exposed to vacuum. It’s the minimum operating pressure for spacesuits—most operate at around 240 mmHg. The Space Shuttle maintained an internal atmosphere at 530 mmHg during missions. These are pretty lethal pressures to try to hold in the lungs.

It’s not just the cardiovascular system that will suffer. The lungs themselves are not designed to support that sort of pressure differential. In the fifth edition of Diving and Subaquatic Medicine, Chapter 6, Edmonds et al. report that a person’s lungs will leak air into the surrounding tissue when subjected to a pressure gradient of 110 mmHg, even if the chest is prevented from expanding using a binder. If the chest is allowed to expand in response to the imposed pressure, the lungs start to leak at just 70 mmHg. (Admittedly, this is from a cadaver study, but it’s not the sort of test you can find volunteers for.) What’s going on here is that stretch is bad for your lungs, too—if your chest is blown up like a balloon, the lungs will burst at a lower pressure. This is bad news for anyone tempted to take a deep breath before entering vacuum—the extra stretch moves their lungs closer to the burst point. But that’s probably academic, since these experimental burst pressures are lower than the lowest spacesuit operating pressures.

So if you try to hold a lungful of air on decompression, not only will you squash your heart and cause your blood pressure to fall catastrophically, your lungs will leak—they’ll squeeze air into the lung blood vessels, sending showers of bubbles into your circulation; they’ll squeeze air into the tissues around your heart and then up into your neck; and they’ll squeeze air into the pleural cavities lining your chest, causing your lungs to deflate.

As an added extra, the air held in your middle ears will burst your eardrums.

So breath-holding on exposure to vacuum is a good way to incapacitate yourself. Is there another option?

Robin Tunney in "Supernova"
Robin Tunney exhales (albeit too late) in Supernova

Should people exhale on exposure to vacuum?
Better … but still not great.

If breath-holding will make you lose consciousness and pop your lungs, not breath-holding seems like the only viable alternative. If you can arrange to yawn just as the pressure drops, to open your Eustachian tubes and let the air out of your middle ears, then you’ll also prevent your eardrums bursting.

Exhaling gets rid of that abnormal pressure gradient in the chest, so there’s no interference with blood pressure, and no popped lungs. However, it takes time for the lungs to empty, so if decompression is very fast, lung injury could still occur while the pressure in the lungs remains transiently higher than the pressure in the surrounding tissues. Here’s a theoretical plot from the second edition of the Bioastronautics Data Book, showing that a 250 mmHg decompression (from 350 to 100 mmHg) over 0.3 seconds will produce a brief pressure gradient across the chest wall that reaches what we know to be potentially lung-popping levels:

Pressure differential across lung during decompression
Click to enlarge

Having exhaled to vacuum, there’s now no oxygen at all in the lungs. From the point of view of keeping oxygen circulating in the blood, this is a disaster. Venous blood, returning from the tissues, still contains a considerable residue of oxygen. This is normally topped up by oxygen diffusing from the lungs into the blood. Even if not much oxygen is added (for instance, if you’re holding your breath), at least some oxygen goes back to the tissues. But if there’s no oxygen in the lungs, the normal diffusion gradient is reversed—oxygen leaves the venous blood and diffuses into the space inside the lungs. So very little then gets sent back to the tissues. This means that tissue oxygenation fails abruptly and catastrophically—much faster than it does with simple breath-holding.

We’ve got some experience of how quickly things go badly wrong under this sort of hypoxic insult—some of it comes from pilots depressurizing at very high altitude, and some of it comes from people who have accidentally breathed gas containing no oxygen, usually in an industrial accident.

The USN Aerospace Physiologist’s Manual (50 MB pdf) makes a prediction about how long a person might remain conscious and orientated enough to carry out simple tasks, once exposed to near-vacuum. This is based on the apparent convergence of a number of graphs generated by decompressing various hapless volunteers to various altitude-equivalents during the 1960s:

Time of Useful Consciousness (USN Aerospace Physiologist's Manual 4-28)
Click to enlarge

Somewhere around an altitude of 65,000 ft (with an air pressure of 43 mmHg, and a partial pressure of oxygen of just 9 mmHg) the period during which volunteers remain conscious enough to perform simple tasks converges on 12 seconds.

So a Time of Useful Consciousness (TUC) of 12 seconds following  exposure to vacuum is often quoted, but there are two caveats to it, neither of them encouraging. Firstly, this derived time applies to individuals at rest—that is, not trying to do urgent things in order to stay alive. Secondly, it derives from experiments involving non-abrupt decompression. As Paul W. Fisher reports in the USAF Flight Surgeon’s Guide:

These TUCs are for an individual at rest. Any exercise will reduce the time considerably. For example, usually upon exposure to hypoxia at FL 250 [an altitude of 25,000 ft], an average individual has a TUC of 3 to 5 minutes. The same individual, after performing 10 deep knee bends, will have a TUC in the range of 1 to 1.5 minutes.
A rapid decompression can reduce the TUC by up to 50 percent caused by the forced exhalation of the lungs during decompression …

Exercise not only increases the consumption of oxygen, it also increases the speed at which blood returns to the lungs and is depleted of its oxygen content. So (extrapolating the extrapolations!) if you’re explosively decompressed while exercising vigorously, it looks like you might end up with less than six seconds of useful consciousness. Which would be disappointing.

Michael Ironside in "Scanners"
Michael Ironside‘s blood boils (well, not really) in Scanners

Will a person’s blood boil on exposure to vacuum?

At body temperature, water has a saturated vapour pressure of 47 mmHg. Which means that if the ambient pressure falls below 47 mmHg, the water will evaporate into the gas phase throughout its bulk. Which is the definition of boiling—bubbles forming and expanding within the liquid. The altitude at which the atmospheric pressure drops below 47 mmHg (63,000 ft), and the water in human tissues is in danger of boiling, is called the Armstrong Limit—named for Harry Armstrong, one of the pioneers of aviation medicine.

You’ll find some places on the internet that claim a person’s blood won’t boil, because normal blood pressure (120 mmHg systolic, 80 mmHg diastolic, remember) is higher than that 47 mmHg critical value. In support of this claim, many cite physicist Geoffrey Landis’s otherwise excellent exposition on explosive decompression:

Your blood is at a higher pressure than the outside environment. A typical blood pressure might be 75/120. The “75” part of this means that between heartbeats, the blood is at a pressure of 75 Torr (equal to about 100 mbar) above the external pressure. If the external pressure drops to zero, at a blood pressure of 75 Torr the boiling point of water is 46 degrees Celsius (115 F). This is well above body temperature of 37 C (98.6 F). Blood won’t boil, because the elastic pressure of the blood vessels keeps it it a pressure high enough that the body temperature is below the boiling point …

Now, the fact that Landis gets the notation for arterial blood pressure reversed (his numbers should be written 120/75) is probably a hint that he’s not entirely at ease with the physiology of blood circulation. What he has forgotten about is the blood that’s not in the arteries, which at any given time amounts to about 90% of the blood volume—flowing through the capillaries, veins and lung blood vessels, all of which normally have pressures well below the critical value. The pressure in the central veins, in particular, is usually only a few millimetres of mercury above ambient. So although Landis writes “No” in answer to the rhetorical question “Would your blood boil?” what he really means is that ten percent of your blood wouldn’t boil, but the rest would. Which certainly seems more like a “Yes” to me.

It’s also sometimes claimed that, while the veins do run at low pressures normally, they will tightly contain any rise in pressure, preventing gas bubbles from forming. But veins are so-called capacitance vessels—they adjust their volume according to the volume of the circulating blood. They will reach their elastic limit if overfilled, but in healthy adults they’re continuously adjusting their volume at low pressure. It’s possible to infuse a litre or more of fluid into the veins of a healthy adult without the venous pressure shifting much over 15 mmHg. So there’s likewise room for a litre or more of gas to form in the venous side of the circulation without causing any major pressure rise in the system. This is a problem, because the amount of gas necessary to cause cardiac arrest in a human is estimated at 3-5 ml/kg—if a few hundred millilitres of gas gets into the heart chambers, it forms a compressible volume that stops the heart propelling liquid when it pumps. The heart continues beating, but it moves no blood.

Another variation on this optimistic theme is that the skin and subcutaneous tissues are tight enough to prevent gas expanding. (Presumably the people who make this claim have never thought seriously about the level of tissue stretchiness implied by a yawn or a clenched fist.) There’s a condition called subcutaneous emphysema, in which gas (usually air from a leaking lung) becomes trapped  under the skin. We know that even the relatively low pressures associated with a mechanical ventilator (around 20 mmHg) can squeeze gas out of an already injured lung and into the tissues—which shows that the tissues are unable to immediately generate the necessary counterpressure. Like the veins, the tissues eventually reach an elastic limit and oppose the entry of any further gas, but there is considerable, obvious distension before that happens. So, on exposure to vacuum, the surrounding tissues are not going to be able to prevent the veins expanding.

What happens within the tissues themselves is an interesting question. Dense tissues like tendon and ligament may well be able to contain any tendency for gas bubbles to form within their substance. The loose, soft tissue that lies under the skin (the region affected by subcutaneous emphysema) obviously doesn’t have the structural strength to prevent the spread of gas bubbles once they start forming, but there’s some evidence that it is tightly woven enough to suppress initial bubble formation, for a while.

The third edition of the US Naval Flight Surgeon’s Manual (7.5 MB pdf) discusses a study in which the hands of volunteers were decompressed to very low pressures. Subcutaneous gas bubbles didn’t form at the Armstrong Limit; they didn’t even form at pressures equivalent to the boiling point of water at skin temperature, which is a little lower than the body’s core temperature. The highest pressure at which gas formation occurred was 20 mmHg; three people were decompressed to 5 mmHg, and all showed gas formation. But the onset of visible gas was delayed—it occurred “suddenly and manifested itself by marked swelling”, but after a lapse of between thirty seconds and over ten minutes of decompression time. Once swelling occurred, it could be rapidly abolished by recompression. But, strikingly, gas was generated much more readily when the hand was decompressed again. So this experiment suggests that it might be initially difficult for gas bubbles to form spontaneously in the tissues, but once they do they can spread rapidly. Other body gases (nitrogen, oxygen, carbon dioxide) will diffuse into these gas bubbles once they form, and will persist as tiny bubbles for some time after recompression has caused the water vapour to collapse back into the liquid phase. These tiny bubbles then form nuclei that will quickly re-expand with water vapour if the tissues are decompressed again.

Unfortunately, the Naval Flight Surgeon’s Manual is a little short on detail. It’s not clear if these volunteers’ hands were still being perfused with blood, or if they’d been isolated by tourniquet, for instance—which would make a significant difference to the tissue pressures. The study was reported in the first edition of NASA’s Bioastronautics Data Book (1964), but was dropped from the second edition of 1973, apparently superceded by more recent experiments which I’ll describe in my next post on this topic. So the pressures and timings in this report are interesting, but may not reflect what happens with total-body decompression.


As if all the foregoing wasn’t enough, a few other problems may occur. If the atmosphere you’re breathing immediately before decompression contains nitrogen, some of that nitrogen will bubble out of solution, in the tissues and blood, as the ambient pressure falls. While this can be a major problem for aviators at altitude, it’s probably a relatively minor problem for those exposed to vacuum, given that water vapour bubbles will be forming. And the second edition of the Bioastronautics Data Book notes that, “The symptoms of decompression sickness are rarely observed during the first few minutes of exposure to low pressure.”

I’ve already mentioned the problem of air trapped in the middle ear. Other areas where air can be trapped, causing pain when the ambient pressure falls, are in the sinuses and under dental fillings. The USN Aerospace Physiologist’s Manual puts the experimental incidence of sinus problems on decompression at 1%, and of dental pain at 0.1%.

Finally, there’s the volume of gas that’s sitting in everyone’s stomach and intestines. This will expand as the ambient pressure falls. On slow decompression it generally causes cramping pain, burping and flatulence. But both the third edition of the USN Flight Surgeon’s Manual and the second edition of the Bioastronautics Data Book note that more rapid expansion of gut gas has the potential to intensely stimulate the vagus nerve, causing a profound fall in heart rate and blood pressure, leading to unconsciousness.


You won’t explode, but you may well swell up. You won’t freeze instantly, but your eyes, nose, mouth and airways will experience evaporative cooling. You shouldn’t try to hold your breath. You should breathe out and yawn as the pressure drops, but if decompression is explosive that may not protect your lungs from pressure injury. Your venous blood will boil, and there’s room in your venous circulation to generate enough gas to stop your heart moving blood. The gas in your gut may expand so rapidly it leads to a reflex slowing of your heart and rapid fall in blood pressure. You will have a maximum of 12 seconds of useful consciousness, but if you’re exerting yourself and/or the decompression has been explosive, your period of consciousness may well be considerably shorter.

For Part 2 of this topic, I’m going to look at the data we have from actual vacuum exposure, in humans and animals.

Note: Almost all pressures are quoted in an antique unit of measurement, the millimetre of mercury (mmHg). This is because the most familiar physiological pressure for most people is blood pressure, which is always quoted in millimetres of mercury, and also because a lot of the relevant literature dates back to a time when atmospheric gas pressures were quoted in millimetres of mercury.
If you want to convert, the following are equivalents, in round numbers: 1 atmosphere, 1000 millibars, 760 millimetres of mercury, 100 kilopascals, 15 pounds per square inch.

2 thoughts on “Human Exposure To Vacuum: Part 1”

  1. I seem to recall some discussion of exhalation in Arthur C. Clark’s original 2001 before Bowman decompressed his pod. Clarke, as always, did his research, though I imagine there was also a degree of dramatic license as well.

    1. Yes, by the time of writing 2001 Clarke was aware of some of the animal studies that I’m going to describe in Part 2. He actually wrote an article entitled “A Breath Of Fresh Vacuum” just before 2001 was released, defending the airlock scene and citing references. In it writes, “As long as the main air cavities could be kept open, so that there was no pressure build-up, exposure to vacuum should not cause damage.” Elsewhere (I think in The Lost Worlds of 2001, but I can’t put my hand on it at the moment), he mentions that Dave Bowman should have breathed out, not held his breath.
      But Clarke actually wrote about the possibility of people surviving vacuum exposure long before there were any experiments; indeed, before there were any humans in space at all – in the novel Earthlight in 1954, and the short story “Take A Deep Breath” in 1957.

Leave a Reply

Your email address will not be published. Required fields are marked *