Flying High Unpressurized

Climbing through 16,000 for 23,000 feet, it dawned on me that I was feeling weird. It wasn’t sudden or severe, but once it had my attention, hypoxemia was my first thought. Hypo is low or below. A hypodermic needle goes below the dermis, the skin. Ox is oxygen, of course, and emia refers to blood. Hypoxemia means low oxygen in the blood, and it’s an efficient killer.

flying unpressurized

My wife and I both had nasal prong (nasal cannula) oxygen in place with plans to switch to masks before FL180. Quickly, I used the finger oximeter and found that my blood oxygen saturation, which should be at least 90%, was 73%. I remember asking, “Theresa, how are you?” “Fine, why?” I came back only with “Okay, let’s swap oxygen; if one of us passes out, it shouldn’t be me.” That got her full and eager cooperation. I arrested the climb, swapped our hoses, took some deep breaths, felt better, and then found and fixed the kink in what was now Mrs. Levinson’s oxygen tubing. I diagnosed the oxygen problem quickly because I was ready for it, and if you fly high in an unpressurized cabin, you had better be also.


In 2012, a student pilot flew a PA-24 to FL180 without bothering to use supplemental oxygen. The outcome of his expedition was predictable, and at autopsy he was found to have ingested quite a lot of alcohol with some marijuana, as well, likely explaining his ill-advised expedition.

In late July 2009, a Cirrus SR22 pilot at FL250 requested descent to 12,000 feet and was so cleared, but didn’t comply. The pilot sounded “in distress and out of breath” to ATC. A National Guard interceptor found the pilot “unresponsive” and apparently “unconscious.” They escorted the aircraft until it ran out of fuel and crashed. This pilot was in the habit of using ship’s oxygen with a pulse delivery system connected to an unapproved mask. He had suffered from hypoxia previously, but even so, stuck with his improvised system and only used an oximeter “randomly.”

In August 2014, another SR22 pilot spent an hour at FL210 before requesting 17,000 for no stated reason. After 28 minutes at 17,000, he requested, got and descended to 13,000. Twenty-one minutes later, he called ATC clearly confused and wanting to descend. ATC approved, but he never descended. Intercepting fighters found the Cirrus pilot slumped and unconscious. He was lost at sea.

A similar story, but with a happier outcome involves yet another Cirrus, N591WA, this time from May 17, 2011. At 17,000 feet, the pilot-husband didn’t lose consciousness, but was completely out of it. His wife saved the day by working with ATC and another aircraft on frequency to get down to breathable air. It took a while, but the pilot got better and landed the plane safely. The audio transcript is online and will hold your full attention.

We think of hypoxemia as something that happens all at once leading to unconsciousness, but it’s often not like that. The victim can be mildly to severely confused and even combative for a period of time. Loss of consciousness may occur late or not at all. Importantly, it can take real time for the mind to clear with restoration of oxygen, especially if oxygen is only partially restored, as with a descent to 17,000 or even 13,000 feet.

Why Fly High

For the last decade or so, new piston singles are frequently sold with turbochargers or turbonormalizers, which enable piston engines to develop sea-level power in the thin air of the flight levels. The same power setting that gets a Mooney Acclaim S to a true airspeed of 191 knots at 2,000 feet MSL will get as fast as 242 knots at FL250. Go higher to go faster and get more miles per gallon all at the same time. Go higher to glide farther should the engine fail over open water or inhospitable terrain. Fly on top of a good deal of weather. What’s not to love about that?

Well, there are a few things not to love. Headwinds generally increase with altitude, so flying high into the wind may be a net loser, though in my experience that’s not common. A sick passenger or fire in flight will want you on the ground real soon, like now, and that takes time from altitude. Far and away, though, the greatest concern when flying high is that for pilots, as for their piston engines, oxygen isn’t optional, and it’s in relatively short supply up there.


On a standard day, the barometer at sea level is 29.92 inches of mercury, or 14.7 psi; 21% of air is oxygen, so the partial pressure of oxygen (pO2) at sea level is 21% of 14.7, or 3.1 psi. At 18,000 feet, overall air pressure is about 7.34 psi, so the pO2 is about 1.54, about half that at sea level.

The oxygen-carrying capacity of blood, though, doesn’t vary linearly with the pO2, and that’s because the amount of oxygen dissolved in the blood is trivial. It’s pretty much all carried by the hemoglobin (hb) in red blood cells, so we need to know how much oxygen binds to hb as a function of the pO2, and that’s a curve every medical and nursing student learns early on.

Look at the Oxyhemoglobin Dissociation Curve, which shows the oxygen saturation of hb on the Y axis as a function of pO2 on the x axis. You’ll see this system is beautifully designed. As you go from very high oxygen on the right to lower oxygen on the left, you don’t lose very much oxygen in your blood for quite a while. That’s why we do okay with a mild to moderate pneumonia or flying without supplemental oxygen at 5,000 or 10,000 feet. Unfortunately, as environmental pO2 drops, blood oxygen must follow sooner or later; and by the time you get to a pO2 in the 60 range, the curve steepens, so blood oxygen is dropping fast. That’s the science. The practical implication is that when your finger oximeter shows a saturation of 90%, you’re fine, but even a little lower at 80% to 85%, and you’re not far from a world of pain.

flying unpressurized

This same curve explains why the time of useful consciousness doesn’t fall off slowly and steadily as you climb through the atmosphere. Look at the Time Of Useful Consciousness” Table. At 15,000 feet, you may be a little bit wifty, but most healthy nonsmokers will function pretty well more or less indefinitely. At 20,000 feet, you’re in trouble fast, and at 25,000, you’re hosed.

flying unpressurized

One last thing on physiology. As we get older, we tend to accumulate medical diagnoses and, of course, some of us smoke. Smokers will do less well at any altitude, because with every drag on a cigarette, smokers inhale carbon monoxide, which binds to hb 210 times more strongly than does oxygen, and it stays there a long time, crowding out the oxygen. Smokers have plenty of hb, but it’s not available to carry oxygen. Anemia also decreases oxygen-carrying capacity, but it does that by decreasing the hb level. An anemia would have to be severe in a healthy person to make an important difference. Most cardiac patients do fine at altitude unless they’re in active heart failure or otherwise severely compromised. Patients with even moderate lung disease, however, may not do so well. Think about these things when you consider which passengers to get high and how high to get them.

Diagnosis & Monitoring

One of the things that makes doctoring fun and challenging is the degree to which different people experience the same disorder differently. If you’re to fly high safely, you really must learn your own symptoms of incipient hypoxemia.

The first time I flew my Mooney above 15,000 feet was with a CFII. At FL230, I briefed him on my plan to learn my own symptoms of hypoxemia. He took the controls and assured me he would attend to me and descend the aircraft if needed. All I had to do was loosen the lower strap and tilt my mask up away from my face for about one minute before I felt odd. Nothing drastic, but decidedly not normal. I did my best to memorize that feeling, but in addition, on every flight to 15,000 feet or above, I remind myself before takeoff to be alert to any strange feeling at all and think of oxygen first.

flying unpressurized
John Levinson (above, right ) in his Mooney at FL230, with CFII Steve Davis (above, left).

Two things are essential for safe high-altitude flight without pressurization: an appropriately high index of suspicion and a finger oximeter. Your threat-o-meter should go up with altitude, from attentive above about 10,000 feet to concerned at 15,000 and then fully spring-loaded from the lowest flight levels. Above 12,000 to 14,000 feet, I check my O2 saturation every 5 or 10 minutes, and I’m thinking about my thinking. I’m monitoring myself for stupidity. Do I miss an ATC call? Do I have to think twice about a fresh clearance?

One thing that’s not needed, but that adds greatly to safety in the flight levels is the presence of a second and properly briefed human in the cockpit. This need not be a pilot, but someone calm enough to understand the concern without fear. You keep an eye on each other at altitude.

Systems To Permit High-Altitude Flight

One can enable high-altitude flight by increasing cabin pressure back toward 14.7 psi, and that’s what happens in a pressurized airplane. Alternatively, one can accept the lower pressure, but increase oxygen percentage from 21% to something higher by adding purified oxygen. Both strategies increase the pO2, but adding oxygen is less expensive.

There are two ways to add oxygen. For many decades, you could bring an oxygen tank along, and that’s still the most common approach. Now, though, you can also produce it on the fly (sorry) with a concentrator. These concentrators are not yet common in aviation, but they will be. Lighter than tanks and never in need of a refill, concentrators use ship’s power with battery backup to purify the oxygen from ambient air. I’ve seen units permitting flight to 15,000 and have heard that one is approved to 17,999 feet.

Oxygen tanks, the de facto standard, can be built in or portable, but they’re useless when empty, so they need preflight attention. Oxygen is expensive, but again, there are options, at least for Part 91 operators. It used to be that medical, industrial and aviation breathing oxygen (ABO) were importantly different, but since the early 1970s, all are produced by the same process, called liquefaction. However, medical oxygen is tested for aromatic hydrocarbons, ABO is tested for moisture content, and industrial oxygen isn’t tested at all, used mostly for welding. Moisture content needs to be very low for aviation applications to prevent ice formation and line clogging at altitude. Part 91 operators aren’t required to use ABO, but I prefer to do so because it’s easier to get a fill and because I conserve oxygen with a pulse delivery system, so my annual oxygen cost is tiny.

Standard GA oxygen systems send a steady stream of oxygen to the cannulas or masks. Because you need oxygen only when you inhale, most is wasted by such systems. There are probably other vendors, but Mountain High makes wonderful and easy-to-use 1-place and 2-place pulse demand systems. They deliver oxygen to each hose only after sensing the negative pressure caused by the user initiating a breath. They can be set to activate only above certain altitudes and to deliver more or less with each puff. I’ve had this system since I got my Mooney, and it has long since paid for itself. It also enhances comfort, blowing less of that dry oxygen over the sensitive nasal mucosa. Several companies make nasal cannulas that attach to your headset instead of looping around the ears, and that makes them a lot more comfortable under a headset.

Oxygen masks can be used instead of nasal cannula systems and, of course, they’re required by the FARs in the flight levels. Masks are unpleasant, especially if you wear glasses, and they certainly hamper cockpit and radio conversation. Interestingly, my passengers and I have noticed consistently that conversation above 12,000 to 15,000 feet isn’t much better with nasal prong oxygen. You get tired quickly and oxygen saturation drops off with conversation. Even so, with another pilot acting as PIC, I experimented once with nasal prongs as high as FL230 and found they work fine at high oxygen flow settings, but I wouldn’t trust them to keep working up where the oxygen supply is so tenuous, and I think the FARs got this one right.

Another kind of treatment is worth knowing about because it’s smart and we’ll be seeing similar and smarter systems come along. The envelope protection mode in the Cirrus Perspective avionics system (by Garmin) has an optional “Hypoxia detection with automatic descent mode.” The system functions only above 14,900 feet with the autopilot on. It monitors pilot alertness by looking for the kinds of key presses and knob turns that characterize normal flight. It purposely ignores push-to-talk and audio panel interactions, which a confused pilot might perform. For inactivity to count, it must persist for 30 minutes at 15,000 feet, 20 minutes at FL180 and 5 minutes at FL250.

flying unpressurized
Keep an eye on your flight levels. Without oxygen, at 15,000 feet, you may be a bit fuzzy, but most healthy nonsmokers will function pretty well more or less indefinitely. At 20,000 feet, you’re in trouble fast, and at 25,000, you’re hosed.

If inactivity is detected, the pilot receives a “HYPOXIA ALERT,” and if he or she fails to reset the system by pressing any softkey or turning a knob within one minute, the autopilot will command a descent to 14,000 feet, where it will hold level for four minutes following which if the pilot is still inactive, it will descend to 12,500 feet.

If I were the system designer, I would decrease the default allowable inactivity periods somewhat and I would make them user-configurable to be even shorter (but not longer), if the pilot wishes. More importantly, given the possibility of hypoxemia leaving the pilot conscious but confused, once the HYPOXIA ALERT is generated, I would cue the pilot to function at a higher level before assuming all is well. For example, the pilot might be instructed by the system to press the push-to-talk three times rapidly or a defined sequence of softkeys to prove not mere consciousness, but purposeful thought before resetting the system. You could imagine a system that also connects to a continuously monitored oximeter. I would be a lot more relaxed flying solo in the flight levels with systems like these watching my back.

John Levinson, MD, PhD, practices and teaches Cardiology and Medicine at Massachusetts General Hospital and Harvard Medical School in Boston. An instrument-rated private pilot, he uses his Mooney for business and personal transportation, flying mostly with his non-pilot wife whose very different perspective adds greatly to all he sees and writes.


Every pilot balances risk against benefit with many decisions on every flight. High-altitude flight in an unpressurized aircraft is a wonderful arrow for your aviation quiver, and safety can be maximized with the following steps.

  • In a hypobaric chamber or while flying as a passenger and after a careful briefing, do the experiment to learn your own symptoms of hypoxemia.
  • If you ever fly over 10,000 feet, keep a finger oximeter in the cockpit, and use it often and regularly. You can purchase one for under $20 to $30 if you do a little looking.
  • Above 10,000 feet, you should be concerned, and above 15,000, very concerned about hypoxemia.
  • Over FL180, the slope is slippery. Keep a close eye on your thinking and your oxygen saturation. Use a mask as required by the FARs.
  • The additional risk in ascending from FL180 or FL190 to FL250 is large. Don’t make even part of that climb without a strong specific reason such as weather, wind or fuel.
  • ABO isn’t required for Part 91 operations, so you can save yourself some money there, but you should also look at pulse delivery systems.
  • Consider carrying a disposable bottle of emergency oxygen to use in case of a system failure. These can be purchased online from common aviation supply houses.
  • Think hard before taking passengers with lung disease or who smoke above even 5,000 feet.


Every pilot is taught to warn passengers not to apply lipstick and other lip balms, makeup, sunblock, moustache wax (really?) or other petroleum-based products before flights with supplemental oxygen because of a fire risk. I brought this up to a non-pilot friend who practices pulmonary critical care medicine and is an expert in this area. He politely pressed, “You’re thinking about spontaneous combustion?” and then inquired about any unprescribed medications I might be taking.

A literature review was interesting. The risk of these substances causing fire is given weight without evidence by every aviation source I found. Two nonaviation sources, WEBMD and the NY State Department of Public Health, also advise avoiding petroleum-based products around oxygen. A few reports from the nursing literature suggest this risk isn’t real. The best is from the American Journal of Nursing in November, 1998, “Dispelling the Petroleum Jelly Myth.” The authors limited their scope to petroleum jelly, however, and excluded “oil, grease, and other flammable substances.”

Deep dredging finally revealed a proper scientific study by Dille, et. al., published by the FAA, November, 1963, entitled “The Flammability of Lip, Face and Hair Preparations in the Presence of 100% Oxygen.” The authors studied many products in the presence of “high concentrations and pressures of oxygen and of static sparks.” They concluded, “A wide margin of safety was found for their use at or below one atmosphere of pressure,” but not so at hyperbaric pressure as low as two atmospheres. Myth explained—and busted.

Our resident medical doctor and private pilot John Levinson, MD, PhD, is a regular contributor to Plane & Pilot. Read his article “Anxiety and Pilots” for more insight.