Quantifying the Incredibly Unlikely Aerosol Transmission of COVID on Airplanes – Cranky Flier

There’s a good chance you know someone who has told you, “oh yeah, I know someone who got COVID on an airplane.” That’s most likely not true. Not only is it unlikely that someone could pinpoint transmission like that in the US, but there’s also been plenty of work done showing that it’s probably not happening even remotely often. Studies on transmission via droplets show that masks are very effective at preventing transmission, but what about smaller particles allowing aerosol transmission? We have a new study which gets to that point. The chance of you getting COVID via aerosol transmission on an airplane is remarkably tiny.

The study, thrillingly called TRANSCOM/AMC Commercial Aircraft Cabin Aerosol Dispersion Tests, was commissioned by the federal government in order to test how safe troops would be during transit on commercial aircraft. It’s a worthwhile read, but if you don’t speak science, it might be somewhat… dense. Fortunately, I was able to connect with co-authors David Silcott from S3i, LLC and Sean Kinahan from the National Strategic Research Institute to get my questions answered.

First, let’s get this out of the way. The research was funded entirely by the government. United donated a 777-200 and 767-300 for 8 days of testing, but that’s it. Those aircraft presumably were chosen because they were lying around at Dulles with nothing else to do, but I don’t know that for sure. They are also likely troop transporters, so they are good airplanes to use for this purpose. While David and Sean were highly complimentary of United’s cooperativeness, they made it clear that United had no input in anything except for ensuring the safety of the aircraft.

The Set Up

During those 8 days at the end of August, they did over 800 tests in multiple configurations. Two days per aircraft were spent in the hangar as well as attached to a jet bridge to simulate boarding and deplaning. The other two days per aircraft were spent at altitude.

There were only a handful of researchers in the cabin, but they had a whole lot of instrumentation. Here’s how it looked:

Image via United Airlines

What you see is 42 sensors strapped into seats along with a tape substance on surfaces… all surrounding Ruth. Who is Ruth? Well, that’s the biblically-inpsired name — as in “book of” — given by people at a church where they were doing some testing to the disembodied mannequin with the vacant stare who was patient zero. Here she is looking rather like a guest star on Futurama.

Image via United Airlines

Bad news for Ruth… she was sick with COVID. But fortunately for her, she was also asymptomatic. The assumption was that someone with symptoms would not be allowed to board the aircraft. (Remember, we’re talking about troops here, so the likelihood of someone traveling and hiding symptoms is probably pretty low.)

Fixing the Variables

The study looked at a variety of people with other coronaviruses and some with COVID-19 in the hospital to try to determine how Ruth’s sickness would look in aerosol form. They settled on the two most common sizes of particles: 1 micron and 3 microns. Aerosols can come in other sizes, but these were the most likely to match reality.

The 1 micron particles were given an ultra-bright fluoresence that was easily detectable by these specialized sensors. The sensors would, in real time, measure the particles getting to each seat and then also be able to track dissipation. The 3 micron particles were tagged with a unique string of DNA that would use PCR tests (what the most accurate COVID tests use) to test their presence. A couple tests were done for person to person aerosol transmission with the larger particles, but those were more useful at helping track how they stuck to surfaces.

There were some other tough decisions to make. For example, how much virus would Ruth be shedding? There isn’t a defined science showing that a person will expel x amount of virus. It can vary greatly, but using those sample patients that were measured in other studies, they settled on 4,000 virions per hour. A virion sounds like the latest name for a sports car, but it’s really just a unit of the virus particles.

At the same time, they used 1,000 virions inhaled as being an infectious dose. While there is no silver bullet here, they did say they expect to test out different rates once they get to peer review.

For the tests themselves, they spent a lot of time moving Ruth and her 42 sensor-friends around. On the 767-300, they did a set-up in the front, middle, and aft. On the 777-200, they did the front, mid-front, mid-aft, and aft. Ruth moved around to different seats in each zone so they could measure this from many angles. It sounds exhausting. I hope Ruth was able to lie down and rest when she was up in Polaris.

Oh, and did I mention that Ruth was asymptomatic? I lied. She was mostly asymptomatic, but on the 767, she caught a cough for 30 of the inflight tests. That means the velocity of the particles that came out of her gaping mouth was higher than normal breathing. It didn’t change the particle make-up.

As if we haven’t already talked about enough variables, there were more, including:

  • No mask versus wearing a standard surgical mask
  • Leaving gaspers (air vents) open versus keeping them closed
  • Cabin air system as is with just the 10 people onboard versus with heating blankets over temperature sensors to try to simulate the load on a full airplane and kick the system into high gear
  • Using onboard-generated air on the ground (the APU) versus using ground-provided air

A Positive (or is it Negative?) Result

So, what happened? Not much, frankly, and I mean that in the best possible way. Despite all these different configurations, the results were mostly the same. There was very little aerosol dispersion on the airplane. And why? Well, they confirmed that there were there “dominant protection factors” onboard.

  1. Frequent replacement of air (32 times per hour on the 767 and 35 times on the 777)
  2. Use of a HEPA filter during circulation to weed out particles
  3. The downward flow where air comes in at the top of the cabin and gets sucked out at the bottom

Any airplane that has these features — any modern commercial airplane with more than 50 seats today — will benefit from this. That being said, they only tested widebodies, so could the results be different on a narrowbody? Maybe, but they don’t see why that would be the case. Still, testing would need to be done to confirm.

Digging Into the Numbers

Let’s get into real numbers here. Because of the frequent replacement of the air, the particles in the cabin hung around for less than 6 minutes on average. If this happened in your house, they would hang around for 1.5 hours. When you consider that time is a component of how much virus to you take on, this is a big deal.

You can see full details in the report, but let’s just take a look at the results from a handful of tests on the 767. What you see here is the percent of the virus that was recorded by each sensor when Ruth was sitting in seat 37E at altitude. The three tests in the upper left had her wearing no mask and breathing normally. The three in the upper right had her with a mask, again breathing normally. The four in the bottom left had her coughing with no mask while the four in the bottom right had her coughing with a mask.

The red is Ruth since that’s where 100 percent of the particles start. Keep in mind that the shading is really just meant to give a visual here. That doesn’t mean those people have a chance of being infected or anything like that. It’s just highlighted to show the direction.

What we see here is that in the back of the airplane, the virus particles drifted more toward the back, but we’re still talking tiny numbers here. There is no sensor that recorded more than 0.02 percent of the particles. I should also note that if you look at other test results, those in the front of the plane saw the particles drift more forward while those in the middle saw no favoring either direction. This wasn’t due to seat types or anything like that but just airflow on the aircraft.

For the surfaces, they found that arm rests and table tops collected more than vertical surfaces like seatbacks, but the numbers were still miniscule.

What Does This Mean?

What does this mean for you? Well, they translated that, fortunately. Assuming that you need 1,000 virions to get infected, it would be nearly impossible for you to get sick. The absolute worst outcome they found was in the mid-aft section of the 777. If you had the maximum amount the sensor collected during all the tests — and if no masks were being worn — then you’d have to sit in that seat for 54 hours to get sick. This also assumes that every particle a person expels contains the virus and that every particle that reached the sensor would infect the person. Neither is likely true. Like I said, this is a worst case scenario that still is not very scary.

When you look at the average amount the sensor collected instead of the outlier maximum, nearly all of these would require sitting for 1,000 hours next to someone in order to get sick. The lowest was actually 870 hours. If people are wearing masks, as required, that means aerosols released would reduce by about 90 percent, so you’re talking an even more impossibly high number of hours to get sick.

Limitations of the Study

This is all good news so far, but do keep in mind that no study is perfect. A good study will highlight the limitations, and this is no exception in that regard.

  • This tests only one person having COVID on each flight. If you have multiple people, then chances of infection would increase. But David and Sean noted that there’s no reason to believe that you wouldn’t just add numbers together. The numbers are so small that the chance is still incredibly tiny even with multiple sick people.
  • This hasn’t been tested on a narrowbody, so in theory it could test differently. We don’t know.
  • The numbers showing 4,000 virions being shed per hour and needing 1,000 virions to get sick aren’t fixed in the real world. But considering how low the rate of dispersal is, it can vary significantly and not change the outcome.
  • They did not look at bigger particles which could be expelled especially with coughing or sneezing. Bigger particles have been shown more clearly in other studies to be trapped by masks at higher rates, so that wasn’t the focus.
  • This is about aerosol transmission, not actual even larger droplets. So if someone coughs globs of spittle all over you, or if there is fecal matter sprayed in the lav, those can all end up with different outcomes. Wear a mask.
  • There wasn’t movement in the cabin during these tests. When you go to the lav or the galley, you’re going outside the study parameters. Again, WEAR A MASK.
  • Ruth is stiff. She only stared forward, riveted by that new episode of The Unicorn, or whatever IFE consists of these days with no new movies coming out. Normal people — sorry Ruth, but you aren’t normal — move their heads, turn around, etc so the actual direction of the virus can shift.

The Takeaway

If you’ve stuck it out this long, good for you. Now you’ve reached the payoff. What does this mean for travelers? As long as you’re on an airplane with the three dominant protection factors, it’s exceedingly unlikely that you’ll get sick. Worry more about the rental car shuttle or the airport environment. But that doesn’t mean there aren’t things you can do on the airplane to make it even safer.

I should note that some of these are in the paper, but others are my interpretation based on the study data, so keep that in mind.

  • Use sanitizers to clean the areas around your seat, especially the flat surfaces like arm rests and tray tables with less focus on vertical surfaces. Even if there isn’t much virus at all, it can’t hurt.
  • If you walk around, we know less, so make sure to use your masks (as you must everywhere onboard) and use sanitizer after your return to your seat.
  • Gasp away! There was almost no difference between when people had their air vents on or off.
  • Don’t worry about where you sit. You can’t be sure where a sick person would be in relation to you anyway, and while there were minor differences in where the virus went, it wasn’t much no matter what.
  • Blocked middle seats really don’t seem to matter other than as a marketing exercise. It’s great for travelers to have an empty middle, but there doesn’t appear to be medical reason to actually continue with the practice.
  • Oh, and…wear a mask. I’m a broken record.

I’m traveling on an airplane for the first time since the pandemic began shortly, and I know this study makes me feel a whole lot more comfortable about doing so. The airlines are certainly hoping a whole lot more people feel the same way.

Recommended Posts