Impact of Breathing Your Own Air When Wearing COVID-19 Mask, an Experiment Showing

Important: This is NOT an anti-masking post. It is a scientific experiment to show facts regarding what a COVID-19 mask wearer might have to deal with in partly breathing their own breath back in with every breath, so the wearer can be informed and decide when they would want to do it and not within the law, not against the law. The decision is not to obey or disobey the law, but rather to obey the law and do something requiring mask wearing, or abstain from it because the impact isn’t worth the activity. For example, go shopping in a mall with a mask (required where I am), or not and do online shopping or window shopping from the street, NOT go shopping in a mall without a mask.

In wearing a mask for COVID-19, you protect others from your droplets. That much is proven and clear. What’s also clear is that you breathe back in some of your own air trapped between the mask and your face with every breath. You can probably smell it. But what is in this breath and how might it impact you? Those are two broad questions without singular answers for any group. However, the general nature of answers to both may have enough similarities that an experiment might be able to hint at some common factors for most people. We are, after all, all humans who breathe the same air, more or less aside from quality where we each are, in and out to survive.


I have an Awair air quality monitor, edition 2. It is no longer on sale due to more advanced models of late, but I have found it to be a pretty decent monitor so I used it to test the air quality of the air I was breathing back in while wearing a mask, to compare to the ambient indoor air of my apartment. The Awair monitor tracked:

  • Temperature (in degrees C or F, I go with C but it doesn’t matter much here)
  • % Humidity
  • CO2 levels in parts per million or ppm
  • Volatile organic compounds or VOCs in parts per billion or ppb
  • Particulate matter 2.5 microns and larger in micrograms per cubic meter (ug/m3)
  • An aggregate overall score from 1 to 100 based on actual not proxy readings of the above, with 0-59 being poor, 60-79 being fair, and 80-100 being good.


On a Saturday night, last night as I write this, I did an experiment to see the difference between air I was breathing at home, and air from under the mask. Here’s how I did it.


Below is the air I was breathing late Saturday night. 85 score. A little hot, with some CO2 (600 range ppm) from the slightly stuffy air holding on to my CO2 output all afternoon and evening in my apartment. A little VOC (350 range ppb) from my Instant Pot cooking aroma of supper and Sunday’s lunch. Nice smelling, but probably not ideal to be breathing all the time. 🙂


Here’s what I used to do my experiment.

  • Awair Edition 2 air quality monitor (with app on phone taking results pics), faced up on a cup to make it easier to take pics
  • One government issued mask
  • Two bendable straws to channel air between face and mask to air quality monitor as I breathed (see next pic for assembled set up)


How I got a representative sample of air I breathed in, and not just my pure exhalation, was to put the two straws into my mask space from above, and beside my nostrils. When I breathe out, I should generally displace the air in the pocket first (what would have been there partly for the last breath along with some external air drawn in so it’s actually a cleaner sample). I’m not giving the straws mostly what I breathe out, in other words. Also, in breathing in, some of the airflow going into my nostrils may get into the straws to be pushed through with the next breath out. It’s a reasonably fair sampling without cutting a hole in my nose and sticking in a straw! I’m not willing to go THAT far for citizen science! lol

The far end of the straw is placed hear the air intake of the air quality monitor so as not to dilute the air sample coming out, as much as can be helped. The sample to the monitor is probably slightly diluted than what is coming out, for interpretation of results, but I’ll ignore that and just take results as they are. There is leeway for error in these steps but I think it’s fair to say they’re not all one way, and not great.


This is the overall table of results over 5 minutes of breathing with a mask on tight (preferred for effectiveness), loose (neither acceptable nor good), and medium (acceptable). Note the incredibly high numbers for CO2 (maxxed out at 100000 ppm at one point) and VOCs, to be discussed later. The overall scores aren’t too devastating, but not exactly great, either! There’s an impact that can’t be automatically ignored, is the message. Whether you ultimately ignore it or not is another matter, but it should be worthy of consideration rather than dismissed as insignificant. A pic is available later with each row of results, as well as discussion, but this is the summary of the outcomes.


This reference table is the international standard for CO2 levels for healthy and unhealthy exposure to CO2. Note the range of ppm values categories and consequences for many, if not most people.

VOCs are another matter since that is much more complicated with literally thousands of possible compounds, most of which have some degree of harm, albeit little, from prolonged (long duration at one time) and/or chronic (repeated times, possibly constantly) exposure. The general rule, though, is that higher concentrations are worse than lower concentrations. But for the complexity involved, I will just keep it to that as there are no tables to show anything of any meaning.


Within 15 seconds of strapping on a mask and pulling it tight to my face, you can see the CO2 and VOC levels spike already. Not surprising since it’s a relatively small volume between your face and mask. One breath is more than enough to fill it. The saving grace is that there is some exchange with the air around the mask since a tight mask didn’t mean an air proof mask, just one that felt pretty snug on my head and face. The cloth mask is also porous for some exchange through the double fabric layers.


And within half a minute, CO2 levels were beyond the 5000 ppm limit for prolonged exposure at work over eight hours. Keep in mind that is for eight hours. However, given these readings are well over 5000 ppm, I’m not sure what the acceptable duration for exposure should be. I just know it isn’t healthy to be breathing this stuff for a long time. Just look at all the potential symptoms for concentrations half this number on the list a few pictures above!


And within a minute of nasal breathing within a well-strapped on mask, CO2 levels were almost double the internationally allowable limit for eight hours of exposure on the job! You should NOT interpret that as being OK to do for four hours, though. It’s not a linear situation like that. Increase the amount to maybe 8 times this and you might be dead soon rather than being able to get through it for an hour with the same outcome as eight hours for eight times lower concentration!


With enough sampling of nasal breathing, I then went to mouth breathing, where I could let more air out in a given period of time, and where there would be more volatile organic compounds than through the nose due to all the mouth bacteria and such that can give people halitosis. Not surprisingly, both CO2 and VOCs went up. What may be surprising was by how much. The CO2 maxxed out at 10000 ppm as for this Awair Edition 2 monitor, which is completely fair given the prolonged exposure limit for work is only 5000 ppm! VOCs started to approach doubling the previous levels… and I didn’t even have halitosis! I ate a simple dinner of steamed jasmine rice and steamed yu choi vegetable, dipped in soya sauce, and rinsed my mouth with water every 30 minutes for 3 hours before this experiment. I didn’t want to use toothpaste because it comes with its own VOCs, even if sort of good ones for fresh breath, at least. Besides, if you wore the mask throughout the day, you wouldn’t be breathing in freshly brushed teeth breath for much of it. but wow! Holy cow readings, man! Do I really want to subject myself to this air for prolonged periods unless I had to?


I went back to nasal breathing just to see those results again compared to the mouth breathing. CO2 and VOC levels calmed down a bit but were still generally way over healthy limits for prolonged exposure.


And back to mouth breathing again for another look at those readings. Boom! Took no time for things to change and the overall score to match the previous period of mouth breathing at a miserable 43 out of 100 for overall air quality being breathed in. The prolonged tight mask wearing also built up humidity from droplets kept in that I now had a damp atmosphere in the space between my face and mask, which is going to be terrible for when the weather gets cooler.


With reproducible mouth and nasal breathing on a tightly strapped mask sampling done, I then opted to go to a loose mask before 3 minutes, still breathing through my mouth as before. This is the sort of thing someone might do to try and cope with the bad air they’re inhaling and not processing well, but it wouldn’t be very effective at protecting others as the cloth masks are meant to do. Yet, as you can see, the readings are still high for CO2 and VOCs. However, a significant drop from not keeping in a lot of droplets took the humidity down to a good level of 47% to help the overall score be just on the Fair side of Fair/Poor which happens at a score of 60. In other words, even just barely having a mask on doesn’t leave you in a great place for the air you’re breathing back in from yourself…. and you’re not helping people much if your droplets were getting out significantly more than before.


Without a nasal breathing reading for a loose mask, my next task was to get that, which I did at 3.5 minutes. That improved things drastically with a big drop in CO2, probably in part aid with more time of the mask being loose rather than just nasal breathing compared to mouth breathing. CO2 values were now barely within the work eight hour limit, but that’s hardly anything to be happy about since limits aren’t meant to be approached or broken, ideally. As for the VOCs, the drop was pretty much all nasal versus mouth breathing since the VOCs from the mouth are significantly more than through the nose. These changes gave a slightly improved overall score of 64 out of 100, which is still not far into the Fair zone from Poor with the dividing line at 60.


So now what I had left to do was to get the mask on at some practical level of tightness where it should reasonably do its job to keep enough droplets from getting out, but without necessarily trapping so much of my own breath that I’d be inhaling most of every last breath back in. Where would that leave me, closer to the tight mask or the loose mask? Turns out, it was almost in the middle, at an overall score of 54 compared to 43 low and 63 high, with all three of humidity, CO2, and VOCs contributing to the score rather than any one overwhelming value, which is good so one factor doesn’t control acceptable exposure. Still, the overall air quality is on the Poor side of the Fair / Poor threshold, and that’s for nasal breathing.


Breathing through the mouth again at 4.5 minutes, the score went up because my mouth breathing was able to blow open the mask just a tad more than my nasal breathing was able to. The increase in overall score was due to the drop in humidity from better ventilation between the mask and my face. Maybe I should have had it on a tad tighter, but I think you know what that does to the readings. Still, there was enough CO2 coming out to be at just the limit for workplace eight hour exposure, which is not something you’d want to gamble being around for your safety. VOCs were also a little bit higher, as should have been expected.


I ended the breathing test part by just leaving the mouth breathing for another 30s to see what a slightly longer period of time would do to the readings. Turns out, it didn’t change much, and didn’t change the overall score. It doesn’t take long to establish what equilibrium there would be in that small space between the mask and one’s face, which was why I was also able to rifle through all these settings on 30s turn overs. This 57 reading is probably what I can expect to be the prolonged exposure I would be getting wearing the mask for hours or much of a day, with CO2 being above the international work limit by over 1000 ppm (20% more), and VOCs also fairly high. For those reasons, really, anything I don’t really have to partake in where I’d need to have a mask on, I will take a pass!


With the experiment done, it only took 5 minutes, probably quite a bit less, actually, for my air quality monitor to return to its reading from 10 minutes before. I only have a reading for 5 minutes after getting some water on my face to rid of some of the haziness in my head from breathing all that crap for just 5 minutes, and to grab a bite to take back to my desk. I had only been used to doing the medium to loose mask wearing to this point, and for not much more than 10 minutes at a time from how I’ve managed my life to avoid situations requiring prolonged mask wearing.


Finally, some averages over that 5 minute experiment. As you can see below, average overall score was 58. That was a mix of all those conditions of mask tightness and breathing through nose and mouth, but it was pretty much the same as middle ground readings for the last minute that was 57.


And what did that overall score generally entailed? Well, for temperature, it was slightly warmer than the heat wave going on around me now. That breath temperature should be fairly consistent so without the heat wave, the gap between ambient air and mask air would be even greater since heat wave isn’t exactly ideal for ambient air quality.


As for relative humidity, it was 58%, on average, during the 5 minute experiment. Unless you had a really humid or rainy day outside with the windows open, or it was somehow humid a lot in your place, your breath will probably be more humid than your space. However, the humidity readings seen should be consistent for other times unless you had a cold or runny nose or something, in which case there’d be a LOT more droplets coming out in your breathing. The midday spike was from a long shower after a tough morning workout.


As for CO2, that was a crazy 6500 ppm range for the 5 minute experiment. As shown, you’re going to be hitting that 5000 ppm max limit allowed for workplace safety on an eight hour shift when you wear any decent fitting mask and breathing back in your own air. I wear it when I must, but otherwise, I will take a pass as much as possible by altering my life to avoid such places and activities requiring mask wearing, rather than disobeying the law because this isn’t about disobeying the law, just making informed decisions within the law.


As for VOCs, the average was a fairly high1800 ppb range value. This one could really change a lot, but likely higher rather than lower because I had a pretty clean mouth for the experiment as mentioned before of what I had eaten and how often I had rinsed my mouth prior to starting. If you ate and drank throughout the day, and didn’t brush your teeth afterward, and had spicy or really aromatic foods, you could be in for a much higher score with much grosser breath! Enjoy your halitosis! The peak earlier in the day was from almost burnt toast with chia in it. The rise in the evening was due to cooking of supper and next day’s lunch, especially from the Instant Pot and its lovely steamed food pressure released.


Finally, the particulate matter 2.5 did not change much as breaths don’t usually come with dust. The spikes seen were a fabric stash reorganization in the early afternoon. I can’t recall what might have happened in the evening to cause a slight peak there, but its slightness was probably why it was not memorable.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.