Air Filtration in the Time of COVID-19
Why we got air purifiers, how they work, and how we selected ours.
By Anna Kramer
In this post we want to share with you what we learned about air purification after we did our deep dive to equip the office and treatment rooms at Earth + Sky. This information will help you understand why we feel good about our set-up and will, hopefully, allow you to also feel good about it, too. It will also be helpful if you choose to buy your own air purifier. Below is information on the role of air purification with respect to the coronavirus, general information about how air filters work, and parameters and formulas to help select an air purifier that is correctly sized for a specific space. The issue of airborne transmission of the virus i.e. transmission via droplets small enough to float in the air and not drop to the ground within minutes, is an evolving scientific discussion. While the discussion proceeds, we have chosen to assume that airborne transmission is possible, and we are implementing what we think to be the best precautions available to us.
Please, note that, while we have several science and technical degrees among us, we are not air filtration experts, atmospheric chemists, infectious disease doctors, or epidemiologists, so we cannot guarantee the accuracy of our interpretations of the information we have gathered, nor the accuracy of the sources themselves. We have educated ourselves in good faith and we hope to be of genuine service to all, but please, retain agency and critical thinking, and please do tell us if we have made any errors or missed anything important. We have included in-text links to sites for units conversion or for verifying an air purifier’s compliance with safety standards. Links to our sources are provided as endnotes and are indicated by [ ] in the text.
Finally, while I may mention some air purifier companies by name, neither I nor Earth + Sky are suggesting that these are the right ones for you. It really depends on your specific needs, for example, the size of the rooms to be filtered, acoustics and sound needs for long-term use, the physical size of the unit, and additional pollutant concerns besides the virus.
I. What is the role of air purification in protecting against coronavirus?
The National Institute for Occupational Safety and Health (NIOSH), a division of the Center for Disease Control (CDC), offers a hierarchy of controls [1] framework for dealing with occupational hazards. The hierarchy of controls is an inverted pyramid of interventions, arranged with the potentially most effective and efficient intervention at the top. When applied to the coronavirus, interventions such as air filtration and ventilation rank above personal protective equipment (PPE) with respect to efficiency. In fact, PPE is at the very bottom of the pyramid (indeed!). This does not mean that PPE is not effective. It means that it is less efficient because it only affects the individual wearing it and the people interacting with that individual, whereas something like air filtration would affect everyone in the space. A viable solution involves as many layers of the pyramid as are needed in order to manage the hazard. As we began to research reopening our office, we noted that the public health conversation primarily focused on masks and sanitizing surfaces, whereas here was a government agency (NIOSH) clearly stating that ventilation and air purification were important, possibly further reaching interventions. The realization that air filtration was the highest level of the pyramid that we could hope to influence started us off on our exploration of air purifiers.
II. How do air purifiers work?
Air purifiers use varying technologies. At Earth + Sky, we chose to stick to the HEPA-filter-plus-fan method, in which a fan pulls air from the room, through the filter, and back out into the room. Other air purification techniques, such as ultraviolet light (UV-C), ionizing, PECO, and high heat units (disinfection through high heat) have not been tested as much as the HEPA and some have potential for negative effects, such as the unwanted production of ozone by ionizers [2], and the potential damage to human tissue from UV-C light [3]. If you do wish to use a purifier’s UV or ionizing function, make sure that the machine has been certified as safe by the California Air Resources Board and make sure to buy a unit where this function can be turned on/off separately.
A HEPA filter is usually made by folding a sheet of very small interlaced glass fibers into a complex pleated shape. There are various grades of filters [4], but most of the units we looked at have HEPA 13 filters, defined as able to capture 99.97% of all particles larger than 0.3 microns (= μm = micrometers) in diameter (= 0.000012 inches). If a filter cannot do this, then it cannot call itself a HEPA 13 filter. For reference, the coronavirus diameter is approximately 0.125 μm [5], bacteria are 0.3 – 60 μm, household dust is 0.05- 1000 μm, mold spores 3-40 μm, pollen 10-1000 μm [6] (see Figure 1 to better visualize relative sizes, but note that the range values are not the same across these two sources, so consider them all as approximations). Please, note that filters labeled “HEPA-like” or “HEPA style” are not true HEPA filters, so you do not know how well they filter. Also, do not be dazzled by “military HEPA”: HEPA technology was originally developed by the military (for the Manhattan Project), as were many other things that we now use in the everyday.
Figure 1: Airborne particulate size chart. (I am assuming the x-axis is in microns; note that the scale is not linear, but logarithmic). Jisaac9 (Own work) / CC BY (https://creativecommons.org/licenses/by/3.0). https://upload.wikimedia.org/wikipedia/commons/4/47/Airborne-particulate-size-chart.jpg
Despite the definition, which seems to say nothing about filtration of particles smaller than 0.3 μm, HEPA filters are actually very efficient at capturing particles below 0.1 μm, but the mechanism varies for different sized particles and the relationship of percent captured to particle diameter is complex. The information on this comes from a 2016 NASA analysis of HEPA filter efficiency across particle size by Perry et al [7]. All of the figures in the paper are very instructive and evocative, so do take a look when you have a chance.
For very large diameters the particle is simply too big to pass between the fibers and is strained (see Figure 2). In order to prevent many very large particles, like pet hair, from quickly clogging the HEPA, most air purifiers come with pre-filters similar to the kind you may have seen on a window air conditioning unit. The pre-filter is often washable or separately replaceable, which prolongs the life of the HEPA filter. As an aside, we plan to wear PPE when replacing our filters and pre-filters, even if the manufacturer instructions say nothing about it.
Figure 2: The four filtration mechanisms of a HEPA filter: A-straining, B-inertia impaction, C-interception, and D-diffusion. From the November 1997 Issue of "Infection Control Today" by Gary D. Messina, via RGF BioControls, http://www.biologicalcontrols.com/800400.shtml#hepa
For somewhat smaller particles (certain types of dust, mold, and pollen that are larger than 10 μm), picture a large truck that slams into an obstacle because it was going fast and didn’t see the obstacle soon enough to be able to swerve to avoid collision. This is referred to as inertial impaction because the particle has too much inertia (from the combination of its mass and the speed of the air flow that is carrying it) to alter its direction and get around the fiber, so it slams into it and gets stuck there.
Still smaller particles of less than 10 μm diameter are like a smaller, more maneuverable car that is going at the same speed and does manage to swerve, but grazes the side of the obstacle so strongly that it gets stopped. Such particles can follow the air flow around the fiber, but are still so large that they stay very close to the fiber (within one radius of the particle) as they travel around it, and end up sticking to it because of various adhesion forces that operate at these small distances. This is called interception.
The smallest particles (less than 1 μm diameter) are like a bunch of tiny, out-of-control bumper cars. They’re constantly bumping into each other and ricocheting off of each other, so that sooner or later they, too, collide with that same obstacle. These particles are so small that their path is less influenced by the air stream and more influenced by collisions with other particles, i.e. diffusion or Brownian motion. The randomness of the resulting pathway allows these particles to also come in contact with filter fibers and get stuck.
Based on the above discussion and on Figure 3, it looks to me like diffusion would be the primary mechanism of filtering out bare virus particles, and a combination of impaction and interception would work on aerosols (defined currently as droplets of diameter less than 5 μm). Also notice that there is a valley in the graph for particles approximately 0.2-0.3 μm, where the filtration efficiency is 99.97%--this is the number that defines the HEPA standard. What we are not usually told and what this paper shows is that the filtration rate is even higher for the rest of the particle diameter range (good news!). Based on this, a HEPA filter’s efficiency would be more accurately described as “able to filter at least 99.97% of all particles in the range 0.01 – 10 μm.” It is not clear to me why that is not the current definition. The Environmental Protection Agency (EPA), in any case, concurs: “The diameter specification of 0.3 microns responds to the worst case; the most penetrating particle size (MPPS). Particles that are larger or smaller are trapped with even higher efficiency. Using the worst case particle size results in the worst case efficiency rating (i.e. 99.97% or better for all particle sizes).”[8]
Figure 3. Filter efficiency as a function of particle diameter. (Note that the x-axis scale is, again, logarithmic). From Perry et al, 2016, Submicron and Nanoparticulate Matter Removal by HEPA-Rated Media Filters and Packed Beds of Granular Materials, NASA. https://ntrs.nasa.gov/citations/20170005166
The smallest particles of all are gases or Volatile Organic Compounds (VOCs), and include smoke. The HEPA filter cannot capture these. VOCs are captured by layers of activated carbon, usually placed after the HEPA filter. The more carbon, the more VOCs it can capture. Since our current concern at Earth + Sky is virus, we didn’t give this too much time. If this is an important concern for you, I suggest you look at Austin Air, or any brand that proudly advertises the high carbon content of its filters in pounds (lbs).
Overall, there are two competing factors in the filtration mechanism. On the one hand, the more powerful the fan, the more air it can propel through the filter and the faster it can process a room. On the other hand, the more slowly the air passes through the filter, the more thoroughly it will be filtered. In addition, higher fan speeds favor impaction mechanisms i.e. the capture of larger particles, while lower fan speeds favor diffusion i.e. the capture of smaller particles. As a result, we plan to regularly alternate fan settings throughout the day. Our intent is to balance getting all of the room’s air through the filter multiple times per hour with optimizing the filtering of various ranges of particle size.
There are additional factors of air flow dynamics that we are not able to account for, such as the effect of the specific configuration of our treatment rooms, but we have done the best we can with what we have understood so far. For potentially more information on this we plan to look into a visually splendid document called the Manual of Physical Distancing that summarizes a lot of the research to-date, and proposes case studies of how to arrange spaces to be in compliance with what is known so far [9].
This brings us to the question: how often does all of the room’s air need to pass through the filter in order to potentially offer some protection against coronavirus?
III. How quickly should air be purified to help protect against coronavirus?
For this section, I relied on a May 20, 2020 blog post by Erin Bromage, comparative immunologist at the University of Massachusetts Dartmouth [10]. He calculates an estimated number of virus particles released into the air by coughing, sneezing, talking, and breathing, and the amount of time it would take to inhale enough virus to become infected under each scenario. There are MANY assumptions and estimates in these calculations which are yet to be confirmed experimentally and which make the results VERY approximate, but they provide a starting point. We won’t look at sneezing and coughing, which release a lot of particles. We will look at nose breathing and speaking, which is what we would mostly be doing while giving and receiving a massage.
Based on the blog post’s estimates, as few as 1000 SARS-CoV2 virus particles might be needed to become infected. Using influenza as a guide, breathing releases an approximate 33 virus particles per minute (ppm). Assuming a person is inhaling and retaining every virus being exhaled by an infected individual (very unlikely), it would take 30 mins for a person to inhale 1000 viruses (1000p / 33ppm = 30 mins). For speaking, the estimate is 330 ppm, which, again assuming inhalation of every particle, would require only 3 mins to reach 1000 particles (1000p / 330ppm = 3 mins). Note that these calculations are for unmasked exposure.
These very approximate calculations suggest that all of the air in a room should be purified every 3 mins, i.e. 20 times per hour. However, detailed simulations of ventilation in hospital settings seem to indicate that 9 air exchanges per hour (ACH), i.e. every 6.67 mins, adequately minimizes aerosoled virus particles and maximizes energy efficiency [11]. In the context of massage, if we minimize talking, this rate of filtration is more than adequate by all the above calculations, and is what we have chosen to aim for at Earth + Sky.
IV. How to select an air purifier that is correctly sized for the space?
In order to determine whether an air purifier’s fan can pull the air through fast enough to create the needed air exchanges per hour (ACH), you need to know its Clean Air Delivery Rate (CADR). The volume of air a filter can pull through itself is given either in units of cubic feet per minute (CFM) or cubic meters per hour (CMH). Here is a site for converting between the two systems. The manufacturer must tell you the units used when reporting an air flow rate. If they don’t, ask them. If their customer service does not understand your question, consider buying from another company.
Technically, the CADR is the volume of air the fan can pull through the filter multiplied by the efficiency of the filter [12]. This is why you will sometimes see three numbers given for the CADR: one each for dust, pollen, and smoke. As we saw above, the efficiency does vary according to particle size. However, given that it should be close to 99% for a proper HEPA filter across the entire particle range, in our calculations we assume that the efficiency is 100% and treat the air flow rate as equivalent to the CADR. I will refer to CFM instead of CADR in the formulas and calculations.
The CADR is usually given only for the highest fan setting, although some companies do provide it for all fan settings for some of their models. This is problematic since you need to know what your ACH is at all settings. I’ve had varying success obtaining this information from manufacturers: Coway provided it for the Airmega 200M that we are using, and Austin Air provides it for all their models. Some companies have had to contact their engineering headquarters, which can take a month or more. I encourage you to ask this question of all the companies you interview, in the hopes that the industry will eventually comply.
Many companies claim that a unit clears X sq ft, but that information is incomplete because it does not tell you how many times per hour it clears the air in that X sq ft. Some companies will be more forthcoming and say that this represents two or four ACHs, which is too low for our purposes. They also do not tell you the ceiling height, although it seems they are mostly assuming an 8 ft high ceiling. It is therefore much better to obtain the CFM rate, so you can calculate the ACH for any space.
V. Noise
Noise is important for us since we want to create a peaceful atmosphere for our clients and for ourselves. Noise levels in decibels (dB) are mostly listed by the sellers, either as a range or for each setting. We prefer to have the noise level for each setting, and request that information. Noise tolerance varies, but we have found anything above 50 dB to be uncomfortable for our purposes.
VI. Energy Costs
The amount of electricity the air purifier consumes at its highest fan setting is usually provided in Watts (W). Units with fans capable of 200- 300 CFM seem to require about 40 - 80W. The U.S. Department of Energy helps you calculate how much you will pay using a state-by-state average estimate of price per kilowatt hour: Your electricity bill will of course give you a more precise figure.
VII. Our purifiers, calculations, and plans at Earth + Sky.
We have placed a Coway Airmega 200M (aka AP-1512HH Mighty) in each treatment room, and an Austin Air Junior in the waiting area. Below are our calculations for these units. This way we (and you) always know what ACH we are providing.
Example calculation:
Treatment room 1 (front): 10.5 ft ceiling, 93.44 sq ft (or ft2)
Coway air purifier’s CFM for the High fan setting: 247.3 ft3/min
The rest of the calculations are summarized in the table below. The waiting area includes the bathroom, but does not include the skylight space in the calculation, so the actual ACH for the waiting area is slightly lower. As you can see below, the air in our treatment rooms is cycled through 13-15 times per hour, or about every 4 to 5 minutes when the units are on the high setting, and every 7-9 minutes when the units are on medium.
VIII. Final Thoughts
After all this research, we are confident that we have put together a viable air purification regimen that will positively contribute to supporting the health of everyone who comes into the space. We note that the public discussion on air purification is there, albeit still on the sidelines, with editorials asking why no one is discussing it, health practitioners blogging about it, articles about the need for better ventilation in commercial indoor spaces, and an article about the very fast air flow rate (18 ACH) in the NYC subway [13] (unfortunately, the filter used is rated MERV-7, meaning it captures less than 20% of particles 0.3 – 3.0 μm, and 50-69% of particles 3 - 10μm [14], which is far inferior to a HEPA). As the science around SARS-CoV-2 and the public discussion of the role of air purification and ventilation evolve, we will adapt our regimen to reflect our most current understanding.
I, personally, would encourage you to use the European CDC website [15] to keep abreast of COVID-19 related scientific developments vs. the US CDC. The European CDC provides detailed summaries of studies and is better organized. It does have some elements you may not understand immediately, and some things for which you need a bit of familiarity with biostatistics, but the information is all there and the primary sources are easily accessible (references!).
For those interested in pursuing the links provided in this post and in its endnotes, I suggest saving them and their text to your hard drive (with proper citation) because the internet is an ever-changing forum and some things just disappear after a while.
Finally, if you are a non-union worker who wishes to negotiate with your employer about air filtration or any other safety measure you feel is necessary for you, please, be aware that you are legally protected by the National Labor Relations Act [16] only if you and your fellow workers make such requests as a group, or if you act as a representative of such a group. In this way your activities fall under "concerted activities." There may also be certain provisions of the Americans With Disabilities Act that are applicable if you have predisposing conditions. This is information I gleaned from a webinar by the excellent Cornell University’s Worker Institute [17].
We hope this information has been of reassurance, of interest, and perhaps of help to you in your own air purification arrangements. We seek to support and empower you and we look forward to seeing you at Earth + Sky.
Breathe well,
Anna
Acknowledgements:
My heartfelt thanks go to Jenny Swadosh from the New School for sharing amazing sources of information. May you all be blessed, as I am, with the friendship of an archivist and librarian.
Many thanks to Julie Tudor for her post in the private Facebook group “Massage, Health Practitioners and COVID-19”. It was a super quick and thorough orientation to air purifiers that saved me tons of time—community in action!
Endnotes:
1 https://www.cdc.gov/niosh/topics/hierarchy/default.html discusses of cost vs. effectiveness of each layer of the hierarchy of controls. https://ehs.cornell.edu/campus-health-safety/occupational-health/covid-19/covid-19-hierarchy-controls applies the hierarchy of controls specifically to the coronavirus.
2 https://www.epa.gov/indoor-air-quality-iaq/ozone-generators-are-sold-air-cleaners
3 http://www.uvresources.com/blog/uv-c-lamps-staying-safe
4 https://www.homedetoxing.com/how-are-hepa-filters-rated/
5 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4369385/ Coronaviruses: An Overview of Their Replication and Pathogenesis
6 https://www.oransi.com/page/particle-size nice graph of relative sizes of common household particles ; https://www.engineeringtoolbox.com/particle-sizes-d_934.html more detailed list of particle diameters.
7 https://ntrs.nasa.gov/citations/20170005166 see Fig 1 and the paragraph following it for filtration mechanisms, and Fig 3 and the preceding paragraph for filter efficiency as a function of particle diameter. Article reference was found on https://www.nytimes.com/wirecutter/blog/can-hepa-air-purifiers-capture-coronavirus/
8 https://www.epa.gov/indoor-air-quality-iaq/what-merv-rating-1
9 https://issuu.com/djlewis72/docs/200622_manualphysicaldistancing_draft
10 https://www.erinbromage.com/post/the-risks-know-them-avoid-them
11 https://doi.org/10.1177%2F1420326X16631596
12 https://www.oransi.com/page/cadr
13 https://www.nytimes.com/interactive/2020/08/10/nyregion/nyc-subway-coronavirus.html
14 https://www.lakeair.com/merv-rating-explanation/
15 https://www.ecdc.europa.eu/en/covid-19/latest-evidence
16 https://www.nlrb.gov/about-nlrb/rights-we-protect/your-rights/employee-rights
17 Worker Institute at Cornell University’s School of Industrial and Labor Relations, June 24, 2020 webinar entitled “Securing a Safe Workplace and Workers Rights.” This is where I first found out about the hierarchy of controls, NIOSH, and the National Labor Relations Act.
