Once the role of airborne/aerosol transmission of COVID-19 became more recognized, lots of places starting putting HEPA filter devices into offices, classrooms, and various other locations. HEPA (High Efficiency Particulate Air) filters were initially created in the 1940s to help remove radioactive materials from air in labs and manufacturing spaces (during the development of the atomic bomb). Since then they have found common use in labs, manufacturing and other spaces were fine particles need to be controlled, and this includes removal of biological pathogens from air. Generally, a HEPA filter is one that can remove at least 99.97% of 300 nm (or 0.3 micrometre) sized particles from air that travels through it.

At first glance, 99.97% efficiency seems quite impressive and a good level of protection from bacteria and viruses. However, the reality is somewhat more complicated. The basic question is whether your HEPA device sitting in the room is significantly reducing pathogen exposure or not? Like many engineering questions, it depends on the context and here we will explore some of those factors.

First, let’s put one concern aside. HEPA is good at capturing 300 nm particles, but SARS-CoV-2 (the virus causing COVID-19) is only about 90 to 100 nm in diameter. So will HEPA capture that? The short answer is yes, because the viruses don’t normally float around in the air all alone. They are usually contained within larger aerosols discharged from people when they talk, sing, cough, sneeze, etc., and those aerosols are typically in the 500 to 10,000 nm size range (there are larger droplets too, but those typically don’t stay suspended in air for very long). So HEPA should be quite effective in removing these airborne aerosols containing viruses coming from infected people.

So what’s the problem? The key issue is that the contaminated air has to move through the filter at some reasonable rate, versus the volume of contaminated air. Think of it this way: if we put a single small HEPA filter device in the middle of an enclosed stadium (like the Toronto Rogers Centre, or the Houston NRG Astrodome), would we really expect it to benefit anyone standing hundreds of metres away? No.

So the filter efficiency is important, but so is the device throughput, i.e. the amount of air it is capable of filtering, and the volume of air in the room that needs filtering. Filter manufacturers commonly quantify this throughput using the Clean Air Delivery Rate (CADR), usually expressed as cfm (cubic feet per minute) in North America. So the question comes down to how to select a device (or several) that has a CADR that will be useful for your specific situation.

To select device(s) with suitable CADR, there are several recommendations from various organizations. The Harvard School of Public Health suggested that there should be a CADR of 100 cfm for every 250 square feet of space in a room. This is a fairly simple way to size and select your HEPA unit(s) for an individual room or office.

There are slightly more complicated ways to select device capacities, based on the concept of “Air Changes per Hour” (ACH). Essentially, ACH tells us how frequently the entire volume of air in a room is replaced with fresh/clean air. A typical home will have about 1 ACH, due to normal leakage around doors and windows, and of course much higher if the windows are open. Typical classrooms and offices might have around 3 to 6 ACH, but it entirely depends on how the ventilation (HVAC) system is working and set up. The U.S. CDC recommends at least 6 ACH for medical examination and treatment rooms, and at least 12 ACH for emergency waiting and triage rooms.

The Harvard recommendation of 100 cfm CADR for every 250 square feet is equivalent to about 3 ACH in typical rooms with ceilings that are about 9 feet high (2.7 m). So in the worst case, for a room with very poor ventilation you would achieve at least 3 ACH, which is a very good improvement. If the room already has 6 ACH coming from the ventilation system, you’ll boost it to about 9 resulting in about a 50% improvement. For heavily ventilated spaces, there is a diminishing return to adding any more HEPA treatment. Knowing what your room’s ventilation performance is like is another problem, which can really only be determined by measuring the flows from the vents.

So, identifying HEPA devices with sufficient capacity (CADR) to have an impact is the first step. Another problem is layout. Putting a single device at one end of a long room is unlikely to help much at the other end. Putting a single device in the middle is better. For larger rooms, or non-rectangular rooms, spreading several HEPA devices around is recommended, such that the total CADR of the devices meets the goal.

The final issue is related to operation of the HEPA devices, and this is an issue that I suspect makes the HEPA units almost useless in some classrooms and offices. The issue is this: the manufacturer’s cited CADR is at the top fan speed for the unit (unless they specifically indicate otherwise). The top fan speed is also the noisiest setting, and I wonder how many classrooms and offices have turned down the fan speed because of the noise? Some manufacturers will cite the air throughput at different fan speeds, but many do not. So if you turn it down, you may only be getting 10 to 20% of the CADR that you thought you would. That means you should have bought maybe five times the number of HEPA units for your room that you did, based on the Harvard recommendation or other similar ones, because you’re operating them at much lower actual CADR than what the box indicated.

I suspect that a lot of offices and classrooms are getting nowhere near the recommended HEPA air cleaning capacity, simply because of incorrect sizing of the units and/or operating them at low speeds. Possibly a lot of money has been wasted on ineffective deployment of HEPA units. It would be interesting to do a survey of HEPA unit usage and see for sure. It also illustrates that buying the cheapest HEPA unit (which tend to be the noisiest too) is not necessarily a good approach. Often the manufacturer won’t tell you what the noise levels are, or if they do it is at the lowest fan speed. You need to consider HEPA air throughput as well as noise levels that are appropriate for the specific environment. The more expensive units tend to be designed to achieve high throughput with minimal noise impact (although there will always be a bit of noise).

So, a bit of an overview of HEPA filtration and its use these days. Do you use HEPA filters? Let me know in the comments what your experience has been, or if you have any questions. I’m happy to provide advice if time and circumstances permit.