Author Archives: geek

In response to some discussions I had seen about the use of HEPA filters to help with the COVID-19 crises, I wrote some thoughts on how effective I thought HEPA might be. Several people on Twitter stated they agreed with my statements. An HVAC technician (@JSTootell) provided some thoughts that I had never even considered such as the energy requirements on the buildings where a HEPA filter is installed as HEPA requires more energy (i.e. electricity) to run than normal HVAC filters. He also said the normal air velocity is super low because if you increase the air velocity and hence get more circulation, people complain about the noise of the air through the vents.

Some others have noted that HEPA filters, on a whole HVAC system or portable units in each room, won’t hurt to which I agree. One said they should be a part of a multiple layer approach to prevent the spread of COVID-19 to which I also agree. In fact, while I did not say it, that is part of my argument, HEPA filters alone will not solve the problem of COVID-19 transmission. I want to take a step back though and discuss this from an engineering perspective.

The basic, general idea in engineering is you find out what your design specifications are, you make some calculations and draw some designs to comply with those specifications based on proven information, you throw in some safety factors, and then you build whatever it is to comply with your design and calculations. If you want to build a bridge, you need to know before hand what are the design specifications. Is it for trains, vehicles (cars and trucks), pedestrians, or something else entirely? How many of the intended type of users will be crossing the bridge daily? What is the span of the bridge? What is the height of the bridge? What type of weather will the bridge will be exposed to? There are far more questions, but that is the general idea. You can’t design a bridge until you know what you are designing.

I currently work in human health risk assessment related to exposure to hazardous chemicals. It is not the same as risk assessment related to exposure to infectious agents, but there are similarities. With hazardous chemicals, the goal is to reduce people’s exposure such that they are not at undue risk to the chemical exposure. You can’t reduce risk to zero; it is simply impossible. With chemicals that cause cancer, generally you are trying to get the risk below one in a million chance of cancer caused by exposure to that chemical. Another part of this is who is at most risk. With chemicals, the people we are generally most concerned with are children or pregnant women as they can be more susceptible to harmful effects than healthy, non-pregnant adults. The risk requirement is one of your design requirements. If a person can be exposed to 100 mg/l per day (via ingestion) or 100 mg/g (via inhalation) of a certain chemical and not be above one in a million risk of cancer, then you have to figure out what needs to be done at a contaminated site or with contaminated drinking water to get their exposure (and thus risk level) below that number. This could mean filtering water or removing topsoil at site (to avoid incidental ingestion of contaminated soil or to avoid breathing in soil particles). What kind of treatment and how much treatment is needed to get below that concentration? That is one of your design requirements. Similarly the design and operation of water treatment plants is based on cleaning water such that the water has less than some amount of a contaminant before it is sent via pipes to the customers. Design requirements from water treatment plants is generally based less on risk calculations and more on state and federal requirements for contaminant levels in drinking water. These federal requirements are called Maximum Contaminant Levels. Water treatment plants must meet these requirements, and they are designed to prevent the people who drink that water from getting sick from microorganisms or chemicals in the water.

This leads me to designing a HVAC system with HEPA filters or the use of portable HEPA filters in buildings to protect against COVID-19. In order to design a system, you have to know the design requirements. It is absolutely fine to say you want to reduce virus particles in the air and reduce transmission, but that is not a design requirement. Reduce is a vague, qualitative word. Engineering requires quantitative requirements. If you only want to reduce particles in the air, then you will only reduce the risk by an unknown amount with no clarity on if that reduction is an acceptable amount to the occupants of the building. Reducing risk could mean that instead of 30% of the occupants of a building getting sick, only 20% do. I personally don’t find that to be an acceptable reduction. A design requirement is based on what concentration of virus particles can be in the air and no person gets sick from COVID-19. Perhaps your requirement would not be that stringent, perhaps you would be ok with one in a thousand people getting sick from COVID-19 based on the design. The design requirement can be based on people wearing a mask or not wearing a mask. Maybe with everyone wearing a mask indoors, they can be exposed to 10 virus particles per hour, but without a mask, they can only be exposed to 2 virus particles per hour. This is where infectious disease experts are needed to provide information as to the pathogenicity and virulence of the pathogen, which in this case is COVID-19. An engineer designing a HVAC or some other filtering system for a building is not the person to decide what those design requirements are. They need the infectious disease experts to state what concentration of a pathogen a person can be exposed to without getting infected. The concentration may be zero. The problem at this point is I don’t think we know how much COVID-19 a person can be exposed to without getting sick. Thus, if we don’t know how much COVID-19 a person can be exposed to without getting sick, how can we possibly design a system to prevent a person from getting sick.

I can already hear arguments that we just need to do something. We need to accept some risk but do some things to reduce risk, so we can get things back to normal. I don’t think most business owners are going to be willing to spend a non-negligible amount of money on some design that will simply reduce risk to an unknown and unproven amount. For a place of employment or a school, is it reasonable to ask people, especially children, to return to a building with an unknown risk if a system has been put in place that reduces the risk an unknown amount? How much money should employers and educational boards spend to reduce risk an unknown amount? If you are willing to accept some risk, then why spend money on something that may reduce risk by some unknown amount? Everyone is already spending money on masks, gloves, hand sanitizer, etc. which at least has been proven to reduce risk, but not eliminate it, by a reasonable degree from a cost benefit perspective. I spent $20 or something on two reusable cotton masks that I wash after use. That is a very reasonable cost benefit amount from my perspective even though I can’t calculate the risk reduction of the mask. How much money is reasonable to invest in either a whole system HEPA filter or portable HEPA filters when the risk reduction is unknown? An extremely quick internet search provides options for portable HEPA filters from $200 to $1200. Should schools buy one per classroom, even at the low price end, when there is no data to show they would reduce risk at all? The point is, reducing risk is good, but if you going to invest money to reduce the risk, it would be prudent to determine how much the risk is actually going to be reduced before you do it.