Post #810: An air-curtain approach to indoor restaurant safety in the coronavirus pandemic

Posted on September 12, 2020

Edit:  See Post #854.  The simple method that I describe below was recommended by experts for use in the Vice Presidential debate.  Engineers refer to this as a “Corsi box”, after the air quality expert who first advocated for its use.

Original post follows.

The CDC has finally started doing some epidemiology that will help ordinary citizens to judge the risks of certain activities.  And the very first thing their research highlighted was the risk from dining in restaurants.  In their small-scale study of a sample of individuals with COVID-like symptoms,  persons who tested positive for COVID-19 were twice as likely to have gone to restaurants in the past two weeks, compared to those who tested negative.

I started to write up my own analysis of this issue of risk.  But I got sidetracked by an email discussion about the coming crisis that restaurants in the Town of Vienna are likely to face, if COVID-19 gets any worse, or if the economy gets any worse.  We don’t need any more analysis.  We need a solution that would plausibly increase the safety of indoor dining.

So, instead of just analyzing the situation, I’m going to offer you my cheap solution for safer indoor dining.  In a nutshell, use ceiling-mounted box fans, with high-end air filters, to bathe each table in its own individual stream or “curtain” of clean air.  And to sweep any virus-laden air down to floor level, for further filtering and recirculation.

Detail follows.


The near-field problem.

It’s no news that restaurants are facing tough times.   Right now, the reduction in business is tempered somewhat by the presence of outdoor dining.  If patrons feel unsafe seated indoors, they can be served outdoors.  Most places — Town of Vienna included — have made some accommodate to allowing outdoor seating where it would not otherwise have been permitted, in light of the pandemic.

But as we approach winter, those restaurants are going to lose the opportunity for outdoor seating.  That, along with any resurgence of coronavirus, or any further deterioration in the economic climate, is going to make hard times even tougher.

Many restaurants are using a combination of intense surface cleaning and air treatment, to try to convince customers that they are safe.

Intense efforts aimed at disinfecting surfaces are more-or-less pure theater, as the CDC has now said that “fomite transmission” of COVID-19 (getting infected via touching inanimate objects) is almost unheard-of.   “Deep disinfection” cleanings may make customers feel better, but they don’t actually do anything material from a public health standpoint.  See Post #766, Hygiene Theater, for some details and references on that.  These intense disinfection efforts are more-or-less completely unnecessary.

For air treatment, many restaurants are turning to UV-C systems.  Most notably, local chain Silver Diner has installed UV-C.  My considered opinion is that these, too, are mostly theater.  While it is true that UV-C light can kill viruses, it takes either a very long exposure, or a very high intensity, and I don’t think that cheap off-the-shelf systems provide either. I think the equipment you need to achieve (e.g.) 90% kill rate in a single pass, and high rate of air exchange, is exotic, expensive, and frankly dangerous to be around.

But unlike surface cleaning, there is a true need for anything that will prevent aerosol transmission of COVID-19 in restaurants.  Unlike surface cleaning, restaurants installing air-cleaning systems are attempting to address a true public health need.  I just don’t think that off-the-shelf cheap UV-C systems actually do much to protect the public health.

In particular, neither UV-C systems nor other forms of enhanced whole-room air filtration deal with the “near field” travel of the virus.  If you’ve got an infected aerosol super-emitter sitting at the table next to you, you’re going to get sprayed with those aerosols no matter what sort of fancy whole-room air filtration is installed.  Whole-room air filtration, or UV-C systems, if they are effective, might be able to reduce “far field” travel of virus-laden aerosol.  They might prevent the whole room full of people from getting infected.  But they don’t do anything for you if you’re sitting in the “near field”, that is, e.g., if you’re unlucky enough to be sitting a table next to a COVID-19-infected super-emitter of aerosols.

And, if you look at various cases where COVID-19 was spread in restaurants, the problem was primarily in near-field spread.  Sometimes individuals were infected in the “far field”, and that caught the attention of researchers.  But if you count the infections, almost all the infections occur among restaurant diners that are seated near the infected super-spreader.  Almost all of the problem is near-field infection.


A proposed air-curtain system to suppress near-field spread of coronavirus in restaurants.

That’s a fancy title for a solution that is very much high-tech redneck.

Here’s the deal.  I propose to bathe each table in a restaurant with its own separate curtain of filtered air.  You would take well-mixed air off the ceiling of the room, pass each table’s air through its own aerosol-capturing air filter, and blow that table’s unique air “curtain” gently down onto each table.  Each table’s  air curtain would isolate it from every other table.  This downward-flowing air curtain would also carry any aerosol emissions of the table occupants to floor level, where the air curtains would mix, be picked up by the room HVAC air intake, and be recycled back into the room after standard air filtration.

In a nutshell:  Each table would sit in its own clean air stream, and those air streams would only meet and mix at floor level.  People sitting at a restaurant table would never breathe unfiltered air originating from another table in the restaurant.

And I’d achieve that for $45/table.  Here’s how.

Filtrete ™ air filters are designed to capture aerosol-sized particles.  In particular, Filtrete ™ 2500 filters capture 76% of 0.3 micron particles.  (For comparison, N95 masks capture 95% of 0.3 micron particles.).  Two Filtrete ™ filters in succession capture 95% of 0.3 micron particles — the same as an N95 mask.

The huge benefit of Filtrete ™ over other media is that it generates very little “back pressure”.  This is why you can take your plain-vanilla furnace filter, replace it with Filtrete, and your HVAC system will continue to function normally.  The whole point of Filtrete ™ is that if filters those tiny particles without straining the fan on your existing HVAC air handler.

In fact, Filtrete ™ is so breathable that you can attach a Filtrete ™ to the back of a $20 box fan, and the fan can easily pull air through the filter.  I run such a fan at night in my bedroom, and have pretty much since the start of the pandemic.  I have one sitting next to me right now, running on low.  It’s pictured at the top of the page.

A 20″ x 20″ Filtrete 2500 filter sells for about $25 retail, and should last at least three months.  The box fan, as mentioned, runs about $20.

So, my solution is to hang a box fan, at ceiling height, over every table.  With a Filtrete ™ 2500 taped to the top of each box fan.  It’s the same idea as a standard ceiling fan, only it pushes well-filtered air down onto each table.

The up-front cost would be $45/table at full retail, plus labor.  Maybe throw in another $8 per fan to replace the front plastic grill with something fancier, made for use in ceilings.  Maintenance would cost perhaps $25/table/quarter for filter changes, or $8/table/month.  Along with maybe $2.50/table/month in electricity use (about 50 watts for 10 hours/day).

The use of box fans for cheap air filtration is hardly a new idea.  You can find plenty of products to do exactly that.  Such as this, or this , or these.

The only thing novel about this suggestion is to employ them en masse, as ceiling-mounted fans, with high-end aerosol-filtering material, to allow each table at a restaurant to have its own separate “air curtain”, uncontaminated by any near-field aerosol emissions from nearby tables.

This is, as far as I can tell, the only proposed solution for restaurants that appears capable of dealing with near-field transmission of coronavirus.  You literally don’t let the air from different tables mix until it hits floor level.  And you do that by bathing each table in its own separate curtain of filtered air.


Addendum:  Could you do this with regular ceiling fans, and no individual table-level filters? 

I think the answer to that is a qualified yes.

I think there would be considerable safety benefit in blowing a separate curtain of well-mixed room air from ceiling to floor over every table.  This, by itself, ought to limit near-field transmission by isolating each table from its neighbors.  That is, the aerosol emissions from the table next to yours get driven to the floor and mixed with other air streams, instead of spreading to your table as an undiluted cloud.

But even with excellent whole-room air filtration, the air curtains with under this approach would be less clean than with the box-fan-plus-Filtrete approach.  The reason is that virus-laden air will mix within the room and will make it up to ceiling level.  Overall room air filtration would limit the overall virus concentration in this mixed air.  But the mixing within the room would result in a “dirtier” air curtain blown down on each table, compared to using individual filters over the tables.

To see this, suppose that a restaurant installs Filtrete 2500 filters in its HVAC system, and they work as advertised.  The air coming out of the HVAC vents will have had 76% of aerosol particles removed from it.  But this will then mix with the remaining un-filtered air in the room before being blown down onto the tables by standard ceiling fans, resulting in an air curtain with a higher load of aerosol particles.  By contrast, a filter at every table would be like having a filtered HVAC vent over every table.  The air curtain would have had 76% of aerosol particles removed, without dilution by the “dirty” room air.

There is probably considerable merit in combining standard ceiling fans (set to downdraft) with enhanced HVAC whole-room air filtration (by, e.g., installing high-end Filtrete ™ filters in place of standard air filters).  This would surely be a cost-effective solution for a restaurant that already has standard ceiling fans in place.  And it might be adequate to prevent (or greatly reduce) near-field spread of COVID-19.  But it would not be quite as effective as filtered air provided directly to every table.


Second addendum, and an easy-to-grasp visualization

Taking this to its ultimate conclusion, there is yet another way to achieve the same result:  Put small fans on the tables, blowing upwards.  That’s about as cheap a solution as you can possibly get.

The idea here is that anything that keeps the air streams from individual tables separate, until they can mix well away from where people can inhale them, will reduce or eliminate the risk of spread of COVID-19 infection.

Why?  Because it’s all about the dose.  If you can dilute those aerosols enough, individuals will inhale too few virus particles to cause infection.

The link below is to an excellent piece of epidemiology, from China, on transmission on trains.  It shows this effect with considerable rigor.  It is a summary of thousands of contract tracing events, where an infected person rode a train, and they tested the people who were on those trains.  Conceptually, there’s not a lot if difference between sitting on a train for an hour, and sitting at a table for an hour.

https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1057/5877944

Guess that’s a hard slog for most readers.  To boil it down, the results are exactly what you’d expect:  The closer you sat to the infected person, the higher your likelihood of getting infected.  And that’s because the further away you get, the more dilute that stream of aerosols gets.

Here’s a useful analogy.  Visualize what I’m after by thinking of a person, sitting at a restaurant table, holding a lit cigarette.  Cigarette smoke particles are aerosols, and except for the initial upward thrust from heat, the smoke plume will behave much as COVID-19-laden aerosol plume will behave.  The plume from the lit cigarette is your proxy for the spread of  COVID-19 aerosol.  The goal is not to let that undiluted plume of smoke drift into somebody else’s face, and never to blow that undiluted smoke plume into somebody’s face.

Any of these setups would plausibly dilute that smoke plume before anyone else breathed it in.  Either via downdraft (with box-fan-and-filter, or with classic “Casablanca” ceiling fan), or updraft (with fan-on-table).  Truly, the only advantage of filtering the air just before blowing it down on the table would be a slight improvement in the quality of the air you are using.  And possibly some sort of marketing advantage, to assure your customers that you’re giving them the cleanest air possible.

So, upon reflection, I think the analogy to the smoldering cigarette is a good one.  The bozo with the lit cigarette is the COVID-19-infected individual.  The undiluted smoke plume is the plume of COVID-19 aerosols that will spread infection if somebody else inhales it full strength.  Your task is to dilute that smoke plume by ensuring that it mixes with room air well away from where people might inhale it directly. Put some additional filtration into the system — either above each table, as I initially suggested, or just by swapping Filtrete 2500s for standard filters in the HVAC — to keep the overall level of second-hand smoke (COVID-19 aerosol concentation) down, and I think you’ve gone a long way toward preventing the spread of disease in restaurants.


Coda:  The term “air curtain” is a real thing, used in real restaurants They are being touted as safety features in the COVID-19 epidemic.  It’s just that commercial “air curtains” are used as that — they serve in place of a curtain or door in a doorway opening.  They isolate spaces without imposing a physical barrier.  I’m just suggest that the same concept be used to isolate individual tables within a restaurant.

Second coda:  My wife tells me that the box-fan-with-high-end-air-filter is now a thing on the West Coast, for indoor air quality, in response to the forest fires.  As described here, or here, or here.  Smoke particles are aerosol-sized (that’s why smoke floats).  If they’ll clear smoke aerosols out of the air, they’ll clear other aerosols out of the air.

Edit:  Third Coda:  An air-curtain approach similar approach has already been tried in a hospital, to keep patients from infecting physicians.   Air curtains have also been proposed for air travel, to separate the passengers, in response to the coronavirus pandemic.

Final coda:  And now I’m wondering whether you could just reconfigure the ductwork of an existing HVAC system to get an “air curtain” effect.  For restaurants with exposed ductwork in an unfinished ceiling (e.g., Chipotle), could you achieve this “air curtain” production by directing the HVAC air handler output into temporary ductwork designed to “wash” the tables with a downdraft? 

I don’t think it will work.  And the limitation isn’t the ability to move air per se.  Every horsepower of motor in the HVAC air handler is equivalent to 15 box fans running on low (750 watts).  The problem is that a fan is moving room-temperature air, but the HVAC system is moving heated or cooled air.  In a properly-operating gas furnace, I think the air going into the ducts is around 170F.  You can blow a strong steady stream of room-temperature air over your guests without substantial discomfort, but you couldn’t blow a steady draft of 170F air over them.  

You might be able to centralize your air curtain system, using a single large blower and ductwork. But it would have to be separate from your HVAC system.