Air Filtration Options for the Social Hall, Revisited

WHAT ARE OUR OPTIONS for providing safe ventilation in the Social Hall? When it comes to preventing the spread of covid-19 infections, how does the performance and cost of the following options compare? 

1. Off-the-shelf filtration units sold for commercial use
2. Off-the-shelf filtration units sold for residential use
3. Do-it-yourself filtration units (assembled per a proven design)

This post offers some tentative answers.

Executive Summary

The ballpark up-front costs of the three main approaches are:

1. Off-the-shelf filtration units sold for commercial use:  $8.8K
2. Off-the-shelf filtration units sold for residential use:  $10.3K
3. Do-it-yourself filtration units (assembled per a proven design):  $1.2K

This is to achieve the goal of 6 air changes/hour at a tolerable noise level. That ventilation target, however, would not be adequate for a truly crowded room. And because the filtration units would be placed along the walls, the degree of the mixing of air at the center of this large room might be somewhat less.

The figures above do not include installation (electrical cords and mounting). Nor any difference in longevity, filter life and cost, ease of changing filters, aesthetics, warranty, and so forth. Nonetheless, the up-front costs are so different that they can give us a decent initial comparison. 

It’s Hard to Say

Getting a professional engineering analysis of the Social Hall would cost at least $1500, judging from a proposal that I received in September for the sanctuary. Short of that, I can estimate of the up-front costs by using a generally accepted target of 6 air changes per hour. But first, a caution about applying it in this case, namely a large room intended to hold a large number of people, with a ceiling that varies between 12'4" and 13'8" in height: 

The utility of [this] approach to ventilation is most useful in small rooms with ceiling heights generally less than 12 feet. In rooms with higher ceilings (eg, gymnasiums, atria), aerosols will dilute into the larger space and volumetric flow per area or per person [rather than per hour] would be a more appropriate measure that takes into account occupant density and activity level, which also influence aerosol emission rates.…
Air exchange rates are useful under typical or low-occupant-density scenarios, as should be happening during a pandemic. In places with large occupancy limits, …ventilation needs to scale up accordingly.

—Joseph G. Allen and Andrew M. Ibrahim, “Indoor Air Changes and Potential Implications for SARS-CoV-2 Transmission,” JAMA Insights; in Journal of the American Medical Association 325(20) [April 16, 2021]: 2112–13. doi:10.1001/jama.2021.5053

I honestly don't know whether the per-area or per-person approaches would yield a greater target air flow or a lesser one, compared to the per-hour method. (Dilution argues for needing less mechanical ventilation, whereas crowding argues for needing more.)

Room Dimensions and Methodology

According to the architectural drawings supplied in the 2003 Historic Structure Report, the relevant Social Hall dimensions are as follows: 87 ft long, 46 ft wide, and 13 ft high (on average). 

ROOM VOLUME  =  87  x  46  x  13  =  52,026 cubic feet

Contribution of each purifying unit (given its Clean Air Delivery Rate) in that space: 

CADR divided by ROOM VOLUME  =  air change rate

PLACEMENT.  For a large-room system, a ventilation synergy is achieved when the filtration units are close enough together that the air flowing out of each one reaches (just barely) the intake for the next one in the series. Such a pattern will keep the air circulating throughout the room.

Commercial Filtration Units

I solicited a bid from Air Cleaning Specialists, based in Missouri, which designs and sells commercial and industrial air filtration systems. Unlike with the sanctuary, they did a straightforward calculation that used the entire room size, including the high ceiling. To achieve at least 6 air changes per hour, the room would require 3 of their high-efficiency air cleaners (rated at 1875 CFM, containing both a MERV-15 filter and a HEPA filter).

The suggested placement is along the perimeter, with one unit on each of three walls. Units should be placed 35'–45' apart. They would be mounted on L-shaped brackets ($144 per set) at a 12' height (1' below the ceiling):

Placement Plan
(As usual, click/tap on an image to enlarge it.)

The noise level of the unit alone is 62 dBA (at 5' away), which seems a bit on the noisy side, although less so at a 12' height. A silencer attachment is available to reduce that level to 55 dBa, at an additional $193 ea.

Contribution of each air cleaner: 

1875 cu ft/min * 1 AC / 52,026 cu ft. * 60 min/hr  =  2.2 AC/hr

The cost of the 3 Blue Ox OX2500-HE HEPA Air Cleaner units with 3 silencers and one pressure gauge (to know when to change filters), but without the mounting brackets or electrical wiring, and not including freight charge, is $8,781.

To see the quote online with photos and links to additional product literature, click HERE and enter customer number THA9029.

Residential Filtration Units

An engineering research study by Srikrishna and published in May reported on lab-test performance comparisons between a few well-regarded commercially available units made for residential use and various DIY configurations. It is the basis for what I describe in this section and the next one.

Coway Airmega 300 — tested by Srikrishna for the highest fan speed: CADR for 0.3 micron aerosols = 265 cfm, at 59 dBA, which is probably acceptable for the Social Hall (about the level of conversational speech or an air conditioner). 

Contribution of each air cleaner: 

265 cu ft/min * 1 AC / 52,026 cu ft. * 60 min/hr  =  0.30 AC/hr

In this case, 20 units would produce 6 ACH, at a cost of $470 apiece via Amazon.com (2-day Prime) = $517 incl. tax, for a total of $10,340. Seems highly regarded on various websites. (Although Consumer Reports rated it poorly, those results were due to two issues that are of less interest to us: lackluster performance at low speed, and relatively high filter costs.)

These units stand on the floor and blow air upward, so the room-wide air-circulating synergy would not be achieved. Also, the perceived noise would be greater because the units are closer to ear level.

DIY Filtration Unit

The DIY air filtration unit that I assembled is discussed HERE. Its noise level (roughly 54 dBA, judging from Srikrishna’s tests, twice as loud as a typical refrigerator) is probably acceptable for the Social Hall at a 12' height and located along the walls. 

When the urgent call came in September for me to build the first DIY units for the Chapel, I had to use off-the-shelf 1"-thick hospital-grade MERV-13 filters from Home Depot. Since then, I have found sources for 2"-thick filters. Other things being equal, thicker filters perform significantly better because they don’t restrict the air flow as much.

Srikrishna measured the filtration efficiency and calculated a Clean Air Delivery Rate (CADR) of 293 cfm with the 2" MERV-13 filter.

With Nordic Pure filters (as in the published test): $166 for a box of 12  => $15.21 ea. (incl. tax). With FilterBuy filters: $156 for a box of 12. Performance difference unknown.

Cost per unit is $47: $30 for the box fan (incl. tax); $16 for the filter (including tax); 55¢ for 80" = 6.7' of clear packing tape (Office Depot: 120' for $6 incl. tax). 

Contribution of each air cleaner: 

293 cu ft/min * 1 AC / 52,026 cu ft. * 60 min/hr  =  0.34 AC/hr

In this case, 18 units would produce 6 ACH, but actually more would probably be needed in order to achieve the air-circulation synergy. If each unit’s exhaust travels about 10' (a testable assumption), the room would need roughly 26 units around the periphery.

TOTAL COST: $47 ea. x 26 units = $1,222. (Add another $7 apiece for the dust filters like on the Chapel's units, but then add another 2 units to compensate for a likely performance penalty, for a total of about $1,512.)

Note that the installation cost would be significant—at least $30 per unit, not including labor. And that assumes that electrical conduit is not needed.