In my last post, I noted that yesterday’s readings for airborne particulates in my area were high, but would not have been hugely abnormal in the 1960s. Yesterday, the air was somewhat visibly smoky, and by mid-afternoon the level of PM 2.5 particles was about 200 (micrograms per cubic meter of air), and PM 10 was about 300 (same units).
Today, the air is visibly clearer, and we’re down to a PM 2.5 of 35, and a PM 10 of 50.
Source: Accuweather.com, reading for Vienna VA, accessed 11 AM EDT 6/9/2023
As is my habit, I have a few more related questions and observations.
- What, exactly, is the Air Quality Index (AQI)?
- How does today compare to a typical summer day in NoVA, for air pollution health risk.
- How good are my zero-effort home-made air cleaners, in this situation?
What, exactly, is the AQI?
In the picture above, you’ll notice that the numbers for the “air quality scale” are not the same as the raw readings in micrograms per cubic meter of air. What are those “air quality scale” numbers, and what is the “Air Quality Index”.
As a health economist, I would have bet that the Air Quality Index was set up to combine the estimated health impact of all major air pollutants. You’ll note that a microgram of PM 2.5 — a relatively dangerous form of air pollution, because it can get deep into your lungs — counts more than a microgram of PM 10, a somewhat less-dangerous pollutant. This type of scaling is a very common thing to do, in health care. The goal is to give you a single scale that combines all the various pollutants in a way that captures their overall expected impact on health.
That’s close, but it’s not exactly right. And the details on the actual construction of the AQI are a bit stranger than I would have guessed.
First, the AQI tracks the levels of five categories of pollutants:
- ground-level ozone
- particulate matter (PM 2.5 and PM 10).
- carbon monoxide
- sulfur dioxide
- nitrogen dioxide.
Why five? Which five? Turns out, that’s based in legislation. These are the five major air pollutants that are regulated by the Clean Air Act. That means that a) the EPA has standards for them, b) the EPA has to monitor them to see whether or not areas are in compliance with the Clean Air Act (reference, reference). By law, every city of 350,000 people or more must publicly report its AQI at least five days a week.
But exactly how the levels of those pollutants become the AQI is not obvious. For that, I have to go back to the literal 1999 Federal Register listing for the rule establishing the Air Quality Index can be accessed at this link.
Based on that document, for each of the pollutants, the EPA used the scientific literature to establish two levels. First, there a level that the EPA uses for regulation, based presumably on some demonstrable negative health impact. That’s the level which should not routinely be exceeded. The EPA labels that as the point where the pollutant becomes “unhealthy for sensitive groups”. That level determines the value of 100 on the AQI scale for each pollutant.
Second, there is a level at which the scientific literature demonstrates extreme health hazards. That is set to a level of 500 on the AQI scale for each pollutant. The EPA labels the top category of the scale as “hazardous”, starting with a level of 300, going up to 500.
So, for each pollutant, an AQI reading of 100 is where air pollution has an impact on sensitive individuals (e.g., kids with asthma, people with heart or lung disease). It’s also determines whether or not you’re in violation of the Clean Air Act. By contrast, a reading of 500 is where air pollution is hazardous to all, and probably killing some. Glancing at the detailed documentation, it appears that the scales themselves are non-linear, which I can only guess reflects estimated health hazards.
Everything else is scaled accordingly. So, roughly speaking, if a pollutant is at half the level at which health effects appear for sensitive individuals, the AQI score for it would be somewhere around 50.
Then, the overall AQI for an area is simply the maximum score for any of the pollutants measured. And, typically, AQI reporting will tell you which one. So, today, the AQI in my area is 72, for fine particular matter (PM 2.5). Because, of all the pollutants in the overall AQI, that one has the highest AQI score.
So, to summarize the AQI:
- The level of each regulated pollutant is scored on a scale that roughly (but only roughly) reflects equal health impacts.
- At 100, the pollutant is known to have negative impact on sensitive groups.
- At 500, the pollutant is hazardous to everyone.
- The overall AQI for an area, at any point in time, is the score for the worst pollutant at that time.
So, to be clear, the AQI is not specifically an index that addresses overall expected health impact of all the regulated pollutants in the air. In no sense does it “add up” the different pollutants. It only reflects one air pollutant at a time, whichever one is estimated to have the most significant health impact at that moment.
How does today compare to a typical summer day in NoVA, for air pollution health risk?
The funny thing about risks is that you don’t notice them until you notice them. Then it’s hard NOT to notice them.
I’ve run a home-made air cleaner for years now. And so, I can guarantee you that the air in the DC area is always dirty to some degree.
How do I know? I’ve got pictures. Below you see one of my simple air cleaners — a box fan and a Filtrete (r) 1900 filter — along with the filter that had been on it for roughly three months. As I noted in an earlier post, the schmutz does not lie (Post #906). The discoloration on the filter is from stuff what was filtered out of the ambient air.
So the air is always somewhat dirty. How does today’s PM 2.5 reading compare to a typical summer day in my area? Is today’s air quality an extremely rare event? Or is the air roughly this dirty all the time, I just never stopped to notice it?
I can easily find data for the past few weeks. And for that brief period, the answer is that today’s PM 2.5 is only modestly higher than what we would experience once a week or so. Like this:
Source: IQair.com
This graph spans the last 30 days. Today’s PM 2.5 reading (right hand bar) is just slightly higher than the readings on May 11 or May 22, which is well before people started talking about the smoke from these Candian forest fires. By contrast, the reading from yesterday, when the air was visibly smoky (next-to-last bar on right) was far in excess of what we would normally see in this area.
If I want a multi-year perspective, I have to expand my area to all of Fairfax County, VA. If I do that, I find that we have typically have nearly this amount of PM 2.5 in the air, a few days every year. Like so. Looking back at 2022 (left graph), there are a few dots (days) for which PM 2.5 was very nearly at the current level, on the AQI scale.
Source: EPA.gov website, air pollution plots for your choice of area.
The upshot is that the level of particulates in my area today is high, but not unprecedented. We seem to have a few days every year with fine particulates (PM 2.5) near today’s level.
Filtrete (r) fanboy, or, the easy way to clean your air.
First, let me state the obvious. If you want to reduce the amount of particulates you are breathing, wear an N95 respirator (mask). This is, apparently, common behavior out West, during wildfire season. The test standard for an OSHA-certified N95 respirator is that, if properly fitted, it will stop 95% of particles 0.3 microns in size. Those are the hardest particles to stop. So it will stop at least 95% of PM 2.5 particles.
That said, for keeping the air in the house clean, my go-to solution is to place a high-end 3M Filtrete (r) filter behind a cheap box fan. As above. The suction from the running fan holds the air filter in place. And the fan draws relatively little electricity.
This is my version of the so-called “Corsi box”, a much more complex contraption that uses multiple standard high-air-resistance high-Minimum Efficiency Reporting Value (MERV) filters.
The advantage of Filtrete (r) is that it provides high filtration without much back pressure. This is why you can draw a considerable volume of air through a single filter using a cheap box fan. (And it’s why the “Corsi box”, using standard high-resistance high-MERV filters, has to use several taped together in a box arrangement, to provide enough surface area to allow the cheap box fan to pull air through those high-resistance, non-Filtrete (r) filters.)
Here’s the labeling on that Filtrete (r) 1900 filter. (The 1900 refers to 3Ms measure of the filtering capacity — the higher the better.) This qualifies as a MERV 13 filter. In a single pass through the filter, it removes
- 62% of the tiniest particles (0.3 to 1.0 mircons)
- 87% of the mid-sized particles (1.0 to 3.0 microns)
- 95% of the larger particles (3.0 to 10 microns).
If this were a mask, it would be an N62 mask. And the upshot is that for a single complete air exchange in a room, this removes as much of the larger particles as an N95 respirator would.
(Just as an aside, I went through a lot of the basics of filtration in my posts during the pandemic (Post #593). These high-efficiency filters operate in a completely non-obvious way, using permanently charged micro-fibers (plus Brownian motion) to filter particles out of the air. )
One of the damnable things about cheap box fans is that nobody will give you reliable information on how much air they move. Manufacturers will list their estimate of cubic feet per minute (CFM) on high only. To guess the air flow at lower settings, I have to measure the wattage, and assume air flow is roughly proportional to wattage.
For the particular fan I use, Howe Depot lists it at 1800 CFM (on high, 68 watts). Based on the wattages (via a Kill-a-Watt meter), it should generate about 1400 CFM on medium (54 watts), and 1100 CFM on low (43 watts).
I’m running it in a fairly large space — about 6400 cubic feet. Even with that, on medium, with ideal circulation, this will pass all the air in the room through that filter in less than five minutes. That’s far faster than any plausible exchange of air between the inside and outside of my house. Hence, I judge that it’s pretty effective at keeping the particulate levels down in the interior of my home. And I judge that a cheap 20″ box fan provides more-than-adequate circulation, as long as you use it with a low-resistance Filtrete (r) filter.
If I ran it on medium, it would consume about 1.3 KWH per day. For me, at about 12 cents per KWH, if I ran it 24/7, that would cost about $55 per year for the electricity, and would create about 300 pounds of C02 emissions (at Virginia’s generating mix).
The downside of Filtrete (r) is the cost. As with almost all products from 3M, they work great, and 3M charges accordingly. At present, a single 20″ x 20″ Filtrete 1900 costs about $22 at Lowes. It appears that you can buy knock-off electrostatic low-back-pressure filters on Amazon for half that. (Of that, the low back pressure is key to using this effectively with a cheap box fan.)
If you go with genuine 3M, the trick is that they seem to price the filters virtually the same, regardless of size. The 20×30 filter costs just a dollar or two more than the 20×20. So I tend to stick with Filtrete, but buy the larger filters and rotate them when in use. That way I basically get a filter-and-a-half, for the same price as a filter.
