Page 1 of 1

Particle count to mass conversion

PostPosted: Sun Dec 05, 2010 8:31 pm
by Wilberforce
The colors are based on the US-EPA zones of air quality index (which is based on mass)


Re: Particle count to mass conversion

PostPosted: Wed Jan 19, 2011 7:55 pm
by Wilberforce
The calculations given assume a specific particle size: 2.5 micron diameter. In reality, there exists
a statistical distribution of particle sizes. I am presently working on such a chart, based on the fact
that particle diameters are distributed over a range of sizes, from 0 to 5 um, centered at 2.5 um.
So far the results of these calculations are close to the "single-sized" calculation, but have a value
closer to the natural distribution of total mass value: f(n) = 0.105n (rather than 0.084n)

I am hoping to complete the calculations soon, and will post them.

Re: Particle count to mass conversion

PostPosted: Fri Jan 21, 2011 9:05 pm
by Dorre
Most researchers simply work out the conversion by comparing two instruments located in the same place. Tasmania's BLANKET researchers report PM2.5 pollution measured by DustTraks - see for pictures of the monitoring setup and a comparison graph of optical and gravitmentric PM2.5 measurements.

Broadly speaking, If your nephelometer shows scattering coefficients, 1 neph coefficient is about 22 ug/m3 of PM2.5 pollution.
If you have a DustTrack reporting PM2.5 pollution using the Arizona Dust calibration, multiplying the results by 0.42 should be close to the true PM2.5 for woodsmoke.

An earlier reference Report:
DEH, 2005. Department of Environment and Heritage. Use of Particle Counters in Woodsmoke Affected Areas: ... nters.html).

Hope this helps,

Re: Particle count to mass conversion

PostPosted: Mon Jan 24, 2011 8:56 pm
by Wilberforce
Thank you Dorothy.

The second link is to a study which has made measurement comparisons of particle count to particle mass.

I've been working on an update to my original, simplistic chart, and have been calculating gaussian distribitions for various particle sizes. The problem is that wood smoke, unlike vehicle effluent, is of widely-ranging particle diameters. The largest particles originate from cool, smoldering wood fires (such a OWBs) and the smallest particles originate from very hot-burning wood fires. Measuring a typical woodsmoke-contaminated environment's PM level gets complicated, given these wide range of particle sizes.

What looks promising is a strongly positively-skewed curve (as mentioned in the study) where the smallest particles are by far the most numerous. If we use an integral calculus formula, we can calculate results which are similar to those in the measurement study: f(n) = 0.4n (that is, ug/m3 = 0.40 of PM2.5 count)

Re: Particle count to mass conversion

PostPosted: Mon Oct 22, 2012 8:50 pm
by bushfood
Hi guys
This may sound dumb, but I have happened across a used Dusttrak 8530 and have been trying to measure local brushfire conditions here in Australia.

I am a total amateur... this is a strange hobby of mine I suppose. So forgive the dumb questions.

1) User calibration factor.... I have read that Dusttraks tend to overcount PM 2.5 mass... is this true? What would be an appropriate user calibration setting for a dusttrak 8530 measuring smoke PM 25?
I assume if the Dusttrak is overestimating mass, then the user calibration factor would be < 1.0... correct?

2) Flow rate.... Also what should the Flow Rate be set at? Currently it's set at 1.0. Is this appropriate for woodsmoke? The manual says something about the factory flow rate set at 3L... how come it seems the device is set at 1.0 by default?

In other ways the unit seems reliable (plausible results compared to comparable results I've googled...) for PM2.5 measurements (e.g. average 7 ug/m3 overnight ambient outdoor air, slightly elevated near a freeway to aboutr 15-20ug/m3, spikes to 100-200 ug/m3 during cooking; shoots up to 30-50 ug when there's smoke on the breeze from burning, etc.... seems to be reasonable)

Over the weekdn we were subjected to a massive and completely unnecessary "controlled burn" which resulted in readings > 2000 ug/m3 for hours on end. Lovely...


Re: Particle count to mass conversion

PostPosted: Sat Oct 27, 2012 8:39 pm
by Wilberforce
Hi There

I had meant to post a count-to-mass calculation, but encountered a conundrum when I realized that lacking actual PM10 count data, I might have an inaccurate conversion. This is because I do not have a PM-10 measuring instrument; I only have a PM-0.5/2.5 meter.

Consider that 10-micron size is by far the most massive in the gamut of airborne particulates. These must be counted in order to ascertain their quantity. Lacking a PM10 meter, I can only estimate the total count-to-mass calculation. This is done by mathematically "fragmenting" a 10-micron particle into 2.5-micron (which would be a larger number) The statistical distribution of smoke particles probably follows a chi-square distribution, where there are many times more ultra fine particles than large ones.
These ultra fine particulates (those less than 1 micron) make up only a very small percentage of total mass. Factoring in this count will not alter total mass contribution significantly. Only the larger sizes really matter.

As time permits, I will resume these calculations and post them (perhaps in a month or so)

thanks for your comment


Re: Particle count to mass conversion

PostPosted: Fri Jan 24, 2014 4:06 am
by David Helman
Following on the message I sent when I joined this board, I'd be very grateful if your could continue your work on your calculation that will allow for an estimation of mass based on a specific PM size (PM2.5 is of particular interest to me).

I wish I had the means to buy another instrument that I could compare my MetOne GT-521 with for mass conversation, I don't.

I mostly measure indoor air in a heavily polluted urban setting (Beijing to be specific)

Looking forward to whatever you come up with, and if I do have a chance to compare my findings with someone that does have a mass measuring instrument, I'll let you know how it compares.

Re: Particle count to mass conversion

PostPosted: Fri Dec 18, 2015 7:58 pm
by Wilberforce
Here is a new paper on what we may be looking for: