[Stoves] PM measurement with Speck

[Crispin Pemberton-Pigott]

Dear Frank And All

>So if I understand what you are saying the sample is diluted enough such
that only single particles are measured and they are not entered into the
detector as in groups. I'm thinking more of dirty water that I am familiar
with. The dirty water is paced into a vial and tilted (shaken) just before
making a measure of reflected light from the group. But if we want counts
the water is diluted and number of each particle determined. This to count
cells etc.
>is this much the same as measuring particles in air? 

The particles are not measured individual in the sense of walking them
through one by one - though there are instruments that work that way. They
have multiple scatterings which are reflected off mirrors and land on a
detector. It is something like a CCD camera so it can hit multiple places
and still get counted, like dots on a picture, of different sizes.

The instruments that count one by one are able to give a shadow shape, like
a worm or a ball, and a reflection value to give albedo, and a calculated
volume which can be turned into a mass if a density is assumed. Some can
calculate very large numbers of particles per cubic cm but most are not very
high. If you want shape and colour, you have to dilute more. 

There is a machine that will do all that and analyse a billion particles per
cubic cm (!) but it costs 79,000 Euros.

There is a new machine from Grimm that will report nanoparticles using
optics. I don't know how it works yet. It is brand new. Costs about $40,000
but that was a flexible number. I think it is based on their R-11 series
which I am using in one place and it is really good. Separates particles
into 31 sizes and you can assign a different density to each.

>Does white colored particles give more reflected light than black
particles?

People are clever. They don't use visible light. Visible colours (to us)
means visible light. Colour is different in other colours of light. It will
give more 'light' but the reflection is not necessarily larger in diameter.
So you can choose to measure the diameter or the brightness and get both
colour and size.

>
>What is the converting factor going from reflected light to mass? 

The measure the area of the reflected dot and convert it to a circle and
estimate the volume of a similar sized globe. A density is found practice
and calculation. Wood fire particles have a density of about 1.00.  If
someone says it is different, make them prove it. You can assume PM5-PM10
particles are 2.5 or 2.3. 

Wood fires don't make large particles unless there is a fan, but even then,
they are about PM4-5. I can show you the difference of fan-on and fan-off if
you really want to see it.  I have data from a fan stove. Then it is on low
and high is clearly visible. With no fan (low power) it looks just like the
distribution of a natural draft fire which is all below PM1 and most below
PM0.6.

>When the reported value is 2.5 um is there a range like from 3um to 2um and
above and below that not included?

All PM numbers include 'everything smaller'.
You can divide the measurements into any ranges you want but you need a
machine like a Grimm 11-R or 11-C to do that. There is a Grimm 320 that the
SeTAR Centre has which will do 31 bins from PM0.22 to PM100.  The Grimms
have the big advantage of having size selectable brackets of your own
choosing. So you can shift the available 31 slots into regions of interest
and see quite fine divisions if you want. There is a tab for putting in the
sizes you want counted in each channel - they can even overlap I think. 

>Is the angle of reflected light used to determine the size of the
particles? 

No. I think the angle is used when working with particle physics to
determine mass directly, not so?

Another method is static attraction to one side. The particles are sped
through a portal at a known velocity and charged. A plat on the side is
divided into separate landing zones. If the particles are light, they swing
to the side quickly to the first landing zone. Heavier particles land
further on. The nanoparticles can be grown by condensing vapour on them and
detected in that way, with the growth known very accurately. The mass of the
condensation is deducted and the balance is the nanoparticle.

The new Grimm technology is not like that.

A Tapered Element Oscillating Microbalance is a filter mounted on the end of
a vibrating beam. The sample is drawn through the filter and its mass change
moves the vibrating frequency. This gives real time mass. By using a cyclone
a particular size (like <2.5) the mass of that fraction can be watched.
Fortunately for wood stoves, without fans anyway, there is no need to use a
cyclone because there are virtually no particles larger than PM1.0 so you
get PM2.5 automatically. This also affects the sampling. If the sample is
well mixed and not blasting down a tunnel, you don't have to have isokinetic
sampling either. That saves a lot of complexity and checking. 

If you use a SeTAR type diluter where there is no 'blow-by' in the tunnel,
the whole sample is taken into the instrument so there is no influence in
terms of kinetics. There is a letter somewhere explaining all this to the
JICA Team in Mongolia. It is possible to use particular aspects of wood or
coal combustion to eliminate a lot of the complexity in PM measurement that
is assumed to be necessary in EPA methodologies. Power stations and outdoor
are sources very different from stoves.

Regards
Crispin