Thread: Particle Size Distributions 101 (warning - only invisible dust discussed)

Next my drive home from work.



On the way out of my office I walk through two car parks to get to my car. On the way home I cross a major suburban road so I pulled over into a shoppers car park and took a reading downwind by the side of the road. The reading of 22+ million particles per cubic metre is not unusual. On a hot summers day this can get up to over 30 million.

This fine dust is almost certainly combustion products and tyre dust. We are of course breathing all this crap in all the time.

BTW this busy suburban road is about 300 m from my place.

Now for some domestic results.



Note how the air in my backyard is not that different from the external air at work in the first graph.

None of the rooms in our house have carpets and I'd say the results are higher than normal because our cleaners have been crook and have not come for 2 weeks.

Take a geezer at the results for the so called HEPA filtered vacuum cleaner. This is a Nilfisk unit so it's not exactly a cheapie. This is how I did the test - just monitoring the exit air.

Here is the readings direct off the counter

This demonstrates how a vacuum cleaner can make more invisible dust than is already in the room air.
I strongly suspect this vacuum cleaner may be damaged because this is about double what it was reading a few years ago.

Now we get to find out what is going on in the shed. The last time I used the shed for woodwork was last Sunday and no one has been in there since then so all visible dust and some invisible should have settled. My shed is not all that well sealed so some external air and dust can get inside relatively easily.



The first thing to note that the shed dust levels are similar to the external air but if outside air was the determinant it should be similar to or less than the air inside the house where people and pet have been moving around and generating dust . If anything the shed is dustier with dustiest place is right in the middle of the shed this suggests that the very fine dust still has not settled since Sunday. This is why it is very hard to determine if a specific dust grabbing process makes any kind of difference. One thing that does surprise me is that the external enclosure in which the DC is located is no different to external air, except perhaps the very finest particles where the DC housing has the lowest of all the shed readings.

I also have a HEPA filter up in the ceiling of my shed and I checked it and up close it reads zero. I then ran it for a few minutes and checked 1m below the level of the HEPA and the results are the black squares. I will hopefully be able to reduce the invisibles load inside the shed and the do something for a short period and see what happens. I cannot undertake a process involving the DC for too long because it will quickly drag too much external air into the shed.

What I will have a crack at is seeing if I can measure a noticeable increase in invisibles and how long it takes to flood the shed and then to bring it down to outside air levels. It might not be possible but I think it is worth a try.

Here is a graph demonstrating how the invisible particle distribution of dust changes over 3 days in a shed that is not being used. For comparison I also show the variation in the distributions in the air immediately outside the shed.

MOS = middle of shed (dashed lines)
BY = back yard (solid lines)



This shows a 55 % variation in outside air, and a 40% variation in the numbers of particles inside the shed. The number of small particles can be sometimes correlated with humidity (ie particles smaller than 0.3 microns act as nucleating site for water vapour and the particle can grow to >0.3 microns) which is probably why morning air may have more particles than evening air.

Yesterday the cleaners came and cleaned the house.
Cleaning includes dusting and then vacuuming (using that dirty Nilfisk vacuum cleaner)
Two hours after the finished cleaning I came home and measured the dust levels



This graph shows that there appears to have been a small amount of invisible dust removed for the 0.7 - 2 micron range but all other dust levels appear to remained the same or increased.

Reasons may include
- the vacuum cleaner is dirty/damaged (most likely cause)
- changes in external air dust particle distributions (would not cause this much of a change for the 0.3 - 0.5 micron range
- "cleaning" stirs up more dust than it removes - also likely

The reason I am showing all this is to illustrate why invisible dust particle distribution measurements are very difficult to perform.

My children will never clean their rooms if they get their hands on these graphs!

Regards

Peter

Originally Posted by The Hun

My children will never clean their rooms if they get their hands on these graphs!

Regards

Peter

Sorry about that
This highlights problems with many vacuum cleaners that merely rearrange the dust and add to the problem by mincing up some of the bigger chunks into small ones.

Originally Posted by BobL

Here is a graph demonstrating how the invisible particle distribution of dust changes over 3 days in a shed that is not being used. For comparison I also show the variation in the distributions in the air immediately outside the shed.

MOS = middle of shed (dashed lines)
BY = back yard (solid lines)



This shows a 55 % variation in outside air, and a 40% variation in the numbers of particles inside the shed. The number of small particles can be sometimes correlated with humidity (ie particles smaller than 0.3 microns act as nucleating site for water vapour and the particle can grow to >0.3 microns) which is probably why morning air may have more particles than evening air.

Yesterday the cleaners came and cleaned the house.
Cleaning includes dusting and then vacuuming (using that dirty Nilfisk vacuum cleaner)
Two hours after the finished cleaning I came home and measured the dust levels



This graph shows that there appears to have been a small amount of invisible dust removed for the 0.7 - 2 micron range but all other dust levels appear to remained the same or increased.

Reasons may include
- the vacuum cleaner is dirty/damaged (most likely cause)
- changes in external air dust particle distributions (would not cause this much of a change for the 0.3 - 0.5 micron range
- "cleaning" stirs up more dust than it removes - also likely

The reason I am showing all this is to i;llustrate why invisible dust particle distribution measurements are very difficult to perform.

Very interesting Bob, I'm going to share this information with my next door neighbour, whose child suffers from dust realted issues really badly.

Intriguing series of posts, Bob. Wondering where you will be headed next.

Originally Posted by BobL

This graph shows that there appears to have been a small amount of invisible dust removed for the 0.7 - 2 micron range but all other dust levels appear to remained the same or increased.

Reasons may include
- the vacuum cleaner is dirty/damaged (most likely cause)
- changes in external air dust particle distributions (would not cause this much of a change for the 0.3 - 0.5 micron range
- "cleaning" stirs up more dust than it removes - also likely

The reason I am showing all this is to illustrate why invisible dust particle distribution measurements are very difficult to perform.

A couple of additional explanations might include:

* the vacuum concentrates the smaller particles - sucking them out of the carpet/rugs/off floor and exhausting them into the air.

* Vacuum filters stop some of the larger particles, p>=5, but smaller ones pass through it.

* Vacuum generates small particles by chomping up lager ones.

* Vacuum exhausts last weeks particles from collection bag.

Have you considered monitoring the air before, during and after a solid woodworking session in your shop? After all, that is what this is all about.

Keep up the great work.

Fair Winds

Graeme

PS: Hope everyone realises that you have used a logarithmic y-axis, and its implications. (Between 100 and 1,000 times more small particles than large particles in your tests.)

Originally Posted by GraemeCook

Intriguing series of posts, Bob. Wondering where you will be headed next.

Stay tuned

A couple of additional explanations might include:

* the vacuum concentrates the smaller particles - sucking them out of the carpet/rugs/off floor and exhausting them into the air.

* Vacuum filters stop some of the larger particles, p>=5, but smaller ones pass through it.

* Vacuum generates small particles by chomping up lager ones.

* Vacuum exhausts last weeks particles from collection bag.

All good points. but I think the vacuum cleaner has a leak that I need to check first

Have you considered monitoring the air before, during and after a solid woodworking session in your shop? After all, that is what this is all about.

That is where I am headed but I want to demonstrate that assessing the background is devilishly difficult to deal with.

PS: Hope everyone realises that you have used a logarithmic y-axis, and its implications. (Between 100 and 1,000 times more small particles than large particles in your tests.)

Good point - I should have mentioned this.

Am I missing something...? What's happening to particles between 2 - 5 microns (or are you just skipping those figures)?

Originally Posted by Master Splinter

Am I missing something...? What's happening to particles between 2 - 5 microns (or are you just skipping those figures)?

?? which graph are you referring to?

OK I think I know what you are asking and I agree it is a bit confusing.

The first 4 posts I show what the Particle counter shows.
It shows 6 readings
>0.3 microns
>0.5 microns
>0.7 microns
>1 micron
>2 microns
>5 microns

After that I'm taking the difference between any two adjacent micron size readings and this translates into
0.3 - 0.5 microns
0.5 - 0.7 microns
0.7 - 1 microns
1 and 2 microns
2 - 5 microns
>5 microns

OK I finally had a crack at doing a practical experiment in real wood dust measurements.

I first measured the dust profile outside and inside the shed.
The dust profile inside the shed I called the Shed Start
Then I made some dust for 2 minutes by sanding a piece of sheoak with 240 grit without using the DC.
Then I measured the dust profiles over a period of time 30 minutes
These measurements were made at the height of the operator's face while standing at the lathe
Then I measured the dust profile outside the shed again.
The dust values over time after sanding have the Shed Start values subtracted from them.



By making a few estimates and a bit of arithmetic the Particles/L can be converted to mg/m^3.
The concentration over time is the red line towards the bottom of the graph - the scale for the numbers of particles/L also applies to the conc
The OHS exposure limit for hardwoods is 1 mg/m^3
Within 2 minutes of sanding the levels are 7 times above the limit
It takes about 15 minutes for just 2 minutes of sanding to drop below this limit.
Now imagine turning and sanding for a couple of hours?

I am open to suggestions for follow up experiments but please don't suggest anything too long or too many different experiments as I cannot do anything else in the shed while I am profiling.

I haven't thought about a tolerance for the measurements yet, probably around +/-10%?

Obviously I will repeat this experiment with the DC on.

I had another look at the data for the 2 minute sanding experiment and replotted the particle count data (minus the shed background) in the usual way instead of the versus time method shown in the previous post.



Here we are looking at the usual particle distribution profiles and can see how the profiles changes with time. Where data is missing that is because the values have dropped below the shed background and since log-scales can't handle negative numbers I just left them out.

One thing that shows up for this experiment is how quickly the numbers of the smallest particles decrease over time. I think this is because a 2 minute sanding effectively represents a pulsed injection into any medium sized shed or larger. Since the smallest particles act more like gasses the smaller they are the more quickly they will diffuse from the injection site while the heavier particles hang a for longer.

Once the shed is full of small particles the rate of diffusion will slow down. I don't think I really don't want to demonstrate just now that although I wouldn't mind doing that once to see how well the DC vents the shed.

Good Morning Bob

This latest graph is not quite what I would have expected in a couple of aspects.

First, the four minute line appears to be consistently too low. Is there a possibility of a data measurement or data processing anomaly??

Second, the abrupt change in curve shape for the 20 and 30 minute curves may be questionable? If your dispersion argument is valid, then would not the start of this dispersion have been reflected in the 15 and even 10 minute curves?

Thirdly, was there any external change to the shed environment during the experiment? (eg Some one opened shed door.)

Fair Winds

Graeme