We use a lot of products that are
sprayed. Everything from hair spray to
furniture spray to window cleaner are spray-applied either from a pressurized
can or a trigger spray. The
manufacturers of these items want ALL of the product to go from the container
to the target. It most cases, it mostly does; however, not
all of the material goes to the target, some remains in the air as an aerosol of particulates. This remaining
airborne fraction has been called “overspray” or in some cases, "bounce back". Whatever it is called this airborne fraction is a potential health concern
because the vast majority of spraying occurs in what we call the “near field”. This is the volume of air that includes the
spraying product and the breathing zone (nose and mouth) of the person doing the
spraying. The smaller the near field is,
the higher is the potential for significant concentrations and exposure. Since most spraying occurs at literally at arm’s
length, with some products like hair-spray at considerably closer distances to
the breathing zone, the possibility for at least acute inhalation exposure exists.
In doing a risk assessment for
these types of products, manufacturers have historically depended on some cursory data and “rule
of thumb” estimates for the amount of overspray that they would use for their
analysis. A few years ago, one of my
clients at Procter and Gamble asked if we could do a better job of quantitatively
estimating overspray. That effort
ultimately culminated in a paper published in 2012. I would be happy to send this paper to
anyone requesting it at mjayjock@gmail.com.
What I want to do in the remainder
of this post it to provide you with the primary learning from that work:
The predominate mechanism for
particle overspray from sprayed products is the failure of relatively small particles
in the sprayed stream to impact the surface because of their tendency to remain
in the flow lines of the air stream. The figure below illustrates this effect
which, by the way, is the primary mechanism used by particle size impactors
like the Anderson Impactor:
Thus, even when sprayed at 90
degrees to the surface, some of the smaller particles will following the
streamlines and escape capture. As
such, the term “overspray” is a bit of a misnomer. In this case, everything is being sprayed directly at the target and nothing is being sprayed over
the target. Another term used to describe this loss, “bounce
back” is also a misnomer since studies and data indicate there is essentially
no bounce associated with wet aerosol particles sprayed against a solid
surface, all the wet particles that hit the surface stick.
Other conclusions include:
1. Small aerosol particles (less than 15 μm MMAD) make up the vast
majority of measured airborne overspray from sprayed products.
2. Larger wet particles will have a much stronger tendency to impact and
stick to the receiving surface. Relatively large particles that are indeed “oversprayed”
past the target and do not stick tend to quickly settle out of the air.
3. As a worst case, wet particles less than or equal to 30 μm could rapidly evaporate to respirable size in a time frame relevant to the exposure event.
This size range should be considered in estimating the potential respirable
mass.
4. Also, almost all particles greater than 30 μm MMAD that remain airborne after
spraying will settle 200 cm (2 meters) downward and be on the floor within 1
min. Thus, between impaction loss and settling, few particles are left to
become or remain airborne in a size much above 30 μm for any time 1 min after spraying.
5. Any future work on evaluating this exposure/risk should focus on
respirable and near thoracic particle sizes, that is, particles with an
ultimate aerodynamic diameter below 15 μm.
6. Any reduction in the mass of the low-end particle size distribution tail or
bimodal “hump” from spray products will directly and significantly reduce any
overspray potential.
7. Overspray potential from sprayed products is best estimated using a
real-time laser particle sizer. A rough average from available data would
indicate a reasonable worst-case respirable overspray potential of 5% of the emitted
mass and a worst-case total aerosol overspray potential of 6% for trigger
sprayed products.
8. Theoretical considerations indicate that hard surfaces (e.g., metal or
glass) should react in a manner similar to soft surfaces, such as cloth with a
low pile. That is, they should produce essentially the same amount and type of
airborne overspray from sprayed liquid aerosol.
9. The spreadsheet model developed for this work should be useful for
estimating the amount of potential overspray based on particle size
distribution of the spray. This model will overestimate the overspray potential
for thick pile targets such as hair and carpet.
As usual, I welcome hearing about any of your thoughts or any experience
you have in this matter.
No comments:
Post a Comment