See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/306106362

Wildfire Smoke Exposure: A Comparative Study Wildfire Smoke Exposure: a

Comparative Study between Two Analytical Approaches; Particle Assemblage
Analysis and Soot, Char and Ash Ana...

Article · August 2016

CITATIONS READS
0 260

3 authors, including:

Brad Kovar Ernest Crutcher

Safeguard EnviroGroup
40 PUBLICATIONS   57 CITATIONS   
9 PUBLICATIONS   12 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Wildfire Smoke Exposure: Comparative Study between Particle Assemblage Analysis and Soot, Char and Ash Analysis View project

All content following this page was uploaded by Ernest Crutcher on 15 August 2016.

The user has requested enhancement of the downloaded file.

Wildfire Smoke Exposure: A Comparative Study

Wildfire Smoke Exposure: a Comparative Study between Two Analytical

Approaches; Particle Assemblage Analysis and Soot, Char and Ash Analysis
By: Brad Kovar1, Russ Crutcher2, and Heidie Bettes2
1
Safeguard EnviroGroup, Inc; 2 Microlab Northwest, LLC

ABSTRACT

The characterization of combustion particles is an essential part of establishing the

presence of debris from a wildfire. That characterization is much more than
identifying charred wood, soot, or black material. It involves the identification of the
combustion products from the fuel. That includes characterizing the types of charred
wood, phytoliths from the plants that made up the fuel, skeletonized cell structure of
those plants in the ash, the burnt soil from the updrafts, and even aerosolized fire
retardant. Forest, chaparral, and savannahs are not made up of one or two plant types
but dozens of different plants. These all contribute to the smoke and form a signature
for that fire. This signature forms the assemblage of particles marking the specific
wildfire. This paper includes a comparison of two different approaches used to assess
wildfire smoke exposure. One is based on the presence of black particles and the
other is based on the unique assemblage of signature particles. Duplicate samples
were taken and were sent to two different laboratories for analysis in one of the case
studies shown here. In another case study proximity to major roads improperly
correlated to an increase in the level of reported exposure to fire debris.

INTRODUCTION

Correctly identifying exposure to wildfire smoke is critical to assessing the need and the extent
of remediation that can be justified on an insurance claim. Millions of dollars are at stake for the
insurance companies and tens of thousands of dollars are at stake for the home owner. Both
parties need an accurate assessment of exposure to begin the process of negotiation. The first
step in any assessment is the collection of samples. Where they are collected, how they are
collected, and the medium used to collect them is important1. Ideally, this is done soon after the
suspected exposure. In reality, this is often done months after the exposure. The particles
identifying smoke from a specific wildfire have been identified in some cases as much as
eighteen (18) months after exposure2. This would not have been possible without assemblage
analysis3.

Assemblage analysis is widely used in a number of scientific pursuits though it is not always
called assemblage analysis. It’s applied by medical doctors when they use symptoms to
diagnose a health condition. The diagnosis is not made based on a single symptom but on a

1
Wildfire Smoke Exposure: A Comparative Study

“suite” (assemblage) of symptoms that characterize an illness. It’s used by petrologists to

identify a probable oil-bearing rock formation4,5. It is used by paleontologists and archeologists
to characterize ancient climates and identify ancient cultures6,7. It is also frequently used to
assess water quality8. Assemblage analysis has many applications, including the characterization
of environmental quality and the sources of contaminants in the environment9,10,11,12,13.
Assemblage analysis is basically the use of multiple pieces of information to arrive at a
conclusion not justified by any single component of the assemblage.

As an example, consider a family consisting of a man, a woman, and a child living together in a
home. What is the assemblage? It would be a man and a woman of approximately similar age,
the woman old enough to have a child of that age but not too old to have had that child. It would
include the home and the proximity of the three individuals to one another and to the home.
Toys in the yard consistent with the age of the child would increase the probability of
association. Assemblage analysis is a Bayesian statistical approach to allocating the elements in
an environment to a particular relationship or source. It is possible that the man, woman, child,
and home, with their proximity was a coincidence. Even if the man and woman were kissing and
holding the child that would only increase the probability of our assessment. It is still possible,
though increasingly unlikely, that they are unrelated in any way. But what would be the
probability of the appearance of a man on the street to assuming he was married, lived on this
street, in that house, with his wife, and they had a child if no woman, child, or particular house
seemed related to the man. The assemblage of facts makes a much stronger case for knowing
how the man relates to this neighborhood. The same is true for charred wood.

Environmental Assemblage Analysis takes advantage of the fact that particle sources rarely
generate a pure, single, particle type. It takes a lot of effort and is very costly to make a pure
compound. It doesn’t happen by accident. Emissions from processes tend to show how different
materials exposed to that process are modified by that process.

In the case of a wildfire, leaves may be burned to white ash while wood is splintered and charred
into black fibers or even carbon circles that used to be pores in the wood (Photograph 1).
Grasses may be reduced to carbon darkened silica
phytoliths. Whewellite phytoliths (CaC2O4-H2O,
calcium oxalate monohydrate) may be converted to
calcium carbonate, then calcium oxide, and then back to
calcium carbonate in the plume. Soil may go from
colorless, tan, or slightly bluish to brick red. Each of
these particles not only indicate a fire but the plants that
were burning, what parts of the plants were burning,
how hot the fire was, the weather being created by the
fire, and how fast it was moving. This information is
carried in the crystal structure, morphology, color, Photograph 1: Charred Pine Pores
associations, and reflectivity of the particles. The
2
Wildfire Smoke Exposure: A Comparative Study

crystal structure is not the shape, it is the optical crystallographic information only visible with
polarized light. The morphology is the crystal habit, cellular alignment, pore structure, outline,
texture, internal structure, and other features that relate to the outward and internal shape of the
particle. Color is both the color with transmitted light and with reflected light. These colors may
be very different. Associations relate to different phases in the same particle and to proximity of
related particles. The reflectivity of char increases as the hydrogen in the particle decreases. The
reflectivity may change with linear polarized light. The reflectivity also changes with increased
crystallinity. These are not things that can be measured or detected with an electron microscope.
Aerosolized fire retardant is another particle that is often found associated with the plume of a
wildfire.

Light microscopy is particularly suited to this type of analysis. It is capable of analyzing a

greater variety of particle types than any other single instrument. Thousands of particles can be
characterized in a relatively short period of time as surface areas of a few square inches are
examined. This capability will become evident as we consider the application of the light
microscope to the assemblage analysis of wildfire emissions.

Environmental assemblage analysis using light microscopy in the case exposure to wildfire
smoke can only be performed on tapelifts. Tapelifts are 95% efficient at collecting particles
down to one micrometer. They collect fragile particles and retain vital particle associations.
Tapelifts reliably represent the area sampled. That allows for dependable quantification. Wipe
samples destroy these associations and pulverize delicate particles. Wipes are about 75%
efficient at collecting particles from a surface resulting in a poorly defined surface area1. For
analysis the particles must be removed from the wipes which presents another sampling
dilemma; some particles are easily removed and others are more difficult. Consequently, wipe
samples have very little value for this type of particle analysis.

METHODS

The assessment of smoke exposure has two parts. The first is identifying the presence of smoke
debris from the suspect fire. The second is the quantification of that exposure. Both of these
separate parts of the analysis are quite different using “Assemblage Analysis” compared to the
“Soot, Char, and Ash Analysis” method used by many laboratories for this purpose.

Assemblage analysis has been used as a major tool to identify the source of airborne particles for
since the time of Robert Hooke in the second half of the 1600’s. It has been an important part of
urban air pollution studies since Hooke’s time. In the 1960’s through the 1980’s it was applied
by the United States Environmental Protection Agency (EPA) and by Department of Ecology in
the United Kingdom to identify the sources of urban and indoor air pollution. It was during the
EPA studies of the 60’s and 70’s that charred wood from fireplaces was first identified as a
major source of urban air pollution. This shouldn’t have been much of surprise since peat
burning space heaters were identified as a major cause of polluted air in London during the reign

3
Wildfire Smoke Exposure: A Comparative Study

of Edward the Second at the beginning or the

1300’s. Assemblage analysis was used by the
National Air and Space Agency (NASA) in
the United States to identify sources of
contamination on space craft to better control
the cleanliness of optics in orbit and for deep
space probes. References were all cited in the
introduction. The application of assemblage
analysis to assess wildfire smoke exposure is
an easy extension of this methodology.

Soot, char, and ash (SC&A) analysis seems to

have grown out of laboratories doing fungal
spore identification and/or arson Photograph 2: House Tapelift, No Fire
investigation. Wipe samples are a critical part
of arson investigations but destroy much of This field of view contains 4.3% black
the information needed to identify the source particles by image analysis. Skin flakes,
of the combustion particles. In an arson trichome (plant hair), fragment of plastic,
investigation the source of the combustion natural minerals, and cellulose are also
particles is known, it is the presence of present. The black particles include tire
accelerates that is in question. Wipe samples wear, charred wood from fireplaces, a paint
are not acceptable for wildfire smoke sphere, magnetite inclusions, and toner
exposure investigations where it is the fragments.
particles that must be identified.

Fungal spore identification doesn’t require the use of polarized light or reflected light
examination. The use of clear tape is
apparently adequate for that analysis. The
limitations of applying this technique to the
identification of wildfire emission will shortly
become evident.

Identification of Smoke Debris

Indoor environments are impacted by a wide

variety of combustion sources. Consider first
the internal sources. Cooking is a pyrolysis of
mixed fuels. The house dusts that settle on a
stove top will be burned and become a
pyrolysis product in the home when the unit,
gas or electric, is turned on. Photograph 3: Cigarette Ash, 20.4% Black

4
Wildfire Smoke Exposure: A Comparative Study

Stoves are not the only source of combustion

products in the home. Electric baseboard heaters
have exposed coils that burn whatever has settled
on them. The odor generated when the heat is first
turned on is the result of baking these
accumulated particles. In addition, incandescent
bulbs get very hot and burn skin flakes, paper Photograph 4: Gnat Fecal Pellet (Frass)
fibers, clothing fibers, etc., that then become
airborne and settle on surfaces in the indoor This is frass from a mold eating Gnat on
space. Fireplaces (Photograph 2), cigarettes a tapelift from an attic.
(Photograph 3), cigars, pipes, ovens, electrostatic
filters, heat exchange surfaces in gas or oil
furnaces, use of power saws or drills, candles or
oil lamps, printers and copy machines, and other
activities or products contribute to the combustion
sources in the indoor environment, including
toasters with or without burnt toast.

Outdoor ambient particles are a major contributor

to indoor particle concentrations. Studies have
shown the relationship between indoor and
outdoor particle concentrations is influenced by
Photograph 5: White Ash with Cell
the specific indoor/outdoor (I/O) ratio,
Structure
ventilation, infiltration, and penetration factors15.
These factors vary from one home to another. This type of fragile particle is destroyed
The largest single source of outdoor fine particles when a wipe or vacuum sample is taken.
(PM2.5) entering homes in many American cities This particle is white with reflected light.
is the neighbors’ fireplace or woodstove 16,17
.
They also contribute significantly to the larger particles, ten micrometers and larger, in a home
(see Photograph 2). Another mode for migration is soil adherence to footwear, clothing, pets,
and anything else brought in from outdoors. These particles are brought into the home, released,
and distributed throughout the space.

Combustion products are not the only black particles in an environment. Tire wear, shoe wear,
fretting metal wear, dark minerals, insect debris, fungal debris, decayed plant material, insect and
arachnid frass (see Photograph 4), newspaper ink, toner, cosmetics, pencil debris, etc. are always
or often black and are not uncommon in indoor environments. The products of combustion are
not all black or even dark. When all of the carbon is consumed the result is generally a white,
yellow, or red ash. These particles still retain structure and optical characteristics sufficient to be
identified as ash if they are carefully lifted from a surface (see Photographs 5 and 6).

5
Wildfire Smoke Exposure: A Comparative Study

The identification of these materials requires a mount of high optical quality. That is not
possible with a tapelift unless the plastic backing is removed. The plastic backing is too stiff to
conform to particles thicker than about three quarters of the adhesive film thickness. The
adhesive film is typically ten to twenty micrometers thick (0.01 to 0.02 millimeters). As a result,
particles thicker than about seven micrometers (0.007 millimeters) are often associated with
pockets of air that mask their morphology and their optical properties. If the plastic film is
removed without significantly disrupting the particles in the adhesive layer then the adhesive will
conform to the particles and no optical gaps will be present. With the clear tapes used for SC&A
analysis the plastic backing cannot be removed without serious disruption of the particles in the
adhesive layer. A tape commonly used when a detailed analysis of surface dust is required is 3M
Scotch Brand frosted Magic Tape. This tape has an acrylic adhesive layer for high adhesion and
a cellulose ester plastic film that can easily be

Photograph 8: Laser Speckle

Photograph 7: The Point Pattern of Processed “Magic Tape”
Photograph 9: Laser Speckle
Pattern without a Slide in Pattern of a Typical SC&A
The sharp image of the beam
the Path analysis Tape
shows very little scatter, even
The diameter of the spot is when particles are present. The
The flare and lack of definition
about 3 millimeters with refractive indices of the glass,
in the beam diameter illustrates
some scatter due to the resin, and the adhesive are very
the poor optical quality of the
pointer optics. close, which minimizes aberration
plastic and the interfaces
of the particle image.
between the plastic, adhesive,
and glass slide.
removed using acetone. High
adhesion is required for efficient
particle collection. Acetone has little effect on most environmental particles. The refractive
index of the acrylic adhesive is about 1.48. With a mounting medium of about 1.49 the adhesive
nearly disappears. The result when a glass coverslip is added is a high quality optical mount.

6
Wildfire Smoke Exposure: A Comparative Study

Photographs 7, 8, and 9 show the laser speckle pattern of a laser pointer with no slide present,
with a processed environmental tapelift using frosted Magic Tape, and with an environmental
tapelift using a common tape provided by laboratories using the SC&A analysis technique. The
white spot in the center of these
images is the result of
overexposure.

The laser speckle pattern of the

SC&A analysis tape loses all
coherence and the speckle
pattern expands by a factor of
10 if particles associated with
bubbles or gaps around the
particles are included in the
field. Photograph 8 represents a
Photograph 10: A Processed Magic Tape Tapelift
very clean area of the tape.
This is a photograph taken through a processed Magic
Photograph 10 shows a ruler
Tape tapelift of a plastic metric ruler. The area available
under the slide that generated
for analysis is shown here. It is 4 centimeters by 2
the speckle pattern shown in
centimeters approximately.
Photograph 7. This is the full
area under the coverslip. The
coverslip is a 20X40 millimeter, #0 coverslip. The entire 800 square millimeters are available
for analysis as needed.

Photograph 11 shows a tapelift

typically used for SC&A
analysis on the same ruler. The
tape is colored except for a small
10X10 millimeter window. The
photograph was taken at the
same scale as Photograph 10.
The analyst is restricted to
inspecting the 100 square Photograph 11: Typical Tapelift for SC&A analysis
millimeter area. A clean area of
this sample with no air bubbles This is a photograph taken of the same ruler through a tape
or gaps was used to generate the provided by a typical laboratory doing SC&A analysis.
speckle pattern shown in The analyst is limited to examining an area within the 1
Photograph 8. The plastic film centimeter by 1 centimeter window.
is a polyethylene. The optical
characteristics of polyethylene are inferior to glass, which results in reduced image quality. The
tapelifts shown in Photograph 10 and 11 were taken from clean surfaces.
7
Wildfire Smoke Exposure: A Comparative Study

It is critical to remove air gaps and bubbles around particles. A gap between a particle and the
mounting medium or the mounting medium and the glass slide adds a number of optical effects
that mask the optical properties of the particle. These interfaces cause more light scatter, change
the polarization of the beam of light, and can even completely overwhelm detail of the particle of
interest.

Image aberration due to light scatter is not the only problem with the plastic film. The plastic is
also optically active, meaning that it has different refractive indices in different directions. That
prevents the proper characterization of particles in a number of ways. A critical part of the
analysis of a particle under the microscope is determining how it affects a beam of polarized
light. When the particle is fixed in a plastic film that is optically active subtle but critical effects
can’t be detected or measured. This defect is also present when plastic slides are used. That
includes the depolarization at the interface of a conductive particle and the mounting medium,
rotation of the polarization caused by internal reflection, rotation caused by molecular
asymmetry, anomalous interference effects, polarization on reflection, and other effects.

One final comment on the tape

shown in Photograph 10. The area
available for analysis is inadequate.
A critical member of an important
assemblage may only be present at
parts per thousands. Looking at less
than a few thousand particles in an
environmental analysis is generally
not sufficient to characterize
exposures. This often requires much
more than one square centimeter of a
tapelift.

Wildfire “Assemblage Analysis”

Photograph 12: Forest After a Wildfire (Las Conchas)
Wildfire assemblage analysis is
based on the differences between Most of the wood is still present but the leaves,
wildfires, slash burns, fireplace fires, needles, small twigs and some of the bark have been
burned and were carried away as smoke particles.

8
Wildfire Smoke Exposure: A Comparative Study

backyard plant waste fires, vehicle

emissions, and about forty other
sources of combustion products.
Wildfires are unique and each
biome has its own characteristics.
The aftermath of a forest fire is
shown in Photograph 12. Most of
the wood is still present. The
leaves, twigs, and much of the bark
is missing. Some of the trunk is
burned but much of the trees’ wood
remains. Wood is one of the minor
fuels in a typical forest fire.

Trees are not the only fuel in a Photograph 13: Douglas Fir Needle Char
forest fire. Shrubs, herbs, and This shows the pseudo-cubical calcium oxalate
grasses also burn, along with the phytoliths of Douglas fir and the cell structure of the
duff, degraded biological debris charred needle.
that has accumulated on the forest
floor. Each of these fuels have
distinctive markers as charred
fragments and phytoliths, both
silica and calcium oxalate
(Photographs 13 and 14). The
amount of these other fuels that
burn will vary depending on their
availability in a particular forest
environment and the nature of the
forest fire. A crown fire may miss
most of the ground fuels, at least
initially.

Much of the combustion debris

from wildfires isn’t black. Much Photograph 14: Pine Needle Ash
of the ash is white (see Photograph
This shows part of the needle cell structure, the silica
14). The silica and calcium oxalate
phytoliths (colorless, transparent elongated plates), and
phytoliths are generally not dark
the bipyramidal pyrolyzed calcium oxalate phytoliths.
though the silica phytoliths are
The orange color is due to light scatter. With reflected
sometimes coated with soot. The
light this particle looks white.
calcium oxalate phytoliths are

9
Wildfire Smoke Exposure: A Comparative Study

generally white though they may be associated with charred plant material, as in Photograph 12.

What is burned in a fireplace? The purpose of the fireplace is to maximize the burning of the
wood. Leaves and twigs are not typically burned in the fireplace and the amount of bark burned
is minor compared to the amount of wood. Charred wood is a major combustion product of
fireplaces and woodstoves.

Burning of yard waste, agricultural field burning, and the burning of orchard pruning are all
distinctive sources. Even slash fires or back fires are often distinctive because their purpose is to
remove the “ladder fuels” and tinder from the forest floor. Shrubs, herbs, and grasses are the
main fuel, along with deadfall and duff. The products of these intentional fires are not the same
as the uncontrolled burn of a wildfire.

Forest wildfire smokes are relatively easy to identify due to the characteristic pyrolysis products
of the calcium oxalate phytoliths
that are concentrated in the needles,
leaves, and bark of trees. Chaparral
wildfires leave a different mark (see
Photograph 15). After a chaparral
wildfire the charred trunk and larger
branches of the vegetation are still
present. Just as in the forest
wildfire it has been the leaves,
twigs, dry flower heads, and bark
that are the main fuel. Chaparral is
a mix of plants just as is the forest Photograph 15: Leaf and Twig Char, Adenostoma
but the mix is different.
Adenostoma species, Oak, Adenostoma charred resin droplets are on the left and
Brittlebush, Agave, Sumac, etc. are charred vessel cells from the stock of the plant are on
common in Southern California. At the right. Diamond shaped calcium oxalate phytoliths
higher elevations and in other States are also typical of Adenostoma (not shown). Two
Mountain Mahogany, Feltbush, and Adenostoma species dominate much of the Southern
Sycamore may be added as other California Chaparral. One is known as Red Shanks,
species become less common. Adenostoma sparsifolium, and the other is Chamise,
Every biome is different, just as in Adenostoma fasciculatum.
the case of forest fires.

10
Wildfire Smoke Exposure: A Comparative Study

Savannah or grassland fires also have a

distinctive assemblage (see Photograph
16). The silica phytoliths of the grasses
and the charred fragments of leaves are
easily identified and are not like any
other common non-wildfire source.
Sagebrush, Brittlebush, Rabbit bush, and
dozens of other low shrubs and herbs
add to the assemblage depending on the
location of the savannah. Grasslands in
Southern California are not the same as
the grasslands of Eastern Washington
State. Each location has its own
Photograph 16: Charred Brittlebush Following a
characteristic assemblage. That is why
Grassland Wildfire
knowing the location of the fire is
important to the analysis. This was from a wildfire near Banning, California.
Notice that the grasses and smaller herbs have been
The calcium oxalate phytoliths are
burned to the ground level but shrubs and small trees
thermally sensitive and change their
will be largely intact.
optical properties as seen using
polarized light even though their shape
stays typical of that plant type and that
part of the plant (see Photographs 17 Photograph 17: Pine Calcium Oxalate Phytoliths
and 18). The silica phytoliths retain
their shape but are often coated with a Calcium oxalate phytoliths from pine have this
layer of coked plant material or soot characteristic elongated bipyramidal shape.
(see Photograph 19).

Photograph 19: Charred Silica Phytolith

Photograph 18: Burnt Pine Calcium Pine needles also have characteristic silica
Oxalate Phytoliths phytoliths. These can be a common particle
in pine forest soils but when coated with
Calcium Oxalate passes through a number
soot, as in this case, are the product of
of chemical reaction as it is exposed to heat.
combustion.
It may then go through some additional
chemical reactions in the plume, but the
shape is still recognizable.

11
Wildfire Smoke Exposure: A Comparative Study

A characteristic of wildfires is that they create

their own weather. The powerful convective
air currents carry particles of soil with them.
As these particles are carried through the
combustion zone the iron they contain is
oxidized. Most soils contain biogenic or
leached iron that changes the clay and other
soils brick-red when they are oxidized (see
Photograph 20: Burnt Clay
Photograph 20). This burnt clay becomes
another part of the wild fire assemblage. A fragment of charred Ponderosa pine is
above the two Burnt Clay particles.
Air dropped fire retardant is often used to help
control wildfires. These fire retardants are a
mixture of phosphate and sulfate salts with iron oxide as
a pigment. The pigment is added so that past
applications can be seen from the air. That helps guide
the next application so that the same area is not covered
twice. Photographs taken of an aircraft making these
drops show the main body of the material dropping
vertically and a pink trail of very fine aerosolized fire
retardant. These fine red aerosols become part of the
plume from the fire and can often be found in tapelifts
from homes exposed to the smoke from the wildfire (see
Photograph 21). Photograph 21: Fire Retardant
Aerosol
This collection or assemblage of particles becomes the
unmistakable signature of a wildfire2,14. Fire retardant This particle is from a tapelift
may be absent on occasion. In some environments the collected in a home exposed to
natural red clay can complicate the analysis. In these smoke from a wildfire.
cases it becomes more important to identify more of the
plant types as part of the assemblage to increase the confidence in the conclusion of exposure to
wildfire or the lack of that exposure.

Wildfire “SC&A analysis”

Wildfire SC&A analysis is described by most laboratories as the identification of the presence of
charred biomass, soot, and ash. The biomass is not characterized as to the type other than it is
assumed to be plant, based on visible cell structure. The microscopists in many of these
laboratories are highly trained and very good at what they do but there are a number of
interferences in any sample that can complicate the identification of a particle as a plant
pyrolysis product. These laboratories typically use clear tape applied directly to a microscope
slide. Clear tape presents a number of problems for an analysis of opaque or dark particles. It is
not as clear as optical resins typically used for microscopical analysis. ASTM D7910 requires
12
Wildfire Smoke Exposure: A Comparative Study

only “reasonable optical quality”. Figure 1 and 2

illustrate some of the differences between a clear tape 1
mount and a prepared frosted tape mount. Figure 1 Plastic Film of Tape
shows the optical interfaces in a clear tape mount. The
first interface is between air and the top layer of the 2,3,4
tape. The second layer is between the plastic film and 5,6 Particle
the adhesive. These two materials are reasonably close Microscope Slide
in refractive index and so don’t introduce much scatter.
The third layer is between the adhesive and trapped air. 7
The particle is not in a bubble but with a complex
Figure 1: Clear Tape Interfaces
shape the particle is not uniformly wetted by the
adhesive. The result are gaps that distort the image of Light is scattered by the intrinsic
the particle and create light scatter. The presence of optical quality of the material it passes
these gaps creates another interface, four, between the through and at every interface with a
gap and the particle. The fifth and sixth interfaces are significant difference in refractive
the result of the gap between the particle and the glass index across the boundary. There are
slide. A simple example of this effect can be seen by typically seven interfaces through a
applying a piece of frosted or clear tape to a glass slide, clear tape mount: 1 air to Plastic film,
turning the slide over and looking at the reflection from 2 plastic film to adhesive, 3 adhesive to
the tape-slide interface. The tape will appear darker air gap, 4 air gap to particle, 5 particle
where it has wetted the surface of the slide. In the to air gap, 6 air gap to slide, and 7 slide
other areas the additional interfaces result in reflected to air.
light and areas that are brighter. For particles with
simple shape that are smaller than about five
1
micrometers in thickness the air gaps may not be
present but for many particles the presence of these 2 Coverslip and
gaps is a major limitation in image quality. Optical Resin
3,4
Figure 2 shows the optical interfaces in a mount made 5 Particle
with frosted 3M Scotch Brand Magic Tape. The first
interface is the same with the exception that the top Microscope Slide
layer is an optical glass coverslip and not plastic. The 6
interface between the coverslip and the optical resin is
Figure 2: Prepare Frosted Tape Mount
a near refractive index match as is the interface
between the optical resin and the adhesive. There is no The interfaces: 1 air to glass and, 2
gap between the particle and the adhesive because the matching optical resin, 3 optical resin
acetone soak allows the adhesive to completely wet the to near match adhesive, 4 adhesive to
particle. That removes the presence of any air gap. wetted particle, 5 particle to slide, 6
The result of these cleaner interfaces shows in the laser slide to air.
light scatter shown in Photographs 8 and 9.

13
Wildfire Smoke Exposure: A Comparative Study

When reflected darkfield illumination is used there is more light scatter with clear tape than with
optical resin and a glass coverslip because the light must travel twice through interfaces 1, 2, 3,
and 4. Interfaces 3 and 4, the air gaps between the adhesive and the particle, are the major
problem. Each pass creates more light scatter. Light scatter is like looking though fog. The fine
features that distinguish charred bark from tire wear become impossible to see. The mount is
further compromised by not being flat. The proper identification of dark particles becomes very
difficult. Tire debris, dark minerals, cenospheres, magnetite spheres, and many other interfering
particles can be misidentified as char, false positives. Char can be misidentified as tire wear or
rotted biologicals rather than charred biologicals, false negatives.

The definition of ash creates more uncertainty in the analysis. One reputable laboratory defines
“Ash” as “a high carbon containing particulate that does not maintain its original form”. In other
words, a black particle that has no definitive shape. That would be very hard to distinguish from
tire wear or any number of other black particles that are not the result of combustion. That has
led to problems in some cases (see Case Histories below). That definition of ash also excludes
white ash that may be a considerable part of the total wildfire emissions at a given location (see
Photographs 4, 5, and 13).

The definitions of “Soot” and “Char” are no more well defined than “Ash” which results in the
inclusion of many more possible interferences. Many sources produce charred plant biomass
that are not wildfire related. Charred wood from fireplaces is much more easily identified than
charred bark or fine ash from leaves or other plant parts. Field burning, slash burning, and other
non-wildfire combustion sources produce much more of the type of material identified by this
approach. The burning of candles is identified in these reports as a possible interference for soot
identified as being from wildfire. Basically, the reports indicate that the source of the particulate
matter should not be assumed based on this analysis and that they are not responsible for how
this information is interpreted. Further, the results are unreliable in that any change in the
method could produce very different results of at least equal validity. That leads to the common
and appropriate precaution added to most of these reports shown below. It should also be added
that there are no “published standard methods” for wildfire analysis although their disclaimer
suggests that some exist and they comply18. That is a misstatement.

A Common Precaution Attached to “SC&A analysis” Results

The results are obtained using the methods and sampling procedures as
described in the report or as stated in the published standard methods, and are
only guaranteed to the accuracy and precision consistent with the used methods
and sampling procedures. Any change in methods and sampling procedure may
generate substantially different results. The laboratory assumes no responsibly
or liability for the manner in which the results are used or interpreted.

14
Wildfire Smoke Exposure: A Comparative Study

The limitations of this approach are not necessarily due to the lack of skill on the part of the
analyst but are limitations imposed by the way the analyst is required to perform the analysis.

Quantification of Smoke Debris

The purpose of quantification after a wildfire is to assess the extent of exposure to the emissions
from the fire. In order to determine the amount of combustion material present it must be
quantified based on its amount per unit area and not on the amount of other particulate matter
present. The amount of other debris in that space relates to the general cleanliness of the surface
and has no significance relative to the impact of smoke from the wildfire. The error of
quantification based on a percent of the total particle population is pointed out in another
disclaimer often present on SC&A analysis reports. It points out that the percentage of SC&A
will decrease with time due to the accumulation of other dusts even if no remediation has been
done. The approach of “Assemblage Analysis” and “SC&A analysis” is very different in this
regard. Only the “assemblage analysis” method is independent of the background particle
loading and other combustion sources.

Another complexity associated with Wildfire SC&A analysis compared to Particle Assemblage
Analysis, investigators need benchmarks or thresholds to which they compare their results. There
may be no published data against which field data may be compared. Comparison data may be
based on in-house research or other published information. However, in many cases internal
research and published information may not be appropriate, as this data may not account for
regional variations, site-specific characteristics, variations in collection media or the presence of
alternative combustion sources. Scant agreement among hygienists and their organizations on a
concentration level that would constitute “damage” and remedial action only further complicates
the task18.

Particle Assemblage Analysis suffers from a related problem. Correlating “damage” to the level
of wildfire debris is not part of the analysis. The advantage of assemblage analysis is that it
represents the actual exposure to debris from the wildfire and not simply combustion products or
black particles as a percent of other particles at the site.

“Assemblage Analysis” Quantification

“Assemblage Analysis” Quantification is based on the area of a tapelift that must be examined in
order to see the required wildfire assemblage. It has the advantage of being independent of the
particles that are not related to wildfire. Whether the surface was clean or dirty prior to exposure
doesn’t matter. The presence of a fireplace or woodstove won’t interfere with the results because
such sources don’t contain the wildfire assemblage and the addition of charred wood doesn’t
decrease or increase the area that needs to be scanned in order to find the wildfire assemblage.
The time since exposure with the accumulation of other particles is irrelevant. The sole criterion
is how much of the surface is covered with particles from the wildfire.

15
Wildfire Smoke Exposure: A Comparative Study

The area examined is based on the number of scans across the tapelift required to identify the
wildfire assemblage or to establish that it is not present at a significant level. In some cases, the
exposure is so great that nearly every field of view, about 1/25th of a single scan, contains the
entire assemblage. In this case the exposure is reported as “high”. If a full scan is required, then
the exposure is reported as “moderate”. If a few scans are required, then it is reported as “low”.
If many scans, up to ten or a few more, are required then “trace” is reported. For samples at the
trace level other combustion sources often exceed the amount from wildfire but wildfire debris is
still present. If the assemblage is still not complete after twelve scans, then “non-wildfire” is
reported. That doesn’t mean that there was no exposure but that the exposure at this surface was
so low as to be inconsequential or has already been remediated. The home may be
recontaminated if the environment at large has not been remediated. That includes the exterior
environment that is remediated naturally by weathering conditions. Weathering occurs when
exterior temperatures cycle above and below the dew point, by precipitation, and by aging of the
combustion products that tend to fix them in place.

“SC&A analysis” Quantification

Laboratories using “SC&A analysis” use one of three referenced methods. The most common
method is “Visual Estimate”. Visual estimate involves looking as the sample and estimating the
relative amount of area covered by SC&A compared to the area covered by other types of
particles. This is a notoriously inaccurate method when percentages are under ten percent, even
when the material being quantified is well defined and being done by “experts”19,20. Differences
of at least a factor of two are not uncommon between experts at a level of ten percent and the
difference increases as the percentage decreases. Size also affects estimation. The reader is
encouraged to look at Photograph 1 and 2. The amount of black material in those images was
measured using an image analysis program. Does the amount of black materials in Photograph 2
look like five times as much as the black material in Photograph 1? The human eye is not good
at estimating percent coverage in a field of view.

A second method of quantification used by some laboratories using “SC&A analysis” is to count
particles and generate a percentage based on count. Although it might sound more scientific than
a visual estimate in reality it is not. If all of the particles were the same shape and the same size
it would be fine, but they are not. The analysis may be improved a little by only counting
particles larger than a certain size, say three micrometers or larger. If the particles were all
spheres that would help but do we mean three micrometers in length, in width, in equivalent
spherical diameter, an average of six ferrets, or some other measure? Is there an upper limit
restriction? Is one three micrometer particle the equivalent of a thirty micrometer particle?
Visually estimated area coverage is beginning to sound better.

A third approach attempting to compensate for the defects in the second method is to use a
random point array as the basis for counting. That sounds better but it also suffers from

16
Wildfire Smoke Exposure: A Comparative Study

problems related to sample size, especially for materials at low percentage coverage or for
particles of small size widely distributed 21.

The problem with all of these methods is that they compare the materials called wildfire debris to
the total particle loading in the sample. If a surface was dirty before exposure to the wildfire,
then the wildfire contribution may be small even with a significant exposure. Similarly, if the
surface was clean prior to exposure but the sample wasn’t collected until ten months later, a
common situation, then the contribution of wildfire debris is a lower percent than if the sample
had been collected shortly after the fire, a few months after the fire or even five months after the
fire. This approach is not a measure of the exposure to wildfire debris. It’s a measure of how
clean the environment is relative to the debris from wildfire. That is assuming that the debris
from wildfire has been accurately identified to begin with.

There are better ways of estimating the quantity of wildfire related particles in a sample but they
are also more expensive and are still subject to the accuracy with which the particles are
identified. They are also limited by the accuracy of the estimation of the third dimension of the
particle. The third dimension, thickness, will vary depending on the type of particle and even the
specific particle involved. A flake of mica will be different than a rhomb of a carbonate mineral.
A skin flake will be different than a hair or feather barbule. A rat hair will be different than a
dog or cat hair because their cross-section is different to say nothing or their relative density. A
number percent may seem more accurate but in reality it is still a guess no matter how it is
generated and has little to do with the actual exposure per unit area in the home.

CASE STUDIES

The best way to see how these differences in approach, assemblage analysis or SC&A analysis,
is to compare results with real world examples. All of the laboratories involved in the following
case studies claim to be the experts in the identification of wildfire smoke exposure. They all
claim to be able to defend or have successfully defended their results in a court of law. They will
not be identified by name but all are generally considered to be reputable laboratories. The
purpose of these studies is not to challenge the ability of the analysts but rather the analytical
method used and how the results of the analyses were interpreted.

Case Study 1: Roadside Smoke Debris

A study of sixty-four (64) homes was conducted to assess the impact of the Las Conchas wildfire
to homes in New Mexico22. Each home was sampled using alcohol moistened one inch by one
inch pads. The samples were collected over the late Fall and Winter of 2011-2012, four to nine
months post-fire. The surfaces sampled included windowsills, tops of fan blades, tops of
shelves, and tops of door jams. The area sampled was not recorded. The wiper was applied to
the surface until discoloration was observed. That may have been one wipe or multiple wipes.
These samples were then sent to a laboratory using the “SC&A analysis” procedure described
above. But before this procedure could be applied the particles had to be removed from the

17
Wildfire Smoke Exposure: A Comparative Study

wiper and applied to a microscope slide. How the particles were extracted and how they were
mounted on the slide was not provided.

Fifty of the sixty-four homes were reported as being impacted at some level. The homes that
showed the least impact were the ones furthest from any major road. Most of the homes that
showed a significant impact were on a major highway or within a few hundred feet of a major
highway. This pattern of proximity to a major roadway and what was interpreted as exposure to
the Las Conchas fire was a dominant feature in these results. Distance from the fire had no
effect. Along the same trajectory proximity to a major road was the only correlation to what was
interpreted as exposure to wildfire. Tire wear and vehicle emissions would appear to be what
was being interpreted as SC&A from the wildfire.

Twenty-six homes in this same area were sampled a few months later using Magic Frosted tape
tapelifts and “Assemblage Analysis”. None of these homes had been examined in the first study.
The Las Conchas fire was a forest fire burning Ponderosa pine, some Spruce, shrubs, herbs, and
grasses. This was a large fire and the plume was still evident hundreds of miles away in Texas.
The assemblage associated with this fire included pyrolyzed pine and spruce calcium oxalate
phytoliths, needle ash, bark char, wood char from pine, spruce, and hardwood char from shrubs,
carbon coated silica phytoliths from grasses and needles, deciduous leaf ash and char, burnt clay,
and fire retardant aerosol particles. In some instances, residue from the fire was detected in
homes sampled as much as eighteen months after the fire was controlled.

Of the two hundred thirty-two (232) individual tapelifts analyzed one hundred sixty (160)
showed no wildfire. That was sixty-nine percent of the samples. All of these non-wildfire
tapelifts had tire wear, charred wood, insect debris and other black particles on them. Seven of
the twenty-six homes showed no wild fire emissions on any of the tapelifts collected in the
home. These homes included homes sampled eleven to eighteen months after the fire. Of the
remaining nineteen homes only two showed any tapelifts with levels above trace exposure. Only
one of ten in one home and three of ten tapelifts in the other showed values above a trace level.
One of these homes was in Ojo Caliente and the other was in Truchas. These tapelifts were
collected on surfaces that had never been cleaned since the fire. The tapelifts were from the top
of a light fixture in a bathroom, the top of freezers in the garage, and the top of a beam in the
living room. Both of these homes were near major roadways but many of the homes that showed
no presence or only trace exposure to the wildfire were on or within a few hundred feet of major
roadways.

Finding the assemblage of particles that indicated the Las Conchas wildfire on a few scans across
the tapelifts certainly provides more confidence in the results than the detection of black
particles. The fact that the high SC&A analysis results correlated so well with proximity to
major roadways strongly suggests that it was tire wear and vehicle emissions that were being
interpreted as debris from the Las Conchas wildfire.

18
Wildfire Smoke Exposure: A Comparative Study

Case Study 2: The 1% Rule

In the State of California, the presence of 1% SC&A has been accepted in litigation as proof of
exposure to smoke from a wildfire. Black particles that fit the poorly defined description of
SC&A exceed 1% in most homes with no exposure to wildfire. This is based on studies done by
laboratories that used their SC&A criteria. Another problem with the 1% rule is that it has no
basis in any scientific study. Indoor environments often contain black particles far in excess of
1% regardless of how that 1% might be measured (see below). There are a large number of
environmental combustion sources that are not wildfire. These sources often exceed 1% as
reported in the literature. So 1% doesn’t establish wildfire exposure. Is there some other
percentage that would work? That introduces the next problem.

Reporting a percentage indicates that something has been measured as a baseline and a subset
has been separated and measured to establish a relationship between the subset and the total.
The question is subset of what and how is it measured? It is usually interpreted as 1% of the
total particle load. Is that 1% by area, by weight, by count, or some other parameter. In some
soot reports it is by count, in others it is by estimated total particle area or by estimated weight
assuming some density for the particles in the sample, in still others it is by area sampled. Each
of these would provide very different numbers. Count provides what appears to be scientifically
reliable numbers but what do the numbers mean? One particle 25 micrometers in diameter
should count more than a particle 3 micrometers in diameter. The percent based on count will be
strongly biased toward the smaller particles and in the case of a wipe sample, where delicate char
particles are crushed, would significantly overestimate the amount of soot and char present.

Wipe samples couldn’t be used for estimated particle percent by area either. The collection
efficiency for a wipe on a perfectly smooth surface is at best around 75% but that collection
percentage varies by particle type and by particle size. For surfaces that are not perfectly smooth
the collection efficiency drops rapidly. Wiping a given fixed area doesn’t mean that that area
was sampled with equal efficiency for different particle sizes and types or even that the
collection efficiency for the same type and size of particle was the same over the area wiped.
Wipe samples are quite unreliable as a particle collection technique if quantification is important.

Tapelifts with an adhesive pull strength greater than 30 pounds per square inch (acrylic adhesive)
are about 98% efficient at collecting particles from surfaces whose particle area coverage is 20%
or less and is still about 90% efficient with area coverage of 60% for relatively smooth surfaces.
Tapelifts are the desired collection technique of soot char and ash because they are efficient and
they retain particle configurations and associations that are important for the accurate
identification of the particles. Percent of total particle area or of area sampled could be
determined using tapelifts.

The next problem with the 1% rule is that it is often based not on any measurement but rather on
estimation. Humans are poor at estimating percent coverage of anything. They are much better

19
Wildfire Smoke Exposure: A Comparative Study

at comparing coverage. Some laboratories use a set of standard images representing different
levels of coverage for this reason. It improves the reliability of the estimate in proportion to how
closely the standard look to the sample being examined and the ability of the analyst to integrate
estimates over a number of fields of view. It is still an estimate. Image analysis is not a
significant improvement although it sounds better, because of the additional time required to
determine if a given particle actually is from the fire in question. Image analyzers see the world
in black and white. The greys still have to be interpreted by a human analyst, something humans
are very good at with sufficient training. But the human judgement is still part of the analysis
and the analyst must decide which particles to include. That is the next problem.

Particles in the smoke from wildfires are not all black.

The environment is full of black particles. Near heavily traveled roads tire wear alone can be as
much as 10% of the total particle area on indoor surfaces. Emissions from fireplaces,
woodstoves, and Fall and Winter slash burns can dominate the atmospheric particle burden in the
Fall and Winter. Charred wood and soot and ash from combustion of wood in fireplaces and
wood stoves have been estimated to account for 40% to 60% of the cool weather air pollution in
temperate climates. That is exactly when most wildfire investigations are conducted. Charred
wood is a very common environmental contaminant in homes without exposure to wildfire.
Because of this background charred wood is not an indicator of wildfire exposure by itself.
Charred wood can exceed 1% in a home even without wildfire. Homes with fireplaces or
woodstoves often exceed 1% charred wood as a result of blow-back from wind conditions or
inadequate draft. Even homes without fireplaces or woodstoves can contain more than 1%
charred wood due to the fireplaces and woodstoves of neighbors. The 1% criteria has no
scientific validity.

Case Study 3: Duplicate Analyses, “Soot” vs. Assemblage Analysis

A number of homes with suspected smoke damage from wildfire emissions were sampled with
side by side tapelifts. One set was sent out for “standard SC&A analysis” and the other set was
sent out for assemblage analysis. There were a total of sixty-one paired samples. In each case
the SC&A analysis was performed on clear tape looking at a subset of a 1 centimeter by 1
centimeter area. The assemblage analysis was performed on a tapelift using a cellulose ester
frosted tape (Scotch Brand Magic Tape). After the tape was placed on a microscope slide the
plastic cellulose ester layer removed with acetone. The adhesive with the particles was then
covered with a synthetic resin that was a near match for the refractive index of the adhesive and
the added coverslip. The area examined for assemblage analysis ranged from 0.4 square
centimeters to 8 square centimeters, depending on the area required to find the members of the
wildfire assemblage or determine that it was absent. The chart below shows how the results
compared. The assemblage analysis results were converted to numerical values by assigning
“Non-Wildfire” to “0”, “Trace” to “1”, “Low” to “2”, “Moderate” to “3”, and “High” to “4”.
The assemblage analysis results are shown along the X-axis and the matching SC&A analysis

20
Wildfire Smoke Exposure: A Comparative Study

results are indicated by the Y-axis. All 61 of the matched tapelifts are represented on the chart.
Many of the points represent more than one pair of analyses.

Asemblage Analysis vs Estimated %

25%

20%
Estimated % Char

15%

10%

5%

0%
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
0 = Not Wildfire, 1 = Trace, 2 = Low, 3 = Moderate, 4 = High Exposure

The chart demonstrates that when assemblage analysis detects a high level of exposure to
wildfire smoke the probability of the SC&A analysis also finding a high level increases. It also
indicates that a SC&A analysis value of under 5% can occur at any level of exposure determined
by assemblage analysis, non-wildfire (0) to high exposure (4). Similarly, a value of 5% or more
can occur at any level determined by assemblage analysis.

There were six tapelifts determined to be high wildfire exposure by assemblage analysis that
were reported by SC&A analysis to be under 5% SC&A. Four of these six were from one home
exposed to the Las Pulgas fire. The Las Pulgas fire was finally controlled in May of 2014. The
samples from this home were collected in February of 2015, ten months after the fire was
controlled. All of them were windowsill samples and the total particle loading determined by
obscuration on the tapelifts was in excess of 20% of the total area on some parts of each of the
four tapelifts. The combustion debris on these tapelifts was dominated by white ash and calcium
oxalate phytoliths that had been exposed to high temperatures. The black particles present
included tire wear, fungal debris, charred wood, charred bark, charred leaves, charred resin, and
insect frass and fragments. The charred plant debris include herbaceous plants, some shrubs, and
desert plants. Most of the white ash was from the herbaceous plants and shrubs. The relatively
heavy loading of particles and the moderate loading of black particles was consistent with the
SC&A analysis values reported for other tapelifts with a similar ratio of black particles to total
particle loading. The higher than normal level of white ash from this fire at this location
contributed to the low estimate in two ways. First, white ash didn’t comply with any of the
combustion particle definitions used by the SC&A analysis. It was excluded from the wildfire
particle estimate. Second, since it wasn’t included with the wildfire particulate matter it added to

21
Wildfire Smoke Exposure: A Comparative Study

the particle population of things other than SC&A. The presence of white ash actually reduces
the estimate of combustion related particles using the soot, char and ash protocol.

In another home well away from the home above but also reportedly exposed to the Las Pulgas
fire the SC&A analysis results were about the same as was found in the home cited above but
assemblage analysis indicated far less exposure. These tapelifts had also been collected about
ten months after the fire was controlled. The total amount of black particulate matter was about
the same or a little less than in the case above, but most of the black particles were tire wear.
The tire wear particles were apparently influencing the quantification. The assemblage analysis
results for this home ranged from non-wildfire to low exposure for these matching tapelifts. The
amount of ash and pyrolyzed phytoliths was very much lower or absent on this set of tapelifts.

The SC&A analysis results and the assemblage analysis results were in good agreement at a
home exposed to the Etiwanda fire. The Etiwanda fire was controlled in May of 2014 and the
tapelifts were collected in January of 2015, eight months after the Etiwanda fire was controlled.
Much of the material from the wildfire in this home was charred plant material along with the
rest of the fire assemblage.

A home reportedly exposed to the Colby fire was found by assemblage analysis to have at most a
trace exposure while SC&A analysis reported high values for SC&A exposure. These samples
had been collected four months after the Colby fire had been controlled. The surfaces with the
highest exposure according to the SC&A analysis were taken from a shelf in the garage and the
headboard of a bed in a second floor bedroom. The black particles in the garage were dominated
by tire wear, tailpipe emissions, high hydrocarbon content soot, wear metal, magnetite spheres,
and charred plant material. The black material on the headboard included charred wood, charred
cotton, charred skin flakes and charred plant material. Assemblage analysis identified that there
was some ash and rare pyrolyzed phytoliths on these tapelifts. In preparation of this paper the
notes from the home inspector were revisited. This home had two fireplaces in the home, they
burned incense, and there were two barbeques and a fire pit outside very near the exterior of the
home. That was consistent with the assemblage analysis results that detected combustions
sources that were not from the Colby fire. The SC&A results responded to these combustion
sources, like the incense, but also to the elevated tire wear in the garage.

The limitations imposed by the use of clear tape significantly limited the analytical capability of
the SC&A analysis. That analysis was further restricted by the limited area available for
inspection. The SC&A analysis concentrated on black particles and completely missed cases in
which white ash was a significant part of the combustion particle assemblage. It also tended to
indiscriminately ascribe black particles to wildfire even when wildfire debris was absent. As a
result, SC&A analysis over estimated exposures to wildfire smoke in some cases and under
estimated exposures in other cases, based on paired tapelifts.

DISCUSSION

22
Wildfire Smoke Exposure: A Comparative Study

Assemblage analysis actually looks for the markers that indicate what is burning. Every fire has
different markers and the markers may change as the fire progresses, but they are still the same
plants. What was char from a plant may become white ash as the conditions change. The part of
the plume to which a home is exposed may affect how the assemblage presents itself. In one
investigation the home was clearly exposed to a fire consuming plants similar to those in the
wildfire plume but the concentration of tree related particles was inconsistent with the plume
from the wildfire. A check of activities in the area of the home discovered a backfire had been
set in an area that had been logged a few years before. Many of the understory plants had grown
back but the trees were largely absent. The backfire was to protect this neighborhood. This
home had been impacted by the backfire, not the main wildfire.

Investigating the exposure of a home to wildfire smoke requires some basic familiarity with the
materials being burned and how these materials burn. Photographs 12 and 16 are an example.
Wildfires are very selective in what they burn. Some basic familiarity with the plants in the area
of the wildfire is required. That means the analyst must know the name of the fire involved.
Then reference materials can be reviewed to characterized the biome23,24,25,26. Knowing the
major plants indicates the fuels and what markers should be present. An additional source of
information on the plants that are burning is often available on the internet in area specific
publications27,28. Searching for information on the fire provides photographs of the fire and the
fuel. A review of plants in the area provides more detail on the fuel. Microscope slide libraries
or photographs of debris from wildfires of different types provide the reference materials29.

Applying assemblage analysis to environmental samples requires special training not typically
available from a University curriculum. As in any specialized area of analysis, the analyst
requires specialized training and experience under a mentor. It is not the laboratory that has the
skill, it is the analyst. The analyst must recognize the thermally modified calcium oxalate
phytoliths30. The analyst must recognize the cross-field pitting, the bag cell pores, the structure
of conifer pores, the fine structure of hardwood tracheids, the leaf cell-island structure, grass
silica phytoliths, etc.31,32,33,34.

As was shown in Case 1 and Case 3, assemblage analysis was able to identify the debris from a
specific wildfire and to determine its semi-quantitative concentration on surfaces in the home.
SC&A analyses were basically black particle analyses. It is not for lack of potential skill but the
methodology used and the definitions applied to the particles that severely limit the quality of the
results. The materials generated by a wildfire range from black particles difficult to distinguish
from tire wear or fireplaces emissions, to ghostly white ash that still retains some cell structure or
pyrolyzed calcium oxalate phytoliths. A tapelift sample of high optical quality is required to
identify these materials. Some of the materials that mark the wildfire are present at low
concentrations. It may be necessary to examine many square centimeters of a tape lift to find the
assemblage require to have confidence in the level of exposure or lack thereof.

23
Wildfire Smoke Exposure: A Comparative Study

The clear tape used for SC&A analysis lacks the require optical quality and introduces optical
artifacts due to poor wetting of the particle surface. As a result, the analysis is largely limited to
identifying black particles with slightly more discernment than determining that the particle is
black or at least opaque. There is no provision under the definitions used to identify white ash or
the white calcium oxalate phytoliths characteristic of various plants involved in the fire.

Charred wood is always present to some extent in indoor environments. It comes from
fireplaces, fire pits, wood stoves, distant wildfires, hog-fuel boilers, agricultural burns,
construction activities, yard waste burns, etc. Finding charred wood in a sample without the
other materials characteristic of a specific fire has little significance in regard to exposure to that
fire. It is not surprising that there is little correlation between the results of a SC&A analysis and
an assemblage analysis of the same surface in a home.

A SC&A analysis doesn’t require any information about the fire that is supposed to be the source
of the combustion products. Assemblage analysis requires that the analyst know the source.
Looking for a pine forest fire exposure won’t help if the smoke was from the house next door
that burned down. The answer “no exposure to pine forest fire” is irrelevant to potential damage
from house fire smoke.

In one case a set of samples was provided with a number of sample sets from a wild fire. The
assumption was that this one set of samples was from the same wildfire as the other sets of
samples. The analysis indicated a trace level of exposure to wildfire but the plant assemblage
was different. The analyst added in the e-mail that went with the report that it looked like this
home owner was burning trash, including plastic, in their fireplace. There was a significant
amount of combustion products that were not from wildfire. The conclusion in the report was
that there was a trace of wildfire exposure but it was not consistent with the wildfire in question.

The client was unhappy with the report because they wanted to know about exposure to a
plumbing warehouse fire. The fire was so intense that it had burned the plants in a parking strip
adjacent to the parking lot, including shrubs and trees. This parking strip was between the home
in question and the warehouse fire. This new information explained all of the combustion
products, including the plastic from all of the plastic pipe that had burned. It also explained the
assemblage of plant combustion products that didn’t match the wildfire of the other samples.
The exposure to the warehouse fire smoke was now reported as “high”. The probability that the
home owner would have tolerated the odor of burning plastic in the fireplace was low. The
amount of burned plastic and tar in the home was consistent with the debris that would come
from this plumbing warehouse. The assemblage was there but it was not what the analyst was
looking for. The point of this example is that to control the cost of the analysis the focus is on
one assemblage. The combustion products that don’t belong to that assemblage can be excluded
and don’t need to be characterized beyond the point where it is determined that they don’t
belong. A much more detailed analysis can be conducted but the cost of the analysis increases as
does the time required for the analysis. Cost, inevitably, is a main driver in any analysis.

24
Wildfire Smoke Exposure: A Comparative Study

CONCLUSION

Assemblage analysis is the only approach that actually identifies the emissions from a specific
wildfire as being present in an indoor environment. The initial expectation was that there would
be a high number of false positives using the SC&A criteria. Instead, there were as many false
negatives as there were false positives. Assemblage analysis tended to detect the presence of
emissions from a wildfire in homes when SC&A analysis missed its presence. White ash and
pyrolyzed calcium oxalate photoliths were generally the problem in these cases. Particles that
are not black that come from a wildfire are not recognized in the definitions used for soot, char
and ash analysis. SC&A analysis has never claimed to identify emissions from a specific
wildfire or even to determine if the combustion products are from a wildfire. It appears that the
SC&A protocol isn’t sensitive enough to determine if the majority of black particles in a sample
are combustion or non-combustion particles.

This study demonstrated that false positives and false negatives can dominate the results of a
SC&A analysis. The chart on page 20 shows that values of 5% SC&A are possible when the
presence of the wildfire assemblage is absent. It also shows that values of 5% or less SC&A are
possible when the wildfire assemblage is present at high levels. The problem begins with the
poor definition of what constitutes the SC&A particles. Quantification is the next problem.
How the percentage of those particles attributed to wildfire in each field of view are integrated
across those fields and the sample as a whole into an estimated percentage of the total particle
area is no easy task. No testable procedure is provided except for the counting methods that
suffer from the diverse size problem.

The quantification criteria for exposure used in the assemblage analysis method is independent
of other particles that may be on the surface sampled. It often requires that a surface area in
excess of that used for SC&A analysis be used in order to determine that there has been no
exposure. SC&A analysis often reports exposure to wildfire smoke when assemblage analysis
finds no exposure to that wildfire because combustion products and other black particles are
always present indoors, even without exposure to a wildfire. SC&A analysis can also under
estimate the amount of combustion particles because the quantification procedure used relates
the amount of suspected black combustion particles to the total particle population in the sample.

A requirement for assemblage analysis is that the prepared sample must be of high optical
quality. The tape used has a cellulose ester plastic film with an acrylic adhesive. Acetone is then
used to remove the plastic film and a resin approximately matching the acrylic adhesive’s
refractive index is applied along with a glass coverslip.

The use of clear tape where the plastic is not removed doesn’t have sufficient optical quality for
this type of analysis. The sample typically used for SC&A analysis is of inferior quality. That
significantly limits the confidence in the identification of specific particles.

25
Wildfire Smoke Exposure: A Comparative Study

REFERENCES

1. Crutcher, E. Russ, Ken Warner, and H. K. Crutcher, PARTICLES AND HEALTH:

ENVIRONMENTAL FORENSIC ANALYSIS, PDC 10, PCIH 2007, Louisville,
Kentucky, October 20, 2007, pp. 77-97.
2. Crutcher, Ernest R. and Heidie K. Crutcher-Bettes, Assessing Wildfire Exposure in
Homes near Wildfires, ResearchGate Publication, page 7, 2015.
https://www.researchgate.net/publication/282132868_Assessing_Wildfire_Exposure_in_
Homes_near_Wildfires
3. Crutcher, E. Russ, Ken Warner, and H. K. Crutcher, PARTICLES AND HEALTH:
ENVIRONMENTAL FORENSIC ANALYSIS, PDC 10, PCIH 2007, Louisville,
Kentucky, October 20, 2007, pp. 11-24.
4. Jiang, Dexin, Eleanora I. Robbins, Yongdong Wang, and Huiqiu Yang,
PETROLIPALYNOLOGY, Springer, Beijing, 2016
5. Helgeson, Harold C., Annette M. Knox, Christine E. Owens, and Everett L. Shock,
“Petroleum, oil field waters, and authigenic mineral assemblages are theu in metastable
equilibrium in hydrocarbon reservoirs”, GEOCHIMICA ET COSMOCHIMICA ACTA,
Vol. 57, Issue 14, July 1993, pp. 3295-3339.
6. Martin, Paul S., The Last 10,000 Years: Fossil Pollen Record of the American Southwest,
University of Arizona Press, 1970.
7. Klein, Richard G., Man and Culture in the Late Pleistocene: A Case Study, Chandler
Publishing Company, pp. 8-19, 1963
8. Palmer, C. Mervin, Harold J. Walter, Sharon Adams, Ronald L. Lewis, Algae and Water
Pollution, EPA-600/9-77-036, 1977
9. Crutcher, E. R., “Assemblage analysis—identification of contamination sources”, 2nd
Spacecraft Contamination Conference, NASA-CP-2039, pp. 763-778, 1978.
10. Crutcher, E. R., “Light microscopy as an analytical approach to receptor modeling”, In
Receptor Models Applied to Contemporary Pollution Problems, Air Pollution Control
Association, pp. 267-284, 1982.
11. Hopke, Philip K., Receptor Modeling in Environmental Chemistry, Wiley-Interscience,
1985.
12. Crutcher, E. R., “The characterization of particles on spacecraft returned from orbit”, In
Particles on Surfaces (ed. K. L. Mittal), Marcel Dekker, Inc, NY, pp. 219-252, 1995.
13. Crutcher, E. R., “Microscopy of indoor environmental particles”, Northwest
Occupational Health Conference, Seattle, WA, Oct 8, 1998. Available at
https://www.researchgate.net/publication/265510256_Microscopy_of_Indoor_Environme
ntal_Particles
14. Numerous locations on Microlabgallery.com under “Wildfire”, “Pyrolyzed Phytoliths”,
and other alphabetized titles lead to photographs of these materials.
15. Chen, Chun and Bin Zhao, “Review of relationship between indoor and outdoor particles:
I/O ratio, infiltration factor and penetration factor”, Atmospheric Environment, Volume
45, Issue 2, January 2011. ISSN 1352-2310
16. Ott, Wayne, Stanford University, http://burningissues.org/car-www/index.html

26
Wildfire Smoke Exposure: A Comparative Study

17. Thatcher, Tracy L., Thomas W. Kirchsetter, Christopher J. Malejan, and Courtney E.
Ward, “Infiltration of Black Carbon Particles from Residential Woodsmoke into Nearby
Homes”, Open Journal of Air Pollution, No. 3, pp.111-120, 2014.
18. Kovar, Brad, Martin L. King, and Dr. Priyotosh Chakravarty, “Suggested Guidelines for
Wildfire Smoke Damage Investigations and Remediation”, IICRC Journal of Cleaning,
Restoration and Inspection, June 2015.
19. Sherwood, R. T., C. C. Berg, M.R. Hoover, and K.E. Zeiders, “Illusions in visual
assessment of Stagonospora Leaf Spot of Orchardgrass”, PHYTOPATHOLOGY, Vol.
73, No. 2, 1983.
20. Tonteri, Tiina, “Inter-observer variation in forest vegetation cover assessments”, SILVA
FENNICA, Vol. 24, No. 2, pp189-196, 1990.
21. Dethier, Megan N., Elisabeth S. Graham, Sara Cohen, and Lucinda M. Tear, “Visual
versus random-point percent cover estimations: ‘objective’ is not always better”, Marine
Ecology Progress Series, Vol. 96, pp. 93-100, 1993.
22. Ward, Tony J., “Residential wipe sampling results from the Summer 2011 Las Conchas
Fire”, INTERMOUNTAIN JOURNAL OF SCIENCES, (in press)
23. USDA, NATURAL VEGETATION OF OREGON AND WASHINGTON, General
Technical Report PNW-8, 1973.
24. Collingwood, G.H. and Warren D. Brush, KNOWING YOUR TREES, Revised and
Edited by Devereux Butcher, The American Forestry Association, 1964.
25. Colorado Department of Natural Resources, NATIVE PLANT REVEGETATION
GUIDE FOR COLORADO, Caring for the Land Series, Vol III, October 1998
26. Hanes, Ted L., THE VEGETATION CALLED CHAPARRAL,
http://sheltonherbert.tripod.com/bio1/native_habitats.pdf
27. Plants of the Cajon Pass Area, California State University, San Bernardino,
http://biology.csusb.edu/PlantGuideFolder/LonePine.htm
28. Plants of the Coastal Sage Scrub and Chaparral, California State University, San
Bernardino, http://biology.csusb.edu/PlantGuideFolder/
29. http://www.microlabgallery.com/PyrolyzedCaOxFile.aspx This is just one entry of use on
this website. Photographs of the structure of over 25 different species of trees, calcium
oxalate phytoliths from many other plants, char and ash from many plants, and other
information on both combustion products and non-combustion black particles are
available here.
30. Hoyas, Carmen, Jordi Juan, and Enric Villate, Airborne particles: an approach to calcium
oxalate phytoliths”,
http://www.univeur.org/cuebc/downloads/Pubblicazioni%20scaricabili/PACT%2033%20
Airbonne%20particles%20and%20Gases,%20and%20their%20Impact%20on%20the%2
0Cultural%20Heritage%20and%20its%20Environment/04%20C.%20Hoyas,%20J.%20J
uan%20and%20E.%20Villate.pdf
31. Kukachka, Francis, “Identification of coniferous woods”, TAPPI, Vol 43, No. 11, pp.
887-896, November 1960.
32. Core, H.A., W.A. Cote, and A.C. Day, WOOD STRUCTURE AND IDENTIFICATION,
2ND Ed., Syracuse University Press, 1979.
33. Elmore, Francis H., Shrubs & TREES OF THE SOUTHWEST UPLANDS, Western
National Parks Association, 1976.

27
 Wildfire Smoke Exposure: A Comparative Study

34. Harlow, William M. and Ellwood S. Harrar, TEXTBOOK OF DENDROLOGY:

COVERING THE IMPORTANT FOREST TREES OF THE UNITED STATES AND
CANADA, 4TH Ed., McGraw-Hill, 1958.

28

View publication stats