Full text of DOI:10.1080/10473289.2000.10464163

This article was downloaded by: [Hospital Miguel Servet]
On: 21 October 2014, At: 00:37
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer
House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of the Air & Waste Management Association
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uawm20
Systematic Biases in Measured PM10 Values with
U.S. Environmental Protection Agency-Approved
Samplers at Owens Lake, California
D.M. Ono a , E. Hardebeck a , J. Parker a & B.G. Cox a
a Great Basin Unified Air Pollution Control District , Bishop , California , USA
Published online: 27 Dec 2011.
To cite this article: D.M. Ono , E. Hardebeck , J. Parker & B.G. Cox (2000) Systematic Biases in Measured PM10 Values
with U.S. Environmental Protection Agency-Approved Samplers at Owens Lake, California, Journal of the Air & Waste
Management Association, 50:7, 1144-1156, DOI: 10.1080/10473289.2000.10464163
To link to this article: http://dx.doi.org/10.1080/10473289.2000.10464163
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose
of the Content. Any opinions and views expressed in this publication are the opinions and views of the
authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should
not be relied upon and should be independently verified with primary sources of information. Taylor
and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses,
damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection
with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Ono, Hardebeck, Parker, and Cox
TECHNICAL PAPER
ISSN 1047-3289 J. Air & Waste Manage. Assoc. 50:1144-1156
Copyright 2000 Air & Waste Management Association
Systematic Biases in Measured PM Values with U.S.
Environmental Protection Agency-A1p0proved Samplers at
Owens Lake, California
D.M. Ono, E. Hardebeck, J. Parker, and B.G. Cox
Great Basin Unified Air Pollution Control District, Bishop, California
ABSTRACT
significantly lower than 10 µm under certain conditions.
However, these results are opposite of the bias found for
TEOM samplers in areas that have significant amounts of
volatile particles, where the TEOM reads low due to the
vaporization of particles on the TEOM’s heated filter. Coarse
particles like fugitive dust are relatively unaffected by the
filter temperature. This study shows that in the absence of
volatile particles and in the presence of fugitive dust, a dif-
ferent systematic bias of up to 35% exists between sam-
plers using dichot inlets and high-volume samplers, which
may cause the Graseby and Wedding PM10 samplers to
undermeasure PM10 by up to 35% when the PM10 is pre-
dominantly from coarse particulate sources.
From 1993 through 1998, Wedding or Graseby high-
volume PM10 samplers were collocated with tapered ele-
ment oscillating microbalance (TEOM) samplers at three
sites at Owens Lake, CA. The study area is heavily im-
pacted by windblown dust from the dry Owens Lake bed,
which was exposed as a result of water diversions to the
city of Los Angeles. A dichotomous (dichot) sampler and
three collocated Partisol samplers were added in 1995 and
1999, respectively. U.S. Environmental Protection Agency
(EPA) operating procedures were followed for all samplers,
except for a Wedding sampler that was not cleaned for
the purpose of this study. On average, the TEOM and
Partisol samplers agreed to within 6%, and the dichot,
Graseby, and Wedding samplers measured lower PM10 con-
centrations by about 10, 25, and 35%, respectively. Sur-
prisingly, the “clean” Wedding sampler consistently
measured the same concentration as the “dirty” Wedding
sampler through 85 runs without cleaning. The finding
that the Graseby and Wedding high-volume PM10 samplers
read consistently lower than the TEOM, Partisol, and dichot
samplers at Owens Lake is consistent with PM10 sampler
comparisons done in other fugitive dust areas, and with
wind tunnel tests showing that sampler cut points can be
INTRODUCTION
The dried bed of Owens Lake in Inyo County, CA, is the
largest single source of PM pollution in the United States.
The lakebed covers an area of approximately 110 sq mi
(285 km2) and was a saline lake with no outlet at the ter-
minus of the Owens River. When the Owens River was
diverted to the city of Los Angeles, the lake became virtu-
ally dry by 1930. A small permanent brine pool is present
in the lowest part of the basin, surrounded by the ex-
posed dry alkali soils and crusts. Hundreds of thousands
of tons of PM10 are lofted from the dry lakebed every year,
with daily PM10 concentrations approaching 4000 µg/m3
at monitoring sites at the historic shoreline. Hourly aver-
age wind speeds typically range from 20 to 40 mph (32–
64 km/hr) at 10 m during dust storms when high PM10
concentrations are measured at the monitoring sites. On
average, about 19 PM10 National Ambient Air Quality Stan-
dards (NAAQS) violations occur per year at Keeler, which
is the most frequently impacted monitor site.1
IMPLICATIONS
The results of this study indicate that the Wedding and
Graseby samplers used to determine compliance with the
federal PM10 standards may undermeasure PM10 concen-
trations by up to 35%. This could affect PM10 nonattainment
designations and particulate control strategies in many ar-
eas. Epidemiologic studies that use multi-site and multi-
city monitor data with a mix of PM10 sampler types should
consider incorporating this coarse particle bias into their
statistical analyses to properly compare ambient PM10 and
coarse particle exposure effects on an equal basis. This
study shows that the coarse particulate concentration (par-
ticles between 2.5 and 10 µm) cannot be determined by a
simple concentration difference between a PM10 high-
volume sampler and a PM2.5 sampler.
The Great Basin Air Pollution Control District (Dis-
trict) has been sampling PM in the vicinity of Owens Lake
since 1979. Figure 1 is a map of Owens Lake and the moni-
toring sites. Total suspended particulate (TSP) loadings at
Owens Lake were so high that the material on the filters
would restrict sampler airflow and flow rates would fall
1144 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
Figure 1. Map of Owens Lake and surrounding monitoring stations.
below the U.S. Environmental Protection Agency (EPA) rec-
ommended minimum. Nevertheless, TSP concentrations
for samples that overloaded the samplers were more than
10 times the federal standard of 260 µg/m3, even when it
was conservatively assumed that the total flow volume for
those runs was the air volume for a normal 24-hr run.
Accurate concentration data for Owens Lake can be
obtained only from a sampler that can handle the high
filter loadings. When the Wedding sampler was developed,
it had a volumetric flow design that included a critical
orifice to maintain the flow rate regardless of increased
filter loading. This seemed to promise a better tolerance
for high loadings than the mass flow controller design of
the Andersen Instruments size-selective inlet (SSI) sam-
plers (formerly Sierra-Andersen, Graseby-Andersen, and
General Metal Works—called Graseby in this report). How-
ever, very high PM10 loadings also caused the Wedding
sampler flow rate to decrease. The District purchased three
Rupprecht & Patashnick (R&P) tapered element oscillat-
ing microbalance (TEOM) samplers in order to have a sam-
pler that would not restrict airflow during the heavy
loadings and would allow continuous monitoring. The
TEOM samplers were installed in 1993 in Lone Pine and
Keeler, and in 1994 in Olancha. They were collocated with
Wedding samplers at Keeler and Olancha and with a
Graseby/GMW (G1200) sampler at Lone Pine. In 1995,
Volume 50 July 2000
Journal of the Air & Waste Management Association 1145

Ono, Hardebeck, Parker, and Cox
an Andersen Instruments dichotomous (dichot) sampler
belonging to the California Air Resources Board (CARB)
was installed at Keeler. To further test whether the differ-
ences observed might be due to the PM10 inlet design,
three R&P Partisol samplers were collocated at Keeler over
5 weeks in the spring of 1999, and two Partisols remained
through December 1999.
It was noted that there were systematic differences in
the measurements from the different samplers for the same
location and time period. All of these sampler types are
approved by the EPA for determining compliance with
the federal PM10 standards (40 CFR 50, Appendix J and 40
CFR 53, Subparts C and D). The Wedding, Graseby,
Partisol, and dichot samplers are all Federal Reference
Method (FRM) samplers, and the TEOM is an equivalent
method sampler with an inlet head design identical to
the dichot and Partisol.2-7 Although they are all EPA-
approved samplers for PM10, different samplers incorpo-
rate distinctly different designs, as shown in Figure 2. The
dichot, Partisol, and TEOM samplers use the same inlet
head design, which may be the reason they agreed most
closely in this test.
began sampling at a time other than midnight and ran for
24 hr, TEOM sampler data for the same 24 hr were used.
For the 1-year cleaning study, one Graseby sampler and
a second Wedding sampler were installed at Keeler, and all
sampler heads were cleaned on a weekly basis, except for
one Wedding sampler that, for the purposes of this test, was
never cleaned during the test year. Cleaning was performed
in accordance with the manufacturer’s recommendations.
For the Wedding sampler, this consisted of running a stiff
bristle brush through the inner tube of the sampler inlet. In
addition, the protective housing, insect screen, and vanes
were cleaned at the same time to ensure that particles were
not intercepted prematurely. All samplers were operated in
accordance with EPA operating procedures,8 except for the
“dirty” Wedding sampler that was not cleaned.
Unusual factors in the Owens Valley that might in-
crease experimental errors for sampler measurements are
the large diurnal temperature variations that exist in all
desert environments and the heavy filter loadings. The
In addition, the District investigated the hypothesis
that some of the observed difference may be due to insuf-
ficient cleaning of the heads. Historic data were exam-
ined, and in 1997 the District completed a 1-year study
to assess the influence of cleaning.
A
B
METHODOLOGY
The Wedding, Graseby, and dichot samplers were all sched-
uled to run every sixth day, midnight to midnight, and
in general they have done so since the onset of sampling
at each of the District’s Owens Lake sites. This schedule
should produce approximately 60 runs per year per site,
although runs are sometimes lost due to instrument mal-
functions or other unforeseen events. In an effort to bet-
ter characterize the magnitude and frequency of the severe
Owens Lake dust storms, additional sampling was done
on days when high PM10 concentrations were predicted,
even if they did not fall on the 1-in-6-day schedule. TEOM
samplers provide continuous PM10 data, except during
power or mechanical failures. Two or three collocated
Partisols were run for 52 24-hr periods from April to Oc-
tober 1999 at the Keeler site.
C
D
Five and one-half years of data for collocated TEOM
and high-volume SSI samplers exist for the Keeler site, 5 years
for Lone Pine, and 4 years for Olancha. Comparison of TEOM
sampler data with data collected by the other sampler types
was made for dates where there were valid 24-hr SSI, Partisol,
or dichot sampler runs, and when the collocated TEOM sam-
pler had collected valid data for the same 24-hr period. For
most dates, midnight-to-midnight sampling is compared,
but on dates when the SSI, Partisol, or dichot samplers
Figure 2. PM10 samplers for this monitor comparison study: (a) an
Andersen 241 dichotomous sampler with 246b inlet was operated at
Keeler; (b) R&P TEOM 1400a samplers were operated at Keeler, Lone
Pine, and Olancha; (c) Graseby/GMW G1200 samplers were operated
at Keeler and Lone Pine; and (d) Wedding samplers were operated at
Keeler and Olancha. Not shown are the R&P PM10 Partisol samplers
that were operated at Keeler and have the same PM10 inlet design as
the dichotomous and TEOM samplers in (a) and (b).
1146 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
District uses monthly, as opposed to seasonal, set points to
better compensate for the diurnal temperature variations.
Since the systematic differences between the samplers
persist into the low concentrations, the effects of heavy
filter loadings on flow rates do not appear to be a signifi-
cant source of error. Weighing errors would be reduced,
rather than increased, by the large sample weights. All fil-
ters were carefully installed in the samplers, removed
promptly, and stored and transported in an upright posi-
tion. Samples were hand-delivered to the District’s labora-
tory for processing. Field blanks showed little to no passive
sampling. The District’s filter laboratory is inspected and
certified by the CARB on an annual basis.
although it is heavily influenced by a few high points
and is not as good as the Keeler relationship. Figure 5
plots 270 24-hr periods at Lone Pine in which the Graseby
SSI samples are 76% of the TEOM samples, with a 0.96 R2
value. Therefore, over this 4- to 5-year period, the Wed-
ding samples have been consistently about 35% lower
than the TEOM samples, and the Graseby samples have
been consistently about 25% lower than the TEOM
samples. Note that the TEOM is used as a comparative
reference for these analyses only because it is a common
sampler type at all three sites used in the study.
In Figure 6, the relationship between the collocated
Keeler TEOM and the CARB’s dichot sampler runs is de-
picted. It shows a comparison of dichot and TEOM sam-
pler data for 158 runs when the exposed dichot filters
were weighed by the District before shipping to CARB. A
comparison of dichot filter weights taken before and af-
ter mailing showed that shipping the filters caused mate-
rial to fall off the filter into the filter container. CARB
then either invalidated the sample or recorded a low read-
ing. Some of the coarse filters in the CARB and District
comparison study showed they lost 40–50% of their mass
as a result of shipping the filters in the mail. The study
could not account for particulate mass that could have
been lost between the sampler and the District labora-
tory, so there is no certain way to know the absolute mass
RESULTS
Analysis of All Data
Figure 3 is a plot of data from the Keeler site for 341 24-hr
periods for which there were valid simultaneous collo-
cated runs of Wedding and TEOM samplers. Linear re-
gression analysis shows the Wedding sampler readings are
about 63% of the TEOM readings. The R2 value of 0.98
indicates an excellent linear relationship. The dashed line
(1:1) represents perfect agreement of samples. Figure 4 is
a similar plot of 193 24-hr periods at Olancha in which
the Wedding samples are 57% of the TEOM samples, with
an R2 value of 0.95. This is also a good fit to the data,
Keeler
TEOM vs Wedding SSI
1000
900
800
700
600
500
400
300
200
100
0
1:1
y = 0.63x + 2.38
R2 =
0.98
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
TEOM PM10 (µg/m3)
Figure 3. PM10 measurements from the Keeler Wedding sampler were 63% of the TEOM (341 runs, March 1993–December 1998).
Volume 50 July 2000
Journal of the Air & Waste Management Association 1147

Ono, Hardebeck, Parker, and Cox
Olancha
TEOM vs Wedding SSI
1000
900
800
700
600
500
400
300
200
100
0
1:1
y = 0.57x + 0.65
R2
= 0.95
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
TEOM PM10 (µg/m3)
Figure 4. PM10 measurements from the Olancha Wedding sampler were 57% of the TEOM (193 runs, November 1994–December 1998).
Lone Pine
TEOM vs Graseby SSI
1000
900
1:1
800
700
600
500
400
300
200
y = 0.76x + 0.75
100
R2
= 0.96
0
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
TEOM PM10 (µg/m3)
Figure 5. PM10 measurements from the Lone Pine Graseby sampler were 76% of the TEOM (270 runs, October 1993–December 1998).
1148 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
Keeler
TEOM vs Dichot
1000
900
800
700
600
500
400
300
200
100
0
1:1
y = 0.89x - 2.74
R2 = 0.99
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
1400
1500
TEOM PM10 (µg/m3
)
Figure 6. PM10 measurements from the Keeler dichot were 89% of the TEOM (158 runs, September 1996–December 1998).
loss.9 In this study, the District-weighed dichot samples
approximately the same ratio as in Figure 6 for all data. The
are about 90% of the TEOM samples with a 0.99 R2 value.
relationship between the Keeler TEOM and the average of
the collocated Partisol sampler runs is plotted in Figure 10.
The samplers were cleaned within 2 weeks of each run. For
52 24-hr runs, the average of the Partisol samples is 6% greater
than the TEOM values, with an R2 value of 1.00.
Comparison of Frequently Cleaned Samplers
In order to eliminate sampler cleaning frequency as a
possible cause of the systematic difference between sam-
plers, the District conducted a year-long study of a num-
ber of PM10 sampler designs collocated at the Keeler site.
The samplers consisted of a TEOM, two Weddings, a
dichot, and a Graseby (G1200), and all samples were run
simultaneously. All sampler heads were cleaned weekly,
except for one Wedding sampler that was not cleaned at
all for the purpose of this test.
Figure 7 shows the comparison of the TEOM sampler
and one of the Wedding samplers for 93 24-hr periods dur-
ing the time in which both were cleaned on a weekly basis.
The Wedding sampler values are 60% of the TEOM values—
very close to the 63% derived from the comparison in Fig-
ure 3 for all data regardless of cleaning. Figure 8 shows the
comparison for 111 24-hr periods for the Keeler TEOM and
Graseby samplers. The Graseby sampler values are 79% of
the TEOM values—nearly the same ratio as in Figure 5, which
was for all TEOM and Graseby sampler data at the Lone
Pine site regardless of cleaning. Figure 9 shows the results
for 63 24-hr periods for the Keeler TEOM and dichot sam-
plers. The dichot values are 87% of the TEOM values—
The most unexpected result from this special study
was the comparison of the “clean” (head cleaned weekly)
and the “dirty” (head never cleaned) Wedding sampler
data. Figure 11 shows the results for 85 24-hr periods for
the Keeler clean and dirty Wedding samplers. The read-
ings are identical (slope 1.00, intercept 0.34) with an R2
value of 1.00. The cumulative exposure for filters during
that year was large. By the end of the year, samples
amounting to over 4500 µg/m3 were collected. Figure 12
is a graph of the difference (dirty minus clean) between
the two samplers as a function of cumulative exposure.
All but 5 of the 85 samples fall within 10 µg/m3 of each
other, which is probably within experimental error. The
largest discrepancy occurred early in the testing period
when both samplers were clean.
DISCUSSION
Owens Lake Study
Analysis of hundreds of samples taken at Owens Lake
shows that there is a significant systematic difference
Volume 50 July 2000
Journal of the Air & Waste Management Association 1149

Ono, Hardebeck, Parker, and Cox
y
x
R2
Figure 7. Comparison of Keeler Wedding sampler PM10 measurements to the TEOM when both monitors were cleaned weekly, showing Wedding
equal to 60% of TEOM (93 runs, November 1997–December 1998).
y
x
R2
Figure 8. Comparison of Keeler Graseby sampler PM10 measurements to the TEOM when both monitors were cleaned weekly, showing Graseby
equal to 79% of TEOM (111 runs, November 1997–December 1998).
1150 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
y
x
R2
Figure 9. Comparison of Keeler dichot PM10 measurements to the TEOM when both monitors were cleaned weekly, showing the dichot equal to
87% of TEOM (63 runs, February–December 1998).
y
x
R2
Figure 10. Partisol vs TEOM. All monitors were cleaned within 2 weeks. Partisol equals 106% of TEOM (40 runs of 3 Partisols, April–June 1999; 12
runs of 2 Partisols, August–October 1999).
Volume 50 July 2000
Journal of the Air & Waste Management Association 1151

Ono, Hardebeck, Parker, and Cox
y
x
R2
Figure 11. Comparison of Keeler clean Wedding to dirty Wedding PM10 measurements. Readings are identical to within 1% (85 runs, December
1997–December 1998).
Figure 12. The difference in measured PM10 concentrations between dirty and clean Wedding monitors as a function of cumulative exposure of
dirty monitor.
1152 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
Table 1. Summary of PM10 monitor comparison at Owens Lake, CA.
between the Wedding, Graseby,
TEOM, Partisol, and dichot PM10
samplers. The Partisol and dichot
samplers agree with the TEOM to
within 10%, and the Graseby and
Wedding sampler readings are, re-
Sampler Comparison
& Operating Condition
Ratio of Sampler Measurements (No. of Samples)
Olancha Lone Pine
Keeler
All
DICHOT
Dichot/TEOM
Clean Dichot/Clean TEOM
0.89
0.87
(158)
(63)
–
–
–
–
0.89
0.87
(158)
(63)
spectively, 25 and 35% lower than GRASEBY
Graseby/TEOM
–
–
0.76
–
(270)
0.76
0.79
(270)
(111)
the TEOM readings. A summary of
the results is contained in Table 1.
The District’s study of frequently
cleaned samplers shows that more
frequent cleaning of the Wedding
sampler does not affect the data at
Owens Lake. Even with high cu-
mulative loadings, the Wedding
Clean Graseby/Clean TEOM
WEDDING
Wedding/TEOM
Clean Wedding/Clean TEOM
Clean Wedding/Dirty Wedding
PARTISOLS
0.79
(111)
0.63
0.60
1.00
(341)
(93)
(85)
0.57
–
–
(193)
–
–
–
0.63
0.60
1.00
(534)
(93)
(85)
Clean Partisol/Clean TEOM
1.06
(52)
1.06
(52)
samplers did not appear to be sensitive to cleaning fre-
quency. The relative bias in all the sampler comparisons
appears to be consistent throughout the range of 0–1000
µg/m3, so it does not seem likely that the difference is
caused by high concentrations alone.
The monitor bias that is observed over the range from
zero to over 1000 µg/m3 is consistent even at low values.
A comparison of the monitor bias for PM10 values below
40 µg/m3 at Keeler shows that the monitor ratio for the
Wedding sampler to the TEOM is 0.63 (R2 = 0.72, n = 256),
which is the same ratio as for all data pairs from zero to
over 1000 µg/m3. For the Graseby sampler operating at
Keeler, the ratio to the TEOM is 0.74 (R2 = 0.94, n = 65) for
values less than 40 µg/m3, and 0.79 for all data pairs to
over 1000 µg/m3. This clearly indicates that the bias oc-
curs at low as well as high concentrations and cannot be
attributed to the high concentrations at Owens Lake. This
is also supported by the similar observed biases in New
Mexico (Figure 13) and Arizona PM10 monitor compari-
sons, which do not experience the high concentrations
measured at Owens Lake.10,11
To earn EPA approval as an equivalent method moni-
tor for determining compliance with the federal NAAQS
for PM10, a monitor such as the TEOM must demonstrate
agreement with an FRM sampler to within 10%, or 5
µg/m3, with an R2 value of 0.97 or better (40 CFR, Part 53,
Subpart C). Although the TEOM could pass these toler-
ances at Owens Lake when compared to the Partisol or
PM10 Methods Comparison
Data from collocated TEOM
and Wedding monitors at
Sunland Park, NM and from
collocated monitors at
Anthony, NM
TEOM-1400a
Figure 13. Data from Sunland Park and Anthony, NM, January 1995–March 1997, show the Wedding sampler measurements to be 60–70% of the
TEOM measurements. (Courtesy of New Mexico Environment Department)
Volume 50 July 2000
Journal of the Air & Waste Management Association 1153

Ono, Hardebeck, Parker, and Cox
the dichot samplers, which are approved FRM samplers,
the TEOM would likely fail the equivalency test when
compared to the Wedding and Graseby high-volume sam-
plers. Ironically, the Owens Lake data also show that the
Partisol or dichot samplers would not pass the equiva-
lency test if they were tested against the Wedding or
Graseby samplers, which are also FRM-approved samplers.
reduction in the 50% cut point at higher wind speeds.
The Wedding and Graseby samplers decrease to 8.0 and
8.3 µm, respectively, at 48 km/hr (30 mph),13 which is a
typical high wind condition for wind erosion and PM10
episodes. This cut point decrease would significantly af-
fect sampling for fugitive dust that has a large proportion
of particles in the 6- to 10-µm range.18 The dichot inlet
was not tested at 48 km/hr, but had a cut point of 10 0.5
µm at 8 and 15 km/hr. The dichot inlet should be tested
at 48 km/hr to determine the effect of higher wind speeds
on the cut point.
Monitor Comparison Studies in
Other Fugitive Dust Areas
The result that the Graseby and Wedding high-volume
PM10 samplers read consistently lower than the TEOM,
Partisol, and dichot samplers at Owens Lake is consistent
with PM10 sampler comparisons done in fugitive dust ar-
eas in Arizona and New Mexico. The study in New Mexico
involving collocated Wedding and TEOM samplers
showed the Wedding sampler measuring about 60–70%
of the TEOM values, which agrees well with the compari-
sons at Owens Lake. These data were collected January
1995–March 1997, and consisted of approximately 120
measurements taken at Sunland Park and Anthony, NM.
Figure 13 is a reproduction of Figure 93 of a preliminary
study report.10
In a studyin Rillito, AZ, an area dominated by fugitive
dust, a Wedding sampler produced PM10 concentrations
that were 54% of those produced by a collocated dichot,
and the dichot was about 10% higher than the Sierra-
Andersen 1200 sampler (same as Graseby).11 The slope of
the relationship between the Wedding and dichot samplers
was 0.537; the same relationship (since the dichot and the
TEOM agree within 10%) for our study was 0.63.
The Wedding sampler was also found to have a sig-
nificant cut point decrease to 6.6 µm (at 8 km/hr) when
the inlet was dirty.19 For comparison, a clean Wedding
sampler may have a 50% cut point at 8.6 µm.20 How-
ever, the Owens Lake study did not show a significant
difference between the clean and dirty Wedding sam-
plers, even though they both started out clean and the
dirty sampler was allowed to sample more than 4500
µg/m3 of PM10 over the 1-year test (Figures 11 and 12).
The expected result would have been for the difference
between the monitors to increase as the dirty Wedding
sampler became dirtier. Although a slight 5-µg/m3 dif-
ference may have occurred over the test period, this is
well within the expected precision for properly operated
samplers. A possible explanation is that even though the
cleaning procedures exceeded the manufacturer’s recom-
mendation, the clean sampler was not clean enough. The
clean sampler may have had enough surface roughness
from long-term normal use to induce the same cut point-
lowering effect as the dirty sampler. This could be due to
the increasing dirtiness and viscosity of the oil-coated
surface, which traps large particles that would make the
Wedding sampler inlet more difficult to clean over time.
If this theory is true, the required inlet cleaning for the
Wedding sampler does not improve sampler performance
and the effective cut point of the “used” Wedding sam-
pler is 6.6 µm. This could explain the consistent 35%
lower readings measured at Owens Lake with the Wed-
ding samplers as compared to the dichot-type inlets that
have a 50% cut point around 10 µm. Cut point testing
of used Wedding samplers should be performed to test
this theory.
PM10 Cut Point Differences
Researchers found that high ambient winds and “dirti-
ness” of the Wedding sampler will lower the high-vol-
ume sampler cut points.12-15 This may help explain the
consistent systematic differences among the Graseby,
Wedding, and dichot sampler PM10 measurements found
by other researchers in ambient studies that appear to
show a coarse particle sampling bias.16,17 Dichot studies
are relevant to TEOM sampler performance, since these
two samplers have the same inlet design and, as shown
by this study, agree to within 10%.
In all the studies cited above that were conducted in
areas dominated by fugitive dust, the Wedding sampler
consistently yielded the lowest concentration of all the
PM10 monitors when collocated. Wind tunnel studies to
test inlet performance show that the 50% cut point of
high-volume monitor inlets is lower than 10 µm at high
wind speeds or with dirty monitor inlets. Table 2 shows a
summary of the cut point changes observed in wind tun-
nel tests for different wind speeds and different loadings
or dirtiness of the samplers. All the samplers show a
It has been suggested that the dichot-type inlet could
have particle bounce or blow-off that could cause sam-
plers using those inlets to overmeasure PM10, thus caus-
ing the observed bias. More specifically, it has been
suggested that a dirty inlet would be more likely to expe-
rience particle bounce or blow-off. However, weekly clean-
ing of the monitors during the study did not significantly
affect the observed monitor bias from the bias observed
when monitors were cleaned less frequently. In addition,
the authors did not find any documented evidence in the
1154 Journal of the Air & Waste Management Association
Volume 50 July 2000

Ono, Hardebeck, Parker, and Cox
Table 2. Fifty percent cut points for PM10 sampler inlets based on particle
size testing.
study shows that, in the absence of volatile PM, a system-
atic difference exists between the dichot-type inlets, which
are used on a number of EPA-approved reference method
and equivalent method samplers, and Wedding and
Graseby PM10 inlets. This systematic bias may be caused
by different monitor inlet designs that affect the coarse
particle sampling cut points in the 6- to 10-µm range. For
most fugitive dust sources, this particle size range com-
prises a significant portion of the particulate emissions.
Sampling bias in the PM10 samplers could be attrib-
uted primarily to two factors: a reduction of mass due to
loss of volatile particulates (TEOM sampler), and a reduc-
tion of collection efficiency in the 6- to 10-µm range due
to sampler inlet design (Graseby and Wedding samplers).
These two systematic biases could offset each other in ar-
eas that have simultaneous PM10 contributions from fugi-
tive dust and volatile particulates. For example, the
Graseby or Wedding samplers may measure the fugitive
dust component lower than the TEOM due to a lowered
coarse particle collection efficiency, while the TEOM may
measure the volatile particulate component lower than
the Graseby or Wedding samplers due to the high filter
temperature, resulting in apparent agreement between the
TEOM and Graseby samplers. These offsetting biases could
explain why the TEOM agrees with the Wedding sampler
or the Graseby sampler in some cases where it is known
that volatile particles are present that should be volatil-
ized when captured on a 50 °C filter.
Although the TEOM and Partisol samplers in this
study read about 10–15% higher than the dichot mea-
surements, they should theoretically provide the same
readings for nonvolatile particles since they use the same
inlet design. The small difference between the TEOM/
Partisol samplers and the dichot sampler readings at
Owens Lake may be due to two minor factors. The testing
performed for the TEOM sampler to earn a designation as
an EPA equivalent method resulted in the actual TEOM
reading being increased by 3% to match the reference
method measurements (using a Wedding sampler) at the
equivalency test sites. This 3% increase is calculated in-
ternally for the TEOM output and may represent the re-
sidual effect of compensating for coarse and volatile
particles (slope = 1.03, intercept = 3 µg/m3).24 If the ad-
justed TEOM reading at Owens Lake did not include the
3% increase since the particles are nonvolatile, the differ-
ence between the TEOM and dichot samplers would be
less. Another compensating factor that would bring the
TEOM/Partisol and dichot sampler readings closer is that
it has been visually observed and gravimetrically docu-
mented that for high loadings, particles are lost from the
dichot filters, and the measured dichot mass in these cases
is low.9 Even though dichot filters are carefully handled
and weighed at the District laboratory, some PM10 mass
50%
PM10 Sampler/Condition
Wind Speed
Cut Point
Sierra-Andersen Dichotomous
Clean13
Clean14
Graseby
Clean25
Clean14
Clean13
Dirty14
8 km/hr
15 km/hr
10.3 µm
9.9 µm
8 km/hr
15 km/hr
48 km/hr
15 km/hr
10.0 µm
9.8 µm
8.3 µm
9.1 µm
Wedding
Clean20
Clean14
Clean13
Dirty19
8 km/hr
15 km/hr
48 km/hr
8 km/hr
8.6 µm
9.0 µm
8.0 µm
6.6 µm
7.8 µm
Dirty14
15 km/hr
published literature to show that the dichot-type inlets
oversample PM10. Based on the laboratory tests of the
monitors, however, there is good corroboration that the
Wedding and Graseby high-volume-type samplers may
undermeasure PM10 based on the lowering of the cut
point.14,19,20,25
Monitor Comparison Studies in
Areas with Volatile PM
The results discussed above, which show that the dichot,
Graseby, and Wedding sampler PM10 readings are consis-
tently lower than those of TEOM samplers, are opposite
to the results observed from other sampler comparisons
performed in areas that have significant amounts of vola-
tile PM. Those studies found that the TEOM reads lower
than other samplers due to its elevated filter temperature,
which caused a portion of the volatile PM mass to be va-
porized. A previous TEOM sampler study performed in
Mammoth Lakes, CA, a wood smoke-dominated area,
showed that volatilization of particulates on the TEOM
filter at 50 °C reduced the wintertime PM10 readings by
45% as compared with a Sierra-Andersen 1200 (Graseby)
sampler.21 Another study showed that in the case of the
Los Angeles basin, TEOM readings were up to 24% lower
than those of a Sierra-Andersen sampler.22 This was at-
tributed to ammonium nitrate loss on the TEOM filter
operated at 50 °C. That study also found similar differ-
ences in other areas depending on the amount of volatile
particles in the ambient samples. The TEOM manufac-
turer recommends operating the TEOM sampler at 30 °C
in areas with volatile particulates.
Reconciliation of Observed Monitor Differences
At Owens Lake, PM10 is predominantly from nonvolatile
lakebed soil23 and is relatively unaffected by the elevated
TEOM sampler temperature at 50 °C. The Owens Lake
Volume 50 July 2000
Journal of the Air & Waste Management Association 1155

Ono, Hardebeck, Parker, and Cox
7. Operating Manual TEOM Series 1400a Ambient Particulate (PM10) Moni-
tor (AA Serial Numbers); Designation No. EQPM-1090-079; Rupprecht
& Patashnick Co., Inc.: Albany, NY, Revision A– October 1995, Revi-
sion B–May 1996.
8. Quality Assurance Handbook for Air Pollution Measurement Systems: Vol-
ume II: Ambient Air Specific Methods; Section 2.11; EPA/600/R-94/038a;
U.S. Environmental Protection Agency; U.S. Government Printing
Office: Washington, DC, April 1994.
may be lost from dichot samples. Collocated Partisol and
TEOM samplers at Owens Lake were found to provide
consistent readings with only about 5% difference, which
is within the measurement uncertainty for identical col-
located samplers.
9. Cowell, C.P.; California Air Resources Board, Sacramento, CA, Memo
to C. Lanane, February 27, 1998.
10. Aaboe, E. Analysis of PM-10 Exceedances January 1995-March 1997,
Dona Ana County, New Mexico; Preliminary Report of the Air Quality
Bureau; New Mexico Environment Department: Santa Fe, NM.
11. Thanukos, L.C.; Miller, T.; Mathai, C.V.; Reinholt, D; Bennett, J.
Intercomparison of PM-10 Samplers and Source Apportionment of
Ambient PM-10 Concentrations at Rillito, Arizona. In PM-10 Stan-
dards and Nontraditional Particulate Source Controls; A&WMA: Pitts-
burgh, PA, 1992; pp 244-261.
12. Countess, R.J.; Watson, J.G.; Chow, J.C. Assessment of the PM-10
Measurement Process. In Transactions on PM-10 Implementation of Stan-
dards; Mathai, C.V., Stonefield, D.H., Eds.; Air Pollution Control As-
sociation: Pittsburgh, PA, 1988; pp 179-190.
13. McFarland, A.R.; Ortiz, C.A. Response to Comment on “A Field Com-
parison of PM-10 Inlets at Four Locations;” J. Air Pollut. Control Assoc.
1985, 35(9), 950-953.
14. John, W.; Wang, H. Laboratory Testing Method for PM-10 Samplers:
Lowered Effectiveness from Particle Loading; Aerosol Sci. Technol. 1991,
14, 93-101.
15. Rodes, C.E.; Holland, D.M.; Purdue, L.J.; Rehme, K.A. A Field Com-
parison of PM-10 Inlets at Four Locations; J. Air Pollut. Control Assoc.
1985, 35, 345-354.
16. Purdue, L.J.; Rodes, C.E.; Rehme, K.A.; Holland, D.M.; Bond, A.E.
Intercomparison of High-Volume PM-10 Samplers at a Site with High
Particulate Concentrations; J. Air Pollut. Control Assoc. 1986, 36(8),
pp 917-920.
17. Hoffman, A.J.; Purdue, L.J.; Rehme, K.A.; Holland, D.M. 1987 PM-10
Sampler Intercomparison Study. In Transactions on PM-10 Implemen-
tation of Standards; Mathai, C.V.; Stonefield, D.H., Eds.; Air Pollution
Control Association: Pittsburgh, PA, 1988; pp 138-149.
18. Wilson, W.; Suh, H. Fine Particles and Coarse Particles: Concentra-
tion Relationships Relevant to Epidemiologic Studies; J. Air & Waste
Manage. Assoc. 1997, 47, 1238-1249.
19. McFarland, A.R.; Ortiz, C.A. Sampling Anomalies of the 40-CFM Wed-
ding Aerosol Sampler; Department of Civil Engineering, Texas A&M:
College Station, TX, June 1985.
CONCLUSIONS
At Owens Lake, a fugitive dust area without significant
volatile PM, TEOM sampler PM10 readings are within 10%
of Partisol and dichot sampler readings, and 25–35%
higher than Graseby and Wedding sampler readings. This
study showed that the observed differences are not af-
fected by cleaning schedules or high ambient concentra-
tions. The systematic bias between the samplers is
consistent for 24-hr concentrations up to 1000 µg/m3.
Laboratory studies have shown that the 50% cut points
for the Graseby and Wedding samplers may decrease to
8.3 and 8.0 µm, respectively, at wind speeds of 48 km/hr,
and that the Wedding sampler cut point decreases to 6.6
µm when the inlet is dirty. This cut point decrease would
significantly affect sampling for fugitive dust that has a
large proportion of particles in the 6- to 10-µm range.
There are no reports of significant changes to the dichot
inlet cut point at high wind speeds or when the inlet is
dirty. This study showed no significant difference between
clean and dirty Wedding sampler readings. The results
show that the PM10 inlets used on the Sierra-Andersen
dichotomous sampler, the R&P Partisol sampler, and the
TEOM sampler provide consistent results. Based on cut
point studies by other researchers, these dichot-type sam-
plers may provide more accurate concentrations for PM10
generated from fugitive dust sources.
20. Ranade, M.B.; Kashdan, E.R. An Evaluation of GMW Wedding PM-10
Size Selective Inlet; EPA Wind Tunnel Test Report No. 4; U.S. Environ-
mental Protection Agency; Research Triangle Institute: Research Tri-
angle Park, NC, 1984.
21. Meyer, M.B.; Lijek, J.; Ono, D. Continuous PM-10 Measurements in a
Wood Smoke Environment. In PM-10 Standards and Nontraditional
Particulate Source Controls; A&WMA: Pittsburgh, PA, 1992; pp 24-38.
22. Allen, G.; Sioutas, C.; Koutrakis, P.; Reiss, R.; Lurmann, F.W.; Roberts,
P.T. Evaluation of the TEOM Method for Measurement of Ambient
Particulate Mass in Urban Areas; J. Air & Waste Manage. Assoc. 1997,
47, 682-689.
ACKNOWLEDGMENTS
The authors thank all the staff of the Great Basin Unified Air
Pollution Control District for their assistance. In particular,
we wish to acknowledge Christopher Lanane and Mike Horn
for quality assurance; Chris Rumm, Cheryl Harry, and Scott
Weaver for filter weights; Guy Davis for sampler operation;
and Mike Slates and Ted Schade for graphics.
23. Chester LabNet. Report on Chemical Analysis of Ambient Filters; Report
No. 95-085; Prepared for Great Basin Unified Air Pollution Control
District; Tigard, OR, June, 1996.
24. Patashnick, H.; Rupprecht, E.G. Continuous PM-10 Measurements
Using the Tapered Element Oscillating Microbalance; J. Air & Waste
Manage. Assoc. 1991 41(8), 1079-1083.
25. Ranade, M.B.; Kashdan, E.R. An Evaluation of Andersen Samplers Model
321A 10-mm Size Selective Inlet; EPA Wind Tunnel Test Report No. 3;
U.S. Environmental Protection Agency; Research Triangle Institute:
Research Triangle Park, NC, 1984.
REFERENCES
About the Authors
1. Owens Valley PM10 Planning Area Demonstration of Attainment State
Implementation Plan; Great Basin Unified Air Pollution Control Dis-
trict: Bishop, CA, November 1998.
Duane Ono, Ellen Hardebeck, Jim Parker, and Bill Cox are,
respectively, deputy air pollution control officer, air pollution
control officer, director of data processing, and director of
technical services for the Great Basin Unified Air Pollution
Control District, 157 Short St., Bishop, CA. The authors have
extensive experience in monitoring air quality, estimating
emissions, and developing control strategies for PM gener-
ated from Owens Lake. In 1990, the authors conducted a
study in Mammoth Lakes, CA, that was the first study to
demonstrate that the heated 50 °C TEOM inlet significantly
lowered PM10 measurements in an area with volatile particles.
2. List of Designated Reference and Equivalent Methods; U.S. Environmen-
tal Protection Agency; National Exposure Research Laboratory: Re-
search Triangle Park, NC, June 1999.
3. Operations and Maintenance Manual, The Wedding & Associates’ PM10
Critical Flow High-Volume Sampler, Model 600; Designation No. RFPS-
1087-062; Wedding & Associates, Inc.: Fort Collins, CO, October 1987.
4. Operator Manual PM10 High Volume Air Sampler, Model 1200; Designation
No. RFPS-1287-063; Graseby Andersen/Graseby GMW: Smyrna, GA, 1987.
5. Operating Manual Partisol-FRM Model 2000 PM10 Air Sampler; Designa-
tion No. RFPS-1298-126; Rupprecht & Patashnick Co., Inc.: Albany,
NY, October 1998.
6. Andersen Instruments Dichotomous Sampler Operator’s Manual, Model
SA241 (SA246b inlet); Designation No. RFPS-0789-073; Graseby
Andersen/Graseby GMW, Smyrna, GA, 1989.
1156 Journal of the Air & Waste Management Association
Volume 50 July 2000