United States
                    Environmental Protection
                    Agency
Environmental Monitoring
Systems Laboratory
Research Triangle Park NC 27711
                    Research and Development
EPA/600/S4-87/004 Apr. 1987
vvEPA         Project Summary
                   The  PM10  Sampler  Evaluation
                   Program:  January  1985  to
                   July  1986

                   Mark Woods, Fu-Lin Chen, and M. B. (Arun) Ranade
                     Test inlets designed to measure the
                   atmospheric concentration of particu-
                   late matter smaller than a nominal 10
                   micrometers (|im) aerodynamic diame-
                   ter (PM10) were tested at the EPA wind
                   tunnel test facility in Research Triangle
                   Park, North Carolina. The Andersen
                   Samplers Model 321A and the Wedding
                   IP10 size selective inlets (SSI) were eval-
                   uated following the procedures pro-
                   posed in 40 CFR Part 53 (Ambient Air
                   Monitoring Reference and Equivalent
                   Methods). The tests consisted of meas-
                   uring the inlet effectiveness over a
                   range of particle sizes (3 to 20 |im nom-
                   inal aerodynamic diameter) using both
                   liquid and solid test particles at two
                   windspeeds (2 and 8 km/h). Addition-
                   ally two field-tested Wedding IP10 SSI's
                   (EPA field study Phoenix, Arizona, May
                   1986) were compared with  a clean
                   "reference" Wedding IP10 SSI.
                     Before conducting the  above inlet
                   tests, the procedures and requirements
                   of 40 CFR Part 53 were revised to make
                   the test procedures more practical.
                   Several shortcomings in the sampling
                   and analysis procedures as previously
                   practiced were eliminated to ensure
                   compliance with requirements of 40
                   CFR Part 53. At the time the tests were
                   conducted, the wind tunnel  and the
                   wind tunnel test procedures fully met
                   the requirements of 40 CFR Part 53.
                     The Andersen Samplers Model 321A
                   and the Wedding IP10 SSI's were both
                   found to meet the 50-percent cutpoint
                   requirement at 2 and 8 km/h for both
                   liquid and solid particles.  The "used"
                   Wedding SSI's did not meet the 50-
                   percent cutpoint criterion. However, all
                   samplers met the expected mass ratio
                   requirement.
  This Project Summary was devel-
oped by EPA's Environmental Monitor-
ing Systems Laboratory, Research Tri-
angle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).

Introduction
  On March 20, 1984, the U.S. Environ-
mental Protection Agency (EPA) pro-
posed revisions to the National Ambi-
ent Air-Quality Standards (NAAQS) for
paniculate matter. New primary stand-
ards were proposed for particulate mat-
ter measured as PM10 (particles with
aerodynamic equivalent diameters less
than a nominal 10 (im). A new Federal
Reference  Method (FRM) for the deter-
mination of PM-ig in the atmosphere and
provisions for the designation of refer-
ence and equivalent methods for PM10
determination also were proposed.
Under the  provisions of these proposed
regulations, candidate  reference and
equivalent methods for PM10 determi-
nation would have to be tested  in ac-
cordance with the explicit procedures
contained  in 40 CFR Part 53 (Ambient
Air Monitoring Reference and Equiva-
lent Methods). Following satisfactory
completion of all test  requirements,
candidate methods would be desig-
nated formally as reference or equiva-
lent methods. The proposed testing re-
quirements of Part 53 include wind
tunnel tests for sampling effectiveness
and a 50-percent cutpoint. Sampling ef-
fectiveness is the ratio of the mass col-
lected by a test inlet compared to the
mass collected by an isokinetic high-
volume sampler. The 50-percent cut-

-------
point is the particle size that is collected
by the test inlet with 50 percent effi-
ciency.
  Sampling effectiveness is determined
in the following  manner.  A mono-
disperse aerosol of known size is gener-
ated using a Berglund-Liu vibrating ori-
fice aerosol generator. The aerosol  is
introduced into the wind tunnel so that
a uniform distribution of particles  is
achieved in the test section. The unifor-
mity is evaluated by using an array  of
isokinetic samplers (referred to as an
isokinetic rake) to measure the aerosol
concentration  profile in the sampling
zone of the test section. To be accept-
able, the  concentration at each sample
point on the rake must be within  10 per-
cent of the mean concentration of all
sample points on the rake. After sam-
pling the concentration  profile, a test
inlet is placed in the wind tunnel and
operated  at a  nominal flow rate of 40
cubic feet per minute (ft3/min) for 20
min. A high-volume isokinetic sampler
(40 ft3/min) is then run for 20 min  to
determine the true  wind tunnel aerosol
concentration. The aerosol  mass  con-
centration collected by both the test
inlet and  the isokinetic sampler is deter-
mined  using fluorometry. Liquid aero-
sol particles contain  uranine and solid
particles are  ammonium fluorescein,
both of which are fluorescent materials.
All  inlet  tests covered in this project
summary included two  inlets run se-
quentially during a single test. The pro-
cedure was to first run the isokinetic
rake and then three replicates of the
series consisting of test inlet one, the
isokinetic high-volume inlet, and test
inlet two.
  During  most of calendar year 1984,13
tests were performed on various inlets
at the EPA test facility. During most  of
calendar  year 1985, shortcomings in the
sampling and analysis procedures were
identified and eliminated, and in  1986
inlet tests were resumed.
  The aerosol distribution system  used
during 1984 worked well at windspeeds
of 2 and  8 kilometers per hour (km/h).
However, at a windspeed of 24 km/h,
the aerosol concentration was high  at
the center of the wind tunnel and low
near the  sides. The aerosol distribution
system was modified (largely by trial
and error) until acceptable performance
was achieved  at each of the three re-
quired windspeeds.
  After the aerosol distribution  system
was redesigned,  it was necessary  to
measure again the velocity and turbu-
lence profiles in the test section. The ve-
locity and turbulence profiles  were
measured using a TSI Model 1050-1
Anemometer and a ruggedized probe
along with  a  TSI  Model  1056 RMS,
mean-square,  direct current (dc) volt-
meter. The probe was calibrated on a
TSI Model 1125 calibrator, and the cali-
bration  curve  was linearized using a
fourth order least squares polynominal
fit to the original data. The reproducibil-
ity of the velocity profiles also was  de-
termined  over the period  of  several
weeks.
  The isokinetic  high-volume sampler
was used  as the aerosol concentration
reference for all sampling effectiveness
calculations. Because a single isokinetic
high-volume measurement was used as
a reference for two test inlets (one  run
before the isokinetic reference and one
run after the isokinetic reference), it was
necessary to determine the repeatabil-
ity of duplicate isokinetic concentration
measurements. This was accomplished
by making a series of four or five con-
secutive isokinetic high-volume  sam-
pler concentration  measurements.  Ad-
ditionally, it was necessary  to stop  the
wind tunnel fan to replace inlets in  the
wind tunnel. Therefore, the effect on
concentration of stopping the wind tun-
nel fan between isokinetic sampler runs
also was investigated.
  The rake was used to measure aero-
sol uniformity, and the  high-volume
sampler was used to measure the wind
tunnel aerosol concentration. Both  are
isokinetic sampling devices, and there-
fore the two methods should yield iden-
tical concentration values. In reality,  a
discrepancy exists, and a  record  has
been kept on the quality of agreement
obtained when both measurements are
made.
  Fluorometry is the analytical tech-
nique used to determine the mass con-
centration of aerosol particles collected
by a test inlet. An SLM Aminco Fluoro-
Colorimeter II is used at the EPA wind
tunnel.  A calibration curve was devel-
oped by preparing serial dilutions  of a
uranine solution (liquid particles) or of
an ammonium fluorescein solution
(solid particles). The precision of the in-
strument was determined by measuring
the fluorescence of five samples  of the
same solution concentration.
  After  the test protocol was modified,
the inlet tests were resumed. First, the
Wedding  IP10 SSI  was run at  a  wind-
speed of 8 km/h using liquid and solid
aerosol particles.  This was a  prelimi-
nary test designed chiefly to evalual
the new test protocol. Following thi
preliminary test, the Wedding IP10 SJ
and  the Andersen Samplers Mode
321A SSI were tested  using both soli
and  liquid aerosol particles  at wine
speeds of 2 and 8 km/h. A "dirty" fiek
tested Wedding IP10 SSI (Phoenix, Ar
zona, May 1986) then was tested wit
the "clean" Wedding SSI used in prev
ous wind tunnel tests at a windspeed c
2 km/h with liquid particles. Liquid part
cle tests then were conducted  at 2 km/
on a cleaned, field-tested Wedding IP-
SSI and the "clean" Wedding  referenc
inlet.

Results
  The final aerosol distribution configi
ration consisted of a six-point injectio
system, a 4 ft x 4 ft baffle located jus
upstream of the injection points, a 16-i
diameter mixing fan located just dowr
stream of the injection points (and d
reeled into the bulk air flow),  and flov
straighteners located just upstream c
the 4 ft x 4 ft baffle. Using this configu
ration, the aerosol concentration a
each rake nozzle was found to  be withii
±10 percent of the mean concentratioi
at 2, 8 and 24 km/h.
  The velocity and turbulence profile
again were measured. Atypical velocit
and  turbulence  profile obtained at
windspeed of 8 km/h  is shown in Fie
ure 1. The air velocity is within 10 pei
cent of the mean velocity, and  the turbi
lence intensity is less than 5 percent a
all points measured in the test sectior
The  maximum deviation in day-to-da
widespeed measurements at a give
point was 1.7 percent, and the max
mum deviation in turbulence intensit
was 1.0 percent.
   Concentration stability was detei
mined by making a series of  isokineti
high-volume sampler measurements £
each of the three required windspeed
(2, 8, and 24 km/h). The coefficient c
variation (the standard deviation d
vided by the mean) was less than 3 pei
cent at  each of the three windspeeds
The isokinetic measurements takei
after stopping the wind tunnel fan be
tween runs all fell within one standan
deviation of their respective means ex
cept in  one  case where the concentra
tion was within two standard deviation
of the mean.
   A record of the agreement betweei
concentration values  indicated by th
isokinetic rake and the isokinetic  high
volume sampler has been kept for run

-------
10


 9
  I'
              Velocity Profile
              fb
                                                  * = 6" above centerline
                                                  9 = Centerline
                                                  O = 6" below centerline
                                                 •no
                           I
                                                 j_
                          18         30         42
                       Distance from East Wind Tunnel Wall (in)
                                                        54
      5.0
   a
     4.0
      3.0
               Turbulence Intensity
                                                              O
                                                              +
         •
         O
                                      I
                                                 j_
              6           18         30          42          54

                       Distance from East Wind Tunnel Wall (in)

Figure 1.    Velocity and turbulence profiles in the test section at 8 km/h.
in which  both measurements were
made. The concentration indicated by
the isokinetic rake has been on average
7 percent higher than the concentration
indicated by the isokinetic high-volume
sampler.
  The relative error of the fluorometer is
calculated using the equation
        RE = (p x 100%)/(x)        (1)
where
  p = instrument precision (fxg/mL)
  x = measured concentration (n,g/mL).
                                   The precision of the fluorometer was
                                 determined over its range of operation
                                 using uranine solutions of 1, 0.1, 0.01,
                                 and 0.001 p,g/mL. Six samples were pre-
                                 pared at each concentration mimicking
                                 the procedures used to handle filter
                                 samples from an actual inlet test. Fluo-
                                 rometer readings were obtained for
                                 each solution concentration and con-
                                 verted to uranine concentrations using
                                 a current calibration curve.  The  stand-
                                 ard deviation of replicate uranine con-
                                 centration values was  taken to be the
precision of the fluorometer. The rela-
tive error (required to be less than 5 per-
cent in 40 CFR Part 53) then was calcu-
lated using Equation 1. A plot of uranine
concentration versus relative error
shows that a  uranine concentration of
0.014  (jig/mL corresponds to a relative
error of 5 percent.  Uranine concentra-
tions encountered  in the wind tunnel
are typically much greater than 0.014
|j.g/mL except when small particles (3 or
5 n-m) are used at high windspeeds (24
km/h). A new rake flow system is being
installed to provide for the collection of
greater quantities of uranine in the
same amount of sampling time.
  The two most important parameters
obtained from the inlet tests are the 50-
percent cutpoint and the expected mass
ratio. The 50-percent cutpoint was de-
fined  previously. The expected mass
ratio is the mass collected by the test
inlet (found by integrating the  inlet's
sampling effectiveness curve against an
assumed aerosol size distribution) com-
pared  to the mass predicted for an
"ideal" sampler. The sampling effec-
tiveness values for the ideal sampler
and the aerosol size distribution are
both found in 49 FR, 10461, Table  D-3.
The sampling effectiveness curve must
be such that the test inlet has the 50-
percent cutpoint between 9.0 and  11.0
(jim and  an expected mass ratio be-
tween 0.90 and 1.10. The results of the
inlet tests are summarized in Table 1.
Except where the test inlet is designated
as a field test inlet, the same Wedding
IP10 SSI and the same Andersen Model
321A SSI were used.
  The  two field-tested Wedding  IP10
inlets  do not  meet the 50-percent cut-
point criterion; however,  all  inlets
tested meet the expected mass ratio re-
quirement. Both the Wedding IP10 SSI
and the Andersen Samplers Model
321A SSI meet the 50-percent cutpoint
requirement at windspeeds of 2 and 8
km/h using both solid and liquid aerosol
particles.

Conclusions and
Recommendations
  The changes made in the test protocol
at the EPA wind tunnel test facility have
made it possible to meet all require-
ments set forth in 40 CFR Part 53 regard-
ing wind tunnel tests.
  The wind tunnel  tests performed on
the Wedding  IP10 and Andersen Sam-
plers Model 321A SSI's show that both
inlets  meet the proposed specification
for 50-percent cutpoint and sampling ef-

-------
   Table 1.    Summary of Inlet Test Results
Test inlet
Wedding IPW
Wedding IPW
Wedding IPW
Wedding IPW
Wedding IPW
Wedding IPW
Wedding IPW
("Dirty" field-
tested inlet)
Wedding IPW
Wedding IPW
("Cleaned" field
tested inlet)
Model 321 A
Model 321 A
Model 321A
Model 321 A
Windspeed
(km/h)
a
8
8
2
2
2
2


2
2


8
8
2
2
Particle
type3
L
L
S
L
S
L
L


L
L


L
S
L
S
50-percent
outpoint
(pm)
9.20
9.20
9.30
9.35
9.35
9.35
6.95


9.35
8.70


10.40
10.60
10.50
10.20
Slope
(Vg)"
1.37
1.32
1.37
1.28
1.28
1.30
1.68


1.28
1.36


1.44
1.48
1.39
1.40
Expected
mass
ratio
0.976
0.985
0.990
0.989
0.987
0.989
0.936


0.989
0.965


1.021
1.025
1.030
1.026
   aL = Liquid, S = Solid.
   bThe slope of the sampling effectiveness curve is estimated by the parameter cra =
                                                U.S. Environmental  Protection Agency
                                                under Contract  Number 68-02-3992 to
                                                Research Triangle Institute (RTI). It has
                                                been subjected to the Agency's peer
                                                and administrative review, and  it has
                                                been approved for  publication  as an
                                                EPA document.
   fectiveness at windspeeds of 2 and
   8 km/h using both solid and liquid aero-
   sol particles. The two field-tested Wed-
   ding IP10 SSI's met the sampling effec-
   tiveness criterion but not the 50-percent
   cutpoint criterion.
     The new rake flow system should be
   completed before tests are run at 24 km/
   h, to ensure adequate quantities of ura-
   m'ne are  collected for fluorescence
   measurements.
     The  sampling effectiveness  curves
   generated for the test  inlets  have not
   been corrected for the presence of mul-
   tiplets. Multiplets occur when  two or
   three particles aggregate forming a sin-
   gle larger particle. At present, no single
   algorithm for the correction  of multi-
   plets has been widely accepted. A single
   correction method should be developed
   and applied consistently to all sampling
        effectiveness  measurements involving
        the PM10 test  inlets.
          Th§ information in this document has
        beerrfiunded  wholly or in part by the
          Mark Woods, Fu-Lin Chen, and M. B. (Arun) Ranade are with Research Triangle
            Institute (RTI), Research  Triangle Park, NC 27709.
          Kenneth A. Rehme is the EPA Project Officer (see below).
          The complete report,  entitled "The PMto Sampler Evaluation Program: January
            1985 to July 1986." (OMer No. PB 87-145 801/AS;  Cost: $18.95. subject
            to change) will be available only from:
                  National Technical Information Service
                  5285 Port Royal Road
                  Springfield, VA 22161
                  Telephone: 703-487-4650
          The EPA Project Officer can be contacted at:
                  Environmental Monitoring Systems Laboratory
                  U.S. Environmental Protection Agency
                  Research Triangle Park. NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
'"••"-
:  :c-^~  0,22
Official Business
Penalty for Private Use S300
EPA/600/S4-87/OC4

-------