PM Sampling IssuesPM Sampling Issues
Dr. Michael BuserBiosystems & Agricultural Engineering Department
Oklahoma State UniversityOklahoma State UniversityStillwater, OK
(405) [email protected]
Perspective
Size Selective Ambient Samplers
Size Selective Stack Samplers
PM10 Samplers – Theoretical Errors0.012
100% 0 04
0.006
0.008
0.010
Den
sity
Ambient PM (MMD - 10 m; GSD 1.5)
Mass 2
60%
80%
100%
atio
n Ef
ficie
ncy
0.025
0.03
0.035
0.04
Mass
Penetration Curve
Mass of particles < 10 m that are captured by the pre-collector (Mass 1)
True Cut
Uniform Particleman
ceid
elin
es
ronm
ent
0.002
0.004
Mas
s
PM captured by the pre-collector (Sampler Cutpoint - 10 m; Slope 1.5)
Mass 1
Note: Mass 1 = Mass
20%
40%
Cum
ulat
ive
Pene
tra
0.005
0.01
0.015
0.02
s Density
Mass of the particles > 10 m that are NOT captured by the pre-collector (Mass 2)
Common Assumption:Samplers produce a "nominal" cut,because it is commonly assumed thatMass 1 = Mass 2. In other words, theerrors offset one another.
The assumption is only valid when thePSD's are described by a uniformdistribution and encompass a sufficient
Uniform ParticleSize Distribution
PA P
erfo
rmC
riter
ia G
ui
deal
Env
ir
0.0000 5 10 15 20 25 30
Particle Diameter (m)
0.012
0.014
Ambient PM (Urban)(MMD - 5.7 m; GSD 2.25)
Note: Mass 1 = 0.65 Mass 2
0%1 10 100
Particle Diameter (m)
0
d s bu o a d e co pass a su c erange of particle diameters.
0.005
0.006
Ambient PM(MMD - 20 m; GSD 1.5)
E C Idm
ent
0.006
0.008
0.010
Mas
s D
ensi
ty
0.002
0.003
0.004
Mas
s D
ensi
ty
PM captured by the pre-collector(Sampler Cutpoint - 10 m; Slope 1.5)
Mass 2
n onm
ent
Env
iron m
0.000
0.002
0.004
0 5 10 15 20 25 30
Particle Diameter (m)
PM captured by the pre-collector(Sampler Cutpoint - 10 m; Slope 1.5)
Mass 2
Mass 1
0.000
0.001
0 5 10 15 20 25 30
Particle Diameter (m)
Mass 1
Note: Mass 1 Mass Urb
anE
nviro
Rur
al
Characteristics of Various Types of Particulate Matter
Particle Source
MMD (m)
GSD
ParticleDensity (g/cm3)
Reference
Urban Urban Dust 5.7 2.25 NR USEPA (1996a) AgriculturalAgricultural Rice 21.75 NR NR Plemons (1981) Rice 12.10 2.24 1.46 Parnell et al. (1986) Corn 19.57 NR NR Plemons (1981) Corn 13.70 NR NR Wade (1979) Corn 13.60 1.80 1.50 Parnell et al. (1986) So beans 25 17 NR NR Plemons (1981)Soybeans 25.17 NR NR Plemons (1981)Soybeans 30.00 NR NR Martin (1981) Soybeans 15.50 NR NR Wade (1979) Soybeans 14.80 1.87 1.69 Parnell et al. (1986) Wheat 32.97 NR NR Plemons (1981) Wheat 14.70 2.08 1.48 Parnell et al. (1986) S h 36 92 NR NR Pl (1981)Sorghum 36.92 NR NR Plemons (1981)Sorghum 15.70 2.16 1.43 Parnell et al. (1986) Cotton Gin (Combined Streams) 20 - 23 1.82 – 2.00 1.8 - 2.0 Wang (2000) Cotton Lint Fibers 12.94 2.25 NR Parnell and
Adams (1979) Cattle Feedlot (Downwind) 14.2 2.25 1.71 Sweeten et al. (1989) S i Fi i hi H (A i l) 14 3 2 02 NR B b t l (1991)Swine Finishing House (Aerial) 14.3 2.02 NR Barber et al. (1991)Swine Finishing House (Settled) 18.4 1.99 NR Barber et al. (1991) Swine Production Facility 17.97 NR NR Barber et al. (1991) Poultry Production Facility 24.0 – 26.7 1.6 NR Redwine and Lacey (2001) Typical Soil 25 2.0 2.5 Pargmann et al. (2000) NR – Data not reported in the reference.
Theoretical Ratios of Ambient PM10 SamplerTheoretical Ratios of Ambient PM10 Sampler to True Concentrations (PSD – GSD = 2.0)
1.8R ti f 5 7 MMD PSD
1.6Ratio range for a 10 m MMD PSD0 9 R i 1 0 ( R i d)
Ratio range for a 5.7 m MMD PSD0.92 < Ratio < 0.99 (a < Ratio < b)Acceptable PM10 sampler measurement to meet PLC138 < x < 149 g/m3 (Ratio * 150 g/m3)
1.4
ncen
trat
ion
entr
atio
n
0.95 < Ratio < 1.05 (c < Ratio < d)Acceptable PM10 sampler measurement to meet PLC142 < x < 158 g/m3 (Ratio * 150 g/m3)
Ratio range for a 20 m MMD PSD1.05 < Ratio < 1.39 (e < Ratio < f)A t bl PM l t t t PLC
f
1.2
Sam
pler
Con
True
Con
ce Acceptable PM10 sampler measurement to meet PLC158 < x < 209 g/m3 (Ratio * 150 g/m3)
d
1.0a < ratio < b, c < ratio < d, and e < ratio < f are the acceptable ratio ranges for 5.7, 10 and 20 m particles, respectively based on the interaction of the PM10 sampler performance characteristics and particle size distribution.
c
eb
a
0.80 5 10 15 20 25 30 35 40
MMD (m)Cutpoint = 10.5 µm; Slope = 1.6 Cutpoint = 9.5 µm; Slope = 1.6Cutpoint = 10.5 µm; Slope = 1.4 Cutpoint = 9.5 µm; Slope = 1.4
Regulated PM10 property line concentration (PLC) = 150 g/m3
Theoretical Ratios of Ambient PM10 SamplerTheoretical Ratios of Ambient PM10 Sampler to True Concentrations (PSD – GSD = 1.5)5.8
4.8Ratio range for a 10 m MMD PSD0 92 R ti 1 07 ( R ti d)
Ratio range for a 5.7 m MMD PSD0.87 < Ratio < 0.96 (a < Ratio < b)Acceptable PM10 sampler measurement to meet PLC131 < x < 144 g/m3 (Ratio * 150 g/m3)
3.8
ncen
trat
ion
entr
atio
n
0.92 < Ratio < 1.07 (c < Ratio < d)Acceptable PM10 sampler measurement to meet PLC138 < x < 161 g/m3 (Ratio * 150 g/m3)
Ratio range for a 20 m MMD PSD1.81 < Ratio < 3.43 (e < Ratio < f)Acceptable PM sampler measurement to meet PLC
2.8
Sam
pler
Con
True
Con
ce Acceptable PM10 sampler measurement to meet PLC271 < x < 514 g/m3 (Ratio * 150 g/m3)
f
1.8
d
e
b
a < ratio < b, c < ratio < d, and e < ratio < f are the acceptable ratio ranges for 5.7, 10 and 20 m particles, respectively based on the interaction of the PM10 sampler performance characteristics and particle size distribution.
0.80 5 10 15 20 25 30 35 40
MMD (m)Cutpoint = 10.5 µm; Slope = 1.6 Cutpoint = 9.5 µm; Slope = 1.6Cutpoint = 10.5 µm; Slope = 1.4 Cutpoint = 9.5 µm; Slope = 1.4
Regulated PM10 property line concentration (PLC) = 150 g/m3
c
a
p
Theoretical Ratios of Ambient PM2 5 SamplerTheoretical Ratios of Ambient PM2.5 Sampler to True Concentrations (PSD – GSD = 2.0)
5
4Ratio range for a 10 m MMD PSD
Ratio range for a 5.7 m MMD PSD0.92 < Ratio < 1.34 (a < Ratio < b)Acceptable PM2.5 sampler measurement to meet PLC32 < x < 47 g/m3 (Ratio * 35 g/m3)
3
once
ntra
tion
cent
ratio
n
0.99 < Ratio < 1.77 (c < Ratio < d)Acceptable PM2.5 sampler measurement to meet PLC35 < x < 62 g/m3 (Ratio * 35 g/m3)
Ratio range for a 20 m MMD PSD1.18 < Ratio < 2.80 (e < Ratio < f)Acceptable PM sampler meas rement to meet PLC
f
2
Sam
pler
Co
True
Con
c Acceptable PM2.5 sampler measurement to meet PLC41< x < 98 g/m3 (Ratio * 35 g/m3)
d
eb
1a < ratio < b, c < ratio < d, and e < ratio < f are the acceptable ratio ranges for 5.7, 10 and 20 m particles, respectively based on the interaction of the PM2.5 sampler performance characteristics and particle size distribution.
ca
00 5 10 15 20 25 30 35 40
MMD (m)
Cutpoint = 2.7 µm; Slope = 1.33 Cutpoint = 2.3 µm; Slope = 1.27
Proposed PM2.5 property line concentration (PLC) = 35 g/m3
Theoretical Ratios of Ambient PM2 5 SamplerTheoretical Ratios of Ambient PM2.5 Sampler to True Concentrations (PSD – GSD = 1.5)
200
160Ratio range for a 10 m MMD PSD2 85 R ti 13 14 ( R ti d)
a < ratio < b, c < ratio < d, and e < ratio < f are the acceptable ratio ranges for 5.7, 10 and 20 m particles, respectively based on the interaction of the PM2.5 sampler performance characteristics and particle size distribution.
f
Ratio range for a 5.7 m MMD PSD1.24 < Ratio < 2.96 (a < Ratio < b)Acceptable PM2.5 sampler measurement to meet PLC43 < x < 104 g/m3 (Ratio * 35 g/m3)
120
once
ntra
tion
ncen
trat
ion
2.85 < Ratio < 13.14 (c < Ratio < d)Acceptable PM2.5 sampler measurement to meet PLC100 < x < 460 g/m3 (Ratio * 35 g/m3)
Ratio range for a 20 m MMD PSD14.81 < Ratio < 183.5 (e < Ratio < f)Acceptable PM2 5 sampler measurement to meet PLC
80
Sam
pler
Co
True
Con
Acceptable PM2.5 sampler measurement to meet PLC518 < x < 6,423 g/m3 (Ratio * 35 g/m3)
40
d
ceb
00 5 10 15 20 25 30 35 40
MMD (m)
Cutpoint = 2.7 µm; Slope = 1.33 Cutpoint = 2.3 µm; Slope = 1.27
Proposed PM2.5 property line concentration (PLC) = 35 g/m3
a
Theoretical Ratios of Stack PM Sampler toTheoretical Ratios of Stack PM10 Sampler to True Concentrations (PSD – GSD = 2.0)
1.8Ratio range for a 5 7 m MMD PSD
1.6Ratio range for a 10 m MMD PSD0 91 < Ratio < 1 08 (c < Ratio < d)
f
Ratio range for a 5.7 m MMD PSD0.87 < Ratio < 1.0 (a < Ratio < b)Acceptable PM10 sampler measurement to meet PLC131 < x < 150 g/m3 (Ratio * 150 g/m3)
1.4
once
ntra
tion
cent
ratio
n
0.91 < Ratio < 1.08 (c < Ratio < d)Acceptable PM10 sampler measurement to meet PLC137 < x < 162 g/m3 (Ratio * 150 g/m3)
Ratio range for a 20 m MMD PSD1.0 < Ratio < 1.60 (e < Ratio < f)Acceptable PM10 sampler measurement to meet PLC
1.2
Sam
pler
Co
True
Con
c Acceptable PM10 sampler measurement to meet PLC150 < x < 240 g/m3 (Ratio * 150 g/m3)
d
eb
1.0a < ratio < b, c < ratio < d, and e < ratio < f are the acceptable ratio ranges for 5.7, 10 and 20 m particles, respectively based on the interaction of the PM10 sampler performance characteristics and particle size distribution.
c
b
a0.8
0 5 10 15 20 25 30 35 40
MMD (m)Cutpoint = 11.0 µm; Slope = 1.79 Cutpoint = 10.0 µm; Slope = 1.9Cutpoint = 9.0 µm; Slope = 1.87
Regulated PM10 property line concentration (PLC) = 150 g/m3
PM10 Stack Sampler Performance Criteria
80
90
100
60
70
cien
cy
30
40
50
Perc
ent E
ffi
PM10 Efficiency EnvelopPM10 Sampler Collection Curves Cutpoint 9 m; Slope 1.87 Cutpoint 10 m; Slope 1.90
Cutpoint 11 m; Slope 1 7617 < Gas Velocity < 27 m/s
10
20
30 Cutpoint 11 m; Slope 1.76y
9 < Gas Velocity < 17 m/s
Gas Velocity < 9 m/s
01 10 100
Aerodynamic Diameter (m)
PM2.5 Stack Sampler Performance Criteria
Theoretical Ambient Particle Size Distribution (Vanderpool, 2010)
120
140
) 90100
Coarse
80
100
120
p (μ
g/m
3 )
607080
on (%
)
Mode
PM2.5Curve
40
60
80
ss/V
dlnD
304050
Pene
trat
io
0
20Mas
01020
P
FineMode PM10
Curve
0.001 0.01 0.1 1 10 100
Aerodynamic Diameter (micrometers)2.5
Stack Sampling - Field Evaluation Results
1.2Feeder Dust
{Based on Method 5 filter PSD}MMD = 6.0 m
GSD = 1.6PM10 = 86.3%PM2.5 = 3.29%
1.2
B Overflow Dust{Based on Method 5 filter PSD}
MMD = 8.0 mGSD = 1.7
PM10 = 66.9%PM2 5 = 1 26%
1.2
#1 A & B Stick Machine{Based on Method 5 filter PSD}
MMD = 6.4 mGSD = 1.8
PM10 = 78.3%PM2.5 = 4.89%
0 4
0.8
Vol
ume
(%)
0.8Vo
lum
e (%
)PM2.5 1.26%
0 4
0.8
Volu
me
(%)
0
0.4
0
0.4
0
0.4
Exhaust % < 10m % < 2 5m % < 10m % < 2 5m % < 10m % < 2 5m
CTM-039 Results PSD Analysis of Method 5 Filter Sampler/True
1 10 100Particle Diameter (m) 1 10 100
Particle Diameter (m)1 10 100
Particle Diameter (m)
Note: PSDs are in t f ESD tExhaust % < 10m % < 2.5m % < 10m % < 2.5m % < 10m % < 2.5m
Stick Machine 73% 28.3% 78% 4.90% 93% 579%
Overflow 67% 16.8% 67% 1.30% 100% 1335%
Feeder 81% 36.0% 86% 3.30% 93% 1095%
terms of ESD not AED (conservative estimates)
AERODYNAMIC DIAMETER: describes a particle’s inertial behavior(V d l 2010)
Dp = 4.0 μm (equiv. physical diameter)ρ = 2 g/cc (particle density) A d i
inertial behavior(Vanderpool, 2010)
ρp = 2 g/cc (particle density)К = 1.3 (dynamic shape factor)
AerodynamicDiameter
Da = 5.0 μmρa = 1 g/cc
Vs = 2.8 m/hrVs = 2.8 m/hr
Da = Dp (ρp/ К ρa)0.5
2004 NRI Grant – Errors Associated2004 NRI Grant Errors Associated with PM Stack Samplers
Ambient Sampler Errors – Field Studies
350%
In Field Measurement
250%
300%
Rat
e (%
)
In Field Measurement
Theoretical (D50=10.5, Slp=1.6)
150%
200%
ampl
ing
R
50%
100%
Ove
rsa
0%1 2 3 4 5 6 7 8 9 10 11 12
Run
Ambient PM10 Sampler – Actual 10
Errors {Cotton Gin}1600
1400
1600
m)
Theoretical Errors - Assuming the Sampler Performance Characteristics
Remain within the EPA defined tolerances
SourceMMD = 12.3 mGSD = 1.94MMD = 13.4 mGSD = 2.0D50 = 24.1 mSlope = 2.9
1000
1200
trat
ion
( g/
acm
T PM S l PM
600
800
PMC
once
nt True PM10 = Sampler PM10
True PM10 = 0.55 * Sampler PM10
R2 = 0.81200
400
True
P
00 200 400 600 800 1000 1200 1400 1600 1800
PM10 Sampler Concentration (g/acm)
Ambient PM10 Sampler – Actual 10 pErrors {Cattle Feed Yard}
True PM10 = Measured PM10
800
900
1000MMD = 18.7 mGSD = 2.2D50 = 19.2 m
True PM10 Conc. = 0.797*Sampler Measured Conc.R2 0 998
600
700
800
atio
n (
g/m
3 ) Slope = 2.4
R2 = 0.998D50=10.5, Slope = 1.6
300
400
500
PM10
Con
cent
ra
True PM10 Conc. = 0.540*Sampler Measured Conc.R2 = 0.602
100
200
300
True
P
00 100 200 300 400 500 600 700 800 900 1000
Sampler PM10 Concentration ( g/m3)
Ambient PM10 Sampler – Actual ErrorsAmbient PM10 Sampler Actual Errors {Almond Orchard – Harvesting}
True PM10 Conc. = 0.84 * Sampler Measured Conc.2000
2500
pR2 = .995
1500
2000
tratio
n (
g/m
3 )
MMD = 15.1 mGSD = 2.0D50 = 11.3 mSlope = 3.3
1000
PM10
Con
cent
True PM10 Conc. = 0.72 * Sampler Measured Conc.R2 = 0.91
500
True
P
00 500 1000 1500 2000 2500
PM10 Sampler Concentration (g/m3)
So why are the actual differences largerSo why are the actual differences larger than the theoretical differences?
3Concentration (g/m3) MMD (m) GSDTSP 1,207 13.4 2
PM 812 11 3 1 80.8
TSP Filter 347 - PSD DataLognormal Fit (MMD = 11.8; GSD = 2.02)
PM10 812 11.3 1.8
Bottom Line!C t i t 24 1
0.4
0.6
al V
olum
e (%
)
Cutpoint = 24.1 m {compared to 10 m}
Slope = 2.9
0.2
Diff
eren
tia{compared to 1.5}
Causes High Concentrations
01 10 100
Particle Diameter (m)
High Concentrations PSD Characteristics Poor sampler placement
Effects of Varying PM10 SPCEffects of Varying PM10 SPC(PSD: MMD = 5.7 m; GSD = 2.25)1.5
( ) R ti 0 93
1.3
(a) Ratio = 0.93Acceptable PM10 sampler measurement to meet PLC = 139 g/m3 (Ratio * 150 g/m3)(b) Ratio =0.99 Acceptable PM10 sampler measurement to meet PLC = 149 g/m3 (Ratio * 150 g/m3)(c) Ratio = 1 05
(d) Cutpoint = 14 m Slope = 2.0
(b) Cutpoint = 10 m
0 8
1.0
once
ntra
tion
cent
ratio
n
(c) Ratio 1.05 Acceptable PM10 sampler measurement to meet PLC = 159 g/m3 (Ratio * 150 g/m3)(d) Ratio = 1.14Acceptable PM10 sampler measurement to meet PLC = 171 g/m3 (Ratio * 150 g/m3)
(c) Cutpoint = 14 m Slope = 1.2
(b) Cutpoint = 10 m Slope = 2.0
0.5
0.8
Sam
pler
Co
True
Con (a) Cutpoint = 10 m
Slope = 1.2
0.3
0.01 3 5 7 9 11 13 15
Sampler Cutpoint (m)
Slope = 1.2 Slope = 1.5 Slope = 2.0 Slope = 2.5
Proposed PM10 property line concentration (PLC) = 150 g/m3
Effects of Varying PM10 SPCEffects of Varying PM10 SPC(PSD: MMD = 20 m; GSD =1.5)
10 0
8.0
10.0(a) Ratio = 1.4Acceptable PM10 sampler measurement to meet PLC = 204 g/m3 (Ratio * 150 g/m3)(b) Ratio =4.4 Acceptable PM10 sampler measurement to meet PLC = 666 g/m3 (Ratio * 150 g/m3)(c) Ratio = 4.8 Acceptable PM10 sampler measurement to meet PLC = 725 g/m3 (Ratio * 150 g/m3)
(d) Cutpoint = 14 m Slope = 2.0
6.0
ncen
trat
ion
cent
ratio
n
(d) Ratio = 7.5Acceptable PM10 sampler measurement to meet PLC = 1,128 g/m3 (Ratio * 150 g/m3)
(b) Cutpoint = 10 mSlope = 2.0
4.0
Sam
pler
Con
True
Con
c
( ) C
Slope 2.0
2.0
(c) Cutpoint = 14 mSlope = 1.2
(a) Cutpoint = 10 m Slope = 1.2
0.01 3 5 7 9 11 13 15
Sampler Cutpoint (m)
Slope = 1.2 Slope = 1.5 Slope = 2.0 Slope = 2.5
Proposed PM10 property line concentration (PLC) = 150 g/m3
Comparing Material Collected from PM10 samplers
1
Both samples were collected using Method 201a (PM sampler)
0.8
Plant A (M-201a)MMD = 3.7 m
GSD = 1.8PM10 = 96%PM2 5 = 27%201a (PM10 sampler)
Filter comparison only Concentration based on
filter mass only
0.6
e (%
) Plant B (M-201a)MMD = 12.9 m
GSD = 1 7
PM2.5 27%
filter mass only Plant A – 48 mg/dscm Plant B – 60 mg/dscm
Concentration < 10 m
0.4Vol
umGSD = 1.7
PM10 = 56%PM2.5 = 1.3%
Concentration < 10 m Plant A – 46 mg/dscm
{48*0.96=46}
Plant B – 34 mg/dscm{60*0 56=34}
0.2
{60*0.56=34}
1 10 100Particle Diameter (m)
0
Questions1) Health based studies – are the PM data
used in the studies comparable?
A. Are we comparing apples to apples?
1
Plant A (M-201a)MMD = 3.7 m
2) If I stand at the property line that separates Plant A and B will Plant B’s (higher PM10 sampler based concentration) emissions more negatively impact my health? 0.6
0.8
%) Plant B (M 201a)
GSD = 1.8
PM10 = 96%PM2.5 = 27%
C=54 mg/dscm
g y p y
3) If I’m evaluating regional PM air quality models using FRM PM sampler concentrations, how good are my modeling results?
0.4Vol
ume
(% Plant B (M-201a)MMD = 12.9 m
GSD = 1.7PM10 = 56%PM2.5 = 1.3%
C=60 mg/dscmg
A. Garbage in – garbage out
4) Are these plants being equally regulated?
0.2
5) How will you answer the same questions for PM2.5?
1) The PSD differences are greater 1 10 100Particle Diameter (m)
0
Dispersion Modeling
30o
~300 feet ~300 feet ~300 feet
Tower Sampler {PM sampler heads and anemometers located at 1.0, 2.0, 3.0, 4.5, 7.25, and 10.0 meters}
Stand Alone Sampler {PM sampler head located at 2.0 meters}
2 Tower Sites:1 – TEOM2 PM10 S l2 – PM10 Samplers2 – PM2.5 Samplers1 - Tower
Recommendations
Development of alternative ambient and stack sampling methodologiesstack sampling methodologies TSP or total particulate matter sampling
coupled with particle size analysisp p y
Development of ambient PM10 and PM2.5sampler placement guidelinessampler placement guidelines
Development of dispersion modeling correction factors for low level sourcescorrection factors for low level sources
Top Related