United States Office of Air Quality EPA-450/4-83-002
Environmental Protection Planning and Standards May 1981
Agency Research Triangle Park NC 27711
__
v>EPA Field Study to
Determine Spatial
Variability Of Lead
From Roadways
-------
EPA-450/4-83-002
Field Study To Determine
Spatial Variability Of
Lead From Roadways
by
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68023013
Task Order No. 7
PN 3366-G
EPA Project Officer: Mr. David Lutz
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Monitoring and Data Analysis Division
Research Triangle Park, N.C. 27711
May 1981
-------
DISCLAIMER
This report was written for the U.S. Environmental Protection Agency
Monitoring and Data Analysis Division by PEDCo Environmental, Inc., Cin-
cinnati, under contract No. 68-02-3103, Task Order No. 7. The contents of
this report are reproduced,herein as received from the contractor. The
opinions, findings, and conclusions are those of the author and do not neces-
sarily reflect the views of EPA. Mention of company or product names is not
to be considered as an endorsement by the author or the EPA.
VI
-------
CONTENTS
Page
Figures iv
Tables v
Acknowledgment vi
1. Introduction 1
2. Sampling Design 2
2.1 Site selection 2
2.2 Vehicle Density 2
2.3 Site characteristics 6
2.4 Location of monitors 6
2.5 Sampling procedures 9
2.6 Laboratory procedure 9
3. Results 11
3.1 Average total suspended particulates 11
3.2 Average lead concentrations 14
3.3 Lead as a percentage of total suspended particulate 14
3.4 Relative lead concentrations 20
4. Conclusions 22
References 23
Appendix A Laboratory Procedures 24
Appendix B Analysis of Roadway Lead Data Using Analysis of
Variance Techniques 28
111
-------
FIGURES
Number Page
2-1 Map of Route 562 Relative to 1-75 and 1-71 3
2-2 Schematic of Traffic Contributing to Route 562 4
2-3 Photograph of Sampling Location Showing Monitors Situated
on Towers 7
2-4 Schematic of Sampling Locations 8
3-1 Wind Rose Indicating Frequency of Hourly Average Direction 12
3-2 Average 24-Hour Concentration of Total Suspended
Particulates at Various Elevations and Setback Distances 13
3-3 Average 24-Hour Concentration of Lead at Various Elevations
and Setback Distances 15
3-4 Percentage of Lead in Total Particulate Samples at Various
Elevations and Setback Distances 19
3-5 Average 24-Hour Concentration of Lead Obtained at the
Ground Level Monitor Compared to Concentrations Obtained
at Elevated Monitors 21
-------
TABLES
Number Page
2-1 Typical Entrance and Exit Traffic Volume on Route 562
Ramps 5
3-1 Total Suspended Particulate and Lead Concentrations at
Three Heights and 2.8 Meters Setback from the Road 16
3-2 Total Suspended Particulate and Lead Concentration at Three
Heights and 7.1 Meters Setback from the Road 17
3-3 Total Suspended Particulate and Lead Concentration at
Three Heights and 21.4 Meters Setback from the Road 18
-------
ACKNOWLEDGMENT
This report was prepared by PEDCo Environmental, Inc., for the U.S.
Environmental Protection Agency under Contract No. 68-02-3013. Mr. David
Lutz was the Project Officer from the Monitoring and Data Analysis Division.
The PEDCo Project Director was Mr. Charles Zimmer and the Project Manager was
Mr. David Armentrout. Mr. Anthony Wisbith was director of field monitoring
and Mr. Craig Caldwell was director of laboratory procedures. The principal
author of the report was Mr. Douglas Orf.
-------
SECTION 1
INTRODUCTION
The United States Environmental Protection Agency (U.S. EPA) promulgated
National Ambient Air Quality Standards (NAAQS) for lead on October 5, 1978.
Compliance with these standards is determined by measuring the concentrations
of lead in the ambient air. In support of the measurement programs, the U.S.
EPA is promulgating regulations for selection of appropriate lead monitoring
sites. The guidelines specify vertical distances and setback distances from
roadways for lead monitoring sites.
The EPA requested PEDCo Environmental to perform a limited field moni-
toring study to determine horizontal and vertical lead distribution in the
area of expected maximum lead concentrations along roadways. The intent was
to show relative distributions over specific distance ranges to provide
support for the monitor siting ranges specified in the regulations. These
ranges are necessary in order to provide monitoring agencies with flexiblity
to consider practical factors such as availability of utilities, protection
of instruments from vandalism, etc. in monitor siting. While inferences can
be drawn from existing studies of lead and total suspended particulate
2-7
distributions and relationships, the previous studies do not address
adequately both horizontal and vertical lead concentration distributions
within the ranges specified in the guideline. Since the study has a narrowly-
defined purpose, it was not designed to provide data for predictive models,
for explaining traffic volume and meteorological impact on lead concentra-
tions, for correlation with particle size data, or for similar applications
that would require more extensive sampling and experimental design.
-------
SECTION 2
SAMPLING DESIGN
2.1 SITE SELECTION
The following criteria were applied in locating an appropriate monitor-
ing site: (1) an average daily vehicle volume of at least 40,000 vehicles,
(2) an average vehicle speed of at least 35 to 45 miles per hour, (3) rea-
sonable distance from any topographic obstructions to air flow, and (4)
availability of utilities and security for equipment. The site selected for
the monitors was the parking area of an abandoned drive-in theatre on State
Route 562, also called the Norwood Lateral. This roadway is the major
connecting route between Interstate 7E and Interstate 71 in the Norwood area,
a few miles north of downtown Cincinnati, as shown in Figure 2-1.
2.2 VEHICLE DENSITY
Information obtained from the City of Cincinnati Traffic Engineer's
Office indicates an average of 58,500 vehicles per day in the area of the
monitoring site. Figure 2-2 indicates; contributions to the total traffic
volume at the various entrance and exit ramps; Table 2-1 shows a typical
hourly breakdown of traffic volume. The table shows definite peak periods of
traffic during the hours of 3 to 5 p.m. and 7 to 8 a.m.
Leaded gasoline is the primary contributor of lead emissions from motor
vehicle traffic. The Ohio, Kentucky, Indiana Regional Council of Governments
indicated that 62 percent of the vehicle miles traveled in Hamilton County
(which encompasses the monitoring area) represent vehicles of model year 1975
or later.*
*
Telephone communication with a representative of the Ohio, Kentucky, and
Indiana Regional Council of Governments on May 19, 1980.
-------
EJ LOCATION OF SITE
Figure 2-1. Map of Route 562 relative to 1-75 and 1-71
-------
CM
VO
in
3
O
O
-M
cn
-Q
+J
c
O
(J
O
•r—
<4-
(*-
re
4-
o
O
(O
O
oo
CM
I
CM
-------
00
a.
cc.
C\J
UD
tn
O
O
UJ
«t_ -^"^
^) (O
O (B
X
UJ S_
a>
a a.
cu
UJ r—
(U
CVJ
CO
r— CM CM *— *— CM
CM CM CM ^" *T f> CM CM r~ •—• F~ CM *t CM P— »» »
O
CM^W tn«n ^•^•fnrocMCMCMr-*- .— ro m m CM CM *
c
o ;
CM _ _ _ -_ __ _
r-. O. »— f— CM CM CM CM p- CM ^ CM »
O
h-
-• — CM CM ^- \D ^- r- r— P— P— ^ CM CM •— ^ CM
.O
C3 tfl -
fSiocoooCTt^^ro^cst^j-^-t^-f— ~ p^«- J5 ^- eb v
Li.
LO
I
CM«— COinCMCM^ _.-_,. .-, _-
SE gg^2J^p^|P«^«u>rr!°"^-o^p*^«*ft®<'1 ^*
*** P
_ . CM O O\ r- CM rs. CM r«- O OO r- Ift »O in CM «*> CM l/> CM 00 CM O CJ» O
O toi* io ^o O cvj
-------
This is significant because, beginning -in 1975, most U.S. manufactured light-
duty vehicles were designed to operate on lead-free fuel. However, not all
vehicles traveling in Hamilton County are U.S.-manufactured or light-duty.
Some foreign-manufactured vehicles still burn leaded gasoline; moreover, some
heavy-duty vehicles burn leaded gasoline and others burn diesel fuel, which
is lead-free. Also, some owners of newer cars may have altered their vehicles
so that they can burn the less expensive leaded fuel.
The average speed of the vehicles on Route 562 was assumed to be greater
than 35 to 45 miles per hour (a study siting criterion), since the posted
speed limit on this section is 50 miles per hour.
2.3 SITE CHARACTERISTICS
The abandoned theatre site is used as a holding area for newly manu-
factured General Motors automobiles (upwind of the monitoring site); for this
reason, the entire facility was secured by a fence with a locked gate.
Utilities were available on site for monitor operation.
As shown in Figure 2-3, the site provides unobscured exposure to the
Route 562 traffic flow, with no topographical interruptions.
Ten high-volume (Hi-Vol) ambient air samplers were operated at the site.
The samplers were placed at three elevations and three setback distances from
the roadway. One Hi-Vol was used as a control. The control was co-located
at the second setback distance and the middle elevation.
Setback distances were measured from the north edge of the four-lane
divided road. No attempt was made to measure Hi-Vol setback distances from
individual lanes. The prevailing meteorology, together with the effects of
traffic volume and speed, were assumed to keep the particulates airborne such
that contributions from all four lanes could best be measured from the north
edge of Route 562.
2.4 LOCATION OF MONITORS
Three towers, each with three tiers at 1.1, 6.3, and 10.5 meter heights
were constructed and oriented as shown in Figure 2-4. Each tier provided a
secure platform for at least one Hi-Vol sampler. Tower No. 1 was located 2.8
meters from the road. The third tier included instruments to measure wind
-------
OJ
o
c
o
o>
1/1
S-
o
o
C
o
ro
U
O
O
4->
O
JZ
D.
CO
I
CM
OJ
s_
3
cn
-------
LEGEND
HEIGHT TO AIR INTAKE:
MONITOR NOS. 1,4,8 - 1.1 meters
MONITOR NOS. 2,5,6,9 - 6.3 meters
MONITOR NOS. 3,7,10 - 10.5 meters
X PENCE
i:: ... ..... MEDIAN STRIP
LANE DIVIDER
TOWER HO. 3
MONITOR NO. 10 i ,[~| ^
MONITOR NO. 9
MONITOR NO. 8
Jtu
TOWER HO. 2
MONITOR NO. 7 ,Tj|,
MONITOR NOS. 5 & 6
MONITOR NO. 4
JDU
TOWER NO. 1
MONITOR NO. 3 FT)
(WITH WEATHER M' " H
INSTRUMENTS)
MONITOR NO. 2
MONITOR NO. 1
JaJ
7 1 m
SETBACK
-X—
17.8 m SET
BACK
-X-
21 4 m
SETBACK
23.2 m
I
ROUTE 562 (NORWOOD LATERAL)
Figure 2-4. Schematic of sampling locations.
8
-------
speed and direction. Tower No. 2, set 7.1 meters from the north edge of the
road, held four Hi-Vol monitors. Placement was the same as on Tower No. 1,
except that the second tier (6.3 meters above ground) held two samplers, one
of which was used as a control. The filters in the control sampler were
handled exactly like the filters of the other nine samplers, but no power was
supplied to the Hi-Vol. The third tower was 21.4 meters from the road.
2.5 SAMPLING PROCEDURE
The filter media used in this study were Schleicher and Schuell No. 1 HV
of spectro quality grade. During each of the 21 consecutive sampling days, at
approximately 10:00 a.m., ten filters from the previous 24-hour sampling
period were removed from the filter housing and replaced with unexposed
filters. Hi Vols were calibrated and operated as specified in the Quality
o
Assurance Handbook for Air Pollution Measurement Systems, Volume II. The
exposed filters were then placed in an envelope and taken to the laboratory
for analysis. Each filter was weighed twice and handled according to the
o
procedures described in the Quality Assurance Handbook. As a control in the
laboratory, a laboratory filter blank was included daily and was handled in
the same manner as the other filters. The laboratory blank provided infor-
mation on the background lead levels for this type of filter. Additionally,
the EPA supplied 20 audit filter strips with known lead content, which were
also analyzed. Sampling and analysis quality assurance data are included in
Appendix A.
As indicated earlier, the measurements of wind speed and direction were
obtained from instruments located atop the third tier of Tower No. 1. This
information was recorded continuously throughout the study on a strip chart.
The stripchart data were then reduced to hourly readings.
2.6 LABORATORY PROCEDURE
The laboratory procedure involved gravimetric analysis of all filters
for particulate matter with a Torbal EA-1 AP analytical balance. The filters
were equilibrated in a controlled environment of 20° to 25°C +3 percent and
relative humidity of less than 50 +5 percent for at least 24 hours before
weighing. When equilibration was reached, the filters were weighed imme-
diately after removal from the controlled environment. Each filter was tare
-------
and gross weighed twice. If the difference between the weighings exceeded
o
the requirements specified in the Quality Assurance Handbook the filters
were weighed again. The original and check weighings were performed by
different analysts.
After the filters were weighed the lead fraction of the particulate
sample was analyzed with a Perkin-Elmer Model 560 atomic absorption spectro-
photometer. Samples were prepared by a hot extraction procedure as described
o
in the Quality Assurance Handbook. The filters were digested in batches of
25, and all samples were analyzed for lead on the same day. Ten percent of
the samples were analyzed in duplicate, including the laboratory and field
blanks. Strips measuring 1.9 by 20.3 cm were cut from the exposed filter.
Lead normally is considered to be uniformly distributed across a filter. ' '
911
This has not proved true, however, in measurement of roadside emissions. '
Therefore, several cuttings were made at various locations on the filter.
10
-------
SECTION 3
RESULTS
Despite efforts to place the monitoring site at the point of optimum
impact relative to wind direction, Figure 3-1 indicates that the overall
impact was primarily from the southwest, west, and west-northwest rather than
the southeast, the direction toward which the monitors were oriented. The
monitors were oriented southeast to catch the full impact of the plume from
the nearest traffic lane. The intent was to maximize lead emission impact as
opposed to providing data to characterize traffic emissions.
There were no days during the study when the wind was blowing directly
toward the monitoring site with appreciable speed (daily average in excess of
1.4 meters per second).
3.1 AVERAGE TOTAL SUSPENDED PARTICULATES
Figure 3-2 depicts the average 24-hour concentration of total suspended
particulate matter (TSP) obtained from each of the field monitors. The
average TSP concentration decreased with increasing elevation of the monitors
at each setback distance. For all three setback distances the TSP concen-
tration is highest at the lowest elevation (1.1 meters). This is the posi-
tion nearest the vehicle emission point and closest to the road level where
reentrained dust can be picked up. The average particulate level at ground
level is highest at the monitor nearest the roadway, and it decreases with
distance from the roadway. At an elevation of 6.3 meters, average particu-
late concentrations from the sampler on the second tower (7.1 meters setback)
were higher than that at the 2.8 meter setback. The sampler at 6.3 meters
elevation and 21.4 meters setback distance recorded lower average particulate
concentrations than did the sampler setback 7.1 meters at the same elevation.
It may be that the site nearest the roadway was located too close to the
source for the elevated monitors to collect the maximum portion of the
11
-------
WWW
ssw
ESE
Figure 3-1. Wind rose indicating frequency
of hourly average wind direction.
12!
-------
140
130
120]
no
100
90
fc 60-
£
50-
40-
30-
20-
10
TOWER
NO. 1
I
J_
I
_L
TOWER
NO. 3
6 8 10 12 14 16 18
MONITOR SETBACK DISTANCE, meters
20 22
Figure 3-2. Average 24-hour concentration of total suspended
particulates at various elevations and setback distances.
13
-------
dispersing plume of traffic emissions. The 7.1 meter setback may have been
in a better position to catch a larger portion of the dispersing plume. The
average concentrations at an elevation of 10.5 meters indicate a slight
increase between 1.1 and 7.1 meters setback distance, but remained virtually
unchanged between 7.1 and 21.4 meters. Tables 3-1, 3-2, and 3-3 show the 24-
hour TSP and lead concentrations at three elevations and at 2.8 meters, 7.1
meters, and 21.4 meter setbacks, respectively.
3.2 AVERAGE LEAD CONCENTRATIONS
The average 24-hour concentrations of lead are plotted in Figure 3-3.
Supporting data are in Tables 3-1, 3-2, and 3-3. All of the measured concen-
trations of lead were adjusted to account for a mean background lead concen-
tration on the filters. This was done by analyzing laboratory blanks for
each day of the sampling program. Results from duplicate filter analysis for
lead indicated a mean coefficient of variation of only 0.044 (4.4%) for the
exposed filters.
As was the case with TSP, the highest lead concentrations for all set-
back distances were at 1.1 meters elevation. The data show that lead con-
centrations are higher at the lower elevations for each setback distance.
The average concentrations at both the 6.3 meter arid the 10.5 meter
elevations decrease slightly between the 7.1 meter and the 21.4 meter setbacks.
These concentration gradients are less than for the 1.1 meter elevation. The
average lead concentrations as a function of height do not converge as
rapidly with increased distance frim the road as they did for TSP.
The data also show that the concentrations at both the 6.3 and 10.5
meters elevation are lower at the 2.8 meters setback than for the 7.1
meters setback. The wind speed, wind direction, and turbulence created by
the vehicular traffic are not sufficient to transport as many of the lead
particles to the monitors at 6.3 and 10.5 meters elevation close to the
roadway (2.8 meters) as farther from the roadway (7.1 and 21.4 meters).
The concentrations at the 1.1 meter elevation show the normal decreasing
trend as the distance from the roadway is increased.
-------
1.40
1.30-
1.20-
1.10-
1.00
>
L
•0.90
5 0.80
UJ
§0.70
0.60-
0.50-
0.40-
0.30-
0.20-
0.10-
-1 1 1 T
TOWER
NO. 1
10 12
14
TOWER
NO. 3
16 IB 20 22
MONITOR SETBACK DISTANCE, meters
Figure 3-3. Average 24-hour concentration of lead at various
elevations and setback distances.
15
-------
TABLE 3-1. TOTAL SUSPENDED ^ARTICULATE AND LEAD CONCENTRATIONS
AT THREE HEIGHTS AND 2.3 METERS SETBACK FROM THE ROAD
Date
4/17/80
4/18/80
4/19/80
4/20/80
4/21/80
4/22/80
4/23/80
4/24/80
4/25/80
4/26/80
4/27/80
4/28/80
4/29/80
4/30/80
5/1/80
5/2/80
5/3/80
5/4/80
5/5/80
5/6/80
5/7/80
X
Concentration (yq/m )
TSP
1.1 m
79
125
124
79
124
159
121
119
100
80
89
86
127
122
145
199
157
182
209
202
162
133
6.3 m
63
105
95
63
103
101
96
74
86
72
61
59
93
92
124
166
136
158
139
160
120
103
10.5 m
53
113
97
53
57
97
94
68
91
72
54
53
61
89
121
160
131
156
114
143
104
95
Lead
1.1 m
0.64
2.45
2.30
0.69
1.98
1.08
0.70
1.03
0.41
0.18
1.31
0.97
1.80
0.75
1.53
2.49
1.97
2.28
1.33
0.87
1.09
1.33
6.3 m
0.43
1.98
1.71
0.53
1.35
0.51
0.59
0.53
0.28
0.18
0.59
0.60
1.27
0.55
1.24
1.99
1.64
2.13
0.71
0.66
0.62
0.96
10.5 m
0.41
2.11
1.56
0.42
0.77
0.32
0.51
0.38
0.28
0.19
0.11
0.40
0.83
0.49
1.14
1.75
1.64
1.94
•0.35
0.73
0.70
0.81
16
-------
TABLE 3-2. TOTAL SUSPENDED PARTICULATE AND LEAD CONCENTRATION
AT THREE HEIGHTS AND 7.1 METERS SETBACK FROM THE ROAD
Date
4/17/80
4/18/80
4/19/80
4/20/80
4/21/80
4/22/80
4/23/80
4/24/80
4/25/80
4/26/80
4/27/80
4/28/80
4/29/80
4/30/80
5/1/80
5/2/80
5/3/80
5/4/80
5/5/80
5/6/80
5/7/80
X
o
Concentration (yg/m )
TSP
1.1 m
77
125
115
82
118
160
112
112
98
83
83
81
115
113
138
197
150
177
189
183
142
126
6.3 m
72
125
112
74
120
118
111
86
100
83
71
71
101
104
134
197 .
162
183
164
181
135
119
10.5 m
-
116
101
65
108
99
96
70
95
76
57
55
85
91
123
163
143
160
124
149
113
104
Lead
1.1 m
0.56
2.33
2.18
0.71
1.49
1.01
0.72
0.68
0.37
0.22
1.06
0.76
1.36
0.82
1.14
2.26
1.85
2.16
1.09
0.97
0.68
1.16
6.3 m
0.44
2.13
1.72
0.57
1.67
0.51
0.70
0.53
0.40
0.43
0.94
0.63
0.95
0.59
1.12
2.27
1.92
2.35
0.78
0.98
0.89
1.07
10.5 m
-
2.03
1.78
0.38
1.65
0.32
0.52
0.37
0.30
0.31
0.60
0.34
0.65
0.42
1.42
1.69
1.91
2.22
0.41
0.61
0.71
0.93
17
-------
TABLE 3-3. TOTAL SUSPENDED PARTICIPATE AND LEAD CONCENTRATION
AT THREE HEIGHTS AND 21.4 METERS SETBACK FROM THE ROAD
Date
4/17/80
4/18/80
4/19/80
4/20/80
4/21/80
4/22/80
4/23/80
4/24/80
4/25/80
4/26/80
4/27/80
4/28/80
4/29/80
4/30/80
5/1/80
5/2/80
5/3/80
5/4/80
5/5/80
5/6/80
5/7/80
X
•3
Concentration (yg/mj
TSP
1.1 m
70
120
103
68
108
131
95
90
96
64
58
67
96
101
123
202
141
168
161
161
126
112
6.3 m
70
118
105
70
113
113
94
80
98
80
67
60
91
95
129
179
151
168
142
156
118
109
10.5 m
68
114
101
66
107
104
95
72
96
78
61
55
87
95
129
169
157
166
133
154
117
106
Lead
1.1 m
0 39
2.18
. 1.91
0.53
1.53
0.71
0.50
0.59
0.28
0.26
0.90
0.68
1.03
0.63
1.16
1.85
1.81
1.88
0.85
0.69
0.90
1.01
6.3 m
0.39
2.22
1.67
0.42
1.57
0.52
0.50
0.46
0.29
0.34
0.83
0.56
0.79
0.54
1.18
2.12
1.85
1.82
0.77
0.74
0.83
0.97
10.5 m
0.41
2.07
1.68
0.44
1.49
0.36
0.48
0.36
0.31
0.31
0.62
0.39
0.86
0.43
1.20
1.75
1.96
1.98
0.52
0.60
0.77
0.90
18
-------
1.10
1.00
0.90
0.80
I
'0.70
I
JO.60
i
i 0.50
0.40
0.30
0.20
0.10
1 T
TOWER
NO. 1
TOWER
NO. 2
_L
1.1 METER ELEVATION
10.5 METER ELEVATION
J_
_L
TOWER
NO. 3
_L
6 8 10 12 14 16 18 20 22
MONITOR SETBACK DISTANCE, meters
Figure 3-4. Average percentage of lead in total suspended
particulate samples at three elevations and three
setback distances.
19
-------
The data show that sampler height, is critical in terms of concentration
range uniformity closer to the roadway (2.8 meters setback), but it becomes
less critical between 7.1 meters and 21.4 meters setback.
3.3 LEAD AS A PERCENTAGE OF TOTAL SUSPENDED PARTICIPATES
The lead fraction as a percentage of the particulate concentration
was calculated for all samples. The results are illustrated in Figure 3-4.
A 1977 report by PEDCo^ indicated that the average fractions of lead
in particulate samples were 1 to 2 percent, with none less than 0.2
percent and none higher than 5.0 percent. The data in the current study
indicate an average of 0.80 to 1.00 percent. The highest percentage of
lead was measured at monitors located at 1.1 meter elevation. Figure 3-4
indicates that lead as a percent of TSP decreases with elevation. Distance
from the roadway appears to have minimal effect on lead concentration
expressed as percent of TSP concentration.
3.4 RELATIVE LEAD CONCENTRATIONS
Figure 3-5 shows the effects of locating lead monitors in positions
that are less than optimal for measuring maximum concentrations, where
breathing level (1.1 meters elevation) would be considered optimal. The
average concentrations at elevations of 6.3 and 10.5 meters are expressed
relative to the concentrations at 1.1 meters for each of the three setback
distances. At both elevations the maximum relative lead concentration is
obtained when the setback distance is 21.4 meters. Relative concentrations
at setback of 7.1 meters from the roadway and less than 6.3 meters
elevation represented 96 percent of the maximum. Relative concentrations
at 21.4 meters from the roadway and 10.5 meters elevation represented 89
percent of the maximum lead concentration.
20
-------
1.0
3 0.8
UJ
tt
UJ
: »—
> UJ
':'- 0.6
i (—
> «t
5z
00
0.4
LU U-
a: o
o
gO.2
§
TOWER
NO. 1
i I
j I
TOWER
NO. 3
i I
6 8 10 12 14 16 18 20 22
MONITOR SETBACK DISTANCE, meters
Figure 3-5. Average 24-hour concentration of lead at the
ground level monitor compared to concentrations at elevated monitors.
21
-------
SECTION 4
CONCLUSIONS
The data from this study show that both TSP and lead concentrations are
greatest at the 1.1 meter breathing level height for each of three setback
distances studied. The TSP and lead concentrations at three vertical heights
are different at each setback distance from the roadway. The concentration
differences between each height are greater between 2.8 and 7.1 meters set-
back.than between 7.1 and 21.4 meters.
The EPA performed a statistical analysis of the monitoring results
(Appendix B) to determine if the data support the siting criteria for micro-
scale and middle scale lead monitoring stated in the regulation. The cri-
teria allow microscale monitors to be placed between 2 and 15 meters from the
roadway and at a vertical height of 2 to 7 meters. Monitors at middle scale
sites should be between 15 and 100 meters from the roadway and at 2 to 15
meters high. The analysis concludes that the siting criteria for both
monitoring sites are reasonable both in terms of height and setback distance.
22
-------
REFERENCES
1. Federal Register, Vol. 43, No. 194, Thursday, October 5, 1978, pp.
46246-46247.
2. Danies, R. H., H. Motto, and D. M. Chilko. Atmospheric Lead: Its
Relationship to Traffic Volume and Proximity to Highway. Environ. Sci.
Technol. 4 (4):318-322, 1970.
3. PEDCo Environmental, Inc. Lead Analysis for Kansas City and Cincinnati.
Environmental Protection Agency, Contract 68-02-2515. June 1977.
4. Bryan, R. J., R. J. Gordon, and H. Menck. Comparison of High Volume Air
Filter Samples at Varying Distances from Los Angeles Freeways, Presented
at the 68th Annual Meeting of the Air Pollution Control Association,
Chicago. June 24-28, 1973.
5. Barltrap, D., and C. D. Strelow. Westway Nursery Testing Project.
Report to the Greater London Council. August 1976.
6. Creason, J.P., et al. "Roadside Gradients in Atmospheric Concentrations
of Cadmium, Lead, and Zinc," in Trace Substances in Environmental Health,
V. 5., A Symposium, D.D. Hemphill, ed. U. of Missouri, 1972.
7. Record, F., et al. Philadelphia Particulate Study, G.C.A. Report to EPA,
Report No. GCA-TR-78-02-6, 1978.
8. U.S. Environmental Protection Agency Report No. 600/4-77-027a, May 1977.
Quality Assurance Handbook for Air Pollution Measurement Systems Volume
II.
9. Scott, R. K., et al. Atomic Absorption and Optical Emission Analysis of
NASN Atmospheric Particulate Sampler for Lead. Environ. Sci. and Technol,
10, 877-880, 1976.
10. Zdrojewaki, A., et al. The Accurate Measurement of Lead in Airborn
Particulates. Inter. J. Environ. Anal. Chem. £, 63-77, 1972.
11. U.S. Environmental Protection Agency Report No. 600/4-77-034, June 1977.
Los Angeles Catalyst Study Symposium, pp. 223.
23
-------
APPENDIX A
LABORATORY PROCEDURES
The methods described in the following references were used to determine
the total suspended particulate (TSP) arid lead (Pb) concentrations:
1. Reference Method for the Determination of Suspended Particulates in
the Atmosphere. 40 CFR 50.11, Appendix B, July 1, 1975.
2. Reference Method for the Determination of Lead in Suspended Par-
ticulate Matter Collected from Ambient Air. 43 CFR 194, Appendix
G, October 5, 1978.
The Quality Assurance procedures used are described in:
Quality Assurance Handbook for Air Pollution Measurement Systems,
Volume II. U.S. EPA Publication No. EPA-600/ 4-77-027a, May 1977
(The lead analysis procedure is in draft form).
The following deviations were made from the published QA procedures.
All procedures are more rigorous than required by the manual.
1. Schleicher & Schuell Type 1-HV spectoquality filters were used.
2. All filters were tare and gross weighed twice. The original and
check weighings were performed by different analysts.
3. 10% of the samples were analyzed in duplicate,. 10% of each of the
field and lab blanks were also analyzed in duplicate. Each of the
9 sites had at least 2 duplicates run several days apart.
4. 20 audit strips of known Pb content, supplied by the U.S. EPA, were
also analyzed.
The filters were digested in batches of 25. The hot acid method de-
scribed in the Reference Method was used. Table A-l details the distribution
by filter type of each batch. All samples were analyzed for lead on the same
day.
24
-------
Summaries of the values obtained from the analyses of the field blanks,
lab blanks, and audit strips are attached. All lead values reported have
been corrected for the 23 yg Pb/filter of the lab blank.
The audit strip summary shows decreasing recovery of Pb with increasing
concentration. No sample, however, contained more than 500 yg Pb per strip
(1/12 filter), and most contained less than 200 yg Pb.
The summary of replicates shows a mean coefficient of variance of 0.044
(4.4%).
25
-------
LABORATORY QUALITY CONTROL FOR
SPECIAL MONITORING PROGRAM FOR LEAD
PN 3366-G
Each filter for this project was prepared according to method 87 except
that each filter was weighed at least twice before sampling. Each day samples
consisted of a set of 11 filters - 9 sampling filters, 1 laboratory blank and
1 field blank.
All filters were equilibrated, weighed, and then stored in their original
container. The filters were stored and loaded in a clean area prior to their
delivery to the sampling site. The field blank was placed in a shelter
similar to those used for the samplers.
Upon return of each set of recovered filters to the laboratory a clean
filter from the stock was added as a laboratory blank. Each set of filters
was delivered and logged in the laboratory using the standard procedure.
Each set of filters was equilibrated and weighed according to method 87
except that each filter was weighed at least twice. When all filters in a set
had been weighed and met the specified criteria, they were prepared for lead
analysis.
Lead analysis was done according to the method list in the Q.A. Manual
Volume II for Ambient Methods. All eleven filters in the set were analyzed
along with one audit strip for Pb. One sample filter was extracted and
analyzed in duplicate.
26
-------
C/)
»—I
C/1
oo
LjJ
a:
oo
oo
<£.
cr
OL
O
o
co
O
cc
oo
I
i— CTl
CO i— CM CM CM CO
co i — ro co CM r^-.
CO OJ CM . CM i — -
i — in
f^ CM CM i — i — CM "
r— OO
CM
r^ CM i — i — CM CM »
to to
(O to (/) 03 O
a> ^* j*z Q. cu -i-
Q. C ' C 4/1 -r~ CX-M
ojftjna -i^-*: s-oiro
i- i — i — to c ci/> -i-> s--t->
to -QX3-M(CrO-l-JtO i — E
Q-CLi — i — Q. Q.-t-Q_O
E E cu cuauj^jDcu-o ES-
(O ITS -i- T- S- (O to i. 3 (O4-
OOOOU-LL. I I CCOO
CO
CM
ID
in
CM
in
CM
in
CM
in
CM
in
CM
in
CM
in
CM
CM
IT)
CM
in
CM
in
CM
(O
o
I—
27
-------
APPENDIX B
ANALYSIS OF ROADWAY LEAD DATA USING ANALYSIS OF VARIANCE TECHNIQUES
William F. Hunt, Jr.,
Thomas Cur^an and
Eve Sneed
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, ^orth Carolina 27711
The lead data, collected by PEDCo Environmental, Inc. in the report, Field
Study to Determine Spatial Variability of Lead from Roadways , have been reana-
lyzed using the Analysis of Variance (ANOVA). The data were collected by PEDCo
Environmental to perform a limited field monitoring study to determine the
relative horizontal and vertical lead distribution in support of the monitoring
siting ranges in the Part 58 regulation. The application of the ANOVA to the
lead and TSP data is presented, along with the results of the analysis.
Statement of Objectives
(1) To determine whether there are significant differences in lead
concentrations measured at varying setback distances from roadways
and at different vertical heights.
(2) If there is a difference between setback distances or vertical
heights or combinations of both, we wish to determine the optimum
location for monitoring the expected maximum lead concentrations
along roadways with consideration being given to safety, vandalism,
and averaging time of the standard.
Descriptions of the Experiment
The experiment is a factorial design, where the setback distances (L^) and
vertical heights (H,) are fixed and the effects of week (Wk) and day (D^ are
random. The location for the experiment was the parking area of an abandoned
drive-in theatre on State Route 562, in the Norwood area of Cincinnati. Three
towers, each with three tiers at 1.1, 6.3, and 10.5 meter heights (H^), were
28
-------
constructed. One high-volume sampler for measuring TSP and lead was located at
each tier of each tower. The towers were located at setback distances of 2.8,
7.1, and 21.4 meters from the road. Both lead and TSP data were collected for
three weeks between April 17 and May 7, 1980. The procedures taken to ensure
data quality are described in the report and will not be discussed here.
Mathematical Model
An additive model was used to describe the experimental design as follows:
Xijkl = *+ Li + Hj + LHij + Wk + LWik +HWjk + Dl + WDlk + Eijkl
where X. .., = the lead or TSP measurement
u = the overall average
Li = the effect due to setback distance
H. = the effect due to vertical height
*J
LH. . = the effect due to the interaction of setback distance and
J vertical height
Wk = the effect due to differences between weeks
LW.jk = the effect due to the possible interaction between setback
distances and weeks
HW.. = the effect due to the possible interaction between vertical
J heights and weeks
D, = the effect due to days
WD,. = the effect due to the possible interaction between weeks and days
E. .. , = the undesigned variability or random error
The term E.-^-i is made up of the following 2 and 3 way interactions which
were assumed not to exist: LHWi ,k> LD^, HD^, LHD.^, LWD^, HWD,kl and
It must be kept in mind that the authors of this Appendix did not
design the original experiment, but instead applied the ANOVA after the experiment
had been run and the data collected.
29
-------
Results
In performing the ANOVA, no tranformation Of the data was taken. The data
were assumed to be approximately normally distributed. The ANOVA for lead is
shown in Table 1 and the ANOVA for TSP is shown in Table 2. In both analyses,
all the sources of variation were statistically different from 0.0 with the
exception of the interactions of setback distance by week and vertical height by
week. The mathematical model explains 95.5% of the lead variability and 95.3% of
the TSP variability.
Of particular importance is the result that-there is a significant interaction
between setback distance and vertical height. This is illustrated in Figures 1
and 2 for lead and TSP, respectively, which summarize the interactions by calculating
the means associated with each combination of setback distance and vertical
height, along with their 95% confidence Intervals. Where the confidence intervals
overlap, the means are not significantly different from one another. Because
multiple comparisons are being made, the Just Significant Confidence Interval
(JSCI) has been calculated using the Tukey "q" statistic.
Generally, at each setback distance, the mean of lead or TSP decreases as
the vertical height increases. At the setback distance of 2.8 meters, the mean
associated with a vertical height of 1.1 meters is the highest recorded and is
significantly different from the recorded means at the vertical heights of 6.3
and 10.5 meters. (The confidence intervals do not overlap.) As the vertical
height increases to 10.5 meters, the lowest mean is recorded. At each setback
distance, the decrease in both lead and TSP levels as the vertical height increases,
is different with the greatest drop shown at the setback distance closest to the
roadway (2.8 meters). This difference, in the relative change at each of the
setback distances, is why the interaction exists.
An examination of Figures 1 and 2 shows that multiple comparisons can be
made. Of particular interest is whether or not this analysis supports the EPA
recommended siting criteria for the microscale and middle scale roadway sites
to measure the area of maximum lead concentration. The EPA recommendation for
the microscale sites is that the lead monitor be placed between 5 and 15 meters
30
-------
from the roadway with a vertical height of 2 to 7 meters. The recommendation for
the middle scale sites is that the lead monitor must be placed between 15 and 100
meters, depending on the average daily traffic, with a vertical height of 2 to 15
meters. From Figure 1, the maximum concentration is observed at the monitor
closest to the roadway (setback distance of 2.8 meters and vertical height of 1.1
meters). In some cases it may not be permissible to establish such a site or it
may not be practical to locate the monitor so close to the roadway, because of
potential vandalism, and problems in servicing a monitor so close to the flow of
traffic.
Eliminating the monitor closest to the roadway, the next highest recorded
mean of ambient lead levels occurs at a setback distance of 7.1 meters and a
vertical height of 1.1 meters. This mean is not significantly different from the
mean of the ambient level recorded at the same setback distance, but at the
higher vertical height of 6.3 meters.
The mean recorded at this combination of vertical height (6.3 meters) and
setback distance (7.1 meters) is of interest, because it is the only monitor
located within the EPA criteria for the microscale roadway type site. It is
important to note that the confidence interval about the mean of this monitor
overlaps the confidence intervals of the means of all other monitors with the
exception of the means of the monitors located at a setback distance of 2.8
meters and vertical heights of 1.1 meters (the highest mean) and 10.5 meters (the
lowest mean).
All three monitors at 21.4 meters from the roadway were located within the
EPA criteria for the middle scale roadway type site. The means for these monitors
are not significantly different from each other since the confidence intervals of
the means overlap. Also, it should be noted that these confidence intervals
overlap the confidence interval of the mean of the monitor located at 7.1 meters
from the road and 6.3 meters high.
Since the monitor closest to the roadway with the highest mean (vertical
height of 1.1 meters) is not practical because of potential vandalism and problems
in servicing a monitor so close to the flow of the traffic, the EPA recommended
31
-------
siting criteria for microscale and middle scale sites for distance from roads
and height above ground are reasonable.
Reference
1. W.J. Dixon and F.J. Massey, Jr., Introduction to Statistical
Analysis. 440-442, McGraw-Hill Book Co.,, Inc., New York (1957).
32
-------
TABLE 1. ANOVA TABLE FOR LEAD
i
Source of
Variation
Distance (L)
Height (H)
L x H
Week (W)
L x W
H x W
Day (D)
W x D
Error
Degrees of
Freedom
2
2
4
2
4
4
6
12
151
Sum of
Squares
0.247
2.729
1.080
19.005
0.021
0.101
15.925
35.980
3.506
Mean
Square
0.124
1.365
0.270
9.503
0.005
0.025
2.654
2.998
0.023
F Statistic
5.39*
59.35***
11.74***
413.17***
<1
1.09
115.39***
130.34***
* Probability less than 0.01
** Probability less than 0.001
*** Probability less than 0.0001
33
-------
TABLE 2. ANOVA TABLE FOR TSP
Source of
Variation
Distance (L)
Height (H)
L x H
Week (W)
L x W
H x W
Day (D)
W x D
Error
Degrees of
Freedom
2
2
4
2
4
4
6
12
151
Sum of
Squares
1660.290
15850.626
7141.353
168137.959
81.719
479.111
27471.217
29238.256
12239.233
Mean
Square
830.145
7925.313
1785.338
84068.979
20.430
119.778
4578.536
2436.521
81.055
F Statistic
10.24**
97.78***
22.03***
1037.19***
<1
1.48
56.49***
30.06***
* Probability less than 0.01
** Probability less than 0.001
*** Probability less than 0.0001
34
-------
00
CM
X
o
Sj
1 1
I 1 I I
o:
UJ
1
-.
00 O
I
O
*P
LU
~. CO
^ OJ
QV31
o
re
a) -c
-H en
C i-
i—i CD
ac
at
O i—
C (O
c_>
•o
-P C
C
O (U
$- U
LO «/)
cn -r-
o
-o
c -^
(O O
(O
0)
GO
u o
•r~
+J to
tt) C
e o
J^ •!-
-(J •»->
•r- (O
s- c
et ••-
35
-------
o
£
^^£
^^ ** * ^u
~. c-l »O
TT
CM
OJ
cc
UJ
Bi
(P —
i
Q
"
CM UJ
~~
_
CO
p
UJ
to
«0
>
&_ 0}
CD o
4J C
C (C
to
(^ "^~
O 0
•o o
t- »0
H j^
C +J
O QJ
4J «4- +
cox:
O) O1
O CO f-
i_ C O>
o» o oc
CL. -i-
-P r-
ir>
c E i-
na O Ol
o s»
I/)
C J- -O
<0 O C
Q) M- ra
in
O -M
•r- C
+•> O)
S-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/4-83-002
4. TITLE AND SUBTITLE
Field Study to Determine Spatial Variability of Lead
from Roadways
5. REPORT DATE
1981
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
3. RECIPIENT'S ACCESSION NO.
May
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
10. PROGRAM ELEMENT NO.
A24A2F
11. CONTRACT/GRANT NO.
68023013
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Monitoring and Data Analysis Division
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Field Study - 1980
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A short-term field monitoring study was conducted to determine the horizontal
and vertical lead distribution along roadways. Results are presented for three
heights and three horizontal setback distances from roadways.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Lead Monitoring
Horizontal and Vertical Lead Distribution
Ambient Air Quality Measurements
b. IDENTIFIERS/OPEN ENDED TERMS
~| 19 SF"3\.R~T- CLASS (ThisJieponT
Unclassified
c. COSATI Field/Group
10
21 NO. OF PAGES
46
i Release Unlimited
EDA Form 2270-.; !Sov. 4->7\
- = .-', ^ T ' CL 133 ,''"
Unclassified
22. PRICE
E V I C' L- S ED' ."ION ! 5 O B b O L £_ T t
------- |