FINAL REPORT
ON
. ป
STUDY OF AIR POLLUTION ASPECTS
OF
VARIOUS ROADWAY CONFIGURATIONS
SEPTEMBER 1, 1971
Submitted To
New York City Department of Air Resources
By The
General Electric Company
3198 Chestnut Street
Philadelphia, Pennsylvania 19101
-------
TABLE OF CONTENTS
Page No.
1.0 Introduction 1
1.1 Study Objectives 1
1.2 Study Participants 2
1.3 Roadway Configurations Evaluated 2
1.4 Study Report 3
2.0 Summary of Results 4
2.1 Data Characteristics 4
2.1.1 Brief Site Descriptions 4
2.1.2 Average Weekly Diurnal Traffic Curves for
Each Site 9
2.1.3 Horizontal and Vertical CO Concentration
Distribution and Concentration Isopleths 16
2.1.4 Meteorological Patterns at the Study Sites 23
2.1.5 Hydrocarbon Concentrations at the Study Sites 32
2.1.6 Particulate Concentrations and Trace Metals at
the Study Sites 32
2.2 Traffic Air Quality Relationships 37
2.3 Site Ranking by Air Pollution Severity 51
2.3.1 Comparison with Standards 51
2.3.2 Vehicular Pollution Factor 52
2.3.3 Ranking by Concentration in Breathing Zone of
Nearest Receptor 52
-------
TABLE OF CONTENTS. Continued
Page No.
2.5 Recommendations for Future Research 64
2.5.1 Recommendations for Further Measurements . 64
2.5.2 Recommendations for Further Model Developments 66
2.5.3 Recommendations for Health Effects Research 67
3.0 Study Methodology 68
3.1 Measurement Techniques 68
3.1.1 Overall Methodology 68
3.1.2 Air Quality Measurements 71
3.1.3 Traffic Measurements 73
3.1.4 Meteorological Measurements 74
3.2 Regression and Correlation Analyses 77
3.2.1 Carbon Monoxide 77
3.2.2 Hydrocarbon 93
3.3 Model Development 99
3.3.1 Superposition Method 100
ซ ,
3.3.2 Model Validation 105
4.0 Site Description & Measurements 120
4.1 Summary Tables 120
4.2 Franklin D. Roosevelt Drive at. Sutton Place 122
4.3 Brooklyn Battery Tunnel 130
4.4 Brooklyn Queens Expressway at Hicks Street 135
4.5 Cross Bronx Expressway at Grant Circle 145
4.6 Cross Bronx Expressway at Jessup Avenue 152
-------
TABLE OF CONTENTS. Continued
Page No,
4.7 Bruckner Expressway near White Plains Road 160
4.8 Grand Central Parkway at Parsons Boulevard 167
4.9 Brooklyn Queens Expressway at Park & Navy Streets 175
4.10 Canal Street between Church & Mercer Streets 181
4.11 Trans-Manhattan Expressway at George Washington Bridge
Plaza 187
5.0 Highway Planning Factors 192
Appendices
A Carbon Monoxide Measurements & Statistics
B Hydrocarbon Measurements & Statistics
C Meteorological Measurements & Statistics
D Traffic Measurements & Statistics
-------
LIST OF ILLUSTRATIONS
Figure No. Title Page No.
2.1-1 Open Cut Roadway Sites " 5
2.1-2 Tunnel Roadway Sites 6
2.1-3 Cantilever Covered Roads 7
2.1-'4 Open Roadway Configurations 8
2.1-5 Intermittently Covered Roadway 9
2.1-6 Average Weekday Diurnal Traffic Flow Rates at
Three Open Cut Sites 11
2.1-7 Diurnal Traffic Flow Rate Characteristics in Two
Tunnels 12
2.1-8 Diurnal Characteristics of Traffic Flow Rate at Two
Cantilever Covered Roadways 13
2.1-9 Diurnal Characteristics of Traffic Flow Rate at Two
Open Roadway Configurations 14
2.1-10 Average Weekday Diurnal Traffic Flow Rate on An Inter-
mittently Covered Roadway 15
2.1-11 Horizontal & Vertical Concentration Profiles in Three
Open Cuts 18
2.1-12 Horizontal & Vertical Concentration Profiles in Two
Tunnels 19
2.1-13 Horizontal & Vertical Concentration Profiles on Two
Cantilever Covered Roads 20
2.1-14 Horizontal & Vertical Concentration Profiles at Two
Open Roadway Sites 21
2.1-:15 Horizontal & Vertical Concentration Profiles at an
Intermittently Covered Roadway 22 :
2.1-16 Concentration Isopleths at Two Open Cuts 24
2.1-17 Concentration Isopleth in a City Street 25
2.1-18 Concentration Isopleths in Two Tunnels .. 26
-------
LIST OF ILLUSTRATIONS, Continued
Figure No. Title Page No.
2.1-19 Concentration Isopleths at Two Cantilever Covered
Roads 27
2.1-20 Concentration Isopleths at Two Open Roadways 28
2.1-21 Concentration Isopleth at an Intermittently Covered
Road ' 29
2.1-22 Hydrocarbons vs. Traffic Flow Rate 33
2.1-23 Participate Concentration vs. Distance 34
2.1-24 Particulate Concentration vs. Traffic Volume 35
2.2-1 Profile of CO vs. Traffic Flow Rate at FDR Drive 39
2.2-2 Profile of CO vs. Traffic Flow Rate at Brooklyn
Battery Tunnel 40
2.2-3 Profile of CO vs. Traffic Flow Rate at Hicks St. 41
2.2-4 Profile of CO vs. Traffic Flow Rate at HughJ. Grant
Circle 42
2.2-5 Profile of CO vs. Traffic Flow Rate at Nelson &
Jessup 43
2.2-6 Profile of CO vs. Traffic Flow Rate at Bruckner
Expressway 44
2.2-7 Profile of CO vs. Traffic Flow Rate at Grand Central
Parkway 45
2.2-8 Profile of CO vs. Traffic Flow Rate at Park & Navy
Streets 46
2.2-9 Profile of CO vs. Traffic Flow Rate at Canal Street 47
2.2-10 Profile .of CO vs. Traffic Flow Rate at G.W.B. Plaza 48
2.2-11 CO vs. Traffic Flow Rate 49
2.2-12 Average Weekday Diurnal CO at Three Points Across BQE
-------
LIST OF ILLUSTRATIONS, Continued
Figure No. Title Page No.
2.3-1 Relation Between Vehicular Pollution Factors Taken
at Exhaust Plane & Traffic Speed 54
2.4-1 Diurnal Indoor & Outdoor Carbon Monoxide Concentra-
tions in Air Rights Building at Sutton Place 59
2.4-2 Diurnal Indoor & Outdoor Carbon Monoxide Concentra-
tions at Grand Central Parkway 60
2.4-3 Diurnal Indoor & Outdoor Carbon Monoxide Concentra-
tions at Hicks Street 61
i -
2.4-4 Diurnal Indoor & Outdoor Carbon Monoxide Concentra-
tions on Canal Street 62
2.4-5 Diurnal Indoor & Outdoor Carbon Monoxide Concentra-
tions near George Washington Bridge Plaza 63
3.1-1 General Air Pollution Trailer Layout 69
3.1-2 Flow Diagram for CO and Hydrocarbon System 70
3.1-3 Sign Conventions for Horizontal Wind Components 76
3.2-1 Hydrocarbon Concentration for BQE at Hicks Street 97
3.3-1 Superposition Method for Predicting Traffic Generated
Pollution Concentrations 101
3.3-2 Method of Images for FOR Drive 102
3.3-3 Relation Between Vehicular Pollution Factors Taken
at Exhaust Plane & Traffic Speed 104
4.2-1 Franklin D. Rooselvet Drive at Sutton Place - Pictor-
ial View 123
4.2-2 Franklin D. Roosevelt Drive at Sutton Place - Eleva-
tion View 124
4.2-3 Franklin D. Roosevelt Drive at Sutton Place - Plan
View 125
-------
LIST OF ILLUSTRATIONS, Continued
Figure No. Titie Page No.
4.2-4 Carbon Monoxide vs. Distance from West Wall at FDR 128
Drive
4.3-1 Brooklyn Battery Tunnel - Pictorial View 131
4.3-2 Brooklyn Battery Tunnel - Elevation View 132
4.3-3 Brooklyn Battery Tunnel - Plan View 133
4.4-1 Brooklyn Queens Expressway at Hicks Street - Pictor-
ial View 136
4.4-2 Brooklyn Queens Expressway at Hicks Street - Eleva-
tion View 137
4.4-3 Brooklyn Queens Expressway at Hicks Street - Plan View 138
4.4-4 Illustration of Eddy over the Shallow Cut at Hicks St. 140
4,4-5 Comparison of Wind Azimuth Between Central Park Sta-
tion and Hicks StreetSite 141
4.4-6 Carbon Monoxide Levels on Two Days Under Different
Wind Regimes 142
4.5-1 Cross Bronx Expressway at Hugh J. Grant Circle - Pic-
torial View 146
4.5-2 Cross Bronx Expressway at Hugh J. Grant Circle - Eleva-
tion View 147
4.5-3 Cross Bronx Expressway at Hugh J. Grant Circle - Plan
View 148
4ป5-4 Comparison of Traffic and Turbulence Curves at Hugh J.
Grant Circle 150
4.6-1 Cross Bronx Expressway at Nelson & Jessup Avenues -
Pictorial "View 153
4.6-2 Cross Bronx Expressway at Nelson & Jessup Avenues -
Elevation View 154
-------
LIST OF ILLUSTRATIONS. Continued
Figure No. Title " Page No.
4.6-3 Cross Bronx Expressway at Nelson & Jessup Avenues -
Plan View 155
4.6-4 Relationship of General Wind Flow to Azimuth Recorded
in Deep Cut at Jessup Ave. Site 157
4.6-5 Sigma Azimuth vs. CO Concentration - 5 Weekday
Averages - CBE at Jessup 158
4.7-1 Bruckner Expressway Near White Plains Road - Pictor-
ial View 161
4.7-2 Bruckner Expressway Near White Plains Road - Eleva-
tion View 162
4.7-3 Bruckner Expressway Near White Plains Road - Plan View 163
4.7-4 Turbulence - Traffic - Pollutant Relationship -
Bruckner Expressway 165
4.8-1 Grand Central Parkway at Parsons Boulevard - Pictor-
ial View 168
4.8-2 Grand Central Parkway at Parsons Boulevard - Eleva-
tion View 169
4.8-3 Grand Central Parkway at Parsons Boulevard - Plan View 170
4.8-4 Sigma Azimuth and Total Traffic at Grand Central Park-
way 173
1
4.9-1 Brooklyn Queens Expressway at Park & Navy Streets -
Pictorial View 176
4.9-2 Brooklyn Queens Expressway at Park & Navy Streets -
Elevation View 177
4.9-3 Brooklyn Queens Expressway at Park & Navy Streets -
Plan View 178
-------
LIST OF ILLUSTRATIONS, Continued
Figure No. . Title Page No.
4.10-1 Canal Street Between Church and Mercer Streets -'
Pictorial View 182
4.10-2 Canal Street Between Church and Mercer Streets -
Elevation View 183
4.10-3 Canal Street Between Church and Mercer Streets -
Plan View 184
.4.11-1 Trans Manhattan Expressway at G.W.B. Plaza -
Pictorial View 188
4.11-2 Trans Manhattan Expressway at G.W.B. Plaza -
Elevation View 189
4.11-3 Trans Manhattan Expressway at G.W. B. Plaza -
Plan View " 190
5-1 Proposed Highway to be Evaluated for Air Quality 193
-------
LIST OF TABLES
Table Title Page No.
2.1-1 Truck Hourly Averages 10
2.1-2 Meteorological Summary 30
2.1-3 Trace Metals 36
2.3-1 Summary of Salient Data at Each Site 52
2.3-2 Maximum Hourly CO Concentration at Nearest Receptor 55
2.3-3 Ranking of Sites by Breathing Zone of the Nearest
Receptor 57
3.2-1 Numerical Results of 54th & Button Place CO Regression
Analysis 80
3.2-2 Numerical Results of BQE at Hicks Street CO Regression
Analysis 81
3.2-3 Numerical Results of Hugh J. Grant Circle CO Regression
Analysis 82
3.2-4 Numerical Results of Nelson & Jessup CO Regression
Analysis 83
3.2-5 Numerical Results of Bruckner Expressway CO Regression
Analysis 84
3.2-6 Numerical Results of Bruckner Expressway CO Regression
Analysis 85
3.2-7 Numerical Results of Grand Central Parkway CO
Regression Analysis 86
3.2-8 Numerical Results of Grand Central Parkway C0_
Regression Analysis 87
3.2-9 Numerical Results of BQE at Park & Navy CO Regression
Analysis . 88
3.2-10 Numerical Results of BQE at Park & Navy CO Regression
-------
LIST OF TABLES, Continued
Table . Titie Page N(3.
3.2-11 Numerical Results of Canal Street CO Regression
Analysis 90
3.2-12 Numerical Results of Canal Street CO Regression
Analysis 91
3.2-13 Numerical Results of BOB at Hicks Street CO Re-
gression Analysis 94
3.2-14 Numerical Results of Hydrocarbon Regression Analysis 96
4.1-1 Summary of Monitoring Activity at Each Site 121
-------
-------
1. Introduction
1.1 Study Objectives
The increasing cost of real estate, demand for better vehicular trans-
portation and number of registered vehicles in urban areas suggest that
the urban dweller will live close to major arterial highways and will
be subjected to high doses of pollutants such as carbon monoxide, par-
ticulate matter, oxidants, and oxides of nitrogen. The advent of the
Clean Air Act of 1970 projects relief in the year 1975. Achievement
of these clean air goals, however, is dependent upon knowledge and use
of that knowledge, of the relationship of vehicular generated air
pollution and the configuration of urban roadways. Therefore, it is
necessary to:
1. Assess the potential effect of alternate roadway configurations
on ambient air pollutant concentrations in and about such roadways,
surrounding working and living areas and to relate indoor and out-
door concentrations.
2. Develop generalized guidelines for use by urban and transportation
planners to guide the future design of urban expressways to results
that are consistent with national and local air quality standards
and objectives.
This experimental study was undertaken therefore to:
1. Ascertain the current air quality in the immediate vicinity of
various urban roadway configurations.
2. Determine how the selected urban roadway configurations aid or
hinder the diffusion of the pollutants emitted by urban traffic.
3. Develop mathematical relationships between traffic, traffic speed,
pollutant concentration, meteorological parameters and roadway
configuration.
The study has shown that it is possible to assess the impact of the
1970 Clean Air Act by future monitoring programs which determine the
extent to which the clean air goals are achieved. It is possible to
identify those areas in and around a given configuration that do not
meet the National Air Quality Standards under certain meteorological
and traffic conditions. With this information, it might be possible to
modify traffic flow rates and speed such that, for a given roadway con-
figuration and meteorological situation, the National Air Quality
Standards will be met. Similarly, it is possible for urban and trans-
portation planners to compute in advance the pollution anticipated for
a proposed roadway design in order to insure that dwellings adjacent
to the proposed design will not be exposed to pollutant concentrations
in excess of the National Air Quality Standards.
-------
The quantity and the quality of the data taken as well as the models
developed for the wide range of urban configurations evaluated, in one
of the most complex urban communities, namely New York City,
lead toward a better understanding of the diffusion of pollutants and
meeting the study objectives.
1.2 Study Participants
This study was conducted under the leadership and guidance of the New
York City Department of Air Resources. The field work and data analysis
was performed by the General Electric Company. Five other city, state
and federal agencies participated in the study.
The Department of Transportation, New York City, provided services for
site selection, site use permits, equipment set up and detector in-
stallations and site security. '
The Department of Transportation, New York State, using its own person-
nel and equipment, conducted traffic data collection on a 24 hour basis
throughout the field phase of the study.
The Department of Health, New York State, provided technical assistance
by use of its own staff to the New York City Department of Air Resources.
The Air Pollution Control Office, E.P.A. (formerly National Air Pollution
Control Administration, H.E.W.) provided funding of the air pollution
monitoring, data analysis and report writing phases of the study through
a contract with the New York City Department of Air Resources.
The United States Bureau of Public Roads, New York Division Office,
provided on a matching fund basis, support of the traffic data collection
activities of the New York State Department of Transportation.
1.3 Roadway Configurations Evaluated
Ten urban roadway configurations were evaluated. These configurations
and their location are tabulated below in the study sequence:
Roadway Configuration Location
Covered on Top-Open at Side Franklin D. Roosevelt Drive at Button
Place
Long Tunnel (ventilated) Brooklyn Battery Tunnel
Shallow Cut Brooklyn Queens Expressway at Hicks St.
Short Tunnel (unventilated) Cross Bronx Expressway at Hugh J.
Grant Circle
Deep Cut Cross Bronx Expressway at Nelson &
Jessup Avenues
Grade Road : Bruckner Expressway near White Plains
Road
Cantilever Cover Grand Central Parkway at Parsons Blvd.
-------
Roadway Configuration Location
Viaduct Brooklyn Queens Expressway at Park
and Navy Streets
City Street Canal Street between Church and
Mercer Streets
Intermittent Span Trans-Manhattan Expressway at George
Washington Bridge Plaza
These sites were chosen because they represent typical roadway configura-
tions one would expect in any urban area. They range from simple grade
configurations to complex city streets. The city street was chosen to
show levels existing on one of the more heavily travelled streets in New
York City, and was the major lower Trans-Manhattan artery between the
Holland Tunnel and Brooklyn.
The ventilation characteristics for the sites ranged from good (open
configurations, such as the Viaduct site which was surrounded by a
large park and apartment/school courtyard) to poor (closed configurations,
such as tunneIs).
The peak traffic flow rates for the configurations studied ranged from
12,000 vehicles/hour to 1,500 vehicles/hour, thereby encompassing a wide
spectrum of the traffic flow rates that would be encountered in any urban
area.
1.4 Study Report
This report describes the study, its results and how they apply to meet-
ing the above objectives. For purposes of improving readability, this
report is divided into seven volumes of which this is the first. Volumes
2 through 7 consist entirely of data. Volumes 2, 3 and 4 contain carbon
monoxide data. Hydrocarbon data is in Volume 5. Volume 6 and 7 re-
spectively, are Meteorological and Traffic data. It is hoped that this
report will be of service in helping to meet the ultimate objective of
cleaner air and still provide for better vehicular transportation.
-------
-------
Fig. 2. 1-1A
Deep cut section on Cross Bronx
Expressway at Nelson & Jessup Ave.
6 lane road carrying 118,000 vehicles/
day
Fig. 2. UB
Shallow cut section on Brooklyn
Queens Expressway at Hicks St.
carries 107,000 vehicles/day.
Fig. 2. 1-1C
City Street - Canal Street in lower
Manhattan carries 28,000 vehicles/day
Fig. 2. 1-1 - OPEN CUT ROADWAY SITES
-------
B. Tunnels - There w&re two of these as follows:
ฉ The long,, blower ventilated Brooklyn-Battery Tunnel
A short tunnel without blower ventilation on the Cross .
Bronx Expressway at Hugh J. Grant Circle
Sketches of the geometry of these tunnels appear in figures
2. 1-ZAandB.
Fig. 2. 1-2A
The long, blower ventilated Brooklyn-
Battery Tunnel. 2 lanes eastbound
carry 26, 000 vehicles/day. 2 lanes
west-carry 25,000 vehicles/day.
Fig. 2. 1-2B
A short, unventilated tunnel on the
Cross Bronx Expressway at Hugh
Grant Circle. 3 lanes eastbound
carry 44,000 vehicles/day. 3 lanes
westbound carry 42,000 vehicles/day.
Fig. 2. 1-2 - TUNNEL, ROADWAY SITES
C. Cantilever Covered Roads - There were two of these as follows:
e A fully covered^ open at one side configuration on the Franklin
D. Roosevelt Bs?ive at Sutfcon Place carrying six lanes of traffic
and 9?6000 cars per weekday.
A doubly cantilevered;, open in the top middle configuration on
the Grand Central Parkway at Parsons Boulevard. The six lane
road carries 110,000 cars /weekday. A pair of parallel
-------
Sketches of the geometry of these two sites appear in figures
2. 1-3A and B.
Fig. 2. 1-3A
Six lane road carrying 97, 000 cars/day.
Fully covered open at one side, passing
beneath an air rights building on
Franklin D. Roosevelt Drive at Sutton
Place.
Fig. 2. 1ป3B
Six lane road carrying 110, 000 cars/day.
Doubly cantilevered covering, open in .
the top-middle configuration on Grand
Central Parkway at Parsons Boulevard.
Parallel service roads carry 23,000
vehicles/day.
Fig. 2. 1-3 - CANTILEVER COVERED ROADS
D. Open Configurations - There were two of these as follows:
A grade road on Bruckner Express way near White Plains Road
and carrying six lanes of traffic. Traffic volume on highway
and parallel service road averaged 55S000 vehicles/weekday.
An elevated roadway on the Brooklyn Queens Expressway at
Navy Street carrying six lanes of traffic.
Sketches of the geometry of these sites appear in figures
2. l-4AandB.
-------
Fig. 2. 1-4A
6 lane open roadway, with parallel
service roads' in each direction, on
Bruckner Expressway near White Plains
Road. Traffic volume averaged
55,000 vehicles/day.
Fig. 2. 1-4B
6 lane viaduct roadway carrying
108,000 vehicles/day on Brooklyn
Queens Expressway at Navy Street.
Parallel service roads at grade carry an
additional 11,500 vehicles/day.
Fig. 2. 1-4 - OPEN ROADWAY CONFIGURATIONS
E. Intermittently Covered Roadway - These was one such site as
follows:
A twelve lane highway passing beneath a group of four air
rights apartment buildings on the Trans-Manhattan Express-
way. These buildings, known as the George Washington
Bridge Apartments, straddle the highway as shown in figure
2. 1-5. The Expressway carries 175,000 vehicles/weekday. Two
parallel city streets carry an additional 15,000 vehicles/
weekday.
-------
Fig. 2.1-5
Twelve lane highway passing beneath
a succession of four air rights buildings
on the Trans-Manhattan Expressway.
Road carries 175,000 vehicles/weekday.
Two parallel city streets carry an addi-
tional 15, 000 vehicles/weekday.
Fig. 2. 1-5 - INTERMITTENTLY COVERED ROADWAY
2.1.2 Average Weekday Diurnal Traffic Curves for Each Site
Figures 2. 1-6 through 2. 1-10 show the average weekday diurnal
curves for traffic.flow at each of the sites. For convenience of comparison
the curves for sites having similar geometry appear on the same page. Thus
the traffic curves for the three open cuts appear as figures 2. 1-6A, 2. 1-6B
and 2. 1-6C. Since traffic flow was bi-directional at all sites, a flow rate
curve for each direction is given. Table 2. 1. 1 gives the hourly' averages of
truck traffic included in the diurnal traffic curves.
It will be noted from figures 2. 1-6 through 2. 1-10, that in general the
traffic at all sites showed the two characteristic rush hour peaks, the slight
dip between rush hours and the pronounced relative minimum between two and
four A.M.
-------
TABLE 2.1-1
TRUCK HOURLY AVERAGES
Open Cuts
Time
M-l
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-N
N-l
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10
10-11
11-M
Canal
St.
30
30
30
30
70
140
230
330
350
350
310
290
270
280
310
250
250
170
120
70
50
30
30
30
CBE at
BQE Nelson
at and
Hicks Jessup
St. Avenue
330
No 370
Data 420
Taken 500
550
610
690
770
870
850
880
930
830
750
900
670
590
580
630
470
320
270
220
230
Tunnels
CBE at
Hugh
Brooklyn Grant
Battery Circle
137
No 160
Data 179
Taken 189
225
305
319
260
312
317
299
324
319
285
317
301
255
242
; 220
200
150
145
147
137
Open Sites
Bruckner
Expy.
60
70
70
110
120
150
190
250
300
350
360
350
340
360
330
320
250
180
160
120
100
70
70
60
10.
BOE at
Park &
Navy St.
276
361
440
344
517
663
574
940
1039
939
850
999
951
921
1017
830
697
609
433
463
384
316
289
210
Cantilever
Covered
FDR
Drive
at
Sutton
Place
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 .
0
0
GCP at
Parsons
Blvd.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Inter-
mittently
Covered
Trans-Man-
hattan Expy
At. G.W.Bo
Plaza
330
330
350
390
460
560
670
750
870
930
860
920
890
870
970
940
710
590
520
420
340
230
210
-------
00
4>
4)
4)
13
OS
i
n
S
u
5
0)
>
I
V
o
t I
b
u
0)
pH
u
I
4)
13
rt
I
H
H
1000
10 N 2 46
Time of Day
8 10 M
M 2 4 b
FIG. 2. 1-6A Shallow Cut on Brooklyn-Queens Expressway at Hicks Street
1000
800
600
400
200
M
8 10 N 246
Time of Day
FIG. 2. 1-6B _ City Street - Canal Street in Lower Manhattan
8 10 M
5000
4000 rg;
3000
2000
1000
M
8
8 10 M
10 N 246
Time of Day
FIG. 2. 1-6C Deep Cut on Cross Bronx Expressway at Nelson & Jessup Aves.
FIG. 2. 1-6 AVERAGE WEEKDAY DIURNAL TRAFFIC FLOW
RATES AT THREE OPEN CUT ROADWAYS
-------
0)
0)
I
V
13
tf
I
u
X
ซW
At
1*
H
o>
4)
u
n)
Rt
M
H
8
8
10 N 2 4 6
Time of Day
FIG. 2. 1-7A Average Weekday Diurnal Traffic Flow Rate
Characteristics in a Short Unventilated Tunnel -
Cross Bronx Expressway at Hugh Grant Circle
10
M
2500
2000
1500
?
ฃ 1000
500
M246810N 246
Time of Day
FIG. 2. 1-7B Average Weekday Diurnal Traffic Flow
Rate Characteristics in the Long, Blower
Ventilated, Brooklyn Battery Tunnel
8
10 M
FIG. 2. 1-7 DIURNAL TRAFFIC FLOW RATE
CHARACTERISTICS IN TWO TUNNELS
-------
n
v
o
a
4)
I
4)
o
0)
h
H
n
o>
XI
4)
I
4)
*>
n)
ซ
tป
u
N
4000
3500
3000
2500
2000
1500
1000
500
* 4 f t ! ซ *ss*"zs* zs mZmEปS3Sซซszzr*"
iซ ป i * i!S**""*S*l***!'
!* ;i"5f" ! งง!ป*
i *
* ! fง*ซ***ซ
SS8SS5SS !ป! ! SS !! I SS ml ง !!! ! -3 S!a!tZ!l"l"S*SMw'v
{ii ง ง ii *! ป "" *ป" iป*3*ซซ ซ l^ftM * 1* ?*WtT**ฃC2!S2
i ปi -^7*' !..ป* f i ปง 31 IM^Mปซ iiซป lb*fซii' :i .:*!" i tป JS222H
I *'* 4 , h . . I ' il ป .ซป. *m , >. iปit . ri i* - k 4* *_
mmmmmmmm * m * ""55' ^"ffc '* 1 ' -ซ S^SSS!! SS SSIIIIBSSSSSrTESS
: * ซ I * ipJfซBซiik^Bi ^ '*!
{I ซ ii ! ! ,JiซBซiซi^ii ซ "5SS3S55*g**S****S
!ป! ป ง * j*ซ'*iป"""lBปii,M"3 22SS"*25 SS ปgi!8*ง"?ง
* > !! ซ a^fcซ* BMWIMM***
ป
mm mi
mmmm
!
:!!!! E ::
a
a
**
!ง
::s::s::::=25:::::
10
M2 4 6 810N2 468
FIG. 2. 1-8A Average Weekday EKurnal Traffic Flow Rate
Characteristics on a Covered Roadway Open at One
Side - Franklin D. Roosevelt Drive at Sutton Place
M
7000
6000
5000
4000 :::: = ::g
3000
2000 ฑ
1000
M
m
FIG. 2. 1-8B Average Weekday Diurnal Traffic Flow Rate Characteristics
on a Doubly Cantilever Covered Roadway, Open in the Top
Middle - Grand Central Parkway at Parsons Boulevard
FIG. 2. 1-8 DIURNAL CHARACTERISTICS OF TRAFFIC FLOW
RATE AT TWO CANTILEVER COVERED ROADWAYS
-------
i
ซj
w
W
w
H
3
M
10 N 2
t
TIME OF DAY
8
10
O
a
h
O
3500
3000
2500
ฃJ 2000
1500
1000
500
FIG. Z. 1-9A Average Weekday Diurnal Traffic Flow Rate
Characteristics on Grade Road on Bruckner
Expressway Near White Plains Road
bwJBwaM ปป I *' * ) * JkซซB ปซป*ซ** VWH
i , 'JMBBk '' ^riซซBซซ^^l
I < r ^*ปป *i tm''' i
a*<ซ r ^m ''n
1 !* im' > ซ!* .ป 4ซซป-,
.ปซ *' *< <'ง'*'*'
! i*flp '..r ซ> ! **%*.,
' ซป ^M ปซ ?'&'**ซ!ซ(
!ซซ< *ป;ซ*?ซ
!ป i ป' ^fc/ซMซซซBi>ซซซ
ซ! i ปปป"ปซ;ซ*.^ซ
i' ซ ! ''^ป'.ซ*;.
!* .ซซ ^'^(IIซIBซ* ?*.;-ซ,*.
i mmmmt
:::::::K:BS
tm\ t ซป
IBI *(
ป tซ 31 ' !ป
.ซป*
M
8
10 N 2
TIME OF DAY
10 M
FIG. 2.1-9B Average Weekday Diurnal Traffic Flow Rate
Characteristics on Viaduct Roadway on
Brooklyn-Queens Expressway at Navy Street
FIG. 2. 1-9 DIURNAL CHARACTERISTICS OF TRAFFIC FLOW
RATE AT TWO OPEN ROADWAY CONFIGURATIONS
-------
n
o>
,-4
(J
a
0)
A
1
o
M
8
8
10 M
10 N 24
Time of Day
FIG. 2. 1-10 Average Weekday Diurnal Traffic Flow Rate Characteristics
on an Intermittently Covered Roadway - Trans Manhattan
> Expressway at George Washington Bridge Plaza
-------
2.1.3 Horizontal and Vertical Carbon Monoxide Concentration
Distribution and Concentration Isopleths
2.1.3.1 Horizontal and Vertical Concentration Profiles
Figures 2.1-11 through 2.1-15 are plots of the horizontal and
vertical distributions of carbon monoxide concentration for each
of the ten sites. As with the traffic curves, the plots for
sites with similar geometry appear on the same page for con-
venience of comparison. Thus the concentration curves for the
three open cuts are given in figures 2.1-11A, 2.1-11B and 2.1-11C.
In all cases the concentration values plotted are the weekday
means of the hourly averages over the entire monitoring period.
Generally, in figures 2.1-11 through 2.1-15, hourly concentrations
for four hours of the day were chosen to illustrate the daily
spread of values. The horizontal profiles represent sensor values
across the road at a height chosen to best illustrate the shape of
the profile. The vertical profiles generally represent sensor
values in the vertical plane which showed the highest concentration.
The direction of traffic flow at each site is shown on the horizontal
profile plots. The plane of vertical profiles are indicated on the
vertical plots. The average weekday diurnal concentration curves for
each sensor, not shown in this section, exhibit characteristics
similar to their traffic counterparts with two characteristic rush
hour peaks, a slight dip between rush hours and a pronounced relative
minimum between two and four A. M.
Before examining figures 2.1-11 through 2.1-15 in detail, it is
well to summarize what the data might reasonably be expected to
show. This summary will then serve as a basis for comparison with
the measured data.
The horizontal concentration profiles should tend to show peak
values at impermeable walls because the pollutant will tend to ac-
cumulate there. Thus in an open cut one would expect relative
maxima at the two walls and accordingly a relative minimum at or near
the center of the road. The two maxima will not, in general, be of
equal magnitude, but rather their relative magnitudes can be ex-
pected to depend on the traffic volume on each side of the median
strip and on the cross roadway component of wind.
The horizontal profile at a site with a wall on side of the road
and the other side open to ventilation, can be expected to show a
peak at the wall and a decay in concentration with distance from the
wall approaching the background level, in the absence of other
sources, when the distance from the wall becomes very large.
The horizontal concentration profile at a site which is open to
ventilation at both sides can be expected to show a relative maximum
at or near the middle of the road and a decay in both directions
across the road. It is obviously much easier for the pollutant to
diffuse away from the source than across the median strip into the
lanes of traffic travelling in the opposite direction.
-------
The vertical concentration profile above a road which is open to
ventilation at the top can be expected to-exhibit a decay in con-
centration with height above the mean exhaust plane of the vehicles
and to approach the background level when the height above the
source becomes large. It would not be unreasonable to expect an
exponential decay of concentration with height. Indeed Georgii et
al (Ref. 1) have reported such a relation in measurements made in
Frankfort, Germany.
In covered roadways, the presence of the impermeable ceiling
would require that the vertical component, of the concentration
gradient be zero at the ceiling. This would imply a relative maxi-
mum at the ceiling if no other means of ventilation exists. In the
case of a site such as Sutton Place which is covered, but has one
side open to ventilation, it is possible that a relative minimum of
concentration could exist at the ceiling.
With this summary of what might be expected at the various site
configurations, an examination of the concentration profiles at
the three open cut sites in figure 2.1-11 shows that at Hicks
Street and Nelson and Jessup Avenues, the horizontal profiles indeed
indicate relative maxima at the walls and a relative minimum at or
near the center of the road. The vertical concentration profiles at
these two sites show a decay with height above the road and the
Nelson and Jessup Avenues site actually indicates a vertical profile
shape that is consistent with an exponential decay.
The Canal Street site infigure 2.1-11C does not appear to behave
quite like the other two open cuts. The horizontal-profile appears
to be roughly sinusoidal in shape. The vertical concentration pro-
file shows a decay with height for the first 27 feet, then an in-
crease and then a decay. This seems to indicate that an inversion
layer exists along the building face.
Figure 2.1-12A shows the horizontal and vertical concentration
profiles in the short, unventilated tunnel at Hugh Grant Circle. As
can be seen from the figure, the horizontal profile shows relative
maxima at the walls and a relative minimum at mid-road for most hours
of the day. This behavior is as expected. The vertical profile in-
dicates an increase of concentration with height for the first ten
feet and then a leveling off. This indicates a fairly thick accumu-
lation at the ceiling more or less as expected.
The Brooklyn Battery Tunnel which is blower ventilated does not
exhibit a consistent concentration profile as can be seen from
figure 2.1-12B. This is almost certainly the result of the forced
ventilation blower cycle.
The two cantilever covered roadway sites exhibit the concentration
profiles that one would expect. The covered on top, open at one side
configuration at Sutton Place shows the expected maximum at the wall
and a roughly Gaussian shaped horizontal profile decaying towards the
-------
80
10 15 20 25
0 ..;. .10 .20 . _30 40. 50 ...... 60 70
Distance, y, across road ; Height, z, above mid- road
Figure 2.1-11A - Brooklyn Queens Expressway at Hicks Street
9 A.M.
4 P.M.
1 P.M;
3 20 40 60 . 80. 100 120 _ . ' ; 0 10 20 30 40 50
Distance, y, across road Height, z, above mid-road
Figure 2.1-1IB - Cross Bronx Expressway at Nelson & Jessup Avenues
20
o
10
0 20 40 60 80
Height, z, above mid-road
0 20 40 60 80
Distance, y, across road
Figure 2.1-11C - Canal Street between Church and Mercer Streets
FIGURE 2.1-11 HORIZONTAL & VERTICAL CONCENTRATION PROFILES IN THREE OPEN CUTS
-------
ฃ
i
10 SI-PS
50
Distance, y, across road
Horizontal Concentration Profile
Across Ceiling 16' Above Road
5 10 15
Height, z, above road
Vertical Concentration Profile
Along Left Lane tSouth) Wall
Figure 2.1-12A - Cross Bronx Expressway at Hugh Grant Circle - Westbound Tube
120 :ฑ
100
i
3& 120 ::ฃ
100
I
o
0 5 10 15
Distance, y, across road
Horizontal Concentration Profile
Across Ceiling 13' Above Road
Height, z, above road
Vertical Concentration Profile
Along Right Lane (South) Wall
Figure 2.1-12B - Brooklyn Battery Tunnel - Brooklyn Bound Tube
FIGURE 2.1-12 HORIZONTAL AND VERTICAL CONCENTRATION PROFILES IN TWO TUNNELS
-------
9 20
Height, z, above road
Vertical Concentration Pro-
file Along Wall
Figure 2.1-13A - Cantilever covered road, open at one side - FDR Drive at Sutton
Place
10 20 ...30 40 50 60
Distance, y, across road
Horizontal Concentration Profile Across Ceiling
o
o
c
o
V
8
o
O
30
25
20
15
10
5
0
100 .
0 2'0 4'0 60 8*0
Distance, y9 across road
Horizontal Concentration Profile
Across Roadway at 10 foot level
Height, EJ above road
Vertical Concentration Profile
At Middle of Road
Figure 2,1-13B - Doubly cantilever covered roadway, open in the top middle.
Grand Central Parkway at Parsons Boulevard
FIGURE 2.1-13
HORIZONTAL 6s VERTICAL CONCENTRATION PROFIUSS ON TOO
CANTILEVER COVERED ROADS
-------
10
50 40 20 0 20 40 5)
Distance, y, from mid-road
Horizontal Concentration Profile 10 ft.
above road
Figure 2.1-14A - Viaduct road - Brooklyn Queens Expressway at Park & Navy Streets
0 10
Height, from road
Vertical Profile at Edge of Road
20
::::::::::Sl|3:Sฑ:"::::::::::::I-.ฃ|3
:ฃj:ฑ:;|::.ฑ:i:::;:: + :::::::::g::2
ilil;!!!ili!ii;;i;;!!!l;;l!!;i!li;!i!
::gง:!ฑg:::::::::::::::::::::::::g!
..__. iZ... . . IL- i | .! it , - . ..... ..... . . ^ . ....
l:::i::.::ฃ3;l;::;:;:|:il;ป::;:;:i
i:::::::::::::::::!:::::::::!:^;?::.:"!
BlilliliBHiMJ
;^5!:]ln&;;;;;;; ;;;:;;;;^;;;:;;i
or\ ฃ.f\ J.f\ nr\ s\
HilBliiil I2B
+ ( ; i< i ! :::; ::::
:::::!i;::::::::::;;i!:l:::::::::::::::::::: ::::::::
ii!iiii!|!!;iiiiiiii=i:i|i!!ij!|ii:i= -10 i|i!|i
::!i ;::::::::: :::::::::; ;;>: 1 :: :;i !::::::::: ....^..i.
BiilllliiB g
1 1 1 1 1 1 MJ 1 i 1 1 1 LM" 1 1 1 1 1 1 1 1 II 1 1 1 1 I 1 1 1 1 1 1 ( I ! 1 1 1 1 1 1 1 rV I 1 1 f I
iffliii 11 " 11
1 ' f- Itttttmf
:::::::::::::: :::::::: ::::::::::::::; O -.;
^-i ::.'. ;:::
::::.::::::::::::::::::: :::::::::::::::::: u ::::::!:
ill II 1 fttll ttltll 1 ttttl 11 1 1 1'll I'l 1 m mtffli s gjjM
a)
IliiliiiH i *
i:::::::::::::: :::::::::::::::::::::::::::::: o ::::::::-
-HtTTl 1 1 1 1 1 1 1 Wh[MfMyi^?ffl Fffffffli
::::::::::::::':::::;; ::::::::::::. I: ::: r\
nf\ f.f\ f f\ n^ s* f
l:ttf|ฑiฑtjฑii|
fflffir^
;;;i;|:;^
] m
jlMjiiuA
:^;:|;:ฑi]
MlWff
^:|;::::-g
--V ->- -|f-
:::3Sig:r
l#
*ij*
, _; j_ i . __i
ttti ฃ4 tiiฃ
L-ttiiffiSh^!
IHPif
sk
-^L^j-
Hif
tea.
itt
"P-r^
-H--
tHฑ
Asi
^
f5-<
J* ^
^+4
^ซ
IZ."~4
k -:f:-.
--.^^ . . .
j- -
( -
r t
,^.f . .
I * '
H-
"BETM;
;rfcfMr
^^ftM.
~ -fi" M
-t -
- ii.: " .
."".: i ;
- i -
- - - f -
^_^}:.::
Distance, y, from mid-road Height, zt above road
Horizontal Concentration Profile 25 ft. Above Road Vertical Concentration
Profile in Middle or Road
Figure 2.1-14B - Grade Road - Bruckner Expressway near White Plains Road
#
FIGURE 2.1-14 HORIZONTAL & VERTICAL CONCENTRATION PROFILES AT TWO OPEN ROADWAY SITES
-------
to
50 .100 150 200
Distance, y, across road
Horizontal Concentration Profile at 3 foot level
10 20 30 40 50
Height, z, above road *-
Vertical Concentration Profile above
middle of westbound lanes
FIGURE 2.1=15 HORIZONTAL & VERTICAL CONCENTRATION PROFILES AT AN INTERMITTENTLY
-------
open side. The vertical concentration profile decays with height
and has its smallest value at the ceiling. The doubly cantilevered
road, open in the top middle, at Parsons Boulevard shows a hori-
zontal concentration profile with maxima at the walls and a minimum
in the middle. The vertical profile in mid-road shows an exponen-
tial-like decay with height below the opening in the roof. The
concentration values in the two upper corners were the highest
concentrations recorded at the site.
Both the open roadway configurations in figure 2.1-14 show
horizontal concentration profiles which peak at mid-road as expected.
The rise in concentration at both sides of the road at the Bruckner
Expressway site is almost certainly caused by the heavy traffic on
the service roads on each side. The vertical concentration profile
at the Bruckner Expressway shows the expected exponential-like
decay. The vertical gradient of the viaduct road at Park and Navy
Streets exhibits a linear'characteristic at the edge of the roadway.
It is entirely possible that the concentrations measured in this
vertical plane are predominantly background influence.
The vertical concentration profile at the intermittently covered
road in figure 2.1-15 exhibits an exponential-like decay as ex-
pected for an open top configuration. The horizontal concentration
profile does not appear to have a consistent pattern.
2.1.3.2 Concentration Isopleths
In addition to the actual concentration spatial profiles presented
in figures 2.1-11 through 2.1-15, a set of concentration isopleths
has been plotted for each of the sites. The isopleths were computed
from an empirical model which assumes that each lane of traffic con-
stitutes a line source of carbon monoxide. Then the concentration
at any point near the roadway consists of the superposition of con-
tributions from each of the line sources. It is further assumed in
the empirical model that concentration decays exponentially with
radial distance from the line source in each lane, and that the
horizontal plane in which the vehicle exhausts lie, coincides with
a plane three feet above the road containing the lowest level probes
at each site. Based on these assumptions and an empirically de-
termined decay factor, the concentration of each of the postulated
line sources was determined. From these concentrations and the
radial exponential decay law mentioned above, the isopleth plots
were computed for several hours at each site. This model and the
isopleths are presented and discussed in detail in Section 3 of this
report. Figures 2.1-16 through 2.1-21 are presented here, however,
as a brief summary showing the isopleth plot for a single hour at
each site. The direction of traffic flow at each site is as shown
on the plots. The values actually measured by the probes are
indicated on the figures for reference. It can be seen that reason-
ably good agreement exists between measured and computed values.
2.1.4 Meteorological Patterns at the Study Sites
Table 2.1.2 gives a summary of the general meteorological picture
at each of the sites. More complete descriptions appear in Section 4.
-------
29 3-3ป-33-3krr3d^3a 30 30 30 31 31 32 33 3435 36 37^33 39 39 3S
3J _32-ซ3 3^3.1 31 31 31 31 31 32-J233J34 35 36 3 _ .
33 33 33 32 32^32-32-02"32 33 34 35 36 37/39 4.3
35 35 34. 33 33 32 32 32 32 33 33 3.4 35 3.5 33 39/41/4^/46' 43 .
-x \ ! \\ll\fS
37 3^ 35 3/4 33 33 32 32 32 33 3^3 3^4 35 36 3,8 4;3 4.2 4,4 4.6/49/52
3
Northbound Lanes , Southbound Lanes
Grid Size = 4 feet
Shallow Cut - Brooklyn Queens Expressway at Hicks Street - 1600-1700 hours
1CU1CT11 11 11 11 11 11 11 10U10_10 10 9
12^2 12 12-12 12 12 12 <111 IP^IO 10 9 9 9
13 13 14-14^14^14^13 13 13 12 12 12>41 11 10.10 9
14 15 15 15 15 15 15 14 lV>l3 13 13 12 12O1 11 10
17 17 17 17 16 16^15 15 14 14^13 13 12
11 10OO
19 19 18 18^17 16 16^J5 15 A4 14>.13 12 12
>^
21220^19 18 17 17 16 16.5 15 14>13 13
11 11 11 10
12 12. 11 11
23 22 21 20X19 18 18^17 17
15 14X13
13 12 12 12>11
13 13 13 13 12
13 13 13 13
26-26 24 22 21 20 19 18 18
7 J25\ 23\ 21 201 19
17 16 15 14.
18 18 18
Westbound Lanes
: Grid Size = 6 feet
Deep Cut - Cross Bronx Expressway at Nelson & Jessup Avenues
Eastbound Lanes
- 0700-0800 hours
Figure 2.1-16 Concentration Isopleths at Two Open Cut's
(Measured Values enclosed in 0)
-------
/6 7 7 7 7
888
8999 10xป10 10 11
9 9 10^10 11 11 11 11
11 11 10 fOxl 099
ri 1111 o^o
13 13 12 12O2 11 11
'1414,13 13
15 15 14 14,13 13 it
15 14\j A 1 3
17 17 17 16 15
19 1SV8
20^20 19
9 10MO 11 11 J2T2 12
10 10 11 11 12^12 13 13 13
/ /
10 11 11 12 13 13 14^14 14
11 11 IB 13 13 14'15 15 15
^i 12 13 14-14 15 16,16^16
^fl* 14jftL 15 16^17
/ /U2J /
12 13 14 T5 16 17 18"W 18
12 14 15 16^7 18 19 19 19
^L 14 15 16 18 19 19 20 20"
h> 14 15/17 18 E.Q 20^20 20
19
20
20
20
19 20
20 21 21
21 21 22-
21 22' 22
17 16 15
21 2Kl<
> O ~^V> O O L
CO C. O ฃi ฃ- C-U A 7* * ' ^^S
23/5^23121 \ 19 117 if
19. 17
Westbound ' Eastbound
Grid Size = 5 feet
Canal Street between Church & Mercer Streets - 1200-1300 hours
Figure 2.1-1? Concentration Isopleth in a City Street
(Measured Values Enclosed in 0)
-------
15)
11 11 11 11 11 11 11 11 11 11 Vr 11 11 11 i
\
2 12\12 12
ฉ
13 13 13 T3
11 11 11 11 11 11 11 11J1T1 11 1U11 11 11 12 12 12 12 13 13/13 13
JJ 11 11 11 11 11 11^11 11 11 11 11 lf\ll 11 11 12 12 12 13 13 13
11 1111 11 1111 11 11 11 11 11 11 1\11 11 12 12 12 13 13 1
121212 11 11 11 11 11 10 10 10 10 11 11 11 11 12 1212 13 13 13 13
11/1
1 11 10 10
12 12 12 1
12 11 11 11 10 10,
/10
(L 10 11 1111 11
1212 13 13 14 i
10 10 10 10 10 10 11 II 11 12 12 12
13 14 !
4
.15,
South
Wall
Grid 'Size = 2 feet
North
Wall
A Short Unventilated Tunnel - Cross Bronx Expressway at Hugh J. Grant Circle
- Eastbound Tunnel - 0700-0800 hours
Ji, 6 i 7 7*7 7 7 o 6
ฃ> '
s-' r. 7 v '7 i 7 7 7 6 8
(
6 t 7 7 7 i 7 7 7 8 b
ป 6 6 7 7 | '7 7 7 3 b
\
6 f-. 6 7 7 !,7 7 7 b o
!'' 6 6 7 7 V7 7 7 d 8
!o 5 677 7. 7738
\\ \
it. <:, 6 1 7 7\ 7 7 7 b
\ \
6 \f, 6 7 7 7 \7 7 7 ti
i * \
('.^66777778
|r. 6 \<5 6 7 7 7\7 7 7
South
Wall
'i 8 i<<; 9 J v 9 9 y 1 0
> 6 b^""9 9 j 9 9 9 910
o c b 9 9 f 9 9 9 910
y b <: 9 9 i 9 9 9 9 10
\
b o b y 9 | 9 9 9 1 u 1 u
8\ 3 b 'J 9 19 9 9 10 10
8 \3 G 9 9 9 9 9 10 10
\ \
tt b b y y y y y i u i u
8 A t; M 9 y\ 9 910 10
W o V: fi 9 9 V' 9 10 10
\ \
b' d b\ b b 9 9 9 910
1
0
10
1
l'
/
h
/
0
(j
I"
1 1
i i
i
V
1
0
u
0
0
0
1
/f
f 1
1
(
i
1
1
i
1
1
' 1
0
p
0
0
0
0
0
u
0
0
0
10
/7>
1(j,
10;
10'
10
10
1
lu
1
10
U)
1 i.'
ฉ
10
1(A
10
10
10
10
10
10
10
10
9
'
9
i
Vo
10
10
10
1
10
1
10
10
10
< -, 1
(D
9
9
9
ฎ
10
10
10
9
North
Grid Size = 1 foot
Wall
Long, Blower Ventilated, Brooklyn Battery Tunnel - Brooklyn Bound Tube
Values x 10
- 1600-1700 hours
Figure 2.1-18 Concentration Isopleths in Two Tunnels
-------
Y39V".
4~I'j"i3c 36 |3 i 39 k>J ^Tli-H 22 -.'; 1 v 17"
41 4Q 33 36 134 32 29 27 ?5 S3 2"' 19 17
3 41\ 3
4ฃ A3 43 44 44 4>^i3 43 42 42 4
'13 43 44 44 44 44 43 43 43 42
43 44 44 44 44 44 44 43 43 43 42
i '.
15
39 37 35 32 30 ft7 25 23.P!\ 19 17
^/i /i/i ,'J4 /i 3 43 /i2 Vซ 35 35 33 3'i y? 2^ 23 2.1119 17
47 47 46 46 45 45 45 4>ป45 44 4 3^-4 i\ 3 9 36 33 30 2o 25 23 21\rj 1-3
\ ฉ \ \ I I - \
49 48 47 46 45 45 45 45 46 46 45 43 49 37 33 S'-'i 23 25 23 21 2' IS
\ \ I / I
52 5,1\49 47 4o 45 45 45 45 46 -43 4S 45 41 37
i9 47 45 44 44 44 45 46 4d 5
ซ 4r> 41 3
1^461 411 3
63)
33 3? 27 25 23 21 2? 13
33 -J:D 27/24 22 21/19 to
17
17
Southbound Lanes Northbound Lanes
Grid Size = 3 feet
A Cantilever Covered Roadway Open at One Side - Franklin D. Roosevelt Drive at
Button Place - 0700-0800 hours
22 .22*22 ^3-25 '^(<^i
18'l 8 19' 19xป20 21-ป2
1919 20 2x 2 1,2 2
^
20 21 22 2323
^ c . .
19 19 20/20 21 22
27 27 27 27 2<28 297 3132
9 20 20 20 20 2121 22 23 24
/ / '
21 200 20 20 2121 2,2 ,23
23
M^dh
28 27 27 27 28 2,9 30 3,1 /3,3f 3,5/^
31) ' 69
Eastbound Lanes Westbound Lanes
Grid Size = 5 feet
Double Cantilever Covered Roadway, Open in the Top Middle-Grand Central Parkway
at Parsons Boulevard - 0700-0800 hours
Figure 2.1-19 Concentration Isopleths at Two Cantilever
-------
Southbound Lanes
Grid Size = 5 feet
Grade Roadway - Bruckner Blvd. near White Plains Road - 1600-1700 hours
Northbound Lanes
ho
oo
0 0
0 0
;s)o o
o
o o
0 0
O 0 0
00 1
001
o tfs
O 0
O 0
1
I
1
1
i i
1 I
100000
i 0, 0 0 00
0 0 0 0(4
0
1 0
1
0 0
6| 5 /3
Eastbound Westbound
Grid Size = 20 feet >
Viaduct Roadway - Brooklyn Queens Expressway at Park & Navy Street - 1600-1700 hours
Figure 2.1-20 Concentration Isopleths at Two Open Roadways
-------
Ill 115
I'A 16 17 18 19-20-20-19
18 17 16 IS 14 14 13 13 12 12
19 21 22
8 18 17 17 16 15
6 26 25"*24 23 21
22 22 22 2
27 26 27 272
. "^"
9 33534 32 3! 31 343
Westbound
Eastbound
Grid Size = 10 feet
Figure 2.1-21 Concentration Isopleth at an Intermittently
Covered Road - Trans Manhattan Expressway at
George Washington Bridge Plaza - 1600-1700 hours
(Measured Values enclosed in 0)
-------
TABLE 2. 1-2
METEOROLOGICAL SUMMARY
Zero Degrees Azimuth Is Parallel To The Road; Zero Degrees Elevation Is Horizontal
Vane 1
Prevailing Azimuth'**'
Prevailing Elevation
Mean Sigma Azimuth
Mean Sigma Elevation
Mean Wind Speed (mph)
Vane 2
(8)
Prevailing Azimuth
Prevailing Elevation
Mean Sigma Azimuth
Mean Sigma Elevation
Mean Wind Speed (mph)
FDR Dr. at
Sutton Place
Over S. B.
Lane
z = 14'
85ฐ
+15ฐ
11.5ฐ
6.3ฐ
6.4
On Trailer
5ฐ
+ 5ฐ
14.3ฐ
9.1ฐ
3.6
BQE at
Hicks St.
Median
z = 14'
165ฐ
- 5ฐ
15.9ฐ
12.5ฐ
5.4
Over
Queens -
Bound
Lanes
z = 18'
175ฐ
M<5>
17.4ฐ
M
M
CBE at [CBE at
Grant uessup
Circle Uve. (ฐ)
Over [Median
Tunnel
Exit
z = 14' La = 14'
305ฐ 5ฐ
+ . 5ฐ I 5ฐ
35.9ฐ 19.8ฐ
19.7ฐ 13.9ฐ
4.4 5.5
Median 1
z = 14'
325ฐ M
M M
30.8ฐ M
M M
5.3(3 M
Bruckner
Expressway
South of
Road
z = 30'
175ฐ
o
- 15
11.2ฐ
7.6ฐ
8.1
Median
z = 14'
115ฐ
+ 15ฐ
14.5ฐ,
10.0ฐ
4.9
Grand
Central
Parkway
Service
Road
z = 25'
5ฐ
- 5ฐ
11.9ฐ
8.9ฐ
3.7
Median
z = 14'
125ฐ
- 15ฐ
o
15.4
8.8ฐ
2.5
BQE at
Navy St.
North
Edge of
Road
z = 20"
75ฐ
- 5ฐ
9.3ฐ
7.1ฐ
3.3
Median
z = 14'
15ฐ
- 5ฐ
19.8ฐ
9.9ฐ
2.9
Canal St.
North
Roof
z = 75'
15ฐ
-15ฐ
10.8ฐ
6.8ฐ
2.9
North
Trailer
z = 14'
5ฐ
+55ฐ(7)
9.2ฐ
7.2ฐ
2.7
Geo. Wash.
Bridget2)
Westbound
Divider
z = 14'
M
+5ฐ
M
7.2ฐ
2.9
.
.
M
M
M
M
-------
TABLE 2. 1- 2 (NOTES) METEOROLOGICAL SUMMARY
1. For exact vane locations, see Section 4. 0
2. This vane was designated vane 4
3. Based on only 1 day of available data
4. Bimodal with peaks of identical magnitude - most other azimuth
distributions were also bimodal
5. "M" indicates missing or insufficient data
6. Data at this site was taken with makeshift vector vane tail and
therefore sigma channels should not be used in inter site comparisons
^- 7. Broken and unbalanced tail section caused large positive values
-------
2.1.5 Hydrocarbon Concentrations at the Study Sites
The data pertaining to the hydrocarbon concentrations at the
study sites are tabulated in Appendix B. In general, the diurnal
pattern at all sites showed increased concentrations during the daytime
and a general reflection of rush hour traffic peaks. The highest con-
centrations occurred at the Canal Street and George Washington Bridge Plaza
sites. Bruckner Expressway and the Brooklyn Queen Expressway at Nelson
and Jessup Avenues showed the lowest hydrocarbons. The average daily
concentrations at each of the sites is shown on Table 2.3-1.
Figure 2.1-22 shows the effect of the average hourly traffic flow
rate on the change in hydrocarbon concentrations at the nine sites
measured. The city street, Canal Street between Church and Mercer Streets,
exhibits a marked change in hydrocarbons with small traffic variations.
The other sites show much lower changes in concentration with increases in
traffic flow. The curves for all nine sites suggest that the hydrocarbon
concentrations at low traffic flow rates are more a function of background
concentration than traffic flow at the site.
2.1.6 Particulate Concentrations and Trace Metals at the Study Sites
The range of particulate concentrations recorded at each site is
included in Table 2.3-1. These data represent differing number of samples
at each site and were taken at varying distances from the test roadways. The
average particulate concentration for all sites was 164 jug/m3 at an average
distance of 108 feet.
Since the particulate sampler was physically located outside the con-
fines of the test roadway, no firm conclusions can be derived on the influence
of site configuration on particulates. Figure 2.1-23 shows the average concen-
tration recorded at each of the sites (except BQE at Park and Navy at which 122
jug/m3 were measured, 500 feet from the roadway) versus the horizontal distance
between the Hi Vol Sampler and the middle of the roadway. Valid data at two
distances from the test roadway is available from only the FDR site. A reduc-
tion in particulate concentration did occur at that site with distance. As
can be seen from Figure 2.1-24, configuration appears to be more significant
than average daily traffic volume. (Two way traffic data was used for all
sites except the Brooklyn Battery Tunnel).
The analysis of the particulates for trace metals showed that lead was
higher than all other metals at all test sites. As seen in Table 2.1-3 Canal
Street with 3.7 ug/m3 was highest, while the 3.4 wg/m3 measurement at the George
Washington Bridge Pla?a was the next.
-------
VEHICLES PER HOUR
-------
it
ฑ
T l
.:tti
ill
T1|
Ill
:ฑฑ
-f}
fj.
4ฑt
ilrj
EJ
. 3 !>
' ie
j .
ji.
itl
it}-
i.tr.
t
tt
m
TF
41
-------
+ -
IT
I
' i ;ti T :
H
::ni
ฑ
uu
f-H-f
Y-
Ij
j-
fru
i
i
4 -i-Ht
i-H.:-
1
Ttn
IVS'
1
1
Itt
i. .4
hi
IT:.-
J.
tit
il-ti
lit
tff
I
rt-fh
..if.-
ii
ffl
11;
tn;
't!
I , |
lit
-------
Site Trace Metals
Pb Ni Cu Fe Cd Mn Cr Zn
FDR Drive 2.31 T .101 1,49 .002 .006 T
BBT 3.07 .021 .060 1.04 .005 .016 T .271
Hicks St. 2.69 .006 .046 0.91 .002 .022 T
Grant Circle No Data
Nelson & Jessup 3.02 T .179 0.98 .006 .019 T .257-
Bruckner 2.38 T .103 1.27 .004 .022 T .154
Grand Central Ey. 2.53 T .240 1.14 .005 .031 T .118
Park & Navy 2.31 T .111 .067 .006 .016 T
Canal St. 3.66 .029 .154 1.27 .005 .017 T
GWB Plaza 3.40 .013 .152 1.64 .003 .030 T
Ave. 2.82 .016 .127 1.16 .004 .020 T .200
Table 2.1-3 Trace Metals ug/m3
-------
2.2 Traffic Air Quality Relationships
The basic intent of this section is to present nomographs which
provide highway designers with gross estimates of the CO concentration
to be expected for various roadway configurations. Figures 2.2-1 to
2.2-10 present representative relationships of CO concentration vs
traffic flow rate for the ten sites monitored during this study. The CO
values shown in the curves are those of the vertical plane of probes
reflecting the maximum values. The traffic values shown are those of
the total traffic involved. The Brooklyn Battery Tunnel and Hugh J.
Grant sites show the one way traffic passing thru the westbound tubes.
The other eight sites show the two way traffic traveling on the road-
ways.
Sites such as FDR Drive, Brooklyn Battery Tunnel, Hugh J. Grant
Circle, Canal Street and Hicks Street show curves with steep slopes
indicating strong sensitivity of roadway configurations to traffic
volume. This is to be expected from configurations which tend to
limit the dilution effect of local meteorology. Conversely, the
more open sites, such as Bruckner Expressway, show curves with a
small slope.
The Hugh J. Grant site shows an interesting pattern of an
inversion and small vertical gradient. This is probably due to the
lack of forced ventilation within the tunnel. The Grand Central Park-
way curves show a higher CO concentration at the 10 foot level than at
the 3 foot level. This results from the lack of ventilation in the .
corner of the double cantilever structure. FDR Drive at Sutton Place
and the Trans Manhattan Expressway at George Washington Bridge Plaza
present the highest concentrations of the 10 sites. This is due both
to the geometry of the sites and to the volume of traffic involved.
The traffic air quality relationships for the nine sites at which CO
concentrations were monitored 3 feet above the roadway surface is shown
on Figure 2.2-11.
Figure 2.2-2 for the Brooklyn Battery Tunnel shows the relationship
between traffic flow and CO concentration for uni-directional traffic within
a tube. Data for the two way traffic periods (when rush hour traffic is
handled by three lanes of the tunnel) are omitted to more accurately relate
to the westbound traffic count. Bi-directional traffic within a tube is
marked by an increase in CO concentration in that tube consistent with
opposite direction rush hour periods.
-------
Sites that are characterized by traffic slow downs such as Canal
Street and Hicks Street show increased CO concentrations disproportionate to
traffic flow. Figure 2.2-12 presents the average weekday diurnal CO concen-
trations measured across the road at Hicks Street. Both the Hicks Street
wall and median strip curves follow the variation in traffic volume. The
west wall curve, however, does not relate to the quantity of either the south-
bound or total traffic. The high afternoon and evening CO concentrations at
this wall reflect the slow down of southbound traffic approaching the Brooklyn
Battery Tunnel turnoff.
-------
-------
-------
-------
^VEHICLES PER HgUI
-------
VEHICLES: PER HOUR,-,-;-
-------
-------
-------
-------
2.3 Site Ranking by Air Pollution Severity
This sub-section contains a comparison of monitored concentrations
with EPA standards in 2.3.1, Vehicular Pollution Factors in 2.3.2 and Site
Ranking by Concentration in Breathing Zone of Nearest Receptor in 2.3.3.
2.3.1 Comparison with Standards
The ten sites evaluated were examined from the viewpoint of the
National Air Quality Standards which have been designed to protect public
health and welfare. These standards are more directly applied to those
areas where human health of the general public would most likely be affected;
namely, dwellings, parks, sidewalks, etc. instead of on the actual roadway,
inside tunnels and on shoulders of roadways. These latter areas could have
health effects on automobile drivers, toll attendants, and maintenance men.
Table 2.3-1 presents the salient data obtained at the ten sites.
Besides the obvious result that the maximum concentration was found directly
over the roadway the following conclusions are drawn in light of the data
obtained during the study.
1. Receptor sites adjacent to two of the test roadways, as shown in
Column 6, were found to exceed the National Air Quality Standards
for carbon monoxide* (9 ppm-maximum 8 hour concentration).
a. Covered on top - open at side site, Figure 2.1-3A (Sutton Place
over F.D.R. Drive). The closest ground level receptor site was
in a nearby park. In addition, carbon monoxide measurements were
made inside and outside a fifth floor apartment.
b. Short tunnel site Figure 2.1-2B - Cross Bronx Expressway at
Hugh J. Grant Circle. The closest receptor site was in a park
30 feet above the roadway.
2. Two sites, as shown in column 10, were found to exceed the National
Air Quality Standards for particulates (260 micrograms/M3 - maximum
24 hours).
a. Short tunnel site, Figure 2.1-2B on Cross Bronx Expressway at
Hugh J. Grant Circle.
b. Deep cut site, Figure 2.1-IB on Cross Bronx Expressway at Nelson
and Jessup Avenues.
3. At a distance of 3' off the roadway, all sites (column 7) exceeded the
National Air Quality Standards for carbon monoxide except the viaduct and
grade road sites.
* Hourly data averaged over a minimum of 14 days.
-------
12
14
Site
Covered -on-Top ,
Open on One Side
Franklin D. *
Roosevelt Drive
@ Button Place.
Long Tunnel
(Ventilated)
(Brooklyn
Battery Tunnel)
Short Tunnel
(Cross Bx.
Expressway)
Deep Cut (Cross
Bx. Expressway)
Shallow Cut
(Brooklyn-Queens
Expressway)
City Street
(Canal St., NYC)
Grade Road
(Bruckner
Expressway)
Viaduct (Brook-
lyn-Queens
Expressway)
Cantilever
(Grand Central
Parkway)
Intermittent
Span (Trans-Man-
hattan Expwy.)
Dir-
ec-
tion
SB
NB
EB
WB
EB
WB
EB
WB
NB
SB
EB
WB
NB
SB
NB
SB
EB
WB
EB
WB
Vehicular
Pollution
Factor
( $ )
(PPM-hr.)
(veh.)
X 103
21
21
6
Not
Calcu-
lated
it ti
2.4
4.8
8.4
12.7
2.1
20.0
19.8
\2.6
3.3
8.1
3.0
1.8
5.2
.8
6.5
3.6
4.8
4.9
7.5
7.5
Dilution
Volume
X 10-6
.11
.11
.40
Not
Calcu-
lated
M it
7.5
3.8
2.1
.15
.1
2.1
1.9
5.5
6.0
1.0
1.9
3.8
5.0
3.7
4.4
2.8
2.8
Regression
Slope
(PPM-hr.)
(veh.)
x io3
Direction-
al Traf-
fic/Total
Traffic
26 13
11
8 4
30 20
20
10 5
(Blower)
16
14
3 1
1.8
3 2
.6 .4
2
.9 .2
2
1.8
1.7
7.6
4.8
7.2
Maximum
Hourly CO
Concentra-
tion at
Nearest
iReceptor
(PPM)
28
Back -
Ground
26
8
11
17
9
6
15
15
Peak
Hourly CO
Concentra-
tion on
Roadway
(PPM)
164
200
200
71
120
64
34
26
42
85
Average
Daily
Hydro -
Carbon
Concentra-
tion
(PPM)
7.5
No .
Data
Taken
6
4
5
14
3.5
11
9
14
Soiling
Index
(COHS/
1000
Lineal
ft.)
1.7
No Data
Taken
1.6
2.3
1.9
1.0
0.7
0.6
1.1
1.0
Particulate
Concentra-
tion
(/4A /M3)
Hi Vol.
150-212
142- 180
137-327
88-352
82-214
183-232
130-214
115-129
67-232
82-200
Percentage
of
Partic-
ulate
Site
Less
Than 1 JA.
72
42-72
No Data
Taken
70-72
66-68
No Data
Taken
71
No Data
Taken
75
No Data
Taken
Average
Peak
Hourly
Traffic
Volume
3370
3820
2500
2430
2860
2570
3590
4010
3430
2980,
670
860
2550
2140
3910
3750
6000
6120
6250
6140
Average
Total
Daily
Traffic
Volume
47,007
49,615
25,676
25,224
43,933
41,957
58,096
60.027
53,968
53,009
12,645
15.032
29,232
25 ,043
56,448
51,780
56,934
52,911
8/,S88
87,688
Avera0,
Daily
Traffic
Speed
(MPH)
(35)
(35)
(25)
(25)
42
48
49
47
38
38
17
15
48
48
38
38
46
44
jy
37
-------
2.3.2 Vehicular Pollution Factor
Eight bf the ten sites were examined on the basis of the vehicular
pollution factor*, as shown in column 3, Table 2.3-1. When the vehicular
pollution factor, 0 computed over all sites is plotted against the average
traffic speed for all sites a linear relation results, as shown in Figure 2.3-1.
Thus traffic speed seems to be the dominating influence on vehicular induced
turbulence, since (p is defined at the exhaust plane level (31 off the roadway).
At the exhaust plane, vehicular speed is far more significant than the
local meteorological effects. The peak hourly CO concentrations on the roadways
(column 7 of Table 2.3-1) all occur at the exhaust plane level during rush hours
when vehicle speed is well below the daily average. The obvious conclusion to
be drawn by urban transportation planners from the data presented in Figure 2.3-1
is that an express roadway moving at higher average traffic velocities can accommo-
date more cars and produce lower carbon monoxide concentrations than a clogged city
street. This result is a consequence of higher vehicular efficiency/lower emission
of pollutant per travel distance and vehicular induced turbulence at the higher
speeds. The linear relationship shown in Figure 2.3-1 between traffic speed and
vehicular pollution is one of the important discoveries of this study. This is
probably due to the quantity of data and the close proximity of the sampling probes
to the roadway.
When the vehicular pollution factor is compared on the basis of constant
traffic volume, the influence of site configuration is seen. The average daily
total traffic speed at 3 sites, i.e., Hicks Street, Park & Navy and George
Washington Bridge Plaza, Was 38 mph. The total vehicular pollution factor is
lowest for the viaduct site and highest for the shallow cut site. Thus, the more
open the site, the lower the pollution at the roadway.
2.3.3 Ranking by Concentration in Breathing Zone of Nearest Receptor
The breathing zone concentration of the nearest receptor is another method
by which the various sites may be ranked. Table 2.3-2 lists the maximum averaged
hourly carbon monoxide concentration at the nearest receptor location, the linear
distance from the roadway to the receptor location and the total number of days
over which the data was averaged. As expected, those receptor locations that are
the closest to the roadway yield the highest values, with the exception of the via-
duct site where the ventilation was particularly good and the city street site where
the traffic volume was small.
. ["(ppm-hr)!
* Vehicular pollution factor, 0,1 vehicleJ is the hourly average concentration
produced per vehicle at the exhaust plane, corrected for cross-roadway diffusion
as explained in Section 3.3. Thus the factor is equivalent to an emission rate.
** The viaduct site was surrounded on one side by a park 2600 ft. wide on the one
side and a playground apartment building courtyard 350 ft. away on the other
side.
-------
Q
W
o
n
60
55
50
45
40
35
30
25
ZO
15
10
5
RELATION BETWEEN
VEHICULAR POLLUTION FACTORS TAKEN AT EXHAUST PLANE
fe '
TRAFFIC SPEED
O
o
12 14 16 18 20 22 24 26
PPM - HR
8 ' 10
VEHICULAR POLLUTION FACTOR, $,
VEH
Figure 2.3-1
-------
Site
Maximum Averaged
Hourly CO
Concentration at
Receptor Location
(PPM)
28
Hour
0900-1000
Covered on top
Open at Side
Shallow Cut
Short Tunnel
Deep Cut
Grade Road
Cantilever
Viaduct
City Street
Intermittent Span
* Nearest Receptor Location directly under Viaduct
Shortest Distance
from Roadway to
Receptor Location
(Feet)
20
Number of Days
over which Data
Was Averaged
11
11
26
8
9
15
6
17
15
1600- 1700
1500-1600
1700- 1800
0700-0800
1700- 1800
0700-0800
1700- 1800
1700- 1800
1700- 1800
62
30
82
98
30
10*
6
45
13
12
11
11
11
13
11
8
Table 2.3-2
-------
Rigorous comparisons are difficult since both traffic speed and volume
flow rates vary from site to site. However, using the breathing zone of the
nearest receptor, the ranking and the distances between receptor and roadway
and the sites is presented in Table 2.3-3. It is interesting that the higher
carbon monoxide concentrations at the receptor are associated with the two sites
which are covered. Also, it is these sites which exceeded the Federal Air Quality
Standards. This suggests that caution in locating parks and air rights structures
in the immediate vicinity of covered sites should be observed.
-------
Maximum Averaged Hourly Distance Between Nearest
CO Concentration at Re- Receptor & Roadway in
Site ceptor in PPM : feet
Viaduct
Deep Cut
Grade Road
Shallow Cut
Cantilever
Intermittent Span
City Street
Short Tunnel
Covered on top-
Open at side
6
8
9
11
15 .
15
17
26
28
10
82
98
62
30
45
6
30
20
Table 2.3-3 Ranking of Sites by Breathing Zone of the Nearest Receptor
-------
2. 4 Indoor-Outdoor Relationships
Figures 2.4-1 through 2.4-5 show the hourly mean values for carbon
monoxide at each of five sites where both indoor and outdoor concentrations
were monitored.
In general the outdoor concentrations tend to be higher but only seldom
does this difference exceed 4 PPM. The graphs show that, as the outdoor
values increase, the indoor ones also go up but with a time lag of up to two
hours. This is also true as the outdoor concentrations decrease although
sometimes, if ventilation of the building is not good, the inside levels will
remain elevated above the outside values for longer periods of time. Time
of year, because of open windows and doors, will affect these ventilation rates.
Both indoor and outdoor concentrations follow each other closely and
have the same general shape. In the case of air rights structures, the indoor
weekend levels were generally higher than outside reflecting increased cooking
and household activities. At the George Washington Bridge Apartments the
weekday indoor levels were mostly higher than those outside at the 3rd floor
level. This could be a result of the flow of air in the building from the ground
floor to the roof. Weekday values at the 9th floor level show higher than the
3rd floor and this results both from the upward air flow plus the addition of
carbon monoxide from apartment activities on the intervening floors.
Three weekday and two weekend concentrations at four sites for an
eight-hour period exceeded the National Primary Ambient Air Quality Standard
of the Federal Government. At Canal Street, the Brooklyn Queens Expressway
at Hicks Street and Button Place the levels were greater during weekdays
indoors than the 9 PPM allowed. At Sutton Place and Grand Central Parkway
at Parsons Boulevard, the weekend levels also exceeded this value. As shown
on the graphs, these concentrations are dependent upon outside levels.
-------
c
o
o
U
O
o
c
(4
4)
I
ft.
ง
4)
O
O
O
O
e
2400
0400
0800 1200 1600
Time - Hours
Fig. 2. 4-1A Weekday Indoor/Outdoor CO Concentration
2000
2400
2400 0400 0800 1200 1600 2000 2400
Time - Hours
Fig, 2.4-1B Weekend Indoor/Outdoor CO Concentration
Fig. 2.4-1 Diurnal Indoor & Outdoor Carbon Monoxide Concentrations
in Air Rights Building at Sutton Place
-------
e
ft
%
te
8.
O:
o
1200 1600
Time - Hours
2A Weekday Indoor/Outdoor CO Concentration
I
0,
ฃ3
O
2000
2400
1200 1600
4
Time - Hours
Weekend Indoor/Outdoor CO Concentration
Diurnal Indoor & Outdoor Carbon Monoxide Concentrations
Hear a Double Cantilever Covered Open Top Middle at
Central Parkway at Parsons Boulevard
-------
I
4)
o
O
O
O
a
10 ::
2400
0400
0800
1200 1600
Time - Hours
Fig. 2.4-3A Weekday Indoor /Outdoor CO Concentration
2000
2400
fc
c
o
25
20
fi 15
o 10
O
I 5
4)
^ 0
2400
0400
0800
1200 1600
Time - Hours
2000
2400
Fig. 2.4-3B Weekend Indoor /Outdoor CO Concentration
Fig. 2.4-3 Diurnal Indoor & Outdoor Carbon Monoxide Concentrations
Near an Open Cut on Brooklyn- Queens Expressway at
Hicks Street (in a School)
61
-------
I
ti
o
u
c
o
u
O
U
C
n)
0)
25
20
10
2400 0400 0800 1200 1600
Time - Hours
Fig. 2.4-4A Weekday Indoor/Outdoor CO Concentration
2000
2400
5
cu
c
o
1-4
s
1
0)
- o
a
o
U
O
O
a
nJ
(I)
2
25
20
10
2400
2000
2400
0400 0800 1200 1600
Time - Hours
Fig. 2. 4-4B Weekend Indoor/Outdoor CO Concentration
Fig. 2.4-4 Diurnal Indoor & Outdoor Carbon Monoxide Concentrations
on Canal Street (in a Building)
-------
c
V
u
o
c
g.3
v ri
i
C
V
u
e
o
o cm
v nt
o>
8
o ซ5
uฃ
oou
^ c
S ฐ
ro +J
4) nl
15
10
2400
0400
0800
1200 . 1600
Time - Hours
Fig. 2.4-5A Weekday Indoor/Outdoor CO Concentration
2000
2400
15
10
2400
0400
0800
1200 1600
Time - Hours
Fig. 2.4-5B Weekend Indoor/Outdoor CO Concentration
2000
2400
2400
0400
0800
1200 1600
Time - Hours
Fig. 2.4-5C Weekday Indoor/Outdoor CO Concentration
2000
2400
2400
0400
0800
1200 1600
Time - Hours
Fig. 2.4-5D Weekend Indoor/Outdoor CO Concentration
2000
2400
Fig. 2.4-5 Diurnal Indoor & Outdoor Carbon Monoxide Concentrations
' . in an Air Rights Building Over the Trans-Manhattan
Expressway Near the George Washington Bridge Plaza
-------
2. 5 Recommendations for Future Research
This section consists of three parts. Recommendations for further
measurements of traffic, air quality and meteorology factors are given in
2.5. 1. Recommendations concerning model development based on data from
this study and on further measurements are in Section 2.5.2. Recommenda-
tions for health effects study are given in Section 2.5.3.
2. 5. 1 Recommendations for Further Measurements
Two fundamental questions in the development of good predictive
models for traffic generated carbon monoxide are:
The form of the law of decay of concentration with distance
from the source
The role of traffic generated turbulence in providing a
diffusive mechanism for the transport of the traffic
generated carbon monoxide
Analysis of the data gathered in the course of this study strongly suggests
that the vertical decay with height above the source plane is exponential in
character when the top of the site is open to ventilation. The situation is
not as clear in the case of tunnels. The law of decay of concentration with
.horizontal distance from the source is also not entirely clear. With the
exception of the Sutton Place site, there were not generally enough probes
at a given height to obtain a good estimate of the shape of the horizontal
concentration profile at that height. While it is true that the suspicion
exists that the horizontal profiles are Gaussian in shape, the need remains
to verify that this is true. Because the concentration gradients are often
quite small, they tend to be masked by instrument error. This makes it
difficult to obtain good estimates of the shape of the concentration profiles,
even with a larger number of probes, if absolute values of the concentration
are taken at'each point. Instead, the procedure that should be used is to
obtain the difference concentration between each pair of probes employing a
single long path infra-red instrument fitted with a floating reference cell to
replace the fixed reference cell. The(Intertech, URAS-2 instruments used
in this study can employ floating reference cells marketed by the Intertech
Corporation.:
Indications of turbulence were obtained, at each of the sites in this
study, by means of the Sigma Meters discussed in Section 3. However the
need exists for a more direct measure of the quantities u^c^ป v^c* anc*
wlcl at each of the points at which a carbon monoxide probe is located.
These quantities which define the turbulence at a given point over some
averaging time have the following meaning.
-------
1
T
L
T
v cdt
T
t - T
where:
T
w*cldt
t
* " 2
T = averaging time
c = fluctuation of the concentration, C, about the mean value c" at
any instant during the averaging time
1
u , v*, w = fluctuation of the x, y, z components of wind, u, v, w,
about the means U, v", w at any instant during the
averaging time.
Since it can be shown that
vlcl = vc - v c
w
it is possible to determine these correlation functions directly at any point
where a carbon monoxide probe and a device which determines wind speeds
wind elevation and wind azimuth angles can be co-located. A simple, hybrid
circuit, turbulence cross correlation computer has been designed at General
Electric and could be built inexpensively. The solid state circuit elements
are similar to those used in the Data Converter developed by G. E. for APCO
in the Indoor /Outdoor Study. With such a device it would be possible to obtain;
turbulence measurements for various averaging times; diurnal variation of
turbulence; correlations of turbulence with traffic volume and speed, concen-
tration gradient, wind speed and direction and other pertinent variable So
-------
2.5.2 Recommendations for Further Model Development
The Superposition Model described briefly in Section 2. 1.3 and in
more detail in Section 3 has been reasonably successful in estimating traffic
generated carbon monoxide concentrations. It is felt, however, that the model
can be improved by modifying the assumed law of exponential decay with radial
distance from the source. It is believed that if the decay law assumed expo-
nential decay with height above the source and a gaussian decay horizontally,
that predictive accuracy would be improved. Accordingly, it is recommended
that such a modified Superposition model be exercised on the data gathered
during this study.
Because the superposition model does not contain wind and turbulence
terms explicitly, it is further recommended that a more theoretical model be
pursued. The basic approach would consist of attempting a solution to the basic
differential equation
using the data gathered in the turbulence experiment recommended in Section
2.5.1.
Although the superposition model was applied to the two tunnel config-
urations, the air quality data suggests that tunnels do not show a continuous
line source. Close to the ceiling exhaust fan ventilation and the "piston" effect
of the vehicles in tunnels provide a controlled environment whereby the
Vehicular Pollution Factor, (p, can be evaluated more vigorously. Also in a
tunnel environment, vehicular induced turbulence can be measured as a func-
tion of speed which will allow for better predictive techniques. This latter
point is a, key one since no speed data was taken in the tunnel and the speed
data taken for the rest of the sites was sparse. It is recommended that a
tunnel study be undertaken in which wind speed, wind turbulence, vehicular
speed, vehicular volume, and vehicular mix be measured in order to develop
a more adequate tunnel model, a better definition of vehicular induced turbu-
lence, and a better definition of the Vehicular Pollution Factor, <ฃ.
Finally, the speed taken did not provide sufficient range to define the
Vehicular Pollution Factor, $, for speed ranges above 50 mph and below 15
mph. There is also a range between 35 mph and 17 mph where the data is_
meager. Also, the strong linear correlation between ^> and vehicular speed
has not been found by previous workers and it is recommended that further
investigations be undertaken to determine the universality of this relation.
-------
2.5.3 Recommendations for Health Effects Research
It is recommended that a medical team do a continuing follow-up study
of the population residing near each of the sites in this urban expressway study.
Special attention should be directed toward the tenants in the two air rights
buildings.
-------
3.0 STUDY METHODOLOGY
-------
3.1 Measurement Techniques
3.1.1 Overall Methodology
In order to monitor the various traffic-derived pollutants and
wind parameters at the ten selected site configurations, a 30' aluminum
van was equipped with a variety of air pollution measuring devices.
This mobile laboratory provided the capability, as shown in Fig. 3.1-1, of
sensing, measuring and recording carbon monoxide, hydrocarbons, and other
related parameters. A New York State Department of Transportation Traffic
trailer was used to obtain traffic measurements at the test sites. These
measurements included daily traffic volumes, hourly vehicle counts, average
speeds, and some indication to the mix ratio between trucks, buses and
automobiles.
The GE Air Pollution Van and the New York State Traffic trailer
were parked adjacent to the roadway at each site. Both trailers were
supplied the necessary electrical power, from an external source, to
operate all equipment used in this study.
In order to study the relationships between pollution levels, traffic
patterns, and meteorological data, each parameter had to be simultaneously
monitored and recorded for final data analysis. Various pollution levels
and meteorological variables were continuously monitored at each site for
an extended period of the time to insure that a sufficient amount of relevant
data was obtained. Data was collected for several weeks at each site to
obtain both weekday and weekend coverage.
A sampling technique was utilized that allowed for the largest number
of point samplings to be made using the instruments available to systematically
relate pollution measurements with all other variables. A total of sixteen
possible probe locations at each site were monitored to obtain hourly CO
and hydrocarbon concentration averages. The probe locations were divided into
sets of eight probes each. Each set of eight probes was sampled alternately
for a period of five minutes. In a hour, each probe location was monitored
six times for a total of thirty minutes. Fig. 3.1-2 shows the flow diagram
in the trailer for CO and hydrocarbon measurements.
Polyethylene tubing was directed from an Intake manifold at the GE
van to specified positions over the roadway. The probes were all positioned
in a plane perpendicular to the roadway under evaluation to develop a two
dimensional CO profile of the site. This mapping technique displayed CO
contour concentrations as related to horizontal and vertical distances at the
roadway. This profile, in conjunction with data on traffic and meteorological
conditions recorded at the same time, gives valuable information as to vehicular
emission concentrations/CO diffusion characteristics, wind effects, and
calibration data for formation of theoretical mathematical site models which
can define similar site situations without actual investigation.
-------
o\
I.-"
ป."*
**
*
SeUw..Ji -- r^t
<ปปซป**ป
TAp*
.Uv
Q
7V6x/,"c
-7
\ c <;
1
-.
J / ป A
^
n i. v -
-^
! ซ K ฃ
1
*
1
J
ฐ
o
0
C,AS>
n
00
PO **ฃ v * ซ
3 **t T"~ev
-------
ฉ
g) <3 /// S/>**t C
I
-------
One Important study objective was to determine the pollution levels
to which local community residents were subjected as a result of automobile
emissions generated on the roadway. For this study, probes were placed
inside and outside nearby buildings in order to define the Indoor-Outdoor CO
concentration relationships at various distances from the roadway.
Inside the GE van, instruments continuously measured all important
wind parameters. The sensing units were vector vanes capable of measuring
wind speed, azimuth direction, and elevation. These vanes were strategically
positioned at various site positions in order to define the local wind
characteristics. Usually one vane was placed on the medial strip between the
traffic lanes in order to sense traffic-generated turbulence and other wind data
at the roadway. Another vane was placed away from the roadway in order to
get overall wind information at the particular site. The number of vanes used
and the locations selected, depended on site configuration and operator accessi-
bility to vane for general maintenance and repair. In addition high volume
air and tape samplers were positioned at the site to measure suspended
particulates and trace metals (especially lead).
3.1.2 Air Quality Measurements
3.1.2.1 Carbon Monoxide
Carbon monoxide was measured using an infrared analysis technique.
The measuring principle of the CO analyzer makes use of the specific
radiation absorption band of carbon monoxide in the infrared range. A
total of eight analyzers (Intertech Corp. Princeton, N.J.) were used in
this study. The instrument was usually operated on the 0 to 100 ppm CO
range and had the capability of measuring concentrations of less than one
ppm CO in the sampled gas. The inherent zero and span drift for the
instrument was ^~_ 27o of full scale per week. The analyzers were calibrated
daily (except weekends) to ensure maximum accuracy. Nitrogen gas (zero grade)
and a standard carbon monoxide (80 -Jy90 ppm) in Nฃ mixture were used for
calibration. Each analyzer sampled two probe locations for five minute
intervals. Each CO analyzer was equipped with a pump which pulled air from
one of two probe locations, through the polyethylene tubing, to the intake
manifold of the air pollution van. From the manifold, the sampled gas from
each location passed through a solenoid valve that was electrically controlled
by a relay connected to a five minute timer. From this point, the gas flowed
either to a CO analyzer or was by-passed to a ballast tank for mixing depend-
ing on whether the solenoid was or was not actuated. The timer controlled
a total of 16 solenoid valves which in turn controlled the direction of the
individual gas flows. During one five minute timer interval, eight probe
locations were analyzed for carbon monoxide while the other eight were fed
into a ballast tank where they were mixed and analyzed for total hydrocarbons.
A Hi Capacity pump was connected to the ballast tank to ensure that
delay time due to solenoid switching was kept to a minimum.
Before going to the CO analyzer, the sample gas was passed through
a porous filter in order to remove all particulate matter. Next, the gas
was passed through a refrigerator containing a condensing coil which kept
the amount of water vapor in the sample stream at a constant level. Interfer-
ences from water bands were thus eliminated since the amount of water was
-------
the same in both sample and calibration gases. Finally the gas passed
through a flow regulator and into the CO analyzer for analysis. All exhaust
gas is vented to the outside of the trailer.
Principal source errors in CO system were as follows:
1. Errors due to inaccurate instrument calibration
2. Inherent span and zero drifts in the instrument
3. Large humidity changes causing water vapor to interfere
with CO absorption bands
4. The difficulty in reading hourly CO averages from the
recording charts
5. Inability to differentiate between CO levels of two sampled
locations due to small concentration differences
6. At any particular time, eight probe concentrations are not
being sampled due to 5 minute cycling procedure
3.1.2.2 Hydrocarbons
Hydrocarbon concentrations were measured by a flame ionization method
of detection. The sensor is a burner where a regulated flow of sample gas
passes through a flame sustained by regulated flows of hydrogen and air zero.
Hydrocarbon components in the sample stream undergo a complex ionization that
produces electrons and positive ions. The ions produce an ionization current
proportioned to the rate which carbon atoms enter the burner. Thus hydrocarbon
concentrations are proportional to this ionization current. The hydrocarbon
analyzer (Beckman Instruments, Fullerton, Calif.) that was used in this study,
had a full scale sensitivity of 1 ppra CH4 to 2% CH4. The electronic stability
at maximum sensitivity was + 1% Full Scale. Reproducibility was 1% of full
scale, for successive identical samples. All readout concentrations are
expressed as ppm of CH4, since CH4 was the particular hydrocarbon present in
the span gas.
Since only one hydrocarbon analyzer wa.s available, a sampling scheme
was used to obtain an average value for all probes sampled during each five.
minute cycling interval. The eight gases from probes not being analyzed for
CO were mixed in the ballast tank from which an average hydrocarbon concentra-
tion was obtained.
3.1.2.3 Particulates and Trace Metals
In order to determine the amounts of particulates and trace metals in
the air at each site, a 24 hour air sample was passed through a weighed porous
filter. At the end of the sampling period, the filter was weighed again and
the total weight of particulates deposited was calculated by subtracting the
total weight from the initial filter weight. A High Volume air sampler was
utilized for this portion of the study. Each sampler had its own calibrated
flowmeter which indicated air flow rate through the porous filter. By taking
a flowmeter reading before and after the test, an average flow rate was
estimated. The total weight of particulate matter obtained divided by the
-------
volume of air sampled, gave a weight per cubic meter of particulate matter
in the sampled air. The filter with the deposit was analyzed for lead and other
trace metals by using an atomic absorption technique.
3.1.2.4 Tape Sampler
A tape sampler was a!so utilized to indicate the amount of
particulates in the air at each site. A two hour air sample was pulled
through a portion of tape by a pump. After the 2 hour sampling, the tapa
was automatically fed to a new unexposed portion of tape. A darkened
circular spot was obtained whose darkened appearance is proportional to
the amount of particulates in the air.
3.1.3 Traffic. Measurements
Traffic measurements and pollutant concentrations are mutually
interdependant. To secure the traffic parameters associated with vehicle
emissions being recorded, data for volume, classification and speed were
recorded at the location of each pollution monitoring site.
Traffic volumes were obtained by means of a Fischer Porter Punched
Tape Recorder utilizing pneumatic road tubes for detection of vehicles.
These vehicle impulses were stored on a memory code dish and digitally punched
on a tape at 15 minute intervals. The punched tape was then run through
a translator, the output of which was used to prepare the hourly traffic
summary sheets.
Traffic volume counts are a product of axle impulses. The Fischer
Porter counter automatically records one vehicle count for each two impulses
received. On the F.D.R. Drive and the Grand Central Parkway, where commerical
traffic is prohibited, no adjustment for multi-axle vehicles was necessary.
At the eight remaining sites, vehicle counts were recorded in duplicate;
the punched tapes were compared for correlation; the resulting hourly readings
edited to eliminate extraneous values and hourly values selected which conform
to the established prevailing traffic pattern for each site. Classification
counts of multi-axle vehicles were obtained manually and the machine counts
were adjusted to reflect the reduction in transit of vehicles occasioned by
the passage of multi-axle trucks over the detectors.
j
At site 3, the Brooklyn Queens Expressway at Hicks Street, classifica-
tion counts were obtained only between the hours of 8:00 a.m. and 4:00 p.m. and
at site.8, the Brooklyn Queens Expressway at Park and Navy Streets, classifica-
tion counts were obtained between the hours of 7:00 a.m. and 6:00 p.m. The
corresponding hourly traffic volumes were adjusted to reflect the affect of
the multi-axle vehicles. The uncorrected 24 hour traffic summaries are avail-
able for comparison.
-------
For the purpose of classification, those vehicles having three or
more axles or two ascle vehicles having a capacity of two tons or more,
were considered in the category of trucks. All light trucks, pick ups,
small vans and delivery wagons were included in the count by passenger
cars. The multi-axle vehicles were identified as gas or diesel powered.
Under normal operating conditions, vehicular speeds approximated
the posted speed limits. During periods of congestion, speeds dropped to
minimum values. To measure these conditions, directional speeds were
recorded at 15 minute intervals by observation and stop watch check. The
beginning and ending times of traffic tieups were recorded. Radar speed
recorders were also used to monitor speed. At Site 10, speed was also
measured by recording the length of time required for passage of a vehicle
thrcagh a 10 foot trap. The trap consisted of two pneumatic tubes spaced
10 feet apart which started and stopped 1 K Hertz signal which was recorded
on magnetic tape. Speed readings were sampled at 5 minute intervals. The
magnetic tape was then run through a Beckman Highspeed Digital Analyzer from
which vehicle speeds were tabulated.
3.1.4 Meteorological Measurements
Meteorological s^ensors were incorporated into the monitoring scheme
in recognition of the important effect of weather on the distribution and
dispersion of pollutants. Vector Vane equipment, manufactured by Meteorology
Research, Inc. of Altadena, Calif, was used to monitor wind conditions. The
Vector Vane is a lightweight, quick response, low threshold instrument which,
with its associated electronic package, measures wind azimuth (0ฐ - 540ฐ),
wind elevation angle (+60ฐ to -60ฐ), and wind speed (0-80 mph). The response
characteristics of the Vector Vane sensor are as follows:
Starting Threshold Speed 1.0 mph
Direction 1.0 mph
Response Distance Speed 2-3 feet
Delay Distance Direction 2-3 feet
Damping Ratio Direction 0.4-0.7
Overshoot Direction less than 10%
Linearity Speed + 1% of pitch
Direction + 17,
The Sigma Meter portion of the Vector Vane system continuously calculated
the standard deviation (sigma) of the fluctuations of the azimuth and
elevation angles. A five-second sampling time for the sigma computations was
used almost exclusively since the main interest lies in the higher frequency
components of turbulence. The Vector Vane Sigma Meter/Transmuter provided
analog outputs for each of the above wind functions in the form of a 0-5 VDC
signal.
-------
Two complete Vector Vane systems were used during the course of
this study. One sensor was usually located near the probe plane and on the
median of the road being studied. It was expected that this vane, being
as close to the traffic as possible without being in danger of damage,
would record any turbulence generated by the motion of the vehicles them-
selves^ as well as natural turbulence. Turbulent atmosphere motions
generated by passing traffic were expected to be of very limited scale in
both time and space. It is for this reason that a sampling time constant
of 5 seconds was chosen for the sigma computations for this vane. The
remaining system was installed, whenever practical, at a higher elevation
and away from nearby obstructions, to obtain a record of the more general
wind conditions in the area. A five second constant was also used for
this system to permit comparison with the road wind system.
The azimuth, sigma azimuth, elevation, and sigma elevation channels
from each system were recorded on four channel Esterline Angus strip chart
recorders. The wind speed channel from each system was recorded on one
channel of a shared two channel Esterline Angus panelgraph recorder. The
wind systems were operated twenty-four hours per day at each site except
for shutdowns due to damaged or malfunctioning equipment. The strip chart
recorders were operated at a speed of three inches per hour.
Strip chart traces were manually digitized with data averaged and
recorded at one hour intervals. The calculation of the sigma azimuth function
from the strip chart trace was somewhat impaired by the electrical "crossover"
experienced in the azimuth channel. Although the 540 azimuth format and a
special switching system reduced these crossovers to a minimum, their effect
on sigma azimuth was significant, especially during turbulent weather. The
contributions of this"crossover to the sigma azimuth value were subjectively
removed, whenever possible, during digitization of the chart record. After
digitization, the data was keypunched and processed on the GE-635 computer
with specially developed software. The program translated the voltages read
from the strip charts to angles and velocities as well as performing a
statistical analysis on the data. The azimuth angle was translated according
to the following relation:
Azimuth Angle = (X - C) .540ฐ for upright sensor
= [I - X-C)] .540ฐ for inverted sensor
where: X is the number read from the strip chart
C is an orientation factor used to translate the
axes so that 0ฐ is parallel to the road.
A value equal to 360ฐ was added or subtracted from the answer as needed to
keep the angles in the 0ฐ to 360ฐ range. The sigma azimuth and sigma
elevation values were determined by simply multiplying the strip chart readings
by 45ฐ and 15ฐ respectively, the full scale values of the functions. Eleva-
tion angles were determined by the following formula:
Elevation Angle = (X - 120ฐ) - 60ฐ
where: X is again the number read from the strip chart (range 0 to 1.0)
-------
/
Wind speed values were read directly from the charts and required no trans-
lation. All missing data was designated by -1. or (-99. where positive and
negative values inherent in the data would make -1. ambiguous).
A statistical analysis of each wind function was performed and
means,medians, and standard deviations were printed. In the case of sigma
azimuth, sigma elevation, and wind speed, the means and standard deviations
were plotted ys time of day. However, azimuth and elevation angle data did
not easily lend itself to this type of analysis so a table of frequency of
occurrence and an associated histogram were printed instead.
Finally, the wind vector was resolved into components and the
statistical analysis was again performed. The hourly means and standard
deviations of each of the three components was plotted. On the print-outs,
the vertical velocity is identified as W (+ up, - down), the component
parallel to the road is identified as U (+ when blowing from direction of
0ฐ azimuth), and the component perpendicular to the road is V (+ when azimuth
angle is in the range of 0ฐ to 180ฐ). The sign conventions used are
portrayed in figure 3.1-3. The data reduceH and processed in this manner
was used in the regression and correlation analysis discussed in section
3.3.2. The complete set of wind observations and the analysis can be found
in Appendix C.
Axis of Road
Wind from 0
270C
180*
Axis of Road
Figure 3.1-3 Sign Conventions for Horizontal Wind Components
-------
3.2 Regression and Correlation Analyses
Regression and correlation analyses were performed to determine
whether the relationship between environmental variables and carbon monoxide
and hydrocarbon concentrations could be predicted mathematically.
3.2.1 Carbon Monoxide
The CO study was aimed at determining the effects of the various
wind parameters on the CO concentrations using linear regression.
3.2.1.1 Regression Equation
The relationship between the wind parameters; i.e., sigma azimuth,
siftma elevation, and wind speed, and the CO concentration was assumed to be
of the form
C = K0 AZ EL
Where
C = average hourly carbon monoxide concentration
Q^Z ป average hourly standard deviation of the wind azimuth angle
0EL = average hourly standard deviation of the wind elevation angle
V = average hourly wind speed
Kb,K^, K2, Kg = unknown constants to be estimated
To use this equation in linear regression analysis, it is necessary
to take the natural logarithm of both sides of the equation. Thus
In C = In [ Ko CTAZK1 CTELK2 V*3 3.2-2
and by using the properties of the natural logarithm we obtain
In C = In KO+ Kx In CfAZ + K2 In tf EL + K3ln V 3.2-3
If we now let
Y
x2
X3
Bo
Bl
B2
B3
= In C
= In &AZ
= InCTfiL
= In V
- lnK0
e Kj_
= K2
= K3
-------
then the above equation becomes
B
B
B2 X2
B3 X3
3.2-4
which is the form used in the regression analysis.
A computer program was developed to do the numerical computation in
the regression analysis on the G.E. 635 computer system. The program ac-
cepted as input a set of N observations for p + 1 variables of the form
12
21
X
22
IP
X.
n
X
nl
Vn2
X
np
,Xi0 ) represents one particular obser-
where the row (Yi X-ji X*
vation of the p + 1 variables.
For a particular run, one of the columns is selected as a dependent
variable (Column one in the above matrix is chosen for convenience) and the
remaining columns as the independent variables. Then using the linear model
Y = B X
Where
Y= <
r ^
Yl 'B
1 1 i ' -DO
Y2
YN
> -D / Bl
, D '
i
j
;BP
/ *. ;
> , x= <
1 Xn Xi2 ' ' ' " XIP
1 X21 X22 "" X2p
1 XN1 XN2 '*" XNP
* y
3.2-5
the least squares estimates of
B
o, B-^, ,...,Bp (iiijn means transpose),
namely b = bQ, bi ..., bp are computed
using the equation
T
X Y
("-1" means inverse)
3.2-6
and the predicted values of Y, namely, Y are computed using the relation
Y = Xb
3.2-7
-------
A measure of the success of the regression equation in explaining
the variation in the data is given by the multiple correlation coefficient,
R, and is computed using the equation
/bT XT Y - N Y2
R
- NY2
3.2-8
N
where Y =
- Y
N L^
Y is the mean value of the" N dependent variable observa-
tions. The values of R lie between zero and one. The closer to unity R is
the more successful the regression equation explains the data variation.
A measure of the linear relationship between the dependent variable
Y and any independent variable X^ is given by the simple correlation coef-
ficient, rjj and is computed from the equation
N
Cฃฑ - Y)(Xi:j - Xj)
N
> *
=1
, - 1/2 r- N
[ij ' V
ฃ
|_i=l
_ 2"
(Yฃ - Y)
1/2
3.2-9
where X*
variable
""l
is the mean value of the N observations of the j independent
The values of rj* lie between -1 and 1. The magnitude of rjj de-
termines the strength of the linear relationship between the variable, whereas
the sign of rjj tells whether Y tends to increase, or decrease, with increasing
The value of r^ will be equal to + 1 if, and only if, the points
..
lie on some straight line.
Since Y represents the natural logarithm of the CO concentration and
Xj_t X2, and Xg represent the logarithms of the sigma azimuth angle, sigma
elevation angle, and wind speed respectively, rjj , r^2 > an^ ri3 represent the
simple correlation coefficients between the logarithms of the CO concentrations
and the wind speeds respectively. The correlation coefficient, as stated pre-
viously, is a measure of the linear relationship between two variables. How-
ever, it is also a measure of how well an equation of the form
C = AZM
describes the relationship between the variables, C and Z where Z is either
Oaz (sigma azimuth angle), e^ (sigma elevation angle), or V (wind speed).
3.2.1.2 CO Regression & Correlation Analyses
Correlation analyses were performed, using the CO and meteorological
data gathered at each site, to evaluate the effectiveness of the postulated
regression equation. The numerical results of these analyses and the location
of the data sensors, are shown in Tables 3.2-1 to 3.2-12. The traffic flow
rate data provided by the New York State Transportation Administration on
April 13, 1971 was used.
-------
NUMERICAL RESULTS OF 54th AND BUTTON PLACE CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (Veh/Hr)
4130-4570
4420-4880
4700-5200
4990-5510
5270-5830
5560-6140
5840-6460
6130-6770
6410-7090
6700-7400
# of
Sample
Points
4
10
10
7
6
4
5
7
6
4
b b
o 1
Too few
9.55 -3.48
5.85 -1.60
-.28 -.75
,84 -.59
Too few
2.01 1.06
-.38 -.38
2.19 1.83
Too few
b
2
sample
.98
1.45
3.12
2.03
sample
-1,37
1.33
-.57
sample
b r r r
3 11 12 13
points
^.85 -.76 -,07 .38
-1.08 .04 .22 -.37
-.10 -.10 .35 -.23
.01 .31 .81 -.37
points
.45 .65 .53 -.36
.97 .63 .73 .59
-.93 .94 .89 .59
points
R
.88
.62
.53
.86
.74
.78
.95
CO data obtained 12' above the sidewalk alongside the instrumentation ran it the southwest
corner of the tunnel
Meteorlogical data obtained from Vane 1
4
Traffic flow rate intervals selected from total traffic flow rate data.
Table 3.2-1
-------
NUMERICAL RESULTS OF BROOKLYN QUEENS EXPRESSWAY AT HICKS STREET CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE # of
INTERVAL (Veh/Hr.) Sample b
Points ฐ
2840-3140
3410-3770
3700-4080
3980-4400
4270-4710
4550-5030
4840-5340
5120-5660
5410-5970
5690-6290
5980-6600
6260-6920
3
5
5
5
10
10
12
15
15
8
5
6
-8.24
-24.8
6.44
4.51
3.24
1.34
5.35
1.66
1.48
-3.05
3.67
b
Too few
.24
-3.91
-.86
1.42
.52
-.05
-.87
-1.96
.63
-.52
-.15
b
2
b
3
rn
'ป
'ป
R
sample points
2.32
13.9
.31
-2.63
-.55
.94
.42
2.96
-4.52
2.93
.08
3.04
2.15
-.98
.70
-.18
-.38
-.49
-.41
-.91
-.19
-.28
.61
-.02
-.37
.09
.22
.31
-.11
-.31
-.36
-.19
-.37
.66
.02
-.55
-.23
.09
.29
.04
-.14
-.48
.45
-.30
.34
.47
-.27
-.09
-.27
-.46
-.29
-.15
-.13
-.36
-.44
.99
.85
.57
.55
.35
.50
.47
.65
.65
.68
.50
CO data obtained 14' above the medial strip'
Meteorological data obtained fron Vane 2
Traffic flow rate intervals selected from total traffic flow rate data
Table 3.2-2
-------
NUMERICAL RESULTS OF HUGH J. GRANT CIRCLE CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (Veh/Hr)
1580-1740
1770-1950
1960-2160
2150-2370
2340-2580
2530-2790
2720-3000
2910-3210
3100-3420
# of
Sample
Points
2
7
8
14
14
8
14
11
4
D *
o
bl
b2 -
Too few sample
.21
11.00
7.88
7.54
10.30
8.02
4.87
-5.28
-.44
-.39
1.64
-1.26
.30
.35
Too few
-1.40
-1.93
-1.32
-3.62
-.93
-2.29
-1.27
b-j i
points
3.01
-.27
.32
.33
-.01
.41
.32
:11
-.60
-.16
-.44
.48
-.72
-.36
-.14
r!2
-.68
-.43
-.32
-.45
-.43
-.57
-.40
r!3
.50
-.14
.44
.29
.69
.32
.46
R
.88
.49
.56
.78
.74
.64
.63
sample' points
CO data obtained 16* above the westbound lanes near the outer wall at the mouth of the
tunnel
Meteorological data obtained from Vane 1
Traffic flow rate intervals selected from westbound traffic flow rate data
Table 3.2-3
-------
NUMERICAL RESULTS OF NELSON AND JESSUP AVES CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (Veh/Hr)
4550-5030
4930-5450
5310-5870
5690-6290
6070-6710
6450-7130
6830-7550
7210-7970
7590-8390
7970-8810
# of
Sample
Points
7
14
15
25
16
2828
22
22
14
5
b
o
13.20
1.50
1.89
3.98
5.70
5.37
2.51
-5.15
-2.35
2.36
bl
-1.22
-.36
-.94
-.64
-.73
-1.11
-.42
-.57
-.66
-.46
b2
-1.26
1.08
1.18
.42
.25
1.33
.66
-2.84
2.16
.59
b3
-2.15
-.71
-.01
-.57
-1.06
-.77
-.49
.78
.38
-.19
'11
.22
.12
-.47
-.05
.06
.29
.45
.25
.13
.52
r 12
.06
.51
.001
.05
.17
.42
.54
.56
.44
.64
r!3
-.88
-.54
.12
-.32
-.50
-.46
-.58
-.38
-.22
-.69
R
.96
.68
.65
.49
.69
.63
.63
.73
.65
.89
CO data obtained 26* above the medial strip
Meteorological data obtained from Vane 1.
Traffic flow rate intervals selected from total traffic flow rate data
Table 3.2-4
-------
NUMERICAL RESULTS OF BRUCKNER EXPRESSWAY CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE # of
INTERVAL (Veh/Hr)
1880-2080
2170-2390
2450-2710
2740-3020
3020-3340
3310-3650
3590-3970
3880-4280
Sample b
Points
3
12
22
15
10
16
21
12
bl
b2
/
b3
rll
r!2
r!3
R
Too few sample points
1.26
3.01
1.08
3.29
2.76
1.97
3.94
-1.11
.69
-.16
.69
.14
-.21
.20
1.82
-1.00
-.59
-1.32
-.37
.43
-.67
-.50
-.67
-.20
-.23
-.32
-.33
-.67
-.07
.23
.41
-.05
-.03
.16
.01
.21
.31
.48
-.35
-.08
.29
.01
-.75
-.62
-.34
.36
-.57
-.56
-.64
.86
.65
.65
.51
,59
.57
.73
CO data obtained 14" above the medial strip
Meteorological data obtained from Vane 2
Traffic flow rate intervals selected from total traffic flow rate data
Table 3.2-5
-------
NUMERICAL RESULTS OF BRUCKNER EXPRESSWAY CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (Veh/Hr)
1880-2080
2170-2390
2450-2710
2740-3020
3020-3340
3310-3650
3590-3970
3880-4280
# of
Sample
Points
6
14
24
21
19
19
21
12
bo
.004
2.58
3.31
4.18
33.65
3.42
2.73
4.40
bl
-1.81
-.26
-.58
-1.99
-1.67
-.50
-.67
-2.07
b2
2.36
.05
.31
1.13
1.63
.12
.60
1.71
b3
.36
-.26
-.41
-.25
-.82
-.38
-.23
-.43
rll
-.07
-.28
-.13
-.62
-.09
-.26
.02
-.32
r!2
.18
-.25
-.11
-.46
.03
-.28
.08
-.30
r 13
.26
-.26
-.34
.10
-.41
-.34
-.27
-.08
R
.83
.40
.52
.70
.64
.57
.41
.57
CO data obtained 25' above the ground at the south side of the road
Meteorological data obtained from Vane 1
Traffic flow rate intervals selected from total traffic flow rate data
Table 3.2-6
-------
NUMERICAL RESULTS OF GRAND CENTRAL PARKWAY CO REGRESSION ANALYSIS
rRAFFIC FLOW RATE # of
INTERVAL (Veh/Hr)
3190-3530
3950-4370
4710-5210
5470-6050
6230-6890
6990-7730
7750-8570
8510-9410
9270-10,250
Sample
Points
9
8
8
12
10
11
17
11
7
2
-6
2
1
9
-1
5
6
4
b
o
.78
.56
.87
.93
.90
.60
.20
.23
.92
bl
.03
.78
2.40
1.18
.27
.73
-.08
.71
-.60
b2
-.09
3.04
-2.98
-.89
-3.32
.51
-.55
-2.11
.02
b3
-.90
-.30
-.44
-.56
-.04
.71
-.96
-.63
-.67
rll
-.02
.88
.74
.61
.01
.32
.62
.24
.29
r!2
.43
.88
.47
.47
-.71
.23
.31
-.13
.03
r!3
-.84
-.84
-.73
-.39
.05
.57
-.78
-.49
-.82
R
.84
.93
.84
. .69
.72
.65
.79
.80
.91
CO data obtained 10'above the medial strip
Meteorological data obtained from Vane 2
Traffic flow rate interval selected from total traffic flow rate data
Table 3.2-7
-------
NUMERICAL RESULTS OF GRAND CENTRAL PARKWAY CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE NO. OF SAMPLE
INTERVAL (VEH/HR.) POINTS bo bt b2 b3 rn r12 r!3 R
3190-3530 9 .27 -1.55 3.40 -2.32 -.24 .09 -.35 .68
3950-4370 7 4.68 -4.70 4.08 - .42 -.94 -.82 -.40 .97
4710-5210 8 1.23 -4.46 5.61 - .45 -.74 -.53 -.35 .94
5470-6050 12 6.15 .89 -2.99 .25 -.43 -.51 -.36 .55
6230-6890 10 13.21 -1.33 -2.70 - .67 -.81 -.68 -.34 .84
6990-7730 14 1.91 - .21 - .04 .46 -.10 -.08 -.46 .50
7750-8570 16 .70 -2.83 -3.68 .07 -.36 -.03 -.02 .78
8510-9410 11 4.09 -1.99 1.27 .31 -.64 -.32 -.13 .83
9270-10,250 6 5.65 -1.70 .09 .55 -.56 -.17 .26 .90
CO data obtained 20" above the center of the northbound service road
Meteorological data obtained from Vane I
"Traffic flow rate intervals selected from total traffic flow rate data
TABLE 3.2.8
-------
NUMERICAL RESULTS OF BROOKLYN QUEENS EXPRESSWAY AT PARK & NAVY CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE NO. OF SAMPLE
INTERVAL (VEH/HR) POINTS bf
r!2 r!3 R
3040-3360
3800-4200
-3.56 1.86 -1.87 2.64 .84 .34 .01 .94
Too few sample points
4180-4620
15
3.59 .10 - .01 -1.12 .28 .34 -.56 .56
4560-5040
21
-1.92 - .08 1.97 - .36 .46 .49 -.33 .52
4940-5460
17
-1.35 1.91 - .95 - .20 .64 .52 -.30 .66
5320-5880
22
-1.39 1.03 .28 - .31 .48 .33 -.34 .51
5700-6300
17
1.72 .09 .36 - .43 .22 .18 -.26 .29
6080-6720
10
- .71 -2.25 4.72 - .92 .50 .68 -.84 .96
6460-7140
4.21 .31 - .42 -1.27 .47 .60 -.89 .90
6840-7560
Too few sample points
CO data obtained 10' above the center of the road
Meteorological data obtained from Vane 2
Traffic flow .rate intervals selected from total traffic
flow rate data
TABLE 3.2.9
-------
NUMERICAL RESULTS OF BROOKLYN QUEENS EXPRESSWAY AT PARK AND NAVY CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (Veh/Hr)
3040-3360
3800-4200
4180-4620
4560-5040
4940-5460
5320-5880
5700-6300
6080-6720
6460-7140
6840-7560
# of
Sample
Points
0
2
14
27
21
21
15
9
6
2
b
o
Too
Too
.96
-1.26
.33
1.45
.12
1.39
3.31
bl b
few sample
few sample
-.05
.73
1.33
.01
-.54 1.
1.33 -1.
-.60
2 b3 rll
points
points
74 -.33 .83
93 -.39 .73
57 -.27 .68
54 -.48 .54
72 -.32 .67
34 .002 .46
45 -.58 .07
r!2
.79
.72
.55
.46
.70
.06
-.23
r 13
-.71
-.26
-.17
-.73
-.60
-.71
-.84
R
.88
.81
.76
.77
.79
.81
.87
Too few sample points
CO data obtained 20" above the west side of the road
Meteorological data obtained from Vane 1
Traffic flow rate intervals selected from total traffic flow rate data
Table 3.2-10
-------
NUMERICAL RESULTS OF CANAL STREET CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE NO; OF SAMPLE
INTERVAL (VEH/HR) POINTS b
r!2 r!3 R
890-990
990-1090
1080-1200
1180-1300
1270-1410
1370-1510
1460-1620
1560-1720
1650-1830
20
16
10
_o _i _2 _1 _1
.42 .05 1.22 -.32 .11 .53 .21 .59
-.94 -.42 2.16 .19 .05 .83 .73 .92
2.47 -.11 -.46 .78 .04 -.13 .46 .53
2.02 .09 -.10 .33 .19 -.27 .32 .39
1.15 .86 -.16 -.30 .36 -.24 .11 .50
-1.74 .99 1.16 -.75 .85 .58 .72 .92
Too few sample points
Too few sample points
Too few sample points
CO data obtained 20' above the westbound lanes
Meteorological data obtained from Vane 2
Traffic flow rate intervals selected from total traffic flow rate data.
TABLE 3.2-11
-------
NUMERICAL RESULTS OF CANAL STREET CO REGRESSION ANALYSIS
TRAFIC FLOW RATE NO. OF SAMPLE
INTERVAL (VEH/HR) POINTS
b'2
r12
r13 R
890-990
990-1090
1080-1200
1180-1300
1270-1410
1370-1510
1460-1620
1560-1720
1650-1830
4.98 -1.00 .18 - .53 -.79 -.57 -.43 .97
18
28
29
22
15
1.88 2.97 -4.28 1.65 .59
1.07 - .14 .85 - .42 .12
3.50
1.71
.58 -1.61
.83 -1.13
1.81 - .03 - .18
.26 .16
.43
.25 -.04 -.08
.07
.19 .69
,10 -.09 .28
-.01 .15
.07 .23
.26 -.03 -.005 .10 .15
1.45 - .06 .29 - .43 -.02 -.09 -.23 .24
-1.36 2.20 -.31 -1.11 .43 -.16 -.53 .65
3.05 1.35 -3.90 1.88 .18
,13
.27 .99
CO data obtained 50' above the ground, 16" morth of the road
Meteorological data obtained from Vane 1
Traffic Flow Rate intervals selected from total traffic flow rate data
TABLE 3.2-12
-------
In order to determine the direct effects of the various wind
parameters on the CO concentrations, the influence of the traffic flow
rates on the concentrations had to be eliminated. This was effectively
achieved by choosing a traffic flow rate interval "small" enough so that
the values of the interval for our purpose could be considered constant
and then using only those data points for the wind parameters and CO
concentrations in the analysis that corresponded to the traffic flow rates
in the chosen interval. Sirice, in general, the traffic flow rates for
each site varied considerably, several traffic flow rate intervals were
chosen and a separate regression and correlation analysis was performed
on the data that corresponded to each interval.
The traffic flow rate intervals were selected for each site to
provide approximately ten intervals spanning the range of total traffic.
The analysis was performed by checking all the reliable traffic flow rate
values to see if they fell in the traffic flow rate interval chosen for
the particular run. Whenever a value was in the interval, the day and the
hour of that value was noted. Then the mean CO concentration, the mean
sigma azimuth angle, the mean sigma elevation angle, and the mean wind
speed for that particular day and hour were checked to see if any of the
values were missing. If one or more were missing, the computer would dis-
regard all those values for the day and hour and check the values of the
CO and meteorological data for the day and hour that corresponded to the
next traffic flow rate that fell in the interval. If none of the values
were missing, the logarithm of each value was computed and stored as the
first observation in the regression array. This process was continued
until the CO concentration and wind parameters for the day and hour that
corresponded to the last traffic flow rate that fell into the interval
were checked and either discarded or else the logarithm of each value was
computed and the four values stored as the last observation in the regres-
sion array.
If the number of observations in a regression array was four or
less, no regression or correlation analysis was performed since the results
would provide little or no information regarding how well the proposed
regression model describes the variation in the data. This is because the
regression curve would pass through each data point and the multiple cor-
relation coefficient would always be unity.
By examination of the data in Tables 3.2-1 to 3.2-12 it can be
seen that the simple correlation coefficients, r^^, r^, r!3 an^ the
multiple correlation coefficient, R, show large variations. It must be
concluded, therefore, that the postulated regression equation does not ade-
quately describe the effects of the wind parameters on the carbon monoxide
concentrations.
Another regression model was formulated to see if a better fit of
the data could be obtained. The new postulated model was
t/EL V )K1
-------
where C = mean hourly CO concentration
OAZ = mean hourly standard deviation of the
wind azimuth angle
= mean hourly standard deviation of the
wind elevation angle
V = wind speed
Kp KI = constants to be estimated
The linearized form of the model is
In C = In KQ + Kj^ In ( 0^z (JEL V)
where In is the natural logarithm function
Letting Y = In C
Xl = ln
B0 - In
and substituting in the above equation yields
Y = B0 + BiXl
which is the form used in the computer program
Table 3.2-13 shows the numerical results of the revised regression
equation, using data gathered for the Brooklyn Queens Expressway at Hicks
"Street site. It can be seen that the correlation coefficient, TH between
the CO concentration and the product of the sigma azimuth, sigma elevation,
and wind speed, varies between -.77 and .95. The multiple correlation
coefficient, R, varies between .05 and .95. Thus the new postulated model
does not explain the data variation as well as the original postulated
model.
3.2.2 Hydrocarbon
The hydrocarbon study was aimed at determining the effect of the
change in traffic flow rate on the hydrocarbon concentrations.
3.2.2.1 Regression Equation
The relationship between the hydrocarbon concentration and the
traffic flow rate was assumed to be linear so that the relationship could be
written in the form
H = B0 + B! N
-------
NUMERICAL RESULTS OF BROOKLYN QUEENS EXPRESSWAY AT HICKS STREET CO REGRESSION ANALYSIS
TRAFFIC FLOW RATE
INTERVAL (VEH/HR)
NO. OF SAMPLE
POINTS
2840-3140
3410-3770
3 700-4 0180
3980-4400
i
1
4270-47lo
4550-5030
4840-5340
5120-5660
5410-5970
5690-6290
5980-6600
6260-6920
3
5
5
5
10
10
12
15
15
8
5
6
.74
31.10
46.35
13.19
20.91
23.14
23.30
13.60
7.18
11.53
23.60
19.59
3.2 X 104
-8.27 X 102
-1.80 X 104
1.20 X 104
1.00 X 103
- .07 .
2.10 X 103
1,32 X 104
1.97 X 104
1.84 X 104
-3.18 X 103
1.41 X 103
.95
-.77
-.46
.68
.05
-.05
.06
.34
.56
.55
-.22
.34
.95
.77
.46
.68
.05
.05
.06
.34
.56
.55
.22
.34
CO data obtained 14" above the medial strip
Meteorological data obtained from Vane I
Traffic flow rate interval selected from total traffic
flow rate data.
TABLE 3.2-13
-------
where H = mean hourly hydrocarbon concentration
N = mean traffic flow rate
BO, BI = unknown constants to be estimated
Since the equation is linear, it .was used directly in the computer
program for the analysis. Therefore, in the program, p, the number of in-
dependent variables, was equal to one and that variable was the traffic
flow rate. The hydrocarbon concentration was designated as the dependent
variable in this analysis- Bo and B^ were the constants in the postulated
regression model and their least squares estimates were designed bo and b,
respectively. The simple correlation coefficient, r, between the hydrocar-
bon concentrations and the traffic flow rates was, in itself, a measure of
the linear relationship between the variables for each site.
3.2.2.2 Results of the Hydrocarbon Regression and Correlation Analysis
The hydrocarbon concentration and traffic flow rate data for all
twenty-four hours' of each day was used for seven of the nine sites
analyzed. Since the only reliable traffic flow rate data at B.Q.E. at
Hicks Street was for the hours of 0800 through 1500 and for the hours of
0700 through 1700 at B.Q.E. at Park and Navy Streets, the analysis for
these sites was done using the hydrocarbon concentration and traffic flow
rate data for those hours only.
As can be seen from Table 3.2-14 which summarizes the numerical
results of the analysis, the magnitudes of the correlation coefficients
between the hydrocarbon concentrations and the traffic flow rates for
eight of the nine sites analyzed are very, very small. Since the magnitude
of the correlation coefficient provides a measure of the usefulness in the
regression model of the term containing Bj_, it can be concluded that a model
of the form
H = B0
Where H = hydrocarbon concentration
B = constant
would describe the data variation for eight sites as well as the postulated
model. In fact, the model
H = H
where H is the mean value of the hydrocarbon measurements, seems to provide as
good of fit of the data at B.Q.E. at Hicks Street as the regression model.
Figure 3.2-1 illustrates this. It can also be seen from Figure 3.2~1 that
the relationship does not appear to be linear. Hence the correlation coef-
ficient is only .19.
The coorelation coefficient between the hydrocarbon concentrations
and the traffic flow rates .for the George Washington Bridge is .65.
-------
NUMERICAL RESULTS OF HYDROCARBON REGRESSION ANALYSIS
SITE
1. 54th 6t Sutton Place
2. Brooklyn Battery Tunnel
3. B.Q.E. at Hicks St.
4. H.J. Grant Circle
5, Nelson & Jessup Aves
6. Bruckner Expy.
7. Grand Central Pky.
8. B.Q.E. at Park & Navy
9. Canal Street
10. Geo. Washington Bridge
Legend
bo Least squares estimate of Bo in the regression model
i
bi Least squares estimate of Z\ in the regression model
r Correlation coefficient between hydrocarbon concentrations and traffic flow rates
TABLE 3.2-14
# OF SAMPLE
POINTS
e 111
"unnel
:. 50
192
es. 210
226
258
lavy 137
160
idge 323
bo
6.97
No Analysis
3.05
5.68
3.36
3.30
10.01
8.10
8.51
8.70
bl
.00005
Performed
.00040
.00026
.00007
-.00005
.00030
.00058
.00284
.00063
r
.05
.19
.21
.23
-.13
.30
.29
.26
.65
-------
TRAFFIC FLOW RATE (VEH/HR)
Figure 3.2-1 - Hydrocarbon Concentration vs. Traffic Flow
Rate for Brooklyn Queens Expressway at
Hicks Street
-------
This value is also too small to conclude that there is a linear relation-
ship exhibited between the variables. Hence, there does not appear to be
a linear relationship between the hydrocarbon concentrations and the
traffic flow rates for the nine sites analyzed and for this reason the
postulated regression model does not adequately describe the data varia-
tion of the two variables for these sites.
-------
3.3 Model Development
Exponential decay of carbon monoxide concentration with increasing height
has been previously reported'1' and has been theoretically justified by Sontowski
and Dworetzky^ '. These workers have found that the following relation may be
applied to define the concentrations as a function of vertical distance:
C =
aW (D + V)
where:
M
Mass
Vehicle mile
N = Traffic flow rate (vehicles/hr.) v
a = Constant found to be .015 - 025 ft."1
W = Width of cut
D = Eddy Diffusion Coefficient
V = Wind velocity
z = Height above vehicle exhaust plane
C = Concentration
The applicability of the exponential decay considering dimensions other
than the vertical dimension as suggested in the previous two references was
attempted for the ten roadway configurations. The modification of Equation
3.3-1 utilizes the radial distance, (3, in place of the vertical distance z.
The mathematical equation describing the decrease in concentration is of the
form:
C = COX e-a 3.3-2
where :
rn = Q o >
Cฐx aW (D + V) 3<3-3
a
\ = Radial distance between source and receptor >
Georgii, H.W. , Busch, E., Weber, E., "Investigation of the Time and Space
Distribution of Carbon Monoxide Emission Concentrations in Frankfurt AM
Main," Report of the Institute for Meteorology and Geophysics of Frankfurt
AM Main University, No. 11.
(2)
Sontowski, J.F. and Dworetzky, L.H., "Transport of Traffic Generated Carbon
Monoxide in Long City Streets," GE TIS 70SD236 (1970)
-------
3.3.1 Superposition Method
It is assumed that any given roadway can be divided into independent
line sources as represented in Figure 3.3-1 and each probe located at the
roadway measures the concentration from its immediate line source plus the
concentration emanating from the other (x-1) line sources. Further, it is
assumed that the decrease in concentration from each probe location follows
Equation 3.3-2. As a first approximation the effects of wind are implicitly
stated in the coefficient, COX, as shown in Equation 3.3-3. This is somewhat
justified since all of the computation will be made on concentration data
that has been averaged over a minimum of 14 days, and provided COX is defined
close to the roadway where vehicular induced turbulence predominates as
opposr.d to wind induced turbulence.
COX is defined only for those probes located at the exhaust plane level
which measure the highest concentration. These were located 3' off the roadway
for all sites excepting the Viaduct site. Therefore, the concentration measured
at any 3' probe when the cross roadway diffusion has been properly accounted for
may be expressed by a set of linear equations which expressed in matrix algebra
has the following form:
[Cx] = [ exp (-a pxj)] [COX] 3.3-4
where :
Cx = Concentration measured by a probe located 3' off the xth line source
= Distance between center of xth line and jth line source
CO = Concentration that would be measured at 3' if xth line source
were unperturbed by concentrations diffusing over the other
(x-1) source
a = Same constant as defined in Equation 3.3-1.
The value, COX, is defined as the 'unperturbed concentration resulting from the
xth line source since the diffusion from the other (x-1) line sources has been
now accounted for in Equation 3.3-4.
For the open cut or grade road sites the coefficient matrix is a 3 x 3
matrix since there were only 3 probes which measured the concentration of 3
line sources. It is quite impractical to measure the concentration of the
middle line sources. When the model is applied to covered sites like Grand
Central Parkway or Sutton Place, the method of images was used. This method
considers the covering as a reflector, which means that there exist virtual
sources on the other side of the reflector which must be considered. A
figure depicting the application of this method to the Sutton Place site is
presented in figure 3.3-2. Here the two additional sources are x' and x'
-------
ฃT8aปซ 3-3-1
gonce
Tjoe 2oni.ce y fctope
-------
X
O
to
\
/
/
r-?
Receptor, R
Figure 3.3-2 Method of Images for FDR Drive
-------
which are mirror images of x. and x2, respectively. For this particular site
the coefficient matrix becomes a 5 x 5 matrix. It was also found, for covered
sites, that some adjustments must be made in the decay constant, a, to provide
an adequate representation of the concentrations measured.
The model was further generalized by defining the Vehicular Pollution
factor, (b , by dividing the unperturbed concentration, COX, by the traffic
flow rate in the immediate vicinity of the probe location.
The vehicular pollution factor, C ,
3.3-5
For eight of the configurations evaluated was found that . shows a
strong linear relation with traffic speed (correlation of .81) as shown by
figure 3.3-3. The viaduct site exhibited pronounced meteorological effects
which resulted in the two lower points to the extreme left in the figure.
The low velocity points in the lower right hand is data collected in a city
street. Data from the tunnel sites was not included in figure 3.3-3.
The linear relation between / and traffic speed was found to be:
(fฑ = [ -0.51 (Tฑ) + 26.9] x 10"3 3.3-6
for T: 15 mph.^ T H49 mph
= Apparent emission factor (PPM-Hr)
veh
T^ = Average Traffic Speed in lane i
It is seen from Equation 3.3-6 that traffic speed is the predominant factor
and that the ranking of the sites must be based upon traffic speed, as well
as site configuration. Stated in other terms, the line source strength
decreases with increasing automobile speed due to the measure in vehicular
induced turbulence and engine efficiency.
Using Equation 3.3-6 and the total traffic volume it now is possible
to compute the concentration in and around a given expressway by super imposing
the various line sources as given below:
_
\
tCO]R- 1 i . 3.3-7
i - 1
where
[CO] = Carbon Monoxide concentration at receptor, R
K.
0 = Vehicular pollution factor for line source, i, from
1 Equation 3.3-6 (PPM-Hr)
veh
-------
o
J>
Q
W
W
&
w
U
ii
1=4
60
55
50
45
40
35
30
25
20
15
10
O
RELATION BETWEEN
VEHICULAR POLLUTION FACTORS TAKEN AT EXHAUST PLANE
TRAFFIC SPEED
o
o L
8 10 12 14 16 18 20 22 24 26
PPM - HR
VEHICULAR POLLUTION FACTOR, <$,
VEH
X 10
-3
-------
JL = Traffic flow rate from roadway source, i,
a = Empirical constant derived from New York City data and
reference 1. (.02 ft"1 for open sites and .045 ft"1
for covered sites) :
= Radial distance between roadway source, i., and receptor, R
i = Roadway source which may be defined by direction, i.e., one
roadway may be considered as being two sources moving in
opposite directions.
S = Total number of sources
It is felt that the particular reason why the high degree of linear correlation
between traffic speed and (h exists is due directly to the application of
Equation 3.3-4, where the cross roadway diffusion is accounted for in a gross
sense. Consequently, the concentration measured can be directly related to the
traffic volumes and speed in the immediate vicinity of the measurement. The
Vehicular Pollution factor, (n ^, may now be used as the source strength, Q,
in area models to project the concentrations to longer distances than those
evaluated in this study. One of the particular weaknesses of Equation 3.3-7
is when the distance, PiR, becomes large and the concentrations approach the
background. Certain model modifications are possible but were not attempted
in this study.
3.3.2 Model Validation
The model was utilized to define the concentration isopleths over the
test roadway by programming Equation 3.3-7 on a desk side computer and
calculating the concentration in discrete grid points over the roadway.
Although not shown in the following figures, the model was used to project
the.concentration to the zone of the nearest receptor. Figures 3.3-4 through
3.3-16 present the comparison between the measured and predicted concentration
at each site for selected hours during the course of the day. The hours chosen
represent those periods of peak traffic volume.
The model has the capability o,f distinguishing the influence of varying
traffic speed and volume in each direction as shown in Figure 3.3-7, where the
major flow during the periods between 0800 - 0900 hours is Manhattan or Queens
bound and, between 1600 - 1700 hours, is homeward bound toward Brooklyn. In
figure 3.3-4 the ventilation characteristics of the covered on top - open on
one side site is demonstrated. The contours for this covered on top - open at
side site correctly show the concentration gradient vector pointing in the
direction of the open side. The effect of a cover over the roadway is shown
in figures 3.3-8 and 3.3-9 for the unventilated tunnel where the anticipated
high CO concentration appear along the walls and ceiling.
Model validation was not as good for the two sites which possessed the
highest ventilation namely, the Grade road and the Viaduct sites. It is seen
-------
T - , - . .,. - . -
51j 4o 45 42 vJSN=vW''i :>y Vd^ *7 y
I 1 I \ \ \
r55 57 5o 5rf 59 59 59 5K 5t> 58 57 55 53 51149/16-43 40 37 35 3?. 3ฃi 25 'ir
56 L>r
:>6 5b 59 59 59 59 59 59 59 56 57 56
59 5o
J *J \J ..I A 1 *-t * ' * V_ป ' -ง *^ -I V^ V' f vj ^/ ^ (.< ' * '- '- '
=>\54 53 Vi9 46\/.'3 41\ 3o 36 33 31\ซ9 H'
> 5oV53 5^) 47*44 41 \39 3^\:^t 3-.> :-ป i v
61 61 60 61 6a 63 63
68 60 65 63 61 6M 6w 61 62
I
65 6ซ 60 59 59 60 61 63 66 7fe>
76\ (73
5 42 39 37 ^5 33 3<> ^
\ \ \ (29
a 63 63 63^i;J 56\ 52 Vj 45 42 4H 37 H6\S.:! 33 M
2- 64 66 66 63\ 58 |53 [*9 4:5 42 ;V9 37 36 3r5
feU$'3/ 57/5?/48 /i4 4 1/C
39 37 36 35 3!
Southbound lanes
0700 - 0800 hours
Northbound lanes
(39) (28)
inzn^tn!;;;;
3 43 /.'2 4\39 37 35 32 33 27 25 2" 1\ 1 O IV
ซ/* ซ ซ\, 33 ^33.1, ,7 J P-,P-\,o ,,
:3 ''.'3 43 44 44 *i'-> ** ,J *-* i o
5 46 /.?o 51-X161 41J3
63)
3,3 33- P.I
23 21 ft"-; I-- i/
, .
33 :^:.r) S 7/^.4 ซ'-! 21/19 \ >1 U
Southbound lanes
1600 - 1700 hours
Calculated Values - Decay Constant = -.045 - Grid Size = 3 ft.
Measured values included in 0
Northbound lanes
- Contour interval=5PPM
Figure 3.3-4 Measured & Calculated Mean Carbon Monoxide Contours -
F.D.R. Drive at Sutton Place Covered on top - open at side
-------
23 24 24 25 26126 27 27 28 29129 30JL30 31
III I I
23 24 24 25 26126 27 27 28 29129 3Q
23 24 24 25 26 126 27, 27. 28 29 29 30
23
I
24 25 26 26 27\27 28 28 29 30
30 31
30 3l
32 32
/
31 32 32
32 33 33 33 33 33
32 33 33 <3333\33
32
?
\ \
23 24 24 25 26 26 27
28 28 29 30 30 31
v \ r
. _. __ _ _ _. 28\28 29 30130 31
i\24 25\25 26 27 27 28 28 29\29 30 31
\ii> co\iio a7. a7 ฃ3 kiy 29 30 30
33 33 33 33 33 33
33 33 33 33 33 33
33/33 33 33 33 33'
3/
I
2ซ28 29
27 28 28 29 129 30
,23 24 24 25 25 26 26 27
\ \ \ \ \
23 24 24\ 24 25X25 26 26 2'. __ __ .
\ \ \ \ \ \ .
23 23 24 24 25 25 26 26 27 27 2b\ 28 29 3
* * _ * " \ *
12
3\23 24 2\24 25\ฃ
26\26 27> 27
\
9\2
33 33 33 33 33 33
33 33 34 J}4,34 33
33 33 34 34 34 34
33\33 34 34 34 34
\ I *
33 33 34 35""35L 3.4
\ \ ( \\
32 33 34 35 36 J34
\ ป ป i /^^\
6e)
33
North Wall
South Will
0700 - 0800 hours
Calculated Values - Decay Constant = -.045 - Grid Size = 1 foot - Contour Interval = 1 PPM
Measured Values included in 0
Figure 3.3-5 Measured and Calculated Mean Carbon Monoxide Contours Brooklyn Battery Tunnel,
-------
o
00
70 71 7,1 72 72 12 72 71
^ I I '
^j 70 71 71 72 72 72 72 71
70 71 7
1/71 72/72 72/72 71
70 71 71 72 12 12 7.2 72 71
70 71 71
71'71
71 71 .71 71^70 70/70 69 69 68
/ (65) / /
71 71 71 7r"70 70 70 69 69 68
'0/0
72 72 72 71 ,71 71
71.71 71 70 70 /JO 70 69/59 68
71 71 71 70 70 70 69 69/69 68
71 71 70 70 70 70 69 69 69 68
> V /U:
oo
72*72 71
72y/2 72 72' 72 7.1/71 71 70
72 73 72 72 72/71 70/70 70
/^ / / / -
73 73 72/72 71 70/70 69^69
J / C* m id f A I \J M I
74 73 72 71/70 7,0 69" 68 68
71 70 7o~ 7o 7u ov
70,70 70 69 69 69,69 69 68 68
70 69 69 ฃ9 *69 68 68 68 68 6i<
68 6868^68-^8 68 68
68 68^68^68 676^67^67-167*^57
68 67 67"*67 6(S 66 66 66 66 66
South Wall
Calculated Values -
North Wall
0700 - 0800 hours
Decay Constant = -.045 - Grid Size =ป 1 foot - Contour Interval = 1 PPM
Measured Values Included in 0
Figure 3.3-6 Measured and Calculated Mean Carbon Monoxide Contours Brooklyn Battery Tunnel,
-------
"'(19)
"ffiV
33 32^ป3ฃ 31 30^ 29 26^ 27 2~6_ 25 24vฃ4 23 23
^
3636^35 3*5x33
313^38 37 35X34 32 31 30 29 2ฃ 27 '26 25 25 24 ซ*4s24 23 23
\ \ C
_41 4"'3 38 36 35\33.\J2 3^29 tt\>7 27^26 25 25 25 24 24^24^
'15 '44 ^2 40 \57 \35 3*4 3^ 31 30 29 2*8 27 26 86 25 25 25 25 25 25
ft & %i
27 26\26 25 25 25 25x2J^-e8 29 29 29 29
29 30^30-r30"-30"-3"030 30 30 3
31 3g 32 33^,34 35 J36^36
8 39 39 39
0-OCT31 32 32 33^34 35.:
I 31 32 33 3435 36 37^
31 31 31 31 32JJ2 33 34 35 36 3
2*32 73 34 35 36 37
35 35 34.33 33 32 32 32 32 33 33 3,4 35 3.6 38 39/41/43/46' 43' 49
\ I I I I f f I f S
5 34 33 33 32 32 32 33 33 3,4 35 36 38 40 42 44 4.6(49/52
3^ "$ -11 g?
33 33 33 32 3232-<32-3J
Northbound Lanes
Southbound Lanes
1600 - 1700 hours
Calculated Values - Decay Constant = -.02 - Grid Size = 4 feet - Contour
Interval = 2 PPM Measured Values enclosed in 0
Figure 3.3-7 Measured & Calculated Carbon Monoxide Contours
Brooklyn Queens Expressway at Hicks Street - Shallow Cut
-------
2) ฉ \
n 11 11 11 11 11 11 11 11 11 11 rf 11 11 M 12 i2\
12 12 13 13 13
11 11 11 II 11 11 II
11 11 12 12 12 12 13 13/13 i:
1} 11 11 11 11 11 ll'll 11 11 11 11 11\J1 11 11 12 12 12 13 13 13
\ / U2J
12 12 12 13 13 13
X \
11 11 11 11 11 if 11 11 11 11 11 11 11 1111 11
1212-12 11 11 11 11 11 10 10 10 10 11 11 11 11 12 12118 13 13 13 13
2 \yi-
15 12 11 11 11
S)
12 12 12
1 11 10 10 10 10 10 10^,10 11 11\11 11 12\12 13 13
10 loXo 10 10 10 10 10 10 11 11 11 12 12
iy
v r
12 13 14 1
South Wall
0700 - 0800 hours
North Wall
9 9
99
9999
999
10 10 10\10 9
9999
9 9 10 10/10 10
9 9 10 lOf 10 10
11 11/11 11 O
11 11 11 11 11
9 9 9 10 101 10 10 11 11 11 11
. . t I do.
999.9 99999999 10\10 10 11 11 11 11
9999
9999
9 98 8 8
V
(
9 9 10 10 10 11 11 11 1S-12
999 10\10 10 11 11 12 12J
\ \ '
10 i\\ii 12X12!
U3J
South Wall
North Wall
1200 - 1300 hours
Calculated Values - Decay Constant = -.045 - Grid Size = 2 feet - Contour Interval
1 PPM - Mean Measured Values Included in 0
Figure 3.3-8 Measured & Calculated Carbon Monoxide Contours
Cross Bronx Expressway at Grant Circle, Eas.tbound Tunnel,
Short Tunnel
-------
24 24 24 23 23 23 23 ?3 23123 23 24 24 24 25 25 25 25
24 25 25 24 24\24 23 23 23 23 23 23 23 23 23 24 24 24 25 25 25 25 25]
2SX2525 25 24124 23 23 23 23 22 22 23 23^3 23 24 24 25 25 26"2i
25 25 25 24/24 23 23 23 22 22 22 22 23 23 23 24\24 25, 25/26 26
I I ' / x ^N \ * 1 /""*>>
26 26. 25 25 24 24 23/23 22 22 22 22 22 22 23 23 23 24 251 25126 f27 ?
) / / 7 s \ \ \ \ \ \
24 24
26 26 25
J6 2
325
5 24
i / ' \ \
24 23/23 22 22 22 22 22 22 22 22 23v2
>/22 21 21 21 21 21 21 22> 22 2
23 23 22
24\25\2
,25 26 27 27
3 23 24
!4 25 26 27
North Wall
South Wall
1600 - 1700 hours
nf 1 3 1 3 13
13 13
L3 13
V-3 13
13-13
14-^4
B4/4
ED
13 13
\
13 13
13 13
13 13
13 13
13 12
12 12/12
12 12*
12 12
12 12
12/12
J
12* 12
12 12
12
12
12
12
11
1 1
1211
1211
11 11
11 11
11 \lj
/
1 1.11
11 11
11 11
11 11
11 11
11/11
n i 1 1
u 10
10 10y
'
W
TT
1 1
ซ
,1 1
11
10
u 11
11 11
^-
u 11
u 11
10 10
x^
'10 10
10 10
1 1
1 1
-x
1 1
1 1
10
^N
10
10
11 11 11 11 11
11 11 11 11 11
1 1NJ1 11 11 11
11 1111 11 11
11 11 1 1\ 1 11
\
10 10 11 1 U 1 1
10\0 10 1 1 \ 1
1
11 11
11 11
11 11
11 11
12 J 2-
(
12 M2
1 1 12
TT
1 1
-------
(f) ^ <ฎ ฃL_
9 9 9^-10"TO 10 10 10 10 10 10 10 10^ 9 9 9 9 8 8~
8 8
,10' 10 10 11 11 11 11 11 11 11 11 11 11 10 1010^ 9 9 9
' "
11 11 12 12,~12*~f2 12 12 12 12 12 12 12^12 11 11 10^10 10
12^12^13 13 13 13 13_L3 14-U3_13 13 13 13 12s 12 11 11 1
13 14 14-tl4'*"T515 15 15 15 15 15 15
^^ ^____ _>_
14 LS. 1
-------
7 T 7 7^N7 7 6 : 6\ 6 S^^S 5 \ 5 . 5 5 4 4 44
8-V8 7 7.7!'6 6ป6 5 5 5\5 5 5 5 5 4
8X8 7 "7 ' 6 6\6 5 5 55 5 5 5 5
87\J766\6555\5555
1098\8 7\7 66 \6 5555555
Southbound Lanes
Northbound Lanes
0700 - 0800 hours
Southbound Lanes
1200 - 1300 hours
Northbound Lanes
Calculated Values - Decay Constant = -.02 - Grid Size = 5 feet - Contour Interval = 1 PPM - Mean Measured Values
enclosed in 0
-------
"15x15 14**14ซ.14
^
M6 15 15-1515-15
N * ^
7*17.16 IfL 16 16-166
4 13 133 12 12
>48 1
"^5,14 14,13 13 13 .^ -,
i IsSs 14*14 13 13 O3 13 lb
14 J4 13
17 17 17 17 17 17-17 -. .. -
\ / \
19 18 18 18 17.17.17 18^.1818^17 17 16 15 15 14 14 14^14 14 13
18
17 17 18
18
14 14 14 14 14
Eastbound Lanes
Westbound Lanes
1700 - 1800 hours
8888 8JL.8^ 8888
88888
888888888
8 8 8^8, 8
88888
88 _ 8 88 8
9888888
Eastbound Lanes
Westbound Lanes
2000 - 2100 hours
Calculated Values - Decay Constant == -.02 - Grid Size = 5 feet - Contour
Interval = 1 PPM Mean Measured Values Enclosed in 0
Figure 3.3-12 Measured & Calculated Mean Carbon Monoxide Contours
Grand Central Parkway at Parsons Boulevard - Cantilever
-------
0 0 O 0
0000
Eastbound Westbound
0700 - 0800 hours
0 0
0
0
0(7
0
0
0
3 5
3 4 I 45T
Eastbound Westbound
2000 - 2100 hours
Calculated Values - Decay Constant = -.02 - Grid Size = 20 feet - Contour Interval =
2 PPM Mean Measured Values Enclosed in 0
Fteure 3.3-13 Measured & Calculated Mean Carbon Monoxide Contours
Brooklyn Queens Expressway at Park & Navy Streets - Viaduct
-------
10 11 12 13 14
6 66 __6_ 6 --6. 6 6-^ 6. 66555
66666666 6^b>*'ป6>6 6 5
7777777777766
77788888777776
888888 ^8^*8^8 877
89999999998 3,^3 7
9 9 10 UL-10-.lO-.lO-JO.JlO 9998
lO-HO 10 10 11 11 11 10 10 Tb\JO 999
11 11 11 11 11 11 11 11 11 11 11
11 12 12-12 12 12 12 12*-12^12 11111
12 13' 13 13 13 13 13 13 13 13 12 J 2 11 1
%14 13
15 1514 To \-2 IP
12 13 14 15/
13 13 \*
^\J f12
IbSlti T5 15 15 15 15 . _ .. .. _ . . . ... .
15 15 is isis^is-'in*- rr~T6 15 14 13 12
16*^6 16 16 16 16 16-17 17 16 15\13 12
\
1 1
i 1
16
1C- 16 16 16 16 17 17
16- 15 14 1ฃ
Westbound
Calculated Values
Interval = 2 PPM
0700 - 0800 hours
- Decay Constant = -.02 - Grid Size
Mean Measured Values Enclosed in 0
Eastbound
= 5 feet - Contour
Figure 3.3-14 Measured & Calculated Mean Carbon Monoxide Contours
Canal Street between Church & Mercer Streets - City Street
-------
10 10 11 11 11
8 3899988 ซ8^8 8877
9 ' 9 9 9 9 9 9 9 9 9 9 ~S%^3
1010 10 10 10 10 10 10"*10ป40 999
111111111111111111101010 9 9 9
11 11- 11 12
12_12-12 12 12 12 12^12ซxl2 11 11 11 10 10 9
13 13 13 13 13 13 13 13 12 12s!2 11 11 11 10
JJ 12x12 13 13
fS 13 13 !ฃ, 1
13 14-14 15 15
141 14-14~~14"""l4"-14'--i4
14 13 13 12X12 11 11
15 1-5 15 15 15 15 15 15 15 14*! 4 13 13 12 11
*^ 5
1A 15 15 16 17
15 16 16 17 18^
16 17 18 19
16-16 16 17 17 17 16 16 16.15 15 U>14 1 3 J.
17 17 1^.18**18 18^18,^,17 17 17 16.15 15
13 19 19 19 19 19 19.19 18
20 20 20 21 21 ฃ0 acTl^EO 19 18. 17 1
8 17 1 ? 6 1 5 : -
21 22*^2 22 22 22 ฃ2^ 21 20^19 lฃv!7 If
^ 07
23 23 >3 23 23 23 23 2~? '2>^Z ฃ l^tf 13 17 ff
\ \
24 25 25 ฃ4 ฃ4 24 ฃ^"*<24>IP4 23 '22 21\19 I*. If
X \ \ \
25 25 25 25 25 ฃ5 24 ฃ3*21 ฃ0 I)-' J^
ฃ. f~, /i * O O O
25 S5 25 25 25 25
Westbound
Calculated Values
Contour Interval
Eastbound
1600 - 1700 hours
- Decay Constant = -.02 - Grid Size = 5 feet -
2 PPM Mean Measured Values Enclosed in 0
Figure 3.3-15 Measured & Calculated Mean Carbon Monoxide Contours
Canal Street between Church and Mercer Streets - City Street
-------
>2L 28 2_3 28
3
34 34-35'-3ฃ~35 35^35
44 42 41 42 ^39
34 335 38
Wes cbound
Eastbound
0700 - 0800 hours
11 12 12 13 13 13 13 13 12 12 12 12 lilt 11 1
16 17 17 17 17 16 16 1 6
5.J4 14 14 13 13
18 18 18 13 1 7J
23 22 22 23 22
29 2929^33
32 31 31^29 28 23 29 27 26 26 23 29
34 jj/-j_7 V3i i^g /^.^ Tซ; 29 g7
We stbound
1200 - 1300 hours
Calculated Values - Decay Constant = -.032 - Grid Size = 10 feet -
Contour Interval = 5 PPM Mean Measured Values Enclosed in 0
Figure 3.3-16 Measured & Calculated Mean Carbon Monoxide Contours
Trans-Manhattan Expressway at George Washington Plaza
Intermittent Span
-------
for these sites (figures 3.3-11 and 3.3-13) that the levels measured are very
near background, which is not accounted for in the model. Also, the measurement
accuracy is questionable at these low levels due to relative humidity effects
and calibration accuracy. To measure these small gradients instrumentation
modifications are necessary. (Ref. to Intertech Corp. Operators Manual CG-G57
3e pg. 13, 1960).
Similarly, model validation was not as good for the long tunnel
configuration due in part to the small differences in the concentration gradient
and the forced ventilation of the tunnel. Generally, the gradient vector points
to the walls, which suggests the major air flow in the tunnel is in the direction
of vehicular travel. During the course of probe installation for both tunnels,
this was found to be the case. No meteorological gear was installed in the roof
of the tunnel to verify general flow in the tunnel. The so called "piston"
effect for tunnels suggests that the assumption of a constant line source is not
a valid assumption. Data collected for the long tunnel shows pronounced increase
in concentration for those periods of time of counter current traffic. However,
no data was obtained to prove the existence of a non-uniform line source. If
the line source is non-uniform, model modification is necessary as perhaps an
another model should be developed for the covered forced ventilated configuration.
For all of the open sites, the model was found to be quite general since
a constant decay constant of .02 ft"-*- was used. Further, the model was found
not to be configuration limited due to the high influence of traffic speed
except for those cases of unusual high ventilation.
Considering the general nature of the model, the comparisons between
measured and predicted data are remarkably good and should have wide application
in defining urban and transportation planning parameters in meeting the 1970
Air Quality Standards.
-------
-------
4.0 SITE DESCRIPTION AND MEASUREMENTS
4.1 Summary Tables
4.1.1 Pollutants
Table 4-1 presents the quantity and quality of data collected at
each site. The overall summary results of the study are presented in
Section 1. The majority of the data collected was carbon monoxide data
since there were 8 instruments operating simultaneously at each site
evaluated.
>
*
One Hydrocarbon analyzer, two High Volume samplers, one Tape
sampler and two MRI vector vane metorological sensors were used for the
other data. All of the above data was collected for 24 hours over the
time durations shown in Table. 4-1 The data is tabulated in the Ap-
pendices .
4.1.2 Traffic Characteristics
The Diurnal curves for traffic volumes are shown in Section 2.1.
Table 2.3-1 gives data oh speeds, total volumes, and hourly volumes for
traffic at each site.
All traffic measurements were gathered by the New York State
Transportation Administration in their specially equipped trailer. Pneumatic
tubes were placed on the roadway in each traffic lane, and by measuring
the time lapse between tube depressions as well as the number of de-
pressions, both the speeds and traffic volumes were determined.
As expected, the largest traffic volumes were in the express roads -
Trans Manhattan Expressway, Cross Bronx Expressway, Grand Central Parkway,
and the Brooklyn-Queens Expressway. These roads also registered fairly
high average speeds - about 45 miles per hour - with the exception of the
Trans Manhattan Expressway because of the merging traffic patterns for the
George Washington Bridge approach and exits. City street speeds and
volumes (Canal Street) predictably were lowest of all sites. Average peak
hourly traffic volumes were greatest for the Trans Manhattan Expressway
and Grand Central Parkway, reflecting interstate road travel and transport
between Long Island and the Bronx and Manhattan respectively.
-------
TABLE 4-1 - SUMMARY OF MONITORING ACTIVITY AT EACH SITE
Site
Covered on Top-Open at Sides
54th & Sutton Place-FDR Drive
Long Tunnel
Brooklyn-Battery Tunnel
Brooklyn-Queens Expressway
at Hicks Street Shallow Cut
Short Tunnel
Hugh Grant Circle
Deep Cut
Nelson and Jessup
Grade Road
Bruckner Expressway
Grand Central Parkway at Queei
Hospital Cantilever Site
Brooklyn-Queens Expressway
on Navy Street Viaduct Site
City Street
Canal Street
Intermittent Span
George Washington Bridge
Number of Sampling Days \
Carbon Monoxide
Long Path
Infra-red
Week- Week-
days ends
23 6
7 3
15 4
12 4
11 3
11 4
s
11 4
13 4
11 4
17 6
Hydrocarbons
Flame
lonization
Week- Week-
days ends
8 4
0 0
12 4
8 2
10 2
11 4
11 4
11 4
9 2
17 6
Particulates
High Volume
Air Sampler
Week- Week-
days ends
4 2
2 1
7 1
7 0
7 2
4 2
9 2
3 0
3 0
10 3
Particulates
Tape Sampler
Week- Week-
days ends
10 3
0 0
12 4
13 4
13 4
16 6
10 4
13 4
10 4
6 2
Meteorological Data
\
Week- Week- /
days ends
94
0 0
15 6
12 4
9 4
11 4
11 4
13 4
12 4
14 6
-------
4.2 Site Activity
4.2.1 Franklin D. Roosevelt Drive at Sutton Place
4.2.1.1 Site Description
In order to study the carbon monoxide profile within a cantilever
covered roadway, a unique site configuration along Manhattan's Franklin D.
Roosevelt Drive was selected to be the testing area. At this point along
the Drive, a large apartment building was built directly over six lanes
of traffic in such a way as to form a covered roadway open at one side.
Through this structure ('N- 3/8 miles long) flows a total of three northbound and
three southbound traffic lanes. The outermost lane heading south is an exit
lane which tends to slow down traffic in that direction. The counter flow
lanes are separated by a 2' high medial barrier on which 2' x 2' support
pillars are evenly spaced (r-^151 between pillars). Concrete pillars are also
positioned along the open side of the structure but for all practical purposes
that side is considered open. Fig. 4.2-1 shows the test location at the south
end of the cantilever structure.
Frankling D. Roosevelt Drive carries approximately 98,000 cars per day.
Average speed is 35 miles per hour. The northbound lanes carry 50,000 cars
with an average weekday peak volume of 3,820. The southbound lanes carry
47,000 cars daily at a 3,370 average peak. No truck traffic is permitted.
There are no exhaust fans within the structure, the majority of the
ventilation being provided by the open side. The site is isolated, in
that all CO concentrations can be considered to have totally evolved from
the traffic flowing to and from the site structure.
This site has two important features that make it interesting for
evaluation. First, it is a cantilever covered roadway with one side open
and secondly, people are living directly above the roadway where pollution
levels may at times reach an unhealthy level.
4.2.1.2 Instrumentation Arrangement
The GE Air Pollution Van and the New York State Traffic Trailer
were parked to the west of the structure at the corner of 54th Street and
Sutton Place, approximately 15 feet above the surface of the test roadway.
Figs. 4.2-2 and 4.2-3 show the locations of the trailers and the data
sensors.
A total of 15 probes were installed to monitor the CO concentrations.
Eleven probes were positioned/-^70' within the structure in order to monitor
a two dimensional CO profile perpendicular to the roadway. Three of these
were located along the west wall, three at the medial barrier, and three
along the open side. These probles were 31, 9", and 20' from the roadway.
One probe was located 20* above the roadway in the center of each direction
of traffic.
Three additional probles were positioned adjacent to the apartment
building outside the structure at 12', 27', and 55' respectively from the
sidewalk. One probe was located inside the apartment building at the 55'
level. These four probes provided data on the indoor-outdoor CO relationship.
-------
ro
T>
2o' GS
O
O
A
n
--O
I
I
I
9
i
i
6
30.
M
2'
r -O
20*
O
I
I
I
I
I
ra
Cantilever Covered Roadway - Open At One Side
-------
o
ฃ> VetT-OW.
A M i - v/o I
Q ซSo\.ซv Rec
Ul
a<**)'T~fO/**
ป*ป*
t ovo'*>*
f f^ ซ * *^ fv^
Qj ฃ V/*
i
j
*s
?
^^^
J.
A
a
.
y
^^ _
'
LJK.'
^ **~H5
!
I
i
1
1
i
1
H
i
i
i
i
!
I
1
i
1
ซn ,
\& '
8O ,
(.a
(> ^7 1
* " 1
'ซ' 1
IrnrtFF'C.
Tnttii-ttt -
A
r' I
M
3ซ
IVfl
'
1
1 /J
1
1
pfer-
iO
( ^-2
1
T
I
i
1
i
1
1
1
\ ,
1
i
j:
~*A. "~f
r*
1
pr
1
\
L
//
P'
I
1
|T;
t
(-
i
t
> t
i
i
1 '
|
i - i
j ,'
i
1 ''
^i" ""'', ~ซ
SA ! ,
1 k
1 ' H
^ * ... . i
^-2 U-C.K ! ,
iปปซ4ปซ.V- *TtViO , __
: 61
1 : P
i i
i
1 i
i . "*
Soft^oซ"^ if
'!ปป_" "-'-
'^~ p\\Wv i
1 ' i .
. - ,. ,
;
i 1
. !
I
!
,| * j
j
30
w/3
S-Q
\
1
i
i
j
N
Cantilever Covered Roadway - Open At One Side
-------
Two weather vanes were installed at this site to measure meteoro-
logical conditions. One sensor (designated vane 2 on all figures and tables)
was installed on a mast about 7' above the roof of the
instrumentation van which was parked at curbside on Sutton Place. The other
vane (vane 1) was mounted on a cantilevered support fastened to the fence
enclosing the F.D.R. Drive and extending over the outer southbound lane of
traffic. This sensor was 16' above the level of F.D.R. Drive. Although it
would have been desirable to have this sensor either on the medial barrier or
inside the tunnel itself, these locations were impossible due to mounting
problems and danger of traffic damage to the sensor or its supports.
One Hi Volume Air Sampler and one Tape Sampler were located on the
roof of both the GE Air Pollution Trailer and the New York State Traffic
Trailer.
Other instruments used for monitoring the general air quality ad-
jacent to the trailer were as follows:
1. Hydrocarbon analyzer (checking average concentrations of all
probes)
2. Anderson Impactors & Membrane Samplers (top of trailer)
3. Solar Recorder - (top of trailer)
4. Nephelometer (particle counter)
5. Climate Particle Counter
6. Hygrothermograph (top of trailer)
4.2.1.3 Meteorological data was obtained at Sutton Place between December 3
and December 10, and those observations taken between December 18 and
December 22. This data is documented in Appendix C. During the intervening
period, December 11 through 17, both Vector Vane systems were out of service
due to storm damage and the unavailability of replacement parts.
The first monitoring period was characterized by very stable meteoro-
logical conditions. Winds over the F.D.R Drive blew consistently from a
direction of 70ฐ to 80ฐ relative to the roadway. Sigma values were also
quite low. The December 18 through 22 period was, in contrast, very turbu-
lent. This can be seen by comparing sigma azimuth and sigma elevation
readings for the two periods as shown in Tables C-3, C-6, C-13 and C-16.
These tables show the turbulence parameters measured at both vanes to be two
to three times higher during the latter period. No significant effect of
this increased turbulence can be found on the carbon monoxide averages for
those particular days. '
There is, however, a dramatic change in concentration levels cor-
responding to a change in wind azimuth. Wind asimuth, as recorded over the
road was quite constant through most of the monitoring period. Winds blew
across the road from the river except for two periods: December 18 and 19,
1900 - 0400 hours; and December 21, 0300 to 2400 hours. Unfortunately,
December 21 was a Sunday and a considerable amount of CO data is missing.
Also, being a portion of a weekend, the hourly concentration means are cal-
culated on a small sample. Comparisons with available mean data show a
large reduction in CO levels under the westerly wind regime.
This can be seen by examining the data recorded for probe 2A in
Appendix A. During the afternoon and evening of the 21st, CO levels
-------
recorded were only about one quarter of the mean values and vere, during
some hours such as 1800, only one-ninth those recorded at the same hour on
a previous Sunday (6 PPM on December 21 vs. 56 PPM on December 7). Since
both sigma azimuth and sigma elevation show a large increase on December 21
as compared to December 7, one might conclude that this increased turbu-
lence was the cause of the decreased concentration. However, looking at
wind and CO information for December 22, a day with turbulence almost as
great as the 21st and certainly higher than average but with a wind from
the easterly quadrants, it can be seen that the CO levels are higher than
average. The CO levels on this Monday are also generally higher than those
for the same hours on previous Mondays with lower turbulence. Thus, it
appears that wind direction has a stronger influence on CO levels inside
the partially closed tunnel than turbulence.
Figure 4.2-4 shows CO concentration plotted as a function of dis-
tance across the road for the one remaining day with a west wind which is
December 18. The hours plotted are 1000 and 1100 and the curve stops at
mid-road due to missing CO data on the "B" probe side. Curves derived
from the mean values are also plotted for comparison. Even though the
data is not complete, comparison of the curves reveals the magnitude of
the effect of the west wind in lowering pollution levels in the area of
the probe plane. The December 18 curve also illustrates another important
effect of the westerly wind: that of shifting the maximum concentration
away from the wall.
Wind directions recorded by the vane on the instrumentation van differ
somewhat from those recorded over the road, A histogram of wind azimuth as
recorded by this vane is displayed in Appendix C on Fig. C-ll. Due to the
restriction that tall buildings impose on the air flow, the air motion was
almost always from the north and parallel to Button place. The wind also
blew from the opposite direction of a few occasions, making the distribution
bimodal. This is a characteristic distribution of any location in which the
wind is restricted by walls of buildings or a roadway cut. The fact that
the wind blew from slightly east of south instead of being restricted to
exactly 180ฐ from the prevailing direction was due to the absence of
buildings to the southeast of this sensor.
No diurnal pattern is evident in the averages of the sigma azimuth
or sigma elevation wind functions for the sensor over the F.D.R. Drive.
(See Appendix C, Figs. C-2 and C-4). Sigma azimuth and elevation were,
on the average, higher at the trailer than over the road. This is explained
by the close proximity of the road vane to large walls which restricted the
free passage and size of eddies.
The majority of elevation readings from the trailer vane centered
close to zero. Eighty-five percent of all readings fell between -10 and
l"North" and "South" directions are used for convenience, realizing that
Sutton Place does not run true north and south. All calculations were
made and data is presented such that a wind blowing parallel to Sutton
Place from Uptown to Downtown is defined as a 0ฐ wind or a "north wind.'
The true compass direction of this wind would be about 330ฐ. Similar
conventions are used for subsequent sites.
-------
1'PM
Figure 4.2-4 Carbon Monoxide vs Distance from West Wall
Franklin D. Roosevelt Drive at Sutton Place
-------
+10 degrees. This is as expected, since the sensor was located some
distance from obstacles. The remaining 15% of the readings were negative
and were probably caused, at least in a large part, by the unbalancing
effects of precipitation on the sensor surfaces. The vane over the road,
being located beside a retaining wall, would be expected to show the
effect of air either ascending or descending the wall face. The elevation
angle measured by this sensor did, in fact, show a dependence upon azimuth
angle. When the prevailing wind was from the west, the elevation angle
was negative indicating air descending over the wall from street level to
the level of the F.D.R. Drive. When the wind was from one of the easterly
quadrants, the elevation angle was positive, indicating air ascending the
wall. Because there was such a preponderance of days with winds from
easterly quadrants, and thus rising air, the distribution seen in Figure
C-3 is shifted toward positive angles so that it centers about +10ฐ.
The mean wind speed curve for vane 1 presented in Fig. C-5 does
show a diurnal pattern. The minimum occurs between three and four in the
morning and a broad maximum is in evidence during the afternoon and even-
ing hours. The highest one hour average wind recorded by this vane was
13.5 mph on December 11 between 2400 and 0100 hours. Mean wind speeds
recorded at the trailer were generally lower than those recorded over the
road. A possible explanation for this is that the means calculated for
the trailer vane were based on fewer points than the means for the road
vane. The highest one hour average speed recorded by this vane was 15.7
mph between 0100 and 0200 on December 11. The relatively high winds
recorded on this date and the subsequent failure of the vane systems
coupled with the small number of sample points produced the abnormally
high standard deviations seen during the first five hours of Fig. C-13.
The diurnal wind speed variation pattern is not as pronounced for the
trailer location, probably because of fewer points and the effect of the
high winds of December 11 on the morning means.
-------
4.3.1 Brooklyn Battery Tunnel
4.3.1.1 Site Description
The Brooklyn Battery Tunnel, one of four tunnels connecting the
island of Manhattan with New Jersey and the boroughs of Brooklyn and
Queens, was the long, blower ventilated test site. The tunnel extends
from Lower Manhattan to Brooklyn and consists of two tubes carrying a
total of 4 lanes of traffic. The roadway cross section is rectangular
even though the tunnel itself is tubular. Adjacent to the inside lane
of each tube is a 3' wide walkway. The total width of each road is 25';
each lane being 12.5* wide. The height of the tube is <^* 13'. In order
to ventilate the tunnel, three ventilation towers were built. One such
tower is located in Brooklyn and sits directly above the tunnel supplying
clean air to each tube at various points along the roadway. Fresh air en-
ters from both sides of the roadway (^ 2* from roadway) and exhaust vents
are located in the roof directly above the center of each lane. The venti-
lation to the tunnel is manually controlled depending on CO concentrations
continuously monitored within. Figure 4.3-1 shows the test locations on
the Brooklyn side.
Approximately 51,000 vehicles travel through the Brooklyn Battery
Tunnel each day at an average speed of 25 mph. Normally two lanes of
traffic occur in both tubes. The daily eastbound and westbound traffic
volume is essentially equal. The hourly traffic volume each way varies
considerably and peaks at rush hours. Eastbound traffic peaks at 2500 vehi-
cles; the westbound peak is 2430. The rush hour peaks are accommodated
by three lanes of traffic in the direction of heavy traffic; limiting the
opposite traffic to one lane. This results in two way traffic in each
tube at selected hours of the day. During these rush hour peaks, Manhattan
bound traffic is slow entering the tunnel. This is the result of merging
of traffic from 8 toll booths to the 3 lanes of the tunnel. Brooklyn
bound traffic at rush hour, however, is not slowed because the 3 lanes of
the tunnel are served by 8 toll booths.
4.3.1.2 Instrumentation Arrangement
The GE Air Pollution Trailer was parked above the tunnel adjacent
to the ventilation tower on the Brooklyn side. All wiring and tubing for
data gathering were fed through the ventilation tower to both tubes of the
tunnel. This routing necessitated using all available tubing to connect
the probes in one tube to the instrument van. The CO concentrations were
monitored in both tunnel directions by moving the tubing back and forth
between the probes mounted in the Brooklyn bound and the Manhattan bound
tubes. The van and data sensor locations are shown in Figs. 4.3-2 and
4.3-3.
Fourteen probes were used to monitor the CO concentrations in the
tunnel. All the probes were 200' from the opening on the Brooklyn side
and in a plane perpendicular to the roadways. Three probes were placed on
each wall, at 3', 8', and 13' above the roadway. One probe was located
-------
. ,
. . '";.- . ,;. --v... t -*
-------
-r/0
O
-O-
/
0
3ft
o
\
> c
o--
Long Blower Ventilated, Tunnel
Figure 4.3-2 Brooklyn Battery Tunnel - Elevation View
-------
Co
Co
C./3
9/3
kf
'3 > o a
O
(V
L
_ ifv-f \ ซ> v
HiVol
i i
n
A
sr -
Long Blower Ventilated Tunnel
-------
in the center of each tube, above the two traffic lanes. Two additional
probes monitored CO concentration at an exhaust port in each tube at
various times during the study.
No weather vanes were used at this site due to the unavailability
of a sufficient length of cables to reach the level of the tunnel.
One Hi Vol sampler was located in the tunnel, sampling one tube
for several days and then switching to the other tube. The Hi Vol sampling
locations were 250" from the exit of the Brooklyn bound tube and 300" from
the entrance to Manhattan bound tube. One Anderson Impactor and one Mem-
brane Sampler were also placed at these locations within the respective
tubes.
No hydrocarbon measurements were obtained in the Brooklyn Battery
tunnel due to the usage of all available tubing for CO data gathering.
4.3.1.3 Meteorological Conditions
No wind data was taken at this site due to an insufficient amount
of cable to reach the chosen monitoring position in the tunnel. Personal
inspection of the probe plane area resulted in the only estimations of
wind speed in the tunnel.
Wind speed was estimated to exceed 20 mph in the Brooklyn bound
tube during one-way traffic flow. Air movement was in the direction of
traffic motion, that is, toward the Brooklyn exit. Air movement in the
Manhattan bound tube was estimated at less than 10 mph during periods of
two-way traffic flow. No observation was available during periods of one
way Manhattan-bound traffic flow.
-------
4.4.1 Brooklyn Queens Expressway at Hicks Street
4.4.1.1 Site Description
The shallow cut site for this study was chosen along the Brooklyn
Queens Expressway along Hicks Street between Kane and DeGraw Streets. At
this point, the expressway runs in a North/South direction. The site is
a few blocks north of the traffic exchange with the Brooklyn Battery Tunnel.
Here in Brooklyn, the elevation of the expressway is 21' below the adjacent
roadway levels. The site was an excellent example of shallow cut formation.
Two service roads ran parallel to the roadway at this point. A total of
six (6) traffic lanes flowed within the shallow cut. Fig. 4.4-1 shows
the configuration of the test location, looking north.
Approximately 107,000 vehicles travel daily on the Brooklyn Queens
Expressway at Hicks Street. Average traffic speed is 38 mph. The total
traffic for each direction is roughly equal (^x,53,000 each) however, the
average peak hourly traffic northbound is 3,430 vehicles while the south"
bound peak is only 2,980 vehicles.
The elementary school facing the expressway was an ideal place to
check indoor CO pollution levels resulting from heavy traffic on the express"
way and local traffic on the service road. School representatives were very
eager to assist because of their concern for the children studying within the
school.
4.4.1.2 Instrumentation Arrangement
The GE Air Pollution Van was parked along Hicks Street directly
in front of the Sacred Heart Elementary School which faces the expressway.
The locations of the trailer and the data sensors are shown on Figs. 4.4-2
and 4.4-3.
All 16 probes were installed to monitor the CO concentrations.
Eleven of these were positioned in a plane perpendicular to the roadway.
Three (3) vertical probes were placed down each wall and three down the
center to the medial strip. One probe was centrally located, 21 ft.
above each direction of traffic. The vertical probes were positioned
3", 14' , and 21" from the roadway. Another probe was located 20' above
Hicks Street, and monitored the CO emission above the service road. Two
CO probes were placed inside,the school at 15' and 35' respectively. Two
other probes were located outside the school at the same levels. These
four probes provided data on the indoor-outdoor CO relationship.
Two vector vane systems were used to monitor wind conditions at
this site. One vector vane was placed on the medial strip 14' off the
roadway and approximately 15' north of the plane of the probes. Another
vane was positioned 181 above Hicks Street on a pole at the intersection
of the east wall plane and the plane of the probing. A Hi Vol sampler,
solar recorder, and a hygrothermograph were located on the roof of the
GE trailer. One tape sample was placed inside the school on the 1st
floor. Hydrocarbon concentrations monitored were average values derived
from a summation of all probes.
-------
OJ
Fig. 4.4-1 Brooklyn Queens Expressway at Hicks Street
-------
1st
to
Q A
26,
8*
3/9
ซ *0ปi
MHO
PC.ซ
1
Oeta
?
1
i^TT
V
1
1
i
I
rO* T .
1 tf/MMC % .* ,
i 1 ;
w1 | l^1
. ^ 1
1 3' 0ซป Y
i
JTA
<
I
4
in
ฃฑ l-A i O o
f
Shallow Cut - Looking South
-------
00
oe
o
*
^
Q
W
fO
ie
a
GE
v/u
k1
VI
g
3
80
Figure 4e4-3
1>1
O
O
7&
8A
C*A
MrS
V
j
Shallow Cut
^r. Queens Expressway at Hicks Strsst - Flan View
-------
4.4.1.3 Meteorological Conditions
The wind sensors were installed January 16, 1970 but usable data
exists only for the period January 20 to February 5, due to vandalism of
sensors and accumulations of road salt on sensors in connectors. Vane 2
was severely and repeatedly affected by salt spray from passing vehicles
on the nearby service road, rendering all channels except azimuth and
sigma azimuth inoperative for the duration of monitoring. The azimuth
angle as recorded by Vane 2 shows that, for the main portion of the monitoring
period, the air flow was parallel to the Brooklyn Queens Expressway towards
Queens (see Fig C-25 in Appendix). Air sometimes moved across the top of
the cut from east to west as can be seen in the same figure but almost never
flowed in the opposite direction. This cross-highway wind is in reality, the
lower branch of an eddy formed over the Brooklyn Queens Express by the
influence of nearby buildings. Figure 4.4-4 illustrates this eddy which was
observed during westerly and northwesterly wind flow only. This figure
depicts a smoke plume from a nearby factory which made the eddy visible several
times during westerly flow. A plot of wind direction vs. time at the Central
Park National Weather Service observation point, as well as the direction
recorded by Vane 2 just above the cut at Hicks St., is presented in Fig. 4.4-5.
The six day period chosen to illustrate the eddy effect exhibits several
changes in wind direction. On January 24, as the general wind flow over the
city trends towards north through west, the wind recorded above the cut trends
toward north but through the easterly directions. This occurs again on
January 26. General southerly winds in the area also produce southerly winds
over the cut and elimination of the eddy (January 25 and January 28). This
same elimination of the eddy would be expected of northerly winds over this N-S
street but northerly winds were too infrequent to illustrate this point.
The eddy, as can be seen in Figure 4.4-4 may be expected to drag
clearner air from aloft down the face of the school building and across the
expressway. This fact may be demonstrated by comparing measured CO pollution
levels on days when the eddy was present with levels recorded on days when the
wind flow was parallel to the road. Figure 4.4-6 is a plot of carbon monoxide
concentration vs. time of day for two days, January 26 and 27. An examination
fo the curves discloses that the levels on January 26 were trending downward
during the day and the levels on January 27 were trending upward. Referring to
Figure 4.4-5, it can be seen that the eddy (showing as an easterly component
on the Brooklyn Queens Expressway data) makes its appearance at about nocn on
the 26th and vanishes at about the same time on the 27th. The carbon monoxide
concentrations rise and fall in concert with these changes in direction, being
generally high until about noon on January 26, then falling to comparatively
low levels over the ensuing 24 hours. Pollutant levels become elevated again on
afternoon of January 27.^ It is interesting to note the magnitude of the eddy
effect on pollutant levels on these two days. (It should be noted that traffic
volumes were approximately equal on the two weekdays.) During the early morning
hours, carbon monoxide concentrations were almost 4 times higher during southerly
wind regimes than during easterly flow. During the late evening, the CO levels
occurring when the flow as parallel to the street were about twice as high as
those influenced by the eddy circulation.
The sigma azimuth readings from Vane 2 were markely higher during
periods of easterly flow at the vane level. These high values promote
-------
-------
ISO1
270ฐ
I
8
rl
3
90'
180<
j
/
Central Park Data
BQE Data - Vane *2
8 24 06 12 18 24 06 12 18 24 06 12 ~18
Jan. 24 Jan. 25 Jan. 26
24 06 12 18
Jan. 27
24 06 12 18 24 06 12 18
Jan. 28
Date and Hour
-------
"
o
H
g
o
ง
ซ
TJ
1
I
3
30
28
26
24
22
20
18
16
14
12
10
8
6
Probe 05B
January 26, 1970
Probe #SB
January 27, 1970
24 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
21 22
Hour
23
-------
vigorous mixing and diffusion of the CO and would tend to generally
lower concentrations at the probe locations. The amount of reduction in
concentration on such days as January 26th and 27th, due simply to the
increase in turbulence and not to a different air source, cannot be
estimated from the present data, since all turbulent days were also eddy
days and all non-turbulent days were during periods with winds parallel
to the street. Both the increased turbulence and the transport of
cleaner air from roof level down into the street lead to decreased con-
centrations.
An examination of the mean sigma azimuth plot (Figure C-26) dis-
closes very little resemblence to the expected diurnal pattern. The
curve is quite flat with only a hint of a minimum in the early morning
hours. This is undoubtedly due to the small sample size and considerable
missing data. Weather systems were quite active during this period and a
latger sample would be needed to obtain a representative diurnal curve.
The azimuth frequency of occurrence chart for Vane 1 (median strip)
found in Figure C-21, shows very little remnant of the eddy circulation in
the cut itself. The distribution shows two maximums, one near 0ฐ and the
other near 170ฐ. In other words, the air in the cut almost always moves
parallel to the cut. A wind from any northerly quadrant above the cut
generally becomes a northerly wind in the cut and a wind from any southerly
quadrant becomes southerly. An additional illustration of the restricted
flow in the cut may be found by comparing Figure C-23, the plot of the
mean u component with Figure C-24, the plot of the mean v component. The
wind component parallel to the road (u) is consistently at least twice
that of the component across the road (v).
Mean sigma azimuth as recorded on the median is of approximately
the same magnitude as that recorded above the cut. The mean values
(Figure C-18) again display only slight resemblence to the expected diurnal
pattern. A minimum occurs early in the morning, as would be expected;
however, another minimum occurs at 6 PM. The maximum occurs at 9 AM. Any
effect of traffic generated turbulence on sigma azimuth appears to have
been effectively masked by large weather variations and, possibly, the in-
clusion of weekend data in the means. That large day to day variations
were the rule in this function may be seen by examining the standard devi-
ation curve on Figure C-18. The magnitude of the standard deviation is 1/3
that of the mean values for almost all hours.
The wind elevation angle as recorded on the median is presented in
Table C-24 and a frequency histogram may be found in Figure C-19. The
average angle is very near, but just slightly negative of zero degrees
(horizontal). 83% of all readings fell between 0ฐ and -10ฐ. Although this
negative angle would appear to indicate a slight downdraft condition, the
unbalancing effects of precipitation as well as salt and dirt deposits on
sensor tail surfaces may have contributed some inaccurate measurements. As
access to sensors, after initial installation, was generally restricted by
traffic and safety factors, the quantitative effects of deposits could not
be determined by removing the sensor and measuring the balance point in a
quiet atmosphere. In any case, the angle is so small that the mean verti-
cal velocities are usually only .5 mph. (See Table C-38 and Fig. C-30).
-------
Sigma elevation readings from Vane 1 were among the highest
encountered at the nine sites for which wind data is available. However,
the mean hourly values (Figure C-20) again exhibit the strange minimum
near 6 PM and rising values thereafter that was found in the sigma
azimuth curve. The diurnal curve is again more the reflection of frontal
passages and storms than turbulence generated by passing vehicles.
The highest wind speed recorded in the cut (hourly average) was
18 mph from the south. This value was recorded on February 2 at 1500.
Gusts during that hour reached over 40 mph and the wind system failed
shortly thereafter.
-------
4.5.1 Cross Bronx Expressway at Hugh J. Grant Circle
4.5.1.1 Site Description
The short unventilated tunnel situated beneath Hugh J. Grant
Circle on the Cross Bronx Expressway was the fourth site visited in the
Urban Expressway study. A total of 6 lanes of traffic flow through this
tunnel. A wall divides the 3 west bound lanes from the 3 east bound
lanes. The tunnel is not ventilated so high pollution levels were ex-
pected. The test roadway is once again lower than the surrounding ground
level as shown in Figure 4.5-1.
Approximately 86,000 vehicles pass thru the tunnel daily. The
eastbound lanes carry 44,000 vehicles at an average speed of 42 mph.
42,000 vehicles travel westbound at 48 mph. The eastbound traffic volume '
rises during both the morning and evening work hours and peaks at 2860
vehicles about 5 PM. The westbound traffic is relatively constant between
8 AM and 6 PM, with a 2570 vehicle peak near 3 PM..
4.5.1.2 Instrumentation Arrangement
The GE trailer was parked on the north side of the Cross Bronx
Expressway at the west end of the tunnel approximately 25* above the
surface of the roadway.
A total of 16 CO probes were in the tunnel, as shown in Figures
4.5-2 and 4.5-3. Six probes were placed in each side of the tunnel, 100"
from the west end, in a plane perpendicular to the traffic flow. Three of
these probes were 3', 91, and 16' off the roadway along the inner walls.
Two probes were located along the outer wall at 31 and 16'. One probe was
also placed at the 16' height over the geometer center of the roadway in
each direction. Another probe was placed in each side of the tunnel at
the 16* level along the north wall at a distance of 50" from the west bound
exit. One probe was placed adjacent to the west bound opening of both
sides of the tunnel at 16'.
One vector vane (vane 2) was installed on the median strip of the
Cross Bronx Expressway 14' above the road surface and 32* west of the Grant
Circle tunnel. This was approximately in the middle of the open space
formed between Grant Circle and the White Plains Road overpass. The other
wind sensor (vane 1) was placed, in an inverted position, over the exit of
the westbound tunnel. It was 14* above the road surface.
A Hi Vol and a tape sampler were located on the median strip,
adjacent to vector vane 2. A Hi Vol sampler, a solar recorder, and the
hygrothermograph were located on the roof of the trailer.
4.5.1.3 Meteorological Conditions
Wind monitoring commenced at this site on February 19, and ended
on March 6. Again at this site, the amount of data collected was reduced
by equipment breakdowns and malfunctions as well as difficulty of access
to equipment in the middle of an expressway. All channels on vane 1 (over
the tunnel entrance) except for wind speed failed on February 27 and re-
mained inoperative for the duration of monitoring at this site. Azimuth
-------
-------
O- .QO-#' O
|8B
'68
no
(3oui40
10
O
O
o o
Mft
ISA tT
<*'
I
O CO
U. Oo\
r~\
Short Unventilated Tunnel
-------
OE
O
D
\
O
60
4
ฐFF
I?
* ' ^ **
00
O
2/4
----- O ---
1
-------
and sigma azimuth functioned fairly well on vane 2, however, elevation
and sigma elevation were missing from February 27 through March 6.
Virtually no wind speed data was obtained from sensor #2. All wind data
and statistics for this site are presented in Figures C-33, through C-47,
and Tables C-41 through C-60 in Appendix C.
Wind azimuth as recorded on the medial barrier shows a bimodal
distribution as can be observed in:the histogram in Figure C-33. This
type distribution occurs because the portion of the site in which the
vane was located was actually a shallow open cut. The sensor over the
westbound, tunnel exit indicated that the wind blew from the east during
the majority of recorded hours. This is probably due to the traffic in
the tunnel inducing a wind in the direction of traffic motion.
The plot of sigma azimuth vs. time as recorded by the median vane
showed only slight evidence of a diurnal pattern. Also, the values recorded
at this vane are lower than those recorded by the tunnel vane. This is due
to several factors. First, the sample period and number were not the same
for both sensors and a certain amount of the data is doubtful due to missing
and damaged tail surfaces. The effect of missing tail surfaces on the
median vane would normally be evident in the elevation data due to the re-
sulting sensor imbalance. Data presented in Tables C-54, C-55 and Figure
C-43 does show the highest elevation angle frequency of occurrence to be
almost 50 from the horizontal. This is very unexpected since the sensor
was not placed near any obstacle. Close examination of the strip chart
records disclosed the presence of an electronics malfunction in the eleva-
tion channel which masks any difference between good and bad data received
from the sensor. For this reason, there is no way to edit the sigma azi-
muth data to remove incorrect readings.
The elevation channel malfunction naturally affected the input
to the elevation sigma computer. This resulted in values far below those
recorded at the mouth of the tunnel and also far below what they actually
should be. The wind speed channel in this sensor operated for only one
day, thus precluding any analysis.
The sensor over the tunnel exit (#1) provided fairly reliable
sigma azimuth data over a period slightly in excess of one week. As this
sensor was in close proximity to the traffic, the effect of vehicle motion
may be seen in the sigma azimuth measurements. Figure 4.5-4 presents
diurnal plots of the mean sigma azimuth as recorded at vane 1 and the traf-
fic flow leaving the tunnel. The similarity in shape of the two curves
illustrates the effect of moving vehicles on the nearby atmosphere. The
absolute minimum and relative minima and maxima of the two curves occurs
during the same hour in may instances, such as 0300, 0700, 0800, 1700,
2200 and 2300 hours.
The effect of traffic generated turbulence is not seen as clearly
in the sigma elevation channel. Turbulence was so extreme during a large
number of hours that the sigma meter constantly read full scale. The lack
of sufficient range in the instrument caused the range of average values
to be compressed. Although the minimum values still appear in the early
morning (see Figure C-36), the values for the remainder of the day are
constant, at or near the full scale reading of 22.5 .
-------
I
Figure 4.5-4 Comparison of Traffic and Turbulence Curves at
Hugh J. Grant Circle
-------
The diurnal variation in the average wind speed curve, as de-
picted in Figure C-37, is slight with the range between maximum and
minimum equal to 1.1 mph. The highest average wind speed recorded at
this site was 10 mph as noon on February 23. The wind speed at this
sensor never dropped below 2.0 mph even during the early morning hours.
This is probably due, at least in part, to the motion of traffic beneath
the sensor generating some local air motion even when the lower atmosphere
was calm.
-------
4.6.1 Cross Bronx Expressway at Nelson and Jessup Avenues
4.6.1.1 Site Description
The Cross Bronx Expressway between Nelson and Jessup Avenues
is a perfect example of a deep cut site. The roadway cuts through
solid rock, 55' lower than the north service road where the GE air pollu-
tion trailer was located. A total of eight lanes of two way traffic
flowed on the expressway beneath the Jessup Avenue Bridge. The bridge is
<^210 ft. long and spans the expressway at this point. The one wall
running down to the expressway is brick while the other side is solid
rock. The brick wall is not perpendicular but is angled toward the
roadway. The other side is completely irregular in appearance composed
completely of rock and gravel. One lane in the westbound direction is
an exit lane while in the eastbound direction, one lane is an entrance
lane. There is a slight upward grade when traveling westbound on the
expressway. Figure 4.6-1 shows the site, looking westbound.
Approximately 118,000 vehicles use the Cross Bronx Expressway
at this point each day. Traffic peaks occur in both directions during
both the morning and evening rush hours. The eastbound traffic peaks at
3590 vehicles while the westbound peak reaches 4010 vehicles. 58,000
vehicles travel eastbound at an average speed of 49 mph. The westbound
traffic of 60,000 vehicles averages 47 mph.
4.6.1.2 Instrumentation Arrangement
The GE Van was located on the north service road 55* above the
surface of the roadway and about 83' west of the Jessup Avenue Bridge.
The site and probe configuration are shown in Figures 4.6-2 and 4.6-3.
Sixteen probes were used to monitor the CO concentration in
and adjacent to the site. Three sets of probes were positioned from
the bridge down to the roadway. One set went down the brick wall along
the westbound lanes, one down the middle and the last down the stone
wall. Each set contained three probes. These probes were located 3',
26" and 55* above the roadway.
Another probe was located over the westbound lanes, 55' off
the roadway. Three additional probes ran down the brick wall (north
side) 83* west of the bridge. One probe monitored the CO concentrations
at the top of the cut on the north side 45* from the bridge. Two probes
sampled ambient air in the trailer
A Membrane sampler, an Anderson Impactor, a tape sampler, Hi
Vol Particulate sampler, a Solar recorder, and a hygrothermograph were
all located on the trailer roof.
One vector vane was located on the medial strip 45' west of the
Jessup Avenue Bridge. The vane was 14' off the roadway.
-------
Ln
u>
PHOTOGRAPH #12
-------
Deep Cut
Figure 4.6-2 Gross Bronx Expressway at Nelson & Jessup Avenues
Elevation View
O co
>Ja\
Q
-------
4$'
i
o
A
oo
841
HO
(DO
I
V
"A
o
HA
Oqa
V
i
t
X
i
0
k
V)
V
,_L
ZA
Vi
Dsep Cut
SA
ป*
-------
4.6.1.3 Meteorological Conditions
Due to severe sensor damage, mechanical failure problems, and
vendor inability to supply immediate replacement parts, it was necessary
to use makeshift components, including a hastily fashioned plexiglass
tail section. These makeshift components resulted in the sensor being
considerably heavier than the design weight. The resultant increase in
the inertia of the sensor caused the sigma parameters to be in error
for this site. Although values obtained at this site should not be
compared to those obtained at other sites, intra-site relationships
should remain qualitatively valid.
Wind circulation at this site was very similar to the circula-
tion at the shallow cut site. The wind almost always blew parallel to
the sides of the cut as can be seen on Figure C-49. (0 on this chart
indicates a wind parallel to the road and blowing toward the George
Washington Bridge). Figure 4.6-4 illustrates the pronounced effect of
the deep cut in translating the general air motion into components paral-
lel to the cut. For example, for all hours that a southeast wind was re-
corded at Central Park, 94% of the time the wind in the cut was from the
east and 6% from the north. In general, flows from the east, southeast
and south become easterly at the 14* level of the cut at least 80% of the
time. Winds from the southwest, north, and northwest are translated into
west for a majority of hours, however their behavior in the cut suggests
the presence of a spiral-type eddy under these regimes. General north-
easterly winds become disorganized and appear distributed equally between
east, south, and southwest. It is possible that tall buildings located
along the north edge of the cut disrupt the northeasterly flow over the
cut and generate a complex eddy system causing this effect.
Sigma azimuth recorded in the cut, and presented in Figure C-50,
is relatively constant over the day. A minimum occurs between 0200 and
0300 but no pronounced maximum can be found during the daylight hours.
As expected, and shown on Figure 4.6-5, CO concentration at the site in-
creases as sigma azimuth decreases. (In order to determine the effect of
of changes in the horizontal turbulence parameter on carbon monoxide con-
centration, one average value each for CO at probe #6B /55 feet above the
median/, total traffic flow rate, and sigma azimuthwere determined for
the five weekdays when complete data was available. The daily averages
of carbon monoxide concentration and the sigma azimuth parameter were used
in plotting the line seen in Figure 4.6-5. This line was determined by
linear least squares fit. The total traffic averages were used only as a
check to determine whether varying traffic volume was causing the changes
in CO levels. Over the five day period, traffic held fairly constant, its
This figure consists of 8 bar graphs representing the eight octants of
a circle, each being centered about the directions: N, NE, E, etc.
The "general flow" in each graph represents the wind recorded at the
Central Park National Weather Service Station which is assumed to be
representative of the general wind direction in the area of the site.
The wind directions recorded in the cut are plotted in terms of per-
centage of total hours from a given direction;for each of the eight
general wind flows.
-------
E SE S SW W NW N
07.
fc li S l W
Wind Direction in Cut
WinJ Direction in Cut
Figure 4.6-4 Relationship of
at Jessup Avenue
General Wind Flow to Azimuth Recorded in Deep Cut
Site
-------
Figure 4.6-5 Sigma Azimuth vs CO Concentration - 5 Weekday Averages - CBE at Jessup
-------
range being 5.2% of the largest value. In contrast, CO and sigma azimuth
varied over 47% and 35% of their highest values, respectively. The
changes in traffic can therefore be considered small compared to the
changes in the other two quantities. This allows traffic to be con-
sidered a constant for this particular set of points. The exact form of
the relationship at this site, however, cannot be determined from these
few points.)
The prevailing wind elevation angle recorded in the cut is about
-5 , indicating that air moved downward more often then up and out of the
cut. This angle, however, is very small and the previously discussed
emergency repairs to the wind system as well as precipitation effects
cast doubt upon its validity. An additional indication of a slightly out-
of-balance sensor is the tendency for elevation angles to become larger
in a negative direction during the calm hours of the early morning such as
on March 19, 21, and 23. (See Table C-64).
Sigma elevation recorded at this site is plotted in Figure G-66.
The curve is very similar in shape to the sigma azimuth curve. Hourly
data on most days also follows the same pattern and thus retains about
the same relationship to concentration as that for sigma azimuth.
Winds were generally light with only 10 hours out of more than
200 in excess of 10 mph. The average wind speed was 5.5 mph. The hourly
wind speed curve, presented in Figure C-53, has a minimum between 0300 and
0400 and a maximum between 1400 and 1500.
-------
4.7.1 Bruckner Expressway Near White Plains Road
4.7.1.1 Site Description
Bruckner Expressway extends from the Tri-borough Bridge to the
Throgs Neck Bridge in the Bronx, carrying six lanes of two-way traffic.
The "Grade Road" site chosen for this study was located along this
expressway between White Plains Rd. and Castle Hill Rd. Figure 4.7-1
shows the test site, looking east.
The most important characteristic of this site was the fact
that it is completely open. There were no natural or man-made obstacles
such as walls, buildings, etc., to interfere with the wind currents and
normal CO diffusion patterns. There was, however, a metal sign structure
spanning the entire roadway at this point. A total of 8 lanes of traffic
flow beneath the sign structure with two of the lanes being an exit and
an entrance lane to and from parallel service roads adjacent to the ex-
pressway.
Approximately 55,000 vehicles travel on BrucknerExpressway daily
at an average speed of 48 miles per hour. The traffic flow southbound
peaks at an average of 2140 vehicles during the morning rush hour. The
evening rush hour produces a northbound average traffic peak of 2550
vehicles. Average total daily traffic southbound is about 25,000 vehicles
while the northbound average exceeds 29,000 vehicles daily.
4.7.1.2 Instrumentation Arrangement
Figs. 4.7-2 and 4.7-3 show the probing scheme and associated site
information. The GE trailer was parked on the south service road which
ran parallel to the expressway. The overhead sign structure was utilized
to position the various CO probes relative to the roadway.
Twelve probes were used to monitor CO concentration adjacent to
the roadway. CO probing ran overhead from the GE trailer to the sup-
porting metal framework. Three probes were positioned at the North, Center
and South sides of the roadway at 3', 14' and 25' above the roadway. One
probe was also placed at the center 25' above each direction of traffic.
One probe was 20' above the service road on which the trailer was situated.
A Hi Volume Particulate Sampler, solar recorder, Tape Sampler,
and hygrothermograph were on the roof of the trailer.
Hydrocarbon concentrations were monitored by utilizing all the
probes and were average site values.
Meteorological readings were taken with two vector vane systems.
Vane 1 was located on the south corner of the sign structure, 30" above
ground level. Vane 2 was located, as was customary, 14' above the medial
barrier and approximately 20' east of the sign bridge. The sign structure
was of an open type construction and should have had only minimal effect
on the wind parameters as measured by Vane 2.
-------
-------
Gra.de Road
O Co
-------
V)
v
u
, -O'
t.
20'
I
o e/=ป
Grade Road
Fig. 4.7-3 Bruckner Expressway Near White Plains Road - Plan View
O co
"ฃฑ H.
V
-------
4.7.1.3 Meteorological Conditions
Wind data was taken at this site from April 8 through April 22.
Minimum problems were experienced with the measuring equipment. All data
taken at this site is presented in Tables C-71 through C-90 and Figures
C-57 through C-72. (Zero degrees azimuth at this site indicate a wind
parallel to the Bruckner Expressway and blowing from the west.) The
prevailing azimuth direction" recorded at vane 1 was from 175ฐ; or from
the east. The hourly directions recorded at this vane agree very well
with the general wind over the New York City area. This is due to the
openness of the site and the height of the sensor above ground level.
No consistent southerly winds occurred during the operational
period of vane 1, eliminating the possibility of a quantitative investi-
gation of eddy effects caused by a large apartment structure to the
south of the monitoring location. This eddy was visually observed, however,
in an experiment performed on April 9 when the wind was generally south to
southwest over the site. A small gas filled balloon of approximately
neutral buoyancy was released in an area to the north of the apartment
building where rising air had been noted earlier. This balloon was carried
to the top of the building and then downward and away from the building,
finally to return to its release location by moving near ground level
back toward the building. This circuit, describing the size and shape of
the eddy at the lee side of the building was repeated several times until
the balloon finally left the eddy and entered the main southerly air flow
by penetrating the separation layer at roof level. It was then carried
downward and toward the northeast, reaching ground level at the approximate
location of the median on the Bruckner Expressway. The fact that the eddy
does not quite reach the location of the wind sensor on the median strip
may be seen in the data from April 8 and 9 which shows southerly winds at
the medial barrier location.
The sigma azimuth curve for vane 1, shown in Figure C-58, is quite
flat with small maxima near 0900 and 1800 and a minimum near 0400. Sigma
azimuth for vane 2, closer to the ground and to moving traffic, was, on the
average, higher but still maintained the same flat characteristic.
The sigma elevation turbulence parameter at the medial location
also shows very small range in its mean values. The readings were virtu-
ally the same from 0600 through 2400. Sigma elevation recorded at the
higher vane showed considerable more variation with time of day, having a
broad maximum during the daylight hours and a minimum between 0300 and
0400 hours.
Determination of any relationship between turbulence parameters
and pollutant concentrations at this site was very difficult due to the
extremely small values and limited range of the CO concentrations and the
small range- of turbulence measurements. Turbulence, however, was con-
siderably above the mean on four days during the monitoring period. On
one of those days, Saturday April 18, meteorological conditions which were
unusual may be used to gain insight into the meteorology-pollutant re-
lationships at this site. Figure 4.7-4 is a plot of sigma elevation from
vane 1, carbon monoxide as measured at the trailer, and total traffic all
-------
500
45C
Figure 4.7-4 Turbulence - Traffic - Pollutant Relationship
Saturday, April 18, 1970 - Bruckner Expressway
-------
plotted against time. (The trailer probe location was chosen in this
case because it is near a living area and serves to illustrate the
adverse effects that may occur downwind from a source under certain meteor-
ological conditions. Upwind probes, those on the north side of the ex-
pressway, showed the same pattern but to a lesser extent.) As can be
seen in this figure, carbon monoxide levels in front of the apartment
structure appear virtually independent of traffic volume on this particu-
lar day. This is a result ofJthe immense change in atmospheric stability
occurring at 0600, masking any traffic effect on CO levels. Accompanying
the very sharp increase in sigma elevation was a corresponding increase
in sigma azimuth and wind speed. The sharp decrease in CO concentration
was due to vigorous mixing and not a different air trajectory. It is
interesting to note that the value of 13 ppm reached during the calm at
0500 was met or exceeded only 6 hours at this location during the entire
monitoring period.
The diurnal wind speed patterns recorded at the two vector vanes
are similar. A minimum occurs in the early morning and a broad maximum
over several hours at mid-day. A small indication of a double peak
occurs in this maximum but it is probably not traffic induced. The range
of mean wind speed values was among the largest of any site, being 4.6
mph for the 30' vane (vane 1). The highest wind speed recorded was 20.8
mph at 0900 and 1000 on April 15. The average wind speed of 8.1 mph was
the highest wind speed of any site. The large ranges and high averages
are due to the open configuration of the area. Wind speed decreased from
the 30' vane to the 14* median vane. Apparently any traffic generated
wind velocity at this site is not enough to overcome the normal downward
decay of wind speed associated with the logarithmic wind profile.
-------
4.8.1 Grand Central Parkway at Parson's Boulevard
4.8.1.1 Site Description
The Grand Central Parkway, at Parson's Boulevard in Queens,
becomes a subsurface roadway and is part of a unique roadway
structure not often encountered. The roadway is partially
covered on both sides by two elevated service roads. A total
of six lanes of traffic flow within this configuration. The
four inner lanes are uncovered while the two outside lanes
are sheltered by two cantilevered sections. Adjacent to each
traffic direction is a 8* pavement and a 4' medial barrier.
Figure 4.8-1 shows the roadway configuration at the test site.
The Parkway carries approximately 110,000 cars per day. Average
speed is 45 miles per hour. The eastbound lanes carry
57,000 cars with an average weekday peak volume of 6,000. The
westbound lanes carry 53,000 cars daily with a 6,120 average
peak. No truck traffic is permitted. The parallel service
roads carry 23,000 vehicles per day.
4.8.1.2 Instrumentation Arrangement
The GE Air Pollution Van and the New York State Traffic Trailer
were parked along the North Service Road adjacent to the Queens
Hospital. Nine probes were used to monitor the CO concentra-
tion on the parkway as shown in Figures 4.8-2 and 4.8-3. These
probes were 3', 10' and 20' off the roadway. One set was along
each site of the double cantilever structure while the third
set was along the median strip. Three other probes sampled CO
concentration 20' over the adjacent service roads. CO was also
monitored directly outside the GE trailer (probes 6B & 7B).
Two probes were used to provide data on the indoor-outdoor CO
relationship at the hospital. One of these was placed inside
the nurses quarters on the 4th floor. The other was located on
the outside of the hospital at the same elevation.
Two weather vanes were installed at this site to measure meteor-
ological conditions. The road vector vane (vane 2) was 25' west
of the probe plane along the median strip at a height of 14".
Another vane (vane 1) was located on a pole on the wall separating
the North Service Road from the Parkway across the street from the
GE trailer at a height of 25' above the service road.
One Tape Sampler and one Hi Vol were located on the GE trailer and
a second tape sampler was placed inside the nurses quarters on the
3rd floor.
4.8.1.3 Meteorological Conditions
Wind data was acquired at this site for the period April 27 through
May 11. The data and associated statistics for this site are pre-
-------
cr.
co
-------
0
i
so
}
b
O
_Q__ _ Q
V3 30 T
t
, Se.u\>
-------
1
1
1
1
1
1
1
r
MO 3O
> Ob
Ifi
1*0
^
ts
***
** *
o
1
1
1 1
D
88
T
i"
^ *n
^^
1 ^ฐ
v- ^
IT v
;
i
D
(oP,
J?"
T
r
/
/
i
Double Cantilever Covered
O
f
*.
f
1
1
1
I
1
1
1
1
1
i
0 q c/-*- -
V/9|
t
1
1
1
1
1 *Ok
1 o*/ J
'*'
i
I
<
il
oe
VAN
OA
a
r^rซ
~-
_
^
_5V=ป
- D
i.
A
^
r-
yv
^
i >
fl
s
Q
V
I N
i
i
i
i
V
-O f o pvoV>-ป
A U.ฐ vol
-------
sented in Figures C-73 through Cr88 and Tables C-91 through
C-110 in Appendix C.
The predominant azimuth reading recorded at the service road
was 0ฐ. (The 0ฐ reference for both vanes at this site repre-
sented a wind parallel to the Parkway and from a generally
westerly direction.) Virtually no wind was recorded from 50ฐ
to 160ฐ, as Table C-92 "illustrates. This is due to a combina-
tion of factors. First, Queens Hospital, located just to the
north of the service road, shielded the sensor from any direct
exposure to a northerly wind. Second, wind flow over the city
during the monitoring period was from the north on only 11 out
of 360 hours, as determined from National Weather Service
records from the Central Park Station. Finally, northeasterly
winds over the area were channelled parallel to the street and
easterly in the locality of the sensor and northwesterly winds
were chanelled parallel to the street, and therefore westerly.
Winds from the southerly quadrants did occur, as the topography
was open to the south of the sensor.
The sigma azimuth curve for the service road sensor shows a
typical diurnal pattern. The minimum occurs between 0400 and
0500 hours and, after a fairly steady increase in turbulence
through the morning and early afternoon, the maximum occurs
near 1600 (see Figure C-74). A calm period similar to that
discussed for the Bruckner Expressway was encountered early in
the morning of April 25. Values of sigma azimuth dropped to
less than one degree early that morning. Wind velocity also
dropped to near zero. This stagnation of atmospheric motion
caused a considerable buildup of carbon monoxide in the air above
the Parkway. This is evidenced in the concentration measured at
probe 3B, over the eastbound service road. Concentrations at
this probe over these morning hours were among the highest re-
corded any day at those same hours and were approximately twice
the mean values. The same effect was also seen inside the cut
where concentrations at probe 4A, north wall - 3 foot level,
ran between two and three times the mean levels. This was not
a traffic-induced phenomenon, as the total traffic volume in the
early morning hours of April 29 was at or below mean values.
The elevation angle frequency of occurrence data in Table C-95
indicates that, on the average, the wind at the vane located
over the service road had a slight downward component of about
5ฐ. No dependence of the elevation angle on horizontal wind
direction can be found for this sensor. Some dependence on
wind speed is noted, but this is probably due to the magnifica-
tion of errors associated with the sensor balancing in low-wind
velocity conditions.
The vertical turbulence parameter, sigma elevation, as recorded
by vane 1, is similar in its trends to the sigma azimuth channel.
The minimum of the sigma elevation curve occurs between 0300 and
0400, and the maximum near 1600.
-------
Wind speeds over the service road were, on the average, quite
low compared with previous sites. The average for all hours
was 3.7 mph. Calm or near calm conditions occurred several
mornings, including April 29, May 1, 9, and 11. Except for
the morning of May 1, these stagnant conditions were associated
with significantly elevated carbon monoxide concentrations.
Figure C-77, which is a plot of wind speed over an average day
at this sensor location, shows a typical diurnal curve with
the minimum during the early morning and the maximum occurring
during the time of maximum surface temperature, at about 1500.
The prevailing azimuth angle recorded by vane 2, located in the
partially covered cut, was 125ฐ. Wind never blew from this
direction above the cut. The azimuth frequency distribution
was evenly spread from 130ฐ through 300ฐ; however, there were
virtually no hours of wind between 0ฐ and 110ฐ. These facts
indicate that the azimuth at this site behaved differently
from that at any other cut. This is undoubtedly due to the
fact that this cut was partially covered. It appears that a
great deal of the external influences on the wind in the cut
were effectively removed by the partial cover.
Sigma azimuth values in the cut were higher than those measured
over the service road by about 2 . It is also noted that the
azimuth traces tended to wander over large angular deviations -
direct result of the lack of a strong steering effect by synoptic
winds. This is one of the causes of the higher sigma azimuth.
Although the weather-induced trends that were measured by vane 1
are apparent to a limited extent, the calm periods of the early
morning hours are no longer as calm - an effect of moving traffic
close to the vane. The diurnal pattern of the sigma azimuth
values in the cut also differs from the pattern recorded on the
service road. No longer is there a smooth transition from mini-
mum to maximum. Instead, there is a new second maximum between
0800 and 0900. The minimum remains near 0400, but the afternoon
maximum is shifted one hour later to 1700. The morning maximum
is real. An examination of the standard deviation curve in
Figure C-82 (dashed line), as well as comparisons with the
standard deviation curve for vane 1, support this statement.
The double-peaked feature is caused by the influence of moving
traffic on the surrounding atmosphere. Figure 4.8-4 is a plot
of the total traffic and sigma azimuth diurnal cures. There is
a great deal of similarity between the two curves with minima and
maxima occurring within one hour of each other.
The Grand Central Parkway is the only site where turbulence pa'ra=- -
meters showed a distinct double peak attributable to the traffic
maxima, during the morning and evening rush hours. Other closed
sites, such as the two tunnels, should also show a very strong
correlation between traffic and sigma azimuth but, as was indi-
cated previously, it was impossible to monitor inside the tunnels.
-------
t>0
Figure 4.8-4 Sigma Azimuth and Total Traffic at Grand Central Parkway
-------
The stgma elevation curve fi.br the medial barrier location is
similar to the sigma azimuth curve. The double maximum is
again present in the morning and afternoon hours. The minimum of
the curve occurs between 0300 and 0400, and a relative minimum
occurs near noon.
Elevation data recorded: by this sensor indicates the center of
the frequency distribution to be near -15ฐ. Only 0.7 percent
of the monitoring hours had elevation angles greater than 0ฐ.
Wind velocities recorded in the cut were quite low. The mean
wind speed for all monitoring hours was 2.5 mph. This compares
with 3.7 mph measured on the service road. The highest hourly
wind velocity recorded was 7.0 mph between 1000 and 1100 on
May 6. Mean wind speed shows no evidence of a traffic-induced
double peak. The maximum occurs between 1500 and 1600 hours,
and the minimum between 0300 and 0400.
-------
4.9.1 Brooklyn Queens Expressway at Park and Navy Streets
4.9.1.1 Site Description
At this point in Brooklyn, the Brooklyn Queens Expressway becomes
an elevated six lane highway. The roadway is^^^30 ft. above
ground, being supported from beneath by heavy steel pillars. The
total width of the structure isx-^/93 ft., having 4 ft. curbs and
a 7 ft. medial strip. Two service roads ran parallel to the ele-
vated highway at ground level. The test location is shown in
Figure 4.9-1.
108,000 vehicles travel on the expressway at this point each day
at an average speed of 38 mph. The northbound lanes carry approxi-
mately 56,000 vehicles per day, reaching an average peak hourly
traffic volume of 3910 vehicles. About 52,000 vehicles travel the
southbound lanes daily. The southbound peak traffic is 3750 vehicles.
The parallel service roads at grade carry an additional 11,500 ve-
hicles per day.
The site is relatively open. An elementary school is located/-/ 200'
to the south of the expressway and a park area is situated to the
north.
4.9.1.2 Instrumentation Arrangement
The G. E. trailer was parked beneath the viaduct between the support
pillars. The sampling probe array used for CO is shown on Figures
4.9-2 and 4.9-3.
Five CO probes were placed above the highway to monitor CO concentra-
tions in a plane perpendicular to the roadway. These highway probes
were 10* and 20' off the surface of the roadway. Other probes were
placed at various vertical and horizontal distances from the roadway
in order to determine CO levels at points remote from the highway.
The open area surrounding the site permitted probes to be placed as
far as 200' from the roadway.
Two wind sensors were placed on the elevated portion of the BQE at
this site. One sensor, vane 1, was located on the probe supports at
the north edge of ,the road. This vane was elevated 20' above the
road surface which, itself, was 27' above the ground. Vane 2 was
located on the medial barrier and 14' above the road surface. Both
wind sensors were in the probe plane. One tape sampler was located
on the roof of the GE trailer while another sampled inside a nearby
hospital at roof level. A high Volume Air Sampler was located on
the roof of the hospital. No Hi Vol samples were taken at the trailer
due to.vandalism of equipment.
4.9.1.3 Meteorological Conditions
The wind monitoring period comprised 17 days from May 18 through
June 3. Reliable data was obtained from most wind channels during
-------
-------
j?Q-
fj
Q I
s4 ,
^ 1
,
s
c
?C-I
tO
'A V/5) *
1 D_ _ _ i
|* so' I
O' c:
^ซ
ป SCWiC^L V*tt^
2ft \
D ' eaO
. . 1 -ปj 7k
r^^ x ' i
-_JJ '^ \ A, ~
J ^ ' ^ / 1
, /^
t) > 1 0 jr
2ป7
v"N8flj
,
3t
o
1 /o'
. . Y t _
- - '
(
_ .
^
s
ft
1^- . . . *7i ' .. .-^-1 1 ^ ^
h _ 6. i o
f~ --ซ-->-
/o ' r^ ^"0/ *J
^ /C/C/ 4,
_
^
. X0-*? M /< /=)/=? f/:
.TV 4 ปป"><ป vซ*4 t
4
>~in
Viaduct Roadway
O
=. \-0
-------
V
"7/i
C,f)
ฉ
d
00
o
0
o- o
0
^
^
Q
c?
Btt
Viaduct
t
\A
U
s
30
29
/G
T ฉ o- e
+/v
=- IM
-------
this period, broken only by short duration interruptions. Some
data was missed in the wind speed channel of vane 1 and the eleva-
tion and sigma elevation channels of vane 2 due to malfunctions
requiring a longer time to correct. The means of the ventilation
parameters, (i.e. wind speed, sigma azimuth and sigma elevation),
were comparatively low at this site, indicative of the approaching
summer weather conditions.
The azimuth angles as recorded by vane 1 at the 20' elevation show
the frequency of occurrence to be spread evenly over all 360 ex-
cept for one spike at 75ฐ (Figure G-89). (A 75ฐ wind in the co-
ordinate system of this site would correspond to a synoptic wind
from 360ฐ or north.) The majority of readings taken at Central
Park during the monitoring period indicate a wind from northerly
quadrants, agreeing with measurements taken at this site.
The mean sigma azimuth values recorded at the 20* location vary
very little over a 24 hour period. The minimum of 6.9ฐ occurs be-
tween 0300 and 0400 hours. The maximum of 11.6ฐ occurs between
1200 and 1300 hours (Table C-113).
The frequency distribution for the elevation angle, as recorded by
vane 1, shows the prevailing elevation to be centered about -5ฐ
(Table C-115). The effect of precipitation on the balance of the
vector vane sensor may be seen by examining the elevation data
from this vane. Rain was reported by the field crew on May 23, 24,
25, and 26. The elevation data for those days showed several hours
with strongly negative readings such as -30.4ฐ at 1100 on the 25th
and -52.6ฐ at both 9300 and 0400 on the 26th. Low wind speeds such
as .8 and .4 mph at 0300 and 0400 on the 26th allowed the wet sensor
to hang with the tail at almost maximum negative elevation.
Sigma elevation recorded by vane 1 had a somewhat limited range of
3.2ฐ in the mean values. The minimum occurred at 0400 and the maxi-
mum at 1500. As expected, because of this sensor's distance from
traffic and openess of the site configuration, the sigma curves bear
little resemblance to the double peaked traffic curves for this lo-
cation.
Wind speed at the vane 1 location was light, with the average for
all hours being,3.3 mph. The highest speed was 10 mph on May 27 at
both 1300 and 1700. The wind velocity plot, as seen in Figure C-93,
is different from most others in that, while the maximum is where
it should be in the afternoon, the minimum occurs about four
hours earlier than expected at 0100.
Wind azimuth as measured by vane 2, on the median, was even more uni-
formly distributed than that measured by vane 1. No 10ฐ interval
contained more than 9 percent of the total observations (Table C-122).
It is likely that a lack of definite and extended duration synoptic
wind patterns during the monitoring period caused this randomness and
not any pecularity of the site configuration.
-------
Sigma azimuth at the median showed a considerable increase from
that measured 20' above the road edge. In the mean values given
in Table C-123, this increase approached 100 percent for most
hours. At least a portion of this increase was caused by nearby
traffic. This is a reiteration of the same traffic induced pattern
found at the Grand Central Parkway and at Bruckner Expressway. The
diurnal pattern is also more evident in the mean values from this
sensor, with the minimum occurring between 0400 and 0500 gnd the
maximum during the evening rush hour. The range was 16.1 or 3 times
the range at the higher vane.
^
The prevailing elevation recorded by vane 2 was -5 . As in the case
of vane 1, very large negative values occur on days with precipi-
tation and light winds such as 0300 and 0400 on May 26.
Sigma elevation computed from the elevation angle fluctuations of the
median vane averages 9.9ฐ for all hours of monitoring. This is 2.8ฐ
or 39 percent higher than that determined from the vane along the
road shoulder. The 39 percent sigma elevation increase compares with
a 113 percent sigma azimuth increase. The sigma azimuth parameter
also increased to a greater degree than sigma elevation at the Grand
Central Parkway site. The increases from the farther to the closer
(to traffic) vane at Bruckner were of the same order. Sigma azimuth
experiences a greater contribution from moving traffic than does a
sigma elevation when the measurement is made on a medial barrier
with opposing traffic on the sides. The opposing traffic lanes
generate a high degree of small scale turbulence between them re-
sulting in highly variable azimuth angles. The effect is mainly in
the horizontal and tends to increase the sigma azimuth parameter
considerably more than sigma elevation.
Mean wind speeds for the median sensor are plotted in Figure C-101.
This curve shows a minimum occurring between 0300 and 0400 and a
maximum between 1600 and 1700. The wind speeds were generally higher
at the 20' vane than at the 14' vane. The difference in the averages
is .4 mph. This decrease of wind velocity from the reference vane to
the roadway vane was also seen at Bruckner Expressway and Grand Central
Parkway sites. Any additional wind velocity generated by passing traf-
fic is too small to compensate for the difference in wind velocities
caused by the elevation difference of the two sensors.
-------
4.10.1 Canal Street between Church and Mercer Streets
4.10.1.1 Site Description
Canal Street, in Lower Manhattan, along with its adjacent buildings,
forms the typical city street site. This street carries a large
volume of traffic, especially trucks, to and from the Holland
Tunnel. During the rush hours, traffic tie-ups are a common oc-
currence. Figure 4.10-1 shows the two way traffic passing the
probing plane at this site.
28,000 vehicles transverse Canal Street daily. Approximately 13,000
vehicles move eastbound while about 15,000 vehicles travel west-
bound. The average daily traffic speed is 16 mph; however, in high
traffic average speed drops to between 5 to 7 mph. The average peak
hourly traffic volumes in both directions occur during the morning
rush hour.
4.10.1.2 Instrumentation Arrangement
The GE Air Pollution Trailer and the New York State Traffic Trailer
were parked parallel, at curbside, on opposite sides of Canal
Street. A probe plane to monitor CO emissions of passing vehicles
was established between the trailers, as shown in Figure 4.10-2
and 4.10-3.
Thirteen probes were used to monitor CO concentration in the probe
plane. Four of these were positioned 20' above Canal Street, one
over each traffic direction and one at the edge of the sidewalk
at each side. One probe was located outside each trailer, 3' above
the street. Outside one building on the south side were positioned
four probes at 8', 27', 42' and 80' levels. On the north side, a
similar building was equipped with 5', 25', and 50' level probes.
One probe was placed inside the building adjacent to the N. Y8
State trailer on the north side at the second floor level.
Two vector vanes were again used to determine the meteorological
characteristics for this configuration. Vane 1 was located ap-
proximately 5 feet above roof level on the north side of Canal
Street. This sensor was on a mast attached to the building fire
escape and, as such, was slightly out over the sidewalk. Vane 2
was located on the southwest corner of the trailer parked on the
north curb of Canal Street. Both vanes were within a few feet of
the probe plane.
A Tape Sampler and a Hi Volume Air Sampler were placed on the roof
of the "GE trailer.
Measured hydrocarbon concentrations were average site values (aver-
age of all probes, including probes 1A and IB at the GE trailer).
-------
CO
N3
-------
7.ACV
oo
u>
fcfl
_O
\
St
80
41
ฐl
&
City Street
-------
^ . Q
N
1
A
1 f\
63
CANAL
0
1
A
/ / / / /
City Street
/ /
-------
4.10.1.3 Meteorological Conditions
The monitoring performed at this site was divided into two
periods: June 10 to June 15, and June 23 through July 2.
Weather conditions before and after the interruption in monitor-
ing were essentially the same. These conditions were generally
characterized by low wind velocities and low turbulence.
The frequency distribution .of the azimuth angle at the roof level
(Figure C-105) shows two peaks similar to the distribution re-
corded at the deep and shallow cut sites. The two peaks are
near 175ฐ and 15 , or easterly and westerly winds, respectively.
(0ฐ at this site is a wind exactly parallel to the road and
blowing from the west.)
Sigma azimuth averaged for all hours at the roof location was 10.8C
one of the lowest averages of any site. The plot of hourly mean
sigma azimuth shows this turbulence parameter to be virtually con-
stant over the average 24 hour day. The range is only 4.7 .
Wind elevation angles recorded at the roof level were generally
negative indicating descending air. The prevailing elevation
direction was -15ฐ.
Sigma elevation at roof level was virtually constant over the
average 24 hour day as shown in Figure C-108. The minimum occurs
near 0500 and the maximum near 1400. The average of 6.8ฐ is lower
than any other site with the exception of 54th and Sutton Place.
This is a reflection of the settled, summer type weather which
predominated at this site.
The mean wind speed of 2.9 mph was also among the lowest of any
site. The diurnal wind speed pattern shows a minimum near 0600
and a maximum near 1500.
A bimodal distribution is again evident in the frequency of oc-
currence histogram (Figure C-113) for the vane located on the
trailer parked at curbside on the north side of Canal Street.
The peaks are at 5ฐ and 165ฐ which indicates a counterclockwise
displacement of about 10ฐ from the sensor at roof level. This
displacement may not be real however, since 10 is within the
estimated error involved in orienting the sensor.
Sigma azimuth was higher at the roof level than in the street
but only by about 1.2ฐ. The diurnal pattern for the vane in the
street is also flat with a weak maximum occurring at 1700. There
was apparently only a negligible contribution to sigma azimuth
by passing vehicles because the turbulence parameter curve is
unlike the traffic curve and the values of sigma azimuth de-
creased from the roof to the road. It is likely that traffic
velocity at this site was so slow that any turbulence generated
by passing vehicles was too small to be detected.
-------
The same discussion may be applied to the sigma elevation
turbulence parameter. Although this quantity does increase
from roof to road, the increase is so small (.4ฐ) as to be
insignificant. In addition, the diurnal curve is so flat
that the minimum and maximum are practically unrecognizable.
Although traffic patterns are different than at any other site,
there is still at least one pronounced maximum in the morning.
This maximum is not reflected in the sigma elevation curve of
the wind sensor closest to the traffic.
Wind speed in the street averaged 2.7 mph over all monitoring
hours. This is .2 mph less than at the roof level. The wind
speed curve has essentially the same shape as the curve for the
roof level vane.
Canal Street may be considered an open cut from a meteorological
standpoint since the azimuth angle is restricted to a bimodal
distribution. The normal increase in turbulence expected from
moving traffic, however, is missing probably due to low
vehicular velocities.
-------
4.11.1 Trans Manhattan Expressway at George Washington Bridge Plaza
4.11.1.1 Site Description
George Washington Bridge Plaza is formed by four hi-rise apartment
buildings built directly over the Trans Manhattan Expressway in
upper Manhattan. The expressway is the main traffic artery which
connects upper Manhattan and the Bronx with New Jersey. The two
apartment building complex between St. Nicholas Ave. and Wadsworth
Ave. on 179th Street was the structure directly involved in this
study. Beneath the buildings were twelve (12) lanes of two-way
traffic. A large vent area open to the roadway was located be-
tween the buildings. This vent area was 100" long and 180' wide.
The vertical distance from the roadway up to the street level
(179th St.) was /-^~- 45'. The test site is shown in Fig. 4.11-1.
The Trans Manhattan Expressway carries 175,000 vehicles per week-
day. Two parallel city streets carry an additional 15,000 vehicles
each weekday. Approximately 87,500 travel the expressway each way
at an average speed of 38 mph. The eastbound average hourly peak
of 6,250 vehicles occurs during the morning rush hour. The west-
bound peak during the evening rush hours reaches 6,140 vehicles.
4.11.1.2 Instrumentation Arrangement
The GE Air Pollution Van was parked on West 179th Street, approxi-
mately 12' from the north edge of the Trans Manhattan Expressway.
16 probes were used to monitor the CO concentrations of the road-
way and to bbtain indoor-outdoor CO data. Initially, five sets of
three probes each were positioned as shown in Figs. 4.11-2 and
4.11-3. These probes were 3', 29' and 50* above the roadway. Two
of these probes (5A & 5B) were relocated on July 21 to an apartment
building adjacent to the vent area. One probe was installed inside
a 3rd floor apartment s~**-s 60* above the top of the expressway and
the other outside the same apartment. On July 27, these probes
were moved to a 9th floor apartment /~^100' above the expressway.
These indoor-outdoor probe locations were 32' east of the main
probing plane and over the north-most lane of the expressway. The
sixteenth probe was placed 50' west of the other probes along the
north wall, 29' above the roadway.
i
One vector vane system was in operation at the George Washington
Bridge site. This vane was designated number four and is shown as
such in Appendix C. The vector vane sensor was located beneath the
apartment structure and approximately 15* east of the end of the
"tunnel" formed by the roadway passing under the building. It was
located on the divider separating the six eastbound lanes into two
groups of three and was 14* above the road surface.
Hi vol samples were taken from the trailer roof and the outside
balcony of an apt. (23rd floor) in the adjacent building.
-------
00
oo
-------
oo
sr.
GE
O
.SB -5*
O O
O
O-
/? -p ~t~. >3 ui<- a nJ &
eo
(K
rv
Vt'-
43
14
0
o
_!
T
Intermittently Covered Roadway
Fig. 4,11-2 Trans-Manhattan Expressway at George Washington Bridge Plaza - Elevation View
\
O- 1>vo-J-
-------
,
o ซ
VI
\
M
%
X
Q
k
*
k
SA
t
\
i
*
sa
oio
3(3
40' yl
'
J50'
3 3fป
G
J
5
C2
V
V
f,
H
Fig. 4.11-3 Trans-Manhattan Expressway at George Washington Bridge Plaza - Plan View O
X
o
7
Q
tO"J
-------
A tape sampler sampled the air adjacent to the trailer. A
hydrocarbon analyzer, inside the trailer, sampled all probes
and recorded an average concentration for the site.
4.11ol.3 Meteorological Conditions
The meteorological monitoring period at this site encompassed
the days between July 16 and August 5. General meteorological
conditions were again quite calm in comparison with earlier
sites. No usable azimuth data was obtained due to a malfunction
in the transmitter electronics. The missing azimuth data, of
course, precludes any sigma azimuth data.
Elevation angle data was available for most of the period. The
elevation frequency of occurrence histogram shows over 98.5% of
all recorded data was in the range of 0 - 10 . This is as would
be expected since any wholesale vertical movement is effectively
prevented by the roof over the road.
Sigma elevation is very constant over the average 24 hour day with
a range of only 1.6ฐ. At first, this may seem surprising con-
sidering the experience at the partially covered Grand Central
Parkway site. The explanation is that the sensor at this site was
placed between two streams of traffic moving in the same direction
while at the cantilever site, the sensor was between opposing
traffic lanes.
-------
-------
5.0
HIGHWAY PLANNING FACTORS
The purpose of this section is to apply the Superposition model
directly to a hypothetical roadway in order to ascertain whether the roadway
will produce carbon monoxide levels at the receptor locations that are in
excess of the National Air Quality Standards.
Problem //I
Given: A four lane highway (Figure 5-1) carrying traffic volumes
of 3000 vehicles/hr. at an average speed of 30 miles/hr.
in one direction and 4200 vehicles/hr. at an average speed
of 10 mph traveling in the other direction. The proposed
highway is to be located such that the nearest receptor
will be 75 ft. from the roadway.
Solution: Equation 3.3-6 is employed to determine (h for each side
of the highway.
d) . = [-.54 (T ) + 29.6] X 10~3
', = [-.54 X 30 mph + 29.6] X 10"3 = 13.4 X 10~3 PPm~hr
rl veh.
ฃ = [-.54 X 10 mph + 29.6] X 10~3 = 24.2 X 10~3 PPm~hr
~2 veh.
Equation 3.3-7 is employed to determine the CO concentration
at receptor locations 1 and 2. The concentration at
receptor 1 is computed as follows:
2
[CO] = (h N. exp [-ฃ
Rl / rฑ i fฑR
i = 1
= (13.4 X 10~3 ppm-hr. . 3000 veh.) . exp [-.02 ft.'1 . 87 ft.] +
veh. hr.
(24.2 X 10~3 ppm-hr. . 4200 veh.) . exp [-.02 ft.'1 . Ill ft.]
veh. hr.
= [40.2 . exp (-1.74) +101.6 . exp (-2.22)] ppm
= [40.2 (.175) + 101.6 (.108)] ppm
= (7.0 + 11.0) = 18 ppm
-------
vo
CO
DWELLINGS
(ซ)
1
v,umpuueu v^unctjntration
18 ppm ! |
75 r
42'
fclosest dwelling location where
hourly AQ Standard can be met
12'
J l ^ 30 mph
Line Source ฎ
36'
36'
12
4200^vehicles/hr.
10 mph
r
Line Source
^> __ -L
-2
75*
Computed Concentration
=
DWELLIN(
Proposed Highway to be evaluated for Air Quality
Figure 5-1
-------
The concentration at receptor 2 is computed as follows:
[CO] = (24.2 X 10~3 ppm-hr. . 4200 veh.) . exp [-.02 ft."1 . 87 ft.] +
R2 veh. hr.
(13.4 X 10~3 ppm-hr. . 3000 veh.) . exp [-.02 ft."1 . Ill ft.]
veh. hr.
- 101.6 exp (-1.74) + 40.2 exp (-2.22)] ppm
= [101.6 (0.175) + 40.2 (.108)] ppm
= (17.8 + 4.3) = 22.1 ppm
Therefore, it is seen that if this traffic volume and speed persists
for a minimum 4 hours and then drops to zero, the National Air Quality Standards
would be exceeded.
Problem #2
At what distance from each side of the highway described in
Problem #1 will the CO concentration equal 35 ppm-max. 1 hr.
Solution: Equation 3.3-7 is employed, by trial and error to determine
the distance from the edge of roadway at which the concentration
at the receptors, R-^ and R2> will equal 35 ppm.
[35] = (13.4 X 10-3 ppm-hr. . 3000 veh.) . exp [-.02 ft.'1 . (12 + X,)] +
KJ. , _ -*-
veh. hr.
(24.2 X 10~3 ppm-hr. . 4200 veh.) . exp [-.02 ft.'1 . (36 + X )]
veh. hr.
[35] = (24.2 X 10~3 ppm-hr. . 4200 veh.) . exp [-.02 ft.'1 . (12 + X )] +
R2 veh. hr. 2
(13.4 X 10~3 ppm-hr. . 3000 veh.) . exp [-.02 ft.'1 . (36 + X )]
veh. hr>
It 'is found that X-j^ = 42 ft.
and X2 = 52 ft.
-------
Problem #3
It is proposed that a 6 lane highway, 100 feet wide, with
a cantilever cover over the outside lanes be constructed.
The cover will be 10 ft. above road level at each edge.
It is estimated that peak traffic volume will be 8500 cars/
hr. (no truck traffic permitted). The rush hour traffic
peak will be 6000 cars/hr. inbound in the morning and out-
bound in the afternoon. What, CO concentration can be
expected on the roadway?
\
Solution: Using Figure 2.2-7, the CO concentration at the 3 ft. level
on the median strip for 8500 vehicles/hr. is found to be
28.5 ppm. At the 10 ft. level, for 8500 vehicles/hr. the
CO concentration is 20 ppm.
The horizontal concentration profile in Figure 2.1-13B is
used to extrapolate from the middle of the road to the edge
of the road. From Figure 2.1-8B, it can be seen that the
two-way traffic shown during the 7 - 8 AM period approxi-
mates the anticipated AM rush-hour condition. Therefore,
assuming the westbound lanes are inbound, a line from 20
ppm at y = 50 feet with a slope which is halfway between
that shown for the 6 - 7 AM and 8 - 9 AM curves, indicates
approximately 32 ppm at 100 feet. Similarly, a 22 ppm
concentration is expected at the edge of the outbound lanes.
Problem #4
It is expected that southbound traffic on the F.D.R. Drive
will increase by 500 vehicles/hr. and northbound traffic
will increase by 300 vehicles/hr. What will be the anti-
cipated CO concentration along the west wall at rush-hour
periods?
Solution: Using Figure 2.1-8A, it is seen that the morning traffic
peak occurs between 8 - 9 AM and the afternoon between
4-5 PM. The morning traffic consists of 3300 vehicles
southbound and 2700 vehicles northbound, or a total of
6000 vehicles. The anticipated morning peak, due to
increased two-way traffic, is 6800 vehicles. Using the
z = 3' curve of Figure 2.2-1, it is seen that 83 ppm CO
can be expected for a two-way traffic flow of 6800 vehicles.
By a similar process, the anticipated afternoon peak of 7600
vehicles will produce approximately 99 ppm CO. (Note that it
is necessary to project the z = 3' curve)
-------
Problem #5 /
/'
Determine the anticipated CO concentration at the edge of
the westbound lane of the Viaduct road at noontime.
Solution: From Figure 2.1-9B, it can be determined that the total
traffic is approximately 4600 vehicles/hr. Using Figure
2.2-8, it is found that the CO concentration on the median
strip at the 10 foot level is approximately 10.2 ppm.
Using the horizontal profile curve of Figure 2.1-4A, it
is found that the concentration decays from 10.2 ppm in
mid road at the 10 foot level, to about 6.8 at the edge
of the westbound lane. The vertical profile shows a
decay from 6.8 at the 10 foot level to about 6.0 ppm at
the 0 foot level.
------- |