United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-78-034
August 1978
Air
Carbon  Monoxide
Hot Spot
Guidelines
Volume  II: Rationale

-------
                                            EPA-450/3-78-034
Carbon Monoxide Hot Spot Guidelines
               Volume II:  Rationale
                             by

                         Frank Benesh

                        GCA Corporation
                      GCA/Technology Division
                        Burlington Road
                    Bedford, Massachusetts 01730
                      Contract No. 68-02-2539
                  EPA Project Officer: George J. Schewe
                          Prepared for

               U.S. ENVIRONMENTAL PROTECTION AGENCY
                   Office of Air, Noise, and Radiation
                Office of Air Quality Planning and Standards
               Research Triangle Park, North Carolina 27711

                          August 1978

-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers.  Copies are
available free of charge to Federal employees, current contractors and
grantees,  and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35),  U.S. Environmental  Protection Agency,
Research Triangle Park, North Carolina 27711; or, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
This report was furnished to the Environmental Protection Agency by
CCA Corporation, CCA/Technology Division, Burlington Road, Bedford,
Massachusetts 01730, in fulfillment of Contract No. 68-02-2539. The contents
of this report are  reproduced herein as received from CCA Corporation.
The opinions,  findings, and conclusions expressed are those of the author
and not necessarily those of the Environmental Protection Agency.  Mention
of company or product names is not to be considered as an endorsement
by the Environmental Protection Agency.
                    Publication No. EPA-450/3-78-034

-------
                               ABSTRACT

This report presents the rationale used in developing the analytical tech-
niques for the carbon monoxide hot spot guidelines.

Discussed in this report are the technical aspects of the guidelines, in-
cluding the assumptions used in developing hot spot procedures.   Since the
guidelines were based largely on EPA's Indirect Source Guidelines,  these
are discussed in some detail, as well.
                                iii

-------
iv

-------
                               PREFACE

This document is the first in a series comprising the Carbon Monoxide Hot
Spot Guidelines.  The purpose of this series is to provide state and local
agencies with a relatively simple yet accurate procedure for assessing
carbon monoxide hot spot potential on urban street networks.  Included
in the Hot Spot Guideline series are:

    Volume I:    Techniques
    Volume II:   Rationale
    Volume III:  Summary Workbook
    Volume IV:   Documentation of Computer Programs to Generate Volume I
                 Curves and Tables
    Volume V:    Intersection-Midblock Model User's Manual
    Volume VI:   Modified ISMAP User's Manual
    Volume VII:  Example Applications at Waltham/Providence/Washington, B.C.

Hot spots are defined as locations where ambient carbon monoxide concen-
trations exceed the national ambient air quality standards (NAAQS).   For
both the 1-hour and 8-hour averaging times the assumption is made through-
out these guidelines that a CO hot spot is primarily affected by local
vehicle emissions, rather than areawide emissions.  Studies have shown
that for the 1-hour CO concentration, local sources are the dominant
factor.  Accordingly, representative urban worst-case meteorological,
traffic, and background concentration conditions are selected as those
corresponding to the period of maximum local emissions — usually the
period of peak traffic.  For 8-hour concentrations evidence indicates
that neither the local nor the areawide contributions can be assumed to
be dominant in every case.  However, for the purpose of analysis discussed
in these guidelines, local source domination of CO hot spots is assumed

-------
for 8-hour averages.   This allows some consistency between assumptions in
relating the 1-hour and 8-hour CO estimates.
                                   vi

-------
                               CONTENTS

                                                                   Page
Abstract                                                           ill
Preface                                                            v
List of Figures                                                    viii
List of Tables                                                     ix
Acknowledgments                                                    x
Sections
I      Introduction                                                1
           Introduction                                            1
II     Development and Assumptions of the Verification Procedure   3
           Introduction                                            3
           Derivation of the Normalized Concentration Curves       5
           Derivation of Street Canyon Curve                       21
           Derivation of Emission Rates                            22
III    Development of the Hot Spot Screening Process               34
           Introduction                                            34
           Assumptions Regarding Background Concentrations         36
           Uninterrupted Flow Conditions                           39
           Interrupted Flow Conditions - Signalized Intersections  43
           Interrupted Flow Conditions - Nonsignalized
             Intersections                                         46
           Effects of Variation in Parameter Assumptions           47
IV     References                                                  52
                                vii

-------
                              FIGURES

No.                                                                 Page

 1   Intersection Geometry                                             8

 2   Normalized CO Concentration Contribution from Excess              9
     Emissions on Approach 1

 3   Variation of the Normalized CO Concentration with Roadway        10
     Length, Road/Receptor Separation, Stability, Wind/Road Angle

 4   Normalized CO Concentration Contributions from Excess            12
     Emissions on Approaches 2, 3, and 4

 5   Intersection Geometry - Crossroad                                13

 6   Distance Correction Factor for Excess Emission Contributions     15
     at Intersections

 7   Normalized CO Concentration Contributions from Free-Flow         16
     Emissions on Each Lane of Roadways at Intersections
  8   Values of Xu/Q  (lO"^"1) for Various Roadway/Receptor            17
     Separations and Wind/Roadway Angles; Infinite Line Source

  9   Distance Correction Factor for Free-Flow Emission Contri-        18
     butions at Intersection Locations

 10   Normalized CO Concentration Contribution from Each Traffic       20
     Lane at Locations of Uninterrupted Flow

 11   Examples of Effects on Screening Curves of Variations in         50
     Parameter Assumptions
                                Vlll

-------
                             TABLES

No.                                                                 Page

 1   Assumed Operating Speeds, Levels of Service and Demand-         42
     Capacity Ratios for Major Streets and Corresponding Emission
     Factors for Free Flow Conditions

 2   Assumed Operating Speeds, Levels of Service and Demand-         42
     Capacity Ratios for Urban Expressways and Corresponding
     Emission Factors for Free Flow Conditions

 3   Assumptions for CO Concentration Computations                   48

 4   Variation in Estimated Parameter Values for Sensitivity         49
     Analysis

 5   Changes in Parameters versus Changes in Allowable Volumes       51
                                IX

-------
                           ACKNOWLEDGMENTS
We wish to acknowledge the significant contributions made early in the
development of the Hot Spot Guidelines by previous GCA/Technology Division
staff members, including Dr.  Robert Patterson, Messrs.  David Bryant,
Alan Castaline, and Walter Stanley.  We are especially  indebted to the
EPA Project Officer, Mr. George J. Schewe of the Source Receptor Analysis
Branch, who provided overall project direction and performed extensive
technical and editorial review of the final reports.
                                 x

-------
                               SECTION  I
                               INTRODUCTION
 INTRODUCTION

 This  volume,  Volume  II,  is  part of a seven-volume report providing guide-
 lines for  identifying  and analyzing carbon monoxide "hot spots"-locations
 with  the potential for experiencing violations of the National Ambient
 Air Quality Standards  (NAAQS)  for CO.  These guidelines are intended for
 engineers, planners, and others who must consider the air quality effects
 of traffic management  decisions and who are responsible for traffic plan-
 ning  to control  CO hot spots.

 The guidelines present two  levels for screening potential hot spots.  One
 is a  screening procedure to  identify potential carbon monoxide hot spots
 using only data  on automobile  traffic volumes, thus obviating time-
 consuming and costly monitoring of air quality at potential hot spots.
 The other is  a hot spot  verification procedure that uses more detailed
 input  data.  This procedure  provides the capability of accounting for a
 number of additional conditions beyond those assumed in the screening
 procedure, and it provides a worst case quantitative estimate of hot spot
 potential.

 Volume I  discusses in detail  the concepts of hot spot screening and
 verification as well as  providing analytical techniques and procedures.
Two companion volumes, this  volume and Volume III, present the technical
rationale behind the guidelines and a workbook.  Volumes IV, V and VI

-------
provide documentation respectively on an intersection midblock traffic
and dispersion model, the IMM; a traffic assignment and dispersion model,
the modified 1SMAP model; and the computer programs used to generate the
curves and tables in Volume I.  Volume VII presents the results of sev-
eral application studies.  The remainder of this volume describes the
background and techniques in Volume I.

Section II provided a discussion of the rationale for developing the
verification procedures described in Volume I.  Also provides are discus-
sions of the technical aspects of the guidelines including traffic,
emission, and dispersion used and the basis for the assumptions.  The
emphasis is on how the various nomographs and tables were constructed
rather than on how to apply the procedure.  A similar discussion is
provided in Section III concerning development of the screening procedures,

-------
                               SECTION II
                   DEVELOPMENT AND ASSUMPTIONS OF THE
                        VERIFICATION PROCEDURE
INTRODUCTION

The verification process in Volume I is a followup to the initial screen-
ing of an urban area.  The intent is to perform a more thorough evaluation
of the hot spot potential of a street section or intersection using a tech-
nique that permits  input of parameters specific to that location rather
than assumed parameters.  While the initial screening process focused on
identifying potential hot spot locations anywhere within a city or town
(thus requiring a very general approach), the verification process involves
analysis of specific locations, and a more detailed analysis of each loca-
tion is feasible.   Since the screening curves are developed from the same
methodology as the  verification procedure, using a more detailed set of
assumptions, the development of the verification procedure is discussed
first.

                          3 4
During an earlier project, '  analysis of CO hot spots was initially en-
visioned as a two-step process:  screening and detailed dispersion modeling.
Development of the  hot spot screening guidelines, originally intended to be
the only hot spot guidelines, involved many assumptions and generalizations
in order to achieve the simplicity that was desired.  In these original
guidelines the assumptions were such that the screening was thought to be
overly conservative and thus limit their utility.  Therefore, the test
cases in Waltham, Massachusetts of the Hot Spot Guidelines were reanalyzed
utilizing the Indirect Source Guidelines,  ISG (1975) to check the degree
of conservativeness.

-------
 The application  of  the  Indirect Source Guidelines to the Waltham study
 locations  led  to the  following conclusions:.
     •   The hot  spot  screening guidelines are not overly
         conservative, as demonstrated by the fact that the
         screening guidelines identified only a few more
         potential hot spots in the Waltham case study than
         did the  application of the more detailed Indirect
         Source Guidelines.

     •   The Indirect Source Guidelines are a workable method
         for analysis of potential hot spots,  and allow the
         use of more data specific to conditions at individual
         locations.  Results are quantitative, rather than
         qualitative.
     •   The Indirect Source Guidelines can,  in some cases,
         be used for assessment  of alternative improvement
         measures, in lieu of detailed computer modeling.

 These  conclusions led  to a recommendation for using the Indirect Source
 Guidelines  as  the foundation for  a second,  verification stage of hot  spot
 potential.  However, the Indirect  Source  Guidelines  have  been revised2
 and  they (the  original Guidelines  )  are no longer suitable for the
 level  of analysis desired  for hot  spot verification.   The Revised In-
                         2
 direct  Source  Guidelines  allow much more latitude in the selection of
 values  of  input variables  than did the original Guidelines,  and their
 application is  similarly much more complex.   In developing revised veri-
 fication procedures  based  on the  Revised  Indirect Source  Guidelines,  the
 intent  was  to  deep the requirements  on the user at about  the same level
 as in the old  verification  method.
The remainder of this  section discusses how  the new verification proce-
dures were derived using the Revised  Indirect Source Guidelines, the
                                                 12
Automobile Exhaust Emission Modal Analysis Model,   and the Compilation
of Air Pollutant Emission Factors (AP-42).   In particular, three specific

-------
Assumptions


Before we begin a detailed discussion of the derivation of the normalized

concentration curves, used in Volumes I and III, we list a number of the

more important assumptions upon which these curves are based.  The ratio-

nale for these assumptions are presented in the  following sections.

    •   Queue lengths for each of the four approaches to the
        intersection are assumed to be equal to  the queue
        length on the approach adjacent to the receptor (designated
        as approach 1).  Criteria for the selection of a given
        intersection approach for the placement  of a receptor
        are presented in Volume I.  It must be remembered that
        lane volumes are used in the calculation of queue
        lengths, but approach volumes are used to calculate
        excess or free flow emissions.

    •   Each of the two intersecting streets are 10 meters wide.

    •   The distance from the receptor to the centerline of
        approach 1 is 10 meters.

    •   Consistent with worst case urban conditions, the
        atmospheric stability is chosen to be category D,
        windspeed is assumed to be 1 m/sec, and  the initial
        vertical dispersion (°"7n) is assumed to  be 5 meters.

    •   The optimum wind angle and location of the receptor along
        approach 1 (at a perpendicular distance  of 10 m from the
        approach centerline) will depend upon the queue length
        calculated for this approach.  The resultant wind angle
        and receptor location is then used for the calculation
        of normalized concentration contributions for excess
        emissions from the three remaining queues and the free
        flow emissions from all four approaches.


Intersections - Queueing Vehicles
The main technical problem in deriving the hot spot verification procedures

was to develop curves relating traffic parameters to normalized CO con-

centrations at the critical receptor location for the interrupted flow,

or intersection, case.  The location of the point of maximum concentra-

tion varies along a line parallel to the roadway depending on the queue
length and the wind angle.

-------
Figure 1 shows the intersection configuration for the determination of
the receptor location and wind angle  (6).  For the discussion that fol-
lows, all measurements  in the direction of the approach in question will
be taken from a reference plane located 20 m behind the receptor.  For
a wind perpendicular to the road, the reference plane establishes the
extent of the line source emissions that significantly affect the concen-
tration at the most distant receptor  considered.  Its use will become
apparent in the following discussion.  As shown in Figure 1 the distances
from the reference plane to the downwind and upwind boundaries of the
queue are Yd and Yu respectively, so  that:

                              Yu = Yd + Le

where Le = queue length (m).

                               -v                 *
Figure 2 depicts the normalized  concentration (X ) contribution from
approach 1 as a function of effective queue length, Le, on that approach.
It is assumed that all  lanes comprising approach 1 develop queues of
equal length.  The CO concentration at the receptor site is maximized
when the wind angle 6 is such that the contribution from the nearest
lane is maximized.  Figure  2 was developed from Figure 3 which is taken
                                           2
from the Revised Indirect Source Guidelines  (a   = 5m, stability class D).
                                              f-i\J
Figure 3 is based on sequential runs  of HIWAY^^ and treats queuing vehicles
                          *
as finite line sources.  X  is a function of the distance from the receptor
to the emission source(x), the roadway/wind angle (6), the length Yu, and
                 *        / *         *    \
the length Yd.  X  equals ^X (Yu) - x (Yd))) and Yu equals (Le + Yd) where
Le is the effective queue length.  For a given queue length Le and dis-
          -k
tance x, X  is maximized when the following condition is satisfied:
                                  *
                                dX
                *
 that  is,  when X  is  at a peak value in the curves in Figure 3.
 Normalized with respect to wind speed and emissions,  such that X  = x  U
                                                                  e    e

-------
                                                            LEG  4
 LEG  I
                                                             LEG  3
Reference Plane
             Yu -  Distance from reference plane  to upwind end of queue (m)
             Yd •  Distance from reference plane  to downwind end of queue (>0)
                  (Distance is denoted positive  to windward) (m)
             iY -  Distance between receptor  and  reference plane (20m)
             Le •  Effective excess emissions length  (m)
              V •  Wind vector
              6 •  Wind/roadway angle (acute)
              x •  10m
                          Figure 1.    Intersection geometry

-------
•—
T
i
10
1
e
-XI
a
c
t
c
H
:
«.
i
V
«
i
ii
«
«
u
1.
>
b
i
<
•DO
800
1
> T60
E
>
i TOO
i
!
, 660
i
>
i
600
>
i
i
| 660
>
o
l 600
3
h-
£ ...
1
1
U
i
3
a 400
a
*
j
t
c 360
e
300
260
200(


































/
/
















/
















/
















/

















1
















,

















'














/
/
t















1
















/
















/















1
f















/















,
f
















1
















/

































9 20 60 100 200 300 600 10
                          QUEUE LENGTH , L, (m)
Figure 2.  Normalized CO concentration contribution from excess
           emissions on approach 1

-------
      (A)   ROAD/RECEPTOR SEPARATION - 10.0
           SIGMA  ZO = 5.0 m  STABILITY - D
(B)   ROAD/RECEPTOR SEPARATION - 15.0
     SIGMA  ZO • 5.0 m  STABILITY - D
(C)   ROAD/RECEPTOR  SEPARATION = 20.0 m
     SIGMA ZO - 5.0 m  STABILITY  =- D
  1000
  1000
  800
      (D)  ROAD/RECEPTOR SEPARATION - 40.0 m (£)   ROAD/RECEPTOR SEPARATION = 80.0 m  (F)   ROAD/RECEPTOR SEPARATION •= 160.0 m
          SIGMA 2O = 5.0 m  STABILITY - D         SIGMA ZO = 5.0 m  STABILITY  = D          SIGMA ZO = 5.0 m  STABILITY - D
  600
o
p
6
  400 -
a
   200
      0    10  20  30  40  50  60  70  80  90  0  10  20  30  40  50  60  70   80  90
            WIND/ROAD ANGLE — degrees              WIND/ROAD  ANGLE — degrees
                                           10  20  30  40  50  60  70  80  90
                                             WIND/ROAD ANGLE — degrees
                                                                    SA-4429-15
     Figure 3.   Variation  of the  normalized  CO concentration with roadway  length,  road/rectptor separa-
                  tion,  stability,  wind/road angle

-------
In terms of Figure 3a, one would move the receptor downwind of the queue
(subject to the constraint of a 10 m road receptor separation), thereby
decreasing the wind/road angle and increasing both Yu and Yd.  This max-
imization procedure is actually a two-step process in which the distance
Yd and therefore Yu are first selected and then the wind angle which
            i\        if
maximizes (X (Yu) - X (Yd)) is determined.  A larger value of Yd is then
obtained.  In practice, a value of Yd of approximately 20 m was found to
          if
maximize X  in most cases.
          e
                                   *
Figure 4 is used to determine the x  contributions from legs 2, 3, and
4 of the intersection.  This graph was also developed from Figure 3.
This curve was formulated by assuming the road center/receptor seperation, x
of the leg 2 approach is 15 m and    Yu      =    Yu      + Le + 10 m.
                                  approach 2   approach 1
Inherent in the latter assumption is the fact that the queue which de-
velops on approach 2 is the same length as that which develons on
approach 1 (Le).      Yd     =     Yu     - Le, and 6 is the same as de-
                  approach 2   approach 2
termined previously for approach 1, at the corresponding Le.  Hence, all
                             *
variables are specified and x  for approach 2 can be derived from Figure 3(B)

The procedure is slightly different for determining the excess emissions
from the crossroad, approaches 3 and 4.  The reference plane must be
rotated 90° to yield the intersection geometry depicted in Figure 5.

The actual receptor location (where the contribution from approach 1 is
maximized) remains unchanged, as does the wind direction.  However, the
roadway/wind angle 0' is 90° - 6.  Yu = 27.5 m and Yd = 0 for approach 3.
The roadway/receptor separation distance x =     Yu     - 12.5 m.  Thus,
                                             approach 1
x varies as Le changes on approach 1.
                                 11

-------
IO

CJ

CO
kl
8
0.
a
CO
10
CO
CO
o
X
o
ae.
u.
 ffi
 K
 O
 O
 H
 ac
 o
 u
 a
 UJ
 Z
 K
 O
     ISO



     140



     130



     120



     110



     100



?E    90
 o

^_«   eo



-•—-  TO



      60



      50



      40



      30



      20



      10
                          50          100    190   200

                      EFFECTIVE  QUEUE LENGTH, L, (m«t«rt)
                                                           50O
Figure 4.  Normalized CO concentration  contributions from excess
            emissions on approaches 2, 3,  and 4
                             12

-------
                      LEG 4
     LEG I
    RECEPTOR
  REFERENCE  PLANE
                     LEG 3
          XrYu, -12.5
                                            V
                                  LEG 2
(o) APPROACH 3  CONTRIBUTION
Yu =27.5
                                       Yd =0
                                                       X=Yu-l7.5
                LEG  I
                                                         -Le
                                                                     LEG 4
•J RECEPTOR
             REFERENCE  PLANE
                                                                    LEG 3
                                               LEG  2
                                                  . 5
            (b) APPROACH  4 CONTRIBUTION
                        Figure 5.   Intersection geometry  - crossroad

-------
                 *                         o
For approach 4, x  is calculated at 0' = 90  - 9, Yu = Le + 37.5 m,
Yd = 37.5 m, and x =     Yu     - 17.5 m.  Note that slight errors will
                     approach 1
result for crossroad flow configurations and lane widths which would yield
values of x that differ from that of the estimated values.  However, this
                                         *
error will be extremely small since the xo value at large 9 changes very
                                  *
slowly with changes in x; i.e., dxe (9) « Q at e>45°
                                  dx    ~
                                                              *
Figure 6 gives the distance correction factors (Cxe) for the  X  terms
previously determined.  These curves were derived by calculating
the X  values in the HIWAY model at road/receptor separation  distances
other than 10 m, thus at different optimum wind angles.  The  ratio of
these X  values to X  at x = 10 was plotted.  The curves are  given for
       e            e
queue lengths only up to 180 m, since longer queue lengths produce curves
that vary only slightly from the 180 m queue length curve.
 Intersections - Through Vehicles (Free-Flow)

 In making CO concentration estimates at a receptor near an intersection,
 the contribution from free-flowing vehicles, that is, those that do not
 stop, on each leg must also be considered.

 Figure 7 from the Volume 1 procedures depicts the normalized concentration
 contribution from free-flowing traffic, Xf  as a function of queue length
 Le.  Actually X  is a function of the 6 which maximizes the excess emis-
 sion contribution, X   at that Le.  Excess emissions contributions are
                    e,
 generally higher than free-flow, hence the use of the  6 and Le that max-
 imizes X .  Figure 8, from the Revised ISG, was used in generating the
 curves by taking a given Le and the associated roadway/wind angle, 0,
 as determined previously in constructing Figure 3, and finding xf-  Curves
 are plotted for both the main road and crossroad at each intersection.
Distance correction factors similar to those derived in Figure 6 are
shown for free-flow traffic contributions at intersections in Figure 9.
                                  14

-------
10
    20
        30  40
                 50
                     60  70  80   90   100  110   120  130
                      x  ROAD/RECEPTOR  SEPARATION, meters
                                                       140  ISO  160   170  ISO
    Figure 6.  Distance correction factor for excess emission
                contributions  at intersections

-------
 in
 z
 g
 
 in

 i
 UJ



 o
 u
 UJ
 ir
 b.



 O
 a.
 g>
 ir
 o
 o
 z
 o
 o
 cc
 o
 z
700
600
500
400
300
250
200
150
100
90
80
70
6O
50
40
30
20
10






















MAIN ROAD
^
^

CROSSR













^^


OAD













^
x



L__^












^




~~—~.












^





— --










^






— •— .










/






	 ,










"""






~ —









^








' — - — —








*••'








—







0 20 30 40 50 60 70 8090100 150 2C
EFFECTIVE QUEUE LENGTH, Le ( meters )
Figure 7.  Normalized CO concentration contributions  from free-flow

           emissions  on each lane of roadways at  intersections
                              16

-------
          NEUTRAL STABILITY (D)
10


Figure 8.
20     30   40 50 60 70 80 90 100        200
   ROADWAY/RECEPTOR  SEPARATION __m
Values of
                    "1
             (HT"1) for various roadway/receptor
           separations and wind/roadway angles; infinite line
           source
                         17

-------
00
                       O
                       (J
                          1.0
                       
-------
Uninterrupted Flow

Curves for concentration estimates from uninterrupted flow roadways
(expressways, midblock locations, etc.) in Volume I are derived similarly
to the above from the infinite line source curves (Figure 8) in the
                                   2
Revised Indirect Source Guidelines.   Figure 10 gives the normalized
                                                                     *
concentration contributions  from each traffic stream.  In this case X^
is a function of the road/receptor distance, x, and the road/wind angle,
0, and are taken from the maximum concentration estimates in Figure 8.
          #
Because Xf is a function of  x no further distance correction is needed.

Further Comments on Curves

The procedure just  described is based upon a number of limiting assumptions.
The most  important  of these  is that the position of the critical receptor
location  at  an  intersection  is determined by the queue length on approach 1,
Ideally,  the receptor should be located so as to maximize the joint con-
tributions from the free  flow emissions and the queue emissions on all
approaches but  this approach would be quite difficult to  implement on a
graphical basis.  Using the  queue on approach 1 to locate the critical
receptor  is  a reasonable  approximation  for the  following  reasons:
    •   The  relative  concentrations of  approaches 2, 3 and  4 are
        small with  respect to approach  1.  For  example, a queue
        length  of  60  meters  (approach 1) will contribute  approxi-
        mately  70  percent  of the contribution of all queues combined.
    •   Although  the  normalized contribution of queue and free
        flow emissions  are comparable,  the excess emissions
        assigned  to the queue are often several times higher
        than the  free flow emissions.
    •   If  free flow emissions were allowed to  influence  the
        choice  of  the critical receptor and the wind angle,
        the  wind  angle which would have been selected for the
        10 meter  road-receptor configuration would have been
        too  small  in terms of finite queues or  finite line
        sources.   The assumption of an  "infinite"  line  source  for
        free-flow  traffic  does not apply  very well near  inter-
        sections,  hence the  use of the  limiting wind/road angle
        dependent  on queues.

                                  19

-------
LJ
(J

IT
   tO 00
    900
    800
 =  700
 ac
 o
 E

10
i
O
 z
 o
 4
 o:
 UJ
 u
 z
 o
 u


 o
 UJ
 a:
 o
    600
500
    400
    300
200
    100
                                                               -I-
       10
              15       20     25    30   35  40  45  50    60


                      ROADWAY/RECEPTOR  SEPARATION, meters
                                                             70  80  90 100
    Figure 10.   Normalized CO concentration  contribution from  each traf-

                 fic stream at locations of uninterrupted flow
                                    20

-------
 The  green phase length is a fraction of the signal cycle time minus the
 total .amber time.  ,A:3-second amber time is assumed for all green phases.
 The  green phase length of phase j is given: by the following equation:
                               Max (V   /Cs.  )
                                j    x r j   1 »J
                    G. = Cy 	±	3              (6)
                     J        E   Max (V. ./Cs^ .)
                            all j  i    *•*    *>*
 where    Max (V. ./Cs. .) is the maximum V/Gs ratio on all approaches i
           i     '•*         moving on greeii phase j.
                         3 is an assumed 3-second amber time.
        2   (V^ ./Cs.  .)  is the sum of the V/Cs  ratios  that control
       all j     'J     >J   the green phase durations.
The  approach capacity,  C,  is found by multiplying the approach capacity
service  Volume by the appropriate green to cycle ratio and summing for
all  applicable phases.   The capacity of an approach is given as follows:

                             Ci = j Csi,j Gj/Cy                        (7)

where  j  are  those green signal phases that allow traffic  to move on inter-
section  approach  i.

Unsignalized Intersection  Traffic Movement

At unsignalized intersections,  the number of queued vehicles,  N,  is given
simply by
                                  N -
The queue length  is  found using  equation  (3)  or  40  meters  as  before,  and
the idle time is

                                      360  ON
                                 Rq =  ~ n -                           (9)
                                   25

-------
For an unsignalized intersection, the approach capacity, C,  is estimated
differently than for a signalized intersection.  Instead of depending on
the G/Cy ratio, it is a. function of the traffic flow on the cross street
and the time gap between cross street vehicles that is acceptable to a
driver wanting to cross or turn onto the cross street.  In this case C
is found from
                                  -T(V  + V )/3600
                                 e    m
                        i      ,     -T(2V)/3600
                               1 - e
where   V is the volume in one direction on the cross street in vph.
       2V is the assumed two-way volume on the cross street.

The capacity on the cross street  is assumed to be equal to  the free flow
capacity on that street.  It  is also assumed that no vehicles on the
cross street stop at the intersection.  The parameter T is  the acceptable
time gap (seconds) between cross  street vehicles.  It was assumed to be
4 seconds in developing Table 11  of Volume I.  While this value is  lower
than that given in the Revised Indirect Source Guidelines and hence produces
a higher capacity and fewer queued vehicles, it was chosen  as being more
appropriate for congested, potential hot spot locations where more  aggres-
sive driver behavior would previal.

Excess Emission Calculation

To calculate the excess emissions produced due to the queue length, Le,
and the idle time, Rq, emissions  must be known as a function of driving
mode as well as speed.  Thus, rather than using average speed as in AP-42,
an emissions estimating technique called the Modal Model    is used.  The
Modal Model calculates total  emissions over a user-specified driving sequence
by adding the emissions from  each 1-second time interval of the driving
sequence.  The emissions during each interval are found as  the product of
the "instantaneous" emission  rate and the 1-second time interval.  The
instantaneous emission rate,  e, during deceleration and acceleration modes
is a function of speed, v, and acceleration or deceleration, a:

                                  26

-------
           e  (v,a) or e  (v,a) = b- + b v + b a + b.av + b,-v

                                     2      2       2       22
                                + b,a  + b,v a + b_a v + b.v a .      (11)
                                   o      /       o       y

where for this application it is assumed a = 2.5 mph sec" , a typical value.

During cruise mode it is a function of speed only:

                      ec (v,o) = b1Q + buv + b12v2.                  (12)

These two equations were modified to calculate modal emissions for hot
spot analysis as described below.

All Modal Model emission estimates above are from a user-specified vehicle
age mix.  The effect of different vehicle age mixes is to change the b. co-
efficients of equations (11) and (12).  These empirical coefficients were
derived using 1975, warmed-up light-duty vehicle emission data deteriorated
to 1977 emission levels.

Excess emissions are those occurring over and above those which would have
occurred had the vehicle not stopped.  This may be expressed by:

                     EE = V ED - EC + EID Rq                        (13)
where E  = total excess emissions per vehicle
       E
      E  = total emissions due to acceleration per vehicle
       A
      E  = total emissions due to deceleration per vehicle
     E   = total idle emissions per vehicle/per second
      R  = average queueing time (seconds)
       q
      E  = total cruise emissions per vehicle
       c

The emissions during acceleration or deceleration are found by relating
speed to acceleration and time.  Under constant acceleration or deceleration
the speed can be expressed as a function of time as:

                                 27

-------
                             v = v  + at                              (14)
                                  o
where  v  is the initial speed of the vehicle .
        v is the vehicle speed after time, t
        a is the rate of acceleration or deceleration
        t is the time of travel

Substituting Equation (14) in Equation (11), and integrating over the time
    to come to a stop or to reach cruise speed:
                    /T
                      e   (v(t),a) dt, grams per vehicle
                       A
                    /T
                      e_  (v(t),a) dt, grams per vehicle
                   ..
                                                   (15)
•T
 Implicit in these equations is that the  initial approach  and  the  final  de-
 parture speeds are assumed to be equal.

 The  emissions due to  idling are not estimated using  the Modal Model,  but
 are  calculated using  AP-42 (1978) mobile source emission  factors  since  the
 estimate of idle emission from the Modal Model, that is the use of  b^g, is
 less accurate than the AP-42 factor.  This  is because  AP-42,  is from  observed
 data and the Modal Model from an empirical  fit.   The cruise emissions may be
 estimated  by using the vehicle speed  in  Equation  (12).  This  is the estimated
 emissions  per vehicle on an uninterrupted roadway.   The cruise emissions
 used in Equation (13), however, must  only be over the  acceleration  and
 deceleration distance so that only those emissions had the vehicle  not
 stopped are subtracted.

 When 'constant acceleration is assumed as in this  analysis, the time (T)
 to cover the distance a vehicle travels  accelerating from 0 mph  to  cruise
 speed, is  traveled by a cruising vehicle in %T.   Hence, to calculate  the
 equivalent emissions  of a cruising vehicle  during acceleration and  decelera-
 tion, Equation  (12) is modified so that  v is a  function of time  (as in
 Equation (14)) and integrated with respect  to i>T:
                                28

-------
           E
                   .T/2
            c
/'
   e£ (v(t),0) dt, grams per vehicle
                                                                       (16)
This  is  two  times  the  integral  to  allow  for both  the acceleration and de-
celeration portions.

In deriving  the  excess emissions for  the hot  spot guidlines, the estimated
idle  emissions from AP-42  times the idle time plus the  integrated Modal
Model estimates  in Equations  (15)  and (16) are summed as in Equation (13).

                         -1   -1
Converting to units of gm  sec   by considering the number of vehicles that
stop, N, the total cycle time, C , the average running time for all vehicles,
                                y
Rq, the volume demand,  V,  and the  queue  length, Le, (and letting
E   = E  + E  -  E  ) Equation  (13)  becomes:
 AJJ    A    JD    L»
        (EN
         AD
        ~
                                       R
                      EE-

which gives the emissions occurring per unit length of queue from acceler-
ating, decelerating, and idling vehicles.

Free Flow Emissions

Similarly, the cruise emissions, e  , in Equation (12) may be used to esti-
mate free flow emissions for all vehicles travelling through an intersection
or for uninterrupted flow.  Considering the number of vehicles per hour, V,
and the distance traveled by those vehicles in 1 hour, X, (i.e., the speed,
mph), the following equation results:
                           e  V        ,     .
                            -c    ,__ _-l  ___-!)                       (lg)
                      c   X 1609
where 1609 is the conversion from miles to meters.
                                 29

-------
Derivation of Emission Correction Factors

As stated in the last section, the Modal Model   is used instead of AP-42
to generate emissions estimates for these guidelines for the acceleration,
deceleration, and cruise modes of vehicle travel.  The Modal Model is used
because it can more accurately estimate emissions over variable driving
sequences, such as occur at intersections, than can AP-42.   One drawback
to using the Modal Model is that the emissions are only applicable to one
set of emission conditions, viz:
                               (100 percent stable)
     calendar years = 1977
     75 F ambient temperature
     0 percent cold starts  )
     0 percent hot starts   (
     low altitude
     non-California
     light duty vehicles.
These will be called base conditions.  In order to combine the best features
of the Modal Model (variable driving sequences) with the best features of
AP-42 (variable average speed, cold starts, hot starts, temperature, calen-
dar year, and region) it is necessary to make an assumption relating the
two procedures.  The assumption is essentially that the ratio of estimated
emissions under other than base conditions to those estimated under base
conditions are equal for AP-42 and the Model Model; i.e.:
                  (AP-42) Scenario = (MM) Scenario
                    (AP-42) Base       (MM) Base                      (19)
where AP-42 estimates are calculated assuming the average vehicle speed
      of the driving sequence in the Modal Model
and   (MM) Base are calculated using any driving sequence under base
      conditions in the Modal Model
and   (MM) Scenario is the unknown being solved.

Hence,  to correct for calendar years, temperatures, cold starts, hot starts,
average speed, and regions other than the base conditions, total emission
                                30

-------
factors using AP-42 (henceforth called composite emission factors) are
calculated ,for (AP-42) Scenario, and divided by the AP-42 emissions for
light duty vehicles for base conditions.  This ratio is multiplied by the
Modal Model composite emission factor under the same base conditions and
thus solves for (MM) Scenario.  This emission factor reflects adjustments
for both variable driving sequence and variable environmental and calendar
year conditions.

In the following subsections the above procedure will be discussed as it
applies to each vehicle category and finally how to apply a total correction
factor to the Modal Model emissions using vehicle proportions as weighting
factors.

Light Duty Vehicles (LDV)

To adjust the Modal Model emissions estimates to reflect the user specified
light duty vehicle emissions conditions, the ratio technique discussed
above must be applied.  The AP-42 emissions may be more specifically defined
by:
                      (AP-42)   = S E    . M. R.                       (20)
                             cy      cy,i  i  i
where (AP-42)   is the composite emission factor for a given calendar
              ^ year (composite meaning  total of all model years still
                running in that calendar year)
          E    . is the emission factor for each model year in the given
           cy>1 calendar year
             M  is the fraction of annual vehicle travel by model year
              i
             R. is the correction factor for cold starts, hot starts,
              1 temperature, and speed by model year.

Thus (AP-42) Scenario and (AP-42) Base are estimated and multiplied by  the
Modal Model emissions for the base conditions to yield the AP-42 adjusted
Modal Model estimate for the desired scenario.  For convenience this AP-42
adjustment will be called R*DV Qr.
                                  31

-------
                     (AP-42)    _      .     £E    . M  R
                                _      .           .
              R*   = _   cy. Scenario      cy,i  i  i _            (  1}
              *LDV       (AP-42),            2E,.   . M. R    .            ^
                                by            by,i  i  by,i
where the subscript, by, refers to base year.

Light Duty Trucks (LPT) and Motorcycles (MC)

The derivation of correction factors for light duty trucks and motorcycles
is similar to that for LDV's.  The only difference of note is that  the E
                                                                        cy,i
AP-42 estimates are for trucks (or motorcycles).

Heavy Duty Trucks (Gas-HDG and Diesel-HDD)

Correction factors for heavy duty trucks only apply to adjustments  for
speed and model year.  The correction factors (C*) are calculated by
taking the ratio of the composite AP-42 emission factors for the calendar
year of interest and the composite AP-42 emission factors for LDV's under
the base conditions.  Thus:

                                2(EHDG. M.  V   .)
                        r-.-    =    cy,i  i  s,i                        (22)
                        C'HDG -    LDV      v   }
                                  v by, i  i  s,i
                                                        -I-
and a similar expression for heavy duty diesel trucks (c"  ).
                                                        HDD

Application of Correction Factors

A composite correction factor must be calculated combining all vehicle types
at the roadway under analysis in order to make the CO concentration esti-
mates reflect all vehicle categories.  Thus a total composite correction
factor C  may be given by:
    C  = P    R*   + P    R*   + P    C*   + P    C*   + P   R*
     T   rLDV KLDV    LOT *LDT    HDG  HDG    HDD  HDD   *MC  MC

where PLDV, PLDT> PHDG> PHDD' PMC are the Pr°Portion of each vehicle type
      R*    R*    R*   C*    C*
       LDV  LDT'  MC:  HDG'  HDD are composite correction factors as
      described previously.
                                  32

-------
Since the verification methodology employs precalculated tables of cruise
and excess emission rates to compute emissions on each link, a separate
Crp must he applied to both the cruise emissions and the excess emissions.

With these basic correction factors in hand, base emissions may now be
varied to other scenarios.  Since the cruise component of emissions (i.e.,
free flow) has a known speed, s, the corrected cruise emission can be
calculated, viz :
                     (C )  (E ) = (E ) corrected                      (24)
                       J- S   (j      \_»
where E  is the cruise emission rate estimated from the Modal Model, speed,
       LI
       in
      discussed above for given cruise speed.
and volume of traffic and (C )  is the composite correction factor as
The corrected excess emissions is the difference over the queue length (or
acceleration and deceleration distance), between the cruise emissions and
the emissions to decelerate, idle, and accelerate.  The average speed of
the cycle of deceleration, idling, and acceleration is very low - in almost
all situations less than 5 miles per hour.  For the purposes of screening
and hot spot verification the R-factors of AP-42 are calculated at 5 mph
to correct the excess emissions at intersections.

Thus, corrected excess  emissions are:

         CT,5(EAD  + EC  + V  ~  CT,S(y = (V C°rreCted            (25)
Note that the correction factor is applied to  total queue emission rates
(E  is added back  into  E  ) and not  to  excess  emissions.  This allows the
  C                     AJJ
cruise emissions to be  separately accounted for  at  the actual cruise  speed
and then subtracted to  yield corrected  excess  emissions.

It is important to note that the cruise emissions  in Equation  (25)  is the
cruise emissions that would have occurred for  the  vehicles  that  stop, if
they had not stopped, and not  the total cruise emissions.   It  is  for  this
reason that the tables  of excess emission rates  in Volume  I list  two
numbers, the total queue emissions and  the cruise  emissions that  would
have occurred had  the vehicles in the queue not  stopped.
                                  33

-------
                              SECTION III
             DEVELOPMENT OF THE HOT SPOT SCREENING PROCESS

INTRODUCTION

This section describes the development of the hot spot screening techniques
presented in Volume I.  In essence these screening procedures graphically
protray the results of repeated applications of the verification procedure.
The primary input parameter is traffic volume.  For each roadway/receptor
situation, there is a critical traffic volume above which the potential
for a violation of the CO standard exists (according to the model implicit
                                   2
in the Indirect Source Guidelines).   Thus,  utilizing traffic volumes and
several simplifying assumptions about traffic and dispersion, a determina-
tion can be made as to whether a  given location has hot spot potential.

Separate techniques have been developed for analyzing three broad cate-
gories of roadway facilities, including:  signalized intersections, non-
signalized intersections, and uninterrupted flow locations.  Furthermore,
separate provisions are provided for considering particular types of loca-
tions within each category, such as freeways versus nonfreeways for un-
interrupted flow locations, two-lane versus four-lane approaches in the
intersection categories, and so on.

Because the effort here was directed toward development of a general guide-
line for identifying carbon monoxide hot spots, consideration had to be
given to several issues that have an influence on the methodology being
utilized.   First, the guidelines will be used to evaluate literally
hundreds of street sections and intersections within any municipality;
                                   34

-------
therefore the parameters considered must be general enough to require the
absolute minimum of data input, yet the process must yield a reliable as-
sessment of hot spot potential.  Second, the process should be relatively
simple and capable of being accomplished quickly, utilizing data that are
ordinarily available from  state or city agencies.  Third, the process is
intended to be applicable  to any city or town where the existence of a hot
spot problem is suspected.  These factors, plus the fact that traffic op-
erating characteristics are often highly varied among similar locations
(for example, among signalized intersections), indicated that the screening
process had to involve a very general approach, relying to a large extent
on the validity of applying an assumed set of conservative traffic emission
and dispersion parameters  in order to reduce to a minimum the number of
variables considered in the process.

Consequently, the screening guidelines were developed utilizing generalized
assumptions regarding several of the variables in the verification pro-
cedure, thus simplyifying  the amount of data and computing needed for CO
           *
assessment.   A sensitivity analysis described below verifies the reason-
ableness of the simplifying assumptions and shows the direction of the
effects on air quality of  variations in the parameter assumptions.

The screening guidelines were first presented in previous documents de-
                                                        3 4
scribing a procedure for identifying hot spot locations. '   Subsequent
to the publication of these volumes, the Indirect Source Guidelines,  and
AP-42 emission factors  were revised.  These revisions are reflected in
both this volume and Volume I of the Guidelines.
 The screening curves are generated by a computer program described in
Volume IV of these Guidelines.  The program  is set up so that new curves
may be computed for assumptions other than those made in this section.
                                  35

-------
ASSUMPTIONS REGARDING BACKGROUND CONCENTRATIONS

As mentioned previously; ambient concentrations of carbon monoxide at any
location on an urban highway network are actually a result of emissions
from the nearby source (highway) plus a background concentration resulting
from emissions generated at more distant sources.  Consequently, it was
necessary to account for background concentrations in the screening guide-
lines.  Various methods are suggested in the Indirect Source Guidelines
and Volume I of the  Hot Spot Guidelines for determining the background
concentration for a particular area; however, these methods require a certain
amount of short-term, local air quality monitoring and also assume that
historical ambient air quality data are available from a permanent monitor-
ing station within the general area of the site.  Given the cost and time
requirements for short-term monitoring, and the extent of current long-term
air quality monitoring programs, it is highly unlikely that these procedures
could be used in the context of the hot spot screening.  In an attempt to
develop a simple yet reasonable method for identifying background concentra-
tions, the results of diffusion modeling (using the APRAC diffusion model)
efforts in three New England cities were analyzed.   The data available for
analysis included estimates of background concentrations at 20 receptors in
•each city computed from local traffic data, and local meteorological data
for a 1-year period.  The averages of the maximum background concentrations
                                                        3                   3
computed for the 20 receptors in each city were 2.9 mg/m  (2.5 ppm), 3.3 mg/m
                      2
(2.9 ppm), and 5.9 mg/m  (5.1 ppm) (averaged over 8 hours).  These data were
developed for conditions during 1973-74.  If these averages are projected
for 1982-83 conditions,  the results (conservatively) are 1.4 mg/m3 (1.2 ppm),
1.6 mg/m3 (1.4 ppm), and 2.9 mg/m3 (2.5 ppm).  Clearly, this shows that there
may be significant variations among cities with regard to background concen-
trations.   However, it can be postulated that a value of about 2.9 mg/m3
(2.5 ppm) may be representative of the upper portion of the range in which
8-hour average background levels occur.  Assuming that this is a reasonable
conclusion, then the corresponding 1-hour average background can be estimated
                                   36

-------
by applying the correlation factor (0.7)  developed for relating 8-hour to
1-hour average concentrations.  This results in an estimated 1-hour average
background concentration of about 4.1 mg/m3 (3.6 ppra).

The need for this "standard value" is again stressed because of anticipation
that sufficient local data will not be available in all instances to permit
a determination of actual background concentrations.  Therefore, a standard
value of background will be used in the generation of the screening curves.
                                                     2
Data and procedures in the Indirect Source Guidelines  are oriented to the
maximum 1-hour average concentration of carbon monoxide while the 8-hour
average concentration is of interest here.  Analysis of air quality data
                                               8 9
from a number of continuous monitoring stations '  indicate that the re-
lationship between the maximum 1-hour and 8-hour average concentrations
can be expressed by:

                             X8 = PXX                                 (26)

where  x0 = highest expected 8-hour average concentration
        o
       X, = highest 1-hour average concentration (in same 8-hour period)
        P = 8-hour correlation factor.

While the value of the correlation factor can be expected to vary depending
on local traffic and meteorological characterists, data analysis described
below indicates that a value of 0.7 may be considered appropriate, es-
                                                                      3
pecially in the range of 1-hour average concentrations of from 10 mg/m  ,
          3
to 20 mg/m .   In this analysis, monthly summaries of carbon monoxide con-
centrations measured at several monitoring stations in three cities were
reviewed.  The analyses consisted of determining the ratios of the maximum
daily 8-hour average concentrations to the maximum daily 1-hour average
concentrations for various monitoring sites in each city.  The maximum
8-hour average concentration used in the analyses included the maximum
 Discussed later in this section.
                                  37

-------
daily 1-hour average concentration in its averaging period.  Further,
these ratios were examined for three ranges of maximum daily 1-hour average
concentrations - where the maximum 1-hour average was (1) between 0 and
      3                            33
7 mg/m  (6 ppm), (2) between 7 mg/m  (6 ppm) and 11.5 mg/m  (10 ppm), and
                          3
(3) greater than 11.5 mg/m .   Considering the higher range separately, the
analysis indicated that for moderately active downtown locations (locations
typified by occasional traffic congestion) with apparently good atmospheric
ventilation (that is, fairly wide streets and sidewalks and building heights
generally not exceeding five or six stories), that the ratio of the maximum
8-hour average concentration to the maximum 1-hour average concentration
ranges from about 0.6 to 0.7.  For downtown areas where heavy traffic con-
gestion occurs throughout much of the day and where ventilation is somewhat
restricted by narrow streets and tall buildings, values of from 0.7 to 0.8
were indicated for this ratio.  Conditions in most areas where hot spot
analyses are conducted are like to be somewhat less severe (with regard to
traffic conditions and ventilation) than those in the congested area re-
ferred to here.  Given this assumption, the value of 0.7 is considered
"reasonable" for use as a "standard value" to describe the ratio of the
maximum 8-hour average concentration to the maximum 1-hour average concen-
tration, for cities and towns lacking sufficient data to permit development
of a more specific value.  The verification procedure does allow the use
of a locally-derived 8-hour correlation factor.

The maximum 8-hour average concentration of carbon monoxide for a particular
location, then, can be expressed by the general equation:

                          Xft = (X,  + Xn) P                            (27)
                           o     ID

where x0 = tne estimated maximum 8-hour concentration
       o
      X,  = the total estimated concentration contributed by a nearby
           source (roadway)
      XT, = the 1-hour average background concentration, assumed to be
           3.6 ppm (4.1 mg/m3)
       P = 8-hour correlation factor =0.7
                                  38

-------
If Xg is set equal to the National Ambient Air Quality Standard for 8-hour
concentrations (10.0 mg/m  (9.0 ppm)) then the only unknown in the equation
becomes x-, > or :

              _ X8 " XB(P)    (10.0) - (4.1)(0.7)   .- _   /3
           X1	p	 = 	Q-^	 = 10.2 mg/md

Therefore, for every roadway  condition where the calculated 1-hour average
concentration contributed by  the roadway is about 10.2 mg/m3, there is a
potential  for violations to the 8-hour standard, given the above
assumptions.
UNINTERRUPTED FLOW CONDITIONS

General

For conditions of uninterrupted flow on streets and highways, vehicle speed
is a key determinant of emissions intensity.  Because average travel speeds
vary to such a large extent throughout a highway network, it is necessary
to separately consider facilities where speed- characteristics are expected
to be quite dissimilar.  For this analysis, two facility-types were con-
sidered, these being (1) freeways, expressways, or other limited access,
high speed highway classes; and (2) arterial streets and highways.

The verification procedure presents a technique for expressing an empirical
relationship between air quality and various combinations of roadway volume
and capacity.  Again, other parameters, mainly meteorology, emission factors,
and certain operating characteristics, have been included, but as nonvari-
ables (the reader is referred to the previous section for a discussion of
the assumptions made regarding these parameters).  An implication of the air
quality-volume-capacity relationships expressed by the verification procedure
is that for every given value of lane capacity there is a critical volume
demand which, once reached, will generate concentrations of carbon monoxide
that are in violation of the National Ambient Air Quality Standards.
                                  39

-------
 Representative  vehicle  operating  speeds  are  needed to  estimate  the  emissions
 from  free-flowing  traffic.   The assumptions  used  here  are  based  on  relation-
 ships between volume-capacity  ratios  and operating speeds  on various  types
 of roadways with specified  average  highway speeds as estimated  from the
 Highway  Research Board's  L965  Highway Capacity  Manual.11   Traffic on  the
 roadway  is assumed to be  accommodated at the maximum level  of service con-
 sistent  with the indicated  v/c ratios.   This procedure  is  followed  in de-
 riving the screening curves for freeways and for  arterials.  For the  free-
 flow  curves in  the screening procedures,  demand-capacity ratios  are
 assumed  identical  with  volume-capacity ratios.  This assumption  implies
 that, under free-flow conditions, traffic _is_ moving at  the  maximum
 level of service consistent with  the  specified  volume-capacity ratio.
 Thus, the operating speeds  on  a roadway  are  possible to estimate given
 the demand-capacity ratio.   The combinations of operating speeds and
 demand-capacity ratios  derived for  arterial  streets and expressways,
 along with the  corresponding levels of service appear in Tables  1 and 2.

 To be representative of as  large  a  portion of the  country as possible and
 to insure usability under representative  adverse  emissions  conditions,
 about 20 percent of the vehicle population is assumed to be operating under
 cold-start conditions and 0 C  is  assumed  appropriate for worst case temper-
 ature in the winter months  (the season where CO concentrations have most
 commonly been high).

 An additional adjustment was to adjust the curves  to apply  to the winter
 of 1982  to 1983, to correspond with the  statutory  attainment date for the
 carbon monoxide NAAQS.  The adjustment is identical to  that described pre-
viously  for the verification procedure.

 Curve Generation and Discussion

The result of the  analyses  performed  was  the identification of the  critical
 (minimum) ADT which could result  in a violation to the  National  Ambient
Air Quality Standard for 8-hour average  carbon monoxide concentrations.

                                  41

-------
Table 1.   ASSUMED OPERATING SPEEDS, LEVELS OF SERVICE AND DEMAND-CAPACITY
          RATIOS FOR MAJOR STREETS AND CORRESPONDING EMISSION FACTORS FOR
          FREE FLOW CONDITIONS
Assumed
operating
speed
(mph)
30
25
20
15
15
Demand -
capacity
ratio
^0. 60
0.70
0.80
0.90
1.00
Level
of
service
A
B
C
D
E
Description
Completely free flow
Stable flow (slight delay)
Stable flow (acceptable delay)
Approaching unstable flow
(tolerable delay)
Unstable flow (congestion,
intolerable delay)
Table 2.  ASSUMED OPERATING SPEEDS, LEVELS OF SERVICE AND DEMAND-CAPACITY
          RATIOS FOR URBAN EXPRESSWAYS3 AND CORRESPONDING EMISSION FACTORS
          FOR FREE FLOW CONDITIONS
Assumed
operating
speed
(mph)
57
55
53
50
47
45
42
40
37
30
Demand -
capacity
ratio
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Level
of
service
A
A
A
A
B
C
C
C
D
E
Description
Completely free flow
Completely free flow
Completely free flow
Completely free flow
Stable flow (upper speed
range)
Stable flow
Stable flow
Stable flow
Approaching unstable flow
Unstable flow
      Average highway speed assumed to be 60 mph.
                                 42

-------
Figures in the screening guidelines show these critical volumes for various
configurations of limited and uncontrolled access facilities.  The figures
provide the basis for screening roadways where conditions of uninterrupted
flow prevail.  The resulting procedure provides a "go/no-go" type of anal-
ysis.  The procedure simply indicates that a hot spot potential exists or
does not exist.

The procedure for screening highway sections where uninterrupted flow con-
ditions prevail requires only very basic data regarding the roadway net-
work.  Essentially, the data required involve traffic volumes and traffic
flow characteristics, and general physical data including number of travel
lanes, and estimates of other capacity determinants, such as lateral lane
clearance.  The data required allow estimates to be made of the facility's
lane capacity which, through the use. of the curves presented in Volume I,
can be related to a corresponding "critical" ADT which, potentially, would
result in a violation to the National Ambient Air Quality Standard (NAAQS)
for 8-hour average concentrations of carbon monoxide.  This critical ADT
is then compared with the estimated ADT on the facility, and a potential
hot-spot is indicated when the estimated ADT for the facility equals or
exceeds the critical ADT.

INTERRUPTED FLOW CONDITIONS - SIGNALIZED INTERSECTIONS

General

Near signalized intersections, emissions intensity is affected by vehicle
operating characteristics including acceleration and deceleration rates,
time in idle mode, volume of vehicles that stop, and total volume.  These
operating characteristics, in turn, are influenced by such elements as
intersection capacity, amount of green (signal) time allocated to each
approach, location of the intersection (e.g., rural, residential, or down-
town), and the proximity of other signalized intersections.  Emissions  in-
tensity is also related to the physical layout of the intersection with
respect to lane configuration as well as receptor location relative to  the
                                  43

-------
emissions source.  Since there is a Large number of variables that can
significantly affect carbon monoxide concentrationsj 40 prototypical inter-
sections were developed as the basis for the screening process.

A number of assumptions in addition to those in the verification procedures
were required in developing a general procedure for assessing the hot spot
potential of signalized intersections.  These concern mainly traffic volume
distribution, receptor distance, and general intersection operating charac-
teristics, as discussed below.

Assumptions

It was assumed, again, that the basic parameter of traffic volume would be
expressed as ADT.  Also, the following assumptions were used regarding ADT.
It is recognized that conditions vary widely in urban areas; the following
conditions are assumed to be typical:
    •   peak hour traffic represents 8.5 percent of the ADT
    •   an even directional distribution occurs on two-way
        facilities during the peak hour
    •   for multilane facilities, the volume on the lanes are
        equally distributed.

The distance from the edge of the roadway to the receptor was assumed to
be 5 meters.

Two different cruise speeds were used in developing the screening procedure
the first was 15 miles per hour used for conditions where congestion is
highly likely (e.g., within urban business districts), while 30 miles per
hour was used for noncongested areas.

Assumptions regarding cold operation and temperature correction for sig-
nalized intersections were the same as those described in the previous
section dealing with uninterrupted flow conditions.  These assumptions
                                  44

-------
were £hat the ambient temperature representative of winter operation is
0 C and that 20 percent of the vehicles passing any point in the highway
network would be operating under cold conditions.  These issues are dis-
cussed in detail in a previous portion of this section.  Data were also
adjusted to 1982 to 1983, as before.

Curve Generation and Discussion

The results of the analyses were the identification of the critical (mini-
mum) ADT's for several basic configurations of intersecting streets that
could result in violations to the National Ambient Air Standard for 8-hour
average carbon monoxide concentrations.  In the analysis, eight general
intersection approach configurations were considered, including:
    (a) 4-Lane, 2-Way in congested areas
    (b) 4-Lane, 2-Way in noncongested areas
    (c) 3-Lane, 2-Way in congested areas
    (d) 3-Lane, 2-Way in noncongested areas
    (e) 2-Lane, 2-Way in congested areas
    (f) 2-Lane, 2-Way in noncongested areas
    (g) 2-Lane, 1-Way
    (h) 3-Lane, 1-Way
For each of these, five configurations of intersection street-types were
analyzed and corresponding "critical" ADT's determined.  Thus, a total of
40 separate intersection configurations were evaluated; results are presented
in the screening procedures.

For screening signalized intersections, traffic volume data, physical lay-
out and traffic operational characteristics are of primary importance.
The traffic volume and physical/operational characteristics of each approach
to the intersection are then related to the physical/operational charac-
teristics of the opposing roadways and, from these, the  "critical" volumes
are determined for the opposing traffic.  If the actual volumes on these
cross streets are greater than or equal to the "critical" volumes, a hot
spot potential is indicated for the approach being analyzed.

                                  45

-------
INTERRUPTED FLOW CONDITIONS - NONSIGNALIZED INTERSECTIONS

General

Although nonsignalized, at-grade intersections are by far the most common
type of intersection, they are also the least studied with regard to op-
erational characteristics and capacity.  Nonsignalized intersections are
very seldom a critical factor of capacity or level of service on through
routes although they are of great significance with regard to the minor
cross route.    Typically, when a point is reached where traffic flow on
the major route is affected to any degree by traffic from a minor cross
                              •,<
street, a signal is installed.   To date, most of the research that has
been conducted on capacity and operational aspects of nonsignalized inter-
sections has tended to produce data representative of local conditions
only.

In perspective, however, nonsignalized intersections are also probably the
least likely hot spot locations since they are generally characterized by
relatively low volumes on the minor cross streets, and produce little, if
any, interference to main street flow.  This being the case,  the main
stream legs of many nonsignalized intersections can be assessed for hot
spot potential using the technique described earlier in this section for
uninterrupted flow conditions.  For the minor street legs, however, a dif-
ferent analysis must be performed.

The concept, then for develping a screening procedure for nonsignalized
intersections was the same as that used in developing the screening pro-
cess for both uninterrupted flow conditions and signalized intersections.
In essence , this involved identifying volume relationships which result
in concentrations (based on the verification procedure) high enough to po-
tentially violate the NAAWS for 8-hour average concentrations of carbon
monoxide.
 Four-way stop sign controlled intersections are excluded to keep the scope
jf these guidelines manageable.  Such intersections, with their typical low
traffic volumes,  are unlikely hot spots.

                                  46

-------
Assumptions

Assumptions used regarding peak hour volume, directional split, and lane
distribution were the same as were used for development of the screening
procedure for signalized intersections.  The minimum distance from the edge
of the nearest lane to a receptor was assumed to be 5 meters.

Curve Generation

The verification procedure and the assumption noted above were utilized
to develop critical volumes  for various configurations of STOP-sign con-
trolled intersections, as shown in the guideline volume.  Included are the
following:
    •   2-lane, 2-way minor; 2-lane major  (congested area
    •   2—lane, 2-way minor; 2-lane major  (noncongested area)
    •   2-lane, 2-way minor; 4-lane major  (congested area)
    •   2-lane, 2-way minor; 4-lane manor  (noncongested area)
    •   4-lane, 2-way minor; 4-lane major  (congested area)
    •   4-lane, 2-way minor; 4-lane major  (noncongested area)
    •   2-lane, 1-way minor; 2-lane major
    •   2-lane, 1-way minor; 4-lane major

EFFECTS OF VARIATION IN PARAMETER ASSUMPTIONS

The analysis presented here  is performed to show the sensitivity of the
computed CO concentrations to changes in the input parameters for the
screening curves.   Assumptions for six traffic and air quality parametert
were among those required to compute the values for the plotting of the
hot spot screening curves.  These assumptions are calculated in Table 3
for the analysis of a signalized intersection composed of a 4-lane/2-way
roadway in a noncongested area which is crossed by a 2-lane/2-way facility.
                                  47

-------
       Table  3.  ASSUMPTIONS FOR  CO CONCENTRATION  COMPUTATIONS
                                                               a
                    Parameter
      Directional volume split on main  street
      Roadway edge to receptor distance
      Percent ADT during peak hour
      Background concentration,  1 hour
      Acceleration-deceleration
      8-Hour correlation factor
Value assumed in
screening curves
   50% - 50%
           3
  5 meters
    8.5%
  5.8 mg/m
2.5 mph sec"
    0.70
       Analysis of a 4-lane/2-way roadway in a noncongested area
       crossed by a 2-lane/2-way roadway.  Intersection is  fully
       signalized.

The estimates for the parameters are regarded as producing an accurate
model  of the proposed generalized conditions.  Since all estimates, how-
ever,  are subject to some amount of uncertainty, a sensitivity analysis,
discussed in the following text, helps in analyzing and understanding
the effects of variations in the values of the control parameters.

The process selected for the sensitivity analysis was to vary one param-
eter at a time and recompute the CO concentration screening curve for
that new set of parameters.   The variations in the parameter values se-
lected for use in the sensitivity analysis are listed in Table 4.  The re-
sults of the analyses show that most of the selected changes (Table 4) in
most parameters can have substantial effects on the curves.  The only excep-
tion is acceleration, in which a change of 0.5 mph sec'1  had no effect.
The basic screening curves and the seven variations are plotted on Figure 11
for comparison.   The  percentage variation in the allowable volumes are
presented in Table  5, along  with the differences in the parameter values.
                                48

-------
         Table 4.  VARIATION IN ESTIMATED PARAMETER VALUES FOR
                   SENSITIVITY ANALYSIS
                  Parameter
   Value used in
sensitivity analysis
      Directional volume split
      Roadway edge to receptor distance
      Percent ADT during peak hour
      Background concentration, 1 hour
      Deceleration-acceleration
      8-hour correlation factor
    60% to 40%
     2 meters
        10%
     4.8 mg/nf
    2 mph sec
        0.9
-1
The significance of this analysis of possible variations in the parameters
is that it exhibits the effects on expected CO concentration of differences
between the assumed conditions and conditions at an actual site.  In ad-
dition, it shows the direction and relative magnitude of the effect on air
quality of changes in site conditions that might, be brought about through
some of the control measures.

The results shown in Table 5 and plotted on Figure 11 exhibit the variations
possible with example changes in the estimated values of the parameters.   It
is considered that the set of values for the generalized conditions used in
the screening analysis yield satisfactory results for average conditions.
Uncertainty in the use of these estimated values will undoubtedly occur at
unusual or complex locations.  The uncertainty for these special cases,
however, would be eliminated during application of the verification process.
                                49

-------
        <" g
        CO g
        O 2
        SI
        z •-
        O ~*
       o
       <
                                                              CURVE
                                                                     LEGEND
                                                                     VARIED PARAMETER
                                                                    BASE CASE
                                                                    BACKGROUND
                                                                    8 »o I hour FACTOR =0.9
                                                                    ACCELERATION, DECELERATION =2.0, -2.0
                                                                    ROAD-RECEPTOR = 2 meteri
                                                                    DIRECTIONAL SPLIT = 75% , 25 %
                                                                    PEAK HOUR FACTOR = 10%
                                                              NOTE -REFER TO TABLES  3,4  AND 5
                         2     3    4    5     6     7     8     9     10    II    12    13    14-
                          ADT ON STREET OF ANALYSIS : 4-lone/2-way (NON-CONGESTED  AREA)
                                              (in thousands of vehicles)
15
Figure 11.   Examples of effects  on screening curves  of variations in parameter assumptions

-------
        Table 5.  CHANGES IN PARAMETERS VERSUS CHANGES IN ALLOWABLE VOLUMES
            Parameter
 Difference in value
 of parameter between
   screening curve
   assumption and
sensitivity analysis
 Approximate change
in allowable volumes
     (percent)
Background concentration
8-Hour correlation factor
Acceleration, deceleration
Roadway edge to receptor distance
Directional volume split
Percent ADT during peak hour
     - 1 mg/m
        0.2
    0.5 mph/sec
     -3 meters
        25%
       +2.5%
       +35%
        40%
     no change
       -15%
       -25%
       -25%

-------
                               SECTION IV

                               REFERENCES
1.   Carbon Monoxide Hot Spot  Guidelines.   Volume I:   Techniques  and
     Workbook.   GCA Corporation,  GGA/Technology  Division,  Bedford,
     Massachusetts.  Prepared  for U.S.  Environmental  Protection Agency,
     Research Triangle Park, North Carolina.   GCA Report No.
     TR-77-02-GU).  June 1978.

2.   Guidelines for Air Quality Maintenance Planning  and Analysis.
     Volume 9 (Revised):  Evaluating  Indirect  Sources.  OAQPS No. 1.2-
     028R, EPA-450/4-78-001, U.S.  Environmental  Protection Agency,
     Research Triangle Park, North Carolina.   September 1978.

3.   Guidelines for Identification and  Evaluation of  Localized  Violations
     of Carbon Monoxide Standards.  Final  Guideline Report.  GCA  Corpora-
     tion, GCA/Technology Division.   Bedford,  Massachusetts.  Prepared  for
     U.S. Environmental Protection Agency,  Region I Office, Boston,
     Massachusetts.  Publication  Number EPA-901/9-76-001.  January  1976.

4.   Development of Guidelines for Identification and Evaluation  of
     Localized Violations of Carbon Monoxide Standards.  Final  Summary
     Report.  GCA Corporation, GCA/Technology  Division, Bedford,  Massachu-
     setts.  Prepared for U.S. Environmental Protection Agency  Region I
     Office.  Boston, Massachusetts.  Publication Number EPA-901/9-76-002.
     January 1976.

5.   Guidelines for Air Quality Maintenance Planning  and Analysis.
     Volume 9:   Evaluating Indirect Sources.   U.S.  Environmental  Pro-
     tection Agency, Research  Triangle  Park, North Carolina  27711.
     Publication Number EPA-450/4-75-001.   January 1975.

6.   Mobile Source  Emission Factors.  U.S.  Environmental Protection
     Agency.  Washington, D.C. EPA 400/9-78-005.   March 1978.

7.   Midurski,  Theodore P., et al.  GCA/Technology Division.  Development
     and Evaluation of a Transportation Control  Plan  for the Massachusetts
     Portion of the Hartford-New  Haven-Springfield Air Quality  Control
     Region.  Volume I:   Proposed Transportation Control Plan.  Prepared
     for U.S.  Environmental Protection  Agency, Boston, Massachusetts.
     Publication No. EPA-901/9-75-002a.  September 1975.   p. 321.
                                52

-------
8.   Massachusetts Department of Public Health, Bureau of Air Quality
     Control.  Computer Summaries of Air Quality Data from the Waltham,
     Massachusetts Monitoring Site, 1-Hour and 8-Hour Average Carbon
     Monoxide Concentrations for 1972 through 1974.

9.   Massachusetts Department of Public Health, Bureau of Air Quality
     Control.  Computer Summaries of Air Quality Data From the Monitoring
     Sites in Springfield, Massachusetts, 1-Hour and 8-Hour Average
     Carbon Monoxide Concentrations for 1972 through 1974.

10.  Kunselman, P. et al.  Automobile Exhaust Emission Modal Analysis
     Model.  U.S. Environmental Protection Agency Report No. EPA-460/3-74-
     005.  January 1974.

11.  Highway Capacity Manual.  Highway Research Board, National Academy
     of Sciences, National Research Council.  Special Report No. 87.  1965.

12.  Users Guide  for HIWAY.  U.S. Environmental Protection Agency.
     Research Triangle Park, North Carolina.  EPA/650/4-74/008.
     February 1975.

13.  Mancuso and  Ludwig.  Users Manual for the APRAC-1A Urban Diffusion
     Model Computer Program.  Stanford Research Institute, Menlo Park,
     California.  September  1972.
                                 53

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 RfcPORT NO. 2.
EPA-450/3-78-034
-J. TITLE AND SUBTITLE
Carbon Monoxide Hot Spot Guidelines
Volume II: Rationale
7. AUTHOR(S)
Frank Benesh
9. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA/Technology Division
Burlington Road
Bedford,, Massachusetts 01730
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, N.C. 27711
15. SUPPLEMENTARY NOTES
3. RECIPIENT'S ACCESSI ON- NO.
5. REPORT DATE
August 1978
6. PERI-ORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-78-32-G(3)
10. PROGRAM ELEMENT NO.
2AF643
11. CONTRACT/GRANT NO.
68-02-2539
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE

16. ABSTRACT
This report presents the rationale used in developing the analytical techniques
for the carbon monoxide hot spot guidelines.
Discussed in this report are the technical aspects of the guidelines, including
the assumptions used in developing hot spot procedures. Since the guidelines
were based largely on EPA's Indirect Source Guidelines, these are discussed in
some detail, as well.
17 KEY WORDS AND DOCUMENT ANALYSIS
.i DESCRIPTORS b.lDENTIFI
Carbon
Air Qua
Transpc
•i - - r ^ < i JTI O\ STATEMENT 19.SECURI
Pelease Unlimited UNCLASE
20. SECURI
UNCLAS5
ERS/OPEN ENDED TERMS C. COSATI Field/Group
Monoxide
ility Models
>rtation Planning
TY CLASS (This Report) 21. NO. OF PAGES
5IFIED 64
TV CLASS (This page) 22. PRICE
JIFIED
EPA Porrn 2220-1 (9-73)

-------