CMA For Signalized Intersections by Elena Prassas

M23_ROES5739_04_SE_C23.
QXD
6/1/10
8:51 PM
Page 545
CHAPTER
23
Critical Movement
Analysis of Signalized
Intersections
In Chapter 21, a detailed methodology for establishing
optimal timing of isolated pretimed signals was presented.
In this chapter, a relatively simple methodology for analysis
of a signalized intersection is discussed; in Chapter 24, the
complex model of the Highway Capacity Manual [1] is
presented and illustrated.
Signal timing and analysis of signalized intersections
are simply inverse applications of what should be the same
model. In signal timing, demand flow rates are used with a
known geometric design to establish the green times needed
to accommodate the stated demands. In analysis, the signal
timing, demand flows, and geometry are known, and the
capacity (and/or level of service) of each approach or lane
group is estimated. In essence, signal timing applications
set green times to provide the necessary capacity. In analysis, the known signal timing is used to estimate the existing
capacity.
Unfortunately, signal timing and signalized intersection analysis models are rarely the same and are often quite
different from each other. The signalized intersection
methodology of the 1965 Highway Capacity Manual [2],
for example, used a methodology that was completely
inconsistent with the signal timing methodologies of
the day. This often presented the traffic engineer with an

inexplicable conundrum: A signal timing designed to provide equivalent service to all major movements could produce wildly different levels of service for these same
movements.
Today, the difference is in the level of complexity
included in the models used. The HCM model for signalized
intersection analysis is extremely complex and includes
many iterative elements. It is generally recognized that the
methodology cannot be implemented by hand. Software is
the only practical means to implement the methodology.
Signal timing must be more straightforward to be effective.
Therefore, signal timing methodologies, although based on
the same fundamental concepts as the HCM analysis model,
incorporate many defaults and simplifications to allow for a
relatively rapid and understandable approach to setting an
appropriate timing pattern.
In 1980, the Transportation Research Board issued a
set of preliminary capacity and level of service analysis
methodologies (Interim Materials on Highway Capacity)
in advance of the anticipated 1985 Highway Capacity
Manual [3]. It included a relatively straightforward method
for analysis of signalized intersections called Critical
545
M23_ROES5739_04_SE_C23.QXD
546
6/1/10
8:51 PM
Page 546
CHAPTER 23
CRITICAL MOVEMENT ANALYSIS OF SIGNALIZED INTERSECTIONS
Movement Analysis [4]. The methodology could be easily

implemented by hand in a reasonable amount of time, and
based level-of-service (LOS) determinations on the sum
of critical-lane flows, a concept discussed in Chapter 20.
Although general delay estimates for each LOS for the
intersection as a whole were provided by verbal description, there was no attempt to estimate average delays for
individual movements or approaches.
The earliest work on critical-lane analysis of signalized
intersections goes back to the original concepts of saturation
headways and lost time developed by Bruce Greenshields. Its
application to analysis originally appeared in 1961 in a paper
by Capelle and Pinelle [5]. Messer and Fambro [6] produced a
definitive methodology in 1977 that was adapted for inclusion
in the Interim Materials. By the time the 1985 Highway
Capacity Manual was published, however, Messer had built
the basic procedure into a far more complex analysis methodology. Subsequent revisions have served to further complicate
the approach.
As a result, there are traffic agencies that still (at this
writing) use the methodology of the Interim Materials to
analyze signalized intersections, including the California
Department of Transportation. Because of the great complexity of the HCM approach, there are frequent calls for a
return to a simpler approach similar to that of the Interim
Materials.
23.1 The TRB Circular 212

Methodology
The methodology of TRB Circular 212 (Interim Materials on
Highway Capacity) provides two levels of analysis: planning
and operations/design. The first is done entirely in units
of mixed veh/h with few adjustments. Average or typical
conditions are assumed. The planning approach is as close to
a back of the envelope approach as can be accomplished.
The most complex signalized intersections could be analyzed
in minutes. The operations and design model provides more
detailed adjustments for heavy-vehicle presence, local bus
presence, lane width, parking conditions, and other prevailing
conditions. Even with these adjustments included, the most
complex intersection could be analyzed by hand in less than
15 minutes.
Unlike the HCM, critical-lane or critical movement
analysis applies all adjustment factors to the demand volume
or flow rate. Thus volumes in veh/h are inflated by adjustment
factors to reflect through passenger car units (tpcus), as is at

least partially done for signal timing.
This chapter presents a design and operations methodology based on the concept of the Interim Materials. It has
been updated, however, to reflect more recently calibrated
adjustment factors, and for consistency with the signal timing methodology presented in Chapter 21. A delay prediction
module has been added to the operations/design methodology to allow for better level of service estimates.
Although the operations/design model is not the same
as that of the Interim Materials, the conceptual approach is the
same: Only the numbers have been changed to reflect more
modern conditions. This chapter does not present the planning level analysis of the Interin Materials, which should be
consulted directly for a detailed description of that approach.
23.2 A Critical Movement Approach

to Signalized Intersection
Analysis
The methodology presented in this chapter is conceptually the
same as the operations and design methodology of the 1980
Interim Procedure. It has, however, been updated to be more
consistent with the signal timing methodology of this text and
to reflect adjustment factors of the HCM, where it is appropriate to do so. A module for estimation of delay is added, based
on theoretical models (as opposed to the more complex models
of the HCM).
Unlike the HCM methodology, this critical-lane analysis approach makes all adjustments to demand volumes, as
opposed to capacity or saturation flow rate. In the HCM
methodology, saturation flow rates are reduced to reflect nonideal prevailing conditions. In this methodology, demand
volumes are increased to reflect them.
23.2.1
Steps 1 Through 3
Steps 1 through 3 of the operations and design methodology

are straightforward (1) Identify the lane geometry and use, (2)
identify hourly demand volumes (in veh/h under prevailing
conditions), and (3) specify the signal timing (a complete ring
diagram is required).
In addition to the hourly demand volumes of Step 2,
the prevailing conditions affecting each should be described
23.2
6/1/10
8:51 PM
Page 547
A CRITICAL MOVEMENT APPROACH TO SIGNALIZED INTERSECTION ANALYSIS
in as much detail as possible. At a minimum, the following

characteristics of each movement should be specified:
Proportion of heavy vehicles

Proportion of local buses
Lane widths
Approach grade
Parking conditions on approach
Pedestrian interference levels (either specify pedestrian
volume in each crosswalk, or generally describe by level)
A heavy vehicle is any vehicle with more than four

wheels on the ground, except for local buses, which are handled
separately. In terms of defining buses, a through bus is one that
is not making a stop to pick up or discharge passengers within
the confines of the intersection (either near side or far side); a
local bus is one that does make a stop to pick up or discharge
passengers within the confines of the intersection (near side or
far side). Thus a through bus is treated as a heavy vehicle.
The signal timing specified as part of Step 3 should
include the green, yellow, and all red intervals. For the purpose
of the operations and design approach, vehicles are assumed to
move during the effective green time, which for each phase
may be computed as:
g = G + y + ar - /1 - /2
(23-1)
For the vast majority of cases, the assumption that (y + ar) =

(/1 + /2) may be made. In this case, g = G. In cases where
this assumption is known to be inaccurate, Equation 23-3 is
used to estimate effective green time for each phase.
Step 4: Convert Demand Volumes

to Equivalent Passenger-Car
Volumes
Each movement demand volume is converted to a volume

expressed in equivalent passenger-car units, using the following
equation:
Vpc = (V PHV EHV) + (V PLB ELB)
+ V (1 - PHV - PLB)
where: Vpc = volume in passenger-car equivalents, pc/h

V = volume in vehicles per hour under prevailing
conditions, veh/h
PHV = proportion of heavy vehicles (trucks, through
buses, and RVs) in the movement, decimal
PLB = proportion of local buses in the movement,
decimal
EHV = passenger-car equivalent of one heavy vehicle
in movement
ELB = passenger-car equivalent of one local bus in
movement
The passenger-car equivalent for all heavy vehicles is 2.0.
Passenger-car equivalents for local buses are more complex and
shown in Table 23.1.
Local bus equivalents vary with a number of underlying
conditions, including the percentage of local buses in the traffic stream, the total volume in the affected traffic stream, and
Table 23.1: Passenger-car Equivalents for Local Buses

at Signalized Intersections
% Local
Buses
No. of
Lanes
Volume in Lane Group, veh/h

200 400 600 800
1,000
Buses Stopping in Travel Lane
where: g = effective green time, s

G = actual green time, s
y = actual yellow time, s
ar = actual all-red time, s
/1 = start-up lost time, s
/2 = clearance lost time, s
23.2.2
547
(23-2)
2%
1
2
3
1.8
1.4
1.3
2.7
1.8
1.5
3.5
2.2
1.8
4.4
2.7
2.1
5.3
3.1
2.4
5%
1
2
3
1.8
1.4
1.3
2.7
1.8
1.5
3.7
2.3
1.8
4.8
2.7
2.1
6.0
3.2
2.4
10%
1
2
3
1.9
1.4
1.3
2.9
1.9
1.6
4.2
2.4
1.9
5.7
2.9
2.2
7.7
3.5
2.5
Buses Stopping Off-Line or in Parking Lane

2%
1
2
3
1.3
1.2
1.1
1.7
1.3
1.2
2.0
1.5
1.3
2.4
2.7
2.1
2.7
1.8
1.6
5%
1
2
3
1.3
1.2
1.1
1.7
1.3
1.2
2.1
1.5
1.3
2.4
1.7
1.5
2.8
1.9
1.6
10%
1
2
3
1.3
1.2
1.1
2.1
1.5
1.3
2.5
1.7
1.5
3.0
1.9
1.6
1.7
1.3
1.2
548
6/1/10
8:51 PM
Page 548
CHAPTER 23
the number of lanes in the affected lane group. Exclusive LT

lanes are generally not included as affected lanes.
The type of bus stop also affects these equivalents. If
the bus stops in a travel lane, the impact of the blockage is
more severe than when the bus stops in a parking lane or in an
offline bus bay. In general, local buses disrupt traffic primarily in the right-most lane of the affected lane group. Thus the
overall impact of buses is reduced as the number of lanes in
the lane group increases.
It is noted that the equivalents of Table 23.1 are generally
based on the implied equivalents of the HCM signalized intersection analysis procedure. It is assumed that a bus stopping in a
travel lane effectively blocks it for 14.4 seconds. A bus stopping in
an offline position is assumed to block the right-most lane for 6.0
seconds as it enters and leaves the offline bus bay or parking lane.
23.2.3
Step 5: Convert Passenger-Car

Equivalents to Through-Car
Equivalents
After completing Step 4, all movement volumes have been converted to passenger-car equivalents. Now, left-turning vehicles
and right-turning passenger-car volumes must be converted to
equivalent through-car units.
Through-car equivalents for left-turning passenger cars
are given in Table 23.2. These are the same equivalents used in
timing traffic signals (Chapter 21). All protected left turns have
an equivalent of 1.05, whereas the equivalents for permitted left
turns depend on the opposing volume through which they turn
and the number of lanes on the opposing through lane group.
Table 23.3 shows the through-car equivalents for rightturning passenger cars. These depend on the pedestrian volume
in the conflicting crosswalk.
Table 23.2: Through-Car Equivalents for Left-Turning
Vehicles, ELT
Opposing Flow
Vo (veh/h)
0
200
400
600
800
1,000
1,200
Number of Opposing Lanes, No

1
2
3
1.1
1.1
1.1
2.5
2.0
1.8
5.0
3.0
2.5
10.0*
5.0
4.0
13.0*
8.0
6.0
15.0*
13.0*
10.0*
15.0*
15.0*
15.0*
ELT for all protected left turns = 1.05.

*The LT capacity is only available through sneakers.
Table 23.3: Through-Car Equivalents for Right-Turning

Vehicles, ERT
Pedestrian Volume in Conflicting
Crosswalk (peds/h)
Equivalent
None (0)
Low (50)
Moderate (200)
High (400)
Extreme (800)
1.18
1.21
1.32
1.52
2.14
It is appropriate to interpolate for intermediate values in

Table 23.2, but this is not suggested for the values in Table 23.2,
which are more approximate. The values in these tables are
much simplified versions of adjustments in the HCM 2000.
They eliminate the need to iterate solutions and greatly simplify
the consideration of left and right turns.
Special Case: Compound LT Phases (Protected
Permitted and Permitted Protected)
It is difficult to analyze compound phases in any detail in a
simplified analysis approach such as CMA. In the HCM
methodology (Chapter 24), protected and permitted portions
of the phase are initially treated as two separate phases. This,
in turn, requires that the left-turn demand flow be split
between the two portions of the compound phase.
Two potential approaches for such an allocation are to
(1) assume the full capacity of the initial portion of the phase is
fully consumed with all remaining left turns assigned to the second
portion of the compound phase, and (2) assume left-turn demand
is assigned to the two portions of the phase in proportion to the
capacities of two portions of the phase. Both are iterative because
the initial assignment will influence the total tcu/h, which when
allocated most often will violate the assumed initial distribution.
A simple methodology must find some more simplistic
approach. The approach recommended here is that the entire
left-turn demand flow be converted to tcu/h using a single
equivalent that is an average of the equivalents for the protected
portion of the phase and the permitted portion of the phase. It is
further recommended that the average be approximately
weighted using the relative green times in the compound phase.
Thus the left-turn equivalent, ELT, would be estimated as:
ELTC =
A ELTPT * gLTPT) + (ELTPM * gLTPM B

gLTPT + gLTPM
(23-3)
where: ELTC = left-turn equivalent for compound LT phasing

ELTPT = left-turn equivalent for protected portion of
the compound LT phase
23.2
6/1/10
8:51 PM
Page 549

ELTPM = left-turn equivalent for permitted portion of
the compound LT phase
gLTPT = effective green time for protected portion of
the compound LT phase, s
gLTPM = effective green time for permitted portion of
the compound LT phase, s
This average equivalent is then used to convert the full

left-turn flow to tcu/h. For compound phases, this approach is
highly approximate but keeps the analysis in the range of a
simple and straightforward methodology. Where more detailed
analysis is required, the HCM methodology would be applied
directly. Note that the effective green time for the first portion
of the compound phase includes the yellow and all red intervals between the two portions.
23.2.4
Step 6: Converting Through-Car

Equivalents Under Prevailing
Conditions to Through-Car
Equivalents Under Ideal Conditions
At this point in the methodology, all movement volumes have

been converted to volumes in through-car equivalent units
(tcu/h). They still represent, however, a number of prevailing
conditions noted previously. Each movement volume must
now be converted to a flow rate (for a 15-minute period) under
equivalent ideal conditions. This conversion is computed as:
v =
Vtcu
PHF * fw * fg * fp * f LU
(23-4)
where: v = movement flow rate under equivalent ideal conditions, tcu/h

Vtcu = movement volume, tcu/h, reflecting some prevailing conditions
PHF = peak-hour factor
fw = adjustment factor for lane width (see Table 23.4)
fg = adjustment factor for grade (see Table 23.5)
fp = adjustment factor for parking (see Table 23.6)
fLU = adjustment factor for lane utilization (see
Table 23.7)
Table 23.4: Adjustment Factors for Lane Width

Lane Width (ft)
fw
10
11
12
13
14
15
0.90 0.93 0.97 1.00 1.03 1.07 1.10
549
Table 23.5: Adjustment Factors for Grade

Grade (%)
-6
-4
fg
1.03
1.02
-2
1.01 1.00
0.99
0.98 0.97
Table 23.6: Adjustment Factors for Parking Conditions

Parking Activity
High
Medium
Low
No parking lane
Number of Lanes
in Lane Group
Adjustment
Factor fp
1
2
3
1
2
3
1
2
3
All
0.825
0.913
0.942
0.850
0.925
0.950
0.875
0.938
0.958
1.000
Table 23.7: Adjustment Factors for Lane Utilization

Lane Group
Movements
Through or
shared
Exclusive left
turn
Exclusive right
turn
No. of Lanes
in Lane Group
Adjustment
Factor fLU
1
2
3
1
2
1
2
1.000
0.952
0.908
1.000
0.971
1.000
0.885
(Source: Highway Capacity Manual, 3rd Edition, Transportation

Research Board, Washington DC, 2000, modified from Exhibit 10-23,
p. 1026.)
The peak-hour factor (PHF) converts a full-hour

volume to a flow rate within the worst 15 minutes of the
hour under consideration (usually one of the peak hours of
the day). The adjustment factors account for nonstandard
conditions of lane width (12-foot lanes are the base condition), grade (level grade is the base condition), parking
activity adjacent to the lane group (no parking is the base
condition), and lane use (uniform use of lanes is the base
condition).
550
6/1/10
8:51 PM
Page 550
CHAPTER 23
Lane widths less than 9 feet are not recommended; in

fact, lane widths less than 11 feet are problematic where any
appreciable number of heavy vehicles are present. A lane
width of 16 feet or more will be used by drivers as two eightfoot lanes under the pressure of heavy traffic.
The parking adjustments of Table 23.6 focus on curb
parking spaces within 250 feet of the stop line of the intersection approach. If there is no curb parking on the last 250
feet of the approach, there is no adjustment to be made. The
extent of the adjustment is related to the number of parking
maneuvers made (expressed as an equivalent hourly rate)
and the number of lanes in the affected lane group. The
number of movements (each parking maneuver and each
unparking maneuver is separately counted) is generally
characterized as low (0 to 7 mvts/h), medium (8 to 13
mvts/h), and high (more than 13 mvts/h) for the purposes
of this methodology. Because parking maneuvers mainly
affect the right-most lane (adjacent to the parking lane),
the impact is dissipated by having multiple lanes for
moving vehicles.
The lane utilization adjustment recognizes that where
multiple lanes are present, vehicles will not divide themselves
equally among all available lanes. The factor increases the
equivalent ideal flow rate such that when the adjusted demand
flow rate is divided equally into all available lanes of the lane
group, the resulting average will represent the largest expected
lane flow within the lane group.
23.2.5
Assigning Lane Flow Rates
23.2.6
Finding Critical-Lane Flow Rates

for Each Signal Phase
Simple Case: No Overlapping or Compound Phases

The process of finding the critical-lane flow rate for each
signal phase is relatively straightforward where there are no
overlapping or compound phases. All lane flows moving
during a given phase are examined. The critical-lane flow is
the lane with the highest flow rate in adjusted tcu/h. When all
critical-lane flows are identified, they are added to determine
the sum of critical-lane flow rates. There will be one criticallane flow rate for each phase of the signal.
Overlapping Phases
Finding the critical-lane flow rates for a signalization with
overlapping phases must use the same technique described in
the signal timing methodology of Chapter 21. A ring diagram
of the signal timing must be drawn, with the appropriate
critical-lane flow rates inserted into the proper parts of the
diagram. An example is shown in Figure 23.1.
The ring diagram of Figure 23.1 depicts a standard
NEMA phase sequence in which exclusive LT phases are
followed by leading green phases. Each box of the ring
Ring 1
148*
Each movement volume is now expressed in through-car units

under equivalent ideal conditions. It is now time to assign lane
flows within each lane group. The following rules govern this
assignment:
1. Where a separate LT lane (or lanes) exist, assign
all LT tcus to this lane group. If more than one
lane exists, divide the tcu/h equally among the
lanes.
2. Where a separate RT lane (or lanes) exist, assign
all RT tcus to this lane group. If more than one
lane exists, divide the tcu/h equally among the
lanes.
3. For all mixed lane groups (i.e., LT/TH/RT, LT/TH, or
TH/RT) divide the total tcu/h equally among the
lanes, except that (a) all LT tcus must be in the lefthand lane, and (b) all RT tcus must be in the righthand lane.
Ring 2
203
A1
A2
420*
330
120
PHASE
220*
380
A3
B1
B2
250*
B3
Figure 23.1: Sample Ring Diagram for Overlapping

Phases
23.2
6/1/10
8:51 PM
Page 551
diagram shows the critical-lane flow rate (under equivalent

ideal conditions) for the movements shown.
The critical path through the ring diagram is that path
that results in the highest sum of critical-lane flow rates. It is
found as follows:
For Phases A1 to A3, the critical path can run down
Ring 1 or Ring 2. The Ring 1 sum of critical-lane
flow rates is 148 + 420 = 568 tcu/h. The Ring 2 sum
is 203 + 330 = 533 tcu/h. The critical path runs
through Ring 1 and the sum of the critical-lane flow
rates for Phases A1 through A3 is 568 tcu/h.
For phases B1 to B3, the same choice is available.
The Ring 1 sum of critical-lane flow rates is 120 +
380 = 500 tcu/h. The Ring 2 sum is 220 + 250 = 570
tcu/h. Ring 2 has the largest sum and therefore represents the critical path. The sum of critical-lane flow
rates for Phase 2 is 570 tcu/h.
The critical path through the signal timing is indicated by
asterisks, and the sum of critical-lane flow rates for this example is 568 + 570 = 1,138 tcu/h.
Examination of the ring diagram identifies the critical
path through the signal timing, and it also depicts the criticallane flow rate for each phase and lane group in the intersection.
where: ci = capacity of critical lane in lane group i, tcu/h/ln

gi = effective green time for lane group i, s
C = cycle length, s
In the case of compound phasing, the same equation is used,
employing the effective green time for the entire compound
phase (including the yellow and all red between the two portions of the phase). This is possible because the demand flow
rates have been adjusted to reflect the different values of ELT
that apply in each portion of the phase. Capacity is stated in
terms of tcu/h under ideal conditions.
The maximum sum of critical-lane flow rates that
can be accommodated by the signalized intersection as a
whole is:
where: cSUM
gi
n
C
=
=
=
=
cSUM = 1900 P
a gi
i=1-n
(23-6)
maximum sum of critical-lane flow rates, tcu/h

effective green time for critical phase i, s
number of phases
cycle length, s
Once capacities for all critical lanes are determined, a

v/c ratio can be computed for each:
Compound Phases
When a protected + permitted or permitted + protected LT
phase exists, it will be necessary to divide the demand between
the two portions of the phase to determine the sum of criticallane flows. The total LT flow in these cases has been converted
to tcu/h using an average ELT reflecting both portions of the
phase. The converted flow can be divided in the proportion to
the length of the effective green times for each portion of the
phase. Note that a yellow transition between a leading
protected phase to a permitted phase would be counted as
effective green in this distribution.
23.2.7
551
Capacity and v/c Ratios
At this point in the analysis, the capacity of each critical lane

must be estimated and compared to the demand flow rate
using the v/c ratio.
Because all demand flow rates are expressed in tcu/h
under equivalent ideal conditions, capacities must be stated in
the same units. The recommended ideal saturation flow rate
used to make these estimates is 1,900 tcu/h. The capacity of a
critical lane within a lane group is:
ci = 1900 A giC B
(23-5)
Xi =
vi
ci
(23-7)
where: Xi = v/c ratio for critical lane i

vi = demand flow rate in critical lane i, tcu/h
ci = capacity of critical lane i, tcu/h
In the case of compound phasing, a single value of X is computed based on the sum of the component phase demand
flows, v, and the capacity of the compound phase, taken as a
whole.
A critical v/c ratio for the entire intersection may also
be computed as:
Xc =
a vci
i = 1 - n c
SUM
(23-8)
where: Xi = critical v/c ratio for intersection

vci = critical-lane flow rate i (in critical signal path), tcu/h
cSUM = sum of critical-lane flow rates, tcu/h
Note that the v/c ratios produced in this step are analogous to those produced by the HCM 2000 methodology. The
only difference is that several critical computations (notably
the impact of turning movements) have been greatly simplified
552
6/1/10
8:51 PM
Page 552
CHAPTER 23
to avoid multiple iterations and to allow for a manual solution

if desired.
At this point, each critical lane has been identified, and the
sufficiency of the capacity provided to handle it (v/c ratio) has
been assessed. A v/c ratio in excess of 1.0 signifies that the
capacity is insufficient to accommodate the stated demand.
A v/c ratio of 1.00 or less indicates that the capacity provided is
sufficient to accommodate the demand. Although these computations are all based on a single critical lane within the intersection, when the v/c ratio is 1.00 or less, this means that all other
lanes moving during that phase also have sufficient capacity.
The critical v/c ratio, Xc, is an important variable. If Xc is
less than 1.00, it signifies that the existing signal cycle length,
phase sequence, and the physical configuration of the intersection and its approaches are sufficient to handle all of the stated
demand flows. If Xc is more than 1.00, the situation cannot be
rectified without doing one or more of the following:
Increasing the cycle length,
Making the phase sequence more efficient (generally
means adding protected or compound left-turn phasing),
and/or
Adding lanes to critical approaches and/or changing
lane use restrictions for greater efficiency.
It is possible, however, that Xc is less than 1.00, but that
some individual lane groups have Xi values that exceed 1.00. In
this case, the difficulty can be resolved by reallocating green time
within the existing phase sequence, cycle length, and geometric
design. To do this, the methodology for signal timing found in
Chapter 21 can be used, or (alternatively) the methodology for
readjusting signal timing found in Chapter 24. The latter uses
values of Xi directly, as opposed to critical-lane flow rates.
23.2.8
Delay and Level of Service
The planning approach estimated LOS on the basis of the sum of

critical-lane volumes. Although this is sufficient for the most
approximate of analyses, it contains a basic flaw. The presumption is that lower sums of critical-lane volumes produce higher
quality operations. Although not exactly the same, use of the sum
of critical-lane volumes is analogous to using v/c ratios (Xi) to
determine level of service, and it implies that lower v/c ratios
yield better operations.
The difficulty is that for signal timing, this is essentially
not correct. In the analysis of uninterrupted flow facilities,
lower v/c ratios lead directly to lower densities and higher
speeds (i.e., higher quality operations). The critical difference
between interrupted and uninterrupted flow, however, is in the
determinants of capacity. For uninterrupted flow, capacity is

based solely on the number of lanes present and the specifics
of geometry (for a given traffic composition). For interrupted
flow, it is based on the number of lanes present, the specifics
of geometry, and the timing of the signal. In uninterrupted
flow, we strive for low v/c ratios to allow for growth in traffic
over time and to ensure that the road we build has sufficient
space (lanes) to serve traffic for 20 to 30 years. When the time
element is considered, however, the bottom line is a constant:
There are 3,600, and only 3,600 seconds in an hour. There
will never be more; there will never be less. Thus, although it
is logical to provide excess capacity in terms of space,
it is quite illogical to provide significant amounts of it in a
signal timing.
If delay is used as the principal measure of LOS at intersections, very low v/c ratios often produce relatively high delays
because vehicles are waiting on a RED signal while there are no
vehicles present to use the GREEN on the other street. Unused
green time is a cause of delay at a signalized intersection, not a
cure for it. In general, v/c ratios at signalized intersections will
be relatively high when optimal delay is achievedoften in the
range of 0.80 0.95. Large amounts of unused green time occur
when v/c ratios are low, and they generally indicate that the
cycle length and green times are too long. Cycle lengths that
are too long increase delay. On the other side of the coin, cycle
lengths that are too short, pushing the v/c ratio close to 1.00, also
increase delaythis time because of individual cycle failures
that are likely to occur.
This methodology suggests that levels of service be directly
related to estimated delay, just as in the more complex HCM
methodology. It requires, therefore, that delays be estimated for
each lane group in the intersection. To avoid the complexity of the
HCM methodology, a more simplistic approach is taken using
theoretical equations (see Chapter 20), with a simplified adjustment for progression quality.
The delay for a lane group is computed as:
di = d1i * PF + d2i
(23-9)
where: di = approach delay for lane group i, s/veh

d1i = uniform delay for lane group i, s/veh
d2i = overflow plus random delay for lane group i,
s/veh
PF = progression adjustment factor
This equation is analogous to that used in HCM 2000, except
for the following:
It predicts approach delay, not control delay. The
difference is slight. Approach delay is the sum of
stopped delay plus acceleration and deceleration
23.2
6/1/10
8:51 PM
Page 553

delay. Control delay is time-in-queue delay plus
acceleration and deceleration delay. The difference is
the time spent in queue while moving.
It assumes there is no unserved queue present at the
beginning of the analysis period. The HCM method
adds a third element of delay when such queues are
present.
The progression factor recommended in this methodology is a much simplified version of those factors
than used in the HCM.
The equation for uniform delay is the same as in the
HCM 2000. The equation for overflow delay is based
on Akceliks theoretical approach [7] as modified in
the HCM 2000. Some simplifications of terms are
introduced.
The uniform delay term, d1, is computed using

Websters original delay equation:
d1i =
0.5 C C 1 -
AgiCBD 2
1 - Cmin A 1, Xi B *
AgiCBD
(23-10)
where all terms are as previously defined. The min function

is used to emphasize that for uniform delay, the value of Xi
cannot be higher than 1.00.
The overflow delay term is computed as follows:
d2i = 225
i
2
J A Xi - 1 B + C A Xi - 1 B + ci Ni K (23-11)
16 X
where Ni is the number of lanes in lane group i. This is essentially

the HCM formula for the combination of overflow and random
delay. Several things have been modified for this methodology:
The analysis period is automatically set to 15 minutes
(0.25 hour).
Capacities ci are computed per lane in this methodology. They are multiplied by Ni to provide a total lane
group capacity, as needed for the HCM equation.
From the HCM equation, adjustment factor k is set to
0.50, and adjustment factor I to 1.00.
Recommended progression factors (PF) are shown in
Table 23.8. These are taken from the HCM 2000 but are very
much simplified for this methodology. Instead of six arrival types
specified in the HCM, this methodology uses only three terms:
Good progression: Platoons arrive within the green
phase, and the majority of vehicles move through the
green without stopping.
553
Table 23.8: Progression Adjustment Factors for Lane

Group Delay
Progression Factor (PF)
Quality of
Progression
Pretimed Signals
Actuated Signals
Good
Random
Poor
0.70
1.00
1.25
0.60
0.85
1.06
Random arrivals: The signal is isolated, or uncoordinated with adjacent signals; the proportion of arrivals
on red and green approximates the proportions of red
and green in the cycle.
Poor progression: Platoons arrive when the signal is
red, and/or the majority of vehicles must stop before
proceeding through the intersection.
These definitions are somewhat imprecise, and they rely
on the ability of the traffic engineer to make appropriate
judgments on observing the arrival pattern or studying the
time-space diagram(s) for a future situation.
The HCM varies this factor with g/C. To simplify for
this methodology, values for a g/C of approximately 0.50 are
used. For actuated signals, an additional adjustment if 0.85 is
also included to roughly estimate the greater efficiency of
these in allocating green time.
Once lane-group delays have been estimated, Table 23.9 is
used to determine the level of service. A separate level of service
is assigned to each lane group in the analysis. Where there is
more than one lane group on an approach (such as in the case of
an exclusive LT lane group or exclusive RT lane group), delays
may be aggregated to obtain an average for the entire approach
and a LOS assigned to the approach. Aggregation of approaches
to determine an overall intersection delay and level of service is
Table 23.9: Levels of Service for Signalized Intersection
Lane Groups and Approaches
Level of Service
Delay (s/veh)
A
B
C
D
E
F
10
7 1020
7 2035
7 3555
7 5580
7 80
(Source: Used with permission of Transportation Research Board,

National Research Council, Highway Capacity Manual, 4th Edition,
Washington DC, 2000, Exhibit 16-2, p. 16-2.)
6/1/10
554
8:51 PM
Page 554
CHAPTER 23
not recommended because this could easily mask serious problems on a particular approach or lane group.
Equation 23-12 is used to compute an average delay for
an approach with more than one lane group.
a vi di
dapp =
(23-12)
a vi
where: dapp = approach delay, s/veh

di = delay for lane group i, s/veh
vi = demand flow for lane group i, pc/h
23.2.9
A Worksheet for Critical

Movement Analysis
Although the critical movement analysis method recommended in this section is far less complex than the HCM it is
relatively detailed and includes many steps. Because of this a
worksheet is provided to help guide computations. It is shown
in Figure 23.2. Instructions for the use of this three-page
worksheet are given here:
1. Enter all identifying information concerning the
location and time period of analysis in the top portion
of page 1.
2. In the space provided (A), sketch a diagram of the
intersection with demand volumes shown. Indicate
(in parentheses) the percentage of heavy vehicles
and percentage of local buses in each volume
shown.
3. In the space provided (B), sketch a ring diagram of the
signal timing, with complete timing details shown.
4. In the space provided (C), enter pedestrian flows for
each crosswalk in peds/h, or generally characterize
pedestrian flows as none, low, moderate, high, or
extreme.
5. In the space provided (C), for each approach, enter the
lane width (feet) and grade (%). Characterize parking
activity as high, medium, low, or no parking lane.
6. Part D of the spreadsheet is on page 2 and consists
of converting demands in veh/h to pc/h expressed as
a flow rate under equivalent ideal conditions:
a. For each approach and movement, enter the
number of passenger cars (PC), heavy vehicles
(HV), and local buses (LB). In the next line, enter
the equivalents for PC, HV, and LB. All PCs have
an equivalent of 1.000 (already shown on the

worksheet), and all HVs have an equivalent of
2.000 (already shown on the worksheet). The
equivalent for LBs is obtained from Table 23.5.
b. Multiply all demand volumes in veh/h by the
appropriate equivalent to determine the volume
in pc/h. From this point on, only the Total
column for each movement is pursued.
c. For each movement, enter the left-turn equivalent
from Table 23.6 or the right-turn equivalent from
Table 23.7. The equivalent for through vehicles is
always 1.000. Where a compound left-turn phase
exists, use Part (E) of the worksheet to estimate
the weighted average equivalent for use.
d. For each movement, multiply the volume in pc/h
by the appropriate equivalent to obtain the
volume in tcu/h.
e. For each movement, enter the following adjustment factors: lane width (Table 23.4), grade
(Table 23.5), parking (Table 23.6), lane utilization
(Table 23.7), and the peak-hour factor (PHF).
Divide the volume in tcu/h by all of these factors
to estimate the demand flow in ideal tcu/h.
7. In the space provided (F), sketch the intersection,
and assign lane flow rates as appropriate to each
lane shown.
8. In the space provided (G) sketch the signal ring
diagram, assigning the appropriate critical-lane flow
rate to each cell of the diagram. Find the critical
path through the ring diagram and determine the
sum of critical-lane flows.
9. Go to page 3 of the worksheet (H). The worksheet
allows for up to 12 lane groups (one for each movement). In most cases, there will be fewer because
many movements will share lanes. Where left turns
and/or right turns share lanes with through
movements, draw an arrow in the LT and/or RT cells
pointing to the through lanes, where the appropriate
critical-lane flow rate will be shown. Where separate
LT and/or RT lanes exist, show the critical flow rate
for each separately in the appropriate column.
10. Compute and enter the maximum sum of criticallane flow rates and the critical v/c ratio as shown (H).
11. For each lane group, compute delay (I) as shown, and
determine the level of service for each lane group
(Table 23.9). Where desired, aggregate multiple lane
groups to obtain an overall delay and LOS for the
approach.
23.2
6/1/10
8:51 PM
Page 555
Worksheet for CMA: Operations and Design Method
555
Pg 1
LOCATION:
TIME PERIOD OF ANALYSIS:
TO
ANALYST:
DATE:
(B) Ring Diagram with Signal Timing
(A) Diagram of Site and Demand Volumes

(Show % HV and % LB)
(C) Pedestrian Flows Conflicting With:
EBRT
NBRT
WBRT
SBRT
Pkg Act
Other Characteristics:
Ln Width Grade
EB
WB
NB
SB
(D) Conversion of Volumes to Equivalent Ideal Flow Rates in tcus/h (See Pg 2 of Worksheet)
(E) Compound Phase LT Equivalents
ELTC =
COMPOUND PHASE LT EQUIVALENTS

SB
WB
EB
NB
(ELTPT * gLTPT) + (ELTPM * gLTPM)

gLTPT + gLTPM
(F) Diagram with Lane Flows
(G) Ring Diagram with Critical Lane Flows
Sum of Critical Lane Flows:
Figure 23.2: Worksheet for Critical Movement AnalysisOperations and DesignPage 1
6/1/10
556
8:51 PM
Page 556
CHAPTER 23

Item
PC
EB
veh/h
equiv (Tab.23.1)
pc/h
ELT/ERT (Tab. 23.2/3)
tcu/h
fw (Tab. 23.4)
fg (Tab. 23.5)
fp (Tab. 23.6)
fLU (23.7)
PHF
ideal tcu/h
WB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
NB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
SB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
LEFT
HV
LB
1.000
2.000
PC
HV
LB
1.000
2.000
PC
HV
LB
1.000
2.000
PC
HV
LB
1.000
2.000
THROUGH
HV
LB
TOTAL
PC
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
Pg 2
RIGHT
HV
LB
TOTAL
PC
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
TOTAL
PC
HV
LB
1.000
2.000
Figure 23.2: (Continued) Worksheet for Critical Movement AnalysisOperations and DesignPage 2
TOTAL
TOTAL
TOTAL
TOTAL
23.2
6/1/10
8:51 PM
Page 557
557
Pg 3
(H) Critical Lane Computations

LANE GROUP
EBLT
Critical Lane Flow, vi
Capacity of Critical Lane, ci
v/c Ratio of Critical Lane, Xi
EB
EBRT
WBLT
WB
WBRT
NBLT
NB
NBRT
SBLT
SB
SBRT
EBRT
WBLT
WB
WBRT
NBLT
NB
NBRT
SBLT
SB
SBRT
Maximum Sum of Critical Lane Flows:

gi
i=1n
cSUM = 1900
C
Critical v/c Ratio:
vci
Xc = i = 1 n
cSUM
(I) Delay and Level of Service Computations

LANE GROUP
Input Variables:
Cycle Length, C
Effective Green, g
v/c Ratio, X
Number of Lanes, N
Capacity, c
Delay Computations:
Uniform Delay, di
Prog. Factor, PF (Tab. 23.12)
EBLT
EB
Overflow Delay, d2
Total Delay, d
Level of Service (Tab. 23.13)
Approach Delay
Approach LOS (Tab. 23.13)
DELAY EQUATIONS:
di = d1i * PF + d2i
Q5C 1
gi
d2i = 225 (Xi 1)+ (Xi 1)2+16 Xi

ci N i
d1i =
g
1 min(1, Xi)* i
Figure 23.2: (Continued) Worksheet for Critical Movement AnalysisOperations and DesignPage 3
6/1/10
558
8:51 PM
Page 558
CHAPTER 23
23.2.10 Summary
The critical movement analysis (CMA) methodology recommended in this section provides the same level of results as
does the HCM. Its major simplifications include (1) greatly
simplifying the determination of left- and right-turn adjustments, (2) greatly simplifying the approach to analysis of
compound (protected + permitted or permitted + protected
phasing), and (3) greatly simplifying delay computations.
Another significant difference between this method and the
HCM is that many of the adjustments are applied to the volume side of the equation, not the saturation flow rate (and
therefore, capacity) side of the equation. This, however, is a
difference of form, not substance.
These simplifications will mean that the CMA
approach is less sensitive to the details of these portions of
the analysis. Many of the complex additions to the HCM
methodology have not been tested against field data, so
although additional sensitivity is provided, no clear evidence indicates that the results are more accurate than a
simpler approach. It is also true that the recommended
approach here was not tested against field data either.
Often, however, analysis of signalized intersections is
done to test the impacts of various proposed schemes and
approaches. The results of interest are the changes that
occur in delay (and therefore LOS) and in v/c ratios. It is
believed that the recommended approach is sufficiently
detailed to reflect the vast majority of signal timing
designs and physical designs that might be compared
using it.
23.3 Sample Problems Using

Critical Movement Analysis
23.3.1
Sample Problem 1: A Relatively

Simple Problem
approach. The one-way street has three lanes. There is no

parking on any of the approaches, but the EW arterial has a
bus stop serving 20 buses per hour.
The information in Figure 23.3 is more detailed than
would normally be necessary for a CMA solution. It is given
so that the more detailed analysis using the HCM methodology (Chapter 24) can also be applied, and the results compared. Key information for this solution is the % heavy
vehicles, local bus volumes, pedestrian flows, lane widths,
grades, and the arrival type. For this methodology, the
arrival type cited is equivalent to a random progression
quality.
The signalization is typical for a three-leg intersection
with only one opposed left turn: a leading green for the WB
approach, followed by a EW through/right turn phase, and a
phase for the NB approach.
Placing Input Data on Worksheet
Part (A) through (C) of page 1 of the worksheet provides
space to inscribe the input information onto the worksheet.
Note that the intersection diagram is simplified because a
more detailing drawing is available in Figure 23.3.
Lane Group Decisions
Before starting, it is always beneficial to establish the lane
groups to be analyzed. In this case, identifying lane groups is
relatively easy. First, there are only three approaches. There is
no SB approach on the one-way street. All vehicles flow from
shared lanes on the WB and NB approaches, so each approach
constitutes a single lane group. Because an exclusive LT
lane exists on the EB approach, it will be treated as a separate
lane group, with through and right-turning vehicles sharing
the two lanes of the primary EB lane group. At the end of
the analysis, a weighted average of delays on the two EB
lane groups can be computed to characterize the entire
EB approach.
Intersection Description
Entering Demand Volumes (veh/h)
Figure 23.3 shows a relatively simple intersection that will be

analyzed using the recommended CMA methodology. It is
noted that this same problem is repeated in Chapter 24, using
the HCM methodology, so that the results may be compared.
Figure 23.4 shows the worksheet with computational results
summarized.
Figure 23.3 shows the intersection of a two-way arterial
(EW) with a one-way arterial (NB). The two-way arterial has
two lanes in each direction, with a left-turn lane in the WB
Demand volumes are entered on the top line of Part (D), page 2,
of the worksheet for each movement. Note that volume is
entered separately for the three classes of vehicles for each
movement on each approach. This is done to simplify the
conversion computations that follow.
For example, the EB through movement is 1,100 veh/h.
It contains 10% heavy vehicles and 20 buses per hour. Thus,
there are 1,100 * 0.10 = 110 heavy veh/h, 20 local buses/h
(given), and 1,100 - 110 - 20 = 970 pc/h.
23.3
6/1/10
8:51 PM
Page 559
SAMPLE PROBLEMS USING CRITICAL MOVEMENT ANALYSIS
PHF = 0.92
10% trucks (all movements)
Non-CBD location
100 peds/h in each xwalk
xwalks = 8 ft (2 ft set-back
from curb line)
Arrival Type 3 (all lane groups)
No pedestrian signals
559
120
950
Level Grade
2 @ 12 ft
5 @ 11 ft
7 ft
2 @ 12 ft
Level Grade
1000
200
N
1100
50
+3% Grade
150
Bus Stop
20 buses/h
3 @ 13 ft
Phasing Diagram
A1: G = 10 s
y = 2 s, ar = 2 s
A2: G = 22 s
y = 2 s, ar = 2 s
B: G = 16 s
y = 2 s, ar = 2 s
Figure 23.3: Intersection for Sample Problem 1
Conversions to Equivalent Passenger Cars (pc/h)
Conversions to Through-Car Equivalents (tcu/h)
Equivalents are entered into the next line of Part (D). Passenger
cars always have an equivalent of 1.000, heavy vehicles always
have an equivalent of 2.000. The local bus equivalents are found
in Table 23.1. The 20 buses/h in the EB direction operate in a
two-lane group, with a lane group volume of 1,100 veh/h. The
20 buses represent (20/1,100) 100 = 1.82% of the lane-group
demand. The buses stop in a travel lane. From Table 23.1, this is
an equivalent of 3.1. The equivalent for WB local buses is found
in a similar way from Table 23.1. Each equivalent is multiplied
by the volume to obtain the passenger-car equivalents for each
movement, which are then totaled. The results are entered
into the third line (for each approach) of Part (D) of the worksheet. At this point, for each movement, only the total columns
are used.
Left-turn equivalents (Table 23.2) and right-turn equivalents (Table 23.3) are now entered. There are two left turns.
The EB left turn is fully protected and has an equivalent of
1.05. The NB left turn is a special case. Because it is made
from a one-way street, it is, by definition, unopposed. It is
not protected in the sense that it has no separate phase.
Moreover, left-turners must immediately go through a
pedestrian crosswalk, just like right-turning vehicles. In this
case, it is appropriate to treat it as another right turn, taking
the equivalent from Table 23.3. All crosswalks have a
pedestrian volume of 100 peds/h. No interpolation is
permitted in this table, so the value for 50 peds/h is used
(closest tabulated volume): The equivalent for all right
turns, therefore, is 1.21.
6/1/10
560
8:51 PM
Page 560
CHAPTER 23

LOCATION:
Pg 1
SAMPLE PROBLEM 1
TIME PERIOD OF ANALYSIS: 8:00 AM
TO 9:00 AM
DATE: September 22, 2008
ANALYST: Roger P. Roess

120
200
Phase A1:
G = 10 s
y = 2 s; ar = 2 s
950
Phase A2:
G = 22 a
y = 2 s; ar = 2 s
1100
1000
150
50
Phase B:
G = 16 s
y = 2 s; ar = 2 s
10% HV (all mvts)

EB: 20 LB/h
WB: 20 LB/h

EBRT
NBRT
No EBRT
100 peds/h
EB
WB
NB
SB
Pkg Act
None
None
None
x
WBRT
SBRT
100 peds/h
No SBRT
Ln Width Grade
11 ft
Level
12 ft
Level
13 ft
3%
x
x
(E) Compound Phase Computations
ELTC =

SB
WB
EB
NB

gLTPT + gLTPM
NO COMPOUND PHASE IN THIS PROBLEM.

183
529
Phase A1:
G = 10 s
y = 2 s; ar = 2 s
712
259
712 + 259 = 971
Phase A2:
G = 22 a
y = 2 s; ar = 2 s
259
737
712
737
737
79
458
538
302
236
Phase B:
G = 16 s
y = 2 s; ar = 2 s
538
538
Sum of Critical Lane Flows:
Figure 23.4: CMA Worksheet for Sample Problem 1Page 1
971+538 = 1509 tcu/h
23.3
6/1/10
8:51 PM
Page 561
561

Item
EB
WB
NB
SB
PC
veh/h
180
equiv (Tab.23.1)
1.000
180
pc/h
tcu/h
fw (Tab. 23.4)
fg (Tab. 23.5)
fp (Tab. 23.6)
fLU (23.7)
PHF
ideal tcu/h
PC
veh/h
N/A
1.000
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
PC
45
veh/h
1.000
equiv (E)
pc/h
45
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
PC
N/A
veh/h
1.000
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
LEFT
HV
LB
20
0
2.000
N/A
40
0
HV
LB
N/A
N/A
2.000
HV
5
2.000
10
HV
N/A
2.000
LB
0
N/A
0
LB
N/A
TOTAL
200
220
1.05
231
0.97
1.00
1.00
1.00
0.92
259
TOTAL
N/A
TOTAL
50
55
1.21
67
1.03
0.985
1.00
0.908
0.92
79
TOTAL
N/A
PC
970
1.000
970
PC
835
1.000
835
PC
900
1.000
900
PC
N/A
1.000
THROUGH
HV
LB
110
20
2.000
3.100
220
62
HV
LB
20
95
2.000
3.100
62
190
HV
LB
100
0
N/A
2.000
200
0
HV
LB
N/A
N/A
2.000
Figure 23.4: (Continued) CMA Worksheet for Sample Problem 1Page 2
TOTAL
1100
1252
1.00
1252
0.97
1.00
1.00
0.952
0.92
1474
TOTAL
950
1087
1.00
1087
1.00
1.00
1.00
0.952
0.92
1241
TOTAL
1000
1100
1.00
1100
1.03
0.985
1.00
0.908
0.92
1298
TOTAL
N/A
Pg 2
PC
N/A
1.000
RIGHT
HV
LB
N/A
N/A
2.000
PC
108
1.000
108
PC
135
1.000
135
PC
N/A
1.000
HV
12
2.000
24
HV
15
2.000
30
HV
N/A
2.000
LB
N/A
LB
0
N/A
0
LB
N/A
TOTAL
N/A
TOTAL
120
132
1.21
160
1.00
1.00
1.00
0.952
0.92
183
TOTAL
150
165
1.21
200
1.03
0.985
1.00
0.908
0.92
236
TOTAL
N/A
6/1/10
562
8:51 PM
Page 562
CHAPTER 23
Pg 3

LANE GROUP
EBLT
259
317
0.817
EB
EBRT
737
1140
0.647
WBLT
WB
WBRT NBLT
712
697
1.022
NB
NBRT
538
507
1.061
SBLT
SB
N/A
SBRT
WBLT
WB
NB
SBLT
SB
SBRT

gi
i=1n
cSUM = 1900
Critical v/c Ratio:

vci
Xc = i = 1 n
1520 tcu/h
0.993
cSUM

LANE GROUP
Input Variables:
Cycle Length, C
Effective Green, g
v/c Ratio, X
Number of Lanes, N
Capacity, c
Delay Computations:
Uniform Delay, di
Overflow Delay, d2
Total Delay, d
Approach Delay
EBLT
EB
EBRT
WBRT
NBLT
60
10
0.817
1
317
60
36
0.647
2
1140
60
22
1.022
2
697
60
16
1.061
3
507
24.1
1.00
20.3
44.9
D
7.8
1.00
1.40
9.3
A
42.0
D
19.0
1.00
29.8
48.8
D
22.0
1.00
45.9
67.9
E
NBRT
DELAY EQUATIONS:
di = d1i * PF + d2i
Q5C 1
gi
C
d1i =
g
1 min(1, Xi)* i
d2i = 225 (Xi 1)+ (Xi 1)2+16 Xi

ci Ni
C
Figure 23.4: (Continued) CMA Worksheet for Sample Problem 1Page 3
23.3
6/1/10
8:51 PM
Page 563
563
Conversions to Through-Car Equivalents Under

Ideal Conditions (tcu/h)
Determining Critical-Lane Flows and the Sum

of Critical-Lane Flows
The next several entries involve the PHF and adjustments

for nonideal conditions in the following categories: lane
width (Table 23.4), grade (Table 23.5), parking conditions
(Table 23.6), and lane utilization (Table 23.7). All lane groups
have no parking ( fp = 1.00). The EB approach has a lane
width of 11 feet ( fw = 0.97), the WB approach has 12-foot
lanes ( fw = 1.00), and the NB approach has 13-foot lanes
( fw = 1.03). The EBLT lane group has one lane ( fLU = 1.00),
the EB TH and WB lane groups have two lanes each ( fLU =
0.952), and the NB lane group has three lanes ( fLU = 0.908).
The through-car equivalent volumes for each movement are now divided by the PHF and all adjustment
factors, resulting in the final determination of demand flow
rates in tcu/h under equivalent ideal conditions, for each
movement.
The highest single-lane flow moving in each signal phase is

identified in bold typeface in Part (F) of the worksheet. These
are then entered into the ring diagram. The highest sum of
critical-lane flows passes through Phases A1, A2, and B
on Ring 1 of the diagram. The sum of critical-lane flows is
1,509 tcu/h as shown.
Assigning tcu/h to Specific Lanes

Converted volumes must now be assigned to specific lanes
and lane groups as follows:
EB: The EBLT lane group handles all of the 259
left-turning tcu/h. The two through lanes of the EB
approach handle 1474 tcu/h divided equally737
tcu/h each.
WB: The WB approach carries 183 tcu/h (right turn)
and 1,241 tcu/h (through), for a total of 1424 tcu/h.
Assuming this is uniformly split between the two
lanes of the lane group, 712 tcu/h use each lane. In the
left-most lane, this is 712 tcu/h of through vehicles
only. The right-most lane carries all 183 right-turning
tcu/h and the remaining 1,241 - 712 = 529 tcu/h
through vehicles.
NB: All NB flows operate from a single lane group.
The lane group has a total flow rate of 79 + 1,298 +
236 = 1,613 tcu/h. This does not evenly divide into
the three lanes available. One lane will carry
538 tcu/h; the other two will carry 537 tcu/h.
The center lane can carry only through vehicles and
is assigned 538 tcu/h. The right lane must accommodate all right-turning vehicles (236 tcu/h) and
537 - 236 = 302 through tcu/h. The left lane
handles all 79 tcu/h of left-turning vehicles and
535 - 79 = 458 tvu/h.
The lane assignments are shown in the diagram in Part (F) of
page 1 of the worksheet.
Capacity and v/c Ratios

Part (H) on page 3 of the worksheet summarizes computations of capacity and v/c ratio for each lane group. The
demand flow on each lane group is transferred from page 2
of the worksheet. There are four lane groups: EBLT, EB,
WB, and NB. The capacity of each is computed using
Equation 23-5. It will be assumed that actual green times are
equal to effective green times. The EBLT phase is 10
seconds, the WB phase is 22 seconds, and the NB phase is
16 seconds. It must be remembered that the EB through
movement is permitted during the EBLT phase, the WB
phase, and the 4 seconds of yellow plus all red between
them: 10 + 22 + 4 = 36 seconds. The cycle length is 60
seconds. Capacities are then computed as:
c = 1900 A gc B
cEBLT = 1900 A 1060 B = 317 tcu / h /ln
cEB = 1900 A 3660 B = 1,140 tcu / h /ln
cWB = 1900 A 2260 B = 697 tcu / h /ln

cNB = 1900 A 1660 B = 506 tcu / h
The critical v/c ratio for the intersection is computed

using Equation 23-6:
gi
10 + 22 + 16
cSUM = 1900 a a b = 1900 a
b
60
C
= 1,520 tcu /h
With both demand flows and capacities now expressed
in tcu/h for critical lanes, v/c ratios can be directly computed:
XEBLT = 259317 = 0.817
XEB = 7371140 = 0.647
XWB = 712697 = 1.022
6/1/10
564
8:51 PM
Page 564
CHAPTER 23
XNB = 538507 = 1.061

Xc =
d2EBLT = 225 B(0.816 -1) +
259 + 712 + 538

= 0.993
1520
These results are significant. Of the critical lanes, two

operate at v/c ratios in excess of 1.00, signifying that demand
exceeds the available capacity (WB, NB). The critical v/c
ratio, however, is under 1.00 (though barely so), indicating that
a reallocation of green time might provide a workable signal
timing (i.e., without lengthening the cycle length, providing a
more efficient phase plan, or adding lanes to a lane group or
lane groups). This situation is addressed subsequently.
(0.816 - 1)2 + a
16 * 0.817 R
b
317 * 1
= 20.3 s/veh
d2EB = 225 B(0.647 - 1) +
(0.647 - 1)2 + a
16 * 0.647 R
b
1140 * 2
= 1.4 s/veh
d2WB = 225 B(1.022 - 1) +
(1.022 - 1) + a
16 * 1.022 R
b
697 * 2
= 29.8 s/veh
Average delays per vehicle can now be computed using
Equations 23-9, 23-10, and 23-11. Because the stated arrival
pattern is random, the progression adjustment factor for all
lane groups is 1.00. Total delays per vehicle are computed as:
d = d1 PF + d2
d1EBLT =
d1EB =
0.5 C C 1- A gC B D 2
1- A 0.817 * 1060 B
= 24.1 s/veh
d1WB =
0.50 * 60 * A 1- 2260 B 2
1- A 1 * 2260 B
= 19.0 s/veh
d1NB =
0.50 * 60 * A 1- 1660 B 2
1- A 1 * 1660 B
= 22.0 s/veh
The second term, d2, is overflow delay, and is computed

= 45.9 s/veh
Total delays for each lane group are:
dEB = 7.8(1) + 1.4 = 9.3 s/veh
dNB = 22.0(1) + 45.9 = 67.9 s/veh
0.50 * 60 * A 1-3660 B 2
= 7.8 s/veh
1- A 0.647 * 3660 B
d2 = 225 B(X - 1) +
16 * 1.061 R
b
507 * 3
dWB = 19.0(1) + 29.8 = 48.8 s/veh
1-min [(1, X)] * A gC B
0.50 * 60 * A 1- 1060 B 2
(1.061 - 1)2 + a
dEBLT = 24.1(1) + 20.3 = 44.4 s/veh
The first term, d1, is uniform delay, and is computed

d1 =
d2NB = 225 B(1.061 - 1) +
(X - 1)2 + a
16 X R
b
cN
From the criteria of Table 23.13, the levels of service

are EBLT (LOS D), EB (LOS A), WB (LOS D), NB (LOS E).
Analysis of Results
Any analysis must consider both the LOS designation and
the v/c ratios. Interestingly, none of the lane groups have a
LOS of F, even though two of them have a v/c ratio in
excess of 1.00. This is because of the 15-minute analysis
period and the assumption of no preexisting unserved queue
at the beginning of the analysis period. If a v/c ratio of
greater than 1.00 persists for more than 15 minutes, the
delay will quickly increase to values that would be classified as LOS F.
The critical v/c ratio is less than 1.00, suggesting that
green time could be reallocated within the existing cycle
length to achieve better results. The critical v/c ratio, however,
is 0.993, meaning that such a signal retiming would still
consume 99.3% of the available capacity. This is obviously
higher than one would like.
23.3
6/1/10
8:51 PM
Page 565
Table 23.10: Comparison of CMA and HCM Analysis:

Results for Sample Problem 1
v/c Ratios
Delays (s/veh)
Lane
Group
CMA
HCM
CMA
HCM
EBLT
EB
WB
NB
0.817
0.647
1.022
1.061
0.819
0.653
1.030
1.059
44.9
9.3
48.8
67.9
47.9
9.7
53.8
64.8
Comparing to the HCM Model

As we noted previously, this sample problem is re-solved in
Chapter 24 using the more complex HCM model. The resulting v/c ratios and delays are compared in Table 23.10.
In this case, both the v/c ratios and delays are
quite comparable. In general, the CMA results show somewhat less delay than the HCM methodology. The differences are small, however, and none of the levels of service
would have been different from those assigned as a result
of CMA.
Improving the Signalization
Because the critical v/c ratio is less than 1.00, it is possible to
reduce all of the v/c ratios to less than 1.00 by simply reallocating the green time. Using the approach of Chapter 21,
effective green time would be reallocated in the ratio of critical-lane flow rates for each phase. The critical-lane flow rates
estimated in this problem are as follows: Phase A1: 259 tcu/h,
Phase A2: 712 tcu/h, and Phase B: 538 tcu/h. The signal has
a 60-second cycle length, and it will be assumed that the
seconds of yellow and all red intervals remains
unchanged. It is also assumed (as was done in the analysis)
that effective green equals actual green.
Then:
gTOT = 60 - 12 = 48 s
gA1 = 48 *
gA2 = 48 *
gB = 48 *
A 2591509 B = 8.2 s
A 7121509B = 22.7 s
A 5381509B = 17.1 s
As might be expected, the retiming reduces the green time for

Phase A1 slightly while increasing it in Phases A2 and B.
Although this will produce an acceptable signal timing, it is
barely so because the critical v/c ratio was 0.993, with virtually
no unused green time.
565
It might be wise to consider an increase in the cycle

length, as well as a reallocation of green time. A cycle
length of 90 seconds would lead to the following signal
timing:
gTOT = 90 - 12 = 78 s
A 2591509B = 13.4 s
gA2 = 78 * A 7121509B = 36.8 s
gB = 78 * A 5381509 B = 27.8 s
gA1 = 78 *
The full worksheets are not shown here to conserve

space, but Table 23.11 shows the resulting v/c ratios and
delays for each lane group under these two signal timing
options.
The two reallocations result in v/c ratios that are all
below 1.00; that is, no lane group fails, leading to unserved
queues at the end of the 15-minute analysis period. With a
60-second cycle length, the delay for the EBLT lane group
reaches more than 80 s/veh, or LOS F. High delays are
always expected for short exclusive LT phases because
stopped vehicles must wait a significant time to reacquire the
green. Delays in the WB and NB lane groups is significantly
reduced, even at a cycle length of 60 seconds because they no
longer fail.
The results improve for a 90-second cycle length. Delay
to the EBLT lane group is reduced, and the LOS is improved
to E. Delay in the WB and NB lane groups is further reduced,
primarily because fewer isolated cycle failures are expected
with the longer cycle length. This would probably be the recommended timing, assuming there are no system constraints
on selecting a cycle length.
Table 23.11: Delay, v/c Ratio, and Delay Results from

Reallocation of Green
Lane Group
Item
EBLT
Cycle Length = 60 s
v/c
0.977
d (s/veh)
88.50
LOS
F
Cycle Length = 90 s
v/c
0.916
d (s/veh)
78.1
LOS
E
EB
WB
NB
0.667
10.40
B
0.990
43.00
D
0.994
50.70
D
0.644
13.2
B
0.916
36.5
D
0.917
44.3
D
6/1/10
566
8:51 PM
Page 566
CHAPTER 23
23.3.2
Sample Problem 2: Example with

Compound Phasing
The EB left turn faces an opposing flow of 700 veh/h

in two lanes. From Table 23.6, this equivalent is 6.5.
Then:
Intersection Description
Figure 23.5 is the worksheet for Sample Problem 2. A full
description of the intersection and signal timing are found in
sections (A), (B), and (C) of page 1. The intersection consists
of two four-lane arterials, with left-turn lanes for the EB and
WB approaches. The signalization includes a protected +
permitted phase for the EB and WB left turns. The intersection may be considered to be isolated; that is, vehicles arrive
randomly on all approaches.
Conversion of Volumes to pc/h
Page 2 of the worksheet shows all volume conversion computations. Each movement volume contains 10% heavy vehicles
and no local buses. The volumes are segregated for conversion, with results rounded to the nearest whole vehicle. There
are no equivalents to look up because all passenger cars
have an equivalent of 1.00 and all heavy vehicles have an
equivalent of 2.0.
Conversion of Volumes to tcu/h
In this step, left-turn and right-turn volumes are converted to
through-car units. Through cars have an equivalent of 1.00.
Right-turn equivalents are selected from Table 23.3 for
50 peds/h in each crosswalk. This equivalent is 1.21 and
applies to all right turns on all approaches.
Left-turn equivalents are found in Table 23.3:
NB left turns are permitted, and face an opposing
flow of 650 veh/h in two lanes. Interpolating, the
equivalent for NB left turns is 5.75.
SB left turns are permitted, and face an opposing flow
of 500 veh/h in two lanes. Interpolating, the equivalent for SB left turns is 4.0.
The EB left turns have a protected + permitted compound phase. The equivalent is estimated as the
weighted average of the equivalents for protected LTs
and permitted LTs. The weighting is proportional to
the effective green time in the protected and permitted portions of the phase. In the signalization shown,
the protected portion has 8 seconds of green and
2 seconds of yellow (which counts as effective green),
or 10 seconds. The permitted portion of the phase is
40 seconds. Protected LTs have an equivalent of 1.05.
ELTEB =
(10 * 1.05) + (40 * 6.5)

(10 + 40)
= 5.41
The WB left turns share the same compound phase

as EB left turns. The equivalent for the protected
portion of the phase is again 1.05. The WB left turns
face an opposing flow of 800 veh/h in two lanes.
From Table 23.8, the equivalent for this LT is 8.0.
Then:
ELTWB =
(10 * 1.05) + (40 * 8.0)

(10 + 40)
= 6.41
These equivalents are used to convert volumes in pc/h to tcu/h

on the second page of the worksheet in Figure 23.5.
Converting Volumes in tcu/h to tcu/h
Under Equivalent Ideal Conditions
The final step in converting demand volumes involves adjustments for the PHF and for lane width, grade, parking conditions, and lane utilization. None of the lane groups have
parking, and all are on level grades. Thus the adjustment factors for these two conditions are both 1.00. Lane width adjustment factors are found in Table 23.4. All EB and WB lane
groups have 12-foot lanes, and the adjustment factor for these
is 1.00. The NB and SB lane groups, however, have 11-foot
lanes, and an adjustment factor of 0.97 must be applied. The
lane-utilization factor is found in Table 23.7. The EB and WB
left-turn lane groups have one lane, for which the adjustment
factor is 1.00. All other lane groups have two lanes, and the
adjustment factor for these is 0.952. The PHF for all lane
groups is given as 0.96.
Assigning Demand Flows to Lanes
and Determining Critical Lanes
In section (F) on page 1 of the worksheet (Figure 23.5), lane
volumes are assigned to individual lanes. The highest lane
volumes for each signal phase are identified in bold type.
Determining the critical-lane flows, however, requires that the
EB and WB left-turn demands be split between the protected
and permitted portions of the compound phase (in proportion
to the length of the portions of the phase). Because the compound LT phase has 10 seconds of protected green (which
6/1/10
8:51 PM
Page 567

LOCATION:
SAMPLE PROBLEM 1
TIME PERIOD OF ANALYSIS:

ANALYST:
4:30 PM
TO 5:30 PM
DATE: 10/1/08
R. P. Roess
SB
25
Pg 1
650
120
700
15
100
5% HV
0% LB
All Mvts
PHF = 0.96
WB
Phase A1
G = 8.0 s
y = 2.0 s
Phase A2
G = 40 s
y = 2.0 s
ar = 2.0 s
Phase B
G = 32 s
y = 2.0 s
ar = 2.0 s
EB
30
110
500
20
800
c = 90 s
NB
100
EBRT
NBRT
Pkg Act
None
None
None
None
EB
WB
NB
SB
PHF = 0.96
WBRT
SBRT
50
50
50
50
Ln Width Grade
12 ft
0%
12 ft
0%
11 ft
0%
0%
11 ft
SB
WB
EB
NB
5.41
6.61 NA
NA
(E) Compound Phase LT Equivalents
ELTC =

gLTPT + gLTPM
438
439
A1:
35
654
403
367
143
72
A2:
166
316
482
529
390
139
208 206
143
523
482
529
578
439
414
578
723
386 28
B:
439
414
131
414
Sum of Critical Lane Flows: 143+578+439 = 1,160 tcu/h
Figure 23.5: Sample Problem 2Page 1
567
6/1/10
568
8:51 PM
Page 568
CHAPTER 23

Item
EB
veh/h
equiv (Tab.23.1)
pc/h
tcu/h
fw (Tab. 23.4)
fg (Tab. 23.5)
fp (Tab. 23.6)
fLU (23.7)
PHF
ideal tcu/h
WB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
NB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
SB
veh/h
equiv (E)
pc/h
ELT/ERT
tcu/h
fw
fg
fp
fLU
PHF
ideal tcu/h
PC
104
1.000
104
PC
95
1.000
95
PC
28
1.000
28
PC
14
1.000
14
LEFT
HV
LB
6
0
2.000
12
0
HV
LB
5
0
2.000
10
0
HV
LB
2
0
2.000
4
0
HV
LB
1
0
2.000
0
2
TOTAL
110
116
5.41
628
1
1
1
1
0.96
654
TOTAL
100
105
6.61
694
1
1
1
1
0.96
723
TOTAL
30
32
5.75
184
0.97
1
1
0.952
0.96
208
TOTAL
15
16
4
64
0.97
1
1
0.952
0.96
72
Figure 23.5: (Continued) Sample Problem 2Page2
PC
760
1.000
760
PC
665
1.000
665
PC
475
1.000
475
PC
617
1.000
617
THROUGH
HV
LB
40
0
2.000
80
0
HV
LB
0
35
2.000
0
70
HV
LB
25
0
2.000
50
0
HV
LB
33
0
2.000
66
0
TOTAL
800
840
1
840
1
1
1
0.952
0.96
919
TOTAL
700
735
1
1
1
1
1
0.952
0.96
804
TOTAL
500
525
1
525
0.97
1
1
0.952
0.96
592
TOTAL
650
683
1
683
0.97
1
1
0.952
0.96
770
Pg 2
PC
95
1.000
95
PC
114
1.000
114
PC
19
1.000
19
PC
24
1.000
24
RIGHT
HV
LB
5
0
2.000
10
0
HV
LB
6
0
2.000
12
0
HV
LB
1
0
2.000
2
0
HV
LB
1
0
2.000
2
0
TOTAL
100
105
1.21
127
1
1
1
0.952
0.96
139
TOTAL
120
126
1.21
152
1
1
1
0.952
0.96
166
TOTAL
20
21
1.21
25
0.97
1
1
0.952
0.96
28
TOTAL
25
26
1.21
34
0.97
1
1
0.952
0.96
35
23.3
6/1/10
8:51 PM
Page 569
569
Pg 3

LANE GROUP
EBLT
694
1055
0.62
EB
EBRT
529
844
0.627
WBLT
723
1055
0.685
WB
WBRT
482
844
0.571
NBLT
NB
NBRT
414
676
0.612
SBLT
SB
SBRT
439
676
0.649
WBLT
WB
NBLT
NB
SBLT
SB

gi
i=1n
cSUM = 1900
Critical v/c Ratio:

vci
Xc = i = 1 n
1731 tcu/h
0.67
cSUM

LANE GROUP
Input Variables:
Cycle Length, C
Effective Green, g
v/c Ratio, X
Number of Lanes, N
Capacity, c
Delay Computations:
Uniform Delay, di
Overflow Delay, d2
Total Delay, d
Approach Delay
EBLT
EB
EBRT
WBRT
NBRT
90
50
0.62
1
1055
90
40
0.627
2
844
90
50
0.685
1
1055
90
40
0.571
2
844
90
32
0.612
2
676
90
32
0.649
2
676
13.6
1
3.1
16.7
B
19.3
1
2
21.3
C
18.7
B
14.4
1
4.1
18.5
B
18.6
1
1.6
20.2
C
19.1
B
23.9
1
2.4
26.2
C
24.3
1
2.8
27.1
C
DELAY EQUATIONS:
di = d1i * PF + d2i
Q5C 1
gi
d2i = 225 (Xi 1)+ (Xi 1)2+ 16 Xi

ci N i
d1i =
g
1 min(1, Xi)* i
Figure 23.5: (Continued) Sample Problem 2Page3
SBRT
6/1/10
570
8:51 PM
Page 570
CHAPTER 23
includes the yellow), and 40 seconds of permitted green, the

EB and WB left-turn flows are allocated as follows:
vEBLTprot = 654 *
A1050 B = 131 tcu/h
vEBLTperm = 654 - 131 = 523 tcu/h

vWBLTprot = 723 *
A1050 B = 145 tcu/h
vSBLTperm = 723 - 145 = 578

In part (G) of the worksheet, these and all other demand
flows are assigned to the appropriate portion of the signal ring
diagram. Note that in Phase A2, there are four lane flows
moving: the portion of the EB and WB left turns in the
permitted phase, and the lane flows on the through and rightturn lane groups. The selection of critical-lane flows is shown
on the worksheet. Of particular note is that the critical-lane
flow rate for Phase A2 is in the WB left-turn lane, which is
generally not a desirable condition.
Section (I) of the worksheet (page 3) shows the results
of delay computations. These computations use Equations
23-9, 23-10, and 23-11. The intersection is isolated, so the
progression factor, PF, is 1.00. The resulting levels of service are B for the left-turn lane groups and C for all other lane
groups. It is noted that the EB and WB through-lane groups
have delays that are just above the maximum boundary for
LOS B. When the EB and WB lane groups are averaged, the
aggregate EB and WB delays yield LOS B for the
approaches.
In general, this intersection is operating well within
acceptable limits. Most of the lane-group v/c ratios are relatively
low. A shorter cycle length might be possible and might yield
lower delays. The EB and WB protected LT phase might be proportionally increased by several seconds to eliminate the critical
nature of the WB LT flow in the permitted portion of the phase.
TRB Circular 212, published in 1980. The planning approach is

valid only for the most general of estimates and should not be
used where a more detailed analysis of conditions is needed.
The planning approach might be used, for example, to generally
evaluate the basic viability of a future intersection configuration
and signal timing. No decision on any specific design or signal
timing would ever be made using the planning approach.
The operational analysis and design approach to critical
movement analysis presented in this chapter is an updated
model, taking into account some of the latest developments in
signal timing methodologies and in the HCM methodology for
analysis of signalized intersections. Although it is not a back
of the envelope process, it can be done by hand in a reasonable
amount of time. It is sensitive to many of the important factors
affecting the operation of signalized intersections but is not as
detailed as the HCM methodology, described in Chapter 24.
It does yield delay estimates and can provide detailed insight
into how key features of the intersection and signal timing
affect overall performance.
References
1.
2.
3.
4.
5.
6.
23.4 Closing Comments
7.
Critical-lane analysis has been used to evaluate the performance

of signalized intersections for many years, and it has been the
basis for signal timing for even more.
In this chapter, a very basic back of the envelope planning approach has been presented and was taken directly from
8.
Highway Capacity Manual, Transportation Research

Board, Washington DC, 2000.
Highway Capacity Manual, Special Report 87, Transportation Research Board, Washington DC, 1965.
Highway Capacity Manual, Special Report 209, Transportation Research Board, Washington DC, 1985.
Interim Materials on Highway Capacity, Circular 212,
Transportation Research Board, Washington DC, 1980.
Capelle, D.G., and Pinell, C., Capacity Study of
Signalized Diamond Interchanges, Highway Research
Bulletin 291, Transportation Research Board, Washington
DC, 1961.
Messer, C.J., and Fambro, D.B., A New Critical Lane
Analysis for Intersection Design, 56th Annual Meeting of the Transportation Research Board, Washington
DC, January 1977.
Akcelik, R., SIDRA for the Highway Capacity
Manual, Compendium of Papers, 60th Annual Meeting
of the Institute of Transportation Engineers, Washington
DC, 1990.
Reilly, W., et al., Signalized Intersection Capacity
Study, Final Report, NCHRP Project 3-28(2), JHK &
Associates, Tucson AZ, 1982.
6/1/10
8:51 PM
Page 571
PROBLEMS
571
changes in the physical design or signal timing
that will result in better operations. It is not necessary to reanalyze the modified intersection to
demonstrate the effectiveness of the recommended
improvements.
(e) Comment on the differences between the results of
the Planning Approach and the Operations and
Design Approach.
Problems
23-1. The intersection shown here is to be analyzed using CMA.
The following information in available for this intersection: (a) PHF = 0.90. (b) Progression quality = Good
(EB), Poor (WB), and Random (NB). (c) 5% HV in all
movements, no local buses. (d) no pedestrians present
(overpasses are provided). For the intersection as shown:
(a) Analyze the intersection using the CMA Planning
Approach.
(b) Using the Operations and Design Approach, determine the existing v/c ratio, delay, and level of service for each lane group in the intersection.
(c) Aggregate the lane group delays for each approach
and determine the existing level of service for each.
(d) Evaluate how well the intersection is operating,
based on these results. If necessary, suggest
23-2. Analyze the intersection shown using the CMA Operations and Design Approach to determine the v/c ratio,
delay, and LOS for each lane group and each approach.
Assess the operation and recommend changes to the
physical design or signalization that you think are
needed to obtain acceptable operations. Use CMA to
assess the effectiveness of your recommended
improvements.
500
210
22 ft
26 ft
510
30 pkg
moves/h
400
45
26 ft
Phase A1
G=25 s
y=3 s; ar=1 s
Intersection for Problem 23.1
Phase A2
G=20 s
y=3 s;ar=1 s
Phase B
G=22.5 s
y=3 s;ar=1.5 s
380
6/1/10
572
8:51 PM
Page 572
CHAPTER 23
Other Characteristics
PHF = 0.95
Moderate Pedestrian Activity
Level Grades
8% HV in all mvts
LB included in through volumes
Good Prog SB, NB
Random Prog EB,WB
36 ft
100
1350
60
200 30
LB/h
500
11
45
40
11
30
LB/h
450
50
A1
G = 12
y = 2 ar = 2
A2
G = 10
y = 2 ar = 2
A3
G = 31
y = 2 ar = 2
G = 25
y = 2 ar = 2
Phase Diagram Actuated (Average Timing)

C = 90 s
Intersection for Problem 23.2
125
900
90

CMA For Signalized Intersections by Elena Prassas

Uploaded by

Copyright:

Available Formats

CMA For Signalized Intersections by Elena Prassas

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CMA For Signalized Intersections by Elena Prassas

Uploaded by

Copyright:

Available Formats

M23_ROES5739_04_SE_C23.

the day. This often presented the traffic engineer with an

CRITICAL MOVEMENT ANALYSIS OF SIGNALIZED INTERSECTIONS

Movement Analysis [4]. The methodology could be easily

23.1 The TRB Circular 212

factors to reflect through passenger car units (tpcus), as is at

23.2 A Critical Movement Approach

Steps 1 through 3 of the operations and design methodology

A CRITICAL MOVEMENT APPROACH TO SIGNALIZED INTERSECTION ANALYSIS

in as much detail as possible. At a minimum, the following

Proportion of heavy vehicles

A heavy vehicle is any vehicle with more than four

For the vast majority of cases, the assumption that (y + ar) =

Step 4: Convert Demand Volumes

Each movement demand volume is converted to a volume

where: Vpc = volume in passenger-car equivalents, pc/h

Table 23.1: Passenger-car Equivalents for Local Buses

Volume in Lane Group, veh/h

Buses Stopping in Travel Lane

where: g = effective green time, s

Buses Stopping Off-Line or in Parking Lane

CRITICAL MOVEMENT ANALYSIS OF SIGNALIZED INTERSECTIONS

the number of lanes in the affected lane group. Exclusive LT

Step 5: Convert Passenger-Car

Number of Opposing Lanes, No

ELT for all protected left turns = 1.05.

Table 23.3: Through-Car Equivalents for Right-Turning

It is appropriate to interpolate for intermediate values in

A ELTPT * gLTPT) + (ELTPM * gLTPM B

where: ELTC = left-turn equivalent for compound LT phasing

A CRITICAL MOVEMENT APPROACH TO SIGNALIZED INTERSECTION ANALYSIS

This average equivalent is then used to convert the full

Step 6: Converting Through-Car

At this point in the methodology, all movement volumes have

where: v = movement flow rate under equivalent ideal conditions, tcu/h

Table 23.4: Adjustment Factors for Lane Width

0.90 0.93 0.97 1.00 1.03 1.07 1.10

Table 23.5: Adjustment Factors for Grade

Table 23.6: Adjustment Factors for Parking Conditions

Table 23.7: Adjustment Factors for Lane Utilization

(Source: Highway Capacity Manual, 3rd Edition, Transportation

The peak-hour factor (PHF) converts a full-hour

CRITICAL MOVEMENT ANALYSIS OF SIGNALIZED INTERSECTIONS

Lane widths less than 9 feet are not recommended; in

Assigning Lane Flow Rates

Finding Critical-Lane Flow Rates

Simple Case: No Overlapping or Compound Phases

Each movement volume is now expressed in through-car units

Figure 23.1: Sample Ring Diagram for Overlapping

A CRITICAL MOVEMENT APPROACH TO SIGNALIZED INTERSECTION ANALYSIS

diagram shows the critical-lane flow rate (under equivalent

where: ci = capacity of critical lane in lane group i, tcu/h/ln

maximum sum of critical-lane flow rates, tcu/h

Once capacities for all critical lanes are determined, a

Capacity and v/c Ratios

At this point in the analysis, the capacity of each critical lane

where: Xi = v/c ratio for critical lane i

where: Xi = critical v/c ratio for intersection