Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.

Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 41

41
CHAPTER 6
Forecast Future Performance
6.1 Introduction For a trip on a given route at a given time of day, over
several different facility types and segments of facilities, it is
The purpose of this chapter is to provide guidance on the es- necessary to estimate the travel time for each road segment
timation of trip travel times, delays, and reliability when direct and sum them. The estimated travel time for each road
measurement of these performance measures is not feasible. segment should include any delay incurred when transition-
Two methods are provided: a sketch planning method that re- ing from one segment to the next (for example the delay in
quires minimal data and a planning method that requires data making a left turn from one street to another).
to compute the capacity and FFS of the facility. The road segment travel time can be estimated by dividing the
This chapter focuses on planning applications. Methods segment length by the mean speed of the traveler over the length
provided here are designed for rapid application to large of the segment and adding in any segment transition delay.
batches of facilities requiring minimal field data collection on
TT = Ti (Eq. 6.1)
the characteristics of the facilities. Methods are applicable to i
the estimation of travel-time, delay, and reliability for trips
made on all types of highway facilities. The methods provided Li
Ti = + di (Eq. 6.2)
here are not as precise as field measurements of travel time, Si
delay, and reliability. They also are not as accurate or as sensi-
tive to variations in traffic controls as methods employing mi- where
crosimulation or HCM (5) techniques. Since this chapter deals TT = Trip Travel Time (hours);
with forecasting future performance, the same procedures Ti = Road Segment Travel Time (hours);
apply whether or not the agency is data-rich (with continuous Li = Length of road segment i (miles);
surveillance on some of its facilities), or data-poor. Si = Mean speed of vehicle over length of road segment i
The chapter is split into three major sections: estimating/ (mph); and
forecasting travel time, estimating delay, and estimating relia- di = Transition Delay. The delay incurred moving from
bility. The travel time section is split into two parts: estimation the end of segment i to the start of the next segment.
of trip times, and estimation of road segment travel times. The
road segment travel times section provides two methods for 6.2.1 Identification of Road Segments
estimating segment travel times, a sketch planning method and The route taken by a given trip must be divided into road seg-
a planning method. The sketch planning method requires only ments. A road segment is the portion of the road or highway
average daily traffic, signal spacing, and number of lanes. The over which neither demand nor capacity varies by more than
planning method requires additional information on the road 10 percent of their average value for the segment. The selected
segment to estimate capacity and FFS. route for the given trip is divided into a series of road segments,
and each can be presumed to have relatively uniform demand
and capacity over its length.
6.2 Estimating/Forecasting
Travel Time 6.2.2 Sketch Planning Techniques
for Estimating Segment Speeds
To estimate trip travel time you need to know the time of
day when the trip will be made, the starting point, the end The following equations for estimating road segment
point, and the route taken for the trip. speeds are designed for sketch planning purposes. They do

OCR for page 41

42
not require that the capacity be known, nor the PSL. They re- to the posted speed limit on the highway. Similarly, if the signal
quire information only on the average daily traffic (ADT), density (number of signals per mile) is less than the minimums
lanes, the number of signals per mile, and the number of in- listed above for each arterial class, the equations for the appro-
terchanges per mile. priate speed limit may still be applied, but the results should be
For freeways: used with caution.
A superior approach to using the equations listed above
Speed (mph) = 91.4 - 0.002 [ADT/Lane]
would be to measure speeds in the local region using the
-2.85 [Access Points Per Mile] (Eq. 6.3)
methods described in Chapter 3 and using linear regression
For Class I arterials: to fit equations that are more accurate for conditions in the
area. Assuming good data collection practices, locally devel-
Speed (mph) = 40.6 - 0.0002 [ADT/Lane]
oped equations almost always will be superior in accuracy to
-2.67 [Signals Per Mile] (Eq. 6.4)
the national average equations presented in this Guidebook.
For Class II and Class III arterials:
Speed (mph) = 36.4 - 0.000301 [ADT/Lane] 6.2.3 Planning Method
-1.56 [Signals Per Mile] (Eq. 6.5) for Estimating Speeds
where The planning method for estimating speed is designed to
Speed = Mean speed during weekday peak be applied to specific hours of the day and to specific street
hours, average of both directions (mph); segments. The method is sensitive to the posted speed limit,
ADT = Average daily traffic (total of both volume, and capacity. A method is provided to estimate
directions for weekdays) for road capacity based on signal timing and the physical characteris-
segment; tics of the facility. This method is taken from NCHRP Report
Lane = number of through lanes (total of both 387: Planning Techniques for Estimating Speed and Service Vol-
directions) on road segment; umes for Planning Applications (7).
Access Points = Average number of freeway interchanges If it is desired to estimate mean speed over a 24-hour pe-
Per Mile per mile for freeway segment. Treat par- riod, then one can either use the "one-hour" method below,
tial interchanges as full interchanges for applying it 24 times, for each hour of the day, or one can use
purpose of computing the average access the following equations from NCHRP Report 398.
points per mile; The recommended speed estimation technique is an update
Signals Per Mile = Average number of traffic signals per of the Bureau of Public Roads (BPR) speed-flow curve. The
mile on road segment; new curve has been fitted to updated speed-flow data con-
Arterial Class I = Signalized arterial street with at least tained in the HCM and has been validated against speed flow
1 signal every 2 miles, and with posted data for both uninterrupted flow facilities and interrupted
speed limit in excess of 40 mph; flow facilities.
Arterial Class II = Signalized arterial street with at least The facility space mean speed is computed in three steps:
1 signal every one-quarter mile, and
with posted speed limits of 30 to 40 1. Estimate the FFS;
mph, inclusive; and 2. Estimate capacity; and
Arterial Class III = Signalized arterial street with at least 3. Compute the average speed.
1 signal every one-quarter mile, and with
posted speed limit less than 30 mph. Look-up tables of defaults can be used to skip the first two
The above equations, taken from NCHRP Report 398, steps, but poor choices of the FFS and capacity can seriously
Quantifying Congestion (1), were developed by applying linear compromise the accuracy of the technique.
regression to various data sets available to the researchers. As
such, these equations are unlikely to be accurate for extreme Step 1. Estimate Free-Flow Speed
situations not covered in the original NCHRP 398 data sets,
such as roads in mountainous terrain. FFS of a facility is defined as the space mean speed of traffic
The above NCHRP 398 equations also are not designed to be when volumes are so light that they have negligible effect on
applied to multilane-rural highways or two-lane rural roads. speed. The best technique for estimating FFS is to measure it in
However, in the absence of better information, the analyst may the field under light traffic conditions, but this is not a feasible
cautiously use the equivalent arterial speed equations according option when several thousand street links must be analyzed.

OCR for page 41

43
The paragraphs below provide a recommended set of equations where
for estimating FFS in the absence of field measurements of FFS.
D = The total signal delay per vehicle (seconds);
g = The effective green time (seconds);
Option 1a. Equations for Facilities Without Signals C = The cycle length (seconds);
Two separate linear equations are provided for estimating If signal timing data is not available, the planner can use
free-flow speed for facilities with less than one signal every the following default values:
2 miles (3.2 km). One equation is for facilities with posted C = 120 seconds; and
speed limits in excess of 50 mph (80 kph). The other equation g/C = 0.45.
is for facilities with lower posted speed limits. DF = (1 - P)/(1 - g/C)
High-Speed Facilities [PSL in excess of 50 mph (80 kph)].
where P is the proportion of vehicles arriving on green.
Sf (mph) = 0.88 * Sp + 14 (Eq. 6.6)
If P is unknown, the following defaults can be used for DF:
Sf (kph) = 0.88 * Sp + 22 (Eq. 6.7)
DF = 0.9 for uncoordinated traffic actuated signals;
Low-Speed Facilities [PSL is 50 mph (80 kph) or less]. DF = 1.0 for uncoordinated fixed time signals;
DF = 1.2 for coordinated signals with unfavorable pro-
Sf (mph) = 0.79 * Sp + 12 (Eq. 6.8) gression;
Sf (kph) = 0.79 * Sp + 19 (Eq. 6.9) DF = 0.90 for coordinated signals with favorable progres-
sion; and
where DF = 0.60 for coordinated signals with highly favorable
progression.
Sf = FFS in either mph or kph; and
Sp = PSL in either of mph or kph.
Option 1c. Default FFS
Planners may wish to develop a look-up table of FFSs
Option 1b. Equations for Signalized Facilities
based upon the facility type and the area type where it is lo-
FFS for signalized facilities must take into account both the cated in order to simplify the estimation of FFSs. Depend-
FFS measured mid-block between signals and the signal ing upon local conditions, the planning agency may wish to
delays along the street (which occur even at low volumes). add terrain type (e.g., level, rolling, mountainous) and
The mean FFS (including signal delay) is computed using the frontage development types (commercial, residential, un-
following equation that adds together the free-flow travel developed) to the general development types used in
time between signals and the delay time at signals (under free- Exhibits 6.1 and 6.2.
flow conditions). The accuracy of the speed estimation procedure is highly
dependent on the accuracy of the FFS and capacity used in the
L
Sf = (Eq. 6.10) computations. Great care should be taken in the creation of
( 3600 )
L +N* D local look-up tables that accurately reflect the FFSs present in
Smb
the locality.
where
Step 2. Estimate Link Capacity
Sf = FFS speed for urban interrupted facility (mph or kph);
L = Length of facility (miles or km); The HCM (5) provides a set of procedures for estimating
Smb = Mid-block FFS (mph or kph); facility capacity for operations analysis purposes. These pro-
= 0.79 (PSL in mph) + 12 (mph); and cedures vary by facility type and generally require a great deal
= 0.79 (PSL in kph) + 19 (kph); of information on the facility. The following equations sim-
N = Number of signalized intersections on length L of plify the application of the HCM methods for use in planning
facility; and applications.
D = Average delay per signal per Equation 6.11 below
(seconds). Option 2a. Capacity Equation for Freeways
The average delay per signal is computed using the follow- The following equation is used to compute the capacity of
ing equation: a freeway at its critical point:
D = DF * 0.5 * C(1 - g/C)2 (Eq. 6.11) Capacity (vph) = Ideal Cap * N * Fhv * PHF (Eq. 6.12)

OCR for page 41

44
Area Type Freeway Expressway Arterial Collector Local
Central Business District 50 45 40 35 30
Urban 55 50 45 40 35
Suburban 60 55 50 45 40
Rural 65 60 55 50 45
Exhibit 6.1. Example default FFSs (in miles per hour).
where where
Ideal Cap = 2,400 passenger cars per hour per lane (pcphl) Ideal Cap = 2,200 (pcphl) for multilane rural roads with
for freeways with 70 mph (110 kph) or greater 60 mph FFS;
FFS; and = 2,100 (pcphl) for multilane rural roads with
= 2,300 (pcphl) for all other freeways [FFS 55 mph FFS;
< 70 mph (110 kph)]; = 2,000 (pcphl) for multilane rural roads with
N = Number of through lanes. Ignore auxiliary 50 mph FFS;
N = Number of through lanes; ignore exclusive
lanes and exit only lanes.
turn lanes;
Fhv = Heavy vehicle adjustment factor.
Fhv = Heavy vehicle adjustment factor;
= 100/(100 + 0.5 * HV) for level terrain;
= 100/(100 + 0.5 * HV) for level terrain;
= 100/(100 + 2.0 * HV) for rolling terrain; and = 100/(100 + 2.0 * HV) for rolling terrain;
= 100/(100 + 5.0 * HV) for mountainous terrain. = 100/(100 + 5.0 * HV) for mountainous terrain
HV = The proportion of heavy vehicles (including HV = The proportion of heavy vehicles (including
trucks, buses, and recreational vehicles) in the trucks, buses, and recreational vehicles) in the
traffic flow. If the HV is unknown, use 0.05 traffic flow. If the HV is unknown, use 0.05
heavy vehicles as default. heavy vehicles as default;
PHF = Peak-hour factor (the ratio of the peak 15- PHF = Peak-hour factor (the ratio of the peak 15-
minute flow rate to the average hourly flow minute flow rate to the average hourly flow
rate). If unknown, use default of 0.90. rate). If unknown, use default of 0.90;
Option 2c. Capacity Equation for
Option 2b. Capacity Equation for Two-Lane Unsignalized Roads
Unsignalized Multilane Roads
The following equation is used to compute the capacity (in
The following equation is used to compute the capacity of a one direction) for a two-lane (total of both directions) road
multilane road with signals (if any) spaced more than 2 miles with signals (if any) more than 2 miles apart:
apart:
Capacity (vph) = Ideal Cap * N * Fw * Fhv *
Capacity (vph) = Ideal Cap * N * Fhv * PHF (Eq. 6.13) PHF * Fdir * Fnopass (Eq. 6.14)
Area Type Freeway Expressway Arterial Collector Local
Central Business District 80 72 64 56 50
Urban 88 80 72 64 56
Suburban 96 88 80 72 64
Rural 104 96 88 80 72
Exhibit 6.2. Example default FFSs (in kilometers per hour).

OCR for page 41

45
where HV = The proportion of heavy vehicles (including
trucks, buses, and recreational vehicles) in the
Ideal Cap = 1,400 (pcphl) for all two-lane rural roads;
traffic flow. If the HV is unknown, use 2 percent
Fw = Lane width and lateral clearance factor;
heavy vehicles as default;
= 0.80 if narrow lanes and/or narrow shoulders
are present; PHF = Peak-hour factor (the ratio of the peak
= 1.00 otherwise; 15-minute flow rate to the average hourly flow
Narrow lanes are less than 12 feet (3.6 m) wide; rate). Use 0.90 as default if PHF not known;
Narrow shoulders are less than 3 feet wide Fpark = On-street parking adjustment factor;
(1.0 m); = 0.90 if on-street parking present and parking
Fhv = Heavy vehicle adjustment factor; time limit is one hour or less;
= 1.00/(1.00 + 1.0 * HV) for level terrain; = 1.00 otherwise;
= 100/(100 + 4.0 * HV) for rolling terrain; FBay = Left-turn bay adjustment factor;
= 100/(100 + 11.0 * HV) for mountainous terrain; = 1.10 if exclusive left-turn lane(s) (often as a
HV = The proportion of heavy vehicles (including left-turn bay) are present;
trucks, buses, and recreational vehicles) in the = 1.00 otherwise;
traffic flow. If the HV is unknown, use 0.02 FCBD = Central Business District (CBD) Adjustment
heavy vehicles as default; Factor;
PHF = Peak-hour factor (the ratio of the peak = 0.90 if located in CBD;
15-minute flow rate to the average hourly flow = 1.00 elsewhere;
rate). If not known, use default of 0.90; g/C = Ratio of effective green time per cycle;
Fdir = Directional Adjustment Factor; If no data available, use following defaults;
= 0.71 + 0.58 * (1.00 Peak Direction Propor- Protected left-turn phase present: g/C = 0.40;
tion); Protected left-turn phase NOT present: g/C =
Peak Direction Proportion is the proportion 0.45;
of two-way traffic going in peak direction. If Other defaults may be developed by the local
not known, use default of 0.55 peak direction; planning agency based upon local conditions.
Fnopass = No-Passing Zone Factor; Additional defaults might be developed based
= 1.00 for level terrain; upon the functional classes of the major and
= 0.97 0.07 * (NoPass) for rolling terrain; crossing streets; and
= 0.91 0.13 * (NoPass) for mountainous terrain; Fc = Optional user specified calibration factor nec-
NoPass is the proportion of length of facility essary to match estimated capacity with field
for which passing is prohibited; and measurements or other independent estimates
If NoPass is unknown, use 0.60 NoPass for of capacity (no units). Can be used to account
rolling terrain and 0.80 for mountainous ter- for the capacity reducing effects of left and right
rain. turns made from through lanes.
Option 2d. Capacity Equation
Option 2e. Construction of Localized Capacity
for Signalized Arterials
Look-Up Table
The following equation is used to compute the one di-
The accuracy of the speed estimates is highly dependent
rection capacity of any signalized road with signals spaced
on the quality of the estimated capacity for the facility. Con-
2 miles or less apart:
sequently it is recommended that each planning agency use
Capacity (vph) = Ideal Sat * N * Fhv * PHF * Fpark * FBay capacities specific to the critical point of the selected study
* FCBD * g/C * Fc (Eq. 6.15) section whenever possible. However it is recognized that this
is not always feasible for planning studies. Consequently the
where
following two tables show a procedure for selecting default
Ideal Sat = Ideal saturation flow rate (vehicles per lane per values and computing a look-up table of capacities by facil-
hour of green) = 1,900; ity type, area type, and terrain type. Other classification
N = Number of lanes (exclude exclusive turn lanes schemes may be appropriate, depending on the nature of
and short lane additions); local roadway conditions.
Fhv = Heavy vehicle adjustment factor; Exhibit 6.3 shows a set of selected default parameters for
= 1.00/(1.00 + HV); the calculation of capacity for freeways, divided arterials,

OCR for page 41

46
Left-
Functional Area Terrain Free Lane % Heavy Direction % No Turn
Clas s Type Type Lanes Speed Width PHF Vehicles Split Pass Parking Bay G/C
Freeway Rural Level All > 70 mph 0.85 5%
Rolling All > 70 mph 0.85 5%
Mountain All < 70 mph 0.85 5%
Urban All All < 70 mph 0.90 2%
Divided Rural Level >2 60 mph 0.85 5%
Arterial Rolling >2 55 mph 0.85 5%
Mountain >2 50 mph 0.85 5%
Suburb All All 0.90 2% No Yes 0.45
Urban All All 0.90 2% Yes Yes 0.45
CBD All All 0.90 2% Yes Yes 0.45
Undivided Rural Level 2 Standard 0.85 5% 55% 0%
Arterial Rolling 2 Standard 0.85 5% 55% 60%
Mountain 2 Narrow 0.85 5% 55% 80%
Suburb All All 0.90 2% No No 0.45
Urban All All 0.90 2% Yes No 0.45
CBD All All 0.90 2% Yes No 0.45
Collector Urban All All 0.85 2% Yes No 0.40
Exhibit 6.3. Example default values for computing capacity by functional class and area/terrain type.
undivided arterials, and collectors. Each facility type is further in this table would then be multiplied by the number of lanes
subclassified according to the area type (urban or rural), ter- (in one direction) at the critical point to obtain the critical
rain type (level, rolling, mountainous), and number of lanes point capacity for the facility.
(total of two lanes both directions, or more). A separate set of
default parameters is then selected for each subclassification
Step 3. Compute Average Speed
of each facility type.
For example, a rural freeway in level or mountainous ter- If it is desired to compute mean speed for each hour of a
rain is assumed to have a FFS in excess of 70 mph (112 kph), day, then once the link capacity and free-flow speed are
5 percent heavy vehicles, and a peak-hour factor of 0.85. An known, the updated BPR equation (Equation 6.16) can be
urban freeway is assumed to have a FFS below 70 mph (112 used to predict the space mean vehicle speed for the link at
kph), 2 percent heavy vehicles, and a peak-hour factor of 0.90 forecasted traffic volumes. The same equation is used for both
to reflect the lower design speeds, heavier passenger car vol- metric and customary units. This method requires that ca-
umes, and flatter peak volumes in urban areas. pacity be measured or estimated.
Divided arterials in rural areas are assumed to have FFS sf
that decrease as the difficulty of the terrain increases. The as- s=
1 + a(v / c )b (Eq. 6.16)
sumed FFS for level terrain is 60 mph (96 kph), for rolling
terrain 55 mph (88 kph), and for mountainous terrain where
50 mph (80 kph).
Any road in a rural area is assumed in this table to have s = predicted space mean speed;
signals (if any) spaced farther than 2 miles apart. Urban area sf= FFS;
roads are assumed in this table to have signals at least 2 miles v = volume;
apart. The local planning agency should modify these c = capacity;
assumptions if they are not appropriate for its particular a = 0.05 for facilities with signals spaced 2 miles or less
jurisdiction. Exhibit 6.3 shows assumptions only for two-lane apart, and
rural undivided arterials, but the planning agency can add ad- = 0.20 for all other facilities; and
ditional rows of data for multilane rural undivided arterials. b = 10.
Exhibit 6.4 shows the computation of the capacities by fa- The two keys to success in applying the updated BPR curve
cility type based upon the assumptions contained in Exhibit are to have an accurate estimate of the FFS and the capacity
6.3. The results have been rounded off to the nearest 50 or 100 for the facility. Once those two key parameters are accurately
vehicles per hour per lane. The capacities per lane contained known, the updated BPR curve can estimate speeds for both

OCR for page 41

47
Functional Area Terrain Ideal Cap/
Class Type Type Lanes Cap PHF Fhv Fw Fdir Fnopass Fpark Fleft Fcbd G/C Lane
Freeway Rural Level All 2400 0.85 0.98 2000
Rolling All 2400 0.85 0.91 1900
Mountain All 2300 0.85 0.80 1600
Urban All All 2300 0.90 0.98 2000
Divided Rural Level >2 2200 0.85 0.98 1800
Arterial
Rolling >2 2100 0.85 0.91 1600
Mountain >2 2000 0.85 0.80 1400
Suburb All All 1900 0.90 0.98 1.00 1.10 1.00 0.45 850
Urban All All 1900 0.90 0.98 0.90 1.10 1.00 0.45 750
CBD All All 1900 0.90 0.98 0.90 1.10 0.90 0.45 650
Undivided Rural Level 2 1400 0.85 0.95 1.00 0.97 1.00 1100
Arterial
Rolling 2 1400 0.85 0.83 1.00 0.97 0.93 900
Mountain 2 1400 0.85 0.65 0.80 0.97 0.81 500
Suburb All All 1900 0.90 0.98 1.00 1.00 1.00 0.45 750
Urban All All 1900 0.90 0.98 0.90 1.00 1.00 0.45 700
CBD All All 1900 0.90 0.98 0.90 1.00 0.90 0.45 600
Collector Urban All All 1900 0.85 0.98 0.90 1.00 1.00 0.40 550
Exhibit 6.4. Example computation of default capacities by functional class and area/terrain type.
arterials and freeways with accuracies approaching those of where
the HCM and simulation models.
di = Average left-turn delay at signal at end of segment i
(hours).
6.2.4 Estimating Transition Delays Ci = average cycle length of signal at end of segment i (sec-
onds). If cycle length is not known, assume 120 seconds
Delay incurred by a traveler moving from the end of one for suburban intersection, 90 seconds for downtown
segment to the beginning of the next segment usually can be intersection.
neglected in the case of freeways, highways, and rural roads.
The segment to segment transition delay also can be neg-
6.2.5 Verification and Calibration
lected for through travel along a signalized arterial street since
of Travel-Time Estimates
the methods described for estimating segment speeds include
a nominal delay per signal in their estimates. The above equations for estimating trip travel time are
If the analyst is evaluating a route that involves left turns based on national average conditions. It is good practice for
on signalized streets, the analyst may wish to improve the the analyst to verify the estimates produced by these equations
accuracy of the travel time estimate by adding a nominal for a select sample of trips in the local area. Chapter 3 suggests
delay per left turn to the estimated total travel time for the the appropriate methods for developing local measurements
trip. of travel time for verifying the estimates produced by the
Actual field measurements of delay are best. HCM is the methods in this chapter.
next best method for estimating left-turn delay, but this re- If the selected field measurements of trip travel times are
quires a great deal of information on signal timing and turn- within an acceptable range of the estimated trip travel times,
ing movements at the intersection. In the absence of this data, the method can be considered to be verified against local
an estimate of one-half the cycle length for the signal may be conditions. The methods described in Chapter 3 can be used
used for the left-turn delay. (This assumes that a left turner is to determine the acceptable range for the results.
equally likely to arrive at any point in the signal cycle, so the If the results are not acceptable, the analyst should check
average wait for the left-turn arrow will be half the cycle for errors in the data used to estimate the travel times. Once
length of the signal.) the possibility of input data error has been ruled out (or at
least reduced to an acceptably low probability), the analyst
Ci should calibrate the estimated trip travel times to better
di = (Eq. 6.17)
2 * 3600 match observed times in the field.