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 68
68
TABLE 25
BEFORE AND AFTER RESULTS OF SFMTA BUS STOP REDUCTION IN SEVEN CORRIDORS
Before After Change
Street Stops per Avg. Bus Stops per Avg. Bus Stops per Avg. Bus
Mile Speed Mile Speed Mile Speed
Haight 10.7 8.2 mph 7.1 9.4 mph -3.6 +14.6%
Union 11.0 9.1 mph 7.1 10.0 mph -3.9 +9.9%
Van Ness 10.6 6.2 mph 8.2 6.5 mph -2.4 +4.8%
Polk (NB) 12.0 9.1 mph 7.8 9.5 mph -4.2 +4.4%
Mission (NB) 10.4 6.1 mph 5.2 6.8 mph -5.2 +11.5%
Sacramento/ 13.2 5.4 mph 7.3 5.8 mph -5.9 +7.4%
Columbus
(NB)
Source: SFMTA Transit Preferential Streets Program--19851988 Final Report (34). NB = Northbound.
corridor (33). Route 4Fessenden/104Division, which pro- gram. Bus stops were reduced from 2.5 to 5.9 stops per mile,
vides radial service interlined through downtown Portland, with average bus speeds increasing from 4.4% to 14.6% (32).
was the subject of the analysis. Both control and "with treat-
ment" segments were evaluated, each comprising a length of TCRP Reports 26, 90, and 118 (16,4,5) all addressed the
two miles. The "with treatment" segment had a net reduction of impact of stop spacing on arterial bus travel time. Table 26
four inbound and six outbound stops, whereas the 104/Division relates the average arterial bus travel time rate (minutes per
route had a net reduction of five inbound and seven outbound mile) to the number of bus stops and the average dwell per
stops. The net reduction in stops resulted in an increase in aver- stop. Using such a table, with knowledge of how dwell time
age spacing of 6% for inbound and 8% for outbound stops. might change with a stop consolidation strategy, the travel
A 5.7% reduction in bus running time attributable to stop time savings associated with stop consolidation along a bus
consolidation was identified. route can be estimated.
In the late 1980s, MUNI in San Francisco undertook a sys- ANALYSIS METHODS
tematic evaluation of the impact of bus stop reduction and relo-
cation in seven bus corridors: Haight Street, Union Street, Van There are various methodologies for assessing the impacts of
Ness Avenue, Polk Street, Mission Street, Sacramento Street, different transit preferential treatments on both transit opera-
and Columbus Avenue. Table 25 shows the results of this pro- tions (change in delay, operating speed, on-time performance)
TABLE 26
BASE ARTERIAL BUS TRAVEL TIMES WITH DIFFERENT STOP SPACING
AND DWELL TIMES
Average Dwell Time Stop Made Per Mile
Per Stop (sec) 2 4 5 6 7 8 9 10 12
10 2.40 3.27 3.77 4.3 4.88 5.53 6.23 7.00 8.75
20 2.73 3.93 4.60 5.3 6.04 6.87 7.73 8.67 10.75
30 3.07 4.60 5.43 6.3 7.20 6.20 9.21 10.33 12.75
40 3.40 5.27 6.26 7.3 8.35 9.53 10.71 12.00 14.75
50 3.74 5.92 7.08 8.3 9.52 10.88 12.21 13.67 16.75
60 4.07 6.58 7.90 9.3 10.67 12.21 13.70 15.33 18.75
Source: TCRP Reports, 26, 90, and 118 (16,4,5).
OCR for page 69
69
and general traffic operations. This section describes the use of TABLE 27
field surveys, application of refined data and guidance from ESTIMATED TRAVEL TIME RATE
REDUCTION WITH ARTERIAL BUS
different documents, and micro-simulation to identify these LANES--ANALOGY
impacts. Which analysis method to apply will usually be dic-
Minutes per Mile
tated by the desired information and complexity of the eval- Location Reduction
uation, as well as funds available. In certain cases, basic ana-
Highly Congested CBD 3 to 5
lytical models with typical values for certain cases may be
Typical CBD 1 to 2
adequate (particularly for earlier planning-level evaluations),
Typical Arterial 0.5 to 1
where simulation modeling would be more appropriate when
the effects of a system of treatments are desired to be evalu- Source: TCRP Report 118 (5).
CBD = central business district.
ated, and where different scenarios (such as alternate signal
timing settings for TSP) need to be evaluated.
Estimated travel time rate reductions based on analogy
Exclusive Transit Lanes (analysis/synthesis of experience) are shown in Table 27.
These values can provide an initial order of magnitude estimate
Analysis of the travel time implications of new dedicated tran- of time savings. More refined estimates of travel time savings
sit lanes can address all persons traveling in the respective cor- and speed increases can be obtained from the values shown
ridor, including auto drivers and passengers, not just existing in Table 28 and Figures 46 and 47, as developed through the
and future transit passengers. Historic information on changes TCRP A-23A research.
in transit travel times from implementation of bus lanes can
be obtained from a variety of sources, including the FTA
document Characteristics of Bus Rapid Transit for Decision- The top half of Table 28 shows the estimated speed changes
Making (32) and TCRP Report 90 (4). resulting from installing a curb bus lane for various initial
speeds. Figure 46 graphs the speed before and after bus lane
Highway Capacity Manual 2000 (19) can be used to cal- installation. Given the initial bus speed, the chart may be used
culate the impact of removing a general traffic lane from an to estimate the benefits of a curb bus lane. The gain in speed
arterial and dedicating it to the exclusive use of transit. When ranges from 1.5 mph for speeds lower than 6 mph to 2 mph for
an analysis of the effect of removing a lane from general traf- greater speeds.
fic use is done, any route diversion for existing highway users
must be accounted for. For example, if the corridor is part of The bottom half of Table 28 and Figure 47 show the time
a continuous grid of major arterials, some general traffic may savings in minutes per mile resulting from installing a bus
divert to parallel streets after a lane is removed. lane. The percent of time saved declines from approximately
one-third at the lowest speeds to about 20% at speeds that are
The likely changes in travel times resulting from installing
typical for an arterial bus (or BRT route).
a bus lane can be estimated in three basic ways:
1. Analogy (an estimate based on a synthesis and analy- The actual time saved depends on the length of the bus
sis of actual operating experience; see subsequent lane. For example, based on Figure 47, a bus traveling at
discussion), about 5 mph (12 min per mile) before bus lane installation may
2. Application of the Highway Capacity Manual Signal- expect a savings of about three minutes per mile after bus lane
ized Intersection Delay Analysis, and installation. If the bus lane is 5 miles long, the total savings
3. Computer simulation. would be 15 s.
TABLE 28
ESTIMATED TRAVEL TIME RATE REDUCTION WITH ARTERIAL
BUS LANES--FOR SPECIFIC CASES BASED ON ANALOGY
Item Case A Case B Case C Case D Case E
Initial Speed (mph) 3.0 4.0 6.0 8.0 10.0
Speed with Curb Bus Lane 4.4 5.7 8.0 10.2 12.2
(mph)
mph Gain 1.4 1.7 2.0 2.2 2.2
% Gain 47.0 42.0 33.3 27.5 22.0
Initial Minutes/Mile 20 15 10 7.5 6.0
Minutes/Mile with Bus 13.5 10.5 7.5 5.9 4.0
Lane
Minutes/Mile Gain 6.5 4.5 2.5 1.6 1.1
% Gain 32.5 30.0 25.0 21.3 18.3
Source: TCRP Report 90 (4).
OCR for page 70
70
25
Bus Speed on Bus Lane (MPH)
20
15
10
5
0
0 5 10 15 20 25
Initial Bus Speed (MPH)
Initial Bus Speed with Bus Lane
FIGURE 46 Impact of curb bus lanes on bus speed [Source: TCRP Report 118 (5 )].
Transit Signal Priority time and schedule adherence through field data collection. An
on-board transit travel time and delay survey is the most appro-
Field surveys and both analytical and simulation modeling priate tool to be applied. Measuring changes in general traffic
can be used to estimate the reduction in bus delay and, hence, delay associated with TSP is much more cumbersome because
reductions in overall travel time associated with the applica- extensive staff is required to manually record vehicle delays in
tion of TSP. A description of the potential application of sur- the field, videotape general traffic conditions, and then deci-
veys and simulation follows. pher changes in delay through video observations.
Field Surveys Analytical Model
The most accurate yet perhaps most time-consuming and As mentioned previously, TSP advances or extends the green
expensive means to identify the impact of TSP is to conduct a time whenever a transit vehicle arrives within the designated
"before" and "after" evaluation of changes in transit travel windows at the beginning or end of the cycle. This has the
25
Minutes per Mile with Curb Bus Lane
20
15
10
5
0
0 5 10 15 20 25
Minutes per Mile
without Bus Lane with Bus Lane
FIGURE 47 Travel time savings with curb bus lane [Source: TCRP Report 118 (5)].
OCR for page 71
71
FIGURE 48 Signalized intersection delay (60-second cycle and 50% effective green) [Source: TCRP
Report 118 (5 )].
effect of reducing the red time that transit vehicles incur. comparing the delays for the initial g/C value with those for an
Delays to transit vehicles with and without TSP can be approx- appropriate curve with a higher value (e.g., comparing the
imated by using delay curves for signalized intersections that curves in Figures 48 through 51).
relate intersection approach green time available (g/C ) to the
v/c ratio of the approach. Such signalized intersection delay Figure 52 gives an example of how priority for transit can
curves are presented in Figures 48 through Figure 51 for dif- reduce delay. A 90 s cycle with a g/C of 0.4 is assumed as a
ferent signal cycle lengths. Therefore, assuming 10% of the base with a v/c ratio of 0.8. The base delay is 33 s. An increase
cycle time for a TSP window, the delay savings for any given in g/C to 50% would result from TSP. The longer green period
v/c for the particular intersection approach can be estimated by would result in a 26 s delay, which is a savings of 7 s or 21%
FIGURE 49 Signalized intersection delay (60-second cycle and range of effective green) [Source: TCRP
Report 118 (5 )].
OCR for page 72
72
FIGURE 50 Signalized intersection delay (90-second cycle and range of effective green) [Source: TCRP
Report 118 (5 )].
FIGURE 51 Signalized intersection delay (120-second cycle and range of effective green) [Source: TCRP
Report 118 (5 )].
OCR for page 73
73
FIGURE 52 Effect of TSP on signalized intersection delay (90-second cycle) [Source: TCRP Report 118 (5 )].
per signalized intersection. This savings compares with an operate, accounting for any delays associated with turning traf-
average of 5 to 6 s saved per bus found along Wilshire fic. With a queue jump signal, some increased general traf-
Whittier and Ventura Boulevards in Los Angeles and along fic delay would occur as a result of the reduction of green time
San Pablo Avenue in Oakland. typically from the parallel through traffic phase to create a sep-
arate bus signal phase.
Simulation Modeling Figure 53 presents a graph that identifies the travel time sav-
ings associated with a queue jump treatment assuming (1) the
Another method to identify TSP impacts is to develop a simu- queue bypass lane is long enough to function effectively and
lation model of before and after conditions at an intersection or (2) an advance green of about 10% of the cycle length is pro-
along a corridor and measure the change in bus travel time and vided. The example assumes an initial g/C (effective green
delay and general traffic delay. The model is normally cali- time per cycle) of 50% and v/c of 0.8. After the bypass is
brated to field conditions through some level of field data col- installed, the g/C is assumed as 0.6 and the v/c at 0.2. In this
lection of bus travel times and bus and general traffic delays. example, a bus travel time savings of 17 s would result. Com-
Given the time to develop a simulation model plus added field parative benefits for other values of g/C and v/c can be obtained
data collection for calibration, this analysis approach tends to either by interpolation or by application of the delay equations.
be more expensive. However, simulation modeling does allow
for the testing of the impact of different traffic volume, con- Simulation modeling can also be applied to identify impacts
troller setting, and degree of lateness conditions in the most to both bus travel time and general traffic delay associated
economical manner in evaluating the sensitivity and overall with queue jump or bypass lane application. As with TSP,
impact of TSP on intersection and corridor operations. before and after conditions can be modeled using existing
field data.
Queue Jumps and Bypass Lanes
Curb Extensions
The reduction in bus delay and, hence, travel time associated
with the provision of queue jumps or bypass lanes can be The travel time savings associated with individual curb exten-
estimated by using procedures in the 2000 Highway Capacity sion treatments can be estimated through the transit vehicle
Manual (19). Intersection approach delay for general traffic clearance time savings identified from the analytical model
can be identified for a condition where buses would be in the reflective of the values in Table 24. As for other transit pref-
general traffic stream with no queue bypass treatment being erential treatments, simulation modeling can also be applied
provided. The delay to buses with the queue bypass treatment when it is desired to assess the impacts of a series of curb
can then be estimated in the separate lane where buses would extensions on overall general traffic travel time in a corridor.
OCR for page 74
74
FIGURE 53 Effect of queue bypass with advanced green on signalized intersection delay (90-second
cycle) [Source: TCRP Report 118 (5 )].
Stop Consolidation demand, transit service characteristics, traffic flow character-
istics, and elements of geometric feasibility (e.g., roadway
The travel time savings associated with different bus stop cross section).
spacing along an urban street can be estimated for planning
applications using Table 26. Simulation modeling can also The evaluation process as defined would identify bus
be applied if desired. preferential treatments based on the following steps:
Cumulative Impact Assessment 1. For each location (i.e., corridor segment, intersec-
tion, or bus stop), evaluate the factors described in
In addition to analyzing the impact of individual transit pref- Figure 54.
erential treatments, many times there is a need to compare 2. If all of the thresholds are met for a potential improve-
and prioritize different preferential treatments within a corri- ment at a given location, assign the weights for that
dor or at an intersection. One potential analysis methodology potential improvement to the corridor for four differ-
involves scoring and weighing different preferential treat- ent factors--increasing ridership, increasing travel
ments for potential application. Such a methodology was speed (or decreasing delay), increasing passenger com-
developed for a study for HART in Tampa to identify transit fort, and increasing service reliability).
improvements in certain corridors. 3. Sum the weights for each location in the corridor for
use in corridor prioritization.
The evaluation framework that was developed is a planning-
level tool that is intended to both prioritize corridors and iden- The weights identified were based on a scale of 0 to 10,
tify specific "hot spots" where there is a compelling need for a where 0 means that it would have no positive impact and 10
particular type of transit improvement. Three categories of means it would have a significant positive impact.
improvements, service improvements, bus preferential treat-
ments, and facility improvements, were considered. Figure 54 To properly compare corridors given that each corridor
presents the bus preferential treatment worksheet that lists HART had evaluated has a different length, number of bus
potential bus preferential improvements that can be applied to stops, and number of intersections, total scores (i.e., tallied
a corridor, bus stop, or intersection. This worksheet was devel- weights) for the bus preferential treatment improvement cat-
oped to help determine if a certain location meets the identified egory were normalized (divided by the number of signalized
thresholds to warrant the improvement or improvements. The intersections in a corridor) so that a consistent unit compari-
framework's factors reflect existing and potential passenger son among corridors could be made. Table 29 identifies the
OCR for page 75
75
FIGURE 54 Scoring/weighing system for bus preferential treatments--HART study [Source: "Transit Corridor
Evaluation and Prioritization Framework," 2006 TRB Annual Meeting (14)].
OCR for page 76
76
TABLE 29
SCORING OF BUS PREFERENTIAL TREATMENTS BY HART CORRIDOR
AND NUMBER OF WARRANTED IMPROVEMENTS
Raw Scores Normalized Scores
Corridor Direction 1 Direction 2 Total No. of Signals* Score
Florida Ave. 555 547 1,102 27.5 40.1
Nebraska Ave. 760 751 1,511 39 38.7
Columbus Dr. 706 720 1,426 38.5 37.0
M.L. King, Jr. Blvd. 462 535 997 34 29.3
Hillsborough Ave. 741 775 1,516 39.5 38.4
* This
is the average of both directions. Directions may not be symmetric.
Source: "Transit Corridor Evaluation and Prioritization Framework," 2005 TRB Annual Meeting (14).
TABLE 30
EXAMPLE OF SCORING EVALUATION OF TRANSIT PRIORITY TREATMENTS--ROUTE 5
CORRIDOR IN SEATTLE
Transit Implemen-
GP Delay Parking
Location Cost Delay tation
Savings Impacts
Savings Times
Fremont Ave N & N 39th St TSP 2 5 4 3 5
Phinney Ave N & N 46th St
Option 1 TSP 3 5 4 3 5
Option 2 Parking Restriction 5 5 4 1 5
Phinney Ave N & N 50th St TSP 3 5 3 3 5
Phinney Ave N & N 60th St TSP 3 3 2 3 5
Phinney Ave N & N 65th St
Option 1 TSP 3 4 2 3 5
Option 2 Queue Jump 4 5 2 2 4
Greenwood Ave N & N 80th St TSP 2 5 5 3 5
Greenwood Ave N & N 85th St
Option 1 TSP 3 5 1 3 5
Option 2 Parking Restriction 5 5 4 5
Greenwood Ave N & N 87th St TSP 2 3 2 3 5
Greenwood Ave N & N 105th St TSP 3 5 1 3 5
Westminster Way N & Dayton Ave N TSP 3 4 4 3 5
Evaluation Criteria Definitions
Cost General Purpose Traffic Delay Savings
1. Over $100,000 1. Over 5 seconds per vehicle degradation
2. $50,000 - $100,000 2. 1 to 5 seconds per vehicle degradation
3. $25,000 - $49,999 3. No measurable change in delay time
4. $10,000 - $24,999 4. 1 to 5 seconds per vehicle trip improvement
5. Minimal Cost (Less than $10,000) 5. Over 5 seconds per vehicle trip improvement
Transit Delay Savings Parking Impacts
1. Over 10 seconds per bus trip degradation 1. Greater than 50% utilization of removed parking
2. 1 to 10 seconds per bus trip degradation 2. Up to 50% utilization of removed parking
3. No measurable change in delay time 3. No change
4. 1 to 10 seconds per bus trip improvement 4. Up to 50% utilization of added parking
5. Over 10 seconds per bus trip improvement 5. Greater than 50% utilization of added parking
Implementation Time
Worse Better 1. Greater than 24 months to implement
2. 19 to 24 months to implement
1 2 3 4 5 3. 13 to 18 months to implement
4. 7 to 12 months to implement
5. 0 to 6 months to implement
Source: King County Route 5 Corridor Evaluation Report, DKS Associates (30).
