Safety aspects of centrally situated bus Literature findings public - - PDF document

SLIDE 1

Safety aspects of centrally situated bus public transport routes

Shalom Hakkert Victoria Gitelman

28th ICTCT Workshop in Ashdod, Israel

• n 29th – 30th October 2015

Transportation Research Institute Technion – Israel Institute of Technology

Introduction Literature findings Analysis of central Public Transport Routes (PTR) and

pedestrian crossing configurations (study 1)

Comparison of safety levels of various bus systems

configurations: central PTR versus other types of bus-lanes (study 2)

SPIs: the concept (III)

Layout of presentation

SPIs: the concept (III)

Literature findings Mostly based on Duduta N., Adriazola C., Hidalgo D., Lindau L.A., Jaffe R., 2015 Traffic Safety on Bus priority systems, EMBARQ, the Bloomberg Philantropies

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SLIDE 2

SPIs: the concept (III)

Literature findings: safety impact of bus priority

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings: safety impact of common infrastructure changes

associated with implementing bus priority systems SPIs: the concept (III)

Literature findings: Results of safety impact assessment on bus priority

systems SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SLIDE 3

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Literature findings

SPIs: the concept (III)

Study 1: Analysis of central Public Transport Routes (PTR) and pedestrian crossing configurations Characteristics of PTR studied:

Central PTR, including left-side bus lane with stops on

median

Main arterials with heavy traffic and pedestrian activity At least two traffic lanes per direction A PTR lane in each direction Signalised 3- and 4-legged intersections SPIs: the concept (III)

Study 1: Analysis of central PTR and pedestrian crossing configurations The problem:

Many crashes with pedestrians at intersections on PTR

Source of problem:

Unfamiliar situation for crossing pedestrians: Having to cross three traffic streams in opposing

directions

Long waiting times for crossing pedestrians The PTR is bi-directional with relatively little traffic

SLIDE 4

SPIs: the concept (III)

Study 1: Pedestrian crossing configurations at signalized intersections with central PTR

Type 1 – a direct three-routes crossing Type 2 – a gradated right-right crossing Type 3 – a gradated left-left crossing Type 4 – a gradated-crossing with mixed-shifting: right-left or left-right Type 5 – a direct two-routes crossing SPIs: the concept (III)

Study 1: Examples of pedestrian crossings at signalized intersections with central PTR, in Israel

Type 1: Rothshild boulevard, in Haifa Type 2: Haatzmaut road, in Haifa SPIs: the concept (III)

Study 1: Examples of pedestrian crossings at signalized intersections with central PTR, in Israel

Type 3: Jabotinsky road, in Petah Tiqwa Type 4: Balfour street, in Bat - Yam Type 5: Haatzmaut road, in Haifa SPIs: the concept (III) Characteristic Distribution acc. to categories Junction configuration 18 three-legged (53%), 16 four-legged (47%) Pedestrian crossing configuration 11 of type 1 (32%), 6 of type 3 (18%), 15 of type 4 (44%), 2 of type 5 (6%) Speed limits 6 with 70 km/h (18%), 28 with 50 km/h (82%) All injury accidents Pedestrian accidents Accidents involving buses Accidents involving a bus and a pedestrian Value Total Serious Fatal Total Ser Fatal Total Ser Fatal Total Ser Fatal Average 6.9 0.8 0.3 1.7 0.4 0.2 1.5 0.3 0.2 0.7 0.2 0.2 s.d. 5 1 0.8 2.3 0.8 0.7 2 0.7 0.7 1.4 0.5 0.7

Accident indices, per junction, in 2010 - 2012

SPIs: the concept (III)

Study 1: Analysis of central PTR and pedestrian crossing configurations Type of analyses: 1.Safety indices for various intersection configurations 2.Comparison of crash indices for intersections by: a.Type of crossing configuration b.Number of intersection legs

• 3. Comparison of indices with comparison sites

SPIs: the concept (III)

Study 1 findings: Characteristics affecting the number of accidents at the study sited

Crossing type 3 gradated left-left had a

consistent negative correlation with accidents

Crossing type 4 gradated right-left had a

consistent positive correlation with accidents Regression models:

For all types of crashes – Crossing type 4 is

associated with an increase in accidents Non- parametric comparison of average no. of accidents:

Differences between type 4 and type 3 crossings

is significant and, in some cases, between type 3 and type 1 (straight across – is better than type 4 but worse than type 3)

Type 3 Type 4

SLIDE 5

SPIs: the concept (III)

Study 1: Comparing accident indices# at the PTR sites, by junction configuration

All accidents Pedestrian accidents Accidents involving buses Sites groups Total Serious Fatal Total Serious Fatal Total Serious Fatal Average accident indices: 3-legged 5.7 0.6 0.2 1.5 0.4 0.1 1.3 0.2 0.1 4-legged 8.2 0.9 0.4 1.9 0.5 0.3 1.8 0.4 0.4 Differences between the site groups, estimated by means of T-statistics (p-values): 3-legged vs 4- legged 0.007* 0.281 0.162 0.400 0.627 0.216 0.264 0.251 0.054* * Significant difference

# in 2010-2012

SPIs: the concept (III)

Study 1: Comparing accident indices# at the PTR sites and comparison-group (CG) sites, by junction configuration

All accidents Pedestrian accidents Accidents involving buses Sites Total Serious Fatal Total Serious Fatal Total Serious Fatal PTR junctions 5.7* 0.6 0.2 1.5 0.4 0.1 1.3 0.2 0.1 CG junctions 2.3 0.3 0.9 0.2 a – Three-legged junctions b – Four-legged junctions All accidents Pedestrian accidents Accidents involving buses Sites Total Serious Fatal Total Serious Fatal Total Serious Fatal PTR junctions 8.2* 0.9* 0.4* 1.9* 0.5 0.3 1.8 0.4 0.4* CG junctions 5.0 0.3 0.1 0.4 1.2 0.1 0.1 * Significant difference

# in 2010-2012

SPIs: the concept (III)

Study 1: Main conclusions

• Crossing type 4 gradated right-left, was consistently found to perform

worst, type 3 gradated left-left - to perform best. In addition, the accident analysis results provide an indication that a direct crossing (type 1 and type 5) is safer than a mixed-shifting (type 4)

• Pedestrian crossing configurations of types 1, 3 and 5 are recommended
• Behaviour observations at PTR junctions: a significant proportion of

pedestrians cross on red: 7-27%. It is important to create consecutive green phases for the pedestrians to minimize waiting times. Type 1 Type 3 Type 5 SPIs: the concept (III)

Study 2: Comparison of safety level of central PTR with

• ther types of bus lanes

Typical bus-lane layouts:

• Central PTR
• Right (curb-side) bus lane
• Left side bus-lane

Problem: a central PTR on Jabotinsky road, in Petah Tiqwa, experienced a lot

• f pedestrian accidents at intersections

Study aims: (1)To compare the safety level of Jabotinsky road with that of similar streets (2)To compare the safety level of central PTR with those of other bus-lane configurations Main findings: Arterials with intensive public transport have intensive land-use, vehicle and pedestrian activities. They generally have accident concentrations Jabotinsky road appears in the list of such streets but is not among the worst

Allenby str, Tel Aviv Hebron road, Jerusalem Keren Hayesod str, Jerusalem

Study 2: safety level of Jabotinsky road with central PTR

• vs. similar streets with public transport

Left bus-lane with PTR at stops Right bus-lane Central PTR

Data: 26 arterials with bus-lanes, in 9 towns Data on traffic volumes were collected Crash data for years 2010 – 2013 137 intersections and 92 road segments

Study 2: Detailed analysis of various bus-lane layouts

SLIDE 6

Road and traffic characteristics: Type of bus-lane layout (central PTR, right-side, left-side) Type of area (urban, interurban) Number of traffic lanes in each direction PTR with/without overtaking lane Level of pedestrian activity (low, high) Traffic volumes entering intersection in 14 hours Numbers of buses on the main roads Types of accidents analysed:

• All injury accidents
• Serious and fatal accidents
• Pedestrian accidents
• Accidents with bus involvement
• Accidents with pedestrian and buses

Study 2: Street characteristics examined

• A. Comparison of accident indices by PTR type:

annual number of accidents and accident rates per exposure (10 million vehicles per year, 1 million passing buses) or per section length T = ln(θ)/√(1/N1+1/N2)* where: θ = R1/R2 N1

• total number of accidents per study group

N2 - total number of accidents per comparison group R1 - accident index in study group R2 - accident index in comparison group Ho :θ =1, rejected when p<0.05

• B. Fitting explanatory models to predict the number of accidents at

intersections/sections, using available characteristics Multivariate Analysis of Variance (MANOVA) in SPSS v.22

Study 2: Methods of analysis

* Griffith, M. S. (1999) Statistical Analysis Techniques. Chapter 4 in: Statistical Evaluation in Traffic Safety Studies, Institute of Transportation Engineers

Study 2: Junctions - comparing accident indices by PTR type

Accident rates per 10 million entering vehicles, per junction, per year

Bus-lane type Index Total acc’s Severe acc’s Pedestrian acc’s Bus acc’s Acc’s with pedestrians and buses

Accident rates Central PTR

Mean

2.04 0.35 0.71 0.61 0.38

SD 1.60 0.47 0.86 0.92 0.66

Right bus-lane

Mean

1.03 0.15 0.26 0.06 0.02

SD 0.81 0.20 0.30 0.09 0.05

Left bus-lane

Mean

0.92 0.08 0.25 0.23 0.03

SD 1.17 0.18 0.37 0.27 0.08

Differences between the site groups

Central PTR vs right bus-lane T 8.30 3.98 6.53 8.99 7.22 p-value 0.000 0.000 0.000 0.000 0.000 Central PTR vs left bus-lane T 4.21 1.98 2.62 2.47 2.61 p-value 0.000 0.048 0.009 0.014 0.009

Study 2: Junctions - comparing accident indices by PTR type

Accident rates per 1 million passing busses, per junction, per year

Bus-lane type Index Total acc’s Severe acc’s Pedestrian acc’s Bus acc’s Acc’s with pedestrians and buses

Accident rates Central PTR

Mean

7.69 1.15 2.44 1.94 1.09

SD 7.06 1.34 2.92 2.31 1.63

Right bus-lane

Mean

4.49 0.70 1.11 0.29 0.13

SD 4.27 1.02 1.30 0.47 0.37

Left bus-lane

Mean

4.49 0.28 0.79 0.74 0.09

SD 6.14 0.59 1.17 0.89 0.28

Differences between the site groups

Central PTR vs right bus-lane T 6.53 2.30 5.00 7.35 5.10 p-value 0.000 0.022 0.000 0.000 0.000 Central PTR vs left bus-lane T 2.84 1.97 2.85 2.42 2.49 p-value 0.005 0.049 0.004 0.015 0.013

Study 2: Junctions - comparing accident indices by PTR type

Average accident numbers per junction, per year

Bus-lane type Index Total acc’s Severe acc’s Pedestrian acc’s Bus acc’s Acc’s with pedestrians and buses

Accident rates Central PTR

Mean

2.25 0.38 0.75 0.64 0.40

SD 2.22 0.60 1.12 1.18 0.87

Right bus-lane

Mean

1.55 0.20 0.34 0.09 0.03

SD 1.51 0.30 0.45 0.14 0.08

Left bus-lane

Mean

0.85 0.07 0.23 0.20 0.09

SD 0.93 0.14 0.35 0.27 0.18

Differences between the site groups

Central PTR vs right bus-lane T 4.84 3.16 5.31 8.70 7.09 p-value 0.000 0.002 0.000 0.000 0.000 Central PTR vs left bus-lane T 7.27 4.04 4.67 4.02 3.46 p-value 0.000 0.000 0.000 0.000 0.001

Study 2: Explanatory models for accidents at junctions

Explanatory variables B Sig.

Total accidents

Intercept

• .667

.488 Right bus-lane vs. central PTR

• 1.457

.023 Left bus-lane vs. central PTR

• 1.145

.006 Traffic volume on secondary road

↑

.786 .000 No of buses on the main road ↑ .261 .022 Traffic volume on the main road .201 .147 Pedestrian activity (1 – high) .054 .904 Type of area (0 - urban, 1 - interurban)

• .573

.312 Number of traffic lanes .047 .918 PTR overtaking lane (1 – yes) .568 .431 Year .092 .344

Pedestrian accidents

Intercept .531 .230 Right bus-lane vs. central PTR

• .471

.108 Left bus-lane vs. central PTR

• .349

.066 Traffic volume on secondary road ↑ .103 .056 No of buses on the main road .057 .276 Traffic volume on the main road .091 .152 Pedestrian activity (1 – high)

↓

• .480

.020 Type of area (0 - urban, 1 - interurban)

↓

• .809

.002 Number of traffic lanes

• .061

.767 PTR overtaking lane (1 – yes) .234 .480 Year .061 .170

SLIDE 7

Study 2: Explanatory models for accidents at junctions – main findings

Right- and left- bus-lane configurations are associated with accident decrease compared to central PTR, in all accident types Increase in the number of buses leads to an increase in total, bus and bus-pedestrian accidents Higher pedestrian activity is associated with a decrease in pedestrian, bus and bus- pedestrian accidents PTR overtaking lane is associated with an increase in bus and bus-pedestrian accidents Interurban area decreases the number of pedestrian accidents Higher traffic volumes on secondary road increase the number of total and severe accidents The impact of traffic volumes on the main roads, number of lanes not significant

Study 2: Sections - comparing accident indices by PTR type

Accident rates per 10 million passing vehicles, per year

Bus-lane type Index Total acc’s Severe acc’s Pedestrian acc’s Bus acc’s Acc’s with pedestrians and buses

Accident rates Central PTR

Mean 0.10 0.10 0.10 0.10 0.10 SD 0.33 0.33 0.33 0.33 0.33

Right bus-lane

Mean 0.79 0.13 0.39 0.06 0.04 SD 0.91 0.43 0.80 0.18 0.17

Left bus-lane

Mean 1.71 0.79 1.09 0.44 0.44 SD 3.55 1.84 1.82 0.76 0.76

Differences between the site groups

Central PTR vs right bus-lane T

• 2.09
• 0.28
• 1.38

0.44 0.71 p-value 0.036 0.779 0.169 0.660 0.477 Central PTR vs left bus-lane T

• 2.70
• 1.83
• 2.22
• 1.33
• 1.33

p-value 0.007 0.067 0.026 0.184 0.184

Study 2: Sections - comparing accident indices by PTR type

Average accident numbers per road km, per year

Bus-lane type Index Total acc’s Severe acc’s Pedestrian acc’s Bus acc’s Acc’s with pedestrians and buses

Accident rates Central PTR

Mean 0.06 0.06 0.06 0.06 0.06 SD 0.23 0.23 0.23 0.23 0.23

Right bus-lane

Mean 0.86 0.14 0.37 0.07 0.04 SD 1.09 0.47 0.81 0.22 0.20

Left bus-lane

Mean 0.50 0.22 0.35 0.14 0.14 SD 1.04 0.58 0.61 0.34 0.34

Differences between the site groups

Central PTR vs right bus-lane T

• 2.58
• 0.74
• 1.71
• 0.03

0.33 p-value 0.010 0.462 0.088 0.979 0.738 Central PTR vs left bus-lane T

• 1.93
• 1.07
• 1.56
• 0.65
• 0.65

p-value 0.054 0.285 0.119 0.514 0.514

Intersections situated on central PTR have a lower safety level compared with other bus-lane configurations Much attention has to be given to the detailed design features at PTR intersections, with particular emphasis on pedestrian arrangements Road sections of central PTR have a higher level of safety compared with other bus-lane configurations, mainly due to the extensive presence of median barriers/ fencing

Study 2: Safety level of central PTR vs. other bus-lane configurations - Conclusions

Main focus on pedestrian safety Behaviour studies may supply initial insights More simple engineering solutions are preferable Remember to keep a balance between promoting public transport use and pedestrian safety

Final remarks

Main questions remain with regard to particular design features of PTR intersections, bus stops, pedestrian settings and their impacts

• n road safety

More research efforts are required to ascertain safety effects

Thank you!