Directorate of Road Network Performance
P2Ws in bus lanes Main Report
`
P2Ws in bus lanes study
Main Report
Transport for London - December 2007
Page 1
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
Contents
1.0 Executive Summary
4
2.0 Introduction
7
3.0 Trial Objective
8
4.0 Report Structure
9
4.1
Background
9
4.2
Original trial report
9
4.3
Extended trial report
10
4.4
Final trial report
11
5.0 Collision Analysis - Before and after data
13
comparison
5.1
Vulnerable Road Users
14
5.1.1 Powered Two Wheelers
14
5.1.2 Pedal Cycles
15
5.1.3 Pedestrians
15
5.1.4 Bus occupants
16
5.2
Conclusions – Casualty analysis
16
6.0 User and Attitudinal surveys
17
6.1
Opinion Surveys
17
6.2
Speeds and bus journey times
17
6.3
Vehicle usage and Migration
19
7.0 Experiences from other schemes
20
7.1
Bus Lane trials in London – M4 motorway
21
7.2
Other London borough schemes
21
7.3
UK schemes outside London
22
7.4
Conclusions – experiences from other schemes
22
Transport for London - December 2007
Page 2
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
8.0 Conclusions
25
8.1
Context
25
8.2
Control method changes
25
8.3
Findings
26
8.3.1 Opinion surveys
26
8.3.2 Cyclists
26
8.3.3 Pedestrians
26
8.3.4 Bus occupants
27
8.3.5 Powered Two Wheelers
27
8.4
Change in use by modes
27
8.5
Summary
27
Appendices
Appendix I
Casualty analysis using Original control method
Appendix II Casualty analysis using Tanner control method
Appendix III User and attitudinal surveys
Appendix IV Comparison of two control methods against trial data
Transport for London - December 2007
Page 3
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
1.0 Executive Summary
I. Collisions involving Powered Two Wheeler (P2W) riders generate comparatively high
numbers of casualties in London, relative to the size of this group among road users.
Accordingly, P2W riders are identified as highly Vulnerable Road Users (VRUs), and
share this category with cyclists and pedestrians.
II. Whilst the Mayor’s Transport Strategy recognises that P2Ws “can generate more
pollution and noise” than cars it also recognises their use as a “quick, relatively low
cost private transport that are more space and fuel efficient than cars” (Chapter 4G.26)
and in response to the P2W lobby committed to implementing a trial of P2W usage in
bus lanes (Proposal 4G.1).
III. The primary objective for this study was to investigate and offer evidence to show
whether the safety of P2W users could be enhanced by allowing them access to bus
lanes (the measure), The assessment was to involve comparing the casualty history of
all vulnerable road users (VRUs) to ensure the measure does not create a negative
impact on themselves or other road users.
IV. In September and October 2002, TfL introduced three pilot schemes on the Transport
for London Road Network (TLRN) whereby P2Ws were permitted to use bus lanes
along the three routes (A41, A13 and A10) during the hours of operation.
V. The report has evolved since the original study was commissioned and complex arrays
of factors and concerns have emerged during the course of the trial. For ease of
reference the three iterations of the trial’s development are:
• Original trial – (Duration 18 months) instigated as part of the Mayor’s commitment in
his transport strategy to reduce P2W casualties.
• Extended trial – (Duration 36 months) undertaken as a result of the findings of the
original trial report.
• Final trial – (Duration 36 months) Involved re-writing the extended trial to address the
issues raised by the stakeholders.
VI. This report addresses issues raised by stakeholders at a seminar held by TfL in
September 2006, where preliminary findings from the extended trial were presented.
These included the exclusion of one trial route as the collision history was adversely
affected by roadworks during the trial, and an investigation to see whether the potential
effect of migration to and away from the trial routes could be investigated.
VII. Analysis of the casualty history on the trial routes for a 36 month period before and
after implementation of the measure has been compared with two different control
methods, the original control route data and a statistically more robust “Tanner control”
method introduced after 18 months.
Transport for London - December 2007
Page 4
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
VIII. The casualty history on each trial route was compared to both controls using four
different scenarios (two during all hours and two during bus lane operating hours). This
resulted in eight comparison analyses for each VRU group.
IX. The assessment of
P2W collision numbers showed six of the eight analyses as being
beneficial to the safety of P2W riders and two of the eight showed a disbenefit to this
group. All four assessments during operational hours showed a safety benefit.
X. The
pedal cycle casualty numbers showed three of the eight analyses as being
beneficial to the safety of pedal cyclists and five of the eight showed a disbenefit to this
group. Two of the four assessments during operational hours showed a safety benefit.
XI. The
pedestrian casualty numbers showed two of the eight analyses as being
beneficial to the safety of pedestrians, five returned a disbenefit and one showed no
change.
XII. None of the figures produced are statistically significant.
XIII. It was intended to undertake an assessment of casualty rates (number of collisions per
P2W or cycle journey made), however the only data available for this purpose is the
Department for Transport’s Average Annual Daily Flow (AADF) counts which are
produced at over 15,000 locations across the UK using a variety of different methods.
These different methods can result in inaccuracies and variability in the data and
generally proponents of the data limit its use to studies of local areas and
understanding patterns.
XIV. The DfT is not aware of widespread use of the data for studies of trials or schemes and
whilst they would not be concerned about the use of AADFs for the purposes of this
study, there are some significant variations in the data that might produce spurious
casualty rate results. This report, therefore, does not consider it suitable to pursue the
issue of migration and casualty rates using the available data.
XV. Figures from the Original Trial Report demonstrate that the bus speeds on all trial
routes increased.
XVI. Opinion surveys on behalf of TfL were made available for this study. Studies targeted
motorcyclists, cyclists, bus drivers, car drivers, pedestrians and the general public.
XVII. The surveys revealed that only motorcyclists and car drivers (who are not bus users)
were the only two user groups who approved of the measure.
XVIII. Amongst other VRUs the main reason for disapproval of the measure was their
perception of the compromise to safety.
XIX. The comparisons of collisions involving VRUs using the Tanner control showed neither
a benefit nor disbenefit from the introduction of the measure. None of the results from
any of the user group comparisons were statistically significant.
When the VRU collisions were assessed against the original control routes (excluding the
A13) during bus lane operating hours, a slight net benefit was returned. However, included in
this reduction were localised increases in pedestrian casualties on the A41. None of the
results were statistically significant.
Transport for London - December 2007
Page 5
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
2.0 Introduction
Collisions involving Powered Two Wheeler (P2W) riders generate comparatively high
numbers of casualties in London, relative to the size of this group among road users.
Accordingly, P2W riders are identified as highly Vulnerable Road Users (VRUs), and share
this category with cyclists and pedestrians.
An extensive range of measures is being developed and deployed specifically to improve the
safety of cyclists and pedestrians. In contrast, apart from the general use of safety cameras,
trials of P2Ws in bus lanes and access to advanced stop lines at controlled junctions, no
practical traffic management measures have been introduced in London specifically to
enhance the safe use of the P2W mode.
The Mayor and TfL recognise that the P2W has a positive role to play in reducing congestion
and associated pollution by offering a practical and efficient alternative to four wheeled
motorised modes – especially cars and vans for courier and light freight. Other benefits to
transport in London stem from low financial costs associated with the mode. Many P2Ws,
including small motorcycles, mopeds and scooters, are cheap to buy and run, and are
increasingly recognised by TfL and nationally by the Department for Transport (DfT) as an
important component in tackling social and economic exclusion.
Currently, the extent of net benefits that the use of P2Ws offers to the population of London
and its visitors is significantly limited by the high costs of relatively large numbers of collisions
and casualties.
The number of trips and kilometres travelled by P2Ws have increased in London steadily from
1995 to 1999 and has remained around the 1999 level since. Correspondingly, the
development of innovative traffic management measures to improve the safe use of this mode
is becoming an increasingly important goal.
Proponents of P2W use consider that allowing this mode to use bus lanes would dramatically
improve the safety of riders, and all other road users
1. If these expectations prove well
founded, roll out of the measure has the added benefit of involving relatively simple and small
changes to streets with existing bus lanes. Apart from identifying where to introduce the
improvement first, the main task and cost will be to renew bus lane signage. Should it be
shown that allowing motorcycles access to bus lanes would improve the safety of all road
users, the costs of implementing this measure would be minimal in comparison to other major
schemes.
Whilst the Mayor’s Transport Strategy recognises that P2Ws “can generate more pollution
and noise” than cars it also recognises their use as a “quick, relatively low cost private
transport that are more space and fuel efficient than cars” (Chapter 4G.26) and in response to
the P2W lobby committed to implementing a trial of P2W usage in bus lanes (Proposal 4G.1).
The outcome from such a trial was deemed to be of incalculable value to the cause of
improving the safe use of P2Ws, and road safety in general throughout the capital and the
UK.
Transport for London - December 2007
Page 6
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
3.0 Trial Objective The primary objective for this study was to investigate and offer evidence to show whether the
safety of P2W users could be enhanced by allowing them access to bus lanes (the measure),
The assessment was to involve comparing the casualty history of all vulnerable road users
(VRUs) to ensure the measure does not create a negative impact on other road users.
Relatively high casualty numbers associated with use of mopeds, scooters and motorcycles
provide an ongoing cause for concern in London. P2W Killed or Seriously Injured casualties
(KSI) for the Capital have been the road user category that has shown least progress towards
the 2010 casualty reduction targets. However, it is recognised that this has been against a
background of growth in ownership and usage.
The Mayor's Transport Strategy
2 includes a specific response to concerns about P2W
casualties. Proposal 4G 1, committed TfL to consider trials allowing P2Ws into bus lanes as a
means of potentially reducing the exposure of P2W riders to general traffic in order to improve
safety.
TfL took these factors into account in designing a study which considered the safety of all
road users. The primary objectives for the trial were to:
• Investigate whether allowing P2W use of bus lanes would be an effective way to
improve the safe use of the P2W mode in London.
• Discover whether significant disbenefits would arise, with particular concern for other
vulnerable road users, especially cyclists and pedestrians.
• Publish the trial data and results of comparative calculations to quantify the tangible
positive and negative impacts of the measure.
• Draw conclusions from the experiment data where possible, to provide a basis for
future action by TfL, and to assist other transport authorities in considering the merits
of the measure on test.
4.0 Report Structure
4.1
Background
The report has evolved since the original study was commissioned and complex arrays of
factors and concerns have emerged during the course of the trial. For ease of reference the
three iterations of the trial’s development are:
• Original trial – (Duration, 18 months) instigated as part of the Mayor’s commitment in
his transport strategy to reduce P2W casualties.
• Extended trial – (Duration 36 months) undertaken as a result of the findings of the
original report.
• Final trial – (Duration 36 months) Involved re-writing the extended trial to address the
issues raised by the stakeholders.
These are detailed in the sections below.
Transport for London - December 2007
Page 7
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
4.2
Original trial report
In September and October 2002, TfL introduced three pilot schemes on the Transport for
London Road Network (TLRN) whereby P2Ws were permitted to use bus lanes along the
three routes during the hours of operation.
The original trial was reported on after 18 months and the report published on 19th November
2004
3.
The report concluded that
“further casualty data is needed in order to make a robust
assessment of the trials.”
After consultation with stakeholders, the trial was extended for a further 18 months.
4.2.1 Original trial – casualty analysis scope and plan
The original procedure for the trial required the identification of ‘trial’ sections of highway in
which P2Ws were allowed in bus lanes, and ‘control’ sections of highway without P2W access
to bus lanes.
Detailed records of all reported collisions and casualties were gathered from trial and control
corridors throughout the eighteen month duration of the experiment. Data were also gathered
to establish an accurate record of reported collisions and casualties on the corridors before
the experiment began.
4.2.1.1
Methodology - Key criteria for selection of trial and control sites3
• Part of TLRN
• Known locations of motorcycle usage
• Standard with-flow bus lane
• Mix of frontage (residential, commercial)
• Mix of location, but outside congestion charge area
• High but not excessive casualty rates
4.2.1.2
Sites chosen for trial routes:
• A13 East India Dock Road, between Leamouth Road and Butcher Row East - from 9th
Sept 2002;
• A23 Brixton Road, between Camberwell New Road and Streatham Common South; -
from 20th Oct 2002; and
• A41 Finchley Road, between Queen’s Grove North and Platt’s Lane;
- from 20th Oct 2002.
4.2.1.3 Sites chosen for control routes:
• A5 Rondu Road to Summit Close and Humber Road to Staples Corner
• A10 Pasteur Gardens to Ostliffe Road and Wilbury Way to Laburnum Avenue
However, there were a number of concerns about this method, namely:
Transport for London - December 2007
Page 8
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
• The section of the carriageway of the A13 used for this trial underwent
considerable roadworks during the after period of the trial. The extent of works
culminated in a 22% increase in KSIs compared to a 35% fall in the KSI rate
across the TLRN network. It is for this reason that the final report does not
consider this data to be valid and therefore the comparisons in this final report
disregard the A13 data when the assessment involves either the original control
data or the Tanner Test.
• The A10 may not be a suitable control as it is in North London and generally in
outer London whilst the trial sites on the A41 and A23 are generally in inner
London.
• The A5 may not be a suitable control for the A41 as it was a parallel route and
therefore could be directly influenced by the trial.
• The control sites were considered too small for meaningful comparison
4.3
Extended trial report
To avoid the methodological problems listed above the method of control was changed from
the route based comparison, to an area based control that satisfied the requirements of the
Tanner Test, a formula devised by J C Tanner in work for the Transport Research Laboratory,
last published in the early 1980s.
The ‘Tanner Test’ formula became a new element in the method by which control data could
be generated, and a detailed description of how the Tanner Test was applied can be found in
Appendix II. In summary it involved collision figures from the entire TLRN divided into three
sets of figures with the results for the boroughs nearest to the trial corridors being used as the
control. The disadvantages of this method include the fact that it uses a control ratio rather
than absolute figures,
4 which means that collision rates are not able to be compared.
The publication of the extended report was proposed to be in the form of two types of test
procedure which could be used to assess the impact of introducing the measure.
• A practical experiment – to quantify the tangible impact on collisions and casualties
during a thirty-six month before and after trial period (using the Tanner Test).
• Attitudinal surveys – to gauge the opinions and feelings of road users about the
measure.
4.3.1 Extended trial – Casualty analysis scope and plan
The extended trial used the same start date as the original report (October 2002).
The key criteria and the sites chosen for the trial routes remained consistent with the original
scope (detailed in 4.2, above), but two changes were made to the trial structure
Firstly, the duration of the before and after casualty studies was extended from 18 months to
36 months.
The second change involved the method of assessing the trial corridors against the control
data. To satisfy the Tanner Test, collision figures from the entire TLRN were divided into three
sets of figures from the boroughs nearest to the trial corridors.
Transport for London - December 2007
Page 9
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
4.4
Final trial report
However, a number of concerns were raised by stakeholders when a verbal account of the
draft report on the casualty analysis (using Tanner) and an Executive Summary of the user
and attitudinal surveys were circulated at a seminar held by TfL in September 2006. It was
agreed that the issues raised at the seminar would be addressed in the final report.
As a consequence of the issues raised by stakeholders at the seminar held in September
2006 (section 4.3), the following areas of contention were considered in the final report:
• As a result of extending the trial period to 36 months before and after analysis of the
casualty data the method of control altered in mid trial from control routes to the use of
the Tanner Test which allowed for a more rigorous statistical analysis. However, the
stakeholders were not informed of the change.
• As a result of this change the new control method did not enable collision rate
comparison (due to possible migration, which was a factor not considered during the
original scoping of the trial) to be assessed.
• The Original Trial remit did not consider experiences from other authorities in the UK
and abroad.
• The use of the A13 as a trial corridor was flawed due to the presence of roadworks for
the duration of the after period. This skewed the collision numbers upwards.
• There was an element of subjectivity in the attitudinal and user surveys
• The impact of congestion charging (which started 4 months after this trial) was not
considered.
Obviously, the most important issue to be considered when assessing the use of bus lanes by
P2Ws is the safety of all road users. The views of the stakeholders are reflected in the new
suite of documents which focus on the importance of safety. All existing data has been re-
assessed and the casualty history for the trial routes has been assessed against the original
control routes and also the revised control method.
The data considered meaningful from the original user survey and attitudinal surveys are also
re-visited and presented here.
The report structure is as follows:
MAIN
REPORT
Appendix I - Collision
Appendix II - Collision
Appendix III - User
Analysis using original
Analysis using revised
and attitudinal
control data
control (Tanner)
surveys
Appendix IV –
Comparison of two
control methods against
trial data
Transport for London - December 2007
Page 10
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
This document (the main report) pulls together the evidence contained within the supporting
documents (which are appended to this report) and assesses potential benefits and
disbenefits of the trial. The report widens the scope of the trial to enable the concerns raised
by the stakeholders to be addressed. This has enabled factors previously not considered to
be considered, such as the experiences of some of the other schemes that have been
introduced in the UK and overseas, with an investigation of the issue surrounding migration of
P2Ws between bus lanes.
4.4.1 Final trial – casualty analysis scope and plan
All data have been re-assessed for relevance and accuracy for the purposes of assessing the
viability of this measure. As requested at the stakeholders meetings, two parallel casualty
reports have been produced (see Appendix I – Casualty analysis using the original control
method, and Appendix II – Casualty analysis using the Tanner control method). A summary of
the comparison of the two control methods against the trial data is contained within this
report, with further, more detailed analyses in Appendix IV.
5.0 Collision Analysis - Before and after data comparison
The summary comparisons made in this section use the casualty data collected during the
trial period and documented in Appendices I and II. These two documents share the same
trial route data but differ in the way the control data are derived. Full details of the comparison
analyses can be found in Appendix IV.
Appendix I derives its data from the empirical evidence gathered from the original trial control
sites casualty histories. These have been produced following the most recent review
commissioned after the October 2006 Stakeholder seminar.
Appendix II has been compiled from the same before data for comparison with figures
resulting from extrapolations of TLRN statistics (see section 4.3), and including use of the
Tanner formula. Data for the A13 is presented in this document but the assessments
contained in this section only use that for the A41 and A23.
This report focuses on the impact of the measure on casualties in the VRU groups that may
be affected as this is seen as essential to the overall outcome of the report. To complement
this analysis and help understand general trends, all casualties are also considered.
For each VRU group assessment is carried out of the combined trial routes against the
combined control routes and the Tanner control areas.
As stated in section 4.4.1, concerns have been noted that the reliability and validity of data
from the A13 trial site were adversely affected by the extensive disruption caused by a major
redevelopment program of works. Consequently the review using both control data
concentrates on the trial as a whole but discounts the A13 route. The comparisons between
sums of data from the A5 + A10 control sites, and sums of data from the A23 & A41 trial sites
offer the next most useful combination of comparable data, after analysis of the A41 and A5
results.
Transport for London - December 2007
Page 11
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
Each VRU group is also assessed using data from the single trial route, the A41 and
compared to the single control route the A5 as the A41 trial site runs parallel to the A5 control
site. In many respects this makes data from these sites the
most directly comparable and a potentially useful gauge of the impacts of the measure.
The VRU groups reported on are:
• P2Ws
• Pedal Cycles
• Pedestrians
• Bus occupants
5.1
Vulnerable Road Users
The assessments contained within sections 5.1.1 to 5.1.3 below consider all collision types
where the casualties (of all severities) have been the respective vulnerable road user.
Each assessment investigates four scenarios
1. A41 trial route against control route (A5) and control area (using the Tanner Test
control) for all times of day.
2. A41 trial route against control route (A5) and control area (using the Tanner Test
control) for operating hours only.
3. A41 & A23 trial routes combined data against A5 & A10 control routes combined
data and the Tanner Test for all times of day.
4. A41 & A23 trial routes combined data against A5 & A10 control routes combined
data and the Tanner Test for operating hours only.
The assessments provide a very useful picture of how the measure performed in relation to
VRUs and are summarised in the tables below. A full analysis of all the figures in the tables
below can be found in Appendix IV.
5.1.1 Collision Analysis - Impact on Powered Two Wheeler Riders
Using the original control data, three of the four graphs (taken from graph numbers 1 to 4,
Appendix IV and summarised in Table 1) show a safety benefit. Using figures from the Tanner
Test calculations, the graphs also reflect a safety benefit in three cases. There is a safety
benefit in all cases when bus lane operating hours are active.
%age change in
P2Ws
collisions, trial against
(Appendix IV
controls
Graphs 1 - 4)
(
Reduction, Increase)
Scenario
Original
Tanner
(section 4.1)
Control
Control
1
1
7
2
21
1
3
8
1
4
8
3
Table 1 – Percentage change in “Before” and “After” casualty numbers from the four scenario comparisons for P2Ws
Transport for London - December 2007
Page 12
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
5.1.2 Collision Analysis - Impact on Pedal Cyclists
The analysis of the four scenarios (see section 5.1) showed mixed results. The original
control data comparison (see Table 2) shows safety benefits in scenarios 2 and 4 (operating
hours) with disbenefits in 1 and 3 (all hours). The Tanner control comparison shows a
disbenefit in three of the four scenarios. The changes are not statistically significant.
%age change in
Pedal cycles
collisions, trial against
(Appendix IV
controls
Graphs 5 – 8)
(
Reduction, Increase)
Scenario
Original
Tanner
(section 4.1)
Control
Control
1
20
19
2
19
4
3
4
11
4
13
3
Table 2 – Percentage change in “Before” and “After” casualty numbers from the four scenario comparisons for pedal cycles
5.1.3 Collision Analysis - Impact on Pedestrians
Analysis of the trial routes against the original control data showed an increase in collisions
involving pedestrians in three out of four scenarios and in two out of four scenarios in the
Tanner Test comparison. (See Table 3).
Of the two that showed a benefit, both were during operating hours.
These results are not statistically significant.
%age change in
Pedestrians
collisions, trial against
(Appendix IV
controls
Graphs 5 – 8)
(
Reduction, Increase)
Scenario
Original
Tanner
(section 4.1)
Control
Control
1
27
32
2
40
66
3
3
0
4
15
5
Table 3 – Percentage change in “Before” and “After” casualty numbers from the four scenario comparisons for pedestrians
5.2
Conclusions - Impact on Vulnerable Road Users
The
VRU casualty number assessments were carried out using four scenarios. Each
scenario compared the trial against both types of control, resulting in eight analyses for each
VRU group.
Transport for London - December 2007
Page 13
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
The assessment of
P2W collision numbers showed six of the eight analyses as being
beneficial to the safety of P2W riders and two out of eight showed a dis-benefit. All four
assessments during operational hours showed a safety benefit.
The
pedal cycle casualty numbers showed three of the eight analyses as being beneficial
to the safety of pedal cyclists and five out of eight showed a disbenefit. Two of the four
assessments during operational hours showed a safety benefit.
The
pedestrian casualty numbers showed two of the eight analyses as being beneficial to
the safety of pedestrians, five returned a disbenefit and one showed no change.
None of the figures produced are statistically significant.
6.0 User and Attitudinal surveys
To understand fully how riders of P2Ws and pedal cycles would adjust to the measure,
vehicle counts before and after the implementation of the measure were assessed. The
attitude of all road users towards the measure was also investigated using opinion poll
surveys
The full report and data sets used can be found in Appendix III, and this section summarises
the results. In order to address the concern raised by stakeholders about the casualty rates
on the trial and control routes, it is necessary to explore whether any robust vehicle usage
data is available. Any such data would allow a casualty rate to be derived by simply dividing
the casualty numbers by the usage to produce a casualty rate. However, the only available
data that could potentially be used retrospectively for this exercise is the estimated Annual
Average Daily Traffic (AADF) flows from DfT counters which exist extensively across the
TLRN network. The validity of these counts is explored in this section.
6.1
Opinion surveys
Opinion surveys on behalf of TfL were made available for this study. Studies targeted
motorcyclists, cyclists, bus drivers, car drivers, pedestrians and the general public and details
of all surveys can be found in Appendix III.
The surveys did not reveal many surprises in their findings with only motorcyclists and car
drivers (who are not bus users) being the two user groups who approved of the measure.
Amongst the VRU groups, the main reason for disapproval of the measure was their
perception of the compromise to safety.
The findings also showed that pedal cyclists are viewed by bus drivers as being the most
problematic of all road users.
It must be noted that the small size of the response groups to some of the surveys might not
be representative of that groups’ opinion.
Transport for London - December 2007
Page 14
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
6.2
Speeds and bus journey times
One concern for the introduction of the measure relates to the potential for the delay to buses,
which obviously contradicts the commitment in the Mayor’s Transport Strategy for the
expeditious movement of public transport.
The assessment of the impact of the measure on the speed of buses has to rely on the only
“before and after” comparisons that were available. These were presented in the Original trial
report
3 and are detailed below.
These “before” and “after” data (Table 5 - shown below) only provides 18 months figures but
is sufficient enough to make a valid comparison.
Table 5 – Before and After speed data from Original Trial report
The table demonstrates that the bus speeds increased in all cases but when assessing this
data the following must be considered,
• The “before” P2W speeds in the trial bus lanes and the “before” and “after” speeds
from the control routes are based on illegal usage of the lanes by P2Ws and cannot
be considered representative.
The Original Trial Report also reported that bus journey times fell by 1 to 2 minutes on the
A13 and A41, but increased by 1 minute on the A23. The document states that the following
must be considered,
• The A13 bus routes were affected by roadworks
• The A23 journey times were affected by changes to the bus route.
The Extended Trial Report carried out a more detailed assessment of the bus journey times
on the three trial and two control routes and is included in Appendix III.
The data sets were collected using the “Marquis” roadside beacon infrastructure that records
the time of each bus that passes.
The report states that there was
• “
remarkable consistency in average speed provided by bus lanes”
• “
Most routes have seen an improvement in reliability” culminating in a “
range in
journey time of just two minutes on a 30 minute journey”,
though it wasn’t stated that this was as a result of the trial.
Transport for London - December 2007
Page 15
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
However the reliability of this data is subject to the following:
• The start date for the collection of the data is April 2004, 18 months after the
implementation of the trial routes.
• There is no before and after comparison of journey times from these data sets.
• The journey times recorded do not consider the effect of the length of time spent at
each stop due to the number of passengers alighting or disembarking.
6.3
Vehicle Usage and Migration
The potential for migration of P2Ws and indeed cyclists to or away from the trial routes must
be understood in order to gauge the impact this might have when comparing casualty
statistics.
The scope of the original trial did not account for the collection of meaningful vehicle usage
data and so alternative data sets were sought.
The only available data available retrospectively for the six year period of the trial is the
Department for Transport’s (DfT) Average Annual Daily Flow (AADF) data. Consequently,
investigation of the DfT AADF data was undertaken for both trial routes and comparison
AADFs from a similar route that follows each trial route in relatively close proximity. The data
of the A41 were compared to the control route A5 corridor. However, the path of the A23 in
south west London bears no similarities to that of the other control route, the A10 which is in
north London. Therefore, for this exercise only, a new data set was introduced to enable a
more accurate comparison and that was collected from AADF flows on the A24 in south west
London.
Tables 6 and 7 show the figures obtained from the three DfT AADF counter positions on each
of the trial and control routes.
Table 6 Estimated AADT flows of P2Ws and cyclists for A41 trial route and A5 control
route
P2W
2000
2001
2002
2003
2004
2005
A41
A5
A41
A5
A41
A5
A41
A5
A41
A5
A41
A5
Outer
712
533
770
577
737
695
762
719
728
367
668
338
Central
1370
658
896
732 1548
685
2440
667 1190
700 2004
698
Inner
768
658
866
732 1121
685
1540
667 1999
700 1894
698
Average
950
616
844
680 1135
688
1581
684 1306
589 1522
578
CYCLISTS
Outer
50
256
55
281
57
253
83
367
72
152
73
155
Central
300
304
218
346
717
352
488
417
252
441
419
569
Inner
191
304
198
346
477
352
350
417
500
441
564
569
Average
180
288
157
324
417
319
307
400
275
345
352
431
Transport for London - December 2007
Page 16
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
Table 7 Estimated AADT flows of P2Ws and cyclists for A23 trial route and A24 control
route
P2W
2000
2001
2002
2003
2004
2005
A23
A24
A23
A24
A23
A24
A23
A24
A23
A24
A23
A24
Outer
624 1000
685 1081
773
659
799
681
632
651 1173
598
Central
1503 1544 1673 2119 1895 1985 1974 2457 2049 2582 2043 1629
Inner
509 1294
567 2281
531 2138
553 1772
582 2923
580 2241
Average
879 1279
975 1827 1066 1594 1109 1637 1088 2052 1265 1489
CYCLISTS
Outer
130
388
315
426
136
205
197
298
332
256
450
262
Central
1152 1455 1312 1193
927 1214 1082 1214
797 1286 1030 1454
Inner
441
732
502 3309
511 3368
596 2088
632 2747
615 2582
Average
574
858
710 1643
525 1596
625 1200
587 1430
698 1433
However, before any meaningful assessment using the data was carried out, the quality and
accuracy of the data had to be established as the data in table 6 the annual production of the
data uses a variety of different methods. These different methods can result in inaccuracies
and variability in the data and generally proponents of the data limit its use to studies of local
areas and understanding patterns.
The DfT is not aware of widespread use of the data for studies of trials or schemes and whilst
they would not be concerned about the use of AADFs for the purposes of this study, there are
some significant variations in the data that might produce spurious casualty rate results. This
report, therefore, does not consider it suitable to pursue the issue of migration and casualty
rates using the available data.
A full explanation of the process for the production of the AADF data can be found in
Appendix III.
7.0 Experiences from other schemes
During the last eleven years, P2W access to bus lanes has been trialled or introduced as a
permanent measure in a number of locations throughout the UK.
The first Local Authority to implement this measure was Bristol, following a trial scheme.
Deployment of this measure in the City of Bristol has continued since 1996.
By the end of 2006, seventeen UK local authorities
6 had introduced this measure in varying
forms, and it is also deployed as a permanent measure throughout Northern Ireland. The
primary reason of all schemes is to improve road safety. Indeed, there are also a number of
schemes that have been implemented in London by some London boroughs.
A growing number of European towns and cities deploy this measure for road safety
purposes. Motorcycles have been allowed in the bus lanes of Sweden’s capital, Stockholm,
Transport for London - December 2007
Page 17
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
since 1986, and the same approach has been successfully adopted in Barcelona in Spain and
some Italian cities.
As far as could be established by the authors, no trial of P2W in bus lanes anywhere in the
UK – or anywhere else in the world – has ever resulted in a rejection of the measure for
safety reasons. Similarly, in all known trials to date, once P2W access to bus lanes has been
granted, such access has never been subsequently rescinded.
Since the TfL trial started, the DfT published ‘The Government's Motorcycling Strategy
7 which
followed a five year study by the Government Advisory Group for Motorcycling (GAGM). In
essence, government and the DfT now formally acknowledge that the P2W mode has vital
contributions to make in developing sustainable transport throughout Britain. The P2W offers
an extremely efficient and low financial cost alternative to four wheeler modes in many
situations where walking, cycling or public transport cannot meet demand for transporting
people and goods.
The national strategy document makes it clear that more can and should be done by Local
Transport Planning authorities to improve the safe use of P2Ws throughout the UK.
The principle aim of this major nationwide policy initiative is to ‘mainstream’ motorcycling. In
particular it says that taking account of motorcycling is to play a greater part in plans for road
design and traffic management.
7.1
Bus Lane trials in London – M4 motorway
Aside from the trials that are the subject of this report, a number of trial and permanent
schemes have been implemented in London.
Perhaps the most high profile of these is the offside bus lane implemented on the eastbound
carriageway of the M4 motorway from Junction 3 to the elevated section.
In July 2002 the eastbound carriageway speed limit was increased to 60mph from 50mph and
motorcycles were allowed to use the offside bus lane. The site was monitored by TfL’s
London Road Safety Unit who undertook a 36 month before and after study.
The study showed that the number of collisions involving injury decreased from 44 in the 36
months to July 2002 to 28 in the 36 months after. This decrease of 36% was found to be
statistically significant at the 10% level using the K test
8.
Collisions involving P2Ws reduced by nearly twice as much – from 12 to 4 which equates to
67%.
7.2
Other London borough schemes
There have been other trials in three London boroughs, Westminster, Kingston and
Richmond. Table 8 gives a summary of the performance of each scheme.
In September 2005 the City of Westminster introduced the measure in the form of a trial on a
total of ten lengths of bus lane. The overall impact suggests that there are safety benefits for
all vulnerable road users. In the 14 months of the measure, the figures have returned a 24%
reduction in pedestrian casualties, and 17% reduction for both pedal cycles and P2Ws.
Transport for London - December 2007
Page 18
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
There are four schemes introduced in the Royal Borough of Kingston, with two having
collected 36 months after data sets and two with 31 months data sets. The collective casualty
figures have shown reductions for all vulnerable road users with pedestrian casualties down
by 17%, P2Ws down by 29% and pedal cycle casualties down by 50%.
Two schemes have been introduced in one of Kingston’s neighbours, Richmond. These
schemes have been operating for 31 months and have seen a 33% reduction in pedestrian
casualties, and 67% reduction in P2W casualties. There has, however, been an increase from
0 to 3 in pedal cycle injuries.
None of these schemes has been implemented against a control area so no comparison can
be made with the general trends. However, it does give an indication that the measure in
differing scenarios can provide safety benefits to vulnerable road users in the vast majority of
cases.
7.3
UK schemes outside London
There has not been a trial of the measure in the UK that has undertaken a comprehensive
“before and after” casualty analysis using control sites. The majority of studies failed to
introduce trial routes and supplement the data with vehicle usage surveys and journey times.
However, some useful studies have been undertaken which have been assessed using
differing methodologies over the last decade.
Avon County Council first resolved to introduce motorcycles to bus lanes in Bristol using an
Experimental Order on 14th February 1995. The experimental order came into effect in June
of that year and was confirmed as a Permanent Order on 12th March 1996 when the
Committee, anticipating the imminent Local Government Reorganisation, strongly
recommended that its four successor Unitary Authorities should extend the scheme to their
areas (which each has subsequently done).
There has been much written on the Bristol scheme but the evidence recently given to the
aforementioned Transport Select Committee by the Motorcycle Industry Association (MCIA)
probably sums the results most succinctly.
“During the 36 months prior to the implementation, accidents involving motorcyclists averaged
1.1 per month, compared to 0.8 during the six-months of the experiment, suggesting a 25%
decrease, and that no motorcycle accidents were recorded in the bus lanes and no collisions
with pedestrians or cyclists were recorded”.
The same evidence also stated that “
an 18-month experiment by Sheffield City Council during
2003/04 which also reported a 25% decrease in monthly average motorcycle accidents.”
7.4
Conclusions – experiences from other schemes
No trials in the UK have been removed for safety reasons, though there has not been any
comprehensive casualty analysis trials undertaken.
The trials that have occurred have shown a reduction in casualties by up to 25% though none
of the trials has used a control for objective comparison with collisions after its introduction
Transport for London - December 2007
Page 19
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
Table 8 Before and after collisions for borough P2W in bus lane schemes at all times (by November 2006 data provisional)
N.B. This is for all hours/days and not just the bus lane hours of operation.
Borough
Scheme name
Months
Collisions Before
Collisions After
Change in collisions (Numeric)
Fat
Ser
Sli
Pedn
P2W
PC
Tot
Fat
Ser
Sli
Pedn
P2W
PC
Tot
Fat
Ser
Sli
Pedn
P2W
PC
Tot
Westminster
Cockspur Street
14
0
1
12
6
0
3
13
1
0
7
5
1
0
8
1
-1
-5
-1
1
-3
-5
Westminster
Kensington Gore
14
0
0
3
0
0
0
3
0
1
0
0
1
0
1
0
1
-3
0
1
0
-2
Westminster
Kensington Road
14
0
2
8
2
6
1
10
0
3
3
1
1
2
6
0
1
-5
-1
-5
1
-4
Westminster
Knightsbridge
14
0
0
4
2
3
0
4
0
0
1
0
0
0
1
0
0
-3
-2
-3
0
-3
Westminster
Baker Street
14
0
1
14
6
3
1
15
0
0
10
3
3
1
10
0
-1
-4
-3
0
0
-5
Westminster
Haymarket
14
0
2
10
5
2
2
12
1
1
8
3
1
3
10
1
-1
-2
-2
-1
1
-2
Westminster
Bayswater Road
14
0
2
1
3
0
0
3
0
1
4
0
2
1
5
0
-1
3
-3
2
1
2
Westminster
Piccadilly (SW end)
14
1
0
6
1
2
1
7
0
2
5
3
3
1
7
-1
2
-1
2
1
0
0
Westminster
Piccadilly (NE end)
14
0
5
30
14
7
3
35
0
4
23
13
6
2
27
0
-1
-7
-1
-1
-1
-8
Westminster
Gloucester Place
14
0
1
6
2
1
1
7
0
1
9
3
2
0
10
0
0
3
1
1
-1
3
Kingston
Cambridge Road
36
0
0
8
2
1
2
8
0
0
6
2
2
0
6
0
0
-2
0
1
-2
-2
Kingston
Cambridge Road
36
1
6
25
7
8
2
32
0
8
22
7
9
3
30
-1
2
-3
0
1
1
-2
Kingston
London Road (SB)
31
0
0
6
2
3
0
6
0
0
1
0
0
0
1
0
0
-5
-2
-3
0
-5
Kingston
London Road (NB)
31
0
0
11
1
5
2
11
0
2
4
1
1
0
6
0
2
-7
0
-4
-2
-5
Richmond
London Road, TW10
31
0
1
7
3
1
0
8
0
1
5
2
2
2
6
0
0
-2
-1
1
2
-2
Richmond
Eton Street
31
0
0
10
3
5
0
10
0
0
5
2
0
1
5
0
0
-5
-1
-5
1
-5
Total
2
21
161
59
47
18
184
2
24
113
45
34
16
139
0
3
-48
-14
-13
-2
-45
%
change
0%
14%
-30%
-24%
-28%
-11%
-24%
Total for
schemes in
Westminster
1
14
94
41
24
12
109
2
13
70
31
20
10
85
1
-1
-24
-10
-4
-2
-24
%
change
100%
-7%
-26%
-24%
-17%
-17%
-22%
Total for
schemes in
Kingston
1
6
50
12
17
6
57
0
10
33
10
12
3
43
-1
4
-17
-2
-5
-3
-14
%
change
-100%
67%
-34%
-17%
-29%
-50%
-25%
Total schemes
in Richmond
0
1
17
6
6
0
18
0
1
10
4
2
3
11
0
0
-7
-2
-4
3
-7
%
change
0%
0%
-41%
-33%
-66%
300%
-39%
Transport for London – January 2008
Page 20
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
8.0 Conclusions
8.1 Context
The evidence from this experiment offers more information about the impact of allowing P2W
access to bus lanes than any study to date. Nevertheless, it is vital to recognise the scope and
limitations of this test of a potential enhancement to road safety in a live experiment with
several significant other variables; one of the most important being changes in highway usage
by different modes during the investigation.
Changing the methodology for generating control data for the experiment during the second 18
month phase of the trial gave rise to problems. It rekindled initial questions about the
experiment design, and generated new concerns among some stakeholders that the clarity,
reliability and validity of results and conclusions might be reduced from optimum levels.
In addition to concerns about changes in control method, questions focused on the scientific
value of data collected from attitudinal surveys. Queries also arose regarding the collection of
speed and journey time data and vehicle usage of the actual bus lanes themselves. The largest
gap in the data sets concerned the vehicle usage data. The lack of these data resulted in the
inability to produce any modal migration assessment and consequently any casualty rate
analysis.
TfL has, where possible, addressed such questions and concerns with freshly focused action
and this final report results from that action. The suite of reports and these conclusions are
made in response to input from all concerned with the experiment, inside and outside TfL, and
to optimise the value of the study to all parties interested in improving road safety for all road
users.
8.2 Control method changes
It can be argued that the Tanner Test calculations, to extrapolate figures for control data on
casualties, offer a more statistically robust outcome than use of figures from control sites that
are of correspondingly small size to trial sites.
However, what the Tanner method cannot do is allow for any fluctuations in vehicle usage, and
therefore cannot account for the impact of migration on the results to be used in assessing the
impacts of the measure under test.
The original control data method does allow direct comparisons between adjacent routes for
trial and control, and it allows the influence of migration to be taken into account when
analysing the impact of the measure on casualty numbers and rates. However, the lack of any
vehicle usage data meant that this assessment could not be carried out.
It must be stated that because of the small size of all the data sets (even when the Tanner
control was applied), none of the results returned were statistically significant.
Transport for London – December 2007
Page 21
Directorate of Road Network Performance
P2Ws in bus lanes Main Report
8.3 Findings
8.3.1
Opinion surveys
The views collected from the opinion surveys were as expected.
Powered two wheeler riders responding to the survey favoured the measure as they felt it made
the network safer for them.
Pedestrian and cyclist respondents expressed more negative views, formed by expectations
that they would be less safe. There were indications that some cyclists did not like the idea of
sharing use of bus lanes with P2W riders regardless of any safety factors.
Bus driver respondents indicated a preference for sharing bus lane road space with P2Ws
rather than pedal cycles.
8.3.2
Cyclists
The evidence from casualty and collision data shows that cyclists’ concerns that their casualty
rates would significantly rise, were generally unfounded in practice.
Against the original control data, the pedal cycle casualties reduced during bus lane operating
hours, but increased marginally under the Tanner test assessment, providing a largely neutral
picture overall.
8.3.3
Pedestrians
The sum of casualty evidence shows that two of the four scenarios assessed returned a safety
benefit during operating hours, whilst in all other situations casualties increased.
When looking at the actual casualty numbers from the individual trial routes without considering
the migration issue, the experiences have been conflicting. In one trial site, pedestrian
casualties from P2W collisions halved (A23), and in the other they doubled (A41).
8.3.5
Powered Two Wheelers
Although there has been a reduction in collisions for P2W riders on the trial routes , it has not
been statistically significant.
8.5 Summary
The evidence shows increased collisions for pedestrians and pedal cyclists and a reduction for
powered two wheelers, however none of these results are statistically significant.
Cyclists and pedestrians had concerns about motorcyclists in bus lanes. Car drivers , who were
not bus passengers and motorcyclists did not share the same concerns.
Transport for London – December 2007
Page 22
