INTRODUCTION
ANNUAL SAFETY SAFETY OVERVIEW PERFORMANCE REPORT PROGRESS 2007 AGAINST THE SSP
BENCHMARKING
PASSENGERS
WORKFORCE
PUBLIC
TRAIN ACCIDENTS
LEVEL CROSSINGS
PERSONAL SECURITY
STATION SAFETY
LEARNING FROM ACCIDENTS
EUROPEAN DEVELOPMENTS
APPENDICES
Annual Safety Performance Report 2007
If you would like further information, wish to give feedback on any of the material contained in this report, or if you have any suggestions for future editions, please contact:
Marcus Dacre Senior Safety Intelligence Analyst Rail Safety and Standards Board Evergreen House 160 Euston Road London NW1 2DX 020 7983 6752 [email protected]
The report may be downloaded from the RSSB website: www.rssb.co.uk.
Additional printed copies may be ordered at cost price by contacting the RSSB enquiry desk on 020 7904 7518.
Version 1.0 (29 April 2008)
© Rail Safety and Standards Board 2008
Intentionally blank
Contents
Executive summary vii 1 Introduction 1 1.1 Purpose of the report 1 1.2 How does the ASPR assess safety performance? 1 1.2.1 Scope 1 1.2.2 Grouping risk 1 1.2.3 Fatalities, injuries and FWI 2 1.2.4 Data sources and methodology 3 1.2.5 Data quality 5 1.3 Report structure 6 2 Safety overview 9 2.1 Fatalities and injuries in 2007 9 2.2 Risk profile 11 2.2.1 Fatality risk profile 11 2.2.2 FWI risk profile 12 2.3 Safety diary 15 2.4 Longer-term trends 19 2.4.1 Rail usage 19 2.4.2 Rail safety 19 2.5 Looking to the future 21 2.5.1 Future influences on safety 21 2.5.2 Research and Development Programme 23 2.5.3 Sustainable Rail Programme 23 3 Review of progress against SSP trajectories 25 3.1 Introduction 26 3.2 Review of progress 28 3.2.1 Passengers at stations 28 3.2.2 Passengers on trains 30 3.2.3 Workforce – train crew 31 3.2.4 Workforce – track workers 32 3.2.5 Workforce at stations 34 3.2.6 Engineering – track 35 3.2.7 Engineering – trains 37 3.2.8 Public behaviour – crime 37 3.2.9 Public behaviour – level crossings 39 3.3 Future developments in the trajectory approach to safety planning 40 3.4 Further information 41 4 Benchmarking railway performance 43 4.1 Comparing the railway with other modes of transport 43 4.1.1 Relative safety of travel on different transport modes 43 4.1.2 Whole-journey risk 46
2007 Annual Safety Performance Report i Contents
4.1.3 Total risk associated with different transport modes 47 4.1.4 Inter-modal safety trends 49 4.2 International comparisons 50 4.3 Occupational risk: comparisons with other industries 52 4.3.1 Industry risk 53 4.3.2 Occupational risk 54 4.4 Benchmarking within the industry 55 5 Passenger safety 57 5.1 Passenger risk profile 58 5.2 Passenger fatalities and injuries in 2007 59 5.3 Passenger safety trends 60 5.3.1 Trends in passenger fatalities 61 5.3.2 Trends in passenger injuries 62 5.4 Passenger safety by location 64 5.4.1 Passengers in stations 64 5.4.2 Passengers on trains 64 5.5 Factors affecting passenger risk 66 5.5.1 Alcohol and drugs 67 5.5.2 Age 69 5.5.3 Gender 70 6 Workforce safety 71 6.1 Workforce risk profile 72 6.2 Workforce fatalities and injuries in 2007 74 6.3 Workforce safety trends 74 6.3.1 Trends in workforce fatalities 74 6.3.2 Trends in workforce injuries 77 6.4 Workforce safety by occupation 77 6.4.1 Track workers 77 6.4.2 Train drivers 80 6.4.3 Train guards and other train crew 83 6.4.4 Station staff 85 6.4.5 Shunters 87 6.5 Further information 88 7 Public safety 89 7.1 Risk profile for members of the public 90 7.2 Public fatalities and injuries in 2007 91 7.3 Public safety trends 92 7.4 Trespassers 93 7.4.1 Fatalities and injury trends 94 7.4.2 Near misses and other trespass reports 95 7.4.3 Factors affecting trespass 96
ii 2007 Annual Safety Performance Report Contents
7.5 Level crossing users 102 7.6 Other members of the public 102 7.6.1 Members of the public in stations 102 7.6.2 Falls onto the mainline railway 102 7.6.3 Road vehicle incursions 103 7.6.4 Railway neighbours 103 7.7 Suicide 103 8 Risk from train accidents 107 8.1 Trends in the numbers of train accidents 108 8.2 Train accident risk profile 108 8.3 Train accident fatalities and injuries in 2007 111 8.4 Trends in train accidents 112 8.4.1 Long-term trends in the number of fatal train accidents 113 8.4.2 Trends in the number of potentially higher-risk train accidents 114 8.4.3 Trends in the number of non-PHRTAs by type 122 8.5 Trends in precursors to train accidents 123 8.5.1 The Precursor Indicator Model 123 8.5.2 Trends in the PIM 124 Analysis of PIM subgroups in more detail 126 8.6 Further information 133 9 Risk at level crossings 135 9.1 Level crossing risk profile 136 9.2 Fatalities and injuries in 2007 138 9.2.1 Pedestrians 138 9.2.2 Road vehicle users 139 9.2.3 Suicides 140 9.3 Trends in injuries at level crossings 140 9.3.1 Level crossings FWI since 2003 140 9.3.2 Fatalities by user type since 1998 141 9.4 Trends in collisions and near misses at level crossings 142 9.4.1 Collisions with road vehicles 142 9.4.2 Near misses with road vehicles 143 9.4.3 Near misses with pedestrians 144 9.5 Factors affecting the risk at level crossings 144 9.5.1 User behaviour 144 9.5.2 Equipment failure 146 9.5.3 Trains striking gates or barriers 147 9.5.4 Workforce error 148 9.5.5 Railway crime 149 9.6 Further information 150 10 Personal security 151
2007 Annual Safety Performance Report iii Contents
10.1 Personal security in context 152 10.2 Fatalities and injuries in 2007 153 10.3 Trends in workforce personal security 153 10.3.1 Harm to the workforce from assaults 154 10.3.2 Workforce assaults by type 155 10.3.3 Workforce assaults by severity 156 10.3.4 Time lost as a result of workforce assaults 157 10.3.5 Further analysis of workforce assaults 158 10.4 Trends in passenger and public personal security 161 10.4.1 Personal security incidents 162 10.4.2 Personal security incidents by type 162 10.4.3 Further analysis of passenger and public assaults 163 10.5 Public disorder 164 10.6 Perceived security 166 11 Station safety 169 11.1 Station risk profile 170 11.2 Fatalities and injuries in stations during 2007 170 11.3 Station safety trends 171 11.3.1 Trends in fatalities at stations 172 11.3.2 Trends in major injuries at stations 173 11.4 Passenger safety in stations 174 11.4.1 Boarding and alighting 175 11.4.2 Other accidents at the platform edge 176 11.4.3 Moving around the station 177 11.4.4 Other hazards in stations 178 11.5 Workforce safety in stations 180 11.6 Public safety in stations 181 12 Learning from accidents 183 12.1 Industry processes for learning from accidents 183 12.2 Investigations completed during 2007 184 12.2.1 Lessons learnt during the year 184 12.3 Industry progress against recommendations from investigations 187 12.3.1 Analysis of the key themes from recommendations 190 12.3.2 Formal inquiry recommendation process 192 12.4 Confidential Incident Reporting and Analysis System 193 12.4.1 Who reports to CIRAS? 194 12.4.2 Why do people report to CIRAS? 194 12.4.3 Key issues of concern in CIRAS reports in 2007 195 12.4.4 Additional CIRAS research outputs 196 12.5 Developments in industry strategy 197 12.5.1 Learning from accidents workshop and strategy 197
iv 2007 Annual Safety Performance Report Contents
12.5.2 European developments in learning from accidents 198 12.6 Further information 199 13 European developments 201 13.1 ROGS and Railways (Interoperability) Regulations 202 13.2 Technical Specifications for Interoperability 202 13.3 Train driver licensing 203 13.4 Safety Management Systems and Safety Certification and Authorisation 204 13.5 Common safety targets, methods and indicators 205 13.5.1 Common safety targets 206 13.5.2 Common safety methods 209 13.5.3 Common safety indicators 211
Appendix 1. Key safety facts 215 Appendix 2. Fatalities in 2007 223 Appendix 3. Potentially higher-risk train accidents 225 Appendix 4. Ovenstone criteria adapted for the railways 227 Appendix 5. Level crossing types 228 Appendix 6. Definitions 232 Appendix 7. Glossary 238
2007 Annual Safety Performance Report v Contents
Intentionally blank
vi 2007 Annual Safety Performance Report Executive summary
Executive summary
Introduction This report reviews our industry’s safety performance over the calendar year and assesses progress in the key risk areas identified by the Strategic Safety Plan. It benchmarks rail safety in Britain against other transport modes and overseas railways, reviews the way that our industry learns from accidents, and discusses the European developments that are having an effect on the way safety is managed.
This is the third Annual Safety Performance Report that RSSB has produced in-house, and the first to be based on the revised weightings for shock, trauma and minor injuries in the calculation of fatalities and weighted injuries, which were introduced in April 2008.
Headlines • One passenger died in a derailment at Grayrigg – the first on-board fatality in a train accident since 2004. • Eight other passengers died in individual incidents – one as a result of assault, two jumped or fell from moving trains and five were struck by trains in stations. • Two members of the workforce died in accidents in the trackside environment – the same number as in 2006. • At 49, the number of potentially higher-risk train accidents (PHRTAs) was little changed from the previous year. • The risk from signals passed at danger (SPADs) at the end of 2007 stood at around 13% of its March 2001 level, representing a slight increase in risk during the year. • Nine pedestrians died at level crossings, an increase of four compared to 2006. • There were three accidental fatalities involving road vehicle occupants at level crossings, which is more than the zero recorded in 2006 but comparable with levels seen over the previous five years.
• Fatalities from suicide and trespass totalled 249. This represents a return to more typical levels after an unusually high number in 2006.
Train accidents One passenger was killed and 88 people were injured in a high-speed derailment at Grayrigg on 23 February 2007. The immediate cause of the accident was points failure.
Prior to Grayrigg, the last time a train accident led to on-board fatalities was November 2004. There has been a significant long-term reduction in train accidents with on-board fatalities and the current decade has seen an average of less than one such incident per year.
The types of train accident with the greatest potential to cause harm are known as ‘potentially higher-risk train accidents’, or PHRTAs. There were 49 PHRTAs in 2007, which is very similar to last year, and less than half the total seen 10 years ago.
Because serious train accidents are rare, RSSB also analyses trends in precursors and their associated risk. This analysis shows that the underlying risk from train accidents was relatively static over the year, consolidating the improvements seen earlier in the decade.
2007 Annual Safety Performance Report vii Executive summary
Passenger safety In addition to the fatality at Grayrigg, eight passengers died in individual incidents. Of these, one was assaulted, two jumped or fell from moving trains, and the remainder were struck by trains (one on a station crossing, two from standing too close to the platform edge, one after falling down the gap between the train and platform when alighting a train, and one after falling from the platform). The total of nine passenger fatalities is broadly in line with recent years.
There was a small rise in major injuries to passengers, from 219 to 233. The increase was largely brought about by the accident at Grayrigg. Most major injuries occur in stations, and the underlying trend is downwards, especially when normalised by passenger journeys.
Intoxication plays a role in many passenger injuries, particularly late at night. Other factors which influence risk from certain types of accident include age and gender.
Workforce safety Two members of the workforce were killed in 2007. Both were track workers struck by trains in separate incidents. This is the same number of workforce fatalities as 2006 (although there were no track worker fatalities that year), and less than the average of recent years.
A further two track workers received major injuries from being struck by trains. Overall there were 128 workforce major injuries, which is very similar to last year and a significant reduction from 2004 and 2005. More than half of the workforce major injuries affect track workers, and most of these result from slips, trips and falls or other personal accidents about the track. Track worker major injuries continued to reduce in 2007 and have halved since 2004.
There were 16 major injuries involving train drivers, including one at Grayrigg and two who were struck by objects thrown at their trains. Around 150 drivers need to take time off work with shock or trauma each year; the two major causes are striking persons on the line and near misses.
Shunting has been identified as a high-risk railway occupation, but most injuries are outside the scope of the ASPR because they occur in depots, yards and sidings. In August, a shunter had both legs amputated following an accident in which he was run over by a freight train at Harlow Mill.
Public safety Most risk occurs to members of the public, much of it being a consequence of their own behaviour. There were 43 trespasser fatalities during the year, which is more than last year but in line with the 10-year average. Approximately half of these occurred around stations. Young men make up a high proportion of trespasser fatalities, and peaks occur in the early and late evening.
Most of the other accidental fatalities to members of the public occurred at level crossings, although three people (who were not in the station in connection with train journeys they were making) were hit by trains after getting too close to or falling from the platform edge.
There were 206 suicides (or suspected suicides) over the year, fewer than in 2006 but comparable with the long-term average.
viii 2007 Annual Safety Performance Report Executive summary
Level crossings The harm arising from level crossing misuse increased in 2007, returning to pre-2006 levels.
Nine pedestrians died as a result of being struck by trains at level crossings. One of these was a passenger at a station crossing. This total is higher than 2006, but comparable with the five-year average.
There were 11 collisions between trains and road vehicles, which is unchanged from 2006 and lower than previous years. Four road vehicle users died as a result, one of whom is suspected of committing suicide. Of the three accidental fatalities, one person was killed at an automatic half-barrier crossing (which was the scene of another fatal collision in 2005) and two died in a collision at an automatic open level crossing.
There is also a downward trend in near misses with road vehicles at level crossings. Unusually, one of the collisions in 2007 was the result of signaller error; there were no reported injuries.
Personal security One passenger died after being assaulted in 2007. He was stabbed in the back in an incident at Dewsbury station.
Overall, there was a reduction in the number of assaults recorded both by British Transport Police and by the industry’s Safety Management Information System (SMIS). Harm from workforce assaults also fell for the third year running. Nevertheless, the personal security of rail staff remains an issue of concern. On a typical day, more than 10 members of the workforce are absent as a result of assault or abuse.
Reports of threats and abuse against staff rose again during the year, although it is possible that continued improvements in reporting are behind this trend. Ticket disputes are the most common contributory factor in staff assaults, and alcohol can exacerbate the situation.
Around two-thirds of passengers rate their personal security on trains and in stations as good, and passenger perceptions have shown slight improvements in this area over recent years.
Station safety Excluding trespassers, nine people died in incidents at stations. One was assaulted, the other eight were struck by trains (see the passenger and public sections for further details).
There were no fatalities from slips, trips and falls away from the platform edge, and there was a reduction in the number of major injuries from such accidents. Stairs and escalators continue to account for more than half of all the harm from this cause.
The substantial reduction in boarding and alighting injuries in 2006 has not been maintained. That year now looks like a ‘statistical outlier’ and the number of major injuries in 2007 returned to the level seen in 2005.
The number of trespasser fatalities in stations rose to 24 in 2007, which is the highest number for at least a decade.
2007 Annual Safety Performance Report ix Executive summary
Summary of performance Although 2007 saw a number of serious accidents, the overall story was one of consolidating the progress achieved in recent years, with continued improvements in some areas. Previous Strategic Safety Plans had the principle of ‘maintaining safety in a period of change’ as their overarching objective, and the industry continues to do this against a backdrop of increasing passenger numbers and growth in freight.
The 2007–09 Strategic Safety Plan took a new approach to safety planning by basing itself on companies’ own safety plans. It set out nine trajectories that reflect both the industry’s ambitions in Key Risk Areas and the actions the industry is undertaking to achieve them.
The table below presents the trajectories in each of the Key Risk Areas, together with an assessment of whether, after one year of the three-year plan, progress towards achieving them is apparent from safety data. Progress towards the trajectory is indicated by a tick, and failure to make progress is marked with a cross. Where there is insufficient data to ascertain whether progress is being made, a dash has been placed in the final column.
Key Risk Area Trajectory Progress Passengers at stations Deliver a measurable reduction in the level of risk caused by the behaviour of passengers in stations by reducing or eliminating the 9 typical hazards that result in accidents . Passengers on trains Given the improvements in this Key Risk Area, the aim is to maintain the current level of safety, improve it where reasonably practicable and tak e opportunities to reduce risk when there are train - refurbishments projects and new rolling stock procured. Workforce - train crew Given the improvements in this Key Risk Area, the aim is to maintain the current level of safety, and improve it where reasonably 9 practicable, over the next three years. Workforce - track workers Deliver a measurable reduction in the rate of track work er accidents. 9
Workforce - at stations Deliver a measurable reduction in the risk of accidents to passengers in stations through an improvement in the competency of - station staff. Engineering - track Given the improvements in this Key Risk Area, the aim is to maintain the current level of safety, and improve it where reasonably 9 practicable, over the next three years. Engineering - trains Deliver a measurable reduction in the risk from catastrophic accidents and accidents to passengers from vehicle defects. 9
Public behaviour - crime Deliver a measurable reduction in the risk from crime to passengers and the work force. 9
Public behaviour - level crossings Deliver a measurable reduction in the risk from the misuse of level crossings. -
Overall, there is evidence that the industry is delivering on the trajectories in the strategic safety plan. In six of the nine Areas, tangible progress has been made towards delivering the trajectory. In the remaining three Key Risk Areas, there is insufficient data to yet comment on whether there is any trend. The areas that show the greatest improvements are Workforce – track workers, Engineering – trains and Engineering – track.
x 2007 Annual Safety Performance Report Introduction
1 Introduction
Over 40 years ago, the noted transport historian, L.T.C. Rolt, wrote: ‘A railway accident is always news. Governments may fall or battles be fought and lost but no matter what the crisis of the hour may be, a derailment or a collision involving a relatively small loss of life will be sure to command a banner headline in the popular Press’.1
Intensive media coverage still has the potential to influence the public’s perception of risk on the railway, despite the major improvements that have occurred since the above lines were written. The Annual Safety Performance Report (ASPR) gives a full and detailed account of recent safety performance, so that people both inside and outside the industry can base their views and decisions on the true state of safety on today’s railway.
1.1 Purpose of the report This Rail Safety and Standards Board (RSSB) ASPR presents our railway’s safety trends for the calendar year 2007. It reviews the performance levels achieved during that time across a number of topic areas and considers how key safety issues are being addressed by the industry. The areas covered are those identified in the 2007–09 Strategic Safety Plan (SSP).
The 2007–09 SSP is a departure from its predecessors in that it has been developed by bringing together companies’ own individual safety plans; a link has thus been created between the SSP and the duty holder planning process. This is explored in Chapter 3.
The ASPR’s primary purpose is to provide safety intelligence and risk information to RSSB members. However, it is also intended to inform rail employees, passengers, the government (and its agencies), and the public at large.
1.2 How does the ASPR assess safety performance? The report includes comprehensive statistical analyses on a wide range of safety performance indicators: many concern the actual safety performance level that has been achieved; others provide a measure of the underlying risk.
1.2.1 Scope The analyses herein mainly relate to risk from incidents on or affecting stations and Network Rail managed infrastructure (NRMI). However, they also cover other fatal accidents occurring to railway employees away from these locations, but during working hours.
1.2.2 Grouping risk There are different ways of analysing accident statistics, and the ASPR looks at safety performance from several angles.
1 L. T. C. Rolt, Red for Danger, 3rd edn (Stroud: Sutton Publishing, 1998), p. 15. This well-known and well-respected work is still a standard text for many railway employees.
2007 Annual Safety Performance Report 1 Introduction
The SSP is framed around Key Risk Areas, which reflect the various sources of risk and the controls in place to address them. For example, one area relates to the risk managed by station staff, which covers not only personal accidents to the workers themselves, but also accidents to passengers, such as slips on wet surfaces and errors related to train despatch. In Chapter 3, which assesses safety performance against the SSP, accident data is categorised to reflect the risk from each of the Key Risk Areas.
Most other chapters look at the risk to different groups. The three main groups on which the report focuses are passengers, the workforce and members of the public. The term ‘passenger’ refers not just to those on trains, but to anyone who is on railway property in connection with a train journey that they are making (or intend to make).
Each injury is also categorised by the hazardous event that caused it and the major precursor to that event. The ASPR uses the same set of hazardous events and precursors as RSSB’s Safety Risk Model. There are 125 hazardous events, ranging from slips, trips and falls to collisions between trains, and more than 2,000 precursors. To prevent the charts in the ASPR becoming too complex, similar hazardous events are often grouped together.
1.2.3 Fatalities, injuries and FWI The ASPR looks at the risk in terms of fatalities, injuries and – in a departure from previous editions – shock and trauma. Injuries are categorised according to their seriousness. While many charts focus solely on fatality risk, or major injury risk, others look at the total harm. Fatalities, injuries and shock and trauma are combined into a single figure, termed fatalities and weighted injuries (FWI). In previous ASPRs, 10 major injuries and 200 minor injuries were deemed to be ‘statistically equivalent’ to one fatality.
An important change in this year’s ASPR is that all FWI figures are based on a new set of weightings, which came into effect on 1 April 2008. Use of the new weightings was endorsed by the RSSB Board following a major research project and consultation with the industry, the Office of Rail Regulation (ORR) and the Department for Transport.2 It is the first time that the weights have changed since the use of fatalities and weighted injuries as a measure of railway safety gained widespread acceptance in the late 1980s.
Table 1 shows the different injury classifications and their associated weightings, as used in this report. The figures in the Weight column represent the number of injuries of each type that are ‘statistically equivalent’ to one fatality. For example, if an accident resulted in one fatality and three major injuries, the total FWI arising would be 1.3.
As well as formalising the basis on which shock and trauma is taken into account, the biggest change is that the least severe minor injuries (those that do not require hospital treatment) are given less weight than previously. The research suggested that the old weighting overstated the value that society places on such events in relation to more serious injuries. Because a large number of these less serious minor injuries are reported each year, the change in weightings has a significant effect on the risk profile. A comparison of safety performance trends based on both the new and old weightings is presented in Chapter 2.
2 The full report of the research project on The weighting of non-fatal injuries (T440) is available from the RSSB website: http://www.rssb.co.uk/pdf/reports/research/T440_rpt_final.pdf.
2 2007 Annual Safety Performance Report Introduction
Table 1. Injury degrees and weightings
Injury degree Definition Weight
Fatality Death occurs within one year of the accident. 1
As defined in RIDDOR3 1995 – includes losing Major injury consciousness, most fractures, major dislocations and 10 hospital stays of 24 hours or more.
Physical injuries that are not major, but which result in more RIDDOR-reportable than three days’ absence from work (for members of the 200 minor injury workforce) or require hospital treatment (for passengers and members of the public).
Non RIDDOR-reportable All other physical injuries. 1000 minor injury
Caused by witnessing a fatality or being involved in a Class 1 shock / trauma 200 collision, derailment or train fire.
Other causes, such as verbal abuse, near misses and Class 2 shock / trauma 1000 witnessing non-fatal assaults.
The RSSB Board has recommended that currently approved plans are not amended, but that the new weightings are used with immediate effect for the assessment of future investments and safety improvement schemes.
NB: All historical data in the 2007 ASPR is now reported using the new weightings.
1.2.4 Data sources and methodology Most analyses in the ASPR are based on data from the industry’s Safety Management Information System (SMIS). However, SMIS data is supplemented – where appropriate – with data from other sources, such as British Transport Police (BTP), the ORR and Network Rail. Where a chart or table has been derived from a source other than SMIS, that source is stated.
Risk profile
The ability to manage risk at a level that is as low as reasonably practicable depends on a good understanding of the types of hazardous events that can occur, and the scale and distribution of the residual risk they represent. To support transport operators’ efforts in this area, RSSB has developed the Safety Risk Model (SRM). This is a comprehensive mathematical representation of 125 hazardous events that could lead directly to injury or fatality during the operation and maintenance of the mainline railway. The causes and consequences of each event are modelled in detail, considering the railway as a whole, rather than concentrating on a particular route or railway undertaking. This provides the
3 RIDDOR refers to the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995, a set of health and safety regulations that mandates the reporting of work-related accidents etc.
2007 Annual Safety Performance Report 3 Introduction
context for each company’s management of safety, acting as a guide to the overall risk on the network.
The SRM has been designed to take full account of high-frequency, low-consequence events (such as slips, trips and falls) and low-frequency, high-consequence events (such as fatal train accidents). It was developed using established modelling techniques, together with available information, informed by expert judgement. The use of expert judgement is important, since derivation of risk estimates on data alone could lead to inaccuracy, especially where data is sparse. The majority of the data used to populate the SRM comes from SMIS.
Version 5 of the SRM (SRMv5) formed the basis for analysing the risk profile in the 2006 ASPR, but this was based on the old injury weightings; as discussed in the Introduction, the industry is applying revised weightings from 1 April 2008. The first SRM to take account of the revised injury weightings is Version 5.5 (SRMv5.5), which will be published in May 2008. SRMv5.5 represents an interim, rather than a full, update of the SRM; risk estimates have been based on the new weightings, but only a selection of hazardous events (those related to areas showing the greatest change in risk levels) have been updated. A full revision of the SRM (Version 6) will be produced in time to assess the safety metrics for the High Level Output Specification (HLOS) in March 2009.
Charts based on the SRM are used within the ASPR to set the context for a particular chapter or topic.
Safety trends
Analysis of trends in incident data is provided separately for each topic – usually over five years (for injuries) or ten years (for fatalities and train accidents). When considering such data, it is important to differentiate between changes in underlying safety and the chance fluctuations that can occur from one year to the next. For example, annual numbers of passenger fatalities vary greatly depending on the occurrence (or not) of low-frequency, high- consequence events, such as train accidents. However, a year without a train accident does not necessarily indicate an improvement in passenger safety, just as the occurrence of a single accident does not necessarily imply a deterioration.
To address this, longer-term trends can be assessed using moving averages, for example over five or ten years. Further understanding of changes in the underlying system risk can also be gained by looking at trends in accident precursors or ‘near misses’.
Significance testing Statistical significance testing can help to explain whether a genuine change has occurred or whether the data could be the result of chance fluctuations. Throughout this report, the term significant (represented by 33 in the sections on ‘performance at a glance’) refers to a change that is significant at the 5% significance level4 – that is, we can be reasonably confident that there has been a real improvement or deterioration. An increase or decrease that is significant at the 20% (but not the 5%) significance level – meaning that there is some
4 A change is significant at the 5% significance level if the probability of it occurring by chance alone is less than or equal to 5%. A change is significant at the 20% significance level if the probability of it occurring by chance alone is less than or equal to 20%.
4 2007 Annual Safety Performance Report Introduction
evidence of a genuine change, but it is not particularly strong – is represented by 3. Such changes are sometimes referred to as marginal. Differences that are not significant at the 20% significance level are deemed to be not significant and represented by 8 in the ‘performance at a glance’ sections.
1.2.5 Data quality The value of any safety performance report depends to a large degree on the quality of the data on which it is based. The number of events reported into SMIS each year is large (about 100,000) and there are known data quality issues with some event types, such as level crossing misuse or irregular working.
Poor data quality can be due to a number of factors, including under-reporting or poor supply of information. These issues are more associated with high-frequency, low-consequence events than major incidents. However, because precursors can be valuable in the study of underlying trends in risk, it is important that data quality for all types of event is as good as possible.
RSSB’s Safety Intelligence Team is currently leading a data quality project, backed by the SMIS Programme Board and ATOC Operations Council. This project has various strands, including the following:
• ‘Health check’ visits to each organisation that enters data into SMIS. • Data quality indicators for each company, highlighting areas of concern. • Data quality indices for RSSB outputs, to assess the reliability and robustness of safety measures. • Returning cleaned-up data to SMIS for use across the industry. • Rewriting and reissuing Railway Group Safety Performance Monitoring – Definitions and Guidance. • Improving the recording of train fires in SMIS.
RSSB also employs a number of initiatives on a day-to-day basis, to ensure that the existing data is of the highest possible standard. These include, where possible, validating the information in SMIS and categorising the injuries with the relevant SRM hazardous events and precursors. This information is fed back to the SMIS event owners via specific fields in SMIS (in the case of injuries) and by telephone conversations.
In addition, RSSB uses information from other sources to try to gain as much knowledge of an event as possible. In the case of fatalities, for example, the information in SMIS is supplemented with BTP reports and coroners’ reports, where these are available. In order to generate timely statistics that are as accurate as possible, if a coroner has yet to return a verdict or has returned an open verdict, the industry applies the Ovenstone criteria to determine on the balance of probability whether a fatality was the result of an accident or suicide. During the past year, RSSB and Network Rail have been working together to resolve differences between the two organisations’ trespass and suicide statistics. This has led to a major downward revision of accidental trespass fatalities in 2006 from the figures published in last year’s ASPR (and a corresponding increase in suicides). New information- sharing arrangements have been put in place to ensure that future statistics will be robust,
2007 Annual Safety Performance Report 5 Introduction
but there are likely to be continued minor changes to the numbers of suicides and trespass fatalities because classifications are reviewed when a coroner’s verdict or new evidence comes to light. See Appendix 4 for more information on this topic.
Another cause of changes to previously published numbers is late reporting into SMIS. Chart 1 shows, as a proportion of the injuries that were eventually reported from 2006, the volume of events that SMIS contained at the end of that year and at subsequent month ends.
Much of the increase during January 2007 was the result of accidents from late December being entered into SMIS. Some lag is inevitable; the Railway Group Standard (GE/RT8047) specifies that injuries should be entered into SMIS within 10 working days. By the end of February, SMIS contained over 99% of the injuries that were eventually recorded for the previous year. For passengers, this figure was slightly lower, at 98.8%. The chart shows that some passenger injuries came to light more than 12 months after the end of the year. Such delays usually occur when the industry is only made aware of an event a long time after it occurred, for example if the injured party makes a claim for compensation.
Chart 1. Lag in the reporting of injuries into SMIS
100%
99% Public Passenger 98% Workforce All 97%
96% Reports entered into SMIS 95% Jul 2007 Oct 2007 Oct Jun 2007 Jun Jan 2007 Jan 2007 Apr Nov 2007 Mar 2007 Mar Aug 2007 Feb 2007 Sep 2007 Dec 2007 Dec 2006 2007 May
Taking all these factors into account, RSSB bases the analyses in the ASPR on the latest and most accurate information available at the time of production. We also continually update and revise previous years’ data in the light of any new information received. The data cut-off dates for the 2007 ASPR were 29 February 2008 for SMIS data and 25 January 2008 for data from BTP.
1.3 Report structure Our aim has been to produce a document that all readers find engaging, although certain sections will obviously be of greater relevance to some readers than others.
As in previous ASPRs, the Safety overview immediately follows the Introduction. This sets the context by presenting the overall risk profile and performance trends, reviewing the year’s main events, and looking at long-term changes in railway usage and safety.
A new chapter then follows – Review of progress against SSP trajectories. Included in response to stakeholder feedback, this chapter assesses industry progress against the trajectories set out in the 2007-09 SSP. It also summarises the actions that companies are taking to deliver the required levels of safety.
6 2007 Annual Safety Performance Report Introduction
Benchmarking comes next, and we take a step back to compare British railways with other modes of transport, railways in other countries, and other employers. The chapter also discusses the steps being taken to enable companies to benchmark their own performance against the rest of the industry.
The risk to passengers, the workforce and members of the public is dealt with separately (in Chapters 5–7 inclusive). However, there will be a degree of overlap with the specialised sections on train accidents, level crossings, personal security and station safety, each of which is covered in a separate chapter.5
The Learning from accidents chapter (which replaces RSSB’s Annual Recommendations Report) tracks progress made against recommendations from old Formal Inquiries and more recent reports from the Rail Accident Investigation Branch (RAIB). It reflects on some of the lessons learnt in 2007, and discusses how RSSB is planning to meet industry needs in this area.
The report closes with a review of European developments and their implications for the management of rail safety in Great Britain.
Various appendices have been incorporated, including Key safety facts sheets, which provide summary tables of key safety statistics. A definitions list and glossary have also been prepared to assist the reader. These may be found at the back of the document.
We would like to thank everyone who returned the feedback forms supplied with the previous edition of the ASPR, and who gave feedback prior to our starting work on this year’s report.
The ASPR is a document of considerable length but, even so, it has obviously not been possible to include all our data. Therefore, if you are unable to find the answers to your safety performance questions here (or in our other publications), please do not hesitate to get in touch with us – we will be happy to be of assistance. Contact details are provided on the title page.
5 More information on some of these issues may be found on our website – www.rssb.co.uk – which contains freely downloadable versions of safety performance reports on various topics.
2007 Annual Safety Performance Report 7 Introduction
Intentionally blank
8 2007 Annual Safety Performance Report Safety overview
2 Safety overview
The derailment at Grayrigg in February 2007 was the first train accident to result in an on- board fatality for more than two years – the longest such period since rail became an established means of transport. Overall, safety trends for passengers and the workforce continue to improve. Safety trends for members of the public remain more variable, with their exposure to risk being greatly influenced by their own behaviour, rather than being under the direct control of the industry.
2.1 Fatalities and injuries in 2007 Sixty-eight people died on the railway in 2007, excluding suicides and suspected suicides. Nine were passengers, two were members of the workforce, and the remaining 57 were members of the public (43 of whom were engaged in acts of trespass). A further 206 people died as a result of suicide or suspected suicide. (In total, there were 274 fatalities.)
Chart 2 shows how 2007’s accidental fatalities and weighted injuries (excluding those due to suicide or suspected suicide) compare with each year since 2003, for passengers, workforce and the public.
Chart 2. Accidental fatalities and weighted injuries, 2003–2007
100 Shock and trauma 90 Weighted minor injuries 80 Weighted major injuries 63.8
Fatalities 61.9 70 59.7 51.0
60 48.8 47.2 46.8 43.9 42.9 42.5
50 39.5 35.4 32.0 40 26.7 25.4
30 59 54 57 45
Fatalities and weighted injuries and weighted Fatalities 20 42
10 10 13 889 9 5 0 3 22 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
Passenger Workforce Public
Trends in the total number of accidental public fatalities should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See section 7.3 for more information.
The Introduction described the recent revision to the weightings applied to minor injuries and shock/trauma, which came into effect from April 2008. RSSB publications from this date on, will, wherever possible and appropriate, use the new weighting system; this includes the current ASPR. However, it is worth looking at the effect that adopting the new weightings has had, in terms of a both absolute and relative change.
2007 Annual Safety Performance Report 9 Safety overview
Chart 3 presents the difference in passenger and workforce FWI levels over the past five years, using the new and old weighting values. Levels of harm to members of the public are not shown. Public risk is composed mainly of fatality data; changes to the weightings therefore make minimal difference. The chart shows that, on average, the new weightings result in a level of passenger harm that is 27% lower, and an estimate of workforce harm that is 46% lower. The relative changes in risk over time are, however, very similar. When public risk is taken into account, and averaged over the last five years, moving to the new injury weightings results in a level of overall harm that is 25% lower than that based on the old weightings.
Chart 3. Difference in passenger and workforce risk based on new and old injury weightings, 2003–2007
100 New weightings Old weightings 90
80 73.6
70 64.1 66.0 64.1 61.6 58.3 59.0 60 55.8 54.3 51.2 Average difference of 27% 50 Average difference of 46% 47.2 48.8 40 42.9 42.5 43.9 39.5 30 35.4 32.0 20 26.7 25.4
Fatalities and weighted injuries per year per injuries and weighted Fatalities 10
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Passenger Workforce
Table 2 presents information on major injuries and fatalities to members of the public due to suicide, suspected suicide and attempted suicide, over the same period.
Table 2. Fatalities and major injuries due to suicide, suspected suicide and attempted suicide, 2003–2007
2003 2004 2005 2006 2007 Fatalities 186 197 202 254 206 Major injuries 30 30 23 42 26
10 2007 Annual Safety Performance Report Safety overview
2.2 Risk profile This section presents the risk profiles for passengers, the workforce and members of the public arising from the risk area groups identified in the 2007-09 SSP. Section 3.1 contains further information on the make-up of the five risk area groups, and their relationship with the nine Key Risk Areas.
2.2.1 Fatality risk profile Chart 4 and Table 3 show the fatality risk profile. The inner ring of the chart shows the breakdown of the risk occurring to each of the person types. The outer ring shows the breakdown of where the risk arises from. Fatalities due to actual and suspected suicide are not included in the chart, although any shock/trauma suffered by the workforce in connection with these events is incorporated. To give a complete picture of risk on the railway, the information also includes the estimated risk from assaults.6
Based on SRMv5.5, it is estimated that the risk from all accidents on the railway equates to 70.9 fatalities per year. More than three-quarters occurs to members of the public, as a result of their own behaviour. Seven per cent of the total fatality risk occurs to workforce, and again, the majority of this is identified as being within the responsibility of the workforce. Sixteen per cent occurs to passengers, and whereas some of this is caused by passenger behaviour, more than half is from sources outside their control.
Chart 4. Fatality risk profile, excluding suicide (70.9 FWI)
Risk to : (inner circle) Passengers 15.8% Workforce 7.3% 7.3% 15.8% Public 76.8%
Risk from : (outer circle) Passenger behaviour Public behaviour Workforce responsibilities Engineering 76.8% The environment
Source: SRMv5.5 – total fatality risk of 70.9 FWI, excluding suicide and suspected suicide.
6 Whilst physical assaults on members of workforce are believed to be well recorded in SMIS, physical assault on passengers and members of the public are known to be significantly under-reported. BTP figures indicate that the number of passenger and public assaults per year is much higher than that recorded in SMIS (in the region of around 9 FWI per year). The risk from assaults on passengers and members of the public are not included in the overall risk predictions in the Risk Profile Bulletin (RPB) because of the uncertainty in the estimates. The risk profiles presented in the ASPR and the RPB therefore differ. Research has begun to resolve this issue prior to the publication of version 6 of the SRM.
2007 Annual Safety Performance Report 11 Safety overview
Table 3. Fatality risk profile, excluding suicide (70.9 FWI)
Proportion of Risk to: Risk from: Fatalities total Passengers 4.29 6.0% Public behaviour 2.83 4.0% Passengers Workforce 2.21 3.1% Engineering 1.75 2.5% Environment 0.15 0.2% Workforce 4.65 6.6% Engineering 0.34 0.5% Workforce Public behaviour 0.20 0.3% Environment 0.02 0.0% Passengers 0.00 0.0% Public behaviour 53.77 75.8% Engineering 0.34 0.5% Public Workforce 0.33 0.5% Environment 0.06 0.1% Passengers 0.00 0.0% Grand Total 70.93 100.0%
2.2.2 FWI risk profile Chart 5 presents a similar risk profile for passengers, the workforce and members of the public, this time based on both fatalities and weighted injuries. Neither fatalities nor injuries due to actual/attempted suicide and suspected suicide are included in the chart, although, as highlighted in the previous section, any shock/trauma suffered by the workforce in connection with these events is incorporated. Again, to give a complete picture of risk on the railway, the information includes the estimated risk from assaults.
The total system risk is estimated by SRMv5.5 to be 146.3 FWI per year, 36% of which occurs to passengers, 23% to the workforce, and 41% to members of the public.
Chart 5 shows that, the largest proportion of the risk to passengers is classified by the SSP as arising from passenger behaviour. This covers much of the risk from events such as passenger slips, trips and falls. However, a further notable proportion is under the responsibility of the workforce. This includes of the risk from some train accidents (such as a collision following a SPAD caused by driver error), but also covers some slips, trips and falls, since the SSP assumes these to be partly due to workforce station management issues. Most of the risk to passengers arising from the public is due to assaults,7 with a much smaller part arising from road vehicle drivers at level crossings. However, it should be noted that the risk estimate for assaults on passengers has a fairly high level of uncertainty associated with it, as the data on which it is based is not as robust as for most other parts of the profile. A small part of passenger risk arises from the SSP area of engineering (eg track defects); and an even smaller proportion arises from events related to the environment.
7 The 2007–09 SSP assigns all assaults to the ‘public’ risk area, even if the offender was a passenger.
12 2007 Annual Safety Performance Report Safety overview
Most of the risk to the workforce is classified by the SSP as arising from the workforce itself. Much of it involves events such as slips, trips and falls, and individual accidents at the trackside. Around 3.5% is due to assaults from the public,8 with a lesser proportion arising from engineering causes.
The risk to members of the public is almost entirely the result of their own actions, with only a very small proportion due to other causes.
Chart 5. Risk profile based on new FWI weighting system (146.3 FWI)
Risk to : (inner circle) Passengers 36.1% 22.6% Workforce 22.6% 41.3% 36.1% Public
Risk from : (outer circle) Engineering The environment 41.3% Passenger behaviour Public behaviour Workforce responsibilities
Source: SRMv5.5 – total risk of 146.3 FWI, excluding suicide, suspected suicide and attempted suicide.
For comparison, Chart 6 presents the risk profiled based on the old FWI weightings. The most noticeable shift in proportions is between workforce and members of the public. This is because public risk is dominated by fatalities, and is therefore only minimally affected by the weighting change, whereas workforce risk is comprised of notable numbers of non-reportable minor injuries and Class 2 shock/trauma events. Under the new weighting system these incidents receive one-fifth of their previous weighting.
The change in risk between the old weightings and new weightings is 57.1 FWI, which represents a decrease of 28%. This is close to the reduction of 25% observed from actual performance data over the past five years, quoted above in relation to Chart 3.
Table 4 presents the difference in FWI numbers for the old and new weightings, for comparison.
8 Ibid.
2007 Annual Safety Performance Report 13 Safety overview
Chart 6. Risk profile based on old FWI weighting system (203.4 FWI)
Risk to : (inner circle) Passengers 40.7% Workforce 29.1% 29.1% Public 30.2% 40.7% Risk from : (outer circle) Engineering The environment 30.2% Passenger behaviour Public behaviour Workforce responsibilities
Source: SRMv5.5 – total risk of 203.4 FWI, excluding suicide, suspected suicide and attempted suicide.
Table 4. FWI risk profile: a comparison of new and old weightings
Old injury weightings New injury weightings Proportion of Proportion of Risk to: Risk from: FWI FWI total total Engineering 3.91 1.9% 3.04 2.1% Environment 0.41 0.2% 0.39 0.3% Passenger Passengers 30.57 15.0% 22.95 15.7% Public behaviour 25.77 12.7% 10.49 7.2% Workforce 22.05 10.8% 15.99 10.9% Engineering 0.68 0.3% 0.63 0.4% Environment 0.07 0.0% 0.07 0.0% Public Passengers 0.00 0.0% 0.00 0.0% Public behaviour 60.33 29.7% 59.28 40.5% Workforce 0.37 0.2% 0.37 0.3% Engineering 3.97 2.0% 1.99 1.4% Environment 0.05 0.0% 0.05 0.0% Workforce Passengers 0.06 0.0% 0.04 0.0% Public behaviour 12.69 6.2% 5.16 3.5% Workforce 42.48 20.9% 25.86 17.7% Grand Total 203.4 100.0% 146.3 100.0%
14 2007 Annual Safety Performance Report Safety overview
2.3 Safety diary This section lists some of the more high-profile safety-related events of 2007.
January
Rail saboteur in court A former railway worker appeared in court, charged with causing £4 million worth of damage to rail equipment during a 17-month campaign of sabotage and arson in the West Midlands. He was traced by police after dropping a supermarket receipt at a crime scene.
Landslide results in derailment A landslide derailed one carriage of a passenger train near Merstham on 13 January (South East). The carriage remained upright and there were no reports of serious injuries, although services were disrupted as a result.
Elsenham inquest closes The inquest into the deaths of the two teenagers killed at Elsenham station level crossing in December 2005 closed. It concluded that both deaths had been accidental.
Level crossing misuse leads to fatality On 24 January, a passenger was struck on Wokingham station manually controlled barrier level crossing (South East). The young man had climbed over the correctly functioning barriers to join a train in the Up platform and was struck by a Down train.
Passenger platform-edge fatality A 15-year-old boy was struck by a passenger service at Treorchy station, South Wales, on 25 January (Western). He had been leaning over the platform edge at the time of the incident. He was taken to hospital but later died of his injuries.
February
Level crossing incident at Delny On 2 February, a road vehicle struck the side of a passenger train at Delny automatic open level crossing (locally monitored), near Barbaraville (Scotland). Two young men in the car died; the vehicle driver received major injuries.
Passenger fatality while alighting train On 13 February, a passenger died after falling between the train and the platform whilst alighting at Haddenham & Thame Parkway station (London North West).
Train derails at Grayrigg, Cumbria On 23 February, a passenger was killed when a passenger train derailed at Grayrigg in Cumbria (London North West). Eighty-eight people (including the driver) were injured. The immediate cause of the incident, which is subject to an RAIB investigation, was deemed to be the stretcher bar arrangement at the points, which resulted in one of the switch blades being able to move to an unsafe position.
2007 Annual Safety Performance Report 15 Safety overview
Fatal incident at Little Bowden On 23 February, a man was found with head injuries at the trackside near Little Bowden (London North East). The circumstances behind the fall have yet to be determined, but appeared consistent with a passenger falling or jumping from a train.
Fatal level crossing crash On 23 February, an empty coaching stock (ECS) formation collided with a van on Gailes automatic half-barrier level crossing (Scotland), killing the driver who was standing near the road vehicle. The crossing was working correctly at the time of the incident. The event is being treated as a suspected suicide.
March
Fallen tree results in derailment On 6 March, a passenger service derailed on the Up line near Godstone, having struck a fallen tree near Crowhurst Junction (South East). None of the eight passengers on board were injured, although there was major disruption to subsequent services. The tree had been uprooted by severe gales.
April
Track worker killed near Twyford On 29 April, a track worker was struck and killed by an ECS formation at Ruscombe Junction, near Twyford, Berkshire (Western). It was reported that he failed to respond to warnings to move clear of the train. RAIB launched an investigation into the accident, which was published in February 2008.
May
Motorist using satnav drives into path of train On 10 May, a motorist leapt clear of her car after apparently following satellite-navigation system directions that led her into the path of a passenger service at Ffynnongain user- worked level crossing (Western). The train struck the vehicle at over 60mph, shunting it half a mile down the track. BTP has launched an investigation into the accident.
June
Elsenham footbridge approved Network Rail received clearance to erect a pedestrian footbridge at Elsenham station, where two teenagers were struck by a train in December 2005.
TPWS PSR exemption granted Following approval from the Office of Rail Regulation for an exemption to the 1999 Railway Safety Regulations allowing for the removal of TPWS fitments at permanent speed restrictions (PSRs), Network Rail is currently identifying and verifying those that can be considered for TPWS removal.
16 2007 Annual Safety Performance Report Safety overview
Bridge derailment leads to long-term disruption Services in Suffolk and Cambridgeshire were severely disrupted after a ballast train derailed near Ely (Anglia) on 22 June. Eleven wagons fell onto their sides on a bridge over the River Ouse. The bridge sustained significant damage, but there were no reported injuries. RAIB has launched an investigation into the incident.
August
Rail worker struck by train in Leatherhead A passenger service struck a member of a Permanent Way team near Leatherhead station (South East) on 29 August. He sustained severe head, arm and leg injuries and was taken to hospital by helicopter.
September
Vehicle plunges into train's path On 17 September, a car, which had driven through a boundary fence in Muirhead, was struck by a passenger service (Scotland). The train was not derailed by the impact and there were no reported passenger injuries. However, the driver sustained minor facial injuries and was later taken to hospital for treatment. The sole occupant of the vehicle was able to escape after coming to rest on the line; he called the emergency services from safety at the top of the cutting. An investigation has been launched into the cause of the incident.
October
Vandalism causes train accident in East Renfrewshire A train carrying nine passengers struck a tree trunk which had apparently been deliberately left on the line at Busby in East Renfrewshire (Scotland) on 22 October. The unit sustained minor damage but there were no reported injuries. A driver also reported seeing several pieces of concrete near the same location earlier in the day.
November
Ufton inquest returns suicide verdict The inquest into the Ufton Nervet accident of 2004 concluded that the motorist who parked his car on the level crossing did so to commit suicide. The deaths of those on the train were found to have been caused by unlawful killing, as a result of the presence of the car on the interface.
Passenger struck by train On 7 November, a passenger was struck and fatally injured by a train approaching the platform at Cambuslang station (Scotland) when attempting to flag it down. The passenger was reported to be under the influence of alcohol at the time of the incident.
Super-Railer jumps rails in Cornwall On 12 November, a Road-Railer and trailer derailed and fell 20 feet down an embankment at St. Ives onto a service road below (Western). There were no reported major injuries, but the operator was taken to hospital, having fallen with the machine. Initial investigations revealed
2007 Annual Safety Performance Report 17 Safety overview
that the RRV was standing on a viaduct, lifting bullhead rails; as it swung around, it toppled and fell.
Staff charged over Hednesford Three railway workers face court action following an incident in September 2004 at Hednesford, which resulted in the deaths of two track workers. The two men were relaying track sections when they were hit by an RRV. Three of the work group have been charged with endangering the public by a wilful omission, neglect or improper use of a rail vehicle. Two of the three have also been accused of failing to ensure the health and safety of themselves and their colleagues while at work. Their employer at the time, V G Clements, is charged with failing to ensure the safety of its workers and the public. Network Rail and Carillion are accused of health and safety breaches.
Passenger fatality at Glengarnock On 21 November, a passenger apparently under the influence of alcohol collapsed and fell from the platform at Glengarnock station (Scotland). As he tried to climb back up, he was struck by a passing freight train.
Man dies after falling from train A 19-year-old man died after falling or jumping from a passenger train heading for Swansea on 24 November. He is thought to have joined the train at Newport and is believed to have been travelling with a group of friends. The body was found at the lineside near Llansamlet (Western).
Track worker killed near Reading A track worker was struck and killed by a train at Kennet Bridge, near Slough, on 29 November (Western). Prior to the incident, the Down and Up Relief lines had been blocked by engineers’ possession. RAIB has launched an investigation into the incident.
December
Purley conviction overturned A train driver involved in the Purley train accident had his manslaughter convictions overturned. The incident involved a London-bound train, which collided with another service outside the station in March 1989, following a SPAD. There were five fatalities and over 80 injuries. Following the incident, the driver admitted two counts of manslaughter, and was jailed in September 1990 for 18 months, 12 of which were suspended.9 However, following later information coming to light that there had been four previous SPADs at the same signal in the five years prior to the incident, the appeal court quashed the original conviction.
Passenger fatal assault at station On the afternoon of 15 December, a 17-year-old male was stabbed on the platform at Dewsbury station (London North West). Two teenagers have been charged with murder.
9 The term was later cut to four months on appeal.
18 2007 Annual Safety Performance Report Safety overview
2.4 Longer-term trends 2.4.1 Rail usage Chart 7 shows the changing levels of rail usage over the last five decades. Figures fell between the mid-1950s and early 1980s, affected by increasing ownership of road vehicles. However, since the 1980s, there has been a gradual re-growth in all four measures, reflecting changes in society, transport policy and the economic climate.
Chart 7. Trends in train kilometres and passenger usage 1956–2006
160 Passenger journeys 140 Passenger km Freight moved (tonne km) 120 Train km (passenger and freight)
100
80
60
Index (base year 1956 = 100) 1956 = year (base Index 40
20
0 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006
Sources: ORR for train km, DfT (Transport Statistics Great Britain) for other measures 2.4.2 Rail safety Against the background of changing traffic and usage, the railway has become safer for passengers and the workforce. Chart 8 shows the relevant trends in fatalities since 1947.
Chart 8. Reductions in number of fatalities since 1947
300 600 Passengers
250 Workforce 500 Public (trespass, non-trespass, suicide and suspected suicide)
200 400 - public Fatatlities
150 300
100 200
Fatalities - passengers and workforce passengers - Fatalities 50 100
0 0 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Source: Passengers and workforce – ORR data for mainline railway up to 1993, RSSB data from 1994 onwards. Public – ORR data (including LUL) up to 1993, RSSB data (excluding LUL) from 1994 onwards.
2007 Annual Safety Performance Report 19 Safety overview
Immediately after World War II, the railway was operating with equipment that had (from necessity) been overworked and under-maintained. As technologies improved, further safety schemes were developed, such as multi-aspect signalling (MAS) and the Automatic Warning System (AWS). The many initiatives devised in more recent years to address SPAD risk, including the Train Protection and Warning System (TPWS) – together with improvements in the crashworthiness of rolling stock – have led to further improvements in passenger fatality risk.
The greatest improvement, in percentage terms, has been in the number of workforce fatalities, which stood at remarkably high levels after the war. Clearly, the amount of maintenance work being performed during this time, as well as the more rudimentary (and labour-intensive) methods used, contributed to the higher-risk environment. Furthermore, after the 1955 Modernisation Plan, greater mechanisation, and the introduction of continuously welded rail (with improved quality steel), not only reduced the railway’s maintenance requirement, but also helped create better working conditions for staff.
The largest variation, and least overall improvement, is seen in the public trespass and suicide fatality rates. This is understandable, given that the causes of trespass and suicide will not be directly influenced by technological or methodological advancements in railway operations.
Chart 9 shows the clear progress that has been made in the area of train accidents. Over the last 60 years, the number of train accidents resulting in fatalities to passengers and/or members of the workforce has steadily reduced. The average number per year (based on a rolling 10 years) is currently less than one. (The chart excludes train collisions with road vehicles that resulted solely in public fatalities.)
Chart 9. Reduction in number of fatal train accidents with passenger/workforce fatalities since 1947 12 Number of train accidents with passenger/workforce fatalities 11
10 Average number over preceding 10 years
9
8
7
6
5
4
3 Number of fatal train accidents train fatal of Number 2
1
0 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
Source: ORR data for historic statistics, SMIS for recent data.
20 2007 Annual Safety Performance Report Safety overview
2.5 Looking to the future The future management of safety on the railway will be influenced by a number of factors, both external and internal to the industry.
2.5.1 Future influences on safety Politics, the economy and society
Over the last 10 years, the railway has seen a 40% increase in passenger journeys, and the HLOS predicts a further 22% increase in passenger numbers by 2014.
In July 2007, the government released Delivering a Sustainable Railway,10 which presents future ambitions for the industry:
‘Our ambition for the future is for a railway that:
• Can handle double today’s level of freight and passenger traffic. • Is even safer, more reliable and more efficient than now. • Can cater for a more diverse, affluent and demanding population. • Has reduced its own carbon footprint and improved its broader environmental performance.’ The document then discusses how these ambitions can be achieved, covering everything from getting the basics right in the present day, to enabling growth and anticipating the longer-term future (ie, the situation in 2030). As well as a UK agenda, there have also been developments on a Europe-wide scale; these are discussed in Chapter 13, European developments.
The forecast growth in rail travel presents safety challenges. For example, increasing traffic might cause more wear and tear on the track, and the consequences of a train accident are likely to be higher if the trains involved are carrying more people. Another potential effect of increased usage is crowding on trains and in stations. Although more than 60% of passengers are satisfied that there is sufficient space to sit or stand on trains,11 there are problems on some of the busiest routes, where 1,300 new carriages will be introduced to address the issue of crowding. The option of lengthening trains to up to 16 carriages is also being considered as a possible solution.
As well as passing through stations in connection with train journeys, passengers and public are also using the retail facilities in stations in greater numbers. The railway industry is encouraging this trend in some areas by promoting the concept of the ‘destination station’. Such was seen recently when St. Pancras International opened: the marketing was not just aimed at rail travellers, but anyone interested in using its shops and bars.
Demographic change, and particularly the fact that the UK population is ageing, is also likely to have an impact on passengers, the workforce and members of the public.
10 Department for Transport, Delivering a Sustainable Railway (DfT, July 2007). 11 Passenger Focus – National Passenger Survey, Autumn 2007.
2007 Annual Safety Performance Report 21 Safety overview
In terms of customer demand, there is likely to be an increase in the percentage of mature and elderly travellers, which will raise issues of health and safety, customer mobility and access. To combat this, upgrades to over 260 stations will be completed between 2012 and 2015, to make rail travel more accessible for the disabled and elderly, as well as parents with pushchairs and young children. The ‘Access for All’ project will provide step-free access from entrance to platforms, plus better lighting, hearing induction loops and passenger- information screens. It also plans to install obstacle-free routes. Research studies to understand the impact of changing demographics on the railway are currently under way by RSSB.12
It is hard to make accurate predictions about the way that societal changes will impact on the railway. Although the industry can take measures to prevent them, and mitigate their effects, deliberate actions, such as assaults, acts of vandalism, suicides and level crossing violations are closely related to the wider social context.
Environmental change
There is now a worldwide consensus that a global temperature rise has begun and will continue in the future. This is predicted to have an effect, not only on the average temperature in the UK, but also on the occurrence of extreme weather events. These have the potential to affect the running of a safe railway in a variety of ways:
• An increase in flooding events could affect the railway directly, or could cause landslips, when cuttings or embankments fail. Droughts can also cause instability of the earth on cuttings/embankments due to the loss of moisture. • The predicted rise in sea levels could cause more coastal erosion; this will have an impact on railway lines that follow the coastline. • Hotter summers could cause an increase in track buckles, which might increase the potential for derailments. • More storms with higher winds could have many effects, such as damage of rail vehicles, overhead lines, buildings or other infrastructure. Technology
The railway continues to explore the use of new technology for improving the operational railway, both in terms of safety and/or performance. For example, developments currently on the horizon include GSM-R (Global System for Mobile communications – Railway), ERTMS (European Rail Traffic Management System) and the use of obstacle detectors at level crossings.
GSM-R The GSM-R National Voice Radio Programme is a cross-industry programme led by Network Rail. A trial began in Strathclyde on 23 October 2007 and will be rolled out nationally from 1 January 2009, with the majority of the country being completed by the end of 2012. GSM-R will allow direct communication between the signaller and train driver throughout the country.
12 RSSB research project T661, Scoping research on ageing for passengers and the railway workforce RSSB research project T559, Health, safety and employability of an ageing workforce.
22 2007 Annual Safety Performance Report Safety overview
This will reduce the risk introduced through third-party communication, and address many public and formal inquiry recommendations. GSM-R also introduces a driver-initiated emergency call that alerts the controlling signaller and other drivers in the vicinity, allowing other drivers to react immediately. This reduces the risk from collisions with other trains and obstacles on the line.
ERTMS This National ERTMS Programme is also being led by Network Rail. It commences with the Cambrian Lines in 2009; migration schemes follow until national roll-out starts around 2017. ERTMS includes an Automatic Train Protection (ATP) system, reducing the risk from train collisions, and continual speed supervision, reducing the risk arising from overspeeding. The programme is also developing a mobile protection system that allows greater signalling protection to track workers and allows engineering possessions and protection arrangements to be set up more quickly, reducing the reliance on Red Zone working. ERTMS developments also include enhancing protection at user-worked level crossings, although this is currently at the early stage of development.
Obstacle detectors In 2006, RSSB examined the options of using obstacle detection systems, based on radar technology, to improve safety at level crossings.13 Such devices are already used elsewhere in Europe to detect obstructions capable of causing significant damage to a train, or to assist the signaller in charge of a CCTV-controlled crossing. However, such a system has to be sensitive enough to distinguish between a significant threat to a train (such as a car), from an insignificant one (like a shopping basket or a small animal) in order to avoid unacceptably high levels of safe-side (false) activations. Further work is in progress to define the size criteria that the system might apply.14
2.5.2 Research and Development Programme RSSB manages a programme of research and development (R&D) on behalf of the government and the railway industry. It is funded by the Department for Transport (DfT) and aims to assist the industry and its stakeholders in achieving the key objectives of improving performance and increasing capacity and availability whilst reducing cost. These objectives should be used to compete effectively with other transport modes (or complement them as appropriate) and deliver a sustainable future for the railway.
For more information on the R&D programme, including the projects and reports referred to in this chapter, please see the R&D section of the RSSB website (www.rssb.co.uk).
2.5.3 Sustainable Rail Programme The Sustainable Rail Programme (SRP) was formed to support the UK government Sustainable Development Strategy launched in March 2005, along with the subsequent DfT Sustainable Development Action Plan (which included the railway).
Central to the SRP is the development of a 30-year sustainable development (SD) strategy for the rail industry. This consists of two parts:
13 RSSB research report T522, Research into obstacle detection at level crossings. 14 RSSB research project T729, Determining whether it is ‘ALARP’ to restrict the size of objects to be detected by obstacle detectors at level crossings.
2007 Annual Safety Performance Report 23 Safety overview
• A route map, completed early in 2007 and discussed in the 2006 ASPR. • The second stage, now complete, which sets out four possible future scenarios. These are on the extremes of two scales, with one scale relating to the amount of travel in the future and the other relating to how much government policy drives change. The scenarios are designed to help people take a different view of the choices facing them now, and to make better long-term decisions as a result. It should improve strategic decision-making when the future seems complex or uncertain.15
More information can be found on this subject by contacting RSSB at the following address: [email protected].
15 This is also covered in RSSB research project T713, Foresight studies and futures planning for the rail sustainable development strategy.
24 2007 Annual Safety Performance Report SSP trajectories
3 Review of progress against SSP trajectories
This chapter investigates safety performance in each of the nine Key Risk Areas (KRAs) identified in the 2007–09 Strategic Safety Plan (SSP). The SSP covers safety performance over a three-year period; the safety performance presented in this chapter contains data for 2007, which corresponds to the first year of the plan.
2007 Headlines • Overall, there is evidence that the industry is delivering on the commitments made in the SSP; in six of the nine KRAs, tangible progress towards delivery of the trajectory has been achieved.
• In the remaining three KRAs, there is insufficient data to comment on how the presence of a trend (ie, there are no areas where there is evidence that the trajectory will not be met). • KRAs that show the greatest improvements are Workforce – track workers, Engineering – trains and Engineering – track. Performance at a glance The table below presents the trajectories in each of the KRAs, together with an assessment of whether, after one year of the three-year plan, progress towards achievement of the nine trajectories is apparent from the data. Progress towards the trajectory is indicated by a tick, and failure to make progress is marked with a cross. Where there is insufficient data to ascertain whether progress is being made, a dash has been placed in the final column.
Key Risk Area Trajectory Progress Passengers at stations Deliver a measurable reduction in the level of risk caused by the behaviour of passengers in stations by reducing or eliminating the 9 typical hazards that result in accidents . Passengers on trains Given the improvements in this Key Risk area, the aim is to maintain the current level of safety, improve it where reasonably practicable and tak e opportunities to reduce risk when there are train - refurbishments projects and new rolling stock procured. Workforce - train crew Given the improvements in this k ey risk area, the aim is to maintain the current level of safety, and improve it where reasonably 9 practicable, over the next three years. Workforce - track workers Deliver a measurable reduction in the rate of track worker accidents. 9
Workforce - at stations Deliver a measurable reduction in the risk of accidents to passengers in stations through an improvement in the competency of - station staff. Engineering - track Given the improvements in this Key Risk area, the aim is to maintain the current level of safety, and improve it where reasonably 9 practicable, over the next three years. Engineering - trains Deliver a measurable reduction in the risk from catastrophic accidents and accidents to passengers from vehicle defects. 9
Public behaviour - crime Deliver a measurable reduction in the risk from crime to passengers and the work force. 9
Public behaviour - level crossings Deliver a measurable reduction in the risk from the misuse of level crossings. -
2007 Annual Safety Performance Report 25 SSP trajectories
3.1 Introduction Effective safety planning requires a detailed understanding of the activities or circumstances that result in the greatest risk to passengers, the workforce and members of the public. To identify the focus areas for the 2007–09 SSP, the sources of risk were categorised into 18 risk areas and the SRM was used to estimate the contribution of each to the industry’s risk profile. Nine risk areas accounted for more than 95% of the total risk (in terms of fatalities and weighted injuries); these were designated Key Risk Areas (KRAs).
Chart 10 shows the amount of risk that is caused by, or is under the control of, each KRA. The greatest contribution comes from Public behaviour – crime. This reflects the danger in which trespassers place themselves, but also includes the risk to other people from assaults and vandalism. The three KRAs relating to the workforce reflect both the risk from individual accidents to the workers themselves, and the risk to passengers that can result from workforce errors or oversights (for example, when carrying out train despatch duties, or failing to properly maintain a safe station environment).
Chart 10. Risk profile by Key Risk Area (sources of risk*)
Passengers - at stations Darker shades represent fatality risk Lighter shades represent weighted injuries Passengers - on trains
Workforce - train crew
Workforce - track workers
Workforce - at stations
Engineering - track
Engineering - trains
Public behaviour - crime
Public behaviour - level crossings
Other sources of risk
0 10203040506070 Fatalities and weighted injuries
Source: SRMv5.5. * The Key Risk Areas relate to sources of risk so, for example, the bar corresponding to Passengers – on trains shows the risk that arises from passenger behaviour on trains, rather than the risk to passengers on trains.
The 2007–09 SSP also introduced the concept of trajectories in each of the KRAs. Trajectories are a way of illustrating expected changes in the level of safety risk as a result of the industry’s actions. They have as their starting point the level of risk at January 2007 and the actions put in place to address that risk; a reasonable judgement is then made by each duty holder to assess the changes in safety performance that can be realistically expected over the three years of the plan as a result of the actions. The development of the trajectories is therefore rooted in the industry’s safety planning process, because each is calculated from the industry’s own assessment of the actions it intends to take over the three-year lifetime of this SSP.
26 2007 Annual Safety Performance Report SSP trajectories
Figure 1 shows the 18 risk areas considered for the 2007–09 SSP. The KRAs are highlighted in colour, and the charts that are used in this chapter to monitor progress against the trajectories are shown in the third column.
Figure 1. Risk areas, Key Risk Areas and their corresponding monitoring measures
Risk area group Risk area Monitoring measures
Slips, trips and falls At stations Platform-train interface Passengers Passenger struck against object Fall from train in On trains running
Structures Slips, trips and falls Environment Adjacent Personal accidents properties
Train crew Slips, trips and falls
Train despatch Shunter
Struck by object Depots Other occupational accidents Workforce Track workers Struck by train Signaller / crossing Slips, trips and falls keeper Station management Station staff Train despatch
Assaults Crime Trespass Road vehicles Public behaviour (not crossings) Road vehicles drivers at level crossing Level crossings Pedestrians at level crossing Structures
Track Track PIM indicator
Engineering Level crossings
Trains Train PIM indicator
Signalling and telecomms
Key Risk Areas are shown in colour. Risk areas that were not selected as Key Risk Areas are shown in grey.
2007 Annual Safety Performance Report 27 SSP trajectories
The safety performance trajectories set out in the SSP are each presented in terms of risk. Hence, ‘fatalities and weighted injuries’ (FWI) is the basis of all data in this chapter. For most of the charts, the FWI risk is calculated directly from injuries reported into SMIS as there are a large number of incidents. However, injuries from reported accidents are not a reasonable indicator of low-frequency, high-consequence accidents, such as collisions and derailments. This mainly affects the KRAs related to Engineering – track and Engineering – trains. For these two areas, the information has been taken from the Precursor Indicator Model (PIM), which is described in detail in Chapter 8, Risk from train accidents. To indicate the current trend, safety performance information for 2007 together with the previous four years is provided.
For the 2007–09 SSP, the projected safety performance in each KRA is qualitative rather than numeric. Subsequent versions of the SSP contain specific quantitative trajectories. This chapter monitors performance against the qualitative trajectories contained in the 2007– 09 SSP. It also summarises the key industry safety initiatives that were undertaken in 2007, many of which are still ongoing.
In other chapters of this report, the risk profile is presented in terms of risk area groups rather than risk areas. The 18 risk areas can be combined into the five risk area groups (which therefore cover 100% of the total risk). The groups, and their constituent risk areas, are shown in Figure 1. The only risk area group that does not contain any of the KRAs is Environment.
3.2 Review of progress A significant difference between the presentation of safety performance in the SSP and this chapter, compared to other chapters in the ASPR, is that the breakdown of risk is based on the cause of the risk, rather than the group exposed to it. For example, the precursors that give rise to passenger slips, trips and falls in stations within this chapter are covered by both Passengers at stations and Workforce – station staff (as station staff have a responsibility to ensure stations are free from obstacles and provide a reasonable underfoot surface). A judgement has been made for each precursor regarding the KRA that gives rise to the risk. This means that it is not possible to draw a direct comparison between charts in this chapter and charts in other chapters.
3.2.1 Passengers at stations The Passengers at stations KRA is dominated by slips, trips and falls. There is also a significant contribution from accidents at the platform–train interface. Depending on the cause of the accident, some platform–train interface risk is also allocated to train crew and station staff as they provide some 14 management of the risk during their train Passengers at stations: slips, trips and falls despatch activities. Assaults on passengers 12 are covered under Public behaviour – crime. 10 8 The chart for slips, trips and falls shows 6 that, for 2007, the risk is at a lower level 4 than in any of the previous four years. 2
The chart – and subsequent charts relating FWIper billion passenger journeys 0 to the risk associated with passengers – is 2003 2004 2005 2006 2007
28 2007 Annual Safety Performance Report SSP trajectories
normalised by the number of passenger journeys to account for the increase in railway usage in recent years.
12 The chart showing the risk at the platform– Passengers at stations: train interface shows a reduction after 10 platform-train interface 2003, which coincides with the phasing-out 8 of slam door stock. Subsequently, there is 6 a relatively high degree of fluctuation from
4 year to year. This is not due to major fluctuations in underlying safety 2 performance; the trend is dominated by the
FWI per billion passenger journeys 0 occasional fatality during boarding and 2003 2004 2005 2006 2007 alighting.
There were three such fatal incidents in 2007. Therefore, there is no statistical significance to the relatively high level seen in 2007.
The trajectory contained in the 2007–09 SSP for this Key Risk Area is to: ‘Deliver a measurable reduction in the level of risk caused by the behaviour of passengers at stations by reducing or eliminating the typical hazards that result in accidents.’
Given the recent reduction in slips, trips and falls and the generally static performance at the platform–train interface, there is evidence that progress towards the trajectory is being made.
Initiatives to support the trajectory
The following initiatives were undertaken in 2007 to support the trajectory:
• A considerable amount of work has been completed to reduce the risk from slips, trips and falls. In particular, some operators have applied a slip, trip and fall toolkit to prioritise remedial work in stations. • Safety on stairs has been addressed by marking the nosing on stairs at high-risk stations. This has been an ongoing initiative over recent years. • Strategies to improve the walking surfaces at stations have been developed based on coefficient friction measurements. • RSSB’s good practice guide Way-finding at stations and the ‘Way-finding toolkit’ has been widely adopted to assist operators with optimising signage at stations. • Southeastern has improved the way-finding and signage at 20 stations, which should help reduce crowding and rushing for trains. • South West Trains (SWT) has developed a safety management system specifically to control the risks likely to arise during major project improvements at stations. • There are also examples of stations being upgraded during refurbishment work. An example of this is a service underpass at York, serving all platforms, which was resurfaced with slip-resistant material and the stairs reconstructed. The underpass and the lift access were also refurbished. The underpass was then brought into use as an additional passenger access route.
2007 Annual Safety Performance Report 29 SSP trajectories
3.2.2 Passengers on trains The risk from passenger behaviour on trains comprises passengers being struck against objects (either parts of the train interior or luggage), slipping, tripping and falling on board the train, and falling from trains in running.
2.0 Passengers on trains: 1.6 Passengers on trains: struck against object slips, trips and falls 1.5 1.2
1.0 0.8
0.5 0.4 FWI per billion passenger journeys passenger billion per FWI
FWI per billion passenger journeys passenger billion per FWI 0.0 0.0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
The risk from being struck by objects and slips, trips and falls has remained static; this is compatible with the SSP trajectory, which states: ‘Given the recent improvements in this Key Risk Area, the aim is to maintain the current level of safety, improve it where reasonably practicable and take the opportunity to reduce risk when trains are refurbished or procured.’
Passengers falling from trains in running 2.0 had been virtually eliminated with the Passengers on trains: fall from train in running removal of slam-door stock; however, there 1.5 were two isolated incidents in 2007, one where a fare evader escaped from a train 1.0 and another where a passenger jumped from a train following an argument with 0.5 another passenger. Both incidents were largely outside the railway’s control. The FWI per billion passenger journeys 0.0 two incidents account for the peak in 2007 2003 2004 2005 2006 2007 on the chart.
Given the performance in the three topics that comprise the trajectory, it is not yet considered possible to establish whether progress has been made towards delivery of the three-year trajectory.
Initiatives to support the trajectory
The widespread replacement of rolling stock over recent years has reduced the scope for major initiatives in this area as the design of the new trains incorporates improved crashworthiness, improved door design, enhanced interior design and compliance with modern standards. Hence, initiatives for dealing with passenger behaviour on trains tend to be local rather than network-wide. Examples include:
• National Express East Coast has implemented a holistic approach to the handling of luggage. This looks at the booking arrangements, the arrival at the station, the movement onto and about the platforms, storage on the train and the arrangements at the destination station. This reduces the risk from passengers being struck by luggage and reduces the potential for slips, trips and falls.
30 2007 Annual Safety Performance Report SSP trajectories
• Hull Trains train crew have been coached in safe handling awareness in relation to their passengers, empowering them to assist customers in the correct stowage of their luggage through correct lifting techniques. This should also reduce the risk from passengers being struck by luggage and reduces the potential for slips, trips and falls. • A number of train operating companies have also increased their staff presence by deploying on-board security staff. 3.2.3 Workforce – train crew There are four main aspects to the risk from errors by members of the train crew:
• SPADs. • Train despatch. • Slips, trips and falls. • Other personal accidents. The last two mostly affect the members of the train crew themselves. The category other personal accidents includes accidents involving catering staff on trains (such as incidents involving the trolley and hot beverages) and driver and guard manual handing accidents.
16 10 Train crew: slips, trips and falls Train crew: miscellaneous 8 personal accidents 12
6 8 4
4 2 FWI per billion train miles train billion per FWI miles train billion per FWI
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2.5 Most of risk from train despatch errors is to Train crew: train despatch passengers. 2.0
The risk from SPADs affects both train crew 1.5 and passengers (with passengers accounting 1.0 for the majority in FWI terms). 0.5 The trend in SPAD risk is presented in FWI per billion passenger journeys passenger billion per FWI 0.0 Chapter 8, Risk from train accidents. 2003 2004 2005 2006 2007
The SSP trajectory in this KRA is: ‘Given the improvements in this Key Risk Area, the aim is to maintain the current level of safety, and improve it, where reasonably practicable, over the next three years.’
The trend in the risk from both train crew personal accidents and train despatch is static, whereas the trend in train crew slips, trips and falls is decreasing. The downward trend in slips, trips and falls since 2003 is statistically significant. SPAD risk has also been managed
2007 Annual Safety Performance Report 31 SSP trajectories
to a historically very low level, which is currently being maintained; thus there is initial evidence that the trajectory in this KRA will be achieved.
Initiatives to support the trajectory
• There has been an increased use of driver simulators at several TOCs over recent years. Attention is now being focused on their application and the development of training programmes and integration with the driver assessment programmes. • There has been widespread application of the fatigue index to assist with the rostering of train crew and other safety critical staff. • There have been widespread initiatives to improve communication with drivers on technical and operational safety issues. • There has been an industry programme to enhance driver–signaller understanding through cab rides for signallers and signal box visits for train drivers. • At the national level, an Operations Focus Group (OFG) replaces the former National SPAD Focus Group and its associated steering group. It also encompasses relevant aspects of the former Track Safety Strategy Group and the Safety Critical Communications Focus Group. The purpose of the OFG is to facilitate the progressive improvement of operational safety through the identification, discussion, development and promotion of justifiable effective campaigns, programmes and tools. Specifically, it aims to monitor and review industry performance trends in relation to operational safety and consider changes to priorities and strategies, making recommendations to the RSSB Board as appropriate. • In addition to the national OFG, the industry has expanded the remit of the former, local SPAD strategy groups to encompass a wider range of operational incidents. The outputs of these new OPSRAM groups are being implemented through action plans. The development of these groups is in recognition that operational risk is a wider subject than just SPADs. • GNER (now National Express East Coast) introduced new instructions for the positioning of staff during train despatch to ensure that the guard can be contacted in an emergency and can bring the train to a halt during the train despatch process. • There has been continuous improvement to monitoring of the conductor’s door operating procedures. One recent initiative has been implemented by Arriva Train Wales (ATW) to improve the safety of train despatch at unmanned stations. • South West Trains (SWT) has undertaken manual handling refresher training for guards. 3.2.4 Workforce – track workers Four measures are used to assess the safety performance for track workers: struck by objects, struck by train, slips, trips and falls, and a ‘catch-all’ category, miscellaneous occupational hazards.16
Each of the categories is showing a generally reducing trend since 2003, with 2007 being at relatively low level compared to the average of the previous four years. The exception to this is track workers struck by train. However, the risk in this category is dominated by a
16 For track workers, miscellaneous occupational hazards include incidents such as crushing injuries, burns and cuts.
32 2007 Annual Safety Performance Report SSP trajectories
relatively low number of fatalities; therefore little statistical significance can be attributed to the two fatal events that occurred in 2007.
The trajectory for this Key Risk Area is to: ‘Deliver a measurable reduction in the rate of track worker accidents.’ After one year of the three-year trajectory, there is evidence that progress is being made towards delivering a measurable reduction in track worker risk.
0.15 0.30 Track workers: Track workers: struck by objects 0.25 slips, trips and falls 0.10 0.20
0.15
0.05 0.10
0.05
FWI per thousand track workers track thousand per FWI 0.00 workers track thousand per FWI 0.00 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
0.30 Track workers: miscellaneous 0.25 occupational hazards Track workers: 0.25 0.20 struck by train 0.20 0.15 0.15 0.10 0.10
0.05 0.05
FWI per thousand track workers track thousand per FWI 0.00 workers track thousand per FWI 0.00 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Initiatives to support the trajectory
• All Maintenance Delivery Units have developed and implemented their own, local accident reduction plans. These are designed to focus on issues that have been identified through work activity risk assessments and local accident investigations. They give local ownership to local issues and empower people to deliver local resolution of safety issues. The plans are owned by Infrastructure Maintenance Managers and Maintenance Unit Delivery Managers. Achievements against the plans are monitored within the line through the Monthly Business Review process. • In order to improve the availability of suitable and sufficient welfare facilities for track workers, a phased plan has been developed and implemented for the provision of permanent welfare facilities at strategic locations. This was initiated in 2007 and will continue through 2008. • Work has been undertaken with Network Rail’s small plant and tool suppliers to reduce exposure of track workers to noise and vibration, and to reduce the likelihood of musculoskeletal disorders. This has included the redesign and modification of existing plant and tools, as well as the sourcing of new plant and tools. The methods of work employing the plant and tools have also been examined with a view to reducing exposure to hazards while maintaining or improving productivity. • A national risk-based programme to refurbish access points, removing hazards, reducing risk and creating improved access for track workers to the infrastructure has been instigated. The programme has been allocated £14 million during this control period (3). Maintenance Territories have prioritised work using risk criteria. Delivery of work
2007 Annual Safety Performance Report 33 SSP trajectories
commenced in 2005/2006 and is ongoing. The number of refurbishments completed to date is 650; by the end of the control period, some 1,500–1,800 will have been refurbished. • A national risk-based programme was initiated in 2007 to install fixed lighting at 70 junctions where a high risk to track workers from slips, trips and falls has been identified. A variety of lighting equipment is being tested at a designated trial site. The testing explores the actual levels of light required on site to perform particular tasks and will ensure that all technical issues are identified prior to a national roll-out (due to start in 2008). • New Lookout Operated Warning System (LOWS) technologies were trialled during 2007 and £2.5m has been allocated for the purchase of equipment during 2008. The new equipment will enhance the safety of both the lookout and the group being protected by providing both audible and visual warning throughout the worksite using modern radio technologies. • A revision of the COSS (Controller of Site Safety) forms has been undertaken. This included considerable user consultation and testing. The revised forms are simpler to complete and are much clearer. • To increase the effectiveness of workforce briefings, a training module on ‘effective briefing’ was developed and introduced to the COSS training course. The object of the module is to raise the delivery standards for COSS briefings and increase knowledge transfer. The aims are to raise workforce awareness of the hazards and control measures in place at the worksite and to raise the workforce’s expectations of a good briefing. • The ‘Safety 365’ safety awareness scheme continued throughout 2007, covering a variety of track worker and general safety topics using a variety of media. Subjects included: safety critical communications; slips, trips and falls; the use of mobile telephones on the infrastructure, and buried services. The media used included: briefing packs for use by line managers; DVDs; posters; booklets and pocket cards. An impact survey was carried out in the latter half of 2007 to assess the effectiveness of the ‘Safety 365’ campaigns. This demonstrated that the messages of the campaigns were being communicated and identified those campaigns that had been particularly successful. This information is being used in the development and delivery of the ‘Safety 365 – Healthy 2008’ campaigns. 3.2.5 Workforce at stations Two measures are used to assess the safety performance of workforce at stations: station management and train despatch.
Accidents related to station management 14 Workforce at stations: station management include such things as passenger slips, trips 12 and falls on wet surfaces. Station staff have a 10 responsibility to clear up spillages or water from leaking roofs. Many accidents in stations have 8 contributory factors that can arise either from 6 the way the station is managed or from the 4 behaviour of passengers (or a combination of 2 the two). Much of the risk from this Key Risk FWI per billion passenger journeys 0 2003 2004 2005 2006 2007 Area is therefore shared with passengers at stations.
34 2007 Annual Safety Performance Report SSP trajectories
Accidents at the platform–train interface 2.5 Workforce at stations: can be affected by the behaviour of 2.0 train despatch passengers or train crew (who also undertake train despatch) as well as the 1.5 actions of platform staff. As well as 1.0 boarding and alighting accidents, the platform–train interface topic encompasses 0.5 staff activities to ensure passengers stand
FWI per billion passenger journeys 0.0 away from the platform edge. Slipstream 2003 2004 2005 2006 2007 safety will be the subject of a forthcoming RSSB research project (T749).
The risk from station management, which is the biggest risk in this KRA, is showing a reducing trend with the risk in 2007 being the lowest in the five years of data analysed. The trend in train despatch is relatively static, with significant year-on-year variation. The main cause of the large variation is the occasional fatal accident.
The trajectory in this KRA is to: ‘Deliver a measurable reduction in the risk from accidents to passengers in stations through an improvement in the competency of station staff.’ As the risk is dominated by station management, rather than train despatch, there is evidence that progress is being made towards delivery of the three-year trajectory.
Initiatives to support the trajectory
The initiatives in this section are related to station management issues and therefore in common with the Passengers at stations initiatives (as station management is largely about dealing with the behaviour of passengers). These have not been repeated here. Additionally, the following tasks have been undertaken:
• There have been many initiatives to improve train despatch, including new instructions for the positioning of staff during train despatch to ensure that the guard can be contacted in an emergency and can bring the train to a halt during the train despatch process. The improved monitoring of train despatch at unmanned stations has also been covered. • SWT has revised its risk assessment procedures for stations to include greater involvement of staff that carry out the tasks in the risk assessment. SWT has also developed a toolkit for assessing the ergonomic risk of using new ticket issuing equipment within ticket offices. Note that initiatives to address assault risk to both passengers and staff are included in section 3.2.8, Public behaviour – crime.
3.2.6 Engineering – track As mentioned in the introduction, for track risk, the PIM measure17 has been used to monitor safety performance rather than the actual number of injuries. This is because the risk is dominated by infrequent, but high-consequence, accidents.
17 The PIM Indicator is described in detail in section 8.5.1.
2007 Annual Safety Performance Report 35 SSP trajectories
The PIM measure is indicative of the level of 8 Engineering - track: risk, but is not in units of FWI (although it is 7 PIM indicator normalised by train miles). The PIM risk 6 indicator for track includes precursors such 5 as track twisted, broken rail and track 4 buckle. The chart shows a decreasing trend 3 since 2003, with a notable step downward in indicator risk PIM 2 2007. More details of the calculation of the 1 PIM can be found in Chapter 8, Risk from 0 train accidents. 2003 2004 2005 2006 2007
The trajectory in this Key Risk Area states: ‘Given the improvements in this Key Risk Area, the aim is to maintain the current level of safety, and improve it where reasonably practicable, over the next three years.’ Given the safety performance improvement achieved in this KRA, the initial indication is that the three-year trajectory can be achieved.
Initiatives to support the trajectory
The following initiatives have been implemented by Network Rail in the areas of track and trackside structures:
Track • There has been increased ultrasonic testing of the rails with approximately three-quarters of track in Categories 1A to 3 being covered by the ultrasonic test unit.18 This regime, together with targeted renewals, rail grinding and treatment of dipped joints, has delivered a reduction in both the number of broken rails and the number of high-risk rail end breaks. • A national programme has been delivered to identify sites with increased risk potential for track buckles and implement a package of control measures, including the fitting of lateral resistance plates to selected points. • New arrangements for the management of derailment risk at switches have been developed, documented in a new standard (NR/L2/TRK/0053) and briefed out to front- line teams. • Further progress has been made in reducing the number of joints with dip angle exceedances resulting in the national figure being reduced by almost one-fifth. • Enhanced management of track gauge spread has reduced the number of derailments from this cause. Trackside structures • Standard design details for earthworks have been distributed to engineers and contractors to spread good practice and to promote easily maintainable designs. • Rock Slope Hazard Index and Soil Slope Hazard Index systems have been introduced as part of the company’s specification for the examination of earthworks – these are having benefits in the prioritisation of rock scaling and netting works, and soil cutting and embankment repair work.
18 Track is classified into seven categories, numbered 1a to 6, based on the speed of the line and the type and quantity of traffic, where Category 1A is for the highest train speeds. The rating is connected with the use of the line and the rate of wear and tear.
36 2007 Annual Safety Performance Report SSP trajectories
• A new hazard-reporting system has been introduced, allowing analysis of earthwork failures in order to recognise developing failure patterns at an early stage. This will aid the planning of repair work to counter these trends. • A new standard (NR/GN/CIV/801) has been introduced for the application of the observational approach to the design of remedial works to earthworks, supplementing the existing standard (NR/SP/CIV/071) for the design of earthworks. These measures have provided for improved targeting of maintenance work. • Good progress has been made in assessing tunnel management risks. • Improved co-ordination has been achieved with local authorities and the County Surveyors’ Society on the management of bridge strikes.
3.2.7 Engineering – trains 6 Engineering - trains: Similar to the risk from track, the risk from 5 PIM indicator rolling stock failures is monitored using the 4 PIM. The PIM measure shown in the chart is based upon fires due to rolling stock failures, 3 hot axle box failures, brake failure and other 2 PIM risk indicator risk PIM safety related defects. The chart shows a 1 progressive reduction in risk since 2003. 0 2003 2004 2005 2006 2007 The SSP trajectory in this KRA is to: ‘Deliver a measurable reduction in the risk from catastrophic accidents and accidents to passengers from vehicle defects.’ Given the evident improvement in this area, it is concluded that there is tangible progress towards delivery of the three-year trajectory.
Initiatives to support the trajectory
Given the widespread replacement of slam-door carriages, and the diverse range and age of rolling stock that exists across the network, initiatives in this area tend to be local rather that national programmes. Specific initiatives include:
• National Express East Coast has undertaken modifications to its HST fleet and has continued with improved suspension and improved interior layout design, lighting and use of materials. • Train operating companies have, where appropriate, modified their train horns to ensure compliance with the noise directive. • ATW has introduced improved underframe cleaning of trains and is currently implementing door modifications to prevent wrongside door opening. • EWS has made two significant modifications to its Class 66 fleet. First, the parking brake has been modified to provide a clearer indication to the driver that the brake has been correctly applied / released, and to achieve full application without the need for a driver to maintain control of push buttons. Secondly, the brake frame has been modified to reduce the possibility of incorrectly preparing a locomotive for dead haulage. 3.2.8 Public behaviour – crime Public behaviour – crime is the highest-risk KRA. The two main topics associated with crime are assaults and trespass events. The trajectory does not cover trespass events as they are largely outside of the control of the railway, hence only the trend in assaults is used to
2007 Annual Safety Performance Report 37 SSP trajectories
monitor performance against the trajectory. The risk from assaults is to both passengers and the workforce19. 10 Public behaviour: The trajectory in this KRA is to: ‘Deliver a assault 8 measurable reduction in the risk from crime to passengers and the workforce.’ There is an 6 evident downward trend in the risk from 4 assaults; therefore it is concluded that FWIyear per progress is being made towards delivery of 2 the trajectory. 0 Initiatives to support the trajectory 2003 2004 2005 2006 2007
• The DfT operates a national secure station scheme where stations that fulfil specific criteria can achieve secure station accreditation. The scheme has been ongoing for a few years; 2007 saw a significant increase in the number of stations that have achieved Secure Station status. • In addition to the stations that have achieved Secure Station status, operators have also achieved Park Mark Safer Parking status for numerous car parks.20 The Park Mark is awarded to parking facilities that have met the requirements of a risk assessment conducted by the Police. The requirements mean the parking operator has put in place measures that help deter criminal activity and anti-social behaviour, thereby doing everything reasonably practical to prevent crime and reduce the fear of crime. For customers, using a Park Mark Safer Parking facility means that the area has been vetted by the Police and has measures in place to create a safer environment. • Security in stations and on trains has been enhanced by the increased use of Police Community Support Officers (PCSOs) and specialist security personnel to police high- risk locations. • Mosquito devices have been installed to deter youths under 25 from congregating in specific areas. These devices emit ultrasonic sound, which can usually only be heard by young people. They have been shown to be effective in dispersing gatherings of youths and hence preventing antisocial behaviour at known trouble spots. • At particular locations, barriers have been installed to reduce verbal and physical assaults on staff carrying out revenue protection duties. • Security action groups have been developed to implement and monitor initiatives to deter and reduce the number and effect of staff assaults. One such initiative is the Security Workplace Action Group (SWAG) that has been implemented by Northern, which has held roadshows to seek feedback and suggestions from staff on how to reduce assaults. • There has been increased joint planning between BTP and train operators to increase the effectiveness of BTP and railway resources. • The concept of a ‘Safer Zone’ on stations was trialled in a number of locations. Safer zones are areas within stations which have heightened levels of security through, for
19 Assaults on passengers and members of the public are not well recorded in SMIS. The data in the chart is therefore largely made up of assaults on rail staff. Assault data and trends in assaults are covered in more detail in Chapter 10. 20 Note that assaults in car parks are outside the scope of this report. However, this initiative is included as improved car park security will have a positive effect on overall station security.
38 2007 Annual Safety Performance Report SSP trajectories
example, enhanced CCTV coverage and lighting. Some operators some have implemented the scheme whereas other are evaluating the results from trials. • Enhanced training programmes are being implemented to help customer facing staff handle conflict, personal safety and security (an example of this is the SWeRvE DVD). • ‘Airwave’ radios have been implemented by some operators, which are capable of communicating on emergency services’ frequencies to assist with complementary policing. • Poster campaigns have been run to publicise commitment to prosecuting offenders and supporting staff. • Forward-facing cameras have been fitted to trains on certain routes to monitor and record route crime. • Mobile CCTV units have been used to target trespass and vandalism at hotspots. • Motion sensing cameras have been fitted at sidings and platform ends. The cameras activate a sound alarm when a trespass event occurs. 3.2.9 Public behaviour – level crossings Two measures are used to monitor performance in this KRA: pedestrian behaviour at level crossings and road vehicle driver behaviour at level crossings. Trends for each of these are shown in the charts below. The risk from pedestrian behaviour is predominantly to the pedestrians involved, with a smaller risk to the train driver from the shock/trauma associated with being involved in the incident. For road vehicle driver misuse, the majority of the risk is typically to the road user, but incidents such as that at Ufton Nervet in 2004 illustrate the risk to passengers and the workforce.
35 Public behaviour at level crossings: pedestrians 40 Public behaviour at level crossings: road users 30 35 30 25 25 20 20 15 15 10 10 FWI per billion trainmiles 5 miles train billion per FWI 5 0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
The trajectory for this KRA is to: ‘Deliver a measurable reduction in the risk from the misuse of level crossings.’ For both pedestrian and road users, misuse accidents are infrequent; however, when incidents occur, they are likely to be fatal, especially for pedestrians. These two factors mean that it is not possible to assess the whether there is a significant trend in the injuries based on performance in year one of the three-year trajectory.
Initiatives to support the trajectory
• Network Rail has continued to implement and evolve the Don’t run the risk public awareness campaign to educate users on how to use level crossings correctly and warn them of the dangers of misuse. This included a hard-hitting television advertisement on prime-time television. Local radio, regional press adverts, outdoor posters and direct marketing to local residents were used to target ‘hot spot’ level crossings with the highest
2007 Annual Safety Performance Report 39 SSP trajectories
levels of misuse. Farmers were also sent education packs reminding them of how to correctly use their crossings. In 2007, the campaign specifically targeted the groups known to be at highest risk from crossing misuse: 18–24 year olds and the over-65s. In addition to its regular advertisements, Network Rail has also produced an online viral music video, which aims to educate people on the dangers of misusing level crossings. The broadcast and online aspects are one part of a campaign specifically aimed at ‘hot spot’ areas. All media channels were, and will continue to be, used – including television, radio, cinema, online, outdoor posters, newspaper advertisements and direct marketing.
• During 2007, Network Rail continued its strategy for reducing risk at level crossings based on: a programme of risk assessment to identify reasonably practicable measures for risk reduction; the continued reduction in the numbers of level crossings where justified; effective operation and maintenance; and education of the public on the risks of level crossing misuse. • A level crossing risk management toolkit was launched in September 2006 and updated during the summer of 2007. It is used by the industry to supplement level crossing risk assessments by providing detailed information about the human factors issues underlying crossing risk, and suggesting appropriate risk mitigation measures.
3.3 Future developments in the trajectory approach to safety planning The trajectories against which the safety performance is monitored in this chapter are those presented in the 2007–09 SSP. The 2008–10 SSP was issued in February 2008 and contains an enhanced set of trajectories, many of which are quantitative. These will be monitored in subsequent editions of the ASPR.
For the 2009 Plan, the following changes have already been identified:
• The Railways Act 2005 requires the Secretary of State for Transport (for England and Wales) and Scottish Ministers (for Scotland) to present to ORR a High Level Output Specification (HLOS) that they want the railway to deliver, and a statement of funds available (SoFA) for future investment. ORR then determines the outputs that Network Rail must deliver to achieve the HLOS, the cost of delivering them in the most efficient way, and the implications for the charges payable by train operators to Network Rail for using the railway network. Amongst other performance requirements, there are two specific to safety, one for passengers and one for the workforce. The specification is for: ‘A 3% reduction in the national level of risk to passengers and rail workers from 2008/09 to 2013/14’. The opportunity will be taken to align the SSP with the periods to which the ORR’s HLOSs apply. This means that, commencing in the 2009 Plan, the SSP will cover a period through to March 2014, with annual reviews.
40 2007 Annual Safety Performance Report SSP trajectories
SRM version 6, which is due in March 2009, will be used to provide an indicative benchmark level of risk for the industry at that time. Performance against HLOS safety measures will then be gauged against the risk estimates from subsequent versions of the SRM. • Efforts will be made to deliver quantified trajectories for the remaining KRAs. • All new passenger franchises, which have taken over since this plan was developed, will be integrated into the process to develop the Plan.
3.4 Further information The SSP can be downloaded from the RSSB website: www.rssb.co.uk.
2007 Annual Safety Performance Report 41 SSP trajectories
Intentionally blank
42 2007 Annual Safety Performance Report Benchmarking
4 Benchmarking railway performance
This chapter takes a step back and looks at railway safety in the wider context. It compares the risk associated with train travel with other modes of transport, assesses how Britain compares with its European neighbours in terms of railway safety, and compares occupational safety on the railway with other industries.
The chapter closes with a discussion on the progress that is being made to help train operators benchmark their own safety performance, in order to assess how they compare with the rest of the industry.
2007 Headlines • Rail travel is more than 20 times safer than car travel in terms of risk per km travelled. • In recent years, railway safety has improved more rapidly than the safety of car travel. • The safety of Britain’s railways compares well with the rest of Europe. • Track work and shunting are relatively high-risk occupations, compared with other jobs on the railway and other sectors of British industry.
4.1 Comparing the railway with other modes of transport The railway has experienced recent growth in both passengers and freight, but competition between different modes of transport remains intense. The factors influencing transport choices range from speed, cost, comfort and convenience, to safety and – increasingly – environmental impact. Many regard the relative safety of rail travel as one of its strengths.
4.1.1 Relative safety of travel on different transport modes From the user’s perspective, the risk from using a mode of transport can be assessed on the basis of fatalities per traveller kilometre.21 In theory, this allows him/her to compare the risk from undertaking the same journey using different modes.
Chart 11 compares the safety of different modes of transport relative to rail.22 For each mode, the figures represent the risk per traveller kilometre on a typical journey as a multiple of the risk associated with rail travel.
In summary:
• By far the most dangerous mode of popular transport is the motorcycle, with a fatality rate around 1,000 times greater than rail travel. • Car travel is more than 20 times more dangerous than making the same journey by rail.
21 The term traveller is used in this section to refer to a person travelling by any means (including on foot), with the exception of professional drivers and other public transport staff. It differs from the definition of a passenger that is used elsewhere in the ASPR, in that while people boarding and alighting public transport vehicles are classed as travellers, other people who may be intending to travel (for example, those waiting on station platforms, at bus stops, or in airport lounges) are not.
22 The risk associated with rail travel is estimated to be 0.10 fatalities per billion passenger kilometres. To put this in context, one could expect to make a daily return trip between London and Edinburgh for around 22,000 years before being fatally injured in an accident on the railway.
2007 Annual Safety Performance Report 43 Benchmarking
• Bus and coach travel is nine times safer than travel by car, but around three times more dangerous than travel on the mainline railway. • Commercial air travel and rail travel are estimated to have broadly similar risk levels. However, the risk estimate for aviation is subject to a substantial degree of uncertainty and should be viewed as providing only a rough indication of the true risk.
Chart 11. Traveller fatality risk for different transport modes (relative to rail)
Railway 1
Airline 1
Bus / coach 3
Car 27
Pedestrian 373
Cycle 337
Motorcycle 1125
0 200 400 600 800 1000 1200 Fatalities per billion traveller km as a multiple of rail
Sources: SRMv5.5 (RSSB) for the railway. Transport Trends 2007 (DfT) for road vehicles (three years’ accident data for bus and coach, one year for other sources). Various sources, including CAP 763 Aviation Safety Review 2005 (CAA), for aviation.
Rail
The risk estimate for rail travellers covers train accidents and individual accidents that occur on board trains, while boarding and alighting trains, or in falls from trains. To allow a like-for- like comparison with rival modes, other elements of individual risk, such as slips, trips and falls in the station, have been excluded. The estimate was derived from the accident risk as modelled by SRMv5.5. The SRM provides a more robust estimate of the underlying risk than accident statistics because it takes into account the expected frequency and consequence of rare multi-fatality accidents. At current usage levels, the risk of 0.10 fatalities per billion traveller km corresponds to less than five fatalities per year.23
23 Transport Trends 2007 also presents fatality rates for rail. The five-year average from 2001-2005 is around 0.23 fatalities per billion passenger km. This is higher than that derived from the SRM because: (i) some accidents that are included in the DfT figures, such as falls from the platform, have been excluded from the analysis in this chapter in an attempt to make a like-for-like comparison between different transport modes (someone run over while waiting at a bus stop would not be included in the bus and coach statistics); and (ii) fatality risk has reduced since 2001, for example as a result of the removal of Mark I rolling stock and the introduction of TPWS.
44 2007 Annual Safety Performance Report Benchmarking
Road
More than 3,000 people are killed each year in road traffic accidents. This reflects the widespread usage of road transport (which accounts for more than 90% of the total distance covered on journeys within Britain) as well as its safety. The volume of data means that fairly robust risk estimates can be derived from a single year’s accident statistics for travel by car, cycle, motorcycle and foot. Three years’ data was used for bus and coach travel because of the relatively small number of traveller fatalities on these modes. The estimated risk from bus and coach travel relative to rail has increased since the 2006 ASPR; most of this is the result of a relatively high number of fatalities (17) in 2006.
The risk estimates apply to an ‘average person’ making an ‘average journey’ by each mode. Car drivers, cyclists and pedestrians typically have more control over their destinies than travellers on trains and aeroplanes. Differences in risk levels can be seen in the accident statistics for different demographic groups. For example, per head of population, around five times as many 18 and 19-year-olds are killed in car accidents as those in the 40–59 age group. Likewise, some journeys are safer than others. Driving on motorways is around six times safer than driving on urban roads on a per km basis. Therefore, the relative advantage of rail over road reduces on motorway journeys (although not to the extent that it is safer to travel by car).
Air
Civil aviation in Britain has had a very good safety record in recent years. The last major loss of life occurred in January 1989, when 47 people died after a Boeing 737 crashed on the embankment of the M1 near Kegworth. However, the accident at Heathrow Airport on 17 January 2008, when a Boeing 777 lost thrust and made contact with the ground some 1,000 feet short of the runway, is a reminder that the risk has not been entirely eliminated. There were no fatalities among the 152 passengers and crew, but the aircraft was damaged beyond economic repair and there was a significant fuel leak.
The risk from air travel is dominated by accidents that are very rare but of potentially very high consequence, so safety cannot be satisfactorily estimated using historical data alone. The figure in the chart is based on worldwide accident rates, adjusted to account for the superior safety records of Europe, North America and the South-West Pacific and an improving trend over time. It represents the average risk on all flights made by Britons (both domestic and international). However, the figure must be treated with some caution. For example, weather and communication problems are two common causes of air accidents, but the relatively clement British weather and the widespread use of English in aviation (as well as the lack of high ground near airports and a greater use of landing aids) may render travel to and from British airports safer than the ‘developed world’ average. Conversely, shorter journeys (over which air and rail travel most obviously compete) are likely to carry more risk per kilometre because much of the risk from flying occurs during take-off and landing.
2007 Annual Safety Performance Report 45 Benchmarking
Alternative risk metrics Although risk per distance travelled is probably the best metric for comparing different modes of transport for most purposes, alternative measures include the risk per hour, or the risk per journey.
Table 5. Comparing fatality risk using different metrics
Fatality risk per billion traveller… km hours trips Railway 0.1 5 4 Airline 0.1 76 290 Bus / coach 0.363 Car 2.6 100 36 Cycle 32.0 380 120 Pedestrian 36.0 150 40 Motorcycle 110.0 4,300 1,900
Source: National Travel Survey (DfT) for average journey times and lengths (otherwise, same sources as Chart 11).
Table 5 presents a comparison of traveller fatality risk using the three metrics. Rail travel remains in either first or second place whichever metric is used. Although the figures for aviation are subject to a substantial degree of uncertainty and should be viewed only as being broadly indicative of true risk, air travel falls out of the top three when comparisons are made on a per journey or per hour basis because it is used to cover large distances quickly.
When a journey has to be made to a given destination, the risk per hour is not a sensible metric because it merely penalises the fastest modes of transport. However, the risk per journey or per hour might be a relevant metric to the leisure traveller, who will often select the destination and the mode of transport in tandem, taking into account such factors as time, cost and safety.
4.1.2 Whole-journey risk The safety of rail travellers before they arrive at the station and after they leave it is largely beyond the industry’s control. However, from the passenger’s perspective, the safety of the entire journey, from ‘door-to-door’, is important.
Most journeys by private transport do not involve additional ‘access stages’ (other than very short walks to and from the vehicle). However, a person travelling by public transport needs to get to and from the bus stop, railway station or airport, and these legs of the journey carry their own risk.
The average distance walked to or from stations in association with national rail journeys is 0.9km, and the average distance travelled to or from stations using the various modes of road transport (predominantly cars) is around 3km.24
24 Based on an analysis of National Travel Survey data by A W Evans and J D Addison as part of an EPSRC-funded research project on Interactions between rail and road safety (2005).
46 2007 Annual Safety Performance Report Benchmarking
Chart 12. Cumulative risk profile of a representative door-to-door journey
5 in 100 million
4 in 100 million
3 in 100 million 73%
2 in 100 million Cumulative fatalityrisk
<1%
1 in 100 million <1% 8%
18%
Car (3km) Boarding train On the train (45km) Alighting train Walk (0.9km)
Source: RSSB for rail, Transport Trends (DfT) for car and foot.
Chart 12 shows the risk profile for a typical journey with national rail as the main mode. It follows a person who leaves home, drives 3km to the station, takes the train for 45km and then walks 900m to the office. The y-axis shows the cumulative risk associated with the journey, with the contribution of each stage shown in colour (the percentages reflect the proportion of the whole-journey risk that each stage accounts for). Only the risk associated with transport accidents is included, so the risk from slips, trips and falls on pavements or in stations is not represented.
Most of the risk associated with the door-to-door journey occurs in the last (and shortest) stage: the walk to the office. Because pedestrians are a relatively high-risk group, it is likely that far more people are killed in road traffic accidents (including pedestrians struck by vehicles) on the way to and from the station than on the train itself.
When whole-journey risk is considered, the safety advantage that rail enjoys over car travel is reduced (although not to the extent that it would be safer to make the entire journey by road).
The analysis in this section also illustrates the importance of an integrated approach to transport planning. Travellers’ safety depends on the interfaces between the railway and other modes of transport, as well as the safety in stations and on-board trains.
4.1.3 Total risk associated with different transport modes A transport policy maker is likely to be interested in the costs and benefits that each mode of travel brings to society as a whole, as well as to its users. Again, safety is one of many important factors, but it is important to understand the extent to which accidents involving
2007 Annual Safety Performance Report 47 Benchmarking
cars affect not only car drivers and passengers, but also other road users, such as pedestrians, cyclists and motorcyclists.
Chart 13. Total risk for different transport modes
Including trespassers Fatalities Travellers using named mode 1.6 Pedestrians and other transport users Railway Major injuries 0.7 Travellers using named mode Excluding trespassers Pedestrians and other transport users
Airline 0.1
Bus / coach 5.6
Car 7.8
0123456789 Fatalities and weighted major injuries (per bn traveller km)
Source: RSSB for rail, Road Casualties in Great Britain (DfT) for road vehicles, and various sources for aviation, including CAP 763 Aviation Safety Review 2005 (CAA). One major injury has been given one-tenth the weighting of one fatality.
Chart 13 shows the risk (in terms of both fatalities and weighted major injuries per billion traveller km) to both travellers and other people from accidents involving four of the main modes of passenger transport. In summary:
• Buses and coaches present the highest risk to pedestrians and other transport users on a traveller km basis. These are heavy vehicles that often operate on busy streets. Bus and coach travellers also have a substantially higher risk from major injury than those travelling by rail. • Cars also kill and injure more people than trains (even when normalised by usage). Interactions between trains and people are restricted to level crossings and stations, whereas most public highways are open to pedestrians, cyclists and other road users. • When trespassers are included in the railway statistics, rail and car transport have similar third-party fatality risks, but rail has a much lower rate of major injuries. • Aviation poses little risk to third parties. Air crew make up the majority of third-party casualties, with a small number of injuries on the ground. The risk from air travel is dominated by fatality risk; major injuries make only a small contribution.
48 2007 Annual Safety Performance Report Benchmarking
4.1.4 Inter-modal safety trends Safety has improved on most modes of transport – and many other areas of life – over recent decades. Chart 14 shows long-term trends in fatality rates (normalised by usage) for rail and car travel.25
Chart 14. Historical safety trends for car and rail travel Rail traveller risk as a proportion of car and taxi traveller risk 12 0.4 Railways (main line and metros) 11 Cars and taxis 10 Relative safety of car and taxi vs train 9 0.3
8
7
6 0.2
5
4
3 0.1
2
Traveller fatality risk (per billion traveller km) traveller billion (per risk fatality Traveller 1
0 0 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005
Source: ORR for rail, DfT for road.
It is clear that there have been substantial long-term improvements in both forms of transport – and developments in medical care have also helped save lives. Nevertheless, it is interesting that car travel today is still less safe than rail travel was half a century ago. The dashed red line on the chart shows the relative safety of car travel and rail travel: the higher the point, the smaller rail’s relative safety advantage in that year.
• Car occupant risk reduced at a faster rate than rail safety between the late 1970s and the early 1990s, eroding the railway’s comparative safety advantage (although never getting close to parity). Among other changes, this period saw legislation mandating the wearing of seat belts (1982 for front seats) and a clampdown on drink-driving. • Since the early 1990s, the gap has widened again. Recent years have seen significant safety improvements on the railway, while the safety of car occupants has been improving at a much slower rate (around 1% per year).
25 The data plotted in the chart is somewhat different in scope to that presented in the earlier analyses in this section. Metros and modern tram systems are included, and the data covers all fatalities arising from train accidents and train movement accidents. The figure plotted is a five-year average and, to smooth the effect of high-consequence accidents further, the underlying risk from train accidents for each year has been estimated as the five-year average number of fatal accidents multiplied by the average consequence of fatal accidents over the 50-year period, 1957–2006.
2007 Annual Safety Performance Report 49 Benchmarking
While we might expect reductions in fatality and injury rates to become progressively harder to achieve, most transport modes show the capacity to keep improving. The railway will have to work hard to remain competitive as other modes of transport improve their services and safety levels.
4.2 International comparisons Countries across Europe submitted their first common safety indicators to the European Rail Agency in 2007. Once this process is established, the availability of safety statistics based on a consistent set of definitions will make it easier to compare the safety performance of different railway networks. Further discussion on European developments can be found in Chapter 13.
Eurostat and the Union Internationale des Chemins de Fer (UIC) have been collecting information on European rail accidents for some time. Although there are problems with data quality (such as some countries using national definitions), it is possible to make reasonable comparisons between safety levels in different countries.
Chart 15 shows the number of fatalities to passengers and rail staff over the three-year period 2004–06. The figures have been normalised by train kilometres to account for the different scales of European railways. The European Union (EU) is expanding, and average fatality rates are shown for both the EU 15, which comprises the member states prior to 1 May 2004, and the EU 25, which includes the 10 accession countries.26 The countries in the EU 15 are marked with asterisks on the chart. The GB figures presented in this section may be slightly different to those shown elsewhere in the ASPR because, for consistency, they exclude events that are outside the scope of Eurostat or UIC statistics.
Chart 15. Passenger and workforce fatalities on European railways
0.16 Workforce 0.14 Passengers
0.12
0.10
0.08
0.06
0.04 Fatalities per million train km 0.02
0.00 GB* Italy * Latvia EU 25 EU Poland Spain * Spain Estonia EU 15 * EU Hungary Ireland * Ireland Austria * Austria France * France Slovenia Finland * Finland Lithuania Greece * Sweden * Sweden Belgium * Portugal * Portugal Denmark * Denmark Germany * Germany Slovak Rep Czech Rep. Netherlands * Netherlands Luxembourg * Luxembourg Source: RSSB data for Great Britain, and Eurostat for other European countries – all based on the period 2004–06. * Members of the EU 15 are marked with an asterisk.
26 Malta and Cyprus have been omitted from the chart because they no longer have railways.
50 2007 Annual Safety Performance Report Benchmarking
Passenger and workforce fatality rates in Great Britain were lower than both EU averages over the three years. Britain’s recent safety record places it within a group that includes France, Belgium, the Netherlands and Germany. In general, countries in the northern and western parts of Europe have safer railways than those further south and east.
Chart 16 shows the trend in passenger fatality rates for GB and Europe (with the EU 15 and EU 25 shown separately). There have been significant safety improvements across Europe over the period plotted, and fatality rates are now a fairly small fraction of what they were 40 years ago. Safety performance in Britain appears to have improved at a broadly similar rate to the rest of Europe.
Chart 16. Passenger fatality rates on European railways 2.5 Europe EU-25 Moorgate (LUL) Clapham Junction EU-15 2.0 GB
1.5 Ladbroke Grove
1.0
0.5 GB figures include metro
passenger fatalities per billion passenger km passenger billion per fatalities passenger systems and trams until 1984
0.0 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source: RSSB, ORR and Transport Statistics Great Britain (DfT) for Great Britain; UIC and Eurostat for other countries.
British accidents with more then 30 fatalities have been labelled. The effect that one large accident can have on that year’s fatality rate is clear. The upturn in the all-Europe fatality rate in 2006 was the result of an accident at Bioče in Montenegro. A train derailed and fell into a ravine, killing 45 people.
The worst European accident in 2007 occurred on Sardinia, where three people died in a collision between a passenger train and a freight train on 15 June. Worldwide, there were a number of serious accidents, three of which resulted in more than 50 deaths (in India, Congo and Pakistan). It is more than 50 years since Britain witnessed an accident of this magnitude (Lewisham, 1957 – see section 8.4.1).
2007 Annual Safety Performance Report 51 Benchmarking
Table 6. International multi-fatality accidents in 2007 (5 deaths and above)
Place and Total Accident Reported cause Date Description of event country fatalities type (if known) Collision Hatay Province, A freight train struck a lorry carrying farm workers on a level Human error (road 04/01/2007 7 with road Turkey crossing. vehicle driver). vehicle. Jakarta, One carriage of a passenger train derailed and plunged into 16/01/2007 5 Derailment. Indonesia a dry river bed. Collision Khanh Hoa, A train struck a bus which had slammed through the barriers Human error (road 08/02/2007 13 with road Vietnam of a level crossing. vehicle driver). vehicle. Deewana, 19/02/2004 68 Explosion. The ‘Friendship’ express was subject to a bomb blast. Terror attack. India Collision Harare, A bus collided with a freight train at a level crossing. All Human error (road 06/03/2007 34 with road Zimbabwe victims were on board the bus. vehicle driver). vehicle. A passenger service struck a minibus at an unmanned level Collision Kanchipuram, crossing. It was reported that the minibus, which was driving Human error (road 16/04/2007 11 with road India along a parallel road, was trying to overtake the train when vehicle driver). vehicle. the accident occurred. Kerang, Collision A lorry struck a passenger train at an unprotected level Human error (road 05/06/2007 11 Aus tralia with road crossing. vehicle driver). Driver failed to A locomotive on trial collided with a commuter train travelling locate signal / 15/06/2007 Tali, Taiwan 5 Collision. in the opposite direction. equipment failure (ATP). Benaleka, A passenger train derailed; many of the dead had been 01/08/2007 100 Derailment. Congo riding on the roof. Rio de Janeiro, 30/08/2007 8 Collision. A commuter train struck an ECS formation. Brazil Collision Nr. Manzanillo, A bus was struck by a passenger service on an open level 06/10/2007 28 with road Cuba crossing, and subsequently plunged into a ravine. vehicle. Southern 06/10/2007 5 Derailment. Five carriages of a busy passenger train derailed. Bangladesh Collision Muridke, Human error (road 10/10/2007 13 with road A train struck a bus struck a train at a level crossing. Pakistan vehicle driver). vehicle. Collision North-East A van collided with a freight train on a level crossing. All the Human error (road 11/11/2007 8 with road China victims were road vehicle occupants. vehicle driver). vehicle. Collision Human error (road A passenger train stuck a school bus, which had crashed 14/12/2007 Northern India 18 with road vehicle driver and through barriers in fog, at a level crossing. vehicle. crossing keeper). Mehrabpur, Twelve carriages of a crowded express derailed before 19/12/2007 56 Derailment. Track defect. Pakistan plunging down an embankment into water. Collision A mini-bus was involved in a collision with a locomotive at a Human error (road 24/12/2007 Tajikistan 9 with road level crossing. vehicle driver). vehicle.
Table 6 lists the train accidents which resulted in five or more fatalities in 2007, from all regions of the world. Around half were caused by the drivers of road vehicles, showing that this problem is not confined to Britain. Railways differ in terms of infrastructure, rolling stock, working practices and the external hazards they are exposed to, but it is still possible to learn lessons from international events. They can show accident scenarios that are rare in Britain, identify possible vulnerabilities and indicate the potential for harm if effective controls are not maintained.
4.3 Occupational risk: comparisons with other industries Railway work is often regarded as relatively high-risk compared to other occupations. Track workers, shunters and fitters can be exposed to moving trains and unprotected electricity supplies as well as more general hazards associated with difficult site access and the
52 2007 Annual Safety Performance Report Benchmarking
physical nature of the work. Train drivers, on the other hand, are exposed to the risk from train accidents, and can be vulnerable when they need to venture trackside to change ends or inspect their trains.
The Health and Safety at Work etc Act 1974 (HSWA) requires employers to ensure, so far as is reasonably practicable, the health, safety and welfare at work of employees. This applies to all workers, regardless of the risk they face, but the priority and effort applied to analysing the contributors to risk and developing measures to mitigate risk should increase in line with the level of individual risk.
4.3.1 Industry risk Chart 17 compares the level of fatality risk for rail industry staff with those working in other higher-risk industries, based on three years of accident statistics.
Chart 17. Industry risk comparison
1 in 3,950 Mining of coal and lignite, extraction of peat 1 in 5,800 Recycling of scrap and waste 1 in 10,700 Agriculture, hunting, forestry & fishing 1 in 12,600 Sewage and refuse disposal 1 in 15,350 Other mining and quarrying 1 in 16,900 Extraction of crude petroleum & natural gas 1 in 23,000 All extractive and utility supply 1 in 25,400 All construction 1 in 28,000 Land transport & transport via pipelines 1 in 29,350 Manufacture of wood and wood products 1 in 30,000 Manufacture of non-metalic mineral products 1 in 34,100 All main line rail 1 in 45,000 Manufacture of rubber and plastic products 1 in 56,700 Renting of machinery, equipment & other goods 1 in 77,200 All manufacturing 1 in 126,400 All industries 1 in 152,500 Electricity, gas, steam & hot water supply 1 in 327,400 All services
0 5 10 15 20 25 30 Fatalities per 100,000 workers per year
Source: RSSB for railway statistics (based on the three-year period January 2005–December 2007) and Statistics of Fatal Injuries (HSC) for other industries (based on the three-year period April 2004–March 2007). Orange bars represent the five main sectors and the grey bar represents all UK industry.
The labels on the bars show the estimated probability that an individual will have a fatal accident at work in a given year. Some of the estimates are subject to a high degree of statistical uncertainty because they are based on a very small number of fatalities. The individual risk estimate for rail industry workers is substantially lower than the figure presented in the 2006 ASPR because 2004, which saw a relatively high number of workforce fatalities, has dropped out of the three-year period on which the statistics are based.
Overall, the risk to workers on the main line railway is around four times higher than the all- industry average. It is broadly similar to – and slightly lower than – the risk faced by those in the construction industry.
2007 Annual Safety Performance Report 53 Benchmarking
4.3.2 Occupational risk Within any industry, some occupations are exposed to a significantly higher risk than others. The estimates in Chart 17 represent the average risk for all staff, including people working in low-risk environments (such as offices, for example). It is useful, therefore, to look at the safety of specific occupations, especially those with a relatively high exposure to risk.
Chart 18 compares individual fatality risk for track workers, shunters and train drivers with comparator high-risk occupations in other industries. Again, some figures are subject to a large amount of statistical uncertainty. For example, the risk estimate for shunters might be very different if it were based on a different three-year period, and that for track workers is much lower than the figure presented in the 2006 ASPR.27
Chart 18. High-risk occupation comparison
1 in 23,100 Train driver 1 in 5,600 HGV driver 1 in 4,750 Van driver 1 in 47,200 Bus & coach driver
1 in 18,300 Track worker 1 in 3,000 Shunter 1 in 8,050 Road construction 1 in 11,650 Lines repair & cabling 1 in 18,500 Plant & machinery operator 1 in 9,000 Scaffolders, stagers, riggers 1 in 8,050 Farm manager 1 in 8,150 Agricultural machinery drivers 1 in 8,150 Farm worker 1 in 4,750 Forestry workers 1 in 4,850 Refuse & salvage occupations
1 in 126,400 All industries
0 5 10 15 20 25 30 35 Fatalities per 100,000 workers per year
Source: RSSB for railway statistics (based on the period January 2005–December 2007) and Statistics of Fatal Injuries (HSC) and Road Casualties Great Britain (DfT) for other industries (based on the period April 2004–March 2007).
Although no other jobs are exactly comparable with railway occupations, road construction has some similar causes of risk to track work, and professional bus and lorry drivers are exposed to some hazards that are similar to those faced by train drivers.
Train drivers have a lower level of risk than the drivers of heavy goods vehicles and vans, but a higher level of risk than professional bus and coach drivers. Track workers have a lower level of risk (based on the three-year average) than the comparator occupations shown on the chart.
27 For comparison, SRM v5 estimated individual fatality risks of one in 8,300 for track workers and one in 15,300 for train drivers. SRM estimates are usually more representative of the ‘underlying’ level of risk because the contribution of low-frequency, high-consequence events is included. Estimates of individual risk were not updated for version 5.5.
54 2007 Annual Safety Performance Report Benchmarking
Track work appears to be safer than road construction work, although if the SRM estimate is used rather than the estimate based on three years’ data, the two occupations looks very similar in terms of safety.
It is difficult to estimate the risk faced by shunters with a high level of accuracy, but it is clear that shunting is a high-risk occupation. RSSB published a special topic report on shunter safety in January 2008, which includes a more detailed analysis of occupational risk. This is available from the RSSB website: http://www.rssb.co.uk/safety/spr/spreports.asp.
4.4 Benchmarking within the industry It is useful for the companies that make up the British railway to be able to benchmark their own safety performance against the rest of the industry. This can help them to identify in which areas they are industry leaders, and in which areas they should focus on improving.
Of course, benchmarking can be fraught with difficulties. For example:
• Comparisons must be based on consistent definitions. • Incident rates can reflect company reporting regimes as well as the true level of safety. A small number of accident reports could reflect either strong safety performance, or a poor safety culture where few incidents are recorded. • Different companies might be expected to have different safety profiles for reasons beyond their control, such as the type of services they run and the areas in which they operate. • If the number of events is small, there is a large degree of statistical uncertainty in the estimated rate, and it may not be a true representation of real system safety.
In early 2008, RSSB began producing safety data profiles for train operating companies. The profile provided to each company shows how it compares with the national average for a range of around 30 safety measures. These measures relate to sub-categories of the following incident types:
• Assaults. • Personal accidents to members of the workforce. • Passenger injuries in stations, on trains and when boarding or alighting trains. • Train fires. • Trains running into objects or being hit by missiles. • Accidents and near misses at level crossings. The primary purpose of the profiles is to help companies benchmark their safety performance against others in the industry, identifying strengths and weaknesses. However, they should also generate thought and discussion around areas such as data quality and the comparability of different types of operation.
Chart 19 shows examples of internal benchmarking measures, using data from 2007. Train operating companies have been grouped fairly crudely into three sectors: long distance operators, regional operators and those operating in London and the South East. These
2007 Annual Safety Performance Report 55 Benchmarking
sectors (and the way in which companies have been assigned to them) follow the approach used by the ORR in National Rail Trends. Note that franchise changes over recent years have led to a blurring of the boundaries, and some companies’ operations span more than one sector.
Chart 19. Benchmarking safety performance within the industry
Passenger boarding and alighting injuries Passenger slips, trips and falls in stations 1.2 3.0
1.0 2.5
0.8 2.0
0.6 1.5
0.4 1.0
0.2 0.5
0.0 0.0 Per million passenger journeys passenger million Per journeys passenger million Per Long distance Regional London & SE Long distance Regional London & SE
Train fires Workforce assaults on trains 0.6 20 Lighter shades represent fires from train defects. Lighter shades represent 0.5 Darker shades represent fires from arson. 15 verbal abuse and threats. 0.4 Darker shades represent physical assaults. 0.3 10
0.2 5 0.1 Per million train miles Per million train miles 0.0 0 Long distance Regional London & SE Long distance Regional London & SE
The chart raises interesting questions. For example, why do operators in London and the South East have significantly lower rates of passenger injuries than those in the other sectors? Are the passengers that travel on their trains less accident-prone, or less likely to report accidents to a member of staff? Or is it related to the way they manage their stations, or record accident data? Similarly, is there a reason why these operators have higher rates of arson, but relatively few fires caused by train defects? Do differences between operators reflect their different passenger profiles, differences in operations and rolling stock, or differences in company reporting systems? Long-distance operators might expect to record more workforce assaults on trains because they tend to have more staff on board. This raises the question of whether ‘train miles’ is the most appropriate normaliser for this data set. But the ratio of verbal to physical assaults also varies markedly between the sectors: does this reflect a genuine difference or merely different reporting cultures?
RSSB is working with the industry to further develop individual safety data profiles, and to interpret what the profiles are showing. For further information, contact Paul Sizer, Safety Intelligence Delivery Manager at RSSB, on 020 7904 7496, or email [email protected].
56 2007 Annual Safety Performance Report Passenger safety
5 Passenger safety
For the purposes of this report, a passenger is any person on railway infrastructure who either intends to travel, is travelling or has travelled, regardless of whether or not he or she has a valid ticket. The exception is travellers who trespass or attempt to commit suicide. These people are classed as members of the public and are covered in Chapter 7.
2007 Headlines • 2007 saw the first passenger fatality in a train accident since 2004. This was a result of the Grayrigg accident (23 February), where one passenger was fatally injured.
• Eight other passengers died in individual incidents. • There were 233 passenger major injuries, which is more than 2006, but below previous years. • The total number of passenger FWI in 2007 was 42.5. This is 8% higher than 2006. • 76% of total passenger harm in 2007 occurred at stations and was mainly attributed to personal accidents while moving around the station. There was a significant decrease in the number of major injuries whilst moving around stations, compared with 2006. * Performance at a glance
Passenger fatalities Passenger major injuries
14 13 300 275 259 12 254 250 233 10 219 10 9 200 88 8 150 6 100 4
2 50
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 12% higher than 20062 6% higher than 2006 2 8% lower than 2003-06 average2 8% lower than 2003-06 average 2
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 57 Passenger safety
5.1 Passenger risk profile Although passenger risk and the risk from train accidents are strongly linked in the public mind, passengers are more likely to suffer harm from a wide range of other hazardous events, many of which are not particular to the railway, such as falling down stairs or tripping over.
Chart 20 shows the average annual injury risk arising from various types of incidents to which passengers are exposed. The highest source of FWI risk arises from moving around stations, and will typically be due to slip-, trip- and fall-type events that might occur in any public area. Assaults are estimated to be the next highest single cause of risk, although it must be noted that the figure of 8.5 FWI per year is a best attempt at estimating the true figure; data on passenger assaults is not normally reported to SMIS, is are recorded by other means, such as through the BTP. It is also true that assault is not a risk issue specific to the railway.
The next highest sources of FWI risk are specific to the railway, and comprise accidents whilst boarding and alighting trains, incidents involving the platform edge, and train accidents. Train accidents comprise 8% of the total passenger risk, measured in FWI, but a higher proportion (28%) of the passenger fatality risk.
Chart 20. Passenger risk by accident type
Fall or jump from train / leaning 0.1 out of train Fatal Major Minor Shock/trauma
Other 0.3
On board injury 3.4
Other platform edge 4.2
Train accident 4.2
Boarding / alighting Type of passenger accident passenger Type of 6.3
Assault 8.5
Moving around stations 25.9 (including foot crossings)
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 Average risk per year Source: SRMv5.5.
Chart 21 shows passenger risk broken down by the SSP risk area group from which it originates. The greatest level of harm to passengers is identified by the SSP as arising from passengers, with the next greatest level identified as arising from the workforce. Most passenger harm comes from accidents whilst moving around stations (for example, slips, trips and falls), boarding and alighting accidents, and accidents involving the platform edge. The SSP assumes that many these types of incidents are equally under the control of the
58 2007 Annual Safety Performance Report Passenger safety
passengers themselves and the workforce (who have responsibilities for maintaining a safe station environment, and for train despatch).
The next highest level of risk arises from the area that the SSP describes as public behaviour. Most of the contribution from this risk area group is from assaults. Even if these are committed by passengers, the SSP describes the risk as arising from public behaviour.
Chart 21. Passenger risk by SSP risk area group (source of risk*)
Environment 0.39 Fatal Major Minor Shock/trauma
Engineering 3.04
Public behaviour 10.49
Workforce 15.99 SSP area fromSSP area which passengerarisesrisk Passengers 22.95
0 2 4 6 8 101214161820222426 Average risk per year Source: SRMv5.5. * The risk area groups relate to sources of risk so, for example, the bar corresponding the Passengers risk area group shows the risk that arises from passenger behaviour, rather than the risk to passengers.
5.2 Passenger fatalities and injuries in 2007 Fatalities On 23 February, a passenger was killed when a passenger train derailed at Grayrigg in Cumbria (London North Western). The immediate cause of the derailment was deemed to be the stretcher bar arrangement at a set of points, which resulted in a loss of gauge integrity at the switch blade.
Eight other passengers died in separate incidents:
• On 24 January, a passenger was struck and fatally injured at Wokingham manually controlled barrier (MCB) level crossing (South East). The young male climbed over the barriers to join a train in the Up platform and was struck by a Down service. • On 25 January, a child was struck by a train at Treorchy station (Western). The 15-year- old boy was leaning over the platform edge wearing earphones and failed to see or hear the train’s approach. • On 13 February, a passenger fell between the train and the platform whilst alighting at Haddenham & Thame Parkway station (London North West).
2007 Annual Safety Performance Report 59 Passenger safety
• On 23 February, a male was found with head injuries beside the line near Little Bowden (London North East). The circumstances behind the fatality have yet to be determined, but the injuries appear to be consistent with a person falling or jumping from a train. • On 7 November, a passenger was struck by a train approaching the platform at Cambuslang station (Scotland) when attempting to flag it down. The passenger was reported to be under the influence of alcohol at the time of the incident. • On 21 November, a passenger reported to be under the influence of alcohol collapsed and fell from the platform at Glengarnock station (Scotland). As he tried to climb back up, he was struck by a passing freight train. • On 24 November, a passenger fell from a train near Llansamlet (Western). Reports suggest that the passenger may have jumped out of the HST window, but police investigations are ongoing. • On 15 December, a passenger was stabbed on the platform at Dewsbury station (London North East). Two men, aged 17 and 18, were later charged with murder. Major injuries There were 233 passenger major injuries in 2007; 77% occurred at stations, of which the most common cause was slips, trips and falls, particularly on stairs. Boarding and alighting accidents were the next greatest cause of major injuries.
The derailment at Grayrigg caused 55 on-board passenger injuries, of which 28 were major.
5.3 Passenger safety trends Chart 22 presents the levels of passenger harm since 2003. The total passenger harm increased slightly in 2007, compared with the previous year.
Chart 22. Passenger fatalities and weighted injuries
Shock/trauma 50 48.8 Minor 47.2 0.2 0.3 Major 45 Fatal 9.7 42.9 42.5 9.4 0.2 39.5 0.3 40 0.3 9.2 9.9 35 9.2
30
25.9 25 27.5
20 25.4 23.3 21.9
15 Fatalities and weighted injuries weighted and Fatalities 10 13 5 10 889
0 2003 2004 2005 2006 2007
60 2007 Annual Safety Performance Report Passenger safety
The rise in harm is mainly a result of the accident at Grayrigg, which resulted in total passenger harm of 4.09 FWI. Weighted major injuries have dominated the total passenger harm in most years, but 2007 saw the largest volume of passenger harm from minor injuries in the five years from 2003. The introduction of revised weighting for minor injuries has reduced the contribution of minor injuries to the total passenger harm when compared with previous safety performance reports.
5.3.1 Trends in passenger fatalities Train accidents, which are low-frequency, high-consequence events account for 8% of the risk to passengers. It therefore makes sense to look at the passenger risk over a period longer than one year. For example, although no passengers were killed in train accidents in 2006, the risk to passengers from this source was not zero, as the subsequent Grayrigg derailment demonstrated.
It also makes sense to normalise accident data by the number of passenger journeys undertaken, since changes in passenger usage will affect the total number of fatalities and injuries, without necessarily affecting any passenger’s individual safety level.
Chart 23 presents trends in passenger fatality rates over the longer term. The overall annual rate of passenger fatality is currently just under 10 fatalities per billion passenger journeys.28
Chart 23. Trends in the underlying passenger fatality rate
Fatalities in train accidents (10-year moving average) 20 Fatalities in individual accidents (5-year moving average) 18
16 Including benefits 14 of TPWS
12
10 Reduction due to accident at Increase due to accident 8 Clapham Junction (1988) at Ladbroke Grove falling out of the 10-year (October 1999) 6 average.
4 Rate per billion passenger journeys passenger billion per Rate 2
0 Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Mar Mar Mar Mar Mar Mar Mar Mar Mar Mar Nov Nov Nov Nov Nov Nov Nov Nov Nov Nov May May May May May May May May May May Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
The chart shows the reduction in fatality risk brought about by the introduction of TPWS. This was achieved by excluding all fatalities from train accidents that would have been
28 This overall rate is based on the rate of fatality from train accidents averaged over 10 years, and the rate of fatality in individual accidents (such as falls from the platform) averaged over five years. It is higher than the estimate presented in Chapter 4, which excludes the risk from ‘non-movement’ accidents, such as slips, trips or falls.
2007 Annual Safety Performance Report 61 Passenger safety
prevented by TPWS from the calculation of the 10-year moving average from 2004 onwards. The TPWS fitment programme was completed at the end of 2003.
Chart 24 presents the number of passenger fatalities since 1998, split by accident type. Because they are low-frequency but potentially high-consequence events, the actual number of fatalities in train accidents per year can differ greatly from the SRM estimate of just above three. It is possible for single train accidents to result in many fatalities, as at Ladbroke Grove in 1999; conversely, there have been a number of years with no train accident fatalities.
Chart 24. Passenger fatalities by accident type
32 Train accident 30 29 28 Moving around stations (including foot crossings)
26 Other platform edge 24 Boarding / alighting 22 20 19 Fall or jump from train / leaning out of train 18 Assault 4 16 14 12 12 12 1 8 1 10 10 3 1 8888 8 6 Number of passenger fatalities 66 6 1 6 5 3 5 4 3 3 4 7 5 3 4 2 4 2 2 3 1 3 1 2 2 1 2 1 2 2211 1 3 2 0 1 11 1 1 1
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
For fatalities other than those due to train accidents, there is no clear trend in the numbers since 1998. The chart shows that 2007 saw the first fatalities from leaning out of or falling from moving trains since 2003. There were two fatalities in this category. These were isolated incidents and the risk associated with falls from moving trains has reduced since the early part of the decade, largely as a result of the removal of Mark I (slam-door) rolling stock.
5.3.2 Trends in passenger injuries Chart 25 presents the annual moving average of passenger weighted injuries, normalised by the number of passenger journeys made.
The chart shows that there has been a clear downward trend in passenger harm from major injuries since 2003. Over the same period of time, the passenger harm from shock/trauma and minor injuries has remained relatively constant.
62 2007 Annual Safety Performance Report Passenger safety
Chart 25. Trends in weighted passenger injury rates (annual moving average)
0.045 Shock/trauma 0.040 Minor 0.035 Major 0.030
0.025
0.020
0.015 Grayrigg 0.010
0.005 FWI per million passenger journeys passenger million per FWI 0.000 Jul Jul Jul Jul Jul Apr Apr Apr Apr Apr Oct Oct Oct Oct Oct Jun Jun Jan Jun Jan Jun Jan Jun Jan Jan Aug Feb Mar Aug Feb Mar Aug Feb Mar Aug Feb Mar Aug Feb Mar Sep Nov Dec Sep Nov Dec Sep Nov Dec Sep Nov Dec Sep Nov Dec May May May May May 2003 2004 2005 2006 2007
The figures at the end of 2007 equate to an average of around:
• One fatality per 100 million passenger journeys (see Chart 23). • One major injury per 4.8 million passenger journeys. • One minor injury (resulting in hospital treatment) per 550,000 passenger journeys. • One minor injury (not resulting in hospital treatment) per 35,000 passenger journeys.
Chart 26. Passenger major injuries by accident type
Moving around stations (including foot crossings) Other platform edge Boarding / alighting On board injury Fall or jump from train / leaning out of train Assault 300 Train accident 275 12 259 2 254 250 10 1 28 5 9 233 4 21 19 219 1 29 10 54 38 200 46 26 9 10 21 11 19 11 12 34 150 14
100 168 177 Number of major injuries 162 151 126 50
0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 63 Passenger safety
Chart 26 shows the numbers of passenger major injuries occurring for each of the past five years, broken down by accident type.
The drop in the number of passenger major injuries seen in 2006 has not continued into 2007. This is mainly due to the higher contribution from train accidents, with 28 major injuries as a result of Grayrigg.29 Other key points to note include the following:
• The number of major injuries sustained when boarding and alighting trains rose significantly compared with 2006, although it remains lower than in earlier years. This accident category includes falls between the train and the platform and getting caught in train doors. Most of the injuries in 2007 were sustained by passengers alighting from trains who missed their footing or tripped over objects. More detailed analysis of this subject can be found in Chapter 11. • 2007 saw a significant decrease in the number of major injuries from individual accidents (ie, those not involving train accidents), when compared with the average for the previous four years. This is due to a reduction in major incidents involving movement around the station, particularly slips, trips and falls. • Though not statistically significant, the number of major injuries resulting from platform- edge incidents increased in 2007, to its highest levels in the five-year period. Half of the accidents involved intoxicated passengers, but other accidents included passengers standing too close the platform edge due to ill health30 or losing their balance. The consequences of intoxication are discussed further in section 5.5.1.
5.4 Passenger safety by location 5.4.1 Passengers in stations Seventy-six per cent of the total harm to passengers occurred in stations in 2007. This included six fatalities and 179 major injuries. In general, there has been a decrease in total passenger harm at stations in the five years from 2003 from 40.7 FWI to 32.4 FWI.
• In 2007, 55% of the total passenger harm at stations was a result of slips, trips and falls, many of which occurred on stairs and escalators. Another key contribution to passenger harm came from boarding and alighting, which contributed 19% of the total passenger harm at stations. • Falls from or standing too close to the platform edge contributed to about 14% of the passenger harm at stations. In 2007, there were two fatalities in this category, one of which involved a passenger reported to be intoxicated. See Chapter 11 for further information on station safety.
5.4.2 Passengers on trains Train accidents
The SRM estimates that the annual passenger fatality risk from train accidents lies at just above three. In 2007, there was one passenger fatality and 29 major injuries as a result of train accidents, with all but one of the major injuries occurring at Grayrigg. The year saw
29 The 29th major injury in train accidents was due to a train striking a tree, which smashed through the bodyside window and caused a passenger to become unconscious. 30 One passenger was struck by a train while leaning over the platform edge to vomit, another fell onto the track after fainting on the platform.
64 2007 Annual Safety Performance Report Passenger safety
seven passenger train derailments in total, mostly caused by landslips and/or trees on the line.
In addition, there were six reportable collisions between trains in 2007 – all at low speed. There were also three reportable buffer stop collisions at stations; two involved passenger trains, one involved an ECS formation. There were no reported major injuries in any of these accidents.
For more information on train accidents, please refer to Chapter 8.
Falls or jumps from trains
Accidents involving passengers falling, jumping or leaning out of trains caused at least one fatality in most years prior to the complete phasing out of Mark I rolling stock. Last year (2007) saw the first two accidents since these withdrawals were made (see Chart 24). Both occurred on HSTs in motion. HSTs utilise Mark III coaching stock, the doors of which are centrally locked, but which also have sprung droplights (out of which it is possible to climb).
Other accidents on trains
Chart 27. Trends in on-board passenger harm Slip, trip or fall Injuries due to train movement Overheating and train defects (including sharp objects) Caught by internal train doors (either slam or power) Struck by object (including luggage) Others Assault 6 5.7
5.1 5 1.7 0.6
0.3 4.2 4 3.9 0.7 0.6 3.6 0.4 0.6 0.5 0.4 0.3 0.7 0.3 3 0.4 0.3 0.5 0.4 0.3 0.4 0.6 0.7 0.2 0.6 2 1.0 0.6 1.1 0.9 0.8
Fatalities and weighted injuries Fatalities 0.6 0.8 1 0.3
1.1 1.3 0.9 0.8 1.0 0 2003 2004 2005 2006 2007
Chart does not include injuries from train accidents or falls from train/leaning out of train.
In 2007, just below 10% of passenger harm on the railway was attributed to personal accidents on-board trains. Chart 27 gives a breakdown of the types of passenger accidents that have occurred on board trains (excluding train accidents). It is clear that there is no single dominant accident type.
2007 Annual Safety Performance Report 65 Passenger safety
Excluding train accidents, the last on-board fatality occurred when a passenger was stabbed on a train in 2006.31 This was classed as ‘assault’ in Chart 27. Notable points relating to 2007 include:
• The harm due to train movement reached its lowest level in the five-year period. The majority of these injuries involve passengers striking parts of the train as a result of trains lurching. It should be noted, however, that a proportion of other accidents, particularly slips, trips and falls and being struck by objects, may also be a result of train movement (though this is often difficult to confirm). External factors such as the track condition will affect the quality of the ride; ongoing work is being conducted to establish the optimum track quality.32 • There has been an increase in passenger injuries resulting from overheating and train defects (including sharp objects). Four out of the six major injuries in this category were to passengers who fainted due to overheating, either because the heating and ventilation systems were defective, or because of crowding. This increase seen in 2007, albeit small, highlights the importance of ensuring that there is adequate ventilation in the saloon areas in order to offset uncomfortable crowding conditions.
5.5 Factors affecting passenger risk An increasingly modern and sustainable railway has meant that a greater number of people are choosing rail as their preferred choice of transport. The railway faces challenges to ensure that it can cope with the increased volume and diversity of its users. Some of the factors affecting passenger safety are related to demographics, and examples of these factors include the following:
• The ageing passenger population and consequent issues related to reduced mobility. • The diversity of the population, leading to issues related to clarity of signage and customer information systems. • The improved interfaces at international terminals and issues related to increased baggage on the railways. It is difficult to quantify the impact of these factors on passenger safety, chiefly because the details often go unreported. By looking at injuries by age, gender, time of day and whether intoxication was involved, however, one can begin to build a picture of where the risk lies and how passengers are affected.
Chart 28 presents the variation in passenger injury numbers by time of day and day of week, (not normalised by passenger numbers travelling). It reveals the following key points:
• As expected, Sunday has the smallest number of passenger injuries of any day of the week. This is followed by Saturday, but the latter is more comparable with weekdays. • Most passenger incidents relate to movement around the station. As is to be expected, there are more injuries during the peaks, owing to greater passenger numbers at these times. The injury profile at the peaks is, however, much less defined in Chart 28 than the usage profile would be. This indicates that off-peak passengers may suffer a higher
31 A 19-year-old student was fatally wounded near Oxenholme (LNW) on 27 May 2006 while travelling on a Glasgow–Paignton service. A man was later convicted of murder. 32 Further information on the Vehicle/Track Systems Interface Committee is available at http://www.rssb.co.uk/sysint/sic/vt_sic.asp.
66 2007 Annual Safety Performance Report Passenger safety
incidence of injuries per journey (or at least they may have a higher propensity to report them than regular commuters travelling at peak times). • Injuries from assaults are most likely to occur in the evenings and appear to become progressively more frequent later in the week. Friday and Saturday nights have the highest number of incidents.
Chart 28. Passenger injury profile throughout the week (2003–2007)
Assault On board injury 400 Boarding / alighting Other platform edge Moving around stations (including foot crossings) 350
300
250
200
150 Number of incidents 100
50
0 00:00 - 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 01:00 - 00:00 07:00 - 06:00 13:00 - 12:00 19:00 - 18:00 Sunday Monday Tuesday Wednesday Thursday Friday Saturday
Note: Chart is not normalised by passenger numbers.
5.5.1 Alcohol and drugs A number of passenger fatalities and major injuries occur as a result of intoxication. Chart 29 shows the proportion of harm involving intoxication for different accident types.
The effect of intoxication is most prevalent for platform-edge incidents. Such incidents are relatively rare, but can have serious consequences, as evidenced by the two passenger fatalities attributed to this cause in 2007.
What is perhaps surprising, however, is that alcohol and drugs apparently contribute only 7% of the total harm resulting from assaults. This may be due to the amount of information recorded in relation to these incidents, since the analysis is based on whether intoxication (or the presence of alcohol or drugs) is mentioned in the description of the event. The peaks in assaults on Friday and Saturday evenings – when intoxication is most common – suggests that the true figure may be more than 7%. More detailed analysis of assaults is available in Chapter 10, Personal security.
2007 Annual Safety Performance Report 67 Passenger safety
Chart 29. Passenger harm involving intoxication by accident type (2003–2007)
100
90 Not intoxicated 80 38 Intoxicated 70
60 86 84 93 94 50
40
30 62
20 Proportion of harm by accident type (%) type accident by harm of Proportion 10 14 16 7 6 0 Moving around Boarding / alighting Other platform Assault On board injury stations (including edge foot crossings)
The classification of whether or not an accident involved intoxication is based on a word search of the narrative recorded in SMIS. Because some narratives provide little information about the circumstances surrounding the accident, the chart is likely to underestimate the true incidence of intoxication.
Further analysis of the relationship between passenger harm and intoxication is shown in Chart 30, which presents the proportion of passenger harm that results from intoxication at different times of the day.
Chart 30. Passenger harm involving intoxication by time of day (2003 – 2007)
100
90
80
70
60
50
40
30 56% 44% 46% 20 41% 32% 24% 25% 10 24% 23% 14%13% 16% 10% 2% 11% 9% 9% 12% 0% 1% 1% 4% 5% 4% 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 :0 :0 :0 :0 9:0 2:0 5:0 8:0
Proportion injuries involving intoxication by time of day (%) of by time intoxication involving injuries Proportion -07:0 -10:0 -13:0 -16:0 -20:0 -23:0 02:0 05:0 0 0-0 0 0-1 0 0-1 0 0-1 0-19 0 0-22 0 0-01 0- 0-04 0- 8:0 1:0 4:0 7:0 8:0 1:0 0:0 3:0 06:0 07:00-08:000 09:0 10:00-11:001 12:0 13:00-14:001 15:0 16:00-17:001 1 19:0 20:00-21:002 22:0 23:00-00:000 01:0 02:00-03:000 04:0 05:00-06:00 The classification of whether or not an accident involved intoxication is based on a word search of the narrative recorded in SMIS. Because some narratives provide little information about the circumstances surrounding the accident, the chart is likely to underestimate the true incidence of intoxication.
68 2007 Annual Safety Performance Report Passenger safety
It is clear from Chart 30 that intoxication has a marked impact on passenger safety from 19:00 onwards, reaching a peak between 22:00 and 23:00, when the majority of passenger harm is due to intoxication. These times are consistent with usual social drinking hours.
5.5.2 Age
Chart 31. Passenger harm by age group (2003–2007)
Less than 16 0.20 16-20 years 21-30 years 31-50 years 51-70 years Over 70 years 0.15
0.10
0.05 FWI per million passenger journeys passenger million per FWI
0.00 Moving around Boarding / Other platform Assault On board injury Grand Total stations (including alighting edge foot crossings)
This chart is based on accidents where victim’s age was recorded (comprising approximately 50% of all FWI) extrapolated to estimate the total passenger harm from each cause over the five years. Source: Age proportions from DfT National Rail Travel Survey 2007 (provisional).
Chart 31 presents the rate of passenger harm by age for different types of accident. It takes into account the proportion of passengers in each age group; 67% are aged 21-50 years – many of them will be regular commuters. Only 3% are over 70 years and the same proportion is under 16.
• Elderly passengers are the most susceptible to harm on the railway. They are subject to particularly high risk when boarding or alighting trains and moving around stations. Reasons for this might include impaired mobility and a lack of familiarity with rail travel. The results may also indicate that leisure passengers are better at reporting injuries than time-pressed commuters and business passengers, or indeed, that reporting of age (into SMIS) is better for injuries affecting people at either end of the age spectrum. • Passengers aged 21–50 are susceptible to the least proportion of harm, compared with other age groups. The majority of this group are commuters. • Passengers under 16 are susceptible to harm on board trains and moving around stations, while passengers between 16–20 years are susceptible to harm from platform- edge incidents and assaults. The changing population demographic means that people are living longer. In order to improve safety for the elderly in the future, much work is being done to improve the station
2007 Annual Safety Performance Report 69 Passenger safety
environment and provide for step-free access. Some of these initiatives are discussed in section 2.5 (Chapter 2, Safety overview).
5.5.3 Gender
Chart 32. Passenger harm by gender (2003–2007)
0.045 Female Male 0.040 Darker shades represent the contribution of harm from fatalities. 0.035
0.030
0.025
0.020
0.015
0.010 FWI per million passenger journeys passenger million per FWI 0.005
0.000 Moving around Boarding/alighting Other platform edge Assault On board injury Grand Total stations (including foot crossings)
This chart is based on accidents where victim’s gender was recorded (comprising approximately 90% of all FWI) extrapolated to estimate the total passenger harm from each cause over the five years. Source: Gender proportions available from DfT National Rail Travel Survey 2007 (provisional).
Chart 32 presents the rate of passenger harm by gender for different types of accident. It takes into account the proportion of rail passengers in each gender; roughly 53% are male and 47% female.
Over the five-year period (2003–2007), there were twice as many male passenger fatalities as there were female ones, but females have slightly higher levels of total harm. This implies that females were involved in (or report) a higher number of lower-severity accidents than males.
The chart also shows that:
• Female passengers are more susceptible to harm from moving around the station, boarding and alighting and on-board train injuries. This is similar to the profile for elderly passengers, detailed in section 5.5.2, and it may be because a relatively high proportion of passengers over 70 years are female. The types of accident to which females are more susceptible also indicate that footwear (eg, heels and sandals) may be a factor. • Males are relatively more susceptible to harm from platform-edge incidents and assaults.
70 2007 Annual Safety Performance Report Workforce safety
6 Workforce safety
A person is classed as a member of the workforce if he or she is working for the industry on railway operations, either as a direct employee or under contract. Accident and injury data is collected in SMIS on all events occurring at stations or on NRMI. Fatalities occurring off NRMI but during working time (for example, whilst in depots or sidings, or as a result of a road traffic accident) are also included. Major or minor injuries occurring similarly are, however, not included.
This chapter investigates the range of risk that impinges on a wide variety of staff occupations, from track workers to train drivers, train crew to station staff and shunters.
2007 Headlines • Two members of the workforce were killed in accidents during 2007. Both were track workers struck by trains in a trackside environment. • The long-term monthly moving average rate of train driver fatalities is at an historically low level; the long-term monthly moving average rate of track worker fatalities continued its recent downward trend during 2007. • There were 128 workforce major injuries in the year, lower than each of the past four years. Two of these were serious injuries to track workers who were struck by trains. • There has been a downward trend in total train-crew FWI, which now stands at around two-thirds of the 2004 level. Performance at a glance*
Workforce fatalities Workforce major injuries 10 250 9 219
8 200 182 155 6 150 130 128 5
4 3 100 22 2 50
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 PerformanceSignificant change? 2007 Performance Significant change? Same as 2006 2 2% lower than 2006 2 58% lower than 2003-20062 25% lower than 2003-2006 33
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 71 Workforce safety
6.1 Workforce risk profile Working on the railway covers a wide range of occupations and activities that involve a variety of hazards, some of which are particular to the rail industry, others of which are not.
Chart 33 shows that most of the fatality risk arises from workers being struck by trains. As well as infrastructure maintenance personnel, other staff can be exposed to moving trains. Train drivers, for example, sometimes have occasion to be in the trackside environment to use a signal post telephone or change driving ends of a train.
The greatest FWI risk arises from slips, trips and falls, and contact with objects. These accidents are rarely fatal, but they dominate the major injury statistics. The category ‘contact with objects’ includes accidents when using tools, dropping equipment onto feet, and striking parts of the body against fixed objects.
Chart 33. Workforce risk by accident type
RTA 0.2 Fatal Major Minor Shock/trauma
Manual handling 1.1
Sudden train movement 1.1
Electric shock 1.4
Train accident 1.6
Boarding / alighting 1.9
Other 2.6
Struck by train 2.7 Type of workforce accident Assault 3.5
Contact with object 8.1
Slips, trips and falls 8.8
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Average risk per year
Source: SRMv5.5.
Chart 34 shows workforce risk again, this time broken down by the SSP area from which it arises. The SSP identifies the greatest level of risk as arising from actions by workers themselves. This would cover most personal injuries, such as slips, trips and falls or contact with objects. It would also cover some types of train accidents, where the cause is identified as being due to workforce error, such as (some) SPADs or misrouting events.
A further notable proportion of risk arises from the SSP area of public behaviour. This includes assaults on staff, but also covers other incidents, such as train collisions with objects placed on the track by vandals.
72 2007 Annual Safety Performance Report Workforce safety
Chart 34. Workforce risk by SSP area group (source of risk*)
Fatal Major Minor Shock/trauma Passengers 0.04
Environment 0.05
Engineering 1.99
Public behaviour 5.16 SSP area from which workforce risk arises from which workforce SSP area Workforce 25.86
0 2 4 6 8 10121416182022242628 Average risk per year
Source: SRMv5.5. * The risk area groups relate to sources of risk so, for example, the bar corresponding the Passengers risk area group shows the risk that arises from passenger behaviour, rather than the risk to passengers.
Chart 35 shows workforce profile again, broken down this time by the type of worker to whom the harm occurs. This information is based on SMIS data over the past five years, rather than SRM data, since the SRM precursor breakdown does not allow this profile to be easily generated.
Chart 35. Workforce harm by worker type (2003 to 2007)
Shunter 0.50 Fatal Major Minor Shock/trauma
Other 1.58
Train driver 3.86
Station staff 4.31
Workforce risk by worker type Other on-board train crew 7.43
Trackworker 15.00
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 Average FWI per year
Source: SMIS.
2007 Annual Safety Performance Report 73 Workforce safety
The chart shows that the largest proportion of both fatalities and injuries occur to track workers. On-board train crew (not including drivers) and station staff account for the next highest FWI levels, although neither group has suffered a fatality during the five-year period.
Shunters make up a small proportion of the total FWI shown, and most of this proportion is made up of fatalities; there is no requirement for non-fatal shunter injuries to be entered into SMIS. Despite accounting for a low proportion of the total FWI, this does not mean that shunting is a low risk occupation because the population of shunters is relatively small. Indeed, the opposite is true (see section 4.3).
6.2 Workforce fatalities and injuries in 2007 There were two workforce fatalities in 2007:
• On 29 April, a track worker was struck and killed by an ECS formation at Ruscombe Junction (Western). It is reported that he failed to respond to warning horns from the power car and two lookouts. • On 29 November, a track worker was struck by a train at Kennet Bridge near Reading (Western). The incident occurred after the possession was given up in the early hours of the morning. The Rail Accident Investigation Branch launched investigations into both accidents. That into the Ruscombe fatality has now been published and can be downloaded from the RAIB website: www.raib.gov.uk.
Of the 128 major injuries, most involved slips, trips or falls about the track. More details of some of the incidents can be found in section 6.4.
6.3 Workforce safety trends 6.3.1 Trends in workforce fatalities There was a significant upward trend in the track worker long-term fatality rate from 1998 to 2004, although there has been a reduction in the rate in the last two years. Much of the increase in 2004 can be attributed to the multi-fatality events at Tebay and Hednesford. However, there has been a significant downward trend in the long-term rate of train driver fatalities since 1998. As one might expect, the reduction in train accidents has had a positive effect on this trend.
74 2007 Annual Safety Performance Report Workforce safety
Chart 36. Trends in the underlying workforce fatality rate
Train driver (10-year moving average) 0.35 Track worker (5-year moving average)
0.3
0.25
0.2
0.15
Including benefits of TPWS 0.1
0.05 Rate of fatalities per month per 20,000 employees 20,000 per month per fatalities of Rate
0 1998 1999 2000 2001 2002 2003 2004 2005 200620072007 From 2004 onwards train driver fatalities in accidents that could have been prevented by TPWS have been excluded from the trend, since risk from these accidents has been mitigated by the TPWS system.
Chart 37 shows the fatalities that have befallen members of the workforce since 1998. The two track workers killed in 2007 were both struck by trains.
Chart 37. Workforce fatalities by occupation
12 Station staff 11 Other on-board train crew Train driver Other 10 Shunter 9 Track worker
8
6 6 55 5
Fatalities per year 4 4 3
22 2
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 75 Workforce safety
The chart shows that:
• There have been at least two workforce fatalities every year since 1997 (when there were none). • Most workforce fatalities occur to track workers (there have been 31 in the last 10 years). • 2004 was the worst year for track worker fatalities in the analysis period, with two multi- fatality events: Tebay (four track workers) and Hednesford (two track workers). There is no significant trending behaviour in total workforce fatalities over the 10-year period, although the last two years have seen the lowest levels in recent times. Table 7 shows the numbers of workforce fatalities by location and occupation over the past decade.
Table 7. Workforce fatalities by location and occupation 1998–2007
Travelling Trackside location Worksite - not trackside to site Total Travelling between sites Running line In depots In sidings pary (3rd owned) (NRMI) sidings In - infrastructure railway On not trackside On trains In stations Track maintainence 1 17 1 19 Civil structure inspection 1 1 S&T renewal/upgrade 1 1 Engineering Supervisor 1 1 CoSS 2 2 Hand Signaller 2 2 Track workers Track Lookout 4 4 Machine operator 1 1 Train drivers 2 5 7 Onboard train crew 1 1 Train
operators Train guards 1 1 Station staff 1 1 Station cleaners 1 1 Shunters 1 2 1 4 Other Non-railway personnel delivering to site 1 1 Fitter/MOM 2 2 1 5 Total 43033119152
Clearly, the majority of fatalities occur to track maintenance staff on or about the running line. This is a consequence of the high number of employees in this group and their exposure to a high-risk environment. Lookouts are also an ‘at risk’ group.
76 2007 Annual Safety Performance Report Workforce safety
6.3.2 Trends in workforce injuries
Chart 38. Workforce major injuries by occupation
Station staff 250 Other on-board train crew 219 Train driver Other 200 Shunter 182 Track worker
155 150 130 128
100 Major injuries per year per injuries Major
50
0 2003 2004 2005 2006 2007
Note that 2007 saw significantly fewer workforce major injuries than the average over the previous four years. Indeed, there has been a significant downward trend in all major injuries since 2004.
The number of track worker and other on-board train crew major injuries in 2007 is significantly lower than the average of the previous four years, while train driver major injuries were marginally higher than both the 2006 figure and the average of the previous four years. Track workers still account for more than half of workforce major injuries.
In 2007, there was only one shunter major injury within the scope of this report. Chart 38 excludes most of the major injuries to shunters since they are often in depots and sidings.
6.4 Workforce safety by occupation 6.4.1 Track workers There were 70 major injuries to track workers in 2007. Notable among these were the following:
• On 25 March, a track worker dropped a rail onto his middle and index fingers at Trent South (London North East); both fingers were later amputated in hospital. • On 13 May, some coping fell onto a track worker at Clapham High Street (South East), resulting in a fractured pelvis and tailbone. • On 29 August, a track patrolman was struck by a train at Leatherhead (South East); he suffered two broken femurs, fractured ribs, skull and vertebrae.
2007 Annual Safety Performance Report 77 Workforce safety
• On 13 November, a track worker received a glancing blow at Victoria Sidings (South East), after failing to acknowledge a passing train. He received multiple injuries, including electrocution from the third rail.
Chart 39. Trends in track worker fatalities and weighted injuries
Shock/trauma 30 Minor Major 25.2 Fatal 25
20
15.6 15 13.6
10.8 9.8 10 Fatalities and weighted injuries and weighted Fatalities 5
0 2003 2004 2005 2006 2007
As Chart 39 reveals, there has been a downward trend in track worker minor and major injuries since 2004. However, there was a slight increase in track worker FWI in 2007, due to the fatalities at Ruscombe Junction and Kennet Bridge. There were no track worker fatalities in 2006.
Major injuries contribute an average of 60% to the harm experienced by track workers.
The contribution of shock and trauma to harm profile is very small. This is likely to reflect a combination of factors, including the relatively small weight given to cases of shock arising from near misses in the calculation of FWI and the culture in which track workers operate. Nevertheless, near misses involving trains can have a profound effect on members of the workforce, and such incidents are reported more than once per week on average (see Chart 42).
78 2007 Annual Safety Performance Report Workforce safety
Chart 40. Track worker fatalities by cause
Electric shock 9 Struck by train 8 Slips, trips and falls 8 Other personal accidents
7
6 5 5 4 4 3 3 Fatalities per year per Fatalities 3 2 22 2 2
1 0 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Chart 40 shows that most track worker fatalities involve being struck by trains. Sometimes, however, these incidents are more unusual in nature, as occurred in 2004, when a specialist abseiler sustained fatal head injuries after falling down an 80ft tunnel vent shaft onto the running line at Fareham (South East).
Chart 41. Track worker major injuries by cause
Electric shock 160 Struck by train 143 Slips, trips and falls 140 Other personal accidents
120
100 100 87 76 80 70
60 Major injuries per year per injuries Major 40
20
0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 79 Workforce safety
Chart 41 shows that:
• The majority of major injuries occur due to slips, trips and falls and other personal accidents (which generally constitute injuries from using tools, dropping equipment on toes or striking parts of the body against fixed objects). Trains striking track workers has resulted in 10 major injuries in the past 10 years, compared to 24 fatalities. This reflects the likely severity of such accidents. • There have been significant downward trends over the period in track worker major injuries resulting from electric shock and other personal accidents (mainly comprising contact with objects and manual handling). • The number of major injuries in 2007 is significantly lower than the average of the previous four years, and less than half the total recorded in 2004. Chart 42 indicates that the monthly numbers of reported near misses exhibits a high degree of variation. However, overall, the annual moving average number of near miss incidents has fallen over the period and was at a level of six per month at the end of 2007 – less than half the level in 2004.
Chart 42. Near misses with the workforce
25 Near misses Near misses AMA
20
15
10 Near misses per month per misses Near
5
0 Jul Jul Jul Jul Jul Jan Jan Apr Jun Oct Jan Apr Jun Oct Jan Apr Jun Oct Jan Apr Jun Oct Apr Jun Oct Mar Mar Mar Mar Mar Nov Nov Nov Nov Nov Feb Feb May Aug Feb May Aug Feb May Aug Feb May Aug May Aug Dec Dec Dec Sep Dec Sep Dec Sep Sep Sep 2003 2004 2005 2006 2007
6.4.2 Train drivers There were 16 major injuries to train drivers in 2007, three of which occurred in train accidents:
• As noted elsewhere in this report, the driver was injured when a passenger train derailed at Grayrigg in Cumbria on 23 February (London North West).
80 2007 Annual Safety Performance Report Workforce safety
• On 30 April, a brick was thrown at a train near Salisbury (South East). The windscreen was smashed, glass entering both of the driver’s eyes. The relief driver on board took the train forward to Salisbury station. • On 21 August, a brick was thrown at a train from an overbridge at Philips Park (London North Western). It smashed though the windscreen and hit the driver, resulting in a broken shoulder.
Chart 43. Trends in train driver fatalities and weighted injuries
Shock/trauma 4.5 Minor 4.1 Major 4 3.8 3.8 3.9 Fatal 3.7
3.5
3
2.5
2
1.5
Fatalities and Fatalities and weighted injuries 1
0.5
0 2003 2004 2005 2006 2007
There were no train driver fatalities in 2007, although there was an increase in major injuries. Despite this, the total train driver FWI does not show a significant trend over the analysis period portrayed in Chart 43.
There have been seven train driver fatalities in the last 10 years, the details of which are as follows:
• Two train drivers were killed in the collision, derailment and fire at Ladbroke Grove (1999). • Two train drivers were killed in the collision and derailment at Great Heck (2001). • A train driver was killed when his HST collided with a car that had been parked on the level crossing at Ufton Nervet (2004). • A train driver was struck and killed by a train whilst changing ends at Edgely Junction (2005). • A train driver was electrocuted on the third rail whilst investigating a potential wagon fire at Deal (2006). Chart 44 presents information on train driver major injuries, with a third of those in 2007 involving boarding and alighting.
2007 Annual Safety Performance Report 81 Workforce safety
Chart 44. Train driver major injuries by cause
Other 18 Assault Boarding / alighting 16 16 Contact with object Slips, trips and falls 14 Struck by train 12 Train accident 12 10 10
8 77
6 Major injuries per year per injuries Major
4
2
0 2003 2004 2005 2006 2007
There were more than twice as many major injuries to train drivers in 2007 as there were in 2006. A decreasing trend thus appears to have reversed; however, the annual totals are not large enough to attach any statistical significance to change. These were all isolated incidents, and there is no common factor behind the increase.
Chart 45. Train driver shock and trauma leading to lost time
Train strikes person Near miss 200 Other train accident (eg hits object) Assault / abuse 179 Train struck by missile Other 180 158 153 160 148
140 119 120
100
80
60
40
Shock / trauma incidents resulting in lost time lost in resulting incidents trauma / Shock 20
0 2003 2004 2005 2006 2007
82 2007 Annual Safety Performance Report Workforce safety
When a train strikes a person, or is hit by an object, the shock or trauma that the driver suffers can have far-reaching consequences.33 Chart 45 indicates the magnitude of the problem, confirming that most of train driver shock/trauma incidents result from trains striking a person.
Around 20% of shock/trauma to train drivers leading to lost time occurs after a near miss incident, where the train was very close to striking and possibly killing the person.
After a low number of reports in 2004 compared to 2003, the number of reported incidents in the past three years has remained fairly constant at around 150 per year.
In a quarter of the cases shown on the chart, the driver was absent from work for a week or more, and in a further quarter of cases, he/she was off work for more than a month.
6.4.3 Train guards and other train crew There were 13 major injuries to train crew in 2007. Notable among these were:
• On 9 May, a member of the on-board train crew was struck in the eye by a piece of metal which flew through an open window as a result of air turbulence from a passing train. • On 31 July, a member of the on-board train crew fell between the train and platform when alighting from the crew door of a coach stabled in Platform 12 at Euston. He sustained a fractured ankle. The driver later advised that, while preparing the train, he had found the Track Access Device deployed and had retracted it. This meant that the door step had not deployed as the door had not been closed and reopened. The incident is subject to a formal investigation.
Chart 46 show FWI to train crew over the last five years.
• There have been no train crew fatalities since 2001, when a train guard and a member of the on-board train staff were killed in the accident at Great Heck. • The number of major injuries to train crew in 2007 is around half the number in 2006, and approximately one-third of the number in 2005. • The numbers of minor injuries and shock/trauma incidents have remained fairly constant over the period.
33 Within railway safety reporting, ‘shock’ refers to a momentary or short-term distress or disturbance, whereas ‘trauma’ describes more severe psychological problems that require specialist medical treatment. Instances of non-clinical shock within the railway are not uncommon, but occasions of clinically diagnosed trauma remain exceptional.
2007 Annual Safety Performance Report 83 Workforce safety
Chart 46. Trends in train crew fatalities and weighted injuries
Shock/trauma 10 Minor 8.9 9 Major 8.3 Fatal 8 7.5 7.1 7
6 5.3
5
4
3 Fatalitiesand weighted injuries 2
1
0 2003 2004 2005 2006 2007
Drilling down into the causes of the fall in train crew major injuries (Chart 47), it can be seen that there were no major injuries due to sudden train movement in 2007, a large contributor to the recent overall downward trend. Sudden train movements can occur on pointwork, but also on poor quality plain line. Recent track improvement work may thus have helped diminish the problem.
Chart 47. Train crew major injuries by cause
Accidents about the track 45 Assault Boarding / alighting 39 39 40 Other personal accidents in stations Other personal accidents on trains Slips, trips and falls in stations and trains 35 Sudden train movement Train accident r 30 28 26 25
20
Major injuries per yea 15 13
10
5
0 2003 2004 2005 2006 2007
84 2007 Annual Safety Performance Report Workforce safety
The chart also shows that:
• In 2007, there were more major injuries resulting from boarding and alighting accidents than 2006. However, this is roughly equal to the average over the period. • The number of major injuries resulting from assault has continued to fall, to just one in 2007. 6.4.4 Station staff There were 20 major injuries to station staff in 2007. Notable among these were the following:
• On 11 January, a member of the station staff at Preston (London North West) slipped on water that had leaked onto the platform as a result of a fridge being switched off. The operative suffered a broken arm and tailbone. • On 17 March, a member of staff was injured whilst trying to extinguish a dustbin fire at Slough (Western). The person slipped on the foam and suffered 20% burns (mainly to his left leg) as a result of coming into contact with molten plastic. • On 30 March, a member of the platform staff fell to the floor whilst restraining an assailant who had punched another member of staff in the face at Birmingham International (London North West). A broken elbow was sustained in the fall.
Chart 48. Trends in station staff fatalities and weighted injuries
6 Shock/trauma Minor Major 4.9 5 Fatal 4.6 4.4 4.2
4 3.4
3
2 Fatalities and weighted injuries and weighted Fatalities 1
0 2003 2004 2005 2006 2007
There have been no fatalities to station staff since 2002, when a member of the station staff at Purley died from an asthma attack triggered by a train fire at the station. Chart 48 also shows that:
2007 Annual Safety Performance Report 85 Workforce safety
• There were twice as many major injuries to station staff in 2007 as in 2006; however, the numbers involved are small so this is not significantly higher than the previous year or the average over the period. • There is no significant trend in total FWI over the period, the variation being explained by the yearly fluctuations in the number of major injuries to station staff.
Chart 49. Station staff major injuries by cause
Assault 25 Boarding / alighting 22 22 Manual handling 20 Other personal accidents 20 20 Slips, trips and falls
15
12
10 Major injuries per year per injuries Major
5
0 2003 2004 2005 2006 2007
The increase in major injuries in 2007 over 2006 (Chart 49) can be explained by higher numbers of assaults, slips, trips and falls and other personal accidents (mainly comprising walking into obstacles and incidents such as being struck by passengers’ luggage).
There is no significant trending behaviour over the period, with high degrees of variation year on year.
Chart 48 indicates that minor injuries account for almost half of the risk to station staff. Chart 50 shows that much of this can be attributed to walking into obstacles, being struck by passengers’ luggage, slips, trips and falls and assaults.
• The number of minor injuries resulting from contact with objects in 2007 is significantly higher than in 2006. • The number of minor injuries resulting from assault in 2007 is significantly lower than in 2006. • The total number of minor injuries in 2007 is marginally lower than the average of the previous four years; however, there is no significant trend over the period.
86 2007 Annual Safety Performance Report Workforce safety
Chart 50. Station staff minor injuries by cause
Assault Boarding / alighting Contact with object 1800 Manual handling Other personal accidents 1600 1531 1532 Slips, trips and falls 1456 1446 1412 1400 r 1200
1000
800
600 Minor injuries per yea per injuries Minor
400
200
0 2003 2004 2005 2006 2007
6.4.5 Shunters Many of the injuries to shunters occur off NRMI, so are beyond the scope of the ASPR.
• There was one major injury to a shunter on NRMI in 2007. This occurred at Kettering (London North East) in November when the operative slipped on some wooded boarding at the bottom of a platform ramp and broke his leg. • At Harlow Mill (South East, not on NRMI) in August, a shunter fell under a train whilst walking on the ‘four foot’, resulting in both legs requiring amputation. The operative had not been facing the flow of traffic. In February 2008, RSSB published a special topic report on shunter safety. This was based not only on SMIS data but also on accident records from companies’ own reporting systems, thereby including accidents that occurred in depots, yards and sidings. A summary of the findings is presented below:
• There have been four freight shunter fatalities in the past 10 years. All the shunters involved had been carrying out shunting duties ‘on the ground’. Three of the fatalities were attributed to getting trapped in between two vehicles whilst carrying out manual coupling. This rate compares to seven shunter fatalities in the previous 10-year period (1988–97). • The rates of shunter fatalities and RIDDOR-reportable injuries indicate that shunters face a high level of individual risk, compared with other railway occupations. • There is no evidence of any significant trend in shunter incident rates in recent years. • On average, each FOC shunter loses 0.7 days each year as a result of injuries sustained whilst working. This compares with 0.2 for TOC shunters.
2007 Annual Safety Performance Report 87 Workforce safety
• The most common cause of lost time injuries is slips, trips or falls whilst moving around or between work areas. Strains and sprains are the most common types of injury resulting in lost time. Human factors have a critical impact on shunter safety. In particular, management visibility, effective resource planning and a robust risk-based competence management system have been identified as some of the essential elements for maintaining shunter safety.
A number of industry conferences and workshops focused on the risk from shunting in 2007. These sought to identify control measures, including plans for effective management of staff, equipment, procedures and morale.
6.5 Further information The aforementioned shunting safety special topic report may be found on the RSSB website: http://www.rssb.co.uk/pdf/reports/shunter_safety_full_feb_08.pdf.
88 2007 Annual Safety Performance Report Public safety
7 Public safety
A person is considered to be a member of the public if he or she is classed as neither a passenger nor a member of the workforce. The highest level of risk on the railway occurs to people in this grouping. In the majority of cases, the risk is the direct result of their own behaviour, either deliberate or accidental, rather than the operation of the railway.
While the behaviour of members of the public causes risk to themselves, it sometimes also contributes to the risk to others. Misuse of level crossings by road vehicle drivers can result in train accidents, as can objects being left on the line by vandals. These sources of risk are discussed in Chapter 8, Risk from train accidents.
2007 Headlines • There were 263 public fatalities (including suicides) in 2007, which is fewer than in 2006 but close to the 10-year average. Of these, 57 were the result of accidents, and there were 206 suicides or suspected suicides. • 43 trespassers were killed in the year, 24 of whom were at stations (the highest count in 10 years). • There were 14 accidental fatalities to members of the public not engaged in acts of trespass; 11 of these were at level crossings (eight pedestrians and three road vehicle occupants). • Three other non-trespassers were struck and killed by trains in stations due to being too close to the platform edge. (They were not considered passengers, as they were not intending to travel.) * Performance at a glance
Non-trespass accidental fatalities All public fatalities 18 17 350 16 299 300 14 263 14 13 13 261 250 240 239 12 11 10 200
8 150 6 100 4 50 2
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 7% higher than 2006 2 13% lower than 2006 3 3% higher than 2003-06 average2 1% higher than 2003-06 average 2
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 89 Public safety
7.1 Risk profile for members of the public The risk to members of the public is dominated by fatality risk, weighted injuries accounting for a very small part the FWI total. This is partly because injuries to the public are not likely to be reported into SMIS, and partly because the hazards that account for most of the risk (in particular, being struck by trains) are more likely to result in fatality than injury.
Chart 51 presents the accident risk profile for members of the public (ie, excluding suicide and suspected suicide). The chart shows that most risk occurs to trespassers, either as a result of being struck by trains, or being electrocuted.
A notable proportion of risk also occurs to level crossing users, both road vehicle drivers and pedestrians. Most of this risk does not arise during proper use of the crossing, but during misuse, either as a result of honest error, or deliberate violation.
Chart 51. Risk to members of the public by activity and cause
Fall or jump from moving train 0.9
Fall or jump from height 2.7
Electric shock 7.3
Struck by train 31.6 t Other 0.1
Electric shock 0.1
Other platform edge 0.2
Assault 0.4
Other third party risk 0.4 Fatal
Type of public acciden public Type of Train accident 1.3
Slips, trips and falls 1.4 Major
Falls / incursions onto the Railway 3.3 Minor
Motor vehicle 3.0 Shock/trauma user Other Trespasser Level Pedestrian 7.7 crossing
0 5 10 15 20 25 30 35 40 45 50 Average risk per year
Source: SRMv5.5.
Chart 52 shows that most of the public fatality risk arises from a person’s own behaviour (whether deliberate or not). However, there is a small (but identifiable) level of risk to the public from structural collapses and train accidents. An example of this occurred in the Potters Bar train accident of 2002, when a member of the public was struck by falling debris from a rail-over-road bridge. These incidents, though extremely rare, still have the potential to cause harm to those not directly using the railway.
90 2007 Annual Safety Performance Report Public safety
Chart 52. Risk to members of the public by SSP risk area group (source of risk*)
Fatal Major Minor Shock/trauma Passengers 0.00
Environment 0.07
h public risk arises Workforce 0.37
Engineering 0.63 SSP area from whic from area SSP
Public behaviour 59.28
0 5 10 15 20 25 30 35 40 45 50 55 60 65 Average risk per year
Source: SRMv5.5. * The risk area groups relate to sources of risk so, for example, the bar corresponding the Passengers risk area group shows the risk that arises from passenger behaviour, rather than the risk to passengers.
7.2 Public fatalities and injuries in 2007 There were 57 accidental fatalities and 206 suicides and suspected suicides in 2007.34
Eight non-trespass fatalities occurred to pedestrians at level crossings. One of these was a child, with reports suggesting he was involved in a game of ‘chicken’.35 Three other non- trespass fatalities were road vehicle occupants, who died in collisions with trains at level crossings. (See Chapter 9, Risk at level crossings, for more details.)
The remaining three non-trespass fatalities concerned accidents involving the platform edge, these were:
• In January 2007, an intoxicated man who was running beside a train fell between it and the platform. • In February 2007, a young person died as a result of being too close to the platform edge. A train struck him and amputated his legs. • In August 2007, two 16-year-olds fell from the platform at Shoreham (South East) during a fight. One was struck by a train and died of his injuries, the other suffered major burns after an electric shock from the conductor rail.
34 This number is likely to change, due to upcoming coroners’ reports (see section 7.3).
35 Pedestrians on level crossings are always considered level crossing users, never trespassers.
2007 Annual Safety Performance Report 91 Public safety
7.3 Public safety trends Chart 53 shows the number of public fatalities over the last 10 years.
Chart 53. Trends in public fatalities
350 Suicides Trespassers At level crossings (non-trespass) Other (non-trespass) 299 300 282 8 270 5 5 11 266 261 263 4 254 32 3 6 8 4 2 250 237 5 240 239 11 11 11 3 3 4 56 61 8 10 47 14 46 43 47 31 37 200 56
150
Public fatalities 254
100 204 205 207 202 206 191 186 197 167
50
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Trends in the total number of trespass fatalities should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See below for more information.
It is clear from this chart that trespasser fatalities make up the bulk of all accidental public fatalities, accounting for 43 from a total of 57 accidental fatalities in 2007. For non- trespassers, the majority of accidental fatalities occur at level crossings (except in 2006). The three accidental non-trespasser fatalities in 2007 were all members of the public struck by trains after standing too close to the platform edge. They were not classed as passengers as they did not intend to travel.
There was a significantly higher number of public fatalities in 2006, mainly due to a significant increase in suicides. However, in 2007, the total number of public fatalities returned close to the 10-year average of 261, due to a reduction in suicides to more ‘usual’ levels.
Distinguishing between suicide and accidental death
When categorising public fatalities, it is useful to distinguish between suicides and accidental deaths, because the means of addressing these issues will be different.
For the rail industry, determining that a fatality was a suicide is straightforward where this was the conclusion of a coroner’s report. Similarly, where a coroner’s report concludes that a death was accidental, the industry classes the fatality accordingly. The difficulty lies in incidents where the coroner has yet to return a verdict, or returns an open verdict.
92 2007 Annual Safety Performance Report Public safety
Most coroners’ reports take around six months to complete, and some verdicts are not returned until several years after the event. A coroner will then only return a suicide verdict if there is evidence that proves beyond reasonable doubt that the deceased intended to take his or her own life. If the cause of death cannot be established, an open verdict is returned. The industry’s own investigations suggest that the majority (around four-fifths) of these fatalities are most likely to have been suicides.
In order to generate timely statistics that are as accurate as possible (if a coroner has yet to return a verdict or has returned an open verdict), the industry applies rules known as the Ovenstone criteria (see Appendix 4) to determine on the balance of probability whether a fatality was the result of an accident or suicide. The decision is based on all the information available, which might include evidence gathered by the local Network Rail manager and a BTP report. Fatalities that have been judged by the industry to have been suicides, but have not been classed as such by the coroner, are referred to as suspected suicides.
The 2006 ASPR reported that there were differences between the numbers of suicides and accidental trespass fatalities reported by RSSB and those being used internally by Network Rail. During the ensuing year, the two organisations have worked together to reconcile these differences and have established a process of better information sharing to ensure that fatality classifications are agreed in future. This has led to a major downward revision of accidental trespass fatalities in 2006 from the figures published in last year’s ASPR (and a corresponding increase in suicides). Because of the difficulties in revisiting the way in which historic fatalities were classified, there may be inconsistencies in historic trespass and suicide statistics. For this reason, the separate trends in accidental trespass fatalities and suicide fatalities should be interpreted with caution. Although future statistics will be more robust, there are likely to be continued minor changes to the numbers of suicides and trespass fatalities because classifications are reviewed when a coroner’s verdict or other evidence comes to light.
7.4 Trespassers A trespasser is someone who goes where they are never authorised to be (for example, someone who accesses the track from a station platform). The term is never applied to level crossing users, even if they are behaving inappropriately. Passenger trespassers are classed as members of the public for the purposes of this report, and are included in the analyses in this section.
There were 43 trespass fatalities during 2007; none were under 16-years-old. There were also 30 major injuries to trespassers in 2007, including two under-16s:
• On 29 May, a child (exact age unknown) fell from Crayford Road Bridge after walking along the coping stones on top of the bridge and slipping off. He fell on to the track and sustained a suspected broken ankle. This person was classified as a trespasser due to the fact he had climbed up on top of the bridge. • On 30 October, a 14-year-old boy climbed onto the glass roof of Battersea Park station and fell though the glass to the booking hall floor below. This resulted in a broken arm and leg, as well as serious head injuries.
2007 Annual Safety Performance Report 93 Public safety
A range of other circumstances surround trespasser fatalities, such as people taking short- cuts across the railway, vandalism activities, and one person who was electrocuted while riding on top of a freight rail vehicle (an activity known as ‘train surfing’).
7.4.1 Fatalities and injury trends Chart 54 shows that, although the number of trespass fatalities was close to the 10-year average rate of 45.6 in 2007, their general location has changed. The 24 trespass fatalities occurring at stations in 2007 (the highest of the last 10 years) represent a significant increase on earlier years. Taken as a proportion of the total number of fatalities, the percentage of trespass fatalities occurring at a station in 2007 was 56% (the 10-year average is 35%). Note that the location is deemed to be ‘at a station’ if the person was injured within 200m of the station limits.
Chart 54. Trespass fatalities by location
70 Not at station At station 61 60 56 56
50 47 47 46 43
40 45 37 38 36 32 28 31 19 31 30 32 23
Trespasser fatalities Trespasser 19 19 20
24 10 20 19 18 16 16 14 12 14 13
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Trends in the total number of trespass fatalities should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See section 7.3 for more information.
Being struck by a train is the most common cause of death for trespassers on the railway (see Chart 55). The next most common cause of death is electrocution.
94 2007 Annual Safety Performance Report Public safety
Chart 55. Trespass fatalities by cause of death
70 Struck by train Fall or jump from height 61 Fall or jump from moving train Electric shock 60 1 Other 56 56 2 1 14 5 50 47 47 2 12 46 1 1 1 43 6 7 5 3 1 40 2 37 2 22 2 8 32 1 8 31 1 1 30 4 2 10
Trespasser fatalities Trespasser 46 45 20 1 38 37 37 38 33 29 25 10 20
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Trends in the total number of trespass fatalities should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See section 7.3 for more information.
In 2007, the number of major injuries to trespassers returned to a level close to the average for the last five years, as shown in Table 8.
Table 8. Trespass major injuries Injury Degree 2003 2004 2005 2006 2007 Mean Major Injury 30 25 20 37 30 28.4
Some of the injuries sustained by trespassers are particularly severe. They include, for example, serious burns from electric shock, such as when a man jumped onto the top of a freight train in April 2007, coming into contact with the overhead line. He was taken to hospital in a critical condition, suffering serious burns. Other injuries include multiple fractures and serious head injuries as a result of being struck by a train while trespassing.
7.4.2 Near misses and other trespass reports The accuracy of trespass and near miss data is reliant on how well the events are reported. The probability that an event will be reported might depend on many factors, such as the age of the perpetrator, if they were seen (affected by daylight) and how close the near miss was. It is therefore difficult to distinguish changes to reporting rates from changes to actual trespass rates. Chart 56 shows the trends in reported trespass and near miss incidents.
2007 Annual Safety Performance Report 95 Public safety
Chart 56. Trends in reported trespass and near miss
14000 Other trespass incidents 12432 Trespass near misses 11928 11823 12003 12000 11075
10000
8000
11961 11354 11354 11594 6000 10607
4000 Trespass incidents/near misses incidents/near Trespass
2000
574 0 469 409 471 468 2003 2004 2005 2006 2007
7.4.3 Factors affecting trespass Age
A breakdown of trespass fatalities by age for the past five years is shown Chart 57. The peak age for trespass fatalities is 17; when normalised by population, nearly four times as many young adults (aged between 16 and 29) are killed trespassing than older adults (aged 30 and over).
The fact the children account for only 8.5% of trespass fatalities but as much as 50% of near misses (see Chart 60) seems inconsistent. A possible reason for this is that drivers are more likely to report near misses with children. However, it may be children are somehow ‘better’ at avoiding the dangers of trespass than those who are older, possibly as a result of initiatives that have been introduced to educate children of the dangers and possibly as a result of alcohol and drugs being a factor in adult fatalities, especially in the evening. Another possibility is that, as most child near misses occur during the day (see Chart 60), drivers are more likely to see them. A higher proportion of near misses with adults occur at night, and it is likely many of these events are simply not seen by drivers (discussed later in this section).
96 2007 Annual Safety Performance Report Public safety
Chart 57. Breakdown of trespass fatalities by age (2003–2007)
14 900 UK population
800 Average UK population- 2003-2007 ('000s) 12
Young Adults 700 10 47.6% 600
8 500
6 400 Children Older adults 8.5% 43.9% 300 4 200 Trespasser fatalities (2003-2007) fatalities Trespasser 2 100
0 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Age Source: Trespass fatalities – RSSB fatalities database (derived from SMIS). UK population – Office for National Statistics.
Chart 58 shows the number of child fatalities split by trespass and non-trespass. There were no child trespass fatalities in 2007 – the first time this has occurred in a decade. Indeed, there have been very few child fatalities in the last two years, a fact which may point to the success of the various initiatives to warn children about trespassing on the railway.
Chart 58. Public fatalities occurring to children
7 Girls 66 6 Boys 6 1 5 5 5 2 1 4 4 4
3 3 55 1 2 2 Accidental fatalities Accidental 2 4 4 4 1 1 11 1 1 2 2 2 1 111 1 00000 0 0 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Trespass Non- trespass
2007 Annual Safety Performance Report 97 Public safety
Time of day
Chart 59 presents the number of trespass fatalities occurring at every hour of the day, using data from 2003–07. It is broken down by age group.
The total number of near miss events since 2003, together with the number of trespass incidents, can be seen in Chart 60. The exact age of the trespasser is not generally known in a ‘near miss’ event, but it is often possible to make a rough classification based on the driver’s description. The three age categories used in the chart, adult, young person and child, do not correspond to well-defined age bands.
Chart 59. Trespasser fatalities by time of day and age (totals for the years 2003–2007)
25 Age unknown Older adults 20 Young adults 20 Children
16 1 15 10 14 13 2 11 11 2 8 10 1 10 10 9 99 9 1 4 31 75 2 3 3 7 5 6 1 8 4 1 3 Trespasser fatalities (2003-2007) fatalities Trespasser 4 5 44 3 44 4 5 9 4 1 33 7 3 3 1 1 3 3 1 3 2 6 4 6 1 1 4 2 2 2 4 1 3 3 3 3 313 2 2 2 2 22 1111 1 1 0 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00
The high number of fatalities seen in the early evening can be explained by the fact that this is the peak time for trespass incidents (see Chart 60). Many incidents at this time involve children and young adults. This is the time when children are most likely to be active outside school and is also when the highest number of children are killed while trespassing. However, the high fatality rate continues into the night, and is highest at midnight, when reported trespass incidents are comparatively low. Some of these could be caused by individuals taking short-cuts on their way home after a night out and it is likely that alcohol may have been a factor in many of these incidents. Alcohol can affect a person’s perception of danger and their reaction times.
Chart 59 also shows a peak in the number of fatalities occurring early in the morning, between 05:00 and 09:00 and consisting mainly of young adults. At this time there are, again, few trespass incidents (see Chart 60), so there must be another explanation for the peak in fatalities. The circumstances surrounding early morning fatalities show several factors combining to give a higher than expected rate.
98 2007 Annual Safety Performance Report Public safety
• 15 fatalities involved a person being struck by first, or very early, passenger or freight trains. It could be that the victims were less vigilant than they might ordinarily have been, because they were not expecting trains to be running at that time. • 17 fatalities were reported between 05:00 and 09:00, but actually happened earlier. Many were later traced to a train running the night before, but were not seen (due to the darkness) until the morning. They were therefore subject to causes similar to incidents that occurred at night. These two factors account for more than half of all early morning fatalities occurring in the last five years and, together, probably explain the unusually high rate at this time.
In the time between 10:00 and 17:00, the number of fatalities is lower and consists mainly of older adults (60%).
Chart 60. Near misses with adults, young people and children by time of day, and reported trespass incidents (totals for the years 2003–2007)
300 6000 Adult Young person 250 5000 Child 233 226 Reported trespass incidents 217
197 Trespass incidents 191 200 185 4000 86 76 87 154 66 146 150 85 69 3000 130 110 65 80 77 100 92 97 65 88 2000 Near misseswith train 75 77 77 58 59 67 66 62 53 51 46 47 44 60 50 37 43 1000 42 42 49 38 28 67 18 18 20 32 21 53 55 53 30 18 43 49 48 7 105 26 20 4 4 2 4 124 15 3 5 7 12 20 24 26 17 2 12 15 16 2 0 4 1 3 1 222 8 3 8 0 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00
It is possible that children and young people constitute a different ‘type’ of trespasser, who are more likely to be on the railway for the purpose of spending time there rather than, say, simply taking a short-cut. This would mean that young trespassers are more likely to be seen and, therefore, reported.
A final point of interest is that there have been eight fatalities to trespassing dog-walkers over the last five years. Four occurred between 05:00 and 09:00, with a further two around lunch time and two in the early evening.
2007 Annual Safety Performance Report 99 Public safety
Gender
In 2007, 93% of trespasser fatalities were male. Over the last 10 years, 90% were male.
Trackside vandalism
Vandalism on the railway encompasses any kind of deliberate damage or defacement to the property of the railway.
• ‘Superficial’ vandalism, like graffiti, gains in seriousness when committed extensively, violently or as an expression of hatred and intimidation. It can cause fear among passengers, fostering doubts about the safety of public transportation. • ‘Structural’ vandalism has the real potential to result in safety risk (either deliberate or unintentional) to rail workers and users. For example, damage caused by throwing missiles at trains can cause injury to passengers or workforce from the missile itself or from shattered glass. Missiles can also distract the driver. With all kinds of vandalism, however, there is also the personal risk that the vandals themselves run simply by trespassing into dangerous areas. For example, in March 2007 a train struck a person at Bethnal Green. The man was found with paint on his hands and spray cans in the area suggest he may have been involved in vandalism. Also, in July 2007, a young male fell from a bridge to the track after trying to deface the bridge. He survived, but suffered a fracture to the skull and severe burns due to coming into contact with the conductor rail.
• The trends in trackside vandalism over the last five years are shown in Chart 61.
Chart 61. Trends in vandalism at the trackside
1400 Missiles thrown or fired
1200 Obstruction on the line Arson
1000 Vandalism (other)
800
600
Number of incidents of Number 400
200
0 Jul Jul Jul Jul Jul Jan Jan Apr Jun Oct Jan Apr Jun Oct Jan Apr Jun Oct Jan Apr Jun Oct Apr Jun Oct Mar Mar Mar Mar Mar Nov Nov Nov Nov Nov Feb Feb May Aug Feb May Aug Feb May Aug Feb May Aug May Aug Dec Dec Dec Sep Dec Sep Dec Sep Sep Sep 2003 2004 2005 2006 2007
100 2007 Annual Safety Performance Report Public safety
The chart is broken down into different types of vandalism:
• Arson. • Obstructions on the line, which includes serious obstructions that have the potential to derail a train. (This hazard is dealt with in more detail in Chapter 8, Risk from train Accidents.) • Missiles thrown or fired. • Vandalism (other), which includes interference with trains, lineside equipment, damage to property at stations and any other incidents such as graffiti or cable theft. A strong seasonality is evident in the occurrence of trackside vandalism incidents. There is a very sharp peak, in late spring, caused especially by an increase in incidents where missiles are thrown or fired. After the spring peak, a high level of vandalism continues through to early summer. It appears the spring peak coincides with the Easter holidays. At this time, the children and young people36 who are responsible for a many incidents of vandalism are not at school and it is warm enough and light enough in the evenings to make it more attractive to be outside. There is smaller peak in the summer holiday period, and another possibly coinciding with half-term in autumn. It is not known why these peaks are smaller, though it is possible that many potential trespassers are away from home in the summer holiday period.
Not all types of vandalism have the same seasonality. Chart 61 shows that Vandalism (other) varies much less throughout the year. One possible reason for this could be that graffiti crimes can be perpetrated inside, where offenders are less likely to be put off by the winter cold. Another reason could be that the acts of vandalism in these categories (such as cable theft) are more likely to be committed by adults, who are less likely to be affected by school term times. It certainly appears that children and young people are involved more in missile throwing (38% of cases where the identity of the perpetrator is known), compared with Vandalism (other) (19%).
Chart 61 also shows that there is a general downward trend of all types of vandalism except Vandalism (other), which began to increase in 2006, pushing the overall number of incidents above that of 2005. Much of this is due to the increase in cable theft incidents from the beginning of 2006 linked to the rise in value of copper.
Many initiatives aimed at improving security at high risk areas are currently under way at the moment. These are detailed in Chapter 3, Review of progress against SSP trajectories. There are also many research reports relating to vandalism on the RSSB website: www.rssb.co.uk.
36 Out of all incidents of vandalism, children and young people are reported as the perpetrators in 31% of cases. The perpetrator is not described for much of the other 69%, but it is likely many of these are children and young people.
2007 Annual Safety Performance Report 101 Public safety
7.5 Level crossing users As seen in Chart 62, fatalities account for most of the total FWI figure for members of the public at level crossings.
Chart 62. Fatalities and weighted injuries to level crossing users (MOP, excluding suicides)
12 Weighted injuries Fatalities 10 9.8 0.8 8.6 8.2 0.6 8 0.2
5.8 6 5.5 5.6 0.5 0.8 0.6 9.0 4 8.0 8.0 3.3 3.1 3.1 0.1 0.3 0.1 5.0 5.0 5.0 2 3.0 3.0 3.0 Number of fatalities and weighted injuries weighted and fatalities of Number
0.01 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
Pedestrian Road vehicle occupant
Public fatalities at level crossings show no clear trend over the last five years.
For more information on level crossings, see Chapter 9, Risk at level crossings.
There are many initiatives being implemented at the moment to improve the safety of pedestrians and road vehicle occupants at level crossings, as well as the passengers and workforce that have contact with level crossings. These are covered in detail in Chapter 3, Review of progress against SSP trajectories.
7.6 Other members of the public 7.6.1 Members of the public in stations Though people have always met and seen off friends and loved ones, the rise in retail outlets at stations has brought an increase in non-travelling members of the public at stations. Now people enter the station’s confines to shop, dine or simply browse through the merchandise. In this, they carry with them a risk profile similar to passengers. There were three public fatalities at stations in 2007, as well as seven major injuries. This subject is discussed in more detail in Chapter 11, Station safety.
7.6.2 Falls onto the mainline railway For only the third year out of the last 10, there were no accidental fatalities from falls or jumps onto the railway in 2007 (see Table 9).
102 2007 Annual Safety Performance Report Public safety
Table 9. Accidental fatalities from falls and jumps onto the railway
Year 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Mean Fatalities 33320130802.3
7.6.3 Road vehicle incursions Road vehicle incursions are when road vehicles break through barriers or fences and end up on railway property. These can put many people at risk, and in different ways. For example, a road vehicle occupant could be at risk from the initial accident itself, especially if the vehicle collides with an object. The road vehicle occupant may then incur similar hazards as trespassers, such as electric shock if they exit the vehicle. If a road vehicle blocks the path of a train, there is the potential for a train accident, putting the train occupants at risk, especially if the train derails. The last major accident of this type, involving multiple fatalities, occurred at Great Heck in 2001, where 10 people were killed and 76 more were injured. There were six incidents in 2007 where a train struck a road vehicle as a result of an incursion. (Chapter 8, Risk from train accidents, covers this subject in more detail).
7.6.4 Railway neighbours There is a small risk to those that live in, work in, or travel though, areas that adjoin the railway. An example of the consequences of this type of event happened as a result of the derailment at Potters Bar in May 2002, where a pedestrian walking beneath an underbridge was killed after debris fell from it to the road below. Another example was seen in Amagasaki, Japan in 2005, when more than 100 people died after a high-speed train derailed and crashed into an apartment block; many of the victims were residents in the block or bystanders in the surrounding area and car park.
7.7 Suicide The railway uses the Ovenstone criteria to differentiate between suicides and accidental fatalities (see section 7.3 for a full discussion). Note that passengers who commit suicide are classed as members of the public for the purposes of this report, and are included in the analysis in this section.
The highest number of suicides and suspected suicides on the railway in the last decade was seen in 2006 (254, as shown in Chart 63). This is a significantly higher number than the other nine years in that period. However, it appears that it might be simply a ‘statistical outlier’; railway suicides in 2007 have returned to a level closer to the 10-year average of 202 per year.
Over recent years, around 10% of suicides have occurred at level crossings, roughly one- third at stations, and the remainder at other locations (mostly on open track, but some elsewhere on railway property).
2007 Annual Safety Performance Report 103 Public safety
Chart 63. Suicides and suspected suicides by location
300 In stations At station level crossings At level crossings Other 254 250
207 204 205 202 206 197 191 200 186 142 167
150 113 122 121 104 112 143 110 97 Suicides 100 100 22 1 19 18 14 23 11 12 18 1 2 14 14 50 89 72 69 68 75 73 67 73 53 47
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Trends in the total number of suicides should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See section 7.3 for more information.
Table 10 shows that, as with suicides, numbers of major injuries from attempted suicides have fallen back to close to average levels in 2007 after a significantly higher level in 2006. It may be surprising to note the number of failed suicide attempts that result in major injury. The major injuries arising from attempted suicide are often very severe, including whole body injuries, and amputations.
Table 10. Major injuries from attempted suicide
Year 2003 2004 2005 2006 2007 Mean Major Injuries 30 30 23 42 26 30.2
As with suicide in general in the UK, many more men than women commit suicide on the railway; however, the ratios are slightly different. The average ratio over the last 10 years has been 4:1 on the railway, compared to the overall average37 of around 3:1.
The suicide rate on the railway varies considerably with age as well as gender. The rate of male suicide increases with age for under-18s and remains relatively high between 18 and 45, after which the rate drops to a low level by the age of 60. The female rate of suicide on the railway remains similar for all ages although the highest rates can be seen in women in their 50s.
Chart 64 presents the national rate for suicides, together with the number of suicides occurring on the railway. The national level of suicides can be seen to drop steadily almost
37 According to Office for National Statistics.
104 2007 Annual Safety Performance Report Public safety
every year since 1998. This could be due to the national suicide prevention strategy for England, which aims to reduce the number of suicides by 20% by 2010. However, the number of suicides on the railway has not exhibited the same pattern. While the proportion of the total number of suicides occurring on the railway is small (around 4%), it has been increasing.
Chart 64. Railway suicide trends in the wider context
8000 National Statistics (UK) On railway property 7000 % suicides occuring on railway property
6000 200 property railway on Suicides
5000
4.6% 4000 254 3.6 3.6% 3.3% 3.3% 3.2 3.2% 3.4% 207 206 205 204 202 197 191 2.6% 186 6354
3000 6255 100 6124 168 5932 5850 5826 5731 5671 5554
2000
1000 Suicides in the UK (including open verdicts) open (including UK the in Suicides
0 N/A 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Source: Railway Suicides – RSSB fatalities database (derived from SMIS). National Suicides – Office for National Statistics. Trends in the total number of suicide fatalities on the railway should be interpreted with caution because of possible inconsistencies in the way that historic fatalities have been classified as accidents or suicides. See section 7.3.
2007 Annual Safety Performance Report 105 Public safety
Intentionally blank
106 2007 Annual Safety Performance Report Train accidents
8 Risk from train accidents
This chapter covers all types of train accident. Those most likely to lead to harm are termed potentially higher-risk train accidents (PHRTAs), a group which includes derailments, trains striking road vehicles, buffer stop collisions and collisions between trains (excluding roll backs and open doors). However, the chapter also includes analysis of ‘non-PHRTA’ incidents, such as train fires and trains being struck by missiles. More information on the precise definitions may be found overleaf, in section 8.1.
2007 Headlines • The derailment at Grayrigg on 23 February resulted in the first on-board fatality in a train accident since November 2004. • There were 49 PHRTAs during the year, which is very similar to 2006. • There were 25 derailments. Seven involved passenger trains, most of which were caused by landslips and/or trees on the line. • There were six passenger train collisions, all of which occurred at low speed. These involved roll backs, station arrivals and uncoupling errors. There were no reported major injuries. • There were eleven collisions between trains and road vehicles at level crossings. Three resulted in fatalities to road vehicle occupants. • The Precursor Indicator Model (PIM) remained fairly static during the year. Based on this measure, train accident risk has roughly halved over a five-year period. • There were 785 RIDDOR-reportable train accidents in 2007, compared to 821 in 2006. * Performance at a glance
RIDDOR reportable PHRTAs per mtm PIM indicator 0.25 90 84.8 80 0.20 0.19 0.18 70 66.0 55.8 0.15 0.15 60 0.15 0.14 47.9 50 46.7
40 0.10 30
0.05 20 10 0.00 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 7% higher than 20062 2% higher than 2006 2 11% lower than 2003-06 average2 25% lower than 2003-06 average 33
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 107 Train accidents
8.1 Trends in the numbers of train accidents Train accident risk cannot be eliminated entirely. However, incidents of a more serious nature are few and falling, and the likelihood of being injured in an accident remains very low.
The nature of what is regarded as a train accident varies enormously in scale. It can range from vandals throwing objects at trains to a collision between two high-speed passenger trains. In fact, the term ‘train accident’ applies to eleven types of RIDDOR-reportable event (shown in Table 11). Those not reportable under RIDDOR are generally excluded from the analysis in this chapter.
Although some reportable train accidents have the potential for serious consequences, most carry little risk. For this reason, RSSB focuses on a small set of train accidents, known as potentially higher-risk train accidents (PHRTAs), to cover events that have the most potential to result in harm to any or all person types. The scope of the different subsets is contained in Table 11, along with the SRM hazardous events to which each relates. Table 11 also shows which RIDDOR-reportable accidents are covered by the Precursor Indicator Model (PIM) – see section 8.5.
Table 11. Train accident scope Covered by the Train All train accidents SRM hazardous events PHRTA accident PIM Train derailments (excludes striking road vehicles on level crossings) HET-12, HET13 All All Collisions between trains HET-01, HET-02, HET-03, HET- (excluding roll backs) 06B, HET-06C All All Roll back collisions HET-06A No All Train striking buffer stops HET-09 All All Trains striking road vehicles at level crossings (including derailments) HET-10, HET-11 All All Train fires HET-17, HET-18, HET-20 No All Trains running into road vehicles not at level crossings & no Used as a precursor in derailment Part of HET-04 All the PIM Open door collisions HEM-04, HEM-18 No No Trains running into other objects with no derailment (incl gates/ Used as a precursor in barriers at level crossings) Part of HET-04 No the PIM Trains striking animals with no Used as a precursor in derailment Part of HET-04 No the PIM Trains being struck by missiles HEM-04, HEM-18 No No (for all categories, only those accidents that meet the RIDDOR 1995 definition are included)
8.2 Train accident risk profile Chart 65 shows train accident risk by accident type and by the type of person affected. Most of the risk to passengers and the workforce occurs in train derailments, whilst trains striking road vehicles at level crossings lead to most public harm.
108 2007 Annual Safety Performance Report Train accidents
Chart 65. Train accident risk by accident type
Train derailments (excluding collisions with road 3.71 vehicles on level crossings)
Collision with road vehicle on level crossing 3.54 (including derailments)
Train collisions (including roll-backs) 1.15
Train striking objects (with no derailment) 0.87
Struck by missiles 0.25
Other 0.15
Passenger Train accidentTrain riskby accidenttype Train fires 0.15 Public Workforce Buffer stop collisions 0.14
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Average risk per year
Source: SRMv5.5.
Chart 66 breaks down train accident risk by person type and precursor group, as used by the Precursor Indicator Model (described later in the chapter). It shows that, whilst level crossing misuse contributes to the highest public risk, it has a relatively low impact on passenger and workforce safety.
Chart 66. Train accident risk by precursor type
3.5 Passenger Workforce Public
3.0
2.5
2.0
1.5
Average risk per year per risk Average 1.0
0.5
0.0 1 - Infrastructure 2 - Irregular working 3 - Level crossing 4 - Objects on the 5 - SPAD 6 - Trains & rolling failures misuse line stock Precursor grouping
Source: SRMv5.5. Note that the risk covered by this chart equates to 8.55 FWI, which is less than the risk of 9.97 FWI covered by Chart 65 and Chart 67. The difference arises because this chart does not show the risk from train accidents that are not covered by the Precursor Indicator Model – notably collisions with road vehicles that do not take place at a level crossing and do not derail a train (where most of the risk is to the road vehicle occupants), and trains being struck by missiles.
2007 Annual Safety Performance Report 109 Train accidents
Chart 67 presents train accident risk to each person type, by SSP risk area group. The greatest risk arises from public behaviour, and also affects members of the public. This is mostly due to level crossing misuse, or road vehicle incursions. Although these sources of risk also have the potential to harm passengers and the workforce, it is not the greatest source of risk for these person types. For passengers, the greatest fatality and FWI risk arises from causes identified by the SSP as being in the area of engineering, with workforce causes (eg, SPADs) being next greatest.
Chart 67. Train accident risk by SSP risk area group (source of risk*)
Engineering 0.52 Fatal Environment 0.05 Major Minor Public behaviour 0.48 Shock/trauma Workforce 0.60
Engineering 0.27
Environment 0.06
Public behaviour 3.48
Workforce 0.27
Engineering 1.93 Train accident risk by SSP area by risk accident Train
Environment 0.15
Public behaviour 0.99 Passenger Public Workforce
Workforce 1.16
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Average risk per year Source: SRMv5.5. * The risk area groups relate to sources of risk so, for example, the bar corresponding the Workforce risk area group shows the risk that is under workforce control, rather than the risk to members of the workforce. Passenger behaviour does not contribute to train accident risk, so the Passenger risk area group has been omitted from the chart.
The SRM can also be used to predict the average number of years between train accidents. Table 12 gives the number of years between events that result in five or more fatalities and 10 or more fatalities. For each consequence level, the number of years has increased since version 1 of the SRM was published, indicating that the likelihood of multi-fatality accidents is decreasing. This reflects the industry’s success in tackling train accident risk, taking into account recent system improvements such as TPWS, Mark I slam-door stock removal and improvements in track quality.
Table 12. SRM estimated frequency of train accidents by severity
Number of fatalites >=5 >=10 (passengers, workforce and public) SRM v5.5 / SRM v5 5.3 9.1 Average number of SRM v4 3.8 7.9 ye ars be twe en SRM v3 2.4 5.6 accidents SRM v2 / SRM v1 1.4 3.1
110 2007 Annual Safety Performance Report Train accidents
8.3 Train accident fatalities and injuries in 2007 There was one passenger fatality in a train accident during 2007:
• On 23 February, a passenger was killed when a passenger train derailed at Grayrigg in Cumbria (London North West). The immediate cause of the derailment was deemed to be the stretcher bar arrangement at the points, which resulted in a loss of gauge separation at the switch blade. In addition to the fatality, the train driver and 28 passengers received major injuries; there were also 59 minor injuries. There were also three accidental fatalities involving members of public where road vehicles were involved in collisions with trains at level crossings.38 In fact, the majority of train accident fatalities over the last five years have involved members of the public.
There were six further PHRTAs which led to physical injury:
• On 11 January, a passenger train struck the buffer stops at West Kirby station (London North West). Extensive damage to the station forced its closure. There were no passenger injuries, although the driver and guard suffered minor injuries. • On 12 June, a passenger train struck a road vehicle on Chapel automatic open level crossing (Western). There were no injuries to the passengers or staff on board the train. The road vehicle occupant was taken to hospital with minor injuries. • On 3 September, an empty coaching stock formation derailed whilst moving from the sidings at Exhibition Centre on to the Up Argyle line (Scotland). The rear coach was derailed by all wheels and was leaning against a tunnel. There was extensive damage to the overhead line and the infrastructure. A member of the revenue protection staff suffered minor injuries. • On 17 September, a passenger train collided with a road vehicle at Markinch (Scotland). The car driver suffered a minor injury. • On 10 October, a passenger train collided with the buffer stops at Sheringham (Anglia). Four passengers sustained minor injuries. • On 28 November, a low-speed collision occurred between two passenger trains at Manchester Airport station (London North West). The incident occurred due to a coupling error when the driver failed to stop short before coupling. Two train crew members received minor injuries. Three non-PHRTAs led to injuries in 2007:
• Two train drivers were seriously injured by missiles through train windows. • A passenger was knocked unconscious when a train ran into a tree, which then smashed through the bodyside windows. The train did not derail.
38 In addition, there was one suicide.
2007 Annual Safety Performance Report 111 Train accidents
8.4 Trends in train accidents
Chart 68. FWI in train accidents over past five years
14 Weighted Injuries 11.9 Workforce 12 Public Passenger Fatalities Workforce 10 Public 8.6 Passenger 8.1 8
6
3.8
FWI from train accidents train from FWI 4
1.8 2
0 2003 2004 2005 2006 2007
The level of harm from train accidents varies from year to year. A single major accident can dominate that year’s figures, particularly for passengers and members of the workforce. However, most fatalities tend to be road vehicle occupants.
Table 13 lists the seven train accidents that have resulted in on-board passenger or workforce fatalities over the past decade.
Table 13. Fatal train accidents over past 10 years (on-board fatalities only)
On-board Date Location Nature of accident Cause fatalitites 05/10/1999 Ladbroke Grove Collision - 2 passenger trains SPAD 31 17/10/2000 Hatfield Derailment - passenger train Broken rail 4 Passenger train derailment then 28/02/2001 Great Heck Car on track 10 collision with freight train 10/05/2002 Potters Bar Derailment - passenger train Points failure 6 Passenger train collision with 06/11/2004 Ufton AHB Suicide 6 road vehicle on level crossing Collision - on-track plant in 05/03/2004 Ancaster Irregular working 1 possession 23/02/2007 Grayrigg Derailment - passenger train Points failure 1
112 2007 Annual Safety Performance Report Train accidents
8.4.1 Long-term trends in the number of fatal train accidents
Chart 69. Train accidents leading to on-board fatalities
12 On-board fatalities 10 50+ 25 to 49 8 10 to 24 5 to 9 6 1 to 4
4 Number of accidents of Number
2
0 1945 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 Lewisham Clapham Jn Bourne End Bourne Hither Green Ladbroke Grove Ladbroke Harrow & Wealdstone South Croydon Jn / Goswick CroydonSouth Jn /
Source: ORR for historic data; SMIS for recent statistics. Accidents that resulted in more than 24 fatalities are labelled.
As Chart 69 reveals, the fatal accident count has fallen significantly since World War II. Over the period, eight accidents resulted in 25 or more on-board fatalities.
Bourne End (1945) was caused by excessive speed on a crossover. The consequences of the resulting derailment were worsened as the locomotive and carriages plunged down an embankment. Goswick (1947) was also caused by a driver overspeeding on a crossover and 27 passengers and one train attendant lost their lives in the resulting derailment. The derailment at South Croydon (1947), however, occurred after a rear-end collision between two heavily loaded passenger trains, following a signaller’s error.
Harrow & Wealdstone (1952), second only to Quintinshill (1915) in term of lives lost,39 involved three trains, a SPAD leading the first to rear-end the second, with a third striking the wreckage soon after. The incident at Lewisham (1957) also arose from a SPAD, which also led to a rear-end collision. In this case, the consequences were compounded by a bridge strike and collapse.
At Hither Green (1967), a passenger train was derailed by a piece of rail which broke away as it passed. The derailed wheels ran along the sleepers for a quarter of a mile before they struck a crossover. Clapham Junction (1988) occurred when a wiring fault led to a signal displaying a proceed aspect when it should have been at danger; this caused two trains to collide; an ECS formation, running on the adjacent line, also struck the wreckage.
39 Britain’s worst crash occurred on 22 May 1915, when a troop train collided with a passenger service at Quintinshill (near Gretna Green). There were 227 fatalities and 246 injuries.
2007 Annual Safety Performance Report 113 Train accidents
Like its predecessors, Ladbroke Grove (1999) was covered extensively by the media. It involved a departing suburban train, which passed SN109 signal at danger before colliding with an incoming HST. There were 31 fatalities.
The most recent rail accident involving a fatality took place on 23 February 2007, when an express derailed on defective point work at Grayrigg in Cumbria.
Our industry has made many safety improvements over the years to reduce both the frequency and the consequence of train accidents. The withdrawal of coaching stock with gas lighting will have played a part in mitigating the risk from post-collision fires, for example, while the use of buckeye couplings,40 continuous braking and the adoption of more crashworthy designs will all have had an effect. Indeed, the interim RAIB report into the Grayrigg incident has suggested that the death toll was prevented from being higher by the crash-resistant characteristics of the Class 390 EMU involved.
Yet many accident causes have remained prevalent over the years. Some precursors, like the road vehicle incursions that led to Great Heck (2001) and Ufton Nervet (2004) are largely beyond the railway’s control.
Three of the worst accidents shown on Chart 69 resulted from SPADs; two of the incidents (Harrow & Wealdstone and Lewisham) involved fatality counts in excess of 50, while the most recent (Ladbroke Grove) was above 30. The risk from SPADs has been significantly reduced since the fitment of TPWS, but the broken rail scenario, similar to that which led to Hither Green (1967), has caused problems more recently, with Hatfield (2000) being the most recent incident of this type. This shows the need for continued vigilance, and the need to go on learning from accidents.
8.4.2 Trends in the number of potentially higher-risk train accidents There were 49 PHRTAs in 2007 – a very similar level to 2006, but many fewer than were occurring in the late 1990s.
Each year derailments account for the highest number of PHRTAs. However, we can see from Chart 70 that train derailments have more than halved since 1998. The biggest step- change occurred in 2001, after the Hatfield accident of the previous year. To quote the 2000/01 ASPR: ‘The derailment at Hatfield […] not only reinforced rail breaks and track quality as a major safety concern, but provoked a recovery programme to address gauge corner cracking that severely disrupted services for months.’
Ironically, HSE and the (then) Rail Regulator had already commissioned the Transport Technology Centre Inc. (TTCI) to investigate broken rails and Railtrack’s management of them in August 2000. The report was published in the November of that year, one month after the accident. It (and of course Hatfield itself) led to the establishment of a taskforce to research metallurgy, wheel–rail interaction, brakes and suspension design, and ultrasonic rail flaw detection. The fruits of this initiative were borne in subsequent years, as Chart 70 shows.
Collisions that occur between trains have also reduced over the 10-year period.
40 Buckeyes prevent override and serve to maintain a derailed train in an upright and in-line position.
114 2007 Annual Safety Performance Report Train accidents
Chart 70. Trends in the numbers of PHRTAs
140 Trains striking road vehicles at level crossings Trains striking buffer stops 119 120 Trains running into road vehicles not at level crossings & no derailment 109 Train derailments (excludes striking road vehicles on level crossings) Collisions between trains (excluding roll backs) 100 100
80 71 66 64 59 60 51 49 46
40 Number of RIDDOR reportable of RIDDOR Number PHRTAs
20
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
8.4.2.1 Derailments There were 25 train derailments in 2007 – seven involving passenger and 18 involving non- passenger trains. One incident at Grayrigg resulted in a passenger fatality. The picture (courtesy of RAIB) shows how the structural integrity of the Class 390 was retained – a fact which probably minimised the fatality count.41
41 For more information on the Grayrigg accident, see RAIB, Rail Accident Report: Progress Report – Derailment at Grayrigg, Cumbria, 23 February 2007 (RAIB, October 2007).
2007 Annual Safety Performance Report 115 Train accidents
Chart 71 shows the number of derailments that have occurred on a yearly basis since 1998. It also highlights the spilt between passenger and the various types of non-passenger trains.
Chart 71. Trends in derailments by train type
60
50
40
30
20 RIDDOR reportable derailments RIDDOR reportable 10
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Passenger ECS Freight Light locomotive or RRV, OTM and other locomotives
The chart shows that freight trains remain the largest contributor to derailments. This is emphasised further when normalised by train miles. While freight trains account for 9.8% of all train miles, freight trains are involved in around half of derailments. The reasons for this disparity are varied and include a number of engineering and operational factors.
There has been a strong downward trend in freight derailments over the last 10 years. From the chart we can see that dramatic drops have occurred in 2001 and 2003. This may be due to the introduction of new, high-capacity bogie wagon types and the withdrawal of older two- axle designs, which are more prone to derail. A case in point may be seen in the derailment which occurred on King Edward Bridge, Newcastle, on 10 May 2007.42
Note that 2007 was the first year since 2000 to see no light loco derailments.
Historically, track defects have caused a high proportion of train derailments; as improvements have been made to the quality of the track, the number of consequent derailments has fallen. In 2007, five derailments were caused by cutting failures (these are shown under the Environment category in Chart 72). See section 8.5.2.1 for further details of track faults and other infrastructure failures.
42 This incident involved a two-axle HAA coal hopper, the left wheel flange of which rode over rail as a result of both a twisted rail and of ‘frame twist’. ‘Frame twist’ is a phenomenon more likely to affect shorter wheelbase vehicles, particularly the older variety (HAAs were built between 1964 and 1982). For details of the King Edward Bridge incident, see Rail Accident Report: The derailment of a freight train at King Edward Bridge, Newcastle 10 May 2007 (RAIB, 2008).
116 2007 Annual Safety Performance Report Train accidents
8.4.2.2 Derailments by cause
Chart 72. Derailments by cause
16
14
12
10
8
6
4 Number of RIDDOR reportable derailments reportable RIDDOR of Number 2
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Category A SPAD Environment Hitting object / Irregular w orking Miscellaneous / Possession w ith Rolling stock Track Train runaw ay Vandalism animal unknow n injury
8.4.2.3 Collisions between trains There were six passenger train collisions in 2007, all of which occurred at low speed. Three were roll backs, two involved station arrivals and one was due to an uncoupling error. No major injuries were reported.
There was also one non-passenger train collision, which was due to shunter error. This more serious incident occurred on 11 October 2007, when a freight train collided with a stabled wagon at Parkandillack (Western). Three wagons derailed, one ending up on its side, foul of the unprotected adjacent line. There was extensive track and sleeper damage, but no reported injuries.
Chart 73 shows the number of train collisions occurring since 1998 up to 2007. Again the split between passenger trains and non-passenger trains is demonstrated. The colours on the bars indicate the type of collision.
2007 Annual Safety Performance Report 117 Train accidents
Chart 73. Trends in collisions between trains by train type
14 Shunting in station 12 12 12 12 Setting back at station 12 Roll back 11 Possession 10 In running open track 10 Coupling on open track 9 Coming into station 8 8 8
6 6 5 4 4 333 RIDDOR reportable train collisions 2 2 2 1 11
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Involving a passenger train Not involving a passenger train
Though freight trains derail more frequently than passenger trains, most collisions involve the latter. However, most passenger train collisions are low risk, occurring at very low speeds as a result of roll backs or when pulling up to another unit in stations. There has been a drop in the number of passenger train collisions, with 2007 showing the lowest number over the period plotted.
An example of a roll back collision occurred on 15 October, when a Class 314 rolled back and collided with a Class 156 at Glasgow Central. The driver of the EMU released the brakes when preparing to depart and allowed his train to roll back into the DMU. There were passengers on both trains, but there were no reported injuries and no damage caused to the units.
8.4.2.4 Trains striking road vehicles Two of the most recent train accidents with passenger fatalities were caused by trains striking road vehicles: one was on the track after veering off an adjacent road (Great Heck), the other had been parked on a level crossing (Ufton Nervet).
There were eleven collisions between trains and road vehicles at level crossings in 2007. Three collisions led to three accidental fatalities to road vehicle users (and one suspected suicide).
• On 2 February, a road vehicle struck the side of a passenger train at Delny automatic open level crossing (locally monitored), near Barbaraville (Scotland). Two young men in the car died; the vehicle driver received major injuries. • On 1 March, a road vehicle was struck by a passenger train at Swainsthorpe automatic half barrier level crossing (South East), fatally injuring the driver. The barriers were down
118 2007 Annual Safety Performance Report Train accidents
when the car attempted to zigzag between them. The same crossing was the scene of a fatal collision in 2005. • The third fatal accident is being treated as a suspected suicide. On 23 February, an empty coaching stock train collided with a van on Gailes automatic half barrier level crossing (Scotland), killing the driver who was standing near the road vehicle. The crossing was working correctly at the time of the incident. See Chapter 9 for a detailed assessment of trends in level crossing safety.
Chart 74. Trends in vehicle incursions
60 Darker colours refer to vehicle incursions being struck by trains 50
40
30
20 Number of vehicle incursions
10
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Access point Bridge Fence Level crossing
Chart 74 shows that incursions through fences account for 55% of all incursions on the network. The level crossing category does not include collisions on level crossings, rather just incidents where the road vehicle user turns off the level crossing onto the track or where the road vehicle is stranded on the level crossing.
Vehicle incursions through access points have fallen in recent years, though 2007 saw that trend reverse. The events shown in the darker section of the bars indicate the vehicle incursions that were then hit by a train. These are also shown as not at level crossing incidents on Chart 75.
The numbers of train collisions with road vehicles, both at and away from level crossings, are shown in Chart 75. Vehicles can also gain access to the railway via the road or boundary fence. As well as the eleven accidents at level crossings there were a further six vehicle strikes at other locations.
2007 Annual Safety Performance Report 119 Train accidents
Chart 75. Trends in trains striking road vehicles
30
No derailment
25 24 Derailment 23 22
20 17 17 16 16 16
15
11 11
10 Trains striking road vehicles road striking Trains 6 6 5 5 5 4 4 22 2 1
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 At a level crossing Not at level crossing
• There were no derailments from incursions in 2007. The most recent incursion that resulted in a train derailment was at Copmanthorpe on 25 September 2006, when a vehicle crashed through a boundary fence and onto the line.
8.4.2.5 Buffer stop collisions There were three reportable buffer stop collisions in 2007:
• On 11 January, a passenger train stuck the buffer stop at West Kirby station (London North West). Extensive damage to the station forced its closure. No passengers were harmed, although the driver and guard suffered minor injuries. • On 10 October, a passenger train collided with the buffer stops at Sheringham (Anglia). Four passengers sustained minor injuries. • On 29 October, two empty coaching stock vehicles were coupled together and then struck the buffer stops at Liverpool Lime Street (London North West). There was damage to one of the couplers. The most recent fatal buffer stop collision occurred at Cannon Street in 1991. A passenger service collided heavily with the hydraulic buffer stops and the fifth carriage partially over- rode the sixth. Two people were killed and a further 542 injured. Since World War II, there have been five fatal buffer stop collisions, as Table 14 shows:
120 2007 Annual Safety Performance Report Train accidents
Table 14. Fatal buffer stop accidents since 1945
Date Location Passenger fatalities Staff fatalities 08/01/1991 Cannon St 2 0 26/02/1971 Sheerness 1 0 18/09/1953 Guildford 0 1 10/12/1947 Manchester Vic 0 1 02/09/1945 Haywards Heath 0 2
In the accident at Sheerness, the passenger who died was standing in the ticket office.
Prior to the war, the most serious buffer stop collision occurred on 27 July 1903: a passenger train struck the buffer stops at Glasgow (St. Enoch) at 10mph. Sixteen passengers were killed and 64 were injured.
Chart 76 shows the number of buffer stop collisions that have occurred annually during the last 10 years. The numbers (compared to other train accident types) are relatively small; indeed, there was only one incident in 2005 and again in 2006.
Chart 76. Buffer stop collisions
8
7 1 6
5 Non passenger train Passenger train 4
3 6 5 1 2 44 33 RIDDOR reportable buffer stop collisions stop buffer reportable RIDDOR 1 2 11 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Buffer stop collisions can still be a problem even on modern railways. For example, on 5 April 2007, between 30 and 50 people suffered minor injuries after a commuter service struck the buffers as it pulled into the Gare de l'Est, Paris. The train was travelling at low speed, but several people were thrown to the floor.
2007 Annual Safety Performance Report 121 Train accidents
In the UK, the overall risk from this type of incident has been reduced significantly by the introduction of TPWS at terminal platforms (fitment was completed at the close of 2003). However, incidents of this type still have the potential to lead to more serious consequences.
8.4.3 Trends in the number of non-PHRTAs by type Each year, some non-PHRTAs result in injuries. In 2007, seven train fires, one train striking an animal, four trains striking objects and five trains being struck by missiles led to physical injuries.
Two of the missile strikes resulted in major injuries to the workforce:
• On 30 April, a brick was thrown at a train near Salisbury (South East). The windscreen was smashed, glass entering both of the driver’s eyes. • On 21 August, a brick was thrown at a train from an overbridge at Philips Park (London North West). It smashed though the windscreen and hit the driver, resulting in a broken shoulder There was one passenger major injury, caused by a train running into an obstacle.
• On 18 January, a Sheffield-Huddersfield service struck a tree near Barnsley (London North East). It smashed a large body-side window and one passenger was knocked unconscious. Chart 77 shows that the number of non-PHRTAs has fallen by over 50% since 1998. This is due to a reduction in train fires and trains being struck by missiles. Major reductions took place each year between 2000 and 2003. A further reduction took place between 2004 and 2005, which has since been maintained.
Chart 77. Trends in number of non-PHRTAs
2000
Train striking object 1800 1717 Train struck by missile 1669 1669 Train striking gate or barrier at LC 1600 1539 Train striking animal Train fire Open door collision 1400 1304 Roll back collision
1200 1038 1006 1000
782 775 800 736
600 RIDDOR reportable train accidents
400
200
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
122 2007 Annual Safety Performance Report Train accidents
Reduced vandalism is one of the reasons for the reduction shown in train fires. Note also that most modern passenger stock is finished with fireproof materials, which make the ignition of fires more difficult. However, recent analysis of other non-SMIS data sources suggest a high level of under-reporting of train fires. Work is currently under way to identify and update these records. The reasons for the under-reporting are being investigated.
8.5 Trends in precursors to train accidents 8.5.1 The Precursor Indicator Model RSSB’s Precursor Indicator Model (PIM) provides a measure of the underlying risk from train accidents by tracking changes in the occurrence of accident precursors. It was first developed in 1999, and has since been subject to a series of modelling improvements, the last being a realignment with updated risk information from SRMv5.
Within the PIM, a train accident is defined as one of the following: train derailment, train collision, train striking buffer stops, train fire and train striking road vehicle at a level crossing. The SRM identifies 84 precursors to these accidents, which are categorised in the PIM into 26 subgroups. These fall into six main precursor groups, as shown in Table 15.
Table 15. PIM groups and subgroups
Overall indicator
3 - Public 1 - Infrastructure 2 - Irregular 4 - Objects on the 6 - Trains & rolling behaviour at level 5 - SPAD failures working line stock crossings
Level crossing Irregular loading of Misuse due to public Category A SPAD - Animals Brakes failures freight trains actions passenger train
Irregular working at Misuse due to Category A SPAD - Structural failures Non-rail vehicles Hot axle box level crossings weather non -passenger train
Irregular working by Track Objects blown onto Other rolling stock signallers the line failures
Wrongside signal Irregular working in Objects on the line Fires due to rolling failures stations due to vandalism stock failures
Environmental Runaway trains Fires due to vandalism
Train speeding Other train fires
Irregular working - other
The following sections look at the results and trends at the overall PIM level and at each precursor group level. The PIM value is an annual moving average, so at any point in time it reflects precursors that have occurred during the previous 12 months. It is also normalised by train miles, to account for changes in the level of activity on the railway.
The PIM tracks, from 1998 onwards, the underlying risk from a train accident every month. It uses the simple formula: risk = frequency x consequence.
The SRM models hazardous events (that is, those that could lead to harm on the railway). Each is broken down into the precursors that could lead to its occurrence. The SRM then
2007 Annual Safety Performance Report 123 Train accidents
estimates risk (in terms of the frequency and consequences) associated with each hazardous event and presents the results in terms of FWI per year.
Each month, the number of events for each precursor is derived from the data, and multiplied by the average consequences per event for that precursor to give a risk estimate for it. The risk from all 84 precursors is then summed and normalised per million train miles. The normalised figures are subsequently rebased against the annual average at March 2002. The risk level at the end of March 2002 is taken as the reference level for the PIM and is set at 100.
No changes were made to the modelling of the PIM during 2007, as the update of the SRM to version 5.5 was did not include an update of the train accident hazardous events.
To govern the PIM a practitioner working group, reporting to the Safety Policy Group, was created. Its two main aims are to ensure that the outputs meet the needs of the industry and to oversee developments to the PIM.
8.5.2 Trends in the PIM
Chart 78. Trends in PIM categories
120 6 - Trains & rolling stock 5 - SPAD 4 - Objects on the line 3 - Level crossing misuse 100 2 - Irregular working 1 - Infrastructure failures
80
60
46.7 47.9 PIM indicator
40
20
0 Jul Jul Jul Jul Jul Oct Oct Oct Oct Oct Apr Apr Apr Apr Apr Jan Jan Jun Jan Jun Jan Jun Jan Jun Jun Mar Mar Mar Mar Mar Feb Feb Feb Feb Feb Aug Sep Nov Dec Aug Sep Nov Dec Aug Sep Nov Dec Aug Sep Nov Dec Aug Sep Nov Dec May May May May May 2003 2004 2005 2006 2007
The overall level of the PIM, as shown on Chart 78, has been consistently just below 50 for the last 18 months. During 2007, there was an increase in the SPAD risk due to a couple of high risk-ranked incidents.
The following chart (Chart 79) breaks the PIM indicator down into person type and shows the trend in train accident risk for passengers, workforce and the public. It shows that the greatest reduction has been made to the risk to passengers, mainly due to the introduction of TPWS. When looked at alongside Chart 66, it clearly shows that different precursors affect
124 2007 Annual Safety Performance Report Train accidents
differing person types. Nowhere is this clearer than in level crossing misuse, where the majority of its impact is on the level crossing user.
Chart 79. Trends in PIM indicator by person type
60 Passenger indicator Workforce indicator Public indicator 50
40
30 PIM indicator 20
10
0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 125 Train accidents
Analysis of PIM subgroups in more detail Table 16 shows the value of the PIM indicator for each precursor group and subgroup at the end of 2007. It also shows the change to each value over the past year, and over the past four years. Those areas which have seen improvements over the year are shown in green, and those where the indicator value has risen are red.
Table 16. Changes in PIM indicator by precursor group
Change Change 2007 from 2006 from 2003 (PIM (absolute (absolute Group Sub group indicator) value) value) 1 - Infrastructure failures Environmental 2.9 0.8 1.8 Level crossing failures 0.3 0.0 0.0 Structural failures 3.9 2.2 2.2 Track 3.4 -1.5 -3.7 Wrongside signal failures 0.8 -0.1 -1.4 1 - Infrastructure failures total 11.3 1.4 -1.1 2 - Irregular working Irregular loading of freight trains 0.1 0.1 -0.3 Irregular working - other 1.2 0.0 -4.7 Irregular working at level crossings 0.9 0.0 -0.4 Irregular working by signallers 3.3 0.0 -0.4 Irregular working in stations 1.3 0.0 -4.1 Runaway trains 0.9 -0.6 -1.8 Train speeding 0.1 0.0 -0.2 2 - Irregular working total 7.9 -0.5 -11.9 3 - Level crossing misuse Near misses due to public actions 18.2 -1.2 -9.9 Misuse due to weather 0.1 -0.3 -0.4 3 - Level crossing misuse total 18.3 -1.5 -10.3 4 - Objects on the line Animals 0.3 -0.1 -0.4 Non-rail vehicles 1.9 0.2 -0.5 Objects blown onto the line 1.8 0.8 1.3 Objects on the line due to vandalism 1.8 0.4 -2.3 4 - Objects on the line total 5.9 1.3 -1.9 5 - SPAD Category A SPAD - non-passenger train 0.3 0.0 -0.8 Category A SPAD - passenger train 2.6 1.3 -7.2 5 - SPAD total 2.8 1.3 -8.1 6 - Trains & rolling stock Brakes 0.1 -0.2 -0.4 Fires due to rolling stock failures 0.4 0.0 -0.4 Fires due to vandalism 0.1 0.0 -0.6 Hot axle box 0.1 -0.1 -0.6 Other rolling stock failures 1.2 -0.3 -1.5 Other train fires 0.0 -0.1 -0.1 6 - Trains & rolling stock total 1.8 -0.8 -3.6 Total 47.9 1.2 -36.9
8.5.2.1 Infrastructure failures Last year was a year that saw an increase in the PIM value of the infrastructure failure group. Whilst there continued to be reductions in the number of broken rails (Chart 81) and ‘Level 2’ exceedances,43 there was also an increase in the number of cutting failures (Chart 82), due to adverse weather conditions.
43 The track is frequently checked for deviations in the geometry and any faults that could potentially cause a derailment are termed ‘Level 2’ exceedences.
126 2007 Annual Safety Performance Report Train accidents
In a six-day period in January 2007, there were four derailments due to running into landslips and/or trees. There was also an accident in early March:
• On 13 January, a passenger train derailed after striking a tree stump which had fallen down the embankment smashing through the timber retaining wall and onto the track with spoil and other debris at Merstham (South East). • On 15 January, a tree fall and landslip caused a passenger train to derail at Duncraig Halt (Scotland). • On 15 January, a retaining wall collapsed and caused a landslip which derailed a passenger train at Kemble (Western). • On 18 January, a passenger train struck a tree at Hookagate (Western) and derailed. • On 6 March, a passenger train struck a tree and derailed at Godstone (South East). The front bogie derailed and the train remained upright. All five of the events in 2007 involved passenger trains, which is more than the total number of passenger train derailments in 2006 (four).
In 2006, there were two derailments due to the environment: one was due to the subsidence under the track, the second due to a train running into trees blown onto the line. Both involved non-passenger trains.
The following chart shows the number of derailments since 2003, identifying those with environmental causes:
Chart 80. Derailments caused by environmental factors
25 Other derailments Environment related 2
20
15
23 10 21 21 17 18 113
RIDDOR reportable derailments reportable RIDDOR 5 5 6 6 5 4 2 0 Passenger Non Passenger Non Passenger Non Passenger Non Passenger Non train passenger train passenger train passenger train passenger train passenger train train train train train 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 127 Train accidents
Chart 81. Trends in track failures
1000 939 890 Broken rails 900 Buckled rails
800 717 700 658
600
500 444
400 379 333
Number of rail defects rail of Number 317 300 227 191 200 137 85 100 60 48 56 29 32 18 22 4 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Source: Network Rail
On Chart 81, the downward trend in broken rails is clear. The 191 that occurred in 2007 were just 20% of the number that occurred in 1999, representing the eighth continuous year of improvement. The number of track buckles was also very low in 2007 – due, in part, to last year’s relatively cool, wet summer.
Chart 82 shows the five-year trend in structural failures. There was an increase in most categories in 2007, mostly caused by adverse weather conditions.
Chart 82. Structural failures 140 bridges, tunnels etc conductor rail 128 cuttings OHLE 120 other structures station supporting embankments trees down 100
80 68 67 64 66 60
40 RIDDOR reportable structural failures structural reportable RIDDOR 20
0 2003 2004 2005 2006 2007
128 2007 Annual Safety Performance Report Train accidents
8.5.2.2 Irregular working In late 2006, the method for recording irregular working in SMIS was changed to capture information on activity and consequence. Currently, within the PIM, the irregular working measure has not been updated to use this new information; this is planned for later in 2008. Chart 83 shows the number of irregular working events recorded in SMIS during 2007, by activity. Most reported events occur during the signalling of trains (63%), whilst a further 16% occurred within possession and protections.
Chart 83. Irregular working events by activity for 2007
General Infrastructure Yard/Depot Working Electrical Isolations Maintenance (No 1% <1% Protection Needed) Train Working Operations 3% & Management 16% Red zone working 2% T4 possession <1%
T3 possession 12%
T2 protection 3% T12 protection Signalling trains <1% 63%
Even though the majority of events occur whilst signalling trains, very few of these result in any consequence; most are zero-risk wrong-routing events. To help identify the potential seriousness of events, work has started on a risk ranking tool based on the same principles as the SPAD risk ranking tool. Each irregular working event will be ranked and given two measures. A ranking category will assigned based on the probability that the event could have escalated to an accident. Secondly, the potential consequence will be ranked, so for example an irregular working event that could have resulted in a high-speed collision will have a high consequence ranking.
With these two pieces of information the irregular working event will then be categorised as potentially severe, potentially significant or low risk.
This new methodology is currently being piloted, with industry roll-out expected to be in April 2008.
8.5.2.3 SPADs Historically, train accidents resulting from category A SPADs have caused high numbers of fatalities and injuries. Recently this has not been the case. There was one derailment from a category A SPAD during 2007 and that was on a set of trap points. There were no resulting injuries – indeed, the last category A SPAD that caused a derailment leading to an injury was in late 2002. Two minor injuries resulted.
2007 Annual Safety Performance Report 129 Train accidents
Chart 84. Trend in SPAD risk
120% 600 SPAD risk as % of March 2001 baseline (two-year moving average) 100% Number of SPADs 500 (annual moving averagel) Annualnumber of SPADs
80% 400
60% 300
40% 200
SPAD risk as percentage of benchmark of as percentage risk SPAD 20% 100
13%
0% 0 2001 2002 200320042005 2006 2007
Source: Risk data is based on outputs from the SPAD risk ranking tool.
Chart 84 indicates that by the end of 2007 SPAD risk had reduced by 87% of the March 2001 baseline. The train protection and warning system (TPWS) fitment programme was completed at the end of 2003, and has been responsible for much of the reduction in SPAD risk.
Although SPAD risk remains low by historical standards, it rose slightly during the year. Each SPAD is assessed in terms of the potential risk that it represents, and given a score of between zero (no risk) and 28 (very high risk). The rise in SPAD risk during 2007 was largely the result of two SPADs with a particularly high risk ranking score (26).
• On 22 August, a passenger train passed SB2209 signal without authority at Didcot North Junction (Western). The overrun was 276 yards. Another passenger train was passing across the junction ahead of the signal at the time of the SPAD. The SPAD train had traversed the overspeed sensor for SB2209 (which has a set speed of 41.5mph) at 67.4mph, which caused an emergency brake application to be made. The train subsequently came to a stand beyond the conflict point shortly after the other train had cleared it. This incident remains under investigation by the RAIB. • On 14 September, a passenger train passed LU17 signal without authority at Lugton (Scotland). The overrun was 250 yards. Although TPWS operated as designed, the train passed the potential conflict point and there was the possibility of a head-on collision between two passenger trains where the maximum permissible speed is 70mph. Chart 85 shows the monthly number of TPWS interventions and activations since the end of 2003. An intervention occurs if the TPWS applies the brakes before the driver doing so (or in the absence of any braking by the driver). If the driver had already applied the brakes before TPWS operates, then it is an activation.
130 2007 Annual Safety Performance Report Train accidents
In addition to activations (green) and interventions (yellow), this chart shows the 15 incidents since January 2004 where the driver has reset the TPWS and continued forward without the signaller’s authority. The (purple) trend-line indicates the monthly total numbers of category A SPADs. The white sections on the bars relate to those incidents where it has not been possible to ascertain whether the driver or the TPWS equipment first initiated the brake application.
Chart 85. TPWS interventions and activations
Reset & continue Intervention I/A not known 50 Activation Category A SPADs 45
40
35
30
25
20
15
10
5
0 2004 2005 2006 2007
Chart 86 suggests that focussed attention, after a number of reset and continue events have occurred, reduces the rate of occurrence, but in time, as the message tires and begins to be discounted as a regular communication, the rate of occurrence rises again. It therefore remains important for driver briefings to be varied and continue to focus on minimising the potential for reset and continue.
The strategy for TPWS, which was established in 2004, identified eight areas where action was needed to minimise the number of unnecessary TPWS brake demands, simplify the reporting procedures, improve the ability of a driver to correctly diagnose a TPWS brake demand and ensure that TPWS is fitted at locations that provide safety benefit. Since 2004, significant progress has been made in all eight areas identified for action. It is anticipated that further progress will be made during 2008, particularly in the areas of optimising TPWS fitments at permanent speed restrictions and reaching conclusions on the potential for modifications to improve the in-cab TPWS indications and controls.
There were five SPADs which became reset and continue incidents during 2007. This compares to five in 2006 and two during 2005:
• 13 February, VR509 signal on the Down Main line at Taffs Well (Western). • 11 May, E486 signal at Princes St. Gardens, Edinburgh (Scotland).
2007 Annual Safety Performance Report 131 Train accidents
• 18 June, L868 signal at Marks Tey (Anglia). • 14 November, at HT270 signal on the Down Main line at Fratton (Wessex). • 27 November, at AD22 signal at Chislehurst (South East).
Chart 86. TPWS ‘reset and continue’
40 40% Number of days between interventions 35 % reset & continue events in previous 50 interventions Proportion of interventions that are interventions that Proportion of 30 30%
25 continue & reset
20 20%
15
10 10%
6%
Number of days between interventions 5
0 0%
3 4 5 6 7 0 0 0 0 0 - - - - - n n n n n a a a a a J J J J J Month and year
8.5.2.4 Public behaviour at level crossings There were 11 collisions on level crossings during 2007, the same number as occurred in the previous year. None resulted in an injury to passengers or the workforce. However, there were four fatalities to level crossing users (including one suspected suicide).
About 95% of the risk at level crossings occurs from the behaviour of the public (rather than workforce errors or equipment failures). This is measured in the PIM by near misses between trains and road vehicles, which are considered in detail in the level crossing chapter.
8.5.2.5 Objects on the line There was an increase in the number of trains striking objects blown onto the line, which was mainly associated with the adverse weather conditions encountered at the beginning of the year (these conditions also led to five passenger train derailments after the trains ran into landslides and trees).
Additionally, there was a marginal increase in the number of trains striking objects placed on the line by vandals.
132 2007 Annual Safety Performance Report Train accidents
Chart 87. PIM – objects on the line
9 8.5 Objects on the line due to vandalism Objects blown onto the line 8 7.7 Non-rail vehicles Animals 7
5.9 6 5.4
5 4.6
4 PIM indicator PIM
3
2
1
0 2003 2004 2005 2006 2007
8.5.2.6 Trains and rolling stock Trains and rolling stock defects contribute the smallest amount of train accident risk out of the six groups (just 3.7% in 2007). During the year, one derailment was caused by a rolling stock defect:
• A freight train derailed on 14 May at Stapleford & Sandiacre (London North Eastern) due to severe wheel flats. Additionally a twisted frame contributed to the derailment of a freight train at King Edward Bridge in May 2007, which passed over track with a twist in it.
The last fatal accident from a rolling stock defect occurred at Rickerscote in 1996. A freight train derailed because an axle fitted to one of its wagons completely fractured. Derailed wagons blocked the adjacent line and were struck by a post office train running in the opposite direction. The locomotive and front four coaches of this train became totally derailed and a Royal Mail employee who had been working on the train was killed.
8.6 Further information More information on category A SPADs and TPWS interventions may be found in our quarterly SPAD and TPWS reports, available on www.opsweb.co.uk.
2007 Annual Safety Performance Report 133 Train accidents
Intentionally blank
134 2007 Annual Safety Performance Report Level crossings
9 Risk at level crossings
This chapter investigates the risk that arises where roads, bridleways and footpaths meet the railway at level crossings.
Specifically, it looks at injuries and fatalities to passengers, staff and members of the public, near misses and misuse at level crossings and trains striking road vehicles at level crossings. It also includes relevant data on equipment failures, workforce errors and railway crime.
2007 Headlines • Nine pedestrians were struck and killed at level crossings in 2007; one was a passenger on a station crossing, eight were members of the public. This represents an increase compared to 2006, and is slightly higher than the long-term average. • Of the 11 collisions between trains and road vehicles at level crossings in 2007, three resulted in road vehicle occupant fatalities (including one accident where two car occupants died, and one suspected suicide). • For the third consecutive year, no trains derailed as a result of collisions at level crossings and there were no on-board fatalities. • Collisions and near misses with road vehicles remain at historically low levels. • There is no evidence of a reduction in collisions and near misses with pedestrians, although there was a fall in reports of pedestrians using crossings when it was not safe to do so. * Performance at a glance
Accidental pedestrian fatalities Road vehicle collisions 12 30
10 10 25 24 9 9
8 20 16 16 6 6 15 5 11 11 4 10
2 5
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 80% higher than 2006 2 The same as 2006 2 20% higher than 2003-06 average 2 35% lower than 2003-06 average 2
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 135 Level crossings
9.1 Level crossing risk profile Different types of level crossings offer different protection to users. There are two broad groups:
• Active crossings – where the road vehicle or pedestrian is warned of the approach of a train through closure of gates or barriers, or by warning lights and/or alarms. • Passive crossings – where no warning of a train’s approach is given, the onus being on the road user or pedestrian to determine whether it is safe to cross the line or not. Instructions for proper use must be provided at each location, along with other appropriate signage. Table 17 shows the different types of level crossing that make up these groups. Information on the level crossing population of Great Britain, along with an illustrated guide to the same, may be found in Appendix 5.
Table 17. Level crossing categories by class and type
MCG Manually controlled gate MCB Manually controlled barrier
Manual MCB-CCTV MCB protected by closed-circuit television AHB Automatic half-barrier
Active ABCL Automatic barrier locally monitored AOCL/R Automatic open crossing locally or remotely monitored
Automatic UWC-MWL User-worked crossing with miniature warning lights UWC-T User-worked crossing with telephone UWC User-worked crossing OC Open crossing Passive FP Footpath crossing
Chart 88 presents level crossing risk broken down by the cause of the risk and the type of level crossing user it affects. It can be seen that train occupants suffer a comparatively small proportion of the risk, nearly all of which is due to improper use of the crossing by road vehicle drivers. Improper use is the dominant cause of risk for all user types, and accounts for more than 90% of level crossing risk overall.
136 2007 Annual Safety Performance Report Level crossings
Chart 88. Level crossing risk by cause and user type
Improper use 2.39
Workforce error 0.10
Railway crime 0.02
Equipment failure 0.03
Other (proper use) cause 0.23
Improper use 7.82
Workforce error 0.00
Railway crime 0.00
Equipment failure 0.04 Fatal Other (proper use) cause 0.43 Major Improper use 0.68 Minor Shock/trauma Risk by type of user and cause of risk of and cause user of by type Risk Workforce error 0.03
Railway crime 0.03
Equipment failure 0.01 Train occupants Pedestrians Road vehicle occupants Other (proper use) cause 0.06
012345678910 Average risk per year
Source: SRMv5.5.
The distribution of level crossing risk by type of crossing, and type of user, is shown in Chart 89. The risk is shown in absolute terms, and normalised by the number of crossings of each type.
Overall risk reflects the number of crossings of each type on the network, the hazards associated with them, and number of people that use them. Footpath crossings account for most of the overall risk to pedestrians. Automatic half-barrier crossings (AHBs) and user- worked crossings with telephones account for most of the overall risk to road vehicle occupants. There are a large number of user-worked crossings, but most are installed on private roads. AHBs are typically more heavily used and are often sited on busier, high- speed main lines. The barriers on AHB crossings span only half the highway. This was ‘designed in’ to provide an escape route for stranded motorists, but additional risk arises from road vehicle drivers deliberately zigzagging in an attempt to ‘beat the train’.
Risk per crossing is a useful metric for safety management purposes, as it can help identify relatively high-risk sites. Automatic open crossings have the highest risk per crossing for road vehicle users because they are located on the public highway with no physical barrier to road traffic (although train speeds are lower than for AHBs and the train driver is responsible for ensuring that the crossing is clear on AOCL crossings).
2007 Annual Safety Performance Report 137 Level crossings
Chart 89. Level crossing risk by crossing and user type
14 14 Train occupant 11.86 12 Pedestrian 12 RV occupant 10 10
8 8
5.69 6 6 4.53 4.72 4.60 4 4 Average risk per year per risk Average 2.07 1.58 1.85 1.90 1.75 1.71 2 1.23 1.21 1.16 2 0.60 0.53 0.73 0.09 0.02 0.27 0 0 FP FP OC OC AHB AHB UWC UWC ABCL ABCL AOCL AOCL UWC+T UWC+T MCB/MCG MCB/MCG UWC+MWL UWC+MWL All level crossings level All All level crossings level All Absolute risk per year Risk per 1000 crossings
Source: SRMv5.5 for risk; Network Rail for crossing numbers.
9.2 Fatalities and injuries in 2007 9.2.1 Pedestrians Nine pedestrians were struck and fatally injured at level crossings in 2007; one was a passenger intending to catch a train, eight were members of the public. Details of the fatalities are as follows.
• On 1 January, a male was struck on Paggetts footpath (FP) level crossing (South East). It was reported that he may have been playing ‘chicken’ whilst under the influence of alcohol. • On 15 January, a 14-year-old girl was struck at Johnstown FP level crossing, near Ruabon (Western), whilst trying to retrieve her shoe. She was part of a group who appear to have been playing ‘chicken’. • On 24 January, a passenger was struck at Wokingham manually controlled barrier (MCB) level crossing (South East). The young male climbed over the barriers to join a train in the Up platform and was struck by a Down service. • On 25 January, an elderly woman was struck on Ballast Hole FP crossing, near Doddington Road (London North Eastern). • On 31 July, an elderly woman walking her dog was struck at Sandringham Avenue user- worked crossing with telephone (UWC-T) level crossing (London North West). • On 16 August, a young male jumped over the barriers at Horsham Road closed circuit television (CCTV) level crossing and was struck by a train (South East). • On 16 September, a member of the public was struck on Windwhistle FP crossing near Weston-super-Mare (Western).
138 2007 Annual Safety Performance Report Level crossings
• On 12 November, a young male was struck on Green Drift FP crossing near Royston (South East). • On 20 November, a man with a dog was struck by a train on Brimscombe FP UWC with miniature warning lights (UWC-MWL) crossing (Western). The man failed to acknowledge the warnings. Two pedestrians also received major injuries on level crossings in 2007. One occurred when a man chased his dog, which had broken loose from its lead, onto a footpath crossing. He was struck a glancing blow by a passing train. The other happened at Green Drift UWC- MWL crossing where a woman was struck by a train, receiving head injuries.
In addition, pedestrians (including pedal cyclists) incurred 23 minor injuries at level crossings. These were the results of slips, trips and falls and people being struck by crossing barriers.
9.2.2 Road vehicle users There were 11 collisions between trains and road vehicles at level crossings in 2007 (see Table 18). None of these caused a train to derail, or resulted in major injuries to those on board.
Table 18. Collisions between trains and road vehicles at level crossings in 2007
Date Location Territory Type Description A car driver - a local resident - claimed she did not see the 02 Jan Dingwall Middle Scotland ABCL flashing lights. Two young men in the car died and the vehicle driver received major injuries when a car struck the side of a passenger train. 02 Feb Delny Scotland AOCL The crossing, near Barbaraville, was working correctly, and the car driver admitted in court to causing the deaths by dangerous driving. The train struck a parked van. The fatality of the vehicle user - an 23 Feb Gailes Scotland AHB off-duty railway employee - is being treated as a suspected suicide. A car driver, who was lost, drove her car onto the crossing after 24 Feb Ffynnongain Western UWC-MWL opening one of the gates. After opening the other gate, she heard the train's whistle and vacated the area. A car driver died when he was hit by a train while attempting to 01 Mar Swainsthorpe South East AHB zig-zag between the crossing barriers, which were down. The same crossing was the scene of a fatal collision in 2005. The crossing was working correctly when the train struck a car. 12 Jun Chapel Western AOCL The car driver was airlifted to hospital with major, but not life- threatening, injuries. The train driver had stopped at the stop board and sounded his 21 Jul Beaver Hill Western OC horn, but was struck by a van while passing over the crossing.
A tractor struck the back of a train after the signaller gave the 30 Aug Abbey Farm LNE UWC-T farmer permission to cross, not realising the train was still in section.
A tractor driver did not stop to check that the crossing was clear 13 Sep Sibster Burn Scotland UWC (as instructed by the signage) and was unable to hear the train horn from the tractor cab.
18 Sep Beaver Hill Western OC A car drove into the side of the train.
A motorcyclist failed to see the flashing red lights in bright 03 Dec Bramfield South East ABCL sunlight, crashed through the barriers and was struck by the train.
Collisions resulting in fatalities are shown in red.
2007 Annual Safety Performance Report 139 Level crossings
However, two accidents resulted in fatalities to road vehicle occupants. These occurred at Delny AOCL crossing (Scotland), where the car driver sustained major injuries and two of his passengers were killed, and Swainsthorpe AHB crossing (South East), where the car driver received fatal injuries. A third fatal incident, at Gailes AHB crossing (Scotland), is being treated as a suspected suicide.
In addition to the car driver at Delny, there was one other major injury to a road vehicle user in 2007, at Chapel AOCL crossing (Western).
There was also one reported minor injury, to a motorcyclist who came off her scooter (no train was involved).
Two of the collisions occurred at the same location: Beavers Hill open crossing (Western). Prior to 2007, the previous collision at this crossing occurred on 24 February 2005.
9.2.3 Suicides There were 20 suicides or suspected suicides on level crossings in 2007. All were pedestrians, except one: the man who parked his van and was struck and killed by an ECS formation at Gailes AHB crossing.44 This method can lead to fatalities beyond the suicide (as happened at Ufton Nervet in 2004). Any suicide can severely traumatise the train driver involved.
9.3 Trends in injuries at level crossings 9.3.1 Level crossings FWI since 2003 Chart 90 shows the number of fatalities and weighted injuries at level crossings over the last five years. All the fatalities in the period plotted involved a train, with the exception of an incident in 2003 (in which an elderly man died after falling and striking his head on a crossing post).
The chart shows that most fatalities and injuries occur to crossing users, as opposed to train occupants.45
44 Though the victim in this case was outside his van at the time of the incident, he is still classed as a road vehicle occupant as he remained with said van while it was on the crossing. 45 Most users are members of the public; the incident at Wokingham on 24 January and the two girls fatally injured at Elsenham (2005) are classified as passengers because of their intent to travel; the youth killed at Gomshall (2004) is also deemed a passenger as he had alighted from a train just prior to the incident.
140 2007 Annual Safety Performance Report Level crossings
Chart 90. Harm at level crossings (excluding suicides)
Weighted injuries 20 Public Workforce 18 Passenger (on crossing) 16.9 Passenger (on train) Fatalities 16 15.5 Public 14.0 Workforce 14 Passenger (on crossing) Passenger (on train) 12.4 12 8 10
8 11 14 1 5.8 11 6 1 Fatalities and weighted injuries 4 5 5 2 2 1 0 2003 2004 2005 2006 2007
9.3.2 Fatalities by user type since 1998 Chart 91 shows the number of fatalities that have occurred at level crossings since 1998. The data has been broken down by pedestrians, road vehicle occupants and train occupants. The majority of fatalities over the last ten years (70%) involve pedestrians. (Note that suicides and suspected suicides are not included.)
Chart 91. Fatalities at level crossings (excluding suicides) 16 15 Pedestrian Road Vehicle 14 14 On board train 13 Total 12 12 11 11 11
10
8 8
6 6 5
4 Number of level crossing fatalities
2
0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 141 Level crossings
The year saw a number of accidental level crossing user fatalities that is just above the 10- year average of 10 (excluding train occupants). The total of 12, comprising nine pedestrians and three road vehicle occupants, is in contrast to the low total of five for 2006. Two of the road vehicle fatalities occurred in the same incident at Delny (AOCL) in Scotland.
9.4 Trends in collisions and near misses at level crossings 9.4.1 Collisions with road vehicles The risk from road vehicles being struck at level crossings includes the possibility of derailment (and consequent injury to the train occupants). Three accidents at level crossings in the last 50 years have resulted in passenger fatalities,46 and a further four caused fatal injuries to train crew. However, it is generally the road vehicle drivers or their passengers who suffer most harm when collisions occur.
• There were 11 collisions between trains and road vehicles at level crossings in 2007. This is the same number as 2006, and fewer than previous years. • For the third consecutive year, none of the collisions caused a derailment. Chart 92 shows a general downward trend in the number of collisions between trains and road vehicles over the last decade.
Chart 92. Collisions between trains and motor vehicles by crossing type
30 Passive Automatically protected 25 24 24 Manually protected 22
6 20 12 17 17 16 16 16 11 15 5 4 5 4 7 11 11
10 18 44
11 13 Train collisions with road vehicles road with collisions Train 12 11 11 12 5 8 5 7
2 0 11 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
The majority of collisions occurred at automatically protected crossings – AHBs, ABCLs, AOCLs and UWC-MWLs. The three ‘A’ types generally have higher volumes of traffic compared with UWC-MWLs. AHBs are typically sited on high-speed lines. ABCLs are found on low-speed lines and have a maximum speed of 55 mph on the approach to the crossing;
46 At Ufton Nervet (2004), Lockington (1986) and Hixon (1968).
142 2007 Annual Safety Performance Report Level crossings
they require the train driver to check if the crossing is clear before proceeding. This is also the case for AOCLs (although they do not have a physical barrier over the crossing).
Passive crossings (UWC, UWC-T, OC and FP) see the second highest number of collisions. These particular interface types do not provide audible or visual warnings. It is the responsibility of the user to detect the presence of an approaching train and decide whether or not it is safe to cross the line. This could be achieved by contacting the signaller by the telephone provided, or by visually checking that there are no trains in the area.
There were no collisions at manually protected crossings in 2007. These are controlled with gates and barriers that extend over the entire interface and are operated by staff locally (MCG, MCB) or via remote location (MCB-CCTV).
9.4.2 Near misses with road vehicles Because of the relatively small number of accidents that occur at level crossings, it is difficult to monitor trends and identify patterns using accident data alone. Therefore, the industry also studies data on near misses. Typically, near misses are reported by a train driver who has to take action to avoid a collision, or feels that he or she came close to striking a road vehicle or pedestrian. There is therefore an element of subjective judgement regarding what constitutes a near miss, and the prevailing light and visibility conditions may also affect the likelihood of an event being observed.
Chart 93 shows reported near misses with road vehicles at level crossings since 2003. There has been a statistically significant downward trend over the five-year period. One of the initiatives during this time has been Network Rail’s Don’t run the risk awareness campaign, which began in May 2006. Its initial focus was mainly on vehicle users, and it may have contributed to a reduction in near misses over the last two years.
Chart 93. Trends in reported near misses with motor vehicles
250 222 Unknown Passive Automatically protected 200 Manually protected 173 176 89 160 145 150 56 67 56 50
100 91 80 72 69 64 50
Near misses between trains and road vehicles and road trains between misses Near 41 34 39 33 31 0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 143 Level crossings
9.4.3 Near misses with pedestrians The number of reported near miss incidents involving pedestrians has fluctuated over the five-year period shown in Chart 94.
The chart shows that there had been an increase in the totals between 2003 and 2005. However, whilst the reports filed in 2006 represented a small decrease, 2007 saw a return to the level seen in 2005. This is mainly due to the rise in near misses reported at passive crossings.
In May 2007, Network Rail’s Don’t run the risk campaign began to focus on pedestrian safety. As more data becomes available, it will be interesting to see whether the apparent downward trend in near misses (and collisions) with road vehicles can also be achieved for pedestrians.
Chart 94. Trends in reported near misses with pedestrians
300 Unknown Passive 250 Automatically protected 240 235 Manually protected 221 209 203 200 100 88 101 70 82 150
71 100 64 66 70 80 Near miss with pedestrians with miss Near
50 68 67 64 54 53
0 2003 2004 2005 2006 2007
9.5 Factors affecting the risk at level crossings 9.5.1 User behaviour Most collisions are caused by errors or, sometimes, deliberate violations by crossing users.
Misuse refers to a variety of situations in which crossing users attempt to traverse a crossing when it is unsafe to do so, or otherwise fail to use it correctly. Typical reports involve road users jumping red lights and trying to beat the barriers, pedestrians ignoring warnings (including cases of people playing ‘chicken’), and, at user-worked crossings, gates being left open, telephones left off the hook, and users failing to report when they are clear of the
144 2007 Annual Safety Performance Report Level crossings
tracks. Misuse is thought to be considerably under-reported, particularly at crossings that are not monitored.
Chart 95 shows the percentage of reported crossing misuse by crossing type and the nature of the misuse involved. User-worked crossings with telephones account for almost one-third of reports of misuse. Most incidents at these crossings relate to users either failing to report when the crossing is clear or leaving the crossing gates open. These acts do not necessarily endanger the misuser first hand, but they have the potential to cause disruption and to endanger future users of the crossings.
Chart 95. Reported level crossing misuse by misuse category and crossing type
footpath 6%
UWC-MWL 5%
UWC-T 31%
UWC 9%
OC 1%
AOCL/R 6%
ABCL <1% Road vehicle crosses when unsafe AHB 11% Pedestrian crosses when unsafe Road vehicle strikes / struck by crossing equipment MCB-CCTV 15% User fails to report clear Phone left off the hook MCB 12% Gate / barrier left open / raised MCG 4% Other
0% 5% 10% 15% 20% 25% 30% 35% Percentage of misuse reports 2003 - 2007
At the more serious end of the misuse spectrum are incidents where a pedestrian or vehicle occupant uses the crossing when it is unsafe to do so. Examples include crossing users trying to beat the lights or pedestrians running across when the barriers are lowering. Such incidents are not always classified as a ‘near miss with a train’. The number of reports of motorists and pedestrians misusing crossings in this way fell during 2007, as Chart 96 illustrates.
Manually protected crossings are likely to have relatively high reporting rates, as they are monitored by railway personnel. It is difficult to gain a realistic picture of the misuse problem as many incidents go unreported at other crossing types.
Network Rail has put a lot of effort into raising users’ awareness of safety at level crossings, and the Don’t run the risk moved its focus towards pedestrians in 2007. This could explain the significant decrease in the number of pedestrians crossing when it is unsafe. It is worth remembering, however, that near miss incidents involving pedestrians saw a slight increase in 2007 (see Chart 94).
2007 Annual Safety Performance Report 145 Level crossings
Chart 96. Trends in reports of users crossing when unsafe (including near misses)
900 Unknown Manually protected 800 780 777 766 Passive 723 Automatically protected 700 637 640 638 648 600 354 379 349 212 354 500 249 480 257 245 400 400 169 189 223 140 300 200 181 152 274 261 277 200 206 170 235 100 199 191 211 216 140 129 87 101 97 0 Reports of users crossing when unsafe (Incl. near miss) near (Incl. unsafe when crossing users of Reports 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
Road vehicle Pedestrian
9.5.2 Equipment failure Equipment failure accounts for a small proportion of the risk at level crossings (see Chart 88), the risk being mitigated by the fact that equipment is designed to ‘fail safe’; that is, it automatically defaults to a position ensuring safety. For example, if the equipment fails at an automatic level crossing, the warning lights operate and the barriers lower.
RIDDOR-reportable wrongside failures (ie, where equipment fails in a dangerous condition) are recorded in SMIS. Over the last few years, the industry has increased such reporting because of changes in:
• The definition of a failure. • The definition of situations considered potentially hazardous to the user. Network Rail issued guidance on the reporting of level crossing equipment failure in 2005; this might also have had an effect on the statistics by increasing the number of incidents recorded.
Chart 97 shows the number and type of reportable equipment failures over the last five years. For the reasons stated above, the significant increase since 2004 should not be interpreted as more frequent equipment failures; it is almost certainly the result of changes in reporting.
The causes of equipment failure range from minor component defects, to more serious disruptions caused by power cuts and technical faults. There is also the issue of vandals causing damage and stealing cable (see section 9.5.5), as well as damage caused by road
146 2007 Annual Safety Performance Report Level crossings
traffic accidents. Environmental influences play a further part in failures, particularly when strong winds damage equipment, and flooding or lighting strikes cause electrical faults.
Chart 97. Trends in RIDDOR-reportable equipment failure (excluding telephone faults)
136 140 Other (including general system failures) CCTV system (including floodlights) 120 Lights (including white lights) Gates 100 Barriers Crossing surface 86
80
62 60
Number of reports of Number 41 40 33
20
0 2003 2004 2005 2006 2007
The number of reported telephone faults can be seen in Table 19.
Table 19. RIDDOR reportable telephone faults 2003 2004 2005 2006 2007 12 9 165 444 475
9.5.3 Trains striking gates or barriers In general, trains only strike barriers when a previous incident, such as a road traffic accident, has caused them to be foul of the line immediately prior to the train’s appearance. Crossing gates are usually struck when high winds cause them to blow open, either due to defective clasps, or users failing to close or secure them properly after passing.
Chart 98 shows the number of trains striking level crossing gates or barriers. There is some evidence of a reduction over time, but the numbers fluctuate from year to year.
There were no injuries reported from trains striking gates or barriers in 2007. The last incident to cause injuries occurred on 15 November 2004, when a passenger train struck the gates at Rowston MCG level crossing (London North East); there was one major injury to a passenger.
2007 Annual Safety Performance Report 147 Level crossings
Chart 98. Trains striking gates or barriers at level crossings
12 Passive 10 Automatically protected 10 Manually protected
8 7 7 5
6 6 6 2 5 5 4 3 4 3 4 2 2 3 2 4 4 1 2 2 2 33 3
Trains striking barrier or gate at level crossing level at gate or barrier striking Trains 2 2 1 1 1 1 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
9.5.4 Workforce error At the end of 2006, a new field was added to SMIS to allow for the categorisation of level crossing incidents caused by irregular working. These have been grouped into 12 main categories (see Chart 99).
Chart 99. Workforce errors affecting level crossings in 2007
Crossing not clear when Crossing not clear when gate/barriers opened gate/barriers closed 3% 20%
Crossing cleared for Signaller error use when not safe (wrong route set) 17% 10%
Marker board Technician reverts signal wrongly placed to danger in error 3% 3% Signaller error (SPAD) Signaller reverts signal to 7% danger in error 13% Signaller failed to caution trains 7% Hand signaller error 3% Signaller Signaller failed to miscommunication protect crossing 7% 7%
148 2007 Annual Safety Performance Report Level crossings
There were 30 reported workforce error incidents in 2007. The highest proportion (20%) occurred due to signallers completely lowering the barriers before the crossing was clear (either trapping pedestrians or vehicles on the interface). Most incidents of irregular working relate to errors made by signallers and crossing keepers.
Irregularities also occur when individuals fail to follow rules/regulations or when ambiguous/incomplete information is supplied. Near miss incidents, or even collisions, can occur when a signaller gives permission to cross when it is not safe to do so. On 30 August, for example, the rear end of a train was struck by a tractor as it went over Abbey Farm UWC- T. The signaller had given the farmer permission to use the crossing, but failed to notice that a passenger train was still in the section.
There has not been a fatality due to workforce error at a level crossing for 38 years. On 15 July 1970, a lorry was struck on Shalmsford Street level crossing near Canterbury (South East). The accident occurred when the crossing keeper failed to close the gate to road traffic and a lorry was stuck by an EMU. The lorry driver and a train guard were both killed.47
9.5.5 Railway crime Crime at level crossings is a serious issue, which has the potential to cost lives, as well as cause delays and unnecessary cost to the industry. Table 20 shows that the number of recorded vandalism events at level crossings has significantly decreased since 2003.
Table 20. Number of recorded vandalism events at level crossings 2003 2004 2005 2006 2007 249 208 127 179 105
Generally, these incidents involve members of the public interfering with level crossing equipment. This is achieved by either defacing signs or by causing extensive and expensive damage to gates, barriers, telephones and the like. In some cases, it can involve theft of railway property, such as copper cables. This can have serious implications on the safety of legitimate crossing users. An example can be seen in the photograph; the board becomes difficult to interpret for those who are not familiar with level crossing safety signs.
Suicides at level crossings show no significant trend since 2003. (See Chapter 7, Public safety, for more detail on railway crime and suicides.)
47 In fact, the accident ‘was caused by the crossing keeper who, contrary to his instructions, telephoned the signalman that the gates were closed, knowing that the Down side gate was still open’. This is according to the inquiry led by Lieutenant-Colonel A. G. Townsend-Rose, who added: ‘That [the keeper] drank his tea before going to close the gate made his omission the more reprehensible. […] I believe that [his] lack of concern when he drank his tea was probably due to the fact that he was not expecting [a] special train.’ Department of the Environment, Railway Accident: Report on the Accident that occurred on 15th July 1970 at Shalmsford Street Occupation Level Crossing near Canterbury (HMSO, 1971), p. 6, para. 22.
2007 Annual Safety Performance Report 149 Level crossings
Table 21. Number of suicides recorded at level crossings 2003 2004 2005 2006 2007 14 20 23 23 20
9.6 Further information Further information on the risk at level crossings (and from road vehicle incursions onto the railway) can be found in RSSB’s latest Road–rail Interface Safety Performance Report, which was published in January 2008 and can be downloaded gratis from the RSSB website http://www.rssb.co.uk/pdf/reports/road-rail_interface_spr_full.pdf.
150 2007 Annual Safety Performance Report Personal security
10 Personal security
This chapter discusses crimes against the person, including physical assaults, threats and verbal abuse aimed at members of staff, passengers and members of the public. The analyses provided here are based on data from SMIS and CRIME, the BTP information system.
2007 Headlines • One passenger was fatally injured in an assault during 2007, compared with two in 2006. • The total harm from assaults suffered by workforce members fell for the third successive year in 2007, registering a 16% drop compared to 2006. • BTP-recorded crimes against the person of robbery, assaults and sexual offences all recorded fewer incidents in 2007, compared with 2006.
* Performance at a glance
Workforce assaults (harm in FWI) Assaults recorded by BTP 5 12000 4.38 10783 10720 4.5 4.09 4.02 10104 10000 9564 4 9337 3.44 3.5 8000 2.89 3 2.5 6000 2 4000 1.5 1 2000 0.5 0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 16% lower than 20063 11% lower than 2006 33 28% lower than 2003-06 average33 7% lower than 2003-06 average 33
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 151 Personal security
10.1 Personal security in context For workforce assaults, SMIS contains good data on the numbers of incidents and injuries, and has been used to estimate the risk for SRMv5.5. Figures quoted in the RPB regarding HEN-65: Workforce assault are thus considered to be a good estimate of the risk.
For passenger and public assaults, BTP data (recorded in the CRIME system) is a good source of information on the overall numbers. However, it does not contain information on injury degree (fatalities, major and minor injuries).
Conversely, SMIS contains information on injury type and exposed groups, but only for a subset of assaults (passenger and public assaults being under-reported in SMIS). Therefore, to estimate the risk associated with passenger and public assaults for SRMv5.5, it was necessary to use a combination of the information contained in SMIS (for injury degree) and CRIME (for total numbers of assaults). This resulted in an estimate of around 8.5 FWI per year (which should, of course only be used as a broad indication of the likely risk profile).
Millions of passenger journeys take place every day and the likelihood of being involved in an assault of any type is very small. To put it into context, using the number of passenger assaults reported to BTP, the assault rate is around 1 in 115,000 passenger journeys.
Because data on assaults to passengers is limited, Chart 100 looks only at the workforce risk profile. The chart shows that:
• 11% of all risk to members of the workforce is due to assaults. • The problem is more prevalent in stations than on trains. • The risk profile for is fairly evenly split between major injuries, minor injuries and shock/trauma. • Nearly half of all assaults occur to train crew.
Chart 100. Workforce assault risk in the context of workforce risk as a whole
Trackworker 0.6% Other Train driver 1.2% 4.6% 100% Shock Trains trauma Other 41.4% 30.1% worker 46.0% Other workforce risk Assault Minor 89% (29.6 FWI) 11% (3.5 FWI) 37.5% Other Stations train 57.4% Major crew 32.4% 48.8% 0% 0.0% 0.0% Location Severity Worker type
Source: SRMv5.5.
152 2007 Annual Safety Performance Report Personal security
10.2 Fatalities and injuries in 2007 There was one passenger fatality resulting from assault in 2007 (one fewer than the 2006 number):
• On 15 December, a 17-year-old male was fatally assaulted at Dewsbury station (London North East). The man was stabbed to death after being set upon by a group of youths. Two teenagers have been charged with murder. There were also 18 major injuries suffered by assault victims in 2007, seven of which were to members of the workforce. Examples may be found in the next section.
10.3 Trends in workforce personal security Violence at work is a significant issue, but it is not unique to the rail industry. As with other public-facing services (the NHS, for example), staff assaults occur every day. Attacks can take the form of verbal abuse and threats, which can result in shock, or actual physical assault, which can also result in shock, but carries the added potential for physical injury. Absence from work may result from any type of assault.
As noted above, seven members of the workforce suffered major injuries from assault in 2007. The following three examples show the serious nature of these incidents:
• On 9 January, a member of station security staff sustained a perforated ear drum and a broken shoulder after being assaulted by a member of the public who had been ejected by BTP and station staff the previous night. • On 24 November, a member of station staff confronted a group of youths and was struck by one of them, causing him to fall backwards down a flight of steps. The employee sustained a fractured skull. • On 30 March, a train manager was slapped and punched in the face when refusing to allow a male customer to board as he was closing the doors during train despatch. Two members of platform staff who were assisting in restraining the assailant fell onto the platform, one sustaining a fractured elbow in the process. Workforce assault data
Incidents are recorded in SMIS when ‘in circumstances related to their work, a member of staff is assaulted, threatened or abused, thereby affecting their safety or welfare.’ This is wider than the definition of assault used by the police; SMIS therefore provides a richer source of information for the analysis of workforce personal security issues. In particular, verbal abuse is unlikely to be included in BTP assault figures (although it may appear in other categories, such as harassment).
Analysis on the total number of workforce assaults alone can be misleading. Although there has been an increasing trend in the number of assaults reported over the analysis period (2003–2007), there is no indication that the risk from workforce assault has risen. Indeed, the decreasing number of reported assault-related injuries evident in the charts provided here would tend to imply the opposite. Having said this, any level of workforce risk from this source is unacceptable, and the continued willingness of staff to report all assaults – no matter how small or insignificant they may seem – is vital, as it allows industry members to gain a full insight into the extent of the problem and help to clarify the actions required to address it.
2007 Annual Safety Performance Report 153 Personal security
Although the reporting of staff assaults has improved in recent years, there is evidence of uneven practice across the industry. Whilst many factors are likely to influence the relative numbers of physical and non-physical assaults to which a given operator’s staff are subjected, the magnitude of the difference suggests that different reporting regimes are in use.
10.3.1 Harm to the workforce from assaults Chart 101 looks at the harm arising from SMIS-recorded workforce assaults. The chart shows that:
• The overall harm suffered by members of the workforce from assaults has fallen to under three FWI for the first time. There was an overall decrease of 16% in 2007, compared with the previous year. • The harm from physical injuries has fallen every year since 2004, recording a drop of 34% over this period.
Chart 101. Harm from workforce assaults
5 Shock/trauma 4.38 4.5 Minor 4.09 4.02 Major 4 1.01 3.44 3.5 1.13 1.12
2.89 3 1.03
1.56 2.5 1.07 1.76 1.50 2 1.41 1.5 fatalities and weighted injuries 1.12 1 1.8 1.4 1.2 0.5 1 0.7
0 2003 2004 2005 2006 2007
154 2007 Annual Safety Performance Report Personal security
10.3.2 Workforce assaults by type Chart 102 shows the number of assaults on railway employees that were reported into SMIS between January 2003 and December 2007. Note that:
• Physical assaults have fallen for the second year in succession. There was a 9% fall in 2007; the drop since 2005 is 16%. • The number of reports of verbal abuse and threats continued to rise. This reflects the success of industry initiatives to improve the reporting of all assaults, although it is possible that there has also been an actual increase in the occurrence of these incidents. • The total number of workforce assaults reported into SMIS has remained at almost the same as the level seen in the previous two years; however there are still some inconsistencies in reporting these incidents among industry stakeholders. • On average, around 13 assaults on members of the railway workforce were reported each day in 2007; of these, around five per day were physical.
Chart 102. Workforce assaults by type
6000 Verbal abuse Threat Physical 4903 4865 5000 4838
4068 3880 4000 2045 2206 2230
1647 1462 3000
543 607 295 235 736 2000 Total number of assaults workforce 2123 2186 2250 2090 1000 1899
0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 155 Personal security
10.3.3 Workforce assaults by severity Chart 103 highlights the total number of workforce assaults, focusing on the severity level of the resultant injuries. It indicates that:
• At seven, the number of workforce major injuries from assault is the lowest over the analysis period. Indeed, 2007 is the third consecutive year where an improvement has been seen. • There has been a significant decreasing trend in the number of minor injuries over the same period, with a fall of 16% being evident in 2007, compared with 2006. • The incidence of shock/trauma has shown no real trend over the five-year period, remaining largely stable, while the recorded number of non-injury assaults has shown a dramatic increase. This illustrates the success of industry efforts to improve the reporting rate of the less significant assaults in order to gain a clear picture of the true extent of these incidents. It also suggests that staff are becoming better equipped to deal with conflict situations, as fewer incidents proceed to physical conclusions.
Chart 103. Workforce assaults by injury severity
3500 3001
3000 2929 2680
2500
2000 1998 1632 1500 1134 1118 1102 1070 1038 1030 1026 1014
1000 934 787 Number of workforce assaults workforce of Number 500 12 18 14 0 10 7 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Major Minor Shock/trauma No injury
156 2007 Annual Safety Performance Report Personal security
10.3.4 Time lost as a result of workforce assaults Physical injury frequently requires a victim to take time off for recovery – a broken leg from being pushed downstairs or stab wound from a knife attack will not heal overnight, for example. But neither will the mental scars that often accompany these events: imagine the fear that returning to the scene of the crime might evoke, or the fear of seeing one’s attacker again. Most companies offer their employees a support package, in an attempt to minimise the traumas that go all too frequently with working on the front line.
Chart 104 illustrates the number of days lost due to staff absence as a result of assault or abuse in each of the years since 2003. It shows that:
• Most time lost is due to physical assault. However, much time is also lost via shock/trauma suffered by the victim in incidents where no actual physical injury was sustained. It is also important to remember that shock or trauma following a serious assault often takes longer to rehabilitate than the physical injuries sustained during the attack. • Days lost in 2006 fell 9%, compared with the number recorded in 2005. It is important to remember that the duration of long-term absence is only recorded in SMIS once the affected staff member returns to work. It may thus take many months to appear. For this reason, the number of days lost in 2007, which seems much lower than previous years, should not be considered significant at this early stage of analysis. • On a typical day, more than 10 rail staff are absent from work recovering from an assault.
Chart 104. Trends in time lost from assault
6000
Verbal 4942 Threat 5000 4654 Physical 4484
4096 4000 3702
3000
2000
1000 number of days lost as a result of workforce assault workforce of a result as lost days of number
0 2003 2004 2005 2006 2007
2007 Annual Safety Performance Report 157 Personal security
10.3.5 Further analysis of workforce assaults 10.3.5.1 Location Chart 105 looks at SMIS-recorded workforce assault trends on trains and in stations. The chart only shows incidents where an injury has resulted from the assault, either physical or shock. It reveals that most physical injuries are sustained in stations.
There are two possible explanations for this. First, there may be a belief on the part of the perpetrator that escape will be easier from a station. Secondly, many stations have revenue protection staff, and ticket disputes are the number one reason for assaults, as Chart 107 highlights. BTP believes that a correlation exists between the number of checks performed at stations and the assault levels therein.
Both locations are showing signs of improvement in terms of actual harm suffered (this being a better measurement of the true trend, as it is influenced less by changes in reporting).
Cases of shock and trauma (without physical injury) are more frequently reported on trains than in stations. These typically result from verbal abuse, threats and less severe physical assaults.
Looking at the harm suffered by victims of these attacks shows that the workforce risk from assault has fallen significantly at stations in the five-year period under consideration.
Chart 105. Trends in workforce assaults by location
1000 3.5 Major 900 Minor 3 Shock/trauma 800 Total FWI
700 2.5
600 1.8 2 500
1.5 400 (FWI) Harm 1.1
Workforce assault injuries assault Workforce 300 1
200 0.5 100
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Station On-train
10.3.5.2 Occupation Many assaults occur on stations, where platform staff are prone to disputes over train times and ticketing – often fuelled by alcohol. Revenue protection officers, train crew, in fact any
158 2007 Annual Safety Performance Report Personal security
person involved in face-to-face contact with the public, is at risk; this can even include level crossing keepers (as occurred in 2006).
Chart 106 illustrates the occupations of members of the workforce who have been the victim of an assault from 2003 to 2007. It shows that:
• Members of the on-board train crew are victims of an assault in just under half of all incidents. It is worth mentioning that the segment labelled ‘revenue protection staff’ will also include people working on trains. • One in five assaults on workforce members involve station staff. Again, a proportion of the revenue protection staff will be working in stations, so this number is likely to be higher. • A higher proportion of staff working on trains is involved in verbal abuse incidents that result in no injury, usually following a ticket dispute.
Chart 106. Assaults by occupation: 2003–2007
Office staff Train drivers 3.0% Not entered 3.8% 0.8%
Other staff 11.6%
Onboard train crew (excluding drivers) Revenue protection 43.0% staff 18.0%
Station staff 19.7%
10.3.5.3 Instigating event Chart 107 illustrates the factors recorded by staff that they believed triggered the assault.
The chart shows that the greatest factor by far relates to ticketing disputes, with just under 40% of all workforce assaults deemed to have taken place for this reason. According to BTP, the number of crimes reported at ticket barriers and incidents reported as revenue- related have remained fairly static in 2007. Train operators have made concerted efforts to improve their revenue collection methods, through increasing the use of revenue protection officers and widespread education campaigns to ensure the traveling public are aware of the requirement to travel with a valid ticket. It is possible that improvements in the recording of crime and an increased police presence as a result of intelligence led operations have led to a higher proportion of revenue-related assaults being recorded by the BTP.
2007 Annual Safety Performance Report 159 Personal security
Alcohol also plays a significant role in workforce assaults. While 17% is a high percentage, it is worth noting that alcohol is also recorded as an aggravating factor in many of the ticketing dispute incidents.
The challenge faced by stakeholders is illustrated by the number of different assault catalysts. Among the many potential causes are train delays, racism, sexual motivation, thwarted vandalism sprees, crowding, and the replacement of services by buses. Even a missed train can become the subject of an outburst against blameless staff members.
Chart 107. Contributory factors to assaults on rail staff: 2003–2007
Service Not entered problems/missed 3% Racial connection 2% 6% Anti-social behaviour 3%
Unknown 9%
Drugs/alcohol related Ticket Dispute 17.0% 39.7%
Other 20.8%
‘Other’ includes: Sexual, domestic, homophobic, overcrowding, and other.
10.3.5.4 Temporal variation Chart 108 counts all of the workforce assaults that took place between 2003 and 2007 and shows the profile of when they took place during the week. The chart shows that:
• The highest rate of workforce assault takes place on Saturday evenings, followed closely by the Friday rush-hour and Friday evenings. It is widely accepted that alcohol plays a significant part at these times. Although the initial assault may have started due to a ticket irregularity, in many cases alcohol serves to exacerbate the incident. • There is also a ‘mini-spike’ during the morning, at which time many passengers are rushing to catch trains to work and overcrowding is also prevalent in some areas. Service disruptions, missed trains and ticket irregularities are often contributory factors in these incidents.
160 2007 Annual Safety Performance Report Personal security
Chart 108. Workforce assaults by time of day and day of week: 2003–2007
400 8pm-9pm 350 5pm-6pm 300
250
200
150
100
Workforce assaults 2003-2007 assaults Workforce 50
0 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midnight - 1am Midnight - 1am Midnight - 1am Midnight - 1am Midnight - 1am Midnight - 1am Midnight - 1am Sunday Monday Tuesday Wednesday Thursday Friday Saturday
10.4 Trends in passenger and public personal security Reflecting the wider social context, fatalities and very serious injuries resulting from crime unfortunately do occur on the railway from time to time, as the fatal assault that occurred in 2007 illustrates. However, there is a danger that the level of media coverage that follows these events could distort public perceptions of personal security to a point where they are not in line with the actual risk. BTP data indicates that the probability of being a victim of violent crime during the average journey is less than 1 in 100,000. The issue of public perception will be explored later in this chapter.
BTP assault data
Assaults on passengers and members of the public are not well recorded in SMIS. British Transport Police’s CRIME database provides a much richer source when analysing these incidents. CRIME data is therefore examined when looking at the incidence of assaults as a whole, with SMIS data being used for the purely workforce-related analyses.
RSSB is currently undertaking a research project (T723) to gain greater understanding of the differences between the events reported into these systems and to provide guidance to the industry on understanding crime data. The findings from this research are likely to be produced towards the end of 2008.
It should be noted that the way in which police record violent crime has changed over the analysis period, largely in response to the introduction of the National Crime Recording Standard (NCRS) in April 2002. The Home Office believes that the apparent increase in less serious violent offences in the years immediately following this were largely the result of
2007 Annual Safety Performance Report 161 Personal security
changes in the reporting and recording of crime in response to NCRS and ongoing audit and improvement schemes.
10.4.1 Personal security incidents Chart 109 provides an overview of recent trends in BTP-recorded incidents of robbery, assault and sexual offences. It shows that:
• The total BTP-recorded robberies fell in each of the last five years; the 41% fall in recorded robberies observed in 2007 means that recorded robberies are now at their lowest level over the analysis period. BTP believes that the reduction in robberies is due to a combination of factors, including actionable intelligence, increased use of covert human sources, better management of the robbery detection and prosecution process and the expansion of digital CCTV across the network (particularly on-board CCTV). • The number of assaults also fell in 2007. The 11% decrease in these incidents is the first significant drop in the recorded level of assault in the analysis period. There had been a significant increase in the recorded numbers of these incidents over the five years up to 2005.48 The increase in PCSOs and the development of Neighbourhood policing teams have had a positive effect on the reduction of violence levels, as has the deployment of Operation Shield, which has recovered 40 weapons during the review period, reducing levels of violence involving knives and other weapons on the rail network. • There has been no significant trend in sexual offences over the five-year period, although 2007 did see a 16% drop in the number of these offences reported to BTP.
10.4.2 Personal security incidents by type
Chart 109. Trends in assault, robbery and sex offences
14000 other sex offences Assault indecent exposure weapons offences 12000 violent assault (incl_ homicide) 10783 10720 racially aggravated harassment 10104 other offences 10000 9564 9337 common assault assaulting police robbery 8000
6000
Robbery Sexual Offences
Crimes recorded by BTP recorded Crimes 4000
2433 2363 2184 2077 2000 1223 1242 1151 1099 1151 963
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
48 As noted previously, it is likely that much of this increase was the result of changes in the way that police recorded violent crime following the introduction of the National Crime Recording Standard in 2002.
162 2007 Annual Safety Performance Report Personal security
10.4.3 Further analysis of passenger and public assaults 10.4.3.1 Location Chart 110 looks at where assaults occurred in 2007; it also illustrates the change from the assaults recorded in 2006.
Most assaults recorded by BTP take place inside stations. The breakdown of assaults by location in 2007 was little changed from the previous year; there was a small rise the proportion of assaults that took place in stations, and a corresponding reduction in the proportion that took place on trains.
Chart 110. Assaults recorded by BTP – where they occurred in 2007 (2006 in brackets)
3rd Party/Tenant Outside Station 4% (5) 9% (6) Other 5% (8)
On Train 25% (28)
Inside Station 53% (51)
Lineside 4% (2)
10.4.3.2 Temporal variation Chart 111 reveals there to be a clear pattern in the times of day when assaults take place.
The chart shows that the highest rate of assaults occurs on Friday evenings; this is closely followed by Saturday evenings. Alcohol is an important contributory factor at these times.
According to the British Crime Survey 2006/07, in nearly half (46%) of all violent incidents, victims believed offenders to be under the influence of alcohol; this figure rose to 58% in cases of attacks by people victims did not know personally. However, a Home Office survey found that less than 40% of alcohol-related assaults were reported to the police. Incidents that lead to violence on the railway can be sparked by ticket disputes, delays, crowding and the behaviour of other passengers. Weekend drinking is likely to exacerbate this problem.
On weekdays other than Friday, assaults peak between 18:00 and 19:00, and are at a fairly high level from between 15:00 and 16:00 (when school children make up a higher proportion of passengers) throughout the rush hour and into the late evening. There is an additional small peak during the morning rush hour.
2007 Annual Safety Performance Report 163 Personal security
Chart 111. Assaults by time of day and day of week
250 11pm-midnight
200 10pm-11pm
150
100
50 Assaults recorded by BTP in 2007 in BTP by recorded Assaults
0 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm 6am - 7am 6pm - 7pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midday - 1pm Midnight -1am Midnight -1am Midnight -1am Midnight -1am Midnight -1am Midnight -1am Midnight -1am Sunday Monday Tuesday Wednesday Thursday Friday Saturday
The BTP Violence Reduction Strategy states that BTP will aim to reduce the level of violent crime and hate crime on the railway network by implementing clear tactical options with partners aimed at tackling the causes of violence. BTP will further develop information- sharing practices with the industry, and work more closely with local Crime and Disorder Reduction Partnerships (CDRPs) to identify trends and seek to reduce violent crime.
BTP’s Neighbourhood Policing programme is one way of engaging with partners and the wider community to address problems and provide long-term solutions.
10.5 Public disorder The term ‘public disorder’ refers to behaviours such as drunkenness, harassment and affray. Unruly behaviour can contribute to an atmosphere that is threatening to both passengers and staff, and can also escalate into much more serious incidents.
RSSB groups BTP-recorded disorder offences into those considered to be the most serious (namely bomb threats, violent disorder and affray, and incidents of harassment, alarm and distress) and those that – while they can be threatening to other passengers or staff – are typically less serious in nature.
Chart 112 shows that 2007 saw a small rise in BTP-recorded public disorder offences. The total grew by just 4% following a large rise of 27% in 2006. This was brought about by a particular increase in harassment, alarm and distress offences.
164 2007 Annual Safety Performance Report Personal security
Chart 112. Trends in disorder on the railway
12000 violent disorder, affray passenger control 9441 9801 10000 hoax calls harassment and distress drink offences breach of peace 7448 8000
6320 5904 6000
4000 Crimes recorded by BTP by recorded Crimes
2000
0 2003 2004 2005 2006 2007
The increase from earlier years is likely to be due, in large part, to greater police activity throughout 2006 and 2007, with BTP having targeted anti-social behaviour in an attempt to raise awareness by increasing cautions and prosecutions.
The offence of harassment/alarm/distress is wide-ranging and appears to be increasingly well used by officers to counter low-level disorder. The continued impact of fixed penalty notices (FPNs) may have played a part in this increase, as officers become more familiar with their use.
According to BTP, the rate of increase in these offences has slowed down in 2007. There has been a real effort by practitioners in policing, security and revenue protection to continue to encourage staff to see anti-social behavior as a precursor to workplace violence, and to manage conflict before it escalates. This, plus the RSSB SWeRvE DVD, aims to raise awareness of the dangers of anti-social behaviour escalation, and manage out attitudes and behaviours before police involvement is required.
Some TOCs have also deployed the ‘Mosquito’ device to hotspots and have reported real benefits for staff and passengers. The correlation between anti-social behaviour and deployment of the device will need to be mapped to gain a better understanding of its success or otherwise.
Alcohol is seen to be a key contributory factor in crime and violent behaviour. While only 14.6% of crimes in the anti-social behaviour category were flagged as alcohol-related in BTP records, it is believed that the poor use of the flagging system continues to mask the true influence of alcohol.
2007 Annual Safety Performance Report 165 Personal security
10.6 Perceived security People’s perception, or fear, of violent crime can deter them from becoming passengers, even when the actual risk from such offences is low. Research suggests that introducing measures that are seen to enhance security on public transport could increase patronage by around 11%.49
Passenger Focus, the independent national rail consumer watchdog, carries out the National Passenger Survey (NPS) twice a year (autumn and spring) to provide a network-wide picture of passengers’ satisfaction with rail travel. Passengers at a representative sample of 650 stations across the country (at different times of the day) are asked to complete and return a questionnaire about a host of topics connected with rail travel – one of which is personal security. National results are based on approximately 25,000 correctly completed questionnaires.
The most recent survey, conducted in the autumn of 2007, suggest that the overall public view of security, in both stations and on trains, has improved over the five-year period between 2003 and 2007.
Chart 113 depicts this, and also shows that:
• The majority of passengers rate their personal security in the station and on the train as good (about two-thirds) or neither good nor poor (just under a third), with respondents that believe the security to be poor at these locations currently around 6%, on average. • Passengers seem to be more confident about their personal security on trains than in stations.
Chart 113. Trends in perceived security
In Stations On trains
80% 80% Good 70% Neither 70% Poor 60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%
3 4 6 4 5 6 7 05 06 0 0 0 005 007 003 0 0 007 200 20 2 20 200 2 2 200 2 2 20 2 g 2004 g g g n in ri mn mn ring 2006mn rin tumn r p tu p tu S u Spring 2005 S u Sprin AutumnSpringAutumn 200 Sp 2004Au Sp AutumnSpringAutumn 2007 Autumn A Autu A Autumn
Source: Passenger Focus (National Passenger Survey).
49 People’s perceptions of personal security and their concerns about crime on public transport. Research findings. Prepared by Crime Concern for the Department for Transport, 2004.
166 2007 Annual Safety Performance Report Personal security
The NPS differentiates between three different types of train operator: regional, long distance, and London & South East. However, this categorisation is becoming less clear-cut, as Passenger Focus itself notes:
‘With changes to the specification of the boundaries of train operating companies, “sector” definitions are becoming less straightforward and less meaningful as train operating company boundaries increasingly do not relate to the traditional sectors. Passenger Focus and the Office of Rail Regulation are currently reviewing the ‘sector’ reporting and following consultation with stakeholders may make some changes to these in future.’
Table 22 shows how the latest perceptions of overall safety compare for the different NPS- operator groupings. It indicates that:
• Passengers using long-distance services have the most confidence about their personal security in stations and on trains. This may be because these services are relatively busy and operate from large stations. • Travellers in London and the South East show the least satisfaction in perceived personal security, both in stations and on trains.
Table 22. Passenger perceptions of personal security: Autumn 2007 NPS In the station On the train Good Neither Poor Good Neither Poor Long distance 71% 26% 3% 83% 15% 2% London and South East 60% 31% 8% 68% 26% 6% Regional 65% 27% 9% 75% 21% 4% National Total 62% 30% 8% 70% 24% 6% Source: Passenger Focus (National Passenger Survey).
The survey also asked respondents to provide details of the nature of their personal security concerns. The top five for ‘in stations’ and ‘on trains’ are listed in Table 23.
Table 23. Reasons for passenger concerns about personal security
On trains In Stations
1. Anti-social behaviour by others (76%). 1. Anti-social behaviour by others (66%).
2. Lack of on-train staff (50%). 2. Lack of station staff (48%).
3. Lack of other passengers (19%). 3. Lack of other passengers (21%).
4. Witnessing acts of vandalism or violence. 4. Poor station lighting.
5. Lack of information. 5. Lack of information.
Source: Passenger Focus (National Passenger Survey).
The table shows that:
• Anti-social behaviour witnessed in stations and on trains is the prime source of personal security concern for passengers interviewed for the survey. In fact, more than two-thirds of passengers who expressed concerns about personal security listed this as a concern.
2007 Annual Safety Performance Report 167 Personal security
• Around half of the passengers who had concerns about personal security pointed out that a lack of staff in stations and on trains caused concern. • One-fifth said that a lack of other passengers was a personal security concern. Safety fears are influenced by an array of factors, not all of which reflect the actual probability of falling victim to violent crime. Persons committing minor offences that pose no threat (such as begging) can also contribute to an overall sense of insecurity – as could the presence of graffiti or litter. Other issues that have the potential to contribute to the overall sense of ill in stations and on trains are:
• Poor lighting. • A lack of information. • The absence of staff, police and other passengers. • The menacing presence of people loitering on or about stations, etc.
Public opinion about personal security on the railway is also shaped by the way in which improved recording levels are sometimes reported by the media as increases in crime levels, together with media focus on violent offences that have occurred on or near railway property.
Chart 114 shows survey respondents’ answers to the question of what issues should BTP address with priority. People were asked to select six areas from a list of 12. The chart shows that around 80% of respondents want BTP to give assaults on passengers and incidents of theft with violence the highest priority. Assaults on rail staff were also rated as a priority, coming higher than structural vandalism, anti-social behaviour and terrorism.
Chart 114. Areas identified as needing high prioritisation by BTP
Assaults on passengers
Theft with violence
Assaults on rail staff
'Structural' vandalism
Anti-social behaviour
Terrorism
Theft without violence
Hate crime
Managing fatalities
Football hooliganism
Graffiti
Areas named by respondents as priority for BTP for priority as respondents by named Areas Theft/vandalism of vehicles
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Proportion of respondents selecting area
Source: Passenger Focus (National Passenger Survey, autumn 2007).
168 2007 Annual Safety Performance Report Station safety
11 Station safety
This chapter looks at all stations on the mainline railway. It focuses mostly on the risk to passengers, but also covers workforce and members of the public (including trespassers).
2007 Headlines • There were 33 accidental deaths at stations in 2007, the highest for eight years. Twenty- four were trespassing.
• Of the nine non-trespass fatalities, six were the result of people standing too close to or falling from the platform edge (three passengers and three members of the public). Alcohol played a part in four of the cases. • There was a significant reduction in the number of major injuries at stations in 2006 and this was maintained in 2007. • Slips, trips and falls remain the dominant cause of injuries at stations. * Performance at a glance
Accidental fatalities at stations (excl. trespass) Accidental major injuries at stations (excl. trespass) 12 300 11 283 276 280 10 10 9 9 250 222 225 8 7 200
6 150
4 100
2 50
0 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007
2007 performanceSignificant change? 2007 performance Significant change? 28% higher than 20062 1% higher than 2006 2 3% lower than 2003-06 average2 16% lower than 2003-06 average 33
* A difference in the performance measure that is statistically significant at the 5% significance level is indicated by 99, a significant difference at the 20% significance level by 9, and 8 denotes a difference that is not significant at the 20% significance level. A difference is significant at the 5% (or 20%) level if the probability of it occurring by chance alone is less than or equal to 5% (or 20%).
2007 Annual Safety Performance Report 169 Station safety
11.1 Station risk profile Chart 115 presents the breakdown of harm at stations, by person type and accident type. The profile is based on actual injury data from the past five years rather than SRM data, as the SRM does not lend itself easily to producing analysis based on accident location.
For passengers, an average of nearly three fatalities per year have occurred as a result of platform-edge incidents, with a further two from slips, trips and falls and just over a half from boarding and alighting incidents. An average of one passenger a year has been fatally assaulted. There has been just under one passenger fatality a year on crossings at stations. When non-fatal injuries are also considered, more than 60% of passenger harm in stations is due to slips, trips and falls.
Public station fatalities – which occur mostly to trespassers who use the station as a means of accessing the tracks – are dominated by being struck/crushed by trains and electrocution.
A smaller proportion (13%) of risk at stations has occurred to the workforce and, over the past five years, there have been no workforce fatalities in stations.
Chart 115. Station harm by person and accident type (2003–2007)
Other 0.5 Electric Shock 3.7 Public Struck/crushed by train 11.9 (trespass) Other 0.1 Struck/crushed by train on level crossing 0.2 Assault 0.5 Platform edge incidents 0.6 Public (non- Public tresspasser) Slips, trips and falls (includes falls from height) 0.7 Fatal Other 0.4 Major Manual Handling 0.4 Contact with objects 1.2 Minor Boarding and alighting 1.4 Shock/trauma Slips, trips and falls (includes falls from height) 2.2 Assault 2.4 Type of station accident station of Type Other 0.1 Struck/crushed by train on level crossing 0.8 Contact with objects 1.1 Assault 1.7 Platform edge incidents 4.5 Passenger Workforce Boarding and alighting 6.2 Slips, trips and falls (includes falls from height) 22.3
0 2 4 6 8 10 12 14 16 18 20 22 24 Average FWI per year
Source: SMIS.
11.2 Fatalities and injuries in stations during 2007 Excluding cases of trespass, suicide, suspected suicide, and death by natural causes, there were nine fatalities at stations during 2007. Six were passengers and three were members of the public.
Of the passenger fatalities, one resulted from assault, one involved being struck by a train at a station crossing, three were a result of passengers standing too close to the platform edge
170 2007 Annual Safety Performance Report Station safety
and one involved a passenger falling between the train and the platform. Further details may be found in Chapter 5, Passenger safety.
All public fatalities were caused by platform-edge incidents, two of the accidents involved intoxication and one involved a youth who had been fighting on the platform. Further details may be found in section 11.6.
Excluding cases of trespass, there were 225 major injuries at stations in 2007; 80% of these affected passengers, 17% affected workforce and 3% affected members of the public.
11.3 Station safety trends Accidental harm at stations in 2007 was split by person type in the following proportions:
• 47% affected passengers. • 11% affected workforce. • 36% affected trespassers. • 6% affected other members of the public. Chart 116 shows the total accidental harm occurring at stations in each of the last five years, split between trespass and non-trespass.
Chart 116. Fatalities and weighted injuries in stations (excluding suicide)
50.6 49.3 50.0 Shock/trauma 50 0.7 0.9 0.7 Minor 43.7 Major 11.3 11.2 11.2 40.7 0.7 Fatal 40 0.6 11.6 10.9
30
0.0 24.9 27.6 0.0 0.9 28.3 28.0 20 22.5 22.2 0.0 0.0 14.5 0.0 14.3 0.0 13.0 13.8 0.0 0.5 0.1 0.8 0.9 Fatalities and weighted injuries weighted and Fatalities 0.2 24 10 0.0 0.0 14 14 13 11 12 9 10 9 7 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Non - trespass Trespass
The overall level of harm for non-trespassers remained fairly constant until 2005, with any variations being largely explained by a small number of fatalities. However, 2006 saw a notable decrease, due to a reduction in major injuries. Though the levels of harm have risen slightly in 2007 (mostly as a result of an additional two fatalities), the lower rate of major injuries has been sustained.
2007 Annual Safety Performance Report 171 Station safety
The overall harm for trespassers saw a large increase in 2007, compared with the previous four years. This is due to a significant increase in fatalities, which comprise the overwhelming proportion of trespass harm at stations.
11.3.1 Trends in fatalities at stations Chart 117 presents the rate of accidental fatalities at stations per billion passenger journeys, split into passengers, members of the public and trespassers. The normaliser is not ideal for public and trespass fatality trends, but public usage (and even to some extent, trespass) is likely to reflect the number of people using stations in connection with train journeys.
Chart 117. Long-term trends in the underlying fatality rate at stations (3-yr moving average)
25 Passenger Public Trespass 20
15
10
5
Number of fatalities per billion passenger journeys passenger billion per fatalities of Number 0 Jul Jul Jul Jul Jul Jul Jul Jul Jul Jul Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Mar Mar Mar Mar Mar Mar Mar Mar Mar Mar Nov Nov Nov Nov Nov Nov Nov Nov Nov Nov May May May May May May May May May May Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
The following points can be concluded from the data:
There is a long-term decreasing trend in the rate of passenger fatalities at stations and the rate in 2007 is the lowest seen in the 10 years since 1998. Conversely, over the past two years, there has been an upward trend in trespass fatalities at stations after nearly a decade of improvement. For members of the public, excluding trespassers, the fatality rate has remained relatively constant over the 10-year period.
The average fatality rates at the end of 2007 are as follows:
• One passenger fatality in 158 million passenger journeys. • One trespasser fatality per 65 million passenger journeys.
172 2007 Annual Safety Performance Report Station safety
Chart 118. Fatalities in stations by person type (excluding suicides)
40 39 Trespass Public Workforce 35 33 Passenger 30 29 20 28
24 24 24 25 23 23
20 24 20 18 16 1 1 12 14 16 14
Fatalities per year per Fatalities 15 19 13 1 3 10 3 17 2 3 3 5 10 1 9988 7 6 5 4 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Chart 118 shows the actual number of accidental fatalities per year over the 10-year period from 1998. Note that:
• The downward trend in passenger fatalities seen over the past six years continued in 2007. • The number of station trespass fatalities in 2007 reached the highest level over the 10- year period. This has brought the total number of fatalities at stations to the highest level since 1999. 11.3.2 Trends in major injuries at stations Chart 119 presents the number of accidental major injuries at stations per year, split by person type. The overwhelming majority of these injuries affect passengers, with a smaller, proportion occurring to the workforce and a handful related to the public and trespassers.
The total number of accidental major injuries occurring at stations in 2007 was approximately the same as the previous year and is significantly less than the average of the previous four years.
2007 Annual Safety Performance Report 173 Station safety
Chart 119. Major injuries in stations by person type
Trespass
300 288 Public 285 282 5 2 Workforce 2 9 6 3 41 Passenger 250 48 48 230 234 8 9 6 7 200 28 39
150
240 222 229
Major injuries per Major year 100 188 179
50
0 2003 2004 2005 2006 2007
Key issues that describe the trends in major injuries are as follows:
• There was a significant decrease in the number of passenger major injuries in 2006; 2007 saw a further small reduction. Passenger slips, trips or falls on stairs, predominantly resulting in fractures, was the biggest single contributor to the major injuries, followed by accidents when alighting trains. • The number of workforce major injuries at stations has marginally increased compared with last year, but levels are relatively consistent with the average for the previous four years. • For members of the public, most of the major injuries were due to falls on or around the station concourse. • Only a small proportion of the major injuries at stations affect trespassers. However, this group accounts for most fatalities in stations, as shown in Chart 118.
11.4 Passenger safety in stations There were six passenger fatalities at stations in 2007.
One of the fatalities involved a young male being struck by a train on the station crossing at Wokingham. Over the previous five years, two children died on the station crossing at Elsenham in 2005, and one man on a crossing at Gomshall station in 2004. Safety performance at level crossings is analysed further in Chapter 9, Risk at level crossings.
Another passenger fatality was the result of an assault. Details are discussed further in Chapter 5, Passenger safety.
The other four passenger fatalities involved an alighting accident and three platform-edge incidents.
174 2007 Annual Safety Performance Report Station safety
11.4.1 Boarding and alighting
Chart 120. Passenger harm from boarding and alighting accidents
9 Shock/trauma 8.3 8 Minor Major 7
Fatal 6.5 6.4 6.1 6
5
4 3.7 3.5
3 2.9 2.5 2.4 2.4 2.1 2.1
2 1.7 Fatalities and weighted injuries and weighted Fatalities 1.4 1.4 1.2 1.1 1.0 0.9 0.8 0.8 0.8
1 0.7 0.5 0.5
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Fall between train Caught in train Other boarding Other alighting Total and platform doors
Chart 120 presents the passenger harm from boarding and alighting incidents at stations. These made up 14% of the total harm at stations in 2007 (excluding trespass and suicides). The number of fatalities and weighted injuries in each category and in total (when all boarding and alighting accidents are considered together), was higher in 2007 than the previous year, though there does not seem to be a specific reason for the increase.
The one fatality in 2007 was an alighting incident on 13 February at Haddenhan & Thame Parkway, where a passenger fell between the train and the platform. In general, the types of major injuries seen were predominantly fractured wrists and ankles. Most harm occurred from alighting trains, predominantly as a result of passengers losing their balance or missing their footing. A high proportion of major injuries involved elderly people, or those with impaired mobility – mainly due to carrying luggage or travelling with young children.
There are a number of major injuries in this category that suggest that crowding at stations can be an issue. These include injuries attributed to passengers being caught by obstructions such as luggage and their fellow travellers. Another example in 2007 occurred when a passenger was caught by the train doors and fractured her wrist whilst attempting to enter a crowded train.
Chapter 3 sets out some of the good practice that is being adopted throughout the industry to further maintain and improve passenger safety when boarding and alighting trains.
2007 Annual Safety Performance Report 175 Station safety
11.4.2 Other accidents at the platform edge
Chart 121. Passenger harm due to falls from and standing too close to the platform edge
9 8.3 Shock/trauma 8 7.6 Minor Major 7 Fatal
6
5 4.6 4.4 4.1 4 3.2 2.9 3 2.4 2.2 2.2 2.0
Fatalities and Weighted injuries and Weighted Fatalities 2
1 0.6 0.2 0.3 0.2 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Fall from platform edge Standing too close to platform edge Total
Passenger harm due to falls from and standing too close to the platform edge made up just over 14% of all passenger FWI in stations in 2007. Chart 121 presents the passenger harm from these accidents in the five years from 2003. Despite these types of accidents being relatively low-frequency, they tend to have very severe consequences. However, because the figures are dominated by a relatively small number of fatalities, it is not possible to identify trends with any degree of confidence.
In 2007, there were three fatalities involving passengers being struck by oncoming trains. Two of these were attributed to standing too close to the platform edge – the first fatalities in this category since 2002:50
• On 25 January, a 15-year-old child was fatally injured at Treorchy station (Western) as he leaned over the platform edge whilst wearing headphones. • On 7 November, an intoxicated passenger was struck by a train whilst standing too close to the platform edge at Strathclyde (Scotland). • On 21 November, an intoxicated passenger fell from the platform at Glengarnock (Scotland). It is evident from the fatalities in 2007 that intoxication can play an important part in these types of accidents. On average, 62% of the total passenger harm due to falls or standing too close to the platform edge over the five-year period resulted from intoxication. More detailed analysis on the effects of intoxication on passengers is included in section 5.5.1.
50 On 18 May 2002, a man was struck by a train whilst sitting at the platform edge with his legs reportedly dangling over the track. He later died of his injuries.
176 2007 Annual Safety Performance Report Station safety
11.4.3 Moving around the station Most accidents in stations – even those that result in major injuries – are of the type that could occur in any public building. Chart 122 presents the number of major injuries incurred whilst moving around stations, broken down by cause.
Chart 122. Passenger major injuries while moving around the station
Other 180 177 1 168 10 Contact with object 3 162 160 5 151 Slips, trips and falls 5 140 126 120 6
100
165 166 80 157 146
60 120 Number ofmajor injuries
40
20
0 2003 2004 2005 2006 2007
The overwhelming majority of passenger major injuries whilst moving around stations are as a result of slips, trips and falls. These injuries are analysed further in section 11.4.3.1. There is also a small contribution to the total passenger harm that comes from contact with objects. During 2007, this included one injury where a passenger was struck by automatic ticket barriers, causing fractured ribs, and three injuries whereby passengers have fallen after colliding with other people.
The number of major injuries resulting from slips, trips and falls fell for the second consecutive year in 2007 and is significantly lower than the average over the previous four years. The reduction in passenger harm comes despite the changing passenger demographics and an evolving, more reliable railway, both of which are continuing to bring in higher volumes of both leisure and commuting passengers into our stations.51
The industry is putting much effort and resource into ensuring that stations are designed to meet passenger needs, for example as part of the Network Rail-led National Station Improvement Programme.
51 Approximately 22¼% passenger growth is predicted by 2014 in accordance with HLOS (White Papers), DfT, July 2007.
2007 Annual Safety Performance Report 177 Station safety
11.4.3.1 Slips, trips and falls
Chart 123. Passenger harm from slips, trips and falls
Shock/trauma
25 24.7 Minor 23.8 22.8 22.2 Major 20 Fatal 17.9
15 10.3 9.9 9.8
10 9.2 7.4 7.1 6.6 6.1 5.9 Fatalities and weighted injuries weighted and Fatalities 5 4.9 3.6 3.5 3.4 3.3 3.3 2.7 2.5 2.4 2.4 2.3 2.1 0.8 0.8 0.8 0.6 0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Stairs Escalators Platform Concourse Other Total
Events listed in ‘Other’ include passenger slips, trips and falls on ramps and from chairs, and falls from station over-bridges.
Passenger slips, trips and falls amount to 55% of the total passenger harm at stations in 2007 (excluding trespass and suicides). Chart 123 presents the passenger harm from slips, trips and falls for different locations around the station.
Last year (2007) saw less harm from passenger slips, trips and falls than the previous four years. This is partly because 2007 was the first year since 2000 where there were no fatalities, but major injuries also fell (see Chart 122).
In the five years analysed, most passenger harm due to slips, trips and falls in stations occurred on stairs. Indeed, passengers are more likely to be fatally injured by slips, trips and falls on stairs than any other location. The industry has committed to the redevelopment of stations to provide step-free access from the station concourse to the platform. The ‘Access for All’ programmes are discussed in Chapter 2, Safety overview.
Many slips, trips and falls have no particular cause assigned to them and are often described in SMIS as ‘missed footing’. When a specific reason is given, however, running, wet or uneven surfaces and obstructions are the most likely precursors to slips, trips and falls at stations. Alongside these are other contributory factors such as crowding and intoxication. There are a number of current initiatives that aim to reduce the harm from slips, trips and falls at stations – notably, resurface work on platforms and restructuring stairs – which are discussed further in section 3.2.
11.4.4 Other hazards in stations Other hazards in stations that may affect passenger safety include assaults and crossing the tracks on station foot crossings.
178 2007 Annual Safety Performance Report Station safety
In general, stations are the prime locations for assaults, with 53% of all assaults recorded by BTP occurring inside their confines.52 Assaults are often the result of ticket disputes, delays, crowding and the behaviour of other passengers. Intoxicated passengers also provide an important contribution and this is evident from the fact that the most incidents of assault occur on Friday nights, closely followed by Saturday nights.
Since 2004, there have been four passenger fatalities at station crossings and one major injury related to crossing the line at stations. The fatalities included one on 24 January 2007, when a passenger was struck and fatally injured at Wokingham manually controlled barrier (MCB) level crossing (South East). The young male climbed over the barriers to join a train in the Up platform and was struck by a Down service.
11.4.4.1 Catastrophic risk Catastrophic risk covers those events that have the potential to cause multiple fatalities.
Crowding The risk from crowding amounts to around 0.1 FWI per year (according to SRMv5.5); crowding can also increase the risk from other types of accident, such as slips, trips and falls, or even falls from the platform.
If crowds are not properly managed, serious overcrowding has the potential to result in large- scale loss of life. The catastrophic effect of crowding was seen in Minsk station (Belarus) in 1999, when 52 people were fatally injured and 78 people were hospitalised as a crowd of 2,500 rushed to get out of the heavy rain during the city’s beer festival. A more recent example occurred in October 2007, when 14 women were crushed to death whilst changing trains on their way to a Hindu festival in northern India.
In Britain, TOCs adopt strategies to handle large volumes of passengers at stations. RSSB has carried out research into these strategies (particularly related to the management of football fans on the railways)53 and provides suggestions on how to minimise the risk from crowding.
Station fires Fatal fires are relatively uncommon on Britain’s railways. According to the SRMv5.5, the risk from fires at stations is around 0.2 FWI per year.
However, historically, some station fires have resulted in large-scale loss of life, although the worst incidents have tended to occur at underground stations, which can be difficult to evacuate. For example, 31 people died in the fire at King’s Cross underground station in 1987, after a match was dropped onto a wooden escalator. The fire led to smoking being banned on London Underground and the gradual replacement of all wooden escalators.
A more recent international example occurred in February 2003 at Daegu subway station (South Korea), where 192 people were killed and 148 people were injured when a train was set on fire by a mentally ill person. The entire station was filled with poisonous smoke from
52 See Chapter 10, Personal Security, for further analysis. 53 Project T592: Management of Football Fans on the Railways’, September 2006, Virago Consulting Ltd (for RSSB). Available at http://www.rssb.co.uk/research/index.asp
2007 Annual Safety Performance Report 179 Station safety
the flammable material on board the train. In addition, the station’s fire detection systems were not working properly and the officials on duty ignored the fire alarm and proceeded to route another train into the station.
Structural collapse Though the risk from structural collapse at stations is small (0.04 FWI per year), there is a continuing threat from terrorism.54 The UK rail industry has actively encouraged workforce and passengers to be vigilant in order to keep the risk to a minimum.
Structural collapse can also be caused by poor design or materials, or unusual weather. An example of structural collapse from the airline industry occurred in May 2004, when a portion of the ceiling collapsed at Charles de Gaulle Airport in Paris, killing four people. Though the construction was new, the investigation concluded that the design, in general, had little margin for safety.
Train accidents Train accidents contribute a small, but not insignificant, amount of passenger harm at stations. The nature of these accidents can be split into four main categories: derailments, train fires, train collisions and buffer stop collisions.
An example occurred in May 2002, when a train derailed at a set of points on the approach to Potters Bar station. Seven people were killed (six on the train, and one person who was walking under a railway bridge) as the rear carriage of the train, which was running at 100mph, flipped across two platforms and became wedged beneath the canopy of the station.
Buffer stop collisions also have the potential to injure passengers in the station, as well as those on the train. For example, in 1971, a passenger standing in the ticket office at Sheerness station was killed as the result of a train striking the buffer stop.
For further information on train accidents in stations, see Chapter 8, Risk from train accidents.
11.5 Workforce safety in stations In 2007, 17% of all reported station injuries occurred to members of the workforce. Predominantly, train crew and station staff (including cleaners) are affected by hazards at stations. The relatively high proportion of harm is despite no workforce fatalities at stations since 1999, when a cleaning contractor died after falling and banging his head due to ill health at Manchester Victoria Station.55
54 The SRM figure does not include the risk from terrorism. 55 In addition to this, on 5 July 2002, a member of station staff was fatally injured after tackling a fire on a train at Purley Station.
180 2007 Annual Safety Performance Report Station safety
Chart 124. Workforce major injuries in stations by accident type
3.0 Shock/trauma 3 2.6
2.6 Minor 2.5 2.5
2.3 Major 2.2 2.1 2.1 2 1.9 1.8 1.7 1.7 1.6
1.5 1.5 1.4 1.3 1.3 1.3
1 1.0 0.7 0.6 0.5 Fatalities and weighted injuries weighted and Fatalities 0.5 0.5 0.3 0.3 0.3 0.3 0.2 0.2 0.2
0 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 2003 2004 2005 2006 2007 Boarding / Slips, trips and Assault Contact with Awkward Other alighting falls object movement / manual handling
Chart 124 gives a breakdown of all workforce injuries at stations by injury cause. The total sum of weighted injuries in 2007 has increased from 2006, but is still less than seen in 2003– 2005.
The largest contribution of workforce harm at stations in 2007 came from slips, trips and falls and assaults. Slips, trips and falls mainly occur on the platform area, with relatively few taking place on stairs (unlike the situation with passengers). One in five of all workforce assaults affected station staff. However, the reduced harm from assaults seen in 2006 has continued into 2007. More detailed analysis on assaults can be found in Chapter 10, Personal Security.
There has been an increase in workforce injuries from boarding and alighting since 2006. This has been caused by an increase in major injuries, which are back to the levels seen between 2003 and 2005. Most of the major injuries in this category occur as a result of train crew or platform staff missing their step and falling between the train and platform.
A more detailed analysis of the risk of harm to workforce is available in Chapter 6, Workforce safety.
11.6 Public safety in stations Though people have always met and seen off friends and loved ones, the increase in retail outlets at stations has brought an increase in non-travelling members of the public. Now people enter the station’s confines to shop, dine or simply browse through books and magazines.56 This gives them a risk profile similar to passengers in the station environment.
56 Injuries that occur within station retail outlets are outside the scope of this report.
2007 Annual Safety Performance Report 181 Station safety
Members of the public accounted for 9% of the total harm at stations in 2007 (excluding trespass and suicide). This year has seen the first three public fatalities as a result of being too close to the platform edge since 2002.57 These are summarised below:
• On 11 January, a member of public fell between the train and platform whilst banging on the train windows as the train was pulling out of Gidea Park (South East). The person was reported as being under the influence of alcohol at the time of the incident. • On 23 February, a member of public was struck by a non-stopper at Seven Kings, having been too close to the platform edge (South East). • On 22 August, a member of public fell onto the track into the path of an oncoming train at Shoreham (South East). The driver reported that a group of youths had been fighting on the platform when two of them fell to the track. The other youth was electrocuted by the third rail, but survived with major injuries. In 2006, no members of the public were fatally injured at stations (excluding suicides), but in 2005, two children were killed along with their mother, who jumped in front of a train to commit suicide, and in 2004, two members of the public died as a result of falls onto the railway and one was fatally injured on a station crossing. Chart 118 presents the fatality rate over the 10-year period since 1998.
In the five years from 2003, the overwhelming majority of public weighted injuries at stations have been attributed to slips, trips and falls; the levels of harm appear to be increasing. This may be indicative of the rise in members of the public within the station environment, or increased vigilance in the classification of person types when reporting incidents into SMIS.
Assaults have also contributed to public harm, but to a lesser extent. Examples from 2007 include a stabbing between two females at a major station and a parent falling while attempting to stop his daughter from being assaulted by three drunken men.
Over recent years, about one-third of suicides have occurred at stations. In 2007, there were 75, with a further two at station crossings. More detailed analysis of suicide trends is available in Chapter 7, Public safety.
57 For each accident, the person involved did not appear to have any intention to travel; thus they are regarded as members of the public, despite being on the platform area.
182 2007 Annual Safety Performance Report Learning from accidents
12 Learning from accidents
This chapter provides details of RAIB investigations carried out during 2007, progress made against recommendations from both new and old investigations and some of the lessons learned.
Specifically, it looks at:
• Industry processes. • Investigations published. • Progress against recommendations. • Confidential Incident Reporting and Analysis System. • Industry strategy. • European developments in learning from accidents. 2007 Headlines • A total of 158 recommendations were issued to railway group members (RGMs) from 22 RAIB investigations of incidents occurring on NRMI. • In January, the report of the BP US Refineries independent safety review panel was published; this identified lessons that are transferable to almost any industry. • RSSB continues to track industry progress on outstanding recommendations from past formal inquiries; more than half the recommendations open at the start of 2007 have now been closed.
12.1 Industry processes for learning from accidents The Rail Accident Investigation Branch (RAIB) was established in 2005 to investigate train accidents and incidents on UK railway infrastructure without apportioning blame or liability. It is independent of the rail industry and the ORR, with the Chief Inspector of Rail Accidents reporting directly to the Secretary of State for Transport. RAIB’s recommendations are addressed to the ORR, which is then required to ensure that they are duly considered by duty holders and that, where appropriate, action is taken.
In addition to the investigations carried out by RAIB, duty holders have their own arrangements for carrying out internal investigations. This includes a capacity to undertake independently chaired investigations when appropriate. The output from internal investigations is managed by the duty holders concerned with actions being picked up by their own tracking systems. The results of duty holder-led formal investigations are also summarised in SMIS for others to learn from.
Whoever undertakes an investigation, the industry seeks to learn the resulting safety lessons, while duty holders consider and respond to recommendations appropriately. RSSB tracks the progress of investigation recommendations made by RAIB, producing reports which allow the ORR to report the progress against said recommendations back to RAIB.
The tracking is carried out by regular correspondence with RGMs, in which details of progress made against allocated recommendations are sought. The SMIS recommendation
2007 Annual Safety Performance Report 183 Learning from accidents
tracking tool is also used for this purpose. The SMIS recommendation tracking tool is visible to the ORR and used to assess the actions taken in response to recommendations before formal closure. It is therefore important that this tool is used, so that a complete and up-to- date picture of industry progress is available to all the parties concerned.
RAIB, as the national investigation body, submits an annual report on its activities to the European Rail Agency.
12.2 Investigations completed during 2007 Last year (2007), RAIB published 47 reports from its independent accident investigation activities. These investigations covered the following rail categories:
• Heavy rail (27), including two in Northern Ireland and three non-NRMI. • Heritage railways (10). • Light Rail (6). • Metro (4). A total of 158 recommendations were issued to RGMs as a result of 22 RAIB investigations in 2007 for incidents on NRMI. This is compared to 100 recommendations from 15 investigation reports in 2006.
12.2.1 Lessons learnt during the year It is not the intention of this chapter to give an account of RAIB output, as it publishes its own annual report that, together with all its published investigation reports, is available on the RAIB website (www.raib.gov.uk). However, the following gives an overview of a few of the events that occurred on the UK heavy rail network which are of particular interest in terms of the lessons learnt.
Derailment of a freight train at Brentingby Junction, near Melton Mowbray on 9 February 2006 At 05:31 on 9 February 2006, 6Z41, the 05:17 Mountsorrel–Barham freight, operated by EWS, derailed at trap points at the end of the Up Goods Loop at Brentingby Junction, near Melton Mowbray. The derailment of the Class 66 locomotive and the first three wagons occurred after the train passed signal 53 at the end of the Up Goods Loop at danger. No one was injured as a result of the accident.
Ten recommendations were made, relating to the following areas:
• Implementing napping as part of a fatigue management system. • Providing the facilities for napping. • Research into sleeping before a first night shift. • Screening train drivers for sleep disorders as part of regular medical surveillance. • Providing improved guidance to drivers about how they can maximise alertness while at work. • Implementing a system that re-briefs the guidance at intervals.
184 2007 Annual Safety Performance Report Learning from accidents
• Research into implementing a system of personal responsibility statements and/or sleep contracts. • Implementing improvements to the procedure to check drivers’ fitness for duty concerning fatigue. • Improving the trap points beyond signal 53. • Repositioning the signal post telephone fitted to signal 53.
Lessons
The investigation highlighted the importance of management of fatigue on both a personal level as well as at company process level.
Possible causes of fatigue can be the shift pattern, workload, individual characteristics and social circumstances. A person who has not had enough sleep may feel sleepy if their workload is boring and undemanding. An individual’s diet, age, personality and fitness can all have a bearing on fatigue, and if their social circumstances are such that they cannot get sufficient sleep at home, then the onset of fatigue will be exacerbated.
Most company fatigue indexes (FIs) provide a means to assess the short-term, daily fatigue and cumulative fatigue risks associated with shift work, such as time of day, shift duration, rest periods, breaks within a shift, and cumulative fatigue. The FI can take account of day, night and early shifts and can be used to compare different shift patterns. It can also be used to identify peaks of fatigue within a shift pattern so that suitable control measures can be put in place.
However, these indexes can have some shortcomings if they take no account of individual factors such as age, and fitness (and so on). Significantly, an index may also take no account of an individual’s lifestyle outside work, including the amount of sleep obtained preceding a shift.
This is where individuals need to be aware of the issues, including those relating to lifestyle, so that they can take action to guard against the causes of fatigue that may then affect their ability to remain alert whilst working.
Near miss involving a track worker at Tinsley Green Junction on 17 March 2007 This incident occurred on the morning of Saturday 17 March 2007, at Tinsley Green Junction, near Gatwick Airport. The driver of 1M20, the 08:55 Brighton–Watford Junction service, reported to the signaller that a member of track maintenance staff had dived clear of his train with only seconds to spare. The incident had occurred as 1M20 was being routed via a series of high-speed crossovers.
The train struck some welding rods that had been left by the welder as he jumped clear of the approaching train. None of the staff concerned were injured, though all were shaken by the event. The train was not damaged.
RAIB led investigation made a number of recommendations relating to the following areas:
• Competency of welders to act as COSS for work in a Red Zone.
2007 Annual Safety Performance Report 185 Learning from accidents
• The rules and training related to working at locations beyond facing points. • Identification, recording and briefing of hazards when working at locations beyond facing points. • The efficacy of existing business processes for the planning of safe systems of work. • Checking of data provided on safe system of work forms. • Improving the presentation of information in operating documentation. Lessons
This incident highlights the issues around Red Zone working in multi-tracked areas near to points with various routing options approaching the site of work. Trains can travel through the area on adjacent lines within view, but without approaching the site of work.
Everyone in the planning and execution of the work has to be aware of the hazards arising from working in the vicinity of points and the directions trains can approach from. Also, everyone working on site has to be clear on the action to be taken on each occasion warning of an approaching train is given.
Another event occurred at Ruscombe Junction involving a welder working in a Red Zone near to points on 29 April 2007, unfortunately with fatal consequences for the track worker involved on this occasion. The RAIB investigation report for this accident was published in February 2008.
Passenger door open on a moving train near Desborough on 10 June 2006 On Saturday 10 June 2006, a passenger on 1D17, the 10:30 London St Pancras–Sheffield service, reported to on-board staff that an exterior power operated door was open, in the first class portion, while the train was in motion. The train was formed of a Class 222 ‘Meridian’ DMU (222 009).
The door opened just north of Kettering. The train was finally brought to a stand at Desborough summit, 5 miles 79 chains north of Kettering station.
The RAIB-led investigation made a number of recommendations which relate to the following areas:
• Review and modification of the algorithm used to control the door. • Procedures for specification, development and verification of train and door system software. • A review of the design of the door-locking switch and the associated manufacturing process. • Fault alarms from the train management system and their treatment by the on-board staff. • Improvements to the training of on-board staff. • The ergonomics of indication lamps in the driver’s cab and the passenger communication apparatus emergency brake handles.
186 2007 Annual Safety Performance Report Learning from accidents
• A review of the operational rules relating to passenger communication apparatus, power- operated doors and use of the emergency brake override.
Lessons
This incident highlights the issues surrounding the ever-increasing complexity of the systems used on trains to operate, manage and monitor all kinds of on-board equipment.
There are often many different bodies involved in the process, such as vehicle owners, operators, maintainers, manufacturers, designers and software engineers. It is imperative that changes/improvements to any part of the system be assessed with sufficient rigour to ensure that the effects on other parts of the system or processes are fully understood.
12.3 Industry progress against recommendations from investigations The content of this chapter is based on information reported either to RSSB directly or entered into SMIS and subsequently extracted. Where the data is used to provide percentages for comparative purposes, it is reliant on the accuracy of information both recorded into, and retrieved from, SMIS.
This chapter indicates the status of recommendations at the time of data extraction, mainly in January 2008, to reflect the status at the end of 2007. However, RSSB has an ongoing dialogue with some RGMs over some responses and progress of actions reported here. It is a dynamic process, in which information can be updated on a daily basis. RSSB carries out reports on the status of industry recommendations twice a year (in June and December). When the evidence is provided by an authorised manager or technical expert, the status of a recommendation can be moved towards a formal closure. In the case of RAIB recommendations, closure in SMIS can only be made once endorsed by the ORR.
Recommendations may be rejected with justification by a duty holder.
Table 24 details the RAIB investigations published during 2006 and 2007 with the resultant recommendations. It illustrates the industry progress in closing out those recommendations which have been accepted by the ORR.
2007 Annual Safety Performance Report 187 Learning from accidents
Table 24. Summary of industry progress against RAIB recommendations in 2007
Incident Recs RAIB Investigation Total Recs in Recs closed Total Recs date Issued Recs progress in 2007 closed 18/10/2005 14/07/2006 Freight train derailment at Hatherley, near Cheltenham Spa 5 1 2 4 19/10/2005 21/07/2006 Collision at Black Horse Drove Crossing, near Littleport, Cambridgeshire 4 2 1 2 26/10/2005 11/08/2006 Derailment near Liverpool Central 8 7 0 1 26/10/2005 25/08/2006 Track worker fatality at Trafford Park 9 3 5 6 28/10/2005 28/03/2006 Derailment at Watford Junction Yard 4 1 0 3 02/11/2005 02/11/2006 Runaway manually-propelled trolley between Larkhall and Barncluith Tunnel 16 9 4 7 04/11/2005 02/11/2006 Runaway manually-propelled trolley between Larkhall and Barncluith Tunnel 6 2 3 4 26/11/2005 29/11/2006 Derailment of a train at Moy 10 8 1 2 03/12/2005 11/12/2006 Investigation in station pedestrian crossings initiated by a fatality at Elsenham Station 10 10 0 0 05/01/2006 20/12/2006 Broken rails at Urchfont and Kennington after transit of freight train 61 5 5 11/01/2006 18/08/2006 Cutting of rail from a line that was still open to traffic, near Thirsk station 8 6 1 2 12/01/2006 20/07/2006 Station over-run at Haywards Heath 2 0 1 2 18/01/2006 14/11/2006 Derailment at York 4 4 0 0 27/01/2006 20/12/2006 Collision between train and buffer stops at Sudbury 2 1 1 1 06/02/2006 19/09/2006 Derailment of a Ballast Plough Brake Van at Carlisle 6 4 2 2 Recs issued 2006 Total 100 59 26 41 25/11/2005 08/01/2007 Autumn Adhesion incidents 2005 including Esher and Lewes. Reports 1, 2 and 3 25 19 6 6 14/01/2006 30/01/2007 Derailment incident at Edinburgh Haymarket 3 3 0 0 21/01/2006 30/01/2007 Derailment incident at Waterside, East Ayreshire 7 3 4 4 31/01/2006 23/01/2007 Derailment incident at Cricklewood 6 6 0 0 09/02/2006 23/01/2007 Derailment at Brentingby Junction, near Melton Mowbray 10 10 0 0 15/02/2006 30/04/2007 Train door incident at Huntingdon 6 6 0 0 21/02/2006 21/02/2007 Unsecure load incident at Basford Hall 5 0 5 5 19/03/2006 25/07/2007 Near miss involving track workers at Manor Park 3 3 0 0 07/04/2006 30/04/2007 SPAD incident at Camden Road 9 0 9 9 01/05/2006 29/05/2007 Near miss incidents at Crofton Old Station Level Crossing 6 6 0 0 22/05/2006 26/04/2007 Train collision with a road vehicle at Bratts Blackhouse Level Crossing 8 8 0 0 10/06/2006 30/08/2007 Power door incident at Desborough 9 9 0 0 28/06/2006 18/07/2007 Derailment at Maltby Colliery 4 2 2 2 29/07/2006 29/05/2007 Fatal accident at Deal 9 9 0 0 18/08/2006 08/08/2007 SPAD incident at Purley 5 5 0 0 27/08/2006 24/05/2007 Runaway incident at East Didsbury 8 8 0 0 08/09/2006 21/09/2007 Derailment at Washwood Heath 4 4 0 0 11/09/2006 18/12/2007 Derailment at London Waterloo 14 14 0 0 12/09/2006 13/09/2007 Derailment at Epsom 3 3 0 0 25/09/2006 05/09/2007 Collision at Copmanthorpe 2 2 0 0 31/10/2006 22/08/2007 Collision at Badminton 4 3 1 1 17/03/2007 18/12/2007 Near miss at Tinsley Green 8 80 0 Recs issued 2007 Total 158 131 27 27 Total 258 190 53 68 Indicates investigations where all recommendations have been closed.
Of the 100 recommendations arising from RAIB investigations in 2006, a further 26 were closed out during 2007. This leaves a total of 41 closed and 59 still in progress.
During 2007, RAIB investigations produced 158 recommendations. Of these, 27 (17%) have been closed, leaving a further 131 (83%) in progress.
188 2007 Annual Safety Performance Report Learning from accidents
There were a further 16 events that occurred on NRMI during 2007 which were still under investigation by RAIB at the end of the year. The associated reports will be published in 2008 (see Table 25).
Table 25. Ongoing RAIB investigations (at 31 December 2007)
Incident date RAIB Investigation
05/12/2007 Collision between two road rail vehicles at Glen Garry
29/11/2007 Member of staff hit by train at Reading East
13/11/2007 Member of staff hit by a train at London Victoria
04/11/2007 & Two runaway engineering machine incidents, between Brentwood and 31/10/2007 Romford, Essex, and at Snow Hill, Birmingham
03/09/2007 Derailment at Glasgow
29/08/2007 Overspeeding incident at Ty Mawr
29/08/2007 Workforce accident at Leatherhead
27/08/2007 Single line incident at Aylesbury North
22/08/2007 Incident at Didcot North Junction
10/08/2007 Derailment at Lawley Street
19/07/2007 Train division incident at Camden Road
22/06/2007 Derailment at Ely
29/04/2007 Fatal accident at Ruscombe
23/02/2007 Derailment at Grayrigg, Cumbria
28/01/2007 Runaway wagon at Armathwaite
15/01/2007 Derailment at Kemble
2007 Annual Safety Performance Report 189 Learning from accidents
12.3.1 Analysis of the key themes from recommendations Recommendations issued tend to reflect not only the nature of events occurring, but also the weighting given to particular events by the investigating organisation. In other words, only the tip of the accident/incident/unsafe act or condition pyramid is represented here. It should be noted, therefore, that numeric analysis of recommendation trends would have little statistical validity as it may not reflect actual safety trends. However, the investigations with which we are concerned here have been carried out to a high standard. Thus the ‘qualitative’ information contained therein provides a good indication of current safety management issues within the industry.
In the interests of continuity, we have continued to use the categorisation process applied in previous years to RAIB and formal inquiry recommendations. The details are shown in Table 26.
Table 26. Recommendation categories
Recs Category Description Signalling system Lineside SPAD controls, signal sighting issues, train planning and regulation, operation of the signalling equipment. Competence management Training and development, driver management, competence systems, briefing, assessment, staff selection procedures, drugs and alcohol, fitness for duty, fatigue. Rules, standards and instructions Modification /development of rules and predefined standards for operation, Standards /process change management. Vehicle operation and integrity Train-borne safety equipment, fire protection, vehicle maintenance, train data recorders, crashworthiness, in-cab ergonomics. Infrastructure asset management Managing contractors, track /signalling maintenance operations, work planning, technical specifications, method statements. Event management/ investigation/ reporting SPAD management, public accident investigation, site investigations, post accident management, formal investigations, formal inquiries, public inquiries, fault reporting, emergency procedures. Monitoring and audit Monitoring activities, safety performance monitoring, follow-up processes. Research and development Suggested research topics/ specific areas of research. Safety communications Defining and communicating safety responsibilities, general safety related communications, meetings, techniques, methods and equipment.
Culture Management commitment, organisational change.
190 2007 Annual Safety Performance Report Learning from accidents
Chart 125 illustrates the percentage of the 2007 recommendations falling into each of the category types.
Chart 125. Percentage of RAIB recommendations by category 2007
Signalling system Culture 0.4% 2.5% Safety communications Competence management 14.2% 16.7%
Research and development 5.0%
Monitoring and audit 3.9%
Event mgmt/ investigation/ reporting 5.7% Rules, standards and instructions 32.3% Infrastructure asset management 10.6% Vehicle operation and integrity 8.9%
The areas with most recommendations targeted at them are rules, standards and instructions (32.3%), competence management (16.7%) and safety communications (14.2%). Together they account for almost two-thirds of all recommendations. All three categories relate to the human element and highlight the significant part that people play in most incidents (rules, standards and instructions being the tools to be used to work safely, whether people choose to use them correctly or not, competence management being the person spec and work standard expected to do the job, and safety communication being the interface with other people involved in the process).
Chart 126 presents a comparison between the recommendations issued in 2006 and 2007.
Chart 126. Yearly recommendation categorisation
35 32.3 Recs issued in 2006 Recs issued in 2007 30
25 22.0 20.3 20 16.7 14.6 15 13.8 14.2
10.6 11.4 10 8.9
Percentage of total categories total of Percentage 6.5 5.7 5.0 4.1 5 3.33.9 3.3 2.5 0.8 0.4 0 t t n y it e it en ent r me ions ions ct pm ultu system ru gem cat C nage t and aud ng ns nd integr g uni li i a mana on gnal and i set and develo comm Si s y rds a etence ma perat e Monitorin et p o tur Saf anda t cle Com s ruc Research s, le Vehi ast nfr Ru I Recs Category Event mgmt/ investigation/ reporting
2007 Annual Safety Performance Report 191 Learning from accidents
There are a number of substantial differences in the figures for four of the categories. Recommendations relating to rules, standards & instructions and safety communications have risen in 2007, whereas those relating to culture and vehicle operation & integrity have fallen in 2007.
However, with only two full years of RAIB operation and the numbers and types of investigations on NRMI being relatively low as a sample base (14 in 2006 and 22 in 2007), it is not possible to make any inference from the figures or identify trends at this stage.
12.3.2 Formal inquiry recommendation process RSSB continues to track industry progress on outstanding recommendations from past formal inquiries. This is achieved through regular correspondence with RGMs, asking them to report on progress made against recommendations, or by use of the recommendation tracking tool in SMIS. In 2007, the system of recommendations tracking was improved and many recommendations were found to be closed. For the purposes of this report, a recommendation was marked closed in 2007 when it was communicated to RSSB, even if the work had been completed before 2007.
There are a number of ways that recommendations can be counted and grouped to illustrate different points. Overall, the total number of recommendations (not counting how many RGMs have been allocated that recommendation) outstanding at the beginning of 2007 was 90. Of these, 48 (53%) were closed during the year, leaving 42 recommendations. Three hundred and fifty-five recommendations have been issued since RSSB started conducting formal inquiries, so the 42 outstanding recommendations represent 12% (or an 88% completion of all formal inquiry recommendations).
Chart 127. Number of recommendations outstanding per inquiry before and after 2007
35 31 Before 2007 30 After 2007
25
20
15
Number of Recs of Number 10 10 10 8 7 5 4 4 4 5 3 3 2 2 2 2 2 2 2 1 11111 1 1 111111 111 1 1 000 0 0 000 0 00 0
l d a n d n air n st io am iel l Bar bay on oo nne tf e St ilse Ea e ft Acton a edon nctio a nctio U enh lv Kinc u u T Blake Stichester unct s Tu H J Lim N J h J m eat Heck Ke tters Prideaux C CranberryEarlsw El Harlow Mill m e ley Gr Hednesford g Po e reha ha d Southall g a ig bri Stewarts Lane F w Ed Le Liverpool e N
192 2007 Annual Safety Performance Report Learning from accidents
Chart 127 shows that most inquiries still open at the beginning of 2007 had five or less outstanding recommendations. Now, most have just one or two (the average number of recommendations now open per formal inquiry is in fact 1.33).
Chart 128 details the number of recommendations closed in 2007 and the number still open as a percentage of the total recommendations for each inquiry.
Chart 128. Percentage of recommendations closed per formal inquiry
100% 8 10 11 7 10 11 9 7711 13 17 15 13 17 13 17 22 22 25 25 25 13 33 30 13 18 75% 22 17 38 25 13
50% 92 90 89 93 90 89 93 89 83 83 88 78 78 80 75 75 75 73 % Total Recs Total % 63 67 67 57 25% 50 49
0%
l k ill n y St er n rd t rry io M o Bar ba on s ood tfield f e ft Acton be w Hec a nctio rs U he ls enham t es Kinclairunction u e T Blake ran r s rlow H J NailseaJ Junct a Kelvedon Prideauxthall East Chic C Ea El H ol Lime St e u ley Grea Hedn o Pott e reham Tunne rp So igham e bridg Stewarts Lane Fa e v w Edg L Li e N Closed before 2007 Closed in 2007 In Progress
12.4 Confidential Incident Reporting and Analysis System Learning does not always have to occur after an accident has happened. Valuable lessons can be learnt from near misses, or even what might be seen as an ‘accident waiting to happen’. One way to learn in this manner is to capture the opinions of the workforce, since they have most contact with the operations on the railway, and can quite often be rightly placed to spot possible upcoming problems. Such information is captured via the Confidential Incident Reporting and Analysis System (CIRAS).
Outputs from CIRAS differ from SMIS and conventional accident investigations in that they rarely describe actual accidents and incidents. Reporters to CIRAS focus mainly on ‘near miss’ events or perceived deficiencies in safety systems and arrangements.
Because of the confidential nature of CIRAS, reporters may be able to state their real concerns and describe underlying causes more openly than they would to their manager or at any formal investigation or inquiry. The result is that CIRAS has the potential to give additional insights into the causes. Because of this, CIRAS reports not only serve to supplement evidence from more conventional analyses, but they can describe the potential circumstances for accidents in some detail. CIRAS therefore offers information that can be
2007 Annual Safety Performance Report 193 Learning from accidents
used in a proactive way to reduce the risk of such accidents and incidents occurring in the first place.
12.4.1 Who reports to CIRAS? Chart 129 shows the distribution of reports categorised by occupation of the reporter. It shows that train drivers and track workers are the occupations most likely to report, followed by station staff and signallers, and then a number of categories including conductors, managers and LUL staff. Train drivers have consistently submitted the largest number of reports, but it is particularly positive to note that other groups now raise safety issues too, including office staff and cleaners.
Chart 129. Number of reports in 2007 by job category
80
70 70
60
50 49
40
30 29 28 24 23 22 Number of reports of Number 20 18
11 10 8 5 4444 221 0
r ff ff r ff r er ler to er a ce or ive rk l ta t S T i ne pot ther o Sta s uc s ician e Dr Off lea r erat D O C tion Signa LUL Manag LUL Engineer Op b Cond Elect ound StaffSupervisore nce Sta in Gr h na Track W Su c e a M Maint
12.4.2 Why do people report to CIRAS? Many staff report to CIRAS because they find it difficult to raise sensitive issues with their own company management and they want their concern progressed but in a way that protects their anonymity. However, the majority of issues (78.2%) reported have previously been reported through company channels, a figure almost identical to last year’s figures. Chart 130 shows the reasons why – when the issue had already been reported internally – the reporter felt it necessary to take the issue to CIRAS. More than half (52.3%) believed the response from their company was inadequate; 23.2% claimed that they had received no response whatsoever; and 12% believed the response to be adequate but had not seen any changes or implementation at the work site. However, it should be noted that concerns which get as far as CIRAS may represent a small proportion of all the issues that are pursued through a company’s internal processes.
194 2007 Annual Safety Performance Report Learning from accidents
Chart 130. Reporters’ views of company responses following internal reports (prior to submitting the report to CIRAS) in 2007
Not stated Other 3% 9%
Adequate response, but not implemented 12%
Inadequate response 53%
No response 23%
12.4.3 Key issues of concern in CIRAS reports in 2007 In 2007, CIRAS received 308 reports covering a wide range of topics. The majority (60%) related to issues with the potential to affect workforce safety, and a further group (27%) related to train accident risk.
CIRAS reporters predominantly focus on the potential for accidents or incidents. The reports therefore represent the perspective of the reporter about risks that they identify in the course of carrying out their jobs.
Issues are identified as being ‘key’ based on the perception and concerns of the reporter. However, a crucial part of the CIRAS role is the identification of such potential risk issues, so that they can be analysed and examined further by the rail industry. Some of the individual issues that were raised in 2007 are listed below:
Workforce behaviour • Contractor staff engaged in double and even treble shifting. • Track workers using mobile telephones whilst on lookout duties. • Welders having to cross open lines with heavy equipment. • Staff at maintenance depots not keeping good records. • Manual handling risks for various staff. • Staff taking drugs.
2007 Annual Safety Performance Report 195 Learning from accidents
• Station staff dispatching trains unsafely whilst distracted.
Engineering • Signalling & Telecommunications staff reporting working as two-man teams instead of in preferred three-man teams, putting pressure on lookouts to engage in work as well as carry out lookout duties. • Reports of freight trains with poor braking efficiency.
Public behaviour • Reports of trespassing on rail infrastructure and vandalism. • Members of the public ignoring warning lights and near-misses at level crossings, and motorists similarly abusing rules at user-worked crossings.
Environment • Adverse weather conditions (eg, snow and ice) causing station overruns and inadequate communication. • Flooding risks at stations and on access roads. • Train horns freezing.
Passenger behaviour • Reports of actual and potential verbal abuse of station staff and train crew at various locations. • Confrontations between members of the public and ticket inspectors. • Overcrowding at stations leading to abuse.
12.4.4 Additional CIRAS research outputs An important recent extension of the system in 2007 followed a number of specific approaches from industry bodies and individual companies to go beyond what was already held on the CIRAS database. These requests invited CIRAS to use its guarantee of confidentiality to access staff safety concerns directly, identify key areas of concern and compile more comprehensive profiles of issues. As a result, CIRAS has conducted a number of confidential workshops, interviews and surveys during 2007.
One such survey was carried out for the Infrastructure Safety Liaison Group (ISLG), where a number of staff members answered questions based on their experience over the previous two years. Preliminary results from those questioned reveal the following risks and concerns:
• Nearly 40% of respondents have been instructed to break rules and procedures in order to get the job done, and some on a number of occasions during the last two years. (This finding both supports and substantially extends the findings of the CIRAS report into non-compliance with rules published last year.) • Around 8% of respondents reported having been instructed not to report an incident. This figure is close to 10% for particular job categories.
196 2007 Annual Safety Performance Report Learning from accidents
• There are significant concerns over shift lengths and unexpected overruns, but the greatest concerns about ‘working hours’ relate to travelling time.
12.5 Developments in industry strategy 12.5.1 Learning from accidents workshop and strategy A number of existing activities across RSSB are relevant to the process of learning from accidents – for example, the collection of safety data through SMIS and its analysis through safety performance reports. However, it is true that the term ‘learning from accidents’ is one that is not well-defined, and one that potentially means a variety of things to members across the industry. For this reason, it was decided to hold an industry workshop on the subject, so that any future strategy proposed by RSSB could be shaped by industry input.
RSSB is currently developing a five-year business plan to have effect from April 2009 and is actively seeking views to shape its activities leading up to and including that period.
The industry workshop was held on 24 January 2008, with the overall aim of establishing RSSB’s role in the wider industry approach to learning from accidents. Representatives included train operators, Network Rail, the ORR, RAIB, the European Railway Agency (ERA), engineering and leasing companies, and trades unions. The two objectives of the workshop were to:
• Identify and document the wider industry roles and relationships in the context of learning from accidents. • Develop an agreed and achievable strategy for RSSB’s work in the learning from accidents area, which recognises and meets stakeholder needs and expectations. Based on the outcome of the workshop, RSSB proposes a framework for considering the process of learning from accidents, as follows:
2007 Annual Safety Performance Report 197 Learning from accidents
The core themes of the learning from accidents (LfA) framework are:
Data collection Covers all aspects of gathering information from accidents or safety-related incidents.
Analysis Covers all aspects of how the information gathered is turned into intelligence.
Dissemination Covers the means and methods by which the intelligence is directed to those who need to be aware of it.
Action Covers the range and type of response to the conclusions and recommendations from intelligence gathering, analysis and investigations.
Review Covers the processes by which actions are assessed for their effectiveness.
The two remaining themes are inextricably linked to the core themes:
Strategy Covers the ideas and initiatives that are devised and applied to one or more of the core themes, with the aim of improving its effectiveness or efficiency.
Culture Refers to the social and organisational beliefs and behaviours, which will have an influence on how the core themes are performed.
The framework diagram shows the links between strategy, culture and the core themes as bi- directional. Information coming to light in any core theme has the potential to influence developing strategy or changing culture, as well as strategy and culture affecting how core themes are carried out. For a similar reason, the diagram also depicts a two-way relationship between strategy and culture, since either aspect may inform or influence the other.
Following review by the workshop participants, RSSB will be seeking wider industry endorsement of the work programme arising from the LfA workshop. The RSSB strategy for supporting the industry will then be considered for the 2008/09 work programme and the subsequent five-year 2009–14 Business Plan.
12.5.2 European developments in learning from accidents
European Network of Investigation Bodies The activities of the ERA are based on support from the whole railway sector in Europe. The development of the detailed measures introduced by the Safety Directive requires a consensus-building process with all the stakeholders. Besides working groups for defined subjects the ERA has established networks to support national bodies.
Many of the national investigation bodies (RAIB in the UK) are now established or are currently being built up with the implementation of the Safety Directive in national legislation. They are obliged to exchange views and experience with the support of the ERA.
The aims of this exchange are in particular:
• To develop common investigation methods. • To draw up common principles for the follow-up of safety recommendations in investigation reports.
198 2007 Annual Safety Performance Report Learning from accidents
• To adapt investigations to the development of technical and scientific progress.
The Network meets regularly to exchange information on ongoing investigations and on methodological issues. It will establish internal task forces to elaborate certain subjects in more detail. The reports of the bodies on occurrences that are investigated and the investigation reports will be published on the ERA website.
12.6 Further information
Report of the BP US Refineries independent safety review panel – published January 2007 On 23 March 2005, the BP Texas City refinery experienced a catastrophic process accident. It was one of the most serious US workplace disasters of the past two decades, resulting in 15 deaths and more than 170 injuries.
In the aftermath of the accident, BP followed the recommendation of the US Chemical Safety and Hazard Investigation Board and formed this independent panel to conduct a thorough review of the company’s corporate safety culture, safety management systems and corporate safety oversight at its US refineries. The report makes specific and extensive recommendations. If implemented and sustained, these recommendations can significantly improve BP’s process safety performance.
Although the report is directed to BP, it was intended for a broader audience, as the deficiencies in process safety culture, management or corporate oversight are not limited to that company. Other firms and their stakeholders can benefit from this report. Companies and organisations are urged to regularly and thoroughly evaluate their safety culture, the performance of their safety management systems, and their corporate safety assurance for possible improvements. Many may already have done so, but companies are also urged to review carefully the reports findings and recommendations for application to their own situations.
Preventing process accidents requires vigilance. The passing of time without an accident is not necessarily an indication that all is well and may contribute to a dangerous and growing sense of complacency. When people lose an appreciation of how their safety systems were intended to work, safety systems and controls can deteriorate, lessons can be forgotten, and hazards and deviations from safe operating procedures can become accepted.
Workers and supervisors can increasingly rely on how things were done before, rather than rely on sound principles and other controls. People can forget to be afraid. When systems and controls deteriorate, everything can come together in the worst possible way.
A full copy of the final report can be obtained via the link below: http://www.csb.gov/completed_investigations/docs/Baker_panel_report.pdf
2007 Annual Safety Performance Report 199 Learning from accidents
Intentionally blank
200 2007 Annual Safety Performance Report European developments
13 European developments
In the overall vision for European railways, particular importance is placed on increasing the competitiveness of rail for passengers and freight to allow European railways to grow and expand. While recognising the importance of maintaining current levels of safety performance, safety improvements should only be considered when it is reasonably practicable to make them.
In 2006, the European Railway Safety Directive (RSD) 2004/49/EC was enacted within the UK through the Railway and Other Guided Transport System (Safety) Regulations (ROGS). Infrastructure managers (IMs) and railway undertakings (RUs) are responsible for managing their operations and are required to cooperate where they interface with other IMs or RUs in order to deliver safe operation.
At the same time, the implementation of the High-Speed and Conventional Rail Interoperability Directives by the Railways (Interoperability) Regulations 2006 means that all the Technical Specifications for Interoperability (TSIs) now apply in Great Britain to infrastructure or rolling stock that is new, upgraded or renewed as a part of a major project and will be used on the routes identified as forming part of the Trans-European Network58 (TEN). Prior to these new Regulations, only the High-Speed TSIs had been implemented in this country.
During 2007, considerable progress was made in the development of common approaches to the management and reporting of safety in Europe. This will, in the future, enable the practical development of a legally and technically integrated European railway area.
The establishment of an integrated European railway area requires the development and implementation of a number of TSIs and a common approach to questions concerning railway safety, which are being addressed by the European Rail Agency (ERA).
The ERA was set up (under EC Regulation No 881/2004) to help create the integrated railway area by reinforcing safety and interoperability. Its main task is to develop economically viable common technical standards and approaches to safety, working closely with railway sector stakeholders, national authorities and other concerned parties, as well as with the European institutions.
This section presents an overview of the:
58 Interoperability is being applied, by means of the Railways (Interoperability) Regulations, to the UK part of the High-Speed rail system and the Conventional Trans-European rail system. This includes the infrastructure, and all rail vehicles – passenger, freight, locomotives and maintenance – that use those systems. The Regulations do not apply to the remainder of the heavy rail UK rail system, or to other systems such as metros and light rail. The TEN in the UK is managed by three infrastructure managers: • Network Rail is the infrastructure manager for the majority of the mainland network, comprising approximately 30,100 track km, of which approximately 70% is on the TEN and the remainder domestic railway outwith the scope of the Regulations • Northern Ireland Railways manages the infrastructure in Northern Ireland – a total of 330 km • The Channel Tunnel rail link is managed by NR(CTRL) with 100% of its 230 track km on the TEN.
2007 Annual Safety Performance Report 201 European developments
• Implementation of ROGS and the Railways (Interoperability) Regulations. • Development of TSIs. • Driver licensing. • Progress towards a common approach to questions concerning safety. • Development of Safety Management Systems and Safety Certification and Authorisation. • Development of common safety targets, methods and indicators.
13.1 ROGS and Railways (Interoperability) Regulations The ROGS regulations came into force in 2006. They are made under HSWA and are administered and enforced by the ORR.
ROGS implements a large part of the RSD, which is intended to harmonise the approach to regulating railway safety across the EU and ensure improved access to the market for rail transport services. Other parts of the Directive are implemented by other legislation such as the Railways (Interoperability) Regulations 2006 and the Rail Accident Investigation Regulations 2005.
In addition to implementing much of the RSD, the ROGS regulations extend broadly similar requirements to railways not covered by the RSD (non-mainline railways), as well as to some other guided transport systems. The RSD, together with other European Commission (EC) initiatives, such as the Directives on Interoperability, aims to promote a more integrated European railway and a standardised railway equipment market.
Some transport operators, for example in industrial premises, are excluded from all or part of ROGS. All are, however, subject to the general duties under HSWA to ensure, so far as reasonably practicable, the health and safety of employees and others, including members of the public, and other health and safety requirements such as the Management of Health and Safety at Work Regulations 1999. A list of other relevant publications is available on the ORR’s website at www.rail-reg.gov.uk.
The Railways (Interoperability) Regulations 2006 came into force that same year. The Interoperability Regulations do not impose an obligation that work must be undertaken, but when major change is made to infrastructure and rolling stock, that is part of or operates on the TEN rail system, those parts of the rail system will have to comply with the Regulations.
13.2 Technical Specifications for Interoperability The development of the third priority Conventional TSIs and any future amendments to the remaining TSIs, is the responsibility of the ERA (http://www.era.europa.eu/). The subsystems being addressed by the third priority Conventional TSIs include Infrastructure, Energy, Rolling Stock (locomotives, traction units and passenger carriages) and Telematic Applications for Passengers.
During 2007, a significant amount of work was completed on the drafting of the third priority Conventional TSIs, which is continuing into 2008. In addition, the ERA has also commenced work on a revision to the Freight Wagons TSI.
202 2007 Annual Safety Performance Report European developments
The RSSB-facilitated Standards Committees provide a forum where representatives of the GB mainline railway can exchange information and views to inform their contributions, through the various European sector organisations, in respect of European standards activities. This includes views on draft TSIs and the identification and development of appropriate GB mainline railway specific cases which need to be submitted to the ERA.
The following TSIs were adopted by the European Commission on the dates indicated and published in the Official Journal of the European Union:
• Revised High-Speed Infrastructure TSI on 20 December 2007. • Safety in Railway Tunnels TSI on 20 December 2007. • Persons of Reduced Mobility TSI on 21 December 2007. • Revised High-Speed Operations TSI on 1 February 2008. • Revised High-Speed Rolling Stock TSI on 21 February 2008. The revised High-Speed Energy TSI was adopted on 6 March 2008 and is yet to be published in the Official Journal.
The translation process now appears to be taking a minimum of 18 months for most TSIs, which means that TSIs are taking significantly longer to come into force following approval at Article 21 Committee than has been the case in the past.
The industry-agreed process for identifying which documents should be proposed to the Department for Transport (DfT) for notification as National Technical Rules (NTRs) to fill ‘open points’ in TSIs is being amended following discussions between RSSB and the Industry Standards Coordination Committee (ISCC). Once the process is agreed, further information about the industry-agreed process for managing NTRs will be available on the RSSB website at: http://www.rssb.co.uk/europe/nntr.asp.
The current status of all the TSIs, links to the most recent publicly available version of the documents and the current set of notified NTRs are contained in the ‘TSI status summary’ file, which is downloadable from the following page of the RSSB website: http://www.rssb.co.uk/europe_tsi.asp.
Further information on the development of the TSIs may be obtained from the ERA website (http://www.era.europa.eu/) or the website of the Directorate General for Energy and Transport (http://ec.europa.eu/transport/rail/index_en.html).
For more information on the TSIs and the British representatives providing input to their development, please contact Adam Milligan, Head of Systems Coordination at RSSB, on 020 7904 7629 or email [email protected].
13.3 Train driver licensing The European Directive on the certification of train drivers (2007/59/EC) was published in the Official Journal of the European Union on 3 December 2007 and member states now have two years to transpose the Directive into national regulations. The requirements of the Directive will be phased in and it requires all drivers to hold a two-part licence: one part with validity across the European Union covering general aspects of train driving competencies
2007 Annual Safety Performance Report 203 European developments
and health issues and a national certificate specifying the routes and types of rolling stock the holder is authorised to drive.
The licences will initially be introduced for drivers employed on cross-border operations. The Directive includes a provision for a member state to apply for a derogation for drivers operating exclusively on the territory of that member state if a cost benefit analysis shows that the costs of applying the provisions outweigh the benefits.
An industry group has been established to work with the DfT and ORR in developing and implementing the UK regulations; the DfT will make the case to the European commission for a derogation for domestic drivers if the evidence from industry substantiates the case.
For more information please contact Graham Arkwright, Safety Planning Manager at RSSB, on 020 7904 7963 or email [email protected].
13.4 Safety Management Systems and Safety Certification and Authorisation The RSD requires all railway undertakings (train and freight operating companies and companies that operate maintenance machines outside possessions) to have a safety certificate issued by the National Safety Authority (NSA) in order to function. In Britain, the ORR has the role of NSA. To be allowed to manage and operate rail infrastructure (including stations), an infrastructure manager must obtain a safety authorisation.
The introduction of the Safety Certificate/Authorisation replaces the previous Safety Case regime.
A Safety Certificate/Authorisation comprises two elements: part A, which comprises the Safety Management System (SMS), and part B, which describes the arrangements for operating on/managing the infrastructure in a particular country. The structure of the SMS is detailed in Annex III of the Directive and is reproduced in Schedule 1 to ROGS.
For an RU, the part A certificate, once granted by the National Safety Authority, will be valid in all member states of the European Union. The part B certificate will only be valid on the infrastructure for which it was granted. Should a railway undertaking with a part B certificate wish to operate in a country and on infrastructure other than that for which it was issued, it must apply for a further part B certificate relevant to the said infrastructure from the NSA of the country in question.
The ERA has established the Safety Certification working group (comprising industry and government representatives) to develop proposals for common safety certification acceptance criteria. The purpose of the group is to draft and recommend to the Commission harmonised formats for both safety certificates and their applications, including the essential details to be provided, and the common assessment criteria involved. The ERA is due to complete these tasks by 2010, but existing national processes will continue in the meantime.
In the second half of 2007 and early 2008, the working group moved on to scrutinise ERA- prepared drafts of the next position papers in the series:
II. Part A – SMS Assessment Procedures (for NSAs):
204 2007 Annual Safety Performance Report European developments
Deals with assessment principles focused on the actions and obligations of the National Safety Authorities and is to include definitions of key terms such as ‘Safety Critical Tasks’.59 III. Part A – SMS Assessment Procedures (for RUs & IMs): Is complementary to Position Paper II with a focus on the duties required of railway undertakings and the infrastructure manager. The issues associated with the certification of train drivers and the implication of the certification of wagon keepers is also being addressed by the ERA safety certification team.
IV. Part B – Harmonised Requirements: Provides a general introduction to the Part B issues and concentrates on the requirements described in the Safety Directive – Annex IV and makes links to Safety Directive Annexes II (national rules) and III (Safety Management Systems).
As these tasks are to be completed later in 2008, the group will then turn its attention to:
V. Part B – Harmonised Formats for application guidance documents.
VI. Part B – Assessment Criteria and Procedures.
During 2007, the ERA published the results of a research study on Cross Acceptance using the Class 66 locomotive as an example and the working group finalised the recommendations that ERA will take to the Commission on Position Paper I, dealing with Part A – Safety Management System Assessment Criteria.
The ERA has continued to advocate a ‘Global Approach to Safety Certification’ and that a CEN60 standard for an SMS should be produced. Industry opposition to this proposal in the UK is unabated as the SMS should be a description of how the company has selected to manage its safety risks and is not a ‘tick the box’ template to gain certification. In December, DG TREN called the ERA and sector stakeholders together to review the position but has yet to advise the stakeholders if it has decided to instruct CEN to draft the standard.
13.5 Common safety targets, methods and indicators The RSD requires that common safety targets (CSTs) and common safety methods (CSMs) be introduced gradually to ensure that a high level of safety in Europe is maintained and further improved when reasonably practicable, improve cross-acceptance, and remove barriers to entry.
The CSTs and CSMs will provide tools for the assessment of the safety level and performance at Community level, as well as in the member states. Across Europe, information on the safety of the railways is scarce and, unlike in Britain, not generally available to the public. There is, therefore, also a requirement to establish common safety
59 Of some surprise has been the decision of the ERA to call for tenders on the production of a glossary of railway safety related terms when such a list has already been prepared by the operators through the UIC Safety Platform – Systems Safety Management Group.
60 European Committee for Standardisation.
2007 Annual Safety Performance Report 205 European developments
indicators (CSIs) to facilitate the monitoring of railway safety performance and check that the system complies with the CSTs.
The development of the CSTs, CSMs and CSIs is being led by the ERA, with input from the railway industry and NSAs via structured working groups. The CST and CSM working groups have full cross-industry representation, whereas the CSIs working group only has representatives from the NSAs, with industry input being provided by a single representative from the UIC.
There are a number of representatives from Britain on each of the working groups. If you wish to know who they are, which organisations they represent and how to get in touch with them, please contact Colin Dennis, Head of Safety Knowledge & Planning at RSSB on 020 7904 7499 or, email [email protected].
13.5.1 Common safety targets Article 3(e) of the Safety Directive defines CSTs as:
‘The safety levels that must at least be achieved by different parts of the rail system (such as the conventional rail system, the high speed rail system, long railway tunnels or lines solely used for freight transport) and by the system as a whole, expressed in risk acceptance criteria.’
During 2007 there has been a significant development in relation to the above definition in that the ERA has concluded that it is not practical at present to apportion the CSTs/national reference values (NRVs) to different parts of the rail system (such as the conventional rail system, the high-speed rail system, long railway tunnels or lines solely used for freight transport). This will not therefore feature as part of the current work in the area of CSTs, but may be included in future sets of CSTs when more knowledge of safety performance is available. The CSTs will currently only apply to the whole system at the member state level.
The CSTs are being developed in two sets. The first is to be based on an examination of existing targets and safety performance in the member states, and will ensure that the current safety performance of the rail system is not reduced in any of them. The set is required to be adopted by the Commission before 30 April 2009.
The second set of draft CSTs will be based on the experiences gained from the first set (and its implementation), and will reflect any priority areas where safety needs to be improved further. It is required to be adopted by the Commission before 30 April 2011.
All proposals for draft and revised CSTs will reflect the obligations on member states and will be accompanied by an assessment of the estimated costs and benefits. They will contain a timetable for gradual implementation and, where necessary, take account of the nature and extent of the investment required for application.
During 2007, the working group developed the proposals for the first set of CSTs. The underlying assumption behind the first set, as required by the Safety Directive, is that even the worst-performing railway in Europe is acceptably safe. This level will thus form the basis for establishing the minimum requirements for the CSTs. However, to ensure that safety is not compromised in any of those member states already operating at a higher level of safety, the first set of CSTs will be defined in terms of the safety levels in each member state not
206 2007 Annual Safety Performance Report European developments
being reduced. Each member state will therefore have its own set of ‘national reference values’61 (NRVs) for the CSTs, against which it will be judged. The implications of a move towards genuine CSTs that have common values applicable to all member states will be considered for the second set of CSTs.
The CSTs are required to reflect the underlying level of risk on the railways, accounting for both the frequency and consequences of significant accidents. For an accident to be considered as a significant accident it must have caused:
• At least one fatality or one serious injury.62 • Significant damage to rolling stock, track, other installations or environment that is equivalent to EUR 150,000 or more. • Extensive disruptions to traffic whereby train services on a main railway line are suspended for more than six hours. It should be noted that the CSTs and CSIs will only relate to the significant accidents associated with ‘rolling stock in motion’ (collisions, derailments, fires, persons struck by trains, etc) and therefore only represent a subset of the overall risk on the railway.
The risk represented by the CSTs will be measured in terms of the number of fatalities and weighted serious injuries (FWSI)63 per train kilometre, where 10 serious injuries are deemed to be equivalent to one fatality.
Table 27 shows the definitions that have been developed by the working group for the NRVs.
Towards the end of 2006 and early 2007, concerns were raised within the industry as to the specific purpose of CSTs/NRVs.
The ERA has indicated that the CSTs/NRVs are aimed at progressively creating trust between member states in terms of their safety performance and thus avoiding safety being used as a barrier to market opening. Unfortunately, the fact that the NRVs will probably vary considerably between member states makes this much more difficult to achieve in a short timeframe. The evolution of the CSTs is regarded as a multi-step process that will require several phases before there are more harmonised safety levels and, consequently, a truly facilitated ‘interoperability of RUs’.
61 NRVs will quantify the current safety levels of mainline rail transport in each member state. The NRVs will act as a commonly defined baseline for the first set of CSTs.
62 Any person injured who was hospitalised for more than 24 hours as a result of an accident, excluding attempted suicides. This definition differs substantially from the UK’s definition of a major injury. 63 FWSI is not the same as the term FWI used in the rest of this report because FWSI excludes the contribution from minor injuries and shock/trauma, and serious injuries are defined differently from major injuries – see above
2007 Annual Safety Performance Report 207 European developments
Table 27. Proposed NRV definitions
NRV Category NRV Definition number
Passengers NRV 1.1 Number of passenger FWSI per train km. For the purposes of the CSTs, passengers only relate to persons on
board a train at the time of an accident and incidents relating to boarding and alighting are excluded. Persons holding a valid ticket
and intending to travel that are involved in a rolling stock in motion- related accident, eg person falls from platform and is struck by train, will be classified as Others (see below). This is different from the definition of a passenger used to report safety performance in the preceding sections of this report – see definitions in Appendix 7.
NRV 1.2 Number of passenger FWSI per passenger km.
Employees NRV 2 Number of employee FWSI per train km.
Level crossing NRV 3.1 Number of road vehicle occupant and pedestrian FWSI per train users traverse over a crossing. Note: This excludes any on-train employees or passengers killed or injured in level crossing accidents.
NRV 3.2 Number of road vehicle occupant and pedestrian FWSI per train km over a crossing. Note: This excludes any on-train employees or passengers killed or injured in level crossing accidents.
Others NRV 4 Number of other person FWSI per train km, where others are classified as not passengers, employees, level crossing users and unauthorised persons or suicides.
Unauthorised NRV 5 Number of unauthorised person FWSI per train km. Note: This persons on excludes suicides. railway premises
Whole society NRV 6 Total number of passenger, employee, level crossing user, other and unauthorised person FWSI per train km.
The main steps in the evolution of CSTs are as follows:
1) Through the first set of CSTs/NRVs, all member states will get used to measuring and monitoring their safety performance in terms of risk levels in a harmonised way (ie, using common definitions and measurement units). This will enable benchmarking between member states.
2) The second set of CSTs may point out areas of concern where safety should be improved, in order to strive further towards a more harmonised European safety performance.
3) Meanwhile, as more railway subsystems and constituents are built or updated, and new operators enter the market, the application of provisions and measures related to CSMs, TSIs and Safety Certification will facilitate a progressive reduction of the differences in safety levels. This is because:
208 2007 Annual Safety Performance Report European developments
a. The same technical specifications will be used when subsystems and constituents are built or significantly changed, or other decisions covered by the TSIs are taken (the TSIs will, in fact, remain the reference technical documents for the interoperability of subsystems and constituents).
b. For each ‘significant change’ introduced into the railway systems, the harmonised CSM framework (see section 13.5.2) will be applied, regardless of the member state in which the change is taking place.
c. The Safety Certification process will ensure that all RUs will have to go through the same level of scrutiny in all member states and have the same level of procedures and safety management systems, etc, in place.
4) Hence, progressively we might arrive, in the longer term, at a situation where the outcomes of all the above-mentioned activities, and their synergistic effects, have brought the European railway system up to a status in which a truly common European safety target and similar performances and procedures can be achieved.
Last year (2007) also saw the development of the methodology to be used to calculate and monitor the NRVs taking due account of the potential for low-frequency, high-consequence accidents. The recommendation relating to the methodology is to be presented to the European Commission during 2008.
The methodology proposed is considered to be a pragmatic approach taking into account: a four-year weighted moving average of FWSI, a tolerance range, whether a single high consequence accident has occurred, the failure to meet the NRV more than once in three years, the underlying number of significant accidents and extensive discussion with the member state involved. The manner by which exceedence of an NRV will be handled by the European Commission is still to be finalised, but a managed safety enhancement plan process is being proposed by the ERA.
An ERA task force is now being established to prepare the guidance that will support the overall recommendation.
13.5.2 Common safety methods Article 3(d) of the Safety Directive defines CSMs as:
‘The methods to be developed to describe how safety levels and achievement of safety targets and compliance with other safety requirements are assessed’.
The CSMs will include: a) Risk evaluation and assessments methods to be used, for example, for managing system change. b) Methods for assessing conformity with requirements in safety certificates and safety authorisations. c) As far as they are not yet covered by TSIs, methods to check that the structural subsystems of the trans-European high-speed and conventional rail systems are operated and maintained in accordance with the relevant essential requirements.
2007 Annual Safety Performance Report 209 European developments
A first set of CSMs, covering at least the methods described in (a) above, are required to be adopted by the Commission before 30 April 2008, with a second set of CSMs, covering the remaining part of the methods as described in (b) and (c), being required to be adopted by the Commission before 30 April 2010.
The CSM (b) is being developed by the Safety Certification working group (see section 13.4 above). Methods for assessing the safety performance of the railway system at the member state level, based on safety indicators, are being dealt with by the CST Working Group, as described in section 13.5.1, above.
The first CSM concerns methods for risk evaluation and assessment. The methods developed by the ERA (with the support of the CSM working group) are predictive methods to be applied in order to assess a priori the safety of any significant changes brought to the railway system. The CSM applies to the whole rail system and covers the assessment of the following changes:
• Construction of new lines or significant changes of existing lines. • Introduction of new and substantially modified technical systems/sub-systems/ equipment. • Operational changes (such as new or substantially modified operational rules and maintenance procedures). • Significant changes within RU/IM organisations. In the CSM, ‘system’ refers to all aspects of a system, including (among others) its development, operation, maintenance and decommissioning.
The CSM is designed to facilitate the access to the market for rail transport services by supporting cross-acceptance through:
• Harmonisation of the risk management processes used to assess the safety levels and the compliance with safety requirements. • Harmonisation of the exchange of safety relevant information between different actors within the rail sector in order to manage safety across the different interfaces which may exist within this rail sector. • Harmonisation of the evidence resulting from the application of a risk management process. The recommendation and guidance for the first set of CSMs were published in draft for comment by the working group at the end of January 2007.
One of the main purposes of this first CSM is to harmonise and facilitate the safety approval process for significant changes introduced in the railway system. The CSM will not prescribe in a detailed manner the methods for risk assessment, but rather identify common principles and evidences used within the risk assessment process.
The overall process for the proposed CSM is shown in Figure 2 at the end of this chapter. It can be seen that the methodology relies on hazard identification followed by an assessment of acceptability via a demonstration that the identified hazards are controlled by:
210 2007 Annual Safety Performance Report European developments
a) The application of codes of practice (TSIs, NTRs, recognised standards, etc).
b) Comparison with similar reference systems (cross-acceptance),
and if neither a) or b) apply,
c) The application of explicit risk assessment and comparison with risk-acceptance criteria.
During 2007, there have been significant concerns raised within the GB rail industry relating to:
a) The degree to which explicit risk assessment will be required as part of the interoperability process and the appropriateness of the use of specified risk- acceptance criteria.
b) The absence of a clear definition of significant changes that require the use of the CSM.
c) The roles, responsibilities and competency requirements for Independent Safety Assessments.
Industry representatives, in close co-operation with the ORR and DfT, are liaising with the ERA to attempt to address these issues during 2008.
13.5.3 Common safety indicators To facilitate the assessment of the achievement of the CSTs, and to provide for the monitoring of the general development of railway safety, member states are required to collect information on common safety indicators and report them via annual reports to the ERA.
The initial CSIs are outlined in Annex 1 of the Safety Directive, covering:
• Indicators relating to accidents. • Indicators relating to incidents and near misses. • Indicators relating to consequences. • Indicators relating to technical safety of infrastructure and its implementation. • Indicators relating to the management of safety. The first Annual Safety Report was submitted by the National Safety Authorities in September 2007 (reporting on the calendar year 2006). To enable the ORR to prepare its reports, the railway companies must notify the ORR in June of each year. Where ROGS relate to the annual provision of data for the production of annual safety reports (Regulation 20(1)(c)), the strategy is for RSSB to undertake SMIS data extraction and provide company- specific data reports for review by the industry party concerned prior to its submission to the National Safety Authority (ie the ORR) as part of its annual safety report. RSSB, in addition, sends a system-level data report to the ORR once the duty holder review period has passed. The annual safety reports that the NSAs submitted to the ERA can be downloaded from the ERA website (http://www.era.europa.eu/).
2007 Annual Safety Performance Report 211 European developments
RSSB’s role in performing the data extraction avoids duplication of effort within the industry. Duty holders retain the responsibility to respond to all other parts of Regulation 20. The ORR has developed guidance material for duty holders on what other information should be contained within the annual safety report to fulfil these requirements. The guidance is available to download from the ORR website http://www.rail-reg.gov.uk/.
Further progress has been made within the CSIs working group during 2007 to agree the common definitions for the CSIs. Reporting for the calendar year 2007 will, where possible, be based on the CSI definitions. Where this is not possible, however, the ERA has stated that national definitions of the indicators can be used.
A review of the CSI requirements in Annex 1 of the RSD is now taking place.
For more information on this on the CSIs, please contact Liz Davies, Strategic Safety Intelligence Manager, at RSSB on 0207 904 7493, or email [email protected].
212 2007 Annual Safety Performance Report European developments
Figure 2. Risk management framework for CSMs
2007 Annual Safety Performance Report 213 European developments
Intentionally blank
214 2007 Annual Safety Performance Report Appendices
Appendix 1. Key safety facts Passengers
Annual totals Passenger fatalities and injuries 2003 2004 2005 2006 2007 Fatalities 10 13 8 8 9 Train accidents 0 5 0 0 1 Moving around station 1 4 5 3 1 Boarding / alighting 1 0 1 0 1 Fall or jump from / leaning out of train 1 0 0 0 2 Other platform edge 7 1 2 3 3 Assault 03021 Major injuries 275 259 254 219 233
Minor injuries 5009 5184 4743 4861 4959 RIDDOR-reportable 1101 1086 1126 1090 1160 Non-RIDDOR-reportable 3908 4098 3617 3771 3799 Incidents of shock 312 218 241 317 325 Class 1 10244 Class 2 311 218 239 313 321 Fatalities and weighted injuries 47.2 48.8 42.9 39.5 42.5 Train accidents 0.1 6.2 0.2 0.2 4.3 Moving around station 23.3 26.0 28.6 24.0 20.1 Boarding / alighting 8.3 6.4 6.5 3.5 6.1 On board injury 4.4 3.6 3.1 4.0 3.3 Fall or jump from / leaning out of train 1.2 0.4 0.0 0.0 2.0 Other platform edge 8.3 2.2 3.2 4.4 4.6 Assault 1.6 3.9 1.3 3.4 2.2 Other passenger FWI 0.0 0.0 0.0 0.0 0.0 FWI rate per million passenger journeys 0.047 0.046 0.040 0.035 0.038
2007 Annual Safety Performance Report 215 Appendices
Workforce
Annual totals Workforce fatalities and injuries 2003 2004 2005 2006 2007
Fatalities 39522 Track workers 28302 Train drivers 0 1 1 1 0 Train crew 0 0 0 0 0 Station staff 0 0 0 0 0 Other workforce 1 0 1 1 0 Major injuries 155 219 182 130 128 Track workers 87 143 100 76 70 Train drivers 12 10 7 7 16 Train crew 28 39 39 26 13 Station staff 22 20 22 12 20 Other workforce 6 7 14 9 9 Minor injuries 6761 7034 6600 6428 5837 RIDDOR-reportable 903 862 741 706 596 Non-RIDDOR-reportable 5858 6172 5859 5722 5241 Incidents of shock 1855 1497 1577 1434 1446 Class 1 324 247 261 254 227 Class 2 1531 1250 1316 1180 1219 Lost time injuries 1705 1748 1645 1537 1338 Fatalities and weighted injuries 32.0 43.9 35.4 26.7 25.4 Track workers 13.6 25.2 15.6 9.8 10.8 Train drivers 3.8 4.1 3.8 3.9 3.7 Train crew 7.5 8.9 8.3 7.1 5.3 Station staff 4.9 4.4 4.6 3.4 4.2 Other workforce 2.2 1.2 3.0 2.6 1.4 Near misses with workforce 124 127 103 98 74
216 2007 Annual Safety Performance Report Appendices
Members of the public
Annual totals Public fatalities and injuries 2003 2004 2005 2006 2007 Fatalities 240 239 261 299 263 Accidental 54 42 59 45 57 Level crossing users 14 8 11 5 11 Pedestrians 95858 Road vehicle occupants 53303 Trespassers 37 31 46 32 43 Trespass at station 14 12 14 13 24 Trespass not at station 23 19 32 19 19 Trespasser struck by train 29 25 38 20 33 Child trespasser fatalities 62510 Other non-trespass fatalities 33283 Suicides 186 197 202 254 206 Suicides at level crossings 14 20 23 23 20 Suicides at stations 75 74 67 90 75 Major injuries 85 74 68 98 72 Accidental 55 44 45 56 46 Level crossing users 135873 Trespassers 30 25 20 37 30 Other non-trespass major injuries 12 14 17 12 13 Attempted suicides 30 30 23 42 26 Minor injuries 104 146 124 165 149 RIDDOR-reportable 32 52 52 61 42 Non-RIDDOR-reportable 59 88 64 94 96 Attempted suicide 13 6 8 10 11 Fatalities and weighted injuries 248.8 246.8 268.2 309.3 270.6 Accidental 59.7 46.8 63.8 51.0 61.9 Level crossing users 15.4 8.6 11.9 5.8 11.3 Trespassers 40.1 33.6 48.1 35.8 46.1 Other non-trespass FWI 4.3 4.5 3.8 9.4 4.5 Suicide 189.1 200.0 204.3 258.3 208.7
2007 Annual Safety Performance Report 217 Appendices
Train accidents
Annual totals Train accidents 2003 2004 2005 2006 2007 RIDDOR-reportable train accidents 1065 1102 833 821 785 Involving passenger trains 886 921 680 672 662 Potentially Higher Risk Train Accidents 31 33 27 18 27 Collisions between trains 2 3 1 3 3 Derailments 7 7 8 4 7 Striking road vehicle not at level crossing 1 4 4 2 5 Striking buffer stops 0 6 1 1 2 Striking road vehicle at level crossing 21 13 13 8 10 Other train accidents 855 888 653 654 635 Open door collisions 36 17 3 1 1 Roll back collisions 6 9 7 7 3 Striking animals 97 104 121 121 115 Struck by missiles 305 276 232 215 218 Train fires 240 272 139 134 105 Striking level crossing gates / barriers 0 4 3 3 3 Striking other objects 171 206 148 173 190 Not involving passenger trains 179 181 153 149 123 Potentially Higher Risk Train Accidents 28 31 24 28 22 Collisions between trains 3 3 2 1 1 Derailments 2123172318 Striking road vehicle not at level crossing 1 1 2 1 1 Striking buffer stops 0 1 0 0 1 Striking road vehicle at level crossing 3 3 3 3 1 Other train accidents 151 150 129 121 101 Open door collisions 1 0 0 0 0 Roll back collisions 0 0 0 0 0 Striking animals 7 7 19 13 13 Struck by missiles 73 66 57 70 52 Train fires 46 37 27 6 10 Striking level crossing gates / barriers 0 1 0 3 2 Striking other objects 2439262924 Fatalities (excluding suicides) 7103 1 4 Passengers 0 5 0 0 1 Workforce 0 2 0 0 0 Members of the public 7 3 3 1 3 FWI (excluding suicides) 8.6 11.9 3.8 1.8 8.1 Passengers 0.1 6.2 0.2 0.2 4.3 Workforce 0.9 2.6 0.6 0.6 0.7 Members of the public 7.6 3.1 3.0 1.0 3.1
218 2007 Annual Safety Performance Report Appendices
Train accident precursors
Annual totals Precursors 2003 2004 2005 2006 2007 Infrastructure failures Environment: adhesion 169 104 191 77 80 Environment: flooding 36 92 64 45 113 Environment: landslips 1420222230 Level crossing failures 2455 2091 2260 2630 2442 Other structural failures 61 70 58 43 99 Track: broken rails 379 333 317 227 191 Track: buckled rails 137 32 56 85 4 Track: level 2 exceedences per mile 1.25 1.06 0.95 0.84 0.70 Wrongside signalling failures 1194 855 690 610 548 Irregular working Runaway trains 19 9 13 10 6 Train speeding 181 203 116 83 78 Other irregular working 3261 3026 3628 3431 3469 Level crossing incidents Level crossing incidents due to weather 23 19 18 19 6 Level crossing road vehicle near misses 222 173 176 158 145 Objects on the line Animals on line 2664 2364 2786 2472 1989 Non-rail vehicles 83 83 66 77 86 Trains striking objects blown onto the line 124 248 154 177 319 Trains striking objects due to vandalism 123 132 98 92 113 Category A SPADs Category A SPADs (total) 392 336 308 337 321 Risk ranked 16+ 179 134 119 115 89 Risk ranked 20+ 54 29 22 19 21 Trains and rolling stock Brakes 35 31 47 52 13 Hot axle box 1656 1617 1139 1003 526 Train fires due to rolling stock failures 126 127 76 70 62 Train fires due to vandalism 133 163 83 60 48 Other rolling stock failures 195 150 109 99 76 Other train fires 24 18 7 20 3 Trespass incidents 11927 11823 12003 12431 11075 Dangerous goods incidents 324 241 144 125 142 Bridge strikes 1855 2057 2027 2098 1604 Rail over road: serious 88994 Rail over road: potentially serious 24 38 22 21 7 Rail over road: not serious 1687 1876 1843 1940 1505 Road over rail: serious 3 6 7 10 2 Road over rail: potentially serious 16 16 23 14 12 Road over rail: not serious 117 113 123 104 74 Vandalism All SMIS reportable vandalism 12436 11129 10098 10391 9539 Line of route vandalism 10426 8970 8352 8753 8142
2007 Annual Safety Performance Report 219 Appendices
Level crossings
Annual totals Level crossings 2003 2004 2005 2006 2007 LTotal fatalities 28 35 36 28 32 L Accidental fatalites 14 15 13 5 12 L Pedestrian fatalities 9 6 10 5 9 L Passengers 0 1 2 0 1 L Members of the public 9 5 8 5 8 L Road vehicle occupants fatalities 5 3 3 0 3 L Train occupants fatalities 0 6 0 0 0 L Passengers 0 5 0 0 0 L Workforce 0 1 0 0 0 L Suicide fatalities 14 20 23 23 20 LCollisions with road vehicles 24 16 16 11 11 L Leading to a derailment 2 3 0 0 0 L No derailment 22 13 16 11 11 LCollisions with gates 0 2 3 4 4 LCollisions with barriers 0 3 0 2 1 LAccidental FWI 15.5 16.9 14.0 5.8 12.4 L Road vehicle FWI (proper use) 0.1 0.0 0.1 0.0 0.0 L Road vehicle FWI (Misuse - violation) 1.0 7.1 0.1 0.0 1.0 L Road vehicle FWI (Misuse - error) 4.5 3.2 3.1 0.0 2.1 L Pedestrian FWI (proper use) 1.4 0.2 0.3 0.5 0.0 L Pedestrian FWI (Misuse - violation) 2.0 2.1 2.1 3.1 5.1 L Pedestrian FWI (Misuse - error) 6.4 4.3 8.3 2.2 4.1 LPedestrian level crossing near misses 203 209 240 221 235 LPedestrian crosses when unsafe 400 480 638 766 648 LRoad vehicle level crossing near misses 222 173 176 160 145 LRoad vehicle crosses when unsafe 637 640 780 777 723 LRIDDOR reportable equipment failures 53 42 87 530 611
220 2007 Annual Safety Performance Report Appendices
Personal security
Annual totals Personal security 2003 2004 2005 2006 2007 PAssaults leading to fatalities 03221 P Workforce fatalities 0 0 0 0 0 P Passenger fatalities 0 3 0 2 1 P Public fatalities 0 0 2 0 0 PWorkforce major injuries 12 18 14 10 7 Workforce minor injuries 1102 1038 1026 934 787 P RIDDOR reportable 164 131 118 118 83 P Non-RIDDOR reportable 938 907 908 816 704 PDays lost from workforce assault 4654 4096 4942 4484 3702 PTotal workforce assualts 3880 4068 4838 4903 4865 P Workforce assualts: physical 2123 2186 2250 2090 1899 P Workforce assualts: threat 295 235 543 607 736 P Workforce assualts: verbal 1462 1647 2045 2206 2230 PBTP recorded assaults 9337 10104 10783 10720 9564 P BTP recorded violent assaults 1295 1326 1590 1654 1497 P BTP recorded common assaults 3984 4110 4392 4282 3952 PBTP public disorder 5904 6320 7448 9441 9801 PBTP robbery 2433 2363 2184 2077 1223 PBTP sexual offences 1151 1242 1099 1151 963
2007 Annual Safety Performance Report 221 Appendices
Stations
Annual totals Stations 2003 2004 2005 2006 2007 SAccidental fatalities 2323242033 S Passenger fatalities 9 8 8 7 6 S Workforce fatalities 0 0 0 0 0 S Trespass fatalities 14 12 14 13 24 S Public non-trespass fatalities 0 3 2 0 3 SMajor injuries at stations 288 285 282 230 234 S Passenger major injuries 240 222 229 188 179 S Workforce major injuries 41 48 48 28 39 S Trespass major injuries 5 9 2 8 9 S Public non-trespass major injuries 2 6 3 6 7 SMinor injuries at stations 6007 6372 6149 6048 6295 S Passenger minor injuries 3653 3909 3737 3708 4020 S Workforce minor injuries 2319 2389 2356 2254 2173 S Trespass minor injuries 10 16 15 13 9 S Public non-trespass minor injuries 25 58 41 73 93 SAccidental FWI at stations 63.9 63.5 64.2 54.6 68.7 S Accidental passenger FWI 40.7 38.1 38.8 33.5 32.4 S FWI from boarding and alighting 8.3 6.4 6.5 3.5 6.1 S FWI from platform edge incidents 8.3 2.2 3.2 4.4 4.6 S FWI from slips, trips, falls 22.2 23.8 24.7 22.8 17.9 S FWI from contact with objects 0.8 1.0 1.4 1.0 1.1 S FWI from other movement accidents 0.1 0.0 0.0 0.1 0.0 S Accidental workforce FWI 8.4 8.8 8.8 6.5 7.5 S FWI from boarding & alighting 1.6 1.7 1.7 0.7 1.5 S FWI from slips, trips & falls 2.3 2.1 2.5 2.1 2.2 S FWI from assaults 2.6 3.0 2.6 1.9 1.8 S FWI from contacting objects 1.3 1.3 1.3 1.0 1.4 S FWI from manual handling 0.3 0.5 0.3 0.6 0.5 S FWI from other incidents 0.3 0.2 0.3 0.2 0.2 S Accidental public FWI 14.8 16.7 16.7 14.6 28.8
222 2007 Annual Safety Performance Report Appendices
Appendix 2. Fatalities in 2007 Accidental fatalities
Accidental fatalities
Passenger 9 Station Date Location Territory Description Operator A passenger was struck and fatally injured at Wokingham manually controlled barrier South West 24/01/2007 Wokingham SE (MCB) level crossing. The young male climbed over the barriers to join a train in the up Trains platform and was struck by a down service
ARRIVA A child was struck by a train at Treorchy station. The boy was leaning over the platform 25/01/2007 Treorchy station W Trains Wales edge wearing earphones and failed to see or hear the train’s approach. Haddenham & Chiltern A passenger fell between the train and the platform whilst alighting at Haddenham & 13/02/2007 Thame Parkway LNW Railway Co. Thame Parkway station. station A male was found with head injuries beside the line near Little Bowden. The N/A - running 23/02/2007 Little Bowden LNE circumstances behind the fatality have yet to be determined, but the injuries appear to be line consistent with a person falling or jumping from a train.
A passenger was killed when a passenger train derailed at Grayrigg in Cumbria. The 23/02/2007 Grayrigg LNW N/A - on train immediate cause of the derailment was deemed to be the stretcher bar arrangement at a set of points, which resulted in a loss of gauge integrity at the switch blade.
A passenger was struck by a train approaching the platform at Cambuslang station when ScotRail 07/11/2007 Cambuslang Sc attempting to flag it down. The passenger was reported to be under the influence of Railways alcohol at the time of the incident.
A passenger reported to be under the influence of alcohol collapsed and fell from the ScotRail 21/11/2007 Glengarnock Sc platform at Glengarnock station. As he tried to climb back up, he was struck by a passing Railways freight train.
N/A - running A passenger fell from a train near Llansamlet. Reports suggest that the passenger 24/11/2007 Llansamlet W line jumped out of an HST window following a domestic dispute on board the train.
TransPennine A passenger was stabbed on the platform at Dewsbury station. Two men, aged 17 and 15/12/2007 Dewsbury station LNW Express 18, were later charged with murder.
Workforce 2 Date Location Territory Employer Description A track worker was struck and killed by an empty coaching stock train at Ruscombe Ruscombe 29/04/2007 W Network Rail Junction. It is reported that he failed to respond to warning horns from the power car and Junction two lookouts A track worker (blockman) was struck by a train at Kennet Bridge near Reading. The 29/11/2007 Reading W Network Rail incident occurred after the possession was given up in the early hours of the morning
Public (not including suicides) 57 At stations 3 Station Date Location Territory Description Operator London A member of public fell between the train and platform whilst banging on the train windows 11/01/2007 Gidea Park SE Eastern as it was pulling out of Gidea Park. The person was reported as being under the influence Railway of alcohol at the time of the incident. London A member of public was struck by a non-stopper at Seven Kings, having been too close to 23/02/2007 Seven Kings SE Eastern the platform edge. Railway A member of public fell onto the track into the path of an oncoming train at Shoreham. The South Eastern driver reported that a group of youths had fighting on the platform when two of them fell to 22/08/2007 Shoreham SE Trains the track. The other youth was electrocuted by the third rail, but survived with major injuries.
2007 Annual Safety Performance Report 223 Appendices
Level crossings 11 Date Location Territory LC type Description A male was struck on Paggetts footpath level crossing. It was reported that he may have 01/01/2007 Paggetts SE FP been playing ‘chicken’ whilst under the influence of alcohol.
A 14-year-old girl was struck at Johnstown level crossing, near Ruabon, whilst trying to 15/01/2007 Johnstown W FP retrieve her shoe. She was part of a group who appear to have been playing ‘chicken’.
25/01/2007 Ballast Hole LNE FP An elderly woman was struck on a foot crossing near Doddington Road.
02/02/2007 Two young men in the car died and the vehicle driver received major injuries when a car Delny Sc AOCL struck the side of a passenger train. The crossing, near Barbaraville, was working 02/02/2007 correctly, and the car driver admitted in court to causing the deaths by dangerous driving.
A car driver died when he was hit by a train while attempting to zig-zag between the 01/03/2007 Swainsthorpe SE AHB crossing barriers, which were down.
Sandringham An elderly woman walking her dog was struck at Sandringham Avenue user worked 31/07/2007 LNW UWC-T Avenue crossing with telephone level crossing.
A young male jumped over the barriers at Horsham Road closed circuit television level 16/08/2007 Horsham Road SE MCB-CCTV crossing and was struck by a train.
16/09/2007 Windwhistle W FP A person under the influence of alcohol was as struck on Windwhistle level crossing.
12/11/2007 Greendrift LNE FP A young male was struck on Greendrift level crossing near Royston.
A man with a dog was struck by a train on Brimscombe level crossing with miniature 20/11/2007 Brimscombe W UWC-MWL warning lights crossing. The man failed to acknowledge the warnings. Trespass 43 At stations 24 On or about the track 19
Total accidental fatalities 68
Non-Accidental fatalities
Suicides (Public) 206 (currently 145 of these are suspected)
224 2007 Annual Safety Performance Report Appendices
Appendix 3. Potentially higher-risk train accidents Passenger trains
Collisions between passenger trains 3 Date Location Territory Train Operator Description Collision category: coming into station. 28/08/2007 Altrincham LNW Serco Metrolink Two trams had a low speed collision on at Altrincham; one was at the stops and the driver of the approaching tram misjudged his braking and collided with it.
Collision category: coming into station. 02/09/2007 Manchester Airport LNW Transpennine Express A train collided at low speed with a stationary train at the stops. The investigation is ongoing. Collision category: coming into station. A train collided at low speed with another train at the stops, when the driver of the 28/11/2007 Manchester Airport LNW Transpennine Express approaching train incorrectly attempted to couple the units, without first stopping short.
Buffer stop collisions 2 Date Location Territory Train Operator Description The driver reportedly lost consciousness sometime between entering the platform 11/01/2007 West Kirby LNW Merseyrail and colliding with the buffer stop and was unable to bring the train to a stand in sufficient time. The driver failed to apply sufficient brake application to stop the train at the required National Express East 10/10/2007 Sheringham SE point. In addition, there is the possibility that the driver was distracted and that he Anglia aimed to stop too close to the stops due to short platforms.
Derailments (excluding at level crossings) 7 Date Location Territory Train Operator Description 13/01/2007 Merstham SE Southern Train ran into a landslip. 15/01/2007 Duncraig Halt Sc First ScotRail Train ran into a landslip and a tree.
15/01/2007 Kemble W First Great Western Train ran into an obstruction.
18/01/2007 Hookagate W Arriva Trains Wales Train ran into a tree.
Deterioration of components in the stretcher bar system led to the left hand switch 23/02/2007 Grayrigg LNW Virgin West Coast rail becoming disconnected from the right hand switch rail, leading to an uncommanded movement towards the left hand stock rail.
The front bogie of the train had become derailed after running into a tree that had 06/03/2007 Godstone SE South Eastern Trains fallen onto the line as a result of high winds.
One wheelset in the leading vehicle was derailed because the points had been left 15/12/2007 Worcester Park SE South West Trains in the wrong position.
Collisions with road vehicles (and no derailment) 15 Struck road vehicle on level crossing (see Table 17 for details) 10 Struck road vehicle at locations other than level crossing 5 Date Location Territory Train Operator Description Train struck a two-wheel permanent way scooter that had been used during T3 20/05/2007 Heaton Norris Jcn LNW Transpennine Express possession the preceding night. The train was the first through the section since a T3 possession was given up. 29/05/2007 Keith Sc First ScotRail Train struck a car that had come off the road following a road traffic accident. Garnqueen North Train struck a van that was parked less than 2m from the line. The van belonged to 19/08/2007 Sc First ScotRail Jcn a plant fitter.
Train struck a car that had come off the road and through the fencing after trying to 03/09/2007 Normanton LNW Central Trains avoid collision with a lorry.
17/09/2007 Markinch Sc First ScotRail Train struck a car that had come off the road following a road traffic accident.
Total passenger PHRTAs 27
2007 Annual Safety Performance Report 225 Appendices
Non-passenger trains
Collisions between trains 1 Date Location Territory Train Operator Train type Description Category: In running open track. 11/10/2007 Parkandillack W EWS Freight Collision with a stabled wagon occurred during a shunt move, resulting in the derailment of three wagons.
Buffer stop collisions 1 Date Location Territory Train Operator Train type Description ECS multiple An empty coaching stock train collided at low speed with the buffer 29/10/2007 Liverpool Lime Street LNW Northern Rail unit stops, following a coupling move.
Derailments only 18 Date Location Territory Train Operator Train type Description Loco Mechanical failure of the track led to spreading of the gauge at a set of 04/01/2007 Edinburgh Waverley Sc First ScotRail hauled/push- points, resulting in the 3rd, 4th and 5th vehicles of the train becoming pull ECS derailed at low speed. ECS multiple The driver proceeded past a signal without authority, leading to the 13/01/2007 Aberdeen Sc First ScotRail unit derailment at a set of points. The locomotive of a freight train derailed on a set of points while 28/02/2007 Lincoln High St LNE EWS Freight propelling into a siding. A runaway wagon, which was not attached to a train, was derailed by 05/04/2007 Washwood Heath LNW EWS Freight catch points. A trolley which was attached to a rail-grinder, derailed when a securing On-track 06/04/2007 Wood Street SE Network Rail bar became detached. Approximately 1.5 miles of track was damaged. machine
The rear vehicle of an ECS train derailed on pointwork when shunting ECS multiple 07/04/2007 Canterbury West SE Southeastern into a siding unit
The last axle of a freight train became derailed on ground-frame points. 16/04/2007 Calvert South LNE Freightliner Freight The 22nd to 24th wagons of an empty coal train derailed whilst King Edward Bridge traversing a set of points due to the combination of a minor track defect 10/05/2007 LNE EWS Freight South Jcn and a distorted wagon frame.
The leading wagon of a freight train derailed on a set of points while Stapleford and 14/05/2007 LNE EWS Freight being propelled into a siding. The vehicle concerned had bad wheel- Sandiacre flats, which caused the derailment. The wagon immediately behind the locomotive derailed on junction 12/06/2007 Madeley LNW EWS Freight pointwork.
The wagon involved in the Madeley derailment was running on a wheel- 12/06/2007 Basford Hall Jcn LNW EWS Freight skate when it became derailed on pointwork.
Eleven wagons of a freight train became derailed on plain line, causing 22/06/2007 Ely Dock Jcn SE EWS Freight considerable damage to a rail-over-river bridge. The line was closed fiTwo vehicles tht towards the thbid middle of the formation of a freight train Lawley Street (Landor derailed whilst traversing point work. 10/08/2007 LNW Freightliner Freight St) terminal
ECS multiple An ECS train became derailed at trap points following a category A 30/08/2007 Paignton W CrossCountry unit SPAD. An ECS train derailed whilst moving from the sidings to the main line. ECS multiple 03/09/2007 Exhibition Centre Sc First ScotRail The rear coach was derailed by all wheels within a tunnel. unit One of two road/rail vehicles which was involved in unloading rails, On-track 12/11/2007 St. Ives W Network Rail derailed and fell from a viaduct to the road approximately 20 feet below. machine
Two wagons, ran away and collided with buffer stops, subsequently 19/11/2007 Mountsorrel LNE EWS Freight becoming derailed. Two wagons from a freight train derailed whilst exiting Tyseley depot. 18/12/2007 Tyseley LNW Freightliner Freight The incident is still under investigation.
Collisions with road vehicles 2 Struck road vehicle on level crossing (see Table 17 for details) 1 Struck road vehicle at locations other than level crossing 1 Date Location Territory Train Operator Train type Description Loco Mechanical failure of the track led to spreading of the gauge at a set of 04/01/2007 Edinburgh Waverley Sc First ScotRail hauled/push- points, resulting in the 3rd, 4th and 5th vehicles of the train becoming pull ECS derailed at low speed.
Total non-passenger PHRTAs 22
226 2007 Annual Safety Performance Report Appendices
Appendix 4. Ovenstone criteria adapted for the railways
Requirement: Every railway fatality in Great Britain (including Scotland) is classified as:
• A suicide (that is, in accordance with the coroner’s verdict – or Scottish equivalent), • A suspected suicide (using the criteria provided), or
• Accidental.
A suspected/attempted suicide requires objective evidence of suicide (other than a coroner’s verdict). It is a managerial assessment, based on applying the Ovenstone criteria adapted for the railways.
Without this positive evidence, the fatality should be deemed accidental. A classification should always be reviewed whenever new evidence comes to light (such as during investigations or at a coroner’s inquest).
Whose decision? The classification is a matter for local railway management judgement, based on all available evidence (for example, eyewitness accounts of the person’s behaviour – which may be the train driver’s own account – BTP findings or the coroner’s findings). The classification is wholly for management statistical purposes and is not:
• Passing judgement on the particulars of any case.
• For use outside the Railway Group.
• For any other purpose.
The criteria for suspected or attempted suicide Each of the following, on its own, may be treated as sufficient evidence of suspected suicide (unless, of course, positive evidence that the fatality was accidental exists, or the coroner gives an accidental verdict):
• Suicide note.
• Clear statement of suicidal intent to an informant.
• Behaviour demonstrates suicidal intent.
• Previous suicide attempts.
• Prolonged depression.
• Instability; that is, a marked emotional reaction to recent stress or evidence of failure to cope (such as a breakdown).
2007 Annual Safety Performance Report 227 Appendices
Appendix 5. Level crossing types
The level crossing population on NRMI (as at 31 December 2007)
London London South North North Scotland Western TOTAL East East West MCG 107 17 3 50 14 191 MCB 9135164745234 MCB-CCTV 145 48 18 134 35 380 Total active manual 343 100 37 231 94 805 AHB 187 20 28 172 45 452 ABCL 11 3 3 19 12 48 AOCR 001001 AOCL 299232237120 UWC-MWL 38 9 5 26 10 88 Total active automatic 265 41 60 239 104 709 UWC 320 100 101 226 233 980 UWC-T 391 201 272 309 451 1624 OC 14 6 2 13 16 51 footpath 593 385 87 663 755 2483 Total passive 1318 692 462 1211 1455 5138 TOTAL 1926 833 559 1681 1653 6652
Source: Network Rail. The table excludes disused crossings on mothballed lines and ‘sleeping dogs’ (see Appendix 6). The category of footpath crossings comprises footpath crossings (86%), bridleway crossings (6%) and station foot and barrow crossings (8%). These are analysed as a single category in the ASPR because the data in SMIS is not always precise enough to differentiate between them. They have been collectively grouped under ‘passive’ crossings, but in reality some have automatic protection: 2% (including some at stations) have miniature warning lights and 1% are station crossings with white lights. A further 4% are equipped with telephones and around one-third have whistle boards.
ACTIVE CROSSINGS
Manual crossings
Manually controlled gate (MCG) This crossing is equipped with gates, which are manually operated by a signaller or crossing keeper either before the protecting signal can be cleared, or with the permission of the signaller or signalling system. At the majority of these crossings, the normal position of the gates is open to road traffic, but on some quiet roads the gates are maintained ‘closed to the road’ and opened when required if no train is approaching.
228 2007 Annual Safety Performance Report Appendices
Manually controlled barrier (MCB) MCB crossings are equipped with full barriers, which extend across the whole width of the roadway, and are operated by a signaller or crossing keeper before the protecting signal can be cleared. Road traffic signals and audible warnings for pedestrians are interlocked into the signalling system.
Manually controlled barrier protected by closed circuit television (MCB-CCTV) Similar to MCB crossings, except that a closed circuit television (CCTV) is used to monitor and control the crossing from a remote location.
Automatic crossings
Automatic half-barrier (AHB) AHB crossings are equipped with barriers that only extend across the nearside of the road (so that the exit is left clear if the crossing commences operation when a vehicle is on it). Road traffic signals and audible warnings are activated a set time before the operation of the barriers, which are activated automatically by approaching trains. The barriers rise automatically when the train has passed, unless another train is approaching. Telephones are provided for the public to contact the signaller in case of an emergency or, for example, to ensure it is safe to cross in a long or slow vehicle. These crossings can only be installed where the permissible speed of trains does not exceed 100mph.
Automatic barrier locally monitored (ABCL) As far as the road user is concerned, this crossing looks identical to an AHB crossing. The difference is that train drivers must ensure that the crossing is clear before passing over it. Train speed is limited to 55mph or less.
2007 Annual Safety Performance Report 229 Appendices
Automatic open crossing remotely monitored (AOCR) The AOCR is equipped with road traffic signals and audible warnings only: there are no barriers. It is operated automatically by approaching trains. Telephones are provided for the public to contact the signaller in an emergency. Only one crossing of this type remains on NRMI, at Rosarie in the Scottish Highlands.
Automatic open crossing locally monitored (AOCL) Like the AOCR, this crossing is equipped with road traffic signals and audible warnings only and is operated automatically by approaching trains. The only difference is that no telephone is provided for crossing users: train drivers must ensure that the crossing is clear before passing over it and train speed is limited to 55mph or less. If a second train is approaching, the lights continue to flash after the passage of the first train, an additional signal lights up, and the tone of the audible warning changes.
User-worked crossing with miniature warning lights (UWC-MWL) This crossing has gates or full lifting barriers, which the user must operate prior to crossing. Red/green miniature warning lights, operated by the approach of trains, inform the user whether it is safe to cross.
PASSIVE CROSSINGS
User-worked crossing (UWC) This crossing has gates or, occasionally, full lifting barriers, which the user must operate prior to crossing. The user is responsible for ensuring that it is safe to cross; hence there must be adequate visibility of approaching trains. Once clear, the user is required to close the gate or barriers. These crossings are often found in rural areas, for example providing access between a farm and fields. They often have an identified user, some of whom keep the crossing gates padlocked to prevent unauthorised access.
230 2007 Annual Safety Performance Report Appendices
User-worked crossing with telephone (UWC-T) These are similar to the standard user worked crossing, but a telephone is provided. In some circumstances (for example when crossing with livestock or vehicles) the user must contact the signaller for permission to cross, and report back when they are clear of the track. They are provided where visibility of approaching trains is limited, or the user needs to move livestock over the railway on a regular basis.
Open crossing (OC) At open crossings, which are sited when the road is quiet and train speeds are low, the interface between road and rail is completely open. Signs warn road users to give way to trains. Road users must therefore have an adequate view of approaching trains. The maximum permissible speed over the crossing is 10mph or the train is required to stop at a stop board before proceeding over.
Footpath crossing These are designed primarily for pedestrians and usually include stiles or wicket gates to restrict access. The crossing user is responsible for making sure that it is safe to cross before doing so. In cases where sufficient sighting time is not available, the railway may provide a ‘whistle’ board, instructing drivers to sound the horn to warn of their train’s approach, or miniature warning lights. A variant is the bridleway crossing, which is usually on a public right of way, although some are private and restricted to authorised users. Some footpath crossings are in stations and these can be protected by a white light generally used by railway staff only (which extinguishes when a train is approaching). All these crossing types, some of which clearly have automatic protection, are analysed as a single group in this report because of concerns over the accuracy of crossing type data in SMIS.
2007 Annual Safety Performance Report 231 Appendices
Appendix 6. Definitions
Term Definition In the ASPR, this term refers to an event that causes harm or damage that Accident was not intended by its victims. Suicides are not therefore classed as accidental fatalities. However, injuries sustained as a result of other people’s behaviour (for example, from assaults or trains striking objects that have been deliberately placed on the line) are classed as accidental if the injured party did not intend to come to harm.
Train accidents are accidents occurring to trains and rolling stock. See Chapter 8 (Table 11) for further details. Individual accidents are accidents to people on railway premises or on trains, but excluding injuries sustained in train accidents.
Assault SMIS records incidents in which ‘in circumstances related to their work, a member of staff is assaulted, threatened or abused, thereby affecting their safety or welfare.’ BTP records and categorises criminal assaults in accordance with Home Office rules. In the ASPR, BTP crime codes have been grouped into higher level categories. Violent assaults are the most serious physical assaults (for example, those resulting in actual or grievous bodily harm) and threats to kill. Common assaults are physical assaults that result in more minor injuries and threats in which the victim perceives him or herself to be at immediate risk. Weapons offences cover possession offences as well as incidents where a weapon was used.
Child This term is used in the ASPR to describe a person aged 15 years or below.
Fatalities and weighted An overall measure of safety harm, taking account of injury and fatalities in injuries (FWI) the following way: One FWI = one fatality = 10 major injuries = 200 RIDDOR-reportable minor injuries or class 1 shock/traumas = 1,000 non-RIDDOR-reportable minor injuries or class 2 shock/traumas.
Fatality Death within one year of the causal accident.
Hazardous event An event that has the potential to lead directly to death or injury.
Irregular working Irregularities affecting, or with the potential to affect the safe operation of trains or the safety and health of persons. The term irregular working applies to a disparate set of human actions involving an infringement of relevant rules, regulations or instructions.
Key Risk Area (KRA) A concept introduced by the Strategic Safety Plan (see Chapter 3). There are currently nine KRAs, covering engineering, human error and public behaviour causes of risk. Individually, the KRAs make a significant contribution to the overall safety risk profile of the railway; collectively they represent over 95% of the residual risk on the railway.
232 2007 Annual Safety Performance Report Appendices
Term Definition
Level crossing The ground-level interface between a road and the railway. The different types of crossing are defined in Appendix 5. Improper use refers to occasions when users cross when a train is imminent, but are either honestly mistaken about its proximity and the warnings given by signs, sirens and so on (error), or deliberately disregard them (violation). Proper use refers to occasions when users begin to cross entirely legitimately, but unforeseen events lead to a transgression (as when a motor vehicle breaks down half-way across a crossing, or the level crossing fails due to an error outside the user's control). RIDDOR-reportable level crossing equipment failures relate to any failure of equipment at a level crossing that could endanger users, where the level crossing is on a running line. Note it does not include misuse of equipment. Sleeping dogs are crossings that have fallen into disuse, although individuals may still have the legal right to use them.
Major injury An injury to a passenger, staff or member of the public as defined in Schedule 1 to RIDDOR 1995 (including most fractures, amputations, losses of consciousness), or where the injury resulted in hospital attendance for more than 24 hours.
Minor injury Physical injuries to passengers, staff or members of the public that are not major injuries. For workforce, minor injuries are RIDDOR-reportable if they result in greater than three days’ lost time. For passengers and members of the public, minor injuries are RIDDOR-reportable if the injured person was taken from the accident site direct to the hospital. Other minor injuries are not reportable under RIDDOR.
Network Rail managed This falls within the boundaries of Network Rail’s operational railway and infrastructure (NRMI) includes the permanent way, land within the lineside fence, and plant used for signalling or exclusively for supplying electricity for operational purposes to the railway. It does not include stations, depots, yards or sidings that are owned by, or leased to, other parties. However, it does include the permanent way at stations and plant within these locations.
Ovenstone criteria Explicit set of criteria, adapted for the railway, which provides an objective assessment of suicide where a coroner’s verdict is not available. The criteria are based on the findings of a 1970 research project into rail suicides and cover aspects such as the presence (or not) of a suicide note, the clear intent to commit suicide, behavioural patterns, previous suicide attempts, prolonged bouts of depression and instability levels. See Appendix 4.
Passenger A person on railway infrastructure, who either intends to travel, is travelling or has travelled. Note this does not include passengers who are trespassing or who commit suicide – they are included as members of the public.
Passenger train A train that is in service and available for the use of passengers.
Pedestrian A person travelling on foot. Note that the category also includes cyclists in Chapter 9 (Level crossings).
2007 Annual Safety Performance Report 233 Appendices
Term Definition
Potentially higher-risk Accidents that are RIDDOR-reportable and have the potential to result in train accidents (PHRTA) harm to any or all person types on the railway. They comprise train derailments, train collisions (excluding roll backs), trains striking buffer stops, trains striking road vehicles at level crossings, and trains running into road vehicles not at level crossings (with no derailment).
Precursor A system failure, sub-system failure, component failure, human error or operational condition which could, individually or in combination with other precursors, result in the occurrence of a hazardous event.
Precursor Indicator An RSSB-devised model that measures the underlying risk from train Model (PIM) accidents by tracking changes in the occurrence of accident precursors. See section 8.5.1 for further information.
Public (members of) Persons other than passengers or workforce members (that is, trespassers, persons on business and other persons). Note this includes passengers who are trespassing (when crossing tracks between platforms, for example).
RIDDOR (Reporting of RIDDOR 1995 is a set of health and safety regulations that require any major Injuries, Diseases and injuries, illnesses or accidents occurring in the workplace to be formally reported to the enforcing authority. It defines major injuries and lists notifiable Dangerous Occurrences diseases – many of which can be occupational in origin. It also defines Regulations) notifiable dangerous occurrences, such as collisions and derailments.
Running line A line that is ordinarily used for the passage of trains, as shown in Table ‘A’ of the sectional appendices.
Safety Management A national database used by railway undertakings and infrastructure Information System managers to record any safety-related events that occur on the railway. SMIS data is accessible to all of the companies who use the system, so that it (SMIS) may be used to analyse risk, predict trends and focus action on major areas of safety concern.
Safety Risk Model (SRM) A quantitative representation of the safety risk that can result from the operation and maintenance of the GB rail network. It comprises 125 individual models, each representing a type of hazardous event (defined as an event or incident that has the potential to result in injuries or fatalities).
Shock/trauma Shock or traumatic stress affecting an employee, passenger or member of the public who has been involved in, or a witness to, an event. Class 1 refers to shock or traumatic stress related to being involved in or witnessing fatality incidents and train accidents (collisions, derailments and fires). Class 2 refers to shock or traumatic stress related to all other causes of shock/trauma, such as verbal assaults, witnessing physical assaults, witnessing non-fatal incidents and near misses.
Signal passed at danger An incident when any part of a train has passed a stop signal at danger (SPAD) without authority or where an in-cab signalled movement authority has been exceeded without authority. A category A SPAD is a SPAD that occurs when the stop aspect, end of in- cab signalled movement authority or indication (and any associated preceding cautionary indications) was displayed correctly, in sufficient time for the train to be stopped safely at the signal or end of in-cab movement authority.
234 2007 Annual Safety Performance Report Appendices
Term Definition
SPAD risk ranking tool A tool that gives a measure of the level of risk from each SPAD. It enables the industry’s total SPAD risk to be monitored and can be used to track performance and inform SPAD investigations. The score for each SPAD ranges from zero (no risk) to 28 (a very high risk) and is based on both the potential for the SPAD to lead to an accident and the potential consequences of any accident that did occur. SPADs with risk rankings between 16 and 19 are classified as potentially significant, and those with risk rankings of 20 and above are classified as potentially severe.
Statistical significance A concept used to determine whether a change in accident statistics implies that the safety of the system has really altered, or whether the change could be explained by ‘statistical variation’.
Suicide and suspected A fatality is classified as a suicide where a coroner’s verdict is suicide. It is suicide classified as a suspected suicide where the coroner has yet to return a verdict or returns an open verdict, but where objective evidence of suicide exists based on the application of Ovenstone criteria.
Strategic Safety Plan This is a joint statement by the companies responsible for Britain’s mainline rail network setting out an agreed industry approach to managing safety. The 2007-2009 plan was developed by bringing together commitments made by industry companies in their own individual safety plans, thus creating a linkage with the duty holder planning process.
Track worker A member of workforce whose responsibilities include engineering or technical activities on or about the track. This includes track maintenance, civil structure inspection, S&T renewal/upgrade, engineering supervision, acting as a controller of site safety (COSS), hand signaller or lookout and machine operation.
Trackside This is a collective term that refers to the running line, Network Rail managed sidings and depots.
Train accident See Accident – Train accident. RIDDOR-reportable train accidents are defined in RIDDOR 1995. To be reportable under RIDDOR, the accident must be on or affect the running line. There are additional criteria for different types of accident, and these can vary depending on whether or not the accident involved a passenger train.
Derailment This includes all passenger train derailments, derailments of non-passenger trains on running lines and any derailment in a siding that obstructs the running line. Accidents in which a train derails after a collision with an object on the track (except for another train or a road vehicle at a level crossing) are included in this category, as are accidents in which a train derails and subsequently catches fire or is involved in a collision with another rail vehicle.
Train fire This includes fires, severe electrical arcing or fusing on any passenger train or train conveying dangerous goods, or on a non-passenger train where the fire is extinguished by a fire brigade.
Train striking road All collisions with road vehicles on level crossings are RIDDOR-reportable. vehicle Collisions with road vehicles elsewhere on the running line are reportable if the train is damaged and requires immediate repair, or if there was a possibility of derailment.
2007 Annual Safety Performance Report 235 Appendices
Term Definition
Collision between trains This term describes collisions involving two (or more) trains. Accidents in which a collision between trains results in derailment or fire are included in this category. Roll back collisions occur when a train rolls back (while not under power) into a train on the same line (including one from which it has decoupled). Setting back collisions occur when a train making a reversing movement under power collides with a train on the same line, usually as part of a decoupling manoeuvre. Shunting movement / coupling collisions arise when the locomotive or unit causing a collision is engaged in marshalling arrangements. While they characteristically occur at low speed and involve the rolling stock with which the locomotive or unit is to be coupled, accidents may involve a different train that could be travelling more quickly. Coming into station collisions occur between two trains that are intended to be adjacent to one another (for example, to share a platform) but are not intended to couple up or otherwise touch. Normally, but not always, the collision speed will be low, because one train is stationary and the approaching train will be intending to stop short of the stationary train (rather as for a buffer stop). This operation is known as permissive working. In running (open track) collisions occur in circumstances where trains are not intended to be in close proximity on the same line. The speed of one or both of the trains involved may be high. Collisions in a possession occur where there is a complete stoppage of all normal train movements on a running line or siding for engineering purposes. These collisions are only RIDDOR-reportable if they cause injury, or obstruct a running line that is open to traffic.
Open door collision This occurs when a train door swings outward, coming into contact with another train.
Buffer stop collision This occurs when a train strikes buffer stops. Accidents resulting in only superficial damage to the train are not reportable under RIDDOR.
Trains running into This includes trains running into or being struck by objects anywhere on a objects running line (including level crossings) if the accident had the potential to cause a derailment or results in damage requiring immediate repair..
Trains striking animals This includes all collisions with large-boned animals and flocks of sheep, and collisions with other animals that cause damage requiring immediate repair.
Trains being struck by This includes trains being struck by airborne objects, such as thrown stones, missiles if this results in damage requiring immediate repair.
Train Protection and A safety system that automatically applies the brakes on a train which either Warning System (TPWS) passes a signal at danger, or exceeds a given speed when approaching a signal at danger, a permissible speed reduction or the buffer stops in a terminal platform. A TPWS intervention is when the system applies the train’s brakes without this action having been taken by the driver first. A TPWS activation is when the system applies the train’s brakes after the driver has already initiated braking. TPWS reset and continue incidents occur when the driver has reset the TPWS after an activation (or intervention) and continued forward without the signaller’s authority.
236 2007 Annual Safety Performance Report Appendices
Term Definition
Trajectory A concept developed for the Strategic Safety Plan. There are three aspects to a trajectory: a statement of current safety performance in a particular risk area, details of the actions being taken to address the risk and an estimation of the safety performance improvement that the actions are expected to deliver.
Trespass Trespass occurs when people go where they are never authorised to be, rather than where they behave inappropriately (either from error or violation) at places where they are allowed to go at certain times and under certain conditions, such as level crossings.
Workforce Persons working for the industry on railway operations (either as direct employees or under contract).
2007 Annual Safety Performance Report 237 Appendices
Appendix 7. Glossary
Acronym Expansion ABCL automatic barrier crossing locally monitored AEIF Association Européenne pour l’Interopérabilité Ferroviaire AHB automatic half-barrier crossing ALARP ‘as low as reasonably practicable’ ALCRM all level crossing risk model AMA annual moving average AOCL automatic open crossing, locally monitored AOCR automatic open crossing, remotely monitored ASPR Annual Safety Performance Report ATOC Association of Train Operating Companies ATP Automatic Train Protection ATW Arriva Trains Wales AWS automatic warning system BTP British Transport Police CAA Civil Aviation Authority CCTV closed-circuit television CDRP Crime and Disorder Reduction Partnership CEN European Committee for Standardisation CIRAS Confidential Incident Reporting and Analysis System CIRIA Construction Industry Research & Information Association COSS controller of site safety CP3 control period 3 CPS Crown Prosecution Service CR conductor rail CSI common safety indicator CSM common safety method CSPG Community Safety Partnership Group CSSG Community Safety Steering Group CSSU Community Safety Support Unit CST common safety target CTRL Channel Tunnel Rail Link DB DeutscheBahn DfT Department for Transport DMU diesel multiple unit EC European Commission ECS empty coaching stock EEA European Environment Agency EMU electric multiple unit ERA European Railway Agency ERTMS European Rail Traffic Management System EU European Union EWS English Welsh and Scottish Railway FC footpath level crossing FI fatigue index FOC freight operating company FP footpath level crossing FPN fixed penalty notice FWI fatalities and weighted injuries FWSI fatalities and weighted serious injuries GB Great Britain GSM-R Global System for Mobile communications – Railway HEM hazardous event movement
238 2007 Annual Safety Performance Report Appendices
HEN hazardous event non-movement HET hazardous event train HLOS High Level Output Specification HMRI Her Majesty’s Railway Inspectorate HSC Health and Safety Commission HSE Health and Safety Executive HST High Speed Train HSWA Health and Safety at Work etc Act 1974 IM infrastructure manager ISCC Industry Standards Co-ordination Committee ISLG Infrastructure Safety Liaison Group KRA Key risk Area LC level crossing LfA learning from accidents LNE London North East LNW London North West LOR line of route LOWS Lookout-Operated Warning System LUL London Underground Ltd LX level crossing MAS multi-aspect signalling MCB manually controlled barrier crossing MCG manually controlled gate crossing MOM Mobile Operations Manager MOP member of the public MORI Market and Opinion Research International MPJ million passenger journeys MSL level crossing with miniature stoplights MTM million train miles MWL miniature warning light NCC National Control Centre NCRS National Crime Recording Standards NIR National Incident Reporting NLXSG National Level Crossing Safety Group NPS National Passenger Survey NR Network Rail NRMI Network Rail managed infrastructure NRV national reference values NSA National Safety Authority NTR National Technical Rules OC open crossing OFG Operations Focus Group OHLE overhead line equipment OPSRAM operations risk and mitigation ORR Office of Rail Regulation OTM on-track machine OTP on-track plant PCSO Police Community Support Officer PHRTA potentially higher-risk train accident PICOP Person in Charge of Possession PIM Precursor Indicator Model PINS Police Information System PSR permanent speed restriction RAIB Rail Accident Investigation Branch RGM railway group member RGS Railway Group Standard RIDDOR Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995
2007 Annual Safety Performance Report 239 Appendices
RISAS Railway Industry Supplier Approval Scheme RIV Regolamento Internazionale Veicoli RMMM rail-mounted maintenance machine ROGS Railways and Other Guided Transport Systems (Safety) Regulations 2006 ROSCO rolling stock leasing company ROTS Railway and Other Transport Systems (Approval of Works, Plant and Equipment) Regulations 1994 RPB Risk Profile Bulletin RPI revenue protection inspector RPSG Rail Personal Security Group RRV road–rail vehicle RSD (European) Railway Safety Directive RSSB Rail Safety and Standards Board RTA road traffic accident RU railway undertaking RV road vehicle S&T signal and telecommunications Sc Scotland SCCFG Safety Critical Communications Focus Group SD sustainable development SDO selective door opening SE South East SIT safety improvement team SMIS Safety Management Information System SMS safety management system SoFA statement of funds available SPAD signal passed at danger SRM Safety Risk Model SRP Sustainable Rail Programme SRRT SPAD risk ranking tool SSP Strategic Safety Plan STA significant train accident STUD safety training update day SWAG Security Workplace Action Group SWT South West Trains TEN Trans-European Network TO train operator TOC Train operating company TPWS train protection and warning system TSI Technical Specification for Interoperability TTCI Transport Technology Centre Inc. UIC Union Internationale des Chemins de Fer UWC user-worked crossing UWC-T user-worked crossing with telephone UWG user-worked gated crossing W Western WCML West Coast Main Line
240 2007 Annual Safety Performance Report