Transportation Safety Board of Canada
www.bst-tsb.gc.ca
Rail Statistics
Foreword
This document provides users of Canadian railway safety data with an annual summary of selected statistics on rail occurrences. It covers federally regulated railways only. Provincial data reported to the Transportation Safety Board of Canada (TSB) are not included in this report. Information in this summary is also posted on this Internet site.
Users of these statistics are advised that, in a live database, the occurrence data are constantly being updated. Consequently, the statistics may change slightly over time. Further, as many occurrences are not formally investigated, information recorded on some occurrences may not have been verified. Therefore, caution should be used when utilizing these statistics. The 2003 statistics presented here reflect the TSB database updated as of 15 March 2004.
To enhance awareness and increase the safety value of the material presented in the TSB Statistical Summary, Railway Occurrences 2003, readers are encouraged to copy or reprint the data presented in whole, or in part, for further distribution (with acknowledgement of the source).
The TSB is an independent agency operating under its own Act of Parliament. Its sole aim is the advancement of transportation safety.
Comments on this document may be forwarded to the following address:
Transportation Safety Board of Canada
Information Strategies and Analysis Directorate
Place du Centre
200 Promenade du Portage
4th Floor
Gatineau, Quebec
K1A 1K8
Telephone: (819) 994-3741
Facsimile: (819) 997-2239
E-mail: communications@bst-tsb.gc.ca
Minister of Public Works and Government Services Canada 2004
Cat. No. TU1-2/2003
ISBN 0-662-68152-5
ACCIDENTS
Overview of Accidents and Casualties (Tables 1, 2 and 3)
In 2003, 1030 rail accidents were reported to the TSB (Figure 1), a 5% increase from 2002 but a 3% decrease from the 1998-2002 average of 1062.
The largest proportion of reported rail accidents are non-main-track related. In 2003, these accounted for 48% of the total (Figure 2). Typically, most non-main-track accidents are minor, occurring during switching operations at speeds of less than 10 mph.
Main-track derailments and collisions accounted for 15% of all accidents in 2003, up from 13% in 2002.
In 2003, 24% of accidents involved vehicles or pedestrians at highway-rail crossings, a proportion which has remained relatively unchanged over the past five years.
In 2003, 225 accidents involved rolling stock or vehicles carrying (or having recently carried) dangerous goods, comparable to the 2002 total of 221 and the five-year average of 228. Of these, 78% were non-main-track accidents. Eight accidents resulted in a dangerous goods release, compared to five in 2002 and the five-year average of six.
Rail fatalities reached a 21-year low of 79 in 2003, down from 96 in 2002 and the five-year average of 98. This decrease is mainly due to a reduction in crossing-related fatalities which totalled 27 in 2003 (Figure 3), a 41% and 31% decrease respectively from the 2002 total of 46 and the five-year average of 39. Trespassing fatalities totalled 46 in 2003, down from 50 in 2002 and the five-year average of 57. In 2003, six employees were fatally injured, up from the five-year average of two.
A total of 81 serious injuries resulted from rail occurrences in 2003 (Figure 4), up from 73 in 2002 but equal to the five-year naverage. Crossing-related injuries totalled 50 in 2003, a 19% increase from the 2002 total and the five-year average of 42.
Freight trains accounted for 80% of trains involved in rail accidents in 2003, followed by single cars/cuts of cars and passenger trains with proportions of 8% and 5% respectively. Although crossing and trespasser accidents continued to account for the majority of passenger train accidents in 2003, the number of passenger trains involved in non-main-track accidents increased to 19, up from the five-year average of six. As most of these accidents occurred while switching equipment in maintenance centres, which meant the passenger cars were empty, none resulted in passenger injuries or fatalities.
Accidents by Type (Tables 4a and 4b, 5a and 5b, 6a and 6b, 7 and 8, and 9, 10 and 11)
Main-track accidents: Main-track accidents (accidents other than crossing or trespasser accidents which occur on main track or spurs) reached a peak of 322 in 1996 (Figure 5). Since 1998, the number of main-track accidents has been relatively constant, with an annual average of 224. Rail activity on main track decreased by 2% over 2002, which resulted in a 13% increase in the accident rate from 2.4 main-track accidents per million main-track train-miles in 2002 to 2.7 in 2003. Notwithstanding, an analysis of the frequency of main-track accident rates using linear regression indicates a statistically significant downward trend1 (p<.05) over the past 10 years.
Main-track collisions and derailments (e.g., where passenger trains are involved or dangerous goods are released from trains that derail while travelling at high speeds in populated areas) are the most serious categories of rail accidents in terms of financial loss and potential risk to the public.
There were six main-track collisions in 2003, fewer than the 2002 total of eight (Figure 6) and the five-year average of 10. No fatalities or serious injuries resulted from main-track collisions in 2003.
A total of 148 main-track derailments were reported in 2003, a 28% increase from the 2002 total of 116 and a 25% increase from the five-year average of 118. This increase was most significant in Ontario where 57 main-track derailments occurred, up from 35 in 2002 and the five-year average of 33. Half of main-track derailments in 2003 involved the derailment of one or two cars, while 17% involved the derailment of more than 10 cars.
In May 2003, two crew members were fatally injured following a freight train derailment on a wooden trestle in McBride, B.C., and two were seriously injured following a freight train derailment in Melrose, Ontario, in February 2003.
In 2003, 38 main-track derailments involved dangerous goods, up from 24 in 2002 and the five-year average of 23. Four of these resulted in a release of dangerous goods, with three leading to an evacuation of the surrounding area.
In 2003, 38% of factors assigned2 to main-track derailments were equipment-related, primarily axles and trucks, compared to 47% in 2002 but equal to the five-year average. Track-related factors, mainly track geometry and rail, accounted for 41% of assigned factors, up from 35% in 2002 and the five-year average of 37%. Factors assigned in an accident are considered to have acted in combination to contribute to the occurrence.
Non-main-track accidents: Non-main-track collisions totalled 104 in 2003, down from 112 in 2002 (Figure 7) but comparable to the five-year average of 105. Derailments occurred in 40% of non-main-track collisions, 77% of which involved the derailment of one or two cars.
No fatalities or serious injuries resulted from non-main-track collisions in 2003.
Dangerous goods were involved in 36% of non-main-track collisions, none of which resulted in a release of product.
Factors assigned to non-main-track collisions are primarily rules-related (89%) (e.g., non-compliance with prescribed procedures). Failure to protect, such as improper positioning of movements and handling of switches, was assigned most often.
There were 388 non-main-track derailments in 2003, up 12% from 2002 (Figure 7) and 2% from the five-year average of 382. Sixty-nine percent of these accidents involved the derailment of one or two cars.
One railway employee was fatally injured in a non-main-track derailment in 2003.
Dangerous goods cars were involved in 36% of non-main-track derailments, two of which resulted in a release of product.
In 2003, 46% of factors assigned to non-main-track derailments were rules-related, comparable to 2002 and the five-year average. Track-related factors also accounted for 39% of assigned factors, a proportion which has been relatively stable over the past 10 years.
Crossing accidents: Crossing accidents represent one of the most serious types of rail accidents in terms of casualties; typically, 25% result in either serious or fatal injuries. Although crossing accidents do not usually result in substantial damage to railway property or equipment, the motor vehicles involved are usually heavily damaged or destroyed.
Crossing accidents reached a 21-year low of 247 in 2003, down from 261 in 2002 and the five-year average of 272. This decrease was due primarily to a reduction in accidents at public passive crossings, from 96 in 2002 to 71 in 2003. The proportion of accidents occurring at public passive crossings decreased from 37% in 2002 to 29% in 2003 (Figure 8). Although there are more than twice as many public passive crossings as public automated ones, more than half of the accidents occurred at automated crossings due in part to higher vehicle and train traffic volumes at these crossings. Accidents at public crossings equipped with flashing lights and bells, totalling 84 in 2003, decreased by 6% from the 2002 total of 89 and by 11% from the five-year average of 94. On the other hand, accidents at gated crossings totalled 49 in 2003, a 26% increase from the 2002 total and the five-year average of 39.
Fatal crossing accidents totalled 23 in 2003, down from 41 in 2002 and the five-year average of 35. Although crossing accidents involving pedestrians account for an average of 5% of all crossing accidents, they account for an average of 20% of fatal crossing accidents. Crossing-related fatalities totalled 27 in 2003, down 41% from the 2002 total of 46 and 31% from the five-year average of 39.
Failing to stop was the most common motor vehicle driver behaviour contributing to crossing accidents (73%), followed by vehicles stopped, stalled or stuck on the track (9%). Accidents involving abandoned vehicles accounted for 6% of crossing accidents.
Crossing accidents were equal to, or lower than, respective five-year averages in most provinces (Figure 9). The most notable decrease was in Alberta, where crossing accidents reached a 21-year low of 38, down from the five-year average of 52. Although the number of crossing accidents in British Columbia and New Brunswick was higher than the five-year average, none resulted in fatalities. However, serious injuries in British Columbia increased to eight in 2003, up from the five-year average of three.
Trespasser accidents: Trespasser accidents involve persons, primarily pedestrians, not authorized to be on railway rights-of-way and who are struck by rolling stock other than at railway crossings. They totalled 68 in 2003, down 7% from the 2002 total of 73 and 16% from the five-year average of 81.
Nearly three-quarters of trespasser accidents occurred in Ontario and British Columbia, which accounted for 59% and 13% of accidents respectively (Figure 10).While the number of trespasser accidents in most provinces has been relatively stable over the past five years, Quebec has seen a decrease in such accidents, from an average of 15 to six in 2003.
In the past five years, virtually all trespasser accidents resulted in casualties (i.e., 68% in fatalities and 30% in serious injuries).
INCIDENTS
Overview of Incidents (Tables 10 and 11)
In 2003, reported rail incidents reached a 21-year low of 295, down from 303 in 2002 and the five-year average of 345.
Statistical analysis using linear regression indicates there has been a significant downward trend (p<.001)3 of reported railway incidents over the past 10 years (Figure 11), due mainly to the large decrease in the number of reported dangerous goods (DG) leaker incidents.
A DG leakage is the unintentional release of a hazardous material while in transportation and does not involve an accident. The vast majority of these incidents involve small quantities of products. There were 151 DG leaker incidents in 2003, accounting for 51% of reported incidents. This total represents a 10% and 24% decrease respectively from the 2002 total of 167 and the five-year average of 198 (Figure 12). In 2003, there were 103 incidents where the movement exceeded the limit of authority, up from 93 in 2002 but comparable to the five-year average of 102.
Factors assigned in non-dangerous goods incidents were primarily operational or rules-related (70%), with the most frequent involving an overlap of authorities or a failure to protect.
Please refer to the Table of Contents for other sections of this summary.
1. It is agreed by convention that, for a result to be considered statistically significant, its probability must be lower than 1 in 20 (i.e. p<.05).
2. Factors assigned are conditions and/or acts that may have played a role in an occurrence.
3. It is agreed by convention that, for a result to be considered statistically significant, its probability must be lower than 1 in 20 (that is, p<.05).