A1/1996R Train accident at Jokela, on April 21, 1996

On Sunday morning 21 April 1996, a fatal train accident took place at Jokela, Finland. The fast train P82 heading for Helsinki from the north derailed at a switch 35 minutes before its scheduled arrival in Helsinki. The locomotive turned around facing its arrival direction and turned over on the railway embankment with the result of two of the eleven coaches of the train, crashing against the locomotive, four other coaches derailing and two failing off the rails.

The fatal train carried 144 people of whom 4 men died: the engine driver, a passenger travelling in the driver’s cabin, and two passengers in the second coach (a day coach). Moreover 75 passengers were injured, the majority of the Injuries being slight.

An Investigation Commission was set up to investigate the details of the accident.

Catastrophes are regularly a result of several factors. In this case, one factor missing in the chain of events would probably have prevented the accident.

The Jokela accident was caused by the engine driver running in thick fog at an overspeed into the switch. This again was due to the following factors:

The engine driver had admitted a passenger to travel in the driver's cabin without a corresponding obligatory permit. The passenger, a friend of the driver, was in a state of heavy intoxication. The presence of this man at the departure preparations, hampered the driver's possibilities to concentrate on his work.

At the presence of his friend, the engine driver carried out the departure preparations with less care than normally. He neglected to set the train category switch at the correct position, which as for the accident was no decisive importance. He furthermore failed to study carefully enough the written instructions on on-going track work sites, i.e. the so-called weekly instructions, which as such lacked in logic and precision. Two days prior to the accident, the driver had run same section and switch with the data (in the weekly instructions) on a locked western track as not being valid. It is nevertheless to be noted that the weekly instructions only serve as an additional tool in the driving work. The train is driven as conforming with the visible signals.

The engine driver used a slight overspeed at several sections, probably in order to avoid delays in the timetable. The Investigation Commission considered this speeding caused by the driver's endeavour, as required by tradition, to strictly keep up to the timetable, regardless of the running conditions.

At the time of the catastrophe, the Jokela environment was surrounded by thick fog. Due to fog and the excess speed of the train, the driver failed to see or saw incorrectly the distant signal with a green flashing light indicating 35 km/h as the maximum admissible speed for coming cross-over. The train had a speed of 116 km/h at the distant signal point.

In practice, the duration of the flashing light pulse of the distant signal was 0,325 sec and the flashing frequency 1,05 times per second. If the meteorological visibility was 100 m and the luminous intensity produced by daylight, 651 lx, the optimal visibility distance of the green flashing light of the distant signal would have been at 160 m and its poorest visibility distance at 85 m, depending on the driver's preparedness for the viewing. Should there have been double intensity of the daylight and the driver failing to locate the distant signal with any precision, it would have been possible that the signal would not have been perceived at all. If the visibility was, instead of 100 m, only 75 m and the luminous intensity 1300 lx, the distant signal may have been visible 50 - 110 m earlier, or in the worst case, not visible at all.

On the basis of the above, the distant signal light may have flashed as visible in the locomotive 0 to 4 times. There is a significant difference between the visibility of a used and a new distant signal optical element. In the worst case, the visibility distance of a flashing light shown by a used and dirty distant signal element is only half of that of a corresponding new element.

It is however probable that due to the thick fog and the high speed of the train, the driver failed to see the distant signal at all. He looked out very attentively anticipating the distant signal which in fact he already had passed. At this stage, the speed of the train attained even 133 km/h, which the driver preset in the speed adjuster without any further consideration. This particular case involves an assessment error of half a minute in such conditions which another engine driver described as driving " the head under a hood".

According to the Commission, the presence of the passenger in the driver's cabin didn't affect the course of events immediately before the accident. The passenger was probably asleep and was of no assistance in the viewing task.
The engine driver having perceived correctly the lights of the main signal and having actuated the brakes immediately, the accident was nevertheless unavoidable. The maximum admissible speed of the fatal train was 120 km/h. The train approached the main signal at 133 km/h and with the emergency braking on, entered the switch (permitting a maximum speed of 35 km/h) at 124 km/h.

On 12 June 1996 the Investigation Commission addressed VR-Group Ltd a proposal on the promotion of safety, in conformity with paragraph 10, Act on Accident Investigation. The proposal suggests more reasonable and logical weekly instructions to be drawn up. It is also suggests the new programme version under work to be prioritised as an urgent project.

The Investigation Report by the Investigation Commission also includes other recommendations for the promotion of train safety. In addition to the weekly instructions, the recommendations pertain to train driving in exceptional conditions, travelling in the driver's cabin, an accelerated introduction of ATP (Automatic Train Protection) system and equipment.

In view of any eventual accidents, the Commission moreover suggests certain technical modifications to be made in the equipment and structures of the train coaches and cars, and some improvements to be carried out in the rescue apparatus and the corresponding instructions. The data transmission and communication methods designed for accident situations also call for improvement and development.

The damage produced by the Jokela accident corresponds to a total of USD 6 million.

Here you can download the report. Unfortunately only the summary is in English.

Alkusanat (Preface), tiivistelmä (summary in Finnish), summary, sisällysluettelo (Table of contents) (pdf, 0.18 Mt)

A1/1996R Report (pdf, 1.19 Mt)

Appendices (pdf, 0.85 Mt)

Appendix photos (pdf, 3.69 Mt)

  • Recommendation S1
  • Recommendation S2
  • Recommendation S3
  • Recommendation S4
  • Recommendation S5
  • Recommendation S6
  • Recommendation S7
  • Recommendation S8
  • Recommendation S9
  • Recommendation S10
  • Recommendation S11
  • Recommendation S12
  • Recommendation S13
  • Recommendation S14
  • Recommendation S15
  • Recommendation S16
  • Recommendation S17
  • Recommendation S18

Published 21.4.1996