S1/2005R Säkerhetsutredning av plankorsningsolyckor
At the request of VR-Group Ltd, in December 2005 the Accident Investigation Board of Finland commenced a safety study on road/railway level crossing accidents and appointed a commission therefor. The safety study included seven recent level crossing accidents, the first one of which had been subject to investigation before the commencement of the safety study referred to. Moreover the commission investiagted other level crossing accidents having occurred in 2003, 2004 and 2005, on the basis of data collected by VR-Group Ltd. The investigation also included fatal level crossing accidents in 1991–2004 as based on investigation documents produced by the Traffic Safety Commission of Insurance Companies (VALT), statistics from 1991–2004 on level crossing accidents and railway and road traffic accidents, international statistics on level crossing accidents and railway and road traffic accidents, as well as investigation reports on individual accidents in certain countries and documentation pertaining to projects on the development of level crossing safety in some countries.
Among recent railway level crossing accidents, the collision of a Pendolino train with an animal carrying truck on November 10, 2005 on an unguarded level crossing in Kälviä, was investigated, as well as the collision of a freight train with a private car on February 14, 2006 on a level crossing equipped with a light and sound warning system in Kouvola, the collision of a freight train consisting of two locomotives with a private car on March 16, 2006 on an unguarded level crossing in Tornio, the collision of a freight train with a private car on May 5, 2006 on an unguarded level crossing in Närpiö, the collision of a freight train with a private car on May 5, 2006 on an unguarded level crossing in Raahe, the collision of a Dm7 DMU museum train with a private car on June 17, 2006 on a level crossing equipped with half-barriers in Alavus, and the collision of a freight train with a private car on June 21, 2006 on an unguarded level crossing in Ylistaro.
There were altogether three fatalities in the seven accidents investigated: two persons died in one accident and one person in one of the accidents. Five persons suffered mild injuries in the accidents; among them one was a train passenger and one was an engine driver. Four car passengers or drivers were unharmed. Among the accident trains, the Pendolino train was severely damaged while the other trains only suffered minor damage. Among the accident cars, six were completely wrecked and one car was severely damaged.
In all of the individual level crossing accidents investigated by the commission, the travelling of the road the vehicle to the level crossing without stopping was disclosed as the immediate cause of the accident. In three cases, the driver's vigilence in terms of the traffic situation was insufficient as due to his absent-mindedness. In two cases the driver essentially focussed on his manipulation of the road vehicle and in two other cases, the driver's attention was concentrated on other objects than the level crossing. Among environmental factors having contributed to the occurring of the accident, there was in four cases too high a speed limit before the level crossing, in three cases a rising gradient to the level crossing, in two cases the road and railway meeting angle, in one case insufficient visibility, in one case darkness and rain, in one case glaring sunshine, and in one case slippery road.
According to VR's accident irregularities reports, in 2003 and 2004 altogether 52 level crossing accidents took place and in 2005, 64 such accidents occurred. In 2003 seven accidents took place on level crossings equipped with half-barriers, five accidents occurred on level crossings with light and sound warning systems, and the remaining fourty accidents took place on unguarded level crossings. In 2004 five accidents occurred on level crossings with half-barriers, four on level crossings with light and sound warning systems, and the remaining fourty-three accidents on unguarded level crossings. In 2005 four accidents took place on level crossings with half-barriers, four on level crossings with light and sound warning systems, and the remaining fifty-six accidents on unguarded level crossings. In 2003 there were six fatalities in level crossing accidents and 26 persons were injured. Among the injuries, six were serious. The injuries occurred in a total of 22 different accidents. In 2004 there were seven fatalities in level crossing accidents and 14 persons were injured. Among the injuries, three were serious. The injuries occurred in a total of 15 different accidents. In 2005 there were eight fatalities in level crossing accidents and 20 persons were injured. Among the injuries, five were serious. The injuries occurred in a total of 22 different accidents. In almost all level crossing accidents in 2003–2005, the accident was caused by the road vehicle running to the level crossing without stopping.
According to the data produced by VALT, in 1991–2004 a total of 110 fatal level crossing accidents took place in the country, that is, on average 7.9 fatal accidents per year. Over the earlier years under investigation (1991–1997), on average 9.7 fatal accidents took place while thereafter (1998–2004) 6 fatal accidents were recorded on average each year. Among the accidents, 22 occurred on level crossings with half-barriers, 5 on level crossings with light and sound warning systems and 78 on unguarded level crossings (no relevant data was available on 5 cases). Over the investigated period, there were altogether 142 fatalities in level crossing accidents, that is, 10 fatalities on average per year, with a variation ranging from 4 (in 2002) to 20 (in 1991). In almost all fatal level crossing accidents in 1991-2004, the road vehicle ran to the level crossing without stopping. The driver's inappropriate behaviour was in most cases explained by erroneous observation (58%) and anticipation and assessment errors (28%). The driver's inappropriate behavious was moreover explained by the familiarity of the level crossing and the driver as being in a hurry. Among the environmental factors, the most general background factors included the rising gradient of the road to the level crossing, the slipperiness of the road and obstructed sightline. Among factors hampering observation, there were moreover the glare caused by sunshine, fog, darkness and the road-railway crossing angle.
According to the 1991-2004 statistics, over the investigated period a total of 818 level crossing accidents took place in Finland. Among the accidents 106 (12 in 1991, 7 in 2004) occurred on level crossings with half-barriers, 74 (5 in 1991, 4 in 2004) occurred on level crossings with light and sound warning systems, and a total of 638 (80 in 1991, 41 in 2004) on unguarded level crossings. Over the period under investigation, the number of level crossings decreased from 6,634 to 4,635: the number of level crossings with half-barriers diminished from 816 to 772, that of level crossings with light and sound warning systems respectively from 154 to 114, and the number of unguarded level crossings decreased from 5,859 to 3,749.
According to international statistics of 2004, 9.06 level crossing accidents as per a thousand track-km took place in Finland. The rate was only slightly poorer than the average figure of 8.44; however the number of these accidents was twofold in comparison with Sweden and threefold in comparison with Norway and Denmark. Also in terms of train-kilometres, the frequence of level crossing accidents in Finland was clearly greater than in the other Nordic countries and higher than the overall average (in Finland, 1.07 accidents; 0.79 accidents on average as per one million train-km).
As a rule the investigation commission considers the removal of level crossings as regularly being the fundamental and foremost means of improvement of safety; the number of level crossings needs to be drastically reduced, the removal of level crossings has to follow an expedient systematic plan with the authorities maintaining and pursuing a relevant strategy, in the removal work safety has to be prioritized, and finally sufficient public funds are to be allocated to the removal of level crossings and the enhancement of safety.
In view of the removal of level crossings, the investigation commission recommends that:
- stopping should be adopted as the regular model of behaviour at unguarded level crossings where the sightline along the railway is only attained at a distance of 8 metres from the railway;
- the visual range requirements should be modified so as to also consider the possibility of crossing a level crossing without stopping;
- the perceptibility of a train and a level crossing should be improved;
- the speed limit before a level crossing should be reduced to max. 50 km/h;
- the wait platforms at level crossings in poor condition should be upgraded to meet the specifications set forth in the Track Technological Rules and Regulations (RAMO);
- maintenance instructions for level crossings should be drawn up;
- the authority responsible for track maintenance and the safety authority should have the possibility to restrict motor vehicle traffic on level crossings;
- a study should be conducted on the use of train whistles on level crossings and a relevant decision on their use should subsequently be made;
- avoiding of crossing of level crossings ought to be adopted as a basis for route planning for private cars and company road transports;
- in planning of land use, special attention should be paid to safe railway crossing, and the building of new level crossings should be avoided.
The investigation commission reiterates three recommendations published earlier:
- level crossings should be equipped with signboards indicating the name and the location of the level crossing;
- also the general emergency telephone number should be directly called from the scene of a level crossing accident;
- the information system of the Emergency Response Centre should also incorporate the track-km data.