REVIEW OF LEVEL CROSSING SAFETY
Types of Railway Level Crossing
In Australia, there are about 9,400 railway level crossings, of which 2,650 (30%) are active crossings and remaining 6,060 are passive crossings (Ford and Matthews et al., 2002). In addition to that there are occupational, cane and private railway level crossings. 'Active' level crossings are the crossings which active protection such as signals and/or boom gates which operate automatically when a train is approaching. 'Passive' level crossings are the crossings which have only signs and/or pavement markings. In Australia there are about 2,400 locomotives in service (ARRB Transport Research et al., 2002).
Cases of crashes
It's clear that the pedestrians are worst effected at the railway level crossings followed by motor vehicles (ATSB et al., 2002). This also explains distinctive crash events, but a major crash involving a bus or a passenger train could result in considerable loss of life and property.
For the past years, the increase in passenger demand, speed of trains, sound proofing, train quietness, size of trucks, freight demand, were expected to increase the severity of crashes. These changes in the courses make it more difficult to reduce these accidents and their effects.
It is known that 60% of the reported deaths at RLCs are the pedestrians but there is no immediate information available readily on a national basis. Therefore at this stage it is little bit difficult to justify improvements on safety, except on a case by case basis. And it becomes critical when disabled persons are involved For example, in Queensland, these safety improvements are generally undertaken in a system wide strategic approach.
There are some cases where the road users are likely to be exposed to risks. Some of them were:
Where motorcycle or bicycle slips on the rails,
Where wheelchair or bicycle wheels may get stuck in the rail flange gap, and Hazards for disabled people.
Therefore at the time of developing Countermeasures, the needs of people with disabilities and other vulnerable users should be particularly considered when developing countermeasures to ensure railway level crossing safety.
Factors of Level Crossing accidents
Whenever accidents occur at RLCs road users tends to blame the traffic control devices. Therefore the engineers need to consider the road user factors in order to plan and design control devices or making any improvements at RLCs. They should be aware of road user characteristics, capabilities, requirements, needs and obligation of users because they will help in designing proper method or better improvements at RLCs (Tustin et al., 1986). There are 3 main factors contributing to accidents at RLC in basic safety engineering studies. They are:
This is one of the major factors responsible for RLCs. This is mainly due to the driver's violation of rules.
This factor is of about the geometric design of the railway track making a clear vision of arriving train.
This factor deals with the environmental conditions prevailing at the RLCs in different seasons and their contribution to accidents.
Caird (2002) reported that many studies related with human factors contributing to accidents at RLC were conducted by many researchers mainly from Australia (Wigglesworth et al., 2001); Sweden (Aberg et al., 1988); Israel (Shinar et al., 1982) and the US (Klein et al., 1994; Lerner et al., 1990).
Familiarity of crossings is found to be the one of the main reason for accidents at RLCs. And on this, Wigglesworth (1978) conducted a case study of accidents occurred in Australia from 1973 to 1977 and he found that 87% of accidents were occurred due to the familiarity of the crossings.
Violation of rules by drivers is another reason for accidents. National Transport Safety Board (NTSB), US investigated in 1998 about 60 accident cases, out of which they found 49 cases were due to driver error. Of those 49 cases, 29 cases include driver's disregard for the stop sign and failure to look for a train. And the remaining cases are related to roadway and track conditions and affecting the ability of the driver to realize the passive crossing ahead and the attendance of approaching a train. Documented evidence (West Net Rail and Australia Western Railroad et al., July 2002) from train drivers indicates many situations where drivers ignore the signs or signals
Also risky behaviour is also one of the reasons for accidents. A case study was done by Witte in 2000 on 891 residents who are selected randomly in Michigan, and he found that 10 to 20 percent of them tried to beat the train which is considered a risky behaviour.
Slowing down of vehicles when approaching these RLCs is also another contributing factor for accidents (Moon et al., 2003, Ward et al., 1996). This occurs because of the misjudgement of the drivers whether to cross or not at passive crossings due to proper lack of vision.
Other factors such as long times may lead drivers to engage in riskier behaviour at crossings (Berg et al., 1982). This 'deliberate risk taking behaviour' results in major risks, particularly where heavy, long or slow vehicles are involved.
Research on Human Factor
This analysis was done by M W Pickett and G B Grayson, who are Researchers at Transportation Laboratory in Berkshire, UK.
Their study has examined a number of aspects of driver behaviour at level crossings.
A preliminary study was conducted by M W Pickett and G B Grayson in 1996. Their study was carried out on a sample of 419 witness statements which they obtained from the British Transport Police, North East Area. The majority of these statements were taken from drivers who had been observed by British Transport Police Officers violated activated warning systems at level crossings protected by automatic half barriers, and at open crossings.
The statements were analysed and classified according to whether the drivers claimed to have been unwilling to stop, unable to stop, or unaware of the crossing.
The results show that over a half of drivers (55%) were unwilling to stop at level crossings when the warning systems were activated. 13% of drivers were unable to stop, while just over one quarter (27%) claimed to be unaware of either the crossing or the lights, and 5% could not be classified. The reason given for doing this ranged from being late for work to simply not wanting to stop. Drivers that were apprehended for crossing while the lights were flashing red, and had claimed they crossed because someone was travelling too close behind, were also caused as unwilling.
Drivers who were on the crossing as the warning system came into operation were classed as being unable to stop. Similarly, those drivers who stated they had been followed too closely by other cars to be able to stop safely were classed as unable to stop.
Drivers were classified as being unaware of the crossing if they did not remember the incident, did not recall seeing the warning lights activated, or did not recall the crossing. Seven percent of violators stated they could not remember the incident.
Fifteen percent of the violators stated that they did not notice that the lights had been activated. One percent of violators reported that the position of the sun prevented them knowing that the lights were flashing.
This analysis has given some insight into why some drivers violate activated warning systems. However it is difficult to make generalisations about behaviour from this sample because of the possibility of misleading statements. It is also difficult to know from this data whether any changes to crossing design would raise driver's awareness, or whether improved driver education would reduce the number of offenders.
From the above data three categories of driver have been identified which are likely to be involved in accidents at level crossings.
those who are unwilling to stop,
those who are unable to stop, and
those who are unaware of the signals.#p#分页标题#e#
In the analysis of witness statements, over half of the drivers were unwilling to stop when the warning systems were activated, and thus continued to cross. Just over one quarter crossed having maintained that they were unaware of the crossing or warning lights, while one in eight drivers held that they were unable to stop at the crossing.
It is necessary for drivers to understand and obey the warning signals in order for level crossing to be effective.
It is of some concern that such a large proportion of drivers are willing to risk crossing against the warning signals. This is not a consequence of any fault with the warning systems, these drivers are aware of the warning signals, but simply choose to ignore them.
Drivers who told they were unaware of the status of the warning signals as they drove over the crossing are a group worthy of more study. The most common reason given for crossing against the warning signal was that drivers said they had not noticed the lights were activated.
Many studies showed that engineering factors includes highway and railway characteristics which are the contributing factors to accidents at RLCs.
This factor comes in to play at the time of designing a railway track. There are about nine components need to be considered. They are There are Annual Daily Traffic (ADT), Number of Passenger Trains, Stopping sight Distance vs. Recommended Sight Distance, Approached Sight Distance vs. Recommended Sight Distance, Speed of Train, Total Numbers of Train, Speed of Highway Traffic, Number of Quadrants Sight is Restricted from and the Clearance Time (Qureshi et al., 2005).
The number of collisions increases with an increase in volume of traffic on the road. An unambiguous relationship was shown by Saccomanno (2003) between the number of collisions and the volume of traffic on the road. Surface width of the road is also another factor with RLCs as it affects vehicle�train collisions as well as among the vehicle-vehicle collisions. As the surface with increases it renders into higher volume of traffic on the road which is directly proportional to the number of accidents at RLCs.
Inadequate lines of sight and warning times generate a tense situation whether to cross the track or not especially in the case of long and heavy vehicles.
Harwood (1990) reported that 11% of the accidents relative to heavy vehicles have been increased to 20% in US. Similarly Tardiff (2001) reported that the percentage of heavy vehicles involved in RLC accidents were increased by 4% from 1990 to 2000 though the original accident rates have been dropped half from 1983 to 2000 in Canada. However, motorcycles have a higher fatality rate. This is due to the lack of driving skills of the driver of the motorcycle. Also angle of the track, number of trains play a significant role in engineering design.
Saccomanno (2003) reported that number of collisions increase with an increase in train speed. It is different as RLCs equipped with gates. It was shown that number of tracks has an effect on collisions at RLC equipped with gates. With the increase in number of trains daily, the number of collisions is expected to increase at RLCs.
This factor contributes to a little extent for the accidents at RLCs. Caird (2002) reported that weather is also an important factor of accidents at passive RLCs. The environmental factors affecting the visibility are snow, fog, heavy rain or mist. The sun can also blind the driver's vision due to its reflection caused by sunrise and sunset. Earlier studies by Meeker (1989) in US showed that, over 57 accidents occurred at RLCs activated by flasher, 56 cases involved visibility problems. This is due to the heavy storm at the location.
Caird (2002) also showed that 40% of the accidents occur between 0930 1530 during Monday Friday and that too in rush hours.
Research on Engineering and Environmental factors
An analysis was done by J.K. Caird, J.I. Creaser, C.J. Edwards, & R.E. Dewar in University of Calgary, Canada in 2002.
Their analysis was a theoretical one and it was totally based on the data which they collected from Transportation Development Centre (TDC), Canada.
Usually the angle at which the highway and railway track meet may sometimes cause visibility problems. Regarding this factor, they obtained the data from Rail Occurrence Database System (RODS) which showed that the accidents are frequents for those crossings which have an angle of less than or equal to 80� or greater than 100�. They crosschecked with the Transport Canada's Integrated Rail Information System (IRIS) database to obtain angles for the crossing. They found that angles provided by both the RODS and IRIS are usually the crossing angles and not the accident angle.
The IRIS showed that more than half of the accidents occurred where the intersection angle was 80or less and this 80could also be sometimes 100(opposite angle).
However, the intersection angle should indicate the approach direction of the highway user and also the direction of the travel. Therefore, the information provided by RODS and IRIS is marginally useful. Collection of accurate angles and the direction of travel for both the train and road vehicle would have improved a resolution of the answer to this line of enquiry.
A higher frequency of accidents in winter months (i.e., November to February) could be attributed to several factors such as fog, and snow. This is due to the driver's incapability to adjust the speed of their vehicles to these unsafe environmental conditions. As they reach a crossing, drivers may attempt to stop but they slide into the crossing and are struck by or slide into the train. A second factor that may increase accident risk during the winter is reduced visibility due to fewer daylight hours, blowing snow, ice fog, and so forth. Thus, trains travelling through crossings are missed for a variety of reasons, including conspicuity. Slow-moving trains under these conditions may pose especially difficult conspicuity issues.
Darkness and fog in combination most likely obscured the visibility of the train, but the extent of the conditions is relative. For example, one investigator may cite “heavy fog” while another simply states “foggy conditions”. The presence of “heavy” in the first case does not adequately differentiate from the “foggy conditions” in the second case.
Sun glare poses a problem for drivers at railway crossings and also it may hinder the detection of cross bucks and flashing light signals at crossings.
Snow can occur throughout the year at any time. Snow conditions may prevent a driver's ability to see the approaching train. Failure to adequately adjust vehicle speed to snow or fog conditions is a common accident contributor.