Dutch Safety Board

The Dutch Safety Board performs independent, comprehensive investigations into the causes or probable causes of individual – or categories of – ‘incidents’. An incident is not only defined as the term ‘incidents’ to include not only disasters and accidents but also ‘incidents that could have turned out badly’. The Dutch Safety Board is an autonomous administrative body set up under a Kingdom Act The Board is authorised to investigate incidents in any conceivable field but in practice is currently active in the following sectors: aviation, shipping, rail transport, road transport, defence, healthcare (human and animal welfare), industry and networks, pipelines, construction and services, water, and crisis management and aid provision.

The organisation consists of a Board with five permanent members, in addition to a number of standing committees. Special guidance committees are set up for the purpose of conducting specific investigations. The Dutch Safety Board is supported by a bureau consisting of investigators and support staff. The Safety Board conducts independent investigations into the causes of incidents. Its investigations look for any systematic safety-related shortcomings and it issues appropriate reports to the parties involved and to the general public. Accordingly, investigations constitute our primary process, with the product being a report in all cases. The key goal of this investigation is to establish the truth rather than to apportion blame.

The purpose of the Dutch Safety Board’s work is to ‘prevent incidents or to limit their after-effects’. Accordingly, the Board’s investigation aims not only to uncover the actual causes of incidents but also – and in particular – to bring to light the underlying causes of the incident, so that any shortcomings in the applied system can be revealed. If the investigation reveals any systematic safety-related shortcomings then the Board can formulate recommendations so that these shortcomings can be put right. Any recommendations are usually addressed to the authorities but others may be intended for individuals, organisations or companies.

The reports from the Dutch Safety Board are mainly in Dutch but some are published in English. For the full set of reports please see the Dutch Safety Board Railway Reports. The available English reports are available here for your convenience.

Documents

Order by : Name | Date | Hits [ Ascendant ]

Fire in High-speed Tram at the Weesperplein Metro Station, Amsterdam on 12 July 1999

Dutch Safety Board

On 12 July 1999, a fire occurred in a high-speed tram operated by the Municipal Transport Company of Amsterdam, in the Weesperplein metro station. The smoke development during this fire was such that the entire station, consisting of two underground levels, had to be evacuated. Fire and smoke in underground transport systems represent a particularly major hazard. Smoke rapidly reduces visibility and light levels. As a consequence, the evacuation process is seriously hindered, whilst in general the assistance provided by the fire brigade, due to the special circumstances, can be no more than marginal. For these reasons, the Council decided to initiate an investigation into this fire, in order to learn the necessary lessons.

The investigation clearly indicates that the events of 12 July 1999 are less coincidental than they would appear at first glance. Traction faults which occur frequently on warm summer days, such as on 12 July, interrupt the transport process, reduce the level of trust amongst drivers in the disruption signals issued by the high-speed tram, and contribute to communication disturbances between traffic controllers and tram drivers. Due to a concurrence of circumstances, a fire erupted on 12 July in a high- speed tram, at the Weesperplein metro station. Once the tram driver of the high-speed tram involved and his traffic controller had eventually established a clear picture of the seriousness of the situation, ad hoc decisions were taken, primarily aimed at as far as possible maintaining the level of the transport operation, in a situation where passenger safety in fact calls for other priorities. In addition, the communication infrastructure and equipment available to the traffic controller at that moment, proved insufficient and unsuitable for the tasks the traffic controller is required to carry out, in such circumstances. Furthermore, the distribution of tasks amongst traffic controllers is not geared towards crisis situations.

It also emerged that unlike normal metro stock, the high-speed tram has wooden floors, and that when the stock was delivered, insufficient testing was carried out aimed at tracing shortcomings in the design. The frequently occurring traction disruptions in warm weather have proven to be a technically solvable problem.

Underground transport systems are complicated systems, whilst at the same time the risks to passengers and metro workers are considerable. In the organisation responsible for this system, safety must be built in structurally and systematically. At present, this is not sufficiently the case. The company culture is unilaterally geared towards the transport process.

The Board and Management of the Municipal Transport Company in Amsterdam are advised at both Board and Management level, to focus greater time and attention on safety, as a result of which limiting conditions will be established, for improving safety at all levels in the company. It is also recommended that safety in the transport process be raised, by in the immediate short term improving the fire safety of high-speed tram stock, improving the maintenance process, fully integrating high-speed tram line 51 into the metro system, developing scenarios, and improving the short-circuit procedure for the third rail.

Shunter under rake of wagons in Rotterdam Waalhaven on 20 August 1999

Dutch Safety Board

In rail transport, shunting is unavoidable. The shunting of rail vehicles, unlike all other train traffic, is carried out ‘by sight’. This means that the person with responsibility for shunting, speeds and braking distances to stationary, must make estimates on the basis of personal observations. Generally speaking, the shunter does not have signals to assist him. The shunting process for freight transport takes place prior to and following the transport of freight trains. The process starts with the delivery of empty freight wagons to the customer, via an industry track or harbour line. Once the wagons are loaded, they are collected and transported to a marshalling yard. In this case, the Waalhaven marshalling yard in Rotterdam. At the Waalhaven, the wagons are sorted on the various tracks, according to destination. Once the wagons have been sorted, they are coupled and the air hoses from the continuous train pipe are connected. To complete this work, the shunter must step over the rail, beneath the buffers, in order to come between the wagons. Here, he can carry out the coupling and connection work.

Shunters and radio-controlled train drivers have a hazardous profession. The Framework Memorandum on Railway Safety, submitted by the Minister to the Lower Chamber, demonstrates that it is one of the most hazardous professions in our country. The high levels of risk for this group of professionals are often caused by three tasks: passing level crossings, coupling vehicles and push-shunting. On 20 August 1999, an accident occurred during a push-shunting movement, at the Waalhaven. Due to interference, probably caused by a telerail call, a rake of wagons, consisting of 19 freight wagons pushed by a locomotive, came to an emergency stop. The shunter, who according to the regulations was standing on the footboard of the flat wagon at the front, fell as a result. His injuries left him permanently disabled. In hindsight, this will not prevent him from participating in the employment process, in some other way, in the future. The radio-controlled train driver, who in an improvised manner had joined him on the front of the same wagon, was left uninjured.

The investigation demonstrated that Railion, the company in which the accident occurred, does have manuals and instructions covering shunting work, but does not have effective rules of behaviour for the process of push-shunting using locomotives controlled by radio signals. The manuals and instructions are based on the rules from Railned, which are highly formalised in character, and are above all geared to authorities. It could have been expected that a company like Railion, of which the former NS Cargo is now part, would have translated these rules according to its own safety philosophy, into practical, implementable rules of behaviour, for this process.

It further emerged that the footboards on freight wagons, intended to be ridden on, offer too little protection against the risk of falling. The design of these footboards is laid down internationally in a specification sheet from the Union Internationale des Chemins de fer.

Railion is recommended to:

review the design of radio control for locomotives, making the radio control less sensitive to external interference sources;
in close collaboration with all those involved, lay down rules of behaviour for push-shunting, coupling and the passage of level crossings, with a view to working as safely and efficiently as possible;
ensure that the planning and implementation of the shunting process are supervised and managed by the shunting controller, in order to make implementation of the work as safe as possible;
as soon as possible, introduce a ‘hands-free’ communication system for shunters, for whom the ability to hold on is of life and death importance.

The Minister is recommended to:

take steps within the European Union and the Union Internationale des Chemins de Fer, to improve the standards for footboards for shunters on freight wagons, laid down by the European and international institutions;
instruct Railinfrabeheer, when carrying out alterations to the infrastructure, to install reasonable facilities to prevent or keep to a minimum the need for push-shunting.
[Communications Aspects: EMC]

Passenger Train Derailment at Baarn - 20 August 1999

Dutch Safety Board

On 20 August 1999, a passenger train type Materieel 64, derailed in Baarn. This train had just left the station at Baarn, heading towards Utrecht, crossing the main Amsterdam-Amersfoort tracks. The train derailed at the last set of points where the single track towards Utrecht branches off from the main Amsterdam-Amersfoort tracks. The speed at the moment of derailment was relatively low, approximately 40 km per hour. As a result, the consequences were fairly minor.

When constructing the Dutch railway network, it was assumed that trains would not derail spontaneously. Trains pass obstacles located close to the rails, at speeds of 140 km per hour. As a result, derailments could have very serious consequences. The accident in Eschede in Germany demonstrates this fact. The cause of the derailment in Baarn was the failure of a wheel, due to fatigue cracks. Materieel 64 has been in existence for more than 30 years. The average service life of rolling stock is approximately 30 years. For this reason, additional emphasis on this phenomenon would seem obvious. Fatigue cracks in construction are as such not unavoidable and permissible. However, at least two conditions must then be met. It must be possible to guarantee the tracing of the cracks, and the speed with which the crack is expanding must be precisely known. The standard means of monitoring these factors consists of inspections and monitoring the number of changes in load. For wheels, a rough guide to the latter monitoring is the number of kilometres travelled.

As a carrier, NS Passengers is fully responsible for the stock deployed in passenger transport. NS Passengers does not itself carry out the maintenance on rolling stock. This maintenance is subcontracted to NedTrain. This company, which is part of the NS Holding company, is a maintenance company certified by Railned, on behalf of the government.

The investigation carried out demonstrated that the method used for tracing cracks is not truly reliable. In addition, until the derailment in Baarn, the rule applied that a wheel had to be investigated for cracking, once every two years. In fact, this rule was not complied with. The speed with which cracks developed in the wheels was not known, and no research was carried out into this subject. The scope of the changing loads was also not monitored. As a result, the safe running of the wheel sets could not be guaranteed, and a train was derailed, in Baarn. The conclusion which must be drawn is that the workshops carrying out these investigations were incorrectly unreservedly certified by Railned.

Collision between two passenger trains in Dordrecht 28 November 1999

Dutch Safety Board

On 28 November 1999, two passenger trains travelling in the same direction, collided with each other, in Dordrecht. The trains in question were the Benelux train and a double-decker passenger train. During the collision, 6 passengers were injured, and the trains derailed, whereby a section of the derailed Benelux train came onto the adjacent track. Fortunately, the driver of the train approaching the site on the adjacent track noticed that the track was blocked, and was able to halt his train in time. The nature and seriousness of the collision, which could have been far more serious given the approaching train, led to the decision by the Council for Transport Safety to further investigate this accident.

The investigation by the Council showed that the double-decker had incorrectly passed a signal at red, which was operating correctly. Safety on the Dutch railway network is based on red light discipline. A red signal may not be missed by a driver, because the consequences can be catastrophic. Unfortunately, human errors still occur, and drivers do miss signals at red.

Until 1962, the only safety system on the railway network was the signal beside the track. However, the tragic accident in Harmelen in 1962, changed this situation.

The then Railway Accident Council advised the Minister of Transport and Public Works to equip the Dutch network with an Automatic Train Influencing system (ATB). One key consideration of the Railway Accident Council was that the intensive use of the railway network was no longer commensurate with the safety system employed. This system consisted exclusively of signals alongside the track, which had to be closely monitored by the drivers. The then Minister followed the advice. The ATB system, which (over a period of 35 years) was gradually introduced after 1962, can today be viewed as a safety net, if certain signals are missed by drivers. The designers had placed priority on avoiding train collisions at high speeds. (In this connection, it should be taken into consideration that the level of technology at the time was limited.) In the event of the passing of a signal which calls for a reduction in speed, the ATB automatically initiates rapid braking, if the driver himself does not brake.

The system is subject to at least two major limitations. On the one hand, the system accepts ‘light’ braking by a driver as a suitable reaction. The system does not check whether the braking is sufficient to bring the train to stationary in time. On the other hand, the system does not operate at speeds below 40 km per hour.

The limitations of the existing Automatic Train Influencing system are now starting to cause problems for the railways. Until 1995, on average 150 trains a year passed a red signal. Over the last few years, a considerable and unexpected rise has been noted, reaching 280 occurrences in 2000. These numbers indicate that on average, every driver misses a red signal one to two times, during his career. The passing of a red sig- nal involves considerable risks. A modern double-decker can carry 1000 passengers at speeds of up to 140 km per hour. Above all near stations, situations arise in which trains travelling at low speed, and therefore effectively without ATB monitoring, in pass- ing red signals can enter the route being travelled by a train at high speed.

Safety on the railway is to a considerable degree dependent on the correct interaction between technical facilities (such as the ATB system) and human task implementation. There are a number of indications which suggest that this interaction is not always smooth. The question therefore unavoidably arises, whether the safety management systems of the organisations in question, and above all the coordination between them, is sufficient.

Acrylonitrile Leakage - Amersfoort Station - 20 August 2002

Dutch Safety Board

The transport of hazardous substances engenders unavoidable risks. However, this does not mean that all the risks arising from the current transport of hazardous substances should simply be accepted. Such transport, and the majority of rail transport, should be organised in such a way that only exceptional factors represent any risk. The investigation by the Board for Transport Safety into the accident involving acrylonitrile, which occurred at the Amersfoort station complex on 20 August 2002, has shown that in practice, the transport of hazardous substances is very far from being organised in this way, as yet.

On this day, at Amersfoort Station, a goods train was present, with a destination in Germany, with immediately behind the locomotive a tank wagon filled with more than 70,000 litres of extremely hazardous acrylonitrile. The train had been parked in this location from five o’clock in the morning. Following arrival, the driver had returned home, so that the train was effectively parked in the centre of the city without any monitoring or supervision. At 11:03 hours, the discovery was made that the tank wagon was leaking acrylonitrile. At 11:28 hours, the decision was taken to evacuate an area with a radius of 100 metres. Not until a further 25 minutes had passed was train traffic passing within only a few metres of the tank wagon also halted. The fire brigade called in the regional hazardous substances officer. The municipal crisis team which by this time had been appointed, opted for the safest course of action, on the basis of the limited information available, and at 13:20 hours had the area within a radius of 500 metres around the train shut off.

The major risks accompanying the transport of hazardous substances led the Board for Transport Safety to investigate this accident. The results of the investigation thus initiated both surprised and disturbed the Board. The system of international rules for the transport of hazardous substances by rail: the RID are an impressive example of professionalism, expertise and consistency. Nonetheless, on the basis of this investigation, the Board has reached the conclusion that within these highly valuable and important international agreements, there remain a number of fundamental lacunas. The central theme in this connection is above all the integrated duty of care, a basic term from the ISO 9000 standard approach. The duty of care approach, which demands permanent attention at all levels within a company, from grass roots through to management, goes far further than merely complying with the rules imposed. The safety of the transport of hazardous substances is at present above all based on rules which are assumed to be complied with. Supervision of compliance with these rules and inspections by government represent an essential element in this system, for the maintenance of a responsible level of safety. However, such supervision and such inspections are marginal. The Hazardous Substances Inspection Agency (Korps Controleurs Gevaarlijke Stoffen) responsible solely for inspection of compliance with the rules for hazardous substances, was done away with years ago. Such inspections are now one of the tasks of the officers of the Inspectorate for Transport, Public Works and Water Management (IVW), an organisation that supervises compliance with a large number of laws and regulations.

Within safety management systems, too, the duty of care plays a key role. These aspects have come to occupy an important place within companies, because many experiences have shown that the principle “compliance with rules” offers no guarantee for achieving the level of safety demanded by society today. The duty of care calls for far greater efforts from those involved than merely complying with rules. For a complex system such as the transport of hazardous substances, however, the duty of care is an absolute precondition if a safer system is to be achieved.

Derailments at RandstadRail 2006

Dutch Safety Board

A new public transport system, RandstadRail is a light rail network which connects The Hague, Rotterdam and Zoetermeer. A large scale project involving the construction of new infrastructure, modification of existing infrastructure, purchase of railway vehicles and establishment of a trans- port and management organisation was required for its realisation.

Five derailments occurred in the Haaglanden region within a month of RandstadRail’s operational launch. Following the derailment at the Forepark stop on 29 November 2006, which resulted in injury to 17 passengers, the railway undertakings ceased operations and the Transport and Water Management Inspectorate formally disallowed these to continue by withdrawing the operating per- mit. Four more derailments occurred following the resumption of services on sections of the Rands- tadRail network.

A total of nine RandstadRail derailments occurred. Based on their respective causes, these derail- ments can be divided into four categories.

The derailment of a RandstadRail vehicle belonging to RET on 29 November 2006, which resulted in injury to 17 passengers, took place on a damaged switch close to the Forepark stop in Leidschenveen. The RandstadRail project bureau of the municipality of The Hague had not taken the possibility of damage to the newly installed switches sufficiently into ac- count. These switches remained in use during replacement of the railway section between Zoetermeer and The Hague and sustained damage as a result.
On the same day, a RandstadRail vehicle belonging to HTM derailed in a curve close to the Ternoot stop, which is near The Hague’s central station. HTM was aware of the risk of derail- ment in this curve but had underestimated the likelihood of actual occurrence. The curve proved in reality to have less favourable characteristics than HTM had assumed and, in addition, the minimum speed of 50 km/h recommended by the vehicle manufacturer in connec- tion with the guarantee issued could not always be achieved. The risk of derailment on this particular railway section was greater at lower speeds.
Three derailments occurred on the Muzenviaduct close to The Hague’s central station on a section of worn out railway track, one of which involved a city tram and the other two Rands tadRail vehicles belonging to HTM. The abrasion had been caused by a modification to The Hague’s city trams and the use of rails with a different hardness, causes that had not been recognised in time by HTM.
Five derailments involving RandstadRail vehicles belonging to HTM occurred at a vehicle activated switch in the city tram network. The vehicle activated switch was introduced as part of the RandstadRail project and was of a type new to HTM. The switch is used to change tracks at a railway end point so that the vehicle can subsequently travel the route in the op posite direction. Investigation of the derailments revealed that the drivers of the vehicles had not been able to clearly see whether they had fully passed the switch and had therefore turned too early. Following these derailments, HTM introduced signs and marking at the turning point in question.

Goods train derailment near Apeldoorn on 30 April 2003

Dutch Safety Board

On 30 April 2003, a goods train loaded with steel coils derailed close to Apeldoorn, leading to considerable havoc. The investigation by the Council for Transport Safety has demonstrated that the derailment was caused because the goods train passed through a set of points at a speed of 70 kilometres per hour, whilst a maximum speed of 40 kilometres per hour was permitted. This excessive speed was due to a lack of alertness on the part of the driver, caused by sleepiness. There is even a possibility of ‘microsleep’, a brief period of sleep (up to 30 seconds) followed by a period of ‘sleep-induced sluggishness’ during which it is possible to react, but less rapidly and less accurately. Another factor which probably contributed to the derailment was the insufficient securing of the steel coils as a result of which they were able to shift.

The report before you clearly shows that this derailment cannot be considered as a one-off accident, but that it was due to structural safety shortcomings. The most important safety shortcoming relates to insufficient control of the speed of goods trains. A second safety shortcoming relates to the loading process.

The most important safety shortcoming exposed by the derailment near Apeldoorn is that the speed of goods trains is insufficiently controlled. The crucial safety task, namely the control of speed, is almost entirely left to human intervention. If a driver fails, for whatever reason, there are no technical provisions to step in. Certainly at night, drivers (as is the case with all people) can fail due to sleepiness, because working at those moments conflicts fundamentally with the natural biorhythm. In addition, at night, circumstances are more monotonous which promotes the occurrence of (micro) sleep.
Control of speed consists of two aspects: monitoring maximum speed and slowing the train in time. For both aspects, the current system for Automatic Train Control (ATB) has proved insufficient as a catch net in the event of driver failure, as was the case in Apeldoorn as a result of reduced alertness. Firstly, the ATB system is set merely to monitor the locally-applicable track section speed, and not the maximum speed set for a specific train. This is primarily a problem for goods traffic, because goods trains generally have a maximum speed of 100 kilometres per hour or lower, because of their weight and braking force, whilst many track section speeds have been set at 130 or 140. The ATB then does not respond to a violation of the maximum train speed, but only if the locally- applicable maximum track section speed is exceeded. Secondly, the ATB system is? unable to check whether the initiated braking is sufficient to reduce the speed to the level indicated by the signal (in this case) 40 kilometres per hour (the so-called ‘braking curve monitoring’). If these monitoring possibilities had been included in the design of the ATB system, the speed violation in Apeldoorn would have been detected at an earlier stage, and the derailment could have been prevented.

The derailments at Amsterdam Central Station on 6 and 10 June 2005

Dutch Safety Board

On 6 June, 10 June and 15 August 2005, trains derailed on the western emplacement of Amsterdam Central Station. Initially, the Board did not intend to carry out a separate investigation into the derailments which occurred on 6 and 10 June but wanted to wait until it had received the relevant reports from the Inspectorate for Transport, Public Works and Water Management. However, after the third derailment on 15 August, the Safety Board decided to start its own investigation into these accidents because it was aware of the degree of anxiety that had been caused due to three derailments taking place in a short space of time on one emplacement. The main issue on which the investigation focused was therefore the extent to which a common cause could be found for the three derailments in the local infrastructure.

The Safety Board concluded that no common cause for the three derailments could be found which was related to the emplacement at Amsterdam Central Station. The cause of the first derailment was a defective wheel on a goods wagon. The wagon could have derailed at an entirely different location. The second derailment was caused by a set of points which had been damaged as a consequence of deficient repairs to the infrastructure after the first derailment. However, these repair-related deficiencies could also have occurred at a different emplacement. Therefore, a direct link exists between the first and second derailments, but this link has nothing to do with the safety of the railway emplacement concerned. The third incident, the derailment of a shunted passenger train on 15 August, is the most complex of the three, with a great many factors playing a role . However, the idea that deficiencies in the railway emplacement were the common cause of the three derailments can be rejected.

Even so, because structural safety defects were the reason for the separate derailments, the Safety Board did decide to issue reports on the derailments. This report focuses on the first two derailments (6 and 10 June 2005). The report on the third derailment is to follow in due course because the complexity of that incident means more time is needed for the investigation.

Through Red at Amsterdam Central - 21 May 2004

Dutch Safety Board

On 21 May 2004, in the eastern complex at Amsterdam Central station, an empty double decker train collided with an intercity passenger train (Heerlen – Haarlem). The immediate cause of the collision was that the empty double decker train, which was being transferred from the station to a stabling zone, had passed a red signal, shortly before the collision. As a result of the collision, 19 passengers from the intercity train were so severely injured that they had to be transferred to hospital. The damage to the track and the trains was massive.

On 27 May 2004, in a letter to the Lower Chamber of the Dutch Parliament, the Council for Transport Safety stated that "in the case of such a serious accident, the Council will always launch and investigation into the causes". However, if it emerges that this collision was once again due to an STS passage, the question emerges as to whether a fundamental investigation into the underlying causes is still necessary. These underlying causes have been described in detail in investigation reports for comparable train collisions as a result of the passing of stop signals (Eindhoven Railways Accident Board 1992, Dordrecht Council for Transport Safety 1999).

Following the train accidents in Dordrecht and Eindhoven, the Council issued recommendations to the parties involved to take measures to reduce the number of red light passages, the so-called Stop Signal (STS) passages. In these recommendations, reference was made to the sharp rise in the number of STS passages. Since 1995, the number of STS passages has been rising, year on year. Until 1995, this number had averaged 150 a year. In 1999, the number had risen to 230 a year. Since that time, the rise has continued unabated. The number of identified STS cases in 2004 totalled 283. As a result of passing an STS, in the period since 1999 (since the collision in Dordrecht), to date, 14 serious accidents have occurred. To date, after passing a stop signal, there has been one fatality, and very considerable material damage. However, STS passages engender such a serious potential risk that large numbers of fatalities are a clear possibility.

The recommendations of the Council following the train accidents in Dordrecht and Eindhoven referred to three aspects, which turn out once again to play an important role in the investigation into "Amsterdam".

A first aspect on which the Council and Railways Accident Board have issued recommendations is the railway control system. In the investigations into the accidents in Eindhoven and Dordrecht, it was determined that the existing control system, the so-called ATB system First Generation, installed on a large proportion of the Dutch railway network, is outdated and offers insufficient protection. Following the serious railway accident in Harmelen (1962), it was decided to introduce this ATB system in the Netherlands, which had been developed in the 1950s. The ATB system however suffers from two major limitations. The ATB system (i) does not monitor whether sufficient braking has been undertaken, and (ii) does not operate at speeds below 40 km/hour. To eradicate these limitations, an ATB system New Generation has been developed, equipped with continuous automatic speed control, which in normal circumstances makes it impossible to pass a stop signal. On those track sections in the Netherlands which are equipped with an ATB system New Generation, not a single STS passage has been identified. One of the recommendations arising from "Dordrecht" to the Minister was to lay down the phasing for the implementation of such a modern control system on the Dutch railways. In response to the recommendations from Council, the Minister suggested (on 28 April 2003) ".. it can be concluded from this that the replacement will not take place only on the basis of safety considerations .." and that "we cannot issue any concrete plans".

A second aspect which was investigated at both "Eindhoven" and "Dordrecht" is the influence of the setting of subroutes, on the passing of stop signals. This aspect refers to the setting of part of the overall route from platform to open track. Because being offered a relatively short route often does not tie in with the expectation pattern of the driver, in the Dordrecht report, the Council recommended reticence in setting such subroutes, as had been the case before 1995. The accident in Amsterdam also involved a subroute.

A final aspect relates to the 'red light discipline' of drivers. In the investigations into "Eindhoven" and "Dordrecht", it was specified that strict compliance with signals at red by drivers has since the birth of the railways represented a key factor in the safety of railway traffic. Because to a considerable extent safety on the railways is determined by this 'red light discipline' of drivers, it was recommended that 'red light discipline' should be given a central focus in the railway sector. It was recommended that the way in which this discipline be dealt with should be improved.

Goods train derailment Amsterdam-Muiderpoort, 22 November 2008

Dutch Safety Board

On November 2008 a goods train derailed at Amsterdam-Muiderpoort station. The train, comprised of an electric locomotive and twenty-five wagons loaded with chalk/quicklime, was on its way from Belgium to the steelworks in Beverwijk, the Netherlands. The train derailed at the moment that it was passing the Amsterdam-Muiderpoort emplacement. In the first instance the derailment was restricted to the front wheel set of the eleventh wagon. However, an escalation occurred about five hundred metres further on, when the other three wheel sets of this wagon derailed1. This occurred as the wagon passed over a set of points. The eight following wagons then also derailed. Some of the derailed wagons came to rest on the adjacent track and four of the wagons fell on their side.

Although there were no casualties, a large amount of damage was caused to the rail infrastructure and the derailed wagons. The direct financial damage to the infrastructure and rolling stock amounted to almost three million euros. The damage to the rail infrastructure caused an extensive and lengthy disruption of rail traffic in the Randstad conurbation that resulted in process damage of about 2 million euros2. Consequently, the total financial damage caused by the derailment amounted to almost five million euros.

At the time of the derailment the adjacent tracks were being used by passenger trains. One passenger train passed the location of the accident shortly before the derailment: two other passenger trains had approached to a relatively short distance from the derailment but were able to stop in time.

The Board, on the basis of the damage and the marks, concluded that the derailment was caused by an overheating axle box on the eleventh wagon3 that in turn caused an axle journal to break off. The overheating of the axle box was caused by the seizure of one of the two bearings in the axle box. The Board was unable to reach a definitive conclusion on the cause of the seizure of the bearing due to the damage caused by the overheating. However, it is clear that the bearing cage failed at an early stage of the seizure process. The nature of the damage also excludes a number of potential causes (such as an assembly error, lack of lubrication and overloading).