Author: petervandermark

I am a former driver and active Transport Museology Consultant and Writer. I have an avid interest in anything rail safety related and have published the book "An unexpected end to the Journey" and am in the process of publishing a book on the history and predecessors of the Dutch 1200 class loco.

De geschiedenis van de 1200

Het gekke is dat het plan was om zo’n 75 stuks 1200 aan te schaffen, maar dat het er uiteindelijk maar 25 werden. Er blijken daar een aantal redenen voor te zijn die voor een deel technisch en een ander deel financieel-politiek zijn. Overigens, de NISM type WH 1’Bo’Bo’1’ e-loks uit 1924 voor de electrische dienst op Java waren wel degelijk een reden voor de Nederlandse Spoorwegen om bij Baldwin en Westinghouse te gaan praten. Een andere goede reden was dat op die wijze, door licentiebouw in Nederland, de Marshall fondsen grotendeels in Nederland besteed zouden worden. Het originele type loks uit de USA zouden overigens ook 3,000 pk 1’Bo’Bo’1’ machines zijn geweest voor f 700,000.- per stuk. Technisch gesproken vreemd, omdat de US industrie normaal geen hogere vermogens als 500 pk per tractiemotor inbouwde. Er werd daar, vanwege de gewoonlijk veel hogere aslasten, zonder wielslipbeveiliging gewerkt. Dit zou, ondanks de relatief lage vermogens per as, het leven op de 1200den voor machinisten bij vlagen moeilijk maken tot er in 1959 alsnog wielslipbeveiliging voor bij het aanzetten werd toegepast. Niettemin zag ik gedurende mijn dagelijks reizen tussen Arnhem en Utrecht gedurende de jaren zeventig en tussen Amsterdam en Utrecht gedurende de jaren tachtig in Utrecht bij slecht weer regelmatig een Zandvoort naar Limburg trein met een 1200 met doorslaande voorassen in de draaistellen vertrekken.

De vreemde situatie doet zich voor dat de industrieel en technisch tamelijk meer gestandaardiseerde 1200 machines op een prijs van f 800,000.- per stuk uitkwamen. We spreken bovendien over wat uiteindelijk in US termen maar een halve lokomotief was omdat alleen het gelijkstroom tractiegedeelte met de weerstandsinstallatie werd ingebouwd; de transformator en de gelijkrichters vielen weg. Ik denk dat de reden voor die hogere prijs van politieke aard is. Nog voordat de onderhandelingen met Baldwin en Westinghouse rond waren bestelde NS bij Alsthom de hele serie 1100. De reden hiervoor was dat er een overeenkomst met Frankrijk was dat zij in Nederland binnenschepen lieten bouwen en dat Nederland als tegenprestatie spullen in Frankrijk bestelde. Hiermee was echter wel 2/3rde van de markt voor electrische locomotieftractie verdwenen en werd voor de Baldwin Westinghouse Werkspoor Heemaf combine het hele project veel minder interessant, dus vandaar die hogere prijs. Het wordt openlijk beschreven dat Werkspoor en Heemaf gewoon behoorlijk kwaad waren. De 22/2300den vielen overigens ook onder deze regeling, die werden voor een behoorlijk deel bij Schneider, Jeumont en Creusot in de Elzas besteld.

Een andere reden voor de bestellingen in Frankrijk, echter, is naar alle waarschijnlijkheid dat de tractiemijnheren bij NS al snel door hadden dat de Baldwinmachines er dan wel leuk uitzagen maar dat ze in termen van werkvermogen nogal aan de lage kant zaten. Met 500 pk per as zaten ze ver onder de 645 pk per as van de 1100, die veel van het 1200 werk dan ook zonder te veel moeite kon overnemen, en de 1300 die met een gelijk vermogen per as als de 1100 voor zover het de zware goederendienst betrof tot veel meer in staat was. Daar waren twee redenen voor: a) beide Franse types hadden wielslipbeveiliging en b) hun draaistellen, zeker die van de 1300en, waren van een veel geavanceerdere constructie op het gebied van het tegengaan van weight-transfer effecten. De Franse CC 7100 machines waarvan de 1300 werd afgeleid zaten op een vermogen van zo’n 4700 pk, terwijl in Duitsland de bekende DB electrische standaardlocomotieven voor wat betreft de zwaardere Bo’Bo’ types (E10 en E40) al op 1250 pk per tractiemotor terechtkwamen en daar geen problemen mee gaven. Aan de andere kant; doordat de Baldwinmachines technisch nooit overbelast werden waren ze buitengewoon betrouwbaar. Dat wel.

Those Euro-American electric locomotives

January 2017

Being an enthusiast of the slightly manic variety, forever on the prowl for more books and more information, fate worked its wonders once again and about a week ago I found a 1980’s book about electric locomotives of the world for 1 pound Sterling in a charity (thrift) shop. Not in faraway London town, where you may kind of expect this thing to occur, but here in Winscombe of all places. More precisely, in the shop where Lyn works her Friday-afternoons as a volunteer. To some among us this kind of thing happens! It’s happened before (ask me how I obtained John H. White’s marvellous The American Railroad Passenger Car, both volumes, for only 45 Euro!) and hopefully it will happen again.

1) I found further info about the Pennsylvania Railroad Ep-3’s. The picture above shows it clearly: next to Tri-Bo’ axle configurations they came as Co’Co’s as well. Welcome the real father of the Dutch class 1200 Co’Co’s. It simply is the case that heavy-haul machines, especially for winding track, were in the Baldwin/Westinghouse catalogues as Tri-Bo’s (the Spanish machines), whilst machines intended for higher speed came as Co’Co’s (the Dutch machines) because the ride and behaviour on the track of the 3-axle bogies/trucks was much more stable.

If you want proof of that issue then try and have a look at pictures of how French Bo’Bo’s left the track between Lamothe and Morcenx after the 1950’s 180 mph speed trials in comparison to the Co’Co’s. It was then found that A) 1.5 kV DC was not the sort of traction current you should use for that sort of job as an eternal flame on the contact point between pantograph and wires was lighted, it was in fact the moment that thoughts on French railways turned to high-tension 25 kV 50 Hz AC with well-known results. B) That the contemporary 2-axle bogies/trucks left the track in a state ready for thorough reconstruction. It also was in fact the reason why early in the 1900’s the bogies of passenger cars were extended to 3-axle units (see the perfect illustration of a 1900’s US trimount under a British Pullman camping coach vehicle on the beach at Marazion, Cornwall, before rescue and retirement to a museum). In fact, it was only in the 1960’s that we got bad bogie behaviour under control with suitable motion-dampers between the vehicle body and the bogie-frame. The Dutch firm of Koni was involved in those developments before moving out to the USA.

2) What kept me wondering about the Westinghouse/Netherlands Railways link after the war is: how it was established. Obviously there was the post-war “close co-operation” that is described when Dutch signalling engineers went after automatic colour light signalling (and later Automatic train Protection) and ended up with General Railway Signalling kit from Rochester, NY. But that really is quite a few miles away from deciding about obtaining locomotives. Especially since by then the French 1.5 kV products from what then was Alsthom (Alsace Thomson) were proven machines with an excellent track record, even if the Bo’Bo’s were certified bastards to ride at speed (see remarks made above).

Well, the book mentions that the Netherlands East-Indies Railway Company ordered 6 WH type box cab 1’Bo’Bo’1’ electric locomotives for the 3 ft 6 inch Cape gauge from Werkspoor in Amsterdam (who would build and assemble the 1200 series Co’Co’s after the war), with electric kit from Westinghouse and Heemaf in Hengelo NL (who would deliver and assemble the traction kit for the Co’Co’s) as early as 1924. There simply was an established relationship!

I marvel at the fact, incidentally, that US rail buffs, seeking bits of their rail history, have to come to Europe to see it in the flesh. Why is it that items of such contemporary historical value were scrapped so assiduously over there? Incidentally, the book mentions that in the 1980’s in Brazil there still were five of the 1946 built General Electric 2’Do’Do’2’ “Little Joe’s”, re-gauged for the 5 ft 3 ins, in operation. What is left of them? I mean, there is some really fascinating interworking of technical matters and world politics mentioned here. The Joe bit is Josef Stalin: need I say more?

3), Where European and American traction matters parted company is when the Swiss commissioned their celebrated high-power Ae6/6 Co’Co’ electrics. They had found a way to beat weight transfer issues by bringing bolster and secondary suspension down to the underside of the bogie. This bogie was subsequently also used in Sweden, Norway and Belgium and probably I missed a few others out somewhere. See the two green machines, the last ones only very recently were taken out of use as three axle bogies are hard on curving track. It was after all in Switzerland that the word “Schienenfresser” (track munchers) for three-axle bogies was coined. But 6,000 horse power in the early 1950’s; some achievement. One can be seen in that beautiful Verkehrshaus der Schweiz in Luzern. Lovely location too.

US/European electrification links

December 2016

In the US there really were only four major users of electric traction. That were the Great Northern Railway for the line through the Cascade Mountain at Steven’s Pass, The Pennsylvania Railroad around New York, Washington DC and stretches inland in-between, The New York, New Haven & Hartford between New York and Hartford CT and the Virginian Railroad in West Virginia, a big coal-haulier. Most electric traction indeed came from the Westinghouse stables, with Baldwin from Philadelphia PA as the main supplier of the non-electric technical bits. General Electric similarly was a supplier of electrical traction systems, notably to the New Haven.

If the loco’s ran on AC-wired systems, rectifying the line voltage to DC and then feeding that as normal through resistances (resistors) to the series wound DC traction motors was the done thing all along. The problem mainly was that when such a series wound DC motor lost adhesion on the track and it was not noticed the motor would go into an uncontrolled spin and eventually self-destruct. For that reason no more than about 500 hp per motor was allowed, also because the effects of weight-transfer in the bogies, unloading the front axle when accelerating, was far from being resolved. Most US traction discussed here used the mercury-arc “ignitron” rectifiers, a somewhat iffy method using enclosed mercury jets to rectify the AC to DC. On the bouncing, rolling and shaking rail vehicles a somewhat contentious way to deal with this process and problems, among others an extra source of heat next to the in those days massive transformers to get the 11 to 20 kV AC down to 1 kV DC suitable for traction. This notably was the downfall of the New Haven EP-5 “Jets”, who experienced several fires due to the closely packed kit on board and who derived their nickname from the fact that they sounded like the 50’s and 60’s jets when pulling away. In our days the Eurostar power cars perhaps may have warranted a similar nickname, especially when using their rheostatic dynamic brakes without being able to get rid of the generated juice into the supply grid as on the English Southern. Working at Stewart’s Lane depot in London you certainly knew when a Eurostar set passed by even when you couldn’t see it.

In any case, the situation improved mightily when AC rectification moved on to using silicon solid-state rectifiers. Unfortunately the US by then had decided to go for the in its initial costs far cheaper diesel option. The railways in Europe, in the sixties starting to look for high-speed application of electric traction with the accompanying need for high-tension traction power from the wires above, explored the AC and DC options further, ditched resistance DC and tap-changer AC traction control altogether and moved on to GTO chopper control initially and then to IGBT AC and –chopper DC control. At the moment a serious traction installation is fed with 50 or 60 Hz AC, that is transformed to 3 kV AC, rectified to 3 kV DC and then inverted to variable frequency AC to feed the brushless induction traction motors. The clever bit is the variable frequency here; those traction motors follow the frequency that is being fed to them. Vary the frequency, vary the speed. It isn’t quite as simple as said here but the principle holds. Because already in the initial power phase an inherent wheelslip control is being exerted on the motor, added to which very precise further wheelslip control has been added, 2000 hp per motor is the norm. Hence those 80 tonne European electric loco’s in pairs that happily drag 5,400 tonne freights in-between the mad timetabled passenger trains in the large conurbations. Or run at 200 km/h (125 mph) with a passenger train on their next run.

To see where the USA got to in their last stage of serious electric rail operations, which was the period from which the Dutch 1200s and the Spanish 278s derive, a few pictures of what was to be seen in those days. 1 is the famous Art Deco 2’Co’Co’2’ Pennsy GG1, the Raymond Loewy (among much else the Coke bottle) styled iconic US electric. Number 2 is the much more modern Co’Co’ New Haven EP-5 “Jet”. Its influence on the styling of the Dutch machines is noticeable.

Baldwin-Westinghouse electric locomotives

December 2016

Recently I got fired up to chase the family tree of the Netherlands Railways class 1200 Co’Co’ locomotives. Machines that I erroneously thought were a sort of one-off, the USA assisting the Dutch in the period just after WWII under the Marshall agreements. The Dutch with their 3/4ths destroyed railways that they so badly needed to get the country moving again. The time of the British loaning an LNER Manchester-Wath 1.5 kV dc Bo’Bo’ locomotive “Tommy” and the French loaning a number of their pre-war dc-electrics, before permanent replacements for lost electric traction started to be commissioned. Those Baldwin designed and part US-fabricated 1200 class locomotives held out till 1998 with NS Netherlands Railways and occasionally a few are still about. They were very reliable, liked by the crews and good machines to look at, a thrill when they come past at speed whilst you’re waiting in a car at the closed barriers of a level crossing, or stand on a platform in a station. The latter especially at the seaside resort of Zandvoort aan Zee, as the view of the big US machines in that terminal station in the dunes was mingled with the pleasure of watching Amsterdam cooling itself in the North Sea, dressed from very scantily to not at all. One 1200, incidentally, is on display in the Utrecht railway museum.

I discovered that I was wrong on two counts, however. In Spain there were locomotives from the same stables, purchased under the same Marshall-plan facilities to get their railways functioning again and put together in very much the same fashion. So the 1200s were not the only US electric locomotives exported to Europe and they were not unique loners but part of a family. Whilst in Barcelona I learned about the big museum in Vilanova I le Geltrú and its residents. Lo and behold, there was one of the class 278 Bo’Bo’Bo’ machines, locally named “Panchagro’s”. I thought then that the Dutch machine was the standard and that the Spanish machine was a clever adaptation to heavy mountain hauling. I was wrong on that score too, the Tri-Bo’s were the standard. However, nothing remains of the endeavour in real life in the Eastern USA, all were scrapped with the demise of Baldwin, of Pennsylvania Railroad and its being merged with the New York Central Railroad into Penn-Central. You received pictures of both types.

Recently I discovered on a few websites about the Pennsy E-3C’s, single cab electric ac machines of the Tri-Bo’ arrangement. Baldwin and Westinghouse at the time were trying to chip out a corner in the market for electric traction and a six-axle ac electric with ignitron rectifiers for the East Coast ac networks of the New Haven and the Pennsylvania Railroads was what these machines were meant for. Two of the machines back to back among the four of them or working with 4-axle variants was the idea, it did not get picked up big time. Fuel was cheap, diesels a fiercely contested market where the purchase prices were low (hence the glut of GM-EMD technology and even complete diesels making their mark in Europe at the time, the ALCO-PA Co’Co’s in Spain are part of that drive). Furthermore, the Pennsy was too happy with their legendary GG1 2’Co’Co’2’ machines that seemed happy on any job. It didn’t work out for Baldwin-Lima-Hamilton and the manufacturer folded quite soon after the delivery of the straight dc loco’s to Europe. Other US machines with a sideways connection to this story are the not altogether successful New Haven EP-5 “Jets”, Co’Co’ machines the design of which provided the template for the 1200’s in the Netherlands

Above a picture of Pennsy 4995. A piece of most interesting history long scrapped, to see US electrics of the period you’ll have to come to Europe. To see the GG1’s there are various locations in the US where you can see them. Incidentally, have a look at details such as that ladder to the roof just behind the entrance door. The protection lid coming from off the roof to prevent anyone coming under the juice and being exposed to live bus-bars on the roof. No doubt a key of the sort used to connect the train to hotel power and get the pantographs up here was necessary to remove that lid. Take that key out of the dashboard and the pan comes down. Also have a look at the registration of the wire above the track. High speed of any kind well out of the question there.

Een visite aan een Spaans spoorwegmuseum

December 2016

Lang dacht ik dat de NS 1200 locomotieven de enige na-oorlogse US export van electrische locomotieven naar Europa waren. Dat bleek echter niet waar te zijn, de Spaanse spoorwegen (RENFE) waren (rijkelijk late) ontvangers van deze US goodwill. Net als de Nederlandse machines werden zij ontworpen en gedeeltelijk gebouwd door Baldwin/Lima/Hamilton, met een door General Electric ontworpen gelijkstroom tractie installatie. In beide gevallen was lokale industrie bij de bouw betrokken en werden de locomotieven ter plaatse als bouwpakketten in elkaar gezet.

De Spaanse locomotieven waren Bo’Bo’Bo’ 3 kV breedspoormachines, dus net als de NS machines met zes tractiemotoren. Deze asindeling was een logisch gevolg van het feit dat deze machines voor bochtige bergtrajecten bedoeld waren, waar zesassers op twee draaistellen als de NS 1200 machines de neiging hebben om het spoor nogal hard aan te pakken. Dat het bovendien om breedspoor gaat maakt het allemaal nog weer ernstiger.

Bovenstaand vier foto’s, twee keer NS 1200 uit 1954 en twee keer RENFE 278 uit 1963.

ZS1 Signal example

Burgsteinfurt, Germany. Shown are the Westbound home signal (Einfahrsignal) and the distant signal for the Westbound platform starting signal (Ausfahrsignal, with the possibility to show which of the two platform starters (main line or passing loop) will be encountered. This is indicated by the possibility to show the different entry/passing speed limits. Above the yellow disc of the distant signal can be seen the small black triangle of the Zs-1 Ersatzsignal. This is the pass on sight aspect that played a main role in a series of accidents.

Staff responsibility in case of rail accidents gets further attention

July 2016

Referring to a previous article in European Railway Review about responsibility attributable to key staff in case of an accident, an interesting development occurred last October when the European Railway Agency (ERA) expressed concern about the CIAF accident report that deals with the 24-07-2013 high-speed Santiago de Compostela derailment in Spain and asked for a review by an independent body of expertise. The ERA noted that the independence from vested interests in the official Spanish accident report appears questionable, given that against European ruling staff from infrastructure provider ADIF, train operator RENFE, and consulting transport engineers Ineco were involved in the investigation by the official transport accident investigation body CIAF. As a result the CIAF investigation and report do not comply with the requirements of European law as expressed in articles 19, 21.1, 21.2 and 22.3 of the Railway Safety Directive. The accident report put the full responsibility for the occurrence on the driver, despite the failure to intervene by an Automatic Train Protection system. As the driver could have no responsibility for this ATP issue it had to rest somewhere else, which inadvertently drew attention to potentially structural safety flaws elsewhere on the Spanish high-speed rail network. Which had to be addressed urgently in the weeks after the accident by fitting series of ASFA Digital speed traps.

According to the ERA the report does not satisfactorily explain the roles that the infrastructure, signalling issues and the type of train used played in the run-up to the accident. This accident highlights inherent lack of essential safety measures, as on a modern high-speed railway line the driver’s  errors should have been warded off through intervention by an Automatic Train Protection system before spinning out of control and causing a crash. These underlying causes, pointing at responsibilities lodged with Ineco, ADIF and RENFE, are not explained to the extent that relevant conclusions can be drawn and measures for improved safety can be proposed. This in turn invalidates judgment on responsibilities for the accident and disables proposing relevant measures to prevent a similar accident from happening again. Confirming receipt on the 8thof July last, EU Transport Commissioner Violeta Bulc said that transparency is the key and that the Spanish authorities have taken steps in the right direction. 

This crash no doubt is widely remembered as a platform video camera recorded it and the video was seen by millions on television and the Internet. What happened was the following: ETCS level 2 ATP was switched off on the class 730 electro-diesel set as, after the conversion from class 130 electric sets, ETCS problems occurred that were yet to be dealt with. As a result this high-speed trainset came off the high speed section at 200 km/h travelling toward a 80 km/h permanent speed restriction in a sharp curve only under protection of the ASFA classic warning system, a system that advises about permitted speed but does not intervene. It is not an Automatic Train Protection system suitable for high-speed use, as for instance ETCS or a later ASFA version called ASFA Digital are. There also were no distant speed warning boards and the ASFA warning about the speed restriction ahead came when the driver was distracted by a phone call from his senior conductor. A “leader driver” in fact commented on potential risks emanating from this situation before the accident occurred, but no initiatives to improve on it were taken. With a decent risk analysis this lay-out would not have been considered safe as no ATP safety net was available at a location where an obvious and potentially fatal train overspeed risk existed.

Level Crossings and the Lorries of Doom; instalment 2

May 2016

Perhaps ERR readers remember my previous article about the remarkable regularity with which collisions between heavy road transport and trains occur on level crossings. A reader from The Netherlands pointed out that in fact it is quite a job to find the right emergency telephone number to call in most countries and that even in case they do get through to rail traffic control, structural possibilities to find and stop approaching trains comfortably turn out to be limited. The usual train describers do not give that accurate a view of just where trains are along plain line, certainly away from the main rail traffic corridors or in ‘black territory’, and many non-controlled automatic signals do not actually have what is called ‘aspect replacement to red’ facilities. In case used in a stress situation, such signal aspect replacement facilities also might potentially trigger a category C SPAD, which is not something one would like to inflict too regularly on train drivers. This obviously is an international view. For instance in the UK aspect replacement to stop is often included in plain line signalling, even if in cases with a key used at location. Not a very usable proposition in this case either. An official placing track circuit clips is another alternative with questionable merits.

There are in fact two key issues at play. One is a contracted heavy transport that is being planned and that, in order to obtain the necessary permits and papers, is being scouted for the best route from origin to destination in the weeks prior to departure. Obviously there should be obligatorycontact between the road transport organiser and the rail traffic authorities, which surprisingly is not a legal requirement in all countries. The NTSB in Washington DC, USA, in fact did get this issue included on a Federal level in the run-up to legalising a heavy transport, but found that level crossing crashes occurred nevertheless and that people involved (including police escorts) afterwards still professed to be unaware of what was required. In Europe this clearly will have to be organised at Continental level, as nationally restricted activities are bound to lead to irritating discrepancies and uncertainty, especially when such a transport is crossing borders. Furthermore, once on their way the transport organisers and escorting authorities (private security or police) must have documentation in their vehicles that spells out the permitted route and contains the necessary telephone numbers to report progress and arrival at level crossings or, indeed, other possible intersections such as drawbridges across waterways and cross-town and city-centre routes. Reading this, logically speaking a kind of one-stop shop from the side of the railway should be provided, to contact and negotiate the rail crossing occupations. One actually wonders how come this internationally isn’t so yet. Is that because blaming road transporters is easier?

The other situation resembles a kind of spur of the moment issue, or an emergency, more than anything. It is what played a role during e.g. the recent fatal level crossing collision at Dalfsen in The Netherlands, where a slow tracked vehicle crossed a single line on a user worked crossing when an Electric Multiple Unit train smashed into it at speed, killing the train driver and completely derailing away from the track. In such a case the contractor working the vehicle should be able to contact rail traffic control from the crossing site using a mobile telephone. Experience on a number of networks has it that lineside telephones, whether at signals or at crossings, lead to vandal abuse, but Finnish experiences with remote level crossing accidents led to installing signs that show the name and location of the level crossing, together with the telephone number of the local traffic controller to report accidents. That, exactly, is what should internationally become an obligatory requirement. Also used to remind anyone with an awkward transport, or getting stuck, to immediately call the rail traffic controller. Again, in the day and age of the mobile telephone it should not pose intractable problems, provided that the sign caters for foreign language speakers through being internationally standardised (local language and the transport lingua franca English, for instance) and placed high enough to escape the attention of an average infantile with spray-paint. Nationally such signs exist; I have seen and photographed them in France and in the UK.

Part 2 of the argument is relevant, objective-referenced training for anyone involved professionally in road freight transport. In many international accident reports dealing with such heavy road transport level crossing crashes, the utter and complete lack of understanding in road freight operators of just what an automatic level crossing, or indeed a completely unguarded one, might mean in terms of accident risk to the precious cargo (and the safety of employees and passengers on trains!) is actually quite stunning. It is what could be seen in action during the level crossing crash at Studenká in the Czech Republic not so long ago. And to mention but one regularly re-occurring issue: It isn’t good practice to see whether your vehicles and their cargo are going to ground, or whether your heavy cargo is clearing the electric overhead wires, when you arrive at a level crossing. This is something the transport contractor should have found out way before leaving and should have prepared his trip accordingly, e.g. through negotiating assistance from the railway for a safe crossing. Especially BEA-TT, the French transport safety authorities, were very clear on this point after the bad accident at Tossiat near Bourg-en-Bresse. Obviously, another issue is making accidents like the described very expensive in the courts of justice when lacking preparation is evident. This helps to focus priorities as well.

Shouting from the moral high ground is not what is going to get this problem sorted out. Being well-introduced to the problem and conversant with four European languages, I hereby offer to assist with any attempt to prepare international proposals (EU-wide) for improvement on this life-threatening and costly problem, aimed at urgent implementation on European level crossings. Reactions please!

The misuse of an associated pass on sight (POSA) signal aspect

May 2016

The misuse of an associated pass on sight (POSA) signal aspect to enable passing of failed main signals without oral communication between signaller and train driver.

Misuse of signal aspects provided for the purpose of keeping traffic moving when a signalling system failure has occurred can lead to very serious consequences, as was illustrated early in 2016 by a fatal accident in Germany. The accident was determined to be the direct result of misuse of such an aspect  by the Bad Aibling signaller  and the issue is of significance and merits early consideration when developing similar systems and when writing or reviewing the rules surrounding their use. 

The accident.

On the 9thof February 2016 at around 06:47, a 174 tonne ET325 six-car and a 111 tonne ET355 three-car Stadler Flirt3 electric multiple unit (EMU), very modern EMU sets built to crash norm DIN/EN 15227 and operated by the French owned Bayrische Oberland Bahn (BOB, trading as Meridian), collided head-on with an impact speed of around 150 km/h (90 mph) near Bad Aibling in Germany.  Investigations found  that the signaller had been at fault due to diverting his attention to playing an internet game on his mobile telephone at least until very shortly before impact. Soon the expression “schlamperei” (sloppiness) appeared in several German media comments.  indicating irritation with bad job-related discipline on the railways that once again went undetected until an accident occurred. This issue concerned use of the Zs-1 Ersatzsignal (Pass on Sight Aspect), presumably to assist the 4 minute late-running Southbound service with timekeeping by sending it into a single-line section that in fact was already occupied by another service. Not only was it found in the Bad Aibling Zs-1 recorder that this had been done two months as well as six days before this accident,  but there have been three well-documented previous fatal accidents in which errors with the Zs-1 signal aspect played a role: Berlin Wannsee on 09-05-1993 (signaller at fault), Brühl near Cologne on 06-02-2000 (non-compliant use of the Zs-1 aspect, overspeed derailment, driver blamed) and Schrozberg in Bavaria on 11-06-2003 (signaller at fault). No other European network that uses Pass on Sight Aspects (to keep trains rolling with a modicum of expedience in case of signal failures), i.e. France (“oeilleton” signal), Austria and former Austrian / Hungarian railway operations, Switzerland and Britain, lists such a repeated series of fatal accidents. In Britain such incidents go back to the erroneous use of the interlock release key (e.g. Battersea Park in 1937 and South Croydon in 1947) and predates current signalling systems.

This article in no way tries to pre-empt the official inquiry in to the recent accident or to make recommendations regarding that accident, but attempts to expose the wider issues arising from misuse of such special signalling arrangements and the potential for consequential errors.

The railway line.

The accident occurred near the Bad Aibling Kurpark platform halt in a curve with a 100 km/h (60 mph) speed restriction (from 120 km/h – 75 mph permitted maximum speed) along the river, where dense lineside vegetation hinders forward view, along the Mangfalltalstrecke. This is a 37 km (23.10 mile) long single-track electrified railway along the river Mangfall between Holzkirchen and Rosenheim in Bavaria. The line has five double-track crossing stations that are located approximately 5 kilometres (3 miles) apart and from Heufeld via Bad Aibling to Kolbermoor is controlled by the signaller at Bad Aibling station from a 1970’s SpDrS60 push-button electric relay NX panel. The line is no quiet backwater; in times of disruption on the electrified double-track (Innsbrück – Kufstein, Austria) – Rosenheim – Munich main line this is the primary diversionary route. Furthermore, booked as well as diverted freight services regularly use this line.

The trains involved.

As booked the two services involved were to cross at Kolbermoor station. The westbound service M79506 from Rosenheim to Holzkirchen entered the Kolbermoor station loop on time and was booked to wait 5 minutes for the opposite service to arrive, but it left on time on a proceed aspect. The eastbound service M79505, from Munich via Holzkirchen to Rosenheim and due to come into Kolbermoor from Bad Aibling, ran 4 minutes late. The signaller stated that he noticed that the signalling from Bad Aibling to Kolbermoor did not accept his input for M79505. He expected what he called a “Phantomstörung”, a spurious fault, and then used the ZS-1 facility to override the PZB90 train protection to get the train going to adhere to the booked meet at Kolbermoor. Somehow he expected the on time westbound service M79506 to wait for the booked crossingdespite having given it the road to Bad Aibling which therefore disabled his choice to set the road for M79505. The Zs-1 “Ersatzsignal” as defined in the German railway signalling handbook, is a small triangle of steady white lights under the main signal, meant to allow a train to pass a failed main signal without the need for oral contact between signaller and driver. M79506, having already passed its signal, now caused that driver to be unaware of any issue with the route ahead. The delayed driver on M99505 at Bad Aibling adhered to regulations on departing on a ZS-1 aspect by passing the PZB ATP magnet at the signal at 40 km/h (25 mph) till clear of the single-line turnout and from there he accelerated, still according to the rules, to 100 km/h (60 mph). The signaller, realising his first mistake, then made a second mistake by sending out an emergency stop message on the GSM-R train radio, but in his stressed confusion used a wrong call destination field on his GSM-R computer screen and sent the emergency message to station staff along the line. Shortly thereafter the mistake was noticed and a second successful attempt was made , by now too late. There initially were 11 fatalities, whilst 85 people were injured of whom 24 severely, but two months later one of the injured passed away in hospital. 

The signal aspect.

The Zs-1 Ersatzsignal is a small white triangle of steady white lights under a main aspect. The rules require that s after checking that the line ahead is clear the aspect may only be used when the associated main signal cannot show a proceed aspect  because there is: 

1) A defective signal semaphore or light bulb,

 2) A defective turnout detection,  but only after the turnout has been secured in the proper position,

 3) A defect in the signalling block system, e.g. a cable fault, and it has been established that the line is clear,

 4) A proven clear track circuit that persistently keeps showing occupied,

 5) No normal possibility to set up a signal-protected route for a necessary special move that is protected according to the relevant rules.

 The use of Zs-1 is logged on an automatic counter and the signalling book has to be filled with the logged number and an explanation as to why the Zs-1 signal was used. When the signal is illuminated it will show its aspect for 90 seconds and then automatically extinguish. On receipt of the Zs-1 the driver must reset his PZB’90 train protection by pushing and holding a button until a 4 lights up and then drive past the signal, to avoid being stopped by the PZB’90 magnet there. The train must travel at 40 km/h (25 mph), prepared to stop before any obstruction, until a next relevant signal is reached, from where the driver adheres to the appropriate aspects shown. However, in case of a starting signal (Ausfahrsignal) exhibiting a Zs-1 aspect, the driver must adhere to the 40 km/h rule until the train has passed the last turnout of the yard or station (Ende Weichenbereich) and is then permitted to accelerate to 100 km/h in section. In the case of the February 2000 accident at Brühl the driver, lacking route knowledge, erroneously thought that he had passed the last turnouts and (against the rules) then accelerated to 120 km/h (75 mph) after passing a Temporary Speed Restriction board with a 12 that was in fact not meant for him. In the case of Bad Aibling the driver passed the last turnout on entering the single line section and then accelerated to 100 km/h (60 mph) as permitted by the rules.

During the week following the accident reports various media reports suggested that Zs-1 signals were being used for rather different purposes to those intended, as was in fact clear from other accidents as well. Whilst rail officials denied it, a signalman’s union official stated in the media that the aspect was occasionally used after “serious deliberation” as a kind of “get out of jail” card. That is what it is there for, according to this statement.

Issues with use of this signal and with emergency communication.

From the accidents mentioned an emerging picture of use, communication problems and the reaction of drivers after receiving the aspect, point at the following safety issues.

  1. The apparent ease of deciding to use the aspect.The decision to use the aspect was taken by a signaller on his own, without further recourse to either another person or equipment checking the justification and validity of the decision in terms of train safety. There is, incidentally, reason to take hints at fairly regular misuse of the aspect in the media at face value.
  2. The apparent ease with which it is possible to overlook section occupation.Like any other electro-magnetic relay signalling installation the Siemens SpDrS60 panel shows the set route with a string of yellow lights in the graphical representation of the track layout. If a track circuit gets occupied the string of yellow lights change to a single red occupied section light. If a particular track section as shown on the diagram is long, as in a single track section between stations, it may be possible to overlook that the section is showing track occupied.
  3. The permitted speed of travel under what is in fact degraded signalling integrity.There is reason to query the safety aspect of allowing a service departing on a Zs-1 aspect to travel at 100 km/hr (60 mph) when the Zs-1 aspect is a case of degraded signalling. Even if the rules prescribe several types of signaller checks that the track ahead of the signal is clear, there is evidence from at least three accidents that indicate uncertainty whether those checks will be properly completed. Certainly when one person controls a longer section of line. 
  4. The issue of checking the use of the aspect from signalling records.Checking of the signalling book, the signalling fault book and the Zs-1 signal occurrence counter figures by supervisors does not appear tohave been either frequent or thorough. That in turn might be interpreted as either a lack of show of authority from the side of supervisors or, worse, a tacit agreement with this type of misuse.
  5. The wrong radio screen destination field used to distribute the emergency message.It is surprising to find that the emergency stop message to the train drivers, which always is a matter of stress and therefore should be very obvious, appears to have been a matter of clicking the “right” button amidst a number of message destination fields. This is where the local GSM-R radio MMI design failed, certainly when it was stated that the collision could have been avoided had the first message been received in the cabs.
  6. Technical aspects.One could also question the technical merit of providing a Zs-1 signal for where it is the final protection of a single line section, or at least use without some supplemental conditions to eliminate the risk of an oncoming train. 

Level Crossings and the Lorries of Doom

april 2016

Ah yes, level crossings! Those locations where ill-disciplined road users interact in fascinating ways with railways. My readers may have noticed that I entertain a few hang-ups with regard to this subject; my number one bugbear is lorries involved in destructive accidents at those locations. The point is: Lorries shouldn’t! Because they are not driven by ill-disciplined private motorists, but by professionals.

This rant in fact is triggered by the widely viewed YouTube videos, from no less than four camera angles, of the destruction of a lorry at Studénka railway station in the Czech Republic on 22-07-2015. It was hit and instantly demolished by Czech Railways Pendolino set 681 003 at approximately 140 km/h impact speed. The Polish driver of the articulated vehicle with a cargo of aluminium sheet material was avoiding motorway tolls through taking local routes, which required considerable extra travel time and caused him to encounter a level crossing. Despite displaying flashing warning lights for at least thirty seconds already, for reasons to be guessed at he decided to enter the crossing anyway, only to find all four barriers coming down at the same moment and trapping him. That, incidentally, points at an issue I have with this level crossing: Had the barriers come down in pairs (as they should these days, the entrance ones first and the exit ones a couple of seconds later) and had the crossing deck been scanned by obstacle detecting equipment that opened an exit barrier in case of detecting an obstacle, this truck would have escaped. 3 lives would have been saved and several injured (the train driver lost both lower legs) would have been spared their ordeal. But also, had the trucker been on the ball in his job, he should have known that closed crossing barriers easily snap off by design when driving through them. But on second thought, is that really something you want society at large to know of? Saturday evening fun on the way home from the pub? Anyhow, the trucker escaped serious physical harm and is presently awaiting trial.

Was this something that occurred once every so many years and for entirely understandable reasons, one could shrug the shoulders and mumble bad luck. But it isn’t like that, because only a year later an electric multiple unit train at speed hit a slowly moving tracked machine on an Automatic Half Barrier (AHB) level crossing at Dalfsen in The Netherlands. That train driver lost his life and the damage was considerable. What is surprising is that the international road transport fraternity, who by now have quite a number of accidents they caused through ignorance to study, appears unable to cotton on to the fact that automatic railway crossings actually are dangerous venues. Certainly if you’re using big vehicles with expensive loads, yet are not prepared to study level crossing functioning and abide by a few rules. That steady ignorance, identifying educational as well as discipline issues, is actually disturbing. Have a look at the short listing below, it is a quick grab and not pretending to be anywhere near complete:

  1. Hixon, UK,06-01-1968, 11 dead, widely publicised and anyone with a professional interest in heavy transport could have known about the dangers of automatic level crossings after it. Main causes: bad scouting and, crucially, no contact with rail operator.
  2. Kissimmee, FL, USA,30-11-1993, 0 dead. Amtrak passenger train crashes into a gas turbine electric power generator being delivered to Cane Island power plant. Main causes: bad scouting, bad preparation and no contact with rail operator.
  3. Kissimmee, FL, USA,17-11-2000, 0 dead. Absolutely unbelievable yet true: exactly the same accident with exactly similar equipment happens again 7 years later. Short memory span at work.
  4. Tossiat, F, 30-01-2003, 1 dead. Heavy road paving machinery hit by a TGV at approx. 150 km/h. Driver of road freight vehicle dead. Convoy was off permitted route, badly scouted and failed to contact rail traffic control.
  5. Domène, F,18-10-2006, 0 dead. Oversize transport hit by a local rail service and destroyed. Main cause: failure to contact rail traffic control.
  6. Kerang, Vict, AUS,05-06-2007, 11 dead. Trucker overlooks dissatisfactory displayed warning indications of an Automatic Open Crossing and runs at 100 km/h into 2ndof 3-coach train. Right hand front corner of trailer slices into the side of the last two coaches.
  7. Halifax, NC USA, 09-03-2015, 0 deaths 55 injured. Amtrak “Carolinian” hits heavy road transport attempting to cross whilst negotiating sharp curves.
  8. Studénka, CZ,22-02-2016, 3 dead. Pioneering foreign trucker makes a mess of passing a level crossing. Vehicle hit by express train at 140 km/h.
  9. Dalfsen, NL,23-02-2016, 1 dead. Tree Surgeon Company contracted by the rail network operator rolls a heavy and slow tracked vehicle across the level crossing. An approaching local train cannot be stopped and hits at 100 km/h. Yet again, simple contact with rail traffic control would have identified a safe moment to cross.
  10. Chester, PA USA,03-04-2016, 2 construction workers dead. Electrically hauled Amtrak “Palmetto” hits erroneously present heavy engineering equipment at 106 mph (170 km/h).

A short trawl through video’s available at LiveLeak and YouTube on the Internet will present four times this amount of heavy road vehicles all over the world getting smashed up on level crossings. Given that the situation merely requires a smidge of operational insight when organising the trip, to contact among others the relevant rail traffic controllers, this listing points at a lack of imaginative yet practical insight. Then there is that dangerous habit among heavy plant movers to solve problems as and when they occur, which is another way of saying that scouting must have been a bit of a joke. But if your vehicle with its precious cargo is stuck and straddling tracks of a busy railway line, it is the last and very costly mistake you made. From the side of the railways, therefore, knowing that growing international awareness from the side of road users cannot be awaited, it could be beneficial, save lives, if they took the initiative on Europe-wide level to inform a large number of people involved in heavy road transport, in offices and behind steering wheels, about the dangers of automatic level crossings.