LRTA logo Light Rail Transit Association
Light Rail for better public transport

Making LRT systems safe

Part Two — The Infrastructure

Major Christopher 'Kit' Holden has been the public image of safety for promoters, builders and operators of 'new generation' light rail for many years. Now retired from the Railway Inspectorate, and a busy consultant, he explains how the modern safety regime has evolved.

At the beginning of this second article I must remind you of the Government Health Warning; the views which follow are strictly my own and are not, necessarily, Her Majesty’s Railway Inspectorate’s (HMRI) current policy. In the first article, I explained that one of the ‘Victorian’ principles upon which the formal attitude towards safety policy rested was that, if a railway had been properly constructed and thence had been inspected and approved, it would be as near as possible accident free. I also wrote that, coupled with the other two principles, this still conditioned HMRI’s attitude today. Hence one of the keys to safe operation is the provision of appropriate clearances both in the electrical sense and in the gap between fixed and moving objects. Note the use of the word appropriate. It is possible, perhaps not always fairly, to summarise the Victorian view of engineering safety as being - “make it thicker, wider, stronger etc. and you will make it safer”. They had not heard of risk assessment. Part of this article will try to analyse why and how the clearances given in Railway Safety Principles & Guidance Part 2 (Section G), generally referred to as RSPG 2(G), came about.

In the discussion of clearances I like to use the analogy of the mouse and the hole. Whilst this is in many ways more applicable to an existing railway or tramway it is, nevertheless, a useful concept for planning entirely new systems because it illustrates the very necessary dialogue between the civil and the mechanical engineer. The analogy is quite simple. If you have a mouse of a given size, which includes length of tail and whiskers, then the required clearances dictate the required size of hole at any point in the mouse’s travels. If, on the other hand, the constraint is the maximum size of hole you can afford, then the clearances dictate the maximum size and shape of the mouse. There are a number of measures which can be taken to reduce the size of the mouse in certain circumstances ( this will be explored in my next article ) but by and large if you fix one, mouse or hole, then you fix the other. Clearly the Victorians had the principles of clearances in the back of their minds, but in places they were beguiled into setting maximum sizes, particularly of vehicles, without perhaps fully appreciating the flexibility of only prescribing clearances. This can be illustrated by taking some of the dimensions in the Tramways Act 1870. For example the overall width of a tram was effectively set by stipulating in that Act that, unless otherwise prescribed in the Provisional Order and later turned into the Special Act, the gauge was to be 4 ft 8½ in and that the side of the carriage should not overlap the wheels by more than 11 in. Hence if the wheel treads were, say, 4 in wide then the maximum carriage width was 7 ft 2½ in ( 2.2 m ). So the 1870 Act has defined the mouse, but only set rules for the hole by giving powers to the frontagers to dissent if a 30 ft ( 9.144m ) length or more of the rails were closer than 9 ft 6 in ( 2.9 m ) to the kerb. It was left to the 1926 Memorandum to set the 15 in ( 380 mm ) minimum clearance between the sides of passing cars and between a car and a structure. Thus the minimum lane width required for a maximum sized tram was 9 ft 8½ in ( 2.96 m ) and, if the 9 ft 6 in requirement had to be observed, then the minimum width of carriageway for a double-track tramway worked out at 32 ft 1 in ( 9.78 m ).

It was not apparent from the Memorandum whether the 15 in minimum between passing cars was to be measured statically or by assuming an effective width when they were in motion. 15 in ( 380 mm ) was also the minimum clearance required by the Blue Book between passing trains on a new mainline railway but the latter could be reduced to 100 mm on existing railways At the outset, for the PGN (The Provisional Guidance Note on the Highway and Vehicle Engineering Aspects of Street Running Light Rapid Transit Systems), it was decided to follow the mainline concept of a Kinematic Envelope (KE) for a tram; that is to define the maximum envelope taken up by a tram in motion on a straight and level track. Because this takes account of sway and any tolerances, amongst others, in the width between gauge faces of the wheel treads, it is speed dependant; i.e. the greater the speed the greater the width of the KE. This was not a concept well understood by the tram builders. The use of the properties of the KE had two benefits. It was possible to define the swept path taken by a tram as it moved along the tramway and hence to mark a safe distance, originally defined as the old 380 mm, from it and also, by making use of the narrower KE at low speeds through platforms at tramstops, bring the platform edge much closer to the tram-door threshold than is the case on mainline railways. This has obvious advantages for the mobility impaired. One downside of this approach was that it led to much greater clearances being provided when Manchester Metrolink was being constructed than existed either on the Blackpool system or on the continent. Although some of the required clearances, particularly to centre traction poles, were reduced for Sheffield it was only later that the argument was advanced that, far from reducing the danger to pedestrians from being squashed by a passing tram, it was in fact increasing it because the trams normally appeared to be sufficiently far apart for a safe space to exist between them. It was therefore deemed possible, without loss of safety in abnormal conditions, to reduce the required clearances, not by much, in RSPG 2(G).

Perhaps one of the strongest precepts was to ensure that, in all its aspects, the tramway was visible and its route obvious. It was felt necessary to cater for a number of different types of running within one system; the old legislation had recognised but one, the street-running tramway. Purely arbitrarily three distinct definitions were coined, the simplest being effectively the old, street-running tramway where trams ‘mixed it’ with other road users. At the other extreme was any part of a system which, to all intents and purposes, was a railway, and to which most of the requirements of the old Blue Book were appropriate. There was, however, a slight concession in that if the trams were driven on line-of-sight and could therefore be stopped within sighting distance of an obstruction, both animate and inanimate, then the tramway did not need to be fully fenced. Lastly there was the middle ground where the tramway was segregated to a greater or lesser extent from other road users but, under certain circumstances, its track might have to be used by vehicles other than trams and pedestrians, whilst not being excluded, were definitely discouraged. I think that, in the final analysis, whilst initially being of help in clarifying some of the technical issues, these definitions became too rigidly lodged in people’s minds and the objective of providing an appropriate solution of ensuring a safe environment for all who used the tramway and those who shared the tramway alignment became blurred. RSPG 2(G) goes some way towards correcting this. I do not have space in this article to enlarge on all the techniques which were devised to mark the separation between the categories of tramway or of the ways of marking the actual tramway itself on-street, with particular reference to those who have some form of visual impairment. You will either have to read RSPG 2(G) or go and visit Manchester, Sheffield, West Midlands, or Croydon.

Signalling perhaps deserves an article all to itself but this is not possible. Again it was quite easy to define what was necessary at the extremes. A ‘light railway’ would need railway-type signalling; the signals and points being fully interlocked and the signals themselves probably only 2-aspect (red/green). On-street of course ordinary traffic signals would suffice, or would they? The use of red/green signals off-street is complicated when running parallel to a main line railway because of the requirement to co-locate them to avoid misreading by drivers on the other system. Signal spacing for tramways is very different to that for railways because of the very different braking characteristics. The different meaning attached to a green railway signal as opposed to a green road traffic signal had also to be considered. The visit to the continent crystallised the view that was beginning to form that tramway signals had to be totally different in aspect to either and would have the benefit that trams could be given exemption from observing red road traffic signals. There was no standard continental tram signal but most used a white bar to indicate stop or proceed if the route was clear, with varying forms of prepare to stop (and sometimes to go) aspect. The Department of Transport was eventually persuaded to prescribe (1) the aspects of the tramway signal for use on the highway which are now used in this country. However the actual signal head design has caused some problems. There was a strong desire to provide all three aspects in the one head so that its integration with the existing road traffic light signals could follow the rules for filter arrows etc.. Two prototypes were tried out in Blackpool, one an early version of the ‘21-eye’ signal currently in use in Manchester etc. and the other a fibre-optic version. The latter probably gave a better aspect but was too complicated and too expensive to fit into a standard 300mm diameter signal head. However the former has not been an unmitigated success, having a fairly high maintenance liability because of the short lamp life. An experimental LED signal head was demonstrated whilst the Sheffield system was being built, but, at that time, it was not possible to achieve the true white colour demanded by the Department of Transport. Advances in LED technology will probably overcome that objection in due course but until they do LRT signals will continue to be a high-maintenance item. It is, however, still open for someone to challenge the use of a single head and to persuade DETR that the simpler, three-individual-aspect heads are equally acceptable.

From the outset it was made plain that it was highly undesirable for points to be detected in their correct lie, let alone controlled by, the controller used to control the traffic signals at a road junction. The reasons are too complex to go into detail here but arise from the response to failures by and within the controller. This has meant that a separate ‘points indicator’ has to be provided with the attendant, albeit small, risk that the tram driver will fail to notice any conflict. In order to preserve the integrity of concept where signals are needed on segregated sections which are otherwise driven on line-of-sight, standard ‘21-eye’ signals have been used but differenced by a plate to indicate that, whilst it is still necessary to proceed only as far as the line is clear, the route ahead is secured. Examples of this are on the ramp up to G-Mex and on the depot entrances and single line sections in Sheffield.

Some of the concepts foreseen in the PGN have not materialised, extensive use of ‘trambaan’ for one. A fairly radical review of clearances, both structural and electrical, has meant that these have been reduced. I am sure that these are steps in the right direction, but a start had to be made somewhere. Only the initial system in Manchester had the disbenefit of having to use over- generous clearances in the city centre, they were of less consequence on the former railway lines to Bury and Altrincham. It was fortunate that the experience in Tuen Mun of trying to graft a tramway signalling system onto an existing road traffic light control system which was not designed to cope with it enabled a rather better system to be used in Manchester. This in turn was further developed using different techniques but preserving the same underlying principles in Sheffield. Later systems should be able to develop these techniques even further.

Note: (1) The word 'prescribe' was incorrectly written as 'proscribe' - Corrected 14 May 1999

Making LRT systems safe Pt. 2: top of page