The Railways

  Many will remember what happened when the culture of safeguards on Britain’s railways was last seriously compromised. The culmination came on 17 October 2000, when a defective rail fractured on a curve just south of Hatfield as a Leeds-bound express passed over it at 115 mph, killing four passengers. By that date the maintenance of the tracks had passed into the hands of private contractors engaged by Railtrack, the privatised infrastructure company. These contractors worked under regimes of incentives and penalties for completing work on time, even as the railways came under growing strain from unexpected increases in traffic. The privatised system created in 1993–6, with its separation of operating companies, rolling stock companies, infrastructure and maintenance, was characterised by displaced or uncertain responsibilities. The changes went against the grain of the existing operation of BR, which had been successfully reorganised not long before into separate business sectors: InterCity (by then no longer hyphenated), Network SouthEast, Regional Railways and Railfreight. These were able to keep tighter control of their operating costs than the old regional divisions had done, while sustaining a close relationship with the engineering and infrastructure side of the network. Privatisation made no serious attempt to engage with the special character of railway operation and engineering; instead, it drew on the one-size-fits-all ideology of the free marketeer, by which every exchange must be monetised, and competition within an agreed legal and financial framework is the universal ideal. Co-operative relationships were thus replaced – deliberately and knowingly – by adversarial ones.

  So it came about that the contractors who had detected early faults in the rail at Hatfield were required to compete for the job of replacing it, and another firm won the work instead. The victors then became bogged down in negotiations with Railtrack over how much time they should be allowed for possession of the track in order to set things right. Over at Railtrack, cost-effective management involved a constant quest for ways of minimising the penalty charges it would have to pay the regulator for delays to services arising from necessary maintenance. Meanwhile the suspect rail remained in place, until it shattered. With profit as the benchmark, efficiency had been fatally confused with financial advantage, and financial advantage with genuine economy – for the administrative and legal costs of running a railway in this way can easily be imagined.

  Different priorities now took hold at Railtrack. It emerged that the company had shrugged off much of its former knowledge of the system when it was divested of the task of maintaining it. The result was that Railtrack no longer had a coherent picture of the physical state of its own domain. As the company rushed to catch up, faulty rails were identified in frightening numbers all across the network. Emergency speed restrictions were slapped on line after line, to a total of well over a thousand, and travel times on many routes collapsed to those of the 1850s. A month after the accident, trains were taking almost nine hours to reach Newcastle from London. One significant figure within the industry at the time summarised the situation thus:

  The railway as a system, under BR it was totally integrated and one person or group of people were able to balance the system. Performance, safety, efficiency, capacity, growth, it is all one system. I think that privatisation did fragment that system into over 100 different parts. That fragmentation did mean that the accountabilities were diffused and many of the different parts were set up with an economic architecture which by definition pointed them in different directions.

  The man thus releasing cat from bag was Gerald Corbett, chief executive at Railtrack, in response to questions from Lord Cullen, who had chaired the enquiry into the railway disaster at Paddington in 1999. Corbett resigned his post shortly afterwards, taking with him a payoff valued at £1.4 million, and Railtrack buckled down to the task of addressing the backlog of repairs to its own sweated assets. This pushed the company’s finances deeply into the red, and its share price into a nosedive.

  By 2002 the position that Railtrack should still be treated as a profit-making business had become impossible to sustain, and it was taken back into public ownership under the name of Network Rail. While the process was under way, another seven passengers were killed when an express was derailed at Potters Bar, a few miles south of the Hatfield crash. Once again, inadequate maintenance by a private contractor was the cause. Not long afterwards it was announced that Network Rail would itself undertake all routine maintenance of the track. Costs fell sharply as a result. Safety has also benefited: there have been no fatalities to passengers since 2007, and none to employees working on the line since 2009.

  Other intrinsic contradictions remain in place. At the most basic level, Network Rail must compensate the operating companies for disruption to the timetable caused by improvements whose only purpose is to improve the companies’ own services. The privately owned railway also needs much more public money in order to function at all. In the last days of British Rail, the system received a direct annual subsidy of between £1 billion and £1.5 billion. Subsidies since that time have reached as high as £6.3 billion (in 2006–7); the figure for 2013–14 was £5.3 billion, despite the doubling of passenger numbers since privatisation. Much of this money comes nowhere near the operating side of the railway, but is sucked straight out again in dividends, administrative and legal costs, inflated salaries and bonuses. Nor is the system cheap for its users. All of these features are intrinsic, not accidental, parts of the business model under which the railways were privatised – a process that, let it be remembered, was meant to address the supposed scandal of a publicly owned system which required high subsidies in order to operate. It has proved an extremely expensive way of saving money.

  More recently, ‘deep alliances’ – joint management of assets – have been contracted between Network Rail and certain operating companies, beginning in 2012. They represent another reversal of the 1990s structure, the outcomes of which were described by Prime Minister David Cameron in a speech to the Institution of Civil Engineers in March 2012: ‘Our railways are crowded and expensive, compared to French, Dutch and Swiss railways, our fares are 30 per cent higher, our running costs 40 per cent higher and our public subsidy is double.’

  The latest twist to the story concerns the finances of Network Rail. Its income derives from the access charges that the operating companies pay to use the track. These have never reflected the true costs of maintenance and upgrading, and the shortfall has been piled on in the form of debt, with the government as guarantor. By the end of 2013 Network Rail was spending more on interest charges than on track maintenance and renewal. Shortly afterwards, European directives swung into play, requiring greater transparency in the classification of public borrowing. So the sum – an astonishing £34 billion by August 2014 – will now be accounted part of the national debt. Such is the cumulative result of having rigged the system so that its private constituents can return a guaranteed profit. Mr Gladstone would have seen through this rigmarole very quickly.

  Footnotes

  * One tradition places some of the blame on rats chewing holes in the leather. Recent historians have been more sceptical, although anyone who thinks that a greasy leather flap could remain in place at ground level across miles of country without once receiving the attention of inquisitive incisors has surely underestimated the English rat.

  ** Stranded, sold-off bodies of this kind should not be confused with so-called camping coaches, a fondly remembered holiday facility using converted stationary carriages. These were introduced on all four of the big railway companies in the 1930s, and a few lingered until 1971.

  *** The term platelayer is a fossil from the early years of railways: the ‘plates’ were the L-shaped metal guide rails used to control the flangeless wheels of vehicles on horse-drawn waggonways. Likewise, the permanent way is so called to distinguish it from the temporary tracks laid down when railways were under construction.

  – 10 –

  SIGNALS AND WIRES

  Thanks to the attentions of the platelayers,
the typical railway line was a more animated place in former days. A traveller in the 1860s would also have been made aware of another aspect of railway work, represented by a new type of structure that was springing up alongside the lines: the signal box.

  Dickens was bang up to date in setting his ghost story of 1866 in one of these boxes. His title, ‘The Signal-Man’, itself had a new-minted character, for signals developed in the wake of the early railways, not in step with them. At first, responsibility for controlling traffic was shared between the stationmaster or his subordinate who saw off the train, the policemen (as they were known) who stood alongside or patrolled up and down the line, and the pointsmen who switched the trains between routes. Time, not space, was the safeguard: distances between trains were maintained by fixed intervals between departures, regulated by watches or hourglasses supplied to the men. Information and instructions were conveyed by means of hand signals, flags or lamps; drivers and guards kept a constant lookout for trains in front or behind. Even when the telegraph began to thread its wires from station to station, it was some time before these were used consistently to control traffic on the railway.

  An integrated system of signalling came about only by stages. It combined inventions that were made separately, and which at first were of limited application. The earliest fixed signals – commonly rotating discs, boards or crossbars – were located at stations and junctions and were effectively the amplified hand-signals of their operators, who stood alongside. Signals that could be worked remotely developed in the late 1840s, beginning with an installation by a practical-minded pointsman on the North British Railway, who rigged up the signals under his control with wires so that he could operate both of them from his cabin. A later advance was to deploy two signals, one to govern possession of the upcoming station, junction or section of track, the other to give advance warning of a stop ahead. These eventually became known as ‘home’ and ‘distant’ signals. To allow both departures and arrivals to be regulated, the home type was commonly repeated at the end of a platform (known as a starting signal). The default setting for these early signals was ‘clear’; only when the line was obstructed were they turned to show danger.

  For ease of working, the levers by which the signals were controlled might be concentrated together, commonly at an open-air platform next to the track. A further efficiency was to operate the points remotely too, a process perfected by the use of heavy jointed rods placed close to the ground. The levers attached to these could then be worked from the same places as those for the signals. An installation of this kind became known as a ground frame.

  Early ground frames lacked any mechanism to prevent conflict between different signals, or between what the signals showed and the position of the points ahead. Safe passage therefore depended on the vigilance and memory of the operator. Various attempts were made to address this weakness in the 1840s and 1850s by linking the controls together. The culmination, achieved first in 1860 at Kentish Town Junction on the North London Railway, was to interlock the operation of all points and signals over a given extent of track, grouping the controls within an iron frame in such a way that none could be moved except in the precise order that would ensure safe operation.

  Interlocking lever frames were hefty things and required the equivalent of a single-storey building to house them. The operating platform sat on top, usually reached by an external staircase. It was an easy matter to put a roof or canopy over this platform and to enclose it within screens or large-windowed walls. The result was an instantly recognisable new building type, its basic form dictated entirely by function. Early examples were often combined with signal posts, which typically poked up through the roof. Later, as the convention took hold that each signal should stand alongside the point where a train should stop, the physical union between signals and box was ended in favour of remote connection by operating wires. A front balcony was a common feature of the box, less to help signalmen speak to train crews than to make it easier to clean soot deposits off the windows.

  Inside a signal box at London Bridge, from the Illustrated London News, 1866. Lever frame, telegraph instruments and register desk would all have been familiar to a ‘mechanical’ signalman a century or more later

  These early boxes were soon linked by telegraph. They communicated by a coded system of long and short signals that were reproduced by means of bells, by electro-magnetic deflections of a needle to left or right within an indicator box, or by the two in combination. Dickens’s signalman is possessed by extra vibrations from the bell of his cabin, which only he can hear. His is a one-man box, the most common type. At large stations and junctions the boxes were much bigger, with staff working in shifts round the clock. Some 13,000 of these cabins, large or small, were required to operate the network around 1900.

  The mature system thus combined three elements, none of which yet existed on the pioneering railways of the 1830s: the signals themselves, the boxes that controlled them and the telegraph wires strung out on poles between. Continuously present, and mostly taller than the trains they served, all three elements played a part in making the railway more conspicuous in the landscape. In functional terms, they were also standing reminders of its autonomy, of the self-contained, self-governing character of the ribbons of railway land that had carved up the ancient estates and smallholdings of the kingdom. Even if no trains were running, the alien sound made by the telegraph wires whenever the wind rose proclaimed the existence of the railway; the effect was compared to that of an Aeolian harp. For the passenger looking out to one side, a visual rhythm now accompanied the railway journey: the repeated downstrokes of the poles every fifty or sixty yards, and the incessant sweep and dip of the wires between, calibrating the distance travelled as surely as the staccato noises from the joints in the rail. It is a sight that has disappeared almost completely, as the once-ubiquitous poles bearing wires mounted on little ceramic insulators have made way for multi-core signal cables boxed within trunking along the ground. One of the last strongholds was the line between Ely and Norwich, converted in 2012–13. In their final years the poles here leant at crazy angles, the burden of their multiple crossbars reduced to two or three wires.

  Why were there once so many wires strung alongside the tracks? Part of the reason lies with the symbiosis between the railways and the telegraph system, the precursor of the pre-satellite telecommunications network. Most of the early circuits along railways were installed by arrangement with private operating companies, eager to exploit the potential of these long, secure routes under single ownership. By setting up public message offices at railway stations, the telegraph companies also enhanced these buildings as local centres of communication. The earliest commercial electric telegraph anywhere shadowed the Great Western’s route almost as soon as it opened: the first stage from Paddington to West Drayton in 1839, an extension to Slough in 1843. Its wires gleamed with the lustre of royal patronage in 1844, when they carried tidings to London of the birth of Prince Alfred, Duke of Edinburgh, at Windsor Castle. One year later, a murderer called John Tawell was caught after fleeing to the capital, when a description was telegraphed from Slough in time for detectives to intercept the train at Paddington. Further publicity came from stunts such as the nine-hour chess match staged by the London & Southampton Railway between adversaries in London and Gosport in 1845. Thackeray’s Jeames, awestruck at encountering the invention at Gloucester, caught the general mood; another observer hailed the electric telegraph as ‘the railway of thought’.

  The potential of telegraphy as an aid to railway working seems obvious, yet the Great Western held aloof from using the technology for its own purposes until the 1850s. It was a similar story elsewhere. When the railways did adopt the telegraph, separate wires and circuits were employed for internal communications, so that these did not conflict with the public, revenue-earning messages. The use of telegraph codes to regulate the passage of the trains themselves took several more years to become widespread, and required e
xtra wires in turn. The later advent of telephony and electricity supply meant that extra wires were again needed, physically heavier than the thin telegraphic kind.

  Interspersed with the telegraph poles were the signals themselves. Britain’s mechanical semaphore signals all derive ultimately from the founding installation in 1841 at New Cross Gate, on the busy London & Croydon Railway. As standardised in the twentieth century, the two main kinds are instantly recognisable: the home signal with its long arm painted bright red on the approach side (signal red, indeed), with a white vertical bar close to the far end, the distant signal arm a strong yellow with a smart black chevron, echoing the outline of the fishtail cut-out at the outer end. The end of the arm next to the post is pierced with two big openings filled with coloured glass: red and green for a home signal, yellow and green for a distant. As the arm pivots, one or other of these panes aligns with a lamp housing fixed to the post, so that a coloured light repeats the message of the signal arm in the hours of darkness. The information presented by the pairing of a home and a distant signal on a single post is thus like that of a road traffic light, except that a yellow light on the distant serves to warn that the next home signal along the track is presently at danger and may require the train to stop in due course.

  Signals of this kind are still surprisingly widespread on secondary lines across Britain. The last mechanical signal box to be built, at Uttoxeter in Staffordshire, opened as recently as 1981, and new semaphores are put up from time to time even now, as at Heckington (Lincolnshire) in 2013. A target for their abolition was announced by Network Rail in the same year, under which the 800 remaining signal boxes and control offices across the country will dwindle to just fourteen. The time-span for this project – 20 per cent of the reduction is not due until after 2026 – suggests that the old order will endure here and there for a good while yet.