Clapham
GREG MORSE looks back at the multi-train, multi-fatality accident that took place 30 years ago at Clapham, and what it means for today’s railway
RAIL recalls the tragic events of the fatal accident at Clapham in 1988, and what it means for today’s railway.
Waiting. That’s the lot of the driver when the signal ahead is red: waiting in a loop for an express to pass by; waiting in a yard for the ‘off’.
Red means danger. It also means safe - red signals warn drivers of points set against them, of obstructions up ahead, and (of course) other trains. If a driver passes a signal at ‘danger’ (a SPAD), an accident can occur. It happened at Slough in 1900, it happened at Norton Fitzwarren in 1940, and it happened at Southall and Ladbroke Grove in the final years of the 20th century.
But what if a red signal actually shows green? It should be impossible, and is very rare. Yet it can happen, as the events of December 1988 were to prove all-too-clearly…
Back then - with work well under way on the redevelopment of Liverpool Street, with the East Coast Main Line being electrified, with increasing productivity, more stations, rising passenger numbers, and falling subsidies - it seemed that sectorisation was working, and that a golden age for rail might just be dawning.
It was around this time, amid this spirit of improvement, that resignalling schemes had also been authorised or completed at Brighton, Newcastle, York, Leeds, Leicester and Waterloo. The last was a particularly big operation, involving replacement of equipment on the busiest stretch of railway in Britain.
On the evening of November 27 1988, an
overtired, under-trained technician left a bare live wire dangling in a relay room at Clapham Junction ‘A’, a huge signal box on a gantry that spanned a sea of lines at the station throat. Two weeks later, further work jolted the wire, causing it to touch a terminal, make a connection and prevent a signal from returning to ‘danger’ after the passage of a train.
Just after 0800 on Monday December 12, the 0614 from Poole was heading for the cutting where that signal stood. On board, the crowded passengers were deep in conversation, deep in their morning papers, novels and thoughts. Green light followed green light - the usual story, except the last one should have been red. As the train rounded the curve into the cutting, driver John Rolls saw another train - the 0718 from Basingstoke - blocking the line ahead. He applied the emergency brake, but it was too late.
The collision forced the leading coach to the side, where it struck an empty unit passing on the opposite line, killing conversations… thoughts… and 35 passengers (33 on the day, while two succumbed to their injuries later in hospital). It was, as RAIL would point out, the worst accident since Hither Green in 1967.
All the fatalities occurred in the front two carriages of the Poole train, which had been ripped open down their left-hand sides. Many were trapped in and beneath the wreckage.
Ronald Arlette, who had been on the Basingstoke train, told The Times that there had been “an almighty bang, like an explosion. We went up and we flew over and over. We ended up on an embankment.”
He added: “I know it sounds strange, but we just lay there and we had a laugh and a bit of a joke with each other. It just seemed the best thing to do to keep our spirits up.”
Mark Barthel, who had been on the Poole train, described how a man sitting in front of him “died instantly in the impact”, and how another man who had been “standing in the gap between the first and second carriages had also been crushed to death”. Jim MacMillan, Assistant Chief Officer of the London Fire Brigade, would call the scene “sheer bloody hell”. Few would have disagreed.
Among the emergency services and a number of senior BR staff, Chairman Sir Robert Reid visited the crash site and accepted full responsibility on behalf of the organisation with honesty, dignity and professionalism. The Board’s own investigators were soon on the scene, too. They quickly established the facts… and the cause
It transpired that the ex-Basingstoke had been travelling at around 60mph as it approached WF138, a new signal commissioned on November 27 as part of the resignalling scheme. Driver Alexander McClymont saw it was green, and got the appropriate Automatic Warning System (AWS) indication ( bell plus black disc) to confirm it.
However, when the train was around 30 yards away, the aspect changed from green to red. Something was wrong, and McClymont knew it. He rammed on the brakes, but then realised he’d be brought up short of WF47 signal, where he’d have to climb down and call the signalman.
So he coasted, and made the call to ‘Clapham A’. The phone was picked up by Patrick Cotter, who told McClymont that the indications in his box suggested there was nothing wrong with WF138. The driver was irked, but as he put down the receiver there was indeed “an almighty bang”, and his train was pushed forward several feet in front of him. He called Cotter again and told him that something definitely was wrong, and that casualties were now involved.
Cotter quickly threw all his signals back and sent ‘six bells’ to the neighbouring boxes. At the same time, the alarms went off at
Raynes Park Electrical Control Room, where Electrical Control Operator Ronald Reeves quickly switched off the current to the relevant sections.
This caused the 0653 Wimbledon-Waterloo to start to coast. Its driver, Barry Pike, thought the power failure might be on board his own train, and decided to let it roll into Clapham Junction. It passed a green, it passed a double yellow, and was then approaching WF138. As his train neared the signal, Pike saw the wreckage and was able to bring the unit to a stand some 190 feet from it.
He went back to WF138 and phoned Cotter, telling him that the signal was now at yellow. Cotter said it should be red. “Red aspect be damned,” said Pike. “There are three trains standing in front of it, and it’s still showing one yellow.”
The instruments were wrong, and the problem wasn’t far away. In fact, it was in the relay room right next to ‘Clapham A’…
It wouldn’t be long before BR found the answer - the live wire that had been left dangling. In short, a wrongside failure - a failure where the system had failed to ‘fail safe’.
This discovery was made the same day as the accident, and there was a feeling that the then-Secretary of State for Transport Paul Channon had bowed to media pressure when he launched a public inquiry under Section 7 of the Railways Act 1871.
He probably had (ministers often do), although it wasn’t quite as simple as that. The Tay Bridge disaster had been the first such inquiry, because the Railway Inspectorate (RI) had been involved in the bridge’s approval (over 70 people had been killed by its collapse on December 28 1879). Hixon had been the next, the RI having been involved in the approval process for automatic half-barrier crossings, which had played a large part in the collision that killed 11 people at the start of 1968.
In fact, the RI had been overseeing the Waterloo Area Resignalling Scheme ( WARS) too - only partially, although its role was confused by the (somewhat ambiguous) Road and Rail Traffic Act of 1933, which led to a situation whereby it was aware WARS was being carried out, but had received a submission to approve only part of the work. This meant it seemed as though it had condoned the non-submission of the rest. Arguably, this put the RI in a difficult position. Arguably, a public inquiry it had to be.
Chaired by Anthony Hidden QC, the formal hearings began on February 20 1989 and would receive evidence from 122 witnesses in 56 days. By this time, there had been two further accidents - both fatal, both resulting from SPADs (see panel). It was decided that although Clapham was not a SPAD, the inquiry would also look at any common issues that arose from all three.
Hidden would write that his purpose was “to seek to establish both the immediate and the underlying causes of the accident”. BR had beaten him to the immediate cause, but the report clarified that the underlying cause of the wrongside signal failure had been a series of errors (as most accidents almost always are).
In brief, the end of the wire in question should have been cut back, tied and insulated. The other end should have been disconnected. The accident would not have happened if either precaution had been taken.
These errors - or omissions - had been made by Brian Hemingway, who had joined BR in 1972, and who had been a signalling technician for almost eight years, but who was “over-tired” and “under-trained”. Regarding the latter, he had undertaken five courses in the 1970s, but did not rise to the rank of Senior Technician Scale A in January 1981 by passing an examination, but because of a “change in staffing arrangements”.
Hidden records that this lack of training exacerbated a number of poor working practices that should have been picked up by proper supervision, which at that time was reliant on Peter Dray, a ‘displaced’ engineer filling the vacant supervisory role on a temporary basis (his own job having disappeared in a mist of reorganisation).
Hidden described Dray as a man with “little liking” for the job and ( because of the travelling he was now forced to undertake) “less enthusiasm” for it. That said, Dray himself had not been properly briefed - in particular, on a crucial instruction for an independent wire count to be carried out ( by him, in the case of the WARS scheme at that time).
Partly this was down to poor channels of communication within the Southern Region’s S&T Department. More fundamentally, it was down to a complacent attitude to safety - and BR at that time was complacent. In fact, in reorganising for sectorisation it had bled, just as Gerry Fiennes had warned 20 years before.
As Hidden would write: “The appearance of a proper regard for safety was not the reality. Working practices, supervision of staff, the testing of new works… failed to live up to the concept of safety. They were not safe, they were the opposite.”
The reorganisation did not cause this, but it had failed to “come to grips” with the situation, which is how Dray came to omit the need to ‘count the wires’ and why it would have been exactly the same had his predecessor still been in post - the endemic culture of the company allowed the instruction
to be virtually ignored in the hope it would go away (to paraphrase Hidden).
Part of that attitude was also responsible for the ‘over-tired’ side of the Hemingway equation - the side that led him to make a number of ‘uncharacteristic errors’ as a result of undertaking constant, repetitive work and excessive levels of overtime, both of which had “blunted his working edge”. To be more explicit, he had had just one day off in the previous 13 weeks.
Among the report’s many recommendations was one to “ensure that overtime is monitored so that no individual is working excessive levels of overtime”. This led to the development of criteria for what was considered acceptable levels of working (known as ‘Hidden 18’), and a process to monitor it. Fatigue management has developed even more since then - indeed, the rail industry now has a whole new standard on managing the risk from fatigue, and continues to place much focus on this vital area.
New processes and instructions were also introduced regarding the installation and testing of signalling works. On safety culture, it was recommended that “British Rail continue to press ahead with its Total Quality Management Initiative and the application of British Standard BS5750”. Originally termed ‘Organising for Quality’, this led to a greater focus on business-led ‘profit centres’ within the Sectors, but also (to quote historian Terry Gourvish) “involved the identification of very clear lines of responsibility for safety… validated by the Safety & Standards Directorate”. BR’s ‘holistic’ structure made this a relatively straightforward process, and thus a generally safe railway was handed over when Railtrack took control of the track and signalling at the start of privatisation in 1994.
However, for breaching the Health and Safety at Work etc Act 1974, the British Railways Board was fined £ 250,000, and there was some frustration at the time that there was no prosecution for manslaughter. That said, Clapham was a catalyst for change, in that (in 1996) it was one of the events that led the Law Commission to form the Corporate Manslaughter and Corporate Homicide Act of 2007.
Regarding the SPAD situation, Clapham (or rather Purley and Bellgrove) also helped the industry down the path towards Automatic Train Protection (ATP) - or its more costeffective variant TPWS ( Train Protection and Warning System), which (like ATP) also improves on AWS by automatically applying the brakes on a train that has passed a fitted signal at ‘danger’ or is approaching one too fast.
It has successfully helped put SPAD risk down to its lowest-ever level, and is merely one in a long line of safety improvements - from the adoption of ‘lock, block and brake’ by 1900, to the increasing mechanisation that has helped train accident fatalities fall since 1950.
And it certainly won’t be the last - not in an industry where staff retire, move on or move in from elsewhere, and where the human element still sits alongside technologies old and new. For this reason, the railway will go on monitoring trends, go on innovating, keep remembering the lessons of the past, and keep sharing those of the present - until the future means it no longer needs to. And yet…
The corporate memory exists only while we remember it, and over the Christmas and New Year period of 2016-17, we seemed to forget. Extensive resignalling and track
remodelling work was being carried out in and around Cardiff Central at this time, with some of the new layout being brought into use on December 29. At 0837 that day, the driver of a Treherbert service noticed that the points his train was about to take were not in the correct position. He stopped the train just before reaching them.
The Rail Accident Investigation Branch (RAIB) would conclude that the points had been left in this “unsafe condition” because they hadn’t been identified as needing to be secured by the points securing team. Furthermore, no one had checked that all the points that needed to be secured during the works over the Christmas period had actually been secured. Route-proving trains had also been cancelled, and a culture had developed between long-standing members of the project team that led to “insular thinking about methods of work and operational risk”, meaning that team members “relied on verbal communications and assurances”. RAIB also believed ineffective fatigue management to be a possible underlying factor.
RAIB Chief Inspector Simon French drew a clear line from Cardiff back to Clapham, pointing out “how easily things can go wrong when railway infrastructure is being upgraded and renewed”, and the importance of managing the working hours of people doing the job “when organising intensive periods of commissioning work”.
He continued: “Back in 1988, the disastrous collision at Clapham Junction happened in part because working for weeks on end without any days off was part of the culture in some areas of the railway.” The events at Cardiff showed “how easy it is to forget the lessons of Clapham and slip back into those habits under the time pressures of a big commissioning”.
The Rail Delivery Group’s Director of Planning, Engineering and Operations Gary Cooper agrees, telling RAIL: “Clapham changed BR’s culture and the culture of those involved in BR at the time, who are still in the industry, forever. An issue is that there are fewer and fewer left who went through that change, which we need to address as an industry collectively and as separate businesses and organisations.”
Indeed. But there’s more… a few months later, in August 2017, a train departed Waterloo on a green aspect, but was incorrectly routed and collided with an engineer’s train on the adjacent line. Luckily, the driver saw the way the points were set and managed to brake, meaning that the collision occurred at low speed and resulted in no injuries. Modification to the wiring of the point detection circuits meant that a ‘desk’ set up to aid testing no longer simulated the detection of the points in question correctly because it hadn’t been modified to account for changes made to the detection circuit.
On the weekend of August 12-13 2017, while trains had been stopped from running on the lines leading to the points, a temporary wiring “mod” was made in the relay room in an attempt to restore the correct operation of the relevant switch on the test desk. But the mod wasn’t reviewed by a signalling designer, and was wrongly left in place when the railway was returned to operation on the morning of August 14.
Not only could we quote Simon French again here, we could also quote Anthony Hidden again. We can also re-remind ourselves that
we combat complacency with continued vigilance, and that corporate memory exists only while we remember it. After all, we have a generation of railway people coming up who will not have lived through Clapham… and not have lived through Southall, Ladbroke Grove, Hatfield, Ufton Nervet or Potters Bar either.
We can also talk about railways being whole systems - something we don’t always seem to consider when the work of one sector cuts across that of another. Everyone needs to understand how their action (or inaction) can affect safety - from senior managers to people insulating wires, tightening fasteners, checking, testing… actually doing the work.
At the same time, we need to be aware that what constituted a risk in 1988 may not be one in 2018 - or rather 2028, as our reliance on digital technology grows and puts a correspondingly greater emphasis on software integrity and cyber security. ( We do seem to keep throwing a lot of our eggs into an ever-decreasing number of baskets, as RAIB’s ongoing investigation into the loss of temporary speed restrictions on the ERTMSfitted Cambrian line may suggest.)
In short, it’s all about understanding and managing risk. Hidden suggested that BR had become almost blind to the risk from wrongside failures, contrasting it with its focus on SPAD risk.
BR was probably right to put proportionately more focus on SPAD risk in the late 1980s, but not to the exclusion of wrongside failures (or any other hazard, come to that).
Indeed, there had been a number of ‘Claphams in the making’ that might have been highlighted by a greater emphasis on learning from operational experience. More specifically, there had been a ‘cluster’ of wrongside failures in Oxted, Northfleet and East Croydon in November 1985, during the installation of new signalling.
Of these, the Oxted incident was the most worrying, as a signal had shown green when it should not have done because a relay had been energised irregularly. This fault would have been discovered by a wire count, but (as with Clapham three years later) no such count had been undertaken. Worse still, the resulting ‘flurry of paperwork’ provided important information, but was shared with very few people and therefore did not feature in anyone’s thinking during WARS.
A similar wrongside failure occurred at Queenstown Road on June 14 1988, where a signal cleared to green instead of yellow. In this case, there had been a design error (a drawing being issued which omitted a track circuit from the signal’s controls). The error was not picked up by the Design Office, nor during the testing of the signal on the ground.
Of course, it’s totally rational to concentrate most risk reduction effort into the areas where risk is highest or where there is the most scope for risk reduction, but it’s also vital that risk assessments are kept fresh. It’s also important to avoid being blinded by solid trend lines and to look sometimes at the outliers (the ‘weak signals’), where perhaps the data points are fewer, but where the consequences might be great if the situation is allowed to persist. There is no better argument for taking such a holistic view of risk than Clapham.
We all know that we can increase the accuracy of our risk picture by collecting, analysing and learning from information - not just about accidents, but also their precursors and the activities that prevent them.
Hidden was damning about BR’s failure to collect information on wrongside failures systematically. After Clapham, however, it tightened up its safety arm and centralised recording by bringing together previously disparate sets of information into its British Rail Information Management System - a computerised database, and the forerunner of today’s Safety Management Intelligence System.
But as Cardiff and Waterloo remind us, data and information (from the past and the present) are only any use if we analyse results, understand what they mean and act on them - out on the railway, not just on paper…
Grateful thanks are due to Roger Badger (RSSB), Matt Clements (RSSB), Gary Cooper (RDG), Andy Fawkes ( VolkerRail), Derek Hotchkiss (RSSB), Neil Massey ( VolkerRail), Stuart Webster Spriggs ( VolkerRail), Michael Woods (RSSB), and RAIB.