Search this site
Delude Retreat Indemnify Prevaricate Squander

Talk a good game

TV producers offered up their usual fayre over Christmas, with programmes from both ends of the quality spectrum. A few tasty mince pies were overpowered by a glut of complete turkeys, notably Victoria Wood’s Christmas Eve offering. Thankfully, rubbish becomes more palatable when viewed through the bottom of a beer glass except, obviously, when Noel Edmonds is involved.

Network Rail has once again ventured into the world of moving pictures with its latest disaster epic. And I heartily commend it. As with its other recent efforts, the production values are top class - great photography complemented by a terrific soundtrack - whilst its powerful message is communicated without the moralistic preaching of first-generation safety films.

It’s set in a worksite where an S&T team turns up to test a point motor. Alongside them is a p-way gang with a roadrailer. When one of the S&T team forgets his meter and has to return to the van, you just know that he will come a cropper as he passes behind the RRV. Except he doesn’t. He asks someone whether the machine has stopped - “yes” says the man as he ties up a bag of pandrols. The technician crosses safely but, moments later, the RRV is on the move and the other chap ends up beneath it. Fade to black.

The video will doubtlessly form part of forthcoming safety briefings; it is also available to the wider world through Network Rail’s ‘Safety Central’ website. Access requires registration but it’s worth it as the site now provides some useful resources - amongst them are animations recreating two notable incidents. One of these also features roadrailers, describing the events leading up to a runaway at Glen Garry that could so easily have ended in tragedy. Sadly injury did attend Kennington Junction in May 2008. Six months later, Network Rail had the whys and wherefores ready to download. RAIB’s findings were only unveiled last November, 18 months after the event.
Network Rail's latest safety video features an incident involving a roadrailer.

At 2147 on Friday 23rd May 2008, two signalling technicians were carrying out a facing point lock test on 207A points at Kennington Junction, on the route from Didcot to Oxford. The junction is the starting point for the freight-only Cowley branch. Immediately to its north are Goods loops for both Up and Down traffic which connect to the sidings at Hinksey Yard. The main line to the south curves gently to the right with lineside trees and hedges limiting the sighting of approaching trains.

The leader of the three-man group, a 55-year-old with 25 years railway experience, was also to act as COSS. Before going to site, he had discussed that evening’s work - three facing point lock tests - with the signaller at Oxford who had indicated that there would probably be a suitable ‘margin’. The pre-planned safe system, as detailed on the COSS form, indicated that red zone working should be adopted.

Assisting the team leader was a relatively inexperienced technician but the group’s third member was a proper time-served railwayman, having joined British Rail in 1968. Now though, for personal reasons, he only ever acted as lookout, a role he would undertake on that particular evening.

According to Network Rail's presentation, the group arrived on site at Kennington at 2040; RAIB states that it was 2100. Either way, the sun set at 2105. Their first task was to test 207A and 207B points, comprising the Up and Down crossover. Familiar with the location, the lookout crossed the tracks to take up his customary position in the Up cess or four-foot of the Up Goods loop. Although this position gave optimum sighting of Down trains, round the curve, he would lose sight of his colleagues in the Down cess if a train passed on the Up.
The view looking south around the curve from Kennington Junction; the Down line is on the right.

Photo: Steve Daniels

At 2116, a margin was obtained for the work with three signals being placed to danger. Eight minutes later, the protection was retaken after being given up to allow a train to pass. Then the test failed with the points set in the reverse position. The team leader and his assistant started to make the necessary adjustments to the rods and, at 2128, a request was made for a longer margin - this was declined as a train was already approaching the protecting signal on the Down. The points were wound back and control returned to the signaller. At 2133, with two more trains due, the signaller turned down a further request for extra time to complete the work. The failure of the points in the reverse position did not affect safe running as trains would pass over the points in the trailing direction whilst they were set to normal.

By now, dark clouds were gathering and the light was starting to fade. Whilst waiting for the trains, the team leader and assistant went to collect a torch and more tools from the van. The lookout crossed the line to donate his lamp and, before returning to his post, provided some guidance on how the failure might be corrected.

At 2141, after another train had gone by, the signaller granted a two-minute margin. The light level was now too low for close work so the torches were deployed. Just as the test was successfully completed, after the allotted time, the signaller rang and control was handed back to him.

4O97, a southbound freight service, was slowing for the protecting signal on the Up line when the signaller cleared it. The driver began to accelerate. The lookout spotted the train’s headlight and shouted a warning to the two technicians. They reportedly replied “OK” but, instead of moving to a position of safety, made preparations to replace a missing cover from the point motor. As the train went by, the lookout lost sight of his colleagues. He then saw light reflected off the ends of containers on 4O97 - he concluded that another train was heading in their direction.

The team leader was crouching over the point machine, reattaching the covers, whilst his assistant searched the ballast for the main cover’s padlock. 1D73 was approaching on the Down, travelling at 89mph. As the driver rounded the left-hand curve, he saw two trackworkers on or close to the line ahead of him. He sounded the horn twice. Neither the team leader nor his assistant heard this warning. The latter became aware of the train as it passed him. He shouted a warning but it was too late - the team leader was struck and, as a result of his injuries, had to have a leg amputated.
A Down train approaches Kennington Junction.

Photo: Steve Daniels

The area around Kennington Junction is heavily trafficked and no time for maintenance access had been built into the timetable. As a result, the inquiry found that routine testing there was always carried out red zone; this reality was accepted by everyone and had never been managerially challenged. It was further asserted that arranging possessions would have been administratively complex and the booked time would regularly go to waste as teams could be redirected for faulting work. The form-filling associated with T12/T2 was deemed to act as a disincentive to their use as it would impact on the already significant workload of the signallers in Oxford box. These issues drove the adoption of informal margins for testing activities.

The form-filling associated with T12/T2 was deemed to act as a disincentive to their use...

The accident was found to have a number of underlying causes. There was no proper COSS briefing and the COSS form was not completed - this resulted in the safe system lacking a formalised structure. Both the team leader and his assistant had been conditioned not to move to a position of safety when a train approached on an adjacent line, leaving them open to considerable risk if the lookout’s view became obscured. The poor light conditions resulted in the lookout being unable to tell whether workgroup members had moved clear following his warning or were still in danger. Despite both men having concerns about the safe system, neither the lookout nor the assistant spoke to the team leader about its perceived deficiencies.

It is clear that there was a host of cultural issues affecting the way track work in the Oxford area was planned, managed and implemented. It’s fair to presume that these issues also raise their heads elsewhere on the network. Paperwork was often incomplete and incorrect, with no system in place for checking it. COSSs implemented pre-planned safe systems without their suitability being critically analysed. Briefings didn’t happen and forms, more often than not, were filled-in after the job.

More widely, the inquiry found an indiscriminate approach to the use of certain Rule Book phrases, together with inconsistent interpretation of some instructions. This created confusion for those required to apply them and could lead to incorrect implementation. RAIB also acknowledges Network Rail’s failure to drive the adoption of ATWS/LOWS and develop a simple, fixed means of providing protection or warnings as part of signalling upgrades. The truth is that NR talks a good game when it comes to trackworker safely but it doesn’t support that with a meaningful investment of time and money. Both signallers and on-track staff feel under considerable pressure not to delay trains - a truth that manifested itself at Kennington.

The inquiry lists a job-lot of actions taken by Network Rail to address some of the factors underlying this accident. Sadly, most of these involve reminders and clarifications; not one is truly substantial. RAIB’s recommendations also lack bite. It repeats a number from previous investigations that NR seems intent on ignoring - clearer rules for work in multi-track or junction areas, improved paperwork of value to the end-user, and the implementation of automatic warning systems for red zone inspections. Also advocated is “the development and subsequent adoption of practical alternative working methods” to protect staff involved in routine maintenance. There are no clues as to what this means.
An aerial view of Kennington Junction, with Down trains travelling from bottom-right to top-left.

Picture © GoogleEarth/Infoterra Ltd/Bluesky 2009

Kennington was one of those events caused by a collection of disparate circumstances coming together to wreak havoc. Many of them will - individually or in smaller groups - be active at other sites across the network as you read this. It could well be argued, with some legitimacy, that such conspiracies are an unwelcome fact of life when fallible people are present in an environment with the railway’s potential dangers. But tragedy is not inevitable. It is incumbent on those further up the industry to see beyond their empty process-driven bureaucracy and instead harness technology to deliver robust protection at higher-risk sites. This is not beyond the wit of man - TOWS has been with us for ages. It would though demand motivation and tenacity - two words not readily associated with today’s middle management. And then there’s the issue of cost.

Let’s leave it there…until next time.

Story added 1st January 2010

The glossary at the back of RAIB Kennington inquiry contains several errors, inconsistencies or omissions. They are all relatively minor but the fact that the investigating authority seems unable to correctly define certain basic railway terms does call its expertise and checking procedures into question. In continuing to rely on Ellis’ British Railway Engineering Encyclopaedia for some of its definitions, it demonstrates no faith in the Rule Book.

Incorrect or incomplete definitions

Fenced green zone “...fencing the area off with blue netting or yellow and black tape...”
No mention is made of a safety barrier, probably the most commonly used method.
Separated green zone “…appointing a site warden…”
No mention is made of the two methods that don’t involve a site warden.
On or near the line “...within 3 metres (9 feet 10 inches) of the nearest rail...”
An interesting imperial conversion there!
Position of safety “...1.2 metres (4 feet) at speeds up to and including 100mph...”
And an interesting metric conversion!
T12 protection “...a temporary stoppage of rail traffic for 30 minutes of [sic] less…”
The time limit has been 60 minutes for a couple of years now.

Suspect definitions

Green zone “...a safe place of work, free from trains...”
Is it really free from trains? This has the hallmarks of RAIB spin.
Blockage of the line “...prevents trains from moving...”
How does it do that - by draining their fuel?
“...records are kept by the signaller...”
No mention is made that the IWA/COSS/PC also completes the Line Blockage Form.
Down direction “...a direction away from London...”
This is not always true - check out the route into Grimsby. The Down direction is generally 'away from the railway company's headquarters'.

The important issue here is that RAIB has decided to define these terms differently - albeit only slightly - to the Rule Book but it's the Rule Book that railway staff use as their bible. Why has it chosen to do that?

RAIB: Accident at Kennington Junction
Link to RAIB's Kennington Junction inquiry page, from where you can download a full copy of the report.

Page Top

Front Page | Safety Valve | Jungle Ron | Newshound | Red Tape | On The Line
Four by Three | Forgotten Relics of an Enterprising Age | God's Own County | Image Library

© Four by Three 2014