Hunting at Thirsk

It’s 1967, it’s the last day of July and a freight is pulling in to Skelton en route from Cliffe to Uddingston. It’s a regular turn, a ‘block train’, the sort of train that would’ve made Beeching proud.

The guard exchanges greetings with the man about to take over. A wagon examiner plies his trade, but the new guard looks at the couplings himself. Towards the rear of the rake he finds a slack one, tightens it up and prepares for the off. When it comes, he settles so he can keep an eye on things from the brake van’s ducket window. He always does this for cement trains. Cement trains – ‘Cemflo’ wagons, at least – can be trouble.

He’s right, they can be trouble, and – shortly after 15:15 – trouble comes as one ‘Cemflo’ starts to sway, its movement increasing until its wheels leave the rails near Thirsk. To the guard, it looks like a bulge, but quickly becomes much more as a coupling breaks and eight wagons disappear down the embankment. All but one come to rest clear of the adjacent line.

Looming large on that line is the 12:00 King’s Cross-Edinburgh, its driver enjoying a clear run after a temporary – frustrating – speed restriction near Tollerton. He could be powering up; he should be able to see the next signal. He can’t. Something’s in the way. A mist? A haze? He closes the controller, but as he starts to apply the brake, out of the haze comes the wagon. He throws out all the anchors. He works the sander to aid adhesion. His secondman gets three detonators ready to protect the line if they survive. If...

...the wagon tears into the side of the locomotive and six of the carriages. The sound is excruciating, but they’re alive to hear it. Not so seven of the passengers.

In a nearby field, a farm worker saw the freight, saw its tail swing down the embankment, saw the guard running back, saw the collision that couldn’t be avoided. He ran too. Ran home. Ran home to call an ambulance. The ambulance came. The ambulance crew tended.


Forty-five people would eventually be taken to hospital, 15 of them with serious injuries. One of the survivors told the press that the cement wagon had been ‘like a can opener’, lacerating the express as it screeched to a stand. The driver of that train later admitted he’d shut his eyes and hoped for the best at the last moment. ‘It probably would have been much worse if it had been head-on,’ he said.

Published later that year, the investigation report began with a description of the accident and its aftermath, before going on to describe the area, the infrastructure, the signalling and the trains involved. More room was reserved for a consideration of the ‘Cemflo’ wagons, which were privately owned by Associated Portland Cement Manufacturers Ltd, had been built by Metro-Cammell, approved by BR, had started to appear from 1961, and had been blighted by problems with broken springs and underframe cracks, such that their permitted speed had been cut from 60mph to 45 when laden. Tests had also shown that ‘hunting’– a strange lateral wheel oscillation – could develop at around 25mph and was exaggerated if couplings were slack. The guard’s unease had been well founded, it seemed.

Examinations on site found the twelfth ‘Cemflo’ in the consist – LA233 – to have been the first to come off, its trailing wheels carrying ‘the typical marks that occur from having run in a derailed state for some distance’. When new, the UIC ‘link type’ suspension used on these wagons had a damping effect on hunting, but this diminished with wear – and the links on LA233 had worn. Examinations also found a slight twist in the track, which – while ‘within specified tolerances’ – created the conditions for derailment when combined with the increased tendency of LA233 to hunt (despite the fact that the train had been ‘properly driven at a speed not exceeding the laid down maximum of 45 mph’).

The UIC suspension had been used successfully on other wagon types both in Britain and on the Continent, but as McMullen admitted, the rapidity of wear – and the reasons it was so rapid on ‘Cemflo’ wagons – were ‘still not fully understood’. ‘Until this accident occurred,’ he went on, ‘the rate of wear [...] was regarded by all concerned as an economic problem rather than a safety one’. Previous accidents involving ‘Cemflos’ had been caused by broken springs, but checks for spring defects were well developed and well known. Suspension link wear, on the other hand, could not be seen without dismantling the suspension. McMullen therefore recommended reducing the speed of loaded ‘Cemflo’ wagons to 35 mph, though he recognised such a limit to be ‘highly restrictive from a traffic point of view and thus [could not] be regarded as a satisfactory long-term solution to the problem’. He also recommended running the wagons with their buffers in slight compression, in order to damp any possible hunting, and reported that tests had indicated ‘that there would probably be considerable advantages in fitting a single friction link suspension’ to all ‘Cemflo’ wagons, as this would provide a greater bearing area ‘to reduce the rate of wear’ and ‘ensure a close running fit to control the friction force’. The abrasive quality of cement powder doubtless exacerbated wear in wagons that carried it. The trouble was, hunting wasn’t just about cement wagons…

Indeed, freight train derailments had become an unnerving part of the Sixties railway in Britain. There had been 259 in 1966 alone, most involving the traditional short wheelbase wagon, most occurring on plain line. These track-train combinations had been a relatively trouble-free part of the railway for years, so what had changed? Arguably, the answer lay in modernisation. Modernisation had brought in more steel-framed wagons, which – while easier to maintain – intensified the hunting phenomenon, as they were less flexible than their older counterparts and therefore more prone to a build-up of oscillations after any imperfections in the track. Continuously welded rail could introduce even more rigidity, while its lack of joints meant fewer natural breaks to disrupt the effect. When these factors combined with the sustained – and more immediate – high speeds possible with diesel traction and continuous braking, incidents increased alarmingly.

For BR, accidents like Thirsk spurred research into the wheel-rail interface that made us safer, but also gave us the series 75-mph air-braked wagons that went on to launch the ‘Air Braked Network’ (rebranded ‘Speedlink’ in 1977). The knowledge gained was also applied to bogie vehicles, and materialised in the ‘Mark III’ coach, which was capable of 125 mph, and are still in use today. Less prevalent today, thankfully, is hunting, which ebbed as more and more short wheelbase wagons were withdrawn during the ‘70s. But that didn’t mean the end of freight train derailments. Indeed, those that came at Gloucester and Camden Road West Junction in 2013 focused minds – not because they both occurred on the same day (October 15), but because they raised a number of issues that had been building for some time…rather like they had in the 1960s. The difference now is that, while the relationship between track and train remains the crucial factor, the spotlight has turned to container wagons.

The incident at Gloucester occurred at 69 mph, and caused a container to fall, and the rear wagon, four miles of track, signalling cables, four level crossings and two bridges to be damaged. The track was suffering from cyclic top, whose dips had initially formed due to water flowing underneath the formation, while the susceptibility of the wagon (an IDA) to it – especially when loaded with the type of empty container it was carrying – had not been identified when the wagon was tested or approved for main line use.

Camden Road West, however, was caused by an unevenly loaded train running on track suffering from excessive twist. At about 02:40, one wagon came off and ran derailed until the train reached the junction, where an empty container toppled off and damaged the overhead line equipment. RAIB noted that the rules on the loading of such wagons (FEAs) had been relaxed following a derailment at Duddeston Junction in 2007, and wrote to the (then) Office of Rail Regulation (ORR) to point out that this would have the effect of increasing the risk associated with dynamic loading. But, again, the issue was not considered as part of the process for accepting FEAs for operation on the main line.

RAIB’s annual report for 2014 noted that ‘[m]any in the railway industry argue that as long as track condition, and wagon design and operation remain close to that in the past (sometimes referred to as the ‘historical norm’), the risk associated with wagon derailments will not vary a great deal from its current level’. However, it went on, the industry needed ‘to better understand how container wagons and their payloads are interacting with the type of track faults that might sometimes be encountered.’ Indeed, a number of RAIB investigations suggested reliance on the ‘historical norm’ to be no longer sufficient. Compared with the number of train kilometres run by freight trains every day, the risk was – and is – low. Nevertheless, it was decided to establish a ‘Cross-Industry Working Group on Freight Derailments’, including representatives from Network Rail, freight operators, RSSB, Interfleet, Huddersfield University, Lloyds Register Rail and the ORR, to enable rolling stock and infrastructure experts to consider all the issues together.

Facilitated by RSSB, the group is looking closely at derailment data, but will also review the origin of current requirements, with a view to ascertaining what could be done differently to manage the situation in light of both the risk and the changes the industry has seen during the last decade – changes like increases in bogie wagon usage and changes to the standards covering their structural strength (the latter having resulted in vehicles which ‘flex’ less and may therefore be less tolerant of track twist).

At the same time, there has been a 50% rise in container traffic, while the containers themselves are getting both taller (8’ to 9’ 6”) and longer (20’ to 40’, 45’ and 50’). These increases may have a bearing on stability, although two incidents involving containers being blown from freights (Scout Green 07/03/15, and Deeping St Nicholas 31/03/15) have also implicated the reliability of the retaining spigots used on trains of this type.

Of course, just as change was a factor fifty years ago, the importance of research is just as acute as it was back then. But that’s why work is under way to try to understand the extent and frequency of overloaded and unevenly loaded containers transported by rail, a wagon’s propensity to derail whilst negotiating twisted track, and the impact of lateral and longitudinal offset loadings on derailment risk.

I should probably add that offset loads have never really been looked at by the industry before, and on their own – a bit like CWR, rigid framed cement wagons and non-steam traction – they are not dangerous. But – in classic James Reason fashion – when they combine with track twist, we can have a problem.

The results of this work are likely to lead to a Standards change that will help ensure we keep tabs on this issue. And that – along with all the other workstreams coming out of the cross-industry group’s work will help us minimise the risks…because the rail industry doesn’t want to be complacent; because the industry doesn’t want another Thirsk.

Join or log in to keep reading
Need some help?
To talk to us about accessing RSSB content or corporate membership:
Cookies help us improve your website experience.
By using our website, you agree to our use of cookies.