Nigel Harris examines the causes and effects of the Carmont crash.
13 years, five months, 18 days since a passenger died on a train
“In the finest railway tradition, off-duty Conductor Nicola Whyte… climbed from the wreckage onto the track and walked for over a mile, to raise the alarm at Carmont signal box.”
AUGUST 12 2020: a beautiful summer’s day in Leicestershire, near Melton Mowbray. I was at the 10.25in gauge Stapleford Miniature Railway (venue for the Institution of Mechanical Engineers’ annual ‘Railway Challenge’) when I heard the appalling news.
Staring at his phone, Railfilms’ Nick Dodson called to me: “Nigel - are you aware that there’s been a fatal HST crash in Scotland…?”
It had been 13 years, five months and 18 days since 84-year-old Margaret Masson died on February 23 2007, in the 95mph Virgin Pendolino derailment at Grayrigg (Cumbria).
Within hours, it was confirmed that Driver Brett McCullough, Conductor Donald Dinnie and passenger Christopher Stuchbury all died in the wreckage of the two power cars and four Mk 3 carriages running from Aberdeen.
Full details are in our 16 pages of coverage - but, in summary, the ‘2+4’ refurbished Inter7City set ran into a landslip in extreme weather, derailed, veered to its left and stripped away a viaduct parapet before three carriages piled up on the double-track main line. One carriage slid down the bank. The leading power car, its cab torn off, rolled left and came to rest on its roof on a steep, wooded slope, where it caught fire.
In the finest railway tradition, off-duty Conductor Nicola Whyte, who was ‘on the cushions’ en route to her next shift, climbed from the wreckage onto the track and walked for over a mile, to raise the alarm at Carmont signal box. With Driver McCullough and Conductor Dinnie both killed, a member of the public, spotting billowing smoke, called Police Scotland. It was by these means that the massive emergency response was triggered at 0943. See pages 6-13 for News, 14-15 for Analysis (Gareth Dennis on earthworks and climate change), 50-51 (Wolmar) and Philip Haigh (56-57)
Torrential rain that had already caused chaos across Scotland was immediately suspected of playing a part, given that the 0638 ex-Aberdeen had been reversed and was en route back to the Granite City, having been alerted by Carmont’s signaller to a landslip between Drumlithie and Carmont.
The high-energy wreckage suggests that the HST was running at (or approaching) the line speed of 75mph, when it hit a landslide at a location which had been clear just two hours earlier, when the HST had passed without incident, running south en route to Glasgow.
RAIL understands that Network Rail Scotland was compliant with all procedures required by the official weather warnings in place at the time. This was repeatedly confirmed by Scotland’s Cabinet Secretary for Transport Michael Matheson in broadcast media interviews, following the accident.
Investigators will inevitably focus sharply on these official weather warnings, which trigger NR’s crucial safety procedures, to assess ‘fitness for purpose’ and any need for revision. These are the same formal weather warnings which dictate NR operational decisions throughout the year, to decide speed restriction/train cancellation responses to (for example) excessive winds or very high railhead temperatures that might cause tracks to buckle.
This is a really difficult balance to strike: not stringent enough and disaster follows; too stringent and criticisms of ‘throwing in the towel after a bit of wind and rain’ follow critical public, politicians and media.
In a challenging Channel 4 News Live interview on the evening of the accident, I was questioned about the Office of Rail and Road’s recent warnings about the need for NR to ramp up earthworks monitoring/maintenance in response to threats posed by climate change and extreme weather ( RAIL 910). An implication that NR is dragging its heels or not doing enough hung in the air. This derailment is obviously proof enough that there’s always more that can be done - but this debate needs to be had in a measured and nuanced manner, for it is far from straightforward.
For instance, NR can only spend on earthworks what the ORR’s economists will allow it to spend. £750 million was allowed by ORR economists in Control Period 5 (CP5, 2014-19), and when NR asked for more for CP6 (2019-24) ORR allowed £900m. If ORR wants NR to do and spend more, then it must revisit its decision, because CP6’s £900m will allow NR to focus on just 4% of its assets.
A good start was made some years ago - there was a doubling of work on earthworks in CP4, while drainage renewal and maintenance nearly quadrupled. Awareness of, and action on, this problem is not new, but Government and ORR are going to have to play their part - not only by making more money available, but also by giving NR ‘air cover’ for some inevitably disruptive works.
NR assesses safety risk on its infrastructure from A to E, with E being the most vulnerable to failure. Bearing in mind that the Carmont landslip site was only moderately rated at B/C, it is crucial that this fatal landslip is very clearly understood, given the implications for NR’s Risk & Analysis model as applied to the rest of the network. Existing processes neither detected nor predicted any potential problems at Carmont.
We are seeing sudden, widespread bursts of torrential rain of maybe 40mm in a day (maybe a month’s rain in just 24 hours), with consequent and unavoidable violent flash floods which overwhelm Victorian drains that were designed and maintained for an earlier age of much more temperate weather. Extremes and new records are now frequent - February 2020 was the wettest UK February on record while May was the sunniest and driest.
It is also important to recall that in the last two decades of BR stewardship and at least the first decade of the post-Railways Act era, drain maintenance was often seriously neglected by financial planners.
It was an easy and (at the time) physically invisible cost saving, but the legacy of decades of neglect, when combined with heavier traffic and extreme weather, have created an enormous problem.
Mainland Britain is nearly 900 miles as measured from John O’ Groats to Lands End. We have a rail network of 11,000 miles. That’s 22,000 miles of trackside infrastructure - the equivalent of 24 British mainlands laid end to end. There is no quick fix. This is going to be a long slog over many years to make our pick and shovel-built Victorian infrastructure properly resilient for the heavy traffic and extreme weather of the 21st century.
And there is no room here for political or media game-playing. This is a huge job and it’s going to be difficult enough without such timewasting distractions.
Remember also that while NR renewal
“Government and ORR are going to have to play their part - not only by making more money available, but also by giving NR air cover for some inevitably disruptive works.”
budgets were deferred (reduced!) in England in CP5, to get spending under control, there were no such restrictions in Scotland, where maintenance and renewal spending was not restricted in any way. And yet this disaster still happened - a cruel irony.
This should prompt serious consideration by government south of the border, where there have also been plenty of landslips in 2020. We average 100 such incidents a year. It was around 60 a decade ago. In recent years, the figure has been as high as 140. The problem is evidently worsening - with its attendant threat to safety.
It’s significant that at the accident site we saw both Westminster and Holyrood politicians passing comment, because constitutionally this is complex. ORR/RAIB report to the English Secretary of State for Transport with regard to their accident, safety and potential prosecution investigations, but any consequent legal proceedings will play out under Scottish law, rather than in English courts.
As usual, RAIB/HMRI reports will go to the SoS in London in due course - but, as I understand it, ORR will be required to separately submit any proposals for prosecution to the Office of Scotland’s Procurator Fiscal.
This divided accountability will be constitutionally and politically tricky, and will need sensitive management. Neither Westminster nor Holyrood ministers will find themselves to be as fully ‘in charge’ as they might have believed themselves to be!
In the aftermath of the Carmont derailment, there were the usual spate of social media and lineside red herrings on the one hand, accompanied by some truly appalling national news reporting on the other.
There was some social media gossip about significant Scottish Government investment on the A9 road to the north, while the railway retains ‘Victorian’ Absolute Block operation controlled by traditional semaphore signalling. Given that Scotland spends £1 billion or so a year on its railways, there’s simply no case that we’re seeing under-investment in rail - probably quite the reverse, if you look at spend per passenger, compared with road users. Also, it doesn’t look to me as if signalling played any major role in the accident, although I’ll leave that one to the Rail Accident Investigation Branch and HMRI investigators.
As is often the case with rail accidents, national media coverage was patchy, with islands of excellence in a generally woeful sea of very poor-quality reporting. The BBC (both in London and, sadly and surprisingly, in Scotland) was in the joint lead with Channel 4 News in plumbing the depths of dismal, dire reporting.
BBC Radio 4 Today’s flagship 0800 bulletin on August 13 told us: “The driver is thought to have been trying to use an alternative track.”
BBC Scotland’s live news page claimed that “…the train reversed and switched tracks… and apparently continued on its route to Glasgow”. No, it didnt - utter claptrap.
Also on August 13, a BBC Scotland reporter solemnly ‘informed’ viewers: “…we understand that at one point the driver had stopped his train and requested to be switched to another track, before continuing his journey, with catastrophic consequences…”
Unbelievably ignorant - simply dreadful reporting. And publicly funded, too!
The frequently rather arrogant ‘now hear this’ tone of misplaced authority and superior knowledge was also as irksome as the crass inaccuracy of the reporting itself.
Channel 4 was guilty of both poor tone and ill-informed editorial approach. I was shocked (disgusted, actually) to have to answer in a live interview that the reason for any perceived delay in reporting the derailment was because the train crew lay dead in the wreckage.
C4 was bizarrely obsessed about the alleged ‘missing hours’ - as if the railway had dragged its feet. There was a slight delay in sounding the alarm for the reason I’ve just outlined - the crew were dead - but C4 seemed incapable of understanding that a slight delay in routeing the train back north after it had been turned back, but before the derailment, was not incompetence. It was because a Network Rail mobile manager had to travel by road and foot to the train to physically clip the points (a crucial safety measure) to enable the HST to run safely in the wrong direction over facing points, to regain the Down road a mile or so south of the accident site.
This considerable degree of technical ignorance in a general editorial office is understandable. The refusal or inability to find out the ‘who, what, where, why, when and how’ basics of a news story, however, was shameful. It was often embarrassing to either listen to or watch this abysmal output.
Some newspapers were even worse. The most disgraceful and distasteful example was the Scottish Sun’s crass front page with its ‘Death Express’ headline, alongside a picture of the late Driver McCullough. It was ill-judged, callous and insensitive. The editor paid a heavy price - he was subsequently compelled to offer a grovelling and humiliating personal apology, following an angry tsunami of protest.
A ray of light amid all this nonsense was Gwyn Topham’s piece in the Guardian, which was timely, measured, accurate and of appropriate tone. That this single report stood out so clearly as an exemplar is a scathing comment in itself on the volumes of drivel surrounding it.
A final important point about landscape management and the challenges ahead. Over the past few years, we have seen Government and ministers either dithering or prevaricating with regard to a fundamental aspect of trackside management on which we must now establish clear policy - lineside trees. The time for Governmental fence-sitting is over.
There are many reasons why there should generally be no large trees between the boundary fences. Signal sighting problems and the annual delays, slow running and operational chaos caused by deciduous leaf fall are just a few. Frequent disruption, inconvenience and high economic cost when they crash onto the tracks are others.
But the damage mature trees do to earthworks also plays a key role in creating the instability issues we must now tackle with yet greater urgency. You often hear claims that tree roots bind earthworks together and enhance stability.
This is only true with regard to grasses and low shrubs - it is quite the opposite for large trees, whose root activity not only makes earthworks unstable but can also distort track alignments as well - yet another potential cause of derailment.
It’s now time to be really clear about trees - there’s a reason why you don’t see them at airports and on motorway verges, and the same should apply to railways. It has applied to HS1 from Day One - there are no trees between its boundary fences.
It is now time for Government to stop running scared. This is about safety. Time for Government to be bold and ‘front up’. ORR’s excellent safety regulator in the shape of Her Majesty‘s Railway Inspectorate has already provided all the ‘air cover’ NR needs to firmly resolve this issue and Government must now do likewise.
Sure, RAIB needs to get its investigation out as soon as possible, but that need not stop Government ‘rolling the pitch’ in terms of both strategy, messaging and investment to allow NR to crack on as rapidly as it can on improving earthwork resilience.
In this issue, engineers Gareth Dennis and Philip Haigh both look at how technology might be applied to monitor earthworks.
NR is already ‘on the case’ here - its man in Inverness was trained some time ago (at his request, incidentally - good man!) in drone monitoring, which enables much longer aerial patrols than are possible on foot.
We need much more of this. Notwithstanding this new technology, we already know a great deal about what needs to be done with our earthworks - but the railway needs the extra resource to do the work. Government must now give the railway the tools it needs to accelerate progress - urgently.