Nigel Harris
“I started reading the farewell messages from the relatives of those who had gone to work, just like any other day, but who never came home. I can still recall the saddest messages word for word.”
Her Majesty’s Railway Inspectorate, founded in 1840 and empowered at George Stephenson’s instigation in 1841, celebrates its 180th birthday in 2020. NIGEL HARRIS mulls over editing A New Illustrated History of Her Majesty’s Railway Inspectorate from 1840, written by Her Majesty’s Chief Inspector of Railways IAN PROSSER CBE ( left) and former Deputy Chief Inspector DAVID KEAY ( right) and published by Steam World in December 2019
Having taken over as editor of RAIL in summer 1995, in the very next year I was plunged into the public, political and media mayhem of a major railway catastrophe with the November 1996 Channel Tunnel fire.
This was followed by what was reported in the wider media as a ‘spate’ of rail tragedies - first at Southall (1997), and then Ladbroke Grove (1999), Hatfield (2000) and Potters Bar (2002).
It was an appalling time. Rail privatisation was still very new and highly controversial, and in the face of strident criticism by its opponents, measured and objective explanation, assessment and lesson-learning became well-nigh impossible.
In those four catastrophes there were a total of 49 fatalities, including 31 at Ladbroke Grove when a Thames Turbo collided head-on with a High Speed Train bound for Paddington. A terrible fireball had raged through Coach H at the front of the HST, ignited by a spray of atomised diesel fuel from the leading power car’s ruptured fuel tank.
Seven people died at Southall after a London-bound HST driver passed a red signal and collided with a stone train crossing his path.
Four were killed at Hatfield when a GNER Leeds train derailed after rails long overdue for replacement shattered and the buffet car rolled on its side. Its roof was ripped open by a catenary mast on the Welham Curve.
Finally, seven lives were lost at Potters Bar when a down commuter train derailed at nearly 100mph on badly maintained points south of the station.
I reported on all of these accidents from the tracksides and will never forget the scenes I witnessed.
In October 1999, today’s Network Rail Head of Media Kevin Groves was a Railtrack press officer conducting me around the Ladbroke Grove crash site. The very public ‘shrine’ of flowers in the Sainsbury’s car park on the down side was shown repeatedly on news bulletins. What was not broadcast or published was a second floral tribute away from the cameras - near the fence on the down side, in the Eurostar depot alongside the wrecked trains, where the emergency teams were based.
“What are those, Kev?” I asked, pointing to the flowers.
He hesitated. “They’re flowers from the families of those who died on the trains,” he said quietly. “Come and look.”
I started reading the farewell messages from the relatives of those who had gone to work, just like any other day, but who never came home. I can still recall the saddest messages word for word - but I won’t quote them. The flowers were placed privately then, and I won’t breach that privacy now.
The most searing little cards, nestled amid the blooms, were the simple words of love, sadness and loss from victims’ children. They broke my heart. Kev and I stood shoulder to shoulder and our tears flowed.
The burned-out wreckage of Coach H was just a few metres away.
I shed plenty more tears (some of rage and frustration, as well as emotion) sitting through Lord Cullen’s public inquiry into this appalling tragedy, all of which played out over many months in the full glare of political and public condemnation, all intensified by 24-hour rolling news. Editing this book by Ian Prosser and David Keay brought it all flooding back.
At this time, Her Majesty’s Railway Inspectorate itself was controversially part of
The Inspectorate is 180 years old in 2020, and HMRI is now our second oldest great railway institution.
Only the British Transport Police as a national railway body can trace its roots back further.
the Health & Safety Executive and made its share of mistakes, all of which were exposed in the inquiry coverage. Ian Prosser and David Keay are to be congratulated for their candid telling of this part of the HMRI story.
I was reporting on wrecked trains again a year later, at the fatal derailment at Hatfield in October 2000. And again at Potters Bar in 2002, when contractor Jarvis, which had been maintaining the tracks, blamed a facing point failure on sabotage.
It was no such thing, of course, but this bizarre allegation ensured a further period of high-profile and understandable critical coverage for the railway. You’ll find all of this in the pages of this new book.
Notwithstanding these tragedies, privatised rail safety steadily improved from the days of BR, when the conventional experience had been a fatal passenger train accident every 18 months or so.
Collisions, crashes and derailments have, of course, been taking place for as long as railways have existed, and Prosser and Keay tell the story of developing (sometimes slowly) rail safety.
Passenger trains had not even completed their first day in service on the world’s first inter-city railway before the first fatality occurred - and a very high-profile loss of life it was, too.
Liverpool Member of Parliament William Huskisson was famously run over by Rocket, and within hours died from his injuries amid the razzamatazz of the opening day celebrations.
George Stephenson himself drove the locomotive Northumbrian to rush the grievously injured Huskisson (he lost a leg) to hospital in Manchester - but in vain, given that Huskisson died the same day. It is perhaps not surprising, therefore, that it was Stephenson himself who fought so consistently a decade
later for the creation of an empowered Railway Inspectorate to both inspect new railways and investigate accidents.
I have always been fascinated by the forensic rail accident investigation reports that followed. They are wonderful documents from which much can still be learned today - even those from the earliest days, with their archaic language.
This is an interest that took root for me in 1968 (I was 11 years old), just as my love of toy trains (clattering three-rail ‘OO’ gauge Hornby-Dublo and Hornby tinplate clockwork ‘O’ gauge) developed into a more mature interest in model railways, which in turn evolved when I was 16 into working as a volunteer (driver, eventually) on steam footplates on the Lakeside & Haverthwaite and Great Central Railways.
This hands-on restoration, maintenance and operating experience, incidentally, proved to be enormously useful in sharpening my understanding of rail mishaps - especially from the driver’s perspective.
This fascination was triggered on one of those seemingly interminable Sunday afternoons of the late 1960s/early 1970s, which those who lived through them may also remember with a sinking feeling!
No shops were open. Little moved. Every Sunday then felt like Christmas Day does now - everyday life came to a standstill. At least one of the only two principal TV channels always seemed to be screening an afternoon Gregory
Peck US Air Force Second World War movie.
There was little to look forward to after that, other than canned salmon sandwiches followed by tinned cling peaches (in syrup) and a can of Carnation evaporated milk at tea time, a black-and-white BBC Dickens dramatisation ( Great Expectations seemed very common), followed by Captain Pugwash and then Sing Something Simple on the radio after
Pick of the Pops with Alan ‘Fluff’ Freeman. It was then all downhill to school on Monday morning…
Nowadays, it’s hard to convey just how excruciatingly tedious these Sunday afternoons were. To escape this stultifying boredom, my dad and I would often jump in the family Ford Cortina and motor ‘over the tops’ (as Burnley parlance had it) into Yorkshire for an afternoon of steam on the Worth Valley Railway.
And it was there, in the souvenir shop at Haworth station in 1968, that it happened. My dad shelled out five shillings (only 25p now, but not insignificant back then for a casual treat) for my much prized - and now very tattered - Pan paperback copy of L.T.C. Rolt’s 1955 (and repeatedly reprinted) classic Red for
Danger - one of the finest railway books ever published.
I read it in less than a week and loved every word, even though I understood hardly any of its technicalities. Red for Danger had a profound impact. I was baffled by Absolute Block working, Automatic Train Control and continuous train brakes - but Rolt’s transformation of those formal accident reports into dramatic, beautifully crafted storytelling gripped me. I was hooked.
I’ve read Red for Danger numerous times since 1968, and each time my advancing railway technical, operational and human knowledge enables me to understand more about accidents whose names are tragically etched into our collective consciousness - Armagh, Tay Bridge, Quintinshill, Hawes Junction, Harrow & Wealdstone…
My pre-teen eyes widened not just in horror, but also in great sympathy (and empathy, although I wouldn’t have understood that as a concept at the time, as an 11-year-old) with Midland Railway Hawes Junction signalman Alfred Sutton, whose wretched mistake in forgetting a pair of light engines standing at his down main line signals caused the terrible 1910 disaster just north of Garsdale, in the high fells, in which 12 people died. They were burned to death in the blazing, splintered wreckage of wooden, gas-lit coaches.
When Sutton realised the full horror of what he had done, he turned to Driver George Tempest, who was in his box on the platform at Garsdale, and said: “Go to Bence [the stationmaster] and tell him I am afraid I have wrecked the Scotch express...”
Those words haunt me still. It is impossible to conceive how the poor, wretched man felt. But what Rolt did - in taking the formal, technical language of railway accident reports and turning them into such compelling prose - was to ensure that the human element was properly and powerfully captured and
Passenger trains had not even completed their first day in service on the world’s first inter-city railway before the first fatality occurred - and a very high-profile loss of life it was, too.
conveyed. Rolt made certain that a trainobsessed 11-year-old would ever after appreciate that human beings are as important as mechanical, system or operating failures. Rolt wisely and cleverly made sure that our interest in rail catastrophes wasn’t prurient.
Even so, The Times was cautious about enthusing about tragedy. Its review of Red
for Danger said: “At the risk of being thought callous, one must praise Mr Rolt for making death and disaster on the line most attractive reading.” The Listener really ‘got it’, too: “Intensely human story. A remarkable book.”
It certainly is. And this new history of Her Majesty’s Railway Inspectorate, by current HMRI Chief Inspector Ian Prosser CBE and retired Deputy Chief Inspector David Keay, honours the relatively small group of technical inspectors who produced those incredible accident reports. If you’ve never read one of these HMRI reports, you really should - their forensic examination of what happened and their skillfully drawn, measured conclusions make powerful reading. No wonder Rolt was inspired.
Dip into any one and you’ll be fascinated, but I’d recommend you look at the report by Lieutenant Colonel G.R.S. Wilson and Brigadier C.A. Langley into the fatal Doncaster derailment and bridge collision of 1951, which happened at just 25mph.
The detail of the investigation into every aspect of this incident - including repeated re-created test runs - is a classic example of these superb reports. You can find this report, and many others online, at The Railways Archive ( https://www.railwaysarchive.co.uk).
We all ask and debate ‘why’ and ‘what happened’ as we pore over rail accidents.
One of the first to appreciate these important questions was George Stephenson. He felt so strongly that accidents should be investigated, and lessons learned, that he wrote to the Board of Trade in 1841, urging rapid development of an investigative body to look into rail accidents in order to improve safety in this fast-growing new industry. Stephenson’s letter is reproduced in full in the book.
The Inspectorate is 180 years old in 2020, and HMRI is now our second oldest great railway institution. Only the British Transport Police as a national railway body can trace its roots back further - to 1826, the year after the Stockton & Darlington Railway opened.
Ian Prosser CBE is surely unique in occupying a position effectively proposed by George Stephenson himself! Who else in the railway industry can make this claim? The full story of how the Railway Inspectorate developed from 1840 is told with the considerable expertise born of decades of front line HMRI experience.
Prosser and Keay tell the RI’s story from investigating its first accident, on August 7 1840, three days before it was even formally established. Four passengers had been killed on the Hull & Selby Railway.
Not every accident since then can either be covered or illustrated in this book, so the authors have chosen those illustrating specific points where major catastrophes triggered major changes, or (with regard to the pictures) those which simply caught their eye. So, plenty of small and lesser-known mishaps are also covered alongside ‘the big ones’ in words, photographs, track plans and diagrams.
Some broad common themes run through nearly two centuries of accident investigation history in these pages. The three principal causes of death and disaster on the railway were:
Mistakes by drivers.
Errors by signalmen.
Excessive speed.
The biggest problem for drivers was overlooking signals - especially crucial distants at caution, warning that the next stop signal was likely to be at danger. Drivers then came horribly to grief when it became impossible to stop before either colliding with another train or derailing because of excessive speed.
Technology to support driver vigilance was slow to arrive in Britain - other than on the Great Western Railway. The GWR’s Automatic Train Control (ATC) of 1906 was a primitive but highly effective precursor to the later BR Automatic Warning System (AWS) and then, later still, the privatised railway’s Train Protection Warning System (TPWS).
Evidence for the life-saving success of
GWR ATC is unequivocal and unarguable. Between 1923 and 1947 (excluding the war years), there were just two fatal GWR train accidents with a single passenger killed in each - an unparalleled safety record, largely as a consequence of ATC.
Compare this with the series of fatal accidents on and around the LMS West
Coast Main Line, which relied on driver vigilance alone. These included: Carlisle Canal ( January 3 1931); Leighton Buzzard (March 22 1931); Bletchley (October 13 1939); Bourne End (September 30 1945); and Polesworth (November 19 1951). There were 56 deaths (and 227 injuries) in these five accidents alone, all of which could probably have been prevented by some form of GWR-style ATC.
That said, it would be wrong to suggest that the LMS had ‘sat on its hands’ and done nothing. It had actively sought improvement via widespread installation of track circuits and block signalling control, as well as by replacing difficult-to-see (especially during bad weather and at night) semaphore signals lit by paraffin with brighter, electric colour light signals.
While this was indeed an improvement, accidents such as those mentioned earlier at Bourne End (1945) and Polesworth (1951) still proved that without the further support of some kind of ATC, colour lights alone provided only marginal and inconsistent improvement. Even with colour light signalling, safe running still hung on the fragile thread of driver vigilance.
This very slow progress to assist drivers is surprising, given that (in parallel) the proportion of accidents resulting from signalling errors was in steady decline, owing to increasingly common track circuits and other applied technology.
Track circuits made it impossible for the presence of a train on his ‘patch’ not to be clearly visible to the signalman - either by lights on his panel, or by instruments on the block shelf at eye level, above his levers.
In conjunction with the old faithful lever collar, and further protection by Rule 55 (which required firemen to visit the signal box and remind the signalman of his train’s presence after a couple of minutes standing at a stop signal), Hawes Junction (1910)-style accidents declined steadily as a direct result.
Of 29 fatal crashes between 1969 and 1989, only three were a consequence of signalling errors. Drivers, however, were responsible for 15 accidents in the same period.
This disparity of management approach is hard to fathom. It is perplexing that drivers remained the ‘poor relation’ in terms of assistance by technology.
As recently as December 4 1984, there was a fatal collision in Manchester (at Eccles), when a locomotive-hauled Liverpool-Scarborough express collided at speed with the rear of an oil tanker train - on a route with no AWS.
This was fully 32 years after Harrow (1952), when 112 people died and 340 were injured in
Rolt made certain that a train-obsessed 11-year-old would ever after appreciate that human beings are as important as mechanical, system or operating failures.
Ian Prosser CBE is surely unique in occupying a position effectively proposed by George Stephenson himself! Who else in the railway industry can make this claim?
a three-train pile-up which would have been prevented by AWS.
It is commonly claimed that Harrow was the spur to widespread fitting of AWS. Well, maybe it played a part, but as Eccles proved it was a slow-burn fitment programme. The 1984 fatal collision was on the route of the world’s first inter-city railway (the Liverpool & Manchester), and yet fully 78 years after the GWR introduced its pioneering ATC, the Liverpool-Manchester main line was still unprotected. These facts tell me that the 1952 Harrow three-train collision was nowhere near as influential in encouraging nationwide AWS installation as has been argued.
Likewise gas lighting. Even 38 years after the terrible deaths by fire in gas-lit wooden coaches at Hawes Junction proved the horrific scale of the risk, fully one in ten of our coaches - more than 4,000 vehicles - were still gas-lit at Nationalisation in 1948.
This was despite the report by Colonel Pringle into the Hawes Junction disaster of 1910 - he recommended that our railways should be using electricity and not gas to light passenger coaches. Yet nearly two decades later, in 1928, roughly half the LMS and LNER passenger fleets were still gas-lit, while on the SR it was one-third. The GWR, so significantly ahead in encouraging safety for drivers by providing ATC, lagged even more severely - around two-thirds of its carriages were still gas-lit as 1930 approached.
ATC was still the exception rather than the rule, by the way, other than on the GWR. Safe running everywhere other than Swindon’s empire continued to rely 100% on driver vigilance.
Why did progress in driver assistance via ATC lag so far behind the steady improvements implemented to assist signalmen?
Maybe it was a hangover from the bizarre corporate belief of the 19th/early 20th century that effective, cheap measures such as wooden lever collars (costing mere pennies) were not a good idea.
These collars were a sturdy, bored wooden block which could be dropped over a signal lever, obstructing the release catch. They were a simple but unmissable physical reminder to a busy signalman that a train was standing at the signal controlled by that lever, which therefore couldn’t be pulled without lifting the collar off.
Had Signalman Sutton had a lever collar at Hawes Junction on that dark and fateful Christmas Eve morning in 1910, he would probably not have forgotten the two light engines standing at his Down main line home signals.
At this time, railway managers regarded lever collars as undermining concentration and encouraging complacency - they perversely believed that lever collars made it too easy for signalmen not to pay attention.
It almost beggars belief now, but this cannot have been an objection rooted in cost - they really did believe that lever collars would compromise and not enhance safety!
Was the widespread slowness to adopt ATC another example of this fear of ‘mollycoddling’ drivers, who might thereby be encouraged to take their eye off the ball?
If so, the human cost of this misjudgement was appalling. Of 1,093 train deaths in accidents between 1912 and 1952, 28% might have been prevented by ATC. That would have been 300 lives saved.
Another obstacle to safety investment was that railways actually had a pretty good general safety record, with only the occasional bad year.
In 1842, there were 20 million journeys made but only one passenger death. Between 1916 and 1922, there were only eight passenger deaths a year on average - even in wartime. There were no passenger deaths at all in
1925, leading even the RI’s Sir John Pringle to comment that the case for ATC was “less urgent and that the expenditure involved could not be justified”. Incidentally, history repeated itself in the mid-1990s, when Her Majesty’s Railway Inspectorate came to precisely the same public conclusion about Automatic Train Protection.
Only a single passenger was killed on a train in 1930, with just 17 more in the next three years - to set this in context, this was just over half of the death toll in the single terrible collision at Ladbroke Grove on October 5 1999.
Did this relatively good safety record lull government and railway into a false sense of security? It certainly made it very difficult to justify heavy capital investment on ATC. And it was especially the case in the 1920s/30s, when railways were hit hard by massive (and unfair) competition from road haulage, plus a world economic slump - and with aviation making its first inroads into the top-end of the express passenger market.
But there were also some very bad years. In 1928, for example, 48 passengers were killed in the worst peacetime year since 1906. And in 1947, the worst ever peacetime year for rail safety thus far, there were 88 passenger deaths.
The safety investment argument has often been muddied and politicised, with the accusation (sometimes still levelled today) that the pre-1948 private companies focused on profits rather than safety. This argument collapses after 1948, when the Labour Government which nationalised our railways nevertheless failed to implement ATC nationwide and simply perpetuated private company policy.
In that worst ever year thus far for peacetime rail safety (1947, when 88 lives were lost), no one died on the GWR, where ATC had protected drivers for many years. It could not have been more obvious that ATC was urgently needed - yet all the new Labour Government did was make vague promises of reviews and ‘consideration’. It kicked the can down the road, as did subsequent governments.
As late as 1986 (two years after the fatal collision at Eccles, where there was no AWS), the Railway Inspectorate’s Major Freddie Rose was still urging widespread installation of
AWS (as ATC had evolved into) in his annual report.
It is important to understand the incredibly challenging context in which the pioneering Railway Inspectors of the 19th century established their new role. The network was extended with breathtaking speed in the
1840s, when more than 250,000 navvies were furiously building our railways:
1841: 1,556 miles of railway constructed 1845: 2,746 miles
1850: 6,621 miles
1890: 20,073 miles
1914: 23,701 miles
In the 19th century, Britain’s railway companies were among the world’s largest and most complex commercial concerns on the planet - consider them as having a similar transformative impact back then as Google and Facebook have had in the 21st century.
According to respected railway historian Terry Gourvish, the London & North
Western Railway was capitalised at more than £ 29 million as early as 1851. By 1894, that capitalisation had risen to £111,457,620, according to Bradshaw’s Railway Manual that year. Of that total, £ 81,123,233 was in stocks and shares with the remainder in loans. Comparable figures were £ 82,726,271 for the GWR, £101,593,763 for the MR and £ 64,430.930 for the NER.
These enormous figures for the day give an indication of the major corporate power which the RI’s handful of army engineer inspectors were ranged against. Britain’s rail industry continued to grow rapidly in size, impact, power and complexity. By the time of the Grouping in 1923, there were more than 24,000 locomotives, 51,500 carriages, 722,639 goods wagons and nearly 700,000 staff - including 71,400 drivers and firemen and 24,000 guards.
Part of the Office of Rail and Road since 2006, HMRI today is very different to the RI of the 19th/20th centuries, and its own history very much mirrors that of the industry it has regulated for almost two centuries. This new book is offered by its authors in gratitude, respect and tribute to successive generations of railway inspectors who strove to understand where it went wrong and how safety could be improved as a consequence.
Of 29 fatal crashes between 1969 and 1989, only three were a consequence of signalling errors. Drivers, however, were responsible for 15 accidents in the same period.