How the Ufton Nervet incident changed level crossing safety measures

Greg Morse considers the causes and consequences of this tragic rail crash from November 2004.

Warning: this article contains references to acts of suicide. 

In this article:

• An off-duty officer attempted to prevent a tragic train crash in 2004 but couldn’t save the trapped driver.
• Seven lives were lost, sparking inquiries and safety improvements at automatic level crossings across the UK.
• Network Rail continues enhancing crossing safety, while initiatives address suicide prevention on railway lines.

Off duty. Glorious words to all who serve, whoever they serve. Off duty means freedom - to go out, stay in, mow the lawn, anything.

But is anyone who serves ever really off duty? Take a police constable, for example. And take Saturday November 6 2004.

The constable was driving down a country lane near Ufton Nervet, in Berkshire. A pleasant pursuit, about which books have been written and television programmes made. Today was different. Today, there was a car on a level crossing, three vehicles ahead. A silver-grey Mazda.

The officer watched as the warning lights flashed and the barriers came down around the vehicle. He flashed his headlights. He jumped from his car. He “scream[ed] at the male motorist to run away,” as The Times later described it.

But the man in the car stayed put. The officer rattled the car doors, desperately trying to open them. They would not yield. He ran to the emergency telephone at the crossing, but it was too late. A train was coming.

On that train, the 1735 Paddington-Plymouth, many were coming back from Christmas shopping in the capital. Among them were Jonathan Stace (21), a student from London, Ian Horler (37) from Newbury, and Brian Knapman (64), who was travelling home with his wife.

The train, a former British Rail HST formation, had left London on time and made a single stop at Reading, which it left one minute late at 1803. It was running just under the line speed of 100mph when it passed the ‘strike-in’ point for Ufton level crossing eight minutes later. It was this that started the closure sequence witnessed by the officer.

The train passed under the Tyle Mill Road overbridge. Its Exeter-based driver, Stan Martin, who had joined the railway in 1965, who had been a driver since 1974, who had seen the HSTs’ introduction, also saw the car up ahead.

He applied the emergency brake, but the collision still came, lifting the leading wheelset of the leading bogie of the leading power car off the track.

The train remained upright for around 95 metres, until it struck the facing points leading to the Down Goods Loop. This took the bogie to the left, dragging those behind, leaving carriages scattered and lives lost.

There would be seven fatalities in all: the driver of the car, the driver of the train and five rail passengers, including an eight-year-old girl.

Stace told the press how, just after 1810, “the lights went out”, how they “spun over”, how he “could feel bodies going on top of [him] and being thrown in different directions”.

Horler told of how a Royal Marine called Tom had saved his wife’s life, while Knapman described feeling his arm “being scraped along the tracks” before the train came to a stand.

A total of 71 people would be taken to hospital with injuries, among them Knapman’s wife, who suffered a broken jaw and wrist. Some used the light from their mobile phones, others novelty glow sticks, to help guide them to safety.

The pager of the area’s mobile operations manager went off at 1832. By the time he got to the scene at 1846, the emergency services were already there. Soon there would be 180 police officers, 84 fire crew and 50 ambulance crew in attendance, all intent on rescuing, helping and treating.

Deputy Chief Constable Andy Trotter of the British Transport Police was full of praise for his off-duty Thames Valley colleague, who had done “absolutely the right thing”.

The constable, who had seen the full horror of the accident unfold before his eyes, received counselling. Survivors and bereaved families and friends struggled to come to terms with what went wrong, struck with the inner turmoil of the way each hazard had conspired to cause harm and loss: the train, the crossing and the actions of the car driver.

As to why the incident had happened, AEA Technology Rail and members of the Health and Safety Executive attended to investigate the causes. However, it fell to the Rail Safety and Standards Board (RSSB) to conduct the formal inquiry on behalf of the industry. By June 2005, the report of that inquiry had been completed.

Switches and crossings

The crossing at Ufton was an automatic half-barrier (AHB).

Under the Railways Clauses Consolidation Act of 1845, the railway was legally obliged to maintain attendance at all public crossings.

However, it proved to be expensive (and increasingly difficult), particularly after the Second World War, when near full employment made it hard for the railway to find staff for what was a responsible but poorly paid and often dull job.

Furthermore, by 1955, there were more than three million cars on British roads, and the number was rising. Many crossings took time to operate, caused heavy delays, and offered insufficient protection for such an increase in usage.

Clearly, the situation was only going to get worse. Something had to be done, and the answer seemed to lie on the continent.

The Railway Inspectorate (RI), along with representatives from BR and the Ministry of Transport, reviewed European automatic level crossings in detail the following year. The group’s findings led to the creation of the attendant-free AHB, where half-barriers block oncoming traffic but leave the exits clear for trapped occupants.

At the time of Ufton, the typical automatic crossing closure sequence, once an approaching train had activated the treadle, was around 40 seconds, compared with 315 seconds for a staffed crossing or 217 seconds for a crossing controlled via CCTV.

Although there was some initial opposition to the design, which many feared was unsafe, the first AHB came into use at Spath (near Uttoxeter), on February 5 1961. By the end of 1967, there were 207 AHBs, with plans for many more.

Then, on January 6 1968, an express passenger train struck a colossal road transporter carrying a 120-ton transformer at Hixon AHB, on the West Coast Main Line. Eleven people were killed, 120 yards of track were damaged, and the overhead power lines were brought down.

Realising that the RI’s involvement in the AHB’s development compromised its independence, the government chose to hold a Court of Inquiry, appointing Mr E B Gibbens QC as chair.

In the first such inquiry since Tay Bridge in 1879, the Court found that the haulage company had failed to inform BR that it intended to take the transporter over the crossing, which in turn prevented BR from taking appropriate precautions.

The accompanying report also highlighted poor communications between railway and police, and railway and haulier, about telephoning the signaller when large, slow-moving vehicles were being routed. It cited inadequate signs and poor police training as part of the full causal chain.

However, the ‘origin of the accident’ was traced back to “the failure of officers of both the Ministry and British Railways in collaboration to appreciate the measures necessary to deal with a hazard of which they were aware”.

Yet despite finding AHBs to be “a valuable answer to the needs of modern transport” that were “reasonably safe”, the management failings the inquiry unearthed led it to recommend additional safety measures, whose cost stalled the automation policy severely.

Indeed, although better signage and warning lights (known as ‘wigwags’) were introduced from 1969, little progress would be made until 1977, when another working party visited Europe and agreed to relax the recommendation requirements.

This made it slightly easier to create an AHB crossing. Ufton became one on September 19 that year. This would stand it in good stead for the resignalling and acceleration of services that were only a few years away.

Into the new century, Hixon remained the only AHB accident comparable to Ufton in terms of severity. With this and Network Rail’s case-by-case crossing risk assessment strategy in mind, RSSB’s panel did not consider the accident justified the “complete abolition” of all AHBs.

And, as it rightly pointed out, closing them would only move the risk from someone deliberately parking a car across the railway to a user-worked crossing, of which there were over 4,000 at the time.

That said, Network Rail’s preferred first option to reduce level crossing risk at any particular crossing was closure. It kept a list of those where opportunities existed for removal, but Ufton was not on it, even though RSSB’s report pointed out that the crossing was on a minor road and that alternative access to the A4 was “within a reasonable distance”.

While the car had been able to access the line at the crossing, and arguably would have done even if the half-barriers had already been closed, what worsened the consequences of the collision for those on the train were the facing points to the Down Goods Loop, which was frequently used by the many aggregate trains that worked between the Mendips and the capital.

Consideration had been given to shortening the loop, thereby placing the points further away from Ufton crossing, but it was realised that a double-headed aggregates train occupied the full length of the loop as it was. Any change to this situation would “constitute a significant impediment to traffic”.

Most of the passenger services on the route were HSTs like the one involved in the accident. Attention now fell on the design of those trains and their performance in train accidents.

Crashworthiness

Ufton was the third major accident to involve the erstwhile InterCity 125s, the other two being Southall (1997, RAIL 835) and Ladbroke Grove (1999, RAIL 889, 1019).

The story of these trains is well known (RAIL 806), but in brief, the production sets had been in service since 1976, when they had accelerated services on the Great Western Main Line, before doing the same on the East Coast Main Line two years later.

By 1979 they were working from Paddington, through Ufton, and on down to Exeter, Plymouth and Penzance.

In safety terms, the Mk 3 carriages that made up the HST featured a ‘monocoque’ construction, in which the all-welded mild steel stressed bodyside combines with the roof and chassis to form a strong steel tube. This made them more crashworthy than their Mk 2 predecessors, which were in turn more crashworthy than the Mk 1s.

That there had been no passenger fatalities in the Colwich accident of 1986 was in part down to the performance of the Mk 3s involved.

In its report on Ufton, RSSB acknowledged that while the Mk 3 did “not meet all the current loading requirements”, it had indeed proven to be “very resistant to deformation” in accident situations.

However, it also noted that two of the train’s carriages lost survival space because of the impact from a detached bogie.

RSSB’s research project T118 (Whole train dynamic behaviour in collisions and improving crashworthiness), already under way at this point, concluded that to manage the risk most effectively, no changes to the current bogie attachment strength requirements should be proposed.

Some of the fatalities at Ufton involved passengers being ejected through broken windows. The inquiry recommended that laminated glass - mandatory for new trains - should be fitted to existing vehicles when refurbished.

At the same time, research into the use of seat belts found that the forces at play when a train accident occurs would in fact lead to more harm, not less, if they were fitted.

These concepts led to the establishment of a ‘containment’ strategy, with the Ufton findings feeding into ongoing research that confirmed that it was safer to keep people on board a train when a collision or derailment occurred.

Yet research also found it important for the industry to have a clear and consistent approach to escape, as passengers taking the instinctive, human step of breaking windows in a bid to flee the scene of an accident would put themselves (and others) at risk.

As a result, it was recommended in 2007 that all hammers provided for breaking windows be removed, even on vehicles with breakable windows. This has largely been implemented, and today all new rolling stock features fully laminated windows, with most existing vehicles having been subject to a progressive replacement plan to match them.

As to the power cars, the report recommended the retrofitting of obstacle deflectors. While heavy metal brackets fitted immediately in front of the leading wheels of a train (known as lifeguards) had been around since the mid-1830s, these later deflectors were introduced by BR after the Polmont collision of 1984. Both serve to help mitigate the effects of obstacles on the track and reduce the likelihood of post-collision derailment.

Research project T189 (Optimal design and deployment of obstacle deflectors and lifeguards, 2003) had concluded that while obstacle deflectors and lifeguards should continue to be fitted on the leading ends of new rolling stock, retrofitting the former to HST power cars was not “reasonably practicable”, analysis having found there to be a very high cost to potential safety:benefit ratio (ranging from 24:1 to 166:1).

But what of the driver? The leading power car had come to rest having slid along the ballast. This led to a structural failure at the top of the left-hand pillar of the cab door, which was found to be missing after the accident.

Investigators also found “clear signs on the cab bulkhead that large quantities of earth and ballast had entered the cab through the door aperture”. Martin had been killed as a result of this ingress.

The structural integrity of driving cabs and the way their doors open was considered in project T190 (Optimising driving cab design for driver protection in a collision).

A specific supplementary report on debris ingress confirmed that doors on drivers’ cabs should have their hinges at the front - the standards for cab integrity have now been established in GM/RT2161 (Requirements for driving cabs of railway vehicles). The findings were also carried forward to GM/GN2686 (Guidance on bodyshell, bogie and suspension elements), published in 2010.

RSSB noted that the knuckle couplers on the HST had not held the train together after impact. There had been no pre-existing flaws in their manufacture - their design simply allowed them to come apart or fracture when the train struck the car and rode up. It recommended that project T118 be brought forward “to promote the earliest possible implementation of any safety measures” identified.

Unsurprisingly, T118 concluded that coupler strength has “a significant role in the crashworthiness of a rail vehicle”.

While a high axial coupler force can be useful as an energy absorption mechanism in low-speed collisions, at high speeds the coupler can generate sufficient yawing and pitching movements to result in derailment. It said consideration “must be given to the conflicting requirements of protecting vehicles in low-speed accidents and preventing derailment at higher speeds”.

While HSTs would not be retrofitted, this would go on to be embedded into standards and therefore into new designs, such as the Pendolino.

As the Rail Accident Investigation Branch (RAIB) noted in its report on Grayrigg (RAIL 820), the ‘390’ involved in that accident “avoided, almost completely, a number of hazards” in terms of its anti-roll bar links ensuring that most of the bogies remained attached, its couplers generally holding the carriages together and its bodyshell helping resist penetration.

The inquest

Three weeks before RSSB’s inquiry report was released, on June 1 2005, it was announced that an inquest into the accident would be held at Windsor Guildhall.

The proceedings, due to open that October, were delayed because of a dispute over whether the families of the victims should be given legal aid. The inquest finally began in October 2007.

The jury heard the testimony of the witnessing police officer, who reiterated that the car driver did not appear to be acting with any urgency. A forensic investigator determined that the car had been parked on the crossing with its engine and lights switched off. Its fuel tank still contained petrol, which suggested that the car had not broken down.

The car driver was Brian Drysdale, a 48-year-old chef employed at Wokefield Park, around seven miles from Ufton Nervet.

On November 6 2004, he left work early at 1730, saying he was ill. He drove to the level crossing, possibly influenced by a documentary on the Great Heck accident (RAIL 925) which, according to his workmate, he had enjoyed watching a couple of weeks before.

No one can really imagine another’s pain. We think we can but it is difficult, if not impossible, for some of us. On November 1 2007, the inquest returned the verdict that the crash was caused by Drysdale’s suicide and that Driver Martin and the five passengers had been unlawfully killed.

The coroner said “a unique set of circumstances […] had resulted in catastrophic consequences”. Arthur Miller once wrote that “a suicide kills two people”. It’s always much more than that.

Lightning strikes and closure

Further fatalities occurred at Ufton AHB in 2009, 2010, 2012 and 2014.

The incident in 2010 was found to be non-suspicious. The coroner’s inquest into the incident two years later recorded an open verdict, as there was insufficient evidence to be certain that the fatality was the result of suicide. The circumstances surrounding the 2014 fatality were not treated as suspicious.

There was a near miss at the crossing on September 4 2011, when a three-car Class 165 formation heading from Paddington to Bedwyn traversed at 61mph while the barriers were raised and the red stop lights were not flashing.

A car approaching on Ufton Lane had to brake heavily to avoid a collision. The train driver, having seen the car, did the same thing, stopping the train 520 yards beyond the interface.

RAIB found that engineering work meant the crossing was being operated locally by a level crossing attendant, who had not received instruction to operate the closure sequence from the Thames Valley Signalling Centre in Didcot (which had taken over the route from Reading Box in 2010).

The report concluded that the incident was probably caused by work overload on the signaller connected with the engineering work, and the resumption of passenger services after its completion.

RAIB did not recommend closure of Ufton AHB, but the RMT union said that “level crossings on high-speed train lines should be banned and replaced with bridges [or] underpasses”.

In July 2012, Network Rail announced that it was considering either converting the crossing to full barriers or installing a bridge. On the tenth anniversary of the 2004 incident, and shortly after the 2014 fatality, the RMT and survivors of the 2004 accident repeated calls for the crossing to be made safe.

The following April, Network Rail submitted plans for a road bridge to the east of the crossing. The West Berkshire Council approved them in August 2015 and preparatory work began in September.

The bridge was officially opened on December 16 2016 and the site of the old crossing was converted to a Road Rail Access Point (RRAP) for maintenance vehicles.

For the survivors and bereaved, this finally closed off two of the factors in the accident: the survivability on board the train and the threat posed by the level crossing itself.

Network Rail continues to take a proactive approach to level crossing safety, using the bespoke All Level Crossing Risk Model to provide accurate risk assessments for each interface.

RAIB continues to investigate incidents, and RSSB continues to monitor the national risk and investigate possible insights - such as changes in crossing usage from dog walkers, delivery driver behaviours, and so on - to help target mitigations.

RSSB is also developing a standard that will capture level crossing fundamental principles to aid future level crossing design/interface management.

All this information is considered by the Level Crossing Strategy Group, which gets the key players around the table to work together towards solutions to the problems that arise.

The group also oversees RSSB’s Level Crossing Digest, which includes information on past accidents such as Hixon and Ufton, as well as more recent developments, to make sure the past is neither forgotten nor omitted from future thinking.

Suicide, of course, is a much bigger question. The industry has long worked to try to minimise the numbers of people choosing (or trying) to take their own lives on the railway.

One of the main problems is that suicide is linked to so many other aspects of society. One of the most recent is the pandemic, which has led to a rise in mental health issues more generally.

With this in mind, Network Rail has worked with Chasing the Stigma to develop a campaign called There is Always Hope, which seeks to address the issues and aims to encourage anyone suffering to seek help before they reach crisis point.

Network Rail, working with the operators, BTP and RSSB, also runs staff training initiatives and campaigns such as Small Talk Saves Lives, which calls on people to look out for one another and intervene safely if they see someone who they believe might need help.

No one can truly understand someone else’s pain, but that needn’t mean suicide can’t be prevented, that we shouldn’t offer support and empathy.

Samaritans, whose signs stand sentinel at level crossings, are there to help - as they always have been. For anyone suffering with their mental health, downloading the Hub of Hope app on the Chasing the Stigma website or at www.hubofhope.co.uk is also recommended.