Hatfield - October 2000

Hello and welcome back to Signals to Danger. As ever, I'm going to use this brief introduction to thank you for your continued listening and to again as ever, ask that if you are enjoying us to like, subscribe, share or review, you know, depending on where you're listening to us. It's a really quick intro for us this week, I would like to use it to plug somebody else's efforts. I was contacted by a chap called Ben this week. He's launching his own rail themed podcast, and you might as well ask something to listen to for the weeks I'm not releasing. Steel Wheels is available on Spotify, Anchor and Pocket Casts, but there are more on the way the same as Signals to Danger was. The plan is to release in batches of three episodes, one exploring railway history, one exploring railway technology and one exploring incidents and safety right up your street if you're listening to this podcast. I've already subscribed, and I'd love it if you would do the same. The first of episodes should be wheels any day now. With that, we're into today's episode, we're the remastered and tweaked opening credits. The railway is an infernal machine constructed of thousands of moving parts, suspension components that compress and flex, traction motors and wheels which rotate hundreds of times a minute, full crossing barriers which raise and lower to protect the line. Amongst all of these moving parts, there's one which needs to stay in place, solid and strong to keep trains and passengers safe. The track itself. When that fails, things go very wrong indeed. It's the year 2000, and this episode, we're visiting Hatfield. This is signals to danger, a podcast where we look at major rail disasters which have occurred in the UK, explain what happened, how the investigation was carried out and how each of these accidents shaped the industry going forwards. I'm damn. I work within the rail industry my day to day life, but today I'll be the one taking you through this podcast. This episode, we're in the year 2000, specifically the 17th of October. This means that if you're listening to this episode on release day, it's been just over a week since the 20th anniversary of the disaster at Hatfield. This is the reason I've chosen to do with this episode this week, 20 years is significant milestone for those touched by the events of this episode and I've certainly had them in mind while writing it. But now it's time to do what I've done every episode and tried to set the scene for what was happening at the time of this accident. The world dodged the Millennium Bug, moved into what everyone hoped was going to be an exciting new era. It was a vintage year for console gamers and my era of music lovers, with both the PlayStation 2 and Nintendo Gamecube being launched as well as Lincoln Parks debut album Hybrid Theory coming out. Some of the fundamental building blocks of my misspent youth fell into place this year. The year also showed that disasters weren't just limited to the railway, with the crash of a Concorde airliner shortly after taking off from Paris. In the closure of the Mayors Prison in Northern Ireland and the Sydney Olympics and that brings us more or less to the month of October. At 10 minutes past 12 on the 17th of October 2000, train 1 Delta 38 left London's Kings Cross Station. The service was an intercity express 225, operated by Great North Eastern Railway, GNER. This might sound familiar if you've listened to this podcast before, and there's a reason for that. This was the exact type of train which we spoke about in episode 1, Great Heck. Because of this I'll briefly run through the makeup of 1 Delta 38. It was comprised of 9 passenger carriages, 6 of them standard, 2 first class and a buffet car or a service car. At the rear of the train was a driving van trailer, a luggage and storage vehicle which also featured a control cab at one end. Leading the train, we find a class 91 electric locomotive capable of pulling the train set along at the maximum speed on the line of 125 miles an hour. The use of the DVT was to facilitate a push pull operation so trains didn't have to turn around at each end. These intercity train sets were one of the main types of passenger train on the East Coast mainline, one of the two main routes from the capital up to Scotland, as well as other cities up the eastern side of the country. Again, the East Coast mainline or ECML, that's something we've featured twice already in the five previous episodes, so I won't go into vast amounts of detail about it. Suffice to say that the electrification, among other factors, meant that large portions of this route were subjected to high permissible speeds. One Delta 38 left King's Cross at 1210, entirely as it would any other day. A train E sat at the controls of the 81-ton locomotive accompanied by a driver trainer. They slowly left King's Cross, passed through the tunnels at the station throught and out onto the mainline. As the train passed Bounce Green Depot, the driver's carried out a running brake test to make sure that everything was in order, no issues were found. As the line speed increased, the train E driver set up the speed set control, setting it to the maximum municipal speed at this point of 115 miles an hour. 170 passengers and 10 GNER staff sat in the train behind them as the train came up to speed. Just another one of the 1900 passenger trains that would be somewhere on the East Coast mainline that day. The journey had started without any incident or any reason for passengers or crew to be alarmed. People got magazines and paperwork out of handbags and briefcases, snacks would come out of backpacks at some tables and the people at other tables who hadn't bought any, started to make their way down to the buffet car to grab a bite to eat or a drink. At 12.23, less than a quarter of an hour after leaving King's Cross, the train was firmly underway. One of the features which make these speeds on the East Coast mainline so fast are the long, sweeping turns, with inclined or banked rails, which allow these trains to lean into the turn and stay on the tracks. Around 16 miles north of its origin, one Delta 38 entered the sequence of turns just like this. Shortly after their pass beneath a road bridge at Oxley 7U, both drivers noticed the brick pipe pressure gauge. It dropped to zero as they passed through the well and green curve. Over the next 700 metres, the train came to a stand from the 115 mile an hour line speed. Both trainee and instructor had felt significant buffeting from the rear of the locomotive as it came to an emergency stop. Once the movement had ceased, it was clear that something terrible had taken place. The locomotive sat stationary on the down fast, but the further back you looked along the train, the worst situation was. Both the locomon and first two carriages, A and B were upright on the line, looking more or less as they had when they set off from London. The only real sign that anything had gone wrong was on the underside of the carriages, but we'll get into that a little bit later. The story changed as you got to the third coach. Still coupled to the rear of coach B, coach C had derailed. The left hand side showed some structural damage, a window had been broken and both boggies were derailed to the left hand side, tilting the vehicle around 15 degrees over. The whole vehicle was slued towards the down slow, with the rear further towards it in the front. Coach D, the next in line sat more or less upright in the ballast between the two running lines, with an noticeable amount of leading end damage on the left hand side. All of the vehicles in this portion of the train were still connected together as they moved out towards the left hand side of the tracks. Coach E, coupled to the rear of coach D, was tilting over again to the left hand side with its rear end across the down slow line. Its trailing bogie had been ripped from its mountings on the bottom of the coach and lay in the ballast about two metres further back. When we get to the next carriage, carriage F, it was laid almost completely on its left hand side. This was reflected in the damage caused to that side of the carriage. The trailing vestibule was damaged and both boggies were detached, while the rear one sat more or less in place, the other one was found 47 metres away. By far the worst damage occurred in the next carriage, coach G. The service car, or the buffet car, I will use both as we go through this episode. It was lying on its side, half in the cess to the side of the line. Above the seating area of the carriage, the roof was peeled back like a tin opener had been at it, and the walls were gouged down to almost the level of the floor. The catering area was more intact towards the front end of the carriage, but virtually the entire seating area had been compromised. The trailing bogie lay 150 metres further back along the line at the bottom of an embankment. The forces involved in the collision had snapped the floor of the coach, and the last few metres of the carriage were bent back and angle away from the rest of the train. The red coach line of the GNAR livery shows this clearly in the aerial photographs of the incident. If you were to walk along the tracks from the locomotive to the DBT, the length of the wreckage you passed after coach G would be a 245 metre gap before you came to coach H. Along the majority of this gap there was debris, not only was this the odd displaced bogie and suspension components and ballast, but there was always a debris field laid alongside the tracks. This debris was made up of items, tables, chairs, fixtures, fittings from the buffet car, which had been, like I said, opened up like a tin. Tragically, the diagram in the report shows us that the bodies of two of the passengers who lost their lives were found within this debris field too. As well as all the contents and components of the train that could be found in the Gulf between the two portions, two uprooted overhead line structures, the masks which held up the wires above the track, were found lying among the wreckage. After a distance of almost a quarter of a mile, the next carriage of the train could be found. The coach H, one of the two first class carriages of 1 Delta 38, was laid on its side in the sess. Its left hand side showed severe abrasion, broken windows and frames. The leading bogie was free of the vehicle and was on the down slow line some distance ahead. In addition, the rear coupler was ripped clear and the connection between it and the next vehicle was also severed. Touch H, lay there entirely alone from the rest of the train. Five metres further south was coach M, the final passenger carriage. Embedded in the ballast, the lead bogie had detached and pushed around a metre and a half further to the rear of the coach. The leading vestibule was damaged but other than those two factors, most of the rest of the coach was mostly undamaged and immediately behind that still connected was the DVT, the driving van trailer. Again, predominantly unmarked, but both bogies had thoroughly embedded in the ballast and detached but only by a matter of centimetres. As the dust settled and the recovery efforts began, it was clear that Hatfield would join the list of names which resonated with the travelling public. This was a high speed disaster for a new millennium. The alarm was raised by drivers shortly after the accident, with the first emergency services on site after around ten minutes. Both GNER and West Angleian Great Northern staff in the area took the steps needed to protect the D rail train from further incidents. The Hartfordshire Fire and ambulance services assisted British Transport Police and the Hartfordshire Police to evacuate and escort passengers from the site, carry out further rescue work and to provide immediate medical care. Around thirty people were taken to hospital and the rest of the passengers to a reception centre at the University of Hartfordshire. By 1630 that afternoon, the recovery stage of the operation was stood down. As no longer expected to find anyone alive. By the end of the day, all four of those who had been lost at Hatfield were recovered from where they lay. The names of those who passed away during this accident were Robert James Alcon, a 37-year old commercial pilot from Auckland, New Zealand. Steve Arthur, a 46-year-old from Peas Potage in West Sussex, Steve had also been a pilot to run his own aviation business. Leslie Gray, 43 of Tuxwood, near Nottingham, was the next name on the list, and finally, Peter Monkhouse, a 50-year-old advertising executive from Headingly in Leeds, had also died on the day. On the next few days, the vehicles of the train were painstakingly removed from the tracks. The entire scene was fingertip searched and photographed from both ground level and the air. The health and safety executive, working in conjunction with British Transport Police, were charged with investigating the accident under the established work-related deaths protocol. As ever, both were on sight from the very early stages of rescue and recovery to start the crucial work of understanding why this all took place. As is the way every time these investigations start, there were a list of points that needed to be understood, factors that would provide answers, not only for grieving families, but a nation increasingly worried about the state of their railways. Hatfield was now a name alongside Southall, Ladbrook Grove, Clapham Junction, the hundreds of thousands of commuters who took to the rails every day, alongside politicians and the wider public, wanted answers as to how this had happened again. The three main factors from the investigation at Hatfield that we're going to look at are these. Number one, what had led to 11 vehicles of the 115 mile-an-hour train leaving the tracks on the curves approaching Hatfield. If I had one vehicle, the service coach being damaged so much more severely than the others and leading to the four deaths on the 17th. And finally, had opportunities to prevent the accident being missed. Probably not the only things will be exploring as we go through this, but they're really good jumping off points to understand. Many who has ever watched the television programme "Seconds from Disaster" will have heard the following words, "Disasters don't just happen, they're triggered by a chain of critical events." If you haven't watched it and you're enjoying this podcast, I'd probably recommend it. It's possibly your type of programme. That aside, the intro script hits the nail on the head. It would be a very rare accident that was caused by one single factor, especially in the industry's advanced as the UK rail network. It would mean the reports would be a lot easier to read and write and shorter and the podcast probably a bit easier to produce as well. But no, it's just never as simple as all that. Pretty much every accident investigated on the railway yields a number of causes. They're all substantial and all important, however we can group them into some different types. The RAIB, the rail accident investigation branch, groups and classifies these causes as the immediate cause, casual factors or underlying factors. The RAIB describes each of these as this, and I apologise but this is lifted straight from an RAIB guide on report writing, but I think sometimes that's okay to ensure that the information I'm giving you is accurate. The immediate cause of an accident is the factor that directly resulted in the occurrence of the accident or incident. Often simply a statement of the inevitability of the accident for example, the person was standing in a position where they could be struck by the approaching train. Factors that describe the key accident causation themes which contributed to the occurrence of the accident or the incident are referred to as "casual factors", avoiding or eliminating a casual factor would have prevented the occurrence. And factors that are associated with the overall management systems, organizational arrangements or regulatory structure are referred to as "underlying factors". An element discovered as part of the investigation that didn't have a direct or indirect effect on the outcome, but which deserves scrutiny is referred to as an observation. Grouping causes in this way means that we're able to understand the impact each had to the outcome on the day. The first point the investigation needed to ascertain was that immediate cause. As I said earlier, what caused the train to leave the track? If you think back to the episode at Potter's Bar, I made a point then about derailments not happening where tracks are straight and level. I made points about how the majority of derailments occur where their junctions are excessive curves. To follow on from that, we should see how Hatfield fits into this theory. Firstly, there was no point work in the area of derailment. It was a relatively simple section of quadruple tracking, as I said a little while ago, down slow, down fast, up fast, up slow, no crossovers, no junctions. That couldn't be a contributing factor. First of the second question, excessive curvature? Well, there was curvature at Hatfield on the well and green curve, but it wasn't excessive. The fast lines here had a permissible speed of 115 miles an hour. This isn't more worth a few hundred miles up the line. Of course the question was asked as to how fast the train was travelling in relation to that speed, but it doesn't really fit the bill here for the due cause of derailment. In any case, the immediate cause was abundantly clear to every investigator who arrived on the scene of the accident. If you recall the potter's bar episode, I used a phrase, "The unbroken, straight rail is pretty good at keeping train and track together." Forty-three metres north of the bridge, which takes Oxley's avenue over the railway, investigators found something which goes against this statement. The broken end of the left hand running rail. The rail was broken into a large number of pieces. In fact, an unbroken section of line didn't start again until 36 metres later, which was 79 metres away from the bridge face, which was used as the datum point zero metres for all measurements at the accident scene. This 44 metres section of unbroken rail, which ran from 79 metres to 123, had been tipped out of its fastenings, and immediately after that there was another 54 metres section of fragmented rail, running through to a 180 metres past the datum point. No train, regardless of the speed it's travelling, can run on only one rail, so one that was running between 115mph and 117mph didn't really have a chance. This means that the fundamental immediate cause of the accident was that 1.380mph had derailed due to a broken rail. Of course it's not really acceptable to turn around and say the track broke and leave it at that, so the investigators did some extensive surveying of the condition of the track, ballast and sleepers at the scene. What they needed to understand further was two things. What was the sequence of the brake, and what had led to it occurring. As they looked around the sights, some of the things they surveyed were marks left on the other rails, one of which is called a drop-in mark. These are found on the opposite rail to a broken one, and they are marks that show when the wheels slide and drop off that rail into the ballast and sleepers below. 32 wheel sets, two wheels joined by an axle of the train, had been derailed at Hatfield. Investigators were able to identify the exact 32 drop-in marks on the right hand running rail of the downfast, which corresponded to each of them. The last one of these drop-in marks was found around 72m from the Dayton Point, and that meant that all of the derailed wheels of the train had left the track within that first section of shattered rail. Other marks and tracks that proved instrumental in helping investigators piece together the sequence of the derailments were where wheels had run through the ballast, forming deep troughs, concrete sleepers, the horizontal structure that the rails amounted to on the floor, they'd taken the full weight and speed of the derailing train, they'd seen heavy damage. Some of them had been reduced to more or less just the reinforcing wires that had once sat deep inside the concrete. There were also marks visible on some wheels of the train which reflected the damage and the impact they had experienced. Further to that there was damage to the down-slower line, its rails and its concrete sleepers boss guard from the derailment as well. Over a length of time, pulling all of these pieces of evidence together, the investigators were able to say that this was the sequence of events. The locomotive at the head of the train passed over the point of derailment, in fact there's a possibility that it may have crossed over a fracture in the rail as there were some marks left on the wheels of the locom. The left hand rail started fracturing at this point and the locom in first two carriages the D&B remained railed, passing over the rapid disintegration of that left hand rail. There was evidence that debris had struck the bottom of the couches. Coaches C, D, E and F encountered this displaced and fractured track and started to derail one by one. As they impacted it they increased the severity of the fragmentation, leading to this 36 metre section of rail being broken into over 200 pieces. It's likely that several of the bogies may have become detached at this point, only held in place by the suspension components. Coaches C started to displace the left hand side as the locomotive pulled the rapidly derailing train along the line, still at about 150 miles an hour. At this point the service coach, coach G, reached what was probably a gulf of misplaced rail. The evidence suggests that the bogies of this vehicle became detached almost immediately and that this coupled with the turning forces rotated the vehicle onto its left hand side. This managed to uncouple the service coach from coach H and encoupling these two carriages broke the pneumatic connections between the two vehicles which is what caused the brakes to be applied on the train. This next part of the sequence explains the reasons for the second point that we're considering. Why did coach G, fare so poorly compared to the other ones? Paint marks on the slow lines showed that the service coach that had been on its side dragged along by the rest of the train and going along the adjacent down slow line, 75 metres past the point of the derailment. As the derail trains sped around the corner, the tip of the tail which coach G had become swung further out to the left. After another 149 metres, the roof of the coach G made contact with one of the masts for the overhead wires. This mast became lodged in the roof of the coach and it was ripped from its foundations carried along by the train. The damaged cause by this started to spread the debris of the service coach along the side of the track. The train started to slow at this point due to the breaking and the resistance but the service coach still had some of its journey to complete. As it swung further out, the roof met another mast. This time the collision was more considerable and the angle was steeper. The mast, Echo 2709, was torn from its base, folded and penetrated into the coach from the roof down to the floor table and seat level. This opened up the rear half of the coach. The debris field increased substantially at this point and as the mast reached the trailing under the carriage, it lodged, carried along for a distance as the first one was. This impact to the rear of the carriage is likely what snapped the floor. Seeing how these impacts took place shows us clearly why the fatalities all took place in the same carriage. The survivability odds in the seating area were not favourable to anyone who'd been in there. If you do get an opportunity to look at the report there are some photographs in there and it's like a scoop has just been taken out of the rear half of the carriage. Seeing those photographs certainly filled me with a somber mental image of what the derailment must have looked like for those in that space. Understanding the how of how 1.38 derailed was now taken care of, we've ascertained the sequence of events and the immediate cause. Now we need to look to understand why the rail broke. They were open to any possibility, need to be when you investigate something like this. There was the possibility that a rail fracture could have been initiated by an object present on the high downfast rail and that it could have been struck by the locomotive and caused the rail fracture, so the front of the locomotive was inspected for damage. There wasn't any, it was in good order, so it followed that a large substantial object couldn't have been in the path of the locomotive. Did something smaller? Well the leading left hand side wheel sanding unit is located approximately 90mm above the rail and it just sticks out it's there. The sanding unit on the locomotive was inspected and again there was no evidence of damage, so they could say fairly categorically that an object greater than 90mm high was not present on the rail. The foreign body theory was fairly dead in the water. Somewhat at what we see in most accidents there was also a concern early on in the investigation that they may have been some terrorism involved, however after a brief period it was cleared up that this wasn't the case, but the investigations had taken place on that first day. What it would all come down to would be microscopic pieces of evidence. Each of the fragments that the health and safety executive in British transport police could locate was collected and examined. From the initial fractured segment they managed to locate around 200 fragments which they believed equated to around 90% of the section. What these collected fragments told the investigators, well I'll quote it directly from the report. The left hand rail where the derailment occurred was riddled with surface cracking. In fact the examined several hundred metres of track around the derailment and this cracked marked surface was typical of it all. They looked at the section of rail leading up to the derailment point and found an almost continuous network of cracks and also some spalling which is where small chunks of the metal are lost from the surface of the rail due to the cracking. The actual derailment point, that sudden lost brick, there was a brittle fracture accompanied by a large transverse fatigue crack down through the rail. Three more large fatigue cracks were found within the section. In fact the scientists who examined the rail stated there was a high probability of those fatigue cracks leading to a full fracture of the rail. Now the report into Hartfield goes into pages of details on how the rail was examined and what was found but I won't bore you with all of that, I don't necessarily understand all of that, it does make for some good reading if you have the time. But suffice to say, the high rail in particular, the left hand running rail on the inclined corner was in a shocking condition all the way around this corner. And it allows us to answer 0.1 fully. One Delta 38 derailed because of a broken rail that was caused by extensive fatigue cracks leading to an eventual shattering of the rail under load. The point that we need to understand next is the third one. Had opportunities to prevent the accident being missed. Arguably it's the most important question to answer. Hartfield has happened. The four lives tragic as it was had already been lost and the damage and the injuries sustained. Very sadly we've never mastered retrospective prevention, we probably never will, but if an opportunity to prevent it was missed it could have been missed elsewhere, answering this could save lives in the future. To start with I'll quickly introduce you to two phrases now, rolling contact fatigue and gauge corner cracking. Rolling contact fatigue is the fatigue cracking which occurs on the railhead due to the high stresses of the wheels pushing down on the rails. These cracks initiate at or near the surface of the railhead. Gauge corner cracking is when these fatigue cracks morph into deeper cracks down through the rail, specifically in the corner of the railheads which are known as the gauge corners. The fact that we know about these two types of fatigue and we did at the time of Hartfield should tell you that metal fatigue and degradation of the cracks isn't something new. It was expected, of course it was. The strongest steel is you can't expect to run hundreds of tons of metal over the top of it at a hundred miles an hour and not expect some wear and tear. So, like anything that wears down, you inspect it, you monitor it, think of your car tires, you check them every now and then to make sure the tread is looking healthy. We might not always do it as often as we're supposed to, but we check. It's the same for the railway, in fact a whole lot of components on the railway, but checking is a lot more regimented. It's time to look at who is responsible for those inspections. If you recall the episode on Potter's Bar, we briefly discuss the relationship between Railtrack, the precursor to Network Rail and Jarvis, one of their maintenance contractors. Well, we're in the same ear now, Railtrack is the private company responsible for virtually anything which isn't a train on the railway, and they have a contractual relationship within another business for maintaining all infrastructure, track, rail and signalling equipment on the East Coast mainline. This company is Balfour BT Rail Maintenance Limited, or BBRML, I'm going to abbreviate it to BBRML quite a bit. BBRML was responsible under their contract from Railtrack to ensure the lines were safe for the passage of trains. The activities used to acetane fitness of the line could be split into three areas. One, visual track patrolling inspections to examine the external condition of rail. Two, ultrasonic, non-destructive testing, which examines the internal condition of rail, and also the operation of a high speed track recording coach, which is a precursor to today's new measurement train, which we are going to touch on again later on. These activities all overlapped so that missing one of them wouldn't cause a major accident. The track recording coach mainly focused on geometry of the track gauging etc, so it's output wasn't really relevant to the events at Hartfield because the track geometry was fine. The other inspections were carried out under the guidance and instruction of a route section manager working underneath an area maintenance manager. Moving into this section, one thing I'd like to have at the forefront of your mind is how the rails take on even loading corners. You'll often see a greater load on the outside rail, the higher one, as the track is inclined to lean passing trains into the turn. At high speed corners, the high rail is subject to a great deal of load and strength. Firstly, let's take a look at the visual inspections. There was a set time skill for them to take place, based on railway group standards and the contract between BB-RML and rail track. Based on the speed at well-emgreen curve, a visual inspection should be completed at their once every week and all rail defects should be marked recorded and kept under special observation. The railway group standards, or RGS, told us that where tracks are separated by no more than a standard 10-foot gap, two tracks may be visually inspected during a single patrol, but the track walked should be alternated between successive patrols. Basically, you could expect the down fast and the down slow at the same time, but if you walk the slow this week and look over to the fast, next week you walk the fast and look over to the slow. Because of the size of the gap, you can only do the uplines of the downlines. Essentially you need to walk the four track section twice to carry out the visual inspections for this section of the H-Tross main line. And that's where we start to see things fall down already. While in general, the record showed that inspections were carried out weekly, the BTB interviews, and the reports didn't tally up. There were anomalies in the way the lines were inspected. In interview, patrolers said that they inspected two lines in each patrol in accordance with the standards. However, their inspection reports recorded observations which concerned all four lines, and were entered sequentially by distance, not my line direction. This indicated all four lines were being inspected during a single patrol. That wasn't in line with the requirements, and it probably meant that not all of the lines were receiving anything like an adequate inspection. You can't see small surface cracks across the distance of four lines. In another concerning revelation, patrolmen stated that with only one lookout, it was only safe to patrol the curves from the Cess, the very edge of the lines at the edge of the ballast. This too was outside of the requirements of the standards. So not only were you looking at four tracks from the left most one or the right most one, you come back another few feet on top of that and try and examine all four from the Cess. There was a revised track patrol system in the summer of 2002, and it clearly indicated that a second lookout was required so that the track walkers could walk in the Cess. In the fourth up sorry, but this wasn't implemented by October. None of the patrols reports referred to or even identified role in contact for tea in the area of the derailment, despite the fact there was heavily evident damage and spalling, which didn't just show up overnight over the length of this piece of track. The patrolmen quite clearly weren't spotting some of the faults they were meant to be, and the visual checks just were not serving the purpose of their design. And the problems didn't stop with the visual inspections, they were also passed across to the ultrasonic testing regime. At this time there were two main tests that could be carried out, the U3 and the U14 test. U14 test being used to help highlight certain fatigue cracks that may not be seen in the quicker and more routine U3. Some fractures and fatigue defects could prevent U3 from working, and a number of actions could be taken based on the severity of the issues. These ranged from an emergency speed restriction to arranging for the rail just to be ground down and retesting. In November of 1999, there was a break on the high rail at Aiklif on the outside of a corner. This led to emergency action to renew the rail. It triggered a range of further actions and briefings on gauge corner cracking. And the management of these defects within BBRML and rail track. Rail track issued an important instruction as to one action should be taken if the U3 test probe couldn't get the signal that it needed to and the test couldn't be completed if it was a result of gauge corner cracking. So if the defect was isolated and it could be clamped in place, then the rail should receive a 20 mile hour speed restriction and removed in seven days. If the defect couldn't be clamped or if it was considered to be a multiple defect, then a 20 mile hour restriction was to be applied and the rail removed in 36 hours. This was a clear and unambiguous instruction, and it was there to save lives that had been a near miss at Aiklif and rail track didn't want another one. So except for the fact that on the 28th of November 1999, an ultrasonic tester was unable to run test U3 at well and green curve just south of Hatfield station. The form listed GCC, gauge corner cracking, severe loss of rail bottom and requirement to re-rail the corner. The action take and field of the form had nothing in it. Which if it wasn't anomaly would be problematic and it was certainly against the instructions issued after Aiklif, but it may be recoverable, except it wasn't a one off. Tests completed on the 5th of April and 14th of June 2000, both listed severe GCC at the same location. None of the forms generated showed any additional actions or follow up, and in fact on the 6th of October 2000, an attempt was made to carry out U3 testing at well and green curve. The report notes read, "Rail untestable intermittent LRB 1650-1640, all caused by severe GCC gross chipping of rail surface in places. At the top of the form was a remark, re-test. They wouldn't get a chance, 11 days later there would be a little left of the rail after one delta-3-8 destroyed, as it derailed over the top of it." To top all of this off, these defects have been knotted as far back as January of 1999. The rail at well and green curve was identified for grinding in April of 1997, BBRML submitted applications again in 1998, 1999 and 2000. Grinding finally occurred on the 5th of September, and month later the site was ultrasonically tested and the defects remained. Having the rail profile ground to comply with the standard and still being able to unable to examine the inside of the rail, should have indicated the severity of the problem with the railhead. That should have caused a site visit by track engineers and emergency action should have been taken. It just didn't happen. Now, I think it would be fair to say that BAL4BT were not living up to the requirements and then as line infrastructure manager. It also answers quite clearly the third point. There was ample opportunity to prevent this accident. In fact at the end of September, another incident of severe cracking was noted at Stambler Ulrich, but that was re-railed within 11 days. There is a page in the report which shares the conclusions of BBRML's performance and it is to the word "ascading indictment" of it. Some of the points raised were the action part on report forms were not completed. Repeated critical observations regarding the rail at well and green curve were not acted upon. BBRML failed to comply with group standards, line standards and specifications and some of its own procedures with respect to managing defect in the rail at the well and green curve over a considerable period of time. And BBRML identified the state of noncompliance but failed to take reasonably practical precautions to control the risk. Which shockingly is the case, they had noted a backlog of tests and repairs which just needed completing in a set time skill and were not. By the 1st of June 2000, they had identified a number of areas where they had failed to meet those standards. From a set of 10 defects that must be removed within 4 weeks, 7 were over due, 66 defects to be removed within 13 weeks, 19 were over due. And from nearly 90 welds, which needed a repair within 3 months, nearly half 41 had become over due. In response to this backlog this situation, they recruited an additional 16 staff and a manager to produce and manage a backlog recovery plan. However, within that recovery plan there was no evidence of any risk assessment to determine likely rail failures and no action to impose interim speed restrictions at any of the sites with the most serious defects. Their 3 month recovery plan deemed adequate to resolve the problem. They just didn't manage to complete it before the 17th of October. While BOL4BT was certainly the company who was supposed to be inspecting and repairing the line at Hatfield, Railtrack was the one who carried responsibility for the infrastructure. They did not get away scot-free. The inquiry led a portion of the blame firmly at their face. In fact, they ascertain that Railtrack failed to apply thoroughly and effectively all the necessary management monitoring systems and the arrangements at their disposal to ensure that BBRML met their contractual standards on quality of maintenance. And they came to the conclusion that the culture within Railtrack, which conditioned decision-making, was biased towards performance-based decisions, in particular towards minimising train delays and failing to focus on the poor quality of maintenance. Railtrack was thoroughly aware of the failings with BBRML's performance, and in the two-year period prior to the due element, several senior staff took them to task for not achieving the required standards. They've threatened the contractor with millions in financial action to seek improvements in performance. They even threatened to bring in another contractor to fill the shortfall and to charge BOL4 for that work. However, despite all the criticisms in June of 2000, a certificate was signed which informed Railtrack headquarters that this zone of the network was in compliance on all the necessary standards, which it just wasn't. Railtrack also held some more direct responsibility. For months, the track at Wellengring curve had been a priority one repair project. The investigation found that the urgency for re-railing had been identified in February of 2000, on the 29th of March 2000, the Railtrack planning system for re-railing designated the work to take place between the 13th of May and the 29th of May. It didn't. On the 30th of March 2000, BBRML wrote to Railtrack advising that the downman was very poor and on the 3rd of April Railtrack replied. This site is indeed in very poor condition, and as such has been assessed as priority one, which is to be undertaken during the 3rd week in April. This communication, clear indication that both BBRML and Railtrack were aware of the severe problems with gauge countercracking and rolling contact fatigue in the derailment area, they even moved the plan date for the Rail renewal forward by a month. The work was planned to take place alongside another piece, slightly further up the line towards Hatfield Station, but that work overrun, and it used up two of the three weekend possessions planned for the well and green site, and those possessions weren't replaced in a subsequent planning. The well and green site renewal was postponed. On the 30th of August 2000, an application was made for possessions to re-rail the well and green curve on various dates up to the 28th of January 2001. The second application proposed some interim temporary speed restrictions, but no action was taken, no rescheduled possession was ever booked in, and by the end of October, the problem had been taken care of from them in a spectacular fashion. Without Railtrack effectively carrying out their audit role and managing their contractor, many opportunities were missed to prevent the disaster. When you couple this to their failure to push the infrastructure renewal through, despite their knowledge of the issues, Railtrack certainly share a considerable amount of the responsibility. Once the causes had been ascertained, and the tracks we laid, the railway started to get moving again. However, because of the serious shortfalls found at the green curve and deprecited issues both at BBRML and Railtrack, Hatfield received another nickname, the crash that stopped Britain. After the realisation of the scale of unresolved rolling contact fatigue and gauge-counter cracking, it's believed that there were 1867 sites across the country where temporary speed restrictions were imposed as a result. Workers had pages upon pages of additional notices to read every week, alterations, temporary speed limits, restrictions, and by January of 2001 more than 240 miles of track had been replaced across the network. Over Christmas 2000, work was carried out at over 60 locations and used 8,000 man days. It was an incredibly painful time for those working on and travelling around the network. On the 24th of November, BBC reported a drop of passengers of 50 per cent as they struggled to use the knee-capped service. The network did recover, however, and saw a gradual increase of passengers over the next 20 years until you come to now, and that's a complete different picture, but it does sort of illustrate it a little bit. If you think that after Hatfield, it dropped by 50 per cent for a period, and now we're living at 30, it's a strange time to be on the railway and it will have been the same then. Lives were needlessly lost at Hatfield, and people needed to be brought to account over it. On the 9th of July 2003, the Crown Prosecution Service announced that six individuals have been charged with the manslaughter of the four people who died in the Hatfield derailment, and with breaches of the Health and Safety Act. A further six individuals were served with summons for Health and Safety breaches, as well as summons being issued to both network rail and Balfour BT. As it happened, all the charges levied against individuals were eventually dropped. On the 18th of July 2005, Balfour BT pleaded guilty to a Health and Safety charge relating to the derailment, and on Tuesday the 6th of November 2005, network rail, who by this point had replaced rail track, was found guilty of the Health and Safety charges. On the 7th of October, Balfour BT was fined £10 million, which was later reduced slightly to £7.5 million, and network rail was fined £3.5 million, both ordered to pay £300,000 additionally in prosecution costs. The judge stated that the company's failure to abide by safety rules was, I quote, the worst example of sustained industrial negligence in a high-risk industry I have ever seen. While a repair work left a significantly better railway in the short term, perhaps the biggest change came with the departure of rail track. We discussed it slightly during the Partis Bar episode, but one of the main reasons for this happening was the derailment at Hatfield. The company's share price plummeted, as people became more and more aware of the incompetence of the business, and the shortcomings of its ability to resolve the issues facing it. Alongside the issues on the East Coast mainline, the company was also struggling with spiraling costs on a modernisation project on the West Coast mainline, and by 2001, rail track announced that despite making a pre-tax profit of £199 million, the £733 million of costs and compensation paid out over the Hatfield crash plunged rail track from profit to a loss of £534 million. After a controversial shareholder dividend payment following some state assistance, the government had enough. They slammed the door on the private company and placed it into administration in October of 2001. It was bought by network rail in October of 2002, and the new government-owned body set about the task of improving the management and maintenance of the infrastructure. As I mentioned in the Partis Bar episode, private contractors were removed from the day to day maintenance, and testing and inspecting was all brought in house as well. One of the most prominent and now beloved outcome of the derailment at Hatfield was the flying banana. Now, haven't lost it, I'm referring to the new measurement train or the NMT. Two Class 43 locomotive, sandwiching five carriages, full of measuring equipment, cameras, sensors, lasers. You name it, it's got it. It's a bright yellow, high speed train, and it can assess and measure many crucial factors at line speed. It's also cleared to go everywhere at HSTs. The train was formed in direct response to the gauge corner cracking concern after Hatfield. It's been adapted and improved over the years, and it currently runs a four-weekly cycle over a large proportion of the UK rail network. Network rail's website has a really good page on there which details the train, and what it does, and it's actually really good and I'd recommend a look at it. The last point I want to make is the seemingly unlucky nature of the locomotive at Hatfield, number 91-023. For the recovery from the scene, the undamaged locomotive was retained at crew for some time, and then was let back out to the network to take passengers up and down the east coast. Until the 28th of February the next year. Only a few months after Hatfield, 91-023 was pushing, its southbound early morning service passed great heck, when the DVT collided with a Land Rover, and another ten people were killed. If you haven't heard episode one of this podcast, go and have a listen to it, it's the first accident we covered, and the locomotive was the same one. I'd like to close today as we have it other times by talking about the memorial to the accident. About five minutes south of the station at Hatfield in a car, just off the Great North Road, this small memorial garden, plants surrounded paved circle directly adjacent to the lines, at the back of Stornstands holding a plaque to remember those who died and for those who visit to see read and remember. The plaque reads, "In memory of the four men who lost their lives and those who were injured in the Hatfield Rail Tragedy on October 17, 2001. And as a train passes by on the east coast, may align behind it, I'm pretty sure it would be hard not to think about what took place on the day." [Music] [Music] [Music]