I have to disagree with you there. For the Watford crash the driver's extensive history of SPADs would have been noted sooner if the driver monitoring processes that were being mandated had been introduced properly. The HSE report on this goes into quite some detail as to the fact that North London Railways, as it had just become, had yet to properly establish oversight of drivers due to the change going on around that time. The remodelling of Watford South Junction was interrupted by privatisation and there was quite a lot of miscommunication about what PSRs existed where and why. It had substandard overlaps which shouldn't really have been allowed with a remodelling and all sorts of other contributory factors IIRC.
The Hatfield crash quite simply couldn't have happened under the pre-1994 structure, and was largely precipitated by Railtrack's failure to properly supervise and audit its outsourced track maintenance. Not to mention the approach to spending as little money as they could get away with.
The Southall crash quite simply wouldn't have happened if the ATP hadn't been disabled. The mandatory reporting of this never happened and it had become routine to continue with it disabled. It was seen as an unnecessary burden in some ways. I can't see British Rail managed GWT having been so careless as to turn a blind eye to this. One reason cited for this policy was the cost savings achieved in reduced maintenance cancellations.
Ladbroke Grove was a lot down to the driver not being trained properly and being rather inexperienced. He should never have been out on his own on a layout as complex as the Paddington approach with his experience. The training programme was completely substandard - to save time and get trainees out there quicker, to plug the manning gap that had occurred. Signal Sighting Committees were not properly held on the sighting of the signal that was SPAD'd and the impact on visibility of the newly erected OLE wasn't properly considered in the mad rush to get it done.
True, some of these could still have happened under BR, but some of them definitely owed a lot to the process of privatisation.
As much as the profiteering TOCs may want it, they can't break the ORR's safety regime. They know it is one of the few areas where not playing by the rules means big penalties - not least in terms of lawsuits if non-compliance causes injuries or fatalities.
Other countries simply don't have the same safety culture as us.
After being busy on other things I should like to return to this topic.
I stand by my statement that the roots of these accidents do go back to the BR era.
In the case of Watford both driver monitoring and the design of the signalling were factors and they dated back to the BR period. The signalling — including the shorter overlap for signal WJ759 and the consequential permanent speed restriction of 60 mile/h on the approach to it — was commissioned in May/June 1993; the Railways Act 1993 was passed by Parliament on 5 November.
As you correctly said, the existing driver monitoring procedures were being changed - but these also dated back to the BR period. A total of 953 SPADs were recorded in the peak year of 1991/92 and as a result in September 1993 the BRB published Group Standard GO/OTC 508 'Signals Passed at Danger’. This introduced a hazard ranking process and a requirement to carry out a detailed analysis of any SPAD and then take appropriate corrective action.
While all this was happening North London Railways was set up as a state-owned ‘shadow franchise’ (as part of BR) in the run-up to privatisation. National Express was awarded the franchise on 7 February 1997 and commenced operations on 3 March.
The accident occurred on 8th August 1996,
seven months before North London Railways was privatised.
The root causes of this accident, both the signalling issues and ineffectual driver monitoring and training, go back to BR times.
Regarding Hatfield I would point out that train accidents had also occurred under British Railways due to track failure which better maintenance could have avoided. Clearly there were serious and terrible mistakes made in this case, but to say that they could not have occurred under the pre-1994 structure is stretching the limit of credibility.
Specifically the phenomenon of rolling contact fatigue/gauge corner cracking was not well understood at the time nor the influence of vehicle suspension and tyre profiles on its development. A clean rail break might — just might — not have led to such a disaster, but that 30-odd metres of rail shattered and disintegrated was unprecedented.
The issue at Southall wasn’t so much that the ATP was not being used but that the train was running with the AWS out of action.
Professor Uff’s report into the accident makes the following statement in the Preface:
The company (GWT) acquired its franchise on 1 February 1996 and had therefore been operating independently for some 19 months only, when the accident occurred. It would be wrong to see the Inquiry and this Report as an inquiry into privatisation. Nevertheless, the new structure of the industry has inevitably affected the events under consideration. At the same time it will be seen that the new industry is still heavily influenced by procedures and structures inherited from British Rail.
My emphasis.
Specifically it was the inactive AWS which was the immediate cause of the accident, not the lack of ATP. One of the affects of this accident was that AWS, and later TPWS, was added to the list of safety critical items which have to be functional — until then BR considered AWS to be an ‘advisory’ system and trains could be run with it inactive (albeit with other mitigations such as a second driver being present, which wasn't the case here) and this attitude was continued into the early privatisation era.
Uff’s report also gives much background to the development of ATP, specifically in Section 13.6 but which, although interesting, is not directly relevant to the immediate events leading to the crash. Nevertheless I quote it:
By 1994, on the eve of privatisation, the trackside infrastructure was complete (except for the final 12 miles into Paddington) with a total of 358 signals fined. The whole of the GWT HST fleet of 87 power cars had been fitted. The bulk of the installation costs were, therefore, already incurred and what remained was to bring the Pilot Scheme (together with that on Chiltern) into full operation. HMRI were, and continued to be, enthusiastic about ATP. However, just before Railtrack took over as infrastructure controller, BR on 31 March 1994 delivered to the Secretary of State a report on ATP which set out their conclusions on its economic viability. It was reported that the cost of ATP was substantially in excess of normal safety investment criteria (cost per equivalent fatality avoided). The concerns of BR and Railtrack were given public expression at a major conference in July 1994 on Value for Money in Transport Safety. By this date it was becoming increasingly apparent that ATP was unlikely to be fitted nationally. The BR report had been referred by the Secretary of State in May 1994 to the HSC. In December 1994 the Chairman responded, expressing qualified support for the report's conclusion. Further correspondence followed between the Secretary of State and the Chairmen of Railtrack and HSC in which reference was made to new safety initiatives within Railtrack's SPADRAM project, including TPWS. Finally, the Secretary of State on 29 November 1995 made a statement listing the safety measures being pursued by Railtrack and BR, with ATP being limited to the two existing pilot schemes and main line re-signalling projects. Extracts from the above documents are contained in Annex 26. From this point ATP was no longer a national solution. It had been effectively replaced by the SPADRAM programme including TPWS, for which trials were then already under consideration. Railtrack remained committed to the Pilot Schemes as they stood in March 1994, but considered they had no commitment to fit ATP more widely, nor to extend the Pilot Schemes.
The Great Western and Chiltern pilot schemes were thus left in an uncertain position. Both Railtrack and the Secretary of State had stated publicly that the pilots would proceed but the purpose of doing so was not obvious. The operational experience with the ATP pilots was initially poor. Equipment problems combined with deficiencies in driver training and, seemingly, a less than committed management drive, meant that the system was operational for at best less than 30% of the time in the period up to August 1997. This is hardly surprising as two years earlier the trials had officially been declared a dead end and the emphasis had turned to TPWS.
Ladbroke Grove was a lot down to signalling arrangements and sighting which, together with the unusual arrangement of the signal heads and the low sun, set a trap for the driver. But the proximate cause remains that the train failed to stop at a signal at danger in spite of the AWS sounding a warning.
One thing the inquiry should have done at the beginning, but didn’t, was to establish the exact sequence of events leading up to the collision. Because of difficulties in decoding the On Train Monitoring Recorder traces the actual behaviour of the outbound Class 165 was available only shortly before the inquiry closed — long after all the hypotheses, claims and counter-claims had been discussed in public.
These OTMR data were overlaid on a track diagram and they show quite clearly that the AWS warning was cancelled at the single yellow indication before signal SN87 on Gantry 6 (in rear of SN109) and speed was allowed to drift down to 38mph as the train entered points 8051, to be transferred from line 4 to line 3. During the crossing movement power was reapplied although the train was still some 60 to 70 metres before the sighting point for signal SN109 (this being 188 metres from the signal) and about 140 metres before the AWS magnet for that signal. On passing the magnet the AWS warning was cancelled, signal SN109 was passed at danger (all 5 other signals on Gantry 8 were also at red) and the train to continue to accelerate. In other words three driving errors were made in quick succession:
- power was re-applied before signal SN109 was visible
- SN109’s AWS warning was cancelled when the signal was visible, even if the aspects were obscured or unclear, but power was left on and the train accelerated past the signal
- no brake application was made when the train overshot turnout 8057 (about 150 metres in rear of SN109) that would have taken the train (if SN 109 had been ‘clear’) to line 1, the Down Main.
Thames Trains did not follow its own guidelines in recruiting the driver and the training process was admittedly deficient but he had driven into and out of Paddington nearly 100 times during and after his training. How many times does one have to drive over the same stretch before one knows it well?
The most likely explanation for points 2 and 3 is that the very bright low sun was shining into the lenses of the signal head and the driver saw what he thought was a double yellow aspect, so cancelled the AWS and carried on. One can debate for eternity whether this misinterpretation was due to lack of experience or lousy training. It is probably a bit of both — but the position of the signal gantry behind the Golborne Road bridge instead of in front of it where sighting would have been much easier is entirely due to decisions made by BR.
I do not accept that Paddington is unduly ‘complex’, the 6 track layout is only some 2 1/2 miles long. In Germany, in a city that I am familiar with, the approach to Munich Hauptbahnhof is twice as wide with 12 running lines over a similar 2 1/2 mile distance from the station at Laim to the throat of the Hauptbahnhof. It is on a similar alignment to that at Paddington, slightly north of west leaving the station. The sun is also low there in spring and autumn mornings but in addition it can suffer from poor visibility when snow is falling. Paddington is not unique.
We have to thank our safety record on the need for all organisations to have written and robust safety cases - this was not something BR had to do. The requirement for safety cases for each organisation involved with railway and train operation was driven by privatisation. It, and the cluster of accidents discussed above, led to the Railways and Other Guided Transport Systems (Safety) Regulations 2006, aka ROGS; enforcement through HMRI is crucial. Other countries are now catching up.
I rest my case.