How Can You Mend A Broken Railway? Thameslink and the Gibb Report
Last summer the Secretary of State for Transport, Chris Grayling, commissioned a report from an experienced and well-respected railwayman, Chris Gibb. The goal was a framework to enable much-needed improvements to railway services across the Govia Thameslink Railway (GTR) franchise.
Terms of reference – optional extra
Although all of the franchise was to be looked at, it was clear to most people that this was basically a case of asking what could be done in the Southern Railway area as a whole to provide a significant improvement in performance. The terms of reference as to what was off-limits seem to be fairly vague – or possibly even ignored. Gibb has gone into great detail in some areas with ideas possibly worth studying, but unlikely to be helpful in the short term. He has also made various subtle comments on the industrial disputes which have affected the area, whilst sticking rigidly to the term of reference that stated:
The management of industrial relations remain a matter to be managed by Govia Thameslink Railway Limited.
Whether this term of reference was intended to encourage Mr Gibb to steer clear of this area or was simply a case of the government emphasising at every opportunity that the industrial dispute is nothing to do with the DfT is unknown.
The report is invaluable in that it contains a number of original ideas. Equally though, it sometimes states the blindingly obvious. That some of those points need to be stated at all, however, shows how far the GTR network really is from an ideal railway.
In true railway style, the report is delayed
The report was due to be presented to the Secretary of State “by the end of the year” and that deadline was met, although the time constraints naturally mean that the report is not as glossy or well structured as one unconstrained by time. Gibb started work on the 1st September so only had four months to complete the work. Indeed the time taken by the government to publish it was actually greater than the time taken for Gibb to write it.
This report was due to be published in April, following the department’s initial assessment of the recommendations with GTR and Network Rail. The pre-election period meant we could not release the report and today’s publication is as soon after the election as possible.
The election was called on the 18th April, so clearly the government weren’t rushing to publish and obviously they were not expecting to be left without a clear majority and the distractions that would cause. Delayed publication seems to be yet another of the insidious indirect consequences of Brexit.
Hard hitting from the start
Gibb does not pull any punches. He leads by stating what he thinks is the primary issue (spelt out in detail in section 3.2.2). He lists a catalogue of changes taking place and states that the fundamental problem was that all these changes were taking place at the same time with no overall systems integration.
The first three items in Gibb’s list of changes are placed fairly and squarely at the collective feet of the DfT. The fourth item is laid at the doorstep of the unions. Most other items have the hand of government upon them in some way – either through Network Rail or the Office of Rail and Road. The effect of the London Overground timetable on Southern is noted. Indeed the only item for which management (government or TOC) or unions cannot really in some way be held responsible for is:
Rapid growth in passenger demand, resulting in overcrowded stations
The fact that the word “timetable” appears 79 times in the first 76 pages of the document emphasises just how much of a problem the timetable is – and this is before we have a 24tph service running through the centre of London.
Blame the unions
Section 3.2.3 is headlined “The Primary Cause of the System Breakdown in 2016”. In this Gibb argues that the industrial dispute was ultimately the straw that broke the camel’s back. This seems to be an emotive argument, rather than one backed up by facts, although he states fairly reasonably:
But I am convinced by what I have seen that if the traincrew were to work in the normal manner that they have in previous years, the output of the system, a safe and reliable rail service for passengers, would be delivered in an acceptable manner, which would be similar to other commuter rail services in the South East.
This may or may not be true, but it could equally be said that had the other events that caused so much disruption not been present then the effect of industrial action, at least in the later walkouts, would also have been fairly minimal.
Southern commuters have suffered for a long time from industrial disputes but, if emotion is put to one side, it is easy to see that the case is not as clear-cut as suggested. The average Southern commuter would be more likely to have a better journey on a day when the railway ran perfectly, but for the RMT being on strike, than in the reverse situation. This is simply because RMT strike action is no longer very effective in the region as a result of more driver only operation (with a second crew member still present) and more guards (those that still are guards and not Onboard Supervisors) being prepared to work on strike days. Furthermore, the few trains that are cancelled mean that the other trains have a greater chance of running to time.
Gibb implies that if the unions withdrew their action then, despite other problems, a decent rail service would be restored. This overlooks the fact that on the vast majority of days no-one is taking part in an industrial dispute as such – although there must be situations where members are still working to existing practices rather than those that are proposed. Such an argument also fails to explain why the system deteriorated almost as soon as GTR took over the running of Southern and also why things were so bad throughout 2015, most notably at the start of the year, when no industrial action was taking place.
After commenting on the industrial action, the Gibb report then looks at maintenance of the network. As with his comment on a lack of systems integration, here he has probably fairly and squarely identified a major issue. Sad to say, the problem is not new – it is well known but requires a considerable amount of money to fix.
Way back in the 1980s, a senior British Rail manager said that the trouble with the Brighton Main Line was that it had been ‘soled and heeled’ too often. Given further deterioration in the privatisation years and lack of money spent on the Brighton Main Line, it is not surprising that today we have a railway that is not reliable:
Along with the rest of the 2018 Thameslink network the condition of the infrastructure needs to be urgently raised. This is not complicated stuff – it is about rail renewal, switch & crossing renewal, sleeper renewal, ballast renewal / removal of wet beds, drainage improvements, telecoms / signalling cable renewal, axle counter introduction, vegetation, removal of temporary speed restrictions, attention to fencing, structures and earthworks.
Such a comment would be expected from Gibb. It was he who led the transformation of the Virgin West Coast Main Line to reduce the considerable delays by looking in detail at all factors that were causing issues. On this aspect, the results of such an action are clear to see on the West Coast Main Line. It is not surprising therefore that Grayling was quick to release £300million, as suggested by Gibb, to enable this work to be carried out.
Possibly, what led to such a prompt releasing of funds to overcome the backlog of maintenance was Gibb spelling out what he thought needed to happen if this was not done:
If these funds cannot be identified, then I recommend that a decision must be taken by the DfT to reduce the Thameslink 2018 specification to a level that the existing system reliability can support. Such a decision should be taken in January, 2017. If it was decided to do this, there are significant implications […] less rolling stock, drivers and depot facilities will be needed for Thameslink, and older rolling stock can be withdrawn quicker. But Thameslink will remain at 12 tph – half of what has been envisaged under the Thameslink programme.
One can just imagine how politically unacceptable a 12tph peak service on Thameslink would be when a full service of 24tph is expected in December 2018. One can certainly imagine what the attitude of Southern commuters – who have suffered years of disruption on the promise of a better service – would be.
Gibb then goes on to point out the large number of vacancies on Network Rail, the fact that there are no specific plans to fill these vacancies and that there is no plan as to how the backlog of maintenance is to be tackled. So, one of the other things he addresses in this area is how this backlog of work is to be carried out.
The report makes the eminently sensible suggestion of reducing overnight main line services and having earlier last trains Monday-Thursday on Metro services. The 11 pages of Appendix 3 go into considerable detail as to how this is to be carried out, down to exactly which trains should be withdrawn and which should have additional compensating stops. This revised timetable was introduced last May with very little dissent. The withdrawn trains should be reinstated by the end of the year.
A blockade too far
As well as enabling more overnight works, Gibb recommends a two-week closure of the line between Horsham and Three Bridges in the summer of 2017:
During Summer, 2017 I recommend a closure for two weeks of the route between Horsham and Three Bridges, with passengers and trains diverted via Dorking, Epsom and Balham. At the same time, with no trains coming from the Horsham direction, I propose closure of two tracks between Three Bridges and Earlswood, closure of Horley, Salfords and Earlswood stations, and a reduction in service. GTR’s local service between Horsham and Epsom will need to be reduced to provide paths and crews for diverted Arun Valley trains. These two weeks would allow for an intensive and productive period of infrastructure maintenance, on these sections, that are notable for their poor condition. With the support of DfT, Network Rail and GTR should start planning this closure at the start of 2017.
It is here we start to question the ideal over the practical. For a start, six months is actually not a lot of time to prepare for something like this. Although he subsequently points out the dates of Waterloo partial closure (5th – 28th August) which need to be avoided, he does not appear to appreciate the need to avoid the following week as well. For a full week starting on the August Bank Holiday Saturday there will be no Southeastern trains at Charing Cross, Waterloo East and London Bridge. Prior to that, through much of August, is the work at Waterloo to length platforms. The work at Waterloo is expected to be extremely problematic (something we will cover in a future article). Ultimately, on a practical level, doing the work between Three Bridges and Horsham in 2017 was always likely to be a non-starter.
August and early September is really the only summer period where passenger levels are light enough to consider line closure between Horsham and Three Bridges and, even if you could do it in late July or early September, the potential for yet another major engineering blockade to affect already planned ones – either in terms of demands on plant or manpower, or plant breaking down due to lack of maintenance or staff being overcommitted – just seems to be a risk not worth taking. The principle of the idea may not be a bad one, but it really has to be implemented in 2018 if it is to be done. It is notable that this recommendation was not acted upon.
The timetable – it’s always the timetable
Another area where Chris Gibb focuses his attention is the timetable. It is a mantra here at LR Towers that you can never put too much effort into getting the timetable right. In trying to suggest improvements, Gibb looks in great detail at why we are in the situation we are in today – where there is basically no slack in the timetable. The consequence of this is that, on Southern, it is next to impossible to fully recover from a disruptive incident in the morning in time for the critical evening peak period.
Gibb probably goes to the heart of the problem, at least on some lines, when he states that:
It is clear to me that the three previously competing operators deliberately filled up every off peak path, to stop the other operators running additional services and receiving a larger income allocation through ORCATS. I can see no sign of any rationalisation of the contractual obligations occurring prior to the letting of the current franchise, so all the competing services were amalgamated into the new franchise obligations.
This is one of the many clear absurdities that Gibb points out throughout his report. However, in this case, his proposed solution is unlikely to go down well. One of his ideas is a “fire-break” between 12.00 and 14.00 Monday-Friday where, for example, services with 6tph go down to 4tph. Apart from timetabling complications, this seems to almost kill off the idea of ‘turn up and go’. Furthermore, the alternative of the regular clockface timetable is also destroyed. One suspects that a lesser inter-peak service from start to finish would be more practical and more desirable if one were to go down this route. It is also hard to imagine TfL willingly agreeing that London Overground should take part in such a process where it shares track with Southern. The lack of TfL co-operation would add to the complexity and limit the proposal’s effectiveness.
Oh Doctor Gibb, what have you done…
The other timetabling idea is to reduce stops at lightly used stations with an infrequent service. History has told us that this normally leads to a further decline in patronage and eventual closure. Gibb seems to want to remove stops to improve reliability, but the effect could be significant or negligible depending on station and the timetable in use. A blanket change would thus appear to be undesirable. Perhaps it is more a case of wishing to do something rather than achieve something. Whilst he probably produces a strong case with his example of Newhaven Harbour, which has another station nearby, it is disingenuous to categorise it with Southease – a remote, inaccessible village on the South Downs Way where numbers of passengers may be small, but the railway performs a valuable service to the community.
Ultimately, whatever the operational benefits of the proposals to reduce the service to some lightly-used stations at various hours, the result would likely to be politically unacceptable anyway.
Depots – the forgotten problem
Another operational feature that Chris Gibb dwells on is depots. Provision of depots is certainly an area on which, in general, there has not been enough focus. The trouble with depots is that they aren’t exactly ‘sexy’. It is thus hard to get politicians and the public to equate improvements to depots with better journeys. Indeed, a recent proposal to lengthen trains on Southeastern failed to get government approval not because of the cost of the trains, but because the cost of expanding a required depot was deemed too expensive.
With regard to depots, Gibb makes a number of points. The gist of these is that the depots are not in the right place. As a result, Selhurst is required to stable too many trains and is too cramped. He expresses his dislike of closing small depots and centralising train storage. In particular, he regards the decision to concentrate South Coast stabling at Barnham as a factor that led to increased union militancy, which does raise the moral question about whether an operator (or the DfT) should consider potential union power when formulating an operational policy. He is particularly scathing about the incredible amount of diesel light train movements (no passengers aboard) on the Uckfield line as a result of the trains being inconveniently stabled at Selhurst.
Uckfield branch, inevitably, gets a mention
On the issue of the Uckfield branch, the report goes to a lot of effort to put forward a case for electrification based on no further doubling of track, and on overhead electrification.
The report does an extremely good job of explaining why the current setup with diesel units is most unsatisfactory. He cites the considerable empty mileage run, environmental factors, train performance into London Bridge, the inability to swap trains at London Bridge to recover from delays (due to electric trains being unable to run to Uckfield), limited time for both track and train maintenance due to running empty trains back to Selhurst at night and a host of other things. Unfortunately this, along with another related proposal for the Uckfield branch, seems to be impractical.
A flawed solution
There is a fundamental flaw in Gibb’s proposal: he doesn’t seem to realise that most of his arguments are not an argument for electrification, but an argument to eliminate diesel working into London. He does not consider alternatives. So diesel bi-mode, Southern EMUs operated in push-pull mode with a diesel engine from Oxted to Uckfield or even pinning one’s hopes on future battery-electric bi-mode do not get a mention.
When considering electrification Gibb makes the fundamental (and common) mistake of quoting low-cost successful schemes and doing a few calculations on a mileage basis to come up with a figure he thinks is plausible. One would not expect someone of his standing to fall into this trap. Basically, overhead electrification consists of three elements which tend to each average out over multiple schemes as each costing roughly a third of the total cost. If not done in conjunction with signalling then there is probably also considerable costs related to immunising the signalling. At this stage, we haven’t even considered the cost of the electric trains.
The three main factors involved in electrification are wiring the line, enabling clearances in bridges and tunnels and – easily forgotten – making sure that the necessary extra capacity in the national power network (“the grid”) is there in the first place. The proportion of the total cost of these three elements can vary considerably from scheme to scheme.
In the case of the last consideration, unfortunately, in this part of rural Sussex the power supply is limited and what is available is already accounted for. Any apparent spare capacity disappears when you take the requirement for resilience in the network into account. There simply isn’t the extensive, dense housing or the energy-demanding industrial processes already present that would mean that the necessary easily-upgradable infrastructure was available for a cost-effective solution. The cost of getting the necessary high voltage lines to a suitable location would thus be high and the railway would bear all (or most) of that cost.
A further consideration is that politicians are likely to run a mile from further electrification until Network Rail can get costs down. The suggested alternative of someone else doing it does not seem very realistic either – especially as instances of other people claiming to be able to do something cheaper than Network Rail have usually not ended well. In any case, if there are limited resources, it is hard to see an existing delayed project being delayed even further in order to electrify the line to Uckfield for its 2tph peak and 1tph off-peak service.
Gibb would have been on much firmer ground if, having emphasised the benefits of eliminating diesel traction, he had recommended further investigation into the practicality of electrifying the Uckfield branch using 25kV AC with no change to the existing track layout.
A stable solution
Gibb also proposes a depot, or at least stabling facility, at Crowborough – two stops short of Uckfield. He points out that this would do a lot to reduce the considerable empty mileage runs and provide more space at Selhurst. That in itself would have beneficial knock-on effects. This does seem a good idea, but then he seems to think that the depot can be built and operated using the existing ground frame “until such time as the route signalling is renewed.” What is questionable is just how practical this is, given that no resignalling is on the horizon and normally a ground frame means that no other train movement can take place over the adjacent main line track in the relevant signalling section. The 25-mile long route from Hurst Green to Uckfield has just six signalling sections.
With the current setup, there is reportedly a four-minute delay between the signalman at Oxted giving the release and the release on the ground frame being activated. Shunting is not a speedy process and drivers are no longer allowed to reverse trains. As a result, they have to walk the length of the train to the other cab. If two diesel units are coupled together on the Uckfield line it is impossible to walk from one end of the train to the other without going outside. Going outside means going down onto the track and observing safety precautions unless a suitable walkway is built. Changing ends in such a fashion is not a popular procedure in wet weather. It is even less popular when it is snowing and, from the sanctity of suburban London in summer, it is easy to forget how harsh rural weather can be in that part of the world.
Milton Keynes – East Croydon
One of the most intriguing ideas in the report is the idea of handing over the Milton Keynes – East Croydon service to London Overground. Again, much of the reasoning for this is that the service causes a strain on Selhurst depot (technically Norwood depot, which is adjacent to Selhurst depot) with drivers having to have a lot of extra route knowledge and also knowledge of overhead wire procedures. Gibb points out that this route does not fit well into GTR’s operations. With some people arguing the GTR franchise is too big, this must be an obvious route to remove from it.
Again, we have a slight breakdown in logic in that, whilst Chris Gibb produces a good case for removing the Milton Keynes – East Croydon route from GTR, he fails to explain why it should go to London Overground. Milton Keynes is halfway to Birmingham and getting London Overground to run it does not seem to be a logical fit. It is a bit surprising that he did not suggest running this as a micro-franchise which operated just this route. He provides suggestions for depots but this, in many ways, just highlights the depot problem we have in London.
What the report does not mention is that there are often proposals for the Milton Keynes service to be cut back to Watford Junction. If that were to be done then it would make a lot of sense for London Overground to run it, as many of the stations served are already TfL managed stations. Whether TfL would want to run the service is another matter. The service is currently hourly so it does not really fit in with TfL’s turn-up-and-go model. It is an awkward service to run as it frequently gets delayed and it is not easy to recover from delays without affecting other services. Indeed one of the only ways to recover it is to miss out stations on the least served part of the route. These stations, between Willesden Junction and Clapham Junction, are the very stations that, ideally, you want to ensure have a reliable direct service with stations further afield.
There is also the danger that, with TfL running it, the penalty regime could mean that, as on the route to London Bridge, the London Overground trains get a higher priority between Clapham Junction and East Croydon. This would mean that there would be further disruption to Southern metro services.
It may be that, by recasting its existing services, London Overground could integrate a Watford Junction – Clapham Junction service, with some or all trains extended to East Croydon or elsewhere, at a higher frequency than today. However, we are really unlikely to find out as long as Grayling is Secretary of State for Transport, because his opposition to London Overground, and especially London Overground services running beyond the London boundary, is well known.
The report quite reasonably focuses considerable attention on Gatwick Airport and the Gatwick Express service. The suggestion is made that the old underpowered class 442 trains with just two single leaf doors per 23m carriage, completely unsuited to the Southern network, are withdrawn. This was an existing objective and has already happened.
The report makes much of the differential fares at Gatwick Airport and the problems that these introduce. Again, well-known stuff that many have commented on before but it is good to see it emphatically pointed out yet again in this report. Similarly, the well-known problem of extended dwell time on a rainy day caused by the platform canopies not extending the length of the platform is pointed out.
What is bold and different is the suggestion that the airport station should be sold off to Gatwick Airport. Whilst an intriguing idea, it is hard to see this happening. It has not been established that the Airport would want to buy it and it would seem to be almost impossible to rationally establish a fair price.
One concern with selling off the airport station would be that Gatwick Airport Ltd would probably not be keen to carry out work which improved reliability generally but had very little to specifically offer the airport. Another concern would be that once sold, it would be very hard to regain ownership if things weren’t working out. The debacle over access charges to Heathrow that Heathrow Airport wanted to impose on Crossrail shows the potential for conflict.
A plethora of recommendations
The Gibb report is full of recommendations and insights and in this short assessment, we have only been able to cover a few of the more notable of them. In the report, there are 38 specific recommendations (Appendix 11) with the majority requiring action by the DfT. Whilst we are sure some will turn out to be impractical, others politically unacceptable and others somewhat non-specific, there are a lot of interesting ideas and explanations for why we are in the situation we are in.
The Gibb Report comes across as a report that provides much food for thought even though not everything mentioned in it would be practical – or politically acceptable. Nonetheless, the government has made a good start with addressing the backlog of maintenance. It remains to be seen how many other recommendations are acted on.