Turning around Britain’s “worst railway”?

You’ve been in the job for just five months, and tasked with turning around the London south-eastern section of Network Rail – dubbed by some as Britain’s worst railway. What do you do?

Andy Derbyshire, recently appointed as the route’s Chief Operating Officer, has given an outline of plans to turn around the performance of the region, which has been beset by delays and disruptions.

The area has been massively affected by the London Bridge redevelopment, which has affected around 60% of Thameslink/GTR traffic, but also by a 40% surge in passenger numbers over a mere 5 years.

With the best will in the world, it would be difficult for any railway to expand capacity to cope in so short a timeframe, but the fact that capacity at London Bridge was shrunk over the same time frame only compounded the problem.

Passengers standing at London Bridge staring at a wall of cancellations and delays may be surprised to learn that this region of Network Rail has fewer incidents than other parts of the country, on a per-passenger-mile basis. However, the sheer volume of trains in the region, and other issues make recovering from a problem a long process, and that’s what causes all the delays.

An example cited was last December’s signal fault in the Balcombe Tunnel, which occured at 8:15am, but took until after 3pm to fully recover the network so that the correct trains were in their correct places.

The key issues they’ve identified shows, unsurprisingly, that Thameslink and South Eastern train punctuality and reliability is inconsistent. Network Rail has concerns about the Thameslink project for 2018, and that the current timetables are too complex for such a busy rail service.

All these combine to an operation which spends most of its time fire-fighting, and lacks resources to plan ahead more strategically. As he said, they need to “get brilliant at the basics” before they can fix the deeper issues.

The network is essentially a metro railway service, with very high intensive service that is closer to the London Underground in volumes of traffic, than a national railway service.

In that sense, it’s potentially lucky that one of London South East newish Chief’s Operating Officer’s previous jobs was with London Underground, where he worked on driving up the reliability of the service.

On such a complex network, there’s no chance of a single big solution, but a cluster of related issues that need to be looked at — what is being called the Galaxy Plan.

Key to its success is data — the collection of information to drive decisions that can prevent problems occurring, or least ensure that repairs are carried out faster. They’ve hired a whole team of data geeks just to look at this issue.

It’s possibly surprising considering how many thousands of miles of cables run alongside the railways, but there is less feedback to central about the infrastructure than you might expect, save when it breaks. Improving this performance feedback will be critical to helping the railways identify where problems are building up before they break.

Improving information, and standardizing how issues are defined will make it easier to know where problems really lie, and then to focus the maintenance in those problem hotspots.

In essence, what is being developed is a single joint action plan between Network Rail and the train operators, with a single agreed model of what constitutes a delay, in this case “delay minutes”.

If everyone talks about problems using the same parameter of what a delay is, then it becomes easier to work out what needs prioritising for repairs, or preventative maintenance.

The final version of the Galaxy Plan is due to go live with Thameslink at the end of October this year.

This is critical ahead of works in 2018 for the final stages of the Thameslink upgrade through the core of the network – central London. Part of the problem has in fact been down to the lack of investment outside London, and delays outside London can cause delays in the centre, meaning the benefits of the Thameslinke upgrade are not fully realised.

All that investment in the centre, and there could still be problems in London, caused by faults in the regions.

Therefore, a slug of £300 million has been recently handed over by the government to carry out upgrades along the rest of the line. It was candidly admitted that they really need at least £700 million, but there simply isn’t the engineering capacity to spend all the money at the moment.

There’s also a limit on just how much disruption you can ask the passengers to put up with. The travelling public need to see some sort of end in sight, so sometimes upgrades are moved or delayed just to give the passengers a break.

A look at the delays that passengers suffer from shows that around a quarter are due to GTR, a third to Network Rail, and the rest are known as “sub-threshold”, essentially too short to merit a formal investigation.

These sub-threshold delays are often not investigated as they don’t always trigger performance penalties, but it has been acknowledged that focusing on them could deliver a considerable drop is small delays, even though in purely financial terms, they are not a priority.

But they can be an issue, when trains run to such a tight timetable as the do in London, a small delay on one train can cascade down the line to the point that it reaches enough of a delay at the end that drivers and trains are in the wrong place, and there’s no spare capacity to move them to the correct locations.

Delays can be caused by the railway, or by a person running for the train and getting their bag stuck in the doors.

A short delay on one train leaving a platform can mean long delays and cancelled trains half an hour later.

To prevent small delays becoming big ones, the railway is looking to increase its fast response teams. Based on experience at London Underground, where the response team can be larger than the daytime maintenance team, the aim is to have the same impact — faster repairs when problems occur. As with the above mentioned Balcombe Tunnel incident.

Another option that is being considered is upgrades to track that will allow trains to run slightly than is currently permitted, but not then selling that capacity on to the train companies.

This sounds counterintuitive, but if the operators have space to squeeze in an extra train, they might want to do so, but unused capacity means that a potentially late-running train has the ability to run a bit faster and catch up lost time.

All these are about reducing the delays that cascade along the network.

Another planned upgrade is the addition of Automatic Route Setting (ARS) which enables the network to redirect trains using a semi-automatic plan that speeds up planning how to get around a problem, so that the railway delays can be minimized.

However, Andy Derbyshire repeatedly turned to what he evidently sees as one of his biggest problems — the timetable.

In his eyes, it is far to complex for a metro style service, and while the complexity may not in of itself cause delays, it makes recovering from delays that much harder.

A simpler service based on more uniform types of train lengths and routes makes it easier to reroute trains when problems occur. No one wants to see an 8-car train arrive when a 12-car one is expected, but that happens at the moment if the correct train cannot arrive at the correct location.

The other issue is that the timetable is based on an idealised railway, and doesn’t take into account the realities of the service, whether caused by infrastructure problems, or passenger actions.

There are consultations ongoing at the moment to review the timetables, but it’s clear from how much emphasis he put on them, that a lot more work can be expected, and a lot of changes made.

Look forward to hosts of outraged headlines when familiar commutes are shuffled around to fit a new simpler timetable when it is ready. Particularly look out for fewer split trains, if they can be avoided.

In the end, the aim is to have a railway that needs, as he put it “fewer boots on ballast”, because there’s fewer problems to fix.

That’s the medium to long term aim — a railway that just works, and where delays are not regular headline news. But it’s likely that passengers will have to put up with a lot more engineering works to get there.

On the vexed issue of who controls the railways, and whether TfL should have a hand in them, he diplomatically observed that it doesn’t really matter whose logo is on the trains, the plans to improve their reliability would almost certainly be the same.

To a degree, fixing delays on a railway is a thankless task as doesn’t have a “big red button” day where the railway switches from bad to good overnight. It’s a thousand tiny improvements each of which is barely noticable, and because they phase in many years, we the passengers barely notice the improvements.

In a decade, we’ll still moan about delays as if nothing ever changes, but will have forgotten just how much worse it used to be.

—

Andy Derbyshire was addressing a meeting of the Institution of Engineering and Technology in London, on 8th February 2017.