Thursday, 12 July 2018

Thameslink Timetables and Traffic Control

Silence is Broken

Perhaps I’ve waited too long but I have watched the continued disruption to the train services operated by Govia Thameslink Railway (GTR) with dismay and despair. I thought, after three weeks of operation, they would have got it sorted out, at least to a level where some sort of regular and reliable train service could be operated. But they haven’t. Here we are, in the 6th week of the new timetable and there are still serious problems with the train service.

Why three weeks to get it sorted? Because, if you consider the average 24-hour shift pattern of early, late and nights, spread over three weeks of operations, it should allow most of the core staff to get used to the new pattern of services, the new train crew duties and the new routes. But it hasn’t happened like that.

I’ve stayed silent so far as I thought I couldn’t add anything useful to the huge range of complaints, blame, calls for the stripping of “franchises” and largely uninformed nonsense from the popular press and social media. In vain, I waited for some sensible comments asking some basic questions about how the service was planned and how it was controlled on a day by day basis but none came. I waited in vain for a proper analysis of the issues and solutions but there were none. So I’ve decided to break my silence and offer some basic ideas, based on my own experience of timetable preparation in various places around the world and of train service operations.

The Routes

The Thameslink routes are too complex. The old setup with a Bedford to Brighton main line with the Sutton loop tacked on was quite complex enough, without adding such destinations as Peterborough and Cambridge in the north and Orpington, Sevenoaks, Horsham, Rainham, Littlehampton and East Grinstead in the south. In the originally planned December 2019 timetable, they were going to add Kings Lynn and Ashford too. This is not going to work well.

It can easily be solved - just run to two destinations each side of the central core. Bedford-Brighton and Peterborough-Sutton. Yes, you could run some short workings in the peak Luton to Three Bridges, for example, but no new branch line destinations.

Lots of new destinations may look very attractive to passengers and to timetable planners who want to demonstrate the pinnacle of their art but they only need a minor disturbance, like a PIOT (passenger ill on train) at 08:00 in the morning, to throw the whole fragile construct into complete disarray. This is demonstrated every day on the Thameslink routes, which have suffered from such things as late surrender of possessions, heat related TSRs, signal failures and train breakdowns with frustrating regularity on top of the already recognised acute shortage of drivers. The service patterns need to be simple to allow rapid recovery.


The people who operate the trains are, not surprisingly, essential for the operation. This might seem an obvious statement but they do seem to have been rather neglected in the lead up to the introduction of the new timetable. They weren’t given enough time to learn the new trains and the new routes, they were supposed to have ATO operation to learn as well and many of them were moved from one depot to another at the same time. Each one of these events is enough to manage on its own without trying to do them all at once. Then, as if all these weren't enough, the attempt to introduce driver controlled doors (DCO) on Southern routes at the same time, with no agreement with the staff to do it, and well, what could go wrong?

Well, everything did go wrong. There was no agreement on DCO, training for trains and routes wasn’t finished in time and there had already been weeks of disruption due to industrial action and the problems with rebuilding London Bridge station. There was also a shortage of drivers, with some refusing to do overtime because of the industrial dispute.


Running a train service on overtime is a recipe for trouble, as is voluntary Sunday working. The full 7-day week roster requirement should be written into the train crew contract. It will have to be paid for but it will pay back many times in train service reliability and the reduction in income loss. Negotiations with staff reps need to be started immediately.

All rest days and Sundays should be written into the contract with suitable rest day cover duties and holiday leave covers rostered for all fixed leave allowances. This will require about 25% more drivers than the basic rostered train running duties. Anything less will result in an unacceptable level of uncovered duties.


A service as intense as Thameslink must be supported by an adequate number of spare crews.  This will require another 25% staff on top of the train running duties. Spare crews are used to cover lateness, sick leave, casual vacancies and disruptions. Again, skimping on spares will make ensuring an adequate service very difficult.

In addition to rostered spare duties, there should be a pool of drivers ready to take over retirements, special leave (for bereavement, incident recovery, enquiries etc.) and promotions. There should also be a number of drivers on training, either for rule book and traction refreshers or as new drivers transferred in or being qualified. The actual numbers will be at least 10% on top of the running duties but may be more depending on the historical wastage rate.

It is worth remembering that the sickness rate amongst train drivers is about 10 days per year, as opposed to the national average of 4 per year. Train driving is a stressful job, with long periods of concentration and occasional moments of high anxiety. Boredom and loneliness can be an issue.


I’m referring here to train crew depots, where drivers sign on for duty and where they can change, shower, get updates for route information, get help from supervisors and so on. They may be in the same location as a train storage or maintenance depot but not necessarily. From my experience, small is beautiful. Despite the perceived efficiency of a large depot where one set of facilities and offices can be based, much time can be wasted with crews riding on trains or in taxis to locations to meet or collect trains, or to return to their depot on the completion of their duty. This is because the depot location may not have all the trains required for the services being operated and because, once trains are out of the depot, crew changes can normally only take place at a station. That means that a replacement driver has to travel to a pick up point away from the depot where he or she signed on.

This brings me to another point. It is best to change crews only at a terminus. Then, if the replacement driver is not available, a train is not stuck at a platform while a queue of trains forms behind it. It a timetable is properly designed, each terminus should have at least one vacant platform at all times to allow for a train to be temporarily stored while a new crew is found or a technical fault is remedied. Of course, this may not always be possible, but it should be a goal. A siding near the terminal is another option.

I recognise that changing crews mid-trip is often necessary, particularly on long routes, but some elementary precautions should be adopted. First, keep a spare driver or two at each location in case of disruption. Second, choose a relieving location where there are spare sidings or an extra platform available  to store a train that finds itself without a crew. Third, provide facilities that encourage crews to be in the pickup location on time in advance of their train’s arrival. A small booth at the arriving cab end of the platform to protect staff against bad weather is very useful. It should have a phone, a real time running display to show where trains are, a water cooler, heater and air conditioning. Providing these would show that management cares that their staff have, sometimes, to wait in unpleasant conditions and it makes sure the driver has no reason to delay being at the cab when the train arrives.

To return to my theme of small being beautiful, the advantage of a small depot is that everyone knows everyone else and a sense of community develops. This means drivers try to help each other out and, with a good supervisor, they will try to help them too. A certain amount of bargaining can take place and things can be sorted out as if “between friends”. This doesn’t work so well at larger depots, where there are many staff and small groups forms separately from others. The sense of overall cohesion is lost.

Casual absenteeism is often lower at smaller depots too. This is because someone tempted to take a day off usually knows who will have to cover for them and will hopefully not want to put a known colleague on the road unnecessarily. Regular absentees soon get to be known.


All that a controller wants is a train service to operate according to the timetable. If it doesn’t though, he (or she) needs all the help they can get. There are certain basic tools and information that fall into the essential category:
  1. A real time train running display of all the routes under their control and of the trains approaching them.
  2. The identity of the train type and length. If the job is in recovery mode, it will not help to send a 12-car train down an 8-car only route, or an electric train down a non electrified track.
  3. The identity of the crew duty on each train and the details of where the crew changes are due to take place. This must be in real time. Just tagging each train ID with the rostered duty won’t work if the driver on the train is a spare or was put on the train as part of a recovery scenario. It is also necessary to know the ID of the next crew due on that train and whether they are available. All this information is needed so that controllers know if the crews are available for a train to complete its full trip. The controller also needs to know how much time each driver has left before he runs out of the time that he can be on the train and if this is he end of the duty or the PNB (Personal Needs Break). A train should not be allowed to depart a terminus in a trip unless the crews are known to be available for it to complete the trip.
  4. Where trains have to be for maintenance, cleaning, sanding and CET disposal. If the service is disrupted, the controller needs to know which trains must be in a particular location for servicing. Not all will be so there might be some opportunities to reform train diagrams to cover missing services. Formation should be allowed freely outside of these constraints. 
  5. The status of service intervals at each station in real time. This is essential where some trains stop and some pass through a station. If a station is supposed to get a 20-minute service and several trains have been cancelled or diverted, the controller needs to stop a fast train or get an extra train inserted. 

Naturally, not all of this can be controlled by one person but the information must be provided to enable a team, working together, to use it to coordinate train movements and crew utilisation to get the service back on book as quickly as possible.

Fallback Timetables

If serious disruption has taken place, it is wise to have a fallback timetable. This might, for example, remove fast trains and make all trains stop at all stations. It may include a simple end to end service for each route at a reduced frequency. It will require a fallback crew duty roster as well or an agreement that all crews work one trip at a time end to end until normal working is restored.

Fallback timetables are particularly useful for evening working at the end of a day of disruption. It is then that train services tend to fall apart as a result of crews being displaced at the end of the evening peak and not being able to complete their duties due to running out of hours.

A fallback timetable might also be used if there is a technical issue with the infrastructure that cannot be fixed quickly or that requires a temporary possession.

Staff Management

We forget the staff at our peril. Too often, the gulf between staff and management is left to widen as changes in the management strategy and the implementation of new ideas or systems are only communicated to staff through the trades unions or company notices. There is too little contact between managers and front line staff. We (I mean those of us who worked on the front line), used to regard a contact with a manager as only being when you were in trouble. This is wrong. Front line staff need to see their managers on a regular basis, otherwise the impression of neglect and not caring becomes widespread.

Managers need to get out and about. They should ride on trains, visit stations and talk to staff and passengers. A good railway manager should feel the heartbeat of the railway and should know instantly if things are working well or not. It takes hard work but it has to be done and it's why managers get a bid salary. They need to earn it.

1 comment: