Insight from Richard Lungmuss
Signalling has been in the news recently for the wrong reasons. The power cut at the North West Rail Operating Centre (ROC) in Manchester in August caused havoc to services across the region despite the use of Uninterruptible Power Supplies (UPS), when systems did not reset as designed.
Having been the RD in York in when the concept of 14 National ROCs was being implemented, with the re-control of ALL signalling boxes into the new centres, I was interested to see what progress had been made in doing so, especially in reducing the signaller workforce, which had been around 5000 nationally when I left Network Rail in late 2011.
I was heartened to see that there are now only around 80 of the original 845 or so signal boxes still in operation but disappointed to read that there were now more signallers than there were when all boxes were in operation. It appears that the original business case for signaller headcount in the ROCs had not taking into account rostering, sickness, leave etc., which feels like another efficiency saving that has been lost in implementation! (I am also significantly concerned that our ROCs may not be ready for the type of cyber attacks that we have seen across Europe in recent days, and which we can certainly expect).
My origins on the UK railway were as the General Manager delivering the Train Protection and Warning System, designed to stop trains going past signals at danger, from over speeding on specific sections of track and to prevent buffer collisions at terminal stations. Whilst involved with the system between 2001 and completing delivery in Dec 2003 I was aware that a more advanced project the European Rail Traffic Management System (ERTMS) was being developed comprising the European Train Control System (ETCS) and a Traffic Management System (TMS).
By the time I left Network Rail a trial of ETCS was running on the Cambrian line and we were being briefed about the rollout of the system on the southern end of the East Coast Main Line (ECML). Whilst adding significantly to safety the system could be expected to increase capacity by as much ass 40% on plain lines and could increase performance. Furthermore, since signals are not required when track balises are being used there should be a significant cost saving in both deployment and ongoing maintenance.
Fast forward to 2026, some 25 years after I first saw Network Rail staff working on ERTMS(!), and we have reached the point where ETCS has been deployed on the southern part of the ECML. Further deployments are planned, eventually reaching Fletton to Stoke near Grantham in the early 2030s. Eventually it is hoped that ETCS will be deployed throughout the UK, although is it hard to see that this will be before at least the late 2040s. Where ETCS is being deployed all traffic must be equipped with in cab ETCS onboard equipment, necessitating significant time and expense in doing so. For new rolling stock the cost of this will have been included in design but retrofitting stock, e.g. tampers and other OTMs, freight locos etc. will be expensive, estimated at up to 900,000 euros per vehicle by European operators. In the initial stages signals will require to be present too, so the maintenance requirement might actually increase.
As the cost of the Trans Pennine Route Upgrade has shown so graphically, deployment of ETCS will often require upgrading of the infrastructure. One of the benefits of GBR is that the total costs of fitting rolling stock and trackside ETCS equipment will come under one organisation, making the actual amount far more visible to those with financial responsibility for delivering track and trains. Network Rail have made ETCS rollout a feature of their CP7 delivery plan, where they cite ETCS rollout as cheaper to deploy and maintain than existing signalling, meaning they can “replace more life expired signalling in each control period.”
Scotland meanwhile have taken a completely different view on ETCS, which is not included in their CP7 priorities. Basically Scotland considers that the overall costs of ETCS (including retrofitting to existing stock) outweigh the benefits. Scotland’s pragmatic approach [and readers may remember from my article a couple of week’s ago that Scotland has a different – and less expensive – strategy for rail electrification] is that, although recognising its benefits, they will only deploy ETCS if it affordable. For that to be the case:
- Transitions must avoid any disproportionately expensive retro fitment to existing stock
- The overlay of ETCS on existing signalling is affordable
- The deployment must not reduce the flexibility to deploy existing stock (and drivers I assume?)
- Processes to improve software updates as a result of infrastructure improvements must be made far less expensive
I am left with the feeling that once again we are implementing a system (albeit incredibly slowly and at huge expense) which is not going to deliver the savings that are expected and which may prove burdensome as operators try to come to terms with it. There are huge administrative burdens of regulation and authorisation in delivering such a safety system on an operating railway and rightly so, hopefully these can be streamlined for efficiency.
If the stated benefits of ETCS are to be achieved then the rollout should either be postponed until such times as it can be delivered with actual demonstrable efficiencies or we should ramp up a National rollout and set some really challenging targets. Maybe at the same time consideration could be given to 4 tracking the entire ECML from London to Newcastle so mixed traffic can be run properly and the true potential of 140 mph Azuma trains can be exploited. I wonder too whether, if ETCS and TMS was deployed on the “improved ECML” and the WCML, HS2 would be such a key capacity requirement? Perhaps it is truly “either or?”
Throughout my research for this piece, I would like to acknowledge an excellent article in Rail Engineer, written by David Shirres in 19 June 2025, which you can read here.
