Network Rail’s report ‘CP7 delivery plan – Our approach to digital signalling (OADS) in CP7’ states that “ETCS (European Train Control System Level 2) is cheaper to deploy and maintain than conventional signalling and requires less infrastructure out on track,” adding “this means we can replace more life-expired signalling in each control period.”
The OADS report also states that ETCS improves the safety and performance of the network. However, it acknowledges that, although the capacity benefits of ETCS have yet to be quantified at a network level, ETCS offers significant performance opportunities. For example, it avoids limitations associated with signal positioning and signal sighting.
It would therefore seem that there is a strong case for the ETCS programme, especially as Network Rail considers that the Signalling Equivalent Unit (SEU) renewal rate for conventional signalling is £420,000 and that for ETCS installation is £315,000. Yet elsewhere, the OADS report acknowledges that there are fleet modification and capability-building costs. It estimates that these amount to £0.9 billion.
No Scottish ETCS
The OADS report also notes that Scotland’s Railway has no plans to implement ETCS in CP7 as it has other strategic priorities. The priorities of the Scottish signalling strategy are to: maintain safety and operational performance; reduce whole railway system net cost; and avoid high unit costs and premature obsolescence. Rail Engineer understands that these are common priorities throughout Network Rail. Nevertheless, Scotland has no plans for ETCS.
It is therefore important to understand why Scotland’s railway considers that there are other, more cost effective, ways of achieving objectives that are generally common throughout Network Rail.
The Scottish view
Network Rail Scotland’s signalling strategy is described in Appendix B of Scotland’s Railway CP7 Delivery Plan which is entitled ‘Signalling Scotland’s Future’ (SSF). Although this identifies potential benefits of in-cab signalling on some lines, it also considers that large implementation costs currently outweigh the potential benefits.
A significant difference between the Scottish approach and that of Network Rail’s OADS report is that the Scottish report considers that the high cost of retrofitting a wide range of existing rolling stock to the current ETCS specification is disproportionate and far greater than money saved in signalling assets. It reaches this conclusion by drawing lessons from other in-cab signalling projects on comparable networks across Europe. The SSF therefore recognises that, in view of these affordability challenges, ways must be found to extend the signalling infrastructure’s lifespan.
Furthermore, the introduction of new stock may require costly ETCS infrastructure upgrades. This was the case when new Class 197 units were built for the Cambrian line. These have the CAF AURIGA ETCS system compliant with the latest specification. However, before they could be used on the Cambrian line its ETCS had to be upgraded to baseline 2.3.0d at a cost of £3 million.
The SSF report considered an alternative ETCS implementation strategy of restricting ETCS deployment to lines that only operate new fleets. However, as demand for passenger and freight services changes over time, restricting rolling stock deployment options on key lines of route was deemed to be unacceptable.
This report also notes that Scotland’s Railway delivers signalling at much less cost than the national average of £420,000/SEU. For example, on a recent project for Aberdeen and Central Scotland the average SEU cost was £356,000. This further weakens the case for ETCS deployment.
Furthermore, the SSF report notes that once ETCS has been implemented there are high costs associated with future infrastructure changes. For example, Rail Engineer understands that flood prevention works on the Cambrian line requiring the track to be raised required ETCS software changes costing in the order of millions of pounds.
Retro fitting ETCS
A new train ordered today with ETCS onboard equipment will probably use that supplier’s equipment and, broadly, the cost will be ‘lost’ in the overall cost of designing, building, testing, commissioning, and assuring the fleet. Retrofitting ETCS is another matter, and its high cost was recently highlighted in a European Commission (EC) report on ERTMS on-board deployment. This showed that, for a fleet of 20 vehicles, the average retrofitting costs have grown from €450,000 to €900,000 per vehicle, and upgrade costs have increased from €200,000 to €400,000 between 2018 and 2022.
In all the publicity about retrofitting ETCS to the modern steam locomotive Tornado, cost has not been mentioned. Yet Rail Engineer has been advised that it is costing around £10 million to fit ETCS to a Deltic diesel locomotive. It is also reported that it cost £11.3 million to design and fit prototype ETCS to a Class 387 Electrostar train operated by Govia Thameslink Railway. Hence the cost of a first-in-class ETCS retrofit is around £10 million. Of the 1,200 or so cabs that need to be fitted with ETCS equipment for the East Coast Digital Programme (ECDP), 700 will be retrofits for about 20 different classes of locomotives, on-track plant or passenger units with each class requiring a first-in-class design.
ETCS fitment will often require power supply to be reinforced (or provided from scratch), finding the required space and providing mountings for all the equipment and sensors. The EC report on onboard deployment considered that such fitment costs are only 50% of the total deployment costs with the remaining costs being design, validation and assurance. This illustrates that the cost for one off vehicles or very small fleets might be disproportionately high.
New trains may have ETCS or be ‘ETCS ready’ with available space and possibly the required wiring in place. Yet they may incur additional costs if the hardware and software have to be upgraded to the latest baseline, especially if ETCS has to be integrated into the vehicle’s train control and management system. Recently, it cost £33 million, or £175,000 per cab, to upgrade the Thameslink Class 700 fleet.
Network Rail aims for a gradual ETCS rollout around the country starting with the East Coast Digital Programme (ECDP) on the East Coast Main Line (ECML). The next phase is between Biggleswade and Fletton, south of Peterborough where there will be no lineside signals. Hence all passenger trains, freight locos, on-track machines, and monitoring vehicles on this section will have to be ETCS equipped.
Thereafter, Fletton to Stoke Tunnel near Grantham is planned to be completed in the early 2030s. This section will have a conventional signalling overlay so that trains on routes crossing the ECML at Peterborough and Grantham do not need to be ETCS fitted.
Hence the early ETCS programmes will require high numbers of trains to be ETCS-fitted and many of these will require retro fitment. However, in the long term, as ETCS is rolled out and new trains purchased a diminishing number of trains will require retrofit.
Making ETCS affordable
The Signalling Scotland’s Future strategy recognises the benefits of ETCS but considers it will only be affordable if:
- Transitions must avoid any disproportionately expensive retro fitment to any of the rolling stock which operates on the route.
- Overlaying ETCS on conventional signalling during a transition period is affordable.
- ETCS deployment must not reduce the flexibility of deployment of existing rolling stock.
- The process to alter safety critical software and data relating to the railway infrastructure must be improved so that it’s no longer disproportionately expensive.
Network Rail’s OADS report and the Scotland’s Railway SSF report reach quite different conclusions in respect of the affordability of ETCS. This is because the SSF report identifies factors that add to installation and upgrade costs that are not mentioned or are understated in the OADS report. In particular, the SSF highlights the high cost of retrofitting ETCS to existing stock which reflects concerns in a recent European Commission (EC) report. The OADS report allows £0.9 billion for fleet fitment which at emerging prices in the EC report is 1,000 trains or about a quarter of the UK passenger fleet, excluding freight locomotives.
Although ETCS offers various benefits, it is difficult to see how the UK railway can afford a large scale ETCS implementation programme unless the affordability issues that Scotland’s Railway have identified are addressed.
Editor’s note: Our feature “Planning for ERTMS and timetable implications” provides more information about Network Rail’s approach to the implementation of ETCS.
Image credit: Network Rail