Signalling investment: the next six years

There is growing concern about the age of the many signalling systems that were commissioned during the 1970s, 80s, and 90s, and which are becoming increasingly difficult to maintain. Not only is the technology becoming outdated but getting spares and replacement parts can be difficult, particularly the computer driven elements.

The problem is recognised but how to resolve it is causing heads to be scratched both inside and outside the immediate rail industry. An analysis of the options was recently given to the London & SE section of the IRSE by Ian Maxwell and Nicola Machado from the Office of Road and Rail (ORR) who will advise the government on the level of investment required. Ian has worked in the signalling business for over 40 years, so is well aware of the dilemma, and Nicola has 17 years of experience.

Funding periods

As is reasonably well known, funding for the railway from the government comes in six-year chunks, known as Control Periods (CP). The present period (CP6) ends in 2023, whence a new period begins in January 2024 lasting until 2029 (CP7). The amount allocated for operation, maintenance and renewal of the network is £43.1 billion, a very substantial sum which gives the industry a degree of certainty into the future. The aspiration for this expenditure is initially put forward by Network Rail which compiled its wish list early in 2023. The ORR then analyses the content which includes evidence of spend patterns in CP6, before a final sum is agreed. The amount has recently been made public.

The ORR will hold Network Rail to account as the period progresses. Everything possible must be done to demonstrate the efficient spending of the money and obtaining good value. The £43.1 billion covers all of the UK but because of its greater devolvement, the Scottish element is £4.8 billion for which they have their own decision-making powers as to how the money is spent. John Larkinson, the ORR’s chief executive, is on record as saying that Network Rail must concentrate on improving performance for both passenger and freight operations, focusing on renewing core assets and achieving efficiencies.

Signalling options

The long-held view of the ORR is that traditional signalling with lineside signals is too expensive, and that better asset sustainability, capacity, safety, and reliability can be achieved by adopting digital signalling. However, it is recognised that if this is to be the way forward, it must not be a stop-start process. The concept is not universally supported, and Transport Scotland have a different approach more aligned to conventional signalling and life extensions.

How to introduce digital signalling is the big question and it is complicated. It would be nice to shut the railway for a long period of time and do the conversion in one big bang but clearly this would be unacceptable. The logistics are mind blowing. There is 19,000 miles of network and over 4,000 trains so any conversion will be very complicated and requires enormous co-ordination. There is a Network Rail master plan – see issue 202 (May/June 2023) – but murmurings are occurring that this is not bold enough and the suggested timescales must be improved. The real need is for a Rail Network Enhancement Pipeline (RNEP).

The digital railway effectively equates to the adoption of the European Train Control System (ETCS). As has been previously recorded, ETCS is already in operation on the Cambrian Line (commissioned in 2010, some 13 years ago); on the central core of Thameslink (since 2018); on the Heathrow spur from the GW main line which will eventually extend back to Paddington; and the soon to be commissioned suburban section of the ECML from Finsbury Park to Moorgate. The rest of the route from Kings Cross to Stoke Tunnel (near Grantham) is not due to be fully commissioned until 2029 under the banner of the East Coast Digital Programme (ECDP), although parts of the route will be in service before then with Welwyn to Hitchin being the chance to prove how the interfaces between conventional and digital signalling on a main line will operate. Thus, progress is being made but it is not exactly speedy.

A significant challenge is the fitting of rolling stock; this is both expensive and time consuming. A captive fleet exists on the Cambrian, the Class 700 Thameslink fleet is equipped although this has to undergo an expensive upgrade to meet the latest software specification, the Class 800 and variants ICE fleet are fitted as is the Class 717 GN suburban trains. Gradually the freight fleet with an essential requirement that the locomotives must be capable of going anywhere, will be equipped.

The process requires a ‘first in fleet’ development, then time out of service to fit the fleet, all of which takes a lot of time and costs a fortune.

What will it cost?

There will be different ways of funding. For England and Wales, the money will essentially come from the DfT and in Scotland if it ever happens, from the Scottish government. Digital signalling is technically a renewal of assets in modern form but in reality it will mean a totally different way of operating the railway. The draft determination from the ORR covers infrastructure renewal (£742 million), fleet fitting (£699 million), enabling projects (£121 million), CP6 legacy projects (£60 million) and training with the cost yet to be determined. This equates to £1.7 billion but Network Rail will be asked to reduce this to £1.5 billion. Scotland will be asked to help fund fleet fitting of trains that cross the border.

All of this is to reduce the cost of Signal Equivalent Units (SEUs) although exactly how these will be measured in a digital world is not yet determined.

The next steps will be for the ORR to discuss the proposals with Network Rail which is already developing a procurement strategy called the Train Control System Framework. Whether in all of this the Network Rail master plan for ETCS deployment is modified, remains to be seen. The fitting of fleets linked to lines of route that are to be prioritised has to be part of the plan. Indications from Network Rail recently suggest that after the ECDP Phase one, extending northwards from Grantham, the Transpennine Route and the WCML north of Warrington will be the candidates for digital signalling.

The Scottish perspective

At another IRSE event in Glasgow, the question of signalling finance was raised again, with much the same message as for England and Wales but with different thoughts as to priorities. Dan Holder and Lynsey Hunter from the Network Rail asset management team confirmed the need to bring track and train together, which is already underway as Scotrail and Network Rail are already in alliance now that the former is under Transport Scotland control. Improved joining up of projects and any associated rolling stock opportunities should lead to sustainable growth in traffic and minimum customer disruption. Renewals are more difficult with much of the signalling infrastructure approaching life expiry.

Scotland’s rail geography has been split into eight Lines of Route:

• Dundee – Aberdeen.
• Fife.
• Edinburgh area including the Edinburgh Glasgow main line and the ECML.
• Glasgow Central including the WCML.
• Glasgow North Electrics.
• South West including the routes to and onwards from Kilmarnock.
• Highland main line to Inverness.
• Far North, West Highland and Stranraer rural routes.

Each will need a business case for signalling renewals or upgrades. The long-term objective is to have 12 signalling control centres for the whole of Scotland. Traffic management using the Luminate system is being deployed and is demonstrating improvements in train regulation and performance.

For each of the lines of route, there will be a route strategy that considers various options. It is clear there will be no single solution for every route, but the options will take account of likely technical changes and adaptability to changing scenarios. The money spent will be compared to the benefit gained. Of the routes considered, Glasgow central and the WCML gives the worst result, and it may be that the solution here will be ETCS since that route is high priority in England. The Aberdeen route needs to be upgraded with a conventional signalling renewal costed at £300 million. Perth Dundee is another early candidate which would include remodelling of Perth station. In both of these, immunisation ahead of electrification has to be part of the plan.

In the Fife area, replacing the Aster track circuits due to unreliability is a priority. The Glasgow North electrics running from Helensburgh to Bathgate has signalling infrastructure, interlockings, and control panels all reaching the end of their economic life, but a renewal strategy must not fall into the trap of inefficient whole life costs and partial renewals. If combined with new rolling stock, the advent of digital signalling could well be favoured.

In all of this, it must be remembered that passengers do not care as to what signalling system is used, just that the train service is regular and reliable. Tourism is major business in Scotland and serving this market is as important as commuters and intercity travellers.

The radio factor

As has been pointed out in previous articles, the adoption of a digital railway must have a robust radio bearer if the train control system is to be reliable. The ageing GSM-R technology is only guaranteed for support from the radio suppliers until 2030, thereafter a decline in its continued operation can be expected.

Moving to the new European standard of the Future Rail Mobile Communication System (FRMCS) as described by Paul Darlington in issue 204 (Sep/Oct 2023), has to be part of the progression. No mention of this is made in the ORR or the Transport Scotland financial analysis, but the need is recognised. It’s going to be interesting seeing how digital signalling and the associated track to train radio upgrade actually takes place.

Image credit: Siemens Mobility / Network Rail