Return to Archive -by date - by topic.

INFORMED SOURCES November 2007

ERTMS – a long time coming

It will be the middle of the decade after next before ERTMS is in wide scale use

On 3 October DfT Rail published its report to the European Commission on the implementation of ERTMS on the UK rail netwoek. This National Implementation Plan (NIP). Production of the Plan is a requirement of the European Communities' High Speed and Conventional Control Command & Signalling Technical Specifications for Interoperability (TSI)

Giving credit where it is due the NIP, is a hard nosed analysis with the timing driven by the life expiry dates of signalling and traction and rolling stock. As required by EU rules it covers the UK routes in the EU Trans European Networks (TENs). By 2038, when implementation is completed, ETCS will signal 72% of the UK national network.

Radio communications is a vital component of both ERTMS (European Rail Traffic Management System) and its subset ETCS (European Train control system). This will be provided, at least initially, by the railway specific version of GSM, known as GSM-R, currently being installed by Network Rail.

Provision of GSM-R for operational network-wide voice communications is due to be completed by 2014. It will thus be ready to province data communications when installation of ETCS Level 2 starts on the Great Western Main line in 2017.

It has to be emphasised that the plan is a ‘political' response to a European requirement. Don't forget that DfT Rail got a tetchy note from the EU asking why the West Coast Route Modernisation had not used ETCS. NIP has been developed with Network Rail and all the other industry players involved and represents the best view based on the information available.

Caveats

As such, it is, wisely, littered with caveats. DfT Rail warns that updates to the Plan are likely ‘in light of revised demand forecasts; challenges to test the underlying assumptions; changes in investment priorities of major infrastructure, rolling stock renewal and network enhancement programmes; technical developments; capabilities of a global supply market and operational experience gained by all ERTMS operators'. Quite.

Under the Plan rolling stock will usually be fitted with ETCS before trackside equipment. More than 50% of the train fleet will require retrospective-installation, as opposed to fitting during manufacture.

Table 1

Route Type	Route	Rolling Stock Fitment	Infrastructure Fitment
High Speed	Great Western Main Line	From 2013 to 2018	from 2017 to 2035 (majority complete by 2025)
East Coast Main Line	From 2013 to 2022	from 2018 to 2035 (majority complete by 2025)
West Coast Main Line	From 2014 to 2027	from 2027 to 2030
Channel Tunnel Rail Link	From 2025 to 2030	from 2038 to 2042
Conventional	Brighton Main Line	From 2011 to 2022	from 2021 to 2025
South West Main Line	From 2014 to 2025	from 2017 to 2034
Midland Main Line	From 2015 to 2027	from 2021 to 2023
Great Eastern	From 2015 to 2023	from 2027 to 2029

Table 1 shows the current programme. Chart 1 puts it in graphical form

Chart 1 - to be inserted after reformatting to a web friendly version.

Timing

During Control Period 4 (2009-10 to 2013-14)

the NIP assumes that like-for-like renewal of conventional signalling will continue. What the plan doesn't mention is that current interlockings are ETCS ready, with the capability to communicate with a Radio Block Centre built in.

However, CP4 will include two more small ‘migration' pilot schemes on the Great Western and East Coast main lines. This is in addition to the existing migration scheme to replace Radio Electronic Token Block (RETB) signalling in East Anglia .

Signalling Informed Sources agree that using ETCS Level 2 to replace RETB is expensive overkill and on a par with choosing the Cambrian lines for the ETCS Early Deployment Scheme. Everyone knows these two schemes are bonkers, but, predictably, no one in the industry is willing to warn the emperor of his nakedness.

Developing the main line Migration Schemes will not be cheap, or quick. In 2009/10 the single option development stage for the first, assumed to be Great Western, will be completed. This will cover the development and acceptance of the preliminary system design and standards, allowing the start of detailed design. The same process for the second main line migration scheme will follow on. Expenditure during CP4 will be ‘around' £80 million.

Assumptions

NIP also provides a comprehensive list of assumptions behind the dates. It warns that if any of the assumptions prove invalid, or change, ‘especially those with a cost impact', DfT Rail will have to review the economic viability of ETCS and update the Plan.

ETCS Level 2, without lineside signals, is assumed to be the preferred solution. It is also assumed that the signalling and other contractors can meet the programme, both in terms of volume and cost. This equates to delivering a 50% increase in re-signalling volumes.

Another key assumption is that the ERTMS products available by 2011 will have resolved current outstanding issues, known in the jargon as ‘error corrections and open points' Off-the shelf ETCS will also have to be able to provide UK-specific requirements, or ‘functionalities'.

Two of these are second order issues, though vital, namely level crossing control and door control supervision. The other two are a Cold Movement Detector (CMD) and the provision of enhanced data capacity in the communications system for high traffic density areas.

Cold start

Cold Movement Detection is all about starting up a train from ‘cold' under ETCS. Out on the main line an ETCS fitted train determines its location and direction of travel from the balises on the track. But what happens when it has been parked overnight in a siding?

Since the on-board kit has no location data, other than that received when passing over balises, the Radio Block Centre can't provide ‘Full Supervision' Movement Authority to a stationary train in a siding. ‘Full Supervision' is signal-speak for permission to run normally up to the end of the authority and within the speed profiles.

At present when a train in a siding contacts the RBC it receives a ‘Staff Responsible' Movement Authority. This permits the driver to run slowly until the train crosses a pair of balises. When these are passed the train radios its location and direction of travel and the RBC issues a Full Supervision Movement Authority, allowing the train to enter service..

Under such a regime getting a fleet out of the depot in the morning would be a time consuming process. Much better if the train could ‘remember' its location over night and prove that it had not moved since it was parked. Then the driver could set up the train, the RBC would be told that the train was where it was last reported and had not moved since and could issue a Full Supervision Movement Authority immediately.

This is the role of the cold movement detector, which would also solve the Howker conundrum of how you get the trains running again if ETCS goes down and everything comes to a stop, But it is not an easy requirement to satisfy.

Remembering location is easy. But how do you prove that the train has not been shunted with the power off (‘cold movement') into another siding during the night – as often happens in depots? A solution will emerge, but the answer is not yet apparent.

Radio data capacity is another technical requirement. Radio contact between the RBC and the trains in its area is not continuous. Using GSM-R the RBC polls each train in a cycle.

Each time it makes contact, protocols have to be followed to ensure that the train has been identified correctly before data is exchanged. While each transaction takes milliseconds, they soon add up. Try to handle a high density operation and GSM-R runs out of capacity.

So the NIP assumes that something with a greater data capacity will be available by 2016. That something is likely to be the General Packet Radio Service (GPRS).

Hardware

Now comes the big assumption - that ERTMS is available off the shelf, as a reliable product in a commercially mature supplier market with no major concerns on technical capabilities.

A related assumption is that any Specific Transmission Modules (STMs) required have been developed. An STM is the electronic box which enables ETCS to read existing Automatic Train Protection balises.

For instance, the NIP assumes that rolling stock is ETCS fitted before infrastructure to avoid running parallel systems. In the case of the Great Western Main line, for example, the IC125 replacement trains will be running while ETCS is being installed. Since ETCS will have been fitted during manufacture, an STM will be needed to continue use the existing ETP equipment.

A couple of minor assumptions worth noting are that retrofitting an existing train will take it out of service for 10 days, while new stock, built with provision for ETCS, can have the hardware installed within ‘a planned normal maintenance and overhaul regime'.

Risks

The key risks and challenges to the delivery of the (ERTMS)implementation plan are the implementation capability of the industry, delivering on the cost assumptions, European specification development, product development and maturity, engineering and operational integration into the current railway and commitment across multiple control periods and franchises.

ERTMS National Implementation Plan

September 2007

Also in the NIP is a brief run-down on the relationship between ERTMS and major projects such as Thameslink and the Intercity Express Programme. However, given that this column devoted several pages to ERTMS in September and that the Signal Engineers have what looks like the definitive ERTMS conference in November, I will put that part of the NIP on hold.

Wires

Surprisingly, my self denying ordinance on further coverage on electrification for the remainder of the year did not receive the approbation I expected. So I will take this opportunity to slip in a small reference.

According to the NIP, ‘the assumption being made in DfT is that any major electrification scheme would follow ERTMS deployment on the route in question'. So let's apply that to the prime candidate, the Great Western Main Line.

Installation of ETCS will be largely completed by 2025. So, assuming that bionic duckweed vats aren't producing industrial quantities of hydrogen by then, this is what we are looking at.

Delivery of new diesel trains for Great Western is scheduled to occupy 2015 and 2016l. Installation of ETCS will take from 2017 to 2025. After which electrification could be an option, but probably not because of the sunk cost of all those diesel power cars.

As my chum at the other end of the magazine has pointed out, the proper way to modernise a route is to combine electrification, new trains and resignalling in what us old farts call Total Route Modernisation. I know my techie chums (or ex-chums) in DfT Rail are feeling badly bruised by my treatment of their Rail Technical Strategy and White Paper, but what do they expect when they come up with this sort of thing?