With installation of ETCS Level 1 underway in several countries, feedback from pilot projects is helping to fine-tune the technical specifications for Level 2. Growing pressure to upgrade the railways of eastern Europe may trigger the high-volume orders needed to make ERTMS cost-effective

'ERTMS products are mature and available to roll out now and quickly at competitive prices', claimed Alcatel Transportation Automation Solutions President Jean-Pierre Forestier at InnoTrans in Berlin last September. But whilst the installation of Level 1 may be starting to gain ground, the prospects for Level 2 remain uncertain and Level 3 is still a distant possibility.

There are currently more than 20 signalling systems in use in Europe. A cost-benefit analysis published by the UIC's ETCS Migration Strategy Group in May 2004 predicted that, based on current commitments, the European Train Control System will have become the fifth most widely-used technology by 2010. With over 16 000 track-km scheduled to be equipped (Table I), the study group believes that ETCS will soon be established as a 'dominant system', with only PZB, Crocodile, KVB and AWS/TPWS retaining greater coverage, predominantly in Germany, France and the UK.

Note, however, that much of the early progress will only be with Level 1, which effectively forms an overlay to existing lineside signalling, adding automatic train protection or replacing current ATP systems to enhance interoperability.

Level 1 is already operational in Bulgaria and Hungary, and is being installed on the trunk network in Austria (RG 4.04 p217). Luxembourg Railways has also started network-wide installation of Level 1, and many other European railways are planning to migrate to ETCS as their existing ATP comes up to life-expiry.

With standardised equipment becoming widely available, suppliers are starting to offer Level 1 as a cost-effective ATP package elsewhere. For example, Bombardier is nearing the end of a four-year programme to install Level 1 across the entire main-line network in Taiwan (RG 12.04 p844) and also has a contract to fit 760 route-km in South Korea. Alcatel is tendering for Level 1 in India and China, and is in discussions about fitting ETCS in Russia. Forestier says that today 'Europe is the competence continent for the rail mode and drives the evolution of world standards.'

Cost advantages of Level 2

For many railways, however, there are no commercial benefits in moving to Level 1. The business case is better for Level 2, where the switch to cab signalling and GSM-R digital cellular radio should allow the elimination of much lineside signalling with its related cabling infrastructure and all the ongoing maintenance issues.

But critically, Level 2 is not yet ready to roll out. Swiss Federal Railways' original testbed between Olten and Luzern provided valuable data during its 18-month operating life, but it was eventually abandoned because the installation was designed around draft specifications that rapidly became non-standard. Most of the present raft of Level 2 pilot projects have adopted the V2.2.2 specification which SBB is using for its Mattstetten - Rothrist installation, which is now not expected to go live until the end of 2006.

However, it is now apparent that V2.2.2 of the system requirements specification is by no means the final solution. It is known to contain errors and 'open points', 41 of which have been identified as priority items for inclusion in the V3.0 update. An example is 'cold movement detection'. This means finding a way in which on-board equipment can be certain when a berthed train is 'awakened' that it has not been moved to another location while the ETCS has been shut down, invalidating stored information. As further functions are introduced to increase the scope of the standardised product, this will require a V4.0. Meanwhile, the ERTMS working group is close to finalising the specifications for V3.0, which is expected to be signed off by the end of 2005 if all goes well.

As it is completed, each version is incorporated into the ERTMS Technical Specification for Interoperability, known as the Command Control Signalling (CoCoSig) TSI. This is used for specification purposes - in the Cambrian pilot project (p33), for example - but with an option to upgrade to a higher version as this becomes available.

Europe's freight challenge

Whereas many of the Level 2 pilot projects are being conducted on high speed passenger lines, the driving force for widespread installation over the next few years is likely to be freight, where rail's share of the western European market has been falling. Whilst the designated road and rail corridors in the Trans-European Networks programme are approximately equal in route length, statistics show that the respective market shares are currently around 90% and 10%. As road transport has absorbed almost all the growth in business over the past 30 years, rail's market share has effectively declined by 66%.

Nevertheless, there is still strong political support for rail, which has great potential for relieving road congestion. The concept of 'modal rebalancing' has significant support at EU and national levels, and the European Commission's Directorate-General for Energy &Transport believes that rail is best placed to handle the growing exchange of goods between the 25 EU member states.

'ERTMS promises borderless competitive railways in Europe', predicts Dr Alfred Veider of Alcatel TSD Austria, suggesting that this offers 'a tremendous opportunity for the railway industry.' At present the average rail freight haul within the EU is around 300 km, and much of the traffic still does not cross national borders. Veider believes the average haul could shoot up to around 1 000 km 'if interoperability really works'.

Implementation of EU directives and other initiatives should prove a key step in enhancing rail's ability to deliver. However, political favour is a fickle thing, and the rail industry needs to start delivering results. 'If its market share falls below 10%, rail will rapidly lose credibility', he warns.

For a continent dominated by domestic passenger services, boosting rail's share of the freight market rapidly comes up against capacity issues. Veider says the predicted growth in rail freight between now and 2020 means moving from around 100 trains/day on the main corridors, at average headways of 14 min, to 250 trains/day running at 6 min intervals. 'How do we get more trains through the existing network?' he muses. 'Technology offers considerable scope to reduce life-cycle costs, and perhaps move to unattended operation, with a greater payback in productivity terms.'

The main options for raising capacity are to invest in additional tracks or new lines, or to upgrade existing corridors by enhancing the train control systems. Veider says 'it is better to invest in automation than concrete', noting that 'the payback value is four times as much per euro of investment.' And as another benefit, he adds that 'ERTMS is as much about ensuring reliable operation as it is about interoperability'.

The TEN-T programme envisages fitting ETCS to a network of 100 000 route-km by 2020. Veider says 'there is no time to waste', as the timetable for installation 'needs to start now, if deliveries are to be phased over the next 10 to 15 years'.

Installation rates at present are far below those needed to meet the 100 000 km target, with less than 5% of the planned network currently programmed to be equipped. In July 2004 DG-TREN set an intermediate target to achieve 20% of TEN-T coverage by 2008, but this has already slipped and Veider says the current thinking envisages a new target of 2010. The priority programme will start with selected 'one-stop shop' freight corridors such as the Belifret route between Belgium and the Mediterranean or the north-south corridor linking Germany's North Sea ports to Italy.

Migration strategy

The UIC study group review concluded that 'the broader economic benefits will not be maximised by an apathetic unco-ordinated approach'. The railways were exhorted to 'review the current situation constructively and prepare in a united fashion' to develop system specifications and a migration programme, and deal with outstanding issues such as interchangeability and cost of equipment.

The UIC study examined 10 sample corridors totalling 30 563 track-km (Table III). These represent around 20% of the TEN network, roughly equivalent to the length that the EU wants to see equipped by 2010. The working group estimated the cost of fitting ETCS to the 10 routes would be around €638m over and above any current investment in signalling and train control already planned. This works out at €466 000 per track-km.

Far more critically, a total of €8·8bn would be needed to equip all the trains using any part of the corridors concerned. Whereas the selected routes only account for 12% of the rail networks of the countries involved, they are used by 28% of the respective national rolling stock fleets. This would require the fitting of 13 650 traction units, of which only 11% would be involved in cross-border operations. Converting only these trains would cost about €1bn.

Assuming an average life expectancy of 40 years for signalling and train-control equipment, the UIC study team projected that around 35% of lineside equipment across the European network should be converted by 2020. Over the same timescale, between 60% and 110% of the rolling stock would have to be converted, depending on the degree of disruption to operating patterns that could be accepted.

Across the 25 EU states and their CER neighbour networks in Norway and Switzerland, the UIC study group estimates that implementation of ETCS would result in a negative net present value of €12·5bn over 40 years. This, it believes, would have to be met by an injection of funds from the EU or national governments to reflect 'the broader economic factors'.

The study recommends a number of steps towards an integrated strategy:

  • no further development of national train control systems;
  • focus on the busiest corridors and minimise the dual-fitting of routes;
  • systematically reduce the number of on-board systems;
  • create a mass market for ETCS components by replacement of life-expired or obsolete equipment;
  • set a realistic programme for converting traction fleets.

In addition, the group recommends that all new rolling stock should be ordered with ETCS or provision for retrofitting, and it urges each national infrastructure manager to draw up a rolling programme for migration in line with the forecast life-expiry of existing lineside signalling.

Veider agrees that the migration stage will be crucial, as railways determine the optimum mix of Level 1 and Level 2 and how to phase their investment. But he emphasises that the greatest return on investment will come when both the network and fleet have been equipped. Overall, he believes the long-term payback prospects are good. Rail freight revenues in Europe are predicted to rise from €40bn a year today to over €100bn in 2020 - provided that interoperability drives up rail's competitiveness in line with expectations.

Look east for progress

Veider believes that the EU accession states in Eastern Europe are well placed to make the running on ETCS over the next few years. He cites the rapid takeover of GSM mobile telephony as an example of how such countries can leap-frog an entire technical generation when given the opportunity to replace obsolete legacy systems. He points out that GSM coverage is already much higher in Eastern Europe than in many western states.

Many of eastern Europe's existing signalling systems are obsolete, and there is limited provision of automatic train protection. With little investment worth salvaging, Veider believes it will be much easier for these railways to scrap everything and start from a clean sheet. Equally, with no residual signalling to overlay with Level 1, the natural drive will be to move straight to Level 2.

The accession states' railways should also find it easier to raise funding, as the focus of the EU's Cohesion Fund is being re-oriented to help bring the new arrivals up to the standards already enjoyed by their western partners in all aspects of their economies.

'It is generally accepted that ETCS is the right long-term technical solution, but that the migration from existing national systems will be difficult and costly, and many networks are questioning whether the net commercial result will be positive. This study takes as its starting point the assumption that the force of the EU Directives precludes any consideration of non-compliance'.

UIC ETCS Migration Strategy Group report, May 2004

ETCS Level 1 equipment from Alstom has been installed on the new Athens suburban line between Larissa and the

ERTMS testbed. International trials with ERTMS are planned on the Paris - Frankfurt corridor from 2007. Trains will be fitted with on-board equipment for ETCS and the national signalling systems in France and Germany. Photo: Quintus Vosman

Table I. ETCS projects in Europe planned or committed for completion by 2008

Country Project description track-km %*
Austria Level 1 to be installed on the ÖBB Category A main line network. By end of 2008 around 60% will be in operation, including the Austrian part of the Wien - Budapest line used as a pilot. 1 500 33
Belgium Level 2 being installed on the new high speed lines from Liège to the German border (L3) and Antwerpen to the Dutch border (L4). Level 1 is to be installed on the remaining conventional network, 50% to be in operation by end 2008. 2 500 55
Bulgaria Level 1 operational on Sofia - Plovdiv - Burgas main line. By 2006 Level 1 to be fitted on 60 km line from Sofia to Serbian border. By 2008 Level 1 on 160 km Plovdiv - Svilengrad line and probably around 400 km on other BDZ main lines. 1 250 25
Czech Republic Proposal for Level 2 pilot application on the Czech part of the Berlin - Praha - Wien corridor is well advanced, and the intention is for the Czech Pendolino sets (which have been fitted with on-board equipment) to operate with Level 2 when the lineside equipment has been installed. 400 <5
France The TGV Est line under construction between Paris and Strasbourg will be used as a pilot project with dual-fitted trains and track, to test all possible combinations of mixing ETCS with legacy signalling systems. Installation of 'dual-standard' Level 2 equipment plus conventional TVM430 cab signalling is scheduled to start in 2005. 600 <5
Germany A section of DB's Berlin - Halle/Leipzig main line has been fitted with Level 2 equipment, and trials are now in progress towards EBA approval. Level 2 to be fitted on Mannheim - Saarbrücken as part of the TGV Est pilot project. Level 1 to be installed from Aachen to the Belgian border. 580 <5
Greece Under a turnkey signalling and electrification contract awarded by Ergose, Alstom has installed Level 1 on the Athens suburban network, including 600 Eurobalises, 270 LEUs and 12 sets of on-train equipment. All equipment was installed before the Olympics, but formal testing must be completed before it is used in commercial service. 60 <5
Hungary Level 1 being installed on Hungarian section of the Wien - Budapest - Bucuresti corridor (420 km), and will be fitted to the Hungarian parts of the Milano - Ljubljana - Budapest corridor (280 km) and on the Szajol - Nyiregyhàza - Tracz line (160 km). 860 11
Italy Level 2 under test on Roma - Napoli high speed line, where four trains are due to be fitted by June 2005 and commercial service is expected to start in December 2005. Level 2 will also be fitted on the Torino - Milano line expected to open in 2006, and on the Firenze - Bologna - Milano line under construction. 1 200 5·4
Luxembourg Level 1 currently being installed on the entire CFL network. 420 100
Netherlands HSL Zuid (Amsterdam - Belgian border) and Betuwe Route (Rotterdam- German border) to be equipped ahead of opening now expected in 2007. Level 2 being installed on the Amsterdam - Utrecht line where quadrupling is to be completed by December 2006. 640 10
Romania Level 1 to be installed on parts of the corridor from the Hungarian border to Bucuresti and Constanta. In 2002 Siemens was awarded a contract to supply 508 Eurobalises and 140 LEUs for the section of Corridor IV between Bucuresti and Campina; project leader IMSAT was responsible for installation, with commissioning due to start in 2004. 700 <5
Spain Level 1 equipment has been installed between Madrid and Lleida on the Madrid - Barcelona high speed line but is not yet in service. The equipment on the Zaragoza - Huesca branch was due to go live in December. Level 1 will also be fitted to Lleida - Barcelona and Madrid - Toledo. Level 2 is to be fitted to the entire route from Madrid to the French border, plus the Madrid - Segovia - Valladolid, Córdoba - Málaga, and Madrid - Valencia high speed lines now under construction. 3000 20
Sweden Level 2 to be installed on Botniabanan (190 km). 134 km Repbäcken - Malung line selected as pilot for ERTMS-Regional. 320 <5
Switzerland Level 2 being installed on Mattstetten - Rothrist new line, and will be fitted to Lötschberg base tunnel, due to open in 2007. Level 1 (with limited supervision) to be installed on the remaining network; 25% expected to be in operation by end 2008. 2 000 27
UK Tenders to be invited in May 2005 for Level 2 pilot project on the Cambrian lines. 220 <5
Total 16 250
* % of network expected to be equipped by 2008

Table II: Who's doing what in ETCS

Country Pilot projects Suppliers
France Tournan - Marles-en-Brie Alstom/Ansaldo
Germany Berlin - Halle/Leipzig Alcatel/Siemens
Italy Firenze - Arezzo Alstom/Ansaldo
Netherlands Meppel - Leeuwaarden, Heerlen - Maastricht Alstom/Bombardier
Spain Albacete - Villar de Chinchilla Bombardier/Invensys
Switzerland Luzern - Olten (dismantled) Bombardier
UK WCML TCS (now aborted) Alstom
Commercial projects
Austria Wien - Nickelsdorf Alcatel/Siemens
Belgium High speed lines 3 and 4 (Level 1/2) Alstom/Siemens
Bulgaria Sofia - Burgas (Level 1) Alcatel
Czech Republic CD Class 680 Pendolinos (Level 1/2 on-board) Alstom
Greece Athens suburban (Level 1 on-board and wayside) Alstom
Hungary Kimle - Budapest and Zalalövö - Hodos (Level 1) Alcatel
Italy Roma - Napoli (Level 2) Alstom/Ansaldo
Italy Trainborne equipment Alstom
Luxembourg CFL network (Level 1) Alstom/Alcatel
Netherlands HSL-Zuid (L1/2 wayside) Alcatel/Siemens
Netherlands Amsterdam - Utrecht Bombardier
Netherlands Betuwe Route Alstom
Romania Bucuresti - Campina (Level 1) Siemens
Spain Madrid - Lleida (Level 1/2) Ansaldo
Spain Zaragoza - Huesca Alstom
Spain Leida - Barcelona, La Sagra - Toledo, Segovia - Valladolid (Level 1/2 wayside) Siemens/Alcatel/Invensys
Spain Córdoba - Málaga (Level 1/2 wayside) Alcatel/Invensys
Spain Euro-AVE and Lanzaderas trains Alstom
Sweden/Norway/Finland STM development Ansaldo
Switzerland Mattstetten - Rothrist (Level 1/2) Alcatel (L1)/Alstom (L2)
Switzerland Lötschberg base tunnel (Level 2) Alcatel

Table III. Corridors analysed by the UIC migration strategy group

Corridor Track-km
Madrid - Paris - Zürich - Milano 4 966
Paris - Mannheim - Zürich 2 136
Paris - Brussels - Köln/Amsterdam/London 2 124
Genève - München - Milano - Napoli 4 284
Rotterdam - Milano - Genova 3 887
Antwerpen - Bettembourg - Metz - Basel 1 326
Hamburg - København - Malmö 1 040
Berlin - Budapest - Bucuresti 4 391
Milano - Trieste - Ljubljana - Budapest 1 693
München - Budapest - Beograd - Sofia - Thessaloniki 4 716
Total 30 563


ERTMS development reaches the critical point

With several ETCS Level 2 pilot projects starting to deliver results, fine-tuning of the [V2.2] technical specification has begun with the aim of firming up V3.0 by the end of this year. [V4.0 is likely to follow.] Meanwhile, as several infrastructure managers install Level 1 equipment, the growing pressure to upgrade ageing rail infrastructure in the accession countries of eastern Europe offers significant potential for the high volume orders needed to make ERTMS installation cost-effective.

Le développement de l'ERTMS atteint le point critique

Avec plusieurs projets pilotes pour l'ETCS de niveau 2 qui commencent à livrer des résultats, la mise au point des spécifications techniques pour la version 2.2.2 a commencé, l'objectif étant de confirmer la version 3.0 pour la fin de 2005. Pendant ce temps, tandis que plusieurs gestionnaires d'infrastructure installent l'équipement de niveau 1, la pression croissante pour améliorer une infrastructure ferroviaire vieillissante dans les pays arrivant de l'Europe de l'Est, offre un potentiel significatif pour des commandes de gros volumes n‚cessaires afin que l'installation de l'ERTMS vaille la peine sur le plan du coût.

ERTMS-Entwicklung erreicht kritischen Punkt

Da nun die verschiedenen ETCS Level 2-Pilotprojekte langsam Resultate zeigen, hat die Feinabstimmung der Version 2.2.2 der Technischen Spezifikationen begonnen, mit dem Ziel die Version 3.0 bis Ende 2005 zu finalisieren. In der Zwischenzeit, da verschiedene Infrastruktur-Betreiber beginnen, Level 1-Ausrüstungen zu installieren, führt der wachsende Druck zur Erneuerung von alter Bahn-Infrastruktur in den Umländern von Osteuropa zu einem signifikanten Potential für Grossaufträge, welche zur finanziellen Tragbarkeit von ERTMS-Installationen notwendig sind.

El desarrollo del ERTMS alcanza un punto crítico

Con varios proyectos piloto del ETCS nivel 2 comenzando a arrojar sus primeros resultados, ha empezado la puesta a punto de la especificación técnica de la versión 2.2.2 con el objetivo de dejar lista la versi¢n 3.0 antes de finales de 2005. Entretanto, mientras algunos gestores de infraestructuras instalan equipos de nivel 1, la creciente presi¢n para la modernización de las infraestructuras envejecidas en los países del Este de Europa candidatos a la Unión Europea ofrece un significativo potencial para los pedidos de gran volumen que se necesitan para que la instalación del ERTMS resulte rentable.