The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER SIX: SEPARATING INFRASTRUCTURE AND OPERATIONS

Introduction

6.1. The requirement for the separation of infrastructure from operations is fundamental to the European Commission's policy for the restructuring of Europe's railways. This separation has worked with other transport infrastructure including airports and ports and, most obviously, the separate operation, maintenance, renewal and enhancement of roads and the vehicles which use them. Multiple private vehicle operators enjoying open access to a single public infrastructure is the dominant road industry model worldwide. For rail, the rationale for separation appears to have been at least partly based on the assumption that:

• Infrastructure provided by the public sector was under-used, and spare capacity should be made available to new operators at marginal cost
• Services, also typically provided by the public sector, were often inefficient and unattractive and could be subject to competition, either "in the market" if commercially viable or "for the market" otherwise
• Standards and processes should be harmonised to facilitate cross-border services to serve the single market

6.2. In practice, these assumptions are not always justified, for a number of reasons which we discuss below.

6.3. Separation of infrastructure and operations is complete in Scotland and all the European comparators except Ireland and Northern Ireland which, as described above, enjoy a derogation from many of the requirements of the First Railway Package until March 2008. In Hamburg/Schleswig-Holstein, DB continues to own infrastructure and provide services from within the same group. Although technically compliant with European law, this had led to continued complaints by other operators that it favours DB as an operator.

6.4. New Zealand, in contrast, is not bound by European law and its railway industry was until 2004 a vertically-integrated business outside the major cities. However, the infrastructure has now been returned to public ownership to ensure that the extent and integrity of the network is preserved and to provide for future increases in service levels and/or the number of operators if required.

Regulation of access and charges

6.5. The European Commission has based its policy on access charging on the economic principle that operators should pay the marginal cost they impose on the network, recognising that this may include not only direct wear and tear but also other factors including:

• Social and environmental externalities, such as savings in accidents, noise and pollution
• Congestion, whereby operators pay for the delays they impose on other operators
• Scarcity of capacity, when access charges may rise to "opportunity cost" to encourage allocation to the operator that values it most

6.6. In principle, it might be expected that access charges of this type could provide most of the signals required to provide an efficient allocation of capacity even on networks which are crowded or where there are major social or environmental externalities.

6.7. In practice, European law permits a wide range of exceptions to the principle of marginal cost access charges, and in both Scotland and Hamburg/Schleswig-Holstein the policy is broadly that all infrastructure costs should be recovered from operators. In addition, as discussed below, most services are not provided on a commercial basis and their specification is not directly influenced by access charges.

6.8. Ireland and Northern Ireland have still not set any access charges. Northern Ireland Railways proposes to introduce separate accounting of infrastructure and operations and to devise access charges, but does not expect any new operators to appear, and any "charges" applied would therefore merely be an internal accounting device. Ireland has had no applications for passenger open access but open access for freight services will be available from January 2007.

6.9. In New Zealand, access agreements incorporate charges negotiated between ONTRACK and operators from a starting point based on the full recovery of the OMR costs of an efficient infrastructure manager and a reasonable rate of return. A more variable charging regime, which would also include charges specific to each line and additional charges on congested track, is currently being negotiated.

6.10. Sweden's Næringsdepartementet and Denmark's Trafikstyrelsen have set access charges with similar structures, with each vehicle paying a fixed annual charge and a per-kilometre charge which may vary with location and time of day.

6.11. In Sweden, charges are based on detailed calculations intended to estimate marginal cost, but an interviewee commented that the simple structure of charges actually used cannot fully reflect all the variations in social marginal cost. In practice, the only applications by new entrants to operate open access passenger services have been on inter-regional routes and have been turned down under the terms of SJ's monopoly.

6.12. In Denmark, Trafikstyrelsen applies peak charges on the main Copenhagen to Århus line, but Banedanmark acknowledged that these would only affect the behaviour of freight operators and may not be material to longer-distance freight transit flows. Banedanmark confirmed that, while application procedures are set out in their "Network Statement" ( See Annex 4), there have been no applications by new entrants to operate open access passenger services.

6.13. In Scotland, ORR will set access charges for awarding additional access rights at marginal costs, provided that the services are not primarily abstractive of other operators' revenue. Existing franchisees and, in principle, new operators, are therefore free to add services if it is commercially viable to do so. As far as we are aware, however, there have been no applications by new entrants to operate open access passenger services in Scotland.

6.14. In Hamburg/Schleswig-Holstein, charges for access to the DB Netz infrastructure are set by the newly-established Bundesnetzagentur (Federal network agency) but the authorities of the Länder are responsible for charges and access on regional railways.

6.15. However, DB Netz's charges, set on the basis of full recovery of its OMR costs, amount to almost €4 billion (£2.6 billion) a year, although around two-thirds of these access charges are paid by the Länder for their PSO services and only one-third by DB and other operators of commercial services. These high access charges are probably a material barrier to entry to at least some services which would otherwise by commercially viable.

6.16. Even if there are no new applications for access, clear rules are needed for appeals and dispute resolution if operators are dissatisfied with capacity allocation decisions. In Scotland, Network Rail's decisions are subject to approval by the independent ORR. In Sweden, operators may now appeal against Banverket's decisions to Järnvägstyrelsen although the process has not yet been tested. In Denmark, there is also a provision for operators to appeal to the Jernbaneklagenævnet (Railway Appeals Board) appointed by the Ministry of Transport, although we understand that it has not yet been convened to deal with an appeal. New Zealand's National Rail Access Agreement has provision for new entry, even though it is seen as unlikely, but not for independent regulatory oversight.

6.17. Access disputes have not yet arisen within Schleswig-Holstein but interviewees had concerns about the effectiveness of the mechanisms for regulating access to DB Netz's infrastructure. Connex's Nord-Ostsee-Bahn ( NOB) applied for access charges to run services to Neuss in Nordrhein-Westfalen and to Berlin from summer 2006. DB Netz refused to assign the necessary paths, basing its decision on DB's grandfather rights. The EBA decided in favour of NOB but DB Netz referred the case to the administrative court in Cologne. In June 2005 the court ruled in favour of DB Netz on the basis that the new national regulations for resolving conflicts did not come into effect until August 2005, despite the fact that to comply with European law they should already have been in force. EBA has now referred the case to the appellate court.

Incentives to operators

6.18. Access charges set on the basis of economic principles can, in principle, incentivise behaviour by operators, encouraging services which can pay their marginal costs and deterring or pricing off those that cannot. In Sweden, marginal access charges are intended to maximise SJ's incentives to identify and provide commercially-viable routes, and in New Zealand the access charging regime being developed by ONTRACK is intended to encourage the monopoly incumbent, Toll Rail, to provide and maintain the current coverage of its services.

6.19. In Germany, access charges are normally above marginal cost. As a result, the Länder pay higher subsidy to PSO services than would otherwise be the case, while the Bund pays DB Netz less support for its fixed costs. These charges also act as a barrier to entry to operators who would be able to afford to pay their marginal costs.

6.20. However, most passenger operators in Scotland and the comparators are providing services specified by a contract or franchise and will not be incentivised by access charges, which they will merely pass through to their transport authority customer. On the many routes where the incumbent's fares are supported, new entry will be unprofitable whatever level of access charges apply. While high access charges may deter or prevent commercial entry, low access charges at best allow commercial provision of bulk freight services, some long-distance or point-to-point passenger services, or some additional regional or urban/suburban services which can be provided at low marginal cost.

6.21. Access charges can also, in principle, incentivise operators to use rolling stock which imposes the least costs on the infrastructure, but these incentives will normally only be felt at the infrequent occasions on which stock is replaced, and may be small or immaterial compared with other elements of the life cycle costs of the rolling stock.

Incentives to infrastructure managers

6.22. A well-designed structure of access charges can, in principle, also provide incentives to infrastructure managers:

• To enhance the capability, or expand the capacity, of the infrastructure where operators are willing to pay for it through access charges
• To improve the operation and management of the infrastructure to improve the punctuality and reliability of operators' services
• To "reserve" infrastructure capacity for operators who are prepared to pay in advance for its use

6.23. In practice, the extent to which these incentives exist or influence behaviour is often limited, as discussed below.

Paying for capability and capacity

6.24. In Scotland, the underlying principle is that operators contract for access rights from Network Rail and can buy additional rights, such as for more capacity or higher speed, at prices and on terms determined by the Office of Rail Regulation ( ORR).

6.25. Sweden's operators also pay access charges to Banverket and can negotiate with it for additional access rights under Järnvägstyrelsen's supervision, although since access charges are limited to marginal costs the relevant PTA would normally have to pay the capital cost of any enhancements.

6.26. In Denmark, even this limited linkage between operator and infrastructure manager is missing, and access charges are paid direct to the Ministry of Finance. Operators or local authorities requiring additional access rights must negotiate with Trafikstyrelsen to have them added to Banverket's plans.

6.27. In practice, however, the marginal cost of using existing infrastructure is nearly always much lower than the incremental cost of enhancing it or expanding its capacity. Access charges based on marginal cost can sometimes provide sufficient incentive for infrastructure managers to make minor changes, at low cost, which enable service improvement. However, access charges which cover the full cost of incremental provision can rarely be covered by operators' revenues, particularly where fares are regulated, and normally require public sector funds, where they can be justified by social or environmental benefits. We discuss the processes of enhancing and expanding infrastructure further below.

Incentivising performance

6.28. Access charges can also provide incentives to the infrastructure manager if they also include a performance regime under which operators are compensated for delays to their trains. In most cases a benchmark performance is set with access charge premia ("bonus") or rebates ("malus") paid for performance above or below the benchmark.

6.29. Scotland's operators have a benchmarked performance regime with Network Rail, and Denmark's operators can choose between a bonus/malus regime and a simpler guaranteed minimum punctuality. Sweden's Banverket does not normally offer a performance regime, but the Association of Swedish Train Operators (Branschföreningen Tågoperatörerna) is lobbying for a better service and SJ has arranged a one-year test of a regime of additional payments for good performance. The access charging regime being developed by New Zealand's ONTRACK also has provision for a performance regime although details have not yet been agreed.

Forward booking capacity

6.30. Transport Scotland drew our attention to ORR's recent decisions to award access rights within England on the main East Coast and West Coast main lines to Scotland, and to the lack of mechanisms to "reserve" future capacity for, for example, future Anglo-Scottish services. In practice, ORR has already applied a "use it or lose it" approach to access rights which have been awarded but not used. This was applied to paths from Edinburgh and Glasgow to the Channel Tunnel, reserved in 1987 for the provision of "regional" Eurostar connections between Scotland and the continent, and retained in the relevant timetables until proposals to operate the services were effectively abandoned in 2000.

6.31. Arguably the RUS process can anticipate reasonable future needs, but ORR is not obliged to act on these in reaching access decisions. In additional, Transport Scotland is only one of many stakeholders to be considered in the English RUSs.

6.32. The presumption in European law is that operators should not be able to reserve capacity in advance for long periods, primarily to prevent governments from awarding long term access rights to state-owned incumbents to prevent or delay competitive entry. Interviewees in Hamburg, Schleswig-Holstein, Denmark and Sweden all told us there is no provision for capacity to be reserved in advance. This contrasts with ORR's willingness, consistent with European law, to award access rights for up to 15, or in some cases, 20 years ahead where required to justify investment.

Network change

6.33. Any modification to the network to add or remove capability can constitute "network change". One issue is the ease with which the infrastructure manager can make changes to the network, whether requested by operators or public transport authorities or on their own initiative. In both Scotland and Denmark the process is seen as cumbersome.

6.34. In Denmark, the access contract specifies the processes for network change. Banedanmark told us that even to remove unused track it had to consult operators and the local authority and then submit plans to Trafikstyrelsen. To modify a level crossing it would need to consult with Trafikstyrelsen, the local and/or highway authority, and if signalling changes were required, to consult operators whose rule books and safety procedures might be affected. Closure of rail lines requires a Parliamentary decision. DSB noted that the process of network change became progressively more complex as more operators were involved.

6.35. In Hamburg/Schleswig-Holstein, DB Netz can only close a line with the permission of the Eisenbahnbundesamt, but this must be granted if there is no interest in leasing or taking over the line. In New Zealand, arrangements for network change are still evolving, but ONTRACK has responsibility for network change. Operators may propose changes to ONTRACK, but must be prepared to meet, through access charges, the costs of infrastructure improvements. Line closures require ministerial approval.

6.36. While it would probably be desirable for minor network change to be relatively simple, interviewees pointed out the difficulty in distinguishing in legislation at what point this had the effect of total or irreversible closure which should be subjected to ministerial or parliamentary scrutiny.

Planning expansion and investment

6.37. Interviewees generally agreed that separate ownership and management of infrastructure and operations raised, or would raise, no particular problems on lightly-used regional lines operating well within their capacity. In this situation:

• The infrastructure manager carries out maintenance and renewal of the infrastructure at quiet times, such as at night, with no interference with services.
• The service operator is free to plan and operate services without interference, either from the infrastructure manager or from other operators.
• Real time control of service is by the infrastructure manager (although in Great Britain and Sweden it has been found effective for operators and infrastructure manager to operate a joint control room, with final decisions resting with the infrastructure manager).

6.38. These conditions arise in at least some parts of all the comparator networks, and most obviously in rural Sweden, where infrastructure and operations were split without undue difficulty with the establishment of Banverket in 1988. Interviewees generally agreed that separation did not pose any particular extra costs or difficulties in these circumstances. One pointed out that some sort of split was inevitable where services and networks overlapped, and argued that the principal difficulty in such circumstances was coordinating the timetables of different operators to provide connections.

6.39. However, access and charging regimes do not readily deal with situations where investment in both infrastructure and operations is required, for example to install electrification or to change signalling equipment, or where the infrastructure is congested and capacity expansion is required. They are generally least able to influence behaviour effectively in situations with multiple operators, where the solution to congestion and capacity problems may be that:

• Operators adapt their services to the requirements of other operators
• Operators adapt or change their trains to facilitate the requirements of other operators
• The infrastructure manager adapts, enhances or expands the infrastructure to meet the collective requirements of the operators

6.40. Without knowing in advance which solution, or combination of solutions, will be the most cost-effective, it is difficult to devise access and charging regimes which can incentivise or produce the right outcome. This becomes even more difficult if the changes required are unlikely to be self-financing and will require public support on the basis of a social business case. In these circumstances the most appropriate approach to resolving problems is likely to be by joint analysis in which all parties understand the costs and benefits of different approaches.

6.41. In Scotland, British Rail planned investment in operations and infrastructure as an integrated business, and increasingly found that the optimum approach was to plan and renew rolling stock, track, signalling and electrification as a single project when the need arose. When Railtrack was created in 1994, it was envisaged that enhancement investment, if needed at all, would normally be arranged between a single "customer" operator wishing to upgrade its services and a "supplier" Railtrack providing infrastructure, and hence the requisite access rights, if not by negotiation then on a basis imposed by ORR.

6.42. The limitations of this conceptual approach became apparent when renewal or enhancement was needed on infrastructure used by several operators who, even if not direct competitors, had different views of the market, commercial objectives, planning horizons and ability and willingness to pay. It could be inefficient to invest in infrastructure only required for a short period where minor changes in service patterns could release additional capacity. To deal with this, the Strategic Rail Authority developed a process of:

• Regional Planning Assessments ( RPAs), to understand and model future demand
• Route Utilisation Strategies ( RUSs), to set out the design for the infrastructure and the patterns of passenger services together to achieve a commercial or social optimum

6.43. To work correctly, these processes needed the coordinated application of a wide range of skills including:

• Land use planning
• Passenger demand forecasting
• Train operations and timetabling
• Design of trains: length, capacity, speed, acceleration and power requirements
• Design of infrastructure: track, points, signalling, power supply and stations

6.44. A major problem is that, while the first four of these services could be provided by suitably-qualified advisors or consultants, only the infrastructure manager, Network Rail, was close enough to the infrastructure to examine the trade-offs between infrastructure costs and capability. In 2005 responsibility for preparing a Scotland Planning Assessment ( SPA), the RPA for Scotland, was transferred to the Scottish Executive. Responsibility for preparing RUSs was transferred to Network Rail. Network Rail began work on the Scotland RUS in July 2005 and published a Draft for Consultation in August 2006.

6.45. Figure 6.1 compares the former and current arrangements in Scotland with those in Sweden, Denmark, the comparators which have the greatest separation in roles, New Zealand and Ireland, which still retains a highly-integrated structure.

FIGURE 6.1 LEAD ROLES FOR PROCUREMENT, PLANNING AND OPERATIONS

6.46. In Denmark, the role of the infrastructure manager is limited. Banedanmark can put proposals for new infrastructure to Trafikstyrelsen, but the latter takes the lead in service and infrastructure development and makes recommendations to the Ministry of Transport on what plans should proceed, including not only enhancements to the infrastructure but also like-for-like renewals. However, Parliament sets overall investment levels and priorities, which are currently to address a backlog of track and signalling renewals that in one case led to a sudden imposition of low speed limits. There is also a need to increase capacity between Copenhagen and Århus across Sjælland (Zealand) and the Storebælt.

6.47. Trafikstyrelsen remains in control until the end of the planning phase before handing to Banedanmark a specification for renewal and enhancement work and an agreed level of funding. Banedanmark must then construct the new infrastructure in and around its operation, maintenance and renewal of the "live" railway. It must also report regularly on expenditure and progress and inform Parliament of any major overruns.

6.48. In the Danish system, Trafikstyrelsen is both buyer and price-setter, and there is a tension between its desire to provide low cost estimates, if only to secure Parliamentary funding, and Banedanmark's ability to deliver them in a changing environment. There are currently major discussions about the reasonable costs of the enhancements to increase capacity across Sjælland.

6.49. Scotland's original arrangements differed from those in Denmark in two main ways:

• Infrastructure renewal was part of Network Rail's ongoing OMR responsibilities
• The roles of buyer and price-setter were separate: the Strategic Rail Authority specified the infrastructure capability, and hence access rights, it wished to buy, but the reasonable cost of work, whether OMR or enhancement, was determined independently by the Office of Rail Regulation

6.50. Network Rail has now taken over the lead role in preparing the RUSs, but the procurement decision remains with Transport Scotland and the reasonable costs are still determined by the Office of Rail Regulation.

6.51. In Sweden, in contrast, Rikstrafiken's only role is the procurement of inter-regional services which SJ does not wish to operate on a commercial basis. Banverket reports direct to Næringsdepartementet and leads all other aspects of infrastructure planning. This is a simpler system, but some interviewees commented that Banverket is not seen as being as responsive as other stakeholders would like.

6.52. In New Zealand, most responsibilities have been transferred to ONTRACK, which now determines access charges as well as operating and maintaining the network, planning renewals and planning for the requirements of freight services. However, the regional authorities remain responsible for procurement of passenger services which they wish to support and planning for any additional infrastructure they require.

6.53. The Treasury pays the funding for all enhancements direct to ONTRACK, and interviewees noted that this has led to debate over how the regions can plan services and infrastructure on an integrated basis or influence ONTRACK's day-to-day performance.

6.54. Ireland's system is the most integrated, with IÉ carrying out train operations and infrastructure management as well as planning and project development. There is no formal process either for procurement of services or for the regulation of IÉ's infrastructure costs: in effect IÉ develops proposals on which DoT must then make a decision against its evaluation criteria. These arrangements are similar to those that existed in Scotland until 1994, when British Rail responded to national needs primarily through a dialogue with its government shareholder.

6.55. One weakness of this system, as noted above, is in urban areas, where it does readily facilitate integration between rail and other modes, or selection of the most appropriate projects when more than one mode can contribute to congestion or capacity problems. Another is the need for a sufficiently-informed DoT to be able to challenge and probe IÉ's analysis and costs.

56.6. Hamburg and Schleswig-Holstein are subject to the provisions of the new Bundesschienenwegeausbaugesetz, or federal rail infrastructure construction law, which gives the infrastructure manager responsibility for infrastructure OMR, enhancement and expansion, and for developing capacity analyses and capacity enhancement plans as required by European law ( see Annex 4). However, interviewees pointed out that these arrangements are as yet relatively untested and may in any case be changed further with the proposed part-privatisation of DB.

Planning electrification and interoperability

6.57. Two examples of areas in which investment in infrastructure and operations need to be coordinated are:

• Electrification of rail services
• Interoperability and ERTMS

Electrification

6.58. When rail infrastructure is electrified it is possible for non-electrified trains to continue to operate on it, in some cases for long periods. Scotland's infrastructure currently allows electric traction to be used on some, but not all, of GNER and Virgin Trains' cross-border services, the suburban services in Glasgow, whose trains must be electrified to pass through low-level stations in the city centre, and some local services between Edinburgh and North Berwick. Services beyond Glasgow and Edinburgh to Aberdeen and Inverness must be diesel-powered, and GNER trains from these cities to Edinburgh continue under diesel power over almost 600 km of electrified track to London. Where possible, however, it is desirable to coordinate electrification of infrastructure and rolling stock to optimise the overall investment.

6.59. Scotland, at least in comparison to England, is rich in renewable energy sources including hydroelectric schemes and, more recently, wind farms. This begs the question, raised at a recent conference on Scotland's Transport Infrastructure, of whether Scotland's railways could make greater use of renewable energy.

6.60. New Zealand's electricity network is isolated and hence generation, consumption and prices are determined wholly by local factors, but this is not true of any of the other comparators. Scotland's sources of low-cost renewable energy are linked by "interconnectors" to Northern Ireland and to the much larger English market. The privatised Scottish generators currently sell much of their output into this wider market at prices which reflect the balance of demand and costs of the market as a whole. Differences in energy prices between Scotland and England exist primarily because of transmission costs and constraints. Furthermore, under the current arrangements, Network Rail cannot identify which generators provide the energy consumed by individual operators.

6.61. The Scottish Executive could, in principle, ask Network Rail to procure, from renewable sources, electricity equivalent to ScotRail's estimated consumption. Network Rail may already do so, but even if it did not, the net result might merely be that Network Rail bought more renewable energy and other customers bought less. In either case, there might be no material effect on the volume of renewable electricity generated or the price paid by ScotRail for energy. This policy could equally be applied to the Scottish Executive's total energy requirement and need not be confined to energy for rail services.

6.62. Electrification increases the OMR costs of the infrastructure and is most likely to be justified, and installed:

• Where trains often require maximum power to climb steep gradients or to accelerate away from stops
• Where trains are frequent, particularly if the electrification system can transfer power from decelerating trains to accelerating trains by "regenerative" braking
• Where electricity is cheap, such as because of abundant hydroelectric power

6.63. This means that electrification is most likely to be cost-effective on densely-used networks around urban areas, or in mountainous areas, or where electricity is cheap.

6.64. We examined the extent of electrification in the comparator networks, examining both:

• The proportion of the infrastructure which had been electrified, measured in route-kilometres.
• The proportion of the rolling stock fleet which used electric power, which we estimated from the number of multiple units of each type in the operators' fleets. In the absence of details of consumption of electricity and fuel, this serves as a proxy for the proportion of the networks' "workload" which makes use of electrification.

6.65. Figure 6.2 compares these two measures of electrification. This shows that Sweden has the greatest proportion of electrification, around 70% of the infrastructure and 90% of the fleet, whereas Northern Ireland has no electrification. In both Scotland and Denmark, around 30% of the infrastructure and 60% of the fleet are electrified.

FIGURE 6.2 ELECTRIFICATION OF INFRASTRUCTURE AND FLEETS

6.66. We also asked whether programmes of electrification were under way and, if so, what steps were taken to optimise and coordinate investment in infrastructure and rolling stock. In the event, none of the comparators have either a formal electrification programme or a standard approach to managing one.

6.67. Northern Ireland has no electrification and Ireland has none outside urban/suburban areas. However, with integrated railway businesses, any future electrification projects could be planned internally, with the overall programme of investment in infrastructure and rolling stock coordinated.

6.68. In New Zealand, electrification installed on sections of the South Island network in the 1920s was removed in the 1970s and 1990s following a decline in traffic. In 1988 the section of the Main North Island Trunk Line between Palmerston North and Hamilton was electrified on the 25kV AC system, although we were informed that changing locomotives to take advantage of the electrified section is not cost-effective and it may in future be removed. However, the Wellington urban network is electrified on the 1500V DC system and a small extension is currently planned. Electrification of the Auckland network is also under consideration, although as yet no conclusive business case has been made.

6.69. In Denmark, the Ministry of Transport has asked Trafikstyrelsen to focus on renewal of track and signals and it has no specific programme relating to electrification. The Ministry has set up a commission to examine transport infrastructure needs for the coming 30 years, which will consider environmental issues and may propose further electrification. If programmes were to be developed, Trafikstyrelsen would need to coordinate its planning with the operator or operators affected. With increasing franchising within Denmark, it would find it increasingly necessary to deal not only with the current franchisee but also the local authorities who would need to be involved in the specification, funding and possibly procurement of future electrified trains.

6.70. Sweden has the highest proportion of electrification among the comparators. The infrastructure was largely electrified after the Second World War, although the programme was never extended to the Inlandsbanan or some of Banverket's minor rural lines. There is no longer any formal electrification programme, although individual projects take place where a business case arises. However, where the majority of existing rolling stock is in the ownership of SJ, we would expect that suitable electrified rolling stock would often be available through "cascades" from elsewhere in the system.

6.71. One example drawn to our attention is the electrification of 120 kilometres of the Blekinge coastal line between Karlskrona and Kristianstad. When the county took control of local services it cut fares, increased frequencies and introduced new stock, and passenger numbers rose around 20%. The line will reopen in 2007 and the new electrified services will run through the neighbouring county of Skåne to Malmö.

6.72. In summary, there is little scope for further electrification of existing lines in Sweden, few proposals for it in Denmark and New Zealand and none in Northern Ireland. New electrification is most likely to take place on new urban lines and in particular where new stations are in tunnel. Interviewees in New Zealand commented that the debate was now shifting from electrification to alternative fuel sources.

Interoperability

6.73. We noted above that the rationale for separation of infrastructure and operations appears to have been at least partly the assumption that standards and processes should be harmonised to facilitate cross-border services to serve the single market. However, European movement towards greater interoperability between the railways also creates a need to coordinate investment in infrastructure and rolling stock. One example is the installation of new signalling systems and in particular the proposed introduction of the European Rail Traffic Management System ( ERTMS). ERTMS requires the installation of equipment on both infrastructure and trains, and its introduction on a single railway line requires coordinated investment by all the operators whose trains use that line at any point.

6.74. In Denmark, the Ministry of Transport has asked Banedanmark to analyse and report to Trafikstyrelsen on how the system could be introduced by 2018. Banedanmark has appointed consultants to examine the issues, but a key issue is likely to be the need for funding.

6.75. In Sweden, we were told that a key difficulty is arranging funding for operators to equip their rolling stock with ERTMS. Freight operations are no longer likely to be viable if they have to bear the cost of equipping locomotives with ERTMS equipment. In practice this is recognised as a problem throughout Europe, where the costs of installing ERTMS equipment are often large compared with the limited profits obtainable in competitive markets.

6.76. Ireland and Northern Ireland are discussing moves to interoperable standards on the Belfast-Dublin route, potentially beginning with a joint investment in the standard GSM-R mobile communications system and, in the longer term, installation of ERTMS Level 2 signalling equipment. With both businesses retaining integration of infrastructure and operations, the principal need for co-ordination is not between separate infrastructure and operations organisations but between the separate railways.

6.77. Once achieved, interoperability between European railways is intended to permit greater standardisation and economies of scale in manufacturing and fewer operational and compatibility problems at borders, although differences of track and loading gauge will remain unless the railways are completely rebuilt. The principal difficulty faced by all the European comparators is the high costs of converting all their domestic operations to an international standard, especially as there may be no corresponding benefit until their neighbours have also done so.

Planning for integration

6.78. Electrification and interoperability require coordinated investment in infrastructure and operations but, as we have shown above, the comparators provide relatively few current examples of how this is done in practice.

6.79. The European Commission understands that investment may be needed to increase rail capacity, and the First Railway Package makes provision for infrastructure managers to identify capacity constraints and to produce plans to remove them. However, infrastructure managers cannot override legitimate national planning and funding constraints and cannot, on their own, optimise investment in infrastructure and operations or in rail and other modes, which may also often need to meet the requirements of more than one tier of government.

6.80. Our interviewees identified a number of projects which required cooperation and co-ordination of planning and investment across a number of bodies to deal with urban/suburban, interurban and cross-border transport. These were:

• Capacity in the Dublin area
• Capacity in the Stockholm area
• The Øresund link

Capacity in the Dublin area

6.81. We noted above that IÉ, as an integrated business, is able to optimise and coordinate its own planning in investment and infrastructure. In practice, the capacity of the network outside the capital is likely to be adequate for at least the next 10-15 years, and the principal requirements for new investment are in the Dublin area. Strategic transport planning in the Greater Dublin Area ( GDA) is the responsibility of the Dublin Transportation Office ( DTO) and the current transport strategy for the GDA is set out in "A Platform for Change" published by the DTO in 2001. The Railway Procurement Agency ( RPA), established in 2001, is responsible for the planning and provision of individual light rail and metro projects, and IÉ is responsible for the provision of suburban rail projects in the GDA, contained in the DTO strategy.

6.82. While some limited 4-tracking has been approved for the route from Dublin to the southwest between Heuston and Kildare, 50 kilometres from the capital, a major constraint is the 2-track north-south line linking Connolly station with routes stretching to Rosslare in the south and Belfast in the north. This line is becoming increasingly congested with three distinct types of passenger services:

• Cross-border "Enterprise" services from the north operated jointly by IÉ and Translink
• Urban/suburban " DART" and other commuter services, operated by IÉ, from the north and northwest to the south
• Interurban IÉ services operating to a range of destinations outside the Dublin area

6.83. The Enterprise service is becoming heavily used, particularly as the Irish economy has expanded, and it is seen by Northern Ireland as an important link to Dublin. There is a policy intention to increase the service frequency to hourly and it would also be possible to carry out track improvements along the line to improve journey times. However the priority within Dublin is to give greater priority to expansion of the local DART and commuter services.

6.84. Some relief could be provided by a scheme for a branch off the line from the north to a new station at Spencer Dock which would provide capacity for additional trains to terminate without passing through the most congested sections of track at Connolly station. A more major proposal, however, is for a new tunnel which would take DART services from the north through central Dublin and then out to Heuston station and the southwest. This would not only link the currently separate DART/commuter networks north of Connolly with Heuston services, and make much more of the city centre accessible by rail, but also effectively double the capacity into Dublin from the north and northwest, removing capacity constraints for the foreseeable future. However, planning and funding such a major investment will require substantial co-ordination between IÉ, the DTO, the RPA and the DoT (the proposed DTA is expected to take on this function after establishment).

6.85. A second problem was the provision of a rail link to the airport. IÉ developed a proposal for a branch from the existing line north of the city which could, in principle, have offered a point-to-point city service between Connolly station and the airport. The DTO proposed a metro solution in their 2001 transport strategy, which was developed by the RPA and eventually accepted by government. The two rival projects were conceived, developed and evaluated separately and on different bases by the different bodies. The current structure does not automatically provide for a comparison of the proposals on a consistent basis, or for any wider optimisation of proposals for IÉ's network and the metro and LUAS light rail systems.

Capacity in the Stockholm area

6.86. Stockholm faces a similar constraint to Dublin, a single north-south route through the central station, immediately south of which it has remained 2-track since it was constructed in 1871. Used by all urban/suburban, interurban and cross-border trains to and from the south, this now carries around 250 trains each way per day and has become the dominant constraint in the rail network, with minor delays in any service rapidly affecting all operations. One interviewee commented that other problems of co-ordination and integration in Sweden were minor compared to the need to remove this capacity bottleneck. We were told that this constraint meant that there was no opportunity to increase rail services during the recent trial of congestion charges in central Stockholm: instead additional public transport capacity had been provided by an additional 200 buses, equivalent to the entire fleet of Malmö, Sweden's third largest city, many of them used to operate express services to which rail would have been ideally suited. Unlike most rail investment projects, however, funding for these additional bus operations is limited and they may be withdrawn at the end of 2006.

6.87. There are also tensions in allocating the limited capacity which is available. Banverket has queried whether a longstanding agreement that commuter trains should have priority at peak periods is still valid, as it was signed by the "old SJ" which no longer exists. SL is concerned that, if the agreement is no longer valid, too much of the limited capacity within the urban area will be allocated to long distance trains. Banverket itself makes the final decisions on capacity allocation, rather than an independent regulator such as ORR, and the new process of appeal to Järnvägstyrelsen has not yet been tested.

6.88. As in Dublin, the proposed solution to the rail capacity constraints is for a new north-south "Citybanan" 2-track tunnel through the city centre which would make new areas accessible by rail and remove many of the urban/suburban trains from the congested central station. We discussed these proposals with Næringsdepartementet, SJ and SL.

6.89. As Figure 6.1 shows, Rikstrafiken has no role in investment plans, and Banverket is expected to take the lead in planning the new link, using its own funds provided by the Swedish government. The local authorities, in this case Storstockholms Lokaltrafik ( SL), are able to lend it money, but the status of the operators is only that of consultees and they cannot directly or indirectly specify or fund the infrastructure.

6.90. Banverket develops, agrees with Næringsdepartementet and then publishes its medium term expenditure plans covering not only OMR activity but also enhancements agreed with Næringsdepartementet and the PTAs. The current plans envisage expenditure of around SEK 140 billion over the period 2004-2015.

6.91. However the existence of a plan does not necessarily ensure its delivery, for a number of reasons. First, the estimated costs of Citybanan itself have doubled from around SEK 6 billion to SEK 13 billion. Second, even when costs have been agreed and funding put in place, overruns during construction, particularly in major projects, can require delays, cutbacks and cancellations elsewhere. We were told that there had been cost overruns on a tunnel on the west coast linking Malmö and Göteborg and that any overruns on the Botniabanan project, currently under construction in Sweden's north and costing more than Citybanan, are likely to require a reassessment of other expenditure plans. Third, all sufficiently large projects are likely to be subject to political uncertainty, particularly the desire to direct expenditure to, or focus it in, particular regions. In the case of Citybanan, the new government elected in September 2006 is reconsidering whether the project should proceed and whether another solution could be more appropriate.

The Øresund link

6.92. The Øresund link, completed in 2000, comprises a 4 kilometre immersed tube tunnel, an artificial island, and an 8 kilometre cable-stayed bridge with a main span of 490 metres, carrying a motorway on the upper deck and a two-track railway on the lower deck. The link was built by an SPV owned by the two governments and financed by government-backed bonds. These will be repaid by 2035 from a combination of road tolls and (fixed) rail access charges.

6.93. The construction of the link was expected to help integrate the Øresund Region comprising the Danish islands of Lolland-Falster and Sjælland, Skåne in Sweden, and the Danish island of Bornholm, which has 3.6 million inhabitants. It appears to have been successful in generating steadily-increasing volumes of rail commuting, particularly from Malmö and Skåne to Copenhagen, although after only 6 years of operation the long term effect on patterns of activity within the region is not yet clear.

6.94. Arrangements for the operation and management of the infrastructure are now in place. Capacity allocation, operation and maintenance of the completed infrastructure is carried out by Banedanmark in Denmark and Banverket in Sweden.

6.95. Services on the link currently include SJ trains between Stockholm and Copenhagen and "Öresundståg" trains, operated in cooperation between DSB and Skånetrafik, the Skåne PTA, serving the Øresund Region. However, the intention is for Trafikstyrelsen and Skånetrafik to tender new services from 2008 in the form of "Danish" and "Swedish" contracts to the mid-point of the link, but a number of practical issues must be dealt with:

• Skånetrafik has applied to the Swedish government to tender inter-regional services linking Malmö to Göteborg, Kalmar and Karlskrona along the new Blekinge coastal line, but the government ruled to protect SJ's inter-regional monopoly. Skånetrafik will therefore cooperate with Denmark's Trafikstyrelsen to operate services between Sjælland and Skåne, but SJ will also continue to operate services from Copenhagen, through Skåne, to other destinations in Sweden.
• DSB has historically operated "Danish" services in Sweden between Malmö and Ystad, connecting with the ferry to Bornholm, as part of its PSO contract with the Ministry of Transport, but these would now need to be transferred to the "Swedish" contract let by Skånetrafik.
• Skånetrafik normally lets gross cost contracts, but Trafikstyrelsen normally lets net cost contracts. The current proposal is that both contracts will be gross cost.
• It is not yet clear how separate national contracts to the mid-point of the bridge will work if different operators win the two tenders.

6.96. These three examples illustrate how co-ordination of infrastructure and operations are not the only issues facing the comparators' railways, particularly in urban areas where there is a need to deal, on an integrated basis, with services under the control of different bodies:

• Urban/suburban, inter-regional and even cross-border services operating on the same infrastructure
• Rail, metro, light rail and bus services, whether complementary or as alternative means of addressing the same issues

Summary

6.97. Separation of infrastructure and operations can work well when the capacity of the infrastructure is more than adequate, when the infrastructure manager's maintenance and renewal works do not interfere with services, and when performance regimes incentivise the infrastructure manager to deliver the contracted access rights.

6.98. Predetermined access charges can send few useful signals to most operators, whose services are specified in a PSO contract. Access charges can normally only affect choice of rolling stock when it is replaced and may be immaterial compared with its overall life cycle costs. They are also of little help when investment in infrastructure and rolling stock needs to be coordinated, as is the case with electrification and signalling schemes and many of the changes designed to increase cross-border interoperability.

6.99. Even where commercial freight and passenger services exist, and are able to pay marginal access charges, under marginal cost access charging regimes there is no provision for them to pay the incremental costs of specific enhancement or capacity expansion schemes. In practice, they could rarely afford to do so from additional revenues, and capacity expansion therefore needs to be supported by public authorities on the basis of a social or environmental business case.

6.100. The European Commission understands that investment may be needed to increase rail capacity, and the First Railway Package makes provision for infrastructure managers to identify capacity constraints and to produce plans to remove them. However, infrastructure managers cannot override legitimate national planning and funding constraints and cannot, on their own, optimise investment in infrastructure and operations or in rail and other modes, which may also often need to meet the requirements of more than one tier of government.

6.101. A high proportion of railway investment in Scotland and the comparator railways is required where it is most difficult, in urban and suburban areas requiring modal co-ordination. The issues faced are of integrating planning and funding not only of rail operations and infrastructure but also of rail and other modes.

« Previous | Contents | Next »