Splitting railway infrastructure and operations is seen by politicians and economic gurus around the world as a panacea to problems of road-rail competition, poor cross-border service quality, and many other ills. But will the methods being adopted achieve the desired results? In an exclusive survey for Railway Gazette International, David Burns reviews the state of progress

WHY ARE RAILWAYS around the world busy finding ways to separate their infrastructure and operations? There are those that think it is the solution to many of the major problems of the industry. Separation is seen as a means of enabling the railways to compete more fairly with roads, permitting a clearer means of subsidising the industry, allowing competition for business to take place on the track, increasing the efficiency of track maintenance, and encouraging the movement of freight trains across national borders. In other words, a panacea.

The recent initiative for separation has come from European Union legislation which, in a series of directives starting in 1991, has stated that the national railways in EU countries should separate the accounting for infrastructure in such a way that the costs are transparent. It has also directed a limited degree of open access, whereby qualified operators can run on the tracks of any railway within the EU.

Not only have the various EU member governments accepted this policy, a number of other countries have accepted their railway's argument for the need to subsidise track through this approach. However, in most cases, the governments are requiring open access as a condition for providing this kind of subsidy.

It is interesting to note that, whilst the staunchly privately-owned railways of North America are very much opposed to separation and open access, with the latest mergers creating mega-railways there is more and more discussion amongst shippers about the need for open access to ensure competition. In fact, permission to merge is increasingly conditional on the requirement to grant trackage rights to competing operators.

Comparison with other modes

If railway track, from an operating perspective, had the characteristics of a road, separation would be a simple matter. Essentially, complete freedom of access can be granted to a user. Roads - even single lane - are bidirectional. Faster vehicles can easily overtake slower ones, and should everyone want to arrive at 08.30 there will be a traffic jam. The driver will soon take an alternative route or time, making the system largely self-regulating. Pricing is simple: a fixed access tax, a user payment in the form of fuel taxes, and an occasional toll. But it is interesting to note that congestion pricing is starting to become a reality.

The separation methodology

Individual governments, ministries and railway managements, many with the assistance of consultants, have devised the methods used for separation and subsidy. Table I shows that there has been little standardisation in approach. Most railways appear to have adopted the public company structure, and whilst some have indicated the possibility of privatisation in the future, only one has taken this approach from the outset.

There is a saying 'he who controls the signals, controls the railway!' But for this all-important aspect of real-time traffic regulation, the approach varies considerably. It is obvious from the survey that many railways are very reluctant to give up control. But for open access to be successfully implemented, it is equally obvious that control has to be in the hands of the track body or an independent organisation.

Access charges

Accurate pricing of track capacity is a very complex task, given the very high fixed cost element and the almost infinite flexibility in pathing through the payoff between speed range and headway. Years of research have gone into reviewing what Amtrak should pay the private US railways for the use of a track that is primarily carrying freight trains, and there is still no sound conclusion.

It is obvious from Table II that there is a wide range of approaches being used to the question of pricing track access. Some railways are trying to account for all combinations and permutations, while others appear to be taking the approach that the simpler the better. Some publish their charges, others negotiate in secret. Many charging regimes are still in the process of development, redevelopment and further refinement! Given these different approaches, it is quite easy to see the accounting problems facing the potential third party operator of international freight trains. If the railways are truly to compete with roads, there has to be a relatively simple, transparent, and standardised charging structure.

Infrastructure subsidy

It is difficult to determine the exact level of subsidy that the infrastructure organisations are getting without detailed knowledge of the railway's maintenance, renewal and investment programme. It is clear, however, that the range of subsidy paid is from 100% of costs to zero. In one case, an infrastructure company actually made a substantial profit and even paid taxes on it!

In most cases, infrastructure subsidy is subject to annual government appropriations. This makes it subject to the current government budget, and hence long-term planning is difficult or impossible. The British approach of privatising the infrastructure company may provide a way to overcome this short-termism and enable long-term planning.

By placing the subsidy entirely with the operators, any reduction could have an immediate impact on service levels, making it politically more difficult to achieve. Only time will tell what will happen to services in Britain, where subsidies to the train operators are budgeted to fall, some to zero or beyond, over the periods of the franchises.

Infrastructure maintenance

Most railways have taken the separation of infrastructure as an opportunity to restructure their track maintenance. Table III illustrates various aspects of the maintenance task that they have chosen to contract out.

Many railways are now using contractors for various aspects of infrastructure maintenance and renewal, but it appears that only a small minority are using contractors for inspection. Of those companies which did indicate the length of contracts, three to seven years seems a typical duration.

The use of contractors for maintenance and renewal has expanded rapidly, primarily through the privatisation of the railways' in-house track maintenance organisations. Those infrastructure organisations that have given long-term contracts will probably face a significant problem in the future, as when the contract does come up for renewal it is unlikely that any company will have the specialised equipment and personnel to bid against the existing contractor. This trend is already evident from the problems that can be seen with railways that have traditionally used long-term contractors.

Inspection, like any good quality assurance programme, must be independent. Yet in several cases the maintenance contractor is also the primary inspector. Very interesting conflicts can arise when, for example, the track owner does not give a contractor sufficient funds for maintenance and the track deteriorates to the point of being unsafe.

It would appear that the infrastructure owners should keep control of the day-to-day maintenance of the track, so that it is intimately familiar with, and responsible for, its condition.

Marketing the infrastructure

Most railways indicate that they have a specific organisation for marketing train paths (Table IV). However, for those organisations where the infrastructure business is still part of the railway, they are really only 'marketing' capacity. This could be considered a conflict of interest. To a large extent this is the system used in the USA, but there shared access only takes place where it is mutually beneficial.

Some railways have set up a separate 'Network Access Group', while a number of European companies are becoming members of route-specific 'One Stop Shops' where an operator can go to obtain a multi-national train path. Given that our survey indicated that paths can take anywhere from one day to three months to be assigned, and that the method of infrastructure use payment varies considerably, it is important that there be a standard approach to this task. For true multi-modal competition it must be as simple as filing a flight plan.

One of the significant problems appears to be a lack of standardisation in computer simulation models for planning of train operations. A number of railways are using different databases and models, which are almost certainly not compatible, so to be able to schedule multi-national trains quickly will still not be possible.

Many questions appear to be outstanding. For example, certain train paths are of more value than others - an 08.30 arrival in a major city will be worth more to an operator than an 11.30 arrival. How does a new operator obtain that 08.30 slot when the existing company has probably had it for 100 years or more? Possibly paths could be bought and sold, in a similar approach to that used by the airlines.

Other assets

In the separation process, there are important questions about the ownership of related assets, such as stations or rolling stock maintenance depots (Table V). In the majority of cases, our survey found that stations have remained with the operator, which must inevitably create a barrier to any new entrant. In the USA Amtrak now owns many of the stations it serves, but in the days of multiple passenger operators most major stations were owned by a terminal company that was independent or jointly owned by the major users.

Maintenance depots have either remained with the operator, or been sold to third party companies. In Britain, the freight workshops were sold to the new operator, but passenger facilities are leased. The responsibility for developing the infrastructure assets further and generating other sources of revenue seems to be divided. Logically, these developments should contribute to the financial benefit of the railway as a whole. But in some cases the benefit goes directly to the government or to the shareholders.

Train operating companies

Most national railway operations have been transformed into a separate public corporation, whilst a few have remained as a business unit within an integrated railway system, only separating their accounts (Table VI). Only in Britain has the operation of all services passed to the private sector, through passenger franchise concessions and outright sale of the freight services. Most railways have publicly indicated that private operators 'would be welcome', but privately some are quite negative about competition on the tracks that they have controlled for generations.

The most dramatic approach has been in Great Britain, with the breaking up of passenger operations into 25 separate, and in some cases competing, concessions. The franchise terms run from 5 to 15 years, depending primarily on the willingness of the franchisees to acquire new rolling stock. To get around the problems of buying long-life vehicles within a seven year franchise, the government sold all the rolling stock to three separate companies which are leasing them back to the franchisees. It is safe to say that this approach is still in a state of flux, and it will probably be a long time before the situation stabilises.

How to subsidise

One of the clear questions that needs to be resolved is where and how do you subsidise a railway? Should it be for the track or for the operations? Should the money come from central, regional or local governments? Once again, there is a clear difference in approach. At one extreme is Britain, where all the subsidy (except for some grants towards the construction of new routes and freight terminals) is channelled entirely through the operating companies. Other countries are essentially subsidising the provision of infrastructure, and at the other extreme in some cases picking up all the track costs.

Many railways and infrastructure organisations objected to the suggestion of 'subsidy', insisting that they had been awarded a contract for providing a specific service. Yet virtually all railways, with the possible exception of some US freight operations, are in receipt of subsidy in some form or another. The larger systems generally get their subsidy from central government, whilst smaller operators receive funding from central, regional, local and even city administrations.

While not many infrastructure organisations indicated track charges as a percentage of the operator's total costs, several of those that did indicated figures ranging from zero to 40% or even 50%. The very low figures suggest that in these countries some form of subsidy is going into the infrastructure, while the high access charges suggest the subsidy is channelled via the operators. It is difficult to conceive how a rail operator can successfully compete with roads when 50% of his costs are effectively fixed in the form of infrastructure charges.

There is no question that subsidy should be provided in the form of a contract for service. However, to be effective, there must be appropriate incentives and penalties built into the contract. It would seem to be more effective if these contracts are held with the operators rather than the infrastructure owners.

Safety and accidents

Most railways indicated that the cost of an accident would be assigned to the party causing the accident. This is far easier to say than to carry out in practice. When the costs were all carried within a single company, the cause was critical - find out what happened to prevent a recurrence - but financially it made little difference where the costs were assigned.

However, with two, three, four or potentially even more for-profit corporations involved, assignment of blame becomes far more important. As a result, the often difficult task of accurately apportioning blame becomes far more critical, and it can get out of all proportion for a small accident. Some form of no-fault 'insurance' will become a necessity.

In conclusion

It is clear that, particularly in the European case, individual national approaches will not solve the problem of encouraging competition and, more importantly, multi-national movement of freight.

There is some concern with total separation, about who decides whether investments should be made. In one sense co-operation between the companies can solve this problem, although it becomes more complex where the benefits of a project are split between a multiplicity of companies. A recent example is the agreement between Virgin Rail and Railtrack to upgrade Britain's West Coast main line, which will impact on at least seven other passenger franchises and three freight operators, quite apart from any future open access entrants in the market.

Given the need for commercial efficiency and, in certain areas, the need for some form of subsidy, there are many good examples where an appropriate contract for service can result in significant increases in revenue and reductions in costs. It is, however, critical that these contracts be written for the granting agency by people who are very knowledgeable in the detailed operation of a railway.

Many railways have indicated that they are, or plan to be, participating in a 'one stop shop' for allocating train paths, particularly on the so-called Freight Freeways (p313). Unfortunately, it appears that certain key railways either have no near-term intention of allowing any other operator to run on their network, or conversely plan to dominate the international freight train market. It is therefore critical, particularly in the multi-national context, that a neutral organisation with power to penalise should be created as soon as possible to play an active role in creating a central train path database, standardising requirements, and acting as a clearing house for wagon and track use charges. In many respects this organisation could be similar to the Association of American Railroads.

It is apparent that the present approach to operating railways must be changed significantly if they are to play the role that they logically should in the future. It is apparent from this brief review that the need for change is recognised in most countries.

While it may be too early to comment positively on the effectiveness of infrastructure separation, it is clear that the evolution process has a long way to go before the goals can be achieved. What, ultimately, the organisations will be is difficult to say. But it must be recognised that the future demands a different kind of railway than that which has existed before. o

TABLE: Table I. Types of infrastructure organisation

Country Who owns the Who do they Who controls

infrastructure report to the signals

Australia (QR) Railway department Government corp Contracted to operator

Australia (RAC) Public corp 2 shareholding Contracted to ministries one operator

Austria Public corp Ministries of Transport Railway and Finance

Denmark Public corp Government ministry Track company

Finland Public corp Government ministry Contracted to operator

France Public corp Government ministry 'Delegated' to operator

Germany Public corp Government Track company

Great Britain Private corp Shareholders Track company

Netherlands 3 public corps (track, Government Track company

signals & consultancy)

Norway Public agency Railway board Track company

Portugal Public company Secretaries of Will be track company

Transport and Finance

Spain Railway department Railway/Government Railway

Sweden Govt administration Government Track administration

Switzerland Railway department Railway/Government Railway

TABLE: Table II. Access charge regimes

Australia (QR) Under development, will include time-of-day element.

Australia (RAC) Individually negotiated between incremental and fully allocated cost. In future will have price differentiation based on rolling stock characteristics, level of priority, train paths, track quality required, market needs, and perhaps time-of-day.

Denmark Charges are based on track, location, traffic density, track quality requirements, time of day, and fixed links (specific bridges, etc).

Finland Variable fees for passenger-km and gross tonne-km. Fixed fee for net tonnes carried.

France At present based on track, location, traffic density, annual fixed charge, time of day, and track quality, but subject to change.

Germany Charges are based on track quality, commercial importance of routes and train category, with adjustments for frequency, pathing, punctuality and contract length.

Great Britain A complex pricing structure negotiated separately with each operating company.

Netherlands Only for Trans-European Rail Freight Freeways, based on tonne-km.

Norway Long-term marginal cost covered by a tonne-km charge.

Portugal At this time there are no charges.

Spain There are no charges as yet, since it is still a department of the railway and there are no non-Renfe users.

Sweden There are three levels of charges: an annual fixed payment for each locomotive and item of rolling stock, a variable charge for gross tonne-km related to type of vehicle, and ancilliary charges for traffic control, energy, fuel and shunting services.

Switzerland Based on train path price.

United States Can be based on tonne-km, train-km, wagon-km or charge per path used.

TABLE: Table III. Contracting out of infrastructure maintenance

Country TR TM/P TM/D INS SM Safety agency

Australia (QR) New construction only - - Internal / Dept of Transport

Australia (RAC) Policy currently under review Rail Access Corp

Austria X X - - - Internal to railway

Denmark X - - - X Separate agency

Finland X X X X X Track administration

France X X X - X Internal to railway

Germany X X X - X Separate agency

Great Britain X X X X X Internal, with supervision from government agency

Netherlands X X X - X Owning company does inspection

Norway - - ? - - Track administration

Portugal X - - - - To be by Railway Institute

Spain X X - - X Internal to railway

Sweden X X - - X Track administration

Switzerland - - - - - Internal to railway

USA X X - - X Railways and federal agency

TR = Track Renewal; TM/P = Track Maintenance, Periodic; TM/D = Track Maintenance, Day-to-day; INS = Inspection; SM = Signalling Maintenance

TABLE: Table IV. Marketing the infrastructure

Country Specific Path Time to Does original railway: % of trains

Depart- Clearing obtain allocate get priority by new

ment House path paths ? assignment ? operators

Australia (QR) Y N - N Y 2

Australia (RAC) Y P 2 weeks N Y 100

Austria - - - - - -

Denmark Y P - N Y/N 1

Finland N - - N Y 0

France Y/N1 P - Y Y 1

Germany Y Y - N N -

Great Britain Y P 3 days - 1 year N Y 100

Netherlands Y Y 1 day - 3 months N N 1

Portugal P - - - - -

Spain N - - - - -

Sweden Y - - - - -

Switzerland Y Y 1 to 30 days N Y -

USA Either negotiated between railroads or dictated as a condition of merger

1. Department is within SNCF not RFF P = Planned

TABLE: Table V. Who controls and develops non-track assets?

Country Stations Maintenance Workshops Property

Depots Development

Australia (QR) Operators Operators Third party Property division

Australia (RAC) Operator Operators Third party Owner

Denmark Operator Operator Operator Operator

Finland Operator Operator Operator Track company

France Operator Operator Operator Operator/track authority

Germany Operator Operators Operators Railway business

Great Britain Track company Freight operator - Track company

Passenger leased Third party

Netherlands Separate co Operator Railway business Two companiesPortugal Operator Subsidiary Subsidiary -

Switzerland Track authority Operator Third party -

TABLE: Table VI. Structure of principal train operators

Country Type of organisation Track charges as Source of subsidy % of total costs

Australia (QR) Railway department 40% Regional government

Australia (RAC) Public and private - Regional government

Austria Public corp - Central and regional govt

Denmark Public corp ?»20% Central government

Finland Public and private ?»10% Central, local and city (purchase of services)

France Public corp - Central and regional govt

Germany Public and private - Regional government

Great Britain 25 passenger franchisees, 50% Central, local, city

3 private freight companies 25% None

Netherlands Public corp Under development Central government (service contracts)

Norway Public corp - -

Portugal Public corp - -

Spain Integrated - Central and regional govt

Sweden Public corp 15% None

Switzerland Pass: public corp - Central and regional govt

Freight: private corp 1 - -

1. Recently formed alliance with Italian freight railway company

CAPTION: If the track is owned by a subsidised public-owned entity, who decides where track investments should be made, the owner or the track users?

CAPTION: If the station remains the property of the original operating company, it could make life more difficult for a new passenger operator

CAPTION: Modern technology, such as computerised train planning, could make open access and multi-national trains a success. However, to achieve this the computer models must work to a standard format. This example is by Comreco Rail


This article is based on a survey of some of the railways which have already separated their infrastructure, or are in the process of doing so. In addition to the EU countries and those mentioned in the tables, others include Chile, Indonesia and the Czech Republic. We are grateful to all the organisations which responded

Séparation des infrastructures : cela va-t-il fonctionner?

Beaucoup de chemins de fer du monde se séparent en deux entités, l'une contr?€?lant les infrastructures, l'autre l'exploitation, une démarche que les leaders politiques et économiques considèrent comme étant la panacée capable de résoudre les problèmes de concurrence entre rail et route, d'améliorer la qualité des services internationaux et de gommer bien d'autres maux. Mais les méthodes choisies parviendront-elles à atteindre les résultats escomptés? David Burns passe en revue les progrès réalisés, et nous fait découvrir que l'évolution de l'organisation des chemins de fer modernes a encore un long chemin à parcourir

Abtrennung der Infrastruktur: der Weg zum Erfolg?

Auf der ganzen Welt trennen Eisenbahnen ihre Infrastruktur- und Betriebsbereiche. Dies wird von Politikern und Wirtschaftsgurus als Wundermittel gegen die Strassenkonkurrenz, die schlechte Qualität der grenzüberschreitenden Leistungen sowie gegen viele weiteren Übel gepriesen. Aber erreichen die gewählten Methoden die gewünschten Resultate? David Burns gibt einen überblick über den Fortschritt und stellt fest, dass die organisatorische Entwicklung der modernen Eisenbahnen noch einen weiten Weg vor sich hat

Separación de la infraestructura: ?€Funcionar? ?

Las compañías de ferrocarril de distintas partes del mundo est? n dividiendo el control de su infraestructura y operaciones, en una acción considerada por políticos y gurús económicos como una panacea a los problemas de la competencia entre carretera y ferrocarril, la baja calidad del servicio transfronterizo y muchos otros males. Pero ?€conseguir? n los métodos elegidos los resultados deseados? David Burns repasa el progreso realizado y descubre que la evolución organizativa del ferrocarril moderno tiene todavía un largo camino por delante