Maintenance philosophy in a privatised market

Politically inspired developments in the rail infrastructure environment in the 1980s and 1990s have led to varying degrees of privatisation of construction and maintenance activities.

Firstly, politicians and taxpayers realised that they were paying substantial amounts of money for something which was not patently efficient. Secondly, the (European) political strategy to introduce competition into the rail transportation market meant that the primary role of the railways as train operating companies had to put infrastructure availability at arm’s length. Thirdly, it was recognised that privatisation would bring out the best qualities in the infra management side. That is the legal and financial ownership, making safe and dependable slots available for train operating companies, communicating with local government and guaranteeing system safety. It would also allow the dynamics and the detailed local knowledge of the marketplace to optimise production, productivity, direct response and innovation. It would make both infra management and the contractor accountable and transparent in the balance between costs and the product quality rendered.

These developments led to the formation of separate infra management companies with a firm goal of transparent and preferably cheaper pricing.

Having been established, the infrastructure managers were then faced with the choice of how to make their operations, especially maintenance, more cost effective. Due to the long life expectancy of the pprivermanent way itself, renewals are minor compared to maintenance, and new construction is generally limited to line extensions paid from separate budgets. Maintenance was therefore targeted and cost effectiveness was achieved by:

• An in house overhaul of the maintenance department, e.g. Finland
• Sub-contracting back to the train operating company, e.g. RFF and SNCF
• Privatising the maintenance company, as in the UK and The Netherlands

The choices made were often political and dependent on the confidence of the infra manager with the process, Figure 1.
In the meantime, Sweden has outsourced its maintenance to Svensk Banproduktion and Carillion through tendering, as has Finland. Finland is currently considering selling part of its maintenance division. The UK on the other hand has taken maintenance back in house.

The choices in privatisation

What should infrastructure managers look for in privatisation? They should want to safeguard infrastructure integrity and safety, prevent loss of knowledge and preferably achieve durable savings in maintenance. Main choices in privatisation are:

• Privatising all or part of maintenance
• Privatising to one or more contractors
• Privatising slowly or in one big bang
• Privatising in short-term or long-term contracts
• Privatising specific parts of maintenance or comprehensive maintenance

When contemplating the choice between privatising all or part of maintenance, it should be remembered that privatising too small a part of maintenance may lead to a maintenance contractor that is too small to be viable.A certain critical mass is required to allow innovative ideas and practices to develop, and to safeguard sufficient technical expertise for round the clock fault repair.Also, the management of the new maintenance contractor should be able to absorb the principles of railinfra maintenance and to inspire maintenance crews to strive for higher effectiveness, usually meaning fewer jobs. Certainly Sweden has done well in this.

However, privatising all maintenance work may be risky if the new contractors cannot absorb the large number of men, equipment and knowledge.

Privatising to only one contractor allows the infrastructure manager to observe the results of green-field privatisation without the possible effect of cut-throat competition.Where privatisation is more an administrative outplacement action, as it would seem to be in France, there is a question whether this just answers a legal/administrative demand or if this leads to more efficiency plus lower costs or higher quality.When privatising to more contractors, the way in which contracts are let and the contractors’ previous expertise govern the viability of the process. The Netherlands have been fortunate in that respect.

The speed of the privatisation process also has an effect on the durability.Well considered steps, with the appropriate safeguards in place, will lead to a safe and smooth transition. Hurried privatisation, especially under political pressure, can lead to the loss of all the advantages of both in-house and privatised maintenance: safety risks, knowledge drain, underoptimisation, lower infrastructure quality and more faults. It can be argued that the UK has suffered from this scenario with rushed decisions and question marks on the positioning of Railtrack on the stock exchange, coupled with a major maintenance backlog.

A further question is whether to privatise maintenance in the form of short term (less than five years) or long term contracts (upward of fifteen years).A shorter term could be argued when there is sufficient doubt on the quality of the proposed maintenance provider or when there are insufficient contractual safeguards to remove a poor maintenance contractor. A longer period would be suitable when there is proven maintenance quality or when more efficiency is expected from innovative approaches that need time to earn back the invested monies. One scenario could be to have contractors prove themselves over shorter contract durations and then grant them longer contract periods.

Finally, it could be considered to start a maintenance contractor off on more or less piece-work maintenance (e.g. inspection or repairs as in Sweden) and, as his grasp of the infrastructure increases, to broaden the scope of the contract to include all aspects of maintenance, not excluding maintenance engineering and even a prominent role in life cycle management. Starting a new contractor on the high end of maintenance with no track record in maintenance is obviously not a prudent decision.

In all these options, the confidence that the infrastructure manager has in his own professionalism and in his ability to withstand political pressure will help him to choose a route that will bring him, and the community as a whole, the fruits of privatisation: safe, efficient and innovative maintenance.

The Dutch experience

In The Netherlands, prior to privatisation, a number of rail contractors were generally capable of renewal and new construction work. It was decided that, in order to achieve maximum efficiency, all of the maintenance work, including inspection and fault repair, would be privatised. The maintenance division of the NS (Dutch Railways) was first divided in house into four regions in 1994, with the mandate to internally set up a contract type relationship within the region between the maintainers and the infrastructure manager. Ten divisions were formed to allow each to have the critical mass necessary to support signalling, overhead line and other technical crews and trades.

The divisions were then sold in 1997 to the three interested rail contractors, under conditions that were also vetted and approved by trade unions on both sides. The split was uneven with five going to Strukton Railinfra, three to Volker and two to NBM (now BAM). However, there was an agreement that only one contractor would provide sufficient incentive for driving down costs without loss of quality.

The run up to privatisation had been taken in steady stages, without rushing the process too much. From 1997 onwards, the 39 maintenance contracts were granted for an initial five year period, with a schedule for competitive letting of contracts over a five year period from then on. This schedule was postponed to allow the three maintenance providers, in an alliance with the infrastructure manager, to first increase the all-round performance of the rail sector as a whole, focussing mainly on implementing new safety rules, increasing infrastructure quality and forcing down the numbers of faults.Achieving these goals, without undue time pressure, has been one of the saving graces of Dutch privatisation. During this period it has also been shown that the three maintenance providers are capable of managing maintenance including the more subtle fields of maintenance engineering and innovation.

In a mid-term evaluation by Strukton and ProRail as the infra manager it was shown that the net costs did indeed decrease without quality loss. Special programmes have since then led to declining fault levels, efficiency increases and further cost cuts.

Developments in maintenance philosophy

The main development has been the move from time based maintenance to condition and/or use based maintenance. Time based maintenance has had the advantage of being a generic, almost failsafe philosophy that can be drilled into crews during training and planned into schedules well in advance. This philosophy, developed halfway through the previous century, was designed to maintain the most heavily stressed items of infrastructure. This in built redundancy for the rest of the infrastructure meant that a missed maintenance activity need not always be immediately corrected. It also meant that maintenance was by definition too inefficient and therefore too expensive for the delivered quality. Another effect was that there was little scope for Life Cycle Management as most infrastructure was over-maintained. In permanent way maintenance, where the bulk of the money goes in Holland, work schedules were set in a more pragmatic way, with budget earmarked for categories of work, with the locations still to be decided on an annual basis.

In order to pick up maintenance that is specific for each item or class of infrastructure through condition/use based maintenance, each item or class should actually be registered in an Asset register, together with those quality or usage parameters that will trigger maintenance activities. Therefore condition monitoring on both line (e.g. track or overhead line) and point objects (e.g. switches and crossings) in infrastructure is essential to developing maintenance philosophies. There is also a clear requirement for an IT base to hold this information, letting those concerned access the data for decision making or monitoring purposes. Furthermore, the rules have to be available (or developed) that govern what maintenance should be completed at which quality/usage level. Therefore, risk based FMECA (Failure Mode Effect and Criticality Analysis) type specific maintenance concepts are required that can be utilised at both individual infrastructure elements or at infrastructure group level. One of the major steps is the scaling down, under specific conditions, of foot patrols and switch inspections to more reasonable levels, with the advantage of less exposure of inspection crews to safety hazards.

The use of measuring data for optimising track and overhead line maintenance has been covered in Holland through the use of the Eurailscout measuring fleet and the IRISsys software for visualising and comparing results. The measurement trains cover the track and overhead line geometry and condition plus ultrasonic monitoring of track integrity. These are complemented by spot checks on ultrasonic faults and visual inspections and the input from foot patrols. Intelligent use of data and the freedom to optimise maintenance have led to better track geometry in combination with cost reductions. For a contract in the North with this kind of condition based maintenance, a reduction of tamping jobs by 10-20% and a reduction of tamping costs of 30-40% is achieved.

A recent addition to the armoury has been the commissioning of video inspection vehicles (Eurailscout) and systems (HORUS) to patrol and view switches, especially at busy hubs, with off-board evaluation in the office. The safety advantages and the decrease in the use of equipment for inspection are promising.

Developing and implementing condition monitoring has been led through Strukton’s POSS family of systems, allowing critical objects such as switches and crossings and even signalling circuits to be monitored for probable impending failure. This technique has been broadly reviewed in other articles. The proactive approach enables crews to remedy faults before they occur. It is however difficult to compare current practice to the ‘might have been’ situation with no intervention.

As demonstrated above, it is not just having the right equipment that lowers costs. The trick is to implement their use into the daily maintenance planning and execution. This requires a change in culture in the crews and is therefore not an overnight process. Explaining the why’s and wherefores is as important as the how’s and when’s.At the end of the day, more quality for less cost implicitly leads to fewer men to do the work and this subject should not be avoided in discussions with the crews.What should also be discussed is cutting out unproductive and unnecessary work. Finally, it should be realised that on demand maintenance requires better planning and/or higher peaks in maintenance crews.

Challenging generic maintenance rules with risk based/ FMECA type specific maintenance concepts is another development that allows more effective maintenance work patterns. In a joint study by the maintenance providers and the infrastructure manager this philosophy showed the potential for an annual €10 million (5%) saving compared to the current practice. This is caused by tailoring maintenance plans for a group of infra objects or even for specific objects, rather than keeping maintenance to higher generic levels. Several studies confirmed that substantial reductions in maintenance costs are possible when the maintenance provider is given the scope to use his own FMECA concepts. Of course it is the infra manager’s task to guard against overoptimistic maintenance plans.All in all, more effective work is possible.

The Swedish experience

In Sweden, the ambition of the client is to get ‘more railroad for the money’. The belief is that competition drives efficiency and innovation, etc. Tendering maintenance contracts and projects over the past three years has led to better quality, lower costs and a decrease in train delays. The same high safety level was reached as before the privatisation.

Svensk Banproduktion, the maintenance daughter of Storstockholms Lokaltrafik, and partly owned by Strukton Railinfra, was awarded a number of the contracts that were tendered. Comparison between Dutch and Swedish contracts and the underlying maintenance concepts has been an enlightening experience. The distinct impression is that, although the tendered volume is not overwhelming, the exercise has been a useful one to all parties.

Conclusions and recommendations

Some recommendations can be made when managing privatisation:

• Safeguard infrastructure integrity and safety without saying no
• Prevent loss of knowledge and push investment in knowledge
• Achieve durable savings in maintenance by allowing innovation
• Privatise all or most of maintenance, with a call-back option
• Privatise to more than one contractor if the market is big enough
• Privatise slowly in a controlled fashion, using an alliance where feasible
• Privatise in short-term contracts initially, move up (and move back) when feasible
• Similarly privatise most of maintenance first, comprehensive maintenance later
• Don’t rush, allow time for the cultural transition

In retrospect, this looks like a paraphrase of a guideline for managing teenagers.

On the more technical, maintenance management side:

• Move from time based maintenance to condition and/or use based maintenance
• Use an IT system suitable for use as an Asset Management tool
• Establish the relevant static data for your assets
• Use condition monitoring, inspection and measuring systems plus suitable software
• Fit in this dynamic quality, condition and usage information
• Build up your faults (technical) and incidents (safety) databases
• Share information, don’t sit on it
• Develop specific maintenance concepts through FMECA type routines
• Allow maintenance concepts to be challenged
• Allow innovation to develop and (dis)prove itself
• Define and demonstrate the Key Performance Indicators achieved
• Do not underestimate the cultural impact, allow time

It can be concluded that privatisation in Europe is here to stay. Whereas some experiments have backfired, in general the involved infrastructure managers are satisfied with the progress being made and the benefit that privatisation brings. The Netherlands and Sweden have been able to increase quality while decreasing net costs without compromising safety. We will be happy to discuss the chances, the pitfalls and the effort necessary to make privatisation a success.

Figure 1: The state of play in the rail infra market in 2004 drawn by ProRail