article

Winter delays and disruption – combating the cold

Posted: 20 September 2010 | | No comments yet

The hard winter of 2009/2010 led to major disruptions to train services in Sweden. In April 2010, the Swedish Transport Administration initiated an inquiry intended to show how great the delays were that affected passenger and goods services, how great a cost to society they entailed, what caused the delays, the factors that interacted to contribute to the situation that arose, and what the Swedish Transport Administration can do, principally from a macro-economic perspective.

The hard winter of 2009/2010 led to major disruptions to train services in Sweden. In April 2010, the Swedish Transport Administration initiated an inquiry intended to show how great the delays were that affected passenger and goods services, how great a cost to society they entailed, what caused the delays, the factors that interacted to contribute to the situation that arose, and what the Swedish Transport Administration can do, principally from a macro-economic perspective.

The hard winter of 2009/2010 led to major disruptions to train services in Sweden. In April 2010, the Swedish Transport Administration initiated an inquiry intended to show how great the delays were that affected passenger and goods services, how great a cost to society they entailed, what caused the delays, the factors that interacted to contribute to the situation that arose, and what the Swedish Transport Administration can do, principally from a macro-economic perspective. The snowy and cold winter of 2009/2010 caused severe disruptions to rail services and serious consequences for both the National Rail Administration, train operators and the economy as a whole. The extreme cold in combination with a lot of precipitation gave very poor conditions for operating trains. The weather was unusual but not an isolated phenomenon. The total number of delay hours totalled 83,000, where just over two thirds affected goods services and the rest affected passenger services. This is twice as many as a normal winter. The cost to society for the winter’s delays was SEK 3 billion. With measures corresponding to initial investment of SEK 410-450 million, half of the delays would have been avoided. Seen as a whole, about half of the delays in Sweden can be said to be caused by factors that the Swedish Transport Administration can influence and the other half by factors that fall under the control of the train operators, though it is difficult to draw a strict line of demarcation between causal factors and responsibility, as many causes are linked. However, it is clear that collaboration between the Swedish Transport Administration, contractors and train operators is necessary to limit problems caused by both winter conditions and other causes of delay. Deficiencies in four areas contributed to the situation that arose. These areas were:

Infrastructure

The technical standard of the facilities does not fully meet the requirements for effective handling of train services during long periods with large amounts of snow and cold. Moreover, rail yards in some places contain highly sensitive technical apparatus that make snow clearing more difficult and in general, contact lines along the track are not protected from ice formation. It is unreasonable to expect infrastructure that can 100% withstand all types of weather conditions with unchanged capacity. The investments that would be required to achieve that would be extremely large, and of doubtful benefit to the national economy. Consequently, it is a matter of ensuring effective maintenance of existing infrastructure and considering targeted investments where they will do the most good. The low level of maintenance and reinvestment since the mid-1990s has led to an estimated accumulated ‘maintenance gap’ in the order of SEK 15 billion (to establish the exact level a comprehensive investigation of the actual infrastructure would be needed). This is also shown by a fall in the estimated remaining life of the facility.

Interface with contractors

The Swedish Transport Administration has engaged contractors to take care of such things as maintenance, winter preparations and snow clearing. A number of deficiencies in both preparations and snow clearing were revealed during the winter – in some cases all the necessary preparations had not been carried out (for example setting out snow shields), in some cases snow clearing was not started in time, the contractors did not always have sufficient resources or sufficiently experienced personnel, and so forth. The deficiencies had several different underlying causes: necessary winter preparations by the contractors are not always specified by the Swedish Transport Administration; function-based contracts are often designed in accordance with a normal winter rather than allowing a margin for more adverse winter conditions; in many quarters there is not enough collaboration and joint training and planning between contractors and the Swedish Transport Administration; and in some cases it can be discussed whether the contractors really fulfilled all their obligations.

Trafikverket are aware that they must continue to do everything they can to prepare for the next winter

Trafikverket are aware that they must continue to do everything they can to prepare for the next winter

Internal management and processes

Priorities and redistribution of services when capacity is reduced

When weather or other circumstances have led to reduced capacity in the system, there is too little decision support for services in the form of pre-determined plans for train service priorities, reduction of services or redirection of services to alternative routes. In the cases where there are contingency plans, they are often drawn up in different ways in different regions, which complicates operations, as decisions affecting trains often have an impact on several traffic management centres.

National traffic management

The national management structure with national operational management is not dimensioned or trained to take optimal national decisions on service priorities and snow clearing. Moreover the structure, with eight regional traffic management centres – all equipped with different technical systems for train control – makes effective coordination of action and measures more difficult when there are major disruptions. Nor do communication and coordination between different traffic management centres and between traffic management centres and national operational management function optimally due to such things as the lack of guidelines, system support and training.

Flexibility and staffing

In many cases there is a lack of flexibility of resources for traffic management, which leads to a lack of clear potential or plan to increase staffing of critical functions when there are major disruptions.

Information to passengers, operators and the community

The past winter revealed problems concerning both collection of correct information and how to communicate the information available. The situation was made more difficult in that the Swedish Transport Administration is often dependent on information from operators and contractors to be able to compile updates. The problems of information management were partly organisational and partly technical. The organisational problems consisted of a lack of plans and guidelines for how to provide information and what information to provide for different types of events. For example, it was noted that there was a resistance to proactively issuing reasonable estimates of the extent of a delay. Instead, during the winter it was more common to delay expected departure times from a given station by 10 minutes at a time. Moreover, during the winter there was little flexibility in the mandate for approval of information, which in some cases led to information becoming obsolete before it could be published. The technical problems mainly consisted of restrictions in the form of underdimensioned and older technology in information systems, which prevented the communication of correct information to passengers. For example, the system of loudspeaker announcements at train stations was under-dimensioned, which meant that on several occasions it was out of order, resulting in no or incorrect announcements to passengers.

Measures in internal management and processes have the greatest macro-economic effect

By dealing with these deficiencies, many of the delays could have been avoided and negative effects reduced. With measures corresponding to initial investment of approximately SEK 500 million, half of the winter-related delays would have been avoided. In addition, several of these measures also have a positive impact on disruptions not due to extreme winter conditions, which increases the macroeconomic benefit. The greatest positive net effect comes from measures in processes and management, mainly alternative train timetables and plans to adapt services when capacity is reduced due to disruptions, and improved national operational management. Technical solutions also have the potential of solving some parts of the problems, but, due to the extent and geographical distribution of the problems, the cost of implementing technical solutions is often too high to be justified from a macro-economic perspective. It is not possible to assume that the weather conditions of the past winter were an isolated phenomenon, so it is important that the Swedish Transport Administration takes necessary steps to reduce the effects of another hard winter. Since the Swedish Transport Administration cannot implement these measures in a vacuum, it is also important that operators, contractors and others involved participate and contribute as needed.

Trafikverket are aware that they must continue to do everything they can to prepare for the next winter

Trafikverket are aware that they must continue to do everything they can to prepare for the next winter

Positive macro-economic effects

The majority of the positive macro-economic effects are driven by a small number of measures in a limited number of geographical areas. Over 50 different measures to reduce sensitivity to disruptions and avoid delays have been identified and analysed. The measures have also been differentiated on the basis of different regions, since one measure that is macro-economically positive in an area with a lot of traffic could very well be negative in an area with less traffic. For each possible measure, the macroeconomic value added (benefit minus cost) has been estimated and the measures can then be prioritised on the basis of this value added. A number of interesting observations follow such an analysis. For one thing it is shown that measures in the eastern operating area reduce delays and macro-economic costs equivalent to half of the total delays and macro-economic costs for the country. Moreover, it can be established that measures in processes and management such as alternative train timetables, reduction of services and optimisation of national operative management account for the clear majority of the total macro-economic value added. Most measures in processes and management have a positive macro-economic value added in all areas of operation. At the same time, several infrastructure measures, such as rail switch upgrades, are more dependent on being carried out in the most important places in the network in order to be macro-economically motivated. At the same time it can be noted that one technical measure for increased line clearance in the eastern operating area, is the measure with the next greatest macro-economic value added of all measures. All-in-all, an implementation of the positive macro-economic measures would lead to a reduction in delay costs in a normalised year of approximately SEK 590 million.

Arguments in addition to purely macro-economic considerations may justify implementation of more measures

A further reduction in delays is possible through more measures. However, these measures are costly in relation to the economic benefit of the reduced delays, but can of course be considered for other reasons. It is also the case that several conceivable advantages of measures are not fully reflected in a purely macro-economic estimate, for example the possibility that improved punctuality of train services may make commuting more attractive, or that individuals decide to live in new places due to good communications. If in using such arguments there is more willingness to invest in reducing delays than is motivated in purely macro-economic terms, further measures could be considered, including a substantial increase in contractors’ staffing, extensive increase in de-icing facilities and increased upgrading of rail switches.

A marked rise in value of goods services leads to a further four measures becoming macro-economically positive

According to the accepted methodology for calculating macroeconomic costs used in this work, one hour of delay in goods services is assigned a low value in comparison with one hour of delay in passenger services. For the purposes of analysing how sensitive the conclusions are for various delay cost assumptions for goods services, the past winter has also been analysed on the basis of an assumption that the cost of delayed goods services is in parity with the cost of delayed passenger services. This assumption corresponds to a tenfold increase in the established macro-economic estimates for delays in goods services. Using this assumption, the macroeconomic benefit of several measures increases, and measures in four areas that previously had a negative macro-economic benefit (net) become macro-economically positive – increased line clearance in the north, resources for contractors in central Sweden, deicing in central and northern Sweden, and some upgraded rail switches. These measures lead to further reduction in delays of approximately 1,700 hours, or just over 10% for a weighted average winter. The further investments that would be required to achieve this potential amount to approximately SEK 60-80 million. The internal order of priority of measures, however, is not appreciably changed on the basis of the higher assumed macro-economic cost of delays in goods services. The reason for this is that most measures affect both passenger services and goods services positively. However, it can be noted that an increase in the relative cost of delay in goods services from a macro-economic perspective means that an increasing number of measures in the northern traffic management area become relevant.

In conclusion…

In summary, first and foremost the Swedish Transport Administration must continue to do everything within reason to keep the infrastructure ready for difficult weather conditions, then to manage reduced capacity when there are disruptions in as optimal way as possible, and finally to inform the public and third parties as quickly and correctly as possible of unavoidable disruptions.

About the Author

Gunnar Malm

Gunnar Malm has been the Director General of Trafikverket since its inception in April 2010. Previously, Gunnar has worked in many organisations including, among others, Managing Director of Arlandaban Infrastructure AB, BANQUIT AB, Translink AB and SJ.