Estimating the damage and marginal cost of different vehicle types on rail infrastructure: combining economic and engineering approaches

EU legislation requires that European infrastructure managers set access charges based on the marginal cost of running trains on their networks. Two methods have been used in the literature for this purpose. Top-down methods relate actual costs to traffic volumes. Bottom-up methods use engineering models to simulate damage and then translate damage into costs based on assumptions about interventions and their unit costs. Whilst top down methods produce sensible results for marginal cost overall, they have struggled to differentiate between traffic types. The challenge for bottom-up approaches is how to translate damage into cost, with numerous assumptions being required which may be invalid. This paper proposes a new, two stage approach to estimating the marginal cost of rail infrastructure usage. The first stage uses engineering models to simulate damage caused by vehicles on the network. The second stage seeks to establish a statistical relationship between actual costs and damage. It is thus possible to convert damage estimates into costs using actual cost data, rather than through a set of potentially invalid assumptions as in previous approaches. Only the first stage is implemented in this paper. We show that it possible to produce total (annualised) damage measures for three damage mechanisms on five actual track sections in Sweden. Once extended, it will be possible to model the relationship between damage and actual costs for the first time; and thus better understand the relative costs of the different damage mechanisms and in turn inform the level and structure of track access charges.