Transportation of Dangerous Goods by Railway. Costs Imposed on People Living Nearby



Transportation of Dangerous Goods by Railway. Costs Imposed on People Living Nearby

Authors

L Hiselius, Lund Institute of Technology, SE

Description

Abstract

Introduction
In Sweden, a model has been developed for calculating the expected costs of accidents due to transportation of dangerous goods by road and railroad respectively. The cost due to the value of a marginal risk reduction (the so-called human value), that is applied in the model, is however calculated on general road traffic accidents and not on accidents involving dangerous goods for road and railway traffic. Research indicates, however, that the individual perceived risk varies substantially depending on the accident type studied. Based on this result, one may also argue that the value of a marginal risk reduction varies over different accident types. That is, the value of a marginal risk reduction may vary depending on whether it is calculated for an average road accident or a road accident involving dangerous goods. One may furthermore argue that the value of a marginal risk reduction varies depending on whether road or railway traffic is studied.

Consequently, when discussing investments affecting transportations of dangerous goods, it is the cost due to the value of a reduction in the risk of accidents involving dangerous goods that is ought to be used. There is then an interest in estimating this cost for road and railway traffic respectively in order to appraise accident-preventing projects correctly and to improve the model estimating the costs of accidents due to transportation of dangerous goods currently being used.

Formulating the problem
The cost due to the value of a marginal risk reduction is often estimated as the individual willingness to pay for a reduction in the risk. However, it seems like an awkward task to ask affected people how much they are willing to pay for a specific reduction in the risk for an accident involving dangerous goods, since we are discussing small probabilities that may be hard to understand and relate to other risks. The accident outcomes in case of an accident involving dangerous goods are furthermore quit divers depending circumstances around the accident. To avoid this problem the accident risk can be more generally described. A reduction in the accident risk may instead be expressed as a reduction in the quantity of dangerous goods being transported. The amount of dangerous goods is then a measure of the exposure that affected people are subjected to.

Aim
The aim of this empirical study is consequently to estimate the cost due to the value that people, living close to a transport route, place on a reduction in their exposure to dangerous goods. The study presented here will be focusing on transportation of dangerous goods by railway. A second part of the project is however planned that will estimate the cost applied on transportation of dangerous goods by road. This second part will be carried out during the spring.

Method
In this study, the conjoint analysis technique will be used to estimate the individual willingness to pay for a reduction in the exposure to dangerous goods. Conjoint methods were originally developed in marketing research in the early 1970s and have been widely used, especially in transport economics and environmental economics. In the conjoint analysis technique, peoples? statements of how they would respond to different hypothetical situations are studied. The analyst designs a set of hypothetical alternatives based on a limited set of important attributes and obtains from the respondent an indication of the relative preference for each alternative.

In the presented study, the respondent chooses between different transportation alternatives of dangerous goods on the railway nearby. The alternatives are described with the following four attributes: number of carriages per day carrying dangerous goods through the city centre, the period during the day that dangerous goods are transported, the degree of dangerousness of the goods being transported and altered housing costs per month. Preferences for scenarios are obtained by the use of discrete choices and each respondent makes six choices. Within each choice, the respondent chooses between two hypothetic transport alternatives and the situation of today. The chosen alternative is the one that leads to the higher level of utility.

The study is carried out as a postal survey to Lund and Borlänge in Sweden where 1000 respondents living close to the railway are randomly selected. These cities are characterised by rail traffic through the city centre with a relatively high proportion of dangerous goods being transported. Besides estimating the individual willingness to pay for a reduction in the exposure to dangerous goods, different hypothesis are studied. One hypothesis, that is tested, is whether the distance between peoples? home and the railway influence their willingness to pay for a reduction in the exposure to transportation of dangerous goods. In Lund, the respondents are therefore randomly selected on three different distances from the railroad, 400 persons living next to the railway, 100 persons living relatively nearby the railway, and 100 persons living outside the city centre.

Another hypothesis that is tested is whether the amount of information on dangerous goods that a respondent is receiving, affects his or her preferences. Under the assumption that people tend to overrate the risk for accidents involving dangerous goods to occur, we expect the value of a reduction in the exposure to dangerous goods to be higher for those respondents receiving little information on transportation of dangerous goods than for those respondents receiving extensive information. In order to test this hypothesis, 400 persons selected living next to the railway in Lund and in Borlänge receive a postal survey with little information and 400 receive a survey with extensive information.

The data is analysed using regression techniques. Since multiple observations are obtained from each individual, a random effect model is used.

Results
The result of the study indicates that people living close to a railway transport route are willing to pay for a reduction in their exposure to dangerous goods. A reduction in the number of carriages carrying dangerous goods and a reduction in the degree of dangerousness increase their utility. Their utility would also increase if transportation of dangerous goods were only allowed at daytime compared to the situation of today were dangerous goods are being transported all day. If transportation of dangerous goods were allowed at nighttime only, their utility would decrease instead. Hence, the affected individuals would have to be compensated in order to accept this detrimental effect. Information does have effect as well as distance to railway. There is also an impact of general socio-economic variables such as income, age, and gender.

Further analysis will give an indication on whether the cost due to the value of a reduction in affected peoples? exposure to dangerous goods on railways, exceeds the cost due to the value of a marginal risk reduction that is used in the original model of accidents involving dangerous goods on railways in Sweden.

Conclusion
When calculating costs and benefits in the appraisal of proposed accident prevention projects it is of great importance that the incremental costs are as far as possible estimated for the specific situation. The result for this study suggests that the presented method can be used to estimate the cost due to the value of a reduction in affected peoples? exposure to transportation of dangerous goods on railways. A second part of this project is planned which seeks to estimate the same cost applied on road traffic. Interesting comparisons can then be made.

The estimates of this project will improve the economic analysis model of accidents involving transportation of dangerous goods that is used in Sweden today. The model currently used has shortcomings. There is a great uncertainty in the value of a risk reduction (the so-called human value) since the figure used is calculated for general road accidents and not on accidents involving dangerous goods.

Publisher

Association for European Transport