Banning dangerous goods from a tunnel does not eliminate the risks, but modifies them and moves them to a different location, where the overall risk may actually be greater (diverting through a dense urban area for instance). For this reason, a joint OECD/PIARC research project recommended that decisions on authorisation/restriction of dangerous goods in a tunnel should be based on a comparison of various alternatives and should take into account the tunnel route as well as possible alternative routes.
A rational decision process was proposed, with the structure shown in the figure below. The first steps would produce objective risk indicators, based on quantitative risk analysis (QRA). The last steps would take into account economic and other data, as well as the political preferences of the decision maker (risk aversion for instance). These later steps could be based on a decision support model (DSM).
The OECD/PIARC project has developed a risk model as well as a DSM. The DG QRAM model is currently used in a number of countries. It is a system-based risk analysis model and produces indicators of the societal risk (F-N curves for the tunnel users and for the permanent neighbouring population), as well as the individual risk (for people permanently living in the neighbourhood of the tunnel) and damage to the tunnel and the environment. It is applicable both to routes including tunnels and to open-air routes, so that the risks on various alternative routes can be compared. The model is based on 13 scenarios representative of each of the five tunnel categories (although categories D and E cannot be distinguished because they lead to similar risks).
This model is currently being updated by PIARC in two steps, in order to adapt it to a modern hard and software environment and to enhance its abilities based on the comprehensive application experience available. It can be bought from PIARC and is described in more detail on its website. However, it has to be stressed that the procedures and algorithms of DG-QRAM were only developed for assessing risks due to dangerous goods. Hence, they are not appropriate to any other types of risk analysis.
In the course of the implementation of the ADR tunnel regulations, various countries have developed specific risk evaluation procedures for the transport of dangerous goods through road tunnels, in order to optimize the benefit and minimize the expenditure for studies. Typically this is a multistage risk assessment process starting with a simple parameter check, followed by a study of the intrinsic DG tunnel risk and ending with an investigation of alternative transport routes. The risk assessment study can be based on DG-QRAM, but in some countries, for instance in Germany, for some steps other methods are also used (for more information see technical report 2012R23 « Current practice for risk evaluation for road tunnels », appendix 2)
Additional information as well as examples of application can be found in the following PIARC references: