Abstract
In the Netherlands, land use planning around establishments handling, processing or storing dangerous substances is based on a Quantitative Risk Assessment (QRA). The criteria for acceptance of a risk level are laid down in law: no houses are allowed in areas where the individual risk of mortality of 1 × 10-6 per year is exceeded. Since the criteria are very strictly, the QRA results must be transparent and robust. Therefore, it was decided in 2006 to lay down the risk assessment method in a detailed, stringent guideline and to prescribe the use of the software tool SAFETI-NL. A helpdesk was installed to address open issues and a model management organisation was initiated for maintenance and development of the method.
As the guideline and software have now been in use for five years, this is a good time to evaluate experiences and to weigh options for the future. Various evaluation studies have been carried out recently to determine whether the application of the QRA in land use planning fulfils the expectations and requirements. Topics of these studies include: (i) How does the simplified risk assessment method relate to the complexity of process plants? (ii) What is a good balance between stability of the method and incorporation of new (scientific) developments? (iii) Can new emerging risks be incorporated in the risk assessment method? and (iv) To what extent have the goals and objectives of 2006 been reached? The results of the evaluation studies show that important objectives have been fulfilled The uniformity in risk calculations for land-use planning has increased substantially. Competent authorities can better assess the inputs and outputs of the risk analysis because one single guideline and software tool are in place. Additionally, there are side-effects, both positive and negative. The standard, structured approach has reduced the number of initiatives to make a site specific risk assessment. However, the remaining initiatives are justified in a more solid way. The easily accessible guideline and software tool has attracted a large new group of consultants to this market. This is good for competition but lack of experience may hinder the quality of the risk calculations. The largest remaining issue is the conflict between the desire for a stable method and the desire to incorporate new insights and new technologies. This issue requires further improvements to the model management organisation.
