6.89. In particular for existing facilities, for which several feasible options for corrective actions are available, the comparison of corrective actions should usually be performed iteratively:@@- It may be possible to disregard some options for corrective actions very early on, e.g. because of prohibitive costs or because it soon becomes evident that basic regulatory requirements cannot be met. For such options, a detailed analysis of impacts would not make sense and would be a waste of effort.@@- The assessment of the implications of the remaining options for corrective actions with regard to the factors to be considered in the decision making can be very time consuming and resource consuming. The decision making may even face fundamental difficulties if a basis for determining precise estimates does not exist (e.g. with respect to the durability of structures). Instead of investing great efforts in trying to improve estimates for such factors, their relevance for the decisions to be made should first be examined. It may turn out that prevailing uncertainties in some factors will not influence a particular decision because it is dominated by other factors. If this is the case, the uncertainties can be accepted and further assessment efforts are not necessary in this respect. As a justification for the decision can be provided on the basis of assessment results, the uncertainties in these factors will not interfere with the overall requirement to build confidence in the assessment.@@- In accordance with the graded approach, the level of effort invested in improving data and the modelling should be commensurate with the importance of the various factors for the decisions to be taken. Within an iterative process, the implications of the results and their uncertainties for the decision making can be ‘tested’ to identify those aspects that warrant further refinement on the basis of their relevance for decision making.