A comparison of the simplified probabilistic method in R6 with the partial safety factor approach

Abstract

Probabilistic assessments are a useful aid to decision making in areas such as safety analysis, design studies and the deployment of resources on maintenance, inspection and repair. In principle, a full probabilistic assessment requires a complete knowledge of the relevant failure models and the distributions for each of the input quantities. However, in practice, these requirements cannot normally be met in full and it is necessary to employ various simplifying assumptions and approximations in order to make the analysis tractable. The partial factor method and the simplified approach in R6 provide two relatively simple and independent methods of assessing failure probabilities using R6. The two methods have been applied to a set of test cases and the results compared. In the case of the partial safety factor method target reliabilities in the range 10 −3–10 −5 were considered. Sets of partial safety factors for load, defect size, fracture toughness and yield stress were taken from BS 7910 and used for assessments covering different regions on the R6 failure assessment diagram. A calculation of the assessed failure probability was also carried out for each of these sets of conditions using a simplified probabilistic approach developed for the R6 procedure. The assessed failure probabilities were compared with the corresponding target reliability assumed for the partial safety factor calculation. It was found that the partial safety factor assessments were generally conservative compared to the simplified approach. However, in many instances the assessed probabilities were several orders of magnitude smaller than the target reliabilities suggesting that the recommended values of partial safety factors in BS7910 were excessively conservative for some of these conditions.