An improved top-down algorithm combined with modularization as a highly efficient method for fault tree analysis

Methods available for fault tree analysis, based on obtaining minimal cut sets and their advantages, are briefly discussed. An algorithm which makes use of structure optimization, restricted pivotal decomposition for repeated events and modularization techniques is presented. If necessary, modules are broken down into cut sets employing an efficient top-down algorithm provided with a quantitative cut-off procedure, and a selective development sequence of gates. This evaluation scheme has been implemented in the computer code TREEMOD, a part of the program system RISA, capable of computing the dominant cut sets of very large and interconnected fault trees. (author)