Incremental Safety Assessment : Theory and Practice

The need for iterative safety assessment that starts at the earliest – conceptual – stages of the design has been long recognised by academia, industry and standardisation groups. Due to budgetary and time constraints such a process can rarely involve complete reassessment of the system at each iteration. Instead iterative assessment should build-up gradually, reutilising results of previous iterations wherever possible, and allowing engineers to focus on changes between iterations. A framework for tracking and managing changes between iterations is therefore a key aspect of incremental safety assessment. This paper explores the types of change that can occur between assessment iterations. We first illustrate a naïve view that assumes that each assessment iteration can only add new knowledge. However, assessment on a real project is likely to challenge, in addition to adding to, knowledge gained at previous iterations. In order to confidently make decisions about the adequacy of design, analysts and engineers will be required to consider the nature and extent to which the increasing design knowledge available on a project challenges the evolving safety assessment.