Justifying the validity of safety assessment models with safety case patterns

Abstract

Safety analysis is an essential part of the development process of complex systems. However, decisions that are based on flawed safety assessment models, or models used beyond their envelope of validity can negatively impact safety design choices, the effectiveness of certification, and operational practice. Therefore, the justification of assumptions, data sources and analytical methods is necessary for appropriate use of these analysis results. Currently, most of the existing guidance on the evaluation or assessment of safety analysis is concerned with the human aspects of safety reviews. However, there are few recommendations as to how to justify a collection of safety assessment models as part of forming a coherent argument, especially for safety assessments performed using novel safety modelling techniques (such as Failure Logic Modelling). This paper examines the concerns for model validation activities in general and presents an exemplar safety case pattern for the adequacy of safely assessment models. The justification concerns of safely assessment models have been developed in order to provide inspiration and a starting point for future safety case developments utilising novel safety assessment models. (6 pages)