A novel framework to establish optimal failure probabilities for engineering structures

The reliability of engineering structures in the built environment is of crucial importance for society. However, improving structural safety by building more robust structures also increases their monetary cost and environmental burden. This represents a trade-off between different aspects of sustainability, which is not fully considered in current literature and building codes such as ISO2394 and EN1990. To address this, a novel framework is proposed to establish target reliabilities for structures, taking broad sustainability into account. This paper focuses on finding an acceptable failure probability value based on reaching a sufficient relative investment in life safety improvement. In current literature, this investment is quantified using a function for the lifetime monetary costs of structural safety, discounted to their net present value, without consideration of environmental effects. As part of the new framework, this paper proposes taking harmful emissions into account through a shadow-costing scheme, thereby establishing an augmented cost function and accompanying criterion for sufficient safety investment. This proposed approach is tested on a typical structural design case study, where the augmented and conventional approaches have been compared. The augmented approach yields an acceptable failure probability around 50 % larger than the conventional approach, though the difference is sensitive to the utilised shadow price of emissions. In general, the acceptable reliability of a structure is also sensitive to the number of potential fatalities occurring at structural failure, and the applied discount rate.

Keywords: Structural Safety, Sustainability, Optimal reliability, Acceptable reliability, Differential discounting.