Risk and Safety in Engineering

Abstract

Risk assessment of engineered components and systems is addressed from the perspective of supporting engineering decision making on behalf of society. Both time invariant and time variant problems are considered. Specific outlines are provided on the treatment of structural reliability, assessment of existing structures, robustness, inspection and maintenance planning and decision making.

Objective

The aim of the present lecture is to provide the students with a thorough understanding of the role of risk assessment in the process of engineering decision making subject to uncertainties. Based on the course the students will be able to assess a given decision problem, formulate and verify engineering models, assess risks and optimize decisions. Specific knowledge is provided on the aspects of structural reliability, development of design basis, inspection and maintenances planning and assessment of existing structures. In these areas the students will be completely up to date with the present best practice.

Content

The ultimate task of the structural engineer, e.g. in connection with the design, assessment, maintenance and strengthening of structures can be seen as being "to identify the best solution" under the given constraints to safety, functionality, time and budget. However, many uncertain factors, such as inherent natural variability associated with the behaviour of loads and material characteristics together with incomplete knowledge about the considered problems, greatly complicate the decision making. Such problems may be treated within the framework of Bayesian decision theory, risk assessment and modern structural reliability theory. In the daily practice the standard codes for the design and assessment of structures provide sufficient guidance to the engineer in regard to normal decision situations. However, for situations and problems not covered by the design codes it is necessary to be able to analyse the effect of the prevailing uncertainties together with the potential consequences of the decisions in more detail for the case at hand. This is e.g. the case when structures made of new types of materials and structures which due to their size or geometry fall beyond the application domain of the codes. Moreover methods of structural reliability provide the only possible way to update the safety of existing structures based on tests and observations on the condition of the structures and their past performance. Finally decision theory and methods of structural reliability provide the basis for the calibration of modern design codes. In the course the notion of risk is explained and it is shown by examples to what degree different types of engineering activities are associated with risks. Thereafter a basic introduction to probability theory is given and the subject of probabilistic modelling in structural engineering is addressed in some detail. Techniques for the identification and analysis of hazards are provided including FMECA, HAZOP, Risk Screening, fault tree analysis, event tree analysis and decision/event tree analysis. Subsequently methods of probability estimation are explained including classical reliability analysis and modern time invariant and time variant reliability methods for components and systems. Thereafter more refined aspects such as Bayesian Probabilistic Nets which may greatly support risk assessment and decision analysis are introduced together with themes such as reliability updating of structures, planning of experiments, probabilistic fatigue crack growth modelling, risk based inspection planning and finally the issue of acceptable risks. The course is useful both for Dipl. Ing. students and Ph. D. students aiming to reach a higher degree of understanding in regard to the typical engineering decision problems.

Resources