| Managing the safety of engineering systems is an important role of regulatory agencies and standards organizations. One technique for safety management is the use of prescribed standards, which mandate certain design features. Another approach is performance standards, which provide performance requirements for certain system functions. While these techniques have been successfully employed for system safety management, they contribute to inconsistency and inability to evaluate the safety of existing and novel designs, new technologies, and complex systems. Many existing standards provide a statement allowing for safety equivalency, however, the process of assessing and managing risk for safety equivalency determination is not well defined. This research examines the applications of risk-based technologies to a suggested process for performing risk-based compliance approval. This process establishes consistency for risk compliance and safety equivalency decisions.; Risk-based technologies establish techniques that facilitate the proactive evaluation of system safety through risk assessment, risk management, and risk communication. Formal modeling techniques are defined which ultimately determine the probability of some system consequence through identification of possible scenarios for the consequence(s) of concern. Fault/success tree and event tree analyses are applied in this research for developing risk models. Both probabilistic and subjective uncertainty techniques are effective in uncertainty risk modeling. Fuzzy arithmetic and interval analysis are appropriate uncertainty techniques when data is subjective, vague, or cognitive. Risk management requires the application of decision analysis techniques that apply risk analysis results in making decisions. Calibration of codes and standards defines a risk acceptance process determined from existing standards. Risk communication facilitates the exchange of risk information between various stakeholders in the risk analysis process.; This research provides a case study for risk-based compliance of personal flotation devices. System modeling is performed by event and success tree modeling with simplifications through sensitivity analysis. Quantitative risk results for PFD models are demonstrated through probability, fuzzy, and interval numbers. The established risk results may be used for comparing PFD system designs to accepted risk values through a process of code calibration.; The methodology for risk analysis of engineering systems established in this research may be applied to numerous engineering systems to improve the assessment, management, and communication of risk. |
