Selected examples of design optimization using an interactive work process and software tools

Risk assessment involves evaluating both the likelihood and consequences of system failures. An interactive risk assessment of a design requires that both the likelihood and consequences of system failures be quickly available to the design team. The consequence assessment focuses on the key failures and failure scenarios which are the input for the likelihood assessment process. While a screening evaluation of protection layers can provide general design guidance, often a more detailed evaluation of the safety and reliability of a design option can result in a design optimization. Probabilistic Risk Assessment (PRA) is an analytical method used to estimate the likelihood of failure of a system and to determine what the most likely contributors are to that failure once the consequence assessment has identified the key events for analysis. Space, nuclear, medical, chemical and defense industries are among those that have used PRA methods for assessing risks and/or reducing the costs in designing, upgrading, manufacturing, assembling, and operating components, systems, or facilities. When applied at an early stage of a project, PRA can be a valuable design tool. Current PRAs, however, are generally performed to demonstrate safety and are often unsuited for applications aimed at making design or operating decisions. Conventional risk assessment (PRA) tools allow integrated design teams and safety risk (PRA) analysts to make comparisons at design "freeze" points to address and mitigate the risk drivers. Unfortunately, the comparisons are made in isolation-that is, one at a time, and often the opportunities for some late stage design modifications are often limited. Using a work process that encourages an interactive design approach and using software that permits rapid evaluation of different design configurations allows for a real time quantitative evaluation of a design and rapid feedback to the design team. This paper focuses on the interactive work process and examples of actual applications with software tools that allow risk analysts and designers to easily determine the probabilistic implications of different design configurations and operating conditions in various combinations to reduce, control, or eliminate risk by quantitatively identifying risk drivers as the design develops and allow users to do an interactive evaluation of design changes. Because of the ease in evaluating alternate component or system arrangements, dramatic increases in reliability were observed with atypical, unusual, or simply different design configurations compared to the designs using "proven" reliability design practices. The lessons learned can be counterintuitive and significant. With ever shorter design timelines, the work process together with the software tools enable an interactive design optimization from a safety and reliability perspective and also consider the costs of different options. In the interest of being better, faster and less expensive, this is clearly a 21^st century analysis method.