Both industrial experience and probabilistic risk assessments (PRAs) have shown that human errors determine, to a large extent, the level of risk associated with operating nuclear power plants (NPPs). While most current analyses have been focused upon a single operator acting alone, the fact is that most NPP-related decisions are made by teams of operators. The members of the team not only share information from different plant hardware, software, and human sources. but also use their training and knowledge base to contribute to the decision-making process through various forms of communication. Moreover, these group decisions are often made under time constraints and psychological pressures. As the new design of advanced lighter water reactors (e.g., AP600 and System 80 + ) has implemented digital software in their control mechanisms, the impact of these embedded systems on operator performance and the potential new failure modes that these digital software may introduce have not been fully studied and understood. There is an increasing need in the development of a reliability/safety analysis model that can evaluate errors of software, hardware, and humanware in an integrated manner, for separate analyses do not properly identify interface errors and issues. This proposal concerns the extension of the Dynamic Flowgraph Methodology (DFM), a prototyped software developed at ASCA for analyzing hardware and software systems that models cause-effect and timed relationship, to include humanware, i.e., human performance and team effects, so that DFM will be the integrated package for reliability and safety analysis.
Anticipated Results:The ultimate goal of this research is to extend the existing DFM methodology and software to include the human performance and team effects. With the additional module proposed in this Phase I SBIR project, DFM can be used to diagnose and to reduce system faults resulting from combinations of human errors, software logic errors, hardware failures, and environmental conditions. After such extension, DFM will be the first software tool with the capability of conducting reliability/safety assessments on hardware, software, and human ware as integral parts of a complex system.
