Development of tools for safety analysis of control software in advanced reactors

Software based control systems have gained a pervasive presence in a wide variety of applications, including nuclear power plant control and protection systems which are within the oversight and licensing responsibility of the US Nuclear Regulatory Commission. While the cost effectiveness and flexibility of software based plant process control is widely recognized, it is very difficult to achieve and prove high levels of demonstrated dependability and safety assurance for the functions performed by process control software, due to the very flexibility and potential complexity of the software itself. The development of tools to model, analyze and test software design and implementations in the context of the system that the software is designed to control can greatly assist the task of providing higher levels of assurance than those obtainable by software testing alone. This report presents and discusses the development of the Dynamic Flowgraph Methodology (DFM) and its application in the dependability and assurance analysis of software-based control systems. The features of the methodology and full-scale examples of application to both generic process and nuclear power plant control systems are presented and discussed in detail. The features of a workstation software tool developed to assist users in the application of DFM are also described.