Abstract

The central inquiry in many safety-critical systems is to assess the probability of all possible risk consequences that can occur in a system and its subsystems. In this research, we use theorem proving to formalize Event Trees (ET), Cause Consequence Diagrams (CCD) and Functional Block Diagrams (FBD), which are efficient techniques for probabilistic risk assessment at system and subsystem levels. Our approach provides the reasoning support with verified mathematical formulations that can analyze multi-level ETs, FBDs for complex systems, Cause Consequence Diagrams (CCD) based on Fault Trees (FT) as well as on Reliability Block Diagrams (RBD), as a novel approach. Also, the proposed formalizations of ETs/CCDs/FBDs allowed us to accurately determine of reliability indices, such as System/Customer Average Interruption Frequency and Duration (SAIFI, SAIDI and CAIDI) at system and subsystem levels. Moreover, we develop FBD and ET Modeling and Analysis (FETMA) software, which provides user-friendly features and graphical interfaces for industrial planners/designers. We applied our methods and tools on several realistic case studies from the power systems sector, i.e., the standard IEEE 3/39/118-bus electrical power generation/transmission/distribution networks, Quebec-New England High Voltage Direct Current (HVDC) transmission coupling system, multiple interconnected Micro-Grids, a nuclear power plant, transmission distance protection and a smart automated substation. Experimental results showed improvements compared to all existing reliability analysis methods in terms of scalability, expressiveness, accuracy and time.