Reliability analysis of substation automation system functions

This paper presents a case study applying a framework developed for the analysis of substation automation system function reliability. The analysis framework is based on Probabilistic Relational Models (PRMs) and includes the analysis of both primary equipment and the supporting information and communication (ICT) systems. Furthermore, the reliability analysis also considers the logical structure and its relation to the physical infrastructure. The system components that are composing the physical infrastructure are set with failure probabilities and depending of the logical structure the reliability of the studied functionality is evaluated. Software failures are also accounted for in the analysis. As part of the case study failure rates of modern digital control and protection relays were identified by studying failure logs from a Nordic power utility. According to the failure logs software counts for approximately 35% of causes of failures related to modern control and protection relays. The framework including failure probabilities is applied to a system for voltage control that consists of a voltage transformer with an on-load tap changer and a control system for controlling the tap. The result shows a 96% probability of successful operation over period of one year for the automatic voltage control. A concluding remark is that when analyzing substation automation system business functions it is important to reduce the modeling effort. The expressiveness of the presented modeling framework has shown somewhat cumbersome when modeling a single business function with a small number of components. Instead the analysis framework's full usefulness may expect to arise when a larger number of business functions are evaluated for a system with a high degree of dependency between the components in the physical infrastructure. The identification of accurate failure rates is also a limiting factor for the analysis and is something that is interesting for further work.