• The most frequently observed failure mechanisms in the different components of SMs, analyzed on the basis of an in-depth review of the literature, with identification of their root causes and effects. • A review of diagnostic methods based on conventional approaches commonly used in industry, highlighting their effectiveness and limitations. • A critical analysis of the most relevant diagnostic and monitoring methods, including their advantages, limitations, and main applications. Synchronous machines (SMs) are critical components of power systems, playing a vital role in maintaining supply continuity and system stability. Despite their high reliability, the occurrence of defects remains inevitable and can compromise their performance. It is therefore essential to continuously monitor their condition to detect faults at an early stage to prevent major damage. Such damage not only leads to significant maintenance costs, but also causes unexpected downtime, resulting in high financial losses. This paper aims to identify the most common failure mechanisms affecting the main components of SMs by discussing their root causes and their impact on machine performance. It evaluates conventional diagnostic methods, highlighting their applicability and limitations, and critically examines relevant condition monitoring and fault diagnosis techniques, including signal-processing and artificial intelligence–based approaches. These methods are compared in terms of detection capability, fault localization, severity assessment and industrial deployment. The review also identifies key challenges, including sensitivity to operating conditions, data scarcity, and practical deployment constraints, while highlighting promising directions for improving predictive maintenance.
Belhaouzi et al. (Sun,) studied this question.