During continuous operation of energy systems, the performance of components will mostly, gradually deviate away from the reference conditions due to performance degradation, which may eventually lead to malfunctions or operation failure. The complex interconnection among components and the propagation nature of additional irreversibility caused by malfunctions increase the difficulty of malfunction diagnosis. Particularly, in common real-world cases, multiple malfunctions usually happen simultaneously in several different components, imposing additional difficulty for effective malfunction identification and quantification. In this paper, we generalize an effective diagnosis method recently proposed by the authors to accurately locate the malfunction component and quantify the effect caused by anomalies of multiple malfunctions. The generalized method is based on advanced exergy analysis, where exergy destruction within each component is split into endogenous and exogenous parts. The endogenous exergy destruction is due to the irreversibility of the component itself, while the exogenous is caused by the inefficiencies of the remaining components. The exogenous exergy destruction is, in fact, the major obstacle to accurately pinpoint the origins of performance degradation. In the generalized approach, an internal exergy indicator is recommended to be applied first to identify the malfunction components in a fast and effective manner. Then the endogenous exergy destruction of the identified malfunction components under the reference and degradation conditions is calculated and compared for accurate quantification. The generalized diagnosis approach is applied to a complex real-world case studies, in which several malfunctions are introduced simultaneously into different components. The results show that the proposed indicator could fast identify the source of anomalies while the endogenous exergy destruction successfully and effectively quantifies all introduced malfunctions. (c) 2017 Elsevier Ltd. All rights reserved.