Nowadays, Power Transmission Networks are operated through Supervisory Control and Data Acquisition (SCADA) systems. In incident situations, SCADA systems are prone to accumulate huge amounts of information potentially with temporal, non-monotonic and incompleteness problems. To support operator's decisions, on-line fault diagnosis is needed due to the critical nature of their work. In this paper a system -SPARSE II -is presented. SPARSE II is based on the cooperation of knowledge and model-based techniques. It has the ability to perform on-line fault diagnosis despite the information problems of having a SCADA system as an information source.