Proposal for an Improvement in Transformer Diagnosis by Dga Using Expert System

Power transformer is a major component of power system and it’s correct functioning is vital to system operations. It is quite expensive and has no substitute for it’s major role. It is therefore, very important to closely monitor the in-service behavior to avoid costly outages and loss of production. Many devices such as Buchholz relay and Differential relay have been evolved to monitor serviceability of power transformer. However, they respond only to severe power failure requiring immediate removal of transformer from service. Thus, preventive techniques for early detection of faults to avoid costly outages would be valuable. Of all the techniques used for assessing the health of oil filled transformer, none can match the serviceability and reliability of Dissolved Gas Analysis (DGA). Many interpretative methods such as Rogers, Dornerburg and IEC PU 599 based on DGA to diagnose the nature of incipient deterioration have been reported. One drawback of all these methods in its present form is that a significant number of DGA results in service fall outside the proposed codes and cannot be diagnosed by these methods. To overcome this limitation, a prototype of an expert system based on DGA for diagnosis of suspected transformer fault and their maintenance action is proposed in this paper. An attempt is also made to fill the void by expansion of ratio combinations in the ratio method IEC PU 599 so that all practical cases can be dealt with. This expert system is implemented into PC by using ‘ Turbo Prolog’ with rule based knowledge representations. The designed expert system has been tested for about ten years collection of CPRI, BHEL and NTPC (India ) transformer gas ratio records to show its effectiveness in transformer diagnosis.