Various techniques of fault diagnostics in oil-filled electrical equipment have been used in many years and are well documented in literature. The common techniques include; IEEE key gas, IEC 60599 gas ratios, Rodgers ratios, Doernenburg ratios and the Duval triangle. Among all the available techniques, Duval triangle remains the most accurate to date. This paper presents a new fault diagnostic method which combines fuzzy trapezoidal membership functions derived empirically after analyzing faults documented in literature and the evidential reasoning techniques. Like in Duval triangle, three gases; methane, acetylene and ethylene are used. First these gases are normalized against their sum. The normalized values are transformed into fuzzy trapezoidal membership functions whose outcome are synthesized into a particular fault type by the evidential reasoning technique. One hundred and seventeen fault cases visually inspected by experts and documented in IEC TC 10 databases are used for verification. This new technique is easy to understand, uses simple standard mathematical expressions and can easily identify simultaneous faults. Unlike Duval triangle where two or more faults lying at the same spot are not easily recognizable, the new technique identifies faults based on numerical values hence it does not have this drawback.
No takes yet. Share an insight, caveat, or question.
Irungu et al. (2016) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: