Improvement of Rogers Four Ratios and IEC Code Methods for Transformer Fault Diagnosis Based on Dissolved Gas Analysis Ibrahim B. M. Taha 1,3 , Sherif. S. M. Ghoneim 2,3 , Hatim G. Zaini 3 1 Electrical Engineering Department, Faculty of Engineering, Tanta University, Tanta, Egypt, i.taha@tu.edu.sa 2 Electrical Department, Faculty of Industrial Education, Suez University, Suez, Egypt, s.ghoneim@tu.edu.sa 3 Electrical Engineering Department, College of Engineering, Taif University, Taif, KSA, h.zaini@tu.edu.sa Abstract—Early detection of the generated dissolved gases due to different electrical and thermal stresses in transformer oil will prevent the malfunction of the transformer and hence maintain the continuity of the power system network operation. Dissolved gas analysis (DGA) is widely spread technique that is very useful in detecting the incipient faults in the oil filled power transformers. There are many interpretation techniques based on DGA such as Dornenburg method, Key-Gas method, IEC Standard Code, Duval triangle and Rogers methods. The diagnostic accuracy of these techniques is very limited and a significant instability is occurring with data uncertainty. In this paper a modification will be made on the IEC code and Rogers four ratios method based on the case study dataset (320 samples) to improve the accuracy of them. The data is collected from the central chemical laboratory of Egyptian electricity utility and from literatures. The results point out the accuracy improvement using the proposed form of IEC Code and Rogers four ratios methods. Keywords—power transformers, faults, dissolved gas analysis, IEC Code and modified Rogers NOMENCLATURE PD Partial discharge. NF No fault (normal). HED (D2) [High energy Arcing, Arc with high energy density, and Arcing discharge with high energy]. LTH (T1) Low temperature thermal overloading. MTH (T2) [Thermal fault (temperature <700 o C), and Thermal fault with temperature between 300 to 700 o C]. HTH (T3) Thermal fault (temperature >700 o C). LTH1 Thermal fault <150 o C. LTH2 Thermal fault (150 to 200 o C). LTH3 Thermal fault (200 to 300 o C). TFC Thermal fault (Increase in overall temperature in the conductive parts). TFCW Thermal fault (circulating current in windings). TFCT Thermal fault (circulating current between core and tank). LED (D1) Arc with low energy density, Arcing discharge with low energy, and Low energy discharge (sparking). CS Continuous spark]. PDT Partial Discharge with tracking. PDLED Partial discharge with low energy density. PDHED Partial discharge with high energy density. LTH4 Thermal fault with temperature between 150 to 300 o C. IEC International Electro-technical Commission. UD Undetermined fault I. INTRODUCTION The incipient fault deteriorates in power transformer will lead to serious damage in the transformer and a blackout of the power system will be occurred. Aging wave, rising energy consumption and liberation are factors that affect on the transformer condition and result in an increase in loading power transformer which causes an increase in thermal, electrical and mechanical stresses of transformers. Therefore, early stage detection of the transformer faults by predictive maintenance technique will prevent the malfunction of the power transformer. Dissolved Gas Analysis (DGA) in oil is considered a reliable and a wide spread method that is used to identify the incipient faults in oil-filled power transformers. Some electrical and thermal stresses will decompose the insulating transformer oil and several gases are released and dissolved in oil. These gases include hydrogen (H 2 ), methane (CH 4 ), ethane (C 2 H 6 ), ethylene (C 2 H 4 ), acetylene (C 2 H 2 ), carbon monoxide (CO) and carbon dioxide (CO 2 ) whose concentration is expressed in ppm. Some traditional techniques are commonly used for dissolved gas analysis to interpret the incipient transformer faults such as Dornenburg method, Rogers’ method, key gas method, Duval Triangle method, and IEC standard Code [1, 2, 3]. Many DGA interpretation techniques are based on gases ratios between the previously mentioned gases except Duval triangle method which defines the transformer faults based only on the percentage of CH 4 , C 2 H 4 and C 2 H 2 gases according to summation of them. All of these conventional methods are easy to implement, but they provides limited accuracy and stability with uncertainties of DGA data. Recently, great efforts are pushed to develop computational models based on artificial intelligence (AI) techniques to diagnose the incipient faults in transformers with high precision. These AI techniques are: Artificial Neural Network (ANN) [4], [5], Fuzzy Logic [6], [7], Support Vector Machine (SVM) [8], [9], Expert System [10], [11], Hybrid