Kegagalan transformator umumnya dibagi menjadi dua, yaitu kegagalan Thermal dan Electrical. Arcing, Partial Discharge dan Overheating pada sistem isolasi sering terjadi pada tranformator. Dekomposisi dari minyak isolasi dapat menyebabkan pembentukan gas yang berbahaya dan mudah terbakar yaitu Hydrogen, Methane, Ethane, Ethylene , Achetylene , Carbonmonoxide dan Carbon Dioxide. Dissolved Gas Analysis dari minyak merupakan cara efektif untuk memantau kondisi transformator. Metode penelitian ini menggunakan Total Dissolved Gas Consumtible (TDGC) dan TDGC_Rate sesuai standar IEEE C57.104, Roger’s Ratio dan Duval Triangle sesuai standar IEC 60599 dengan pendekatan Fuzzy Inference System menggunakan simulasi MATLAB/Simulink. Hasil interpretasi DGA dari Transformator Daya 200 MVA yaitu Continue Normal Operation (OPA) dan 6 Monthly (SIA) untuk sampling interval menurut TDGC dan TDGC_Rate, Discharge of Low Energy (D1) menurut Roger’s Ratio dan Discharge of Low Energy (D1) menurut Duval Triangle. Sedangkan hasil interpretasi DGA dari Unit Auxiliary Transformator 25 MVA yaitu Continue Normal Operation (OPA) dan 6 Monthly (SIA) untuk Sampling Interval menurut TDGC dan TDGC_Rate, Discharge of Low Energy (D1) menurut Roger’s Ratio dan Discharge of High Energy (D2) menurut Duval Triangle. Berdasar pada hasil penelitian menunjukan bahwa dengan menggunakan pendekatan Fuzzy Inference System penentuan tipe fault dapat lebih akurat dan dan lebih mudah.

Transformer failures are generally divided into two, namely Thermal and Electrical failures. Arcing, Partial Discharge and Overheating in the insulation system often occur in the transformer. The decomposition of insulating oil can lead to the formation of harmful and flammable gases namely Hydrogen, Methane, Ethane, Ethylene, Achetylene, Carbonmonoxide and Carbon Dioxide. Dissolved Gas Analysis of oil is an effective way to monitor the condition of the transformer. This research method uses Total Dissolved Gas Consumtible (TDGC) and TDGC_Rate according to IEEE C57.104 standard, Roger's Ratio and Duval Triangle according to IEC 60599 standard with a Fuzzy Inference System approach using MATLAB/Simulink simulation. The results of the DGA interpretation of the 200 MVA Power Transformer are Continue Normal Operation (OPA) and 6 Monthly (SIA) for sampling intervals according to TDGC and TDGC_Rate, Discharge of Low Energy (D1) according to Roger's Ratio and Discharge of Low Energy (D1) according to the Duval Triangle. Meanwhile, the results of the DGA interpretation of the 25 MVA Auxiliary Transformer Unit are Continue Normal Operation (OPA) and 6 Monthly (SIA) for Sampling Interval according to TDGC and TDGC_Rate, Discharge of Low Energy (D1) according to Roger's Ratio and Discharge of High Energy (D2) according to Duval Triangle. Based on the results of the study, it shows that by using the Fuzzy Inference System approach, the determination of fault types can be more accurate and easier.