Thermal Evaluation of Power Transformers Based on Infrared Thermograms and Image Segmentation Using Fuzzy C-Means: A Case Study of Bangyoh Substation, Champasak Province, Laos PDR

Abstract

Condition-based maintenance of a power transformer is a vital process of conserving reliability of a power system at standard level. This research study assesses the thermal condition of a 25 MVA power transformer installed at Bangyoh substation of Electricite Du Laos by using a thermal image processing technique called thermograms. The procedure is started by thermal photographing of a selected power transformer by using a thermal camera. Consequently, a temperature distributed profile of a transformer tank, radiator and bushing can be conducted by dividing thermal zonals of a thermograms using a fuzzy c-means algorithm. Eventually, the thermal condition is evaluated. It was found that thermal condition of the case study transformer is at a satisfying level. The proposed procedures can be implemented as an on-line monitoring system, which can be used as a guiding framework for transformer maintenance planning in the southern area of Electricite Du Laos.