Duval Triangle-Based Dissolved Gas Analysis using GNU Octave for Transformer Fault Detection

Transformers are vital components in electrical power systems. However, they are also susceptible to various types of failures, including thermal and electrical faults caused by the formation of electromotive force, which, if left unaddressed, may result in degradation of the oil insulation. One effective approach to mitigate such issues is to conduct feasibility testing and oil analysis, commonly known as Dissolved Gas Analysis (DGA), which examines the condition of the insulating fluid within the transformer. In this study, gas concentration levels were identified as follows: C₂H₄ = 9 ppm, CH₄ = 4 ppm, and C₂H₂ = 11 ppm. These values were visualized using the Duval Triangle Method, an established technique for analyzing gas content by measuring the concentration of three primary gases: Methane (CH₄), Ethylene (C₂H₄), and Acetylene (C₂H₂), all of which dissolve in the transformer oil. The advantage of this method lies in its ability to serve as an early fault detection tool for transformer oil. The analysis results indicated an electrical fault categorized as a High Energy Discharge in zone D2, identified by a single plotted point where the three gas lines intersect on the triangle diagram. This type of discharge is predominantly associated with Acetylene gas (C₂H₂) and is typically triggered by intense internal arcing within the transformer. The interpretation was further implemented using an automated data plotting system in GNU Octave, serving as a Transformer Fault Detection tool and computational software that utilizes the C++ programming language for data processing and visualization.