Enhanced health index for power transformers diagnosis

Abstract

Power Transformers (PTs) failures in electrical networks cause significant expenses for power utilities, making the use of assessment techniques essential to efficiently diagnose and estimate the actual operating conditions of such equipment's. Efficient diagnoses allow operation management of this chain of assets aiming at the ideal balance among investments, maintenance costs and operation performance. Taking this into account, this paper presents an enhanced diagnostic methodology to estimate the PTs’ health based on simple and low-cost data which can be easily obtained without interrupting the unit's operation. The methodology uses data from oil sample analysis and acquisition systems of available electrical quantities and proposes the creation of a new diagnostic factor to identify the condition of the transformer's solid insulation. The new diagnostic factor is based on the history of the average daily load curve of a PT, allowing the estimation and accounting for the PT insulation degradation “in service”. Thus, the enhanced diagnostic methodology proposed in this work can estimate and update the transformerś health estimation in shorter time intervals than required by conventional Health Index (HI) methodologies and may be an important tool for strategic planning and resource optimization in power systems’ companies. The effectiveness of the proposed diagnostic methodology was assessed considering data from a population of 204 power transformers installed in Brazil and the direct comparison with the maintenance history from the conventional HI diagnostic methodology.