Simulation-based fluid-thermal analysis of power transformers

Simulation-based fluid-thermal analysis of power transformers

Vol. 5 Issue 1

Abstract

The transformer life and performance strongly depend on winding hot-spot temperature (HST). Various alternative techniques for HST prediction are gaining popularity over the conventional direct-measurement methods. In this context, the application of Computational Fluid Dynamics (CFD) based thermal models is particularly interesting because of their accurate assessment, higher precision and low cost. Besides, it can remarkably evaluate and improve the design efficiency of transformer without overshooting the capital cost. In the present work, a comprehensive understanding of CFD-based fluid-thermal assessment is attempted to encourage the readers to review transformer thermal models. It is also expected that these attempts will progressively assist in correlating various economical and operational parameters of transformer manufacturing and asset management.

Keywords: oil-filled transformers, cooling, hot-spot, thermal assessment, Computational Fluid Dynamics


1.  Introduction

The economic reliability of power transformer changes dramatically with its thermal performance. Renewed interests in improving the heat-induced failure models have steered the manufacturers and researchers towards application of advanced thermal modelling techniques that use highly efficient numerical approximation and visually-aided simulation-based solution approach. Two widely acclaimed methods for transformer thermal assessment are Thermal-Hydraulic Network Modelling (THNM) and Computational Fluid Dynamics (CFD). Their sole objective is to predict the oil-winding temperature rise and improve the winding-HST displacement predictions without compromising the capital cost.

These methods are realistic, cost-effective, scientifically accurate, and efficient in improving transformer performance through real-time simulations of multiple interdependent physical phenomena, which ultimately affects the temperature rise and HST evaluation. The attractive aspects of such methods include reliable estimation of the magnitude of internal heat, effect of coolant oil pattern and transformer configuration, and most significantly, its solution technique. Figure 1 is a representation of key aspects in CFD-based thermal modelling of any transformer.

The upcoming sections of this article will assist the reader to gain a comprehensive understanding about CFD-based thermal modelling of an oil-filled transformer through an in-depth analysis of the concomitant multiphysics, application and limitations.