Optimization of the Cooling System in OD-Type Oil-Immersed Power Transformers
Year: 2026
Type: conference-paper
Abstract: The service life of power transformers depends on the thermal degradation of the insulation, a process accelerated by the hot-spot temperature. This study optimizes the design of directed oil flow (OD) cooling systems by reconfiguring the barriers position in disk windings to minimize hot-spot temperature without increasing material costs. The methodology, applicable to any type OD transformer, integrates a multiphysics simulation that couples electromagnetic (EMF) analysis with computational fluid dynamics (CFD) and conjugate heat transfer (CHT), formulating the problem as a mixed-integer nonlinear optimization (MINLP). This approach can detect complex phenomena such as reverse flows or stagnation; however, it is computationally very demanding. To balance physical accuracy with computation time, the model was simplified, and Bayesian optimization algorithms were employed. The transformer in the presented case study can operate at different power levels. Optimization is performed for two cases: the transformer operating at 16 MVA with natural convection (ON) and at 20 MVA with forced flux (OF). The results demonstrate that optimizing the geometric design of the barriers reduces the hot-spot temperature by approximately 3°C for ON, and 8°C for OF when compared to the original design. According to Montsinger's rule, this theoretical reduction could extend the insulation lifespan or allow an overload capacity without compromising the asset.
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7202380903