Transformer T-joint optimization using particle swarm optimization and hemisphere-shape design of the core

• Main Author: Yehya, Omar Sharaf Al-Deen
• Format: Thesis
• Language: English
• Published: 2017
• Online Access: http://psasir.upm.edu.my/id/eprint/70172/1/FK%202017%20107%20-%20IR.pdf; http://psasir.upm.edu.my/id/eprint/70172/

Transformers are considered as a key in the transmission and distribution of electrical energy. The increases for electricity have encouraged the manufacturers to produce huge numbers of transformers for different sizes and ratings to work on the electric grid in order to meet market demands. The electrical power engineering and transformer experts seek to achieve the best economic and practical operation of electrical power system transformers, which include minimizing losses that are generated inside the transformers. The losses in transformers can be significantly reduced, especially in the core by improving the performance of the joint design. Several factors and parameters contribute to core losses such as shape of joint, gaps in between the joint parts, thickness of laminations, overlapping, orientation and number of laminations per stack. In this study, an intelligent algorithm was carried out using the particle swarm optimization technique (PSO) to propose the optimum design of T-joints for the core in three-phase distribution transformers. This technique was applied to design a new geometry of joint to get the minimum losses and to reduce the temperature in three-phase transformers. The smart algorithm proposed in this study presents the following advantages: (i) the correlation between the angles of the T-joint and gaps, (ii) the core loss profiles with temperature were considered, and (iii) the system was examined under different operational conditions. The transformer was simulated on the basis of real dimensions obtained from the transformer manufacture’s data. Furthermore, a 3D finite element analysis software model for transformer coupling with particle swarm optimization (PSO) technique was used and is validated by corresponding experiments. The simulation results have been validated with the manufacturer’s data of transformer rated at 1000 KVA. Practically, good agreements were obtained between the simulation results and the experimental data. The important parameters in the core joint design were emphasized through a comparison of the losses in various types of T-joint designs. The core losses, total losses were reduced for the new proposed model. The core and oil temperature underwent a good reduction as compared with the conventional T-joint designs. The core losses in the proposed design reduced more than 11% and 7% when using material M5 and M4 respectively. While more than 25% of the core loss reduction occurred when using material M6. The total owing cost for energy saving for different materials in the different T-joint designs indicated a life cycle, saving of RM 1297 for the M5 material and RM 1971 for the M6 material per transformer when compared to a conventional T-joint design of the same rating. Moreover, the proposed intelligent algorithm in this work can improve the transformer core design as well as can be applied to various power and distribution transformers.