Multi-physical Simulations of High Voltage Circuit Breakers

Sammanfattning: This report is the outcome of a 6-month Master’s Thesis carried out at General Electric Grid Solutions in Villeurbanne within the LTDT team in the Circuit Breaker and Bay Development (CB2D) department of the Apparatus Research Center (ARC). It presents both the innovative methodology adopted and the results obtained along the project. The Master’s Thesis deals with high voltage circuit breakers and is focused on one project, which will be called CB2 for confidentiality reasons. The aim of the project is to design a new circuit breaker, the CB2, based on an existing circuit breaker, with a cost reduction of 25% in order to adapt the device to the Indian market. To do so, numerical simulations on a software platform called MC3 were implemented. Simulations are indeed quicker and much cheaper than real tests. MC3 is a 2D axisymmetric solver, thus it is necessary to calculate 2D equivalent sections and volumes of the circuit breaker. The first step of the project is to calibrate the MC3 tool for the original device, called CB1. Before implementing the tests on MC3, a mechanical study is led in order to study both the wear and the velocities in these different tests. The conclusion of the study asserted that the velocities are lower than expected on almost all of the tests, however, the levels of wear are sometimes critical. The calibration –comparison of the overpressure in the thermal volume – is carried out based on one parameter, the FRAC (fraction of the radiative flux involved in the ablation process); each default and arcing time is thus tested with two FRAC values: 0.8 and 1.1. The first step was invalidated due to the high wear of the circuit breaker. FRAC 1.1 shows more accurate results than FRAC 0.8. However, the pressure differentials are lower in all MC3 calculations than in real tests, which may be due to the high wear and low velocities (reduced by 3 m/s) compared to the old device. Finally, some modifications of the CB1’s geometry are recommended based on different studies in order to improve the design of CB2. Those modifications will be tested on MC3 by the next interns at ARC.