Optimization of double-layer stress grading system for high voltage rotating electrical machines by electric field and thermal coupled analysis

Silicon carbide particle-loaded semi-conductive materials with electric field-dependent conductivity have been used for end-turn stress grading (SG) systems of high voltage rotating electrical machines for decades. Particularly, high voltage class turbogenerators use double-layer SG systems, at which two SG materials with different conductivities are applied, to reduce power dissipations within the SG materials. To prevent the excessive local heating of the SG material and resulting flashover, the field-dependent conductivities of the two SG materials, as well as length of the SG layer in the longitudinal direction along a coil, were optimized, by a nonlinear transient electrical field and thermal coupled analysis. The field-dependent conductivities of the SG materials were adjusted to the suitable ones by changing the average diameter of silicon carbide particles. Then, the appropriate length of the SG layer was selected to minimize the temperature rise in the system. As a result, the newly optimized double-layer SG system showed a 20% lower temperature rise, and more than 15% higher flashover voltage than the conventional ones. Therefore, the effectiveness of the optimized double-layer SG system is successfully confirmed.