Sample Size Planning in Experimental Research of Dielectric Stress in Oil-barrier Transformer Insulation

An approach for experimental research planning, based on Monte Carlo experiments, is suggested in this paper. The analysis is performed for AC permissible stress research in oil-barrier insulation system where paper insulated electrodes are used for insulation models production. The influence of sample size and high voltage test censoring on the accuracy of permissible stress estimation is examined as well as the possibility of using Monte Carlo experiments in costs and time-consumption optimization.