Prediction on electrical insulation synergistic effect of gas mixture based on dimer microscopic descriptors

Buffer gas is often mixed with insulation gas to reduce the greenhouse gas consumption. Mixing of different gases could lead to the synergistic effect of electrical insulation strength. In a previous effort, breakdown experiments were used for buffer gas selection. However, the experiment was expensive and time consuming, so there is a necessity to investigate the microscopic descriptor that influences the insulation strength synergistic effect. Thus, six kinds of gas mixture were studied in this paper, namely, SF6/CF4, SF6/N2, C4F7N/CO2, C4F7N/N2, c-C4F8/CO2, and c-C4F8/N2. The molecular and dimer of these gas mixtures were analyzed by the B3LYP-D3 method and 6-311G(d, p) basis set. Finally, a prediction model for the synergetic effect coefficient was established based on the Gibbs free energy change from dimer formation, the volume of the dimer, the maximum of electrostatic potential, the positive average potential, and the positive potential variance. The synergetic effect coefficient calculated by the prediction model is consistent with the experiment result, with a root-mean-square error of 3.27 × 10−4, a minimum error of 1.42%, a maximum error of 17.30%, and an average error of 6.85%.