For the resonant grounding system of the distribution network with the neutral point grounded through the arc suppression coil, there are problems such as inconspicuous fault current characteristics, difficulty in threshold setting and low accuracy of fault line selection when a single-phase grounding fault occurs. In this research, a threshold-free fault line selection method based on Complementary Ensemble Empirical Mode Decomposition (Complementary Ensemble Empirical Mode Decomposition, CEEMD) and sample entropy is proposed. By analyzing the transient characteristics of the zero-sequence current of the fault line, the high-frequency components of the oscillation and attenuation characteristics are selected as the observation objects. A fast extraction method and fault characterization strategy of high-frequency components based on CEEMD algorithm are proposed, and the reliability of sample entropy based on high-frequency components as a threshold-free line selection criterion is explored. The resonant grounding system model under disturbance under multiple scenarios is simulated and established, and the feasibility of the proposed algorithm in engineering scenarios such as different line positions, fault closing angle, grounding resistance and arc suppression coil compensation degree is verified. Among them, the proposed algorithm can realize fast fault line selection that only collects half-cycle current. In addition, the method does not need to set a threshold value, and is not affected by different operating modes of the system, which is easy for engineering application of multi-scenario fault line selection.