With the development of the Faulted Feeder Selection Device (FFSD) with the grounding current technology, lots of FFSDs with different principles have been used in the distribution automation system. Most researches focus on examining the actual performance of different FFSDs and then choosing the suitable ones. However, the actual operation performance of these FFSD is not satisfied in practice. On one hand, the test of on line performance is examined offline, so it is hard to find the overall stochastic fault scenarios of power distribution network. On the other hand a valid assessment methodology is absent without entire testing information, the potential drawbacks of distribution network terminal can not be predicted and removed. In order to solve these two problems, this paper firstly establishes a real distribution network model based on the PTDS, which provides various fault scenarios for the faulty line detection. In addition, by using the results of RTDS test from different faulty line detections, an integrated assessment method for the defect of FFSD based on Logistic regression model is put forward. Finally, the blind area of FFSD with different principles is introduced and assessed with the proposed method, which provides a significant reference values for the selection of accessing network equipment.