Sensitivity analysis of fault impedance measurement algorithm used in protection of three-terminal lines

Abstract

This paper presents a methodology for systematic sensitivity study of fault impedance measurement algorithms applied in protection of three-terminal lines. There is a number of uncertain parameters (i.e. factors) associated with three-terminal lines that will impact performance of the algorithms and overall performance of Intelligent Electronic Devices (IEDs). The proposed methodology is variance-based Global Sensitivity Analysis (GSA) technique that uses Quasi-Monte Carlo (QMC) for computation. The screening technique is also implemented to reduce the factor space dimension and speed-up the QMC computation. Three-terminal line during various fault conditions is simulated using DIgSILENT PowerFactory. The automation program is developed in DIgSILENT Programming Language (DPL) to automate the fault simulations for various points (i.e. samples) in the factors space. For each sample, the fault impedance algorithm has been tested by injecting the simulated voltage and current signals. For large number of samples, output variance as well as portions of this variance (i.e. sensitivities) contributed by each factor and their interactions are calculated using SIMLAB. This technique is applied in practical testing of fault impedance measurement algorithm implemented in SEL-421 protective relay.