**8. Conclusions**

This chapter investigated the reliability of the data used for the power quality disturbances assessment. The main applications were to estimate the network harmonic impedance and to determine the interharmonic source. A set of criteria to state about the data reliability was presented. They consisted in proposing thresholds for the following parameters:


For the network impedance estimation application, it has been found that the energy level for *ΔI(f)* is a useful data filtering, but for *ΔV(f)* it has been found that it does not really make any difference. The Coherence index does not reveal much information about unreliable measurements but clearly identifies the principal frequency components of the transient. Analyses carried out on the quantization error level demonstrated that quantization noise is substantial for high frequencies and that the measurements taken are not significantly affected by quantization noises. The suggested thresholds for data rejection used were determined through extensive experience with handling data and provided more accurate and dependable results. These thresholds can be further adjusted as new data are analyzed and experience is built to improve the engineering judgment.

For the interharmonic source detection, the power direction method is very sensitive to the typical low energy level of the interharmonic currents. It was observed that this low energy level affects the displacement angle between voltage and current, which may prevent using the method to conclude about some frequencies. Interharmonic pairing theory was used to draw a final conclusion for the smaller-magnitude interharmonics in the case study. For the higher magnitude interharmonics, the power direction method could be used with the confidence provided by the reliability criteria.
