Harmonic Modeling, Data Generation, and Analysis of Power Electronics-Interfaced Residential Loads

TL;DR

This paper develops a method to generate harmonic-rich data for residential loads, enabling accurate modeling of harmonic interactions in power networks with electronics-based appliances and distributed energy resources.

Contribution

It introduces an impedance-varying approach to produce harmonic data and demonstrates its effectiveness in identifying harmonic load models from limited data.

Findings

Harmonic-rich datasets improve load model accuracy.

The proposed method captures harmonic propagation and interactions.

Limited data availability impacts model identification accuracy.

Abstract

Integration of electronics-based residential appliances and distributed energy resources in homes is expected to rise with grid decarbonization. These devices may introduce significant harmonics into power networks that need to be closely studied in order to accurately model and forecast load. However, it can be difficult to obtain harmonic-rich voltage and current data -- necessary for identifying accurate load models -- for residential electrical loads. Recognizing this need, first a set of electronics-based end-use loads is identified and modeled in an electromagnetic transients program tool for a residence. Second, an impedance-varying method is proposed to generate harmonic data that captures harmonic propagation to the supply voltage and harmonic interactions among end-use loads connected to the same supply voltage. Third, a harmonic-rich dataset produced via the proposed methodology is demonstrated to successfully identify frequency coupling matrix-based harmonic load models using the least-squares method. Numerical results demonstrate the accuracy of the model. The impact of limited data availability on model identification is also explored.