Details

Abstrakt in Englisch

Harmonic resonances are increasingly observed in central European residential low-voltage networks. These resonances are caused by increased usage of power-electronic-based equipment such as household appliances, electric vehicle chargers, and photovoltaic inverters. These resonances amplify the prevailing disturbance levels produced by the power-electronic-based equipment thus inducing a cause-effect loop. These equipment are connected and disconnected by the users at various time instances in a day resulting in a time-dependent cause and effect of harmonic resonances. This warrants a continuous detection and characterization of harmonic resonance in residential low-voltage networks. The resonance can be detected by both invasive and non-invasive approaches. The invasive approaches are accurate but are oftentimes costlier and bulkier solutions. Furthermore, as they inject a signal to detect resonances, they may disturb the nominal operation of the network and cannot be continuously used. The non-invasive approaches are more suited for continuous detection and characterization of resonance since they measure only harmonic voltages and currents and do not inject any disturbing signal. However, their typical disadvantages such as less reliability and longer measurement durations need to be addressed.

To develop such a non-invasive technique, simulation models that represent the time-dependent characteristics of harmonic resonance are developed. The simulation models were developed based on extensive field measurements carried out in various central European networks. The accuracy of the models in terms of their capability to represent harmonic resonance characteristics was also assessed based on data obtained from the campaign. Using the simulation models, the characteristics of harmonic resonance for various realistic scenarios were analyzed. Based on the results from this analysis, it was evident that a single measurement location–low-voltage busbar of supply transformer–is sufficient to comprehensively detect and characterize the resonance noninvasively.

Using the insights obtained from the analysis, in this work, a non-invasive technique for continuous detection and characterization of harmonic resonance in residential low-voltage networks is presented. The technique is formulated in three stages based on three indices each detecting a harmonic resonance characteristic. The indices are validated using the simulation model developed and presented in this work. The success rate of the detection and characterization technique is validated using short- and long-term measurement campaigns in residential low-voltage networks, which are also part of the application example of the proposed technique. Based on the results presented in this work, it was evident that the proposed technique is successful in detecting and characterizing their respective harmonic resonance characteristics continuously.