Modelling aggregate loads in power systems

Sammanfattning: The load response to voltage and frequency changeshas a considerable impact on the behaviour of the powersystem. Thus, the selection of a load model structure andits corresponding parameters is an important task in orderto study and predict the system behaviour. Currently,the Nordic Transmission System Operators (TSO) use theZIP load model, as it provides an easy and flexible way ofrepresenting the load. The main goal of the thesis has beento test two approaches for deriving ZIP model parameters,namely the component-based and measurement-basedapproaches. The former approach uses predefined parametervalues, and information on the loads electricityconsumption, whereas the latter uses measurement dataand curve-fitting techniques. In order to evaluate themethodology, a case study has been performed, wherethe two aggregation approaches were applied on anevaluation point. It was found that the aggregation bymeans of the component-based approach may result in ZIPparameters lacking physical significance. ZIP parameterswithout physical significance pose a challenge for systemplanners, who may have difficulties in accepting thesevalues as they are less intuitive than physically significantones. Furthermore, the results of the measurement-basedapproach indicate that the ZIP model has some limitationwhen it comes to the sudden load changes that it canaccommodate. This has been the case with the measuredreactive power in the case study. Based on the resultsof applying the methodology, it can be concluded thatthe component-based and measurement-based approachesprovide useful information when understanding powersystem loads.