In this paper, we describe the results from an ongoing project to update the modelling strategies of load models used for planning purposes by the Nordic Transmission System Operators (TSOs) Svenska kraftnät, Fingrid & Statnett. The background to this project and a description of the development of this methodology was published as Part 1 of this set of papers, presented at the Cigré 2016 Session [1]. During 2016, the methodology has been through a verification phase which is partly presented in [2] and [3] with further details presented in this paper.

The load response to voltage and frequency changes may have a significant impact on the dynamic behaviour of the power system. In this sense, the selection of load model structures and load model parameterisation gain an increased interest as power systems are operated closer to their limits. Modelling of loads is however highly complex due to the vast number of load devices in a power system, making it unfeasible to model each device separately as well as making it unfeasible to model each possible loading scenario. Based on this, a methodology for development of load models has been presented in [1]. In the work with validating this methodology, several challenges as well as strengths of the method have been identified. For the voltage dependency of the active power part of the load, the validation successfully provided evidence of the validity of the method. For reactive power, only a partial validation could be performed due to limited level of reactive power in gathered measurements. The load model structure used in this project is identified as a limiting factor for the representation of the non-linear behaviour of this part of the load. Frequency dependency of load has only been addressed to a limited extent, with results illustrating the difficulties to assess this kind of behaviour from measurements gathered during this project.

All in all, from the results of the validation it is found that the method is suitable to be employed on a larger scale with some differences in approach regarding the assessment of the voltage and frequency behaviour of the load. Furthermore, this work has provided valuable input for the understanding of the behaviour of the load.