Impact of HWC DSM on voltage conditions and the choice of load parameters for feeder design

SUMMARY An extensive load research program, known as the NRS Load Research Project, was conducted between 1994 and 2005 in South Africa. It produced a large volume of domestic customer load data based on a 5-minute sampling period. From this data source statistically-based load models were derived, and are used in a probabilistic design procedure for the sizing of LV feeders. The statistical description used is a Beta probability density function. While this approach has received nation-wide approval in South Africa, it does not include the effects of peak-load reduction through the use of hotwater cylinder (HWC) switching. In this paper we investigate the impact of HWC DSM on the voltage performance of a LV feeder and the how it affects the choice of load parameters for design. It describes a procedure for using load data for a community with HWC, but without DSM, and then predicts the effect it would have on the characteristic loads once the HWCs are restored, subsequent to load-levelling.