Probabilistic approach for the determination of connection power using smart meter data for urban households with decentralized generation and storage units

Abstract:
In this paper the conventionally used method to calculate the peak connection power for residential households is compared with a probabilistic approach. This approach uses statistically analysed smart meter data from residential households combined with decentralized generation units, electric storage as well as thermal storage units and the electric mobility expected in the near future for a future-proof dimensioning of electrical grid equipment (e.g. transformers and low voltage lines). The statistical analysis performed is inspired by the principal approach of statistical testing procedures (goodness of fit) and adapted for this bottom-up methodology to generate probabilistic load profiles for the thermal storage unit based on measured smart meter data of residential households. The output of this methodology are probabilistic load profiles based on the behaviour of residential households to dimension electrical grid components in terms of possible peak power quantiles.