Detecting residential power peaks for innovative tariff designs: A living lab experiment

As smart meters become increasingly prevalent in private households, the availability of high-resolution electricity load profiles creates new opportunities for innovative electricity tariffs. Already established in industry, innovative tariffs could consider power consumption in order to monitor and maintain grid stability. This study explores the crucial role of time granularity in detecting power peaks, which are frequently missed when using coarse temporal intervals, such as 15 minutes. Through the analysis of high-resolution electricity load data from a living lab experiment, this research aims to determine the sampling rate necessary for accurately detecting power peaks. Our findings indicate that a sampling rate of 1 minute generally allows for effective peak identification, while certain cases necessitate even higher sampling rates, such as 10 to 5 seconds, to account for brief yet intense power peaks. This highlights the importance of employing a high sampling rate in the design of electricity tariffs that incorporate peak power consumption from households. Moreover, our study demonstrates that the temporal resolution of sampled data significantly impacts the accurate detection of temporary load unbalance, further emphasizing the novelty and significance of considering time granularity in smart meter data analysis. • Analysis of a detailed dataset (1 s resolution) on residential electricity consumption. • Open access publication of dataset (15 min granularity). • A smart meter sampling rate of 15 min is not sufficient for power peak billing. • Granularity of 5 s to 10 s detect peaks, but lead to a significant increase of data processing. • Lower resolution results in a decrease in accuracy of detecting temporary load unbalance between phases.