Base Load Estimation with Smart Meter Data

Research Highlights

- Data from conventional smart electricity meters allows to estimate the consumption of always-on and standby-appliances.

- By means of occupancy detection techniques, a load curve of always-on and standby-appliances is derived.

- A decent performance is achieved with the heuristics Load Duration Curve and Geometric Moving Average.

 

Challenge

Information on the consumption of always-on and standby-appliances would help to develop powerful energy saving campaigns and inform individuals about the benefits of more efficient devices. However, such information is difficult to extract from smart meters that are currently being rolled out to households as such meters only provide aggregated data at a resolution of 15-minute intervals.

 

Approach

We estimate the consumption of always-on and standby-appliances based on 15-minute data utilizing occupancy detection techniques. We implemented and tested six existing and three novel occupancy detection methods that only require conventional smart meter data, and we used the detected not-at-home periods to estimate base load curves. The base load curves are than being compared to calculations that rely on ground truth data for occupancy.

 

Results

The results indicate that unsupervised and threshold-based occupancy detection methods which require only smart meter data can detect absence of residents by an F1 score of 44.5 %. Our analysis shows that this performance is sufficient to estimate base load curves of households with a root means square error below 42 Wh.

Funding

This project has been funded in part by the ERA-Net SG+ (acronym: SmartLoad), the German Federal Ministry for Economic Affairs and Energy (Grant Number: 0350010), and the Swiss Federal Office of Energy (Grant Number: SI/501521-01)    

Team

Andreas Weigert, Konstantin Hopf, Thorsten Staake    

 


CONTACT US

For further information or if you have any questions please do not hesitate to contact us.