Virtual Smart Meter Data Set

Authors: Adam Neale, Michaël Kummert, Michel Bernier
Polytechnique Montréal

Data set description

Existing electricity smart meter data sets lack sufficient details on building parameters to evaluate the impact that home characteristics can have on electricity consumption. An extensive, open-source virtual smart meter (VSM) data set with corresponding building characteristics is provided. The methodology used to develop the VSM data is presented in detail in an accompanying publication.

The data set consists of a variety of homes representative of a subset of the Canadian single-family home building stock generated using building energy simulations. The building characteristics cover a wide range of values that are based on probability distributions developed using a segmentation and characterization process. The resulting model and VSM data set can be used by researchers to develop classification models, verify load disaggregation algorithms, and for a variety of other purposes.

This data set contains:
1) 200 000 single-family homes with associated physical characteristics and load profiles
2) Electricity consumption values for each building at 15-minute intervals (200 000 profiles with 35 040 data each)
3) Annual breakdown of heating, cooling, lighting, equipment and domestic hot water electricity consumption for each building.

Version 1 data

Follow this link to VSM Data set (17 GB file size): VSM_dataset_v1.zip

Citation

Please cite this dataset as:
Neale, Adam, Michaël Kummert, Michel Bernier. 2020. "Development of a stochastic virtual smart meter data set for a residential building stock - methodology and sample data." Journal of Building Performance Simulation, Vol 13, No. 5: 583-605. doi:10.1080/19401493.2020.1800096