TEALD - The Tautological Energy AnaLog Dataset

[Stephen Makonin]

If you are interested… here is a 2-day sample of what will eventually be a new dataset:

http://teald.org/sample/labels.txt 

http://teald.org/sample/TEALD.tar.gz 

Each file is 24-hours of data. Each day has 2 files: 

 - mains_YYYY-MM-DD.csv

 - subs_YYYY-MM-DD.csv

The mains data is collected from a Rainforest EMU2 device that talks via ZigBee to the Iron “smart meter” at an interval of about once every 15 seconds.

The subs data is collected using a DENT PowerScout 24 at 1Hz over Modbus/USB-serial. The circuit number (1 to 24) is explained in the labels text file. Unlike AMPds all loads are monitored (with the exception of the breakers used by the DENT to power the meter and do V-sensing).

Voltage and frequency data in mains is actually from the DENT meter, I wanted to save some storage space. Voltage and frequency as the same for all DENT sub-meters in. Even though there is not L3 — I do not have 3-phase power — I included it as it shows the amount of power noise — L2 is tied to neutral and ideally should be zero.

I also plan to include security system (arm/disarm, motion, and door sensors), HVAC thermostat, and external weather (from Environment Canada).

24-hours data files would continually be uploaded to Harvard Dataverse. The dataset grow over time, but researchers would have access to new data almost right away (~24-hour delay).

I will probably put a draft paper up on arXiv.org at the end of this month.

Any questions or comments are welcome.

[Saeed Aghabozorgi]

Hi Stephan,

May I know which algorithm you have used in order to disaggregate this data? can you share the code with me.

[Stephen Makonin]

I have finally release this dataset but under a different name: RAE -- The Rainforest Automation Energy Dataset.

SUMMARY

A dataset that captures smart meter and sub-meter data. Houses are located in and around Vancouver, Canada. The Rainforest Automation Energy (RAE) dataset to help smart grid researchers test their algorithms which make use of smart meter data. RAE contains 72 days of 1Hz data from a residential house's mains and 24 sub-meters resulting in 6.2 million samples for each sub-meter. In addition to power data, environmental and sensor data from the house's thermostat is included. Sub-meter data includes heat pump and rental suite captures which is of interest to power utilities.

Read more about it: http://arxiv.org/abs/1705.05767

Get the data: http://dx.doi.org/10.7910/DVN/ZJW4LC

Let me know if you have any questions about it.

Best,

Stephen.