Disaggregation of Residential Smart Power Meter Data Using Localized Weather Observations

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 5 January 2015: 5:00 PM
224B (Phoenix Convention Center - West and North Buildings)
Michael Siemann, Earth Networks, Germantown, MD; and C. Sloop

Over the past couple of years Earth Networks has been successfully leveraging their WeatherBug weather observation network to provide energy applications to residential and utility customers. Innovative thermodynamic energy models, co-developed with The University of Maryland, have given customers a great deal of insight into their residential energy consumption. In particular, this presentation will focus on using weather (temperature, relative humidity, wind speed, and solar power) and smart electrical power meter data to build those models and disaggregate total residential loads into Base (appliances that are always on or plugged in), HVAC (heat and cooling systems utilized to maintain comfort), and Variable (non-periodic appliances such as entertainment, food processing, and lighting) sub loads.

The model used for this disaggregation correlates an overall heat transfer from the outdoor to indoor environment to past power data. The primary heat transfer energy flows involved are conduction through the walls and roof, solar loading on the walls, roof, and through the windows, infiltration, internal heat generation, energy stored in the thermal mass of the house, and energy to condense moist air. The results of the disaggregation are offered to customers as part of WeatherBug Home's ScoreCard along with a targeted tip specific to the largest inefficiency identified. Results of this technique will compared to disaggregated data in this presentation as well.