Wednesday, 29 June 2016: 8:00 AM-9:30 AM
Smart Building Systems/Remote Monitoring and Diagnostics
Chair:
Kristen Cetin, Ph.D., P.E., Iowa State University
Technical Committee: 07.05 Smart Building Systems
The number of smart, connected technologies available and implemented in buildings has increased significantly in recent years, as have the number of types of devices and their capabilities to collect data on building performance, energy use and demand. This seminar presents opportunities to utilize these devices and data collected to more intelligently assess current building performance and more intelligently operate building systems. This seminar covers various advances in the collection and use of residential building energy and performance data for more smart assessment and operation of buildings.
1 Energy Use Insights from Inverse Thermodyanmic-Based Modeling of Residential Buildings
 |
Kristen Cetin, Ph.D., P.E., Iowa State University |
This presentation investigates the use of utility billing data for residential buildings, combined with highly granular weather data to disaggregate energy data into end uses, determine the type of HVAC system in use, and predict future months’ disaggregated energy use. This is accomplished through the use of an inverse thermodynamic-based model that uses binned temperature values. This methodology was verified using a dataset of several hundred homes. The resulting information can provide insights to residential building customers to motivate energy savings behaviors.
2 Demand Prediction Using Connected Thermostat Residential Building Energy Models
 |
Ratnesh Tiwari, Ph.D., University of Maryland |
A simple gray box model has been developed for residential buildings to predict the future HVAC energy use given the weather forecast. The model is based on thermostat data (temperature, setpoint, and HVAC on/off information) and the outside weather conditions (outdoor temperature, solar irradiation and wind speed). The accuracy of the future HVAC energy use for over 1000 houses across the United States predicted by the models was compared with the actual data. Predicted energy use from the model is within +/- 10% of the actual HVAC used. The temperature prediction error was within +/- 0.3C.
3 Cornell Temperature Datalogger Project
 |
Weili Xu, Carnegie Mellon University |
The Cornell Temperature Datalogger Project placed temperature dataloggers into 200 people’s homes for two weeks at a low-cost of $5. With 3-minute time scale, we observe temperature setbacks and the rate of thermal decay to reach those setbacks. Using this data, we are able to create a histogram of thermal decay and thermal setpoints for benchmarking the community. Compared to utility-meter-based approaches, temperature analysis more directly isolates shell performance. This helps identify community energy “hot spots” (80/20 rule) where shell retrofits would make the most impact. Correlations with building age, size, and heating fuel type will also be presented.