Seminar: Research into Building Energy Data Analytics
Thu, Mar 14
12:00 PM — 1:00 PM
Steinmann HallThe City College of New York160 Convent Avenue
Steinmann Hall, exhibit room
Speaker: Dr. T, Agami Reddy, Arizona State University, Tempe, AZSeveral electric utilities, as part of the Smart Grid Advanced Metering Infrastructure (AMI) have launched a multi-billion dollar program to install “smart” meters in several million residential and small commercial customers. This program would allow customers to have almost real-time access to their electric use data and a few hours delay on their gas use. The data would typically be in the form of whole building interval data (15 min aggregates for commercial customers) similar to the type of data already implemented for commercial customers with installed load greater than 200 kW. The question now is “What can the customer do with this data?” This lecture will provide a broad overview of the types of engineering analysis being developed, and focus on one example involving automated calibration of detailed building energy simulation programs.
Calibration involves tuning the numerous input variables of a detailed simulation program so that the output of the simulation closely matches the measured system performance. The benefits of such a calibrated tool would be that the tool could be used for a variety of applications such as continuous tracking and detection of excessive energy use, faults and degraded operation. Another potential benefit would be for controlled operations; for example, for optimal operation based on the time dependent tracking of the carbon footprint and operational measures to reduce the impact. Ongoing research to develop automated and robust calibration methods will be presented.
A second example relevant to automated design of buildings will be discussed. Designing buildings to be energy efficient can be described as a multi criteria problem whose complexity originates from the large number of variables involved, the dynamic nature of building loads and processes, and the intricacy of interaction effects among variables. As a result of this complexity, building design decisions based on a designer’s intuition and experience alone are typically inadequate to either achieve high performance results or encompass the large number of possible solutions. Optimization routines employing brute force search methods are able to handle a large number of constraints and variables, However, due to aesthetic issues, program restrictions, or specific owner requirements, often an optimal design is not desirable. Thus, rather than knowing a single “optimal” solution it is more valuable for the designer to know the “latitude” or “variability” they have in changing certain design variables while achieving the stipulated levels of energy performance. Thus the goal of the proposed methodology is directed towards providing such a decision support tool rather than a purely optimization tool.