Lead Performer: ENERlite Consulting – Sacramento, CA
Partner(s): University of Louisiana-Lafayette – Lafayette, LA
DOE Total Funding: $200,000
FY21 DOE Funding: $200,000
Project Term: July 2021 – May 2022
Funding Type: STTR phase I
Project Objective
Retrofitting existing buildings requires models calibrated to current conditions that can serve as baselines for analysis and optimization. Calibrated models also have applications in building operations including fault detection and diagnostics (FDD) and advanced control.
White-box, i.e., physics-based, models take significant effort to develop and calibrate due to their large number of parameters. Grey-box models, which use simpler physical models and fewer parameters, should be easier to construct and calibrate and therefore to deploy at scale. ENERlite Consulting will develop a tool for calibrating grey-box models for retrofit and operation applications. The ENERlite tool will use an interactive approach that will allow the user to prune physically unrealistic—relative to actual building—parts of the search space. This approach should yield more accurate models as well as require significantly less computation than fully automated but “blind” calibration.
Project Impact
Calibrated models are currently not widely used in practice because of the cost of calibration and uncertainty about whether the model represents the physical building or is “getting the right answers but for the wrong reasons.” This project will allow facility managers and building operators to more easily create calibrated building energy models that can then be used to improve building operations and plan energy-efficiency upgrades.
Contacts
DOE Technology Manager: Amir Roth
Lead Performer: Ladan Ghobad, ENERlite Consulting