Paper Conference

2018 Building Performance Analysis Conference and SimBuild co-organized by ASHRAE and IBPSA-USA


A Test Cell to Infer Thermal Response of Building Components for Model Predictive Control Using Building Automation Sensors

Elizabeth LeRiche, J. McArthur
Ryerson University, Toronto, Canada

Abstract: The use of Model Predictive Controllers (MPC) in building HVAC systems has demonstrated significant energy savings. This approach relies on knowledge of occupancy, weather conditions, and building thermal dynamics to predict the future HVAC conditioning requirements. The thermal dynamics is usually modeled using a grey box model, which requires the knowledge of the buildings thermal mass, an unknown characteristic for many older buildings. This study analyzes the ability to determine thermal dynamic behavior without building construction knowledge. A test cell has been constructed to monitor typical building control points consisting of the ambient temperature, supply air conditions, occupancy, lighting, and plug loads along with outdoor air temperature and solar radiation. These are used in conjunction with surface temperature sensors to characterize their thermal response with the outdoor and indoor conditions. This paper presents the setup of the test cell along with the preliminary study investigating the ability to solve for thermal mass using sensor measurements. The analysis conducted looks at the simplest thermal condition of the cell, the nighttime unoccupied condition, for the two most thermally slow elements, the concrete floor and ceiling slabs. The results from this preliminary analysis strongly indicate the potential to calculate thermal dynamic behavior using only BAS points. This will support the overall research goal of creating model predictive control approaches for existing building where exact thermal mass is unknown.
Pages: 815 - 822