Model-based estimation on building envelope infiltration

Abstract

Buildings consumed nearly 40% of total energy in the U.S. Air leaks through the building envelope are one of the factors increasing building energy consumption. The estimated energy use associated with infiltration loss through the building envelope within residential and commercial buildings in the United States for the year 2010 is 4 quads annually, which accounts for nearly 10% of the total energy use in buildings. The U.S Department of Energy published a building technologies program air leakage guide. In this air leakage guide, the DOE proposed five requirements for infiltration measurement method. However, two commonly used infiltration diagnostic approaches, blower door test and tracer gas method, are unable to meet the DOE requirement. Since existing infiltration diagnostic approaches do not meet with the DOE requirements, the building infiltration measurement is a challenge. To address the current challenges, a scalable and low-cost Building Infiltration Estimator with Ultrasonic Thermometry (BLAST) is proposed. The proposed method contains the physical measurements and a model-based estimation. This study is focusing on the model-based estimation part. To estimate building infiltration, a building envelope heat transfer model has to be developed. Recent study shows that a low-ordered three resistance-two capacitance (3R2C) thermal network model is sufficient to describe the building envelope heat transfer. A customized 3R2C thermal network is developed to represent building envelope heat transfer. Based on the 3R2C thermal network, the energy balance equation for building envelope has been applied to get the state-space model. The state-space differential equation is one of the key points to determine the suitable estimation method. This study uses an Extended Kalman filter (EKF) to inversely estimate the building infiltration using measurements of surface temperature and total heat flux, and a low-order state-space model. An EnergyPlus-based emulator is used to generate a virtual building and measurements to test the proposed estimation method. Nearly 80% estimated infiltration resistances are within the 20% error band compared to the calculated infiltration resistance from EnergyPlus. This preliminary study shows the EKF based estimator with the proposed measurements is promising for building infiltration estimation.