A Hybrid Data-Driven and SimulationBased Framework for Indoor Air Quality and Energy Modeling

Abstract

Natural ventilation can promote comfort, health, and productivity of occupants and, at the same time, reduce the operational energy use of buildings. This strategy, however, is usually applied at the cost of indoor air quality (IAQ) by unintentionally bringing in outdoor air pollutants. Hence, estimation of energy and indoor air quality (IAQ) is essential for the design of naturally ventilated buildings through reliable evaluation of trade-offs between energy-saving potential and human health risks. The primary objective of this paper is to propose a framework for integrated energy and IAQ which enable localizing outdoor conditions, including air pollution and airflows in building site resolution through a hybrid engineering simulation and an artificial intelligence-based approach. CFD, CONTAM, and EnergyPlus applications are used in this framework to calculate airflow, IAQ, and energy-saving potential of the building, respectively. The ML algorithms are also extended into CFD application to help facilitate localizing outdoor airflow, effectively. The results are validated against measured on-site observation data. This framework is tested on the DOE’s large-size office buildings located at Federal Campus in downtown Chicago. The outcomes of this research enable designers in an early stage to identify outdoor air determinants and localize them in building site scale, and analyze integrated energy-saving potential and IAQ in a reliable and efficient way.