Integrating building energy simulation with a machine learning algorithm for evaluating indoor living walls’ impacts on cooling energy use in commercial buildings

Abstract

The environmental challenges of climate change increase the energy usage and peak demands of buildings. Most extant studies of greenery systems focus on exterior applications such as green façades and roofs, which indirectly affect the indoor environment. Few studies have focused on quantifying the influence of indoor greenery systems on building energy consumption. The cooling effects of indoor greenery systems such as living walls are largely accounted for by the evapotranspiration (ET) process, in which water is transferred to the ambient environment through the evaporation from and transpiration of plants. Current building energy simulation software such as EnergyPlus does not have a module for modeling indoor greenery systems.In this study, an ET model was created using a machine learning algorithm—Gaussian Mixture Regression (GMR) based on experimental data. An indoor living wall model quantifying sensible and latent loads from the ET process was integrated with the energy simulation software—EnergyPlus through the Python plugin feature. The U.S. Department of Energy (DOE) medium-sized office building reference model was modified and used in this study to evaluate indoor living walls’ impacts on cooling energy use. A parametric study on leaf to floor area ratios (LFAR), orientations, distances from windows, and climates was conducted to evaluate the influence of each factor on indoor living walls’ performance. Cooling effects of indoor living walls were evaluated in three ASHRAE climates with high cooling demands. Observable cooling energy savings were obtained for the south, east, and west perimeter zones while savings for the north perimeter zone was negligible in all three climates. With the consideration of extra electricity use from direct expansion (DX) dehumidification devices for humidity control, the maximum cooling electricity savings in Los Angeles, CA are 25.1 % when LFAR = 1.5 for the south perimeter zone, 14.4 % when LFAR = 0.5 for the east perimeter zone, 0.3 % when LFAR = 0.3 for the north perimeter zone, and 14.5 % when LFAR = 0.5 for the west perimeter zone on the design day.