Studies on Thermal Performance of Green Wall with Full-Scale Experimental House

A shortage of domestic energy production has always been a concern in Taiwan. A well-designed building may effectively alleviate this problem for the daily use of a building outcompetes all other aspects in energy consumption such as construction. Statistics has shown that about 40% of energy consumption in summer can be attributed to air conditioning. Therefore, it is very likely to reduce the total energy consumption if a building is well-designed for humans and the air conditioning can be less frequently operated. Building greenery, including green roofs and green walls, is one of the most effective ways to save energy from the excess use of air conditioning. Most of the recent studies focus on green roofs and green walls. However, of the green-wall studies, few experiment mentioned the effects from seasons, weather, day-and-night and coverage. Additionally, oversimplification of parameters by certain assumptions has usually been made in the studies of green-wall simulation. To assess the effects of environmental factors and coverage on green walls, a full scale study house is set with an evergreen plant (Asparagus sprengeri)-covered green wall for experiments; to build the model with physical based energy and water balance and validate it.. The results show a 27% increase of the Predicted Percentage of Dissatisfied in summer but a 5% decrease in winter in the house with the green wall used in day time. As a result, if the green walls adopt deciduous plants rather than the evergreens, the benefits of thermal comfort could be better. For the buildings used in night time, green walls show only about 5% increase or decrease of the Predicted Percentage of Dissatisfied, implying that, if there is short of funding, it is not necessary to set green wall for these buildings. The results of simulation show that the surface temperatures of foliage layer is the paremeter most close to the experiments, with the normalized root-mean-square error (NRMSE) only 7.3%; water content simulated is the most deviant, with the NRMSE up to 19.8%.