Implementation of optimization techniques for reducing the building energy consumption of a residential building

Abstract This study introduces the Taguchi based different optimization techniques, which aids in determining the optimal energy consumption for a typical residential unit. AutoCAD is used to model the residential unit and then conduct building energy simulations using HAP 4.90 software. Building orientation (A), wall heat transfer coefficient (U-value) (B), window-to-wall ratio (C), window glass U-value (D), glass shading coefficient (E), thermal insulant for roof (F), insulant thickness for roof (G), air leakage (H), cooling coil supply temperature (I), and bypass factor (J) are the ten factors that are chosen during the process of optimizing the air-conditioning (AC) load, each at three levels. An orthogonal L27 design is employed for experimental testing, with the cooling and heating loads for the residential unit calculated during each trial. Initially, an analysis of variance (ANOVA) analysis was performed using the Taguchi method for each response variable. Then, multi-objective optimization is conducted using two different optimization techniques, with the results being compared. The ANOVA method is also used to assess the contribution percentage of each factor to the cooling and heating loads and identify the main parameters affecting them. The findings reveal that the cellular concrete walls with a 10% window-to-wall ratio (WWR), triple-shaded glass (0.7 W m 2 -K −1 and 0.1 shading coefficient), and 75 mm Polyurethane foam (PUF) roof thermal insulant should be chosen as building materials, also, the air leakage rate should be kept at the lowest level of 0.5 ACH, with the wall U-value being the most influential factor for residential dwellings. Additionally, an AC equipment with a 287K CCST and 0.1 bypass factor should be chosen to reduce the cooling and heating loads. Prior research has rarely explored the parametric analysis of building factors on AC load, highlighting the novelty of this study.