Article Preview
TopIntroduction
In recent years, the performance-based architectural design approach has become significant because of climate change impacts. The main causes of climate change are the increase in energy demand due to industrialization, the rapid growth of urban areas, and increased fossil fuel use (Mumovic, 2009). Thus, the design process needs to be energy efficient using analytical observations to apply better adaptation strategies against climate change. Although the building design and construction process involve many stakeholders, the most critical influence belongs to the architects. Although the designers are responsible for making important decisions regarding the building envelope and plan scheme (Granadeiro et al., 2013), many of them fall short of performance-based design and fail to grasp the process's advantages (Morbitzer, 2003). Due to the performance improvements requested or needed in architectural projects' proceeding stages, the design process's early stages are often returned. This situation causes losses of time and money (Hien et al., 2000). Figure 1 points out that the most significant influence on buildings' energy performance comes from the early design process (Attia et al., 2012). For instance, improvements in the orientation of opening and building envelope design can reduce the building's energy demand by 40% (Wang et al., 2005). For this reason, the importance given to the early design stages should be increased by doing technical reviews to understand better the relationship between design decisions and building performance, e.g., building energy simulations, statistical data review.
Figure 1.
Representation for the relationship between the effectiveness of the decisions and the lifetime of the building
The performance-based architectural design allows the designer to achieve better energy use and environmental performance. Because the designer can quantify and visualize the building performance. As an example methodology for performance-based design, genetic optimization algorithms are frequently used among designers (Kampf & Robinson, 2010; Konis et al., 2016). The optimization process leads to creating high-performance design solutions by automating the design process. In the process, the designer only edits the value ranges and constraints of the design variables. Nevertheless, the designer gains a limited understanding of the established design solution's reasons because of point-based performance estimation. Therefore, this approach falls short of meaningful evaluation of alternatives in early architectural design. An excellent solution to this problem can be found in statistical sampling methods to analyze the process's uncertainty and provide a powerful alternative to quantifying the relationship between design variables and performance outputs (Hemsath & Alagheband Bandhosseini, 2015).
This study uses the simulation-based methodology in a statistical analysis framework for an early architectural design feedback methodology by reaching quantitative results of building performance results. The building energy simulations are practical for more realistic decision-making (Morbitzer, 2003). However, energy simulations should be proactive for giving feedback about design variations (Attia et al., 2012; Kanters & Horvat, 2012; Rights, 2016). Unfortunately, most energy simulation tools cannot investigate the relationship between building and design variables in a proactive way (Y. Yildiz et al., 2012). Therefore, in this study, the authors suggest using statistical sensitivity analysis for proactive analysis. A systematic literature review was conducted of studies that statistical sensitivity analysis was proposed as an answer by measuring the output uncertainty and demonstrated the impact (i.e., effect) of the design variables (i.e., independent variable) on the performance outputs (i.e., dependent variable) for unbiased decision-making (De Wit & Augenbroe, 2002; O’Neill & Niu, 2017; Østergård et al., 2015; Ruiz Flores et al., 2012). Consequently, designers could have the possibility to reach effective design variable selection using performance output results.