Article Preview
TopIntroduction
The consumption of energy in the building sector has been an important part of the overall energy usage in the United States and on the global scale. According to the U.S. Energy Information Administration, residential and commercial buildings account for about 40% of the total U.S. energy consumption (EIA, 2020).
Identification and comparison of energy efficiency in buildings is based on utilizing a variety of metrics. For example, a very commonly used metric in the U.S. and globally for building energy performance is the energy use intensity (EUI), which is also applied for Energy Star Certification (Energy Star, 2018). This certification shows that comparing the energy use of buildings with others nationwide helps to effectively identify the opportunities for potential savings and best practices that can be replicated (Energy Star, 2018). Another common index for measuring energy efficiency is the energy efficiency index (EEI), also known as the building energy index. It is defined as the ratio of energy input (kWh) to a factor related to energy used, typically, building floor space (Kavousian et al., 2015). A variety of alternative indices for building energy performance has been also discussed in a paper by Goldstein and Eley (2014).
Measuring building energy performance usually takes into consideration a wide range of different factors, including energy usage, floor space, number of occupants, climate condition, energy efficiency of the equipment used, outdoor temperatures and others (Lee & Lee, 2009). There is also a number of methods applied to identify and compare energy efficiency in buildings. These methods can be summarized into four groups: simulation, regression analysis, neural networks, and data envelopment analysis (Ashuri et al., 2019; Wang et al., 2015; Lee & Lee, 2009; Lee, 2008). Each of these methods contains some benefits over other methods and, at the same time, limitations and drawbacks, a detailed description of which have been discussed in literature sources, e.g., Ashuri et al. (2019).
In this paper, the focus is on applying Data Envelopment Analysis (DEA) for measuring and comparing building energy efficiency. While the other three methods (simulation, regression, and neural networks) are being extensively used for data prediction and data analytics, the DEA method is the only one that is directly applied, and very successfully, to identify, compare, and optimize efficiency of similar units for an organization or between organizations. DEA allows to establish optimal efficiency targets and it does not rely on assumptions of relationships between input resources and output results. DEA is also a highly popular method for comparing and optimizing efficiency of similar units in various industries, for example, health care, hotels, and production units (Zakowska & Maciek, 2020; Nguyen & Nguyen, 2019; Liu and Zhang, 2017).
The DEA application in measuring and comparing building energy efficiency is still a developing area in business data analytics, and, therefore, is a subject of an ongoing discussion in academic literature. Molinos-Senante et al. (2016) consider several issues for DEA approaches. One of these issues focuses on architectural factors (compactness, building shape, etc.), and another – on energy management issues such as equipment efficiency and operations strategy. Some researchers recognize one more important DEA issue of applying a variety of DEA models and software in performing energy efficiency comparison for buildings (Iliyasu et al., 2015; Wang, 2017; Yoon & Park, 2017, Xu et al., 2020).