Uncertainty characterization of building emergy analysis (BEmA)

Abstract

In this article, the authors quantify uncertainty of the building emergy analysis (BEmA). The emergy accounting considers environmental sources from the most fundamental context: the sun and the geobiosphere. However, due to the variety of large-scale parameters, BEmA suffers from lack of assurance in estimation. Accordingly, a source and quantity of uncertainty need to be clarified for its wide practical use. For this purpose, uncertainty of the major BEmA parameters, i.e., unit emergy values (UEVs) of the primary construction materials, life time of assemblies, and weight of the used materials, have been analyzed through a case study―a typical single-family house in Pennsylvania, United States. Data collected from all available literature was sampled to implement two uncertainty analysis methods: Monte Carlo simulation (MCS) and Fuzzy logic. The experiments were conducted with the following two ways: (1) we assumed that that empirical distributions of all the parameters fit into the normal distribution. The propagation of uncertainty was tested by MCS with hyper-cube sampling. (2) Fuzzy logic was employed to estimate distributions of actural values, because, in the reality, replacement intervals of building components and operational energy use tend to be subject to the building occupants’ contingent decisions or subjective judgments. In this case, data cannot be described by a steady-state normal distribution.The experimental findings showed that Monte Carlo simulation gave the mean of 9.57 × 1013 sej/m2yr with the standard deviation of 3.73 × 1013 sej/m2yr. Meanwhile, Monte Carlo simulation coupled with the Fuzzy logic reduced variability, thereby resulting in the mean of 8.42 × 1013 sej/m2yr and the standard deviation of 0.94 × 1013 sej/m2yr. Above two results were compared to each other through the KS (Kolmogorov–Smirnov) fitness test and the baseline of a point-estimate (7.66 × 1013 sej/m2yr). Sensitivity analysis as a post processing examined the ranked contribution level of each parameter. As a result, it became clear that specific emergy values of the large amount inputs such as electricity and concrete block are the most influential variables to BEmA.