Consumo de energia e conforto térmico em campus universitário: uma proposta de benchmarking

Abstract

Benchmarking is a tool used for comparing energy consumption between a group of similar buildings or the same building over time. It is applicable for increasing transparency regarding energy management and the development of policies for energy efficiency. However, in buildings with predominance of natural ventilation, as Brazilian Public Universities, the EUI evaluation may not be sufficient to indicate the efficiency of the buildings in use, given that there is the possibility of occurrence of thermal discomfort. Therefore, the goal of this thesis is to develop an internal benchmarking system of university campus buildings based on electricity consumption and thermal comfort. The buildings of the Federal University of Minas Gerais, in Belo Horizonte, were used as a case study. EUI was considered as the metric for energy consumption analysis. The Percentage Occupied of Hours in Comfort (POHC) and the Degrees of Heat (DHH) were used as metrics related to thermal comfort. A EUI database of 110 buildings was built and focus was directed to 27 buildings of Academic Units, for which architectural characteristics were also raised. The thermal comfort of naturally ventilated classroom environments was analyzed from the development of archetypes and parametric computer simulations of 55 thousand scenarios with the EnergyPlus software. The method of this work, of interdisciplinary character, included the elaboration of four predictive equations via multiple regression models: two equations for the prediction of EUI and two related to thermal comfort, being one for POHC and one for DHH. The equations for EUI were divided for buildings with or without laboratories with high density of installed power. The thermal comfort equations consider the classrooms. A benchmarking was proposed with the classification of buildings on the scale from A to E both for EUI and for thermal comfort. The benchmarking classification was integrated by a matrix which generated classifications in levels “Good”, “Medium”, “Poor” and “Very poor”. The developed equations presented R² from 0.93 to 0.97. It was found that 29% of the analyzed buildings are in the diagnostic zone "Good", 53% in "Medium" and 18% in "Poor". No Academic Unit was classified as “Very poor”. It is expected that this diagnosis can be used to support the prioritization of retrofit measures. Author considers that for buildings of hybrid conditioning or with natural ventilation it is essential to include the indication of thermal comfort in a benchmarking process. The contextualization of the case study suggests that the methodology developed can be replicated in other public universities in the country.