Genetic Algorithm Applied to Multi-Criteria Selection of Thermal Insulation on Industrial Shed Roof

Abstract

The rational use of energy has motivated research on improving the energy efficiency of buildings, which are responsible for a large share of world consumption. A strategy to achieve this goal is the application of optimized thermal insulation on a building envelope to avoid thermal exchanges with the external environment, reducing the use of heating, ventilation and air-conditioning (HVAC) systems. In order to contribute to the best choice of insulation applied to an industrial shed roof, this study aims to provide an optimization tool to assist this process. Beyond the thermal comfort and cost of the insulation, some hygrothermic properties also have been analysed to obtain the best insulation option. To implement this optimization technique, several thermo-energetic simulations of an industrial shed were performed using the Domus software, applying 4 types of insulation material (polyurethane, expanded polystyrene, rockwool and glass wool) on the roof. Ten thicknesses ranging from 0.5 cm to 5 cm were considered, with the purpose of obtaining different thermal comfort indexes (PPD, predicted percentage dissatisfied). Posteriorly, the best insulation ranking has been obtained from the weights assigned to the parameters in the objective function, using the technique of the genetic algorithm (GA) applied to multi-criteria selection. The optimization results showed that polyurethane (PU) insulation, applied with a thickness of 1 cm was the best option for the roof, considering the building functional parameters, occupant metabolic activity, clothing insulation and climate conditions. On the other hand, when the Brazilian standard was utilized, rock wool (2 cm) was considered the best choice.