A comprehensive study on thermal reinforcement of Saudi Arabia buildings considering CO2 emissions and capital cost using machine learning

Abstract

Saudi Arabia can be called a country with a desert and hot climate. Buildings consume more than 80% of the total electricity in this country. About 66% of the electricity in residential houses is wasted for cooling, which consequently emitted a significant CO2. In this study, using computational fluid dynamic methods, the thermal behavior of buildings equipped with insulation was compared with conventional buildings in Saudi Arabia to determine how much adding insulation increases the capital cost and how much diminishes CO2 emissions. For eight important areas (average temperature of 19–31 °C), insulation with a thickness of 1 to 10 cm was added to the envelopes in different positions, and the best installation location was explored in different indoor temperature scenarios using machine learning. There is a trade-off between the capital cost and operational CO2 emissions as any insulation thickness rises, increases the former, and reduces the latter parameters. Using artificial neural network and Genetic algorithm, it was found that if the least emission of CO2 is required, it is recommended to install insulation with a thickness of 10 cm in the outer layers, provided that the inside temperature is regulated at 27 °C. For Abha, the inside temperature should be set at 22–23 °C. The economic results revealed that each dollar investment in envelopes thermal reinforcement could diminish CO2 emissions by up to 29.9 kg. Regardless of the economic issue, adding insulation by 10 cm led to CO2 reduction by 66 kgm2.year