Abstract
A considerable amount of the oil used in Saudi Arabia is consumed to generate power. In an effort to minimize such consumption, the government of Saudi Arabia has established an ambitious part of the 2030 Vision that includes 96 strategic objectives. In recent years, the advancement in information technology has made it possible to develop accurate prediction models of energy consumption for constructed facilities, which can assist with establishing plans to reduce future energy consumption. This research proposes a regression-based model for predicting the energy consumption of one of the most significant energy-consuming types of facilities in Saudi Arabia: schools. The model was developed utilizing 350 actual data points of energy consumption gathered from schools operating in the eastern province of Saudi Arabia. The factors affecting consumption of energy were identified from two sources. The first was a review of the literature, and the second was interviews with local experts. A sensitivity analysis indicated that the most important factors (i.e., inputs/independent variables) affecting energy consumption (i.e., output) were AC capacity and building age. An investigation of the correlations among the independent variables revealed that the highest correlation of 0.864 was found between the total built area and total roof area. The developed model was validated using 35 new data points of school buildings across the eastern province of Saudi Arabia. The results show that the model predicted the energy consumption of school buildings with an accuracy higher than 90%. The findings of this study will assist schools and maintenance managers in effectively managing such facilities by allowing them to allocate the required budget in advance. Also, the predictions can be used economic lifecycle analysis. (c) 2021 Elsevier B.V. All rights reserved.