A prediction model for determining energy performance in existing school buildings

Buildings account for more than 30% of the sectoral energy consumption due to the development of building technology and changes in user conditions. Energy cost increases have accelerated efforts to make energy consumption more efficient. Considering the size of the existing building stock, it is understood that significant energy savings can be achieved through energy improvements in buildings. However, the diversity of existing buildings in terms of their intended use and construction period makes it difficult to make decisions in energy planning. Simple renovation or retrofit planning is a very effective method to improve the performance of existing buildings. It is known that the method's effectiveness can be ensured by correctly determining the parameters affecting energy consumption and optimum cost planning. Energy efficient retrofitting of public buildings within the existing building stock is important for social contribution. School buildings, which are included in the public building portfolio, have a special importance in the mission they undertake. Within the scope of the thesis, a building energy forecasting model is planned to be developed to guide studies for improving school building performance. The implementation of this model aims to create comfortable, sustainable, healthy spaces as a result of improving the energy efficiency of school buildings, as well as informing students about the gains to be achieved through improvement works and raising awareness of energy efficiency. Within the scope of the thesis consists of six chapters, it aims to determine the parameters that are effective in the process of reducing and improving energy consumption in schools and to create a guiding model based on benchmarking in school energy improvement. In the first stage, the definition of the problem, literature review, purpose, original value, scope, boundaries, and methodology of the study are included. In the second stage, decisions, regulations, and guidelines taken at the international level within the scope of increasing building performance as a result of energy-efficient improvements of existing buildings have been analyzed. In this context, legal regulations, strategies, and guidelines based on international conventions related to energy efficiency studies in Turkey have been evaluated. The role of buildings in reducing carbon emissions and the effectiveness of energy-efficient building practices in reducing emissions have been examined. As a result of this review, the principles for reducing the energy consumption of existing schools were evaluated, and the schools where energy improvements are made by these principles and their energy-saving potentials were analyzed. In the third stage, literature research is carried out to determine the conditions and factors affecting energy consumption in buildings, and the accepted parameters have been extracted as a result of the findings. Since there are many criteria for building energy consumption, a model that can simultaneously evaluate a large number of variable data has been investigated. Among the forecasting models, the multiple linear regression model is determined as the analysis method to analyze different variables. In the fourth stage, 38 public schools at different educational levels in the provincial center of Kırklareli are examined as the sampling area for evaluating the variable parameters that make up the estimation model. School and user data and natural gas consumption information have been collected from the relevant institutions. The data pool has been normalized by removing irregular schools in terms of construction, usage, and energy consumption data from the examined schools. School catalogs are created by extracting data and plan sketches of selected schools at different levels. In the fifth stage, the level of relationship between the generated parameters is evaluated using relationship analysis. The degree of efficiency of the variable parameters in the estimation model on the annual average natural gas consumption has been determined by multiple linear regression analysis based on benchmarking. As a result of the study, two estimation equations with acceptable R2 values have been obtained. In the last chapter of the thesis, the results of the study are presented, and recommendations are made to shed light on future studies. In this context, a useful decision-making model is presented for institution managers in improvement initiatives to increase energy efficiency and reduce carbon emissions in school buildings in Turkey.