Study on prediction model of nitrogen emission in the production stage of residential building materials—A case study of Guangdong province

This study aims to reduce nitrogen emissions from residential buildings and establishes a prediction model for nitrogen emissions during the production of building materials. The calculation boundary, content, and method of nitrogen emission in the production stage of residential building materials are accurately analyzed. Based on the nitrogen emission data of 20 residential buildings in Guangdong, the composition and distribution characteristics of nitrogen emission in the production stage of building materials are analyzed. The coupling relationship between building design parameters and building materials’ nitrogen emissions is established using linear regression and ridge regression. 10 kinds of nitrogen emission prediction models based on the design parameters of residential buildings in the production stage of building materials were established and verified. The results show that the linear model M3, based on the number of floors above and below ground, the area width and depth, and the ridge regression model M5, based on the number of floors above and below ground, the area width and depth, and the total number of main functional rooms, have good fitting and prediction performance, respectively. The linear regression model M6, based on the number of floors above and below ground, the area width and depth, and the total number of rooms, has the best fitting and prediction performance. M3, M5, and M6 can accurately predict the nitrogen emission composition and distribution characteristics of building materials in residential building design and lay a foundation for future research on nitrogen emission evaluation and calculation methods and nitrogen emission reduction technology strategies.