Long asphalt tunnels serve as critical infrastructure for urban public transport, yet their construction entails substantial energy consumption and carbon emissions. This study aims to quantify and predict carbon emissions from asphalt pavement construction in long tunnels. Adopting Hybrid Life Cycle Assessment (HLCA), the asphalt paving process is divided into five stages: material production, transportation, on-site paving, milling, and ventilation. Four construction schemes are formulated via detailed calculation and analysis. A carbon emission factor database specific to long tunnel asphalt pavement construction in Guangxi is established, and a quantitative model is developed using the emission coefficient method to calculate carbon emissions of each scheme, among which Scheme 2 is determined as the optimal low-carbon construction scheme. The calculation model is validated via numerical software, and a carbon emission prediction dataset is constructed. Three prediction models, namely GA-BPNN, conventional SVR, and BPNN, are established. The results indicate that the GA-BPNN model achieves the highest predictive accuracy among the three models. This study further improves the calculation and prediction methods for carbon emissions in long tunnel asphalt pavement construction, providing theoretical support for carbon reduction and low-carbon construction management.
Yuan et al. (Mon,) studied this question.