A Fine-Grained Carbon Emission Accounting Framework for Manufacturing Systems through Model and Data Integration

With the increasing implementation of carbon markets and carbon tariffs across various sectors, accurately accounting for product carbon emissions from cradle to gate has become critically important. Existing carbon emission accounting methods and standards typically rely on industry-average data from life cycle assessment databases or on averages at the enterprise or workshop level. Such averaging approaches often result in low precision and are insufficient to support manufacturing systems in improving their carbon efficiency. Therefore, there is a pressing need for fine-grained carbon emission accounting, which can be enabled by information technologies that facilitate the collection of more detailed and specific data. Although numerous studies have proposed information technology-based carbon emission accounting methods, significant gaps remain in achieving truly fine-grained accounting. To address this, this study proposes a fine-grained carbon emission accounting framework for manufacturing systems based on the integration of models and data. The framework is developed through a comprehensive analysis of existing accounting methods and practical requirements, and consists of three levels: (1) inventory analysis and the separation of foreground and background data; (2) foreground data acquisition methods based on model-data integration; and (3) background data acquisition methods leveraging large language models. Each level of the framework is described in detail. A case study on an aluminum casting production line demonstrates the applicability of the proposed framework. The results indicate that the framework has strong potential for application across a wide range of manufacturing systems.