Abstract: To explore carbon emission reduction potential of straw pyrolysis and carbonization polygeneration engineering, and to support resource utilization of biomass waste and "dual carbon" goal realization, three straw pyrolysis and carbonization polygeneration projects with different processes located in Xingtai City and Bazhou City of Hebei Province, and Changzhi City of Shanxi Province, were selected as research objects．Based on Life Cycle Assessment（LCA）method coupled with emission factor method, a carbon footprint accounting model covering raw material transportation, straw processing, returning to field stages was constructed to quantitatively analyze greenhouse gas emissions and emission reduction effects of projects' whole lift cycle．Sensitivity analysis was also conducted to identify key influencing factors．Results has showed that all three projects achieved net carbon emission reduction throughout whole life cycles, with Changzhi City project demonstrating the highest net emission reduction of 1.07 t CO₂e per ton of raw material．Its advantage stemmed from synergistic effect of central heating with pyrolysis gas replacing coal combustion and carbon sequestration achieved through biochar returning to field．Electricity consumption during straw processing stage is primary carbon emission source of projects and a sensitive emission reduction link, while carbon emissions during transportation stage account for a low proportion and is a low-sensitive emission reduction area．Biochar returning to field is core carbon sequestration pathway．Sensitivity analysis has indicated that regulating pyrolysis process parameters, improving pyrolysis gas collection efficiency, and increasing biochar carbon yield can further enhance projects' emission reduction potential．This study has clarified carbon emission reduction value of straw pyrolysis and carbonization technology, providing data support and technical reference for low-carbon optimization and widespread promotion of such projects．It also points out that future research needs to expand accounting boundary and research area to improve conclusions' universality．