To effectively combat environmental challenges, it is necessary to evaluate urban residential building carbon emissions and implement energy-efficient, emission-reducing strategies. The lack of a specialized carbon emission monitoring system complicates merging macro- and micro-level analyses to forecast urban residential emissions accurately. This study employs a system dynamics (SD) model to examine the influence of social, economic, energy, and environmental factors on carbon emissions in urban residences in Kunming, China. The SD model forecasts household carbon emissions from 2022 to 2030 and establishes three scenarios: a low-carbon scenario (LCS), a medium low-carbon scenario (MLCS), and a high low-carbon scenario (HLCS) to assess emission reduction potentials. It predicts emissions will climb to 4.108 million tons by 2030, significantly surpassing the 2014 baseline, with economic growth, urbanization, residential energy consumption, and housing investment as key drivers. To curb emissions, the study suggests enhancing low-carbon awareness, altering energy sources, promoting research and development investment, and expanding green areas. The scenarios indicate a 5.1% to 16.1% emission reduction by 2030 compared to the baseline. The study recommends an 8.3% to 11.4% reduction in MLCS as a practical short-term target for managing urban residential emissions, offering a valuable SD approach for optimizing carbon strategies and aiding low-carbon development.