Battery swapping mode analysis for shared e-bikes from a carbon perspective

Shared electric bicycles (e-bikes) primarily operate under two battery management modes: centralized charging (CC) and decentralized swapping (DS). However, systematic comparisons of these modes—particularly from a carbon emissions perspective—remain limited. Using Ningbo City, China, as a case study, this study assesses the carbon-emission reduction-potential of shared e-bike systems under CC and DS modes. By integrating life cycle assessment (LCA) with agent-based modeling, carbon emissions are quantified across the system's full life cycle. The results indicate that the DS mode achieves a 14-day net carbon balance of −1761.3 kg CO 2 -eq, whereas the CC mode results in net emissions of 509.0 kg CO 2 -eq. This difference is mainly attributable to lower transportation-related emissions during battery swapping, fewer deployed battery-swapping cabinets, and reduced unmet travel demand. Furthermore, the DS mode has been shown to improve battery-swapping efficiency in regions with high travel demand. Lowering the battery-swapping threshold reduces bicycle unavailability caused by low battery levels but increases the physical shortage of e-bikes. This trade-off is associated with an overall reduction in carbon emissions. The reduction amounts to 348.2 kg CO 2 -eq in the DS mode and 1645.9 kg CO 2 -eq in the CC mode, with the CC mode exhibiting greater sensitivity to threshold adjustments. Furthermore, under the CC scenarios, dynamically adjusting the battery-swapping threshold in response to fluctuations in travel demand reduces unserved trips and shortens battery-swapping intervals. • We propose a methodological framework consisting of an agent-based model and a life cycle assessment model. • The decentralized swapping mode achieves greater carbon emission reduction benefits compared to centralized charging. • A higher battery-swapping threshold increases charging-related unavailability while reducing the incidence of bicycle shortages. • In the centralized charging mode, demand-inverse temporal adjustments to the battery-swapping threshold are associated with improvements in operational performance.