Abstract: Focusing on urban intersections, this research examines traffic operational conditions and new energy vehicle (NEV) penetration. A traffic demand model and microscopic simulation techniques were used to analyze mechanisms affecting intersection carbon emissions. Results show emissions depend on conventional vehicle flow and their per-vehicle emission rate, the latter directly linked to traffic conditions. When replacing, NEVs reduce emissions by decreasing fuel vehicle share; when added to demand, they exacerbate congestion, increasing per-vehicle emission rate. The mechanistic analysis reveals a linear emission-conventional vehicle relationship under smooth traffic, shifting to exponential under congestion (slight, moderate, or severe). Each level of traffic deterioration raises per-vehicle emission rate by 15%~60%, with the sharpest increase (80%~90%) occurring during the slight-to-moderate congestion transition and severe congestion. This study establishes a theoretical foundation for predicting intersection/network emissions and supports precision traffic management strategies.