This study evaluates how the gradual market penetration of autonomous vehicles influences traffic flow and the environmental footprint in an urban setting. The research focuses on the city of Brașov, using a central urban road area as a case study; this area consists of a one-way traffic loop connecting several road junctions and roundabouts, thereby facilitating the analysis of congested traffic flows. It employs PTV Vissim microscopic simulation software, utilizing vehicle-following models (Wiedemann 99) and the technical framework developed within the European CoEXist project. The methodology involves modeling various autonomous vehicle penetration scenarios, ranging from a flow composed entirely of conventional human-driven vehicles and stages of mixed traffic to full automation. The "AV Aggressive" (CoEXist) behavioral profile was selected for the simulation, modeling optimized autonomous vehicles (SAE Level 4/5). This profile reflects high-precision driving based on wireless connectivity (V2X/CACC), minimal reaction times, reduced safety gaps, and platooning capabilities, representing the technology's maximum potential compared to the uncertainties associated with human drivers or cautious driving profiles. Through node- and network-level evaluations, the study analyzes traffic indicators—such as average delay per vehicle, stopped time, and queue length—as well as environmental and energy indicators, specifically fuel consumption and pollutant emissions.