Electric vehicle charging networks face critical vulnerabilities during power outages, where localized infrastructure failures create demand surges that propagate beyond immediate failure zones. This study presents a spatial-temporal assessment framework for electric vehicle charging network quality of service under outage scenarios using real-world Berlin data. We implement a zone-based approach modeling aggregate electric vehicle behavior across 25 Berlin subdistricts, incorporating 61,768 electric vehicles and 1,411 charging stations. The methodology integrates fuzzy logic charging decision systems with Berlin-specific operator tariffs and Markov Decision Process trip generation. Our analysis reveals counterintuitive vulnerability patterns where single-district outages create 227\% waiting time increases and neighboring districts experience more severe impacts than directly affected areas. Multi-district outages demonstrate non-linear escalation effects, generating 330\% waiting time increases with system-wide stress patterns extending beyond 20 kilometers from outage epicenters. These findings provide actionable insights for infrastructure operators and policymakers, demonstrating that coordinated failures pose disproportionate threats to urban mobility systems.