This study presents a Particle Swarm Optimization (PSO)-based scheme for optimal targeted load shedding and contingency severity assessment in the electric power grid (EPG). The IEEE 14 EPG was used as a testbed. The study identified critical branches and quantitatively evaluated the operational performance of an EPG under base case, outage without and with targeted load shedding schemes, utilizing convergence characteristics, voltage magnitudes and angles, and branch load flows as diagnostic metrics. The base case demonstrated excellent numerical stability, with convergence achieved in fewer than 5 iterations, and all bus voltages maintained within the IEEE standard range. A critical outage scenario caused severe difficulty, as evidenced by prolonged convergence (exceeding 15 iterations), a drastic voltage at Bus 1 to 0.7214pu, and overloading of Line 2 to 2.0524pu, approximately 275% of its base case loading. These conditions signified an unstable operational state, posing severe risks to system security. Implementation of targeted load shedding significantly improved system conditions: convergence iterations reduced to approximately 6, Bus 1 voltage restored to 0.9682pu, and Line 2 loading decreased to 0.5683pu. Other buses consistently maintained voltages within acceptable margins, and branch flows on non-critical lines remained insignificant across all cases. Voltage angle profiles further corroborate the systemic stress during outage and the stabilization effect post-load shedding. The proposed technique quantitatively demonstrates that selective load shedding is an effective corrective control strategy, not only restoring voltage stability but also alleviating transmission line overloading, thus enhancing the EPG’s ability to maintain secure and reliable operation under severe contingency conditions.