This study addresses the challenges of grid stability posed by the integration of renewable energy sources, focusing on regions, like Jeju Island, where overgeneration complicates energy management. We introduce a novel sensitivity-based control algorithm for Virtual Power Plants (VPPs) that optimizes the deployment of distributed energy resources (DERs). Unlike conventional curtailment methods, such as Last In, First Out (LIFO) and Pro-rata, which often overlook the nuanced impacts of individual DERs on grid stability, our approach leverages electrical sensitivity analysis. This analysis quantifies the effect of each DER on voltage stability, allowing resources to be grouped and managed based on their specific sensitivity. By deriving the sensitivity matrix from the Jacobian matrix of power flow equations, the algorithm identifies critical DERs that significantly influence grid conditions, enabling more strategic curtailment decisions. The proposed method was validated using real-world data from Jeju Island, demonstrating improved efficiency and reduced curtailment compared to existing techniques. These findings underscore the potential of sensitivity-based control to enhance the efficiency of renewable energy systems, supporting sustainable energy transitions.