The evaluation of grid voltage support capability has long been constrained by localized characterization metrics such as node-based sensitivity and short-circuit ratio

incapable of resolving the dynamic process of reactive power-voltage support propagation from support sources to load nodes through grid structures. This limitation hinders root-cause analysis of structural triggers during metric anomalies and lacks topology-based weak-link identification and early warning methods. To address these challenges

this paper proposes a dynamic assessment and early warning methodology for voltage support capability in high-penetration photovoltaic (PV) distribution networks based on sensitivity propagation theory. First

the essence of node reactive power-voltage sensitivity (RPVS) is revealed as the superposition effect of reactive power regulation from voltage support sources through branch propagation paths in response to node voltage variations. Accordingly

the branch-level Reactive power- voltage sensitivity of the branch to the receiving node RPVS index is established

accompanied by an RPVS branch cascade model to decode equivalent propagation path characteristics from support sources to target nodes. Subsequently

depth-first search (DFS) algorithm is employed to trace RPVS propagation paths

constructing a physically interpretable voltage support topology that achieves dynamic visualization of “source-network-load” sensitivity propagation paths. A partitioned analysis method based on RPVS propagation topology is proposed

along with a distribution network voltage support early warning mechanism. Finally

case studies on modified IEEE 33-node and 185-node distribution systems in a certain city validate the effectiveness of the proposed methodology in dynamically assessing and prewarning voltage support capability for high-penetration PV-integrated distribution networks through sensitivity propagation theory.