Cascading risk assessment and event oriented tracing based on probabilistic accident chain for multi-entity distribution networks

Abstract Existing probabilistic accident-chain methods can rank chain risk and vulnerable branches, but they do not connect explicit ESS/PV/EV operating constraints with target-event-specific backward tracing or quantify the contribution of upstream sources and corridors to a selected event. To address this gap, this paper develops a probabilistic accident-chain framework for cascading risk assessment and event-oriented tracing in distribution networks with energy storage systems (ESSs), photovoltaics (PVs), and electric vehicles (EVs). Physical operating constraints are incorporated into probabilistic power-flow sampling, composite voltage-overload-service risk quantification, vulnerable-branch screening, successive branch expansion, and directed propagation-network tracing. In the IEEE 118-bus case, the dominant path to EVENT_23-24 has a path risk of 500.0461, approximately seven times that of the second-ranked path (71.5972). For EVENT_15-33, high loading pressure and branch 23–24 contribute 40.69% and 35.90%, respectively. These results show that initial trigger branches and propagation-critical corridors only partially overlap and that dominant upstream contributions vary with the selected target event. The framework therefore provides an interpretable, quantitatively supported basis for preventive monitoring, corridor-oriented propagation suppression, and event-specific operational intervention.