The execution of supply process orders in a supply chain is conditioned by different types of disruptive events that must be detected and solved in real time. In this work we present a compound web service that performs the monitoring and notification functions of a supply chain event management system. This web service is designed based on a reference model that we have proposed to improve the event management activity through a deeper analysis of the occurrence and causality of events, leading to anticipate an exception during the execution of a supply process order. The web service composition is defined based on business processes. The ability to proactively detect, analyze and notify disruptive events is given through of a Bayesian network with decision nodes.