In this paper, we have introduced a novel method for failure analysis of complex systems in the chemical process industry based on recent advances in complex networks. We model the engineering system as a large, complex network and choose to investigate failure from the topological perspective. Then, we introduce a model incorporating the load of nodes and the efficiency to probe the mechanism of cascading failure, and analyze the results based on the topological properties. A case study of a sym-tetramethyl benzene (STBZ) production system is provided to illustrate our work. The study results show that the safety performance of a complex system is indeed affected by the structure and topology of the system, and malfunctioning of a single node can inflict substantial damage, due to the system possessing a highly heterogeneous distribution of loads. Although the events have a very low probability of occurrence, their occurrence may collapse the entire systems. The method can help to identify the critical components of the process industry system, and can also be available for failure prevention and design of a safety control system.