Process modeling has remained challenging in the area of automating hazard and operability (HAZOP) analysis of complex chemical processes. Even though signed digraphs (SDG) model has been proposed and widely applied in various HAZOP expert systems in the past one decade. The limitations of SDG have not been recognized and discussed in literature. The drawbacks of SDG models have caused incompleteness of HAZOP analysis results, which poses a significant limitation on wide acceptance of HAZOP expert systems in industrial practices. In this work, layered digraph (LDG) model extended from SDG is introduced to overcome the drawbacks of SDG models. Knowledge storage and management of LDG models are discussed in detail. Industrial case studies are used to illustrate the effectiveness of this new modeling methodology. Since process variables, process deviations, and interactions of deviations are explicitly decoupled in three dimensions, LDG model has shown to provide a much more flexible and comprehensive mechanism for HAZOP analysis. (C) 2008 American Institute of Chemical Engineers Process Saf Prog 27: 293-305, 2008