As modern chemical plants are becoming more complex and bigger in scale, the associated chance of things going wrong is also increasing rapidly. Due to the flammable, explosive, toxic and corrosive nature of chemical process, any single accident may trigger a major catastrophe that brings tremendous environmental, social and economical loss. In order to prevent any accident from happening, hazard and operability (HAZOP) analysis has been brought in to monitor chemical process and provide early warning for signs of accident. However, most existing HAZOP is carried out manually, and there are always obstacles in terms of cost overrun and incompleteness of the analysis. To address the difficulties in current HAZOP method, this paper proposes a signed digraph (SDG)-based HAZOP analysis method. It is used to identify the most likely operating mistakes that may cause certain process variable deviating from its normal value, which is the main source of safety concern. A case study on polyvinyl chloride (PVC) plant is presented to demonstrate the effectiveness of SDG-based HAZOP analysis method in providing complete analysis result.