Abstract
In this paper, a novel methodology regarding to batch plant safety analysis is proposed. Batch structures are elaborated and categorized into several safety objects that are linked to phase level operation. Batch model is implemented using ISA S88 that naturally represents object-oriented approach. Safety object performance determines safety condition during batch process, if it works in low performance, initiating event will rising into hazardous outcome due to propagation of fault. In the other hand, measured process variable and its liability are important factors that affect safety object performance. To implement safety object performance, raw data is analyzed using fuzzy system approach that starting from knowledge acquisition, connection and mapping the sensor data. From the output of the safety object performance, reasoning mechanism for specific phase level operation can be implemented. The output of safety analysis using fault propagation can be considered as performance index that represents the success or failure of the systems. This research gives significance contribution to improve the batch plant safety with integrates safety management system and information technology. For case study, the polyvinyl chloride (PVC) batch plant that process hazardous material vinyl chloride monomer (VCM) is taken for consideration.
