Abstract
HAZOP studies are widely used for identifying hazard and operability problems during plant design. However, HAZOP studies are time consuming, labour intensive and expensive. Therefore, major research projects have been carried out to develop systems to automate the HAZOP technique. The basic technology used in the developed systems is built around the idea of fault propagation using signed-directed graph, together with a range of information to capture the possible failures of the equipment, and the susceptibility of the equipment to deviations propagated from elsewhere. A lot of progress has been made in the implementation of the basic fault propagation system and development of signed-directed graphs for basic plant models. However, the limitations of the prototypes resulting from different research groups are twofold. First, they are standalone systems that cannot work with commercially available process design software. Second, they lack facilities for reviewing the results generated automatically by the hazard identification systems. This paper presents a set of novel tools that overcome the identified limitations. These tools include a "data wizard" for extracting data from a process design system and converting the data into a format suitable for the hazard identification software, a semi-automatic tool for mapping the extracted plant items to the available signed-directed graph models and an interactive tool for selectively viewing the HAZOP results in different ways and to provide explanation on how an identified cause reaches an identified consequence.
