HAZOP Assistant, the New Co-Pilot for Process Safety

Since its introduction over 40 years ago - the Hazard and Operability (HAZOP) study process has changed very little. Despite the introduction of new tools and software in recent years to address the shortfalls, at its very core the HAZOP process is still human driven nowadays. Kairos Technology and Vysus Group have taken the step to 'digitalize' the HAZOP process by developing a software tool called "HAZOP Assistant".

The HAZOP Assistant has been recently applied in a pilot digital HAZOP study (dHAZOP) which consisted in a new gas compression platform during Front End Engineering Design (FEED) stage for a Major Oil and Gas Operating Company.

The objective of the project was to compare the performance of the dHAZOP with that of the traditional HAZOP previously performed for this project. The comparison includes the coverage of hazards, causes and consequences (i.e. completeness), consistency and resources, overall costs among other aspects. The paper will discuss the findings from this comparison.

Methods, Procedures, Process:

A functional twin of the project was created using the HAZOP Assistant tool. Making use of the equipment libraries and the tool´s bespoke programming language Multilevel Flow Modelling (MFM), the HAZOP Assistant was able to effectively reason propagation of causes to consequences in any of the equipment included. MFM is a type of Explainable Artificial intelligence (XAI).

Contrasted and tuned equipment libraries used for digital twin modelling led to an increased cause-consequence coverage. The reasoning capability of the tool allowed the team to save time on the study of the scenarios present on the process.

A set of rules for qualifying causes and their likelihood, as well as consequences and their severity, was introduced during the preparation of the dHAZOP workshop sessions. This led to an increased consistency in the application of likelihood, severity and risk ranking rules, avoidance of errors and reduction of considerable time spent in risk ranking.

Results, Observations, Conclusions:

After completion of the workshop sessions, the time taken for the study was reduced, while the quality of the study was maintained in comparison with a Traditional HAZOP. Continued use of the digital twin created for the dHAZOP can allow targeted updates on the study, training on the plant and compatibility with other applications such as Control Room Assistant.

The functional twin can also be augmented with information from diverse PSM studies such as Operational Risk Assessments (ORAs) and Failure Mode and Effect Analysis (FMEAs).

The pilot highlighted both the strengths and opportunities for improving the tool. On one side, the time reduction and the overall coherence on the study were found as important benefits. On the other side, the use of a high quality and detailed model led to a great number of results which required later filtering for analysis purposes. This pilot project provided some invaluable feedback which helped to improve the tool.

Novel/Additive Information:

The creation of a functional digital twin helps improving the efficiency of the HAZOP as found in the pilot study.

This dHAZOP and digital twin can afterwards be related to other safety studies, making it more useful and helping to achieve a better coherence on the overall PSM of the facility.

A functional twin will allow the continuous update and access to retrieve safety information related with the facility.