Selecting Details from LOPAs to Enable Process Safety Digitalization & Real-Time Risk Management

Back in 2023, we explained at this conference that “Process Safety has a Problem. And we presented a “Modest Proposal” describing changes that process safety professionals – and potentially software developers – should make in order to enable digitalization of process safety “at scale”. Regardless of whether this is done with natural language processing as one input, or corporations have their PHA teams change how they record information, we have found that some pre-processing effort must be applied before datasets in our field can be used together to provide useful insights.

In 2023, we described a far-reaching vision to integrate all the available information data sources including our HAZOPs, our incident reports, our MoC processes, etc. and show how they can be linked with already-digitalized information held by our maintenance and process colleagues’ systems. This presentation will provide a specific example of extracting meaningful data from LOPA scenarios, including their credited safeguards and expected demand rates. Together, these form the “what should happen” part of our risk comparison. And the data extraction enables direct comparison with the “what really does happen” information found in structured data systems like the CMMS and the data historians. Focusing on the LOPA scenarios should reduce the scope to get started on the most significant process safety hazards.

The presentation will show step-by-step how companies can assemble the hierarchical building blocks and suggest the proper sequence, as well as the benefits available as each “block” is placed. It will also describe some changes that PHA facilitators should so their work product can be integrated. This presentation is based on the approach used by “a major international oil company” as well as pilot programs at other forward-thinking operating companies. to digitally transform their process safety program.