USING KNOWLEDGE OF ACCIDENT CAUSATION TO MITIGATE DESIGNERS' AND OPERATORS' MUTUAL MISCONCEPTIONS

EPSRC Funding: £85 855

Abstract:

Accidents in both onshore and offshore process plants have had disastrous consequences – for loss of human life, loss of revenue and severely degraded environment. An analysis of both the immediate and latent causes of these accidents reveals the central role played by misconceptions. Complex hazardous installations are naturally vulnerable to designers designing in a way that impedes operators' reasonable intentions. They are equally vulnerable to operators misunderstanding reasonable intentions on the part of designers. The purpose of this work is to understand both directions of misconception - by designers of operators, and by operators of designers - jointly. The aim is to extract knowledge about misconceptions from investigative reports and direct it at the primary decision making points in both the design process and operating process.

So far we have completed the data collection and analysis stages. The data used was basically secondary data, not first hand observations of accidents. It consisted of investigative reports of accidents in the marine, offshore and onshore process industries, all of which were in the public domain. Marine accidents were based on the MAIB's safety digests. Onshore process accidents were based on Loss Prevention Bulletin reports and case studies described in books. And offshore accidents were based on public enquiry proceedings and HSE incident reports. The first step of the analysis was to develop causal trees to represent more formally the reasoning in the investigative reports. This was not meant to make any inferences beyond those in the reports. The second step was to identify misconceptions within these causal trees – that is, inappropriate beliefs on the part of anyone implicated in the accident. Our primary interest was in designers and operators (where 'operators' can also include maintenance staff), but there were cases where manufacturing and installation staff also appeared to suffer from misconceptions. The third step was to develop a taxonomy of the misconceptions by attempting both to group them and generalise on them. As with most work of this kind, the process was a subjective one that involved substantial iteration.

The project is still underway and the results are still very tentative. Our main output is a taxonomy of misconceptions: assumptions made by designers about operators or the operating environment that imperil the system, and assumptions made by operators about the design or the designer that imperil the system. The taxonomy has been incoporated into a computer package that can be used 1) for training and sensitisation, 2) for prompting and reminding. We're about to evaluate this to find out whether it's informative and how it could best be incoporated into normal risk assessment activities.

GO TO .... | ideo home | SII home | EPSRC SI programme home | SII call for proposals | SII proposers' guidance | SII contacts