Gå til hovedinnhold

“Accident investigations are opportunities for learning”

When investigating accidents and serious incidents, the petroleum industry places greater emphasis on human and technological factors than on organisational factors. The consequence could be that the investigation only identifies direct causes, but never finds the real, underlying causes of the accident.


About 150 people from the industry, management and R&D circles participated in the PSA’s seminar on organisational factors and measures in accident investigations.

The basis for the seminar was a review of the study performed by the Institute for Energy Technology (IFE) on behalf of the PSA, where they have evaluated 20 of the 91 investigation reports the PSA received in 2007/2008, supplemented with experiences from the operators Marathon and Norske Shell and a research-based lecture on learning from incidents by Sintef.

For a link to the IFE report and the presentations from the seminar, see the column on the right. The report is currently in Norwegian only, but will be translated.

Finn Carlsen, the PSA’s Director for Supervisory Activities, opened the seminar by pointing out that accident investigation has been an important topic for the PSA since the agency was founded in 2004, both because it naturally coincides with the agency’s prioritised effort regarding major accident risk, and also because the many and serious accidents which have taken place in recent years keep accentuating the need to focus on this issue.

He referenced the PSA’s cooperation with the IFE in connection with the industry-wide audit activity which was carried out to contribute to the transfer of experience and improvement processes in the petroleum activities. This is part of the PSA’s follow-up of major accident risk.

MTO
Therefore, he was also glad for the industry's presence and contribution of experiences from their own work, together with Sintef. However, he pointed out that the investigation is not a goal in itself, but a good opportunity for learning. Even if the methodology, in this case a Man-Technology-Organisation approach, is important, the will and ability to learn and implement necessary measures is essential.

PSA principal engineers Siri Wiig and Hilde Heber then reviewed and commented on the IFE report “Assessment of organisational factors and measures in accident investigations”. The PSA contracted IFE for a review of a selection of the investigation reports the PSA received in 2007/2008 because the PSA has experienced that the companies’ investigation reports do not identify organisational factors to their full extent. Cultural conditions, management provisions, power relations and framework conditions at different levels are also not fully clarified in the reports.

Industry-wide
The purpose of the work can be summarised in three points: How different categories of organisational factors are assessed (or not), which measures related to organisational factors are suggested and to which degree it is possible to assess the effects of the organisational factors suggested.

This is an industry-wide audit activity, and the PSA will utilise the results from the work in further supervision of the petroleum activities.

Wiig and Heber pointed out that an accident can be explained in different ways, depending on the causal model used in the incident analysis. It is also challenging to gain an adequate perception of the complex and dynamic nature of organisations using a simply integrated method. We have wanted to increase awareness of the fact that investigations and their inherent accident models and methods influence what the investigations are searching for, which is called the “What-you-look-for-is-what-you-find” principle. When you find what you are looking for, the “What-you-find-is-what-you-fix” principle will govern the measures suggested.

Methods
Leif Gunnar Hestholm, HSE manager for Marathon and Morten Andreassen, operational safety advisor for Norske Shell both shed light on the industry’s approach to organisational factors in accident investigations, follow-up and learning. They both reviewed the methods and tools used in the companies’ accident investigations, and how the companies approach organisational factors in this work.

Hestholm pointed out, among other things, that MTO is practiced very differently across the petroleum industry and the common denominator is hardly more than the name. He expressed the need to define standards for an MTO analysis. Andreassen stressed factors such as the importance of the composition and competence of the investigation group, setting aside enough resources for the investigation, and that the group must dare to challenge its own organisation.

Organisation
Ranveig Tinnmannsvik, safety researcher at Sintef opened by referring to Carlsen’s remarks concerning learning and pointed out that organisational factors in an investigation context are a good starting point for learning. “To understand accidents in complex, dynamic organisations, it is important to have a broad understanding of causality related to Man – Technology – and, not least, Organisation”, she said.

Referencing literature following BP’s Texas City accident, we have experienced that, far too often, learning from both your own and others’ incidents and your own supervision simply does not happen. This also applies to measures, which are often implemented to address the structural dimension of organisational efforts, and to a much lesser degree to address the cultural dimension.

The debates following the speeches demonstrated that the topic was perceived as important and relevant, and that the debate will continue in many forums following the PSA seminar.

Contacts in the PSA: