Collecting Human-Factors Data From Accidents and Incidents
- Rachael Gordon (U. of Aberdeen) | Rhona Flin (U. of Aberdeen) | Kathryn Mearns (U. of Aberdeen)
- Document ID
- Society of Petroleum Engineers
- SPE Production & Facilities
- Publication Date
- May 2001
- Document Type
- Journal Paper
- 73 - 83
- 2001. Society of Petroleum Engineers
- 6.2.8 Ergonomics, 6.1.2 HSSE Reorting, 4.1.5 Processing Equipment, 4.5.5 Installation Equipment and Techniques, 6.1 HSSE & Social Responsibility Management, 4.1.2 Separation and Treating, 6.1.5 Human Resources, Competence and Training, 7.2.1 Risk, Uncertainty and Risk Assessment, 4.3.4 Scale
- 1 in the last 30 days
- 213 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
To investigate human and organizational factors in offshore accidents, two human-factors incident reporting forms were developed and evaluated. The Open Reporting Form (ORF) contained 11 open questions and the Comprehensive Reporting Form (CRF) contained 166 questions/potential causal categories. Both forms were tested on five U.K. offshore installations where personnel involved in an incident were asked to complete one. Using statistical techniques, the form with more structured questions was found to provide 66% more information than the participating company's existing reporting form. Results of the research are described focusing on potential benefits for the offshore oil industry.
Previous research has found that human error is a cause in 90% of industrial accidents.1 Despite this, analyses of large accident databases still tend to focus on the types of injuries and operations which were carried out at the time of the accident. Causal analysis, especially of the human and organizational factors which may have led to the accident and are not always easily identifiable, is often overlooked. Collection of accurate accident data is seen as an important step for the improvement of industrial safety. Industries, such as the offshore oil industry, have accident-reporting systems which are vulnerable to underreporting, have incomplete recordings, and may not provide a complete picture of the conditions under which accidents take place.2 There is currently no standard accident-reporting form in existence across the oil industry. Companies tend to develop their own specific forms. Despite this, the majority of companies operating on the U.K. continental shelf (CS) base their accident-reporting system on the Intl. Safety Rating Systems (ISRS) Intl. Loss Control Inst. (ILCI) system,3 which along with other systems in use, lacks a firm, theoretical framework containing psychological factors. Although information produced from current accident-reporting forms is extensive, its quality and quantity concerning the human-factors causes of accidents is generally poor in a number of ways, such as the sparse inclusion of human-factors codes and the lack of understanding of these codes when present. The ultimate purpose of this work is to improve accident analysis to learn from previous incidents and consequently reduce the likelihood of similar incidents recurring. Data collected through this system will be structured in the hope that companies will be able to analyze weaknesses in their safety-management systems and have a greater awareness of the root causes of accidents in the offshore oil industry.
The study reported in this paper was commissioned by a U.K. offshore oil operating company to develop an incident-report form which would gather human-factors data from offshore workers involved in incidents. This study forms part of the third package, "Human Factors Coding In Accident Analysis" in the Aberdeen U. industrial psychology group's project: "Factoring The Human Into Safety: Translating Research Into Practice."4-6 The specific problem which the operating company had with its accident-report system was that the human-factors causal data being extracted from its current reporting form were not providing information which could be used to improve its systems. Therefore, the specific aim of this study was to improve the structure and content of the incident-reporting form regarding the potential human-factors causes of accidents and near misses.
This paper first describes the participating company's original reporting form which is used to benchmark the findings from the two new reporting forms. The paper then describes the development and evaluation of the ORF and the CRF. Finally, overall conclusions about the two new forms are discussed.
Company's Original Reporting Form
Table 1 gives the original codes used by the company to describe the causes of incidents. These codes are similar to those used in the ILCI Model.3 One of the main criticisms which the company had about this incident data-collection system was that the data collected were not detailed enough to be useful in improving its systems.
An informal evaluation of the company's current coding system revealed that it does not cover a comprehensive range of causes, thus limiting its ability to collect detailed information. A closer examination of the system revealed that some important codes had been overlooked, such as "using wrong equipment" in the immediate cause category and "poor communication" in the underlying cause category. To obtain more detailed and possibly more accurate data from accident investigations, it was decided that the new form should be completed by those people who had witnessed the event, and that they should describe the event in their own words.
Open Reporting Form
The ORF uses open questions to obtain human-factors information about incidents from personnel who were directly involved. Open questions allow respondents to write answers in their own words, as opposed to closed questions, which require personnel to respond from a set of choices.
Development of ORF.
The ORF was designed to be used in conjunction with the company's original reporting form which is completed by an investigation team. Using the ORF, individuals involved in an incident were required to describe the events leading up to the incident in their own words with the expectation that more detailed information would be collected. The structure of the ORF is based on a self-report form developed by British Airways for use in their Human Factors Reporting Programme (part of their BASIS system) to collect incident information from flight crews.7
The 3-page ORF contains these 11 open questions.
Narrative description of the activities engaged in before the event.
Description of how the job was planned.
Deficiencies with the tools and equipment.
Contribution of working conditions to the event.
Description of how the procedures worked.
Description of how the individual was feeling at the time of the incident.
Description of others involved in the task and how they responded.
Description of how training prepared them for the situation.
Description of better ways to handle the situation.
Description of how well the situation was handled.
Comments on how to prevent this type of incident
|File Size||136 KB||Number of Pages||11|