Safety

Effective Safety Culture Necessary To Understand Accident Causation

While there is no one way to prevent all accidents from occurring, organizations can develop an effective safety culture that allows them to learn from their mistakes.

ogf-2014-12-pfcr-safety-accident.jpg

Accidents happen in the petrochemical industry. They can be major or minor, their causes vary, and they can have a devastating effect on both the public and organizations.

While there is no one way to prevent all accidents from occurring, Maury Hill, a former investigator for the Transportation Safety Board of Canada, said that organizations can develop an effective safety culture that allows them to learn from their mistakes.

In a presentation hosted by the SPE Human Factors Technical Section, Hill outlined the challenges that organizations face in conducting safety investigations and linked human factors, organizational factors, and accident causation. By understanding the root causes of accidents, a company can put itself in an optimal position to prevent them in the future.

Every organization has a culture of safety, said Hill, who currently heads the consultancy firm Maury Hill and Associates. The trick is knowing whether it is an effective safety culture and knowing how to improve it. A just culture breeds a good reporting culture within an organization. A good reporting culture breeds a flexible culture and a good learning culture.

At the heart of the safety culture is a thorough understanding of human factors. Hill differentiated between human factors and human error. Human factors are the input variables to human performance, while human error is a post hoc term used to describe a performance that does not go as anticipated. Hill said the term “human error” is too vague to hold much significance in a safety investigation.

“While human error is very prevalent, saying it’s a cause of an accident is like saying the reason my grandmother fell and broke her hip is because of gravity. Well, of course, there was human error. What we need to understand is that human error does not explain an occurrence. It demands an explanation, which is then what takes us back to an examination of the antecedent factors that led to that outcome in the first place,” Hill said.

Hill categorized human factors into three areas: physical, cognitive, and organizational. Physical human factors are the anatomical, anthropometric, physiological, and biomechanical characteristics that relate to physical activity. Cognitive human factors are the mental processes, such as memory, perception, or reasoning, that affect people and other elements of a system. Hill cited the effect of fatigue on workplace safety as an example. Organizational human factors are the policies, processes, procedures, and culture that could affect human behavior.

He explained these categories by asking three questions: What are people being asked to do? Who is doing the work? Where are they working?

In examining the human factors that led to an accident, Hill said the investigative process must be system-centered and not person-centered. A good industrial culture should not focus on finding an individual to blame and punish, but rather to identify the latent unsafe working conditions that led to the accident. Hill defined this culture as an error-tolerant culture, one where the people at the top of the organization recognized that people will make mistakes on the job and figure out how to make sure the same mistakes do not happen in the future.

“In the safety world, it’s a very rare day when you’re actually looking at irrational behavior,” Hill said. “If it is, that’s a medical issue and it’s not part of a safety investigation. You quickly go down a different path. What you’re investigating is rational behavior, and the challenge is what is it about that behavior, what is it about those factors that led the person to think the action they did was actually reasonable and rational?”

To better understand the linkage between human factors and accident causation, Hill introduced three guiding principles. The first is that people are not wired to draw conclusions only after gathering all the data from an accident (i.e., given two points of reference, the human brain will automatically try to connect them), and it is not a good course of action to think otherwise.

Second is the concept of fundamental surprise, or the idea that people on course for an accident may not realize they are on the verge of an accident until moments before it happens. The last principle is that there is very little linkage between input variables and outcome variables. A major accident may not have substantial causes.

The archived web event is available at https://webevents.spe.org/products/meeting-the-challenges-of-effective-occurrence-investigations