Organizational Accidents - Chapter 10

The years between 1976-1988 were marked by the years with the most major disasters worldwide. On page 73 of ‘A Life in Error’ there is a list of a few of these major accidents between ’76-’88.

Causes of the accidents

Although all these kind of disasters are different, they share at least these three characteristics:

• Many of the contributing factors were present within the system before the actual catastrophe occurred.
• All the systems had multiple defences, barriers and safeguards designed to prevent knows hazards.
• The disasters occur because an unforeseen links of latent conditions (unsafe acts by humans and local triggers) that defeated the defences.

The Swiss Cheese Model (SC) is another way to explain these views. The system defences are represented as slices of cheese, they intervene between the operational hazards and potential losses. When it would be a perfect system, the cheese slices were intact but in reality it is more like Emmenthaler (with a lot of holes). One hole would not be a problem, but many slices on top of each other with the holes at the same direction create an opportunity for accidents to happen.

Holes in defences can arise for two reasons. Firstly, active failures are errors and violations caused by something or someone that is in direct contact with the system. Secondly, latent conditions are more long lasting and are incorporated in the system, because you can’t correct for all possible scenarios. As you might expect, because of the long-time existence of the latent conditions, there can be acted pro-active. The safety health can be measured and there can also be more attention for the vital signs (planning, scheduling, training, designing, communicating, building, operating and maintaining).

When organizational accidents occur is it important to focus not on who did it, but on how and why did the defences fail? Accidents give an opportunity to identity where things went wrong, for example the work pressure or inadequate training.

Causes of Error

Now, around 70-80% of all the accidents in technology are caused by human error. In 1960 it was only 20%. There are several reasons why it seems that humans became more fallible in the past years. The material and mechanical elements of technology became more reliable in the past years because of more knowledge. Also between 1970-1980 automation increased dramatically, which lead to more control by fewer operators. Lastly, the automation had three consequences: it can lead to a higher burden of the individual due to the workload, technological layers make it less transparent to the operators and although automation reduces slips and lapses it places a greater load upon the individuals reasoning skills.

Two Kinds of Accidents

By now it is safe to say that accidents can be categorized into two types:

Firstly, individual accidents: Frequent and have limited consequences. Occur in systems where there are few defences and arise from limited causes: slips, trips and lapses. These accidents are caused by the failure in the personal protection.

Secondly, organizational accidents: Rare occurrences that are more widespread and have devastating consequences. In these systems there are more defences. These big bangs are low in frequency and high in severity. These accidents are caused by a combination of multiple failures that are linked.

A key question is: Do individual accidents provide a reliable guide to a system’s vulnerability to organizational accidents? The answer is no. The road to a big disaster is paved with declining or low lost time frequency rates (LTIFR’s).