21
Human Factors in Accident Investigation
from one side of the bridge leaving the navigator unaware of the vessel’s
depth. The navigator could not see the indicator from his position because
of the design of the bridge and failed to retract the dome when the vessel
entered shallow water.
• A safety constraint was ignored. Prior to leaving the Port of Zeebrugge, the
boatswain – the last person on G Deck of the Herald of Free enterprise – said
that he did not close the bow doors of the ferry because it was not his duty.
• A safety constraint was inadequate. Mosenkis ( 1994) gives the example of a
contact lens package that contained two, almost identical, bottles: one containing a caustic lens cleaning fluid and the other containing saline to wash
the fluid off when the lens had been cleaned. Small, difficult to read labels
were the only clues as to the contents of the bottles. Eye damage occurred
when users cleaned the lenses with saline and then washed them with cleaning fluid. The example illustrates the failure to take into account the users’
capacity throughout the t ask – at the critical point, the user’s vision is at its
poorest. Shape and colour could be used to distinguish the two bottles ( e.g.
a red, hexagonal bottle for the cleaner and a blue, smooth one for the saline).
In practice, when safety occurrences are being investigated, the identification of failures such as those described above is part of the initial micro-ergonomic analysis of the
event. It is followed by asking deeper questions about how the organisation manages
safety: how effective the constraints were; why the constraints failed and whether similar accidents had happened before ( and if they had, how the organisation had responded,
if at all). Answers to these questions might be found in the organisation’s ‘ Hazard log’.
A Hazard Log is defined in Def Stan 0 0-56 Issue 4 as: ‘ The continually updated record
of the Hazards, accident sequences and accidents associated with a system. It includes
information documenting risk management for each Hazard and Accident’.
In a safe system, a hazardous process is controlled by:
• A hierarchical control system
• Safety constraints
• Feedback about the operation of the process and the operation of the constraint (Figure 2.4).
So, when investigating an accident, we can begin by asking the following questions:
1. Were all the appropriate constraints in place?
2. Was the process controlled correctly?
3. Was the correct feedback received?
ACCIDENTS AND HUMAN ERROR AS A
STARTING POINT IN ANALYSIS
Accidents are often attributed to human error. Statements such as ‘ 90% of accidents
are due to human error’ are common, but they are wrong. As we saw in T able 2.2,
people make errors all the time and normally nothing h appens – when errors result
Human Factors in Accident Investigation
from one side of the bridge leaving the navigator unaware of the vessel’s
depth. The navigator could not see the indicator from his position because
of the design of the bridge and failed to retract the dome when the vessel
entered shallow water.
• A safety constraint was ignored. Prior to leaving the Port of Zeebrugge, the
boatswain – the last person on G Deck of the Herald of Free enterprise – said
that he did not close the bow doors of the ferry because it was not his duty.
• A safety constraint was inadequate. Mosenkis ( 1994) gives the example of a
contact lens package that contained two, almost identical, bottles: one containing a caustic lens cleaning fluid and the other containing saline to wash
the fluid off when the lens had been cleaned. Small, difficult to read labels
were the only clues as to the contents of the bottles. Eye damage occurred
when users cleaned the lenses with saline and then washed them with cleaning fluid. The example illustrates the failure to take into account the users’
capacity throughout the t ask – at the critical point, the user’s vision is at its
poorest. Shape and colour could be used to distinguish the two bottles ( e.g.
a red, hexagonal bottle for the cleaner and a blue, smooth one for the saline).
In practice, when safety occurrences are being investigated, the identification of failures such as those described above is part of the initial micro-ergonomic analysis of the
event. It is followed by asking deeper questions about how the organisation manages
safety: how effective the constraints were; why the constraints failed and whether similar accidents had happened before ( and if they had, how the organisation had responded,
if at all). Answers to these questions might be found in the organisation’s ‘ Hazard log’.
A Hazard Log is defined in Def Stan 0 0-56 Issue 4 as: ‘ The continually updated record
of the Hazards, accident sequences and accidents associated with a system. It includes
information documenting risk management for each Hazard and Accident’.
In a safe system, a hazardous process is controlled by:
• A hierarchical control system
• Safety constraints
• Feedback about the operation of the process and the operation of the constraint (Figure 2.4).
So, when investigating an accident, we can begin by asking the following questions:
1. Were all the appropriate constraints in place?
2. Was the process controlled correctly?
3. Was the correct feedback received?
ACCIDENTS AND HUMAN ERROR AS A
STARTING POINT IN ANALYSIS
Accidents are often attributed to human error. Statements such as ‘ 90% of accidents
are due to human error’ are common, but they are wrong. As we saw in T able 2.2,
people make errors all the time and normally nothing h appens – when errors result