Over the last few decades, the Shell Group has been quite successful in reducing LTIF and TRCF by focusing on these indicators, but this has not resulted in an equally significant reduction in FAR.
One explanation for this is seen to lie in the fact that the incident "iceberg" has a significantly different shape for different classes of activities.
Focusing on reducing LTIF and TRCF can result in a focusing on the wrong class of activities, which actually contribute less to the overall FAR.
During the 80's and 90's the Shell Group has made a significant effort to improve safety standards and performance, focusing notably on reduction of Lost Time Incident Frequency (LTIF) and Total Recordable Case Frequency (TRCF). Although this has been quite successful in that both LTIF and TRCF have reduced significantly, it has also become clear that this has not resulted in an equally significant reduction in Fatal Accident Rate (FAR).
Annual Safety Performance of the Shell Group (Fig. 1) shows an average annual decrease of about 10% in LTIF and TRCF, whereas the average annual decrease in FAR is about 4%.
Similarly, industry wide Safety Performance indicators compiled by the International Association of Oil & Gas Producers (OGP) show that in the last 10 years (1994–2003), LTIF and TRCF have reduced by approximately 66% respectively 56%, whereas the FAR only decreased by 27%. (Ref. 1)
One of the causes may be an over reliance on the classical "Ice Berg Theory". In simple terms, we have assumed that there is a more or less fixed ratio between very serious incidents (e.g. Fatalities) and related incidents with a less serious outcome such as Medical Treatment Cases (MTC's). Based on this assumption, a reduced level of TRC's should result in a similar fall in FAR's.
That this fixed ratio is a fallacy can be seen if we look at the case of drowning. Very few Lost Time Incidents (LTI's) actually occur related to people falling into the water. However, fatalities are all too frequent. Other typical examples are electrocution and incidents of fire/explosions.
The following figure from the US Bureau of Labour Statistics (Fig 2) shows the large variation (note the logarithmic vertical scale) of the ratio between LTIF and FAR for different types of activities.