The US Chemical Safety and Hazard Investigation Board has just released its preliminary findings on the Deepwater Horizon explosion in 2010 that killed 11 workers and caused a massive spill in the Gulf of Mexico. What makes these findings interesting is that they focus not on the technical failures but on management system deficiencies.
Key among these deficiencies was an over-reliance on personal injury rates as indicators of adequate safety. In essence, reducing or eliminating personal injuries was the primary focus of the safety programs and injury rates were used to measure success. However, BP failed to also consider process safety indicators, such as automatic shutoff system failures, activation of pressure relief devices, or loss of containment of liquids and gases, that could signal potential failures.
What makes this particularly egregious is that these same problems were identified in a 2007 report on the explosion at BP's Texas City TX refinery in 2005.
The lesson here is that we need to be careful in the selection of indicators and understand just what they are measuring and how that relates to our overall program goal. This is not just limited to the safety arena. I've seen many performance indicators that rely solely on objective measures that have no correlation to program success. In the long run, poorly thought out indicators can cost you greatly.