Why worry about something that isn't going to happen
June 10, 2019
LOPA - the die is cast
May 20, 2015
Inspired by a lively demonstration of human event trees at IChemE Hazards25, I dug out my 10-sided dice that I previously used to explain LOPA in a simple way to my colleagues as follows:
INITIATING EVENT: 2 white dice can be 00 to 99 (1 in 100 years frequency)
BPCS: 1 yellow die can be 0 to 9 (1 in 10 PFD)
HUMAN RESPONSE (ALARM): 1 blue die can be 0 to 9 (1 in 10 PFD)
RELIEF: 3 red dice can be 000 to 999 (1 in 1000 PFD)
OCCUPANCY: 1 green die can be 0 to 9 (1 in 10 chance)
AVOIDANCE: 1 orange die can be 0 to 9 (1 in 10 chance)
VULNERABILITY: 1 black die can be 0 to 9 (1 in 10 chance)
Roll the 1st dice and (for simplicity) if it's 0 then a hazard scenario chain of events has started. If it's anything else then the 'game' stops (doesn't start).
Roll subsequent dice in order and if (again for simplicity) they are 0,00 or 000 (depending on number of dice) then the hazard develops. Try this and you'll find that it is difficult even to get a short chain of failures to escalate.
Obviously this is just an EXAMPLE and you can adjust the number & colours of dice to suit e.g. add another prevention layer for SIF with 1 die for SIL 1, 2 dice for SIL 2 and 3 dice for SIL 3.
You can get these dice online and they provide a neat way to represent order of magnitude failures.
We've also used Ker-Plunk with straws as IPL but that's not as "numerical".