What are PFDavg Calculations

The Average Probability of Failure on Demand of a device. PFDavg tells you the average probability that a device will fail to protect you at the time a demand is placed on the device. A “demand” being placed on the device when the SIF of which it is a part is called on to carry out its safety function because an unsafe situation has been detected in the process.

A device would fail to protect us only if a dangerous undetected failure had been introduced since the last round of testing. PFDavg describes the probability that this has occurred. Since it is a probability, it is without units. It is easy to see that a device will have a lower PFDavg if it has a low dangerous undetected failure rate (as the failures that would cause the device to “fail to protect” are being introduced at a lower rate) and is tested more often (so there is less time for dangerous undetected failures to happen).

The PFDavg of each device in the SIF must be calculated, then the results are added together (summed) to give the overall PFDavg of the SIF. To repeat – each device in the SIF contributes to the overall PFDavg of the SIF. We calculate the PFDavg of each device and simply add them all together to give the PFDavg of the SIF.

For each SIL band there is a range of (SIF) PFDavg values. Once we know the PFDavg of the SIF we have designed, we can tell if it sits within the required SIL band. If we want to look in more detail at what we achieved with our design, we can find the value of risk reduction we have achieved, by simply taking the reciprocal of the (SIF) PFDavg – this will give us the risk reduction factor.

If we don’t meet the target with this design, then we have to take action to reduce the PFDavg of the SIF, which could be by choosing devices with better (lower) dangerous undetected failure rates, proof testing more often or adding redundancy above the minimum set by the HFT requirements of the device.