Similar to the Probability of Failure on Demand, which expresses the likelihood that a safety function (a safeguarding loop) does not react when required, the Probability of Failure to Safety (PFS) expresses the likelihood that a safety function does act when not required. In other words, it measures the likelihood of spurious trips.
Spurious trips are a major headache for operators and maintenance technicians. First, because they can cause major process upsets. Second, because it can be hard to prove that a trip was indeed spurious and not caused for some other, valid reason.
It is for this reason that we often build redundancy into our systems. Redundancy effectively provides different channels to perform the same safety function. If one of the channels fails, the other can still perform the safety function when required… so no reason to “trip” the entire system. Redundancy, therefore, improves system availability.
Consider, as an example, the arrangement of a number of transmitters, on the same line, measuring the same process variable. In this example we can install up to three transmitters. There is a number of ways in which they can be connected… we use the MooN terminology for voting, whereby M “trip” signals must come via N channels for the system to perform the safety function. For instance, “2oo3” voting means that at least two transmitters must measure an unsafe process variable out of an installed total of three transmitters for the loop to trip. In other words: if one of the channels is broken, or stuck in the “safe” position, then the safety function of the system as a whole is not impaired. At the same time, the likelihood of spurious trips is minimised as two channels have to fail in the “safe” state for this to happen. The downside of such a solution is the cost of installing multiple transmitters:
Availability comes at a cost, safety comes standard
We are aware that in every project there is a trade-off between cost and availability (i.e. lack of spurious trips). We can advise you on the degree of redundancy appropriate for your situation and calculate the improvement you make in terms of Probability of Failure to Safety for each option.