Once the PFD of the SIF is known, then its RRF is simply the inverse of PFD (RRF = 1/PFD). We buy components with published failure rate values that are, therefore, “suitable for use in an SIL environment.” We don’t buy SIL-rated transmitters or SIL-rated control systems. There is no such thing as an SIL-rated device. It is failure rate data that is required as an input to perform SIL calcs for an SIF, not SIL Level data. Companies frequently contract with consultants such as exida to determine failure rate values. The end result is that for each SIF, you end up with an overall Probability of Failure on Demand (PFD).įailure rate data for the numerous pieces of equipment that can make up SIF loops are published by the equipment manufacturers. SIL calcs are somewhat complex and are outside the scope of this article but essentially, the process is to gather failure rate data for the SIF components and account for factors such as test frequency, redundancy, voting arrangements, etc. The overall failure probability of a given SIF is determined by performing SIL calculations (SIL calcs). For this reason, one must know the failure probability the SIF. If this valve is required to close to achieve a particular SIF, it is possible that the valve could stick open and not close upon demand. One very common failure would be an isolation valve that remains open under normal process conditions. Similarly, one must assume that any of the components of the SIF could fail to act upon demand. You cannot assume a hazard will “never” happen. In the PHA discussion, we learned that one must assume that a hazard will occur at some point. It is likely you will find that not every process hazard actually requires the use of a SIS. In other words, once all Process Hazards are identified and Protection Layers assigned, if the PHA/LOPA study concludes that existing protection cannot reduce risk to an acceptable or tolerable level, a Safety Instrumented System (SIS) will be required.