A lot of people have expended a lot of effort in calculating nuisance trip rate. Usually, the nuisance trip rate is calculated on a SIF-by-SIF basis. But is the effort really worthwhile? What are you doing with the results? We at Kenexis often calculate nuisance trip rate two ways. Once on a SIF-by-SIF basis, and upon request we calculate the overall trip rate for an entire facility, unit operation, or piece of major equipment. Know the nuisance trip rate for a single SIF does not provide a lot of valuable information, and metrics that limit that rate based on SIF are of little value. Limiting the nuisance trip rate to an arbitrary number, such as once in ten years, for a SIF tells you nothing about how often an entire plant will shutdown, and tells you nothing about the criticality of that nuisance trip, nor does it consider that some instruments are more important than others with respect to nuisance trips. For instance, a shutoff valve on a fired heater’s failure may cause a total plant shutdown, but the bleed valve failing open will have no effect on the process at all (but may be a safety issue).
While considering nuisance trip rates on an overall plant basis give more insight into the “overall” performance of the plant, one still needs to consider the effect of the various instruments individually. Any instrument that can cause a plant shutdown needs to be looked at in isolation, inclusion of the rest of the SIF (or SIFs) into the analysis can result in engineering the wrong part of the loop. Even if good consideration of these factors is made, what is the basis for even an overall nuisance trip rate number? What is the basis for metrics to benchmark the performance against. Saying that you’d like to limit the nuisance trip rate of a plant to less than its major turnaround interval (let’s say 5 years), often provides a bit of a comfort level in the overall design, but if the plant is risky and requires a lot of safety instrumentation, it might not be achievable.
Some operators using more prescriptive methods for determining the amount of redundancy against nuisance trips are finding these methods are easier, require less analysis effort (e.g., less consulting hours), and provide designs that are as good or often better than calculation oriented methods. The approach goes like this…
Consider each instrument that is connected to the SIS. Analyze, qualitatively, what the consequence will be if that instrument fails. If the answer is that a plant shutdown will occur, that will result in more than, let’s say, $500,000, then one degree of fault tolerance against nuisance trips is included. Otherwise, no fault tolerance against nuisance trips is included.
Based on my review of designs, this simple qualitative method will result in superior designs to calculation-based methods, is more justifiable to stakeholders who have to pay for the equipment, and is easier and more cost effective to perform in terms of engineering. I think that everyone should really take a hard look at what they’re doing with respect to analyzing the nuisance trip rates of their SIS, and give very serious consideration to employing the simplified analysis that I have presented above before going down the road of spurious trip rate calculations with associated performance metrics. Even if you have gone down that road, you might want to consider coming back.