www.SILSafeData.com is a complimentary resource that contains the upper and lower bounds failure rates for many categories of automation equipment, as well as the methodology used by exida to derive the numbers.
Many of you have asked how to use this data. As an example, let’s look at a control engineer who has just received a certificate from a valve manufacturer for a ball valve.
This document claims a dangerous failure rate of 52.8 FITS. To quickly check if this is a realistic value, the control engineer goes to SILSafeData. The application is “close to trip” without tight shutoff requirements in a low demand mode, and the manufacturer’s website shows this ball valve is a floating type. The SILSafeData chart looks like this:
The statistical low limit for several sets of field failure data and FMEDA analysis is 300 FITS. Obviously, 53 FITS is not a realistic failure rate prediction. In this case, the number to be used in any PFDavg analysis must be in the range of 300 to 900. If the engineer has a lot of experience using this valve and feels it has a high-quality rating, then perhaps the 300 FIT number could be used for the PFDavg safety analysis. Otherwise, the 900 FIT number is recommended, as this analysis is for safety purposes and a weak design may not achieve the needed risk reduction.
Where did the 53 FITS number come from? The method used is not explained in the document, and any idealistic assumptions are not described. There are some methods used that might produce such results, the most common being a “cycle test.” However, the cycle test method is invalid for applications where a mechanical product does not move frequently, such as low demand. Whatever method was used, the results are not realistic.
SILSafeData is based upon numbers generated from a database of over 350 billion unit field operating hours in many industries. It is important to us to provide resources that help prevent unsafe designs.