Safety Instrumented Function (SIF) design is based on performance requirements. During design, assumptions concerning SIF equipment failure rates, SIF demand rates, proof test frequencies, and mission time are made. These form the basis for the expected performance.
Are those assumptions realistic? Is the SIF really operating as assumed?
IEC 61511 expects users to periodically compare actual performance with assumed performance. How do we do that? What is needed to get statistically viable results from the collected data? How does the data apply? How do we make sure we avoid drawing unrealistic conclusions?
This presentation will look at answers to these questions and how creating a data philosophy and exida’s SILstat can help with this.