Evaluation of the Impact of Diagnostic Coverage and Common Cause Failures on Multi-Sensor Architectures in SIS Design

The evaluation of a sensor network used in a Safety Instrumented System (SIS) is an essential phase to guarantee the system’s credibility. It’s important to select and implement appropriate sensors, logic solvers, and final elements to ensure the proper functioning of an SIS. Sensors are essential to this system because they provide the data that triggers the safe response. Therefore, it’s important to ensure that they are reliable, accurate, and compatible with the process conditions and logic solvers. When choosing sensors, considerations include the type and range of the variable being monitored, sensor placement and ease of access, environmental and operating conditions that could affect performance, sensor response time and precision, as well as the communication protocol and integration with the logic solver. Using different multi-sensor architectures, redundancy and interdependency should be considered, this will help to avoid common cause failures. In sensor networks, redundancy can be employed to compensate for sensor failures, using M out of N architectures, with more than one hardware fault tolerance (HFT≥1). These architectures ensure measurement reliability and availability. In this paper, we propose to compare different redundant architectures for multi-sensors by calculating the average frequency of dangerous failure per hour PFH, and evaluating the influence of diagnostic coverage, DC, mission time, T
1
, and common cause failures factors, CCF, in the PFH calculation. This assessment is based on the formulas presented in IEC 61508 Part 6. We will also show the importance of using multiple sensors in SIS. This includes recognizing different multi-sensor design configurations, considering common cause failures, and performing the average frequency of dangerous failure per hour PFH and fault tolerance calculations in the context of these multi-sensor architectures.