SIL Follow-Up in the Operational Phase of Safety Instrumented System

Abstract

Safety instrumented systems (SISs) are frequently used in the process industry to detect the undesired or dangerous process conditions and mitigate the harm to humans, environment, and materials. This thesis attempts to address some unclear points of SISs used in industries. Safety integrity level (SIL) follow-up is such a point.

The concepts associated with SIS/SIL have been thoroughly defined and clarified. IEC 61508 and IEC 61511 are two standards for SIS. In these standards, the functional requirements and SIL requirements are decisive measures for a complete safety instrumented function (SIF). Both qualitative and quantitative parts of these requirements are presented.

SIL follow-up is activities associated with SIS/SIF in the operational phases to ensure that experienced performance (e.g. PFD) of each SIF continues to meet the SIL requirements. Standards and regulations related to SIL follow-up have also been presented. Desired (or required) SIL, predicted SIL, actual (or experienced) SIL are three terms related to SIL. Desired SIL is the target that SIS users want the SIS to achieve. Predicted SIL is a forecasting about the SIL performance will perform. Actual SIL is the SIL performance the users actually get when the specific SIS is already utilized in the actual process. The goal for SIL follow-up is to let the actual SIL performance better than the SIL performance that predicted by engineers.

Different approaches (includes CCPS approach, OLF 070 appendix F approach, DNV procedure, and PDS approach) on SIL follow-up found in literature have been presented. The difference between these four approaches has been discussed with different stages in operational phase separately.

Four SIL follow-up methods have been applied on a case study based on sensor part of a high integrity pressure protection system (HIPPS). The PFD is a recognized measure for the SIL performance in the operational phase. It is not reasonable to change the design to make the experienced PFD in accordance with the predicted PFD. Except for CCPS approach; the other three approaches attempt to achieve acceptable PFD by updating the functional test interval. Conclusions are drawn, based on the numerical results.

In addition, for the PDS approach to become more complete, the following suggestions are made:

 Improve data collection mechanisms i.e. discuss how to developing performance indicators and list the suggested indicators

 Improve data quality i.e. data availability and data accuracy

 Conduct data analysis not only by comparing the values, but also can have some other aspects such as tread analysis

Finally, this thesis advances two proposals on future work. Continuous improvement is always necessary to perfect the SIL follow-up procedure.