Safety assessment schemes levels

One consequence of the increasing realisation that competence has a direct impact on health and safety has been the development of conformity assessment schemes covering competence. Conformity assessment, in its generality, is an activity concerned with determining directly or indirectly that relevant requirements have been met, and is most commonly used to determine if a product, system, process or a person s competence meets a defined specification. In the context of this chapter, conformity assessment relates to the process of determining whether people working in the electrical or safety-related control systems sectors can perform to the required level of competence. 

The possibility of repair during accident situation is yet to be included into Probabilistic Safety Assessment (PSA). A PSA method Includes Event tree (ET) to generate accident scenario and Fault Tree (FT) used to quantify the probability of failure of the Safety Barrier (SB). In literature, there are some examples in which FT analysis has been done for systems with repair components. A FT analysis of phased mission systems has been performed with repairable and non-repairable components (Vaurio 2001). The other FT models with repairable basic events have been proposed based on renewal intensity principle and inclusion-exclusion methods (Yuge et al. 2012 Yuge et al. 2013). However, ET analysis of Initiating Events (IE) with repairable SB has not been dealt before. Therefore this paper proposes a scheme of PSA level-1 for systems with repairable components using a simplified combination of Repairable Event Tree (RET) and Repairable Fault Tree (RFT). 

The KARYON project developed an architectural pattern, which allows to react on a degraded functionality by switching to different levels of service, i.e. to differnt control schemes. The decision is based on the assessment of system health in a broad sense. In this paper, we focussed on failures of the sensor system. For a simple example we showed how the notion of validity can be used for design time analysis and also in run-time assessment of sensor data. A rehable safety kernel monitors the validity and takes actions if validity drop)s below a predefined bound. The bounds on validity and the necessary knowledge for the controller can be statically analysed at design time and transformed into safety rules to be executed at run-time for configuring controller functions. 

---