Abstract
When the question is asked, whether the operation is safe, several sources of information can be tapped to provide information for a reply. But do we really know? Risk assessment is a possibility to obtain a general picture, but it focuses much on stored quantities and technical aspects and is basically static. Management quality, organizational aspects, and human reliability remain out-of-sight as well as temporal risk effects. In this paper, the notion of ‘risk factor’ will be introduced with effects on the short, middle, and long term. Short term risk factors may be monitored physically by means of the digital control system using fault diagnosis, and if required, additional special purpose sensors. For fault diagnosis, several methods are available, but the method making use of the Blended Hazid approach followed by a reasoning routine looks attractive. For middle and long-term risk factors, it is proposed to use the values of process safety performance indicators as exponents of the safety level. To keep it practical, aggregation of the indicator data may have to be applied first. Because it is expected to have synergy from a holistic approach combining technical and organizational safety information, a universal cause-effect logic infrastructure was sought. Bayesian networks seem to fulfill this demand; they can be instrumental for performance monitoring and therefore can greatly help to improve situational awareness.
