| dc.description.abstract | The pervasive presence of safety instrumented systems in the process industry indicate the growing need for reliable loss prevention to the minutest degree. The backdrop towards addressing this need is by drawing up a justifiable rationale, through which the nature of the demand under which these safety related electronic control functions are subjected to are determined. This is important because the type and class of the safety function an equipment under control is to be cost effectively equipped with, is based on the demand rate to be anticipated.
The international electrotechnical commission is a non-profit, non-governmental standards
organization which ensures compliance for all industrial fields that employ electrical, electronic and related technologies. It defines a demand rate discrimination of once per year between low and high/continuous demand modes and employs two respective measures (average probability of failure on demand and probability of hazardous failures per hour) to assess the safety systems independent of the demand rates. The hazardous event frequency is a unique measure which takes the demand rates on the equipment to be protected into consideration, and
from this approach, a high-demand mode can be justified empirically when the product of the
demand rate and the proof test interval is significantly greater than one.
A salient necessity as required by the standards for equipments subjected to high/continuous demand mode is that in the event of a dangerous failure that is detected by online diagnostics, the equipment ought to advance to a safe state. This places a heavy reliability burden on failsafe mechanisms, since these nature of failures are considered safe. A comprehensive failure mode effect and diagnostic analysis is necessary to identify, isolate and implement diagnostic checks with minimal systematic and hardware safety integrity issues.
The main thrust of this thesis centred on analytic assessment of the safety functions subsystems of safety instrumented systems, taking into account the effect of dynamic variations on the failure rate, brought about by differing demand durations. As a basis for this line of argument, a postulation was made to relate the dynamic failure rate and the dangerous undetected failure rate using the Arrhenius failure rate analogy, in order to effectively model the transitions within these subsystems.
With respect to the cases considered, analysis of the trends affirmed that as the recovery
rate tended to be higher than the demand rate, the effect of the dynamic failure rate becomes
inconsequential regardless of configuration type. | en |
