Impact of gas detection coverage on SIF SIL Rating

Selecting the technology and knowing where to place gas detectors for maximum coverage effectiveness are probably the two most critical decisions for any fire and gas system (FGS) designer. A wrong choice of technology coupled with an inappropriate positioning of the gas detector may render the gas detection loop completely ‘blind’ to nearby gas leaks. In recent times, the need to estimate the probability of failure upon demand of gas detection related safety instrumented functions, have brought into sharper focus the importance of appropriate detection technology selection and detector positioning, as these are the fundamental factors that significantly determine the level of detection coverage.

Many FGS experts have long realized the difficulty of achieving a high SIL rating for gas detection related safety instrument functions (SIF) unless a very high level of detection coverage (i.e. probability of the gas sensor detecting gas) is achievable. The approach used by many to calculate the average probability of failure upon demand (PFDavg) of a gas detection related SIF simply assumes that the gas sensor has a 100% probability of detecting the passing gas cloud. In reality, a 100% probability of detection is quite likely to be overly optimistic especially when the gas sensor is not located in close proximity to the actual leakage point or if one or more environmental influences such as wind, terrain, stationery objects and area mechanical congestion exists to make gas cloud movement unpredictable or erratic.

To demonstrate how detection coverage could affect the SIL rating of a gas detection related SIF, this article will use a modified event tree called the FGS Risk Model. The FGS risk model was developed by safety systems experts to take into consideration two other factors besides FGS availability. These two factors are detection coverage and mitigation effectiveness. These two factors, together with FGS availability (see definition below), are collectively called FGS Effectiveness. An understanding of the FGS risk model is needed to appreciate the following mathematical illustration.

FGS Risk Model
The simplified FGS risk model shown in Figure 1 has three conditional branches and four consequence outcomes. The definition of each conditional branch is as follows:

Detection coverage =
   probability that a gas leak will be detected
FGS availability =
   1 – (PFDsensor + PFDlogic_solver + PFDfinal_element)
Mitigation effectiveness = probability that the consequence of the gas leak will be mitigated
[Note: Mitigation effectiveness is not covered in this article. It is assumed to be perfect (=1) in all calculations.]


Click chart to enlarge | Figure 1: FGS risk model

The combinational probabilities of all 3 conditional branches will produce 4 likelihood outcomes and the product of each likelihood outcome with its corresponding Consequence will in turn produce a consequence contribution factor. The subsequent summation of all contribution factors will result in a Weighted Average Consequence (CWA) factor.
[Note: The consequence of an event is typically determined separately by a quantitative consequence analysis which is outside the scope of this article. For simplicity, the consequence factor in Figure 1 is shown as either 1 or 0].

The Residual Risk (RR) of the allocated process risk (after the inclusion of a FGS) is then calculated by:
RR = CWA x FU (where FU = unmitigated frequency of gas leak per year)

In the example shown as Figure 1, it is assumed that:
The gas sensor will definitely detect the gas leak
  (i.e. detection coverage = 1);
The FGS hardware is constantly available
  (i.e. FGS availability = 1);
The risk introduced by the gas leak can be mitigated
   perfectly (i.e. mitigation effectiveness = 1).

In this ‘perfect response scenario’, CWA= 0, and RR = 0, which means that the process risk allocated to the FGS is completely mitigated by the FGS with no residual risk.

With this understanding, we can now look at some scenario based illustrations to observe the impact of detection coverage on the SIL rating of gas detection related SIF.

Continue SIL article...

Download PDF copy of SIL article