iec 61508 PFD/PFH?

[jussu]

Hello!
I have so called "continuous process". I'm trying to calculate failure probability of safety function. Process safety function demand is estimated 0.3/year. Does the term system equivalent down time"TGE" in PFD calculation means that for example if i have measurement combination 2oo3 and it is down so my whole process is down. So is it wrong to use PFD calculation and i have to use PFH calculation despite safety function demand says use PFD calculation? I have to think whole process not only demand of safety function, correct?
-jussu

 

[ihg]

Could you expand a little? If you are using a 2oo3 voting system with transmitters, the MTTFS is out in the hundreds to thousands of year range (3 Rosemount 3051S units -> MTTFS of 3311 years, with PFDave 2.6E-14). If you take one of the transmitters off line to check it, you dont take down the safety system - if you are being fussy, you get the logic to switch to 1oo2 on a "bad measurement" from the third transmitter or while it is signalled off-line. A 1oo2 has close to the same PFDave as a 2oo3, it just isn't as reliable for "on line" operation.
With a 0.3/year demand rate you are covered for >0.6/year prooftest for "low demand". The limiting factor is probably the final element prooftesting.