Normally the fail safe position for a field device such as pressure switch, level switch etc. is contact open to alarm. In some cases for shutdown systems and for Fire and Gas systems monitoring I have seen the inverse philosophy being applied. For the F&G systems I came acrossw this being done with an ABB August TMR and the vague explanation had to do with "line monitoring" .Any good instrumentation philosophers out there who can explain?

Rgds J.Wong

 

Responding to J.Wong's Sun, Apr 4, 11.26 am, query:

See my explanation in the Journal of the System Safety Society, "Hazard Prevention"

Title: "Probabilistic Risk Assessment of Equipment Safety Systems"

Issue: Vol 35, No. 3, 3rd Qtr, 1999.

If you are unable to obtain a copy, contact me!

Regards,
Phil Corso, PE {Boca Raton, FL, USA}
[[email protected]] ([email protected]) {[email protected]}

 

For "CONTACTS" this is completely wrong and its probably not what these guys where really doing.

I think they were using proximity switches instead of contacts. For many brands of proximity sensors like Pepprl+Fuchs etc an alarm can be logic "1" and the normal operating state can be logic "0".

In fact proximity switches are all about current and the current either increases when something is near or it decreases. If the absence of some metal is the error condition and the current increases when the metal is removed then this will give a logic "1" for the error condition.

 

These may be supervised inputs. It that case, there are resistors in the circuit at the sensor so that the controller can detect open and/or shorted conditions in addition to normal and alarm.

 

I think randall's reply is fairly correct... Normally all the F&G loops have an EOL resistances (End-Of-Line resistances) in their loop to detect any open or short condition of the sensor and terminating leads... this perhaps would not be possible with an Normally Closed Contact...
Rahul

 

Responding to Rahul's incredulity...

monitoring of normally-closed contacts has been in use for about 25 years now. Open-circuits, short-circuits, and ground-faults (earths) can be (simultaneously) detected!

Such monitoring is essential for effective improvement of SIS reliability. Another benefit is to eliminate... yes, I said eliminate... false-alarms. Furthermore, it can be done without resorting to very expensive TMR solutions!

For proof of my statement(s) read the paper referred to in an earlier posting on the subject.

Regards,
Phil Corso, PE {Boca Raton, FL, USA}
[[email protected]] ([email protected]) {[email protected]}