Our company at the moment is having difficulties in organising the amount of SIF testing that we need to do. For example we have 23 SIF testing to do on a boiler and it is not efficient to bring the boiler 23 times and trip it 23 times.
Here I want to put a discussion out there for people who are familiar with the SIF testing process.
Personally by testing the following few things I think it is sufficient to say that the SIF test is done:
- Simulation test on the code for the SIF functionality. This should really need to be done only once at the very start (e.g. during commissioning) and if the version of the code has not been changed, then this should not need to be test it again, however if it has, then this simulation test should be done when it is due for SIF testing.
- Testing initiating devices, by simulating the field device and confirming the reading in the PLC.
- Testing the end devices, by simulating the output from the PLC and checking the device output in the field.
The usual way to do a SIF testing is to treat the whole system as a black box and simulate the initiating device to get the desire trip event. Here I am wondering if the above method is sufficient to be consider as a SIF test.
All discussion are welcome, but I am particular interested in what other companies now do to make the SIF testing a little more efficient.