Proof Testing of Safety Instrumented Functions: A Beginners Guide (Part 2)
By Dave Green, Engineering Manager, Engineering Safety Consultants Ltd
Following part 1 of our blog on Proof Testing (Proof Testing of Safety Instrumented Functions: A Beginners Guide (Part 1) where we discussed: 1. What is proof testing? 2. Why do we need to do it? 3. Are proof testing and functional testing the same thing?), part 2 goes into a little more detail on the potential issues and challenges in management of proof testing:
4. Can I conduct a Perfect Proof Test?
So, in a perfect world, we would conduct a proof test where EVERY SINGLE undetected failure would be revealed and the device would be restored to an ‘as new’ condition. Unfortunately, in the real world this rarely happens. The reasons for this varies, for example:
- The proof test may not be carried out under exact process conditions
- There is difficulty in testing (i.e. ensuring a valve is gas tight after operation)
- Ensuring that your reference meter (e.g. flow meter) is calibrated correctly
The effect of imperfect proof testing can be seen below (albeit slightly exaggerated!)
What you will find is that over time the impact of imperfect proof testing will result in a gradual increase of your PFD which may eventually lead to your system not meeting the PFD requirement and reaching an unacceptable level. The effect of an overhaul will ‘restore’ the device to as new condition and will reset the PFD to its original value.
5. What is Proof Test Coverage?
Proof Test Coverage (PTC) is the term given to the percentage of dangerous undetected failures that are exposed by a defined proof test procedure. For example, a PTC of 95% tells us that 95% of all possible undetected failures will be revealed during the proof test.
The proof test coverage should be accounted for in the PFD calculation to take in account the effects of Imperfect proof testing.
6. How do I conduct a proof test?
The effectiveness and efficiency of a proof test is only as good as the procedures which specify it. There are 5 key stages in the development and implementation of proof test procedures as below:
- Stage 1: Develop the procedures – Writing the procedure and reviewing the failure modes.
- Stage 2: Verify the procedures – Physically using the procedure to conduct a test
- Stage 3: Modify the procedures – updating the procedure following use, ensuring that the PTC is maintained
- Stage 4: Ensure those undertaking the testing are competent – especially important when new equipment to the plant is installed
- Stage 5: Review and amend once in use – ensuring that the procedure remains valid for the 20+ years of its usable life
Other practical issues such as visually inspecting components, reporting failures, ensuring systems are placed back into the operable state should be considered to ensure that the proof testing is an efficient and meaningful exercise.
7. How much will this cost!?
The answer… it depends! Proof testing can lead to significantly higher costs when you consider production loss, labour costs, cost of testing equipment etc., which is why it is important the correct testing interval is specified. Too often, and it will continually interrupt production and cause huge financial loss, and too infrequent, it may impact the functioning of the safety function. If planned well and is appropriate to do so, proof testing can be scheduled in during plant shutdown and other maintenance works to minimise production losses.
ESC have the knowledge and expertise to assist with producing proof test procedures which will provide the maximum proof test coverage as well as helping in implementing a formal competence assessment strategy, see details of our one day course on the Introduction to Safety Instrumented Systems for Technicians (IEC 61508/IEC 61511).