Blind flanges for test

[Heating88]

Hi folks,
I am reviewing a situation at work where we now issue identity numbers to blind flanges (ANSI B16.5) which we use to seal Raised Face Slip On flanges during hydrotest.

We typically end up with the test pressure equalling 60 to 70 percent if the flange rating during the tests. Although we do not expose the flanges to very high pressures we are aware of the cyclic nature of these repeated tests on different vessels.

We collect data and record the number of times each flange is used so we can calculate when we should be reviewing the flanges for surface detection inspections to ensure there are no sites for initiation of failure. The flanges are cleaned and given a visual inspection before and after every use.
My question is as follows. Is there any guidance or recommendations as to how many times a blind flange should be put through the cyclic nature of pressurisation or are we being overcautious?
We do not expose the flanges to pressures above their official rating in ANSI for Pressure and temperature. However, we are mindful of the number of incidents and fatalities from test blind failures and we just want yo adopt a safe approach to this topic.
Any guidance on these blind flanges (and repeat use of Bolting for tests) would be greatly appreciated.

 

[XL83NL]

My gut feeling; assuming you're hydroing here (and not pneu-testing) and would (max.) use every blind flange once a day, for 20 yrs, you roughly end up with 7000 cycles. Thats the limit in B31.3 for not requiring to include a stress range factor.
If youre really uncomfortable with that, send out all blind flanges for PT examination once a year, and increase that interval once you start finding small surface cracks.

Huub
- You never get what you expect, you only get what you inspect.

 

[LittleInch]

Interesting question.

ASME (not ANSI) B16.5 doesn't mention fatigue or crack checking at all.

My view is that if you're operating at below the maximum working pressure for the blinds then your number of cycles is essentially unlimited, though you may want to look at fatigue cycles from either a pressure vessel code or the limit noted above ( not sure which clause that is).

A bigger issue for me would be the number of times any flange was subjected to a test above the MWP as you could easily test piping sections to 1.5 x MWP in a hydrotest. I would limit those to maybe 10 or 20 max before checking them for cracks and overstress.

I've always argued for test ends and test manifolds that the test pressure should be looked as the working pressure for testing components used time and time again. For blind flanges you don't have the option to use the next flange rating up as the bolt holes wouldn't match so you're limited to just not using them lots of times above their MWP.

SO I think it's good that you perform this record keeping.

Bolting is more difficult as it is easier to approach yield on a bolt without noticing as someone tightens up a bolt to stop the flange leaking. I doubt you use hydraulic bolt tensioners so the risk is much higher. Torque limits are better, but dependant on certain friction values which may not be there is the bolts are well greased. So bolts I think have a much shorter service life of maybe 25 or 30 tests?

All a bit of judgement really.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

Visual inspection is sufficient.
Never use bolts, use STUDS.
You can also check the flange and studs for permanent deformation.

Regards

 

[GD2]

As long as the test pressure is below the yield strength of the flange material, you should be seeing only elastic behavior and shouldn't be seeing any permanent deformation of the flange. When you say the test pressure is @60-70%of the flange rating, you will not exceed the yield strength.

The possible failure could be by low cycle fatigue which could be very difficult to evaluate. Just keep an eye on permanent deformation and initiation of any crack.

GDD
Canada

 

[GD2]

Here is a small simple calculation I will show where the flange stress will be in comparison to SMYS. Let's assume the test is done at room temp of 38 degree C.
Test Pressure = 1.5 X Max allowable pressure at 38C
Test Pressure = (1.5 x 0.7(0.60 SMYS), assumed 0.60SMYS is designed max working pressure.
Test pressure = 0.63SMYS.

It can be seen above that, if you don't overpressure, you will be using the flange in the elastic limit. No fatigue failure is anticipated.

But if you are unsure or the controls are not tight and may exceed the yield strain, a few thousand uses/cycles, can cause it a fatigue failure. Keep an eye on the deformation and cracks.

GDD
Canada

 

[mk3223]

We typically end up with the test pressure equalling 60 to 70 percent OF the flange rating during the tests.

This doesn't make sense for "protecting" the blind flange in the pressure test or the "cycling nature" usage for the testing.

With the safety factor designed for the specific material, the blind flange should be okay as long as there is no damage or corrosion presented on the part. On the other hand, the flange bolting used for the testings should be recorded and replaced as necessary.