Manway burst test/pressure rating

[arbor]

We are doing some burst tests for manway pressure rating. The maximum test pressure is usually limited by leaking from gasket before reaching the desired pressure. As a result, we will get a lower rating pressure for the manway although the steel components may be able to stand much higher pressure.
Is this right? Should the gasket be included in the test? Thank you,

 

[jehan17718]

arbor,

I am unsure as to the purpose of your test aka manway pressure rating. If you are using ASME B16.5 standard flanges, then the pressure ratings will govern and no additional calculations are required.

All I can say is that when one does a convention hydrotest of a pressure vessel it is an acceptable practise to insist that gaskets used in hydrotest will be identical (i.e. same "grade) to the service gaskets.

-jehan

 

[jte]

arbor-

How do you define "failure"?

If you define "failure" as "it failed to contain the internal pressure" then whether or not "the steel components may be able to stand much higher pressure" is irrelevant. Bolted flanged joints usually fail at a gasket sealing surface and, as such, yes, the gasket is an integral part of the system.

If you plan to explain to the client who is left with a leaking flange connection that the leaks are ok since the steel hasn't torn open yet, well, I guess that's another approach...

jt

 

[arbor]

jehan and jt, thank you for the replies. Sometimes, the pressure at which the test was stopped due to leaking is much higher than the desired maximum allowable pressure, maybe two or three times of it. So the hydrostatic test (1.3x) will pass if we could rate the vessel at the desired pressure. However, for a burst test, the factor is around 4. So the computed maximum allowable pressure cannot reach the desired pressure by using the pressure at leaking.

For example, if we want to rate a manway for 100psi. The required burst test pressure needs to be at least 400psi. However the test is stopped at 300psi due to leaking other than rupture of the manway. Then the computed maximum allowable pressure will only be 75psi by using the leaking pressure of 300psi.

I hope I can get some inputs for this kind of cases. Thank you,

arbor

 

[jte]

arbor-

Based on your stating that the hydro test factor is 1.3x, I'm presuming that you are working to ASME VIII Div. 1 (can you confirm?). Now, I'm still a bit unclear on whether you are fabricating an entire vessel or just a manway assembly similar to this:

http://www.reskem.com/pages/manways.php

If you are fabricating the manway assembly, you might find UG-11(c). Read footnote 6. Basically, you as the manufacturer define the part and state what the rating is. There's probably more to it; I don't get into this part of the code often.

If you are fabricating the entire vessel, including the manway, and the design is based on a burst test, then I can see that you would be experiencing difficulty sealing. If a higher rated standard flange is not feasible, you might consider using a pipe nozzle neck to a weld neck flange on a process inlet or outlet. Let's say that you only have a 6" pipe size requirement, use a 20" or so with a reducer. Now define the scope of VIII-1 in accordance with U-1(e) as ending at the nozzle neck (pipe) weld bevel where the WN flange will be welded on (see Interpretation VIII-1-86-193 if you have concerns about this). This is the first circ seam on the pipe which is not welded to the shell. Now for testing purposes, weld an end cap on and blow 'er up. Basically, this places the manway out of the scope of VIII and into B31.3 (or whatever flavor of piping code you need). But you do need access to the vessel, so you need to make sure that you have a removable pipe spool to accomplish this. A 20" WN flange welded to a 20"x6" reducer to a 6" WN flange which connects to your 6" process pipe. Now you are in compliance with UG-46(f)(7). I'm sure you'll get other opinions on this... I'll be interested in how this discussion goes.

jt

 

[jehan17718]

arbor,

You CANNOT "experimentally" rate a manway for a design pressure in the way you have described above.

It is obvious, that in the "vicinity" of the rated pressure i.e. flange MAWP, leakage through the flange will first generallly commence before actual destruction of the flange itself.

Like I have mentioned before, if you are using a standard ASME B16.5 standard flanges for the manway, then the pressure-temperature ratings of applicable "Group" will govern and no additional calculations are required.

HOwever, if your manway is a user defined one and vessel is designed to ASME VIII-1 Code then use Appendix 2. It will involve trial and error, back solving and iterations (depending upon your software used) to determine the manway's MAWP (which you refer to the manway rating).

I RE-STRESS THAT THE METHOD DESCRIBED IN YOUR POST ABOVE IS IN NEVER USED TO EXPERIMENTALLY RATE THE MANWAY OR FOR THAT MATTER ANY FLANGE.

-jehan

 

[jte]

jehan-

I don't mind being told (or shouted at) that I'm wrong. But I will request that you provide more backup to help me understand why you feel that I'm wrong. What part of the code is explicitly or implicly being violated? Do you feel that I've misinterpreted footnote 6 of UB-11(c)? Keep in mind that not all manways are round with flanges, gaskets, and studs/nuts like typical piping flanges. There are plenty of other sealing options.

arbor-

Well, it seems a bit more detail would be useful. Is your manway something like in my link above with no flange bolting - commonly used in some services - or is it a standard bolted flange with cover or a special bolted flanged joint which doesn't fit within VIII-1 App 2?

jt

 

[arbor]

Thanks again for the inputs. I got many valuable information.

We are a manufacturing company and we have designed and made standard prefabricated manways to ASME VIII-1 Code for more than 20 years as I know of. Most of them are stainless steel manways. The manways are similar to these:

http://www.tankcomponents.com/manways/tci_03_manway/index.htm.

The manway cover and collar are connected by lugs and bolts. The seal material is o-ring. I think the design of this falls into U-2.

I don’t know how the manway MAWP was originally determined. But we are doing some burst tests to rerate some manways now and most of tests were stopped due to gasket leaking before reaching the required test pressure for the original WAMP. So that the new MAWP is lower than the original one although the steel components look intact after the test. So I wonder if this method is appropriate.

We have also done some FEA work with no o-ring included. The FEA results show much higher pressure that the manway can stand (the bolt failed first finally). But the FEA can’t reflect how good the seal is. Right now, I personally prefer to use the FEA result for the MAWP as long as the hydrostatic test can pass. I would like to have your opinion.

Thinking of if we don’t have the resource of FEA brought me the original question. Can we still do burst test for the manway MAWP and what is the proper way to do it? I couldn’t think of any way to do the test without o-ring.

Thank you,

arbor

 

[arbor]

More information. The manway has no standard flange.
For the manways I am working on, there are lugs welded onto the cover and trunnion tubes welded onto the collar. The lugs and tubes are connected by bolts and rods. The o-ring is between the contact surfaces of cover and collar. The o-ring groove is on the cover.

Some manways are like the ones in jt's link. Thank you,

arbor