ASME B31.3 piping, hydrostatic pressure test, loose blind flanges 7

[Esbjerg]

When we are manufacturing prefab pipe spools for modification projects (offshore oil/gas), these spools are normally pressure tested onshore in the workshop. Often all flanged connections are blinded off using temporary test flanges. And at least one or several of the blinds used during the test is a temporary test flange.

After being installed in the offshore plant, some of the branch connections (or sometimes also end of line) is blinded off using ASME B16.5 standard blind flanges. These are taken directly from stock - and has not been subject to any hydrostatic pressure test.

In the end, after the completion of the new pipe installation, the entire new piping system is subject to a sensitive leak test to 25% of the design pressure.

But we have recently had comments from an Inspector telling us, that the loose blind flanges taken from stock and being installed as part of the piping hook-up should be subject to a pressure test to 1.5 times the design pressure.

As the written text is in B31.3, the Inspector is probably right. I however do not think it is the intentions of B31.3 to hydrostatically test a standard blind flange to 1.5 times the design pressure. The important issue here must be the leak test of the final flange joinings.

I would be happy to receive comments and also to hear about "common practice" in this field.

Thanks in advance

 

[rneill]

I agree with the inspector. I do believe it is the intent of code that all components used in a piping system are to be properly tested in accordance with code requirements. I don't really see basis for exempting blind flanges.

Note: I routinely apply the same criteria to flanged valves as well. I either test them as part of the assembled system or I have them tested prior to installation as any test completed by the Manufacturer does not fully satisfy the ASME B31.3 requirements.

Having said this, it is fairly common practice for people to bolt on blind flanges or valves, or install blanks, without properly testing those components.

You may want to review Clause 345.2.3 of B31.3 as this language changed a bit in 2008 to clarify that piping components could be pretested and that if pretested, the flanged joints (when installed into the final system) need not be retested.

 

[Esbjerg]

Thanks for the valuable input, rneill. I will however still be happy for receiving any further input. Thanks in advance. For example also if someone has experienced B16.5 blind flanges that actually failed during pressure test (fracture, rupture, etc.).
Best regards Kristian

 

[lsone]

what ansi rating are the flanges? What pipe size and schedule? What is the design pressure? What are you testing the line to?

thank you

 

[Esbjerg]

Hello Isone,
It is a very generic question what I initiated. Basically we talk about all sizes from 1" to 20" and with pressure rating 150# up to 2500#.
Material wise we talk about LTCS (A333, A350LF2, 316L, 22% Cr duplex - and in some few cases 25% Cr super duplex).
The spools are typically operating at design temperatures of max. 50 deg. C, so we are testing them to 1.5 times the design pressure as per the piping specification.
Regards Kristian

 

[stanweld]

If you are not subsequently full pressure testing (1.5 X design) the installed system, the pipe spools together with the test blinds require 1.5 X design pressure testing.

 

[Esbjerg]

@Stanweld: Our problem is mainly that the test blinds used during the onshore workshop pressure test of separate piping spools are not those we are using for the final installation in the piping system offshore. The blinds being installed during final hook-up are new flanges directly from stock, and they were not part of the onshore pressure test to 1.5 times design pressure.

 

[lsone]

I mis read your post. Missed the "after installed offshore" and thought you were doing your hydrotests to 25% design and was worried.

For all of the piping I have personally dealt with, I would not accept it. It is not a "proven connection". Until that bolted connection is tested to 1.5x the design pressure, it is not Proven.

 

[Esbjerg]

@Isone: Thanks for the opinion. The pipe spools are hydrostatically tested to 1.5 times design pressure. However except the blind flanges which will be part of the final installation offshore. After that final assembly no hydrotest is done - just a sensitive leak test using nitrogen-helium as tracer.

 

[stanweld]

Does your test comply with 345.2.3 (b)? If not, the actual installed blinds need to be tested as they, together with the pipe spools, comprise the "piping system" defined in 300.2 as, "interconnected piping subject to the same set or sets of design conditions."

 

[Esbjerg]

@Stanweld. Qoote from B31.3 (2008), sec. 345.2.3(b): Flanged loinfs. Flanged joints used to connect piping components and subassemblies that have previously been tested, and flanged joints at which a blank or blind is used to isolate equipment or other piping during a test, need not be leak tested in accordance with para. 345.1.

I think our problem in relation to this para is that our blind flanges (that was not pressure tested) will be a piping component being connected to the allready tested WN flange on the piping. It is correct, that if the blind flange had been pressure tested separately, we could connect the pipe spool and the blind flange without any problems (without problems in relation to the Inspector).

However, does it make sense at all to make a separate pressure test to 1.5 times design pressure of a loose B16.5 standard flange?

It would be a "strength test", but here we have virgin, forged standard flangesw. I have never heard about such ones fail mechanically during a pressure test. I also have not been able to find any reporting on such ones failing.

 

[XL83NL]

I partially agree with stanweld, however;
Which part of the Code describes those 'new' blind flanges have to be rehydroed? This may well be a case for an interpretation..

Apart from whether or not it's good engineering practice to do a new off-shore hydro, strictly speaking you may argue with the Inspector that B31.3 para 345.1 refers to piping systems; the definition of piping systems in para 300.2 reads

piping system: interconnected piping subject to the same set or sets of design conditions.

This refers to piping:

piping: assemblies of piping components used to convey, distribute, mix, separate, discharge, meter, control, or snub fluid flows. Piping also includes pipe-supporting elements, but does not include support structures, such as building frames, bents, foundations, or any equipment excluded from this Code (see para. 300.1.3).

So, piping on its term refers to assemblies of piping components.
Flanges, as per the Code definition, are piping components, of course.

Is a blind flange an assembly? No.
Is a blind flange + bolts + gasket + other mating flange (thus a bolted flange joint assembly) an assembly in view of above definiton from B31.3? Probably yes.
I think your case is somewhere in the middle, as part of the assembly has been tested.

Consider the following case. You have 2 pipe spools (A1 & A2), of the same size, design conditions, flange types (say WNRF with SPWND's), same test pressure, etc. To hydrotest them and save time, you make a 'temporary piping system' (say system A) in the shop by hooking up both spools, do the hydro, see how they pass, disassemble them, and install em in the plant. You do the same for spools B1 & B2 (which together make system B), C1 & C2 & C3, D...., etc ..... All these spools in view of this case, have flanged end connections.
Now, all spools are installed at other locations in the plant, compared to how they were hooked up for the hydro; say e.g. A1 is connected to B2, B2 to C3, and C3 to A2, etc. Now, do you have to re-hydro everything all over again, just because the bolted flange joint assemblies during hydro are 'different' compared to the installed condition? I dont think so (but I may be wrong due to my little experience).
This example kind of is the same as what you're dealing with, I think.

Now, I wouldnt agree with the inspector without any discussion or reasoning from his side.
If you have reasons to doubt the integrity of the flanged assembly (i.e. the new blind flange), and are sure the Code requires you to re-do the hydro? Do it.
If you dont have reasons to doubt the joint integrity, and believe the Code doesnt require you to do so, but the inspector does have good reason to re-do the hydro? Do it.
Otherwise, dont get overruled just like that, just because he wants so. Even though he may have some 'rights' based on the Code, doesnt mean he's always right.

Ask him why he thinks it has to be done again.

 

[Kiwi2671]

esbjerg,
It is too late now but you will just have to ensure you do not use temporary blinds in place of permanent in future.
If you have an isometric to be fabricated it will show you where permanent blinds are to be installed so you just need to ensure they are installed at the fabrication stage.
Your purchasing department will (or should) look at the material list on the isometric and see eg.2 x W/N Flanges, 2 x elbows, 1 length of pipe, 2 x gaskets, 16 x studs, 32 x nuts and 1 x BLIND FLANGE.
Put it all together with a temporary blind flange on the opposite end to the permanent blind flange and voila! - a code compliant spool piece.
Cheers,
Kiwi

 

[gr2vessels]

esbjerg is right, you don't hydrotest a new blind flange, be that a spare or a replacement blind flange, because it's coming with all the test certificates anyway. As far as testing concerns, you test the spool, not the temporary closure, hence anything can be used as blind flange as long as it does not affect adversely the spool. After installation, do the leak test without re-hydrotesting the system for a loose blind, already certified for service.
Cheers,
gr2vessels

 

[moltenmetal]

The way we interpret B31.3 345.2.3(b), we do not subject flanged in-line instrumentation, valves or pumps to the hydrotest as a matter of course. The hydrotest in our view is intended to prove the methods used to join the pipe and fittings to one another (i.e. threading, welding or brazing), rather than to establish the leak tightness of flange gaskets or to prove the strength of flange studs or blinds etc. Accordingly, we would not expect a flanged joint to be re-tested merely to prove the rating of an unmodified B16.5 blind flange. If a client asked us to do so, we gladly would, but we would not support this as a requirement of B31.3. An assembled system leak test at design (relief) pressure is always recommended to our clients to verify such connections.

 

[lsone]

Consider the following case. You have 2 pipe spools (A1 & A2), of the same size, design conditions, flange types (say WNRF with SPWND's), same test pressure, etc. To hydrotest them and save time, you make a 'temporary piping system' (say system A) in the shop by hooking up both spools, do the hydro, see how they pass, disassemble them, and install em in the plant. You do the same for spools B1 & B2 (which together make system B), C1 & C2 & C3, D...., etc ..... All these spools in view of this case, have flanged end connections.
Now, all spools are installed at other locations in the plant, compared to how they were hooked up for the hydro; say e.g. A1 is connected to B2, B2 to C3, and C3 to A2, etc. Now, do you have to re-hydro everything all over again, just because the bolted flange joint assemblies during hydro are 'different' compared to the installed condition? I dont think so (but I may be wrong due to my little experience).
This example kind of is the same as what you're dealing with, I think.

I will look into this. I understand your point. completely. believe it is not a code requirement but more so a client request. If they are willing to pay, go for it.

 

[moltenmetal]

We'd consider that 326.3 makes it pretty clear that the testing requirements of B31.3 do not apply to standard components such as B16.5 blind flanges. Clients are of course free to ask you to test anything they want tested, or to exceed the code requirements in any other way they choose, as long as they make that clear up front and are willing to pay for it.

 

[stanweld]

For the literally hundreds of flanges meeting B16.5 which have leaked during hydrostatic testing, I fail to find great relevance in a manufacturer's B16.5 certification thereof, especially since hydrostaic testing of flanges is excluded therein. I have seen failures in Italian, Chinese, Korean and US made flanges all purportedly meeting B16.5. I will also state that of those failed flanges only one or two involved blind flanges, all of which had been weld repaired (hidden from the purchaser). The new blind is permanently installed, is a pressure part and has never been subject to testing other than that defined in the base material specification. Falsified documentation is and has been a reality. As an Owner, why would I jeopardize an off shore installation?

While the sensitive leak test of 345.8 is not specifically defined in 345.1, I believe that it may be germane to the case described by Esbjerg provided the Owner accepts.

 

[moltenmetal]

stanweld: I too have seen flanged joints leak during hydrostatic testing, but I have never seen one FAIL.

Your caveats may be excellent reasons for an owner to insist on testing of a scope greater than that of 345.1's required leak test. However, I still don't think that B31.3 requires users to verify by means of hydrotesting or other tests per 345.1, the strength of blind flanges, flanged valves, flanged instruments, pump housings or other such components which have been either previously tested or otherwise certified to standards accepted by B31.3. Do others disagree? I've read the code interpretations as best I could and haven't found anything which would lead me to another conclusion yet.

 

[stanweld]

I have seen them fail; that is, leaking through the weld neck flange wall - not at the joint. I replaced over 600 flanges on one project.