Socket Welded versus Threaded connections in Process Plants 1

[hull6851]

thread404-358123

I am starting this thread to ask for some guidance after reading another thread:

http://www.eng-tips.com/viewthread.cfm?qid=358123

I fully understand that there are differing *opinions* on when it is appropriate and/or cost-effective to use threaded connections or socket welded connections in process plants (for example, when connecting a drain or vent valve or an instrument root valve to a process line). I am NOT interested in hearing more opinions, as interesting as they may be, because I think those opinions are sufficiently documented elsewhere and vary widely.

What I would like to know is if anyone is aware of any sort of guidance from an industry group/society/association/publication that might be considered standard practice. As an engineer with relatively little personal experience in this area, and with so many varying opinions out there, I'd be appreciative if I could be pointed to a somewhat unbiased and objective resource based on facts, research and data, rather than on opinion, that could be used to help me in determining the proper times to use or avoid threads on connections. I know that certain companies have their own standards that also vary widely. In my small company, I am being tasked with creating our company's standard, so I'd like to be able to have a basis for whatever I decide to propose. Folks in my company have varying opinions also, so we don't have much of a consensus here either.

Any help is appreciated. And if you must share your opinion, I guess that's fine as long as you have some basis for it.

Thank you very much,
Don

 

[moltenmetal]

B31.3 permits threaded joints in normal fluid service. What more do you want to know?

 

[hull6851]

I guess I was hoping that there would be a specification that gave guidance with regard to when (or if) threads are ever inappropriate to use. Am I right to interpret B31.3 that threads are appropriate to use in all temperatures and pressures?

With all of the varying opinions on this board, in my office, and in customer's specs, I figured that those opinions must have some basis. If the concerns that people have with threads are really valid, I would expect that perhaps an industry specification would have at least mentioned a caution against threads in some circumstances. But maybe I'm wrong.

Thanks,
Don

 

[BrianPetersen]

That spec pertains to "normal fluid service". I'm not the process-plant expert that you are looking for, but I do know that hydrogen and helium will leak through anything that gives them the opportunity. Some temperature and pressure combinations might be outside the bounds of what would be considered "normal fluid service".

 

[LittleInch]

I kind of think you've answered your own question. Given the wide variety of opinions, it should be fairly clear that there are no accepted guidelines or publication like the one you are looking for. Unfortunately I think you're wrong and that no such publication exists.

Given the significant number of variations and inputs to this - fluid type, pressure, temperature, vibration, consequence of failure, past operating history, personal preference of the person checking / approving to name a few, the best you can do is devise a flow chart / decision tree type of thing to guide people appropriately.

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

 

[hull6851]

Thank you all for the guidance. It really is unfortunate that there is not an authoritative document on this issue that could be used to standardize processing plants/equipment. At the very least, it would make the meetings I am involved in on this topic much shorter. All the best!

 

[Tmoose]

I don't have access to the API specs any more, but have always been impressed with their (sometimes excessive) thoroughness, but I am sure they have evolved with input from folks in industry that have suffered first hand experience with devilish details.

http://www.api.org/~/media/files/pu...catalog/2016 api catalog_full final.pdf?la=en

 

[moltenmetal]

The reason I provided the reference to B31.3 is that the OP specifically said "objective/3rd party only, no opinion". The selection of which subset of the totality of types of joints permitted by a code for a particular fluid service, is suitable for a particular application, is inherently a matter of opinion, even if somebody at PIP is the one whose opinion is being offered. Selection of which joint type is appropriate depends on the nature of the fluid, the size of the joint, the design life, exterior and interior corrosion conditions, external loadings etc. etc. etc. etc. The only way you can codify such decisions is to massively overkill 90% of the applications by saying "no threads, period", which some firms find expedient to do. The sad fact is, by so doing you may not only be making things more expensive, you might actually be making them less safe.

By the way, in B31.3, if it's not nonhazardous category D, or lethal category M, or severe cyclic or high pressure category K, it's normal fluid service. Normal fluid service INCLUDES very, very hazardous services as well as some that are very low hazard class. It's too broad a category to be useful for much- but there are some joint types which are not permitted, or permitted only with "safeguarding" etc. That alone should tell you something.

 

[hull6851]

Thank you all very much for your responses. After more research that was encouraged and inspired by your responses, I am becoming more familiar with the documents that contain industry guidance/requirements for end connections including threading and welding. Perhaps my question was asked out of ignorance or unfamiliarity with these documents.

As moltenmetal suggests, ASME B31.3, paragraph 314, generally allows threaded joints in Category D Fluid Service and Normal Fluid Service (as defined below), with a statement to avoid threads in any service where crevice corrosion, severe erosion, or cyclic loading may occur. Interestingly, it appears the B31.3 has a bit of a gray area between Category M Fluid Service and Normal Fluid Service (see 300.2.b.2).

For those interested in what PIP has to say (debatably an "opinion"), PIP PNC00005 essentially recommends that threads only be used in Category D Fluid Service where accelerated crevice corrosion won't occur, but suggests threads are appropriate in connections after/downstream of a welded or flanged block valve to instruments, at equipment connections, and for terminal vent and drain connections (for Normal Fluid Service). PIP PNC00005 is then essentially more restrictive than ASME B31.3 with regard to thread use.

From ASME B31.3-2014, Section 300.2 (Definitions):
Fluid service: a general term concerning the application of a piping system, considering the combination of fluid properties, operating conditions, and other factors that establish the basis for design of the piping system. See Appendix M (Guide to Classifying Fluid Services).

(a) Category D Fluid Service: a fluid service in which all of the following apply:
(1) the fluid handled is nonflammable, nontoxic, and not damaging to human tissues as defined in para. 300.2
(2) the design gage pressure does not exceed 1035 kPa (150 psi)
(3) the design temperature is not greater than 186°C (366°F)
(4) the fluid temperature caused by anything other than atmospheric conditions is not less than −29°C (−20°F)

(b) Category M Fluid Service: a fluid service in which both of the following apply:
(1) the fluid is so highly toxic that a single exposure to a very small quantity of the fluid, caused by leakage, can produce serious irreversible harm to persons on breathing or bodily contact, even when prompt
restorative measures are taken
(2) after consideration of piping design, experience, service conditions, and location, the owner determines that the requirements for Normal Fluid Service do not sufficiently provide the leak tightness required
to protect personnel from exposure

(c) Elevated Temperature Fluid Service: a fluid service in which the piping metal temperature is sustained equal to or greater than Tcr as defined in Table 302.3.5, General Note (b).

(d) High Pressure Fluid Service: a fluid service for which the owner specifies the use of Chapter IX for piping design and construction; see also para. K300.

(e) High Purity Fluid Service: a fluid service that requires alternative methods of fabrication, inspection, examination, and testing not covered elsewhere in the Code, with the intent to produce a controlled level
of cleanness. The term thus applies to piping systems defined for other purposes as high purity, ultra high purity, hygienic, or aseptic.

(f) Normal Fluid Service: a fluid service pertaining to most piping covered by this Code, i.e., not subject to the rules for Category D, Category M, Elevated Temperature, High Pressure, or High Purity Fluid Service.

 

[racookpe1978]

No, you are again trying to "write the book" from "looking at the (previous) books".

Don't do that.

Yes, threaded joints are "cheaper" and (sometimes) they are even "faster" than socket-weld joints.
BUT.
Both induce "crud traps" which will allow contaminates to remain the pipe even after dozens of flushes, which trap residue after cooling or corrosion "pops" off wall deposits on the upstream pipes. Threaded pipes cannot be precisely predicted for section length for initial assembly nor for re-assembly, and thus you need "loose" connections like unions. Thus, more joints, more joints that are very likely to leak, more joints and fittings that need to be purchased and made-up. More joints that have to to be made up precisely and properly after EVERY maintenance action or outage period.

A socket weld is accepted for most services, is simple to make up and fit and to test with NDE and with the pressure tests. They allow easily welded backed joints with almost always simpler weld qualifications than butt-welds. They allow a modest amount of mis-alignment and mis-cutting to be accepted without rework and waste so the pipe fab is faster than butt-welding. But they too are not acceptable for all applications.

 

[Duwe6]

PIP specs are the 'opinion' of major industrial customers. This informed position was arrived at by looking at millions* of connections in-service for many decades. Ignoring PIP's is akin to trying to reinvent the wheel -- it is possible you will succeed, but incredibly improbable.

*'millions' is not hyperbole. Think if the number of valves, tees, flanges, and direction changes in one line in a process unit, multiply by the number of pipe circuits [fluids] in that unit, Now multiply by the number of units in a refinery, power plant or chemical plant. Now look up how many refineries, chem plants, and power plants exist in the USA.