threaded joints for pipes and valves 8

[rvivac]

I work for a big Oil & Gas firm that prohibits the use of threaded joints for pipes carrying flammable fluids (gasoline, diesel, naphtha, natural gas, LPG or whatever) subjected to any pressure or any temperature.

For example, when we acquire a machine, we have to make them seal all the threaded connections with welding or any other method. This gives both the client and EPC staff lots of headaches. For example: when we have a PSV, it must be connected with flanges: threaded joints on them no way!

I believe that this is due to some accidents or maintenance problems that occurred in the past, but since there are this practice, the problem could be solved in another way.

My question is: how people work around the world with threaded joints on machines, PSV’s and so one: the maintenance problems and the on-field corrections or on-field improvements. Since the machines come with threaded joints, I believe that this is a common practice around the world but, for us it is like a taboo and I have no argument against that.

Thanks

 

[Mechomatic]

I am reading your question to be, "What is the reason for [your] company not allowing threaded joints on pipelines carrying flammable fluids in equipment, despite the fact that it is common practice (exhibited by the fact that so much equipment from other companies use threaded joints)?" If I am misinterpreting that, please correct me.

Some of the more experienced engineers can probably give a more definite response (thinking of LittleInch and zdas04), but I believe the resistance to using threaded joints comes down to a) corrosion, b) proper joint sealing, and c) ability to remake joints.

Cut threads diminish the wall thickness and expose additional surface area compared to the same length of unthreaded pipe, reducing the amount of corrosion that section of pipe can take before failing (higher surface area = more corrosion, less wall thickness = less material the corrosion has to chew through to reach a critical thickness for failure to occur). Even in materials not usually in danger of significant corrosion (stainless steeel, brass, etc), the surface stresses caused during the machining process to cut the threads can bring about stress corrosion cracking, which can eventually lead to sudden, catastrophic failure.

In addition to corrosion concerns, joints leaking due to poorly cut threads, improper use of thread sealant, or (see above) corrosion can become a significant concern. In some companies, I understand that it is common to backweld threaded joints to seal the connection against leaking. Some companies prohibit this practice because there is concern that the heat from welding will diminish the strength of the threads and - since the weld is not a load-bearing weld and may not be manufactured as such - the joint will eventually fail even though it ought to hold as a threaded joint.

Finally, threads can become damaged from mishandling and their lifespan and quality of seal can be reduced. If a connection has to be taken apart and remade any number of times, a flanged connection is preferable since you can just slip a new, inexpensive gasket between the faces when the connection is remade and know you have a good-as-new joint.

There's a possibility I'm putting too much or too little emphasis on something, though, so this is just my opinion as a young engineer. If some of the eng-tips champs weigh in, I'd take their word before mine.

 

[HPost]

It's quite straightforward, just specify parallel threads or flanges.
Aggressive or pressurised fluid systems should never have tapered threads.

 

[moltenmetal]

We've seen this many times before. It's a result of a company having a hard time establishing easy-to-understand but broader limits for threaded joints based on commonsense. It's easier to establish a total ban on NPT threads for hazardous services than it is to teach people the circumstances under which threads are acceptable. Particularly for instrument connections, these outright bans against NPT threads can lead to some very stupid looking, awkward and unnecessarily expensive assemblies.

Below 150# class, below 2" and below 300 F, in services with low corrosion risk, without excessive vibration or extremely rapid and frequent thermal cycles, and most importantly with the right sealant system, NPT threads are seldom a problem. Machined threads such as those on compression fittings and threaded valves (to say 1/2" NPT) can go a lot higher than 150# class and give long term reliable sealing as long as you are below 300 F and have properly applied a good sealant system.

While avoiding threads in run-of-pipe applications even within those limits might be justified, an outright ban against NPT threads under all circumstances for hazardous fluids absolutely is not. A few NPT branches, particularly for instrument connections? Usually a very low risk, unless the fluid is very hazardous.

As to your other points: you probably want flanges on your larger relief valves anyway, so you can take them off easily to bench calibrate them. Stainless steel unions are notoriously difficult to seal reliably, so we don't use them, whether threaded or socket-welded, so we prefer to use flanges.

For small lines, tubing and compression fittings do away with most of the leakage risk, make every joint a reliable union for maintenance, and take less labour than threading.

 

[zdas04]

As Engineers we try to find logical reasons for things, even when logic does not exist. I've dug into this issue (I've seen threaded connections on 10,000 psig wellheads that have lasted 50 years without issues, I keep wondering why I can't use threads on an ANSI 150 system at most companies these days) and the best I can tell:
[ul]
[li]Several threaded connections failed on compressor skids over a couple of years in the late 1990's[/li]
[li]Most of the failed threaded connections resulted in fires and a couple of them resulted in personal injury[/li]
[li]Companies needed a sound byte to deal with the issue and it became "no threaded connections on compressor skids"[/li]
[li]Then a threaded connection on a pump skid failed in crude-oil service and resulted in a fire and a spill[/li]
[li]Sound byte modified to "no threaded connections in vibrating service"[/li]
[li]Then a pig launcher came apart in mid-stream service and there were threaded connections (no casual linkage between the threads and the failure, but ok)[/li]
[li]Finally the sound byte was modified to "no threaded connections, anywhere, ever"[/li]
[li]Several SPE and ASME presentations were done on what a horrible thing NPT threads are[/li]
[li]All the majors banned them[/li]
[li]ASME B31.3 put in language that can be interpreted to ban threads (it can also be interpreted to not ban them)[/li]
[/ul]

When I look at the corrosion failures that I've investigated in my career (probably close to 100), failures in threads have been pretty rare. While cutting threads does remove metal mass, it also work hardens the steel. The corrosion modalities that I've experienced seem to not work real well on the hardened steel. I've seen a lot of surface corrosion (rust) in threaded connections since we can't effectively coat the threads, but pitting in threads has been pretty rare in my experience.

I owned the standard for this at BP in 2003 and I fought every attempt to restrict their use in 3-inch and smaller (4-inch just takes too much strength to make up properly and they don't get made up properly very often) onshore, upstream. I retired in September, 2003 and the new person came from a refinery that had banned threaded connections a couple of years earlier. My carefully protected standard was discarded before the ides of September, 2003.

I wrote an SPE paper recently that tries to explain some of this thought process. You might find it interesting.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[rvivac]

GTME12, moltenmetal and zdas04,

I really agree with you and I believe that the practice to seal the threaded joints is just a way to solve a problem easily. It’s easier to prohibit rather than teach the issues and cautions during assembly and maintenance.

The ASME B31.3 (2008), para. 314 is really not clear on determining the use or not of threaded joints. It only says when the use is definitely forbidden.

The question that still stands for me is why when we buy, for example: an equipment or machine, they send it to us with some threaded joints even if we specify "no threaded joints". This makes me argue the sealing threaded joints practice. Many times it is sure easier to remove a flanged joint than a threaded joint.

I suspect that the real problem is a project management issue: lack of communication between client, contractor and supplier.

I would like to know your opinion.

 

[LittleInch]

I think ZDAS04 has outlined the thinking and history very well.

My thoughts are that screwed connections are just too dependant on good workmanship - the fact that a single person can actually strip the joint is a vulnerability that is difficult to deny or monitor. Design has something to do with it as well - if you don't allow some sort of free rotation with a screwed union, you get fitters exerting excessive force so that either your elbow goes the right way or your pressure guage is facing you and not the wall. I accept that most threaded joints are made from petty tough and thick steel which can stand it, but sometimes it didn't and the inherent vulnerability of threaded joints compared to flanged joints combined with de-skilling of the guys installing this and loss of supervision has led us to the point we are now.

There have been also failures where excessive use of PTFE and other jointing compounds have resulted in pressure gauges etc flying off and causing injury, then leading to a ban on PTFE and a lot more leaking joints....

Maybe the way to go is socket weld or weld small diameter stubs and then weld on screwed unions (see attached), especially where make / break occurs on a frequent basis and a change of direction is involved. Would that joint be classified as a "threaded joint"? I don't like seal welds as its neither one thing nor another. If you're going to weld it then weld it properly, not this half hearted thing which often doesn't really increase the integrity of the joint at all.

The simple fact is that corporations and government safety organisations especially have simply become much more sensitive to risk of personnel injury and when they look at Root Cause Analysis for incidents, the screwed joint just doesn't stand up to scrutiny.

All just my opinion, but hope it helps in the debate.


My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[zdas04]

I prefer to use threads over flanges in 2-inch and I don't really hate them in 3-inch. I heard an argument against threads on code vessels last year that finally made some sense to me. It was pointed out to me that putting a 2-inch NPT thread on a code vessel runs the risk that if the threads get damaged then fixing the threads requires an R-stamp. That can be pretty hard to come by in many of the places I work. Since that conversation I have spec'd a flange with a threaded blind where I historically would have specified a 2-inch NPT collar since I can replace the blind without a code repair. Everywhere else I still try to spec threads in 2-inch.

David Simpson, PE
MuleShoe Engineering

Law is the common force organized to act as an obstacle of injustice Frédéric Bastiat

 

[moltenmetal]

Bad workmanship can and will cause serious problems with any method used to join components in pressure retaining service. Some people don't like threads merely because other than VT during assembly or leakage testing, there's no NDE you can do to assess the quality of the joint post assembly- but just how is that different from a socket weld?

NPT thread leakage and failures usually occur, in my experience, when the wrong sealant system is used, or when the fitter thinks NPT is a dryseal system and thinks that the solution to all problems is a bigger pipewrench. That can be addressed by training. I can see why some companies can't rely on the training of their suppliers to prevent problems, probably because their procurement people rather than their engineers are the ones making the supplier selections.

Tape only or paste only leads to trouble. Tape only can work but IS very dependent on good thread surface finish (i.e. no dull threading dies, plenty of lubricant while cutting), tape of the correct density, and correct application. Recent experience with anaerobic pipethread sealant only on larger threaded pipe (1.5" and 2") was a disaster- the tape does serve an important function. Tape PLUS paste, properly applied, has been proven to give years of reliable service, through repeated thermal cycles etc., when applied properly and kept within its temperature limits. Contrary to the assertions of some, paste WILL still cure properly if applied on top of tape. Max service temperature for the good pastes is about 175 C (350 F), and it is a mistake to consider those pastes "liquid Teflon"- you need to look at the chemical compatibility of the pastes no differently than you do when selecting any other material.

 

[rocketscientist]

Where I've worked in the past, socket-welds have replaced threading for pipe less than 2" in size. Another accepted approach is bent tubing but these often have the same problem as threaded. As for larger threaded pipe (>3"), I can't see it. Like ball and globe valves larger than 3" they are difficult to install and operate, but then, so are flanges. Lap flanges were once a common thing because they go around these company requirements and allowed less precise pipe construction.

 

[rvivac]

I can conclude from here that threaded joints are:

[ol 1]
[li]Directly dependent on workmanship: either on the assembly and in the maintenance;[/li]
[li]More dependent on fabrication issues and more difficult to be monitored in terms of quality;[/li]
[li]ASME 31.3 does not determine its use or not in most of the conditions;[/li]
[li]It may have higher costs of maintenance and monitoring than welded-flanged joints.[/li]
[/ol]

I still do not understand that some machine suppliers still supply some systems with threaded joints. I believe that it is due to miscommunication between all the parts.

Thanks for your knowledge sharing!