Pipe to Flange Welding 1

[svm81]

We are fabricating gas pipe lines as per ASME B31.3. There is a requirement of thermowell to a primary pipeline, so can we weld a socket weld flange directly to primary pipe, or can we weld a weldolet on a primary pipe and weld a flange directly to the weldolet. kindly advise.

 

[pennpiper]

My advice:

Do this = weld a weld-o-let on a primary pipe and weld a flange directly to the weld-o-let

Don't do this = weld a socket weld flange directly to primary pipe

 

[europipe]

And check the lenght of the thermowell (immersion lenght)
Maybe You have to add a piece of pipe between the weldolet and the flange.

 

[racookpe1978]

What if he tried to weld a RFWN flange to the primary pipe?

He'd have to re-prep the butt-weld end of the RFWN, but he would be able to eliminate one field weld joint.

 

[europipe]

Racookpe, the welder would love your idea

 

[C2it]

There is a component called a weldoflange which is an integral weldolet and a weld neck flange. Might be a good choice since the other options are rubbish. It would account for the thermowell length too.

SO flange .... ugh.
WN direct onto wall .. what about any area replacement reinforcement ??

 

[MJCronin]

Usually, in a real company, experienced pipline engineers develop a "Fabrication and Installation" Specification.

In this specification, all permissable installation details and configurations are specifically detailed. As the companies become more experienced, and since the senior engineers are retained for thier knowledge, the specifcation becomes more detailed and more useful

There should be no creative installation details left up to "field guesswork" ...... especially for something so common as a thermowell installation

http://www.red-bag.com/jcms/design-guides/365-bn-dg-k01-14-instrument-connections-on-vessels.html

 

[ripaldee]

The most conservative design is to weld a Weldolet to the main pipe and flange to the Weldolet.

 

[DLite30]

If you're installing a thermowell in a flanged connection, it's going to be fairly long and possibly susceptible to failure due to flow induced vibration (they're normally screwed into a threadolet).

I would consider that you ponder your application and determine if flow conditions can cause the thermowell to tune up.

Welker engineering could help you with that, or possibly the manufacturer of your specific component.

I've seen this several times, that the thermowell tuned up and cracked...funny thing is you could never design that to happen, but it happens by chance all the time.

 

[europipe]

DLite, strange story, I never saw a broken thermowell, made in China perhaps?

 

[DLite30]

I couldn't believe it the first time I saw it. They never completely broke, but cracked at the threads and leaked gas into the building, and subsequently issued a gas-detect alarm.

Operations installed another identical thermowell, and within a week, it failed the same way.

Ended up installing a shorter well one haven't had the issue again, 6 years later.

Same thing happened at two other stations within about 5 year period, all on new installations.

When you buy thermowells or insertion probes from Welker, that's one thing they like to check for good reason.

 

[DSB123]

DLiteE30,
It's not only Welker who checks for VIV of thermowells, any self respecting thermowell supplier does the same and specifies the unsupported length. If they did not then I would not buy a thermowell from them. However they need the information on the flow media from the purchaser to be correct in the first place. The Formulae are all in PTC 19.3

 

[786392]

Should there not be adequately selected material annular dampening/ sealing rings designed in place to take care of such situation?

If for merely shortening of length, as cited above accurate measurement gets sacrificed!

Best Regards
Qalander(Chem)

 

[Duwe6]

Industry standard at the refineries, powerplants, and chemical plants in my experience:

Weld in thermowells, directly into the pipe, on high-reliability systems like hot hydrocarbons that will aout-ingnite when it hits the air, or high-pressure steam, liquid chlorine and ammonia, etc.

I would consider a gas line to be in the above category, and would buy a TW made in the 1st-world and install it directly into the pipe. Angle the TW 45° off perpindicular, so that the tip is downstream from the hole. If the TW takes up more than a few percent of the internal volume of the pipe, use buttweld reducers to increase the pipe one or 2 sizes in the region of the TW. Completed line should look like a python that swallowed a small pig.

Only use O'lets and other fittings for what they were designed for. Using a socket-weld fitting as a buttweld item is poor engineering. If you really want a flange to put your TW into, ask your welder if using a RFWN buttweld flange will give him enough room to make a full-pennetration weld. If yes, 'go for it'. Otherwese build it using an O'let onto the pipe, whatever length of nipple required by the TW, and finish with a socket-weld / slipon / buttweld flange.


Unless your drawing has selected the wrong materials for the pipe and the TW, the TW will be in Good to As-New condition when the pipe has rusted/eroded to oblivion.