Thread unzippering due to excessive PTFE tape application

[johnchrc]

I have a premium downhole thread that was pressure tested with Test Cap but they put PTFE tape on to the crest of the thread (buttress type). At 12,500 psi the connection below apart and left a large hole and thread deformation along the pin. Mag particle inspection shows now cracks or flaws, the cert looks good for a 4140 110 MYS 36 Rc Low allow steel. Could the PTFE tape actually cause the Box and Pin to have excessive hoops stress and limited engagement of thread to cause part to fail?



- CJ

 

[Compositepro]

The wall thickness where the thread tore looks very thin for 12,500 psi. That would also contribute to the hoop buckling seen elsewhere. Lubrication due to the Teflon may be a factor but I doubt it was the cause.

 

[johnchrc]

It wan't the material. That telon tape must nave balled up and sealed along that thin section. Another Test Cap blew off after they did the same thing.

- CJ

 

[hydtools]

Run test without Teflon tape.

Ted

 

[IFRs]

Could they have over-tightened the pipe cap? With Teflon, the torque required to tighten goes down but sometimes the torque applied does not.

 

[johnchrc]

They should not have used Teflon tape to begin with. This thd has mtm seal and once you reach my loss torque spikes. I think they made this up with 60" until it would mu no more.

They should use rec. pipe dope and make to min. Specified torque and I believe there would be no issue. The Problem is they make and break test caps 30 to 40 times on different tools. Both materials are P110 grade. The pin is probably peened and then kendex is applied. In this case a second tool was made up to test two tools and they parted at the threa so I know the box was phosphated. Thread was Ultra SX with buttress type thread so unzipping is unusual and it happened twice. A test plug blew off another. I'm fairly certain the Teflon tape was the root cause. It was applied to the crest of the lower pin thread (2 step thread)
I have performed mag particle and no surface cracks. Thread vendor will check microstructure but certs are good and tool has never been downhole or tested other than pressure test.

I want to put TDC on test fixtures and torque to
200 ft-lbf over spike at mu loss. Seals will be engaged. Keep contact stress down and TDC help reduce gallon and extend life is test fixtures.

Ultimately an internal fixture that doesn't mu to this is required but management in Maple Leaf country will not let us. Says they are developing a test fixture. Still waiting...

- CJ

 

[tbuelna]

Looking at the picture, a couple things I noted were 1) a fairly large L/D ratio between the end of the thread and the location where the thread fractured, 2) a very thin wall section behind the external threads, 3) buckling of one side of the externally threaded section above the fracture location. The buckling would seem to indicate there might have been some bending applied to the thread joint.

A threaded connection having a large number of engaged thread pitches and a large pitch diameter can result in significant interference in the threads.

The PTFE tape used to seal pipe threads works by filling the small gap existing between the tips of the external threads and the roots of the internal threads.

 

[3DDave]

What does this phrase mean? I keep looking at it but can't decide what 60" and mu stand for.

"I think they made this up with 60" until it would mu no more"

 

[johnchrc]

It oilfield jargon for 60" pipe wrench and make-up loss.

This is not a standard buttress connection and does not use PTFE tape. It has metal to metal seal. The Teflon tape was applied to the crest of the thread at the thin section (large D/t) section so if the PTFE tape held then the thin thread section would see internal differential pressure, axial load from pressure area (small relative to load capacity of entire thread) and load from torquing up with Teflon filling void not made for containing PTFE tape.

The entire problem is we do not want to gall the casing threads so we ask for low torque connections (no interference) but vendors will not do it because of liability so we are going to hard face the plugs and torque until the thread reaches make-up loss and metal to metal seals are engaged.

The failure on this pin is unusual. Usually they unzip and pop off. Here you can see that a portion of the thread took an axial load that yielded the thread on either side of the buckled hole.

Material has been analyzed and is OK. Low alloy Steel with about 120 ksi yield and 36 RC hardness. Ultimate is near 140 ksi.

- CJ

 

[3DDave]

I just read up some on drilling terminology. Interesting.

If I'm seeing this right then the threads are tapered buttress and for no good reason the assemblers decided to use PTFE tape and then tightened it as much as they could with a 5ft pipe wrench. If the insertion distance (make-up?) was not decreased to reflect the tape thickness then a larger than expected hoop stress would result.

The teflon tape possibly lowered the friction in the joint enough that the pipe deformed from external compression. If it failed before the test started, it would leave a stress concentration on only one or two turns of thread at the exposed end of the connection to take all the load, hence the peculiar rupture.

Had the joint been undamaged before the pressure test, then fluid entering would have pressure balanced on the inside and outside surface of the pipe leaving no net force to deform it, though the additional load could have expanded to outer part (box?) causing a leak.

I think you described that the tape was applied to a small number of threads. This would also prevent the remaining threads from taking the load, causing a failure near where the tape was used.

You could try to assemble with PTFE tape, overtorque, and then disassemble to see what damage that does to see if my reasoning is correct. You might have to use a short section of pipe or a coupler and visually inspect or measure the pipe ID to determine if the pipe is deforming non-uniformly due to using PTFE.

I don't understand the use of tape at all in this application. Drill rigs don't use tape; at the best it would be time consuming and difficult to repeat, at worst people losing fingers applying it during assembly. The various pipe catalogs I just looked at indicated sealing occurs at the ends of the pipe (pins?) either as metal-to-metal face seals or with some compound that is applied.

My final thought on this is that conventional pipe wrenches apply a very large localized compression load that could also contribute to collapse of the pipe on the inside. You don't say where the loads are applied in assembling this.

 

[tbuelna]

Where is the metal-metal seal in relation to the thread joint shown in the picture? The external threads shown in the picture are definitely not buttress threads, and appear to be some unconventional type of thread. The crests of the external threads are severely truncated, and the thread roots seem to have a very wide flat with a small fillet radius. This would not seem to be a thread form that has good load capacity.

 

[3DDave]

tbuelna,

I have no idea of your experience in this, so this may be something you already know.

I had to Google some of the terms and look at diagrams about down-hole pipes and found the threads in the pictures here look very much like those in the literature from pipe suppliers. For example

http://www.stewarttubular.com/

An industry that has stabbing guides as a normal part of business has got to be an interesting one to work with.

 

[johnchrc]

3D Dave,

I think you hit on what happened. API thread use pipe dope to seal and even have Teflon seal rings that can be machined in buttress threads. They normally apply PTFE tape but only to API connections in the shop. PTFE tape can cause problems when you back off the joint because it can ball up and cause the thread to gall. On a rig using PTFE tape would not be practical or allowed because bits of PTFE in a well would could plug tools downhole, perforations and cause problems. The tape is never applied to premium threads and that is what happened here.

We are pressure testing assemblies in a shop prior to sending tools to the field. We would normally apply the appropriate pipe dope and torque to the triangle stamped on the pin thread and sealing would not be a problem. The pipe dope would seal in the space between the root and crest. The reason we don't is to prevent galling of our Test Cap and Plug.

When we make up the test plugs over and over onto casing threads they deform and wear and begin to gall. If that happens then we can ruin the tool we are testing. Th problem is casing threads are not made for repeated use. So, we get by by applying PTFE tape and torquing to a much lower torque to limit contact stress on flanks. However, you can only do this on API buttress, LTC and STC threads that rely on pipe dope to seal. Premium connections, like the one that failed only use dope to reduce friction during make-up and prevent galling. They have metal to metal seals and design thread interference into the thread so it requires torque to make-up so it will not back off down hole (some companies make low torque versions of thread without the interference so we can pressure test by pumping the metal to metal seal with about 50 ft-lbf torque and it will seal). We have no choice but to make these up completely (full make-up loss)but as the torque spike we can stop and the thread should seal without applying additional torque the thread vndors recommned for downhole use.

In this case, the Pressure Test bay crew made a mistake when they put PTFE tape on a premium connection like this. The tape definitely caused external compression on box and internal compression on pin. The tape probably roped up and started the localized deformation where the failure occurred. When 12,500 psi was applied, you then have an additional axial load too but applied non uniformly to the deformed threads. Because is sealed, there was differential pressure across part of the thin walled section of the pin thread and it burst (or parted in tension and burst) although it appears to be collapse. Pressure up to the seal would be balanced so there would be no external pressure differential.

Very good analysis for not knowing the subject. I'm hoping fo the thread vendor to provide their explanation but I'm 99% sure it was a result of the PTFE tape on the lower pin thread.

- CJ