Private Forums
Casing failure
16 September 2014

Anyone have a similar experience or possible explanation for the following:  Trying to avoid leaving this an oilfield mystery.

13â…œ"68# N-80 BTC casing ran on semisub to 4350' in vertical hole section.  Landed out and cemented with no issues.  Bumped plug and began pressuring to planned casing test of 3500 psi.  At 2100 psi the string jumped and pressure went to 0 psi.  POH and found pin on 15”² pup joint below SS hanger had released from BTC coupling on 1st full joint of casing. 

  1. Several observers confirmed connection was made up within 1/16" of base of triangle with 12,000 ft lbs torque.  Weatherford Lubeseal thread compound used.
  2. Calculated tensile load on connection at time of failure was 475 kips.  Conn rating is 1586 kips.
  3. Dimensional check of pin and coupling were within API spec (Both pin and coupling were recovered and sent in for inspection)
  4. Multiple calibrated gauges confirmed pressure did not exceed 2100 psi.  Tong torque gages were calibrated.
  5. Pin and coupling had minimal thread damage.  No stripped threads.
  6. Pup joint was cut from joint of same batch of casing run in well.  Inspection after manufacture within API spec.
  7. Plan to send pin and coupling for failure analysis.
Any recommendation for service provider?
12 answer(s)
fmarkus
Drilling Engineer
PT Saka Energi Indonesia
Total Posts: 2
Join Date: 13/11/14

Exactly the same like what I encountered before...specially when I read that casing string jumped after bumped plug :). Hopefully this wont be your mystery of oilfield again....

In my case, casing string parted from the bottom of ECP which we incorporated into the casing string. No sign of any damage whatsoever on the bottom connection of the ECP.

After incident we sent that ECP to one of the machines shops who has a BTC gauge and that machine shop confirmed that the BTC was out of spec. Well... you can figure out the rest of story yourself. 

Visually that BTC connection has no different with other BTC connections.

PaulHowlett
CEO
Sudelac
Total Posts: 90
Join Date: 10/04/08
Was there any conclusions to the investigation of the casing separating?
mmrgpc
Sr. Drilling Engineer
Kuwait Oil Company
Total Posts: 4
Join Date: 09/11/14
The M/U torque not everything, did the recorded and checked the Jam operation profile for the particular cnnection??
Companyrep
Drilling Specialist/Well Engineer/Training Consultant
Kingdom Drilling
Total Posts: 360
Join Date: 10/01/05
Thomas,

i was going to state that likeliness is quality control of pup and if it was cut locally.

i had a similar experience with a 20" combined string that was lost when String was picked up out of slips. The string parting at one of the home made pups. Luckily we fished it with a spear over torqued the back up to assure it did not fall off and completed job. 

history of this was that make up looked ok etc. 

a quality control check of tools, thread gauges etc, cutting process may likely eek out the latent causes.
iankev04
Engineer
Aker Solutions
Total Posts: 2
Join Date: 18/09/14
Hi I found this discussion interesting.

I was working in oilfield design equipment manufacturer.

Lately I have question regarding make-up between two "different grade" casing.

Ex. a L-80 casing grade is made up to a P-110 coupling or hanger.

I look up to premium thread manufacturer catalog like VAM TOP, Tenaris, etc. They likely have thier on method on defining torque value for the "different grade" situation. Usually we have to follow the lower grade thread, which in Example above is L-80 grade.

However, The case is if we have to make up two different casing with BTC (Buttress Casing) Thread.
I didn't find any clue on how API specify "their rule" in making up. In API 5B, API 5B1 or API 5TP.

I do find about torque up common practice in wipertrip.com
However does not specify anything regarding different grade torque up.
Source:
www.wipertrip.com/casing-operations/casing.html
Do anyone in the forum have any experience regarding this?

Thank you
Documents uploaded by user:
make up diff grade.jpg
AndyPua
Drilling Consultant
PT Drilling Services
Total Posts: 62
Join Date: 15/09/14
No, I do not think that making up to Base of Diamond with 12k ft-lb torque is indication of cross threading or under torque.
Actually the diamond is the indication of the make-up. I was with an Internationl Oil Company before (I don't want to mention name in case I get into some legal problem). And after their research, they came out with the Base and Tip of the diamond as the guide for making up, as long as the Make Up torque is within the range of torque (12-15k ft-lb, I think).
I don't think I ever make it up to 15k ft-lb, except for Bakerlok joints.  
If the make up torque was up to 12k ft-lb after the first couple of thread, I would say it was cross thread. 
Even if it was under make-up by 3k ft-lb, the force of 258 kips (the pressure of 2100 psi x cross sectional area), would not strip the threads apart, since it was rated for over 1.5 million pounds for this string of casing. 
I had used Weatherford Lube Seal many times, and it was quite reliable. 
Can't think of anything wrong in what you were doing. 
I know API gauge does have a tolerance of 1/32 for the newly cut thread. But if you had make it up to slightly above Base of Diamond, what you had done should be right. 
Unless there is some internal vertical crack or something on the coupling as one gentleman suggested. Hopefully QC the coupling could give us a satisfying answer.
TEE-1961
Trainer
Addax Petroleum
Total Posts: 11
Join Date: 12/11/12
Thanks for all the comments.  Good to hear from you Fernando.  We have not concluded the investigation and it will take time to have a failure analysis completed.  Here are a few additional notes:
1) The average m/u torque for the first 12 joints was 10,000 ftlbs which is consistent with previous jobs
2) The pup joint pin was initially torqued to 10,000 ftlbs and found to be â…›" short of the base of triangle.  It was backed out, inspected, and made up to 12,000 ftlbs with 1/16" from base of triangle. This was considered acceptable and run in hole.
3) We did not pull the entire string so did not confirm integrity of wiper plugs.
4) The casing should not have been in compression at the point of failure.  Estimated at 475 kip tensile load considering string wt and pressure (2100 psi).
5) Pup joint was cut from full joint of casing.  Failed pin believed to be factory cut and not from machine shop.
6) Casing was from China.  I am not sure of specific supplier.
7) See attached picture of pin
Documents uploaded by user:
Csg hanger pup joint.JPG
mullenmf
Technical Services Representative
Hunting PLC
Total Posts: 5
Join Date: 20/04/09
Looking at your torque figure of 12k, I would say you were crossthreaded.what was the average torgue for the first 12 joints, for this size I would have said it should have been around 15k.
Fernando
Drilling Superintendent
SPREADAssociates
Total Posts: 3
Join Date: 09/09/13
Hi Thomas, I hope you are doing well,

It is interesting to realize the failure involves one connection localy cut which may suggest quality control issues.

Potential reasons for the failure to be considered among others could be the actual dimensions and roundeness of the casing used to localy cut the buttress connection and the quality control of the thread cutting process. Even consideing you used a piece of casing from the same batch run in the well and the API specs passed the inspection (unfortunately for buttress connection, many times it is not performed with the same level of diligence compared to more sophisticated connections); the actual depth of the threads could be compromised; I have seen measurable variations of the OD and ovality along the lenght of one joint, specificaly in 13â…œ" and above.

In the other hand, buttress cutting process and quality control tend to be considered with less attention compared to other connections; almost every single oilfield tubulars machine shop is licensed to cut buttress and not always within the required quality control.     
Augusto
Consultant [retired Shell staff]
SPREADAssociates
Total Posts: 245
Join Date: 02/09/05
Checking my data, most of the unzipping of Buttress Thread & Coupled connections, stemmed from "COMPRESSION". Are you sure, that for landing the casing it was not compressed?
The other failures I knew about were:
- Special Clearance Couplings, particularly with an "O" Ring.
- Couplings with obsolete tin plating.
Outside the 3 weaknesses mentioned above, I would check the Friction Correction Factor of the "running compound" used and eventual tapers of the two engaging threads.
Assume cross-threading is ruled out.
PaulHowlett
CEO
Sudelac
Total Posts: 90
Join Date: 10/04/08
Thomas,

If you recovered the whole string, was the plug still on the float collar, and did the casing pull wet as a result? Or had the plug or float collar failed at 2100pai? The sudden pressure drop of a plug or collar failure might explain the connection at the top releasing.......
whitehead
Senior Completions Engineer / Well Examiner
Engie
Total Posts: 13
Join Date: 06/06/11
Check the coupling is not micro-cracked vertically.
Jump to top of the page