Private Forums
Pressure Testing a Corroded Pipe/Casing
12 June 2018

The company is planning to pressure test the A annulus. In 2017, Multi-barrier imaging (Empulse) was deployed to detect metal loss due to corrosion and metal wear. The maximum metal loss in specific intervals ranged between 13% and 18%.  One of the wells was found to have intensive and significant corrosions in 7” Production Tubing of 34% below casing tie back shoe.  The API SPEC 5CT (Specification for Casing and Tubing) specifies that the minimal permissible pipe wall thickness is 87.5 % of the nominal wall thickness, which in turn has a tolerance of-12.5 %. The metal loss higher than 12.5% requires that the casing is downgraded or corroded pipe replaced. How is the remaining strength of corroded pipe determined? I.e. to what proportion does the loss of metal reduce the strength of the casing when the metal loss is above the specified 12.5%?

Documents uploaded by user:
Fundamental Aspects of Casing Design.pdf
6 answer(s)
cmcandrew
SPREADAssociates
Total Posts: 23
Join Date: 16/08/10
David, I had been thinking the 12.5% tolerance was obsolete with new manufacturing processes as calipering new casing always gives ID very close to tolerance. Thanks for the explanation on "thin wall" and "thick wall" :  is that still the case in 2018 ? but it still doesn't reconcile calipered IDs  with the mass and OD tolerances... that I can see.

Tiny:"The API SPEC 5CT (Specification for Casing and Tubing) specifies that the minimal permissible pipe wall thickness is 87.5 % of the nominal wall thickness, which in turn has a tolerance of-12.5 %."  
I see that is taken verbatim from the attachment- not sure where attachment author got that from 5CT or on what he means !. 
RuariTruter
SPREADAssociates
Total Posts: 18
Join Date: 17/09/07
I agree with the response on Barlow et al but I think the best way to look at this is what you need the annulus test for and what is the minimum acceptable for safe operation of the well?  What is the well integrity requirement?

Calculation might be able to tell you what the rating should be but it won't take into account any unknows, for example a specific area of extreme corrosion that has been 'missed'.

  • Define why you are testing it - upcoming re-entry? Can you in fact live with not testing it?
  • Work out the minimal acceptable pressure test (With whatever safety factors you wish), as defined by your reason for the test.
  • Put together a test program that allows for stop and review points.
  • Detail the risk associated with a failure be that loss of production, loss of containment, requirement for immediate workover.  What is the failure mechanism and consequence?
Run a risk assessment with a couple of stakeholders (Including your friendly Well Examiner) and decide the way forward.



sandeepdhawan
Principal Well Engineer
WellPerform Aps
Total Posts: 7
Join Date: 17/05/13

Hi,

Supplementing to David and Doug's responses, additional points might help:

1. Was the pipe ordered with API tolerance? Meaning is the Kwall 12.5% or less?

2. Wall loss is at pipe body only or connections also? Connection wear and integrity assessment is another complex subject

3. Potential collapse risk of tubing w/ 34% wall loss. For collapse strength a linear deration can be considered.

4. If you are not able to figure out how logged WT (wall thickness) is/was computed, assume wall loss at minimum wall. Meaning add Kwall to reported wear (this is worst case) to check if the recalc. burst strength suffices.

Hope this helps!

whitehead
Senior Completions Engineer / Well Examiner
Engie
Total Posts: 13
Join Date: 06/06/11
Tiny, I can see your question is around an annulus test. I am assuming it is a tubing/casing test and your metal loss is in the tubing.

I agree with others that you need to understand how the log is interpreted. Is it based on a measured actual wall thickness or merely based on a nominal ID (such as a caliper log would be).

You ask how the strength (axial) of the pipe is affected, probably very little. Calculations for the collapse value however are quite complex (see http://petrowiki.org/Strength_of_casing_and_tubing)
for details.

I don't know your exact requirement, but I would be inclined to fill the annulus, check for losses. I assume there is no current leak from tubing to annulus and leave well alone until a leak manifests itself, then plan remedial action.





david.vavasseur
HPHT Drilling Engineering Team Lead
TOTAL
Total Posts: 25
Join Date: 22/07/12
Don't get it wrong with the API tolerance on Kwall ! It is there to represent a manufacturing process and the presence of a "Thin wall" on one side of the pipe against a "thick wall" at the same location on the opposite (Hence no deration on Axial specs but only on the hoop for Burst calculations using either Lame or Barlow equations)
Your pipe strength is also capped by PPF and tolerances on ID and OD.
This is a common misconception and definitely a wrong assomption taken by logging companies which would perform wear logs (Assuming you start with a bigger ID due to the Kwall...)
What you need to understand is how the wear provided by your log is calculated. What normally happens for sonic based wear logs is they use pipe OD as reference and then compute your wear with more or less accuracy. (in fine more or less similar to a mechanical MFCT depending on sizes of course.) 
Once you have figured out, how the WT is computed (as some will add 34% to the 12.5% Kwall - typically when you use a MFCT) but some wont (sonic logs) and 34% is your final result.. 

Then you can do a simple uniaxial burst calculation and check if test will work...

Good Luck
DougP
Retired
Marathon Oil
Total Posts: 18
Join Date: 18/08/05
Loss of strength depends on type of corrosion to some extent, it would be easy if corrosion was even over complete pipe surface as you could just calculate using the reduced wall thickness.  For example burst = 2Ypt/D where Yp = Yield Strength of material, t = wall thickness and D = outside diameter.  For new API pipe t used is nominal pipe wall thickness and the answer is multiplied by 0.875 to allow for the API minimum wall thickness of 87.5%.  So if your casing is worn / corroded to 50% of API nominal wall thickness you would multiply by 0.5 or alternatively use the remaining wall thickness in the base calculation.
Jump to top of the page