Private Forums
Fatigue analysis with very short fatigue life
22 April 2013
Fatigue analysis seems to now be undertaken more commonly for subsea exploration / appraisal (P&A) wells in the North Sea. The results of a fatigue analysis we've had done are suggesting a requirement for much heavier equipment than expected or has been used in the past. It appears some operators are now using heavy conductor for fatigue resistance (i.e. 36" x 2" instead of 30" x 1.5") and wellhead equipment in similar water depths (400ft), but other operators nearby in similar water depths have much lighter equipment. Our concern is that higher specs (including weld quality and extension joint length) could restrict our ability to source equipment. Have those of you who have had similar experiences with fatigue modelling found any learnings that can be taken to mitigate against potentially overdesigning the well due to fatigue modelling?

It could be that this well actually does require heavy equipment, but fatigue analysis has a reputation for being a bit of a dark art with large safety factors (e.g. DnV safety factor of 10); are there any learnings yet from actual measurements made against modelling?

7 answer(s)
Total Posts: 7
Join Date: 20/07/10
We have not had a 30" conductor fatigue failure in the 47 year history of ultra deepwater drilling.  There have been some 20" wellhead failures.  Once the stack is landed the 20" takes the bending.

Yes the conductor moves.   Thats because the soil has low strength in deepwater.  You can make the conductor 6" wall and it will still move.

I recommend giving this analysis to someone who knows what they are doing.  Stress (Ken Bhalls) ran a detailed analysis on this issue for BP some 15 years ago.  They can re-run it for this generation of engineers.

As a rule a thinner wall and smaller diameter conductor is better.

Senior Drilling Engineer
Total Posts: 19
Join Date: 04/05/06
Arun Depends on the enviroment really as well as the points that David provided above- I have never had a failure in 30 years but this is what we tend to do now: Core analysis of the soil Fatigue Assessment for service life loads 2jnts of 36" x 2" WT with Merlin Connectors 2jnts 1.5" WT 30" 2jnts 1" WT 30"
Drilling Specialist/Well Engineer/Training Consultant
Kingdom Drilling
Total Posts: 353
Join Date: 10/01/05

something else I have just experienced. The effect of a mud mat as a fatigue resistor enhancer or more detrimental or not.

I.e. the mud mat we know does not support the conductor that much but is commonly run as a hydrate barrier so they say. Note: However I can show you photos why this does not prevent hydrates either. So it's a complete waste of time for me for this. 

however recent well, too much conductor was over engineered where we could have sat the titanic on it. Mud mat resulted was 2m above seabed and we then experienced conductor movement evident at seabed ( all clay) based on jetting data and penetration test we did.

note dp 6th gen vessel with metocean periods of quite strong current of 3 knots and more. so rigs rao would create a pattern of movement and conductor movement if subsurface conditions allow. I.e soft watery clay in first 25 m jetted at seabed. this is common.

Movement is not as bad as it looks however surely? i.e. soil is plastic and acts like a series of springs according to py curves etc is my understanding. Moving point of fixivity down is good offering more area and load distribution?

with mud mat bedded 2m in then likely car parked with concrete from surface job as our norm, we have never see movement in two years with almost 15 wells that got me thinking on this subject.

does the cemented in mud mat shift the point of fixivity to the seabed where it is. 

When we cement, is point of fixivity where top of cement is?

how does this add or subtract fatigue life!

interesting discussion where a lot of my questions I am sure can be addressed and answered to provide clarification for us all.


Drilling Engineer
Total Posts: 14
Join Date: 01/09/12

As you mentioned Fatigue Analysis is a bit of dark area and very sensitive to inputs. Especiallially those of soil properties, metocean data. Small difference in these inputs can make a huge effect on the results.
We have recently got results of 0.5 hrs - 1 day fatigue lifetime for some of our projects which were obviously not true.
Drilling Specialist/Well Engineer/Training Consultant
Kingdom Drilling
Total Posts: 353
Join Date: 10/01/05
David covers the points that is my understanding.

General description of this is covered well in the IADC deepwater well control guidelines and follow API2A guidelines with large safety factors.

Jetted or drilled and cemented will offer a completely different set of numbers. Soil understanding is paramount if jetted. 

Since the 90's, we always ran a 36"x 2in wall thickness west of Shetland (UK North Sea) for the wellhead and cover joint. Then 30" below this.

Then 30 below.

A 42in hole opener was run behind the BHA and did not present any problems.

Some prefer and state that motors spud better, but my experience of rotary with a stiffened more rigid BHA with stabs, has yet to fail me. 

Verticality being an essential requirement particularly at the seabed and first 20m 

Total Posts: 7
Join Date: 20/07/10
The 20" carries the stack load.  A thicker conductor actually causes higher bending loads.  Any analysis showing otherwise is too simple.  The plastic-elastic soil resistance dominates.
HPHT Drilling Engineering Team Lead
Maersk Oil
Total Posts: 25
Join Date: 22/07/12
Hi Arun

The new generation of BOP stacks (6th and 7th generations) combined with the potential capping stack weight who now has to be considered is the main driver for the 2" CP.
Be carefull as this requirement for 2" is not over the entire Conductor but generally just for the WH extension and first joint below, as most of the studies generally show that stresses are concentrated within the first 15m below mud line... (This can vary slightly according to soil properties on locations but is unlikely to go much deeper than 20m bml.)
Hence, you are ending with WH extension + 1 or 2 jts in 2 ", XO to 1" for the remaining of your string.
The 10x DNV safety factor is not shocking and is a common industry safety factor for structures (we are dealing with a lot of uncertainties here, such as self frequency for BOP stack, riser assy...).
The resistance to fatigue of your connectors on CP has also to be properly evaluated (Companies like 2H/Stress/Atkins/Aquaterra, and I am probably missing some others) can give you good recommendations.
Recent article published in the WO

can be of some interest for you.

Hope this helps,


Jump to top of the page