T&D and Buckling Modeling

This Section Will Cover:

• How T&D Models Work
• What Friction Factors Are
• What is necessary to set up a model for
planning future wells (aka, “forecasting”)
• How to calibrate a model using actual data
from previous wells (aka, “hindcasting”)

© K&M Technology Group 2015 Page 1 of 48

1/18
 T&D and Buckling Modeling
T&D modeling is necessary to:
• Establish feasibility
• Establish equipment (pipe, rig) requirements
• Optimize design and strategies
• Identify sensitivities

© K&M Technology Group 2015 Page 2 of 48

2/18
 Common Observations
Many engineers prefer to rely on “gut feel” &
“experience”, rather than T&D calculations
− “T&D programs aren’t to be taken literally - they are just a
guideline”
• Can you imagine designing a $30 – 40 million dollar building with
no engineering planning?
• Complex wells can (and should) be modeled with confidence.
• Gut feel & intuition will betray you.

© K&M Technology Group 2015 Page 3 of 48

3/18
 Common Observations
Engineers tend to be optimistic with T&D planning
– Even when conservative in other areas of planning wells
• E.g. no safety factor is applied to FFs seen on offset wells
• Sensitivity for “what-if” FFs is rarely performed to understand
consequences of being wrong
Drilling teams assume that past experience is relevant
– “Why should we have to do anything different? We drilled a
similar well a few years ago and it went OK”
• Unless this well is identical in all respects, past experience may be
more dangerous than helpful
• This is “linear thinking”, and very dangerous
– Eg. Cosine(80o) is 33% different than Cosine(75o)
– And additional 1o is 11% difference again

© K&M Technology Group 2015 Page 4 of 48

4/18
 Common Beliefs
• T&D friction factors from offset wells are appropriate
for planning high angle wells
• Use of Rotary Steerable BHAs improves T&D
• That cased-hole FF is slipperier than open-hole FF
• That cased hole FF should be used for T&D calc’s
• That stiff-string models are more accurate than soft-
string models

© K&M Technology Group 2015 Page 5 of 48

5/18
 T&D and Buckling Modeling
Part 1 – How do T&D Models work ?
– What is a “Friction Factor”?
– How are calculations performed?
– What does the “little black box” assume?
– What inputs are important?
– What types of basic calculations are performed?
– And how to interpret them
– Common mistakes

© K&M Technology Group 2015 Page 6 of 48

6/18
 T&D and Buckling Modeling
What is a “Friction Factor”?
• Usually confused with “Coefficient of Friction” (CoF)
• F.F. ≠ C.o.F.
• The CoF is measured in a laboratory, often to compare various mud
systems or lubricants under controlled conditions
• The CoF value measured in the lab is almost never the same as what is
measured in the field

© K&M Technology Group 2015 Page 7 of 48

7/18
 T&D and Buckling Modeling
Why is there a difference?
• A Friction Factor is a “catch all”– it captures many unknowns
that are un-measurable with current technology
– Hole geometry – Ledges, spiraling, washouts, filter cake, etc.
– Pipe stiffness – Connection effects, centralization, pipe wear
– Cuttings Beds – Thickness, roughness, sand/shale content
– Differential Sticking Effects
– Pipe Weight errors
– Tooljoint / coupling interaction

© K&M Technology Group 2015 Page 8 of 48

8/18
 T&D and Buckling Modeling
What is required for Accurate T&D Modeling?
– Good input data
– Wellpath
– Drill-string (especially pipe weight, and pipe OD-ID for buckling calc)
– Block Weight
– Hole size description (affects buckling only)
– Mud weight (for buoyancy only)
– FF inputs
– Cased hole, Open hole, or Average FF
– Understanding of What scenarios to Investigate
– Understanding of How to Interpret Model Output

© K&M Technology Group 2015 Page 9 of 48

9/18
 T&D and Buckling Modeling
How do the T&D Software Models Work?
• Most models use common algorithms for T&D modeling
– C.A. Johancsik, et al – SPE 11380 (Exxon, 1984)
• Buckling models are more specialized
• Stiff string models also tend to vary
– WellPlan (Landmark) – Soft String or Stiff String mode
– Drilling Office (Schlumberger) – Stiff String only
– Advantage (Baker Hughes) – Soft String or Stiff String mode

© K&M Technology Group 2015 Page 10 of 48

10/18
 T&D and Buckling Modeling
How do the T&D Software models work?
• Most T&D models assume “flexible member” theory
– Does not allow for stiffness or geometry of the pipe
(this is the 3rd source of side forces that we ignored in the previous section)
– Explains why:
• Casing has different friction factors than drillpipe
• Torque increases in 8½” hole compared to 12¼”

• Stiff string models work differently

– Attempts to normalize friction factors by allowing for stiffness
• Should get the same FFs for drilling, as for running casing

© K&M Technology Group 2015 Page 11 of 48

11/18
 T&D and Buckling Modeling

Soft String: Assumes pipe follows the shape of the hole (like spaghetti)

Stiff String: Attempts to account for additional side forces caused by stiffness / relative hole size

© K&M Technology Group 2015 Page 12 of 48

12/18
 T&D and Buckling Modeling
Is a stiff-string model better for ER wells?
• K&M contend that such models are not any more accurate than
soft string for ER wells.
• Are actually invalid and dangerous if higher friction factors are
not used for “stiffer” operations
– Unknowns that are critical to accuracy can never be known
• Hole size & shape
• Cuttings bed height, and how it interacts with pipe
• Doglegs between surveys
• Pipe weight (new pipe is wrong, let alone used pipe)
• How couplings, centralizers interact with the wellbore

© K&M Technology Group 2015 Page 13 of 48

13/18
 T&D and Buckling Modeling
Understanding Friction Factors
• Friction factors may not be interchangeable between models
– Models can be easily cross-checked
– For basic calculations, typically with 10%, but variation comes with
more advanced calculations

• Note that for any model, the FF output when hind-casting is

very sensitive to the input data
– Any error in inputs will change the resulting FF output
– And errors are almost always present in:
• Block weight
• Pipe weight
• How CHFF is applied

© K&M Technology Group 2015 Page 14 of 48

14/18
 T&D and Buckling Modeling
Types of T&D Calculations
– Most programs offer 1 or 2 types of calculations
• Tripping-type (or the “driller’s view”)
• Element-type (or the “snap shot view”)
• Results may be graphical, or simply a number

© K&M Technology Group 2015 Page 15 of 48

15/18
 T&D and Buckling Modeling
Snapshot-type Calculations… continued
– Note: be careful of snapshot interpretation
– The program is not telling you that you can get to this point, but
rather, “if you can get this deep, this is what will happen”

© K&M Technology Group 2015 Page 16 of 48

16/18
 Example of a single point or “SNAPSHOT type calculation
• Shows that the casing run is OK at TD

…However, the TRIPPING type calculation

shows something quite different
• The worst-case load is at an intermediate
point in the well
• The string stops running in the ground long
before reaching TD!

© K&M Technology Group 2015 Page 17 of 48

17/18
 T&D and Buckling Modeling
How does the T&D model use:
• Block Weight
• Cased Hole FF (CHFF)

• Both of these are almost always wrongly used, and it can

have a big impact on interpretation of results

© K&M Technology Group 2015 Page 18 of 48

18/18
 T&D and Buckling Modeling
How is Block Weight used in the T&D calc ?
1. A “hookload” or “weight” measurement is taken on the rig
2. But the T&D program doesn’t work with hookload or weight – only tension is
important
• How to convert hookload to tension ?
• Block weight is subtracted from the hookload…
• The FF is then calculated from this block-adjusted number

• This is where the error occurs, because the block weight is usually wrong
• And you can’t back-calculate the correct FF with an incorrect BW

© K&M Technology Group 2015 Page 19 of 48

19/18
 T&D and Buckling Modeling
• Block weight measurements
– BW must be measured separately weight in each direction
– The BW will be different for SO, PU and stationary
• Due to sheave friction, hoses, etc
• Typically 5-8 kips (2-4 mT), but up to 25 k-lbs (11 mT) difference

• It can make a big difference in ER wells

© K&M Technology Group 2015 Page 20 of 48

20/18
 T&D and Buckling Modeling
How is CHFF used in the T&D calc ?
Consider this well, drilling 8½” hole
• 9⅝” casing set at 6,000m (20,000’)

Prior to drilling out, cased-hole T&D

measurements taken to establish CHFFs
• CH TQ = 34 k … for CHFF = 0.25

© K&M Technology Group 2015 Page 21 of 48

21/18
How is CHFF used in the T&D calc?
Consider Torque (but it’s the same for SO & PU)
• At 9⅝” shoe, CHTQ = 34k  CHFF = 0.25
1. Drilling open hole commences
After 1 stand, say TQ increases by 2k (to 36 k)
• What is OHFF?
• Well, Surface TQ = CHTQ + OHTQ
• We don’t know OHTQ, but we can calculate
CHTQ from the CHFF
• Based on CHFF, what torque is generated in cased- This is the CH
hole with current pipe in cased hole? TQ vs depth, for
• CHTQ = 34 (same as before) FF = 0.25
• If Surface = 36, and CHTQ = 34, then OHTQ = 2 k
CH Torque
• OHFF is then calculated, based on what FF is measurement
necessary to generate 2k torque in one stand
• Therefore OHFF must be > 0.5 (say 0.7)

But it is just as likely that the TQ will decrease after

drilling out the shoe (say now 31 k)
• OHTQ is now -2k
• OHFF is therefore now negative – which is impossible
© K&M Technology Group 2015 Page 22 of 48
22/18
 So what is really happening, to explain absurd OHFF?
• The CHFF is changing with each stand drilled
• Why?
• Some polishing
• Also the influence of cuttings bed moving up into casing
• With each stand drilled, the cuttings bed will change
(volume and type of cuttings mixture)
• CHFF will not remain constant while drilling ahead

© K&M Technology Group 2015 Page 23 of 48

23/18
 T&D and Buckling Modeling
Cased Hole vs. Open Hole FF’s: … continued
• CHFF and OHFF are not independent variables
• Cannot have wrong CHFF, and expect meaningful OHFF results

© K&M Technology Group 2015 Page 24 of 48

24/18
 T&D and Buckling Modeling
Cased Hole vs. Open Hole FF’s:
Is Cased Hole FF a valid concept?
– Only when you can measure a meaningful CHFF
• For casing runs, etc.
• NOT for drilling calculations (or other dynamic environments)
– For Drilling calculations use average FF

© K&M Technology Group 2015 Page 25 of 48

25/18
 T&D and Buckling Modeling
Part 2 – Choosing FFs for quality planning
– What drives FFs
– Separating myth & reality
– How conservative should you be?
– When to allow for static FFs

© K&M Technology Group 2015 Page 26 of 48

26/18
 T&D and Buckling Modeling
Understanding Friction Factors
– 3 types of friction factors (not just one)
• Slack-off, pick-up and torque should use different numbers
• Usually only dynamic torque is used or reported
– Common misconception
– Leads to under-estimating Slack-off & Pick-up drag by up to 50%

© K&M Technology Group 2015 Page 27 of 48

27/18
Typical Torque FF’s for 12.25” hole in OBM or SBM
If Torque FF was used for planning SO or PU
• Shows that the TQ-FF = 0.15 – 0.22 (0.13 at TD)
• Plans would under-estimate FF by >70%
• Generally, this is the only number reported or recorded.
• Most projects simply assume that Slack-off and Pick-
up FF’s are the same ….
© K&M Technology Group 2015 Page 28 of 48
• SOFF is 0.30 – 0.35, PUFF = 0.30 – 0.35
28/18
 T&D and Buckling Modeling
Friction Factors
– Typical “Drilling FF’s” for 12¼” in OBM / SBM
– Torque FF = 0.16 – 0.18
– Slack off FF = 0.25 - 0.30
– Pick up FF = 0.20 - 0.25

© K&M Technology Group 2015 Page 29 of 48

29/18
 T&D and Buckling Modeling
Friction Factors
• Why three types of friction factors ?
– If it was only C.o.F., then they would be the same
• But remember FF is much more than C.o.F.
• We are not sliding a block across a polished table …
…we are dragging or rotating pipe through a tunnel
– Tunnel is unknown shape & size,
– We usually don’t know the pipe weight
– Path is only partly known
– Filled with an unknown amount of dirt,
– and we don’t know how the connections, centralizers, stabilizers
interact with any of these

© K&M Technology Group 2015 Page 30 of 48

30/18
 T&D and Buckling Modeling
What drives dynamic FFs ?
• Annular clearance
– Small clearance situations create step change in FF
– Looking for “small-pipe in big-hole”, or “big-pipe-in-small-hole”
• Tortuosity
– But not like people assume
• Fluid Lubricity
• Wellbore & cuttings “material”
– steel, shale, sand, or mixture

© K&M Technology Group 2015 Page 31 of 48

31/18
 T&D and Buckling Modeling
The key issue that drives FF is the pipe / hole clearance
– Is it a “small pipe & big hole” situation
– Or a “big pipe & small hole” situation
• If the latter, a step change in FFs occur

– Different FFs need to be used for different operations

– Running casing has much higher FFs than for the drilling operation (in
the same hole section)
• Drilling SO typically 0.25 – 0.30
• Casing / Liner / Screens run may be 0.4 – 0.5 (for a good run)
• 8½” drilling FFs are typically higher than for 12¼” hole (especially TQ)
• Say 0.30 – 0.35 vs. 0.18 – 0.22

© K&M Technology Group 2015 Page 32 of 48

32/18
ECD, Friction, Hole Cleaning Problems

Optimum Range: 3.25 to 3.75

Hole Optimum DP
12 ¼” 6 ⅝”
10 ⅝” 5 ½” or 5 ⅞”
9 ½” 5”
8 ½” 4 ½”
6 ½” 3 ½”

© K&M Technology Group 2015 Page 33 of 48

33/18
Example of 12¼” Drilling Hookloads Same wellbore, with resultant 95/8” casing run
• P-HAR=4.96 • P-HAR = 1.62
© K&M Technology Group 2015 Page 34 of 48
• From data, SO & FF = 0.2 – 0.25
34/18 • Run was good, but FF’s now 0.4 – 0.5+
 Example running 6⅝” sand screens (OD = 7.5”)
into 8½” open hole
• P-HAR = 1.28
• CHFF = 0.15, but OHFF’s high
• Note OBM and Rotary Steering Tools used

© K&M Technology Group 2015 Page 35 of 48

35/18
 Example running 4” sand screens (OD = 5”) into 7” liner
• CHFF above liner top = 0.31 (P-HAR=2.89)
• CHFF inside liner = 0.5 (P-HAR=1.44)
• Highlights big-pipe / small-hole influence on FF’s

© K&M Technology Group 2015 Page 36 of 48

36/18
 T&D and Buckling Modeling
• Key Message:
– You cannot estimate or assume FFs for a casing / liner run
based on drilling T&D, unless annular clearance is large

– Note that this can be used to your advantage too

• Enlarging a hole (say from 8½” to 9½”, or 12¼” to 13½”) can
have a significant FF benefit when running casing

© K&M Technology Group 2015 Page 37 of 48

37/18
 T&D and Buckling Modeling
Myths / Misconceptions about what drives FFs
• That improved hole cleaning will reduce T&D
– Not true for torque (often neutral or worse as hole get’s cleaner)

• Reducing contact area reduces T&D (i.e. using centralizers)

– Not true.
– Casing still weighs the same, but is now point-loaded
– Contact area is a non-issue, except when differential sticking is present
• Then centralizers are a critical stuck-pipe prevention tool

© K&M Technology Group 2015 Page 38 of 48

38/18
 T&D and Buckling Modeling
Myths / Misconceptions about what drives FFs
• That cased-hole is more slippery than open hole
– Sometimes not true, especially for drilling FFs
• Who said steel-on-steel is slippery?
• Regularly see drilling FFs increase (torque by 50% - 70%) when hole is cased

– Running casing does tend to have lower CHFF, but not drilling operations

© K&M Technology Group 2015 Page 39 of 48

39/18
 T&D and Buckling Modeling
Myths / Misconceptions about what drives FFs
• That friction factors don’t change while drilling
– Lithology changes often effect the natural friction factor (even in
“clean” hole)
• Claystone is often slippery
• Sands can be slippery or very high friction
• Carbonates can be high or low
– Local experience is required to know what “normal” FF behavior is…

© K&M Technology Group 2015 Page 40 of 48

40/18
 Notice how friction factors correlate to lithology:
• 0.15-0.20 when mostly sands have been drilled
• 0.25-0.30 in claystones
• 0.20-0.25 at TD

© K&M Technology Group 2015 Page 41 of 48

41/18
 Notice how Torque friction has a different response…
• TQ FF reduction initially (when P/U and S/O are
increasing)
• No reaction when the claystones are drilled

© K&M Technology Group 2015 Page 42 of 48

42/18
 T&D and Buckling Modeling
Myths / Misconceptions about what drives FFs
• Rotary steerables will result in reduced T&D
– Reducing “macro” tortuosity (i.e. big & sharp doglegs) is critical
– But micro-tortuosity & spiraling is over-rated for ER wells
• Gauge hole actually increases FF for casing runs
• Same can be said for using OBM vs WBM

© K&M Technology Group 2015 Page 43 of 48

43/18
 T&D and Buckling Modeling

Consider this world record well drilled in 1992…

• 12¼” hole drilled with old technology
• 9⅝” liner successfully run to TD

Consider this well drilled from the same platform in late 2005…
• 12¼” hole drilled with new technology
• 9⅝” liner failed to reach TD twice, due to much higher FFs

© K&M Technology Group 2015 Page 44 of 48

44/18
 T&D and Buckling Modeling

And this well drilled in late 2005 (same platform)

• 9⅝” Liner Run, based on previous experience
•With 25% safety factor (0.30 vs. 0.24)
•Now drilled with new technology
Consider this world record well drilled in 1992
• 9⅝” Liner Run
• 12¼” hole drilled with steerable motor &OBM
Resulting OHFF = 0.35 – 0.45
Failed twice, because FF design
Resulting OHFF = 0.18 – 0.23 basis was too optimistic

© K&M Technology Group 2015 Page 45 of 48

45/18
 T&D and Buckling Modeling
So how do you decide what FFs to plan for?
– Is there relevant offset data to establish a minimum design basis?
• Is experience local, or regional?
• Were the circumstances similar?
– Is it a big-clearance or small-clearance situation?
– What sort of lithology is involved?
– What sort of fluid is used?
– How much cased hole?
– Does static-friction need to be allowed for?
– What are the consequences of being wrong?

© K&M Technology Group 2015 Page 46 of 48

46/18
 T&D and Buckling Modeling
Friction Factors … continued
– Planning friction factors should be based on regional
experience (if possible)
• E.g North Sea is very slippery
– Norway experience will lead to under-estimating in other regions
• Allow for lithology differences
– Clay can be very slippery
– Sand can be slippery or coarse
– Well Design (overbalance pressure)
– Expected hole cleaning performance

© K&M Technology Group 2015 Page 47 of 48

47/18
 T&D and Buckling Modeling
Friction Factors… continued
– How conservative should we be for planning friction factors?
• Is is it a question of feasibility or optimization?
• What are the consequences of being wrong?
– Is the casing run a one-way trip?
• Is static friction an issue?
– Your offset information is dynamic…static friction must be allowed for
when neutral weight or the top-drive limit is approached

© K&M Technology Group 2015 Page 48 of 48

48/18