Journal of Petroleum Science and Engineering 73 (2010) 258–266

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p e t r o l

Friction factors: What are they for torque, drag, vibration, bottom hole assembly and
transient surge/swab analyses?
Robello Samuel
2107 Citywest Blvd., Building 2, Halliburton, Houston, TX 77042, USA

a r t i c l e i n f o a b s t r a c t

Article history: The term “friction factor,” conventionally defined for fluid flow engineering calculations, is also used for solid
Received 24 February 2010 mechanics calculations in the drilling industry. At first glance, it may seem to be a misnomer, but there are
Accepted 6 July 2010 underlying facts and reasons for the adoption of the term “mechanical friction factor” that is unique to the
industry. This term is used as a proxy for the coefficient of friction (COF) or the Coulomb friction, a
Keywords: dimensionless scalar value, in the estimation of torque, drag, and side forces for the directional tendency
friction factor
prediction, static displacement of string for vibration analysis, and moving pipe elastic force for transient
torque and drag
drilling
swab/surge pressure estimation. Because of larger uncertainties in the modeling of a long drillstring,
including cuttings bed, tortuosity, mud properties, fluid viscous effects, wellbore temperature, pressure, and
non-uniform geometrical interferences between the drillstring and wellbore, the COF is termed as a friction
factor by including the uncertainties involved without defying the mechanics involved. Another important
parameter used during the planning stage in combination with the friction factors is the tortuosity factor to
provide artificial undulation to the wellbore. The tortuosity factor causes an additional problem, which
requires a different friction factor for the analysis, even though the COF between the drillstring and the
wellbore has not changed. Because there is no industry standard for quantifying these parameters, a great
deal of confusion exists between these parameters and results in the overestimation or underestimation of
the frictional drag. This paper describes the details, clear definition, calibration, pitfalls, and context in which
they can be used. It presents the simple guidelines and several example calculations. This paper also provides
methods for evaluating these parameters with real-time data.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction estimation of bed height are critically important in many phases of the
well construction. During the design phase these calculations are
Friction, an important source of wear and energy loss in the affected by the coefficient of friction (Maidla and Wojtanowicz, 1987;
tubular system, is a resistive force that retards the motion of an object. Locke et al., 1988; Samuel, 2010). The difference between the planned
Because the tubulars and downhole tools are moved in and out during smooth well profile and the actual well drilled can have a major effect
the well construction process, the resistive force is observed in the on the torque and drag losses for the well. In order to account for this
form of drag. Because the resistive force depends upon the surface, the effect, during the planning phase, the hole spiraling or oscillation of the
drag force depends on the smoothness of the wellbore and on several wellbore can be described by applying tortuosity to the wellbore with
other parameters. Various operations, such as tripping in, tripping out, the desired tortuosity factor which further coupled with friction factor
with and without rotation, rotating on and off bottom, and sliding (Luo et al.). Another area where design challenges are faced is running
further complicate the modeling; as a result, the friction factors that casing to its desired target depth using floated casing. This requires a
are used also differ. more comprehensive in-depth analysis. If proper friction factors are
With the advent of new and complex downhole assemblies, there is not used, this results in significant underestimation of frictional drag
a growing need to monitor the performance of these tools, as well as further resulting in casing not reaching its target depth (Johan Eck-
the quality of the wellbore. Alternating between the sliding and rotary Olsen). So in order to account for these effects, the friction factor
modes results in hole spiraling; wellbore oscillation becomes more should be used carefully to estimate various forces so that the limits of
pronounced because of frequent slide drilling. Qualitative assessments the operation can be clearly defined as the well is planned or drilled.
and quantitative estimates of the torque, drag, surge pressures,
vibration intensity, drill ahead prediction, maximum overpull, and 2. Background

Several studies and models have been proposed in the past to

E-mail address: [email protected]. estimate the torque, drag, post buckling prediction, static displacement,

0920-4105/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.petrol.2010.07.007
 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266 259

directional tendencies, etc. The reliability of the prediction of the model

mostly depends on the use of coefficient of friction. There have
attempts to quantify the calculations by way of providing a range of
coefficient of frictions for different operation conditions. The guidelines
provided may be inadequate and the calculated values could differ
considerably from actual data. Because of larger uncertainties in the
wellbore and non accounting of all the contributing parameters,
calculation of torque, drag forces, etc. result in underestimation or
overestimation of the calculated values. To obtain reliable and
consistent values of estimating the interacting mechanical forces, the
COF is termed as a friction factor. To fully understand this factor the
following paragraphs describe the basics of coefficient of friction.
The coefficient of friction (COF) is defined as the ratio of the


frictional force, Ff to the normal force ðFn Þacting at the point of
contact; it is given as:

Ff
μ= : ð1Þ Fig. 1. Friction force as a function of pulling force.
Fn

2.3. Rolling friction

The COF is a scalar dimensionless value that depends on the
surface but is independent of the surface area. Table 1 shows typical
Rolling friction is the friction that occurs when one object rolls on
COFs for various materials. The friction force acts in the direction
the other and is one form of sliding friction. These COFs are based on
opposite to the direction of motion of the object. The COF can be
empirical tests that must be measured and cannot be calculated.
classified on the basis of the following two types:
Typically, the coefficient of rolling friction is less than the coefficient of
• Static friction kinetic friction.
• Sliding friction.
Frf
2.1. Static friction μr = ð4Þ
Fn

Static COF is the ratio of the resisting impending force to the

normal force. The friction force required to move an object beginning
from the “limited displacement” is greater than the force generated 2.4. Slide/Roll friction
when it is moving.
There are conditions in which the object may slide and roll
Fsf simultaneously. Few experiments and studies (Ginzburg, 1989) have
μs = ð2Þ been performed in the mechanical industry to measure the friction
Fn
during simultaneous rotation and reciprocation. In a gearing applica-
tion, the word “Walzreibung” relates to the simultaneous rolling and
sliding of gears that require a different type of lubrication. Based on
2.2. Sliding friction whether the object will slide, roll or slide, and roll, there are three
types of static friction: static sliding friction, static rolling friction, and
Kinetic friction or sliding friction is defined as the friction static roll and slide friction. In the same explanation, there are three
generated when the two objects slide or move relative to one types of kinetic friction: kinetic sliding friction, kinetic rolling friction,
another. This motion does not include rolling or spinning. Usually, the and kinetic slide and roll friction. Kinetic friction tends to be less than
kinetic friction is less than the static friction. A typical static and the static friction.
kinetic coefficient plot is shown in Fig. 1.
2.5. Angle of friction

Fkf
μk = ð3Þ Tangent of the angle of friction is equal to the coefficient of static
Fn friction. It is also called the friction angle or the angle of repose. It is
the angle between the resultant force acting on the surface and the
normal force when the sliding motion of the surface is impending and
Table 1
is given as:
Typical coefficient of friction (Rabbat, 1985).
−1
Material 1 Material 2 Dry Lubricated φ = tan μ s: ð5Þ
Static Sliding Static Sliding

Steel Steel 0.78 0.42 0.05–.11 0.29–0.12 Eq. (5) is valid if the object is sliding at a constant velocity; for an
Aluminum Aluminum 1.05–1.35 1.4 0.3
accelerating object, the kinetic friction and the friction angle are
Aluminum Mild steel 0.61 0.47
Copper-lead Steel 0.22
related by Eq. (6) as follows:
Diamond Diamond .1
Diamond Metal 0.15
Steel Concrete 0.57–0.75 0.45 ax
Steel Embedded sand 0.7 μ k = tan φ− : ð6Þ
g sinφ
260 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266

3. Complexity of friction for a pair of materials, such as casing, drillpipe, formation, and
drillpipe is very complex for long strings. In the absence of continuous
After the explanation and definitions of friction force and COF, they survey data, the actual shape of the wellpath is not known, even if the
may seem simple to use. For complex operations, they are much more surveys are taken 90 ft apart. The following describes the calculations
complicated to analyze because they depend on various other factors, in which the COF is used.
including temperature, asperities between the surface, type of materials, The drillstring can be simultaneously rotated and tripped in or out,
sliding speed, and rolling speed in the case of an object rotating relative and the drag force can be given as:
to another. Also, the real unknown is the measurement of the hookload
at the surface. It is usually measured by using the tension in the drill lines jVts j
Fd = μ v × Fn × : ð8Þ
or from the drill line anchor. The friction effects of the pulleys are jVrs j
neglected. In addition, it has been found that the coefficient of sliding
friction is high at low sliding speeds and decreases as the speed The drillstring can be simultaneously rotated and reciprocated;
rotational speed increases. The relationship between the rotational consequently, Eq. (8) can be modified to include both the operations,
speed and the COF is linear for some materials and non-linear for other which results in:
materials. The same observation can also be made for sliding speed. As a
jωj
result of temperature variation between the surfaces the COF also T = μ v × Fn × r × : ð9Þ
jVrs j
changes. Experiments have shown that the COF increases with increases
in temperature for constant string or rotational speed, and has minimal
Several experimental data and laboratory experiments have
effect on dry surfaces. For certain materials with titanium the effect is
shown that the COF has a strong dependence on sliding and rolling
very minimal (Zuev, 1965). Because the work performed by friction is
speeds (Kireenkov, 2008). Simultaneous tripping and pipe rotation
released in the form of heat, the temperature increases and thereby
are often encountered during tripping in or tripping out operations.
reduces the COF. It can be affected as a result of geometrical
The following empirical equation provides a good representation and
nonlinearities, material nonlinearities, and contact nonlinearities.
coupling of the friction effects and drillstring rotating speed, as well as
Consequently, the friction factor can be given in terms of the COF
tripping speed:
between the materials, lubricity coefficient of mud ðLÞ, pipe sticking
coefficientðSÞ, pipe rotational speedðNÞ, temperature ðt Þ, wellpath −kjVrs j
μv = μs × e : ð10Þ
profile which includes the curvature and borehole torsionðτÞ (Samuel
and Liu, 2009a,b), and other uncertainties.
To calculate the values in the above equations, for drill collars and
μ v = f ðμ; L; S; N; t; E; τÞ ð7Þ coiled tubing, the outside diameter of the drill collar or coiled tubing is
used; for drillpipe, heavy weight and casing, and the outside diameter
of the tool joint is used; and for stabilizers, underreamer and hole
Because the COF is a dimensionless scalar value, it can also be
openers, the outside diameter of the blade or expanded arm diameter
expressed in the form of dimensionless parameters by using the
is used to calculate the radius.
variables expressed in Eq. (7).
During the rotation, the pipe tends to climb  to the high side of the
wellbore, and reaches an equilibrium angle of ϕ ϕ = tan−1 μv when
4. Drilling
the friction force is balanced. Consequently, Eq. (9) can be generalized
(de Zwart) as below:
Friction is an important part of drilling operations. As the work-
string is tripped in or out, or rotated on or off bottom, the friction force jωj
must be considered. It plays an important role in the solid mechanics T = μ v × Fn × r × cos ϕ × : ð11Þ
jVrs j
calculations, such as torque and drag, as well as in the hydraulics
calculations, including surge, swab, and hookload estimation during Eq. (11) can also be written in terms of friction factor (Mitchell and
cementing. Simulation of drilling operations with friction force is Samuel, 2007) as:
very complex because many uncertainties are involved. Some of the
uncertainties include the following: μv jωj
T = qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi × Fn × r × : ð12Þ
• jVrs j
Mud lubricity 1 + μv2
• Cuttings bed
• Dogleg/keyseat In addition, by defining the angle αrt between the trip speed and radial
• Wellbore curvature component as cos αrt = jjVωrsjj, the equilibrium angle can be written as:
• Borehole torsion φ = tan−1 ðμ v cosαrt Þ. The equilibrium angle depends on the friction
• Wellbore tortuosity (Samuel and Liu, 2009a,b) coefficient and the ratio of the tangential velocity and the resultant
• Viscous effect velocity.
• Borehole diameter Based on a recent study (Mitchell, 2006; Samuel, 2007), the
• Asperity between the drillstring and wellbore equilibrium angle attributable to pipe rotation with perfect contact in
• Flexural stiffness of the workstring terms of effective tension of the string Fe is given by:
Although there are uncertainties, the friction factors obtained or !
calibrated are within reasonable limits, even for long horizontal wells −1 Fe Δα Sinθavg −1
θ = tan + tan ðμ v Þ: ð13Þ
(Sonowal et al., 2009). The following descriptions explain the Fe Δf + Wb Sinθavg
complexity associated with modeling various operations. Field
experience has shown that axial drillstring drag is reduced when Eq. (13) is more comprehensive because it includes the azimuth
the drillstring is rotated. Torque and drag models account for this and direction change and is adjusted to include the trip and rotational
mathematically by the use of velocity vectors. The resultant velocity speeds of the pipe as defined earlier in Eq. (10) and Eq. (11).
Vrs of a contact point on the drillstring is the vector sum of two Because the pipe and other downhole tools become embedded
components: circumferential velocity VC (caused by rotation) and into the formation as well as in the cutting bed, a more comprehensive
axial velocity Vts (affected by drilling rate or tripping speed). The COF friction factor, proposed by Sextro (2002), can be used; which has the
 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266 261

dependency on the side force, kinematics, temperature, and geomet- over the length of the wellbore does not remain constant because of
rical parameters of the contacting surfaces. various factors, such as temperatures, cutting bed, and suspended
particles. Even if they are considered in the model, local variations will
μ
μv =  s  ð14Þ still exist. Because the hydrostatic and gravity loads are part of the
μ s σn equilibrium conditions, the side force and consequently, the drag
1+ jVrs j
kt Δt forces along the string, are affected. To account for these local
variations, the correction can be distributed over the length of the
well in the form of the uncertainty calibration factor (UCF). This UCF
4.1. Critical well inclination will provide adjustments to the effects of local hole conditions and
profile, including curvature and wellbore torsion, as well as the low
The friction factor between the string and the hole is an important contact conditions in vertical or near vertical wellbores. Friction
factor for high-angle wells and can be altered by changing the drilling fluid calibrations should adjust the mud weight as well as the string weight.
or lubrication. Normally, the sail inclination should be less than the critical
inclination, to ensure that the drillstring or casing slides downward. For
6.1. Rotating on/off bottom operations
sliding pipe or motor drilling, there is a critical inclination angle ðαcr Þfor
gravity-driven drilling given by (Liu and Samuel, 2009):
How can the friction factor for rotating off bottom condition be
  calibrated, because it is assumed to be without drag for this operating
−1 1
αcr = tan : ð15Þ condition? Regardless of the friction factor used, the measured weight
μ
prediction will always remain constant. The main contributing factors that
would cause deviation from the actual hookload include the mud weight
variability along the hole, drillstring mass, and the wellpath. An
5. Friction reduction
adjustment of the mud weight and drillstring mass will satisfy this
condition. To avoid the global change that will affect other calculations,
Friction is an important part of the drilling operations. As the
such as hydraulics and swab surge, a calibration factor can be used for mud
workstring is tripped in or out, or rotated on or off bottom, friction
weight and drillstring mass as a correction factor distributed over the
generates a great amount of heat and drag forces. A common form of
length that requires calibration. Because hydrostatic and gravity loads are
friction reductions are performed by using mechanical devices and
part of the equilibrium conditions, contact forces and hence drag force
lubricants for a given wellpath and borehole condition. The heat
along the length of the string required will affect the hookload
generated as a result of friction can be given as:
calculations.
H = μk ∫Fn ðsÞds: ð17Þ
6.2. Tripping in/tripping out operations
Torque and drag or friction reduction tools (FRDs) are downhole
For tripping in and tripping out operations, the up and down drag
drillstring tools developed and intended to reduce rotating friction,
forces are not linearized reversals of one another for various reasons,
casing, and pipe wear (Samuel, 2007). These tools feature a non-
such as drillstring geometry, formation geometry, position, and the
rotating drillpipe protector that includes a sleeve on a lubricated
path traversed when the pipe is tripped; the contact points are
bearing surface, which becomes the effective contact point for torque
different, which results in different friction factors. Various studies
generation. There are various types that can be installed on the pipe or
have also indicated this condition (Maidla et al.).
between the pipe connections. FRDs are used in various drilling and
wellbore construction applications for a variety of reasons, including
rig limitations, complex wellpaths, differential sticking, buckling 7. Case studies
casing wear, torque reduction, axial drag, and ultimately operating
costs. They can be classified as either a fixed or roller type. FRDs are The friction factor effects are illustrated with the following analyses
best used in the initial building section of the hole or in the deviated which demonstrate the importance and establish the objectives of
portion of the wellbore where the contact forces are excessive. Several analysis.
friction-reducing tools, including rollers and non-rotating sleeves, are
available. They improve drilling by increasing available weight and 7.1. Static friction
minimizing slip stick. The placement and spacing to achieve optimum
performance depends on the well profile. Static-down drag is the amount of additional slackoff force required
to move the workstring. Similarly, static-up drag is the amount of
6. Friction calibration factor overpull required to move the casing when picking up. Normally, static
friction is automatically entered when using real-time data. If real-time
Because of the uncertainties previously explained in calculating data is being loaded, the static friction can be manually entered if there is
friction factor, it is often very difficult to match the calculated values a means of recording 1, 2, or 3 second initial drag with a surface recorder.
with the actual values. Friction factor also depends on where the When casing is released from the slips, the string may have to overcome
hookload is measured. Pick up and slackoff load will be under- an initial force before it will begin to move (Fig. 2). This resisting force is
estimated if the sheave friction is not taken into consideration; they related to static friction and occurs when running a workstring through
may be less than the indicated for slackoff and greater for pick up a differential sticking environment. Static friction is a particular risk
measurements, which may result in low friction factors for pick up associated with running long casing strings in high-angle and extended
and higher for slackoff loads. For example, in the vertical section, there reach wells. Torque and drag models only model dynamic friction effects
is no alternative option other than adjusting the block weight or the (i.e., when the pipe is moving). Static friction effects are not currently
weight of the string, which may result in additional changes in the accounted for in the calculations. Static friction occurs in varying
non-deviated portion. Because it is assumed that there is no frictional degrees of severity on most extended reach wells. Sometimes this factor
drag, the measured weight calculated will remain the same, is not observable and, at other times, it can be so severe that casing fails
regardless of friction factor; this scenario does not reflect the to reach to depth. Factors that influence the severity relate to hole angle,
downhole conditions in other deviated portions. The mud weight wellbore tortuosity, casing weight and connection type, formation
262 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266

Fig. 2. Static drag.

against the workstring are contacting, centralization strategy, amount of 7.3. Swab and surge analysis
mud overbalance, and hole condition.
For the calculation of dynamic swab and surge pressures due to pipe
movement, the mechanical friction forces are accounted in the correct
7.2. Drill ahead analysis direction to estimate the axial tension using the following equation

ξ
As the bit progresses in the wellbore, forces are generated in three Fμ ξ Fμ ðξ−sÞ
directions (i.e., axially (hole axis), vertically (inclination axis), and F ðξÞ = F0 e + ∫ λðsÞe ds ð16Þ
0
horizontally (direction axis)). In addition, the bending characteristics
of the bottom hole assembly confined in a wellbore will result in a bit
where
tilt in both the inclination and azimuth axes. The interaction of these
variables will determine the wellbore trajectory. While calculating ΔD
these forces, the friction factor used in the form of COF is important λ = −Rwbð−tFμnÞb; ξ = and b = t × n:
R
and, if they are not used properly, will skew the prediction. Figs. 3 and
4 show the side force at the bit, build rate, and walk rate predictions The influence of COF is evident in the overall results in estimating the
for various friction factors. It can be seen from the figures the effect of position of the pipe, as well as in solving the dynamic pressure-flow
friction factor on the side force at the bit which in turn affects the equations. Pipe elasticity is important in the calculation of the dynamic
build and walk rates of the assembly as it drills ahead. So, when a swab and surge pressures because the surface tripping velocity may not
theoretical model is used to predict the build, drop and walk rates it is be the same at the bottom of the workstring velocity. Consequently, the
essential to calibrate the friction factors for accurate prediction. The pipe and annulus pressures must be combined through pipe velocity.
directional and inclination side forces show a large variation, and it Fig. 5 shows the effect of friction factor on the dynamic wellbore
can be noted that the effect on walk rate is less compared to the build pressures. These results are for a J-type directional well with the
rate for the example considered in the analysis. maximum angle of 60° and dogleg 3°/100 ft, and show simulated swab

Fig. 3. Effect of friction force on side forces at the bit.

 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266 263

Fig. 4. Effect of friction force on build and walk rates.

results at the bottom of the well at 20,000 ft with and without friction slackoff and overpull operations when tripping in; Fig. 7 shows the
forces. The wave form completely switches in direction and may result in condition during the slackoff and overpull operations when rotating the
the overestimation or underestimation of the swab/surge effects. The off bottom condition (Haduch and Samuels, 1994).
effects will be more pronounced in highly deviated and extended reach
wells. Frictional forces will play an important role in the estimation of the
composite elasticity of the system while planning ultra-deep extended 7.5. Real-time friction factors
reach wells or in wells where the frac and pore margins are very narrow.
When accurate friction factors are derived from existing field data,
soft-string models yield reasonably accurate results. However,
7.4. Stuck force estimation because the soft-string model does not consider the stiffness of the
drillstring, its accuracy will degrade as the drillstring diameter and/or
The drilling string is assumed to be a series of blocks with weightless hole curvature increases. Increases in pipe stiffness and hole curvature
strings attached. Torque and drag calculations are usually used to result in high normal forces and therefore, in an increase in torque/
predict the overpull and slackoff forces for freeing stuck pipe and to drag. One common cause of tight hole is spiraling, which is difficult to
apply force and torque for backing off. In modeling, the friction force detect with conventional MWD survey data. Because the local or
reverses depend on the loading conditions (Haduch and Samuels, 1994). micro-level of well geometry is often difficult to obtain with
While estimating the friction forces, it may not be conservative to conventional MWD survey data, empirically derived, macro level,
assume the same friction factor. Fig. 6 shows the condition during the friction factors are often used for predictive analyses because these

Fig. 5. Effect of friction force on swab/surge pressures.

264 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266

Fig. 6. Stuck pipe calculation: trip in condition.

factors are calculated from wells that include similar localized factor during the operation phase is important for the tuning of
geometry. Friction factors should be derived from analogous case friction factor using actual real-time data. The following case history
histories. The properties of the drilling fluid used in the baseline wells using the data is provided to substantiate the practical usefulness of
and the planned well should be similar. However, the ranges in the model. The data is from a well where options were planned so that
Table 2 can be used as starting points if prior experience is unavailable the casing running in and cementing operations can be completed
(Johancsik et al., 1984; Rasmussen et al., 1991). Calibration of friction without difficulty. The total measured depth of the well is 12,800 ft

Fig. 7. Stuck pipe calculation: rotating off bottom condition.

 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266 265

Table 2 pulled, if necessary. In addition, the torque and tension/compression

Range of friction factors. loads at any location in the drillstring must be used to compare with
Fluid type Friction factors the torsional, tension, and buckling capabilities of the drillstring and
tool joints.
Cased hole Open hole

Oil-based 0.16–0.20 0.17–0.25

9. SI Metric Conversion Factors
Water-based 0.25–0.35 0.25–0.40
Brine 0.30–0.4 0.3–0.4
Polymer-based 0.15–0.22 0.2–0.3
Synthetic-based 0.12–0.18 0.15–0.25 cP × 1.0* E-03 = Pa.s
Foam 0.30–0.4 0.35–0.55 ft × 3.048* E-03 = m
Air 0.35–0.55 0.40–0.60 in × 2.54* E + 00 = cm
lbf × 9.869 233 E-00 = N
md × 6.894 757 E-04 = μm2
with the open hole of 2400 ft. Fig. 8 shows the actual hookload for a psi × 6.894 757 E + 00 = kPa
tripping in operation against the calculated hookload. For the analysis
Conversion factors exact.
it was found that a friction factor of 0.28 inside the casing and an open
hole friction factor of 0.32 resulted in the best match. It can be clearly
Nomenclature
seen from the figure that once the friction factor is used to match the
a Acceleration
calculated value with that of the actual value, the calculated value can
b Unit binormal vector of wellbore trajectory
be used to predict ahead. Deviation from the prediction may provide
D Measured depth
valuable information such as cuttings build, well collapse, etc.
E Wellpath energy
The following simple procedure can be used as a guideline to estimate
F Tension, force
the friction factors:
g Gravitational constant
H Heat generated
Step 1: Collect the relevant drilling data. These include the well plan
k Heat transfer coefficient
(casing program, survey data, drillstring data, etc) and actual
L Lubricity coefficient of mud
drilling data (fluid data, mud logging data, hookload, etc)
N Pipe rotational speed
Step 2: For each hole section, calibrate the friction factor through
n Unit principal normal vector of wellbore trajectory
casing based on the actual drilling data while drilling out the
r Radius of the component
shoe.
R Radius of curvature
Step 3: Repeat Step 2 for the open hole in each hole section. Note the
t Unit tangent vector of wellbore trajectory
friction factor through the casing should be that from Step 2.
S Sticking coefficient
t Temperature
8. Conclusions T Torque
|Vts| Trip speed q
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Friction factor values should be used carefully in the calculation. They |Vrs| Resultant speed; = Vts2 + ω2
can be tuned with the available data. When calibrating friction factor with Wb Buoyed weight per unit length
limited actual data, a range bounded by minimum and maximum values |ω| Angular speed = diameter × π × 60 N

should be used. α Wellbore inclination,

Well planning should include torque and drag modeling with μ Coefficient of friction (COF)
worst-case friction factors to ensure that the drillstring can be μv Friction Factor
advanced, rotated, slid if oriented drilling is necessary, and pulled out σn Normal stress at the contact
of the hole. Similar modeling should be used to ensure that friction φ Friction angle
will not prevent the casing from being run, and that the casing can be θ Equilibrium angle

Fig. 8. Friction factor calibration.

266 R. Samuel / Journal of Petroleum Science and Engineering 73 (2010) 258–266

Δα Change in the wellbore inclination Ginzburg, V.B., 1989. Steel-Rolling Technology: Theory and Practice. Marcel Dekker.
Haduch, G.A., Samuels, D.A., 1994. Solution of common stuck pipe problems through
Δϕ Change in the wellbore azimuth the adaptation of torque drag calculations. Paper 27490-MS presented at the SPE/
Δt Average contact temperature IADC Drilling Conference, Dallas, Texas, USA, 15–18 February.
Johancsik, C.A., Friesen, D.B., Dawson, R., 1984. Torque and drag in directional wells —
prediction and measurement. J. Petrol. Technol. 36 (6), 987–992. doi:10.2118/
11380-PA.
Superscript Kireenkov, A.A., 2008. Coupled models of sliding and rolling friction. Dokl. Akad. Nauk
419 (6), 759–762.
k Constant
Liu, X., Samuel, R., 2009. Catenary well profiles for extended and ultra-extended-reach
kt Temperature dependent constant wells, Paper SPE 124313. Presented at the SPE Annual Technical Conference and
Exhibition, New Orleans, Louisiana, USA, 4–7 October.
Locke, H.A., Johnson, J.B., Jewkes, D.C., 1988. “Extended reach appraisal well: a case
history” SPE 17564. International Meeting on Petroleum Engineering, 1–4
Subscripts November 1988, Tianjin, China.
avg Average Maidla, E.E., Wojtanowicz, A.K., 1987. Field comparison of 2-D and 3-D methods for the
borehole friction evaluation in directional wells. SPE 16663, 62nd Annual Technical
cr Critical Conference and Exhibition of the SPE. Dallas TX Sept 27–30.
e Effective Mitchell, R.F., 2006. The Effect of friction on initial buckling of tubing and flow lines.
f Friction Paper SPE 99099 presented at the IADC/SPE Drilling Conference, Miami, Florida,
USA, 21–23 February., 2008, Miami, Florida, USA.
n Normal Mitchell, R., Samuel, R., 2007. How good is the torque drag model? Paper SPE 105068
k Kinetic presented at the SPE/IADC Drilling Conference, Amsterdam, The Netherlands,
kf Kinetic force 20–22 February. DOI: 10.2118/105068-MS.
Rabbat, B.G., 1985. Friction coefficient of steel on concrete or grout. J. Struct. Eng.
r Rolling
111 (3), 505–515. doi:10.1061/(ASCE)0733-9445(1985) 111:3(505.
rf Rolling force Rasmussen, B., Sorheim, J.O., Seiffert, E., Angeltvadt, O., Gjedrem, T., 1991. World record
rf Resultant speed in extended reach drilling, well 33/9-C10, Statfjord Field, Norway. Paper 21984-MS
presented at the SPE/IADC Drilling Conference, Amsterdam, The Netherlands,
s Static
11–14 March.
sf Static force Samuel, R., 2007. Downhole Drilling Tools — Theory and Practice for Students and
sf Trip speed Engineers. Gulf Publishing, Houston, Texas.
x x-direction Samuel, R., 2010. Friction factors: what are they for torque, drag, vibration, bottom hole
assembly and transient surge/swab analyses? IADC/SPE Drilling Conference and
v Variable Exhibition, 2–4 February 2010, New Orleans, Louisiana, USA.
0 Initial Samuel, R., Liu, X., 2009a. Advanced Drilling Engineering — Principles and Designs. Gulf
Publishing, Houston, Texas.
Samuel, R., Liu, X., 2009b. Wellbore tortuosity, torsion, drilling indices, energy: what do
they have to do with wellpath design? Paper SPE 124710 presented at the SPE
Acknowledgement Annual Technical Conference and Exhibition, New Orleans, Louisiana, USA, 4–7
October.
Sextro, W., 2002. Dynamical Contact Problems with Friction — Models, Methods,
The author would like to express appreciation to Halliburton for Experiments and Applications. Springer, Berlin.
the opportunity to publish this paper. Sonowal, K., Bennetzen, B., Wong, P., Isevcan, E., 2009. How continuous improvement
lead to the longest horizontal well in the world. SPE 119506 presented at the SPE/
IADC Drilling Conference and Exhibition, Amsterdam, The Netherlands, 17–19
References March.
Zuev, A.M., 1965. Effects of temperature on the coefficient of dry friction of titanium and
de Zwart, M.T.W. Theory on Loading, Torque, Buckling, Stresses, Strain, Twist, Burst and alloys VK6, T15K6, and VT15 on titanium. Russ. Phys. J. (New York: Springer) 8 (2),
Collapse. Shell Doc. FS920101. 103–104. doi:10.1007/BF00838598.