Journal of Petroleum Science and Engineering 127 (2015) 229–244

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

CFD–DEM approach to investigate the effect of drill pipe rotation on

cuttings transport behavior
Siamak Akhshik a, Mehdi Behzad a,n, Majid Rajabi b
a
Department of Mechanical Engineering, Sharif University of Technology, Iran
b
School of Mechanical Engineering, Iran University of Science and Technology, Iran

art ic l e i nf o a b s t r a c t

Article history: Increasing cuttings bed height is a serious concern during extended-reach well drilling. In order to
Received 27 August 2014 predict and prevent cuttings bed height increase, it is essential to study how the critical parameters
Accepted 9 January 2015 influence the cuttings transport, especially the drill pipe rotation effects on the cuttings transport
Available online 22 January 2015
process. In conventional models for cuttings transport, the dynamic behavior of particles due to drill pipe
Keywords: rotation is neglected or empirically simplified. This paper presents a coupled Computational Fluid
horizontal well drilling Dynamics and Discrete Element Method (CFD–DEM) approach to simulate the cuttings transport
cuttings transport considering the dynamic collision process. The fluid phase is treated as an Eulerian continuum and
CFD–DEM method described by CFD method. The cutting phase is modeled by DEM using a soft sphere approach for the
drill pipe rotation
particle collision dynamic. The continuum and dispersed phases are strongly coupled via the interaction
cutting–cutting collisions
forces such as the drag force, lift force and pressure gradient force, which is taken into account in the
developed CFD–DEM model. The model takes into account the collisions of cutting–cutting, cutting–drill
pipe and cutting–wall. Simulations are carried out for a number of laboratory-scale configurations,
showing good agreement with experiment data reported in literature. The numerical simulations show
that the drill pipe rotation builds cuttings presented in non-symmetric distribution along the hole and
significantly decreases the cutting volume concentration at the medium and high fluid inlet velocities
and lower range of drill pipe speeds. There is no additional contribution of drill pipe rotation after
reaching a critical speed at high fluid inlet velocities. At low fluid inlet velocities, increasing the drill pipe
rotation from 0 to 80 rpm dose not significantly affect the cuttings concentration; although an increase
in drill pipe rotation speed from 100 to 120 rpm provided a remarkable improvement. Furthermore, at
low inlet fluid velocities and horizontal inclinations up to 401, drill pipe rotation can significantly
decrease the cuttings concentration. The effect of drill pipe rotation on decreasing the cuttings
concentration is almost negligible at high inlet fluid velocities for any inclination.
& 2015 Elsevier B.V. All rights reserved.

1. Introduction well drilling have been extensively investigated in the literature

(Tomren et al., 1986; Ford et al., 1996, Nguyen and Rahman, 1996;
A key issue in extended-reach well drilling is the transport Larsen et al., 1997; Martins et al., 1999; Kamp and Rivero, 1999;
mechanism of the cuttings by drilling fluid. This difficulty arises Masuda et al., 2000; Ramadan et al., 2001; Li and Walker, 2001;
from the natural tendency of cuttings for leaving the state of Kelessidis and Bandelis, 2004; Ozbayoglu et al., 2005; Capo et al.,
suspension layer and forming a bed. This condition is caused by 2006; Chen et al., 2007; Espinosa-Paredes et al., 2007; Ozbayoglu
the gravitational force, which makes the solid fragments to be et al., 2010, 2012). Despite the importance of the drill pipe
deposited at the bottom of the annular section. Experimental and rotation, very limited studies are available in the literature for
numerical studies have shown that cuttings transport is affected cuttings transport under drill pipe rotation.
by many parameters, such as drill pipe diameter, drill pipe In experimental investigations, Bassal (1996) indicated drill
rotation, well inclination, drill pipe eccentricity, drilling fluid pipe rotation enhances hole cleaning more when the used mud
rheology, etc. The cuttings transport problem in extended-reach had a higher viscosity with smaller cuttings sizes or for a hole
angle of 651. Sanchez et al. (1999) examined the effect of drill pipe
rotation on hole cleaning during directional well drilling through
n
Corresponding author. Tel.: þ 98 21 66165509; fax: þ 98 21 66000021. over 600 tests. The drill pipe rotation effects were maximum at
E-mail address: [email protected] (M. Behzad). 901 inclination and minimum at 401 inclination. It also was shown

http://dx.doi.org/10.1016/j.petrol.2015.01.017
0920-4105/& 2015 Elsevier B.V. All rights reserved.
230 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

Nomenclature t time, s
Ta Taylor number
Cc particle concentration in specified region, m  3 u velocity vector, m s  1
CD drag coefficient uf ;inlet fluid inlet velocity
C DR rotational drag coefficient V cell volume of CFD mesh cell, m3
dp cutting diameter, m Vp volume of the p'th cutting, m3
Dp drill pipe diameter, m vt;pq tangential component of the relative velocity between
Dh hole diameter, m two contacted cuttings, m s  1
Drp dimensionless ROP vn;pq normal component of the relative velocity between
e coefficient of restitution two contacted cuttings, m s  1
E Young's modulus, Pa
E* equivalent Young's modulus, Pa Greek
FD drag force upon p'th cutting, N
FM rotational lift force (Magnus) exerted on p'th αf volume fraction of fluid phase
cutting, N αp volume fraction of particles
Fn;pq normal contact force between cuttings p and q, N δn;pq normal overlap between cuttings p and q, m
Fdn;pq normal damping force between cuttings p and q, N δt;pq tangential overlap between cuttings p and q, m
Ff ;i force exerting on the fluid in a CFD mesh cell, N θ angle of inclination
Ft;pq tangential contact force between cuttings p and q, N μ fluid viscosity, Pa s
Fdt;pq tangential damping force between cuttings p and q, N μr coefficient of rolling friction
Fr Froude number μs sliding friction coefficient
G shear modulus, Pa ρf density of fluid phase, kg m  3
Gn equivalent shear modulus, Pa ρp cutting density, kg m  3
g acceleration of gravity, m/s2 υ Poisons ratio
k consistency index, Pa sn ωdrillpipe drill pipe rotation speed
mp mass of cutting p, kg ωp angular velocity of cutting p, rad s  1
mn equivalent particle mass, kg ωpq relative angular velocity of cutting p to cutting q,
N number of cuttings in computational cell rad  s  1
n fluid behavior index ωf curl of the fluid velocity, rad s  1
p pressure, Pa Ω relative angular velocity of the cutting to the fluid,
rpq vector from the center of mass of cutting p to the rad s  1
contact point, m
Rn equivalent radius, m Subscripts
ReHB cutting Reynolds number
Res Reynolds number for shear flow
D drag
Rgen generalized Reynolds number
p cutting p
ROP rate of penetration, m/h
q cutting q
rpm revolutions per minute
f fluid
s eccentricity ratio
HB Herschel–Bulkley
Sf source term in the momentum equation, N m  3
M Magnus
Sn;pq normal stiffness between cuttings p and q, N m  1
n normal
St;pq tangential stiffness between cuttings p and q, N m  1
R Rayleigh
Tpt;q torque between cuttings p and q generated by tangen-
r rolling
tial forces, N m
s sliding
Tpr;q torque between cuttings p and q generated by radial
t tangential
forces, N m
z axial
TpDT drag torques acting on cutting p due to the fluid
μ viscosity
velocity

that drill pipe orbital motion is required for considerable devel- et al. (2008) experimentally investigated the effect of drill pipe
opment in cuttings transport and the drill pipe rotation decreased rotation on hole cleaning for water-based drilling fluids in hor-
the time needed to remove the cuttings from the wellbore. izontal and deviated wells. It was observed that the drill pipe
Mendez (2002) performed an experimental study on cuttings rotation had a significant effect on cuttings transport ability of the
transport with aerated muds and drill pipe rotation. The study fluid. It was also noticed that the mud viscosity may affect the hole
was focused on horizontal wells. Avila et al. (2004) performed cleaning at low rotation speeds. As the rotation speed was
experiments with cuttings transport in aerated mud drilling for increased, this effect diminished. A direct relation between the
different inclined angles and different drill pipe rotation speeds. hole inclination and drill pipe rotation speed was not identified.
They presented an empirical correlation to predict the critical gas Ahmed and Miska (2008) performed an experimental and theore-
flow rate and volumetric cuttings concentration. Duan et al. (2008) tical study on laminar flows of yield power-law fluids in concentric
studied the transport behavior of small cuttings in extended-reach and fully eccentric annulus with inner drill pipe rotation. Sorgun
drilling. In this study, the effects of cuttings size, drill pipe rotation, et al. (2011) carried out an experimental investigation on the
fluid rheology, flow rate, and hole inclination were investigated on effects of drill pipe rotation and cuttings transport on pressure
cutting transport experimentally and theoretically. Ozbayoglu loss. They observed that the drill pipe rotation enhanced the
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 231

cuttings removal process, significantly decreased the cuttings bed a continuum phase described by the local averaged Navier–Stokes
thickness for both low-viscosity and high-viscosity fluids, espe- equations on a computational cell scale, while the particles are
cially if the drill pipe was in a fully eccentric position. modeled as a discrete phase, solved by Newton's laws of motion
In modeling studies to predict the cuttings transport perfor- through the calculated flow field. The paper addresses the CFD–
mance considering the drill pipe rotation, Hemphill and Ravi DEM approach and its application for the modeling of fluid–
(2006) presented the effects of drill pipe rotation on pressure cuttings flows considering the drill pipe rotation. The various
drop and tangential velocity. They indicated that the drill pipe interaction forces that occur in cuttings transport, including shear
rotation may improve hole-cleaning when the drill pipe is lift force, rotational lift force, drag force and pressure gradient
eccentric. Naganawa and Nomura (2006) developed a cuttings force are discussed and formulated as sub-models.
transport simulator which predicts the transient behavior of
annular pressure, cuttings bed height, suspended cuttings con-
centration and phase velocities, along the whole trajectory of an
extended-reach drilling. Li et al. (2007) presented a one- 2. Mathematical formulation
dimensional transient mechanistic model for sensitivity analysis
of cuttings transport with conventional (incompressible) drilling A three dimensional (3D) model is prepared to study the flow
fluids in horizontal wells. The sensitivity analysis showed that the of cuttings in an deviated well drilling based on the computational
effects of drill pipe eccentricity, cuttings diameter and drilling fluid fluid dynamics (CFD) and discrete element method (DEM). In this
density on the cuttings bed height are not significant. Mishra model, the shapes of cuttings are generalized as spherical type.
(2007) showed that cuttings transport efficiency is significantly The fluid phase is considered as an incompressible fluid based on
improved for small cutting, more than large cutting as the drill an Eulerian approach. The dispersed phase (cuttings phase) is
pipe rotation is increased. Furthermore, the drill pipe rotation at a treated as a collection of individual cuttings whose movements are
fixed velocity reduced the pressure loss. Wang et al. (2009) used governed by applying Newton's second law. The continuum fluid
Computational Fluid Dynamics (CFD) to simulate solid–liquid two- phase and the dispersed cuttings phase are strongly coupled via
phase flows in wellbore annulus. The results indicated that the the interaction forces such as the drag force, lift force and pressure
quasi-spiral flow is the main flow pattern for liquid–solid trans- gradient force which are taken into account in the developed CFD–
port in the horizontal annulus. The influence of drill pipe rotation DEM model. Furthermore, the cuttings collisions are simulated by
on the cuttings transport was also investigated, as the drill pipe DEM. In the following sections, the governing equations of motion
rotation increased the disturbance of liquid to solid in the annulus. for the fluid and cutting phases are presented, followed by a
Sang-Mok et al. (2010a, 2010b) examined the solid–liquid mixture description of the input data for CFD and DEM simulations used in
upward hydraulic transport of solid particles in vertical and the present study.
inclined annuli with a rotating inner cylinder. In their numerical
study, the Eulerian approach for granular flow was used, which
allows the determination of the pressure and viscosity of the solid 2.1. Fluid phase
phase by using the commercial software program FLUENT instead
of using empirical correlation. The effect of drill pipe rotation on The 3D governing equations of motion of unsteady, viscous and
hole-cleaning during foam drilling was investigated by Duan et al. incompressible fluid phase are described by the local averaged
(2010). They found that the drill pipe rotation not only decreased Navier–Stokes equations on a computational cell scale with source
the cuttings concentration in a horizontal annulus but also terms. The mass conservation equation is expressed as
resulted in a considerable reduction in frictional pressure loss.


Guo et al. (2010) developed a three-layer unsteady cuttings ∂ αρf

transport model considering the drill pipe rotation for all sections þ ∇: αρf uf ¼ 0; ð1Þ
∂t
of extended-reach wells. The SETS method was used to solve the
numerical model. Ozbayoglu and Sorgun (2011) observed that the where uf is the fluid velocity, ρf is the fluid density and α is the
CFD model can estimate pressure drop better than slot flow volume fraction of the fluid phase. The momentum conservation
equations when compared with experimental data. Ofei et al. equation is given as
(2014) employed a CFD method to analyze the effects of fluid

velocity, annular diameter ratio (ranging from 0.64 to 0.90), drill ∂ αρf uf

pipe rotation, and fluid type on the prediction of pressure losses þ ∇: αρf uf uf ¼  α∇p þ α∇:τ  Sf þ αρf g; ð2Þ
∂t
and cuttings concentration for solid-fluid flow in eccentric hor-
izontal annular geometries. Sun et al. (2014) presented a CFD where p is the fluid pressure, τ is the viscous stress tensor and Sf is
simulation of the effects of drill pipe rotation on cuttings transport the volume-averaged (on a cell) interaction forces.
behavior in the complex structure well using an Euler multiphase The source term Sf for a particular computational cell was
model. They concluded that the pipe rotation has an important calculated by summing the fluid interaction forces on all cuttings
effect on the distribution of annular cuttings in the inclined present within that cell, and dividing by the volume of the fluid
section. computational cell
Although the positive effect of drill pipe rotation on hole
cleaning performance has been shown (Sanchez et al., 1999), most X
N
Sf ¼ Ff ;i =V cell ; ð3Þ
previous numerical models have not considered this influence, i¼1
specially considering the dynamic behavior of the cuttings. A
recently developed category of multiphase flow models uses the where N is the number of cuttings in the cell and V cell is the
Discrete Element Method (DEM), which determines the motion of volume of a fluid computational cell.
each individual particle separately. The instantaneous event of The relationship between the shear stress and shear rate in the
various kinds of movements and interactions of the particles with fluid phase is described by an appropriated rheological model.
each other and with the surfaces of the boundary are considered in Among the various rheological models proposed for the fluid flows
this method. In the combined computational fluid dynamics and through a circular section or an annulus, the Herschel–Bulkley
discrete element method (CFD–DEM), the fluid phase is treated as model is used. In this model, the dynamic viscosity is considered
232 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

as (Huilgol and You, 2005) 2.2.2. Forces and torques

8 τ0 2.2.2.1. Contact forces and torques. The contact forces on the p'th
< μ ¼ μyield ; when γ o μyield
> _
cutting from the q'th cutting is expressed as (see Fig. 1) (Di Renzo
n 
τ0 þ k γ_  ðτ0 =μyield Þ
n ð4Þ and Di Maio, 2004)
>
:μ¼ ; when γ_ Z μτ0
γ_ yield
Fpc;q ¼ Fn;pq þ Fdn;pq þ Ft;pq þ Fdt;pq ; ð7Þ
where k is the consistency factor, n is the power law exponent,
τ0 denotes the yield stress threshold, μyield means the yielding where Fn;pq denoted the normal contact force given as (Di Renzo
viscosity and γ_ is the strain rate magnitude. and Di Maio, 2004)

4 pffiffiffiffiffi 3=2
Fn;pq ¼ En Rn δn;pq ; ð8Þ
3
2.2. Cutting phase
in which δn;pq is the normal overlap, E* is the equivalent Young's
 h

 i  1
2.2.1. Cutting motion modulus En ¼ 1  υ2p =Ep þ 1  υ2q =Eq , R* is the
The translational motion of p'th cutting in fluid flow is conducted

1
by gravity buoyancy, contact forces (e.g., cutting–cutting, cutting–drill equivalent radius 2=dp þ 2=dq with Ep , υp , dp and Eq , υq , dq
pipe, cutting–wall) and interaction forces (e.g., drag force, shear lift being Young's modulus, Poisson ratio and radius of each element
force, rotational lift force) described as (Mei, 1992; Oesterleé and Bui
in contact. Fdn;pq means the normal damping force given by (Di
Dinh, 1998; Sommerfeld, 2000; Laín and Sommerfeld, 2008)
Renzo and Di Maio, 2004)
! !
∂up ρf X p rffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
mp ¼ mp g 1  þ Fc;q þ FD þFS þ FM þ Fp ; ð5Þ 5 ln e
∂t ρp q
Fdn;pq ¼  2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Sn;pq mn vn;pq ; ð9Þ
6 ln e þ π 2
2

where mp is the mass of p'th cutting, ρp is the cutting density, Fpc;q is


1
where mn is the equivalent cutting mass 1=mp þ 1=mq with mp
the contact force acting from q'th cutting on p'th cutting, FD denoted
the fluid drag force, FS refers to the shear lift force, FM means the and mq being the mass of each element in contact,
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rotational lift force or Magnus force, and FP is the fluid pressure Sn;pq ¼ 2E n
Rn δn;pq is the normal stiffness vn;pq is the normal
gradient force.
component of the relative velocity of contact point, and e is the
The rotational motion of mentioned p'th cutting is described as
coefficient of restitution.
(Kuang et al., 2009)
The tangential component of the contact force, Ft;pq , is
d X
p  expressed as (Di Renzo and Di Maio, 2004)
Ip ωp ¼ Tt;q þ Tpr;q þ TpDT ; ð6Þ 8
dt q <  δt;pq St;pq ; for Ft;pq o μs Fn;pq
Ft;pq ¼ ð10Þ
: μs Fn;pq jvt;pq j; for Ft;pq Z μs Fn;pq
vt;pq
where Tpt;q and Tpr;q are the torque vectors produced by the
tangential and normal contact force acting from q'th cutting on qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
the p'th cutting, respectively. Ip and ωp are the moment of inertia where St;pq ¼ 8Gn Rn δn;pq is the tangential stiffness in which Gn is
tensor and rotational velocity of p'th cutting, respectively. The the equivalent shear modulus, δt;pq is the tangential overlap, μs is
rotational motion was also affected by the drag torque, TpDT , that it sliding friction coefficient and vt;pq is the relative tangential
is produced by the slip-rotation. velocity of contact point.

Fig. 1. Contact forces acting on p'th particle in contact with q'th particle during collisions.
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 233

The tangential damping forces, Fdt;ij , is described as (Di Renzo 2000)

and Di Maio, 2004) 8
< Re0:5 þ Re ; 32 r Rer o 1000
12:9 128:4
r
C DR ¼ π
r
ð19Þ
: 64
Re ; Rer o 32
rffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi r
5 ln e
Fdt;pq ¼  2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi St;pq mn vt;pq ; ð11Þ in which the Reynolds number of cutting rotation is given by
6 ln2 e þ π 2
Rer ¼ ρdp Ω =μ.
2
Accordingly, the tangential torques acting on p'th cutting due to
cutting collision (q'th cutting) is expressed as (Kuang et al., 2009)

 2.2.2.3. Lift force. The lift forces including the shear lift force
Tpt;q ¼ rpq  Ft;pq þFdt;pq ; ð12Þ (Saffman) and the rotational lift force (Magnus) are perpendicular
to the direction of the relative velocity between the cuttings and
and the torque to resist rolling acting on p'th cutting due to cutting
fluid. The shear lift force (Saffman) exerted on p'th cutting is given
collision (q'th cutting) is expressed as (Kuang et al., 2009)
by (Mei, 1992; Sommerfeld, 2000; Saffman, 1965)
ωpq
Tpr;q ¼  μr rpq Fn;pq ; ð13Þ ρf π 3  
ωpq FS ¼ C LS dp uf  up  ωf ; ð20Þ
8
where rpq is a vector from the center of mass of cutting p to the where ωf is the curl of the fluid velocity (ωf ¼ ∇  uf ) and
contact point, μr is the rolling friction coefficient and ωpq is the
  (Sommerfeld, 2000)
relative angular velocity of cutting p to cutting q, ωp  ωq . The
torques Tpt;q and Tpr;q are generated by the tangential contact forces and 4:1126
C LS ¼ f ðReHB ; Res Þ; ð21Þ
the rolling friction, respectively. Re0:5
s

8

2.2.2.2. Drag force and drag torque. The drag force upon p'th < 1  0:3314β0:5 e  ReHB =10 þ 0:3314β0:5 ; ReHB r 40
f ðReHB ; Res Þ ¼
cutting can be represented by : 0:0524β :Re 0:5 ; ReHB 4 40
  HB
FD ¼ Ap uf  up ; ð14Þ
ð22Þ
where uf  up means the slip velocity and Ap denotes the fluid –  
in which β ¼ 0:5Res =ReHB 0:005 o β o0:4 and Reynolds number
cutting exchange coefficient, expressed as (Shook and Roco, 1991)
for shear flow is Res ¼ ρf dp ωf =μ.
2

3 ð1  αÞρl uf  up  1:65 The rotational lift force (Magnus) exerted on p'th cutting is
Ap ¼ C D α ; ð15Þ
calculated as (Oesterleé and Bui Dinh, 1998; Sommerfeld, 2000)
4 dp
 
where dp is the mean diameter of p'th cutting. In this study, the π 3 Ω  uf up
FM ¼ dp ρf C LM uf up ; ð23Þ
cuttings are assumed to be spherical and rigid, and thus the drag 8 Ω
coefficient, C D , is
where C LM is the coefficient of rotational lift (Sommerfeld, 2000),
24


given by
CD ¼ 1 þ 0:15 αRe0:687 ; ð16Þ
αReHB HB
 
Rer 0:4 0:3

in which the cutting Reynolds number is defined as C LM ¼ 0:45 þ  0:45 e  0:5684Rer ReHB ; ð24Þ
ReHB
RePL
ReHB ¼ ; ð17Þ
1 þ ð7π =24ÞBiHB
   2.2.2.4. Pressure gradient force. The pressure gradient force is
2n n
where BiHB ¼ τ0 =k dp = uf  up , RePL ¼ ρf uf up dp =k, τ0 is expressed as
the yield stress, k is the consistency factor and n is the power-law
FP ¼  V p ∇p; ð25Þ
component of the model.
Accordingly, the drag torques acting on cutting p due to the where V p indicates the volume of the p'th cutting and ∇p is the
fluid velocity is expressed as (Sommerfeld, 2000) gradient of the static pressure of the fluid phase at the location of
  cutting.
ρp dp 5
TpDT ¼ C DR Ω Ω; ð18Þ
2 2
2.2.3. Cutting–drill pipe collisions
where C DR is the rotational drag coefficient and Ω is the relative During the cutting–drill pipe collision, the forces and the
angular velocity of the cutting to the fluid (Ω ¼ ∇  uf =2  ωp ). torques on the cuttings are similar to those in cutting–cutting
The rotational drag coefficient, C DR , is defined as (Sommerfeld, collisions, as shown in Eq. (7) through Eq. (12).

Fig. 2. Computational diagram of CFD–DEM framework.

234 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

3. Computational conditions and information exchange in the 4. Model geometry

coupling CFD–DEM framework
The configuration of the problem consists of a finite length
Fig. 2 shows a diagram of the CFD–DEM framework. The fluid eccentric annulus created by two cylindrical bodies, deviated from
flow is simulated in the CFD model. When the solution converges, the vertical by the angle of θ. The interior cylinder rotates with a
the CFD results are transferred to the CFD–DEM coupling interface constant rotating velocity around its axis.
which leads to the calculation of the drag forces, lift forces and The two-phase flow including dispersed particles in a progres-
pressure gradient forces acting on particles. The calculated inter- sive fluid is entered interior the annulus from one end and exited
action forces are then transferred to the DEM solver which extracts from the other. Boundary conditions of the problem include the
the particle motion. Then, the CFD–DEM coupling interface takes specific value for the velocity inlet and the pressure outlet (see
the data of the new particle position and velocity from the DEM Fig. 3). The contact between the particles and the walls are taken
solver, updates the particle volume fraction in each computational no slip. The process is considered isothermal.
cell and finally, the data are transferred to the CFD solver, for the Values of the coefficient of restitution and the friction coefficients
next CFD time step. The CFD solver again iterates over the next for these particles were assumed to be 0.9 and 0.09, respectively
time step until the solution converges. The DEM time step should (Sang-Mok et al., 2010a, 2010b). The same friction coefficient was
be always less than the CFD time step. Here, the authors suggest also considered for interactions between the walls and the particles.
that the DEM time step be selected as 0.01 of the CFD time step for The geometrical parameters, the rheological constants and the
stability of solutions. operational characteristics are inspired from the experimental study
In this paper, the CFD model of fluid flow has been solved by by Osgouei (2010) and Tomren et al. (1986) as given in Table 1.
the Semi-Implicit Method for Pressure Linked Equations (SIM-
PLEC) algorithm and momentum equations of the fluid phase has
been solved by the second-order implicit time integration. The 5. Results and discussion
equations of motion of the dispersed particles in DEM simulation
has been solved by the explicit time integration method. 5.1. Validation of CFD–DEM computational model
The time steps for CFD and DEM solvers for satisfactory
convergence are selected 0.001 s and 0.00001 s, respectively. The In this section, the CFD–DEM model is compared with the
coupled framework has been run for a time period of 40 s which traditional CFD model of Osgouei and Ozbayoglu (2013), the
has took approximately 30 days of CPU time, running on a 32 core experimental data of Osgouei (2010) and the correlation formula
Intels-Xeons CPU processor (3.4 GHz) (Intel Corporation, Santa of Yu et al. (2007) in which the cutting concentration is explicitly
Clara, CA, USA) with 64 GB of RAM. related to the Froud Number, Taylor Number and Rate of Penetra-
The mesh independency of the problem is obtained by increas- tion (ROP) as follows:
ing the number of mesh elements in a trial and error manner in
Cc ¼ k1 F br 1 Rebgen
2
T br 3 Drpb4 tanhð1 þ b5 TaÞ; ð26Þ
order to obtain adequate convergence of the computations.
where Cc is the cuttings concentration, F r is the Froude number,
Rgen is the generalized Reynolds number, Drp is the dimensionless
ROP and Ta is the Taylor number. The other correlation parameters
are shown in Table 2.
Fig. 4 depicts the cutting concentration for selected fluid inlet
velocities, 0.61 m/s, 0.914 m/s, 1.219 m/s and 1.524 m/s, and for
two Rates of Penetration (ROP), 18.28 m/h and 24.38 m/h. Compar-
ison between the CFD–DEM model and the experimental data
illustrates the better prediction of cutting concentration extracted
from the CFD–DEM model in comparison with the CFD model and

Table 2
Parameters for cuttings concentration correlation.

k1 b1 b2 b3 b4 b5

0.062  1.31 0.157 0.165 0.045  0.0043

Fig. 3. Configuration of problem.

Table 1
Data used for verification of numerical solution and the experimental data reported by Tomren et al. (1986) and Osgouei (2010).

Parameter Variable Tomren et al. (1986) Osgouei (2010) Units

Drill string length L 12 7 m

Angle of inclination θ 00,20,40,60,80 90 deg
Pipe diameter Dp 48.26 47 mm
Hole diameter Dh 127 73.91 mm
Particle diameter dp 6.35 2 mm
Particle density dry density ρp 2619 2300 kg/m3
Fluid behavior index n 0.65 1 –
Consistency index k 0.28 0.001 Pa sn
Fluid inlet velocity uf ;inlet 0.58(1.9), 0.72(2.39), 0.1.01(2.86), 1.165(3.34), (3.82) 0.610(2), 0.914(3), 1.219(4), 1.524(5) m/s (ft/s)
Drill pipe rotation speed ωdrillpipe 0,50 0,80,100,120 rpm
Eccentricity ratio s 0.5 0.623 —
Rate of penetration ROP — 18.28 (60), 24.35 (80), 30.48 (100), 36.57 (120) m/h (ft/h)
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 235

Fig. 4. Comparison among experimental data and different models results. (a) ROP¼18.28 and (b)ROP ¼24.38.

Fig. 5. The Snapshots of cutting deposit along the well-hole.

the correlation formula results, especially for higher ROP. More- 5.3. Cutting flow pattern
over, the correlation formula demonstrates generally too higher
values of cutting concentrations for all fluid inlet velocities in Fig. 6 illustrates the cutting flow pattern through the well-hole.
comparison with the experimental data which makes its predic- Similar to the experimental reports (Doron and Barnea, 1993,
tions unreliable. 1995), three distinguishable regions may be identified.
The first region I is a stationary bed with small magnitude
stochastic cutting velocity fluctuations. The cuttings in the bed are
compactly packed, in persistent contact with each other, and
5.2. Cutting deposit along the well-hole clearly, do not contribute to the main flow.
The second region II, located above the stationary bed region I,
Fig. 5 demonstrates the process of cuttings deposit through the describes a moving bed consists of highly collided cuttings. Finally,
well hole. It is seen that the cuttings gradually accumulate and the region III is located above the region II with mean cuttings
bed grows up to a specific height that the velocity of fluid flow concentration considerably smaller than region II. In this region,
reaches a critical value. At this point, the fluid velocity is so high the cuttings remain suspended in the fluid and a homogeneous
that the cuttings can not deposit and the horizontal cutting flow pattern is observed.
transport is guaranteed.
As the time passes, the phenomenon of cutting deposit con-
tinues along the total length of well-hole. Ultimately, a steady- 5.4. Swaying phenomenon
state condition is achieved and the bed height becomes constant
throughout the entire length of the hole. Similar behavior has been The visualization of cuttings accumulation clearly illustrates
reported by experimental work of Osgouei (2010). the swaying of cuttings from left to the right of the flow domain in
236 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

Fig. 7. This circumferential buildup of cuttings (i.e., effective bed) is

due to the drill pipe rotation.
Furthermore, the velocity profile of cuttings depict a non-uniform
pattern which the maximum values belong to the cuttings in the
right-side and vice versa. The simulation result is in good agreement
with the experimental observations of Tomren et al. (1986).

5.5. Impact of fluid inlet velocity on fluid velocity pattern

The fluid inlet velocity is a key factor for controlling the

formation of cuttings bed height. The velocity magnitude contours
of fluid phase for a constant drill pipe rotation speed (80 rpm)
and for four different fluid inlet velocities, 0.61 m/s, 0.914 m/s,
1.219 m/s, 1.524 m/s, are shown in Fig. 8. As it is seen in Fig. 8(a),
the velocity magnitude core is shifted counter clockwise in
consistency with the direction of drill pipe rotation, but for higher
fluid inlet velocities in Fig. 8(b)–(d), this shift is not considerable.
This is due to this fact that for lower inlet velocities, the cutting
concentration (bed height) is higher so that the influence of drill
pipe rotation (swaying phenomenon) on the profile of bed and the
Fig. 6. Three distinguishable regions for horizontal well drilling.
contour of fluid velocity is more observable.

5.6. Impact of fluid inlet velocity on cutting volume fraction

The cuttings volume fraction is characterized by the number

of cuttings per unit volume. The distribution of cuttings volume
fraction for a constant drill pipe rotation speed (80 rpm) and
different fluid inlet velocities are shown in Fig. 9. Obviously,
the increase in fluid inlet velocity leads to decrease in the height
of cuttings bed. Furthermore, the drill pipe rotation at low fluid inlet
velocities significantly affects the distribution of cuttings volume
fraction, resulting in a higher buildup of cuttings bed on one side.
This result is in consistency with the swaying occurrence and the
Fig. 7. Visualization of the cuttings swayed with the drill pipe rotation. fluid velocity contours depicted in Fig. 8.

Fig. 8. Effect of fluid inlet velocity on velocity magnitude contour in annulus for constant drill pipe rotation speed (80 rpm). (a) 0.610m/s, (b) 0.914m/s, (c) 1.219m/s and
(d) 1.524m/s.
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 237

However, increasing the fluid inlet velocity reduces the effect of As this figure shows, the cutting concentration decreases
drill pipe rotation speed on cuttings concentration distribution. It significantly for rotating drill pipe in comparison with stationary
is noteworthy that the drill pipe rotation increases the thickness of drill pipe. As the drill pipe starts to rotate, a circumferential
the cuttings moving region (II) on the side of the annulus in motion is created in the cutting–fluid two-phase flow. This
opposite to the direction of rotation. By increasing the fluid inlet circumferential motion changes the cutting–fluid two-phase dis-
velocity this thickness is further increased. tribution in the annulus and higher buildup of cuttings bed is seen
on one side.
5.7. Impact of drill pipe rotation speed on the cuttings concentration Fig. 11 shows the effect of drill pipe rotation speed on the
cuttings concentration for different fluid inlet velocities and
Fig. 10 depicts the cutting volume concentration for different constant rate of penetration (ROP ¼ 24.38 m/h). As it is observed,
drill pipe rotation speeds, 0 rpm, 80 rpm, 100 rpm, 120 rpm and the drill pipe rotation improves the hole-cleaning, especially for
for constant fluid inlet velocity 0.914 m/s. low fluid inlet velocities.

Fig. 9. Effect of fluid inlet velocity on cuttings concentration in annulus for constant drill pipe rotation speed (80 rpm). (a) 0.610m/s, (b) 0.914m/s, (c) 1.219m/s and
(d) 1.524m/s.

Fig. 10. Distribution of cutting volume fraction at different drill pipe rotations. (a) RPM¼ 0, (b) RPM¼ 80, (c) RPM¼100 and (d) RPM ¼120.
238 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

For stationary drill pipe, a sudden drop is observed for cutting is due to the significant reducing effect of drill pipe rotation on the
concentration due to the transition of fluid inlet velocity from cutting concentration at low fluid inlet velocities.
0.914 m/s to 1.219 m/s, both in experimental data and the CFD–DEM As all the figures show, the cutting concentration predicted
model results. This trend is annihilated for rotating drill pipe which by the CFD–DEM model depicts higher values in comparison

Fig. 11. Cuttings concentration vs. fluid inlet velocity for ROP ¼24.35 m/h and rpm¼ 0, 80, 100 and 120.

Fig. 12. Fluid phase velocity contours for 0.914 m/s fluid inlet velocity and different drill pipe rotations.
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 239

Fig. 13. Distribution of cutting volume fraction at different ROP and constant drill pipe rotations (80 rpm). (a) ROP ¼ 18.28 m/hr, (b) ROP ¼ 24.35 m/hr, (c) ROP ¼ 30.48 m/hr
and (d) ROP¼36.57 m/hr.

Fig. 14. Cuttings concentration vs. fluid inlet velocity for 80 rpm, ROP¼18.28, 24.35, 30.48 and 36.57 (m/h). (a) ROP¼18.28 m/hr, (b) ROP ¼ 24.35 m/hr, (c) ROP ¼ 30.48 m/hr
and (d) ROP¼36.57 m/hr.

with the experimental data. This systematic error might be due to 5.8. Impact of drill pipe rotation speed on contours of fluid velocity
the overestimated values for the coefficient of restitution and the magnitude
friction which needs to be truly identified or updated in future
works. Unfortunately, Osgouei (2010) did not measure and report The contours of velocity magnitude of fluid phase for three
these critical characteristics of particles. different drill pipe rotation speeds are shown in Fig. 12. Comparing
240 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

the fluid phase velocity distributions for drill pipe rotation speeds 5.9. Impact of rate of penetration (ROP) on cutting phase volume
reveal that the maximum velocity of fluid flow decreases for fraction
higher drill pipe rotation speed. This may be interpreted by this
fact that for higher rotation speeds, the smaller bed layers exist Fig. 13 depicts the contour of cutting concentration for selected
and therefore, the effective section of fluid flow increase. ROPs. As this figure show, there is a direct relation between the

Fig. 15. Moving cutting velocity vs. fluid inlet velocity for ROP¼24.35 m/h (80 ft/h) and different drill pipe rotation speeds. (a) RPM¼0, (b) RPM¼80, (c) RPM¼ 100 and
(d) RPM¼ 120.

Fig. 16. Moving cutting velocity vs. fluid inlet velocity for 80 rpm and different ROP. (a) ROP¼ 18.28 m/hr, (b) ROP¼ 24.35 m/hr, (c) ROP¼ 30.48 m/hr and (d) ROP¼36.57 m/hr.
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 241

cuttings concentration present in the hole with the injection rate

of cuttings or rate of penetration (ROP). As ROP is increased, the
cuttings volume fraction increases in the hole. This demonstrates
that it is necessary to compromise between penetration rates and
hole-cleaning.

5.10. Impact of ROP on the cuttings concentration

Fig. 14 illustrates the ROP effects on the cuttings concentration

for different fluid inlet velocities. The drill pipe rotation speed is
kept constant as 80 rpm. The results are compared with the
experimental data reported by Osgouei (2010). As it is seen in this
figure, the cuttings concentration in the hole increase for higher
values of ROP. This effect is negligible for higher values of the fluid
inlet velocity. It is also noteworthy that the differences between
the CFD–DEM results and the experimental data increases for
lower fluid inlet velocities and higher ROPs.

5.11. Impact of drill pipe rotation speed on moving cutting velocity

One of the abilities of the CFD–DEM model is the prediction of

the moving cuttings velocity in the region just above the surface of
the bed. This characteristic is rarely provided in the literature due
to the previous models limitations. Fig. 15 illustrates the moving
cutting velocity for previously selected values of fluid inlet
velocities and drill pipe rotation speeds. The results are compared
with the experimental data reported by Osgouei (2010). Generally,
the CFD–DEM model overestimates the moving cutting velocity
with respect to experimental data.
The drill pipe rotation has a minor effect on the moving
cuttings velocity for a constant ROP. As the drill pipe rotation Fig. 18. Effects of inclination on cuttings transport mechanism. (a) Inlet Fluid
Velocity¼ 0.582 m/s and (b) Inlet Fluid Velocity ¼1.164 m/s.
speed increases, a slight decrease in moving cuttings velocity is

Fig. 17. Comparison of the cuttings concentration predicted by CFD–DEM model with the measured ones by Tomren et al. (1986). (a) Fluid Inlet Velocity¼ 0.58 m/s, (b)Fluid
Inlet Velocity ¼0.72 m/s and (c) Fluid Inlet Velocity¼ 1.16 m/s.
242 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

observed which can be physically interpreted due to the decreas- the cutting moving velocity increases both in CFD–DEM results
ing influence of rotation speed on the bed height which leads to an and the experimental data. This may be linked to the increase of
increase in the effective fluid flow area and therefore, causing a bed height for higher ROPs and therefore, increase in fluid flow
decrease in the fluid flow velocity. The decrease in fluid flow velocity.
velocity due to the increase in drill pipe rotational speed is shown It is observed that for higher ROPs, the difference between the
and discussed in Section 5.8. CFD–DEM results and the experimental data increases which may
be due to the cumulative effects of errors generated by un-true
values of friction and restitution coefficients in the model, for
5.12. Impact of ROP on moving cutting velocity higher ROPs.

Fig. 16 illustrates the moving cutting velocity for previously

selected fluid inlet velocities and ROPs. The drill pipe rotation 5.13. Impact of well inclination on cuttings transport
speed is kept constant as 80 rpm. The results are compared with
the experimental data reported by Osgouei (2010). Generally, the The comparison of the cuttings concentration predicted by the
CFD–DEM model overestimates the moving cutting velocity with current model with the measured data reported by Tomren et al.
respect to experimental data. (1986) is shown in Fig. 17. As seen in this figure, the model
Obviously, the cutting moving velocity increases with the predictions are in good agreement with the experimental data for
increase of fluid inlet velocity. Moreover, as the ROP increases, a fluid inlet velocity greater than 0.78 m/s. For the fluid inlet

Fig. 19. Cuttings concentration vs. annulus inclination for different drill pipe rotation speed and fluid inlet velocity. (a) Fluid Inlet Velocity ¼0.58 m/s, 0RPM, (b)Fluid Inlet
Velocity ¼0.58 m/s, 50RPM, (c) Fluid Inlet Velocity ¼0.72 m/s, 0RPM, (d) Fluid Inlet Velocity ¼ 0.72 m/s, 50RPM, (e) Fluid Inlet Velocity ¼ 1.16 m/s, 0RPM and (f) Fluid Inlet
Velocity ¼1.16 m/s, 50 RPM.
 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244 243

velocities lower than 0.52 m/s and well inclination greater than The cuttings concentration and velocity contours have been
401, the model slightly overestimates the cuttings concentration. computed, showing good agreement with the experimental data
Three significant type of cutting flow have been reported in reported in literature.
experimental investigation on the effects of well-bore inclination This developed numerical model has been used to better
(Clark and Bickham, 1994): suspension, rolling and a combination understand the cuttings transport mechanism and to describe
of rolling and suspension. The suspension mechanism is occurred the hole cleaning process more accurate than the other CFD
in inclination angle near the vertical state (lower angles). The models and correlation formula due to its much more real
rolling mechanism is dominated in inclination angles near the modeling of discrete suspended particles via coupled DEM. The
horizontal state (higher angles). The transition of the suspension model has been shown to be reliable as a prediction tool for
mechanism to the rolling one takes place in medium inclination understanding the physical phenomenon of cuttings transport and
angles. Up to now, the theoretical models just consider a pure can therefore be used as an assessment tool for future cuttings
rolling or a pure suspension type of particle flow and there is no transport studies.
prediction of transition which includes the combination type of Also, the results demonstrated that the CFD–DEM method
motion (i.e., more consistent with the practice). As seen in Fig. 18, offers a cost-effective approach to study the complicated multi-
the solution performed by the proposed CFD–DEM method can phase flow in extended-reach well drilling. However, it is worth
address this scenario for our mentioned example. Particularly, it is mentioning that the results presented in this study have been
noticed that the range of medium inclination angles that the obtained under simplified conditions; for instance, the drill pipe
transition phenomenon between the suspension mechanisms to strictly rotates around its own axis, the cuttings have spherical
rolling one occurred, is influenced by the inlet velocity of fluid. shape, etc. Further studies on model refinement and investigation
As the inlet velocity increases, this transition happens in higher are needed to better understand the cuttings transport process.
angles. The future direction of this study will be toward the investiga-
Fig. 19 examines the cuttings concentration for selected incli- tion of the effects of cuttings shape and size on the cuttings
nation angles (01,201,401,601,801), different inlet velocities (cases transport process.
(a) 0.58 m/s, (b) 0.72 m/s, (c) and 1.16 m/s), non-rotating and
rotating velocity ω ¼ 50 rpm. For lowest inlet velocity in the cases
(a) and (b), the cutting concentration faces with a jump (remark- References
able increase) at inclination angles near 401 which this happening
has been reported experimentally by Tomren et al. (1986). This Ahmed, R., Miska, S., 2008. Experimental study and modeling of yield power-law
fluid flow in annuli with drillpipe rotation. SPE Paper 112604. In: Proceedings of
occurrence is also observed in qualitative demonstration of cutting
the IADC/SPE Drilling Conference in Orlando, March 4–6.
concentration in Fig. 19(a) and (b). Avila, R., Pereira, E., Miska, S., Takach, N., Saasen, A., 2004. Correlations and analysis
For any inlet velocity with 01 and 401 inclination (from the of cuttings transport with aerated fluids in deviated wells. In: Proceedings of
vertical axis), as the drill pipe rotation speed is increased, the the IADC/SPE Drilling Technology Conference and Exhibition. Dallas, Texas,
USA, March 2–4.
cuttings concentration remains nearly constant. As seen in this Bassal, A.A., 1996. The Effect of Drillpipe Rotation on Cuttings Transport in Inclined
figure, at an inlet fluid velocity of 0.58 m/s and inclinations Wellbores (MS Thesis). University of Tulsa.
40–80o, the drill pipe rotation significantly decreases the cuttings Capo, J., Yu, M., Miska, S., Takach, N.E., Ahmed, R., 2006. Cuttings transport with
aqueous foam at intermediate inclined wells. SPE Drill. Complet. 21 (2), 99–107.
concentration. However, the additional contribution of drill pipe Chen, Z., Ahmed, R.M., Miska, S., Takach, N.E., Yu, M., Pickell, M.B., Hallman, J., 2007.
rotation on decreasing the cuttings concentration is almost neg- Experimental study on cuttings transport with foam under simulated horizon-
ligible when the inlet fluid velocity is 0.72 m/s or 1.164 m/s and the tal downhole conditions. SPE Drill. Complet. 22 (4), 304–312.
Clark, R.K., Bickham, K.L., 1994. A mechanistic model for cuttings transport. Paper
inclination is close to horizontal. SPE 28306. In: Proceedings of the SPE Annual Technical Conference and
Exhibition. New Orleans, LA, September 25–28.
Di Renzo, A., Di Maio, F.P., 2004. Comparison of contact-force models for the
simulation of collisions in DEM-based granular flow codes. Chem. Eng. Sci. 59
6. Conclusions (3), 525–541.
Doron, P., Barnea, D., 1993. A three-layer model for solid–liquid flow in horizontal
In the presented study, a coupled CFD–DEM approach has been pipes. Int. J. Multiph. Flow 19 (6), 1029–1043.
Doron, P., Barnea, D., 1995. Flow pattern maps for solid–liquid flow in pipes. Int. J.
presented to simulate the effects of the drill pipe rotation, well
Multiph. Flow 22 (2), 273–283.
inclination, rate of penetration and fluid inlet velocity on the hole Duan, M., Miska, S., Yu, M., Takach, N., Ramadan, A., 2008. Transport of small
cleaning process. In this approach, DEM and CFD have been used cuttings in extended-reach drilling. SPE Drill. Complet. 23 (3), 258–265.
to solve the Newton's law and the locally averaged Navier–Stokes Duan, M., Miska, S., Yu, M., Takach, N., Ahmed, R., Hallma, J., 2010. Experimental
study and modeling of cuttings transport using foam with drill pipe rotation.
equations for the particle/fluid flow, concurrently. SPE Drill. Complet. 25 (3), 352–362.
The interaction between the fluid and particles have been Espinosa-Paredes, G., Salazar-Mendozab, R., Cazarez-Candia, O., 2007. Averaging
considered by calculating the exchange of forces such as shear lift model for cuttings transport in horizontal wellbores. J. Pet. Sci. Eng. 55 (3–4),
301–316.
force, rotational lift force, drag force and buoyancy force via two- Ford, J.T., Oyeneyin, M.B., Peden, J.M., Larrucia, M.B., Parker, D., 1996. A new MTV
way coupling approach. Moreover, the interactions of particle– computer package for hole-cleaning design and analysis. SPE Drill. Complet. 11
particle, particle–wall and particle–drill pipe are taken into (3), 168–172.
Guo, X., Wang, Z., Zhi-hui, L., 2010. Study on three-layer unsteady model of cuttings
account via Hertz–Mindlin model. transport for extended-reach well. J. Pet. Sci. Eng. 73 (1-2), 171–180.
It has been shown that the proposed model is successful in Hemphill, T., Ravi, K., 2006. Pipe rotation and hole cleaning in an eccentric annulus.
prediction of some experimentally reported phenomena in the In: Proceedings of the IADC/SPE Drilling Conference. Miami, Florida, USA,
February 21–23.
particle transportation process such as Huilgol, R.R., You, Z., 2005. Application of the augmented Lagrangian method to
steady pipe flows of Bingham, Casson and Herschel–Bulkley fluids. J. Non-
 The transportation mechanism of particles: suspension type at Newton. Fluid Mech. 128 (1), 126–143.
Kamp, A.M., Rivero, M., 1999. Layer Modeling for cuttings transport in highly
low well inclination angles (near vertical), rolling type at high
inclined wellbores. In: Proceedings of the SPE 53942, Latin American and
inclination angles (near horizontal) and their combination at Carribean Petroleum Engineering Conference. Caracas, Venezulea, April 21–23.
medium inclination angles. Kelessidis, V.C., Bandelis, G.E., 2004. Flow patterns and minimum suspension
 Jump behavior in particle concentration for relatively low fluid velocity for efficient cuttings transport in horizontal and deviated wells in
coiled-tubing drilling. SPE Drill. Complet. 19 (4), 213–227.
inlet velocities at medium inclination angles (approximately Kuang, S.B., Yu, A.B., Zou, Z.S., 2009. Computational study of flow regimes in vertical
40o in our configuration). pneumatic conveying. Ind. Eng. Chem. Res. 48 (14), 6846–6858.
244 S. Akhshik et al. / Journal of Petroleum Science and Engineering 127 (2015) 229–244

Laín, S., Sommerfeld, M., 2008. Euler/Lagrange computations of pneumatic con- Ozbayoglu, M.E., Saasen, A., Sorgun, M., Svanes, K., 2008. Effect of pipe rotation on
veying in a horizontal channel with different wall roughness. Powder Technol. hole cleaning for water-based drilling fluids in horizontal and deviated wells.
184 (1), 76–88. In: Proceedings of the ADC/SPE Asia Pacific Drilling Technology Conference and
Larsen, T.I., Pilehvari, A.A., Azar, J.J., 1997. Development of a new cutting transport Exhibition. Jakarta, Indonesia, August 25–27.
model for high angle wellbores including horizontal wells. SPE Drill. Complet. Ozbayoglu, M.E., Saasen, A., Sorgun, M., Svanes, K., 2010. Critical fluid velocities for
12 (2), 129–136. removing cuttings bed inside horizontal and deviated wells. Pet. Sci. Technol.
Li, J., Walker, S., 2001. Sensitivity analysis of hole cleaning parameters in directional 28, 594–602.
wells. SPE J. 6 (4), 356–363. Ozbayoglu, M.E, Sorgun, M., 2011. Predicting frictional pressure loss during
Li, Y., Bjorndalen, N., Kuru, E., 2007. Numerical modeling of cuttings transport in horizontal drilling for non-Newtonian fluids. Energy Sources, Part A 33 (1),
horizontal wells using conventional drilling fluids. J. Can. Pet. Technol. 46 (7), 631–640.
9–15. Ozbayoglu, M.E., Osgouei, R.E., Ozbayoglu, M.A., Yuksel, E.H., 2012. Hole-cleaning
Martins, A.L., Santana, M.L., Gaspari, E.F., 1999. Evaluating the transport of solids performance of gasified drilling fluids in horizontal well sections. SPE J. 17 (3),
generated by shale instabilities in ERW drilling. SPE Drill. Complet. 14 (4), 912–923.
254–259. Ramadan, A., Skalle, P., Johansen, S.T., Svein, J., Saasen, A., 2001. Mechanistic model
Masuda, Y., Naganawa, S., Doan, Q., Yonezawa, T., Kobayashi, A., Kamp, A., 2000. for cuttings removal from solid bed in inclined channels. J. Pet. Sci. Eng. 30 (3–4),
Experimental Study to Determine Critical Flow Rate of Cuttings Transport 129–141.
During Underbalanced Drilling, IADC/SPE 62737. Asia Pacific Drilling Technol- Saffman, P.G., 1965. The lift on a small sphere in a slow shear flow. J. Fluid Mech. 22
ogy, Kuala Lumpur, Malaysia, September 11–13. (1), 385–400.
Mei, R., 1992. An approximate expression for the shear lift force on a spherical Sanchez, R.A., Azar, J.J., Bassal, A.A., 1999. Effect of drill pipe rotation on hole
particle at finite Reynolds numbers. Int. J. Multiph. Flow 18 (1), 145–147. cleaning during directional-well drilling. SPE J. 4 (2), 101–108.
Mendez, J., 2002. An Experimental Study of Cuttings Transport in Horizontal Wells Sang-mok, Han, Nam-sub, Woo, Young-kyu, Hwang, 2010a. Solid–liquid mixture
with Aerated Muds and Drill Pipe Rotation (M.S. thesis). The University of Tulsa. flow through a slim hole annulus with rotating inner cylinder. J. Mech. Sci.
Mishra, N., 2007. Investigation of Hole Cleaning Parameters using Computational Technol. 23 (2), 569–577.
Fluid Dynamics in Horizontal and Deviated Wells (M.S. thesis). West Virginia Sang-mok, Han, Nam-sub, Woo, Young-kyu, Hwang, 2010b. Solid–liquid hydro-
University. dynamics in a slim hole drilling annulus. J. Pet. Sci. Eng. 70 (3–4), 308–319.
Naganawa, S., Nomura, T., 2006. Simulating transient behavior of cuttings transport Shook, C.A., Roco, M.C., 1991. Slurry Flow: Principles and Practice, 3rd Edition
over whole trajectory of extended-reach well. In: Proceedings of the IADC/SPE Massachusetts, Boston.
Asia Pacific Drilling Technology Conference. Bangkok, Thailand, November 13–15. Sommerfeld, M., 2000. Theoretical and Experimental Modelling of Particulate
Nguyen, D., Rahman, S.S., 1996. A three-layer hydraulic program for effective Flows. Technical Report Lecture Series, von Karman Institute for Fluid
cuttings transport and hole cleaning in highly deviated and horizontal wells. Dynamics, 20–23.
IADC/SPE 36383, Asia Pacific Drilling Technology Kuala Lumpur-Malaysia, Sorgun, M., Aydin, I., Ozbayoglu, M.E., 2011. Friction factors for hydraulic calcula-
September 9–11. tions considering presence of cuttings and pipe rotation in horizontal/highly-
Oesterleé, B., Bui Dinh, T., 1998. Experiments on the lift of a spinning sphere in a inclined wellbores. J. Pet. Sci. Eng. 78 (2), 407–414.
range of intermediate Reynolds numbers. Exp. Fluids 25 (1), 16–22. Sun, X., Wang, K., Yan, T., Shao, S., Jiao, J., 2014. Effect of drillpipe rotation on
Ofei, T.N., Irawan, S., Pao, W., 2014. CFD method for predicting annular pressure cuttings transport using computational fluid dynamics (CFD) in complex
losses and cuttings concentration in eccentric horizontal wells. J. Pet. Eng. 2014 structure wells. J. Pet. Explor. Prod. Technol. 4 (3), 255–261.
(1), 1–16. Tomren, P.H., Iyoho, A.W., Azar, J.J., 1986. Experimental study of cuttings transport
Osgouei, R.E., 2010. Determination of Particles Transports Properties of Gasified in directional well. SPE Drill. Eng. 1 (1), 43–56.
Drilling Fluids (Ph.D. thesis). Middle East Technical University. Wang, Zhi-ming, Guo, Xiao-le, Li, Ming, Hong, Yu-kui, 2009. Effect of drill pipe
Osgouei, R.E., Ozbayoglu, M.E., 2013. CFD simulation of solids carrying capacity of a rotation on borehole cleaning for extended reach well. J. Hydrodro 21 (3),
Newtonian fluid through horizontal eccentric annulus. In: Proceedings of the 366–372.
ASME Fluids Engineering Division Summer Meeting. USA, July 7–11. Yu. M., Takach, N.E., Nakamura, D.R., Shariff, M.M., 2007. An experimental study of
Ozbayoglu, M.E., Miska, Z.S., Reed, T., Takach, N., 2005. Using foam in horizontal hole cleaning under simulated downhole conditions. In: Proceedings of the SPE
well drilling: a cuttings transport approach. J. Pet. Sci. Eng. 46 (4), 267–282. Annual Technical Conference and Exhibition. USA, November 11–14.