Journal of Petroleum Science and Engineering 70 (2010) 308319

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 e v i e r. c o m / l o c a t e / p e t r o l

Solidliquid hydrodynamics in a slim hole drilling annulus

Sang-Mok Han a, Young-Kyu Hwang a, Nam-Sub Woo b, Young-Ju Kim c,
a
b
c

Department of Mechanical Engineering, Sungkyunkwan University, 300, Cheoncheon-dong, Jangan-gu, Suwon, 440-746, Republic of Korea
Korea Institute of Construction and Technology, 1199, Simondae-ro, Ilsanseo-gu, Goyang, 411-712, Republic of Korea
Korea Institute of Geoscience and Mineral Resources, 30, Gajeong-dong, Yuseong-gu, Daejeon, 305-350, Republic of Korea

a r t i c l e

i n f o

Article history:
Received 13 April 2009
Accepted 3 December 2009
Keywords:
Slim hole drilling
Drilling uids
Helical ow
Inclined annulus
Rotating cylinder

a b s t r a c t
We examine solidliquid mixture upward hydraulic transport of solid particles in vertical and inclined annuli
with a rotating inner cylinder. Lift forces acting on uidized particles play a central role in many important
applications such as the removal of drill cuttings in horizontal drill holes, sand transport in fractured
reservoirs, and sediment transport. Annular uid velocities in our study varied from 0.4 m/s to 1.2 m/s. The
effect of annulus inclination and drill pipe rotation on the carrying capacity of drilling uid, particle rising
velocity, and pressure drop in a slim hole annulus were measured for fully developed ows of water, and for
aqueous solutions of sodium carboxymethyl cellulose (CMC) and bentonite. For higher particle feed
concentrations, the hydraulic pressure drop of the mixture ow increased due to friction between the wall
and solids, or among solids. Comparing our numerical and experimental results allowed the assessment and
further development of existing two-phase numerical models and grid constructions.
2009 Elsevier B.V. All rights reserved.

1. Introduction
Among the various industrial operations involved with multiphase
systems, agitation of solidliquid systems is quite common. Examples
include catalytic reactions, oil well drilling, sand transport in fractured
reservoirs, etc. Although there are many industrial applications of
solidliquid mixture ows in technology, the available knowledge
about particle ows is incomplete due to difculties encountered in
analyzing these complex systems.
Due to the variety of parameters related to solidliquid mixture
ow, many conditions must be included when considering the
inuence of variables related to the transportation of drill cuttings.
The study of transportation drill cuttings is an engineering problem
that is classied as multiphase ow involving solid particles, liquids,
and gasses. The study of drill cuttings transportation ability can be
divided into the dynamics of the particles within the uid, and the
action of drilling irrigation. When an oil well is drilled, it is necessary
to transport the cuttings up to the surface. To this end, drilling uid is
pumped through the center of the drill pipe and back up to the surface
through the annular gap between the drill pipe and the drilled hole.
The ow up to the annulus can be laminar or turbulent depending on
the situation.
With the increase in the use of deviated, highly deviated, and long
reach drilling, greater consideration must be given to the uid
mechanics of transportation in situations where none, or only a small
Corresponding author. Tel.: + 82 42 868 3092; fax: + 82 42 868 3220.
E-mail address: kyjp7272@kigam.re.kr (Y.-J. Kim).
0920-4105/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.petrol.2009.12.002

component, of the bulk ow acts against the tendency for cuttings to

drop out of suspension and form a bed on the low side of the annulus.
This is even more important in slim hole drilling, which has small gap
between the drill pipe and drilled hole. Therefore, numerous
mathematical and empirical models for the prediction of cuttings
transport in horizontal and directional wells have been developed.
Tomren et al. (1986) and Ford et al. (1991) performed experimental
studies on cuttings transport in an inclined annulus. They observed
the existence of different layers that might occur during the mud ow,
and cuttings in the annulus. They noted that rotation of the drill pipe
has little effect unless the annulus is eccentric and the drill pipe is on
the low side of a horizontal geometry.
Interest has been growing in the interaction between particles and
the local ow structure in particulate solidliquid mixture ow. Pigott
(1941) discussed the application of Stoke's law for laminar ow, and
Rittinger's formula for turbulent ow, to a particle settling velocity
calculation. He concluded that high uid viscosity was not necessary
and suggested that laminar ow in the annulus would lead to more
efcient cleaning. For trouble-free operation, he also recommended
that the volumetric cuttings concentration in the annulus be kept to
less than 5%.
Sifferman et al. (1974) found that annular velocity and uid's
rheological properties are the most important factors inuencing the
transportability of a uid. According to their study, other variables
such as particle size, drill pipe rotation, and drill pipe eccentricity have
only moderate effects on carrying capacity.
Sellgren (1982) and Ozbelge and Beyaz (2001) discussed the
solidliquid mixture ow pressure drop and the choice of operating

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

Nomenclature
Dh
dp/dz
e
m
N
P
R1
R2
Re
vz
z
CVT

hydraulic diameter, 2(R2 R1)

pressure loss (Pa/m)
eccentricity (mm)
ratio of the eccentricity to the difference of radius
rotational speed (rev/min)
difference of pressure (Pa)
radius of inner cylinder (mm)
radius of outer cylinder (mm)
bulk ow Reynolds number, vz Dh/v
velocity in the z-direction (m/s)
distance between pressure taps (mm)
particle concentration(%)

Greek Symbols

ratio of radius, R1/R2

absolute viscosity (Pas)

kinematic viscosity (m2/s)

density of uid (kg/m3)

angular velocity of rotating cylinder (rad/s)

a highly deviated slim hole annulus. The ow characteristics of solid

liquid mixtures owing upward through a vertical annulus in a closedloop system is determined by measuring the volumetric concentration
of particles and the pressure drop in the test section. We expect that the
results of this study will be useful in explaining the effect of solids on
mixture ow transport phenomena in annular geometry.
2. Experimental method
2.1. Fluid property
Drilling uids usually have non-Newtonian properties. NonNewtonian uids are uids for which the strain rate and stress curve
are not linear; i.e., the viscosity of non-Newtonian uids is not
constant at a given temperature and pressure, and depends on other
factors such as the rate of shear in the uid, the apparatus in which
the uid is contained, or even the previous history of the uid. As
shown by Lauzon and Reid (1979), the power law model adequately
describes most drilling uids at shear rates normally encountered in
wellbore annuli during normal drilling operations. The carrier uids
(0.4% CMC and 5% bentonite solutions) used in the experiment are
shear-thinning.
In the case of CMC and bentonite solutions, n b 1, and the power
law relating the shear stress to the shear rate is given by
n

velocities in the vertical upward pipe ows of solidliquid mixtures.

They reported that additional turbulence was created due to the
relative velocity between the solid and uid phases.
This paper concerns a study of fully developed solidliquid mixture
ow of Newtonian uid and non-Newtonian uids through a concentric
annulus with combined bulk axial ow and inner cylinder rotation.
For an inclined annulus, we note that steady state conditions, which
are known to be prevalent in most drilling operations, commence after
the bed has already formed. Due to the constant interchange of particles
between the bed and the ow stream, the exact mathematical relations
describing free particle and steady bed dynamics will be formidable.
In an inclined annulus, bed formation of particles occurs only on the
low side of the annulus due to gravity effects, as shown in Fig. 1. Any
decrease in uid velocity in the annular space in this vicinity will worsen
the bed formation tendency. Although increasing inner pipe rotational
speed generally improves particle transport, it is more pronounced at
lower annular uid velocities, and appears to be negligible at high
velocities.
This study relates the physics of particle transportation and effective
drilled cutting transportation in highly deviated wells. In addition, this is
one of the rst experimental studies conducted on cuttings transport in

Fig. 1. Cuttings bed dynamics in an inclined annulus.

309

= k

where n is the ow behavior index and k is the consistency factor. The

apparent viscosity a for a power law uid may be expressed in terms
of n and k as
n1

a = k

The effective viscosity of 0.4% CMC solution for a ow rate of 6 l/

min becomes 16.5 cp at 0 rpm and 14.6 cp at 200 rpm, and n is

Fig. 2. Schematic diagram of experimental apparatus.

310

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

2.2. Experimental analysis

Fig. 3. Computational grid of concentric annulus.

measured as 0.75. The effective viscosity of 5% bentonite solution

becomes 40.6 cp at 0 rpm and 35.9 cp at 200 rpm, and n is measured
as 0.73. The density of 0.4% CMC solution is 998.5 kg/m3 and the
density of 5% bentonite solution is measured as 1041.1 kg/m3.

The objective of our solidliquid mixture ow experiments was to

provide high quality data on the effects of pipe rotation, ow regime,
uid properties, and particle concentration on pressure loss in a slim
hole annular ow. The set-up used in the experiments is a closed-loop
system consisting of a centrifugal slurry pump with a bypass line, a
vertical annulus, and a feed slurry tank, as shown in Fig. 2. The ow
conguration and instrumentation are described by Kim and Hwang
(2003). Our equipment consist of a cylinder part, a supporting part, a
uid-providing and rotating part, and a measuring part which
measures ow rate, pressure loss, and temperature.
The outside diameter of the inner pipe Di and the inside diameter
of the outer pipe Do are 30 mm and 44 mm, respectively, yielding a
radius ratio of 0.7 for the annulus. To ensure fully developed ow in
the measuring section, the length of straight annulus upstream of the
test section is 1.8 m, corresponding to 126 hydraulic diameters.
A cylindrical head tank was used for the preparation of feed
particles. The tank has a conical bottom in order to avoid settling of
the particles, and it is connected at the bottom to a centrifugal pump
with a exible hose.
Pressure drop and average ow rate were measured in the ow
range of 0.4 m/s b vm b 1.2 m/s. The ow rate was measured with a
magnetic ow meter (accuracy 0.5%). The inner cylinder can be
rotated at any speed up to a maximum of 600 rpm by means of an
alternating current (ac) motor.
The axial velocity of solid particles was measured using a high
speed charge-coupled device (CCD) camera installed outside of the

Fig. 4. Variation of particle rise velocity with rotational speed and ow rate at (a) 0, (b) 20, (c) 40 and (d) 60 in water.

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

cylinder. The outer cylinder was made of transparent acryl pipe. The
ow eld was captured by an XC-55 CCD camera with 640 480 pixels at 8 bit resolution. The shutter speed was 1/1000 s. The travel
length of the particle was measured in a BMP-type photograph
captured by the CCD camera. The velocity of the particle was obtained
by dividing by time (30 frames/s). The experiment was repeated ten
times for the same experimental conditions, and the particle rise
velocity was averaged.
Static pressure was measured using a static pressure tap with a
0.5 mm diameter, distributed longitudinally in the outer cylinder. Two
static pressure taps were installed along the ow direction in the
measuring section. The static pressures were read from a calibrated
manometer bank with 1 mm resolution. The specic gravity of the
manometer uid (CCl4) was 1.88, and it gave a height in the range of
20 600 mm.
A short horizontal sampling line with a ball valve was constructed
at the exit of the test section to measure volumetric particle
concentration. In each experimental run, a sample uid mixture of
approximately 500 cm3 was collected from this sampling port without
diverting the whole ow through the annulus. This procedure was
repeated at least ve times for the same experimental conditions, and
the measured data were averaged.
First, experiments with water were performed to determine the
accuracy of the experimental set-up. Later, experiments with the
solidliquid mixtures were performed in a similar manner.
The head tank was lled with water up to a marked level, and the
pump was started. Particles at the desired feed particle concentration

311

were prepared in the head tank by adding uniformly-sized sand

particles. The mixture ow rate was adjusted manually by the bypass
valve installed after the outlet of the pump.
When the mixture ow became stable, the axial pressure drop, the
velocity of solid particles, and the particle volume concentration were
measured simultaneously. Sand particles of 0.1 cm mean diameter
and 2.55 g/cm3 material density were used in the experiment.
3. Numerical method
Traditionally, the annulus between the drill pipe and the borehole
has been represented as a concentric annulus, as shown in Fig. 3. In
our numerical study, the EulerianEulerian approach for granular
ow was used, which allows the determination of the pressure and
viscosity of the solids phase by using the commercial software
program FLUENT instead of by using empirical correlations.
In a numerical study of solidliquid mixture, a control volumebased nite volume method was used to solve the equations of
motion. The problem reduced to solving the conservation of mass and
momentum equations with the appropriate boundary and initial
conditions. The governing equations are as follows:
Mass conservation equation:

q q + q q vq = m
pq
t
p=1

Fig. 5. Variation of particle rise velocity with rotational speed and ow rate at (a) 0, (b) 20, (c) 40 and (d) 60 in 0.4% CMC solution.

312

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

Momentum conservation equation for the uid:

+
4
v + q q vq vq = q p +
q
q qg
t q q q

+ q q F q + F lift;q + F vm;q

+ Kpq vp vq + m pq vpq

Momentum conservation equation for the solid:

+
v + s s vs vs = s pps +
s
s s g 5
t s s s

+ s s F s + F lift;s + F vm;s

+ Kls vl vs + m ls vls

The governing set of partial differential equations was discretized

using a nite volume technique. The discretized equations with the
initial and boundary conditions were solved using FLUENT. Particle
particle interactions via friction were also included.
The solids volume fraction in a domain of known volume was
specied at the beginning of each simulation to correspond to the
desired solids loading. The particles used in the simulation were solid
spheres with a density of 2.55 g/cm3. Values for the coefcient of
restitution and the friction coefcients for these particles were
assumed to be 0.9 and 0.09, respectively. The same coefcient of

friction was also assumed for interactions between the walls and the
particles.
No-slip boundary conditions were imposed on all the solid surfaces
for the continuous phase. The same conditions were also applied to the
discrete phase and imposed on the corresponding momentum
equations. The inlet liquid velocity and the outlet pressure were
specied. No-slip boundary conditions were assumed at the walls for
the liquid phase. Interactions between the particles and the walls were
modeled with the same formulation used for solids pressure and
granular viscosity for the particleparticle interactions.

4. Results and discussion

4.1. Experimental results
Hydrodynamic characteristics of single phase ows are well
known, but these characteristics are not valid for solidliquid
mixtures and multiphase ows. In this study, the ow characteristics
of solidliquid mixture ows were investigated experimentally at
different operating conditions.
The experimental parameters are the feed particle concentration,
mixture velocity in the annulus, rotational speed of the inner cylinder,
and inclination angle of the annulus. The important characteristics of
solidliquid ows in an annulus are the particle transport performance and the pressure drop versus mixture velocity relationship.
Experiments were performed in turbulent ow using water, and in
laminar ow using 0.4% CMC and 5% bentonite solutions. The ow

Fig. 6. Variation of particle volume concentration with annulus inclination and ow rate at (a) 0 rpm, (b) 200 rpm, (c) 400 rpm and (d) 600 rpm in water.

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

313

Fig. 7. Variation of particle volume concentration with annulus inclination and ow rate at (a) 0 rpm, (b) 200 rpm, (c) 400 rpm and (d) 600 rpm in 0.4% CMC solution.

regime might be an important factor in cuttings transport due to

changes in the velocity prole from laminar to turbulent ow.
Drilling uids are able to transport cuttings to the surface principally
by means of the uid axial velocity. Due to the gravitational forces, the
cuttings tend to slip downward or settle through the uid medium as
they are transported through the annulus.

4.1.1. Particle velocity

Using a high speed CCD camera, we measured the average particle
rise velocity vp with mixture ow, rotation of the inner cylinder, and
inclination of annulus, as shown in Figs. 4 and 5. The particle velocity
measurement experiment was carried out only in water and CMC
solution.
As the annulus inclination increased to over 20, an increasing
number of particles were forced toward the low side of the annulus,
eventually resulting in the formation of a cuttings bed, especially at
low ow rates due to increased radial slip velocity.
For both water and 0.4% CMC solution, the axial velocity of
particles was gradually increased with the mixture velocity and
rotation of the inner cylinder. The solid velocity showed an irregular
pattern in the vicinity of 40 degree annulus inclination and similar
data in the high velocity. The minimum mixture uid velocity for
particle transport in 0.4% CMC solution is smaller than that of water.
This is because the viscosity of 0.4% CMC solution is about 16 times
higher than that of water. Higher viscosity is effective for particle
transport in a certain range.

Unlike vertical particle transport, the use of average particle rise

velocity to evaluate particle transport performance in a directional
well would be misleading due to bed formation and the associated
reduction in effective ow area, leading to high effective uid and
particle velocities, as shown in Figs. 4 and 5. As the inclination of the
annulus is increased to over 40 from the vertical, the velocity of
particles closes with the mixture uid velocity.
As shown in Figs. 4 and 5, the particle rise velocity is faster at
400 rpm than at 0 rpm for the same ow conditions. Also, the
particle volume concentration (CVT) is slightly decreased at 400 rpm
compared to 0 rpm. From this, rotation of the inner cylinder generally improves the transport of solid particles, is more pronounced at
lower mixture velocities, and appears to be negligible at high
rotational speeds over 300 rpm. One explanation for this may be that
at high ow rate, the mixture velocity is the dominating factor
in particle transportation, with rotational speed having marginal
effects.
4.1.2. Particle concentration
Successful and economic drilling of highly deviated slim hole wells
requires the efcient transportation of cuttings to the surface. Figs. 6
8 show graphical correlations of CVT in terms of several relevant
drilling parameters. Using these correlations, the total particle
concentration can give a good indication of the drilling uids' transport efciency.
The lower the particle concentration is, the better the transport
performance. From this point of view, we note from Figs. 68 that

314

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

Fig. 8. Variation of particle volume concentration with annulus inclination and ow rate at (a) 0 rpm, (b) 200 rpm, (c) 400 rpm and (d) 600 rpm in 5% bentonite solution.

drilling uids in all ow regimes perform best at low angles of annulus

inclination. Particle concentration CVT is dened as
CVT =

Net volume occupied by particles

100:
Total annulus volume

Fig. 6 shows the variation of CVT as a function of ow rate and

inclination of the annulus. CVT decreased as the ow rate increased.
However, the CVT increased with the inclination of the annulus. That
is, particle transport performance declined. The uid velocity of 0.4%
CMC solution is much smaller than that of water for solids transport
for the conditions shown in Fig. 7. (i.e., the transport efciency of 0.4%
CMC solution is much better than that of water). Moreover, the higher
viscosity uid gives better transport performance than the lower
viscosity uid. The effective viscosities of 0.4% CMC and 5% bentonite
solutions are much higher than that of water, and show a shearthinning behavior. Thus, viscoelastic uids are more able to mobilize
and transport particles. As shown in Figs. 7 and 8, 0.4% CMC and 5%
bentonite solutions can transport sand particles at lower mixture
velocities compared to water.
Fig. 8 shows the transport efciency of 5% bentonite solution
compared to 0.4% CMC solution. The transport efciency of 5%
bentonite solution is clearly better than 0.4% CMC solution. The main
reason is that the specic gravity of 5% bentonite solution is several
times higher than that of 0.4% CMC solution. From this, we conclude
that uid viscosity and density are important elements in the design
of drilling uid.

Viscosity is effective for particle transportation in a certain range.

However, particle transport performance does not improve continuously with viscosity, as shown in Figs. 6 and 7.
The effect of inner pipe rotation on particle concentration is shown
in Fig. 6. Rotation of the inner cylinder helped prevent a bed of
cuttings from being formed. The effects of pipe rotation, based on this
graph, are minor at best and is generally negligible. Repeated
experiments show 5% error in the solid particle velocity measurements, and 9% error in the particle concentration measurements.
4.1.3. Pressure loss
Accurate pressure calculations are crucial for safely controlling
formation pressures and protecting wellbore integrity. These calculations are more difcult for slim hole drilling than for conventional
drilling. Variations in annular geometry, annulus inclination, and pipe
rotational speed strongly affect pressure loss of a uid owing in the
narrow annulus of a slim hole well. In slim hole annular ow, because
of the reduced annular clearance, the effect of pressure loss is different
than that experienced in conventional drilling. In conventional
drilling, 90% of the pressure losses occur in the drill pipe and through
the bit nozzle. In slim holes, up to 60% of the pressure losses occur in
the annulus (Saggot and Dupuis, 1994, pp. 8190).
Pressure drop versus mixture velocity and annular inclination are
shown in Figs. 9 and 10 for a mean particle size of 0.2 cm and a sand
volume fraction of 4%. The carrier uids are water, 0.4% CMC, and
bentonite solutions. Unlike single phase ow, the pressure drop of solid
liquid mixture ow changes with the change of annulus inclination. The

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

315

liquid mixture ow increased due to the friction between the wall and
solids, or among solids.
It is difcult to compare directly the pressure drop among the
three uids because the mixture uid velocity of water is greater than
the other uids. However, the pressure drop of the solidliquid
mixture ow increased with the increase in viscosity of the carrier
uid. The inclination of the annulus affected the magnitude of the

Fig. 9. Variation of pressure drop with annulus inclination in added sand (4%):
(a) water, (b) 0.4% CMC solution, (c) 5% bentonite solution at 0 rpm.

additional effect of the gravitational force results in a monotonic change

in pressure drop as the angle of inclination increases. This ow pattern is
more prevalent as the annulus inclination increases because of the bed
formation and increased friction between particles and wall or among
particles.
As shown in Fig. 9, the pressure drop increased along the increase
in mixture uid velocity and annulus inclination, but the characteristic shape of the pressure drop curve did not change. Also, the
pressure drop of the solidliquid mixture ow increased abruptly
compared to the single phase uid ow. The pressure drop of solid

Fig. 10. Variation of pressure drop with rotational speed and solid concentration in
(a) water, (b) 0.4% CMC solution, and (c) 5% bentonite solution. (Inclination = 0).

316

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

distribution and pressure eld of the ow eld inside the annulus

were found for various operating conditions.
Eccentricity of the annulus usually occurs while digging an oil
well. Therefore, this investigation considered a ow with changes in
the eccentric ratio for a xed radius ratio. When the number of
inner cylinder rotations inside the eccentric annulus increases, the
grazing ow velocity is disturbed at the point where the gap is
narrowed. Due to these difculties in calculation, there are few
investigations of matters such as ow rate, eccentricity, and inner
cylinder rotation.
To obtain a more realistic representation of the drilling operation,
the solidliquid mixture ow in concentric and eccentric annuli for a
Newtonian uid and a power law uid was considered. Numerical
calculations were extended for the case in which the inner cylinder
rotates about its own axis at a constant rotational speed. Detailed
calculations were performed for a radius ratio of 0.7 covering
eccentricities up to 0.2, and a rotational speed of the inner cylinder
of up to 100 rpm. The computational domain consisted of 15, 96, and
500 nodes in the radial, azimuthal, and axial directions, respectively.
Calculation results were compared with experimental data in terms of
pressure drop and were compared to experimental data to validate
the results of the numerical model. The calculation results showed
good agreement with experimental data for pressure drop and
particle transport velocity in solidliquid two-phase ow, as shown
in Fig. 12.

Fig. 11. Comparison of pressure loss with sand concentration at 0 rpm: (a) water,
(b) 0.4% CMC solution.

pressure drop. The pressure drop increased as the viscosity of carrier

uid increased.
Fig. 10 shows the pressure drop of solidliquid mixture ow with
rotation of the inner cylinder in a vertical annulus. As shown in Fig. 9,
the pressure drop increased with increasing axial ow rate, and it
increased with the rotational speed of the inner cylinder. The pressure
drop in mixture ow increased much more than pure liquid ow as
shown in Fig. 10. The friction between the wall and solids, or among
solids, increased with the rotation of the inner cylinder. The effect of
the inner cylinder rotation on pressure drop was the greatest for
water, and became weak in 0.4% CMC and 5% bentonite solutions. This
is because the resistance against the ow of shear-thinning uid
decreases with increasing shear rate.
The pressure drop versus mixture velocity and annular inclination
are shown in Fig. 11 for a mean particle size of 0.2 cm and sand
volume fractions of 4% and 8%. The carrier uids were water and 0.4%
CMC. The pressure drop increased along the increase of particle
injection and rotation speed of inner cylinder. Also, the total pressure
drop increased along with the increase in volume concentration of
solid particles and mixture velocity. For higher volume fractions, the
hydraulic pressure drop of the two-phase mixture increased due to
the friction between the wall and solids, or among solids.
4.2. Numerical results
To validate our computer program with FLUENT, our numerical
results were compared with the experimental results. The velocity

Fig. 12. Comparison of pressure loss in water and 0.4% CMC solution as a function of
ow rate.

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

317

Fig. 13. Contour of solid volume fraction in inclined annulus (0.4% CMC solution velocity = 0.5 m/s, sand 8%).

Volume fraction contours of solid particles in the z-section for an

inclined concentric annulus are shown in Fig. 13. The carrier uid was
0.4% CMC solution and the uid velocity was 0.5 m/s. The ow regime
was laminar. The annulus was inclined 30 60 from the vertical. In
the case of non-rotation, as shown in Fig. 13(a), the prole of particle
volume fraction is axisymmetric, and the particles were located on the
bottom of the annulus because of the gravitational effect. The particle
volume fraction decreased as the rotational speed of inner cylinder
increased up to 400 rpm, as shown in Fig. 13(b) and (c). However, the
effect of rotation appeared to weaken above 200 rpm. Fig. 14 shows
the simulated solid content proles from the FLUENT results for
several ow conditions. As seen in Figs. 13 and 14, rotation of the
inner cylinder disturbed the formation of the bed. Generally, this
effect is useful for cuttings transport in an inclined annulus.
Fig. 15 shows the effect of eccentricity. In the case of the concentric
annulus shown in Fig. 15(a), the prole of the particle volume fraction
is axisymmetric. The particle velocity in the narrow gap decreased as
the eccentricity increased (Fig. 15(c)). Thus, the particle volume
fraction increased in the narrow gap.
Calculations were run for 0.4% CMC solution as the carrier uid,
owing at a rate of 8.5 l/min through an annulus with a radius ratio of

0.7, as shown in Fig. 16. The calculated axial pressure drop ratio is
plotted versus eccentricity for inner cylinder rotation speed. The ratio
of the axial pressure drop to the axial pressure drop was obtained
when the inner cylinder was stationary. A pattern in the results is
clearly visible: for small excursions from a concentric annulus, shearthinning effects have ample opportunity to dominate inertial effects.
Thus, the axial pressure drop falls with increased eccentricity.
5. Conclusions
In this study, steady laminar and turbulent upward ows of solid
liquid mixtures through concentric and eccentric annuli were experimentally and numerically investigated.
Numerical calculations were run for 0.4% CMC solution as carrier uid
and the calculation results showed a good agreement with experimental
data for the pressure drop, but numerical results are slightly lower than
experimental values.
Our numerical calculations predicted that almost all sand particles
settle down to the lower part of the annulus, and some particles are
suspended in the upper part of the annulus. As the hole inclination
increases from the vertical, the volume fraction of sand is increased.

318

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

Fig. 15. Contour of volume fraction of solid particle for Re = 1050 in vertical annulus.
(0.4% CMC solution, sand 8%): (a) = 0, (b) = 0.1, (c) = 0.2.

Fig. 14. Volume fraction of sand in 0.4% CMC solution with inclination: (a) 30, (b) 45,
(c) 60 at vz = 0.43 m/s and Ri = 22 mm.

For small excursions from a concentric annulus, the axial pressure

drop falls with the increase in eccentricity due to the shear-thinning
effects having ample opportunity to dominate inertial effects.
The use of an average particle rise velocity to evaluate particle
transport performance in an inclined annulus would be misleading
due to bed formation and concomitant reduction in effective ow
area, leading to high effective uid and particle velocities.
The particle transport performance of 5% bentonite solution is
superior to that of 0.4% CMC solution. There is no great difference in

viscosity of two the uids, but the specic gravity of 5% bentonite

solution is larger than that of 0.4% CMC solution. Viscosity and density
of uids are the important factors in the design of drilling uids.
The pressure drop in a solidliquid mixture ow increases with the
mixture ow rate, inclination of annulus, and rotation of the inner
cylinder. For a solidliquid mixture ow, the hydraulic pressure drop
increases largely compared to single phase ow due to the friction
between the wall and solids or among solids.
Rotation of the inner cylinder generally improves the transport
performance of particles. It is more pronounced at lower mixture
velocities and appears to be weakened at high ow rate. The effect of
the inner cylinder rotation on the pressure drop is largest in water and
it becomes weak in 0.4% CMC and 5% bentonite solutions. This is
because the resistance against the ow of shear-thinning uid
decreases with increasing shear rate.

Acknowledgement
This work was supported by the Brain Korea 21 Project in 2008, Faculty
Research Fund, Sungkyunkwan University, 2008 and Development of

S.-M. Han et al. / Journal of Petroleum Science and Engineering 70 (2010) 308319

319

References
Ford, J.T., Peden, J.M., Oyeneyin, E.G., Zarrough, R., 1991. Experimental investigation of
drilled cuttings transport in inclined boreholes. SPEAnnual Technical Conference,
New Orleans, Sept.2326, SPE 20421.
Kim, Y.J., Hwang, Y.K., 2003. Experimental study on the vortex ow in a concentric
annulus with rotating inner cylinder. KSME Int. J. 17 (4), 562570.
Lauzon, R.V., Reid, K.I.B., 1979. New rheological model offers eld alternatives. The Oil
and Gas J. 5, 5157.
Ozbelge, T.A., Beyaz, A., 2001. Dilute solidliquid upward ows through a vertical
annulus in a closed loop system. Int. J. Multiphase ow 27, 737752.
Pigott, R.J.S., 1941. Mud ow in drilling: drilling and production practice. API 91103.
Saggot, A.M., Dupuis, D.C., 1994. A major step in ultra slimhole drilling. SPE Annual
Technical Conference, New Orleans, Sept. 2528, SPE 28299.
Sellgren, A., 1982. The choice of operating velocity in vertical solidwater pipeline
systems. BHRA Fluid Eng. 211226 (Paper D3).
Sifferman, T.R., Myers, G.M., Haden, E.L., Wahl, H.A., 1974. Drill-cuttings transport in
full-scale vertical annulus. J. Pet. Tech. 12951302.
Tomren, P.H., Iyoho, A.W., Azar, J.J., 1986. An experimental study of cuttings transport in
directional wells. SPE Annual Technical Conference, San Francisco, Oct. 58,
SPE 12123.

Fig. 16. Flow of solidliquid uid in eccentric annuli with inner cylinder rotation:
(a) water, (b) 0.4% CMC solution.

Coiled Tubing Drilling System Project (Energy Technology Innovation) of

the Korea Institute of Geoscience and Mineral Resources (KIGAM) funded
by Korea Energy Management Corporation. This support is gratefully
acknowledged.