Professional Documents
Culture Documents
E. GhasemiKafrudi
Computational Fluid Dynamics
Research Laboratory,
of Drilling Mud Rheological
School of Chemical Engineering,
Iran University of Science
and Technology,
Properties on the Transport
Tehran 16846, Iran
of Drilling Cuttings
S. H. Hashemabadi1
Associate Professor Inaccurate prediction of the required pressures can lead to a number of costly drilling
Computational Fluid Dynamics problems. In this study, the hydrodynamics of mud-cuttings were numerically studied
Research Laboratory, using the Mixture Model. To this end, an in-house code was developed to calculate the
School of Chemical Engineering, velocity and pressure fields. The mud velocity profile using of Herschel–Bulkley model
Iran University of Science and solid phase volume fraction were locally calculated; moreover, pressure drop
and Technology, through the annulus was taken into account. The effects of velocity, mud properties, and
Tehran 16846, Iran solid phase volume fraction on pressure drop were discussed and a new correlation was
e-mail: hashemabadi@iust.ac.ir proposed for calculating friction factor based on corresponding parameters.
[DOI: 10.1115/1.4031450]
1 Introduction reliable analytical model which predicts the pressure drop for lam-
inar fluids in concentric annuli. Sorgun et al. [7] presented a math-
Drilling cost optimization has always been an essential subject
ematical model for predicting flow characteristics and frictional
in the petroleum engineering. Drilling cost depends on various
pressure losses of Newtonian fluids in concentric horizontal annu-
factors particularly mud cost, safety, and operation efficiency.
lus with drill pipe rotation. Wei et al. [8] used theoretical, experi-
Pumping and increasing mud pressure are crucial operating cost
mental, and field data for considering the effects of parameters on
sources. The mud pressure helps clean the well and in some cases
annular pressure loss due to friction for Power law fluids in lami-
provides power for the bit. Thus, it is necessary to study the pa-
nar regime. Wang et al. [9] performed an experimental investiga-
rameters influencing pressure drop and hydrodynamics of mud in
tion for finding the effective parameters of drilling in a slim hole
drilling well. The bottom hole pressure, as a critical parameter in
and concluded that the annular mud flow regime relies on not only
any drilling operation, is a function of the hydrostatic head of the
the Reynolds number but also the Taylor number. Ozbelge and
mud and annular pressure loss. Hence, accurate prediction of the
Beyaz [10] discussed the pressure drop in solid–liquid flows and
annular pressure loss leads to a good estimation of the bottom
the alternative of operating velocities in the vertical pipe for
hole pressure that prevents kicks and circulation losses. In addi-
solid–liquid mixtures. Escudier et al. [11] studied the rheological
tion, drilling mud carries cuttings from the bottom of the well to
behavior of xanthan and carboxymethyl cellulose (CMC) solu-
the surface which has crucial effects on the drilling mud flow
tions mixture and experimentally investigated pressure drop and
field. Therefore, understanding the hydrodynamic behavior of
velocity profiles in the annuli. Zhou and Shah [12] experimentally
drilling fluids through the extraction process from drilling wells
investigated the rheological properties and friction pressure losses
can be used to optimize the extraction performance and reduce
of several common well-drilling, completion, and stimulation flu-
pumping cost of drilling mud [1].
ids. Woo et al. [13] conducted an experimental study of fully
developed laminar flows of Newtonian and non-Newtonian fluids
1.1 Background. Many investigations have been done exper- through a concentric annulus with an inner rotational cylinder.
imentally and numerically on drilling fluids flow through the well- Their results showed that the pressure drop increases at higher
bore. Sorgun [2] developed simple empirical frictional pressure rotational speeds of the inner cylinder. Kelessidis et al. [14] meas-
losses and cuttings bed thickness correlations including pipe rota- ured the viscosity of bentonite suspension in water as the drilling
tion for solid–liquid flow in horizontal and deviated wellbores. fluid and obtained single-phase pressure drop of this drilling fluid
Ozbayoglu and Sorgun [3] developed correction factors for calcu- with two-dimensional analytical solution regardless of the drill bit
lating frictional pressure losses in eccentric horizontal annuli for rotation. Pereira et al. [15] simulated the single-phase flow in
non-Newtonian fluids with the effect of pipe rotation and showed eccentric and concentric annuli and validated the velocity profiles
that pipe rotation affects frictional pressure loss particularly at with Escudier’s data [16] and, then, studied the effects of viscos-
lower flow rates. Sorgun et al. [4] proposed a mechanistic model ity, drilling string rotation and axial velocity on pressure drop.
for predicting the frictional pressure losses of Newtonian fluid in Ahmed and Miska [17] carried out an experimental and theoreti-
concentric annuli. Additionally they studied pipe eccentricity cal study on laminar flows of yield power-law fluids in concentric
effect on tangential velocity inside the annulus. Erge et al. [5] and fully eccentric annuli. They found that the frictional pressure
made a horizontal well setup with drilling string, to investigate loss is mainly affected by fluid rheological properties, flow re-
influence of rotation on frictional pressure losses of yielded power gime, diameter ratio and annulus eccentricity. Zhi-ming et al. [18]
law fluids. They observed most effect of rotation is in the transi- simulated the transport of drilling cuttings in a horizontal well.
tion region from laminar to turbulent flow. Founargiotakis et al. They assumed a Newtonian behavior for the drilling mud and a
[6] investigated laminar, transitional and turbulent flow of spherical shape for particles with a mean diameter of one milli-
Herschel–Bulkley fluids in a concentric annulus. They offered a meter. Han et al. [19] designed an experimental annulus system
with a rotational inner shaft. The fluid was a solution of CMC (4%
1
Corresponding author. volume fraction) in which the fluid was assumed to be a shear
Contributed by the Petroleum Division of ASME for publication in the JOURNAL
OF ENERGY RESOURCES TECHNOLOGY. Manuscript received January 16, 2015; final
thinning fluid. In their study, CMC solution with 4% volume
manuscript received May 23, 2015; published online September 16, 2015. Assoc. fraction of solids passed through the annuli. They concluded that
Editor: Christopher J. Wajnikonis. pressure drop increases with raising mud flow rate, particles
rheological properties and cuttings transport on pressure loss and þ qm g þ F þ r ak qk vdr;k vdr;k
k¼1
profile velocity through a vertical annulus are numerically investi-
gated. Herschel–Bulkley model is to be used to predict the rheolog- (3)
ical properties of the drilling mud. The mud and cuttings velocity
profile and the volume fraction of solids in the annulus cross section The last two terms stand for the effect of second phase (cutting
are to be calculated. Moreover, the effects of mud flow rate, drilling phase) comes in the momentum equation
string rotational speed, the volume percent of solid phase and fluid
type on pressure drop will be investigated. Finally, a correction fac- 2.2 Relative (Slip) and Drift Velocities. The slip and the
tor will be proposed to accurately determine the frictional pressure drift velocity in the momentum equations show the interdepend-
loss in concentric vertical annuli for non-Newtonian fluids includ- ence of drilling fluid hydrodynamics and dispersed phase (cutting
ing the effect of inner cylinder rotation. phase). These two velocities (drift and slip) are related through
the following expression [28]:
1.3 Statement of the Problem. In the drilling process, mud
is pumped down through the drill string so that the mud with high X
n
ak q k
vdr;p ¼ vpc vck (4)
pressure breaks out the bit and carries cuttings to the surface of k¼1
qm
ground through the annulus. In this study, the bit size was selected
according to the stiffness of the well’s walls. Typically, the diam- Here, cuttings and mud are the secondary and primary phases,
eter can be in the range of 0.311–0.445 m and almost 0.127 m for respectively. The form of the slip velocity is given by
wellbore and drilling string, respectively. The cuttings are classi-
fied as low density solid with the density range of 2400–2900 kg/
qp dp2 sp ðqp qm Þ
m3. It is too hard to find out the optimum shapes, sizes and size vpc ¼ a (5)
distribution in each circumstance. However, sphericity or shape 18Fd lc qp
factor (u), which is defined as the ratio of the surface area of a
sphere with the same volume of the particle (a) to the actual sur- 2.3 Mud Force on Cuttings. When a particle is carried by a
face area of the particle (S), has been reported to be in the range moving fluid, different forces can affect its trajectory (Fig. 1(b)).
of 0.75–0.85 in the literature [2,24]. The velocity of mud which is The crucial forces are weight, buoyancy and drag forces and other
pumped into the bore string depends on annulus diameter and ones are negligible. The drag force of drilling fluid on the cuttings
varies from 0.4 to 1.4 m/s [19]. The mud pressure must be high (term F in Eq. (3)), is given [27]:
enough to prevent reservoir fluids (oil and gas) from entering the
well but does not break the formation [1]. !
18lm Cd Re
F¼ (6)
qp dp2 24
2 Computational Methodology
A finite-volume based in-house code was developed using the The drag coefficient, Cd, is
mixture model and was then employed for performing the simula-
tions in this study. The mixture model [25,26] was implemented
for simulating solid–liquid flow field and the interaction of the
24
b Re
3 p
two phases was applied by drag coefficients according to Cd ¼ 1 þ b1 Rebp2 þ (7)
Morsi–Alexander’s law [27]. Rep b4 þ Rep
where shear stress prediction versus shear rate by the model can be
accepted. The mud apparent viscosity is calculated by
b1 ¼ exp ð2:3288 6:4581u þ 2:4486u2 Þ
b2 ¼ 0:0964 þ 0:5565u
lmud ¼ jsj=jcj
_ (9)
b3 ¼ exp ð4:905 13:8944u þ 18:4222u2 10:2599u3 Þ
where shear rate tensor and the magnitude of the shear rate are
b4 ¼ exp ð1:4681 þ 12:2584u 20:7322u2 þ 15:8855u3 Þ given by [30,31]
rffiffiffiffiffiffiffiffiffiffiffiffi
† 1
_c ¼ rvc þ rvc ; j_c j ¼ c_ : c_ (10)
2
2.4 Constitutive Equation. The Herschel–Bulkley [29]
model extends the simple power-law model to include a yield 3 Numerical Solution
stress term as follows:
3.1 Computational Domain and Boundary Conditions.
sy The three-dimensional computational domain used in this study is
s ¼ Kjcj_ n1 þ c_ for jsj sy (8)
jcj
_ presented in Figs. 1(a) and 1(c). All of the boundary conditions
are also shown in Fig. 1(a). In all simulation cases, the inlet veloc-
where n is power law index and K is consistency coefficient. The ity to the annulus was assumed to be uniform. The drill string
experimental data [14] can be predicted by Herschel–Bulkley rotated at constant angular velocity and the wall casing was sta-
model relation and the rheological characteristics of each fluid are tionary. The wall boundary was subjected to no slip conditions
tabulated in Table 1. It can be seen in Table 1 that R2 values are (zero velocity on the wellbore). To reduce computational time and
close to 1 and the biggest difference with 1 is 0.012. Therefore, false diffusion error, the totally orthogonal structural hexahedral
Samples sy K n R2
Fig. 7 Pressure drop versus average velocity for five different Fig. 9 Pressure gradient versus input velocity and increased
fluid types (Table 1) (qp 5 2550 kg/m3, ap 5%3) drilling string rotation velocity. With D.F-1 fluid ap 58%.