Please fill in the name of the event you

Annual Technical Conference and Exhibition – ATCE 2020

are preparing this manuscript for.

Please fill in your 6-digit SPE

SPE-201524-MS
manuscript number.

Please fill in your manuscript title. New Scale Prediction Approach Through Numerical Simulations

Please fill in your author name(s) and company affiliation.

Given Name Surname Company
Marcus Vinicius Duarte Ferreira Petrobras
André Leibsohn Martins Petrobras
Rodrigo Simões Maciel UFES
Rômulo Fieni Fejoli UFES
Fabio de Assis Ressel Pereira UFES
Vinicius Gustavo Poletto UTFPR
Fernando Cesar De Lai UTFPR
Silvio Luiz De Mello Junqueira UTFPR
This template is provided to give authors a basic shell for preparing your manuscript for submittal to an SPE meeting or event.
Styles have been included (Head1, Head2, Para, FigCaption, etc) to give you an idea of how your finalized paper will look
before it is published by SPE. All manuscripts submitted to SPE will be extracted from this template and tagged into an XML
format; SPE’s standardized styles and fonts will be used when laying out the final manuscript. Links will be added to your
manuscript for references, tables, and equations. Figures and tables should be placed directly after the first paragraph they are
mentioned in. The technical content of your paper WILL NOT be changed. Please start your manuscript below.

Abstract
One of the most concerns regarding the development of the Brazilian pre-salt cluster is due to scale
issues. The huge carbonate reservoirs have a high potential for salt depositions while the produced fluid
flow along the well. Furthermore, emerging wellbore configurations aiming the well construction cost
reduction and improving reliability, also brings some drawbacks, such as the inability for downhole
chemical injection in the open-hole full-eletric intelligent completion schemes, for instance.
Scale prediction is worldwide traditionally performed using the formation water thermodynamical
evaluation under static conditions through commercial softwares, leading to conservative results that may
not distinguish the scaling risk of two different wellbore configurations. The approach neglects th
geometry of completion accessories and the fluid dynamics, both substantial factors that influence
precipitated-crystals process such as agglomeration and adhesion to the surface. This paper aims to show
two different methodologies under development, in two different universities in Brazil, for enhancing
scale prediction studies. Both uses Computational Fluid Dynamics (CFD) techniques to describe the fluid
flow through a completion accessory applying an Euler-Lagrange approach is applied. The first approach
coupled the CFD study to the Discrete Phase Method (DPM). The second approach coupled the CFD to
the Discrete Element Method (DEM). Results are explored by applying these two methodologies.

Introduction
The brazilian pre-salt reservoirs development has shown scale issues even on the beginning of wells
production with very low watercut measures. It has led Petrobras to develop new strategies for preventing
and remediating salt depositions and for scale predicting enhancement. This paper proposes different
manners to evaluate salt deposition on completion equipments under downhole conditions using CFD
techniques. This approach is increasing in both industry and academy showing very promising results.
Recently, Maciel et al. (2019) analyzed the influence of fluid flow and geometric variations on the salt
crystals deposition. They used the CFD-DPM approach to model the transport of calcium carbonate
crystals until the adhesion on the Inflow Control Valve (ICV) walls. Bouamra et al. (2019) developed a
methodology to evaluate the risk of calcium carbonate deposition along a well production string focus on
understanding the scale in an ICV. Guan et al. (2018) highlighted the growing demand for intelligent
completion wells given the increasingly reservoir complexity of subsea wells. The study assessed the risk
2

of scale occurrence during the well lifetime and showed evidences that ICVs have the higher likelihood
of calcium carbonate deposition. Particularly, the authors also pointed out the importance of using more
reliable simulation tools to support the selection of laboratory test conditions. CFD approach has also been
used to address other aspects of scale mitigation, such as in the optimization of squeeze placement
operations (Droppert et al, 2020).
The CFD also features the potential of being coupled to the Discrete Element Method (Zhu et al., 2007),
empowering the method with the capability of computing contact particle interactions such as collision,
friction and adhesion. The CFD-DEM may also experience four-way coupling, with the possibility of the
solid phase imposing a pressure drop over the flow (Oweis et al., 2006). A number of works have
employed the DDPM-DEM approach to simulate the liquid-solid two-phase flow to seal discrete fractures
(Barbosa et al., 2016; Barbosa et al., 2015, 2019; De Lai et al., 2014), analyzing the influence of the
particles properties (diameter and specific mass ratio), the flow parameters (Reynolds number, fluid
invasion flow rate) and the geometry of the fracture. The particles accumulation in a channel filled with
an heterogeneous porous medium has also been simulated (Lima et al., 2017), resembling a static mud
cake build up process. The dynamic filtration filter cake build has also been simulated (Poletto et al.,
2020) regarding the influence of the flow parameters. Recently, the liquid-solid flow with adhesion to
simulate scale deposition in Sliding Sleeve Valves (SSV) has been presented (Poletto et al., 2020)
regarding the influence of the particle size distribution.

Methodology
As described in the previous paper from Maciel et al. (2019) the methodology to investigate scale in
completion equipment follows the workflow in the Figure 1. Briefly describing, the the approach starts
with the study of the thermodynamics of the formation water employing a commercial software. Then, it
is proposed the use of CFD coupled with a Lagrangian approach to describe calcium carbonate crystals
transport. With the velocity, pressure and temperature fields along with the particle tracking throughout a
completion tool, for instance a sSSV, it is coupled an adhesion model which is responsible to point out if
the particle will deposit or not at the completion tool walls.

Figure 1. Scale prediction workflow.

CFD-DPM Approach
The Euler-Lagrange approach allows modelling the fluid flow and the particle transport in a diluted
system, i.e. volumetric fraction lower than 10 % (ANSYS, 2017). The liquid phase is treated as a
continuous phase, so the velocity and pressure fields are obtained from the solution of the mass and
momentum conservation equations. In the Lagrangian phase, that is, the dispersed one, the velocity and
the particle position are obtained individually for each particle from the differential equation arising from
 3

the Newton second law so describing the movement of a theoretic particle (Equation (1))

𝑑𝑢 ⃗ 𝑔⃗ 𝜌 − 𝜌
= 𝐹 𝑢⃗ − 𝑢 ⃗ − + 𝐹⃗ (1)
𝑑𝑡 𝜌

More details can be obtained in ANSYS, 2017.

The interaction between the particle and the boundaries of the control volume leads to a boundary
condition for Equation (1). The condition Trap (ANSYS, 2017) corresponds to the total adhesion of
particles in the boundary regions, defined, for instance, by a completion equipment surface walls.
However, to better represent this interaction this condition was modified via UDF (User Defined
Function). In the adhesion phenomenon it can be accounted for the location and number of particles on
the surface, which are used in the post-process step of the simulation resulting in the adhesion rate. At this
point it is important to define the adhesion critical velocity, that is, the threshold velocity for the particle
adherence on the surface. Above this velocity the phenomenon does not take place. Abd-Elhady et al.
(2009) models the adhesion critical velocity of calcium carbonate crystals based on the maximum surface
impact velocity. Above this velocity particles are elastically or semi-elastically reflected. The adhesion
critical velocity is defined as follows:

⁄
𝛤
𝑣 = 1,184 𝑅 (2)
𝜌 𝐸*

where 𝛤 represents the bodies interaction surface energy and is worth 2.29 J/m² for the carbon-steel and
the calcium carbonate. R is the particle radius (23 𝜇𝑚, obtained experimentaly). 𝐸 * corresponds to the
Reduced Young Modulus obtained from Equation (3) and 𝜌 is the calcite density (2,970 kg/m³).

1−𝜐 1−𝜐
𝐸* = − (3)
𝐸 𝐸

where 𝐸 and 𝐸 correspond to the calcium carbonate (0.35 x 1011 J/m²) and the carbon-steel (2.15 x 1011
J/m²). elasticity modulus, respectively. 𝑣 and 𝑣 represent the Poisson ratio of the calcium carbonate
(0.27) and the carbon-steel (0.28), respectively.
Figure 2 represents the control volume, with somewhat simplifications, of a SSV mockup (reduced
scale) tested in a laboratory flow-loop to quantify calcium carbonate deposition. For this test, the set up
was arranged in the horizontal. Although it is not a usual arrangement, it works for validation purposes.
Two gauges were placed in the SSV setup ends to measure the differential pressure in the valve.
The adhesion critical velocity acts as a filter, in which the adhesion is only accounted if there is a
collision between the particle and the SSV control volume surface with impact velocities lower than the
threshold value. The following steps were performed in order to evaluate the effect of the computational
grid discretization on the scale rate accuracy:
i. Initially two different tetrahedral computational grids were generated and then coverted in
polyhedral grid. Such grids were obtained considering a discretization of a suitable boundary layer
with five prismatic elements. The calculation of the first layer takes into account the k- turbulent
model, the average fluid velocity, the parameter y+ of 30 and the hydraulic diameter of the cross
section. The refinement parameter is the maximum thickness of the cell, being 10-4 m for the fine
grid and 8 x 10-4 m for the coarse grid (see Figure 3).
4

ii. Two simulations are performed, both with the target of evaluating the effect of the mesh in the
quantification of the adhesion through the UDF. Thus, it is considered for those tests the critical
velocity of 0.036 m/s.
After the selection of the suitable computational mesh in terms of accuracy, precision and
computational cost, the computational study is then continued. To fit the appropriate simulated scale rate
to that one obtained from the experimental test it was proposed a sensitivity analysis varying the adhesion
critical velocity. The critical velocity of 0.018 m/s is obtained from Equation (2) considering the calcium
carbonate parameters according to (Abd-Elhady et al., 2009).

Figure 2. Control volume of the simulated SSV mockup, close of the TRIM region.
 5

Figure 3. One of the SSV computational mesh used in the mesh convergence study.

CFD-DEM Approach
The scale formation process is envisioned as a two-phase liquid-solid flow constituted of a liquid
medium with discrete solid particles dispersed within (Crowe et al., 1998). Particles mimic the precipitated
crystals of inorganic scale that, supposedly, have already precipitated far away from the valve. The crystal
formation process is schematically illustrated in Figure 4, which begins with a supersaturated solution
with the formation of small nuclei (approximately 10-10 m). These nuclei will grow or undergo secondary
nucleation, and under certain conditions, build clusters that will further grow to larger sizes (100 µm), and
start to agglomerate, break up and form new particulate clusters (Mersmann, 2001). Larger particles may
also adhere to surfaces, building solids deposits that represent the scale formation. The CFD-DEM
approach considers the particle-particle interactions to occur by particles whose size is in the order of 100
µm, as highlighted in Figure 4.

Figure 4. Two-phase liquid-solid flow for the scaling process including adhesion and growing effects.

A hybrid Lagrangian-Eulerian approach is applied to model the liquid-solid flow, with fluid balance
equations defined in an eulerian reference frame and the particles being individually tracked in a
lagrangian reference frame. An index j assigned for each particle assures the identify of the body
throughout the simulation. The evaluation of the movement of each particle is deterministically, with
interactions with other bodies as well with the fluid being computed by forces and torques. The force
balance, presented in equation (4), depends on the summing of fluid-related forces FFLUID[j] [N] body
forces FBODY[j] [N] and contact forces FCONTACT[j] [N], with up[j] standing for the linear velocity and mp[j]
[kg] the mass. Similarly, equation (5) defines the particle angular momentum balance, with Ip [kg.m2]
being the moment of inertia, ωp[j] [1/s] the j-particle angular velocity, TCONTACT[j] [N] the contact torque
and TFLUID[j] the fluid-related torque. The relation displayed in equation (6) allows the determination of
6

the position xp[j] [m].

[ ]
𝑚 = ∑𝐹 [ ] + ∑𝐹 [ ] + ∑𝐹 [ ] (4)

𝑑𝜔 [ ]
𝐼 =𝑇 [ ] +𝑇 [ ] (5)
𝑑𝑡
[ ]
𝑢 [ ] = (6)

The fluid-related forces comprise the drag (Crowe et al., 1998), the virtual mass (Odar & Hamilton,
1964), the Saffman lift (Saffman, 1965) and the Magnus lift (Oesterlé & Dinh, 1998). Additionally, there
is the possibility to address other relevant forces, like the dispersion force in case of turbulent flows. The
expressions to compute each one of the forces arise from experimental and analytical expressions. The
Morsi and Alexander (1972) drag law is suitable to compute the drag coefficient for spherical particles
immersed in viscous liquid flow. The body force, in the present problem, represents the net action of the
weight and buoyancy, but there is also the possibility to include other field-related forces (e.g. magnetic,
electric). Analogously, the fluid-related torque is a consequence of the fluid rotational drag (Dennis et al.,
1980), being fundamental for the computation of the Magnus lift.
The contact forces derive from the interactions with other particles and surfaces (Tsuji, 2006), being
substantial for the process of particulate bed formation which represents the scale deposition. The contact
forces unfold into collisional, frictional, and adhesive forces, each one computed through the Discrete
Element Method (DEM). In the present problem, particle-particle and particle-wall interactions relies on
the collision, allowing a diameter superposition to represent the inelastic deformation (Zhu et al., 2007).
From the superposition, the repulsive collision force is computed using a mechanical contact model
(Kruggel-Emden et al., 2007), which in the present problem being the hysteretic linear spring (Walton &
Braun, 1986). The collision force is commonly used by frictional force models (Kruggel-Emden et al.,
2008), like the Coulomb model, to estimate friction.
The constant force adhesive model, expressed in equation (7), relies in the magnitude of the adhesive
force fad [N], the i and j particle supersposition δ[ij][m] and the adhesive sphere of action δad [m].

0: 𝛿[ ] ≤ 𝛿
𝐹 [ ] = (7)
−𝑓 𝑛[ ] : 𝛿[ ] >𝛿

Regarding the continuous fluid phase, the balance equations defined in an Eulerian reference frame are
adapted to account for the effects of dense particulate flow (Popoff & Braun, 2007). Equation (8) and
equation (9) represent, respectively, the mass and momentum balance, where ρβ [kg/m3] stands for the
fluid specific mass, uβ [m/s] represents the fluid velocity, pβ [Pa] is the pressure, μβ [Pa.s] is the fluid
dynamic viscosity and g [m/s2] expresses the gravity. The particle effects over the flow regards to the
particle-fluid coupling term, denoted by fpβ [N/m3], and the fluid volumetric fraction εβ [-]. The terms fpβ
and εβ characterize CFD-DEM as a two-way coupling, enabling momentum transfer either way from the
fluid to the particles or from the particles to the fluid (Oweis et al., 2006). The term fpβ is responsible for
accounting the particulate bed pressure drop over the flow. Noteworthy, for εβ=1 and fpβ =0, the balance
equations become the Navier-Stokes monophasic flow equations.

𝜕 𝜀 𝜌
+𝛻⋅ 𝜀 𝜌 𝑢 =0 (8)
𝜕𝑡
 7

𝜕 𝜀 𝜌 𝑢
+𝛻⋅ 𝜀 𝜌 𝑢 𝑢 = −𝜀 𝛻𝑝 + 𝛻 ⋅ 𝜇 𝜀 𝛻𝑢 + 𝛻𝑢 +𝜌 𝜀 𝑔+𝑓 (9)
𝜕𝑡

Assuming that uβ is known throughout the domain, Rocky DEM 4.3 computes the forces and the
torques acting on the particle j, keeping the tracking process until all the particles on the domain have their
equations solved. Afterward, Ansys Fluent 2020 R1 receives information concerning the fluid volumetric
fraction εβ, computed from the solid volume fraction, and the momentum source term, fpβ, computed by
volume-averaging the forces and torques on each j particle on nodes of the fluid grid. The fluid flow field
is computed segregately for each flow direction (Chorin, 1968), with advection terms interpolated by the
second-order UPWIND scheme (Patankar, 1980). The PC-SIMPLE scheme (Vasquez & Ianov, 2000)
calculates the pressure fields from the velocity components.

Results
The following two sections will show the main results obtained from the numerical simulations using
the CFD-DPM and CFD-DEM approaches.

CFD-DPM Approach
Firstly it is important to notice that the difference between using a coarse (2,974,998 finite volumes)
or a fine mesh (7,369,659 finite volumes) was only 4.38 %, in terms of the adhesion rate, considering a
particle injection rate of 1.37 x 10-5 kg/s. So, it was decided to employ the coarse mesh in further
simulations aiming to reduce the computational cost.
The experimental scale test was carried out considering a total flow rate of 600 L/h of a supersaturated
solution constituted of 9.1511 g/L of HCO3 0,15 mol/L and 2.244 g/L de Ca2+ of 0.056 mol/L. An average
adhesion rate of around 3.93 x10-7 kg/s was obtained from 4 scale tests performed under the same
conditions and an average adhered mass of 2.8302 g of calcium carbonate.
According to Ansys (2020) and Nichols (2008) the k- turbulent model requires a suitable
computational mesh, specially designed in the boundary layer. Figure 5 shows the y+ profile for the two
important regions of the control volume. It was observed a maximum y+ of 8.47, indicating that the
computational mesh is appropriate to describe turbulence effects in the boundary layer.

Figure 5. y+ profile along the SSV external surface and the flow trim region.

Figure 6 illustrates the fluid velocity profile along the SSV. It can be seen low velocities recirculation
zones arising from the flow direction changes and turbulent effects in the flow trim region. In such regions
solid particles could be more susceptible to sedimentation and possible adhesion to surface due to the low
hydrodynamic drag.
8

Figure 6. Fluid velocity profile plotted at the plane (X=0, Y, Z).

In the flow trim region, the low-velocity zone approaching a high-velocity zone. Such effect is an ideal
scenario for the induction of a local pressure and velocity gradient, causing the transport of the crystals
that are in the fluid body by turbulent diffusion mechanisms to the internal surfaces of the SSV. From the
turbulent kinetic energy (see Figure 7) it is possible to confirm that the fluid flowing through the SSV,
under the simulation conditions, is capable to modify the flow pattern.

Figure 7. Turbulence kinetic energy profile plotted at the plane (X=0, Y, Z).

The anisotropy of the flow yields a gradient of turbulent kinetic energy dissipation leading to a
propagation of the vorticity in a diffusive way, generating the phenomenon of turbulent diffusion. This
phenomenon favors the movement of the crystals in the fluid generating a concentration gradient of
crystals. Figure 8 highlights the scale rate map in three different regions of the SSV. The SSV mockup
employed in the experimental evaluation is also shown in the figure. The CFD-DPM simulation
considering a critical velocity of 0.036 m/s shows that for each adhered particle out of the Flow Trim
region there are about 300000 particles adhered at this location. Overall, this simulation supports the
discussion showing that the adhesion rate is augmented when the fluid stream flows through the trim
microchannels and later diverges creating recirculation and turbulence zones. In terms of scale rate, the
region downward to the flow trim is more affected by scaling. There are also hot spots on the external
sleeve of the SSV as can be seen in Figure 8.
Table 1 shows the estimation of the adhesion rate varying the adhesion critical velocity in the CFD-
DPM simulations. The experiments reference case is an adhered mass of 2.83 g obtained as an average of
four tests considering the same conditions.
 9

Figure 8. Numerical and experimental visualization of the calcium carbonate deposition on the SSV walls.

Table 1 - Estimation of the adhesion rate as a function of the critical velocity

Adhered mass
Adhesion critical Adhesion rate considering the
Simulated time [s]
velocity [m/s] [kg/s] experimental test
conditions
0.018 226.6 2.3727E-07 1.7087
0.036 131.5 2.2910E-07 1.6499
0.042 154.6 3.8541E-07 2.7755
0.046 100.1 4.1635E-07 3.0010
0.056 127.9 5.0190E-07 3.9200

It was observed an increase of 56.4 % of the adhesion rate when the critical velocity is reduced from
the higher to the lower value simulated. Important to note that the critical velocity of 0.018 m/s was
10

calculated using Equation (2) from the work of Abd-Elhady et al. (2009). The best result was reached
when a critical velocity of 0.042 m/s was employed considering the reference adhesion rate of 3.92925e-
7 kg/s and adhered mass of 2.83 g. This result suggests that the critical velocity of 0.042 m/s can be
employed in future studies with considerable accuracy to describe the scale phenomenon in SSV taking
in to account experimental conditions.

CFD-DEM Approach Results

The concept of the SSV follows the interpretation and simplification of fabrication and assembly draws
of a commercial valve from an undisclosed provider. The simplified geometry, shown in Figure 9, consists
of the outer sleeve (part in red), assembled within an open-hole-completed wellbore (colored in light gray).
The valve trim consists of 16 holes arranged in two rows and 32 grooves in the inner sleeve (green
colored). The flow inlets the geometry through the annular space defined between the well wall and the
outer sleeve, either way upward or downward. The fluid accelerates in the adjacency of the hole-groove
interface and reaches the inner sleeve, flowing upward toward the production tubing.

Figure 9. Isometric view of the SSV assembly with the indication of flow boundary conditions and particle
injection surfaces.

Problem parameters listed in Table 2 comprise the influence of solids concentration varying the solid
mass flow rate for each one of the cases. The particulate system is monodisperse with 0.5 mm spherical
particles of specific mass 4500 kg/m3. One should note that DEM is versatile for particle injection,
allowing the prescription of size distribution functions which may also be time-dependent if necessary.
The fluid properties resemble equivalent-properties of an oil-in-water emulsion subjected to downhole
conditions. Properties have been evaluated by Petrobras through PVT-testing fluid samples production
sites. The fluid flow rate is constant at the inlet surfaces throughout the simulation time.

Table 2. Liquid-solid flow with adhesion parameters for the simulation with CFD-DEM.
Case 1 Case 2 Case 3
Solid mass flow rate [kg/s] 3.5·10-4 7.0·10-4 1.4·10-4
Total particles injected [# particles] 37,433 74,867 149,734

CFD-DEM results are displayed in Figure 10 at the same instant. Incrementing the solids mass flow
rate at the injection surfaces, shown in Figure 9, increases the number of particles in the annular region,
bringing particles close to one another. The proximity among particles enables the action of the adhesive
force, resulting in a scenario prone to cluster build up. The particulate clusters tend to adhere to the valve
 11

surfaces, remaining bounded. It is possible to visualize particle deposition on the perimeter of the holes
and also in the hole-groove trim. The detail of the trim shows the reduction in the available flow area for
each case.

Figure 10. CFD-DEM results for the liquid-solid flow with adhesion in terms of solids concentration.

The deposited particles, shown in Figure 10, are a particulate bed, resembling a porous medium that
imposes a pressure drop over the flow field depicted in Figure 11 by the plot of the dimensionless pressure
P [-] over the dimensionless time τ [-], with pref [Pa] standing for the pressure difference of the one-phase
flow (without particles) and tref [s] as the reference time. The pressure P starts to increase at time 0.2,
exactly when the first particles that have been released reach the SSV trim, and keeps going up as more
12

clusters deposit on the grooves. The higher the solid mass flow rate, the higher is the pressure P at the end
of the simulation. The increase is approximately proportional, as for 3.5 10-4 kg/s P increases roughly
25%, whilst for 1.4 10-4 kg/s P nearly doubles.

2.25
3.5 10-4 kg/s
-4
7.0 10 kg/s
2 -3
1.4 10 kg/s

1.75
P=p/pref

1.5

1.25

1
0 0.2 0.4 0.6 0.8 1
=t/tref

Figure 11. Dimensionless pressure over time demonstrating the influence of the solid particle adhesion over
the flow field.

Conclusions
A preliminary study of solids deposition on the surfaces of a sliding-sleeve valve using the CFD-DEM
numerical technique has been presented. The employment of the DEM accounted for the adhesive force,
which led to agglomerates build-up and particle adhesion on the surfaces. The four-way-coupling scheme
enabled the evaluation of the scaling process over the flow field. Spite of the necessity to refine DEM
calibration to properly represent adhesive interactions between crystals, the present work envisions the
inclusion of the hydrodynamic effect over the scale formation observed in wellbore configurations and
production facilities.
From a simple adjustment of the parameter adhesion critical velocity could obtain an adhesion rate
similar to that one observed in the experimental study. Although the literature provides a procedure to
quantify the threshold velocity (Equation (2)), the calculated value, 0.018 m/s, lead to an underestimation
of the adhesion rate. One shoul be noted that the CFD-DPM approach coupled with the adhesion model,
neglects effects acting in the physics of the problem, such as electrostatic factors, more complex fluid
dynamics effects and also the estimation of the precipitation rate provided by the thermodynamics
software. Considering all these effects, the results are satisfactory from the point of view of empirical
adequacy and quantitative reproduction of the calcium carbonate scaling phenomenon in modeled and
simulated SSV.
The complete representation of the full scaling process requires the consideration of other aspects, such
as the presence of oil and gas in the flow and the effect of CO2 in the equilibrium of the calcite formation.
Both experimental and numerical tasks are ongoing.

References
Abd-Elhady, M. S., Rindt, C. C. M., & Van Steenhoven, A. A. (2009). “Optimization of flow direction to
minimize particulate fouling of heat exchangers”. Heat Transfer Engineering, 30(10-11), 895-902.
Barbosa, M. V., De Lai, F. C., & Junqueira, S. L. M. (2016, July). Numerical Analysis of Particulate Flow
Applied to Fluid Loss Control in Fractured Channels. ASME 2016 Heat Transfer Summer
Conference, HT 2016, Collocated with the ASME 2016 Fluids Engineering Division Summer Meeting
and the ASME 2016 14th International Conference on Nanochannels, Microchannels, and
 13

Minichannels. https://doi.org/10.1115/FEDSM2016-1028
Barbosa, M. V, De Lai, F. C., Franco, A. T., & Junqueira, S. L. M. (2015). Parametric analysis of
particulate flow for fluid loss control to sealing fractured channels. 23rd ABCM International
Congress of Mechanical Engineering - COBEM 2015.
Barbosa, M. V, De Lai, F. C., & Junqueira, S. L. M. (2019). Numerical Evaluation of CFD-DEM Coupling
Applied to Lost Circulation Control: Effects of Particle and Flow Inertia. Mathematical Problems in
Engineering, 2019. https://doi.org/10.1155/2019/6742371
Bouamra, R., Carneiro, G., Machado, P., Feliciano da Silva, M., Franquiz, G., Guan, H., & Lindvig, T.
(2019, October). ScaleProTect–Scale Deposition Modeling in Pre-Salt Reservoir. In Offshore
Technology Conference Brasil. Offshore Technology Conference.
Chorin, A. J. (1968). Numerical solution of the Navier-Stokes equations. Mathematics of Computation,
22(104), 745–745. https://doi.org/10.1090/S0025-5718-1968-0242392-2
Crowe, C., Sommerfeld, M., & Tsuji, Y. (1998). Multiphase flow with droplets and particles. CRC Press.
De Lai, F. C., Franco, A. T., & Junqueira, S. L. M. (2014). Numerical simulation of liquid-solid flow in a
channel with transversal fracture. 15th Brazilian Congress of Thermal Sciences and Engineering -
ENCIT 2014.
Dennis, S. C. R., Singh, S. N., & Ingham, D. B. (1980). The steady flow due to a rotating sphere at low
and moderate Reynolds numbers. Journal of Fluid Mechanics, 101(2), 257–279.
https://doi.org/10.1017/S0022112080001656
Droppert, V.S., Hatscher, S, Djayapertapa, L., Byrne, M.T (2020), Improved Scale Squeeze Placement in
Horizontal Wells with Pressure Variations Along the Well using Non-newtonian Fluids with Inflow
Control Devices, SPE 200675, presented at the Virtual SPE Oilfield Scale Conference, june 2020
Guan, H., Bouamra, R., Lindvig, T., & Vernus, J. C. (2018, June). Scale Risk Assessment and Novel
Coating for Smart Completion: Scale Simulation and CFD Modelling Approach. In SPE
International Oilfield Scale Conference and Exhibition. Society of Petroleum Engineers.
Kruggel-Emden, H., Simsek, E., Rickelt, S., Wirtz, S., & Scherer, V. (2007). Review and extension of
normal force models for the Discrete Element Method. Powder Technology, 171(3), 157–173.
https://doi.org/10.1016/j.powtec.2006.10.004
Kruggel-Emden, H., Wirtz, S., & Scherer, V. (2008). A study on tangential force laws applicable to the
discrete element method (DEM) for materials with viscoelastic or plastic behavior. Chemical
Engineering Science, 63(6), 1523–1541. https://doi.org/10.1016/j.ces.2007.11.025
Lima, G. H., De Lai, F. C., & Junqueira, S. L. M. (2017). Numerical simulation of particulate flow for
static mud cake process over heterogeneous porous media. 24th ABCM International Congress of
Mechanical Engineering - COBEM 2017. https://doi.org/10.26678/ABCM.COBEM2017.COB17-
5950
Maciel, R. S, Pereira, R., de Assis, F., Fieni Fejoli, R., Leibsohn Martins, A., & Duarte Ferreira, M. V.
(2019, September). Enhancing Scale Prediction in Pre-Salt Wells Using Numerical Simulation.
In SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers.
Martins, A. L., Castro, B. B., Schluter, H. E. P., Ferreira, M. V. D., Achy, A. R. A., & Pepe, I. M. (2020,
May). Offshore Field Experience with Non-Chemical Oilfield Scale Prevention/Remediation
Strategies in Brazil. In Offshore Technology Conference. Offshore Technology Conference.
Mersmann, A. (2001). Crystallization Technology Handbook - Second Edition Revised and Expanded. In
Marcel Dekker Inc.
Morsi, S. A., & Alexander, A. J. (1972). An investigation of particles trajectories in two-phase flow
systems. Journal of Fluid Mechanics, 55(2), 193–208. https://doi.org/10.1017/S0022112072001806
Odar, F., & Hamilton, S. (1964). Forces on a sphere accelerating in a viscous fluid. Journal of Fluid
Mechanics, 18(2), 302–314. https://doi.org/10.1017/S0022112064000210
Oesterlé, B., & Dinh, T. B. (1998). Experiments on the lift of a spinning sphere in a range of intermediate
Reynolds numbers. Experiments in Fluids, 25(1), 16–22. https://doi.org/10.1007/s003480050203
Oweis, G. F., Ceccio, S. L., Matsumoto, Y., Tropea, C., Roisman, I. V, Tsuji, Y., Lyczkowski, R., Troutt,
14

T. R., Eaton, J. K., & Mashayek, F. (2006). Multiphase interactions. In C. T. Crowe (Ed.), Multiphase
Flow Handbook2. CRC Press.
Patankar, S. V. (1980). Numerical Heat Transfer and Fluid Flow. McGraw-Hill.
Poletto, V. G., De Lai, F. C., Ferreira, M. V. D., Martins, A. L., & Junqueira, S. L. M. (2020). Numerical
Simulation to Study Scale Formation on SSV Valves. SPE Virtual International Oilfield Scale
Conference and Exhibition. https://doi.org/10.2118/200689-ms
Poletto, Vinicius G., De Lai, F. C., & Junqueira, S. L. M. (2020). CFD–DEM simulation of mud cake
formation in heterogeneous porous medium for lost circulation control. Journal of the Brazilian
Society of Mechanical Sciences and Engineering, 42(7). https://doi.org/10.1007/s40430-020-02450-
y
Popoff, B., & Braun, M. (2007). A lagrangian approach to dense particulate flow. 6th International
Conference on Multiphase Flow, ICMF 2007, 510–521.
Saffman, P. G. (1965). The lift on a small sphere in a slow shear flow. Journal of Fluid Mechanics, 22(2),
385–400. https://doi.org/10.1017/S0022112065000824
Tsuji, Y. (2006). Particle-particle collision. In C. T. Crowe (Ed.), Multiphase Flow Handbook. CRC
Taylor & Francis.
Vasquez, S. A., & Ianov, V. A. (2000). A phase coupled method for solving multiphase problem on
unstructured meshes. 2000 ASME Fluids Engineering Division Summer Meeting.
Walton, O. R., & Braun, R. L. (1986). Viscosity, granular‐temperature, and stress calculations for shearing
assemblies of inelastic, frictional disks. Journal of Rheology, 30(5), 949–980.
https://doi.org/10.1122/1.549893
Zhu, H. P., Zhou, Z. Y., Yang, R. Y., & Yu, A. B. (2007). Discrete particle simulation of particulate
systems: Theoretical developments. Chemical Engineering Science, 62(13), 3378–3396.
https://doi.org/10.1016/j.ces.2006.12.089