You are on page 1of 39

Accepted Manuscript

Experimental analysis of pressure drop in the flow of Newtonian fluid in coiled tubing

Caroline Eulino Gonçalves Pereira, Gabriel A. da Cruz, Leonidas Pereira Filho,


Lorena R. Justino, Eduardo C. Paraiso, José M. Rocha, Luis A. Calçada, Cláudia M.
Scheid

PII: S0920-4105(19)30419-X
DOI: https://doi.org/10.1016/j.petrol.2019.04.082
Reference: PETROL 6021

To appear in: Journal of Petroleum Science and Engineering

Received Date: 21 November 2018


Revised Date: 1 March 2019
Accepted Date: 23 April 2019

Please cite this article as: Gonçalves Pereira, C.E., da Cruz, G.A., Filho, L.P., Justino, L.R., Paraiso,
E.C., Rocha, José.M., Calçada, L.A., Scheid, Clá.M., Experimental analysis of pressure drop in the flow
of Newtonian fluid in coiled tubing, Journal of Petroleum Science and Engineering (2019), doi: https://
doi.org/10.1016/j.petrol.2019.04.082.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT

EXPERIMENTAL ANALYSIS OF PRESSURE DROP IN THE FLOW OF


NEWTONIAN FLUID IN COILED TUBING

Abstract

In order to study the fluid flow in the coiled tubing system, an experimental unit with 376 meters

PT
of wrapped copper tube with ½ in diameter, and similarity to an actual coiled tubing system was
built. Pressure transmitters and thermocouples were installed in this unit to monitor pressure and
temperature along tube length. Tests were carried out with water, Newtonian fluid, in the flow

RI
range between 0.05 and 0.65 m3/h, in order to predict pressure drop in each double layer and in
the whole coiled tubing system. Flow rate and pressure drop data obtained in the experimental

SC
unit allowed the evaluation of critical Reynolds number, friction factor correlations in laminar
and turbulent regime, and the influence of the curvature and tube length on pressure drop. A
methodology for predicting pressure drop in the coiled tubing system was proposed and the
percentage error between the experimental and calculated pressure drop values was less than 5%.

U
AN
Keywords: Experimental apparatus; pressure drop; coiled tubing; water.
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

1. Introduction

Coiled tubing system has many uses in the oil industry. Among them, the plugging and
abandonment (P&A) process, through cementing. This system consists mainly of a long, flexible,
continuous steel tube that has part of its length wound on a reel while another part is directed to
the well by an injector and a rotary table. Tube’s length can be approximately 5,000 meters with

PT
outer radius of 1.5 in, and reel diameter of 1.0 meter. For being a continuous pipe, the coiled
tubing allows operation time reduction compared to drill pipes, which need to be connected in
order to be used. This advantage makes the coiled tubing system very attractive, especially when

RI
the activity is the abandonment of an oil well, which is an operation that has no financial return.
Fluid flow in coiled tubing has a different behavior from that found in straight tube. The
existence of the centrifugal force in fluid flow through coiled tubing contributes to the increase of

SC
flow resistance. This increase requires the supply of greater mechanical energy to the fluid,
compared to a straight section with the same dimensions.
Due to the frictional pressure loss and heat transfer which occurs during the flow of the

U
fluid in this system, the physical, chemical, and rheological properties of cement slurries change
over time, and, with temperature increase, decreasing the hardening time. To avoid the risk of
AN
setting the cement slurry before reaching the desired location, additives such as mid-temp
retarder is added in excess. The importance of predicting pressure drop is due to its influence on
temperature profile, and, consequently, on the rheological properties of the fluids. Besides,
M

pressure drop prediction enables the design of pump pressure requirements to pump fluids
through coiled tubing systems.
Pressure drop in the curved section must be adequately predicted for pump pressure
D

planning. Taking the design of pump pressure into account, this work focused on studying the
flow of Newtonian fluids in coiled tubing in order to predict the pressure drop in this flow. Thus,
TE

it can be highlighted: the determination of flow regime of Newtonian fluids, the evaluation of
critical Reynolds correlations, friction factor correlations for laminar and turbulent regime, and
the influence of tube curvature on pressure drop.
EP

The experimental study of pressure drop in coiled tubing has been carried out by authors
such as McCann and Islas (1996), Medjani and Shah (2000), Rao (2002), Jain et al. (2004), Shah
et al. (2009), Guan et al. (2014), among others. However, this work presents as a differential the
C

analysis of pressure drop in different layers of the same coiled tubing system.
Because of the particularity of curvature in coiled tubing systems, three fundamental
AC

concepts need to be defined in this context: curvature ratio, Dean number, and secondary flow. In
addition, critical Reynolds and friction factor correlations that take into account these concepts
need to be considered.
Curvature ratio is the ratio of the inner radius of the tube to the radius in which the tube is
curved. Medjani and Shah (2000) present a curvature ratio considering the number of layers
formed by the tube
= ,
. (1)
ACCEPTED MANUSCRIPT

where is the inner radius of tube, is the curvature radius, is the reel radius, and is the
number of the layer.
The different layers that a coiled tubing system can have are exemplified in Fig. 1.

Fig. 1. Illustration of different layers of tube wound on a reel.

PT
Dean number is the dimensionless that is a function of the curvature ratio, presenting the

RI
effects of centrifugal forces. Dean (1927, 1928) was the first to present an analytical study of the
flow of incompressible fluids in curved tubes, with a slight curvature, and in a laminar regime
(Shaqlaih and Kamel, 2013; Ghobadi and Muzychka, 2015). By this analysis, Dean (1927)

SC
presents an important criterion for slow flow, which relates the reduction of flow rate due to the
curvature and, from this criterion, the dimensionless known as Dean number was presented.
Considering the definition of this dimensionless, Ghobadi and Muzychka (2015) affirm that

U
several authors present different descriptions. However, the format most applied by several
authors is that presented by White (1929)
AN
= ,
(2)
M

where is the Dean number, is the Reynolds number, is the inner radius of tube, and is
the curvature radius.
D

The third fundamental concept in fluid flow in coiled tubing is the secondary flow. The
secondary movement manifests in the cross-section of a curved tube due to the imbalance
TE

between the inertial and centrifugal forces acting on the radial direction of the coil (Medjani and
Shah, 2000; Jain et al., 2004; Ghobady and Muzichka, 2015). This type of flow is composed of
two parallel vortices and, because they have been described by Dean, they are also known as
EP

Dean vortices (Rao, 1999). Although the presence of two vortices is predominant, this amount
may vary. Hasan (2013) states that at high Reynolds numbers, due to the higher centrifugal force
in relation to viscous force, the vortices can be split in two or more, forming several small
C

vortices.
Taking a fluid flow into consideration, three flow regimes can be stablished: laminar,
AC

transition, and turbulent. In order to determine each regime, Reynolds number is used. The
critical Reynolds number ( ) is determined for verification of the transition state.
Flow transition regime can be verified in graphs of pressure loss (∆ ) versus flow rate ( ).
The transition is defined by the change in the slope of the experimental values curves, being the
critical point, the intercession of these curves. The critical Reynolds number can also be
visualized in a graph of friction factor versus Reynolds number (Ghobadi and Muzychka, 2015;
Zhou and Shah, 2004b).
ACCEPTED MANUSCRIPT

White (1929) conducted water experiments on coiled tubing. The author concluded that the
turbulent flow started from 6000 of Reynolds number, with – curvature ratio of the system. At
another curvature ratio ( ), the turbulent regime started in the range of 2250 to 3200.
Table 1 presents critical Reynolds equations quoted in articles written by Zhou and Shah
(2004b), Pawar et al. (2015), and Ghobadi and Muzychka (2015).

PT
Table 1. Experimentally developed critical Reynolds number correlations for curved pipe

RI
Newtonian fluids flow.
Source: Zhou and Shah (2004b); Kubair and Varrier (1961/2); Schmidt (1967); Ghobadi and
Muzychka (2015); Cioncolini and Santini (2006).

SC
Besides changing flow regime, the secondary flow in curved tubes causes a greater

U
frictional loss, compared to a straight section, at the same flow rate and with the same pipe length
(Jain et al., 2004; Shaqlaih and Kamel, 2013). According to Rao (2002), the intensity with which
AN
this flow affects pressure drop depends on curvature ratio, tube radius, flow regime, and fluid
characteristics, Newtonian or non-Newtonian.
Guan et al. (2014) analyzed pressure drop in the flow of water in a curved tube system with
M

a curvature ratio of 0.0242. The authors observed a pressure drop of 11 to 17% higher in curved
tube compared to straight tube.
In order to determine pressure drop to a specific system, it is necessary to choose a friction
D

factor correlation which describes well the system.


TE

A list of friction factor correlations for Newtonian fluids in curved tubes can be found in
papers such as those of the authors: Ali (2001), with the study of pressure drop in helical tubes,
Zhou and Shah (2004b), with a review article on fluid flow in coiled tubing, and Ghobadi and
Muzychka (2015) with a review of heat transfer and pressure drop in the flow of Newtonian
EP

fluids in curved tubes. Selected correlations are presented in Table 2.


C

Table 2. Friction factor correlations for laminar Newtonian fluid flow.


AC

Source: Zhou and Shah (2004b); Barua (1963); Ghobadi and Muzychka (2015); Ito (1969);
Dennis (1980).

As can be seen in Table 2, the friction factor correlations in the laminar regime are
expressed as the ratio of the friction factor in coiled tubing ( ) and straight tube ( ). In this
work, the friction factor in straight section used was the Fanning friction factor.
All the equations in Table 2 were developed numerically or analytically, except for eq. (10)
that had its development from experimental data.
ACCEPTED MANUSCRIPT

Table 3 shows the friction factor correlations for turbulent Newtonian fluids. Zhou and
Shah (2004b) point out that there is an agreement of results between the eqs. (13), (14) and (15).

Table 3. Friction factor correlations of turbulent Newtonian fluid flow.


Source: Ali (2001); adapted from Zhou and Shah (2004b).

PT
In order to evaluate not only friction factor but also critical Reynolds number correlations,

RI
an experimental coiled tubing unit, with dynamic similarity with an actual coiled tubing system,
was built. This evaluation enabled the prediction of pressure drop, and the determination of flow
regimes. The contribution of this work for a literature review is highlighted on the experimental

SC
unit. The CT unit allows the measurement of pressure drop in different layers of a single coiled
tubing system, besides the possibility of measuring fluid and tube temperature in all CT system’s
layers. This differential permits not only the prediction of pressure drop of the whole system, but

U
also allows checking the contribution of each section.
AN
2. Materials and methods

2.1. Experimental Unit


M

For the experimental evaluation of the flow of Newtonian fluids in coiled tubing, a copper
coiled tube that has eight layers, a length of approximately 376 meters, and ½ in diameter was
D

used. Since the copper coil has eight layers, it therefore has eight curvature ratios which in this
case range from 0.0138 to 0.0177. Temperature and pressure can be measured in the edge of each
TE

layer by thermocouples and pressure transmitters, respectively. Fig. 2 shows the coil with the
installed instruments, which is now known as the coiled tubing unit (CT). This unit was
developed respecting a dynamic similarity to an actual case studied.
EP

Fig. 2. Coiled tubing (CT) unit.


C
AC

In order to monitor the pressure in different layers of the CT unit, five gauge pressure
transmitters were used. The transmitters have a remote seal, which allows the fluid pressure to be
measured without the fluid coming into contact with the transmitter head. The transmitters are
from Yokogawa company, and their calibration range was from 0 to 24 bar. The pressure values
can be read by means of a display, however, for better data acquisition, the pressure results as
well as flow rate, and temperature are observed and saved by means of software connected to a
PLC (Programmable Logic Controller).
ACCEPTED MANUSCRIPT

The CT unit was installed to a pump system, thus composing the entire experimental unit.
Fig. 3 presents a schematic layout of the experimental unit.
For a better understanding of the fluid path during an experiment, the legend items in Fig.
3 will be indicated in brackets in the text.
The fluid is added to the tank [2] which has a stirring system [3] and a jacket for
temperature control. The temperature control system is composed of three resistors, coupled to

PT
the tank jacket, and a chiller [1], with a pump system for this jacket. After controlling the
temperature of the fluid, its flow is initiated by means of a positive displacement pump [5]. The
pump rotation and, consequently, fluid flow is controlled by means of a frequency inverter [4].

RI
The pump moves the fluid to a 2-in. steel pipe which is the fluid circulation line [6], and, through
two valves, the fluid is directed to the CT unit [8]. The flow line valve is also used to aid in flow
rate control, since it allows or not the passage of the fluid to that line, changing the flow that goes

SC
to the CT unit line [7]. After leaving CT unit, the fluid passes through a flow meter [10], and then
returns to the tank. The fluid flows in this closed loop.

U
AN
Fig. 3. Experimental unit layout.
M

2.2. Calculation Methodology


D

2.2.1. Pressure Drop


TE

The experimental Fanning friction factor (eq. (17)) was determined by means of the
mechanical energy balance evaluation from the modified Bernoulli Equation (Fox et al., 2006).
EP


= 〈 〉"
,
(3)

where ∆ is the pressure drop, is the diameter of the tube, $ is the fluid density, % is the length
C

of tube, and 〈&〉 is the average velocity.


AC

The experimental friction factor was used to evaluate the friction factor correlations of
Tables 2 and 3. After the evaluation of friction factor correlations and verification of the best
correlation, pressure drop of the CT unit and between each double layer has been evaluated.
Since the layer of tube changes, the curvature ratio will change which will consequently change
the friction factor. Therefore, for these changes to be taken into account, the calculation of the
total pressure drop has been calculated as a sum of pressure drop in each layer
ACCEPTED MANUSCRIPT

∆ " " "


' ) = '2 +
) + '2 "
) + ⋯ + '2 7
4
) .
( ( ,- ./012 ( 45 ./012 ( 7-8 9:;< (4)

All analyses performed between experimental values and calculated by correlations will be
described by means of the absolute percentage error

PT
=>?@ABC D E FCGH @ % = |KLMN OPN4-QR SQRTNUVWXYZXW[\] ^WXZ\|
KLMN OPN4-QR SQRTN
× .
(5)

RI
The experimental value is the reference value and the calculated value the one to be
checked.

SC
2.2.2. Tube Geometry by Layer

Knowing that the tube has a given wall thickness, it was decided to modify eq. (1) for the

U
calculation of the curvature ratio considering the outer radius of the tube ( `ab ) instead of the
inner radius ( )
AN
= .
.
cT- (6)
M

From the analysis of the organization of the tube wrapped on a reel, the calculation of the
D

length of coiled tube in each layer (% ) was calculated from


TE

% = 2d[ + 2 −1 `ab ] = % di + 2 −1 j,
(
cT- cT- (7)

where % is the reel width, `ab is the outer radius of tube,


EP

is the reel radius, and is the


number of the layer.
Determination of the number of layers that are formed on a reel ( ), from the value of the
total tube length (%k ) or vice versa was calculated as the sum of all terms of an arithmetic
C

progression
AC

%k = d% i' )+ j.
(
cT- (8)
In the case of the CT unit, the number of layers is known and the total tube length is found.
2.3. Experimental Methodology: Pressure Drop

The main objective of pressure drop test is to evaluate friction factor correlations and the
methodology adopted to predict this loss. Since the pressure transmitters are gauge type, the
ACCEPTED MANUSCRIPT

differential pressure was calculated every two layers (1st and 2nd, 3rd and 4th, 5th and 6th, 7th and
8th), as well as the differential pressure of the entire CT unit.
Experiments were performed with water, Newtonian fluid, in the flow rate range of 0.05 to
0.65 m3/h, and in the flow rate range in turbulent flow regime, both with intervals of 0.05 m3/h.
In the experiment of broader range of flow rate the following were evaluated: flow regime, by the
analysis of critical Reynolds number, and correlations of friction factor for the laminar regime. In

PT
the turbulent regime experiment, in addition to evaluating friction factor correlations, the
influence of the curvature ratio and tube length was also analyzed. The experiment in the
turbulent flow range was done in triplicate.

RI
3. Results and discussion

SC
3.1. Critical Reynolds Evaluation (lmn )

In order to find the critical Reynolds number ( ), it was first necessary to determine the

U
critical flow rate ( ). The critical flow rate determination was performed by means of the
pressure drop graph (∆ ) versus flow rate ( ). This evaluation was done for each double layer
due to the fact that the flow regime is influenced by the curvature ratio ( ⁄ ).
AN
Fig. 4 shows the graphs used for the evaluation of the flow regime of the 5th and 6th layer.
The same graphical analysis was performed for the other double layers.
M

As can be observed in Fig. 4, after the evaluation of the slope change of the experimental
data, a linear regression of the points of each region, laminar, transition, and turbulent, was made
and the meeting of the laminar and transition region was determined as the critical flow rate. To
D

assist in the evaluation of flow regimes, the experimental friction factor ( p:77q7( ) versus
TE

Reynolds number ( ) data was plotted (Fig. 4b).

Fig. 4. (a) Pressure drop versus flow rate, and (b) experimental friction factor versus Reynolds
EP

number of 5th and 6th layers.


C

The critical flow rate values for all the double layers were close to each other, and are
between 0.20 and 0.21 m3/h. This similar result may be a consequence of very close curvature
AC

values (0.0138 - 0.0177).


The critical Reynolds number calculated for each double layer and for the entire system
was between 6394 and 6690, the lowest value for the 7th and 8th layers, with a curvature ratio
between 0.0138 and 0.0143, and the highest value, for the 1st and 2nd layer, with curvature ratio
between 0.017 and 0.0177. The critical Reynolds result indicated that the critical point for coiled
tubes is much larger than for straight tubes. It is worth noting that for a Newtonian fluid flowing
through a straight tube, the critical Reynolds is between 2100 and 2300. This greater extension of
ACCEPTED MANUSCRIPT

the laminar regime was justified by Ghobadi and Muzychka (2015), due to the presence of the
secondary flow, which stabilizes the laminar characteristic of the flow.
Faced with the fact that the experimental determination of the critical Reynolds is not
always possible, critical Reynolds correlations for Newtonian fluids were evaluated by looking
for the ones that best represent the experimental data. Fig. 5 shows a graph of experimental
critical Reynolds values and those calculated as a function of the inverse of the curvature ratio
( ⁄ ).

PT
RI
Fig. 5. Experimental and calculated critical Reynolds number versus curvature ratio.

SC
A critical Reynolds correlation was proposed based on the experimental data and the
correlation format of Srinivasan et al. (1970)

U
:
= 2,100 + t x y,
u v w
AN
(9)

where G and > are parameters to be estimated.


From eq. (23), a non-linear estimation of the parameters G and > was performed, taking as
M

data the experimental Reynolds values as a function of the inverse of the curvature ratio. The
estimation was done using the Levenberg-Marquardt algorithm for nonlinear adjustment with the
D

values of the parameters found


TE

,{|
= 2,100 + z •.
(
}.~
u v w (10)

The standard error of parameters G and > were, respectively, 4,965.5 and 0.08.
EP

The black colored line in Fig. 5 represents the values calculated by eq. (24) and the pink
range, the 95% confidence band, that is, the upper and lower confidence limit of a mean value
C

predicted by means of the equation.


The absolute percentage error between the experimental critical Reynolds values and those
AC

calculated by the equations presented in Table 1 and by eq. (24) are given in Table 4. Absolute
errors less than or equal to 20% are highlighted.

Table 4. Absolute percentage error between experimental and calculated


critical Reynolds number.
ACCEPTED MANUSCRIPT

As can be seen in Table 4, from the correlations found in the literature, the proposal by
Schmidt (1967) is the one closest to the experimental data, with a mean error of 15.7%. The
proposed correlation (eq. (24)) presented a percentage error of less than 2%.
One possible justification for high percentage errors is the imprecision in graphic methods.
Ghobadi and Muzychka (2015) compared the results of critical Reynolds number between
the correlations presented in Table 1 and others. The authors observed that for the same curvature

PT
ratio, the result was a range of critical Reynolds number instead of a specific number. That is,
even a correlation developed for a range of curvature ratio which covers the range of other
equation, they not necessarily will present the same critical Reynolds number.

RI
3.2. Friction Factor Correlations Evaluation: Laminar Regime

SC
Once the experimental critical Reynolds was determined, it was possible to identify the
flow rates at which the regimes were laminar and turbulent.
Since the calculated friction factor is a function of the curvature ratio, friction factor versus

U
Reynolds number graphs for each double layer were generated with the experimental friction
factor values and those calculated by the equations presented in Table 2. Fig. 6 presents the graph
AN
to evaluate the friction factor correlations in the laminar regime of the 3rd and 4th layer, but the
same analysis was done for the other double layers.
M

Fig. 6. Experimental and calculated friction factor of 3rd and 4th layers versus Reynolds number,
D

in laminar regime.
TE

Analyzing Fig. 6 it is possible to observe that the correlation proposed by Adler (1934) is
the closest to the experimental data.
EP

For a better evaluation of the correlations in all the studied double layers, absolute
percentage errors between the experimental and calculated friction factor results were determined
and are shown in Table 5. Errors smaller than 5% are highlighted.
C
AC

Table 5. Absolute percentage error between experimental and calculated friction factor
in laminar regime (flow rate: 0.05 – 0.15 m3/h).

It can be seen from Table 5 that percentage errors vary with the double layer, indicating
that a correlation suitable for a given curvature ratio will not necessarily be the most indicated for
another system with a different curvature ratio. However, the percentage errors at a given flow
rate at the different curvature ratios are close to each other, which suggests that, for near
curvature ratios, percent errors are similar. This result is important since in a coiled tubing system
ACCEPTED MANUSCRIPT

several curvature ratios are assumed in the same system, and therefore having close curvature
ratios, it is possible to choose a correlation that predicts the most appropriate friction factor for
the system. This can be verified in the result shown in Table 5, where the errors obtained by
evaluating Adler's correlation (1934) were the lowest compared to the other correlations.

3.3. Friction Factor Correlations: Turbulent Regime

PT
Figure 7 is the graphical representation of the comparison between the experimental friction
factor and calculated by each equation presented in Table 3 of the 1st and 2nd double layer, but the

RI
same analysis was done for all other double layers.

SC
Fig. 7. Experimental and calculated friction factor of 1st and 2nd layers versus Reynolds number,
in turbulent regime.

U
Comparing the behavior observed in Fig. 7, the four correlations have very similar slopes
AN
and the equations of Mishra and Gupta (1979) and Srinivasan et al. (1970) are superimposed
almost in all cases. Still in the evaluation of this figure, it is noticed that the correlation of Ito
(1959) is the one closest to the experimental values.
M

The absolute percentage errors between the experimental and calculated friction factor
results in the turbulent regime for all the double layers analyzed are presented in Table 6. Errors
D

smaller than 5% are highlighted.


TE

Table 6. Absolute percentage error between experimental and calculated friction factor
in turbulent regime (flow rate: 0.20 – 0.65 m3/h).
EP

As can be seen in Table 6, the correlation of Ito (1959) was the one that presented the
smallest percentage errors in all cases, with all errors showing less than 5%, being the most
C

adequate for this geometry, in the water flow in the turbulent regime. Both Mishra and Gupta
AC

(1979) and Srinivasan et al. (1970) correlations can be considered secondary appropriate options
for this system since they present similar results.

3.4. Evaluation of Total Pressure Drop Calculation Methodology: Turbulent Regime

For the evaluation of pressure drop of the whole system, the evaluated correlations for the
calculation of friction factor in the turbulent regime were used in the sum of the calculated
pressure drop for all the layers. The calculated pressure drop was compared to the pressure drop
measured in the CT unit (Fig. 8).
ACCEPTED MANUSCRIPT

Fig.8. Experimental and calculated CT unit pressure drop in turbulent regime.

The absolute percentage errors between the experimental and calculated pressure drop are
shown in Table 7, with values less than or equal to 5% being highlighted.
It is noteworthy, in Table 7, that the calculated pressure drop results had a similarity to the

PT
result of the evaluation of friction factor correlations in the turbulent regime, since the choice of
the equation to predict the friction factor is fundamental for predicting the pressure drop.
However, what is desired to emphasize in the evaluation of Table 7 is that, the choice of the Ito

RI
(1959) correlation presented percentage errors lower than 5% in all the evaluated flow rates. This
indicates that the sum of pressure drop summations of all layers - taking into account the
curvature ratio, diameter, and length of each layer - made in the proposal of eq. (18) were

SC
suitable.

U
Table 7. Absolute percentage error between experimental and calculated pressure drop.
AN
3.5. Influence of Curvature Ratio and Tube Length: Turbulent Regime
M

Due to the influence that the curvature ratio and the tube length in each layer have on the
calculation of friction factor, and pressure drop, it was decided to verify the influence of these
D

two factors on the result of pressure drop.


At the first moment, to evaluate only the curvature ratio, the pressure drop result was
TE

normalized for every 10 meters of tube. This normalization was made in order to be able to
compare the pressure drop result between the double layers without taking into account the
difference in length between these layers.
A graph of pressure drop in every 10 meters (∆ ⁄10€ ) versus flow rate ( ) was generated
EP

in order to evaluate the difference in pressure drop between the double layers studied (Fig. 9).
C

Fig.9. Pressure drop of each double layer each 10 meters.


AC

Fig. 9 shows that with the increase in flow rate, the difference between the pressure drop of
the layers increases. It is also noted that the higher the curvature ratio, the greater the pressure
drop. In a general evaluation, however, the difference between the curvature ratios did not
generate such a significant difference in the pressure drop, since the curvature ratios have close
values in this work. This behavior was also observed in the work of Jain et al. (2004). However,
Jain et al. (2004) have shown that the effects of curvature for non-Newtonian fluids are more
significant than for Newtonians.
ACCEPTED MANUSCRIPT

The effect of length was also evaluated, considering the length of each double layer in the
pressure drop. Thus, a graph of pressure drop vs. flow rate was generated, and is presented in Fig.
10.

Fig.10. Pressure drop of double layers versus flow rate.

PT
Comparing Figs. 9 and 10, without the effect of length, the greater curvature ratio leads to a

RI
higher pressure drop; however, by adding the effect of length, the outermost layers, with lower
curvature ratios, showed higher pressure drop results. That is, by analyzing the pressure drop at
the same flow rate, the increasing order of pressure drop between the double layers is reversed

SC
from Fig. 9 to 10, which shows greater contribution of the length than that of the curvature ratio
in the final pressure drop result.
The outermost layers have longer length compared to the innermost ones because as the

U
radius at which the tube will be curved is greater (curvature radius), a larger tube length is
required to maintain the same number of tube turns in each layer.
AN
3.6. Influence of Curvature in Pressure Drop: Turbulent Regime
M

In the research presented in this paper, the experimental evaluation of pressure drop in a
straight tube was not made. However, in order to compare the pressure drop between the CT unit
D

and a straight section with the same pipe diameter and length dimensions, the Blasius equation
(eq. (16)) – a well established friction factor equation for Newtonian fluid flowing through a
TE

straight tube in turbulent regime – was used. Such a comparison of pressure drop is shown in Fig.
11.
EP

Fig.11. Experimental CT unit pressure drop and ST pressure drop calculated with eq. (16).
C

In Fig. 11 it can be seen that the experimental pressure drop of the CT unit is greater than
AC

the calculated one for a straight tube of the same dimensions. The absolute percentage deviation
of the entire flow rate range was 6.2%. The difference between the pressure drop data is more
remarkable with the increase of flow rate, or Reynolds. Azouz et al. (1998) stated that the higher
the Reynolds number, the greater the difference between the friction factor of the systems. A
higher pressure drop in coiled tubing compared to straight tubing was also observed by Zhou and
Shah (2004a) with non-Newtonian fluids. The authors justified this result with the effect of
secondary flow in coiled tubing, thus increasing the friction and, consequently, the pressure drop.
ACCEPTED MANUSCRIPT

4. Conclusions

The water flow was studied in a laminar and turbulent regime (Reynolds from 1.6x103 to
2.5x104) in a coiled tubing experimental unit. Regarding the evaluations made in the study of
pressure drop, the following stand out:

• the experimental critical Reynolds number was in the range of 6394 to 6690, depending

PT
on the curvature ratio. The higher values of critical Reynolds number compared to straight
tubes is a result of the secondary flow, which stabilizes and prolongs the laminar regime;
• a critical Reynolds correlation for coiled tubing was proposed from parameter estimation

RI
using experimental data since the tested correlations found in the literature did not present
a good fit;

SC
• friction factor correlations were analyzed in the laminar and turbulent regime. From the
equations, we suggest the use of the Adler (1934) and Ito (1959) equations for laminar
and turbulent regimes, respectively, since they presented a percentage error between the

U
experimental and calculated values lower than 9%;
• the proposed methodology for calculating pressure drop for a coiled tubing system with
AN
different layers, and, consequently, with different curvature ratios, was efficient, with a
percentage error between experimental and calculated values of pressure drop of less than
5% using the Ito equation (1959).
M

• the influence of the curvature was significant against a flow in straight tube, generating a
larger pressure drop around 6.2%. The curvature ratios, however, did not show much
difference in pressure drop since the values of curvature ratio are close. However, the
D

length of each double layer showed a difference in the comparison of the pressure drop
between the double layers, with the outer layers being the ones with the greatest pressure
TE

drop, because they had a longer tube length.


• this paper presented a double layer evaluation of pressure drop of a same CT system
which is something new in the literature and that enabled to check the influence of
EP

different layers in the whole system.


C

Nomenclature
AC

CT coiled tubing

Er absolute percentage error

ST straight tube

diameter of the tube, m

Dean number, dimensionless

Fanning friction factor, dimensionless


ACCEPTED MANUSCRIPT

friction factor for coil, dimensionless

friction factor for straight tube, dimensionless

H gravitational acceleration, m/s2

% length of tube, m

PT
% reel width, m

%k total tube length, m

RI
% length of tube in the nth layer, m

SC
number of the layer

pressure, Pa

U
flow rate, m3/s
AN
critical flow rate, m3/s

inner radius of tube, m


M

`ab outer radius of tube, m

⁄ curvature ratio, dimensionless


D

curvature radius, m
TE

reel radius, m

Reynolds number, dimensionless


EP

critical Reynolds number, dimensionless

&, 〈&〉 average velocity, m/s


C
AC

Greek Letters

∆ pressure drop, 1 bar = 105 Pa

$ fluid density, kg/m3


ACCEPTED MANUSCRIPT

Acknowledgements

The authors would like to thank CENPES (PETROBRAS Research Center) and Coordination for
the Improvement of Higher Education Personnel (CAPES) for the financial support offered for
this project development and the scientific support from members of PPGEQ/UFRRJ.

References

PT
Adler, M., 1934. Striimung in gekriimmten rohren, 2.Angew. Math. Mech. 14, 257-275.

RI
Ali, S., 2001. Pressure drop correlations for flow through regular helical coil tubes. Fluid
Dynamics Research. 28, 295-310.

SC
Azouz, I., Shah, S.N., Vinod, P.S., Lord, D.L., 1998. Experimental investigation of frictional
pressure losses in coiled tubing. SPE Production & Facilities. 91-96.

U
Barua, S.N., 1963. On Secondary Flow in Stationary Curved Pipes. The Quarterly Journal of
Mechanics and Applied Mathematics. 6, 61-77.
AN
Cioncolini, A., Santini, A., 2006. An experimental investigation regarding the laminar to
turbulent flow transition in helically coiled pipes. Experimental Thermal and Fluid Science. 30,
M

367-380. doi: 10.1016/j.expthermflusci.2005.08.005.

Dean, W.R., 1927. Note on the motion of Fluid in a Curved Pipe. Philosophical Magazine. 20,
D

208-223.
TE

Dean, W.R., 1928. The stream-line motion of fluid in a Curved Pipe. Philosophical Magazine and
Journal of Science. 5, 30, 673-695.
EP

Dennis, S.C.R., 1980. Calculation of the Steady Flow Through a Curved Tube Using a New
Finite-Difference Scheme. Journal of Fluid Mechanics. 99, 449-467.
C

Fox, R.W., McDonald, A.T., Pritchard, P.J., 2006. Introduction to fluid mechanics, sixth ed. John
AC

Wiley & Sons, Inc.

Ghobadi, M., Muzichka, Y.S., 2015. A review of heat transfer and pressure drop correlations for
laminar flow in curved circular ducts. Heat Transfer Engineering. doi:
10.1080/01457632.2015.1089735.

Guan, F., Ma, W., Tu, Y., Zhou, C., Feng, D., Zhou, B., 2014. An Experimental Study of Flow
Behavior of Coiled Tubing Drilling System. Hindawi Publishing Corporation. ID 935159.
ACCEPTED MANUSCRIPT

Hasan, W.K., 2013. Transient three-dimensional numerical analysis of forced convection flow
and heat transfer in a curved pipe. IOSR Journal of Mechanical and Civil Engineering. 9, 5, 47-
57.

Ito, H., 1959. Friction Factors for Turbulent Flow in Curved Pipes. Journal of Basic Engineering.
123-134.

PT
Ito, H., 1969. Laminar flow in curved pipes. ZAMM. 49, 11, 653-663.

RI
Jain, S., Singhal, N., Shah, S.N., 2004. Effect of Coiled Tubing Curvature on Friction Pressure
Loss of Newtonian and Non-Newtonian Fluids – Experimental and Simulation Study. SPE,
Houston, Texas.

SC
Kubair, V., Varrier, C.B.S., 1961/2. Pressure Drop for Liquid Flow in Helical Coils. Transaction
of Indian Institute of Chemical Engineering. 14, 93-97.

U
McCann, R.C., Islas, C.G., 1996. Frictional Pressure Loss During Turbulent Flow in Coiled
AN
Tubing. SPE/ICoTA, Texas. SPE 36345.

Medjani, B., Shah, S.N., 2000. A new approach for predicting frictional pressure losses of non-
M

newtonian fluids in coiled tubing. SPE, Denver, Colorado. SPE 60319.

Mishra, P., Gupta, S.N., 1979. Momentum Transfer in Curved Pipes. I. Newtonian Fluids, II.
D

Non-Newtonian Fluids. Industrial and Engineering Chemistry Process Design and Development.
18, 130-42.
TE

Pawar, S.S., Sunnapwar, V.K., Tagalpallewar, A.R., 2015. Development of experimental heat
transfer correlations using Newtonian fluids in helical coils. Heat Mass Transfer. doi:
EP

10.1007/00231-015-1544-0.

Rao, B., 1999. Coiled tubing hydraulics modeling. CTES, L.C., Tech Note.
C

Rao, B.N., 2002. Friction factors for turbulent flow on non-newtonian fluids in coiled tubing.
AC

SPE, Houston, Texas. SPE 74847.

Schmidt, D.F., 1967. Warmeubarang and Druckverlust in Rohrshlangen. Chemical Engineering


Technology. 13, 781-789.

Shah, S., Zhou, Y., Bailey, M., Hernandez, J., 2009. Correlations to predict frictional pressure
loss of hydraulic-fracturing slurry n coiled tubing. SPE Production & Operations.381- 395.
August
ACCEPTED MANUSCRIPT

Shaqlaih, A.S., Kamel, A.H., 2013. AIC applications in coiled tubing hydraulics. International
Journal of Petroleum and Geoscience Engineering. 1, 2, 62-81.

Srinivasan, P.S., Nandapurkar, S.S., Holland, F.A., 1968. Pressure Drop and Heat Transfer in
Coils. Chemical Engineering Journal. 218, CE113-CE119.

PT
Srinivasan, P.S., Nandapurkar, S.S., Holland, F.A., 1970. Friction Factors for Coils. Transactions
of the Institution of Chemical Engineers. 48, T156-T161.

RI
White, C.M., 1929. Streamline flow through curved pipes. Proc. R. Soc. Lond. A. 123, 792, 645-
663.

SC
White, C.M., 1932. Fluid friction and its relation to heat transfer. Trans. Inst. Chem. Eng.,
(London). 10, 66–86.

U
Zhou, Y., Shah, S.N., 2004a. Rheological properties and frictional pressure loss of drilling,
AN
completion, and stimulation fluids in coiled tubing. Journal of Fluids Engineering. 126, 153-161.
doi: 10.1115/1.1669033.
M

Zhou, Y., Shah, S.N., 2004b. Fluid flow in coiled tubing: a literature review and experimental
investigation. Journal of Canadian Petroleum Technology. 43, 6, 52-61.
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Experimentally developed critical Reynolds number correlations for curved pipe
Newtonian fluids flow.
Authors Correlations Conditions
.
= 20,000
15 < < 860
(1)
Ito (1959)

PT
,
=
( ⁄ ) .
10 < < 2,000
Kubair and (2)

RI
Varrier (1961/2)

!."
= 2,300 1 + ( # (3)

SC
< 200
⁄ ) .
Schmidt (1967)

(
U
)%&', %() = 2,100 *1 + 1 < 200
Srinivasan et al.
.0 (4)
(1970) +,-, +./

AN
,2
=
( ⁄ ) . 3
30 < < 110
Cioncolini and (5)
M

Santini (2006)

Source: Zhou and Shah (2004b); Kubair and Varrier (1961/2); Schmidt (1967); Ghobadi and
D

Muzychka (2015); Cioncolini and Santini (2006).


TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Friction factor correlations for laminar Newtonian fluid flow.


Authors Correlations Conditions

.
= 0.1064 > 10
Adler (1934)
(1)

PT
.
Barua (1963) = 0.509 + 0.0918
10 < < 10

RI
(2)

.
> 300
Srinivasan et al. = 0.1125

SC
7< < 104
(1968)
(3)

U
.
= 0.388 + 0.1015
Dennis (1980) Alto Dean
AN
(4)
Source: Zhou and Shah (2004b); Barua (1963); Ghobadi and Muzychka (2015); Ito (1969);
Dennis (1980).
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Friction factor correlations of turbulent Newtonian fluid flow.


Authors Correlations Conditions
.
= 0.08 .
+ 0.012 1,500 < < 10
White (1932)
(1)

PT
.
1 .
= 0.029 + 0.304 0.034 < < 300
Ito (1959) 4

RI
(2)

.
0.084

SC
Srinivasan et al. = . < < 14,000
(1970)
(3)

U
. 0.079
= + 0.0075 ! =
! . 4,500 < < 10
Mishra and
AN
Gupta (1979)
(4) (5)
Source: Ali (2001); adapted from Zhou and Shah (2004b).
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Absolute percentage error between experimental and calculated


critical Reynolds number.
Kubair; Srinivasan Cioncolini;
Ito Schmidt
Varrier et al. Santini eq. (24)
(1959) (1967)
Rec (1961/2) (1968/70) (2006)
Layer r/R
exp.
Rec Er. Rec Er. Rec Er. Rec Er. Rec Er. Rec Er.

PT
dim. % dim. % dim. % dim. % dim. % dim. %
0.0177 -
1st/2nd 6690 5467 18 3480 48 7707 15 5421 19 3558 47 6686 0,1

RI
0.0170
0.0164 -
3rd/4th 6488 5335 18 3396 48 7576 17 5297 18 3475 46 6582 1,5
0.0158

SC
0.0152 -
5th/6th 6444 5215 19 3320 48 7458 16 5185 20 3399 47 6488 0,7
0.0147
0.0143 -
7th /8th 6394 5106 20 3250 49 7350 15 5085 20 3330 48 6402 0,1
0.0138

U
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Absolute percentage error between experimental and calculated friction factor
in laminar regime (flow rate: 0.05 – 0.15 m3/h).
Barua Dennis Srinivasan et al. Adler
(1963) (1980) (1968) (1934)
Layer
Er. Er. Er. Er.
% % % %

PT
1st/2nd 11.3 14.6 5.9 3.7
rd th
3 /4 16.0 19.2 9.8 3.9
th th

RI
5 /6 19.7 22.8 12.7 6.6
7th /8th 11.3 14.2 6.0 3.8

U SC
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Absolute percentage error between experimental and calculated friction factor
in turbulent regime (flow rate: 0.20 – 0.65 m3/h).
White Ito Mishra; Gupta Srinivasan et al.
(1932) (1959) (1979) (1970)
Layer
Er. Er. Er. Er.
% % % %

PT
1st/2nd 13.5 1.6 4.7 5.7
rd th
3 /4 14.5 2.5 5.8 6.4
th th

RI
5 /6 14.4 2.7 5.9 6.3
7th /8th 16.0 4.1 7.6 7.6

U SC
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

Table 1. Absolute percentage error between experimental and calculated pressure drop.

Ito Mishra; Gupta Srinivasan et al.


White (1932)
Q ∆PExperimental (1959) (1979) (1970)

∆P Er. ∆P Er. ∆P Er. ∆P Er.


m3/h bar bar % bar % bar % bar %

PT
0.20 2.0 ± 0.34 2.2 7.5 2.0 3.4 2.1 0.4 2.0 0.9
0.25 2.9 ± 0.43 3.3 12.8 2.9 1.1 3.1 5.0 3.0 4.2

RI
0.31 4.2 ± 0.71 4.7 13.6 4.2 1.6 4.4 5.5 4.4 5.2
0.35 5.3 ± 0.65 6.1 14.3 5.4 2.0 5.6 5.9 5.6 5.9

SC
0.41 7.0 ± 0.50 8.1 14.8 7.2 2.2 7.5 6.2 7.5 6.5
0.46 8.5 ± 0.20 9.8 15.1 8.7 2.3 9.0 6.3 9.1 6.9
0.50 9.8 ± 0.41 11.5 16.7 10.2 3.6 10.6 7.7 10.7 8.5
0.55 11.5 ± 0.67 13.5 17.2
U 11.9 3.9 12.4 7.9 12.5 8.9
AN
0.61 13.8 ± 1.28 16.2 17.7 14.4 4.1 14.9 8.2 15.1 9.4
0.65 15.5 ± 0.61 18.2 17.1 16.1 3.5 16.7 7.6 16.9 8.9
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
ACCEPTED MANUSCRIPT

Highlights

1. Experimental analysis of pressure drop in the flow of water in coiled tubing.


2. Evaluation of friction factor correlations in laminar and turbulent regimes.
3. Experimentally determination of critical Reynolds values.

PT
4. Evaluation of critical Reynolds number correlations.
5. Proposal of pressure drop planning for Newtonian fluid flow through coiled tubing

RI
system.

U SC
AN
M
D
TE
C EP
AC

You might also like