Sunteți pe pagina 1din 21

IADC/SPE 163545

Surge and Swab Effects Due to Vessel Heave in Deepwater Wells: Model
Development and Benchmarking
Sharat Chandrasekhar, SPE, Blade Energy Partners, William Bacon, SPE, Blade Energy Partners, Benni Toldo,
SPE, Woodside, P. V. Suryanarayana, SPE, Blade Energy Partners

Copyright 2013, IADC/SPE Drilling Conference and Exhibition

This paper was prepared for presentation at the 2013 IADC/SPE Drilling Conference and Exhibition held in Amsterdam, Netherlands, 5-7 March 2013.

This paper was selected for presentation by an IADC/SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not
been reviewed by the International Association of Drilling Contractors or the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily
reflect any position of the International Association of Drilling Contractors or the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any
part of this paper without the written consent of the International Association of Drilling Contractors or the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is
restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of IADC/SPE copyright.

Abstract
During operations in deepwater wells (such as, for instance, running smart completions), there exist periods when the axial
movement of the string in the well remains uncompensated. During these periods, vessel heave is imposed on the string at the
surface, resulting in transient pressure fluctuations (swab and surge effects) in the fluid downhole. The problem is further
complicated by the presence of multiple flow ports in the string, and of choke and kill lines at surface, as well as float valves
in the string itself. As the margin between pore and fracture pressure is narrow, these fluctuations can create conditions for an
influx or lost circulation, with potentially significant consequences. While existing literature and commercial models address
the classical drill pipe induced swab/surge due to tripping, they do not model multiple flow paths and/or continuous forcing
functions (such as heave) at the surface of the drill string or completion string.
In this paper, we present the development of a semi-analytical approach to model such transient pressure problems. The
transient pressure problem is solved using the method of characteristics, and draws from the seminal work of Lubinski et al.
and employs the electrical analogy first proposed by Bergeron (Water hammer in Hydraulics and Wave Surges in
Electricity", John Wiley, 1961). The method is extended to include Power-Law fluids (as well as Bingham Plastic and
Newtonian rheological models). Arbitrary flow ports can be situated anywhere in the string, with either one or two-way flow.
Both open- and closed-end pipes are considered, in addition to nozzles at the pipe end. Temperature and pressure effects on
the compressibility, as well as the elasticity of the formation, are considered in the calculation of the characteristic
impedance. The approach can accept any arbitrary forcing function at surface, either as a periodic wave or as a tabulated time
function of displacements. Several limitations of the previous models are also addressed in the theoretical development
reported in this work.
The method is implemented in a spreadsheet tool that allows the input of a wide variety of situations, incorporates different
fluid PVT and rheology models, and calculates transient pressure at any point of interest in the annulus. The model is
compared to several benchmark field cases in the literature, and the comparison is shown to be very satisfactory. Finally, a
case of a smart completion with one to four flow ports in the lower sections that can be selectively closed or opened,
subjected to sinusoidal vessel heave at surface, is examined. Results show that the pressure fluctuations can be substantial in
some of the cases, and suggest that with appropriate fluid selection and operational procedure, even large heave can be
sustained without initiating either underbalance or lost circulation.

Introduction
The problem of downhole pressure surges while running a pipe into wellbore and the potential for the accompanying
formation damage has been the object of a considerable amount of attention. Likewise, the problem of swab kicks induced by
the depleted downhole pressure created by pulling the pipe out of a wellbore is also a well known and frequently encountered
problem in drilling operations worldwide. Furthermore, these pressure surges and swabs are propagated throughout the
wellbore through the fluid medium that transmits the pressure pulses as acoustic waves.
As a result, the response of pressure surges and swabs to pipe movement has been documented in several studies over the
past couple of decades. While the problem has been investigated since the 1930s, almost all of the studies leading up to the
2 IADC/SPE 163545

more recent ones of Burkhardt (Burkhardt 1961) and Fontenot and Clark (Fontenot 1974). considered steady state pressure
surges and swabs in response to constant pipe motion. Perhaps the first detailed and systematic approach to modeling
transient pressure surges is that of Lubinski et al. (Lubinski 1977) based on a graphical method due to Bergeron (Bergeron
1961). More recently, Manohar Lal (Lal 1983) used the method of characteristics which is closely relate to Bergeron’s
graphical technique to determine safe tripping speeds based on predicted pressure surges while running pipe. More recently,
Mitchell (Mitchell 1988) studied transient pressure surges and swabs including the effects of pipe elasticity.
Steady state models prove inadequate in a number of scenarios because they do not correctly model the underlying physics
involved. As a consequence, they do not predict the overshoots and undershoots that are typical of surge/swab phenomena.
This paper revisits the classical Lubinski et al./Bergeron approach with the goal of leveraging the simplicity of their
formulation in the context of complex wellbore problems involving transient pressure surges and swab effects.

Problem Description
A long wellbore string consisting of an arbitrary number of tubular crossovers is run into a wellbore consisting of an outer
casing and (possibly) an open hole, as depicted in Figure 1. The string could be either a drill-pipe or a casing run into the
wellbore and is subjected to an arbitrarily specified time-dependent vertical motion. The moving pipe could be either open or
closed at the end and have an arbitrary number of flow ports resulting in communication between the flowing streams in the
moving pipe and the adjacent annulus. Downward and upward motions of the moving pipe results in a phenomena referred to
as pressure surges and pressure swabs, respectively. These surges and swabs subsequently result in pressure and velocity
pulses that traverse the length of the wellbore at the acoustic speed of the media through which the waves propagate. The
objective of the present analysis is to determine the transient velocity and pressure profiles of the fluid medium in the annulus
of the moving pipe and in the case of an open ended pipe, the profiles inside the tubing.

Modeling Assumptions
The two fundamental assumptions underlying analysis in this paper are
1. The movement of the pipe does not alter the geometry of the computational domain – i.e., the distance traversed by
the moving pipe is much smaller than the wellbore depth.
2. The elasticity of the pipe is not considered in this study. Therefore a forcing function in terms of pipe movement
effected at the surface is instantaneously translated to all locations in the pipe.

Mathematical Formulation
The transient pressure and velocity profiles in the (moving) pipe and pipe annulus can be formally described by the
equations of conservation of mass and momentum. This results in a set of hyperbolic equations that need to be solved for
pressure and velocity in space and time. In Bergeron’s approach, the need to solve the PDEs is obviated since the observer
moves with the wave along a characteristic line at the speed of propagation of the wave. The present analysis will adopt the
approach of Lubinski et al., based on Bergeron’s approach and extend it to the special cases with flow ports. In this approach,
the flow conduit is divided into a number of spatial steps, taking into account geometry changes. At the boundaries between
spatial steps, there are coincident nodes designated b (behind, upstream) and a (ahead, downstream) relative to the direction
of fluid flow (and opposite to the direction of pipe motion), as indicated in Figure 2 which shows three locations A, B and C
and the corresponding four associated nodal locations Aa, Bb, Ba and Cb. Following the notation of Lubinski et al., the
equations of propagation between Aa and Bb and Ba and Cb can be written as


Pt , Bb  Pt 1, Aa  S AB Qt , Bb  Qt 1, Aa  (1)
and


Pt , Ba  Pt 1,Cb   S BC Qt , Ba  Qt 1,Cb  (2)

where P and Q denote pressure and volumetric flowrate respectively and the subscript t-1 denotes the previous time step (not
necessarily one unit behind) and SAB and SBC are the characteristic impedances between segments A and B and B and C
respectively. The impedance is defined in terms of the conduit geometry and fluid properties as

a 1 
S 
gA A g B  C  (3)

where
 : Fluid density
IADC/SPE 163545 3

g
a :Velocity of propagation of the acoustic wave (celerity)
 B  C 
A : Conduit Cross-Sectional Area
g : Acceleration due to gravity
B : Conduit Expansibility
C : Fluid Compressibility
Lubinski et al. assumed a fixed value of compressibility of 3 × 10-6 in their analysis. In the present study, the compressibility
is estimated from the fluid PVT behaviour, based on the quadratic model of Zamora et al. (M. B. Zamora 2000). Expressions
for the conduit expansibility B are presented in the appendix.
The entities at the locations A and C are known since they are evaluated at the previous (t-1) timestep. Therefore, there are
then, four unknowns Qt ,Ba , Qt ,Bb , Pt ,Ba and Pt ,Bb in terms of the nodal pressures and flowrates at the nodal locations
ahead and behind the interface corresponding to location B, where there is also a possible change in section geometry. Two
additional equations are therefore required in order to achieve closure. The first of these comes from mass conservation,
where in response to the aforementioned change in geometry and the pipe velocity, we have
Qt , B a  Qt , Bb  Vt AB
(4)

where Vt and AB are the pipe velocity and change in pipe cross-sectional area at the location B. The last equation arises
from the consideration of the pressure drop between A and C, which may be represented in terms of a pressure drop between
Bb and Ba as
Pt , B a  Pt , Bb  P
(5)
where P is one half of the sum of the individual pressure drops between Aa and B, and Ba and Cb. The pressure drop
P can be expressed in terms of the flowrates Qt ,Ba and Qt ,Bb according to some known functional form

 
P  fb Qt , Bb  f a Qt , Ba   (6)

so that we now observe that all the equations above, effectively constitute a set of four equations in the four unknowns Qt ,Ba ,
Qt ,Bb , Pt ,Ba and Pt ,Bb . The equations can be combined to yield one equation to be solved for the unknown flowrate at
location Ba as

      
g Qt , Ba  f b Qt , Ba  Vt AB  f a Qt , Ba  Pt 1, C b  Pt 1, Aa 
    
(7)
 S AB Qt , Ba  Vt AB  Qt 1, Aa  S BC Qt , Ba  Qt 1, C b  0
using a Newton-Raphson iterative technique, with the initial guess corresponding to the known flowrate at Ba at the previous
timestep (t-1). An iterative technique is necessary if the pressure-flowrate relationship is governed by a Power-Law model,
wherein Eq (7) becomes a transcendental equation. If the flow is modeled as a Bingham plastic, then the pressure-flowrate
relationship is linear and so is Eq (7) which can then be solved without recourse to iterations. Details of the flowrate/
pressure drop relationships is relegated to the appendix. For a more rigorous appraisal of Non-Newtonians models, the reader
is referred to Zamora and Lord (M. L. Zamora 1974) and Savins (Savins 1958). In any event, Once Qt ,Ba is known, the
remaining unknowns can be calculated by elementary back-substitution.

Flow with an Open-Ended Moving Pipe


If the end of the moving pipe is fixed, then there is no flow inside the pipe and the above equations hold for the pipe annulus.
If the pipe is open-ended however, then we have two flowing streams, one inside the pipe and one in the annulus. These
streams are coupled at the open end of the pipe, leading to a special set of interface conditions. We will first establish a
corresponding set of governing equations for flow inside the pipe in terms of pressures  and flowrates  inside the pipe, as
4 IADC/SPE 163545

 t ,Bb   t 1, Aa   AB  t ,Bb   t 1, Aa   (8)


 t ,Ba   t 1,Cb   BC  t ,Ba   t 1,Cb   (9)
 t ,Ba   t ,Bb  Vt AB
(10)

 t ,Ba   t ,Bb  
(11)
where the  terms are now the characteristic impedances corresponding to conditions inside the pipe. At the open end of the
pipe, the two streams communicate with each other as indicated in Figure 3. This constitutes a flow junction that requires
special attention. The conservation of mass condition is now expressed as

Qt , Bb  Qt , Ba   t , Ba  Vt ACS
(12)
Across the junction at B, the pressure drops in the pipe and pipe annulus branches are


Pt , Ba  Pt , Bb  P  f b(ann) Qt , Bb  f a(ann) Qt , Ba    (13)

 t , Ba  Pt , Bb    f b
(ann)
Q   ft , Bb a
(pipe)
  t , Ba
(14)
We now have two additional unknowns in terms of the pressure and flowrate inside the pipe at the location Ba at the pipe-
annulus junction to consider. Equations (1) through (13) can be combined to form the simultaneous system of (possibly non-
Q and  t , Ba
linear) equations in the two fundamental unknowns t , Ba according to

 Qt , B ,  t , B   Pt , B  Pt , B  f b(ann) Qt , B   f a(ann) Qt , B   0


a a a b b a

 Qt , B ,  t , B    t , B  Pt , B  f
a a a b b
(ann)
Q   f
t , Bb a
(pipe)
   0
t , Ba
(15)

which are to be solved in conjunction with the relations



Pt , Ba  Pt 1, C b  S BC Qt , Ba  Qt 1,C b 

 t , Ba   t 1, C b   BC  t , Ba   t 1, C b 

Pt , Bb  Pt 1, Aa  S AB Qt , Bb  Qt 1, Aa  (16)

Qt , Bb  Qt , Ba   t , Ba  Vt ACS
The system of equations can be solved using the two-dimensional analogue of the Newton-Raphson iterative method as

 Qt , Ba 
(k )
 Qt , Ba 
( k 1)

  Qt , Ba ,  t , Ba   ( k 1)

  J 1 
 
 t , Ba 

 t , B a  
  Qt , Ba ,  t , Ba   (17)

where (k) and (k-1) denote iteration levels and the Jacobian J is evaluated according to

   
 Q  t , Ba 
J   t , Ba 
    (18)
 Qt , B  t , Ba 
 a 
Finite difference approximations to the partial derivatives will suffice, since the N-R technique requires only an
approximation to the Jacobian. Rapid convergence – typically of the order of two to three iterations is observed with suitable
initial guesses for the unknown flowrates.
IADC/SPE 163545 5

Case with Flow Ports in the Moving Pipe.


Consider now, the scenario (see Figure 4) in which a flow port exists at some location in the moving pipe. The flowing
streams in the pipe and annulus are now in lateral communication and if the flow port is modeled as an orifice plate, then the
fluid exchange between the two streams is given by the pair of equations


Qt , Ba  Qt , Bb  K D sgn Pt , Bb   t , Bb  P t , Bb   t , Bb 
 t , B a   t , Bb  K D sgn P t , Bb   t , Bb  P t , Bb   t , Bb  (19)

in which the discharge coefficient is given by

288g c
K D  CAp
 (20)
where the constant C = 0.7 based on the classical orifice plate theory, Ap is the flow port area and  the fluid density. The
equations of propagation in the pipe and annulus are given by the following systems of equations respectively:

Pt , Bb  Pt 1, Aa   S AB Qt , Bb  Qt 1, Aa 
Pt , Ba  Pt 1, C b   S BC Q t , Ba  Qt 1, C b  (21)

and

 t , Bb   t 1, Aa   AB  t , Bb   t 1, Aa 
 t , Ba   t 1, C b    BC  t , Ba   t 1, C b  (22)

The pressure drops in the pipe and pipe annulus across the location B corresponding to the flow port are given by

 
PPt , Ba  Pt , Bb  P  f b(ann) Qt , Bb  f a(ann) Qt , Ba   (23)
and

 
 t , Ba   t , Bb    fb(pipe)  t , Bb  f a(pipe)  t , Ba   (24)

Equations (19) through (24) may be combined to form the simultaneous system of equations

 Qt , B ,  t , B   Pt , B  Pt , B  f b(ann) Qt , B   f a(ann) Qt , B   0


a a a b b a

 Qt , B ,  t , B    t , B   t , B  f
a a a b b
(pipe)
   f
t , Bb a
(pipe)
   0
t , Ba
(25)

which can be solved using the same Newton-Raphson iterative method described in Eqs. (17) and (18).

Time and Spatial Step Size Determination.


Since the observer moves with the wave in the present approach, the spatial step must be chosen such that in each time step,
the distance travelled by the pressure pulse is equal to the product of the local acoustic speed and the timestep. In order to
achieve this throughout the wellbore, it is necessary to adjust the celerity in Eq. (3). If the moving pipe is composed of N
sections of length Li, i=1,2…N and the celerities in these segments are ai, i=1,2…N, then the timestep is given by

min  Li 
t     imin (26)
i  N min ai 
where Nmin is the minimum number of segments in each section. We next define the nominal step size corresponding to the
segment imin, such that
6 IADC/SPE 163545

Limin
L*  (27)
N min
The number of segments and step size in each section are based on the nominal step size as

 L  L
N j  int  j   1  L j  j (28)
 L*  Nj
Finally, in order to satisfy the criteria that the pulses must travel the exact distance each node to the next in every time step, it
is necessary to adjust the celerity following the recommendation of Lubinski et al., according to

L j
aj  (29)
t

Pressure and Flowrate Boundary Conditions.


The computational domain extends from the surface to the bottom of the wellbore. Therefore the pipe annulus encompasses
the entire computational domain, whereas the moving pipe itself is offset from the base of the wellbore by an amount well in
excess of the distance travelled by the pipe over the time period of interest. At the surface, the pressure is zero. Accordingly,
we have

Pt , z0   t , z0  0 (30)

The base of the wellbore is regarded as impenetrable. Therefore the downhole boundary condition is simply

Qt , z0  0 (31)

Results and Discussion


Three surge/swab problems are chosen to demonstrate the performance of a spreadsheet that was created based on the
methodology outlined in this paper. The first two are problems from the public literature used to benchmark the code. The
third problem involves an actual wellbore configuration with the moving pipe subjected to a sinusoidal vessel heave.
Table 1 shows the data corresponding to the benchmarking test case designated as Problem 1. This corresponds to Example 2
in the paper by Manohar Lal (Lal 1983). A 7-5/8” casing string is tripped into a wellbore consisting of a 14 ppg fluid in a 10-
3/4” X 9-7/8” casing down to a depth of 17,000 ft followed by a 1000 ft 9-7/8” open hole section. The pipe is accelerated to 3
ft/sec from rest in 3 seconds and held at this speed for 12 seconds following which it is deccelerated to rest in a 3 second
period. This pipe movement schedule is indicated by the red curve in Figure 5.
Also shown in 5 are the he pressure histories at the base of the wellbore for the scenarios when the pipe end is closed and
open. The results of the latter case are seen to be in excellent agreement with those of Manohar Lal. The slight deviations
towards the end of the pressure decay are most probably due to slightly different implementations of the power law model
used to model the fluid flow in the conduit. As expected, the significantly lower surge and swab pressures are observed when
the end of the pipe is opened.
The transient pressure profile snapshots in each of the panels of Figure 6 show the effect of the interaction between the
flossing streams inside and outside the pipe through the junction at the end of the pipe. The discontinuities sin the annulus
pressure profile at 3000 ft correspond to the crossover from 10-3/4” to 9-7/8” in the outer casing. The movement of the
pressure fronts in both streams are not as pronounced as would be expected owing to their interaction.
The modeling parameters corresponding to Problem 2 are indicated in Table 2. This problem is designated “Well 3a” in
Lubinski et al (Lubinski 1977). It is structurally identical to Manohar Lal’s Problem #1, except that in the pipe movement
schedule, there is no decceleration involved and the end of the pipe is 500 ft closer to the base of the wellbore. In addition,
the effect of lifting the pipe out of the slips is simulated by the upward motion of the pipe for the first two seconds. The fluid
flow was modeled using the Bingham plastic model in the study of Lubinski et al.
Figure 7 shows the pressure histories at the base of the wellbore when the fluid flow is modeled with the Bingham Plastic and
Power Law models. As expected, both histories asymptotically approach the steady state. It is noteworthy, that there is an
accompanying pressure swab of about 100 psi during the period when the pipe is lifted out of the slips. It is difficult to
facilitate a quantitative comparison between these results and those of Lubinski et al., as their results were only presented
graphically in a small panel of a larger figure and moreover, there remains some uncertainty as to the precise way in which
IADC/SPE 163545 7

their fluid flow model was implemented. However, the results appear to be very similar, notwithstanding the 100 psi higher
steady state pressure predicted by our model when the Bingham plastic model is used.
Figure 8 is a schematic depiction of the smart completion in he subsea well corresponding to Problem 3. The tubing is run
into the wellbore with 4 packers that will be activated when the tubing is in position. The issue however has to do with
pressure surges and swabs due to vessel heave transmitted through the workstring in the marine riser. Table 3 shows how the
tubing and wellbore configuration is built into the spreadsheet implementation of the model. The vessel heave is modeled as
a sinusoidal function with an amplitude (half heave) of 1.64 ft.
When the tubing is run with a plug at the base sealing it off from hydraulic communication with the annulus, the transient
history at the base of the open hole indicates surge and swab pressures of the order of 350 psi as seen in Figure 9. These
estimates suggest a moderate degree of overshoot in comparison to steady state model predictions. When the tubing base is
open, a substantial reduction in both surge and swab pressures is observed with a chaotic initial period when the streams
inside and outside the tubing interact in response to the acceleration from rest. However a stable oscillatory pattern quickly
develops. What is noteworthy from Figure 10 is that the pressure profile inside the tubing is more irregular than that in the
annulus despite the fact that the annulus has more area changes between the packers.
Finally a scenario with flow ports is considered. The flow ports are located in the tubing between the packers as indicated in
the configuration of Table 4. In the set of simulations to study the effect of flow ports, the diameter of the ports is varied from
an eighth of an inch to half an inch and the pressure histories of 11 indicate that the pressure surge behaviour is consistent
with expectations in that larger flow ports result in reduced surge pressures. However the swab response to flow port
diameter is intriguing . A partial explanation can be volunteered in terms of fluid being unloaded from inside the pipe during
a swab resulting in reduced downhole pressures. But this phenomenon needs to be explored in greater detail.

Summary and Conclusion


A mathematical model of surge and swab pressure response to pipe movement in a wellbore has been developed by
implementing the methodology of Lubinski et al to include cases with open-ended tubing and flow ports in the pipes.
Arbitrary pipe movement schedules are prescribed and both Bingham Plastic and Power Law rheological models are
deployed. The methodology is validated by means of a benchmark case adopted from an earlier study by Manohar Lal (Lal
1983) wherein excellent agreement is demonstrated. The performance of the model with respect to both open and closed end
moving pipes is demonstrated including the ability to handle flow ports facilitating communication between the flowing
streams inside and outside the pipe. An example of a complex wellbore involving a smart completion with a tubing string
subjected to a prescribed sinusoidal pipe heave is also modeled. The results of the model are consistent with trends from
simpler steady state models. While further study is required to fully understand the complex interactions involved when flow
ports are involved, the model has the demonstrated capacity to be a useful tool for predicting transient surge and swab effects
in complex wellbores.

Acknowledgements
The authors thank Woodside for supporting the work leading to this paper, and for permitting its publication.

References
Bergeron, L. Water Hammer in Hydraulics and Wave Surges in Electricity. John Wiley (Copyright ASME 1961), 1961.
Burkhardt, J.A. "Wellbore Pressure Surges Produced by Pipe Movement." J. Pet. tech, June 1961: 595-605.
Fontenot, J.E. and Clark, R.E. "An Improved Method for Calculating Swab and Surge Pressures and Calculating Pressures in
a Drilling Well." SPE Journal, Oct 1974.
Lal, Manohar. "Surge and Swab Modeling for Dynamic Pressures and Safe Trip Velocities." IADC/SPE, Feb 20-23 1983:
427-433.
Lubinski, A., Hsu, F.H., and Nolte, K.G.,. "Transient Pressure Surges due to Pipe Movement in an Oil Well." (Revue de
l'Institute Franais du Petrole) May-June 1977: pp. 307-47.
Mitchell, R.F. "Dynamic Surge Swab Pressure Predictions." SPE Drilling Engineering, September 1988: 325-333.
Savins, F.J. "Generalized Newtonian (Pseudo-Plastic) Flow in Stationary Pipes and Annuli." Pet. Trans. AIME 213 (1958):
189.
8 IADC/SPE 163545

Zamora, M., Broussard, P.N., and Stephens, M.P. "The Top 10 Mud-Related Concerns in Deepwater Drilling Operations."
SPE Paper No 59019, February 1-3 2000.
Zamora, M., Lord, D.L. "Practical Analysis of Drilling Mud Flow in Pipes and Annuli." SPE Paper No. 4976, Oct 6-9 1974.

Appendix
Conduit Expansibility
Case 1: Pipe in Casing

2  1  2 Ro2  1 Ri2  1  
B  2  Ra 2  2    s 
Es  Ra  1  Ro  1 Ri  1  

Do,OP Do,IP Di,OP


Ro  Ri  Ra 
Di,OP Di,IP Do,IP
Di,IP : ID of Inner Pipe
Do,IP : OD of Inner Pipe
Di,OP : ID of Outer Pipe
Do,OP : OD of Outer Pipe

Es : Steel Elastic Modulus  s : Steel Poisson's Ratio

Case 2: Pipe in Open Hole

2  Ra2 1  Ri2  1 
B 2  1   f    2  s  
Ra  1  Ef Ef  Ri  1 

DOpen Hole
Ra 
Do,IP

Ef : Formation Elastic Modulus  f : Formation Poisson's Ratio

Fluid Flow Models


Average Velocity Definition
Q
V   Vpipe
Acs
where  = 1 if the conduit ID is zero and otherwise, is given by
1 1
 
    1
2 ln DDoi Do 2
Di
IADC/SPE 163545 9

Power Law Model:


P  L V max C laminar V , C turbulent V
n 1.75

Plaminar  C laminar V
n

1.75
Pturbulent  C turbulent V

n
 1 
 2 
K n 
 0.0208 
 
C laminar   
144000D  d 
1 n

3 1
 4 4
C turbulent 
1396D  d 
1.25

Bingham Plastic Model:

P  LAV  B V 


A  D  d 2
1000

y
B D  d 
1000
10 IADC/SPE 163545

Tables

Table 1:  Data for Problem 1. Example 2 of Manohar Lal. 
IADC/SPE 163545 11

Table 2:  Data for Problem 2. Example Well 3a of Lubinski et al. 
12 IADC/SPE 163545

Table 3: Data for Problem 3. Tubing in Smart Completion Subjected to Vessel Heave 
IADC/SPE 163545 13

Table 4:  Example 3 Modified with Flow Ports between Packers. 

Inner Pipe Outer Pipe Flow Control Valves

From To ID OD ID OD Depth Diam .


Description Fluid
(m ) (m ) (in) (in) (in) (in) (m ) (in)

Workstring in Riser 0 130 4.780 5.500 19.750 21.000 Polymer 1 2,925 0.500
Tubing in 13-3/8" 130 1,950 6.184 7.640 12.347 13.375 Polymer 2 2,975 0.500
Tubing in 10-3/4" 1,950 2,850 6.184 7.640 9.560 10.750 Polymer 3 3,050 0.500
Tubing in 9-5/8" 2,850 2,877 6.184 7.640 8.535 9.625 Polymer 4 3,135 0.500
Upper Packer A 2,877 2,880 3.520 8.300 8.535 9.625 Polymer 5
Tubing in 9-5/8" 2,880 2,900 2.992 4.250 8.535 9.625 Polymer 6
Tubing in 7" 2,900 2,947 2.992 4.250 6.184 7.000 Polymer 7
Packer B 2,947 2,950 2.820 5.995 6.184 7.000 Polymer 8
Tubing in 7" 2,950 2,998 2.992 4.250 6.184 7.000 Polymer 9
Packer C 2,998 3,001 2.820 5.995 6.184 7.000 Polymer 10
Tubing in 7" 3,001 3,123 2.992 4.250 6.184 7.000 Polymer 11
Packer D 3,123 3,126 2.820 5.995 6.184 7.000 Polymer 12
Tubing in 7" 3,126 3,145 2.992 4.250 6.184 7.000 Polymer 13
Sump 3,145 3,300 0.000 0.000 6.184 7.000 Polymer 14
14 IADC/SPE 163545

Figures

Figure 1: Pipe subjected to Vertical Motion in a Casing/Open Hole Wellbore (a) Closed End Pipe; (b) Open Ended Pipe; (c) 
Open‐Ended Pipe with Flow Ports. 
IADC/SPE 163545 15

C C
Cb Cb

QBa QBa
Ba Ba
B B
Bb Bb
QBb QBb
Positive direction
of fluid flow

Positive direction
of pipe velocity

Aa Aa
A A
V V V

a) Pipe flow b) Annular flow

Figure 2: Nodal Analysis in Pipe and Pipe Annulus 

Figure 3: Nodal Analysis at the junction of the Pipe Open End and the Pipe Annulus 
16 IADC/SPE 163545

Figure 4: Nodal Analysis at a Flow Port Enabling Communication between the Inner and Outer Streams. 

Figure 5: Transient Pressure History at Well TD – Problem 1, Comparison with Results of Manohar Lal. 

 
IADC/SPE 163545 17

Figure 6: Evolution of Pipe and Annulus Pressure Profiles – Problem 1 (Open Ended Pipe). 
18 IADC/SPE 163545

Figure 7: Transient Pressure History at Well TD – Problem 2 (Example 3a of Lubinski et al.) 

Figure 8: Smart Completion Configuration Corresponding to Problem 3. 
IADC/SPE 163545 19

Figure 9: Transient Pressure History at Well TD – Problem 3, Pipe Subjected to Vessel Heave. 

 
20 IADC/SPE 163545

Figure 10: Evolution of Pipe and Annulus Pressure Profiles – Problem 3 (Open Ended Pipe). 

 
IADC/SPE 163545 21

Figure 11: Transient Pressure Histories with Flow Ports  – Problem 3 (Closed Ended Pipe). 

S-ar putea să vă placă și