See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/264041832

Water Coning

Conference Paper · May 2014

CITATIONS READS
0 4,592

3 authors, including:

Firdavs Aliev
Istanbul Technical University
13 PUBLICATIONS   60 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Polymer flooding View project

Extra-Heavy Oil View project

All content following this page was uploaded by Firdavs Aliev on 19 July 2014.

The user has requested enhancement of the downloaded file.

 i

ABSTRACT

Water coning is a usual problem that is faced by petroleum engineers in reservoirs

having an aquifer, particularly at the bottom. The critical rate is the subject discussed
mostly in the studies on water coning. This paper presents a simulation study using
RUBIS, a subprogram of ECRIN. Some correlations for critical rate are analyzed and
their results are compared with those from RUBIS. For reasonable comparison,
parameters in the simulation program are set so that the assumptions used in correlations
could be met.

Furthermore, a sample reservoir model is taken into consideration for optimization of oil
production in the presence of water coning. Parameters to be optimized are well length
and position of the well in the reservoir. Vertical and horizontal wells are both
considered in this study. Two scenarios are considered: constant production rate and
constant bottomhole flowing pressure. The optimum alternative is defined as the one
which maximizes the economic profit. Moreover, the effects of some reservoir and fluid
parameters on critical rate are analyzed.
 ii

LIST OF TABLES

page
Table 1 – Field data for the example problem. ................................................................ 25

Table 2 – Critical rate calculation approaches for vertical well....................................... 26

Table 3 – Critical rate calculation approaches for horizontal well .................................. 27

Table 4 – Reservoir and fluid parameters ........................................................................ 32

Table 5 – Data from RUBIS for optimization process at a constant production

rate for the vertical well at 10000 days. .......................................................... 36

Table 6 – Data from RUBIS for optimization process at a constant production

rate for the horizontal well based on the vertical position at 10000 days
(L=4000 ft). ..................................................................................................... 36

Table 7 – Data from RUBIS for optimization process at a constant production

rate for the horizontal well based on the well length at 10000 days
(hL=40 ft). ........................................................................................................ 37

Table 8 – Data from RUBIS for optimization process at a constant bottomhole

flowing pressure for the vertical well at 10000 days. ...................................... 38

Table 9 – Data from RUBIS for optimization process at a constant bottomhole

flowing pressure for the horizontal well based on the vertical position at
10000 days (L=4000 ft). .................................................................................. 39

Table 10 – Data from RUBIS for optimization process at a constant bottomhole

flowing pressure for the horizontal well based on the well length
at 10000 days (hL=40 ft). ............................................................................... 39
 iii

LIST OF FIGURES

page
Figure 1 – Shapes of water coning in vertical and horizontal well. ................................... 1

Figure 2 – Homogeneous formation with upper portion saturated with oil and lower
portion with water............................................................................................. 2

Figure 3 – Partially penetrating well with boundary conditions. ..................................... 13

Figure 4 – Critical rate correlation ................................................................................... 14

Figure 5 – Critical rate correlation with fractional well penetration ................................ 15

Figure 6 – Schematic diagram of a partially penetrating well subjected

to water coning. .............................................................................................. 17

Figure 7 – Bottom water-drive case during production ................................................... 21

Figure 8 – Schematic of coning model, a system study ................................................... 23

Figure 9 – A 3D shape of the oil reservoir with a horizontal well ................................... 30

Figure 10 – A 3D shape of the oil reservoir with a vertical well ..................................... 30

Figure 11 – Relative permeability curves.. ...................................................................... 34

Figure 12 – Completion interval of the vertical well. ...................................................... 34

Figure 13 – Placement of the horizontal well at different vertical positions

(L=4000 ft). ................................................................................................... 34

Figure 14 – Schematic diagram for the optimization of horizontal well length

(hL=40 ft). ...................................................................................................... 35

Figure 15 – Water saturation distribution at qoc = 188 STB / days,

(hL = 40 ft, L = 4000ft). ................................................................................. 40

Figure 16 – Water saturation distribution at qo = 300 STB / days,

(hL = 40 ft, L = 4000 ft). ................................................................................ 40
 iv

Figure 17 – The effect of vertical anisotropy on critical production rate. ....................... 41

Figure 18 – The effect horizontal well length on critical rate. ......................................... 42

Figure 19 – The effect of the density difference on critical rate. ..................................... 43

Figure 20– Effect of recharge constant on critical rate. ................................................... 43

Figure 21 – Effect of vertical position of horizontal well on critical rate. ....................... 44

Figure 22 – Effect of grid number on critical rate. .......................................................... 44

 1

1. INTRODUCTION

Water coning is a term used to describe the upward movement of water into the
perforations of a producing well (Ahmed, 2010). This phenomenon can also be
described as a steady and usually sharp displacement of some or all the oil production
by the bottom water when the critical withdrawal rate from the well is exceeded
(Muscat and Wyckoff, 1935). Water coning may lead to several serious problems. For
example, the water is usually corrosive and its disposal cost is high. The affected well
can be abandoned early. Moreover, there may be loss in total recovery (Ahmed, 2010).

Before the production, petroleum reservoirs have fluid contacts such as water-oil
contact (WOC) and gas-oil contact (GOC). As the production is initiated, these contacts
change in shape and form a cone or a crest. If the field is developed by vertical wells,
the final shape formed is referred to as a cone. In the case of horizontal wells, this
ultimate shape is crest as shown in Figure 1. However, these terms “coning” and
“cresting” may be used interchangeably if quantitative discussion is the case. The first
of these terms is used by most engineers even in the cases of horizontal wells for gas
and water production (Makinde et al., 2011).

Figure 1 – Shapes of water coning in vertical and horizontal well (Url-1).

 2

Critical rate is the maximum production rate which does not allow water to
breakthrough into the production well. When the oil production rate becomes higher
than the critical rate, WOC rises and cone becomes unstable reaching the bottom of the
well (Chierici, 1995). The water cone is said to be stable if the pressure at every point
on the WOC is the same as the reservoir pressure p res (Figure 2).

Figure 2 – Homogeneous formation with upper portion saturated with oil and lower
portion with water (Muscat and Wyckoff, 1935).

Equation 1 is the required condition for water cone to stay in static condition.
p   w gz  pres (1)

where
p : the pressure in the oil immediately above the surface of the water at the point
(r,z).
p res : the reservoir pressure.
w : the specific gravity of water.
g : the acceleration of gravity.
z : the height of the water cone from initial WOC.
 3

This study mainly deals with critical oil production rate and parametric analysis for the
observation of some reservoir and fluid properties on the critical rate in horizontal wells.
Moreover, optimization process for the length and vertical position of the horizontal well
and the completion interval of the vertical well will be discussed. In Chapter 2 the
physics of water coning, its types based on whether it reaches the well or not, and its
mechanisms are discussed. In Chapter 2 studies by some authors on water coning in
horizontal wells and correlations for critical production rate and breakthrough time
calculations are mentioned. Moreover, the differences between horizontal and vertical
wells and advantages of the former over the latter are handled. In addition, one example
problem is solved using RUBIS (Ecrin v4.20, 2013) and compared with the results from
some correlations. Then in Chapter 3, statement of the problem is introduced. In the
problem statement, the objective of the study is explained. Chapter 4 involves the
description of the reservoir model used in RUBIS. Finally, in Chapter 5 simplified
economic analysis for the optimization process and sensitivity analysis results, along
with the effects of some parameters on the critical rate are given.
 4

2. WATER CONING

In order to understand the meaning of water coning, it is important to know the physics
behind this phenomenon occurring in oil fields. Many studies have been dealing with
water coning, especially critical rate and the breakthrough time calculations.
Breakthrough time is the time when water from aquifer reaches the production well.
Most correlations are given in the literature to predict the critical rate and breakthrough
time in both vertical and horizontal wells. Each of these correlations is based on some
assumptions related to reservoir types, reservoir parameters, and flow types. An example
problem is solved so as to see how these assumptions differ and affect the results for
critical rate values.

2.1. PHYSICS OF WATER CONING

One of the primary factors leading to coning is pressure drawdown. There is a

substantial pressure drawdown near the wellbore displayed by a vertical well (Makinde
et al., 2011). Since the flow is radial, the pressure gradient increases drastically around
the well. Another explanation is by Ahmed (2010), who indicates that the primary factor
leading to coning is movement of reservoir fluids towards the zone of least resistance.
Muscat and Wyckoff (1935) point out some significant reasons for coning as well. The
first reason is that the pressure drop between the reservoir boundary and the points
below the bottom of the well is greater than the hydrostatic head of the given water
column. Another reason is related to viscous and gravity forces. The latter is associated
with density difference between the oil and water. When the dynamic, or viscous, forces
exceed the static forces, this brings about coning. The forces that have an effect on
water coning are capillary, viscous, and gravity forces. The first forces usually have
imperceptible influence. The viscous forces, which are associated with the pressure
 5

gradient as described by Darcy’s Law, are related to fluid flow through the reservoir.
Gravity forces act in vertical direction and cause the fluid to rise due to density
difference. At any time there is an equilibrium between viscous and gravity forces.
Once this balance is destroyed, more specifically, when the viscous forces exceed the
gravitational ones, cone will break into the well. However, if the opposite circumstance
is the case, that is, if the gravity forces exceed the viscous forces at steady-state
condition, the water cone will not reach the well. The cone will not move backward
either, and therefore, it is called a stable cone. On the other hand, if the pressure in the
system is in unsteady state, the cone, which is now known as an unstable, will proceed
towards the well until the steady-state condition is reached. If the flowing pressure drop
is adequate to surmount the gravity forces, unstable cone will expand and eventually
break into the well (Ahmed, 2010). The reason for water cone to become unstable is that
upward dynamic force is extremely high and is not possible to balance with the weight
of water below. Stating briefly, the dynamic pressure gradient above the critical point,
which is beneath the wellbore, is greater than the hydrostatic pressure gradient of water.
Thus, the water in the cone above this critical point cannot stay in static condition
anymore and it starts flowing upward to reach another equilibrium. This continues until
the water breaks through into the wellbore (Recham at al., 2000).

2.2. LITERATURE REVIEW

Muscat and Wyckoff (1935) by assuming a homogeneous sand formation with the upper
part saturated with oil and the lower part with water at static condition, determined the
shape of water cones for different pressure drops as a function of well penetration and
oil-zone thickness. Muscat and Wyckoff (1935) deducted that water coning cannot be
eradicated if the oil zone is thin. This can be achieved by reduction of production rate to
significantly low values or decreasing well penetration. They also assumed the flow of
 6

oil in sand between two parallel impermeable boundaries and into the partially
penetrating well.

Smith and Pirson’s (1963) investigation involves development of a method to control

water coning by injection of oil at a point below the producing interval. They showed
that the fluid injection reduces the water-oil ratio, and that this reduction can be
enhanced if the injected fluid has a higher viscosity than the reservoir oil. Existence of
zone of reduced permeability around the injection point also proved to improve the
reduction in water-oil ratio. Impermeable barriers or cement pancakes in their study did
not provide significant benefits. At a particular production rate, the point closest to the
bottom of producing interval not inhibiting the oil production was the most appropriate
point for fluid injection. In the case of higher production rate, the stated point was
situated lower for maximum water coning suppression. Commencement of fluid
injection before the development of the water coning displayed no advantage.

Khan (1970) studied water encroachment in the vertical direction in a three-dimensional

scaled laboratory model. Porous sand pack was used to represent thin oil and water sand
layers. As results indicated, the influence of mobility ratio on water cut and degree of
water coning at a specified production rate was essential. As for the shape of the cone, it
was found that for mobility ratio values less than one, the cones have lower profiles and
larger radial spread. When the mobility ratios are higher, the water cone initially rises
and, afterwards, spreads radially.

Mungan (1979) used layered model in which fluid saturation was recorded as a function
of time and the location of the test bed. The experiments were concerned with the effect
of oil viscosity and production rate on water cone, the effect of heterogeneity on test
bed, and the effect of injecting a polymer slug at the WOC prior to water injection. It
was found that the layered model proved to show lower oil recovery and higher water-
oil ratio. Thus, stratification was detrimental to oil recovery in a coning scenario. Study
results indicated that the water saturations were higher in low permeability strata than in
 7

the higher permeability formation. Imbibition of water into the low permeability slayers
brought about this change in saturation. It was also shown that increase in viscosity and
production rate caused decrease in recovery and escalation in water-oil ratio. Injection
of polymer solution at the WOC suspended development of water cone and contributed
to more efficient oil recovery. Briefly expressing, according to Mungan (1979), in
simulating a non-homogeneous reservoir, stratification is vital.

Kaneko et al. (1973) conducted the numerical simulation study for a light oil reservoir
with bottom water drive. It was found that increase in production lead to decrease in
breakthrough time and increase in water-oil ratio. They also concluded that ultimate
recovery was not affected by the production rate. It was farther shown that the oil
production without water was reduced as the well penetration was increased. Wellbore
radius proved to have negligible effect on water breakthrough time and water-oil ratio.

Blades and Stright (1975) examined water coning phenomenon in an undersaturated

reservoir with the bottom water drive and high viscosity crude oil. They did not ignore
the capillary effects in their simulation for history matching coning behavior in heavy
oil reservoirs that have remarkable oil-water transition zone thickness. They used the
history matching for known field and observed effects of several parameters. For
instance, it was found that increase in horizontal permeability contributes to lower water
cut, which is also affected by oil viscosity. Existence of lower permeable strata impedes
the water coning development thereby reducing the water-oil ration. Fixed parameters
for history matching were horizontal permeability and capillary pressure.

Pirson and Mehta (1967) used the numerical approach to investigate the three methods.
First, they injected the reservoir oil below the production interval and found that water-
oil ratio may be reduced by a factor 4 by oil re-injection. In order for pumping to be
economical, this technique can be enhanced by bottomhole re-injection. The second
measure having probability to affect the water coning that was investigated by Pirson
and Mehta (1967) is selective water and oil production from their respective zones into
 8

a dual-completion string. Finally, they considered placement of a “pancake” of an

impermeable material, which may restrict the upward vertical flow. Their primary
recommendation is re-injection of reservoir oil if no pumping of to the surface is
necessitated. Their opinion about pancake placement is that it does not provide a
promising solution to the water-coning problem and that it can delay water cone merely
up to a particular time.

Recham et al. (2000) used numerical simulation to observe the effects of various
reservoir and fluid parameters on water coning. The results of the experiment indicated
that coning has more serious tendency in heavy oil reservoirs than in light oil reservoirs.
The lower the oil-water mobility ratio means the higher oil recovery and the lower the
coning tendency. Increase in horizontal well length increases both oil recovery and
breakthrough time for water. The horizontal permeability was found to have inverse
relation with coning tendency and direct relation with cumulative oil production.

Menouar and Hakim (1995) used numerical method to investigate water cresting in
horizontal wells. The study is based on the observation of saturation gradient in the
reservoir, which was assumed homogeneous and anisotropic. Sensitivity to parameters
like horizontal well length, anisotropy ratio, mobility ratio position of the horizontal
well etc. was also controlled. They proved that the critical rate is an increasing function
of anisotropy ratio only for the latter being in the interval between 0.5 and 1. For
example, for anisotropy ratio values between 0.01 and 0.1 the production rate is a
strongly decreasing function of the anisotropy ratio. Well length proved to be a crucial
factor in case if well is closer to the water-oil contact. If placed below water-oil contact,
horizontal wells give higher oil recovery; however, this leads to higher water cut.

Wu, Reynolds, and Markitell (1995) considered the Amber Field in the Gulf of Mexico
for their study. The goal of their study was to observe the likelihood of horizontal wells
to delay water coning and enhance oil recovery. They used full-filled model along with
the single-well models. Their results showed that the horizontal wells penetrated in the
 9

reservoirs with gas cap could greatly reduce water coning, increase oil production in a
thin reservoir possessing tolerable in dimensions gas cap.

Jiang and Butler (1998) studied experimentally the movement of bottom water to a
horizontal well using Hele-Shaw cell. They checked the effects of different flow rates
and viscosity ratios. The results indicated that oil recovery is a decreasing function of
flow rate and viscosity ratio. When the viscosity ratio was equal to unity, the interface
was stable. It was also revealed that in some cases where the viscosity ratio was high,
oil recovery proved to be even higher than in case of low flow rates. This resulted from
multiple fingering. As can be seen, significant amount of oil can be recovered from
multiple fingering flow. The results further showed that increase in flow rate and
viscosity ratio makes water-oil interface sharper.

Permadi and Jayadi (2010) proposed a semi-empirical method for prediction of

horizontal well performance after the breakthrough time. Their approach was based on
Darcy’s equation for linear flow and the material balance. The study also established the
correlation for deviation factor to estimate displacement efficiency provided by bottom
water. In the model, it was assumed that all the forces have negligible effect. Partial
completion, changes of fluid properties and wellbore friction losses were also ignored. It
was concluded that higher drawdown leads to lower cresting efficiency. Moreover, the
longer the horizontal well, the higher the cresting efficiency. Increase in oil thickness
causes an increase in cresting efficiency. Furthermore, it was shown that the thicker the
oil column, the lower the drawdown and the longer the breakthrough time. It was also
indicated that reduction in vertical anisotropy ratio brings about slow increase in
cresting efficiency.

Permadi (1996) developed fast horizontal well coning evaluation method. For the
development of the correlation, linear displacement concepts were applied. The
influence of relative permeability on coning was also studied. It is stated in the paper
that when the horizontal well is placed close to the top of the oil zone, oil production
 10

may be more efficient. They assumed a tank model reservoir bounded at the top and
lateral sides. The horizontal well was completed at the midpoint of drainage area. Oil
zone was assumed as thin and thus the reservoir pressure at the water-oil contact was
constant throughout the production. Effects of wellbore radius and capillary and gravity
forces were not taken into account. According to the results end-point mobility ratio and
the location of wellbore with respect to the oil-water contact strongly affect the
performance of horizontal wells.

Chaperon (1986) theoretically estimated the critical flow rates for commencement of
water coning for vertical and horizontal wells. It is pointed out that critical flow rate is
inversely proportional to the vertical permeability in vertical wells, which means that as
the latter decreases, the former increases. Horizontal wells are advantageous in that,
they provide higher critical flow rates than vertical wells do. However, the condition
required here is that the ratio of vertical permeability to horizontal one is not too high
(Rajan & Luhning, 1993). The critical rates in vertical and horizontal wells are also
compared. The behavior of coning, which is called crest in horizontal wells was also
studied with the assumption of anisotropic formation and constant interface rise at an
infinite distance. Due to ignorance of water flow restriction, Chaperon’s (1986) theory
may cause exaggerated optimistic results about critical production rate (Recham et al.,
2000). It was observed in the study that critical cones get usually closer to horizontal
wells than to vertical ones. Chaperon (1986) further claims that the advantage of
horizontal wells over the vertical ones in that the former yield higher critical yields
could be lessened when decrease in vertical permeability is the case.

Makinde et al. (2011) used the ECLIPSE100 black oil simulator to study the water
coning behavior in horizontal wells. The oil column height below perforation was
considered as the critical criterion the coning behavior was related. It was surmised that
increase in production rate causes acceleration in water coning formation in horizontal
wells. Higher horizontal permeability is preferred for delaying water cone. Before
breakthrough vertical permeability does not have significant effect on coning behavior;
 11

nevertheless, after the breakthrough, the vertical permeability gives rise to increase in
water-oil ratio. Porosity proved to have no correlation with the performance after
breakthrough in horizontal wells. However, higher porosity contributes to delay of
water coning into the horizontal well. When the height above perforation in case of well
completion closer to the water-oil contact is increased, the rate of water coning is
accelerated. Longer perforations in horizontal wells make water-oil ratio reduce after
the breakthrough.

Abbas and Bass (1988) studied the performance of water coning using analytical,
numerical, and experimental approaches and for different boundary conditions. No
critical oil rate for unstable cone was obtained. They did not neglect the effect of limited
entry of the wellbore on oil production rate. Their results displayed fractional wellbore
should be penetrated at an interval 0.5 in in oil-water coning systems, which is
considered as the optimum interval.

Henley, Owens and Craig (1961) implemented the first scaled-model laboratory
experiments to observe oil recovery by reservoir with bottom water drive. The influence
of well spacing, fluid mobility, and production rate, capillary and gravity forces, well
penetration and well completion on the well performance was investigated. According
to the results, final sweep efficiency was not greatly affected by well penetration. It was
also indicated that while gravity forces have sustainable effects on sweep efficiency, the
capillary forces proved to have negligible effect. Furthermore, impermeable pancake
barrier can moderately increase the oil recovery efficiency.

Permadi et al. (1997) studied the water coning behavior in case of horizontal well
having stinger. According to the results of the experiment, without stinger recovery is
higher if it is early. Inserting the stinger flattens the summit of the cone and decelerates
the rate in increase of water production, which considerably improves the recovery. It is
also stated that optimum length of the stinger is approximately 0.3 of the horizontal
wellbore section length. The water breakthrough for wells without stinger took place
 12

much earlier than for those wells with stingers. Stinger also makes the displacement of
water more uniform and the leg of water cone not so steep. Thus, oil displacement was
more efficient with wells having stinger.

2.3. CORRELATIONS FOR CRITICAL RATE AND BREAKTHROUGH TIME

As discussed before, if the critical production rate is exceeded, the water cone will break
into the well after a certain period. Hence, many authors attempted to develop
correlations to predict the breakthrough time and critical production rate. Critical rate
and breakthrough time are investigated.

2.3.1. Correlations for Critical Production Rate in Vertical Wells

Empirical correlations have been developed to estimate the critical production rate in
vertical wells in the literature. Some of them are discussed below.

2.3.1.1. Hoyland-Papatzacos-Skjaeveland Correlation

Hoyland, Papatzacos, and Skjaeveland (1989) suggested analytical and numerical

correlations for prediction of critical oil rate. They assume bottom water coning in
anisotropic, homogeneous systems where the well is completed from the top of the
formation (Figure 3).
 13

Figure 3 – Partially penetrating well with boundary conditions (Hoyland et al. 1989).
For analytical method, they benefited from the theory of Muscat and Wyckoff (1935).
Thus, the shape of the cone is neglected. In the procedure for calculation, first
dimensionless radius is calculated from Equation 2. Next, dimensional critical rate for
different fractional penetrations is determined from Figure 4. Then the dimensionless
critical rate, qCD as a function of well penetration is plotted as in Figure 5. Fractional

well penetration hp
( ) is found and plot is extrapolated to find the dimensionless critical
h

rate. Finally, using the Equation 3 critical rate is calculated.

re kv
rD  (2)
h kh

 h 2 ( w - o ) kh 
qoc  0.246  10  -4
 qCD (3)
  o Bo 
where,
Bo : oil formation volume factor, Bo, bbl/STB
h : total thickness of the oil zone, ft
h p : penetration interval

k h : horizontal permeability, md
 14

k v : vertical permeability, md

qCD : dimensionless critical flow rate (it is correlated with the dimensionless radius rD ,
and fractional well penetration ratio ( h p ) , as shown in Figure 4.
h
q oc : critical oil rate, STB/day

re : radius of the reservoir, ft

 o : oil viscosity, cp
 o : oil density, lb/ft3
 w : water density, lb/ft3

Figure 4 – Critical rate correlation (Hoyland et al, 1989).

 15

Figure 5 – Critical rate correlation with fractional well penetration (Hoyland et al,
1989).

For the isotropic reservoir, where k v  k h , the relationship developed is given as in

Equation 4.

ko ( w - o )  h p 2 1.325 2.238 
qoc  0.924  10 -4 [1 - ( ) ] h [ln (re )]-1.99  (4)
 o Bo  h 

2.3.1.2. Chaperon Correlation

Chaperon (1986) assumes anisotropic formation. It is also assumed that the completion
interval is too short. Chaperon’s relationship accounts for the distance between the
production well and the boundary. The relation is given in the following equation.

k h (h - hp ) 2
qoc  0.0783  10 [  w -  o ] qc
-4 *
(5)
 o Bo

qc  0.7311  (1.943 /  " )

*
 16

re kv
" 
h kh

2.3.1.3. Guo-Lee Correlation

Guo and Lee (1992) assume a partially penetrating well in an isotropic formation. The
relation is shown in Equation 6.

1 1
hw ( - )
7.08  10 k v g
-3
 [re - re - re (h - hw ) ]2   
kv r w re
qoc  
2
(6)
o kh  kv
2 2 ln (re / rw )

where,

g  0.433  ( w -  o ) , psi/ft

hw : completed interval of the vertical well, ft

2.3.1.4. Ozkan-Raghavan Correlation

Ozkan and Raghavan (1990) assume an infinitely large reservoir. Equation 7 indicates
the expression they obtained.

k h h 2  h h
qoc  [0.546 - 0.021( w ) - 0.525 ( w ) 2 ] (7)
325.7  o Bo h h

2.3.2. Correlations for Breakthrough Time Prediction in Vertical Wells

Water breakthrough in vertical wells starts earlier than in horizontal wells. Ii is very
important to forecast when the water in the aquifer will start to break into the production
well. The most widely used correlations to predict the breakthrough time for vertical
wells are by Sobocinski and Cornelius (1964), and Bournazel and Jeanson (1971).
 17

2.3.2.1. The Sobocinski-Cornelius Correlation

The authors correlated the breakthrough time with two dimensionless parameters: the
dimensionless cone height and the dimensionless breakthrough time. Figure 6 shows
the schematic illustration of the well.

Figure 6 – Schematic diagram of a partially penetrating well subjected to water coning

(Sobocinski and Cornelius, 1964).
The dimensionless cone height is expressed by Equation 8:

(  w -  o )  k h  h (h - hp )
Z  0.492  10 -4 (8)
 o Bo qo
Where,
Z : dimensionless cone height

Dimensionless breakthrough time is calculated using Equation 9.

4Z  1.75Z 2 - 0.75Z 3
(t D ) BT  (9)
7 - 2Z

Time to breakthrough is calculated using Equation 10.

20.325  o h  (t D ) BT
t BT  (10)
(  w -  o ) k v (1  M  )
 18

where,
t BT : time to breakthrough, days
 : porosity, fraction
(k rw ) sor  o
M : water-oil mobility ratio and is defined as M  ( )
(k ro ) swc  w

(kro ) swc : oil relative permeability at connate water saturation

(krw ) sor : water relative permeability at residual oil saturation

  0.5 for M≤1

  0.6 for 1 <M ≤ 10

Joshi (1991) realized that there will be no breakthrough if Z  3.5 in Equation 9.

Hence, substituting this critical value of Z into Equation 8 and solving for q o will give
the critical production rate as shown in Equation 11.

(  w -  o ) k h h (h - hp )
qoc  0.141  10 -4 (11)
 o Bo
2.3.2.2. The Bournazel-Jeanson Correlation

Bournazel and Jeanson (1971) developed the correlation based on the experimental
studies. They used the same dimensionless group as in the methodology of Sobocinski
and Cornelius (1964). The steps for calculation of breakthrough time are as listed
below:

a) The dimensionless cone height is calculated from Equation 8;

b) The dimensionless breakthrough time is calculated using Equation 12;
Z
(t D ) BT  (12)
3 - 0.7Z
c) The time of breakthrough, t BT is calculated by substituting Equation 12 into
Equation 10. However, now instead of Z  3.5 , the critical value of 4.286 is used for
 19

Z . Inserting this value into Equation 8 and solving for q o yields the result shown in

Equation 13.
(  w -  o ) k h h (h - hp )
qoc  0.1148  10 -4 (13)
 o Bo

2.3.3. Correlations for Critical Production Rate in Horizontal Wells

Three correlations for critical rate calculation in horizontal wells are discussed in this
section. These are Chaperon (1986), Ozkan and Raghavan (1990), and Giger (1989)
correlations.

2.3.3.1. Chaperon Correlation for Critical Rate

Chaperon’s (1986) equation generally include both water and gas coning critical rates,
but here our discussion is restricted only to water coning in horizontal wells. In isotropic
formations with steady-state or pseudo steady-state flow conditions the author proposes
that horizontal wells allow higher critical rates than vertical wells. The equation derived
assumes the well to be located at the bottom of oil zone.

Initially Chaperon (1986) observed the effect of forces on a stable crest. It is determined
that balance between viscous and gravity forces keep the crest stable. The equation
derived for the flow potential is expressed in Equation 14.

q o πx πz
 ( x, z k )  log (cosh - cos k ) (14)
2π L k h h

where,
 : flow potential or gravity potential

q : actual oil rate, m3/hour

L : horizontal well length, m
k : permeability to oil, md
 20

h : initial oil thickness, m

zk : coordinate along vertical axis

x : abscissa (aside from the horizontal well)

Chaperon assumed radial flow around the wellbore in deriving Equation 14.

Finally, Chaperon equated the viscous potential difference to the gravity potential
difference and proposed equation predicting critical rate as in Equation 15.

*
L qc k (D ) 2
qoc  0.0783  10 -4
( w - o ) h b
ye  o Bo (15)

qc  3.9624955  0.0616438 " - 0.000504( " ) 2

*

ye kv
The equation is constrained by 1   " < 70, and 2 ye < 4 L where  " 
h kh

Db : distance between the WOC and the horizontal well, ft

q oc : critical oil rate, STB/day

ye : half distance between two lines of horizontal wells, ft

L : length of the horizontal well, ft

: dimensionless function of Joshi which is a function of  " .

*
qc

2.3.3.2. Ozkan-Raghavan Correlation

Ozkan and Raghavan (1990) assumed an infinitely large reservoir and sufficiently long
well. The well is placed at the top of oil zone. The authors proposed Equation 16 below
to find the critical rate of horizontal wells.

k h h 2   y y 2 y 
qoc  1.0194 - 0.1021( h ) - 0.2807( h )  ( h ) LD (16)
325.7  o Bo  
 21

where,

L kv
LD 
2h k h

2.3.3.3 Giger Correlation

For derivation the critical rate correlation Giger (1989) located the well near the top of
oil zone. In addition, he assumes the well extends throughout the oil zone. The external
boundary of reservoir is closed to lateral flow (Figure 7). For such a case Giger (1989)
proposed Equation 17 for calculating critical rate.

qoc 3  0.49  10 -3 k h D 16 y 2 0.5

 [(1  ) - 1] (17)
L  o Bo 3D 2

where,
D : lateral length of oil zone (D=2re=L), ft
y : vertical distance between initial WOC and horizontal well, ft

Figure 7 – Bottom water-drive case during production (Giger, 1989).

 22

2.3.4. Breakthrough Time in Horizontal Wells

There are various practices available to forecast the time to breakthrough. Critical flow
calculations generally show low results, which would be infeasible to continue
production with a given rate. Thus, most of the time engineers have to produce above the
critical rate. This leads the cone to break through after a specific period, which is called
breakthrough time. It is important to predict the evolution of the cone and the time to
breakthrough so that future completion and production scenarios can be predicted. The
determination of breakthrough time in horizontal wells was expressed by several
approaches. However, the only two base methods, which are Ozkan and Raghavan
(1990) method and Hoyland et al. (1989) method are discussed in the paper.

2.3.4.1. Ozkan-Raghavan Correlation

Ozkan and Raghavan (1990) method is an analytical method to investigate the behavior
of water cone and to predict the breakthrough time. The proposed correlation by authors
is applicable to both vertical and horizontal wells. Moreover, the equation is simplified
to provide quick estimation of breakthrough time for engineers before detail stimulation
results. Authors considered the oil-bearing formation, which is in contact with water at
the bottom. They assumed homogenous reservoir, maybe either bounded or infinite in
the radial direction. In addition, anisotropy property was taken into account. Originally,
WOC is assumed as a flat and an oil zone has a uniform thickness. Capillary pressure
gradients are neglected. The mobility of oil is assumed as identical to the mobility of
water. The water phase is assumed to be in gravity equilibrium for all times and the
initial pressure is preserved at WOC. The model proposed by authors considers an active
aquifer, not dead one (Ozkan and Raghavan, 1990). Schematic of water coning model is
represented by Figure 8.
 23

Figure 8 – Schematic of coning model, a system study (Ozkan and Raghavan, 1990).

Under assumptions stated above the authors came up with the simplified correlation of
time to breakthrough as given in Equation 18.

f d h 3 Es k
t BT [ ]( h ) (18)
5.615Qo Bo k v

f d   (1 - Swc - Sor )
where

Swc : connate water saturation, fraction

Sor : residual oil saturation, fraction

 : porosity, fraction

Es : sweep efficiency, dimensionless

qo : oil flow rate, STB/day

2.3.4.2. Hoyland et al. Correlation

Hoyland et al. (1989) approach on the other hand is in anisotropic, with an assumption
of infinite horizontal well located at the top or at the bottom of oil column. The
 24

correlation for time breakthrough is a semi analytical correlation that could be presented
in several steps:

Step 1. Calculation of dimensionless rate from the Equation 19:

20333.66   o Bo qo
qD  (19)
L h ( w - o )  kv kh

Step 2. Solve for the dimensionless breakthrough time tDBT by applying the relationship
given in Equation 20:

3q D
t DBT  1 - (3q D - 1) ln [ ] (20)
3q D - 1

Step 3. Estimate the time to the water breakthrough t BT by using the water and oil
densities in the following Equation 21:

22758.528  h   o t DBT
t BT  (21)
kv ( w - o )

2.4. COMPARISON BETWEEN RESULTS FROM RUBIS AND

CORRELATIONS

Many studies have been focused on development of correlations for critical rate
calculation. In the present study, an example problem was solved by means of these
correlations and RUBIS (Ecrin v4.20., 2013). Approaches used for the vertical well were
those developed by Chaperon (1986), Ozkan and Raghavan (1990), Guo and Lee (1992),
and Hoyland et al. (1989). For horizontal well, correlations used for comparison were
 25

Chaperon (1986), Özkan and Raghavan (1990), and Giger (1989) correlations. The field
data for the example problem for both vertical and horizontal wells are the same and
shown in Table 1. The results are summarized in Table 2 and Table 3, for vertical and
horizontal wells, respectively.

Several cases were run considering both anisotropic and isotropic formations. RUBIS
was run according to the assumptions in each correlations in order to make a reasonable
comparison. For example, for comparison the results from RUBIS with those of
Chaperon (1986) correlation, the well was placed at the top and other properties were
kept as constant. Then in order for comparison with the results from Ozkan and
Raghavan (1990) correlation, the radius of the reservoir was taken about 22000 ft, which
shows that the reservoir is very large.

As seen from Table 2, results from Guo and Lee (1992) and Hoyland et al. (1989)
correlations are comparable to each other. The ratio of completed interval to reservoir
thickness is 0.571 in their case. Chaperon (1986) correlation yields higher critical rate
since it is assumed that the completion interval of the well is very short. Ozkan and
Raghavan (1990) correlation proved to yield even higher result for critical rate. The
completion interval is the same as for Guo and Lee (1992) and Hoyland et al. (1989)
correlations. However, such a high value for critical rate may be the result of assuming
an infinitely large reservoir.

Table 1 – Field data for the example problem.

Resevoir temperature, °F 212

Reservoir initial Pressure, psia 5000
°API gravity 32.8
Vertical depth, ft 6000
Reservoir oil thickness, ft 42
Horizontal well length, ft 660
Reservoir drainage radius, ft 1053
Wellbore radius, ft 0.29
Vertical anisotropy ratio, fraction 0.1
 26

Horizontal permeability, md 37
Vertical permeability, md 3.7
WOC, ft 6042
Mobility ratio, M 3.27
Oil viscosity, cp 1.44
Water density, lb/ ft 68.36
Oil density, lb/ ft 53.75
Porosity, fraction 0.164
Residual oil saturation, fraction 0.337
Connate water saturation, fraction 0.288
Water salinity, ppm 1.00E+05
Pore compressibility, psi 3.0E+06
Water compressibility, psi 2.5E-06
Oil compressibility, psi 3.43E-6
Initial oil formation volume factor, bbl/STB 1.102
Initial water formation volume factor,
bbl/STB 0.999
Aquifer recharge index, bbl/psi-day 200

As seen from the results in Table 3, critical rate values obtained by Ozkan and
Raghavan (1990) correlation and RUBIS proved to be close to each other. The
horizontal well was positioned at the top of the reservoir and the reservoir is assumed as
infinitely large. Therefore, the critical rate in this case is the highest. Similarly, Giger
(1989) assumes the horizontal well at the top of the reservoir; however, the drainage
radius was used as given in the problem, that is, 1053 ft. That is why, the critical rate in
this case is lower.

Table 2 – Critical rate calculation approaches for vertical well.

Approach Vertical Anisotropy Critical Rate, STB/day

Chaperon 5.4
RUBIS 5.7
Guo-Lee kv 1.53
 0.1
Hoyland et al. kh 2.2
RUBIS 2.62
 27

Özkan-Raghavan 10.73
RUBIS 10.70
Hoyland et al. kv 1.69
1
RUBIS kh 2.39

Chaperon (1986) considers the horizontal well position at about the one third of the
reservoir oil thickness. Although the well is placed at the top for Giger (1989)
correlation, his critical rate is expected to be high than the one obtained from Chaperon
(1986) correlation. Nevertheless, the calculation results are controversial, in other words,
Chaperon (1986) approach yielded higher critical rate. This may result from some
parameters and assumptions that might have been ignored in some correlations and
taken into account in others. Based on the results of present approach, it can be
observed that RUBIS gave the anticipated results. Again, the different result obtained by
Chaperon (1986) might have been caused by implicit and unknown assumptions. For
example, Chaperon (1986) ignores the flow restriction and assumes the existence of
immobile water. Hence, the results may display high values for the critical rate. The
equation derived by Giger (1989) is said to be suitable for the low drainage width, more
precisely, around the wellbore. Thus, the Giger (1989) correlation may not yield
reasonable results in case of higher drainage width.

Table 3 – Critical rate calculation approaches for horizontal well.

Approach Critical Rate, STB/day

Chaperon 25

RUBIS 11.5

Ozkan and Raghavan 47

 28

RUBIS 44

Giger 18.9

RUBIS 14.5
 29

3. STATEMENT OF THE PROBLEM

Wells drilled in the reservoirs with the bottom water drive are usually produced above
the critical rate owing to economic reasons. This leads to water coning, or as called in
case of horizontal wells a water crest, and breakthrough of water into the well.

In this study we consider the optimal placement and optimal length of horizontal wells
for maximizing economic profit. The costs related to the problem are drilling costs and
water disposal costs. To find the optimal parameters, the simulator is run for various
horizontal well lengths and various well placements. The scenario that gives the
maximum profit is chosen as the optimal solution.
 30

4. DESCRIPTION OF THE MODEL

In the present simulation study the subprogram of ECRIN, RUBIS was used to study the
water coning in horizontal and vertical wells. First, a single vertical well was considered
in the study. Then a single horizontal well was completed in the reservoir. The reservoir
is assumed to have a cylindrical shape with the dimensions shown in Figure 9 and
Figure 10.

Figure 9 – A 3D shape of the oil reservoir with a horizontal well.

Figure 10 – A 3D shape of the oil reservoir with a vertical well.

 31

The reservoir has a bottom water drive with 200 bbl/psi-day, as recharge constant for the
Schilthuis water influx model (Craft and Hawkins, 1991). The cases are run for 10000
days for different production rates. Correlations used by RUBIS (Ecrin v4.20, 2013) for
the oil formation volume factor, compressibility of oil, and oil viscosity are Standing,
Vasquez-Beggs and Beggs-Robinson correlations, respectively. As for water, Spivey
and Van-Wingen and Frick correlations are used for water formation volume factor and
water viscosity, respectively. Reservoir and fluid parameters used in the study are given
in Table 4. Capillary pressure is assumed as negligible for more uniform and efficient
oil displacement and for more precise water saturation profile. In order to observe
coning and describe the pressure distribution and fluid flow more accurately, the model
was layered into two sections. The upper layer is discretized into 4 and the lower layer
into 30 grids in z-direction. The Figure 11 shows water and oil relative permeability
assumed in the simulation runs.

Figure 11 – Relative permeability curves.

 32

Table 4 – Reservoir and fluid parameters.

Resevoir temperature, °F 212

Reservoir initial Pressure, psia 5000
°API gravity 26
Vertical depth, ft 6000
Reservoir oil thickness, ft 100
Reservoir drainage radius, ft 8000
Wellbore radius, ft 0.3
Vertical anisotropy ratio, fraction 0.1
Horizontal permeability, md 100
Vertical permeability, md 10
WOC, ft 6100
Oil viscosity, cp 8.58
Water density, lb/ ft 68.36
Oil density, lb/ ft 57.64
Porosity, fraction 0.164
Residual oil saturation, fraction 0.2
Connate water saturation, fraction 0.1
Water salinity, ppm 1.22E+05
Pore compressibility, psi 3.0E+06
Water compressibility, psi 2.5E-06
Oil compressibility, psi 3.43E-6
Initial oil formation volume factor, bbl/STB 0.9731
Initial water formation volume factor, bbl/STB 0.9985
Aquifer recharge index, bbl/psi-day 200
 33

5. DESIGN STUDY

Several cases for two scenarios, a constant production rate and a constant bottomhole
flowing pressure, are run for vertical and horizontal wells. The production rate and the
bottomhole flowing pressure are chosen as 300 STB/day and 4800 psia, respectively for
both types of wells. The effects of completion interval of vertical wells, and the length
and vertical position of the horizontal well in those scenarios are studied. The simulation
is run for 10000 days and all the data in the tables in this chapter are taken at the end of
10000 days.

For economic analysis the oil price and water disposal cost were assumed as 100 $/STB
and 1.5 $/STB, respectively. Costs for vertical and horizontal wells are considered
differently. Cost per foot was taken as 300 $ for vertical wells and 750 $ for horizontal
wells. The cost for the horizontal well accounts for the total measured depth, that is,
vertical section plus horizontal section.

5.1. OPTIMIZATION PROCESS FOR CONSTANT PRODUCTION RATE

SCENARIO

Both vertical and horizontal wells are considered for optimization process. First, the
vertical well is assumed to be completed at different intervals in the reservoir (Figure
12). Then for the horizontal well the vertical position and the length of the well are
considered. For the vertical position a horizontal well having a length of 4000 ft is
placed at different interval in the reservoir (Figure 13). For length optimization, the
horizontal well is placed at 40 ft from the top of the reservoir (Figure 14).
 34

Figure 12 – Completion interval of the vertical well.

Figure 13 – Placement of the horizontal well at different vertical positions (L=4000 ft).
 35

Figure 14 – Schematic diagram for the optimization of horizontal well length (hL=40 ft).

Cumulative oil production and oil recovery factor remain constant for all cases since the
production rate is constant. Change in the reservoir pressure is not considerable since the
reservoir is assumed as infinitely large and a single well is drilled for production (Table
5, Table 6, and Table 7). The bottomhole flowing pressure shows significant change in
case of vertical wells, especially at upper part of the reservoir as can be seen from Table
5. From Table 5 it can be seen that when the vertical well is completed at the top,
cumulative water production is lower and it is increasing with the depth in the reservoir.
From Table 6, when the horizontal well is placed close to the top of the reservoir, again
cumulative water production is low. Completion of the well at the top of the reservoir
delays water coning. When the well is close to WOC, there is a high tendency for the
water cone to breakthrough into the well.

Horizontal well length is also used as one of the significant criteria for optimization in
case of horizontal wells. Increase in the horizontal well length delays water production,
which can also be observed from Table 7. The longer the horizontal well, the lower the
drawdown in the reservoir and bottomhole flowing pressures. Therefore, water influx
 36

decreases if the well length increases, that is, tendency of water to breakthrough into the
well is lower in case of longer wells.

It can be concluded that economic analysis of all the cases for constant production rate
scenario does not yield any considerable differences in results for profit values. As
expected, higher cumulative water production leads to higher disposal cost and gives
less profit (Table 5, Table 6, and Table 7). It is important to note that in the economic
analysis considered here only well cost, water disposal cost, and the money earned from
selling the oil is taken into account in a simple manner. Results could change if other
criteria or economic assumptions are assumed.

Table 5 – Data from RUBIS for optimization process at a constant production rate for
the vertical well at 10000 days.
Cost ($)
Qo Qw pres pwf Recovery Revenues
hw (ft) Water Profit ($)
(MMSTB) (MMSTB) (psia) (psia) Factor Well ($)
qo = 300 STB/day

Disposal
5 3.0 1.9088 4996.56 450 0.0062 1500 5.73E+06 3.0E+08 2.943E+08
15 3.0 1.9170 4996.56 3077 0.0062 4500 5.75E+06 3.0E+08 2.942E+08
25 3.0 1.9349 4996.55 3729 0.0062 7500 5.80E+06 3.0E+08 2.942E+08
40 3.0 1.9781 4996.54 4140 0.0062 12000 5.93E+06 3.0E+08 2.941E+08
50 3.0 2.0170 4996.52 4288 0.0062 15000 6.05E+06 3.0E+08 2.939E+08
70 3.0 2.1152 4996.5 4466 0.0062 21000 6.35E+06 3.0E+08 2.936E+08
80 3.0 2.1753 4996.47 4525 0.0062 24000 6.53E+06 3.0E+08 2.935E+08

Table 6 – Data from RUBIS for optimization process at a constant production rate for
the horizontal well based on the vertical position at 10000 days (L=4000 ft).
Cost ($)
Qo Qw pres pwf Recovery Revenues
hL (ft) Water Profit ($)
(MMSTB) (MMSTB) (psia) (psia) Factor Well ($)
Disposal
qo = 300 STB/day

5 3,0 0,03318 4998,22 4938,11 0,0062 3,00E+06 9,95E+04 3,0E+08 2,969E+08

15 3,0 0,07953 4998,06 4942,95 0,0062 3,01E+06 2,39E+05 3,0E+08 2,968E+08
25 3,0 0,20866 4997,77 4944,14 0,0062 3,02E+06 6,26E+05 3,0E+08 2,964E+08
40 3,0 0,65676 4997,2 4944,25 0,0062 3,03E+06 1,97E+06 3,0E+08 2,950E+08
50 3,0 0,99612 4997,08 4946,91 0,0062 3,04E+06 2,99E+06 3,0E+08 2,940E+08
70 3,0 1,75513 4996,72 4951,58 0,0062 3,05E+06 5,27E+06 3,0E+08 2,917E+08
80 3,0 1,88236 4996,74 4953,77 0,0062 3,06E+06 5,65E+06 3,0E+08 2,913E+08
90 3,0 2,06204 4996,66 4954,82 0,0062 3,07E+06 6,19E+06 3,0E+08 2,907E+08
 37

Table 7 – Data from RUBIS for optimization process at a constant production rate for
the horizontal well based on the well length at 10000 days (hw=40 ft).

Cost ($)
Qo Qw pres pwf Recovery Revenues
L (ft) Water Profit($)
(MMSTB) (MMSTB) (psia) (psia) Factor Well ($)
Disposal
qo = 300 STB/day

500 3.0 2.2556 4996.38 4746 0.0062 375000 6.767E+06 3.0E+08 2.9286E+08
1000 3.0 2.0191 4996.5 4843 0.0062 750000 6.057E+06 3.0E+08 2.9319E+08
2000 3.0 1.6526 4996.66 4902 0.0062 1500000 4.958E+06 3.0E+08 2.9354E+08
3000 3.0 1.1429 4996.92 4928 0.0062 2250000 3.429E+06 3.0E+08 2.9432E+08
4000 3.0 0.6568 4997.2 4944 0.0062 3000000 1.970E+06 3.0E+08 2.9503E+08
5000 3.0 0.2786 4997.6 4958 0.0062 3750000 8.358E+05 3.0E+08 2.9541E+08
6000 3.0 0.0320 4998.22 4969 0.0062 4500000 9.590E+04 3.0E+08 2.9540E+08
7000 3.0 0.0001 4998.44 4976 0.0062 5250000 2.133E+02 3.0E+08 2.9475E+08

5.2. OPTIMIZATION PROCESS FOR CONSTANT BOTTOMHOLE FLOWING

PRESSURE SCENARIO

The bottomhole flowing pressure is set constant as 4800 psia and as in case of constant
production rate scenario two types of wells are considered. The same parameters are
chosen to be optimized. There is a very slight decrease in the reservoir pressure in all
cases. As can be seen from Table 8, both cumulative oil and water production increases
as the vertical well is completed deeper in the reservoir. Cumulative water and oil
production is too low for the vertical well compared to the horizontal one. Although the
recovery factor increases when the true vertical depth is increased, it is still too low.

According to the results summarized in Table 9, again the location of the horizontal well
which minimizes the cumulative water production is the top of the reservoir. It can also
be seen that after 50 ft of completion, cumulative oil and recovery factor start to
decrease.

As can be seen from Table 10, difference between the profits for different lengths is
more noticeable rather than in case of vertical position. Cumulative oil and water
 38

production increases significantly when the horizontal well length increases. However,
in other cases there is no such a remarkable difference in cumulative production.

From Table 8 it can be deduced that the maximum profit is observed when the vertical
well is completed at 80 ft. This might have happened as a result of high cumulative oil
production. In this case cumulative water production is also high; however, its increase
is not as much as to affect the final profit values due to considerable difference between
the oil price and the water disposal cost. Table 9 shows that the optimum portion of the
reservoir thickness the well should be placed, is 50 ft. However, the difference between
the profit values is not considerable. Thus, the well can be positioned at any depth of the
reservoir. The horizontal well length proved to be as the primary criterion for
optimization. Hence, one may decide to drill as long horizontal well as possible
regardless of the vertical completion in order to maximize the profit as can be seen from
Table 10.

Table 8 – Data from RUBIS for optimization process at a constant bottomhole flowing
pressure for the vertical well at 10000 days.

Cost ($)
Qo Qw pres Recovery Revenues
hw(ft) Water Profit ($)
(MMSTB) (MMSTB) (psia) Factor Well ($)
Disposal
pwf = 4800 psia

5 0.177778 4.28E-07 4999.91 3.68E-04 1500 1.282713 1.8E+07 1.778E+07

15 0.420335 3.88E-06 4999.78 8.69E-04 4500 11.64402 4.2E+07 4.203E+07
25 0.629545 0.004548 4999.67 0.0013 7500 13644 6.3E+07 6.293E+07
40 0.860067 0.107888 4999.45 0.00178 12000 323664 8.6E+07 8.567E+07
50 1.00067 0.196634 4999.31 0.00207 15000 589902 1.0E+08 9.946E+07
70 1.263388 0.410167 4999.01 0.00261 21000 1230501 1.3E+08 1.251E+08
80 1.38973 0.532016 4998.85 0.00288 24000 1596048 1.4E+08 1.374E+08
 39

Table 9 – Data from RUBIS for optimization process at a constant bottomhole flowing
pressure for the horizontal well based on the vertical position at 10000 days (L=4000 ft).
Cost ($)
Qo Qw pres Recovery Revenues
hL (ft) Water Profit ($)
(MMSTB) (MMSTB) (psia) Factor Well ($)
Disposal
5 9,24342 8,63991 4988,7 0,01912 3,00E+06 2,59E+07 9,2E+08 8,954E+08
pwf = 4800 psia

15 9,84273 10,1485 4987,2 0,02036 3,01E+06 3,04E+07 9,8E+08 9,508E+08

25 10,1198 11,2632 4986,2 0,02093 3,02E+06 3,38E+07 1,0E+09 9,752E+08
40 10,2552 12,6182 4985,5 0,02121 3,03E+06 3,79E+07 1,0E+09 9,846E+08
50 10,2827 13,4673 4985 0,02127 3,04E+06 4,04E+07 1,0E+09 9,848E+08
70 10,2264 14,889 4984,2 0,02115 3,05E+06 4,47E+07 1,0E+09 9,749E+08
80 10,1785 15,4177 4983,9 0,02105 3,06E+06 4,63E+07 1,0E+09 9,685E+08
90 9,96543 15,3242 4984,1 0,02061 3,07E+06 4,60E+07 1,0E+09 9,475E+08

Table 10 – Data from RUBIS for optimization process at a constant bottomhole flowing
pressure for the horizontal well based on the well length at 10000 days (hL=40 ft).

Cost ($)
Qo Qw pres Recovery Revenues
L (ft) Water Profit ($)
(MMSTB) (MMSTB) (psia) Factor Well ($)
Disposal
500 2.66312 1.72506 4997.3 0.00551 405000 5175180 2.7E+08 2.61E+08
pwf = 4800 psia

1000 4.11902 3.4857 4995.2 0.00852 780000 10457100 4.1E+08 4.01E+08

2000 6.38694 6.70012 4991.7 0.01321 1530000 20100360 6.4E+08 6.17E+08
3000 8.35546 9.71869 4988.5 0.01728 2280000 29156070 8.4E+08 8.04E+08
4000 10.2552 12.6182 4985.5 0.02121 3030000 37854600 1.0E+09 9.85E+08
5000 12.0782 15.3425 4982.6 0.02498 3780000 46027500 1.2E+09 1.16E+09
6000 13.84364 17.8489 4980 0.02863 4530000 53546700 1.4E+09 1.33E+09
7000 15.5105 20.191 4977.6 0.03208 5280000 60573000 1.6E+09 1.49E+09

5.3. EFFECTS OF VARIOUS RESERVOIR AND WELL PARAMETERS ON

THE CRITICAL RATE

The most studies have been concentrated on water coning problem in terms of critical
rate, breakthrough time and prediction of WOR after breakthrough. However, there are
some studies, which give effort to understand the effects of some parameters on
 40

behavior of water coning in horizontal wells. These parameters are horizontal

permeability, thickness of oil reservoir, length of completion, density difference,
mobility ratio, length of horizontal well, vertical position of the well, anisotropy ratio,
etc. In order to understand the effects of some of these parameters on critical production
rate in horizontal wells, several cases were run in RUBIS.

In this study the critical rate was found as 188 STB/days. As can be observed from
Figure 15, if the well is produced at 188 STB/day water cone does not reach the well.
However, Figure 16 indicates the different situation where the critical rate is outpaced.
In this case the cone reaches the well.

Figure 15 – Water saturation distribution at qoc = 188 STB/days, (hL = 40 ft, L = 4000ft).

Figure 16 – Water saturation distribution at qo = 300 STB/days, (hL = 40 ft, L = 4000 ft).
 41

In order to understand the effect of vertical anisotropy on behavior of water coning, six
different values of kv/kh have been used and the result are given in Figure 17. When
kv/kh decreases, this means that the vertical permeability is decreasing, since horizontal
permeability is kept constant. It is observed that the critical rate is increasing with
decreasing kv/kh. This is expected, since if the vertical permeability is reduced, the
upward flow of water is delayed more. This results in a higher critical rate.

Figure 17 – The effect of vertical anisotropy on critical production rate.

For studying the effect of the length of horizontal well on performance of water coning
in horizontal wells, seven cases were run in simulation program. The results of
simulation are shown in Figure 18. As results indicated, the longer horizontal well
provides higher critical production rate. This is owing to the larger area open to flow in
case of horizontal wells. The direct relation between the critical rate and the horizontal
well length is also seen from correlations derived for predicting critical production rate
in horizontal wells. For example, Equations 15, 16, and 17 in Chapter 2 also indicate
that the critical production rate is directly proportional to the horizontal well length.
 42

350

Critical Rate (STB/day)

300

250

200

150

100

50
1000 2000 3000 4000 5000 6000 7000
Horizontal Well Length (ft)

Figure 18 – The effect horizontal well length on critical rate.

Difference between water and oil densities has also an influence on critical rate. Six
cases for horizontal wells with different water and oil densities were run. Other
parameters are kept constant as in base model. The results of simulation are shown in
Figure 19. It can be concluded that increase in difference between water and oil
densities results in increase of oil production and retard of water coning. Therefore, it
can be said that the light oil reservoir has less tendency for coning than heavy oil
reservoir.
 43

Figure 19 – The effect of the density difference on critical rate.

In the study, Schilthuis model of water influx was used as the aquifer recharge. Several
cases were run for the observation of the effect of Schilthuis constant on the critical
production rate. The results are plotted as in Figure 20. Increase in Schilthuis recharge
caused decrease in the critical production rate, which is not surprising, since increase in
Schilthuis constant means increase for water that influx into the reservoir. This increase
requires the production rate to be lowered in order for the water breakthrough to be
avoidable.

240

230
Critical Rate (STB/day)

220

210

200

190

180

170

160

150
50 150 250 350 450 550 650 750 850 950 1050
Recharge Constant (bbl/psi/day)

Figure 20– Effect of recharge constant on critical rate.

From Figure 21 it can be seen that the deeper the well location, the lower the critical
production rate. As the well is placed closer to the WOC, the effect of the pressure drop
becomes more sustainable and tendency of water to break into the well becomes higher.
Thus, production rate should be decreased when the well is produced in deeper parts of
the reservoir so that water tendency for breakthrough cannot become higher.
 44

300

Critical Rate (STB/day)

250

200

150

100

50

0
0 10 20 30 40 50 60 70 80 90
hw (ft)

Figure 21 – Effect of vertical position of horizontal well on critical rate.

The grid number in RUBIS is also crucial in predicting water coning behavior. It is
difficult to rely on the results obtained by simulation using the particular number of
grids. Change in the number of grids will cause change in the results. However, if the
time for project is not limited it is better to work with more grids in order to see clearly
the shape of the cone and the breakthrough time. As seen from Figure 22, there is a
logarithmic relation between the critical rate and the grid number. The former is a
logarithmically increasing funcion of the latter.

190
Critical Rate (STB/day)

185

180

175

170

165

160
5 10 15 20 25 30 35
Grid Numbers

Figure 22 – Effect of grid number on critical rate.

 45

6. CONCLUSIONS

Water coning is one of the severe problems encountered in petroleum engineering.

Therefore, a great importance should be given to the studies on this phenomenon. Many
reservoirs are bottom water drive, and oil from these reservoirs is usually produced at
higher rates than the critical rate. This generally results in early breakthrough of water
from aquifer into the well.

One of the main subjects discussed in this study is development of the field in terms of
optimization of length and position of wells in the given reservoir. The horizontal well
proved to be more cost-effective than the vertical well. In addition, sensitivity analysis
was conducted to observe the effect of some reservoir and fluid parameters on the
critical rate as well as the influence of grid number in RUBIS on the results.

Optimization process was performed for two scenarios: constant production rate and
constant bottomhole flowing pressure. For the first case, lower cumulative water
production occurs at the top of the reservoir. In case of longer horizontal well,
cumulative water production is also low. It means, placement of the horizontal well at
the top and making it as long as possible retard the water coning. For constant
bottomhole flowing pressure scenario, the longer the horizontal well results in higher
cumulative oil production. The true vertical depth of this well should be the middle of
the reservoir thickness, since cumulative oil production is the highest at this interval.

From the study on the effects of some parameters on the critical rate the following
results are obtained.

 Increase in vertical anisotropy causes decrease in critical production rate.

 Critical production rate is directly proportional to the horizontal well length.
 Heavy oil is more vulnerable to water coning than light oil.
 46

 The higher the initial reservoir pressure, the lower the critical rate.
 Higher recharge constant results in lower critical rate.
 As the well is placed closer to WOC, there is a higher tendency for water
breakthrough.
 Increase in the grid number in RUBIS leads to higher results for the critical rate.
 47

NOMENCLATURE

Bo : oil formation volume factor, Bo, bbl/STB

D : lateral length of oil zone, ft
Db : distance between the WOC and the horizontal well, ft
Es : sweep efficiency, dimensionless
fd : dimensionless parameter in Ozkan-Raghavan correlation
g : acceleration of gravity, ft/s2
g : difference between pressure gradients of water and oil, psi/ft
h : oil zone thickness, ft
hL : location of the horizontal well
hp : perforated interval, ft
hw : completed interval of the vertical well, ft
k : permeability to oil, md
kh : horizontal permeability, md
ko : oil permeability, md
(kro ) swc : oil relative permeability at connate water saturation
(krw ) sor : water relative permeability at residual oil saturation
kv : vertical permeability, md
L : horizontal well length, ft
LD : dimensionless horizontal well length
M : water-oil mobility ratio
pb : hydrostatic pressure in the water zone, psia
pres : reservoir pressure, psia
pwf : bottomhole flowing pressure, psia
q : actual oil rate, m3/hour
qCD : dimensionless critical flow rate
qD : dimensionless rate
qo : oil flow rate, STB/day
: dimensionless function of Joshi which is a function of  "
*
qc
 48

Qo : cumulative oil produced, MMSTB

Qoc : critical oil rate, STB/day
Qw : cumulative water produced, MMSTB
rD : dimensionless radius
re : radius of the reservoir, ft
S or : residual oil saturation, fraction
S wc : connate water saturation, fraction
t : thickness of oil horizon, ft
t BT : time to breakthrough, days
(t D ) BT : dimensionless breakthrough time
x : abscissa (aside from the horizontal well)
xA : location of a constant pressure boundary, ft
y : vertical distance between initial WOC and horizontal well, ft
ye : half distance between two lines of horizontal wells, ft
Z : dimensionless cone height
z : height of the water cone, ft
zk : coordinate along vertical axis
Zs : well to cone apex distance, ft

Greek letters:

 : constant for Sobocinski-Cornelius correlation

" : constant in Chaperon correlation
w : water specific gravity
o : oil viscosity, cp
o : oil density, lb/ft3
w : water density, lb/ft3
 : porosity, fraction
 : flow potential or gravity potential
 49

REFERENCES

Abbas, H. H., and Bass, D. M. (1988). The Critical Production Rate in Water-coning
System. Paper presented at the SPE Permian Basin Oil and Gas Recovery
conference. Midland, Texas, pp. 351-360.

Ahmed, T. H. (2010). Reservoir engineering handbook (4th ed.). Burlington:

Elsevier/Gulf Professional, pp.583-646.

Blades, D. N., and Stright, D. H. (1975). Predicting High Volume Lift Performance in
Wells Coning Water. PETSOC Journal Paper, pp. 62-70.

Bournazel, C., and Jeanson, B. (1971). Fast Water-Coning Evaluation Method. Paper
presented at the 46th Annual Fall Meeting of the Society of Petroleum
Engineers of AIME, New Orleans.

Chaperon I., 1986, Theoretical Study of Coning Toward Horizontal and Vertical Wells
in Anisotropic Formations: Subcritical and Critical Rates, SPE 15377,
presented at the 61st SPE fall meeting in New Orleans, Louisiana, October 5-
8, pp. 1-10.

Chierici, G. L. (1995). Principles of petroleum reservoir engineering (P. J. Westaway,

Trans). Vologna: Springer-Verlag, pp. 75-76.

Craft, B. C., and Hawkins, M. F. (1991). Applied petroleum reservoir engineering (2nd
ed.). New-Jersey: Prentice Hall, pp. 272-333.

Ecrin Version 4.20, 2013. KAPPA, Petroleum Exploration ant Production Software
Training ant Consulting, Paris, France.

Giger, F. M. (1989). Analytic two-dimensional models of water cresting before

breakthrough for horizontal wells. SPE, pp. 409-416.

Guo, B., and Lee, R. L. (1992). Determination of he maximum water-free production

rate of a horizontal well with water/oil/interface cresting, SPE, pp. 167-179.

Henley, D. H., Owens, W.W., and Craig, F. F. (1961). A scale-model study of bottom-
water drives. Journal of petroleum technology, pp. 90-98.
 50

Hoyland, L. A., Papatzacos, P., and Skjaeveland, S. M. (1989). Critical Rate for Water
Coning: Correlation and Analtytical solution. SPE Reservoir engineering,
pp. 495-502.

Jiang, Q., and Butler, R. M. (1998). Experimental and Numerical Modelling of Bottom
Water Coning to a Horizontal Well. The Journal of Canadian Petroleum
Technology, 37(10), pp. 82-91.

Joshi S., D. (1988). Augmentation of Well Productivity With Slant and Horizontal Wells,
SPE Phillips Petroleum Corporation, JPT, June, pp. 729-739.

Joshi, S., D. (1991), Horizontal Well Technology. Tulsa, OK: Pennwell Publishing
Company, Tulsa, Oklahoma, USA, pp. 251-252.

Kaneko, T., Mungan, N., Aziz, K, and Settari, A. (1973). Some Practical Aspects Of
Coning Simulation. Paper presented at the 24th Annual Technical Meeting
of the Petroleum Society of CIM. Edmonton.

Karp, J. C., Lowe, D. K., and Marusov, N. (1962). Horizontal barriers for controlling
water coning, pp. 783-790.

Khan, A. R. (1970). A Scaled Model study of Water Coning. SPE-AIME, Texas, pp.
771-776.

Law, D. C. and Jossy, W. E. (1996). Visualization Study of Anti-water Coning Process

Using Gels. Paper presented at techical meeting of the petroleum society,
Alberta, Canada, pp.1-6.

Makinde, F. A., Adelfidipe, O. A., and Craig, A. G. (2011). Water Coning in Horizontal
Wells: Prediction of Post-Breakthrough Performance. International Journal
of Engineering & Technology IJET-IJENS, 11(01), pp. 173-175.

Menouar, H. K., and Hakim A. A. (1995). Water coning and critical rates in vertical and
horizontal wells. Paper presented at the SPE Middle East Oil Show. Bahrein,
pp. 367-379.

Mungan, N. (1979). Laboratory study of water coning in a layered model. The Journal
of Canadian Petroleum, Montreal, pp. 66-70

Muscat, M., and Wyckoff, R. D. (1935). An approximate theory of water-coning in oil

production, pp. 144-157.
 51

Okwananke A., 2008, Analysis of Water Cresting in Horizontal Wells, SPE 119,
presented at the 32nd SPE International Technical Conference and Exhibition
in Abuja, Nigeria, August 4-6, pp. 1-10.

Ozkan E., and Raghavan R. (1990). A Breakthrough Time Correlation for Coning
Toward Horizontal Wells, SPE 20964, prepared to be presented at Europec
90, The Hagus, Nedherlands, October 22-24, pp. 209-214

Papatzacos, P., Herring, T. R. and Martinsen, R. (1989). Cone Breakthrough Time for
Horizontal Wells, SPE 19822, presented at the 64th Annual Technical
Conference and Exhibition, San Antonio, TX. Oct. 8

Permadi, P. (1996). Fast Horizontal-Well Coning Evaluation Method. Paper for

presented at the 19S6 SPE Asia Pacific Oil and Gas Conference, Adelaide,
Australia, pp. 613-622.

Permadi, P. and Jayadi, T. (2010). An improved water coning calculation for horizontal
wells. Paper presented at the SPE/RO&G Russian Oil & Gas Technical
Conference, Moscow, Russia, pp. 1-12.

Permadi, P., Wibowo, W., Alamsyah, Y., and Pratomo, S.W. (1997). Horizontal wells
completion with stingerfor reducing water coning problems. Paper presented
at SPE Production Operations Symposium, Oklahoma City, Oklahoma, pp.
595-602.

Pirson, S. J., and Mehta, M. M. (1967). A Study of Remedial Measures for Water-
Coning by Means of a Two-Dimensional Simulator. Paper presented at the
SPE of AIME, Houston, Texas.

Rajan, V. S. V., and Luhning, R. W. (1993). Water coning suppression. Journal of

Canadian Petroleum Technology, 32(4), pp. 37-48.

Recham, R., Osisanya, S. O., and Touami, M. (2000). Effects of Water Coning on the
Performance of Vertical and Horizontal Wells-A Reservoir Simulation Study
of Hassi R'mel Field, Algeria. Paper presented at the SPE/Petroleum Society
of CIM International Conference, Alberta, Canada, pp. 1-12.

Smith, C. R, and Pirson, S. J. (1963). Water Coning Control in Oil Wells by Fluid
Injection. SPE 613, Texas, pp.314-326.

Sobocinski, D. P., and Cornelius, A. J. (1964). A Correlation for Predicting Water

Coning Time, SPE-894, pp. 591-600.
 52

Thomas, F. B., Shtepani, E., Marosi, G., and Bennion, D. B. (2002). Production well
water coning – is there anything we can do? Paper presented at Canadian
International Petroleum Conference, Alberta, Canada, pp. 1-18.

Wu, G., Reynolds, K., and Markitell, B. (1995). A Field Study of Horizontal Well Design
in Reducing Water Coning. Paper presented at the International Meeting on
Petroleum Engineering. Beijing, China, pp. 329-340.

Url-1 < http://www.slb.com >, 28.03.2014.

 i

TABLE OF CONTENTS

ABSTRACT ........................................................................................................................i

LIST OF TABLES ............................................................................................................ ii

LIST OF FIGURES ......................................................................................................... iii

1. INTRODUCTION ........................................................................................................ 1

2. WATER CONING ........................................................................................................ 4

2.1. PHYSICS OF WATER CONING ............................................................................. 4

2.2. LITERATURE REVIEW ........................................................................................... 5

2.3. CORRELATIONS FOR CRITICAL RATE AND BREAKTHROUGH TIME ..... 12

2.3.1. Correlations for Critical Production Rate in Vertical Wells .................................. 12

2.3.1.1. Hoyland-Papatzacos-Skjaeveland Correlation .................................................... 12

2.3.1.2. Chaperon Correlation .......................................................................................... 15

2.3.1.3. Guo-Lee Correlation ........................................................................................... 16

2.3.1.4. Ozkan-Raghavan Correlation .............................................................................. 16

2.3.2. Correlations for Breakthrough Time Prediction in Vertical Wells ........................ 16

2.3.2.1. The Sobocinski-Cornelius Correlation ................................................................ 17

2.3.2.2. The Bournazel-Jeanson Correlation .................................................................... 18

2.3.3. Correlations for Critical Production Rate in Horizontal Wells .............................. 19

2.3.3.1. Chaperon Correlation for Critical Rate .............................................................. 19

2.3.3.2. Ozkan-Raghavan Correlation ............................................................................. 20

 ii

2.3.3.3 Giger Correlation ................................................................................................ 21

2.3.4. Breakthrough Time in Horizontal Wells ............................................................... 22

2.3.4.1. Ozkan-Raghavan Correlation ............................................................................. 22

2.3.4.2. Hoyland et al. Correlation .................................................................................. 23

2.4. COMPARISON BETWEEN RESULTS FROM RUBIS AND CORRELATIONS

................................................................................................................................... 24

3. STATEMENT OF THE PROBLEM .......................................................................... 29

4. DESCRIPTION OF THE MODEL............................................................................. 30

5. DESIGN STUDY ........................................................................................................ 33

5.1. OPTIMIZATION PROCESS FOR CONSTANT PRODUCTION RATE

SCENARIO ............................................................................................................... 33

5.2. OPTIMIZATION PROCESS FOR CONSTANT BOTTOMHOLE FLOWING

PRESSURE SCENARIO .......................................................................................... 37

5.3. EFFECTS OF VARIOUS RESERVOIR AND WELL PARAMETERS ON THE

CRITICAL RATE ..................................................................................................... 39

6. CONCLUSIONS ......................................................................................................... 45

NOMENCLATURE ......................................................................................................... 47

REFERENCES................................................................................................................. 49

View publication stats