Professional Documents
Culture Documents
Saidi9 and Van Golf Racht 10 for an Iranian carbonate fractured 3.1. Isotropic Fracture Permeability
reservoir (Haft Kel field). The aquifer characteristics are The results of different isotropic cases are shown in Figure
shown in Table 3. The effects of nine different parameters are 8. As we can observe, while the fracture permeability
discussed below. increases, the breakthrough time is also increased. It may be
explained that more fracture permeability causes more
1. Well Penetration horizontal distribution of cone than vertical distribution due to
Since the extent of well penetration into the pay thickness the gravity force and results a more stable phenomenon.
has considerable effect on water coning and water Figure 8 shows that after breakthrough the slope of water-cut
breakthrough time, this parameter is first investigated. curves have a little changes for different cases and the slope of
Our model has a well penetration equal to the half of the water-cut in the case of high permeability is more than the
oil pay thickness from top of the layer. Five cases with others. Also Figure 9 shows that above a crit ical value, the
different well penetrations have been selected while the total effect of permeability on delaying b reakthrough decreases. In
oil pay thickness was constant (300 ft ). The simulat ions have other words in low permeability cases breakthrough time is
been run for 546 days (1.5 years), and the resulting water -cut more sensitive to permeability than in high permeab ility
are shown in Figure 4. models.
Pressure drawdown, which is the source of v iscous force,
is a function of distance between the lowest perforation and 3.2. Anisotropic Fracture Permeability
the water oil contact. If this distance increases, the viscous Vertical Fracture Permeability Variations
force will decrease and water coning or water oil contact will Different vertical permeab ility cases with different vertical
be stabilized at higher production rate. to horizontal permeability ratio fro m 0.01 to 3 are selected
When well penetration percentage is decreased fro m 50% (Figure 10). As vertical permeab ility decreases, (Kv /Kh
to 16.67% (case 1), the breakthrough time is increased from decrease) the breakthrough time will be higher; however,
155.7 days to 296.1 days. Therefore, the mo re well comparing to horizontal fracture permeability, the
penetration, the less breakthrough time takes place. This result breakthrough time change is not considerable. This result is
is mo re obvious for case 5 when the penetration percentage is contrary to Van Go lf-Racht and Sonier studies.8 For all cases,
increased to 83.33% and the breakthrough time has been the slope of water cut curves after breakthrough converges to a
decreased to 8.8 days. unique curve. Figure 11 shows that in low vertical fracture
The trend of water-cut after breakthrough shows that well permeab ilities, the dependency of breakthrough time on the Kh
penetration affects only the breakthrough time and the is more than high vertical permeability models.
increasing trend of water-cut is almost the same for all models.
Also for different cases, the breakthrough time changes Horizontal Fracture Permeability Variations
linearly with penetration percent (Figure 5). In this part, eight simu lation models have been constructed
with Kv /Kh variations fro m 0.33 to 100 by the horizontal
2. Oil Layer Thickness permeab ility variat ions from 20 md to 6000 md. Any
The pay thickness is one of the uncontrollable parameters horizontal permeab ility above the base case increases the
in coning process. Five different cases have been selected to breakthrough time gradually. However, at low permeab ility
study the effect of oil layer thickness in constant penetration cases water cone breakthrough is faster.
percent (50%). The base case has an oil layer thickness of 300 Figure 12 shows the behavior of models after the
ft with 50% penetration (150 ft) fro m the top. The results are breakthrough. This Figure shows that for the lo w horizontal
shown in Figure 6. It is clear that the breakthrough time permeab ility cases, there is a sharp increase in water-cut after
directly depends on the oil pay thickness. By decreasing the the breakthrough. The slope of curves decreases when the
oil pay fro m 300 ft (basic model) to 100 ft for case 1, the horizontal permeab ility increases. The effect of horizontal
breakthrough time has approached to zero and there is an fracture permeability can be exp lained by this fact that any
instant coning in this thin pay thickness. increase in horizontal permeability results in a better
Another important point is the high slope of water-cut distribution of fluid in horizontal p lane; therefore, the water
curves for the cases with lower oil pay thicknesses. migrates better from cone shaped mediu m and the gravity
Figure 7 shows that for the oil pay thickness more than 200 force is able to act stronger. Also Figure 13 shows the
ft, breakthrough time is a linear function of oil layer thickness. breakthrough time as a function of horizontal fracture
But for layers thinner than 200 ft, there is no linear permeability.
dependency because of the fluid flow turbulence in thin oil
reservoirs, especially in fluid entry points. 4. Production Rate
The most controllable parameter in water coning
3. Fracture Permeability phenomenon is production rate. There is a critical rate in
Fracture permeability is the most important uncontrollable conventional oil reservoirs in wh ich lower than this rate, water
parameter in water coning phenomenon, because in fractured coning can be prevented. Different authors have been derived
reservoirs, water cone front in fractures moves faster than analytical equations and empirical correlations to determine
matrixes. First, we studied five different cases with isotropic this critical rate in conventional reservoirs.10,11 However, in
fracture permeability. Second, the orientation of fracture fractured reservoirs, because most of the wells have high
permeability has been studied. productivity and can produce at high rates; the crit ical rate is
very low from which is not economical to produce. Therefore,
SPE 108254 3
optimization of production rate is essential for controlling of the key role such that with a weak aquifer, the water coning
water coning problem by considering the most profitable long - can take place similar to a powerful aquifer.
term economical conditions.
For study of the effect of production rate, ten production 9. Fracture Spacing
rates have been chosen and production duration is selected The effect of fracture spacing on water coning has been
three years. Water-cut curves are shown in Figure 14. studied using six simu lation models. Breakthrough time for all
Decrease of production rate delays the water breakthrough cases is similar. A lso, as it is shown in Figure 22, all models
time significantly. A lso the slope of water-cut curves in h igh have similar behavior after breakthrough. It can be concluded
rates is more. Figure 15 shows the dependency of all cases on that although fracture spacing plays an important ro le in o il
production rate such that the water breakthrough decreases recovery, it has no effect on the water coning phenomenon.
when the production rate increases.
Multi-Well Studies
5. Mobility ratio At the second part of this work, Cartesian mult i-well
Eleven mobility rat ios have been selected by applying models with five-spot pattern have been constructed to
changes in oil v iscosity (Figures 16 and 17). As we expected, determine the differences between single-well and mult i-well
at mobility ratio of 10, the breakthrough occurred very fast studies. Grid sensitivity analysis has been done for both cases
(5.5 days) and the slope of the water-cut curve is high. But at to ensure the applicability of the models.
mobility ratio of 0.1 the breakthrough time increases to 287 Figure 3 shows the Cartesian model of five-spot pattern.
days and the slope of water-cut curve is low. Single -well and multi-well models in the geomet ry and the
One impo rtant point is different distributions of water -cone dimensions of drainage area are similar. The distance between
in different mobility ratios. The simu lation results show that in two wells is about 2000 ft (about 1000 ft drainage bound ary
high mobility ratio cases, although water breakthrough occurs for each well) and the total side length of the cubic model is
sooner, the water cone cannot distribute easily in a horizontal about 7150 ft. The total number of grid blocks is increased to
plane because the cone distribution in horizontal direct ion is 45630. The aquifer and PVT properties are the same as before.
less than vertical direction.
1. Model with Five Production Wells
6. Fracture Storativity In this case all five wells are production wells with the
Based on fracture storativity definit ion in equation 1 with same production rate as the single-well model. The
constant compressibility and applying the changes to the initial breakthrough time has increased to 487 days compared to
porosity values (base case), five different cases have been 155.7 days in the single-well model. Since the assumption of a
constructed. radial drainage area with radius of 1000 ft is not applicable
with four wells around the well No.1, the breakthrough time is
c
2 2 ................................................... (1) increased. Furthermore improvement of pressure drawdown
1c1 2 c2 exerted by new wells has been result in mo re stable rise of
water-oil contact. Figure 23 shows the water saturation
ω: Storativity (Storage Ratio) distribution in the fractures in depth of 205 ft fro m top of the
φ1 : Primary Porosity reservoir and 630 days after production. It is obvious that the
φ2 : Secondary Porosity cone of central well (No.1) has more arial distribution due to
c1 : Compressibility of Primary Porosity (1/psi) more pressure drawdown exerted by side wells. Also the cones
c2 : Compressibility of Secondary Porosity (1/psi) of other wells have been distributed asymmetrically due to the
pressure drawdown of well No.1. At last, Figure 24 shows the
The results show the increase of storativity has delayed the water-cut of the model after the breakthrough time, which has
water breakthrough time (Figures 18 and 19). Since the less increasing trend compared to the case of single-well
increase of storativity results a faster movement of oil fro m the model.
matrix to the fracture, less pressure drawdown takes place and
as a result the fast growth of the water cone into the wellbore 2. Model with One Production Well
is stopped. The side wells have been shut in order to study the effect of
drainage area and side production wells. In this case the cone
7. Matrix Permeability breakthrough time is 1460 days. It appears that the drainage
Figure 20 shows the water-cut versus the matrix permeability. area of the well changes the breakthrough time.
With the increase of matrix permeability, matrix-fracture fluid The water saturation distribution in this case is symmetrical.
transfer increases and the exerted pressure drawdown, which Figure 25 shows the water saturation distribution in the depth
is the cause of coning formation reduces. Dependency of the of 255 ft fro m top of the reservoir at the end of production
water breakthrough on matrix permeability has a linear trend. time (1460 days).
in 834 days, which is sooner than the case of one production 6. Study of water coning can be an important part of IOR,
well but better than the case of five production wells. Figures because some parameters such as oil layer thickness and
27 and 28 show the water saturation distribution in the depth permeab ility are uncontrollable parameters, but they may vary
of 185 ft and 195 ft fro m the top of the reservoir and after in different locations in the reservoir, therefore, the recovery
1297 days of production. factor can be maximized with a precise study resulting in
determination of the best wells locations and production
4. Different Production Rates scenario.
In this section the model with five production wells were 7. Increasing the number of wells increases the production cost
used and simulated for five d ifferent production rates. The but decreases the cone growth because of better pressure draw
breakthrough time is different than the single-well model; down distribution in the reservoir and also water-oil contact
however, the trend of the breakthrough time versus rate is becomes more stable.
similar. 8. The fracture pattern, especially around the well, is very
The simu lation results with different production rates are important in water coning study in fractured reservoirs.
shown in Figure 29. Figure 30 shows the dependency of the Accurate data of fracture pattern can imp rove the study
breakthrough time on the production rate. effectively.
74.914
Production Rate 5000 STB/day Water Density
(lb/ft3)
Bottom
300 0.1 500 20000 Drive
(360°)
6 SPE 108254
80
70
60
C on in gW a t e r - c ut %
b=50 ft
50
b=100 ft
40 b=150 ft (BM)
b=20 0 ft
30
b=25 0 ft
20
10
0
0 100 200 300 400 500 600
Time (day)
Figure 1: Typical Water Coning Situation
Figure 4: Models with Different Penetrations
350
300
B r e a k t hr o ugh T i m e ( d a y )
250
200
150
100
50
0
0 20 40 60 80 100
Figure 2: Schematic of Radial Single -Well Model
Penetration %
100
90
80
h=100 ft
70
C on in gW a t e r - C ut %
60 h=200 ft
50 h=300 ft (BM)
40 h=400 ft
30
h=500 ft
20
10
0
0 100 200 300 400 500 600
Time (day)
Figure 3: Schematic of Cartesian Multi-Well Model
Figure 6: Models with Different Oil Layer
Thicknesses
SPE 108254 7
450 80
400
70
B r e a k th r o u g h T i m e (d a y )
W a te r- cu t %
Kv/Kh=1 (BM)
200 40
Kv/Kh=0. 8 (Kv=1600 md)
150 30 Kv/Kh=0. 5 (Kv=100 0 md)
Figure 7: Breakthrough Time vs. Oil Pay Thickness Figure 10: Models with Different vertical Fracture
Permeabilities
80 250
70
50 150
K=1000 md
40 K=2000 md (BM)
K=3000 md
100
30
K=4000 md
20
50
10
0 0
0 100 200 300 400 500 600 0 1000 2000 3000 4000 5000 6000 7000
Time (day) Vertical Fracture Permeability (md)
Figure 8: Models with Different Fracture Figure 11: Breakthrough Time vs. Vertical Fracture
Permeabilities Permeability
90
250 80
Kv/K h=1 00 (Kh=20 md)
70
Kv/K h=1 0 (Kh=200 md)
200
60
B r e a k thr ou g h Tim e (da y)
250 90
80
MR=10
200 70
B r ea k t hr ou g h T i m e ( d ay)
MR=2
MR=1.3
60
MR=1.15
150
Water-Cut%
50 MR=1
MR=0. 85 (BM)
40
MR=0.7
100
30 MR=0.55
MR=0.4
20 MR=0.25
50
MR=0.1
10
0 0
0 1000 2000 3000 4000 5000 6000 7000 0 100 200 300 400 500 600
Figure 13: Breakthrough Time vs. Horizontal Figure 16: Models with Different Mobility Ratios
Fracture Permeability
100
90
Q=10000 STB/ D
80
Q=9000 STB/D
70
Q=8000 STB/D
60
Water-Cut%
Q=7000 STB/D
50 Q=6000 STB/D
30 Q=4000 STB/D
Q=3000 STB/D
20
Q=2000 STB/D
10
0
0 200 400 600 800 1000 1200
Time (day)
Figure 14: Models with Different Production Rates Figure 17: Breakthrough Time vs. Mobility Ratio
600 100
90
500 80
B r ea k t hr ou g h T i m e (d a y )
ω=0.7692
70
400
W a t e r- C u t %
60 ω=0. 6 (BM)
300 50 ω=0.25
40
ω=0. 0 62 5
200
30
ω=0. 00 66 2
100 20
10
0
0
0 2000 4000 6000 8000 10000 12000
0 100 200 300 400 500 600
Production Rate (STB/day) Time (day)
Figure 15: Breakthrough Time vs. Production Rate Figure 18: Models with Different storativities
SPE 108254 9
500 80
B r eakt h r o u g h T im e ( da y
70 Fract ur e Spaci n g =6 ft
400
60 Fract ur e Spaci n g =7 ft
W a te r - C u t %
300 50
Fracture Spacing=8 ft (BM)
40
Fracture Spacing=9 ft
200 30
Fract ur e Spaci n g =1 1 ft
20
100
10 Fracture Spacing=13 ft
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 200 400 600
Storativity Time (day)
Figure 19: Breakthrough Time vs. storativity Figure 22: Model with Different Fracture Spacing
Figure 20: Model with Different Matrix Permeability Figure 23: The Saturation Distribution in Fractures at
Depth 205 ft from Top of the Reservoir
80 30
70
Basicl Model 25
60 Small Size Aquif er
50
W a t e r -c u t %
Low Porosity
40 High Porosity 15
Low per m ea bili t y
30
High Per m ea bili t y 10
20 Weak Aquifer
Strong Aquif er 5
10
0
0
0 100 200 300 400 500 600 0 200 400 600 800 1000 1200
Time (day) Time (day)
Figure 21: Model with Different Aquifer Figure 24: Cartesian Multi-Well Model with 5
Characteristics Production Wells
10 SPE 108254
Figure 25: Water Saturation distribution in the Figure 28: Water Saturation Distribution in Fractures
Fractures in Depth of 255 ft from Reservoir Top in the Depth of 195 ft
16 50
45
14
40
12
35
W a ter- cut %
W a t e r -c u t %
2 5
0
0
0 100 200 300 400 500 600 700 800 900
0 200 400 600 800 1000 1200 1400 1600
Time (day) Time (day)
Figure 26: Cartesian Multi-Well Model with 3 Active Figure 29: Multi-Well Model with Different Production
Wells Rates
600
500
B r eakt h r o u g h Ti m e ( da y)
400
300
200
100
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
40
35
30
25
Water-cut%
b=200 ft (66.67% )
20
b=150 ft (50% ) BM
15
b=100 ft (33.33% )
10
0
0 200 400 600 800 1000 1200
Time (day)
700
600
500
400
300
200
100
0
0 10 20 30 40 50 60 70 80
Penetration %