data from a long period of waterflooding reveal a low water cut. The openhole logging results on sidetracks also show oil recover- ies in excess of the laboratory ...
Effect of Pressure Gradient and Initial Water Saturation on Water Injection in Water-Wet and Mixed-Wet Fractured Porous Media Guo-Qing Tang,* SPE. and Abbas Firoozabadi, SPE, Reservoir Engineering Research Inst.
Summary A systematic study of the effect of wettability and initial water saturation on water injection and imbibition is made in Kansas outcrop chalk samples. (Kansas outcrop chalk is very similar to the rock matrix of the North Sea fractured chalk reservoirs.) Waterinjection tests were conducted at different pressure gradients to simulate the effect of gravity (that is, negative P,) on recovery. Based on a large number of carefully conducted tests, the following conclusions are drawn: Initial water saturation has a pronounced effect on water injection in an intermediate-wet chalk. This effect is much less pronounced for a strongly water-wet chalk. The effects are also in opposite directions. Pressure gradients (which simulate the negative P, effect) have a very strong effect on water-injection performance in an intermediate-wet chalk. Our interpretation of these experiments leads to the conclusion that recovery from the chalk reservoirs may be nearly independent of wettability state. The results from the experiments also reveal that there is no relation between laboratory measurements of spontaneous imbibition and field performance of mixed-wet reservoirs, even when the wettability state is perfectly restored in the laboratory. Introduction Wettability state and its effect on oil recovery has been the subject of numerous studies since 1928.'-8 However, major issues of oil recovery related to wettability remain unresolved. A major parameter of wettability is contact angle. Some authors have even questioned the usefulness of contact angle in defining ~ e t t a b i l i t y . ~ The state of wettability in some reservoirs can vary significantly with depth and rock properties. Jerauld and Rathmel14 presented data showing that there is a clear dependence of residual oil saturation upon depth. In the Prudhoe Bay reservoirs, residual oil saturation to waterflood decreases with depth while the reservoir wettability changes from less-water-wet to more-water-wet conditions with an increase in depth. In the Ekofisk field, which is similar to the Prudhoe Bay reservoirs, there is a systematic change of wettability with depth; water-wetting increases with depth. Spontaneous water imbibition in the laboratory shows poor recovery in the cores from the upper part of the reservoir. However, field data from a long period of waterflooding reveal a low water cut. The openhole logging results on sidetracks also show oil recoveries in excess of the laboratory imbibition measurements.' Despite wettability variation in the Ekofisk field, residual oil saturation is low and independent of the depth. In the Ekofisk field, from the commencement of water injection in 1987, the oil rate has increased from approximately 75,000 B/D to approximately 250,000 B/D in 1997."' Research concerning the effect of gravity forces on water injection in fractured reservoirs is rather limited. In 1968, Hamon" found ' Now with Stanford U Copyright 0 2001 Society of Petroleum Engineers
This paper (SPE 7471 1) was revised for publication from paper SPE 59291, first presented at the 2000 SPEiDOE Improved Oil Recovery Symposium. Tulsa, >5 April. Original manuscript received for review 12 June 2000 Revised manuscript received 11 July 2001 Paper peer approved 1 October 2001.
516
that oil recovery by water drainage in oil-wet fractured porous media could be significant depending on matrix permeability. Similar results were reported recently by Putra et al." Zhou et ul.I3 observed that B decrease in water-wetness of Berea sandstone by adsorption af polar oil components could increase oil recovery by waterflooding. Graue et aZ.14 obtained similar results for low permeability chalk. The effect of initial water saturation on oil recovery remains controversial. Brown15 studied the effect of initial water saturation on waterflood efficiency. He emphasized that connate water (retained as water film and in small pores) could become re-mobilized when water is invading. His experimental results showed that flow of connate water improves waterflood recovery. Skauge et ul." studied the influence of connate water on oil recovery by gas gravity drainage using chalk samples. The maximum oil recovery was obtained &t approximately 30% initial water saturation. Viksund et al.I7 camed out spontaneous imbibition tests with strongly water-wet chalk. A maximum oil recovery was obtained at approximately 34% of initial water saturation. However, results from Narahara et uZ.l8 are mudh different. They measured gas and oil relative permeability in waterwet and mixed-wet Berea at various initial water saturations and found that gas and oil relative permeabilities are independent Qf initial water saturation. Zhou et a1.I9 observed that for a crud& oil/brine/rock system, imbibition recovery increased with initial water saturation, but waterflood recovery decreased with initial water saturation. A long induction time (ranging from 10 to 1,000 minutes) was observed in imbibition tests after the cores were aged with crude oil at T=88"C for 10 days. The main objective of this work is to understand the mechanisms that lead to a vast difference between laboratory spontaneous imbibition measurements and field performance. For this p u r p o ~ , we have conducted an extensive set of laboratory measurements $0 Kansas outcrop chalk with a porosity of approximately 30% and a permeability of some 0.5 md. Waterflood and spontaneous imbibition performances of the Kansas outcrop chalk are studied before and after wettability alteration. We have used Kansas chalk because of its similarity to the chalk matrix rock from the Ekofisk chalk field. In this paper, we present the experimental results that include wettability alteration by chemical adsorption, water injection, and spontaneous imbibition in strongly water-wet and weakly waterwet chalks. In some of the experiments, an initial water saturation was present. In the injection experiments, the pressure gradient is varied to simulate the effect of gravity on recovery.
Materials and Experimental Setup Fluids and Chemicals. Normal decane (n-Clo) with a density ,of 0.73 g/cm3 and a viscosity of 0.92 cp at 24°C is used as the oil phase. Stearic acid (octadecanoic acid), purchased from Sigma with a purity of 99% and a molecular weight of 284.5, is used a$ a surfactant to alter the chalk wettability. This chemical is dissolved in oil (n-Clo) to prepare the stearic acid solution. Solubility tests at room temperature show that stearic acid dissolves in oil when the concentration is less than 2,000 ppm, but it hardly dissolves in water. NaCl and CaC12 are used to prepare the 0.1 % NaCl +O. 1% CaC12 brine solution, which is used both for injection water and for the establishment of initial water saturation. The viscosity and density of the brine are 1.0 cp and 1.02 g/cm3 at 2 4 T , respectively. December 2001 SPE Reservoir Evaluation & Engineering
r'
- End cap
1 1
Coreholder (ID=8.55 cm)
1
2 3
Chalk samples
t.-- (OD=8.50 cm) -Vertical fracture (0.015 to 0.02 cm)
4 5 6
Diarnete~3.8to 5.1 cm Length=5.2 to 6.6 cm
(a) Configuration A
A
Horizontal fracture (0.005 to 0.01 cm)
Total core lcngth=104.7 cm Overall porosity=30.7% Effective permeability= 14.7 darcies
(b) Configuration B
Fig. 1-Configurations of Kansas outcrop chalks used in the experiments.
Rock. Kansas outcrop chalk is used in all experiments in two different configurations: A and B (Fig. 1). Analysis of Kansas chalk indicates 99% calcite and 1% quartz, which resembles that of clean North Sea chalk lacking significant chert and clay minerals. Kansas chalk shows close similarities with the reservoir chalk from some of the North Sea chalk reservoirs in regard to capillary pressure, porosity, permeability, and relative permeability. Configuration A is a single cylindrical plug with a diameter of 5. I cm and a length of 5.2 to 6.6 cm. The measured air permeability of the chalk is approximately 0.5 md, and the porosity is approximately 30%. Configuration B is a composite core consisting of six cylindrical blocks that measure 8.50 cm in diameter. The chalks are stacked and housed in an aluminum coreholder. The total length of the composite chalk is 104.7 cm. The measured annular aperture (between the chalk and the inside surface of the coreholder) is approximately 250 pm for one set of tests and 150 pm for another set. The overall porosity (fracture/matrix) is approximately 30.7%; the matrix porosity is approximately 30%. The fracture volume is 60 to 70 cm3.Total pore volume (matrix and fracture) is 1740 cm3. The effective permeability (fracturelmatrix) is approximately 14.7 darcies lor the fracture aperture of 250 pm and about 4.5 darcies for the fracture aperture of 150 pm.
Apparatus. Fig. 2a shows the experimental apparatus for water injection. It consists of a water-injection pump, a high-pressure cylinder, a water reservoir, pressure transducers, a vertical aluminum coreholder, an oil and water collector, and a vacuum pump with a trap embedded in dry ice. The system can be used to perform water-injection tests at either a constant injection rate or a constant inlet pressure. The outlet pressure is atmospheric. Fig. 2b shows the apparatus for spontaneous imbibition tests consisting mainly of an electronic balance. Experimental Procedure Establishment of Initial Water and Oil Saturations. We used two different procedures to examine the suitability for the establishment of initial water saturation (S,,J. In one procedure, the initial water saturation is established by oilflooding a water-saturated rock for Configuration A in a coreholder. In the second procedure, evacuation is used to establish initial water saturation. In both approaches, the two sides of the core are used to establish S,,;-in the flooding method, oil is injected from one end initially, then from the other end; in evacuation, both ends of the core are subjected to vacuum. The imbibition results from the two procedures for a core used in evaluation is the same. We also test the uniformity of S,; by cutting a core in three parts-the measured saturation is the same in the pieces (in this procedure, S,v, in the gadwater system is tested). In this work, we used the vacuum procedure to apply the same method for both the unfractured and fractured configurations. Details of the evacuation process are described next. After the chalk sample is 100% saturated with brine, it is evacuated to 20 to 50 mbar to reduce water saturation. To speed up evaporation, the coreholder is heated to 66"C, accompanied by evacuation. When the designated water saturation is established, the coreholder is cooled gradually to room temperature. The total water production is measured by weighing the chalk plug for Configuration A and by measuring the produced water collected by the trap for Configuration B. This step provides the approximate initial water saturation. The core is then vacuumed at 20 to 50 mbar for 4 to 6 hours before the oil-saturation step, which lasts approximately 2 days. The initial water saturation, S,.,, is calculated using the difference between the total pore volume and total saturated oil volume. For Configuration B, the fracture volume is excluded froni the total pore volume for the calculation of initial water saturation.
Wettability Alteration. Stearic acid is used as a surfactant to alter the wettability of the chalk. The solution is prepared by dissolving stearic acid in oil. The chemical treatment includes the followin4 steps: the chalk sample is (1) saturated with the stearic acid soh+
Water-injection pump Electronic balance
ressure transducer
Liquid trap
(a) Water injection
u
-
(b) Spontaneous imbibition
Fig. 2-Schematic of apparatuses for water injection and spontaneous imbibition tests.
December 200 I SPE Reservoir Evaluation & Engineering
517
3
tion (e.g., 500 ppm solution), (2) aged with the same stearic acid solution at room temperature lor I O to 20 days, (3) dried and then aged at 105°C for 3 to 5 days, followed by cooling it to room temperature, and (4) resaturated and re-aged with the same stearic acid solution at room temperature for 10 to 20 days. After chemical treatment, the wettability state is determined by conducting spontaneous imbibition and waterflooding tests. The chemical treatment is repeated when the wettability is not stable. For this work, we repeat the chemical treatment procedure twice to establish a stabilized wettability for most cores used in our study.
Wettability Measurement. The chalk plug saturated with oil is hung in a beaker containing brine (0.1% NaCI+O.l% CaC12).
TABLE I-RELEVANT DATA FOR CONFIGURATION A
Initial Wettability 0 1.0 0 0 1.0 0
6.35 6.35
2a 2b 2c 3a 3b 3c 4a 4b 4c 5a 5b 5c
Effect of C, on Wettability 6.51 30.7 0 100 0.83 0 62.0 6.51 30.7 0 100 0.83 0 62.4 6.51 30.7 0 100 0.83 0 63.0 6.62 30.1 0 53.0 0 200 0.74 6.62 30.1 0 53.4 0 200 0.74 6.62 30.1 0 53.4 0 200 0.74 6.31 30.9 0 500 0.66 0 46.0 6.31 30.9 0 500 0.66 0 46.5 6.31 30.9 0 500 0.66 0 48.5 6.50 30.2 0 1,000 0.07 0 3.2 6.50 30.2 0 1,000 0.07 0 2.7 6.50 30.2 0 1,000 0.07 0 2.7 Effect of Ap on Oil Recovery by Water Injection
6a 6b 6c 6d 6e 7a 7b 7c 7d 8a 8b 8c
5.41 5.41 5.41 5.41 5.41 5.21 5.21 5.21 5.21 5.20 5.20 5.20
31.1 31.1 31.1 31.1 31.1 30.7 30.7 30.7 30.7 30.6 30.6 30.6
8d 8e 8f
5.20
30.6
9a 7b 8b 9b 7d 8f I Oa Ob I oc I Od I Oe I Of I og Ila IIb 12a 12b I2c
5.20 30.6 5.20 30.6 Effect of 5.22 30.0 5.21 30.7 5.20 30.6 5.22 30.0 5.21 30.7 5.20 30.6 Effect of 6.01 29.6 6.01 29.6 6.01 29.6 6.01 29.6 6.01 29.6 6.01 29.6 6.01 29.6 5.31 30.5 5.31 30.5 5.40 30.2 5.40 30.2 5.40 30.2
0 0
0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 500 500 500 500 1,000 1,000 1,000
1.0 1.0 1.0 1.0 1.0 0.55 0.55 0.55 0.55 0.07 0.07 0.07
0 0.09 0.37 0.93 12.9 0 0.96 3.85 13.5 0 0.96 1.92
1,000
0.07
-
66.2 66.2
-
-
-
-
-
-
6.55 0 1,000 0.07 7.69 0 1,000 0.07 12.9 I,, on Oil Recovery by Water Injection 0 200 0.82 0.96 0 500 0.55 0.96 0 1,000 0.07 0.96 0 200 0.82 13.5 0 500 0.55 13.5 0 1,000 0.07 12.9 0
-
-
-
-
-
66.2 66.2 66.2 66.2 66.2 39.0 59.5 66.2 72.0 6.0 16.0 42.0 49.5 73.0 78.0 60.5 59.5 16.0 66.2 72.0 78.0
S ,, on Oil Recovery by Water Injection 0 0 1.0 0.93 - 66.7 10 0 1.0 - 67.2 0.93 0 1.0 21 0.93 - 66.2 0 1.0 32 0.93 - 65.5 0 1.0 33 0.93 - 65.0 0 1.0 39 - 60.0 0.93 45 13 1.0 - 53.2 0.93 10 500 0.55 - 61.3 0.96 20 500 0.55 0.96 - 61.0 10 1,000 0.09 - 38.0 0.96 19 1,000 0.09 - 58.5 0.96 10 1,000 0.09 0.96 - 39.1
Note k = 0 5 md, d = 5 08 cm (Runs 2 through 12) and d = 3 81 cm (Run 1)
518
Water Injection. Water-injection tests are performed either at a constant injection rate or at a constant inlet pressure with the outlet end open to atmosphere. For Configuration A, all tests are performed at a constant inlet pressure (Ap varied from 0 to 13.5 psi/cm). For Configuration B, a constant water-injection rate (5 t~ 68 cm'/min) is used; the pressure gradients (excluding gravity) after water breakthrough, Apbrrare small owing to the vertical fracture. Note that all the experiments are performed at room temperature. The experimental data are listed in Tables 1 and 2.
Results Wettability Alteration. Change in wettability is assessed by both the rate of spontaneous imbibition and the Amott index to water (/u,v). Initial Wettability. Kansas outcrop chalk before wettability
la Ib
30.6 30.6
Change in the weight obtained by reading the electronic balance is recorded vs. time. At the end of the spontaneous imbibition test, the chalk plug is waterflooded at a high rate of 2 cm'/min to residual oil saturation. The endpoint oil recoveries by spontaneous imbibition and waterflood are used to calculate the Amott index to water;' zliM..
alteration is strongly water-wet.2' The residual oil saturation to water from spontaneous imbibition is approximately 34%. To provide a reference for various wettability states, we carried out the spontaneous imbibition test with a chalk plug (Configuration A) before wettability alteration. The result is shown in Fig. 3. Water imbibition occurs as soon as the chalk plug is placed in water; most oil is produced in less than 100 minutes. The final oil recovery by spontaneous imbibition is approximately 66% (OOIP). After the spontaneous imbibition test, the chalk plug is waterflooded; the extra oil recovery by waterflooding is very small, approximately 1 % (OOIP). Therefore, /,, for the chalk plug is close to 1.0; the chalk is strongly water-wet (SWW). A second spontaneous imbibition test under the same test conditions is carried out, and the result is the same as the first test. In this paper, the spontaneous imbibition curve presented in Fig. 3 is used as a reference to assess wettability alteration by adsorption of stearic acid.
TABLE 2-RELEVANT DATA FOR CONFIGURATION B k csn 9 APiz R, Run Core (darcy) SW, (ppm) (cm3/rnin) (psi/cm) (YO) __ ___ __ . ___ _ _ ~ Effect of q on Oil Recovery by Water Injection
-
~~
13
B-1
14.7
0
0
1.0
5.0
0.010 66.2
14
B-I
14.7
0
0
1.0
11.0
0.014 66.3
15a
B-1
14.7
0
0
1.0
26.2
0.020 66.4
15b
B-I
14.7
0
0
1.0
26.2
0.020 66.2
16a
B-2
4.5
0
500
-
10.0
0.025 47.0
16b
8-2
4.5
0
500
-
10.0
0.025 44.7
17
B-2
4.5
0
500
-
30.0
0.046 53.0
18
8-2
4.5
0
500
-
68.0
0.101 60.0
Effect of S , on Oil Recovery by Water Injection 19
B-I
14.7
0
0
1.0
26.2
0.020 66.2
20
B-I
14.7
13.9
0
1.0
26.2
0.020 61.9
21
B-1
14.7
21.2
0
1.0
26.2
0.020 57.0
22
B-I
14.7
36.8
0
1.0
26.2
0.020 52.0
18
8-2
4.5
0
500
-
68.0
0.101 60.0
23
8-2
4.5
10.0 500
-
68.0
0.101 62.6
24
8-2
4.5
20.0 500
-
68.0
0,101 59.8
Note: (1) the fracture aperture was approximately 250 pm for B-1 and 150 prn for 8-2.(2) bp,, is the differential pressure afler water breakthrough (3)d=8.5 cm. L=104.7cm, and "J =30.7%.(4)I,, is not valid for the 8-2 core.
December 200 1 SPE Reservoir Evaluation & Engineering
-
70
L
8 s g $
70
I
60
60
50
50
40
3011I
P)
E
20
is
+Run +Run +Run - S W
40
30
t
04 0
s,,=o
20
lo
1,000
500
1,500
2,000
3,000
2,500
2a 2b 2c
3,500
Imbibition Time, minutes 0
500
Fig. 3-Spontaneous imbibition in the chalk plug before wettability alteration: Configuration A.
Wettability After Chemical Treatment. To reduce waterwetting, the chalk plugs are treated with different concentrations of stearic acid. A series of experiments are conducted to optimize the treatment procedure. It is found that wettability alteration is strongly influenced by aging time, aging temperature, and stearic acid concentration. Fig. 4 shows the results for the chalk plug at room temperature (Configuration A) treated with CSA=100 ppm (&=stearic acid concentration). The aging time is 2 days. Reduction of waterwetting for this chalk plug is small. The final oil recovery is close
1,000
1,500
2,000
2,500
3,000
3,500
Imbibition Time, minutes Fig. 4-Wettability transition toward more water-wetting with repeated spontaneous imbibition tests: Configuration A ( CsA=100 ppm).
to that for the strongly water-wet state. The spontaneous imbibition rate increases as the chalk plug is used repeatedly (from Run 2a through Run 2c). This behavior indicates that the wettability stak is not stable, which may be related to desorption of stearic acid from the rock surfaces. However, when the aging time and aging temperature are increased, the wettability alteration is significant and stable. Figs. 5 through 7 show the results for the chalk plugs 70,
60
% 0 50
'$
II
8
0
2 0 2is
60
-
50
-
L
40
c
30
d-
20
is
10
s,=O
40-
-
+Run 4a Run 4b +Run 4c
3020-
-S W
10 -
0 0
2,000
4,000 6,000 8,000 Imbibition Time, minutes
10,000
0
12,000
4,000
2,000
6,000
8,000
10,000
12,OOD
Imbibition Time, minutes
Fig. 5-Spontaneous imbibition in the chalk plug with altered ppm). wettability: Configuration A (CSA=~OO
Fig. 6-Spontaneous imbibition in the chalk plug with altered wettability: Configuration A ( C ~ ~ = 5 0ppm). 0
60
b%80 s
i;
ss8
B B
50
40 40
stopped for 14 hours
30 30
law=l.o + S,=O
20
-0-
S,=13.9%
++
Swj=36.8%
* S,=21.2%
10 10 0
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
Water-Injection Time, minutes Fig. 7-Spontaneous imbibition of the chalk plug with altered wettability: Configuration A (CSA=l,OOO ppm).
December 200 I SPE Reservoir Evaluation & Engineering
519
treated with CSA=200, 500, and 1,000 ppm, respectively. These chalk plugs are aged at T=24”C for 20 days, dried and aged at T= 105°C for 5 days, and re-aged with the stearic acid solution at room temperature for 10 days. To establish a stable wettability alteration, the chalk plugs are treated using the above procedure twice. Three separate spontaneous imbibition tests are conducted with the same chalk plug to study the stability of the altered wettability. Results presented in Figs. 5 through 7 are in the order in which the tests are carried out (for example, in Fig. 5 , Run 3a is conducted first, and Run 3c is conducted last). Fig. 5 shows the recovery data for the chalk plug treated with cS,4=200 ppm; the induction tirne is 600 minutes for all three runs. After the induction period, the water imbibition begins and the imbibition rate increases quickly. The final oil recovery by spontaneous imbibition is approximately 53% (OOIP), which is about 13% less than before wettability alteration. The results are close for Runs 3a through 3c, which confirms that the altered wettability state for this chalk plug is stable. Fig. 6 shows the recovery data for =500 ppm. The induction time is the chalk plug treated with CSA approximately 1,400 minutes, and final oil recovery is approximately 48% (OOIP). However, increase in imbibition rate is lower than that for CSA=200 ppm, which reflects the effect of the concentration of stearic acid. The reproducibility for all three runs is reasonable. Fig. 7 shows the recovery plot for the chalk plug treated with CSA= 1,000 ppm. The induction time is approximately 2,200 minutes for Runs Sa and 5b. The imbibition rate is very low, and the final oil recovery is 2 to 4% (OOIP). This chalk plug is intermediate-wet. All the chalk plugs are watertlooded at a high rate of 2 cm’lmin to residual oil saturation after the spontaneous imbibition tests. The total oil recovery by spontaneous imbibition and waterflooding is 69.6% for cS,4=200 ppm, 71.2% for Cs~=500ppm, and 67% for CSA=1,000 ppm. Therefore, I,, for these chalk plugs is as follows: 0.74 for CS,4=200 ppm, 0.66 for Cs~=500ppm, and 0.05 for CSA=1,000 ppm. Note that the water-wetting of the chalk decreases systematically with an increase in stearic acid concentration.
Water Injection. We divide the results for water injection into two parts (Part 1 and Part 2) based on the configurations in Fig. 1. Part I-Configuration A. For Configuration A, a core plug is placed inside a horizontal coreholder; there is no fracture space between the core and the coreholder. Water is injected from one end at a constant inlet pressure and the oil (and water) is (are) produced from the other end at atmospheric pressure. Effect of Pressure Gradient. The results presented in Fig. 8 are for a strongly water-wet plug with S,i=0. The simple solid curve is for spontaneous imbibition of a strongly water-wet plug and is presented here as a reference (note that this curve is shown in all figures for Configuration A). The results suggest that the increase in the pressure gradient from 0.09 to 0.37 psilcm does not affect the oil-production rate. A further increase in pressure gradient from 0.37 to 12.9 psilcm, however, increases the oil-production rate. At 80
s
70
&
8
c
S,=O law=l.o - Ap=O -+ Ap=O.O9 psi/crn -D- Ap=0.37 psi/cm t Ap=0.93 psilcm x Ap=12.9 psilcm
$ 2
40
30
0
2-
20
f
10
ow 0
I
200
400
600
800 1,000 1,200 1,400 1,600 1,800 2,000
Water-Injection Time, minutes
Fig. 8-Effect of pressure gradient on oil recovery by water injection: strongly water-wet, Configuration A.
a pressure gradient of 12.9 psi/cm, it takes less than 100 minutes to produce all the recoverable oil. The final oil recovery is not influenced by pressure gradients; it is 66% (OOIP). After water breakthrough, very little oil is produced, even at a pressure gradient of 12.9 p d c m . This result is consistent with the observations by Terez and Firoozabadi.*l The increase in pressure gradient in the 0.09 to 12.9 psilcm range does not affect residual oil saturation for a strongly water-wet chalk. Fig. 9 shows recovery data for a weakly water-wet plug treated with 500 ppm stearic acid solution; 1,,=0.55. The results show that an increase in the pressure gradient affects oil production significantly. The induction time decreases from approximately 300 minutes to zero as the pressure gradient increases from zero to 13.5 psi/cm. At a pressure gradient of 13.5 psi/cm, the recovery is even higher than that for the strongly water-wet plug; the final oil recovery is approximately 72% (OOIP) at Ap= 13.5 psi/cm. Fig. 10 shows the results for the chalk plug treated with 1,000 ppm stearic acid solution; I,,=0.09, and the induction time is approximately 20,000 minutes for the spontaneous imbibition test. When the pressure gradient increases to 0.96 psi/cm, the oil-recovery efficiency is not influenced significantly. Further increase in the pressure gradient to 1.92 psi/cm results in a significant increase in the oil-production rate. When the pressure gradient is greater than 1.92 psi/cm, the oil recovery systematically increases with an increase in pressure gradient. The final oil recovery for this plug at a pressure gradient of 13.5 psilcm is approximately 78% (OOIP). The results presented in Figs. 9 and 10 demonstrate that the pressure gradient may affect oil recovery significantly for weakly water-wet and intermediate-wet chalks. In all the tests, the total liquid-production rate (oil and water) after water breakthrough is less than that before breakthrough. Effect of Wettubility. Three chalk plugs with different wettabilities (I,,=0.09,0.55, and 0.82) are used for water-injection tests at I
I +
~ p = (spontaneous 0 i Ap=0.96 psilcm
80
*
* -
50
50
100 I 0-
60
60
s
I
70
5
80
80 -
60
s> 40
-
c
x + -0-
+
-
al
AD=O (sDontaneous imbibition) ~. Ab=0.96 psi/cm Ap=l.92 psilcm Ap=3.85 psilcrn 8~37.69 psilcm Ap=13.5 psilcrn Ap=O(SWW)
40 -
30
al
!E
is
20
20
d
0 0
OW 1
10
100
1,000
10,000
100,000
Water-Injection Time, minutes Fig. 9-Effect of pressure gradient on oil recovery by water injection: weakly water-wet, Configuration A (Cs~=500ppm). 520
-
10 0
1
10 100 1,000 ImbibitionTime, minutes
10,000 100,000
Fig. 10-Effect of pressure gradient on oil recovery by water injection: intermediate-wet, Configuration A (&=I ,000 ppm).
December 2001 SPE Reservoir Evaluation & Engineering
-
4 00
80
0.09
80 -
80 s
i2 B
2
Ap=l3.5 psilcm
CS,
'a,
70
1,000 pprn
+to55
500ppm
4-0.82
200 ppm
s
0 60 0 $ 50
60 -
6 0
40
>
8 2 -
40 -
20 -
i3
30 20
10
0
OW
0
0
1
10
100
1,000
10,000
200
100,000
400
600
800
1,000
1,200
Water-Injection Time, minutes
Imbibition Time, minutes Fig. 11-Effect of pressure gradient on oil recovery by water injection for three wettability states: Configuration A.
Fig. 12-Effect of initial water saturation on oil recovery by water injection: strongly water-wet, Configuration A.
two pressure gradients: Ap=0.96 and 13.5 psi/cm. The initial water saturation is zero. Fig. 11 indicates that the effect of wettability on oil recovery by water injection depends on the applied pressure gradient. For a small pressure gradient (Ap=0.96 psikm), the oil-recovery efficiency is strongly influenced by wettability; it increases systematically from 16 to 60% (OOIP) with an increase in I,,, from 0.09 to 0.82. However, for a large pressure gradient (Ap= 13.5 psikm), the effect of wettability on oil-recovery efficiency is small and opposite; the final oil recovery is 77% (OOIP) for 1,,.=0.09, 72% (OOIP) for 1,,.=0.55, and 68% (OOIP) for I,,, =0.82. Fig. 11 reveals that the pressure gradient (that is, negative E.) has the most influence on the least water-wet plug. One may conclude that for a mixed-wet chalk, the effect of wettability on oil recovery is strongly dependent on the pressure gradient. An increase in the pressure gradient reduces the effect of wettability on oil-recovery efficiency. Effect of Initial Water Saturation. Strongly water-wet (I,,,.= I .O), weakly water-wet (1,,,.=0.55), and intermediate-wet (I,,.=0.09) plugs are used to study the effect of initial water saturation. The pressure gradient for the tests is 0.96 psikm. For the strongly water-wet plug, the tests are conducted at S,;=0, 10, 21, 32, 33, 38, and 45%. For the weakly water-wet and intermediately water-wet plugs, tests are conducted at S,i=0, 10, and approximately 20%. To examine reproducibility, the test at S , , = 10% is repeated after the test at SW;=19%. The results presented in Fig. 12 are for the strongly water-wet plug. An increase in initial water saturation from 0 to 21% does not lead to obvious changes in oil-recovery performance. The induction time is zero, and all the recoverable oil is produced in less than 120 minutes. The final oil recovery is apprpoximately 67% (OOIP). As initial water saturation increases from 21 to 45%, the
oil recovery decreases systematically. The final oil recovery is 64% (OOIP) for S,i=33%, 59.5% for S,i=38%, and 52% (OOIP) fot S,i=45%. This result is consistent with those reported by Skauge et al.I6 and Viksund et ai." Fig. 13 shows the recovery data for the chalk plug treated with 500 ppm stearic acid solution. For S,,=O, there is a long induction time (&,d=300 minutes) before water begins to imbibe into the chalk plug, even at a pressure gradient of 0.96 psi/cm. The final oil recovery is approximately 38% (OOIP): for Sw,i=IO%, the induction time decreases to approximately 40 minutes, and the final oil recovery increases to about 62% (OOIP); for S,;=20%, the induction time is approximately 15 minutes, and the oil-production rate is similar to that for strongly wdter-wet. Fig. 14 shows the recovery performance for an intermediate-wet plug. A$ the initial water saturation increases from 0 to 19%, the induction time decreases from 20,000 to 200 minutes, and the final oil recovery increases from 7 to 57% (OOIP). The reproducibility of recovery performance for the tests at S,i= 10% before and after the test at Sw;= 19% is another indication of stable wettability (see Fig. 14). The results in Fig. 14 show that the initial water saturation in a mixed-wet chalk may have a drastic effect on recovery. Part 2-Configuration B. For Configuration B, the coreholder is positioned vertically; water is injected from the bottom, and oil is produced from the top. There are vertical fractures around the stacked cores. All the water-injection tests are performed at a con+ stant injection rate. I Effect qf Water-Injection Rate. Three tests are performed at 5, 11, and 27 cm3/min (4.1, 8.2, and 22.3 PVlday). Because of the fractures, the pressure gradients (excluding gravity) across the composite chalks are 0.01, 0.014, and 0.02 psi/cm, respectively (the gravity effect is excluded). Fig. 15 shows the oil recovery vs. time for the strongly water-wet composite system. The results indi-
80 I
sww (S,=O)
70
n
5
60
>
80
70
-0
I
1
t
- sww (Sw,=0)
+ S,=O
*
0
s
I
50
c S,=19%
ApO.96 psilcm
40
/,,=0.09 Ap=0.96 psilcrn
30
5 i5
I
20 10
0
1
lo 0
0
1
10
100
1,000
10,000
100,000
Water-Injection Time, minutes Fig. 13-Effect of initial water saturation on oil recovery by water injection: weakly water-wet, Contiguration A (CSA=500 ppm).
December 200 I SPE Reservoir Evaluation & Engineering
0
TB
----I
1
-
10
100
1,000
10,000
100,000
Water-Injection Time, minutes Fig. 14-Effect of initial water saturation on oil recovery by water injection: intermediate-wet, Configuration A (CSA=1,000 ppm). 521
70
I
I
60
60
n
5 50
'
4
Water injection was stopped for 14 hours
0 50 0
'$
0
40
!2 30
t q,=5
$ cr"-
crn3/rnin,Ap=O.OIO psi/cm
qp11 crnYrnin, Ap=0.014 psi/crn q3:=27crnYrnin, Ap=O.O20 psi/crn
20
5
40 - S W
30
(Ap=0.025psilcm, 9=lO c m h i n )
+Ap=O.O25 psilcrn, q=10 cmYmin +Ap=O.O46 psilcm, q=30 cm3/min
20
*Ap=O.lOl
is
psilcm, 9=68 cmVmin
10
10
I
0
0
500
1,000
1,500
2,000
2,500
3,000
5,000
0
3,500
10,000
1
15,000
I
I
20,000
25,000
I 30,000
Water-Injection Time, minutes
Water-Injection Time, minutes Fig. 15-Effect of pressure gradient on oil recovery by water injection: strongly water-wet, Configuration 6.
Fig. 16-Effect of pressure gradient on oil recovery by water injection: weakly water-wet, Configuration B (CSA=500 pprn).
cate that an increase in water-injection rate from 5.0 to 26.2 cm'/min does not lead to an appreciable change in oil-production rate and final oil recovery. The final oil recovery is approximately 66% (OOIP) for all three injixtion rates. However, the breakthrough (BT) oil recovery decreases systematically with the increase in injection rate. It is 42. I % for ql =5, 25.1% for q2= 1 1, and 12.4% for q3=27 cm-'/niin (41, q 2 , and qs=the injection rates). The results in Fig. 15 suggest that the final oil recovery for the strongly water-wet rock of Configuration B (fractured chalk) is independent of rate. This is consistent with the results for the strongly water-wet rock of Configuration A. Fig. 16 presents the results for the weakly water-wet composite system treated with 500 ppm stearic acid solution using the procedure described earlier. For these tests, the fracture aperture is reduced from 250 to 125 pm to increase pressure gradient. The pressure gradients after breakthrough are 0.025, 0.046, and 0.101 psi/cm (corresponding to water-injection rates of 10.0, 30.0, and 68.0 cm3/min, respectively). Because the oil-production rate is very low, the water is injected for 10 hours and then halted for 14 hours. Each water-injection test lasts approximately 19 days. The discontinuity of the recovery curves in Fig. 16 (and in Figs. 15,17, and 18) is caused by the halting of water injection. The results show that the oil-production rate after water breakthrough is low for the weakly water-wet Configuration B. Before wettability alteration (that is, the SWW state), it takes only 800 minutes to reach the final oil recovery (66% OOIP) at Ap=O.025 psi/cm. However, it takes appproximately 15,000 iminutes to reach the final oil recovery (48% OOIP) at the same pressure gradient for the weakly water-wet system of Configuration B.
Increase in the pressure gradient from 0.025 to 0.101 psi/cm does not affect the oil-production rate at the early time. However, the oil-production rate during the later time and the final oil recovery increase with the increase in pressure gradient. The final oil recovery is approximately 60% (OOIP) for Ap=O.lOI psi/cm, 53% (OOIP) for Ap=0.046 psi/cm, and 48% (OOIP) for Ap=0.025 psi/cm. These results demonstrate that the contribution from the negative P, could lead to improvement in oil recovery by water injection in some fractured rocks. The pressure gradient establishes the effect of gravity (that is, the negative P,. effect) which can lead to appreciable increased recovery. Effect of Initial Water Saturation. Water-injection tests are conducted at different initial water saturations with the composite system before and after wettability alteration. Fig. 17 presents the results for the strongly water-wet composite chalk. The established initial water saturation vanes from 0 to 36.8%. The water-injection rate is kept at 27 cm3/min (22.3 PV/day). The results show that when the initial water saturation increases from 0 to 36.8%, the breakthrough oil recovery is nearly the same, with a variation from 12.4 to 13.9% (OOIP). The oil-production rate after water break+ through and the final oil recovery decrease systematically with an increase in initial water saturation. This result is consistent with that obtained for the strongly water-wet system of Configuration A (see Fig. 12). The effect of initial water saturation on oil recovery for the weakly water-wet composite system (treated with 500 pprn stearic acid solution) is presented in Fig. 18. The established S,i is varied from 0 to 20%, and the water-injection rate is 68 cm'/d (the corresponding pressure gradient is 0.1 psi/cm after breakthrough). Watetr
60
50 40
30
20
4-
*
10
++
law=l.o S,=O S,=13.9% S,=21.2% sWi=36.a%
0
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
Water-Injection Time, minutes Fig. 17-Effect of initial water saturation on oil recovery by water injection: strongly water-wet, Configuration 6. s22
December 200 I SPE Reservoir Evaluation & Engineering
\
'
70 60
We also make the assumption that at equilibrium, the saturation I I is uniform across the core. With this assumption, Eq. 3 can be then
50
40 t S,,,=20%
30
A ~ = O . Ipsi/cm
20 10
0
5,000
10,000
15,000
20,000
25,000
30,000
Water-Injection Time, minutes Fig. 18-Effect of initial water saturation on oil recovery by water injection: weakly water-wet, Configuration B (CSA=5OO ppm).
injection is halted for 14 hours after each 10-hour water injection. Each test lasts approximately 20 days. The results show that the oil-production rate systematically increases with the increase in initial water saturation, although the final oil recovery is not influenced significantly. The time to reach the residual oil saturation is approximately 8,000 minutes for S,,,=20%, 15,000 minutes for S,.,= IO'%, and 24,000 minutes for S,,.,=O.This result is consistent with that observed for the weakly water-wet condition of Configuration A (see Fig. 13). The effect of initial water saturation on the oil-recovery efficiency for the weakly water-wet composite system is opposite to that of the strongly water-wet condition. The result is consistent to that in Prudhoe Bay, as reported by Jerauld and RathmelL4
Capillary Pressure. The coreflooding results under imposed pressure gradients can be used to estimate capillary pressure. The water/oil capillary pressure can be defined as
e
(e.un)
..
=p,,-p,,,
...................................
(1)
where p. and p. =the oil-phase and water-phase pressures, respectively. In the coreflooding, when the inlet pressure is kept constant, only p,, at the inlet remains constant. At equilibrium, when there is no further oil production, the oil-pressure drop across the core is zero. At the core outlet, the gadoil capillary pressure is given by
used to estimate &. For this purpose at each R,, after equilibrium is established, the saturation is measured by weighing the core. Fig. 19 depicts the estimated capillary pressure curves obtained for the chalk plugs with different wettability states. Note that there is no change in saturation for the strongly water-wet condition up to P,,,= - 70 psi. On the other hand, there is a substantial effect of capillary pressure on saturation for the less water-wet conditions to &,= - 20 psi. For LO< - 20 psi, the saturation becomes nearly independent of the wettability state. Note that Fig. 19 is based on the fact that owing to the pressure gradient, one may create a condition for the gravity effect. Also note that the oil recovery from t k bottom of an oil column of 120 ft with a density dfference between water and oil of 0.3 g/cm3 is independent of wettability with data from Fig. 19. The negative side of the capillary pressure in water/oil systems in the laboratory can be created by gravity forces. Immersion of an oil-saturated core in water in a centrifuge test is an example. An alternative is the use of pressure gradients, which is the basis of Fig. 19 in this work.
Discussion and Conclusions The spontaneous-imbibition data and results from the flooding test, with and without wettability alteration for the strongly water-wet and mixed-wet cores, have very different features. First, there is an induction time (that is, delay time) for water imbibition in mixedwet cores. This feature, which is often exhibited in nucleation processes, cannot be simulated with a reservoir simulator. The recovery data for mixed-wet rock also show a low imbibition rate in early time, followed by a high imbibition rate and then a low rate toward the end. In contrast, in strongly wet media, the initial imbibition rate is fast. To the best of our knowledge, no attempt has yet been made to model spontaneous imbibition in mixed-wet media either by scaling or by pore-network modeling. Despite the complexity of imbibition in mixed-wet media, practical implications from all the experiments in Configurations A and B lead to the belief that for some fractured reservoirs, the watecinjection performance can be independent of the wettability state. In such reservoirs, the negative sides of the capillary pressure and the capillary continuity2* establish conditions for efficient waterinjection processes. The main conclusions drawn from this work are: 1. With an increased pressure gradient, the oil-recovery efficiency can increase substantially in a mixed-wet chalk. The same behavior also can be expected in a mixed-wet fractured poroub medium (owing to negative 2. The effect of initial water saturation on oil recovery depends on wettability. For a strongly water-wet condition, oil recovery by water injection can decrease mildly with an increase in initial water saturation. However, for weakly water-wetting conditions, the oil recovery by water injection can increase significantly with an increase in initial water saturation.
(eso)
e p,,=pB-p,, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
(2)
wherep,=gas-phase pressure. We assume no gas flow in the core and, therefore, pS=p,=O. At equilibrium, p. (at inlet)=p, (at outlet)=O, which provides the expression for e.,vc, at the inlet:
e.,.,=-p,, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
f
-30-
In
ti2
-=m
P
'k
+- /,=0.09 . L
-40-
+-
/,,=0.55 /aw=l0
-50-
-60-
0 -70y y
, 0
0.2
(3)
0.6 S,, fraction
0.4
0.8
1
Fig. 1%Estimated capillary pressure curves from water injection: Configuration A.
December 200 I SPE Reservoir Evaluation & Engineering
e.).
Nomenclature C , ~= A concentration of stearic acid, ppm d = diameter, cm Iaw = Amott index to water, fraction k = permeability, md or d L = length, cm p , = gas-phase pressure, psi pi. = injection pressure, psi p 0 = oil-phase pressure, psi po(ln~et) = oil-phase pressure at the inlet of the core, psi pn(our~et) = oil-phase pressure at the outlet of the core, psi ph. = water-phase pressure, psi pw(inlet) = water-phase pressure at the inlet of the core, psi pw,(outlet) = water-phase pressure at the outlet of the core, psi P,,,, = capillary pressure between gas and oil, psi E.,.,, = capillary pressure between oil and water, psi q = water-injection rate, cm'/min (or PV/day) 523
* I
R,,,, = oil recovery by spontaneous imbibition, % (OOIP) R,f = oil recovery by water injection, 3’6 (OOIP) S,., = initial water saturation, % tlnd = induction time, min T = temperature, “C
Ap hpb,
=
pressure gradient, psi/cm
= pressure gradient after water breakthrough, p d c m
@ = porosity, fraction
Acknowledgments This work was supported by the U.S. DOE grant DE-FG2296BC I4850 and the members of the Reservoir Engineering Research Inst. (RERI). Their support is appreciated. We thank R. Jahanian for his assistance in all the experimental work. References I . Bartley, F.E. and Miller, F.I.: “Degree of Wetting of Silica by Crude Petroleum Oils,” Ind. and Eng. Chem. (1928) 20, No. 7, 738. 2. Anderson, W.G.: “Wettability Literature Survey-Part 6: The Effects of Wettability on Waterflooding,” JPT (December 1987) 1605. 3. Morrow, N.R.: “Wettability and Its Effect on Oil Recovery,” JPT (December 1990) 1476; Trans., AIME, 289. 4. Jerauld. G.R.: “Wettability and Relative Permeability of Prudhoe Bay: A Case Study in Mixed-Wet Recervoirs” paper SPE 29576 presented at the 1994 SPE Annual Technical Conference and Exhibition, New Orleans, 25--28 September. 5. Fatt, I. and Klikoff, W.A. Jr.. “Effect of Fractional Wettahility on Multiphase Flow Through Porous Media,’’ Trans., AIME (1959) 216,426. 6. Salathiel, R.A.: “Oil Recovery by Surface Film Drainage in MixedWettability Rocks,” JPT (October 1973) I216; Trans., AIME, 255. 7. Cuiec, L.: “RocWCrude-Oil Inlteractions and Wettability: An Attempt To Understand Their Interrelation,” paper SPE 13211 presented at the I984 SPE Annual Technical Conference and Exhibition, Houston, 16-19 September. 8. Buckley, J., Liu, Y., and Monsterleet, S.: “Mechanisms of Wetting Alteration by Crude Oils,” paper 37230 presented at the 1997 SPE International Symposium on Oilfield Chemstry, Houston, 18-21 February. 9. Sylte, J.E., Hallenheck, L.D., and Thomas, L.K.: “Ekofisk Formation Pilot Waterflood,” paper SPE 113276 presented at the 1988 SPE Annual Technical Conference and Exhibition, Houston, 2-5 October. 10. Hermansen, H. ut ul.: “Twenty-Five Years of Ekofisk Reservoir Management,” paper SPE 38927 presented at the 1997 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 5-8 October. 1 1 . Hamon, G.: “Oil/Water Gravity Drainage Mechanisms in Oil-Wet Fractured Reservoirs,” paper SPE 18366 presented at the 1988 SPE European Petroleum Conference, London, 16-19 October. 12. Putra, E. Fidra, Y.. and Schechter, D.S.: “Use of Experimental and Simulation Results for Ertiniating Critical and Optimum Water Injection Rates in Naturally Fractured Reservoirs,” paper SPE 5643 1 presented at the 1999 SPE Annual Technical Conference and Exhibition, Houston, 3-6 October. 13. Zhou, X. er al.: “The Effect of Crude Oil Aging Time and Temperature on the Rate of Water Imbibition and Long-Term Recovery by Imbibition,” SPEFE (December 1995) 259.
524
’
14. Graue, A,. Viksund, B.G., and Baldwin, B.A.: “Reproduce Wettability Alteration of Low-Permeable Outcrop Chalk,” paper SPE 39622 presented at the 1998 SPEDOE Improved Oil Recovery Symposium, Tulsa, 19-22 April. 15. Brown, W.O.: “The Mobility of Connate Water During a Water Flood,” paper SPE 694-G presented at the 1953 SPE Annual Meeting, 14-17 October. 16. Skauge, A. et a!.: “Influence of Connate Water on Oil Recovery by Gravity Drainage,” paper SPE/DOE 27817 presented at the 1994 SPEDOE Symposium on Improved Oil Recovery, Tulsa, 17-20 April. 17. Viksund, B.G et al.: “Initial Water Saturation and Oil Recovery from Chalk and Sandstone by Spontaneous Imbibition,” Proc., Intl. Symposium of the Soc. of Core Analysts, The Hague (1998). 8. Narahara, G.M., Pozzl, A.L. Jr., and Blackshear, T.H. Jr.: “Effect of Connate Water on Gas/Oil Relative Permeabilities for Water-Wet and Mixed-Wet Berea Rock,” SPE Advanced Technology Series, Richardson, Texas (1993) 1, No. 2, 114. 9. Zhou, X., Morrow, N.R., and Ma, S.: “Interrelationship of Wettability, Initial Water Saturation, Aging Time. and Oil Recovery by Spontaneous Imbibition and Waterflooding,” paper SPE 35436 presented at the 1996 SPE/DOE Improved Oil Recovery Symposium, Tulsa, 21-24 April. 20. Amott, E.: “Observations Relating to the Wettability of Porous Rock,” Trans., AIME (1959) 216, 156. 21. Terez, LE. and Firoozabadi, A,: “Water Injection in Water-Wet Fractured Porous Media: Experiments and a New Model with Modified Buckley-Leverett Theory,” SPEJ (June 1999) 134. 22. Firoozabadi, A. and Hauge, J.: “Capillary Pressure in Fractured Porous Media,” JPT (June 1990) 784; Trans., AIME, 289.
Guo-Qing Tang is a research associate in the Dept. of Petroleum Engineering at Stanford U. His current research interests include multiphase flow in porous media, solution gas-drive in heavy-oil reservoirs, and thermal oil recovery from low-permeable heavy-oil reservoirs. Previously, Tang was head of the EOR Research Center of Dagang Oilfield (Group) Ltd., CNPC, China, and a research engineer at the Reservoir Engineering Research Inst. (RERI) in Palo Alto; California. He holds a BS degree in physical chemistry from the Inst. of Chemical Engineering, East China, Shanghai, China, an MS degree in petroleum geology from Northwestern U,, Xian, China, and MS and PhD degrees in petroleum engineering from the U. of Wyoming. Abbas Firoozabadi is a senior scientist and director at the Reservoir Engineering Research Inst. (RERI) in Palo Alto, California, and a visiting professor at Imperial College in London. His researckl interests include multiphase flow in porous media and thermodynamics of hydrocarbon reservoirs and production. He is the recipient of the 2000 SPE Reservoir Engineering Award and the 2002 SPE Anthony F. Lucas Gold Medal. Firoozabadl holds a BS degree from the Abadan Inst. of Technology (Abadan, Iran) and MS and PhD degrees from the Illinois Inst. of Technology, all in gas engineering. He is currently serving on the SPE Editorial Review Committee and has formerly served on the forum Steering and Western Regional Conference Committees.
December 200 I SPE Reservoir Evaluation & Engineering
: ./
