Professional Documents
Culture Documents
Spe 12170 Pa
Spe 12170 Pa
Summary Introduction
One primary goal of any enhanced recovery project is to As world oil reserves dwindle, the need to develop EOR
maximize the ability of the fluids to flow through a porous techniques to maximize recovery is of great importance.
medium (Le., the reservoir). This paper discusses the Methods such as chemical flooding, miscible flooding,
effect of capillary number, a dimensionless group de- and thermal recovery involve altering the mobility and/or
scribing the ratio of viscous to capillary forces, on two- the 1FT between the displacing and the displaced fluids.
phase (oil-water) relative permeability curves. Specifical- Recovery efficiency was found to be dependent on the
ly, a series of steady-state relative permeability measure- capillary number, defined as
ments were carried out to determine whether the capillary
number causes changes in the two-phase permeabilities p.v
or whether one of its constituents, such as flow velocity, Nc =- ................................ (1)
fluid viscosity, or interfacial tension (1FT), is the con- 'Yep
trolling variable.
For the core tests, run in fired Berea sandstone, a Soltrol The viscous forces were defined as the fluid viscosity,
170™ oil/calcium chloride (CaCI 2) brine/isopropyl flow velocity, and the flow path length. Capillary forces
alcohol (IPA)/glycerin system was used. Alcohol was the vary with the fluid 1FT and the pore geometry of the
1FT reducer and glycerin was the wetting-phase medium.!
viscosifier. Taber defined the capillary number in terms of the
The nonwetting-phase (oil) relative permeability showed pressure drop between two points, the flow length, and
little correlation with the capillary number. As 1FT the IFT.2
decreased below 5.50 dyne/cm [5.50 N/m], the oil
permeability increased dramatically. Conversely, as the IIp
water viscosity increased, the oil demonstrated less ability N c = - . ................................ (2)
to flow. For the wetting-phase (water) relative perme- L'Y
ability, the opposite capillary number effect was shown.
For both the tension decrease and the viscosity increase He concluded that as this ratio increased to a value of 5
(i.e., a capillary number increase) the water permeability psi/ft/dyne/cm [0.2 kPa/m/N/m] the ROS was reduced
increased. However, the water increase was not as great significantly. By decreasing the 1FT by using surface-
as the increase in the oil curves with an 1FT decrease. active agents, or by decreasing the path length by alter-
No velocity effects were noted within the range studied. ing the field geometry, the capillary number could be in-
Other properties relating to relative permeabilities were creased.
also investigated. Both the residual oil saturation (ROS) Others have shown similar results. Melrose and Brand-
and the imbibition-drainage hysteresis were found to ner,3 for example, indicated that as the capillary number
decrease with an increase in the capillary number. The rose to a value of 10 -4, the microscopic displacement
irreducible water saturation was a function of 1FT ten- efficiency, which accounts for the residual saturations to
sion only. both oil and water, increased. The effects of the capillary
A relative permeability model was developed from the number on the recovery of residual oil 'are given by Chat-
experimental data, based on fluid saturations, 1FT, fluid zis and Morrow 4 and by other authors 5 (Fig. 1).
viscosities, and the residual saturations, by using regres- Few studies, however, have shown the effect of the
sion analysis. Both phases were modeled for both the im- capillary number on the two-phase flow between the
bibition and the drainage processes. These models residuals. The variables within this group have been
demonstrated similar or better fits with experimental data researched, but their combined effect on relative
of other water- and oil-wet systems, when compared with permeabilities has been largely ignored.
existing relative permeability models. The applicability Several authors have noted that the viscosity ratio of
of these regression models was tested with the aid of a oil and water alters the oil relative permeability but has
two-phase reservoir simulator. little effect on that of water. 6-8 Few or no changes by
fluid flow velocity were observed, provided that no
Copyright 1985 SOCiety of Petroleum Engineers boundary effects were present during the core tests. 9-11
FEBRUARY 1985 249
alumina, and stainless steel rather than reservoir-type rock
Wagner & Leach samples. 16
1.0 I-~~~S{- -'""-----....... , Also, with the increased use of mathematical reservoir
'C simulators to predict recovery from different EOR proc-
'%\.,- esses, the need to model the various flow properties, such
'i \& as relative permeability, becomes important. Thus, ex-
o \~.
'"~ \9"
perimental results need to be applied to empirical or
0.5 \1:9 statistical models for use in numerical simulators.
'" \~
\:> Materials and Experimental Procedure
f;::,, For this study, a series of steady-state relative permeability
- - - Nonwetting Residual ,
- - - Wetting Residual
" tests were carried out at 77°F [25°C] in 2-fi- [0.6-m- ]
O~~L-~~~---L~~--~~~--~~~ long, 2-in.- [5-cm- ] diameter Berea sandstone cores fired
10.8 10 0
Capillory Number, Nc
at 1,832°F [1000°C]. The cores were baked to prevent
large decreases in their absolute permeability because of
Fig. 1-Recovery of residual oil vs. capillary number.4 clay swelling during and between the experimental runs.
Cores were reused for several tests before being discard-
ed to minimize differences in porosity, permeability, and
lithology. The properties of the cores for each test are
Studies on oil and gas permeabilities showed that as the given in Table 1. The fluids used were mixtures of Soltrol
1FT decreased by increasing the temperature and the 170 oil, * 2% CaCl 2 brine, IPA, and glycerin, as shown
equilibrium pressure between two phases, the relative in Table 2. Later, core measurements were taken using
permeability curves increased and straightened out. 12 The Bradford crude oil, ** Kendex 0837™ , t and 2 % sodium
results of other tests on reducing the tension between oil chloride (NaCl) brine.
and water indicated that (1) few or no relative permeability 1FT effects were studied with the Soltrol170, IPA, and
alterations occurred for tension above 0.1 dyne/cm [0.1 CaCl 2 brine system. Alcohol acted as the tension
N/m]; however, larger increases were observed below 0.1 reducer. CaCl 2 brine was used instead of NaCI brine
dyne/cm [0.1 N/m] for both phases; (2) the curves tend- because it does not form a precipitate with Soltrol and
ed toward linearity; (3) the imbibition-drainage hysteresis IPA as does NaCl. 17 Glycerin was used as the viscosi-
lessened; and (4) the residual saturations to both oil and fying agent to increase the aqueous-phase viscosity.:j:
water decreased. 13-15
'Phillips Petroleum Co., Bartlesville, OK.
One study did show relative permeability increases with "Pennzoil Co., Bradford, PA.
increases in the capillary number; however, these ex- tKendall Oil Co., Bradford, PA.
:j:Crookston, R., Ehrlich, R., and Bae, J.: private communication, Gulf R&D Co.,
periments were run using artificial cores of Teflon® , Pittsburgh (Feb. 13, 1981).
Permeability Permeability
Length PV Porosity (Pre·run) (Post-run) Change
Run Core ~ (cm) 3 (%) (md) (md) (%)
1 1* 24 276 22.46 241.3 115.1 -52.3
2 2* 12 148 23.96 198.1 136.0 -31.3
3 3** 24 240 19.42 240.6 115.1 -52.2
4 4 24 280 22.48 325.3 278.4 -14.4
5 4 24 280 22.48 278.4 247.8 -11.0
6 4 24 280 22.48 247.8 251.0 + 1.3
7 4 24 280 22.48 251.0 219.7 -12.5
8 4 24 280 22.48 219.7 219.7 +0.0
9 4 24 280 22.48 219.7 182.1 -17.1
10 4 24 280 22.48 123.2 128.3 +4.1
11 5 24 287 23.05 365.9 408.2 + 11.6
12 5 24 287 23.05 408.2 355.3 -13.0
13 5 24 287 23.05 355.3 353.4 -0.5
14 5 24 287 23.05 353.4 311.1 -12.0
15 5 24 287 23.05 311.1 259.3 -16.7
16 6 24 279 22.40 433.1 433.1 +0.0
17 6 24 279 22.40 433.1 416.4 -3.9
18 6 24 279 22.40 416.4 384.5 -7.7
19 6 24 279 22.40 384.5
20 7 12 146 23.48 531.6
21 8 24 330 26.50 388.1 398.5 +2.7
22 9 24 276 22.16 357.7
23 10* 24 249 20.15 177.3
24 11* 24 235 19.02 279.1
25 12* 24 272 22.01 160.2
·Unfired core .
• ·Core fired at 250·C.
1FT with
2% CaCI 2 Soltrol170 Specific Viscosity Refractive Oleic/Aqueous
Brine Oil JPA Glycerin Gravity at 25°C Index at 25° C
Runs Phase (%) (%) (%) (%) at 25°C (cp) at 25°C (dyne/cm)
1-7,9,
13,21,22 aqueous 100.0 0.0 0.0 0.0 1.007 0.947 1.3385 37.9
8 aqueous 60.0 0.0 40.0 0.0 0.917 2.761 1.3664 5.50
10,11 aqueous 10.9 14.0 75.1 0.0 0.804 2.613 1.3879 0.335
12,16 aqueous 3.9 43.5 52.6 0.0 0.785 2.475 1.3996 0.0389
14 aqueous 40.0 0.0 0.0 60.0 1.156 13.795 1.4231 30.3
15 aqueous 17.5 0.0 0.0 82.5 1.212 128.58 1.4516 29.7
17 aqueous 0.0 9.6 63.5 27.9 0.914 13.636 1.4115 0.454
18 aqueous 0.0 0.1 38.3 61.6 1.069 126.62 1.4418 2.91
19 aqueous 0.0 20.6 62.8 16.6 0.853 6.100 1.4050 0.118
20 aqueous 0.0 0.0 0.0 100.0 1.258 954.00 1.4735 25.3
23,24 aqueous 100.0' 0.0 0.0 0.0 1.009 0.892 1.3356 24.5
25 aqueous 100.0' 0.0 0.0 0.0 1.009 0.892 1.3356 10.8
1-7,9,
12,21,22 oleic 0.0 100.0 0.0 0.0 0.781 2.363 1.4339 37.9
8 oleic 0.2 95.3 4.5 0.0 0.771 2.246 1.4348 5.50
10,11 oleic 0.9 83.2 15.9 0.0 0.774 2.309 1.4277 0.335
12,16 oleic 1.1 71.5 27.4 0.0 0.775 2.352 1.4196 0.0389
14 oleic 0.0 100.0 0.0 0.0 0.781 2.363 1.4339 30.3
15 oleic 0.0 100.0 0.0 0.0 0.781 2.363 1.4339 29.7
17 oleic 0.0 76.1 22.4 1.5 0.785 2.149 1.4207 0.454
18 oleic 0.0 89.7 10.0 0.3 0.781 2.029 1.4264 2.91
19 oleic 0.0 67.0 30.2 2.8 0.791 2.173 1.4168 0.118
20 oleic 0.0 100.0 0.0 0.0 0.781 2.363 1.4335 25.3
23,24 oleic 0.0 100.0" 0.0 0.0 0.814 5.195 1.4504 24.5
25 oleic 0.0 100.0" 0.0 0.0 0.845 11.297 1.4711 10.8
vocuumt:J
I
I
1
I
Core __ J
Fluid
Reservoirs
Sample
r- Collector
Pump I
Ruska
Pump
Manometer Fluid
Reservoir
Fig. 3-Equilibrium phase diagrams for Soltrol 170/lPN20f0 CaCI 2 and Soltrol 170llPNglycerin
systems.
·Oil·Wet System.
System Notes:
S • Soltrol 170 oil.
B - Bradford crude oil.
K - Kendex 0837 oil.
C - 2% (by weight) CaCI 2 brine.
N - 2% (by weight) NaCI brine.
A - Isopropyl alcohol.
G - Glycerin.
Glycerin mixed completely with the brine and was im- filter before entering the core.
miscible with the oil in all proportions. Combined effects Different procedures were used to prepare the fluids
of interfacial tension and viscosity were studied by using for injection depending on the specific fluid system. For
a Soltrol/lPA/glycerin fluid system. viscosity alterations, the glycerin/CaCl 2 brine solutions
The experimental apparatus is shown in Fig. 2. The were mixed separately from the oil. For the 1FT and the
cores were prepared using the Lapp Pulsafeeder ™ combined tension-viscosity systems, a tie line, on the
pump. Relative permeability tests required two Ruska ternary diagram, yielding the desired properties was
constant-displacement pumps, one for each fluid phase. chosen (Fig. 3). A mixture, lying approximately in the
All fluids were flowed through a O.4-micron [O.4-/-tm] center of the tie line, was selected, and the appropriate
252 JOURNAL OF PETROLEUM TECHNOLOGY
Y = 37.9 diem Run 4 Y = 37.9 diem Run 13
v= 40 ft,day --Drainage v = 40 ft/day Drainage
P-w = 0.9468 ep - - --Imbibition P-w = 0.9468 ep - - - - Imbibition
Ne = 3.09 x 10- 6 Ne = 3.02 x 10- 6
1.0 1.0 1.0 1.0
0.9 0.9 0.9 0.9
0.8 0.8 0.8 0.8
Fig. 4-Oil-water relative permeabilities for the base system and reproducibility of measurements.
amounts of each component were mixed. The resulting Kyte and Rapoport,20 a 50-psi [345-kPa] pressure dif-
solution was shaken, then allowed to sit quietly until com- ferential across the core, was required. Thus, an SO-mLlhr
plete phase separation occurred. [SO-cm 3 /h] minimum flow rate at a velocity of 16 fi/D
The Penn State steady-state relative permeability [4.9 mId] was needed for fluids of high 1FT (greater than
method 19 then was run using the desired fluid system. 1.0 dyne/cm [1.0 N/m]), and 200 mLlhr [200 cm 3 Ih] at
The experimental technique involved flowing both phases 40 fi/D [12.2 mId] was needed for low-tension (less than
simultaneously through the core and calculating the ef- 1.0 dyne/cm [1.0 N/m]) systems. 21
fective permeabilities with Darcy's law applied to each
phase. Fluid saturations were determined by material Discussion of Results
balance. Drainage curves were found by moving the A summary of the experimental core tests is shown in
wetting-phase saturation from 100% to its irreducible Table 3. Two control tests were run (Fig. 4) to measure
value. Imbibition curves then were determined by revers- the reproducibility of the procedure, test the viability of
ing the saturation direction from irreducible water to the using fired Berea cores, and establish a comparative base
ROS. for the altered fluid runs. Soltrol 170 and 2 % CaCl 2
To avoid capillary end effects, the criterion given by brine were flowed at a total fluid flow rate of 200 mLlhr
[200 cm 3 /h] in both imbibition and drainage. These Wetting-Phase Viscosity Effects. The wetting-phase
results compared favorably with one another and with viscosity range was from that of CaCl 2 brine (0.947 cp
those for unfired Berea cores. [9.47 X 10 -4 Pa' s)) to that of glycerin (954.0 cp [0.954
Pa·s)). The fluid velocity was 40 ftlD [12 mid] and the
Velocity Effects. The initial variable altered was the fluid 1FT was maintained at approximately 30 dyne/cm [30
flow velocity. The rates ranged from 80 mLlhr [80 N/m]. The relative permeability curves for four viscosity
cm 3 /h] (16 ftlD [4.9 mid)), the minimum rate to avoid values are shown in Fig. 4 (for brine) and in Fig. 7. As
capillary end effects, to 400 mLlhr [400 cm 3 /h] (80 ftlD the wetting phase (aqueous) viscosity increased, its
~
~ 0.4
6 0.3
0.5
0.2
0.1
0
. .J
00
°
0 ~
"------- o +
6
4i
u
10-3
10-5
'"~
.;
. 02
.
o
0.1
o
o 0
0.0 1°6
10- 10- 5 10-4 10- 3 10- 2
Copi Ilory Number
10~
Fig. 10-lmbibition oil and water relative permeabilities as func-
tions of capillary number at 50% water saturation. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Microscopic Displacement Efficiency (Em)
, <>
40
0, - f------- ---- o
g
.c
30
__ 0-
o
o
---
0
0
~_--
• - -_ + tJ,.
-~
'+
results, the microscopic displacement efficiency, defined
as
J 20
e---4S or (bose; v) ~ - -oSwir (bose; 'I) ....................... (3)
----"5.,IY) ..... --"5 w;, IY)
10 ----Sor ("d ~--DSwir (JA,)
• was plotted and compared with the results of Melrose and
Brandner 3 (Fig. 12). Although Em increased at approx-
imately one order of magnitude lower for the experimental
Capillary Number
capillary numbers than those reported, the data fell within
Fig. 11-Experimental residual saturations as functions of the
the same general range.
capillary number. As the capillary number increased, no matter what
variable was being altered, the imbibition-drainage
hysteresis decreased but never disappeared entirely. This
result held true for both the aqueous and oleic phases. 21
Residual Saturation Changes. The effects on the residual
saturations of both phases also were noted. The ROS Relative Permeability Model. To develop a relative
showed a large reduction from 40 % to 0 % for both permeability model based on the experimental results, the
decreases in the 1FT and increases in the wetting-phase Minitab II@23 statistical computation system was used.
viscosity, corresponding to a capillary number increase This system uses regression analysis to determine the best
from 10- 6 to 10- 2 , respectively. As the tension de- coefficients for an equation and the statistical parameters
creased to 5.50 dyne/cm [5.50 N/m], the irreducible water to evaluate the function.
saturation showed very little change, but at the lowest- The following functional forms were found for both im-
tension value of 0.0389 dyne/cm [0.0389 N/m] it de- bibition and drainage. For the oil (nonwetting-phase)
creased back to 32 %. The residual saturations as func- relative permeabilities,
tions of the capillary number are given in Fig. 11, with
the irreducible water saturation showing changes because kro(dr) =AS* (B+CID ")')(/-tw ) D .............. (4)
of tension changes only. As a check on our experimental /-to
So - Sor
kro(im) 1.56878 1.33874 0.09187 0.08528 96.2 511.2 4.13
1-S or
Sw -Swir
k rw1 (dr) 0.70216 1.25579 0.0 -0.074482 97.9 2070.7 4.89
1-S wir
Sw -Sw;r
k rw1 (im) 0.61135 1.25875 0.0 -0.070812 98.0 1704.1 4.92
1-S wir
Sw-Swif
krw2(dr) 0.70340 0.66596 0.0 -0.071513 97.8 1613.8 4.89
1-S w;r
Sw -SWir
K rw2 (im) 0.61135 0.69580 0.0 - 0.068221 97.4 1659.5 4.92
1-S wir
Sor +5.846x10-' +2.96x 10-' +4.62x 10- 2 + 1.8855 x 10- 3 96.6 122.3 5.74
SWif +4.0214 x 10- 1 + 3.976 x 10- 3 - 7.065 X 10-3 0.0 62.4 6.64 4.46
Bradford Crude-NaCI brine system (water-wet) Bradford Crude-NaCI brine system (water-wet)
Il<w = 0:8918 ep 0--0 Experimental Data Il<w =0.8918 ep 0---0 Experimental Data
FIg. 13-Drainage and imbibition relative permeability curves for water-wet system.
,I
I
0.8 0.8 0.8 0.8
0.7
,I-- 0.7 0.7 / 0.7
kro
0.6
,/
0.6
0.5
0.6
0.5 I
/ 0.5
0.5
0.1
0.1 0.1 0.1
0.0
50 60 70 80 90 108.
0 0.00
.10" /...??;;9?
--0 r----
"
20 30 40 50 60 70 80 90
0.0
100
0 10 20 30 40
Sw (%)
Fig. 14-Drainage and imbibition relative permeability curves for oil-wet system.