Sei sulla pagina 1di 12

SPE 84866

Analysis of Oil Recovery by Spontaneous Imbibition of Surfactant Solution


T. Babadagli, SPE, University of Alberta

Copyright 2003, Society of Petroleum Engineers Inc.


In addition to the above-mentioned qualitative analysis, the
This paper was prepared for presentation at the SPE International Improved Oil Recovery recovery curves were evaluated for up-scaling. Existing
Conference in Asia Pacific held in Kuala Lumpur, Malaysia, 2021 October 2003.
dimensionless scaling groups were tested. The scaling
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
exercise helped identify whether the recovery is driven by
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to gravity or capillary forces and clarify the interaction type, i.e.,
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at co-or counter-current or both. The ultimate recoveries were
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
correlated to the Inverse Bond Number using twenty-five
for commercial purposes without the written consent of the Society of Petroleum Engineers is cases covering different combinations of four rock types, four
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous oil and four surfactant samples. It was shown that the
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836 U.S.A., fax 01-972-952-9435.
inclusion of the wettability factor would make the correlation
more universal.
Abstract Introduction
Depending on rock and oil type, lowered interfacial tension Injection of surfactant solution into naturally fractured
(IFT) by the addition of surfactant to brine may contribute to reservoirs (NFR) has gained a great deal of attention in recent
capillary imbibition recovery with the support of gravity years. Although reported field cases are limited, considerable
drainage in naturally fractured reservoirs (NFR). This paper amount of laboratory studies have appeared in literature.
aims at identifying and analyzing the recovery mechanisms Capillary imbibition recovery can be enhanced by the addition
and performing up-scaling exercises for oil recovery from of surfactant into water to reduce the IFT. Depending on the
different rock types by the capillary (spontaneous) imbibition size and wettability characteristics of the reservoir matrix, the
of surfactant solution. gravitational forces might play a role in the matrix recovery
Laboratory tests were performed using four different rock as well.
types that could possibly be the reservoir rock matrix of the Different aspects of capillary imbibition of low IFT
NFRs (sandstone, limestone, dolomitic limestone and chalk). aqueous phase were studied in the past1-13. More research is
The sandstone sample was surface-coated to create a boundary needed to clarify the effects of surfactant addition to the
condition causing only counter-current interaction. Wide aqueous phase on the capillary imbibition recovery rate and
variety of oils (light and heavy-crude oils, kerosene, and ultimate recovery and understand the mechanics of the low
engine oil) was selected as the oleic phase. Different types IFT imbibition. The changes on the rock-fluid and fluid-fluid
(non-ionic and anionic) and concentrations of surfactants were interactions due to surfactant addition could act in different
used as the aqueous phase as well as the brine as a base case. ways in oil recovery by capillary imbibition. It has been
The samples fully saturated with oil (Swi=0) were exposed to observed that the reduction in IFT yields higher ultimate
static capillary imbibition and the recovery was monitored recovery but slower imbibition rate3-6,10 for water wet
against time. Some experiments on the chalks were repeated sandstone samples. Schechter et al.4 observed the same for
using pre-wet samples (Swi >0) to clarify the changes in the iso-octane recovery from limestones whereas Babadagli13,14
capillary imbibition characteristics of the rock. reported that the capillary imbibition of low IFT solution (with
The changes (positive or negative) in the recovery rate and a non-ionic surfactant) was faster than high IFT case when a
ultimate recovery compared to the brine imbibition were light crude oil is used with a limestone sample. Austad et al.9
evaluated for the rock, surfactant and oil types. It was observed the same for chalks with cationic surfactant. Keijzer
observed, for some rock samples, that the imbibition recovery and deVries1 showed that the surfactant imbibition did not
by surfactant solution was strictly controlled by the yield any incremental recovery for water-wet sandstone but
concentration of the surfactant. On the other hand, the some effects were observed on the recovery rate.
difference in the recovery rate and ultimate recovery between In case of unfavorable matrix boundary conditions, that do
high and low IFT could be due not only to change in the IFT not allow interaction from all sides of matrix, surfactant
but also the change in the wettability and adsorption, which solution may yield lower ultimate recovery and lower
might vary with the rock type. This was also analyzed using recovery rate than those of brine10,15. In addition to the
the shape of the curves that indicates the strength of the physical boundary condition, the interaction could be
capillarity on the recovery and the interaction type, i.e., co- or controlled by the strength of the capillary forces that
counter-current. eventually yields co-or counter-current (or both) interaction.
2 SPE 84866

This might eventually affect the ultimate recovery and After coating and saturating the samples, they were
recovery rate4,7. As the capillary forces become stronger, i.e., immersed into an imbibition cell filled with the aqueous phase
strongly water wet rocks or higher IFT, the capillary (brine or surfactant solution) in vertical position and exposed
imbibition dominates the recovery. Once can expect fully to capillary imbibition. The recovery was monitored
counter-current type interaction for this case. As the capillary against time.
forces weaken, i.e., larger matrix block sizes or less water wet
rock or lowered IFT, the effect of the gravity on the Analysis of the results
interaction between matrix and fracture could be felt. That The results, namely the recovery curves, were analyzed
might cause a fully co-current matrix-fracture transfer. qualitatively first. Then, the same data were evaluated against
Wettability alteration by surfactant addition into brine can dimensionless numbers. This allowed understanding the
be expected depending on the rock and surfactant type16-19. scaling behavior as well as the mechanics of the
Incompatibility of the rock with surfactant type might give rise recovery implicitly.
to adsorption7,8. They both play a role on the recovery rate
and ultimate recovery. Surfactant degradation due to factors Qualitative analysis
other than temperature is another possibility as well. Thus, the The recovery curves for all cases are shown in Figs. 1 to 4.
performance of the surfactant solution as an imbibition fluid The evaluations were made for each rock sample.
could be controlled by the rock properties (wettability,
lithology, matrix boundary condition and adsorption) and the Sandstone. Lower IFT resulted in an incremental recovery for
characteristics of surfactant. Therefore, the selection of the all oil samples (Fig. 1). In accordance with the previous
proper surfactant type and amount for different rock and oil experience3-6 where different sandstones were used as rock
types is the key factor. Due to complexities caused by so samples, brine (high IFT) initially shows a faster recovery
many parameters involved in the process, no procedure has so than that of low IFT. This is true for all cases except the
far been suggested in terms of the selection of the surfactant heavy crude oil (Crude Oil-1). The recovery rate turned out to
type and the application of the technique for effective capillary be substantially higher for the surfactant case as well as the
imbibition recovery in NFR. ultimate recovery (Fig. 1-d). Classical Mattax and Kyte23
One of the key factors that affects the rock-oil-surfactant dimensionless group suggests that the recovery rate becomes
solution interaction towards the oil recovery is the existence higher as the IFT increases:
(an the amount) of initial water. Although the effects of initial
k
water on the imbibition of high IFT cases were studies to some td = t (1)
extent20-22, studies on low IFT imbibition is limited. This is wL 2

expected to affect the wettability and adsorption This observation is in disagreement with this equation. The
characteristics of the rock as well. remarkable change in the recovery rate could be attributed to
The objective of this paper is to present an extensive the changes in wettability16-19. Previously, Alveskog et al.16
analysis of the capillary (spontaneous) imbibition of surfactant observed that anionic surfactants change the wettability of
solutions into oil saturated rocks. The results were evaluated sandstones from strongly water-wet to weakly oil-wet.
in terms of oil recovery rate and ultimate recovery for many Standnes and Austad18 reported that the wettability of chalks is
different combinations of rock, oil, and surfactant samples. changed from water-wet to oil-wet and oil-wet to water-wet
Different scaling groups were also applied to clarify the with cationic and anionic surfactants, respectively. In the
scaling behavior explicitly and the mechanics of the experiments given in Fig. 1, a non-ionic surfactant was used.
recovery implicitly. Other three oil cases yielded a faster recovery with brine in
early times and the low IFT imbibition turned out to be faster
Experimental study in late times (Figs. 2-a, b, and c). One can conclude that the
Experiments were conducted on four different rock samples. recovery is not dominated by the capillary forces for both high
Four different types of oil and four different types of and low IFT cases in Fig.1-d when compared to the ones in
surfactants (solutions) were selected as the oleic and aqueous Figs.1-a and c.
phases. The details about the rock and fluid properties are As mentioned earlier, the matrix fracture interaction is
given in Table 1. expected to be in counter-current manner if the rock is
Each sample was used only once to avoid any possible strongly water wet, i.e., the capillary forces dominate the
alteration in the wettability of rock due to cleaning procedure imbibition process4,7,24. The imbibition front progresses
and chemicals. For counter-current experiments, 1.5 in- uniformly towards to the inner part of the core from all sides20.
diameter sandstone samples were cut halfway vertically and As the capillary forces dwindle, the interaction between matrix
coated using epoxy from all sides except the flat vertical and fracture becomes co-current, i.e., water driving oil from
surface. The samples were dried one day at room temperature bottom part in upward direction. This happens in case of long
and fully saturated with oil under vacuum for 24 hours. All matrix4,7 and oil-wet systems24. If, however, the boundaries of
the samples were exposed to oil saturation only 24 hours to the matrix is close to flow and only one side of the matrix is
ensure that no wettability alteration takes place by aging. For open, then the counter-current flow occurs irrespective of the
the samples with initial water, the same procedure was wettability and size of the matrix. Fig. 2 represents this case
repeated with brine. Then the samples were placed into a for two oil types. Figs. 1-a and 2-a illustrate the same cases
Hassler core holder and the brine was displaced by oil until no except the physical boundary conditions. Both cases are
brine is recovered.
SPE 84866 3

expected to yield a very strong capillary imbibition in counter- rate and the ultimate recovery were lower than those of brine.
current manner due to favorable conditions (strongly water The opposite was observed for IFT=1 dyne/cm case. If the
wet, high IFT, low oil viscosity, and high permeability) even initial water exits (irreducible water saturation is around 20%),
though the counter-current interaction for the latter is due to both the brine and surfactant solutions yielded faster recovery
physical boundary conditions. The shape of the curves proves and higher ultimate recovery with increasing surfactant
a counter-current imbibition driven by strong capillary forces. concentration (decreasing IFT). Note that the surfactant
It is a quick start followed by a steep straight line until it levels concentration yielding IFT=1 is close to the CMC value.
off. Low IFT yields higher ultimate recovery for both cases For the non-ionic surfactant (Fig. 4-c), the ultimate
as well. recovery was much higher than that of the brine and the
Fig. 2-b shows the comparison of fully open cylindrical recovery rate was far faster when there was no initial water in
and semi-cylindrical (one side open) cases for crude oil-2. the system. For IFT=5 dyne/cm the experiment was repeated
The recovery curves imply that the imbibition in the with initial water and the recovery was even lower than that of
cylindrical cores with all sides open is not fully dominated by brine (Fig. 4-c). Note that for both cases, the surfactant
the capillary forces. The same can be said for the low IFT concentration values are below the CMC.
case for counter-current flow (semi-cylindrical sample with We observe two different behaviors with two different
one side open to flow). In fact these three cases followed the surfactant types. Different recovery rates and ultimate
same recovery trend and eventually, the low IFT case yielded recoveries were obtained for different surfactant types (Surf. 3
a higher recovery. Note, however, that the high IFT case of and 4). Surf. 4 yielded higher ultimate recovery even with
counter-current flow is strongly capillary dominated lower IFT (5 dyne/cm) than Surf. 3 (1 dyne/cm). It is
interaction and the recovery trend of the recovery curve is expected that both wettability alteration with the addition of
identical to the kerosene case shown in Fig. 2-a. This surfactant and adsorption play a role in this different behavior.
behavior is in accordance with Eq. 1, i.e., the higher the IFT, Note also that the IFT value of 1 dyne/cm for Surf. 3 is close
the faster the recovery. Apparently, totally counter-current to CMC, as mentioned above. This might affect the type and
type transfer did not occur for crude oil (CO-2) recovery from properties the emulsion that eventually play a role on the
the cylindrical cores and low and high IFT did not show any recovery rate.
remarkable difference in terms of the recovery rate (dotted As a final point, a few more comments need to be made in
lines in Fig. 2-b). As opposed to this case, the recovery rates regards to the behavior of surfactant solution imbibition.
and trends are totally different for the high and low IFT cases Surf.3 is an emulsion breaker whereas Surf.4 is commonly
of the semi-cylindrical sample where the interaction is totally used to create a stable microemulsion. The solubility of the
counter-current for the high IFT case. surfactant in oil and water (hydrophilic and hydrophobic
groups) and the type of emulsion, if exists, need to be taken
Limestone. The recovery curves for two types of limestone into account. This is the on-going part of this research.
and two types of crude oil are shown in Fig. 3. It is clear that
the interaction is not capillary dominated since the recoveries Quantitative analysis
are low and the time required for this recovery is very long. Oil recovery from a reservoir matrix in naturally fractured
The cases shown in Fig. 3-b could be exceptional as the reservoirs is characterized by two parameters: (a) ultimate
ultimate recoveries are rather high that could be an indication recovery (or residual oil) and (b) recovery rate. For the
of the stronger capillary forces. For all the cases, on the other quantitative analysis of the results, the evaluations will be
hand, the trend is the same, i.e., lower the IFT, higher the made for these two parameters.
ultimate recovery and faster the recovery. The difference
between Figs. 3-a and b can be attributed to the wettability and Ultimate recovery. The combined effect of gravity and
adsorption characteristics as the two limestones samples differ capillarity would be more helpful to identify the dynamics of
from each other by origin. the capillary imbibition. First analysis was performed for the
ultimate recovery change with the inverse Bond number.
Chalk. The chalk sample exhibited a similar performance as The Bond Number is defined as the ratio of gravity forces (Eq.
the limestone in terms of the recovery trend, i.e., surfactant 7) to the capillary forces (Eq. 1)4,25,26 and usually expressed as
yielded faster and higher ultimate recovery. The case shown the inverse of this (NB-1):
in Fig. 4-a suggests that the interaction is not fully dominated
by capillarity, as similar to the case shown in Fig. 3-a. The f ( )
ultimate recovery is not as high as the case shown in Fig. 3-b k
N B1 = c (2)
but higher than the cases in Figs. 3-a and c. The chalk sample gH
followed a similar recovery trend to the limestone samples but
not to the sandstone. where c=0.4 for capillary tube models and this constant was
More attention was given to chalk case due to existing used for all cases. In this analysis the wettability effect,
EOR potential in fractured chalk reservoirs2,7-9,11,12,16-19. represented by f(), was not considered. In Eq. 2, all the units
Different surfactant types were tested without changing the oil are in metric system, i.e., k is cm2, is dyne/cm, H is cm, and
type. Figs. 4-b and c show the results for two different is g/cc.
surfactant types. Some of the tests were repeated with initial The ultimate recoveries were plotted against NB-1 in Fig. 5.
water as well. Surf. 3 (an anionic surfactant) showed an IFT Previously, similar plots were done for the same sandstones
dependent imbibition. When IFT was 7 dyne/cm, the recovery
4 SPE 84866

with different permeabilities6 and one kind of sandstone and Characteristic lengths were calculated using Eq. 6. For the
limestone4 without also considering the wettability effect. cylindrical samples open to flow from all directions, the Lcs
Here, wide ranges of rock, oil and surfactant types were are 0.68, 0.87 and 0.88 for the samples sizes of 5, 7.5, and 10
included. Typically, the ultimate recovery decreases with cm in length and 2.54 in diameter, respectively. For the
increasing NB-1. Two trends, however, were observed. samples with 10 cm in length and 3.81 cm in diameter, it
Sandstone and limestone sample-2 (LS-2) followed the same is 1.3.
trend, whereas the chalk samples and other limestone (LS-1) Fig. 6 shows the recoverable oil against time (Fig. 6-a) and
followed another trend. Note that the wettability the dimensionless time given in Eq. 5 (Fig. 6-b). If the matrix-
characteristics are not included and these two groups of rocks fracture transfer is dominated by capillary imbibition, it is
in each trend are expected to represent a similar wettability expected that the curves show the same behavior when plotted
characteristics based on the ultimate recovery measures. With against the dimensionless group. All the chalk cases without
the inclusion of the wettability term, f(), the two group of initial water (solid light blue curves) followed the same trend
curves are expected to follow a single trend as suggested and against the time but they fell apart when plotted against the
shown by Babadagli5,15. dimensionless group. This indicates that the transfer is not
predominantly capillary imbibition. On the other hand, the
Recovery rate. The shape of the recovery curve is an curves for sandstone and limestone samples showed two
indication of the interaction type and the contribution of the groupings. One group contains the all curves for sandstone
capillary and gravity forces. In strongly water-wet systems and limestones except the SS/kerosene and LS-1/CO-1 cases
with light oil, the matrix-fracture interaction is expected to be (the right side of the curve bundles). Other group contains
counter-current for short rock samples. In order for the sandstone with kerosene, LS-2 and the chalk cases with initial
gravity to become effective in the transfer, the sample size water (the left side of the curve bundles). The curves on the
(height) should be big enough or the capillary forces should be right side are the weak capillary imbibition cases whereas the
weak. In this case, the interaction becomes fully or partly co- ones on the left side of the bundle are stronger capillary
current. For example, the shapes of the curves in Fig. 1-d are imbibition cases. The shape of the curves also implies this.
different from the ones in Figs. 2-a, b, and c, indicating that Schechter et al.4 suggested that the shape of the curves
the transfer between matrix and fracture is not predominantly becomes concave as the effect of capillarity decreases. Let us
capillary imbibition and not fully counter-current. take the two extreme examples of the curves. Sandstone with
The recovery rates can also be compared using the kerosene represents the fastest recovery rate and it is capillary
dimensionless groups after normalizing them. The dominated imbibition transfer. The recovery is expected to be
normalization is done by dividing the amount recovered by the due totally to counter-current imbibition. Engine oil case,
ultimate recovery. The dimensionless groups can also be used however, is not fully dominated by capillarity due to higher
to identify the recovery mechanism and interaction viscosity and weaker water wettability characteristics. The
type implicitly. recovery is slower and not fully counter-current imbibition as
Cuiec et al.3 proposed the following modification to the implied by the concave shape.
classical Mattax and Kyte23 dimensionless scaling group (Eq. Once the wettability effect is taken into account, the two
1) replacing water viscosity by oil viscosity: groups of curves might show a grouping. The inclusion of
wettability in the formulation is a crucial issue but the effect of
k surfactant on the wettability should be identified first. Contact
td = t (3)
o L 2
angle () is a measure of the wettability and the numerator of
the above equations is multiplied by cos() in order to
Zhang et al.27, on the other hand, combined the oil and water incorporate the wettability effect. But the quantification of the
viscosities as the geometric mean of the viscosity of both wettability through this measurement is difficult task. Thus,
phases and modified Eq. 2 as follows: another practical measure of wettability should replace the
k cos(). Babadagli5,15 suggested a relative measure of
td = t (4) wettability based on capillary imbibition method and
gm L 2
suggested to use a value between 1 and 0 to
represent wettability.
where gm is defined as o w . If, on the other hand, the gravity dominates the interaction
If L is replaced with the characteristics length, Lc, as between matrix and fracture, i.e., IFT is lowered or much
described by Ma et al.28 based on the boundary condition of longer cores are used, the gravity becomes also effective on
the matrix, then Eq. 3 can be rewritten as follows: the recovery. A scaling group for gravity dominated capillary
k imbibition was defined by Du Prey25 as follows.
td = t (5) k
gm L 2
tg = t (7)
L w
c
The characteristic length, Lc was defined as
1 Vb Cuiec et al.3 proposed a modification to this equation and used
Lc = = (6) o instead of w. This is more suitable for our case as
Fs
n
Ai

i =1 l Ai
different oil types were used. Eq. 7 can be written in the
SPE 84866 5

following form to define the dimensionless time for the These plots are helpful in identifying the transfer type.
imbibition process dominated by the gravity forces: Towards the right direction in these plots, the transfer type
k becomes less counter-current dominated (as shown in the
tg = t (8) schematic between the plots). The two cases of sandstone
Lo with engine and heavy oil are good examples of this type of
The recoverable oil was plotted against these two flow. In this case, the shape of the curves becomes more
dimensionless groups in Fig. 7. If the gravity fully dominates concave as an indication of co-current interaction. These
the interaction, the same trend is expected for all the curves. observations verify Schechter et al.s 4 and Austad and
In both cases, there is no such trend when they are compared Millers7 findings. The sandstone with kerosene case is far
to the curves given in Fig. 6-a. Due to the small length of the away from all the other curves. The sandstone case with
cores and a certain degree of water wettability, one can expect kerosene is a good example of counter current flow and in
no significant effect of gravity in the recovery. But as both plots it did not belong to any other groups. This indicates
mentioned earlier, the interaction type can be co-current that the other curves do not represent a fully counter-
driven from the bottom of the core towards the upper part and current interaction.
the counter-current imbibition can be observed only on the top The way the curves were grouped are very similar in Figs.
portion of the core once the capillary forces become weaker. 8-a and b. This indicates that the interactions are not fully
This is a possibility in the cases of heavy-crude oil, limestone counter-current nor co-current, except the sandstone/kerosene
and chalk samples without an initial water, and engine oil with case, which is fully counter-current imbibition. In some
sandstone. Therefore, other dimensionless scaling groups particular cases, a better grouping of the curves can be
suggested by Schechter et al.4 were tested to identify the observed. For example, two sandstone cases with Surf.1/CO-1
interaction type. and Surf.1/engine oil are closer to each other in Fig. 8-b. This
They defined the following dimensionless groups as is an implication of the stronger contribution of co-current
similar to the Eqs.1, 3, 4 and 5. Characteristic time, tc is interaction than counter-current and Eq. 10 is more convenient
defined for capillary dominated matrix transfer than Eq. 9 for the scaling of these types of cases.
R 2 k Zhou et al.29 suggested an improved version of Eq. 5 using
t c1 = t (9) the mobility () as in the Eqs. 9 and 10. This approach was
C not tested in this study due to difficulties in determining the
Characteristic time, tc is defined for gravity driven flow matrix mobility values.
transfer As a final attempt, another approach was tested. Rangel-
German and Kovscek30 recently suggested that the recovery
H 2 k
t c2 = t (10) scale with time for 2-D plane with line source, the square root
C of time for 1-D geometry, and with t3/2 for 3-D geometry.
In fact, if the capillary imbibition takes place in counter- Both t and t3/2 plot did not exhibit and straight line behavior.
current manner in vertically situated cylindrical cores, it is In all cases except LS, a straight line was obtained when the
controlled by the diameter of the core (R, Eq. 9). If the recovery rate (as cc) was plotted against t1/2. Straight line
transfer is not counter-current, the oil is displaced from bottom relationship between the imbibition rate (q) and t1/2 indicates
towards the upper part of the core in co-current manner. that the imbibition progresses in linear manner from a plane
Counter-current imbibition may take place in the upper parts source30,31. In our experiments, the interaction between matrix
of the core depending on the strength of the capillary forces. and surrounding fluid is planar rather than a point. Also note
In this case, the transfer is controlled by the height of the core that the imbibition progresses in linear manner and this could
(H, Eq. 10). be the case for both strongly capillary dominated (counter-
Fig. 8 shows the plots of recoverable oil against these two current flow) and slightly capillary dominated cases (co-
dimensionless groups. The differences among the cases current flow with some counter-current from the top portion of
representing different characteristics are more distinct the core) indicated by low and high NB-1 values, respectively.
compared to Fig. 6-b. There are mainly three groups of
curves. In Fig. 8-a, the sandstone case with kerosene was Conclusions
separated due to it is very fast recovery rate by strong capillary 1. The conditions yielding increased recovery rate and
imbibition. The second group of the curves (bundle of curves ultimate recoveries by the capillary imbibition of
in the middle) consists of all limestones except the heavy oil surfactant solution were identified. Non-ionic surfactant
case (CO-1) and the chalk cases with initial water except solution increased the recovery rate and ultimate recovery
Surf.3 and CO-3. The third group of the curves (the right of heavy-oil in water-wet sandstone compared to the brine
most one) also represents a good grouping except the engine imbibition (high IFT). No such effect was observed for
oil case. Note that in Fig. 8-b the engine and heavy oil the recovery rate of the counter-current imbibition
sandstone cases, which are not expected to have counter- (imbibition taking place only from one side of the core)
current type interaction even though they have capillary for lighter crude oil even though a slight increase in the
imbibition recovery, grouped along with the low IFT chalk ultimate recovery was obtained by the addition
case (Surf. 3-CO-3). These three cases are expected to be the of surfactant
weakest capillary dominated cases. 2. Except light oil cases such as kerosene and light crude oil
in sandstones, in all cases (limestone, sandstone with
heavy-crude and chalk samples) the same surfactant
6 SPE 84866

solution (a non-ionic surfactant) yielded a higher ultimate R = Matrix radius


recovery and faster recovery rate. This can be attributed t = Time
to the change in wettability by addition of this td = Dimensionless time for capillary imbibition
non-ionic surfactant. recovery
3. When an anionic surfactant was used in chalks, low IFT tg = Dimensionless time for imbibition recovery
(higher surfactant concentration) yielded higher recovery governed by gravity forces
than the brine case. Lower surfactant concentration, Vb = Bulk volume of matrix
however, resulted in even lower recovery than the brine = Porosity
case. A non-ionic surfactant, on the other hand, showed = Mobility
very effective recovery regardless the concentration gm = Geometric mean of oil and water viscosities
of surfactant. o= Oil viscosity
4. Having initial water in the system (irreducible water w= Water viscosity
saturation) resulted in faster recovery of light crude oil in = Density
chalks compared to the samples without initial water for
= Interfacial tension
anionic surfactant. The opposite was observed for non-
ionic surfactant. Despite its very effective recovery for
References
the samples without initial water, the recovery rate and 1. Keijzer, P.P.M. and de Vries, A.S.: Imbibition of Surfactant
the ultimate recovery from the chalk sample with initial Solutions, SPE 20222 presented at the 1990 SPE/DOE 7th
water was even lower than those of brine. Symp. on Enhanced Oil Recovery, Tulsa, OK, April, 22-25.
5. A correlation between the ultimate recovery and the 2. Flumerfelt, R.W. et al.: A Cyclic Surfactant-Based
Inverse Bond Number was obtained. As the Inverse Bond Imbibition/Solution Gas Drive Process for Low-Permeability,
Number (Eq. 2) increases the ultimate recovery decreases. Fractured Reservoirs, paper SPE 26373 presented at the 1993
With the inclusion of the wettability, a more general SPE Ann. Tech. Conf. and Exh., Houston, TX, Oct., 3-6.
correlation would be obtained. 3. Cuiec, L., Bourbiaux, B. and Kalaydjian, F.: Oil Recovery by
Imbibition in Low-Permeability Chalk, SPEFE
6. The shape of the recovery curves was found to be a good
(Sept. 1994) 200.
indicator for the strength of the capillary forces in the 4. Schechter, D.S., Zhou, D. and Orr, F.M.: Low IFT Drainage and
imbibition transfer. The dimensionless scaling groups Imbibition, J. Pet. Sci. and Tech., (1994) 11, 283.
were tested for the applicability of them for scaling and to 5. Babadagli, T.: "Temperature Effect on Heavy-Oil Recovery
identify the mechanics of capillary imbibition. A Imbibition in Fractured Reservoirs," J. Pet. Sci. and Tech., (1996)
modified version of the classical Mattax and Kyte group 14, 197.
was tested without considering the wettability effect. No 6. Al-Lawati, S. and Saleh, S.: Oil Recovery in Fractured Oil
scaling behavior was observed. Obviously, the Reservoirs by Low IFT Imbibition Process, paper SPE 36688
interactions were not fully capillary dominated and not in presented at the 1996 SPE Annual Tech. Conf. and Exh., Denver,
CO, Oct. 6-9.
a counter-current manner. The dimensionless group for
7. Austad, T. and Milter, J.: Spontaneous Imbibition of Water Into
the gravity dominated imbibition did not show any scaling Low Permeable Chalk at Different Wettabilities Using
behavior for any cases either. Surfactants, paper SPE 37236 presented at the 1997 SPE Int.
7. Two dimensionless groups suggested by Schechter et al.4 Symp. On Oilfield Chemistry, Houston, TX, Feb. 18-21.
were tested as an alternative to Mattax and Kyte type 8. Milter, J. and Austad, T.: Chemical Flooding of Oil Reservoirs
dimensionless group. A better scaling, suggesting the 6. Evaluation of the Mechanism for Oil Expulsion by
inclusion of the wettability factor, was observed and it Spontaneous Imbibition of Brine With and Without Surfactant in
was shown that these groups could be used to also Water-Wet, Low-Permeable, Chalk Material, Coll. and Surf. A:
distinguish the interaction type, i.e., the strength of the Phys. and Eng. Aspects, (1996) 113, 269.
9. Austad, T. et al..: Chemical Flooding of Oil Reservoirs 8.
capillarity on the imbibition transfer.
Spontaneous Oil Expulsion From Oil- and Water-Wet Low-
Permeable Chalk Material by Imbibition of Aqueous Surfactant
Acknowledgment Solutions, Coll. and Surf. A: Phys. and Eng. Aspects, (1998)
This study was partly funded by an NSERC Grant (No: 137, 117.
G121210595). 10. Babadagli, T., Al-Bemani, A. and Boukadi, F.: Analysis of
Capillary Imbibition Recovery Considering the Simultaneous
Nomenclature Effects of Gravity, Low IFT, and Boundary Conditions, paper
Ai = Area open to imbibition at the ith direction SPE 57321, 1999 SPE Asia Pacific Improved Oil Recovery
c = Constant Conference, Kuala Lumpur, Malaysia, Oct. 25-26.
11. Spinler, E. A.: Enhancement of Oil Recovery Using a Low
C = Constant
Concentration of Surfactant to Improve Spontaneous and Forced
f() = Certain function of wettability/capillarity Imbibition in Chalk, paper SPE 59290 presented at the 2000
Fs = Matrix shape factor SPE/DOE Imp. Oil. Rec. Symp., Tulsa, OK, April, 3-5.
H = Matrix height 12. Chen, H. L. et al.: Laboratory Monitoring of Surfactant
k = Matrix permeability Imbibition Using Computerized Tomography, paper SPE 59006
lAi = Distance from Ai to no-flow boundary. presented at the 2000 SPE Int. Petr. Conf. and Exh., Veracruz,
L = Matrix size Mexico, Feb., 1-3.
Lc= Modified characteristic length. 13. Babadagli, T.: Selection of Proper Enhanced Oil Recovery Fluid
for Efficient Matrix Recovery in Fractured Oil Reservoirs,
n= Total number of surfaces open to imbibition
SPE 84866 7

Colloids and Surface A: Physicochemical and Engineering Table 1. Rock and fluid properties used in the
Aspects, in print, 2003. experiments.
14. Babadagli, T.: Evaluation of EOR Methods for Heavy-Oil
Recovery in Naturally Fractured Reservoirs, J. of Petroleum k Length Diameter
Science and Eng., (Feb. 2003) 37, 25. (md) (cm) (cm)
Rock Types (%)
15. Babadagli, T.: Scaling of Co-Current and Counter-Current
Capillary Imbibition for Surfactant and Polymer Injection in Sandstone (SS) 400 20 10 & 7.5 3.8 & 2.5
Naturally Fractured Reservoirs SPEJ (Dec. 2001) 6, 465. Limestone-1 (LS-1) 17 19 5 2.5
16. Alveskog, P. L., Holt, T. and Torsaeter, O.: The Effect of Limestone-2 (LS-2) 7 16 7.5 2.5
Surfactant Concentration on the Amott Wettability Index and Chalk (CH) 10 28 7.5 2.5
Residual Oil Saturation, J. of Petr. Sci. and Eng., (1998) 20, 247.
17. Graue, A., Viksund, B.G. and Baldwin, B.A.: Reproducible
Wettability Alteration of Low-Permeable Outcrop Chalk, SPE Oil Phase
(cp) (g/cc)
Res. Eval & Eng. (April 1999) 2, 134. Kerosene 2 0.79
18. Standnes, D. C., and Austad, T.: Wettability Alteration in Chalk Mineral Oil 175 0.88
1. Preparation of Core Material and Oil Properties, J. of Petr. Engine Oil 640 0.89
Sci. and Eng., (2000) 28, 111. Crude Oil 1 (CO-1) 2200 0.95
19. Standnes, D. C., and Austad, T.: Wettability Alteration in Chalk
Crude Oil 2 (CO-2) 6 0.81
2. Mechanism for Wettability Alteration from Oil-Wet to Water-
Crude Oil 3 (CO-3) 19 0.83
Wet Using Surfactants, J. of Petr. Sci. and Eng., (2000) 28, 123.
20. Baldwin, B.A. and Spinler, E.A.: In Situ Saturation
Development During Spontaneous Imbibition, J. of Petroleum Aqueous Phases
Sci. and Eng., (2002) 35, 23. Type
21. Weisbord, N., Niemet, M.R. and Selker, J.S.: Imbibition of Brine 3 % NaCl
Saline Solutions into Dry and Prewetted Porous Media, Surfactant Solution- 1 (Surf. 1) Non-ionic
Advances in Water Res., (2002) 25, 841. Surfactant Solution- 2 (Surf. 2) Non-ionic
22. Li, K., Chow, K. and Horne, R.: Effect of Initial Water Surfactant Solution- 3 (Surf. 3) Anionic
Saturation on Spontaneous Water Imbibition, paper SPE 76727 Surfactant Solution- 4 (Surf. 4) Non-ionic
presented at the 2002 SPE Western Reg. Meet., Anchorage,
Alaska, May, 20-22.
23. Mattax, C.C. and Kyte, J.R.: Imbibition Oil Recovery from
Fractured Water Drive Reservoirs, Trans.AIME (1962),
225, 177.
24. Babadagli, T.: Dynamics of Capillary Imbibition When
Surfactant, Polymer and Hot Water Are Used as Aqueous Phase,
J. Coll. and Int. Sci., 246, no.1, Feb. 2002, 203-213.
25. Lefebvre Du Prey, E.: Gravity and Capillarity Effects on
Imbibition in Porous Media, SPEJ (June 1978) 195.
26. Iffly, R., Rousselet, D.C. and Vermeulen, J.L.: Fundamental
Study of Imbibition in Fissured Oil Fields, SPE 4102 presented
at the 1972 SPE Annual Fall Meet, San Antonio, TX, Oct. 8-11.
27. Zhang, X., Morrow, N. and Ma, S.: Experimental Verification of
a Modified Scaling Group for Spontaneous Imbibition, SPERE
(Nov. 1996), 273.
28. Ma, S., Morrow, N. R. and Zhang, X.: Generalized Scaling of
Spontaneous Imbibition Data For Strongly Water-Wet Systems,
J. Pet. Sci. and Tech., (1997) 18, 165.
29. Zhou, D. et al.: Scaling of Counter-Current Imbibition Processes
in Low-Permeability Media, J. Pet. Sci. and Tech., (2002)
33, 61.
30. Rangel-German, E.R. and Kovscek, A.R.: Experimental and
Analytical Study of Multidimensional Imbibition in Fractured
Porous Media, J. of Petr. Sci. and Eng., (2002) 36, 45.
31. Handy, L.L.: Determination of Effective Capillary Pressure for
Porous Media from Imbibition Data, Trans. AIME (1960)
219, 75.
8 SPE 84866
SPE 84866 9
10 SPE 84866
SPE 84866 11
12 SPE 84866

Potrebbero piacerti anche