Wettability of Reservoir Rocks
Wettability of Reservoir Rocks
Wettability of Reservoir Rocks
of network modification (stress), or both. This point must always be checked, for example
by taking measurements in two or three orthogonal directions.
Rocks with probably negligible microcrack porosity: Permeability > 10 mD, Porosity > 0.15
These are porous sedimentary rocks with almost exclusively pore type porosity. The effect
of stress on permeability may lead to permeability variations of a few percent (very porous
and permeable rocks) up to several dozen percent (low or average permeability rocks).
Although important when considered as absolute values, these variations are not sufficient
to change the nature of the porous space and must be kept in perspective in view of the
wide dispersion of permeability values, as was suggested earlier. A large number of
measurements taken under ambient conditions will be more useful when characterising the
reservoir than a smaller number of measurements in confinement, since only very few of
these will be of any use to demonstrate the basic assumption.
Special cases: intermediate cases and highly porous, low-permeability rocks
In practice, with cases in between the two previous ones (0.05 < < 0.1; 0.1 < K < 10 mD) it
may be difficult to decide. Generally, a petrographic analysis will remove any uncertainty.
At the start of the study, however, it is certainly more cautious to check the effect of stress
on a limited number of samples.
Note also the special case of very porous, low-permeability rocks ( > 0.4; K < 5mD).
They may be easily-identifiable sedimentary rocks with very fine grain, such as chalks or
diatomites. Due to their high porosity, these rocks are extremely compressible and the pore
deformation elastic limit may correspond to low differential pressures. Nevertheless, this
does not make them particularly sensitive as regards permeability, obviously as long as we do
not exceed the elastic limit. We may however observe cases of vuggy porous media with
crack connections. These cases include carbonate rocks, although very rarely, but above all
lava (some types of pumice correspond to this definition). Stress may have a large effect on
permeability and it should be checked. Generally, in case of porous media for which few data
are available, e.g. lava, care should obviously be taken regarding the conclusions drawn.
In petroleum practice, the effects of stress on permeability become extremely important
in deep, low-permeability reservoirs, subject during production to a considerable drop in
pore pressure (and therefore to a significant increase in the effective stress).
(the only important parameters of the fluids are their surface tensions ts and, possibly, the
wettability angles ). By using the normalisation parameter ts cos , we were able, for a
given rock, to process equally well the cases water/oil, water/gas or mercury/vapour.
In nature, this favourable situation is only found in very specific cases, such as that of a
rock which has never been in contact with anything but water (or brine) and which is
invaded by liquid or gaseous hydrocarbons (secondary migration 1-1.2.6, p. 102). This is
also true near the surface, in vadose zone, where the porous space contains both water and
air. But even in this case, it is not uncommon to observe states (at least transient) of non-
water wettability (see Fig. 1-1.27, p. 53), under the effect of biological phenomena such as
microorganisms covering the porous walls.
Petroleum is a mixture of a large number of chemical compounds, including molecules
exhibiting a special affinity for solids. This affinity may be more or less marked for certain
minerals. We can therefore see that when a porous rock is invaded by oil, deposition of these
molecules may have a significant impact on the nature of the relations between the solid and
the fluids, in other words, by definition, the wettability. The example we give for the oil
reservoirs may be almost identical to the situation in surface geology (hydrogeology,
pedology), for example in the event of deliberate or accidental injection of chemical products.
Consequently, the capillary behaviour described in Section 1-1.2 must be reconsidered.
If these phenomena led to a clear wettability reversal, i.e. if the rock became as oil-wet as it
was previously water-wet, then we could simply apply the principles of Section 1-1.2,
exchanging the fluids. This type of wettability reversal can be carried out in the laboratory
by depositing silane molecules, for example, on the entire porous wall. In nature, however,
this is very rarely the case. The modification undergone by the solid/fluid interface is
heterogeneous and, together with the fact that the porous wall is sometimes very rough, it
induces phenomena which are extremely difficult to quantify accurately at microscopic
scale. This explains why the definition of wettability when studying geological materials can
only be based on a phenomenological approach.
To make it easy for non-specialists to understand the wettability phenomena observed in
rocks, a few elementary but useful remarks must be made. The first, and most important, is
that the diagnosis (and quantification) of petrophysical wettability is not the result of a
rigorous physical analysis based, for example, on the interactions of electrostatic forces,
but the result of a particular experiment. Several types of experiment (see below) giving
different results can be performed. It is therefore not surprising that the wettability state of a
rock varies depending on the test conducted.
The second remark is that, unlike porosity or single-phase permeability, wettability is not
an intrinsic characteristic of the rock, but a variable state of the interface between the
mineral solid and the fluids contained in the porosity. When defining this state, the fluids
and the history of their relation with the solid are as important as the solid itself.
Lastly, it should be added that, in everyday language, oil wettability may not be the
reciprocal of water wettability. Oil wettability is often the observation deduced from a
particular experiment discussed below. Although the same may be true of water
wettability, the result is often actually more conclusive, being based on a number of
observations. It would often be more appropriate to speak of a trend towards oil wettability.
4417_ Page 169 Vendredi, 9. novembre 2007 4:25 16 > Apogee FrameMaker Couleur
The debate between innate and acquired is a recurrent feature in some fields of
biology. The same applies in petrophysics with the problem of wettability since it is never
easy to distinguish clearly between the causes intrinsically related to the nature of the porous
space: geometry, mineralogy (innate) and those related to the nature of the fluids and the
particular history of their coexistence (acquired). The debate is of much more practical
importance than it seems, since only the innate parameters can be practically extrapolated
by using geology methods.
For detailed information on wettability, readers can refer to the bibliographic work of
Anderson [1986-87].
Pc B
max
(Capillary Pressure) Oil pressure Water pressure
3
A1
C"
S
C'
B' Water sat. 1 ( Sw )
B" A2
2
Swi
Initial Water Saturation
@ Pc max
Oil
Wet
Sw
0
1
Pc C
min
Figure 1-2.30 Schematic diagram of the USBM method for determining wettability
A sample originally totally saturated with water (point S, Fig. 1-2.30) is immersed in oil in
a centrifuge cup where it is subjected to increasing accelerations, up to very high values. The
sample undergoes a first drainage (phase 1 on Figure 1-2.30; see also Figure 1-1.34, p. 65)
bringing it into a state of irreducible water saturation (point B on Figure 1-2.30). Stopping the
centrifuge does not change this state, even though there is no longer any differential pressure
since the sample is completely surrounded by oil (point B' on Figure 1-2.30).
The measurement as such starts at this stage. The sample is transferred into a cup filled
with water and centrifuged again. We can apply the same calculation methods as with a
traditional drainage (P = h), but since the water phase is now the continuous phase
outside the sample the sign of must be inverted. The pressure in the water is greater than
that in the oil and we are moving along branch No. 2 on Figure 1-2.30 (forced imbibition
under the effect of negative capillary pressures). Note that the negative capillary
4417_ Page 171 Vendredi, 9. novembre 2007 4:25 16 > Apogee FrameMaker Couleur
pressures are only observed from point B on Figure 1-2.30 since the saturation interval
B' B'' corresponds to a spontaneous imbibition phase.
Under very high accelerations, from the saturation corresponding to point C, the oil
mobility becomes zero and we reach the residual oil saturation state, a generalisation of the
notion of residual saturation discussed in paragraph 1-2.3.3, p. 188.
Stopping the centrifuge (point C') does not change the saturation (see above). The
sample once again placed in a cup filled with oil undergoes a second forced drainage phase
(curve 3) which could possibly be preceded by spontaneous drainage (from point C' to point
C'').
The area (A1) between this second drainage and the x-axis corresponds to the energy
required for the oil to penetrate into the system. Equally, the area A2 between curve 2 and
the x-axis corresponds to the energy required for the water to penetrate. If the medium is
water-wet, this energy will be very low (ideally A2 = 0) and conversely for A1 if it is oil-
wet. In the USBM method, wettability is characterised by the logarithm of the area ratio
[WUSBM = log(A1/A2)]. If WUSBM > 0, the medium is preferentially water-wet, if
WUSBM < 0, the medium is preferentially oil-wet. For example, a medium with USBM
index of about 1 is strongly water-wet.
c) Amotts method
Amotts method [Amott, 1959; Cuiec, 1975] is based on estimating the relative importance
of the fluid fractions displaceable during spontaneous and forced operations (imbibition/
drainage).
After first being totally saturated with water, a sample is brought into a state of
irreducible water saturation by a long period of oil flooding (or possibly by centrifuging)
until no more water is produced.
In a first step, this sample is placed in water where it undergoes spontaneous imbibi-
tion, displacing a volume Va of oil (Figure 1-2.31a). The sample then undergoes
forced imbibition (waterflooding or centrifuging in water environment) displacing
an additional volume of oil Vb . Va is the volume of oil displaced spontaneously by the
water and Va + Vb the total volume of displaceable oil (spontaneous and forced). The
index Iw is defined: Iw = Va/(Va + Vb). If the medium is perfectly water-wet, we
obtain the situation of 1-1.2.2D, p. 59, and after the spontaneous imbibition phase,
the residual oil is totally trapped by the capillary forces, forced displacement by water
will produce no oil, Vb = 0 and Iw = 1.
The second step consists in carrying out the reverse operation with oil: the sample in
irreducible/residual oil saturation state (note once again the terminology ambiguity
as soon as the wettability state is no longer fixed) is immersed in oil (Figure 1-2.31c),
drainage spontaneously displaces the volume of water Vc. Forced drainage (flooding or
centrifuging) produces an additional volume Vd. We can therefore calculate an index
Io = Vc/(Vc + Vd). We will once again use the example of perfect water wettability
from 1-1.2. Spontaneous drainage will not displace any fluid, so Vc = 0 and Io = 0.
We can define a unique wettability parameter WIA = Iw Io of value between 1 and 1.
A value of 1 corresponds to the case of perfect water wettability, as mentioned previously.
4417_ Page 172 Vendredi, 9. novembre 2007 4:25 16 > Apogee FrameMaker Couleur
Oil (Va)
Sample
Brine
a b
Brine
Oil (Vb)
Brine
Sample
Oil
Sample
Sample
c d
Oil
Oil
Brine (Vd)
Brine (Vc)
Wettability plays a very important role in the oil recovery processes in reservoirs since it has
a direct impact on the proportion of oil recovered and on the kinetics of this recovery. We
will discuss this point in the End Points paragraph ( 1-2.3.3, p. 186). The obvious
practical importance of wettability has led to numerous research studies. Nonetheless, it is
still difficult to obtain an accurate picture of wettability in reservoirs since numerous
difficulties are encountered when trying to upscale results from laboratory to reservoir. As
4417_ Page 174 Vendredi, 9. novembre 2007 4:25 16 > Apogee FrameMaker Couleur
we have already pointed out, this is due to the fact that wettability is a state which may turn
out to be highly unsteady.
Figure 1-2.32 shows examples of reservoir rock wettability evolution, over the course of
geological time and depending on the laboratory sampling and processing conditions.
Preserved Sample
We must therefore rely on the results obtained on cores. Unfortunately, numerous factors
may disturb the wettability state of a rock sample between the reservoir and the laboratory.
During the coring and storage operations, the rock is subjected to three main types of
process which may have a serious impact on wettability.
Flushing by drilling mud. This is one of the most obvious cases of wettability altera-
tion. The muds contain additives (e.g. soda) which may have a drastic effect. Tech-
niques are available, however, to control the degree of invasion, especially by using
markers. By adding small quantities of deuterium, tritium or special salt to the mud,
the depth of penetration of the filtrate inside the core can be measured on small core
samples. It is therefore possible to demonstrate that if coring is carried out with a suit-
able mud pressure, the core centre is often free from any contamination.
Deposition of organic molecules precipitated during depressurisation. When the core
barrel is brought up to the surface, the pressure drops suddenly and the dissolved gas
returns to gaseous phase, suddenly flushing out the liquids remaining in the core.
More importantly, it may also induce the precipitation of high molecular weight mol-
ecules (e.g. asphaltene) which have a major impact on wettability. This phenomenon
can be limited by carefully controlling the speed at which the core barrel is brought to
the surface. The ideal solution would be to have a pressurised core barrel, in order to
bring up the core without any pressure variations, or at least without lowering the
pressure below the bubble point. Since this aspect represents a major financial stake
(more for preservation of saturations than wettabilities) technical solutions have been
proposed while still, apparently, remaining outside the scope of routine applications.
Drying and alteration during storage and transport. If left untreated, the core charac-
teristics are soon modified. It is easy to remedy this situation, however. As soon as
they are removed from the core barrel, native samples are placed in sealed contain-
ers, refrigerated or sometimes even frozen. They therefore reach the laboratory under
optimum conditions.
The roughness may also play an important role. The effect of roughness was men-
tioned with respect to the capillary pressure hysteresis. It seems likely that it has a
direct impact on wettability, for example through its control on the irreducible
water layer. The roughness of the pore walls may vary considerably. Two contrasting
examples of roughness, observed using epoxy pore casts ( 2-2.1.3, p. 331) are illus-
trated on Figure 1-2.33. The roughness parameter is difficult to quantify absolutely
(numerous microscopic analysis would have to be made on images of the type shown
on Figure 1-2.33). However, since the roughness is directly related to the nature of the
100 m
100 m
200 200
Height above initial Water Oil Contact (m)
Height above initial Water Oil Contact (m)
180 180
160 160
140 140
120 120
100 100
80 80
60 60
40 40
20 20
0 0
0 10 100 1,000 10,000 0 0.2 0.4 0.6 0.8
Permeability (mD) Wettability index to Water (lw)
Figure 1-2.34 Evolution of oil wettability against the distance from the oil-water contact.
Sandstone reservoir. Wettability index measured using a method derived from Amott,
samples subjected to soft cleaning then restoring. Experimental values of G. Hamon [2000]
The relation permeability/distance from the water level indicates no special relation between these two parameters.