Vous êtes sur la page 1sur 13

SPE 56513

Defining Fluid Distribution and Fluid Contacts for Dynamically Charged Reservoirs
Fareed Iqbal Siddiqui, SPE, Pakistan Petroleum Limited

Copyright 1999, Society of Petroleum Engineers Inc.


This paper was prepared for presentation at the 1999 SPE Annual Technical Conference and
Exhibition held in Houston, Texas, 36 October 1999.
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
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
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
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
for commercial purposes without the written consent of the Society of Petroleum Engineers is
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
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.

Abstract
Defining fluid contacts(GWC/OWC/GOC) is one of the major
variables for defining the initial hydrocarbons in place. In
exploratory wells RFT/MDT is often used to define
OWC/GWC. In defining contacts static capillary-buoyancy
equilibrium is assumed. There is ample field evidence that
many of the existing reservoirs are actively being charged from
the source rocks. Under these conditions the assumption of
static fluid distributions may not be valid and dynamic
conditions should be used for defining the fluid distributions
and the contacts.
This paper defines the fluid distributions and fluid contacts
for dynamically charged reservoirs. The method is based on
the recently proposed Dynamic Theory of Hydrocarbon
Migration and Trapping. The results of numerous numerical
simulations show that assuming static saturation profiles may
not be valid for reservoirs having active influx of fluids into
them. This is true even at very small influx rates (10-9 m/s-10-14
m/s). The pressure profiles for the static case (no viscous
forces) are compared and contrasted with the dynamic pressure
profiles. The dynamic pressure drops can be explained on the
basis of Darcys law and relative permeability effects under
extreme saturations. These pressure drops are then used to
calculate the saturation distribution in the hydrocarbon column
and the transition zones. The results show that if the
assumption of static equilibrium is used for the dynamically
charged reservoirs it will lead to the underestimation of the
initial hydrocarbon in place. The non-equilibrium effects can
also explain free water production from zones far above the
transition zone, expected, on the basis of static assumption.
The interpretation of MDT data under both static and dynamic
conditions is also discussed.

Introduction
When exploring for or developing new oil and gas reservoirs,
wells high on a structure may not penetrate the hydrocarbonwater contact. This contact needs to be determined, as soon as
possible to locate delineation wells, plan development drilling,
and, forecast reserves and economics. It is just not enough to
determine the location of hydrocarbon-water contact but it is
also important to determine the saturation distribution above it.
The saturation distribution is used in calculating the original
hydrocarbons in place and is also an input to the reservoir
simulators. As such an accurate determination of the
hydrocarbon-water contact location and the saturation
distribution has a big impact on the development plan and
economics of a field. Therefore, it is very important to
understand the basic physics of fluid distribution in porous
media.
Fluid Contacts (Gas-oil/oil-water/Gas-water) are usually
determined either by using logs or the pressure gradients
determined from some type of downhole formation tester (such
as MDT). In fact, most of the time, both of these methods are
used in conjunction to define the fluid contacts. The basis for
interpreting MDT pressure data is the presence of complete
static capillary-buoyancy equilibrium.1,2 It is assumed that
enough geological time has passed since migration, such that
all the dynamic effects can be neglected. Even in the
interpretation of logs the underlying assumption, at the back of
interpreters mind, is the existence of capillary-buoyancy
equilibrium. We show in this paper that the capillary-buoyancy
equilibrium may not exist for reservoirs that are being actively
charged from the source rock. Many of the existing reservoirs
are being charged by active source rocks, thus preventing the
presence of hydrostatic equilibrium in them. Morrow and
Melrose3 have concluded by comparing, the connate water
saturation data from cores cut with oil-base muds, with the
mercury injection data, that many of the present day reservoirs
may not be in hydrostatic equilibrium. Morrow and Melroses
observations can be explained on the basis of active charging
of these reservoirs.
This paper takes a more fundamental view of the problem
and considers the determination of fluid contacts and
saturation distribution, to be a corollary of the hydrocarbon
migration and trapping phenomenon. The fluid contacts and

FAREED IQBAL SIDDIQUI

distribution are defined on the basis of the recently proposed


dynamic theory of hydrocarbon migration and trapping.4 It is
shown here that for reservoirs actively being charged from the
source rocks, the fluid distributions may be quite different
from the distribution under static conditions.
In this paper, a brief overview of hydrocarbon migration
and trapping process is presented first and then a summary of
the fundamental equations governing the migration and
trapping process is presented. The method of defining the
saturation distribution, based on the static equilibrium
assumption, is then discussed. Then it is shown that how,
under dynamic conditions, the pressure gradients can be
significantly different from those under static conditions. The
saturation distributions are defined on the basis of the dynamic
pressure gradients and the dynamic theory of migration. In the
end the interpretation of MDT data for the static and dynamic
reservoirs is discussed.
Background
As discussed earlier, the problem of determining the fluid
contacts and fluid distribution is the end result of the migration
and trapping process. Therefore, a brief discussion of the
migration and trapping process is given here first:
Hydrocarbons are formed in the organic rich source rocks that
are usually low permeability (usually shales) and migrate
through high permeability rocks to traps or seals.5-7 A seal is
some low permeability rock which stops the further migration
of hydrocarbons. Source rocks produce hydrocarbons under
particular temperature and pressure conditions called oil and
gas generation windows. Once these conditions are not present
or when all the organic matter is exhausted the charging of
hydrocarbons into the reservoir stops. A dynamically charged
reservoir is one that has an active source rock feeding fresh
hydrocarbons into it at the present time. In our usual treatment
of reservoirs it is assumed that in the geologic time, complete
capillary-buoyancy equilibrium has been achieved. This
assumption may not be true if the reservoir has active charging
from the source rock.
Fundamental Equations. Migration of hydrocarbons into
reservoirs is governed by the general equations of multiphase
flow in a permeable media. The fundamental equations
describing the multiphase flow in permeable media are
discussed in standard texts8 and are summarized below:
Darcys law

k k ro P o
(1)
+ o g sin
uo = o x

uw = Continuity

k k rw P w
+ w g sin

w x

S o
u o
+
= 0

t
x
S w
u w
+
= 0

t
x

(2)

SPE 56513

where uo and uw are the oil and water fluxes respectively,


and x and t are space and temporal variables, respectively. x is
positive in the vertical upward direction. k ro and k rw in the
above equations are oil and water relative permeabilities and
are considered to be functions of saturations alone. Capillary
pressure (Pc) is defined as the difference in pressure between
the non-wetting and the wetting phases. For an oil-water
system, assuming water to be the wetting phase, Pc = Po- Pw.
The assumptions used in driving the above equations are:
(1) One dimensional flow (2) Isothermal flow (3) No
adsorption or chemical reactions (4) Incompressible fluids and
rock (5) Two phase flow of oil and water (6) No partitioning
between the phases. Expressing uo in terms of Pc we get for
Eqs. (1) and (3):
P
uo = fut + fkrw g fkrw c
(5)
x

S
P
(6)
o + f ut + f k rw g - f k rw c = 0
t
x
x
where f = ro

(ro + rw ) is the Buckley-Leverett

fractional

flow function, = w g , and ro and rw are the oil


and water relative mobilities, respectively. Equation (6) is the
general equation that describes the unsteady state flow of two
phases in one dimension. This is also the equation that
describes a one dimensional migration and trapping problem.
Most of the migration occurs initially vertically upward and
then along the top of the carrier bed, both of which can be
considered a one dimensional flow. Once the basic physics of
the migration problem is understood in one dimension it can
be easily extended to three dimensions. The solution of the
above equation, with proper initial and boundary conditions,
would give the location of trapped hydrocarbons, the length of
hydrocarbon column and the saturation distribution within the
hydrocarbon column. Equation (6) is a non-linear parabolic
partial differential equation that is difficult to solve
analytically and numerical techniques are needed to solve it.
However, when certain terms in it are neglected it can be
solved analytically.
Pressure and Saturation Distribution in Static Reservoirs.
Most of the time in our consideration of the migration problem
the full form of Eq. (6) is not solved and a static form is used.
In the static form it is assumed that, at the present time,
complete static equilibrium has been achieved and the fluids
are completely segregated. This means uo = u w = 0 at each
point in the reservoir. With this assumption the solution of Eq.
(6) reduces to the solution of Darcys equation and gives:

(3)

Po
= - o g ;
x

(4)

and

Pc
= g
x

Pw
= - w g
x

(7)

(8)

SPE 56513

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

The above equation is integrated above the free-water level,


Pc=0, to give capillary pressure as a function of height in the
reservoir.1,2,10
Pc = gh

(9)

h in the above equation is the height above the free water level
(FWL). The free water level is usually determined by using the
MDT pressure points. It should be noted that the free water
level is not the same as the 100% water saturation level,
usually known as the OWC. Once a capillary pressure profile
is available some type of capillary pressure-saturation (Pc-So)
relationship is required to get the saturation distribution. One
of the most commonly used Pc-So relationship is that proposed
by Leverett.11
Pc =

cos
j ( So )
k

(10)

Different forms of Leveretts j function are available. If


experimental data is available the form of j function that fits
it best should be used.
Migration of hydrocarbons is a drainage process. Figure 1
shows a typical Pc-So curve for drainage which shows that a
threshold pressure is required to initiate the injection of the
non-wetting phase into a permeable medium. This threshold
value of capillary pressure is called the displacement pressure,
Pd . This phenomenon causes the OWC ( Pc = Pd ) to be
different from the free-water level ( Pc =0). The height of
OWC above the free water level (do) can be calculated as
d o = Pd g and, as is shown later, this distance can be
significant for moderate to low permeability reservoirs. Figure
1 shows that the Pc curves becomes asymptotic at the
irreducible water saturation (Swi). In this paper the BrooksCoreys12 drainage form of the capillary pressure-saturation
relationship has been used even though other forms can be
used.

j ( S ) = (1 - S ) n Pc
S=

(11)

S o - S or
1 - S or - S wr

where S is the normalized oil saturation. In the rest of the


paper So is used for the normalized saturation instead of S.
According to equations (10) and (11) the displacement
pressure Pd is given by cos k . n Pc in Eq. (11) is a
positive number that characterizes the distribution of pore
throat sizes in the permeable medium and is usually
determined by fitting the experimental data to Eq. (11). Figure
2 (right axis) plots the height of the 100% water level
(OWC/GWC) above the free water level as a function of
permeability for typical values of oil and gas interfacial
tensions. It is obvious from the figure that for moderate to low
permeability reservoirs (less than 100 md) the difference is not
negligible.

Using Eqs. (9)-(11) the capillary pressures at different


heights are converted into saturations. Figure 3 is a schematic
showing the pressure and saturation distribution in a static
reservoir. Capillary pressure is zero at the FWL and increases
linearly in accordance with Eq. (9). Both oil and water
pressures follow their respective static gradients above the
FWL. Oil saturation remains at zero until Pc=Pd, this point
defines the OWC. As the capillary pressure increases above
the OWC, the water saturation decreases gradually in
accordance with the Pc-So relationship (Fig. 1). The saturation
continues to decrease until it reaches the irreducible water
saturation, Swi. The region where water saturation varies from
100% water to Swi is called the transition zone. Both oil and
water phases are mobile in the transition zone. Above the
transition zone there is no change in the water saturation with
the increase in capillary pressure. The capillary pressure in this
zone keeps on increasing with height above the Free-WaterLevel (FWL) in accordance with Eq. (9) but the saturation
does not change because the capillary pressures lie in the
asymptotic region of the Pc-So curve. This is the main oil zone
for a static reservoir, water in this region is considered to be
immobile and water free production is expected from this
zone.
Another corollary of the static pressure distribution is the
definition of the maximum hydrocarbons that can be trapped
under a given seal. The capillary pressure at the top of
reservoirs should be less than or equal to the displacement
pressure of the seal (PdS in Fig. 3). When the pressure at the
top of reservoir becomes equal to or greater than the
displacement pressure of the seal, the hydrocarbons will start
to leak into the seal. Thus the maximum hydrocarbon column
that can be trapped in a reservoir is defined as:4,6,7
hstatic =

PdS Pd
g

(12)

where Pd is the displacement pressure for the reservoir. In


contrast to this dynamically charged reservoirs have the ability
to transiently trap hydrocarbon columns greater than hstatic.6
Dynamically Charged Reservoirs. Many of the reservoirs are
actively being charged from the source rock and the validity of
static saturation distribution assumption may not be valid for
them. Morrow and Melrose3 have concluded, by comparing,
the connate water saturations for the cores cut with oil based
mud, with the mercury injection residual saturations, that
capillary-buoyancy equilibrium may not be present in many of
the reservoirs. Attainment of capillary equilibrium may take
long time because of relative permeability effects, as discussed
in this paper, and also because of film flow phenomenon at
low wetting phase saturations.
Dynamic Pressure Profiles. The static pressure profiles as
shown in Eqs. (7) and (8) are the limiting case of the general
pressure gradient equations based on the Darcys law. We can
use Eq. (5) to express the oil and water pressure gradient

FAREED IQBAL SIDDIQUI

equations in terms of relative permeabilities and capillary


pressures. For the sake of simplicity we assume the regional
ground water flow, ut, to be zero, then from Eq. (5):
dPc S o
- g )
dS o x
- og

krw + kro w
o

(13)

dPc S o
)
dS o x
- wg

kro + krw o
w

(14)

w
o
uo
Pc
= +
(
) + g
k
k
k
x
1 4 4 4 4 rw
2 4 4 4ro 4 3

(15)

Po
=
x

Pw
=
x

krw (

kro ( g -

Dynamic Effect

The first terms in the above equations denote the dynamic


effects while the second terms are the static effects. Static
effects are independent of the movement of fluids and are not a
function of time. The dynamic effects however are, a function
of both the flux rates of the phases and time. The contribution
of these effects changes as the saturations in the system
change. The dynamic effect is a combination of both the
viscous and capillary effects. Viscous effects are a result of the
drag caused by the movement of fluid elements against each
other and the walls of the pores. They are a function of the
viscosity and saturation of the fluids. Capillary effects are a
function of saturation and heterogeneity and cause the
dissipation of sharp fronts. According to Eqs. (13)-(15), in the
regions where either hydrocarbon or water relative
permeabilities are close to zero, the dynamic effect may not be
negligible even for small flux rates (uo).
A large number of numerical simulations at low influx
rates (10-9 m/s-10-14 m/s) have been conducted and the results
show that the resultant pressure profiles are much different
from the static profiles. Some pure segregation cases were also
conducted, in which the reservoir was allowed to segregate
under the influence of gravity from an initial uniform
saturation. Even in these cases long after when 98% of the
segregation was complete the dynamic effects in the pressure
gradients was not negligible. The results of these simulations
are presented in Reference 14.
Dynamic vs. Static Reservoirs. All the dynamically charged
reservoirs may not show dynamic pressure profiles. According
to the dynamic4 theory of migration and trapping every
reservoir-seal system can be categorized as either static or
dynamic. For the sake of convenience, these will be
abbreviated as static or dynamic reservoirs (instead of static or
dynamic reservoir-seal system). In static reservoirs the
pressure distribution follows the static gradients (defined by
Eqs. (7) and (8)), even if they are actively being charged from
the source rock. While in dynamic reservoirs the pressure

SPE 56513

distribution is given by the dynamic pressure profiles defined


by Eqs. (13)-(15). Whether a particular reservoir will behave
as dynamic or static is a function of the relative permeability
vs. saturation behavior of the reservoir and the seal, the
capillary pressure function of the reservoir and the seal, and
the influx rate from the source rock. Reference 4 presented an
approximate solution for the migration and trapping problem
using the method of characteristics (MOC) solution technique.
The effect of capillary pressure was partly accounted for by
including the continuity of capillary pressure as one of the
boundary conditions. The method of characteristics solution is
based on the flux-saturation relationship with the capillary
pressure term neglected in Eq. (6). Neglecting the capillary
pressure term reduces Eq. (6) from a parabolic to a hyperbolic
partial differential equation solvable by the method of
characteristics. The hyperbolic form of the diffusivity equation
is:
1 duo S o
S o
+
= 0
t
dS o x
According to the method of characteristics the specific
velocity of a constant saturation So is given by the slope of the
flux-saturation ( uo S o ) curve at that saturation:8,13
1 du ( S )
dx
v So =
= o o
dS o So
dt So

(16)

The uo So relationship is highly non-linear which under


certain conditions causes shocks to develop from the
continuous waves. The specific velocity of the shock is given
by:
v shock =

1 u o
1 u o- - u o+
=
S o
S o- - S o+

(17)

where (uo- , S o-) and uo+ , S o+ are the points upstream and
downstream of the shock, respectively. Equation (17) suggests
that v shock is the slope of the chord joining the points

(uo- , S o-)

and

(uo+ , S o+)

on a uo vs. So plot. There are

separate uo S o curves associated with the reservoir and the


seal, depending on their absolute permeability and their
relative permeability functions. Trapping is explained as a
result of reflection and refraction of saturation waves from the
reservoir-seal boundary. The capillary pressure term in the flux
expression (Eq. (5)) is only important when Pc x is large
and that only occurs near a heterogeneity boundary or around
the shock fronts. In the regions away from shocks and the
heterogeneity boundary the MOC solution can be applied.
Below the criteria for defining static and dynamic reservoirs is
described on the basis of, the continuity of flux and continuity
of pressure conditions, and the method of characteristics.
Figure 4(a) shows the flux-saturation function for a
reservoir which is actively being charged from the source rock

SPE 56513

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

at an influx rate of uoJ . SA is the saturation corresponding to


this flux on the higher oil saturation flank of the uo S o
curve. It is also the minimum saturation for which the injected
flux, uoJ , in the MOC solution, can produce a reflected wave
into the reservoir. Figure 4(b) shows the capillary pressuresaturation (Pc-So) function for a reservoir and two seals. The
condition of continuity of capillary pressure dictates that a
minimum saturation S* (S1* or S2*) is established in the
reservoir before any hydrocarbons can leak into the seal. For
saturations less than S* the Pc at the top of the reservoir is less
than the displacement pressure of the seal. Therefore, S* is
defined such that Pc(S*) = PdS.
The behavior of a particular reservoir, whether static or
dynamic, will depend on the relationship of SA with S* of the
seal associated with it. If S* for a seal is greater than SA , then
the reservoir is termed as a dynamic reservoir (shown by S1*
in Fig. 4(b)). In a dynamic reservoir the dynamic terms in the
pressure gradient equation (Eqs. (13)-(15)) are not negligible.
Under these conditions the oil and water pressure gradients in
the reservoir and the seal correspond to the dynamic pressure
gradients defined by Eqs. (13)-(15). However, if S* is less
than SA then the seal is termed as a static seal (shown by S2*
in Fig. 4(b)) because in this case the dynamic terms in the
pressure gradient equation are negligible. The pressure
behavior for such reservoir-seal systems corresponds to the
static pressure profile defined by Eqs. (7) and (8). The
saturation and pressure profiles for the static case have been
discussed earlier and the method of defining the saturation and
pressure distribution for the dynamic reservoirs is presented
next.
Pressure and Saturation Profiles for Dynamic Reservoirs.
The method of characteristics solution technique can be used
to determine the saturation profile for a dynamic reservoir.
Figure 5 is a schematic showing the solution for a dynamic
reservoir on a uo S o diagram. The reservoir is initially
completely saturated with water (point I in Fig. 5). An
injection of flux, uoJ , results in an upward moving shock
wave with saturation S J1 and a specific velocity given by the
slope of chord I-J1. Figure 6(b) shows the solution on a timedistance diagram. The shock with saturation S J1 reaches the
reservoir-seal boundary ( x1 ) at time t. From Fig. 4(b) no oil
can flow into the seal until the saturation at x1 becomes equal
to S1*. This results in the trapping of hydrocarbons beneath the
reservoir-seal boundary and the oil saturation at the top of
reservoir starts to build up. At time t L the saturation at the top
of the reservoir becomes equal to S1* (Fig. 6(b)) and the
continuity of capillary pressure upstream and downstream of
the boundary is satisfied. This starts the leakage of
hydrocarbons into the seal. The continuity of flux condition
defines the flux at the boundary to be uo(S*) = uJ2. The

corresponding saturation S J 2 defines the saturation just


downstream of the boundary in the seal. The difference
between the injected flux and the leaking flux, ( uoJ - uoJ 2 ) ,
continues to accumulate in the reservoir. This is the reflected
shock wave shown by the chord J1-T1 in Fig. 5. The
hydrocarbon saturation in this reflected wave is high (S1*) and
is identified as oil column in Fig. 6(a), which shows the
saturation profile at time t1. S1* in Fig. 5 is shown to be
artificially low to illustrate the construction of the MOC
solution.
The line representing the shock in Fig. 6(a) is curved
between t and tL because of the interaction between the
reflected waves and the upward moving waves from the lower
boundary. The length of the hydrocarbon column continues to
increase after t L even though a part of the flux is leaking into
the seal. The saturation upstream and downstream of the
different fronts are shown in Figs. 6. Figures 6 also shows that
dynamic reservoirs can trap hydrocarbon columns greater than
hstatic (Eq. 12).
The time-distance diagram of Fig. 6(b) can be used to
define the saturation distribution in the reservoir at any time.
These saturations can then be used in conjunction with the PcSo relationship to get the solution in terms of pressures. Figure
6(a) is the saturation profile based on the MOC solution at
time t1 and shows that the reservoir can be divided into two
zones: an oil zone formed because of the reflected shock with
constant saturation S1* and a water zone below the shock with
constant saturation SJ1. These shocks are moving at specific
velocities and are thus changing the constant saturation
regions.
The method of characteristics solution described above
includes the effect of capillary pressure for the determination
of the saturations and velocities of the reflected shocks but
does not include the dissipation effects caused by the capillary
forces. According to Eq. (5) the effect of capillary forces is
proportional to Pc / x . Therefore, the inclusion of capillary
effects, the solution given by the method of characteristics and
shown in Figs. 6(a) and 6(b), has to be modified in the regions
where Pc / x is not negligible. Pc / x is negligible in most
of the reservoir except at the reservoir-seal boundary
(where k / x = ) and at the shock front (where S o / x = ).
Capillary pressure will cause the sharp saturation jumps
(shocks) to disperse around their mean position. With the
dissipation regions included the saturation and pressure
profiles in the dynamic reservoirs can be divided into four
regions as shown in Fig. 7. The presence of these four regions
has also been confirmed by the results of a large number of
numerical simulations run at low influx rates(10 -4-10-9 m/s).14
The differentiating characteristics of each of these regions are
discussed below.
Region I.
The saturation in this region is constant at
SJ1 and is dictated by the saturation of the shock formed
because of the influx, uoJ, from the source rock. For practical

FAREED IQBAL SIDDIQUI

rates of hydrocarbon migration (10-4-10-9 m/s), SJ1 is only


slightly higher than zero (considering migration to be a
drainage process). Therefore, in this region S w 1 , k rw 1,
k ro 0 with S o / x = 0 which gives

Po Pw
=
= w g
x
x

and

Pc
=0
x

Thus in this region both oil and water phase pressure gradient
will be equal to the static water gradient ( w g ).
Region III. This is the main hydrocarbon zone of the
reservoir. It is produced by the reflected shock with saturation
S1* shown in Figs. 5 and 6. The saturation in this region is
constant at S1* and S o x = 0 . Constant saturation implies
the capillary pressure to be constant in this region and
therefore, putting Pc x =0 in Eq. (13) and (14) we get for
the pressure behavior in this region:

Pw Po
=
=
x
x

k rw g
o g

k rw + k ro w
o

(18)

According to Eq. (18) above, in a constant saturation region,


both oil and water pressure gradients will be equal and will in
general vary between o g and w g . The actual pressure
gradients in this region can be calculated by using So=S* in
Eq. (18). For good seals PdS is high which results in the water
saturation in this region to be close to the residual water
saturation (Swi). At these saturations k rw 0 and the water
production from the zone would be minimal. The pressure
profiles under these conditions reduce to

Po Pw
=
- o g
x
x
Therefore, in the oil zones of reservoirs with good seals, the
pressure gradient in both the oil and the water phase will be
equal to the oil static gradient. This is in contrast to the
pressure profiles for static reservoirs where both oil and water
pressures will follow their respective static gradients.
Now let us have look at a poor seal with low PdS
(comparatively higher permeability), which results in a lower
value of S* (say 50%). Figure 4 shows that at low influx
rates such a seal will result in a static reservoir with a small
hydrocarbon column equal to hstatic given by Eq. (12).
However, at higher influx rates the same reservoir will act as a
dynamic reservoir which for continuous influx from the source
rock will trap much more hydrocarbons than hstatic. However,
the saturation S* in the shock will be lower and the water
saturation in the hydrocarbon column will be much higher than
the residual water saturation. Under these conditions the water
relative permeabilities will be high and free water production
will be observed from this zone. Therefore, the production of
free water from wells drilled high on the structure much above
the transition zone can be attributed to the reservoir being
charged at high rates with a high permeability seal. This

SPE 56513

suggests that, in dynamic reservoirs high hydrocarbon


saturations zones should selectively be perforated and high
water saturation layers should be avoided, even if they lie on
the crest of the structure.
Region II.
This region is caused by capillary dispersion
of the reflected shock forming Region III with saturation S*.
As shown in Fig. 7(a) capillary pressure causes the shock to
spread around its mean position defined by the MOC. The
saturation in this region varies from S* to SJ1. This region
constitutes the transition zone in dynamic reservoirs. The
length of this zone is very small and can be neglected for
making initial hydrocarbon in place calculations.
Region IV. This is a very small region immediately
below the reservoir-seal boundary and capillary pressure term
is dominant here because of k /x = . When the initial shock
with saturation SJ1 reaches the boundary, it can not pass on
into the seal, as initially, the Pc in the reservoir is less than PdS.
The saturation at the top of the reservoir then starts to build up
and continues to build until it reaches S*, such that
Pc(S*)=PdS. The saturation at the top then becomes constant
and the hydrocarbons start to leak into the seal with a very low
saturation (SJ2 in Fig. 6) with a flux rate given by uo(S*)=(uoJ2).
The Pc x in this region can be calculated from Eq. (15)
with So=S*. The length of this region is also very small.
Experimental Evidence of Dynamic Pressure Profiles. The
pressure gradient behavior in the oil zone (region III Fig. 7)
seems to be counterintuitive but, as shown earlier, is a result of
the dynamic pressure drops. A literature search indicated that
pressure profiles similar to those shown in Fig. 7, have been
reported, although not identified as such, in the experimental
studies in the hydrology literature.15,16
Figure 8 shows the capillary pressure from Scott and
Coreys15 water drainage experiments. They introduced water
at the top of a soil column and let it drain because of gravity
through a column saturated with air initially. Air is the nonwetting phase and water the wetting phase.
Figure 8(a) shows the results for a homogeneous layer
while Fig. 8(b) shows the results for a two layer sand with the
low permeability layer at the bottom. Water moves down into
the sands leaving the low water saturation regions at the top.
Figures 8(a) and (b) show that in the low wetting phase
saturation regions capillary pressure becomes constant,
i.e., Pc x =0. This is similar to the gradients in the oil zone
(region III Fig. 7) for the dynamic reservoir. Static pressure
profile would require capillary pressure to increase in
accordance with Pc x = w g (air 0) which is not what is
recorded during the experiments.
Dynamic pressure profiles have also been reported in the
results of several numerical simulation studies.17,18 Lujan17
showed such pressure profiles for the drainage of oil spills
through soil layers. Fetsy and Van Gold-Racht18 have also
reported similar profiles for gas gravity drainage in fractured
reservoirs.

SPE 56513

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

Comparison of Saturation distribution for Static and


Dynamic Reservoirs. As shown in Fig. 3 the hydrocarbon
zone in a static reservoir can be divided into two regions: a
transition zone, where the hydrocarbon saturation varies from
0, at the OWC, to 1-Swi, at the end of the transition zone, and;
a region of irreducible water saturation. In comparison, the
hydrocarbon column in a dynamic reservoir can be combined
into just one zone (Region III in Fig. 7). The saturation in this
zone is constant at So=S*. The transition zone because of
capillary dispersion and the zone at the reservoir-seal
boundary are very small and can be neglected.
Figure 9 compares the fluid and pressure distribution for a
given oil reservoir in the case if it was acting as a dynamic
reservoir vs. a static reservoir. Figure 9(a) compares the
saturation profiles and 9(b) compares the capillary pressure
profiles. The reservoir is 100 m thick. Other rock and fluid
properties are as follows:
0.20

k
100 md
10 dynes/cm
ow
0.85 g/cm3
o
0.95 g/cm3
w
The capillary pressure dependence is given by the BrooksCoreys form of j function (Eq. (11)) with nPc=0.3 and an
irreducible water saturation of 0.20. The capillary pressure
versus saturation for this reservoir is plotted in Fig.1 and is
used to convert the capillary pressures into saturations. As
shown in Fig. 9(b) the capillary pressure in the static reservoir
increases linearly above the FWL with a slope of -g, in
accordance with Eq. (9). Pd for the reservoir is about 2.1 psi,
which gives the distance of the OWC from the FWL to be 14
m. In Fig. 1 the Pc curve becomes asymptotic at a pressure of
8-9 psi. This is the capillary pressure at the end of the
transition zone and gives a transition zone of about 42-49 m
above the OWC. Above the transition zone there is about 40 m
of oil column with So=1-Swi.
The saturation distribution for the case if the reservoirs
were dynamic is also shown in Fig. 9. In contrast to the static
reservoir, a dynamic reservoir will have just one continuous oil
zone and the oil saturation in it is defined by the threshold
pressure of the seal. Let us suppose that the threshold pressure
for the seal is 60 psi. At a Pc of 60 psi the oil saturation is
0.8 (1-Swi) from the Pc-So curve for the reservoir. This would
give a 86 m oil zone with a saturation of So=0.8 as compared
to 40 m if the reservoir were assumed to be static.
MDT Measurements. It is obvious from the discussion above
that the pressures and pressure gradients in the dynamic
reservoirs can be significantly different from those of the static
reservoirs. Let us see at the implications of these observations
on the interpretations of the pressure data collected by the
wireline formation testers, such as MDT. These tools measure
the in-situ pressures at different depths in a reservoir. These

pressures are then plotted against depth to determine the type


of fluid and the fluid contacts in the reservoir.
In our usual interpretations of the MDT data we implicitly
assume that the permeable medium is occupied with single
phase only. The transition zone is assumed to be small and the
capillary pressures in it are also assumed to be small (assumed
to be in the range of the accuracy of the pressure gauge, which
is 2 psi for the MDT tool). Figure 2(a) and (b) (left axis) plot
the capillary pressures at the GWC vs. permeability for both
gas and oil reservoirs for a range of interfacial tensions. It is
obvious from Figs. 2(a) and (b) that the capillary pressures
may not be negligible even at the start of the transition zone.
The basic question with regard to the pressures measured
by the MDT probe is that, to which phase does this pressure
correspond to? The question is not important in the regions
where only one phase is mobile, the other phases being at their
residual saturations. It however becomes important in the
transition zone or regions where more than one fluid is mobile.
This is important because it is the pressure gradients and not
the absolute value of pressure that are used for determining the
type of fluid and the FWL. The pressure gradients are greatly
affected by which phase the probe is responding to.
One of the possibilities is that it measures the pressure in
the phase dominant at the point of measurement. It is not clear
how to define the dominant phase in the transition zone, one
possible definition would be the phase with higher relative
mobility at the point of measurement. The other possibility is
that the probe measures the pressure of the phase that wets it.
In this case, if the well is drilled using a water based mud than
the probe will record the pressures in the water phase all the
time. The effect of this discrepancy has not been discussed in
the literature which, as is shown below, will effect the
interpretation of the MDT data.
MDT Tools Response in a Static Reservoir
Figures
10(a) and (b) show a schematic of the possible responses of an
MDT tool in a static reservoir for the different possibilities
discussed above. Figure 10(a) gives the response for the case
when the probe is responding to the wetting phase pressure.
Pressures measured by the tool are indicated by *. Above the
transition zone water is assumed to be immobile and the
measured pressure will be that of the oil phase and in a static
reservoir the gradient through the recorded points will be equal
to the static oil gradient (Eq. (7)), line AB in Fig. 9(a).
However, in the transition zone and in the water zone below it,
the probe will record the pressure in the water phase, assuming
water to be the wetting phase. For a static reservoir, the water
phase pressure gradient in the transition zone, and in the 100%
water zone, would be equal to the static water gradient. In this
case the transition zone will be interpreted as water zone based
on the establishment of water gradient through it. If during the
survey no pressure points were recorded in the oil zone
(irreducible water saturation zone), the presence of
hydrocarbons will be missed by the usual interpretation of the
MDT data. In the case when pressures were measured in the
oil zone the extension of oil gradient will intersect the water

FAREED IQBAL SIDDIQUI

gradient line at the correct FWL. However, the presence of


points falling on the water gradient line above the FWL would
cast doubts on the validity of a perfectly valid interpretation.
Figure 10(b) shows the response for the case when the
probe is responding to the pressure of the dominant phase. As
in the previous case, above the transition zone oil phase
pressures will be measured and the gradient established would
be equal to the static oil gradient. The probe will continue to
read in the oil phase up to some point into the transition zone
(point B in Fig. 10(b)) where water becomes the more mobile
phase. In a static reservoir the pressures measured until this
point will continue to follow the static oil gradient line. Below
this point in the transition zone and into the 100% water zone
the probe will record the pressure in the water phase. This
would result in a drop in pressure from point B to C (Fig.
10(b)). For a static reservoir the measured pressures will
establish a gradient equal to static water gradient. As in the
previous case the intersection of the oil and water gradient
lines would intersect at the correct FWL, however the
interpretation will be doubted because of points above the
FWL falling on the water gradient line. In the light of above
discussion for a static reservoir the oil and pressure gradient
lines should be extended to locate the FWL and the pressure
points on the water gradient line above FWL should be
interpreted as the transition zone.
MDT Tools Response in a Dynamic Reservoir. Figures
11(a) and (b) show a schematic of the possible MDT tools
responses for a dynamic reservoir. In a dynamic reservoir the
oil saturation in the hydrocarbon column is S* and the water
saturation is in general greater than Swi. Even for the case of a
good seal when Sw is close to Swi, the water phase exists as a
continuous phase and is mobile. Figure 11(a) gives the
response for the case when the probe is responding to the
wetting phase pressure. The probe will read the pressure in the
water phase both in the oil zone and in the water zone below it.
Pressures measured by the tool are indicated by *. The
pressure gradient through the measured points will however
not be constant, as shown in Fig. 11. As discussed earlier, for a
dynamic reservoir the pressure gradient in the oil zone will be
different from the gradient in the water zone. The pressure
gradient in the oil zone will, in general, be between the static
water gradient and the static oil gradient (Region III, Fig. 7). If
the reservoir has a good seal overlying it, the pressure gradient
in this zone will be equal to the static oil gradient, even though
the pressures measured are in the water phase. This may
mistakenly be interpreted as the pressures in the hydrocarbon
phase. In the transition zone (which is very small) and in the
water zone below it, the measured points will follow the static
water gradient. The intersection of the two gradient lines will
establish the end of the shock zone or the start of the transition
zone for a dynamic reservoir. Since for a dynamic reservoir the
transition zone is very small, the end of the shock will be
termed as the OWC for a dynamic reservoir. In a dynamic
reservoir the oil saturation below the OWC is not at residual

SPE 56513

but is equal to SJ2, which is defined by the influx rate from the
source rock.
Figure 11(b) shows the response for the case when the
probe is responding to the pressure of the dominant phase. In
the oil zone and up to some point in the transition zone(point
B in Fig. 10(b)), the probe will record the pressure in the oil
phase. For a dynamic reservoir the gradient established will in
general vary between the static oil and water gradients. And if
the seal associated with the reservoir is good the gradient will
be closer to the static oil gradient. Below the point in the
transition zone where water becomes the dominant phase and
into the water zone the probe will record the pressure in the
water zone. This is represented by the jump in pressure from
point B to C in Fig 11(b). The gradient established through the
points below point C will be equal to the static water gradient.
From Fig. 11(b) we can see that the oil and water gradient
lines intersect at a point below the OWC, inside the water
zone. Thus, in this case, the true OWC can not be established
by using the MDT pressure data.
Conclusions
1. The pressure and saturation distribution in a dynamically
charged reservoir can be quite different from the usually
assumed static distribution.
2. Assuming static saturation profiles for dynamically
charged reservoirs will result in the underestimation of the
initial hydrocarbon in place.
3. In dynamic reservoirs high water saturation layers well
above the transition zone will produce free water and
should not be perforated.
4. MDT pressure profile interpretations can be quite
different for dynamic and static pressure cases.
5. It may be possible to detect whether a particular reservoir
is dynamic or static if tools that respond to the pressures
in a particular phase are designed.
*
Nomenclature
* [=] means has units of, 'L' is a length unit, 'F' is force, 'm' is
mass and 't' is time.
g
h
j
k
krj
nPc
P
Pc
Pd
Sji
t
u
v
x

=
=
=
=
=
=
=
=
=
=
=
=
=
=
=

Acceleration due to gravity [=] L/t2


Hydrocarbon column length [=] L
Leverett j-function
2
Absolute permeability [=] L
Relative permeability to phase j
Capillary pressure exponent
2
Pressure [=] F/L
2
Capillary pressure [=] F/L
2
Displacement pressure [=] F/L
Irreducible saturation of phase j
Time [=] t
3
Volumetric flux [=] L /t
Specific wave velocities
Distance, positive in the upward direction[=] L
Inclination angle of the medium from the horizontal

SPE 56513

=
=
=
=
=
=
=

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

Difference operator
Porosity
3
Mobility [=] L t/m
Viscosity [=] m/L-t
contact angle
3
Density [=] m/L
2
Interfacial tension [=] m/t

Subscripts
I
= Initial
J
= Injected
o
= Oil
R
= Reservoir
r
= Residual or relative
S
= Seal
t
= Total
w
= Water
Superscripts
+
= Downstream
= Upstream
References
1.
2.
3.

4.

5.
6.

7.
8.
9.
10.

11.
12.
13.

Dake, L. P.. 1994, The practice of Reservoir Engineering:


Elsevier, Amsterdam
Berg, B. B.; 1975, Capillary Pressure in Stratigraphic Traps:
AAPG Bull., v. 59, no. 6, p. 939
Morrow, N. M. and Melrose, J.C., 1990, Applications of
Capillary Pressure Data to the determination of Connate Water
Saturation: in N. R. Morrow (Ed.), Interfacial Phenomenon in
Oil Recovery: Marcel Dekker
Siddiqui, F. I. and Lake, L. W.: A Comprehensive Dynamic
Theory of Hydrocarbon Migration and Trapping, paper SPE
38682 presented at the 1997 SPE Annual Technical Conference
and Exhibition, San Antonio, TX, Oct. 5-8.
Tissot, B. P. and Welte, D. H., 1984, Petroleum Formation and
Occurrence: Springer-Verlag, Berlin
Roberts, W. H. III and Cordell, R. T., 1987, Problems of
Petroleum Migration: Introduction: in W. H. Roberts, III and R.
J. Cordell, (Eds.), Problems of Petroleum Migration, American
Association of Petroleum Geologists Studies in Geology, No.
10: American Association of Petroleum Geologists, Tulsa,
Oklahoma, p. 1
Schowalter, T. T., 1979, Mechanics of Secondary Hydrocarbon
Migration and Entrapment, AAPG Bull., v. 63, no. 5, p. 723
Lake, L. W.,1989, Enhanced Oil Recovery: Prentice-Hall Inc.,
Englewood Cliffs, NJ
Buckley, S. E., and Leverett, M. C., 1942, Mechanisms of Fluid
Displacement in Sands: Trans. AIME, v. 146, p. 107-116
Amyx, J. W., Bass, D. M. and Whiting, R. L., 1960, Petroleum
Reservoir Engineering: McGraw-Hill Book Company, New
York
Leverett, M. C.,1941, Capillary Behavior in Porous Solids:
Trans. AIME, v. 142, p. 152
Corey, A. T., 1986, Mechanics of Immiscible Fluids in Porous
Media: Water Resources Publications, Littleton, CO
Rhee, Hyun-Ku, Aris, R. and Amundson, N. R., 1986, First
Order Partial Differential Equations, Vol. I: Prentice-Hall,
Englewood Cliffs, N. J.

14. Siddiqui, F. I., 1996, A Dynamic Theory of Hydrocarbon


Migration and Trapping, Ph. D. Dissertation: The University of
Texas, Austin
15. Scott, V. H., and Corey, A. T., 1961, Pressure Distribution
During Steady Flow in Unsaturated Sands: Soil Science Society
of America, Proceedings, v. 20, p. 270-274
16. Arbhabhirama, A., Kridakorn, C., 1968, Steady Downward
Flow to a Water Table: Water Resources Research, vol. 4, no. 6,
p. 1249
17. Lujan, C. A., 1985, Three Phase Flow Analysis of Oil Spills in
Partially Water Saturated Soils: Ph. D. Dissertation, Colo., State
Univ., Fort Collins
18. Fetsy, S. and Van Golf-Racht, T. D., 1989, Gas Gravity
Drainage in Fractured Reservoirs Through New DualContinuum Approach: SPE Res. Eng., v. 4, no. 3, p. 271-278

10

FAREED IQBAL SIDDIQUI

SPE 56513

50

60

25

30 dyne/cm
45 dyne/cm

20

40

60 dyne/cm

P at GWC (psi)

40

P (psi)

15

30
20

10

30

20
10

10

wi

GWC height above FWL(m)

50

0.2

0.4

0.6

0.8

10

0
1000

100

k(md)

Oil Saturation, S

Figure 1. Drainage capillary pressure curve for an oil-water


system with ow =10 dyne/cm, k=100md, porosity=0.2, npc =0.3.

50

Figure 2(b). Pc at GWC and GWCs above the FWL for different
permeability reservoirs for, for different values of gw, w = 0.95
g/cc and g = 0.11 g/cc.

350
PcPds

30

250

200

150

20

P at OWC (psi)

40

100

Irreducible Water Saturation


Zone

10
50

0
1

10

100

0
1000

k(md)
Figure 2(a). Pc at OWC and OWCs height above the FWL for
different permeability reservoirs, for different values of ow, w =
0.95 g/cc and o = 0.85 g/cc.

Seal

300

OWC height above FWL(m)

5 dyne/cm
10 dyne/cm
15 dyne/cm
20 dyne/cm

Pw

Po

Pc= gh

Transition
Zone

Transition
Zone
OWC

OWC

FWL

FWL

Pd
P
do= d
g

Swi

So

Water
Zone

Pressure

Figure 3. Schematic of pressure and saturation distribution in a


static reservoir.

SPE 56513

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

11

(a)

Oil Flux, uo

Oil Flux, u o

Reservoir

Dynamic Seal
uoJ

J1

uo- = uo+

J2

uoJ

I
0.0

SJ2 SJ1

T1

0.2

SA

0
(b)

0.4

SA

0.6

S1*

0.8

1.0

Normalized Oil Saturation, S o

Figure 5. Solution for a dynamic reservoir using the method of


characteristics solution technique. 14

Seal 1
(Dynamic)

Capillary Pressure, P c

Seal 2
(Static)

Reservoir

t= t 1

Seal

PdS1
So=0

PdS2

So=SJ2

x1

hstatic

PdR

O il C o lu m n

Reservoir
OWC

0.0

0.2

S*
0.4 2

SA

0.6

S1*

0.8

1.0

Normalized Oil Saturation, S o


Figure 4. Classification of reservoirs as static or dynamic based
on the influx from the source rock, uo-So curve for the reservoir
and the Pc-So behavior of the reservoir and the seal.

So=SJ1

0
0

S*1 1

S J2 S J 1

Oil Saturation, So
(a)

t/ tL

t1
Time, t
(b)

Figure 6. Solution of migration and trapping problem for a


dynamic reservoir on: (a) saturation profile, (b) on time-distance
diagram.

12

FAREED IQBAL SIDDIQUI

t=t 1

SPE 56513

100
Po

Pw
D yna m ic
E ffect

x1

R egion III
R egion II

OWC

0
0S

R e se rvo ir

R egion I

Pd

S *o 1
J2 S J1
O il S atu ration, S o
(a )

Pressure
(b)

Figure 7. Schematic showing different regions for a dynamic


reservoir on the basis dynamic pressure profiles.

Height above FWL (m)

Static

wi

Irreducible Water
Saturation Zone

80

R egion IV

O il C o lum n

1-S

S*

S e al

P ds

60
Dynamic
Reservoir

40

Transition Zone
Static Reservoir

20
d

00

0.2

0.4
0.6
Oil Saturation, S

0.8

Figure 9(a). Comparison of saturation profiles for a static and


dynamic reservoir.

100
P

dS

Irreducible Water
Saturation Zone

Height above FWL (m)

80

60

40

Transition Zone
Static Reservoir

20
d

Dynamic Reservoir

Free Water Level

0
(a)

(b)

Figure 8. Scaled capillary pressure Pc as a function of scaled


elevation for downward flow (a) through a uniform sand (b)
through a sand into another sand of lower permeability, from
Scott and Correy.15

10

20
30
40
50
Capillary Pressure, P (psi)

60

70

Figure 9(b). Comparison of capillary pressure profiles for a static


and dynamic reservoir.

SPE 56513

DEFINING FLUID DISTRIBUTIONS AND FLUID CONTACTS FOR DYNAMICALLY CHARGED RESERVOIRS

13

*
*

*
Pw

Sw=Sirr

*
Pw *
*

Po

Transition
Zone

Region IV

Region II

WOC

*
*

FWL

Region III

WOC

Po

Region I

*
*

*
Pressure

Pressure

(a)

(a)
A

*
*
* Po
*
*B

Pw
C*

Pw
Transition
Zone

Apparent WOC

WOC

FWL

C*

Region IV

*
*
Po
*
*
*
*B
*

Sw=Swirr

Region III

Region II

Region I

*
*

*
Pressure

Pressure
(b)

(b)
Figure 10. Possible MDT tools response for a static reservoir (a)
probe recording wetting phase pressure (b) probe recording
dominant phase pressure

Figure 11. Possible MDT tools response for a dynamic reservoir


(a) probe recording wetting phase pressure (b) probe recording
dominant phase pressure.