Pressure Transistent Analysis
Uploaded by
SubrahmanyamPressure Transistent Analysis
Uploaded by
SubrahmanyamPressure transient analysis
OBJECTIVES AND APPLICATIONS
Pressure transient testing is a form of systems analysis, in which the basic problem is to
indirectly measure one or more characteristics of the reservoir and the wellbore environment.
The approach to the problem may be considered as a forward formulation (Figure 1 ):
Figure 1
We initiate an input function (I) by changing the production or injection rate at the
test well.
This rate change causes a disturbance in the system (S), or reservoir.
The system disturbance is reflected in an output (O), or pressure response, which is
measured at the test well or at an observation well.
The analysis of the problem follows an inverse solution approach. We develop a mathematical
model—usually a partial differential equation—that (1) describes the physics of the system in
terms of reservoir and wellbore properties, and (2) expresses the pressure response as a
function of these properties. The solution to this partial differential equation is represented by the
pressure response. Thus, when we record pressure during a well test, we are actually
measuring the solution to a partial differential equation at a specific location as a function of
time. Because the pressure response is dependent on the reservoir and wellbore properties, we
can use its measured value to determine these properties indirectly. This process is known as
well test analysis or pressure transient analysis.
In analyzing pressure transient data, selection of the proper equation and the accompanying
boundary and initial conditions is of paramount importance. The accuracy of the solution
depends on the model describing physical reality as closely as possible. At the same time, the
model must incorporate certain simplifying assumptions so that a feasible solution can be
obtained using analytical approaches.
Chapter 1 Page 1
ASSUMPTIONS OF THE CLASSICAL WELL TEST MODEL
The following assumptions form the basis for what is generally referred to as classical well test
interpretation:
Homogeneous reservoir with isotropic property distribution and impermeable upper
and lower boundaries
Laminar flow conditions
Single-phase flow
No gravitational or capillary forces
Constant fluid properties throughout the reservoir
Open hole completion with a fully penetrating well
Isothermal conditions throughout the reservoir
Small pressure gradients throughout the reservoir
These assumptions describe a highly idealized system. But by incorporating certain common
deviations from this ideal reservoir model, we can increase the application range of the solution
to a variety of non-ideal situations.
APPLICATIONS
Well testing has long been recognized as a powerful and versatile formation evaluation method.
In recent years, the development of extremely accurate and precise downhole recording
devices, improved analysis techniques and enhanced computing capabilities have increased its
usefulness and areas of application.
Among all formation evaluation methods, well testing is the only one that provides information on
a reservoir’s dynamic behavior, including but not limited to
Permeability
Initial reservoir pressure
Average reservoir pressure
Near-wellbore conditions
Reservoir limits
Inflow performance response
Communication between different parts of the reservoir
RESERVOIR CHARACTERIZATION
Pressure transient data are measured at a scale appropriate for reservoir characterization. The
characterization of a reservoir from a geological perspective helps us understand the reservoir
environment and the controls that its geological architecture exerts on fluid flow regimes and
patterns. This enables us to develop effective strategies for optimizing hydrocarbon recovery.
Reservoir characterization takes place on a number of levels, depending on the purpose of the
study (e.g., searching for improvements in reservoir management, planning an infill drilling
program or exploring options for implementing improved oil recovery projects). Table 1 below
summarizes the elements of a reservoir characterization study:
Table 1: Elements of a Reservoir Characterization Study
Chapter 1 Page 2
Geological characterization Basic description
Reservoir age
Mode of deposition
Type of hydrocarbon trap
Structural description
Reservoir limits
Formation thickness
Aquifer support
Petrophysical Petrophysical model
characterization
Water Saturation
Water sample analysis
Capillary pressure data
Porosity-permeability crossplots
Net pay cut-off values
Oil-water, gas-oil contacts
Wettability characteristics
Relative permeability
characteristics
Rock compressibility
Fluid property Reservoir pressure and
characterization temperature
PVT data
Compositional analysis
Fluid compressibility
Fluid viscosity
Fluid densities
Pressure transient data may be used to derive such reservoir characterization parameters as
horizontal permeability, reservoir architecture (including reservoir limits, reservoir zonation and
layering) and formation type (e.g., double-porosity, double-porosity/double-permeability). In
interpreting these data, the engineer, geologist and geophysicist must develop a good
understanding of each other’s different approaches. While pressure transient tests can provide
Chapter 1 Page 3
significant clues towards solving the reservoir puzzle, analysis of test data needs to be tempered
with the input of the other team members.
RESERVOIR MANAGEMENT
During its life, a reservoir typically passes through various phases of appraisal, planning,
development and surveillance. At each succeeding phase, we require increasingly descriptive
information. Well test analysis is a pragmatic tool for providing some of this information.
During the appraisal phase, a properly designed, conducted and analyzed well test can
provide information on reserves, production rates and aquifer size.
During the planning phase, well tests can be helpful in formulating an optimal production
strategy and determining the required locations and numbers of wells and/or platforms.
During the development phase, well tests aid in establishing completion and workover
policies and drainage areas for each well.
During the surveillance phase, well tests are helpful in updating reservoir and well
characteristics, analyzing workovers, and revising production and injection plans to improve
performance.
Figure 2 summarizes the practical information obtainable from pressure transient testing
and the utilization of this information in reservoir management.
Figure 2
Understanding the Wellbore Environment
A properly designed and analyzed well test can provide valuable insights into the condition
of the wellbore and the near-wellbore environment. These insights are based on an
understanding of deviations from the classical well test model.
Classical well test analysis assumes ideal conditions at the wellbore, such that test data are
controlled only by reservoir characteristics. This assumption permits us to mathematically
specify the inner boundary condition as follows:
When a well is produced at a specified constant flow rate, that rate is achieved
instantaneously at the wellbore.
Chapter 1 Page 4
Similarly, if a well is produced at a specified constant sandface pressure, that pressure
is achieved instantaneously at the wellbore.
These simplifications are for mathematical convenience. In reality, conditions at and near
the wellbore are never ideal, and they do influence the collected data, especially during the
early part of the test. These influencing conditions include:
Wellbore storage (or unloading): Because of wellbore fluid compressibility, flow from
the formation does not immediately cease when a well is shut in at the surface. Fluid
influx continues for some time following shut-in. This phenomenon is known as wellbore
storage. Similarly, when a well is put on production at the beginning of a drawdown test,
fluid flow from the reservoir into the wellbore is initially zero. It thereafter increases to a
specified rate over time. This is referred to as wellbore unloading. The duration of
storage and unloading effects depends on the tubing volume (and the annular volume, if
there is no packer), the wellbore fluid compressibility and viscosity, and the formation
permeability and thickness.
Phase segregation: In pressure build-up tests, segregation of gas in the tubing during
shut-in can cause the measured shut-in pressure to rise above the reservoir pressure. A
pressure decline usually follows this pressure rise. This effect will be pronounced if the
annulus is isolated with a packer. If the well does not have a packer, the annulus
provides additional volume for the gas to expand without causing any significant
pressure increase.
Partial penetration of the wellbore into the producing formation (Figure 3 ): If a well
does not completely penetrate the formation, then the radial flow regime will be distorted
as hemispherical flow starts.
Figure 3
However, this change in flow geometry takes place only in the immediate wellbore
vicinity. At some distance from the wellbore, radial flow resumes. Thus, in a partially
penetrating well, the test duration should be long enough to permit the development of
radial flow away from the wellbore. Distortion of the streamlines will result in an extra
pressure drop within the flow field, which will create an effect on the pressure transient
data similar to that of formation damage.
Restricted entry due to perforations: In cased-hole completions, the perforations
establish a limited connection between the wellbore and the formation. This limited entry
to flow distorts the radial-cylindrical flow geometry and decreases the well’s productivity.
In the immediate vicinity of the perforations, the flow geometry is almost perfectly
spherical. Pressure transient data from a perforated well might indicate formation
damage where none exists, or a higher degree of damage than what is actually present.
Chapter 1 Page 5
This may lead to such incorrect decisions as planning a well stimulation program when
it is not necessary. Figure 4 shows the increased intensity of streamlines around the
perforations.
Figure 4
Hydraulic fractures: Hydraulic fractures complicate the analysis of pressure transient
data, because the flow geometries that are encountered change during the course of
the test (Figure 5 ). It thus becomes necessary to apply analysis procedures that
conform to the flow geometries.
Chapter 1 Page 6
Figure 5
Formation damage/stimulation: Drilling and completion operations alter formation
permeability near the wellbore .A permeability reduction is referred as formation
damage, and permeability enhancement is referred as stimulation. In either case, this
alteration forms a composite system around the wellbore (Figure 6 ).
Figure 6
Chapter 1 Page 7
If the permeability in the inner zone is less than the original permeability of the formation
(i.e., the near-wellbore region is damaged), the observed pressure drop at the wellbore
is greater than it would be if no damage was present. If permeability of the inner zone is
enhanced (e.g., by acidizing or hydraulic fracturing) then the observed pressure in the
wellbore will be less than that predicted by the classical well test model.
It is important to understand how these conditions affect test data. Durations of these
effects will vary from one wellbore environment to another, but once they are identified,
it is possible to implement the correct analysis procedures.
BASIC CONCEPTS OF PRESSURE TRANSIENT
ANALYSIS
There are a number of concepts that are basic to the formulation and practical application of the
classical well test model. These include the following:
Infinite-acting reservoir
Finite reservoir with closed boundaries
Finite reservoir with constant pressure boundaries
Effective drainage radius
Radius of investigation
Time to stabilization
Wellbore storage
INFINITE-ACTING RESERVOIR
The concept of an infinite-acting or infinitely large reservoir is used in implementing the outer
boundary condition for the classical well test model. It describes the condition where measurable
pressure transients have not traveled far enough to reach the closest reservoir boundary. In
a radial-cylindrical flow geometry, the infinite-acting reservoir concept allows us to
impose the outer boundary condition that in an infinitely large system, there is no
pressure change at the external boundary:
As r à ∞ p = pi For t ≥ 0
This boundary condition simplifies the analytical solution. Moreover, it requires us to use the
solution of the classical well test model (the exponential integral solution) only during the initial
stages of testing, before the measurable pressure disturbance has reached the reservoir
boundaries.
In reality, no reservoir is infinitely large. This concept does not imply, however, that the classical
well test model is unrealistic. No matter how small a reservoir is, there will always be an infinite-
acting period during which no measurable pressure change is observed at a closest point on the
boundary. Figure 1 compares the duration of infinite-acting flow periods for various reservoir/well
configurations (all of the reservoirs are assumed to have the same hydraulic diffusivity).
Chapter 1 Page 8
Figure 1
Figure 2 shows the expected pressure profiles for an infinite-acting reservoir.
Figure 2
Note that the reservoir’s external boundary, located at re, has not experienced any pressure
drop, because the pressure transient at time t=t4 has yet to reach the outer boundary.
FINITE RESERVOIR WITH CLOSED
BOUNDARIES
Chapter 1 Page 9
The idea of a finite reservoir with closed boundaries applies to a volumetric reservoir: a
closed system in which there is no fluid flow across the outer boundaries. In other words,
pressure gradients across the system’s outer boundaries are zero.
Closed boundaries may result from physical permeability barriers such as sealing faults or
permeability pinchouts. A no-flow boundary may also form because of a large number of
well patterns drilled in a homogeneous reservoir segment. In such a case, the hypothetical
no-flow boundary corresponds to the drainage area of a given well. In Figure 3 , for
example, the wells are much more closely spaced along the y-direction, because the
formation permeability in that direction is significantly less than the permeability in the x-
direction).
Figure 3
Figure 4 illustrates the pressure profiles in a finite reservoir with no-flow boundaries before and after the infinite-
acting period.
Chapter 1 Page 10
Figure 4
It also shows the pressure transient signifying that a measurable pressure drop has reached the outer boundary
sometime between t4 and t 5, where t4 < t < t 5. Note how the gradients of the pressure profiles become smaller as
the reservoir is depleted.
FINITE RESERVOIR WITH CONSTANT PRESSURE BOUNDARIES
Constant pressure boundaries are characteristic of reservoirs connected to aquifers (i.e., bottom water drive or
edge water drive), or those with associated gas caps. In these types of reservoirs, pressure remains constant at
aquifer and/or gas cap interface(s).
Hypothetical constant pressure boundaries can be formed if a production well is surrounded by injection wells, as
shown in the waterflooding pattern of Figure 5 .
Chapter 1 Page 11
Figure 5
When a pressure transient reaches the outer boundary of a constant pressure boundary reservoir, the flow across
the boundary must be equal to the flow through the well. Thus, the pressure distribution does not change over
time, and the reservoir attains true steady-state conditions (Figure 6 ).
Figure 6
Under these conditions, the wellbore experiences no further pressure decline, and the pressure at the external
Chapter 1 Page 12
boundary remains at the initial pressure. Note that the tested well experiences infinite-acting flow behavior until
these true steady-state conditions develop.
EFFECTIVE DRAINAGE RADIUS
Effective drainage radius is a hypothetical concept that forces the steady-state solution to represent the transient
solution. Consider, for example, a well producing from a finite reservoir for a certain period such that significant
depletion takes place. If this well is shut in for a sufficiently long period, the reservoir pressure ultimately stabilizes
at an average reservoir pressure of pR (in an infinite-acting reservoir, pR is very close to the initial pressure pi , and
in a finite reservoir pR is equal to pi).
Figure 7 gives a graphical interpretation of effective drainage radius.
Figure 7
If a well is shut in after a moderately short production period, the pressure profile at the shut-in time is considerably
higher and the resulting average reservoir pressure is higher as well. The average reservoir pressure line and the
reservoir pressure profile curve at the shut-in time therefore cross each other at a distance closer to the wellbore.
This indicates that initially, the effective drainage radius starts near the wellbore, and moves away from the
wellbore as production continues. The movement of the effective drainage radius away from the wellbore slows
with time as the reservoir pressure profile becomes flatter. At one point, the movement comes virtually to a stop,
defining the well’s effective drainage radius.
Remember that effective drainage radius is a hypothetical quantity that has no physical reality, and that drainage in
fact takes place from the entire reservoir—albeit at small rates from the sections of the reservoir that lie beyond the
effective drainage radius.
RADIUS OF INVESTIGATION
The radius of investigation is the distance traveled by a measurable pressure transient, as measured from the
tested well.
Consider a test well on which a rate change is imposed, and an observation well located a distance r inv from the
test well. Assume that pressure in the observation well is monitored by a pressure gauge with a precision of 0.1
psi. [.69 kPa] Thus, when we see a 0.1 psi [0.69 kPa] pressure change at the observation well, we can say that at
that time the radius of investigation corresponds to the distance between the test well and the observation well.
The radius of investigation depends on the hydraulic diffusivity constant ( of the reservoir rock:
k
c
Chapter 1 Page 13
where k = permeability,
= porosity,
= viscosity and
c = compressibility
The hydraulic diffusivity constant determines the speed at which pressure transients propagate through the
reservoir. Note that although pressure transients are introduced into the reservoir by perturbing the flow rate at the
well, the speed of the pressure transients does not depend on the magnitude of flow rate changes. The flow rate at
the well simply scales the magnitude of the pressure change (drop or increase) at a given point and time in the
reservoir.
The following equation provides a mathematical description for the radius of investigation:
2 2.637 10 4 kt
2
r for r r
inv c 2
1 inv e
Obviously, the radius of investigation cannot exceed the external radius of the reservoir. When the radius of
investigation has reached the furthest point on the external limit of the reservoir, then it is safe to assume that the
well has communicated with every point in the reservoir.
The radius of investigation concept is very important in helping the well test design engineer plan the duration of
the test.
TIME TO STABILIZATION
When the point on the outer boundary that is farthest from a well experiences a measurable pressure drop, the
well is said to be stabilized. If we consider a well located in the center of a circular reservoir (i.e., all the points on
the external boundary are located at an equal distance from the well), then the stabilization time will be the time
required for the measurable pressure transient to reach the external boundary. Stabilization time is a useful
concept in that—in principle—it marks the end of a well test.
We can now define three important flow periods, or regimes, that are encountered during a well test (Figure 8 ).
Figure 8
Chapter 1 Page 14
1. Infinite-acting or transient flow period: Starts at the beginning of the test, and ends when the closest
point on the external boundary experiences a measurable pressure drop (end of the infinite acting
period).
2. Late-transient flow period: Starts at the end of the transient flow period and continues until the furthest
point on the external boundary experiences a measurable pressure drop (stabilization time).
3. Stabilized or pseudo-steady state flow period: Every point in the reservoir experiences a change in
pressure as a linear function of time.
The following equation provides a mathematical description for the stabilization time for a well located in the center
of a circular reservoir.
1 cre2
ts
4 2.637 10 4 k
Where ts = stabilization time, hours
= porosity
= viscosity, cp
c = effective compressibility, psi-1
re = drainage radius, ft
k = permeability, md
WELLBORE STORAGE
When we conduct a pressure buildup test, we perturb the reservoir by shutting in a flowing well; on a drawdown
test, we perturb the reservoir by putting a shut-in well on production. In either case, we are imposing a significant
rate change at the well. Because this rate change is usually imposed at the wellhead, and because the wellbore
fluid is not incompressible, the pressure disturbance that accompanies the rate change does not instantaneously
transmit to the sandface.
Thus, in a buildup test, production from the reservoir into the wellbore does not immediately cease when the well is
shut in at the surface. This continuation of flow from the sandface into the well is called afterflow. By the same
token, when a well is put on production at the surface during the initial flow period of a drawdown test, production
first comes from the expansion of fluid(s) contained in the wellbore. In other words, the sandface flow rate is
initially zero and then gradually increases to the wellhead production rate. This phenomenon is known as
unloading. Figure 9 shows schematically the delayed response of the sandface to the wellhead-controlled rate
changes in buildup and drawdown tests.
Chapter 1 Page 15
Figure 9
We may examine the wellbore storage concept more closely by considering a volumetric balance over the
wellbore:
[wellhead production rate] = [sandface rate] + [wellbore depletion rate]
Wellbore depletion can be expressed in terms of the wellbore storage constant cs:
d
q wb c s p i p wf
dt
The wellbore storage constant can be defined using the compressibility of the wellbore fluid evaluated at the mean
wellbore pressure and temperature:
cs = cfVw
where cf is the wellbore fluid compressibility and Vw is the total wellbore volume. Therefore, the wellbore storage
constant has the units of bbl/psi if Vw is in bbl and cf is in psi-1.
Chapter 1 Page 16
EXERCISE 1
Well tests are conducted for exploration and reservoir characterization purposes. Does an
exploration test or a reservoir characterization test run longer? Why?
EXERCISE 1—SOLUTION
Exploration tests attempt to provide answers to questions such as:
Should a zone be completed?
Should a platform be set to develop this field?
In answering these questions, it will be necessary to generate information on reservoir size and
well deliverability. Therefore, in an exploration test, the reservoir as a whole is targeted. On the
other hand, in characterizing a reservoir, the drainage area of a well is put under the
microscope. Therefore, tests run to measure the significant reservoir properties to characterize
the reservoir are often shorter than those for exploration.
EXERCISE 2
Identify the reservoir and fluid properties that control the wellbore storage duration directly and
inversely.
EXERCISE 2—SOLUTION
Identify the reservoir and fluid properties that will control the wellbore storage duration directly
and inversely.
The longer the well interval, longer the wellbore storage period.
The larger the compressibility of the wellbore fluid, longer the wellbore storage period.
The larger the fluid viscosity, the longer the wellbore storage period.
The larger the permeability of the formation, shorter the wellbore storage period.
The larger the thickness of the formation, shorter the wellbore storage period.
EXERCISE 3
Mark the following statements as TRUE or FALSE. In either case, justify your answer.
1. In a finite reservoir, the duration of the transient flow period is expected to be shorter if
well is produced at a higher rate.
2. Consider two circular reservoirs, one with a no-flow outer boundary and one with a
constant pressure outer boundary. Assume that both reservoirs have similar rock and
fluid properties. If each of these reservoirs has one well located at the center of the
drainage area at a given time, the radius of investigation will be larger for the well
located in the reservoir with no-flow boundaries.
3. Consider a gas reservoir and an oil reservoir with identical reservoir properties and well
parameters. Duration of wellbore storage is expected to be longer in the gas well.
4. The rate of growth of the drainage radius in a gas reservoir is expected to be faster
than that of an oil reservoir, assuming all the reservoir rock properties and well
parameters are the same in both reservoirs.
Chapter 1 Page 17
EXERCISE 3—SOLUTION
1. FALSE
Duration of infinite acting period (transient period) is determined how long for a
measurable pressure transient will take to reach the closest boundary point. This is
controlled by the hydraulic diffusivity constant, not by the flow rate.
2. FALSE
Again, the position of the radius of investigation at a given time is controlled by the
permeability, porosity of the formation and by the viscosity and total compressibility of
the formation/fluid system.
3. TRUE
Since the compressibility of gas is larger than that of oil, duration of wellbore storage will
be longer in gas wells.
4. FALSE
In a typical gas reservoir, hydraulic diffusivity constant is expected to be much smaller
than that of a liquid reservoir (compressibility of gas is 3 to 4 orders of magnitude larger
while the viscosity of gas is 2 orders of magnitude smaller). Therefore, the rate of growth
of drainage radius in a gas reservoir will be slower.
EXERCISE 4
Consider two reservoirs A and B with identical rock and fluid properties and drainage areas as
shown in Figure 1 . Compare the expected duration of the infinite acting and late-transient flow
periods for both reservoirs.
Figure 1
EXERCISE 4—SOLUTION
In Reservoir B, the closest boundary is only 500 feet away (in Reservoir A, the closest boundary
is 1000 feet away). Therefore, the duration of the infinite acting period will be shorter for the well
in Reservoir B. However, duration of the late transient period will be much longer in Reservoir B
because the furthest points on the boundaries (corners of the reservoir) are at a larger distance
than that of the Reservoir A. Therefore, it will take longer for the well in Reservoir B to reach the
stabilized conditions.
Chapter 1 Page 18
Differential Flow Equations for Pressure Transient
Analysis
INTRODUCTION
The partial differential equations that describe the flow of fluids through porous media are
derived from three physical principles:
1. Conservation of Mass
2. Darcy’s Law
3. Equations of State
The general flow equation is derived from the Conservation of Mass principle and Darcy’s Law,
and applies regardless of fluid phase. The equations of state are specific to individual fluid
phases, and describe the relationship between fluid density, pressure and temperature (in the
classical well test model, isothermal conditions are assumed, and the equations of state are
used to relate density and pressure only).
FLOW GEOMETRY
A necessary first step in representing fluid flow dynamics in porous media is to select the
appropriate flow geometry or coordinate system. The most widely used coordinate systems are
listed below:
Coordinate System Area o
Application
Rectangular Multi-well fie
(Figure 1 ) simulation
studies in 1D
2D and 3D
Figure 1
Chapter 1 Page 19
Radial-Cylindrical Single we
(Figure 2 ) studies
involving
pressure
transient
analysis
(PTA). Mo
of th
applications
are in 1D (
direction
only). 2D (r
direction)
applications
are f
hydraulically
fracture well
3D (r-q
directions)
applications
are layere
systems.
Figure 2
Spherical Special
(Figure 3 ) applications
such a
studying flo
characteristic
around th
immediate
vicinity
perforations.
Figure 3
Chapter 1 Page 20
Elliptical-Cylindrical Single we
(Figure 4 ) studies wi
infinitely
conductive
vertical
hydraulic
fractures.
Figure 4
Most well test applications use radial-cylindrical flow geometry to mathematically describe the
flow problem. This choice is mainly controlled by the simplifying assumptions inherent in this
geometry. However, by using certain well-established transformation rules, it is possible to
transform mathematical formulations as expressed in one coordinate system to their equivalent
expressions in other systems. (For a detailed discussion of transformations, go the IPIMS
subtopic titled Pressure Transient Analysis—Mathematical Tools, which is found under the topic
of Advanced Pressure Transient Analysis.)
GENERAL EQUATION FOR SINGLE-PHASE FLOW
The principle of Conservation of Mass, when applied to a control volume, states that over a
given period of time t:
[net flow of mass out] -[net flow of mass in] = [net change of mass]
Figure 1 represents this principle graphically. Note that the shape of the control volume must
conform to the coordinate system used in writing the transport equation. Furthermore, the sign
convention indicates that
Chapter 1 Page 21
Figure 1
[mass flow out] > 0
[mass flow in] < 0
This implies that
If [mass flow out] > [mass flow in], then [net change of mass]>0 => [mass
depletion]
If [mass flow out] < [mass flow in], then [net change of mass]< 0 => [mass
accumulation]
The Continuity Equation in three-dimensional radial-cylindrical coordinates, obtained after
applying the conservation of mass principle to a control volume in cylindrical coordinates, is
1
rv r 1 rv v z
r r r z t
If we consider the above equation in one dimension only—the r-direction—it simplifies to
1
rv r
r r t
which represents the basic form of the classical well test model.
To make the continuity equation strictly applicable to flow in porous media, we must introduce
Darcy’s law:
k r p
vr
r
The continuity equation in 1-dimensional radial cylindrical coordinates then becomes
1 k r p
r
r r r t
Chapter 1 Page 22
This equation is only applicable to flow in porous media.
SINGLE-PHASE FLOW: SLIGHTLY COMPRESSIBLE
FLUIDS
A slightly compressible fluid is one that exhibits constant compressibility (an accurate description
for liquids but not for gases). Using the definition of compressibility,
1 1
c or
p c p
we can alter the left and right hand sides of the continuity equation as follows:
1 k r
r c
r r r t
In practical applications, it is much more convenient to measure pressure than it is to measure
density. We therefore prefer to write the above equation in terms of p rather than .
Furthermore, if we assume that the reservoir properties kr and are constant, and that the
viscosity of a liquid with small compressibility does not change with pressure, then the slightly
compressible flow equation takes the following form (letting k=kr):
1 p c p
r
r r r k t
In the above expression, we have assumed that is not a function of pressure. If porosity is treated
as a function of pressure, the slightly compressible flow equation becomes
1 p c t p
r
r r r k t
In the above equation, ct represents the total compressibility of the liquid and the pore volume
compressibility (formation compressibility) such that
ct = c + cf
The coefficient on the right hand side of the partial differential equation represents one of the most
important properties of a reservoir, as the group
1
c t
k
is identified as the hydraulic diffusivity constant of the reservoir. This constant controls the speed
at which pressure transients propagate in the reservoir. A dimensional analysis conducted on
the hydraulic diffusivity group yields
Chapter 1 Page 23
L2 L2
2
M Lt t
Lt M
which is consistent with the dimensions of the diffusivity constant (i.e., area covered per unit
time).
SINGLE-PHASE FLOW: COMPRESSIBLE FLUIDS
The principle assumption upon which we base the slightly compressible flow equation—that
compressibility is a constant—is reasonably accurate for liquids, but not for gases. The
compressibility and density of a real gas vary with pressure as described by the following
equations:
1 1 z PMW
c and (real gas law)
p z p zRT
This property of gases introduces a relatively strong non-linearity to the transport equation as
indicated below:
1 p p p p
r
r r z r t z
Non-linearity is introduced to the left-hand side of the equation through the group (p/z). On the
right-hand side of the equation, the term (1/z) introduces additional non-linearity. The strong
pressure-dependency of and z makes application of an analytical solution unfeasible. We can,
however, apply a transformation known as real gas potential or pseudo-pressure to linearize the
compressible flow equation.
DIFFERENT FORMS OF GAS FLOW EQUATION
In linearizing the gas flow equation, we introduce a new variable m(P), defined by
p
p
mp dp
0
z
where m(p) is called the real gas potential and has units [psi2/cp]. During the transformation
process, we need to observe the following relations:
mp p
(i)
p z
mp p p
(ii)
r z r
Chapter 1 Page 24
mp p p
(iii)
t z t
Substitution of (ii) and (iii) into the compressible flow equation gives:
1 mp c mp
r
r r r k t
The above equation is quasi-linear because of the dependency of and c on pressure. In well
testing applications, and c are evaluated either at the initial pressure or mean pressure.
REAL GAS POTENTIAL FORM OF THE COMPRESSIBLE FLOW
EQUATION
1 mp i c i mp
r
r r r k t
This form of the compressible flow equation is especially used at intermediate pressures (1000-
4000psia [6895 -27580 kPa]) where the variation of z with pressure is strongly non-linear (see
Figure 1 ).
Figure 1
Chapter 1 Page 25
PRESSURE-SQUARED FORM OF THE COMPRESSIBLE FLOW
EQUATION
At low pressures (<1000 psia [6895 kPa]), the z product of a hydrocarbon gas remains
constant (see Figure 1). In this low pressure range, the real gas potential transformation
simplifies to:
p
p 1
mp zdp 2z p
2
Applying the above transformation to the compressible flow equation yields the pressure-
squared form of the compressible flow equation as
1 p 2 c p 2
r
r r r k t
PRESSURE FORM OF THE COMPRESSIBLE FLOW EQUATION
At high pressures (>4000 psia [27580 kPa]), the variation of z with p exhibits a linear
relationship. In other words, the p/z representing the slope of this linear relationship remains
constant (see Figure 1). Thus, the real gas potential transformation simplifies to
p
p p
mp zdp z
0
Applying the above transformation to the compressible flow equation, we obtain
1 p c p
r
r r r k t
The pressure form of the equation is similar to the slightly compressible (liquid) flow equation.
This is expected because at high pressures, gases start to behave like liquids.
BOUNDARY AND INITIAL CONDITIONS
To complete the description of a mathematically well-posed problem, we must specify the
boundary and initial conditions imposed on the system. Here we will limit the discussion to initial
and boundary conditions that are commonly associated with the classical well test analysis
model.
INITIAL CONDITIONS
The classical well test analysis formulation requires an initial condition which states a uniform
pressure distribution over the entire domain at t=0.
BOUNDARY CONDITIONS
The inner boundary condition (condition specified at the wellbore) states that the well is
produced at a constant rate.
Chapter 1 Page 26
The outer boundary condition (condition specified at the reservoir limits) requires that the
reservoir is infinitely large (i.e., outer reservoir boundaries are considered to be at infinity) so that
there is no pressure drop at the external boundaries at anytime while the well is on production.
If we apply a finite reservoir formulation, then the outer boundary condition can be either a no-
flow boundary (i.e., pressure gradient across the boundary is equal to zero) or a constant
pressure boundary, which implies a strong communication between the aquifer and the reservoir
across the outer boundaries.
SINGLE-PHASE FLOW REGIMES
In a reservoir with closed outer boundaries, we observe three flow regimes:
Transient: This flow regime takes place during the infinite-acting flow period.
Classical well test analysis can be applied only during this period. This flow period
is also known as ‘early transient state’.
Late Transient: When the presence of the closest boundary is felt at the wellbore,
the transient flow period ends and the late transient flow period starts. The late
transient period ends when the pressure transient moving outwards reaches the
furthest point on the boundary.
Pseudo-Steady State: The end of the late transient flow period marks the start of
pseudo-steady state flow. During this flow period, reservoir pressure will decline at
the same rate everywhere in the drainage region. The pseudo-steady state period
is also referred as the ‘semi-steady state’ or ‘quasi-steady state’.
In reservoirs with constant pressure boundaries, a fourth flow regime—Steady State—will
ultimately develop. When true steady-state conditions are established, the pressure profile
attained in the reservoir will stay intact as long as the boundary conditions are not changed.
MULTI-PHASE FLOW CONDITIONS
For multi-phase flow conditions, it is necessary to write the transport equations separately for
each phase, because multi-phase interactions make the assumptions that are set forth for
single-phase flow invalid.
The presence of more than one flowing phase in a porous medium reduces the flow rate of each
well. Moreover, the relative permeability of each phase—and therefore the net effective mobility
—are functions of fluid saturations that change over time. But if the test is of sufficiently short
duration, we may consider the changes in average pressure and saturation to be negligible. And
if multi-phase flow is occurring only in the immediate vicinity of the wellbore, we may be able to
use single-phase analysis by incorporating the multi-phase flow effects into the composite skin.
(Of course, we may also simplify the problem by testing early in the life of the well, before multi-
phase flow effects become dominant.)
A more accurate way to analyze multi-phase flow data is to history match the test data using a
numerical model. A detailed history match will generate absolute permeability and porosity
values as well as relative permeability and saturation relationships.
TWO-PHASE (LIQUID-LIQUID) FLOW
OIL/WATER EQUATIONS:
Oil phase equation:
Chapter 1 Page 27
1 k ro p S o
r
r r oB o r k t B o
Water phase equation:
1 k rw p S w
r
r r w B w r k t B w
Auxiliary equation:
So + Sw = 1.00
TWO-PHASE (LIQUID-GAS) FLOW
Liquid-gas multi-phase flow conditions arise when the reservoir pressure is initially below the
bubble point, or if it decreases below the bubble point during the test. The presence of solution
gas further complicates the analysis. Therefore, it is quite probable that a numerical model will
be used to analyze a well test that is conducted as the reservoir pressure is passing through the
bubble point. Liquid-gas flow equations are as follows:
OIL/GAS EQUATIONS:
Oil phase equation:
1 k ro p S o
r
r r oB o r k t B o
Gas phase equation:
1 k rg k p S g S
r R so ro R so o
r r gB g oB o r k t B g B o
Auxiliary Equation:
So + Sg = 1.00
THREE-PHASE (LIQUID-LIQUID-GAS) FLOW
We generally do not encounter three-phase flow in well test analysis, and when three-phase flow
conditions are present, we consider only the most dominant phases. Nonetheless, for the sake
of completeness, we present the three-phase flow equations as follows:
OIL/WATER/GAS EQUATIONS:
Oil phase equation:
1 k ro p S o
r
r r oB o r k t B o
Water phase equation:
Chapter 1 Page 28
1 k rw p S w
r
r r w B w r k t B w
Gas phase equation (assuming Rsw = 0):
1 k rg k p S g S
r R so ro R so o
r r g B g o B o r k t B g B o
Auxiliary Equation:
So + Sw + Sg= 1.00
Note that in the gas equation, we assume that the solubility of gas in the water phase is
negligible. Therefore, the transport and storage of gas in the water phase are not represented.
PSEUDO REPRESENTATION OF MULTI-PHASE FLOW
The classical well test analysis model assumes single-phase flow conditions. But in reality, most
wells are tested under two-phase or three-phase flow conditions. To model these actual
conditions, we would need to simultaneously solve the set of partial differential equations that
describe multi-phase flow systems.
As an alternative to this rigorous approach, we can apply some simplified analysis methods for
multi-phase well tests, while at the same time keeping in mind that these methods can lead us to
underestimate or overestimate such reservoir and well parameters as effective permeabilities
and skin factors.
PERRINE’S APPROACH (PRESSURE APPROACH):
In this approach, we define total flow properties to replace single-phase flow properties such as
mobility and compressibility. The total mobility of a multi-phase flow system, for example, is
defined as a sum of the mobilities of individual phases.
k k kg
o
t o g
In the above equation ko, kw and kg represent the effective permeabilities to each phase. They
can be written in terms of relative permeabilities as
ko = kkro; kw = kkrw; kg = kkrg
The total compressibility is defined in a similar fashion as
(c)t = Socoa + Swcwa + Socoa + cr
In the above equation, coa and cwa are the apparent compressibilities for the oil and water
phases, which also contain a certain amount of dissolved gas. Therefore, the definitions of the
compressibilities for each phase can be written as (see Exercise No.2 under this subtopic
heading Multi-Phase Flow Equations:
B g R so 1 B o
c oa
B o p B o p
B g R sw 1 B w
c wa
B w p Bw p
Chapter 1 Page 29
1 B g
c
g
B g p
Note that the terms
1 B o 1 B w
and
Bo p Bw p
represent the compressibilities of the oil and water phases when there is no dissolved gas
present. Accordingly, the first terms in the brackets that define coa and cwa scale the volume
change due to gas dissolution in the oil and water phases, respectively, upon pressure change.
Observe also that for pure oil and water phases (with no dissolution of gas)
B o B w
and
p p
are negative quantities, therefore:
coa > co and cwa > cw .
This is why multi-phase systems containing dissolved gas have large compressibilities; they may
even exceed the compressibilities of systems where only gas is present.
In interpreting multi-phase test data, it is also necessary to express production in terms of the
total fluid production rate:
(q)t = (qoBo)+(qwBw)+(qg-Rsoqo-Rswqw)Bg
where qgBg represents the total gas flow rate,
qoRsoBg represent the solution gas rates in the oil phase, and
qwRswBg represent the solution gas rates in the oil water phases, respectively.
The total mobility and compressibility terms introduce non-linearities to the analysis. These non-
linearities can be removed by introducing assumptions such as:
Saturation changes during the test are negligible
Pressure gradients are small
Saturation gradients are small
No distinction between phase pressure (capillary pressures are negligible)
Perrine’s approach is sensitive to gas saturation and becomes attenuated as the gas saturation
increases. In general, Perrine’s method underestimates the effective permeability values and
overestimates the skin. In order to calculate the absolute permeability (recognizing that the
analysis procedure provides an estimate of the total mobility) one can use the relative
permeability characteristics at the saturation conditions that are assumed stationary during the
testing period.
k k k kg
o w
t o w
g
AL-KHALIFAH ET AL.’S APPROACH (PRESSURE-
SQUARED APPROACH)
This approach assumes that the mobility of the oil phase is a linear function of pressure.
Chapter 1 Page 30
ko
p
B
o o
where a is the proportionality constant. When this assumption is made, a transformation such as:
ko p 2
mp dp pdp
oB o 2
converts the equation into the pressure-squared form. The value of is evaluated at initial
reservoir pressure for drawdown tests and at average reservoir pressure for buildup tests.
RAGHAVAN’S APPROACH (PSEUDO-PRESSURE
APPROACH)
This approach establishes a parallelism with the real gas potential approach developed for gas
wells. The definition of multi-phase pseudo-pressure is
p
ko
mp dp
0
oB o
It is necessary to have the relative permeability relationships in place to carry the transformation
given above. This approach is found to be highly sensitive to the relative permeability data.
Therefore, its application is usually limited to situations where accurate relative permeability data
are available.
EXERCISE 1
State two examples of flow conditions where elliptic-cylindrical flow geometry will develop.
EXERCISE 1 SOLUTION
A well connected to an infinite conductivity vertical fracture will develop elliptic
equipotential lines and hyperbolic streamlines. All of the ellipses and hyperbolas will
be confocal. The foci will be located at the two tips of the fracture. In this
description, the well is considered simply a part of the hydraulic fracture and is not
different from the fracture.
In the presence of strong anisotropy kx>>ky (or ky>>kx) again elliptic flow geometry
will develop. The propagation of pressure transients will be much more rapid along
the more permeable direction. Therefore, the major axis of the ellipse will be in the
same direction with the high permeability direction.
EXERCISE 2
Show explicitly how the compressibility term appears on the right-hand-side of the compressible
flow equation.
EXERCISE 2 SOLUTION
The original entry on the right-hand-side of the equation is:
p
RHS
t z
Expanding the time derivative, one obtains:
1 p 1
RHS p
z t t z
Chapter 1 Page 31
1 p 1 p
RHS p
z t p z t
1 1 p
RHS p
z p z t
1 p z p p 1 1 z p
RHS 2 RHS
z z p t z p z p t
p p
RHS c
z t
EXERCISE 3
Rewrite the pseudo representation of multi-phase flow for Perrine’s approach.
A. For a solution gas drive reservoir.
B. For a three phase reservoir in which dissolution of gases water phase is negligible.
EXERCISE 3 SOLUTION
A.
k k kg
o
t o g
(c)t = (Socoa)+(Swcw)+(Sgcg)+cr
(q)t = (qoBo)+(qg-Rsoqo)Bg
B.
k k k kg
o w
t o w
g
(c)t = (Socoa)+(Swcw)+(Sgcg)+cr
(q)t = (qoBo)+ (qwBw)+ (qg-Rsoqo)Bg
EXERCISE 4
Obtain the definition of coa as used in Perrine’s approach (Hint: Start with 1 STB oil, dissolve
gas in it, and use universal definition of compressibility.)
EXERCISE 4 SOLUTION
Consider an initial volume of 1 STB of oil (at standard conditions with no gas)
Vi = 1 STB
Bring this oil into contact with gas and increase the system pressure by p such that Rso
amount of gas is dissolved in oil. The volume of oil with dissolved gas will be:
Chapter 1 Page 32
Bg
Vf 1 R so c o p
Bo
Note that coDp represents the compression of oil with no gas and
1 B o
co
B o p
Then,
Bg B o
Vf 1 R so co
Bo Bo
Again, using the definition of compressibility:
1 v
co
v p T
obtain a definition for coa
Bg B o
V Vf Vi 1 R so 1
Bo Bo
Bg B o
V R so
Bo Bo
Bg B g
R so
1 Bo Bo
c oa
1 p
B g R so 1 B o
c oa
B o p Bo Bo
In differential form:
B g R so 1 B o
c oa
o p B o
B p
Chapter 1 Page 33
DIMENSIONLESS FORMS OF THE FLOW EQUATIONS
The general form of the radial-cylindrical flow equation, written for a homogeneous and isotropic
reservoir where the well is open to flow over its entire thickness, is:
1 p c t p
r
r r r k t
Expressed in dimensionless terms, this equation becomes:
1 p D p D
r
rD rD r t D
A careful inspection of the equation in its dimensional form indicates that its solution takes the
following form:
p = p (r, t, , , ct, k, h, qsc, rw, re, pi)
Note that the terms h, qsc, rw, re and pi, although they do not appear in the partial differential
equation, show up in the above equation as they are introduced to the solution through boundary
and initial conditions.
In the dimensionless form, the solution to the flow equation involves far fewer parameters:
pD = pD (rD, tD)
Clearly, the dimensionless form of the solution is much more compact and more universal, in
that only two principal independent variables are used to define the variation of the dependent
variable.
DIMENSIONLESS FORM OF BOUNDARY AND INITIAL
CONDITIONS
The outer and inner boundary conditions and the initial conditions of the radial-cylindrical flow
equation are listed here (with their counterparts in dimensionless form) for infinite-acting
reservoirs, finite reservoirs with no-flow boundaries, and finite reservoirs with constant pressure
boundaries.
INFINITE ACTING RESERVOIR
Inner Boundary Condition (constant flow rate specification at the wellbore):
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD 0
rD
Outer Boundary Condition:
p p i as r for t 0 (Dimensional form)
p D 0 as rD for t D 0 (Dimensionless form)
Initial Condition:
Chapter 1 Page 34
p p i at t 0 for rw r (Dimensional form)
p D 0 at t D 0 for 1 rD (Dimensionless form)
FINITE RESERVOIR WITH NO-FLOW BOUNDARY
Inner Boundary Condition (Constant flow rate specification at the well)
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD 0
rD
Outer Boundary Condition:
p
0 for t 0 (Dimensional form)
r r re
p D
0 for t D 0 (Dimensionless form)
rD rD 1
Initial Condition:
p p i at t 0 for rw r re (Dimensional form)
p D 0 at t D 0 for 1 rD reD (Dimensionless form)
FINITE RESERVOIR WITH CONSTANT PRESSURE
OUTER BOUNDARY
Inner Boundary Condition (constant flow rate specification at the well)
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD
rD 0
Outer Boundary Condition:
p p i at r re for t 0 (Dimension al form)
p D 0 at rD reD for t D 0 (Dimension less form)
Chapter 1 Page 35
Initial Condition:
p p i at t 0 for rw r re (Dimensional form)
p D 0 at rD reD for 1 rD reD (Dimensionless form)
Note that the constant flow rate specification at the inner boundary (wellbore) is expressed in
terms of the limit as rD approaches 0. This is purely for simplification purposes and is known as
line source approximation.
ANALYTICAL SOLUTIONS OF THE FLOW EQUATIONS
Infinitely Large Reservoir, Radial Flow, Constant Production Rate (line source solution)
The solution to this problem involves Boltzman transformation, and takes the following form:
1 r2
p D Ei D
2 4t D
Ei (-x) is the exponential integral function, defined as
x
e x
Ei( x ) dx
x
The exponential integral function can be expressed by a series expansion, as shown below:
Ei( x ) Ln(1.781x )
1
n 1
xn
n 1 nn!
From the above series expansion, we see that the Ei (-x) function has a singularity at x=0. The
number of terms required under the summation sign depends on the level of accuracy. However,
for small values of x (usually x < 0.01), we can approximate Ei (-x) as
Ei( x ) Ln(1.781x ) for x 0.01
Note that in our solution, x < 0.01 translates to the condition for 4tD/rD2 > 100, which prompts the
solution
1 4t D2 4t D2
p D Ln 2 for 2
100
2 rD rD
or
1 t D2 t D2
p D Ln 0. 809 for 25
2 rD 2 rD
2
If the argument of the exponential integral function is larger than 5 (x > 5), another accurate
approximation will be
Ei (-x) @ 0 for x >5
Which implies that the solution becomes
rD2
p D 0 for 5
tD
The aforementioned two conditions suggest two simple approximations if the argument of the
exponential integral is less than 0.01 or greater than 5. If x is between 0.01 or 5, one needs to
Chapter 1 Page 36
use the series expansion. Table 1 below—summarized from Nisle (1956)—gives example
values of the exponential integral if 0.01£ x£ 5.
Table 1: Values of the
Exponential Integral, -Ei(-x)
(After Nisle, 1956)
x -Ei(-x)
0.00 +¥
0.02 3.355
0.04 2.681
0.06 2.295
0.08 2.027
0.10 1.823
0.12 1.660
0.14 1.524
0.16 1.409
0.18 1.310
0.20 1.223
0.40 0.702
0.60 0.454
0.80 0.311
1.00 0.219
1.2 0.158
1.4 0.116
1.6 0.0863
1.8 0.0647
2.0 0.0489
3.0 0.0130
4.0 3.78 x 10-3
5.0 1.15 x 10-3
Chapter 1 Page 37
Finite Reservoir, Radial Flow, Constant Production Rate, Closed Boundary (Solution at
the Wellbore)
The solution for this problem is obtained using Laplace transformation. The most common form
of the solution is applied at the wellbore for pseudo steady-state conditions and is given by
2t D 3 2t D
p D well LnreD for 0.25
reD 2 4 reD 2
or
1 4A
p D well 2t DA Ln
2 1.781C A rw2
where CA is the Dietz shape factor—see Table C.1 of Earlougher (1977).
Finite Reservoir, Radial Flow, Constant Production Rate, Constant Pressure Boundary
(Solution at the wellbore)
The following equation represents the pressure drop at the wellbore when steady state
conditions are reached.
(pD)well = Ln(reD) for tD > (reD)2
Figure 1 (after Aziz and Flock, 1963) is a composite chart that describes the dimensionless
pressure drop solutions at the wellbore for the three different reservoirs and boundary conditions
described in this section.
Chapter 1 Page 38
Figure 1
EXERCISE 1
Show that the value of the constant g used in the definition of dimensionless flow rate (qD) is
141.2, as expressed in oilfield units.
(Hint: Start with Darcy Units and implement the necessary conversions for the practical field
units)
EXERCISE 1: SOLUTION
The solution is based on the following observation:
Bq sc M Bq sc M
qD
khp i field khp i Darcy
Then
1 1 STB 5.615 ft
3 3
3 cm D
1 STB
field 1 cp
30.
48
3
D 2 D STB ft 86400 s
1 md1 ft 1 psi
1 md 1 D 1 ft 30.48 cm 1 psi 1 atm
1000 md ft 14.7 psi
Solving for gfield gives
Chapter 1 Page 39
field = 141.2
EXERCISE 2
Explain, within the context of well test analysis, the conditions under which the pressure drop in
a reservoir at a given point and time will be equal to zero.
EXERCISE 2: SOLUTION
The expression that describes the pressure drop in a reservoir in dimensionless terms is given
as
1 rD2
p D Ei
2 4t D
In the above equation, PD will be zero when the exponential integral function is zero. In order for
the exponential integral function be equal to zero, its argument should be larger than 5. In other
words,
r2 r2
Ei D 0 when D 5
4t D 4t D
or
cr 2
5
4 kt
For a given reservoir, f, c, and k are fixed and is a constant. Therefore, the above inequality
will hold when r is large and t is small. This implies that as t gets larger, the value of r needs to
get larger to ensure that the ratio remains greater than 5. Within the context of well test analysis,
this simply states that as time gets larger, pressure transients move further away from the
wellbore, and the areas that have not experienced pressure drop are located further away from
the wellbore.
Chapter 1 Page 40
DIMENSIONLESS GROUPS
Dimensionless groups provide a means of generalizing the mathematical representation of the
reservoir. They simplify the representation by enabling us to express the classical well test
analysis model in terms of two independent variables and one dependent variable. Furthermore,
they allow us to present a universal solution independent of particular reservoir properties, and
independent of a particular set of units. The dimensionless parameters used in the
transformations are based on the assumption that rock and fluid properties are constant.
DIMENSIONLESS PRESSURE
The most common form of dimensionless pressure is represented as a dimensionless pressure
drop group, defined as:
kh
p D (or p D ) p
q sc B
Dp represents the pressure drop at any point in the reservoir, including the wellbore. For
example, at the wellbore Dp is equal to (pi -pwf). in the denominator is a constant and has
different values in Darcy units, practical field units and SI units, as shown in Table 1 below:
Table 1: Numerical Values of Constants and Used in Dimensionless Groups
Darcy Units Oilfield Units SI Units
1 2.637 x 10 -4
2.637 x 10-4
/2 141.2 1.842 x 103
DIMENSIONLESS TIME
There are two different forms of dimensionless time. The first form is more widely used, and is
valid for any geometrical representation and area.
kt
tD
c t rw2
The second form of dimensionless time is based upon reservoir area and is defined as:
kt
t DA
c t A
The relation between tD and tDA is obvious:
rw2
t DA t D
A
The constant l is a unit conversion constant, and its numerical values are given in Table 1
above.
Chapter 1 Page 41
DIMENSIONLESS RADIUS
The dimensionless radius is not dependent on any set of units, and is given by:
r
rD
rw
A dimensionless radius of 100, for example, refers to a position in the solution space that is
100rw away from the wellbore.
OTHER DIMENSIONLESS GROUPS
The dimensionless flow rate group scales the flow rate and is given as:
q B
qD
p i kh
We can use this definition of dimensionless flow rate to define the dimensionless pressure group
in a more compact form such as:
p
p D
piqD
Another common dimensionless group used in linear flow systems gives a definition for a space
variable based on the length of the one wing of the hydraulic fracture:
x
xD
xf
In infinite circular reservoirs, the outer boundary is scaled using the dimensionless group:
re
re D
rw
DIMENSIONLESS FORMS OF THE FLOW EQUATIONS
The general form of the radial-cylindrical flow equation, written for a homogeneous and isotropic
reservoir where the well is open to flow over its entire thickness, is:
1 p c t p
r
r r r k t
Expressed in dimensionless terms, this equation becomes:
1 p D p D
r
rD rD r t D
A careful inspection of the equation in its dimensional form indicates that its solution takes the
following form:
p = p (r, t, , , ct, k, h, qsc, rw, re, pi)
Note that the terms h, qsc, rw, re and pi, although they do not appear in the partial differential
equation, show up in the above equation as they are introduced to the solution through boundary
and initial conditions.
In the dimensionless form, the solution to the flow equation involves far fewer parameters:
pD = pD (rD, tD)
Chapter 1 Page 42
Clearly, the dimensionless form of the solution is much more compact and more universal, in
that only two principal independent variables are used to define the variation of the dependent
variable.
DIMENSIONLESS FORM OF BOUNDARY AND INITIAL
CONDITIONS
The outer and inner boundary conditions and the initial conditions of the radial-cylindrical flow
equation are listed here (with their counterparts in dimensionless form) for infinite-acting
reservoirs, finite reservoirs with no-flow boundaries, and finite reservoirs with constant pressure
boundaries.
INFINITE ACTING RESERVOIR
Inner Boundary Condition (constant flow rate specification at the wellbore):
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD 0
rD
Outer Boundary Condition:
p p i as r for t 0 (Dimensional form)
p D 0 as rD for t D 0 (Dimensionless form)
Initial Condition:
p p i at t 0 for rw r (Dimensional form)
p D 0 at t D 0 for 1 rD (Dimensionless form)
FINITE RESERVOIR WITH NO-FLOW BOUNDARY
Inner Boundary Condition (Constant flow rate specification at the well)
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD 0
rD
Outer Boundary Condition:
p
0 for t 0 (Dimensional form)
r r re
Chapter 1 Page 43
p D
0 for t D 0 (Dimensionless form)
rD rD 1
Initial Condition:
p p i at t 0 for rw r re (Dimensional form)
p D 0 at t D 0 for 1 rD reD (Dimensionless form)
FINITE RESERVOIR WITH CONSTANT PRESSURE
OUTER BOUNDARY
Inner Boundary Condition (constant flow rate specification at the well)
p q B
r sc for t 0 (Dimensional form)
r r rw kh
p D
rD 1 for t D 0 (Dimensionless form)
rD rD 1
or
p D
LimrD 1
rD
rD 0
Outer Boundary Condition:
p p i at r re for t 0 (Dimension al form)
p D 0 at rD reD for t D 0 (Dimension less form)
Initial Condition:
p p i at t 0 for rw r re (Dimensional form)
p D 0 at rD reD for 1 rD reD (Dimensionless form)
Note that the constant flow rate specification at the inner boundary (wellbore) is expressed in
terms of the limit as rD approaches 0. This is purely for simplification purposes and is known as
line source approximation.
ANALYTICAL SOLUTIONS OF THE FLOW EQUATIONS
Infinitely Large Reservoir, Radial Flow, Constant Production Rate (line source solution)
The solution to this problem involves Boltzman transformation, and takes the following form:
1 rD2
p D Ei
2 4t D
Ei (-x) is the exponential integral function, defined as
x
e x
Ei( x ) x dx
The exponential integral function can be expressed by a series expansion, as shown below:
Chapter 1 Page 44
Ei( x ) Ln(1.781x )
1
n 1
xn
n 1 nn!
From the above series expansion, we see that the Ei (-x) function has a singularity at x=0. The
number of terms required under the summation sign depends on the level of accuracy. However,
for small values of x (usually x < 0.01), we can approximate Ei (-x) as
Ei( x ) Ln(1.781x ) for x 0.01
Note that in our solution, x < 0.01 translates to the condition for 4tD/rD2 > 100, which prompts the
solution
1 4t D2 4t D2
p D Ln 2 for 2
100
2 rD rD
or
1 t D2 t D2
p D Ln 0. 809 for 25
2 rD 2 rD
2
If the argument of the exponential integral function is larger than 5 (x > 5), another accurate
approximation will be
Ei (-x) @ 0 for x >5
Which implies that the solution becomes
rD2
p D 0 for 5
tD
The aforementioned two conditions suggest two simple approximations if the argument of the
exponential integral is less than 0.01 or greater than 5. If x is between 0.01 or 5, one needs to
use the series expansion. Table 1 below—summarized from Nisle (1956)—gives example
values of the exponential integral if 0.01£ x£ 5.
Table 1: Values of the
Exponential Integral, -Ei(-x)
(After Nisle, 1956)
x -Ei(-x)
0.00 +¥
0.02 3.355
0.04 2.681
0.06 2.295
0.08 2.027
0.10 1.823
0.12 1.660
Chapter 1 Page 45
0.14 1.524
0.16 1.409
0.18 1.310
0.20 1.223
0.40 0.702
0.60 0.454
0.80 0.311
1.00 0.219
1.2 0.158
1.4 0.116
1.6 0.0863
1.8 0.0647
2.0 0.0489
3.0 0.0130
4.0 3.78 x 10-3
5.0 1.15 x 10-3
Finite Reservoir, Radial Flow, Constant Production Rate, Closed Boundary (Solution at
the Wellbore)
The solution for this problem is obtained using Laplace transformation. The most common form
of the solution is applied at the wellbore for pseudo steady-state conditions and is given by
2t D 3 2t D
p D well LnreD for 0.25
reD 2
4 reD 2
or
1 4A
p D well 2t DA Ln
2 1.781C A rw2
where CA is the Dietz shape factor—see Table C.1 of Earlougher (1977).
Chapter 1 Page 46
Finite Reservoir, Radial Flow, Constant Production Rate, Constant Pressure Boundary
(Solution at the wellbore)
The following equation represents the pressure drop at the wellbore when steady state
conditions are reached.
(pD)well = Ln(reD) for tD > (reD)2
Figure 1 (after Aziz and Flock, 1963) is a composite chart that describes the dimensionless
pressure drop solutions at the wellbore for the three different reservoirs and boundary conditions
described in this section.
Figure 1
EXERCISE 1
Show that the value of the constant g used in the definition of dimensionless flow rate (qD) is
141.2, as expressed in oilfield units.
(Hint: Start with Darcy Units and implement the necessary conversions for the practical field
units)
EXERCISE 1: SOLUTION
The solution is based on the following observation:
Chapter 1 Page 47
Bq sc M Bq sc M
qD
khp i field khp i Darcy
Then
1 1 STB 5.615 ft
3 3
3 cm D
1 STB
field 1 cp
30.48 3
D 2 D STB ft 86400 s
1 md1 ft 1 psi
1 md 1 D 1 ft 30.48 cm 1 psi 1 atm
1000 md ft 14.7 psi
Solving for gfield gives
field = 141.2
EXERCISE 2
Explain, within the context of well test analysis, the conditions under which the pressure drop in
a reservoir at a given point and time will be equal to zero.
Chapter 1 Page 48
TYPE CURVES
TYPE CURVES AS TOOLS IN WELL TEST ANALYSIS
Type curves display in graphical form the analytical or numerical solutions to flow equations for
specified cases of initial conditions, boundary conditions and geometry. Most of the time, these
graphical representations are displayed in terms of dimensionless variables. A family of curves
can be displayed in one type curve, allowing variations in skin factor, wellbore storage, hydraulic
fracture length and other key parameters
Type curves are generated using a forward solution approach. To ensure a complete description
of the problem, we incorporate all of the significant deviations from the simplified problem
description and then obtain the solution, usually by numerical means. Since the problem
description is more complete, and devoid of some curtailing assumptions, the solution will also
contain more complete information on the system properties that are readily available for
extraction. Type curves also provide a continuous solution that incorporates the effects of all of
the different flow regimes that control the entire pressure behavior.
RELATIONSHIP BETWEEN THE DATA PLOT AND THE
TYPE CURVE
Type curves are instrumental in achieving sound pattern recognition, or model matching. They
are usually presented in log-log coordinates. The matching process between the data plot and
the type curve is simply based on logarithmic relationships.
For example, consider a type curve that displays a log-log plot of dimensionless pressure drop
versus dimensionless time. If we plot pressure drop versus time data on a log-log graph having
the same scale as that of the type curve, we can observe the following correspondences
between the two plots.
Since (using oilfield units),
pkh
p D
141.2Bq sc
By taking the logarithm of both sides, we obtain
kh
Logp D Logp Log (A)
141. 2Bq
sc
Similarly, since
2.637 10 4 kt
tD
c rw2 ,
again taking the logarithm of both sides, this time we end up with
2.637 10 4 k
Logt D Log Log(t )
crw
2 (B)
Equations (A) and (B) simply indicate that both the type curve and the data plots are similar, with
the exception being that along the DPD axis, they are shifted from each other by a constant
kh
141.2Bq sc
Similarly, along the tD axis, we see a shift with a magnitude of
Chapter 1 Page 49
2.637 10 4 k
c rw2
Therefore, by sliding the data plot over the type curve (magnitude of shifts as defined above), we
can find a match between the data plot and the type curve.
DATA PLOT PREPARATION
Data plots for type curve matching are plotted using the same scale as that used for the type
curves. In preparing a plot, we need to consider the number and frequency of data points,
because state-of-the-art technology allows high-precision measurements at very high frequency.
Therefore, it is necessary to reduce the number of data points to a manageable size. In reducing
the data, we need to ensure that we do not miss important events. We should avoid sampling
the data at fixed intervals, because the most significant and rapid changes take place during the
early time. Also, in handling the early time data, we may have to shift the time line forward or
backward to synchronize the recorded pressure responses with the recorded time.
LOG-LOG TYPE CURVES
Log-log type curves display the dimensionless pressure function versus the dimensionless time
function, The most commonly used log-log type curves are described below:
Agarwal, Al-Hussainy, and Ramey (1970): This is the most widely used type of log-log curve.
Its general appearance is represented in Figure 1 . The main features of this type curve are as
follows:
Figure 1
A single well in an infinitely large homogeneous reservoir is considered
The well is produced at a constant production rate
Wellbore storage effects are considered
Curves are also characterized by different skin factors
Chapter 1 Page 50
Because different type curves may match the same data set (non-uniqueness of the
match), a quantitative analysis may not be very accurate
Type curves with different wellbore storage constants all converge to the same solution,
signifying the end of the early time region. Therefore, these type curves are very helpful in
identifying the start of the semi-long straight line (beginning of the middle-time region).
Earlougher and Kersch (1974): These type curves represent a different relation between
pressure and time. The reservoir is infinitely large, and there is a single well producing at a
constant rate. Wellbore storage and skin effects are included. This type curve is especially
useful when the semi-log straight line does not develop on the Horner plot. The use of this type
curve does not require a prior knowledge of the wellbore storage coefficient.
Gringarten, Ramey and Rhagavan (1974): Gringarten et al. generated type curves for
hydraulically fractured wells. Some of these type curves are developed for a uniform flux fracture
(fracture loading per unit length of the fracture is constant), and the rest are developed for an
infinite conductivity fracture (no pressure drop in the wellbore is considered). All of these type
curves are generated for reservoirs with a single-well producing at a constant rate; some of them
apply to infinitely large reservoirs, while others are for finite, closed reservoirs.
SEMILOG TYPE CURVES
Among semi-log type curves, the two plots that are used most widely are the McKinley type
curves and the Gringarten, Ramey and Raghavan type curves.
McKinley (1969): McKinley type curves are principally used not for calculating pressure
response, but rather for type curve matching of test data. With McKinley type curves, it is
possible to characterize damage or stimulation around a well with significant wellbore storage
effects during the well test. Analysis of the early time data with McKinley type curves gives the
permeability of the formation around the wellbore, and late time data analysis gives the true
formation permeability. Obviously, comparison of near-wellbore permeability and true formation
permeability values as extracted from McKinley’s type curves provides a direct measure about
the extent of the damage or stimulation.
Gringarten, Ramey and Raghavan (1974): These semilog type curves give the dimensionless
pressure data for a fractured well located in an infinite acting reservoir. The type curve presents
solutions for uniform flux and infinite conductivity conditions.
PRESSURE GROUP TYPE CURVES
In pressure group type curves, when the logarithm of pressure difference is plotted versus the
logarithm of time, it looks exactly like the logarithm of dimensionless pressure drop versus the
logarithm of a dimensionless time function. The difference between the two plots is a simple
linear shift along the y and x-axes of the type curve. Once we achieve a proper match of the two
plots, we can calculate the flow capacity (kh) from the vertical displacement, and the storativity
( ct) from the horizontal displacement.
Using a Log(PD) versus Log (tD) type curve and a Log(P) versus Log t data plot, we can use
the following equations at the match point to calculate the formation capacity and storativity:
p M kh p D
p D M kh q sc B
q scB p M
t M k k t
t D M c t
c t rw rw
2 2
t D M
The subscript M in the above equations refers to the match point.
Chapter 1 Page 51
PRESSURE DERIVATIVE TYPE CURVES
The pressure derivative type curves interpretation method is based on analyzing the derivative
of the pressure function with respect to the time function. Pressure derivative type curves can be
helpful in finding a unique solution, and can provide a better basis for selecting the semi-log
straight line and identifying certain patterns related to reservoir heterogeneity. Figure 1 (after
Bourdarot, 1996—courtesy Editions Technip) shows some of the important events, as
highlighted by the derivative of pressure versus a time function on a log-log plot.
Figure 1
In a pressure derivative plot, the time period during which wellbore storage effects are dominant
will still be seen as a straight line of slope equal to unity, which is followed with a hump. When
the early time region is over and a radial flow regime develops, a flat line with an ordinate of 0.5
develops.
Figure 2 (after Bourdarot, 1996—courtesy Editions Technip) gives the Bourdet derivative type
curves for a well located in an infinite-acting homogeneous reservoir and experiencing skin
effects and wellbore storage.
Chapter 1 Page 52
Figure 2
In preparing pressure derivative data plots, we may encounter a significant amount of data
scattering. The oscillations observed for the pressure function, which are the result of noise in
the raw data, should be smoothed prior to taking the derivative. The oscillations become more
pronounced during the late times when the rate of pressure change is minimal. Therefore, in
smoothing the pressure data, we must be particularly careful not to remove the signals that
represent the pressure response.
TYPE CURVE MATCHING
We can summarize the mechanics of type curve matching as follows:
Based on existing knowledge and your engineering judgment, choose a type curve
that you believe closely represents the test data (e.g., type curves for fractured
wells, double porosity reservoirs, etc.)
Prepare the data plot to display the actual pressure change versus time. Make sure
that the data plot has the same scale as the type curve (you can do this manually,
using tracing paper, or you can use computer software designed for this purpose).
Superimpose the data plot over the type curve. Keeping the two coordinate axes
parallel, slide the data plot to find a type curve that matches the data.
Once a satisfactory match is obtained, pick a match point anywhere on the data plot
and record the coordinates of the match point on both plots. To make it easier to
record the coordinates from the data plot, you should select the intersection of the
major gridlines.
Using the recorded match point coordinates, compute the formation flow capacity
(kh) and storativity constant (fct).
Chapter 1 Page 53
COMMONLY USED TYPE CURVES
There are literally hundreds of type curves available for interpreting well test data under different
reservoir and wellbore conditions. Therefore, a first step in the analysis should include a review
of the technical literature to find the type curve that is most applicable to a particular situation.
Table 1 below lists the type curves that are used most often for general well test analysis. These
type curves are generalized for vertical test wells located in infinitely large homogeneous
reservoirs. For detailed information, refer to the sources cited in the References section.
Table 1: Most Widely Used Type Curves
Type Curve Extractable Information Remarks
Agarwal et al. (1970) Wellbore storage constant; Can also be used as a
Skin factor; Start of radial diagnostic plot.
flow regime; Permeability;
Porosity; Porosity-
compressibility product
McKinley (1969) Wellbore storage constant; Good for short-term tests.
Near-wellbore permeability;
Formation permeability;
Flow efficiency
Earlougher and Kersh Wellbore storage constant; Good for curve-matching of
(1974) Skin factor; Permeability short-term transient data
Gringarten, Ramey Hydraulic fracture length; Assumes square drainage
and Rhaghavan Permeability; Distance to area, uniform flux and
(1974) the side of the drainage infinite conductivity vertical
area fractures
Bourdet et al (1983). Wellbore storage constant; Can also be used as a
Permeability; Skin factor; diagnostic plot.
Beginning of semilog
straight line
EXERCISE 1
Is type curve matching more accurate than traditional well test analysis procedures? Explain.
EXERCISE 1 SOLUTION
Type curve matching is especially useful when the test is too short for the semi-log straight line
to develop. Although type curves present the pressure profile as a function of time in a much
more complete form for the quantitative analysis (because skin factor, wellbore storage, linear
flow are all built into the solution), they should be used only when the classical approach is not
applicable. However, type curves as diagnostic tools provide the most practical approach for
qualitative analysis. The type curve matching technique should be coupled with the classical well
test analysis, as it provides important feedback such as start of the semi-log straight line.
Chapter 1 Page 54
EXERCISE 2
What are the two main drawbacks of type curve matching?
What is the main difficulty with pressure derivative type curve analysis?
EXERCISE 2 SOLUTION
The two main drawbacks of type curve matching technique are:
Non-uniqueness of the match. This becomes problematical especially when the
wellbore storage effects disappear in a short period of time as is the case of high-
permeability reservoirs.
Middle-time and late-time data are compressed in the logarithmic scale. This makes
the recognition and evaluation of certain events more difficult.
The principal difficulty with pressure derivative analysis is the preparation of the data plot. The
presence of noise in the data especially makes late-time data more difficult to analyze. In
preparing the derivative plot, it will be necessary to implement some degree of smoothing.
Chapter 1 Page 55
Pressure Transient Analysis in Anisotropic Reservoirs
INTRODUCTION AND OVERVIEW
Classical well test interpretation assumes that a reservoir is isotropic with regard to its physical
properties—that their magnitudes at a given location will be the same regardless of the direction
in which they are measured. In reality, most reservoirs exhibit varying degrees of anisotropy; at
least one property’s measured value will vary depending on the direction of measurement.
The terms isotropy and anisotropy relate to the individual properties of a porous medium. A
medium can be isotropic with respect to one of its properties, and anisotropic with respect to
another. It is also important to note that these terms apply only to properties that can be
expressed in terms of magnitude and direction (i.e., as vectors); volume-based properties do not
meet this criterion. Porosity, for example, is a volume-based property by definition; it utilizes all
three dimensions, and therefore has zero degrees of freedom in terms of directional variation.
Permeability, by contrast, has the dimensions of area, leaving one direction in which it can vary.
From the standpoint analyzing pressure transient data, we are particularly interested in
anisotropy in the context of directional permeability and/or elliptical flow.
DIRECTIONAL PERMEABILITY
Absolute permeability is the most basic transport parameter of a rock. It can exhibit not only
anisotropy (i.e., varying according to the direction of measurement), but also heterogeneity (i.e.,
varying from one location to another).
PERMEABILITY AS A TENSOR
In a porous material, it is possible to find a direction x of maximum permeability. In a plane
perpendicular to x, we will find another direction y in which permeability is maximum on that
plane, and another direction z that is perpendicular to y, in which permeability has a minimum
value (kx ³ ky ³ kz). In every other direction between x, y and z, the permeability has some
intermediate value. Note that in this discussion, the x, y and z directions are considered as the
potential gradient and flow path coincide. If the flow path and the potential gradient did not
coincide, then the intermediate permeability values between x, y and z would be non-zero, and
the full sensorial representation of the permeability would be as follows:
k xx k xy k xz
k k yx k yy k yz
k zx k yx k zz
The symmetric aspect of the permeability tensor stems from the mass conservation principle if
we consider that the permeability value that provides the transport of mass from x to y must be
the same as that which provides the transport of mass from y to x. This implies that
kxy = kyx ; kxz = kzx ; kyz = kzy
For a two-dimensional flow problem (flow in the x-y plane), these xyz representations collapse
to the following forms:
k xx k xy
k
k yx k yy
and
kxy = kyx
In the above representations, the first subscript represents the direction of flow and the second
subscript refers to the gradient that produces the flow. Therefore, if the flow direction and the
gradient directions coincide, then the cross terms of the permeability become zero and we end
up with the following permeability tensors for 3D and 2D flow, respectively
Chapter 1 Page 56
k xx 0 0
k 0 k yy 0
0 0 k zz
and
k xx 0
k
0 k yy
LATERAL PERMEABILITY
Lateral permeability anisotropy is related to the sedimentation process and, in naturally fractured
reservoirs, to the fracture orientation.
Sedimentation and overburden pressure cause particles to be oriented with their
longest dimensions parallel to each other. This type of grain orientation results in
high permeability flow channels.
In many reservoirs and aquifers, fractures produce a pronounced permeability
contrast, showing very high rock permeability parallel to the fractures, and much
smaller permeability perpendicular to the fractures.
Reservoir rock permeability depends strongly on the pore system geometry. In particular, lateral
permeability contrasts are controlled by the size, shape and interconnection of pore spaces in
different directions, along with the tortuosity of the flow path.
AVERAGE PERMEABILITY IN LATERALLY ANISOTROPIC
RESERVOIRS
The partial differential equation that describes flow in a 2D anistropic porous medium is
2p 2p p
kx 2
k y 2
c
x y t
We can reduce the above equation to an isotropic form by performing a coordinate
transformation
x y
x and y
kx ky
which results in
2p 2p c p
x 2 y 2 k t
where
k k xk y
We can estimate the geometric mean permeability k using conventional drawdown, buildup,
multi-rate and other testing procedures. However, it is not possible to find the components of k
from a single well test. To find the directional properties, we must conduct interference tests
involving at least one producer and three observation wells around the producer.
VERTICAL PERMEABILITY
Chapter 1 Page 57
Because of the way that sediments are deposited, a formation’s permeability is generally greater
along the plane of deposition than it is in the direction normal to the plane of deposition. This
implies that horizontal permeability will be greater than vertical permeability, assuming no post-
depositional tilting or folding of the formation has taken place. Vertical permeability is not a
matter of primary concern in reservoirs that are produced by vertical wells. In highly deviated
and horizontal wells, on the other hand, vertical permeability may become the main control of
productivity.
Vertical permeability complicates the pressure transient analysis of layered reservoirs. In a
layered system with vertical permeability, it is necessary to consider both the lateral fluid flow
within each layer and the vertical flow between the layers. To account for crossflow between
layers, we usually assume an instantaneous pseudosteady-state transfer mechanism, which
simply uses pressure values across each layer and assumes linear flow between the layers. In
some cases—for example, where an impermeable shale layer hydraulically separates two sand
layers—there is no pressure communication between layers, and the sand layers are connected
only in the wellbore. This results in commingled layer production. Figure 1 and Figure 2
Figure 2
illlustrate commingled layers and layers in crossflow.
Figure 1
Chapter 1 Page 58
ELLIPTICAL FLOW
Elliptical flow patterns most commonly develop in
Hydraulically fractured wells
Horizontal wells
Laterally anisotropic reservoirs
HYDRAULICALLY FRACTURED WELLS
Hydraulic fracturing can dramatically increase productivity, and can greatly influence a system’s
pressure transient response. Well test analysts should therefore be aware of fracture-caused
deviations from the classical pressure transient model.
Although hydraulic fractures can be either horizontal or vertical, this discussion focuses on
vertical hydraulic fractures, which predominate in formations deeper than about 2000 ft [610 m].
A hydraulically induced vertical fracture tends to propagate in a direction perpendicular to that of
the minimum principal horizontal stress. Once we have created a fracture and released the
fracture pressure, the inside stresses work to heal or close the fracture. To prevent the fracture
from closing, we pack the fracture space with a relatively coarse proppant. This leaves us with a
high permeability conduit from the formation to the wellbore. In well test analysis models, we
assume that the height of a hydraulic fracture is equal to the formation thickness (i.e., the
fracture is 100% contained within the formation of interest). We also assume that the
propagation of a hydraulic fracture from the well in opposite directions is symmetrical, so that
length of each wing of the fracture is the same. (Figure 1 ).
Figure 1
The mathematical models that represent the pressure transient behavior of hydraulically
fractured wells consider three types of fractures:
Finite conductivity
Infinite conductivity
Chapter 1 Page 59
Uniform flux
In preparation for discussing each of these fracture types, we introduce the following
dimensionless groups:
Dimensionless time based on fracture length:
2.637 10 4 kt
t Dxf
c t x 2f
Dimensionless fracture conductivity:
kfwf
cD
kmx f
FINITE CONDUCTIVITY FRACTURES
There are five different flow regimes observed at different times. Figure 2 and Figure 3
Figure 3
provide a schematic representation of these flow regimes.
Chapter 1 Page 60
Figure 2
1. Fracture Linear Flow: Linear flow in the fracture takes place during very early times,
and can be of extremely short duration. During this period, the fracture is emptied of
fluid.
2. Bilinear Flow: Due to the linear flow in the fracture, the pressure within the fracture is
much lower than at the fracture surface. This generates a linear flow from the formation
into the fracture.
3. Formation Linear Flow: This flow regime is observed when fracture conductivity is so
high such that any fluid entering the fracture is immediately produced at the wellbore
(infinite conductivity fracture).
4. Elliptical Flow in Formation: This flow regime represents a transitional flow regime
between linear and radial flow. Equal-pressure contours are confocal ellipses and
streamlines are confocal hyperbolas. Note that both ellipses and hyperbolas share the
same set of foci, located at the tips of the fracture.
5. Radial Flow in Formation: At late times, pressure around the fracture is lowered over
a relatively large area. This reduced pressure zone serves as a sink, and flow lines
orient themselves towards this circular sink such that full radial flow develops.
INFINITE CONDUCTIVITY FRACTURES
In an infinite conductivity fracture, no pressure drop occurs within the fracture itself. In other
words, the fracture represents the innermost equal-pressure contour. Any fluid entering the
fracture is assumed to be produced instantaneously at the wellbore. For practical purposes, a
fracture is generally assumed to be of infinite conductivity when its dimensionless fracture
conductivity is larger than 300.
UNIFORM FLUX FRACTURES
Chapter 1 Page 61
In this conceptual treatment, we assume that fracture loading from the formation per unit length
of fracture is uniform along the fracture length. Again, flow from the formation into the fracture is
linear.
CHARACTERISTICS PRESSURE TRANSIENT RESPONSES OF
HYDRAULICALLY FRACTURED WELLS:
During the linear flow period, the pressure drop is given by the true linear flow response in the
presence of an infinite conductivity fracture):
p D t dxf
1
2 (A)
The above equation indicates that a ½ slope straight line on a log-log plot will be formed during
the linear flow period, since:
1 1
Logp D Logt dxf Log
2 2
Therefore, on the diagnostic plot the data points that fall on a ½ slope straight line should not be
considered in defining the semilog straight line of the Horner plot of the same data set.
The relationship in Equation (A) above also can be used to calculate the length of the fracture.
Returning to the dimensional form of this equation, we obtain:
qB
p 4.064 t
hx f kc t
Therefore, a plot of Dp versus t1/2 on Cartesian coordinates will yield a straight line with a slope
mf equal to
qB
m f 4.064
hx f kc t
From the slope of Cartesian coordinate plot fracture, length xf, can be estimated. This
procedure, however, requires that the permeability of the formation be available, either from the
Horner plot analysis of the infinite acting/radial flow portion of the test, or from a pre-fracture test
on the same well.
When we analyze the pressure data from the radial flow portion of the test (buildup) using the
Horner plot, we see that the curve with has a continuously increasing slope. If the increasing
slope reaches a maximum value and stays on that value for the remainder of the test, then a
true radial flow regime is developed. The slope of the straight line can be used in calculating the
formation permeability.
Figure 4 (Horne,
Chapter 1 Page 62
Figure 4
1995—courtesy Petroway, Inc.), and Figure 5 and Figure 6
Figure 6
(Daltaban and Wall, 1998—courtesy Imperial College Press), respectively, show examples of a
diagnostic plot, a Horner plot and a t1/2 plot generated from the pressure transient data obtained
in a hydraulically fractured well.
Chapter 1 Page 63
Figure 5
As a further reference, Appendix C of Earlougher (1977) includes a dimensionless pressure plot
for a single, vertically fractured well in an infinite system with no wellbore storage, and type
curves generated for vertically fractured wells located in the center of a closed square, based on
the data of Gringarten, Ramey and Rhaghavan (1974).
HORIZONTAL WELLS
Starting in the 1980s, advances in drilling technology began to bring horizontal well applications
within the group of mature petroleum engineering technologies. This introduction of horizontal
wells into oil field production has made it necessary to develop well test interpretation
methodologies for horizontal wells.
It is generally more challenging to design and interpret tests for horizontal wells than it is to do
so for vertical wells, for several reasons:
Because horizontal wells exhibit three-dimensional flow, the role of vertical
permeability is important.
The absence of radial symmetry gives rise to various flow regimes during the test
period. In formulating a representative well-reservoir interaction, all of these flow
regimes need to be considered.
Since horizontal wells are much longer than vertical wells, zonal variations along
the length of the wellbore make it more difficult to obtain a meaningful interpretation
of the pressure transient data collected.
FLOW REGIMES IN HORIZONTAL WELLS
A number of flow regimes can theoretically be encountered during a drawdown or buildup test on
a horizontal well. Depending on the existing heterogeneities, wellbore storage effects, and/or
boundary effects, some of these flow regimes may be missing. The three important
dimensionless parameters that control the pressure transient response and flow regimes around
a horizontal wellbore are:
Ratio of the vertical to the lateral permeability (kv/kh),
Position of the well with respect to the thickness of the formation (zL/h), and
Ratio of the length of the well to the formation thickness (L/h).
Figure 7 illustrates the horizontal well configuration and its key dimensions.
Chapter 1 Page 64
Figure 7
1. Early time radial flow (Figure 8 ): This flow regime is established as the
horizontal well is first produced, and may be masked by wellbore storage
effects.
Figure 8
It continues until the upper and lower boundaries are felt at the wellbore.
Therefore, the duration of the early time radial flow period is controlled by the
formation thickness and the vertical permeability.
Chapter 1 Page 65
2. Intermediate time linear flow (Figure 9 ): This flow regime develops if the
formation thickness is not large, such that the top and bottom boundaries make
their presence felt in a more pronounced manner.
Figure 9
This intermediate time linear flow does not take place if the vertical permeability
is much smaller than the lateral permeability.
3. Hemi-radial (hemicylindrical) flow (Figure 10 ): This flow regime develops if
the horizontal well is not located at a centered position between the top and
bottom boundaries.
Chapter 1 Page 66
Figure 10
4. Late time radial flow (Figure 11 ): In this flow regime, flow is radial in the
horizontal plane. Late time radial flow develops when there are no nearby
boundaries in the horizontal plane.
Figure 11
Chapter 1 Page 67
During the late time radial flow regime, the well acts as if it is a point source.
For this to happen, the length of the horizontal well must be much smaller than
the lateral dimensions of the formation and well must not be located close to a
lateral boundary.
5. Late time linear flow (Figure 12 ): This flow regime will develop when the
parallel field boundaries are felt.
Figure 12
Identifying these flow regimes is extremely important. The following empirical equations are
recommended for estimating the end of the early time radial flow, intermediate time linear flow,
and late time radial flow.
Early time radial flow:
t erf
190.0 h 2.095 rw0.095 c t
kv
Intermediate time linear flow:
20.8c t L2
t ilf
kh
Late time radial flow:
Begins at:
1230.0c t L2
t blrf
kh
Ends at:
Chapter 1 Page 68
297.0x 1 x 2
2.095
c t L0.095
t elrf
kh
CHARACTERISTICS OF THE HORIZONTAL WELL PRESSURE
TRANSIENTS:
Figure 13 (after Daltaban and Wall, 1998—courtesy Imperial College Press) qualitatively shows
the characteristics of the horizontal well pressure transients when all of the assumptions of an
ideal horizontal well are fulfilled.
Figure 13
These assumptions include conditions such as the duration of wellbore storage effects, and a
well that is perfectly horizontal and parallel to the top and bottom boundaries.
LATERALLY ANISOTROPIC RESERVOIRS
In a laterally anisotropic reservoir, permeability is preferentially higher in one direction. Braided
stream deposits, for example, show greater permeability along the braids, while in faulted
reservoirs, the direction of highest permeability is usually parallel to the fault alignment. The
geometric mean of the measured permeabilities along the major and minor anisotropy axes is
equivalent to the test-derived permeability for a laterally anisotropic reservoir.
The direction of highest lateral permeability is an important consideration in orienting a horizontal
well. The axis of the horizontal section should be perpendicular to the direction of highest
permeability to ensure that the maximum flow rate is directed into (rather than parallel to) the
wellbore.
Chapter 1 Page 69
In laterally anisotropic reservoirs, the movement of the pressure transients is much more rapid in
the direction of highest permeability. The resulting equipotential lines are therefore confocal
ellipses rather than concentric circles.
EQUIPRESSURE CONTOURS AND STREAMLINES
The shapes of the inner boundary (the wellbore) and the outer boundary (the physical limits of
the reservoir) determine the shape of the equipressure contours, provided that both the inner
and outer boundaries are each at a uniform pressure. Therefore, the shape of the inner
boundary represents the innermost equipressure contour and similarly, the shape of the outer
boundary represents the outermost equipressure contour.
From this observation, we may conclude that a circular wellbore is characterized by circular
equipressure contours. By way of contrast, a hydraulically fractured well connected to a high
conductivity fracture has as its innermost equipressure contour the fracture axis.
Streamlines refer to the paths that are taken by the fluid particles. Since streamlines are
orthogonal to the equipotential contours, equipotential contours and streamlines are conjugate
function of each other. In radial flow, streamlines are the rays that emanate from the well or are
oriented towards the well. In a hydraulically fractured flow domain, on the other hand,
streamlines become confocal hyperbolas that share the same foci with the elliptic equipressure
contours.
WELL TEST INTERPRETATION IN ANISOTROPIC
RESERVOIRS
SINGLE WELL TESTING
To express an anisotropic formation in the form of an equivalent isotropic formation, we must
implement a coordinate transformation. In rectangular coordinates, this transformation is
achieved by
x y
X and Y
kx ky
The resulting isotropic permeability in the lateral plane for the new domain is
k h k xk y
Note that using the coordinate transformation changes the shape of the well. The circular
wellbore in the original x-y plane becomes an elliptical wellbore in the new plane. We can
estimate the isotropic permeability following the routine analysis procedures of drawdown, build-
up, etc. But once we calculate the equivalent isotropic k h value, it is not possible to resolve it into
its components using a single well test. Also, note that the skin factor calculated from the
transformed system is not equivalent to the actual skin that exists around the wellbore because
of the change in the geometrical shape of the wellbore. The presence of anisotropic permeability
shows as an additional skin.
INTERFERENCE TESTING
It is not possible to find the components of the permeability tensor from a single-well test, even if
the principal flow directions and the coordinate axes coincide. Interference testing, however, can
be used to estimate the anisotropic permeability. This approach utilizes a producer (or an
injector) and two observation wells. The producer defines the origin of the coordinate system. In
the analysis, the exponential integral type curve (PD versus tD/(rD)2 ) is used with data originating
from each of the observation wells.
A field example of the interference testing for analysis of anisotropic formations can be found in
Ramey (October 1975).
VERTICAL PERMEABILITY DETERMINATION
Chapter 1 Page 70
The vertical permeability of a formation is usually significantly different from horizontal
permeability. This is a natural result of the depositional sequence. Pressure and intensity of
streaks of shale will control the degree of anisotropy between the horizontal and vertical
permeabilities.
Vertical Interference Testing
The principal idea behind vertical interference testing is similar to conventional interference
testing. The most significant difference is in that only one well is used in a vertical interference
test. This necessitates using a packer to separate the formation to be tested into two sections
within the wellbore. In a cased hole, there have to be perforations in the well section both above
and below the packer. These perforations need to be separated from each other to ensure a
good vertical flow domain. Fluid is injected from the upper section of the well and produced
through the perforations in the lower section of the well. Figure 1 a and
href="javascript:figurewin2('../../asp/graphic.asp?code=29876&order=1','2')">Figure 2 show the
schematics of the well configuration and the terminology used in vertical interference and vertical
pulse testing.
Figure 2
In most formations, kh > kv, resulting in an ellipsoidal flow field with its minor axis in the vertical
direction. In analyzing the generated data, we use the analysis proposed by Prats (1970), in
which he suggests a plot of pws (pressure measured at lower potion of the well) versus logarithm
of time, which will be a straight line. Using the slope of the semilog straight line, we can calculate
horizontal permeability from
162.6Bq sc
kh
mh
Note that in the above equation, qsc is negative and m is positive. It will also be necessary to
read the intercept from the semilog straight line (pws at t= 1 hour) as p1hr.
Prats then suggests the following equation for calculating the vertical permeability kv:
Chapter 1 Page 71
h
G*
c t h 2
1 p 1hr p i z wf z ws
kv Log
2.637 10 4 m 2.3025
In the above equation, G* is a geometric factor and can be extracted for a given well
configuration—refer to Prats (1970) or Earlougher (1977). Note that parallel to any other well test
analysis procedures, p1hr must be read off the semilog straight line, if necessary from the
extrapolated section.
Vertical Pulse Testing
One of the disadvantages of vertical interference testing is that it requires relatively extended
testing to form a semilog straight line. In the meantime, boundary effects may enter into the
collected data, making the analysis of the test even more cumbersome.
The well configuration used in vertical pulse testing is similar to that used in vertical interference
testing (see Figure 1 and Figure 2 ). The test involves introducing alternating pulses and shut-in
periods of equal duration, and recording the time lag between the arrival of the pulses and the
amplitude of the pressure response (Figure 3 , after Daltaban and Wall, 1998—courtesy Imperial
College Press).
Figure 3
The amplitude of the pressure response is measured between the two parallel tangent lines that
are drawn through the peaks and valleys of the pressure response. The dimensionless pulse
time and dimensionless pulse response amplitude are defined as:
2.637 10 4 k v t p
t PDV
c t Z R2
Chapter 1 Page 72
2k h Z R p
p PDV
141.2qB
To solve for kv and kh, it is necessary to use the type curves of tPDV versus pPDV generated by
Falade and Brigham (1974), and reproduced as Figures 10.28 and 10.29 in Earlougher (1977).
EXERCISE 1
Estimate the value of 1 Darcy in cm2.
EXERCISE 1 SOLUTION
Using Darcy’s law in Darcy units, we have
kA p
q
L
cm 3 1 g
q Lcm
s 100 cm s
kDarcy
cm 1
A cm 2 1033 980g 2 2
s cm
1 Darcy 10 -8 cm 2
EXERCISE 2
Characterize the following reservoirs in terms of their porosity and permeability distributions.
(a)
kx = 100 md kx = 100 md
ky = 200 md ky = 150 md
= 10% = 10%
kx = 100 md kx = 100 md
ky = 80 md ky = 100 md
= 10% = 10%
(b)
kx = 100 md kx = 100 md
ky = 100 md ky = 100 md
= 10% = 10%
kx = 100 md kx = 100 md
ky = 10 md ky = 100 md
= 10% = 10%
Chapter 1 Page 73
(c)
kx = 80 md kx = 60 md
ky = 30 md ky = 20 md
= 8% = 11%
kx = 100 md kx = 50 md
ky = 40 md ky = 30 md
= 10% = 6%
(d)
kx = 100 md kx = 100 md
ky = 150 md ky = 150 md
= 12% = 12%
kx = 100 md kx = 100 md
ky = 150 md ky = 150 md
= 12% = 12%
EXERCISE 2 SOLUTION
(a) Homogeneous in kx and f; Heterogeneous in ky; Anistropic in permeability
(b) Homogeneous in kx, ky, and f; Isotropic in permeability
(c) Heterogeneous in kx, ky, and f; Anistropic in permeability
(d) Homogeneous in kx, ky, and f; Anistropic in permeability
EXERCISE 3
In an anisotropic formation, if kx = 40 md and ky = 80 md, what is the expected value of lateral
permeability that will be derived from a well test using radial flow theory?
EXERCISE 3 SOLUTION
The well test derived equivalent lateral direction permeability for an anisotropic medium will be
expected to be:
k h k x k y 40 80 57md
EXERCISE 4
In general, would you expect the duration of wellbore storage effects to be longer in horizontal
wells or vertical wells?
Chapter 1 Page 74
EXERCISE 4 -SOLUTION
In horizontal wells, duration of the wellbore storage effects is expected to be longer. This is
because of the greater wellbore volume encountered in horizontal wells, and because of the low
values of vertical permeability.
EXERCISE 5
a. Write the diffusivity equation for a horizontal, two-dimensional homogeneous anisotropic
formation.
After implementation of the coordinate transformation
x y
X and Y
kx ky
b. what will be the form of the diffusivity equation in the equivalent isotropic equation?
EXERCISE 5-SOLUTION
a.
2p 2p p
kx 2
k y 2 c t
x y t
b.
2p 2p c t p
X 2 Y 2 k h t
where
k h k xk y
EXERCISE 6
In a three dimensional anisotropic porous medium ( kx ¹ ky ¹kz), what is the equivalent isotropic
permeability?
EXERCISE 6-SOLUTION
k h 3 k x k yk z
EXERCISE 7
What are the common factors that can corrupt the vertical interference and the vertical pulse
testing techniques?
EXERCISE 7--SOLUTION
Wellbore storage
Fluid communication across the packer
Fluid communication behind the pipe (poor cementing)
Presence of vertical fractures
Chapter 1 Page 75
Upper and lower boundaries of the formation should be totally sealed
Fluctuations in the injection rate
Presence of close boundaries
Presence of shale streaks
EXERCISE 8
The vertical permeability values inferred from pressure transient analysis should be checked
against what kind of other sources of information?
EXERCISE 8-SOLUTION
Cores
Logs
Geological interpretations
Inverse simulation (history matching)
Chapter 1 Page 76
Pressure Transient Analysis in Heterogeneous
Reservoirs
LATERAL HETEROGENEITIES
LINEAR BOUNDARIES
The assumption of an infinite acting reservoir forms the basis for classical well test analysis. But
in reality, a reservoir’s continuity is terminated by boundaries. A reservoir boundary may be one
or more of the following types:
No-flow boundary (e.g., a sealing fault)
Constant pressure boundary (e.g., edge water, gas cap)
Partially sealing boundary (e.g., lateral heterogeneities)
These boundaries can be considered as simple subsets of lateral heterogeneities.
LINEAR NO-FLOW BOUNDARY
The method of images is instrumental in representing the pressure response of a well located
near a no-flow boundary. This treatment applies the principle of superposition in considering two
producing wells (Figure 1 ):
Figure 1
The actual well
An imaginary well on the other side of an impermeable boundary
The imaginary well is a reflection of the actual well.
At the beginning of a pressure drawdown or buildup test, when t is small, the effects of the fault
are not noticeable. As t increases, the fault’s effect on the pressure response of the well
becomes more pronounced. This can be explained more explicitly by the superposition principle,
and by following the changes in the arrangement of the exponential integral.
Referring to Figure 1, we can express the pressure drop at the well during a drawdown test as
follows:
c t 2d
2
70.6qB
p mLog t S Ei 4
kh 4 2 . 637 10 kt
where m is the usual Horner plot slope, defined as
Chapter 1 Page 77
162.6qB
m psi/cycle
kh
At the beginning of the test, when t is small, the argument of the Ei function will be large. Thus,
the value of the Ei function is essentially zero, and the entire equation collapses to:
p = m(log t + S)
As the test continues, t gets larger; consequently, the argument of the Ei function becomes
smaller and eventually can be approximated by the logarithmic approximation. Therefore, at late
time, the pressure drop equation takes the following form:
p = 2m(logt) + (constant)
The above equation indicates that at late time, a second semilog straight line with a slope two
times larger than the first one develops. Figure 2 (Sabet, 1991—courtesy Gulf Publishing
Company) shows a typical Horner plot for a well located near a fault.
Figure 2
The distance to the linear boundary may also be estimated by finding the intersection of the
early and late semilog straight lines tx and using the equation:
1
k 2
d 0.0112 tx
c t
CONSTANT PRESSURE BOUNDARY
Chapter 1 Page 78
A constant pressure boundary represents the support of the reservoir pressure by fluid
encroachment (either water influx or gas cap expansion), and leads to the development of true
steady-state conditions. The analysis of systems with constant pressure boundaries, like those
with linear no-flow boundaries, is performed using the method of images. In this case, the
constant pressure boundary is replaced by an image well of identical strength but with opposite
sign. In other words, if the actual well is a producer, then the image well is an injector, and if the
actual well is an injector than the image well is a producer (Figure 3 ).
Figure 3
When a producing well near a constant pressure boundary is tested, the following equation
describes the variation of pressure drop with time:
c t 2d
2
qB
p mLog t 70.6 Ei
kh 4 2.637 10 kt
4
The above equation differs from the one that was presented for the sealing fault case in that the
sign of the second term is reversed; (this is because the strength of the image well has the
opposite sign. Again, at early times the contribution of the second term is negligible. Therefore,
the basic drawdown equation applies. At late times, the magnitude of the second term
approaches to the magnitude of the first term; and since they have the opposite signs (recall that
Ei(-x) is always negative), DP approaches zero. Figure 4 shows a typical plot of DP versus the
logarithm of time.
Chapter 1 Page 79
Figure 4
From the intersection time tx of the first semilog straight line and the second straight line with
m=0, we can calculate the distance to the constant pressure boundary using the equation:
1
kt x 2
d
1.781c t
PARTIALLY SEALING BOUNDARIES
While geophysical methods can be effective in qualitatively identifying a fault as sealing or non-sealing, pressure transient data can be used to quantify the non-sealing
nature of the fault. The fault zone is considered a semi-permeable barrier with negligible storage properties. Figure 5 shows two different permeability values— k and
kf, where 0 < kf < k—assigned respectively to the formation and the fault block,.
Figure 5
The pressure drop P for this case will be less than the one experienced when kf = 0 and more
than the case when kf = k. Therefore, we need to calculate the extra pressure drop caused by
the fault.
Figure 6 shows the expected pressure response of a well located close to a non-sealing fault.
Note that this response is similar to that of a naturally fractured reservoir, although positions of
the two parallel lines with respect to the time axis are switched.
Chapter 1 Page 80
Figure 6
COMPOSITE RESERVOIRS
Solutions for systems with radial lateral discontinuities are more complex. As shown in Figure 7 ,
the system considered here is composed of two concentric domains with varying hydraulic
diffusivities.
Figure 7
Chapter 1 Page 81
The lateral discontinuities may be caused by moving fluid fronts, as in the case of water, steam
or polymer flooding. In thermal recovery processes, the composite system that reflects the
change in the system’s physical properties is mostly formed by the contrast in the reservoir
properties. In polymer flooding and steam injection, the contrast in fluid properties of different
fluid banks is more significant.
The characteristic behavior of a semilog plot of drawdown data is a straight line of slope m that
represents the conductivity of an inner fluid bank extending from the center of the well to the
interface. When the dimensionless time based on the distance from the well to the discontinuity
is greater than 0.25, we see a transitional curve whose duration and shape are dependent on
the storage ratio of the two regions. The drawdown curve forms a second straight line with a
slope that is inversely proportional to the conductivity of the outer zone when the outside region
provides the produced fluid.
STATIONARY INTERFACE
The use of pressure transient analysis in defining the locations of flood fronts is an efficient way
of evaluating improved oil recovery processes. Most of the analytical and numerical solutions
proposed assume that the location of the interface stays stationary during the test period. In this
way, the interface boundary condition simply states that no matter from which direction one
approaches the interface, the interface pressure is single-valued. In solving the problem, we use
a Laplace transformation with analytical inversion. If the position of the interface is not within the
immediate vicinity of the wellbore and the duration of the test is not extended, then the
assumption of a stationary interface is reasonable.
MOVING INTERFACE
If a well test is extended for a relatively long time when the interface is close to the wellbore,
than the stationary interface assumption becomes attenuated. Extensive testing of composite
system models shows that the assumption of stationary flood fronts is no longer valid if the test
period is on the order of hundreds of days. Since most tests are much shorter than this, the
stationary interface assumption, in general, is not considered restrictive. If we need to consider
the dynamic nature of the interface, then the entire solution needs to be based within the
framework of a moving boundary type solution.
TYPE CURVES AND DIAGNOSTIC PLOTS
Since well test analysis is a systems analysis technique, a large number of diagnostic plots and
type curves are catalogued in the literature to help the well test engineer in diagnosing the
system specific properties. Refer, for example, to the following figures from Daltaban and Wall
(1998—courtesy of Imperial College Press):
Figure 8 ,
Chapter 1 Page 82
Figure 8
Figure 9 ,
Figure 9
Chapter 1 Page 83
and Figure 10 represent the variation of dimensionless pressure for reservoir
geometries that can be approximated in closed polygons.
Figure 10
Figure 11 ,
Chapter 1 Page 84
Figure 11
Figure 12 ,
Figure 12
Chapter 1 Page 85
Figure 13 ,
Figure 13
Figure 14 ,
Figure 14
Chapter 1 Page 86
Figure 15 ,
Figure 15
Figure 16 ,
Figure 16
Chapter 1 Page 87
and Figure 17 , represent some diagnostic plots indicating the pressure response
and pressure derivative response of systems with lateral heterogeneities.
Figure 17
STRATIFIED RESERVOIRS
The majority of oil and gas reservoirs are layered, or stratified. In well test analysis models,
stratified reservoirs are treated as horizontally continuous, homogeneous systems with isotropic
property distribution. The layers exhibit markedly different porosity and permeability values, but
they generally have the same lithologic features.
TESTS WITHOUT FLOW MEASUREMENTS
In some cases, layers are entirely separated and are not in hydraulic communication within the
formation, such that vertical crossflow does not take place. However, these isolated layers are
connected to each other within the wellbore. This reservoir model is known as a commingled
system (see Figure 18 ).
Chapter 1 Page 88
Figure 18
If the layers are not totally separated by the intermediate streaks, then interlayer vertical
crossflow takes place and layers are said to be in pressure communication with each other both
through the wellbore and within the formation (see Figure 19 ). The stratified reservoir can be
even more complex when certain layers are in hydraulic communication and when certain layers
are not. The pressure transient responses of the flow models described are different from each
other.
Chapter 1 Page 89
Figure 19
In a layered system, if only the total flow rate at the wellhead is known and the pressure
transient test is conducted like a single-layered system, then pressure transient data will look
exactly like the one obtained from a single-layer system. The calculated permeability will be
equal to the thickness-weighted average of the permeabilities of the individual layers.
n
k ihi
k calculated i1n
hi
i 1
The above observation is sound even if the contrast between the layers is significant. Note also
that the estimated skin factor will also be an average skin factor that can be assigned to the
system.
TESTS WITH FLOW MEASUREMENTS
In a stratified reservoir, if the well test objective is to determine permeability and skin factor for
each individual layer—regardless of whether the layers are in hydraulic communication only in
the wellbore, or in the formation as well—it will be necessary to measure the flow rates in the
individual layers. The concept of simultaneous measurement of pressure and flow rate downhole
provides a unique opportunity to determine individual layer properties. In tests involving flow
measurements, flow rates are measured at the top of each layer. The measured pressure
responses are combined using convolution, and the total response is matched using non-linear
regression. In other words, the reservoir model is convoluted to reproduce the test results
observed.
VERTICAL CROSSFLOW IN THE FORMATION
To simplify the discussion, we consider a stratified reservoir with two distinct layers, Layer 1 and
Layer 2. If these layers have permeabilities of k1 and k2 where k1 >> k 2, then the movement of
pressure transients in Layer 1 will be much faster than in Layer 2. This will create a vertical
pressure gradient between Layers 1 and 2, causing Layer 2 to charge into Layer 1. The direction
Chapter 1 Page 90
of flow from Layer 2 to Layer 1 becomes almost vertical if the permeability contrast between
these two layers is two orders of magnitude or more. The pressure transient behavior of the
reservoir however will be very similar to the homogeneous and isotropic reservoirs.
In some cases, it is possible to see a pressure transient response similar to the one observed in
double-porosity reservoirs. In order for this to happen, the layer with low permeability needs to
have much higher storage capacity than the layer with high permeability. In other words, the
lower permeability layer provides a non-uniformly distributed source to the high permeability
layer.
VERTICAL CROSSFLOW IN THE WELLBORE
In commingled reservoirs, each layer is considered homogeneous and isotropic, as in the case
of the classical well test analysis model. Therefore, the diffusivity equation can be written for
each layer. The important difference is the specification of the flow rate at the wellhead, which is
n
q total q j
i
Again, in order to simplify the problem, consider a commingled reservoir with only two layers.
Figure 20 shows the typical pressure drawdown response of a two-layer commingled reservoir.
In this figure, we see four distinct segments of the curve.
Figure 20
Segment AB: Infinite acting reservoir behavior in the high-permeability layer
Segment BC: Infinite acting period for the high permeability layer ends. During this
period, the low-permeability layer starts to become more dominant on the quality of
the pressure transient data.
Segment CD: The depletion of the second layer continues.
Segment DE: At this state, the boundaries of the low-permeability layer are felt and
the total system enters the pseudo-steady state regime.
We can identify the following overall characteristics of the pressure transient data from
commingled reservoirs:
Chapter 1 Page 91
The semi-log plot exhibits an early transient period. This portion of the data can be
analyzed to determine a total permeability-thickness product (kh)t. However, the
(kh)t value does not remain constant, and it continually decreases as the depletion
of the high permeability layers takes place.
The late-transient period is much more extended because of the very slow
response of the low permeability layers.
It is almost impossible to obtain the average reservoir pressure from a buildup test
in commingled reservoirs, as the duration of the shut-in time needs to be
unrealistically large due to rapid depletion of the high permeability layers.
One possible alternative test in commingled reservoirs—particularly in two-layered reservoirs
with dual completions—is the separate testing of each zone. In this method, the first layer is put
on production and tested first, and the kh and s values are determined. In the second phase of
the test, both layers are tested simultaneously. The analysis yields an equivalent skin and kh
product for both layers. The characteristics of the second layer can be extracted by examining
the differences between the two tests.
TYPE CURVES AND DIAGNOSTIC PLOTS
Bourdet (1985) generated families of double-permeability curves in semi-log and log-log
representations, from which we can identify three important characteristics of reservoirs
exhibiting vertical crossflow:
Before crossflow is established, the response is the same as for a commingled reservoir.
At late time, the pressure transient behavior characteristics of an equivalent homogeneous
system with a total permeability thickness and a total storativity are established.
At intermediate time, a transition behavior develops.
NATURALLY FRACTURED RESERVOIRS
Some of the world’s most prolific reservoirs are naturally fractured reservoirs. Fractures improve
a reservoir’s production performance by enhancing the rock permeability. A fractured reservoir
has two distinct components: matrix blocks and fractures.
Matrix blocks have the lower transmissivity of the two components. They are referred as the
primary porosity, because they hold the reservoir fluid; in other words, they act as a fluid
storage medium. Matrix permeabilities are very low, ranging from 10 -4 md to a few
millidarcies.
The surface areas of matrix blocks are exposed to natural fractures. The fractures are
adjacent to the matrix blocks, and are referred to as secondary porosity. Because of their
high transmissivity characteristics, fractures serve as conduits that take the fluids from tight
matrix blocks and deliver them to the well. In terms of the storativity characteristics of these
two porous media, matrix porosities can be as high as 30-35%, but fracture porosities can
be less than 0.01%.
NATURALLY FRACTURED POROUS MEDIA MODELS
Naturally fractured reservoirs are also known as dual porosity or double porosity systems. In a
dual porosity system, the matrix is assumed to have no permeability, and flow from the matrix to
the fracture network is by diffusion. If the well does not intersect a natural fracture, or if it is not
connected to it by a hydraulic fracture, it will not produce. If a naturally fractured reservoir is
referred to as a dual porosity, dual permeability system, then it is implied that both matrix and
fracture transport fluid.
The pressure transient behavior of a naturally fractured reservoir is significantly different from
that of a homogeneous reservoir. The double porosity behavior in terms of pressure response
will contain three characteristic segments, as shown in Figure 1 .
Chapter 1 Page 92
Figure 1
In this plot, we observe two parallel semilog straight lines connected by a transitional curve. The
initial and late lines reflect the pressure transient behavior of the fracture and matrix flows,
respectively. The transitional curve between the two parallel lines describes the characteristics
of the matrix and fracture interaction. This interaction is always explained by using a non-
equilibrium model (an equilibrium model will indicate instantaneous transfer from matrix to
fracture, and thus cannot account for the transitional episode). In non-equilibrium models, the
flow from matrix to fracture is treated in three different ways:
The matrix-fracture transfer rate is simply proportional to the pressure difference
between the matrix block and the fracture. This model is known as Warren and
Root model, also referred to as the pseudo-steady state model.
The matrix-fracture transfer rate is proportional to the average pressure gradient
established within the matrix. This is the Streltsova pressure gradient model. In this
gradient model, the transitional portion of the pressure response curve develops
into a straight line, rather than a curve on a semilog plot.
The matrix-fracture transfer rate is determined by examining the flow within the
matrix system. Therefore, flow rate from matrix to fracture becomes an unsteady-
state function of pressure drop across the matrix. This is deSwaan’s model, also
known as the true unsteady-state model. This model preserves the principal
features of the Warren and Root model.
TRANSIENT FLOW BETWEEN FRACTURE AND MATRIX
BLOCKS
In studying the pressure transient behavior of naturally fractured reservoirs, we must consider
two sets of properties: one for the matrix and another for the fracture. Accordingly,
we have:
Chapter 1 Page 93
A Matrix permeability km
Matrix porosity m
Matrix ctm
compressibility
B Fracture kf
permeability
Fracture porosity f
Fracture ctf
compressibility
For naturally fractured reservoirs, the dependent and independent variables dimensionless
pressure pD and dimensionless time tD are defined as:
k fh
pD p
141.2qB
and
tD
2.637 10 k t
4
f
f c tf m c tm rw2
The double-porosity characteristics of a naturally fractured reservoir are quantified by introducing
two new parameters known as the storativity ratio, and the transmissivity ratio, :
f c tf
f c tf m c tm
km 2
rw
kf
The coefficient used in the definition of l is known as the shape factor, and has different values
depending upon the shape of the matrix blocks (cubes, spheres, cylinders, etc). The value of w
is between zero and one. A value =1 indicates a single-porosity reservoir. As becomes
smaller, the fracture porosity f becomes a very small portion of the total system porosity. The
transmissivity ratio is usually a small entry. Its values lie between 10 -3 and 10-10. If is close to
10-3, then the naturally fractured character is not that dominant, and the reservoir can be treated
as a single-porosity system.
The Warren and Root solution that describes the pressure transient behavior of the fracture
network is a line-source solution given by:
1 t D t D
p wD Lnt D 0.809 Ei Ei
2 1 1
Chapter 1 Page 94
For early times, a logarithmic approximation can approximate both exponential integral
functions, and the short time representation of the above equation becomes
1
1 1 2
p wD Lnt D 0.809 Ln
2
For the late times, both exponential and integral functions vanish, and the representation of the
same equation becomes
1
Lnt D 0.809 2
1
p wD
2
Comparing the short time and long-time approximations, we observe that:
Short time and long-time semilog straight lines have the same slopes, and hence two
parallel semilog lines develop.
The second semilog straight line is shifted from the first semilog straight line by
1
1 2
Ln
Therefore, by looking at the separation of two semilog straight lines, it is possible to characterize
the naturally fractured reservoir in terms of its storativity characteristic.
TYPE CURVES AND DIAGNOSTIC PLOTS
The type curves of Figure 2 and Figure 2 (after Bourdet et al., 1984—Courtesy World Oil) are
developed for drawdown tests.
Chapter 1 Page 95
Figure 2
However, the same type curves can be used for buildup tests, provided that the total production
time before the shut-in is sufficiently large. If the total production time before shut-in is not
sufficiently large, the same type curves should be used with an equivalent time as defined by
t
t e
t
1
tp
In using these type curves, the early data should be matched with one of the type curves labeled
cDe2s. The permeability is calculated from the ordinate (pressure match). The final segment of
the data is also matched to one of the curves labeled as cDe2s. The intermediate segment is
matched to one of the curves labeled as:
c D
1
RESERVOIRS UNDERGOING IMPROVED RECOVERY
PROCESSES
During a secondary or tertiary recovery operation, radial discontinuities develop. This is because
the injected fluids have properties different from the properties of the in situ fluids that are being
displaced. Therefore, a saturation discontinuity is created, resulting in a composite system.
Testing of injection wells is important for determining transmissivity, storage capacity and well
conditions.
WATERFLOODING
In a waterflood, we can identify three zones with different characteristics (see Figure 1 ).
Chapter 1 Page 96
Figure 1
In these problems, the hydraulic diffusivity of each zone controls the collected pressure transient
data. In analyzing each region, we assume radial homogenous flow conditions. Depending on
the test duration, one, two or three semilog straight lines will develop. The slope of each line will
be used in estimating the mobility of each fluid in each zone. A proper analysis of the injection
well pressure transient data will also yield information on the location of the flood front (i.e., the
radial discontinuity).
STEAM INJECTION
A multi-zone model, as described in water flooding operations, is also applicable in steam
injection operations. In steam injection, the contrast between the mobilities of each zone is more
pronounced than it is in a waterflooding operation. Calculation of the distance to the radial
discontinuity is important, as this information is useful in determining the amount of heat lost to
the overlying and underlying formations. In some cases, if a system is not concentric (e.g., as in
a fault-bounded reservoir) then it becomes necessary to match the pressure transient data using
a numerical model.
EXERCISE 1
Prepare a qualitative diagnostic plot that represents the pressure response of a well located in a
recently discovered field. The well is tested by producing at a constant rate. The well is
hydraulically fractured and it is known that there is a sealing fault in the vicinity of the wellbore. It
is also known that the entire test is conducted under infinite acting reservoir conditions.
EXERCISE 1—SOLUTION
See Figure 1 .
Chapter 1 Page 97
Figure 1
Note the following segments of the data
Wellbore storage effects (unit slope)
Linear flow in the formation (1/2 slope)
Pseudo-radial flow regime (3/2 cycles)
Radial flow regime (flat portion of the curve)
Boundary effects (last 5 to 6 data points)
EXERCISE 2
Prepare a qualitative Horner plot for the problem described in Exercise No. 1.
EXERCISE 2—SOLUTION
See Figure 1 .
Chapter 1 Page 98
Figure 1
Note the following segments of the plot:
Wellbore storage
Negative skin (due to hydraulic fracture)
Increasing slope (due to hydraulic fracture)
First semilog line (constant slope of m)
Second semilog line (constant slope of 2m)
EXERCISE 3
A reservoir consists of three layers, two of which are in hydraulic communication. The third one
is isolated from the other two by an impervious layer. What kind of well test analysis strategy
would you suggest for such a system?
EXERCISE 3--SOLUTION
The reservoir is made up of crossflow and commingled systems. Therefore, its overall behavior
will be similar to that of a commingled reservoir. The behavior of the crossflow zone will be
similar to the behavior of a homogeneous and isotropic layer. Thus, from a practical point of
view, the best approach will be to test each zone separately as in the case of commingled
systems.
EXERCISE 4
Chapter 1 Page 99
Mark the following statements as TRUE or FALSE. In each case, justify your answer.
1. The early time behavior of cross-flow reservoirs is similar to that of commingled
reservoirs
2. The total flow rate of a cross-flow reservoir is less than the flow rate of a commingled
reservoir of the same storage and flow capacity
3. In layered reservoirs, individual layer flow rates are directly proportional to the layer
permeabilities.
EXERCISE 4—SOLUTION
1. TRUE.
During the early time, vertical pressure gradients between the layers of the
crossflow reservoir are not fully established. Therefore, the early time behaviors for
both reservoirs are similar.
2. FALSE.
Presence of cross flow increases the overall depletion rate of the low permeability
layer. Therefore, the total flow rate of a crossflow reservoir is larger than that of a
commingled reservoir of the same flow capacity and the same storage.
3. FALSE
The statement is only true if layers have no skin or have the same skin.
EXERCISE 5
The two Horner plots shown in Figure 1 are from two wells (Well-A and Well-B). Compare
(qualitatively) the pressure transient data presented for both wells. The wells are located in
naturally fractured reservoirs.
Chapter 1 Page 100
Figure 1
EXERCISE 5—SOLUTION
Both data sets exhibit a common response for the early and late time lines. Therefore, both
of the systems have the same permeabilities.
The duration of the transition from the early-time line to the late-time line is longer for well A.
This indicates that the contrast between the matrix and fracture properties for well A is more
pronounced.
EXERCISE 6
Figure 1 represents the Horner plots for the pressure drawdown data collected for wells A and B,
which are located in the same naturally fractured reservoir. Compare and contrast (qualitatively)
the pressure responses for each set.
Figure 1
Chapter 1 Page 101
EXERCISE 6--SOLUTION
Pressure transient data from both wells display similar naturally fractured reservoir
characteristics.
Earlier departure from the late-time line for well A indicates that well A is situated closer to
the external boundaries of the reservoir than well B.
EXERCISE 7
Consider the first two zones formed in a water flooding operation. How will the injection pressure
transient data be differed if the operation is:
A favorable mobility ratio injection?
An unfavorable mobility ratio injection?
EXERCISE 7—SOLUTION
In a favorable mobility injection operation, the mobility of the inner zone is less than the mobility
of the outer zone. Therefore, the slope of the first semilog straight line will be higher than the
slope of the second semilog straight line.
In an unfavorable mobility injection operation, the radial discontinuity will lose its integrity
because the injected water will finger into the oil zone. If two semilog straight lines are observed,
then the slope of the early semilog straight line will be less than that of the late semilog straight
line.
EXERCISE 8
Mark the following statements as TRUE or FALSE. In each case, justify your answer.
1. In analyzing the pressure transient data from a water or steam injection well, two
semilog straight lines will always be observed.
2. The duration of the transition from the first semilog straight line to the second semilog
straight line does not yield any usable information.
EXERCISE 8—SOLUTION
1. FALSE.
If the well test is terminated before transients reach the radial discontinuity, only
one straight line will be observed.
2. FALSE.
If the transition from the first semilog straight line to the second one takes an
excessive amount of time, this will signal an uneven progression of the flood front.
Chapter 1 Page 102
Pressure Transient Analysis in Multi-Phase
Reservoirs
TRANSIENT SOLUTIONS FOR MULTI-PHASE FLOW
When flow takes place under multi-phase conditions, one of the basic assumptions of classical
well testing theory—that of a single phase fluid—is rendered invalid. To describe multi-phase
flow, the conservation and transport equations must be written for each phase. This requires
consideration of capillary phenomena and of relative permeability, which controls the mobility of
each phase as a function of saturation. Clearly, multi-phase flow dynamics add another
dimension of complexity to the task of well test analysis.
ANALYTICAL SOLUTIONS
Analytical approaches to including multi-phase flow effects involve developing a methodology
that preserves the simplicity and practicality of a single fluid theory. Three approaches, which
are summarized below, are most commonly used. For a more detailed treatment of these
approaches, refer to the IPIMS subject heading Pseudo-Representation of Multi-Phase Flow,
which appears under the subtopic heading Multi-Phase Flow Equations.
PRESSURE APPROACH—PERRINE’S METHOD
In this approach, the single-phase flow equation is utilized in its original form. The total mobility
(sum of mobilities of each phase) and the total compressibility (sum of compressibilities of each
phase as weighted by the respective phase saturations) concepts are introduced and the
equation is still kept in the pressure form. This approach implicitly assumes that pressure and
saturation gradients are small, and that saturation changes during the duration of the test and
capillary pressures are negligible. It is found that the method becomes attenuated as the gas
saturation increases. This is explained by the more non-linear nature of the gas phase.
PRESSURE-SQUARED APPROACH -AL-KHALIFA ET AL.
METHOD
This approach is based on the assumption that the transmissivity of a phase is a linear function
of pressure:
ko
p
oB o
When the logarithmic approximation is applied, the equation describing the infinite acting
(semilog straight line) behavior can be written as:
325.1q o
p i2 p 2wf Log t Log k Log k 3.227 0.87 s
h r 2
c t w
The skin factor equation is:
p i2 p 12hr k 1
s 1.151 Log Log 2 3.227
m c t rw
In the analysis, p2 versus Log t is plotted, and the slope of semilog straight line is defined as:
325.2q o
m
h
Since a at initial conditions can be written as
Chapter 1 Page 103
k 1
o
oB o i p i
Combining the definition of a as written at the initial conditions with the definition of the slope of
the semilog straight line, one obtains:
325.2q o oB o i p i
ko
mh
The above equation is strictly correct for a drawdown analysis. For a buildup analysis, the same
expression can be written as:
325.2qo oB o i p
ko
mh
We may determine the permeabilities of the other phases from the producing gas-oil ratio (GOR)
and water-oil ratio (WOR) data using the definitions of mobility ratios. Once we define the
effective permeabilities, then we can calculate the absolute permeability of the formation using
the definition of total mobility.
PSEUDOPRESSURE APPROACH—RAGHAVAN’S METHOD
This method employs a pseudopressure definition,
p
k ro
p dp
0 oB o
which is analogous to the real gas pseudopressure. However, this approach requires the
knowledge of the relative permeability relations. This method is found to be very sensitive to the
relative permeability data.
SEMI-ANALYTICAL SOLUTIONS
Analytical solutions are applicable for small saturation changes. In the case of highly volatile oils,
this assumption of small saturation changes does not hold. Djatmiko et al. [1995] have
developed a semi-analytical interpretation procedure that is capable of predicting in situ pressure
and saturation distributions and effective permeability. Solution gas drive reservoirs with
immobile water saturation are good candidates for this procedure. The procedure requires the
calculation of total compressibility and total mobility in an iterative manner for every pressure.
This information is then used to generate a pressure distribution. Then, the oil phase saturation
is calculated between S0min (at sandface) and S0max (at the radius where pressure is above the
saturation pressure). Within this envelope of saturation changes a quadratic expression is used
to assign a monotic change to the saturation values. At each iteration, the calculated pressure
and saturation values are used to update the pressure and saturation- dependent entries.
NUMERICAL SOLUTIONS
In complex systems (e.g., multi-phase flow), it may become essential to use a computer and
some sort of mathematical model to simulate system behavior. Numerical solutions are widely
used in solving complex mathematical models. The task of the reservoir simulator is to calculate
the system pressure response resulting from a specific flow-rate history. Special reservoir
simulators structured for well-testing applications are mostly written in radial coordinates to
capture the flow geometry, near wellbore phenomena and wellbore storage effects. With the aid
of these special-purpose reservoir simulators, one can history match the pressure data. The
reservoir input parameters that match the well test data will form the results of the well test
analysis. In using numerical solutions and history matching, engineering judgment should always
be in place as the history matching process suffers from the non-unique character of the
solution.
Chapter 1 Page 104
TRANSIENT SOLUTIONS FOR VARIOUS RESERVOIR
DRIVE MECHANISMS
Multi-phase well test models incorporate changes in effective permeability and fluid properties
such as formation volume factor, viscosity, and gas solubility. These changes are usually
introduced using a pseudopressure function. This approach is based on the premise that by
using an expression similar to the real gas pseudopressure, equations describing the
simultaneous flow of oil and gas through porous media can be simplified considerably.
SOLUTION GAS DRIVE RESERVOIRS
The steady-state flow equation written for the oil phase in differential form is given by:
2kk ro h dp
q
oB o dr
If we integrate the above equation after separating the variables, we obtain (in radial-cylindrical
coordinates) in field units
pe
kh k ro S o
q
r oB o
dp
141.2 Ln e r p
We can define a pseudopressure function, m (p) as
pe
k ro S o
mp dp
p oB o
Calculation of the pseudopressure function is different for pressure drawdown and buildup tests.
In order to perform the transformation, we need to specify a relationship between fluid saturation
and pressure. This relationship originates from:
k g oB o
R Rs
k o gB g
which is the gas-oil ratio (GOR) equation. In a drawdown test, the producing GOR curve is
converted to oil saturation versus pressure curve using the above equation. Then, the
pseudopressure function as a function of pressure can be calculated using a numerical
integration (e.g., trapezoidal rule). On the other hand, in a buildup test, the producing GOR is
assumed to be constant and a relationship between oil saturation and shut in sandface pressure
can be obtained directly from the GOR equation. This is followed by the calculation of the
pseudopressure function, once the pseudopressure function is calculated.
In order to conduct a pressure transient analysis in a solution gas reservoir, it is necessary to
have the PVT relations of the oil and gas and also the relative permeability relationship. In
general, a solution gas drive reservoir going through a drawdown test will experience a slightly
higher pressure drop compared to that of a slightly compressible liquid. This is due to gas
coming out of solution and decreasing the effective permeability to oil. Some studies consider
the formation of gas saturation as a skin effect on the pressure transient data. However, with the
use of the pseudopressure function, the incorporation of a skin effect to account for the gas
saturation is not necessary.
The use of two different pseudopressure functions for the drawdown and buildup tests imply that
in solution gas drive reservoirs, the drawdown and buildup tests are not exactly reversible.
GAS CAP RESERVOIRS
In reservoirs overlain by a large gas cap, the proper well test interpretation methodology will be
to use the gas/oil contact as the constant pressure boundary, with the pressure value being the
average pressure in the gas cap. This application assumes that the gas/oil contact between the
Chapter 1 Page 105
gas cap and the oil zone defines the boundary of the well’s drainage area, meaning that the
tested well will exhibit the conventional transient flow period followed by the steady state flow
period. If two-phase flow conditions (gas and oil) are dominant in the oil zone, the methodologies
described in the previous section will be applicable, with consideration given to the constant
pressure boundary placed at the gas/oil contact.
Above the gas/oil contact, the total mobility exceeds the total mobility in the oil zone. Since the
mobility of the gas zone is higher than the oil mobility, the gas phase can develop fingers of gas
into the oil zone. Therefore, the wells to be tested using the methodology described above
should be located in a lower part of the oil zone.
The general drawdown behavior of a gas-cap reservoir for short duration tests is free of gas-cap
dominance. However, since the early data are generally influenced by wellbore storage and near
wellbore conditions; these data cannot be analyzed by semilog analysis. An analysis based
upon a match of the early-time data to after-flow type-curves is a possibility for estimating the
properties of a gas-cap reservoir.
GAS CONDENSATE RESERVOIRS
In solution-gas drive reservoirs, the gas-oil ratio expression is instrumental in calculating the
pseudopressure function. In gas-condensate systems such an expression (or its analogue) does
not exist. During the early stages of production, however, a single-phase region will exist;
therefore, we can use a single-phase analogy.
In the vicinity of the wellbore, the liquid-phase saturation increases after the well is shut in, as
the liquid phase does not vaporize during the shut in period in the near-well region. Well test
analysis can be instrumental in estimating the size of the two-phase region. The size of the two-
phase zone can be predicted from an estimate of the total skin factor. If the drainage-radius
concept is used, then it will not be necessary to know two-phase skin factor to calculate the
radius of two-phase zone.
EXERCISE 1
If single-phase formulation (for the most dominant phase) is used to analyze a well test data
recorded in a multi-phase system. What would be your expectations of the quality of the
analysis? Explain.
Under what conditions will your expectations be accentuated?
Chapter 1 Page 106
Chapter 1 Page 107
Chapter 1 Page 108
Chapter 1 Page 109
Chapter 1 Page 110
Chapter 1 Page 111
You might also like
- Identification of Well Problems Using Well Testing100% (2)Identification of Well Problems Using Well Testing42 pages
- Testing Design and Analysis: Well StingNo ratings yetTesting Design and Analysis: Well Sting18 pages
- 1.1 Role of Oil Well Tests and Information in Petroleum IndustryNo ratings yet1.1 Role of Oil Well Tests and Information in Petroleum Industry12 pages
- Pressure Transient Analysis A Tool For Reservoir Assessment A Case Study From The Field of Assam Asset Jaipur 2015No ratings yetPressure Transient Analysis A Tool For Reservoir Assessment A Case Study From The Field of Assam Asset Jaipur 20156 pages
- Well Test Interpretation SKPP4323: Azmi Mohd Arshad Department of Petroleum Engineering100% (1)Well Test Interpretation SKPP4323: Azmi Mohd Arshad Department of Petroleum Engineering47 pages
- Lecture 13. Well Test Interpretation WorkflowNo ratings yetLecture 13. Well Test Interpretation Workflow45 pages
- Oil_and_Gas_Well_Testing-Class_Note-1_1737440545558891810678f3d21a73d9No ratings yetOil_and_Gas_Well_Testing-Class_Note-1_1737440545558891810678f3d21a73d9231 pages
- Dokumen - Pub - Modern Pressure Transient Analysis of Petroleum Reservoirs A Practical View 3031288882 9783031288883No ratings yetDokumen - Pub - Modern Pressure Transient Analysis of Petroleum Reservoirs A Practical View 3031288882 9783031288883230 pages
- SPE-102079-PA Gringarten From Straight Lines To Deconvolution100% (1)SPE-102079-PA Gringarten From Straight Lines To Deconvolution22 pages
- Principles of Transient Testing: 1.1.1 Description of A Well Test100% (3)Principles of Transient Testing: 1.1.1 Description of A Well Test23 pages
- From Straight Lines To Deconvolution: The Evolution of The State of The Art in Well Test AnalysisNo ratings yetFrom Straight Lines To Deconvolution: The Evolution of The State of The Art in Well Test Analysis22 pages
- Lesson 1 Introduction To Well Test AnalysisNo ratings yetLesson 1 Introduction To Well Test Analysis3 pages
- Input Data Required For Well Test Analysis Test Data: Flow Rate and Bottom Hole Pressure As A Function of Time. The Sequence of Event Must BeNo ratings yetInput Data Required For Well Test Analysis Test Data: Flow Rate and Bottom Hole Pressure As A Function of Time. The Sequence of Event Must Be4 pages
- From Straight Lines To Deconvolution The EvolutionNo ratings yetFrom Straight Lines To Deconvolution The Evolution23 pages
- Well Testing Analysis Ibrahim Yaquob May 2021 1100% (1)Well Testing Analysis Ibrahim Yaquob May 2021 133 pages
- Systematic Methods of Water Quality Parameters Analysis: Analytical MethodsFrom EverandSystematic Methods of Water Quality Parameters Analysis: Analytical Methods1/5 (1)
- Advances in the iterative coupling between flow-geomechanical simulators: applied to cases with different contour conditionsFrom EverandAdvances in the iterative coupling between flow-geomechanical simulators: applied to cases with different contour conditions5/5 (1)
- +++A Short Survey in Application of Ordinary Differential Equations On Cancer ResearchNo ratings yet+++A Short Survey in Application of Ordinary Differential Equations On Cancer Research5 pages
- Origin of Partial Differential Equations: Chapter-1100% (2)Origin of Partial Differential Equations: Chapter-128 pages
- Differential Equation Models in EpidemiologyNo ratings yetDifferential Equation Models in Epidemiology135 pages
- Lecture Notes On Numerical Methods For Differential EquationsNo ratings yetLecture Notes On Numerical Methods For Differential Equations146 pages
- Vibration with Control 1st Edition Daniel J. Inman - Own the ebook now with all fully detailed content100% (1)Vibration with Control 1st Edition Daniel J. Inman - Own the ebook now with all fully detailed content53 pages
- Data-Driven and Physics-Informed Deep Learning Operators For Solution of Heat Conduction EquationNo ratings yetData-Driven and Physics-Informed Deep Learning Operators For Solution of Heat Conduction Equation8 pages
- Baranov (1957) Pseudo Gravimetric AnomaliesNo ratings yetBaranov (1957) Pseudo Gravimetric Anomalies24 pages
- Re Exam ESE Time Table May -June 2024-2025 Odd & Even SemNo ratings yetRe Exam ESE Time Table May -June 2024-2025 Odd & Even Sem5 pages
- Chapter 8. Numerical Solution of Ordinary Differential EquationsNo ratings yetChapter 8. Numerical Solution of Ordinary Differential Equations18 pages
- UG Curriculum With Course Details Final 130716 PDFNo ratings yetUG Curriculum With Course Details Final 130716 PDF118 pages
- Engineering Mathematics 2 - 2024 SyllabusNo ratings yetEngineering Mathematics 2 - 2024 Syllabus2 pages
- Be Mechanical Engineering Cbcs Syllabus 2015 16 OnwardsNo ratings yetBe Mechanical Engineering Cbcs Syllabus 2015 16 Onwards492 pages
- Syllabus: MA 201-Partial Differential Equations (PDE) : ReferencesNo ratings yetSyllabus: MA 201-Partial Differential Equations (PDE) : References20 pages
- Instant Access to Symmetry Analysis of Differential Equations An Introduction 1st Edition Daniel J. Arrigo ebook Full Chapters100% (8)Instant Access to Symmetry Analysis of Differential Equations An Introduction 1st Edition Daniel J. Arrigo ebook Full Chapters72 pages
