Documente Academic
Documente Profesional
Documente Cultură
2014
JISHAN LIU
Email: Jishan.liu@uwa.edu.au
Tel: 6488 7205
Water Compressibility
Gas Compressibility
Oil Compressibility
Grain Compressibility
Bulk Compressibility
Unit Outline
Reservoir Engineering
PETR3511
TS-L-1, 2014
Campus: Crawley
All material reproduced herein has been copied in accordance with and pursuant to a statutory licence administered by Copyright
Agency Limited (CAL), granted to the University of Western Australia pursuant to Part VB of the Copyright Act 1968 (Cth).
Copying of this material by students, except for fair dealing purposes under the Copyright Act, is prohibited. For the purposes of this fair
dealing exception, students should be aware that the rule allowing copying, for fair dealing purposes, of 10% of the work, or one
chapter/article, applies to the original work from which the excerpt in this course material was taken, and not to the course material
itself
Contact Details
Faculty Faculty of Engineering, Computing & Mathematics
School Mechanical and Chemical Engineering
School website http://www.mech.uwa.edu.au
Unit coordinator Professor Jishan Liu
Unit coordinator email jishan.liu@uwa.edu.au
Unit coordinator telephone number 08 6488 7205
Consultation hours From 4PM to 5PM before lectures
Lecturers
Name Position Email Telephone Number
Jishan Liu Professor jishan.liu@uwa.edu.au 64887205
Tutors
Yan Peng: 21364783@student.uwa.edu.au
Unit Description
Description
This unit introduces the basic knowledge of reservoir engineering necessary to analyse and predict reservoir
performance. Topics include reservoir formation, classification, description, volume; pressure distribution in
reservoirs; reservoir fluid properties; reservoir rock properties; capillary pressure, saturation and relative permeability;
laboratory tests and other means to gather rock and fluid properties; basic phase behaviour, gaseous state, liquid
state, single-phase and multiphase flow and well productivity; material balance equations; recovery mechanisms; water
and gas coning; and water influx.
Learning outcomes
Students develop a basic understanding of the origin of oil and gas reservoirs, properties of reservoir rock, and static
and flow properties of reservoir fluids; gain knowledge to analytically model the hydrocarbon volume in the reservoir,
flow behaviour of hydrocarbon and water through reservoir rock and production performance of vertical and horizontal
wells; and understand various recovery mechanisms, water influx and water coning.
Unit structure
This unit describes (1) the fundamental properties of rocks and petroleum fluids, (2) how these properties change during the
production; (3) different producing mechanisms; (4) generalized material balance equation; (5) flow principles in porous media; (6)
applications of material balance equation and flow principles to the analysis of reservoir performances.
Students develop essential knowledge and skills for the accurate estimation of original hydrocarbons in place, for the performance
evaluation of full range of reservoir fluids, from dry gases to heavy oils, including gas condensates and volatile oils, and for the
evaluation of ways and means of increasing the ultimate recovery. Students fully understand one principle (mass conservation law),
two applications (generalized material balance equation and fluid flow in rocks), and three data sets (rock properties, fluid properties,
and production measurements). Class assignments and the final exam are designed to reinforce these objectives.
Assessment
Assessment overview
PETR3511 (2014) : Unit Information Management System : The University of Western Australia Page 2 of 3
This comprises continuous assessment and an end-of-semester examination.
Assessment mechanism
# Component Weight Due Date Relates To Outcomes
1 Assignments 30 TBA 1&2
2 Final Exam 70 TBA 3
Assessment items
Item Title Description Submission Procedure for Assignments
Assignment I Rock and Fluid Properties Online Submission
Assignment II Reservoir Performance Online Submission
Final Exam Overall understanding Closed book
Course notes: Liu, J. Reservoir Engineering: UWA School of Mechanical Engineering 2014
PETR3511 (2014) : Unit Information Management System : The University of Western Australia Page 3 of 3
PREFACE
We have been offering this 6-credit unit of Reservoir Engineering since 2005. Reservoir
Engineering deals with the reservoir on a gross average basis. The objective of Reservoir
Engineering is to develop essential knowledge and skills for the accurate estimation of
original hydrocarbons in place, for the performance evaluation of full range of reservoir fluids,
from dry gases to heavy oils, including gas condensates and volatile oils, and for the
evaluation of ways and means of increasing the ultimate recovery.
Petroleum Reservoir: A petroleum reservoir is a continuous body of oil and/or gas that
occurs within a single geological trap. A reservoir may be small (a few acres) or it may extend
over many square miles. A field consists of one or more reservoirs.
e p 0
Overburden Weight
p, T
e p
Grain Stress or Liquid
Effective Stress Pressure
External Geometry: Defined by seals or flow barriers which inhibit the migration of
hydrocarbons, forming a hydrocarbon trap.
The reservoir pore-fluid pressure is a fraction of the overburden pressure that is supported by
the fluid system. The other portion is supported by the rock. The overburden pressure is
created by the weight of rocks. Reservoir pressure, p, can be calculated as
p e (1.1)
Fig.1.2 Relation between overburden stress, pore pressure, and grain stress (effective stress).
It is normally assumed that the overburden stress does not change during the life of
production. Under this assumption, Equation (1.1) becomes
p e (1.2)
Where p is the change in reservoir pressure, e is the change in the effective stress or the
grain stress.
e p e
2,000 m
e p 0
A
Density of Gas= 200 kg/m3
Gas 50 m Pressure gradient= 0.002 MPa/m
B
Density of Oil= 600 kg/m3
Oil 100 m Pressure gradient= 0.006 MPa/m
C
Density of Water= 1,000 kg/m3
Water 50 m Pressure gradient= 0.01 MPa/m
D
e p e
2,000 m
p e
A
Density of Gas= 200 kg/m3
Gas 50 m Pressure gradient= 0.002 MPa/m
B
Density of Oil= 600 kg/m3
Oil 100 m Pressure gradient= 0.006 MPa/m
C
Density of Water= 1,000 kg/m3
Water 50 m Pressure gradient= 0.01 MPa/m
D
Pressure
2000 2100 2200 psi
2000
Depth
2100
2200
In this unit, Reservoir Engineering is defined as the underground science of evaluating the
original hydrocarbons in place, the performance of the full range of reservoirs from dry gases
to heavy oils, including gas condensates and volatile oils, and the ultimate recovery. These
evaluations will be conducted based on a single variable, i.e., pore pressure, only.
p
INITIAL RESERVIR PRESSURE pinitial
Reservoir Pressure
OPERATIONAL RANGE
1. WHEN TO START?
2. WHEN TO STOP?
3. MANAGE FROM STRAT
TO STOP
pabandon
ABANDON RESERVIR PRESSURE
KEY QUESTION: WHAT WOULD CHANGE WHEN RESERVOIR
PRESSURE GOES FROM THE INITIAL VALUE TO THE FINAL ONE?
0 Reservoir Temperature T
In this unit, we evaluate the performance of a reservoir based on its response to pore pressure
only. Therefore, we can design the classification schemes accordingly.
Overlying rocks
Reservoir pressure (MPa)
2,000 m
22 MPa
A
Gas 50 m
B 22.1 MPa
pCD hw w
Oil 100 m 50 0.01MPa
C 22.7 MPa
From a technical point of view, the various types of reservoirs can be defined by the location
of the initial temperature and pressure with respect to the two-phase (gas and oil) region as
commonly shown on pressure temperature (PT) phase diagrams. This figure is the PT diagram
of a particular reservoir fluid. The area enclosed by the bubble-point and dew-point lines to
the lower left is the region of pressure-temperature combinations in which both gas and liquid
phases will exist.
Point A: initially at 300oF and 3700psi. Since this point lies outside the two-phase region, it
is originally in a one-phase state.
Reservoir Fluid: If the reservoir pressure declines along path (1), it will remain in the
single-phase or gaseous state.
Production Fluid: Although the fluid left in the reservoir remains in one phase, the fluid
produced through the wellbore and into surface separators, may enter the two-phase region
owning to temperature decline, as along line (2).
Cricondentherm
Reservoir Reservoir
on
B
ducti
3000
Critical
o
C
of Pr
Point
Reservoir Pressure, psi
t B1
2500 Poin
ble
Path
De
ub
w
B
Po
C1
t
B2
80% D
40% A1
1500
20% 5%
1000 10%
0%
A2
B3
Point B: initially at 180oF and 3300psi. Since this point lies outside the two-phase region, it
is originally in a one-phase state.
Reservoir Fluid: As pressure declines along path (1) because of production, the composition
of the reservoir fluid will remain constant until the dew-point pressure is reached at 2700psi.
Production Fluid: Because the condensed liquid adheres to the walls of the pore walls of the
rock, the gas produced at the surface will have a lower liquid content. This process of
retrograde continues until B2 is reached.
Point C: initially at 75oF and 2900psi. Since this point lies outside the two-phase region, it is
originally in a one-phase state.
Reservoir Fluid: As pressure declines along path (1) because of production, the composition
of the reservoir fluid will remain constant until the bubble-point pressure is reached at
2550psi.
Production Fluid: Below this point, bubbles or a free gas will appear. Eventually, the free
gas evolved begins to flow to the wellbore, and in ever increasing quantities.
Ground Surface
Gas
Reservoir
Question: When the reservoir pressure deceases, what would happen to both the pore volume
and the gas volume?
GAS
PORE
p g : p g 0 pa
The major types of energy available for primary petroleum production are
In the thermodynamic sense, the “energy” in the first four items of the list refers to the
potential energy stored in the compressed fluids. It is equivalent to the potential energy stored
in a compressed spring. Through the first law of thermodynamics, such energy can be
converted to the pressure-volume work (force through a distance) needed for fluids to be
produced. The energy in the fifth term is the potential energy caused by the differences
between different distances (elevations) from the Earth’s center of gravity.
We need to know and understand by what mechanism the oil is being produced. These are the
fundamental drive mechanisms. The main drive mechanisms include water drive, oil and gas
expansion drive, compaction drive.
If the volumetric rate of water intrusion into the reservoir approaches the volumetric rate of
fluid withdrawal from the reservoir, then the reservoir is a complete-water drive reservoir. A
complete water drive reservoir experiences a very little pressure decline; however, some
pressure decline must exist else no potential (pressure) difference between the reservoir and
aquifer would exist and no water influx would occur. If the volumetric rate of water intrusion
into the reservoir is substantial but substantially less than the volumetric rate of fluid
withdrawal from the reservoir, then the reservoir is a partial water drive reservoir. When a
water drive exists, the reservoir pressure will always be sensitive to the producing rate. If the
producing rate is too large relative to the water influx rate, the water drive will lose its
effectiveness and the reservoir pressure will decline.
PRODUCING MECHANISMS
The simplest case is that of a single-phase oil field. Such fields can be found among fields in
their very early stage of development, when the gas or oil is produced by simple natural
decompression. This stage ends rapidly, when the pressure equilibrium between the oil field
and the atmosphere is attained: the natural production of oil or gas stops, though only a small
percentage of the total amount of oil or gas has been produced. This first stage is called
primary recovery.
In order to recover part of the remaining oil, we could pump off at the wells, creating a
pressure drop which would draw the oil to these wells. This would have two draw backs: first,
the pressure around the wells could fall below the bubble pressure of the oil so that the wells
would produce almost only gas, and the heavier components would mainly remain trapped in
the field. Second, diminishing the pressure in the fluid phase could lead the rock to collapse,
resulting in a field with lower permeability and hence more difficult to produce. This is why
we use an alternative method called secondary recovery: we divide the available wells into
two sets: one set of injection wells, and one set of production wells. The injection wells are
used to inject an inexpensive fluid (usually water) into the porous medium, in order to push
the oil toward the production wells. During this process, the pressure inside the field is
maintained above its initial level, so that the two above mentioned draw backs may be more
easily avoided.
For secondary recovery process, two cases are to be considered: Either the pressure can be
maintained always above the bubble pressure of the oil. The flow in the reservoir is then of
the two-phase immiscible type, one phase being water and the other being oil, with no mass
transfer between the phases. Or the pressure may drop, at some points, below the bubble
pressure of the oil: then the oil may split into one liquid phase and one gaseous phase at
The above waterflooding technique makes it possible to recover a certain percent (up to 40%)
of the oil. Three reasons for low recovery:
There exists regions which are never flooded by the water, and hence whose oil is not
going to be produced;
Even in the flooded regions, a significant part of oil (up to 20% to 30%) remains
trapped in the pores by the action of the capillary forces;
When the oil is heavy and viscous, the water is extremely mobile in comparison to the
oil. Then instead of pushing the oil towards the production well, the water finds very
quickly its own way to the production well, getting the oil moving very slowly toward
the production well.
The oil industry developed a set of different techniques known under the generic name of
tertiary recovery techniques or enhanced oil recovery (EOR) techniques or improved oil
recovery (IOR) techniques. One of the main goals of those techniques is to achieve
miscibility of fluids, thus eliminating the residual oil saturation, which was one cause of low
recovery with the water flooding technique. This miscibility is sought using temperature
increase or the introduction of other components such as polymers.
50%
Incremental
Oil
EOR
Cumulative Oil
35%
Recovery
Recovery Processes
Without Stimulation 20%
Primary Recovery
TIME
Figure 1.7 Oil Recovery Summary
Gas
Oil
Water
Rock
WATER
GAS
OIL
ROCK
At most, there are four components, gas, oil, water and rock;
At most, there are four phases: gas, oil, water, and solid;
The gas component is defined by the composition of the reservoir gas at the reservoir
conditions;
The oil component is defined by the composition of the reservoir oil at the reservoir
conditions;
Thermodynamic equilibrium exists.
In the following chapters, we will define the fundamental properties of each component and
Tuesday, February 25, 2014
interactions between components. These properties are absolutely essential for the
determination of the volumetric behavior of oil and gas reservoirs as a function of pressure.
We will start with what is porosity and then how does the porosity change as a function of
pore pressure.
Definition: Although a reservoir rock looks a solid to the naked eye, a microscopic
examination reveals the existence of voids in the rock. These pores are the ones where
petroleum reservoir fluids are present. This particular storage capacity is called porosity. The
more porous a reservoir rock material is, the greater the amount of voids it contains, hence
greater the capacity to store petroleum reservoir fluids. From a reservoir engineering
perspective, porosity is probably one of the most important reservoir rock properties.
Porosity, φ is a volumetric fraction defined as the ratio of the pore volume, V pore in a
reservoir rock to the total volume (bulk volume), Vbulk :
V pore
(2.1)
Vbulk
The porosity of a rock is a measure of the storage capacity (pore volume) that is capable of
holding fluids. It may by occupied by a single-phase fluid or mixtures. As the sediments were
deposited and the rocks were being formed during past geological times, some void spaces
that developed became isolated from the other void spaces by excessive cementation. Thus,
many of the void spaces are interconnected while some of the pore spaces are completely
isolated. This leads to two distinct types of porosity, namely:
Absolute porosity
Effective porosity.
Vtotalpore
a
Vbulk
Vint erconnectedpore
e
Vbulk
The effective porosity is the value that is used in all reservoir engineering calculations
because it represents the interconnected pore space that contains the recoverable hydrocarbon
fluids.
V pore pore
Vrock rock
Vbulk bulk
PORE
One important application of the porosity is its use in determining the original hydrocarbon
volume in place.
If the pore is occupied by oil only, the calculated PV is the original oil in place (OOIP);
If the pore is occupied by gas only, the calculated PV is the original gas in place (OGIP).
If the pore is occupied by oil and gas, what would happen?
If the pore is occupied by oil, gas and water, what would happen?
RESERVOIR V pore
Vrock
Vbulk
PV Vbulk
pore
rock
bulk
Question 4– Assuming the pores are occupied by water only, calculate the original water in
place (OWIP) and discuss what factors will change OWIP?
Question 5– If pores are occupied by a mixture of oil, gas and water, discuss how to calculate
the original water, oil, or gas in place (OWIP, OOIP and OGIP)?
b
Fracture
Coal Matrix
s
Fig.2.1 Fractured Coal
Fractured coals are made up of two porosity systems; one matrix formed by void spaces
between the grains of the coal, and a second formed by void spaces of fractures as shown in
Fig.2.1. In a fractured coal reservoir the total porosity ( t ) is the result of the simple addition
of the matrix and fracture porosities:
t matrix fracture
Where
b
Fracture
Coal Matrix
When a gas is injected into coal, the gas will adsorb onto the coal and the coal swells. For a
single gas system the sorption-induced volumetric strain s is fit to a Langmuir type curve
as verified in experiments. A Langmuir-type equation is used to calculate this volumetric
strain, defined as
p
s L (2.2)
PL p
where the sorption-induced strain constant, L , representing the volumetric strain at infinite
pore pressure with the Langmuir pressure constant, PL , representing the pore pressure at
which the measured sorption-induced strain is equal to 0.5 L .
When a coal sample is completely constrained as shown in Figure 5(a), a force and resulting
stress develop within the coal. A simple way to determine the amount of stress is to let the
coal expand freely due to the gas sorption as shown in Figure 5(b), then push it back to its
original positions, as shown in Figure 5(c). In this case, the total volumetric expansion is
equal to zero. Under this condition, the volumetric expansion in the coal solid must be equal
to the volumetric compression in the coal pores. Therefore, the coal porosity decreases as the
coal solid swells, and can be defined as
1 s
1
(2.3)
0 0
p
p
(a) Constant Volume Condition (b) Free Expansion (c) Push-Back Force
Assuming:
pL 3MPa
L 0.02
0 0.10
Equation (2.3) becomes
0.5 p
1
0 p3
0.9
0.8
0.7
0.6
Coal Porosity
0.5
0.4
0.3
0.2
0.1
0
0 2 4 6 8 10 12
This plot shows that coal porosity is not a constant. More information can be found in Liu et
al. 2011.
While porosity represents the maximum capacity of a reservoir rock to store fluids, how can
we know how much of this available capacity, pore volume, or pore space distributed or
partitioned among the typical reservoir fluid phases: gas, oil and water? In order to achieve
this, we need to introduce the concept of fluid saturation.
Saturation is defined as that fraction, or percent, of the pore volume occupied by a particular
fluid (oil, gas, or water). This property is expressed mathematically by the following
relationship:
Assuming S o , S g , S w represent the oil saturation, gas saturation and water saturation,
respectively, and Vo ,Vg ,Vw for the oil volume, gas volume, and water volume, respectively,
and V p the total pore volume, applying the above mathematical concept to each to each
reservoir fluid gives
Vo
So
Vp
Vg
Sg
Vp
Vw
Sw
Vp
Thus, all saturation values are based on pore volume and not on the gross reservoir volume.
The saturation of each individual phase ranges between zero to 100 percent. By definition, the
sum of the saturations is 100%, therefore
So S g S w 1
Equation above is probably the simplest, yet the most fundamental equation in reservoir
engineering, and is used everywhere in reservoir engineering calculations. Moreover, many
important reservoir rock properties, such as capillary pressure and relative permeability, are
actually related or linked with individual fluid-phase saturations.
The fluids in most reservoirs are believed to have reached a state of equilibrium and, therefore,
will have become separated according to their density, i.e., oil overlain by gas and underlain
by water. In addition to the bottom (or edge) water, there will be connate water distributed
throughout the oil and gas zones. The water in these zones will have been reduced to some
irreducible minimum. The forces retaining the water in the oil and gas zones are referred to as
capillary forces because they are important only in pore spaces of capillary size. Connate
(interstitial) water saturation Swc is important primarily because it reduces the amount of
space available between oil and gas. It is generally not uniformly distributed throughout the
reservoir but varies with permeability, lithology, and height above the free water table.
Another particular phase saturation of interest is called the critical saturation and it is
associated with each reservoir fluid. The definition and the significance of the critical
saturation for each phase is described below.
Residual oil saturation, S or : During the displacing process of the crude oil system from the
porous media by water or gas injection (or encroachment) there will be some remaining oil
left that is quantitatively characterized by a saturation value that is larger than the critical oil
saturation. This saturation value is called the residual oil saturation, S or . The term residual
saturation is usually associated with the nonwetting phase when it is being displaced by a
wetting phase.
Movable oil saturation, S om : Movable oil saturation is another saturation of interest and is
defined as the fraction of pore volume occupied by movable oil as expressed by the following
equation:
Som 1 S wc Soc
Where S wc and S oc are connate water saturation and critical oil saturation, respectively.
Critical gas saturation, S gc : As the reservoir pressure declines below the bubble-point
pressure, gas evolves from the oil phase and consequently the saturation of the gas increases
as the reservoir pressure declines. The gas phase remains immobile until its saturation exceeds
a certain saturation, called critical gas saturation, above which gas begins to move.
Critical water saturation, S gc : The critical water saturation, connate water saturation, and
irreducible water saturation are extensively used interchangeably to define the maximum
water saturation at which the water phase will remain immobile.
S w S wr
Vpore 1 S w*
1 S wr Sor
Swr Sor Swm Som 1 Som
1 S w Sor
So*
1 S wr Sor
Irreducible water saturation
Sor S w* So* 1
S wm S w S wm S wr
So Som Sor
S wi So S w 1
In above sections, we address the porosity and saturation both of which are used to define the
storage capacity of reservoir rock for reservoir fluids. However, merely having a large enough
porosity of reservoir rock is not sufficient because the petroleum reservoir fluids contained in
the pore spaces of reservoir rock have to flow, so that they can be produced or brought to the
surface from the reservoir. This particular property of a reservoir rock is called permeability.
Permeability is one of the most influential parameters in determining the production
capabilities of a producing formation.
Unlike porosity, permeability is a flow property (dynamic) and therefore can be characterized
only by flow experiments in a reservoir rock. Permeability is a property of the porous medium
that measures the capacity and ability of the formation to transmit fluids. The rock
permeability, k, is a very important rock property because it controls the directional
movement and the flow rate of the reservoir fluids in the formation. Absolute permeability is
the rock permeability when a reservoir rock is 100% saturated with a given fluid. It should be
noted that the absolute permeability is a property of the rock alone and not the fluid that flows
through it, provided no chemical reaction takes place between the rock and the flowing fluid.
All the equations used to describe fluid flow in reservoirs are based on Darcy’s law. Darcy
(1856), investigated the flow of water through sand filters. He observed the following
relationship between velocity and pressure gradient as
Q k p
ux (2.4)
A x
Where
Basic assumptions:
The negative sign in the above equations indicates that pressure decreases in the direction of
flow. The sign convention is therefore that distance is measured positive in the direction of
flow.
L3
Q
T
A L2
p
force ML
2 2
M
M Mass
area T L LT 2
force M
time
area LT
3
L M LT 1
kL2
T LT 2 M L
k L2
1 Darcy 1 10-12 m2
p1 A 1cm 2
A p2
L 1cm
p 105 pa
Q m3 / s
L Q 1cm 3
0.001 pa s
Figure 2.3 Illustration of 1-D Flow
Q k p
u
A L
QL
k
Ap
1D 10-8 cm 2 10 12 m 2
1mD 10 15 m 2
1D 10 18 m 2
1nD 10 21 m 2
FRACTURE PERMEABILITY
Fracture porosity
Block total volume:
V S3
Volume of each fracture:
V f1 b S S bS 2
Total volume of three
fractures:
Vf3bS 2 3b
fracture 3 V f 3V f1 3bS 2
V S S
For a fixed spacing, we
12Sk f 1/ 3
3
b 3 can plot the relation
kf f between porosity and
12S S
permeability!
COAL/SHALE PERMEABILITY
Where k is the coal permeability, k0 is the initial permeability, is the initial porosity,
3
k L p
1 (2.6)
k0 0 pL p
0.7
0.6 0 0.10
0.5 pL 3MPa
0.4
0.3
0.2
0.1
0
0 2 4 6 8 10 12
Pore Pressure
QUESTION: Assuming 1-D flow, determine the pressure and Darcy velocity distributions.
Basic assumptions:
It is assumed that the porous medium is saturated with a single fluid.
The flowing fluid is incompressible.
The flow takes place under the viscous regime (i.e., the rate of flow is sufficiently
low so that it is directly proportional to the pressure differential or the hydraulic
gradient).
The flowing fluid does not react with the porous medium.
p1
?
p2
Flow-Out
Flow-In
k g pg
ug (2.7a)
g x
ko po
uo (2.7b)
o x
k p
uw w w (2.7c)
w x
Where ug , uo , uw are gas velocity, oil velocity, and water velocity, respectively; k g , ko , kw are
effective gas permeability, effective oil permeability, and effective water permeability,
respectively; pg , po , pw are gas pressure, oil pressure, and water pressure, respectively;
g , o , w are dynamic gas viscosity, dynamic oil viscosity, and dynamic water viscosity,
respectively.
Darcy’s law was considered to apply when the porous medium was 100% saturated with a
homogeneous single-phase fluid. However, petroleum reservoirs having much simple
single-phase fluid systems seldom exist; usually reservoir rock systems are saturated with at
least two or more fluids such as water, oil and gas. These multiphase fluid systems play a very
important role in the reservoir flow processes when petroleum reservoirs are produced by
primary recovery mechanism or immiscible displacement methods involving the injection of
water or gas. It is under these circumstances that more than one fluid phase is flowing or is
mobile through a porous medium; thus the flow of one fluid phase interferes with the other.
This interference is a competition for the flow paths. As a result of the interference, the
following relation holds
k g ko kw k
or
kg ko k w
1
k k k
For example, if the absolute permeability k of a rock is 200md and the effective
permeability ko of the rock at an oil saturation of 80% is 60md, the relative permeability
60
kro 0.3 . Since the effective permeabilities may range from zero to k, the relative
200
permeabilities may have any value between zero and one, or:
0 k rw , k rg , k ro 1.0
It should be pointed out that when three phases are present the sum of the relative
permeabilities k ro k rg k rw is both variable and always less than or equal to unity.
Initially, it might appear that the sum of the phase permeabilities equals the total or absolute
permeability, which would mean that the relative permeabilities should sum to unity. However,
this is not true. When two or more phases are present, capillary forces exist that reduce the
flow rate of each individual phase in a non-linear fashion. This means that the sum of the
phase permeabilities is always less than the total or absolute permeability and the sum of
relative permeabilities is always less than one. Indeed, at irreducible water saturation, the
relative permeability to water becomes zero while the relative permeability to oil or gas is less
than one because the immobile water is occupying some of the flow volume. Similarly, at
residual oil saturation, the relative permeability to water or gas is less than one.
The steady state technique involves injecting two phases at a certain volumetric ratios until
stabilization of both the pressure drop across the core and the effluent volumetric ratios.
Darcy law can be applied to each phase:
qo k dpo ko po
uo o
A o dx o L
q k dpw kwpw
uw w w
A w dx w L
po pw p
Assuming the capillary effect is small, . The phase effective permeability
L L L
can be calculated as
qo o L
ko
Ap
q L
kw w w
Ap
p
WATER WATER
WATER
OIL
OIL OIL
Practical Question: A set of steady state oil/water relative flow rates at different saturations
is given below. Core dimensions are a length of 12cm and a 5cm 2 flow area (i.e. the area
perpendicular to the direction of flow). Oil viscosity is 5cp and water viscosity is 1.2cp.
2.5 COMPRESSIBILITY
The pressure difference between overburden and internal pore pressure is referred to as the
effective overburden pressure. During pressure depletion operations, the internal pore pressure
decreases and, therefore, the effective overburden pressure increases. This increase causes the
following effects:
These two volume changes tend to reduce the pore space and, therefore, the porosity of the
rock. Often these data exhibit relationships with both porosity and the effective overburden
pressure.
Water Compressibility
Gas Compressibility
Oil Compressibility
Grain Compressibility
Bulk Compressibility
1 dVrock
cr
Vrock dp
1 dVB
cB
VB d eff
Where VB is the bulk volume and eff is effective pressure.
Equation above can be expressed in terms of the porosity by noting that porosity increases
with the increase in the pore pressure; or:
1 d
cp
dp
For most petroleum reservoirs, the rock and bulk compressibility are considered small in
comparison with the pore compressibility c p . The formation compressibility c f is the term
commonly used to describe the total compressibility of the formation and is set equal to c p ,
i.e.:
1 dV p 1 d
c f cp
V p dp dp
Typical values for the formation compressibility range from 3 10 psi to 25 10 psi .
6 1 6 1
It should be pointed out that the total reservoir compressibility ct is extensively used in the
transient flow equation and the material balance equation as defined by the following
expression:
ct S o co S w cw S g cg c f
Decreasing Permeability
a r t i c l e i n f o a b s t r a c t
Article history: Laboratory observations have shown that coal permeability under the influence of gas adsorption can change
Received 13 May 2011 instantaneously from reduction to enhancement. It is commonly believed that this instantaneous switching of
Received in revised form 28 July 2011 permeability is due to the fact that the matrix swelling ultimately ceases at higher pressures and the influence of
Accepted 28 July 2011
effective stresses take over. In this study, our previously-developed poroelastic model is used to uncover the true
Available online 4 August 2011
reason why coal permeability switches from reduction to enhancement. This goal is achieved through explicit
Keywords:
simulations of the dynamic interactions between coal matrix swelling/shrinking and fracture aperture alteration,
Coal permeability and translations of these interactions to perrmeability evolution under unconstrained swellings. Our results of
Coal swelling this study have revealed the transition of coal matrix swelling from local swelling to macro-swelling as a novel
Local swelling mechanism for this switching. Our specific findings include: (1) at the initial stage of CO2 injection, matrix
Macro swelling swelling is localized within the vicinity of the fracture compartment. As the injection continues, the swelling zone
is extending further into the matrix and becomes macro-swelling. Matrix properties control the swelling
transition from local swelling to macro swelling; (2) matrix swelling processes control the evolution of coal
permeability. When the swelling is localized, coal permeability is controlled by the internal fracture boundary
condition and behaves volumetrically; when the swelling becomes macro-swelling, coal permeability is
controlled by the external boundary condition and behaves non-volumetrically; and (3) matrix properties
control the switch from local swelling to macro swelling and the associated switch in permeability behavior from
reduction to recovery. Based on these findings, a permeability switching model has been proposed to represent
the evolution of coal permeability under variable stress conditions. This model is verified against our
experimental data. It is found that the model predictions are consistent with typical laboratory and in-situ
observations available in lietratures.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction such as argon and nitrogen (N2) (Chen et al., 2011; Siriwardane et al.,
2009; Somerton et al., 1975). The permeabilities may decrease by as
Gas transport in coal seams is significantly different from that in other much as five orders of magnitude if confining pressures vary from 0.1 to
types of rocks because of the phenomena of gas adsorption/desorption 70 MPa (Durucan and Edwards, 1986; Somerton et al., 1975). Under
and coal swelling/shrinkage. The relative roles of stress level, gas constant total stress, adsorbing gas permeability decreases with the
pressure, gas composition, fracture geometry of coal, and water content increase of pore pressure due to coal swelling (Mazumder and Wolf,
are intimately connected to the processes of gas adsorption/desorption, 2008; Pan et al., 2010; Robertson, 2005; Wang et al., 2010, 2011), and
diffusion, transport, and coal swelling/ shrinkage. As a direct consequence increases with the decrease of pore pressure due to matrix shrinkage
of these complex coal–gas interactions, coal permeability changes in both (Cui and Bustin, 2005; Harpalani and Chen, 1997; Harpalani and
space and time. Schraufnagel, 1990; Seidle and Huitt, 1995). Rebound pressure, which
Significant experimental efforts have been made to investigate gas corresponds to the minimum permeability, was observed for CO2
permeability and its evolution in coal. Laboratory measured coal injection at 1.7 MPa (Pini et al., 2009), and at 7 MPa (Palmer and
permeabilities to adsorbing gasses such as CH4 and CO2 are lower Mansoori, 1996; Shi and Durucan, 2004a, 2004b), recpectively.
than those permeabilities to non-adsorbing or lightly adsorbing gasses Permeability of adsorbing gas in coal is found to be a function of gas
exposure time (Siriwardane et al., 2009). The pemeability is also
influenced by water saturation (Han et al., 2010).
⁎ Corresponding author at: School of Mechanical and Chemical Engineering, The
Based on field and laboratory experimental results, various coal
University of Western Australia, WA 6009, Australia. permeability models have been proposed based on the fundamental
E-mail address: jishan@cyllene.uwa.edu.au (J. Liu). principles of poroelasticity. These models could have different forms
0166-5162/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.coal.2011.07.008
32 J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40
after imposing specific conditions. When total stress has no change, rather than matchstick geometry under biaxial or hydrostatic confining
both coal porosity and permeability are independent of the total pressures. It was also designed to handle changes in permeability caused
stress. Under this condition, they can be expressed as a function of gas by adsorption and desorption of gasses from the matrix blocks. Contrary
pressure and temperature only. If coal sample is in uniaxial strain and to previous models developed for field conditions, their model mainly
under no-changing overburden load, they can also be expressed as a deals with variable stress conditions which are commonly used during
function of gas pressure and temperature only. When the impact of measurement of permeability in the laboratory. Similarly, Connell et al.
coal fractures and gas compositions is considered, coal porosity and (2010) presented two new analytical permeability models representing
permeability models can be linked to fracture parameters and gas for standard triaxial strain and stress conditions. They presented a novel
concentrations. approach to distinguish the sorption strains of the coal matrix, the pores
Permeability models were proposed for some idealized coal (or cleats) and the bulk coal. Based on the theory of poroelasticity, Zhang
structures as well as specific boundary conditions. For example, Reiss et al. (2008) developed a general porosity and permeability model. It
(1980) developed the equations for permeability and porosity based on was shown that current commonly used permeability models could be
a collection of matchsticks and the collections of slabs and cubes. Gray treated as specific examples.
(1987) considered that the changes in the cleat permeability were When laboratory experimental results are interpreted, a matchstick
primarily controlled by the prevailing effective horizontal stresses that or cubic coal model is typically assumed with the matrix blocks
act across the cleats. Under the assumption of uniaxial strain, the completely separated from each other in a stacked structure. In such a
influence of matrix shrinkage on the change of coal permeability was model, the permeability should not change under conditions of constant
first incorporated into a permeability model. Gilman and Beckie (2000) confining (total) stress due to through-going fracutres (Liu et al., 2011;
assumed that an individual fracture reacts as an elastic body upon a Liu and Rutqvist, 2010). This interpretation is not consistent with
change in the normal stress component and proposed a simplified laboratory observations (Harpalani and Chen, 1997; Pan et al., 2010; Pini
mathematical model of methane movement in a coal seam. Their model et al., 2009), which show dramatic reduction in permeability with the
takes into account the following features: a relatively regular cleat injection of an adsorbing gas. Liu and Rutqvist (2010) assumed that coal
system, adsorptive methane storage, an extremely slow release matrix blocks connected each other by the coal-matrix bridges and
mechanism of methane from the coal matrix into cleats, and a significant developed a new coal-permeability model to explicitly consider
change of permeability due to desorption. Seidle and Huitt (1995) fracture–matrix interaction during coal-deformation processes based
calculated the permeability increase due to matrix shrinkage alone by on the concept of internal swelling stress. An alternate reasoning has
using a linear relationship between matrix shrinkage and pore pressure. been applied by Liu et al. (2010a, 2011) on this issue. They regarded that
Their model ignored the impact of coal compressibility and thus was the above phenomena may be due to the ignorgance of the internal
limited to specific conditions in which sorption-induced strain (matrix actions between coal fractures and matrix. Izadi et al. (2011) proposed a
swelling or shrinkage) dwarfs pressure-induced, elastic changes in cleat mechanistic representation of coal as a collection of unconnected cracks
permeability (Robertson and Christiansen, 2008). Based on the in an elastic swelling medium. The cracks are isolated from each other
matchstick geometry model and the relationship between permeability but swelling within a homogeneous but cracked continuum results in a
and porosity proposed by Seidle and Huitt (1995), Shi and Durucan reduction in crack aperture with swelling, and a concomitant reduction
(2004a, 2004b) presented a pore pressure-dependent cleat permeability in permeability. Ma et al. (2011) developed a model based on the
model for gas-desorbing, linear elastic coalbeds under uniaxial strain volumetric balance between the bulk coal, and solid grains and pores,
conditions. In this model, the changes in the cleat permeability of and using the constant volume theory (Massarotto et al., 2009). It
coalbeds were assumed to be controlled by the prevailing effective incorporates primarily the changes in grain and cleat volumes and is,
horizontal stresses normal to the cleats. A widely used theoretical therefore, different from the other models that lay heavy emphasis on
permeability model was derived by Palmer and Mansoori (1996) (called the pore volume/cleat compressibility.
P&M model later). Their permeability in coals is as a function of effective As reviewed above, a large variety of coal permeability models have
stress and matrix shrinkage. The P&M model was improved and been proposed. These span conditions representing constant stress
summarized by Palmer et al. (2007). Similarly, Pekot and Reeves through variable stress conditions. All of these coal permeability models
(2002) have developed another permeability model which does not were derived based on the theory of poroelasticity or equivalent
have a geomechanics framework, but instead extracts matrix strain continuum approach. Current experimental studies have been focusing
changes from a Langmuir curve of strain versus reservoir pressure. It was on the overall behaviors of coal samples and only a few of them focus on
assumed that strain was proportional to the gas concentration curve and the impact of fractures (Siriwardane et al., 2009).
matrix shrinkage was proportional to the adsorbed gas concentration Gas transport in coal seams is commonly represented as a dual
change multiplied by shrinkage compressibility. This model has been porosity system accommodating two serial transport mechanisms:
compared to the P&M model, and concluded that the two models are diffusion through the coal matrix then laminar flow through the cleat
essentially equivalent in saturated coals where the strain versus pressure system (Bai and Elsworth, 2000; Elsworth and Bai, 1992). The
function is proportional to the Langmuir isotherm (Palmer et al., 2007). permeability is primarily determined by the cleat aperture (Wu et
Following the above work, Cui and Bustin (2005) quantitatively al., 2010a, b; Zhang et al., 2008). The change in cleat aperture is a
investigated the effects of reservoir pressure and sorption-induced function of effective stress through poroelasticity, but coal swelling
volumetric strain on coal-seam permeability with constraints from the and shrinkage under a confining stress may also change the cleat
adsorption isotherm and associated volumetric strain and derived a aperture (Izadi et al., 2011; Wu et al., 2010a, b). Thus, the net change
stress-dependent permeability model. Focusing on the full separation of in coal permeability is a function of both the poroelastic response and
strain between matrix and cleat/fracture, Gu and Chalaturnyk (2005a,b, the coal swelling or shrinkage.
2006) proposed a permeability model using an equivalent continuum Over the past few years, a series of advanced modeling tools has been
approach based on geomechanics. Pan and Connell (2007) developed a developed to quantify the complex coal–gas interactions (Chen et al.,
theoretical model for sorption-induced strain and applied to single- 2009, Chen et al., 2010a, 2010b; Connell, 2009; Connell and Detournay,
component adsorption/strain experimental data. Clarkson (2008) 2009; Gu and Chalaturnyk, 2005a,b, 2006; Liu et al., 2010a,b; Wu et al.,
recently expanded this theoretical model to calculate the sorption– 2010a,b, 2011; Zhang et al., 2008). These works have provided a
strain component of the P&M model. coupling approach to represent important non-linear responses of coal
Robertson and Christiansen (2008) derived a coal permeability matrix to the effective stress effects. In this study, we used our poroelastic
model for a fractured, sorptive-elastic medium, such as coal, under models to uncover the true reason why coal permeability switches from
variable stress conditions. This model was derived for cubic geometry reduction to enhancement. This goal was achieved through explicit
J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40 33
simulations of the dynamic interactions between coal matrix swelling/ extensive, roughly planar fractures that run parallel to one another. Butt
shrinkage and fracture aperture alteration, and translations of these cleats are orthogonal to face cleats and often terminate at the face cleats.
interactions to perrmeability evolution under unconstrained boundary Butt cleats are also roughly planar but not as well-developed or as
conditions. continuous as face cleats. The cleat system provides an essential and
effective flow path for gas. Much of the measured bulk or “seam”
2. General permeability model permeability is due to the cleat system, although the presence of larger
scale discontinuities such as fractures, joints, and faults can also make a
It is clear that there is a relationship between porosity, permeability significant contribution. The coal matrix is isolated by the fracture
and the grain-size distribution in porous media. Chilingar (1964) network and is the principal medium for storage of the gas, principally in
defined this relationship as adsorbed form and with low permeability in comparison to the
bounding cleats (Gray, 1987). The surface area of the coal on which
d2e ϕ3 the methane is adsorbed is very large (20 to 200 m2/g) (Patching, 1970)
k= ð1Þ
72ð1−ϕÞ2 and gas is stored at near-liquid densities.The remaining gas is stored in
the natural fractures, or cleats, either as free gas or dissolved in water.
where k is the permeability, ϕ is porosity and de is the effective In this study, we consider the interactions of the fractured coal
diameter of grains. mass where cleats do not create a full separation between adjacent
Based on this equation, one obtains matrix blocks but where solid rock bridges are present, as illustrated
3 2 in Fig. 1. We accommodate the role of sorption-induced swelling
k ϕ 1−ϕ0 strains both over contact bridges that hold cleat faces apart and over
= · ð2Þ
k0 ϕ0 1−ϕ the non-contacting span between these bridges.
Fig. 1 illustrates the effective stress transfer between coal matrix
When the porosity is much smaller than 1.0 (normally less than and fracture induced by the CO2 injection. Prior to the CO2 injection,
10%), the second factor of the right-hand side asymptotes to unity. the gas pressure is equal in both fracture and matrix. This state is
This yields the cubic relationship between permeability and porosity defined as the initial equilibrium state. The initial pore pressure is
for the coal matrix assumed to be zero, pm = pf = 0. Coal permeability at this state is
3 defined as the initial equilibrium permeability. At this equilibrium
k ϕ state, no swelling takes place anywhere. We define this as the starting
= ð3Þ
k0 ϕ0 point for the evolution of permeability
8
Further, the coal porosity ratio evolves with the effective strain >
< k =1
increment as k0 ð7Þ
>
: σ = σ = 0·
me fe
ϕ α
=1+ Δε : ð4Þ
ϕ0 ϕ0 e
When CO2 is injected, the gas occupies the fracture and the gas
Therefore, the permeability ratio is evolved by pressure in the fracture reaches the injection pressure almost
instantly. At this stage, the maximum inbalance beween fracture
3 pressure and matrix pressure is achieved. This pressure inbalance is
k α
= 1+ Δεe ð5Þ defined as
k0 ϕ0
where the effective strain increment is calculated by σme = 0
: ð8Þ
σfe = −αpf
Δp
Δεe = Δεv + −Δεs ð6aÞ
Ks This inbalance diminishes as the gas diffuses into the coal matrix. The
pore pressure in the matrix increases and in turn reduces the effective
or stress in matrix. As an outcome of the diffusion, coal matrix swells due to
both the matrix pore pressure increase and the gas sorption. Initially,
Δσ−Δp
Δεe = − ð6bÞ this matrix swelling is confined in the vicinity of the fracture voids. This
K
localized swelling reduces the fracture aperture. This reduction in
where Δεe is defined as the total effective volumetric strain, Δεv is aperture reduces the fracture permeability. As the gas diffusion
total volumetric strain increment, Δp/Ks is coal compressive strain progresses, the swelling zone extends further into the coal matrix. As
change, Δεs is gas sorption-induced volumetric strain increment, and the swelling zone front moves away from the fracture void, the influence
Ks represents the bulk modulus of coal grains. of matrix swelling on the fracture aperture weekens. As a result of the
Eqs. (4) and (5) are the formula for coal porosity and permeability. widening of the swelling zone, the fracture permeability recovers while
They are derived based on the fundamental principles of poroelasticity the local swelling has become the macro-swelling. When the inbalace
and can be applied to the evolution of coal porosity and permeability between fracture pressure and matrix pressure vanishes completely, the
under variable boundary conditions. As shown in Eqs. (4) and (5), coal final equilibrium state is achieved. At the final equilibrium state, the
porosity and permeability can be expressed as a function of either matrix pressure is equal to the fracture pressure, i.e., pm = pf which was
effective strain (see Eq. (6a)) or effective stress (see Eq. (6b)). defined as the ending point for the evolution of coal permeability.
3. Conceptual model σme = −αpf
ð9Þ
σfe = −αpf
Coal is a typical dual porosity/permeability system (Warren and
Root, 1963) containing porous matrix surrounded by fractures. These Coal permeability at this state is defined as the final equilibrium
natural fractures form a closely-spaced network called cleats. The main permeability in which a uniform matrix swelling is achieved for a
set of fractures, termed face cleats, is comprised of well-developed, homogeneous coal sample.
34 J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40
Fracture pf pf pf
0 pm p f
Matrix pm x, y, t
pf pf pf pf
me t pm fe t pf
Fig. 1. Illustration of effective stress transfer between matrix and fracture and its impact on the evolution of coal permeability.
Fig. 2 illustrates the difference between local swelling and macro Based on the above analysis, the final equilibrium permeability is
swelling. When the coal matrix swelling is localized in the vicinity of a always higher than the initial equilibrium permeability if a uniform
fracture compartment, the effect of swelling acts competitively over swelling state is achieved within coal sample. However, laboratory
these two components: increasing porosity and permeability due to measurements indicate that the coal equilibrium permeability is
swelling of the bridging contacts but reducing porosity and permeability much lower than the initial equilibrium permeability at lower pore
due to the swelling of the intervening free-faces. When the coal matrix pressures. It may recover but rarely exceeds the initial equilibrium
swelling is de-localized from the vicinity of a fracture compartment to permeability even at higher pore pressures. This distinct discrepancy
the external boundary, the coal bridge swelling increases the fracture points that a uniform matrix swelling state has rarely been achieved in
aperture while the coal matrix swelling changes the spacing only. At the real coal samples. This implies that the transient characteristics of coal
uniform swelling state, as shown in Eq. (5), the coal permeability is permeability are likely controlled by the localized swelling near the
determined by the coal bridge swelling only vicinity of coal fracture voids rather than the outside boundaries.
The pressure transient method (Brace et al., 1968; Hsieh et al., 1981)
!3 !3 is normally used to conduct gas flow experiments in low permeability
k pf pf pf
ðt = ∞Þ = 1 + εL − ≈ 1 + εL ð10Þ coal samples. In a typical experiment, the sample is placed into a triaxial
k0 pL + pf Ks pL + pf
core holder and both confining stress and axial stresses are applied at a
rate of loading to establish initial conditions and are then kept constant.
where εL is the Langmuir strain constant, pL is the Langmuir pressure, The sample-reservoir system is then vacuum desaturated to evacuate air
k is the coal permeability, and Ks is the coal grain bulk modulus. As from the system. A pressure increment is then applied to the upstream
shown in Eq. (10), coal permeability at the uniform matrix swelling gas reservoir and discharged through the sample to the downstream gas
state is not related to any elastic properties of coal at all. reservoir. The time taken for the discharging upstream reservoir and the
Fig. 2. Illustration of the difference between local swelling and macro swelling.
J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40 35
4. Quantitative model
me fe 0 Initial Equilibrium State
0
time In order to recover important non-linear responses due to effective
stress effects, mechanical influence must be rigorously coupled with
Modeling Injection Profile the gas transport system. This can be achieved through a full coupling
Change in Effective Stress
Pressure in Fracture
Experimental Injection Profile
fe large system of non-linear coupled partial differential equations)
incorporating all of the relevant physics is solved simultaneously. Full
coupling is often the preferred method for simulating multiple types
me
of physics simultaneously since it should theoretically produce the
most realistic results.
In the following section, a simulation model was constructed to
me fe pf Final Equilirium State investigate the permeability change under stress controlled conditions.
The selected geometry is for a regular array of interacting cracks as
shown in Fig. 4(b). The influence of effective stress and sorption-
Fig. 3. Qualitative relations between complex processes triggered by CO2 injection.
induced swelling response for an rectangular crack are examined. A
single component part is removed from the array where the appropriate
boundary conditions are for uniform displacement along the bound-
aries. This represents the symmetry of the displacement boundary
recharging downstream reservoir to reach a new equilibrium pressure is condition mid-way between flaws as shown in Fig. 4(a). The simulation
measured. This pattern is repeated for multiple cycles using the desired model geometry is 1.0 cm by 1.0 cm with a fracture located at the center.
gas permeant. After multiple repeats of this procedure a relationship The fracure is 5 mm in length and 0.5 mm in width. Roller boundary was
between permeability and pore pressure and effective stress is obtained. applied on the left side and bottom side, and the other two sides are
It is obvious that current laboratory measurements are only for the stress control. No flow boundary was applied on all the outer boundaries
permeability at the final equilibrium state. For homogeneous coal and constant injection pressure was used along the fracrture bound-
samples, a net increase in permeability should be observed theoritically aries. The change in aperture due to the combined influence of soprtion-
in the tests, as shown in Fig. 3. induced swelling and effective stress was examined.
Fig. 3. illustrates the evolution of coal permeability from the initial Input parameters for simulations were listed in Table 1. For the gas
equilibrium state to the final equilibrium state and the associated transport model, a time-dependent injection pressure was specified at
effective stress transfer processes. The upper graph shows how the the boundaries of fracture:
coal permeability ratio evolves from the initial state (k/k0 = 1) to the
final state (k/k0 = [1 + (εLpf)/(pL + pf)] 3). The ultimate permeability is
8 0
calculated based on the assumption of a uniform swelling within the > t−tp 1
>
> −
coal matrix. However, this condition may never be achieved for real >
<P B C
init + Pc @1−e td A t≥tp
coal samples. A difference of the ultimate permeability between an Pin ðt Þ = ð11Þ
>
>
ideal homogeneous coal and a real heterogeneous coal is expected as >
>
:
shown in Fig. 3. Pinit t < tp
No flow
4
Boundary 2
No flow
No flow
0.005cm
0.01m
Boundary 1
No flow
(a) Single Fracture Compartment Model (b) Multiple Fracture Compartment Model
Fig. 4. Numerical model of a single fracture compartment to represent the response of multiple fracture compartment system under the unconstrained swelling condition. For
deformation boundary: 1 and 2 are roller boundary; 3 and 4 are uniform tractions.
36 J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40
Table 1 1.08
Basic parameters in computations.
1.07
Parameter Value
It is difficult to keep the constant but we can keep the same Time
dt
Pc 0.8
= C: ð13Þ
td
0.7
2 MPa
4 MPa
If Pc = 8MPa and td = 1000s, then C = 8 × 10 − 3MPa/s. 0.6 6 MPa
8 MPa
Table 2 shows the characteristic time for different equilibrium 10 MPa
0 1 2 3 4 5 6 7 8 9 10
Table 2
Fracture pressure (MPa)
Fracture pressure and characteristic time.
Pc(MPa) 2 4 6 8 10
(b) Coal permeability ratio with pore pressure
td(s) 250 500 750 1000 1250
Fig. 6. Evolutions of coal permeability for the case of k = 10− 20 m2.
J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40 37
permeability recovery and the transitional period is shorter for higher when local swelling controls; Δεv ≠ 0 when macro swelling controls.
injection pressure . Under this approximation, Eq. (5) can be extended into
The evolution of coal permeablity against the injection pressure is 8
>
> α p 3
shown in Fig. 6(b). Prior to the CO2 injection, the gas pressure in the > 1+
< −Δεs + p≤pc
k φ0 Ks
fracture is equal to that in the matrix, pm = pf = Pinit, the intial reservoir k0
=
> 3 :
>
> α εp p α ε ðp−pc Þ p−pc 3
pressure. When CO2 is injected, the gas occupies the fracture and the : 1+ − L c + c 1+ Δεv − L + p > pc
φ0 pc + pL Ks φ0 p−pc + pL Ks
gas pressure in the fracture reaches the injection pressure almost
ð14Þ
instantly, i.e., pf = pin, where pin is the injection pressure. At this stage,
there exists an inbalance beween fracture pressure and matrix
For the stress controlled experiments, Eq. (14) is simplifed as
pressure. As a result of this inbalance, gas diffuses into the surrounding
matrix block. In this process, the fracture supplies gas mass to the 8
>
> α p 3
matrix block and the matrix pressure always lags the fracture pressure. > 1+
< −Δεs + p≤pc
k φ0 Ks
At the initial stage, gas diffuses only into the vicinity of the fracture viod = 3 :
k0 >
>
> α ε p p α p−pc 3
and local swelling controls the coal behavior. Because of the local : 1+ − L c + c 1+ p > pc
swelling, coal permeability experiences rapid reduction for each φ0 pc + pL Ks φ0 K
loading step from 2MPa to 10MPa. This swelling-genearated perme- ð15Þ
ability reduction is controlled by the internal (fracture) boundary.
When local swelling becomes macro swelling, coal permeability
recovers and this recovery is controlled by the external boundary 5.2. Model verification
condition. For all loading steps, the period of permeability reduction is
much shorter than that of permeability recovery. As shown in Fig. 6(a), In order to verify Eq. (15), an experiment was performed on
the period of permeability reduction is from 1000 s to 3000 s while that anthracite coal from the Northumberland Basin, Pennsylvania.
of permeability recovery from 10,000 s to 40,000 s. As shown in Fig. 6 Detailed description of the experiment can be found in Wang et al.
(b), sharp switches in behavior are observed for lower loading step
pressures from 2 MPa to 6 MPa while smooth switches in behavior are
1.1
observed for higher step pressures from 8 MPa to 10 MPa. All of these
observed behaviors are due to the fact that the matrix pressure always 1 2 MPa
lags the fracture pressure.
k/k0, permeability ratio
4 MPa
6 MPa
Based on the above analysis, the permeability switch from 0.9 8 MPa
reduction to recovery is corresponding to the matrix swelling switch 10 MPa Critical time
case of lower permeability before the local welling is switched to the Time (seconds)
macro swelling.
(a) Temporal evolution of coalpermeability ratio
5. Development and evaluation of permeability switching model 1.1
2.5
(2011). The sample was drilled parallel to the bedding plane. A Constant
roughness of 3 on Barton's scale (Barton, 1976) is estimated for
bedding planes in this coal. The cleat aperture is approximately 2 p Pore Pressure
pc 0
10–50 μm. The mean density of the coal under unconfined conditions
Permeability Ration
was calculated from the mass and volume of the three cylindrical
1.5
cores, and yielded an average matrix density of 1397.9 kg m − 3. In the
laboratory tests, permeability evolution was measured with respect to p Pore Pressure
k k0 1
pore pressure at constant total stresses of 12 MPa. At a total stress of 1
12 MPa the permeability to CO2 first decreases by 86% (pore pressure
Macro-Swelling
of 1.15 to 2.22 MPa) and then rebounds to a net 19% reduction over
0.5
the initial permeability at the final pore pressure of 4.58 MPa. In this
match, the following papameters were assumed: φf = 0.005, Rm = 0, Local-Swelling Completely Constrained Behaviour pc
1.2 6. Conclusions
1 Experimental Data A full coupling approach was applied to recover important non-
linear responses due to the effective stress effects during the
Permeability Ratio
controlled by the external boundary condition and behaves non- coal: anthracite from Qinshui Basin, China. International Journal of Coal Geology 81
(2), 128–138.
volumetrically. Harpalani, S., Chen, G., 1997. Influence of gas production induced volumetric strain on
• Matrix properties control the switch in swelling behavior from local permeability of coal. Geotechnical and Geological Engineering 15 (4), 303–325.
swelling to macro swelling and the associated switch in permeability Harpalani, S., Schraufnagel, R.A., 1990. Shrinkage of coal matrix with release of gas and
its impact on permeability of coal. Fuel 69 (5), 551–556.
behavior from reduction to recovery. A coal permeability model has Hsieh, P.A., Tracy, J.V., Neuzil, C.E., Bredehoeft, J.D., Silliman, S.E., 1981. A transient
been developed to represent this switching behavior in coal laboratory method for determining the hydraulic properties of ‘tight’ rocks–I.
permeability under variable stress conditions. Model predictions are Theory. International Journal of Rock Mechanics and Mining Sciences &
Geomechanics Abstracts 18 (3), 245–252.
consistent with typical laboratory and in-situ observations available in Izadi, G., Wang, S., Elsworth, D., Liu, J., Wu, Y., Pone, D., 2011. Permeability evolution of
the lietrature. fluid-infiltrated coal containing discrete fractures. International Journal of Coal
Geology 85 (2), 202–211.
Kiyama, T., Nishimoto, S., Fujioka, M., Xue, Z., Ishijima, Y., Pan, Z., Connel, L.D., 2011. Coal
Acknowledgments
swelling strain and permeability change with injecting liquid /supercritical CO2 and N2
at stress-constrained conditions. International Journal of Coal Geology 85 (1), 56–64.
This work is a partial result of support from CSIRO Flagship top-up Koperna, G.J., Oudinot, A.Y., McColpin, G.R., Liu, N., Heath, J.E., Wells, A., Young, G.B., 2009.
CO2-ECBM /storage activities at the San Juan Basin's pump Canyon test site. SPE Annual
scholarship, NIOSH under contract 200-2008-25702, Major State Basic
Technical Conference and Exhibition, New Orleans, Louisiana, SPE 124002.
Research Development Program Fund (grant no. 2010CB226801) and the Liu, H.-H., Rutqvist, J., 2010. A new coal permeability model: internal swelling stress and
State Administration of Foreign Experts Affairs, PR China (TS2011ZGKY fracture–matrix interaction. Transport in Porous Media 82 (1), 157–171.
[BJ]019). These sources of support are gratefully acknowledged. Liu, J., Chen, Z., Elsworth, D., Miao, X., Mao, X., 2010a. Linking gas-sorption induced
changes in coal permeability to directional strains through a modulus reduction
ratio. International Journal of Coal Geology 83 (1), 21–30.
References Liu, J., Chen, Z., Elsworth, D., Miao, X., Mao, X., 2010b. Evaluation of stress-controlled
coal swelling processes. International Journal of Coal Geology 83 (4), 446–455.
Bai, M., Elsworth, D., 2000. Coupled Processes in Subsurface Deformation, Flow and Liu, J., Chen, Z., Elsworth, D., Qu, H., Chen, D., 2011. Interactions of multiple processes
Transport. ASCE Press. 336 pp. during CBM extraction: a critical review. International Journal of Coal Geology.
Barton, N., 1976. The shear strength of rock and rock joints. International Journal of doi:10.1016/j.coal. 2011.06.004.
Rock Mechanics and Mining Sciences & Geomechanics Abstracts 13, 255–279. Ma, Q., Harpalani, S., Liu, S., 2011. A simplified permeability model for coalbed methane
Brace, W.F., Walsh, J.B., Frangos, W.T., 1968. Permeability of granite under high reservoirs based on matchstick strain and constant volume theory. International
pressure. Journal of Geophysical Research 73 (6), 2225–2236. Journal of Coal Geology 85 (1), 43–48.
Chen, Z., Liu, J., Elsworth, D., Connell, L., Pan, Z., 2009. Investigation of CO2 injection Massarotto, P., Golding, S.D., Rudolph, V., 2009. Constant volume CBM reservoirs: an
induced coal-gas interactions. 43rd U.S. Rock Mechanics Symposium & 4th U.S.- important principle. International Coalbed Methane Symposium, Tuscaloosa, Alabama.
Canada Rock Mechanics Symposium. American Rock Mechanics Association, Mazumder, S., Wolf, K.H., 2008. Differential swelling and permeability change of coal in
Asheville, North Carolina. response to CO2 injection for ECBM. International Journal of Coal Geology 74 (2),
Chen, Z., Liu, J., Elsworth, D., Connell, L.D., Pan, Z., 2010a. Impact of CO2 injection and 123–138.
differential deformation on CO2 injectivity under in-situ stress conditions. Palmer, I., Mansoori, J., 1996. How permeability depends on stress and pore pressure in
International Journal of Coal Geology 81 (2), 97–108. coalbeds: a new model. 1996 Copyright 1996 SPE Annual Technical Conference and
Chen, Z., Liu, J., Pan, Z., Connell, L., Elsworth, D., 2010b. Relations between coal Exhibition. Society of Petroleum Engineers, Inc., Denver, Colorado.
permeability and directional strains and their application to San Juan Basin. 44th Palmer, I.D., Mavor, M., Gunter, B., 2007. Permeability changes in coal seams during
U.S. Rock Mechanics Symposium and 5th U.S.-Canada Rock Mechanics Sympo- production and injection. International Coalbed Methane Symposium, University of
sium. American Rock Mechanics Association, Salt Lake City, Utah. Alabama, Tuscaloosa, Alabama. Paper 0713.
Chen, Z., Pan, Z., Liu, J., Connell, L.D., Elsworth, D., 2011. Effect of the effective stress Pan, Z., Connell, L.D., 2007. A theoretical model for gas adsorption-induced coal
coefficient and sorption-induced strain on the evolution of coal permeability: swelling. International Journal of Coal Geology 69 (4), 243–252.
experimental observations. International Journal of Greenhouse Gas Control. Pan, Z., Connell, L.D., Camilleri, M., 2010. Laboratory characterisation of coal reservoir
doi:10.1016/j.ijggc.2011.07.005. permeability for primary and enhanced coalbed methane recovery. International
Cherian, B.V., Claugus, A., Dilli, M., Mata, D., Sitchler, J.C., Alatrach, S., Neumiller, J., Journal of Coal Geology 82 (3–4), 252–261.
Panjaitan, M., 2010. An Integrated single-well approach to evaluating completion Patching, T.H., 1970. Retention and release of gas in coal—a review. Canadian Mining
effectiveness and reservoir properties in the Wind Dancer Field. Tight Gas and Metallurgical Bulletin 63 (703), 1302–1308.
Completions Conference, San Antonio, Texas, USA, SPE 137126. Pekot, L.J., Reeves, S.R., 2002. Modeling the effects of matrix shrinkage and differential
Chilingar, G.V., 1964. Relationship between porosity, permeability, and grain-size swelling on coalbed methane recovery and carbon sequestration. U.S. Department
distribution of sands and sandstones. In: Straaten, L.M.J.U.v. (Ed.), Developments in of Energy DE-FC26-00NT40924.
Sedimentology. Elsevier, pp. 71–75. Pini, R., Ottiger, S., Burlini, L., Storti, G., Mazzotti, M., 2009. Role of adsorption and
Clarkson, C.R., 2008. Case study: production data and pressure transient analysis of swelling on the dynamics of gas injection in coal. Journal of Geophysical Research
horseshoe Canyon CBM wells. CIPC/SPE Gas Technology Symposium 2008 Joint 114 (B4), B04203.
Conference, Calgary, Alberta, Canada. Reeves, S., Taillefert, A., Pekot, L., Clarkson, C., 2003. The Allison unit CO2-ECBM pilot: a
Connell, L.D., 2009. Coupled flow and geomechanical processes during gas production reservoir modelling study. Topical Report, U.S. Department of Energy, DE-FC26-
from coal seams. International Journal of Coal Geology 79 (1–2), 18–28. 0NT40924. (February).
Connell, L.D., Detournay, C., 2009. Coupled flow and geomechanical processes during Reiss, L.H., 1980. The Reservoir Engineering Aspects of Fractured Formations. Gulf
enhanced coal seam methane recovery through CO2 sequestration. International Publishing Co., Houston.
Journal of Coal Geology 77 (1–2), 222–233. Robertson, E.P., 2005. Modeling permeability in coal using sorption-induced strain data.
Connell, L.D., Lu, M., Pan, Z., 2010. An analytical coal permeability model for tri-axial SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers,
strain and stress conditions. International Journal of Coal Geology 84 (2), 103–114. Dallas, Texas.
Cui, X., Bustin, R.M., 2005. Volumetric strain associated with methane desorption and Robertson, E.P., Christiansen, R.L., 2008. A permeability model for coal and other
its impact on coalbed gas production from deep coal seams. AAPG Bulletin 89 (9), fractured, sorptive-elastic media. SPE Journal 13 (3), 314–424.
1181–1202. Seidle, J.R., Huitt, L.G., 1995. Experimental measurement of coal matrix shrinkage due to
Durucan, S., Edwards, J.S., 1986. The effects of stress and fracturing on permeability of gas desorption and implications for cleat permeability increases. 1995 Copyright
coal. Mining Science and Technology 3 (3), 205–216. 1995 International Meeting on Petroleum Engineering. Society of Petroleum
Elsworth, D., Bai, M., 1992. Flow-deformation response of dual porosity media. Journal Engineers, Inc., Beijing, China.
of Geotechnical Engineering 118 (1), 107–124. Shi, J.-Q., Durucan, S., 2004a. A numerical simulation study of the Allison unit CO2
Gilman, A., Beckie, R., 2000. Flow of coal-bed methane to a gallery. Transport in Porous -ECBM pilot: the impact of matrix shrinkage and swelling on ECBM production and
Media 41 (1), 1–16. CO2 injectivity. Proceedings of 7th International Conference on Greenhouse Gas
Gray, I., 1987. Reservoir engineering in coal seams: part 1—the physical process of gas Control Technologies.
storage and movement in coal seams. SPE Reservoir Engineering 2 (1), 28–34. Shi, J.Q., Durucan, S., 2004b. Drawdown induced changes in permeability of coalbeds: a
Gu, F., Chalaturnyk, J.J., 2005a. Analysis of coalbed methane production by reservoir and new interpretation of the reservoir response to primary recovery. Transport in
geomechanical coupling simulation. Journal of Canadian Petroleum Technology 44 Porous Media 56 (1), 1–16.
(10), 33–42. Siriwardane, H., Haljasmaa, I., McLendon, R., Irdi, G., Soong, Y., Bromhal, G., 2009.
Gu, F., Chalaturnyk, R.J., 2005b. Sensitivity study of coalbed methane production with Influence of carbon dioxide on coal permeability determined by pressure transient
reservoir and geomechanic coupling simulation. Journal of Canadian Petroleum methods. International Journal of Coal Geology 77 (1–2), 109–118.
Technology 44 (10), 23–32. Somerton, W.H., Söylemezoglu, I.M., Dudley, R.C., 1975. Effect of stress on permeability
Gu, F., Chalaturnyk, R.J., 2006. Numerical simulation of stress and strain due to gas of coal. International Journal of Rock Mechanics and Mining Sciences &
sorption/ desorption and their effects on in-Situ permeability of coalbeds. Journal of Geomechanics Abstracts 12 (5–6), 129–145.
Canadian Petroleum Technology 45 (10), 52–62. Sparks, D.P., McLendon, T.H., Saulsberry, J.L., Lambert, S.W., 1995. The effects of stress
Han, F., Busch, A., Wageningen, N., Yang, J., Liu, Z., Krooss, B.M., 2010. Experimental on coalbed reservoir performance, Black Warrior Basin. U.S.A. SPE Annual Technical
study of gas and water transport processes in the inter-cleat (matrix) system of Conference and Exhibition, Dallas, Texas.
40 J. Liu et al. / International Journal of Coal Geology 88 (2011) 31–40
Wang, G.X., Wei, X.R., Wang, K., Massarotto, P., Rudolph, V., 2010. Sorption-induced Wu, Y., Liu, J., Elsworth, D., Miao, X., Mao, X., 2010b. Development of anisotropic
swelling/ shrinkage and permeability of coal under stressed adsorption/desorption permeability during coalbed methane production. Journal of Natural Gas Science
conditions. International Journal of Coal Geology 83 (1), 46–54. and Engineering 2 (4), 197–210.
Wang, S., Elsworth, D., Liu, J., 2011. Permeability evolution in fractured coal: the roles of Wu, Y., Liu, J., Chen, Z., Elsworth, D., Pone, D., 2011. A dual poroelastic model for CO2-
fracture geometry and water-content. International Journal of Coal Geology 87, enhanced coalbed methane recovery. International Journal of Coal Geology 86 (2–3),
13–25. 177–189.
Warren, J.E., Root, P.J., 1963. The behavior of naturally fractured reservoirs. Society of Young, G.B.C., McElhiney, J.E., Paul, G.W., McBane, R.A., 1991. An analysis of Fruitland
Petroleum Engineers Journal 3 (3), 245–255. coalbed methane production, Cedar Hill Field, Northern San Juan Basin. SPE Annual
Wong, S., Law, D., Deng, X., Robinson, J., Kadatz, B., Gunter, W.D., 2007. Enhanced Technical Conference and Exhibition, Dallas, Texas.
coalbed methane and CO2 storage in anthracitic coals—Micro-pilot test at South Zhang, H., Liu, J., Elsworth, D., 2008. How sorption-induced matrix deformation affects
Qinshui, Shanxi, China. International Journal of Greenhouse Gas Control 1 (2), gas flow in coal seams: a new FE model. International Journal of Rock Mechanics
215–222. and Mining Sciences 45 (8), 1226–1236.
Wu, Y., Liu, J., Elsworth, D., Chen, Z., Connel, L., pan, Z., 2010a. Dual poroelastic response
of a coal seam to CO2 injection. International Journal of Greenhouse Gas Control 4
(4), 668–678.
CHAPTER 3 FLUID PROPERTIES
To understand and predict the volumetric behavior of oil and gas reservoirs as a function of
pressure, knowledge of the physical properties of reservoir fluids must be obtained. These
fluid properties are usually determined by lab experiments on samples of actual reservoir
fluids. In the absence of experimentally measured properties, it is necessary for the petroleum
engineer to determine the properties from empirically derived correlations. The objective of
this chapter is to define fundamental PVT properties for the following reservoir fluids:
Natural Gases
Crude Oil
Water
Ground Surface
1. Gas
2. Oil ? Volumetric Relations
3. Water Gas Reservoir
Reservoir Pressure p p ,T
Reservoir Volume V p ,T
A gas is defined as a homogeneous fluid of low viscosity and density that has no definite
volume but expands to completely fill the vessel in which it is placed. Generally, the natural
gas is a mixture of hydrocarbon and nonhydrocarbon gases. The hydrocarbon gases that are
normally found in a natural gas are methanes, ethanes, propanes, butanes, pentanes, and small
amounts of hexanes and heavier. The nonhydrocarbon gases (i.e., impurities) include carbon
dioxide, hydrogen sulfide, and nitrogen.
Behavior of Ideal Gases: The kinetic theory of gases postulates that gases are composed of a
very large number of particles called molecules. For an ideal gas, the volume of these
molecules is insignificant compared with the total volume occupied by the gas. It is also
assumed that these molecules have no attractive or repulsive forces between them, and that all
collisions of molecules are perfectly elastic. Based on the above kinetic theory of gases, a
mathematical equation called equation-of-state can be derived to express the relationship
existing between pressure p, volume V, and temperature T for a given quantity of moles of
gas n. This relationship for perfect gases is called the ideal gas law and is expressed
mathematically by the following equation:
PV nRT (3.1)
Where ρ is the absolute pressure, psi; V is the volume, cft; T is the temperature, oR; n is the
number of moles of gas, lb-mole; R is the universal gas constant which, for the above units,
has the value of 10.730psi cft/lb-mole oR.
Behavior of Real Gases: Basically, the magnitude of deviations of real gases from the
conditions of the ideal gas law increases with increasing pressure and temperature and varies
widely with the composition of the gas. Real gases behave differently than ideal gases. The
reason for this is that the perfect gas law was derived under the assumption that the volume of
molecules is insignificant and that no molecular attraction or repulsion exists between them.
This is not the case for real gases. Numerous equations-of-state have been developed in the
attempt to correlate the pressure-volume-temperature variables for real gases with
experimental data. In order to express a more exact relationship between the variables p, V,
and T, a correction factor called the gas compressibility factor, gas deviation factor, or simply
the z-factor, must be introduced into Equation 3-1 to account for the departure of gases from
ideality. The equation has the following form:
PV znRT (3.2)
Where the gas deviation factor z is a dimensionless quantity and is defined as the ratio of the
actual volume of n-moles of gas at T and p to the ideal volume of the same number of moles
at the same T and p:
Vactual V
z (3.3)
Videal nRT / p
PRESSURE, PSI
Many gases near atmospheric PT conditions approach ideal behavior (Z=1). All molecules
have two tendencies: 1. to fly apart from each other because of their constant kinetic motion,
and 2. to come together because of electrical attractive forces between molecules. Because the
molecules are quite far apart, the attractive forces are negligible, and the gas behaves close to
ideal. Also at high temperatures the kinetic motion, being greater, makes the attractive forces
comparatively negligible, and, again, the gas approaches ideal behavior. When the gas is
highly compressed, the gas appears to be more difficult to compress.
QUESTION– For an ideal gas under the isothermal condition, prove the following relation
between two P-V states:
p1 V2
p2 V1
p 14.7 psi
V ?
Ground Surface
p 1000 psi
V 106 cf
The gas formation volume factor is used to relate the volume of gas, as measured at reservoir
conditions, to the volume of the gas as measured at standard conditions, i.e., 60°F and 14.7
psia. This gas property is then defined as the actual volume occupied by a certain amount of
gas at a specified pressure and temperature, divided by the volume occupied by the same
amount of gas at standard conditions. In an equation form, the relationship is expressed as
V
Bg
g p ,T
V
(3.4)
g SC
Bg Gas Formation Volume Factor ft 3 / scf
V
g SC Volume of Gas at Standard Conditions
V
g p ,T Volume of Gas at Reservoir Conditions
Where z SC is z-factor, zSC 1 at standard conditions, pSC and TSC are standard pressure and
standard temperature. Assuming that the standard conditions are represented by psc =14.7psia
and Tsc = 520, the above expression can be reduced to the following relationship:
zT
Bg 0.02827 (3.6)
p
Where Bg is gas formation volume factor, z is gas compressibility factor, T is temperature.
For an ideal gas, equation (3.7) becomes
T
Bg 0.02827 (3.7)
p
Under the isothermal condition, the gas formation volume factor is inversely proportional to
pressure as shown in the following Figure.
p
Figure 3.2 Relation between gas formation volume factor and pressure.
Gas Compressibility
By definition, the isothermal gas compressibility is the change in volume per unit volume for
a unit change in pressure or, in equation form:
1 Vg
cg
Vg p (3.8)
nRTz
Vp (3.9)
p
V p 1 z z
nRT 2 (3.10)
p p p p
Substituting (3.10) into (3.8) gives
1 1 z
cg (3.11)
p z p
1
cg (3.12)
p
Dissolved Gas-Oil Ratio
Dissolved gas-oil ratio is defined as the ratio of surface gas to stock-tank oil in a reservoir
liquid phase at reservoir conditions. The volumes of surface gas and stock-tank oil are those
measured at standard conditions. The dissolved gas-oil ratio, Rv is defined as
V
Rs
g SC
(3.13)
Vo SC
For a particular gas and crude oil to exist at a constant temperature, the solubility increases
with pressure until the saturation pressure is reached. At the saturation pressure (bubble-point
pressure) all the available gases are dissolved in the oil and the gas solubility reaches its
maximum value. Rather than measuring the amount of gas that will dissolve in a given
stock-tank crude oil as the pressure is increased, it is customary to determine the amount of
gas that will come out of a sample of reservoir crude oil as pressure decreases.
A typical gas solubility curve, as a function of pressure for an undersaturated crude oil, is
shown in Figure 3.3. As the pressure is reduced from the initial reservoir pressure pi, to the
bubble-point pressure pb, no gas evolves from the oil and consequently the gas solubility
remains constant at its maximum value of Rsb. Below the bubble-point pressure, the solution
gas is liberated and the value of Rs decreases with pressure.
When a reservoir fluid moves into the two-phase region, the vapor phase may contain a liquid
component. This is characterized with a liquid/gas ratio parameter, Rv, that is analogous to the
solution-gas/oil ratio parameter, Rs. While Rs gives the amount of gas in the liquid phase, Rv
gives the amount of liquid in the vapor phase. Rv is relevant only for gas condensates and
Vo SC
Rv
V
g SC
(3.14)
The volatilized oil-gas ratio is defined as the ratio of the volume of stock-tank oil to surface
gas contained in a reservoir vapor phase at reservoir conditions. The volumes of surface gas
and stock-tank oil are those measured at standard conditions.
Rv Rs 1
Petroleum (an equivalent term is crude oil) is a complex mixture consisting predominantly of
hydrocarbons and containing sulfur, nitrogen, oxygen, and helium as minor constituents. The
physical and chemical properties of crude oils vary considerably and dependent on the
concentration of the various types of hydrocarbons and minor constituents present.
The oil formation volume factor, Bo, is defined as the ratio of the volume of oil (plus the gas
in solution) at the prevailing reservoir temperature and pressure to the volume of oil at
standard conditions. Bo is always greater than or equal to unity. The oil formation volume
factor can be expressed mathematically as:
Vo p,T
Bo (3.15)
Vo SC
A typical oil formation factor curve, as a function of pressure for an undersaturated crude oil
(pi > pb), is shown in Figure 3.4. As the pressure is reduced below the initial reservoir
pressure pi, the oil volume increases due to the oil expansion. This behavior results in an
increase in the oil formation volume factor and will continue until the bubble-point pressure is
reached. At pb, the oil reaches its maximum expansion and consequently attains a maximum
value of Bob for the oil formation volume factor. As the pressure is reduced below pb,
volume of the oil and Bo are decreased as the solution gas is liberated. When the pressure is
reduced to atmospheric pressure and the temperature to 60°F, the value of Bo is equal to one.
Oil Compressibility
1 Vo
co (3.16)
Vo p
For a crude oil system, the compressibility of the oil phase is defined for pressures above the
bubble-point by one of the following equivalent expressions
1 Bo
co (3.17)
Bo p
1 o
co (3.18)
o p
1 Bo Bg Rs
co (3.19)
Bo p Bo p
Note that the first term in Eq (3.19) is negative. The second term is necessary because FVF’s
Tuesday, February 25, 2014
contain the effect of solution gas on the change in liquid volume caused by gas going into
solution as the pressure is increased.
WATER
Vw p,T
Bw (3.20)
Vw SC
Two-Component Model
1. At most, there are two hydrocarbon pseudo-components: surface gas and
stock-tank-oil;
2. At most, there are two hydrocarbon phases: gas and oil;
3. The surface gas pseudo-component is defined by the composition of the gas at
standard conditions;
4. The stock-tank-oil pseudo-component is defined by the composition of the stock-tank
oil at standard conditions;
5. The surface gas can partition between the oil and gas phases;
6. The stock-tank oil can partition between the oil and the gas phases;
7. Thermodynamic equilibrium exists.
Assumption 5: The partitioning of surface gas into the oil phase allows for dissolved gas;
Assumption 6: The partitioning of stock-tank oil into gas phase allows for volatilized oil;
Partitioning implicitly exists at all conditions except standard conditions. Standard conditions
to measure a standard cubic food are defined as 14.7psi and 60oF.
GROUND SURFACE
p pbp
p pbp
OIL
1 GAS
RESERVOIR
CONDITIONS
2 OIL
Two-phase or total formation volume factors are secondary PVT properties. They are strictly a
function of the standard PVT relations ( Bo , Bg , Rs ,and Rv ); consequently, they can always be
calculated from the standard PVT.
Two-Phase Oil Formation Volume Factor ( Bto ): the total (liquid-phase plus gas-phase)
volume at reservoir conditions divided by its resulting oil-phase volume at standard
conditions.
V o V g p ,T
Bto (3.21)
Vo SC
1 GAS
2 OIL
Btg 3 GAS
2 OIL
Bto OIL
4 OIL 3 GAS 4
GROUND SURFACE
p pbp
p pbp
RESERVOIR
1 GAS
CONDITIONS
OIL
2 OIL
V Vg p ,T
Btg
o
(3.22)
V o SC
4 OIL
3 GAS 3 GAS
GROUND SURFACE
p pdw
The gas expansivity is defined
analogously: the total expansion of a pi pdw
1 GAS
unit mass of gas phase between two
pressures at the reservoir GAS
temperature, expressed in units of 2 OIL
reservoir volume per unit volume of
gas at standard conditions. The
expansion volume includes liquid
dropout if such condensation occurs. Moveable Piston Moveable Piston
RESERVOIRCONDITIONS
Figure 3.7 Schematic interpretation of two phase gas formation volume factor.
Bo 1 Rsi Rv Bg Rsi Rs
Bto (3.23)
1 Rv Rs
Bg 1 Rvi Rs Bo Rvi Rv
Btg (3.24)
1 Rv Rs
Equations (23) and (24) apply to any fluid whose overall composition is defined by a gas-oil
ratio, Rsi , or by a oil-gas ratio, Rvi .
Bto Bo Bg Rsi Rs
(3.25)
Btg Bg
The oil-phase expansivity is defined as the total expansion of a unit mass of oil phase between
two pressures at the reservoir temperature, expressed in units of reservoir volume per unit
volume of oil at standard conditions. The two subject pressures are a reference pressure and a
terminal pressure. The terminal pressure is invariably less than the reference pressure, hence
the term expansivity. The reference pressure is usually the initial reservoir pressure. The
expansion volume includes a gas-phase volume if one should emerge from the expanding oil
phase.
GROUND SURFACE
p pbp
The oil-phase expansivity is pi pbp
defined as the total expansion
1 GAS
of a unit mass of oil phase
OIL
between two pressures at the 2 OIL
reservoir temperature,
expressed in units of reservoir
volume per unit volume of oil
at standard conditions. Moveable Piston Moveable Piston
RESERVOIRCONDITIONS
The gas expansivity is defined analogously: the total expansion of a unit mass of gas phase
between two pressures at the reservoir temperature, expressed in units of reservoir volume per
unit volume of gas at standard conditions. The expansion volume includes liquid dropout if
such condensation occurs.
Eo p, pi Bto p Btoi pi
(3.26)
E g p, pi Btg p Btgi pi
4 OIL
3 GAS 3 GAS
GROUND SURFACE
p pdw
The gas expansivity is defined
analogously: the total expansion of a pi pdw
1 GAS
unit mass of gas phase between two
pressures at the reservoir GAS
temperature, expressed in units of 2 OIL
reservoir volume per unit volume of
gas at standard conditions. The
expansion volume includes liquid
dropout if such condensation occurs. Moveable Piston Moveable Piston
RESERVOIRCONDITIONS
Bo 1 Rsi Rv Bg Rsi Rs
Bto
1 Rv Rs
Bto ( p 4320) Bo ( p 4320) 1.3647
Bto ( p 4225)
1.3578 1 710 23 106 0.845 103 710 693.5
1 693.5 23 106
1.3800
1
Bo Rs Bg Rs
Rv
Reservoir Oil
Vtotal RC could be treated either as OIL or as GAS.
LIST OF SYMBOLS
F Net fluids undergroun d withdrawn Rsoi initial solution gas oil ratio
E Net fluids expansion Bg Gas formation volume factor
N foi Original free - phase in place Rv Volatile oil gas ratio
N Original Oil In Place Rvi Initial volatile oil gas ratio
G fgi Original free - phase gas in place Bto Two pahse oil formation volume factor
G Original Gas In Place Btoi Inital two phase oil formation volume factor
E o Oil expansion factor Btg two phase gas formation volume factor
E g Gas expansion factor Btgi Inital two phase gas formation volume factor
t During production
a
p 14.7 psi Standard Pressure Produced
Mass
Ground Surface
t During production
p Reservoir Pressure
Total Mass = +
t 0 before production Residual
Mass
p p0 Initial Pressure
Produced
Volume
Expanded
Volume
Initial
Volume
As illustrated above, the total system expansion must be equal to the production. Therefore,
the key to derive the general material balance equation is to define the total system expansion
and the production, respectively.
In this unit, we use E to represent the expansion, F the production. Therefore, the general
material balance equation can be defined as
EF (4.1)
Equation (4.1) is the simplest form of the generalized material balance equation.
GAS
5 GAS
7 GAS G fgi
5 GAS G fgi Rvi
N foi Rsi 8 OIL
N
N foi 6 OIL
4 OIL
GROUND SURFACE
G G fgi N foi Rsi Original Gas In Place (OGIP) pi
N N foi G fgi Rvi Original Oil In Place (OOIP)
1 GAS
To Produce Gp, the reservoir needs to expand:
G fgi E g G fgi Btg Btgi
OIL
2
To Produce Np, the reservoir needs to expand:
Bg Bo Rv B Bg Rs
G p N p o
1 Rv Rs 1 Rv Rs Np 4 OIL
GROUND SURFACE
RESERVOIRCONDITIONS
p
pi
OIL
1
2
GAS
OIL
G fgi E g N foi Eo
B Bo Rv B Bg Rs
G fgi Eg N foi Eo G p g N p o (4.2)
1 Rv Rs 1 Rv Rs
V pore
1 S wi WATER
V pi V pore
1 V p ROCK
c f cw S wi
V pi pi p
Assuming Er and Ew are rock expansivity and water expansivity, respectively, the total
reservoir expansion due to rock compaction and water expansion can be defined as
B Bo Rv B Bg Rs
G fgi Eg N foi Eo N foi Boi G fgi Bgi E fw G p g N p o (4.4)
1 Rv Rs 1 Rv Rs
E fw
c f cw S wi pi p
Reservoir Formation Expansivit y (4.5)
1 S wi
For completeness, we also include the water balance into the MBE. We use We to represent
the water influx from the supporting aquifer beneath the reservoir; WI to represent the amount
We define
I gd I od I cd I wd 1
G fgi E g
I gd Gas Expansion Drive Index
Et
N foi E o
I od Oil Expansion Drive Index
Et
G B gi N foi Boi E fw
(4.8)
I cd
fgi
Compaction Drive Index
Et
We W p WI Bw
I wd Water Drive Index
Et
These drive indices provide a convenient means to rank the different producing mechanisms.
They change with time. For example, the gas drive may dominate early in the life of a
reservoir while the water drive dominates later after water influx commences.
Estimate the original oil and gas in place (OOIP and OGIP);
Estimate the gas-cap size;
Estimate water influx;
Estimate water influx model parameters;
Estimate geophysical parameters;
Confirm producing mechanisms.
The application forms may be different for different reservoir types. In the following, we
present some general principles.
Definitive Tool or Diagnostic Tool : The simplest form of MBE can be written as
FE
where, F and E are the hydrocarbon production and the reservoir expansion in volume. If the
material is fully balanced, F vs. E is a straight line; if the material is NOT fully balanced, F vs.
Tuesday, February 25, 2014
E is a curve, as illustrated in Figure 4.6.
IMBALANCE
BALANCE
E
Figure 4.6 Illustration of Material Balance and Imbalance.
FE 0
Equation above can be used to solve only for one unknown. If We or E fw are not known
and are dropped off from the equation, equation (4.1) becomes
FE0
If the water production data are not available and are dropped off from the MBE, equation
(4.1) becomes
FE0
Recovery Ratios: In practice, we calculate the fraction of the total OOIP recovered. The oil
recovery fraction and gas recovery fraction are defined as
Np
Fo (4.9a)
N
Gp
Fg (4.9b)
G
respectively.
A reservoir that produces only gas and no appreciable hydrocarbon liquids is called a dry-gas
reservoir. Dry-gas reservoir is the simplest type to evaluate. Gas reservoirs often have high
recovery factors irrespective of the drive mechanism. Recovery factors of more than 80% are
not uncommon, and can even be found in volumetric reservoirs. Natural gas is very
compressible and has a very low viscosity, and both factors contribute to the high recovery
factor.
The recovery factor is lower in waterdrive gas reservoirs primarily because waterdrive leaves
behind a residual gas saturation that is unrecoverable. This residual gas saturation is usually at
high pressure. The residual gas saturation at high pressure has a large volume of gas
compressed into a small space. In heterogeneous waterdrive reservoirs, water tends to flow
along the high permeability streaks, bypassing gas that gets trapped in low permeability areas.
This becomes a severe problem if the reservoir is highly fractured and has a strong waterdrive.
The waterdrive flushes the gas out of the fractures and bypasses the great majority of the gas
contained in the matrix.
4000
Single-phase Gas
Dew Point or Retrograde
Bubble Point or Reservoir
Gas –Condensate
Dissolved Gas Reservoir
A
3500 Reservoir B Cricondentherm
Po
o
De
le
Path
bb
Bu
w
2500 C1
Po
in
B2
t
80% D
2000 40% A1
psi
20% 5%
1500 10%
0%
A2 B3
1000
Ah 1 S wi
G (5.1)
Bgi
Where
AhS gr
Ga (5.2)
B ga
G Ga Bgi S gr
FR 1 (5.3)
G Bga 1 S wi
If the reservoir is homogeneous and volumetric (no waterdrive), S wi 0 and S gr 1 ,
equation (5.3) can be simplified as
G Ga B gi
FR 1 (5.4)
G Bga
MBE Analysis
Two conditions:
N foi 0
(5.5)
Rvi Rv 0
B g Bo Rv B B g Rs
W p WI BW
(5.7)
F G ps N p o
1 Rv Rs 1 Rv Rs
We 0
N foi 0
c f c w S wi
E fw cT p p
1 S wi
Tuesday, February 25, 2014
Eowf Eo Boi cT p Eo Boi E fw (5.8)
E gwf E g Bgi cT p E g Bgi E fw
Equation (5.11) is the equation most commonly used to analyze the normal volumetric gas
reservoirs.
P-Z Plot
We note
zTpsc
Bg (5.12)
pTsc
p pi pi G p
(5.14)
z z i z i G fgi
Example 1: A volumetric dry-gas reservoir with an initial pressure of 3000 psi (zi=0.912) and
formation temperature 190F has produced 384MMscf of gas, and the pressure has dropped to
2,876 psi (z=0.907). Determine the original gas in place (OGIP) by material balance.
Solution:
The material balance equation for this problem is
A volumetric dry-gas reservoir has the measured production and pressures given below
P(psi) Gp(MMscf) z
3000 0 0.912
2876 384 0.907
2824 550 0.905
2755 788 0.903
2688 1002 0.902
2570 1445 0.901
2435 1899 0.900
2226 2670 0.901
2122 3113 0.903
1866 3982 0.905
Solution:
Alternatively,
p pi pi
Gp (E2.2)
z zi zi G fgi
We can use both equations to solve for Gfgi.
3500
3000
2500
2000
p/z, psi
40
35 y = 10740x
2
R = 0.9998
30
GpBg 25
20
15
10
5
0
0 0.001 0.002 0.003 0.004
Bg-Bgi=Eg
Example 3:
If we do not know if there is a waterdrive, we can use the material balance equation as a
diagnostic tool. The material balance equation for a waterdrive gas reservoir can be written as
F G p Bg G fgi E g W
F G p Bg W
G fgi
Eg Eg Eg
11500
F 11000
Eg 10500
10000
9500
9000
0 1000 2000 3000 4000 5000
Gp
The graph verifies the volumetric nature of the gas reservoir.
Flow in porous media is a very complex phenomenon and as such cannot be described as
explicitly as flow through pipes or conduits. It is rather easy to measure the length and
diameter of a pipe and compute its flow capacity as a function of pressure; in porous media,
however, flow is different in that there are no clear-cut flow paths which lend themselves to
measurement.
The analysis of fluid flow in porous media has evolved throughout the years along two
fronts—the experimental and the analytical. Physicists, engineers, hydrologists, and the like
have examined experimentally the behavior of various fluids as they flow through porous
media ranging from sand packs to fused Pyrex glass. On the basis of their analyses, they have
attempted to formulate laws and correlations that can then be utilized to make analytical
predictions for similar systems.
The main objective of this chapter is to present the mathematical relationships that are
designed to describe the flow behavior of the reservoir fluids. The mathematical forms of
these relationships will vary depending upon the characteristics of the reservoir. The primary
reservoir characteristics that must be considered include:
A Flow Cross Area m 2 Ax Rock Volume
Q Flow Rate m / s
3 Ax Pore Volume
Density kg / m 3 Ax Liquid Mass
Ax
Porosity Mass Change Rate
t
Ax A
Mass Change Rate
t
Au x Au x
Au x
x
x x
Au x Ax
ux
k dp x
dx x t
Au x Q Volumetric Flow Rate u x
Au x Mass Flow Rate Flow Equation
x t
k p
ux (6.1)
x
k 2 p
(6.2)
x 2 t
We assume that both the rock and the fluid are incompressible:
remains constant;
remains constant;
Furthermore, we assume rock permeability, k , and viscosity, , are also constant. Under these
conditions, Equation (6.2) becomes
2 p
0 (6.3)
x 2
Determine the pressure distribution, px , and the velocity distribution, ux
k 10 15 m 2 0.001 pa s A 1m 2
p1
1MPa p2 0
10m
x
0
k 10 12 m 2 k 10 15 m 2
p1
1MPa p2 0
10m
x
0
We assume that the flow-in mass rate is equal to the flow-out mass rate. Under this condition,
the flow equation becomes
u x
0 (6.4)
x
In the following, we apply Equation (6.4) to the ideal gas flow. For the ideal gas, its density is
directly proportional to pressure:
p (6.5)
Where is constant.
GAS FLOW
Determine the pressure distribution, p x , and the velocity distribution, u x
k 10 15 m 2 0.001 pa s A 1m 2
p1
1MPa p2 0
10m
x
0
If 0 , flow is called unsteady state flow. Three possibilities exist:
t
1. the fluid itself is compressible;
2. the medium is compressible;
3. both the fluid and the medium are compressible.
Assuming both the fluid and the medium are compressible, the right hand side of the flow
equation becomes
p p p
c c p ct (6.6)
t t t t t t
k p k p p
ct (6.7)
x x y y t
p 2 p 2 1 p
(6.8)
x 2 y 2 t
Where
k
(6.9)
ct
is called hydraulic diffusivity. The hydraulic diffusivity determines how fast the flow
regimes transform.
There are basically three types of flow regimes that must be recognize in order to describe the
fluid flow behavior and reservoir pressure distribution as a function of time. There are three
flow regimes:
Unsteady-state flow
Pseudo steady-state flow
Steady State flow
Steady-State Flow: The flow regime is identified as a steady-state flow if the pressure at
every location in the reservoir remains constant, i.e., does not change with time.
Mathematically, this condition is expressed as:
p
0 (6.10)
t i
Unsteady-State Flow: The unsteady-state flow (frequently called transient flow) is defined as
the fluid flowing condition at which the rate of change of pressure with respect to time at any
position in the reservoir is not zero or constant. This definition suggests that the pressure
derivative with respect to timeis essentially a function of both position i and time t, thus
p
f i, t (6.11)
t i
Pseudo Steady State Flow: When the pressure at different locations in the reservoir is
declining linearly as a function of time, i.e., at a constant declining rate, the flowing condition
is characterized as the pseudo-steady-state flow. Mathematically, this definition states that the
rate of change of pressure with respect to time at every position is constant, or
p
Constant (6.12)
t i
It should be pointed out that the pseudo steady-state flow is commonly referred to as semi
steady-state flow and quasisteady-state flow.
Methane in coal seams is an important natural energy resource although ignition and the
resulting explosion hazard remains a major problem during coal mining. Degassing seams is
an important method to mitigate this hazard, and results in the beneficial recovery of a
clean-burning and low-carbon fuel resource. The injection of carbon dioxide to preferentially
dissociate methane has been an effective measure used after primary recovery by
depressurization. Recently, carbon dioxide sequestration in deep coal seams has attracted
more attention as a method of reducing the output of greenhouse gases to the atmosphere.
Gas flow within coal seams is quite different from that of conventional reservoirs. Detailed
studies have examined the storage and transport mechanisms of gas in coal seams. In-situ and
laboratory data indicate that the storage and flow of gas in coal seams is associated with the
matrix structure of coal and the absorption or desorption of gas. Coal is a naturally fractured
dual porosity reservoir, consisting of micro-porous matrix and cleats. Most of the gas is
initially stored within micro-pores in the absorbed state. When gas recovery begins, the gas
desorbs and diffuses from matrix to cleats due to the concentration gradient. The gas flowing
through the cleats is considered to be gas seepage controlled by the permeability of the coal
seam. Experimental results have shown that gas sorption generally follows a Langmuir
isotherm. Desorption plays an important role in both defining the longevity and rate of the gas
supply.
Adsorption of gas, such as carbon dioxide, is the reverse of desorption. When the gas pressure
increases, the gas adsorbs onto the coal matrix. The increase of pore pressure results in a
decrease of effective stress. The reduction in effective stress enhances the coal permeability.
In contrast, the adsorption-induced swelling of the coal matrix reduces the cleat apertures and
decreases the permeability. The net change in permeability accompanying gas sequestration is
also controlled competitively by the influence of effective stresses and matrix swelling, again
controlled by the boundary conditions applied locally between the end-members of null
changes in either mean stress or volume strain.
GOVERNING EQUATIONS
In the following, a set of field equations are defined to prescribe the transport and interaction
of gas flow. These derivations are based on the following assumptions:
a. Coal is a homogeneous, isotropic and elastic continuum.
b. Strains are much smaller than the length scale.
c. Gas contained within the pores is ideal, and its viscosity is constant under isothermal
conditions.
d. The rate of gas flow through the coal is defined by Darcy’s law.
e. Conditions are isothermal.
f. Coal is saturated by gas.
g. Compositions of the gas are not competitive, i.e., one gas component at a time.
where
ga
is the gas density at standard conditions,
c is the coal density and is
Tuesday, February 25, 2014
porosity. VL represents the Langmuir volume constant and PL represents the Langmuir
pressure constant. According to the ideal gas law, the gas density is proportional to the pore
gas pressure and can be described as
Mg
g p
RT (6.15)
Mg
Where is the molecular mass of the gas, R is the universal gas constant and T is the
gas temperature.
Assuming the effect of gravity is relatively small and can be neglected, the Darcy velocity is
defined as
k
ug p
(6.16)
Where k denotes permeability and denotes the dynamic viscosity of the gas.
Substituting equations (6.14), (6.15) and (6.16) into Equation (6.13), we obtain
c paVL PL p k
( p P )2 t p t pp Qs
L (6.17)
where p a is one atmosphere of pressure (101.325 kPa). In Equation (6.17), the permeability
k is dependent on the porosity, . Because the change in porosity is negligible, the final flow
equation becomes:
c paVL PL p k
( p P ) 2 t pp Qs
L (6.18)
We adopt the Palomer-Mansoori model which specifies the permeability expression as
follows:
3
k
1
1
p p0 L K 1 p p0
k0 M M PL p PL p0
(6.19)
E 1 E
M K
Where 1 1 2 and 3(1 2 )
K represents the bulk modulus of coal. E is the Young’s modulus of coal. is Possion’s
ratio of coal.
Abstract
The influence of sorption-induced coal matrix deformation on the evolution of porosity and permeability of fractured coal seams is
evaluated, together with its influence on gas recovery rates. The porosity-based model considers factors such as the volume occupied by
the free-phase gas, the volume occupied by the adsorbed phase gas, the deformation-induced pore volume change, and the sorption-
induced coal pore volume change. More importantly, these factors are quantified under in situ stress conditions. A cubic relation between
coal porosity and permeability is introduced to relate the coal storage capability (changing porosity) to the coal transport property
(changing permeability). A general porosity and permeability model is then implemented into a coupled gas flow and coal deformation
finite element model. The new FE model was used to compare the performance of the new model with that of the Palmer–Mansoori
model. It is found that the Palmer–Mansoori model may produce significant errors if loading conditions deviate from the assumptions of
the uniaxial strain condition and infinite bulk modulus of the grains. The FE model was also applied to quantify the net change in
permeability, the gas flow, and the resultant deformation in a coal seam. Model results demonstrate that the evolution of porosity and of
permeability is controlled by the competing influences of effective stresses and sorption-based volume changes. The resulting sense of
permeability change is controlled by the dominant mechanism.
r 2008 Elsevier Ltd. All rights reserved.
Keywords: Matrix shrinking; Coal permeability; Coal porosity; Coupled model; Simulation
1. Introduction Gas flow within coal seams is quite different from that of
conventional reservoirs. Detailed studies have examined
Methane in coal seams is an important natural energy the storage and transport mechanisms of gas in coal seams.
resource, although ignition and the resulting explosion In situ and laboratory data indicate that the storage and
hazard remain a major problem during coal mining. flow of gas in coal seams is associated with the matrix
Degassing seams is an important method to mitigate this structure of coal and the absorption or desorption of gas.
hazard, and results in the beneficial recovery of a clean- Coal is a naturally fractured dual-porosity reservoir [2],
burning and low-carbon fuel resource. The injection of consisting of micro-porous matrix and cleats. Most of the
carbon dioxide to preferentially dissociate methane has gas is initially stored within micro-pores in the absorbed
been an effective measure used after primary recovery by state [3]. When gas recovery begins, the gas desorbs and
depressurization. Recently, carbon dioxide (CO2) seques- diffuses from matrix to cleats due to the concentration
tration in deep coal seams has attracted more attention as a gradient. The gas flowing through the cleats is considered
method of reducing the output of greenhouse gases to the to be gas seepage controlled by the permeability of the coal
atmosphere [1]. seam [4]. Experimental results have shown that gas
sorption generally follows a Langmuir isotherm [5,6].
Corresponding author. Tel.: +61 8 6488 7205; fax: +61 8 6488 1964. Desorption plays an important role in both defining the
E-mail address: Jishan@cyllene.uwa.edu.au (J. Liu). longevity and rate of the gas supply, and in controlling the
1365-1609/$ - see front matter r 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijrmms.2007.11.007
ARTICLE IN PRESS
H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236 1227
related deformation of the solid matter comprising from the coal matrix into cleats to have a critical influence
the seam. on overall methane production [19]. A model for multi-
A variety of experiments have investigated sorption phase flow, coupled with heat transfer and rock deforma-
characteristics under isothermal conditions [4,7–10] with tion, was used to simulate CO2 injection into a brine
supporting models representing isothermal response formation by Rutqvist and Tsang [20]. In 2004, a non-
[8,9,11–13]. These studies have also noted the dependency linear coupled mathematical model of solid deformation
of volumetric strain of the coal matrix as a non-linear and gas seepage was presented and the methane extraction
function of gas pressure, driven by gas desorption. There is from fractured coal seam was simulated [21]. However, the
an approximately linear relationship between the sorption- constitutive relationships between stress and strain are
induced volumetric strain and the absorbed gas volume similar to conventional poroelastic mechanics in most of
[7,9,10]. This relation holds both during uptake, described the above simulations and the effect of sorption-induced
as sorption, and in discharge, described as desorption, of strain on matrix volumetric strain has not been taken into
gases from the surfaces of the coal matrix. Because of the account although experimental data have noted its
dual-porosity structure of coal seams (i.e., micro-porous significant impact—both on total volumetric strain of the
matrix and macro-porous cleat/fracture network), the coal seam, and the resulting feedback on permeability.
matrix represents the main reservoir for the gas, and the The gas flow in coal seams is a complex physical and
cleats the main fracture pathways. When the pore pressure chemical process coupling solid deformation, gas deso-
declines during methane production, methane desorbs rption and gas movement. Although the influence of
from the coal matrix and the desorbed gas flows through sorption-induced deformation on porosity, and on perme-
the cleats to the producing well. The decline of pore ability has been widely studied, how this in turn affects gas
pressure results in a concomitant increase in effective stress. flow within the seam is not well understood. This is partly
The increase in effective stress reduces the stress-sensitive because no coupled gas flow and sorption-induced coal
permeability of the cleat system. In contrast, the deso- deformation models are available for in situ stress
rption-induced shrinkage of the coal matrix widens the conditions. The primary motivation of this study is to
cleats and enhances permeability. The net change in investigate how sorption-induced coal matrix deformation
permeability accompanying gas production is thus con- affects the gas flow in a coal seam through developing such
trolled by the competitive effects of declining pore pressure a porosity-based model.
decreasing permeability, and the shrinking coal matrix
increasing permeability. The net effect, of permeability loss 2. Governing equations
or permeability gain, is dependent on the mechanical
boundary conditions applied to the system. In the following, a set of field equations are defined
Adsorption of gas, such as carbon dioxide, is the reverse of which govern the deformation of the solid matrix, and
desorption. When the gas pressure increases, the gas adsorbs prescribe the transport and interaction of gas flow in a
onto the coal matrix. The increase of pore pressure results in similar way to poroelastic theory [22]. These derivations
a decrease of effective stress. The reduction in effective stress are based on the following assumptions: (a) coal is a
enhances the coal permeability. In contrast, the adsorption- homogeneous, isotropic and elastic continuum. (b) Strains
induced swelling of the coal matrix reduces the cleat are much smaller than the length scale. (c) Gas contained
apertures and decreases the permeability. The net change within the pores is ideal, and its viscosity is constant under
in permeability accompanying gas sequestration is also isothermal conditions. (d) The rate of gas flow through the
controlled competitively by the influence of effective stresses coal is defined by Darcy’s law. (e) Conditions are
and matrix swelling, again controlled by the boundary isothermal. (f) Coal is saturated by gas. (g) Compositions
conditions applied locally between the end-members of null of the gas are not competitive, i.e., one gas component at
changes in either mean stress or volume strain. a time.
Numerical simulations of gas diffusion, gas flow, and
coupled hydromechanical response have been widely 2.1. Governing equation for coal seam deformation
applied. Finite element methods and a formulation for
modeling mass transport problems in porous media have For a homogeneous, isotropic and elastic medium, the
been applied, including the effects of coupled solid–gas strain-displacement relation is expressed as
response for gas flow in coal seams [14]. This included only
the effect of gas sorption on mass storage [15]. Valliappan ij ¼ 12ðui;j þ uj;i Þ, (1)
and Zhang developed a coupled model incorporating the where eij is the component of the total strain tensor and ui is
effect of diffusion of adsorbed methane gas from the solid the component of the displacement. The equilibrium
matrix to the voids [16]. A dual-porosity poroelastic model equation is defined as
was extended and utilized in solving generalized plane
sij;j þ f i ¼ 0, (2)
strain problems [17,18]. Gilman and Beckie proposed a
simplified model of methane diffusion and transport in a where sij denotes the component of the total stress tensor
coal seam and found a reference time of methane release and fi denotes the component of the body force. The
ARTICLE IN PRESS
1228 H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236
following conventions have been adopted in Eqs. (1) and where Mg is the molecular mass of the gas, R is
(2) and following related equations: a comma followed by the universal gas constant, and T is the absolute gas
subscripts denotes differentiation with respect to spatial temperature.
coordinates and repeated indices in the same monomial Assuming the effect of gravity is relatively small and can
imply summation over the range of the indices (generally be neglected, the Darcy velocity, qg, is given by
1–3, unless otherwise indicated) [23].
k
The gas sorption-induced strain es is presumed to result qg ¼ rp, (9)
in volumetric strain only. Its effects on all three normal m
components of strain are same. On the basis of poroelas- where k is the three permeability of the coal and m is the
ticity [23] and by making an analogy between thermal dynamic viscosity of the gas. Substituting Eqs. (7)–(9) into
contraction and matrix shrinkage [12], the constitutive Eq. (6), we obtain
relation for the deformed coal seam becomes
r p V L PL qp qf k
1 1 1 a s fþ c a þ p r prp ¼ Qs , (10)
ij ¼ sij skk dij þ pdij þ dij , (3) ðp þ PL Þ2 qt qt m
2G 6G 9K 3K 3
where pa is one atmosphere of pressure (101.325 kPa). In
where G ¼ E/2(1+n), K ¼ E/3(12n), a ¼ 1K/Ks, and Eq. (10), the permeability k is dependent on the porosity,
skk ¼ s11+s22+s33, where K is the bulk modulus of coal, f, while f is a function of the volumetric strain, en, and the
Ks is the bulk modulus of the coal grains, G is the shear sorption-induced strain, es. Therefore, Eqs. (5) and (10) are
modulus of coal, E is the Young’s modulus of coal, n is coupled through the porosity–permeability relation.
Possion’s ratio of coal, a is the Biot coefficient, dij is the
Kronecker delta, and p is the gas pressure within the pores.
From Eq. (3), we obtain 2.3. A general porosity model
1 The sorption-induced volumetric strain es is fitted onto
v ¼ ðs̄ apÞ þ s , (4)
K Langmuir-type curves and has been verified through
where ev ¼ e11+e22+e33 is the volumetric strain of coal experiments [7,9,10]. A Langmuir-type equation is used
matrix and s̄ ¼ skk =3 is the mean compressive stress. The to calculate this volumetric strain, defined as
component of effective stress seij is also defined as p
seij ¼ sij+apdij. Combination of Eqs. (1)–(3) yields the s ¼ L , (11)
PL þ p
Navier-type equation expressed as
where the Langmuir volumetric strain, eL, is a constant
G
Gui;kk þ uk;ki ap;i Ks;i þ f i ¼ 0. (5) representing the volumetric strain at infinite pore pressure
1 2n and the Langmuir pressure constant, PL, representing the
Eq. (5) is the governing equation for coal deformation, pore pressure at which the measured volumetric strain is
where the gas pressure, p, can be solved from the gas flow equal to 0.5eL. The authors in the above experiments also
equation as discussed following. assume that the coal permeability varies with porosity as
follows:
2.2. Governing equation for gas flow 3
k f
¼ , (12)
The equation for mass balance of the gas is defined as k0 f0
permeability are defined as [2] where se denotes effective stress. Or under invariant
total stress.
f L 2 p0 p
¼1þ 1þ , (13)
f0 3 f0 P L þ p0 P L þ p dse ¼ dp. (22)
3 According to the principle of effective stress, the induced
k L 2 p0 p coal deformation is determined by the change in effective
¼ 1þ 1þ . (14)
k0 3 f0 pL þ p0 pL þ p stress, dse, which can be replaced by the change in pore
pressure, dp, under the assumption of null change in total
stress. This is why terms representing effective stress or
2.3.1.2. Palmer–Mansoori model. Unlike the Seidle– total stress are absent in all of these existing permeability
Huitt model, the Palmer–Mansoori model considers the models. However, this condition can be violated in a
elastic deformation of coal under uniaxial stress conditions. number of circumstances, including the penetration by drill
The porosity model is defined as [11] holes, the massive stimulation by the injection of fluids,
and the proximity of excavation surfaces. These factors
f 1 L K p p0
¼1þ ðp p0 Þ þ 1 , result in the re-distribution of total stresses in coal seams.
f0 Mf0 3f0 M PL þ p PL þ p0 Therefore, a new porosity and permeability model under
(15) variable stress conditions is needed to quantify the gas flow
3 in coal seams.
k 1 L K p p0
¼ 1 ðp p0 Þ þ 1 ,
k0 Mf0 f0 M PL þ p PL þ p 0
2.3.2. The general porosity model
(16) Considering a porous medium containing solid volume
where M ¼ E(1n)/(1+n)(12n). of Vs and pore volume of Vp, we assume the bulk
volume V ¼ Vp+Vs and the porosity f ¼ Vp/V. According
2.3.1.3. Shi–Durucan model. The assumptions are same to Eq. (4), the volumetric evolution of the porous medium
as the Palmer–Mansoori model [12]: with the load of s̄ and p can be described in terms of DV/V
and DVp/Vp, the volumetric strain of coal and volumetric
k n L E
¼ exp 3cf ðp p0 Þ þ strain of pore space, respectively [23]. The relations are
k0 1n 3 1n
DV 1
p0 p ¼ ðs̄ apÞ þ s , (23)
, ð17Þ V K
PL þ p0 PL þ p
where cf is cleat volume compressibility. DV p 1
¼ ðs̄ bpÞ þ s , (24)
Vp Kp
2.3.1.4. Cui–Bustin model. This model has a general form where b ¼ 1Kp/Ks.
but it is only applied to the same assumed situation as the We assume that the sorption-induced strain for the coal
Palmer–Mansoori model [9]: is the same as for the pore space. Without the gas sor-
f 1 1 ption effect, the volumetric variation of the porous med-
¼ exp ½ðs s0 Þ ðp p0 Þ , (18) ium satisfies the Betti–Maxwell reciprocal theorem [23],
f0 K Kp
ðqV =qpÞs̄ ¼ ðqV p =qs̄Þp , and we obtain
k 3 f
¼ exp ½ðs s0 Þ ðp p0 Þ , (19) Kp ¼ K. (25)
k0 Kp a
where Kp is the bulk modulus of pores. Using the definition of porosity, the following expressions
can be deduced as
2.3.1.5. Robertson–Christiansen model. In this model, the
DV DV s Df
deformation of coal grains is neglected and equal axial and ¼ þ , (26)
radial stresses are assumed [13]. V Vs 1f
k 1 exp½ac ðp p0 Þ 9 1 2n DV p DV s Df
¼ exp 3c0 þ ðp p0 Þ ¼ þ . (27)
k0 a f E Vp Vs fð1 fÞ
0
L PL PL þ p Solving Eqs. (23), (24), (26) and (27), we obtain
In , ð20Þ
3 PL þ p0 P L þ p0
1 1
where c0 is the initial fracture compressibility and ac is the Df ¼ f ðs̄ pÞ. (28)
K Kp
change rate in fracture compressibility.
In all these models, the total stress, s, has been assumed Substituting Eqs. (4) and (25) into Eq. (28) yields
as constant, i.e.,
p
Df ¼ ða fÞ v þ s . (29)
ds ¼ dse dp ¼ 0 ða ¼ 1Þ, (21) Ks
ARTICLE IN PRESS
1230 H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236
0.1 m
non-linear partial differential equations (PDE) of second
order in space and first order in time. The non-linearity
appears both in the space and time domains; and therefore,
these equations are difficult to solve analytically. There-
fore, the complete set of coupled equations is implemented
into, and solved by using COMSOL Multiphysics, a 0.1 m
powerful PDE-based multiphysics modeling environment.
4. Simulation examples
100.0 1.6
Matrix Volumetric
80.0
Strain 1.5
Contribution to Porosity Ratio (%)
Grains Volumetric
60.0
Strain
-20.0 1.1
-40.0
1.0
-60.0
0.9
-80.0 0.00 0.02 0.04 0.06 0.08 0.10
0.0 1.0 2.0 3.0 4.0 5.0 6.0 Gas Flow Distance (m)
Pore Pressure (MPa)
Fig. 4. Example I: spatial and temporal distributions of permeability ratio
Fig. 2. Example I: contributions of different volumetric strain compo- (k/k0).
nents to the coal porosity ratio.
2.2
100
90 2.0
Contribution to Storage Coefficient (%)
80 K/Ks=1/3
1.8
Permeability Ratio (k/k0)
K/Ks=1/10
70
K/Ks=0
term 1 1.6
60 Without Desorption
term 2
Palmer-Mansoori
50 term 3
1.4
term 4
40
1.2
30
20 1.0
10 0.8
0
0.6
-10 0.0 1.0 2.0 3.0 4.0 5.0 6.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pore Pressure (MPa)
Pore Pressure (MPa)
Fig. 5. Example I: impacts of different bulk modulus ratio (K/Ks) and gas
Fig. 3. Example I: contributions of four terms in gas storage coefficient, desorption on permeability ratio (k/k0).
i.e., the volume occupied by the free-phase gas (term 1), the volume
occupied by the adsorbed gas phase (term 2), the mechanical deformation-
induced pore volume change (term 3), and the sorption deformation-
induced pore volume change (term 4). tion capability and 9.2–3.0% is from free gas storage
capacity. The contributions from the remaining two
sources are below 4.5%, and can be neglected.
volumetric strain is more significant than the bulk
mechanical volumetric strain. The contribution from the
grain volumetric strain is not obvious. 4.1.3. Permeability evolution
The spatial and temporal variations of the permeability
4.1.2. Gas storage ratios are shown in Fig. 4. As the pore pressure declines,
As shown in Eq. (37), the storage term consists of four the permeability ratio increases with time. The permeability
contributions: from free gas, from gas absorption, from in the area near the right edge changes more rapidly
coal deformation, and from the combined effect of than the area far from this edge because the pressure
sorption and deformation. Contributions from different gradient close to right edge is far greater. The steady state
sources are presented in Fig. 3. As pore pressure depletes, is reached at about 20,000 s (5 h) when the pressure is
94.6–97.1% of the storage coefficient is from gas absorp- equal to 1 atm.
ARTICLE IN PRESS
H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236 1233
6.0
t=0 s
5.0 K/Ks=1/3, 100 s
K/Ks=1/3, 20000 s
Without Desorption, 100 s
K/Ks=1/3, 1 s t=10 s
4.0
Pore Pressure (MPa)
K/Ks=1/3, 1000 s
Without Desorption, 1 s
Without Desorption, 800 s
t=100 s
3.0
2.0 t=20000 s
1.0
K/Ks=0, 10000 s
K/Ks=0, 100 s
K/Ks=1/3, 100 s istics of the gas desorption from coal as evident in the
4.0
K/Ks=1/10, 1 s experiments [7,10]. One of these characteristics is that the
K/Ks=1/10, 12000 s
gas sorption-induced volumetric strain plays an important
K/Ks=0, 1000 s
3.0 role in the variation of coal porosity and permeability. Our
model results show that under the conditions of uniaxial
2.0 strain and constant overburden load, the influence from the
coal grain deformation can be neglected when estimating
volumetric strain, porosity and pressure evolution. How-
1.0
ever, the bulk modulus of grains may not be simply treated
as infinite when calculating permeability because the cubic
0.0 relationship between porosity and permeability. A slight
0.00 0.02 0.04 0.06 0.08 0.10
change in porosity may result in a much larger change in
Gas Flow Distance (m)
permeability. If gas desorption is not considered, the
Fig. 7. Example I: impact of different bulk modulus ratio (K/Ks) on the permeability may decrease linearly due to the increasing
pore pressure distribution. effective stress. Otherwise, gas desorption may increase the
permeability and the net change in permeability is
4.1.4. Impact of modulus ratios on permeability controlled by the opposite effects from effective stress
As shown in Fig. 5, the permeability ratio (k/k0) and gas desorption. The pore pressure may be under-
increases due to gas desorption when pore pressure estimated dramatically if the effect of gas desorption is
decreases. When the ratio of bulk modulus (K/Ks) changes neglected. Although the permeability may decrease without
from 1/3 to 0, the highest permeability ratio varies from gas desorption and cause gas to flow slowly, the storage
1.55 to 2.03. When the bulk modulus of coal grains (Ks) is coefficient becomes much smaller and pore pressure
assumed to be infinite, the simulation result is identical decreases very quickly. Therefore, the gas absorption
with the data calculated by the Palmer–Mansoori model. If capability dominates the storage coefficient.
the gas desorption is neglected, the permeability ratio
drops linearly to 0.66. 4.2. Example II: gas desorption under constrained plane
strain
4.1.5. Gas pressure distributions
The spatial and temporal variations of the gas pressure The new coupled model is applicable to variable loading
are shown in Figs. 6 and 7. These results demonstrate that conditions. The following example is used to simulate the
ARTICLE IN PRESS
1234 H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236
2.2
K/Ks=1/3
2.0
K/Ks=1/10
K/Ks=0
1.8 Without Desorption
1.4
1.2
0.1 m
p(0) = 6.2 MPa (Ω)
1.0
0.8
0.6
0.1 m 0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pore Pressure (MPa)
Fig. 10. Example II: impacts of different bulk modulus ratio (K/Ks) and
gas desorption on permeability ratio (k/k0).
t=0 s
coal sample under the constrained plane strain state as
shown in Fig. 9. t=10 s
The same geometrical model and the same material
properties (Table 1) are applied as previous. The upper and t=100 s
right sides are unconstrained. The displacements at the left
and bottom sides are constrained in the horizontal and t=4000 s
vertical directions, respectively. A distributed overburden
load of 6.9 MPa is applied on the upper side and remains
unchanged during gas desorption. The initial gas pore
pressure in the coal is 6.2 MPa and the pressure on the right
side remains at 101.325 kPa. Zero fluxes are specified on the
other three boundaries. Fig. 11. Example II: evolution of the coal sample configuration with gas
The relations between gas pressure and permeability desorption.
ratio under different conditions are shown in Fig. 10. As
the pore pressure decreases, the permeability decreases the Palmer–Mansoori model is much greater than our
first, then increases after the pressure comes to a critical model results.
value of about 1 MPa. The permeability ratio (k/k0) varies
not significantly when the ratio of bulk modulus (K/Ks) 4.3. Example III: gas desorption under the unconstrained
changes from 1/3 to 0. The biggest relative difference is plane strain
6.7%. However, the maximum permeability ratios calcu-
lated by using the Palmer–Mansoori model are significantly In this example, we change the left boundary of Example
higher. The relative error is 141–153%. If the gas I from the constrained horizontal displacement condition
desorption is neglected, the permeability ratio decreases to an unconstrained condition, as illustrated in Fig. 12. All
to 0.59 linearly. As shown in Fig. 11, the coal shrinks in other conditions remain the same as Example II.
both directions when the gas pressure declines. The relations between gas pressure and permeability
The impact of the different boundary conditions is ratio under different conditions are shown in Fig. 13. If gas
significant in this example although the property para- desorption is included, the permeability decreases initially
meters are the same as those in the Example I. The when pore pressure declines. Then it rebounds at the
overburden loading and plane strain conditions increase pressure of about 3 MPa and increases to 0.91. The
the effective stress and reduce the permeability in contrast influence of the bulk modulus ratio (K/Ks ¼ 1/3, 1/10, 0)
to the uniaxially loaded condition in the first example. This is not significant. If gas desorption is not included, the
difference is the reason why the permeability calculated by permeability ratio decreases linearly with decreasing pore
ARTICLE IN PRESS
H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236 1235
t=0 s
t=10 s
t=100 s
0.1 m
p(0) = 6.2 MPa (Ω)
t=5000 s
0.1 m Fig. 14. Example III: evolution of the coal sample configuration with gas
desorption.
2.2 5. Conclusions
K/Ks=1/3
2.0
K/Ks=1/10 In this study, a new coupled gas flow and sorption-
K/Ks=0 induced coal deformation finite element model is developed
1.8
Permeability Ratio (k/k0)
Without Desorption
Palmer-Mansoori to quantify the net change in permeability, the gas flow,
1.6 and the resultant deformation of the coal seam. The
coupling between gas flow and coal deformation is realized
1.4 through a general porosity and permeability model. The
general porosity model considers the principal controlling
1.2
factors, including the volume occupied by the free-phase
1.0 gas, the volume occupied by the adsorbed phase gas, the
mechanical deformation-induced pore volume change, and
0.8 the sorption-induced coal pore volume change. More
importantly, these factors are quantified under in situ
0.6 stress conditions. A cubic relation between coal porosity
0.0 1.0 2.0 3.0 4.0 5.0 6.0
and permeability is introduced to relate the coal storage
Pore Pressure (MPa)
capability (changing porosity) to the coal transport
Fig. 13. Example III: impacts of different bulk modulus ratio (K/Ks) and property (changing permeability). The general porosity
gas desorption on permeability ratio (k/k0). and permeability model is then implemented into the new
coupled gas flow and coal deformation finite element
pressure. If the Palmer–Mansoori model is applied, the model.
permeability ratio increases throughout and the maximum The FE model has been applied to compare the
deviation from the current model is about 121%. As shown performance of the new model with that of the Palmer–
in Fig. 14, the coal shrinks both in the horizontal direction Mansoori model. It is found that the Palmer–Mansoori
and in the vertical direction simultaneously when the gas model may produce significant errors if loading conditions
pressure declines. deviate from the assumptions of the uniaxial strain
This example is analogous to Example II and also condition and that of infinite bulk modulus of the coal
demonstrates that the loading and boundary conditions grain. The FE model has also been applied to conduct a
have significant impact on model results. The permeability number of simulation examples. The model results have
of our model is much less than that of the Palmer– revealed the characteristics of the gas desorption from coal
Mansoori model. The difference indicates that the influence as evident in the experiments reported in the literature.
of effective stress is underestimated due to the assumptions When the pore pressure declines during gas desorption,
in the Palmer–Mansoori model. Only one direction of the net change in permeability accompanying the gas
ARTICLE IN PRESS
1236 H. Zhang et al. / International Journal of Rock Mechanics & Mining Sciences 45 (2008) 1226–1236
production is controlled competitively by the effects of [9] Cui X, Bustin RM. Volumetric strain associated with methane
declining pore pressure and increasing shrinkage of coal. desorption and its impact on coalbed gas production from deep coal
seams. AAPG Bull 2005;89:1181–202.
[10] Robertson EP, Christiansen RL. Modeling permeability in coal using
Acknowledgment sorption—induced strain data. In: Proceedings of the 2005 SPE
annual technical conference and exhibition, Dallas, 9–12 October
This work is supported by the Australia Research 2005, paper SPE 97068.
[11] Palmer I, Mansoori J. How permeability depends on stress and pore
Council under Grants DP0342446 and DP0209425 and
pressure in coalbeds: a new model. Soc Petrol Eng Res Eng 1998;12:
by the Australian Department of Education, Science and 539–44.
Training through the Australia-China Special Fund. This [12] Shi JQ, Durucan S. Drawdown induced changes in permeability of
support is gratefully appreciated. coalbeds: a new interpretation of the reservoir response to primary
recovery. Transp Porous Media 2004;56:1–16.
[13] Robertson EP, Christiansen RL. A permeability model for coal and
References other fractured, sorptive-elastic media. In: Proceedings of the SPE
eastern regional meeting, Canton, OH, 11–13 October 2006, paper
[1] Saghafi A, Faiz M, Roberts D. CO2 storage and gas diffusivity SPE 104380.
properties of coals from Sydney Basin, Australia. Int J Coal Geol [14] Zhao C, Xu T, Valliappan S. Numerical modeling of mass transport
2007;70:240–54. problems in porous media: a review. Comput Struct 1994;53:849–60.
[2] Seidle JP, Jeansonne MW, Erickson DJ. Application of matchstick [15] Zhao Y, Jin Z, Sun J. Mathematical model for coupled solid
geometry to stress dependent permeability in coals. In: Proceedings of deformation and methane. Appl Math Modell 1994;18:328–33.
the SPE rocky mountain regional meeting, Casper, Wyoming, 15–21 [16] Valliappan S, Zhang W. Numerical modeling of methane gas
May 1992, paper SPE 24361. migration in dry coal seams. Int J Numer Anal Methods Geomech
[3] Gary I. Reservoir engineering in coal seams: Part 1—The physical 1996;20:571–93.
process of gas storage and movement in coal seams. Soc Petrol Eng [17] Bai M, Meng F, Elsworth D, Abousleiman Y, Roegiers JC.
Res Eng 1987;1:28–34. Numerical modeling of coupled flow and deformation in fractured
[4] Harpalani S, Chen G. Influence of gas production induced volumetric rock specimens. Int J Numer Anal Methods Geomech 1999;23:
strain on permeability of coal. Geotech Geol Eng 1997;15:303–25. 141–60.
[5] Clarkson CR, Bustin RM. The effect of pore structure and gas [18] Bai M, Abousleiman Y, Cui L, Zhang J. Dual-porosity poroelastic
pressure upon the transport properties of coal: a laboratory and modeling of generalized plane strain. Int J Rock Mech Min Sci
modeling study. 1. Isotherms and pore volume distributions. Fuel 1999;36:1087–94.
1999;78:1333–44. [19] Gilman A, Beckie R. Flow of coal-bed methane to a gallery. Transp
[6] Faiz M, Saghafi A, Sherwood N, Wang I. The influence of Porous Media 2000;41:1–16.
petrological properties and burial history on coal seam methane [20] Rutqvist J, Tsang C. A study of caprock hydromechanical changes
reservoir characterisation, Sydney Basin, Australia. Int J Coal Geol associated with CO2-injection into a brine formation. Environ Geol
2007;70:193–208. 2002;42:296–305.
[7] Harpalani S, Schraufnagel A. Measurement of parameters impacting [21] Zhao Y, Hu Y, Zhao B, Yang D. Nonlinear coupled mathematical
methane recovery from coal seams. Int J Min Geol Eng 1990;8: model for solid deformation and gas seepage in fractured media.
369–84. Transp Porous Media 2004;55:119–36.
[8] Seidle JP, Huitt LG. Experimental measurement of coal matrix [22] Biot MA. General theory of three-dimensional consolidation. J Appl
shrinkage due to gas desorption and implications for cleat perme- Phys 1941;12:155–64.
ability increases. In: Proceedings of the international meeting on [23] Detournay E, Cheng AHD. Fundamentals of poroelasticity. In:
petroleum engineering, Beijing, 14–17 November 1995, paper SPE Fairhurst C, editor. Comprehensive rock engineering, vol. 2. Oxford:
30010. Pergamon; 1993. p. 113–71.