Beruflich Dokumente
Kultur Dokumente
6. Physical Properties of Fractured Rocks
By: Sadeghnejad
sadeghnejad@modares.ac.ir
Clastic Rocks
Consist Primarily of Silicate Minerals
- Grain Size
- Mineral Composition
Carbonate Rocks
Consist Primarily of Carbonate Minerals
(i.e. Minerals With a CO3-2 Anion Group)
Limestone - Predominately Calcite (Calcium
Carbonate, CaCO3)
Dolomite - Predominately Dolostone (Calcium
Magnesium Carbonate, CaMg(CO3)2 )
1
3/3/2018
Relative Abundances of Sedimentary Rock Types
and conglomerate
(clastic)
~11%
Limestone and
dolomite
Siltstone ~14%
and shale
(clastic)
~75%
Compositions of Clastic & Carbonate Rocks
Ayers (2001)
Clastic Rocks Carbonate Rocks
Fossils
Sand Quartz Allochemical Pelloids
Feldspar Grains Oolites
Grains Intractlasts
Rock Fragments
Average Average
Sandstone Sparry
Average Limestone
Mudrock Average
(Shale) Micritic
Limestone
2
3/3/2018
Dual Porosity in Sandstone
Sandstone Comp. 1. Primary matrix porosity
• Framework 2. Secondary porosity (Fracture)
• Matrix
• Cement FRACTURE DISSOLUTION
• Pores PORE
PORE
FRAMEWORK
(QUARTZ)
CEMENT
MATRIX
FRAMEWORK
(FELDSPAR)
Framework
0.25 mm
Ayers (2001)
Sandstone Composition, Framework Grains
Ayers (2001)
KF = Potassium
Feldspar
PRF = Plutonic Rock
Q Fragment
Q = Quartz
PRF KF
P = Pore
Potassium Feldspar is
Stained Yellow With a
Chemical Dye
P
Pores are Impregnated With
Blue-Dyed Epoxy
Norphlet Sandstone, Offshore Alabama, USA
Grains ~0.25 mm in Diameter/Length Photo by R. Kugler
3
3/3/2018
Intergranular Pore and Microporosity
Ayers (2001)
Intergranular
Pore
Intergranular Pores
Microporosity Contain Hydrocarbon
Fluids
Quartz Micropores Contain
Kaolinite Detrital
Grain Irreducible Water
4
3/3/2018
Carbonate Porosity‐example
Ayers (2001)
Moldic and
Interparticle Pores
Interparticle
Pores • Combination pore system
• Connected pores
Double Porosity Definition
matrix void vol
1 total bulk vol
t 1 2
fracture void vol
2 total bulk vol
1 2 , 2 1%
There are also alternative definitions i.e. matrix porosity:
matrix void vol fracture void vol
m f , f 2
matrix bulk vol matrix bulk vol
Then:
1 1 2 m
Effective primary porosity: 1 1 2 m 1 S wi
10
5
3/3/2018
Quantitative Evaluation of φ
The total porosity can be evaluated from direct (cores) or
indirect (well log) methods
But, it is difficult to distinguish the primary porosity from
the secondary one
Depending on type of rock & stress state, either the
macrofractures or the microfractures will be more
predominant:
11
Quantitative Evaluation
Empirical Evaluation of φ
The following empirical correlations also relate the maximum
secondary porosity to the total porosity:
The accuracy of the evaluation of φf is of very limited
importance, since it is generally negligible compared with
matrix porosity
When φt <5%, it is important to evaluate φf
12
6
3/3/2018
Quantitative Evaluation
Φf from Core Analysis
Core analysis gives both the fracture
opening and volumetric fracture density:
S
V fD
VB
V oid V olume S * b
f V fD b
Bulk V olume VB
As the fracture opening is irregular, in
practice, an averaged value of opening b is
used
For example
fracture density 40 m2/m3
average fracture opening of 0.08mm
porosity is: Mokhtari et al. 2017
Quantitative Evaluation
Φf from Thin Section
The uncertainty in the measurement increases when the sample size
is smaller than the distance between the two adjacent fractures
b m cos b
θ
h
cos
b m cos a cos
h
b
for a single fracture: f m bm a
Lah L S
n
nb a l b m ,i l T ,i
for a fracture parallel to each other: f m b m A fD b m T i 1
n
S
S S
i
i 1
n
b m ,i l T ,i
for n thin sec tio ns : f i 1
n
S
i 1
i
w here , a : ex tension of a single fracture , lT : total length of fractures , S aL ,T hin S ection S urface 14
7
3/3/2018
Quantitative Evaluation
Φf from Structural Data
From the structural geological analysis relating the
curvature to bed thickness, H (Murray 1977)
During structural folding, fractures normal to the bedding
may result:
2Z
f 0.5H
x 2
Example: highest reservoir curvature in the range of 0.5
to 1*10‐5 1/ft and H=20ft
f min 20 0.5 105 0.1%
f max 20 105 0.2% 15
Quantitative Evaluation
Φf from Structural Data
(R H ) R H 2RH H
2
Bulk Volume :V B
H H
f
2 2
(R H ) R H H 2 2R H 2R
Fracture PoreVolume :V pf
2 2
1
R 2
d z 2
dx
2 Z
f 0.5H
x 2 16
8
3/3/2018
Quantitative Evaluation
Φf from Triaxial Core Test
By reducing the pore pressure, a steep change in the
porosity occurs (first fractures closes) until the point
where the matrix pores are compacted
One can then use extrapolation to have an estimate of
the fracture porosity
Fracture Matrix
o Compaction Compaction
o
reduction in porosity
6
a
4
0
a, f 0
0 2
17
Intrinsic Fracture Permeability
The Intrinsic fracture permeability, Kff is a function of the
fracture width and spacing, when flow passes through
one fracture independent of surrounding matrix
Kff is independent of the matrix permeability
From Navier‐Stocks Eq.:
b 2 P b3 P
for fracture # 1: q ab a
12 L 12 L
b cos
2
P
for fracture # 2: q ab
12 L
By comparing this with the Darcy law:
k ff P b2
q A (Darcy) k ff cos 2
L 12
18
9
3/3/2018
Intrinsic Fracture Permeability
For example: the intrinsic permeability for b=0.05mm is
200 darcy (really a fantastic reservoir )
This can simply be extended to a fracture system or a
fracture network as:
cos2
Parallel group of fractures: k ff
12
b
i
i
2
1 2
Fracture network fromed by fracture systems k ff cos bi cos b i ...
2 2 2
12 i i
19
Conventional Fracture Permeability
Conventional fracture permeability used when both fracture
and matrix form hydrodynamic units
The bulk volume associated to the fractures is considered (ab
is replaced by ah)
k f P ab b b3
q ah (Darcy law) k f k ff k ff
L ah h 12h
Permeability of a fracture‐matrix system (i.e. total
permeability) which depends on the direction of the flow is:
Kt K m K f
20
10
3/3/2018
Notes
Intrinsic/conventional permeability from fracture width is
a very simplistic approach
It takes into account neither the fracture roughness nor
the possible filling inside it
In reality, fracture conductivity might not obey this simple
law.
Sometimes, no reliable measurement of fracture
apertures is available at down hole condition
21
Quantitative Evaluation
K from Core Analysis
Conventional method using a cylindrical core can be used to
get the total permeability
It is difficult to distinguish the fracture and matrix
permeability (only special cases)
Kt depends on the fracture orientation
Oriented Fracture Random Orientation of Fractures
22
11
3/3/2018
Quantitative Evaluation
K from Core Analysis
The Kelton Permeameter:
An essential characteristic is the change in flowing direction
Two K are measured:
Kmax (flow crosses through principal fracture)
K90 (core turned 90°)
23
Quantitative Evaluation
K from Core Analysis
The Kelton Permeameter:
Q
K
p l g ,eff
l g ,eff l g LSF TSF
l g ,eff : Equivalant Gasket Length
LSF : Longitudinal Shape Factor
TSF :Transversal Shape Factor
24
12
3/3/2018
Quantitative Evaluation
K from Core Analysis
Hasler Permeameter:
Measures vertical & horizontal permeability
25
Quantitative Evaluation
K from Core Analysis
Core length
If fracture density is low, K may represent only matrix in laboratories
Fracture Orientation:
Fracture orientation vs. flowing direction substantially changes the results
The radial flow toward the well will have very little in common with the linear
flow in the laboratory
This explains why Kwelltest≈10 to 100 Klaboratory
Reservoir Condition
Making laboratory measurements in the absence of pore pressure, reduces
the fracture permeability values
Mokhtari et al. 2017 26
13
3/3/2018
Quantitative Evaluation
K from Well Test
Simplified models which idealized the flow in a fractured
system are used to estimate the permeability
Km, Kf and Kt can be estimated
The evaluation of K is independent of any knowledge
concerning fracture parameters (opening, Orientation,
etc)
sugar cube
fracture model
Matrix feed fracture, flow only
27
through fractures
Quantitative Evaluation
K & φ from Well Test
For example (Khanin 1969)
r
B ln e S
SS flowing condition : k f PI rw
2 hA
for multi fracture random distribution f 3 29.6K f A fD2
1/ 3
r
B ln e S
1 rw 2
for Oil wells : f PI A fD
557.9 h
f : fraction , PI : m / d / atm , : cP , h : m , A fD :1/ cm
3
28
14
3/3/2018
Quantitative Evaluation
K from Simplified models
In the case of single fracture
For F1: aLb b
f
aLh h
b b3 b2
K f K ff f
h 12h 12
Therefore, porosity & fracture opening can be evaluated
as:
b 12k f
f 2
h b
b 12k f / f
29
Quantitative Evaluation
K from Simplified models
In the case of multiple fractures
L fD 1/ e
K ff P b 3 P nab 3 b3 b2
flow through n fractures: q n ab na Kf AfD f or
L 12 L S ah 12S 12 12
K P
3 3
K nb L b A bK
Darcy law: q ah f
L f
12h
fD
12
fD ff
na lt na n
AfD LfD
S S ah h
nabL nb
f AfDb LfDb
SL h
30
15
3/3/2018
Quantitative Evaluation
K from Simplified models
31
Quantitative Evaluation
K from Simplified models
Other simplified geometries
32
16
3/3/2018
Quantitative Evaluation
K from Simplified models
33
Quantitative Evaluation
K from Simplified models
Example: Cube model (5)
n l 2a 2
A fD
Surface a 2 a
b2
K f f
12 (f a / 2) 2 1 2 3
K f f a f
(a a ) b 2a 2b 12 48
f
2a 3 a
in dimensional unit :
( K : darcy , a : cm , b : micron , : percenct )
K f 2.08a 2f3
K f 8.33*104 b 2f
34
17
3/3/2018
Quantitative Evaluation
Well test & simplified model
Calculation Example: (P 182‐183, Van Golf‐Racht)
35
Quantitative Evaluation
K from Thin Section
b2
for a single fracture: K f f
12
Therefore, it is sufficient to measure fracture opening, no of
fractures and their lengths (for a fracture system)
In the case of random distribution of fracture orientation in a thin
section (Romm 1966)
4b m3 l t b m3 l t
K f C1 C
2S
2
S
S : areaof the thin sec tion
For n thin sections:
n
b 3
l
m ,i t .i
Kf C2 i 1
n
S
i 1
i
36
18
3/3/2018
Quantitative Evaluation
K from Structural Data
2Z
f 0.5H
x 2
f3 3
Kf 1 2 2Z
12A fD2 K e H
192 x 2
f
H H
A fD
S H e
b is avariable (figure )
As b is a variable, K should be
considered as an average equivalent
of half of what result from direct
substitution of the above equations
3
2Z 2Z
K f 5*108 e 2 H 2 (K : md , e : cm , H & 2 the same unit )
x x 37
Quantitative Evaluation
Abgrall 1971 K from Triaxial Core Test
Non‐fractured Limestone fractured Limestone
After a rapid linear reduction (100 After a rapid linear reduction (100
bar) a slow parabolic reduction bar) a slow parabolic reduction
was observed was observed (400bar) where K=0
In decompression K is lower than During decompression,
compression deformation is permanent
K returns to original K at σ‐p→0 Kfinal=25% Kinitial
38
19
3/3/2018
K & φ Relationship
Conventional Reservoirs Fractured Reservoirs
Φm , Km
relationship
Open Fracture
Zone
Closed
Fracture/Stylolite
log K m am b
39
Rock Compressibility
The general definition of rock compressibility (normal
formation):
1 dV B
Bulk: cB
V B d
1 dV r
Rock (matrix): cr
V r d
1 dV p
Pore: cp
V p d
where: effective stress P , V B V r V p andV p V B
1
Cp
In fractured reservoirs we have fracture porosity (f),
cavern (c), and vugs (v) as well as the matrix porosity (m)
c r c m c cc f c f v cv
cv 3c m 40
20
3/3/2018
Relative Permeability
Validity of X‐Shaped Kr
Firoozabadi & Hauge (1990) found Pc= 275 kPa in the fractures, clearly showing
that the assumption of zero capillary pressure not necessarily is correct
Firoozabadi et al. (1990) through an experimental study showed that Pc in
fractures might have a similar form as the matrix
Porte et al. (2005) showed that Pc in fracture could change the final recovery.
(e.g. For a gas‐oil case, the recovery could be increased by a factor two if Pc in
fracture is present)
The fracture capillary pressure is only valid for narrow fractures (<100 µm), for
fractures wider than 100 microns, zero capillary pressure could be used
42
21
3/3/2018
Kr of Fractured Rocks
The straight‐line behavior assumes that there is an ideal
flow of fluids in fractures (phases move past each other
with no resistance inside the fracture)
However various flow regimes in a fracture may result in
different behavior
43
Kr of a Layered Formation
K=100mD
K=10mD
Case 1 Case 2 Case 3 Case 4
The discontinuity of Krg for parallel arrangement is the result of the
saturation in which the low permeable layer starts to produce fluid 44
22
3/3/2018
Kr of a Layered Formation
Corey & Rathjens 1956
K=100mD
K=10mD
Case 1 Case 2
Case 3 Case 4
45
Capillary Pressure vs. Pore Size Distribution
46
23
3/3/2018
Capillary Pressure of Fractured Rocks
In fractures, capillary forces may be ignored in
comparison with gravitational forces
Hence, the GOC/WOC in a fracture system will be in a
very distinct horizontal plane over the entire reservoir
(transition zone disappeared)
47
Fractured Reservoir Initial Condition
In fractured reservoirs, high water saturation zones are unrelated to WOC
If fracturing in a fractured reservoir took place before oil migration, it is
possible to find such a high water zones (A,B,C,D‐small height blocks)
Any correlation of WOC as an effect of the transition zone is worthless
During the migration the capillary threshold forces opposing the entering
non‐wetting phase (i.e. oil)
Pc
Pth hBlock
Pth
hblock
0
Sw 48
24
3/3/2018
Naturally Fractured Vuggy Reservoirs (NFVR)
Vugs affect flow & storage.
Fractures network generally contributes
< 1% of porous volume while Vuggy ϕ
can be high
Vug network normally has good vertical
permeability
Wu et al. 2012
49
http://www.wangsgroup.net
Dissolution Scale in Formation?
50
25
3/3/2018
Carbonate and Sandstone
Dissolved bioclast (G5), marix (E10) Vuggy porosity and partially
and cement (A7) dissolved quartz grains (D8 and G4)
51
Vug Connectivity
Vug size, orientation, connectivity, and distribution are
caused by deposit environment and diagenetic processes
they are difficult to characterize.
Matrix Porosity Isolated Vug
Touching
Vugs
Separate and touching vugs (H6,
Zhang et al., 2007 J6), moldic (F4) and intra‐particle
(G7; USV‐4)
52
26
3/3/2018
NFVR
Increasing ϕ and k may be due to directly connected vugs
and vugs connected through their halos. Vuggy kv may be
> fracture kf.
Phi=24.4 %
K=11.5 md
Phi=25.8 %
K=84.4 md
53
Kr of Vuggy limestone
Abgrall & Iffly 1973
Sgc may become very high
especially when the pressure
decline rate is small
Gas stays in the vugs without
moving until a large amount of the
oil leaves them out
Librated gas in the vugs can not
displace oil in the matrix due to
the Pc effects
54
27
3/3/2018
Vug Modeling by Digital Rock Physics
Zhang 2009 SPE
Sample Scanning Network Extraction
Pore Network Modeling
CCAL & SCAL (Image Analysis)
Vug Connectivity modeling by
56
28
3/3/2018
Developed Algorithm
57
Results
58
29