Affiliates 2006 Hill
Affiliates 2006 Hill
Affiliates 2006 Hill
Heterogeneous Media
Dr. A. D. Hill
Robert L. Whiting Chair in
Petroleum Engineering
Texas A&M University
Acid Stimulation
Matrix acidizing (below frac pressure)
Inject relatively small volumes of acid
Objective is to overcome damage effects
Acid only affects the formation for a few
inches to a few feet
Acid Fracturing
Create a hydraulically induced fracture,
just like proppant fracturing
Conductivity is created by acid etching the
faces of the fracture in a non-uniform way
Acid Stimulation
Matrix acidizing (below frac pressure)
Inject relatively small volumes of acid
Objective is to overcome damage effects
Acid only affects the formation for a few
inches to a few feet
Acid Fracturing
Create a hydraulically induced fracture,
just like proppant fracturing
Conductivity is created by acid etching the
faces of the fracture in a non-uniform way
Is it important?
Yes at all scales
CSM Meeting Oct. 17, 2006
Matrix stimulation of
sandstones
Creation of conductivity in
acid fractures
Matrix Acidizing of
Sandstone
Complex chemical interactions between HF/HCl and
many minerals
Acid typically penetrates only a few inches into the
formation
Propagation of acid through the damaged zone is
critical to success
Models generally treat the formation as 2
homogeneous regions in series damaged zone and
undamaged region
Small scale heterogeneity can change acid
penetration by an order of magnitude
Property Initialization
Property Initialization
Basic Case
HF Front Propagation
(Homogeneous porosity and mineral)
A: 5 PV
PV
C: 25 PV
B: 15 PV
D: 35 PV
Basic Case
A: 5 PV
PV
C: 25 PV
B: 15 PV
D: 35 PV
Heterogeneous Porosity
A: 5 PV
C: 25 PV
B: 15 PV
D: 35 PV
Heterogeneous Porosity
Sandstone
core
small-scale heterogeneity
Comparison
ofwith
Breakthrough
Time (PVbt) in
Cores with Different
Heterogeneities
35
33
31
29
27
25
23
21
19
17
15
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Correlated Porosity
A: 5 PV
B: 15 PV
C: 25 PV
D: 35 PV
Correlated Porosity
3.5
PVbt
2.5
1.5
1
10
correlation scale parameter in axial direction (inch)
100
Matrix Acidizing of
Sandstone
Small-scale heterogeneities allow acid to
penetrate deeper into the formation than
predicted by homogeneous models
Sandstones with correlated permeability in
the main flow direction (laminated systems)
show the greatest effect acid penetrates
ten times as fast
Heterogeneity results in better stimulation
results than predicted by homogeneous
models
Creation of Conductivity in
Acid Fracturing
After hydraulic pressure is released during a
fracture treatment, the fracture will close
under the force of the confining stresses
Acid creates conductivity through
differential etching the acid must remove
rock in an uneven manner to create lasting
conductivity
Paradox Acid creates conductivity by
heterogeneously removing rock; acid frac
models simulate a homogeneous process
2D fracture geometry
Homogeneous formation no
variations in rock properties,
stress state, leakoff conditions
Local conductivity based on total
amount of rock dissolved and
empirical correlation (NierodeKruk correlation)
No effect of rock dissolution on
fracture mechanics
CSM Meeting Oct. 17, 2006
Scaling Up Local
Conductivity - Background
Laboratory conductivity
measurements are for samples a
few inches in size
Fracture simulators have grid
blocks 5 30 feet in size
Channels that create conductivity
are intermediate in size, are
completely missed if we directly
apply lab-scale models
CSM Meeting Oct. 17, 2006
Intermediate scale
corelation used in
frac model
Intermediate
Scale Model:
Simulation
studies to
develop new
correlation
Acid fracture
model
Lab-scale
Conductivity
correlations
Experimental
study
The task:
3D Navier-Stokes (NS)
Acid transport
Acid/rock reaction->acid etching profile
Complications
Fracture domain is irregular
Domain is growing with time
Method:
Front fixing method (body-fitted
coordinate transformation)
CSM Meeting Oct. 17, 2006
Problem Description
z
y
x
y
Mathematical Model
Assumptions
Incompressible
Consecutive steady state (NS calculation)
Newtonian
Gravity is neglected
Mathematical Model
NS
v
u = 0
v
v
2v
(u )u = p + [ u ]
Acid transport
c
c
+ u.c =
De
t
y y
Fracture width
y0
c
vl c Deff
=
t
y y =0
(1 )
y1
c
vl c Deff
=
t
(1 )
y y = w
Coordinate Transformation
Boundaries and of the mesh points x y z moving ->
fixed points at each time step
y
CSM Meeting Oct. 17, 2006
Coordinate Transformation
Boundaries and of the mesh points x y z moving ->
fixed points at each time step
Coordinate Transformation
Irregular moving boundary (physical) -> a
regular fixed boundary (calculation)
x
L
y y0
=
y1 y 0
z
=
H
=t
0 1
(b( x, z , t ) = y1 y 0 )
0 1
y0
y1
CSM Meeting Oct. 17, 2006
0 1
Y (inch)
-0.1
-0.1
10
15
20
X (feet)
25
30
35
Velocity Profile of Ux
Ux at x=10m
0.015
Analytical
Ux (m/s)
0.012
Numerical
0.009
0.006
0.003
0
-0.1
-0.05
0
Y (inch)
0.05
0.1
Velocity Profile of Uy
0.00002
Uy (m/s)
0.00001
0
-0.1
-0.05
0.05
-0.00001
Analytical
Numerical
-0.00002
Y (inch)
0.1
Velocity Field
Velocity Contour
Initializing a Fracture
1. Variables
Permeability and width
2. Characteristics
Not completely random, spatially correlated
3. Model
Semi-variogram model
FFT simulator (Jennings 2003): correlation scale parameters
Permeability Distribution
Z
30
k (md)
140
130
120
110
100
90
80
70
60
50
40
30
20
20
Z (feet)
10
0
10
-0.2
-0.1
20
0.1
Y (in
ch)
0.2
30
X (fe
et)