Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

L 02 Darcy Linear Radial PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Flow Through Porous Media (Internal Circulation Only)

2.0 Darcy’s Law for Linear and Radial Flow

Darcy’s Law for Linear and Radial Flow


Outline
0. Synopsis
1. The Differential Form of Darcy’s Law
2. Darcy’s Law for Anisotropic Permeability
3. Integrated Forms of Darcy’s Law for Incompressible Fluids
4. Integrated Forms of Darcy’s Law for Slightly Compressible Fluids
5. Integrated Forms of Darcy’s Law for Gases

2.0 Synopsis
In the previous lecture, we discussed Darcy’s Law over finite lengths. In this leture, we will give
the differential form of Darcy’s Law. The differential form will be used to compute flow relations
for several different geometries, and for several different types of fluids.

2.1 The Differential Form of Darcy’s Law


Darcy’s Law was based on experiments with the potential (similar to pressure) drop p measured
over a finite length, L, in a sandpack of pemeability, k, and cross-sectional area, A. For linear flow
of an incompressible fluid of viscosity  through this sand pack, the flow rate is related to other
factors by
kAp
q ............................................................. (2.1)
L
q
in consistent units. The superficial velocity is v  . Also, we recognize that the limit of the
A
quotient of the pressure difference and the length of flow is minus the pressure gradient (or its
p( x  x)  p( x) dp
derivative with respect to length, lim  . This gives us the differential form
x 0 x dx
of Darcy’s Law:
k dp
u  vsup   ....................................................... (2.2)
 dx
The change in sign is necessary because the direction of flow is opposite to the direction of
pressure change.

We can integrate Eqn. (2.2) to compute forms of Darcy’s Law for various flow geometries and
fluid types. Before proceeding to that, however, we will first consider a generalization of Darcy’s
Law for anisotropic permeability.

2.2 Darcy’s Law for Anisotropic Permeability


In many reservoir rocks, the permeability depends on a direction. Often, the vertical permeability
is much less than the horizontal permeability because of shale layers or other sedimentary
features. The permeability may vary horizontally because of tectonic stresses, fracturing, or
depositional processes. The result is that we must consider different permeabilities for flows in
these different directions. In a three dimensional system, the components of velocity are

Page 1 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

1  p p p 
ux    k xx  k xy  k xz 
  x y z 
1 p p p 
uy    k yx  k yy  k yz  ......................................... (2.3)
  x y z 
1  p p p 
uz   k zx  k zy  k zz 
  x y z 
In anisotropic systems, permeability is a tensor. You are familiar with tensors from your statics
class; stress is a tensor. Physically, elements like kxy give the flow that occurs in the x-direction
due to pressure gradients in the y-direction. You see, anisotropic porous media can act like
“louvres”, with flow being “shunted” toward the direction of highest permeability (parallel to the
louvers) even if the pressure gradient is not in that direction. For porous media, kij = kji, and one
can always define a coordinate system in which the off-diagonal terms are zero. This is the
principal coordinate space, which is defined in exactly the same way the principal stress space
k p
was defined in statics. In the principal coordinate space (X, Y, Z), u X   X , and so on.
 X

The points you should recall from this discussion are conceptual rather than detailed:
 Permeability may vary with direction in porous media.
 Permeability can be represented as a symetric tensor. One can always find a space in
which the tensor is diagonal.
 In general, 6 factors must be specified to describe permeability in three dimensions (kxx,
kxy, kxz, kyy, kyz, and kzz or kX, kY, kZ, and three angles).
 The vertical permeability is often much less than the horizontal permeability.
 Reservoir engineers seldom have enough data to know what the permeability tensor is,
nor do most solution methods include it. However, engineers often use a lower vertical
permeability compared to horizontal.

2.3 Integrated Forms of Darcy’s Law for Incompressible


Fluids
General
There is a very important point about Darcy’s Law that we haven’t emphasized until this point:
unlike your favorite snack food items, it sells by volume, not weight. That is, the pressure
gradient is proportional to the volume flow rate, not the mass flow rate. For steady-state flow, the
mass flow rate is constant. For incompressible flow, the volume flow rate is also constant.
However, if the fluid is compressible, the volume flow rate varies and the integrated form of
Darcy’s Law will look different (as we shall see below).

We will consider linear, radial, and spherical flow (Fig. 2.1).

We now introduce the formation volume factor, B. The formation volume factor converts a mass
quantity (stock tank barrels) to a volume quantity (reservoir barrels). Therefore, we can rewrite
the differential form of Darcy’s Law (Eqn. 2.2) as
kA dp
q ........................................................... (2.4)
B dx

Page 2 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

Figure 2.1 – Illustration of three different flow geometries and their


variations in cross-sectional area for flow.

Linear Flow
The equation for linear flow over a finite length can be obtained by rearranging and integrating
Eqn. (2.4):
L p ( L ) kA
0 qdx   p ( 0 ) B dp
kA p ( L)  p (0)
q  ( L  0)   ........................................... (2.5a)
B
kA p ( L)  p (0)
q
BL
in consistent units. In field units,
kA p( L)  p(0)
q  0.001127 .......................................... (2.5b)
BL
When the pressure is greater at 0 than at L, the difference in the numerator on the right hand side
is negative, and flow goes from 0 to L. On the other hand, when pressure is greater at L than at 0,
flow goes from L to 0. The sign convention makes sense. Note that Eqn. (2.5) is very similar to

Page 3 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

our ealier form of Darcy’s Law [Eqn. (1.1)], except that we have now included B and are being
much more careful about the sign of the pressure drop and the direction of flow.

Radial Flow
The derivation for radial flow is very similar, except the area is a function of r (Fig. 2.1). The
equation to be integrated is
k dp
u
 dr
....................................................... (2.6a)
q k dp

2rh  dr
Rearranging and integrating,
r q dr p 2kh
r r   p B dp
e e

w w

2kh
qln( re )  ln( rw )   ( p e  p w ) ...................................... (2.6b)
B
2kh( p e  p w )
q
r 
B ln  e 
 rw 
This definition makes sense, in that flow is negative, that is against r, when the pressure increases
with r. However, because most wells are production wells, and reservoir engineers don’t like
dragging minus signs around all the time, the radial flow equation is almost always defined as
(Craft et al., p. 226):
2kh( p e  p w )
q ...................................................... (2.7)
 re 
B ln  
 rw 
in consistent units. In field units,
kh( pe  p w )
q  0.00708
r 
B ln  e 
 rw 
This is just a convention, a matter of convenience. As long as one remembers the convention,
either of Eqns. (2.6b) or (2.7) can be used. I will generally use Eqn. (2.7).

Spherical Flow
For spherical flow (Fig. 2.1), the equation to be integrated is
k dp
u
 dr
....................................................... (2.8a)
q k dp

4r 2  dr
which yields

Page 4 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

q dr
re pe 4k
rw r 2
 
p w B
dp

1 1  4k
 q     ( pe  p w ) ........................................ (2.8b)
 re rw  B
4k ( pe  p w )
q
1 1
B   
 rw re 
Again, this sign convention is often reversed to give the more convenient form,
4k ( pe  p w )
q ...................................................... (2.9)
1 1
B   
 rw re 

Applications
The integrated flow equations [Eqns. (2.5), (2.7), and (2.9)] can be used to predict steady-state
flow of incompressible fluids.

Linear flow occurs in core floods. It may also be a good approximation to flow far from wells in
reservoirs, in aquifers, or between wells in a pattern flood.

Radial flow occurs around most wells.

Spherical flow occurs in wells that do not penetrate the entire formation, at the tips of horizontal
wells, and near perforations.

Observations
The pressure is linear versus a different function of distance for each flow geometry:

Linear: x
Radial: ln(r)
Spherical: 1/r

2.4 Integrated Forms of Darcy’s Law for Slightly


Compressible Fluids
Compressibility Model
The formation volume factor is the ratio of the volume of the fluid at reservoir conditions to its
volume at standard (or reference , R, or stock-tank, sc) conditions. Thus,
V
B
VR
........................................................... (2.10a)
R


where the fluid density is . For a fluid of small and constant compressibility (Craft et al., p. 214),
   R exp c( p  p R ) ............................................... (2.10b)

Page 5 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

1 
where c is the isothermal compressibility, c  . For oil and water the product of fluid
 p
compressibility and pressure change is usually small: typical water compressility is 3 10-6/psi
and (p-pR) is seldom more than 104, so the maximum product is about 10-2 or less. For such
systems, the exonential can be very accurately approximated as
   R exp c( p  p R )
............................................... (2.10c)
  R 1  c( p  p R )
Finally, we get an expression of B by combining Eqns. (2.10a) and (2.10c):
B  1  c( p  p R ) ..................................................... (2.11)

Linear Flow
Now, Eqn. (2.11) for B can be substituted into the differential form of Darcy’s Law [Eqn. (2.4)]
to get an expression for linear flow of an slightly compressible fluid:
kA dp
q
B dx
............................................... (2.12)
kA dp

 1  c( p  p R ) dx
For linear flow, this can be rearranged and integrated as follows:
L p( L) kA
0
q dx   
p (0)  1  c( p  p R )
dp

q ( L  0)  
kA
ln 1  c( p L  p R )  ln 1  c( p0  p R )................... (2.13a)
c
kA  1  c( p L  p R )  
q ln 
cL   1  c( p0  p R )  
in consistent units. In field units,
kA  1  c( p L  p R )  
q  0.001127 ln  ............................. (2.13b)
cL  1  c( p0  p R )  

Assumptions
 Compressibility is constant; needed for Eqn. (2.10b).
 Compressibility is small; needed for Eqn. (2.10c).
 Viscosity and permeability are constant; needed to integrate Eqn. (2.13).

Applications
Craft et al. (p. 220) show that this effect is small for most oil and gas systems; that is not
surprising given the calculations we sketched out above. Most steady-state liquid inflow
calculations can safely neglect compressibility and use Eqns. (2.5), (2.7) and (2.9) as appropriate.
You will have an opportunity to assess the importance for yourself in the homework.

Other Flow Geometries


Beginning with the differential form of Darcy’s Law for slightly compressible flow [Eqn. (2.12)],
you can compute the inflow equations for other geometries using the same methods used above.
These calculations are relatively easy, because the variable are separable and we can solve these
ordinary differential equations by integrating with respect to p and r or x separately.

Page 6 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

2.5 Integrated Forms of Darcy’s Law for Gases


For a gas, the formation volume factor is
p scTz
B ......................................................... (2.14a)
pTsc
in consistent units. In field units,
p scTz
B ..................................................... (2.14b)
5.615 pTsc
We will go ahead and solve this problem in consistent units, and then convert to field units at the
end. Substituting the expression for B [Eqn. (2.14a)] into the differential form of Darcy’s Law
[Eqn. (2.4)],
kA dp
q
B dx
kA dp
 ..................................................... (2.15)
p scTz dx

pTsc
kTsc Ap dp

p scTz dx
This can be integrated as we have done before:
L p kTsc A
0 q dx  p p Tz p dp ............................................. (2.15b)
L

0
sc

This integral (on the right-hand side) is (as mathematicians say) nontrivial --  and z are functions
of pressure. Later, we will actually do this integral to improve our analysis of gas flow. For now,
we will follow Craft et al. (p. 221) and assume they are approximately constant and move them
outside the integral:
L kTsc A p
0 q dx   p Tz p p dp
L

0
sc

kTsc A  1 2
qL     p L  p02  ........................................... (2.16a)
p scTz  2 
kTsc A p L2  p02 
q
2p scTzL
in consistent units. In field units,
kTsc A p L2  p02 
q  0.003124 ......................................... (2.16b)
p scTzL
The constant given in Craft et al. is very slightly wrong.

The viscosity, , and gas deviation factor, z, should be evaluated at the mean pressure,
p  p0
p L .
2

Assumptions
 z, k, Tconstant
  and z evaluated at the mean pressure.

Page 7 of 8
Flow Through Porous Media (Internal Circulation Only)
2.0 Darcy’s Law for Linear and Radial Flow

Discussion
Craft et al. (p. 221) discuss how to better
approximate the integrals evaluated above based
on the behavior at relatively low and relatively
high pressure (Fig. 2.2, from Craft et al.). The
relatively simple low- and high-pressure behavior
can be exploited to derive corresponding
approximations for the flow equations.

Later, we will take a more accurate, more general,


and more reliable approach based on a transform
called real gas pseudopressure. As we will see,
this is the preferred method for analyzing gas flow
– it involves a few more computations, but at the
end of the day analysis is simpler because you
don’t have to use different approximations at high
Figure 2.2 – Behavior of  as a function of and low pressure.
pressure. At low pressure (below 1000 psi),
zis approximately constant; above 6000 psi For an ideal gas, z is one and  does not depend on
z/p is approximately constant. The pressure. Thus, Eqn. (2.16) could be used to
intermediate behavior is more complex. accurately analyze gases whose PVT behavior
near ideal.

At low pressure, flow of gases is much different than flow of liquids – flow rate is proportional to
the difference of the squares of pressure rather than the difference of pressure itself. At high
presssure, the gas compressibility becomes small and high-pressure gas flow is similar to liquid
flow.

------

Page 8 of 8

You might also like