GEOS 4430 Lecture Notes: Darcys

Law
Dr. T. Brikowski
Fall 2012
le:darcy law.tex,v (1.22), printed October 15, 2012
Introduction
1
Introduction
Motivation: hydrology generally driven by the need for
accurate predictions or answers, i.e. quantitative analysis
this requires mathematical description of uid and
contaminant movement in the subsurface
for now well examine water movement, beginning with an
empirical (qualitative) relation: Darcys Law
2
Darcys Experiment
3
Darcys Observations
Henry Darcy, studying public water supply development in
Dijon, France, 1856. Trying to improve sand lters for water
purication
apparatus measured water level at both ends and discharge
(rate of ow
L
3
T
) through a vertical column lled with sand
(Fig. 1)
4
Darcy Experimental Apparatus
L
h
h
1
A
Crosssectional area
Q
2
Figure 1: Simplied view of Darcys experimental apparatus. The central
shaded area is a square sand-lled tube, in contact with water reservoirs on
either side with water levels h
1
and h
2
.
5
Empirical Darcy Law
Darcy found the following relationship
Q
A(h
1
h
2
)
L
or
Q = KA
h
1
h
2
L
(1)
where K is a proportionality constant termed hydraulic
conductivity
L
T
6
Head
7
Head as Energy
Head as a measure of energy (1) was used without much
further consideration until the 1940s when M. King Hubbert
explored the quantitative meaning of head h, and ultimately
derived what he termed force potential = g h. He
demonstrated that head is a measure of the mechanical
energy of a packet of uid
mechanical energy of a unit mass of uid taken to be the
sum of kinetic energy, gravitational potential energy and
uid pressure energy (work).
Energy content found by computing work to get to current
state from a standard state (e.g. z = v = p = 0)
energy units are
kg m
2
sec
2
= Nt m = joule
8
Head Energy Components
Kinetic Energy or velocity work. Energy required to
accelerate uid packet from velocity v
1
from velocity v
2
.
E
k
= m
_
v
2
v
1
vdv =
1
2
mv
2
(2a)
Remember that the integral sign is really just a fancy sum,
saying add up all the tiny increments of change between
points 1 and 2
Gravitational work. Energy required to raise uid packet
9
from elevation z
1
to elevation z
2
.
W =
_
z
2
z
1
dz = mgz (2b)
Pressure work. Energy required to raise uid packet pressure
(i.e. squeeze uid) from P
1
to P
2
.
P =
_
P
2
P
1
V dP = m
_
P
2
P
1
V
m
= m
_
P
2
P
1
1

=
1

_
P
2
P
1
dP =
P

(2c)
where (2c) assumes a unit mass of incompressible uid
10
Final Expression for Head
the sum of (2a)(2c) is the total mechanical energy for the
unit mass (i.e. m = 1)
E
tm
=
v
2
2
+ gz +
P

(3)
in a real setting, energy is lost when ow occurs Fetter [Sec.
4.2, 2001], so E
tm
= constant implies no ow. No ow
(Q = 0) in (1) implies constant head (h
1
= h
2
).
Following the analysis of Hubbert [1940] dene head h such
that h g is the total energy of a uid packet. Then
hg =
v
2
2
+ gz +
P

(4a)
11
h =
v
2
2g
+ z +
P
g
(4b)
= z +
P
g
(4c)
where (4c) assumes v is small (true for ow in porous
media).
head is a combination of elevation (gravitational potential)
head and pressure head (where P = gh
p
, and h
p
is the
pressure head, or pressure from the column of water above
the point being considered)
for a given head gradient (
h
x
) discharge ux Q is
independent of path [Fig. 4.5, Fetter, 2001].
12
Head for Variable-Density Fluids
(4c) indicates that head also depends on uid density. This
is only an issue for saline or hot uids. For constant density
(4c) can be written as
h = z + h
p
where h
p
is the height of water above the point of interest
[eq. 4.11, Fetter, 2001].
This implies that head is constant vs. z (for every meter
change in z there is an equal but opposite change in h
p
).
when isnt constant, fresh-water head (the head for an
equivalent-mass fresh water column, h
f
=

p

f
h
p
) should
13
be used for computing gradients, or in ow models [Pottor
et al., 1987]
14
Rock Properties
15
Applicability of Darcys Law
Darcys Law makes some assumptions, which can limit its
applicability
Assumption: kinetic energy can be ignored. Limitation:
laminar ow is required (i.e. Reynolds number low, R
10, viscous forces dominate)
Assumption: average properties control discharge.
Limitation: (1) applicable only on macroscopic scale (for
areas 5 to 10 times the average pore cross section, Fig.
2)
Assumption: uid properties are constant. Limitation:
(1) applicable only for constant temperature or salinity
settings, or if head h is converted to fresh-water equivalent
h
f
16
Scale Dependence of Rock Properties
, K, k
Average
Homogeneous Medium
P
o
r
o
s
i
t
y
,

P
e
r
m
e
a
b
i
l
i
t
y
,

o
r

C
o
n
d
u
c
t
i
v
i
t
y
Distance
H
e
t
e
r
o
g
e
n
e
o
u
s

M
e
d
i
u
m
Microscopic Macroscopic
Figure 2: Variation of rock properties with scale. Note heterogeneous
media have average properties that vary positively (shown) or negatively
(not shown) with distance. See also Bear [Fig. 1.3.2, 1972].
17
True Fluid Velocity
(1) gives total discharge through the cross-sectional area A
specic discharge (q traditionally, v in the text) is the ux
per unit area q =
Q
A
pore or seepage or average linear velocity (v traditionally,
V
x
in the text) is the velocity at which water actually moves
within the pores v =
K

dh
dl
18
Hydraulic conductivity (K)
the proportionality constant in (1) depends on the properties
of the uid, a fact that wasnt recognized until Hubbert
[1956]
Hubbert rederived Darcys Law from physical principles,
obtaining a form similar to the Navier-Stokes equation
1
.
From this he found the following form for K:
K =
kg

(5)
where is the density of the uid and is its dynamic
viscosity (
M
LT
). k is the intrinsic permeability of the porous
medium Fetter [written as K
i
eqn. 3-19, 2001].
1
http://en.wikipedia.org/wiki/Navier-stokes
19
(5) demonstrates that hydraulic conductivity depends on the
properties of both the uid and the rock
20
Permeability (k)
k (units L
2
) depends on the size and connectivity of the pore
space, and can be expressed as k = Cd
2
, where C is the
tortuosity of the medium (depends on grain size distribution,
packing, etc.; unmeasurable), and d is the mean grain
diameter (a proxy for the mean pore diameter). See also
Fetter [sec. 3.4.3, 2001].
standard units are the darcy, (1 darcy = 9.87x10
9
cm
2
),
which is dened as the permeability that produces unit ux
given unit viscosity, head gradient, and cross-sectional area
[Sec. 3.4.2, Fetter, 2001]
21
Measuring Rock Hydraulic Properties
Porosity is measured in the lab with a porosimeter
sample is dried and weighed
then resaturate sample with water (or mercury), very
dicult to do completely
measure volume or mass of liquid absorbed by sample
yields eective porosity
Permeability/Hydraulic Conductivity are measured in the lab
using permeameters (Fig. 3), which apply Darcys Law to
determine K
Constant Head permeameter best used for high
conductivity, indurated samples (rocks)
22
Falling Head permeameter best used for low-permeability
samples or soils
23
Permeameters
(a) Constant Head
Permeameter: K =
V L
Ath
(b) Falling Head
Permeameter:
K =
d
2
t
d
2
c
L
t
ln
`
h
o
h

Figure 3: Constant and Falling-Head Permeameters. a) best for consolidated, high-permeability samples; b) best for
unconsolidated and low-permeability samples. t is the duration of the experiment, V = Qt is the total volume discharged, and A
is the cross-sectional area of the apparatus. After Fetter [Fig. 3.16-3.17, 2001], see also Freeze and Cherry [1979], Todd [1959].
24
Applications of Darcys Law
nd magnitude and direction of discharge through aquifer
solve for head at heterogeneity boundary in composite (two-
material) aquifer
nd total ux through aquitard in a multi-layer system
25
Derivation of the Flow
Equation
26
The Flow Equation (Continuity Eqn.)
Approach: use control volume, itemize ow in and out,
and content of volume [sec. 4.7, Fetter, 2001], see Fig. 4
Simplications: assume steady-state, constant uid
properties
Mass ux
must describe movement of mass across the sides of the
control volume, i.e. determine
mass
areatime
or
m
tl
3
known uid parameters: density
_
m
l
3
_
, velocity q
_
l
t
_
(really specic discharge, or velocity averaged over a
cross-sectional area; see references such as Bear [1972],
27
Schlichting [1979], Slattery [1972] for discussion of volume-
averaging and REVs)
Then mass ux = q
28
Representative Control Volume
q
x
q
x
x+

z
y
x

x
(x,y,z)
(x+ x, y+ y, z+ z)
Figure 4: Mass uxes for control volume in uniform ow eld.
29
Control Volume Mass Balance
Content of control volume (storage)
Steady-state no change in content with time
then sum of in and outows must be 0 (otherwise content
would change)
Sum of ows for control volume in uniform ow
from above know that form will be:
_
rate of
mass in
_

_
rate of
mass out
_
= 0 (6)
30
mass ux in
_
rate of
mass in
_
= q
x
..
mass ux per unit area
yz
. .
area of cube face
(7)
mass ux out
depends on q
out
, let q
out
= q
in
+ change
express change as distance (rate of change) and rate
of change as
q
x
x
to be most accurate, determine rate of change over a
very small distance, i.e. take limit as x 0
thats a derivative
lim
x0
q
x
x
=
dq
dx
31
since q can vary as a function of y and z as well, we write
it as a partial derivative, holding other variables constant
dq
dx

y,z constant
=
q
x
then in the form required for (6)
_
rate of
mass out
_
=
_
q
x
+ x
q
x
x
_
yz
net ux in x-direction:
q
x
x
xyz =
q
x
x
V
32
Continuity Equation
sum of ows for general case (y and z sums are same form
as for x)
V
_
q
x
x
+
q
y
y
+
q
z
z
_
= 0
_
q
x
x
+
q
y
y
+
q
z
z
_
= 0
q = 0 (8)
this is the divergence of the specic discharge, a measure of
the uids to diverge from or converge to the control volume
assume this equation applies everywhere in problem domain,
33
i.e. that uid is everywhere, and uid/rock properties and
variables are continuous
then rock and uid are overlapping continua
34
Flow Equation
Approach: governing equations (8) and (1) contain similar
variables, can we simplify?
Combined equation
substituting Darcys Law (1) for q in (8):
q =

(Kh)
=
2
h
=
_

2
h
x
2
+

2
h
y
2
+

2
h
z
2
_
= 0
This is our governing equation for groundwater ow
35
Assumptions: many were made above
steady-state: no time variation
continuum: uid and rock properties and variables
continuous everywhere in problem domain
constant density: incompressible uid, no compositional
change, no temperature change
no viscous/inertial eects: low ow velocities
36
Bibliography
37
J. Bear. Dynamics of Fluids in Porous Media. Elsevier, New York, NY, 1972.
C. W. Fetter. Applied Hydrogeology. Prentice Hall, Upper Saddle River, NJ, 4th edition, 2001.
ISBN 0-13-088239-9.
R. A. Freeze and J. A. Cherry. Groundwater. Prentice-Hall, Englewood Clis, NJ, 1979.
M. K. Hubbert. The theory of ground-water motion. J. Geol., 48:785944, 1940.
M. K. Hubbert. Darcys law and the eld equations of the ow of underground uids. AIME
Transact., 207:222239, 1956.
E. J. Pottor, S. J. Erikson, and M. E. Campana. Hydrologic utility of borehole temperatures in
areas 19 and 20, pahute mesa, nevada test site. Report DRI-45060, DOE/NV/10384-19,
Desert Research Institute, Reno, NV, 1987.
H. Schlichting. Boundary-Layer Theory. McGraw-Hill, New York, 1979.
J. C. Slattery. Momentum, Energy, and Mass Transfer in Continua. McGraw-Hill, New York,
1972.
D. K. Todd. Ground Water Hydrology. J. Wiley and Sons, New York, 1959.
38