equations of water flow in soil (seepage equation)

The equation for seepage through soil which forms the theoretical basis for the flow net as well as other
methods of solving flow problems will be derived in this section. The following assumptions are made:

Assumptions made

1. Darcy’s law is valid for flow through soil.

2. The hydraulic boundary conditions are known at entry and exit of the fluid (water) into the porous
medium (soil).

3. Water is incompressible.

4. The porous medium is incompressible. These assumptions have been known to be very nearly or
precisely valid. Let us consider an element of soil as shown in Fig. 1.0, through which laminar flow of
water is occurring:

Let q be the discharge with components q x,qy, and qy in the X,Y, and Z directions respectively.

q=q x +q y + q z obviously.

Fig.1.0 Flow through an element of soil

• q=q +q + q
x y z

By darcy’s law,

q z =k . i. A ,…………………………………………1

Where A is the area of the bottom face and q z is the flow into the face

=k z ( −∂∂ zh )dxdy
Where k z z is the permeability of the soil in the Z-direction at the point (x, y, z) and h is the total head.
Flow out of the top of the element is given by:

∂kz −∂ h ∂ 2 h
(
q z + ∆ q z= k z +
∂z )(
, dz −
∂ z ∂ z2 )
dz dxdy ………………………………………..2

Net flow into the element from vertical flow:

∆ q z=inflow-outflow

−∂ h ∂k ∂ h ∂2h
¿kz ( )
∂z (
dxdy − k z + z dz
∂z
−)(
−∂ z ∂ z 2
dxdy
)
∂2 h ∂ k z ∂ h ∂ k z ∂2 h
:.
(
∆ q z= k z
∂ z2
+
∂ z2
+
∂z
. dz
∂ z2)dz dxdydz ……………………………………..3

Assuming the permeability to be constant at all points in a given direction, (that is, the soil is
homogeneous),

∂kz
=0
∂z

∂2 h
:. :. ∆ q z=( k z ) dxdydz ……………………………………………………………………4
∂ z2

Similarly, the net inflow in the X-direction is:

∂2 h
∆ q x =(k x ) dxdydz
∂ x2

For two-dimensional flow, q y = 0

∂2 h ∂2 h
(
∆ q=∆ q x + ∆ q z = k x 2 + k z 2 dxdydz
∂x ∂z )
∆q may be obtained in a different manner as follows:

The volume of water in the element is:

S .e
V w= dxdydz
(1+ e)

∆q = rate of change of water in the element with time:

 ∂V w S.e
¿
∂t
=(∂ /∂ t) [
(1+ e)
. dxdydz ]
dxdydz
is the volume of solids ,which is constant
(1+e)

dxdydz ∂
:. ∆ q=
(1+e )
= ( )
∂t
( S . e)

Equating the two expressions for ∆q, we have:

∂2 h ∂2 h dxdydz ∂
( kx
∂x 2
+ k z
∂z) 2
dxdydz=
( 1+e )
= ( )
∂t
( S . e )…………………………………..2

This is the basic equation for two-dimensional laminar flow through soil. The following are the possible
situations:

(i) Both e and S are constant.

(ii) (ii) e varies, S remaining constant.
(iii) (iii) S varies, e remaining constant.
(iv) (iv) Both e and S vary.
Situation (i) represents steady flow.
Situation (ii) represents ‘Consolidation’ or ‘Expansion’, depending upon whether e decreases
or increases.
Situation (iii) represents ‘drainage’ at constant volume or ‘imbibition’, depending upon
whether S decreases or increases.
Situation (iv) includes problems of compression and expansion.
Situations (iii) and (iv) are complex flow conditions for which satisfactory solutions have yet
to be found. ( Eq. 2 is applicable only for small strains).
For situation (i), Eq. 2reduces to:

∂2 h ∂2 h
kx +k z =0………………………………….3
∂ x2 ∂ z2

If the permeability is the same in all directions, (that is, the soil is isotropic),

∂2 h ∂ 2 h
+ =0……………………………………………………4
∂ x2 ∂ z2

This is nothing but the Laplace’s equation in two-dimensions. In words, this equation means
that the change of gradient in the X-direction plus that in the Z-direction is zero.
From Eq. 6,
( ∂∂x )(k ∂∂ hx )+( ∂∂z )(k ∂h∂z )=0
x z

∂h ∂h
But k x . =vx and k z =v z by Darcy’s law.
∂x ∂z

∂ v x ∂ vz
:. + =0 ……………………………………………………….5
∂x ∂z

This is called the ‘Equation of Continuity’ in two-dimensions and can be got by setting ∆q = 0
(or net inflow is zero) during the derivation of Eq. 3. The flow net which consists of two sets
of curves – a series of flow lines and of equipotential lines–is obtained merely as a solution
to the Laplace’s equation – Eq. 4. The fact that the basic equation of steady flow in isotropic
soil satisfies Laplace’s equation, suggests that, the flow lines and equipotential lines
intersect at right-angles to form an orthogonal net – the ‘flow net’. In other words, the flow
net as drawn in the preceding sections is a theoretically sound solution to the flow
problems. The ‘velocity potential’ is defined as a scalar function of space and time such that
its derivative with respect to any direction gives the velocity in that direction. Thus, if the
velocity potential, φ is defined as kh, φ being a function of x and intersect at right-angles to
form an orthogonal net – the ‘flow net’. In other words, the flow net as drawn in the
preceding sections is a theoretically sound solution to the flow problems. The ‘velocity
potential’ is defined as a scalar function of space and time such that its derivative with
respect to any direction gives the velocity in that direction. Thus, if the velocity potential, φ
is defined as kh, φ being a function of x and z,

∂∅ ∂h

{ ∂x
∂∅
∂z
=k .

=k .
∂h
∂z
=v
∂x x
=v y }
……………………………………………………….6

From Eq 7 for an isotropic soil and in the view of the definition of the velocity potential, we
have:
∂2 ∅ ∂2 ∅
+ =0……………………………………………………………7
∂ x2 ∂ z 2

This is to say the head as well as the velocity potential satisfy the Laplace’s equation in two-
dimensions.
The equipotential lines are contours of equal or potential. The direction of seepage is always
at right angles to the equipotential lines.
The ‘stream function’ is defined as a scalar function of space and time such that the partial
derivative of this function with respect to any direction gives the component of velocity in a
direction inclined at + 90° (clockwise) to the original direction. If the stream function is
designated as ψ(x, z),

∂Ψ

{ } ∂z
∂Ψ
∂x
=v x

=−v z
……………………………………………………8

And by definition Eq 9 and 11 ,we have

∂Ψ
{ ∂ ∅ /∂ x=
}
∂ z …………………………………………..9
∂ ∅=−∂ Ψ /∂ x
These equations are known as Cauchy-Riemann equations
Substituting the relevant values in terms of ψ in the continuity equation (Eq. 8) and
Laplace’s equation (Eq. 10), we can show easily that the stream function ψ(x, z) satisfies
both these equations .
Functions φ and ψ are termed ‘Conjugate harmonic functions’. In such a case, the curves
‘‘φ(x, z) = a constant’’ will be orthogonal trajectories of the curves ‘‘ψ(x, z) = a constant’’.
Flow nets will be useful for the determination of rate of seepage, hydrostatic pressure,
seepage pressure and exit gradient. These aspects have already been discussed in Sec. 6.2.
The following important properties of the flow nets are useful to remember:
(i) The flow lines and equipotential lines intersect at right angles to each other.
(ii) The spaces between consecutive flow and equipotential lines form elementary
squares (a circle can be inscribed touching all four lines).
(iii) The head drop will be the same between successive equipotential lines; also, the
flow in each flow channel will be the same.
(iv) The transitions are smooth, being elliptical or parabolic in shape.
(v) The smaller the size of the elementary square, the greater will be the velocity and
the hydraulic gradient. These are correct for homogeneous and isotropic soils

SEEPAGE THROUGH NON-HOMOGENEOUS AND ANISOTROPIC SOIL

Although Eq. 5 was derived from general conditions, the preceding examples considered
only soil that does not vary in properties from point horizontally or vertically–homogeneous
soil– and one that has similar properties at a given location on planes at all inclination–
isotropic soil. Unfortunately, soils are invariably non-homogeneous and anisotropic.
The process of formation of sedimentary soils is such that the vertical compression is larger
than the horizontal compression. Because of the higher vertical effective stress in a
sedimentary soil, the clay platelets tend to have a horizontal alignment resulting in lower
permeability for vertical flow than for horizontal flow.
In man-made as well as natural soil, the horizontal permeability tends to be larger than the
vertical. The method of placement and compaction of earth fills is such that stratification
tend to be built into the embankments leading to anisotropy.

Non-homogeneous Soil
In case of flow perpendicular to soil strata, the loss of head and rate of flow are influenced
primarily by the less pervious soil whereas in the case of flow parallel to the strata, the rate
of flow is essential controlled by comparatively more pervious soil. Figure 1.1 shows a flow
channel and part of a flow net, from soil A to soil B. The permeability of soil A is greater than
that of soil B. By the principle of continuity, the same rate of flow exists in the flow channel
in soil A as in soil B. By means of this, relationship between the angles of incidence of the
flow paths with the boundary of the two flow channels can be determined. Not only does
the direction of flow change at the boundary between soils with different permeabilities,
but also the geometry of the figures in the flow net changes. As can be seen from Fig. 1.1,
the figures in soil B are not squares as in soil A, but are rectangles

Fig 1.1 Flow at the boundary between two soils.

qA = q B
But qA = kA.