Professional Documents
Culture Documents
Paper Hpsc2006
Paper Hpsc2006
Abstract - In this paper the authors present a general mathematical model of two
dimensional horizontal flow of seawater intrusion into coastal confined and unconfined
aquifers.
Algorithms and programing of this model are formulated by weak Galrkin finite
element method for prediction of the transient effect of pumping wells on seawater
intrusion into coastal confined and unconfined aquifers.
The validity of the model is tested by using the analytical solutions. An example
was calculated to determine the location of the interface at Hoa Khanh coastal aquifers
when one extracts, by pumping wells, a certain discharge of freshwater, during a long
time, near the interface.
I. Introduction
At coastal aquifers, a zone of transition from fresh aquifer water to seawater, the
water with different concentrations of salt as a result of hydrodynamic dispersion of the
dissolved matter. Under certain conditions, it may be narrow, relative the aquifer
thickness, and may be approximated as a sharp interface. The intensive extraction of
groundwater has upset the long established balance between freshwater and seawater
potentials, causing encroachment of seawater into freshwater aquifers, this phenomenon
is said to be seawater intrusion.
General, the interface is described by three-dimensional mathematical model; but
under certain conditions for simplifying reason, one can described this interface by
regional flow model; with integrating the three-dimensional model over the vertical,
between the upper and the lower boundaries of a considered domain.
In this paper, the authors derive a general mathematical model of two-dimensional
horizontal flow just validitied for coastal confined aquifer and phreatic aquifer.
The algorithms and programing of this model are formulated by standard Garlerkin
finite element technique.
II. Governing equations
The governing system of coupled differential equation of
phreatic aquifer [4]
-.(T f) + .(Tah) = If + Is+ q'f+ q's
Sh/t -.(Ta )h + .(Ta f) = -Is- q's
where:
f = hf/ ; T = K(H1+hf) ; Ta = T(H1 - h)/(H1+ hf)
(1)
(2)
(3)
where:
(4)
f
f
h
h
(Tx ) + (Ty ) + (Tax ) + (Tay ) = If + Is + q'f +q's
x y
y x
x y
y
x
f
f
h
h
h
S0 (Tax ) + (Tay )+ (Tax ) + (Tay ) = Is q,s
y
x y
y x
x y
t x
(5)
(6)
where:
f =
(1 ) h f +
s f
f
; hf = z +
p
fg
Tx = K x [H 1 + (1 )h f H 2 ]; Ty = K y [H 1 + (1 )h f H 2 ] ;
Tax = Tx
(H 1 h)
(H 1 h)
; Tay = Ty
[H 1 + (1 )h f H 2 ]
[H 1 + (1 )h f H 2 ]
and:
+ for the case of phreatic aquifer = 0
+ for the case of confined aquifer = 1
Where:
f : density of freshwater; s: density of seawater; g: gravity acceleration
S0: specific storativity; t: time
hf: piezometric head for freshwater
f ,s: piezometric head (as Hubberts potential) of freshwater and saltwater
respectively
If, If:- supply functions, representing a distributed surface supply of fresh
and salt water into the aquifer
q'f, q's: sources or sinks of freshwater and salt water aquifer respectively
Kx, Ky: x, y directional component of hydraulic conductivity
q 'f
'
s
hf
h
h Freshwater
Saltwater
x
H Saltwater
Interface
'
s
q 'f
H1
h Freshwater
Interface
H
x
H2
N L( P)
i
dA
( 7)
Where:
Ni(e) , i =1,2,3 are the linear shape functions
L(P)m , m =1,2 are the partial differential equations (5) or (6)
A: area of element
The interface depths are linearly interpolated in terms of the shape function as
follow:
n
h ( e ) = N (i e ) h i
(8)
i =1
Where:
h (e): the z coordinate of interface or interface depth at any point of an element
n: total nodes of an element; hi: interface depth at node i of an element
The time derivative is discretized using the finite difference scheme as
h h(t + t ) h(t )
t
t
h( ) h ( t ) + ( t )
Where:
h
h(t + t ) h(t )
= h ( t ) + ( t )
t
t
( t )
t
h( ) = (1 )h(t ) + h(t + t )
A1
R1
3 q f 1 + q s1
.
[(1 ) h f H 2 + h ] .
+
+
[ H 1 + (1 ) h f H 2 ] .
.
R
A n q fn + q sn
n
(9)
f1
.
f n
+ K
.
C
t .
h
]+ [
t+ t
.
=
.
I
(e)
s1
A
3
(e)
sn
A
3
s1
sn
[C ]
t
.
h
(10)
where:
[K a ] = [K (ae ) ]
m
[K b ] = [K (be ) ] ; [C ] = [C ( e ) ]
m
e =1
[q fi
+ q si ] =
[q
m
e =1
e =1
(e)
fi
[K ] = 4TA
( e)
b
[K ]
( e)
a
+ q
(e)
ax
(e)
si
e =1
] ; [q ] = [q ]
m
si
e =1
(e)
si
[R i ] = [R (i e ) ]
b 2i b i b j b i b k
T ( e ) ay
2
b
b
b
b
b
j i
j
j k+
4A
2
b k b i b k b j b k
b 2i b i b j b i b k
Ty(e)
Tx(e)
2
=
b j b i b j b j b k +
4A
4A
2
b
b
b
b
b
k i k j k
( i = 1, n )
e =1
c 2i c i c j c i c k
2
c j c i c j c j c k
2
c k c i c k c j c k
c 2i c i c j c i c k
2
c j c i c j c j c k ;
2
c k c i c k c j c k
[C ] = S
( e)
( e)
0
1 0 0
A
0 1 0
3
0 0 1
in which
So(e): specific storativity of a triangular element
ai, bi, ci: coefficients of a shape function Ni
The domain pattern for studying illustration is a rectangular, symmetric by pass CD
edge; AC is the coast line along the y axis, AB, CD perpendicular to the coast; Q1, Q2 are
the charge of fresh water of pumping well (Fig. 2)
y
A
qn1
B
Q1
Coast
line
Q2
x
C
qn2
100
200
300
400
500
600
700
800
h (t=0 day)
-16.00
-22.80
-25
-25
-25
-25
-25
-25
-25 -25
-25
h(t=8000 d)
-8.91
-14.78
-20.17
-23.95
-24.95
-24.99
-25
-25
-25 -25
-25
h(t=14000 d)
-7.54
-12.36
-16.78
-20.96
-24.1
-24.95
-24.99 -25
-25 -25
-25
h(t=20000d)
-6.76
-10.96
-14.83
-18.56
-22.05
-24.34
-25
Based on the numerical results of Table 1, the location of time dependent sharp
interface is shown in Fig. 3. One can see that the distance of the interfaces toe from the
coast line, at t = 0 day, is 250 m, and from the extraction well is 750 m. After t = 20000
days, this toe will deeply move into the aquifer and these distances respectively are 701.5
m and 298.5m
Q1
-5 0
200
400
600
800
1000
1200
-10
t = 20000 day
-15
-20
t=0
-25
-30
FEM
701,50
Theis
701,46
Strack
701,60
From the results summarized in Table 2, it follows that the results of these
methods differ very little, but the numerical solutions by finite element method can
calculation the shape, location of the sharp interface with different geologies.
V. Application for a real problems
Applying for quartier North of Danang City, with an area of 2.88 km2 , one its side along
the coastal line of Danang Bay (Fig. 4).
Average
layer
thickness
(m)
Phreatic
aquifer
0,74
20
Confined
aquifer
0,85
25
Layer
Hydraulic conductivity
Kx
(m/day)
Ky
(m/day)
Kz
(m/day)
16
16
1,6
2
3
20
12
20
12
1,2
1,2
0,8
Initial condition
At initial time t0, parameters hf, h are determined from steady flow. For phreatic aquifer
and confined aquifer they are showed correspondent in Table 5 and Table 6.
Table 5: Parameters hf, h at perpendicular cross section to the coastal line
and crossing the pumping well HK6 (at initial time t0 = 0).
Node
x
hf
26
43
60
77
94
111
128
145
162
179
196
213
110
200
300
400
500
600
700
800
900
1000
1100
1200
1300
h initial
0.45 0.55 0.64 0.72 0.78 0.85 0.91 0.96 1.01 1.06 1.11 1.15
18.1 20.5 20.5 18.4 20.1 20.9 22.3 19.1 20.1 -20 18.5 20.6
10
27
44
61
248
265
282
299
70 100
200
300 400 500 600 700 800 900 1000 1100 1200 1300 1400
1500
1600
1700
0.51
f
h initial
0.36
78
197
1.35
1.39
1.44
1.48
-22.3 -34.4 -40.6 -44.9 -44.8 -44.6 -45.8 -44.9 -45.9 -44.2 -44.6 -45.8 -45.8 -44.4 -44.9
-47.3
-46.2
-44.6
Boundary condition
For phreatic aquifer
- Boundary AB: qn1 = 0,
- Boundary AC: h = 0,
For confined aquifer
- Boundary AB: qn1 = 0,
- Boundary AC: h = -20m,
1.19
214 231
1.25 1.30
f
=0
y
f
=0
y
Results:
For phreatic aquifer, we study two cases:
Case 1: no supply function I = 0.00 m / day
Case 2: with supply function I = 0.001 m / day
The movement of interface at pumping wells of phreatic aquifer without supply
functions as in Table 7.
Table 7: Distance from the toe to pumping well of phreatic aquifer
without supply functions.
No
Pumping well
1
2
3
4
5
6
7
8
9
10
11
12
13
HK1
HK2
HK3
HK4
HK5
HK6
HK7
DK1
DK2
DK3
DK4
DK5
DK6
1000 day
200
195
195
200
385
389
385
489
480
485
485
685
685
8000 day
98
96
96
98
200
198
200
240
230
230
240
390
390
10000 day
45
40
40
45
45
40
45
160
155
155
155
280
280
The movement of interface at cross section passing pumping wells of phreatic aquifer
with supply function as in Table 8.
No
1
2
3
4
5
6
7
h(t=10000 d)
0 -4.8 -9.1 -12.5 -15.9 -18.5 -20.8 -22.3 -19.1 -20.1 -20 -18.5 -20.6
At plane section across pumping well HK3 (Q6=300m3/day)
Node
10 27 44 61 78 95 112 129 146 163 180 197 214
X
110 200 300 400 500 600 700 800 900 1000 1100 1200 1300
h initial
0 -18.1 -20. -20.4 -17.9 -20.1 -20.9 -22.3 -19.1 -20.1 -19.5 -19.2 -20.7
h(t=1000 day) 0 -14.3 -18.4 -20.4 -17.9 -20.1 -20.9 -22.3 -19.1 -20.1 -19.5 -19.2 -20.7
h(t=8000 d)
0
-7 -11 -15 -17 -20.1 -20.9 -22.3 -19.1 -20.1 -19.5 -19.2 -20.7
h(t=10000 d)
0
-5 -9 -12 -18.5 -20.1 -20.9 -22.3 -19.1 -20.1 -19.5 -19.2 -20.7
.
Q6=300m3/day
0
-5 0
200
-1 0
-1 5
400
t=0
600
1000 8000
800
10000 day
1000
1200
-2 0
-2 5
Q3=300m3/day
0
-5 0
-1 0
-1 5
200
400
t=0
600
1000 8000
800
1000
1200
10000 day
-2 0
-2 5
26
43
60
77
94
111
128
145
196
213
110
200
300
400
500
600
700
800
1300
20.6
20.6
20.6
20.6
h initial
h(t=1000
)
h(t=8000
)
h(t=1000
0)
0
0
0
0
162
179
10
27
44
61
78
95
112
129
110
200
300
400
500
600
700
800
20.1
20.1
20.1
20.1
20.9
20.9
20.9
20.9
22.3
22.3
22.3
22.3
h initial
h(t=1000
)
h(t=8000
)
h(t=1000
0)
0
0
0
0
146
163
180
197
214
1200
19.2
19.2
19.2
19.2
1300
20.7
20.7
20.7
20.7
200
400
t=0
600
1000 8000
800
10000 day
1000
1200
-1 5
-2 0
-2 5
0
-5 0
-1 0
-1 5
Q3=300m3/day
200
400
t=0
600
1000 8000
800
Q6=300m3 /ngy
1000
1200
10000 day
-2 0
-2 5
Table 11: Location of time dependent sharp interface at DK3, DK6 of confined aquifer
Node
10
X
70
h initial
-22.3
h(t=1000 d) -22.3
h(t=8000 d) -22.3
h(t=10000d) -22.3
Node
163
X
900
h initial
-44.2
h(t=1000)
-44.2
h(t=8000)
-44.2
ht=10000)
-44.2
Node
11
X
50
h initial
-19.6
h(t=1000)
-19.6
h(t=8000)
-19.6
ht=10000)
-19.6
Node
164
X
900
h initial
-46.7
h(t=1000)
-46.7
h(t=8000)
-46.7
H(t=10000) -46.7
ht=10000) -46.7
27
100
-34.4
-28.9
-25.1
-23.6
180
1000
-44.6
-44.6
-44.6
-44.6
28
100
-34.4
-28.8
-24.7
-24.1
181
1000
-44.6
-44.6
-44.6
-44.6
-44.6
44
200
-40.6
-37.7
-32.8
-30.1
197
1100
-45.8
-45.8
-45.8
-45.8
45
200
-40.4
-36.3
-32.8
-30.0
198
1100
-43.4
-43.4
-43.4
-43.4
-43.4
61
300
-44.9
-41.5
-37.4
-34.2
214
1200
-45.8
-45.8
-45.8
-45.8
62
300
-44.9
-40.0
-37.4
-34.4
215
1200
-45.3
-45.3
-45.3
-45.3
-45.3
78
400
-44.8
-43.1
-41.2
-37.7
231
1300
-44.4
-44.4
-44.4
-44.4
79
400
-46.7
-43.4
-40.6
-37.3
232
1300
-46.7
-46.7
-46.7
-46.7
-46.7
95
500
-44.6
-44.6
-42.9
-40.0
248
1400
-44.9
-44.9
-44.9
-44.9
96
500
-47.2
-47.2
-44.3
-40.0
249
1400
-44.9
-44.9
-44.9
-44.9
-44.9
112
600
-45.8
-45.8
-44.6
-42.5
265
1500
-47.3
-47.3
-47.3
-47.3
113
600
-47.0
-44.6
-44.0
-42.5
266
1500
-43.3
-43.3
-43.3
-43.3
-43.3
129
700
-44.9
-44.9
-44.7
-44.5
282
1600
-46.2
-46.2
-46.2
-46.2
130
700
-45.0
-45.0
-45.0
-45.0
283
1600
-46.2
-46.2
-46.2
-46.2
-46.2
146
800
-45.9
-45.9
-45.9
-45.8
299
1700
-44.6
-44.6
-44.6
-44.6
147
800
-44.0
-44.0
-44.0
-44.0
300
1700
-44.6
-44.6
-44.6
-44.6
-44.6
On the horizontal plane, the toe of interface corresponds the time period of
phreatic aquifer when it has supply function (Fig. 13a) and has not supply function (Fig.
13b).
On the horizontal plane, the toe of interface corresponds the time period of confined
aquifer (Fig. 14).
Based on the results above, we can see that, during the pumping time period 8000
days ( 22 years), the distance of the nearest toe the pumping well systems (HK1, HK2,
HK3, HK4) of phreatic aquifer is 96m (if this aquifer has not supply functions) and is
275m (if this aquifer has supply function), and the confined aquifer (DK1, DK2, DK3,
DK4) is 130m.
These results are correct with physical phenomena, because when the phreatic
aquifer has a supply function, the goundwater level will arise and the pressure of this
aquifer will arise, too; so the toe of the interface will be pushed toward the sea.
VI- Conclusion
This mathematical model of two-dimensional horizontal flow of seawater intrusion is
generally valid for phreatic aquifer and confined aquifer. Algorithm and programing can
be applied for any coastal aquifer with many pumping wells for predicting the sharp
interface. It is important for the geologists, hydrologists, hydraulic engineers, water
resources planners, managers, and governmental policy makers, who are engaged in the
sustainable development of coastal fresh groundwater resources.
References
1.
2.
3.
4.
Phan Ngoc Cu, Ton Si Kinh, Groundwater Mechanics, ed. Dai Hoc & Trung Hoc Chuyen Nghiep, Ha
Noi 1981.
NGUYEN The Hung, Finite element method in flow problems, Monograph NXB Xay Dung, Ha Noi
2004.
NGUYEN The Hung, Mathematical model of sediment transport two dimensional horizontal flow,
Proceedings of International Conference on Engineering Mechanics Today, Vol1, p.541-548, Hanoi
1995.
Jacob Bear and Arnold Verruijt , Modeling Groundwater Flow and Pollution, D. Reidel Publishing
Company 1979.