Feflow Reference Manual
Feflow Reference Manual
Feflow Reference Manual
cbcilt o
cáåáíÉ=bäÉãÉåí=pìÄëìêÑ~ÅÉ=cäçï
C=qê~åëéçêí=páãìä~íáçå=póëíÉã
oÉÑÉêÉåÅÉ=j~åì~ä
eKJgKdK=aáÉêëÅÜ=
aefJt^pv=dãÄe
`çéóêáÖÜí=åçíáÅÉW
kç=é~êí=çÑ=íÜáë=ã~åì~ä=ã~ó=ÄÉ=éÜçíçÅçéáÉÇI=êÉéêçÇìÅÉÇI=çê=íê~åëä~íÉÇ=ïáíÜçìí=ïêáííÉå=éÉêãáëëáçå=çÑ=íÜÉ=
ÇÉîÉäçéÉê=~åÇ=ÇáëíêáÄìíçê=aefJt^pv=dãÄeK
`çéóêáÖÜí=EÅF=NVVUI=OMMNI=OMMOI=OMMRI=OMMTI=OMMV=aefJt^pv=dãÄe=_Éêäáå=J=~ää=êáÖÜíë=êÉëÉêîÉÇK=
aefJt^pv=~åÇ=cbcilt=~êÉ=êÉÖáëíÉêÉÇ=íê~ÇÉã~êâë=çÑ=aefJt^pv=dãÄeK
aefJt^pv=dãÄeI===
t~äíÉêëÇçêÑÉê=píê~≈É=NMRI==aJNOROS=_ÉêäáåI=dÉêã~åó
mÜçåÉW=HQVJPMJST==VV=VUJMI=c~ñW=HQVJPMJST=VV=VUJVV
bJj~áäW=ã~áä]ÇÜáJï~ëóKÇÉ
áá=ö=oÉÑÉêÉåÅÉ=j~åì~ä
^`hkltibadbjbkq
cbcilt=ö=ááá
áî=ö=oÉÑÉêÉåÅÉ=j~åì~ä
`çåíÉåíë
`çåíÉåíë
m~êí=f==mÜóëáÅ~ä=_~ëáë=çÑ=jçÇÉäáåÖ
NK=_~ëáÅ=bèì~íáçåë K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K NP
OK=cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K QP
PK=_~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ =K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K RN
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë =K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K RT
cbcilt=ö=î
`çåíÉåíë
RK=^åáëçíêçéó K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K= TV
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt =K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K= UT
TK=oÉÑÉêÉåÅÉë=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K= NMP
m~êí=ff==^ééäáÅ~íáçåë
îá=ö=oÉÑÉêÉåÅÉ=j~åì~ä
`çåíÉåíë
NPK=eççéÉë=~åÇ=e~êäÉã~ååÛë=bñéÉêáãÉåíW=qÜÉ=qïç=tÉää=mêçÄäÉãK=K=K=K=K=K=K=K=K=K NQV
cbcilt=ö=îáá
`çåíÉåíë
NUK=d~êÇåÉêÛë=mêçÄäÉãW=råë~íìê~íÉÇ=pçáä=`çäìãå=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K= OOT
îááá=ö=oÉÑÉêÉåÅÉ=j~åì~ä
`çåíÉåíë
OPK=eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãåK=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K ORT
OQK=tÉåÇä~åÇ=~åÇ=eáããÉäëÄ~ÅÜÛë=bñéÉêáãÉåíW=pçäìíÉ=qê~åëéçêí=áå=~=qÜêÉÉJÇáJ
ãÉåëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉãK=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K OSP
cbcilt=ö=áñ
`çåíÉåíë
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë=K=K=K=K= OTN
^ìíÜçê=fåÇÉñ=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K=K= OVN
ñ=ö=oÉÑÉêÉåÅÉ=j~åì~ä
m~êí=f
mÜóëáÅ~ä=_~ëáë=çÑ=
jçÇÉäáåÖ
cbcilt=ö=NN
NO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
_~ëáÅ=bèì~íáçåë
N
density function and ψ ( t ) as an intensive balance
N _~ëáÅ=bèì~íáçåë
• mass conservation of fluid and solid continua All following considerations are related to a spa-
• mass conservation of contaminants and chemical tially fixed Eulerian coordinate system. Accordingly,
constituents the postulate of conservation of property G is stated as:
• momentum conservation of fluid and solid con-
tinua DG D-
Dt Dt ∫ ∫ ρf ( t ) dU
--------- ≡ ----- ρψ dU = (1-2)
• energy conservation (first law of thermodynam-
U U
ics).
Mass, motion and energy-related quantities can be where f corresponds to the specific rate of temporary
defined in a ’microscopic’ (local) volume element production (supply) of G and
(continuum) denoted by U, for which balance laws are
D- ∂ G ∂
postulated. Mass, linear momentum and energy repre- ----- = + vi (1-3)
Dt ∂ t ∂ xi
sent extensive properties G ( t ) (i.e., quantities are addi-
tive over volumes), which are dependent on the volume
of the continuum. On the other hand, intensive proper- is the material (substantial) time derivative, in which
G
ties g ( t ) = ρψ ( t ) concern densities of these extensive v i represents the velocity vector of the considered par-
properties being independent of the balance volume in ticle associated with the property G.
form of mass densities, momentum densities and
energy densities. In this context ρ is introduced as a For all vectorial and tensorial quantities having
cbcilt=ö=NP
NK=_~ëáÅ=bèì~íáçåë
repeated indices Einstein’s summation convention will by diffusion laws. Assuming a linear continuous rela-
be generally employed throughout the paper, e.g., tionship between the diffusive flux of G and the mass-
G ∂ G ∂ G ∂ G ∂
vi
∂ xi
≡ v1
∂ x1
+ v2
∂ x2
+ v3
∂ x 3 in the three-dimensional case.
weighted (barycentric) movement the diffusive flux
can be defined as
The balance law (1-2) can be expressed by using the
Reynolds transport theorem (e.g. Bear4) as follows: G
j i = ρψ ( v i – v i ) (1-6)
G
D- ⎛ D ( ρψ ) ∂v i ⎞ where v i represents the barycentric (mass-weighted)
Dt ∫ ∫ ⎜⎝ ----------------
----- ρψ dU = - + ρψ --------
-⎟ dU = (1-4)
Dt ∂x i ⎠ velocity.
U U
( ρψ )- ∂
∂-------------- G
= ∫ ∂t
+
∂ xi
( ρv i ψ ) dU = ∫ ρf dU Introducing (1-6) into (1-5) the conservation princi-
U U ple results in the general form:
Thus ∂--------------
( ρψ )- ∂ ∂j i
+ ( ρv i ψ ) + = ρf (1-7)
∂t ∂ xi ∂ xi
( ρψ )- ∂
∂-------------- G
+ ( ρv i ψ ) = ρf (1-5)
∂t ∂ xi
related to a miscroscopic (local) continuum, where
individual balance laws for ψ , j i and f of conserving
because the balance expression becomes independent mass, momentum and energy are summarized in Tab.
of the volume U. 1.1:
G
Generally, the velocities v i of property G are not
explicitly measurable and have to describe implicitly
Table 1.1 Microscopic balance laws
Quantity ψ ji f
Mass
Fluid mass 1 0 0
Contaminant mass ωk j ik rk
Momentum vi σ ij gi
1
Energy E + --- ( v i v i ) σ ij ⋅ v j + j iT gj ⋅ vj + QT
2
NQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKO=j~ÅêçëÅçéáÅ=`çåëÉêî~íáçå=bèì~íáçåë
Here i, j = 1, …, NDM are the indices of the Eulerian Phases α = 1, …, M are regarded as material sub-
spatial coordinates (summation convention!) with domains separated by phase interfaces (e.g., solid-
NDM = 3 for three-dimensional (3D) processes, solid, fluid-solid, fluid-fluid) where the system proper-
k = 1, …, N represents the index to indicate a chemical ties undergo an abrupt change (discontinuity). Each
α
species (component) with N as the total number of phase α is composed by N miscible chemical species
encountered species, ω k = C k ⁄ ρ is the mass fraction characterizing a molecular mixture of throughout dis-
with Ck as the concentration of contaminant species k , tributed several identifiable chemical components
α
j ik corresponds to the mass flux vector of the compo- k = 1, …, N , …, N , for which abrupt interfaces are not
nent k , r k is a partial chemical reaction rate, σ ij is the generally present. By definition a chemical species k
stress tensor, g i is the external supply of momentum only exists in one phase α or, if passing through differ-
(e.g., gravity field), E is the internal energy density, j iT ent phases, it is regarded as a phase-pertinent separate
is the heat flux vector, and Q T is the external supply (new) constituent, accordingly it holds N = ∑ Nα (see
due to heat and mass diffusion (for more see also Section 1.3). α
cbcilt=ö=NR
NK=_~ëáÅ=bèì~íáçåë
fault zone
porous aquifer
karstic aquifer
α
γ
β fissured rock
fracture matrix
rock mass
microfractures
multiphase system α−β−γ fractured rock masses
macrofractures
Figure 1.1 Representation of averaging volumes (REV) in the hydrogeologic system (modified from Diersch14
and Kolditz and Zielke50).
∂ ε α α ∂ ε α α α ∂ jα
( ρ E )+ ( ρ vi E ) + ( ) (1-12)
∂t α ∂ xi α ∂ x i iT
Contaminant mass conservation of a phase: α
α
= εα ρ QT
∂ ε α ∂ ε α α ∂ α
( C )+ ( v C )+ ( j ) = εα Rk (1-10)
∂t α k ∂ xi α i k ∂ x i ik
NS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKO=j~ÅêçëÅçéáÅ=`çåëÉêî~íáçå=bèì~íáçåë
α α αβ βα βα
( εα ρ ( wi – vi ) ) ⋅ ni = Q̂ ρi ⋅ n i (1-18)
Fourierian heat flux by a generalized thermodispersion α
approach:
where
α α α α α
j iT = j iT ( ε α, ρ , v i , T ) (1-14)
εα = volume fraction of α -phase, where
α 0 ≤ ε α ≤ 1 and ∑ εα = 1 ;
cond α disp α ∂T α α
= – ( λ ij + λ ij ) --------- ρ = density of α -phase;
∂x j α
vi = velocity vector of α -phase;
α α
α α α α α α v i v j ∂T α wi = macroscopic velocity of phase interface;
= – ε α ( λ + c ρ α T V )δ ij + c ρ ( α L – α T ) -----------
- --------- α
V
α ∂x j Ck = concentration of component k in the α -phase;
α
k ij = permeability tensor of α -phase;
α
Macroscopic reaction rate for homogeneous and heter- p = α -phase pressure;
α
ogeneous chemical reactions of component k: μ = dynamic viscosity of α -phase;
gi = gravity vector;
α het α
Rk = rk + Rk (1-15) E = internal (thermal) energy of α -phase;
α α
Q ρ , Q T∗ = mass and heat supply of α -phase,
respectively;
satisfying its balance constraint βα
Q̂ ρi = mass supply through a β-α-interface (positive
cbcilt=ö=NT
NK=_~ëáÅ=bèì~íáçåë
NU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKQ=jçÇÉäáåÖ=^ééêç~ÅÜ=~åÇ=páãéäáÑáÅ~íáçå
S k=1
b phase indicator ranges then over α = s, f , where s indi-
cates solid and f denotes fluid. In the present context
# S
where v is the decay frequency and b represents a fluid is represented by water.
reaction rate constant. The latter can be expressed by
the well-known phenomenological Arrhenius equation The phase volume fractions are now defined
given in the form according to: ε = ε f and ε s = 1 – ε , where ε repre-
sents the porosity of the porous medium. Additionally,
following relations may be appropriately introduced:
cbcilt=ö=NV
NK=_~ëáÅ=bèì~íáçåë
f f
ρ = ρ ( h, C, T ) (1-31)
The hydraulic head h (piezometric head) related to
f
the reference fluid density ρ o is defined as
which will be thoroughly discussed in Section 1.6.
f
p
h = --------
f
- + xl (1-27)
ρo g
NKQKO `ÜÉãáÅ~ä
under assuming exclusive action of gravity (cf. Eq. (1- Let us focus on a simplified chemical reaction sys-
11)) tem. On the one hand, reversible binary ion exchange
z z
reactions between species A 11 and species A 22 take the
g j = – ge j (1-28) form:
z1 z2 z1 z2
where the subscript l indicates the direction of gravity ν 1 A 1 ( aq ) + ν 2 A 2 ↔ ν 1 A 1 + ν 2 A 2 ( aq ) (1-32)
action in the Cartesian coordinate system, g is the grav-
itational acceleration, and e j denotes the gravitational
where aq indicates the component of the aqueous phase
unit vector.
f, the overbar symbols designate the sorbed compo-
f nents of solid phase s (exchange phase), and z k is the
The Darcy velocity (flux) q i is related to the fluid
f charge number of species (ion) k.
phase velocity v i (often referred to as pore, intrinsic or
kinematic velocity) by
On the other hand, each of these components should
f f be subjected to irreversible decay (disappearance)
q i = εv i (1-29) reactions according to
OM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKQ=jçÇÉäáåÖ=^ééêç~ÅÜ=~åÇ=páãéäáÑáÅ~íáçå
⎛ ⎞ ν1 ν2
s s s f
⎜ ⎟ 12 ( ρ C1 ⁄ CT ) ( f2 C2 )
zk 1 s ⎜ 1 ⎟ K eq = ---------------------------------------------------------
ν2
(1-39)
[ A k ] = ------ ρ ω k ⎜ ---------------------⎟ (1-35) f ν1 s s s
mk ⎜ α ( f1 C1 ) ( ρ C2 ⁄ CT )
Cq ⎟
⎜ ∑ ∑ -------⎟
⎝ mq⎠
α q
In the case of monovalence ν1 = ν 2 = 1 the sorbed
species may be explicitly expressed by the dissolved
where f k corresponds to the activity coefficients (cor- quantities in the fluid phase (restrictedly also for het-
recting factors against ideal mixtures), which can be erovalent chemical equilibrium reactions as more
assumed with f k = 1 (ideal-dilute solution) for the shown in Diersch19). Using the Eqs. (1-37), (1-38) and
sorbed phase, and m k is the molecular weight of spe- (1-39) the well-known (nonlinear) Langmuir adsorp-
cies k. The mass fractions sorbed on the solid phase ω k tion isotherm results directly from Eq. (1-39):
s
are related to the solid phase concentration C k via the
s
ion exchange capacity C T in form of f 1 K eq C T
12 s
C1
f
s f
C1 = f ( C1 ) = -------------------- ------------------------------------------------ (1-40)
f2 f
C T + ⎛⎝ ---- K eq – 1⎞⎠ C 1
f 1 12 f
s s f2
ωk = Ck ⁄ CT (1-36)
where written in the more common form
Ns
f
s s k1 C1
CT = ∑ Ck (1-37) s f
C 1 = f ( C 1 ) = --------------------f-
k 1 + k2 C1
(1-41)
12 s
f 1 K eq C T f 1 12
k 2 = ⎛ ---- K eq – 1⎞ ⁄ C T
f
when N s ≥ 2 denotes the total number of sorbed spe- -,
k 1 = -------------------
f ⎝ f2 ⎠
cies. Moreover, the total solution normality for the f2 CT
fluid phase f can be introduced as:
Admitting for low concentrations
Ns
f f
CT = ∑ Ck (1-38) k2 C1 « 1
f
(1-42)
k
cbcilt=ö=ON
NK=_~ëáÅ=bèì~íáçåë
∂ ( 1 – ε )C
( 2 ) = ( 1 – ε )R 2 (1-47)
where with κ the Henry sorptivity coefficient is intro- ∂t
duced. Otherwise, for high concentrations the sorbed
s
concentration C 1 for the Langmuir isotherm tends to Using the above adsorption isotherms (1-41), (1-43)
the limit k 1 ⁄ k 2 . and (1-44) as well as the definition (1-45) following
transport equation results after summation of Eqs. (1-
In the case of heterovalent equilibrium reactions 46) and (1-47):
ν 1 > 1, ν 2 > 1 polynomial relationships can be
s f a
derived C 1 ∼ f ( C 1 ) (see Diersch19), where a > 1 repre- ∂ εRC ∂ qf ∂ f
( 1) + ( C )+ ( j ) = εR 1 + ( 1 – ε )R 2 (1-48)
sents a corresponding exponent. An empirical approach ∂t ∂ x i i 1 ∂ x i i1
joining such a power character of the exchange reac-
tions provides the so-called Freundlich adsorption iso-
introducing the so-called retardation factor R, which is
therm written in the form:
defined as
s f f b2
C1 = f ( C1 ) = b1 ( C1 ) (1-44) (1 – ε)
R = 1 + ---------------- χ ( C 1 ) (1-49)
ε
OO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKQ=jçÇÉäáåÖ=^ééêç~ÅÜ=~åÇ=páãéäáÑáÅ~íáçå
⎧
⎪
⎪
⎨
⎪
⎪
⎩
leads to a linear expression: D ij
(r) + εRϑC = Q C
Rk = – ϑC k for k = 1, 2 (1-54) kçíáÅÉI= íÜÉ= ÇÉÅ~ó
ê~íÉ= ϑ = Å~å= ÄÉ
where ϑ is the decay rate‡) of component k, and the where C = C1 is the concentration of the essential ÉñéêÉëëÉÇ= Äó= íÜÉ
transport component in the fluid (aqueous) phase and êÉ~Åíáçå=Ü~äÑJäáÑÉ= t 1 ⁄ 2 =áå=íÜÉ
‡)
Referring to radioactive decay processes the decay rate ϑ is
frequently expressed in terms of a reaction half-life t 1 ⁄ 2 , which Q C = εQ C1 + ( 1 – ε )Q C2 ln2
(1-57) ÑçêãW= ϑ = --------
-K
is a specific solution of the reaction equation t1 ⁄ 2
dC k
dt
= – ϑC k represents a zero-order production term encompassing
the fluid and solid phase.
applied to a simple chemical batch reaction (without diffusion/
dispersion and advection). Its analytical solution yields
– ϑt
C k = C ok e
NKQKP qÜÉêã~ä
The time t for the concentration C k to decrease from the initial
concentration C ok to its half value 1--2- C ok is the half-life t 1 ⁄ 2 . For the fluid phase f and the rigid solid phase s the
From above it results individual energy conservation equations are in the fol-
ln2
t 1 ⁄ 2 = -------- lowing forms, respectively:
ϑ
cbcilt=ö=OP
NK=_~ëáÅ=bèì~íáçåë
⎧
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎩
conductive
⎧
⎪
⎪
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎪
⎪
⎩
where T is the system temperature. Moreover, it is dispersive
α
assumed the state of internal energy E of the phases
should be primarily a function of temperature (see also
f f s s
Section 1.6): Q T = ερ Q T + ( 1 – ε )ρ Q T (1-64)
α α α
α ∂E α ∂E α ∂E α α α f
where V q =
f f
q i q i designates the absolute Darcy flux.
dE = ---------
α
dρ + ∑ ---------α- dC k + ---------
α
dT ≈ c dT (1-61)
∂ρ k ∂C k ∂T
for α = s, f
α ∂E
α NKR ^èìáÑÉêJ~îÉê~ÖÉÇ= bèì~J
with the heat capacity c = ---------α of α -phase.
∂T
íáçåë
Accordingly, the heat transport equations (1-58) and Transport process modeling in aquifers possesses
(1-59) can be combined (added). As a result, energy important special cases, in which the horizontal extent
equations for each phase need not be considered. of a regional flow field much extends the thickness B
Rather the total energy conservation equation is pre- of an aquifer. For such conditions vertical variations
ferred. After summation of Eqs. (1-58) and (1-59) it can be often neglected to reduce the full 3D equations
yields: to two-dimensional (2D) essentially horizontal rela-
tionships. Regarding groundwater hydraulic processes
∂ ερ f f s s ∂ ρf f f this procedure is associated with the well-known
[ E + ( 1 – ε )ρ E ] + ( qi E ) (1-62)
∂t ∂ xi
Dupuit assumption (Bear4,5).
∂T
– ∂ ⎛ λ ij -------⎞ = Q T
∂ x i ⎝ ∂x j⎠
There are two distinctly different approaches to
integrate vertically the macroscopic balance equations.
In forming the above total energy conservation equa- A two-step averaging procedure (first step represents a
tion (1-62) the following thermodispersive and heat REV averaging, second step is the vertical averaging)
OQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKR=^èìáÑÉêJ~îÉê~ÖÉÇ=bèì~íáçåë
dS
Vertically averaged energy conservation of a phase:
∂ ( ε ρ α BE α ) + ∂ ( ε ρ α Bv α E α )
i (1-68)
∂t α ∂ xi α
∂ Bj α α α
+ ( iT ) = ε α ρ BQ T
∂ xi
H
B
For the horizontal equations the following defini-
tions are appropriately introduced:
x3
x2 Depth-integrated Darcy flux:
D
x1 f f f
Figure 1.2 Aquifer averaging volume (AREV). q i = Bq i = εBv i (1-69)
cbcilt=ö=OR
NK=_~ëáÅ=bèì~íáçåë
zontal processes. Analogously to the simplifications as a function of the hydraulic head h, the concentration
done in Section 1.4, the resulting balance equations for C, and the temperature T, viz.,
essentially horizontal flow and transport processes are
as follows: f f
ρ = ρ ( h, C, T ) (1-75)
f ∂B
B ∂ ( ερ ) + ερ ------ +
f ∂ ρf f f
( qi ) = ρ Qρ (1-71) The total differential yields then
∂t ∂t ∂ x i
f f f
f ∂ρ ∂ρ ∂ρ
f ∂h dρ = ------- dh + ------- dC + ------- dT
∂h ∂C ∂T
qi = – T ij ------- (1-72)
∂x j
= ⎛ 1 ∂ρ f⎞ f
ρ dh + ⎛ 1 ∂ρ f⎞ f
ρ dC + ⎛ 1 ∂ρ f⎞ f
ρ dT (1-76)
∂ ( εRBC ) + ∂ q f – ∂ B D mol f disp f ∂C ⎜ ----f -------⎟
⎝ ρ ∂h ⎠
⎜ ----f -------⎟
⎝ ρ ∂C ⎠
⎜ ----f -------⎟
( C) ( ij + D ij ) ------- (1-73) ⎝ ρ ∂T ⎠
∂t ∂ xi i ∂ xi ∂x j
⎧
⎪
⎨
⎪
⎩
⎧
⎪
⎨
⎪
⎩
⎧
⎪
⎨
⎪
⎩
γ α ⁄ ( Cs – Co ) –β
+ εRBϑC = Q C
f α f f
= γρ dh + ----------------------- ρ dC – βρ dT
( Cs – Co )
∂ B ερ f f s s ∂ ρf f f
[ ( E + ( 1 – ε )ρ E ) ] + ( qi E ) (1-74)
∂t ∂ xi
∂T
where γ is the fluid compressibility, α is the fluid den-
– ∂ ⎛⎝ Bλ ij -------⎞⎠ = Q T sity difference ratio, which is normalized by the differ-
∂ xi ∂x j
ence ( Cs – C o ) of the maximum concentration Cs and
the reference concentration Co , and β is the volumet-
ric thermal fluid expansion coefficient to be introduced
NKS bèì~íáçåë=çÑ=pí~íÉ as intensive (volume-independent) material coeffi-
cients. If we assume γ , α ⁄ ( C s – C o ) and β are con-
The derived balance equations (1-9) to (1-12), (1- stants the integration of Eq. (1-76) immediately leads
56), (1-62), (1-65) to (1-68) and (1-71) to (1-74) reveal to the state equation for the fluid density in the com-
dependencies upon state variables which have to be mon form
more investigated. In this context the relationships for
f
the fluid density ρ , the porosity ε , the internal energy α
γ ( h – h o ) + ----------------------- ( C – C o ) – β ( T – T o )
α f f ( Cs – Co )
density E , the aquifer thickness B, and the dynamic ρ = ρo e (1-77)
f
viscosity μ of the fluid are of further interest:
f α
≈ ρ o 1 + γ ( h – h o ) + ----------------------- ( C – C o ) – β ( T – T o )
f ( Cs – Co )
NKSKN cäìáÇ=ÇÉåëáíó= ρ
f where suitable reference values appear for the fluid
Generally, the fluid density ρ should be regarded
OS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKS=bèì~íáçåë=çÑ=pí~íÉ
f f
density ρ o = ρ h o, C o, T o as the reference hydraulic head in the range between 0 and 100 °C . kçíáÅÉI= ïÜáäÉ= íÜÉ
h o , the reference concentration C o , and the reference ÅçåÅÉåíê~íáçåJ
temperature T o . The above introduced material coeffi- As a consequence the volumetric thermal fluid î~êá~ÄäÉ=ÑäìáÇ= den-
cients may be estimated by appropriate relations as expansion relation β cannot generally be considered as sáíó= áë= ëìÑÑáÅáÉåíäó
exemplified for the density difference ratio α ‡) a constant coefficient and a more general approach is ÇÉëÅêáÄÉÇ= Äó= ~= äáåÉ~ê
desired. Perrochet57 proposed a formula for computing êÉä~íáçåëÜáé= áå= Ñçêã= çÑ
f
α = [ ρ ( Cs ) – ρo ] ⁄ ρo
f f
(1-78) a nonlinear temperature-dependent expansion coeffi- íÜÉ= ÇáÑÑÉêÉåÅÉ= ê~íáç= α I
cient β ( T ) which is based on the physically ’exact’ Ñçê= íÜÉ= íÉãéÉê~íìêÉJ
fluid density curve (Fig. 1.3) measured for tempera- î~êá~ÄäÉ= ÑäìáÇ= ÇÉåëáíó= ~
f
A first guess of the fluid density ρ at maximum con- ãçêÉ= ÖÉåÉê~ä= ÅçåëíáíìJ
tures in the range between 0 and 100 °C and represents
centration C s can be given by the relation (Baxter and íáîÉ= êÉä~íáçåëÜáé= β = Å~å
a correction parameter for the state equation (1-77) by ÄÉ= åÉÅÉëë~êóI= é~êíáÅìJ
Wallace1) introducing higher-order terms. This approach has been ä~êäó=áÑ=~=ä~êÖÉê=íÉãéÉê~J
f f
implemented in FEFLOW as the option termed as ther- íìêÉ= ê~åÖÉ= Ü~ë= íç= ÄÉ
ρ ( C s ) ≈ ρ o + 0.7 ⋅ C s (1-79) mally variable fluid density expansion. ãçÇÉäÉÇ= çêL~åÇ= íÜÉ= Q
°C =~åçã~äó=áå=íÜÉ=ÑäìáÇ
f
which leads to an estimation of density difference ratio Perrochet57 has fit the temperature-dependent fluid Eï~íÉêF= ÇÉåëáíó= ρ = Å~å
f éä~ó= ~= êçäÉ= EÉKÖKI= ÅçäÇ
α in form of density ρ ranging between 0 and 100 °C by a 6th
order polynomial with a high accuracy as ÖêçìåÇï~íÉê= ~ë= ëÜçïå
0.7 ⋅ C s áå=eçäòÄÉÅÜÉê40FK
α ≈ -----------------
f
- (1-80)
ρo
cbcilt=ö=OT
NK=_~ëáÅ=bèì~íáçåë
2 ⎫
a = 9.998396 ⋅ 10 ⎪
1000 –2 ⎪
b = 6.764771 ⋅ 10 ⎪
–3 ⎪
c = – 8.993699 ⋅ 10 ⎪
⎪
d = 9.143518 ⋅ 10
–5
⎬ (1-82)
⎪
990 e =
–7
– 8.907391 ⋅ 10 ⎪
⎪
f = 5.291959 ⋅ 10
–9 ⎪
⎪
– 1.359813 ⋅ 10 ⎪⎭
– 11
g =
980
where the temperature T is in °C and a represents the
ρf [g/l] 1000
fluid density at T = 0 . Taking also into consideration
of concentration and pressure dependencies we can
970 argue the measured curve (1-81) is related to a refer-
ence concentration Co and reference hydraulic head
h o . Indeed, the coefficients (1-82) has been derived for
960
999
0 5 10
a freshwater condition C o = 0 . Introducing a reference
temperature T o we can then obtain directly from (1-81)
f
variable expansion an expression for the reference fluid density ρ o , viz.,
linear expansion
f f
950 ρ o = ρ ( h o, C o, T o ) (1-83)
0 20 40 60 80 100
2 3
T [oC] = a h , C + b h , C To + c h , C To + d T
h o, C o o
o o o o o o
f
Figure 1.3 Fluid density ρ as a function of tempera- 4 5
+ e h , C To + f h , C To + g h , C To
6
ture T in the range of 0 and 100 °C , close-up view o o o o o o
OU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKS=bèì~íáçåë=çÑ=pí~íÉ
f
ρ around T o , C o and h o : f f α
ρ ( T, C, h ) = ρ o 1 + ----------------------- ( C – C o ) + γ ( h – h o ) (1-87)
( Cs – Co )
f 2 3 4 5
ρo + ( b + 2cT o + 3dT o + 4eT o + 5fT o + 6gT o ) ( T – T o )
f 2 f
∂ρ 1∂ ρ
⎫
⎪
⎪
⎬
⎪
⎪
⎭
f f 2
ρ ( T, C, h ) = ρ ( T o, C o, h o ) + ------- ( T – T o ) + --- ---------2- ( T – To ) + 2 3
+ ( c + 3dT o + 6eT o + 10fT o + 15gT o ) ( T – T o )
4 2
∂T 2 ∂T
T o, C o, h o
T o, C o, h o
2 3 3
3 f 4 f 5 f + ( d + 4eT o + 10fT o + 20gT o ) ( T – T o )
--- ρ-
∂
1 --------- 1∂ ρ
3 4 1 ∂ ρ 5
( T – T o ) + ------ ---------4- ( T – T o ) + --------- ---------5- ( T – To ) +
6 ∂T 3 24 ∂T 120 ∂T 2 4
T o, C o, h o T o, C o, h o T o, C o, h o + ( e + 5fT o + 15gT o ) ( T – T o )
5
f
+ ( f + 6gT o ) ( T – T o )
f
1 ∂ ρ
6 f
6 ∂ρ ∂ρ
--------- ---------- ( T – T o ) + ------
∂C
- ( C – C o ) + ------- ( h – ho )
720 ∂T 6 T o, C o, h o ∂h T o, C o, h o + g ( T – To )
6
T o, C o, h o
⎧
⎪
⎨
⎪
⎩
⎧
⎪
⎨
⎪
⎩
f α f
ρ o ----------------------- ρo γ
( Cs – Co )
(1-85)
From Eqs. (1-87) and (1-84) we finally obtain the
expression for computing the nonlinear thermally vari-
where a 6th order approximation is used for the tem- able fluid density expansion to be used in Eq. (1-84)
perature T while only a linear 1st order approximation
is deemed to be sufficient for C and h dependencies. By 1
β ( T ) = – ----f- [ 2 3
( b + 2cT o + 3dT o + 4eT o + 5fT o + 6gT o ) +
4 5
ρo
utilizing the Eq. (1-81) we can evaluate the above
n f n 2 3
( c + 3dT o + 6eT o + 10fT o + 15gT o ) ( T – T o ) +
4
∂ ρ ⁄ ∂T at T o , C o and h o according to
2 3 2
( d + 4eT o + 10fT o + 20gT o ) ( T – T o ) + (1-88)
f
∂------
ρ- 2 3 4 5⎫ ( e + 5fT o +
2
15gT o ) ( T
3
– To ) +
= b + 2cT o + 3dT o + 4eT o + 5fT o + 6gT o ⎪
∂T T o, C o, h o ⎪ 4
⎪ ( f + 6gT o ) ( T – T o ) +
2 f
∂ ρ- 2 3 4 ⎪
--------- = 2c + 6dT o + 12eT o + 20fT o + 30gT o ⎪ g ( T – To )
5
]
2
∂T T o, C o, h o ⎪
⎪
3 f
∂ ρ- 2 3
⎪
--------- = 6d + 24eT o + 60fT o + 120gT o ⎪ f
∂T
3
T o, C o, h o
⎪
⎪
with ρ o computed from (1-83) and the coefficients
4 f ⎬ (1-86) ( a, b, c, d, e, f, g ) as given in (1-82).
∂ ρ-
--------- 2 ⎪
4
= 24e + 120fT o + 360gT o ⎪
∂T T o, C o, h o ⎪
5 f
⎪
⎪
As the consequence from the nonlinear thermally
∂ ρ- f f
---------
5
= 120f + 720gT o ⎪
⎪
variable fluid density expansion the term ( 1 ⁄ ρ ) ∂ρ ⁄ ∂T
∂T T o, C o, h o ⎪
⎪
in Eq. (1-76) is not longer a constant, β . Instead we
6 f
∂ ρ-
--------- ⎪ have
6
= 720g ⎪
∂T T o, C o, h o ⎭
f
As the result the fluid density ρ (1-85) gives
cbcilt=ö=OV
NK=_~ëáÅ=bèì~íáçåë
∂β ( T )
β ( T ) + --------------- ( T – T o )
α
1 ∂ ρ
f
∂T NKSKP fåíÉêå~ä=ÉåÉêÖó=EÇÉåëáíóF= E
---- ------- = – --------------------------------------------------------------------------------------------
- (1-89)
f ∂T α
ρ 1 + ----------------------- ( C – C o ) – β ( T ) ( T – T o )
( Cs – Co ) The internal energy is generally a state function of
α α
the density ρ , the concentration C k , and the tempera-
α
ture T of α -phase:
where here effects of fluid compressibility can be
neglected. The derivation of the nonlinear thermally
variable fluid density expansion in Eq. (1-89) becomes α α α
α ∂E α ∂E α ∂E dT α
dE = ---------
α
dρ + ∑ ---------α- dC k + --------- (1-93)
∂ρ α
∂β ( T ) 1
--------------- = – ----- [ ( c + 3dT o + 6eT 2o + 10fT 3o + 15gT 4o ) + k ∂C k ∂T
∂T f
⎧
⎨
⎩
ρo
α
2
2 ( d + 4eT o + 10fT o + 20gT o ) ( T – T o ) +
3 c
(1-90)
2
3 ( e + 5fT o + 15gT o ) ( T – T o ) +
2
for α = s, f
3
4 ( f + 6gT o ) ( T – T o ) +
4
5g ( T – T o ) ]
Commonly, it is deemed to be sufficient that the inter-
α
nal energy E only represents a state relation with
NKSKO mçêçëáíó= ε α
respect to the temperature T . Thus
α α α
In changing the porosity of the porous medium dE = c dT for α = s, f (1-94)
compression work of the skeleton (deformation) should
be taken into account, where concentration and temper- α
where c corresponds to the heat capacity of α -phase.
ature effects remain disregarded. Thus
ε = ε(h) (1-91)
NKSKQ ^èìáÑÉê=íÜáÅâåÉëë=_
The differential becomes The thickness of an aquifer B = B ( x i, t ) can be
mapped as an interface equation. Since the interface
∂ε
may be regarded as a material surface of discontinuity
1 ∂ε⎞ ε dh = ϒ ( 1 – ε )dh
dε = ----- dh = ⎛ --- (1-92)
∂h - -----
⎝ ε s ∂h⎠ s
occupied by fluid and solid particles, principles of con-
servation have to be subjected (Bear and Bachmat6):
⎧
⎨
⎩
ϒ
∂-----
B- ∂B ∂B- ∂B- n
+ w i ------- = 0 , ------ = ------ i (1-95)
∂t ∂x i ∂x i ∂x i
where ϒ represents the coefficient of skeleton com-
pressibility.
PM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKS=bèì~íáçåë=çÑ=pí~íÉ
cbcilt=ö=PN
NK=_~ëáÅ=bèì~íáçåë
f f
where μ o = μ Co, To is the reference viscosity with 2.0
respect to the reference concentration Co and the refer- /l]
0 [g
=
ence temperature T o . C 50
0
1.5 10
0
15
The consequences of viscosity and density effects to 200
f
the input parameter of hydraulic conductivity are more μo
-------------------
- 1.0
f
discussed below in Section 1.8. For instance, the vis- μ ( C, T )
f f
cosity relation f μ = μ o ⁄ μ ( C, T ) is displayed in Figs.
1.4 and 1.5 for the temperature range T between 0 and 0.5
]
g/ l
0[
= 0
C 5
0
10.0 10
0
15
f
μo 200
--------------------
f
μ ( C, T )
5.0
0.0
0 100 200 300
T [ °C ]
f f
Figure 1.5 Viscosity relation μ o ⁄ μ as function of tempera-
f
ture T [ °C ] and concentration C [ g ⁄ l ] with μ o for the ref-
erence temperature of T o = 10 °C and reference
concentration of C o = 0 (freshwater).
PO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKT=péÉÅáÑáÅ~íáçå=çÑ=aÉéÉåÇÉåí=_~ä~åÅÉ=qÉêãë
f f
The conservation equation (1-9) for the fluid phase
Figure 1.6 Relative viscosity μ ⁄ μ o as function of tem- f, starting from the basic form
f
perature T [ °C ] and concentration C [ g ⁄ l ] with μ o for
the reference temperature of T o = 150 °C and reference ∂ ερ f ∂ ρf f f
concentration of C o = 0 (freshwater). ( )+ ( qi ) = ρ Qρ (1-102)
∂t ∂ xi
2.5 f
with Q ρ = εQ ρ
2.0
⎧
⎨
⎩
⎧
⎨
⎩
⎧
⎨
⎩
0.5
1 2 3 4
0.0
0 50 100 150 The first term of Eq.(1-103) can be developed by using
T [ °C ] the state equation (1-76) as
f f
Figure 1.7 Relative viscosity μ ⁄ μ o as function of tem- f
∂ρ f ∂h α ∂C ∂T
f
perature T [ °C ] and concentration C [ g ⁄ l ] with μ o for ε ------- = ερ ⎛ γ ----- + ----------------------- ------ – β∗ ------⎞ (1-104)
∂t ⎝ ∂t ( C s – C o ) ∂t ∂t ⎠
the reference temperature of T o = 10 °C and reference
concentration of C o = 0 (freshwater).
cbcilt=ö=PP
NK=_~ëáÅ=bèì~íáçåë
α ∂C ∂T
⎧β for the constant expansion coefficient Q EB ( C, T ) = – ε ⎛⎝ ----------------------- ------ – β∗ ------⎞⎠ (1-110)
⎪ ( C s – C o ) ∂t ∂t
⎪
⎪ (1-105) α ∂C ∂T
– q i ⎛⎝ ----------------------- ------- – β∗ -------⎞⎠
f
⎪ ∂β ( T )
β∗ = ⎨ β ( T ) + --------------- ( T – T o ) ( C s – C o ) ∂x i ∂x i
∂T
⎪ ----------------------------------------------------------------------------------------- for the thermally variable
⎪ - fluid density expansion
α
⎪ 1 + --------------------
⎪
- ( C – C o ) – β ( T ) ( T – T o )
( Cs – Co ) represents an additional term to incorporate mass-
⎩
dependent and temperature-dependent compression
where β is a given constant while β ( T ) and ∂β ( T ) ⁄ ∂T effects as required for the extended Boussinesq approx-
correspond to Eq. (1-88) and (1-90), respectively. imation (see the subsequent Section 1.10 for more),
where compression work of pressure has been
f ∂h
The second term is provided by using the constitu- neglected q i γ ------- ≈ 0 .
∂x i
tive equation (1-92)
Analogously to relationship (1-104), the third term For essentially horizontal problems density effects
yields resulting from concentration and/or temperature influ-
ences do not play a role. Accordingly, the mass conser-
f ∂ρ
f
f f ∂h α ∂C ∂T vation equation (1-73) for the fluid phase f becomes
q i ------- = ρ q i ⎛ γ ------- + ----------------------- ------- – β∗ -------⎞ (1-107)
∂x i ⎝ ∂x i ( C s – C o ) ∂x i ∂x i⎠
f ∂B
B ∂ ( ερ ) + ερ ------ +
f ∂ ρf f f
( qi ) = ρ Qρ (1-111)
∂t ∂t ∂ x i
Accordingly, the continuity equation (1-103) leads to
the following form as actually realized in FEFLOW:
It can be further developed as follows:
f
∂h ∂q i
S o ----- + ------- = Q ρ + Q EB ( C, T ) (1-108) ∂ρ
f
f ∂ε B- + f ∂------
ρ
f
f ∂q i f
f
(1-112)
∂t ∂x i εB ------- + ρ B ----- + ερ f ∂----- q i - + ρ ------- = ρ Q ρ
∂t ∂t ∂t ∂x i ∂x i
⎧
⎨
⎩
⎧
⎨
⎩
⎧
⎨
⎩
⎧
⎨
⎩
⎧
⎨
⎩
where 1 2 3 4 5
PQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKT=péÉÅáÑáÅ~íáçå=çÑ=aÉéÉåÇÉåí=_~ä~åÅÉ=qÉêãë
f f
∂ρ f ∂ε ∂h ∂p- ∂h
= ρ o g ⎛ ------- – e j⎞
f f
εB ------- + ρ B ----- = ρ B S o ----- (1-113) ------ (1-117)
∂t ∂t ∂t ∂x j ⎝ ∂x j ⎠
with the specific storage coefficient S o according to Substituting it into Eq. (1-11) the Darcy’s law in terms
Eq. (1-109). The third term in Eq. (1-112) becomes, of hydraulic head and under consideration of density
after applying the state relation (1-96) effects takes finally the form
f ∂B f ∂h f ∂h f f
ερ ------ = ρ εξ ----- = ρ ε e ----- (1-114) ⎛ ∂h ρ – ρ o ⎞
∂t ∂t ∂t f
q i = – K ij f μ ⎜ ------- + ---------------
- e j⎟ (1-118)
⎝ ∂x j ρo
f
⎠
where ε e = ε ξ at 0 ≤ ξ ≤ 1 is the effective, so-called
drainable or fillable porosity. For confined conditions where the constitutive equation (1-101) for the
in an aquifer the effective porosity vanishes ε e ≡ 0 . dynamic viscosity has been incorporated in introducing
the hydraulic conductivity according to
Neglecting the fourth term in Eq. (1-112) the hori-
zontal fluid mass conservation equation finally results
f f f f
k ij k ij ρ o g μo 1- 1-
f
∂q i - ⋅
----- = -------------- ------ ⋅ -------- = K ij f μ -------- (1-119)
∂h μ
f
μo
f
μ
f
ρo g
f
ρo g
f
( S o B + ε e ) ----- + ------- = Q ρ (1-115)
∂t ∂x i
⎧
⎨
⎩
⎧
⎨
⎩
K ij fμ
NKTKP a~êÅó=Éèì~íáçå The so-called buoyancy term in the Darcy equation (1-
118) can be expanded as follows:
The momentum conservation equation (1-11) of the
f f
fluid phase f is to be expressed by the hydraulic head h, ρ – ρo α -
according to Eq. (1-27), the gravitational vector (1-28),
---------------- = ---------------------- ( C – Co ) – β ( T – To ) (1-120)
ρo
f ( C s – Co )
f
the Darcy velocity q i , and the hydraulic conductivity
f f
K ij . From h = p ⁄ ( ρ o g ) + x l it yields for the fluid pres-
sure where here effects of pressure compression need not to
be regarded for the groundwater flow problems under
f f
p = ρo g ( h – xl ) (1-116) consideration.
and accordingly
cbcilt=ö=PR
NK=_~ëáÅ=bèì~íáçåë
f
NKU aÉåëáíó=~åÇ=sáëÅçëáíó=oÉÑJ ture effects on the fluid density ρ and the fluid viscos-
f
ÉêÉåÅÉ= `çåÇáíáçåë= Ñçê ity μ are implied by the buoyancy term (1-120) and
the viscosity function f μ (1-101), respectively. The vis-
eóÇê~ìäáÅ=`çåÇìÅíáîáíó cosity influence provides an effective (intrinsic) con-
a
ductivity K ij in the form:
In transforming the pressure-permeability-(p-k)-
form of the Darcy equation (1-11) to a hydraulic-head- f
conductivity-(h-K)-form of the Darcy equation (1-118) a
μ( C , T )
K ij = K ij ⋅ f μ = K ij ⋅ -------------------
f
o o
- (1-123)
the hydraulic conductivity has been introduced as μ ( C, T )
f f
k ij ρ o g For comparison purposes the viscosity influence
K ij = --------------
f
- (1-121)
μo may be explicitly excluded in FEFLOW (it is default),
where the viscosity function f μ is forced to be f μ ≡ 1
f for each case independent of the actually appearing
which depends, among the permeability k ij providing a concentration or temperature relations. Then the effec-
property of the porous medium, and the gravitational tive conductivity corresponds exactly to the input
acceleration g accepted as a constant for earth condi- a
parameter K ij = Kij . This special case is generally
tions, upon two additional reference variables of the given for pure flow problems.
f
fluid phase, viz., the reference density ρ o and the refer-
f
ence viscosity μ o .
f f
Both ρ o and μ o are basically related to reference NKV aáîÉêÖÉåÅÉ= ~åÇ= `çåîÉÅJ
conditions for the hydraulic head h o , the concentration íáîÉ= cçêãë= çÑ= qê~åëéçêí
C o and the temperature T o as expressed by the consti- bèì~íáçåë
f
tutive equation (1-77) for the fluid density ρ and by
the Eq. (1-101) for the viscosity relation function f μ . From the conservation principle (1-7) the following
According to Eq. (1-121) the conductivity is to be spec- transport equations have been derived in the above sec-
ified over a range of appropriately referenced magni- tions for the contaminant mass (cf. Eq. (1-56))
tudes of head, concentration and temperature, viz.,
∂ ( εRC ) + ∂ q f – ∂ ⎛ D ------
∂C-⎞
( C) + εRϑC = Q C (1-124)
f f ∂t ∂ xi i ∂ x i ⎝ ij ∂x j⎠
k ij ρ ( h , C , T ) g
K ij = ----------------------------------
o o o
- (1-122)
f
μ( C , T )
o o and for the heat (cf. Eq. (1-62))
PS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKV=aáîÉêÖÉåÅÉ=~åÇ=`çåîÉÅíáîÉ=cçêãë=çÑ=qê~åëéçêí=bèì~íáçåë
∂ ερ f f s s ∂ ρf f f ∂ ερ f f s s ∂ ρf f f
[ E + ( 1 – ε )ρ E ] + ( qi E ) (1-125) [ ( c + ( 1 – ε )ρ c )T ] + ( qi c T ) (1-128)
∂t ∂ xi ∂t ∂ xi
∂T ∂T
– ∂ ⎛ λ ij -------⎞ = Q T – ∂ ⎛ λ ij -------⎞ = Q T
∂ x i ⎝ ∂x j⎠ ∂ x i ⎝ ∂x j⎠
These formulations are called divergence forms due to The mass conservation (continuity) equation (1-9) for
the divergent expressions for the properties to be con- the fluid phase f and the solid phase s, respectively,
served.
∂ ( ερ f ) + ∂ ( ρ f q f ) = ερ f Q f = ρ f Q
i ρ ρ⎫
The divergence form of the heat transport equation ∂t ∂ xi ⎪
⎬ (1-129)
(1-125) can be completely expressed by the system ∂ ( 1 – ε )ρ s
) = ( 1 – ε )ρ Q ρ ≈ 0 ⎪⎭
s s
(
temperature T if the equations of state for the internal ∂t
f s
energies E and E for the fluid phase f and solid phase
s, respectively, are applied. According to Eq. (1-94) can be utilized to substitute the divergent terms in the
α α α
with dE = c dT for α = s, f the state relation after above transport equations (1-124) and (1-125). Notice
integration gives s
that the mass source term Q ρ for the solid phase is
assumed to be negligible. The substitution finally
α α α α
E = c ( T – To ) for α = s, f (1-126) results the following transport equations for the con-
taminant mass
f s
Since T = T = T according to the thermal equi-
∂C f ∂C ∂C-⎞
∂ ⎛ D ------
librium assumption (1-60) as stated above the internal εR d ------ + q i ------- – + ( εRϑ + Q ρ )C = Q C (1-130)
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠
energies for the fluid phase f and solid phase s are given
by
and for the heat
f f ⎫
E = c ( T – To ) ⎪ f f s s ∂T f f f ∂T-
⎬ (1-127) [ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------ (1-131)
s s ∂t ∂x i
E = c ( T – To ) ⎪
⎭ ∂T-⎞
∂ ⎛ λ ------ f f
– + ρ c Qρ ( T – To ) = QT
∂ x i ⎝ ij ∂x j⎠
where T o is the reference temperature as introduced in
Section 1.6. which are called convective forms.
Accordingly, the divergence form of the heat trans- The divergence forms (1-124), (1-128) and the con-
port equation (1-125) can now be written for the tem- vective forms (1-130), (1-131) of the transport equa-
perature T as tions are physically completely equivalent. The
cbcilt=ö=PT
NK=_~ëáÅ=bèì~íáçåë
For the 2D horizontal transport equations (1-73), (1- For the most density-dependent transport phenom-
74) written in their divergence forms ena the so-called Boussinesq approximation (pre-
ciously referred to as Oberbeck-Boussinesq
∂ ( εRBC ) + ∂ ( q f C ) – ∂ B ( Dmol + D disp ) ∂------ ⎫
i
f f
ij ij
C-
+ εRBϑC = Q C ⎪ approximation) is employed (Bird et al.9, Combarnous
∂t ∂ xi ∂ xi ∂x j ⎪
⎬ and Bories13, Nield and Bejan55). On this basis the cou-
∂ [ B ( ερf c f + ( 1 – ε )ρ s c s )T ] + ∂ ( ρ f q f c f T ) – ∂ ⎛ Bλ ------∂T ⎞ ⎪
∂t ∂ xi
i -
∂ x i ⎝ ij ∂x j⎠
= Q T⎪ pled nonlinear system of balance equations can be
⎭
essentially simplified without the elimination of intrin-
(1-132) sic coupling mechanisms significant for free convec-
tion processes. The Boussinesq approximation is valid
equivalent convective forms of the transport equations for small to moderate density variations due to concen-
can be derived as tration and/or temperature changes. It consists of
neglecting all density dependencies in the balance
∂------
C f ∂ C ∂ mol disp ∂ C ⎫ terms, except the vital buoyancy term in the momen-
εR d B + q i ------- –
f f
B ( D ij + D ij ) ------- + ( εRϑ + Q ρ )BC = Q C ⎪
∂t ∂x i ∂ x i ∂x j ⎪ tum conservation (Darcy-) equation. Accepting the
⎬
s s ∂T f f f ∂T ∂T Boussinesq approximation the following coupled non-
B [ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- – ∂ ⎛⎝ Bλ ij -------⎞⎠ + ρ c Q ρ B ( T – T o ) = Q T ⎪⎪
f f f f
∂t ∂x i ∂ x i ∂x j
⎭ linear system of balance equations results:
(1-133)
f
∂h ∂q i
S o ----- + ------- = Q ρ (1-134)
∂t ∂x i
PU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKNN=`Ü~ê~ÅíÉêáòáåÖ=cêÉÉ=`çåîÉÅíáçå=mêçÅÉëëÉë
f ⎛ ∂h ρ – ρ o ⎞
f f ature vanish. However, just the evolving features of a
q i = – K ij f μ ⎜ ------- + ---------------
- e j⎟ (1-135) convection process may be thoroughly affected at
⎝ ∂x j
f
ρo ⎠
∂h α
higher density contrasts (problems of bifurcation, phys-
= – K ij f μ ------- + ⎛ ----------------------- ( C – C o ) – β ( T – T o )⎞ e j
∂x j ⎝ ( C s – C o ) ⎠ ical instability and hydrodynamic pattern formation).
The second term Q EB ( C, T ) of (1-138) only vanishes if
∂ ∂ ⎛ f ∂C the density gradient is essentially orthogonal to the
( εRC ) + q C – D ij -------⎞ + εRϑC = Q C divergence form
∂t ∂ xi ⎝ i ∂x j⎠ flow. This is quite often not a tolerable assumption.
∂C f ∂C ∂C
∂ ⎛ ------
εR d ------ + q i ------- – D -⎞ + ( εRϑ + Q ρ )C = Q C convective form
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠ FEFLOW also enables the computation by employ-
(1-136) ing the so-called extended Boussinesq approximation.
Here the fluid mass conservation equation (1-134) is
∂ ∂ f f f – ∂ ⎛ ------ ∂T used in a generalized form:
λ -⎞ = Q T
f f s s
[ ( ερ c + ( 1 – ε )ρ c )T ] + ( ρ qi c T )
∂t ∂ xi ∂ x i ⎝ ij ∂x j⎠
f
divergence form
∂h ∂q i
∂T f f f ∂T ∂T
∂ ⎛ ------ S o ----- + ------- = Q ρ + Q EB ( C, T ) (1-139)
λ -⎞ + ρ c Q ρ ( T – T o ) = Q T ∂t ∂x i
f f s s f f
[ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- –
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠
convective form
⎧
⎪
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎪
⎩
cbcilt=ö=PV
NK=_~ëáÅ=bèì~íáçåë
QM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
NKNN=`Ü~ê~ÅíÉêáòáåÖ=cêÉÉ=`çåîÉÅíáçå=mêçÅÉëëÉë
2
• 4π ≤ Ra < 390 = Ra crit2 onset of free convection,
stationary conditions, stable convective rolls,
• Ra ≥ Racrit2 ≈ 390 fluctuating free convection, no
more stationary conditions, transient flow oscilla-
tions appear.
cbcilt=ö=QN
NK=_~ëáÅ=bèì~íáçåë
QO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë
O
down or upward in dependence on the movement of the
O cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë
The treatment of free surfaces differs significantly Alternatively to fixed grid models, FEFLOW pro-
between 2D and 3D processes. While the formulations vides an advanced method which is based on 3D mov-
of the basic equations for essentially horizontal prob- ing meshes. It allows a more accurate and rigorous
lems in an unconfined (or phreatic) aquifer are rather modeling of both the flow and, especially here, the
simple, 3D problems can imply complicate conditions, transport processes.
especially due to:
• existence of multiple (more than one) free sur- OKO eçêáòçåí~ä= EOaF= mêçÄäÉãë
faces; áå=råÅçåÑáåÉÇ=^èìáÑÉêë
• effects of location of free surface(s) on the trans-
port process in the depth; The modeling is based on the vertically averaged
• problems arising if parts of the domain of an aqui- equations (1-71) to (1-74), which are solved in the fol-
fer system fall dry and/or the water table rises; lowing form:
• boundary and initial conditions in dependence on
the location of free surface(s). f
∂h ∂q
( S o B + ε e ) ----- + -------i = Q ρ (2-1)
∂t ∂x i
⎧
⎪
⎨
⎪
⎩
∂C-⎞
∂ ( εRBC ) + ∂ ⎛ q f C – BD ------ ⎫
ij ∂x ⎠ + εRBϑC = Q C divergence form ⎪
∂t ∂ xi ⎝ i j ⎪
⎬ (2-3)
∂C f ∂C ∂C
εR d B ------ + q i ------- – ∂ ⎛⎝ BD ij -------⎞⎠ + ( εRBϑ + Q ρ )C = Q C convective form ⎪
⎪
∂t ∂x i ∂ x i ∂x j
⎭
∂ B ερ f f s s ∂ ρ f f f – ∂ ⎛ Bλ ------ ∂T-⎞ ⎫
[ ( c + ( 1 – ε )ρ c )T ] + ( qi c T ) = Q T divergence form ⎪
∂t ∂ xi ∂ x i ⎝ ij ∂x j⎠ ⎪
⎬ (2-4)
f f s s ∂T f f f ∂T ∂ ⎛ ∂T-⎞ f f ⎪
B [ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- – Bλ ------ + ρ c Q ρ ( T – T ) = Q convective form ⎪
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠ o T
⎭
S = So B + εe ( 1 – τ ) ⎫
as exhibited in Fig. 2.1. ⎪
1 – Δh 1 ⎬ (2-6)
⎛ ⎞
τ = --- atan ---------- + --- ⎪
π ⎝ σ ⎠ 2
⎭
top
xl
top
Δh where Δh = x l – h (Fig. 2.1) is the difference
between the aquifer top and the computed head h, and
–3
σ ≈ 10 represents a given smoothing parameter.
confined unconfined h B Accordingly, for Δh ≥ 0 unconfined and for Δh < 0
B bottom
xl confined conditions occur.
QQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
OKP=cêÉÉ=pìêÑ~ÅÉë=áå=Pa
∂C-⎞
∂ ( εRC ) + ∂ ⎛ f C – D ------ ⎫
q ij ∂x ⎠ + εRϑC = Q C divergence form ⎪
∂t ∂ xi ⎝ i j ⎪
⎬ (2-9)
∂C f ∂C ∂C
εR d ------ + q i ------- – ∂ ⎛⎝ D ij -------⎞⎠ + ( εRϑ + Q ρ )C = Q C convective form ⎪
⎪
∂t ∂x i ∂ x i ∂x j
⎭
∂T-⎞
∂ [ ( ερ f c f + ( 1 – ε )ρ s c s )T ] + ∂ ( ρ f q f c f T ) – ∂ ⎛ λ ------ ⎫
i = Q T divergence form ⎪
∂t ∂ xi ∂ x i ⎝ ij ∂x j⎠ ⎪
⎬ (2-10)
s s ∂T f f f ∂T ∂
[ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- – ∂ ⎛⎝ λ ij -------⎞⎠ + ρ c Q ρ ( T – T o ) = Q T convective form ⎪⎪
f f T f f
∂t ∂x i ∂ x i ∂x j
⎭
cbcilt=ö=QR
OK=cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë
f f fs
–qn = –ε vi ⋅ ni (2-13)
∂B ⁄ ∂t ∂h sf
= P o ⋅ n l + ε --------------------- ⋅ n l = ⎛⎝ P o – ε e -----⎞⎠ n l
sf fs
∂x l ⁄ ∂x i ∂t upper free surface
⎛ P – ε ∂-----h⎞ n = – q f ⎫
⎝ o e ∂t ⎠ l n⎪
⎬ (2-14)
h = xl ⎪ OKQ _^pa=qÉÅÜåáèìÉ=~åÇ=jçîJ
⎭ áåÖ=Pa=jÉëÜÉë
where
The adaptation of a finite element mesh to a change-
Po = groundwater recharge or infiltration rate; able free-surface location has the advantage that the
εe = effective (drainable/fillable) porosity; mesh density is maintained in domains which are actu-
nl = normal unit vector (positive outward);
ally discharged. The conditions (2-14) lead to a nonlin-
f ear boundary-value problem due to the a priori
qn = Darcy normal flux of fluid;
unknown free-surface locations. The process has to be
xl = elevation.
solved by an appropriate iterative scheme. According
to the computed hydraulic head at a time level the mesh
For 3D modeling of free-surface problems FEFLOW slices are to be adjusted. Slices of the inner mesh have
provides a number of additional extensions for which to be adapted in accordance with the changed water
the most important features are discussed next. table. However, in moving the meshes a number of data
must be adapted to the new locations. For instance, the
movement of a free surface through a layered aquifer
system requires an assignment of conductivities and
storage coefficients at each time of the new free-sur-
face location. Especially at jump conditions in the
QS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
OKQ=_^pa=qÉÅÜåáèìÉ=~åÇ=jçîáåÖ=Pa=jÉëÜÉë
parameter fields typical for aquifer systems, the adapta- bution in the depth without unnecessary mesh refine-
tion of the spatially varying material data to the moving ment and coarsening if ever attainable. Following
mesh has to be performed effectively and carefully. criteria are employed to position the material layers:
FEFLOW provides here an advanced technique.
• fit layer surfaces (slices) to surfaces of parameter
This technique termed BASD (Best-Adaptation-to- discontinuity as soon located within the domain,
Stratigraphic-Data) transforms and joins the model data • distribute remaining slices proportional to layer
containing the stratigraphic initial structure to a mov- thicknesses to attain as best as possible well-
ing finite element mesh which is appropriately adapted behaved mesh distances in the depth,
to the free-surface locations. In this adaptation process • join parameter for such elements according to par-
the mesh slices are aligned in such a manner that the tial volumes and surfaces, which cross and inter-
adjusted mesh is exactly fitted to parameter discontinu- sect more than one stratigraphic material layer.
ities if they ever exist. Remaining slices can be shifted
and repositioned to achieve a well-spaced nodal distri-
finite element
K1 alignment K1
node
cross element free surface
K2 K2
K2 K3 K2
K3 K3
K3 K3
K3 K3
Figure 2.3 Moving mesh BASD technique of parameter adaptation applied to 3D free-surface
problems: schematized example for a groundwater rise.
BASD has proven to be a powerful and accurate is accordingly shrunk, where the lower two layers com-
computation strategy for complex 3D application with pletely fit into the K3 stratigraphy. However, the upper
free surface(s). The principle of the BASD mesh adap- layer crosses between the K 2 and K 3 stratigraphy and
tation process is illustrated in Fig. 2.3 for a schema- a special treatment is here required. Such cross ele-
tized example. The initial stratigraphy consists here of ments are admitted only if unavoidable. A proper 3D
three layers having different conductivities. At initial interpolation technique has been developed which
time t o the water table is on a lower position. The mesh allows a data joining for elements intersecting an arbi-
cbcilt=ö=QT
OK=cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë
trary number of stratigraphic layers (for more see Dier- head, viz.,
sch and Michels28). If the water table ascends (Fig. 2.3
at time t 1 ) the moving mesh totally fits the K2 - K 3 C ( x i, 0 ) = C I ( x i, h ) (2-16)
stratum while the remaining slice is used to subdivide
the widest nodal spacing, here in the K3 layer. At later A number of special conditions can be raised. The most
time t 2 a further rise of the free surface occurs and the important items are:
moving mesh slices appear to be well aligned to the
data stratification without any need of interpolation. • At initial time the contaminant distribution is
related to the hydraulic head distribution (location
of free surface). When the initial water table does
OKR ^Åíáî~íáåÖ= `çåí~ãáå~åí not meet the location of the contaminants the
fåáíá~ä= `çåÇáíáçåë= áå actual initial condition of contaminants will corre-
aÉéÉåÇÉåÅÉ= çå= íÜÉ= içÅ~J spond to the background value (default is zero).
• At changing position of water table it is constantly
íáçå=çÑ=cêÉÉ=pìêÑ~ÅÉ checked whether further contaminants have to be
Commonly, initial conditions in the model domain conveyed to the corresponding mesh nodes as ’ini-
are spatially dependent: tial conditions’. It obeys the following criteria:
QU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
OKR=^Åíáî~íáåÖ=`çåí~ãáå~åí=fåáíá~ä=`çåÇáíáçåë=áå=aÉéÉåÇÉåÅÉ=çå=íÜÉ=içÅ~íáçå=çÑ=cêÉÉ
pìêÑ~ÅÉ
If water contains a certain dissolved amount of con-
taminants, the computational procedure provides that
further contaminants can be added as soon the water
table arrives at a location, where higher contaminant
concentrations are charged. In this procedure a chemi-
cal equilibrium is assumed. It causes that the contami-
nants are instantaneously dissolved and conveyed to
the computational nodes as soon contacted with water.
However, if this location possesses a lower concentra-
tion in the initial condition, then such a new contribu-
tion is disregarded. The advantages of such an
approach are obvious. On the one hand, it controls that
higher contaminant concentrations can always float up
according to a rising water table. On the other hand, if
the water table descends and higher concentrations
occur in the groundwater, contaminants can also reach
deeper regions of the aquifer system which are previ-
ously clean (or lowly contaminated).
cbcilt=ö=QV
OK=cäçï=~åÇ=qê~åëéçêí=mêçÅÉëëÉë=ïáíÜ=cêÉÉ=pìêÑ~ÅÉë
RM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
_~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ
P
For three-dimensional (3D) and two-dimensional
P _~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ
∂C⎞
∂ ( εRC ) + ∂ ⎛ q f C – D ------ ⎫
ij ∂x ⎠ + εRϑC = Q C
- divergence form ⎪
∂t ∂ xi ⎝ i j ⎪
⎬ (3-3)
∂C f ∂C ∂C-⎞
∂ ⎛ D ------ ⎪
εR d ------ + q i ------- – + ( εRϑ + Q ρ )C = Q C convective form ⎪
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠
⎭
∂ ερ f f s s ∂ ρ f f f – ∂ ⎛ λ ------ ∂T-⎞ ⎫
[ ( c + ( 1 – ε )ρ c )T ] + ( qi c T ) = Q T divergence form ⎪
∂t ∂ xi ∂ x i ⎝ ij ∂x j⎠ ⎪
⎬ (3-4)
f f s s ∂T f f f ∂T ∂T-⎞
∂ ⎛ λ ------ f f ⎪
[ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- – + ρ c Q ( T – T ) = Q convective form ⎪
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠ ρ o T
⎭
cbcilt=ö=RN
PK=_~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ
which are to be solved for the four primary variables: Q EB = term of extended Boussinesq approximation,
Eq. (1-138);
h = hydraulic head; R = retardation factor;
f Rd = derivative term of retardation;
qi = Darcy velocity vector of fluid;
C = concentration of chemical component; D ij = tensor of hydrodynamic dispersion;
T = temperature; ϑ = decay rate‡);
ε = porosity;
f s
where c,c = specific heat capacity of fluid and solid,
respectively;
kçíáÅÉI= íÜÉ= ÇÉÅ~ó f
ρ , ρo
f
= fluid and reference fluid density, respectively; λ ij = tensor of hydrodynamic thermodispersion;
ê~íÉ= ϑ = Å~å= ÄÉ
s Qx = source/sink function of fluid ( x = ρ ), of
ÉñéêÉëëÉÇ= Äó= íÜÉ ρ = solid density;
êÉ~Åíáçå= = Ü~äÑJäáÑÉ= t 1 ⁄ 2 = ~ë So = specific storage coefficient (compressibility); ‡) Referring to radioactive decay processes the decay rate ϑ
ÇÉëÅêáÄÉÇ= áå= íÜÉ= ~ÅÅçãé~J can
K ij = tensor of hydraulic conductivity; ln2
åóáåÖ=ÑççíåçíÉK be expressed by the reaction half-life t 1 ⁄ 2 , viz., ϑ = --------
- using
t1 ⁄ 2
ej = gravitational unit vector;
t 1 ⁄ 2 as an (alternative) input parameter related to a (radioactive)
fμ = constitutive viscosity relation function;
component (for more see Section 1.4.2).
RO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
PKO=eçêáòçåí~ä=cçêãìä~íáçåë
cbcilt=ö=RP
PK=_~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ
ing to the above derivations the vertically averaged bal- PKOKN `çåÑáåÉÇ=~èìáÑÉê
ance equations are in summarized form:
f
∂q
S o B ∂-----h + -------i = Q ρ (3-9)
∂t ∂x i
⎧
⎨
⎩
So
f ∂h
q i = – T ij ------- (3-10)
∂x j
∂C⎞
∂ ( εRC ) + ∂ ⎛ q f C – D ------ ⎫
ij -⎠ + εRϑC = Q C divergence form ⎪
∂t ∂ xi ⎝ i ∂x j ⎪
⎬ (3-11)
∂C f ∂C ∂C-⎞
∂ ⎛ D ------ ⎪
εR d ------ + q i ------- – + ( εRϑ + Q ρ )C = Q C convective form ⎪
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠
⎭
∂ B ερ f f s s ∂ ρ f f f – ∂ ⎛ λ ------ ∂T ⎞ ⎫
[ ( c + ( 1 – ε )ρ c )T ] + ( qi c T ) ij - = Q T divergence form ⎪
∂t ∂ xi ∂ x i ⎝ ∂x j⎠ ⎪
⎬ (3-12)
s s ∂T f f f ∂T ∂ ∂ T
⎛ λ ij ------- + ρ c Q ρ ( T – T o ) = Q T convective form ⎪⎪
⎞
f f f f
B [ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- –
∂t ∂x i ∂ x i ⎝ ∂x j ⎠
⎭
with
top bottom
B = xl – xl ⎫
⎪
f f ⎪
f qi qj ⎪
D ij = ( εBD d + β T V q )δ ij + ( β L – β T ) --------
f
- ⎪
Vq ⎪
⎪
(1 – ε) ( 1 – ε ) ∂[ χ ( C ) ⋅ C ] ⎪
R = B 1 + ---------------- χ ( C ) R d = B 1 + ---------------- ---------------------------- ⎬ (3-13)
ε ε ∂C ⎪
cond disp f f s s ⎪
λ ij = λ ij + λ ij Q T = B [ ερ Q T + ( 1 – ε )ρ Q T ] ⎪
⎪
f f ⎪
cond f s disp f f f qi qj ⎪
λ ij = B [ ελ + ( 1 – ε )λ ]δ ij λ ij = ρ c α T V q δ ij + ( α L – α T ) -------- -
f ⎪
Vq ⎭
RQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
PKO=eçêáòçåí~ä=cçêãìä~íáçåë
PKOKO råÅçåÑáåÉÇ=EéÜêÉ~íáÅF=~èìáÑÉê
f
∂h ∂q i
( S o B + ε e ) ----- + ------- = Q ρ (3-14)
∂t ∂x i
f ∂h
q i = – BK ij ------- (3-15)
∂x j
∂C⎞
∂ ( εRC ) + ∂ ⎛ q f C – D ------ ⎫
ij -⎠ + εRϑC = Q C divergence form ⎪
∂t ∂ xi ⎝ i ∂x j ⎪
⎬ (3-16)
∂C f ∂C ∂C-⎞
∂ ⎛ D ------ ⎪
εR d ------ + q i ------- – + ( εRϑ + Q ρ )C = Q C convective form ⎪
∂t ∂x i ∂ x i ⎝ ij ∂x j⎠
⎭
∂ f f s s ∂ f f f ∂ ⎛ ------∂T ⎫
[ B ( ερ c + ( 1 – ε )ρ c )T ] + ( ρ qi c T ) – λ ij -⎞ = Q T divergence form ⎪
∂t ∂ xi ∂ x i ⎝ ∂x j⎠ ⎪
⎬ (3-17)
f f s s ∂T f f f ∂T ∂ ∂T
⎛ λ -------⎞ + ρ c Q ( T – T ) = Q convective form ⎪
f f
B [ ερ c + ( 1 – ε )ρ c ] ------ + ρ c q i ------- – ρ ⎪
∂t ∂x i ∂ x i ⎝ ∂x j⎠
ij o T
⎭
with
bottom
B = h – xl ⎫
⎪
f f ⎪
f qi qj ⎪
D ij = B ( εD d + β T V q )δ ij + ( β L – βT ) f -
-------- ⎪
Vq ⎪
⎪
(---------------
1 – ε )- (---------------
1 – ε )- ∂----------------------------
[χ(C) ⋅ C] ⎪
R = B 1+ χ(C) Rd = B 1 + ⎬ (3-18)
ε ε ∂C ⎪
cond disp f f s s ⎪
λ ij = λ ij + λ ij Q T = B [ ερ Q T + ( 1 – ε )ρ Q T ] ⎪
⎪
q i q j ⎪⎪
f f
cond f s disp f f f
λ ij = B [ ελ + ( 1 – ε )λ ]δ ij λ ij = ρ c α T V q δ ij + ( α L – α T ) -------- -
f ⎪
Vq ⎭
cbcilt=ö=RR
PK=_~ëáÅ=bèì~íáçåë=çÑ=jçÇÉäáåÖ
RS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
Q
C ( x i, 0 ) = C I ( x i ) in Ω
Q fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
Mass transport:
cbcilt=ö=RT
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
QKOKN cäçï
R
h ( x i, t ) = h 1 ( t ) on Γ 1 × t [ 0, ∞ ) (4-4)
⎛ ∂h ρ – ρ o ⎞
f f ⎫
- e j⎟ n i for 3D and 2D vertical ⎪
R
q n ( x i, t ) = q h ( t ) = – K ij f μ ⎜ ------- + ---------------
h
⎝ ∂x j ρo
f
⎠ ⎪
⎪
⎪
∂ h
for 2D horizontal unconfined ⎬ on Γ 2 × t [ 0, ∞ ) (4-5)
R
q n ( x i, t ) = q h ( t ) = – K ij ------- n i
h ∂x j ⎪
⎪
R ∂ h ⎪
q n ( x i, t ) = q h ( t ) = – T ij ------- n i for 2D horizontal confined ⎪
h ∂x j ⎭
R
⎬ on Γ 3 × t [ 0, ∞ ) (4-6)
q n ( x i, t ) = – Φ h ( h 2 – h ) for 2D horizontal confined ⎪
h ⎭
RU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
tions ( q n h < 0 , e.g. infiltration from a surface water into for 3D and 2D (horizontal), respectively;
in out
the aquifer) and outflowing conditions ( q n h > 0 , e.g. Φh , Φh = directional coefficient of in-transfer and out-
exfiltrating the aquifer into the surface water). Their transfer, respectively, for 3D;
in out
usefulness for river-aquifer interactions is discussed Φ h , Φ h = directional coefficient of in-transfer and out-
in out
further below.
in
The special case Φ h = Φ h = Φ h or
out
transfer, respectively, for 2D (horizontal);
w
Φ h = Φ h = Φ h does not differ between inward and Qρ = well function;
w
outward boundary flux, so it becomes directionally Qm = pumping/injecting rate of single well m;
independent. m
xi = coordinate of single well m;
Po = groundwater recharge (infiltration rate);
4th kind boundary condition (single well type εe = effective (drainable/fillable) porosity;
(singular point source)):
ni = normal unit vector;
xl = elevation.
w w m
Q ρ ( x i, t ) = ∑ Qm ∏ { δ ( xi – xi ) } (4-9)
m i With respect to boundary conditions of 2nd and 3rd
m
for ∀( x i, x i ) ∈ Ω kind special boundary conditions are available for
problems with free (phreatic) surface(s). They are
referred to as integral boundary conditions and are
defined as follows:
Free surface conditions to be satisfied simultaneously:
⎛ P – ε ∂-----h⎞ n = – q ⎫
⎝ o e ∂t ⎠ l nh ⎪
⎬ on Γ 4 × t [ 0, ∞ ) (4-10)
h = xl ⎪
⎭
in which
R R
h1 , h2 = prescribed boundary values of hydraulic head
h;
qn = normal Darcy flux of fluid (positive outward);
h
qn = vertically averaged normal Darcy flux of fluid
h
(positive outward);
R R
qh , qh = prescribed normal boundary fluid flux for 3D
and 2D (horizontal), respectively;
Φ h, Φ h = fluid transfer coefficient (leakage parameter)
cbcilt=ö=RV
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
R
q n ( x i, t ) = q h ( t ) for 3D related to the initial stratigraphic structure ⎫
h ⎪
⎬ (4-11)
R
q n ( x i, t ) = q h ( t ) for 2D horizontal-unconfined as depth-integrated flux ⎪
h ⎭
R
q n ( x i, t ) = – Φ h ( h 2 – h ) for 3D related to the initial stratigraphic structure ⎫
h ⎪
⎬ (4-12)
R
q n ( x i, t ) = – Φ h ( h 2 – h ) for 2D horizontal-unconfined as depth-integrated flux ⎪
h ⎭
Using these integral formulations of flux boundary (1) For 2D problems the fluxes have to be assigned as
conditions it is ensured that a given flux rate on their already depth-integrated. The dimension of these
2 –1
boundary portions becomes independent of the actually fluxes is then L T , similar to a horizontal confined
discharging aquifer thickness and the location of free condition.
surface. This is unlike a default nonintegral boundary
condition where a flux rate is integrated along the (2) For 3D problems the aquifer system is compiled as
effective aquifer thickness, which depends on the an initial stratigraphic layer structure. Boundary condi-
actual (computed) free-surface position, and accord- tions of the integral type are related to these initial
ingly, varying (gross-) discharges may occur through structure and accordingly, the integrated gross dis-
such boundaries. As a result, it may happen that the charges remain independent of the free-surface location
total discharge through such varying boundaries con- during the computation with the BASD technique (Sec-
stantly decreases at a descending water table. Conse- tion 2.4). Notice, the dimension of these boundary
–1 2 –1
quently, such a flow region can inevitably fall dry and fluxes is LT (not L T !).
possibly the problem can ’collapse’ with a zero inflow.
Integral boundary conditions prevents such situations Integral boundary flux conditions have only a dis-
since the gross discharges are not influenced by the tinct meaning for problems with free (movable) sur-
location of free surface. The relation of fluxes, how- face(s). If no free surfaces exist, they are totally
ever, is distinguished in 2D and 3D applications due to equivalent to the nonintegral boundary conditions of
reasons of implementation: 2nd and 3rd kind, in accordance with Eqs. (4-5) and (4-
SM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
As natural boundary conditions appear those, Boundary fluxes (e.g., 2nd kind boundary condi-
which result automatically for the model without any tion) are defined normal to the boundary Γ and posi-
specification on boundaries. It is q n h = 0 implicitly tive outward as illustrated in Fig. 4.1 for both 3D and
satisfied if no additional assignments occur, i.e., the 2D.
boundaries are impervious, independent of their actual
2D 3D
n qnh
Γ .
q nh n
. Γ
Ω
Ω
The formulation of 3rd kind boundary conditions is kind is reduced to a Dirichlet-type (1st kind) boundary
R
based on a general transfer relation between the refer- condition approaching to h = h2 .
R
ence value h 2 on the boundary portion and the hydrau-
lic head h to be computed at the same place. The For flow problems the transfer coefficient Φ h can
R
reference hydraulic head h 2 can also be time-depen- be identified as a specific colmation (or leakage) coef-
R R
dent h 2 = h 2 ( t ) . The dual transfer coefficient Φ h pos- ficient as outlined in Fig. 4.2 for inflowing (infiltration)
in R
sesses the property of a resistance coefficient which conditions ( Φ h → Φ h (h 2 > h) ). An adjacent river bed
constrains the discharge through the boundary and, is clogged (’colmated’) by a layer of thickness d and a
in
additionally, differs between inflowing and outflowing conductivity of K o . Normally, the layer conductivity
in out in
conditions by means of Φ h and Φ h , respectively, K o is much smaller than the conductivity K 1 of the
according to Eqs. (4-7) and (4-8). If Φ h ≡ 0 the bound- aquifer to be modeled. Thereby the model boundary Γ
ary becomes impervious. On the other hand, using a represents the inner boundary of the ’colmation’ layer
very large value Φ h → ∞ the boundary condition of 3rd Γ 3 , where the model domain Ω ends.
cbcilt=ö=SN
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
infiltrating
Γ4
in d
Ko Γ4 exfiltrating
out
Γ3
R
h2 h Γ3 Ω K1 » Ko
Ω in
K1 » Ko out d
Ko h
R
h2
in
Figure 4.2 Transfer coefficient Φ h ( = Φ h ) as ’colmation’
parameter of a river bed. Figure 4.3 Transfer coefficient Φ h ( = Φ h ) as ’colma-
out
in
in in Ko –1
Φ h = BΦ h ≈ B -------- in [ md ] (4-15)
d
SO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
QKOKO j~ëë=íê~åëéçêí
R
C ( x i, t ) = C 1 ( t ) on Γ 5 × t [ 0, ∞ ) (4-18)
⎫
R ∂C ⎪
convective form: q n ( x i, t ) = qC ( t )
= – D ij ------- n i ⎪
C ∂x j ⎪
⎪
⎧
⎪
⎨
⎪
⎩
dispersive flux ⎬ on Γ 6 × t [ 0, ∞ ) (4-19)
⎪
∂C
divergence form: q n ( x i, t ) = q C ( t ) = C 2 q n h – D ij ------- n i ⎪⎪
R† R
C ∂x j
⎪
⎧
⎪
⎪
⎨
⎪
⎪
⎩ ⎭
total flux
⎫
R ∂C ⎪
convective form: q n ( x i, t ) = qC ( t )
= – D ij ------- n i ⎪
C ∂x j ⎪
⎪
⎧
⎪
⎨
⎪
⎩
⎭
total flux
cbcilt=ö=SP
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
SQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
With respect to boundary conditions of 2nd and 3rd referred to as integral boundary conditions and are
kind special boundary conditions are available for defined as follows:
problems with free (phreatic) surface(s). They are
R
q n ( x i, t ) = q C ( t ) for 3D related to the initial stratigraphic structure ⎫
C ⎪
⎬ (4-25)
R
q n ( x i, t ) = q C ( t ) for 2D horizontal-unconfined as depth-integrated flux ⎪
C ⎭
R
q n ( x i, t ) = – Φ C ( C 3 – C ) for 3D related to the initial stratigraphic structure ⎫
C ⎪
⎬ (4-26)
R
q n ( x i, t ) = – Φ C ( C 3 – C ) for 2D horizontal-unconfined as depth-integrated flux ⎪
C ⎭
Referred to as natural boundary conditions the rela- boundary condition, which can cause a specific han-
tions q n C = 0 and q n C = 0 , respectively, are automati- dling of such formulations in the case of unknown
R
cally satisfied in the model if the boundary remains mass concentration C 2 on outflowing boundaries
R
unspecified. In dependence on the formulation of the (rather, C2 is here to be solved). Nevertheless, using
mass transport equation (of convective or divergence the divergence form FEFLOW is capable of computing
form) the normal mass fluxes on the boundaries are automatically the unknown convective part of the mass
suppressed for such natural boundary conditions. Here, flux at outflowing boundaries independent of the
it should be considered, however, either the dispersive imposed types of boundary conditions. This is done by
or the total fluxes are imposed. The differences are dis- evaluating the fluid flux via a budget analysis in a post-
cussed next. processing step which is then involved in modifying
the boundary conditions of the mass flux at such por-
The divergence form is capable of prescribing the tions of boundaries. A more detailed discussion can be
total mass flux along a boundary portion resulting from found in Diersch29.
R R
the convective part C2 q nh (load of concentration C 2 in
the fluid flow q nh ) and the dispersive part On the other hand, the default convective form does
– D ij ( ∂C ⁄ ∂x j )n i . However, regarding this formulation not require a specific handling associated with formu-
all boundaries have to be specified with such a type of lations on outflowing boundaries and should be usually
cbcilt=ö=SR
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
in
the mass transfer coefficient Φ C can be assessed as
in
in D o –1
Φ C ≈ -------- in [ md ] (4-28)
d
SS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
QKOKP eÉ~í=íê~åëéçêí
R
T ( x i, t ) = T 1 ( t ) on Γ 8 × t [ 0, ∞ ) (4-30)
⎫
R ∂T ⎪
convective form: q n ( x i, t ) = q T ( t ) = – λ ij ------- n i ⎪
T ∂x j ⎪
⎪
⎧
⎪
⎨
⎪
⎩
conductive flux ⎬ on Γ 9 × t [ 0, ∞ ) (4-31)
⎪
∂T
q n ( x i, t ) = q T ( t ) = ρ c T 2 q n – λ ij ------- n i ⎪⎪
R † f f R
divergence form:
T h ∂x j
⎪
⎧
⎪
⎪
⎪
⎨
⎪
⎪
⎪
⎩
⎭
total flux
⎫
R ∂T ⎪
convective form: q n ( x i, t ) = qT ( t ) = – λ ij ------- n i ⎪
T ∂x j ⎪
⎪
⎧
⎪
⎨
⎪
⎩
⎭
total flux
cbcilt=ö=ST
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
R
⎬ on Γ 10 × t [ 0, ∞ ) (4-33)
q n ( x i, t ) = – Φ T ( T 3 – T ) for 2D horizontal ⎪
T ⎭
R R R
⎧ in R T1 , T2 , T3 = prescribed boundary values of temperature T;
⎪ ΦT for T3 > T
ΦT = ⎨ (4-35) qn = normal heat flux (positive outward);
⎪ Φ out for
R
T3 ≤ T qn
T
= vertically averaged normal heat flux (positive
⎩ T T
outward);
R R
q T , q T = prescribed normal boundary heat flux of the
Accordingly, in specifying two alternate (if neces- conductive including thermodispersive part;
sary temporal) coefficients different heat transfer con- R† R†
q T , q T = prescribed normal boundary heat flux of the
ditions can be input to distinguish between inflowing total (convective + conductive +
conditions ( q nT < 0 ) and outflowing conditions thermodispersive) part;
in out
( q nT > 0 ).in Theoutspecial case Φ T = Φ T = Φ T or Φ T, Φ T = heat transfer coefficients;
Φ T = Φ T = Φ T does not differ between inward and in out
Φ T , Φ T = directional coefficient of heat in-transfer and
outward heat boundary flux. heat out-transfer, respectively;
in out
Φ T , Φ T = depth-integrated heat in-transfer and heat out-
transfer, respectively;
4th kind boundary condition (single well type QT
w
= heat flux well function;
(singular point source)): Tm
w
= temperature of single well m;
w
Qm = pumping/injecting heat rate of single well m;
m
xi = coordinate of single well m.
SU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKO=_çìåÇ~êó=`çåÇáíáçåë
R
q n ( x i, t ) = q T ( t ) for 3D related to the initial stratigraphic structure ⎫
T ⎪
⎬ (4-37)
R
q n ( x i, t ) = q T ( t ) for 2D horizontal-unconfined as depth-integrated flux ⎪
T ⎭
R
q n ( x i, t ) = – Φ T ( T 2 – T ) for 3D related to the initial stratigraphic structure ⎫
T ⎪
⎬ (4-38)
R
q n ( x i, t ) = – Φ T ( T 2 – T ) for 2D horizontal-unconfined as depth-integrated flux ⎪
T ⎭
in
Referred to as natural boundary conditions the rela- the heat transfer coefficient Φ T can be obtained as
tions q n T = 0 and q n T = 0 , respectively, are automati-
cally satisfied in the model. in
in λ o –2 –1 –1
Φ T ≈ ------- in [ Jm d K ] (4-40)
d
The heat transfer coefficients Φ T associated with
boundary conditions of 3rd kind, Eq. (4-33), represent
and similarly to a horizontal problem as
a heat transition resistance. If Φ T ≡ 0 the boundary
becomes adiabatic (insulated). On the other hand, using in
a very large value Φ T → ∞ the boundary condition of in in λo –1 –1 –1
Φ T = BΦ T ≈ B ------- in [ Jm d K ] (4-41)
3rd kind is reduced to a Dirichlet-type (1st kind) d
R
boundary condition with T = T2 . The heat transfer
in
coefficients can be estimated analogously to the above where λ o represents a heat conduction coefficient of
out
transfer coefficients for mass flux. Considering a thick- the transition
out
layer. Analogous relationships for Φ T
ness d for the a heat transition layer and applying Fou- and Φ T result if the heat transition resistance differs
R
rier’s law for input condition ( T 2 > T ) in form of: between inflowing and outflowing conditions:
in out
λo ≠ λo .
R
in ΔT in T 2 – T
q n ≈ – λ o ------- = – λ o --------------- (4-39)
T Δl d
cbcilt=ö=SV
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
Constraints in relation to free-surface problems are For free-surface problems it may happen that parts
of special significance, for instance in the separation of of the aquifer(s) become dry (water table falls below
a groundwater flow from a surface water if the water the aquifer base). Such a situation referred to as falling
table (or the head difference) falls below a given mag- dry can be considered as a natural constraint condition
nitude and an infiltration from the surface water into as described further below in greater detail.
the aquifer has to be constrained by a maximum influx.
Furthermore, it arises in falling below a critical hydrau- The formulation of constraints is based on the for-
lic head for a well (e.g., below the bottom of a well malism of complementary conditions for a type of
screen) operating with a given pumping rate. If attained boundary condition. Accordingly, a potential condition
a minimum water table has to be held and the pumpage (hydraulic head, concentration or temperature) is con-
is to be correspondingly adapted. strained by maximum and minimum flux relations
(fluid, mass or heat fluxes). On the other hand, flux
Additionally, regarding boundary conditions of relationships are constrained by complementary poten-
transport equations constraints appear here in a com- tial limits, i.e., the fluid flux is constrained by maxi-
bined manner. Limiting conditions are not only raised mum-minimum heads, the contaminant mass flux by
by the contaminant mass or heat regime but should also minimum-maximum concentrations, and the heat flux
be regarded in form of minimum or maximum bounds by minimum-maximum temperatures. Following for-
for the water table. For instance, a contaminant bound- mulations are available for flow conditions:
ary condition may be related to the current hydraulic
head. If the water table is low, apparently the contami-
nant boundary condition is yet inactive, a typical case
occurring in rising groundwater tables if (re-)mobiliza-
tion of contaminants due to an increasing wetting pro-
cess of a contaminant source can be expected.
TM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKP=`çåëíê~áåíë
QKPKN cäçï
⎫
⎛ Max: Q R < Q max 1 ( t ) else
R
Qh = Qh
max 1
( t )⎪
1st kind
R
h1 ( t ) if ⎜ h h
⎪
⎜ R min 1 R min 1 ⎪
⎝ Min: Q h > Q h ( t ) else Q h = Q h ( t ) ⎪
⎪
⎛ Max: h < h 2 ( t ) else h = h 2 ( t ) ⎪⎪
R max R max
R
2nd kind qh ( t ) if ⎜
⎝ Min: h R > h min2 ( t ) R min 2 ⎪
else h = h (t) ⎪
⎬ (4-42)
⎛ Max: Q R < Q max 3 ( t ) else Q R = Q max 3 ( t )⎪
⎪
if ⎜
R h h h h
3rd kind h2 ( t )
⎜ R min 3 R min 3 ⎪
⎝ Min: Q h > Q h ( t ) else Q h = Q h ( t )⎪
⎪
max max ⎪
w ⎛ Max: h R < h 4 ( t ) else h R = h 4 ( t ) ⎪
4th kind Qρ ( t ) if ⎜ ⎪
⎝ Min: h R > h min 4 ( t ) else h R = h min 4 ( t ) ⎪
⎭
R R
where Q h = – ∫ q h dV represents the summed-up, called by Eqs. (4-42) are shown in sketches of Fig. 4.5.
lumped balance fluxes at nodal points to which poten-
tial boundary values are related. Note, due to technical While constraints (4-42) are applied to ’active’ flow
reasons the balance quantities are defined positive boundary conditions there exists one important excep-
inward. These fluxes are computed in a balance analy- tion, which includes all points of the domain Ω and the
sis during the simulation. The minimum-maximum boundaries Γ regardless of acting boundary and con-
R R
bounds for the fluxes Q h and the potentials h them- straint conditions there: The control of the simulated
selves can be time-dependent functions again. Accord- water levels for problems with free surface by using a
ingly, it is possible to consider time-dependent natural constraint in case of falling dry of aquifers and
variations in the existence and influence of boundary a system of aquifers, respectively, in points, in subdo-
and constraint conditions. For instance, these tempo- mains, or in the whole area given by the following cri-
rary capabilities of constraints are very useful in mod- terion:
eling the temporarily varying occurrence of sealing or
drainage activities over a restricted time period, or in ⎛ ⎧ h > x bottom
l ⎫⎞
⎜⎪ ⎪⎟
simulating time-constrained boundary conditions (e.g., ⎜ ⎨ and ⎬⎟
⎜⎪ ⎪⎟ bottom
1st kind) which are associated with certain construction h(t) if ⎜ ⎩ h > h min i ⎭⎟ else h = xl (4-43)
⎜ ⎟
or remedial actions arising only at given times. Typical ⎜ or ⎟
⎜ ⎟
applications of such constraint conditions formulated ⎝ confined ⎠
cbcilt=ö=TN
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
a) b) infiltrating exfiltrating
Δh(t)
Q-(t) Q+(t) Q max
Q(t)
surface
hmax(t)
Q max(t)
le
r tab
wate
hR2 h
hmin(t)
The constraint (4-43) implies that the hydraulic head constraints are subordinated to the natural constraint
h ( t ) is valid as long as the water table lies above the conditions in order to avoid generally ’negative’ aqui-
bottom
kçíáÅÉI= êÉÑÉêêáåÖ aquifer bottom x l (bottom elevation in case of 2D fer thicknesses for free-surface problems (even at
íç= PêÇ= âáåÇ or ’fixed’ aquifer slices in case of 3D), unless other ’wrong’ boundary conditions too). It is important to
`~ìÅÜóJíóéÉ= Ñäçï ’stronger’ conditions for the potential field are to be note that the constraint (4-43) is optional that means it
ÄçìåÇ~êó= ÅçåÇáíáçåë= íÜÉêÉ subjected due to proper constraints with respect to a can be set inoperative whenever an auxiliary hydraulic
min
~êÉ= ÉñíÉåÇÉÇ= Ñçêãìä~íáçåë minimum water table h i for boundary conditions head boundary condition at the bottom cannot be
çÑ= Åçåëíê~áåíë= ïÜáÅÜ= ~êÉ i = 1, 2, 3, 4 according to Eq. (4-42). Furthermore, accepted due to the possible occurrence of inappropri-
íÜçêçìÖÜäó= ÇÉëÅêáÄÉÇ= áå
such a constraint is neither relevant in areas having ate water balance conditions (constraints introduce
íÜÉ= tÜáíÉ= m~éÉêë= J= sçäK= f
fully confined conditions nor in areas where the water additional fluxes). Under such conditions the hydraulic
ÉåíáíäÉÇ= ÛbñíÉåÇÉÇ= ÑçêãìJ
ä~íáçåë= çÑ= Åçåëíê~áåíë= Ñçê table rises above possible aquifer confinements (mixed head can fall below the bottom in an unlimited manner
`~ìÅÜóJíóéÉ= EPêÇ= âáåÇF conditions). Violating this constraint condition the min- while the effective thickness of the aquifer is hold at a
ÄçìåÇ~êó= ÅçåÇáíáçåë= áå imum water table is intermediately set equal to the minimum (as default 1 mm). As the result, it mimics an
bottom
cbciltF= áåíêçÇìÅÉÇ= Äó xl elevation at current time as depicted in Fig. 4.6. ’unsaturated’ condition where the pressure head
íÜÉ=cbcilt=êÉäÉ~ëÉ=QKVK Obviously, the criterion (4-43) have to take prior over becomes a negative suction and the minimum thickness
the constraint condition (4-42) because the prescribed implies a residual water content of an element.
TO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKP=`çåëíê~áåíë
a)
h = xbottom
Q h = xbottom
confinement
falling dry
falling dry
aquifer bottom
confined
b)
Q Q
upper area of testing unconfined conditions
1st aquifer
falling dry
confined unconfined
1st aquitard
2nd aquifer
confined
unconfined
falling dry
confined 2nd aquitard
Figure 4.6 Natural minimum constraints regarding aquifer bottom encountered in free-surface problems:
a) unconfined test area and areas of falling dry in 2D,
b) complex situation for testing unconfined conditions in a 3D aquifer system involving multiple free surfaces
at juxtaposed unconfined-confined relations.
The constraints for the transport equations provide bining features as mentioned above. In more general
more complex criteria resulting from the desired com- form they yield:
cbcilt=ö=TP
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
QKPKO j~ëë=íê~åëéçêí
⎛ ⎧ max 1 ⎫ ⎫
⎜ R
⎪ QC < QC ( t ) ⎪ ⎪
⎜ ⎛ Q R = Q max1 ( t ) as long as h min 1 ≤ h R ≤ h max 1 ⎪
⎪ ⎪ ⎜ C C
⎜ Max:⎨ and ⎬else ⎜ ⎪
⎜ min 1 ⎪
⎜
⎪ ⎪
⎪ h min 1 ≤ h R ≤ h max1 ⎪
R
⎝Q C = 0 if h < h
R
or h R > h max1 ⎪
⎜ ⎩ ⎭ ⎪
R
1st kind C 1 ( t ) if ⎜ ⎪
⎜ ⎪
⎜ ⎧ R min 1 ⎫ ⎪
⎜ ⎪ QC > QC ( t ) ⎪ ⎛ Q R = Q min 1 ( t ) as long as h min 1 ≤ h R ≤ h max1 ⎪
⎜ ⎪ ⎪ ⎜ C C ⎪
⎜ Min: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪
⎪ h min1 ≤ h R ≤ h max1 ⎪
⎝ C
Q = 0 if hR < h min1 or hR > h max1 ⎪
⎜ ⎪
⎝ ⎩ ⎭ ⎪
⎪
max ⎪
⎛ R
⎧ C <C
2
(t) ⎫ ⎪
⎜ ⎛ C R = C max2 ( t ) as long as h min 2 ≤ h R ≤ h max 2
⎪ ⎪ ⎜ ⎪
⎜ Max: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪ R min 2
or h R > hmax2
⎜ ⎝Q C = 0 if h < h ⎪
⎩h min 2
≤ h
R
≤ h
max 2 ⎭ ⎪
R
2nd kind q C ( t ) if ⎜ ⎪
⎜ R min 2 ⎪
⎜ ⎧ C >C (t) ⎫ ⎛ C R = C min2 ( t ) as long as h min 2 ≤ h R ≤ h max 2 ⎪
⎜ ⎪ ⎪ ⎜ ⎪
⎜ Min: ⎨ and ⎬ else ⎜ R min 2 ⎪
⎜ ⎪ ⎪ ⎝Q C = 0 if h < h
R
or hR > h max2 ⎪
⎝ ⎩h min 2 R
≤h ≤h
max 2
⎭ ⎪
⎬ (4-44)
⎛ ⎧ max3 ⎫ ⎪
⎜ R
⎪ QC < QC ( t ) ⎪ ⎪
⎜ ⎛ Q R = Q max3 ( t ) as long as h min 3 ≤ h R ≤ h max 3 ⎪
⎪ ⎪ ⎜ C C ⎪
⎜ Max: ⎨ and ⎬ else
⎜ ⎜ R ⎪
⎜
⎪ ⎪
⎪ h min 3 ≤ h R ≤ h max3 ⎪
⎝ C
Q = 0 if h R < h min3 or h R > h max3 ⎪
⎜ ⎩ ⎭ ⎪
R ⎜ ⎪
3rd kind C 3 ( t ) if ⎜ ⎪
⎜ ⎧ R min 3 ⎫ ⎪
⎜ ⎪ QC > QC ( t ) ⎪ ⎛ Q R = Q min 3 ( t ) as long as h min3 ≤ h R ≤ h max3 ⎪
⎜ ⎪ ⎪ ⎜ C C ⎪
⎜ Min: ⎨ and ⎬ else ⎜ ⎪
⎜ ⎪ ⎪ ⎝Q
R
= 0 if hR < h min3 or h R > hmax3 ⎪
⎜ ⎪ h min 3 ≤ h R ≤ h max3 ⎪ C ⎪
⎝ ⎩ ⎭ ⎪
⎪
max ⎪
⎛ ⎧ C <C
R 4
(t) ⎫ ⎪
⎜ ⎛ C R = C max4 ( t ) as long as h min 4 ≤ h R ≤ h max 4
⎪ ⎪ ⎜ ⎪
⎜ Max: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪ R min 4
or h R > hmax4 ⎪
⎜ ⎝Q C = 0 if h < h
⎩ h min 4 ≤ h R ≤ h max 4 ⎭ ⎪
w
4th kind Q C ( t ) if ⎜ ⎪
⎜ R min 4 ⎪
⎜ ⎧ C >C (t) ⎫ ⎛ C R = C min4 ( t ) as long as h min 4 ≤ h R ≤ h max 4 ⎪
⎜ ⎪ ⎪ ⎜ ⎪
⎜ Min: ⎨ and ⎬ else ⎜ R min 4 ⎪
⎜ ⎪ ⎪ ⎝Q C = 0 if h < h
R
or hR > h max4 ⎪
⎝ ⎩ h min 4 ≤ h R ≤ h max4 ⎭ ⎭
TQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKP=`çåëíê~áåíë
QKPKP eÉ~í=íê~åëéçêí
⎛ ⎧ max 1 ⎫ ⎫
⎜ R
⎪ QT < QT ( t ) ⎪ ⎪
⎜ ⎛ Q R = Q max 1 ( t ) as long as h min1 ≤ h R ≤ h max1 ⎪
⎪ ⎪ ⎜ T T
⎜ Max:⎨ and ⎬else ⎜ ⎪
⎜ min 1 ⎪
⎜
⎪ ⎪
⎪ h min 1 ≤ h R ≤ h max 1 ⎪
R
⎝Q T = 0 if h < h
R
or h R > hmax1 ⎪
⎜ ⎩ ⎭ ⎪
R
1st kind T 1 ( t ) if ⎜ ⎪
⎜ ⎪
⎜ ⎧ R min 1 ⎫ ⎪
⎜ ⎪ QT > QT ( t ) ⎪ ⎛ Q R = Q min1 ( t ) as long as h min 1 ≤ h R ≤ h max 1 ⎪
⎜ ⎪ ⎪ ⎜ T T ⎪
⎜ Min: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪
⎪ h min 1 ≤ h R ≤ h max 1 ⎪
⎝ T
Q = 0 if h R < h min1 or h R > h max1 ⎪
⎜ ⎪
⎝ ⎩ ⎭ ⎪
⎪
max ⎪
⎛ R
⎧ T <T
2
(t) ⎫ ⎪
⎜ ⎛ T R = T max2 ( t ) as long as min 2 R
≤h ≤h
max 2
⎪ ⎪ ⎜ h ⎪
⎜ Max: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪ R min 2 R max 2
⎜ ⎝Q T = 0 if h < h or h >h ⎪
⎩h min 2
≤ h
R
≤ h
max 2 ⎭ ⎪
R
2nd kind q T ( t ) if ⎜ ⎪
⎜ R min 2 ⎪
⎜ ⎧ T >T (t) ⎫ ⎛ T R = T min 2 ( t ) as long as min 2 R max2 ⎪
⎜ ⎪ ⎪ ⎜ h ≤h ≤h ⎪
⎜ Min: ⎨ and ⎬else ⎜ R R min 2 R max2 ⎪
⎜ ⎪ ⎪ ⎝Q T = 0 if h < h or h >h ⎪
⎝ ⎩h min 2 R
≤h ≤h
max 2
⎭ ⎪
⎬ (4-45)
⎛ ⎧ max3 ⎫ ⎪
⎜ R
⎪ QT < QT ( t ) ⎪ ⎪
⎜ ⎛ Q R = Q max 3 ( t ) as long as h min3 ≤ h R ≤ h max3 ⎪
⎪ ⎪ ⎜ T T ⎪
⎜ Max: ⎨ and ⎬ else
⎜ ⎜ R ⎪
⎜
⎪
⎪ h min3 ≤ h R ≤ h max3 ⎪
⎪ ⎝ T
Q = 0 if h R < hmin3 or h R > hmax3 ⎪
⎜ ⎩ ⎭ ⎪
R ⎜ ⎪
3rd kind T 2 ( t ) if ⎜ ⎪
⎜ ⎧ R min 3 ⎫ ⎪
⎜ ⎪ QT > QT ( t ) ⎪ ⎛ Q R = Q min 3 ( t ) as long as h min 3 ≤ h R ≤ h max 3 ⎪
⎜ ⎪ ⎪ ⎜ T T ⎪
⎜ Min: ⎨ and ⎬ else ⎜ ⎪
⎜ ⎪ ⎪ ⎝Q
R
= 0 if hR < h min3 or hR > h max3 ⎪
⎜ ⎪ h min 3 ≤ h R ≤ h max 3 ⎪ T ⎪
⎝ ⎩ ⎭ ⎪
⎪
max ⎪
⎛ ⎧ T <T
R 4
(t) ⎫ ⎪
⎜ ⎛ T R = T max 4 ( t ) as long as min 4 R
≤h ≤h
max4
⎪ ⎪ ⎜ h ⎪
⎜ Max: ⎨ and ⎬ else ⎜ R ⎪
⎜ ⎪ ⎪ R min 4 R max4
⎪
⎜ ⎝Q T = 0 if h < h or h >h
⎩ h min 4 ≤ h R ≤ h max 4 ⎭ ⎪
w
4th kind Q T ( t ) if ⎜ ⎪
⎜ R min 4 ⎪
⎜ ⎧ T >T (t) ⎫ ⎛ T R = T min 4 ( t ) as long as min 4 R max 4 ⎪
⎜ ⎪ ⎪ ⎜ h ≤h ≤h ⎪
⎜ Min: ⎨ and ⎬ else ⎜ R R min 4 R max 4 ⎪
⎜ ⎪ ⎪ ⎝Q T = 0 if h < h or h >h ⎪
⎝ ⎩ h min4 ≤ h R ≤ h max4 ⎭ ⎭
cbcilt=ö=TR
QK=fåáíá~äI=_çìåÇ~êóI=~åÇ=`çåëíê~áåí=`çåÇáíáçåë
max min
where ( … ) typ and ( … ) typ denote the prescribed
maximum and minimum bounds, respectively, for the R f f RR ∂T
R R R qT = ρ c T qh – λ ij ------- n i (4-48)
corresponding type of boundary condition, Qh , Q C, Q T ∂x j
⎧
⎪
⎨
⎪
⎩
⎧
⎪
⎨
⎪
⎩
the computed balance fluxes of fluid, contaminant convective
R R R conductive
mass and heat, respectively, and h , C , T the com-
puted potential values of hydraulic head, concentration
and temperature, respectively, on the boundary. In practice, it has shown to be inappropriate to
include the total (convective plus dispersive/conduc-
R R
The fluxes Q C, Q T represent again summed-up, tive) flux into the procedure of controlling the con-
called lumped mass balance fluxes at nodal points, straint conditions (4-44) and , because the direction of
analogously to the flow problem and defined positive the dispersive/conductive fluxes is ambiguous (e.g., the
inward, viz., dispersive spreading also occurs against the convective
flow direction). Accordingly, the balance-based evalua-
tion of fluxes is exclusively related to the convective
Q C = – ∫ q C dV ⎫
R R
⎪ mass and heat fluxes, respectively:
⎬ (4-46)
R R
Q T = – ∫ q T dV ⎪ R R R R
⎭ Q C = – ∫ q C dV ≈ – ∫ ( C q h ) dV
(4-49)
R R R R
Q T = – ∫ q T dV ≈ – ∫ ( T q h ) dV
to which potential boundary values (concentrations and
temperatures, respectively) are assigned. The mini-
R R
mum-maximum bounds for the fluxes Q C, Q T and the presenting unambiguously directional balance quanti-
R R
potentials C , T can be again time-dependent func- ties.
tions allowing very comfortable rules for constraints.
The transport constraints essentially consist of two
R R
Naturally, the specific balance fluxes q C, q T are parts for the individual types of boundary conditions:
composed by their convective and dispersive parts (see
here also the relationships (4-19) and (4-20), and (4-31) (1) A minimum-maximum bound complementary
and (4-32), respectively) according to for the type of boundary condition (e.g., a con-
centration boundary (1st or 3rd kind) is con-
trolled by an allowable minimum-maximum
R R R ∂C
qC = C qh – D ij ------- n i (4-47) boundary mass flux, or a mass flux boundary
∂x j
magnitude (2nd or 4th kind) is controlled by an
⎧
⎨
⎩
⎧
⎪
⎨
⎪
⎩
TS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
QKP=`çåëíê~áåíë
min max
between h i and h i ) is imposed. If the simu- intermittent pumping regime encountered in remedial fãéçêí~åí= åçíÉW
lated water table lies outside this range the nodal activities for a contaminant plume. The wells inject ^= ÑêÉëÜï~íÉê= ÅçåÇáJ
R
boundary conditions (all types, 1st to 4th kind) clean water ( C = C1 ) as long as an injecting pumpage íáçå= áÇÉåíáÅ~ä= ïáíÜ
w òÉêç= E C R1 ≡ 0 F= áë= áå~ééêçéêáJ
are suppressed. This can easily be realized by occurs ( Q ρ < 0 ) and pump up contaminated groundwa-
R
assigning intermediately a zero flux ( Q C = 0 ter in a time-given pumping operation with a mixing ~íÉ= áå= íÜÉ= éêÉëÉåí= Ä~ä~åÅÉJ
R Ä~ëÉÇ=Åçãéìí~íáçå=íç=ÇáÑÑÉê
and Q T = 0 , respectively), i.e., no additional concentration to be determined. Both wells possess a
ÄÉíïÉÉå= áåï~êÇ= ~åÇ= çìíJ
mass or heat flux then occurs at such nodes and concentration boundary condition of 1st kind for fresh-
R ï~êÇ= ÇáêÉÅíÉÇ= ÅçåîÉÅíáîÉ
the boundary conditions become intermediately water ( C = C1 ) and a minimum mass flux constraint ÑäìñÉëK= fí= Å~å= Ñ~áä= ÄÉÅ~ìëÉ
min 1
hidden). of zero Q C ≡ 0 . íÜÉ=ÇáêÉÅíáçå~ä=ã~ÖåáíìÇÉ=çÑ
Q C = ~ÅÅçêÇáåÖ= íç= bèK= EQJRMF
R
Typical applications of transport constraints are out- The third example shown in Fig. 4.7c describes an áë=åç=ãçêÉ=áÇÉåíáÑá~ÄäÉ=ëáåÅÉ
lined in Figs. 4.7 a-c. Figure 4.7a describes the case of application in modeling a contaminant spreading pro- QC = 0 ≡ QC = 0 >
R min
1
a density-coupled saltwater intrusion problem (flow cess from a deposit associated with rising groundwater ^ÅÅçêÇáåÖäóI= áåëíÉ~Ç= çÑ= òÉêç
over a salt dome) having a boundary on which alternate in a phreatic aquifer (referred to as flooding problem). áí= áë= êÉÅçããÉåÇÉÇ= íç= ìëÉ= ~
boundary concentrations appear in dependence on the The contaminant boundary condition (e.g., modeled as åìãÉêáÅ~ääó=îÉêó=ëã~ää=î~äìÉ
dynamic process: As long as water enters the domain it a 1st kind type) should be active only when the water Eë~ó= C ⁄ Cs ≈ 10–8 FK
should have a prescribed concentration of freshwater. table reaches the contaminant deposit (wetting case),
min
However, if the water releases the domain (along the i.e., a constraint in form of h 1 is prescribed repre-
same boundary) the concentration on the boundary is senting the bottom of the contaminant deposit. More
unknown and should be automatically computed. Such constraints are not necessarily required in such a case.
a description can be easily realized if the entire bound-
ary section is assigned by a freshwater boundary condi-
R
tion of 1st kind ( C = C1 ), and at the same time, the
boundary will be imposed by a constraint condition in
min
form of a null minimum mass flux Q C 1 ≡ 0 (more
constraints are not necessarily to be specified). Such an
arrangement provides that the freshwater condition
remains valid as long as the convective flux, being con-
centration-dependent due to the density coupling phe-
nomena, points into the domain:
R R R min 1
Q C = – ∫ ( C q h ) dV > QC = 0
(4-50)
R
since qh < 0 for inflow
a) b) Q C
freshwater C = 0 releasing unknown concentration C
min1
t
QC =0 t
well 1
rte
wa
C=0
g
min
xin
QC 1 = 0
aquifer contaminant plume
mi
saltwater C=0
min
well 2 QC 1 = 0
salt dome
Q C
c)
t
t
TU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
^åáëçíêçéó
R
R ^åáëçíêçéó
RKN mêáåÅáé~ä= aáêÉÅíáçåë= ~åÇ Accordingly, the conductivity along the principal direc-
m
tions x i represents a diagonal matrix
oçí~íáçå
m
For the tensors of conductivity Kij and transmissiv- K1 0 0
ity T ij anisotropy is to be taken into account. It is m
( K ij ) = m , (5-1)
assumed they have orthotropic properties, i.e., the con- 0 K2 0
m m
ductivities are given along their principal directions x i 0 0 K3
(i = 1,2,3), which are turned against the global Carte-
sian coordinate system x i (i = 1,2,3) by a rigid-body
rotation (Fig. 5.1), see e.g. Goldstein et al.35. which can be mapped by using the transformation in
form of the rotation matrix a ij
x3m x3
x2
a 11 a 12 a 13
x2m
( a ij ) = a 21 a 22 a 23 (5-2)
a 31 a 32 a 33
x1
onto the global Cartesian coordinate system x i , viz.:
x1m
cbcilt=ö=TV
RK=^åáëçíêçéó
m
K ij = a li K lk a kj After rotation, the tensor K ij in the global Cartesian
m
coordinates results, which describes an anisotropic
K 11 K 12 K 13 a 11 a 21 a 31 K 1 0 0 a 11 a 12 a 13 (5-3) conductivity in the form:
( K ij ) = K 21 K 22 K 23 = a 12 a 22 a 32 0 K m 0 a 21 a 22 a 23
2
K 31 K 32 K 33 a 13 a 23 a 33 m a 31 a 32 a 33
0 0 K3
2 m 2 m 2 m m m m m m m
a 11 K 1 + a 21 K 2 + a 31 K 3 a 11 a 12 K 1 + a 21 a 22 K 2 + a 31 a 32 K 3 a 11 a 13 K 1 + a 21 a 23 K 2 + a 31 a 33 K 3
( K ij ) = 2 m 2 m 2
a 12 K 1 + a 22 K 2 + a 32 K 3
m m m
a 12 a 13 K 1 + a 22 a 23 K 2 + a 32 a 33 K 3
m (5-4)
2 m 2 m 2 m
symm. a 13 K 1 + a 23 K 2 + a 33 K 3
The elements of the rotation matrix a ij , Eq. (5-2), Cartesian coordinate system x 1, x 2 by a given angle φ
m
represent the directional cosines of the rotation angles between the first principal axis of K ij and the x 1 -axis
appearing in transforming from the Cartesian coordi- (Fig. 5.2). An anisotropic conductivity is then uniquely
m
nate system x i to the principal direction system x i . defined by their maximum and minimum conductivi-
m m
They are referred to as Eulerian angles. In 3D three ties, K max ≡ K 1 and K min ≡ K2 , respectively, acting
angles ( φ, θ, ψ ) result, while in 2D only one Eulerian along the principal axes, and by the rotation angle φ , as
angle φ has to be specified to perform an unique prin- depicted in Fig. 5.2.
cipal axis rotation. As the result, anisotropy for three-
dimensional applications needs formally six parameters x2
m
x2
m m m
( K 1 , K 2 , K 3 , φ, θ, ψ ) , while three parameters
m m
( K 1 , K 2 , φ ) are required to handle a two-dimensional K
K min m
anisotropy. Considering the effort and the appropriate- x1
ness in computing the anisotropic conductivity (5-4), K max
different strategies are employed for 3D and 2D cases φ
as described subsequently. x1
UM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
RKP=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó
using the rotation formula (5-3) for the two-dimen- missivity Tij , Eq. (1-70), applied to 2D problems of
sional case. Expressing the directional cosines a ij confined aquifers.
θ η' θ η'
In FEFLOW for each discretized (finite) element fol- ζ' x3m x2m
lowing quantities are input:
x2 x1m x2
(1) the maximum conductivity K max , ψ
x1 φ
(2) a factor of anisotropy defined as x1 φ
ξ' Line of nodes
cbcilt=ö=UN
RK=^åáëçíêçéó
cos ψ cos φ – cos θ sin φ sin ψ cos ψ sin φ + cos θ cos φ sin ψ sin ψ sin θ
( a ij ) = – sin ψ cos φ – cos θ sin φ cos ψ – sin ψ sin φ + cos θ cos φ cos ψ cos ψ sin θ (5-9)
sin θ sin φ – sin θ cos φ cos θ
This represents the most general formulation of aniso- RKPKO pÜ~éÉJÇÉêáîÉÇ= Pa= ~åáëçíêçéó
tropy in 3D. This case is optionally available in Äó=péêáåÖÉêÚë=ãÉíÜçÇ
FEFLOW, where the user can directly input the six
m m m
parameters ( K1 , K 2 , K 3 , φ, θ, ψ ) . Applying the rotation matrix (5-9) to the transfor-
mation (5-4) it leads to relatively expensive formulae
The general rotation matrix (5-9) contains the fol- for the resulting conductivity matrix K ij in the three-
lowing three important special cases of axis rotation: dimensional anisotropic case.
(a) Rotation about the x 3 -axis only: Unlike the 2D case, where the additional effort in
prescribing only one Eulerian angle φ for the definition
cos φ sin φ 0 of an anisotropic property of conductivity is still justifi-
( a ij ) = – sin φ cos φ 0 at θ = ψ = 0 (5-10) able, in 3D the Eulerian angles ( φ, θ, ψ ) are often not
0 0 1 utilized as a direct input because, in practice, they are
generally not explicitly known and not given.
(b) Rotation about the x 2 -axis only:
Rather, one can make use of the fact that the princi-
pal axes generally correlate with the geological layer
cos θ 0 sin θ φ = – 90° structure and, accordingly, provided that the three-
( a ij ) = 0 1 0 at (5-11)
ψ = 90° dimensional shape of the layers is known, the spatial
– sin θ 0 cos θ rotation of the principal directions is automatically
accomplished by computation. The conductivity in a
(c) Rotation about the x 1 -axis only: geologically layered structure is often orthotropic inas-
m m
much the higher conductivity K1 = K2 is parallel to
m
the layering and a lower value K 3 is normal to the
1 0 0
( a ij ) = 0 cos θ sin θ at φ = ψ = 0 (5-12)
stratigraphy. It leads to a computational method for
transforming the principal axes of each finite element if
0 – sin θ cos θ
three-dimensional prismatic elements are utilized,
which are capable of fitting their top and bottom faces
UO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
RKP=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó
to the actual stratigraphic layering (Fig. 5.4a). local system ( ξ, η, ζ ) to the global Cartesian system
( x 1, x 2, x 3 ) is achieved by the Jacobian matrix J :
For three-dimensional prismatic finite elements a
method of shape-derived principal directions has been ∂x 1 ∂x 2 ∂x
proposed by Springer66 which is implemented in -------- -------- -------3-
x 1, ξ x 2, ξ x 3, ξ ∂ξ ∂ξ ∂ξ
FEFLOW to compute 3D anisotropy. Using quadrilat-
J = x 1, η x 2, η x 3, η ∂ x1
= -------
∂x 2 ∂x 3
(5-13)
eral or triangular prismatic elements the two opposite - -------- --------
∂η ∂η ∂η
top and bottom faces of each element represent layer x 1, ζ x 2, ζ x 3, ζ
boundaries. A finite element is usually characterized by ∂x ∂x 2 ∂x 3
-------1- -------- --------
its local distorted coordinate system ( ξ, η, ζ ) as shown ∂ζ ∂ζ ∂ζ
in Fig. 5.4b. The coordinate transformation from the
m
u3 K3
a) b)
a ye r
to p l
ζ m
ace u1 K1
to p f
er
lay m
m
tto u2 K2
bo ξ
x3
η .
x2
e
fac
m
tto
bo
x1
Figure 5.4 Layer-oriented principal directions of anisotropy for a prismatic finite element.
cbcilt=ö=UP
RK=^åáëçíêçéó
x 1, η x 1, ξ x 1, ξ x 1, η
u 2 = m 11 ⋅ x 2, η – m 12 ⋅ x 2, ξ (5-15) u 3 = x 2, ξ × x 2, η (5-16)
x 3, η x 3, ξ x 3, ξ x 3, η
x 1, η x 1, ξ x 1, ξ x 1, η
(orthogonalization of x 2, η with respect to x 2, ξ ) (perpenticular to the area spanned by x 2, ξ and x 2, η )
x 3, η x 3, ξ x 3, ξ x 3, η
where
2 2 2
x 1, ξ + x 2, ξ + x 3, ξ x 1, ξ x 1, η + x 2, ξ x 2, η + x 3, ξ x 3, η x 1, ξ x 1, ζ + x 2, ξ x 2, ζ + x 3, ξ x 3, ζ
T
( m ij ) = J ⋅ J = 2
x 1, η +
2
x 2, η +
2
x 3, η x 1, η x 1, ζ + x 2, η x 2, ζ + x 3, η x 3, ζ (5-17)
2 2 2
symm. x 1, ζ + x 2, ζ + x 3, ζ
T 1 0 0
kçíáÅÉI= ãçÇÉäáåÖ ui ⋅ ej
a ij = cos ( u i, e j ) = - , with the basis vectors: e 1 = 0 , e 2 = 1 , e 3 = 0
-------------- (5-18)
ãçî~ÄäÉ= ÑêÉÉ= ëìêJ ui ej
Ñ~ÅÉë=ïáíÜ=íÜÉ=_^pa 0 0 1
íÉÅÜåáèìÉ= EëÉÉ= pÉÅíáçå= OKQF
íÜÉ= Åçãéìí~íáçå= çÑ= íÜÉ to compute the rotation matrix a ij . Finally, it yields:
êçí~íáçå= ã~íêáñ= ERJNVF
Ä~ëÉÇ=çå=íÜÉ=ä~óÉêáåÖ=ÅçåÇáJ
x 1, ξ x 2, ξ x 3, ξ
íáçåë=áë=êÉä~íÉÇ=íç=íÜÉ=áåáíá~ä ------------
- ------------
- ------------
-
m 11 m 11 m 11
EìåãçîÉÇF= ëíê~íáÖê~éÜáÅ
m 11 x 1, η – m 12 x 1, ξ m 11 x 2, η – m 12 x 2, ξ m 11 x 3, η – m 12 x 3, ξ
ãÉëÜ=ëíêìÅíìêÉK ( a ij ) = ---------------------------------------------
- ---------------------------------------------
- ---------------------------------------------
- (5-19)
2 2 2 2 2 2
m 11 m 22 – m 11 m 12 m 11 m 22 – m 11 m 12 m 11 m 22 – m 11 m 12
x 2, ξ x 3, η – x 2, η x 3, ξ x 3, ξ x 1, η – x 3, η x 1, ξ x 1, ξ x 2, η – x 1, η x 2, ξ
--------------------------------------------- --------------------------------------------
- --------------------------------------------
-
2 2 2
m 11 m 22 – m 12 m 11 m 22 – m 12 m 11 m 22 – m 12
In the practical simulation, FEFLOW computes the 3D stratigraphic layer structure. In this way, the aniso-
rotation matrix by using Eq. (5-19) at each Gaussian tropic three-dimensional tensor of conductivity K ij is
integration point of a finite element according to the automatically computed by applying Eq. (5-4).
UQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
RKQ=péÉÅá~ä=`~ëÉë
m ⎫
The key advantages of this method concern that K 11 K 12 K 13 K1 0 0 ⎪
only the conductivities of the three principal directions ⎪
( K ij ) = K 21 K 22 K 23 = 0 K m 0 ⎪
2
along the layer structure need to be input for each dis- ⎪
K 31 K 32 K 33 m ⎬ (5-20)
cretized element, viz.: 0 0 K3 ⎪
⎪
m or ⎪
(1) the conductivity K 1 parallel to layering, ⎪
m
K ii = K i and K ij = 0 for i ≠ j at i = 1,2,3 ⎭
m
(2) the conductivity K2 parallel to layering, and
m This case is to be referred to as axis-parallel aniso-
(3) the conductivity K 3 normal to layering, tropy. Here, different conductivities can be prescribed
in each direction of the x i -axes:
and the 3D shape of the layer geometry is embodied in
the anisotropy relations with a relatively high accuracy. m m m
For 3D: K11 = K 1 , K 22 = K2 and K 33 = K 3 .
cbcilt=ö=UR
RK=^åáëçíêçéó
US=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
S
oÉã~êâ=O=J=qÜÉêã~ä=é~ê~ãÉíÉêëW
S oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
for heat transport to specify a temperature reference tions) within the layering structure of the problem to be
frame, because effects of thermal viscosity can become solved. Principally, they are input at nodal points on
large (e.g., between 10 °C and 150 °C the difference is each mesh slice. In order to do it very effectively, spe-
given by a factor of 7). It is to be noted that T o is to be cific editing tools are available in FEFLOW.
settled in a temperature range as much as possible
which is covered by the initial or boundary conditions Take into account, due to the different procedures in
between maximum and minimum temperatures relating depth coordinates in the 2D case (area-related)
(entrance temperature, background temperature, etc.) and in the 3D case (node-related) small deviations can
as indicated in Tab. 6.13. occur if an unconfined 2D problem is expanded to a 3D
problem and the elemental (centroid) elevations are
oÉã~êâ=Q=J=aÉéíÜJêÉä~íáçåë=~åÇ=~èìáÑÉê=ÉäÉî~íáçåëW automatically adapted to nodal quantities via a patch
The description of the geometric relations in the depth evaluation. If such derivations should not be tolerable it
of an aquifer and an aquifer system differs for 2D and is recommended to redo the assignment process for the
3D applications. elevations in FEFLOW’s problem attribute editor
based on the primary elevation data (measured triplets).
For (essentially) horizontal unconfined 2D condi-
top
tions the top and bottom geometries of the aquifer, x l
bottom
and x l , respectively, are assigned within the flow
material attributing procedures. Accordingly, their ele-
vations are handled as areal properties which are
related to the centroid of a finite element. The relation
of these elevations to nodal points of the finite-element
mesh is performed program-internally in considering
the elevations of elements belonging to the node via a
patch evaluation. For confined 2D problems, vertical or
axisymmetric, the specification of depth relation is not
required. If both confined and unconfined conditions
simultaneously occur the task is to be globally declared
as an unconfined problem and the top elevation of the
top
aquifer x l is appropriately specified as stated in Sec-
top
tion 2.2. For instance, using a numerically large x l -
value (say 1000 m) it always provides a free water
top
table for corresponding hydraulic conditions h < x l .
UU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKN=Pa=cäçï=m~ê~ãÉíÉêë
SKN Pa=cäçï=m~ê~ãÉíÉêë
Table 6.1 Material conditions for 3D flow (with and without mass transport)
Axis-parallel anisotropy:
–4 –1
Conductivity [Kxx] K 11 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kyy] K 22 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kzz] K 33 10 ms 1 5-20, 3-5, 1-30
cbcilt=ö=UV
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
Axis-parallel anisotropy:
–4 –1
Conductivity [Kxx] K 11 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kyy] K 22 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kzz] K 33 10 ms 1 5-20, 3-5, 1-30
VM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKN=Pa=cäçï=m~ê~ãÉíÉêë
Axis-parallele anisotropy:
–4 –1
Conductivity [Kxx] K 11 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kyy] K 22 10 ms 1 5-20, 3-5, 1-30
–4 –1
Conductivity [Kzz] K 33 10 ms 1 5-20, 3-5, 1-30
cbcilt=ö=VN
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
Table 6.4 Material conditions for vertical 2D flow (with and without mass transport)
VO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKO=sÉêíáÅ~ä=çê=^ñáëóããÉíêáÅ=Oa=cäçï=m~ê~ãÉíÉêë
Table 6.5 Material conditions für vertical 2D flow (at heat transport)
cbcilt=ö=VP
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
Table 6.6 Material conditions for vertical 2D flow (at thermohaline transport)
VQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKP=eçêáòçåí~ä=Oa=cäçï=m~ê~JãÉíÉêë
cbcilt=ö=VR
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
VS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKQ=j~ëë=qê~åëéçêí=m~ê~ãÉíÉêë
Table 6.9 Material conditions for 3D, vertical or axisymmetric mass transport
Henry sorption:
Freundlich sorption:
1 – b2
Sorption coefficient b1 ( mg ⁄ l ) 0 1-51, 3-5, 3-7
Langmuir sorption:
cbcilt=ö=VT
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
Table 6.10 Material conditions for 2D horizontal mass transport (confined and unconfined)
Henry sorption:
Freundlich sorption:
1 – b2
Sorption coefficient b1 ( mg ⁄ l ) 0 1-51, 3-5, 3-7
Langmuir sorption:
Sorption coefficient (numerator) k1 1 0 1-52, 3-5, 3-8
VU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKR=eÉ~í=qê~åëéçêí=m~ê~ãÉíÉêë
Table 6.11 Material conditions for 3D, vertical or axisymmetric heat transport
cbcilt=ö=VV
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
Table 6.12 Material conditions for 2D horizontal heat transport (confined and unconfined)
a. For confined conditions and for unconfined conditions at simultaneous application of integral boundary conditions of 3rd
–1 –1 –1
kind the unit for the ’In-transfer rate’ is: J m d K
b. For confined conditions and for unconfined conditions at simulatneous application of integral boundary conditions of 3rd
–1 –1 –1
kind the unit for the ’Out-transfer rate’ is: J m d K
NMM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
SKR=eÉ~í=qê~åëéçêí=m~ê~ãÉíÉêë
Concentration Co mg ⁄ l 0.0
(freshwater condition)
cbcilt=ö=NMN
SK=oÉä~íáåÖ=íÜÉ=cçêãìä~íáçåë=íç=cbcilt
NMO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
oÉÑÉêÉåÅÉë
T
13. Combarnous, M.A. and Bories, S.A. (1975) Hydrothermal con-
T oÉÑÉêÉåÅÉë
cbcilt=ö=NMP
TK=oÉÑÉêÉåÅÉë
21. Diersch, H.-J., Kolditz, O. and Jesse, J. (1989) Finite element 30. Diersch, H.-J.G. and Kolditz, O. (1998) Coupled groundwater
analysis of geothermal circulation processes in hot dry rock frac- flow and transport: 2. Thermohaline and 3D convection systems.
tures. Zeitschr. Angew. Math. Mech. Vol. 69, No. 3, 139-153. Advances in Water Resources Vol. 21, No. 5, 401-425.
22. Diersch, H.-J. (1992) Interactive, graphics-based finite element 31. Diersch, H.-J.G. (2000) Note to the opposing flow regime at
simulation of groundwater contamination processes. Adv. Engi- mixed convection around a heated cylinder in a porous medium.
neering Software Vol. 15, 1-13. Transport in Porous Media Vol. 38, No. 3, 345-352.
23. Diersch, H.-J., Gründler, R, Kaden, S, and Michels. I. (1992) 32. Diersch, H.-J.G. and Kolditz, O. (2002) Variable-density flow and
Toward GIS-based 3D/2D groundwater contamination modeling transport in porous media: approaches and challenges. Advances
using FEM. In: Computational Methods in Water Resources IX, in Water Resources Vol. 25, Nos. 8-12 (Special issue), 899-944.
Vol. 1., Proc. IX. Intern. Conf. on Comp. Methods in Water 33. Elder, J.W. (1967) Transient convection in a porous medium. J.
Resources, June 1992, Denver, Comp. Mech. Publ., Southamp- Fluid Mech. Vol. 27, Part 3, 609-623.
ton, 749-760. 34. Frolkovic, P. and De Schepper, H. (2000) Numerical modelling of
24. Diersch, H.-J. (1993) GIS-based groundwater flow and contami- convection dominated transport with density driven flow in
nant transport modeling - the simulation system FEFLOW. In: porous media. Advances in Water Resources Vol. 24, No. 1, 63-
Praxis der Umweltinformatik Vol. 4 (Rechnergestützte Ermit- 72.
tlung, Bewertung und Bearbeitung von Altlasten), ed. F. Ossing, 35. Goldstein H., Poole, C. and Safko, J. (2002) Classical mechanics.
Metropolis-Verlag, Marburg, 187-208. 3rd edition, Addison Wesly, San Francisco.
25. Diersch, H.-J. (1994) Computational aspects in developing an 36. Gray, W.G. (1982) Derivation of vertically averaged equations
interactive 3D groundwater transport simulator using FEM and describing multiphase flow in porous media. Water Resources
GIS. Intern. Conf. on Groundwater Quality Management (GQM Research Vol. 18, No. 6, 1705-1712.
93), Tallinn, Estonia, 6-9 September 1993, IAHS Publication No. 37. Hassanizadeh, S.M. (1986) Derivation of basic equations of mass
220, 313-326. transport in porous media. Part 1. Macroscopic balance laws, Part
26. Diersch, H.-J., Voigt, R. and Gründler, R. (1994) Visiometric 2. Generalized Darcy’s and Fick’s laws. Advances in Water
techniques in a 3D groundwater transport code. In: Computa- Resources Vol. 9, 196-222.
tional Methods in Water Resources X, Vol. 2, X. Intern. Conf. on 38. Hassanizadeh, S.M. (1988) Modeling species transport by con-
Comp. Methods in Water Resources, July 1994, Heidelberg, Klu- centrated brine in aggregated porous media. Transport in Porous
wer Academic Publishers, 1449-1456. Media Vol. 3, 299-318.
27. Diersch, H.-J.G. (1996) Finite element analysis of three-dimen- 39. Holst, P.H. and Aziz, K. (1972) Transient three-dimensional natu-
sional transient free convection processes in porous media. In: ral convection in confined porous media. Int. J. Heat Mass Trans-
Three-Dimensional Complex Flows, (Proc. Intern. IMACS-COST fer Vol. 15, 73-90.
Conf. Lausanne, September 13-15, Switzerland, 1995), ed. by M.
40. Holzbecher, E. (1998) Modeling density-driven flow in porous
Deville et al., Notes on Numerical Fluid Mechanics Vol. 53,
media. Springer-Verlag, Berlin.
Vieweg & Sohn, Braunschweig/Wiesbaden, 1996, 38-44.
41. Homsy, G.M. and Sherwood, A.E. (1976) Convective instabilities
28. Diersch, H.-J.G. and Michels, I. (1996) Moving finite element
in porous media with through flow. AIChE Journal Vol. 22, No.1,
meshes for simulating 3D transient free surface groundwater flow 168-174.
and transport processes. In: Computational Methods in Water
42. Horne, R.N. and O’Sullivan, M.J. (1974) Oscillatory convection
Resources XI, Vol. 1 Computational Methods in Subsurface Flow
in a porous medium heated from below. J. Fluid Mech. Vol. 66,
and Transport Problems, ed. A.A. Aldama et al., Comp. Mech.
Part 2, 339-352.
Publ., Southampton, 1996, 85-92.
43. Horne, R.N. and O’Sullivan, M.J. (1978) Origin of oscillatory
29. Diersch, H.-J.G. (2002) About the difference between the convec-
convection in a porous medium heated from below. Phys. Fluids
tive form and the divergence form of the transport equation.
Vol. 21, No. 8, 1260-1264.
FEFLOW White Papers - Vol. I, WASY Ltd, Berlin.
NMQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
44. Horne, R.N. (1979) Three-dimensional natural convection in a 58. Robinson, J.L. and O’Sullivan, M.J. (1976) A boundary-layer
confined porous medium heated from below. J. Fluid Mech. Vol. model of flow in a porous medium at high Rayleigh number. J.
92, Part 4, 751-766. Fluid Mech. Vol. 75, Part 3, 459-467.
45. Horne, R.N. and Caltagirone, J.-P. (1980) On the evolution of 59. Rubin, H. (1975) On the analysis of cellular convection in porous
thermal disturbances during natural convection in a porous media. Int. J. Heat Mass Transfer Vol. 18, 1483-1486.
medium. J. Fluid Mech. Vol. 100, Part 2, 385-395. 60. Rubin, H. and Roth, C. (1978) Instability of horizontal thermoha-
46. Horton, C.W. and Rogers Jr., F.T. (1945) Convection currents on line flow in a porous medium layer. Israel Journal of Technology
a porous medium. Vol. 16, J. Appl. Phys., 367-369. Vol. 16, No. 5/6, 216-223.
47. Hsu, C.T., Cheng, P. and Homsy, G.M. (1978) Instability of free 61. Rudraiah, N. and Srimani, P.K. (1980) Finite-amplitude cellular
convection flow over a horizontal impermeable surface in a convection in a fluid-saturated porous layer. Proc. R. Soc. Lond.,
porous medium. Int. J. Heat Mass Transfer Vol. 21, 1221-1228. A 373, 199-222.
48. Knupp, P. (1996) A moving mesh algorithm for 3-D regional 62. Scheidegger, A.E. (1961) General theory of dispersion in porous
groundwater flow with water table and seepage face. Advances in media. J. Geophys. Res. Vol. 66, No. 10, 3273-3278.
Water Resources Vol. 19, No. 2, 83-95. 63. Schubert, G. and Straus, J.M. (1979) Three-dimensional and mul-
49. Kolditz, O. and Diersch, H.-J. (1993) Quasi-steady-state strategy ticellular steady and unsteady convection in fluid-saturated
for numerical simulation of geothermal circulation in hot dry rock porous media at high Rayleigh numbers. J. Fluid Mech. Vol. 94,
fractures. Int. J. Non-Linear Mechanics, Vol. 28, No. 4, 467-481. Part 1, 25-38.
50. Kolditz, O. and Zielke, W. (1994) Modell des geklüfteten porösen 64. Segol, G. (1994) Classic groundwater simulations: proving and
Mediums für Festgesteinsaquifere (Model of the fractured porous improving numerical models. PTR Prentice Hall, Englewood
medium for rock aquifers). Preprint, Institute for Flow Mechan- Cliffs, New Jersey, 531p.
ics, Hannover, 57 p. 65. Simkins, P.G. and Blythe, P.A. (1980) Convection in a porous
51. Kolditz, O., Ratke, R., Diersch, H.-J.G. and Zielke, W. (1998) layer. Int. J. Heat Mass Transfer Vol. 23, 881-887.
Coupled groundwater flow and transport: 1. Verification of den- 66. Springer, J. (1996) Shape-derived anisotropy directions in quad-
sity variable flow and transport models. Advances in Water rangle and brick finite elements. Communications in Numerical
Resources Vol. 21, No. 1, 27-46. Methods in Engineering Vol. 12, 351-357.
52. Lever, D.A. and Jackson, C.P. (1985) On the equations for the 67. Straus, J.M. (1974) Large amplitude convection in porous media.
flow of concentrated salt solution through a porous medium. U.K. J. Fluid Mech. Vol. 64, Part 1, 51-63.
DOE Report No. DOE/RW/85.100. 68. Straus, J.M. and Schubert, G. (1979) Three-dimensional convec-
53. McKibbin, R. and O’Sullivan, M.J. (1980) Onset of convection in tion in a cubic box of fluid-saturated porous material. J. Fluid
a layered porous medium heated from below. J. Fluid Mech. Vol. Mech. Vol. 91, Part 1, 155-165.
96, Part 2, 375-393. 69. Trevisan, O.V. and Bejan, A. (1985) Natural convection with
54. Mercer, J.W. und Pinder, G.F. (1974) Finite element analysis of combined heat and mass transfer buoyancy effects in a porous
hydrothermal systems. Finite Element Methods in Flow Problems medium. Int. J. Heat Mass Transfer Vol. 28, No. 8, 1597-1611.
(Proc. 1st Symp., Swansea, 1974), (ed. Oden, J. T. et al.) Univ. of 70. Wooding, R.A. (1962) Free convection of fluid in a vertical tube
Alabama Press, 401-414. filled with porous material. J. Fluid Mech. Vol. 13, 126-144.
55. Nield, D. A. and Bejan, A. (1992) Convection in porous media. 71. Yeh, G.-T., Cheng, J.-R. and Bensabat, J.A. (1994) A three-
Springer Verlag, New York Inc. dimensional finite-element model of transient free surface flow in
56. Palm, E., Weber, J.E. and Kvernvold, O. (1972) On steady con- aquifers. In: Computational Methods in Water Resources X, Vol.
vection in a porous medium. J. Fluid Mech., Vol. 54, Part 1, 153- 1, X. Intern. Conf. on Comp. Methods in Water Resources, July
161. 1994, Heidelberg, Kluwer Academic Publishers, 131-138.
57. Perrochet, P. (1996) Personal communication, EPFL Lausanne, 72. Younes, A., Ackerer, P., and Mosé, R. (1999) Modeling variable
GEOLEP Laboratoire de géologie, Lausanne (Switzerland). density flow and solute transport in porous medium: 1. Re-evalu-
cbcilt=ö=NMR
TK=oÉÑÉêÉåÅÉë
NMS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
kçãÉåÅä~íìêÉ
U
44), (1-51), (3-7)
U kçãÉåÅä~íìêÉ
cbcilt=ö=NMT
UK=kçãÉåÅä~íìêÉ
2 –1 mol f disp f
D ij L T = D ij + D ij tensor of conductivity, (1-140), (1-141), (5-
hydrodynamic dispersion of fluid 21)
phase, (1-56) –1 3 s
Kd M L = κ ⁄ ρ distribution coefficient,
3 –1
D ij L T depth-integrated tensor of (Chapter 3)
hydrodynamic dispersion, (3-13), K eq 1 equilibrium constant, (1-21)
(3-18) K ij LT
–1
tensor of hydraulic conductivity of
mol α 2 –1
D ij L T tensor of molecular macrodiffusion fluid phase, (1-30), (1-122)
of α -phase, (1-13) K ij
a
LT
–1
effective (intrinsic) conductivity of
disp α 2 –1
D ij L T tensor of mechanical dispersion of fluid phase, (1-123)
α -phase, (1-13) Ki
m
LT
–1
hydraulic conductivity along the
d L thickness, Figs. 1.8, 4.2, 4.3 principal directions, (5-1)
2 –2 –1
E L T internal (thermal) energy density, K max, K min LT maximum and minimum
Tab. 1.1 conductivity, respectively, (5-6), (5-
# 2 –2
E L T activated energy, (1-26) 7)
–1 3
ei 1 gravitational unit vector, (1-28) k 1, k 2 1 and M L Langmuir numerator and
f specific rate of temporary Langmuir denominator sorption
production, (1-2) coefficient, respectively, (1-41), (1-
fk 1 activity coefficient, (1-34) 52), (3-8)
α 2
fμ 1 viscosity relation function, (1-101) k ij L permeability tensor of α -phase, (1-
11)
G extensive property, (1-1)
–2 M number of phases
g LT gravitational acceleration, (1-28) –1
–2 mk Mmol molecular weight of species k, (1-
gi LT gravity vector, (1-28)
34), (1-35)
h L hydraulic head, (1-27)
R R N number of chemical components
h1 , h2 L prescribed boundary values of (species)
hydraulic head h, (4-4), (4-6)
N° number of reactants
hI L initial condition of hydraulic head h,
Ns number of sorbed species, (1-37)
(4-1)
ni 1 normal unit vector (positive
ho L reference hydraulic head, (1-77)
outward), Fig. 4.1
J L Jacobian matrix, (5-13) –1
Po LT groundwater recharge (infiltration
ji diffusive flux vector, (1-6) rate), (2-11), (4-10)
α –2 –1 α
j ik ML T Fickian mass flux vector of α - p ML T
–1 –2
thermodynamic pressure of α -
phase, (1-13) phase, (1-11)
α –3
j iT MT Fourierian heat flux vector of α - p
f
ML T
–1 –2
fluid pressure, (1-116)
phase, (1-14) –1
–1 Q EB T term of extended Boussinesq
K LT coefficient of isotropic hydraulic approximation, (1-110), (1-138)
NMU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
–3 –1 R R† –1 –1
QC ML T mass source term, (1-57) q C, q C ML T depth-integrated prescribed normal
–2 –1 boundary mass flux of the dispersive
QC ML T = B ⋅ Q C depth-integrated mass
source term, (2-3), (3-11), (3-16) and the total (convective plus
QC
R
MT
–1
lumped balance flux of contaminant dispersive) part, respectively, (4-
mass (positive inward), (4-44), (4- 20), (4-25)
R –1
46), (4-49), (4-50) qh LT normal boundary fluid flux (positive
QC
w
MT
–1
mass flux well function, (4-24) outward), (4-5), (4-11)
R 2 –1
R
Qh
3 –1
L T lumped balance flux of fluid qh L T depth-integrated normal boundary
(positive inward), (4-42) fluid flux (positive outward), (4-5),
w 3 –1 (4-11)
Qm L T pumping/injecting rate of single f –1
well m, (4-9) qi LT Darcy velocity of fluid phase, (1-29)
f 2 –1
Qρ T
–1
specific sink/source rate of fluid, (1- qi L T depth-integrated Darcy velocity of
102) fluid phase, (1-69)
–2 –1
Qρ LT
–1
= B ⋅ Q ρ depth-integrated specific qn ML T normal mass flux (positive
C
sink/source rate of fluid, (1-108), (2- outward), (4-19), (4-21), (4-25), (4-
1), (3-9), (3-14) 26)
–1 –1
Qρ
w 3 –1
L T well function, (4-9) qn ML T depth-integrated normal mass flux
C
α –1 (positive outward), (4-20), (4-21)
Qρ T specific mass supply of α -phase, –1
(1-9) qn LT normal Darcy flux of fluid (positive
h
βα –2 –1 outward), (4-5), (4-6), (4-11), (4-12)
Q̂ ρi ML T mass supply through a β-α-interface 2 –1
(positive outward from β to α), (1- qn L T depth-integrated normal Darcy flux
h
cbcilt=ö=NMV
UK=kçãÉåÅä~íìêÉ
NNM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
α –3 –1
α L, α T L longitudinal and transverse λ MLT Θ thermal conductivity of α -phase,
thermodispersivity, respectively, (1- (1-14)
cond α –3 –1
14) λ ij MLT Θ conductive part of thermodispersion
–1
β Θ volumetric thermal fluid expansion tensor of α -phase, (1-14)
disp α –3 –1
coefficient (constant), (1-76) λ ij MLT Θ dispersive part of thermodispersion
β∗
–1
Θ modified thermal fluid expansion tensor of α -phase, (1-14)
function (variable), (1-105) α –1 –1
μ ML T dynamic viscosity of α -phase, (1-
β L, β T L longitudinal and transverse 17), (1-98), (1-99)
dispersivity, respectively, of α 2 –2
μk L T chemical potential of the kth species
chemical component, (1-13) in the α -phase, (1-20)
Γ, Γ i total boundary and boundary f –1 –1
μo ML T reference viscosity of the fluid phase
portion, respectively related to the reference
–1
γ L fluid compressibility, (1-76) concentration C o and the reference
–3 temperature T o , (1-101)
ΔC ML concentration difference, (1-140)
f –1 –1
Δh L difference of hydraulic head, (2-6) μo ML T reference viscosity of the fluid phase
ΔT Θ temperature difference, (1-141) related to the reference
concentration C = 0 and the
δ(…) Dirac delta function
reference temperature T = 150°C ,
δ ij Kronecker tensor ( δ ij = 1 for (1-98), (1-99)
i = j , else δ ij = 0 for i ≠ j )
ν kr stoichiometric number of species k
ε 1 porosity (= volume fraction of fluid and reaction r
phase)
Ξ aniso 1 factor of anisotropy, (5-8)
εα 1 volume fraction of α -phase
ξ 1 inhibition factor, (1-96)
εe 1 effective porosity, (1-114)
ξ local coordinate, Fig. 4.7
ζ local coordinate –3
ρ ML density function
η local coordinate b –3 s
ρ ML = ( 1 – ε )ρ bulk density of porous
θ Eulerian angle medium, (Chapter 3)
–1
ϑ T chemical decay rate, (1-54) ρ
f
ML
–3
density of fluid phase, (1-31), (1-
κ 1 Henry sorptivity coefficient, (1-43), 77), (1-84), (1-87)
(1-50), (3-6) f –3
ρo ML reference density of fluid phase
2 –1 f s f f
Λ L T = ( ελ + ( 1 – ε )λ ) ⁄ ( ρ c ) related to the reference head h o , the
thermodiffusivity, (1-141) reference concentration C o , and the
–3 –1
λ ij MLT Θ tensor of hydrodynamic reference temperature T o , (1-27),
thermodispersion, (1-63) (1-77)
2 –3 –1 s –3
λ ij ML T Θ depth-integrated tensor of ρ ML density of solid phase
–1 –2
hydrodynamic thermodispersion, σ ij ML T stress tensor, Tab. 1.1
(3-13), (3-18)
cbcilt=ö=NNN
UK=kçãÉåÅä~íìêÉ
NNO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
Abbreviations
cbcilt=ö=NNP
UK=kçãÉåÅä~íìêÉ
NNQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=f
m~êí=ff
^ééäáÅ~íáçåë
cbcilt=ö=NNR
NNS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
qÜÉáëÛ=mêçÄäÉã
V
VKP ^ëëìãéíáçåë
V qÜÉáëÛ=mêçÄäÉã
VKN fåíêçÇìÅíáçå
The following assumptions must be introduced to
Theis’ problem examines the transient lowering of obtain an analytical solution:
the water table induced by a pumping well. Theis’ fun- aquifer: confined, infinite areal extend, homogeneous,
damental insight was to recognize that Darcy’s law is isotropic, uniform thickness, horizontal piezometric
analogous to the law of the flow of heat by conduction, surface;
hydraulic pressure being analogous to temperature, well:constant discharge rate, well penetrates the entire
pressure-gradient to thermal gradient. Therefore, the thickness, well storage effects can be neglected.
mathematical theory of heat conduction developed by
Fourier and subsequent writers (see Carslaw and
Jaeger1) is largely applicable to hydraulic theory.’
(cited from Theis5).
cbcilt=ö=NNT
VK=qÜÉáëÛ=mêçÄäÉã
VKR ^å~äóíáÅ~ä=pçäìíáçå
VKQ dçîÉêåáåÖ=bèì~íáçåë The analytical solution of Eq. (9-1) with the condi-
tions Eq. (9-2) and Eq. (9-3) derived by Theis is:
The governing equations of the Theis’ problem are:
Q
h ( t, r ) = h 0 – ---------- W ( u ) (9-4)
2 4πT
--S- ∂-----h – --------
∂ h 1--- ∂h
– = 0 (9-1)
T ∂t ∂r 2 r ∂ r
with the well function:
initial conditions: inf
e ( –ξ )
h ( 0, r ) = h 0 (9-2)
W ( u ) = – Ei ( – u ) = ∫ ----------
ξ
- dξ (9-5)
u
Sym
VKS kìãÉêáÅ~ä=^å~äóëáë=EcbjF
Property Unit
bol
For the numerical analysis of Theis’ problem we
h hydraulic head L use a case study by Segol4, who referred to results from
Pinder and Frind3. They tested different finite element
r radial distance from the well L meshes and element types (cf. Figure 9.2). Table 9.2
S storage coefficient 1 summarizes the parameters for simulation. Figure 9.2
shows the finite element grid used by Pinder and
t time T Frind3. These finite elements are used as superelements
T transmissivity L2 T-1 in our following investigation. The refined mesh is pre-
sented in Figure 9.3.
Q constant discharge rate L3 T-1
NNU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
VKS=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
unsteady flow
well pumping rate Q = 0.5 ft3s-1 = 1.22329x103 m3d-1 q = Q/2 π rb = 638.75425 m2d-1
rb = 1 ft (wellbore radius) rb = 0.3048 m
flow materials:
FEM:
cbcilt=ö=NNV
VK=qÜÉáëÛ=mêçÄäÉã
VKT oÉëìäíë
FEFLOW provides two temporal discretization
techniques: (1) implicit or Crank-Nicolson time step- Figure 9.5 and Table 9.3 show the comparison of
ping with predefined time steps and (2) automatically analytically and numerically calculated drawdown of
controlled time stepping based on predictor-corrector hydraulic head versus time at the distance of r = 9.639
techniques of first (FE/BE) and second order (AB/TR) m from the well.
accuracy, which both are applied in the following
study.
Figure 9.4 Calculated drawdowns at a distance of 31.6228 ft. from the well (from Pinder
and Frind3).
NOM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
VKT=oÉëìäíë
Figure 9.5 Calculated drawdowns at a distance of 9.639 m from the well (FEFLOW).
Constant time
Time in [s] Analytical FE/BE AB/TR
steps
cbcilt=ö=NON
VK=qÜÉáëÛ=mêçÄäÉã
Constant time
Time in [s] Analytical FE/BE AB/TR
steps
VKU oÉÑÉêÉåÅÉë
Up to one day the agreement of analytical and
numerical results is quite well. Afterwards, the numeri- 1. Carslaw, H.S. and Jaeger, J.C. (1959) Conduction of heat in sol-
cal solution is already influenced by the constant head ids. 2nd ed. Oxford: Oxford University Press.
condition at the outer boundary. In contrary, the corre- 2. Kruseman, G.P. and de Ridder, N.A. (1990) Analysis and evalua-
tion of pumping test data, ILRI publication 47, ISBN 90 70754
sponding boundary condition for the analytical solu-
207.
tion, following to Eq. (9-3), is constant head at infinite 3. Pinder, G.F. and Frind, E.O. (1972) Application of Galerkin’s pro-
distance from the well. To improve the agreement of cedure to aquifer analysis. Water Resources Research 8(1), pp.
the drawdown curves for later times the calculation 108-120.
domain must be enlarged. 4. Segol, G. (1994) Classic groundwater simulation: proving and
improving numerical models. PRT Prentice Hall, Englewood
Cliffs, New Jersey.
5. Theis, C.V. (1935) The relation between lowering of the piezo-
metric surface and the rate and duration of discharge of a well
using groundwater storage. Trans. Amer. Geophys. Union 16, pp.
519-524.
NOO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
lÖ~í~=~åÇ=_~åâëÛ=mêçÄäÉãW=NJa=^ÇîÉÅíáçåJaáëéÉêJ
ëáçåJaÉÅ~ó NM
NMKO ^ëëìãéíáçåë
NM lÖ~í~=~åÇ=_~åâëÛ=mêçÄäÉãW=NJa=^ÇîÉÅíáçåJaáëéÉêëáçåJaÉÅ~ó
NMKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
The following assumptions must be introduced to
Figure 10.1 shows typical field situations, which obtain an analytical solution:
permit a reduction of the dimensionality following to aquifer:confined, infinite areal extend, homogeneous,
an 1-D problem isotropic, uniform thickness, horizontal piezometric
surface
well gallery:constant discharge rate, well penetrates the
entire thickness, one-dimensional
Figure 10.1 Typical field situations for the application of 1-D models (from Sauty3 and Kinzelbach1).
cbcilt=ö=NOP
NMK=lÖ~í~=~åÇ=_~åâëÛ=mêçÄäÉãW=NJa=^ÇîÉÅíáçåJaáëéÉêëáçåJaÉÅ~ó
initial condition:
Table 10.1 Model parameters
retardation factor 1
(1 – ε)
R = 1 + ---------------- κ
ε
NOQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NMKQ=^å~äóíáÅ~ä=pçäìíáçå
c- 1 x(1 – γ) x – vγt/R
The analytical solution of Eq. (10-1) with the condi- ---- = --- ⎛⎝ exp ------------------- erfc ------------------------⎞⎠ (10-9)
c0 2 2β L 2 vβ t/R
tions Eq. (10-2) and Eq. (10-3) derived by Ogata and L
Banks2 is the following:
c- 1 x( 1 – γ) x – vγt/R
---- = --- ⎛ exp ------------------- erfc ------------------------ (10-4) NMKR kìãÉêáÅ~ä=^å~äóëáë=EcbjF
c0 2⎝ 2β L 2 vβ t/R L
x(1 + γ) x + vγt/R
+ exp ------------------- erfc ------------------------⎞ The Figure 10.2 shows the finite element grid used
2β L 2 vβ L t/R⎠
for the numerical analysis.
with
γ = 1 + 4ϑRβ L /v (10-5)
cbcilt=ö=NOR
NMK=lÖ~í~=~åÇ=_~åâëÛ=mêçÄäÉãW=NJa=^ÇîÉÅíáçåJaáëéÉêëáçåJaÉÅ~ó
porosity ε = 0.2
longitudinal, trans- β L = 5m
verse dispersivity
FEM
NOS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NMKS=oÉëìäíë
t = 10 d t = 50 d t = 500 d
distance
in [m]
analytical numerical analytical numerical analytical numerical
t = 10 d t = 50 d t = 100 d
distance
in [m]
analytical numerical analytical numerical analytical numerical
cbcilt=ö=NOT
NMK=lÖ~í~=~åÇ=_~åâëÛ=mêçÄäÉãW=NJa=^ÇîÉÅíáçåJaáëéÉêëáçåJaÉÅ~ó
Table 10.4 Comparison of analytical and numerical results (excluding decay) (continued)
t = 10 d t = 50 d t = 100 d
distance
in [m]
analytical numerical analytical numerical analytical numerical
NMKT oÉÑÉêÉåÅÉë
1. Kinzelbach, W. (1992) Numerische Methoden zur Modellierung
des Transports von Schadstoffen im Grundwasser. R. Oldenbourg
Verlag, 2. Auflage, ISBN 3-486-26347-1.
2. Ogata, A. and Banks, R.B. (1961) A solution of the differential
equation of longitudinal dispersion in porous media. Professional
paper No. 411-A, U. S. Geological Survey.
3. Sauty, J.P. (1980) An analysis of hydrodispersive transfer in aqui-
fers. Water Resources Research 16(1): 145-158.
NOU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
NN
ered aquifer situation. The layered aquifer consists of a
NN _êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
NNKN pÅÜÉã~íáÅ= pâÉíÅÜ= çÑ= íÜÉ more permeable upper part (e.g. gravels, quartz or
sands) and a less permeable lower part (e.g. glimmer
^èìáÑÉê= qóéÉ= E`çåÑáåÉÇ sands). Therefore, differential velocities may develop
^èìáÑÉêF within such layered structures.
Figure 11.1 (a) Schematic sketch of the considered aquifer situation and (b) its idealization for modeling pur-
poses (from Thiele and Diersch2).
cbcilt=ö=NOV
NNK=_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
Freshwater and saltwater intrude through the left εR + q x – D xx -------2- – D yy -------2- = 0 (11-1)
∂t ∂x ∂x ∂y
boundary into the model domain. This situation
may be caused e.g. by a well, which is cased in the
upper aquifer section. (A) Bruch and Street considered the following initial
and boundary conditions:
• (B) by Thiele and Diersch2
Transient pollution of an initially precontaminated homogeneous initial condition:
finite layered aquifer:
c ( 0, x, y ) = c 2 (11-2)
Due to the different permeabilities within the lay-
ers the stagnant lower part of the aquifer may be
even more contaminated then the upper part. boundary conditions:
NPM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NNKP=dçîÉêåáåÖ=bèì~íáçåë
⎧ c1 , 0 ≤ y < E ∂
c ( t, L, y ) = 0 ⎫⎪
c ( 0, x, y ) = ⎨ (11-5) ∂x
⎩ c2 , E ≤ y < B ⎪
∂ c ( t, x, 0 ) = 0 ⎪
⎬ (11-7)
∂y ⎪
boundary conditions: ∂ c ( t, x , B ) = 0 ⎪
⎪
∂y ⎭
⎧ c1 , 0 ≤ y < E
c ( t, 0, y ) = ⎨ (11-6)
⎩ c2 , E ≤ y < B
with the following model parameters:
Table 11.1 Model parameters
B aquifer thickness L
retardation factor 1
(1 – ε)
R = 1 + ---------------- κ
ε
cbcilt=ö=NPN
NNK=_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
with
Thiele and Diersch2 extended the analytical solution
4i π
2 2 ofEq. (11-1) for inhomogeneous initial conditions Eq.
Ii = 1 + -------------
2
βL βT (11-9) (11-5) and boundary conditions (Eq. (11-6) - Eq. (11-7)):
B
c c2
- = E
--------------- --- +
c1 – c2 B
∞
1- ⎛ E y ⎛ x x – I i q t/ε x x + I i qt/ε ⎞
+ ∑ ---- sin ⎝ iπ ---⎞⎠ cos ⎛⎝ iπ ---⎞⎠ ⎜ exp ⎛⎝ --------- ( 1 – I i )⎞⎠ erfc ------------------------ + exp ⎛⎝ --------- ( 1 + I i )⎞⎠ erfc ------------------------⎟
iπ B B ⎝ 2β L 2 qβ t/ε 2β L 2 qβ t/ε⎠ (11-10)
L L
i=1
NPO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NNKR=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
Figure 11.2 shows the finite element mesh used for Parameters FEFLOW
the numerical analysis.
flow materials:
unsteady transport
transport initials:
transport boundaries:
cbcilt=ö=NPP
NNK=_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
Parameters FEFLOW
longitudinal, trans- β L = 10 m, β T = 1 m
verse dispersivity
FEM:
solver direct
PCG (ORTHOMIN)
NNKS oÉëìäíë
Figure 11.3 describes the transient saltwater devel-
opment for the case (B) after Thiele and Diersch2 at an
observation point (x,y) = (200m,20m) in the upper
layer. Furthermore, Figure 11.3 answers the question,
how the mixing process is influenced by varying veloc-
ity ratios v 1 /v 2 = ω 1 . As can be seen, there is a signifi-
cant dependence of the saltwater breakthrough
behavior on this velocity relation. Figure 11.3 and
Figure 11.4 compare analytical (Thiele and Diersch2)
and numerical (FEFLOW) results for verification pur-
poses. They reveal a good visual agreement between
the results of both methods.
NPQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NNKS=oÉëìäíë
cbcilt=ö=NPR
NNK=_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
Figure 11.5 Effect of homogeneous (Bruch and Street1) and inhomogeneous ini-
tial conditions (Thiele and Diersch2) within the aquifer.
NPS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NNKT=oÉÑÉêÉåÅÉë
Figure 11.5 illustrates clearly the differences distribution vanishes with proceeding time.
between the models (A) by Bruch and Street and (B)
Thiele and Diersch. In the first case (A) of a homoge- Figure 11.6 shows the temporal contaminant break-
neous initial concentration distribution a common through at different observation points (model B):
monotonic breakthrough behavior of the contaminants 1-3 - within the initially contaminated lower zone
is attained. On the other hand for a stratified initial con- 4 - even at the layer interface
dition of the saltwater concentration typical ’over- 5-6 - within the initially uncontaminated upper zone
shooting’ characteristics develop, which depend on the
velocity ratio. Sooner or later the breakthrough curves In this case the seepage velocity within the upper layer
of both models have to meet at the steady-state solu- is twice as this within the lower one
tions. The temporal impact of the initial contaminant
NNKT oÉÑÉêÉåÅÉë 2. Thiele, M. and Diersch, H.-J. (1986) ’Overshooting’ effects due
to hydrodispersive mixing of saltwater layers in aquifers. Adv.
Water Resources 9: 24-33.
1. Bruch, J.C. and Street, R.L. (1967) Two-dimensional dispersion.
J. Sanit. Eng. Div., ASCE, 93(SA6): 17-39.
cbcilt=ö=NPT
NNK=_êìÅÜ=~åÇ=píêÉÉíÛë=mêçÄäÉãW=i~óÉêÉÇ=^èìáÑÉê
NPU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
NO
NO j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
NOKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
The objective of this chapter is to describe the pro-
cesses important in the transport of solutes through
fractured media. The role of diffusion in solute trans-
port through geologic media has been recognized in
many areas of research. Grisack and Pickens2 reported
that quantitative analysis of solute transport through
fractures with diffusion into a porous matrix can be
considered in several hydrogeologic context. These
include (i) groundwater recharge and associated solute
flux to the water table, (ii) evolution of groundwater
chemistry, secondary fracture mineralogy, and matrix
diagenesis in fractured media, (iii) contaminant trans-
port, (iv) tracer test, (v) groundwater aging, and (vi) ore
deposition in hydrodynamic systems.
Figure 12.1 shows a schematical sketch of the con- Figure 12.1 Schematical sketch of the fracture-matrix
system (from Tang et al.3).
sidered aquifer situation for investigating the contami-
nant transport along a discrete fracture embedded in a
saturated porous rock.
NOKO ^ëëìãéíáçåë
The following assumptions are introduced relating
cbcilt=ö=NPV
NOK=j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
The governing equation balancing the mass of con- with the following parameters:
taminant in the fracture domain ( 0 ≤ y ≤ b ) is:
NQM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NOKQ=^å~äóíáÅ~ä=pçäìíáçå
t 1/2 half-life T
ε kinematic porosity 1
cbcilt=ö=NQN
NOK=j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
c-
----
c0
1
2
ϑRx
= --- exp ⎛ – ----------⎞
⎝ q ⎠ {exp⎛⎝–ε---------------------
ϑR'D' ⎞
bq ⎠
ε R'D'
x ⋅ erfc ⎛ ------------------------------------- x –
⎝ 2bvR t – xR/q
ϑ t – xR/q⎞ +
⎠
(12-5)
ε ϑR'D' ε R'D' ⎫
exp ⎛ --------------------- x⎞ ⋅ erfc ⎛ -------------------------------------- x + ϑ t – xR/q⎞ ⎬
⎝ bq ⎠ ⎝ 2bqR t – xR/q ⎠
⎭
c- 1 ϑRx
---- = --- exp ⎛⎝ – ----------⎞⎠ ⋅
c0 2 q
{exp⎛⎝–ε---------------------
ϑR'D'
bq
x+ ----
D' ⎠
ε R'D' -
R'- ( y – b )⎞ ⋅ erfc ⎛ ------------------------------------
⎝ 2bvR t – xR/q
x + ----
y–b -
R'- --------------------------
D' 2 t – xR/q
– ϑ t – xR/q⎞ +
⎠ (12-6)
ε ϑR'D' ε R'D' -
R'- ( y – b )⎞ erfc ⎛ ------------------------------------
exp ⎛⎝ --------------------- x + ----
y–b -
R'- -------------------------- ⎞⎫
⎠⋅ ⎝ 2bvR t – xR/q + D' 2 t – xR/q + ϑ t – xR/q⎠ ⎬
x ----
bq D' ⎭
the numerical analysis. It is formed by two-dimen- Table 12.2 Simulation parameters (continued)
sional quadrilateral finite elements for the discretiza-
tion of both the porous matrix and the fracture. Parameters FEFLOW
flow materials:
unsteady transport
transport initials:
transport boundaries:
cbcilt=ö=NQP
NOK=j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
NQQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NOKR=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
cbcilt=ö=NQR
NOK=j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
Figure 12.5 presents analytical and numerical con- onstrated, y = 0 corresponds to the fracture centre, y = b
centration history curves at a distance 0.76 m from the corresponds to the fracture surface. For large matrix
contaminant source for different values of matrix diffu- diffusivities the concentration values at the fracture
sivity obtained by Tang et al.3 and Grisack and surface and at the fracture centre (area of maximum
Pickens2, respectively. Especially in the mid range of velocity) are identically. Decreasing the influence of
the diffusivity values considerable discrepancy appears matrix diffusion differences between concentrations at
in breakthrough curves, which result from insufficient the fracture centre and at the fracture surface become
discretization near the fracture interface. obvious. The concentration values differ up to 5 per-
cent. The FEFLOW results for the fracture centre,
Figure 12.6 shows temporal breakthrough curves using the FE mesh illustrated in Figure 12.2, agree
for the same case study as above computed by good with the analytical findings by Tang et al.3, which
FEFLOW at a distance 0.76 m from the source point. based on the concept of average velocity within the
First, the effect of the parabolic velocity profile is dem- fracture.
NQS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NOKR=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
a) Parameters FEFLOW
flow materials:
b) transport materials:
cbcilt=ö=NQT
NOK=j~íêáñ=aáÑÑìëáçå=áå=cê~ÅíìêÉÇ=mçêçìë=jÉÇá~
100
70
– 12 2
D d = 10 m ⁄s
60
50
40
– 11 2
D d = 10 m ⁄s
30
20
– 10 2
D d = 10 m ⁄s
10
0
0 1 2 3 4
time [days]
Figure 12.8 Simulated solute breakthrough in the fracture at a distance of 0.76 m from the
source point; comparison of solutions with and without using fracture elements.
NOKS oÉÑÉêÉåÅÉë 2. Grisak, G.E. and Pickens, J.F. (1980) Solute transport through
fractured media: 1. The effect of matrix diffusion. Water
Resources Research 16(4): 719-730.
1. Diersch, H.-J.G. (2002) Discrete feature modeling of flow, mass
3. Tang, D.H., Frind E.O. and Sudicky, E.A. (1981) Contaminant
and heat transport processes by using FEFLOW. White Papers -
transport in fractured porous media: Analytical solution for a sin-
Vol. I, Chapter 9, WASY Ltd., 147-190.
gle fracture. Water Resources Research 17(3): 555-564.
NQU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
eççéÉë= ~åÇ= e~êäÉã~ååÛë= bñéÉêáãÉåíW= qÜÉ= qïç
tÉää=mêçÄäÉã NP
NP eççéÉë=~åÇ=e~êäÉã~ååÛë=bñéÉêáãÉåíW=qÜÉ=qïç=tÉää=mêçÄäÉã
NPKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
Hoopes and Harlemann1 conducted a number of
experiments using a sand model to measure the distri-
bution of a solute between recharging and discharging
wells. A plane view of their laboratory sand model rep-
resenting a horizontal, confined aquifer is illustrated in
Figure 13.1. Furthermore, Hoopes and Harlemann1
developed an analytical framework to reproduce the
experimental results.
flow isothermal, steady state, 2-D horizontal, con- with the quantities:
stant discharge/ recharge
q Q 2πM 2πM
v = --- = ------------------ ⎛⎝ cosh ⎛⎝ ------------ φ⎞⎠ + cos ⎛⎝ ------------ ψ⎞⎠ ⎞⎠ (13-2)
transport advective, dispersive (along streamlines), dif- ε 2πMεd Q Q
fusive
NRM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NPKQ=^å~äóíáÅ~ä=pçäìíáçå
2 2 c ( 0,φ,ψ ) = 0 ⎫
Q (x + d) + y - ⎫
φ = ------------ ln ------------------------------ ⎪ ⎪
4πM ( x + d ) 2 – y 2 ⎪ c ( t,φ(-d,y),ψ(-d,y) ) = c 0 ⎬ (13-4)
⎬ (13-3) ⎪
Q – 2yd c ( t,0,ψ ) = 0 ⎭
- ⎪⎪
ψ = ------------ arctan ---------------------------
2πM 2 2 2
x +y –d ⎭
and following parameters and quantities:
and the initial and boundary conditions:
Table 13.2 Model parameters
M aquifer thickness L
ε kinematic porosity 1
velocity potential playing the part of the local variable. numerical methods (Huyakorn et al.2). The quality of
However, due to Eq. (13-2) the velocity depends on numerical solutions depends essentially on an appro-
both, the velocity potential and stream function. The priate discretization of the calculation domain because
analytical solution of Eq. (13-1) through Eq. (13-4) of the steep concentration increase (cf. Figure 13.2).
obtained by Hoopes and Harlemann is
We use different meshes for the FE-analysis
c 1 ⎛ I D – t D⎞ (Figure 13.3, Figure 13.4) and observe strong mesh
cD = ----- = --- erfc ⎜ ----------------⎟ (13-5) effects in the following. The wellbores are represented
c0 2 ⎝ 4J ⎠ D
by small circles, which are cut off around the borehole
centers (Figure 13.5)
Qt
t D = --------------------2 ⎫
2πεMd ⎪⎪
φD ⎪
dφ D ⎪
I D = ∫ ---------- ⎪
2 ⎬ (13-6)
v
–∞ D ⎪
φD φ ⎪
DD ⎪
J D = ∫ ------- dφ ⎪
vD
4 D ⎪
–∞ ⎭
NPKR kìãÉêáÅ~ä=^å~äóëáë=EcbjF
Hoopes and Harlemann’s experiments was analyzed Figure 13.4 Finite element grid (mesh B).
by several authors in the literature using analytical and
NRO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NPKR=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
Parameters FEFLOW
unsteady transport
transport initials:
transport boundaries:
constant concentration c ( t, r B ) = c 0
at the injection well
transport materials:
Figure 13.5 Well discretization (detail of supermesh B). aquifer thickness M = 0.089 m
porosity ε = 0.374
The simulation parameters are summarized in the fol-
lowing Table 13.3. sorption coefficient κ = 0, therefore, R = 1
flux at the injection q = Q ⁄ ( 2πr B ) = 6.4327m2d-1 wellbore radius, outer r B = 0.005 m, R o = 1.45 m
well boundary
cbcilt=ö=NRP
NPK=eççéÉë=~åÇ=e~êäÉã~ååÛë=bñéÉêáãÉåíW=qÜÉ=qïç=tÉää=mêçÄäÉã
Because of the extremely small dispersivities within advection-dominated transport well-known problems
the sand-box model, a sharp increase of tracer concen- arise concerning oscillating numerical solutions. Huya-
tration takes place at the observation points. From this korn et al.2 applied the streamline-upwinding method
NRQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NPKS=oÉëìäíë
for damping such oscillations. Unfortunately, from tions). Corresponding to the analytical model by
Figure 13.6 it is not clear whether or not oscillations Hoopes and Harlemann, two different dispersion
occur before and after tracer breakthrough. parameters - βL = 0.0015 m, βL = 0.00294 m - are used.
Applying the streamline upwind method the oscilla-
With help of Figure 13.7 we discuss some numeri- tions disappear. Due to the introduction of additional
cal aspects of modeling with the sparse mesh A (cf. (non-physical) dispersion the concentration fronts are
Figure 13.3). The Galerkin-FEM produces oscillating smoothed. Figure 13.7 provides an idea, how the
solutions. The oscillations grow with decreasing dis- upwind FEM could change the results.
persion (increasing Peclet-numbers, which characterize
the relation of advective and dispersive transport frac-
The discussion of numerical aspects is continued in considerations. The velocity potential and stream func-
Figure 13.8. A refined mesh B (Figure 13.4) was intro- tion produced by a well doublet are given by Eq. (13-3).
duced, which is constructed on the basis of physical The general idea is to refine the mesh in domains of
cbcilt=ö=NRR
NPK=eççéÉë=~åÇ=e~êäÉã~ååÛë=bñéÉêáãÉåíW=qÜÉ=qïç=tÉää=mêçÄäÉã
higher velocities (dominating advection) in the vicinity The remaining differences between analytical and
of the well and otherwise to spare grid points in areas numerical results may be caused by different geometri-
of smaller velocities. cal features. The analytical approach treats the wells as
point-shaped sources/sinks, whereas the numerical
Figure 13.8 documents the success of this operation. model use finite wells corresponding to the laboratory
At first, non-oscillating solutions can be obtained even situation (see Figure 13.5). Furthermore, the analytical
by the Galerkin-FEM. Secondly, using mesh B we can method assumes an infinitely extended sand-box
improve the agreement between numerical and analyti- region, whereas the numerical model reflects the finite
cal results. However, it must recognized that the geometry of the sand-box laboratory.
numerical approximation of extremely steep fronts
requires very expensive local discretizations.
NRS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NPKT=oÉÑÉêÉåÅÉë
NPKT oÉÑÉêÉåÅÉë
1. Hoopes, J.A. and Harlemann, D.R.F. (1967) Wastewater recharge
and dispersion in porous media. ASCE Journal of the Hydraulics
Division 93 (HY5): 51-71.
2. Huyakorn, P.S., Andersen, P.F., Güven, O. and Molz, F.J. (1986)
A curvilinear finite element model for simulating two-well tracer
tests and transport in stratified aquifers. Water Resources
Research 22(5): 663-678,
3. Segol, G. (1994) Classic groundwater simulation: proving and
improving numerical models. PRT Prentice Hall, Englewood
Cliffs, New Jersey.
cbcilt=ö=NRT
NPK=eççéÉë=~åÇ=e~êäÉã~ååÛë=bñéÉêáãÉåíW=qÜÉ=qïç=tÉää=mêçÄäÉã
NRU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
NQ
investigate numerical details for the simulation of
NQ bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
The numerical results of Elder for this problem of NQKO _çìëëáåÉëè= ^ééêçñáã~J
complex natural convection are frequently recom- íáçå
mended for the verification of simulators (e.g.
Diersch7, Voss and Souza14, Kolditz et al.13, Diersch Commonly, the so-called Boussinesq approximation
and Kolditz9, Ackerer et al.1, Frolkovic and De is used for describing steady flow in hydrogeologic
Schepper6). Diersch et al.8, Kolditz and Diersch11 systems. Basically, the Boussinesq approxi-mation sub-
cbcilt=ö=NRV
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
stitutes the conservation of mass by the conservation of The model assumptions are summarized in Table 14.1.
fluid volume only. It takes the form (Bear1):
Table 14.1 Model assumptions
∇q = 0 (14-1)
aquifer homogeneous, isotropic, constant
thickness
The Boussinesq approximation is applicable in many
flow steady state, incompressible,
groundwater problems in which the considered fluid is
2-D vertical
incompressible and its density is only weakly depen-
Boussinesq approximation is not
dent on temperature and solute concentration. recommended
NSM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKQ=dçîÉêåáåÖ=bèì~íáçåë=J=cçêãìä~íáçå=áå=qÉêãë=çÑ=eóÇê~ìäáÅ=eÉ~Çë
μ 0 ∂w ∂h
transport
∂w μ 0
∫ Kij -----
-
μ ∂ xi ∂ xj
dΩ = –∫ K ------ β ( c – c 0 )e j dΩ
∂ x i ij μ c
(14-5)
ρ = ρ0 ( 1 + βc ( c – c0 ) ) ρ = ρ0 ( 1 – βT ( T – T0 ) ) (14-8)
Ω Ω
f f
ρ0 ∂c ρ 0 ∂c
– ∫ w -----f q i β c
ρ ∂x i
dΩ – ∫ wε -----f β c dΩ
ρ ∂t
+ ∫ wQ h dΩ 1 ∂ρ
β c = -----
1 ∂ρ
β T = – ----- (14-9)
Ω Ω Ω ρ0 ∂ c ρ0 ∂ c
– ∫ wq n dΩ
2 3
Ω μ = μ 0 ( 1 + 1.85ω – 4.1ω + 44.5ω ) (14-10)
μ0
The second and the third terms on the right side of μ = -----------------------------------------------------------2-
Eq. (14-5) appear due to the extended formulation of 1 + 0.7063ξ – 0.04832ξ
the Boussinesq approximation. The third term will be
activated only for predictor-corrector time stepping –6 T – 150
ω = 10 c ξ = ------------------ (14-11)
150
schemes.
cbcilt=ö=NSN
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
L x h
q D = ---- q x D = --- h D = ----------------
Λ L Lβ T T 0 with the dimensionless quantities
(14-14)
c T Λ
c D = ----- T D = ----- t D = -----2 t KLβ T T 0
c0 T0 L Ra = -------------------- (14-18)
Λ
g εD d
q D = – Ra ⎛⎝ ∇h D – N ( c D – 1 ) -----⎞⎠ (14-15) Le = --------- (14-20)
g Λ
g
q D = – Ra ∇h D + ( T D – 1 ) -----⎞
⎛ where
⎝ g⎠
f s
∂c D ελ + ( 1 – ε )λ
Λ = -----------------------------------
- (14-21)
ε + q D ∇c D – LeΔc D = 0 (14-16) cρ
f f
∂ tD
εc
f f
ρ + ( 1 – ε )c ρ- ∂T D
s s The following table summarizes the model parameters
--------------------------------------------- + q D ∇T D – ΔT D = 0 (14-17) and its values used in the present study.
cρ
f f ∂ tD
h piezometric head L
NSO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKQ=dçîÉêåáåÖ=bèì~íáçåë=J=cçêãìä~íáçå=áå=qÉêãë=çÑ=eóÇê~ìäáÅ=eÉ~Çë
Le Lewis number 1
M aquifer thickness L
Pe Peclet number 1
Sh Stroughal number 1
w basic function 1
Ω calculation domain L2
cbcilt=ö=NSP
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
∂T D ⎛ ∂ψ D ∂ψ D⎞
From the definition of vorticity one can obtain immedi- = ΔT D – ∂( ψ D ,T D ) = ΔT D – ⎜ , ⎟ ∇T D (14-28)
∂ tD ⎝ ∂ zD ∂ xD ⎠
ately (Elder omits the factor 1/2 commonly used for the
vorticity)
The following table summarizes the data of the ther-
NSQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKR=bäÇÉêÛë=bèì~íáçåë=J=cçêãìä~íáçå=áå=qÉêãë=çÑ=píêÉ~ã=cìåÅíáçåë
mal experiment by Elder4. Unfortunately, Elder reports properties of the silicon-oil Elder uses in his experi-
only the concentrated dimensionless quantities, and the ments. Therefore, we add some material data for sili-
back-transformation to physical quantities is compli- con-oil from the literature.
cated. No data can be reconstructed for the thermal
Table 14.3 Original data from Elder’s thermal experiment
ε kinematic porosity 1 1
cbcilt=ö=NSR
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
steady flow
flow boundaries:
NSS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKS=kìãÉêáÅ~ä=^å~äóëáë=fW=`êçëëJsÉêáÑáÅ~íáçå=ïáíÜ=çíÜÉê=jçÇÉäë
Table 14.4 Simulation parameters (continued) Table 14.4 Simulation parameters (continued)
thermal expansion β T = 2000 x [10-4 K-1] molecular diffusivity D d = 3565 x [10-9 m2s-1]
coefficient longitudinal, trans- β L = 0.0, β T = 0 m
verse dispersivity
transient transport
decay rate ϑ = 0 [10-4s-1]
mass transport initials:
heat transport materials:
homogeneous concen- c ( t=0, x, y ) = 0
tration aquifer thickness M =1m
cbcilt=ö=NST
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
Figure 14.4 Comparison of nondimensional results for the Elder problem, showing 20 and 60% of
maximum concentration for SUTRA (solid curves) and of maximum temperature from Elder (dashed
curves) (from Voss and Souza14).
It is important to point out that Elder simulated ther- agreement between numerical solutions derived by dif-
mal convection using a finite difference scheme, ferent methods and codes for complex, nonlinear prob-
whereas Voss and Souza considered the analogous case lems.
for solute movement in a thin cavity using the finite-
element method. Despite of the different numerical The following table should provide an overview on
methods applied, we feel both solutions agree qualita- the simulation runs presented in order to show the
tively well. However, it is important to realize at the impact of distinct model assumptions and numerical
beginning of our study, that one cannot expect perfect
NSU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKS=kìãÉêáÅ~ä=^å~äóëáë=fW=`êçëëJsÉêáÑáÅ~íáçå=ïáíÜ=çíÜÉê=jçÇÉäë
aspects.
Table 14.5 Simulation overview
Topic Realization
The next table summarizes the simulations carried out time of 10 years all results agree quite well with those
in order to compare our results with former findings by of Elder as well as Voss and Souza.
Elder as well as Voss and Souza. Up to a simulation
Table 14.6 Comparing results
cbcilt=ö=NSV
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
In our numerical investigation of the Elder problem we approximation (BA) and the extended BA appear at
pay intention to the following numerical and physical later simulation times, e.g. after 10 years. As seen
aspects: above only one new term is introduced using constant
time stepping due to local density variations by the
NQKSKN d~äÉêâáå=îÉêëìë=ìéïáåÇJcbj extension of the BA. This formulation encourages
stronger circulation (Figure 14.7).
As expected the upwind-FEM produces in general
smoother contours of concentrations. Furthermore, it Applying the extended BA in connection with pre-
provides better agreement with Elders results at later dictor/corrector time stepping methods the circulation
simulation times, e.g. from 15 through 20 years. pattern for later times return to those of the BA (cf.
Figure 14.6 F with Figure 14.8 F and Figure 14.9 F).
NQKSKO jÉëÜ=ÉÑÑÉÅíë These results compare well with those obtained with
BA using the mesh A (Figure 14.5).
Comparing the Figure 14.5 a-f with Figure 14.6 a-f
and Figure 14.5 A-F with Figure 14.6 A-F one can see, NQKSKQ _çìåÇ~êó=ÅçåÇáíáçåë
that the salt distribution depends on the finite-element
discretization. It seems that changes of mesh density Elder as well as Voss and Souza used different
give rise for changes of the convection pattern. boundary conditions. Elder prescribed zero tempera-
tures on the top and the sides of the convection cell.
NQKSKP ^ëëìãéíáçåë= Ñçê= íÜÉ= ÅçåëÉêî~J Whereas Voss and Souza defined the condition of no
mass flux through the boundary beside the bottom and
íáçå=çÑ=ÑäìáÇ=ã~ëë
the part of the top (see Figure 14.1). These different
boundary condition have no obvious influence on the
Comparing Figure 14.6 through Figure 14.8 it
results.
becomes obvious that differences from the Boussinesq
NTM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKS=kìãÉêáÅ~ä=^å~äóëáë=fW=`êçëëJsÉêáÑáÅ~íáçå=ïáíÜ=çíÜÉê=jçÇÉäë
NQKSKR qáãÉ=ëíÉééáåÖ=ëÅÜÉãÉë
a A
We use semi-implicit (Crank-Nicolson), fully
implicit and predictor-corrector time stepping schemes.
It was found that the results for later simulation times
(after 15 years) depend strongly on the time stepping b B
NQKSKS pçäîÉêë
e E
f F
cbcilt=ö=NTN
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
FEM: standard Galerkin FEM: streamline upwind FEM: standard Galerkin FEM: streamline upwind
time: fully implicit, const. step time: fully implicit, const. step time: fully implicit, const. step time: fully implicit, const. step
Figure 14.6 Model: Boussinesq approximation, mesh B Figure 14.7 Model: extended Boussinesq approximation,
(Figure 14.3). mesh B (Figure 14.3).
NTO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKS=kìãÉêáÅ~ä=^å~äóëáë=fW=`êçëëJsÉêáÑáÅ~íáçå=ïáíÜ=çíÜÉê=jçÇÉäë
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
FEM: standard Galerkin FEM: streamline upwind FEM: standard Galerkin FEM: streamline upwind
time: FE/BE predictor-corrector time: FE/BE predictor-corrector time: AB/TR predictor-corrector time: AB/TR predictor-corrector
Figure 14.8 Model: extended Boussinesq approximation, Figure 14.9 Model: extended Boussinesq approximation,
mesh B (Figure 14.3). mesh B (Figure 14.3).
cbcilt=ö=NTP
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
FEM: standard Galerkin FEM: streamline upwind solver: iterative solver: direct
time: semi-implicit Crank- time: semi-implicit Crank-
Nicolson Nicolson
Figure 14.10 Model: extended Boussinesq approximation. Figure 14.11 Model: extended Boussinesq approximation.
NTQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
NQKT kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëJ
ÅêÉíáò~íáçå= bÑÑÉÅíë= ~åÇ
_~ä~åÅÉ=bèì~íáçåë
NQKTKN jÉëÜ=ÉÑÑÉÅíë
Topic Realization
Figure 14.12 Finite-element mesh C (number of elements - salt mass balance convective form of transport equation
ne = 4400, number of grid points - np = 4539).
fluid mass bal- extended Boussinesq approximation
ance
cbcilt=ö=NTR
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
Inspecting Table 14.8 we realize that fulfilling the The previous study of mesh effects reveals that conver-
standard criteria for the Courant and Peclet numbers gent solutions due to the local discretization can be
does not guarantee accuracy for this type of non-linear obtained by use of mesh C, which provides the basis
differential equations. Therefore, for the investigation for the following considerations.
of stability and accuracy of these non-linear problems,
consecutive refinement of local and temporal discreti- Now we continue the discussion in 14.6 concerning
zations is strongly recommended. the time discretization schemes and the extended
Boussinesq approximation for the fluid mass balance.
Furthermore, we investigate effects resulting from the
Table 14.8 Mesh characteristics
numerical realization of the different formulations of
the transport equation: the convective and the diver-
Topic Mesh B Mesh C Mesh D
gence forms of the salt mass balance.
elements ne = 1100 ne = 4400 ne = 9900
44 x 25 88 x 50 132 x 75 Table 14.9 provides an overview on the presented
simulations. The abbreviations are: BA - Boussinesq
grid points np = 1170 np = 4539 np = 10108
approximation, EBA - extended Boussinesq approxi-
Courant- Crx = 0.089 Crx = 0.178 Crx = 0.268 mation, CFTE - convective form of transport equation,
number Crz = 0.203 Crz = 0.406 Crz = 0.608 DFTE - divergence form of transport equation.
Peclet-num- Pgx = 1.771 Pgx = 0.885 Pgx = 0.590
ber Pgz = 0.779 Pgz = 0.390 Pgz = 0.260
NTS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
cbcilt=ö=NTT
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
NQKTKO qáãÉ=ÇáëÅêÉíáò~íáçå
NTU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
Figure 14.15 Effect of temporal discretization: fully implicit, CN and AB/TR schemes; basic
case: CFTE and BA.
cbcilt=ö=NTV
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
AB/TR
CN
fully implicit
Figure 14.16 Effect of temporal discretization: fully implicit, CN and AB/TR schemes; basic
case: CFTE and EBA.
NUM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
Figure 14.17 Effect of temporal discretization: fully implicit, CN and AB/TR schemes; basic
case: DFTE and BA.
cbcilt=ö=NUN
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
AB/TR
CN
fully implicit
Figure 14.18 Effect of temporal discretization: fully implicit, CN and AB/TR schemes; basic
case: DFTE and EBA.
NUO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
cbcilt=ö=NUP
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
a A a A
b B b B
c C c C
d D d D
e E e
f F f
fluid mass balance: BA fluid mass balance: EBA fluid mass balance: BA fluid mass balance: EBA
Figure 14.20 Basic case: CFTE, Crank-Nicolson time Figure 14.21 CFTE, predictor-corrector AB/TR time
scheme, Galerkin-FEM, iterative solver, mesh C. scheme, Galerkin-FEM, iterative solver, mesh C.
NUQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
fluid mass balance: BA fluid mass balance: EBA fluid mass balance: BA fluid mass balance: EBA
Figure 14.22 Basic case: DFTE, fully implicit time Figure 14.23 Basic case: DFTE, Crank-Nicolson time
scheme, Galerkin-FEM, iterative solver, mesh C. scheme, Galerkin-FEM, iterative solver, mesh C.
cbcilt=ö=NUR
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
∂C
a A ε + q∇C – εD d ΔC = 0 (14-29)
∂t
∂ ( εC ) ∇ qC εD
+ ( – d ∇C ) = 0 (14-30)
∂t
c C
The divergence form of the transport equation has a
distinct advantage to conserve the total (dispersive +
advective) contaminant flux through boundaries. How-
d D ever, the numerical handling of these boundary terms is
complicated.
∂C α ∂C
f F
ε + q∇C – εD d ΔC = C ------ ⎛⎝ ε + q ∇C⎞⎠ (14-31)
∂t Cs ∂ t
fluid mass balance: BA fluid mass balance: EBA This means, both formulations for the transport equa-
tion are equivalent only if the Boussinesq approxima-
tion is consequently employed.
Figure 14.24 Basic case: DFTE, predictor-corrector AB/
TR time scheme, Galerkin-FEM, iterative solver, mesh C.
Comparing the salinity evolutions in the following
figures computed by the different formulations of the
transport equation we find that the DFTE forces the
NQKTKQ cçêãìä~íáçå= çÑ= íÜÉ= íê~åëéçêí saltwater intrusion process. In this case the isohalines
Éèì~íáçåë propagate more rapidly into the depth of the porous
cavity.
FEFLOW permits the computation of two alterna-
tive formulations of the contaminant transport equation
termed as the convective form of the transport equation
(CFTE),
NUS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
convective form of transport divergence form of transport convective form of transport divergence form of transport
equation (CFTE) equation (DFTE) equation (CFTE) equation (DFTE)
Figure 14.25 Basic case: BA, fully implicit time scheme, Figure 14.26 Basic case: BA, Crank-Nicolson time
Galerkin-FEM, iterative solver, mesh C. scheme, Galerkin-FEM, iterative solver, mesh C.
cbcilt=ö=NUT
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
a A a A
b B b B
c C c C
d D d D
e E e E
f F f F
convective form of transport divergence form of transport convective form of transport divergence form of transport
equation (CFTE) equation (DFTE) equation (CFTE) equation (DFTE)
Figure 14.27 Basic case: BA, predictor-corrector AB/TR Figure 14.28 Basic case: EBA, fully implicit time scheme,
time scheme, Galerkin-FEM, iterative solver, mesh C. Galerkin-FEM, iterative solver, mesh C.
NUU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKT=kìãÉêáÅ~ä=^å~äóëáë=ffW=aáëÅêÉíáò~íáçå=bÑÑÉÅíë=~åÇ=_~ä~åÅÉ=bèì~íáçåë
a A a A
b B b B
c C c C
d D d D
e E E
f F F
convective form of transport divergence form of transport convective form of transport divergence form of transport
equation (CFTE) equation (DFTE) equation (CFTE) equation (DFTE)
Figure 14.29 Basic case: EBA, Crank-Nicolson time Figure 14.30 Basic case: EBA, predictor-corrector AB/TR
scheme, Galerkin-FEM, iterative solver, mesh C. time scheme, Galerkin-FEM, iterative solver, mesh C.
cbcilt=ö=NUV
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
Figure 14.31 Streamlines and velocities at t = 20 years for the Elder problem.
Figure 14.32 Hydraulic heads and velocities at t = 20 years for the Elder problem.
NVM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NQKU=oÉÑÉêÉåÅÉë
cbcilt=ö=NVN
NQK=bäÇÉêÛë=mêçÄäÉãW=k~íìê~ä=`çåîÉÅíáçå
NVO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
NR
NR qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
NRKN fåíêçÇìÅíáçå
Thermohaline processes in groundwater systems are
connected with the presence of heterogeneous tempera-
ture and concentration fields. Thus, complex convec-
tions can arise from heat and salinity gradients acting
simultaneously (Murray and Chen4, Shen5, Angirasa
and Srinivasan1). Geophysical applications of thermo-
haline models can be found e.g. in the field of geother-
mics and waste disposal in salt formations (Evans and
Nunn3). In this way thermohaline effects gain impor- Figure 15.1 Definition of the thermohaline Elder problem
tance for the production of mineralized thermal water (adapted from Voss and Souza6).
and groundwater movement near salt domes.
cbcilt=ö=NVP
NRK=qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
h piezometric head L
NVQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NRKQ=kìãÉêáÅ~ä=^å~äóëáë
Le Lewis number 1 1
T temperature K
cbcilt=ö=NVR
NRK=qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
NVS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NRKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
Table 15.2 Simulation parameters (continued) the salinity variations (we calculate also the cases
N = 1000, 100) with the asymptotic (N = ∞ ) pure salt-
Parameters FEFLOW water Elder problem could be used for the plausibility
of the thermohaline models.
longitudinal, trans- β L = 0.0, β T = 0.0 m
verse dispersivity
The next series of Figure 15.3 is dedicated to inves-
decay rate ϑ = 0 [10-4s-1] tigate the onset of the thermohaline convection more in
detail. We consider the results for decreasing buoyancy
heat transport materials: ratios N = 5,4,3,2. Beginning with N = 4 the double-dif-
aquifer thickness M =1m fusive convection system is dramatically influenced by
the superposing thermal convection. There are not
porosity ε = 0.1 monotonic changes in the salinity pattern. Surprisingly
thermal capacity of the
s s
c ρ = 0.0 x 106 J m-3 K-1 we find non-symmetrical salinity pattern for the long-
solid phase term situations if the influence of thermal convective
s
currents becomes stronger (N = 2). We suspect numeri-
thermal conductivity of λ = 1.5914444 J m-1 s-1 K-1 cal effects. Therefore, again we use a finer mesh con-
the solid phase sisting of 9900 elements and 10108 nodes (Figure
14.13). The long-term salinity patterns for small buoy-
ancy ratios are illustrated in Figure 15.4. Now we
derive a symmetric salinity distribution for N = 2. But
NRKR oÉëìäíë=~åÇ=aáëÅìëëáçå there is a quite different salinity evolution with respect
to N = 3,4. The flow system suffers further qualitative
The case of pure saltwater intrusion presented in changes. For equilibrated solute and thermal buoyancy
Figure 14.18 provides the basis for the comparison of effects (N = 1) again non-symmetric salinity pattern are
the results of the thermohaline simulations. This case revealed. Because of exhausting computation time we
corresponds to the best available numerical approxima- have not tried further mesh refinements grid conver-
tion: divergence formulation of the transport equations, gence.
extended Boussinesq approximation, Galerkin-FEM,
predictor-corrector forward Adams-Bashforth/back- It seems, that the numerical solution reflects the
ward trapezoid Rule time integration scheme. physical instabilities which the thermohaline flow sys-
tems encounters if solute and thermal effects are nearly
The series of Figure 15.2 show the salinity intrusion equilibrated. We understand that modeling of such
for different buoyancy ratios N = ∞ , 10, 5, 4. Up to instable thermohaline systems will be very expensive
N = 5 we detect only small changes in the salinity evo- especially in three dimensions. Finally, Figure 15.5
lutions. This agreement of the results for higher buoy- presents the temperature and the salinity distributions
ancy ratios where density alterations are dominated by for the case N = 5.
cbcilt=ö=NVT
NRK=qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
N=4
N=5
N = 10
N= ∞
Figure 15.2 Salinity evolutions (t = 1,2,4,10,15,20 yr) for different buoyancy ratios N starting from the
pure saltwater intrusion Elder problem (from bottom to top: N = ∞ ,10,5,4).
NVU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NRKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
N=2
N=3
N=4
N= 5
Figure 15.3 Salinity evolutions (t = 1,2,4,10,15,20 yr) for different buoyancy ratios (from bottom to
top: N = 5,4,3,2), onset of thermohaline effects.
cbcilt=ö=NVV
NRK=qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
N=1
N=2
N= 4
Figure 15.4 Salinity evolutions (t = 10,15,20 yr) for small buoyancy ratios (from bottom to top:
N = 4,2,1), computed with the very fine mesh (Figure 14.13).
OMM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NRKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
qÜÉ= íÉãéÉê~íìêÉ
salinity C temperature T ~åÇ= ë~äáåáíó= ÇáëíêáJ
Äìíáçåë= ~êÉ= ÇáëJ
éä~óÉÇ= Ñçê= í= Z= NI= OI= QI= NMI
NR= ~åÇ= OM= óÉ~êë= EÑêçã= íçé
íç=ÄçííçãFK=
cbcilt=ö=OMN
NRK=qÜÉêãçÜ~äáåÉ=`çåîÉÅíáçå
NRKS oÉÑÉêÉåÅÉë
1. Angirasa, D. and Srinivasan, J. (1989) Natural convection flows
due to the combined buoyancy of heat and mass diffusion in a
thermally stratified medium. J. Heat Transfer 111, 657-663.
2. Diersch, H.-J.G. and Kolditz, O. (1998) Coupled groundwater
flow and transport: 2. Thermohaline and 3D convection systems.
Advances in Water Resources 21(5), 401-425.
3. Evans, G.E. and Nunn, J.A. (1989) Tree thermohaline convection
in sediments surrounding a salt column. J. Geophysical Research
94(B9), 12413-12422.
4. Murray, B.T. and Chen, C.F. (1989) Double-diffusive convection
in a porous medium. J. Fluid Mechanics 201, 147-166.
5. Shen, C.Y. (1989) The evolution of the double-diffusive instabili-
ties: Salt fingers. Phys. Fluids A1(5).
6. Voss, C.I. and Souza, W.R. (1987) Variable density flow and sol-
ute transport simulation of regional aquifers containing a narrow
freshwater-saltwater transition zone. Water Resources Research
23(10), 1851-1866.
OMO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
eÉåêóÛë=mêçÄäÉã
NS
NSKO pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
NS eÉåêóÛë=mêçÄäÉã
NSKN fåíêçÇìÅíáçå
NSKOKN aÉÑáåáíáçå
Henry’s problem describes the advance of a saltwa-
ter front in a confined aquifer which is initially charged The idealized aquifer for the simulation of Henry’s
with uncontaminated freshwater. Henry3 develops a problem is shown in Figure 16.1. The boundary condi-
solution technique to this problem which can calculate tions for flow consist of impermeable boundaries along
the steady-state solute distribution. To this end he the top and the bottom. Hydrostatic pressure is
applies the Boussinesq approximation implying the assumed along the vertical boundary of the sea side.
existence of stream functions. The aquifer is charged at a constant flux with freshwa-
ter. At the inland side the concentration is zero which
Henry3 derives analytical expressions for the stream corresponds to freshwater. At the coastal side concen-
function and the concentration in the form of Fourier tration of brine is set (Eq. (16-1).
series. The resulting algebraic equations for the deter-
mination of the coefficients of the Fourier series must The steady-state flow pattern and the concentration
be obtained by numerical techniques. The ’mystery’ of distribution derived by Henry3 are shown in the
Henry’s solution is, that no numerical model so far has Figure 16.2 and Figure 16.3 Additionally the sharp-
been reproduced closely his semi-analytical model (cf. interface solution (Ghyben-Herzberg interface) is
Segol5). Nevertheless, because of the absence of any superposed on the isochlors.
other non-numerical technique for this kind of non-lin-
ear problems, Henry’s solutions has become one of the
standard tests of variable density groundwater models.
cbcilt=ö=OMP
NSK=eÉåêóÛë=mêçÄäÉã
NSKOKO aÉåëáíó=çÑ=ëçäìíáçåë
Figure 16.1 Definition of the Henry problem (from Table 16.1 Density, concentration and mass fraction
Kröhn4).
Mass density Mass Mass fraction
concentration
[kg m-3] [kg m-3] [1]
w w w
water ρ = 1000 C = 988.398 ω = 0.9643
s s s
salt ρ = 1700 C = 36.592 ω = 0.0357
b b b
brine ρ = 1024.99 C = 1024.99 ω =1
OMQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NSKP=dçîÉêåáåÖ=bèì~íáçåë
NSKP dçîÉêåáåÖ=bèì~íáçåë
We shortly summarize the basic equations for salt-
water intrusion problems. For transient flow problems
the fluid mass is balanced by
∂h α ∂C
S0 + ∇q = Q ρ – ------ ⎛ ε + q∇C⎞ (16-4)
∂t Cs ⎝ ∂ t ⎠
∂h α ∂C
S0 + ε ------ = Q ρ + KΔh (16-5)
∂t Cs ∂ t
Figure 16.4 Mass concentrations of salt, water and solu-
α α
tion. + K ⎛ ∇h + ------ ( C – C 0 )e z⎞ ------ ∇C
⎝ Cs ⎠ Cs
cbcilt=ö=OMR
NSK=eÉåêóÛë=mêçÄäÉã
h piezometric head L
k -9 m2 L2
permeability of the porous medium 1.019368 x 10
K hydraulic conductivity of the porous medium 10-2 m s-1 L T-1
Ra Rayleigh number 1
OMS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NSKQ=kìãÉêáÅ~ä=^å~äóëáë
NSKQ kìãÉêáÅ~ä=^å~äóëáë
The simulation parameters we use for the computa-
Figure 16.5 and Figure 16.6 present the finite ele- tions are summarized in Table 16.3.
ment meshes for the following computations. The
coarse mesh A corresponds to one that is frequently
Table 16.3 Simulation parameters
used in the literature for verification purposes. How-
ever, this mesh produces a lot of numerical dispersion Parameters FEFLOW
because of its grid Peclet number equal to 20. The finer
mesh B is applied to obtain more accurate results. problem class: flow and mass transport
Galerkin-FEM
Figure 16.5 Finite element mesh A (finite elements
ne = 200, nodal points np = 231). flow data (steady-state)
flow boundaries:
w b
hydrostatic heads at the h = ρ ⁄ρ ⋅z
coast boundary
flow materials:
cbcilt=ö=OMT
NSK=eÉåêóÛë=mêçÄäÉã
OMU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NSKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
a)
b)
c)
Figure 16.7 Positions of isochlors at equilibrium (coarse mesh results) - (a) const. dispersion D d = 6.6 x 10-6 m2s-1,
(b) const. dispersion D d = 18.86 x 10-6 m2s-1, and (c) velocity-dependent dispersion β L = 0.035, β T = 0.035 m; left
hand side: FEFLOW results, right hand side: (a)+(b) Voss and Souza6, (c) Galeati et al.2.
cbcilt=ö=OMV
NSK=eÉåêóÛë=mêçÄäÉã
a)
b)
c)
Figure 16.8 Positions of isochlors at equilibrium (fine mesh results) - (a) const. dispersion D d = 6.6 x 10-6 m2s-1, (b)
const. dispersion D d = 18.86 x 10-6 m2s-1, and (c) velocity-dependent dispersion β L = 0.035, β T = 0.035 m.
ONM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NSKS=oÉÑÉêÉåÅÉë
In Figure 16.9 and Figure 16.10 finally, we present 2. Galeati, G., Gambolati, G. and Neumann, S. (1992) Coupled and
the streamline and hydraulic head distributions com- partially coupled Eulerian-Lagrangian model of freshwater-sea-
water mixing. Water Resources Research 28(1): 149-165.
puted with the fine mesh B.
3. Henry, H.R. (1964) Effects of dispersion on salt enrichment in
coastal aquifers. In: U.S. Geological Survey Water Supply Paper
1613-C, Sea water in coastal aquifers: C70-C84.
4. Kröhn, K.-P. (1991) Simulation von Transportvorgängen im
klüftigen Gestein mit der Methode der Finiten-Elemente. Bericht
Nr. 29/1991, Institut für Strömungsmechanik, Universität Han-
nover, Dissertation.
5. Segol, G. (1994) Classic groundwater simulation: proving and
improving numerical models. PRT Prentice Hall, Englewood
Cliffs, New Jersey.
6. Voss, C.I. and Souza, W.R. (1987) Variable density flow and sol-
ute transport simulation of regional aquifers containing a narrow
freshwater-saltwater transition zone. Water Resources Research
23(10), 1851-1866.
Figure 16.9 Stream functions - fine mesh results.
NSKS oÉÑÉêÉåÅÉë
1. Frind, E.O. (1982) Simulation of long-term transient density-
dependent transport in groundwater. Advances in Water
Resources 5: 73-88.
cbcilt=ö=ONN
NSK=eÉåêóÛë=mêçÄäÉã
ONO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
NT
NT p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
NTKN fåíêçÇìÅíáçå
In this Section we discuss one of the test cases that
was proposed for study by the participants of the inter-
national HYDROCOIN project (Swedish Nuclear
Power Inspectorate 1984) for the verification of
groundwater models. This test case is addressed to
model density variable groundwater flow over a hypo-
thetical salt dome.
Figure 17.1 Definition of the salt dome problem (from
Herbert et al.1).
NTKO pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
The hydraulic head varies linearly on the top of the
The considered test case is an idealization of flow aquifer. All the other sides are impervious for flow. The
over a salt dome, where the geometry is greatly simpli- concentration on the top is taken equal zero on the
fied. The attention is directed to the physics of flow of inflow domain ( 0m ≤ x ≤ 200m ). The middle section of
an non-homogeneous fluid, whose density is depending the base represents the top of the salt dome with rela-
on the salt concentration. The geometry and boundary tive salt concentration to equal unity. All the remaining
conditions used are shown in Figure 17.1. The cross parts of the boundary are closed for diffusive fluxes.
section of the model extends horizontally to 900 m and
vertically to 300 m. The aquifer is considered to be
homogeneous and isotropic.
NTKP dçîÉêåáåÖ=bèì~íáçåë
The governing equations of balancing fluid mass,
salt mass and linear momentum as well as the equation
cbcilt=ö=ONP
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
of state for the solution density are already described in of the corresponding densities. More recently, the salt
detail in Chapters 14 through 16. dome problem has been studied by Younes et al.5.
Herbert et al.1 introduced a slightly different formu- The following table summarizes the model parame-
lation by using mass fractions of fluid and salt instead ters and its values used in the present study.
Table 17.1 Model parameters
h piezometric head L
μ0 fluid dynamic viscosity, where C = C 0 8.9 x 10-4 kg m-1 s-1 M L-1 T-1
ONQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKQ=kìãÉêáÅ~ä=^å~äóëáë
NTKQKN jáñÉÇ= áåíÉêéçä~íáçå= J= ÅçåëáëJ Herbert et al.1 solve the resulting algebraic equa-
íÉåí=îÉäçÅáíáÉë tions by use of direct methods (fast direct frontal
solver). Additionally, a Newton-Raphson scheme is
Herbert et al.1 as well as Voss and Souza4 modify implemented for the iterative solution of the non-linear
the standard Galerkin finite element method by using mass transport equation. The non-linearity arises from
mixed finite element interpolations for the field vari- concentration dependent density variations.
ables - pressure and concentration. In this way they
model concentrations on each element by a polynomial Starting from an initial distribution of zero concen-
of lower degree than that for pressure. In particular, tration this iterative scheme works for high diffusivity
they use linear elements for concentration and qua- situations, such as D = 5 x 10-6 m2s-1. This provides a
dratic elements for pressure. This was done to obtain basic solution. The basic solution is used as an initial
the same order of variation (linear) of the discretized guess for less diffusivity cases. During the iterative
pressure gradients and gravitational terms, too. Other- scheme the diffusivities or dispersion lengths are then
wise, wiggles can arise in regions where the vertical reduced by small steps. These parameter increments
pressure gradient (upward) and the gravitational term must be small enough to obtain satisfactorily conver-
(downward) are nearly balanced. By this technique a gence of the Newton-Raphson scheme
much better point-wise balance of pressure and gravita-
tional forces can be achieved within the finite elements. Kolditz and Diersch3 successfully prove this kind of
Furthermore, Herbert et al.1 computed continuous quasi steady-state solution procedure in connection
velocities by applying a finite-element scheme for the with parameter stepping for highly non-linear hydro-
velocity function instead of substituting the finite ele- thermal convection problems.
ment values for pressure gradient and gravitational
term into the Darcy equation which leads to discontinu- NTKQKP `çãéìí~íáçå~ä=éêçÅÉÇìêÉ
ous velocity fields.
In contrast to Herbert et al.1 we calculate the fully
transient evolution of the convection system. We use
the standard Galerkin FEM in connection with the sec-
ond order AB/TR rule for time stepping with initial
increment of t = 10-5d. This computational procedure
has proved already for the simulation of Elder’s prob-
lem (Chapter 14). Furthermore, the extended Bouss-
inesq approximation for the fluid mass balance and the
Figure 17.2 Finite element mesh A (finite elements
ne = 1920, nodal points np = 2013).
convective formulation of the transport equation is cal-
culated. The finite element mesh we employed is
cbcilt=ö=ONR
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
shown in Figure 17.2. The actual simulation parame- Table 17.2 Simulation parameters (continued)
ters are summarized in Table 17.2.
Parameters FEFLOW
Table 17.2 Simulation parameters
mass transport initials:
Parameters FEFLOW
homogeneous concen- C ( t=0, x, z ) = 0
problem class flow and mass transport tration
ONS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
NTKR oÉëìäíë=~åÇ=aáëÅìëëáçå which has been tackled by several authors and is well
described in the literature (Herbert et al.1, Oldenburg
In our study we calculate different models: diffu- and Pruess2).
sive, dispersive, and diffusive-dispersive ones (Table
17.3). The data are due to the HYDROCOIN test case
Table 17.3 Simulation overview
5 x 10-6 m2 s-1 0m 0m
-7 2 -1
diffusive models 5 x 10 m s 0m 0m
5 x 10-8 m2 s-1 0m 0m
5x 10-8 m2 s-1 20 m 2m
diffusive-dispersive models
1.39 x 10-8 m2 s-1 20 m 2m
dispersive models 0 20 m 2m
In all simulation we detect stratified systems (Fig- Figure 17.3 presents the findings for the steady-
ure 17.3 - Figure 17.10). There is a freshwater region in state salt distribution by Herbert et al.1 and our results
the upper part with higher velocities where flow is for the diffusive models.
driven by the superposed head gradient on the surface.
And there is a brine pool along the bottom with recircu-
lating flow at lower velocities. The outflow of the salt
is focussed on the upper right-hand corner. With
decreasing diffusivities the freshwater zone is dipping
more into the depth of the reservoir or otherwise, the
saltwater rising is more limited.
cbcilt=ö=ONT
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
Figure 17.3 Comparison of different diffusive models: D d = 5 x 10-6, 5 x 10-7m2s-1 (from top to bottom).
Herbert et al.1 begin their computations with molec- overall counterclockwise flow imposed by the head
ular diffusivity of D = 5 x 10-6 m2s-1 which corre- differences along the top. In contrast, for small diffu-
sponds to the largest principal component of the sive fluxes ( D < 10-7 m2s-1) the convective forces are
dispersion tensor occurring within the model region. insufficient to overcome the pressure-driven flow. As a
The largest velocities arise near the top where the flow consequence, the brine is swept forward from left to
is driven purely by the specified head gradient. Starting right with the overall flow.
from the high-diffusivity basic case, they applied
parameter stepping for reducing the diffusivity down to However, the results must be very careful treated by
D = 5 x 10-8 m2s-1. physical interpretation. The considered flow system is
highly sensitive to diffusion. Therefore, especially for
Figure 17.4 presents results for the case with small small physical diffusivities the flow regime may be sig-
diffusivity ( D = 5 x 10-8 m2s-1). Oldenburg and nificantly affected by numerical diffusion. The pre-
Pruess2 compute a quite different salt distribution - they sented findings by the several authors are quite similar
called it ’swept forward’. They argued that for large but a perfect agreement could not be achieved (Figure
diffusive fluxes ( D > 2 x 10-7 m2s-1) the amount of 17.5). We feel that the results strongly depend on the
brine entering rising from the bottom source is suffi- specific numerical schemes.
cient to cause a clockwise recirculation against the
ONU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKR=oÉëìäíë=~åÇ=aáëÅìëëáçå
Figure 17.4 Comparison of salt distributions for the diffusive model: D d = 5 x 10-8 m2s-1.
Figure 17.5 Comparison of salt distributions for the diffusive model: D d = 5 x 10-7 m2s-1.
cbcilt=ö=ONV
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
D d = 5 x 10-8 m2 s-1 187.2 109.8 Figure 17.6 compares the findings by Herbert et al.1
and our results for the diffusive-dispersive models. We
dispersive models β L = 20 m, β T = 2 m 0.75 0.25 observe a quite optimistic agreement.
Figure 17.6 Comparison of different diffusive-dispersive models: β L =20 m, β T =2 m; D d = 5 x 10-8 m2s-1 and 1.39 x
10-8 m2s-1 (from top to bottom).
sive transport dynamics in strongly coupled groundwa-
ter-brine flow systems. In contrast to previous
NTKRKR aáëéÉêëáîÉ=ãçÇÉäë simulations of the HYDROCOIN case they found so-
called swept forward regimes (see above) for small dif-
Oldenburg and Pruess2 recently reported on disper- fusivities (Figure 17.4) and also for the dispersive
model (Figure 17.7).
OOM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKS=qÜÉêãçÜ~äáåÉ=bñíÉåëáçå=çÑ=íÜÉ=p~äí=açãÉ=mêçÄäÉã
Finally, we plot the contours of hydraulic head, Elder problem (see Chapter 14) are derived by use of
streamlines, velocity field, and salt concentrations for thermal Rayleigh numbers,
the dispersive model: β L = 20 m, β T = 2 m (Figure
17.8). Stream functions are computed by the finite ele- KLβΔT
Ra t = ------------------- (17-2)
ment method. Contour plots of streamlines give an Λ
excellent image of the flow system. where
ρ – ρ0 g
q = – K ⎛ ∇h – --------------- -----⎞ (17-4)
⎝ ρ0 g ⎠
NTKS qÜÉêãçÜ~äáåÉ= bñíÉåëáçå
çÑ=íÜÉ=p~äí=açãÉ=mêçÄäÉã we transform the momentum balance Eq. (17-3) to
dimensionless forms by introducing the following
NTKSKN dçîÉêåáåÖ=Éèì~íáçåë change of variables
The dimensionless equations for the thermohaline
cbcilt=ö=OON
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
g
q D = – Ra s ⎛⎝ ∇h D – ( C D – 1 ) -----⎞⎠ (17-6)
g
α
KL ------ ΔC
Cs L
Ra s = ---------------------- = ------ (17-7)
εD βT
Buoyancy Ra s Ra t
Figure 17.8 Hydraulic head, streamlines, velocity field, and ratio N
salinity distribution at t = 100 yr for the diffusive-dispersive Eq. (14-19) Eq. (17-7) Eq. (17-2)
model: β L = 20 m, β T = 2 m; D d = 1.39 x 10-8 m2s-1 (from
top to bottom). 1 150 1114
2 150 557
L x h
q D = ------- q x D = --- h D = ------------------------------
εD L L ( α ⁄ C s )ΔC 3 150 371
(17-5)
C 5 150 223
C D = --------
ΔC
OOO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKS=qÜÉêãçÜ~äáåÉ=bñíÉåëáçå=çÑ=íÜÉ=p~äí=açãÉ=mêçÄäÉã
salinity temperature
Figure 17.9 Salinity and temperature distributions at t = 100 yr for different buoyancy ratios N = 5, 3, 2, 1 (from top to
bottom) as well as velocity fields for the thermohaline salt dome problem.
cbcilt=ö=OOP
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
salinity temperature
Figure 17.10 Salinity and temperature distributions at t = 100 yr for different buoyancy ratios N = 10, 5, 1 (from top to
bottom) as well as velocity fields for the thermohaline salt dome problem. Thermal Rayleigh number is Ra t = 22.
NTKSKO `çåÅäìÇáåÖ=êÉã~êâë
The results presented in Figure 17.10 correspond to
a smaller thermal Rayleigh number of around Rat = In the previous sections we discussed three promi-
22, which is due to a thermal expansion coefficient of nent examples of variable density groundwater sys-
β = 5 x 10-4 K-1 and a geothermal gradient of ΔT ⁄ L = tems: Elder’s problem, Henry’s problem, and now the
0.33 Km-1. salt dome problem. We learn a lot of their physics and
how to simulate numerically this kind of non-linear
problem. There are still numerous questions. We sus-
OOQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NTKT=oÉÑÉêÉåÅÉë
NTKT oÉÑÉêÉåÅÉë
1. Herbert, A.W., Jackson, C.P. and Lever, D.A. (1988) Coupled
groundwater flow and solute transport with fluid density strongly
dependent upon concentration. Water Resources Research 24(10):
1781-1794.
2. Oldenburg, C.M. and Pruess, K. (1995) Dispersive transport
dynamics in a strongly groundwater-brine flow system. Water
Resources Research 31(2): 289-302.
3. Kolditz, O. and Diersch, H.-J. (1993) Quasi steady-state strategy
for numerical simulation of geothermal circulation in hot dry rock
fractures. Int. J. Non-Linear Mechanics 28(4), 467-481.
4. Voss, C.I. and Souza, W.R. (1987) Variable density flow and sol-
ute transport simulation of regional aquifers containing a narrow
freshwater-saltwater transition zone. Water Resources Research
23(10), 1851-1866.
5. Younes, A., Ackerer, P., and Mosé, R. (1999) Modeling variable
density flow and solute transport in porous medium: 1. Re-evalu-
ation of the salt dome flow problem. Transport in Porous Media
35(3), 375-394.
cbcilt=ö=OOR
NTK=p~äí=açãÉ=mêçÄäÉã=Eevaol`lfk=Å~ëÉ=R=äÉîÉä=NF
OOS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
d~êÇåÉêÛë=mêçÄäÉãW=råë~íìê~íÉÇ=pçáä=`çäìãå
NU
solved in two and three dimensions:
NU d~êÇåÉêÛë=mêçÄäÉãW=råë~íìê~íÉÇ=pçáä=`çäìãå
NUKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã
v
An analytical solution for the steady-state capillary
head distribution above the water table in a soil column
(Figure 18.1) has been presented by Gardner2. This z
analytical solution is compared with a steady-state sim-
ulation in FEFLOW.
vadose zone
L g
NUKO dçîÉêåáåÖ=bèì~íáçåë
The governing equations for the unsaturated-satu-
rated flow, contaminant mass and heat transport are water table
cbcilt=ö=OOT
NUK=d~êÇåÉêÛë=mêçÄäÉãW=råë~íìê~íÉÇ=pçáä=`çäìãå
f ∂C q⊗q
s ( ψ )εR d ( C ) ------- + q ⋅ ∇C (18-3) D = ( β L – β T ) ------------- + β T q I (18-5h)
∂t q
f
– ∇ ⋅ [ ( εs ( ψ )D d I + D ) ⋅ ∇C ]
f f f f f
f f Λ = ρ c D + εs ( ψ )λ I (18-5i)
+ [ s ( ψ )εR ( C )ϑ + Q h ]C = s ( ψ )Q C
f f s
Q T ( ψ ) = s ( ψ )εQ T + ( 1 – ε )Q T (18-5j)
f f f s s ∂T f f
[ s ( ψ )ερ c + ( 1 – ε )ρ c ] ------ + ρ c q ⋅ ∇T (18-4)
∂t
f s To solve the nonlinear equations Eq. (18-1) to Eq.
– ∇ ⋅ [ ( Λ + ( 1 – ε )λ I ) ⋅ ∇T ]
(18-4) for h , q , C , and T under unsaturated-saturated
f f
+ ρ c Qh ( T – To ) = QT ( ψ ) conditions constitutive relationships for the saturation-
dependent moisture capacity C ( ψ ) and hydraulic con-
f
ductivity Kr ( s ) have to be specified in form of empiri-
with the definitions and constitutive relationships cal relationships for the capillary pressure head-
f f
f
saturation ψ ( s ) , with its inverse s ( ψ ) , and for the rel-
p f
h = --------- + z = ψ + z (18-5a) ative conductivities Kr ( s ) .
ρo g
S o = εγ + ( 1 – ε )φ (18-5b)
NUKP ^å~äóíáÅ~ä=pçäìíáçå
f
∂s ( ψ ) Assuming the relative conductivity K r ( ψ ) relation-
C ( ψ ) = ---------------- (18-5c)
∂ψ
ship as given by Philip1
f
kρ o g
K = -----------
- (18-5d) ⎧ αψ for ψ < 0
f Kr ( ψ ) = ⎨ e (18-6)
μ0 for ψ ≥ 0
⎩1
f
μ0
f μ = -------------------
- (18-5e) where α is the sorptive number and ψ the capillary
f
μ ( C, T ) pressure head, an analytical solution for the steady-
state capillary head distribution above the water table
f f
ρ = ρo [ 1 + α ( C – Co ) – β ( T – To ) ] (18-5f) (Figure 18.1) has been presented by Gardner2 in the
form
(1 – ε) ⎫
R ( C ) = 1 + ---------------- χ ( C )
ε ⎪ 1 1 –α ( L – z )
⎬ (18-5g) ψ ( z ) = – --- ln ------ ( v + ( K s – v ) e ) (18-7)
(---------------
1 – ε )- ∂----------------------------
[ χ ( C ) ⋅ C -] ⎪ α Ks
Rd ( C ) = 1 +
ε ∂C ⎭
OOU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NUKQ=kìãÉêáÅ~ä=^å~äóëáë=EcbjF
where z is vertical coordinate (positive in the gravita- Table 18.2 FEFLOW parameters (continued)
tional direction g), v is the infiltration/exfiltration rate
(also positive in the gravitational direction g), Ks is the Parameters FEFLOW
saturated conductivity, L is the height of the column, –4
constant infiltration v = 8.64 ⋅ 10 m/d
and ’ln’ represents the natural logarithm.
rate (flux boundary)
Table 18.1 Used parameters
flow materials:
Symbol Value Unit
saturated hydraulic –4
–4 0.001. 10 m/s
v ± 8.64 ⋅ 10 m/d conductivity
–7
Ks 10 m/s Gardner model
parameters
L 1 m
porosity ε = default
residual saturation f
s r = default
NUKQ kìãÉêáÅ~ä=^å~äóëáë=EcbjF
maximum saturation f
The comparison with the analytical solution is per- s s = default
formed by a simulation run using the following param-
eters as summarized in Table 18.2. Table 18.3 Mesh parameters
Parameters FEFLOW
Table 18.2 FEFLOW parameters
mesh
Parameters FEFLOW
depth L=1m
steady state number of nodes in z- 101
direction
flow initials:
cbcilt=ö=OOV
NUK=d~êÇåÉêÛë=mêçÄäÉãW=råë~íìê~íÉÇ=pçáä=`çäìãå
OPM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
`Éäá~= Éí= ~äKÛë= mêçÄäÉãW= råë~íìê~íÉÇ= cäçï= áå= ~
eçãçÖÉåÉçìë=pçáä NV
hydraulic conductivity of 0.922. 10 m/s, boundary
NV `Éäá~=Éí=~äKÛë=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~=eçãçÖÉåÉçìë=pçáä= –4
(18-4)) with the definitions and constitutive relation- ture for the numerical implementation. Under saturated
f
ships of Eq. (18-5a) to Eq. (18-5j). conditions s ≡ 1 the nonlinearities in the parametric
relationships vanish, however, for free-surface condi-
To solve the nonlinear equations Eq. (18-1) to Eq. tions a nonlinear boundary-value problem remains to
(18-4) for h , q , C , and T under unsaturated-saturated be solved.
conditions constitutive relationships for the saturation-
dependent moisture capacity C ( ψ ) and hydraulic con-
f
ductivity K r ( s ) have to be specified in form of empiri- NVKP kìãÉêáÅ~ä=^å~äóëáë=EcbjF
cal relationships for the capillary pressure head-
f f
saturation ψ ( s ) , with its inverse s ( ψ ) , and for the rel- The model has been created in FEFLOW using the fol-
f
ative conductivities Kr ( s ) . Following parametric lowing parameters:
model will be applied1:
Table 19.1 FEFLOW parameters
van Genuchten - Mualem model:
Parameters FEFLOW
⎧ -----------------------------------
1 - for ψ<0
f ⎪ m
transient flow
se = ⎨ [ 1 + ( A ψ )n ] (19-1)
⎪ flow initials:
⎩1 for ψ≥0
initial pressure head ψ 0 = -1.0 m
2
--- ⎧
1 1- m ⎫
---
f 2⎪ f m ⎪ flow boundaries:
Kr = ( se ) ⎨ 1 – 1– ( se ) ⎬ (19-2)
⎪ ⎪ constant head at top of ψ = -0.75 m
⎩ ⎭
column
OPO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
NVKQ=oÉëìäíë
cbcilt=ö=OPP
NVK=`Éäá~=Éí=~äKÛë=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~=eçãçÖÉåÉçìë=pçáä
0 0
-20 -20
-40
Pressure [kPa]
Pressure [kPa]
-40
-60 -60
Celia et al. Δz = 0.50 cm
Present Δz = 2.50 cm
-80 -80 Δz = 16.67 cm
-100 -100
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Depth [m] Depth [m]
Figure 19.2 Pressure profiles at t = 1 day for the dense Figure 19.3 Pressure profiles at t = 1 day for different
grid: Celia et al.’s θ – ψ -based solution3,4 using a mixed- grid spacing Δz computed by the present predictor-cor-
form Picard iteration versus the present predictor-correc- rector technique with a h-based form of the Richards
tor h-based solution using an adaptive time stepping. equation.
NVKR oÉÑÉêÉåÅÉë
1. Bear, J. and Bachmat, Y., Introduction to modeling of transport
phenomena in porous media. Kluwer Academic Publ., Dordrecht,
1991.
2. Diersch, H.-J.G. and Perrochet, P., On the primary variable
switching technique for simulating unsaturated-saturated flows,
Advances in Water Resources 23 (1999) 3, 271-301.
3. Celia, M.A., Bouloutas, E.T. and Zarba, R.L., A general mass-
conservative numerical solution for the unsaturated flow equa-
tion. Water Resour. Res., 26 (1990) 7, 1483-1496.
4. Segol, G., Classic groundwater simulations: proving and improv-
ing numerical models. PTR Prentice Hall, Englewood Cliffs, New
Jersey, 1994.
OPQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å
fåÜçãçÖÉåÉçìë=pçáä OM
erties change gradually with depth. The initial condi-
OM s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å=fåÜçãçÖÉåÉçìë=pçáä
OMKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã tions for the flow are given by ψ 0 = -3.5 m. A constant
h
flux is at the surface q n = -0.25 m/d at t ≤ 1 day (infil-
h
The objective of tration) and q n = 0.005 m/d at t > 1 day (evaporation).
0
this chapter is to Clay Loam (1) At the bottom a drainage gradient-type BC of = 4 m/d
describe the pro- 25
is imposed, accordingly, the bottom boundary can
cesses of true unsat- Loamy Sand (3) freely drain. The parameter in the constitutive relations
(4)
urated flows in (5) (van Genuchten-Mualem model) can be found in4.
Changing to
layered soils to show (6)
the capabilities and Sand (7)
OMKO dçîÉêåáåÖ=bèì~íáçåë
Depth (cm)
cbcilt=ö=OPR
OMK=s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å=fåÜçãçÖÉåÉçìë=pçáä
f ∂h ⎧ -----------------------------------
1
[ S o ⋅ s ( ψ ) + ε ⋅ C ( ψ ) ] ------ + ∇ ⋅ q = Q h (20-1) - for ψ<0
∂t f ⎪ m
se = ⎨ [ 1 + ( A ψ )n ] (20-5)
⎪
⎛ ρ –
f f
ρo⎞ ⎩1 for ψ≥0
f
q = – K r ( s )Kf μ ⎜ ∇h + ---------------
f
- e⎟ (20-2)
⎝ ρo ⎠ 2
1⎧ 1 m⎫
--- ----
∂C f 2⎪ f m ⎪
f
s ( ψ )εR d ( C ) ------- + q ⋅ ∇C (20-3) Kr = ( se ) ⎨ 1 – 1– ( se ) ⎬ (20-6)
∂t ⎪ ⎪
f ⎩ ⎭
– ∇ ⋅ [ ( εs ( ψ )D d I + D ) ⋅ ∇C ]
f f
+ [ s ( ψ )εR ( C )ϑ + Q h ]C = s ( ψ )Q C with the effective saturation of fluid
f f f f
f f f s s ∂T f f s – sr θ –θ
[ s ( ψ )ερ c + ( 1 – ε )ρ c ] ------ + ρ c q ⋅ ∇T (20-4) f
- = ---------------r-
s e = ------------- (20-7)
∂t f f f f
f s ss – sr θs – θr
– ∇ ⋅ [ ( Λ + ( 1 – ε )λ I ) ⋅ ∇T ]
f f
+ ρ c Qh ( T – To ) = QT ( ψ )
OMKP kìãÉêáÅ~ä=^å~äóëáë
with the definitions and constitutive relationships of
Eq. (18-5a) to Eq. (18-5j). The parameter settings of the van Genuchten model
are summarized in the following tables.
To solve the nonlinear equations Eq. (20-1) to Eq.
(20-4) for h , q , C , and T under unsaturated-saturated Table 20.1 Initial and boundary conditions used by
conditions constitutive relationships for the saturation- van Genuchten1
dependent moisture capacity C ( ψ ) and hydraulic con-
f
ductivity K r ( s ) have to be specified in form of empiri- Parameters Value
cal relationships for the capillary pressure head-
f f
saturation ψ ( s ) , with its inverse s ( ψ ) , and for the rel- transient flow
f
ative conductivities Kr ( s ) . Following parametric
model will be preferred3: flow initials:
OPS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OMKP=kìãÉêáÅ~ä=^å~äóëáë
Table 20.1 Initial and boundary conditions used by The conversion of van Genuchten’s example to a
van Genuchten1 (continued) FEFLOW model requires some changes in the units of
the parameters. The finite element model is 1.70 m
Parameters Value long and 0.01 m wide (Figure 20.2). The grid spacing
h is 0.01 m.
constant flux at sur- q n = 0.005 m/d
face at t > 1 day
h
drainage at bottom q n = 4 m/d
cbcilt=ö=OPT
OMK=s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å=fåÜçãçÖÉåÉçìë=pçáä
OPU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OMKQ=oÉëìäíë
OMKR oÉÑÉêÉåÅÉë
Table 20.5 FEM parameters
1. van Genuchten, M.Th., Mass transport in saturated-unsaturated
media: One-dimensional solutions. Research Rep. No. 87-WR-
Parameter FEFLOW 11, Water Resources Program, Princeton Univ., 1978.
Figure 20.2 2. Diersch, H.-J.G., FEFLOW-Physical basis of modeling. WASY
mesh
Institute for Water Resources Planning and Systems Research
Ltd., Berlin, 1997.
width, depth B = 0.01 m, L = 1.70 m
3. Bear, J. and Bachmat, Y., Introduction to modeling of transport
grid spacing: phenomena in porous media. Kluwer Academic Publ., Dordrecht,
1991.
Δx = 0.01m, Δy = 0.01m 4. Segol, G., Classic groundwater simulations: proving and improv-
time stepping regime AB/TR time integration ing numerical models. PTR Prentice Hall, Englewood Cliffs, New
Jersey, 1994.
scheme
OMKQ oÉëìäíë
A comparison between the present solutions and the
results obtained by van Genuchten1 who used a Hermi-
tian-finite element approach is exhibited in Figure 20.3
and Figure 20.4 displays the simulated moisture-con-
tent profiles during the infiltration period at t ≤ 1 day.
The moisture-content histories during the redistribution
phase at t > 1 day of the soil column are seen in
Figure 20.4. As shown the agreement of the results is
nearly perfect.
cbcilt=ö=OPV
OMK=s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å=fåÜçãçÖÉåÉçìë=pçáä
0.00
t=0 .05
.15
0.25 θfr
.25
0.50
.35
.50
Depth [m]
0.75 θfs
1.00 .65
1.25 .80
.95
1.50
OQM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OMKR=oÉÑÉêÉåÅÉë
0.00
0.25 θfr
1.0
1.5
0.50 8.0 4.0
Depth [m]
0.75 θfs
1.00
1.5
1.0
1.25
4.0
1.50
8.0
cbcilt=ö=OQN
OMK=s~å=dÉåìÅÜíÉå=mêçÄäÉãW=råë~íìê~íÉÇ=cäçï=áå=~å=fåÜçãçÖÉåÉçìë=pçáä
OQO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
s~ìÅÜ~ìÇ=Éí=~äKÛë=mêçÄäÉãW=aê~áå~ÖÉ=bñéÉêáãÉåí
ON
the water table descends continuously up to reaching
ON s~ìÅÜ~ìÇ=Éí=~äKÛë=mêçÄäÉãW=aê~áå~ÖÉ=bñéÉêáãÉåí
h
Vauchaud et al.2 reported experimental results qn = –v
which referred to a ditch-drained soil problem. Their unsaturated zone
results are useful for proving and comparing numerical h
qn = 0
schemes applied to a typical drainage problem as initial water table
f ∂C
s ( ψ )εR d ( C ) ------- + q ⋅ ∇C (21-3)
∂t
f
– ∇ ⋅ [ ( εs ( ψ )D d I + D ) ⋅ ∇C ]
q = -0.384 m/d
f f
+ [ s ( ψ )εR ( C )ϑ + Q h ]C = s ( ψ )Q C
h =1.45 m
f f f s s ∂T f f
[ s ( ψ )ερ c + ( 1 – ε )ρ c ] ------ + ρ c q ⋅ ∇T (21-4)
∂t
f s
– ∇ ⋅ [ ( Λ + ( 1 – ε )λ I ) ⋅ ∇T ]
h =0.75 m
f f
+ ρ c Qh ( T – To ) = QT ( ψ )
To solve the nonlinear equations (21-1) to (21-4) for Figure 21.2 FEFLOW finite element grid with boundary
h , q , C , and T under unsaturated-saturated conditions conditions.
constitutive relationships for the saturation-dependent
moisture capacity C ( ψ ) and hydraulic conductivity
OQQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ONKP=kìãÉêáÅ~ä=^å~äóëáë
Parameters FEFLOW
Table 21.3 FEM parameters
transient flow
Parameter FEFLOW
flow initials:
mesh Figure 21.1
initial pressure head ψ 0 = 1.45 [m]
width, depth B = 3 m, L = 2 m
flow boundaries:
grid spacing:
h
inflow on top qn = -0.384 [m/d] Δx = 0.15 m,
hydraulic head at ditch h=z Δy = 0.0583 m - 0.06875 m
from 1.45 m to 0.75 m
time stepping regime FE/BE time integration
hydraulic head at ditch h = 0.75 [m] scheme
from 0.75 m to 0 m
initial time step length 1e-3 [d]
Parameters FEFLOW
ONKQ oÉëìäíë
flow materials:
Numerical results are given by Gureghian3. How-
conductivity 1.11[10e-4 m/s] ever, his solutions are based on a non-zero infiltration
total saturation
f
ss = 1 rate v . Accordingly, the water table descends continu-
ously up to reaching the water level h w of the ditch.
f
residual saturation sr = 0 Figure 21.3 compare the present numerical results with
porosity ε = 0.3 Vauchaud et al.’s experimental data. As seen the agree-
ment is quite well. A comparison of the hydraulic head
cbcilt=ö=OQR
ONK=s~ìÅÜ~ìÇ=Éí=~äKÛë=mêçÄäÉãW=aê~áå~ÖÉ=bñéÉêáãÉåí
contours, the water table and capillary fringe at time of head contours which obviously results from different
t = 1 hour is presented in Figure 21.4 between the description of the infiltration boundary condition.
present solutions and Gureghian’s results. The agree-
ment is satisfactory. Differences appear for the upper
t=0
0.1
1.0
5.0
Figure 21.3 Descending water table of the drainage experiment: simulated free-surface locations (left) and water tables
measured by Vauchaud et al.2 (right), times in hours.
Figure 21.4 Hydraulic head contour and water table location at t = 1 hour: present results (left) and Gureghian’s
solutions3,5 (right).
OQS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ONKR=oÉÑÉêÉåÅÉë
ONKR oÉÑÉêÉåÅÉë
1. Bear, J. and Bachmat, Y., Introduction to modeling of transport
phenomena in porous media. Kluwer Academic Publ., Dordrecht,
1991.
2. Vauchaud, G., Vauclin, M. and Khanji, D., Étude expérimentale
des transferts bidimensionels dans la zone non-saturée - Applica-
tion a l’étude du drainage d’une nappe a surface libre. Houille
Blanche 1 (1973), 65-74.
3. Gureghian, A.B., A two-dimensional finite element solution
scheme for the saturated-unsaturated flow with applications to
flow through ditch-drained soils. J. Hydrol. 50 (1981), 333-353.
4. Nützmann, G., Eine Galerkin-finite-element-Methode zur Simu-
lation instationärer zweidimensionaler ungesättigter und gesät-
tigter Wasserströmungen im Boden. Acta Hydrophysica XXVIII
(1983) 1/2, 37-107.
5. Nguyen, H., WHAT-FEM - A two-dimensional finite element
model for simulating water flow in variably saturated soils. Tech-
nical Note, No. I.96.125, JRC, Ispra Institute, July, 1996.
cbcilt=ö=OQT
ONK=s~ìÅÜ~ìÇ=Éí=~äKÛë=mêçÄäÉãW=aê~áå~ÖÉ=bñéÉêáãÉåí
OQU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
OO
recently, Webb4 presented improved results obtained by
OO tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
OOKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã the TOUGH2 simulator which compare now well with
Ross’ analytical solution.
A capillary barrier arises in unsaturated flow condi-
tions at the contact of a fine layer of soil overlying a infiltrati
on
coarse layer of soil1,2. If such a contact is tilted, infil-
trating moisture in the fine layer will be diverted and Φ
flow down the contact as capillary diversion g
(Figure 22.1). In practical applications, a capillary bar-
fine
rier can be built by placing a fine layer (e.g., fine sand)
over a coarse layer (e.g., gravel) along a sloping con-
tact. It is used to protect underground regions from
wetting by downward infiltration (protective cover for coarse
waste disposals)3.
Ross1 developed an analytical relationship to calcu- Figure 22.1 Schematic of a capillary barrier (from Webb4).
late the steady-state diversion length of a tilted fine
over coarse capillary barrier with constant infiltration.
Oldenburg and Pruess2 presented a first numerical The present paper aims at a comparison of Ross’
study of a 2D tilted fine over coarse capillary barrier by capillary barrier diversion formula with numerical sim-
using the finite difference TOUGH2 code5. To find ulations by using the finite-element simulator
qualitatively reasonable results Oldenburg and Pruess FEFLOW. The numerical investigations refer to a cap-
had used an upstream weighting scheme. From the illary barrier problem using van Genuchten characteris-
quantitative point of view the agreement of their results tic curves as recently studied by Webb4. The
with Roos’ predictions became generally poor. More simulations are performed for both 2D and axisymmet-
cbcilt=ö=OQV
OOK=tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
initial saturation sr
As summarized in Table 22.1 the assumed infiltra-
tion rate is 0.0048 m/d, and the initial saturation is infiltration rate, m/d 0.0048
assumed equal to the residual values for each layer. In
the lateral direction the model domain is 100 m in the Table 22.2 Material properties
length. At the bottom and the right side boundaries a
potential boundary condition is prescribed to realize a Upper Lower
freely drain whenever water reaches these parts of the Parameter layer layer
domain. (fine) (coarse)
ORM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OOKP=kìãÉêáÅ~ä=^å~äóëáë
initial saturation sr
boundary conditions:
OOKPKO ^ééäáÉÇ=cbcilt=jçÇÉä
infiltration rate, m/d 0.0048
FEFLOW solves the governing Richards equation
by the finite-element method. A predictor-corrector
technique combined with an error-controlled automatic
time stepping is used to tackle the strongly nonlinear Table 22.4 FEFLOW material properties
equations. For the present problem the second order
Upper Lower
accurate Adams-Bashforth/trapezoid rule (AB/TR)
Parameter layer layer
scheme is preferred. For the error criterion a relative
–3 (fine) (coarse)
maximum norm of 10 is applied. FEFLOW uses a
mass-conservative solution procedure based on a chord porosity 0.39 0.42
slope approximation for the moisture capacity term and
a mass lumping scheme. The predictor-corrector tech- saturated conductivity, 10e-4 m/s 2.1 1000
nique has proven very powerful also for the present residual saturation 0.3949 0.0286
problem class of unsaturated-saturated flow6. Unlike
the TOUGH2 code, no upstream weighting schemes maximum saturation 1. 1.
become necessary for the present unsaturated problem van Genuchten exponent 5.74 2.19
if using FEFLOW. For the 2D and axisymmetric prob- (0.826) (0.543)
lem both quadrilateral and triangular meshes of linear
finite elements are performed. Mesh refinements are van Genuchten parameter, 1/m 3.9 490.
used along the fine-coarse material interface and at the
bottom of the coarse layer. The used meshes are shown
in Figure 22.2.
cbcilt=ö=ORN
OOK=tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
a)
Table 22.5 FEM parameter
1m
Parameter FEFLOW
5m
triangular mesh 2944 bi-linear elements, 1551
nodes
OOKQ oÉëìäíë
The simulations have been considered as steady-
state if an equilibrium of the flow budget has been
achieved. Practically, it appears after about 100 days.
For the 2D cross-sectional problem Figure 22.3 exhib-
ORO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OOKQ=oÉëìäíë
its the computed streamline pattern. It clearly reveals lary barrier schemes.
the capillary diversion effect in the form of a moisture
movement downdip above the contact of the fine and A comparison of the computed diversion for the 2D
coarse layer. The diversion is maintained up to a certain capillary barrier with Ross’ analytical formula and
distance, called the diversion length, from where an TOUGH2 numerical results presented recently by
amount of water equal to the infiltration flow enters the Webb4 can be expressed as a function of leakage/infil-
coarse layer. The corresponding distribution of the sat- tration ratio. The theoretical value of the diversion
uration can be seen in Figure 22.4. length from Ross’ formula for the presented van Genu-
chten parameter is given by 33.2 m (cf. Webb4). As
For the axisymmetric problem the length of diver- shown in Figure 22.6 there is both a good qualitative
sion increases compared with 2D as displayed in Fig- and quantitative agreement between the analytical
ure 22.5 for the computed streamlines. Up to now, the results and the numerical simulations performed by
axisymmetric problem was not considered by other TOUGH2 and FEFLOW.
authors. Ross’ formula is only applicable for 2D capil-
cbcilt=ö=ORP
OOK=tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
ORQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OOKR=kçí~íáçå
medium is saturated).
1.5 se 1 effective saturation.
sr 1 residual saturation.
ss 1 maximum saturation.
–1
α L van Genuchten curve-fitting
parameter.
Leakage/infiltration
1.0
ε 1 porosity ( 0 < ε ≤ 1 ).
θ 1 s ⋅ ε , volumetric moisture content
( 0 < θ ≤ ε , θ = ε if medium is
saturated).
0.5 Ross' formula
present results θr 1 residual volumetric moisture
Webb's results content.
θs 1 saturated volumetric moisture
content.
0.0 Φ angle of the fine-coarse interface
0 10 20 30 40 50 60 70 80 90 100
with respect to the horizontal.
Distance [m] ψ L pressure head ( ψ > 0 saturated
Figure 22.6 Computed barrier diversion in form of the medium, ψ≤0 unsaturated
leakage/infiltration ratio at the lower coarse layer com- medium).
pared with Ross’ analytical formula and Webb’s numerical
results for the 2D capillary barrier problem.
OOKS oÉÑÉêÉåÅÉë
OOKR kçí~íáçå 1. Ross, B., The diversion capacity of capillary barriers. Water
Resour. Res. 26 (1990) 10, 2625-2629.
–2 2. Oldenburg, C.M. and Pruess, K., On numerical modeling of capil-
g LT gravitational acceleration. lary barriers. Water Resour. Res. 29 (1993) 4, 1045-1056.
Kr 1 relative hydraulic conductivity 3. Selker, J., Design of interface shape for protective capillary barri-
( 0 < K r ≤ 1 , K r = 1 if saturated at ers. Water Resour. Res. 33 (1997) 1, 259-260.
s = 1 ). 4. Webb, S.W., Generalization of Ross’ tilted capillary barrier diver-
–1 sion formula for different two-phase characteristic curves. Water
Ks LT saturated hydraulic conductivity.
Resour. Res. 33 (1997) 8, 1855-1859.
m 1 1 – 1 ⁄ n curve-fitting parameter
5. Pruess, K., TOUGH2 - A general-purpose numerical simulator
(Mualem assumption). for multiphase fluid and heat flow. LBL-29400, Lawrence Berke-
n 1 n > 1 , van Genuchten curve-fitting ley Lab., Berkeley, Calif., May 1991.
exponent (pore size distribution 6. Diersch, H.-J.G., Treatment of free surfaces in 2D and 3D ground-
index). water modeling. Math. Geologie Vol. 2, 17-43, 1998.
s 1 saturation ( 0 < s ≤ 1 , s = 1 if
cbcilt=ö=ORR
OOK=tÉÄÄÛë=`~éáää~êó=_~êêáÉê=mêçÄäÉã
ORS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãå
OP
OP eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãå
⎧ ------------------------------
1 - for ψ<0
OPKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã ⎪
s e = ⎨ [ 1 + αψ n ]
m
(23-1)
⎪
⎩1 for ψ≥0
Modeling of hysteretic flow processes in unsatur-
ated porous media requires knowledge of the main wet-
1 1 m 2
ting and drying retention curves plus the hydraulic ---
2 ⎛ ----⎞
m
conductivity-water saturation relation of the medium. Kr = se 1 – ⎜ 1 – se ⎟ (23-2)
⎝ ⎠
The importance of hysteresis has been shown in labora-
tory studies. Lenhard et al.3 conducted an air-water
with the effective saturation
flow experiment in a 72-cm vertical soil column, where
the water table is fluctuated. This column experiment s – sr θ – θr
having been previously simulated by Clausnitzer1 using s e = -------------- = ---------------- . (23-3)
ss – sr θs – θr
a one-dimensional finite element method, the corre-
sponding FEFLOW solution can be compared with
independently obtained numerical results for both hys- Any hysteresis loop is confined by the main drying and
wetting retention curves which can be expressed
teretic and nonhysteretic conditions.
according to (23-1). The s-ψ retention relation s ( ψ ) is
d w
denoted by s ( ψ ) for the main drying curve and s ( ψ )
for the main wetting curve. The main hysteresis loop is
OPKO dçîÉêåáåÖ=bèì~íáçåë then described by the parameter vector
d d d d d
( s s , s r , α , n , m ) for the main drying curve and
The relations of saturation vs pressure head (s-ψ)
w w w w w
and hydraulic conductivity vs saturation (K-s) are ( s s , s r , α , n , m ) for the main wetting curve. The
described by the van Genuchten parametric model hysteresis loop is closed at residual and maximum
(symbols are summarized under Notation): d w d w
saturation, i.e., s r = s r = s r and s s = s s = s s .
cbcilt=ö=ORT
OPK=eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãå
OPKP kìãÉêáÅ~ä=^å~äóëáë
For the prediction of the scanning curves in the
s ( ψ ) relation Scott’s model4 is used, in which the dry-
OPKPKN iÉå~êÇ=Éí=~äKÛë=mêçÄäÉã
ing scanning curves are scaled from the main drying
curve and wetting scanning curves from the main wet-
A soil column with a length of 0.72 m is considered
ting curves so as to pass through a reversal point
The column is filled by a homogeneous sandy material
(Figure 23.1). If the reversal is from wetting to drying,
with a saturated conductivity of 1.19 m/h and porosity
the main drying curve is scaled by changing the maxi-
of 0.36. The complete set of material parameters is
mum saturation value for that curve such that the rever-
given in Table 23.1.
sal point falls on the resulting curve. Analogously, if
the reversal is in the opposite direction, the main wet- Table 23.1 Material properties
ting curve is scaled by changing the minimum (i.e.,
residual) saturation value. The scanning curves thus Parameter Value
computed are maintained for all nodes.
porosity ε 0.36
–4
saturated conductivity K s , m/s 3.3056 ⋅ 10
d w
100 residual saturation s r = s r = s r 0.17
drying curve
d d
10-1 (m = 1 – 1 ⁄ n ) (0.8095)
w
van Genuchten exponent n of 5.25
main wetting curve wetting curve
w w
(m = 1 – 1 ⁄ n ) (0.8095)
scanning curves d
van Genuchten parameter α of 4.2
-2 drying curve, 1/m
10
w
van Genuchten parameter α of 8.4
wetting curve, 1/m
0.0 0.2 0.4 0.6 0.8 1.0
saturation, s [1]
Initially, most of the column is fully saturated, with the
Figure 23.1 The main characteristic curves for wetting, water table positioned at 0.695 m (25 mm below the
drying, and one set of scanning curves in a hysteresis loop. top of the column). At the bottom of the soil column
the pressure head ψ is varied in time. Its boundary
ORU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OPKP=kìãÉêáÅ~ä=^å~äóëáë
295 0.420
345 0.170
OPKPKO ^ééäáÉÇ=cbcilt=jçÇÉä
395 0.170
Using FEFLOW the column is discretized by 200
505 0.720
quadrilateral bi-linear finite elements with an element
600 0.720 width of Δz = 3.6 mm in the vertical direction. A vari-
a)
able time stepping technique is applied that is based on
changes of pressure head between the
the FE/BE predictor corrector scheme with an initial
time stages are linear. –5 –4
time step of 10 d and an RMS error criterion of 10 .
Results will be compared for five observation points The standard form of the Richards equation with a
located at (P1) 0.69 m, (P2) 0.59 m, (P3) 0.49 m, (P4) Picard iteration strategy is used. The simulation control
0.39 m, and (P5) 0.29 m, measured from the soil bot- parameters are summarized in Table 23.4.
tom. The geometric, initial, and boundary conditions Table 23.4 FEM parameters
are summarized in Table 23.3.
Parameter FEFLOW
width Δz = 3.6 mm
cbcilt=ö=ORV
OPK=eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãå
Parameter FEFLOW
0.3
time stepping regime FE/BE time integration P5
scheme, automatic time steps
P2
initial time step length –5
10 [d]
0.1
final time 0.416667 [d], (10 [h]) P1
Clausnitzer
FEFLOW
0.0
0 2 4 6 8 10
time [h]
OPKQ oÉëìäíë Figure 23.2 History of water content θ simulated at dif-
ferent vertical locations P1 through P5 in the column for
A comparison of the FEFLOW results with the hysteretic conditions: comparison between FEFLOW and
solutions obtained by Clausnitzer1 gives a good agree- Clausnitzer’s1 results.
ment as revealed in Figure 23.2. It is noted that Claus- 0.4
Clausnitzer
FEFLOW
0.0
0 2 4 6 8 10
time [h]
OSM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OPKR=kçí~íáçå
cbcilt=ö=OSN
OPK=eóëíÉêÉíáÅ=cäçï=áå=~å=råë~íìê~íÉÇ=pçáä=`çäìãå
OSO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
tÉåÇä~åÇ=~åÇ=eáããÉäëÄ~ÅÜÛë=bñéÉêáãÉåíW=pçäìíÉ
qê~åëéçêí=áå=~=qÜêÉÉJÇáãÉåëáçå~ä=cê~ÅíìêÉJj~íêáñ
póëíÉã
OQ
The fracture with a mean aperture of > 350 μm divides
OQ tÉåÇä~åÇ=~åÇ=eáããÉäëÄ~ÅÜÛë=bñéÉêáãÉåíW=pçäìíÉ=qê~åëéçêí=áå=~=qÜêÉÉJÇáãÉåëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉã
OQKN pí~íÉãÉåí=çÑ=íÜÉ=mêçÄäÉã the block into two parts. The geometric details of the
fracture plane are shown in Figure 24.2.
Wendland and Himmelsbach3 conducted laboratory 35 mm 90 mm
experiments and numerical computations of solute
transport in a fractured sandstone block. The sandstone
40 mm
outlet
block has a length of 24 cm, a width of 21 cm and a
100 mm
height of 24 cm (Figure 24.1).
outlet sealed
35 mm
240 mm
40 mm
areas
sealed
areas
240 mm
ure
fract
injection
100 mm
210 mm
120
mm
injection
420
mm 120
mm 210
mm Figure 24.2 Geometry of the fracture plane.
Figure 24.1 Schematic representation of the fractured Water is pumped from below flowing upwards at a con-
sandstone block. stant rate of Q = 4.57 ml/h through the fracture plane.
cbcilt=ö=OSP
OQK= tÉåÇä~åÇ= ~åÇ= eáããÉäëÄ~ÅÜÛë= bñéÉêáãÉåíW= pçäìíÉ= qê~åëéçêí= áå= ~= qÜêÉÉJÇáãÉåJ
ëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉã
A multi-tracer experiment with pyranine and cadmium ⊗ q-
⎧ ( εD d + β T q )I + ( β L – β T ) q------------ porous matrix
as solutes was performed. The tracer is injected in the ⎪ q
D = ⎨ (24-5)
fracture at the bottom and the tracer breakthrough is ⊗ q-
⎪ D I + β q------------ fracture
observed at the outlet on the top of the fracture (Figures ⎩ d L q
24.1 and 24.2). and
⎧ε porous matrix
OQKO dçîÉêåáåÖ=bèì~íáçåë ε = ⎨
⎩1 fracture
(24-6)
∇⋅q = Q (24-1)
OQKP kìãÉêáÅ~ä=^å~äóëáë
q = – K∇h (24-2)
In the tracer experiment the injection of the solutes
with
is considered as a pulse directly into the fracture. The
injection of the total tracer mass of 32.2 μg lasted less
⎧K Darcy law in the porous matrix
⎪ than 1 min. For modeling purposes, Wendland and
K = ⎨ b 2 ρg (24-3) Himmelsbach3 smoothed the pulse injection over a
⎪ ------------ I Hagen-Poiseuille law in the fracture
⎩ 12μ time interval of 6 min, which is still small relative to
the duration of the tracer experiment.
The solute transport of a conservative tracer in the
porous matrix and the fracture is governed by the Geometric, initial and boundary conditions for the
advection-dispersion equation (solute mass balance) in present simulation are summarized in Table 24.1.
the following form: Material parameters are listed in Table 24.2.
∂C
ε ------- + q ⋅ ∇C – ∇ ⋅ ( D ⋅ ∇C ) = Q c (24-4)
∂t
OSQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OQKP=kìãÉêáÅ~ä=^å~äóëáë
cbcilt=ö=OSR
OQK= tÉåÇä~åÇ= ~åÇ= eáããÉäëÄ~ÅÜÛë= bñéÉêáãÉåíW= pçäìíÉ= qê~åëéçêí= áå= ~= qÜêÉÉJÇáãÉåJ
ëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉã
further nodes parallel to the vertical fracture are located three-dimensional hexahedral elements for the porous
at a constant horizontal distance of 1 cm. The resulting matrix and 1904 two-dimensional bi-linear fracture
finite-element mesh of the entire sandstone block is elements. In order to account for the sealed areas in the
shown in Figure 24.3. Note that in the FEFLOW simu- fracture plane (Figure 24.2), the corresponding two-
lations only the symmetric half of the domain is con- dimensional elements of the fracture were deleted from
sidered. This leads to a half-mesh consisting of 20160 the mesh.
a) b)
fracture
fracture
Figure 24.3 Finite-element mesh of the sandstone block with a vertical fracture: a) view of the entire block, b)
zoomed mesh at the fracture.
OSS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OQKQ=oÉëìäíë
Figure 24.4 Computed pathlines in the fracture and head Figure 24.5 FEFLOW results of solute distribution within
distributions in the contacted sandstone (half-space view). the fracture at final simulation time t = 660 min.
Figure 24.6 Wendland and Himmelsbach’s3 results of solute distribution within the
fracture at final simulation time t = 660 min.
cbcilt=ö=OST
OQK= tÉåÇä~åÇ= ~åÇ= eáããÉäëÄ~ÅÜÛë= bñéÉêáãÉåíW= pçäìíÉ= qê~åëéçêí= áå= ~= qÜêÉÉJÇáãÉåJ
ëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉã
A comparison of the FEFLOW results with solution 1.75
concentration [mg/l]
agreement (the same concentration levels are used both 1.00
ute diffusion into the matrix and its accurate numerical Figure 24.7 Breakthrough curve at the outlet: FEFLOW
representation in the three-dimensional fracture-matrix results compared to the measurements3.
system is of high importance. The better agreement of
the FEFLOW results can result from the more refined
spatial resolution.
OSU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
OQKR=kçí~íáçå
OQKR kçí~íáçå
b L aperture of fracture.
–3
C ML solute concentration.
2 –1
D L T tensor of hydrodynamic dispersion.
2 –1
Dd L T molecular diffusion.
–2
g LT gravitational acceleration.
h L hydraulic head.
I 1 unit (identity) tensor.
–1
K LT tensor of hydraulic conductivity.
–1
K LT generalized conductivity.
–1
Q T fluid mass sink(-)/source(+) term.
–3 –1
Qc ML T solute mass sink(-)/source(+) term.
–1
q LT specific flux.
–1
v LT fluid velocity vector.
β L, β T L longitudinal and transverse disper-
sivity, respectively.
ε 1 porosity of porous matrix.
ε 1 generalized void space.
–3
ρ ML fluid density.
–1 –1
μ ML T dynamic viscosity of fluid.
OQKS oÉÑÉêÉåÅÉë
1. Diersch, H.-J.G. (2002) Discrete feature modeling of flow, mass
and heat transport processes by using FEFLOW. White Papers -
Vol. I, Chapter 9, WASY Ltd., 147-190.
2. Diersch, H.-J.G. (2002) The Petrov-Galerkin least square method
(PGLS). White Papers - Vol. I, Chapter 13, WASY Ltd., 219-262.
3. Wendland, E. and Himmelsbach, T., Transport simulation with
stochastic aperture for a single fracture - comparison with a labo-
ratory experiment. Advances in Water Resources 25 (2002) 1, 19-
32.
cbcilt=ö=OSV
OQK= tÉåÇä~åÇ= ~åÇ= eáããÉäëÄ~ÅÜÛë= bñéÉêáãÉåíW= pçäìíÉ= qê~åëéçêí= áå= ~= qÜêÉÉJÇáãÉåJ
ëáçå~ä=cê~ÅíìêÉJj~íêáñ=póëíÉã
OTM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
qÜêÉÉJÇáãÉåëáçå~ä= ^åáëçíêçéó= ~åÇ= m~ííÉêåë= çÑ
dêçìåÇï~íÉê=tÜáêäë OR
the center of the box is 150 m long, 20 m wide, and 20
OR qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
ORKN pí~íÉãÉåí=çÑ=mêçÄäÉãë m thick. Inside the anisotropic block, the major princi-
pal directions of the horizontal hydraulic conductivity
Three-dimensional anisotropy in layered aquifers tensor are orthogonal. The general flow direction is
can lead to somewhat unusual flow patterns, even in lengthwise (straight north) and makes an angle of 45°
steady-state flow situations. This was firstly discovered with either of these directions (see Figure 25.1).
and reported by Hemker et al.4, who termed this type of
flow pattern as groundwater whirls. The cause and 25 m
N
appearance of groundwater whirls has been thoroughly 25 m
150 m
70 m 20 m
Hemker et al.4 considered a box-shaped, two-layer,
general flow direction
20 m
homogeneous aquifer. The model box is 200 m long, anisotropy in top layer
70 m wide, and 20 m thick, and consists of only two anisotropy in bottom layer
layers, each 10 m thick. An anisotropic block located at Figure 25.1 The two-layer anisotropic model prob-
lem, example 1 (modified from4).
cbcilt=ö=OTN
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
300 m
K1
K2
α
layer 1
Kz
18 m
layer 9
100 m 100 m 100 m
Figure 25.2 A stratified confined aquifer with a laterally heterogeneous anisotropic central zone (taken from5).
OTO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKO=kìãÉêáÅ~ä=^å~äóëáë
The major and minor principal values of the hori- distributed anisotropy directions (α-angle, see
zontal hydraulic conductivity tensor are 10 m/d and 5 Figure 25.2) were chosen (-45°, -35°, -25°, …, 45°)
m/d in the entire model. The vertical hydraulic conduc- and for each layer these ten values were randomly
tivity is 1 m/d in all layers. The general flow direction assigned to the cells. The resulting distribution is given
is in the direction of the strips (straight north). In the in Table 25.1.
presented model the major principal direction of the
horizontal hydraulic conductivity tensor is also chosen The model size is 300 by 300 m by 18 m. The east
straight north in the two large side blocks, while it var- and west sides are no-flow boundaries, while the south
ies between -45° (N45W, northwest) and 45° (N45E, and north sides are open boundaries with fixed potenti-
northeast) in the 90 cells of the central zone. To obtain ometric heads that differ by 0.3 m in all layers5.
a two-dimensional spatial distribution, ten uniformly
Table 25.1 Principal directions of horizontal anisotropy in all 9 by 12 cells of the model (angle α in °)
ORKO kìãÉêáÅ~ä=^å~äóëáë
For the present FEFLOW simulations we prefer a
ORKOKN bñ~ãéäÉ=N refined 3D mesh of hexahedral ’brick’ elements, where
the vertical cross-section of the two-layer aquifer prob-
Hemker et al.4 analyzed the steady-state problem by lem coincides with the x 1 – x 2 -plane, while the x 3 -
using the finite-element code MicroFEM3. They coordinate coincides with the horizontal north direc-
applied a 3D finite element mesh consisting of 2470 tion. For such an coordinate orientation the Eulerian
nodes and 4800 triangular elements in each layer. The anisotropic angles have to be defined for the general
nodal distance decreases from 4 m at the model bound- rotation matrix of anisotropy according to Eq. (5-11),
aries to 2 m inside the anisotropic block. viz.,
cbcilt=ö=OTP
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
a)
x2
x3 x1
b)
x2
x1
Figure 25.3 FEFLOW mesh used for the example 1 consisting of 664,000 hexahedral elements with 667,521
nodes: a) 3D view, b) cross-sectional view.
OTQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKO=kìãÉêáÅ~ä=^å~äóëáë
The geometric and boundary conditions used for the Table 25.3 Material parameters for example 1
FEFLOW simulation of example 1 are summarized in
Table 25.2. Material parameters are listed in Table major principal conductivity
–5 –1
of inner area, 1.15741 ⋅ 10 ms
25.3. m m
K 1 = K 2 = K max
Table 25.2 Geometric and boundary conditions of
minor principal conductivity
example 1 m
of inner area, K 3 = K min 1.15741 ⋅ 10
–6
ms
–1
impervious boundaries, except at vertical front and back The steady-state problem 2 was computed by
faces Hemker and Bakker5 with the MicroFEM code3, where
a 3D finite-element mesh of 18 layers was built up, i.e.,
Table 25.3 Material parameters for example 1 each 2 m thick layer was subdivided by two computa-
tional layers. They used a nodal spacing along the hori-
Quantity Value zontal direction of exactly 2 m within the central zone,
which increases to nearly 10 m at the east and west
isotropic conductivity of outer
m m
area, K 1 = K 2 = K 3
m
1.15741 ⋅ 10
–5
ms
–1 model boundaries to attain 11,668 nodes per layer
(210,024 nodes in total).
cbcilt=ö=OTR
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
For the FEFLOW simulations we choose a similar Table 25.4 Geometric and boundary conditions of
mesh resolution comparable to Hemker and Bakker5. example 2 (continued)
Again, it is convenient in FEFLOW to use the vertical
cross-section as the x 1 – x 2 -plane so that the horizontal width of inner zone 100 m
extent of the aquifer system is oriented to the x 3 -coor- width of outer zone (left and 100 m
dinate (north) direction. According to Table 25.1 the right)
anisotropic angle is only defined around the vertical
9 layers with a thickness of 2m
axis. In using our coordinate alignments it means - in
each layer
the same manner as for example 1 - that the rotation is
done explicitly via the Eulerian angle θ for the rotation steady-state flow:
around the vertical x 2 -axis (see Eq. (25-1)).
potential difference in north Δh = 0.3 m
direction
The FEFLOW mesh for example 2 is shown in
Figure 25.3. It consists of 259,200 hexahedral elements impervious boundaries, except at vertical front and back
with 273,097 nodes. In the cross-sectional x 1 – x 2 - faces
plane the nodal distances are 2.5 m in the x 1 -direction
and 0.5 m in the x 2 -direction. The nodal distance along Table 25.5 Material parameters for example 2
the horizontal x 3 -direction is constantly 5 m.
Quantity Value
The geometric and boundary conditions used for the conductivity
m
K1 for all ele- 1.15741 ⋅ 10
–4
ms
–1
FEFLOW simulation of example 2 are summarized in ments
Table 25.4. Material parameters are listed in Table m –5 –1
25.5. conductivity K 2 for all ele- 1.15741 ⋅ 10 ms
ments
Table 25.4 Geometric and boundary conditions of m –5 –1
conductivity K 3 for all ele- 5.78704 ⋅ 10 ms
example 2
ments
Quantity Value Eulerian angle φ for all ele- – 90°
ments
geometry:
Eulerian angle ψ for all ele- 90°
domain measure 300 m x 300 m x 18 m ments
vertical plane is oriented in the x 1 – x 2 -coordinate direc- Eulerian angle θ for elements 0°
tions (x - y plane) of the isotropic outer zone
horizontal plane is oriented in the x 1 – x 3 -coordinate Eulerian angle θ for elements α -angle pattern accord-
directions (x-z plane) of the anisotropic inner zone ing to Table 25.1
OTS=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKP=oÉëìäíë
a)
b)
x2
x1
Figure 25.4 FEFLOW mesh used for the example 2 consisting of 259,200 hexahedral elements with 273,097
nodes: a) 3D view with drawn θ -angle pattern of anisotropy, b) cross-sectional view; (exaggeration 3 : 1).
cbcilt=ö=OTT
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
a) b)
1 5 4 3 2
1 3 5
2 4
Figure 25.5 Plan, side, and front view of the two-layer model (example 1) with hydraulic head con-
tours of both layers and five pathlines (1, 2, 3, 4, 5) starting at different levels in the upper layer on the
southern boundary: a) results by Hemker et al.4, b) FEFLOW results.
OTU=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKP=oÉëìäíë
cbcilt=ö=OTV
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
and Bakker5 reveal a rather good agreement. FEFLOW compute the same results, which also agree
Figure 25.8 displays the computed whirl pattern at a very well with the analytical predictions.
vertical cross-section. Evidently, MicroFEM and
a)
b)
Figure 25.7 Groundwater whirls occurring for example 2 simulated by FEFLOW: a) 3D view of computed flow
pathlines, b) pathlines projected on the vertical cross-section (exaggeration 3 : 1).
OUM=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKP=oÉëìäíë
20
a) b)
15
10
0
50 100 150 200 250
c) 0 d)
-2
-4
-6
-8
-1 0
-1 2
-1 4
-1 6
-1 8
50 100 150 200 250
Figure 25.8 Cross-section showing the pattern of groundwater whirls: a) MicroFEM results by Hemker and Bakker5.
Projection of 20 pathlines in each of the 9 layers, pathlines start in the centre of each cell and at equal distances to the
left and right in the homogeneous side blocks. All starting points are located at 100 m north of the southern boundary
and all pathlines run for 100 m due north. Different colors are used for different starting depths. b) FEFLOW results
using the same starting points at 100 m north of the southern boundary, however, pathlines run for 200 m due north. c)
Analytical results of the streamfunction contours in the cross-section. d) FEFLOW results for a refined pathline pat-
tern at the cross-section, pathlines run over the full horizontal distance (300 m) due north.
The same pathlines projected on a side view are section with the analytical (exact) results presented by
depicted in Figure 25.9 comparing the FEFLOW Hemker and Bakker5 we found in Figure 25.10 that
results with the MicroFEM results. Slight differences while the profiles agree rather well, the peaks in the
in the divergent characteristics of the pathlines can be zigzag-profiles of the hydraulic head appear generally
recognized that are obviously caused by the different somewhat smaller in the numerical predictions. More
meshes, velocity computations and pathline tracking refined 3D meshes seem to be needed to improve the
techniques used in the different codes. Comparing agreement with the analytical results.
FEFLOW’s hydraulic head distribution along a cross-
cbcilt=ö=OUN
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
a)
0
b)
-5
-10
-15
0 10 20 30 40 50 60 70 80 90 100
Figure 25.9 The 9 * 20 pathlines projected on a side view: a) MicroFEM results5 b) FEFLOW results.
OUO=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
ORKQ=`çåÅäìëáçå
a) b)
-3 Heads in all layers x 10-3
x 10
3
3
layer 1 layer 1
layer 2 layer 2
layer 3 layer 3
2 layer 4
layer 4 2
layer 5 layer 5
layer 6 layer 6
layer 7 layer 7
layer 8 layer 8
layer 9 layer 9
1 1
Head
0 0
-1 -1
-2 -2
-3 -3
50 100 150 200 250
50 100 150 200 250
Figure 25.10 Hydraulic heads in an east-west cross-section of the layered aquifer: a) analytical profiles5, b) present
FEFLOW results.
cbcilt=ö=OUP
ORK=qÜêÉÉJÇáãÉåëáçå~ä=^åáëçíêçéó=~åÇ=m~ííÉêåë=çÑ=dêçìåÇï~íÉê=tÜáêäë
ORKS oÉÑÉêÉåÅÉë
1. Bakker, M. and Hemker, K. A Dupuit formulation for flow in lay-
ered, anisotropic aquifers. Advances in Water Resources 25
(2002) 7, 747-754.
2. Bakker, M. and Hemker, K. Analytical solutions for groundwater
whirls in box-shaped, layered anisotropic aquifers. Advances in
Water Resources 27 (2004) 11, 1075-1086.
3. Hemker, C.J. and Nijsten, G.J. Groundwater flow modeling using
MicroFem: Version 3. Amsterdam, The Netherlands; Hemker
Geohydroloog Amsterdam, 1996.
4. Hemker, K., van den Berg, E. and Bakker, M. Ground water
whirls. Groundwater 42 (2004) 2, 234-242.
5. Hemker, K. and Bakker, M., Groundwater whirls in heteroge-
neous and anisotropic layered aquifers. In: FEM_MODFLOW
Intern. Conf. on Finite Element Models, MODFLOW, and More:
Solving Groundwater Problems, ed. K. Kovar et al., Proceedings,
Karlovy Vary (Czech Republic), 2004, 27-30; long text version
on the proceedings CD.
6. Hemker, K. and Bakker, M., Analytical solutions for whirling
groundwater flow in two-dimensional heterogeneous anisotropic
aquifers. Water Resour. Res. 42 (2006), W12419, doi:10.1029/
2006WR004901.
7. Meesters, A.G.C.A., Hemker, C.J. and can den Berg, E.H. An
approximate analytical solution for well flow in anisotropic lay-
ered aquifer systems. Journal of Hydrology 296 (2004) 241-253.
OUQ=ö=oÉÑÉêÉåÅÉ=j~åì~ä=J=m~êí=ff
pìÄàÉÅí=fåÇÉñ
pìÄàÉÅí=fåÇÉñ ëáåÖäÉ=ïÉää=RVI=SQI=SU
_çìëëáåÉëè=~ééêçñáã~íáçå=PUÓPVI=NRVÓNSMI=OMP
bñíÉåÇÉÇ=PQI=PVI=NSMÓNSN
íÉêã=çÑ=RO
^ ÄêÉ~âíÜêçìÖÜ=OSU
ÄêáåÉ=PVI=NUPI=OMPI=ONTI=ONU
~Åíáîáíó=ÅçÉÑÑáÅáÉåíë=ON
éççä=OON
~Çëçêéíáçå=ÑìåÅíáçå=OOI=RP
_êìÅÜ=~åÇ=píêÉÉíÛë=éêçÄäÉã=NOV
~åáëçíêçéó=TVÓUSI=OTN
Äìçó~åÅó
~ñáëJé~ê~ääÉä=URI=UT
ê~íáç=NVR
íÜêÉÉJÇáãÉåëáçå~ä=UNÓUR
íÉêã=PR
íïçJÇáãÉåëáçå~ä=UMÓUN
~èìáÑÉê `
ÅçåÑáåÉÇ=RQ Å~éáää~êó=Ä~êêáÉê=OQV
ÉäÉî~íáçåë=UU `Éäá~=Éí=~äKÛë=éêçÄäÉã=OPN
ä~óÉêÉÇ=NOV ÅÉääìä~ê=ÅçåîÉÅíáçå=PV
éÜêÉ~íáÅ=QP ÅÜÉãáÅ~ä
obs=E^obsF=OR ~ÅíáîáíáÉë=NUI=OM
ëóëíÉã=QS ÅçãéäÉñ=NV
íÜáÅâåÉëë=PM ÉèìáäáÄêáìã=NU
ìåÅçåÑáåÉÇ=PNI=QPI=QPÓQQI=RR ã~ÅêçëÅçéáÅ=Äìäâ=ê~íÉ=NV
^êêÜÉåáìë=Éèì~íáçå=NV ãçÇÉäáåÖ=NUÓNV
~ñáëóããÉíêáÅ=ÖÉçãÉíêó=PPI=OQV éêçÇìÅíë=NU
_ êÉ~Åí~åíë=NU
êÉ~Åíáçå=NU
Ä~ä~åÅÉ=ä~ïë
ÅÜÉãáÅ~ä=ëéÉÅáÉë=NR
~èìáÑÉêJ~îÉê~ÖÉÇ=OQÓOS
Åçäã~íáçå=SN
áåÇáîáÇì~ä=NS
ÅçãéäÉãÉåí~êó=ÅçåÇáíáçåë=TM
áåíÉêÑ~ÅÉ=NR
ÅçãéêÉëëáÄáäáíó=PQ
ã~ÅêçëÅçéáÅ=NRÓNU
ÑäìáÇ=OS
ãáÅêçëÅçéáÅ=NQ
ÅçåÅÉåíê~íáçå=RO
_^pa=íÉÅÜåáèìÉ=QSÓQUI=SMI=UQ
ÅçãéçåÉåí=NT
ÄçìåÇ~êó=ÅçåÇáíáçå=RTÓSV
Åçåí~ãáå~åí=NR
`~ìÅÜó=RUI=SQI=SU
áåáíá~ä=QU
aáêáÅÜäÉí=RUI=SPI=ST
ã~ñáãìã=OT
Çê~áå~ÖÉ=Öê~ÇáÉåíJíóéÉ=OPR
ãçä~ê=OM
ÑêÉÉ=ëìêÑ~ÅÉ=RV
êÉÑÉêÉåÅÉ=POI=UTI=NMN
áåíÉÖê~ä=`~ìÅÜó=SMI=SRI=SV
ëçäáÇ=éÜ~ëÉ=ON
áåíÉÖê~ä=kÉìã~åå=SMI=SRI=SV
ÅçåÇìÅíáîáíó
å~íìê~ä=SNI=SRI=SV
~åáëçíêçéáÅ=TV
kÉìã~åå=RUI=SPI=ST
cbcilt=ö=OUR
pìÄàÉÅí=fåÇÉñ
áëçíêçéáÅ=UR ~ÄëçäìíÉ=OQ
êÉä~íáîÉ=OOU ÇÉéíÜJáåíÉÖê~íÉÇ=OR
íÜÉêã~ä=RP åçêã~ä=QSI=RV
ÅçåëÉêî~íáçå ä~ï=NSN
Åçåí~ãáå~åíë=NP îÉäçÅáíó=RO
ÉåÉêÖó=NPI=OQ ÇÉÅ~ó
ÑäìáÇ=ã~ëë=NP áêêÉîÉêëáÄäÉ=OM
ãçãÉåíìã=NP ê~íÉ=OPI=RO
ÅçåëÉêî~íáçå=çÑ=~=éÜ~ëÉ ÇÉåëáíó
Åçåí~ãáå~åí=ã~ëë=NS ~åçã~äó=OT
ÉåÉêÖó=NS Äìäâ=RP
ÑäìáÇ=ã~ëë=NS ÇáÑÑÉêÉåÅÉ=ê~íáç=OS
ãçãÉåíìã=NS ÑäìáÇ=OMI=OSÓPM
îÉêíáÅ~ääó=~îÉê~ÖÉÇ êÉÑÉêÉåÅÉ=PS
Åçåí~ãáå~åí=OR êÉÑÉêÉåÅÉ=ÑäìáÇ=OM
ÉåÉêÖó=OR ëçäáÇ=ROI=RP
ã~ëë=OR ëçäìíáçåë=OMQÓOMR
ãçãÉåíìã=OR íÜÉêã~ääó=î~êá~ÄäÉ=ÑäìáÇ=OT
ÅçåëÉêî~íáçå=éêáåÅáéäÉ=NPÓNR ÇÉêáî~íáîÉ
Åçåëíê~áåíë=TMÓTU ã~íÉêá~ä=EëìÄëí~åíá~äF=NP
å~íìê~ä=TO ÇáÑÑìëáçå
Åçåí~ãáå~åí=ëéÉÅáÉë=NR ãçäÉÅìä~ê=RP
Åçåíáåìáíó=Éèì~íáçå=PQ ÇáêÉÅíáçå~ä=ÅçëáåÉë=UQ
ÅçåîÉÅíáçå ÇáëéÉêëáçå=NT
ÅÉääìä~ê=PV ÜóÇêçÇóå~ãáÅ=RO
ÑêÉÉ=PV ãÉÅÜ~åáÅ~ä=NU
ãáñÉÇ=PV íÜÉêãçJ=NT
å~íìê~ä=PVI=NRV ÜóÇêçÇóå~ãáÅ=RO
íÜÉêãçÜ~äáåÉ=NVP ÇáëíêáÄìíáçå=ÅçÉÑÑáÅáÉåí=RP
ÅçåîÉÅíáîÉ=Ñçêã=PT ÇáíÅÜ=OQP
ÅççêÇáå~íÉ=ëóëíÉã ÇáîÉêÖÉåÅÉ=Ñçêã=PTI=SR
~ñáëóããÉíêáÅ=PPI=RN Çê~áå~ÖÉ=ÉñéÉêáãÉåí=OQP
`~êíÉëá~å=OMI=TV Çê~ïÇçïå=TOI=NOM
bìäÉêá~å=NP ÇêóáåÖ=ÅìêîÉ=ORT
a aìéìáí=~ëëìãéíáçå=OQ
a~êÅó b
Éèì~íáçå=NUI=PR báåëíÉáåÛë=ëìãã~íáçå=ÅçåîÉåíáçå=NQ
ÇáãÉåëáçåäÉëë=NSO bäÇÉêÛë=éêçÄäÉã=NRV
Ñäìñ ÉäÉî~íáçå=QSI=UU
OUS=ö=oÉÑÉêÉåÅÉ=j~åì~ä
pìÄàÉÅí=fåÇÉñ
ÉåÉêÖó d
áåíÉêå~ä=PM d~êÇåÉêÛë=éêçÄäÉã=OOT
áåíÉêå~ä=EíÜÉêã~äF=NT Ö~ë=Åçåëí~åí=NV
ÉåÉêÖó=ÅçåëÉêî~íáçå=NS dÜóÄÉåJeÉêòÄÉêÖ=áåíÉêÑ~ÅÉ=OMP
Éèì~íáçåë=çÑ=ëí~íÉ=OSÓPP Öê~îáíó=OM
ÉèìáäáÄêáìã Öê~îáíó=îÉÅíçê=NT
Åçåëí~åí=NU ÖêçìåÇï~íÉê=Ñäçï
íÜÉêã~ä=OQ ÇÉåëáíó=î~êá~ÄäÉ=ONP
Éêêçê=ÑìåÅíáçå ÖêçìåÇï~íÉê=êÉÅÜ~êÖÉ=QRI=QS
ÅçãéäÉãÉåí~êó=NOR
e
bìäÉêá~å=~åÖäÉ=OTS
Ü~äÑJäáÑÉ=OPI=RO
bìäÉêá~å=~åÖäÉë=UM
e~îÉêâ~ãé=ãçÇÉä=OQQ
Éî~éçê~íáçå=OPR
ÜÉ~í=Å~é~Åáíó=OQI=PM
Éñé~åëáçå
ÑäìáÇ=RO
äáåÉ~ê=OT
ëçäáÇ=RO
î~êá~ÄäÉ=OTI=OV
îçäìãÉíêáÅ=UT
Éñé~åëáçå=ÅçÉÑÑáÅáÉåí=OS
eÉåêó
åçåäáåÉ~ê=íÉãéÉê~íìêÉJÇÉéÉåÇÉåí=OT
~Çëçêéíáçå=ON
Éñé~åëáçå=ÑìåÅíáçå
áëçíÜÉêã=OOI=RP
ãçÇáÑáÉÇ=íÜÉêã~ä=ÑäìáÇ=PQ
eÉåêóÛë=éêçÄäÉã=OMP
ÉñíÉåëáîÉ=éêçéÉêíáÉë=NP
eççéÉë=~åÇ=e~êäÉã~ååÛë=ÉñéÉêáãÉåí=NQV
ÉñíÉêå~ä=ëìééäó=NR
ÜóÇê~ìäáÅ=ÅçåÇìÅíáîáíó=OMI=PRI=PS
c ÜóÇê~ìäáÅ=ÜÉ~Ç=OMI=RO
cbcilt evaol`lfk=ONP
ëáãìä~íáçå=ëóëíÉã=RN ÜóÇêçÇóå~ãáÅ=ÇáëéÉêëáçå=RO
cáÅâá~å=ã~ëë=Ñäìñ=NT ÜóëíÉêÉëáë=ORT
Ñäìñ ÜóëíÉêÉíáÅ=Ñäçï=ORT
cáÅâá~å=ã~ëë=NT
f
cçìêáÉêá~å=ÜÉ~í=NT
áåÑáäíê~íáçå=OPRI=OQV
cçìêáÉêá~å=ÜÉ~í=Ñäìñ=NT
ê~íÉ=QRI=QS
Ñê~ÅíìêÉ=NPVI=OSP
áåáíá~ä=ÅçåÇáíáçå=RT
Ñê~ÅíìêÉÇ=ë~åÇëíçåÉ=OSP
áåíÉåëáîÉ=éêçéÉêíáÉë=NP
ÑêÉÉ=ÅçåîÉÅíáçå=PV
áçå=ÉñÅÜ~åÖÉ=Å~é~Åáíó=ON
ÑêÉÉ=ëìêÑ~ÅÉ=QRÓQSI=TN
áëçíÜÉêã
ãçîÉãÉåí=QS
cêÉìåÇäáÅÜ=OOI=RP
ãìäíáéäÉ=QP
eÉåêó=ONI=RP
cêÉìåÇäáÅÜ
i~åÖãìáê=ONI=RP
~Çëçêéíáçå=OO
áëçíÜÉêã=OOI=RP
cbcilt=ö=OUT
pìÄàÉÅí=fåÇÉñ
g ~èìÉçìë=OM
g~ÅçÄá~å=ã~íêáñ=UP ÑäìáÇ=NV
àçáåáåÖ=QT ëçäáÇ=NVI=OM
îçäìãÉ=Ñê~Åíáçåë=NV
h
éÜêÉ~íáÅ=~èìáÑÉê=QP
hêçåÉÅâÉê=íÉåëçê=NU
éáÉòçãÉíêáÅ=ÜÉ~Ç=OM
i éçêçëáíó=PMI=RO
i~åÖãìáê Çê~áå~ÄäÉ=PR
áëçíÜÉêã=OOI=RP ÉÑÑÉÅíáîÉ=PRI=QS
i~åÖãìáê=~Çëçêéíáçå=ON Ñáää~ÄäÉ=PR
ä~ï=çÑ=ã~ëë=~Åíáçå=Eij^F=NU éêÉëëìêÉ=NTI=PRI=RP
ä~óÉê=QT Å~éáää~êó=OOU
ä~óÉêÉÇ=ëíêìÅíìêÉë=NOV Åçåëí~åí=QR
äÉ~ÅÜáåÖ=é~ê~ãÉíÉêë=SS ÑäìáÇ=PR
äÉ~â~ÖÉ=ÅçÉÑÑáÅáÉåí=SN éêáåÅáé~ä=ÇáêÉÅíáçåë=TV
iÉïáë=åìãÄÉê=NSP ëÜ~éÉJÇÉêáîÉÇ=UP
äçåÖáíìÇáå~ä=ÇáëéÉêëáîáíó=RP éêçÇìÅíáçå=íÉêã
äìãéÉÇ=ã~ëë=Ä~ä~åÅÉ=ÑäìñÉë=TS òÉêçJçêÇÉê=OP
j o
ã~ÅêçëÅçéáò~íáçå=NR o~óäÉáÖÜ=åìãÄÉê=PVI=NSPI=OON
ã~ëë=ÅçåëÉêî~íáçå=NS ÅêáíáÅ~ä=QM
ã~ëë=Ñê~Åíáçå=RP ÜÉ~í=QM
ãÉëÜ ëçäìíÉë=QM
~ÇàìëíÉÇ=QT êÉ~Åíáçå
jáÅêçcbj=OTP ÜÉíÉêçÖÉåÉçìë=NU
ãáñÉÇ=ÅçåîÉÅíáçå=PV ÜÉíÉêçî~äÉåí=OO
ãçáëíìêÉ=Å~é~Åáíó=OOU ÜçãçÖÉåÉçìë=NU
ãçãÉåíìã=ÅçåëÉêî~íáçå=NS áçå=ÉñÅÜ~åÖÉ=OM
ãçîáåÖ=ãÉëÜ=QPI=QSÓQU åçåÉèìáäáÄêáìã=OP
k êÉ~Åíáçå=ê~íÉ=NT
å~íìê~ä=ÅçåîÉÅíáçå=PV êÉéêÉëÉåí~íáîÉ=ÉäÉãÉåí~êó=îçäìãÉ=EobsF=NR
l êÉí~êÇ~íáçå=OOI=RP
lÖ~í~=~åÇ=_~åâëÛ=éêçÄäÉã=NOP êÉí~êÇ~íáçå=Ñ~Åíçê=RO
oÉóåçäÇë=íê~åëéçêí=íÜÉçêÉã=NQ
m êçí~íáçå=ã~íêáñ=TV
mÉÅäÉí=åìãÄÉê=PVI=NSPI=ONV
éÉêãÉ~Äáäáíó=PS
p
íÉåëçê=NT ë~äí=ÇçãÉ=éêçÄäÉã=ONP
éÜ~ëÉ ë~äíï~íÉê=OMP
ÜáÖÜJÅçåÅÉåíê~íáçå=PN
OUU=ö=oÉÑÉêÉåÅÉ=j~åì~ä
pìÄàÉÅí=fåÇÉñ
áåíêìëáçå=TT íÜÉêãçÜ~äáåÉ=ÅçåîÉÅíáçå=NVP
ë~äíï~íÉê=áåíêìëáçå=NRVI=NVTI=OMR íê~åëÑÉê=ÅçÉÑÑáÅáÉåí=RUI=RVI=SOI=SQI=SSI=SUI=SV
ë~íìê~íáçå íê~åëÑÉê=ê~íÉ=RUI=SQI=SUI=UVI=VMI=VNI=VOI=VPI=VQI=VRI=VSI
ÉÑÑÉÅíáîÉ=OPOI=OPS VTI=VUI=VVI=NMM
ÑäìáÇ=OPO íê~åëãáëëáîáíó=ORI=UN
ëâÉäÉíçå=ÅçãéêÉëëáÄáäáíó=PM íê~åëéçêí=Éèì~íáçåë
ëäáÅÉë=QS ÅçåîÉÅíáîÉ=Ñçêã=PSÓPU
ëçáä ÇáîÉêÖÉåÅÉ=Ñçêã=PSÓPU
ÇáíÅÜJÇê~áåÉÇ=OQP íê~åëéçêí=éêçÅÉëë
ÜçãçÖÉåÉçìë=OPN Äìçó~åÅóJÇêáîÉå=PV
áåÜçãçÖÉåÉçìë=OPR íê~åëîÉêëÉ=ÇáëéÉêëáîáíó=RP
ëçäìíáçå=åçêã~äáíó=ON r
ëçêéíáîÉ=åìãÄÉê=OOU ìåë~íìê~íÉÇ=ëçáä=Åçäìãå=OOT
ëçêéíáîáíó=RP
s
ëéÉÅáÉë
î~å=dÉåìÅÜíÉå=J=jì~äÉã=ãçÇÉä=OPOI=OPS
ÇáëëçäîÉÇ=NU
î~å=dÉåìÅÜíÉå=éêçÄäÉã=OPR
ëíçáÅÜáçãÉíêáÅ
s~ìÅÜ~ìÇ=Éí=~äKÛë=éêçÄäÉã=OQP
ÅçÉÑÑáÅáÉåíë=NU
îÉäçÅáíó
åìãÄÉêë=NU
Ä~êóÅÉåíêáÅ=NQ
ëíçê~ÖÉ=ÅçÉÑÑáÅáÉåí
ÅçåëáëíÉåí=ONR
ëéÉÅáÑáÅ=PQ
a~êÅó=OM
ëíê~íáÖê~éÜó=QT
áåíêáåëáÅ=OM
ëíêÉ~ã=ÑìåÅíáçåë=NSQ
âáåÉã~íáÅ=OM
píêçìÖÜ~ä=åìãÄÉê=NSP
éÜ~ëÉ=áåíÉêÑ~ÅÉ=NT
ëìÅíáçå=TO
îáëÅçëáíó
ëìééäó
Çóå~ãáÅ=NTI=PNÓPP
ÜÉ~í=NT
ÅçåÅÉåíê~íáçåJÇÉéÉåÇÉåí=PO
ã~ëë=NT
íÉãéÉê~íìêÉJÇÉéÉåÇÉåí=PO
ëïÉéí=Ñçêï~êÇ=êÉÖáãÉ=OOM
êÉÑÉêÉåÅÉ=OMI=PNI=PS
q êÉä~íáçå=ÑìåÅíáçå=PN
q~óäçê=ëÉêáÉë=OUI=OMR îáëÅçëáíó=êÉä~íáçå=ÑìåÅíáçå=RO
íÉãéÉê~íìêÉ=NUI=OMI=OQI=PMI=RO îçáÇ=ëé~ÅÉ=PN
åçêã~äáòÉÇ=RP îçäìãÉ=Ñê~Åíáçå=çÑ=~=éÜ~ëÉ=NT
êÉÑÉêÉåÅÉ=OTI=PNI=UTI=NMN îçêíáÅáíó=NSPI=NSQ
ëóëíÉã=OQ
t
qÜÉáëÛ=éêçÄäÉã=NNT
ï~íÉê=í~ÄäÉ=QPI=QUI=SMI=TMI=TOI=TTI=NNTI=OOTI=OQP
íÜÉêãçÇáëéÉêëáçå=NTI=RP
ïÉííáåÖ=ÅìêîÉ=ORT
íÜÉêãçÇóå~ãáÅë
ïÜáêäë=OTN
Ñáêëí=ä~ï=çÑ=NP
cbcilt=ö=OUV
pìÄàÉÅí=fåÇÉñ
OVM=ö=oÉÑÉêÉåÅÉ=j~åì~ä
^ìíÜçê=fåÇÉñ
^ìíÜçê=fåÇÉñ d
d~äÉ~íá=OMUI=OMV
d~êÇåÉê=OOTI=OOU
dê~ó=OR
^ dêáë~Åâ=NPVI=NQQI=NQS
^ÅâÉêÉê=QMI=NRV dìêÉÖÜá~å=OQPI=OQRI=OQS
^åÖáê~ë~=NVP e
^òáò=QM e~êäÉã~åå=NQVI=NRMI=NRN
_ e~ëë~åáò~ÇÉÜ=NRI=PN
_~ÅÜã~í=NRI=PM eÉãâÉê=OTN
_~ââÉê=OTN eÉåêó=OMPI=OMQ
_~åâë=NOR eÉêÄÉêí=ONPI=ONQI=ONRI=ONTI=ONUI=ONVI=OOM
_~ñíÉê=OT eáããÉäëÄ~ÅÜ=OSP
_É~ê=NQI=NRI=NTI=NUI=OQI=PMI=NSM eçäëí=QM
_ÉÅâ=QM eçäòÄÉÅÜÉê=OT
_Éà~å=PUI=QM eçãëó=QM
_áêÇ=PU eççéÉë=NQVI=NRMI=NRN
_äóíÜÉ=QM eçêåÉ=QM
_çêáÉë=PUI=QM eçêíçå=QM
_êìÅÜ=NPMI=NPOI=NPS eëì=QM
` eìó~âçêå=NROI=NRQ
`~äí~ÖáêçåÉ=QM g
`~êëä~ï=NNT g~Åâëçå=PN
`Éäá~=OPNI=OPPI=OPQ g~ÉÖÉê=NNT
`ÜÉå=NVP h
`ÜÉåÖ=QM h~ëëçó=QM
`ä~ìëåáíòÉê=ORTI=OSM háåòÉäÄ~ÅÜ=NOP
`çãÄ~êåçìë=PUI=QM håìéé=QR
a hçäÇáíò=NSI=QMI=NRVI=NVPI=ONR
aÉ=pÅÜÉééÉê=QMI=NRV hê∏Üå=OMQ
aáÉêëÅÜ=NRI=NSI=ONI=OOI=ORI=PVI=QMI=QUI=RNI=SRI=NOVI=NPMI hêìëÉã~å=NNT
NPOI=NPQI=NPRI=NPSI=NRVI=NVPI=ONRI=OPN i
b iÉåÜ~êÇ=ORT
bäÇÉê=QMI=NRVI=NSQI=NSRI=NSSI=NSTI=NTR iÉîÉê=PN
bî~åë=NSMI=NSQI=NVP j
c jÅháÄÄáå=QM
cêáåÇ=NNUI=NNVI=NOM jÉêÅÉê=PN
cêçäâçîáÅ=QMI=NRV jáÅÜÉäë=QU
cbcilt=ö=OVN
^ìíÜçê=fåÇÉñ
jìêê~ó=NVP píê~ìë=QM
k píêÉÉí=NPMI=NPOI=NPS
kÖóìÉå=OQP q
káÉäÇ=PU q~åÖ=NPVI=NQNI=NQQI=NQRI=NQS
kìåå=NVP qÜÉáë=NNT
kΩíòã~åå=OQP qÜáÉäÉ=NOVI=NPMI=NPOI=NPQI=NPRI=NPS
l qêÉîáë~å=QM
lÛpìääáî~å=QM s
lÖ~í~=NOR î~å=dÉåìÅÜíÉå=OPRI=OPSI=OPVI=OQMI=OQN
läÇÉåÄìêÖ=ONTI=ONUI=ONVI=OOMI=OONI=OQV s~ìÅÜ~ìÇ=OQPI=OQS
m sçëë=NRVI=NSSI=NSTI=NSUI=NTRI=NVPI=OMUI=OMVI=ONR
m~äã=QM t
mÉêêçÅÜÉí=OTI=OUI=OPN t~ää~ÅÉ=OT
mÜáäáé=OOU tÉÄÄ=OQVI=ORP
máÅâÉåë=NPVI=NQQI=NQS tÉåÇä~åÇ=OSP
máåÇÉê=PNI=NNUI=NNVI=NOM tççÇáåÖ=QM
mêìÉëë=ONTI=ONUI=ONVI=OOMI=OONI=OQV v
o vÉÜ=QR
o~ÑÑÉåëéÉêÖÉê=NSMI=NSQ vçìåÉë=QMI=ONQ
oáÇÇÉê=NNT w
oçÄáåëçå=QM wÉÄáÄ=QM
oçÖÉêë=QM wáÉäâÉ=NS
oçëë=OQV
oçíÜ=QM
oìÄáå=QM
oìÇê~á~Ü=QM
p
p~ìíó=NOP
pÅÜÉáÇÉÖÖÉê=NT
pÅÜìÄÉêí=QM
pÉÖçä=QRI=NNUI=NQVI=OMP
pÜÉå=NVP
pÜÉêïççÇ=QM
páãâáåë=QM
pçìò~=NRVI=NSSI=NSTI=NSUI=NTRI=NVPI=OMUI=OMVI=ONR
péêáåÖÉê=UP
pêáã~åá=QM
pêáåáî~ë~å=NVP
OVO=ö=oÉÑÉêÉåÅÉ=j~åì~ä