Journal of Plant Nutrition, 36:429–442, 2013

Copyright  C Taylor & Francis Group, LLC

ISSN: 0190-4167 print / 1532-4087 online

DOI: 10.1080/01904167.2012.748063

SIMULATION MODEL OF WATER AND SOLUTE TRANSPORT

IN TWO TYPES OF ROCKWOOL

D. Titouna1 and S. Bougoul2

1
Département de Mécanique, Université de Batna, Batna, Algeria
2
Laboratoire de Physique Energétique Appliquée, Département de Physique, Université de
Batna, Batna, Algeria

2 The aim of this study is to present a two dimensional model for describing the movement of
water and solute transport in two types of rockwool slabs – Floriculture, a high density type; and
Expert, which has a low density. Water movement is described by the Richards equation, which is im-
plicitly solved using the control volume finite element method. The governing convection dispersion
equation describing the solute transport is explicitly solved. Numerical solutions are obtained for the
distribution of water and also for the concentration of solute. The model is validated by comparing
the results of water transport developed by this model with numerical results obtained by CFD that
are validated by experimental results. A composite substrate is tested and validated, and is shown
to give good conditions for development of plants.

Keywords: simulation, two dimensional, water transport, solute transport, rockwool

INTRODUCTION
Soil-less crop substrates should be a starting point for management mod-
els that can be used to develop growth systems, with minimal input of water,
nutrients, and pesticides, with a high production of good quality crops and
with minimal harm to the environment. These systems are better defined
and controllable. For these reasons Bougoul et al. (2005) determined the
hydraulic and physical proprieties of two rockwool slab types, Floriculture
that has a high density and Expert that has a low density.
After this, Bougoul and Boulard (2006) introduced these values for sim-
ulating the movement of water by the use of Computer Dynamics software

Received 8 August 2010; accepted 17 January 2011.

Address correspondence to D. Titouna, Département de Mécanique, Faculté des Sciences de
l’Ingénieur, Université de Batna, Rue Chahid Mohamed El Hadi Boukhlouf, 05000 Batna, Algeria.
E-mail: titouna d@yahoo.fr

429
430 D. Titouna and S. Bougoul

(CFD2000, CFD Systems, Sierra Pacific Corporation, Compton, CA, USA).

The numerical results were then compared and validated with experimen-
tal results. In the same context, Bougoul and Titouna (2010) analyzed the
motion of solution in a rockwool slab by a model of sources and sinks, which
is simple, and the results obtained were validated with experimental results.
In this study, the hydraulic and physical properties, and the experimental
values, were used to validate the numerical model developed for simulating
water movement and solute transport.
For describing water movement and solute transport we can proceed
to an experimental solution, an analytical solution or a numerical solution.
The first requires support equipment that is expensive, the second has the
advantage of easily exploring the behavior of the process and examines lim-
iting situations, but all sorts of simplifications have been applied. Numerical
solutions are more flexible and consider more complicated situations. For
this purpose several codes were elaborated: HYDRUS (Simunek et al., 1998)
used finite element method, FUSSIM (Heinen, 2001) used the control vol-
ume finite element method and Bougoul and Boulard (2006) transformed
the Navier- Stockes equations used by the CFD software to the standard
transport equations. In this study we formulate a two dimensional numer-
ical model for describing water movement and solute transport: it is used
to solve implicitly the Richards equation (Richards, 1931) for unsaturated
porous medium with the control volume finite element method, giving water
transport and it solves explicitly the governing convection dispersion equa-
tion (Bolt, 1982; Heinen, 1997) for solute transport, followed by the general
expressions of boundary and initial conditions. We compared and validated
the result of water transport for the two types of rockwool, Floriculture and
Expert, developed by this model with the numerical results obtained by CFD
(Bougoul and Boulard, 2006) which are validated by experimental results.
The model also developed the results of solute transport, giving scenarios of
concentration distribution and profiles of concentration following the depth
and the length of the substrate. In order to have better conditions for growth
and production of plants a composite substrate is tested and validated.

MATERIALS AND METHODS

The field of simulation is two types of rockwool slabs (Bougoul and
Boulard, 2006) used as growing substrates, manufactured by the Grodan
company (Grodan, Denmanrk): Floriculture, with a high density; and Expert
with a low density. The dimensions of each were 7.5 cm depth, 100 cm length.
Two drippers were located on top of the slab and a drain was located at the
lower right corner. Geometry, position of the drippers and drainage slot are
shown in Figure 1.
 Simulation Model of Water and Solute Transport 431

Water input

20cm 10cm 40cm 10cm 20cm

7.5cm

100cm 5cm

Drainage
FIGURE 1 Geometry of the growing media.

Water Transport
Mathematical Model
The general governing flow equation for incompressible water transport
in porous media is obtained by combination of the continuity equation
and Darcy’s law (Darcy, 1856). The h-based form of the Richards equation
(Richards, 1931) in two dimensions is given by:

   
∂h ∂ ∂h ∂ ∂h ∂K (h)
C(h) = K (h) + K (h) − − Sw (1)
∂t ∂x ∂x ∂z ∂z ∂z

Where C(h) is the differential moisture capacity defined as C =

dθ /dh(L−1), h is the pressure head (L), K(h) is the hydraulic conductiv-
ity (LT−1), θ (h) is the volumetric water (L3L−3), Sw is the volumetric sink
(L3L−3T−1), x is a horizontal coordinate (L), z is the vertical coordinate, as-
sumed positive downward (L), and t is the time (T). The hydraulic properties
have been described by the models of Mualem (1976) and Van Genuchten
(1980). The hydraulic conductivity characteristic given by Mualem (1976)
is:

⎧ ⎫
⎪
⎨ 1+|αh|n
m
−|αh|n−1
2
⎪
⎬
h≤0
K (h) = Ks 1+|αh|n
(λ+2)m
(2)
⎪
⎩1 ⎪
h > 0⎭

Where Ks is hydraulic conductivity at saturation (LT−1), λ, n, m are curve

shape parameters (dimensionless), α is the curve shape parameter (L−1).
432 D. Titouna and S. Bougoul

TABLE 1 Physical properties of two rockwool slab types

Type KS (ms−1) θ S (m3m−3) θ r (m3m−3) α(m−1) n(1) m(1)

Fl 0.002 0.975 0.026 15.917 3.061 0.673

Expert 0.006 0.983 0.019 39.212 2.181 0.5415

The water retention characteristic given by the Van Genuchten function

(Van Genuchten, 1980) is:
 
θ (h) − θr 1
(1+|αh|n )m
h≤0
Se (h) = = (3)
θs − θr 1 h>0

Where Se is an effective saturation 0 ≤ Se ≤1, θ r is the residual volumetric

water content (L3L−3), θ s is the saturated volumetric water content (L3L−3).
The differential moisture capacity (Heinen, 1997) is defined as:
 
(θS − θr )nmα n |h|n−1 (1 + |αh|n )−1−m h ≤ 0
c (h) = (4)
0 h>0

Bougoul et al. (2005) experimentally obtained values for physical prop-

erties for the two types of rockwool that are shown in Table 1.
The dimensions given between brackets of the various parameters are
defined as follows:
L is the length, T is the time and M is the mass.

Boundary and Initial Conditions

The initial condition is that of given initial pressure head distribution
(h = −0.5 cm).
The boundary conditions are as follows: the upper boundary condi-
tion is impermeable and the condition is that of no flow (Neumann
condition)(q z = 0), except at the positions where fertigation takes place the
condition is a Neumann or prescribed volumetric flow density (q z = 5.510−5 )
m/s.
The lower boundary is impermeable and the condition is that of no flow
(Neumann condition)(q z = 0), except at the positions where drainage takes
place the condition may be partly of Neumann type and partly of Dirichlet
type in case the drain location is unsaturated or saturated

for h < 0 qz = 0
or for q z ≥ 0 h=0
 Simulation Model of Water and Solute Transport 433

Due to symmetry considerations the left and right boundaries are no

flow (Neumann condition)

∂h
=0
∂x

Numerical Resolution
Equation (1), with equations (2), (3), and (4), is solved numerically
using the control volume finite element method (Patankar, 1980). After
rearrangement of terms the numerically approximated form is:

I,J = AW h I −1,J + A E h I +1,J + AN h I,J −1 + AS h I,J +1 + b

Ah k+1 k+1 k+1 k+1 k+1
(5)

j z J j z J
k k
Kx,i−1, Kx,i,
AW = AE =
xi−1 xi
j −1 xI j xI
k k
Kx,i, Kx,i,
AN = AS = (6)
x j −1 z j
 k
b = AC h k+1
I,J − Kz,i, j − Kz,i, j −1 xI − Sw ,I,J xI z J
k
(7)

Where:
xI z J
AC = CI,J
k
(8)
t
A = A E + AW + AN + AS + AC (9)

The numerical solution gives for any t the θ (x, z) and h (x, z) distribu-
tions. The water results are then used to solve the solute transport equation.

Solute Transport
Mathematical Model
As the rockwool is considered inert, the governing convection-dispersion
equation (Bolt, 1982; Heinen, 1997) for solute transport is written in two
dimensions as:
∂Q ∂q s x ∂q s z
=− − − Ss (10)
∂t ∂x ∂z
The total density of solute per unit volume of substrate (M L−3) is:

Q = θc (11)
434 D. Titouna and S. Bougoul

q s is the solute mass flux density (L−2MT−1), q s x is the horizontal component

of q s (L−2M T−1), q s z is the vertical component of q s (L−2M T−1).

∂c ∂c
q s x = q x c − θ Dxx − θ Dxz
∂x ∂z
∂c ∂c
qs z = q z c − θ Dzz − θ Dz x (12)
∂z ∂x

Where q is the volumetric water flux density (L3 L−2 T−1), q x is the
horizontal component of q (L3 L−2 T−1), q z is the vertical component of q (L3
L−2 T−1), c is the solute concentration (ML−3), Ss is the mass sink strength
for the solute (ML−3 T−1) and θ Di, j is the dispersion-diffusion tensor (L2T−1)
given by Bear and Verruijt (1987) and Simunek et al. (1994):

qi q j
θ Di, j = aT |q | δi j + (aL − aT ) + θ D0 τ (θ )δi j (13)
|q |

Where aL is the longitudinal dispersivity (L), aT is the transversal disper-

sivity (L), |q | is the absolute value of q (L3 L−2 T−1), δ is the Kronecker delta
(δi j = 1 if i = j , δi j = 0 if i = j ), D0 is the diffusion coefficient in free water
(L2 T−1), τ (θ ) is the tortuosity factor (Barraclough and Tinker, 1981).
The above equations can be applied to all individual ions; in this study
NO3 − was considered.

Boundary and Initial Conditions

The initial condition is that of given initial concentration distribution
(c 0 = 15) mmol/l, the upper boundary condition is impermeable and the
condition is that of no nutrient flux (Neumann condition) (q s z = 0), except
at the positions where fertigation takes place where the condition is:

for q z > 0 q s z = c f q z
and for q z ≤ 0 q s z = 0

where c f is the solute concentration in the fertigation water (c f = 15

mmol/l)
The lower boundary is impermeable and the condition is that of no
nutrient flux (Neumann condition) (q s z = 0), except at the drain location
where the condition is:

for q z > 0 q s z = c D q z
and for q z ≤ 0 q s z = 0
 Simulation Model of Water and Solute Transport 435

where c D is the solute concentration at the drain location, assumed to be

constant.
Due to symmetry considerations there is no flow across the left and the
right boundaries (Neumann condition).

∂c
=0
∂x

Numerical Resolution
The continuity equation (10), with equations (12), (13), and (14), is
explicitly solved using the most recent known output of the water parameters
(θ and q).
The total density of solute Qm (M) is defined as:

Qm,I,J = QI,J xI z J (14)

Using equation (10), (14) is written as:

   
t+t
Qm,I,J = Qm,I,J
t
+ q st+t
x,i−1, j − q t+t
s x,i, j z J t + q t+t
s z,i, j −1 − q t+t
s z,i, j xI t
(15)
The concentration can be computed from equation (11):

t+t
Qm,I,J
t+t
c I,J = (16)
θI,J
t+t
xI z J

RESULTS AND DISCUSSION

Water Transport
Numerical calculations obtained by this model were compared with the
results developed by Bougoul and Boulard (2006) and show similar distribu-
tions of moisture for both substrate types (Floriculture and Expert) and we
observe that:

• The saturated zones are always located just below the drippers and at the
base of the substrate and the area just above the drainage slot is always dry.
• The saturated zones are more extensive in the high-density rockwool
(Floriculture) than in the lower-density slab (Expert) (Figure 2).
436 D. Titouna and S. Bougoul

FIGURE 2 Scenarios of water distribution for a) Floriculture; b) Expert (Color figure available online).

Solute Transport
The results validated for water transport are used in the transport model
to obtain the results of solute transport, and we observe that:
 Simulation Model of Water and Solute Transport 437

a)
18.5

18

17.5
Concentration (mmol/l)

17

16.5

16

15.5

15

14.5

0.07 0.06 0.05 0.04 0.03 0.02 0.01

Depth (m)

b)
27
26
25
24
Concentration (mmol/l)

23
22
21
20
19
18
17
16
15
14
13
0.07 0.06 0.05 0.04 0.03 0.02 0.01
Depth (m)

FIGURE 3 Concentration following the depth of the substrate for a) Floriculture and b) Expert (Color
figure available online).

• The concentration is maximum just below the drippers (Figure 3) because

of the accumulations of solutes near the surface due to the prescribed
volumetric flux density across the top boundary.
• The concentration is maximum in the top of the substrate and decreases
to the bottom (Figure 3): this is due to Fick’s law and the phenomenon of
diffusion.
• For the profile of the concentration following the length, we observe that
the concentration is maximum for higher layers, it is minimum for lower
layers and it is constant in the zone of drainage (Figure 4).
438 D. Titouna and S. Bougoul

a) b)

Z = 6.75 cm
Z = 7.20 cm
Z = 7.35 cm Z = 6.75 cm
19 Z = 7.20 cm
28 Z = 7.35 cm
18.75
27
18.5
18.25 26
18
C o n c e n tra tio n ( m m o l/l)

25

C o n c e n tra tio n (m m o l/l)

17.75
24
17.5
17.25 23

17 22
16.75 21
16.5
20
16.25
16 19
15.75 18
15.5
17
15.25
16
15
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
Horizontal distance (m) Horizontal distance (m)

14.8 Z = 0.45 cm
Z = 1.05 cm Z = 0.45 cm
Z = 1.50 cm Z = 1.05 cm
14.6 14 Z = 1.50 cm

14.4
13.5
C on ce n tra tio n (m m ol/l)

14.2
C oncentration (mmol/l)

13
14

13.8 12.5

13.6 12

13.4
11.5
13.2
11
13

12.8 10.5
0 0.25 0.5 0.75 1 0 0.25 0.5 0.75 1
Horizontal distance (m) Horizontal distance (m)

FIGURE 4 Concentration following the length of the substrate for a) Floricultureand b) Expert (Color
figure available online).

• Comparing the two rockwool slab types (Figure 5), we can see that the
concentration in the high density slab (Floriculture) is always lower than
in the lower-density slab (Expert) because:
• The moisture content of Floriculture is higher than that of Expert.
• The porosity of Floriculture is lower than that of Expert and there is
proportionality between the coefficient of diffusion and porosity.

Substrate Optimization
Bougoul and Boulard (2006) proposed a composite slab, made up of two
rockwool types. The top quarter was composed of a high-density rockwool
(Floriculture type) and the remaining part (the bottom three quarters) was
composed of a lower-density rockwool (Expert type). They observed that the
substrate contained a more even distribution of moisture from top to bottom
of the slab.
 Simulation Model of Water and Solute Transport 439

FIGURE 5 Scenarios of concentration distribution for a) Floriculture and b) Expert (Color figure
available online).
440 D. Titouna and S. Bougoul

0.9

0.8 h(m)
0.054
0.049
0.7 0.044
0.039
0.035
0.6 0.030
0.025
0.020
Depth (m) 0.5 0.016
0.011
0.006
0.4
0.001
-0.004
0.3 -0.008
-0.013

0.2

0.1

0
0 0.25 0.5 0.75 1
Horizontal distance(m)

FIGURE 6 Water distribution in composite substrate (Color figure available online).

To have good conditions for development of plants we tested and vali-

dated the substrate moisture in the composite substrate (Figure 6), and we
determined a distribution of solute concentration (Figure 7), which is dif-
fused more across the substrate due to the increased distribution of moisture.

CONCLUSION
This study demonstrates how the simulation model can largely reduce
the number of experiments and helps in finding alternative strategies. After
the determination of hydraulic and physical properties of two rockwool slabs

0.9

0.8 c(mmol/l)
26.055
25.142
0.7 24.229
23.316
22.403
0.6 21.490
20.577
Depth (m)

19.664
0.5 18.751
17.838
16.925
0.4
16.012
15.098
0.3 14.185
13.272

0.2

0.1

0
0 0.25 0.5 0.75 1
Horizontal distance (m)

FIGURE 7 Concentration distribution in composite substrate (Color figure available online).

 Simulation Model of Water and Solute Transport 441

using as growing substrates (Floriculture with a high density and Expert

with a low density) (Bougoul et al., 2005), these values are introduced for
simulating the movement of water (Bougoul and Boulard, 2006) by the use
of Computer Dynamics software (CFD2000). The numerical results were
then compared and validated with experimental results.
In this study a two dimensional numerical model is developed to inves-
tigate the water movement and solute transport in two types of rockwool
slabs. A series of simulations are presented for water distributions and are
compared with numerical results obtained by CFD, which are validated by
experimental results. The distribution of solute concentration is presented
too.
The concentration is maximum just below the drippers and in the higher
layers (top of the substrate): it decreases to the bottom to reach a minimal
value below the drippers in the lower layers (base of the substrate) and it
is constant in the zone of drainage. We observe also that the concentration
in Floriculture is always lower than in Expert. To have better conditions for
growth and production of the plant a composite substrate is tested and val-
idated and we note that the solute concentration is distributed throughout
the substrate. Such a composite substrate would give more efficient utiliza-
tion of water and nutrients, and better growth of crops.

REFERENCES
Barraclough, P. B., and P. B. Tinker. 1981. The determination of ionic diffusion coefficients in field soils.
I. Diffusion coefficients in sieved soils in relation to water content and bulk density. Journal of Soil
Science 32: 225–236.
Bear J., and A. Verruijt. 1987. Modeling Groundwater Flow and Pollution. Dordrecht, the Netherlands: D.
Reidel Publishing Company.
Bolt, G. H. 1982. Movement of solutes in soil: Principles of adsorption/exchange chromatography. In:
Soil Chemistry. B. Physico-Chemical Models, ed. G. H. Bolt, pp. 285–348. Amsterdam: Elsevier Scientific.
Bougoul, S., and T. Boulard. 2006. Water dynamics in two rockwool slab growing substrates of contrasting
densities. Scientia Horticulturae 107: 399–404.
Bougoul, S., S. Ruy, F. de Groot, and T. Boulard. 2005. Hydraulic and physical properties of stonewool
substrates in horticulture. Scientia Horticulturae 104: 391–405.
Bougoul, S., and D. Titouna. 2010. Study of a nutrient solution flows in a saturated rockwool slab using
the model of sources and sinks. Transport in Porous Media 85: 477–487.
Darcy, H. 1856. Les Fontaines Publiques de la Ville de Dijon [The Public Fountains of the City of Dijon].
Paris: Dalmont.
Heinen, M., 1997. Dynamics of water and nutrients in closed, recirculating cropping systems in glasshouse
horticulture with special attention to lettuce grown in irrigated sand beds. Ph.D. Thesis, Wageningen
University, Wageningen, The Netherlands.
Heinen, M. 2001. FUSSIM2: Brief description of the simulation model and application to fertigation
scenarios. Agronomy 21: 285–296.
Mualem, Y. 1976. A new model for predicting the hydraulic conductivity of unsaturated porous media.
Water Resources Research 12: 513–522.
Patankar, S. V. 1980. Numerical Heat Transfer and Fluid Flow. New York: Hemisphere Publishing Corpora-
tion.
Richards, L. A. 1931. Capillary conduction of liquids through porous mediums. Physics 1: 318–333.
Simunek, J., K. Huang, and M. T. van Genuchten. 1998. The HYDRUS code for simulating the one-
442 D. Titouna and S. Bougoul

dimensional movement of water, heat, and multiple solutes in variably – saturated media. Version
6.0. Research Report No. 144. Riverside, CA: U.S. Salinity Laboratory.
Simunek, J., T. Vogel and M. T. Van Genuchten. 1994. The SWMS 2D code for simulating water flow
and solute transport in two-dimensional variably saturated media. Research Report 132. Riverside,
CA: US Salinity Laboratory, ARS, USDA.
Van Genuchten, M. T. 1980. A closed- form equation for predicting the hydraulic conductivity of unsat-
urated soils. Soil Science Society of America Journal 44: 892–898
Copyright of Journal of Plant Nutrition is the property of Taylor & Francis Ltd and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.