Energy and Buildings 43 (2011) 3027–3035

Contents lists available at ScienceDirect

Energy and Buildings

journal homepage: www.elsevier.com/locate/enbuild

Thermal performance of a greenhouse with a phase change material north wall

F. Berroug a,∗ , E.K. Lakhal a , M. El Omari a , M. Faraji b , H. El Qarnia c
a
Laboratoire d’Automatique de l’Environnement et Procédés de Transfert, Université Cadi Ayyad, Faculté des Sciences Semlalia, Département de Physique, Affilié au CNRST, URAC 28,
Marrakech, Morocco
b
Laboratoire de Physique des Matériaux, Microélectronique, Automatique et Thermique, Département de Physiques, Faculté des Sciences Ain Chock, Université Hassan II, Casablanca,
Morocco
c
Laboratoire de Mécanique des Fluides et Energétiques, Université Cadi Ayyad, Faculté des Sciences Semlalia, Département de Physique, Marrakech, Morocco

a r t i c l e i n f o a b s t r a c t

Article history: Solar energy is considered one of the most prospective sources of renewable energy for greenhouse
Received 8 June 2011 heating in cold period for Mediterranean climate. In this paper, the thermal performance of a north
Received in revised form 10 July 2011 wall made with phase change material (PCM) as a storage medium in east–west oriented greenhouse is
Accepted 19 July 2011
analyzed and discussed. CaCl2 ·6H2 O was used as a PCM. A numerical thermal model taken into account
the different components of the greenhouse (cover, plants, inside air and north wall PCM) and based
Keywords:
on the greenhouse heat and mass balance, has been developed to investigate the impact of the PCM on
PCM
greenhouse temperature and humidity. Calculations were done for typical decade climate of January in
Greenhouse heating
North wall
Marrakesh (31.62◦ N, 8.03◦ W). Results shows that with an equivalent to 32.4 kg of PCM per square meter
Numerical analysis of the greenhouse ground surface area, temperature of plants and inside air were found to be 6–12 ◦ C
more at night time in winter period with less fluctuations. Relative humidity was found to be on average
10–15% lower at night time.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction
Due to the high cost of energy, the use of alternative heating
system is important for a greenhouse to provide optimum inside
conditions during winter months. The basic strategy of greenhouse
passive heating system is to reduce the heat losses and at the same
time to transfer excess heat from inside the greenhouse during the
day to heat storage. This heat is used during the night to satisfy the
heating needs of the greenhouse. Several types of passive solar sys-
tems and techniques have been proposed and used by [1–3]. The
most important existing greenhouse heating systems are: water
storage, rock bed storage, mulching, movable insulation and ther-
mal curtain, ground air collector and north wall storage is also used
for raising the greenhouse air temperature. For greenhouse located
in northern hemisphere, east–west orientation is the most suitable,
it favorites a maximum of solar radiations in winter and a minimum
of solar radiations in summer. For east–west oriented greenhouse,
a maximum solar radiation fall on the south wall during winter
months and leaves the greenhouse through north wall. Tiwari [4]
has suggested a term called ‘solar fraction for north wall’ to quan-
tify this loss through north wall. The value of solar fraction will be
more during winter months because of the law altitude angle of the
∗ Corresponding author.
E-mail address: f.berroug@ucam.ac.ma (F. Berroug).
sun and hence the heat losses will be more. Therefore several works
in the literature reveals that the use of north walls for absorption
or reflectance of solar radiations can raise the greenhouse air tem-
perature and reduce the heating needs of the greenhouse. Working
with the concept of solar fraction radiation on the north wall, water
was stored in black steel barrels placed in the north side to raise
the inside air temperature in greenhouse situated at Flagstaff, USA
(35.12◦ N, 111.39◦ W) in which vegetables were grown [5]. The sys-
tem absorbs the incident solar radiations during the day. During the
night, the stored heat is returned to the greenhouse by convection
and radiation. Although the cost of these systems was low but the
water containers occupied valuable ground space. In another appli-
cation of north wall, some reflective material can be painted on the
north wall to reflect solar radiations towards the plants and floor. In
one such study, north wall was constructed having reflective sur-
face in a 50 m2 glasshouse [6]. A high reflective white coating (93%)
was done on plywood to reflect the beam radiation towards the
greenhouse plants and floor. It was observed that the greenhouse
required 14% lesser energy for heating during winter months as
compared to a conventional greenhouse. Gupta et al. [7] replaced
north wall by an inclined surface to maximize the solar radiation on
floor of the greenhouse. It has been found that an inclined reflective
surface is more suitable for utilizing solar radiations transmitting
out through north wall. With a conventional greenhouse solar dryer
of 24 m2 ground area (east–west orientation) and using inclined
reflective north wall covered with aluminized reflector sheet,

0378-7788/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.enbuild.2011.07.020
3028 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035

Nomenclature
εc emission coefficient for thermal radiation of the
Ap plants surface (m2 ) cover
Ag ground area of the greenhouse (m2 ) εw emission coefficient for thermal radiation of the wall
Ac cover surface (m2 )
Aw north wall surface (m2 ) Subscripts
Ai ith inclined surface of greenhouse (m2 ) c cover
ca specific heat of air (J kg−1 K−1 ) p plants
cp specific heat of PCM (J kg−1 K−1 ) i inside air
ea outside water vapor pressure (kPa) w wall
ei inside water vapor pressure (kPa) EW east wall
e(T) saturated water vapor pressure at temperature T NW north wall
(kPa) SW sud wall
Fwc geometrical shape between the wall and the cover SR sud roof
hrc,w heat transfer coefficient due to long wave radi- NR north roof
ation between the cover and the wall radiation WW west wall
(W m−2 K−1 )
hcw,i convective heat transfer coefficient between the
inside air and the wall (W m−2 K−1 ) product received the reflected beam radiation (which otherwise
Ib beam radiation on a horizontal surface at any instant leaves through the north wall) in addition to the direct total solar
(W m−2 ) radiation available on the horizontal surface which increase the
Id diffuse radiation on a horizontal surface at any inside air and crop temperature [8]. Inclination angle of the reflec-
instant (W m−2 ) tive north wall can be optimized for a given width and height of a
Ii total incident solar radiation flux on ith inclined Sur- greenhouse, but it is also found that an inclined reflecting surface
face at any instant (W m−2 ) cannot reduce losses of solar radiation to zero [7]. Many studies on
k thermal conductivity of the PCM (W m−1 K−1 ) sensible thermal storage on massive north wall were reported. The
Ka thermal conductivity of the air (W m−1 K−1 ) concept of opaque north wall is commonly employed for east–west
Lc characteristic length of the cover (m) oriented greenhouse in northern hemisphere. North wall is there-
LAI leaf area index fore insulated externally and painted black internally for thermal
lf characteristic length of the leaf plants (m) storage. During the day time, incident solar radiations impinge on
Lw length of the wall (m) the wall and raise its thermal storage. This stored energy is released
L thick of the wall (m) during the night for thermal heating inside the greenhouse. The
N leakage rate (h−1 ) north wall of a 30 m2 glass covered greenhouse was used for heat-
Nu Nusselt number ing and storage of thermal energy [5]. The greenhouse was situated
Qc total solar radiation incident on the cover (W m−2 ) in Chateauroux (46.85◦ N, 1.72◦ E), France in which tomatoes were
Qp total solar radiation incident on the plants (W m−2 ) grown. The east and west sides were insulated and a north wall
Qw total solar radiation incident on the wall (W m−2 ) of 60 cm thickness was constructed. The system was able to meet
r reflection coefficient of the ground (≈0.2) 82% annual heating needs of the greenhouse. In another appli-
St total solar radiation falling on the Greenhouse at cation, north wall was constructed with stones inside a 350 m2
each wall and roof (W m−2 ) glasshouse situated at Atalia, used for vegetable growing [9]. The
T wall temperature (◦ C) greenhouse was able to maintain 1–2 ◦ C higher inside air tem-
Tw interior face wall temperature (◦ C) perature than the minimum outdoor air temperature. The north
Tm melting temperature of the PCM (◦ C) wall of a 20 m2 polyethylene (PE) covered greenhouses was con-
Ti inside air temperature of the greenhouse (◦ C) structed with 60 cm wide concrete blocks [5]. The wall was used
Ta outside air temperature (◦ C) as a heat storage unit during day and supplied heat during the
Tc cover temperature (◦ C) night by convection and radiation. These greenhouses were pro-
Tp plants temperature (◦ C) moted for raising vegetables under extreme climatic conditions.
Va outside wind speed (m/s) The greenhouses were maintained at 15–20 ◦ C during the winter
V greenhouse air volume (m3 ) conditions when outside air temperature was less than 10 ◦ C. The
impact of north storage wall was studied on the inside air tem-
Greek symbols peratures of three greenhouses measuring 100 m2 PE covered, at
˛sc , sc ,  sc absorptivity, reflectivity and transmissivity of the Athens (37.90◦ N, 23.70◦ E), Greece, 2000 m2 PE covered at Athens,
cover to solar radiation Greece and 1000 m2 glass covered at Chania, Greece [10]. These
˛sw , sw absorptivity, reflectivity of the wall to solar radia- systems satisfied about 35–50% heating needs of the greenhouses.
tion In this works thermal energy was stored in the wall as sensible
˛sp , sp absorptivity, reflectivity of the plants to solar radia- heat, the performance of the wall is affected by the thickness and
tion media used for heat storage. However, the optimum structural
 Stefan–Boltzman constant (5.67 × 10−8 W m−2 K−4 ) thickness of the wall attaint 60 cm in several works [5]. Fang and
 psychometric constant (≈0.0667 kPa K−1 ) Li [11] found that the optimum thickness of the wall for solar
 PCM density (kg m−3 ) passive heated building attain 45 cm. The sensible storage wall
a air density (kg m−3 ) required high volume and temperature difference. The latent ther-
εp emission coefficient for thermal radiation of the mal storage has many advantages over the sensible one: high heat
plants capacity, less volume, low storage temperature, thermal energy is
stored and released at an almost constant temperature. There are
a large numbers of PCMs that melt and solidify at a wide range of
 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035 3029

temperatures, making them attractive in a number of applications,

especially in greenhouse. Organic and inorganic compounds are the
two most common groups of PCMs. Most organic PCMs such as
paraffin waxes are chemically stable and non-corrosives, are com-
patible with most building materials, have a high latent heat per
unit weight and are recyclable. Their disadvantages are low ther-
mal conductivity, high changes in volume during phase change
and flammability. Inorganic compounds such as salt hydrates have
much higher latent heat per unit volume, higher thermal con-
ductivity, are non flammable and lower in cost in comparison to
organic compounds; they are corrosive to most metals and suf-
fer from decomposition and super cooling, which can affect their
phase change properties. Nucleating and thickening agents can be
added to inorganic PCMs to minimize super cooling and decompo-
sition. Use of phase change material capsules assembled as a packed
bed is one of the important methods that have been proposed to
achieve the objective of high storage density with higher efficiency
[12]. The most frequently used PCM in greenhouse are CaCl2 ·6H2 O,
Na2 SO4 ·10H2 O and paraffin. Kern and Aldrich [13] used CaCl2 ·6H2 O
as a PCM in aerosol cans to investigate heat storage possibilities in a
36 m2 ground area greenhouse covered with fiberglass. They found
that phase change storage system can provide a desirable alter-
native to rock storage. In an experimental study by Boulard and Fig. 1. Energy balance of greenhouse components.
Baille [14], CaCl2 ·6H2 O was utilized in a greenhouse with 176 m2
ground areas, double polycarbonate-cover and forced ventilation,
CaCl2 ·6H2 O (2970 kg) was packed in containers and placed along
the north wall. The PCM could provide 30% night heating needs The last term takes into account the latent energy associated with
during the winter period. Ozturk et al. [15] performed an exper- phase change when they occur. The latent heat of melting is Hm ,
imental evaluation of energy and exergy efficiency of a seasonal f represents the liquid fraction of melt and it is given by:
latent heat storage system for greenhouse heating using paraffin 
f =1 if T < Tm
wax as a PCM in a 180 m2 greenhouse ground areas. The system f =0 if T > Tm (2)
consists of five units: (1) flat plate solar air collectors (as heat col- 0<f<1 if T = Tm
lection unit 27 m2 ), (2) latent heat storage unit (6000 kg of paraffin
wax), (3) experimental greenhouse, (4) heat transfer unit and (5) Tm is the melting temperature of the PCM.
data acquisition unit. It was observed that the average net energy Boundary conditions in the wall are:
efficiency of the arrangement was 40.4%. In this study we purpose
the inclusion of phase change material in north wall of the green- • x = 0, interior surface of the PCM wall (facing the greenhouse as
house and analyze the impact of PCM north wall (PCM NW) on shown in Fig. 1:
thermal performance of a greenhouse. This system was already s c r r cond
Qw + Qi,w + Qp,w + Qc,w = Qw (3)
used in building by Sharma et al. [16] and he can reduce both cool-
ing and heating demands. Bourdeau [17] tested two passive storage where Qw s (W m−2 ) is the solar radiation absorbed by the PCM
collector walls using CaCl2 ·6H2 O (melting point 29 ◦ C) as a phase c
wall; Qi,w (W m−2 ) is the convective heat transfer from the
change material, he concluded that an 8.1 cm PCM wall has slightly r
inside air to the wall; Qp,w (W m−2 ) and Qc,w
r (W m−2 ) are the
better thermal performance than a 40 cm thick masonry wall.
net thermal radiation between the wall and respectively the cover
and the plants; Qw cond (W m−2 ) is the conduct if heat transfer
2. Mathematical model through the PCM wall. Eq. (3) can be written as:
Aw Aw
2.1. PCM north wall ˛sw Qw + hci,w (T − Tw ) + hrp,w (Tp − Tw )
Ag i Ag

The north wall of the greenhouse is insulated externally and Aw ∂T  Aw
+ hrc,w (Tc − Tw ) = −k (4)
∂x 
internally received a fraction of total solar radiation transmitted to Ag Ag
the greenhouse and has a radiative and convective heat exchange x=0

with the components of the greenhouse. Al the fluxes of heat are expressed per m2 of ground area of the
Following assumptions were made: greenhouse.
• x = L, exterior surface of the wall:
• PCM is homogeneous and isotropic. 
∂T 
• The mode of heat transfer is conduction only, the convection is = 0 (5)
∂x 
neglected (encapsulated PCM). x=L
• Heat transfer is one-dimensional.
Thermo-physical properties of the PCM are evaluated as:

km = fkl + (1 − f )ks , (cp )m = f (cp )l + (1 − f )(cp )s (6)

Using an enthalpy method, the energy equation in the wall reads
as: For selecting the PCM for the north wall, the following desir-
  able’s properties of calcium chloride hexahydrate were taken
∂T ∂ ∂T ∂f
cp = k −  Hm (1) into account: high latent fusion per unit mass, chemical stabil-
∂t ∂x ∂x ∂t
ity, melting in the desired operating temperature range, small
3030 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035

volume change during phase transition, availability in large quan- relating accurately saturated water vapor pressure to temperature
tities and low price. The north wall was filled with 777.6 kg for the temperature range between 0 and 60 ◦ C [19]:
of calcium chloride hexahydrate, corresponding to 32.4 kg of  
17.27Tp
PCM/m2 of the greenhouse ground surface. e∗ (Tp ) = 0.6108 exp (12)
Tp + 237.3

2.2. Greenhouse components The aerodynamic resistance ra (s/m) of the plants is deduced from
the characteristic of air flow in the greenhouse and the leaf length.
To mathematically describe the thermal behavior of the green- It is expressed by Haxaire [20] as:
house, three components that play important roles in the thermal
lf a ca
balance: the cover, the inside air and the plants are analyzed (Fig. 1). ra = (13)
Nu Ka
The following assumptions are made:
The stomatal resistance rs (s/m) of the plants is derived from a sim-
ple empirical relationship [21] with global radiation (in the first
- Relative humidity and temperature of inside air are supposed
approximation, the thermal and humidity dependences of green-
uniforms.
house plants transpiration were neglected):
- Heat capacity of the enclosed air and the cover are neglected.
- Climate date are hourly invariable. rs = 200 × [1 + exp(0.05(50 − QP ))] (14)
- No radiant heat is absorbed by the ground, which is fully covered
by vegetation: heat and humidity exchanged between the ground 2.2.3. Inside air
and the inside air is neglected. 2.2.3.1. Sensible heat. The sensible heat balance of the air is
- No condensation within the greenhouse. described by the following equation:
- There is no water-stress situation in the plants. c c c c
Qp,i + Qc,i + Qw,i + Qa,i =0 (15)

2.2.1. Greenhouse cover where Qp,i c , Q c and Q c are the convective heat transfer between
c,i w,i
The energy balance of the cover is given by: the inside air and respectively the plants, the cover and the PCM
c represent the sensible heat transfer between the inside
wall. Qa,i
Qcs + Qa,c
c c
+ Qi,c r
+ Qsky,c r
+ Qp,c r
+ Qw,c =0 (7)
air and the outside air due to leakage or ventilation. Eq. (15) can be
written as:
where Qcs (W m−2 ) is the solar radiation absorbed by the cover;
c
Qa,c (W m−2 ) and Qi,cc (W m−2 ) are the convective heat transfer Ap Ac Aw
hcp,i (Tp − Ti ) + hcc,i (Tc − Ti ) + hcw,i (Tw − Ti )
between the cover and respectively the outside and inside air of the Ag Ag Ag
r
greenhouse. Qsky,c (W m−2 ), Qp,c
r (W m−2 ) and Qw,cr (W m−2 ) + hca,i (Ta − Ti ) = 0 (16)
are the net thermal radiation between the cover and respectively
where hca,i
is the sensible heat transfer coefficient between the
the sky, the plants and the PCM NW. Eq. (7) can be written as:
inside air and the outside air, calculated using the formula of Fuchs
Ac Ac Ac et al. [22]:
˛sc Qc + hca,c (Ta − Tc ) + hci,c (Ti − Tc ) + hrsky,c (Tsky − Tc )
Ag Ag Ag a ca × N × V
hca,i = (17)
Ap Aw 3600 × Ag
+ hrp,c (Tp − Tc ) + hrw,c (Tw − Tc ) = 0 (8)
Ag Ag where N is the leakage rate expressed in number of air volume
renewals per hour (h−1 ). In this study, we used the relationship
where Tsky is the sky temperature (◦ C) which was calculated by
between N and Va founded experimentally by means of the tracer
Swinbank [18]:
gas technique for a low cost plastic greenhouse in a Mediterranean
Tsky = 0.0552 × (Ta + 273.16)1.5 − 273.16 (9) area, this expression is given by Baille at al. [23]:
N = 0.29 × Va + 0.76 (18)
2.2.2. Greenhouse plants
The energy balance of the plants surface in steady state condi- 2.2.3.2. Latent heat. In non-condensation conditions (i.e. transpi-
tions is: ration is the only source of water vapor) the latent heat balance of
the greenhouse air is described by the following equation:
Qps + Qc,p
r r
+ Qw,p c
= Qp,i l
+ Qp,i (10)
l l
Qp,i = Qi,a (19)
−2
where Qps (W m ) is the solar radiation absorbed by the plants; l
where Qp,i (W m−2 ) is the latent heat transfer between the plants
r
Qc,p (W m−2 ) and Qw,pr (W m−2 ) are the net thermal radiation l
and the inside air; Qa,i (W m−2 ) represent the latent heat trans-
between the plants and respectively the cover and the PCM NW;
c
Qp,i (W m−2 ) is the sensible heat transfer between the plants and fer between the inside air and the outside air due to leakage or
ventilation. Eq. (19) can be written as:
l
the inside air of greenhouse; Qp,i (W m−2 ) is the latent heat trans-
fer between the plants and the inside air of greenhouse. Eq. (10) a ca × LAI [e∗ (Tp ) − ei ] hci,a
= (ei − ea ) (20)
can be written as:  ra + rs 

Ac Aw 2.2.4. Solar radiations computations

˛sp Qp + hrc,p (Tc − Tp ) + hrw,p (Tw − Tp )
Ag Ag Total solar radiation available on the greenhouse cover is com-
Tp − Ti a ca × LAI [e∗ (Tp ) − ei ] puted for each wall and roof of the greenhouse using Lui and Jordan
= a ca LAI + (11) formula [24] for inclined surface:
ra  ra + rs

The saturated water vapor pressure e*(Tp ) at the temperature of the St = Ai Ii (21)
plants is commonly calculated by the following empirical formula i
 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035 3031

with Table 1
    Parameters used for the computation.
cos( ) 1 + cos(ˇ)
Ii = Ib + Id Parameters Values
sin(h) 2
  Greenhouse
V (m3 ) 57.6
1 − cos(ˇ)
+r (Ib + Id ) (22) Ag (m2 ) 4×6
2 Plants
LAI 3
where i is EW, WW, NW, SW, SR, NR. lf (m) 0.03
is the zenith angle of sun on an inclined surface; h is the alti- εsp 0.3
tude angle of sun with vertical; ˇ is the slope of the surface with sp 0.7
εp 1
horizontal.
Ap LAI × Ag
Total solar fraction F is the ratio of total solar radiation loss Cover
from the greenhouse to total solar radiation transmitted inside the  sp 0.9
greenhouse [4]. The value of F will be more during winter months sp 0.05
because of the low altitude angle of the sun and hence the losses ˛sc 0.05
εc 0.5
will be more. A model has been developed to calculate “total solar
Ac (m2 ) 59.07
fraction”, which quantifies the losses of all transmitted solar radi- PCM (CaCl2 ·6H2 O) [30]
ation coming inside the greenhouse. These losses take place from Tm (◦ C) 29
north wall/roof, and east/west wall for east–west oriented green- kl , ks (W m−1 K−1 ) 0.539, 1.088
house. The fraction falling on the north wall Fn is the most important 1 , s (kg/m3 ) 1560, 1800
cp1 , cps (J kg−1 K−1 ) 2130, 1460
compared to the other face [4], which motive this study and gives Hm (kJ/kg) 187.49
the idea of storing this energy in a MCP NW. ˛sw 0.9
Solar radiations transmitted into the greenhouse are assumed sw 0.1
to be reflected multiply between the greenhouse plants, cover and ˛w 0.99
L (cm) 4
PCM NW. However the total solar radiations incidents on plants,
Aw (m) 1.8 × 6
cover and PCM north wall are given approximately by [25]: Price ($/kg) 2
Toxic effect No
Qp = sc St [(1 − F)(1 + sc sp ) + sc (F − Fn ) + w Fn ] (23)
1

Qc = sc St (1 − F)(1 + sp sc ) + (F − FNW )(1 + sp sc ) + (24) resolve this problem iteratively, the above terms must be con-
sc
sidered constant within each time step. The new value of the
Qw = Fn sc St (25)
obtained temperature Ti and Tj will be used to substitute the

2.2.5. Convective heat transfer coefficients term (Ti + Tj )(Ti2 + Tj2 ) by the term (Ti + Tj )(T  2i + T  2j ) and the term
Empirical equations that were used to calculate the convective |Ti − Tj |a by the term |Ti − Tj |a . The process continues until conver-
coefficients are [26]: gence reached. Note that, the energy balance plants equation was
not linearized because of the presence of the exponential term of
hca,c = 0.95 + 6.76Va0.49 with Va ≤ 6.3 m/s (26)
saturated water vapor pressure. Its numerical solution is performed
Nu Ka Nu Ka Nu Ka by an iterative Newton–Raphson procedure. A finite-volume code
hci,c = , hci,p = , hci,w = (27)
Lc lf Lw has been developed in order to determine the temperature and
the liquid fraction in the PCM wall. Inputs for the model include
Correlation developed by Monteith [27] and Campbell [28] based
solar radiations on horizontal surface (global, diffuse), outside air
on the convection regime and the type of air flow inside the green-
temperature, relative humidity and wind speed. Parameters that
house are used for the computation of the Nusselt number.
need to be specified include the greenhouse configuration (cover
materiel, dimensions), the plants characteristics and the thermo
2.2.6. Radiative heat transfer coefficients
physical properties of the PCM as given in Table 1.
The radiative heat transfer coefficients are given by:

hrsky,c = εc (Tsky

2
+ Tc2 )(Tsky + Tc ) (28) 2.3. Validation of the numerical model
 1 −1
1 1 Ap Numerical code was validated by comparing it with the experi-
hrp,c = (Tp2 + Tc2 )(Tp + Tc ) + −1+ −1 (29)
Fpc εp Ac εc mental results performed by Sethi and Sharma [31]. Experimental
measurements of plants and inside air temperatures per hour are
1 1 Aw
1 
−1
made in evenspan greenhouse with plastic cover and introducing
hrw,c = (Tw
2
+ Tc2 )(Tw + Tc ) + −1+ −1 (30)
Fwc εw Ac εc the same orientation and dimensions as that adopted in the present
model. Using similar parameters of entrance adopted in the experi-
  −1 ence of [31], Fig. 2 illustrates the hourly variation of the measured
1 1 Aw 1
hrw,p = 2
(Tw + Tp2 )(Tw + Tp ) + −1+ −1 and calculated plants and inside air temperature. Results show a
Fwp εw Ap εp
good concordance between both models with difference average
(31)
of 3% for the temperature of plants and 4% for the temperature of
The geometrical shape factors are calculated using formula in [29]. inside air. This allows concluding that our model allows simulating
climatic parameters correctly inside the greenhouse.
2.2.7. Numerical computation and used parameters
The obtained equations are non linear because the radiative 3. Results and discussions
transfer coefficient can appear with the term (Ti + Tj )(Ti2 + Tj2 ) in
the radiative component and the convective transfer coefficient Fig. 3 shows the total solar radiation falling on the greenhouse
appear with the term |Ti − Tj |a in the convective component. To for a typical climate decade of January in Marrakesh (31.62◦ N,
3032 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035

Fig. 4. Hourly variation of the liquid fraction and the temperature of the north wall
Fig. 2. Hourly variation of the predicted and measured plants and inside air tem- for a typical climate day of January in Marrakesh.
peratures.

and leads to a rapid melting of PCM and liquid fraction reaches

8.03◦ W). The figure illustrates also the solar radiation fraction
1 at 13 h, but solar radiation persists after 13 h, what leads to an
transmitted in the greenhouse and incident on the north wall. For a
overheating of the liquid PCM layer (Tw = 54 ◦ C at 17 h). After 17 h,
conventional greenhouse (without PCM NW), these fraction leaves
solar radiation disappears, outside air temperature decrease and
the greenhouse because of the high solar radiations transmission
consequently the inside greenhouse temperature decrease. Heat
of a plastic cover. Also these fraction depend on the time of the day,
transfer changes the direction from the wall (hotter medium) to
solar altitude angle, nth day of the year and solar azimuth angle
the inside greenhouse (cold medium). Liquid PCM and wall losses
[4]. So this fraction is high in winter months due to the low alti-
their heat but liquid fraction remains unchanged (f = 1) because the
tude angle of the sun. For a greenhouse with PCM NW as a medium
liquid PCM was superheated (sensible heat storage). After about
of storage, this energy will be stored during the day time and will
20 h liquid fraction begins to decrease (solidification or discharg-
be released during the night for thermal heating inside the green-
ing period) and we remarks that the wall temperature remains
house. Also the isolation of the external face of the north wall can
relatively constants (latent heat of fusion) until the morning (9 h).
reduce significantly heat losses through the cover.
It was observed that during the night, wall temperature never
Fig. 4 presents the hourly variation of the liquid fraction and the
falls down (27 ◦ C) when the outside air temperature drop until 5 ◦ C.
north wall temperature for the second day from the typical climate
The north wall incorporating PCM plays the role of a heat sink dur-
decade of January in Marrakesh. Analysis for such figure shows that
ing the day and a heat source during the night. The energy stored
liquid fraction increase during the morning (from 9 h to 11 h) due
during the day is released at night. Approximately, the same trend
to the increasing incident solar radiations and the wall tempera-
was observed during the other days of the typical climate decade
ture remains overly unchanged, because all the energy transmitted
of January in Marrakesh (Fig. 5) which leads to a passive heating of
to the wall is completely used to melt the PCM as a latent heat
the greenhouse.
of fusion. After these period (from 11 h to 13 h), wall temperature
The hourly variation of temperature for outside air, plants, cover
start increasing due to the activation of the sensible heat storage
and inside air of greenhouse with and without PCM NW for the

Fig. 3. Hourly variation of the total incident solar radiation on the greenhouse and Fig. 5. Hourly variation of the liquid fraction and the temperature of the north wall
on the north wall for a typical climate decade of January in Marrakesh. for the typical decade climate of January in Marrakesh.
 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035 3033

Fig. 6. Hourly variation of the outside air temperature, plants, cover and inside Fig. 7. Hourly variation of the outside and the inside relative humidity for green-
air temperature with and without PCM NW for a typical climate day of January in house with and without PCM NW for typical climate day of January in Marrakesh.
Marrakesh.

second day from the typical decade climate of January in Marrakesh

have been presented in Fig. 5. From the figure, for a conventional
greenhouse (without PCM NW), it is seen that during the day, the
temperature of plants, cover and inside air are higher than the out-
side air temperature due to the solar radiations absorbed by the
plants and dissipate as sensible heat and latent heat inside the
greenhouse. But during the night the picture is inverted and the out-
side air is warmer than the inside greenhouse due to the nocturnal
losses of polyethylene covered greenhouse by leakage, convection
and radiation during winter periods [23].
The diurnal variation of temperatures for greenhouse with PCM
NW was observed to be 1–2 ◦ C higher from 9 h to 13 h (Fig. 6). This
due to the fraction of solar radiations reflected from the north wall
and incident on the plants (w = 0.1). But from 14 to 20 h, there
occurred more difference in temperature 5–12 ◦ C for plants and
inside air and 2–5 ◦ C for cover. This remarkable temperature dif-
ference is due to the overheating of the PCM NW as was mentioned
above. The increased temperature difference between the wall and
the greenhouse components (plants, inside air and cover) creates
Fig. 8. Performances of the greenhouse components with different thicknesses of
an important convective and radiative heat transfer. From 21 to 9 h,
PCM north wall.
the nocturnal variation of temperature for the greenhouse with
the PCM NW were observed to be 3–8 ◦ C higher for plants and
inside air, and 1–4 ◦ C for cover. This period corresponds to the dis- greenhouse. The PCM system is able to create a passive dehumidifi-
charge process of the PCM (solidification) and as it was mentioned cation process especially at night time due to the increase in inside
above the wall temperature remains approximately unchanged air temperature.
(Tm = 29 ◦ C). The wall can be considered as a heat source and the Fig. 8 shows the time wise variation of the inside greenhouse
heat is transferred to the greenhouse components (plants, inside air and wall temperatures and liquid fraction for various PCM NW
air and cover) by convection and radiation. With the use of the thicknesses. The performances of the greenhouse components for
PCM NW, the greenhouse air temperature was maintained 3–4 ◦ C the colder and the hotter day from the typical decade climate of
higher as compared to outside air during nighttime, this create a January in Marrakesh are presented in Table 2. For the case with
healthy environment of plants during winter period. 2 cm of PCM, liquid fraction decreases to 0.03 (≈0) for the colder
Fig. 7 shows the effect of the PCM NW on relative humid- night in January as shown in Table 2. This thickness is considered
ity. The greenhouse air relative humidity was maintained 10–14% not safe, because it is practical to do not solidify all the PCM for the
lower as compared to the relative humidity in conventional very cold periods. The target was how to maintain a sustainable

Table 2
Performances of the greenhouse components with different thicknesses of PCM North wall.

2 cm 4 cm 5 cm

Cold day Warm day Cold day Warm day Cold day Warm day

None melted PCM 3% 6% 50% 57% 64% 68%

Tw,max (day) 49 ◦ C 67 ◦ C 42 ◦ C 57 ◦ C 41.7 ◦ C 54 ◦ C
Ti ,max (night) 12.33 ◦ C 11.11 ◦ C 12.4 ◦ C 11.25 ◦ C 12.4 ◦ C 11.33 ◦ C
3034 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035

Fig. 9. Daily variations of thermal load leveling, TLL, for the typical decade climate of January in Marrakesh.

heat source for the greenhouse. Also, for this value of the PCM NW The performance of PCM NW as a passive heating system for the
thickness (2 cm), it was observed that, during the hotter day, wall greenhouse has been evaluated in terms of thermal load leveling,
temperature exceeds 67 ◦ C, what can damage the plants located TLL, and heating potential, Qh , using the following equations [32]:
near the wall (Table 2). For the other values of the PCM thickness, ⎧
⎪ Ti,max − Ti,min
liquid fraction never reaches 0, this lead to establish the wall tem- ⎪
⎪ TLL =
perature to the value of (Tm = 29 ◦ C) and the same effect of the PCM
⎨ Ti,max + Ti,min
 24 (32)
thickness on the wall temperature was observed for the three other ⎪
⎪
values of the PCM thickness. For the higher thicknesses (L > 4 cm), ⎪
⎩ Qh = ma ca (Ti (t) − Ta (t))
the liquid fraction varies between 1 and 0.68. So only 32% of the t=1
daily heat stored in the wall was used and 68% of the PCM layer Thermal load leveling, TLL, is a ratio even used to quantify the
seems to be not necessary. This lead to conclude that incorporating fluctuations of temperature inside the greenhouse. The less the
a North wall having 4 cm of PCM thickness is the best and practi- fluctuations, the better is the environment for plants inside the
cal choice for heating greenhouse located in Marrakesh (31.62◦ N, greenhouse. In winter, TLL should have lower value by incorpo-
8.03◦ W). rating heating method due to the increase of (Ti,max + Ti,min ) as well

Fig. 10. Daily variations of total heating potential, Qh , for the typical decade climate of January in Marrakesh.
 F. Berroug et al. / Energy and Buildings 43 (2011) 3027–3035
as decrease of (Ti,max − Ti,min ) as compared to TLL without heating
arrangement for optimal environment for growth and develop-
ment of plants. The results for daily variation of thermal load
leveling for greenhouse with and without PCM NW have been
shown in Fig. 9. It can be seen that the value of TLL is maximum
for greenhouse without PCM NW and it is reduced on average
about half for greenhouse with PCM NW. The lower values of
thermal load leveling indicate the decrease in the fluctuations
of greenhouse air and thereby, there occurs an improvement of
desired environment for plants in the greenhouse. Similarly, the
daily variations of total heating potential, Qh (Eq. (32)) obtained
from PCM NW for typical climate decade (Marrakesh – January)
were calculated and have been shown in Fig. 10. From the results,
it is seen that the heating potentials obtained in greenhouse
with PCM NW were higher as compared to greenhouse without
PCM NW.
4. Conclusion
For east–west oriented greenhouse, maximum solar radiations
falls on the south wall during winter months and a fraction of this
solar radiations leaves the greenhouse through north wall. There-
fore, a phase change material north wall is proposed in this study
for absorption and reflectance of solar radiations. During the day
time, incident solar radiations on the wall raise its thermal storage.
This stored energy is realized to the greenhouse by convection and
radiation. The performance of the system was evaluated in terms of
thermal load leveling and potential heating for a typical decade cli-
mate of January in Marrakesh From the present study, the following
conclusions can be drawn:
- There occurs a 6–12 ◦ C rise of plants and inside air temperature
and 4–5 ◦ C for cover temperature at night time due to the use of
a 4 cm thick PCM NW as a storage medium.
- Fluctuation of temperature for greenhouse air is less in a green-
house with PCM NW.
- Relative humidity is 10–15% lower in a greenhouse with PCM NW.
- PCM Thickness of 4 cm is practical and sufficient for heating
greenhouse located in Marrakesh (31.62◦ N, 8.03◦ W).
Acknowledgements
The present work was accomplished with the financial support
of the CNRST as part of Program URAC, Convention URAC28.
References
[1] M. Santamouris, G. Mihalakakou, C.A. Balaras, J.O. Lewis, M. Vallindras, A. Argiri-
ous, Energy conservation in greenhouses with buried pipes, Energy 21 (5)
(1996) 353–360.
[2] J.R. Barral, P.D. Galimberti, A. Barone, A.L. Miguel, Integrated thermal improve-
ments for greenhouse cultivation in the central part of Argentina, Solar Energy
67 (1–3) (1999) 111–118.
[3] M.N. Bargach, R. Tadili, A.S. Dahman, M. Boukallouch, Survey of
agricultural greenhouses in Morocco, Renewable Energy 20 (2000)
415–433.
[4] G.N. Tiwari, G. Amita, G. Ravi, Evaluation of solar fraction on north partition
wall for various shapes of solarium by Auto-CAD, Energy and Building 1506
(2002) 1–8.
3035
[5] M. Santamouris, C.A. Balaras, E. Dascalaki, M. Vallindras, Passive solar agricul-
tural greenhouses: a worldwide classification and evaluation of technologies
and systems used for heating purposes, Solar Energy 53 (5) (1994) 411–426.
[6] T.K. Hartz, J.A. Lewis, H.A. Hughes, Performance of modified brace institute
greenhouse in Virginia, Horticulture Science 16 (1981) 74–78.
[7] R. Gupta, G.N. Tiwari, Modeling of energy distribution inside greenhouse using
concept of solar fraction with and without reflecting surface on north wall,
Building and Environment 40 (2005) 63–71.
[8] V.P. Sethi, S. Arora, Improvement in greenhouse solar drying using inclined
north wall refection, Solar Energy 83 (2009) 1472–1484.
[9] H. Sallanbus, E. Durceylan, K. Yelboga, Utilization of solar energy in greenhouse,
in: C. von Zabeltitz (Ed.), Greenhouse Heating with Solar Energy. REU Technical
Series 1, FAO, ENEA, Roma, 1987, pp. 152–158.
[10] M. Santamouris, Active solar agricultural greenhouse. The state of art, Interna-
tional Journal of Solar Energy 14 (1993) 19–32.
[11] X. Fang, Y. Li, Numerical simulation and sensitivity of lattice passive solar heat-
ing wall, Solar Energy 69 (1) (2000) 55–66.
[12] X. Wang, J. Niu, Y. Li, X. Wang, B. Chen, R. Zeng, Q. Song, Y. Zhang, Flow and heat
transfer behaviors of phase change material slurries in a horizontal circular
tube, Heat and Mass Transfer 50 (2007) 2480–2491.
[13] M. Kern, R.A. Aldrich, Phase Change Energy Storage in a Greenhouse Solar
Heating System, ASAE Paper No. 79-4028, Am. Soc. Agric. Eng., St. Joseph, MI,
1979.
[14] T. Boulard, J. Baille, Thermal performance and model of two types of green-
houses with solar energy storage, Acta Horticulture 263 (1987) 121–130.
[15] H.H. Ozturk, Experimental evaluation of energy and exergy efficiency of a sea-
sonal latent heat storage system for greenhouse heating, Energy Conversion
and Management 46 (2005) 1523–1542.
[16] A. Sharma, V.V. Tyagi, C.R. Chen, D. Buddhi, Review on thermal energy stor-
age with phase change materials and applications, Renewable and Sustainable
Energy Reviews 13 (2009) 318–345.
[17] L.E. Bourdeau, Study of two passive solar systems containing phase change
materials for thermal storage, in: J. Hayes, R. Snyder (Eds.), Proceedings of the
Fifth National Passive Solar Conference, Amherst, October 19–26, American
Solar Energy Society, Newark, DE, 1980, pp. 297–301.
[18] W.C. Swinbank, Long-wave radiation from clearskies, Quarterly Journal of the
Royal Meteorological Society 81 (89) (1963) 339–348.
[19] O. Tetens, Ueber einige meteorologische Begriffe, Zeitschrift fur Geophysik 6
(1930) 297–309.
[20] R. Haxaire, Caractérisation et modélisation des écoulements d’air dans une
serre, Thesis, Université de Nice Sophia Antipolis, 1999.
[21] T.T. Boulard, A. Baille, M. Mermier, F. Villette, Mesures et modelisation de la
resistance stomatique foliaire et de la transpiration d’un couvert de tomates de
serre (Measurement and modelling of stomatal resistance and tomato transpi-
ration in greenhouse), Agronomie 11 (1991) 259–274.
[22] M. Fuchs, E. Dayan, E. Presnov, Evaporative cooling of a ventilated greenhouse
rose crop, Agricultural and Forest Meteorology 138 (2006) 203–215.
[23] A. Baille, J.C. Lopez, S. Bonachela, J.I. Montero, Night energy balance in a heated
low-cost plastic greenhouse, Journal of Agricultural and Forest Meteorology
137 (2006) 107–118.
[24] B.Y.H. Liu, R.C. Jordan, The interrelationship and characteristic distribution of
direct, diffuse and total solar radiation, Solar Energy 4 (1960) 1–9.
[25] A. Abdel-Ghany, T. Kozai, On the determination of the overall heat transmission
coefficient and soil heat flux for a fog cooled, naturally ventilated greenhouse:
analysis of radiation and convection heat transfer, Energy Conversion and Man-
agement 47 (2006) 2612–2628.
[26] G. Papadakis, A. Frangoudakis, S. Kyritsis, Mixed, forced and free convection
heat transfer at the greenhouse cover, Agricultural Engineering Research 51
(1992) 191–205.
[27] J.L. Montheith, Principles of Environment Physics, Edward Arnold, 1973 (New-
Work, Edition).
[28] G.S. Campell, An Introduction to Environment Biophysics, Springer-Verlag, New
York, 1977 (Edition).
[29] F.P. Incropera, D.P. DeWitt, Fundamentals of Heat Transfer, John Wiley & Sons,
1981.
[30] H. Benli, A. Durmus, Performance analysis of a latent heat storage system with
phase change material for new designed solar collectors in greenhouse heating,
Solar Energy 83 (2009) 2109–2119.
[31] V.P. Sethi, S.K. Sharma, Thermal modelling of a greenhouse integrated to an
aquifer coupled cavity flow heat exchanger system, Solar energy 81 (2007)
723–741.
[32] R.D. Singh, G.N. Tiwari, Thermal heating of controlled environment greenhouse:
a transient analysis, Energy Conversion and Management 41 (2000) 505–522.