Climate assessment of greenhouse equipped with south oriented PV roofs an experimental and computational fluid dynamics study

The climate behavior during summer and winter days inside a greenhouse integrated with PV panels on the roof and a reference was assessed.. A comparative study of PV shading effects on t

Available online 9 March 2021

2213-1388/© 2021 Elsevier Ltd All rights reserved

Climate assessment of greenhouse equipped with south-oriented PV roofs:

An experimental and computational fluid dynamics study

Hela Ben Amaraa,*, Salwa Bouadilaa, Hicham Fatnassib,d, Müslüm Aricic, Amen Allah Guizania

aLaboratory of Thermal Processes, Research and Technology Center of Energy, Borj Cedria, 2050 Hammam-Lif, Tunisia

bINRAE, Univ Nice Sophia Antipolis, CNRS, UMR 1355-7254 Institut Sophia Agrobiotech, 06900 Sophia Antipolis, France

cKocaeli University, Engineering Faculty, Mechanical Engineering Department, Umuttepe Campus, 4001 Kocaeli, Turkey

dInternational Center for Biosaline Agriculture, ICBA, Dubai, P.O Box 14660, United Arab Emirates

A R T I C L E I N F O

Keywords:

Renewable energy

Photovoltaic panels

Greenhouse

Solar radiation

Shading

Economic analysis

A B S T R A C T The present study was undertaken to better understand the effect of the shading induced by south oriented photovoltaic panels on the distributed climate and plant activities in a mono-span greenhouse using CFD tool The climate behavior during summer and winter days inside a greenhouse integrated with PV panels on the roof and a reference was assessed Solar radiation distribution, wind velocity, relative humidity, and ambient tem-perature from the two greenhouses were presented and energy production and plant activity parameters were evaluated The greenhouse equipped with photovoltaic panels provides more favorable climatic conditions during summer season Therefore, results of the paper can be useful for farmers in the Mediterranean area to contain the ability level of this innovative photovoltaic greenhouse crop and to perceive if it can perform a balance between yield crop and PV electricity production Results showed that PV panels can produce around 55 W/m2 for the cultivation period of January During a summer day, the solar radiation of the PV greenhouse was lower than that of the reference greenhouse and exceeded 115 W/m2, and the difference between the inside and the outside reached 220 W/m2 The annual solar energy of the photovoltaic panel was around 5054.4 kWh Accordingly, the shading plays a positive effect on plants in decreasing temperature due to the reduced thermal load of the sun inside the greenhouse In addition, the payback period of PV system was found to be less than 6 years

Introduction

The food security is among the challenges that humanity have much

interest According the International Renewable Energy Agency [1],

food production will need to increase by 60% and water availability by

55% until 2030 To ensure the food security, greenhouse is a potential

alternative to provide needed both energy and food production [2] In

fact, greenhouses cultivation, even if they are required to control the

microclimate of reason with cultivate the desired culture where all the

growth factors of the plants can be maintained on an optimal level, also

permitted a highly-quality production and technology [3,4]

Greenhouses become one of the largest scale of food production in

the agricultural industry [5] because of its ability to increase production

while around 405 thousand ha of greenhouses distributed on all the

zones Hence, the use of renewable energies in greenhouses in

Medi-terranean area was an interesting proposition for growers Solar energy

[6,7], wind energy [8] or geothermal energy [9] have been proposed for providing ideal growth conditions to enhance the greenhouse energy efficiency To limit global warming problems and environmental pollution in the future, innovative renewable energy sources, specif-ically solar energy based on photovoltaic (PV) technology can be used [10] The generation of photovoltaic energy is considered as an effective means, clean and valuable energy source for both indoor [11] and outdoor applications [6] Meanwhile, the exploitation of the solar photovoltaic presents a good solution to generate the electricity regarding the limitation of the traditional energy resources

The dependability of photovoltaic systems in agricultural in-frastructures resides in the use for waste water treatment plants, food and herb drying, producing electricity without carbon and preserving the agricultural resources [12] or its use for agricultural land by combining mobile photovoltaic panels and plants to balance energy and food crops production [13] Compared to ordinary greenhouses, photovoltaic greenhouses can provide higher fruit quality and yield,

* Corresponding author

E-mail address: benamarahela@hotmail.fr (H Ben Amara)

Contents lists available at ScienceDirect Sustainable Energy Technologies and Assessments

https://doi.org/10.1016/j.seta.2021.101100

Received 24 November 2020; Received in revised form 30 December 2020; Accepted 25 January 2021

greater opportunity, and better environmental sustainability of

green-house cultivation Photovoltaic greengreen-houses appear not only to be a

fascinating option to progress horticultural sector economy but also to

generate more energy and develop agricultural activities

In order to improve the management level of the greenhouse

photovoltaic, it is of great importance to study the performance of the

photovoltaic power system and data on both its potential and actual

output In fact, mounting PV modules on the greenhouse roof is an

interesting proposition for growers and it can meet the increasing

de-mand of energy of the agricultural industry and improved crop

pro-ductivity, simultaneously However, PV modules mounted on the roof

could decrease plants growth due to shading especially in winter season

For that reason, the relationship between orientation, position, and

generated energy of the PV cells fixed on the roof greenhouses must be

considered carefully For instance, the effect of the geometrical

ar-rangements of PV arrays can significantly affect the cultivated plants

growth [14] A comparative study of PV shading effects on the internal

microclimate and crop growth during the winter and summer period

under greenhouse that 40% roof area covering by PV has been reported

by [15] Moreover, distribution and variation of the shading percentage

were examined on the 21st day of every month, in order to perceive the

percentage of shading with the PV arrangement adopted during every

month of the year [16] In another work [17], a similar subject has been

applied by placing flexible photovoltaic panels on 10% of the roof area

of a Canarian greenhouse, confirming that the use of PV panels in a

checkerboard arrangement does not seem to produce a significant

impact on the agronomic parameters including quality-yield and height

tomato crop, and all others climatic parameters under the Agrivoltaic

Meanwhile, varying the degree of shading panels seems to be a good

solution to optimizing the energy production and agriculture In a

pre-vious paper [18], a dynamic PV greenhouse prototype with varying the

degree of shading according to the weather conditions was proposed in

order to achieve a better global daily radiation and energy flow

Furthermore, arid and semi-arid regions are unusually fitting for PV

electricity production because of their heat and cold weather A great

research effort has been made for energy sustainable greenhouses in

order to improve their electrical [19] and thermal [20] efficiency and the feasibility of PV technologies to supply the food production chain by affording regenerated electricity from sunlight

Marrou et al [21] showed that air temperature and wind speed were less affected than soil temperature by shading In addition, the impact of energy production and plant growing in the greenhouse shading with the PV panels are evaluated by [22] Theoretical results demonstrated that the PV panels can produce over 50 kWh/m2 for the characteristic agriculture period and can create around 20% of the greenhouse shading That study also indicated that the shading led to better results

on crop growth and yield compared to those of the greenhouse without

PV In particular, the large use of greenhouse roofs covered by PV modules requires an environmental load For that purpose, an economic analysis such as payback period [23], energy payback time [24], social performance [25] of greenhouses have been evaluated The estimation

of the photovoltaic energy need of a 150 m2 greenhouse was conducted

by [26].It was indicated that the photovoltaic greenhouse can meet 33–67.2% of electricity demand in summer The system energy payback time is five years while the crop cultivations payback time is ranged from 7.0 to 7.4 years and the annual PV generated electricity was re-ported as 21510.4 kWh Another study in Agrivoltaic systems was per-formed on tomato in South Eastern Spain, in which the yearly electricity production was 8.25 kWh m− 2 [27] For their side, Fatnassi et al [28] revealed in their CFD numerical study that the use of integrated PV system in greenhouse can decrease automatically the solar irradiation at the top and at the bottom, the values of which were respectively 138 W/

m2 and 20 W/m2 of the cover in summer conditions, and 70 W/m2 and

10 W/m2 of the cover in the winter conditions A similar numerical study was made by Marucci et al [29] to evaluate the electrical energy production of a photovoltaic greenhouse with different shadow rates in summer conditions Modeling tools were also used by Cossu et al [30] They proposed an algorithm which allows to calculate the solar radia-tion distriburadia-tion inside a commercial pitched roof PV greenhouse with East-West orientation The results of the calculation of the light distri-bution on different canopy height showed that the incident solar energy

on the crop changes steadily, depending on the stage of plant growth

Nomenclature

A ratio of the stomatal resistances of the upper and lower

sides of the leaf

a1,a2, b1,b2 coefficients for the determination of stomatal resistance

Cp specific heat, (J kg− 1K− 1)

D saturation deficit of the leaf, (Pa)

f1, f2 functions for the determination of stomatal resistance

ILAv crop stand leaf area index per volume unit

ILAs crop stand leaf area index per unit area

k turbulent kinetic energy, (m2s− 2)

kc extinction coefficient of radiation

Lν latent heat of water vaporization, (J kg− 1)

u0 friction velocity, (ms− 1)

Uh reference velocity, (ms− 1)

Uinl inlet velocity, (ms− 1)

U

→

interior air speed, (ms− 1)

RG global radiation outside the greenhouse, (Wm− 2)

R(z) solar radiation at z, (Wm− 2)

Rgi global radiation inside the greenhouse, (Wm− 2)

Rabs absorbed radiation, (Wm− 2)

rs stomatal resistance of the leaf, (s m− 1)

rsu stomatal resistance of the upper side of the leaf, (s m− 1)

rsl stomatal resistance of the lower side of the leaf, (s m− 1)

rt total resistance for the mass transfer, (s m− 1)

rs,min minimal stomatal resistance for tomatoes, (s m− 1)

Tcrop tomato crop temperature, (K)

Greek symbols

λ thermal conductivity, (W m− 1 K− 1)

v viscosity, (m2s− 1)

Abbreviation

Amaducci et al [31] developed an innovative platform that can be

employed both for predicting the energy and food crops production

under specific Agrivoltaic configurations and for enhancing the design

of the PV infrastructure Their proposed platform was developed by

coupling a radiation and shading model to the generic growth of maize

under Agrivoltaic system It has been shown that yield of crop was more

favorable than in conditions of full light, and the suitable shading under

Agrivoltaic as it can affect photosynthesis, soil temperature and

evapotranspiration, can provide better conditions for crop yield and

water/energy

In this sense, the purpose of our study is to consider the ability of

energy production and plant growing from photovoltaic integrated

greenhouse and to assess their influence on the internal microclimate

under different external climate conditions (winter and summer)

The aim of this study is to achieve a comparison between two types of

greenhouses with a PV roof cover ratio of 50% and a reference

green-house (without PV) by using CFD tools It assesses the shading effect of

panel photovoltaic on the microclimate and energy production

through two typical days of the summer and winter periods when the

intensity light is at these two extremes (maximum and minimum

dura-tion sun) The photovoltaic panels mounted in the straight-line format

on the greenhouse roof surface has specific advantages for farmers in the

Mediterranean area The novelty of this work is to develop sustainable

farming systems on their own farms and to understand the daily

per-formance of these system for large scale agricultural It represents

therefore a modern technology that can both improve the power of high-

tech agricultural production and perform a balance between yield crop

and photovoltaic electricity production

Experimental setup

Greenhouse description

The experimental greenhouse (Fig 1) is a mono-span greenhouse

with a polyethylene plastic cover (100 µm thick) It is located at the

Research Center of INRAE in Sophia Antipolis, France (latitude: 43◦56,

longitude: 7◦18, altitude:100 m) The greenhouse covers an area of

172.8 m2 (18 m length and 9.6 m width) with a maximum height reaches

6 m with Southern-Eastern orientation The ventilation of greenhouse

was achieved by two continuous roofs (1.5 m × 18 m each, total of 54

m2) with each protected by an insect proof Table 1 summarizes the

thermal and spectral properties used in this study

Experimental measurements

This section aimed to explore the internal microclimate of the

reference greenhouse (without PV) and the outside climatic conditions

on summer and winter days (January and June) The distribution of the micrometeorology was measured indoor/outdoor the greenhouse with

an array of sensors and instruments noted as follows:

• The 083E sensors are used to measure the variation of temperature and relative humidity inside the greenhouse at eight equidistant

positions The schematic layout illustrated in (Fig 2) describes these

positions of sensors distributed inside the greenhouse For both measurement periods, temperature and relative humidity were measured in the horizontal plane located at a height of 1.8 m above the ground level

• A meteorological station located at 7 m above the ground in the site

of Sophia Antipoils is used to provide external weather conditions

• The CR1000X data logger (Campell Scientific datalogger) is used to record the internal and external environmental parameters every 10 min and overall daytime and nighttime mean values were calculated for each time

• Analyzer datalogger is used to process and analyze the calculated parameters and measured data

Uncertainty analysis

During the experimental period, obtained measurements may contain several uncertainties An uncertainty analysis is then required to estimate the experiment measurements accuracies and to extend the accuracy performed experiments We note that the experimental error is defined as the difference between an experimental value and the actual value of a quantity and this difference designates the accuracy of the measurement

Calculation of experimental uncertainties came essentially from all available data, like sensor measurements used for the determining wind speed, solar radiation, temperature, and RH distribution inside and outside the greenhouse, and the sensitiveness and collects and stores data uncertainty More details of the sensitiveness of the used sensors and equipment devices required for greenhouse measurements are given

in Table 2

Climatic parameters measurements

The experimental measurements have been taken on two typical days: a clear sky day in winter (January, 10th) and on a clear sky day in summer (June, 21th) The hourly variation of weathers data is shown in Table 3 and Table 4 The solar irradiation, temperature and relative humidity in greenhouse were considered from 6:00 to 18:00

In winter, the indoor and the outdoor temperature difference was lower in the early hours of the day and at the end of afternoon (3–4 ◦C), especially during the central hours when the differential was quite high

Fig 1 Schematic view of the experimental greenhouse

(from 7 to 9 ◦C) due to the effect of greenhouse Greenhouse ventilation

is crucial in summer season because it is necessary to control the airflow

and internal climate which interact with the heat exchanges with the

crops It was observed that the peak value of relative humidity recorded

outdoor was about 99%, observed at dawn and in the late afternoon and

the maximum value of relative humidity recorded in the indoor

envi-ronment was noted 92%, whose average values ranged between 62%

and 85%

Mathematic modeling

Computational fluid dynamic model

Computational fluids dynamic (CFD) tools provide powerful means

for simulating the spatial distribution of the microclimatic conditions

within the entire greenhouses and evaluating the climate variables and

their action in crop progress In this paper, a three-dimensional CFD

greenhouse model consisting continuity, momentum and energy

equa-tions (Eq (1)) was built

⎧

⎪

⎪

⎪

⎪

⎪

⎪

∂ρ

∂ t+

∂(ρ u j)

∂ x i

=0

∂

∂ t(ρ u i) +∂(ρ u i u j)

∂ x j

∂ x i

[

− ρ δ ij+μ

(

∂ u i

∂ x j

+∂ u j

∂ x i

) ] +ρ g i

∂

∂ t(ρ C p T) + ∂(ρ u j C p T)

∂ x j

∂ x j

(λ ∂ T

∂ x j

) =S ϕ

(1)

As the fluid flow in the greenhouse is assumed to be incompressible and turbulent, the two-equation k-ε model was employed to specify the turbulence intensity of air flow by solving two additional governing equations for turbulent kinetic energy and its dissipation rate profiles of energy The solution of the governing equations is obtained by the CFD code ANSYS Fluent (R.16.1 Academic) The radiative heat transfer in-side the greenhouse was also conin-sidered in the analysis The radiation was simulated in Fluent by using Discrete Ordinates (DO) model CFD tool was employed to simulate dynamic fields, solar radiation, and temperature distribution as well as shading effect inside reference greenhouse without PV and that with integrated photovoltaic panels on the roof, for sunny and cloudy conditions Besides, User Defined Func-tions (UDFs) were introduced to the software to calculate crop transpi-ration by addition of evapotranspitranspi-ration and energy equations as chosen

Table 1

Physical properties of the greenhouse components

Characteristics Density

ρ(kgm− 3) Specific

heatC p(Jkg− 1K− 1) Thermal

conductivityλ(Wm− 1K− 1) Under visible

wavelength Under long infra-red wavelength polyethylene film (200 µm

Photovoltaic panels:

= 0αpv =0.1 εpv =0.9

=0.25 αgv =0.75 εgv =

0.9

0 rs=0.2 αs=0.8 εs=0.9

Fig 2 Representation of measurement points inside the greenhouse

The mesh and boundary conditions greenhouse system

The mesh is generated in a three-dimensional computational

domain The choice of suitable numerical factors, grid refinement and

sufficiently accurate boundary conditions to describe flow models

around complex profiles of the greenhouse is of primary importance for

accurate results

The greenhouse was considered as a blockage and included leeward

in a large domain which has 109.6 m wide, 58 m length and 30 m high

The computational domain chosen for the greenhouse was made up 6 m

high and 18 m wide After conducting a grid independent study by

progressively refining the grid, an unstructured grid with around

102,000 grid cells was used for the analysis A finer resolution was

chosen at the vent openings, the crop canopy, all wall and at the level of

photovoltaic panel as described in Fig 3, where sharp gradients in

ve-locity and temperature were expected Tomatoes crop were arranged in

five rows perpendicular to the direction of the wind and the distance between parallel rows was take equal to 1 m and around 0.5 m between plants The modeled domain of each crop row was a rectangular par-allelepiped of 2 m high and 16 m wide

The imposition of accurate and specific boundary conditions in the CFD models is crucial for obtaining reliable results For this study, Fluent was employed with specific input variables associated with the Fluent manual and values of experimental greenhouse to simulate the spatial distribution of different climatic parameters within the greenhouse as well as the action of the photovoltaic system and crop on the flow and energy production The numerical calculations introduced in this study was described with the following boundary conditions to determinate these profiles

The inlet velocity profile consists of a vertical logarithmic velocity distribution taken from the literature [32]:

Table 2

Sensitiveness of equipment required for greenhouse measurements

Relative humidity and Air temperature Internal greenhouse at 1.8 m above the ground level (9 locations in the horizontal plane) ±0.01

◦ C and ± 2%

RH

Meteorological station

greenhouse 8 m above the ground

± 0.01 m/s

Relative humidity and Air

± 0.5 ◦ C and ± 3% RH

CR1000X data logger

Table 3

Weather data on January, 10th

Time

(h) Solar radiation

(W/m 2 )

Outside temperature ( ◦ C)

Inside temperature ( ◦ C)

Outside relative humidity (%)

Inside relative humidity (%)

Table 4

Weather data on June, 21th

Time (h) Solar radiation (W/m 2 )

Outside temperature ( ◦ C)

Inside temperature ( ◦ C)

Outside relative humidity (%)

Inside relative humidity (%)

Uinl(z) = u0

k ln

(

z

z0

)

(2) The friction velocity is expressed as:

The turbulent kinetic energy k (m2s− 2) and dissipation rate profilesε

(m2s− 3) are calculated as functions of the friction velocity u0:

k = u

2

̅̅̅̅̅̅

Cμ

The dynamic boundary conditions prescribed a symmetry boundary

at the greater frontier of the domain The outlet boundary conditions

were consisted of a zero-diffusion flux condition in which an entirely

developed air movement was assumed The Wall type boundary

condi-tions (i.e no-slip condicondi-tions) were imposed a classical logarithmic wall

function along the roof, the ground, the gravel, the polyethylene cover

and the wood The thermal boundary conditions executed fixed

tem-peratures at the side walls, roof, and surfaces and along the ground of

the greenhouse model The boundary profile of the crop canopy is also

considered in this study with its influence of transpiration, convection,

evaporation of water, and radiation on the flow and the internal

greenhouse climate The energy source terms, porous medium, and

water vapor balance equations were activated via a UDF in Fluent for

specifying vegetal biological of each crop row within natural

ecosystems

The greenhouse crop system parameters

To simulate the crop transpiration, we have used a sub-model (i.e

user defined function) in which latent and sensible heat exchanges

be-tween the crop cover and the air inside greenhouse have been taken in

account [33] The absorbed radiation R(z) in each cell of the canopy was

deduced from Beer’s law, as described in [34] The radiation extinction

coefficient that appears in Beer’s law was estimated from PAR radiation

measurements above the canopy at a value 0.75 like in the study of

Goudriaan [35] The five rows of crop arranged in the shape of

rectan-gular parallelepiped with 2 m high, 1 m wide and 16 m length each, are

considered as porous mediums in the model

Crop transpiration and plant activities in greenhouses are estimated

based on the Penman-Monteith approach[36]

The total resistance r t is defined by [37] :

r t=

(

r2+Ar2

sl+ (1 + A)ra r sl 2r a+ (1 + A)rsl

)

(6)

The stomatal resistance of the leaf r sis given by the following rela-tionship [37]:

r s=r s,min f1f2where

{

f1=1 + [exp(a1(R G− b1))]−1

P*(T v)is the saturated water vapor pressure at the canopy temper-ature which is defined as follows [38]:

P *(T v) =6.11exp

(

17.25 T v 237.8 + T v

)

(8)

The aerodynamic resistance, r a (sm− 1), is deduced from the surface and the reference height, and is expressed by the following relationship [39] :

r a= ρ C p 0.288 λ (d v v/

⃦

⃦→U

⃦

R(z)is the solar irradiation received at z (m), which is expressed by

Beer’s law [40] as follows:

The crop temperature T cropis calculated according to:

T crop=T i+ r a

ρ C p

[ 1

2I LA ν

(

dR(z)

dz − ρ Lν ω i− ω a

r t

I LA ν

) ]

(11)

Model validation

The numerical model was validated by comparing the simulation results against measurements carried out in the experimental green-house through two typical days of the summer and winter periods (June, 21th and January, 10th) when the intensity light is at these two extremes (maximum and minimum duration sun) The greenhouse indoor tem-perature and relative humidity in the reference greenhouse were measured and compared with the numerical results obtained in the similar conditions RH values (%) and temperature values (◦C) were obtained for each data point represented in Section 2.2 (Fig 2)

Table 4 and Table 5 show the absolute error of the simulated tem-perature and humidity values inside the reference greenhouse with experimented measured values in two weather condition (summer and winter period) Absolute error to the experimental data of temperature

Fig 3 Computational domain and boundary conditions

(T) and relative humidity (RH) in summer day were summarized in

Table 5 The error of RH and T was ranged between 0.62 and 1.7% and

0.3–1.3%, respectively Experimental errors derived mainly from the

relative humidity and temperature sensor and the measurement

un-certainties, and measurement error based of data acquisition

Experi-mentally measured RH is generally lower than the simulated one, since

the boundary conditions such as air velocity profile, may not strictly

reflect real internal airflow patterns Further infiltration may also affect

the airflow behavior by the air exhaust through the openings and the

door In the simulation, the ventilation through the roof openings will

eventually decrease the RH and temperature in summer day

Absolute error to the experimental data of RH and T in winter day

were also summarized in Table 6 The error was ranged between 2.75%

minimum and 8.42% maximum for the relative humidity and between

0.9% and 2.19% for the temperature Simulated RH values were higher

than the experimental measurements Indeed, in the experiments, the

evaporation crop will eventually increase the RH because of some water

droplets that not evaporated instantly The droplets will stay too longer

to evaporate in cloud day

Results and discussion

The integration of PV panels on greenhouse roofs dramatically

af-fects plants growing and internal climate due to shading of the PV

modules To determine those effects, tests with shading and un-shading

PV greenhouses were carried out The experiments were performed for a

summer day (June, 21st) and a winter day (January, 10th) The obtained

results are presented and discussed below

Internal climate simulation under un-shading PV greenhouse

This section aims at investigating internal climate inside the

green-house without PV, i.e the reference greengreen-house The internal climate of

this greenhouse was simulated in a large greenhouse (172.8 m2)

equipped with two continuous roofs The boundary conditions

corre-sponding to climatic data of the two typical days are described in section

3.4

Case study 1: Summer day

Fig 4 shows the dynamic field parallel to the prevailing wind in a

typical summer day The results exhibit that the air velocity is charac-terized by a high outside flow through the roofs opening on the wind-ward Windward and leeward roof openings improved ventilation capacity of the greenhouse and strongly affect the inside flow mecha-nism Fed both by incoming air jets on the windward and buoyancy forces encouraged by the heat exchanges at crop level, a counter rotating cell develops inside the greenhouse This phenomenon was also described by [41]

Fig 5 presents the temperature distribution in the middle of the unshaded greenhouse in a sunny day The convective developed cell is fed by cold air coming through the roof openings and the thermal heat exchange between walls Because of a big airflow in the region located above the roof, the temperature has nearly the same value of the outside, whereas it decreases in most of the cavity and the temperature auto-matically follows the air profile The air temperature has high gradients near the lateral walls (from + 3 ◦C to + 4.5 ◦C) and roof (+5 ◦C) With low-speed flow near the corner of the opening roofs, an air temperature elevation around 3 ◦C with respect to outdoor temperature is noticed

Table 5

Temperature and Relative Humidity rises measured and simulated during the winter period

X (m) Y (m) Measured T ( ◦ C) Calculated T ( ◦ C) Error (%) Measured RH (%) Calculated RH (%) Error (%)

Table 6

Temperature and Relative Humidity rises measured and simulated during the summer period

X (m) Y (m) Measured T ( ◦ C) Calculated T ( ◦ C) Error (%) Measured RH (%) Calculated RH (%) Error (%)

Fig 4 Dynamic field parallel to the wind speed inside greenhouse without PV

in a typical summer day (2.5 ms− 1 wind speed)

Case study 2: Winter day

In Fig 6, the dynamic field in a vertical section in the middle of the

greenhouse, parallel to the prevailing wind in a typical winter day is

presented The results showed that the flow was characterized by two

contra-rotating cells inside the greenhouse caused by the buoyancy

forces which enhances the homogenization of the temperature

distri-bution Above plants level, a lower airflow due to the high density of

plants is noted

Fig 7 shows the temperature distributions inside the greenhouse

during winter conditions This distribution was homogeneous inside the

greenhouse The internal temperature was around 14 ◦C which was

about 12 ◦C elsewhere, due to the intensification of natural ventilation

In winter conditions, when the greenhouse with crop is completely

closed, the absorbed and stored heat by soil changing during the day

represents an important heat source

Internal climate simulation under shading greenhouse

This section is referred to the application of photovoltaic panel The

photovoltaic panel occupied array with 28.8 m2 cell area, related to half

of the roof area, was mounted in the straight-line format on the

green-house roof surface which has around 172.8 m2 ground area The fixed

installation was studied for a slope of 33◦for PV panel aligned east–west,

mounted on the south roof of an east–west oriented greenhouse, which is

suitable for electricity production For this reason, south roofs shading is

used to mitigate excessive summer sunlight for greenhouses in high

insolation, reflecting important part of the solar energy

Fig 8 shows the dynamic field distributions in a vertical section in the center of the greenhouse equipped with PV panel It was observed that wind affect the airflow patterns under greenhouse designed with buoyancy-driven natural effect Fresh air eliminates the heat absorbed, cooling the rear of the PV cells with the effect of an increase in the electrical conversion efficiency The climate was also more homoge-neous in the greenhouse with PV panel compared to the reference case, i

e the greenhouse without PV

In Fig 9 (a) and (b), the experimental measurements of outside,

inside and under PV of solar radiation in two typical days are presented The photovoltaic system mounted on the roof acts as a passive cooling system The result illustrate that the transmission rate of the external solar radiation was only 16% under the photovoltaic panels On other hand, it was noted that about 80% of the external solar radiation was diffused by polyethylene section in the greenhouse On a clear cold day Fig 9 (a), the effects of shading were maximized causing a big drop in

the cooling load compared to the exposed roof For example, on January 10th, the peak irradiance for the PV covered roof was 55 Wm− 2 at 14:00, which was less than 10 Wm− 2 at the dawn and late afternoon The Fig 9

(b), the increased solar radiation from the panel develops a benefit on

night and on cloudy day The irradiance was high in June, reaching a peak value about 115 Wm− 2 at 14:00 whereas it was nearly zero (0.01 kWm− 2) at dawn and late afternoon

The effect of shading greenhouse at the crop level

The microclimate at the crop level is strongly linked to the climate in the greenhouse The distribution of solar radiation and air temperature

in a greenhouse are two of the main factors influencing the growth and yield of plant PV inherently conflicts with cultivation because both

Fig 5 Simulated distribution of air temperature in a typical summer day

within a reference greenhouse and domain

Fig.6 Dynamic field parallel to the wind speed inside the greenhouse without

PV in a typical winter day (2.5 ms− 1 wind speed)

Fig 7 Air temperature distributions inside the reference greenhouse in a

typical winter day

Fig 8 Flow field inside greenhouse equipped with PV panel

photosynthesis and PV depend on sunlight availability For this reason,

studying the spatial distribution of the internal air temperature and solar

radiation and also the distribution of crop transpiration flux are very

important to provide useful information for knowing the effect of the

shading of photovoltaic panels at the crop level on the climate

param-eters inside the greenhouse

To assess the shading on the microclimate at 1.8 m above the crop

level, we simulated numerically the distributed climate in the reference

greenhouse and photovoltaic greenhouse Fig 10 indicates the

tem-perature distribution at plant level inside greenhouse equipped without

(a) and with (b) panel PV in the summer day The variation of

temper-ature may be explained by plant activities and the shading The

inte-gration of PV modules to the roof was accountable to rise the indoor

temperature of the greenhouse by convection The result showed that a

lower homogeneous climate was attained in the greenhouse with PV

panels, with respect to the reference greenhouse, which provides

suit-able conditions for growing plants [31]

The minimum temperature verified in the shaded greenhouse varied

from 26 ◦C to 27.9 ◦C, while the maximum temperature range was

be-tween 28.9 ◦C and 30.19 ◦C On the other hand, the minimum

temper-ature in the greenhouse without PV panel ranged between 28 ◦C and

29.3 ◦C, and the maximum temperature varied from 31 ◦C to 32.9 ◦C

Therefore, utilization of PV panels results in about 1.6 ◦C reduction in

maximum temperature This phenomenon has also been reported by

[17] Fig 11 presents the distribution of the crop transpiration flux in the reference greenhouse (a) and photovoltaic (b) greenhouse under the same conditions Analysis of this figure reveals that the crop tion was less homogeneous, and the tomato crops had more transpira-tion in the reference greenhouse In this greenhouse, the minimum and maximum transpiration fluxes of tomato crop are 128 W/m2 and 427 W/

m2, respectively, for an ambient temperature of 33 ◦C, while this flux is ranged from 85 W/m2 to 185 W/m2 in the PV greenhouse, for an ambient temperature of 27 ◦C This difference is attributed to the in-fluence of photovoltaic panels shading The strong values of the crop transpiration flux observed in the reference greenhouse can be explained

by the high solar irradiation on hot day of summer

It can be deduced from the above results that photovoltaic system can play a positive and improving effect on plants by decreasing higher temperature due to reduction in the thermal load of the sun inside the greenhouse

Fig 12 illustrates the relative humidity distributed in the horizontal plane located at a height of 1.8 m above the ground level under the greenhouse with 50% and 0% shading Analysis of the figure shows a difference of relative humidity between shaded and an unshaded greenhouse The relative humidity ranged between 81 and 90% in the reference greenhouse and 71–77% in the greenhouse with photovoltaic panels

Fig 9 Experimental measurement of solar radiation in the outside, inside and under PV of solar radiation panel during a typical summer (a) and winter (b) day

Fig 10 Distributions of air temperature at plant level: inside the greenhouse equipped without (a) and with PV panel (b) in summer day

The average relative humidity was about 72.5% inside the PV

greenhouse, which is apparently more stable and lower compared to the

reference greenhouse (83.4%) This reduction caused by the shading of

solar panels, directly allowed tomato plants to grow on sweaty

condi-tions Reducing humidity is beneficial as higher values of humidity

promotes the fungal growth planted in protected environment [15]

Economic analysis

In the sunny condition, with the internal solar irradiation of 115

Wm− 2, the energy demand of the greenhouse was about 528 kWh/m2

The use of ventilator can provide heat to the greenhouse during winter

and cooling during summer The energy use and the cost production in

the greenhouse equipped with PV panel was estimated by comparing the

hybrid system (PV + greenhouse) with the conventional system

(greenhouse without PV panel) The Tunisian production period under greenhouse is spread out between October and May Greenhouse heating

is necessary only for three months (from December to February) what is equivalent to 910 h of operation (10 h /day) The annual solar energy of the photovoltaic panel was around 5054.4 kWh In the cloudy condition, shading PV and electricity production were extremely reduced to be insignificant, as the solar energy that extents the ground is sufficient for cultivated crops Table 7 presents the cost systems (hybrid system and conventional greenhouse) and the payback time It is apparent that the installation cost associated with PV panel covered half roof area is the most expensive (16320 $), accounting for about 51% compared to the cost of the greenhouse without PV (8000 $) In particular, the mainte-nance has less cost (300 $) in the PV integrated system, representing around 40% with respect to the reference greenhouse (500 $) The operation cost reduced by 46% (1180 $ instead of 2200 $) for the

Fig 11 Distribution of crop transpiration flux: inside the greenhouse equipped without (a) and with PV panel (b) in summer condition

Fig 12 Top-view contour maps of air relative humidity: inside the photovoltaic greenhouse (a) and the control greenhouse (b) in sunny day