Energy Conversion and Management: X 18 (2023) 100353

Contents lists available at ScienceDirect

Energy Conversion and Management: X

journal homepage: www.sciencedirect.com/journal/energy-conversion-and-management-x

Optimal design and operation of solar energy system with heat storage for
agricultural greenhouse heating
Parastoo Mohebi , Ramin Roshandel *
Department of Energy Engineering, Sharif University of Technology, Tehran, Iran

A R T I C L E I N F O A B S T R A C T

Keywords: A significant challenge of agricultural greenhouses is their high energy demand which is mainly satisfied by fossil
Design and operation optimization fuels resulting in climate change impacts. In this paper, a joint design-operation linear optimization framework
Heat storage for a solar energy system with heat storage is developed to fulfill the agricultural greenhouse heating load. The
Solar thermal collector
energy system consists of solar collector, backup boiler, and short-long term heat storages. The developed
Agricultural greenhouse
framework is applied to reach minimum-cost solution. Then, the effects of emission reduction policies, green­
house cultivation scheduling, natural gas price, and investment cost scenarios are investigated. Furthermore, a
multi-objective optimization is performed in terms of minimizing CO2 emissions and total annual cost using
epsilon-constraint method. The optimal energy system due to the minimum-cost solution includes a 1065 m2
solar collector and a 1265 kW boiler in combination with 967 kWh and 25 MWh short-term and long-term heat
storages, respectively. The 30 % carbon reduction policy results in a 70.5 % increase in solar collector area. The
selected optimal solution of the Pareto front, which is the closest solution to the ideal point, has 35.3 % more
annual cost and 89.5 % less CO2 emissions compared to the minimum-cost solution.

including greenhouse applications [4,15,16]. However, solar energy is

an intermittent energy source. The highest solar energy production oc­
1. Introduction curs in summer, while the highest heating demand is required in winter
[17,18]. Seasonal thermal energy storage is a promising solution to store
Ensuring availability of clean energy, food, and clean water are three summer heat for winter use [19,20]. Thus, the mismatch between the
of seventeen goals of the 2030 UN sustainable development [1]. The greenhouse heating demand and the availability of solar thermal energy
population of the world is expected to be 8.5 billion by 2030. Population can be compensated by using heat storage systems [13].
growth, along with dietary change, urbanization, and other conse­ Numerous articles described and evaluated the ability of different
quences of economic growth, lead to a considerable increase in energy, thermal energy storage systems to overcome the mismatch between
water, and food demand. This increasing demand puts pressure on energy supply and demand [21–23]. Esen et al. [22] conducted a
supply resources and can cause a resource shortage [2–4]. Agricultural parametric study to determine the time dependency of the stored energy
greenhouses are the junction of energy-food-water nexus by raising the in a tank containing phase change materials. In addition, a theoretical
production yield as well as reducing water demand. However, their model was developed to predict the effects of various thermal and
energy requirement can be a hundred times more than conventional geometric parameters on the melting time for different phase change
cultivation [4,5]. Thus, energy consumption is a significant challenge in materials and tank configurations [23].
expansion of greenhouses [6], while most of the required energy in cold In particular, a number of studies have been conducted to assess the
and even moderate climates is due to heating [7–9]. performance of a solar energy system combined with seasonal heat
Greenhouses conventionally rely on carbon-based fuels, contributing storage for the purpose of heating greenhouses [8,16,24–26]. The po­
to climate change impacts, high production costs, and growing concerns tential of implementing large-scale solar collector system in combina­
over fossil fuel depletion [8,10–14]. These issues necessitate the utili­ tion with seasonal heat storage for greenhouse applications is
zation of renewable energy resources [13,14]. From the environmental investigated by Semple et al. [8]. This study simulates the borehole
and sustainable point of view, solar energy is a clean, renewable, and thermal energy storage system using TRNSYS software. The investigated
necessary component of the sustainable energy future in agriculture,

* Corresponding author.
E-mail address: [email protected] (R. Roshandel).

https://doi.org/10.1016/j.ecmx.2023.100353
Received 28 November 2022; Received in revised form 23 January 2023; Accepted 24 January 2023
Available online 25 January 2023
2590-1745/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Nomenclature T temperature [◦ C]
U overall heat transfer coefficient [mW2 K]
Abbreviations u kWh]
fuel price [EUR
LP linear programming V volume [m3 ]
LTS long-term heat storage y discount rate [%]
MILP mixed-integer linear programming Superscripts
SCO solar thermal collector max maximum
SOC state of charge Min minimum
STS short-term heat storage T length of the time horizon [hour of the year]
Chemical formulas t time step [hour of the year]
CO2 carbon dioxide Subscripts
Variables and parameters a ambient
A area [m2] air air
b1 first-order heat loss coefficient [Km
W
2] ch charge
b2 second-order heat loss coefficient [K2Wm2 ] dch discharge
C specific heat [kgJK] g greenhouse
Cfuel fuel annual cost [Euro] i technology index
Cinvestment investment annual cost [Euro] in index of input power
CO & M operation and maintenance annual cost [Euro] j objective function index
Cap technology capacity [kW] m number of Pareto solutions
CRF capital recovery factor NG natural gas
E stored energy [kWh] n total number of objective functions
e specific emission [kgkWhCO2
] o optical
f objective function p panel
G solar irradiation [W] Greek letters
H length of a side of the pitched roof of greenhouse [m] γ cost coefficient [%]
h height of greenhouse [m] η efficiency [%]
L length of greenhouse [m] θ roof pitch angle [◦ ]
Loss heat loss [kW] Λ loss coefficient of heat storage [h− 1 ]
N infiltration rate [s− 1 ] µ cost coefficient
n system lifetime [year] ξ fraction of incident light absorbed by the canopy [%]
P charge and discharge power of heat storage [kW] ρ density [mkg3 ]
3
Q thermal power [kW] φ ventilation rate [ms ]
Size storage size [kWh]

energy system can reduce annual CO2 equivalent emissions by about 220 [42]. Dorotić et al. [29] developed a multi-objective optimization for
tonnes per acre. Zhang et al. [16] constructed and investigated a sea­ district heating, minimizing economic, ecological, and exergy destruc­
sonal solar heat storage system that provides greenhouse heating de­ tion objective functions. This LP model optimizes the design and hourly
mand. The system proved that seasonal thermal energy storage is a operation of the energy system. Gabrielli et al. [32] developed a multi-
feasible technology that can partially solve the heat load and solar en­ objective MILP methodology to optimize the design and operation of a
ergy mismatch between summer and winter months. multi-energy system, in terms of minimizing total cost and CO2 emis­
A significant challenge of implementing solar collectors and thermal sions objective functions.
energy storage is their high required investment costs. The financial and Regarding the literature review, there is a lack of hourly-based
environmental advantages of energy storage can be guaranteed and operation optimization for a solar energy system, including long-term
increased further by optimization of the design and operation of the heat storage, to cover greenhouse energy demand. Hourly operation of
overall energy system [27,28]. There are various approaches to handle this energy system for a whole year is essential since the greenhouse
the optimization procedure. The most commonly applied methods are heating load has a significant seasonal effect. Secondly, design/opera­
linear programming (LP) [29–31], mixed-integer linear programming tion optimizations of such energy systems are carried out using
(MILP) [28,32–34], and heuristic methods (genetic algorithms and nonlinear techniques, which can not guarantee the global optimality of
particle swarm optimization) [27,35–37]. results. Lastly, previous studies have not evaluated the impact of the
Various researchers optimized energy systems, including solar col­ dynamic greenhouse heating demand on the optimal energy system
lectors in combination with heat storage. Studies considering single- under different growing seasons.
objective optimization mainly aim to minimize total cost [38,39]. The current paper presents an optimization framework for a hybrid
Durao et al. [36] developed a framework based on Matlab/Simulink, solar energy system with long-term heat storage that satisfies the heat­
which can simulate and optimize the sizing of a greenhouse solar ing demand of a greenhouse while minimizing total annual cost. This
heating system equipped with long-term heat storage. In this study, the developed framework optimizes sizing and hourly-based operation of
genetic algorithm was employed as the optimization algorithm, and a the energy system for the time horizon of a whole year to take seasonal
constant rate of 50 kW was considered for heating load in the winter. characteristics into account. The global optimal solution is achieved
In order to generalize the outcomes for decision-making, multi- using Cplex solver. The energy system under investigation consists of
objective optimization is required [40]. Multi-objective optimization short-term and long-term heat storage systems, solar thermal collectors,
can show the trade-offs between conflicting objectives [41]. Different and a backup boiler. Additionally, this research evaluates the impacts of
approaches could be implemented to achieve a whole Pareto front, such carbon reduction policies, greenhouse cultivation scheduling, natural
as the epsilon constraint method [29,32,33] and weighted sum method gas price, and solar collector investment cost scenarios on the optimal

2
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 1. Schematic representation of the investigated energy system.

design and operation of the investigated system configuration. Ulti­

mately, the epsilon-constraint method is applied for multi-objective
optimization to minimize CO2 emissions along with the total annual
cost. The optimal design and energy flows associated with the closest
solution to the ideal point are determined using the LINMAP decision-
making approach.
The proposed method is novel since, to the best of our knowledge, no
previous study has presented a linear methodology for optimizing the
joint design and hourly operation of a solar energy system with long-
term heat storage that meets the hourly heating demand of a green­
house for an entire year. Additionally, no prior research has investigated
the effect of greenhouse heating demand under various cultivation
scheduling scenarios (different growing seasons) on the optimal design
Fig. 2. Overall methodology of the current research.
and operation of the energy system.
In summary, the main objective of the study is to decarbonize agri­
cultural greenhouses through the use of solar energy to supply heating long-term heat storage systems, solar thermal collectors, and a backup
demand, while long-term heat storage is implemented to compensate for boiler. Short-term heat storage acts as a buffer for the system, while
the mismatch between heating load and solar thermal energy avail­ long-term heat storage is charged by solar thermal collectors and is used
ability. The developed framework optimizes decarbonization-cost trade- as seasonal storage.
offs. The scientific contributions of the current paper are as follows: The overall methodology of the current research is depicted in Fig. 2.
The first stage of decarbonization of greenhouse is to (1) analyze the
• Development of LP model for a hybrid solar energy system, including greenhouse heating demand; for this, a climate-based energy demand
heat storage that minimizes total annual cost. This model guarantees model is utilized. Then (2) solar potential availability is determined
the global optimal solution. using historical data of solar insolation in the case study region, and
• Presentation of multi-objective optimization framework, which can finally (3) the proposed hybrid renewable energy system is optimized by
show the trade-offs between conflicting objectives, including total developing an optimization framework. The outcomes of this framework
annual cost and CO2 emissions, using epsilon-constraint and LINMAP are the optimal design of hybrid renewable energy (optimal sizing of
methodologies. solar collector, natural gas boiler, and storage system) in addition to
• Development of joint design and operation optimization considering optimal energy flows in the proposed system.
hourly energy flows as operation decision variables in addition to Fig. 3 presents the overview of the optimization framework, which
solar collector area, boiler, and heat storage capacities as design returns optimal sizing and hourly energy flows of the system for a whole
decision variables. year considering objective function of minimizing total annual cost. The
• Integrated assessment of the carbon reduction policies, greenhouse total cost includes the investment cost of the solar collector, natural gas
cultivation scheduling, natural gas price, and investment cost sce­ boiler, and heat storage system, in addition to operation cost, including
narios on optimal design and energy flows. natural gas and maintenance costs.
The input data of the framework includes hourly profiles of weather
This article is organized as follows: Section 2 provides an overview of conditions, greenhouse heating demand, and specific solar collector heat
the studied energy system, input data, and mathematical formulation of production, as well as techno-economic information. The following is a
the optimization framework. Section 3 describes the case study region, detailed explanation of the optimization framework in terms of input
and Section 4 offers the acquired results and discussions. Finally, the data, decision variables, constraints, objective functions, and imple­
conclusions and future work ideas are presented in Section 5. mented approach for multi-objective optimization.

2. System description and methodology

2.1. Input data
The considered hybrid energy system of this study has the primary
objective of providing heating demand of a greenhouse. As shown in The input data of the optimization framework are assumed to be
Fig. 1, the energy system under investigation involves short-term and constant throughout the lifetime of the system. The greenhouse heating
demand and solar collector heat production models are presented in this

3
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 3. Overview of the optimization framework.

section since their hourly outputs are used in the optimization model as
input parameters. In short, the inputs to the developed framework are:

• The hourly weather conditions, including air temperature and solar

radiation, obtained from publicly available resources [43].
• Physical characteristics of greenhouse structure.
• The technical and economic parameters of solar thermal collector,
short-term and long-term heat storages, and backup boiler.
• The price and emission coefficient of fossil fuel carrier used in the
hybrid energy system.
• The hourly heat demand profile of the greenhouse for a whole year,
which is output of the heating demand model presented in section
2.1.1.
• The hourly heat production profile of the solar collector, which is
output of the solar collector model presented in section 2.1.2.

2.1.1. Greenhouse heating demand

Several researchers have developed approaches to determine
greenhouse energy demand [7,44–46]. In this study, the heating de­
Fig. 4. Representation of a Venlo-type greenhouse. mand of greenhouse is calculated by the model presented in [10,47].
This model is based on energy balance and considers three mechanisms:
energy transfer between indoor and outdoor conditions, heat transfer
due to ventilation, and energy transfer related to solar radiation.
Table 1 Therefore, the energy demand of greenhouse is calculated with Eq. (1):
Input data of the greenhouse heating demand model.
( ) ( )
Parameter Unit Value Reference Qg = Ag U Tg − Ta + Cair φρair Tg − Ta − ξG (1)
N 1 2.1 × 10− 4 [47,48] Where Ag is the total area of the cover, U is the overall heat transfer
s
U W 4 [47,48]
coefficient of the greenhouse structure, Cair is the specific heat of the air,
m2 K φ is the ventilation rate, ρair is the density of the air, ξ is the fraction of
h m 4 [47] the incident light due to absorptivity of the canopy, G is the total solar
30 [47]
irradiation, and Ta is the ambient air temperature. Tg is the greenhouse
◦
θ
H m 2 [47]
L m 100 Assumption interior air temperature considered to be the minimum suitable tem­
Ta ◦
C Hourly distributed [43] perature for 100 % possible growth of crops like tomato, pepper, and
Tg ◦
C 19 (day)- 16 (night) [4] eggplant [4].
G W Hourly distributed [43] By considering Venlo greenhouse structure similar to Fig. 4 and also

4
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

ASABE [48] simplifications for energy demand, Eq. (1) can be approx­ 2.3.2. Short-term heat storage
imated by: The short-term storage is modeled through the following set of linear
equations:
L2 ( ) H ( )
Qg = (4Lh + )U Tg − Ta + 1800(h + )L2 N Tg − Ta − ξG (2)
cosθ 2 Ptsts,dch
t
Ests t− 1
= Ests + (ηch Ptsts,ch − )Δt − Losststs Δt (6)
Where L is the greenhouse length/width considering a square foot­ ηdch
print based on [47], N is the infiltration rate, h is the greenhouse height,
0≤ Etsts ≤ Sizests (7)
θ is the roof pitch angle, and H represents the length of a side of the
pitched roof. The values of input parameters for greenhouse energy t=0
Ests t=T
= Ests (8)
model are presented in Table 1.
Eq. (2) is improved to determine the heating demand at time step t: ( t)
Losststs = Λstorage
sts
t− 1
Ests + Λstatic
sts Sizests g Ta (9)
( ) ( )
L2 ( ) H 2 ( )
Qtg = 4Lh + U Tg − Tat + 1800 h + L N Tg − Tat − ξGt , t
cosθ 2 ( ) T min − Tat
g Tat = max (10)
= 1, 2, ⋯, 8760 (3) T − T min

Active cooling is required whenever the heating load becomes 0 ≤ Ptsts,ch ≤ Sizests (11)
negative, according to Eq. (3).
0 ≤ Ptsts,dch ≤ Sizests (12)
2.1.2. Specific collector heat production
Etsts
is the actual level of energy stored at hour t and Sizests is the ca­
A flat plate collector model is used to calculate solar thermal col­
pacity of heat storage. Eq. (6) states that the energy stored in each time
lector (SCO) heat production according to [49]. The specific thermal
step equals the energy stored in the former time step, increased by
output of SCO at time step t is formulated as follows:
charged energy (Ptsts,ch ) and reduced by energy output, resulted by either
(4) discharging (Ptsts,dch ) or heat losses (Losststs ). Δt is the time interval dura­
2
QtSpecific,SCO = ηo GtSCO − b1 (Tp − Tat ) − b2 (Tp − Tat )
tion (one hour). According to Eq. (7), the storage level is limited by the
Where QtSpecific,SCO represents collector heat production and GtSCO is
storage capacity. Eq. (8) guarantees an identical storage level in the last
specific solar irradiation, assuming a south-oriented collector, tilted 35◦ hour as in the first hour of the year. Heat loss of STS is described by
with respect to horizontal plane, and no shading effect. Tp and Tat are the ( )
considering the influence of ambient temperature through g Tat . Λstorage sts
mean panel temperature and ambient temperature, respectively. GtSCO
and Λstatic in Eq. (9) are heat loss coefficients. The heat flows from/to
and Tat are obtained from publicly available databases [43]. ηo is known sts
heat storage are restricted by a maximum rate, which is a fraction of the
as optical efficiency. b1 and b2 represent temperature dependent co­
storage capacity. As shown in Eq. (11) and Eq. (12), in this research, it is
efficients. In this research, these parameters are assumed, corresponding
assumed that the maximum (dis)charge rate is the storage capacity
to an Arcon Sunmark HT-SolarBoost 35/10 flat plate collector [49].
[29,31,32,51].
When the heat output becomes negative based on Eq. (4), it is set to be
0 W [27,29,34,50].
2.3.3. Long-term heat storage
Long-term heat storage is modeled similarly to short-term heat
storage with Eqs. (13)–(18), widely applied in modeling heat storages in
2.2. Decision variables
LP optimization of energy systems [28,29,32,38]. It is considered that,
unlike short-term storage, long-term storage is buried underground,
The optimization decision variables are as follows:
which leads to ignoring the impact of ambient temperature. Λstorage
represents the self-discharge parameter.
• Design variables, including:
1. Short-term and long-term heat storage sizes (Sizests and Sizelts ) Ptlts,dch
2. Solar collector area (ASCO )
t
Elts t− 1
= Elts + (ηch Ptlts,ch − )Δt − Losstlts Δt (13)
ηdch
3. The capacity of the boiler (Capboiler )
• Operation variables are determined at every hour of the year, t
0 ≤ Elts ≤ Sizelts (14)
including:
1. Natural gas consumed in boiler in each hour t=0
Elts t=T
= Elts (15)
2. The hourly input and/or output heat flows of solar thermal collector,
boiler, and storage technologies Losstlts = Λstorage Elts
t− 1
(16)
3. The hourly stored energy in short-term and long-term heat storages
0 ≤ Ptlts,ch ≤ Sizelts (17)

2.3. Optimization constraints 0 ≤ Ptlts,dch ≤ Sizelts (18)

Optimization constraints can be classified into two categories. The 2.3.4. Boiler
first category includes constraints related to the performance of con­ The boiler generates heat from natural gas and is expressed through
version and storage technologies, while the second category includes Eq. (19) and Eq. (20) [32,52]. Ramping limits and size dependency of
energy balance equations. performance are neglected for simplicity.

(19)
t
2.3.1. Solar collector Qtboiler = ηboiler Qboiler,in
The solar collector output is obtained by Eq. (5):
0≤ Qtboiler,in ≤ Capboiler (20)
QtSCO = ASCO QtSpecific,SCO (5)

Where ASCO represents the collector surface area, the only optimi­ 2.3.5. Energy balance equations
zation variable related to the solar collector. Eq. (21) and (22) formulate the energy balance between supply and

5
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 5. Hourly solar irradiation and temperature for the case study region in 2019.

demand. Eq. (21) states that the greenhouse heating demand must be 2.5. Multi-objective optimization approach
satisfied for every hour of the year by discharged energy from short-term
and long-term heat storage systems and produced heat from the backup In this paper, the epsilon constraint method is implemented for
boiler while short-term storage can be charged. multi-objective optimization. This method translates the problem to a
single objective optimization problem and considers other objectives as
Qtg = Qtboiler + Ptsts,dch − Ptsts,ch + Ptlts,dch (21)
constraints, as shown in Eq. (29) and (30). In order to apply this method,
the range of each objective function has to be recognized to ensure that
Ptlts,ch = Qtsco (22)
assigned epsilon constraints are eligible. The typical approach is to
calculate these ranges by individually optimizing the objective functions
2.4. Objective functions [29,32]. In this research, first, the total annual cost and annual emission
of the system are optimized separately to achieve the upper and lower
The developed multi-objective optimization framework is defined limits of each objective function. Then the emission interval is divided
with two contrasting objective functions: (1) the total annualized cost into equal steps, and the total annual cost is minimized while consid­
and (2) the yearly CO2 emissions of the system. The total annualized ering a maximum threshold for annual CO2 emissions.
cost, (Fannual cost ), consists of three contributions, namely, investment minFannual Cost (29)
cost (Cinvestment ), fuel cost (Cfuel ) and maintenance cost (CO&M ). Notably,
the current ecological approach does not consider life cycle assessment Femission ⩽ε (30)
and focuses on annual CO2 emissions caused by natural gas consumption
in the boiler. The objective functions are expressed as follows: 2.6. Decision-making approach
∑ ∑
Fannual Cost = Cinvestment,i + CO&M,i + Cfuel (23)
i i
After implementation of multi-objective optimization, a decision-
making process is required to select the final solution from the avail­
Cinvestment,i = μi .Capi .CRF (24) able optimal solutions of the Pareto front. Some researchers propose
LINMAP method, which selects the solution with the least distance to the
CO&M,i = γi .Cinvestment,i (25) ideal point, as represented in Eq. (31) and (32) [57]. The ideal point is an
infeasible solution where each objective is optimized, regardless of
∑
T
satisfaction with other objectives [29].
Cfuel = uNG .Qtboiler,in (26)
t=1 min(dm ) (31)

y(1 + y)n √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅

∑n
CRF = (27) dm = (Fmj − F Ideal )
2
(32)
(1 + y)n − 1 j=1 j

∑
T Where dm is the distance to the ideal point, while m stands for each
Femission = eCO2 .Qtboiler,in (28) solution on the Pareto front, and n denotes the number of objectives. In
Eq. (32), FjIdeal is the ideal value for j-th normalized objective function
t=1

Where μi is the cost coefficient and Capi is the size of the i-th tech­ and Fmj is the non-minimum value of the normalized objective function j.
nology. The annual investment cost is calculated using the capital re­
covery factor (CRF) [53]. y and n indicate the discount rate and project 3. Case study
lifetime assumed to be 5 % and 20 years, respectively. In addition, a
constant natural gas price (uNG ) of 0.03 Euro
kWh
is used [54,55]. The annual The developed methodology is applied to meet the heating demand
maintenance cost of each technology is a fraction (γ) of its annual in­ of a one-hectare greenhouse in Tehran, Iran considering meteorological
vestment cost [32]. The specific carbon dioxide emission of natural gas profiles for 2019 [43]. Fig. 5 depicts the meteorological data, including
(eCO2 ), is assumed to be 0.22 kgkWh
CO2
[56]. air temperature and solar radiation. The maximum and minimum

6
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Table 2 Section 4.6 investigates the effect of natural gas price and solar collector
Technology data. investment cost on optimal results. In Section 4.7, a two-dimensional
Parameter Unit Value Reference Pareto front, the characteristics of the selected optimal solution, and
its optimal energy flows are presented. Optimizations are performed
SCO (Solar thermal collector) ηo – 0.839 [49]
b1 W 2.46 [49] using the Cplex solver on an AMD A8-7200P Radeon R5, with a 2.4 GHz
Km2 processing clock and 4 GB of RAM.
b2 W 0.0197 [49]
K2 m2
μ Euro 190 [54] 4.1. Hourly greenhouse heat demand and solar collector heat production
m2
γ % 3 [32]
STS (Short-term heat storage) ηch,dch % 90 [51]
The hourly heating demand of a one-hectare greenhouse, based on
Λstorage % 0.06 [51] the weather condition of the case study region in 2019, is shown in
Λstatic % 0.053 [51] Fig. 6. The total annual heating load is approximately 3 GWh. It can be
Tmin ◦
C 65 [32,51] seen that the heating demand has a significant seasonal effect, peaking
Tmax ◦
C 90 [32,51] at 1945 kW in winter and 249.7 kW in summer. In addition, the distri­
Euro 4.5 [54]
μ bution of hours in different load ranges is illustrated in Fig. 7, which
kWh
LTS (Long-term heat storage) ηch,dch % 99 [38] shows that the heating load exceeds 100 kW during 38.4 % of the hours
Λstorage % 0.05 [54] of the year. The specific solar collector heat production is shown in
day Fig. 8. The maximum hourly specific collector output is 780.9 mW2 . The
μ Euro 0.9 [54]
kWh heat output of solar collector frequently equals zero during winter due to
Boiler ηboiler % 89 [54] the lower temperature and solar irradiance.
μ Euro 60 [54]
kW
γ % 2 [32] 4.2. Minimum-cost solution

In this section, the developed tool is applied to determine the

outside air temperatures are 37.3℃ and − 9.8℃, respectively [43]. The minimum-cost solution by considering two approaches: joint design-
total and average solar radiation are approximately 2063 m2kW.year
and operation optimization and design optimization. Operation cost in­
235.5 m2 W.hour
,respectively [43], making this location suitable for inte­ cludes natural gas consumption and maintenance costs. As illustrated in
grating solar thermal collectors. The technology-related parameters Fig. 9, in the design optimization approach, a 1606.3 kW boiler almost
used in the optimization framework, along with their references, are satisfies the entire heating demand due to the lower cost of boiler than
listed in Table 2. solar collector. While considering the joint design-operation approach,
optimal supply capacities are converted to a 1264.5 kW boiler and a
4. Results and discussion 1065.3 m2 solar collector. Additionally, long-term heat storage capacity
rises from 6.3 MWh to 25.1 MWh. Obtained results indicate that
The results acquired in this paper are represented through six sec­ considering operation in addition to design reduces CO2 emissions and
tions. Section 4.1 describes the outputs of the greenhouse heating de­ total annual cost by 22.4 % and 6.6 %, respectively. This demonstrates
mand and solar collector heat production models. Section 4.2 represents the significance of joint design-operation optimization. Therefore, the
the significance of considering operation by comparing joint design- base scenario in this research is the minimum-cost solution due to the
operation optimization with design optimization results in the joint design-operation approach, which is shifted to evaluate the effects
minimum-cost solution. In section 4.3, a sensitivity analysis is conducted of several scenarios in subsequent sections.
by changing heating demand. The obtained results of the base
minimum-cost solution are shifted due to the implementation of CO2 4.3. Sensitivity analysis on the greenhouse set-point temperature
mitigation policies in section 4.4. Section 4.5 shows the impact of
considering greenhouse cultivation scheduling on optimization results. To demonstrate the validity of the model, a sensitivity analysis is

Fig. 6. Hourly heating load of greenhouse for the case study region in 2019.

7
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 7. Heating load distribution of greenhouse for the case study region in 2019.

Fig. 8. Hourly specific solar collector heat production for the case study region in 2019.

conducted in this section by varying the greenhouse set-point temper­ and 50 % reductions compared to the conventional use of boilers for
ature (heating load). As previously stated, the total heating demand in greenhouse heating. Following this, the results of the two reduction
the base case (19 ◦ C during the day and 16 ◦ C at night) is 3 GWh, and the scenarios and the minimum-cost scenario are compared.
relevant optimal energy system is outlined in section 4.2. As depicted in As shown in Fig. 11, 30 % and 50 % mitigations in CO2 emissions
Fig. 10, as the greenhouse set-point temperature is raised, the required result in a 70.5 % and 220.5 % increase in solar collector area, as well as
solar collector area increases significantly while boiler capacity in­ 9.9 % and 27.8 % boiler capacity reduction relative to the base
creases gradually. In particular, an increase of 1 ◦ C in the greenhouse minimum-cost solution, respectively. This occurs due to decreased nat­
set-point temperature results in a rise of 8.6 % in heating demand, while ural gas consumption. The optimal solution for the 50 % CO2 reduction
the optimal solar collector area and boiler capacity increase by 25.1 % scenario contains a 3414.7-m2 solar collector area in combination with
and 1.9 %, respectively. Moreover, it raises annual expenses by 7.3 % 25 MWh long-term and 3.3 MWh short-term heat storage systems.
and CO2 emissions by 4 %.
4.5. Impact of reducing greenhouse cultivation period
4.4. Impact of implementation of emission reduction policies
As shown in Fig. 6, most of the heating load occurs during winter
According to [58], the case study region will be severely affected by (December to March). The greenhouse cultivation schedule can affect
energy production and CO2 emissions by 2025, emphasizing the need for the optimal energy system due to the required heating load. This effect is
immediate emission reductions. This section evaluates the impact of analyzed by comparing the year-round and two reduced cultivation
emission reduction on the optimal energy system. The amount of period scenarios. Initially, the greenhouse heating demand is calculated
emissions is restricted by applying constraints characterized by 30 % considering the cultivation periods presented in Table 3 based on [59].

8
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 9. Comparison of optimal characteristics of the investigated energy system considering minimum-cost solutions in joint design-operation optimization and
design optimization approaches, including a) Total annual cost and CO2 emissions, b) Decision variable: natural gas consumption, c) Decision variables: boiler
capacity and solar collector (SCO) area, d) Decision variables: long-term heat storage (LTS) and short-term heat storage (STS) sizes.

Fig. 10. Sensitivity analysis on heating load by varying greenhouse set-point temperature including a) Total annual cost and CO2 emissions, b) Decision variable:
natural gas consumption, c) Decision variables: boiler capacity and solar collector (SCO) area, d) Decision variables: long-term heat storage (LTS) and short-term heat
storage (STS) sizes.

Then, the annual cost is minimized by taking the calculated heating 52.2 % while lowering emissions by 51.3 % compared to the year-round
demand into account. Finally, optimal outputs for year-round and scenario. Therefore, by decreasing the cultivation season, both annual
reduced cultivation period scenarios are compared. cost and CO2 emissions are significantly reduced. Meanwhile, vegetable
Under the first and second cultivation period scenarios, the total market demand is typically highest during the winter months.
heating load of the greenhouse is reduced by about 36.1 % and 56.5 %,
respectively. As shown in Fig. 12, the first reduced cultivation period 4.6. Impact of natural gas price and solar collector investment cost
decreases total annual cost by 32.4 % and CO2 emissions by 28.3 %. In
addition, the second reduced cultivation period decreases annual cost by The impact of natural gas price and solar collector investment cost

9
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 11. Comparison of optimal characteristics of the investigated energy system considering different levels of CO2 reduction, including a) Total annual cost and
CO2 emissions, b) Decision variable: natural gas consumption, c) Decision variables: boiler capacity and solar collector (SCO) area, d) Decision variables: long-term
heat storage (LTS) and short-term heat storage (STS) sizes.

Table 3
Different cultivation period scenarios for tomato.
Cultivation scenario Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec MWh Peak demand
Total demand [ ]
yr [kW]

Year-round X X X X X X X X X X X X 3023 1945

Schedule 1 X X X X X X X X X X 1930.8 1945
Schedule 2 X X X X X X X X X 1315.1 1745.2

Fig. 12. Comparison of optimal characteristics of the investigated energy system considering the effect of cultivation scheduling, including a) Total annual cost and
CO2 emissions, b) Decision variable: natural gas consumption, c) Decision variables: boiler capacity and solar collector (SCO) area, d) Decision variables: long-term
heat storage (LTS) and short-term heat storage (STS) sizes.

projections on optimal energy system capacities, annual cost, and CO2 energy is economically feasible in scenarios of rising natural gas price
emissions are investigated in this section. Fig. 13 and Fig. 14 demon­ and decreasing solar collector investment cost.
strate that solar energy is not only used in environmentally friendly As shown in Fig. 13, three different natural gas prices are taken into
scenarios that mitigate carbon emissions but also transitioning to solar account because of the crucial effect of natural gas price on optimization

10
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 13. Comparison of optimal characteristics of the investigated energy system considering different natural gas prices, including a) Total annual cost and CO2
emissions, b) Decision variable: natural gas consumption, c) Decision variables: boiler capacity and solar collector (SCO) area, d) Decision variables: long-term heat
storage (LTS) and short-term heat storage (STS) sizes.

Fig. 14. Comparison of optimal characteristics of the investigated energy system considering different solar collector investment costs, including a) Total annual cost
and CO2 emissions, b) Decision variable: natural gas consumption, c) Decision variables: boiler capacity and solar collector (SCO) area, d) Decision variables: long-
term heat storage (LTS) and short-term heat storage (STS) sizes.

results, including export natural gas prices to Iraq and Turkey, as well as considered: 300 [29], 250 [38], 190 [54], 150, and 100 (to assess the
the EU-average natural gas price, equal to 0.02 [55], 0.03 [55] and effect of lower costs) EUR
m2
. The solar energy share increases as solar
0.055 [54] EUR
kWh. According to Fig. 14, as the price of natural gas rises, the collector cost decreases, particularly when costs are less than 190 EUR .
m2
optimal solar collector area increases, and boiler size decreases. Comparing 100 EUR to the base scenario (190 EUR ), results in 22.5 % and
m2 m2
Consideration of 0.055 EUR
kWh for natural gas price results in an increase of 68.7 % reductions in annual cost and CO2 emissions, respectively.
38.7 % in annual cost and a decrease of 70.8 % in CO2 emissions
compared to the base minimum-cost solution (0.03 EUR kWh).
Fig. 14 illustrates the effect of solar collector investment cost on the 4.7. Multi-objective optimization
optimal energy system. Five investment costs for solar collector are
The trade-offs between two conflicting objectives of minimizing total

11
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 15. The characteristics of Pareto optimal solutions, including a) Total annual cost and CO2 emissions, b) Decision variables: boiler capacity and solar collector
(SCO) area, c) Decision variables: long-term heat storage (LTS) and short-term heat storage (STS) sizes, d) Decision variable: natural gas consumption.

annual cost and CO2 emissions are presented as Pareto frontiers. The objective dominates, the characteristics of the minimum-cost solution
supply capacities, heat storage sizes, and natural gas consumption of are presented with the highest amount of CO2 emissions.
each optimal solution of the Pareto are shown in Fig. 15. The bi- After identifying Pareto optimal points, the LINMAP method is
objective optimization is performed by implementing the epsilon applied to rank these solutions regarding the relative closeness to the
constraint method; therefore, reducing the maximum threshold of CO2 ideal solution (where both objectives are minimized). The selected
emissions increases the total annual cost and share of solar energy. optimal solution is also shown in Fig. 15, marked with a yellow square,
Considering the minimum-CO2 solution, the demand is entirely covered which achieves 63.1 tonnes of CO2 emissions. Long-term heat storage
by solar energy. On the right side of the diagram, where the annual cost reaches a peak SOC of 423.7 MWh in the selected solution. Moreover,

12
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Fig. 16. Accumulated monthly heat production of solar collector (SCO) and boiler considering the selected optimal solution by LINMAP method.

Fig. 17. Short-term and long-term heat storages state of charge considering the selected optimal solution by LINMAP method.

this solution combines a 6540.4 m2 solar thermal collector area with a the reported values are affected by the location, greenhouse area,
5.5 MWh short-term heat storage and a 155.8 kW boiler. greenhouse set-point temperature, different techno-economic parame­
Fig. 16 expresses the optimal monthly heat production of the solar ters, and different solar fractions.
collector and boiler, considering the selected optimal solution. Solar In the following, the optimal collector area and solar fraction in our
collectors are the primary source of heat production, while boilers are work are compared with those reported in similar studies. In Semple
used to provide additional heating during the winter months. The hourly et al. [8], the ratio of solar collector area to greenhouse area is reported
state of charge of short-term and long-term heat storage systems is to be in the range of 21.5–50.2 %, and solar energy is responsible for
shown in Fig. 17. As can be seen, short-term heat storage serves as a 41–70 % of greenhouse total heating demand. The greenhouse location
buffer, while long-term storage has an apparent seasonal pattern. The is in Ontario, Canada with an average temperature of − 0.65 ◦ C [43]. In
long-term storage SOC is nearly zero at the end of May. Solar collectors another research in a different climate with an average temperature of
gradually charge long-term storage during the spring and summer until 17.1 ◦ C [24], the ratio of solar collector area to greenhouse area is re­
the maximum SOC is reached in September; after that, it is depleted to ported to be 21.7 % while the solar energy covers the heating demand
meet the heat load, and the cycle repeats. completely in the entire year (100 % solar fraction).
Table 4 indicates the solar collector area and heat storage sizes re­ Our optimization framework suggests the optimal ratio of solar
ported in previous greenhouse heating system studies. Low and high- collector area to greenhouse area to be in the range of 10.6–65.4 %
temperature long-term thermal energy storage systems are simulated (single and multi-objective functions) with a solar fraction of
using TRNSYS software in [8]. Xu et al. [24] and Kim et al. [26] eval­ 33.3–96.7 %. The average temperature in this case study region is
uated the thermal performance of a solar heating system with under­ 12.8 ◦ C [43].
ground seasonal energy storage for greenhouse applications. However, These analyses and comparisons demonstrate that our optimization

13
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

Table 4
Comparison of optimization results with prior papers.
Study Methodology Greenhouse Location/ Solar collector area Short-term storage Long-term storage Solar
Area capacity1 capacity2 fraction5

[8] Simulation (TRNSYS) Ontario (Canada) 861 m2 11.67 MWh 438.75 MWh 41 %
/ 4000 m2
2009 m2 11.67 MWh 630 MWh 70 %
[24] Evaluation of Shanghai (China) 500 m2 1.17 MWh 111.82 MWh 100 %
performance / 2304 m2
[26] Evaluation of Yeoju (462 m2 SCO)(234 m2 7 MWh (641.25 MWh BTES3) 38.4 %
performance (South Korea) PVT) (84 MWh TTES4)
/ 3429 m2
Current Optimization Tehran 1065 m2 0.97 MWh 25.1 MWh 33.3 %
study (Minimum-Cost) / 10,000 m2
2
Current Bi-objective Tehran 6540.4 m 5.5 MWh 423.7 MWh 96.7 %
study optimization /10000 m2

1 KJ
Capacity of short-term heat storages in prior studies is calculated considering the volume reported in the article and also assuming ρ C = 4200 , ΔT = 50 K
m3 .K
according to[28].
2 KWh
Capacity of long-term heat storages in prior studies is calculated considering the volume reported in the article and also assuming Energy density = 22.5 for
m3
KWh
Borehole type (BTES) and Energy density = 70 for tank type (TTES) according to[60].
3 m3
BTES refers to borehole thermal energy storage.
4
TTES refers to tank thermal energy storage.
5
To compute the solar fraction of the current study, short-term heat storage is assumed as a greenhouse component.

results are within the range of similar studies and could be applied to collector area, as well as 9.9 % and 27.8 % boiler capacity reduction
further investigations considering different climatic conditions and compared to the base minimum-cost solution, respectively.
objective functions. • The reduced cultivation season (from January 1st to October 15th)
decreases greenhouse heating load by 36.1 % compared to year-
5. Conclusion and future work round agriculture. This cultivation schedule lowers annual cost by
32.4 % and CO2 emissions by 28.3 %. However, the vegetable market
This research proposes an optimization framework for the joint demand in the period of October-January is generally the highest.
design and hourly operation of a greenhouse-heating solar energy sys­ • The selected optimal solution of the Pareto front, which is the closest
tem. The hybrid energy system includes solar collectors, long-term and solution to the ideal point, has 35.3 % more annual cost and 89.5 %
short-term heat storages, and a backup boiler. The minimum-cost solu­ less CO2 emissions than the minimum-cost solution. 5.5 MWh short-
tion is reached using Cplex solver. Afterward, optimal results are ach­ term and 423.7 MWh long-term heat storages, 155.8 kW boiler, and
ieved by considering carbon reduction policies, greenhouse cultivation 6540.4 m2 solar thermal collector area comprise this optimal energy
scheduling, natural gas pricing, and solar collector investment cost system.
scenarios. Finally, using the epsilon-constraint method, multi-objective
optimization is performed with the minimization of CO2 emissions and The developed framework in this paper can be implemented in future
annual cost as objectives. LINMAP determines the closest optimal so­ works to investigate how weather condition uncertainties affect the
lution of the Pareto front to the ideal point and its optimal size and optimal results in real-world situations using stochastic optimization.
energy flows. In summary, this study found: Furthermore, it is beneficial to integrate electricity consumption of
greenhouse, as well as the heating demand. It would also be valuable to
• The joint design-operation optimization (decision variables: sizes of consider more efficient heat storage systems with less land footprint,
solar collector, natural gas boiler, and storage system in addition to such as phase change materials or thermo-chemical heat storages.
heat flows) reduces CO2 emissions by 22.4 % and the total annual
cost by 6.6 %, compared to the design optimization approach (de­ CRediT authorship contribution statement
cision variables: sizes of solar collector, natural gas boiler and stor­
age system). This shows the importance of considering operation Parastoo Mohebi: Conceptualization, Methodology, Software,
decision variables in addition to sizing decision variables. Writing – original draft, Validation, Visualization. Ramin Roshandel:
• The base minimum-cost solution (due to the joint design-operation Supervision, Conceptualization, Methodology, Writing – review &
approach) results in 599.9 tonnes of CO2 emissions yearly. The editing.
optimal energy system consists of a 1264.5 kW boiler and a
1065.3 m2 solar collector in combination with 967 kWh and 25.1 Declaration of Competing Interest
MWh short-term and long-term heat storages, respectively. In short,
if the solar collector with an area of 10 % of the greenhouse area is The authors declare that they have no known competing financial
utilized, the fossil fuel input to the greenhouse decreases by 19.7 % interests or personal relationships that could have appeared to influence
compared to the conventional use of boilers. More trade-offs could be the work reported in this paper.
discussed regarding land availability or fossil fuel limitations using
our proposed optimization framework. Data availability
• The sensitivity analysis shows that implementation of decarbon­
ization policy, such as considering 30 % and 50 % carbon emission Data will be made available on request.
reduction constraints, compared to the conventional use of boilers
for heating greenhouses, causes a 70.5 % and 220.5 % rise in solar

14
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

References Conversion System for Greenhouses. Energies 2022:15. https://doi.org/10.3390/

en15186592.
[27] van der Heijde B, Vandermeulen A, Salenbien R, Helsen L. Representative days
[1] United Nations General Assembly,Transforming Our World: the 2030 Agenda for
selection for district energy system optimisation: a solar district heating system
Sustainable Development, 2015, p. 1–5. https://doi.org/1.
with seasonal storage. Appl Energy 2019;248:79–94. https://doi.org/10.1016/j.
[2] Irabien A, Darton RC. Energy–water–food nexus in the Spanish greenhouse tomato
apenergy.2019.04.030.
production. Clean Technol Environ Policy 2016;18:1307–16. https://doi.org/
[28] Renaldi R, Friedrich D. Multiple time grids in operational optimisation of energy
10.1007/s10098-015-1076-9.
systems with short- and long-term thermal energy storage. Energy 2017;133:
[3] de Amorim WS, Valduga IB, Ribeiro JMP, Williamson VG, Krauser GE,
784–95. https://doi.org/10.1016/j.energy.2017.05.120.
Magtoto MK, et al. The nexus between water, energy, and food in the context of the
[29] DorotiĆ H, Pukšec T, Duić N. Economical, environmental and exergetic multi-
global risks: An analysis of the interactions between food, water, and energy
objective optimization of district heating systems on hourly level for a whole year.
security. Environ Impact Assess Rev 2018;72:1–11.
Appl Energy 2019;251:113394. https://doi.org/10.1016/j.apenergy.2019.113394.
[4] Esmaeli H, Roshandel R. Optimal design for solar greenhouses based on climate
[30] Wang H, Yin W, Abdollahi E, Lahdelma R, Jiao W. Modelling and optimization of
conditions. Renew Energy 2020;145:1255–65. https://doi.org/10.1016/j.
CHP based district heating system with renewable energy production and energy
renene.2019.06.090.
storage. Appl Energy 2015;159:401–21. https://doi.org/10.1016/j.
[5] Barbosa G, Gadelha F, Kublik N, Proctor A, Reichelm L, Weissinger E, et al.
apenergy.2015.09.020.
Comparison of land, water, and energy requirements of lettuce grown using
[31] Di Somma M, Graditi G, Heydarian-Forushani E, Shafie-khah M, Siano P. Stochastic
hydroponic vs. Conventional agricultural methods. Int J Environ Res Public Health
optimal scheduling of distributed energy resources with renewables considering
2015;12(6):6879–91.
economic and environmental aspects. Renew Energy 2018;116:272–87. https://
[6] Hassanien RHE, Li M, Dong LW. Advanced applications of solar energy in
doi.org/10.1016/j.renene.2017.09.074.
agricultural greenhouses. Renew Sustain Energy Rev 2016;54:989–1001. https://
[32] Gabrielli P, Gazzani M, Martelli E, Mazzotti M. Optimal design of multi-energy
doi.org/10.1016/j.rser.2015.10.095.
systems with seasonal storage. Appl Energy 2018;219:408–24. https://doi.org/
[7] Costantino A, Comba L, Sicardi G, Bariani M, Fabrizio E. Energy performance and
10.1016/j.apenergy.2017.07.142.
climate control in mechanically ventilated greenhouses : A dynamic modelling-
[33] Morvaj B, Evins R, Carmeliet J. Optimising urban energy systems: Simultaneous
based assessment and investigation. Appl Energy 2021;288:116583. https://doi.
system sizing, operation and district heating network layout. Energy 2016;116:
org/10.1016/j.apenergy.2021.116583.
619–36. https://doi.org/10.1016/j.energy.2016.09.139.
[8] Semple L, Carriveau R, Ting DSK. A techno-economic analysis of seasonal thermal
[34] Carpaneto E, Lazzeroni P, Repetto M. Optimal integration of solar energy in a
energy storage for greenhouse applications. Energy Build 2017;154:175–87.
district heating network. Renew Energy 2015;75:714–21. https://doi.org/
https://doi.org/10.1016/j.enbuild.2017.08.065.
10.1016/j.renene.2014.10.055.
[9] Esen H, Esen M, Yuksel T. Modelling of biogas, solar and a ground source heat
[35] Tulus V, Abokersh MH, Cabeza LF, Vallès M, Jiménez L, Boer D. Economic and
pump greenhouse heating system by using ensemble learning. Austria: New Dev
environmental potential for solar assisted central heating plants in the EU
Mech Mech Eng WSEAS Press Vienna; 2015. p. 74–81.
residential sector: Contribution to the 2030 climate and energy EU agenda. Appl
[10] Burg V, Golzar F, Bowman G, Hellweg S, Roshandel R. Symbiosis opportunities
Energy 2019;236:318–39. https://doi.org/10.1016/j.apenergy.2018.11.094.
between food and energy system: The potential of manure-based biogas as heating
[36] Durão B, Joyce A, Mendes JF. Optimization of a seasonal storage solar system using
source for greenhouse production. Symbiosis opportunities between food and
Genetic Algorithms. Sol Energy 2014;101:160–6. https://doi.org/10.1016/j.
energy system The potential of manure-based biogas as heating source for
solener.2013.12.031.
greenhouse production 2021;25(3):648–62.
[37] Hirvonen J, ur Rehman H, Sirén K. Techno-economic optimization and analysis of a
[11] Esen M, Yuksel T. Experimental evaluation of using various renewable energy
high latitude solar district heating system with seasonal storage, considering
sources for heating a greenhouse. Energy Build 2013;65:340–51. https://doi.org/
different community sizes. Sol Energy 2018;162:472–88.
10.1016/j.enbuild.2013.06.018.
[38] Kotzur L, Markewitz P, Robinius M, Stolten D. Time series aggregation for energy
[12] van Beveren PJM, Bontsema J, van Straten G, van Henten EJ. Optimal utilization of
system design: Modeling seasonal storage. Appl Energy 2018;213:123–35. https://
a boiler, combined heat and power installation, and heat buffers in horticultural
doi.org/10.1016/j.apenergy.2018.01.023.
greenhouses. Comput Electron Agric 2019;162:1035–48. https://doi.org/10.1016/
[39] Pavičević M, Novosel T, Pukšec T, Duić N. Hourly optimization and sizing of
j.compag.2019.05.040.
district heating systems considering building refurbishment – Case study for the
[13] Cabeza LF. Advances in Thermal Energy Storage Systems: Methods and
city of Zagreb. Energy 2017;137:1264–76. https://doi.org/10.1016/j.
Applications. Adv Therm Energy Storage Syst Methods Appl 2014:1–592. https://
energy.2017.06.105.
doi.org/10.1016/C2013-0-16453-7.
[40] Shah SK, Aye L, Rismanchi B. Multi-objective optimisation of a seasonal solar
[14] Forough AB, Roshandel R. Lifetime optimization framework for a hybrid renewable
thermal energy storage system for space heating in cold climate. Appl Energy 2020;
energy system based on receding horizon optimization. Energy 2018;150:617–30.
268:115047. https://doi.org/10.1016/j.apenergy.2020.115047.
https://doi.org/10.1016/j.energy.2018.02.158.
[41] Behzadi Forough A, Roshandel R. Multi objective receding horizon optimization for
[15] Bakirci K, Yuksel B. Experimental thermal performance of a solar source heat-pump
optimal scheduling of hybrid renewable energy system. Energy Build 2017;150:
system for residential heating in cold climate region. Appl Therm Eng 2011;31:
583–97. https://doi.org/10.1016/j.enbuild.2017.06.031.
1508–18. https://doi.org/10.1016/j.applthermaleng.2011.01.039.
[42] Di Somma M, Yan B, Bianco N, Graditi G, Luh PB, Mongibello L, et al. Operation
[16] Zhang L, Xu P, Mao J, Tang X, Li Z, Shi J. A low cost seasonal solar soil heat storage
optimization of a distributed energy system considering energy costs and exergy
system for greenhouse heating: Design and pilot study. Appl Energy 2015;156:
efficiency. Energy Convers Manag 2015;103:739–51.
213–22. https://doi.org/10.1016/j.apenergy.2015.07.036.
[43] https://www.renewables.ninja/. Renewable.ninja (Accessed 16 January 2023).
[17] Hesaraki A, Holmberg S, Haghighat F. Seasonal thermal energy storage with heat
[44] Chen J, Zhao J, Xu F, Hu H, Ai QingLin, Yang J. Modeling of Energy Demand in the
pumps and low temperatures in building projects - A comparative review. Renew
Greenhouse Using PSO-GA Hybrid Algorithms. Math Probl Eng 2015;2015:1–6.
Sustain Energy Rev 2015;43:1199–213. https://doi.org/10.1016/j.
[45] Golzar F, Heeren N, Hellweg S, Roshandel R. A novel integrated framework to
rser.2014.12.002.
evaluate greenhouse energy demand and crop yield production. Renew Sustain
[18] Esen M. Thermal performance of a solar-aided latent heat store used for space
Energy Rev 2018;96:487–501. https://doi.org/10.1016/j.rser.2018.06.046.
heating by heat pump. Sol Energy 2000;69:15–25. https://doi.org/10.1016/
[46] Van Beveren PJM, Bontsema J, Van Straten G, Van Henten EJ. Minimal heating and
S0038-092X(00)00015-3.
cooling in a modern rose greenhouse. Appl Energy 2015;137:97–109. https://doi.
[19] Yang T, Liu W, Kramer GJ, Sun Q. Seasonal thermal energy storage: A techno-
org/10.1016/j.apenergy.2014.09.083.
economic literature review. Renew Sustain Energy Rev 2021;139:110732. https://
[47] Andrews R, Pearce JM, Andrews R, Environmental JMP, Assessment E, Waste G.
doi.org/10.1016/j.rser.2021.110732.
Environmental and Economic Assessment of a Greenhouse Waste Heat Exchange
[20] Antoniadis CN, Martinopoulos G. Optimization of a building integrated solar
To cite this version : HAL Id : hal-02120486 Environmental and Economic
thermal system with seasonal storage using TRNSYS. Renew Energy 2019:56–66.
Assessment of a Greenhouse Waste Heat Exchange 2019.
https://doi.org/10.1016/j.renene.2018.03.074.
[48] Heating ASABE. Ventilating and Cooling Greenhouses 2008.
[21] Ermiş K, Findik F. Thermal energy storage Sustain Eng Innov 2020;2:66–88.
[49] Arcon-Sunmark A, Solar S. Collector Factsheet Arcon-Sunmark HT-SolarBoost
https://doi.org/10.1016/B978-0-08-087872-0.00307-3.
2017;35(10):2–3.
[22] Esen M, Ayhan T. Development of a model compatible with solar assisted
[50] Kubiński K, Szabłowski Ł. Dynamic model of solar heating plant with seasonal
cylindrical energy storage tank and variation of stored energy with time for
thermal energy storage. Renew Energy 2020;145:2025–33. https://doi.org/
different phase change materials. Energy Convers Manag 1996;37:1775–85.
10.1016/j.renene.2019.07.120.
https://doi.org/10.1016/0196-8904(96)00035-0.
[51] Steen D, Stadler M, Cardoso G, Groissböck M, DeForest N, Marnay C. Modeling of
[23] Esen M, Durmuş A, Durmuş A. Geometric design of solar-aided latent heat store
thermal storage systems in MILP distributed energy resource models. Appl Energy
depending on various parameters and phase change materials. Sol Energy 1998;62:
2015;137:782–92. https://doi.org/10.1016/j.apenergy.2014.07.036.
19–28. https://doi.org/10.1016/S0038-092X(97)00104-7.
[52] Sameti M, Haghighat F. Optimization of 4th generation distributed district heating
[24] Xu J, Li Y, Wang RZ, Liu W. Performance investigation of a solar heating system
system: Design and planning of combined heat and power. Renew Energy 2019;
with underground seasonal energy storage for greenhouse application.
130:371–87. https://doi.org/10.1016/j.renene.2018.06.068.
Performance investigation of a solar heating system with underground seasonal
[53] Esen H, Inalli M, Esen M. Technoeconomic appraisal of a ground source heat pump
energy storage for greenhouse application 2014;67:63–73.
system for a heating season in eastern Turkey. Energy Convers Manag 2006;47:
[25] Ataei A, Hemmatabady H, Nobakht SY. Hybrid thermal seasonal storage and solar
1281–97. https://doi.org/10.1016/j.enconman.2005.06.024.
assisted geothermal heat pump systems for greenhouses. Adv Energy Res 2016;4:
[54] Dorotić H, Pukšec T, Schneider DR, Duić N. Evaluation of district heating with
87–106. https://doi.org/10.12989/eri.2016.4.1.087.
regard to individual systems – Importance of carbon and cost allocation in
[26] Kim DW, Kim MH, Lee DW. Economic and Environmental Analysis of Solar
cogeneration units. Energy 2021;221:119905.
Thermal and Seasonal Thermal Energy Storage Based on a Renewable Energy

15
P. Mohebi and R. Roshandel Energy Conversion and Management: X 18 (2023) 100353

[55] Https://www.tinn.ir/fa/tiny/news-186301. Iran export gas price (Accessed 16 [58] Mirzaei M, Bekri M. Energy consumption and CO2 emissions in Iran, 2025. Environ
January 2023). Res 2017;154:345–51. https://doi.org/10.1016/j.envres.2017.01.023.
[56] Buoro D, Pinamonti P, Reini M. Optimization of a Distributed Cogeneration System [59] Forschungsinstitut für biologischen Landbau. Bioland Beratung GmbH.
with solar district heating. Appl Energy 2014;124:298–308. https://doi.org/ Biologischer Anbau im Biolandbau von Tomaten: Kompetenzzentrum Ökolandbau
10.1016/j.apenergy.2014.02.062. Niedersachsen; 2005.
[57] Sayyaadi H, Mehrabipour R. Efficiency enhancement of a gas turbine cycle using [60] Dahash A, Ochs F, Janetti MB, Streicher W. Advances in seasonal thermal energy
an optimized tubular recuperative heat exchanger. Energy 2012;38:362–75. storage for solar district heating applications: A critical review on large-scale hot-
https://doi.org/10.1016/j.energy.2011.11.048. water tank and pit thermal energy storage systems. Appl Energy 2019;239:
296–315. https://doi.org/10.1016/j.apenergy.2019.01.189.

16