Energy Reports 8 (2022) 9506–9527

Contents lists available at ScienceDirect

Energy Reports
journal homepage: www.elsevier.com/locate/egyr

Research paper

Optimal design of Photovoltaic, Biomass, Fuel Cell, Hydrogen Tank

units and Electrolyzer hybrid system for a remote area in Egypt
∗
Hoda Abd El-Sattar a,b , Salah Kamel a , Hamdy M. Sultan c , Hossam M. Zawbaa d,e , ,
Francisco Jurado b
a
Department of Electrical Engineering, Faculty of Engineering, Aswan University, 81542 Aswan, Egypt
b
Department of Electrical Engineering, University of Jaén, 23700 EPS Linares, Jaén, Spain
c
Electrical Engineering Department, Faculty of Engineering, Minia University, 61111 Minia, Egypt
d
Faculty of Computers and Artificial Intelligence, Beni-Suef University, Beni-Suef, Egypt
e
Technological University Dublin, Dublin, Ireland

article info a b s t r a c t

Article history: In this paper, a new isolated hybrid system is simulated and analyzed to obtain the optimal sizing and
Received 13 September 2021 meet the electricity demand with cost improvement for servicing a small remote area with a peak load
Received in revised form 14 June 2022 of 420 kW. The major configuration of this hybrid system is Photovoltaic (PV) modules, Biomass gasifier
Accepted 13 July 2022
(BG), Electrolyzer units, Hydrogen Tank units (HT), and Fuel Cell (FC) system. A recent optimization
Available online 30 July 2022
algorithm, namely Mayfly Optimization Algorithm (MOA) is utilized to ensure that all load demand is
Keywords: met at the lowest energy cost (EC) and minimize the greenhouse gas (GHG) emissions of the proposed
Optimization system. The MOA is selected as it collects the main merits of swarm intelligence and evolutionary
Photovoltaic algorithms; hence it has good convergence characteristics. To ensure the superiority of the selected
Biomass MOA, the obtained results are compared with other well-known optimization algorithms, namely Sooty
Fuel Cell Tern Optimization Algorithm (STOA), Whale Optimization Algorithm (WOA), and Sine Cosine Algorithm
Hybrid system (SCA). The results reveal that the suggested MOA achieves the best system design, achieving a stable
Mayfly
convergence characteristic after 44 iterations. MOA yielded the best EC with 0.2106533 $/kWh, the
net present cost (NPC) with 6,170,134 $, the loss of power supply probability (LPSP) with 0.05993%,
and GHG with 792.534 t/y.
© 2022 The Author(s). 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/).

1. Introduction energy, and marine energy, will contribute approximately 40% of

the electricity supply (Anon, 2021g).
1.1. Motivations Egypt is one of the countries that is working to promote
the use of renewable energy sources by exploiting all avail-
able resources to optimize the efficient use of various renewable
Energy is more than just producing electricity; it is an eco-
sources and developing their technology to contribute to eco-
nomic pillar and a source of long-term prosperity. Renewable
nomic growth and environmental preservation (El-Sattar et al.,
energy utilization is still not growing at the desired rate. The total 2016). A strategy has been developed that includes increasing
share of renewable energy in global electricity production has not the contribution of renewable energy by 20% and 42% of electric
exceeded more than 27% (Anon, 2021f). This indicates that fossil energy production by 2022 and 2035, respectively (Anon, 2021e).
energy is still dominant in energy production, and the reason The objective of this strategy is to provide the necessary energy
can be attributed to the fact that generating renewable energy supplies for development needs, including identifying the optimal
is still relatively expensive. However, it is planned that by the mix of energy, developing reliance on renewable sources and
year 2030, renewable energy sources, which include hydropower, raising the efficiency of its use (Anon, 2021d). According to the
wind energy, solar photovoltaic energy, bio-energy, geothermal data from International Renewable Energy Agency (IRENA) in
2018, Fig. 1 presented the renewable sources generation statues
∗ Correspondence to: Technological University Dublin, Park House, 191 N (GWh) in Egypt.
Egypt is one of the wealthiest countries in the world in the
Circular Rd, Cabra East, Grangegorman, Dublin, D07 EWV4, Ireland.
E-mail addresses: [email protected] (H.A. El-Sattar),
field of solar energy and bioenergy due to its geographical loca-
[email protected] (S. Kamel), [email protected] (H.M. Sultan), tion and the availability of raw materials for biomass, especially
[email protected] (H.M. Zawbaa), [email protected] (F. Jurado). from crop wastes. For the solar energy field, Egypt enjoys direct

https://doi.org/10.1016/j.egyr.2022.07.060
2352-4847/© 2022 The Author(s). 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/).
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Nomenclature WCA Water Cycle Algorithm

WOA Whale Optimization Algorithm
ABC Artificial Bee Colony Algorithm
WTs Wind Turbines
ACO Ant colony optimization algorithm
AEFO Artificial Electric Field optimizer Symbols
AEO Artificial Ecosystem Optimization
BET Total biomass consumption (kWh)
BA Bees algorithm
BGP (t) Generator output power (kW)
BG Biomass Gasifier
Brated (t) Biomass consumption rate per hour
BSA Backtracking Search Algorithm y
CA_Cap Annual interest of the capital cost for
CRF Capital Recovery Factor each unit
CS Cuckoo Search Algorithm CA_fuel Annual fuel cost of the biomass unite
DF Dragonfly Optimizer Algorithm y
CA_Rep Replacement cost for each unit
DG Diesel Generator y
CA Annualized cost of each subsystem com-
DSPBO Enhanced model of the Student Psychol-
ponent
ogy Based Optimizer
CB A uniformly distributed arbitrary coeffi-
EC Energy Cost
cient
EO Equilibrium Optimizer
Cf The controlling coefficient that
FC Fuel Cell decreases from Cf to zero
FFA Firefly algorithm Ci The location of the search agent in STOA
FOA Firefly Optimization Algorithm CO2 Carbon Dioxide
FPA Flower Pollination Algorithm y
COM Operation and maintenance cost of each
GA Genetic Algorithm unit
GHG Greenhouse Gas emissions CRFy The Capital Recovery Factor of the
GWO Grey Wolf Optimizer required unit
y
GWP Global Warming Potential CI_Cap The initial capital cost of the required
HHO Harris Hawks optimizer unit
HSA Harmony search algorithm cω The walk coefficient
HT Hydrogen Tank units DA The movement of the sooty tern in the
IACO Improved ant colony optimization algo- search field
rithm ECon Energy consumption (kWh)
IAEO Improved Artificial Ecosystem F0 No-load fuel consumption (kg/h/50 Kw)
Optimization FBG (t) Average fuel consumption per hour
IRENA International Renewable Energy Agency FCP Generated electrical power from FC
IWO Invasive weed optimization (kW)
JOA Jaya Optimization Algorithm FCmax
P The maximum rating of the FC unit
LPSP Loss of Power Supply Probability Fm Marginal fuel consumption
MBA Mine Blast Algorithm (kg/h/50 kW)
MFO Moth–Flame Optimizer Gi The distance between the search agent
MOA Mayfly Optimization Algorithm and the best one
MOCSA multi-objective crow search algorithm GN Number of biomass generators
NPC Net Present Cost Gmax
N The maximum number of gasifier gener-
NSGAII Non-dominated sorting genetic algo- ators
rithm GPrat Generator rated power (kW)
PSO Particle Swarm Optimizer HHVH Higher heat value of the stored hydro-
PSOGSA Hybrid particle swarm-gravitational gen (kWh/kg)
search Algorithm HTm Stored hydrogen mass in the HT (kg)
PV Photovoltaic HTmax
m The maximum capacity of the HT in kg.
SA simulated annealing HTE Stored hydrogen energy (kWh)
SAA Simulated annealing algorithm k1 and k2 Attraction constants for evaluating the
SCA Sine Cosine Algorithm cognitive and social behaviors
SPEA2 Strength Pareto Evolutionary Algorithm LHVB Feedstock lower heat value (MJ/kg)
2 LHVsy Syngas lower heat value (MJ/kg)
SSA Salp Swarm Algorithm LoadP The load power (kWh)
SSO Social Spider Optimizer Maxit The maximum iterations number
STOA Sooty Tern Optimization Algorithm mB Feedstock mass flow (kg/h)
TAC total annual cost msy Syngas mass flow (kg/h)
TLBO Teaching–Learning Based Optimization PELE/HT Rated power of water electrolyzer
pbestij The best position reached by the ith
mayfly in the ith dimension

9507
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

PDum Dummy load ηELE Electrolyzer efficiency

Pmax
ELE/HT The maximum rating of the electrolyzer ηFC FC overall efficiency
unit ηHT HT efficiency
PEXC Excess energy ηinv Inverter efficiency
Pmax
EXE The maximum rating of the excess ηsy Syngas efficiency (%)
energy c A coefficient describing the nuptial
PHT/FC Power supplied from the HT to the FC dancing of the mayflies.
pi (t) the current iteration (t) f(.) The objective function of the optimiza-
Piti The best individual’s position in SCA tion problem
pij The location of the ith search agent in gbest The global best position
the jth dimension i the position of certain mayfly
Pren Generated power from renewable m A variable that ranges between (0 ≤ k ≤
sources 2π )
Pren/ELE Electrical energy from renewable p The position vector
sources feeding the electrolyzer pbest The best position
PVN Number of PV units Radius The radius of each turn in the spiral
PVmax
N The maximum number of PV modules u and v Factors describing the spiral shape
PVP Generated power of the PV (kWh) v The velocity vector
PVrat
P PV Rated Power (kW) X The control variables of the optimization
PVη PV module efficiency problem
qi (t) The position of a certain individual β The weight factor value of each objec-
female mayfly in the search area tive function
qij (t) The location of the ith female in the jth ∆t Time durations during simulation
dimension during the iteration t ε A random number in the range of [−1,
r1 , r2 , r3 , and r4 The four random control parameters 1]
used to prevent becoming stuck in poor λ Limit the visibility between individual
solutions and keep the exploration and mayflies
exploitation processes in balance ρ Random numbers in Gauss distribution
rfm The distance between the male and
female mayflies
rg The distance between the position of the
individual in the pi (t) and its gbest
Ri The various positions of the search agent
Xi (it) in the search space in its move-
ment towards the best fittest sooty
tern
rp the distance between the position of the
individual in the pi (t) and its pbest
SRint (t) Solar Radiation Intensity at a time (t)
(kWh/m2 )
TA Ambient temperature (◦ C)
TC Cell Temperature (◦ C)
Tnor Cell temperature at normal operating
conditions (◦ C)
Tr Reference cell temperature (◦ C)
uij (t) The velocity of the ith female mayfly in
the jth dimension Fig. 1. The status of renewable sources generation in Egypt.

vij The velocity of the ith search agent in

the jth dimension
Wη Wiring Efficiency solar radiation with an intensity that varies from 2000 to 3200
′ ′
x , y , and z ,
′
The angle of attack kWh/m2 /year from north to south (Anon, 2021b). The average
daily sunshine ranges between 9 to 11 hours/day from North
Xbest Best search agent
to South (Anon, 2021b,a). Currently, solar energy projects in
Xi (it) The current location of the sooty tern at Egypt have become one of the strategic axes of the plan to
the it-th iteration expand reliance on renewable energy. Solar energy represented
Xiti The current individual (i) at iteration (it) about 2% of the total energy generated in Egypt from 2012 to
in SCA 2017 (Anon, 2021b). During 2019–2020, the electricity generated
δT Temperature coefficient of maximum through solar energy reached about 3655 GWh (Anon, 2021d).
power (◦ C−1 ) Bioenergy is the energy produced from organic materials in
which energy is stored. The most important sources are agri-
cultural waste, food waste, and livestock dung (El-Sattar et al.,
2020b). Egypt has a huge potential for bioenergy production,
especially from the crop remains, as about 21 million tons of
9508
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 1
A summary of different hybrid system configurations.
Hybrid System Ref. Location Technique Objective function
Configuration
Diab et al. (2019) Abu-Monqar village in the Western WOA, WCA, PSOGSA, EC
Desert of Egypt MFO LPSP
Fathy et al. (2020) A remote area in Aljouf region in the SSO, HHO, GWO, EC
PV, WTs, DG, and north of Saudi Arabia MVO, ALO
batteries Kharrich et al. (2021a) Dakhla area in Morocco EO, HHO, AEFO, NPC
GWO, STOA
Suresh and Meenakumari (2019) – An improved GA EC, NPC, LPSP, CO2
emissions
Kharrich et al. (2021b) Rabat, Morocco MOPSO, PESA II, Total NPC, CO2
SPEA2 emission
Naderipour et al. (2019) Ahvaz city in Iran GWO and PSO NPC
PV, WTs, and batteries
Sanajaoba (2019) Manipur state in India FOA EC
Tabak et al. (2019) a faculty building in Karabük GWO, GA, and SA Total NPC, LPSP, and
University campus, Turkey EC
PV, WTs, BG, and
batteries Alshammari and Asumadu (2020) A small rural community located at HSA, JOA, and PSO lower cost and
latitude 29◦ 49.794′ N and longitude reduce CO2 emission
39◦ 34.362′ E
Singh and Kaushik (2016) – ABC and HOMER annual cost
Ghosh and Karar (2018) Kaidupur village, in Punjab State, DF and ABC TAC
PV, BG, and battery India
Eteiba et al. (2018) Monshaet Taher village in Beni Suef, FPA, HS, ABC, and NPC
Egypt FOA
PV, WTs, BG, and Alturki and Awwad (2021) Saudi Arabia WOA, FFA, and PSO EC
pumped hydro storage

PV, Biogas generator, Das et al. (2019) A radio transmitter station in India WCA and MFO Total NPC
Battery, and Shi et al. (2021) A radio transmitter station in China DSPBO, WCA, MOA, Reducing the overall
Pumped-storage hydro at Hubei’s Three Gorges Dam GA, FOA, and GWO life cycle cost

agricultural waste are disposed of annually (Hassan et al., 2014). perform the techno-economic analysis of a stand-alone hybrid
This waste can be exploited and invested well in the bio-economy system located in a remote village in west China. The proposed
by using appropriate technologies. system consists of PV, WTs, BG, and batteries.
Based on the Literature review, Table 1 provides a summary
1.2. Literature review of these cited references depending on their studied systems’
configurations. At the same time, a summary of some recent
Currently, there are many papers related to studying the per- hybrid system research based on FC configuration is presented
formance of hybrid renewable energy systems and studying dif- in Table 2.
ferent optimization algorithms. The purpose of such studies is to This paper focuses on studying the optimal sizing of an iso-
assess the system components’ maximum capacity and improve lated hybrid system based on PV, BG, Electrolyzer units, HT units,
the main economic and technical indicators in the design of these and FC to serve a remote area Abu-Monqar in the Western Desert
systems. In Naderipour et al. (2019), the Grey Wolf Optimizer of Egypt. The literature works demonstrate that metaheuristic
(GWO) has been applied to reduce the total net present cost (NPC) algorithms based on natural strategies have been successfully
by considering the reliability constraints for an isolated hybrid applied to a variety of optimization issues. According to the
system in Ahvaz city in Iran. This research is based on study- principle of ‘‘no free lunch’′ (NFL) (Wolpert and Macready, 1997),
ing three different system configurations; the first combination which can be characterized as ‘‘no optimization approach offers
consists of PV panels, wind turbines (WTs), and batteries, the good performance for free’’. This principle supports the argument
second one consists of PV, and batteries, and the third consists that existing techniques can still be improved, and new ones can
of WTs, and batteries. The obtained results from the proposed be implemented. Based on the previous context, MOA has been
GWO are compared with the particle swarm Optimizer (PSO). The applied for solving the optimization problem related to the sug-
obtained results showed the effectiveness of the proposed GWO gested configuration. The suggested population-based approach
in designing a reliable and costless PV, WTs, and batteries hybrid is inspired by adult mayfly behavior, such as crossover, muta-
system. tion, swarm gathering, nuptial dance, and random walk; these
As is evident, most of the papers are concerned with studying behaviors incorporate significant benefits of the proposed algo-
hybrid systems consisting of PV, WTs, and DG with batteries. rithm, which improve exploration. To estimate the effectiveness
Many researchers have currently studied the optimal sizing of of this algorithm and to test the characteristics of exploitation
the hybrid system consisting of biomass generators (BG) with and exploration of the MOA method; a qualitative analysis has
the PV and WTs using different optimization algorithms. HOMER been made, in which this method has been benchmarked against
software is often used in simulating and analyzing the hybrid seven high-quality meta-heuristic optimization algorithms on 25
power system from a techno-economic perspective. In Cano et al. test functions which are classified in 3 groups (unimodal, mul-
(2020), the authors proposed an optimal configuration analysis timodal and fixed-dimension), as well as 13 CEC2017 test func-
using HOMER software for an off-grid hybrid system based on PV, tions, multi-objective optimization, and a discrete classic flow-
BG, hydrokinetic turbines, and batteries banks located in southern shop scheduling problem. Also, it has been compared with other
Ecuador. Also, in Li et al. (2020), HOMER has been applied to popular metaheuristic algorithms, specifically PSO, GA, FOA, DE,
9509
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 2
FC based on different hybrid energy systems.
Ref. Hybrid System Location Technique Objective Findings
Configuration function
Maleki and Pourfayaz (2015) PV/WTs/DG/battery Iran HSA TAC The PV/WT/DG/battery system is more cost-effective
PV/WTs/DG/FC HOMER than the PV/WT/DG/FC system. Nonetheless, using
SAA PV/WT/DG/FC as an energy source provides a reliable
energy source, and with improved FC and
electrolyzer efficiency, may become economically
competitive in the future.
Sultan et al. (2021) PV/WTs/FC Suez Gulf in IAEO,AEO EC The suggested IAEO algorithm shows quick
Egypt PSO, SSA, convergence properties, the best objective function
and GWO minimum values, and the lowest EC compared to the
traditional AEO, PSO, SSA, and GWO algorithms.
Dong et al. (2016) PV/WTs/battery/HT Zhejiang in IACO Annual Cost The hybrid system described in this study is based
China ACO LPSP on the hybrid system of battery and hydrogen.
Under certain constraints, the results of IACO
algorithm presented that the suggested hybrid
system is more cost-effective than others because of
battery leakage and lower FC efficiency.
Attemene et al. (2020) WT/FC/battery/super- Tanda in NSGAII TAC This study aims to figure out which variables had
capacitor Ivory Coast LPSP the biggest influence. The most influential
parameters were the wind speed profile, wind
generator cost, FC cost, and battery cost. As a result,
the suggested design would be quite beneficial in
places with high wind potential.
Khan and Javaid (2020) PV/WT/FC Hawksbayin JOA TAC The PV/FC is the most cost-effective hybrid system
PV/FC Pakistan GA, BSA and compared to the PV/WT/FC and WT/FC systems. The
WT/FC PSO Jaya method’s results showed its superiority
compared to other algorithms used.
Duman and Güler (2018) PV/WT/FC Turkey HOMER EC This system presented the dispatch control strategies
and the optimal ratings of the renewable system
units. According to the techno economic analysis of
this study, showed that battery storage is still
economically superior to hydrogen storage.
Jamshidi and Askarzadeh (2019) PV/DG/FC Kerman in MOCSA NPC According to the design results, incorporating
(off-grid) Iran LPSP hydrogen energy technology will decrease the total
cost of hybrid energy systems. Furthermore,
operating reserve has a greater impact on the Pareto
front than load and solar power uncertainties.
Gharibi and Askarzadeh (2019) PV/DG/FC Kerman in MOCSA EC Based on the design results, when the price of
(grid-connected) Iran LPSP electricity reduces, the EC rises and the value of the
grid factor decrease.
Fathy (2016) PV/WT/FC Egypt MBA, PSO Annual Cost The proposed algorithm based on MBA showed its
ABC, and CS effective for solving the presented optimization
problem of optimal sizing of the hybrid PV/WT/FC
system compared to the other three algorithms used.
Hadidian Moghaddam et al. (2019) PV/WT/FC Iran FPA Total NPC FPA method showed the best results in solving the
PV/FC PSO optimal sizing problem with fast convergence,
WT/FC TLBO minimum Total NPC and better reliability compared
to the other optimization algorithms used. PV/WT/FC
system is the most cost-effective for many situations,
with lower costs and higher reliability indices in
comparison to other suggested hybrid system
combinations.

HS, invasive weed optimization (IWO), and bees algorithm (BA), 1.3. Contribution and paper organization
under the same conditions to achieve fairness in comparison.
The obtained results demonstrate that the proposed MOA has The novelty of this work consists in studying and providing the
optimal sizing and energy management strategy of integrating
superior performance in convergence rate compared to the other
an isolated hybrid system based on PV, BG, Electrolyzer units,
well-known metaheuristic optimization which has been used in
HT units, and FC to serve a rural town called Abu-Monqar in
this paper. The suggested MOA method’s convergence character- the Western Desert of Egypt. The MOA technique was developed
istic is noteworthy since it frequently achieves the best overall with Matlab© and applied to solve the optimization problem
solution in the early iterations. MOA’s findings are satisfactory related to the suggested configuration. According to (Zervoudakis
for discrete and multi-objective optimization problems. To the and Tsafarakis, 2020), the suggested MOA is a new optimization
authors’ knowledge, MOA has never been used before to evaluate method, and it incorporates the best features of PSO, GA, and
FOA. MOA can be considered a modification of PSO; it provides a
the optimal size for the proposed configuration system. In Chen
powerful hybrid algorithmic structure based on mayfly behavior
et al. (2021), MOA has been adopted for analyzing the optimal for researchers attempting to improve PSO performance using
allocation with the minimum installation cost for an Electric crossover and local search techniques. The major contributions
Vehicle Charging Station system in India. of this study can be summarized as follows;
9510
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 2. The proposed hybrid system’s configuration layout.

• A new hybrid energy system, consisting of PV, BG, FC, Elec- The control diagram of the proposed hybrid system is shown in
trolyzer units, and HT units, has been developed for a small Fig. 3. This figure illustrates a general scheme of the system’s
village located in the Western Desert of Egypt; operation. If the energy produced from the PV and BG units covers
• The optimal size of the proposed stand-alone hybrid system the load needed power, then no power from the FC is needed,
is determined using a novel MOA approach; and no power is supplied to the electrolyzer unit. If the energy
• To prove the optimal effective and performance of the MOA produced exceeds the load needed, in this case, this excess energy
methodology, the results of MOA are compared with a will feed the electrolyzer and begin producing hydrogen, which
three optimization algorithms; STOA (Maleki et al., 2020), is used to charge the HT units. If the HT becomes full, the dummy
WOA (Bukar et al., 2019), and SCA (Sultan et al., 2020); load will consume this excess power. If the energy generated by
• The optimal function of this hybrid system is to minimize the PV and BG units is unable to fulfill the load power needs, the
the EC and reduce the CO2 emissions; FC will utilize hydrogen stored in the HT units to compensate for
• Evaluate the optimal decision variables of the components, the power generation shortfall.
which include the number of PV units (PVN ), number of
biomass generators (GN ), the rated power of the FC (FCP ), 2.1. Photovoltaic solar module (PV)
the rated power of the water electrolyzer (PELE/HT ), and the
mass of the hydrogen gas tank (HTm ). PV module is a device that uses the photoelectric effect to
The organization of the rest of the paper is as follows: Section 2 transform light energy directly into electrical energy. It is a simple
provides the construction of the hybrid system, which consists energy provider since it does not require chemical reactions or
of PV, BG, Inverter, FC, Electrolyzer units, and HT components. fuel to generate electrical energy. (Singh and Baredar, 2016). The
Section 3 discusses the formulation of the optimization problem, generated power of the PV array (PVP ) and the cell temperature
which includes the economic and cost analysis, objective func- (TC ) can be evaluated according to the following equations (Geleta
tion, constraints, and operation strategy. Section 4 describes the et al., 2020; Maleki et al., 2020; Bukar et al., 2019; Sultan et al.,
proposed optimization of MOA, STOA, WOA and SCA algorithms; 2020):
Section 5 presents the results and discussion, which includes the
SRint (t)
( )
PVP (t) = P Wη PVη (1 − δT (TC − Tr )),
PVN PVrat
( )
meteorology of the case study area and the discussion of the (1)
optimal results, while Section 6 provides the conclusion of this 1000
( )
study. Tnor − 20
TC (t) = SRint (t) + TA , (2)
800
2. The construction of the proposed hybrid system
where SRint (t), PVN , PVrat
P , Wη , and PVη are the intensity of solar
In this section, the stand-alone configuration of PV, BG, In- radiation at a time (t), the number of PV units, the rated power,
verter, FC, Electrolyzer units, and HT components is discussed in the wiring efficiency, and the PV module efficiency, respectively,
detail. Fig. 2 illustrates the graphical diagram of the proposed δT denotes the PV module temperature coefficient of maximum
hybrid energy system. PV units, electrolyzer, and FC are con- power, TA is the ambient temperature, and Tr denotes the refer-
nected to the DC bus, where the PV modules provide the required ence temperature of the solar cell is equal to 25 ◦ C. Tnor is the
input power to the electrolyzer. The biomass plant is connected cell temperature at normal operating conditions, it is dependent
to the AC bus, to which the main load is connected. The generated on four basic standard reference settings; solar irradiation with
power from the PV and FC systems is converted through DC/AC 800W/m2 , wind velocity with 1 m/s, air temperature of 20 ◦ C,
inverter to meet the demand load. The parameters and operating and inclination angle of 45◦ from the horizontal (Anoune et al.,
characteristics of all system components are presented in Table 3. 2020).
9511
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Where ηsy denotes the syngas efficiency. LHVsy and LHVB are
the lower heat value of the syngas and biomass feedstock, re-
spectively. msy and mB are the mass flow of the syngas and
the biomass feedstock, respectively. BGP (t), GN , Brated (t), GPrat ,
FBG (t), Fm , and F0 are the output power of the generator, the
number of generators used, the biomass consumption rate per
hour, the generator rated power, average fuel consumption per
hour, Marginal and No-load fuel consumption, respectively.
Greenhouse Gas (GHG) emissions from the feedstock residues
are calculated as (Eteiba et al., 2018):

BET = LHVB ∗ Sum(Brated ), (6)

ECon = BET ∗ 0.0002778, (7)
GHG = ECon ∗ GWP ∗ 0.43, (8)

Where BET and ECon are the total biomass consumption and
the energy consumption (kWh), respectively. GWP is the global
warming potential (=1) (Sultan et al., 2021), and 0.43 is the
electricity emission factor (kg CO2 /kWh) (Ray et al., 2007).

2.3. Electrolyzer modeling

The electrolyzer consists of an anode and a cathode separated

by an electrolyte in which an electric current passes through,
Fig. 3. The control diagram of the proposed hybrid system. creating a chemical reaction that decomposes water into its two
elements (hydrogen and oxygen). The generated hydrogen is
collected around the surface of the anode. Based on (Kashefi Ka-
2.2. Biomass system modeling viani et al., 2009), most of the electrolyzers, produce hydrogen
at 30 bar but this hydrogen is transferred directly to the FC
This work uses a small-scale downdraft gasification technol- at a low pressure of 1.2 bar. In most research, the produced
ogy that converts solid biomass into a gaseous fuel (syngas) hydrogen is directly delivered to the hydrogen tank, however, in
used in turbines to generate electricity. The reason for choosing other studies, the pressure is raised to 200 bar with the use of
downdraft gasification technology is that the fixed bed gasifiers a compressor in order to increase the stored energy in the tank
are often simple and low-cost, and it has a high total carbon con- (Kashefi Kaviani et al., 2009). Other works employ two tanks with
version rate and high solids residence time and low gas velocity. A low and high pressure. If this tank is fully charged, a compressor
downdraft gasifier is a co-current reactor in which the gasification is used to transport the hydrogen to the high-pressure tank in
agent is supplied from the top or sides of the reactor, causing the order to waste as little energy as possible because the compressor
feed stock and gasification agent to flow in the same direction. will not be running all the time (Mills and Al-Hallaj, 2004). In
Downdraft gasifiers are commonly utilized in engine applications
this study, Proton-conducting polymeric membrane electrolyzer
because of their simplicity, low investment costs, good product
(PEM) is used, where the produced hydrogen of the electrolyzer
gas quality, and ability to install in small spaces.
is immediately transported to the hydrogen tank at a pressure
Egypt is one of the agricultural countries, which means the
of 30 bar. The energy is transferred to the hydrogen tank from
availability of agricultural residues that can be used as biomass
the electrolyzer (PELE/HT ) is expressed as (Sultan et al., 2021;
feedstock. Egypt produces the largest amount of corn crop residues
annually compared to other agricultural residues, with a to- Kashefi Kaviani et al., 2009);
tal amount of 3.23 million tons/year (Abdelhady et al., 2021). PELE/HT = ηELE ∗Pren/ELE , (9)
Corn stalk is one of the feed stocks with the greatest potential
for energy production and industrial applications because of its Where ηELE denotes the efficiency of an electrolyzer and Pren/ELE
abundance, which is abundant in the middle region of Egypt. In represents the electrical energy feeding the electrolyzer from
this study, corn stalk is used due to its availability in areas near renewable sources.
the study case area. In addition, compared to other agricultural
crop wastes, corn stalk is a possible hydrocarbon feedstock for
2.4. Hydrogen tank storage (HT)
biomass gasification since it contains substantial amounts of
carbon and modest levels of ash and sulphur.
Based on the features of the corn stalk used, the system After the hydrogen and oxygen separation process, the hydro-
performance is expressed as follows (Cano et al., 2020; Eteiba gen gas is transported and stored in HT to be used in the FC for
et al., 2018; El-Sattar et al., 2018): power generation. To ensure the process balance, the HT upper
pressure is adjusted to the same as that of the water electrolyzer
LHVsy msy
ηsy = , (3) pressure, while the HT lowest pressure is adjusted to the same
LHVB mB fuel cell pressure (Attemene et al., 2020). During the operation,
GN ηsy LHVB Brated (t)
( )
the stored hydrogen mass in the HT (HTm ) is constrained by the
BGP (t) = − F0 GPrat , (4)
Fm LHVsy minimum and maximum values (Sultan et al., 2021) according to
LHVsy this formulation:
FBG (t) = (GN F0 GPrat + Fm BGP (t)) , (5)
ηsy LHVB
m ≤ HTm (t) ≤ HTm ,
HTmin max
(10)
9512
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 3
Main parameters of the stand-alone hybrid system PV/Biomass/FC.
Component Parameter Value Unit
PV module capital cost 7000 $
δT 0.0037 –
PVη 15 %
◦
Tr 25 C
Rated power 1 kW
PV System (Diab et al., 2019; Sultan et al., 2021)
Length 1625 Mm
Width 1019 Mm
Thickness 46 Mm
lifetime of PV system 20 Year
PV replacement cost 13885 $
O&M cost 20 $/unit-year
ηinv 95 %
Max. power 1 kW
inverter lifespan 15 Year
Inverter Unit (Sultan et al., 2021)
Inverter Capital Cost 800 $/unit
Replacement cost 750 $/kw
O&M cost 8 $/unit-year
LHVB (El-Sattar et al., 2018) 14.8 MJ/kg
LHVsy (El-Sattar et al., 2018) 4.766 MJ/kg
ηsy 80 %
GPrat 40 KW
No-load fuel consumption (F0 ) (Eteiba et al., 2018) 0.0644 kg/h/50 kW
Gasification System
Marginal fuel consumption (Fm ) (Eteiba et al., 2018) 0.2998 kg/h/50 kW
Capital Cost (Alturki and Awwad, 2021) 23,700 $/kw
Lifespan (Alturki and Awwad, 2021) 15,000 h
Replacement cost (Alturki and Awwad, 2021) 15000 $/unit
Yearly O&M cost (Alturki and Awwad, 2021) 0.05 $/h
Rated Power 1 kW
Lifespan 20 Year
Efficiency 75 %
Electrolyzer Unit (Sultan et al., 2021)
Capital cost 2000 $/unit
Replacement cost 1500 $/unit
Yearly O&M cost 25 $/unit
Power 1 kW
Efficiency 95 %
Capital cost 1300 $/unit
Hydrogen Tank Unit (Sultan et al., 2021)
O&M cost 15 $/unit-year
Replacement cost 1200 $/unit
Lifespan 20 Year
Power 1 kW
Efficiency 50 %
Capital cost 3000 $/unit
FC System (Sultan et al., 2021)
O&M cost 175 $/unit-year
Replacement cost 2500 $/unit
Lifespan 5 Year

The stored hydrogen energy in the HT (HTE (t)) and the stored not only water, but also thermal energy (heat) (Wang et al., 2020,
hydrogen mass (HTm (t)) at time step (t) are expressed as fol- 2011). There are many types of fuel cells, the common types of
lows (Sultan et al., 2021; Baghaee et al., 2016; Vendoti et al., fuel cells are Proton exchange membrane fuel cell, Alkaline fuel
2021): cells, Direct alcohol fuel cell, Molten carbonate fuel cell and Solid
( ) oxide fuel cell (Abdelkareem et al., 2021). In this study, a polymer
PHT/FC (t)
HTE (t) = HTE (t − 1) + PELE/HT (t) − × ∆t, (11) electrolyte membrane (PEMFC) is used. PEMFC is characterized by
ηHT its small size, long lifespan, and operation under relatively low
HTm (t) = HTE (t) /HHVH , (12) temperatures (Samy and Barakat, 2020). The generated electrical
power from PEMFC (FCP ) depends on its overall efficiency (ηFC )
Where PHT/FC (t) indicates the power supplied towards the FC from and can be expressed by Baghaee et al. (2016), Vendoti et al.
the HT, ηHT is the HT efficiency, ∆t is the time durations during (2021):
simulation, and HHVH denotes the higher heat value of the stored
hydrogen gas (kWh/kg). FCP = ηFC ∗ PHT/FC , (13)

2.6. Converter unit

2.5. Fuel cell (FC) modeling

It is a power electronic device used for converting DC power

Fuel cells are electrochemical cells used to produce electri-
from PV modules and FC into AC power to serve the loads. For the
cal energy through an electrochemical reaction by continuously
case study, the efficiency of the inverter (ηinv ) used is 95%, and its
supplying oxygen and hydrogen gases to the cell. At the anode,
lifespan is 15 years.
hydrogen is oxidized into protons (which rotate by the electrolyte
The generated power from renewable sources (Pren ) is ex-
to the anode) and electrons (which rotate from outside the cell
pressed according to the following equation:
to the anode), and after oxygen is collected, it is reduced to form
water. Fuel cells produce electricity while having as byproducts Pren (t) = PVP (t) + BGP (t)/ηinv , (14)
9513
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

3. Formulation of the optimization problem where β denotes the weight factor value of each objective func-
tion, and X represents the control variables of the optimization
3.1. Economic and cost analysis problem that must be optimized using the studied optimization
algorithms, as given in the following formula:
The major goal of this section is to optimize the proposed
isolated hybrid system to achieve a guaranteed supply of power at X = [PVN GN HTm PELE/HT FCP ], (24)
the lowest possible cost, which explains how to evaluate the total
annual cost, the EC, and the net present cost (NPC). The present 3.3. Constraints
value of the principal investment and operating expenses over the
project’s lifetime is depicted by the NPC as (Das et al., 2019): The optimization algorithm is working according to the fol-
NPC = TAC/CRF, (15) lowing constraints and based on the upper and lower limits of
the decision variables;
Where TAC denotes the total annualized cost, which is the sum of
PVmax
y
⎡ ⎤
the annualized cost of each subsystem component (CA ), and the ⎡
PVN
⎤ N
max
capital recovery factor (CRF) is a conversion factor that is used to ⎥ ⎢ GN
⎢ ⎥
⎢ GN ⎥
transform the initial principal cost into the annual capital cost. 1 ≤ ⎢ PELE/HT
⎢ ⎥ ⎢ Pmax
⎥ ≤ ⎢ ELE/HT ⎥,
⎥
(25)
The CRF and CTA are expressed as (Sultan et al., 2021): ⎣ FC
⎣ FCmax
⎦ ⎢ ⎥
P
P
⎦
R(R + 1)S HTm
CRF(R, S) = , (16) HTmax
m
(R + 1)S −1
∑ y
PEXC ≤ Pmax
EXE , (26)
A + CA + CA + CA + CA + CA ,
CA = CPV BG ELE HT FC inv
TAC = (17)
LSPS ≤ 0.06, (27)
Where R is the interest rate, and S represents the lifespan of the
proposed hybrid system. y denotes the subsystem components of PVN and GN are set as integer numbers. PVmax N and Gmax
N are
the hybrid energy system, which are PV, biomass gasifier, elec- the maximum number of PV modules and gasifier generators,
max
y
trolyzer, HT, FC, and power inverter. The CA of each component respectively. Pmax
ELE/HT and FCP are the maximum rating in kW of
y
is composed of annual capital cost (CA_Cap ) of the system, annual the electrolyzer unit and the FC, respectively. HTmax
m denotes the
y
replacement cost (CA_Rep ) of the system, annual fuel cost (CA_fuel ) maximum capacity of the HT in kg. PmaxEXE represents the maximum
y rating of the excess energy.
of BG, and annual operation and maintenance cost (COM ) of the
system, it can be expressed as (Das et al., 2019):
y y y y 3.4. Operation strategy
CA = COM + CA_Rep + CA_fuel + CA_Cap , (18)
y The operation strategy used in the proposed hybrid system is
CA_Cap is the annualized capital cost for each unit which is the
expenses that are spent for each component in the first year of simplified in a flowchart shown in Fig. 4. The operating strategy
the project’s lifetime in order to make the system operational, it can be summarized in three main cases as follows:
can be given by (Suresh and Meenakumari, 2019): When the energy produced from the renewable sources sat-
LoadP (t)
y y isfies the load demand (Pren (t) = ηinv
), in this case, the
CA_Cap = CRF∗CI_Cap , (19) generated power is supplied to cover the required load demand,
y and no power is needed from the FC, and no power is supplied
CI_Capis the initial capital cost of the hybrid system.
The energy production cost (EC) ($/kWh) is the cost of one to the electrolyzer;
kWh produced from the hybrid system, and it is calculated as When the energy produced from renewable sources exceeds
the load demand (Pren (t) > η P , in this case, this excess energy
Load (t)
follows (Sultan et al., 2021): inv
PEXC (t) will feed the electrolyzer, and the produced hydrogen is
NPC ($) * CRF
EC($/kWh) = ∑8760 , (20) used to charge the HT units;
1 LoadP (kWh) When the energy produced from renewable sources is insuffi-
cient to meet the load demand (Pren (t) < η P ), in this case, the
Load (t)
Where LoadP is the load power in (kWh). inv
FC will use hydrogen stored in the HT units to make up for the
3.2. Objective function deficiency in power generation. When the HT capacity reaches its
minimum, this leads to load loss.
This work aims to establish an optimal combination of the Each optimization technique used in this study works by fol-
system components to attain the maximum energy supply of lowing the steps below:
the hybrid renewable energy system. This goal is achieved by
minimizing the EC, maintaining high reliability of the power (1) Run the software of creating the proposed hybrid power
supply, minimizing the loss of power supply probability (LPSP) system, and this step contains the following stages:
based on (22), and reducing the excess energy (PEXC ) absorbed by (a) Generation of the initial population.
the dummy load (PDum ) as to keep the overall system cost down. (b) Run the code for creating the hybrid system, which
The PEXC is defined by (23). The following expressions are used to is shown in the flowchart in Fig. 4, and stated in
calculate the values of the objective functions:
Eqs. (1) through (20).
MinX (F ) = Min(β1 EC + β2 LPSP + β3 PEXC ), (21) (c) The fitness functions are evaluated for all agents/
8760 positions according to Eq. (21).
∑ LoadP (t) − Pren (t) − FCP (t)
LPSP = ( ) (22) (2) Run the optimization technique with the proposed hybrid
LoadP (t)
1 system, which includes the following stages:
8760
(a) Depending on the nature of each optimization tech-
∑
PEXC = (PDum /LoadP ), (23)
1
nique used, the size of the system parts is updated.
9514
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 4. The flowchart of the operation strategy of the proposed hybrid system.

9515
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

(b) Run the code for creating the hybrid system, which pbestij denotes the best position reached by the ith mayfly in
is shown in the flowchart in Fig. 4, and stated in the ith dimension. The best position of the ith search agent is
Eqs. (1) through (20). determined according to the following formula,
(c) The fitness functions are evaluated for all agents/
pi (t + 1) , f (pi (t + 1)) < f(pbesti )
{
positions according to Eq. (21). pbesti = , (30)
(d) Examine whether the proposed system satisfies the pbesti , otherwise
end criterion or not. If so, (Yes), the code will be Where f(.) is the objective function of the optimization problem.
paused and will move to the following step; other- rp and rg are evaluated as follows,
wise (No), the three previous stages from ‘‘a’’ to ‘‘c’’, ⎛ ⎞1/2
will be repeated. d
∑
(e) Display the required results, including hybrid system r2p = ⎝ (pij − pbesti )⎠ , (31)
sizing and EC, NPC, LPSP, and dummy load optimal j=1
values. ⎛ ⎞1/2
d
∑
4. Proposed algorithms r2g = ⎝ (pij − gbesti )⎠ (32)
j=1
4.1. The mayfly optimization algorithm (MOA)
To ensure the optimization algorithm’s best performance, the
best candidates should continue their up and down dancing.
Mayflies are a species of insects that belong to the order
Therefore, the velocity of the best mayflies should continue to
Ephemeroptera that originated from the so old group named
change. The velocity, in this case, is expressed as follows,
Palaeoptera. The MOA algorithm can be called a modification of
the standard PSO (Zervoudakis and Tsafarakis, 2020; Gao et al., vij (t + 1) = vij (t) + c ∗ ε, (33)
2020). The proposed MOA method combines the merits of the
PSO, genetic algorithm (GA), and Firefly algorithm (FFA). The Where ε is a random number in the range of [−1, 1] and c is a
MOA is inspired by the patterns of the mating process of the coefficient describing the nuptial dancing of the mayflies.
mayflies. In the MOA, the males and females are arranged in
different groups. In addition, the male mayflies are considered 4.1.2. Movement of females
stronger than the female individuals, and then the males will do The female mayflies do not fly in swarms like males. For the
better in the optimization process. A group of males and another mating process, female mayflies move towards the location of
female group are randomly initiated. Each mayfly is considered a the males. The position of a certain individual female mayfly in
candidate solution in the search space of the optimization prob- the search area qi (t) is updated as presented in the following
lem. The position of each individual determines the possibility of equation (Zervoudakis and Tsafarakis, 2020; Gao et al., 2020),
obtaining the desired optimal solution. The random location of
qi (t + 1) = qi (t) + vi (t + 1) (34)
the individuals in the search space is presented by the position
vector p = (p1 , p2 , . . . . . . ..., pd ), the performance of each mayfly The authors of the MOA decided to model the attraction be-
is assessed regarding a certain objective function represented by tween males’ and females’ mayflies in a deterministic manner.
f(x). The social and movement experiences of each individual in The fittest female is proposed to be attracted to the fittest male,
the search space determine the new path of the mayfly after the second female towards the second male, etc. Consequently,
updating its position using the velocity vector v = (v1 , v1 , . . . . . . ., the velocity of the female mayflies is determined using the fol-
vd ,). Each individual in the swarm decides its direction regarding lowing equation (Guo et al., 2021),
its best position (pbest), while the best position obtained in the { 2
swarm by any individual is the global best position (gbest). vi (t) + k2 *e−ηrfm pij (t) − qij (t) , f (qi ) > f(pi )
( )
vij (t + 1) =
uij (t) + cω ∗η f (qi ) ≤ f(pi )
4.1.1. Movement of males
Similar to the PSO, the individuals in the swarm update their (35)
current position according to the following equation in which the
position of certain mayfly i, in the current iteration (t) (pi (t)) is Where qij (t) denotes the location of the ith female in the jth
updated by adding its velocity in the next iteration vi (t+1) (Gao dimension during the iteration t. uij (t) presents the velocity
et al., 2020), of the ith female mayfly in the jth dimension. rfm denotes the
distance between the male and female mayflies, and cω is the
pi (t + 1) = pi (t) + vi (t + 1), (28) walk coefficient that is randomly selected in the case when a
female mayfly is not attracted by any male.
The male mayflies continue their nuptial dance up to some me-
ters above the water level, and their velocity is updated as fol-
4.1.3. Mating process
lows (Zervoudakis and Tsafarakis, 2020),
The way for the selection of the parents is the same as that
2 2
vij (t + 1) = vij (t) + k1 ∗ e−λrp (pbestij − pij (t)) + k2 ∗ e−λrg (gbestj − Pij (t)), used for the attraction of females by males’ mayflies. The selec-
tion might be random or according to a certain objective function.
(29) When the mating is based on an objective function, the best fittest
Where k1 and k2 are attraction constants for evaluating the cog- female mates with the corresponding best fittest male, the second
nitive and social behaviors, respectively. λ is used to limit the vis- female in the chain mates with the second fittest male in the
ibility between individual mayflies. rp and rg denote the distance swarm, and so on. Finally, two offspring are generated according
between the position of the individual in the current iteration to the following equation (Zervoudakis and Tsafarakis, 2020; Guo
pi (t) and its best position (pbest) and the global best position et al., 2021),
(gbest), respectively. The values of rp and rg are calculated accord- offspring1 = ρ ∗ male + (1 − ρ ) ∗ female, (36)
ing to Eqs. (30) and (31), respectively. vij and pij are the velocity
and the location of the ith search agent in the jth dimension. offspring2 = ρ ∗ female + (1 − ρ ) ∗ male, (37)

9516
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 4 sults showed that STOA outperforms nine well-known optimiza-

Pseudo-Code of the MOA (Duman and Güler, 2018). tion techniques in terms of performance. In addition, the results
objective function represented by f(x), x = (x1 , x2 , . . . . . . xd )T of the unimodal and multimodal test functions demonstrate the
Initialize the male and female population pi and qi
Compute solutions
STOA algorithm’s exploration and exploitation potencies. On the
Evaluate the global best (gbest) other hand, the results of CEC 2005 and CEC 2015 standard test
Do While stopping criteria are not met functions prove that the STOA is capable of addressing high and
Update solutions and velocities of males and females challenging dimensionality bound-constrained actual situations.
Compute solutions
Statistical analysis has been carried out to establish the statistical
Rank the mayflies
Mate the mayflies significance of the STOA over benchmark test functions. Further-
Calculate offspring more, to verify and demonstrate the performance of STOA on a
Separate offspring to male and female randomly given search space, it has been applied in 6 classic engineering
Replace the worst solutions with the best new ones structure problems, namely; pressure vessel design, speed re-
Update gbest and pbest
ducer design, welded beam design, tension/compression spring
end while
Return design, 25-bar truss design, and rolling element bearing design
problems (Dhiman and Kaur, 2019).
However, sooty terns are species of birds that are found in
all regions of the world and are scientifically called Onychoprion
fuscatus. STOA stimulates the attacking and migration behav-
ior of sooty tern birds in nature. Sooty terns generally live in
colonies, and they use their natural intelligence in attacking their
prey (Dhiman and Kaur, 2019). The following features character-
ize the migration behavior of the sooty terns:

• During seasonal movements, sooty terns fly in groups while

their initial position within the same group is different to
prevent collisions in the group,
• In the same group, all individuals can fly in the direction of
the fittest sooty tern,
• All individuals in a particular group update their location
regarding the best fittest candidate,
• During movement in the air, sooty terns use the spiral-
shaped mechanism surrounding the prey and attacking it.
During the migration stage, the sooty terns should satisfy
some conditions such as avoidance of collision, flying, and con-
verging towards the best sooty tern in the group, updating loca-
tion according to the location of the best fittest sooty tern. The
mathematical representation of the migration behavior is given
in Eqs. (37) to (41) (Dhiman and Kaur, 2019).

Ci = DA ∗ Xi (it), (38)
DA = Cf − (it ∗ Cf/Maxit ), (39)

Where Ci denotes the location of the search agent, Xi (it) denotes

the current location of the sooty tern at the it-th iteration, DA
presents the movement of the sooty tern (search agent) in the
search field, Cf is a controlling coefficient that decreases from Cf
to zero and used for adjusting the movement in the search field,
and Maxit denotes the maximum iteration number.

Ri = CB ∗ (Xbest (it) − Xi (it)), (40)

CB = 0.05 ∗ rand(0, 1), (41)

Fig. 5. Flowchart of the MOA method. Where Ri denotes the various positions of the search agent Xi (it)
in the search space in its movement towards the best fittest sooty
tern, CB denotes a uniformly distributed arbitrary coefficient.
Where ρ is random numbers in Gauss distribution, the flowchart Gi = Ci ∗ Ri , (42)
describing the procedure of the proposed MOA algorithm is
shown in Fig. 5, while the block diagram of MOA used for op- Where Gi denotes the distance between the search agent and the
timizing the proposed hybrid renewable system is illustrated in best one.
Fig. 6, and the algorithm pseudo-code is presented in Table 4. During migration time, sooty terns have the ability to change
the speed and the direction of the attack. During the attack,
sooty terns move towards the prey in a spiral shape that can be
4.2. The Sooty Tern Optimization Algorithm (STOA) presented by the following equations (Dhiman and Kaur, 2019),

x′ = Radius ∗ sin(m), (43)

STOA is a bio-inspired algorithm that has been validated using
44 benchmark test functions (Dhiman and Kaur, 2019). The re- y = Radius ∗ cos(m),
′
(44)
9517
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 6. Block diagram of MOA used for optimizing the proposed hybrid systems.

z′ = Radius ∗ m, (45) obtained at each iteration. There is a variable called A, which is a

control variable that is varied from 2 to 0 to allow exploration
R=u∗e , kv
(46)
and exploitation, respectively. When |A| > 1, a random search
′ ′ ′ agent is picked, while in case |A| < 1, the best solution is chosen
Where x , y , and z , denote the angle of attack; Radius denotes
the radius of each turn in the spiral; m is a variable that ranges for updating the search agents’ positions. To balance the explo-
between (0 ≤ k ≤ 2π ); u and v are factors describing the spiral ration/exploitation capabilities, WOA can switch between a spiral
shape. The updated position of the candidate search agent is and a circular movement based on the value of another variable
computed according to the following equation, called p. The pseudo-code of the WOA technique is indicated in
Table 6.
Xi (t) = (Ri (it) ∗ x′ + y′ + z′ )∗Xbest (it),
( )
(47)

The flowchart of the proposed STOA algorithm is shown in 4.4. The Sine cosine optimization algorithm (SCA)
Fig. 7, while the block diagram of STOA used for optimizing the
proposed hybrid renewable system is illustrated in Fig. 8 and the A meta-heuristic optimization method called sine cosine al-
algorithm pseudo-code is presented in Table 5. gorithm (SCA) has been created by Mirjalili et al. (Mirjalili, 2016).
The SCA optimization method includes the transportation of pop-
4.3. The Whale Optimization Algorithm (WOA) ulation members within the search space, presenting an approach
to the issue. SCA uses trigonometric sine and cosine formulas for
The Whale Optimization Algorithm (WOA) has been proposed this purpose. SCA begins by producing a set of random solutions.
by Mirjalili et al. (Mirjalili and Lewis, 2016). WOA is a recent These solutions are iterated during the optimization process to
heuristic optimization method; it is based on imitating the whale’s improve them. At each step of the calculation, solutions are
social behavior and is inspired by the humpback whale’s bubble- updated based on the next formulas;
net hunting behavior. The WOA technique starts with a set of {
Xiti + r1 sin (r2 ) ⏐r3 Piti − Xiti ⏐ , r4 < 0.5
⏐ ⏐
random individuals. Individuals update their places in relation Xiti +1 = , (48)
Xiti + r1 cos (r2 ) ⏐r3 Piti − Xiti ⏐ , r4 ≥ 0.5
⏐ ⏐
to either a randomly chosen search agent or the best solution
9518
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 5
Pseudo-Code of the STOA (Dhiman and Kaur, 2019).
p p p p p p
Initialize the population Xi = (X1 , X2 , X3 , . . . . . . , XN ) within the limits Xmin
i ≤ Xi ≤ Xmax
i
Initialize parameters DA and CB
Evaluate the fitness of the whole population
Best search agent → Xbest
While (it < Maxit )
for (i = 1: N)
Update the position of the current search agent
end for
Initialize parameters DA and CB
Evaluate the fitness of the whole population
Update Xbest
it = it +1
end while
Return

Table 6
Pseudo-Code of the WOA (Mirjalili and Lewis, 2016).
p p p p p p
Initialize the population Xi = (X1 , X2 , X3 , . . . . . . , XN ) within the limits Xmin
i ≤ Xi ≤ Xmax
i

Evaluate the fitness of each search agent

Best search agent → Xbest
While (it < Maxit )
for each search agent
Update a, A, C, l and p
if1 (p < 0.5)
if2 (|A| <1)
Update the position of the current search agent
else if2 (|A| ≥1)
Select a random search agent (Xrand )
Update the position of the current search agent
end if2
else if1 (p ≥0.5)
Update the position of the current search agent
end if1
end for
Check if any search agent goes beyond the search space and amend it
Evaluate the fitness of each search agent
Update Xbest if there is a better solution
it = it +1
end while
return Xbest

26◦ 30.3′ North and 27◦ 39.8′ East, respectively (Diab et al., 2019).
Table 7 The system average load value used for simulation is 260 kW,
Pseudo-Code of the SCA (Mirjalili, 2016).
and the peak value is 420 kW (Diab et al., 2019). Fig. 9 shows
Start the population of the solution
Initialize r1 , r2 , r3 , and r4 values
the hourly load power over one summer day and one year (Diab
et al., 2019). The monthly average global solar radiation and
Do
Compute the objective function for each solution temperature for one year are taken from the SOLAR RADIATION
Evaluate the best solution AND METEOROLOGICAL DATA SERVICES (Anon, 2021c), which
While (it < Maxit ) is based on the average long-term values over 20 years from
Update r1 , r2 , r3 , and r4 values the NASA Surface Meteorology and Solar Energy website for the
Update the obtained solution
end while
proposed case study area. The hourly average solar radiation on
Return the best solution the horizontal plane and temperature during one year and a
summer day are shown in Figs. 10 and 11, respectively.
Biomass energy is one of the renewable energies that can
be found almost anywhere because it is formed from organic
Where Xiti is the current individual (i) at iteration (it), Piti denotes materials, either directly from plants or indirectly from industrial,
the best individual’s position, and r1 , r2 , r3 , and r4 are the four ran- agricultural, and household products (El-Sattar et al., 2020a).
dom control parameters used to prevent becoming stuck in poor Biomass energy can be converted as a source of electricity or
solutions and keep the exploration and exploitation processes in heat using several technologies such as thermolysis, anaerobic
balance. The pseudo-code of the SCA technique is indicated in digestion, and gasification (Bakheet et al., 2018). Egypt is one
Table 7. of the agricultural countries, which means the availability of
agricultural residues that can be used as biomass feedstock. Rice
5. Results and discussion straw is one of the feed stocks with the greatest potential for
energy production, which is abundant in the north, corn stalk
5.1. Meteorology of the case study area in the middle of Egypt, and sugar cane in the south (Food and
Agriculture Organization of the United Nations, 2017). In this
Abu-Monqar is a remote Egyptian settlement in the Western study, corn stalk is used due to its availability in areas near the
Desert of Egypt. It is a remote oasis located 100 km south of study area. The amount of feedstock used during each month is
Frafra Oasis, 250 km from the El- Dakhla Oasis, and 650 km shown in Fig. 12.
southwest of Cairo. The selected area’s latitude and longitude are
9519
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

• SCA: Nsearch = 20, Maxit = 100, dimension size = 5, a

linearly decreased from 2 to 0.
Fig. 13 displays the convergence curves of the optimization
process based on MOA, STOA, WOA, and SCA techniques. These
optimization algorithms have been applied 50 runs for 100 iter-
ations to assess the best fitness function value to subjugate the
proposed algorithms’ randomness and verify their stability and
robustness. For the proposed case study, each optimizer is used
in the same way. The final values of the proposed MOA’s objective
function are within a small range, demonstrating MOA’s consis-
tency in finding the best solution for the optimization problem.
Fig. 14 shows a comparison between the convergence curves
for the best implementation within the 50 runs of the opti-
mization programs for the different algorithms. According to the
convergence characteristics, MOA has the best results in achiev-
ing the minimum fitness value after 44 iterations, outperforming
STOA, WOA and SCA.
Table 8 indicates the results of the optimal sizing characteris-
tics of the hybrid microgrid for the four optimization algorithms
used in this study. These results include the best fitness function,
the number of iterations to reach the optimal solution, the opti-
mal values of the decision variables, EC ($/kWh), NPC ($), LPSP
(%), and GHG (ton/y). This table shows that the MOA has the
best fitness function with 0.1219998, followed by STOA, SCA, and
WOA with 0.1221296, 0.1224865, and 0.1251143, respectively.
By comparing the results provided in the table, it can be noticed
that the MOA is characterized by the minimum EC, the minimum
NPC, and the lowest GHG, compared with the rest of the used
algorithms, followed by the STOA, SCA, and WOA. The values of
the LPSP obtained from all algorithms are within constraints.
Fig. 15 shows the contribution of the annual cost of each unit
(representing the operating and maintenance costs, which are
costs spent each year) of the proposed hybrid system for the MOA
algorithm. As is evident, the FC represents the highest annual
percentage cost with 36% of the total cost of the proposed system,
followed by PV modules with 19%, then the Electrolyzer units
with 18%, inverter with 14%, biomass system with 11%, and finally
the HT with 2%. The results from the MOA indicated that the
Fig. 7. Flowchart of the STOA method. annual percentage cost of the FC represents the largest portion
of all units.
Fig. 16 reveals the contribution of each unit to the NPC per-
5.2. Discussion of the optimal results centage of the proposed hybrid system using the MOA. The re-
sults revealed that the NPC of the PV modules accounts for the
largest portion with 27% (1,066,906 $), followed by the Elec-
trolyzer unit with 25% (1,006,631$), the FC with 18% (728,728 $),
A new application for a recent MOA is proposed in this work the biomass system with 15% (614,001$), the inverter with 12%
for designing the optimum sizing of stand-alone PV, BG, FC, Elec- (488,132$), and the HT represents the lowest NPC value with 3%
trolyzer units, and HT systems. The proposed algorithm’s results (134,966$).
are compared with those obtained by other recent methods, Fig. 17 indicates the hybrid PV, BG, and FC power system’s
STOA, WOA, and SCA, in order to affirm the proposed MOA’s power output performance over a 24-hour cycle for four different
efficacy in achieving the highest reliability and lowest cost. days with high energy consumption over the year. According to
The studied optimization algorithms are coded and imple- the obtained results, the excess power from PV and BG units (Pren )
mented using MATLAB software. For each optimization algorithm, is used to power the electrolyzer for hydrogen production. When
50 independent runs are conducted, and the results are recorded the power generated by the Pren is not able to meet the load
to compare the algorithms’ accuracy. The parameters used in each power needs, then the FC will use the hydrogen stored in the
algorithm are given as: HT unit to make up for the shortage in power generation. Fig. 18
shows the variation in the stored mass amounts of hydrogen in
• MOA: Population Size (males and females) =20, Maximum
each hour in the HT during one year (8760 h) in case of using
number of iterations (Maxit ) = 100, and dimension size = 5,
the MOA method. It can be noticed that the maximum amount of
Personal Learning Coefficient =1, Global Learning Coefficient
hydrogen mass stored is 63 kg.
=1.5. Distance sight Coefficient = 2, Damping Ratio = 0.8,
Random flight = 1.
• STOA: Number of search agents (Nsearch ) = 20, Maxit = 100, 6. Conclusions
and dimension size = 5, a linearly decreased from 2 to 0.
• WOA: Nsearch = 20, Maxit = 100, dimension size = 5, a This work’s main objective is to assist officials, decision-makers,
linearly decreased from 2 to 0. and the private sector in identifying advantages, opportunities,
9520
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 8. Block diagram of STOA used for optimizing the proposed hybrid system.

Fig. 9. Daily and yearly load profile.

and obstacles to the future of solar and biomass energy in Egypt In this paper, a recent optimization technique called MOA has
and potential profit opportunities. Hence, this will attract in- been utilized to design a small stand-alone microgrid with the
vestments by the private sector to expand the use of solar and aim of meeting the electricity requirements of a small remote
biomass energy as an alternative to traditional energy. area in the Western Desert of Egypt with a peak load of 420 kW.

9521
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 10. Hourly PV average radiation during one year and a summer day.

Fig. 11. Hourly PV average temperature during one year and a summer day.

Fig. 12. The rate of the feedstock used during a year.

9522
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 13. The convergence curves of the studied algorithms for 100 iterations, (a) MAO, (b) STOA, (c) WOA, (d) SCA.

9523
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Table 8
The Optimization Features Parameters for the Proposed Hybrid System Based On Using MOA, STOA, WOA, and SCA.
MOA STOA SCA WOA
Best Fitness Function 0.1219998 0.1221296 0.1224865 0.1251143
Iteration Number 44 41 27 33
PV (units) 47 42 41 51
Nominal PV power (kW) 46.7 41.92 40.1 50.79
Generators (units) 2 3 3 3
Nominal generator power (kW) 72 84.9 86 89.6
Electrolyzer rated (kW) 319.87 326.35 323.419 336.78
Hydrogen tank (kg) 62.48 61.28 66.15 67.41
Fuel cell (kW) 124.71 130.09 134.927 155.03
EC ($/kWh) 0.2106533 0.210961 0.2123484 0.2367564
NPC ($) 6,170,134 6,179,148 6,219,783 6,934,705
LPSP (%) 0.05993 0.059411 0.059187 0.0365246
GHG (ton/y) 792.534 806.654 808.559 849.626

Fig. 14. The convergence curves for the best implements using MOA, STOA, WOA and SCA for 100 iterations.

Fig. 15. Break down of the annual costs of the hybrid system’s units. Fig. 16. Break down of the NPC of the hybrid system units using MOA.

This hybrid power system is based on two renewable resources, techniques; STOA, WOA, and SCA. The simulation results demon-
the PV modules and biomass generating system integrated with strated the proposed algorithm’s effectiveness in identifying the
the FC and an electrolyzer for hydrogen production. This work optimal capacities of the generating and energy storage units
aims to reduce the energy production cost, NPC, and GHG while in the proposed grid-independent hybrid system. The MOA has
maintaining high values of power supply reliability. In order achieved the best optimal solution for the proposed system after
to assess the accuracy of the proposed algorithm, its results 44 iterations with the minimum EC of 0.2106533 $/kWh, NPC of
have been compared with those obtained by other well-known 6,170,134 $, LPSP of 0.05993%, and GHG of 792.534 t/y.

9524
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 17. The operation capacity of the proposed hybrid system for 4 different days using MOA.

9525
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

Fig. 18. The hourly change in the mass of hydrogen stored during a year.

Declaration of competing interest Anon, 2021f. Renewables - fuels & technologies - IEA. Available on-
line: https://www.iea.org/fuels-and-technologies/renewables. (Accessed 6
February 2021).
The authors declare that they have no known competing finan- Anon, 2021g. World energy outlook 2020 – analysis - IEA. Available on-
cial interests or personal relationships that could have appeared line: https://www.iea.org/reports/world-energy-outlook-2020. (Accessed 6
to influence the work reported in this paper. February 2021).
Anoune, K., Ghazi, M., Bouya, M., Laknizi, A., Ghazouani, M., Abdellah, A. Ben,
Astito, A., 2020. Optimization and techno-economic analysis of photovoltaic-
Data availability wind-battery based hybrid system. J. Energy Storage 32, 101878. http://dx.
doi.org/10.1016/j.est.2020.101878.
Data will be made available on request. Attemene, N.S., Agbli, K.S., Fofana, S., Hissel, D., 2020. Optimal sizing of a
wind, fuel cell, electrolyzer, battery and supercapacitor system for off-grid
applications. Int. J. Hydrogen Energy 45, 5512–5525. http://dx.doi.org/10.
Acknowledgments 1016/j.ijhydene.2019.05.212.
Baghaee, H.R., Mirsalim, M., Gharehpetian, G.B., Talebi, H.A., 2016.
The author (Hossam M. Zawbaa) thanks to the European Reliability/cost-based multi-objective Pareto optimal design of stand-
alone wind/PV/FC generation microgrid system. Energy 115, 1022–1041.
Union’s Horizon 2020 research and Enterprise Ireland for their http://dx.doi.org/10.1016/j.energy.2016.09.007.
support under the Marie Skłodowska-Curie grant agreement No. Bakheet, S., Kamel, S., El-Sattar, H.A., Jurado, 2018. Different biomass gasification
847402. The authors thank the support of the Programme of reactors for energy applications. In: Proceedings of the 2018 Twentieth
Requalification of the Spanish University System for 2021-2023. International Middle East Power Systems Conference. MEPCON, IEEE, pp.
660–665.
Bukar, A.L., Tan, C.W., Lau, K.Y., 2019. Optimal sizing of an autonomous
References photovoltaic/wind/battery/diesel generator microgrid using grasshopper op-
timization algorithm. Sol. Energy 188, 685–696. http://dx.doi.org/10.1016/j.
Abdelhady, S., Shalaby, M.A., Shaban, A., 2021. Techno-economic analysis for the solener.2019.06.050.
optimal design of a national network of agro-energy biomass power plants Cano, A., Arévalo, P., Jurado, F., 2020. Energy analysis and techno-economic
in Egypt. Energies 14, http://dx.doi.org/10.3390/en14113063. assessment of a hybrid PV/HKT/bat system using biomass gasifier: Cuenca-
Abdelkareem, M.A., Elsaid, K., Wilberforce, T., Kamil, M., Sayed, E.T., Olabi, A., Ecuador case study. Energy 202, 117727. http://dx.doi.org/10.1016/j.energy.
2021. Environmental aspects of fuel cells: A review. Sci. Total Environ. 752, 2020.117727.
141803. http://dx.doi.org/10.1016/j.scitotenv.2020.141803. Chen, L., Xu, C., Song, H., Jermsittiparsert, K., 2021. Optimal sizing and sitting of
EVCS in the distribution system using metaheuristics: A case study. Energy
Alshammari, N., Asumadu, J., 2020. Optimum unit sizing of hybrid renewable
Rep. 7, 0–9. http://dx.doi.org/10.1016/j.egyr.2020.12.032.
energy system utilizing harmony search, Jaya and particle swarm optimiza-
Das, M., Singh, M.A.K., Biswas, A., 2019. Techno-economic optimization of an
tion algorithms. Sustain. Cities Soc. 60, 102255. http://dx.doi.org/10.1016/j.
off-grid hybrid renewable energy system using metaheuristic optimization
scs.2020.102255.
approaches – Case of a radio transmitter station in India. Energy Convers.
Alturki, F.A., Awwad, E.M., 2021. Sizing and cost minimization of standalone
Manag. 185, 339–352. http://dx.doi.org/10.1016/j.enconman.2019.01.107.
hybrid WT/PV/Biomass/Pump-Hydro storage-based energy systems. Energies
Dhiman, G., Kaur, A., 2019. STOA: A bio-inspired based optimization algorithm
14, 489. http://dx.doi.org/10.3390/en14020489.
for industrial engineering problems. Eng. Appl. Artif. Intell. 82, 148–174.
Anon, 2021a. Egypt - Renewable energy. Available online: https://www. http://dx.doi.org/10.1016/j.engappai.2019.03.021.
trade.gov/knowledge-product/egypt-renewable-energy. (Accessed 7 Febru- Diab, A.A.Z., Sultan, H.M., Mohamed, I.S., Kuznetsov, O.N., Do, T.D., 2019.
ary 2021). Application of different optimization algorithms for optimal sizing of
Anon, 2021b. Egypt - Renewable energy | privacy shield. Available online: https: PV/Wind/Diesel/Battery storage stand-alone hybrid microgrid. IEEE Access
//www.privacyshield.gov/article?id=Egypt-Renewable-Energy. (Accessed 7 7, 119223–119245. http://dx.doi.org/10.1109/access.2019.2936656.
February 2021). Dong, W., Li, Y., Xiang, J., 2016. Optimal sizing of a stand-alone hybrid power sys-
Anon, 2021c. MERRA - SoDa. Available online: https://www.soda-pro.com/web- tem based on battery/hydrogen with an improved ant colony optimization.
services/meteo-data/merra?fbclid=IwAR2vTObCUaC3DpZev3PqLX0FwV- Energies 9, http://dx.doi.org/10.3390/en9100785.
XATjk0E2qDqp1ZRCWIcVxBQBv6eeWTUA. (Accessed 1 April 2022). Duman, A.C., Güler, Ö., 2018. Techno-economic analysis of off-grid PV/wind/fuel
Anon, 2021d. The ministry of electricity and renewable energy. Available on- cell hybrid system combinations with a comparison of regularly and season-
line: http://www.moee.gov.eg/english_new/home.aspx. (Accessed 7 February ally occupied households. Sustain. Cities Soc. 42, 107–126. http://dx.doi.org/
2021). 10.1016/j.scs.2018.06.029.
Anon, 2021e. Renewable energy outlook: Egypt. Available online: El-Sattar, H.A., Kamel, S., Jurado, F., 2020a. Fixed bed gasification of corn stover
https://www.irena.org/publications/{2018}/Oct/Renewable-Energy-Outlook- biomass fuel: Egypt as a case study. Biofuels, Bioprod. Biorefining 14, 7–19.
{Egypt}. (Accessed 7 February 2021). http://dx.doi.org/10.1002/bbb.2044.

9526
H.A. El-Sattar, S. Kamel, H.M. Sultan et al. Energy Reports 8 (2022) 9506–9527

El-Sattar, H.A., Kamel, S., Tawfik, M.A., Vera, D, 2016. Modeling of a downdraft Maleki, A., Nazari, M.A., Pourfayaz, F., 2020. Harmony search optimization for
gasifier combined with externally fired gas turbine using rice straw for optimum sizing of hybrid solar schemes based on battery storage unit.
generating electricity in Egypt. In: Proceedings of the 2016 Eighteenth Energy Rep. http://dx.doi.org/10.1016/j.egyr.2020.03.014.
International Middle East Power Systems Conference. MEPCON, IEEE, pp. Maleki, A., Pourfayaz, F., 2015. Sizing of stand-alone photovoltaic/wind/diesel
747–752. system with battery and fuel cell storage devices by harmony search
El-Sattar, H.A., Kamel, S., Tawfik, M.A., Vera, D., Jurado, F., 2018. Modeling algorithm. J. Energy Storage 2, 30–42. http://dx.doi.org/10.1016/j.est.2015.05.
and simulation of corn stover gasifier and micro-turbine for power gen- 006.
eration. Waste Biomass Valoriz. 10, 1–14. http://dx.doi.org/10.1007/s12649- Mills, A., Al-Hallaj, S., 2004. Simulation of hydrogen-based hybrid systems using
018-0284-z. Hybrid2. Int. J. Hydrogen Energy 29, 991–999. http://dx.doi.org/10.1016/j.
El-Sattar, H.A., Kamel, S., Vera, D., Jurado, F., 2020b. Tri-generation biomass ijhydene.2004.01.004.
system based on externally fired gas turbine, organic rankine cycle and Mirjalili, S., 2016. SCA: A Sine cosine algorithm for solving optimization prob-
absorption chiller. J. Clean. Prod. 260. http://dx.doi.org/10.1016/j.jclepro. lems. Knowledge-Based Syst. 96, 120–133. http://dx.doi.org/10.1016/j.knosys.
2020.121068. 2015.12.022.
Eteiba, M.B., Barakat, S., Samy, M.M., Wahba, W.I., 2018. Optimization of an off- Mirjalili, S., Lewis, A., 2016. The whale optimization algorithm. Adv. Eng. Softw.
grid PV/Biomass hybrid system with different battery technologies. Sustain. 95, 51–67. http://dx.doi.org/10.1016/j.advengsoft.2016.01.008.
Cities Soc. 40, 713–727. http://dx.doi.org/10.1016/j.scs.2018.01.012. Naderipour, A., Abdul-Malek, Z., Zahedi Vahid, M., Mirzaei Seifabad, Z., Ha-
Fathy, A., 2016. A reliable methodology based on mine blast optimization jivand, M., Arabi-Nowdeh, S., 2019. Optimal, reliable and cost-effective
algorithm for optimal sizing of hybrid PV-wind-FC system for remote area in framework of photovoltaic-wind-battery energy system design considering
Egypt. Renew. Energy 95, 367–380. http://dx.doi.org/10.1016/j.renene.2016. outage concept using grey wolf optimizer algorithm - Case study for
04.030. Iran. IEEE Access 7, 182611–182623. http://dx.doi.org/10.1109/ACCESS.2019.
Fathy, A., Kaaniche, K., Alanazi, T.M., 2020. Recent approach based social spider 2958964.
optimizer for optimal sizing of hybrid PV/Wind/Battery/Diesel integrated Ray, M., Sheeran, R., Lewis, E., Cavanagh, J., 2007. The climate change Levy and
microgrid in Aljouf region. IEEE Access 8, 57630–57645. http://dx.doi.org/ climate change agreements a review by the national audit office. Natl. Audit
10.1109/ACCESS.2020.2982805. Off..
Food and Agriculture Organization of the United Nations, 2017. BEFS ASSESS- Samy, M.M., Barakat, S.H., 2020. Biomass and fuel cells multi-objective optimiza-
MENT for EGYPT: Sustainable Bioenergy Options from Crop and Livestock tion of hybrid renewable energy system based on biomass and fuel cells.
Residues. ISBN: 9789251095874. http://dx.doi.org/10.1002/er.5815.
Gao, Z., Zhao, J., Li, S., Hu, 2020. The improved mayfly optimization algorithm Sanajaoba, S., 2019. Optimal sizing of off-grid hybrid energy system based on
the improved mayfly optimization algorithm. J. Phys. Conf. Ser. 1684, 012077. minimum cost of energy and reliability criteria using firefly algorithm. Sol.
http://dx.doi.org/10.1088/1742-6596/1684/1/012077. Energy 188, 655–666. http://dx.doi.org/10.1016/j.solener.2019.06.049.
Geleta, D.K., Manshahia, M.S., Vasant, P., Banik, A., 2020. Grey wolf optimizer for Shi, R., Wang, W., Yuan, Z., Fan, X., Ramezani, E., 2021. A novel optimum
optimal sizing of hybrid wind and solar renewable energy system. Comput. arrangement for a hybrid renewable energy system using developed student
Intell. 1–30. http://dx.doi.org/10.1111/coin.12349. psychology based optimizer: A case study. Energy Rep. 7, 70–80. http:
Gharibi, M., Askarzadeh, A., 2019. Technical and economical bi-objective design //dx.doi.org/10.1016/j.egyr.2020.11.168.
of a grid-connected photovoltaic/diesel generator/fuel cell energy system. Singh, A., Baredar, P., 2016. Techno-economic assessment of a solar PV, fuel
Sustain. Cities Soc. 50, 101575. http://dx.doi.org/10.1016/j.scs.2019.101575. cell, and biomass gasifier hybrid energy system. Energy Rep. 2, 254–260.
Ghosh, S., Karar, V., 2018. Assimilation of optimal sized hybrid photovoltaic- http://dx.doi.org/10.1016/j.egyr.2016.10.001.
biomass system by dragonfly algorithm with grid. Energies 11, http://dx.doi. Singh, S., Kaushik, S.C., 2016. Optimal sizing of grid integrated hybrid PV-biomass
org/10.3390/en11071892. energy system using artificial bee colony algorithm. IET Renew. Power Gener.
Guo, X., Yan, X., Jermsittiparsert, K., 2021. Using the modified mayfly algorithm 10, 642–650. http://dx.doi.org/10.1049/iet-rpg.2015.0298.
for optimizing the component size and operation strategy of a high temper- Sultan, H.M., Kuznetsov, O.N., Menesy, A.S., Kamel, S., 2020. Optimal configu-
ature PEMFC-powered CCHP. Energy Rep. 7, 1234–1245. http://dx.doi.org/10. ration of a grid-connected hybrid PV/Wind/Hydro-Pumped storage power
1016/j.egyr.2021.02.042. system based on a novel optimization algorithm. In: Proc. 2nd 2020 Int.
Hadidian Moghaddam, M.J., Kalam, A., Nowdeh, S.A., Ahmadi, A., Babanezhad, M., Youth Conf. Radio Electron. Electr. Power Eng.. REEPE 2020, http://dx.doi.
Saha, S., 2019. Optimal sizing and energy management of stand-alone hybrid org/10.1109/REEPE49198.2020.9059189.
photovoltaic/wind system based on hydrogen storage considering LOEE and Sultan, H.M., Menesy, A.S., Kamel, S., Korashy, A., Almohaimeed, S.A., Abdel-
LOLE reliability indices using flower pollination algorithm. Renew. Energy Akher, M., 2021. An improved artificial ecosystem optimization algorithm
135, 1412–1434. http://dx.doi.org/10.1016/j.renene.2018.09.078. for optimal configuration of a hybrid PV/WT/FC energy system. Alexandria
Hassan, H.B.A., el Gebaly, M.R., Ghani, S.S.A., Hussein, Y.M.M., 2014. An economic Eng. J. 60, 1001–1025. http://dx.doi.org/10.1016/j.aej.2020.10.027.
study of recycling agricultural wastes in Egypt. Middle East J. Agric. Res. 3, Suresh, M., Meenakumari, R., 2019. An improved genetic algorithm-based
592–608. optimal sizing of solar photovoltaic/wind turbine generator/diesel genera-
Jamshidi, M., Askarzadeh, A., 2019. Techno-economic analysis and size optimiza- tor/battery connected hybrid energy systems for standalone applications. Int.
tion of an off-grid hybrid photovoltaic, fuel cell and diesel generator system. J. Ambient Energy 1–20. http://dx.doi.org/10.1080/01430750.2019.1587720.
Sustain. Cities Soc. 44, 310–320. http://dx.doi.org/10.1016/j.scs.2018.10.021. Tabak, A., Kayabasi, E., Guneser, M.T., Ozkaymak, M., 2019. Grey wolf opti-
Kashefi Kaviani, A., Riahy, G.H., Kouhsari, S.M., 2009. Optimal design of a mization for optimum sizing and controlling of a PV/WT/BM hybrid energy
reliable hydrogen-based stand-alone wind/PV generating system, considering system considering TNPC, LPSP, and LCOE concepts. Energy Sources, Part
component outages. Renew. Energy 34, 2380–2390. http://dx.doi.org/10. A Recover. Util. Environ. Eff. 1–21. http://dx.doi.org/10.1080/15567036.2019.
1016/j.renene.2009.03.020. 1668880.
Khan, A., Javaid, N., 2020. Optimal sizing of a stand-alone photovoltaic, wind Vendoti, S., Muralidhar, M., Kiranmayi, 2021. Techno-economic analysis of off-
turbine and fuel cell systems. Comput. Electr. Eng. 85, http://dx.doi.org/10. grid solar/wind/biogas/biomass/fuel cell/battery system for electrification in
1016/j.compeleceng.2020.106682. a cluster of villages by HOMER software. Environ. Dev. Sustain. 23, 351–372.
Kharrich, M., Kamel, S., Abdeen, M., Mohammed, O.H., Akherraz, M., Khur- http://dx.doi.org/10.1007/s10668-019-00583-2.
shaid, T., Rhee, S.B., 2021a. Developed approach based on equilibrium Wang, Y., Chen, K.S., Mishler, J., Cho, S.C., Adroher, X.C., 2011. A review of
optimizer for optimal design of hybrid PV/Wind/Diesel/Battery microgrid polymer electrolyte membrane fuel cells: Technology, applications, and
in Dakhla, Morocco. IEEE Access 9, 13655–13670. http://dx.doi.org/10.1109/ needs on fundamental research. Appl. Energy 88, 981–1007. http://dx.doi.
ACCESS.2021.3051573. org/10.1016/j.apenergy.2010.09.030.
Kharrich, M., Mohammed, O.H., Alshammari, N., Akherraz, M., 2021b. Multi- Wang, Y., Seo, B., Wang, B., Zamel, N., Jiao, K., Adroher, X.C., 2020. Fundamentals,
objective optimization and the effect of the economic factors on the design materials, and machine learning of polymer electrolyte membrane fuel
of the microgrid hybrid system. Sustain. Cities Soc. 65, 102646. http://dx.doi. cell technology. Energy AI 1, 100014. http://dx.doi.org/10.1016/j.egyai.2020.
org/10.1016/j.scs.2020.102646. 100014.
Li, J., Liu, P., Li, Z., 2020. Optimal design and techno-economic analysis of a solar- Wolpert, D.H., Macready, W.G., 1997. No free lunch theorems for optimization.
wind-biomass off-grid hybrid power system for remote rural electrification: IEEE Trans. Evol. Comput. 1, 67–82.
A case study of west China. Energy 208, 118387. http://dx.doi.org/10.1016/ Zervoudakis, K., Tsafarakis, S., 2020. A mayfly optimization algorithm. Comput.
j.energy.2020.118387. Ind. Eng. 145, http://dx.doi.org/10.1016/j.cie.2020.106559.

9527