Optimal Configuration of a Grid-Connected Hybrid

PV/Wind/Hydro-Pumped Storage Power System

Based on a Novel Optimization Algorithm
Hamdy M. Sultan1, Oleg N. Kuznetsov2 Ahmed S. Menesy Salah Kamel
dept. of Electrical Power Systems dept. of Electrical Engineering dept. of Electrical Engineering
National Research University “MPEI” Minia University Aswan University
Moscow, Russia Minia, Egypt Aswan, Egypt
[email protected], [email protected]
[email protected]
[email protected]

Abstract—In this paper, a novel metaphor-less Rao the generated electrical energy from the hybrid system is
optimization technique is developed to solve the optimization higher than the load demand, the stored water in the lower
problem of optimal configuration of a hybrid system consisting reservoir will be pumped to the higher tank. During peak
of solar photovoltaic (PV), wind turbines and a hydroelectric hours of load, the stored water in the upper tank will fall to
pumped storage (HPS) system. Furthermore, the optimal design the lower reservoir through the hydraulic turbine producing
of this system was conducted based on real-time meteorological
data of Ataka area, Egypt. Minimizing the cost of energy (COE)
electric energy [7]. The HPS system is studied in several
delivered from the proposed system is the essential goal of stand-alone micro-grids and hybrid systems applications [8].
addressed optimization problem. In addition, statistical Also, the economic operations for stand-alone hybrid systems
measurements based on various metrics has been studied to with various energy storage systems types are studied in [7,
validate the accuracy of developed Rao optimization technique. 8], the previous studies proved that the reliability of HPS
Moreover, the simulation results obtained from the developed system is better than battery storage system. Furthermore,
technique are compared with other recent optimization size optimization, modeling and economic analysis of
techniques. Finally, from the obtained simulation results, the PV/pumped storage system have been presented in [9].
developed technique proved effectiveness as a good competitor Recently, several recent metaheuristic optimization
to other optimization techniques in optimal sizing of grid-
connected hybrid system. The simulation process of proposed
techniques have been applied for the optimal sizing of various
algorithm has been implemented using MATLAB software. types of hybrid systems, namely, cuckoo search optimization
algorithm [10], multi-objective self-adaptive differential
Keywords—Hybrid system, Hydro-pumped storage, COE, evolution (MOSaDE) algorithm [11], hybrid particle swarm-
LPSP, Optimization, Rao algorithm. gravitational search algorithm (PSOGSA) [12], particle
swarm optimization (PSO). In Egypt, the load demand per
I. INTRODUCTION day significantly fluctuates and the evening peak load
Although most of the produced electrical energy all over demand is 1.5 times more than the off-peak hours’ load
the world was generated from fossil fuels, they have a demand [13].
negative impact on the environment and the health of In this paper, a recent optimization technique called Rao
human[1, 2]. Due to the disadvantages of fossil fuels in has been proposed to solve the optimization problem of
power applications, it is necessary to find another alternative optimizing the hybrid PV/wind/HPS system of a practical
power sources that have characteristics better than the case study of Egypt. Minimizing the cost of energy (COE)
conventional sources like little cost, clean and high reliability. supplied by the proposed hybrid system is the main objective
The most attractive and clean sources of renewable energy of the studied optimization problem. The following design
are solar photovoltaic (PV), wind turbine and hydropower. constraints must be taken into consideration: acceptable level
Each one of these sources have it is own characteristics and for the power supply reliability, low fluctuations level in
operations, but most of them are suitable in low power energy exchange with the national network, and
application and isolated areas]3[ . Generally, the renewable complimentary use of PV and wind sources. Furthermore,
energy sources such as PV and wind turbine have many statistical analysis is performed to evaluate the stability and
fluctuations during their individually operation in the system, reliability of the proposed Rao technique. In addition, a
and this will cause instability in the system]4[ . Therefore, to comprehensive comparison between the proposed technique
overcome the fluctuations that cause increase in the level of and other recent metaheuristic optimization techniques has
generated voltage and energy storage system size as well as been studied. The rest of the paper is organized as follow;
increasing the system reliability, the individual PV and wind Section II introduces modelling of the sub-systems of the
sources are used in one system, namely, hybrid PV/wind hybrid energy system, the optimization problem is addressed
system]5[ . in Section III, the proposed Rao optimization algorithm is
To solve the problem of irregular nature of renewable briefly reported in Section IV, results of validation of the Rao
energy sources, several types of energy storage technologies algorithm and statistical measurements are provided in
can be used like fuel cells, batteries, compressed air, super Section V, conclusions of the present work are given in
capacitors and hydroelectric pumped storage (HPS) [6]. Section VI.
Newly, HPS is considered the most famous and widely used
II. MODELLING OF THE PROPOSED HYBRID SYSTEM
type of energy storage systems. HPS can be used with a high
efficiency in the range of 70–80% and its capacity is The proposed hybrid renewable energy system includes
independent of the water flow variation during the year. when solar PV power plant, wind farm, and a double stock

978-1-7281-5655-2/20/$31.00 ©2020 IEEE

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
hydraulic pumped storage power plant. Fig. 1 presents a
schematic description of the proposed system. The hybrid 0
 2 2 v (t )  vcut in
system is a grid-connected system. The demand for electric
energy is principally supplied by the renewable energy  NWT WT PWT _ rated (v (t )  vcut in ) vcut in  v (t )  vrated
sources and the utility grid helps only in the cases of deficit  2 2
PWT   (vrated  vcut in )
of electric energy. All possible cases of operation of the vrated  v (t )  vcut off
system are summarized as follow: 
 When the energy supplied by the PV and wind  NWT WT PWT _ rated v (t )  vcut off
0
power plants exceeds the load, the excess energy is 
delivered to the motor-pump set to raise up sea (2)
water to the tank until it full with water, after that where, NWT is the number of wind turbines in the wind farm,
the surplus energy is sold to the external network. ηWT denotes the efficiency of wind farm, PWT_rated is the
 Oppositely, when the renewable energy supplied by nominal capacity of the wind turbine at the rated speed of
the renewable resources is less than the load wind vrated, vcut-in and vcut-off denote the wind speeds at which
requirements, the energy shortage is covered by the wind generating system starts operation and closed for
discharging the water from the upper tank and safety reasons, respectively.
supplied to the turbine-generator set. If the level of
water in the tank reaches it minimum allowable C. Hydro-pumped Storage System
value, the energy deficit is purchased from the grid In this study the storage system consists of only one upper
utility. reservoir, which is assumed to have the cubic shape for easy
mathematical calculations. The depth of the tank is assumed
to be h2, while the effective head of water over the turbine-
generator set is variable with time due to the charging and
discharging process of the upper reservoir during the day and
nighttime. The additional head of water hadd depends on the
volume of water stored in the reservoir V.
The difference between the renewable generation from
the PV and wind power plants with respect to the load
demand ED, EB, determines the way that the hydraulic system
will act. The energy balance EB can be calculated from the
following expression[14],
EB  EPV  EWT  ED (3)

Equations (4) and (5) describe the operation of the

pumped storage system in the generating and pumping
modes, respectively [15].
V (t  1)
E H (t )  min[min( ; QT )T WP  g ( hadd  h3 ); E B ] (4)
3600
Fig. 1. Schematic configuration of the system under consideration.
V  V (t  1) (5)
E pump (t )  min[min( max ; QP ) PWP  g ( hadd  h3 ); E B ]
A. PV System Modelling 3600
The energy supplied by the proposed PV power plant at a where, EH and Epump denote the energy produced by the
certain intensity of solar irradiation and air temperature is turbine-generator and energy drawn by the motor-pump set,
mathematically calculated from the following formula [5, respectively. ηT, ηP, and ηWP are the efficiencies of the turbine
14], system, pumping system, and pipeline, respectively. QT and
E PV (t)  N PV  PPV _ rated   PV QP present the discharging and charging rates of the turbine
and pump, respectively (m3/s). Vmax denotes the volume of the
G (t ) (1) proposed reservoir.
 (1  T (TCell (t )  Tcell _ nom )) When the hydraulic plant is working in the generating
Gnom phase, the volume of water fall down from the tank in an
where, NPV denotes the number of PV modules in the power interval of time (Δt = 1 hour) can be calculated as given in (6)
plant, PPV_rated denotes the nominal power of the PV module and if the volume of water stored in the reservoir reached the
at standard conditions of temperature and solar irradiation, minimum limit, then the deficit covered from the grid utility
ηPV is the efficiency of the PV system including the inverter system is calculated according to (7).
efficiency and also that of the wiring system, G(t) is the solar E H (t )
voldis (t )  (6)
irradiation at any period of time t, Gnom, Tcell denotes the cell T WP  g ( hadd  h3 )
temperature during a certain period of time t, and Tcell_nom
refers to the ambient cell temperature (25°C).
Edef (t )  E B (t )  E H (t ) (7)
B. Wind Farm Modelling
Based on the basics of aerodynamics and wind energy, the On the other hand, when the station is working like a
power generated from the wind farm at a certain wind speed pump, the volume of water raised up in one hour of operation
v(t) can be mathematically expressed as follow [12], is calculated from (8). When the renewable energy generation

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
exceeds the energy needed for the load and the pumping 8760
i 1 [Pload (t i )  (PPV (t i )  PWT (t i )  PH (t i ))]
system, the surplus energy supplied to grid is calculated from LPSP  (15)
(9). 8760
i 1 Pload (t i )
E pump (t ) The exchange of electric energy with the network utility
vol pump (t )  (8)
 PWP  g ( hadd  h3 ) is constrained by the predefined maximum and minimum
limits,
Pg _ min  Pg  Pg _max (16)
ESur (t )  E B (t )  E pump (t ) (9)
The level of fluctuation in the energy exchanged Dgs with
III. FORMULATION OF THE OPTIMIZATION PROBLEM the grid is limited by the value βg and the standard deviation
must be kept minimum [18],
The main object of this study is optimal sizing of the
Pgs _ max  Pgs _ min
capacities of each subsystem of the proposed hybrid Dgs  (17)
renewable energy system in order to receive the electric t
energy from this system with the minimum cost of energy
(COE). At the same time, the reliability of the power supply 2
system must be within the allowable limits (LPSP < βL),
STD 

 iN1 Pgs (t i )  Pgs  (18)
which is taken as 0.05 in this study. In addition, the level of
N 1
energy fluctuation between the hybrid energy system and the
external utility grid within the acceptable limits. Moreover, a where, Pgs_max and Pgs_min present the maximum and minimum
satisfactory use of the allowable renewable energy resources levels of energy fed to grid. Finally, to ensure an acceptable
has to be fulfilled. level of renewable sources utilization, the relative fluctuation
The annual cost Cann_tot of the proposed system includes rate Dload have to be kept around unity, which proves that the
the annual interest of the capital investment costs Cann_cap, the renewable generation is close to the requirements of the load
annual cost due to replacement Cann_rep if the lifetime of the [18],
subsystem is less than that of the whole system, the annual
1
 
2
cost for operation and maintenance Cann_O&M, the penalty cost  iN1 PPV (t i )  PWT (t i )  Pload (t i )
Cpc, which is added as a fine if the reliability or the fluctuation N
level are not in the acceptable ranges. And finally the cost of Dload  (19)
Pload
kWh units purchased and sold to the grid Cgp and Cgs.
The net present cost (NPC) for the proposed energy The objective function proposed in this study is described
system is given in the following expression [16], in (20), which provides the goal of the study to minimize the
Cann _ tot cost of energy (COE) and at the same time controls the
NPC  (10) predefined design constraints,
CRF min f x  min(0.94  COE  0.02  STD
where, CRF denotes the capital recovery factor, which is (20)
0.02  Dload  0.01  Dgs  0.01  LPSP )
calculated mathematically as in (11),
M
r (1  r ) where, x denotes a vector of the parameters have to be
CRF ( r , M )  M (11) optimized (NPV, NWT, Vmax, QP, and QT).
(1  r ) 1
IV. RAO ALGORITHM
where, r presents the rate of interest and M denote the lifetime
of the proposed hybrid system. The COE, which presents the Rao Algorithm is a novel metaphor-less, which is based
cost of the kWh units ($/kWh) generated from the system on the optimal and worst solutions given from the process of
under consideration is calculated as follows [12, 16], optimization as well as the randomly evaluated interactions
among the candidate solutions with each other. The merits of
Cann _ tot NPC
COE  8760  8760  CRF (12) this technique that is requires only common specifications of
1 Pload 1 Pload the algorithm such as the number of iterations and the
population size, while any other specific parameters are not
The optimization of the capacities of the hybrid system included in the algorithm. The proposed algorithm introduces
components will be performed under a group of constraints; three different metaphor-less optimization methods [17].
first of all, the volume of water stored in the tank must be Assume that the objective function that have to be
with the upper and lower limits that have been determined, minimized is f(x). During the optimization process, at any
iteration ‘i’, the number of variables have to be optimized is
Vmin  SOC (t )  Vmax (13)
‘m’ and the number of solutions is ‘n’. The value of the best
The LPSP, which is an indication of the reliability of the candidate solution is given as f(x)best, while the value of the
system must be less than a predefined value βL, 0.05 in this worst candidate solution within the provided candidate
case study. solutions is f(x)worst. The value of variable number j for the
candidate solution number k, in the running iteration number
LPSP   L (14) i is given as Xj,k,i. This value will be modified according to the
following equations for the proposed three algorithms [17]:
where, the LPSP is calculated as follows,
'

X j ,k ,i  X j ,k,i  r1, j ,i X j ,best ,i  X j , worst ,i  (21)

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
 '

X j ,k ,i  X j ,k,i  r1, j ,i X j ,best ,i  X j , worst ,i  Rao optimization algorithm. According to the plan of the
renewable energy sector in the ministry of electricity and
(22)
 
energy in Egypt, Ataka region in Suez city on the shore of the
 r2, j ,i X j ,k,i orX j ,l ,i  X j ,l ,i orX j ,k,i Red Sea (latitude 30.0, longitude 32.5) is selected as the
location of the proposed hybrid system. The hydro-pumped
'

X j ,k ,i  X j ,k,i  r1, j ,i X j ,best ,i  X j , worst ,i  storage system is suggested to be constructed on the
mountains along the shore of the sea and sea water is used for
(23) operation. According to the reports of the Egyptian
 
 r2, j ,i X j ,k,i orX j ,l ,i  X j ,l ,i orX j ,k,i  Electricity Holding Company, most of the planned projects
of renewable energy in the future are suggested in the Suez
where, Xj,best,i and Xj,worst,i denote the the value of the variable Gulf region thanks to its suitable environmental conditions of
j for the best and worst candidate solutions in the running solar irradiation and speed of wind. Fig. 3 presents the
iteration i, respectively. while r1,j,i and r2,j,i denote randomly monthly averaged values of input parameters in the selected
distributed numbers between [0, 1] for the control variable j site; solar potential, wind speed, and temperature. The system
in the iteration number i. cost included in the optimization problem are divided into;
The term (Xj,k,i or Xj,l,i) included in (22) and (23) refers annual interest of initial cost for installation, running
that the value of the kth candidate solution is compared with operating and maintenance cost, cost of replacements, cost of
the value of another candidate solution l, which is selected kWh units sold and bought to the external grid utility, and the
randomly. Based on the fitness value of this term the penalty cost, which presents the shortage in the reliability of
information is updated. Thus, if the value of fitness function the whole supply system in covering the load demand.
belongs to the candidate solution k is better than that of the lth Based on the data introduced in Fig. 3, the selected region
candidate, then the term in (22) and (23) will be updated to for the proposed system is characterized by an average solar
be Xj,k,i. else while, if the fitness of the candidate solution l is potential of 5.69 kWh/m2/day on the horizontal surface and
better that the corresponding value of the kth candidate average wind speed of 6.72 m/s. Table I provided the design
solution, the term will be changed to Xj,l,i. The same principle and economic specifications of all subsystems involved in the
is applied for the second term (Xj,l,i or Xj,k,i). the proposed proposed energy system. In the current work, the lifespan of
three algorithms (Rao-1, Rao-2, and Rao-3) introduced in this the whole system is taken as 25 years, while the rate of
paper are depending only on the best and worst candidate interest is assumed to be 6%.
solutions as well as the interactions between the candidates, MATLAB simulation package is used for validating the
which is randomly selected. The flowchart describing the proposed optimization technique for optimal sizing and
procedure of the proposed Rao optimization technique is design of the proposed hybrid energy system. The
presented in Fig. 2. optimization program was launched 40 individual times and
the optimal design of the system is taken based on the run,
which result in the minimum value of the proposed objective
function. The three proposed algorithms Rao-1, Rao-2, and
Rao-3 are operated and statistical measurements based on
different metrics is demonstrated, from which the best
algorithm is determined. For all the algorithms involved in
this study, the maximum number of iterations is adjusted at
100 iterations, while the number of search agents is taken as
20 agents.

Fig. 3. Monthly averaged values of input parameters in the selected site;

solar potential, wind speed, and temperature.
Fig. 2. Flowchart of the proposed Rao algorithm.

V. RESULTS AND DISCUSSION

In this section, a case study of a renewable PV and wind
power system including hydro-pumped storage system is
introduced for validating the effectiveness of the proposed

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
 TABLE I. SPECIFICATIONS OF THE COMPONENTS OF THE SYSTEM. in Table III for the three algorithms. The results obtained
PV module parameters WT parameters from three algorithms are compared with that obtained from
the application of other recent optimization techniques,
Model PV- Model Fuhrländer
MLT260HC FL 30 Whale optimization algorithm (WOA) and Salp swarm
Max. power 260 Watt Rated power 30 kW algorithm (SSA) [15], used for optimal design of the same
Length 1625 mm Swept area 13 m2 hybrid energy system. Within all optimization techniques,
Width 1019 mm Cut-in speed 2.5 m/s Rao-2 reached the minimum fitness value, which results in
Thickness 46 mm Rated speed 12 m/s the minimum COE and least NPV at an acceptable value of
Operating 47 ºC Cut-off speed 25 m/s the reliability index less than the standard value (LPSP =
temperature 0.01079 < 0.05).
Temperature 0.45% Initial cost 58564.79 $
coefficient
Initial cost 112 $ O&M cost 3%
O&M cost 1% Replacement 34553.226
cost $
Lifetime 25 year Lifetime 20 year
Turbine-generator set Pump-motor set Parameters
parameters
Power 1 kW Rated power 45 kW
Efficiency 86 % Efficiency 86 %
Initial cost 1000 $ Initial cost/kW 10749 $
Replacement 750 $ Replacement 10749 $
cost cost
O&M cost 1% O&M cost 1%
Lifetime 10 year Lifetime 10 year (a)
Parameters of the tank
Initial cost/m3 170 $ Lifetime 35 year

A. Statistical Study
In this section, the accuracy and stability of the proposed
Rao optimization algorithms are examined. The optimization
program of optimal design of the suggested renewable energy
system has been operated 40 individual runs and the end
values of the fitness function for all executions have been
reported. The convergence curves of the objective function
for Rao-1, Rao-2, and Rao-3 are shown in Fig. 4. Parametric
and non-parametric statistical study upon the reported values
of the fitness functions for the three algorithms are listed in (b)
Table II. The statistical study and comparisons are based on
different metrics such as, best minimum value of the
objective function, maximum worst values, average value of
the fitness function. The stability of the proposed algorithms
is measured based on various metrics such as relative error
(RE), mean absolute error (MAE), root mean square error
(RMSE), and efficiency (Eff). The values of these metrics for
all algorithms are shown in Table II. Based on the obtained
results, the three algorithms proved their stability of
operation, while Rao-2 algorithm gives the best minimum
value of the objective function within the three algorithms.
B. Results of Application of the Proposed Algorithm
The optimal configuration of the proposed hybrid (c)
renewable energy system includes the number of PV modules Fig. 4. Convergence curves for the 40 executions for proposed optimization
included in the PV power plant, the number of wind turbine algorithms; (a) Rao-1, (b) Rao-2, (c) Rao-3.
generators in the suggested wind farm, capacity of the
hydraulic pumped storage system, which contains three main TABLE II. COMPARISON OF THE STATISTICAL RESULTS OF THE RAO-1,
parts; the volume of only the upper tank as the lower one is RAO-2, AND RAO-3.
taken as the sea water, the rating of the turbine-generator unit, Rao-1 Rao-2 Rao-3
and the capacity of the motor-pump unit. The convergence Min 0.4291344470460 0.4291286744737 0.4291287268953
curves of the objective function of the three algorithms for Max 0.4292780869238 0.4292037804796 0.4395966208002
the executions resulted in the minimum value of the objective Mean 0.4291937003471 0.4291855142802 0.4297385458317
function are shown in Fig. 5. It is clearly evidence from this Median 0.4292019693242 0.4292017748325 0.4292004431812
SD 0.0023557450639 0.0023810178252 0.2301342762122
figure that Rao-2 algorithm successes to reach the best value RE 0.005523052414 0.0052981597200 0.0568425181747
of the fitness function in 84 iterations. The values of the MAE 0.000059253301 0.0000568398064 0.0006098189364
optimized sizes of the subsystems as well as other design RMSE 0.000063655583 0.0000615102842 0.0023527968848
factors such as COE, LPSP, NPV, and STD are summarized Eff. 99.98619456889 99.986756654936 99.860828965475

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
C. Application of the Optimal Design PV and wind systems and the water in the tank reaches the
The optimal size of the system components obtained from minimum allowable limit) energy exchanged with the grid
Rao-2 algorithm listed in Table III are applied to the model utility are presented in Fig. 6(b). For clear understanding of
of the hybrid renewable energy system. The operation of the the operation of the system under study, 24-hour operation
whole system under the application of the obtained values for for a certain day in the summer is shown in Fig. 7.
a complete year with one-hour interval is shown in Fig. 6. The
annual load demand, annual production from the PV power
plant and wind farm as well as the difference between the
renewable energy generated and the consumption are shown
in Fig. 6(a). While Fig. 6(b) shows the power absorbed by the
motor-pump unit to catch up the water to the tank during the
periods of excess power from PV and wind generating
stations, the power produced from the turbine-generator set
when the load requirements exceed the power delivered by
the power plants (PV and wind farms). The volume of water
raised to the upper reservoir during the excess power hours
and water discharged in the low power hours are also shown
in Fig. 6(b). The state of charge (SOC) of the upper reservoir,
which reflects the ratio of the water stored in the tank with
respect to the its maximum capacity is shown in Fig. 6(b).
Finally, the surplus (when the renewable generation exceeds
the load demand and the reservoir is full of water) and deficit Fig. 5. Convergence curves for Rao-1, Rao-2, and Rao-3.
(when the load demand is greater than the generation from

TABLE III. OPTIMIZATION RESULTS OF THE THREE ALGORITHMS COMPARED WITH RECENT OPTIMIZATION TECHNIQUES

Rao -1 Rao -2 Rao -3 WOA [18] SSA [18]

Best objective function 0.42913444704 0.42912867447376 0.42912872689535 0.4343083313823 0.434313141546035
PV (units) 1095.84447930 1095.72701606910 1095.72754414961 1164.63546710256 1167.28803229316
Wind (units) 10.0437584648 10.0444018120054 10.0444017411556 49.9799457125606 49.9936243045239
Best
Reservoir (m3) 10000 10000 10000 2500 2500
solution
3
Pump (m /s) 0.200 0.200 0.200 0.2 0.2
Turbine (m3/s) 0.200 0.200 0.2000 0.2 0.200000019043594
COE 0.13597455263 0.13594291878508 0.13594790824843 0.216089326188 0.216225093593638
NPV 3282897.34325 3282133.59248917 3282343. 6809451 6329357.5510906 6333334.24195501
LPSP 0.01079803765 0.01081680945102 0.01080180664347 0.00085956003946 0.000858397281621
Dgs 2.87467932366 2.87467932366818 2.87467932366818 3.48751477512181 3.48751477512181
STD 12.3841870436 12.3865177620646 12.3856647318195 8.88109547057805 8.87514581663334
DLoad 1.31619601342 1.31518997116154 1.31712998601186 0.93364597696127 0.940500816521039

(a) (b)

Fig. 6. A complete year (8760 hours) operation based on the results of Rao-2 algorithm.

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.
 algorithm to reach the best optimal value of COE. Finally, the
obtained results ensured that the designed hybrid system
provides an acceptable value of COE and LPSP. It is
recommended for repeating such hybrid systems for
achieving the desired penetration level of renewable
generation in the plans of the ministry of electricity and
energy in Egypt. The optimization model introduced in this
paper universal and could be applied on any case study.
REFERENCES
[1] A. Arabali, M. Ghofrani, M. Etezadi-Amoli, and M. S. Fadali,
"Stochastic performance assessment and sizing for a hybrid power
system of solar/wind/energy storage," IEEE Transactions on
Sustainable Energy, vol. 5, no. 2, pp. 363-371, 2013.
[2] M. Goedecke, S. Therdthianwong, and S. H. Gheewala, "Life cycle
cost analysis of alternative vehicles and fuels in Thailand," Energy
Policy, vol. 35, no. 6, pp. 3236-3246, 2007.
[3] K. Kusakana and H. J. Vermaak, "Hybrid renewable power systems for
mobile telephony base stations in developing countries," Renewable
Energy, vol. 51, pp. 419-425, 2013.
[4] M. Elhadidy and S. Shaahid, "Parametric study of hybrid (wind+ solar+
diesel) power generating systems," Renewable energy, vol. 21, no. 2,
pp. 129-139, 2000.
[5] J. L. Bernal-Agustín and R. Dufo-Lopez, "Simulation and optimization
(a) of stand-alone hybrid renewable energy systems," Renewable and
Sustainable Energy Reviews, vol. 13, no. 8, pp. 2111-2118, 2009.
[6] O. Kizilkan and I. Dincer, "Borehole thermal energy storage system for
heating applications: Thermodynamic performance assessment,"
Energy Conversion and Management, vol. 90, pp. 53-61, 2015.
[7] T. Ma, H. Yang, and L. Lu, "Feasibility study and economic analysis of
pumped hydro storage and battery storage for a renewable energy
powered island," Energy Conversion and Management, vol. 79, pp.
387-397, 2014.
[8] T. Ma, H. Yang, L. Lu, and J. Peng, "Technical feasibility study on a
standalone hybrid solar-wind system with pumped hydro storage for a
remote island in Hong Kong," Renewable energy, vol. 69, pp. 7-15,
2014.
[9] T. Ma, H. Yang, L. Lu, and J. Peng, "Optimal design of an autonomous
solar–wind-pumped storage power supply system," Applied Energy,
vol. 160, pp. 728-736, 2015.
[10] M. A. Mohamed, A. M. Eltamaly, A. I. Alolah, and A. Hatata, "A novel
framework-based cuckoo search algorithm for sizing and optimization
of grid-independent hybrid renewable energy systems," International
journal of green energy, vol. 16, no. 1, pp. 86-100, 2019.
[11] M. A. Ramli, H. Bouchekara, and A. S. Alghamdi, "Optimal sizing of
PV/wind/diesel hybrid microgrid system using multi-objective self-
adaptive differential evolution algorithm," Renewable energy, vol. 121,
pp. 400-411, 2018.
[12] A. A. Z. Diab, H. M. Sultan, I. S. Mohamed, O. N. Kuznetsov, and T.
D. Do, "Application of Different Optimization Algorithms for Optimal
Sizing of PV/Wind/Diesel/Battery Storage Stand-Alone Hybrid
Microgrid," IEEE Access, vol. 7, pp. 119223-119245, 2019.
(b) [13] H. M. Sultan, O. N. Kuznetsov, and A. A. Z. Diab, "Site selection of
large-scale grid-connected solar PV system in Egypt," in 2018 IEEE
Fig. 7. One day of operation (24 hours) operation based on the results of Conference of Russian Young Researchers in Electrical and Electronic
Rao-2 algorithm. Engineering (EIConRus), 2018, pp. 813-818: IEEE.
[14] D. Manolakos, G. Papadakis, D. Papantonis, and S. Kyritsis, "A
VI. CONCLUSION simulation-optimisation programme for designing hybrid energy
systems for supplying electricity and fresh water through desalination
In this paper, the optimal arrangement of a hybrid to remote areas: case study: the Merssini village, Donoussa island,
renewable energy system including solar PV/wind/ HPS Aegean Sea, Greece," Energy, vol. 26, no. 7, pp. 679-704, 2001.
system for a practical case study in Ataka region, Suez, [15] A. A. Z. Diab, H. M. Sultan, and O. N. Kuznetsov, "Optimal sizing of
Egypt, based on several factors like LPSP, lower fluctuations hybrid solar/wind/hydroelectric pumped storage energy system in
level in energy interchanged with the national network as Egypt based on different meta-heuristic techniques," Environmental
Science and Pollution Research, pp. 1-23, 2019.
well as complete utilization of renewable resources has been [16] H. M. Sultan, A. A. Z. Diab, N. K. Oleg, and S. Z. Irina, "Design and
provided. In this study, a novel Rao optimization technique evaluation of PV-wind hybrid system with hydroelectric pumped
has been proposed to get the optimal configuration of the storage on the National Power System of Egypt," Global Energy
suggested hybrid system. Moreover, statistical measurements Interconnection, vol. 1, no. 3, pp. 301-311, 2018.
of the proposed Rao technique have been studied to validate [17] R. Rao, "Rao algorithms: Three metaphor-less simple algorithms for
solving optimization problems," International Journal of Industrial
the accuracy and stability of the proposed algorithm. The Engineering Computations, vol. 11, no. 1, pp. 107-130, 2020.
results obtained from the application of Rao have been
compared with other novel optimization techniques. The
obtained results proved the goodness of the proposed

Authorized licensed use limited to: Auckland University of Technology. Downloaded on May 25,2020 at 09:56:08 UTC from IEEE Xplore. Restrictions apply.