Renewable energy integration/techno-economic feasibility analysis, cost/benefit impact on islanded and grid-connected operations: A case study

Abstract

On-grid and off-grid renewable energy sources have emerged as a more efficient way to meet large-scale urban and rural needs. The integration of renewables into the utilities/island mode as well as into the techno-economic feasibility studies determines the financial viability, technical and operational viability of the project. This study was conducted to integrate renewables and techno-economic feasibility analysis using the utility grid-connected and islanded mode to meet the demand for community and commercial loads with these strategic objectives. Three main analytical tasks have been performed, including: optimization, simulation, and sensitivity analysis. We selected four sensitivity cases and further divided each sensitivity case into four sensitivity loops, including different combinations of renewable energy sources, e.g., photovoltaics, wind turbines, battery banks, power grids, converters, and diesel generators. The following parameters were used to conduct a techno-economic feasibility analysis: capital costs, replacement costs, operating and maintenance costs, salvage costs, fuel costs, real-time commercial/community load, and climate data. The best results of sensitivity case-2 were selected for the techno-economic feasibility analysis of the remaining four sensitivity cases. The islanded mode net present and annualized sensitivity loop-1 costs in sensitivity case-2 were noted at $1381080.42 and $106832.62, respectively. Net present and annualized sensitivity loop-1 costs in sensitivity case-2 of the grid-connected utility mode were noted at $512792.1 and $39,666.71, respectively. The proposed utility grid-connected mode cost is two times lower than that of the islanded mode, so we have chosen sensitive case-2 for hourly, seasonal, and annual electrical load flow and techno-economic feasibility analysis to meet the community and commercial load demand.

Full text

UNCORRECTED PROOF
Renewable Energy xxx (xxxx) 1–25
Contents lists available at ScienceDirect
Renewable Energy
journal homepage: www.elsevier.com/locate/renene
Renewable energy integration/techno-economic feasibility analysis,
cost/benefit impact on islanded and grid-connected operations: A case study
Tanveer Ahmad a,∗, Dongdong Zhang b
aEnergy and Electricity Research Center, International Energy College, Jinan University, Zhuhai, Guangdong Province, 519070, China
bSchool of Electrical Engineering, Guangxi University, Nanning, 530004, China
ARTICLE INFO
Article history:
Received 17 February 2021
Received in revised form 21 July 2021
Accepted 12 August 2021
Keywords:
Renewable energy integration
Techno-economic feasibility analysis
Sensitivity analysis
Photovoltaic and wind power generation
Islanded & utility grid-connected mode
Cost optimization
ABSTRACT
On-grid and off-grid renewable energy sources have emerged as a more ef^cient way to meet large-scale urban
and rural needs. The integration of renewables into the utilities/island mode as well as into the techno-economic
feasibility studies determines the ^nancial viability, technical and operational viability of the project. This study
was conducted to integrate renewables and techno-economic feasibility analysis using the utility grid-connected
and islanded mode to meet the demand for community and commercial loads with these strategic objectives.
Three main analytical tasks have been performed, including: optimization, simulation, and sensitivity analysis.
We selected four sensitivity cases and further divided each sensitivity case into four sensitivity loops, including
different combinations of renewable energy sources, e.g., photovoltaics, wind turbines, battery banks, power
grids, converters, and diesel generators. The following parameters were used to conduct a techno-economic feasi-
bility analysis: capital costs, replacement costs, operating and maintenance costs, salvage costs, fuel costs, real-
time commercial/community load, and climate data. The best results of sensitivity case-2 were selected for the
techno-economic feasibility analysis of the remaining four sensitivity cases. The islanded mode net present and
annualized sensitivity loop-1 costs in sensitivity case-2 were noted at $1381080.42 and $106832.62, respec-
tively. Net present and annualized sensitivity loop-1 costs in sensitivity case-2 of the grid-connected utility mode
were noted at $512792.1 and $39,666.71, respectively. The proposed utility grid-connected mode cost is two
times lower than that of the islanded mode, so we have chosen sensitive case-2 for hourly, seasonal, and annual
electrical load _ow and techno-economic feasibility analysis to meet the community and commercial load de-
mand.
© 2021
Nomenclature
AC Alternating current
CC Capital cost
CO2 Carbon dioxide
CO Carbon monoxide
DC Direct current
DG Distributed generation
COE Energy costs
FC Fuel costs
HOMER A hybrid optimization model for electric renew-
ables
ISDM Islanded mode
kW Kilowatt
∗Corresponding author.
E-mail address: tanveer.ahmad.pk11@gmail.com (T. Ahmad).
MG Microgrids
NREL National Renewable Energy Laboratory
NO2Nitrogen oxides
O&M Operating and maintenance
PV-1 Photovoltaic 1
PV-2 Photovoltaic 2
PV Photovoltaics
RES Renewable energy sources
RC Replacement cost
SGC Salvage cost
SO₂ Sulfur dioxide
WT-1 Wind turbine 1
WT-2 Wind turbine 2
Air density at standard pressure and temperature
( )
Anemometer height (m)
Anemometer height at different wind speed (m/s)
Derating factor of PV-1 and PV-2 (%)
https://doi.org/10.1016/j.renene.2021.08.041
0960-1481/© 2021

UNCORRECTED PROOF
2T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 1. Proposed scheme for integration and cost optimization of demand for commercial and community loads with two different scenarios of utility GCD and ISDM.
Table 1
Speci^cation for PV systems.
Description PV-1 PV-2 Unit
Type of panel Flat plate Flat plate Quantity
Power rated capacity 200 59.95 kW
Coef^cient of temperature −0.4 −0.4100 –
Operating PV temperature 25 26
Ef^ciency 15 17.20 %
Total CC 56000 16786 $
Total RC 20000 15000 $
The O&M cost 500 200 $/year
Derating factor 80 96 %
Different current time steps PV cell temperature
(°C)
Fuel mass _ow rate (kg/hr)
Generator output in this time step (kW)
Generator rated capacity (kW)
Intercept coef^cient of generator fuel curve (L/hr/
kWrated)
Lower heating values of the generator (MJ/kg)
Natural logarithm function
Power coef^cient of temperature (%/°C)
The power output of WT-1 and WT-2 (kW)
PV module ef^ciency under test conditions (%)
PV temperature of at standard test conditions
(20 °C–25 °C)
PV-1 and PV-2 capacity (kW)
PV-1 and PV-2 modules surface area
Rate of fuel consumption (L/hour)
The rated output of PV-1 and PV-2 modules (kW)
Real air density ( )
Represents the curve slope of the generator (L/hr/
kWoutput)
Represents the output power at standard pressure
and temperature (kW)
Solar radiation incident on PV modules )
Speed of wind at the hub height of WT-1 and WT-2
(m/s)
Standard radiation test condition
Standard test incident radiation )
Surface rough patch (m)
WT-1 and WT-2 hub height (m)
1. Introduction
The energy sector moves into microgrids (MG) and the age of dis-
tributed generation [1]. By 2040, total energy consumption is expected
to increase by approximately 30.1% over 2015 [2]. Almost 75% of the
world's electricity is generated using fossil fuels referred to as conven-
tional energy sources [3]. Globally, energy ef^ciency [4] and renew-
able sources have been recognized as a panacea and an ef^cient and op-
timal way to overcome fossil fuel consumption [5]. Wind turbines and
photovoltaic cells are the most frequently used energy sources that are
integrated into the main grid as distributed generators [6,7]. The mi-
crogrid can operate independently or in conjunction with the power
grid. The grid-connected mode of operation implies that the MG is con-
nected to the utility distribution networks and exchanges energy via
bidirectional metering [8]. ISDM refers to power plants/generators that
operate independently of local and national electricity distribution net-
works [9]. Researchers and energy planners are increasingly focusing
on renewable energy resources as a result of diminishing conventional
energy resources (e.g., fossil fuels) and, more specifically, environmen-
tal and economic challenges [10,11]. Additionally, as demand for elec-
tricity increases, the load on transmission networks is increasing unex-
pectedly. Due to the economic dif^culties associated with expanding
transmission networks, MG has been viewed as a ^nancially viable al-
ternative [12].
Improving energy ef^ciency/reliability becomes even more critical
in today's era when electricity demand is gradually increasing. The in-
stallation of MG in various distributed generation units throughout the
electricity grid aims to resolve a variety of issues with today's power
systems [13]. To implement distributed generation successfully, many
technological challenges must be overcome to ensure that existing reli-
ability standards are not jeopardized and that the potential bene^ts of
distributed generation are fully realized [14]. Before successfully im-

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 3
Fig. 2. Daily PV-1 and PV-2 output trends at varying periods of the day and on different days of the year.
Table 2
Wind turbine speci^cation parameters.
Description WT-1 WT-2 Unit
Rated capacity of each WT 25 25 kW
Rotor diameter 15.8 12.6 Meter
Cut-in wind speed 2.75 3.5 m/s
Cut-out speed 20 25 m/s
The height of the hub 23 30 Meter
Lifetime 25 25 Years
Total CC 9000 9000 $
Replace cost 4000 4000 $
O&M cost 200 200 $/year
plementing distributed generation, it is critical to understand the feasi-
bility of the site, cost optimization, cost comparisons, and the renew-
able potential of the site. This study proposed two modes for accurate
site identi^cation in MG, correct cost optimization of various loops in
various Sensitivity Cases (SC) in ISDM and utility grid-connected (GCD)
mode. The use of MG in islanded and utility GCD mode has a number of
advantages for the environment, performance, power quality, invest-
ment, marketing, and cost savings [11]. By improving power suppliers'
power quality and reliability, network congestion can be significantly
reduced, as can the need for bulk transmission systems. Our proposed
MG is capable of operating in both islanded and utility GCD modes.
Recent research studies: Several state-of-the-art studies have
been conducted on integrating and optimizing renewable energy costs.
For example, MG control trends [14], MG control strategies in ISDM
[15], distributed generation control [16,17], current and voltage con-
trol strategies for microgrids [18], improvement of microgrid perfor-
mance during ISDM with storage batteries [19], power-sharing ap-
proach to MG [20], and detection of islanded mode for distributed
generation has been studied in the past [21]. These studies have fo-
cused on MG control strategies, current and voltage regulation, in-
creasing storage capacity, and controlling distributed generation para-
meters. In our study, the MG's complex cost analysis aims to reduce
MG production costs and the cost of carbon emissions, while satisfying
disparities, and constraints on equity. The multiple route costs set out
in this study include operating, and maintenance (O&M) costs, fuel
costs (FCs), and depreciation costs resulting from wear and tear of dis-
tributed generation (DGs) sources after deployment.
Off-grid power generation using renewable energy technology has
evolved into a more ef^cient method of meeting urban and rural needs
on a limited budget without relying on traditional resources [22]. Re-
newable energy technology has the potential to impact the economic
development of any country. Additionally, as a result of rising fuel
prices and the problems caused by developing nations' energy crises, re-
newable resources have become critical. Solar and wind energy are ex-
amples of technologies and energy resources that are clean, environ-
mentally friendly, and micro-useable [23]. Sovacool and Enevoldsen in-
vestigated the challenges and opportunities associated with developing
a 100% self-su`icient renewable energy source (RES) on one of Myki-
nes's islands [24,25].
Several other studies on techno-economic design have been con-
ducted in recent years, such as electri^cation for irrigation and agricul-
ture [26], microgrid feasibility of different fuel options [27], Wind-PV
sustainability of hybrid systems [28,29], Wind-PV size in remote is-
landed mode [30], a feasibility study of MG in urban centers [31], off-
grid wind/PV/Hybrid fuel cell techno-economic analysis [32], wind
and photovoltaic optimization (PV) and techno-economic analysis
[33,34], rural electri^cation with stand-alone hybrid power systems
[35], techno-economic feasibility study of Wind-PV-Diesel-Hybrid bat-
tery systems for telecommunications applications [36,37], integration
of renewable energy into various off-grid rural electri^cation schemes
[38], autonomous desalinization schemes [39], city-integrated design
of renewable energy for climate-resilient and low-carbon communities
[40], size optimization and techno-economic analysis, and off-grid pho-
tovoltaic hybrid system [41], and so on. Most of these studies have been
optimized for the utility grid and the rest of the ISDM. In addition, dif-
ferent SCs were not considered in the utility grid and ISDM. First, we
used the combined island and utility GCD mode in this study. It was
then divided into four SCs and each case sensitivity was further divided
into four Sensitivity Loops for the integration of renewable energy and
a techno-economic feasibility analysis for community and commercial
loads. This type of study has not been carried out before, according to
our best knowledge.

UNCORRECTED PROOF
4T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 3. Fuel consumption of the proposed diesel generator on various days of the year.
Fig. 4. Commercial loads on weekdays and weekends.
We used the Hybrid Optimization Model for Electrical Renewables
(HOMER) tool to integrate GCD mode and techno-economic feasibility
analysis of renewable sources. HOMER is one of the growing energy op-
timization and planning software [23,42]. This involves three main an-
alytical types: optimization, simulation, and sensitivity analysis
[43,44]. The National Renewable Energy Laboratory (NREL) has been
developed. This commercial energy planning software is particularly
suited for designing and optimizing stand-alone GCD energy systems
and utilities [45].
Our integration and techno-economic feasibility studies were de-
signed to assess the project's ^nancial viability and technical feasibility.
For example, it covers the following aspects: (i) markets: includes ex-
pected potential project profits based on projected overall sales and
costs; (ii) fuel: calculates the feasibility of the fuel quantity and quality
of the project, project forecasts and project costs; (iii) facilities and po-
sition: assesses the infrastructure required; (iv) technical concept: MG
capacity, storage capacity, equipment size, diesel generator capacity,
automation & control; (v) environmental impacts; (vi) requirements
and costs; (vii) proposed islanded; and (viii) utility grid-connected prof-
itability modes. The advantage of this is that it helps researchers to
achieve technical feasibility, investment decisions, and ^nancial vari-
ability.
The main contribution to the study is described below:
⁃We propose an integration and techno-economic feasibility analysis
of RES and an analysis of its overall cost/benefit impact in ISDM and
utility GCD mode operation. We analyze the percentage of
renewable energy used in commercial and community loads. We
proposed a bi-directional metering system in the utility GCD mode
that operates in two ways: energy estimation, which determines how
much electricity we receive, and electricity given back to the grid.
⁃For islanded and utility GCD mode, we selected four SCs, and each
sensitivity case was further divided into four Sensitivity Loops
( ) which included RES, e.g., PV, wind turbines, batteries,
power grids, and diesel generators. Hourly, monthly, and seasonal
load _ow analyses are further conducted for the integration of
renewable energy.
⁃The design system is proposed for two di`erent types of loads,
including community loads and commercial loads. Data analysis,
data collection, projects, and reporting of proposed modes have been
conducted in a real-time location in Guangzhou, China.
⁃The proposed modes' cost optimization factors are calculated,
including capital cost (CC), replacement cost (RC), operation and
maintenance cost (O&M), salvage cost (SGC), ISDM FC, and utility
GCD electrical costs. Additionally, it compared island and utility
GCD modes and chose the most cost-effective solution for both
residential and commercial loads.
⁃Carbon dioxide (CO2), unburned hydrocarbons, carbon monoxide
(CO), sulfur dioxide (SO2), particulate matter, and nitrogen oxides
(NO2) emissions from islanded and utility GCDs are compared and
measured.
⁃We choose di`erent input parameters for integration and techno-
economic feasibility analysis, e.g., wind turbines (for example,
power curve and lifetime, cut-out and cut-in speed), PV systems
(e.g., derating factor, PV life, temperature effect, nominal cell
temperature operating range), diesel generator (e.g., price of fuel,
minimum to maximum load ratio), battery storage (e.g., loading and
charging status, nominal capacity and life), load demand (e.g.,
community and commercial load requirements), economics, (e.g.,

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 5
Fig. 5. Community loads on weekdays and weekends.
Fig. 6. Climate variables and their impact on the load at different setpoints.
project life of 25 years, net current costs and annualized costs),
technical requirements (e.g., hourly reserve operational conditions
and hourly reserve operational RES), resources (e.g., PV-1, PV-2,
WT-1, and WT-2), emissions (e.g., emissions of greenhouse gasses),
and reliability (e.g., permissible storage capacity).
The remaining parts of this study are divided into six main sections.
Section 2provides a brief overview of the islanded and utility GCD
mode. It covers an overview of renewable resources with explicit para-
meters. Section 3brie_y analyzes the four-sensitivity scenarios and four
in each scenario. In section 4, we conducted a thorough study of
commercial, community, and environmental variables. In section 5, the
^rst part of section 5explains the electrical load _ow analysis at differ-
ent time horizons, monthly, seasonal, and annual. The second part of
section 5describes a comprehensive cost optimization analysis of is-
landed and utility GCD modes. Section 6gives an overview of the re-
search discussions, the study challenges, and the real-time applications
of this study. Section 7concludes the study.
2. Proposed schemes
This section describes the proposed schemes in two parts: i) ISDM;
and (ii) GCD. The generator was used in ISDM with renewable sources,
and no power was taken out of the power grid. Energy is imported from
the utility grid in GCD mode. A detailed description of the two schemes
can be found in Fig. 1. The details of each part of the proposed method
are described below.
2.1. Proposed scheme for islanded mode
The operation of ISDM is based on two key perspectives:
1. The number of stand-alone generators that are not connected to the
utility grid.
2. A total number of generators connected to the parallel power grid
can be generated separately when a power failure occurs.
The trend of stand-alone generators that are not connected to the
power grid is followed in this section. These trends have the advantage
of minimizing the cost of building an external power supply site con-
nection. In the past, ISDM has been chosen where there is no power
grid. It is also used in rural areas, large construction sites, large com-
mercial centers, industries, oil ^elds, etc. In this study, the ISDM is used
as a backup generator with RES. When the load demand exceeds certain
RES limits, such as PV and Wind, the generator will be used as a backup
to meet the load de^cit demand. The selection of renewable sources is
described in more detail below.
2.1.1. Selection of PV systems
Selecting the optimal size and con^guration of a photovoltaic sys-
tem requires consideration of three major factors: economics, energy
loads, and esthetic or architectural considerations. There are various
types of photovoltaic systems that are cost-effective but require a large
installation area. In this case, the cost of PV installation and the amount
of space required must be balanced. To determine the building's opti-

UNCORRECTED PROOF
6T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 7. Wind speed and PV-1 and PV-2 power output.
mal photovoltaic system power level, it is necessary ^rst to assess the
building's electrical speci^cations. Architectural considerations shall
include the available surface area and dimensions for physical PV sys-
tem installation. We chose two distinct photovoltaic systems for this
study, as shown in Table 1. The PV-1 capacity is 200 kW, and the PV-2
capacity is 59.95 (~60 kW). The capacity of PV-1 and PV-2 is selected
based on the load demand. The ef^ciency of PV-1 and PV-2 was noted
at 15% and 5.30%, respectively. Both PV systems' CC, RC, and O&M
costs are determined using the most recent local market rates.
The ef^ciency of the photovoltaic module is determined by the
equation below:
(1)
represents the ef^ciency of the PV module under test condi-
tions (%), are the rated output of PV-2 and PV-1 modules
(kW), is the PV-2 and PV-1 modules surface area , and
described the standard radiation test condition .
The following equation is used to determine the rated power capac-
ity of the PV-1 and PV-2 benchmarks:
(2)
are the rated PV-1 and PV-2 capacity (kW), rep-
resents the derating factor of PV-1 and PV-2 (%), is the radiation in-
cident of solar on PV modules ), standard test incident
radiation ), is the energy coef^cient of temperature
(%/°C), is the different current steps of time of PV cell temperature
(°C), and designates the PV temperature at standard test condi-
tions (20°C 25 °C). Appendix-I described the mathematical formula-
tion of PV cell temperature measurement and temperature coef^cient
in our case, e.g., PV-1 and PV-2. The daily output of PV-1 and PV-2 is
shown in Fig. 2. The left vertical axis represents the hours of the day.
The vertical right axis shows the total PV output of PV-1 and PV-2, and
the horizontal axis shows the day of the year. Black parts indicate no
power generation at night. The generation of solar energy varies in the
daytime. For example, 11 a.m. Up to 5 p.m., the daytime power genera-
tion of PV-1 is greater than 170 kW/h. Similar trends in power genera-
tion can be seen in the output of PV-2, but its capacity is lower than PV-
1 due to its rating capacity.
2.1.2. Selection of wind turbines
Two distinct types of wind turbines have been chosen, and their
speci^cation parameters are listed in Table 2. Each turbine has the
same nominal capacity, but the difference is considered to be the cut-in
wind speed, the cut-out wind speed, the hub height, and the rotor diam-
eter. The increased height of the hub will result in increased wind en-
ergy production. For example, it maximizes bene^ts, reduces interfer-
ence from buildings and trees, and eliminates the need for additional
space clearance for longer blades. Wind output is affected by the differ-
ence in wind speed and rotor diameter of each turbine. For example,
the cut-in speeds of wind turbine 1 (WT-1) and wind turbine 2 (WT-2)
are 2.75 m/s and 3.5 m/s, respectively. The selected rotor diameters
for the WT-1 and WT-2 are 15.8 m and 12.6 m, respectively. The WT-1
and WT-2 cut-off speeds are set to 20 m/s and 25 m/s, respectively.
The following equation is used to determine the heights of WT-2 and
WT-1:
(3)
It's Equ. (3) is adopted following the Logarithmic Law. Where is
the speed of wind at the hub height of WT-1 and WT-2 (m/s), rep-
resents the anemometer height at different wind speeds (m/s), is
WT-1 and WT-2 hub height (m), is the anemometer height (m),
surface rough patch (m), and explicates the natural logarithm
function. The following equation takes the WT-1 and WT-2 output
power:
(4)
Where, explicates the power output of WT-1 and WT-2 (kW),
represents the output power at different standard pressure
and temperature (kW), is air density ( ), and described the

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 7
Fig. 8. Monthly power generation with grid connected and ISDM.
Fig. 9. Monthly analysis of renewable energy generation and load _ow.
density of air at standard pressure and temperature ( ). The
calculation of the diurnal pattern strength of WT-1 and WT-2 is shown
in Appendix I (Equ. 10).
2.1.3. Selection of converter and diesel generator
Any system that contains both alternating current (AC) and direct
current (DC) components require a converter. The capacity of the con-
verter is determined by the peak energy generated by wind and photo-
voltaic systems. Energy input and output through the converter are also
considered: operating hours, capacity factor, converter losses, and the
mean, minimum, and maximum output passing through the converter.
An 80-kW generator is used in conjunction with renewable energy
sources as a backup, and when renewable energy production falls short
of the total load requirement, the generator is used to meet the addi-
tional load demand. The consumption of the fuel generator is deter-
mined by the need for commercial and community load following the
total production of renewables, as illustrated in Fig. 3. Fuel consump-
tion is low daily, averaging between 6 and 12 L per hour. This is be-
cause renewable energy accounts for a sizable portion of the system's
load. The following equation is used to determine the generator's fuel
curve:

UNCORRECTED PROOF
8T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 10. Each month, the primary load is served in commercial and community
areas at varying times (loop 1 6 ms).
(5)
Where explains the rate of consumption of fuel (L/hour), intercept
coef^cient of fuel curve of the generator (L/hr/kWrated), represents
the curve slope of the generator (L/hr/kW output), is the generator
rated capacity (kW), and represents the generator output in this
time step (kW). The generator output curve is derived from the follow-
ing equation:
(6)
fuel mass _ow rate (kg/hr), and represents the lower heat-
ing values of the generator (MJ/kg). Factor 3.6 comes because
1 kWh = 3.6 MJ. The equation of is given in Appendix-I (Equ. 11).
2.2. Proposed scheme for grid-connected utility mode
In this scheme, we have met the additional load requirements of the
utility grid. The RC grid is zero due to the utility grid. The grid CC cor-
responds to the PV-1, PV-2, WT-1, and WT-2 interoperability rates. The
grid's O & M cost is equivalent to the annual cost of purchasing power
from the grid (demand cost + energy cost) by subtracting any profit
from the sale of energy to the grid. Energy charges are calculated in two
different ways: bi-directional metering and net metering. In bi-
directional metering, where the load demand exceeds the actual pro-
duction capacity of renewable energy, electricity is taken from the dis-
tribution grid; otherwise, additional renewable energy is supplied to
the grid. In net metering, only the power grid is used to meet the re-
quired load demand. We used the bi-directional metering method in
this proposed mode. The electricity rate was taken at a local market
rate of $0.085 per kWh. The additional selling price of renewable pro-
duction to the grid is $0.045 per kWh.
3. Sensitivity analysis and cost optimization of proposed schemes
Sensitivity analysis and mathematical formulation for cost opti-
mization are described in this section.
3.1. Selection of di>erent sensitivity levels for the penetration of renewable
energy
There are numerous reasons for determining the sensitivity of an in-
put parameter by assigning multiple parameters to it. By specifying a
set of values, we can determine the significance of a parameter and how
it changes in response to its significance. Sensitivity levels can be in_u-
enced by a variety of factors, including the price of diesel generator fuel
at various times of the day, wind speed, and solar radiation at various
times of the day. This study used only different wind speed sensitivity
levels to account for the fact that the wind blowing pattern varies over
time. Solar radiation and diesel fuel prices have no discernible effect on
the study area.
3.1.1. Sensitivity optimization case 1
In case 1 of sensitivity optimization, wind speed is taken as 5 m/s.
Each case selected is optimized by its architecture (e.g., PV-1, PV-2,
WT-1, WT-2, converter, diesel generator for ISDM, the grid for utility
GCD mode, batteries), cost (e.g., initial costs, operating costs, NPC and
COE), renewable fraction, renewable pentation, and O & M costs. Each
case of sensitivity optimization is further divided into four loops. Each
loop consists of different electrical components such as PV-1, PV-2, Bat-
tery-1, Battery-2, WT-1 and WT-2, a diesel generator, and a power grid
combination of a battery-1 and a battery-2 with a capacity of
1000 kWh, which is a combination of different batteries. Sensitivity op-
timization case 1 for ISDM with the following equation is proposed:
(7)
(8)
(9)
(10)
Where SC represents the sensitivity cases, represent the of
each SC, and are the photovoltaic systems 1 and 2, and
represents the wind turbines 1 and 2, explicates the diesel gen-
erator and different digits/numbers in Equ. (7–10) represent the se-
lected quantity of these components. The equations for utility GCD
modes are taken from the following equations:
(11)

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 9
Fig. 11. Seasonal electricity generation with grid connected and ISDM.

UNCORRECTED PROOF
10 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 12. Seasonal renewable energy generation and load _ow analysis.
(12)
(13)
(14)
3.1.2. Sensitivity optimization case 2
In sensitivity optimization case 2, wind speed is selected at 6 m/s.
The proposed islanded and utility GCD sensitivity optimization case 2
equations are taken as follows:
(15)
(16)
(17)
(18)
Fig. 13. Seasonal primary load served at varying hours throughout the month in commercial and community areas (loop 3, 6 ms).

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 11
Table 3
Total energy production, grid purchases, and renewable energy penetration as a percentage of total load demand.
Sr.
#
Sensitivity
cases
Sensitivity
loops
Energy resources Islanded mode Utility grid-connected
Total energy generation
(kWh/year)
% Ratio of total
generation
Total energy generation
(kWh/year)
(%) ratio of total
generation
2 SC-2 PV-1 334,293 33.1 334,293 30.8
PV-2 110,452 10.9 110,452 10.2
Diesel generator 276,439 27.3 – –
WT-2 118,767 11.7 118,767 11.0
WT-1 170,995 16.9 170,995 15.8
Grid purchases energy – – 349,828 32.3
Total 1,010,946 100 1,084,335 100
Total commercial &
community load
942,225 100 942,225 100
Excess electricity 13,384 1.32 257 257
Renewable fraction 70.6 % 67.1 %
PV-1 334,293 33.1 334,293 32.2
PV-2 110,452 10.9 110,452 10.7
Diesel generator 276,439 27.3 – –
WT-2 118,767 11.7 118,767 11.5
WT-1 170,995 16.9 170,995 16.5
Grid purchases energy – – 302,349 29.2
Total 1,010,946 100 1,036,856 100
Total commercial &
community load
942,225 100 942,225 100
Excess electricity 13,384 1.32 47.6 0.00460
Renewable fraction 70.6 % 69.0 %
PV-1 334,293 33.5 334,293 32.1
Diesel generator 432,549 43.4 – –
WT-2 59,384 5.95 118,767 11.4
WT-1 170,995 17.1 170,995 16.4
Grid purchases energy – – 416,714 40.0
Total 997,220 100 1,040,769 100
Total commercial &
community load
942,225 100 942,225 100
Excess electricity 278 0.0278 8.62 0.000800
Renewable fraction 54.0 % 59.1 %
PV-1 334,293 33.6 334,293 32.1
Diesel generator 515,301 51.8 – –
WT-2 59,384 5.97 118,767 11.4
WT-1 85,498 8.60 170,995 16.4
Grid purchases energy – – 416,714 40.0
Total 994,475 100 1,040,769 100
Total commercial &
community load
942,225 100 942,225 100
Excess electricity 0 0 8.62 0.000800
Renewable fraction 45.2 % 59.1 %
(19)
(20)
(21)
(22)
3.1.3. Sensitivity optimization case 3 and case 4
Wind speed was selected at 7 m/s and 8 m/s for sensitivity opti-
mization of case 3 and case 4, respectively. The GCD utility equations
are kept the same in case 3 and case 4. The ISDM cases 3 and 4 were
taken differently. The main reason for this is that the additional storage
capacity was chosen because additional renewable energy is not sup-
plied to the power grid. But this option cannot be used in ISDM, and the
following equation is used for optimizing ISDM sensitivity in case 3 and
case 4:
(23)
(24)

UNCORRECTED PROOF
12 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 14. Percent of renewable penetration at different times of the day.
(25)
(26)
(27)
(28)
(29)
(30)
As can be seen in the sensitivity mentioned above of the optimiza-
tion equations case 3 and case 4, the combination of WT-1 and WT-2 is
the same, but the battery bank has been changed. The reason for select-
ing more wind turbines is to indicates higher wind speeds at a glance.
3.2. Mathematical formulation for cost optimization
As stated in section 2.2, the bidirectional metering system is used for
estimating energy costs. To do this, we used the following equation:
Where is the utility grid prices for rates ($/kWh), rep-
resents the sellback rate to the utility grid ($/kWh), and
explicates the utility grid purchases minus utility
grid sell during the rate time such as applied. Equ. (31) represent the
monthly energy cost estimation, and the following equation takes the
estimate of the annual costs:
The present and annualized costs of the different components (e.g.,
PV-1, PV-2, WT-1, WT-2, battery bank, converter, and generator) are
measured by the following equation:
(33)
Where represents the yearly discount rate ($), represents the net
present cost ($), is the returning function of the capital recovery
factor, and represents the project's lifetime (year). Calculation of
demand charges, RC, SGC, O&M costs, and greenhouse gas emissions is
provided in Appendix II (Equ.12-17).

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 13
Fig. 15. Renewable energy penetration into the utility grid and system load.
4. Commercial and community load analysis
Two distinct types of loads were used in this study: commercial and
community loads. The location (Guangzhou, China) was chosen due to
the commercial and community loads and climate conditions. Daily av-
erage load demand, peak load demand, weekday and weekend load de-
mand are identical for the island and utility GCDs. Below is a detailed
description of the commercial and community loads.
Table 4
Optimization results for WT-1 and WT-2.
Sensitivity
cases
Sensitivity
loops
Description WT-1 WT-2 Units
Quantity
SC-2 Maximum output 58.4 49.3 kW
Wind penetration 18.1 12.6 %
Hours of operation 6983 6356 hours/year
Levelized cost 0.0105 0.0151 $/kWh
Total rated capacity of
WT-1 and WT-2
50 50 kW
Mean power output 19.5 13.6 kW
Capacity factor 39.0 27.1 %
Maximum output 58.4 49.3 kW
Wind penetration 18.1 12.6 %
Hours of operation 6983 6356 hours/year
Levelized cost 0.0105 0.0151 $/kWh
Total rated capacity of
WT-1 and WT-2
50 50 kW
Mean rated output 19.5 13.6 kW
Capacity factor 39.0 27.1 %
Maximum output 58.4 24.6 kW
Wind penetration 18.1 6.28 %
Hours of operation 6983 6356 hours/year
Levelized cost 0.0105 0.0151 $/kWh
Total rated capacity of
WT-1 and WT-2
50 25 kW
Mean rated output 19.5 6.78 kW
Capacity factor 39.0 27.1 %
Maximum output 29.2 24.6 kW
Wind penetration 9.04 6.28 %
Hours of operation 6983 6356 hours/year
Levelized cost 0.0105 0.0151 $/kWh
Total rated capacity of
WT-1 and WT-2
25 25 kW
Mean rated output 9.76 6.78 kW
Total capacity factor 39.0 27.1 %
Table 5
Optimization results of PV-1 and PV-2.
Sensitivity
cases
Sensitivity
loops
Description PV-1 PV-2 Units
Quantity
SC-2 Maximum output 194 60.0 kW
PV penetration 35.3 11.7 %
Operation hours 4404 4404 hours/year
Total levelized cost 0.0143 0.0136 $/kWh
Total rated capacity 200 60.0 kW
Mean power output 38.2 12.6 kW
Mean energy output
(kWh/d)
916 303 kWh/d
Capacity factor 19.1 21.0 %
4.1. Commercial load
Fig. 4 illustrates the trend in demand for commercial loads on
weekdays and weekends. During the day, there is a high demand for
loads. At night, the load demand is extremely low during the week-
days. However, demand for night loads is slightly higher on weekends
than on other weekdays. Between the hours of 8 a.m. and 4 p.m., en-
ergy demand is extremely high on weekdays and weekends. From 4
p.m. to 7 p.m., the energy consumption requirement gradually in-
creases. On the other hand, while the load requirement for the summer
season (e.g., April, May, June, July, and August) is higher, the demand
for the remainder of the year is low. The baseline average, annual aver-
age/day, and peak demand for commercial load are 101.1 kW,
2426.4 kW/day, and 405.71 kW.

UNCORRECTED PROOF
14 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 16. Purchased energy and sold additional renewable energy to the utility grid.
4.2. Community load
On weekdays and weekends, the demand for community loads is
distinct from the demand for commercial loads. Peak load demand is
depicted in Fig. 5, but it will vary after 5 p.m. Weekends have a slightly
lower total load demand than weekdays. For community load, the base-
line average, annual average/day, and peak demand are 6.9 kW,
165.59 kWh/day, and 23.31 kW, respectively. The time step size is 1 h
(60 min) per month and year for both community and commercial
loads. The year is further divided into four seasons: spring, summer,
winter, and autumn. In Guangzhou, China, the summer season is
longer; for example, the spring season (February–March), the summer
season (April–September), the autumn season (October–November),
and the winter season (December–January) all begin and end during
the summer season.
4.3. Climate variable
Climate variables are closely linked and directly impact the energy
requirements of buildings, and accurate forecasting of these variables is
a key task. Actual site readings of climate variables have been obtained
for this study. These variables include wind speed (WS), solar radiation
(SR), visibility and clearance index, low, average, and high ambient
temperature, solar zenith, solar azimuth, and angle of incidence used
for integration and cost optimization analysis, and some of the vari-
ables are shown in Fig. 6. The different ambient temperature points, for
illustrate, peak increasing load point 1 ( ), , , peak de-
creasing load point 1 , and are considered to cal-
culate the accurate load ramp at different hours of the day. The point
and represent the lower load demand at low point 1 and
point 2 in the winter season. The point explicates the peak load
demand in the summer duration.
Fig. 7 depicted wind speeds, photovoltaic (PV-1), and photovoltaic
(PV-2) power output at various times of the day. While the wind speed
is slightly higher during the day, the overall pattern for the year is
nearly identical, with solar energy generation beginning at 8 a.m. and
ending at 6 p.m. Between 10 a.m. and 3 p.m., PV-1 and PV-2 outputs
are significantly higher than the rest of the day. PV-1 and PV-2 have the
same hourly and daily solar generation trends, but their overall capac-
ity is different.
5. Optimization results and discussion
This section compares and analyzes the electrical load _ow analysis
in ISDM and GCD modes, the overall cost optimization analysis, the cost
comparison, and the environmental impact of greenhouse gas emissions
due to the selection of two different modes.
5.1. Electrical load =ow analysis in islanded and grid-connected mode
This section discussed the amount of electricity generated by renew-
able sources, the amount generated by diesel generators, and the
amount extracted from a GCD utility grid. Additional information about
these modes is provided below:
5.1.1. Analysis of the electrical load =ow on a monthly basis
As mentioned in section 3.1, we have chosen four SC and four
for integration and optimization analysis. The SC-2 results were se-
lected in detail in this study from all the selected cases. The main reason
for selecting only SC-2 results is that local site monitoring and weather
conditions are the most suitable solution we have discovered. However,
the other cases are compared in the cost optimization analysis. The
electrical load _ow analyzes of SC-1, SC-3 and SC-4 are attached and in-
cluded in the supplementary materials. The monthly generation of elec-
tricity in the GCD and ISDM utility SC-2 is shown in Fig. 8. In GCD

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 15
Table 6
Battery-1 and Battery-2 optimization results for SC-2.
Sensitivity
loops
Description Islanded mode Utility grid-
connected
Battery-1 Battery-2 Battery-1 Battery-
2
Units
Batteries –5.00 –1 Qty.
The wear
cost of
storage
–0.0158 –0.0158 $/kWh
Rated
capacity –5000 –1000 kWh
Useable
rated
capacity
–4000 –800 kWh
Throughput
lifetime –4,087,295 –12,671 kWh
Expected life –15.0 –15.0 Year
Mean value
of energy
cost
–0.109 –0 $/kWh
Energy-in to
battery –283,267 –47.2 kWh/year
Energy-out
from battery –258,503 –801 kWh/year
Storage
depletion –3756 –800 kWh/year
Total battery
losses –28,519 –45.8 kWh/year
Annual
throughput
_ow
–272,486 –845 kWh/year
Batteries 5.00 –1–Qty.
The wear
cost of
storage
0–0–$/kWh
Rated
capacity
5000 –1000 –kWh
Useable
rated
capacity
4000 –800 –kWh
Throughput
lifetime
4,087,295 –3,763,700 –kWh
Expected life 15.0 –15.0 –Year
Mean value
of energy
cost
0.109 –0.0592 –$/kWh
Energy-in to
battery
283,267 –263,759 –kWh/year
Energy-out
from battery
258,503 –238,037 –kWh/year
Storage
depletion
3756 –689 –kWh/year
Total battery
losses
28,519 –26,411 –kWh/year
Annual
throughput
_ow
272,486 –250,913 –kWh/year
Batteries –6.00 –1 Qty.
The wear
cost of
storage
–0.0158 –0.0158 $/kWh
Rated
capacity –6000 –1000 kWh
Useable
rated
capacity
–4800 –800 kWh
Throughput
lifetime –3,753,083 –12,671 kWh
Expected life –15.0 –15.0 Year
Mean value
of energy
cost
–0.154 –0 $/kWh
Table 6 (continued)
Sensitivity
loops
Description Islanded mode Utility grid-
connected
Battery-1 Battery-2 Battery-1 Battery-
2
Units
Energy-in to
battery –259,140 –47.2 kWh/year
Energy-out
from battery –237,366 –801 kWh/year
Storage
depletion –4364 –800 kWh/year
Total battery
losses –26,138 –45.8 kWh/year
Annual
throughput
_ow
–250,206 –845 kWh/year
Batteries 8.00 –1–Qty.
The wear
cost of
storage
0–0–$/kWh
Rated
capacity
8000 –1000 –kWh
Useable
rated
capacity
6400 –800 –kWh
Throughput
lifetime
3,518,803 –12,671 –kWh
Expected life 15.0 –15.0 –Year
Mean value
of energy
cost
0.177 –0–$/kWh
Energy-in to
battery
242,283 –47.2 –kWh/year
Energy-out
from battery
222,549 –801 –kWh/year
Storage
depletion
4737 –800 –kWh/year
Total battery
losses
24,471 –45.8 –kWh/year
Annual
throughput
_ow
234,587 –845 –kWh/year
mode, electricity generation from renewable sources, e.g., PV-1, PV-2,
WT-1, and WT-2 of is reached more than power generation
from the power grid. The electricity generation from renewable sources
in May, June, July, and August is higher than in other months. The out-
put of PV-1 in is almost equal to the remaining renewables
sources like PV-2, WT-1, and WT-2. The main reason for this is the dif-
ference in the nominal capacity of the selected renewables. In ,
imports of electricity from the grid are higher; the main reason for this
is the lack of PV-2 power generation. In June, July, and August, elec-
tricity imports from the electricity grid are higher due to electricity's
peak demand.
The trend in electricity generation in the ISDM is somewhat differ-
ent. This differentiation applies to electricity generated by the genera-
tor, not to renewable energy. The most important reason is that the
storage capacity is such as selecting a large number of batteries. In
ISDM, the electricity production in January, February, March, April,
November, and December of is very low. Renewables charge
the battery banks during the day. The batteries meet peak load require-
ments during the peak hours in the evening and night. The diesel gener-
ator is used more for electricity generation in the evening when power
generation from PV-1 and PV-2 stops. The peak demand for commercial
and community loads is always met. Electricity generation from diesel
is higher in and compared to the utility grid mode of
and . This is because there is no PV-2 in the generator
system, and the generator charges the battery bank.
Fig. 2 depicts the generation of electricity and the associated load
_ow at various times of the day. Commercial and community loads are
included in the total load. For example, on 01/08/2019, between 7

UNCORRECTED PROOF
16 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Fig. 17. The charging and discharging states of a battery on various dates throughout the year.
Fig. 18. The power curve of batteries charging and discharging at various times
of the day.
a.m. and 12 p.m., the aggregate demand for PV-1, PV-2, and battery
discharging is met (B-Discharging). Meanwhile, the generator gener-
ates no power. On the same day, the situation varies between 1:00 p.m.
and 6:00 p.m. Meanwhile, electricity generation has begun at WT-1
and WT-2. The PV-1, PV-2, WT-1, and WT-2 meet the total load de-
mand during this time and begin charging the battery (B-Charging). At
Table 7
Consumption of fuel.
Sensitivity loops Description Quantity Units
Total fuel consumption 96,226 Litter
Average fuel consumption per day 264 Litter/day
Average fuel consumption per hour 11.0 Litter/hour
Total fuel consumption 96,226 Litter
Average fuel consumption per day 264 Litter/day
Average fuel consumption per hour 11.0 Litter/hour
Total fuel consumption 145,846 Litter
Average fuel consumption per day 400 Litter/day
Average fuel consumption per hour 16.60 Litter/hour
Total fuel consumption 171,308 Litter
Average fuel consumption per day 469 Litter/day
Average fuel consumption per hour 19.60 Litter/hour
night, total load demand decreases, while commercial load demand in-
creases during daylight hours.
On the one hand, PV-1 and PV-2 generate no electricity, but WT-1
and WT-2 generate electricity at night. Charges to the battery increase
during the night hours. In Fig. 9, two days (01/08/2019 and 01/09/
2019) are depicted. In the meantime, the diesel generator produces no
electricity because electricity production from the proposed renewable
sources exceeds the load demand. Still, this trend is not consistent
throughout the year. The monthly primary load of, . is distrib-
uted to commercial and community areas every month. Fig. 10 illus-

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 17
Fig. 19. Fuel consumption trends for different months of the year.
trates the various hours of the year. During the day, load demand for
the proposed islanded and utility GCD modes are high. Additionally,
the different months' days are denoted by distinct colors.
5.1.2. Seasonal electrical load =ow analysis
Fig. 11 depicts a seasonal analysis of load _ow. With two distinct
loops, four distinct seasons (Winter, Spring, Summer, and Autumn) and
two proposed modes (Utility GCD and ISDM) are visualized. In the
monthly load _ow analysis.
electricity generated from all components, e.g., PV-1, PV-2, WT-1,
WT-2, generator, gird) was collected annually. We chose a single day to
illustrate the seasonal load _ow analysis results for each season. The
supplementary material contains all seasonal trends (Excel ^les). Wind
energy generation from WT-1 and WT-2 is extremely low during ISDM's
winter season. Additionally, the total load demand is low. However, the
generator generates additional electricity in the interim. This is because
the generator's excess power is being used to charge the battery. From 1
a.m. to 9 p.m., the generator generates 80 kW of electricity. Between 6
p.m. and 11 p.m. during the ISDM Spring season, renewable energy
sources meet 100% of load requirements. During the autumn season of
ISDM, diesel (Dsl) generator output is extremely low, and renewable
energy sources meet the entire load. During the day, PV-1, PV-2, WT-1,
WT-2, and B-Discharging are visible to meet the load demand. The bat-
tery charge was low on this day (10/30/2019) due to the previous day's
full charge.
A power grid replaces the diesel generator in GCD mode. In the GCD
utility mode, bidirectional metering is enabled. When total load de-
mand exceeds total renewable energy production, the grid purchase
mode is activated. The grid sales mode is maintained in the event of a
decrease in total load demand and increased production from renew-
able energy sources connected to the power grid. As previously stated,
the GCD utility mode utilizes a minimal number of battery banks.
As a result, surplus electricity is sold to the grid, except for the bat-
tery used for charging. Additional renewable energy is sold to the
power grid during the winter season between 11 a.m. and 4 p.m. Be-
tween 11 a.m. and 4 p.m., this amount is noted as 74.83506 kW,
152.81 kW, 86.78 kW, 162.70 kW, 48.23 kW, and 45.60 kW. The trend
in load _ow and renewable energy generation varies by season.
Fig. 12 depicts the seasonal load _ow analysis's entire stacked
area. All components, including energy generation, charging and dis-
charging, exhibit a consistent trend. We quantitatively depicted the
seasonal load _ow trend in Fig. 11 at various points in time. How-
ever, in Fig. 12, the layers of various renewable energy sources repre-
sent an increase and a decrease in demand. For example, at 9 a.m. on
25/06/2019, the PV-1, PV-2, WT-1, WT-2, and B-Discharging power
outputs are measured at 50%, which is equal to the community and
commercial load demand. This type of trend can be compared at any
hour during the speci^ed days. The bene^t of this plot is that each
component's layer contains accurate data on the load _ow in hours.
The demand for hourly loads is depicted in Fig. 13 for the winter,
spring, summer, and autumn seasons. The vertical axis depicts the daily
Fig. 20. Overall net present and annualized ISDM cost summary.

UNCORRECTED PROOF
18 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Table 8
Total net present and annualized cost summary of the grid-connected mode.
Component Capital
Cost
RC O&M Cost Fuel
Cost
SGC Total Cost
Net present cost
Battery-2 45000 19092.32 6463.76 0 −3593.37 66962.71
Converter 2359.88 752.35 1016.91 0 −423.99 3705.14
Grid 0 0 314745.5 0 0 314745.5
PV-1 56000 0 6463.76 0 −798.53 61665.23
PV-2 16786 0 2585.5 0 0 19371.5
WT-1 18000 0 5171.01 0 0 23171.01
WT-2 18000 0 5171.01 0 0 23171.01
Annualized cost
Battery-2 3480.95 1476.87 500 0 −277.96 5179.86
Converter 182.55 58.2 78.66 0 −32.8 286.61
Grid 0 0 24346.94 0 0 24346.94
PV-1 4331.85 0 500 0 −61.77 4770.08
PV-2 1298.47 0 200 0 0 1498.47
WT-1 1392.38 0 400 0 0 1792.38
WT-2 1392.38 0 400 0 0 1792.38
Net present cost
Battery-2 45000 19092.32 6463.76 0 −3593.37 66962.71
Converter 2588.25 825.15 1115.32 0 −465.03 4063.7
Grid 0 0 413267.53 0 0 413267.53
PV-1 56000 0 6463.76 0 −798.53 61665.23
WT-1 18000 0 5171.01 0 0 23171.01
WT-2 18000 0 5171.01 0 0 23171.01
Annualized cost
Battery-2 3480.95 1476.87 500 0 −277.96 5179.86
Converter 200.21 63.83 86.28 0 −35.97 314.34
Grid 0 0 31968.05 0 0 31968.05
PV-1 4331.85 0 500 0 −61.77 4770.08
WT-1 1392.38 0 400 0 0 1792.38
WT-2 1392.38 0 400 0 0 1792.38
hourly energy consumption, while the horizontal axis depicts the
monthly readings for each season.
The legendary bar on the right side of the vertical axis depicts the
various energy demand ranges for each hour. The dark red color de-
notes the peak demand for electricity. The trend in overall energy con-
sumption indicates that between 8 a.m. and 5 p.m., load demand is high
due to the high commercial load during the day. Each season has a dis-
tinct trend in terms of energy consumption. We record the highest en-
ergy consumption in the summer and the lowest in the winter.
5.1.3. of commercial and community loads powered by renewable energy
The renewable energy penetration of GCD and ISDM is shown in
Table 3 and Fig. 14. The total energy production of PV-1, PV-2, WT-1,
and WT-2 in ISDM of is 334,293 kWh/year, 110,452 kWh/
year, 118,767 kWh/year, and 170,995 kWh/year, respectively. This
represents the share of 33.1%, 10.95%, 11.7%, and 16.9% of the total
demand for renewable energy from PV-1, PV-2, WT-1, and WT-2. In this
, the power production from the diesel generator is
276,439 kWh/year, which has a 27.3% share of total energy produc-
tion. In utility GCD mode, the grid purchases energy for the load was
taken instead of diesel generator. The total commercial and community
load for the whole year is 942,225 kWh/year. While in utility GCD and
ISDM, the total load of all renewables, diesel generators, and grid
power purchases is 1,010,946 kWh/year and 1,084,335 kWh/year,
which is more than the total commercial and community load. In fact,
this difference is noted due to the charging of batteries and the loss of
power lines. The quantity of excess electricity remaining after charging
the battery banks is 13,384 kWh/year and 257 kWh/year, respectively,
for ISDM and the utility GCD.
Due to the minimum number of batteries, excess electricity in GCD
mode is low. In and , the total renewable energy pro-
duction of ISDM and the utility grid is the same; it is due to the similar
amount of PV-1, PV-2, WT-1, and WT-2. But the difference can be seen
in and . The basic reason for this is the difference in the
total number of wind turbines and batteries. In these SCs, there is a high
level of energy generation from diesel generators and grid purchases.
The absence of PV-2 is another major reason. The excess energy in
and is noted as lower than in and be-
cause of limited renewables sources. The renewable fraction is the frac-
tion of the electricity generated by the RES and supplied to the load.
The renewable fraction of islanded and utility grid modes of ,
, and are recorded at 70.6%, 67.1%, 70.6%,
69.0%, 54.0%, 59.1%, 45.2% and 59.1%, respectively. A clearer view
of renewable energy penetration of islanded and utility GCD mode of
randomly chosen of different can be seen in Fig. 14. Of course,
the daytime penetration of renewable energy is higher due to the exis-
tence of PV-1 and PV-2. Renewable energy penetration of the system
load has less variation during the day. There is quite a high variation at
night due to a change in air blowing.
The percentage of renewable penetration at different times of the
day and year is shown in Fig. 15. Renewable penetration is more than
85% on most days of the year. The red color indicates a high degree of
renewable penetration. There is an unusual trend in the penetration of
renewable energy from 7 a.m. to 6 p.m. every day that represents PV-1
and PV-2. The ISDM, , utility GCD mode at a wind
speed of 8 m/s produces a large amount of wind power generation due
to the overall renewable penetration. Another reason for this is that
more electricity is being generated from the wind due to high wind
speeds. On different days of the year, the trend is different. Some days
of renewable penetration are high, and some days of renewable pene-
tration are relatively low.
The optimization results for WT-1, WT-2, PV-1, and PV-2 are shown
in Tables 3 and 4. The net capacity factor is the dimensionless ratio of
the actual energy generated over a given period to the average electric-
ity generation over that period. The levelized cost is the average cost
per kWh of useable electrical power of the system. The calculated wind
penetration, maximum output, operation hours, levelized cost, total
rated capacity, mean output, and a capacity factor of WT-1 and WT-2 in
are 18.1%, 58.4 kW, 6983 h/year, 0.0105$/kWh, 50.0 kW,
19.5 kW and 39.0% which is equal to . The reason for this is
that both loops have the same number of wind turbines. Calculated
maximum output, wind penetration, operating time, levelized cost, to-
tal rated capacity, mean output, and WT-2 capacity factor in
are measured 24.6 kW, 6.28%, 6356 h/year, 0.0151$/kWh, 25 kW,
6.78 kW and 27.1%, respectively.
Like in Table 4, the comparison in Table 5 shows the optimization
results for PV-1 and PV-2. The readings of PV-1 and PV-2 of maximum
output, PV penetration, operation hours, levelized cost, total rated ca-
pacity, mean output (kWh/d) and capacity factor are 194 kW, 35.3%,
4404 h/year, 0.0143 $/kWh, 200 kW, 38.2 kW, 916 kWh/d, 19.1%
and 60.0 kW, 11.7%, 4404 h/year, 0.0136$/kWh, 60.0 kW, 12.6 kW,
303 kWh/d and 21.0% noted of in SC-2. The optimization re-
sults of PV-1 in , and are the same because of
the similar nominal rated capacity of PV-1. The operating hours of both
PV-1 and PV-2 are the same, but the penetration rate of PV-1 is more
than doubled compared to PV-2.
5.1.4. Energy purchased and sold to the utility gird
Battery consumption is minimal in GCD utility mode, and excess
power is exported to the utility grid, as illustrated in Fig. 16. Sold en-
ergy is renewable energy that is supplied to the utility grid in addition
to the energy that is generated on-site. The difference between the en-
ergy purchased from the grid and the energy sold to the grid is the net
energy purchased. In Fig. 16, we considered various wind speed levels,
including 5 m/s, 6 m/s, and 7 m/s, intending to determine the addi-
tional power capacity at various times and wind speed levels.
For example, the gird energy purchased between January
and December 2019 at a wind speed of 6 m/s is 26,054 kWh,

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 19
Table 9
Comparison of total net present value and annualized ISDM costs.
Sensitivity cases Sensitivity loops Duration Capital Cost ($) RC ($) O&M Cost ($) Fuel Cost ($) SGC ($) Total Cost ($)
SC-1 NPC 333412.31 100055.68 57746.49 1322397.47 −19943.18 1793668.78
Annual cost 25790.9 7739.75 4466.94 102293.23 −1542.69 138748.13
NPC 477543.63 156609.5 80336.64 1186571.94 −29975.8 1871085.92
Annual cost 36940.09 12114.43 6214.39 91786.53 −2318.76 144736.69
NPC 461522.37 157802.81 77997.04 1463796.38 −30747.28 2130371.32
Annual cost 35700.78 12206.74 6033.41 113231.06 −2378.44 164793.55
NPC 542073.36 196147.76 89879.63 1588383.56 −37487.26 2378997.05
Annual cost 41931.75 15172.89 6952.58 122868.42 −2899.8 184025.84
SC-2 NPC 342942.45 98859.55 58547.62 900633.35 −19902.55 1381080.42
Annual cost 26528.1 7647.22 4528.91 69667.93 −1539.55 106832.62
NPC 342942.45 98859.55 58989.34 900633.35 −19842.66 1381582.02
Annual cost 26528.1 7647.22 4563.08 69667.93 −1534.92 106871.42
NPC 362031.1 119327.63 61925.86 1365048.58 −23572.59 1884760.58
Annual cost 28004.69 9230.51 4790.24 105592.48 −1823.44 145794.48
NPC 443006.92 157862.31 73928.6 1603365.59 −30379.92 2247783.51
Annual cost 34268.52 12211.34 5718.7 124027.35 −2350.02 173875.89
SC-3 NPC 252681.94 59528.33 44383.15 703956.72 −12275.99 1048274.16
Annual cost 19546.05 4604.78 3433.23 54454.13 −949.6 81088.6
NPC 252681.94 59528.33 44648.18 703956.72 −12240.05 1048575.12
Annual cost 19546.05 4604.78 3453.73 54454.13 −946.82 81111.88
NPC 416435.94 137314.35 69509.08 968717.72 −26995.02 1564982.06
Annual cost 32213.14 10621.87 5376.83 74934.56 −2088.18 121058.21
NPC 416435.94 137314.35 70127.48 968717.72 −26911.17 1565684.31
Annual cost 32213.14 10621.87 5424.67 74934.56 −2081.7 121112.54
SC-4 NPC 344968.69 97890.94 56568.96 452536.57 −20373.66 931591.51
Annual cost 26684.84 7572.29 4375.86 35005.69 −1575.99 72062.68
NPC 344968.69 97890.94 57010.67 452536.57 −20313.77 932093.11
Annual cost 26684.84 7572.29 4410.03 35005.69 −1571.36 72101.48
NPC 462719.75 155580.66 74939.84 730863.03 −30559.65 1393543.63
Annual cost 35793.4 12034.85 5796.92 56535.46 −2363.92 107796.7
NPC 462719.75 155580.66 75646.59 730863.03 −30463.83 1394346.21
Annual cost 35793.4 12034.85 5851.6 56535.46 −2356.51 107858.78
19,713 kWh, 22,392 kWh, 23,411 kWh, 31,406 kWh, 36,030 kWh,
40,868 kWh, 43,532 kWh, 33,410 kWh, 30,181 kWh, 24,044 kWh,
18,785 kWh, and 349,8 From January to December, 13,047 kWh,
11,782 kWh, 14,150 kWh, 12,120 kWh, 8345 kWh, 4908 kWh,
3294 kWh, 3402 kWh, 6570 kWh, 8458 kWh, 13,363 kWh,
16,619 kWh, and 116,058 kWh of energy were sold in this mode. Be-
tween January and December, the net energy value purchased was
13,006 kWh, 7931 kWh, 8242 kWh, 11,292 kWh, 23,061 kWh,
31,123 kWh, 37,574 kWh, 40,130 kWh, 26,841 kWh, 21,723 kWh,
10,681 kWh, 2166 kWh, and 233,769 kWh. Between January and De-
cember, the estimated energy charges are $1613.90, $1134.68,
$1254.14, $1432.31, $2279.65, $2826.56, $3308.92, $3529.51,
$2529.63, $2170.91, $1429.55, $837.18, and $24,346.94. The rate at
which energy is purchased and sold back to the utility grid has already
been stated. Between May and September, grid-purchased energy de-
mand is higher, owing to the summer season. From November to
March, the energy sold back to the power grid can be seen. As wind
speeds increase, the _ow of energy (energy sold) through the power
grid increases as more wind generates more electricity.
5.1.5. Battery charge/discharge status
The detailed description of Battery-1 and Battery-2 in ISDM and
utility GCD modes is shown in Table 6.“The cost of battery storage is
the cost of cycling energy through the storage bank.”The number of
batteries in each loop varies. The quantity of batteries in
ISDM is more than in the GCD utility mode. In utility GCD mode, the
battery's use is lower, which is shown in the form of energy in, energy
out, and total battery losses. Table 6 represents the yearly battery en-
ergy charging capacity, which is 272,486 kWh/year, 845 kWh/year of
in ISDM and GCD, respectively. In ISDM, the average cost of
the battery is higher. The term “storage depletion charges”refers to the
difference between the battery's year-end and year-end charging status.
Increased battery life accelerates the depletion of stored energy. Battery
life is estimated to be 15 years and is dependent on the type of battery
and its characteristics. Each battery has a nominal rated capacity of
1000 kWh.
Figs. 17 and 18 depict a graphical representation of the battery
charging/discharging power curve and the battery charging/discharg-
ing state on various days of the year and at various hours of the day.
Battery charging and discharge states vary at different times of the
day. For example, the battery charge is low and is about 20% between
200 and 240 h of in the ISDM. The yellow color indicates a
charge rate of between 70% and 80%. This style varies in GCD mode
due to the battery's low dependence on charging. The charging and dis-
charge curves for Battery-2 are depicted in Fig. 18. The discharging
curve steepens during the day as Battery-2 supplies energy to the load,
but Battery-2 is also recharged using renewable energy. When the
charging curve of the battery remains zero, the discharge curve reaches
its maximum. The difference in two battery discharge curves is used to
investigate various, rates.
5.1.6. Fuel consumption in islanded mode
The total fuel consumption of the generator is shown in Table 7 and
Fig. 19. The fuel consumption of and is the same but
differ in the case of and . Due to the low penetration of
renewable energy in and , total fuel consumption de-
mand increases. The pattern of fuel consumption can be seen in the dif-
ferent months of the year in Fig. 19. White space indicates low con-
sumption of fuel. Dark blue lines show commercial and community
load demand variability due to dispersion in generator ON and OFF
hours. Summer fuel consumption is higher than during reset months.

UNCORRECTED PROOF
20 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Table 10
Comparing the utility grid-connected mode's total net present and annualized costs.
Sensitivity cases Sensitivity loops Duration Capital Cost ($) Replacement Cost ($) O&M Cost ($) SGC ($) Total Cost ($)
SC-1 NPC 156222 19868.94 424239.09 −4829.57 595500.46
Annual cost 12084.46 1536.95 32816.75 −373.59 46064.57
NPC 156222 19868.94 424327.43 −4817.59 595600.78
Annual cost 12084.46 1536.95 32823.58 −372.66 46072.33
NPC 139436 19868.94 524405.89 −4829.57 678881.26
Annual cost 10785.99 1536.95 40565.09 −373.59 52514.44
NPC 139436 19868.94 524494.24 −4817.59 678981.59
Annual cost 10785.99 1536.95 40571.93 −372.66 52522.2
SC-2 NPC 156145.88 19844.67 341617.45 −4815.89 512792.1
Annual cost 12078.57 1535.07 26425.61 −372.53 39666.71
NPC 156222 19868.94 341150.63 −4817.59 512423.97
Annual cost 12084.46 1536.95 26389.49 −372.66 39638.24
NPC 139588.25 19917.48 437652.38 −4856.92 592301.19
Annual cost 10797.76 1540.7 33854.33 −375.7 45817.09
NPC 139588.25 19917.48 437652.38 −4856.92 592301.19
Annual cost 10797.76 1540.7 33854.33 −375.7 45817.09
SC-3 NPC 156222.00 19868.94 268391.74 −4829.57 439653.11
Annual cost 12084.46 1536.95 20761.28 −373.59 34009.09
NPC 156374.25 19917.48 260682.85 −4844.94 432129.64
Annual cost 12096.23 1540.7 20164.96 −374.78 33427.12
NPC 139740.5 19966.01 361095.3 −4884.28 515917.54
Annual cost 10809.54 1544.46 27932.3 −377.82 39908.48
NPC 139740.5 19966.01 361183.65 −4872.3 516017.86
Annual cost 10809.54 1544.46 27939.14 −376.89 39916.24
SC-4 NPC 156298.13 19893.21 208906.35 −4843.24 380254.44
Annual cost 12090.34 1538.83 16159.82 −374.65 29414.35
NPC 156222 19868.94 198871.7 −4817.59 370145.05
Annual cost 12084.46 1536.95 15383.6 −372.66 28632.34
NPC 139740.5 19966.01 299133.41 −4884.28 453955.65
Annual cost 10809.54 1544.46 23139.28 −377.82 35115.46
NPC 139436 19868.94 295111.52 −4817.59 449598.86
Annual cost 10785.99 1536.95 22828.17 −372.66 34778.44
5.2. Analysis of cost optimization
This section discusses the cost optimization of the utility grid-
connected mode, ISDM, cost comparisons, and the environmental im-
pact of greenhouse gas emissions.
5.2.1. Optimization of the proposed cost of the system in islanded mode
The optimization of the proposed cost of the ISDM system can be
seen in Fig. 20. It shows the optimization results of ,
and in SC-2. The cost of capital and the cost of replacement are
determined using market rates in the local area. The component's life-
cycle cost, or net present value, is the present value of all costs associ-
ated with the component's installation and operation (e.g., PV-1, PV-2,
WT-1, WT-2, batteries, diesel generator) over the project's lifetime, less
the present value of all revenues earned over the project's lifetime. The
component's annualized cost is the cost if it occurs on an equal basis
during each project life year. Estimated net present total costs of bat-
tery-2, converter, diesel generator, PV-1, PV-2, WT-1, and WT-2 in
is noted 334813.56, $ 4955.8, $ 913932.31, $ 61665.23, $
19371.5, $ 23171.01, and $ 23171.01, respectively. The estimated an-
nualized total cost of Battery-2, converter, diesel generator, PV-1, PV-2,
WT-1 and WT-2 in is noted as 25899.29, $ 383.35, $ 70696.66,
$ 4770.08, $ 1498.47, $ 1792.38 and $ 1792.38, respectively. If this
cost is compared to and then cost is observed
to be lower. The main reason for this difference is the lack of renewable
sources in and and the more dependency of high-
power generation comes from the diesel generator. For example, the
annualized total production cost of electricity from the diesel generator
in , and at SC-2 is calculated at $ 69667.93, $
105592.48, and $124027.35, respectively. The net percentage cost and
the annualized cost of both and are the same, mainly
due to the similarity of batteries and wind turbines. The SGC of batter-
ies is higher in and . The cost of O&M is increased by in-
creasing the component CC.
5.2.2. Optimization of proposed system costs in grid-connected mode
The optimization cost of the proposed system in GCD mode is shown
in Table 8. The FC in utility GCD mode is zero because there is no diesel
generator available. The PV-1, PV-2, WT-1, and WT-2 CCs values are
noted the same as the ISDM. But there is a difference between the cost
of O&M, RC, and SGC. In this mode, the cost of , ,
, and is the same because the number of power compo-
nents are the same. The numbers of power components are the same in
calculating the overall cost estimation in and , and
and of SC-2, but the sequence is different from each
component in each loop, such as already discussed in section 3.1. Un-
availability of PV-2 in and of SC-2 does not have much
effect on the total cost if compared with and . But in the
ISDM, there was a different trend noted. The main reason for this is the
low CC of batteries/quantities and the import of cheap electricity from
the electricity grid.
5.2.3. Cost comparison between islanded and utility grid-connected mode
Table 9 and Table 10 show the overall cost comparison between
the islanded and GCD. We discussed each sensitivity case (SC-1, SC-2,
SC-3, and SC-4) and four sensitivity loops for each sensitivity case.
Each loop shows the total cost of CC, RC, O&M, FC, SGC, and each
component's total cost ($). There is no FC portion in the GCD utility
mode. The total NPC and annual cost including CC, RC, O&M cost, FC,
SGC, and total cost in , , and of SC-2 is
measured at $1793668.78, $138748.13, $1871085.92, $144736.69,
$2130371.32, $164793.55, $2378997.05, and $184025.84, respec-
tively. But the price in SC-2 is lower. SC-3 and SC-4 have lower costs
than SC-2. The reason for this is that the SC-2 has the highest FC and the

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 21
Table 11
Comparison of COE and NPC to previously published studies.
HOMER planning Ref. Area Region COE
($/kWh)
NPC
(Million
$)
WT/PV/Battery/Converter [40] Busan South
Korea
0.399 26.09
DG/Battery/PV/WT/Converter [26] Dongola Sudan 0.387 24.16
DG/Battery/PV/Converter [46] Rafha KSA 0.170 28.5
Battery/Converter/PV/DG/Natur
al gas Generator/Biomass
[47] Vancouver Canada 0.285 29.3
Our study (Utility grid-connected mode)
PV-1/PV-2/WT-1/WT-2/Utility
Grid/Battery-2 for (SC-2,
)
–Guangzhou China 0.0373 0.5128
PV-1/PV-2/WT-1/WT-2/Utility
Grid/Battery-1 for (SC-2,
)
–Guangzhou China 0.0407 0.5124
PV-1/WT-1/WT-2/Utility
Grid/Battery-2 for (SC-2,
)
–Guangzhou China 0.0450 0.5923
PV-1/WT-1/WT-2/Utility
Grid/Battery-1 for (SC-2,
)
–Guangzhou China 0.0450 0.5924
Our study (Islanded mode)
PV-1/PV-2/WT-1/WT-2/Utility
Grid/Battery-2 for (SC-2,
)
–Guangzhou China 0.114 1.38
PV-1/PV-2/WT-1/WT-2/Utility
Grid/Battery-1 for (SC-2,
)
–Guangzhou China 0.114 1.38
PV-1/WT-1/WT-2/Utility
Grid/Battery-2 for (SC-2,
)
–Guangzhou China 0.155 1.88
PV-1/WT-1/WT-2/Utility
Grid/Battery-1 for (SC-2,
)
–Guangzhou China 0.185 2.25
Fig. 21. CO2 emissions from ISDM's sensitivity case-1.
lowest battery storage capacity. The number of batteries in the SC-4 is
eight, nearly double that of the SC-1. The measured annual FC of
, , and of SC-1 is calculated
$102293.23, $91786.53, $113231.06, and $122868.42, respectively.
This is the lowest cost in . The optimal renewable energy com-
bination includes PV-1, PV-2, WT-1, WT-2, and eight batteries. As SC
grows, the amount of renewable energy produced will grow as well.
Fig. 22. Carbon dioxide and nitrogen oxide emissions from grid connected and
ISDM systems.
This is primarily due to the fact that as wind speed increases (see Table
11).
Compared to ISDM, the overall net present and annualized costs of
the GCD utility mode are low. In FC, CC, and RC, minor adjustments
may be made. Reduced total battery count in GCD mode is a significant
factor in achieving low CC and total costs. In the utility GCD mode of all
SGC and all , only one battery is taken. Reducing the CC of the
batteries reduces the overall cost. Additional batteries are required in
ISDM because additional storage capacity is required to store additional
renewable energy. However, additional renewable energy can be sold
to the utility grid in the GCD utility mode via bi-directional metering. In
comparison to ISDM, the replacement costs in GCD utility mode are
low. However, O&M costs remain higher in utility GCD mode. This is
primarily due to the grid's capital, which is zero, and the fact that RC is
zero. Our analysis excludes Grid CC and RC, but we calculated electric-
ity imports using the most recent local utility rates. Costs associated
with O&M and SGC are offset by the community and commercially im-
ported load demand from the utility grid.
5.2.4. Emissions of greenhouse gases in islanded and grid-connected modes
CO2 emissions are negligible in the GCD utility mode. CO2, CO, Un-
burned Hydrocarbons, Particulate Matter, SO2, and NO2 in utility GCD
and ISDM of SC-2 are depicted in Figs. 21 and 22. The amount of CO2,
CO, Unburned Hydrocarbons, Particulate Matter, SO₂, and NO2in
of ISDM are calculated at 253,396 kg/year, 625 kg/year,
69.3 kg/year, 47.2 kg/year, 509 kg/year and 5581 kg/year, respec-
tively. The values of CO2, CO, Unburned Hydrocarbons, Particulate
Matter, SO₂, and NO2in of ISDM are measured at 451,112 kg/
year, 1114 kg/year, 123 kg/year, 83.9 kg/year, 906 kg/year and
9936 kg/year, respectively. The values of of greenhouse gas
emissions are approximately doubled to .
This is due to a low level of renewable energy production and a high
level of fossil fuel consumption. The use of diesel generators to generate
electricity results in an increase in greenhouse gas emissions. The dis-
persion of greenhouse gas emissions from each utility grid loop is very
small. As a result, the GCD utility mode is more advantageous for re-
newables and community and commercial loads.
5.2.5. Validation of study results through comparison to previous studies
We compared our ^ndings to those of previous studies. The energy
costs (COE) and net present value (NPV) of current studies are higher
than those of our proposed utility GCD mode and ISDM. However, it is
worth noting that while the methodology and application of existing
studies vary, the overall goal is to reduce COE and NPC. COE and NPC
are significantly less expensive in our study, demonstrating the solu-

UNCORRECTED PROOF
22 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
tion's perfect integration and economic viability for community and
commercial loads in the GCD and ISDM utility modes .
6. Discussion, challenges, and real-time applications
This study aimed to create a hybrid of ISDM and GCD modes, ana-
lyze it, and determine which combination is suitable for execution. In
this study, two types of loads, including commercial and community
loads, and four SC, were selected. Further, divide each SC into four
. The objective was to determine the optimal mix of renewable
energy sources for commercial and community loads and the optimal
cost-optimization mix. The best SC-2 was chosen from all SCs. Further,
and results are adopted in SC-2. Renewables in combi-
nation with utility GCD mode is more appropriate than ISDM. The cost
of utility GCD mode is approximately two-thirds that of ISDM. For ex-
ample, the net present value and the annualized cost of, SC-loop1. in
ISDM at SC-2 are $1381080.42 and $106832.62, respectively, and
these costs are $512792.1 and $39666.71 for utility GCD mode. The
cost savings are noticeable in this scenario, and there are also fewer
greenhouse emissions, which is critical for maintaining a clean environ-
ment. The overall cost is even lower in SC-3 and SC-4, both in islanded
and utility GCD modes, but that combination was not chosen. The rea-
son this combination was not determined is that the weather conditions
are inadequate. For example, wind speeds of 7 and 8 m/s are not avail-
able for the entire year at the proposed site at a glance. Readers can
choose between SC-3 and SC-4, which have higher wind speeds and
more sunlight for year-round wind and solar energy generation. Addi-
tionally, the study faces some challenges. The primary obstacle is secur-
ing suf^cient space for wind turbines and solar photovoltaic installa-
tions. Maintaining an adequate supply of fuel in ISDM is also critical. In
GCD utility mode, bidirectional metering is also a significant challenge.
Maintaining voltage, power, and frequency stability is also a dif^cult
task. The application of this study in real-time is extensive. On week-
days and weekends, we detailed the overall community and commer-
cial load demand regarding structural factors. Further, the optimization
cost is compared between different SCs with different at both is-
landed and utility GCD mode. Real-time component rates and weather
conditions are chosen according to the location of the local modes. As a
design package, researchers can use it and prepare their estimates, con-
sidering any climatic conditions.
7. Conclusion
This study presents integration and techno-economic feasibility
analysis of renewable sources and their overall cost impact in the con-
text of islanded and GCD operations. In this study, the real-time com-
munity and commercial loads and climatic parameters were taken from
Guangzhou, China. Four different SC, for example, SC-1, SC-2, SC-3,
and SC-4, and four loops of each SC case (e.g., , ,
and ) are taken. Daily average load demand, peak de-
mand, weekday load demand, and weekend load demand are the same
for islanded and utility GCD mode analysis. Hourly, seasonal, and
yearly load _ow analysis is performed on different proposed SC and
. The total electricity generation of PV-1, PV-2, WT-1, and WT-2
in ISDM of is noted at 334,293 kWh/year, 110,452 kWh/year,
118,767 kWh/year and 170,995 kWh/year, respectively. It represents
33.1%, 10.95%, 11.7%, and 16.9% of the total demand for renewable
energy from PV-1, PV-2, WT-1, and WT-2. The bidirectional metering
method is used for estimating energy costs in utility GCD mode. In this
scenario, additional energy is imported from the power grid to meet the
required load demand. The surplus energy generated by renewables is
sold to the utility grid. The storage of batteries is vital and plays a key
role in ISDM, and greatly impacts the overall cost. For example, the net
present cost of batteries in the ISDM and utility GCD mode of SC-2
( ) is calculated at $536504.28 and $ 66962.71, respectively.
Hot weather conditions increase the demand for loads, which in turn in-
creases fuel consumption, but the higher the penetration of renewables,
the need for fuel consumption can be overcome. We found the best cost
optimization solution for SC-2 using different parameters, including
CC, RC, O&M, SGC, and FC. The impact of greenhouse gases is lowest in
the GCD mode. For example, the amount of CO2, CO, Unburned Hydro-
carbons, Particulate Matter, SO₂, and NO2in of ISDM are noted
as 253,396 kg/year, 625 kg/year, 69.3 kg/year, 47.2 kg/year, 509 kg/
year and 5581 kg/year, respectively. These greenhouse gas values are
low enough to show more appropriate greenhouse emission solutions in
grid-connected utility mode. This integration and techno-economic fea-
sibility study is a more reliable and effective way of carrying out the
feasibility study at a glance. By promoting urban and rural areas
through these systems, provincial and local governments will play an
important role in addressing energy shortages and ^nding possible so-
lutions. These systems help reduce the power grid's overall peak load
and result in cost savings for end-users. Suppose local governments or
regional governments start subsidizing end-users for the deployment of
renewable energy resources. In that case, the grid load will decrease
^rst, and the end-users will generate more of their electricity demand.
This would also reduce the end user's total energy costs and sell the sur-
plus energy to the utility grid. The proposed modes in rural and urban
areas can be used as a design package, making them autonomous with
grids and without grids. In the future, we will propose renewable en-
ergy integration/techno-economic feasibility analysis, cost/bene^t im-
pact on islanded and grid-connected for large scale industries.
CRediT authorship contribution statement
Tanveer Ahmad: Conceptualization, Investigation, Methodol-
ogy, Formal analysis, Software, Validation, Data cura-
tion, Writing –review & editing, Visualization. Dongdong
Zhang: Methodology, Resources, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing ^nancial
interests or personal relationships that could have appeared to in_u-
ence the work reported in this paper.
Acknowledgment
This work was supported in part by the National Key Research and
Development Program of China (Grant No. 2019YFE0118000), the Key
Laboratory of Special Machine and High Voltage Apparatus (Shenyang
University of Technology), Ministry of Education, Grant KFKT202006,
and Guangxi Young and Middle-aged Scienti^c Research Basic Ability
Promotion Project, 2020KY01009.
Appendix.
The temperature of PV cells is measured by the following equations:
(1A)

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 23
is the solar transmittance (%), is the solar absorptance (%), is the striking of solar radiation of PV-1 and PV-2 modules ), is the con-
version ef^ciency of electrical for PV modules arrays (%), explicates the transfer of heat coef^cient to the surroundings ), PV-1 and
PV-2 module temperature at ( ), and explicates the total ambient temperature ( ).
The above equation describes the electrical output and the solar energy absorbed by the PV-1 and PV-2 arrays. For the cell temperature, the equa-
tion is further extended:
(2A)
To measure the value of directly is a challenging task and could substitute these into Equ. (2) above to estimate the :
(3A)
, and are noted the nominal cell operation temperature (°C), ambient temperature (20 °C–25 °C) and solar radiation
, respectively.
If the is assumes a constant; we could replace Equ. (3) with the module temperature equation or cell temperature and found:
(4A)
The Homer software always measures the cell ef^ciency is equal to the maximum power point tracking ef^ciency.
(5A)
explicit the PowerPoint maximum tracking ef^ciency of PV-1 and PV-2 modules. In this regard, is replaced with we get Equ. (4) in a new
form:
(6A)
The parameter depends on PV cell temperature . If we assume, the PV-1 and PV-2 cell ef^ciency vary linearly; we used the following equa-
tion:
(7A)
is the maximum number of PowerPoint tracking ef^ciency (%), explicates the coef^cient of power temperature (%/°C) and repre-
sents the cell temperature (20 °C–25 °C). Finally, we could substitute Equ. (7) with the previous cell temperature equation and solve it to yield the
cell temperature:
(8A)
The following equation measures the temperature coef^cient:
(9A)
Where is the open circuit temperature coef^cient (V/°C) and is described as the voltages (V) at maximum power point tracking.
The following equation measures the diurnal pattern strength of WT-1 and WT-2:
(10A)
Where represents the average wind speed in different hours of day (m/s), is the overall average speed of the wind (m/s), describes the diur-
nal strength pattern, and it varies between 1 and 0, and illustrates the peak wind speed hours, and it varies between 0 and 23 h.
The equation of :
(11A)
Where explicates the fuel density (kg/m3).

UNCORRECTED PROOF
24 T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25
Appendix-II
Demand charges are measured using the following equation:
(12A)
represents the utility grid hourly load demand in month while the rates applied (kWh) and explicates the rate demand
($/kW/month).
The following equations measure the salvage and RC:
(13A)
(14A)
(15A)
Where is the components remaining life till the end of the project, is the duration of RC, represents the RC ($), is the lifetime of the
project (25-year), is the lifetime of components (year), and is an integer function that returns the integer amount.
The O&M cost is measured using the following equation:
(16A)
represents the capacity storage penalty ($/yr), is the penalty for emissions (($/year), and represents the ^xed O&M cost
($/year). The green gas house emissions are measured from the following equation:
(17A)
is the emission's penalty ($/t), is penalty of emissions ($/t), is unburned hydrocarbons ($/t), particulate matter penalty of
emissions ($/t), emissions penalty for of ($/t), yearly emissions of (kg/year), yearly emissions of (kg/year), yearly emis-
sions of the total unburned amount of hydrocarbons (kg/year), yearly particulate matter emissions of (kg/year), yearly emissions
of (kg/year), and yearly emissions of (kg/year).
References
[1] S.M. Ashabani, Y.A.R.I. Mohamed, A _exible control strategy for grid-connected and
islanded microgrids with enhanced stability using nonlinear microgrid stabilizer,,
IEEE Trans. Smart Grid 3 (2012) 1291–1301, https://doi.org/10.1109/TSG.2012.
2202131.
[2] V. Indragandhi, V. Subramaniyaswamy, R. Logesh, Resources, configurations, and
soft computing techniques for power management and control of PV/wind hybrid
system, Renew. Sustain. Energy Rev. 69 (2017) 129–143, https://doi.org/10.1016/
j.rser.2016.11.209.
[3] I. Renewable, E. Agency, Global Energy, Aroadmap To 2050 By Irena, n.d.
[4] K.R. Genwa, C.P. Sagar, Energy ef^ciency, solar energy conversion and storage in
photogalvanic cell, Energy Convers, Manag 66 (2013) 121–126, https://doi.org/10.
1016/j.enconman.2012.10.007.
[5] S. Kosai, Dynamic vulnerability in standalone hybrid renewable energy system,
Energy Convers, Manag 180 (2019) 258–268, https://doi.org/10.1016/j.enconman.
2018.10.087.
[6] O. Palizban, K. Kauhaniemi, Hierarchical control structure in microgrids with
distributed generation: island and grid-connected mode, Renew. Sustain. Energy
Rev. 44 (2015) 797–813, https://doi.org/10.1016/j.rser.2015.01.008.
[7] D. Mazzeo, N. Matera, P. De Luca, C. Baglivo, P. Maria Congedo, G. Oliveti,
Worldwide geographical mapping and optimization of stand-alone and grid-
connected hybrid renewable system techno-economic performance across Köppen-
Geiger climates, Appl. Energy, 276 115507 (2020), https://doi.org/10.1016/j.
apenergy.2020.115507.
[8] H. Mehrjerdi, M. Bornapour, R. Hemmati, S.M.S. Ghiasi, Unified energy
management and load control in building equipped with wind-solar-battery
incorporating electric and hydrogen vehicles under both connected to the grid and
islanding modes, Energy 168 (2019) 919–930, https://doi.org/10.1016/j.energy.
2018.11.131.
[9] M. Ganjian-Aboukheili, M. Shahabi, Q. Shafiee, J.M. Guerrero, Seamless transition
of microgrids operation from grid-connected to islanded mode, IEEE Trans. Smart
Grid 11 (2019) 2106–2114, https://doi.org/10.1109/tsg.2019.2947651.
[10] P. Borazjani, N.I.A. Wahab, H.B. Hizam, A.B.C. Soh, A review on microgrid control
techniques, 2014 IEEE Innov. Smart Grid Technol. - Asia, ISGT ASIA 2014 (2014)
749–753, https://doi.org/10.1109/ISGT-Asia.2014.6873886.
[11] M.H. Andishgar, E. Gholipour, R. allah Hooshmand, An overview of control
approaches of inverter-based microgrids in islanding mode of operation, Renew.
Sustain. Energy Rev. 80 (2017) 1043–1060, https://doi.org/10.1016/j.rser.2017.
05.267.
[12] M.S. Mahmoud, S. Azher Hussain, M.A. Abido, Modeling and control of microgrid:
an overview, J. Franklin Inst. 351 (2014) 2822–2859, https://doi.org/10.1016/j.
jfranklin.2014.01.016.
[13] M.H. Andishgar, A. Fereidunian, H. Lesani, Healer reinforcement for smart grid
using discrete event models of FLISR in distribution automation, J. Intell. Fuzzy
Syst. 30 (2016) 2939–2951, https://doi.org/10.3233/IFS-151814.
[14] I.T. Force, C.D.E. Olivares, S. Member, A. Mehrizi-sani, A.H. Etemadi, S. Member, C.
A. Cañizares, R. Iravani, M. Kazerani, S. Member, A.H. Hajimiragha, O. Gomis-
bellmunt, M. Saeedifard, S. Member, R. Palma-behnke, S. Member, G.A. Jiménez-
estévez, S. Member, N.D. Hatziargyriou, Trends in microgrid control, IEEE Trans.
Smart Grid 5 (2014) 1905–1919.
[15] T.L. Vandoorn, J.D.M. De Kooning, B. Meersman, L. Vandevelde, Review of primary
control strategies for islanded microgrids with power-electronic interfaces, Renew.
Sustain. Energy Rev. 19 (2013) 613–628, https://doi.org/10.1016/j.rser.2012.11.
062.
[16] A.M. Bouzid, J.M. Guerrero, A. Cheriti, M. Bouhamida, P. Sicard, M. Benghanem, A
survey on control of electric power distributed generation systems for microgrid
applications, Renew. Sustain, Energy Rev. 44 (2015) 751–766, https://doi.org/10.
1016/j.rser.2015.01.016.
[17] H. Han, X. Hou, J. Yang, J. Wu, M. Su, J.M. Guerrero, Review of power sharing
control strategies for islanding operation of AC microgrids, IEEE Trans. Smart Grid 7
(2016) 200–215, https://doi.org/10.1109/TSG.2015.2434849.
[18] E. Serban, H. Serban, A control strategy for a distributed power generation
microgrid application with voltage- and current-controlled source converter,, IEEE
Trans. Power Electron. 25 (2010) 2981–2992, https://doi.org/10.1109/TPEL.2010.
2050006.
[19] R.M. Kamel, A. Chaouachi, K. Nagasaka, Enhancement of micro-grid performance
during islanding mode using storage batteries and new fuzzy logic pitch angle
controller, Energy Convers, Manag 52 (2011) 2204–2216, https://doi.org/10.1016/
j.enconman.2010.12.025.
[20] S.J. Ahn, J.W. Park, I.Y. Chung, S. Il Moon, S.H. Kang, S.R. Nam, Power-sharing
method of multiple distributed generators considering control modes and
configurations of a microgrid, IEEE Trans. Power Deliv. 25 (2010) 2007–2016,
https://doi.org/10.1109/TPWRD.2010.2047736.
[21] M. Ahmadipour, H. Hizam, M. Lutfi Othman, M.A.M. Radzi, N. Chireh, A novel
islanding detection technique using modi^ed Slantlet transform in multi-distributed
generation, Int, J. Electr. Power Energy Syst 112 (2019) 460–475, https://doi.org/
10.1016/j.ijepes.2019.05.008.

UNCORRECTED PROOF
T. Ahmad and D. Zhang / Renewable Energy xxx (xxxx) 1–25 25
[22] M.K. Shahzad, A. Zahid, T. Rashid, M.A. Rehan, M. Ali, M. Ahmad, Techno-
economic feasibility analysis of a solar-biomass off grid system for the electrification
of remote rural areas in Pakistan using HOMER software, Renew. Energy 106 (2017)
264–273, https://doi.org/10.1016/j.renene.2017.01.033.
[23] R. Sen, S.C. Bhattacharyya, Off-grid electricity generation with renewable energy
technologies inIndia: an application of HOMER, Renew. Energy 62 (2014) 388–398,
https://doi.org/10.1016/j.renene.2013.07.028.
[24] P. Enevoldsen, B.K. Sovacool, Integrating power systems for remote island energy
supply: lessons from Mykines, Faroe Islands, Renew. Energy 85 (2016) 642–648,
https://doi.org/10.1016/j.renene.2015.06.065.
[25] H. Zahboune, S. Zouggar, G. Krajacic, P.S. Varbanov, M. Elhafyani, E. Ziani, Optimal
hybrid renewable energy design in autonomous system using modi^ed electric
system cascade analysis and homer software, energy convers, Manag 126 (2016)
909–922, https://doi.org/10.1016/j.enconman.2016.08.061.
[26] M.R. Elkadeem, S. Wang, S.W. Sharshir, E.G. Atia, Feasibility analysis and techno-
economic design of grid-isolated hybrid renewable energy system for electrification
of agriculture and irrigation area: a case study in Dongola, Sudan, Energy Convers,
Manag 196 (2019) 1453–1478, https://doi.org/10.1016/j.enconman.2019.06.085.
[27] F.K. Abo-Elyousr, A. Elnozahy, Bi-objective economic feasibility of hybrid micro-
grid systems with multiple fuel options for islanded areas in Egypt, Renew, Energy
128 (2018) 37–56, https://doi.org/10.1016/j.renene.2018.05.066.
[28] S.M. Dawoud, X.N. Lin, J.W. Sun, M.I. Okba, M.S. Khalid, A. Waqar, Feasibility
study of isolated PV-wind hybrid system in Egypt, Adv. Mater. Res. 1092–1093
(2015) 145–151, https://doi.org/10.4028/www.scientific.net/amr.1092-1093.145.
[29] Y. Sawle, S.C. Gupta, A.K. Bohre, Socio-techno-economic design of hybrid
renewable energy system using optimization techniques, Renew. Energy 119 (2018)
459–472, https://doi.org/10.1016/j.renene.2017.11.058.
[30] T. Ma, M.S. Javed, Integrated sizing of hybrid PV-wind-battery system for remote
island considering the saturation of each renewable energy resource, Energy
Convers, Manag 182 (2019) 178–190, https://doi.org/10.1016/j.enconman.2018.
12.059.
[31] A.M. Abdilahi, A.H. Mohd Yatim, M.W. Mustafa, O.T. Khalaf, A.F. Shumran, F.
Mohamed Nor, Feasibility study of renewable energy-based microgrid system in
Somaliland’s urban centers, Renew. Sustain. Energy Rev. 40 (2014) 1048–1059,
https://doi.org/10.1016/j.rser.2014.07.150.
[32] A.C. Duman, Ö. Güler, Techno-economic analysis of off-grid PV/wind/fuel cell
hybrid system combinations with a comparison of regularly and seasonally occupied
households, Sustain. Cities Soc. 42 (2018) 107–126, https://doi.org/10.1016/j.scs.
2018.06.029.
[33] A. Al-Sharafi, A.Z. Sahin, T. Ayar, B.S. Yilbas, Techno-economic analysis and
optimization of solar and wind energy systems for power generation and hydrogen
production in Saudi Arabia, Renew. Sustain. Energy Rev. 69 (2017) 33–49, https://
doi.org/10.1016/j.rser.2016.11.157.
[34] A. Haghighat Mamaghani, S.A. Avella Escandon, B. Naja^, A. Shirazi, F. Rinaldi,
Techno-economic feasibility of photovoltaic, wind, diesel and hybrid electrification
systems for off-grid rural electrification in Colombia, Renew, Energy 97 (2016)
293–305, https://doi.org/10.1016/j.renene.2016.05.086.
[35] S. Mandal, B.K. Das, N. Hoque, Optimum sizing of a stand-alone hybrid energy
system for rural electrification in Bangladesh, J. Clean. Prod. 200 (2018) 12–27,
https://doi.org/10.1016/j.jclepro.2018.07.257.
[36] M.J. Khan, A.K. Yadav, L. Mathew, Techno economic feasibility analysis of di`erent
combinations of PV-Wind-Diesel-Battery hybrid system for telecommunication
applications in di`erent cities of Punjab, India, Renew. Sustain. Energy Rev. 76
(2017) 577–607, https://doi.org/10.1016/j.rser.2017.03.076.
[37] K. Elmaadawy, K.M. Kotb, M.R. Elkadeem, S.W. Sharshir, A. Dán, A. Moawad, B.
Liu, Optimal sizing and techno-enviro-economic feasibility assessment of large-scale
reverse osmosis desalination powered with hybrid renewable energy sources,
Energy Convers. Manag, 224 113377 (2020), https://doi.org/10.1016/j.enconman.
2020.113377.
[38] A.B. Kanase-Patil, R.P. Saini, M.P. Sharma, Integrated renewable energy systems for
off grid rural electrification of remote area, Renew. Energy 35 (2010) 1342–1349,
https://doi.org/10.1016/j.renene.2009.10.005.
[39] I. Padrón, D. Avila, G.N. Marichal, J.A. Rodríguez, Assessment of hybrid renewable
energy systems to supplied energy to autonomous desalination systems in two
islands of the canary archipelago, Renew. Sustain. Energy Rev. 101 (2019)
221–230, https://doi.org/10.1016/j.rser.2018.11.009.
[40] M. Bagheri, S.H. Delbari, M. Pakzadmanesh, C.A. Kennedy, City-integrated
renewable energy design for low-carbon and climate-resilient communities, Appl.
Energy 239 (2019) 1212–1225, https://doi.org/10.1016/j.apenergy.2019.02.031.
[41] Techno-economic analysis and size optimization of an off-grid hybrid photovoltaic,
Fuel Cell and Diesel Generator System, (n.d.).
[42] K.H. Chua, Y.S. Lim, S. Morris, Cost-bene^t assessment of energy storage for utility
and customers: a case study in Malaysia, Energy Convers, Manag 106 (2015)
1071–1081, https://doi.org/10.1016/j.enconman.2015.10.041.
[43] M.S. Adaramola, M. Agelin-Chaab, S.S. Paul, Analysis of hybrid energy systems for
application in southern Ghana, Energy Convers, Manag 88 (2014) 284–295, https://
doi.org/10.1016/j.enconman.2014.08.029.
[44] Case study: Simulation and Optimization of Photovoltaic-Wind-Battery Hybrid
Energy System in Taleghan-Iran Using Homer Software, (n.d.).
[45] NREL The HOMER Pro® microgrid software by HOMER Energy https://www.
homerenergy.com/products/pro/index.html 2019
[46] S.M. Shaahid, I. El-Amin, Techno-economic evaluation of off-grid hybrid
photovoltaic-diesel-battery power systems for rural electrification in Saudi Arabia-A
way forward for sustainable development, Renew. Sustain. Energy Rev. 13 (2009)
625–633, https://doi.org/10.1016/j.rser.2007.11.017.
[47] S. Baek, E. Park, M.G. Kim, S.J. Kwon, K.J. Kim, J.Y. Ohm, A.P. del Pobil, Optimal
renewable power generation systems for Busan metropolitan city in South Korea,
Renew. Energy 88 (2016) 517–525, https://doi.org/10.1016/j.renene.2015.11.058.